Fishery Bulletin f?*ari*« H t Sr ATES O* + 1983 r Wooos Hote, Jv»ass. Vol. 81, No. 1 January 1983 SMITH, TIM D. Changes in size of three dolphin (Stenella spp.) populations in the eastern tropical Pacific 1 LANGTON, RICHARD W. Food habits of yellowtail flounder, Limanda ferru- gi>iia (Storer), from off the northeastern United States 15 DUNN, JEAN R. Development and distribution of the young of northern smooth- tongue, Leuroglossus sckmidti (Bathylagidae), in the northeast Pacific, with com- ments on the systematics of the genus Leuroglossus Gilbert 23 KATZ, S. J., C. B. GRIMES, and K. W. ABLE. Delineation of tilefish, Lopholatilus chamaeleonticeps, stocks along the United States east coast and in the Gulf of Mexico 41 RICHARDS. R. ANNE, J. STANLEY COBB, and MICHAEL J. FOGARTY. Effects of behavioral interactions on the catchability of American lobster, Ho- marus americanus, and two species of Cancer crabs 51 PARSONS, GLENN R. The reproductive biology of the Atlantic sharpnose shark, Rhizoprionodon terraenovae (Richardson) 61 APPELDOORN, RICHARD S. Variation in the growth rate of Myaarenaria and its relationship to the environment as analyzed through principal components analysis and the a> parameter of the von Bertalanffy equation 75 SHAKLEE, JAMES B., RICHARD W. BRILL, and ROBIN ACERRA. Bio- chemical genetics of Pacific blue marlin, Makaira nigricans, from Hawaiian waters 85 BARTOO, NORMAN W., and KEITH R. PARKER. Stochastic age-frequency estimation using the von Bertalanffy growth equation 91 JOHNSON, ALLYN G., WILLIAM A. FABLE, JR., MARK L. WILLIAMS, and LYMAN E. BARGER. Age, growth, and mortality of king mackerel, Scom- beromorus ca valla, from the southeastern United States 97 HAN AN, DOYLE A. Review and analysis of the bluefin tuna, Thunnus thynnus, fishery in the eastern North Pacific Ocean 107 SWARTZMAN, GORDON L., and ROBERT T. HAAR. Interactions between fur seal populations and fisheries in the Bering Sea 121 DURBIN, ANN GAIL, EDWARD G. DURBIN, THOMAS J. SMAYDA, and PETER G. VERITY. Age, size, growth, and chemical composition of Atlantic menhaden, Brevoortia tyrannus, from Narragansett Bay, Rhode Island 133 (Continued on back cover) V. Seattle, Washington U.S. DEPARTMENT OF COMMERCE Malcolm Baldrige, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION John V. Byrne, Administrator NATIONAL MARINE FISHERIES SERVICE William G. Gordon, Assistant Administrator Fishery Bulletin The Fishery Bulletin carries original research reports and technical notes on investigations in fishery science, engineering, and economics. The Bulletin of the United States Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries in 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates were issued as documents through volume 46; the last document was No. 1 103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate appeared as a numbered bulletin. A new system began in 1963 with volume 63 in which papers are bound together in a single issue of the bulle- tin instead of being issued individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a periodi- cal, issued quarterly. In this form, it is available by subscription from the Superintendentof 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. EDITOR Dr. Carl J. Sindermann Scientific Editor, Fishery Bulletin Northeast Fisheries Center Sandy Hook Laboratory National Marine Fisheries Service, NOAA Highlands, NJ 07732 Editorial Committee Dr. Bruce B. Collette Dr. Donald C. Malins National Marine Fisheries Service National Marine Fisheries Service Dr. Edward D. Houde Dr. Jerome J. Pella Chesapeake Biological Laboratory National Marine Fisheries Service Dr. Merton C. Ingham Dr. Jay C. Quast National Marine Fisheries Service National Marine Fisheries Service Dr. Reuben Lasker Dr. Sally L. Richardson National Marine Fisheries Service Gulf Coast Research Laboratory Mary S. Fukuyama, Managing Editor The Fishery Bulletin (ISSN 0090-0656) is published quarterly by Scientific Publications Office, National Marine Fisheries Service, NOAA, 7600 Sand Point Way NE, Bin C15700, Seattle, WA 98115 Second class postage paid to Finance Department, USPS, Washington, DC 20260. 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 i| 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 through 1 April 1985. , Fishery Bulletin CONTENTS Vol. 81, No. 1 January 1983 SMITH. TIM D. Changes in size of three dolphin (Stenella spp.) populations in the eastern tropical Pacific 1 LANGTON, RICHARD W. Food habits of yellowtail flounder, Limanda ferru- ginia (Storer), from off the northeastern United States 15 DUNN, JEAN R. Development and distribution of the young of northern smooth- tongue, Leuroglossus schmidt i (Bathylagidae), in the northeast Pacific, with com- ments on the systematics of the genus Leuroglossus Gilbert 23 KATZ, S. J., C. B. GRIMES, and K. W. ABLE. Delineation of tilefish, Lopholatilus chamaeleonticeps, stocks along the United States east coast and in the Gulf of Mexico 41 RICHARDS, R. ANNE, J. STANLEY COBB, and MICHAEL J. FOGARTY. Effects of behavioral interactions on the catchability of American lobster, Ho- marus americanus, and two species of Cancer crabs 51 PARSONS, GLENN R. The reproductive biology of the Atlantic sharpnose shark, Rhizoprionodon terraenovae (Richardson) 61 APPELDOORN, RICHARD S. Variation in the growth rate of Mya arenaria and its relationship to the environment as analyzed through principal components analysis and the w parameter of the von Bertalanffy equation 75 SHAKLEE, JAMES B., RICHARD W. BRILL, and ROBIN ACERRA. Bio- chemical genetics of Pacific blue marlin, Makaira nigricans, from Hawaiian waters 85 BARTOO, NORMAN W., and KEITH R. PARKER. Stochastic age-frequency estimation using the von Bertalanffy growth equation 91 JOHNSON, ALLYN G., WILLIAM A. FABLE, JR., MARK L. WILLIAMS, and LYMAN E. BARGER. Age, growth, and mortality of king mackerel, Scom- beromorus cavalla, from the southeastern United States 97 HAN AN, DOYLE A. Review and analysis of the bluefin tuna, Thunnus thynnus, fishery in the eastern North Pacific Ocean 107 SWARTZMAN, GORDON L., and ROBERT T. HAAR. Interactions between fur seal populations and fisheries in the Bering Sea 121 DURBIN, ANN GAIL, EDWARD G. DURBIN, THOMAS J. SMAYDA, and PETER G. VERITY. Age, size, growth, and chemical composition of Atlantic menhaden, Brevoortia tyrannus, from Narragansett Bay, Rhode Island 133 (Continued on next page) Seattle, Washington 1983 For sale by the Superintendent of Documents, U.S. Government Printing Office. Washington. DC 20402— Subscription price per year: #21.00 domestic and $20.25 foreign. Cost per single issue: $6.50 domestic and $8.15 foreign. ( 'ontents— continued Notes VREELAND, ROBERT R., and ROY J. WAHLE. Homing and fisheries contri- bution of marked coho salmon, Oncorhynchus kisutch, released at two Columbia River locations 143 COLTON, DOUGLAS E., and WILLIAM S. ALE VIZON. Movement patterns of bonefish, Albula vulpes, in Bahamian waters 148 KLEPPEL, G. S., and E. MANZANILLA. Analyses of feeding in two marine copepods from Santa Monica Bay, California 154 SHENKER, JONATHAN M. Distribution, size relationships, and food habits of juvenile king-of-the-salmon, Trachipterus altivelis, caught off the Oregon coast 161 BOND, CARL E., TING T. KAN, and KATHERINE W. MYERS. Notes on the marine life of the river lamprey, Lampetra ayresi, in Yaquina Bay, Oregon, and the Columbia River estuary 165 MENZ, FREDRIC C, and DONALD P. WILTON. An economic evaluation of the St. Lawrence River-eastern Lake Ontario bass fishery 168 The National Marine Fisheries Service (NMFS) does not approve, recommend or endorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to NMFS, or to this publication fur- nished by NMFS, in any advertising or sales promotion which would indicate or imply that NMFS approves, recommends or endorses any proprietary prod- uct or proprietary material mentioned herein, or which has as its purpose an intent to cause directly or indirectly the advertised product to be used or pur- chased because of this NMFS publication. ^ m- AWARD AWARD Best NMFS publications for 1981 The Publications Advisory Committee of the National Marine Fisheries Service has announced the best publications authored by the NMFS scientists and published in the Fishery Bulletin and the Marine Fisheries Review for 1981. Only effective and interpretive articles which significantly contribute to the understanding and knowledge of NMFS mission-related studies are eligible, and the following papers have met this requirement. For the Fishery Bulletin, the paper "The spawning energetics of female northern anchovy, Engraulis mordax" by J. Roe Hunter and Roderick Leong was awarded as the best publication, appearing in the Fishery Bulletin 79(2): 215-230. Hunter and Leong are both fishery biologists with the Southwest Fisheries Center La Jolla Laboratory, NMFS, NOAA, La Jolla, Calif. For the Marine Fisheries Review, the paper "Low temperature preserva- tion of seafoods: A review" by Louis J. Ronsivalli and Daniel W. Baker II was awarded the best publication, appearing in the Marine Fisheries Review 43(4):1-15. Ronsivalli, now retired, was the former Director of the Northeast Fisheries Center Gloucester Laboratory, NMFS, NOAA, Gloucester, Mass.; Baker is a mechanical engineering technician with the same laboratory. N fr CHANGES IN SIZE OF THREE DOLPHIN (STENELLA SPP.) POPULATIONS IN THE EASTERN TROPICAL PACIFIC Tim D. Smith 1 ABSTRACT Dolphins from three populations, one of Stenella attenuata and two of S. longirostris, have been killed incidentally in the yellowfin tuna purse seine fishery in the eastern tropical Pacific, two populations since about 1959 and the other since about 1969. Size changes in these populations are estimated from numbers killed each year, population size estimates in 1979, and net recruitment rates. Ranges of values for some parameters are considered, accounting for some uncertainties. Assuming central values of the ranges of maximum net recruitment rate (3%) and the population level giving maximum net productivity (65%), one S. longirostris population, the eastern spinner dolphin, is near 20% of pre-exploitation levels; the S. attenuata population, the northern offshore spotted dolphin, is between 35 and 50%; and the second S. longirostris population, the whitebelly spinner dolphin, is between 58 and 72% of pre-exploitation levels. Purse seine fishing for tuna in the eastern tropi- cal Pacific often involves dolphins found in asso- ciation with yellowfin tuna. Tuna fishermen pursue and capture the dolphin-yellowfin tuna complex, releasing the dolphins from the net while retaining the tuna (Green et al. 1971). Mor- tality of dolphins occurs incidental to this fishing process. Purse seine fishermen were using dolphin schools to catch tuna by 1959; there is anecdotal information suggesting limited use as early as the 1940's (anonymous reviewer). Starting in the mid-1960's the Bureau of Commercial Fisheries, predecessor of the National Marine Fisheries Service (NMFS), conducted limited research to document the situation and to collect data on numbers and kinds of dolphins killed. This re- search expanded in the 1970's, especially after passage of the Marine Mammal Protection Act (MMPA) of the United States in 1972, and con- tinues. Substantial research efforts were mount- ed to assess the status of the dolphin stocks and to develop procedures for reducing incidental mor- tality and injury. Two assessments of the condition of dolphin populations involved in the yellowfin tuna purse seine fishery have been completed in recent years. 2,3 1 describe the results of the latest assess- ment of the three populations most affected by the fishery; calculation of population sizes from 1959 through 1978 is emphasized, based on esti- mates of the population size in 1979, on annual numbers killed from 1959 through 1978, and on net recruitment rates. These results, based on data available through the end of 1980, do not necessarily represent NMFS policy, which in- volves additional considerations. A third assess- ment of these populations is scheduled for 1984 and will include information since 1980. POPULATION MODEL Methods developed in 1976 (footnote 2) for esti- mating pre-exploitation abundance are based on a simple recursive relationship Nm = N t - K t + R t (N t -%K t ), where t denotes the year; N, the abundance; K, the number of animals killed; and R, the net re- cruitment rate. This model assumes that the population size in the next year is simply the present population size, minus the present inci- dental kill, plus the net number of individuals re- cruited to the population during the year. This latter quantity is taken to be the net recruitment 'Southwest Fisheries Center, National Marine Fisheries Service, NOAA, P.O. Box 271, LaJolla, Calif.; present address: Northeast Fisheries Center, National Marine Fisheries Ser- vice, NOAA, Woods Hole, MA 02543. 2 Southwest Fisheries Center La Jolla Laboratory. 1976. Report of the workshop on stock assessment of porpoises in- volved in the eastern Pacific yellowfin tuna fishery. Natl. Mar. Fish. Serv., NOAA, Admin. Rep. LJ-76-29, 53 p. 3 Smith, T. D. (editor). 1979. Report of the workshop on status of porpoise stocks, La Jolla, Calif., 27-31 August 1979. Southwest Fish. Cent. La Jolla Lab., Natl. Mar. Fish. Serv., NOAA, Admin. Rep. LJ-70-41, 120 p. Manuscript accepted June 1982. FISHERY BULLETIN: VOL. 81, NO. 1. 1983. FISHERY BULLETIN: VOL. 81. NO. 1 rate (birth rate less natural death rate) multi- plied by the number of animals actually repro- ducing in a given year. The number of repro- ducing animals is approximated by assuming that one-half the animals killed in a year repro- duce before dying. Solving this relationship for N t , one obtains (i) Repeatedly applying this equation to estimate the population size for any number of years (s) prior to the year (c) for which an independent estimate of population size {N ) is available yields in general N s = N c -z 1 - . (2) n(i + r\ j=1 n/i + r) The 1979 workshop (footnote 3) extended this procedure by calculating the recruitment rate Ri, i years prior to the present, using the density- dependent relationship (Allen 1981) Ri — -R TO <1 (3) N p is the estimated population size at the begin- ning of the first year of exploitation, p years earlier; R,„, the maximum net recruitment rate; Z, the density-dependent exponent; and N,- and N p , estimated from Equation (2). Because N p in Equation (3) is not known until the series in Equation (2) has been calculated, an iterative procedure is required to solve the equations for historical population size. Equations (1) and (3) together form a special case of the generalized production model of Pella and Tomlinson (1969). In Equation (3) the net recruitment rate is maximum (R m ) when the population size ap- proaches zero, decreasing to zero as the popula- tion size approaches N p . Z determines the popu- lation size at which the rate of change of the population is maximum, the maximum net pro- ductivity level (MNPL). The values of Z corre- spond to the MNPL approximately as (Polacheck 1982) MNPL N P (1 + Zf z (4) If Z=l, then the MNPL is one-half the equilibri- um population size; if Z is >1, then MNPL is greater than one-half the equilibrium size. The fraction of the maximum reproductive rate, R m , realized at a given population size, increases as the value of Z increases. Statistical properties of the estimate of N p and the ratio N c /Np are examined in detail in Smith and Polacheck (1979), wherein methods are de- veloped for calculating the variances. Tests of sensitivity of the estimates of N p to the values JV C , K t , and R m show that the estimates are most sen- sitive to the value of present abundance and least sensitive to the net recruitment rate. Examina- tion via simulation of the shape of the sampling distribution shows that if N has a symmetrical sampling distribution, then so does the estimate Np. Several estimates of each parameter required by the model are available in working documents and technical memoranda prepared by the staff of the Southwest Fisheries Center. I rely on the most current estimates, primarily minor revi- sions of those used by the 1979 workshop, with reference to papers describing earlier estimates as needed to document methods. POPULATIONS Populations affected most by the yellowfin tuna purse seine fishery are of the genus Stenella, and are found in the area from just south of the Equator to an approximate lat. 20°N and west from the Mexican and Central American coasts to an approximate long. 150°W. Two populations of spotted dolphins, S. attenuata, and three popu- lations of spinner dolphins, S. longirostris, are found in this region. The two spotted dolphin populations are re- ferred to as "offshore" and "coastal" forms. The coastal spotted dolphin population occurs near- shore and around islands, while the offshore spotted dolphin ranges from nearshore to an ap- proximate long. 150°W. The two forms overlap in range near the coast. Perrin (1975) distinguished these two forms of S. attenuata morphologically. He noted that 1) the larger coastal form occurs seaward to 50 km while the offshore form occurs as nearshore as 20 km, and 2) the coastal form was involved in only 7 of 1,373 purse seine sets on dolphins observed be- tween 1971 and 1974. Additional data collected since then, including reexamination of speci- mens collected during sets in the years 1971-74, SMITH: SIZE CHANGES OF THREE DOLPHIN POPULATIONS indicate that through 1978, a total of 22 sets have been observed on coastal spotted dolphins, out of a total 9,672 observed overall (about 0.2%). The yellowfin tuna purse seine fishery was concentrated nearshore in the early 1960's, and many sets were made in the area probably occu- pied by both coastal and offshore spotted dol- phins. Direct observations in the 1960's distin- guishing these forms of the spotted dolphin are not available. Based on observations made in the 1970's where these forms were distinguished, however, it appears that the coastal form has never been significantly involved in this fishery. Since 1978, sighting data collected by scientific observers aboard tuna vessels have been edited, using consistent criteria of school size, body size, and coloration for distinguishing coastal and off- shore spotted dolphins. In 1979, for example, within about 50 km of the coast there were 46 sightings of coastal spotted schools, 160 sightings of offshore spotted schools, and 25 sightings of spotted dolphin schools which could not be dis- tinguished to form with the available data. These three school types were subsequently set on in 2, 73, and 6 instances, respectively. Even assuming that all schools not identified to form were coastal spotted dolphins, the proportion of sighted coast- al spotted dolphin schools subsequently set on is much smaller than that of the offshore form (0.11 vs. 0.46, P<0.001). This differential selection exists even though the catch of yellowfin tuna in sets on coastal spotted dolphins has been approxi- mately twice that on offshore spotted dolphins. If coastal spotted dolphins were a significant part of this fishery, one would expect their involve- ment in sets to be proportional to the rate at which they are encountered. In addition, 18 of the 22 sets on coastal spotted dolphins occurred in 1973, and, except for one set, these were made by two vessels in the Gulf of Nicoya, a small area off the Costa Rican coast. Based on this information, I have assumed that the coastal spotted dolphin has been involved only rarely in this fishery. Two spinner dolphin populations, referred to as the "eastern" and "whitebelly" forms, are in- volved in the yellowfin tuna purse seine fishery. A third form, termed the Costa Rican spinner, occurs near the coast from Mexico to Panama, but is not involved in the fishery. The eastern and whitebelly forms overlap broadly in range, with the whitebelly spinner dolphin generally occur- ring more seaward. The eastern form has been involved with this fishery since 1959, whereas the whitebelly spinner dolphin population ap- parently became increasingly involved as the fishery expanded seaward in the 1960's. The whitebelly spinner and the offshore spotted forms have Southern Hemisphere populations (Perrin et al. 1979). These populations have been involved only recently with the yellowfin tuna purse seine fishery, as it has expanded south- ward, and are only lightly exploited. Data on reproductive condition of these southern popula- tions are used as estimates of reproductive rates for unexploited or equilibrium populations. 1979 POPULATION SIZE ESTIMATES Holt and Powers (1982) gave estimates of abun- dance based on aerial and research-vessel sight- ing surveys and data from scientific observers aboard fishing vessels. Estimates of the size, N,, of the i th population in their survey area are based on the equation N, = P t S t D P, A, (5) where P, denotes the proportion of dolphin schools containing dolphin of the genera Stenella, Delphinus, and Lagenodelphis; S t , the mean size of these schools; D, the estimated density of all dolphin schools sighted; P,, the fraction of schools containing dolphins of the ith population; and A, the area inhabited. This equation is applied to 1) a nearshore stratum, surveyed using both an air- plane and research vessels, and 2) an offshore stratum, surveyed only by research vessels. The nearshore stratum extends seaward from the coastline about 800 km, and from lat. 22°N to 12°S. The offshore stratum extends from the outer edge of the nearshore stratum to the bound- ary of the dolphin range. Approximate areas of the maximum historical range of the three dolphin populations are used for the area inhabited, A in Equation (5). These are estimates of the area enclosed by a smooth curve which includes most locations where dol- phins of different species have been reported by both fishing vessels and research vessels, as de- scribed in Holt and Powers (1982). While occasional sightings of dolphin schools have been reported outside these areas, the areas are overestimated in that ". . . at any point in time it is likely that each of the various dolphin species FISHERY BULLETIN: VOL. 81, NO. 1 only occupies a portion of its historical range." 4 Overlap between coastal and offshore forms of spotted dolphin is not reflected in the population estimates given by Holt and Powers (1982). Due to the large differences in areas inhabited, how- ever, adjustments to account for the unknown degree of overlap would increase the offshore spotted dolphin population estimate by 3% at most, which is insignificant for the general re- sults being presented here. The density estimate for the nearshore area is obtained from line transect theory applied to aerial survey sighting data. This follows earlier applications (Smith 1981), but with several im- provements. For instance, the aircraft we used had superior downward visibility; right-angle distance from the aircraft trackline to the sighted dolphin schools was determined directly, either by electronic navigation equipment or visually for shorter distances, rather than being calcu- lated from visual estimates of range and bearing; and the originally used negative-exponential sighting model was replaced with the superior Fourier series model (Burnham et al. 1980). The density estimate for the offshore area, which could not be surveyed by air, is obtained by com- paring relative dolphin school sighting rates from research vessel surveys in nearshore and offshore areas with absolute density estimates from the nearshore area. The resulting density estimate of all dolphin schools of >15 animals in the nearshore area is about 3.6 schools/ 1,000 km 2 , while the density estimate in the offshore area is about one-half that value. The school size estimate is about 200 animals, based on visual estimates of the size of schools seen during the aerial survey. The accuracy of these visual estimates has been confirmed by counts of individual dolphins from aerial photo- graphs, and the accuracy of the counts from these photographs has been confirmed by counts of dolphins released from a purse seine (Allen et al. 1980). This estimated school size also includes an adjustment for the tendency of larger schools to be more readily visible at greater distances from the aircraft, and hence to be overrepresent- ed in the sample. Allen et al. (1980) also demonstrated that accu- rate school size estimates could be made from ships. Although not used by Holt and Powers (1982), the mean school size estimated from re- search vessel sighting data was about 180, not significantly different from the value derived from aerial data described above. In contrast, the mean school size estimated from tuna vessel sighting data collected by scientific technicians was about 580, significantly higher (P<0.001) than the other two values. This difference im- plies either nonrandomness of the sample of dol- phin schools encountered by the tuna vessels, or biases in the estimation techniques used by the technicians. P ( for each of the 22 populations involved in the yellowfin tuna purse seine fishery can be esti- mated from data collected aboard either tuna vessels or research vessels. Fishing vessels en- counter significantly more schools composed pri- marily of spotted and spinner dolphins than do research vessels. The reason for this difference is not known, but it is possible that fishing vessels encounter spotted and spinner dolpin schools more frequently than would be expected under random search because they are searching for tuna, which occur with these two schools more frequently than with other species of dolphins. Studies of the searching process of tuna fishing vessels are being conducted which should help resolve this question. Because the proportions P, are different for unknown reasons, Holt and Powers (1982) gave several sets of estimates of total abundance, depending on the estimates of Pi from different combinations of the research vessel and tuna vessel data. Two sets of estimates are considered here (Table 1), one using research vessel data alone and the other using combined tuna vessel and research vessel data. Aerial survey procedures used in the present population-size estimates are still being refined. For instance, a field study was completed in mid- Table 1.— Population size estimates (thousands) at the beginning of 1979 for three populations of dol- phins in the eastern tropical Pacific, using estimates of the species mix from research vessel data alone, and from combined tuna vessel data and research vessel data, with standard deviations in parentheses (Holt and Powers 1982). Population Research vessel data only Fishing and research vessel data 4 Hammond, P. S. (editor). 1981. Report of the Workshop on Tuna-Dolphin Interactions, Managua, Nicaragua, April 1981, p. 5. IATTC Spec. Rep. 4, Inter-Am. Trop. Tuna Comm., c/o Scripps Inst. Oceanogr., La Jolla, CA 92093. Offshore spotted 1,682.0 (471.8) 2,775.0 (761.4) Eastern spinner 292 7 (71.0) 2929 (64.4) Whitebelly spinner 216.0 (67.4) 3804 (134.9) SMITH: SIZE CHANGES OF THREE DOLPHIN POPULATIONS 1980 to determine the effect of sea state and sun position on the visibility of dolphin schools di- rectly on the trackline. Data from this experi- ment, which have not yet been completely ana- lyzed, will be of use in the design of future surveys and in the evaluation of earlier sur- veys. INCIDENTAL KILL ESTIMATES Incidental kill (K, ) of dolphins in year t is esti- mated by multiplying the mean kill of dolphins per set in year t ( KPS,) by the total number of net sets involving dolphins made by the tuna fleet in year t (NSETS,), as K, = KPS, NSETS,. (6) These estimates are obtained for each year with the data stratified by vessel fish-carrying capac- ity, amount of tuna caught in the net set, and geographic location of the set, following the gen- eral approach described by Lo et al. (1982). Kill rate information is available from a lim- ited set of tuna fishing trips in the 1960's and from a more extensive set in the 1970's collected by scientific observers placed aboard a large pro- portion of the U.S. fishing vessels. To illustrate the data, some mean kill rates, stratified by amount of tuna caught, are shown in Table 2. Higher kill rates are apparent in successful (>% ton tuna caught) than in unsuccessful (<% ton tuna caught) sets, as are marked declines in kill rates over time. Numbers of dolphin sets and fishing trips on which observations of numbers of dolphins killed were made are shown in Table 3. Observations of the numbers of dolphins killed in the 1960's were made by both the crew and the scientists. Although few observations were made, there is no consistent difference between kill Table 2.— Observed mean kill of dolphins per net set (KPS) by U.S. tuna purse seiners 1964-78, for successful and unsuccessful net sets, with sample sizes (iV), from NMFS records. Successful sets Unsuccessful sets (>% KPS t tuna) N (< 'A t tuna) Year KPS W 1964-72 55.7 343 79 25 1973. 226 576 06 130 1974 15.7 753 24 261 1975 18.9 778 2.5 169 1976 16.1 627 57 126 1977 3.4 2,706 0.9 495 1978 4.2 1,434 3.9 249 rates reported by both types of observers (59 and 52, respectively); this suggests the presence of a noncrew-member observer had no significant effect on the kill rate of dolphins in the 1960's. All data on kill rates of dolphins for the period 1971-78 were collected by noncrew-member sci- entists, precluding a direct comparison of kill rates between fishing trips with (observed) and without (unobserved) scientific observers for this period. Groom, 5 however, reported dolphin kill rates on a fishing trip in 1979 with no scientific observer on board; his kill rates were about 4 times higher than the average rate in 1979 for scientist-observed trips and were approximately 20 times higher than on previous and succeeding observed fishing trips by the same vessel and captain. This difference in mean kill rates was due to the significantly lower proportion of sets with few dolphins killed on Groom's trip than on the scientist-observed trips. For instance, the proportion of sets with zero dolphins killed was 0.23 on Croom's trip with 0.76 for observed trips. Although limited information is available, it appears that kill rates on some unobserved ves- sels were higher in the late 1970's, and that this could result in the observed kill rates being lower than the actual rates. If there has been an "ob- server effect," it most likely occurred in the late 1970's, because regulations were adopted in the United States in 1976 requiring the use of cer- tain dolphin-release procedures, and because sci- entific observers were then used to collect regu- lation compliance information. If kill estimates for the last few years were revised with this in mind, it would only slightly affect the calcula- tions presented here, since the large number of animals killed through 1975 tends to dominate in Equation (2). However, such revisions to the kill estimates could markedly change our perception of the current rate of change of these populations. In addition to the known direct kill of dolphins in the fishery, research has been conducted to estimate both the number of dolphins injured and released alive from the purse seines, and the possible number of dolphins which, while not ex- hibiting injuries, die or suffer reduced viability from stress of capture and handling in the purse seining operation. Observations of the number of injured dolphins have been made aboard tuna vessels since 1975; estimates of the number in- 5 Croom, M. M. 1980. The tuna-porpoise problem: Man- agement aspects of a fishery. M.S. Internship Rep., Marine Resour. Manage. Program, Oreg. State Univ., Corvallis, 41 p. FISHERY BULLETIN: VOL. 81. NO. 1 jured fluctuate around 4.8% of the number killed directly, ranging from 3 to 7%. The problem of stress-induced mortality or debility was explored in a workshop of experts on large mammal physi- ology and pathology, and research plans to ap- proach this problem were developed. 6 Subse- quently one aspect of this problem was examined with dolphin specimens collected aboard tuna vessels. 7 Reproductive tracts were examined for evidence of spontaneous abortion, and muscle tissue for myopathy; no evidence of either was found. No estimates of the magnitude of such effects have been made, and currently no re- search is underway to investigate stress-induced mortality. As a conservative measure, given our limited knowledge, I assume in the estimates of total dolphin mortality given here that all of the injured dolphins subsequently die of their in- juries. Thus estimates of total kill of dolphins are the sum of the estimated numbers killed directly and the numbers injured. Numbers of net sets made by the tuna purse seine fleet have been recorded by the Inter- American Tropical Tuna Commission (IATTC) from logbooks kept by the fishermen (Table 3). In the logbooks the type of each net set may be re- corded, along with tuna catch, location, and other information. The three major types of sets are 1) those known to involve dolphins, 2) those known not to involve dolphins, and 3) those for which the data indicate neither the presence nor absence of dolphins. Types of sets not involving dolphins include "floating object sets" (e.g., a rope, board, log, etc.), a "school fish set" (i.e., a net set on tuna sighted at or near the surface), and a "porpoise set." The logbook data are incomplete, however, because some members of the fleet do not report and because, in some cases, only lim- ited information was recorded by the fishermen. The logbook coverage rate, however, is high. The data in columns D, N, and U in Table 3 have only recently become available, and analy- ses are proceeding to use this information di- rectly to estimate the total number of sets made on dolphins. Preliminary results for the total numbers of sets for each year 8 are similar to 6 Stuntz, W. E., and T. B. Shay. 1979. Report on capture stress workshop, La Jolla. California, May 1979. Southwest Fish. Cent. La Jolla Lab., Natl. Mar. Fish. Serv., NOAA, Admin. Rep. LJ-79-28, 24 p. 7 Cowan, D., and W. Walker. 1979. Disease factors in Ste- nella attenuata and Stenella longirostris taken in the eastern tropical Pacific yellowfin tuna purse seine fishery. South- west Fish. Cent. La Jolla Lab., Natl. Mar. Fish. Serv., NOAA, Admin. Rep. LJ-79-32, 21 p. 8 Inter-American Tropical Tuna Commission. 1981. Tuna- Table 3.— Number of tuna purse seine sets. 1959-78, (D) known to have been made on dolphins in the eastern tropical Pacific, (N) known not to have been made on dolphins, and ( U) unknown if made on dolphins (IATTC text footnote 8), with (is) estimates of the total number of sets made on dolphins (Smith text foot- note 3). Also shown are the numbers of observed fishing trips and purse seine sets on porpoise from NMFS records. Sets made Sets observed Trips Year D N U £ observed 1959 132 759 2,985 1.037 1960 1.644 1,256 7,390 5,696 1961 3,617 3.825 8,694 8,247 1962 2.886 8.830 4,337 4,060 1963 3,290 9.266 6,322 4.687 1964 5,933 7.681 4,745 8.090 67 2 1965 6,172 7,176 5,631 7,981 1966 5,443 7,001 5,247 7,250 28 1 1967 3,510 10,018 3,594 4,478 1968 3,833 8,988 1,642 4,271 15 1 1969 7,664 6.552 2,055 8.678 1970 7,912 9.692 1,664 8,552 1971 4.816 10,728 3,404 5,039 78 5 1972 8.193 4,682 3,514 9,036 272 12 1973 8,686 9,463 3,672 9,998 752 22 1974 7.955 11.669 4,835 8,539 1,120 36 1975 8.172 13,396 4,902 8,951 1,049 31 1976 7.481 17,789 5,184 7,910 1,295 45 1977 7,485 15,005 7,643 9.757 3,335 94 1978 5,174 21,527 5,639 5,910 1,771 102 those given in column E of Table 3, but the re- sults are not yet available in the stratified form needed to estimate numbers killed, described be- low. Earlier estimates of the total number of sets made on dolphins (column Uoi Table 3) were ob- tained indirectly for the years prior to 1970, based on the catches of tuna, and include an adjustment for nonreported sets. For the Period 1959-72 Estimating the annual rate of dolphin kill dur- ing the period 1959-72 is difficult because obser- vations were few, especially in the early part of the period; consequently, extrapolation of infor- mation on kill rates is necessary. One effect on rate of kill is the development and improvement of the "backdown" dolphin-release procedure (Coe and Sousa 1972; Barham et al. 1977), by which the vessel moves in reverse during a short portion of the purse seine retrieval, thereby pulling the net out from under the dolphins. Barham et al. (1977) reported that the "backdown" dolphin-re- lease procedure was developed aboard one vessel in 1959 and 1960, and transferred to a second ves- sel in 1961. Subsequently, the use of the proce- dure expanded rapidly within the fleet, although dolphin investigations. Background paper 6, prepared for the 39th meeting of the IATTC, Paris. October 1981. Inter- Am. Trop. Tuna Comm., c/o Scripps Inst. Oceanogr., La Jolla, CA 92093, 17 p. 6 SMITH: SIZE CHANGES OF THREE DOLPHIN POPULATIONS full use was not evident even by 1964. Comparing kill rates with and without "backdown" is compli- cated however, because the effectiveness of the release procedure has increased over time. No information is available on kill rates from non-U. S. vessels during 1959-72, but the non- U.S. fleet was small. There is little reason to sus- pect that these kill rates were different, because fishermen of both fleets were still learning how to use purse seine gear for catching tuna in asso- ciation with dolphins and how to release the caught dolphins. The available kill rate data for this period were stratified, for use in Equation (6), by amount of tuna caught, size of the vessel, and fre- quency of use of the "backdown" procedure. The data were pooled across the years 1964-72 and extrapolated back to the years 1959-63 when no kill rate data were collected. These stratified kill rates were multiplied by the number of sets made on dolphins in each stratum to estimate the total number of dolphins, of all populations, killed directly in this fishery. Estimating proportions of the total kill of dol- phins from each population for this period is dif- ficult because the yellowfin tuna purse seining was expanding westward and because data on the species of dolphins observed killed are avail- able only for 1971 and 1972. Prior to 1969 this fishery operated shoreward of the range of the whitebelly spinner dolphin, primarily within the range of the eastern spinner and offshore spotted dolphins. The total kill estimates are prorated to population for the years 1959-72, based on ob- served proportions in the 1971-72 data of 70, 23, and 3% for offshore spotted, eastern spinner, and whitebelly spinner dolphins, respectively. The other 4% consisted of several species, primarily common dolphins, Delphinus delphis, which are not considered in this study. Although the tuna purse seine fishery was expanding seaward throughout the 1960's toward the range of the whitebelly spinner dolphin, a major seaward shift occurred in 1969. Lacking detailed data, I assume this year to be the first significant in- volvement of the whitebelly population. Some additional data on the species of dolphins involved in each set has recently become avail- able from the IATTC, suggesting a declining proportion of sets involving spinner dolphins and an increasing proportion involving spotted dol- phins throughout the 1960's. Preliminary exami- nation of these data indicates that the overall proportions of sets involving each species are not greatly different from the 1971-72 observer data. Direct use of these new data will involve making a number of assumptions about species-specific kill rates. Using the above proportions based on the 1971- 72 data and increasing the estimates of total number killed by 4.8% to account for those dol- phins possibly dying of injuries, I estimated the total numbers of dolphins killed, by population (Table 4). These are revisions of estimates used by the 1979 workshop (footnote 3). Table 4. — Estimates of numbers (in thousands) of dolphins killed by all fleets in the eastern tropical Pacific, 1959-78, for three populations of dolphins (Smith text footnote 3). Offshore Eastern Whitebelly Year spotted spinner spinner 1959 71 27 1960 357 133 1961 402 150 1962 167 62 1963 183 69 1964 306 115 1965 337 126 1966 306 115 1967 206 77 1968 178 67 1969 365 122 15 1970 355 118 14 1971 176 59 7 1972 288 96 12 1973 131 32 33 1974 95 26 47 1975 105 45 34 1976 47 9 20 1977 22 5 5 1978 19 2 4 For the Period 1973-80 Substantially more data exist on kill rates for the period 1973-78 than for the period 1959-72. The 1973-78 data are more reliable because they were collected by NMFS employees trained spe- cifically for obtaining kill information. Starting in 1974 fishing trips were randomly selected for observation to obtain a representative sample. Greater cooperation by the fishing fleet resulted in an increasing proportion of selected trips actually observed from 1974 to 1976. However, it was not until 1976 that fishing trips begun after July were sampled. In the early 1970's fish- ing tended to occur farther offshore later in the year; because kill rates are generally higher in the offshore areas, the failure to collect data from late-season trips probably resulted in an under- estimate of actual dolphin kill rates in those years. This problem is partially accounted for by stratifying the data by area. The species composi- FISHERY BULLETIN: VOL. 81. NO. 1 tion of the kill was also recorded, allowing direct estimates of total kill of dolphins, from Equation (6), for each population. The number of dolphins killed per set from 1973 to 1976 for successful and unsuccessful sets was about 18 and 3, respectively, a decrease from the 1964-72 levels of 56 and 8. The number killed in successful and unsuccessful sets in 1977 and 1978 was again lower, about 4 and 2, respectively (Table 2). These decreases occurred as U.S. regu- lations were developed and eventually imple- mented, and as methods for more effective use of backdown and other dolphin-release procedures were developed and used. The decreases in kill rates were apparently due, at least in part, to wider adoption of procedures for dolphin release. The non-U. S. tuna purse seine fleet increased markedly during this period. First observations of the kill rate for this fleet were in 1979, which showed that the rate was very similar to that of the U.S. fleet (Allen and Goldsmith 1981). Given this similarity in 1979, it is reasonable to assume that during the earlier part of the 1970's the non- U.S. kill rate declined, as did the U.S. kill rate (Table 2), as dolphin-release technology devel- oped by the U.S. fleet became known. If such a decline in the non-U. S. kill rate occurred, how- ever, it would probably have been somewhat slower than that for the U.S. fleet, because of lack of legal pressure to reduce the incidental kill and time lags in technology transfer. Following the procedure developed by the 1979 workshop (footnote 3), I estimated the non-U. S. kill by assuming 1) the same kill rate in 1971-72 for the non-U. S. fleet as that observed aboard U.S. ves- sels in those years; 2) the same kill rate in 1973 for the non-U. S. fleet as that of the U.S. fleet in 1975; and 3) a linear convergence of the two rates toward the 1979 U.S. rate. Estimates of numbers of dolphins killed by non-U. S. vessels obtained under these assumptions are used here. How- ever, additional study is needed, especially since the recorded kill rate for the non-U. S. fleet in 1980 was somewhat higher than that for the U.S. fleet (Allen and Goldsmith 1982). These kill rates, stratified by vessel size, amount of tuna caught, and area fished, are used in Equation (6), along with the estimated num- ber of sets on dolphins, to estimate total direct kill by population for each year. These estimates are then increased by 4.8% to account for dol- phins assumed to die of their injuries (Table 4). The results in Table 4 are slight revisions of the estimates used by the 1979 workshop (footnote 3). NET RECRUITMENT RATE ESTIMATES Maximum net recruitment rate (i?,„) is re- quired to estimate historical abundance. This is calculated as the difference between gross pro- duction of calves and the natural mortality rate, assuming that natural mortality does not change, when a population is reduced substantially be- low its equilibrium level. Gross Reproductive Rates Gross recruitment rates can be estimated as the product of the female fraction of the popula- tion, the mature female fraction, and the annual pregnancy rates. Estimates of these parameters are given in Table 5, based on samples of dol- phins collected by scientific observers aboard tuna vessels from 1973 to 1978. Two methods were used to estimate the annual pregnancy rate: The first method (I) is the observed proportion of pregnant females in the population divided by the gestation period; the second method (II) is similar, but uses additional information on fre- quency of nursing calves in the samples from each net set (Perrin et al. 1977a, b, c). There are known sampling biases in these data for spotted dolphin because of the fishing pro- cess, partly accounted for by using data for spot- ted dolphin recruitment rates from only those sets where more than 40 dolphins were killed. In addition, the observed fraction of the mature, pregnant female dolphins has varied among years, with a general decline in offshore spotted dolphin and a large degree of variability in east- ern spinner dolphin. Age-specific effects are not accounted for in the analyses so far, however, particularly the Table 5. — Proportion of sampled dolphins (female and ma- ture) of three populations and estimates of annual pregnancy rate (P) and gross reproductive rate (G), using two methods. 1 See text for details. Annual production Proportion Method I Metf P lod II Population Female Mature P G G Offshore spotted 0.56 0.56 0.38 0.119 0.32 0.100 Eastern spinner 0.51 0.43 0.34 0.075 0.45 0.099 Whitebelly spinner 0.51 0.52 036 0.096 0.33 0.088 'Henderson, J. R..W. F. Perrin, and R. B. Miller. 1980. Ratesof gross annual production in dolphin populations (Stenella spp and Delphinus delphis) in the eastern tropical Pacific, 1973-1978 Southwest Fish. Cent. La Jolla Lab., Natl. Mar Fish. Serv , NOAA. Admin Rep. LJ-80-02. 51 p. 8 SMITH: SIZE CHANGES OF THREE DOLPHIN POPULATIONS lower pregnancy rate that probably occurs in older animals. New methods are being developed for age determination, and an effort is being made to apply these methods to age the samples of dolphins. With accurate data on age of ani- mals, a more detailed examination of sampling biases will be undertaken. Natural Mortality Rates No direct estimates of natural mortality rates exist for the eastern Pacific dolphin populations, as might be obtained from tagging data or from a sampled age structure. Ohsumi (1979) presented a statistical relationship between natural mor- tality rate and body length for cetaceans, from which can be derived an annual, natural mortal- ity rate of around 0.14 for the eastern Pacific dolphin populations. However, this estimate is obtained by extrapolating the relationship out- side the range of his data, and consequently is unreliable. Another method of estimating natural mortal- ity rate is from information on gross reproduc- tive rate for a population in equilibrium with its environment, assuming natural mortality does not change with population size. This approach was used in the 1976 workshop (footnote 2). An estimate of gross reproductive rate of 0.09 (Ka- suya et al. 1974) for a population off Japan, thought to be lightly exploited, was used as the natural mortality rate estimate for the eastern tropical Pacific populations. It now appears that the population off Japan had, in fact, been ex- ploited to a greater degree than was thought, and that there is segregation of prepubertal dolphins into separate schools (footnote 3, p. 41). The as- sumption, consequently, of a natural mortality rate of 0.09 is probably not valid. In the 1979 workshop (footnote 3), estimates of the gross reproductive rate of lightly exploited Southern Hemisphere populations of spotted and spinner dolphins in the eastern tropical Pacific were used as estimates of natural mortality rates. These rates were 0.098 and 0.067 for spot- ted and spinner dolphins, respectively. Net Rates Net recruitment rates for the offshore spotted, eastern spinner, and whitebelly spinner dolphin populations can be estimated as the differences between the gross reproductive rate estimates, listed in Table 5, and the corresponding natural mortality rate estimates given above. Using method I estimates of pregnancy rates, one ob- tains estimated net reproductive rates of 0.021, 0.008, and 0.029 for these three populations, re- spectively. Using method II estimates of preg- nancy rates, one obtains estimates of 0.002, 0.032, and 0.021, respectively. These highly variable estimates are unsatisfactory, because they are based on data with known sample biases, and they differ among populations in unexpected ways. In particular, it is not expected that the net reproductive rate of the whitebelly spinner dol- phin, which has been relatively less exploited, should be higher than that of the more heavily exploited eastern spinner dolphin popula- tion. Due to these uncertainties, specific point esti- mates were not obtained by the 1979 workshop participants. Rather, a range of values from 0.0 to 0.04 were considered equally likely, given the available information. The lower value of 0.0 was selected by the 1979 workshop to reflect uncer- tainties about unexpected changes in some repro- ductive rates, and the small magnitude of the estimates of net reproductive rates. This range compares with the estimates from the 1976 work- shop of 0.02-0.06, with a midpoint estimate of 0.04. Although higher rates of increase of ceta- cean populations have been reported, contrary to the conclusions in the 1979 workshop report, there are no reliable estimates of rates of increase for dolphin populations which can be used with confidence. Pending better information, the range of estimates considered in the two work- shops will be used here, recognizing that higher rates may be possible. Rate Dependent on Population Size The evidence on which to base an estimate of the value of Z in Equation (3) for dolphin popula- tions is limited. Fowler (1981) argued that for large, long-lived mammals, Z is greater than unity. He based this conclusion on a review of empirical data, primarily from terrestrial popu- lations, and on an analysis of the demographic constraints which come with long life and ex- tended parental care. McCullough (footnote 3, p. 8) gave preliminary estimates of maximum net productivity level (MNPL), and hence Z, for four large terrestrial mammal populations. His esti- mates agree with Fowler's conclusions that Z is greater than unity, and that later reproducing animals would have higher values of Z. FISHERY BULLETIN: VOL. 81. NO. 1 The 1976 workshop (footnote 2) concluded that the available information implies MNPL is with- in the range of 50-70% of the equilibrium popula- tion size, corresponding to values of Z from 1 to 5.1. The 1979 workshop recognized that "There had been a shift of scientific opinion in recent years [since 1976] towards accepting the idea that relative net productivity in mammals, especially large, K-selected species, is a non-linear func- tion of population size," (footnote 3, p. 7) and con- cluded that MNPL for these dolphin populations is probably in the range of 65-80% of the equilib- rium population size (Zfrom 3.5 to 11.5). I con- sider the values for MNPL of 50-80% (Zfrom 1 to 11.5) of equilibrium population size in order to explore the sensitivity of the calculations to this uncertainty. HISTORICAL TRENDS IN ABUNDANCE Estimates of population sizes prior to 1979 from Equations (2) and (3) for each population are shown in Table 6. Values are given using 1) two different estimates of present (1979) abun- dance (from combined research vessel and fish- ing vessel data, and from research vessel data alone), and 2) the parameters MNPL = 65% and R,„ — 0.03. For this range of parameter values, the offshore spotted dolphin population in 1959 Table 6.— Estimates of population size (in thousands) of off- shore spotted, eastern spinner, and whitebelly spinner dolphins from 1979 back to 1959, using Equations (2) and (3) and param- eters MNPL = 65% and R,„ = 0.03. 1979 estimates are based on species proportions from (FR) combined research vessel and fishing vessel data and (R) research vessel data alone. Offshore spotted Year FR Eastern spinner FR = R Whitebelly spinner FR 1979 1978 1977 1976 1975 1974 1973 1972 1971 1970 1969 1968 1967 1966 1965 1964 1963 1962 1961 1960 1959 2,775 2,719 2,668 2,673 2,675 2,679 2,754 2,967 3,064 3,340 3,264 3,720 3,844 4.071 4,335 4,574 4,695 4,803 5,169 5,519 5.590 1,682 1,653 1,628 1,629 1,686 1,732 1,824 2,046 2,164 2,457 2,756 2,865 3,001 3,239 3,520 3,754 3,879 3,991 4,358 4,708 4,779 293 287 283 284 320 336 358 443 488 591 695 742 799 893 998 1,092 1,141 1,185 1,323 1.454 1.481 380 376 373 386 434 456 486 494 499 512 527 527 527 527 527 527 527 527 527 527 527 216 215 214 229 258 301 331 340 345 358 373 373 373 373 373 373 373 373 373 373 373 was between about 4,800,000 and 5,600,000 ani- mals. The eastern spinner dolphin population in 1959 numbered about 1 ,500,000, while the white- 1.0 0.9 LU 0.8 N (/) ( Z o 1- 06 < _l 3 Q. Oh o a. LU 04 > 1- < 3 _l LU OC 0.2 0.1 0.0 Whitebelly \Spinner _l I L_ Eastern Spinner _L 1960 1965 1970 YEAR 1975 Figure 1. — Relative population sizes of whitebelly spinner, offshore spotted, and eastern spinner dolphins, 1959-79, using population estimates based on species proportions from com- bined research and fishing vessel data, and assuming R „, = 0.03 and MNPL = 65% of equilibrium abundance. Population sizes are relative to estimated population sizes in 1959. 1.0 0.9 0.8 LU N (/) Z o 0.7 0.6 a 0.5 O a. iu 0.4 > < 0.3 _i LU K 0.2 0.1 0.0 Whitebelly \Spinner \ Eastern N — _ Spinner _L_I I i_ 1960 1965 1970 YEAR 1975 Figure 2.— Relative population sizes of whitebelly spinner, offshore spotted, and eastern spinner dolphins, 1959-79, using population estimates based on species proportions from re- search vessel data alone, and assuming R ,„ = 0.03 and MNPL - 65% of equilibrium abundance. Population sizes are relative to estimated sizes in 1959. 10 SMITH: SIZE CHANGES OF THREE DOLPHIN POPULATIONS belly spinner dolphin population in 1969 num- bered between 400,000 and 500,000. The offshore spotted and eastern spinner dolphin populations declined rapidly in the 1960's and early 1970's in the face of kills which were, for example, on the order of 7-12% of the 1965 population sizes. The whitebelly spinner dolphin population declined most rapidly in 1974 when the kill was between 11 and 16% of its population size. These estimates of absolute population sizes are shown in Figures 1 and 2 relative to the equi- librium population size (N t /N,,), so that the trend in abundance of these populations can be exam- ined. For all of the parameter values considered, these dolphin populations have declined substan- tially relative to their pre-exploitation sizes. The ratio of 1979 to pre-exploitation popula- tion sizes for different values of R,„ and MNPL (and hence Z) shows the sensitivity of the calcula- tions to changes in parameter estimates (Table 7; Figs. 3,4). The value of MNPL when R,„ is zero is not meaningful, as the estimate of pre-exploi- tation population size (Equation (2)) collapses to the sum of the present population size estimate and the total numbers killed over all years. This is reflected in Figures 3 and 4 in the convergence of the lines when R m is zero. Table 7.— Estimates of 1979 relative population sizes of off- shore spotted, eastern spinner, and whitebelly spinner dolphin populations, using two estimates which differ in species pro- portions from (FR) combined fishing and research vessel data and from (R) research vessel data alone, for ranges of maxi- mum net recruitment rate (R,„ ) and maximum net productiv- ity level (MNPL). Offshc re Eastern Wh itebelly MNPL (%) spotted spinner FR = ft S| Dinner Rm FR ft FR ft 0.00 — 0.40 0.29 0.17 066 053 0.03 50 0.45 0.32 0.18 0.69 0.55 65 0.50 0.35 020 0.72 0.58 80 0.53 0.37 0.21 0.77 0.61 0.06 50 049 0.35 0.20 0.71 0.57 65 060 0.42 023 0.78 0.63 80 0.68 0.47 0.25 086 0.69 1.0 0.9 - UJ N 0.8 7 O 0./ 1- < -J 0.6 3 0. O 0.5 LU > P 0.4 < -1 LU 0.3 0.2 0.1 0.0 0.00 0.03 R m 0.06 1.0 i- 0.9 - B 0.8 i °' 7 H < 0.6 - Q. O 0.5 Q. LU > 0.4 I- < uj 0.3 h K 0.2 - 0.1 - 0.0 m 5>- ef*a •-- =B) Easternjg-^^ESS 0.00 0.03 'm 0.06 Figure 3.— Population size of offshore spotted dolphins in 1979 relative to 1959 (Nn/Nss) as a function of maximum recruit- ment rate (i?,,, = 0, 3, 6%) using two current population estimates which differ in species proportions from (FR) combined fish- ing and research vessel data and from (R) research vessel data alone. MNPL values of 50% (dashed lines), 65% ( solid lines), and 80% (dot-dashed lines) are shown. Figure 4.— Population sizes of eastern spinner and whitebelly spinner dolphins in 1979 relative to 1959 (Ni 9 /N 59 ) as a function of maximum recruitment rate(.R„, =0, 3, 6%) using two current population estimates which differ in species proportions from (FR) combined fishing and research vessel data and from (R) research vessel data alone. MNPL values of 50% (dashed lines), 65% (solid lines), and 80% (dot-dashed lines) are shown. 11 FISHERY BULLETIN: VOL. 81. NO. 1 DISCUSSION AND CONCLUSIONS The three populations of dolphins involved with the yellowfin tuna purse seine fleet in the eastern tropical Pacific have declined since 1959 and the decline was not arrested until recently (Figs. 1, 2). Assuming the historical kill level, and the central values for R m and MNPL, the whitebelly spinner dolphin population has de- clined to between 58 and 72% of its pre-exploita- tion levels; the offshore spotted dolphin popula- tion has declined to between 35 and 50% of its pre-exploitation size; and the eastern spinner dolphin population has declined to around 20% of its pre-exploitation size. Examination of Figures 3 and 4 shows that the numerical values of the estimates of relative abundance in 1979 for offshore spotted dolphin and whitebelly spinner dolphin are relatively more sensitive to changes in the maximum net recruitment rate and the maximum net produc- tivity level parameters than are the estimates for the eastern spinner dolphin. Also, the sensitivity of these calculations to the maximum net pro- ductivity level increases markedly as the value of the maximum net recruitment level increases. The sensitivity (in percent change) in the ratio of present to pre-exploitation abundance, however, is largest for the offshore spotted dolphin and least for the whitebelly spinner dolphin. This is due in part to the shorter time span over which the whitebelly spinner dolphin has been exploit- N t ed, and in part to the lower — — ratio for the east- ern spinner dolphin ratio, which makes smaller differences result in a larger percentage. Although there are a number of uncertainties about specific parameter estimates used in these calculations, the general declines in abundance change relatively little over the ranges of param- eter estimates explored. For example, rather rapid declines in the 1960's, followed by decreas- ing rates of decline in the 1970's, are evident for all parameter values considered. Specific aspects of these declines in abundance, however, depend to a greater degree on the actual parameter val- ues. For example, the estimated changes in popu- lation sizes from 1975 to 1978 vary with the spe- cific values of maximum net recruitment rate, while the estimated changes in population sizes in the 1960's are relatively insensitive to this parameter. In order to improve our estimates of reproduc- tive and mortality rates, a complete review of vita! rates for these dolphin populations and for cetaceans in general should be carried out. Sev- eral approaches to this problem have been identi- fied, including a detailed review of the eastern tropical Pacific dolphin data and of the existing data for other cetacean populations. Given the gaps in our knowledge of cetacean reproductive processes, analyses of alternate mathematical models of such processes will be fruitful. Although improvements in estimates of abun- dance and kill levels are needed, these areas are generally much better understood than the re- cruitment process. Population-size estimation techniques are still being improved upon; cur- rent emphasis is on testing the assumptions needed in applying line transect theory to aerial sighting survey data and in estimating dolphin school size. Future work will emphasize im- proved shipboard sighting methodology for pos- sible application of line transect theory. Marked improvements in the estimates of num- bers of dolphins killed are not anticipated; key areas needing additional information are the kill rates both in the non-U. S. fleet and on unobserved fishing trips. Neither of these areas is readily amenable to study, although further analysis of the kill rates on unobserved trips may provide some basis for exploring this uncertainty. The possible levels of indirect mortality or debility due to the stress of chase and capture are also of concern. Because of the large numbers of dol- phins captured and released each year, even very low rates of indirect mortality could have a sig- nificant effect on the population. ACKNOWLEDGMENTS This assessment of the status of the dolphin populations is built on data collected by many in- dividuals. The collection of these data has been made possible in large measure by the coopera- tion of the U.S. tuna fishing fleet. In addition, many individuals have contributed to the analy- sis of the data, including National Marine Fish- eries Service staff and numerous scientists from various organizations. It is not possible to ac- knowledge the contributions of specific individ- uals to information presented here because of the large numbers of people who have been involved, but without their efforts the present analysis would not be possible. I also wish to acknowledge the very helpful reviews of an earlier draft of this paper by Douglas Chapman, John Gulland, Linda Jones, and Jeff Breiwick. 12 SMITH: SIZE CHANGES OF THREE DOLPHIN POPULATIONS LITERATURE CITED Allen, K. R. 1981. Application of population models to large whales. In C. W. Fowler and T. D. Smith (editors), Dynamics of large mammal populations, p. 263-275. John Wiley, N.Y. Allen, R. L.,D. A. Bratten.J. L. Laake, J. F. Lambert, W. L. Perryman, and M. D. Scott. 1980. Report on estimating the size of dolphin schools, based on data obtained during a charter cruise of the M/VGina Anne, October 11-November 25, 1979. In- ter-Am. Trop. Tuna Comm., Data Rep. 6, 28 p. Allen, R. L., and M. D. Goldsmith. 1981. Dolphin mortality in the eastern tropical Pacific incidental to purse seining for yellowfin tunas, 1979. Rep. Int. Whaling Comm. 31:539-540. 1982. Dolphin mortality in the eastern tropical Pacific incidental to purse seining for yellowfin tunas, 1980. Rep. Int. Whaling Comm. 32:419-422. Barham, E., W. K. Taguchi, and S. B. Reilly. 1977. Porpoise rescue methods in the yellowfin purse seine fishery and the importance of Medina panel mesh size. Mar. Fish. Rev. 39(5):1-10. Burnham, K. P., D. R. Anderson, and J. L. Laake. 1980. Estimation of density for line transect sampling of biological populations. Wildl. Monogr. 72, 202 p. Coe, J., and G. Sousa. 1972. Removing porpoise from a tuna purse seine. Mar. Fish. Rev. 34(11-12):15-19. Fowler, C. W. 1981. Density dependence as related to life history strat- egy. Ecology 62:602-610. Green, R. E., W. F. Perrin, and B. P. Petrich. 1971. The American tuna purse seine fishery. In H. Kristjonsson (editor), Modern fishing gear of the world 3, p. 182-194. Fish. News (Books) Ltd., Lond. Holt, R. S., and J. E. Powers. 1982. Abundance estimation of dolphin stocks involved in the eastern tropical Pacific yellowfin tuna fishery de- termined from aerial and ship surveys. U.S. Dep. Com- mer., NOAA Tech. Memo. NOAA-TM-NMFS SWFC- 23. Kasuya, T., N. Miyazaki, and W. H. Dawbin. 1974. Growth and reproduction of Stenella attenuata in the Pacific coast of Japan. Sci. Rep. Whales Res. Inst. (Tokyo) 26:157-226. Lo, N. H., J. Powers, and B. Whalen. 1982. Estimating and monitoring incidental dolphin mortality in the eastern tropical Pacific tuna purse seine fishery. Fish. Bull., U.S. 80:396-401. Ohsumi, S. 1979. Interspecies relationships among some biological parameters in cetaceans and estimation of the natural mortality coefficient of the southern hemisphere Minke whale. Rep. Int. Whaling Comm. 29:397-406. Pella, J. J., and P. K. Tomlinson. 1969. A generalized stock production model. [In Engl. and Span.] Inter-Am. Trop. Tuna Comm.. Bull. 13: 421-496. Perrin, W. F. 1975. Distribution and differentiation of populations of dolphins of the genus Stenella in the eastern tropical Pa- cific. J. Fish. Res. Board Can. 32:1059-1067. Perrin, W. F., J. M. Coe, and J. R. Zweifel. 1977a. Growth and reproduction of the spotted porpoise, Stenella attenuata, in the offshore eastern tropical Pa- cific. Fish. Bull., U.S. 74:229-269. Perrin, W. F., D. B. Holts, and R. B. Miller. 1977b. Growth and reproduction of the eastern spinner dolphin, a geographical form of Stenella longirostris, in the eastern tropical Pacific. Fish. Bull., U.S. 75:725- 750. Perrin, W. F., R. B. Miller, and P. A. Sloan. 1977c. Reproductive parameters of the offshore spotted dolphin, a geographical form of Stenella attenuata, in the eastern tropical Pacific, 1973-1975. Fish. Bull., U.S. 75:629-633. Perrin, W. F., P. A. Sloan, and J. R. Henderson. 1979. Taxonomic status of the "southwestern" stocks of spinner dolphin, Stenella longirostris and spotted dol- phin, S. attenuata. Rep. Int. Whaling Comm. 29:175- 184. Polacheck, T. 1982. Local stability and maximum net productivity levels for a simple model of porpoise population size. U.S. Dep. Commer., NOAA Tech. Memo. NOAA-TM- NMFS SWFC-17. Smith, T. D. 1981. Line transect techniques for estimating density of porpoise schools. J. Wildl. Manage. 45(3):650-657. Smith, T. D., and T. Polacheck. 1979. Analysis of a simple model for estimating histori- cal population sizes. Fish. Bull., U.S. 76:771-779. 13 FOOD HABITS OF YELLOWTAIL FLOUNDER, LIMANDA FERRUGINEA (STORER), FROM OFF THE NORTHEASTERN UNITED STATES Richard W. Langton 1 ABSTRACT Stomachs of 1,021 yellowtail flounder caught in 1973-76 contained primarily polychaetes (43%) and crustaceans (18%) as a percentage weight of total contents. The most important prey were Spiophanes bombyx (9.68%) and Unciola sp. (13.65%). Predator size had little effect on diet com- position whereas geographic distribution did. Spiophanes bombyxwas three times more important as prey on Georges Bank than in southern New England, and amphipods were more important in southern New England than on Georges Bank. From the middle Atlantic to southern New England to Georges Bank the total weight of stomach contents increased from 0.12% to 0.14% to 0.21% of the fishes' body weight. Year-to-year differences were inconsistent; however, fish stom- achs from spring cruises contained more food, 0.20%, than those from autumn cruises, 0.14% body weight. During a composite 24-hour day, peak stomach content weight occurred in the afternoon to early evening. Polychaetes accounted for less of the stomach contents at night while amphipods increased in importance during the night. Sex of the fish had no effect on diet composition although the stomachs of females were fuller than males, 0.15% vs. 0.11% body weight. Neither diet composition nor the percentage of empty stomachs were related to gonadal maturity stages, but stomachs from spawning fish contained the least amount of prey, 0.06%, while resting-stage fish contained the most, 0.24% body weight. Over a 12°C temperature range there was little change in diet composition, but between 3° and 8°C a greater percentage of stomachs contained prey and a larger quantity of prey than between 9° and 15°C. Over a 220 m depth range the stomach content weight increased with depth for smaller fish (<15 cm), while the percentage of empty stomachs increased for larger fish (>21 cm). Diet composition showed the greatest effect of depth with S. bombij.r dominating the diet in the 74-110 m depth zone (26.6% of the stomach content weight) and Crangon septemspinosa, also being dominant in a single depth zone, comprising 39.6% of the diet at 147-183 m. The yellowtail flounder, Limanda ferruginea (Storer), is a right-handed, thin-bodied flounder that occurs along the eastern seaboard of North America from Labrador to Chesapeake Bay (Bigelow and Schroeder 1953; Royce et al. 1959). It has contributed significantly to the total flat- fish catch, primarily from southern New Eng- land and Georges Bank, since about 1935 (Royce et al. 1959; Sissenwine et al. 1978 2 ). Biological in- formation has been summarized by Bigelow and Schroeder (1953) and updated by Lux and Liv- ingston (in press). These summaries qualitatively describe the diet as consisting of small crusta- ceans, worms, and molluscs. Quantitative work ■Northeast Fisheries Center Woods Hole Laboratory, Na- tional Marine Fisheries Service, NOAA, Woods Hole, Mass.; present address: Maine Department of Marine Resources, Ma- rine Resources Laboratory, West Boothbay Harbor, ME 04575. 2 Sissenwine, M. P., B. E. Brown, and M. M. McBride. 1978. Yellowtail flounder (Limanda ferruginea): Status of the stocks, January 1978. Northeast Fisheries Center Woods Hole Laboratory Reference No. 78-02, 27 p. on the diet is limited. Inshore yellowtail flounder have been examined by Libey and Cole (1979) off Cape Ann in Massachusetts while Efanov and Vinogradov (1973) surveyed the offshore feeding pattern of yellowtail flounder in southern New England and on Georges Bank. Langton (1979 3 ), Grosslein et al. (1980), and Langton and Bowman (1981) described the diet of fish from the middle Atlantic to western Nova Scotia, and Pitt (1976) conducted a study on the Grand Banks. These papers generally agree that crustaceans, par- ticularly amphipods, and polychaetes are major prey items. However, the absolute quantities of prey in the stomachs differ, being influenced by both biological and abiotic factors. Only one of the studies lists the stomach contents by predator size (Pitt 1976) and none of the studies evaluate comprehensively all factors influencing the diet Manuscript accepted June 1982. FISHERY BULLETIN: VOL. 81, NO. 1. 1983. 3 Langton, R. W. 1979. Food of yellowtail flounder, Li- ma ndaferruginea( Storer). International Council for Explor- ation of the Sea. CM. 1979/G:54, 10 p. 15 FISHERY BULLETIN: VOL. 81. NO. 1 of this predator. The purpose of the present paper is to describe the stomach contents of yel- lowtail flounder and quantitatively evaluate fac- tors influencing the quantity and composition of the animal's stomach contents. METHODS Yellowtail flounder stomachs were collected on eight bottom trawl survey cruises conducted from 1973 through 1976. The dates of the cruises are as follows: 16 March-15 May 1973; 26 Sep- tember-20 November 1973; 12 March-4 May 1974; 20 September-14 November 1974; 4 March- 12 May 1975; 15 October-18 November 1975; 4 March-8 May 1976; 20 October-23 November 1976. Fish collections were made from the RV Albatross IV or RV Delaware II, using a #36 Yankee otter trawl for autumn surveys and a #41 Yankee otter trawl for spring surveys. A scheme of stratified random sampling was carried out in the continental shelf waters between Nova Scotia and Cape Hatteras, N.C. For survey purposes this region has been divided into five geographic areas, which are further subdivided into depth strata as depicted in Clark and Brown (1977) and described by Grosslein (1969 4 ). Yellowtail flounder were selected from the catch in, primarily, two of the five geographic areas, i.e., southern New England and Georges Bank which include the three major fishing grounds and major yellowtail flounder stocks in U.S. waters (Lux 1963). Stomachs were labelled according to vessel, cruise, station, length, sex, and sexual maturity and were preserved individ- ually in a gauze wrapping in 10% Formalin 5 . The sampling strategy was designed to collect fish, more or less at random, from the population with- out bias towards a specific length, except as de- scribed below. We attempted to collect 50 fish per geographic area per cruise for fish both above and below 12 cm TL (total length). Twelve centimeters in length approximates the length of 1- to 2-yr-old fish, and these smaller fish were preserved intact after the body cavity was cut open to insure fixation of the contents. In the laboratory, individual stomachs were opened, and the contents emptied onto a fine mesh screen and rinsed with seawater. The vari- 4 Grosslein, M. E. 1969. Groundfish survey methods. Northeast Fisheries Center Woods Hole Laboratory Reference No. 69-2, 34 p. 5 Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. ous items were sorted and identified to the lowest possible taxa. Each distinct group was blotted dry and immediately weighed. In the text and tables these weights have been expressed as a percentage of the total weight of stomach con- tents. In the text these percentages are often given in brackets after the mention of taxa to quantify their relative importance. Twelve percent of the fish collected fell into the three smallest size classes (Table 1) with a mean length of 7.6 cm. Fish >15 cm TL were equally distributed around the 31-35 cm size class with 70% (h =715) of all fish examined falling between 26 and 40 cm TL. The average length of all fish comprising this peak is 32.8 cm. For some analy- ses two size-related groupings of fish, represen- tative of this bimodal distribution, have been differentiated while in other cases the data are presented by 5 cm length classes or expressed as a percentage of the fishes' body weight according to the length/weight equation in Wilk et al. (1978). [W=aL h where a = 0.4514" 5 , 6 = 3.1257, and L is in millimeters.] RESULTS Food Of the 1,021 stomachs examined, 684 contained prey which weighed in total 422 g. The overall mean fish length and standard deviation was 29.4+10.5 cm. The prey were allocated into 148 different categories, which included all taxonom- ic levels of identification and such miscellaneous categories as sand and unidentifiable animal re- mains. The most important major taxonomic groupings were polychaetes and crustaceans (Table 1). Polychaetes accounted for 43% of the stomach contents. The families Spionidae (13.27%), Lum- brinereidae (1.90%), Sabellidae (1.42%), and Nephtyidae (1.19%) were all of some importance. Spiophanes bombyx was the major prey, making up 9.68% of the weight of the total stomach con- tents. Other polychaetes (17.24%) and polychaete tubes (7.94%) accounted for the remainder of the prey in this taxon. Crustaceans (18.0%) were second in impor- tance, the amphipods (13.65%) being the major prey group. Unciola sp. (4.41%), Leptocheirus pinguis (2.25%), and Byblis serrata (1.72%) were important amphipod prey. Other gammarids (1.92%), ampeliscids (1.56%), and corophiids (0.3%) made up most of the remaining amphipod 16 LANGTON: FOOD HABITS OF YELLOWTAIL FLOUNDER Table 1.— Principal items in stomachs of yellowtail flounder, Limandaferruginea, by 5 cm length classes. Data are expressed as a percentage of the total weight of stomach contents (+ indicates present but <0.01%). Fish length intervals in centimeters Stomach contents 1-5 6-10 11-15 16-20 21-25 26-30 31-35 36-40 41-45 46-50 51-55 Anthozoa 1.09 0.69 4.24 8 10 3.69 971 Other Cnidaria 0.01 Sptophanes bornbyx 11.74 5.82 0.93 1028 11.97 11.99 11.06 Spionidae 0.60 1.77 2.71 4.91 9.80 Sabellidae 2.02 0.97 3.34 088 0.16 Annelida tubes 1.01 4.46 12.12 916 14.35 Lumbrinereidae 0.54 3.05 349 1.05 0.11 2.22 Nephtyidae + 0.54 0.28 280 0.92 0.23 039 Other polychaetes 10.82 11.91 13.64 29.19 2335 21.83 14.51 1646 1226 0.33 Byblis serrata 1.73 2.02 0.59 1.24 1.24 2.35 2.50 2.34 0.03 Other Ampeliscidae 1471 0.79 1.81 0.71 2.45 2.25 1.40 0.54 0.92 Unciola sp + 2.22 3.64 1528 1286 1203 5.36 3.44 0.26 0.60 Other Corophndae 4.02 3.31 0.93 0.33 1.34 0.32 0.07 0.01 Gammandae + 3.77 4.96 1.47 1.61 421 2.14 1.35 2.79 0.12 Leptocheirus pmguis + 8.51 4.77 3.81 1.92 3.11 1.96 Other Amphipoda 6.24 0.93 1.43 1.27 2.55 1.03 1.53 0.18 3.87 Crangon septemspmosa 41.38 9.09 11.98 3.25 7.14 2.53 0.71 001 Dichelopandalus leptocerus 31.83 0.64 1.96 066 009 Other crustaceans 58.62 19.16 4.37 1928 6.94 2.32 1 96 0.72 1065 0.16 Animal remains + 15.39 7.28 28.42 2446 19.30 1609 15.47 3679 0.59 Sand 0.93 0.60 1.81 10.19 822 10.22 17.34 23.02 1030 8845 Other groups + 0.19 18.40 5.39 0.31 2.28 3.38 1.25 1.99 0.93 0.88 Number examined 39 77 21 23 63 187 337 191 60 18 2 Number empty 22 25 6 9 16 63 108 62 20 4 Mean weight per stomach (g) 00001 0.021 0072 0.103 0230 0242 0375 563 0.950 2.445 9652 Mean length (cm) 4.0 8.3 120 18.1 23.4 283 32.7 377 42.3 47.5 51.0 prey. Only two other crustaceans were of signifi- cance in the yellowtail flounder's diet, namely, the shrimps Crangon septemspinosa (1.89%) and Dichelopandalus leptocerus (0.94%). All other taxonomically distinct groups con- tributed only 4.96% of the weight of stomach con- tents. Unidentifiable animal remains (17.18%) and sand ( 16.92%) accounted for the remainder of the total weight of stomach contents. Size-Related Feeding Habits Amphipods were the most important prey for the smaller yellowtail flounder although stom- achs from every size class of fish contained am- phipods (Table 1). Polychaetes comprise a greater percentage of the stomach contents of the larger fish but, like amphipods, they occur in stomachs from most every size class. The occurrence of anthozoans in the larger size fish (>26 cm) might reflect a tendency for larger yellowtail flounder to be selecting "wormlike" prey. Geographic Comparison Composition of the diet of yellowtail flounder in southern New England and on Georges Bank was similar, with polychaetes and amphipods accounting for 50 to 70% of the total weight of stomach contents in both areas (Table 2). Poly- chaetes were the major prey in both regions with TABLE 2.— Principal items in stomachs of yellowtail flounder, Limandaferruginea, by geographic area in the northwest At- lantic. Data are presented as a percentage of the total weight of stomach contents (+ indicates present but <0.01%). Southern Middle New Georges Stomach contents Atlantic England Bank Anthozoa 1.93 3.52 Other Cnidaria 0.01 Spiophanes bornbyx 4.35 13.18 Spionidae 4.41 3.12 Sabellidae 3.30 0.25 Annelida tubes 5.47 7.53 824 Lumbrinereidae + 2.57 1.50 Nephtyidae 266 0.28 Other polychaetes 9.80 22.74 13.89 Byblis serrata 8.15 2.18 1.34 Other Ampeliscidae 1.55 2.44 1.01 Unciola sp. 1.50 7.01 2.81 Other Corophiidae 5.36 0.36 0.19 Gammaridae 0.77 2.25 1.72 Leptocheirus pm- guis 4.38 3.25 1.58 Other Amphipoda 0.54 1.38 1.58 Crangon septemspi- nosa 2.16 1.75 Dichelopandalus leptocerus 1.90 0.35 Other crustaceans 1.82 1 34 Animal remains 41.57 15.53 1785 Sand 20 09 7.75 22.65 Other groups 0.83 2.46 1.86 No. fish examined 16 502 502 No empty stomachs 4 163 169 Mean weight per stomach ±SD (g) 0242±0.324 0.323±0.578 0.512±1.452 Mean length ±SD (cm) 28.2±8.2 29.2±9.2 29.7±11.7 Spiophanes bornbyx being the most important species identified. On Georges Bank S. bornbyx was three times more important as prey than in 17 FISHERY BULLETIN: VOL. 81. NO. 1 southern New England. The other major differ- ence between areas was in the quantity of other polychaetes, but a large percentage of this group was unidentified remains (12.93% in southern New England and 11.61% on Georges Bank). The diversity of polychaete prey was very similar in the two areas; 27 families of polychaetes in the stomach contents of fish from southern New Eng- land and 24 different families on Georges Bank. Eleven different genera of polychaetes were iden- tified in each area. Six of these were common to both regions, but only Spiophanes contributed >1% to the total stomach contents weight. Amphipods made up almost twice the percent- age of the weight of stomach contents in southern New England than on Georges Bank (18.87% vs. 10.23%). The same species were important in both areas (Table 2). There was, however, a slightly greater reliance on Unciolasp. and Lep- tocheirus pinguis in southern New England than on Georges Bank. The diversity of amphipod prey was greater on Georges Bank, 16 genera as opposed to 11 genera, although yellowtail floun- der from the two areas preyed on 9 of the same genera. Crustaceans such as C. septemspinosa and D. leptocerus played a minor role in the diet of yel- lowtail flounder as did all other arthropod groups except the amphipods. The only other category of stomach contents that differed substantially be- tween areas was the quantity of sand in the stom- achs. This might be related to the heavy preda- tion on «S. bombyx on Georges Bank, since this polychaete is reported to prefer a fine sand sub- strate (Light 1978). The percentage of empty stomachs was virtu- ally the same in southern New England and on Georges Bank, but was less in the Middle Atlan- tic (Table 2). The mean weight per stomach in- creased from the Middle Atlantic to Georges Bank and the mean fish length also increased from south to north(Table2). This size difference did not counterbalance the increase in stomach content weight. The mean weight of stomach con- tents ranged from 0.12% in the Middle Atlantic to 0.14% in southern New England and 0.21% body weight on Georges Bank. Yearly, Seasonal, and Diurnal Variation Data were collected over a 4-yr period in both the spring and autumn and throughout the day- night cycle. It is, therefore, possible to examine the influence of the time of capture on the com- position of the diet as well as on changes in the absolute quantity of prey in the stomachs. On a year-to-year basis, polychaete worms were always the most important prey, between 36 and 44% of the diet, followed by amphipod crusta- ceans, 10 to 33% of the diet. Within these two taxa the actual percentage composition of the various groups fluctuated, but no systematic changes in diet were discernible. Within the Polychaeta, for example, S. bombyx made up between 2 and 12% of the diet from 1973 to 1974 and ranged from 9 to 11% between 1975 and 1976, respectively. At the family level, Spionidae, the range increased from 2 to 16% for the first 2 yr and 9 to 18% for the lat- ter 2 yr. When spionids were most important, there was also a very large percentage of sand in the stomachs, 20 and 27% for 1974 and 1976, re- spectively, which probably relates to predation on these particular polychaetes. Among the am- phipods, Unciola sp. showed the greatest fluctua- tion, ranging from 16% of the diet in 1973 to 1% in 1975 but increasing to just under 5% in 1976. The mean weight of prey showed an increase from 1973 to 1976, but when this was corrected for fish size, there was no pattern evident in these changes. The slightly larger mean fish lengths occurring in 1975 and 1976 counterbalanced the increase in the mean weight of stomach contents. The percentage of empty stomachs also showed no consistent yearly change, fluctuating around the overall mean value of 33%. Species composition of the diet showed no dras- tic shift between spring and autumn. Polychaetes were more important in the spring (49%) than in the autumn (35%), and the same was true for am- phipods, 19% vs. 13%. Both of the changes may, however, simply reflect the higher percentages of unidentified animal remains and sand in the fish stomachs collected in the autumn. In all years, except 1976, stomachs collected on spring cruises contained a greater mean weight of prey than stomachs from fish collected in the autumn. Although the mean length offish in the spring was only slightly larger (30.0 cm vs. 28.8 cm). The 4-yr mean weight of prey in the stom- achs was 0.505 g (0.20% body weight) for the spring and 0.298 g (0.14% body weight) in the autumn. The percentage of empty stomachs was also lower in the spring than the autumn; 22.7% of the 574 stomachs examined from spring cruises versus 46.3% of 447 stomachs examined from autumn cruises. An examination of the data for a composite 24-h day revealed a diurnal feeding pattern. 18 LANGTON: FOOD HABITS OF YELLOWTAIL FLOUNDER Although there was a certain degree of hour-to- hour variability, a peak in the weight of stomach contents occurred during the afternoon-early eve- ning period (Fig. 1). In Figure 1 the day has been divided into four periods— dawn (0300-0800 h), day (0900-1400 h), dusk (1500-2000 h), and night (2100-0200 h)— which accounts for seasonally variable day length in the dawn and dusk period. Despite a seasonal change in day length, the com- position of the diet also changed over a 24-h peri- od. Polychaetes were less important prey during the night than during any of the other three time periods. They dropped from values ranging from 41-47% to 24% as a percentage of the weight of stomach contents. Conversely, crustaceans, am- phipods in particular, were more important at night (values ranging from 15 to 23% vs. 34%). Unidentifiable animal remains also accounted for their smallest percentage of the diet (13.0%) in the dusk period when the fish stomachs were fullest. The greatest percentage of empty yellow- tail flounder stomachs was found during the night (46%) and the smallest ( 19%) during the day with intermediate levels occurring at dawn (34%) and dusk (26%). To evaluate whether the diurnal feeding pat- tern shown in Figure 1 is statistically significant, an analysis of variance, including time of day and seasonal factors, was conducted. The results of this analysis are given in Table 3 for trans- formed data using an inverse hyperbolic sine transformation (Y' = sin h" 1 i\fY)) to account for the extreme skewness of the data (i.e., a large number of empty and almost empty stomachs) (see Bartlett 1947). Both time of day and season are significant factors in determining the weight of stomach contents for yellowtail flounder. This analysis confirms that there are statistically sig- nificant differences in stomach content weight over a 24-h period. These results are, however, influenced by the level of interaction between time of day and season, such that it is not clear which of these two factors is the most important Table 3.— Analysis of variance of the weight of stomach contents for yellowtail flounder, ex- pressed as percent body weight, for time of day and season. See text for details. Source df Sum of squares F Time of day Season Interaction Error 3 1 3 1,011 0.03376 0.02464 0.00894 0.78005 1459" 31 93* 386* 'Significant, P = 0.05 "5- 0.30 cu 5 O CD 3 ^Z cd -t-» CO -C CD he S 3 3 cd c-3, 3 C D> O CD- £ CO. » CO 3 « C Ol CO o 3 « C o> CO £ en CO PS I — .O LI O ai (1) C u a.— LU CD >.T3 III 3 CO £ o 5 « c I ai •5£ o Ol c c C0.3> 3 «£ o o 2 ai <= ~ £ c CO ™ » c£-p £ c E Q. CD ~ E~2 oi> E c eg p O' o CO CO Ol TT CO LO Tf in in in cd N N O) CM CM Tf LO CO Tf Tf CM CM CM CM CO CO Tf 1- — ^ CO C m co co co O 00 CO -^ CM Tf ° 2 CO o LO CO CD Tf CO CO Tf LO LO Tf o co m o CM co co yf Tf m CD I s - r» r- CO I s - Ol I s - Oi o> O CM r- O y- CM CM CM CO CO •~ — 5 C-O CM y- 0> CO ? 5" S CM CO CO CO CO Ol CO O i- CO o 00 CD T- CM TT o CD CM co r- o Ol co in co o m CO N N MO O ■- >- CM r- t- CM CM CM CM C- CO CO CM ■** co co o co in co -^ in in lo co *-r CM CM CO CO CO f CD f- in CM Ol LO CM CO Tf CO CO CM CO CO co in CM CM CO CO Ol Ol Tf CM o CO LO CM O CM Ol LO CM TT 01 CD CO CD O O CM CM CO Tf Tf LO O CO CM CO OWSO"- CM CM CM CO CO cm in co cm cd CM CM CO CO CO d d d d d >-OCJ)i-f n coir in m ■a o I s - w O) 00 Tt Tt d d ■<* ■<»■ CD o a CO CD O o o o o CO LO CO O 0) CO CO CO O CO o p CO Ol CM p >- CO m CD CM O CD o co co in id co s CO CM o CM CO CD Cli LO CM CM CO CO CO CM CM O CO Tf CM CM CO I-- CM LO Oi CM Ol Ol CO CO CO CO CM CO Tf Tf lh co Tf r- co cm co -- T- T- CM CO Ol CM o o in o r- co CM CO CO CD Ol CO CD CO i- CO o o o o o o o o o CM CM CM CM CM CM CO CO CO CO CO Tf Tf Tf Tf Tf Tf m cd co oi o c\j in in oi r-- co cm t- | co 1- in CO 5 Tf in t- C0 CD in m in CO LO LO CD CO f-- CM ai o in r— co o i- T" CO CO CO CO Ol 1— co i in ^f in in CO CO CO CO CO Ol CD CO CO CO CM CO f- Ol CO y- 00 O o o o o o o o o o O o O O CM CM CM cm co Tf Tf in o cm" in in CM CO C\J CD cm 1 eg co co CO CD CO in" P r^ co ■a- un CO LO o CD o. LO Ol CD 1^ CM t^ CO 1— t^- y- CO Ol Ol s ^ CO O T- CM t- Tf CO in co CO LO 1 CO Tl- m in ai in O LO O o Tf m oo o d odd o o o o O o O o O o o CM CM CO co co Tf Tf m cm in in O O) CO Tf CO C\J i- CM CO CO o d d (6 d CO O) co tt Tf co co r- cm co CO CC CO o Ol Tf in i- co o in in d d o in CO o 35 CD O m o LO o CO d CO O Ol Ol Ol o o> CO CO O CO Ol o CT> T-; in ■- f i- CM CO 5" n in co cm ^ in ^ 9 9 t CO CO Tf CD O CO Cli I s Tf in O CM LO CM LO LO CM CO LO CM CM CO CD CO CM r^ CO o cq CO in CD Ol Tf ai CM CM r- o O CM O <- CO o co in -^ co lo ^r oi in co o cd in CO CD TT CO Tf t- CO Tf CM CM CM CM CM CO CO CO •» i T in T in in co co cd r- r- r— f- r- r-- o CO Ol o O y- Tf in CD co m in tt co co in co r- CM J- >- CO CO Ol CO in P CD -- CM i- o CM I s - CM Ol CM CM Tf Ol CM in CM m Ol CM CO CM CO CO Tf Tf CM CD Tf CD CM CO O O - tt in in co I s - o o in co un CO m CM 1 CO LO o LO CD O CD CM CD CD O 00 CO CM 1- CM CO LO CM CO CD Tj- CM CO ^t CO CO Tf CD r- Tf CM CO Tf lO o o CO LO CO i-ooioo TT in CD N CO Ol O 1- CM CO ^r LO CD h~ CO O) O r~ CM CM CM CM CO Tf in CD N CO CM CM CM CM CM CM Ol O CM CO CO CM CO CO CO Tf LO CO CO CO CO CO Tt CO CD N N t- in co co Tf in lo *- *- co «- h- co O CO CM CO in LO CO LO in in in in in m in co in in in in in CO o> a> Ol o> Ol Ol Ol o> O CT> Ol Ol Ol Ol Ol Ol Ol Ol Ol Ol Ol Ol Ol O) Ol Ol Ol Ol Ol Ol Ol Ol Ol Ol oi oi oi oi oi 't in cbsco Ol o >- CM CO ^3- LO CD r~ cd cr> O r- CM CM CM CM co tj in co s co CM CM CM CM CM CM Ol o CM CO CO CM CO CO CO tt m CO CO CO CO CO Tf CO cd r-- r*- y- in co co Tf in in o o o o o o o o O o O O O O o o O O O o o o o o o o o o O O o o o o o o o o o o ^r in co s co Ol O T- CM CO •t LO CD r^ cd O r- CM CM CM CM co tt in co s co CM CM CM CM CM CM Ol o CM CO CO CM CO CO CO Tf in coo CO CO CO Tf CO CD N N t- m co co Tf lo in 3) and rays develop. Rays develop in the pectoral and pelvic fins. Distinctive "muscle bands," which overlap the base of the caudal fin rays de- velop in transforming and juvenile specimens (Figs. 2C, 3). The adipose fin develops after trans- formation (Fig. 3). 29 FISHERY BULLETIN: VOL. 81. NO. 1 Osteology c - — - 3 > C sj ■D a> c E _o o '-£ .5 > CO c CO . 13 H CD CD E 2 T3 3 <° T3 C rt *j m c a) cu CO £ CD to 0) 3 > CO cfi CO > 0) be c c § CO 0) > be c/) 0) 3 e« . to > -s CD 3 0) CD en 03 CO •~ 0) > > •*- .c -a ts CO •~ c pare c o X CD LL so CD CD CO ■° 3 _o E "> "Si 3 w £ s S3 ^ «** o CO 0) CO 'S 11) CD 3 Hi CO > 3 > CO CO > '"— i T3 c o C cfl X 0) 0) "oj oa > c3 a CD £ ■° 3 o CO is C .o +j >-, o ft o i~ ft >, -a o E m CD I — oj w j 09 < H cocom"<*r--h--^.intpm CO CO CO CM CM i- cocomi-ocor-co co o cq p ^r -^ in •r- t- CM CM LO if" T^ -H -H -H -H -H -H -H O CM CO CM O ^ CM r*- m r^ io in *- O) OO CO -<* CD O o o i- 1 r 1 d ^ -H -H -H -H -H -H co n n ^r s o OO)0)O1CT>C7>O")CJ)O)O)O)O)CJ> o co ^ COO) ^tco^- mo cb ow in to ' r ~ cococmcoco co-^tco i- q co coco aiO) 7 cb^rt^r^^ cm co ^ i-h-T-i-r- t-CMCO CO ^T if) ~ oS cm a> u m co in 00 ' ' ' i t-ttO ip'-cOt-cocO'-Nmcoa) r-' C\i (N (O ^' CO i- O i- t- r-' -H-H+l-H+l+l-H-H+i-H-H MOO)TtlD(DOONOlO NCNi^cicb'-OCDCDO)Tf N CM CNJ CM CM CO i- in S in I I CO N CO t CM CO ^T coscpcoscomi T CO t ID <£> t- *? IT) T ty ^ t-IOt-CVJi-t-i-CD^ I I I I I IT) TJ CO t- CD r- r- ] co i-cocMco-sroo -H-H-H-H-H-H-H-H oiownit onoi Loaicbc3cDcbc6-^ h-. t- CM CM CM C\J CVJCMCMCMCMCMCMCMCM — O ^ CO CD ^ co cm en oi CMCOO ' CO "^

^ f~- . a> co ^ciSCMCO-^4 T CDS -- '^r 't CO 'CD O) CO I I I I CO o CO i- O co o m CMCOi-r-.-CO.-0 -H-H-H+t-H+l-H-H cosooicococtico S CD S CO CO CO _1 CO > CO > CO c O) c CD CO c £ - I c o>£ oi £ "O I co CO — c £ ^ I _ — CO o>£ £| 2 O c 0)~ - E we? a" CD t - •D « o <3> Q. CO © 0) _ CJ o <0 ™ ;• c C m O O O §3-D J =5 23££333 OcOcDCDCDOQ.Q.000 COCOlLULUicOCIOcOCOCO 3 3 T3 T3 CD CD a. o. TO TO T3 -D 3 3 CO CO O O (Tables 3, 4; Figure 4) Although a few structures ossify in relatively small L. schmidti larvae, a number of skeletal elements do not calcify until the larvae trans- form into juveniles. Other portions of the skele- ton do not completely ossify until well into the juvenile stage. The general sequence of ossifica- tion is as follows: cleithrum; dentary and vomer- ine teeth; pharyngeal teeth; parasphenoid, den- tary, maxillary, vomer, premaxillary; most other bones of the cranium; certain elements of the caudal fin and, at or near transformation, axial skeleton; dorsal, anal, and paired fins; and gill rakers and secondary caudal fin rays. Teeth Teeth on the dentary and vomer begin to form in 8 mm larvae (Table 3), increasing in number as the larvae grow and doubling in number at transformation. Dentary teeth appear to increase in number with growth in transformed juveniles, but vomerine teeth remain relatively constant in number. In specimens about 12 mm long, one pharyngeal tooth develops on each plate (Table 4); after transformation the teeth then increase to three. Borodulina (1969) also reported three teeth on the pharyngeal plate in adult L. schmidti. Teeth on the glossohyal and palatine develop in larvae 16-18 mm long (Table 3). Glossohyal teeth disappear during transformation and are absent in juveniles and adults: hence, the common name "smoothtongue" for L. schmidti. A single palatine tooth is present during the larval stage; the num- ber of palatine teeth increases after transforma- tion (Table 3). Skull The dentary, maxillary, parasphenoid, and operculum begin to ossify in 14-15 mm larvae (Table 3). In larvae 18-21 mm long, the premaxil- lary, pre-, sub-, and interopercle bones, vomer, symplectic, branchiostegal rays, and urohyal begin ossifying in some larvae (Tables 3, 4). These structures, however, are not consistently calci- fied until larvae reach about 28 mm long (32 mm for the urohyal). Certain bones of the olfactory, orbital, otic, and oromandibular regions and much of the hyoid arch begin to ossify in some larvae 24-26 mm 30 DUNN: DEVELOPMENT AND DISTRIBUTION OF LEUROGLOSSUS SCHMIDTI Table 3.— Development of meristic and other structures in Leuroglossus schmidti larvae and juveniles. Mean data are given for specimens in the specified length range. Specimens between dashed lines are undergoing notochord flexion; those between solid lines are juveniles. SL (mm) 5.9-7.9 8.0-8.9 9.0-9.9 10.0-10.9 11.0-11 9 120-129 Sam- ple size Dorsal fin rays Anal fin rays Pectoral fin rays Pelvic fin rays Principal caudal rays Secondary caudal rays Upper Lower Upper Lower Neural spines Haemal spines Centra Abdom. Caudal Total 12 6 5 1 1 2 13.0-13.9 14.0-14.9 15.0-15.9 16.0-16.9 17.0-17.9 18.0-18.9 19.0-19.9 20.0-20.9 21.0-21.9 22.0-229 23.0-23.9 24.0-24.9 250-25.9 26 0-26.9 27.0-27.9 28.0-289 29 0-299 30 0-309 31.0-31.9 32.0-32.9 33.0-33.9 34 0-349 31.0-31 9 33.0-33.9 35.0-35.9 37.0-37.9 44.0-449 51.0-51 9 SL (mm) 5.9-7.9 8.0-8.9 9.0-9.9 10.0-10.9 11.0-11 9 12.0-12.9 22 2.0 4.3 4.5 0.3 0.3 10.0 9.0 20 2.0 20 1.8 0.4 0.4 4.0 3.6 1.0 1.2 0.4 5.0 4.5 1.0 1.3 0.3 80 7.2 1.2 18 02 6.7 60 0.3 03 67 6.0 0.7 1.0 07 7.5 6.8 3.0 3.0 0.8 100 9.0 3.0 3.3 0.5 08 100 9.0 3.3 48 08 2.0 10.0 90 3.0 4.0 0.5 10.0 9.0 2.0 1.0 10.0 9.0 5.0 5.0 2.0 2.0 10.0 9.0 4.0 5.0 1.0 0.3 03 05 1 1.0 10.0 10.0 10.0 10.0 100 100 13.0 13.0 12.5 11.0 11.0 12.0 9.0 7.0 7.5 90 6.0 9.0 8.0 9.0 9.0 9.0 9.0 9.0 100 10.0 10.0 10.0 10.0 10.0 9.0 9.0 9.0 9.0 9.0 9.0 13.0 13.0 14.0 15.0 15.0 15.0 14.0 12.0 13.5 15.0 13.0 15.0 490 51.0 495 490 49.0 50.0 230 24.0 23.5 240 23.0 24.0 27.0 28.0 27.0 27.0 27.0 27.0 230 24.0 23.5 24.0 23.0 24.0 Sample size Hypurals Epurals Uroneurals Branchios- tegal rays Gill rakers Teeth 0.7 04 2.0 0.7 0.4 1.0 0.3 '12.8 0.5 1.0 1.0 50.0 52.0 50.5 51.0 50.0 51.0 Upper Lower Total Dentary Glossohyal Vomer Palatine 13.0-13.9 14.0-14.9 15.0-159 16.0-16.9 17.0-17.9 0.3 2.0 3.0 2.0 5.0 3.7 2.0 1.7 1.0 1.0 0.7 3.0 1.3 0.3 180-189 19.0-19.9 20.0-20.9 21.0-21.9 220-22.9 23.0-23.9 24.0-24.9 25.0-259 26.0-269 27.0-27.9 280-28.9 29 0-29.9 30 0-30.9 31.0-31.9 32.0-329 33.0-33.9 34.0-349 31.0-31.9 33.0-33.9 35.0-35.9 37.0-37.9 44.0-449 51.0-51 9 1.8 7.0 1.4 2.8 3.0 1.2 2.3 5.3 4.8 5.0 6.0 7.0 7.0 0.5 2.0 04 1.0 1.3 1.6 1.3 1.3 1.5 20 2.0 20 20 2.0 2.0 0.7 0.7 1.0 20 0.4 1.2 2.0 12 1.7 1.3 2.0 2.0 2.0 2.0 2.0 2.0 2.0 1.0 2.8 38 4.0 4.3 10.0 10.0 9.0 10.4 10.8 11.8 12.0 11.0 13.0 11.8 14.0 13.0 14.0 12.0 11.0 1.3 1.3 2.5 3.0 3.0 2.8 23 24 2.3 3.0 2.8 28 3.0 2.5 3.0 3.0 3.0 2.5 2.0 3.0 30 3.0 3.0 3.0 3.0 30 3.0 30 3.0 30 3.0 3.0 3.0 3.0 7.0 7.0 7.0 7.0 7.0 7.0 0.2 7 08 2.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 2.0 2.0 7.0 150 22.0 240 20 6.0 3.0 2.0 20 7.0 14.0 21.0 28.0 9.0 4.0 2.0 2.0 7.0 14.5 21.5 27.5 6.5 4.0 1.0 2.0 2.0 7.0 14.0 21.0 34.0 8.0 7.0 2.0 2.0 7.0 17.0 24.0 32.0 10.0 4.0 1.0 2.0 2.0 9.0 17.0 260 46.0 9.0 5.0 'Haemal spines not fully differentiated on one specimen with 50 centra ossifying. It was therefore not possible to determine the number of precaudal vertebrae in that specimen 31 FISHERY BULLETIN: VOL. 81, NO. 1 c — a> en c o> .22 c c 09 E £ o o c c C3 e« o _o "5 *5 c <£ • ~ en OB ° c .c — u TJ IE Ji 5 C/j *j « n! <£ >>~ .Q 0) O) w en .5 =3 .2 .S *e .* 22 «t- ._ •— ««-. 02 ■- en en e g !s >> .■tt *3 J to H c _: o ■C o -« c ^ O) 00 E = _o; &« V- to s « o -a en C a> •- ~ cv c c """* — T3 O C m a! cd QJ .. > ro f*> O cr> n pj q tn pi p] □ 09 o o PI p] c * E ■S 8 c a. O en '-£ : aS S o a! 5 - 0) 3 £ o" * CO .5 ^.2 > » gjs < c H .22 3 K X c> I- c r c- c- c* c- Olc o s! E n J m id- gCL(/)CLC0C0Q.a. O u. D y co a. 0- UJ UJ = £ £ ci a 5 fl « " s « Q- K « 3 a O lu m co .|Er' i P K i y E 0> « c m E of J £ £ a O 5q.lu05<20 mm. In a 23.1 mm speci- men (Fig. 4e) all seven hypural bones, 10+9 prin- cipal rays, and 2+2 secondary rays are ossifying. Both pairs of uroneurals are ossified. A single unossified epural and a number of unossified neural and haemal spines can be observed. The 34 HS HY, HY 2 HY3 HY 4 HY 2 HY 3 HY 2 H Y 3 Figure 4.— Development of the caudal fin of Leuroglossus schmidti: A, 6.4 mm SL; B, 11.5 mm SL; C, 15.3 mm SL; D, 18.4 mm SL; E, 23.1 mm SL; F, 30.6 mm SL; G, 35.1 mm SL; H, 51.6 mm SL. Ossified elements are stippled. NC = noto- chord; HY = hypurals; EP = epural; HS = haemal spine; UN = uroneural; NS = neural spine; U = ural centra; PU = preural centra; SNP = specialized neural process; PCR = principal caudal rays; SCR = secondary caudal rays. DUNN: DEVELOPMENT AND DISTRIBUTION OF LEUROGLOSSUS SCHMIDTI £ NS UN, EP U 7 N 2, HS HY, NS Ui U , Ni U 2 EP UN 2 UN, EP SCR- UN1 EP UN 2 hy 7 SCR 35 FISHERY BULLETIN: VOL. 81, NO. 1 adult complement of 10+9 principal caudal rays is consistently ossified in 29-30 mm larvae, but secondary caudal rays are not completely de- veloped until after transformation. In a 30.6 mm specimen (Fig. 4f), ural centra 1 and 2 are ossify- ing as are two neural spines associated with preural centra 1 and 2 and haemal spines associ- ated with preural centra 1-4. The bases of pre- ural centra 1-8 are beginning to ossify. In a 35.1 mm transformed juvenile, all centra of the caudal complex, preural and ural, are ossi- fied as is the specialized neural process (Hollister 1936) dorsad to Ui (Fig. 4g). Ural centra 1 and 2 are still separate. Neural and haemal spines and principal and secondary caudal rays are fully ossified. The hypural bones are separate and autogenous in this specimen. A 51.6 mm juvenile has ural centra 1 and 2 fused into a single uro- style (Fig. 4h). The single epural is beginning to ossify and hypural bones 1-3 are fused together at their bases and ankylosed to the urostyle. Hypural bones 4-7 are still autogenous in this specimen. In this specimen, the 10 superior prin- cipal caudal rays are distributed as follows: hypural 7, one ray; hypural 6, two rays; hypural 5, four rays; hypural 4, three rays. The inferior nine principal rays included one ray on hypural 3, five rays on hypural 2, two rays on hypural 1, and one ray on the posteriormost haemal spine. During the development of the caudal fin, the ventralmost principal caudal ray is associated with hypural 1; it apparently is displaced to ar- ticulate with the ultimate haemal spine in some, but not all, juvenile specimens. Dorsal and anal fin rays ossify rapidly during transformation as do pectoral and pelvic fin rays (Table 3). It was not possible to follow the se- quence of ossification of individual rays in these fins. The adult complement of dorsal, anal, and pelvic fin rays is present on the smallest trans- formed specimen. The adult complement of eight to nine pectoral rays, however, is not present in all transformed specimens examined. Scales are not present in a 55.0 mm juvenile, but they are reported to be deciduous (Hart 1973), and hence, may have been lost during capture. COMMENTS ON THE VALIDITY OF THE GENUS LEUROGLOSSUS GILBERT AND THE SPECIFIC STATUS AND NAME OF L. SCHMIDTI The results of this study offer support for the validity of the genus Leuroglossus Gilbert which has been questioned. Gilbert (1890) erected the genus Leuroglossus in the family Argentinidae for specimens from the Gulf of California that he described as Leuroglossus stilbius. Chapman (1943) reviewed Leuroglossus Gilbert, removed it from Argentinidae, and placed it in Bathylagi- dae. Cohen (1964) synonymized Leuroglossus Gil- bert with Bathylagus Giinther. Borodulina (1968) stated Leuroglossus lacked an orbitosphenoid bone, although Cohen (1964) said it was present in Leuroglossus and used its presence as a gener- ic character for Bathylagus. Borodulina (1969) later described the osteology of L. schmidti and compared it with B. paeificus (based on Chap- man 1943). She stated that Leuroglossus, in con- trast to Bathylagus, lacked an orbitosphenoid, possessed teeth on the palatine, had three den- ticles on the last pharyngobranchial, and pos- sessed antorbitals. Ahlstrom (1969) described differences in the movements of oil globules be- tween Bathylagus and Leuroglossus eggs which, with the lack of an orbitosphenoid in Leuroglos- sus (as reported by Borodulina 1968), he felt, lent additional support to the validity of Leuroglossus as a genus distinct from Bathylagus. My samples of Leuroglossus lacked an orbito- sphenoid, whereas those cleared and stained spec- imens I examined of Bathylagus did possess an orbitosphenoid. Specimens of B. paeificus, B. ochotensis, and B. milleri I examined lacked teeth on the pharyngobranchials, whereas Leuro- glossus possesssed three teeth on the fourth pharyngobranchial. Based in part on the lack of an orbitosphenoid in Leuroglossus, the presence of one in Bathyla- gus and the differences in the movements of oil globules in the eggs of these two genera, as re- ported by Ahlstrom (1969), I follow Borodulina (1969) and Ahlstrom (1969) in considering the two genera distinct. The number of valid genera in the Bathylagidae and analysis of their rela- tionships, however, await further study of the entire family. This study provides additional evidence to rec- ognize L. schmidti as a species distinct from L. stilbius. Rass (1955) described a northern sub- species, L. stilbius schmidti, from the Kurile- Kamchatka Trench, based on morphometric measurements which differed from measure- ments described by Gilbert (1890). Cohen (1956) synonymized L. s. schmidti with L. stilbius, as- serting that the proportions used by Rass to de- scribe L. s. schmidti were size dependent. Boro- dulina (1968) pointed out that L. stilbius had 36 DUNN: DEVELOPMENT AND DISTRIBUTION OF LEUROGLOSSUS SCHMIDTI 39-42 vertebrae, whereas L. schmidti possessed 49-51 vertebrae. Borodulina (1968) considered L. schmidti to be a subspecies of L. stilbius, although she suggested that L. schmidti might subse- quently be recognized as a separate species, a suggestion that she did not pursue because of insufficient material. She also considered L. uro- tranus (Bussing 1965), described from the Peru- Chile Trench, to be another subspecies of L. stil- bius. Ahlstrom (1968) noted the differences in vertebral counts in L. stilbius and L. schmidti. Subsequently Ahlstrom (1969) pointed out the differences in egg size, pattern in migration of oil globules during embryonic development, larval pigment, and body proportions between L. stil- bius and L. schmidti which, along with differ- ences in vertebral counts, he felt, enabled recog- nition of L. schmidti as a distinct species. Peden (1981) examined vertebral numbers in Leuroglossus from samples collected from Mexi- co to the Aleutian Islands and westward to Japan. He noted that samples of Leuroglossus from Brit- ish Columbia waters had an average of 8.5 more vertebrae than those samples collected off Ore- gon. He therefore recognized L. schmidti as distinct from L. stilbius stilbius. As presently known, the geographical ranges of the two spe- cies do not overlap, as discussed below. Based on the differences in vertebral counts in the two nominal species reported by Borodulina (1968, 1969) and Peden (1981), the evidence presented by Ahlstrom (1968, 1969), and the results of this study, I consider L. schmidti specifically distinct from L. stilbius. The valid name of the northern smoothtongue is considered here to be L. schmidti, rather than Therobromus callorhini or Leuroglossus callo- rhini. Therobromus callorhini was described by Lucas (in Jordan and Gilbert 1899) from bones extracted from fur seal stomachs collected in the Bering Sea. He noted that the specimens had 26 precaudal and 23 caudal vertebrae and placed the species in Osmeridae. Chapman (1941) showed that T. callorhini was notanosmerid and later (Chapman 1943) he suggested that T. callorhini (emended to callorhinus) was most likely identical with either Bathylagus pacificus or B. alascanus (= B. milleri). Cohen (1964) syn- onymized Therobromus Lucas with Bathylagus Gunther. Ahlstrom (1968, 1969) suggested that the correct name of L. schmidti was Leuroglossus callorhini, but did not formally propose such a synonomy. If the two names do refer to the same species, then L. schmidti is a junior synonym of T. cal- lorhini. The type material of Therobromus cal- lorhini Lucas apparently no longer exists (ac- cording to D. M. Cohen 7 ). However, as the name T. callorhini has apparently not been used as a senior synonym in more than 50 yr (Chapman 1943), T. callorhini constitutes a nomen oblitum according to the International Code of Zoological Nomenclature (Stoll et al. 1964). Hence, I con- sider the valid name of the northern smooth- tongue to be Leuroglossus schmidti. OCCURRENCE OF EGGS AND LARVAE OF LEUROGLOSSUS SCHMIDTI (Figure 5) Eggs and larvae of L. schmidti are broadly distributed in near-coastal waters from about southern Vancouver Island, British Columbia, to the central Bering Sea. In midocean they are apparently distributed as far south as lat. 46°N, since eggs and larvae of L. schmidti were col- lected in 1951 and 1955 from about lat. 46°-57°N and long. 149°-179°W (Ahlstrom 1969; Moser 8 ). They have apparently not been found in coastal waters off Oregon, however, as the relatively few specimens of Leuroglossus larvae, collected off Oregon by Oregon State University, consisted exclusively of L. stilbius stilbius (Richardson 1973; Washington 9 ). The results of an ichthyo- plankton survey conducted in October-November 1971 from off Washington (lat. 46°45'N) to Dixon Entrance, British Columbia (lat. 54°30'N), were reported by Naplin et al. (footnote 4). Eggs of L. schmidti were found only north of lat. 53°N off Queen Charlotte Islands, whereas only L. stilbius stilbius eggs were collected south of lat. 51°N off Vancouver Island and coastal Washington. The few Leuroglossus larvae collected during this cruise were all L. schmidti, and they were taken only north of lat. 54°N. Possibly the eggs identi- fied as L. stilbius stilbius off Vancouver Island 7 D. M. Cohen, National Systematics Laboratory. National Marine Fisheries Service, National Museum of Natural His- tory Washington, D.C. (present address: NWAFC, NMFS, NOAA, 2725 Montlake Blvd. East, Seattle, WA 98112), pers. commun. July 1980. 8 H. G. Moser, Southwest Fisheries Center La Jolla Labora- tory, National Marine Fisheries Service, NOAA, La Jolla, CA 92038, pers. commun. March 1980. 9 B. B. Washington, School of Oceanography, Oregon State University, Corvallis, Oreg. (present address: Gulf Coast Re- search Laboratory, Ocean Springs, MI 39564), pers. commun. November 1980. 37 FISHERY BULLETIN: VOL. 81. NO. 1 65 00N ^\£ Naplin et i Larvae Naplin el al Eggs Matson and Wing 11978) Larvae □ Kendall et al Larvae O Kendall et al Eggs Waldron 119811 Larvae 1 I Moser Eggs ^ Moser Larvae # Ahlslrom 119691 Eggs 60 00N - 55 00N 50 00N ▲ • 45 OON 175 00E 175 OOW 165 00W 155 00W 1 45 OOW 135 OOW 125 OOW FIGURE 5. — General areas where eggs and larvae of Leuroglossus schmidti have been reported. Key: Naplin et al. (text footnote 4); Matson and Wing (1978); Kendall et al. (text footnote 3); Waldron [1981]; Moser (text footnote 8); Ahlstrom (1969). couid have been transported northward from more southerly spawning areas. During this cruise, surface geostrophic currents indicated a 6 cm/s northward flow offshore from about lat. 47°N to 51 °N (Naplin et al. footnote 4). Leuroglossus schmidti larvae were the third most abundant fish larvae collected in plankton samples from coastal waters of southeastern Alaska (lat. 56°50'-59°28'N, long. 133°10'-135° 23'W) in April-November 1972 (Mattson and Wing 1978). This species accounted for 4.5% of the total catch of fish larvae; abundance was high from May to August, peaking in June and July. Plankton sampling in Kodiak Island shelf waters from November 1977 through March 1979 re- vealed that eggs of L. schmidti were found prin- cipally at the shelf break (water depth >200 m); abundance was greatest in the fall, but eggs were found in small numbers in summer and winter (Kendall et al. footnote 3). Larvae were also most abundant over the shelf break in the fall, but sea- sonal abundance was not determined. Waldron [1981] summarized available distribution data on larvae and juveniles of L. schmidti occurring in the eastern Bering Sea from 1955 to 1978. Based on plankton sampling conducted by the United States, U.S.S.R., and Japan (primarily during summer) utilizing a variety of sampling devices, larvae identified as L. schmidti were most frequently reported over the shelf break. ACKNOWLEDGMENTS A number of people at N WAFC assisted in this study. Arthur W. Kendall, Jr., engaged in help- ful discussions with the author, constructively reviewed the manuscript, and assisted in numer- ous other ways. Beverly Vinter illustrated the larvae and caudal fin development and shared her knowledge of larval bathylagids. Ann C. Matarese reviewed the manuscript and made many useful comments. Bernie Goiney and Jay Clark rendered technical assistance. Eleanor S. Uhlinger efficiently and rapidly processed sev- eral drafts of the manuscript. Bruce W. Wing, NWAFC, Auke Bay, Alaska, loaned larvae of L. schmidti. H. Geoffrey Moser, SWFC, La Jolla, Calif., kindly made available unpublished data on the occurrence of L. schmidti and provided unpublished illustrations of this species; he also made available his collection of radiographs and cleared and stained bathylagid specimens. T. W. Pietsch, University of Wash- ington (UW), Seattle, loaned bathylagids. Alex 38 DUNN: DEVELOPMENT AND DISTRIBUTION OF LEUROGLOSSUS SCHMIDTI Peden, British Columbia Provincial Museum, Victoria, kindly made available unpublished data on vertebral counts of Leuroglossus spp. Kevin M. Howe, UW, shared his knowledge of fishes, radiographed specimens, made literature available, and reviewed the manuscript. Daniel M. Cohen, National Systematics Laboratory, NMFS, Washington, D.C. (present address: NWAFC, Seattle), imparted generously his knowledge of bathylagids, provided unpublished data on L. schmidti, and made helpful comments on the manuscript. LITERATURE CITED Ahlstrom, E. H. 1965. Kinds and abundance of fishes in the California Current region based on egg and larval surveys. Calif. Coop. Oceanic Fish. Invest. Rep. 10:31-52. 1968. What might be gained from an oceanwide survey of fish eggs and larvae in various seasons. Calif. Coop. Oceanic Fish. Invest. Rep. 12:64-67. 1969. Remarkable movements of oil globules in eggs of bathylagid smelts during embryonic development. J. Mar. Biol. Assoc. India 11:206-217. 1972. Distributional atlas of fish larvae in the California Current Region: six common mesopelagic fishes — Vinci- guerria lucetia, Triphoturus mexicanus, Stenobrachius leucopsarus, Leuroglossus stilbius, Bathylagus wesethi, and Bathylagus ochotensis, 1955 through 1960. Calif. Coop. Oceanic Fish. Invest. Atlas 17, 306 p. Ahlstrom, E. H., and H. G. Moser. 1976. Eggs and larvae of fishes and their role in system- atic investigations and in fisheries. Rev. Trav. Inst. Peches Marit. 40:379-398. BORODULINA, O. D. 1968. Taxonomy and distribution of the genus Leuroglos- sus (Bathylagidae, Pisces). Probl. Ichthyol. 8:1-10. 1969. Osteology of Leuroglossus stilbius schmidti Rass (Bathylagidae). Probl. Ichthyol. 9:309-320. Bussing, W. A. 1965. Studies of the midwater fishes of the Peru-Chile Trench. In G. A. Llano (editor), Biology of the Antarctic Seas II, p. 185-227. Antarct. Res. Ser. 5. Natl. Acad. Sci. Nat. Res. Counc. Publ. 1297. Chapman, W. M. 1941. The osteology and relationships of the osmerid fishes. J. Morphol. 69:279-301. 1943. The osteology and relationships of the bathypelagic fishes of the genus Bathylagus Giinther with notes on the systematic position of Leuroglossus stilbius Gilbert and Therobromus callorhinus Lucas. J. Wash. Acad. Sci. 33:147-160. Cohen, D. M. 1956. The synonymy and distribution of Leuroglossus stilbius Gilbert, a North Pacific bathypelagic fish. Stanford Ichthyol. Bull. 7:19-23. 1964. Suborder Argentinoidea. In Fishes of the West- ern North Atlantic, Part 4, p. 1-70. Mem. Sears Found. Mar. Res., Yale Univ. Dingerkus, G., and L. D. Uhler. 1977. Enzyme clearing of alcian blue stained whole small vertebrates for demonstration of cartilage. Stain Technol. 52:229-232. Dunn, J. R. 1983. The utility of developmental osteology in taxonomic and systematic studies of teleost larvae: a review. U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circ. 450. 19 p. Gilbert, C. H. 1890. A preliminary report on the fishes collected by the steamer Albatross on the Pacific coast of North America during the year 1889, with descriptions of twelve new genera and ninety-two new species. Proc. U.S. Natl. Mus. 13:49-126. Hart, J. L. 1973. Pacific fishes of Canada. Fish. Res. Board Can. Bull. 180, 740 p. Hollister, G. 1936. Caudal skeleton of Bermuda shallow water fishes. I. Order Isospondyli: Elopidae, Megalopidae, Albulidae, Clupeidae. Dussumieriidae, Engraulidae. Zoologica (N.Y.) 21:257-291. Jordan, D. S., and C. H. Gilbert. 1899. The fishes of Bering Sea. In D. S. Jordan, The fur seals and fur-seal islands of the North Pacific Ocean, Part 3:433-492. Gov. Print. Off., Wash.. D.C. Mattson, C. R., and B. L. Wing. 1978. Ichthyoplankton composition and plankton vol- umes from inland coastal waters of southeastern Alaska, April-November, 1972. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-723, 11 p. Monod. T. 1968. Le complexe urophore des poissons teleosteens. Mem. Inst. Fond. Afr. Noire 81, 705 p. Moser, H. G. [1981]. Morphological and functional aspects of marine fish larvae. In R. Lasker (editor), Marine fish larvae: Morphology, ecology, and relation to fisheries, p. 89-131. Univ. Wash. Press, Seattle. Norden, C. R. 1961. Comparative osteology of representative salmonid fishes, with particular reference to the grayling (Thy- mallus arcticus) and its phylogeny. J. Fish. Res. Board Can. 18:679-791. Peden, A. E. 1981. Recognition of Leuroglossus schmidti and L. stil- bius (Bathylagidae, Pisces) as distinct species in the North Pacific Ocean. Can. J. Zool. 59:2396-2398. Rass, T. S. 1955. Glovokovodnye ryby Kurilo-Kamchatskoi vradin. (Deep-sea fishes of the Kurile-Kamchatka Trench). [In Russ.] Akad. Nauk SSSR, Tr. Inst. Okeanol. 12: 328-339. (Transl. by Nat. Hist. Mus., Stanford Univ., Calif.). Richardson, S. L 1973. Abundance and distribution of larval fishes in wa- ters off Oregon, May-October 1969, with special empha- sis on the northern anchovy, Engraulis mordax. Fish. Bull., U.S. 71:697-711. Russell, F. S. 1976. The eggs and planktonic stages of British marine fishes. Acad. Press, Inc., Lond., 524 p. Sanzo, L. 1931-1933. Salmonoide. Fauna Flora Golfo Napoli, Mongr. 38(4 parts), 1064 p. Schmidt, J. 1906. On the larval and post-larval development of the Argentines {Argentina silus [Ascan.] and Argentina 39 FISHERY BULLETIN: VOL. 81, NO. 1 sphyraena Linne) with some notes on Mallotus villosus adopted by the XV international congress of zoology. [0. F. Muller]. Medd. Komm. Havunders. (Fisk.) 2(4): Int. Trust Zool. Nomencl., Lond., 176 p. 1-20. Waldron, K. D. 1918. Argentinidae, Microstomidae, Opisthoproctidae. [1981.] Ichthyoplankton. In D. W. Hoodand J. A. Calder Mediterranean Odontostomidae. Rep. Dan. Oceanogr. (editors), The eastern Bering Sea Shelf: oceanography Exped. Mediterr. 2(A5):l-40. and resources. Vol. I. p. 471-493. U.S. Dep. Com- Stoll, N. R., R. P. Dollfus, J. Forest, N. D. Riley, C. W. mer.. Natl. Oceanic Atmos. Admin., Off. Mar. Pollut. Sabrosky, C. W. Wright, and R. V. Melville (editors). Assess. 1964. International code of zoological nomenclature 40 DELINEATION OF TILEFISH, LOPHOLATILUS CHAMAELEONTICEPS, STOCKS ALONG THE UNITED STATES EAST COAST AND IN THE GULF OF MEXICO S. J. Katz, 1 C. B. Grimes, 2 and K. W. Able 3 ABSTRACT Tilefish, Lopkolatilus chamaeleonticeps, are an important commercial species in the Mid-Atlantic Bight and the focus of developing fisheries in the South Atlantic Bight and the Gulf of Mexico. Attempts were made to delineate stocks over this range by analyzing for variation in morphology (28 meristic and morphometric characters) and electrophoretic migration of eye, liver, and muscle proteins. Morphological and electrophoretic data (liver isocitrate dehydrogenase and liver esterase) consistently supported a separate Mid-Atlantic Bight stock. Electrophoretic data suggested that South Atlantic Bight and Gulf of Mexico samples belonged to a separate, single stock. This was not consistently supported by the more variable morphometric characters. It was suggested that Mid- Atlantic Bight populations be treated as a separate stock and, as a working hypothesis, that South Atlantic and Gulf of Mexico populations be considered as a second stock. Tilefish, Lopholatilus chamaeleonticeps, are dis- tributed from southern Nova Scotia (Leim 1960; Markle et al. 1980) south to off Surinam, South America, (Wolf and Rathjen 1974) and through- out the Gulf of Mexico (Bigelow and Schroeder 1947; Hoese and Moore 1977) but exclusive of the Caribbean Sea (Dooley 1978). The tilefish is the basis for a valuable bottom longline fishery in the Mid-Atlantic Bight (Grimes et al. 1980), and this fishery is developing elsewhere along the east coast of the United States and in the Gulf of Mexico. This paper investigates tilefish popula- tions to determine if separate stocks can be iden- tified over this range. There are several reasons to suspect that dis- tinct stocks of tilefish may occur. Tilefish prob- ably have a restricted habitat. They are reported from rather narrow temperature ranges (9°- 14°C) at the edge of the continental shelf along the east coast (Goode 1884; Rathburn 1895; Bige- low and Schroeder 1953) and in the Gulf of Mexi- co (Nelson and Carpenter 1968; Wolf and Rath- jen 1974). Also, preliminary tagging studies (Grimes et al. in press) suggested that individual 'Ecology Graduate Program, Rutgers University, New Brunswick, NJ 08903. 2 Forestry and Wildlife Section and New Jersey Agricultural Experiment Station, Rutgers University, New Brunswick, NJ 08903. 3 Biological Sciences and Center for Coastal and Environ- mental Studies, Rutgers University, New Brunswick, NJ 08903. tilefish moved <2 km in over 1 yr. These obser- vations are supported by submersible observa- tions which suggest that tilefish are resident in temporally stable burrows of their own con- struction (Able et al. 1982). In the Mid-Atlan- tic Bight, tilefish are caught the year-round which also suggests that these may be resident populations. In addition, the prevailing current patterns, temperature regimes, and species dis- tribution patterns along the east coast suggest that important faunal boundaries may exist at Cape Hatteras and around the Florida peninsula (see Briggs 1974 for discussion). This study re- ports on morphological and electrophoretic char- acteristics of tilefish from the U.S. east coast and the Gulf of Mexico. The distribution of the char- acters were used to test the null hypothesis that there are no differences among these popula- tions. MATERIALS AND METHODS Tilefish samples were obtained from commer- cial fishermen or collected by hook and line on exploratory fishing cruises (National Marine Fisheries Service RV Oregon ID during 1978 and 1979 (Fig. 1) (Katz 1982). Information on physi- cal conditions at collection were unavailable, but temperature is known to be relatively constant throughout the range (see above). Fish were transported fresh, on ice, or frozen, depending on distance of collection from the laboratory. Manuscript accepted July 1982. FISHERY BULLETIN: VOL. 81. NO. 1. 1983. 41 FISHERY BULLETIN: VOL. 81. NO. 1 Figure 1.— Sample locations for tile- fish along the U.S. east coast and the Gulf of Mexico. Submarine canyons are identified in the inset. Electrophoresis Eye, liver, and muscle tissues were removed from individual fish and frozen as soon as pos- sible. Vertical starch gel electrophoresis was used to detect protein variation. Initially only tis- sues of fish from the most distant collection lo- calities (Hudson Canyon and off Texas) were screened for 28 enzymes to maximize the chance of finding polymorphic enzymes. Of the 28 en- zymes screened during the initial electrophore- sis, several were scorable; however, most ap- peared monomorphic (malate dehydrogenase, lactate dehydrogenase, xanthine dehydrogenase, creatin kinase, adenylate kinase, peptidase, alco- hol dehydrogenase, malic enzyme, 6-phosphoglu- conate dehydrogenase, and glyceraldehyde 3- phosphate dehydrogenase) and only two [liver isocitrate dehydrogenase (IDH) and liver ester- ase (EST)] were polymorphic. Liver tissues from all collections were then run for both IDH and EST with an amine citrate buffer (pH 6.0) (Clay- ton and Tretiak 1972) for 17 h at 140 V and 40°C, and allelic frequencies were determined for all populations. Allelic frequencies were compared with their Hardy-Weinberg expectations by a chi-square test (Spiess 1977). We evaluated dif- ferences between sample locations, by chi-square contingency tests of electromorph distribution between sample locations. This test does not as- sume Hardy-Weinberg equilibrium and com- pares n samples with k classes to determine whether the individual A" classes are in the same relative proportion throughout the n samples. Length (age)-related differences in genotype distribution were tested (chi-square) on the largest sample with a wide range of sizes ( n = 40, west side of Hudson Canyon). Fish were divided into two size classes (<550 mm fork length and >550 mm) based on the approximate size at sex- ual maturity. Morphology Seven meristic (number of dorsal fin spines and rays, anal fin spines and rays, pectoral fin rays, upper and lower gill rakers on the first arch) and 21 morphometric (fork, standard, total, pectoral fin, pelvic fin, upper jaw, snout, adipose flap, barbel, snout to vent, snout to anal origin, snout to dorsal origin, snout to incurrent nostril, lengths; orbit diameter, interorbital width, head width, height of first, second, and third dorsal fin spines, caudal peduncle depth, and suborbital depth) characters were counted or measured following Hubbs and Lagler (1967), 42 KATZ ET AL.: DELINEATION OF TILEFISH STOCKS with two exceptions: Barbel length was mea- sured from its posterior tip to the junction with the lower lip, and the suborbital depth was mea- sured from the lower margin of the infraorbitals to the junction of the articular and interopercu- lar bones. Morphometric characters were mea- sured to the nearest millimeter with dividers and a tape measure. These characters were chosen on the basis of a preliminary study of two specimens of tilefish by Bigelow and Schroeder (1947) and a systematic study of the Branchiostegidae by Dooley (1978). Morphological data was determined from fish of dissimilar lengths (Fig. 2), so we used analysis of covariance to remove the size effects as sug- gested by Atchley et al. (1976). A linear relation- ship to standard length (SL) was determined for most morphological characters with the excep- tion of adipose flap length where an additional coefficient of standard length squared was in- cluded in the model because of allometry. For the final size-corrected comparisons between sam- ple locations we used sample location least square means for each morphological character ( Barr et Gult of Mexico- Campeche Banks MALES Jl FEMALE ,ll s N 10 -40 -20 Gulf of Mexico- off Texas J. L 15 -20 Gulf of Mexico- off West Florida .dk k 9 -40 -20 South Atlantic Bight- off South Carolina .iL ,h 10 -40 ^20 Mid-Atlantic Bight- west side of Hudson Canyon J 33 1-6O -40 -20 Mid-Atlantic Bight- i- east of Hudson Canyon ^_| i 26 hLL 12 -20 Mid-Atlantic Bight- east of Block Canyon L 4, ii 54 -40 -20 Mid-Atlantic Bight - west of Veatch Canyon IX. " li. 24 -40 -20 — n J- _L 200 400 600 800 1000 200 400 600 800 1000 Standard Length (mm) Figure 2.— Length-frequency histograms of tilefish samples used to conduct the morphological analysis. See Figure 1 for approximate locations. al. 1976). Least square means are estimates of arithmetic means that would be predicted had samples with the same size composition been ob- tainable from each sampling location. We conducted analysis of covariance on each morphological character to test for differences between sampling locations. Sex, sample loca- tion, and all interactions were initially included in the covariance model, but all nonsignificant (P<0.01) interactions were removed from the final model. The difference between sample loca- tion least square means for each morphological character for each sex was tested by comparison with the west Hudson Canyon sample using a t test. Significant differences were determined conservatively, using a high significance level (P<0.001), because the possibility of finding dif- ferences increases with the number of tests run. To further test for differences between sample locations we used discriminant function analysis (Jolicoeur 1959; Seal 1964) to determine the level of distinctness of fish from each location. The discriminant function was computed using both raw and size-corrected data for males and fe- males separately, because the analysis of covari- ance indicated sexual dimorphism. Only linearly related morphological characters were used in the raw discriminant function (Seal 1964). Size correction of morphological characters was ac- complished using the average value of standard length (SL) of all samples, and linear and quad- ratic regression coefficients (Bi, B 2 ) obtained from covariance analysis for each morphological character according to the following formula: corrected = raw - B x (SL-SL) - B 2 (SL-SL) 2 . This correction removed size effects by displac- ing each morphological observation towards the average, while allowing sample location and in- teraction effects to remain. RESULTS Electrophoretic Data The genetic basis of protein variation in tile- fish was implied from the electrophoretic band- ing patterns. IDH showed a dimeric pattern (heterozygote was three banded) with medium, slow, and fast bands. The rare fast form occurred only as a heterozygote in 10 out of 226 fish in the Mid- Atlantic Bight samples; therefore it has been left out of the statistical analysis. The EST 43 FISHERY BULLETIN: VOL. 81, NO. 1 locus exhibited a monomeric pattern (heterozy- gote was two banded) with fast and slow bands. Our interpretation of the dimeric and monomeric nature of these enzymes is consistent with past studies of their molecular structure (Manwell and Baker 1970). Distribution of EST and IDH electromorphs was not significantly different than expected from Hardy-Weinberg equilibri- um (Tables 1 , 2), which is additional support for a single-locus, two-allele genetic model. However, it should be noted that the chi-square test is not very sensitive at small sample sizes(<200)(Fair- bairn and Roff 1980). There was no significant difference in geno- type distribution as a function of length (age) for both enzymes (EST X 2 = 1.16, P>0.6, n = 20; IDH x 2 = 2.93, P>0.2, n = 20) in the sample ex- amined (west side of Hudson Canyon). There were distinct patterns of variation among the populations sampled (Fig. 3). Chi- square contingency tests revealed no significant differences in genotype distribution within the Mid- Atlantic Bight or southern sampling loca- tions (South Carolina, west Florida, Texas, and Campeche) (within Mid- Atlantic Bight EST X 2 5 , 3 = 8.77, 0.25 CL O ^ 0) ° II T3 en * D) o C - _i O — i j ■■3 P 6 e« o 7 §y CL) T, a - a V n CL a » 3 £5 e« en £ c 1.1 »- o T3 CD £ E en Q. C en _CD _C at en o. .2 g as «> £ +3 =■> ™ O <= en o. c (71 a; a «<-> £ .2 o 5 oj £ - en -r a) £ 5 C en m 3> .S c en O C >} CD CO o CL~ X c b s cfi -<-» 2 to ra .c 03 0; o c: o> ^ 3 N t* o CD CD .— CS Q_ CO • — *- is O to _ > C en C . r ai (/) TO r- c o -Com ndep 15 o o en i H- c CO en Q. W 'en E J >> CO oa <$ in < c H co 17) c I ns i o m o X r- X 0) o CD Q. ^ o E o t- 505" O CD CO 00 ID CO - CO in CM 00 CO 00 CO CO CD oS i- ,- 1- CO CD o >. c CO i O i co .*' 7: j= c H c 2 ■DCDuOlOoioCJ) ) w ,. CO i^ ^ '" — 10 -C C I c o >s c o o cm O CM co d o •xt o CO CD O CO a> LD o r~ LO o o en o ■J CO CO CD r- o CD CO b d ° I ra i. O U £, i 'OffliicfiOoiocn a>-^ 17) >< "- 1= CO cu X 3 C5 UqQJT3lli^lu urn „ q;^^o^o3o°o» „ ra £ - x - I = CD - > •5 V> ^. ^ ^ ~- c_ c^c^ cu .f5oJ5oi5C) o 45 FISHERY BULLETIN: VOL. 81. NO. 1 o O a 03 < to >> ccj o £ ■o 2 Q. C ■O <"i> mi? < _ co Q) I CD CO -D ' O 1> n Q. il t 1) i- CO *~ -So* 2 <= o) o — c CD 0) CL — CD CO CO CO t- o o ■* ^ m C\J C\J OJ CJ) CO CM 1 *r CO W b CD CO CO T CD O CO O) in O CM o og CO o ID CO *r CO CO CD r- CM r- CO CO CO CO o r^ C\J o CO CO CO CO O) o CO CO CO CD O r- o in O) o o CO t- CO t- d en t- o) CD CD o CM CO in o CO co o 1 O) in o The nature of the variation in the morphomet- ric characters examined varied between sexes and locations (Tables 3, 4). For several characters the least square mean values appeared to vary clinally. This was most evident for male adipose flap height and orbit diameter as seen in plots of raw data (Figs. 4, 5), female interorbital width and male head length. The values for other char- acters showed less consistent patterns and in some cases could be interpreted to suggest two distinct groups with the South Carolina samples most similar to Mid-Atlantic Bight groups (Tables 3, 4). This was most obvious for male pec- toral fin length and female pectoral fin length, caudal peduncle depth, and head length. Clinal variation was also suggested by the increasing number of significantly different morphological characters with increasing geographic distance between compared samples. The discriminant function analysis was con- ducted with both raw and size-corrected data. In each case the results were virtually identical with two exceptions (males, east Hudson Can- yon - 60% correct classification with size correct- ed vs. 23% raw data, and Campeche - 86% correct classification vs. 43% raw data). We believe neither of these significantly affects the overall interpretation of the results, and we report the raw data results here (Tables 6, 7). The discriminant function analysis suggests a similar clinal pattern of variation for both males and females (Tables 6, 7). There was generally low differentiation within the Mid-Atlantic Bight samples, and where misidentification oc- curred it was to other Mid-Atlantic Bight or South Carolina samples and infrequently to west Florida and the Gulf of Mexico off Texas. Gulf of Mexico samples naturally had higher percent- age correct classification (sample locations were more widely separated geographically) and in- correct classifications were usually to other Gulf of Mexico samples. Classifications for South Carolina samples had a high correct classifica- tion, and where misclassification occurred it was to both Mid-Atlantic Bight and Gulf of Mexico locations. DISCUSSION For purposes of interpreting the significance in allelic frequencies observed for IDH and EST we are assuming that the genetic variation ob- served is neutral (Allendorf and Phelps 1981; Ihssen et al. 1981). Thus, based on the patterns 46 KATZ ET AL.: DELINEATION OF TILEFISH STOCKS E E en c a. CO Q.d)CU<0 -. E o t- o * *- 03 c o>_ o O 5 u to o o y o X U- c ^ re •o CD CD Ol o CD CD o CD -g o ° o. re ^5 _ro o co--*;--;-';- From sample locations 5O5050 go O O O co < w < 5 5 Gulf of Mexico- Campeche Banks Gulf of Mexico- off Texas Gulf of Mexico— off west Florida South Atlantic Bight— off South Carolina Mid-Atlantic Bight- west of Hudson Canyon Mid-Atlantic Bight- east of Hudson Canyon Mid-Atlantic Bight- east of Block Canyon Mid-Atlantic Bight- west of Veatch Canyon N 43 14 43 _7 17 83 _6 18 12 70 V7_ 9 75 8 8 12. 9 55 18 9 9 22 4 15 23 23 19 16 26 2 5 7 5 54 27 41 12 4 4 17 50 13 24 Table 7.— Percent female tilefish classified to sample loca- tions by discriminant function analysis. To sample locations From sample locations Gulf of Mexico- Campeche Banks Gulf of Mexico- off Texas Gulf of Mexico- off west Florida South Atlantic Bight- off South Carolina Mid-Atlantic Bight- west of Hudson Canyon Mid-Atlantic Bight- east of Hudson Canyon Mid-Atlantic Bight- east of Block Canyon Mid-Atlantic Bight- west of Veatch Canyon c c c c c ~ re •c c , re . O c ~ re 1 re O c CD P t SZ c -C O r re 1 , "O rr rn rr O CD x c o> re 8b 1) — y - 1 CD O 5= CD m CO cu > >s 5 x CD 5 "> z: 3 < O CO O re ™ CO 'o b E ^_ re 30 TO TO < 0] < w J, re T3 a) < to J, re T3 0) < 0) 5 O O C3 CO > ^ :> 2> N 20 20 87 22 50 13 78 10 70 10 10 10 1°. 15 _9 10. 3 6 45 3 33 9 33 17 17 33 33 12 2 4 11 6 3 67 9 54 4 17 8 5 33 33 24 Mexico populations. In the discriminant function analysis South Carolina samples of both sexes classified correctly a high percentage of the time but misclassification occurred to both Mid-At- lantic Bight and Gulf of Mexico samples. The variability in the pattern of morphological char- acters can be accounted for by clinal variation in these characters, or, less likely, by two distinct groups that are only weakly differentiated. The interpretation of the morphological data may also be hampered by the small samples for more southern populations and the great distances be- tween them. Other life history data for tilefish in the Mid- 48 KATZ ET AL.: DELINEATION OF TILEFISH STOCKS Atlantic Bight are in accord with the concept of a separate stock. As we have previously mentioned, they are resident because they are taken year- round in the fishery (Grimes et al. 1980), appar- ently move short distances in the course of a year (Grimes et al. in press), and construct temporally stable burrows that may be occupied for the life of a fish (Able et al. 1982). In addition, they are known to reproduce in the Mid-Atlantic Bight because gonads show seasonal patterns of devel- opment and decline (Idelberger et al. 1981) and eggs and larvae have been collected (Fahay and Berrien 1981). The prevailing current patterns and hydro- graphic regimes over the study area are consis- tent with our delineation of the stocks. While there is a southwesterly drift of shelf water with- in the Mid-Atlantic Bight (Miller 1952; Bumpus 1973) that would provide mixing of eggs and lar- vae, it is unlikely that egg or larval transport occurs between the Mid-Atlantic and South At- lantic Bights. The Gulf Stream turns eastward at Cape Hatteras so that its axis is located 250 km east of the shelf break in the Mid-Atlantic Bight (Emery and Uchupi 1972). This difference in Gulf Stream effects produces distinct northern and southern continental shelf water masses (Stefansson et al. 1971; Emery and Uchupi 1972). Thus it is unlikely that egg and larval transport between these two areas would commonly occur, although Cox and Wiebe (1979) have suggested that anticyclonic eddies could provide a mech- anism for transporting oceanic larvae across the Gulf Stream to Mid-Atlantic Bight waters. Prevailing current systems in the southern United States may provide the means for larval mixing between the Gulf of Mexico and the South Atlantic Bight as suggested by the similarities in allelic frequencies for samples from these two areas. The Gulf of Mexico Loop Current (Maul 1977) provides a means for tilefish larvae to be transported out of the Gulf of Mexico and into the South Atlantic Bight as it joins the Florida Cur- rent and eventually forms the Gulf Stream. In addition to prevailing currents, periodic mass mortality may have contributed to the dif- ferences between distinct stocks. Following their discovery by a cod fisherman off southern New England in 1879, tilefish experienced a mass mortality in 1882 (a few billion fish reported floating at the surface; Bumpus 1898) probably caused by a sudden temporary intrusion of cold water (McLellan et al. 1953; Hachey 1955). This mortality may have resulted in a "founder effect" phenomenon and thus be responsible for stock differences we have noted. In summary, we believe that the available data suggest that Mid-Atlantic Bight tilefish popula- tions represent one unit stock and that South Atlantic Bight and Gulf of Mexico populations be considered another stock, at least as a working hypothesis. However, the wide geographic sepa- ration of the latter two areas may necessitate managing them as two stocks. Because the elec- trophoretic results suggest that gene flow may occur between Gulf of Mexico and South Atlantic Bight populations, this should be done with cog- nizance that Gulf of Mexico populations could serve as a source of recruits to South Atlantic Bight populations. ACKNOWLEDGMENTS This work was initiated upon the urgings and assistance of our friend, the late Lionel Walford, and could not have been accomplished without the cooperation and assistance of other persons whom we wish to gratefully acknowledge. S. Turner and M. Horvath assisted in processing samples. R. Trout provided valuable statistical council and R. Vrijenhoek availed us of his elec- trophoresis facilities and expertise. Fran and Lou Puskus and commercial tilefish fishermen in Barnegat Light, N.J., helped us obtain Mid- Atlantic Bight samples. M. Godcharles, Marine Research Laboratory, Florida Department of Natural Resources, helped us obtain samples from west Florida. Samples from off Texas, South Carolina, and the Yucatan Peninsula (Campeche Bank) were obtained from National Marine Fisheries Service RV Oregon //cruises. Financial support was provided by a New Jersey Sea Grant (R/F-2), a small grant from Rutgers University Research Council, the New Jersey Agricultural Experiment Station (Project No. 12409), and the Center for Coastal and Environ- mental Studies, Rutgers University. LITERATURE CITED Able, K. W., C. B. Grimes, R. A. Cooper, and J. R. Uzmann. 1982. Burrow construction and behavior of tilefish, Lo- pholatiluschamaeleonticeps, in Hudson Submarine Can- yon. Environ. Biol. Fish. 7:199-205. Allendorf, F. W., and S. R. Phelps. 1981. Use of allelic frequencies to describe population structure. Can. J. Fish. Aquat. Sci. 38:1507-1514. Atchley, W. R., C. T. Gaskins, and D. Anderson. 1976. Statistical properties of ratios. I. Empirical re- sults. Syst. Zool. 25:137-148. 49 FISHERY BULLETIN: VOL. 81. NO. 1 Barr, A. J., J. H. Goodnight, J. P. Sall, and J. T. Helwig. 1976. A users guide to SAS 76. SAS Inst. Inc., Raleigh, N.C., 329 p. BlGELOW, H. B., AND W. C. SCHROEDER. 1947. Record of the tilefish, Lopholatilus chamaeleonti- ceps Goode and Bean, for the Gulf of Mexico. Copeia 1947:62-63. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53:1-567. Briggs, J. C. 1974. Marine zoogeography. McGraw-Hill, N.Y., 475 p. BUMPUS, D. F. 1973. A description of the circulation of the continental shelf of the east coast of the United States. In B. War- ren (editor), Progress in oceanography, Vol. 6, 4:111-157. Pergamon Press. Bumpus, H. C. 1898. On the reappearance of the tilefish (Lopholatilus chamaeleonticeps). Bull. U.S. Fish Comm. 18:321- 333. Clayton. J. W., and D. N. Tretiak. 1972. Amine-citrate buffers for pH control in starch gel electrophoresis. J. Fish. Res. Board Can. 29:1169- 1172. Cox, J., AND P. H. Wiebe. 1979. Origins of oceanic plankton in the Middle Atlantic Bight. Estuarine Coastal Mar. Sci. 9:509-527. Dooley, J. K. 1978. Systematics and biology of the tilefishes (Perci- formes: Branchiostegidae and Malacanthidae), with de- scriptions of two new species. U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circ. 411, 78 p. Emery, K. 0., and E. Uchupi. 1972. Western North Atlantic Ocean: topography, rocks, structure, water, life and sediments. Am. Assoc. Pet. Geol., Tulsa, Okla., 523 p. Fahay, M. P., and P. Berrien. 1981. Preliminary description of larval tilefish (Lopho- latilus chamaeleonticeps). In R. Lasker and K. Sher- man (editors), The early life history of fishes: recent studies, p. 600-602. Rapp. P.-V. Reun. Cons. Int. Ex- plor. Mer 178. Fairbairn, D. J., and D. A. Roff. 1980. Testing genetic models of isozyme variability with- out breeding data: can we depend on the X 2 ? Can. J. Fish. Aquat. Sci. 37:1149-1159. Goode, G. B. 1884. The tile-fish family— Latilidae. In The fisheries and fishery industries of the United States. Section I. Natural history of useful aquatic animals, p. 360-361. U.S. Comm. Fish Fish. 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. Grimes, C. B., S. C. Turner, and K. W. Able. In press. A technique for tagging deepwater fish. Fish. Bull.. U.S. 81(3). Hachey, H. B. 1955. Water replacements and their significance to a fishery. Pap. Mar. Biol. Oceanogr., Deep-Sea Res., Suppl. to Vol. 3, p. 68-73. Hoese, H. D., and R. H. Moore. 1977. Fishes of the Gulf of Mexico: Texas, Louisiana, and adjacent waters. Texas A&M Univ. Press, College Sta- tion, 327 p. Hubbs, C. L., and K. F. Lagler. 1967. Fishes of the Great Lakes region. Univ. Mich. Press, Ann Arbor, 186 p. Idelberger, C, C. B. Grimes, and K. W. Able. 1981. The reproductive biology of tilefish. Lopholatilus chamaeleonticeps in Middle Atlantic and southern New England waters. (Abstr.) Bull. N.J. Acad. Sci. 26(2): 67. 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. Jolicoeur, P. 1959. Multivariate geographical variation in the wolf Canis lupus L. Evolution 13:283-299. KATZ, S. J. 1982. Identification of tilefish, Lopholatilus chamaeleon- ticeps, stocks along the east coast of the United States and the Gulf of Mexico. M.S. Thesis, Rutgers Univ., New Brunswick, N.J. Leim, A. H. 1960. Records of uncommon fishes from waters off the maritime provinces of Canada. J. Fish. Res. Board Can. 17:731-733. Manwell, C, and C. M. A. Baker. 1970. Molecular biology and the origin of species; hetero- sis, protein polymorphism and annual breeding. Univ. Wash. Press, Seattle, 394 p. Markle, D. F., W. B. Scott, and A. C. Kohler. 1980. New and rare records of Canadian fishes and the influence of hydrography on resident and nonresident Scotian Shelf ichthyofauna. Can. J. Fish. Aquat. Sci. 37:49-65. Maul, G. A. 1977. The annual cycle of the Gulf Loop Current Part I: Observations during a one-year time series. J. Mar. Res. 35:29-47. McLellan, H. J., L. Lauzier, and W. B. Bailey. 1953. The slope water off the Scotian shelf. J. Fish. Res. Board Can. 10:155-176. Miller, A. R. 1952. A pattern of surface coastal circulation inferred from surface salinity-temperature data and drift bottle recoveries. Woods Hole Oceanogr. Inst. Ref. 52-28, 14 p. Nelson, W. R., and J. S. Carpenter. 1968. Bottom longline explorations in the Gulf of Mexi- co — A report on "Oregon IPs" first cruise. Commer. Fish. Rev. 30(10):57-62. Rathbun, R. 1895. Physical inquiries. Off coast of southern New Eng- land and the Middle States. In Report upon the in- quiry respecting food-fishes and the fishing-grounds, p. 32-35. U.S. Comm. Fish Fish., Part 19, Rep. Comm. 1893. Seal, H. L. 1964. Multivariate statistical analysis for biologists. Metheun and Co., Ltd., Lond., 209 p. Spiess, E. B. 1977. Genes in populations. Wiley, N.Y., 780 p. Stefansson, U., L. P. Atkinson, and D. F. Bumpus. 1971. Hydrographic properties and circulation of the North Carolina shelf and slope waters. Deep-Sea Res. 18:383-420. Wolf, R. S., and W. F. Rathjen. 1974. Exploratory fishing activities of the UNDP/FAO Caribbean Fishery Development Project, 1965-1971: A summary. Mar. Fish. Rev. 36(9):l-8. 50 EFFECTS OF BEHAVIORAL INTERACTIONS ON THE CATCHABILITY OF AMERICAN LOBSTER, HOMARUS AMERICANUS, AND TWO SPECIES OF CANCER CRAB R. Anne Richards, 1 J. Stanley Cobb, 1 and Michael J. Fogarty 2 ABSTRACT Intraspecific and interspecific behavioral interactions may affect the probability of capturing Cancer irroratus, C. borealis, and Homarus americanus in lobster traps. To test this hypothesis, the catch per unitof effort(CPUE)of eachof these species in trapsstocked with C. irroratus, C. borealis, or H. americanus was compared with that obtained from empty baited traps (controls). In traps stocked with lobsters, the catch of all three species was significantly reduced. Traps stocked with 8 lobsters caught significantly fewer crabs than traps containing 3 lobsters. The only effect of stocking traps with crabs was to increase the catch of C. borealis in traps stocked with 3 crabs of either species. Results of laboratory experiments comparing crab CPUE in control traps with crab CPUE in traps stocked with 8 lobsters concurred with the field results. When H. americanus was stocked in the holding section (parlor) of the trap, a greater proportion of the crab catch was found in the entrance section (kitchen ). This behavioral response may facilitate escape of crabs from traps containing H. americanus. The distribution of the lobster catch was un- affected by stocking H. americanus or Cancer crabs in the parlor. Behavioral mechanisms underlying reductions in crab CPUE were investigated by laboratory observation of an actively fishing trap. When H. americanus was stocked, C. borealis avoided entering traps. Cancer irroratus entered the kitchen of traps containing H. americanus, but the proportion entering the parlor was reduced. The escape rate of both crab species increased in traps stocked with H. americanus. The position underneath the entrance to the parlor was preferred by all species. When both H. americanus and Cancer crabs were present in the trap, H. americanus occupied that position. A number of environmental and biological fac- tors are known to affect the probability of cap- turing crustaceans in traps. Water temperature and salinity are positively correlated with cap- ture rates of rock lobster, Panulirus cygnus, (Morgan 1974), and a linear relationship between temperature and the catchability of American lobster, Homarus americanus, was found by McLeese and Wilder (1958). Biological rhythms and physiological changes, such as those asso- ciated with the molt cycle (e.g., Chittleborough 1975), may affect feeding and other activities (e.g., Bennett 1974; Morgan 1974) and thus cause fluctuations in catchability. In addition, behav- ioral attributes such as avoidance of dead con- specifics (Hancock 1974; Morgan 1974; Chapman and Smith 1979), intraspecific attraction (re- viewed in Hancock 1974), or competitive relations 'Department of Zoology, University of Rhode Island, King- ston, RI 02881. 2 Rhode Island Department of Environmental Management, Division of Fish and Wildlife. 150 Fowler St., Wickford, RI 02852; present address: Northeast Fisheries Center Woods Hole Laboratory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543. 0.05, Table 1). Therefore the catches from both field locations were com- bined according to treatment. The number of trap hauls for each stock species was made equal by randomly deleting observations. The hypothe- sis that the CPUE of C. irroratus, C. borealis, and H. americanus is not affected by the pres- ence of other animals inside traps was tested by comparing the total catch of each species in stocked traps with the total catch in control traps. Catches obtained after 24 h immersion time were compared using a x 2 goodness of fit test (Zar 1974). In traps containing 8 or 3 lobsters, the total catch of C. irroratus, C. borealis, and H. ameri- canus was significantly reduced (x 2 <2> = 277.8, 35.1, 18.2, respectively, P<0.001) (Table 2). In addition, the catch of both species of crabs was significantly lower in 8-lobster treatments than in 3-lobster treatments (C. irroratus, \ 2 w = 22.9, Table 1.— x 2 values for 3 X 2 contingency tables comparing strings of each treatment type for Homarus americanus (Ha), Cancer irroratus (Ci), and C. borealis (Cb) between locations. A separate contingency table was made for each species caught. * = P<0.05, @ = expected frequency of one cell was <5. Corr parison of locations for Species caught Ha treatments Ci treatments Cb treatments C. borealis C. irroratus H. americanus 0.980 @ 3880 0.348 @ 0.920 48.357* 0146 2.675 2.594 1.816 P<0.001; C. borealis, x \d = 6.1, P<0.025). The catch of lobsters was not affected by the density of stocked lobsters (x 2 u> = 2.42, P>0.05). The only effect of stocking traps with crabs was to increase the catch of C. borealis in traps stocked with either 3 C. borealis or 3 C. irroratus (for both treatments, x 2 (d = 8.6, P<0.005). Stocking traps with crabs had no effect on the catch of lobsters (P>0.05). The average size of animals captured did not differ between treatments for any of the species (Student's t test, P>0.05) (Table 3). The results of the laboratory experiments in which lobsters were stocked concurred with those from the field. The catch of both C. irrora- tus and C. borealis was significantly reduced when H. americanus was in the parlor (Table 4). Behavior Location Within Trap The spatial distribution of animals caught in a trap may be affected by behavioral interactions among the trap occupants. To test this hypothe- sis, the proportion of the catch found in the entry section, or "kitchen," in control traps was com- pared with the proportion in the kitchen in stocked traps. All comparisons of proportions were made using the normal approximation for differences between two proportions (Zar 1974). Stocked animals were placed in the parlor. In both field and laboratory experiments, a Table 2.— Total numbers of Cancer irroratus, C. borealis, and Homarus americanus caught after 24-h immersion time in field experiments. Catch per trap haul is indicated in parentheses; control = empty baited traps; treatment refers to species stocked; n = no. of trap hauls for each treatment level. H. amer/canus-stocked C borea//s-stocked C. /rrorafi/s-stocked Species caught Control 3 8 Control 3 8 Control 3 8 C. irroratus 319(7.60) 100(2.38) 42(1.00) 300(8.82) 371(10.91) 300(8.82) 342(9.50) 365(10.14) 355(986) C. borealis 70(1.67) 36(0.86) 17(0.40) 61(1.79) 99(2.91) 78(2.29) 65(1.81) 102(2.83) 70(1.94) H. americanus 54(1.29) 31(0.74) 19(0.45) 23(0.68) 21(0.62) 33(0.97) 29(0.81) 29(0.81) 29(0.81) n 42 42 42 34 34 34 36 36 36 In = 336 Table 3. — Average size (mm) and standard deviation (SD) of Homarus americanus. Cancer bore- alis, and C. irroratus caught in all traps, locations combined. Size of crabs is carapace width; size of lobsters is carapace length. H. amencanus-stocked C. borea//s-stocked C. /rrorafus-stocked Species caught Control 3 8 Control 3 8 Control 3 8 C. irroratus X 91.7 91.8 92.2 90.6 91.1 92.2 91.5 89.5 92.1 SD (10.1) (11.5) (13.2) (9.9) (11.3) (10.4) (10.6) (118) (8.6) C. borealis X 92.8 94 8 94.8 93.3 94.5 92.3 94.4 94.6 92.7 SD (9.5) (9.2) (6.5) (10.6) (8.4) (8.1) (7.9) (6.9) (9.0) H. americanus X 68.3 73.4 74.8 722 71.1 73.2 71.2 72.2 71.8 SD (7.9) (6.8) (8.1) (6.7) (9.1) (96) (7.6) (7.0) (12.5) 54 RICHARDS ET AL.: BEHAVIORAL INTERACTIONS OF AMERICAN LOBSTER AND CANCER CRABS Table 4. — Total number of Cancer irroratus or C. borealis caught in 10 laboratory trials of each treatment and catch spe- cies. ** = P<0.001, \ 2 goodness of fit test. Treatment Species caught Control 8 Homarus amencanus x 2 C. irroratus C. borealis 49 66 15 20 238" 14.9" greater proportion of the crab catch was found in the kitchen of 8-lobster treatments than of con- trols (Tables 5, 6). Stocking traps with 3 lobsters had no effect on the distribution of crabs, and lob- sters were unaffected by either stock density of lobsters (P>0.05). Interspecific interactions between C. irroratus and C. borealis apparently influenced the distri- bution of these species inside traps. In traps stocked with either 3 or 8 C. irroratus, the pro- portion of the C. borealis catch found in the kitchen was significantly greater than in con- trols (Z = 2.50, P<0.01). In traps containing 3 C. borealis, the proportion of the C. irroratus catch found in the kitchen was significantly greater than in controls (Z = 2.50, P<0.01), but no effect was seen in traps stocked with 8 C. borealis (P>0.05) (Fig. 2). LJ .14 X o I- .12 X. .10 z .08 z o .06 I- cr .04 o Q_ o .02 rr *" C. irroratus a 5^ C 4) ^S ■o ro u) CO w o *-* ** o / V/ \ ' realis s d s\ IC.bo stocke ** 1 - z 8 C. bore trol — o / , O / / O o u / / n = n = -n = - n = 'rU' "n s^ 35 -34' -37^ 38 A0\36. C. borealis SPECIES CAUGHT Figure 2.— Proportion of Cancer irroratus and C. borealis found in the kitchen of traps stocked with congeners in field experiments. All data obtained after one setover day are in- cluded, n = number of trap hauls, ** = significant difference (P<0.01) between treatment and control, using normal ap- proximation for differences between two proportions (Zar 1974). Table 5.— In field experiments, spatial distribution of Cancer irroratus, C. borealis, and Homarus americanus catch in traps stocked with H. americanus (Ha). All data obtained after one setover day are included. Proportion of catch found in the kitchen of stocked traps was compared with controls using normal approximation for differences between two propor- tions (Z) (Zar 1974). n = number of trap hauls; * = P<0.05, ** = P<0.001. Proportion Species cau ght Treatment n in kitchen z C. irroratus 8 Ha 42 0.29 8.67" 3 Ha 54 0.06 0.91 ns Control 51 0.03 C. borealis 8 Ha 42 0.35 2.00' 3 Ha 54 009 67 ns Control 51 0.13 H americanus 8 Ha 42 000 0.94 ns 3 Ha 54 00 1 25 ns Control 51 003 Table 6. — In laboratory experiments, spatial distribution of the Cancer crab catch in traps stocked with Homarus ameri- canus (Ha). Proportion of catch found in the kitchen of stocked traps was compared with controls using normal approxima- tion for differences between two proportions (Z) (Zar 1974). n = no. ot trap hauls, * = f< U.UUU1. Species cau ght Treatment n Proportion in kitchen Z C. irroratus C. borealis 8 Ha Control 8 Ha Control 10 10 10 10 0.27 0.08 0.70 009 5.19* 527- Competition Inside Traps To further investigate how the location of ani- mals in a trap is affected by behavioral interac- tions, competition for preferred areas in the trap was studied in the laboratory. Frequency of occu- pation was used as an index of preference and was measured as the number of times a given position was occupied when censused every 15 min. The observed distribution of animals was compared with an expected uniform distribution using a x 2 goodness of fit test. For lobsters and for each crab species in the absence of lobsters, the preferred position in the parlor was underneath the entry head (C. irroratus, x 2 w = 202.0, P< 0.001; C. borealis, x \v =51.8, P<0.001; H. ameri- canus, x 2 (4> = 744.2, P<0.001). When lobsters were present, the number of crabs in the parlor decreased sharply, so comparisons between lob- ster-stocked and control traps were made using proportions. In the presence of lobsters, the preference of both crab species changed (C. irro- ratus, Z = 2.26, P<0.01; C. borealis, Z = 5.97, P<0.001). Cancer irroratus occupied the middle of the parlor, and C. borealis occupied the cor- ners most frequently when H. americanus was 55 FISHERY BULLETIN: VOL. 81. NO. 1 present (C. irroratus, x \v = 82.3, P<0.001; C. borealis, x 2 (4, = 52.5, P<0.001) (Table 7). Space inside the trap was partitioned into ver- tical strata. Both crab species showed a signifi- cant increase in occupation of the top part of the trap when lobsters were present (C. irroratus, 0.47 vs. 0.79, Z = 4.87, P<0.001; C. borealis, 0.21 vs. 0.38, Z = 1.76, P<0.05). This contrasts with 99% occurrence of lobsters in the bottom portion of the trap. from the parlor did not increase in lobster- stocked traps for either species (C. irroratus, Z= 1.37, P>0.05; C. borealis, Z = 0.37, P>0.05). DISCUSSION Trap Efficiency The results of the field and laboratory experi- ments demonstrate that the presence of lobsters Table 7.— Laboratory-observed frequency and relative frequency of occupation of positions in the parlor by Cancer irroratus, C. borealis, and Homarus americanus. Counts were weighted to com- pensate for unequal availability of positions due to trap design. * = significant (P<0.01) x 2 val- ues for frequency of occupation and preferred positions; + = significant (P<0.01) differences in occupation of a particular position in lobster-stocked traps and controls; ctl = control; lob = 5 lob- sters stocked. Position i sccupied Corner Under head & arner by head Side Middle Species caught ctl lob ctl lob ctl lob ctl lob ctl lob C. irroratus Frequency 129* 27 12 9.3 32 8 9 8 60 57' Relative frequency 0.53 0.25 + 005 0.09 0.13 0.07 0.04 0.07 025 0.52 + C. borealis Frequency 93* 3 32 42.5* 42.5 17.3 23 9 42 15 Relative frequency 0.40 0.04 + 0.14 0.49 + 0.18 0.20 0.10 010 0.18 0.17 H. americanus Frequency — 555* — 204.8 — 38.6 — 109 — 471 Relative frequency — 0.40 — 0.15 — 0.03 — 0.08 — 0.34 Trap Entry and Escapement Laboratory observations revealed that C. ir- roratus and C. borealis respond differently to traps stocked with H. americanus. The presence of H. americanus did not affect the number of C. irroratus entering the kitchen (39 vs. 33, x 2 u> = 0.35, P>0.05); however, significantly fewer C. borealis entered when H. americanus were stocked (35 vs. 8, x 2 (d = 18.2, P<0.001). The proportion of C. irroratus which moved from the kitchen to the parlor was significantly reduced in lobster-stocked traps (0.81 vs. 0.23, Z= 2.73, P<0.0001). The proportion of C. bore- alis entering the parlor did not decrease signifi- cantly when H. americanus was present (0.53 vs. 0.31, Z = 0.58, P>0.05); however, the number of C. borealis that had entered the kitchen was relatively low. The proportion of both C. irroratus and C. bo- realis which escaped the kitchen increased sig- nificantly in the presence of H. americanus (C. irroratus, 0.23 vs. 0.55, Z = 2.86, P<0.005; C. bo- realis, 0.26 vs. 0.63, Z= 1.97, P<0.025). Escape reduces the CPUE of crabs, and provide a pos- sible explanation for the inverse relationship be- tween lobster and crab catches seen in other studies (e.g., Stasko 1975; Krouse 1978; Fogarty and Borden 1980). This effect appears to be den- sity-dependent since fewer crabs were captured when a large number of lobsters were present. Factors other than behavioral interactions could cause negative correlations between lob- ster and crab catch rates. Cancer irroratus is often spatially separated from C. borealis and H. americanus in Narragansett Bay (Jeffries 1966; Fogarty 1976). Such discontinuous distributions could result in inverse catches of C. irroratus and H. americanus, or of C. irroratus and C. borealis, but do not explain the differences seen in the catch of adjacent traps in this study. Other fac- tors known to affect catchability (e.g., size, sex, reproductive condition, molt stage) were held constant among stocked animals used in the dif- ferent treatments. Temperature changed little over the course of the study (average surface tem- perature, 21.9°±2.15°C). This and other environ- mental variables would have affected all treat- 56 RICHARDS ET AL.: BEHAVIORAL INTERACTION'S OF AMERICAN LOBSTER AND CANCER CRABS ments equally. The nonrandom arrangement of treatment levels within strings could have biased catch rates through gear competition. However, we feel the assumption that equal numbers of animals were attracted to all traps is valid for the following reason. If gear competition caused the reduced crab catches in lobster-stocked strings, a similar pattern of catch rates would have been seen in crab-stocked strings. This was not the case. Cancer irroratus is a prey item for lobsters (Squires 1970; Weiss 1970; Scarratt and Lowe 1972; Ennis 1973), suggesting that the decreased catch of this species in traps containing lob- sters may be the result of predator-avoidance be- havior. Cancer borealis and H. americanus are thought to compete for shelter space in rocky subtidal habitats (Stewart 1972; Fogarty 1976; Cooper and Uzmann 1977; Wang 1982). In labo- ratory studies (Fogarty 1976), H. americanus dominated C. borealis for possession of shelter. This dominance appeared to be the result of avoidance by C. borealis rather than overt ag- gressive interactions. Such behavior may cause reduced catches of C. borealis in traps containing lobsters. The reduction in lobster CPUE when lobsters were stocked is not surprising since lobsters are known to be highly aggressive and generally in- habit shelter alone under natural conditions (Cobb 1971; Cooper and Uzmann 1980). Trap sat- uration apparently becomes important for lob- sters at relatively low catch levels since traps stocked with 8 and 3 lobsters were equally effec- tive in reducing the lobster catch. In a laboratory experiment reported by Smolowitz (1978), a re- duction in trap entry was seen with only 1 or 2 lobsters in the trap. Reduced entry was probably important in the present study since escapement of stocked lobsters was low (10.1%). Stock rates used for crabs were low compared with crab catches in control traps. At higher den- sities, crabs might have had a more significant effect on the catch of lobsters. An increased lob- ster catch might be expected in traps containing C irroratus, a lobster prey item (Squires 1970; Weiss 1970; Ennis 1973; McLeese 1974). How- ever, the presence of live prey may not signifi- cantly increase the attractiveness of an already baited trap. No evidence was seen of lobster predation on crabs in traps. Similarly a decrease in lobster catch might be expected in traps con- taining a competitor (C. borealis). However, C. borealis is less aggressive than H. americanus (Fogarty 1976; Wang 1982) and occupies mutu- ally desirable shelters through passive means rather than active displacement, as shown in Stewart's (1972) study. Trap saturation apparently was not an impor- tant factor for crabs at the stock levels used, since crab catches in crab-stocked traps were not re- duced below the level of control traps. In labora- tory observations, Miller (1978, 1979a, 1980) noted that intraspecific agonistic interactions among C irroratus, Hyas araneus, and C. pro- ductus aggregating downstream from baited traps often resulted in departure from the trap area. He suggested that trap saturation in these three species was due in part to "intimidation" of crabs outside the trap by those inside. However, at relatively low catch densities, the effects of aggression may be minimal. The increased C. borealis catch in traps stocked with 3 crabs of either species is difficult to ex- plain. Release of attractants from the bait by feeding activity could enhance trap entry. As crab density inside the trap increases, such enhancement may be countered by increased aggression, reducing trap entry rates and in- creasing escapement. These speculations do not explain why the C. irroratus catch was not simi- larly increased by a low stock density of either crab species. Behavior Location Within Trap Behavioral interactions apparently affected the spatial distribution of animals in traps. A greater proportion of the crab catch was found in the kitchen when 8 lobsters were stocked in the parlor. This may have been the result of the avoid- ance responses discussed above and may enhance escapement of crabs from traps containing lob- sters. Cancer borealis shifted to the kitchen in both density levels of C. irroratus-stocked traps, but the distribution of C. irroratus changed sig- nificantly only in traps stocked with 3 C. bore- alis. Perhaps the generally greater activity of C. irroratus (Jeffries 1966; pers. obs.) serves as a deterrent to parlor entry by C. borealis. Both spe- cies may be influenced by prior residence effects in which an advantage is conferred upon the in- dividual^) initially utilizing a resource (e.g., Sinclair 1977; Davies 1978; O'Neill and Cobb 1979). Such an effect may have been caused by the stocking procedure. 57 FISHERY BULLETIN: VOL. 81, NO. 1 Competition Inside Traps During scuba diving observations of lobster traps, Pecci etal. (1978) noted an apparent domi- nance of crabs over lobsters in occupation of mutually desirable "niches" in traps. They re- ported that when both crabs and lobsters were present in traps, crabs always occupied positions that were evidently preferred by both species. The observations of this study contradict those of Pecci et al. Both crab species were displaced by lobsters. It is possible that our results re- flect a prior residence advantage conferred on lobsters by the stocking procedure. However, our findings agree with what is known of the relative aggressiveness of H. americanus, C. borealis, and C. irroratus (Fogarty 1976; Wang 1982). Escapement could be a significant factor in re- ducing the efficiency of traps. Skud 5 considered this the most likely explanation for declining catch rates for lobster over time. High escape rates for two species of Cancer have been ob- served by Miller (1979b) and High (1976). In this study, escape of both crab species from the kitchen increased when lobsters were present in the parlor, probably due to the behavioral inter- actions described above. Escape of crabs from the parlor did not increase when lobsters were stocked. This may reflect both the design of the parlor head, which makes escape more difficult, and the small sample size resulting from a low rate of entry to the parlor. In summary, the behavioral mechanisms in- volved in reducing crab catches in traps contain- ing lobsters were Trap Entry and Escapement In the laboratory, the presence of H. america- nus in a trap did not affect the number of C. irro- ratus entering the kitchen, but did decrease the number of C. boreal is entering. Just the opposite might have been expected in light of the preda- tor-prey relationship between C. irroratus and H. americanus. We observed no interactions be- tween animals inside the trap and those outside; thus the sensory basis for avoidance by C. bore- alis of traps containing lobsters is unknown. The proportion of C. irroratus moving from the kitchen to the parlor was reduced in lobster- stocked traps. The decrease in parlor entry rate for C. borealis was not statistically significant; however, the number of C borealis that had entered the kitchen was relatively low. Reduced parlor entry appeared to be the direct result of interactions between animals in the two trap compartments. These typically consisted of a lob- ster displaying (meral spread) or lunging at a crab climbing up the parlor head, resulting in retreat to the kitchen by the crab. In several in- stances, crabs hanging from the parlor head con- tacted a lobster, which responded by displaying or attacking the crab. The crab then pulled back up into the parlor head and returned to the kitchen. General lobster activity (fighting, ex- ploring, etc.) had a similar effect on crabs in the parlor head. Only 24% of C. irroratus and 10% of C. borealis entering the parlor head actually entered the parlor when lobsters were stocked. Parlor entrants increased to 60% and 67%, re- spectively, in control traps. 1 ) For C. borealis, entry to the trap is reduced, and escapement of those that enter the kitchen is increased. 2) For C. irroratus, trap entry is not reduced, but entry to the parlor decreases and rate of escape from the kitchen increases. SUMMARY This study demonstrated that behavioral inter- actions between animals attracted to traps can have significant effects on the probability of their capture. The CPUE of American lobsters and of two species of commercially harvested Cancer crabs was significantly reduced in traps containing lobsters. Such effects may be density- dependent, since significantly fewer crabs were caught in traps containing 8 lobsters than in traps containing 3 lobsters. The proportion of captured crabs occupying each trap section changed significantly when lobsters were stocked, and behavioral observations indicated that lobsters occupy the mutually preferred posi- tions in traps. The behavioral mechanisms re- sponsible for decreased crab catches included both reduced entry (C. borealis) and increased escapement (C. irroratus and C. borealis). These results reflect the behavioral and ecological rela- tions of the three species. 5 Skud, B. E. 1976. Soak-time and the catch per pot in an offshore fishery for lobsters (Homarus ameriranus). ICES Special Meeting on Population Assessments of Shellfish Stocks, No. 8, 25 p. 58 RICHARDS ET AL.: BEHAVIORAL INTERACTIONS OF AMERICAN LOBSTER AND CANCER CRABS ACKNOWLEDGMENTS We are grateful to Saul B. Saila and H. Perry Jeffries for their helpful suggestions throughout the project. The Rhode Island Department of En- vironmental Management generously provided a boat and laboratory facilities. Additional lab- oratory space was provided by the University of Rhode Island's Graduate School of Oceanog- raphy. Funding was provided by the University of Rhode Island Sea Grant Development fund. The thoughtful reviews of Jay K rouse, Robert Elner, and a journal reviewer were much appre- ciated. LITERATURE CITED Bennett. D. B. 1974. The effects of pot immersion time on catches of crabs. Cancer pagurus (L.) and lobsters, Honiara* gam- marus (L.). J. Cons. Int. Explor. Mer 35:332-336. Chapman, C. J., and G. L. Smith. 1979. Creel catches of crab, Cancer pagurus L. using dif- ferent baits. J. Cons. Int. Explor. Mer 38:226-229. Chittleborough, R. G. 1975. Environmental factors affecting growth and sur- vival of juvenile western rock lobsters Panulirus longi- pes (Milne-Edwards). Aust. J. Mar. Freshw. Res. 26: 177-196. Cobb, J. S. 1971. The shelter-related behavior of the lobster, Homar- us americanus. Ecology 52:108-115. Cooper, R. A., and J. R. Uzmann. 1977. Ecology of juvenile and adult clawed lobsters, Ho- marus americanus, Homarus gammarus, and Nephrops norvegicus. In B. F. Phillips and J. S. Cobb (editors), Workshop on lobster and rock lobster ecology and physi- ology, p. 187-208. CSIRO, Aust., Div. Fish. Oceanogr., Circ. 7. 1980. Ecology of juvenile and adult Homarus. In J. S. Cobb and B. F. Phillips (editors), The biology and man- agement of lobsters, Vol. II, p. 97-142. Acad. Press, N.Y. Davies, N. B. 1978. Territorial defense in the speckled wood butterfly (Pararge aegeria): the resident always wins. Anim. Behav. 26:138-147. Ennis, G. P. 1973. Food, feeding, and condition of lobsters, Homarus americanus, throughout the seasonal cycle in Bonavista Bay, Newfoundland. J. Fish. Res. Board Can. 30:1905- 1909. FOGARTY, M. J. 1976. Competition and resource partitioning in two spe- cies of Cancer (Crustacea, Brachyura). M.S. Thesis, Univ. Rhode Island, Kingston, 94 p. FOGARTY, M. J., AND D. V. D. BORDEN. 1980. Effects of trap venting on gear selectivity in the inshore Rhode Island American lobster, Homarus amer- icanus, fishery. Fish. Bull., U.S. 77:925-933. Hancock, D. A. 1974. Attraction and avoidance in marine inverte- brates—their possible role in developing an artificial bait. J. Cons. Int. Explor. Mer 35:328-331. High, W. L. 1976. Escape of Dungeness crabs from pots. Mar. Fish. Rev. 38(4):19-23. Jeffries, H. P. 1966. Partitioning of the estuarine environment by two species of Cancer. Ecology 47:477-481. Kennedy, D., and M. S. Bruno. 1961. The spectral sensitivity of crayfish and lobster vision. J. Gen. Physiol. 44:1089-1102. Krouse, J. S. 1978. Effectiveness of escape vent shape in traps for catching legal-sized lobster, Homarus americanus, and harvestable-sized crabs, Cancer borealis and Cancer ir- roratus. Fish. Bull, U.S. 76:425-432. McLeese, D. W. 1974. Olfactory responses of lobsters (Homarus ameri- canus) to solutions from prey species and to seawater extracts and chemical fractions offish muscle and effects of antennule ablation. Mar. Behav. Physiol. 2:237- 249. McLeese. D. W., and D. G. Wilder. 1958. The activity and catchability of the lobster (Ho- marus americanus) in relation to temperature. J. Fish. Res. Board Can. 15:1345-1354. Miller, R. J. 1978. Crab (Cancer irroratus and Hyas araneus) ease of entry to baited traps. Can. Fish. Mar. Serv. Tech. Rep. 771, 8 p. 1979a. Entry of Cancer productus to baited traps. J. Cons. Int. Explor. Mer 38:220-225. 1979b. Saturation of crab traps: reduced entry and es- capement. J. Cons. Int. Explor. Mer 38:338-345. 1980. Design criteria for crab traps. J. Cons. Int. Ex- plor. Mer 39:140-147. Morgan, G. R. 1974. Aspects of the population dynamics of the western rock lobster, Panulirus cygnus George. II. Seasonal changes in the catchability coefficient. Aust. J. Mar. Freshw. Res. 25:249-259. O'Neill, D. J., and J. S. Cobb. 1979. Some factors influencing the outcome of shelter competition in lobsters (Homarus americanus). Mar. Behav. Physiol. 6:33-45. Pecci, K. J., R. A. Cooper, C. D. Newell, R. A. Clifford, and R. J. Smolowitz. 1978. Ghost fishing of vented and unvented lobster, Ho- marus americanus, traps. Mar. Fish. Rev. 40(5-6):9-43. RlCKER, W. E. 1975. Computation and interpretation of biological sta- tistics of fish populations. Fish. Res. Board Can. Bull. 191, 382 p. Scarratt, D. J., and R. Lowe. 1972. Biology of the rock crab (Cancer irroratus) in Northumberland Strait. J. Fish. Res. Board Can. 29: 161-166. Sinclair, M. E. 1977. Agonistic behavior of the stone crab, Menippe mer- cenaria (Say). Anim. Behav. 25:193-207. Smolowitz, R. J. 1978. Trap design and ghost fishing: Discussion. Mar. Fish. Rev. 40(5-6):59-67. Squires, H. J. 1970. Lobster (Homarus americanus) fishery and ecol- ogy in Port au Port Bay, Newfoundland, 1960-65. Proc. Natl. Shellfish. Assoc. 60:22-39. 59 FISHERY BULLETIN: VOL. 81. NO. 1 Stasko, A. B. 1975. Modified lobster traps for catching crabs and keep- ing lobsters out. J. Fish. Res. Board Can. 32:2515-2520. Stewart, L. L. 1972. The seasonal movements, population dynamics and ecology of the lobster, Homarus americanus (Milne- Edwards), off Ram Island, Connecticut. Ph.D. Thesis, Univ. Connecticut, Storrs, 112 p. Wang, D. 1982. The behavioral ecology of competition among three decapod species, the American lobster, Homarus ameri- canus, the Jonah crab, Cancer borealis, and the rock crab, Cancer irroratus in rocky habitats. Ph.D. Thesis, Dep. Zoology, Univ. Rhode Island, Kingston, 105 p. Weiss, H. M. 1970. The diet and feeding behavior of the lobster, Ho- marus americanus, in Long Island Sound. Ph.D. The- sis, Univ. Connecticut, Storrs, 104 p. Zar, J. H. 1974. Biostatistical analysis. Prentice-Hall, Englewood Cliffs, N. J., 620 p. 60 THE REPRODUCTIVE BIOLOGY OF THE ATLANTIC SHARPNOSE SHARK, RHIZOPRIONODON TERRAENOVAE (RICHARDSON) Glenn R. Parsons 1 ABSTRACT Atlantic sharpnose sharks, Rhizoprionodon terraenovae (Richardson), were collected in the north central Gulf of Mexico from June 1979 to May 1980. The principal sampling devices employed were longline, trawl, and rod and reel. From a total of 215 Atlantic sharpnose sharks obtained during the study, 144 were female and 71 were male, ranging from 30 to 107 cm total lengths. The reproductive anatomy of both male and female sharpnose sharks is described. Atlantic sharpnose sharks differ from other carcharhinids in that the ovary is developed on the left side in females and overlapping siphon sacs are present in males. Clasper development suggests that males mature at about 80 cm total length, while ovarian egg diameters show that female maturation occurs at about 85 cm. Matings occur primarily between mid-May and mid-July. Embryonic growth is rapid immediately after fertilization during summer and fall but declines during winter and spring. Gestation requires 10 to 11 months and parturitions probably peak in June. Pups are released near shore at an average total length of 32 cm. Statistical analyses reveal a positive relationship between adult total length and litter size, with the largest individuals being the most fecund. An inverse relationship was observed between the numbers of embryos per uterus and embryo size. Mechanical "packing" within the uterus is proposed to explain the relationship. The seasonal distribution of sharpnose sharks was found to be determined by an inshore-offshore migration. The data indicate that during winter months in deeper offshore waters, aggregates of predominately adult female sharpnose sharks may be encountered. The sex ratio at birth was found to be 1:1 but among adults collected a 1:2.8 ratio was observed. Studies dealing with the reproductive biology of elasmobranchs have fallen far behind the volu- minous amount of data that have accumulated on reproduction in the teleostean fishes. The north- ern Gulf of Mexico has been an area of particu- lar neglect with only a few rather generalized studies (Springer 1938, 1940, 1950; Baughman and Springer 1950). Springer's (1960) classic work on the natural history of the sandbar shark, Careharhinus milberti (Eulamia milberti), con- tains a great deal of reproductive information that might be applied to carcharhinid sharks in general. Likewise, Clark and von Schmidt's (1965) survey of the sharks of the central gulf coast of Florida provided valuable reproductive data. The understanding of the life history of the blue shark, Prionace glauca, was furthered by Pratt's (1979) examination of its reproductive biology. Data concerning the life history of Rhizoprion- odon terraenovae are scarce. Rhizoprionodon spe- cies are believed to be born in the late spring and •Department of Biological Sciences, University of South Alabama, Mobile, AL 36688; present address: Department of Marine Science, University of South Florida, 140 Seventh Ave- nue South, St. Petersburg, FL 33701. Manuscript accepted June 1982. FISHERY BULLETIN: VOL. 81, NO. 1. 1983. summer. Bigelow and Schroeder (1948) reported that recently born specimens can be collected from Florida in July and that they were also present off the mouth of the Mississippi River in August. Skocik (1969) reported that pups are usually born in the spring but no data were avail- able on mating season or gestation period. Rhizoprionodon species are viviparous, the embryos obtaining nourishment via a placental connection (sometimes called a "pseudo- or yolk- sac placenta") between mother and embryo. Fe- cundity in Rhizoprionodon has been variously reported. Baughman and Springer (1950) report- ed four embryos for R. terraenovae. Bass et al. (1975) found an average of 4.7 embryos with a range of two to eight in R. acutus. Skocik (1969) reported a litter size of 12 for R. terraenovae, while Bigelow and Schroeder (1948) reported the same number for R. terraenovae taken around Cuba. Clark and von Schmidt (1965) briefly sur- veyed R. terraenovae off Englewood, Fla., and found one 83 cm female with five eggs. They also reported that all adult females examined had functional left ovaries. Compagno (1978) report- ed a range of one to four embryos for R. porosus. The pups of R. terraenovae have been reported to be 11 to 16 in (27.9 to 40.6 cm) at birth (Baugh- 61 FISHERY BULLETIN: VOL. 81. NO. 1 man and Springer 1950). Bigelow and Schroeder (1948) reported that specimens from Texas showing traces of the umbilical scar were from 280 to 407 mm long. Among R. terraenovae populations, adults are commonly 26 to 30 in (66 to 76 cm) total length (TL) (Baughman and Springer 1950), but the size at which male and female Atlantic sharp- nose sharks mature is unknown. In his revision of the genera Scoliodon, Loxodon, and Rhizoprion- odon, V. G. Springer (1964) reported that insuffi- cient information was available to establish the size at which males first mature but it appeared that maturation occurs at >640 mm TL. Bass et al. (1975) reported that male R. acutus mature between 68 and 72 cm and females at 70 to 80 cm TL. The present study is an attempt to clarify some of the known aspects of R. terraenovae reproduc- tive biology as well as to provide additional in- formation. The reproductive "strategy" of the Atlantic sharpnose shark is also examined. METHODS AND MATERIALS Atlantic sharpnose sharks, Rhizoprionodon terraenovae (Richardson), were collected in the north central Gulf of Mexico from June 1979 to May 1980. The principal sampling devices em- ployed were longline, trawl, and rod and reel. Floating longline generally gave the best re- sults (Table 1). The technique, as used by Japa- nese fishermen, is described by Lopez et al. (1979). Because of the hazard to navigation that a floating longline represents, longlining opera- tions were undertaken exclusively in deep waters offshore (Fig. 1). Longline sets were made in 10 to 28 fathom (18 to 51 m) depths, approximately due south of Dauphin Island, Ala. A trawl was used to collect specimens both inshore as well as offshore. Rod and reel, gill net, and seine were used exclusively inshore. Specimens were immediately weighed and sexed. Total, fork, and standard lengths were 30 u 30' 30" 15' 30 u 00' 29 u »5' 29 u 30' MOBILE O. O 0^2) DAUPHIN q ISLAND fort Morgan' o ••-S GULF OF MEXICO 10 fm 20 fm 30 fm 15' 00' 87° k5' Figure 1.— Coastal Alabama study area of the Atlantic sharp- nose shark. Offshore points (closed circles) represent longline and trawl sites. Inshore points (open circles) represent trawl, gill net, rod and reel, and seine sites. measured to the nearest 0.1 cm. Lengths of the claspers and siphon sacs were measured on all male specimens. All specimens were dissected immediately in the field by an incision starting at the cloaca and extending to the midpectoral region. Notes on reproductive condition in males Table 1. — Landings of Atlantic sharpnose sharks by month and by method. Longline and trawl produced more than 60% of the sharpnose shark specimens. Sharpnose sharks were collected in 10 of the 12 mo of the study period. — indicates no collections; indicates collections attempted but no sharks landed. Jan. Feb Mar Apr May June July Aug. Sept. Oct. Nov. Dec Totals Longline — 1 — 2 — — 14 — 4 19 35 75 Trawl — 1 — 21 8 — 6 8 6 9 59 Rod/reel — — 8 1 38 1 2 50 Gill net — — — 2 15 4 4 — — — — 25 Seine — — — — — 6 — — — — — 6 Totals — 2 — 25 31 7 48 27 8 4 28 35 215 62 PARSONS: REPRODUCTIVE BIOLOGY OF ATLANTIC SHARPNOSE SHARK were taken, using those indicators of maturity reported by Clark and von Schmidt (1965). Dis- sections of males allowed examinations of the re- productive systems and measurements of testicu- lar length, weight, and volume. Testes and epididymides were removed from some specimens, preserved in 10% Formalin 2 , and returned to the laboratory. Histological sec- tions of testes as well as epididymides were pre- pared. The tissues were embedded in paraffin, sectioned at 7 jum, stained with hematoxylin and eosin, and examined with phase contrast micros- copy. Sperm smears were also examined under the microscope. After obtaining weight and total, fork, and standard lengths, female specimens were dis- sected and their reproductive organs examined. Ovarian lengths as well as the number of ovarian eggs and their diameters were recorded. When embryos were present, the number, sex, total length, and wet weight were determined for each uterus. When appropriate, the data were keypunched and statistically evaluated, using the McGill University System for Interactive Computing (MUSIC) time sharing system. The STATPAK computer program, a statistical package con- taining 23 statistical analyses and data modifica- tion routines, was used to analyze the data. RESULTS AND DISCUSSION Reproductive Anatomy Ovarian Structure Forty-two Atlantic sharpnose shark ovaries were examined during the study period. E lasmo- branchs possess a great deal of variability in the structure of the ovary (Dodd 1972). The ovary of the adult Atlantic sharpnose shark is an un- paired, tear-shaped organ, 6 to 10 cm long and 3 to 5 cm wide. Unlike other carcharhinids, the ovary of the sharpnose shark is developed on the left side only. Structure and location of the sharp- nose shark ovary (aside from its position on the left side of the body cavity) are similar to that found in the blue shark (Pratt 1979). The adult sharpnose shark's ovary, during most of the year, is filled with many small (ca. 2.0 to 5.0 mm) oocytes embedded in dense connective tissue. Outside the breeding season the ovary of the adult female contains an average of about 30 oocytes greater than ca. 2 mm in diameter. These oocytes serve as a "pool" from which the next gen- eration of eggs will be drawn. In some ovaries, unusual, bright red, fluid-filled structures were found, ranging from about 2 to 8 mm in diame- ter (Fig. 2). These structures are assumed to be oocytes in a state of atresia that had failed to ovu- late during the most recent breeding period. These preovulatory structures may be "corpora atretica," which are derived from egg-containing follicles. In Cetorhinus maximus the corpora atretica are believed to arise from follicles that have attained a diameter of about 1.0 mm (Dodd 1972). The corpora atretica consist of vacuolated peripheral cells and a central cavity and are well vascularized (Dodd 1972). ANTERIOR Normal Developing Ovum Atretic Ovum 2 Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. POSTERIOR FIGURE 2.— Diagram of an Atlantic sharpnose shark ovary taken in December from a 93 cm gravid female. A red, fluid filled (atretic?) ovum can be seen in the center of the ovary. Ovulation As ovulation approaches, rapid yolk deposi- tion occurs in four to eight of the many smaller oocytes. The "selected" oocytes are preferentially yolked, while the others undergo atresia. At or near ovulation the ovary appears highly vascu- larized and the large, yellow oocytes fill the en- tire ovary (Fig. 3). Measurements of both ovarian and uterine oocytes suggest that ovulation occurs at an egg diameter of about 20 mm. After ovulation, the eggs move through the body cavity into the ostium tubae which forms the anterior end of the oviduct. In most cases 63 FISHERY BULLETIN: VOL. 81, NO. 1 FULL SIZE OVARIAN EGG ANTERIOR I cm POSTERIOR OVARY Figure 3.— "Ripe" ovary of an Atlantic sharpnose shark. The ovary contains ca. 20 mm ova that are ready to ovulate. an equal number of ova enter both oviducts, although in some instances greatly dispropor- tionate numbers of embryos were found between right and left uteri. The eggs move through the oviducts to the oviducal gland where fertilization probably takes place. The oviducal gland (Fig. 4) in the Atlantic sharpnose shark is a paired struc- ture located at the forward end of the oviduct. The oviducal glands are the source of the egg case, and in some sharks the glands may be the cranial oviduct lumen opening caudal oviduct ANTERIOR POSTERIOR 1.0mm Figure 4.— Diagram of an oviducal gland taken from a mature female Atlantic sharpnose shark. site of long-term sperm storage (Pratt 1979). Viable sperm can be found within the lumen of those tubules within the gland which secretes the egg shell (Wourms 1977). As no histological sec- tions of adult sharpnose sharks' oviducal glands were prepared, the question of sperm storage in sharpnose sharks remains unresolved. Prasad (1944), however, noted the presence of spermato- zoa in the oviducal glands of Scoliodon sorra- kowah, a closely related Indian Ocean species. This observation suggests that the oviducal gland may have at least a short-term storage capacity. After moving through the oviducal gland the fertilized eggs then move to the uterus where they become implanted in depressions in the uterine wall. At this point the eggs are found en- cased in a thin, yellowish shell with pointed ends (Bigelow and Schroeder 1948). Within the uterus the eggs are elongate, averaging about 18 mm wide and about 32 mm long. Fertilization is ap- parently very efficient since in examination of 315 embryos only two unfertile eggs were noted (0.6%). Placentation and Structure of the Umbilical Cord During the first 2.5 to 3.0 mo of gestation, the Atlantic sharpnose shark embryos depend upon the yolk sac for nourishment. After about 3 mo the yolk sac has become intimately associated with the uterine wall to form a yolk-sac placenta. October embryos, i.e., 3 mo old, were ca. 16 to 20 cm and had well-developed placentas with little yolk material remaining. By November, 4 mo into gestation, embryos were 19 to 23 cm long and no yolk material remained in the placenta. In a related Indian Ocean species, Scoliodon sor- rakowah, Mahadevan (1940) described a very thick vascularized area of the uterine wall, re- ferred to as a trophonematous cup, which forms to receive the yolk sac of the foetus. This vascu- larized area was also noted in the Atlantic sharp- nose shark. Development of the umbilical cord closely par- allels placentation. The umbilical cord is con- nected on the embryo's ventral surface in the midpectoral region. Very early in development the umbilical cord is virtually naked. By the time the embryos have grown to about 6.0 cm TL the umbilical cord has developed many knoblike ap- pendages which give it a "pipe-cleaner" appear- ance. The appendages are about 1 mm long, and terminate in one or a cluster of several grapelike 64 PARSONS: REPRODUCTIVE BIOLOGY OF ATLANTIC SHARPNOSE SHARK distentions. Budker (1971) suggested that in ad- dition to placentally derived nutrients, these appendages may allow the embryo to absorb di- rectly nutritive substances that are secreted by the uterine lining. This type of nutrition is termed histotrophic. As gestation progresses the append- ages of the sharpnose shark's umbilical cord lengthen and change morphologically. Full-term embryos possessed umbilical cords about 10 to 12 cm long with appendages about 10 mm. The pro- jections at this time have a foliose appearance, i.e., flattened, extensively branched, and termi- nating in rounded, flat expansions. This differs from the fingerlike shape described for the pro- jections found on the umbilical cord of Sphyrna tiburo (Schlernitzauer and Gilbert 1966). Structure of Claspers and Siphon Sac The paired claspers of the adult male Atlantic sharpnose shark are much the same as those of other carcharhinid sharks. The claspers are rigid, calcified, intromittent organs that rotate freely around their attachment base. The tip, or rhipidion, expands whereupon the rigid carti- lages of the tip are directed at right angles to the main axis of the clasper. This expansion is be- lieved to function as an anchor, holding the clasp- er in the oviduct during copulation. Under nor- mal circumstances the claspers are directed posteriorly. Springer (1960) has suggested that just prior to mating the claspers of large carcha- rhinid sharks such as Eulamia milberti (Carcha- rhinus milberti) rotate in and forward. Expan- sion of the rhipidion occurs independently after insertion of the clasper into the oviduct of the fe- male. This apparently also occurs in the Atlantic sharpnose shark, since a live specimen captured in December had one clasper oriented in this fashion, with the rhipidion expanded, probably a result of trauma. The clasper gradually returned to normal after about 3 min. The siphon sac in the adult Atlantic sharpnose shark is a muscular, subdermal organ which be- gins at the base of the claspers, extends anteriorly along the ventral surface, and ends just short of the coracoid bar. The sac in adults ranges from about 20 to 28 cm long and 1 to 2 cm wide. Unlike other shark species which have paired separate siphon sacs, Atlantic sharpnose sharks possess overlapping sacs which communicate with the claspers via an opening located at the base of each clasper. Springer (1960) suggested that the siphon sac is filled with water just prior to mating and is used to flush sperm along the clasper groove and into the oviducts during copulation. The clasper siphon of adult spiny dogfish, Squa- lus acanthias, has been found to be a rich source of serotonin. This suggests that the siphon-sac secretion may play a role in affecting the mech- anism of copulation and ejaculation in the male, or by eliciting contractions of the female repro- ductive tract, thus influencing passage of sperm and fertilization (Mann 1960). Structure of the Testes and Epididymides The testes in the adult male Atlantic sharpnose sharks are paired, elongate, flattened organs (Fig. 5). Depending on the season and the size of the adult, the testes range from 13 to 20 cm long, 1 to 2 cm wide, and 0.5 to 1.0 cm thick. The testes are located dorsal to the lobes of the liver at the anterior end of the peritoneal cavity. The organs are supported here by a mesorchium. Microscopic examination of a mature testis of the sharpnose shark shows that the organ is filled ANTERIOR ^i POSTERIOR cm TEST I S Figure 5.— Diagram of a "ripe" Atlantic sharpnose shark tes- tis. The testis is turgid indicative of the reproductively active condition. 65 FISHERY BULLETIN: VOL. 81. NO. 1 with spherical seminiferous ampullae, much the same as are found in spiny dogfish (Simpson and Wardle 1967) and blue shark (Pratt 1979). Histo- logical sections of mature testes demonstrate that these ampullae contain spermatozoa in vari- ous stages of development (Fig. 6). Viewed in cross section, the heads of the mature spermato- zoa are arranged in discrete groups around the periphery of the spherical ampullae. The spermatozoa leave the testis by way of the efferent ductules and enter the epididymis. The epididymis is a paired organ located above the testis against the dorsal wall of the abdominal cavity. The sharpnose shark's epididymis is about 15 cm long, 1.0 cm wide, and 0.5 cm thick. Histological sections of an epididymis from a re- productively active sharpnose shark reveal great numbers of spermatozoa present in the tubules of the organ (Fig. 7). Maturation Males Maturity in animals can generally be deter- mined by comparing external secondary sex characters in adults with the same characters in smaller individuals. Using two indicators of sex- ual maturity (i.e., clasper growth and siphon-sac development), it was determined that matura- tion of the male Atlantic sharpnose shark begins at about 60 to 65 cm TL and is complete at about 80 cm. At <65 cm TL the clasper length represents about 2.5% of the adult total length. Regression analysis shows that the claspers undergo a period of rapid growth with a major inflection in the line occurring at 65 to 70 cm TL (Fig. 8). The claspers quickly elongate, growing 3 cm within a short period of time to represent 7 to 8% of the total length. The smallest mature males exam- ined were about 80 cm long and their claspers represented about 7.8% of total length. There were many individuals examined between 75 and 80 cm TL that possessed elongated claspers, but incomplete calcification of the claspers indi- cated that the specimens were not mature. The clasper grows faster than the total length at the onset of maturation and for a short period into adult life. Regression analysis indicates that from about 85 to 95 cm TL the relationship is un- changing, but after 95 cm there is a period of Figure 6.— Histological section of a testis from a mature Atlantic sharpnose shark (X440). The cross sections show that the heads of the mature spermatozoa are arranged in discrete groups around the periphery of the spherical seminiferous ampullae. 66 PARSONS: REPRODUCTIVE BIOLOGY OF ATLANTIC SHARPNOSE SHARK ¥!*?& . FIGURE 7.— Histological section of an epididymis from a mature Atlantic sharpnose shark (X 140). Large num- bers of spermatozoa are present within the tubules of the structure. negative allometric growth. The claspers, after attaining their functional length, do not continue to grow or at least grow very little. This is a ten- able hypothesis since continued growth would not necessarily enhance the claspers' utility. Development of the siphon sacs coincides close- ly with the rapid increase in clasper length (Fig. 9). This muscular, subdermal organ is nonexis- tent until the onset of maturity. The siphon sacs develop quickly and represent about 28% of the 88t 74 5 59 0. 4 5 31 •6t 30 o OBSERVED • REGRESSION ESTIMATED '8-8- o o -o OO o • O \Q 8 7 O 8\D 2b o/°° o ' ° \° _l o < 20 in 15 z o o I Q. 35 7 44 1 52-4 60 8 69 2 77 5 85 9 94 2 102 6 TOTAL LENGTH (cm) Figure 8. — The maturation of male Atlantic sharpnose sharks as evidenced by clasper development. The regression line indi- cates that maturation occurs between 80 and 85 cm total length, AT = 70. 10 5 35-7 44 I 52-4 60-8 69-2 77-5 85 9 94 2 102 6 TOTAL LENGTH (cm) Figure 9.— The maturation of male Atlantic sharpnose sharks as evidenced by siphon-sac development. The scatter diagram suggests that maturation occurs at about 80 cm total length, AT = 35. 67 FISHERY BULLETIN: VOL. 81. NO. 1 total length at maturity. The smallest mature individuals were about 80 cm and possessed si- phon sacs about 23% of total length. Females Maturation in females was determined by ex- amining the developing ovary and ovarian eggs. Females were found to mature at a greater total length than males. The ovary does not begin to develop until the individual reaches about 60 cm TL. Figure 10 shows that development reaches an asymptote between 85 and 90 cm TL. Even among individuals of the same size taken during the same month there is a high degree of varia- tion in ovarian length. For this reason ovarian length is not considered a good indicator of ma- turity in Atlantic sharpnose shark. Changes in the diameter of ovarian eggs were found to be a reliable indicator of the beginning of maturation. Figure 11 shows the first genera- tion of ovarian eggs produced by the subadult population. Increase in egg diameter begins at 60 to 65 cm TL, at about the same time the length of the ovary begins to increase. The eggs increase in diameter until the first ovulation, which oc- curs at about 85 to 90 cm TL. Most female sharp- nose sharks mature within this size range. Several female sharpnose sharks that had re- cently matured were examined. One individual of 88 cm TL, collected in late May, had full-sized ovarian eggs and had apparently recently mated due to the numerous mating scars that were ob- served in the region between the first and second I0-4T >- cr 88 I 7-3 + 5 57 + 2 6 o o o o I . • o o o o o o oo •••••• . ooo o • I o o o g oo o o o o o OBSERVED • REGRESSION ESTIMATED 50 60 70 80 90 TOTAL LENGTH fcrW 100 110 Figure 10.— Regression analysis showing development of the ovary in Atlantic sharpnose sharks, N = 42. Maturation is esti- mated to be complete at 85 to 90 cm total length. CT (li H ld < Q (a to < cr o 25 T 20 15- 10 • 1st OVULATION •-• •- • • -+- ■+- -+- 40 50 60 70 80 TOTAL LENGTH (cm) 90 100 Figure 11. — Maturation of female Atlantic sharpnose sharks as evidenced by the increase in ovarian egg diameter. Hand-fit curve approximates the increase in ovarian egg diameter from juvenile to first ovulation, N = 63. dorsal fins. An 86 cm individual, collected in early July, possessed six ova (8 to 10 cm), while another 89 cm female, collected in mid-July, pos- sessed uterine eggs. In late August, all mature females examined contained embryos. The small- est gravid specimens were 87, 88, and 89 cm TL and contained 11, 8, and 6 cm embryos, respec- tively. These observations further support the 85 to 90 cm estimated size at maturity. Mating Season Twenty-three reproductively active male At- lantic sharpnose sharks were examined to delin- eate the mating season. A gonadosomatic index (GSI), testis weight expressed as percent total body weight, was found to be the best indicator of mating season. The GSI provided a defined mating season for male sharpnose sharks (Fig. 12). Reporting on central gulf coast of Florida populations, Clark and von Schmidt (1965) suggested that small shark species (such as Mustelus norrisi and Scoli- odon terraenovae = Rhizoprionodon terraenovae) mate and bear young in the late winter and early spring. In the north central gulf, contrary to Clark and von Schmidt's findings for Florida, male sharpnose sharks appear to be reproduc- tively active during late spring and summer. From about September to March, the GSI was found to be low, 0.2 to 0.37. During these months specimens were observed to have reduced testes 68 PARSONS: REPRODUCTIVE BIOLOGY OF ATLANTIC SHARPNOSE SHARK 0.6t 0.5 0.4 o 0.3 < 5 0.2 0.1 I 2 HAND FIT CURVE Figure 12. — Mating season of adult male Atlantic sharpnose sharks as evidenced by the seasonal increase in gonadosomatic index (GSI). The data suggest that male sharpnose sharks are reproductively active during late spring and summer. The closed circles represent mean values and the numbers indicate sample sizes, N= 20. and no visible sperm or semen in the seminal vesicles. In late April the GSI had risen to 0.51, but there was little sperm present in the seminal vesicles. During mid- to late May the GSI aver- aged 0.47. All mature individuals had enlarged testes, turgid seminal vesicles, and copious amounts of sperm present in the claspers as evi- denced by microscopic examination. This condi- tion was found to persist through June and July with GSI equalling 0.59 and 0.57, respectively. Several adult males examined in August were found to have large quantities of sperm in the seminal vesicles. A single GSI determination in- dicated a slight decline from previous months. The mating season in female sharpnose sharks was evidenced by an increase in ovarian egg di- ameter (Fig. 13). From August to December the average egg diameter increased from ca. 3.0 to 4.2 mm. In almost every ovary examined during November and December, a few eggs were be- ginning to visually dominate the other oocytes. In February, the mean oocyte diameter equalled 5.0 mm, with some eggs reaching 11 mm. In Feb- ruary, all mature ovaries contained four to eight oocytes that were noticeably larger than sur- rounding eggs. From mid-February to late May or June, there was a rapid increase in egg diame- ter to about 20 mm at ovulation. The information indicates that the mating sea- son for male and female sharpnose sharks in the northern Gulf of Mexico coincides, although male sharpnose sharks are reproductively active earlier in the year. Assuming that females do not mate when gravid and that ovulations occur after copulation, then the mating season must occur between mid-May and mid-July. Most adult females still carried near-term embryos in mid-May, and by mid-July all females examined had uterine eggs. Considering the peak of partu- rition for gravid females (see Embryonic Growth and Development section), the subsequent ap- pearance of uterine eggs, and the occurrence of the first detectable embryos, the peak of mating most likely occurs from mid-June to mid-July. Embryonic Growth and Development Embryos representing various stages of devel- opment were weighed, sexed, and measured in total length. Conceptions were estimated to be at a peak in early to mid-July. At this time several sharpnose sharks that possessed recently ovu- lated uterine eggs but no visible embryos were examined. In late August, gravid females were collected, and they contained embryos ranging from about 4 to 11 cm TL. The smallest embryos examined were still dependent upon the yolk sac. They had prominent branchial gill filaments, undeveloped fins, and the anterior end was en- larged in relation to the rest of the body. Pratt (1979) suggested that growth of embryonic Prio- nace glauca is linear. Increase in length of sharp- nose shark's embryos approximates a sigmoid curve as evidenced by polynomial regression 25 1 IT LU < O O 19 LlI z < o 20- 15- 10 5- UTERINE • MEAN I RANGE •H-l Figure 13.— Mating season of adult female Atlantic sharpnose sharks as evidenced by the seasonal increase in ovarian egg diameter, N = 1,260. The data suggest that the mating season for females occurs from mid-June to mid-July. 69 FISHERY BULLETIN: VOL. 81. NO. 1 analysis (Fig. 14). After conception there is a pe- riod of rapid growth through the remainder of the summer and fall. By November the embryos have attained an average of 21.3 cm and appear almost completely developed. There is a notice- able inflection in the regression line in Novem- ber. The increase in length declines through the winter and spring months, although a slight in- crease may occur just before parturition in May or June. Pups are born at an average of about 32 cm TL. Skocik (1969) reported a total length of 25 cm for sharpnose shark at birth, and Bigelow and Schroeder (1948) stated that newborn sharp- nose sharks are generally about 275 to 400 mm long. The largest embryo recorded during the study period was 36 cm TL and the smallest free- living specimen was 32 cm. Increases in weight of the sharpnose shark's embryo differed from the increases in total length (Fig. 15). Embryo weight increased slowly during the period from estimated conception (mid-July) to October. Thereafter, however, until parturition in late May or June, an almost linear increase of about 16 g/mo occurred. Parturition occurs most likely between about 95 and 150 g. By using the above information, it was possible to estimate the gestation period. Atlantic sharp- nose shark's embryos require a 10 to 11 mo gesta- tion period, beginning in July or August and ending in May or June of the following year. Relationships Between Adult Females and Embryos A significant relationship was observed be- tween total length of the gravid female and the number of offspring produced. This is note- worthy since other works have failed to show such a relationship among carcharhinids (Springer 1960; Clark and von Schmidt 1965). Figure 16 shows that the total length of the adult is correlated with litter size ( ANOVA significant at <0.01). There is a direct relationship between fecundity and the size of the adult with the largest individuals being the most fecund. Grav- id females produce an average of 5 pups/litter per year (one to seven), but in most cases either four or six embryos will be present. It was anticipated that a relationship between litter size and embryo size could be detected. An optimal clutch size has been demonstrated in some species of birds (Lack 1954, 1966, 1968). Compared with small and large clutches, inter- mediate-sized clutches leave proportionately 36 5 28 2 199 11-6 o > a. 33 -4-9 1 * • REGRESSION ESTIMATE I RANGE Figure 14.— Growth of embryonic Atlantic sharpnose shark. Regression analysis shows the increase in embryo total length from fertilization to parturition, N = 300. 145-6 124-7- I03-8-- 829- x S2 Id S o £ 62-0 m 2 Id 412- 203- -0 6-»-« • REGRESSION ESTIMATE I RANGE -t 1 f -i 1 1 1 Figure 15. — Growth of embryonic Atlantic sharpnose shark. Regression analysis shows the increase in embryo weight from fertilization to parturition, N = 300. more offspring that survive to maturity. Birds from large clutches are smaller in size than birds from intermediate-sized clutches. After evalu- ating the data, an "optimal litter size" could not be demonstrated for the Atlantic sharpnose sharks. However, when the right and left uteri of adults collected during a single sampling trip (to cancel out seasonal differences) were treated separately, an inverse relationship was observed between the numbers of embryos per uterus and 70 PARSONS: REPRODUCTIVE BIOLOGY OF ATLANTIC SHARPNOSE SHARK oz 4 LU 3 2 • MEAN I RANGE N = 78 85-90 90-95 95-100 100-105 105-110 ADULT TOTAL LENGTH (cm) Figure 16.— Relationship between adult total length and litter size of the Atlantic sharpnose sharks. The plot indicates that fecundity increases significantly as adult total length increases (F = 9.216, P<0.00001). 24.1 + 22.9 I i- o z LU _l _l s 2 1.7- *~ 20.4 - o CE HI s LU 19.2 18. X • MEAN 1 RANGE EH 95% CONFIDENCE INTERVAL N = 300 I 3 4 No. PER UTERUS Figure 17.— Relationship between numbers of embryos per uterus and embryo total length of the Atlantic sharpnose sharks. Embryo total length decreases significantly with in- creasing number per uterus, AT = 89. embryo size (Fig. 17). The figure indicates that at the 95% confidence limits significant differ- ences exist between the total lengths of the em- bryos. Embryos were found to be largest when one or two are present per uterus. However, in only one case was there a single embryo found within a uterus. It is conceivable that mechanical "packing" within the uterus causes "intra-uterine competi- tion" for nutrients. As already discussed, in addi- tion to placentally derived nourishment, sharp- nose shark embryos may be able to absorb directly nutrients which are produced by the uterine epithelium. An increase in the number of embryos within the uterus above some optimal value might result in competition for this "uter- ine milk" and a decrease in embryo size. In sharpnose sharks, the parents that produce what might be termed an "optimal" number of embryos per uterus are producing the largest embryos. If we assume that these size differences are retained until birth, and thereafter, these larger embryos will result in progeny of highest individual fitness. Larger offspring cost more to produce, but they are also worth more (Pianka 1978). It would be interesting to examine the repro- ductive strategy of tropical sharpnose shark populations, since these sharks have been report- ed to have litters with as many as 12 embryos (Bigelow and Schroeder 1948; Skocik 1969). Based on this study, it would be a logical extra- polation to predict that these litters would result in smaller offspring. A litter of 12 must be approaching maximum fecundity for sharpnose sharks. Seasonal Distribution In this study it was determined that migratory behavior of the Atlantic sharpnose shark is pri- marily limited to an inshore-offshore movement. From late April to September of 1979, 93 sharp- nose sharks were collected from shallow inshore waters. During the period from late October 1979 to April 1980, despite numerous attempts, no sharpnose sharks were collected inshore. Sharpnose sharks may be encountered offshore year-round; however, the data indicate that the concentration of sharks is greatest during the fall and in particular, winter months. From Octo- ber 1979 to February 1980, 59 sharpnose sharks were collected during offshore longlining. Fig- ure 18 shows that the number of sharpnose sharks landed in deep water, as well as the catch per unit effort (CUE), is low in spring and sum- mer (CUE = 1.2 and 2.4, respectively) and in- creases to a high in winter (CUE = 7.3). The above data suggest that the migration from inshore to offshore begins around October or November. Atlantic sharpnose sharks appar- ently remain in deeper waters during the colder months and return inshore again in April and May. 71 FISHERY BULLETIN: VOL. 81, NO. 1 10 1 LjJ o 5 T 40 23. 14 12 IA ia. Li 2£l 30 ■20 •10 D to o z S z < Id co O z Q. < I co SPRING FALL Figure 18.— Catch per unit effort (CUE) in sharks/100 hooks per hour and number of Atlantic sharpnose sharks landed dur- ing longline operations. Ninety percent of these offshore land- ings were gravid females. Since adult female Atlantic sharpnose sharks were collected inshore only during summer months, the data suggest that females migrate inshore in late spring or summer to pup and mate, whereupon they return offshore again to overwinter. During June and July sharpnose shark pups with a fresh umbilical scar (in some cases the scar was actually an open slit) could be collected from the littoral zone. It is likely that special nursery areas exist for many shark spe- cies (Springer 1967), although the existence of specific pupping or nursery grounds for the At- lantic sharpnose sharks could not be conclusively established from this study. However, since new- born pups were never taken from deep waters in spite of intensive trawling, it is reasonable to suppose that the pups were born in shallow water. Perhaps the shallows of the northern Gulf of Mexico's extensive barrier island system serve as pupping/nursery grounds for the Atlantic sharpnose shark. Sex Ratio Sex of the Atlantic sharpnose sharks could be determined by clasper examination in embryos as small as 5.0 cm TL. The sex ratio through most of gestation could therefore be determined. The sex ratio early in development and of near-term embryos was found to be 1:1. One-hundred and fifty male and 155 female embryos were exam- ined. These data suggest that the sex ratio at par- turition is also 1:1. Among adults sampled, the sex ratio was found to be one sided in favor of females. During this study 33 adult male and 91 adult female sharpnose sharks were collected representing a 1:2.8 ratio. During offshore longlining 90% of the catch consisted of gravid adult female sharpnose sharks. This condition in sharpnose shark is not without precedent, as it has been observed in other shark species. Springer (1940), discussing Carcharhinus milberti and Carcharhinus ob- scurus, stated that in both species females out- number males. Clark and von Schmidt (1965) found a similar situation in Galeocerdo cuvieri. ACKNOWLEDGMENTS This research was funded by a fellowship awarded to the author by the Dauphin Island Sea Laboratory. Robert Shipp provided helpful as- sistance. Special thanks are due David Nelson for reviewing the manuscript, Snead Collard for supplying shark rodeo data, Benny Rohr for sup- plying NMFS data, and John Gourley for supply- ing data. I am also very grateful for the assistance pro- vided by the faculty and staff of the Dauphin Island Sea Laboratory. Likewise, I wish to thank the graduate students of the University of South Alabama, especially Rick Blaise, Steve Branstet- ter, Don Marley, and Austin Swift. LITERATURE CITED Bass, A. J., J. D. D'Aubrey, and N. Kistnasamy. 1975. Sharks of the east coast of southern Africa. III. The families Carcharhinidae (excluding Mustelus and Car- charhinus) and Sphyrnidae. South Afr. Assoc. Mar. Biol. Res., Invest. Rep. 38, 100 p. Baughman, J. L., and S. Springer. 1950. Biological and economic notes on the sharks of the Gulf of Mexico, with especial reference to those of Texas, and with a key for their identification. Am. Midi. Nat. 44:96-152. Bigelow, H. B., and W. C. Schroeder. 1948. Sharks. In A. E. Parr and Y. H. Olsen (editors), Fishes of the western North Atlantic, Part 1, p. 59-546. Mem. Sears Found. Mar. Res., Yale Univ. 1. Budker, P. 1971. The life of sharks. Columbia Univ. Press, N.Y., 222 p. Clark, E., and K. von Schmidt. 1965. Sharks of the central Gulf coast of Florida. Bull. Mar. Sci. 15:13-83. Compagno, L. J. V. 1978. Sharks. In W. Fischer (editor), FAO species iden- tification sheets for fishery management purposes; west- ern central Atlantic, Vol. 5, unpaginated. Lack, D. 1954. The natural regulation of animal numbers. Ox- 72 PARSONS: REPRODUCTIVE BIOLOGY OF ATLANTIC SHARPNOSE SHARK ford Univ. Press, Lond., 343 p. 1966. Population studies of birds. Oxford Univ. Press, Lond., 341 p. 1968. Ecological adaptations for breeding in birds. Methuen, Lond., 409 p. Lopez, A. M„ D. B. McClellan, A. R. Bertolino, and M. D. Lange. 1979. The Japanese longline fishery in the Gulf of Mexi- co, 1978. Mar. Fish. Rev. 41(10):23-28. Mahadevan, G. 1940. Preliminary observations on the structure of the uterus and the placenta of a few Indian elasmobranchs. Proc. Indian Acad. Sci., B, 11:1-44. Mann, T. 1960. Serotonin (5-hydroxytryptamine) in the male re- productive tract of the spiny dogfish. Nature (Lond.) 188:941-942. PlANKA, E. R. 1978. Evolutionary ecology. Harper and Row, N.Y., 397 p. Prasad, R. R. 1944. The structure, phylogenetic significance, and func- tion of the nidamental glands of some elasmobranchs of the Madras coast. Proc. Natl. Inst. Sci. India, Part B, Biol. Sci. 11:282-302. Pratt, H. L. 1979. Reproduction in the blue shark, Prionace glauca. Fish. Bull., U.S. 77:445-470. Radcliffe. L. 1916. The sharks and rays of Beaufort, North Carolina. Bull. U.S. Bur. Fish. 34:239-284. SCHLERNITZAUER, D. A., AND P. W. GILBERT. 1966. Placentation and associated aspects of gestation in the bonnethead shark. Spin/nut tiburo. J. Morphol. 120:219-231. Simpson, T. H., and C. S. Wardle. 1967. A seasonal cycle in the testis of the spurdog, Squalus acanthias, and the sites of 3/3-hydroxysteroid dehydro- genase activity. J. Mar. Biol. Assoc. U.K. 47:699-708. Skocik, R. 1969. The sharks around us. Star Publ. Co. Inc., 206 p. Springer, S. 1938. Notes on the sharks of Florida. Proc. Fla. Acad. Sci. 3:9-41. 1940. The sex ratio and seasonal distribution of some Florida sharks. Copeia 1940:188-194. 1950. A revision of North American sharks allied to the genus Carcharhinus. Am. Mus. Nat. Hist. Novit. 1451, 13 p. 1960. Natural history of the sandbar shark, Eulamia mtiberti U.S. Fish Wildl. Serv., Fish. Bull. 61:1-38. 1967. Social organization of shark populations. /« P. W. Gilbert, R. F. Mathewson, and D. P. Rail (editors), Sharks, skates, and rays, p. 149-174. John Hopkins Press, Baltimore. Springer, V. G. 1964. A revision of the carcharhinid shark genera Scoli- odon, Loxodon, and Rhizoprionodon. Proc. U.S. Natl. Mus. 115:559-632. WOURMS, J. P. 1977. Reproduction and development in chondrichthyan fishes. Am. Zool. 17:379-410. 73 VARIATION IN THE GROWTH RATE OF MYA ARENARIA AND ITS RELATIONSHIP TO THE ENVIRONMENT AS ANALYZED THROUGH PRINCIPAL COMPONENTS ANALYSIS AND THE co PARAMETER OF THE VON BERTALANFFY EQUATION Richard S. Appeldoorn 1 ABSTRACT Age-length data and environmental parameters were obtained for 25 populations of the soft-shell clam, Mija arenaria. Growth rates were analyzed for 20 of the populations and variations in the growth rates were related to differences in the environment. The analysis of growth was based on Gallucci and Quinn's w parameter for the von Bertalanffy equation. Environmental variability was analyzed, using principal components analysis which yielded three environmental factors: North- ness, siltiness, and sedimentary hydrocarbons. Growth was found to be significantly related to each of the three components. A distinct latitudinal growth relationship was observed, with growth de- creasing towards the north. Temperature, tidal height, tidal position, and edaphic conditions sys- tematically varied with latitude, with temperature being the dominant factor affecting growth. Growth was negatively correlated to both siltiness and sedimentary hydrocarbons. The growth of the soft-shell clam, Mya a renaria, has been studied by many investigators (Wilton and Wilton 1929; Belding 1930; Newcombe 1936; Swan 1952; Brousseau 1979; and others), and much work has been done in assessing the im- portance of various environmental factors in the growth process. These factors include water cur- rent and quality, food, temperature, salinity, various edaphic parameters, and pollution. In the past, investigators were obliged to study these factors individually even though it was realized that many were interrelated (Belding 1930). Because of local variations researchers often disagreed on the relative importance of each of these factors, and overall trends have not been firmly established. The purpose of this study was to investigate various factors contributing to growth rate vari- ations in soft clam populations and to demon- strate a methodology incorporating the analysis of multiple factors applicable to the above inves- tigation. Of specific interest was the demonstra- tion of a latitudinal trend in growth and the fac- tors responsible for it, since such a relationship had yet to be quantified (Brousseau 1979). Prin- cipal components analysis was used to analyze 'Graduate School of Oceanography, University of Rhode Island, Kingston, R.I.; present address: Department of Marine Sciences, University of Puerto Rico, Mayagiiez, Puerto Rico 00708. multivariate environmental data, and the von Bertalanffy model was used for the analysis of growth, using the recently introduced growth rate parameter w of Gallucci and Quinn (1979). This study represents one of the first applica- tions of w to investigate growth rate variations. MATERIALS AND METHODS Samples of Mya arenaria and environmental data were obtained from 25 sites located along the east coast of North America, from Maryland to Nova Scotia (Fig. 1). The sites were initially chosen and sampled as part of a study to investi- gate the relationship between environmental quality and neoplasia (Brown 1980), and as a re- sult 1) the sites varied greatly in their environ- mental quality, 2) the sampling design employed was not specifically designed for the present study, and 3) it was therefore necessary to use proxy data in some cases to represent certain en- vironmental characteristics. These drawbacks were not severely limiting, since the particular statistical techniques used could control much of the induced variability in the data. Estimates of the following environmental parameters were obtained: Salinity, tidal position, tidal range, average annual temperature, sedimentary grain size, dispersion and skewness of grain sizes, per- cent silt-clay, percent organic matter, and total sedimentary hydrocarbons. Manuscript accepted June 1982. FISHERY BULLETIN: VOL. 81, NO. 1, 1983. 75 FISHERY BULLETIN: VOL. 81. NO. 1 75°W 70° W 65 8 W 45° N - 40° N - r l" A / RS \ hzcr / PY \ ^k ^2r V £ NORTH AMERICA RB-~^^ BN, nbA WC\\ AH-^ SH ^^ SP-^\. \ : T \ gc \\vS PT Dl \ji/ / pi JL \ WK 3R y / oi * 02 \ WP TS — J _jZ U' A NR ATLANTIC OCEAN ^AR 1 l - 45°N - 40° N 75°W 70°W 65°W Figure 1.— Location of sampling sites. Site codes are given in Table 1. Salinity, at low tide, was measured by a refrac- tometer; tidal position was estimated on a scale of 0-1, where = subtidal and 1 = full exposure. Estimates of the average annual temperature near each site were obtained from various litera- ture sources, and estimates of the tidal range were obtained from National Ocean Survey (1978). Sediment samples (composites of two surface cores 21 cm 2 X 8 cm depth) were collected and analyzed to determine grain size distribution and organic content. The sand fraction was ana- lyzed by dry sieving; silt-clay by the hydrometer method (American Society for Testing Materials 1963). The particle size distributions obtained from the two analyses were pooled, and the cumu- lative frequency versus grain size () was plotted for each sample. From the graphs the following summary statistics were obtained: Median grain size (Md4>), quartile deviation (QD), and skew- ness (Skq) (Buchanan 1971). The results were reported in phi notation rather than millimeters [(f) = — log2(mm)], as this scale is commonly used to describe grain size characteristics and be- cause it allows for greater discrimination in the silt-clay range which may be more meaningful biologically. The percent organic matter was determined by measuring the percent weight loss of a small aliquot upon ignition at 550°C for 4 h (Buchanan 1971). Estimates for total sedimentary hydrocar- bons through infrared analysis were obtained from C. Brown. 2 The sites and their environmen- tal parameters are given in Table 1 along with their dates of collection, latitude, and code. Clams were sampled from one or more trenches dug by a standard clam hoe or shovel, with the exception of the Chesapeake Bay sites where a commercial hydraulic escalator dredge was used. All clams excavated were retained for analysis. For each individual, shell length (maximum shell dimension) was measured by vernier calipers to the nearest millimeter. Age structure was determined via length-fre- quency analysis for 19 of the populations. Simi- lar information for six of the sites (BN, WF, SP, GC, PY, JL) was available from Appeldoorn (1981), though only the West Falmouth (WF) growth data were used in subsequent analyses since major growth interruptions resulting from pollution events occurred at the other sites. 2 C. W. Brown, Professor, Department of Chemistry, Univer- sity of Rhode Island, Kingston, R.I., pers. commun. May 1979. 76 APPELDOORN: VARIATION IN GROWTH RATE OF MYA ARENARIA ccj u — c E c o S» '> c S- '53 -1-3 T3 C eS be c "a. E I w < E Q. to «- > O i_ c £ TO — CT <° O E # = JS Ed I- CL — 3?I CD a> _ d o> co t; ^ 5 = ra o co Q. — E .1 CD ~ a; q. 5 E Q co 2% w 8 i-'-^'T»-'-T-(DS(\jcoc\j(MO)oinsc\joifiu)ocoo)a) i-i-CMC\JCOCOCOi-CMi-C\JCOC\J^COC\JC\JCOC\JC\JC\ICOC\JCMCM i- o cm 0'-oina)(Dcjc\j 1—1- m^i-coi-m^-m Oi -^ -COCMCOC\ICO cbbcoinr-^moScob r-a)omtx)r^aoa> | tom<3)Oono)'-c\Ttrioco y- cvi i- ^ co W c\i ^ *t ' b i- cm b ^CONSincOCOlDCO '-cocotoommtm i-i-i-i-c\ji-i-b^-" tDO'-OJCDOJi-^S mc7)mm(\cvjijc\j(o i-i-Oi-OOOOCNJO I I I o o o m o m in m m in co m co co o o o b o ooinooooooinmmmmm mmcocMmmmcocococococococo ooooboboobbbooboobb f^r^coinojc\jc\jc\jincoco^-co^rr^inotnc\ji-i-cocococo ssococDiocpO)itc\j( , )incNini-sc\i^'itininit;'tc\ic\j ooT-^obbbi-i-i-i-i-i-i-i-cocococococbbi- • *— intninoooooh-omooocor^-^-cvii-mmoooo (NicNicocnwcNJwnir^^in^-incoinO'-'-'-t-'-'-Trit i- i- i- i- i- CM CNJ oooococococoino^j-inmmmi-i-^r»-i-i-r^cooo inincocowwcviwo^cNjdor-dr-c^sNsscoicDcTiai ^~ t— **"" ^~ t - t- r- i - T~ T~ ^- t- v— ^- i— ^- (0010000>0>- r- "-•-C>JCN*<*i(*)c*)OV(OVlDiDVlO , W .-.-.- .- .- WT-SC^NcOCOCOCONOrOOlCOCAJC^CDCOSCO^fOCDCOa) ininNinojcorocNjcONCNJcooincocncooNCDcoNOincD 0)oco^cococo^rininco(0- h ca ro "- 5 QC ~> - CL CO O CO o ~ C/) o « CO O CO z s c Z 5 en CD CD o CO -* - O a; O > 0)1 w 3 - r o t= o S 3 __ O lli m C- CD E cn O CO E o CO CL CD a s r 5 o D 2? E CO ^ "D CD CC C f~ CO O CN CD CD O) CD CO ^ CO ^> co^ „E CO o — CD i- C CM _-i2» 0) o> c o c -^ >• o = H o E ^ CD -^- c° 5ilu0ll10$ o m , m co tr co SM£M m - <-> *z — *i : ^ in co r» 77 FISHERY BULLETIN: VOL. 81, NO. 1 Length-frequency analysis was chosen because it could be applied to all samples, thus facili- tating the comparison between samples. The use of shell annuli is unreliable south of Cape Cod (Mead and Barnes 1904; Shuster 1951), and MacDonald and Thomas (1980) found little sup- port for the technique in a Prince Edward Island population. Constraints on the sampling design precluded mark-recapture methods. For each population the modes on a length-fre- quency histogram were broken down into a se- ries of normal curves (Tesch 1971; MacDonald and Pitcher 1979) by a Dupont 3 310 Curve Re- solver, an analog computer which allows one to break down a complex distribution into its basic components in a graphical fashion (Appeldoorn 1981). From the resulting graphs the mean and standard deviation of the curve which represents each mode of the histogram can be obtained. The curve resolver also determines the percentage of the whole sample under each curve. Length-frequency analysis assumes that spawning and settlement are discrete relative to growth such that the length distributions of co- horts are separable. Ropes and Stickney (1965), Pfitzenmeyer (1962), and Brousseau (1978) found that periods of both spawning and settlement of each cohort were discrete events. In the latter study, closely spaced cohorts within the same year were separable by length-frequency analy- sis using probability paper. In the present study, discrimination of cohorts within a year class was also possible. By inspection of the histograms and subse- quent age-length curves and through considera- tion of local recruitment processes and sampling efficiency, ages were assigned to each cohort (Brothers 1980; Schnute and Fournier 1980). When possible, results were corroborated by comparing them with previously published age- length data for the same or nearby areas (e.g., Belding 1930; Pfitzenmeyer 1972; Mead and Barnes 1904; Gilfillan and Vandermuelen 1978; Brousseau 1979), by comparison of adjacent areas (e.g., the two Quonochontaug Pond sites), by comparison of multiple samplings (Allen Har- bor, Deer Isle), and by counts of shell annuli (Portland, Deer Isle). The ages assigned were relative rather than absolute; the time beyond the last yearly incre- ment represents the fraction of expected yearly growth already obtained (Appeldoorn 1981). This process results in a smoother growth curve, since it linearizes seasonal growth variations which would otherwise necessitate the use of a more complex growth model (Cloern and Nichols 1978). The analysis of growth differences can be sim- plified by comparing model parameters rather than the direct age-length observations (Rao 1958). Growth was modeled by fitting the von Bertalanffy growth function (VBGF) to the age- length data. The VBGF is described by the equa- tion: L, = U (1 _ -ftt-to) 3 Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. where t = time, L, = length at time t, L^ = maxi- mum asymptotic length, K = growth constant, and to = time when L, = 0. The single growth parameter of Gallucci and Quinn ( 1979) is obtain- ed by w = KL X . Recent studies on the statistical comparison of VBGF's (Allen 1976; Bayley 1977; Gallucci and Quinn 1979; Kimura 1980; Misra 1980; Kappen- man 1981) and on the VBGF's biological basis (Pauly 1979, 1981) have removed most of its past criticism (Roff 1980). Dickie (1971) considered the VBGF applicable for modeling population growth even when individual growth did not fit the model. The VBGF has been previously ap- plied to Myo armaria by Munch-Petersen(1973), Brousseau (1979), and Brethes and Desrosiers (1981). The co parameter was chosen for analysis be- cause, as a single parameter, it was easily calcu- lated, tractable to further analysis, statistically comparable, interpretable in both a biological and statistical sense, and more robust than either Kor L^ (Gallucci and Quinn 1979). A major bene- fit of applying the VBGF is that only estimates of length at known time intervals are required to determine K, L tt , and hence co. Absolute age at length is only required to estimate to. However, to is of less importance here, since it is not a mea- sure of growth, but only a location parameter. The VBGF was fitted to the data according to the methods of Gallucci and Quinn (1979), using the NLIN procedure of SAS79 (Helwig and Council 1979) which yielded estimates of the parameters, their asymptotic standard errors, and the correlation coefficient of /f and L x . From these estimates the co parameter and its variance were calculated (Gallucci and Quinn 1979). The regression procedure incorporated the size and 78 APPELDOORN: VARIATION IN GROWTH RATE OF MYA ARENARIA variance of each age mode. Therefore, variation in the original data is reflected in the variance estimates of the model parameters, and poorly represented age modes, where estimates of mean length and variance might be subject to error, are weighted less. The resulting growth curves are based on the assumption that growth varies from year to year only to the extent expected owing to normal fluctuations in growing condi- tions. Hence, they are an estimation of "average" growth within a population, representing an in- tegration of several variable processes affecting growth. The environmental data listed in Table 1 were used to characterize Mya armaria habitats. These data were subjected to principal compo- nents analysis (PCA) to reduce the observed var- iables to a more meaningful and manageable number of factors without excessive loss of infor- mation. PCA locates hidden components which have generated dependence in the observed var- iables (Morrison 1976). Each resulting compo- nent is a composite variable — a linear combina- tion of the original variables. The components are independent and ordered, so that the first component accounts for most of the observed var- iation, the second for most of the residual varia- tion, and so on. The loadings given for each com- ponent represent the correlation coefficient (r) between a variable and a component. The analy- sis was run on the Pearson product-moment cor- relation matrix of the environmental parameters (to allow for standardization of the units of mea- sure) by using the CORR, FACTOR, and SCORE procedures of SAS79 (Helwig and Council 1979). The components produced by PCA are limited by the input data and can only reflect the factors represented by those data. In the present study the selection of factors was constrained by the sampling design, and no direct measurements were made on a number of factors which would be expected to influence growth (e.g., current flow, food concentration). However, several of the factors represent an integration of processes, incorporating factors not measured directly. For example, current flow is represented to some de- gree by tidal range, tidal position, and sediment characteristics (see Discussion). This integration effect will help offset the limitations of the input data. The growth rate parameter was transformed to log 10 (co) for the analysis of growth variations. Since logio(ZO and logio(L J are inversely propor- tional (Pauly 1979), it isfeltthatlogio(aj)isamore suitable measure of growth (Appeldoorn in press). A difference in logio(co) would then indi- cate a fundamental difference in growth — not just a reciprocal change in Kand L^. [See Pauly 1979, 1980 for a discussion of the analogous P = \ogw{K-W x ) parameter of the VBGF for weight.] Variations in growth rate were analyzed using a stepwise functional regression of logio(a;) on the components generated by PCA, where the resid- uals of the regression of the logio(ou) on Compo- nent 1 were regressed against Component 2 and so on. The geometric mean functional regression was deemed appropriate because of variability in both ai and the components, small sample size, and uncertainties about the distribution of the data (Ricker 1973; Laws and Archie 1981). In normal predictive regressions the regression co- efficient (slope) is b\ functional regression yields a coefficient of v — b/r where r is the correlation coefficient. The standard error of v (SE,) equals the standard error of 6(SE/,) and 95% confidence limits on rare approximated by v± 2SE, (Ricker 1973). Estimates of b, r 2 , and SE* were obtained using the GLM procedure of SAS79 (Helwig and Council 1979) and used to calculate v and its 95% confidence limits. The significance of the regres- sion is tested by determining if the confidence limits bracket v = 0. If not, the null hypothesis Ho: v = is rejected. RESULTS The mean lengths at age as determined through length-frequency analysis are given in Appendix Table 1 for the 19 populations analyzed here. The parameters of the VBGF and logio(to) are given in Table 2. Using the 95% confidence limits around logmM, statistically significant growth differences become readily apparent. A functional regression of logioMon latitude yield- ed: logio(to) = 4.8184 - 0.0878 latitude with r = 0.8220. Although the regression accounts for the majority of the observed variation in growth, it does not indicate what underlying processes may be responsible for this relationship. The results of the PCA are shown in Table 3. The terms used in the table follow the definitions in Morrison ( 1976). In order to simplify the table, those loadings <0.30 have been left out, although all variables contribute to all components to some degree. The first five components have been retained and account for 88% of the observed variation. Of these, the first three were exam- ined in greater detail. 79 FISHERY BULLETIN: VOL. 81. NO. 1 Table 2.— Estimates and standard errors for the von Bertalanffy constants. Site code K L^ to Logio(w) + 95% confidence interval TS AR NR RB WP Q1 Q2 SR WK CR AH WF NB EG WC PT Dl SH RS PI 0.2530 0.2740 3016 0.1829 02992 0.1175 0.1069 0.2119 0.1811 0.1997 0.0903 0.0917 0.1532 0.1377 0.1411 1468 0.1255 00565 0.1623 00986 (00597) (0.0520) (0.0162) (00986) (0.0114) (0.0194) (0.0134) (0.0229) (0 0155) (0.0114) (0.0184) (0.0162) (00198) (0 0425) (00246) (0.0077) (0.0114) (00083) (00287) (0 0248) 111.05 107.13 7969 81.50 73.27 93.23 1 1 1 .00 72.34 111 80 97.75 113.20 13673 8928 91.95 87 18 67.91 67 96 135.71 73.13 81 55 (1118) ( 7.22) ( 1.10) (22.08) ( 0.89) ( 9.20) ( 696) ( 2.48) ( 4.21) ( 1.60) (13.11) (1488) ( 4.30) (18.20) ( 7.37) ( 1.39) ( 2.46) (12.34) ( 4.52) (1078) -1.188 -1.440 -0.718 -1.450 -0 400 -1.104 -1.205 -0.445 -0436 -0 990 -1.668 -1.357 -1.571 -0.914 -1.549 0.836 -0.781 -0980 -0.745 -0.171 (0263) (0.268) (0.095) (0.558) (0.058) (0.148) (0.191) (0.225) (0.127) (0.143) (0.288) (0.184) (0.304) (0.186) (0.236) (0.122) (0.218) (0.336) (0.434) (0.432) 1.4486 1.4677 1 3808 1.1734 1.3418 1.0396 1.0743 1.1855 1.3066 1.2905 1.0095 1 0982 1.1360 1.1025 1 .0899 9986 0.9311 08847 1.0754 09053 1 3839 1.4166 1.3473 1.1202 1.3253 09848 1.0045 1.1417 1.2724 1.2512 0.9147 1 0056 1.0902 1 0439 09965 09778 08974 07663 1.0275 07839 1.5050 1.5134 1.4119 1.2207 1 3577 1.0882 1.1344 1.2253 -1.3383 1.3265 -1.0873 -1.1746 -1.1774 -1.1541 1.1668 1.0186 -0.9623 •0.9776 1.1167 09674 Table 3.— Results of the principal components analysis on environmental data. Loadings <0.30 have been omitted for clarity. Environmental Principa components Commu- parameter 1 2 3 4 5 nality Average temperature -0 938 0925 Tidal range 0.806 383 0.819 Tidal position 0817 0.855 Md0 -0.503 0.725 0.891 QD0 0.808 -0.382 0872 Skq0 0.386 -0.858 0.933 % silt-clay 609 0.675 0880 % organic matter 0.521 0.765 0.942 Total hydrocarbons 0.802 0855 Salinity 0.396 0.750 0.840 Eigenvalues 3 588 1.875 1.363 1.134 0.851 % variance 35.9 18.7 13.6 11.3 8.5 % cumulative variance 35.9 546 68 3 796 88.0 The first component is interpreted as repre- senting latitude, since the major contributing variables vary with latitude. Average annual temperature, as might be expected, shows the highest correlation. It decreases with latitude. To avoid confusion the first component will be re- ferred to as "northness." The second component is sediment siltiness. Grain size (negatively cor- related) and percent silt-clay (positively cor- related) are the main contributing variables. The high correlation of salinity may reflect the role of flocculation and estuarine circulation in the distribution of silts and clays in estuarine sediments (Krumbein and Sloss 1963; Knauss 1978). The third component, positively correlated with hydrocarbons and percent silt-clay, is sedi- mentary hydrocarbons. The higher silt-clay com- ponent (also reflected to some degree by positive skewness) provides a greater sedimentary sur- face area for the retention of hydrocarbons (Ly tie and Lytle 1977). The first three components were used for the further analysis of growth to try and deduce fac- tors which could have contributed to the latitudi- nal trend and to point out secondary growth affecting factors. Several sites were omitted from this analysis because missing values pre- cluded the calculation of the component scores. The results of the stepwise regression analysis are given in Table 4. As expected, growth was found to be negatively correlated with northness. The second regression showed a negative rela- tionship between siltiness and growth. The last Table 4.— Results of the stepwise regression of growth (w) on the first three principal components. The slope of the predictive regression (6) can be found by 6 = vr. Regression Intercept v = slope Approximate 95% confidence limits Log,o(w) vs. northness 0.693 1st residual vs. siltiness 0.184 2d residual vs sedimentary hydrocarbons 0217 1.1137 -0.1653 -0.2269 parameter proved useful, both for its de- scription of growth rate and ease of manipula- tion in further analyses. Incorporating both the sample size and variance of the age-length deter- minations into the nonlinear regression protected oj against random errors in the location of the age-length modes. In addition, since w is more robust than either Kor L x , it is further protected against inaccuracies in their estimation — an advantage over using K during the subsequent growth analyses. The trends resulting from the subsequent regressions were also protected from random inaccuracies in the estimation of to by virtue of the large number of sample sites used. However, it is important to recognize that such random errors do exist and not to continue the analysis past its potential limits. ACKNOWLEDGMENTS The author wishes to express his gratitude to those people who assisted in the clam collection, particularly R. L. Dow, M. Richards, J. M. Hick- ey, M. L. H. Thomas, and M. Worobec; and to those people who were involved in the sample processing, notably R. S. Brown, J. Keller, A. Miller, S. Polofsky, C. Brown, and E. Franklin. S. B. Saila, A. N. Sastry, R. C. Bullock, and two journal reviewers gave helpful criticism. Finan- cial support for sample collection was provided by the American Petroleum Institute. LITERATURE CITED Allen, R. L. 1976. Method for comparing fish growth curves. N.Z. J. Mar. Freshwater Res. 10:687-692. American Society for Testing Materials. 1963. Grain-size analysis of soils. D422-63. A.S.T.M. Standards 11:205-216. Appeldoorn, R. S. 1981. Response of soft-shell clam (Mya arenaria) growth to onset and abatement of pollution. J. Shellfish Res. 1: 41-49. In press. Relationships between the parameters of the von Bertalanffy growth function among bivalves as re- vealed with the auximetric grid. [Abstr.] J. Shellfish Res. Bayley, P. B. 1977. A methodfor finding the limits of application of the von Bertalanffy growth model and statistical estimates of the parameters. J. Fish. Res. Board Can. 34:1079- 1084. Beaven, G. F. 1960. Temperature and salinity of surface water at Solo- mons, Maryland. Chesapeake Sci. 1:2-11. Belding, D. L. 1930. The soft-shelled clam fishery of Massachusetts. Mass. Dep. Conserv., Div. Fish. Game, Mar. Fish. Sect., Mar. Fish. Ser. 1, 65 p. Brethes, J.-C. F., and G. Desrosiers. 1981. Estimation of potential catches of an unexploited stock of soft-shell clam (Mya arenaria) from length com- position data. Can. J. Fish. Aquat. Sci. 38:371-374. Brothers, D. T. 1980. Age and growth studies on tropical fishes. In S. B. Saila and P. M. Roedel (editors), Stock assessment for tropical small scale fisheries, p. 119-136. Int. Cent- Mar. Resour. Dev., Kingston, R.I. Brousseau, D. J. 1978. Spawning cycle, fecundity, and recruitment in a population of soft-shell clam, Mya arenaria, from Cape Ann, Massachusetts. Fish. Bull., U.S. 76:155-166. 1979. Analysis of growth rate in Mya arenaria using the von Bertalanffy equation. Mar. Biol. (Berl.) 51:221- 227. Brown, R. S. 1980. The value of the multidisciplinary approach to re- search on marine pollution effects as evidenced in a three-year study to determine the etiology and patho- genesis of neoplasia in the soft-shell clam, Mya arenaria. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 179:125-128. Buchanan, J. B. 1971. Sediments. In N. A. Holme and A. D. Mclntyre (editors). Methods for the study of marine benthos, p. 30- 52. Blackwell Sci. Publ., Oxford. Cloern, J. E., and F. H. Nichols. 1978. A von Bertalanffy growth model with a seasonally varying coefficient. J. Fish. Res. Board Can. 35:1479- 1482. Dickie, L. M. 1971. Addendum: Mathematical models of growth. In W. E. Ricker (editor). Methods for assessment of fish production in fresh waters, p. 126-130. Blackwell Sci. Publ., Oxford. Dow, R. L. 1975. Reduced growth and survival of clams transplant- ed to an oil spill site. Mar. Pollut. Bull. 6:124-125. Dow, R. L., and J. W. Hurst, Jr. 1975. The ecological, chemical and histopathological evaluation of an oil spill site. Part I. Ecological studies. Mar. Pollut. Bull. 6:164-166. Dow, R. L., and D. E. Wallace. 1961. The soft-shell clam industry of Maine. U.S. Fish Wildl. Serv., Circ. 110, 36 p. Gallucci, V. F., and T. J. Quinn II. 1979. Reparameterizing, fitting, and testing a simple growth model. Trans. Am. Fish. Soc. 108:14-25. Gilbert, M. A. 1973. Growth rate, longevity and maximum size of Ma- coma balthica (L.). Biol. Bull. (Woods Hole) 145:119- 126. Gilfillan, E. S., D. Mayo, S. Hanson, D. Donovan, and L. C. Jiang. 1976. Reduction in carbon flux in Mya arenaria caused by a spill of No. 6 fuel oil. Mar. Biol. (Berl.) 37:115-123. Gilfillan, E. S., and J. H. Vandermeulen. 1978. Alterations in growth and physiology of soft-shell clams, Mya arenaria, chronically oiled with Bunker C 82 APPELDOORN: VARIATION IN GROWTH RATE OF MYA ARENARIA from Chedabucto Bay, Nova Scotia, 1970-76. J. Fish. Res. Board Can. 35:630-636. Helwig, J. T., and K. A. Council. 1979. SAS user's guide, 1979 ed. SAS Inst. Inc.. Ra- leigh. N.C., 494 p. HiCKS, S. D. 1963. Physical oceanographic studies of Narragansett Bay, 1957 and 1958. U.S. Fish Wildl. Serv., Spec. Sci. Rep.— Fish. 457, 30 p. Jeffries, H. P. 1962. Environmental characteristics of Raritan Bay, a polluted estuary. Limnol. Oceanogr. 7:21-31. Kappenman, R. F. 1981. A method for growth curve comparisons. Fish. Bull., U.S. 79:95-101. Kellogg, J. L. 1905. Conditions governing existence and growth of the soft clam (Mya armaria). Rep. Comm. year ending June 30, 1903, 29:195-224. KlMURA, D. K. 1980. Likelihood methods for the von Bertalanffy growth curve. Fish. Bull., U.S. 77:765-776. Knauss, J. A. 1978. Introduction to physical oceanography. Prentice- Hall, Inc., Englewood Cliffs, N.J., 338 p. Krumbein, W. C, and L. L. Sloss. 1963. Stratigraphy and sedimentation. W. H. Freeman, San Franc, 660 p. Laws, E. A., and J. W. Archie. 1981. Appropriate use of regression analysis in marine biology. Mar. Biol. (Berl.) 65:13-16. Lytle, J. S., AND T. F. Lytle. 1977. Sediment hydrocarbons as environmental indi- cators in the northeast Gulf of Mexico. In D. A. Wolfe (editor), Fate and effects of petroleum hydrocarbons in marine ecosystems and organisms, p. 404-412. Per- gamon Press, N.Y. MacDonald, B. A., and M. L. H. Thomas. 1980. Age determination of the soft-shell clam Mya are- naria using shell internal growth lines. Mar. Biol. (Berl.) 58:105-109. MacDonald, P. D. M., and T. J. Pitcher. 1979. Age-groups from size-frequency data: A versatile and efficient method of analyzing distribution mixtures. J. Fish. Res. Board Can. 36:987-1001. Marine Research. 1975. Charleston study annual report. April 1974 - March 1975. Mar. Res. Inc., Falmouth, Mass., 241 p. Mead, A. D., and E. W. Barnes. 1904. Observations on the soft-shell clam (fifth paper). R.I. Comm. Inland Fish., 34th Annu. Rep., p. 29-68. MlSRA, R. K. 1980. Statistical comparisons of several growth curves of the von Bertalanffy type. Can. J. Fish. Aquat. Sci. 37:920-926. Morrison, D. F. 1976. Multivariate statistical methods. 2d ed. Mc- Graw-Hill, Inc., N.Y., 415 p. Munch-Petersen, S. 1973. An investigation of a population of the soft clam (Mya arenaria L.) in a Danish estuary. Medd. Dan. Fisk. Havunders., New Ser. 7:47-73. National Ocean Survey. 1978. Tide tables 1979: High and low water predictions. East coast of North and South America, including Greenland. Natl. Oceanic Atmos. Admin., Natl. Ocean Surv., Rockville, Md„ 293 p. Newcombe, C. L. 1936. A comparative study of the abundance and the rate of growth of Mya arenaria L. in the Gulf of St. Lawrence and Bay of Fundy regions. Ecology 17:418-428. Newcombe, C. L., and H. Kessler. 1936. Variations in growth indices of Mya arenaria L. on the Atlantic coast of North America. Ecology 17: 429-443. Pauly, D. 1979. Gill size and temperature as governing factors in fish growth: a generalization of von Bertalanffy's growth formula. Berichte aus dem Institut fur Meere- skunde 63, 156 p. 1980. A new methodology for rapidly acquiring basic in- formation on tropical fish stocks: growth, mortality, and stock recruitment relationships. In S. B. Saila and P. M. Roedel (editors), Stock assessment for tropical small scale fisheries, p. 154-172. Int. Cent. Mar. Resour. Dev., Kingston, R.I. 1981. The relationships between gill surface area and growth performance in fish: a generalization of von Ber- talanffy's theory of growth. Meeresforschung 28:251- 282. Pfitzenmeyer, H. T. 1962. Periods of spawning and setting of the soft-shell clam, Mya arenaria, at Solomons, Maryland. Chesa- peake Sci. 3:114-120. 1972. Tentative outline for inventory of molluscs: Mya arenaria (soft-shell clam). Chesapeake Sci. 13(suppl.): S182-S184. Rao, C. R. 1958. Some statistical methods for comparison of growth curves. Biometrics 14:1-17. RlCKER, W. E. 1973. Linear regressions in fishery research. J. Fish. Res. Board Can. 30:409-434. Ropes, J. W., and A. P. Stickney. 1965. Reproductive cycle of Mya arenaria in New Eng- land. Biol. Bull. (Woods Hole) 128:315-327. ROFF, D. A. 1980. A motion for the retirement of the von Bertalanffy function. Can. J. Fish. Aquat. Sci. 37:127-129. Sameoto, D. D. 1972. Yearly respiration rate and estimated energy bud- get for Sagitta elegans. J. Fish. Res. Board Can. 29: 987-996. Schnute, J., and D. Fournier. 1980. A new approach to length-frequency analysis: Growth structure. Can. J. Fish. Aquat. Sci. 37:1337- 1351. SHOREY, W. K. 1973. Macrobenthic ecology of a sawdust-bearing sub- strate in the Penobscot River estuary (Maine). J. Fish. Res. Board Can. 30:493-497. Shuster, C. N. 1951. On the formation of mid-season checks in the shell of Mya. [Abstr.] Anat. Rec. 111:543. Swan, E. F. 1952. The growth of the clam Mya arenaria as affected by the substratum. Ecology 33:530-534. Tesch, F. W. 1971. Age and growth. In W. E. Ricker (editor), Meth- ods for assessment of fish production in fresh waters, p. 98-130. Blackwell Sci. Publ., Oxford. 83 FISHERY BULLETIN: VOL. 81, NO. 1 Thomas, M. L. H. 1978. Comparison of oiled and unoiled intertidal com- munities in Chedabucto Bay, Nova Scotia. J. Fish. Res. Board Can. 35:707-716. Turner, H. J., Jr. 1948. Report on investigations of the propagation of the soft-shell clam, Mya arenaria. Rep. Invest. Shellfish. Mass. 1:3-9. Welch, W. R. 1981. Monthly and annual means of sea surface tempera- ture: Boothbay Harbor, Maine 1905 through 1980. Maine Dep. Mar. Resour., Res. Ref. Doc. 81/8. Wilton, M. H., and H. I. Wilton. 1929. Conditions affecting the growth of the soft shell clam, Mya arenaria L. Contrib. Can. Biol. Fish. 4:81- 93. Appendix Table 1. — The age (yr), length ±1 standard deviation (mm), and percent of sample in each age class. Percentages may not total 100 due to roundoff or exclusion of some outliers from the analysis. Sample size is given in parentheses. — = undefined. % of % of % of Age Length sample Age Length sample Age Length sample Navesink River (103) Coonamessett River (124) Portlanc (367) 0.67 26.012.3 3 1.15 340 — 0.5 1.5 19010.8 1 1 33 35.2+25 14 2.15 42.8+2.0 4 2.5 26.711.3 4 1.67 42.5+2.0 23 285 52.011.0 7 3.5 31.511.6 17 2.33 47.311.8 18 3.15 56.112.0 7 4.5 37.111.6 21 2.67 51.8+1.4 10 4 15 64.011.4 18 5.5 41.311.6 29 3.33 55.311.4 11 5.15 69.411.2 9 6.5 44.511.2 10 3.67 59.611.6 2 600 72.811.2 16 7.5 48.111.2 15 4.50 62.1 + 1.4 2 7 15 77.411.2 18 95 52.710.7 3 5.50 65.411.2 2 8.15 82.311.4 17 10.5 55.010.5 1 6.50 69.711.5 4 10.15 88211.1 5 11.5 56.710.4 1 8.50 74.711 2 4 11.15 94 - 0.5 13.5 60 210.5 1 9.50 78.011 2 6 Quonochontaug Pond-2 (146) Saugatucket River (140) Deer Isle (318) 2.15 33.712.7 41 20 29 011.3 2 3.0 25.711.9 4 3.15 41.6+2.0 28 26 33.810.9 5 4.0 31.211.1 7 4.15 49.0+1.3 6 3.0 37.611.6 9 5.0 35.211.3 21 5.15 55.111.8 6 3.8 43.312.2 42 6.0 39 2+1 3 18 6.15 60.211 4 4.8 48.911.5 7 7.0 42.1 + 1 1 9 7.15 65811.3 6 58 52.811.1 12 8.0 44.611.0 10 8.15 70.510.8 4 7.0 56.110.7 6 9.0 47.8±1.0 12 9.15 73.510.8 3 8.0 60 011.2 6 10.0 51.5+1.0 9 10.15 79 011.1 4 9.0 64.511.2 3 Quonochontaug Pond-1 (198) Allen Harbor (144) Potato I sland (201) 1.33 23.9+2.1 33 1.85 30 011.8 11 2.5 25.310.7 2 2.33 30.7+2.1 43 2.85 37.712.6 24 3.5 30.911.3 8 3.33 38.311.9 13 3.85 44.511.5 15 4.5 34.611.6 25 4.33 44.911.3 7 4.85 50.411.8 18 5.5 39 211.5 41 5.33 49.611.2 2 5.85 55.211.3 14 6.5 43 7±1.0 13 6.33 54.211.4 3 685 59.711.4 8 7.5 47 110.8 6 733 58.611.4 1 7.85 65.811.4 9 8.5 49.410.9 3 Watchemoket Cove (90) Big Annemessex River (177 Robinston (190) 1.15 27 612.9 24 1.33 57.011.3 14 3.67 37.911.6 8 2.15 34.812.0 31 1.80 60 911.8 30 4.67 42 511.6 16 3.15 42.4±1.7 19 2.33 688+2.0 33 5.67 47.211.6 41 4.15 48 212 5 11 280 737+0.8 4 6.67 51 711.0 20 6.15 57.111.4 8 3.33 77.211.3 7 7.67 54.610.9 7 7 15 62.511.0 6 4.33 86 811.5 2 8.67 57 011.2 4 New Bedford (180) Stockton Harbor (164) Tangier Sound (166) 1.85 32 011.0 05 3.0 31.011.8 2 1.33 546±2.3 36 2.85 43.711.2 3 40 38.211.8 2 1.80 62.011.6 26 3.56 48.311 6 8 5.0 44.8+1.4 11 2.33 67.911.2 13 3.85 51.210.8 15 6.0 49211.1 13 2.80 73.411.3 11 4.56 53 910.9 16 7.0 54.1 ±1.5 17 3.33 78.210.9 3 4.85 56 .210.9 21 8.0 58 211.2 21 3.80 81.610.8 1 5.56 58911.2 16 90 62 811 3 13 4.33 86.210.7 2 5.85 61.210 8 11 10.0 66.511.4 11 6.85 64.410.9 6 11.0 70.311.2 3 Raritan 3ay (200) 7.85 69.010.6 3 12.0 75.011.3 5 1.33 30410.8 3 Winnapau 3 Pond (229) Wickford (203) 1.67 36.411.8 23 09 24.211.6 3 0.20 7.0 — 05 2.33 40.211.7 39 1.5 31.511.0 2 2.00 37.311.1 2 2.67 43811.6 25 19 37.411.2 5 2.67 48.711.4 4 3.33 47 410.8 10 25 41.4+1.2 7 3.00 53.911.4 10 2.9 46.211.3 14 380 60411.8 18 East Greenwich Cove (192) 3.5 50.611.8 24 4.80 68 1 ±2.5 36 1.0 20.813.5 30 4.5 56.0+1.6 29 580 75.1 + 1.8 13 2.0 30811.0 9 5.5 60.511.2 8 600 80211.1 7 3.0 38.412.8 49 6.5 64.111.4 4 7.00 84.710.9 2 4.0 45.011.5 11 7.5 67.210.7 4 800 885106 2 5.0 50.911.1 3 84 BIOCHEMICAL GENETICS OF PACIFIC BLUE MARLIN, MAKAIRA NIGRICANS, FROM HAWAIIAN WATERS 1 James B. Shaklee, 2 Richard W. Brill, 3 and Robin Acerra 4 ABSTRACT An electrophoretic survey of 35 enzyme-coding gene loci in Pacific blue marlin was accomplished to determine levels of genetic variation and the feasibility of using electrophoresis to study stock structure in this species. Polymorphism (P99) in the marlin was 0.26 and the average heterozygosity (H) was 0.06. Allele frequencies at 11 variable loci were determined for a sample of 95 fish from Kona, Hawaii. The observed levels of polymorphism and heterozygosity suggest that a biochemical genetic analysis of blue marlin stock structure is possible and may reveal stock heterogeneity. The Pacific blue marlin, Makaira nigricans, is the predominant billfish species in the central tropical Pacific. As such, it is an important com- mercial species and the object of a considerable sport fishery. The average annual catch of this species in the Pacific exceeds 14,000 t (metric tons) (Shomura 1980). The Pacific blue marlin is primarily distributed in equatorial areas, al- though Japanese longliner catch records indi- cate that its range extends from lat. 48°N to 48°S. During the Southern Hemisphere summer (December through March) a center of concen- tration occurs in the western and central South Pacific (between lat. 8°S and 26°S). In the North- ern Hemisphere summer (May through October) a center of concentration occurs in the central North Pacific (between lat. 2°N and 24 °N). Dur- ing April and November the fish appear to be concentrated equatorially between lat. 10°N and 10°S (Rivas 1975). There is currently no direct evidence of migration of blue marlin within the Pacific. However, a general movement to the northwestern Pacific during the Northern Hem- isphere summer and to the southeastern Pacific during the Southern Hemisphere summer has been postulated by Howard and Ueyanagi (1965) 'Contribution No. 652 from the Hawaii Institute of Marine Biology. 2 Hawaii Institute of Marine Biology and Department of Zool- ogy, University of Hawaii, Honolulu, Hawaii; present address: Division of Fisheries Research, CSIRO Marine Laboratories, 233 Middle Street, Cleveland, Queensland 4163, Australia. :i Pacific Gamefish Foundation, Honolulu, Hawaii; present address: Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 3830. Honolulu, HI 96812. 4 Hawaii Institute of Marine Biology, University of Hawaii. Honolulu, Hawaii; present address: Southampton College, Long Island University, Southampton, NY 11968. on the basis of the shifting abundance patterns of the fish. Little is known about spawning, other than that Pacific blue marlin appear to spawn throughout the year in an area 10°-20° on either side of the Equator, and up to 30° on either side of the E qua- tor during the Northern and Southern Hemis- pheres' respective summer months. In general, the highest spawning densities occur in the west- ern Pacific, with the density decreasing eastward (Strasburg 1970; Matsumoto and Kazama 1974; Rivas 1975). Because of the apparently single equatorial Pacific spawning area, it has been as- sumed that the species consists of a single unit stock (Yuen and Miyake 1980; Yoshida 1981), yet there has been no direct test of this assumed stock structure. The most recent report available on the condition of the Pacific blue marlin stock considers it to be badly overfished. Yuen and Miyake ( 1980) calculated that the present fishing effort (commercial longliner effort only, since no data are available on recreational fishing effort) is about twice that suitable for maximum sus- tainable yield. Because the catch per unit effort of Pacific blue marlin has steadily declined over the past 10 yr, in spite of a fairly constant level of effort, Yuen and Miyake (1980:19) concluded "...that continued fishing at high levels will con- tinue to reduce the abundance of the stock and a recruitment failure will become a distinct possi- bility." The importance of being able to define sub- populations or stocks of fishes with respect to the formulation of appropriate fishery management schemes has long been recognized (Marr 1957). This problem is especially acute for species (such as Pacific blue marlin) which are highly migra- Manuscript accepted June 1982. FISHERY BULLETIN: VOL. 81. NO. 1, 1983. 85 FISHERY BULLETIN: VOL. 81. NO. 1 tory, subjected to an oceanwide multinational fishery, and which, because of relatively low catches, are not well suited to tag-recapture studies. In fact, the pressing need to understand blue marlin stock structure has been recognized for some time (Shomura 1980; Yoshida 1981). The electrophoretic analysis of protein poly- morphisms in natural populations can be a pow- erful approach for analyzing genetic aspects of population structure in sexually reproducing or- ganisms. For this reason, the technique has been applied to the study of racial or subpopulation differentiation in numerous invertebrates and vertebrates (Ayala 1976). Because of the basic importance of information on subpopulation or stock structure to fisheries management (Berst and Simon 1981), population genetic studies have been conducted for many species of fishes [re- viewed by de Ligny (1969) and Allendorf and Ut- ter (1979)]. Most of the fishes investigated to date have been freshwater species or marine forms which are either inshore shallow-water species or demersal species. Stock heterogeneity for oceanic species has not generally been reported (but see Fujino 1976; Fujinoet al. 1981). Although open water, pelagic species may be characterized by large panmictic cosmopolitan populations, this pattern has not yet been clearly established. One problem in test- ing this hypothesis has been the unusually low levels of genetic variability observed to date in several large marine vertebrates such as skip- jack tuna (Fujino 1970) and seals (McDermid et al. 1972; Bonnell and Selander 1974). Indeed, Selander and Kaufman (1973) have even sug- gested that large, mobile vertebrates may gener- ally have low levels of heterozygosity— a charac- teristic which, if true, would preclude definitive stock analysis using electrophoretic techniques (but see Ryman et al. 1980). The general lack of progress in defining stock structure in oceanic fishes, such as scombroids, using electrophoretic methods, is attributable to several factors. Many of the reports in the literature have been prelimi- nary in nature dealing with small samples of fish and few variable loci. Although such small sam- ple sizes are not unexpected given the remote, far-seas nature of many of the commercial fish- eries, they severely limit the subsequent statisti- cal treatment of the data. Similarly, the analysis of only one or two polymorphisms reduces the likelihood of demonstrating any population sub- division which may exist. Finally, the schooling and/or highly migratory nature of many of these fishes makes it difficult to plan and execute ade- quate sampling programs. The study described in the present report was designed to determine the suitability of utilizing electrophoretic techniques to study stock struc- ture in the Pacific blue marlin. Three specific questions were addressed: 1 ) How much and what kind of electrophoreti- cally detectable genetic variation is there in the Pacific blue marlin? Specifically, is there enough genetic variation to allow an electro- phoretic analysis of stock structure in this species? 2) What combinations of enzymes, tissues, and buffer systems can be utilized in a study of genetic variation in this species? 3) What allele frequency distributions charac- terize the population of Pacific blue marlin in Hawaii? MATERIALS AND METHODS Muscle, liver, heart, eye, and brain samples were dissected from Pacific blue marlin landed at the Hawaiian International Billfish Tourna- ment held at Kailua-Kona, Hawaii, in August 1980. All tissue samples were taken immediately after each fish had been weighed, and all fish had been dead for at least 1 h but <8 h. The dissected tissues were initially placed on ice and subse- quently transferred to a freezer within 12 h. The time delay between fish capture and the freezing of dissected tissues did not seem to adversely affect any of the polymorphic enzymes screened with the possible exception of L-iditol dehydro- genase which could only be scored in 84 of the 95 fish analyzed. Tissues were stored frozen at -20°C until extracted. Tissue extracts were prepared by homogeniza- tion using a loose-fitting, motorized stainless steel pestle in polycarbonate centrifuge tubes. The extraction buffer consisted of 0.1M Tris-HCl pH 7.0 containing 1 X 10" 3 M EDTA and 5 X 10~ 5 M NADP\ After homogenization, the ex- tracts were centrifuged at 25,000 X gfor at least 30 min. Supernatants were transferred to indi- vidually labeled glass vials, capped, and stored at — 75°C until the electrophoretic analysis was completed. The supernatants were subjected to horizontal starch gel electrophoresis (modified from Selan- der et al. 1971), using some 15 different buffers. The gels were made using Lot 60F-0558 starch 86 SHAKLEE ET AL.: BIOCHEMICAL GENETICS OF PACIFIC BLUE MARLIN (Sigma Chemical Co., St. Louis, Mo.) at a concen- tration of 12% w/v. After electrophoretic separa- tions, enzyme patterns were visualized using standard histochemical staining recipes modi- fied from Shaw and Prasad (1970), Selander et al. (1971), and Siciliano and Shaw (1976). All zymograms were photographically recorded. Patterns of enzyme variation which were con- sistent with the subunit structure of the enzyme (when known) and simple models of Mendelian inheritance were scored and recorded as geno- types. Names of enzymes and Enzyme Commis- sion numbers follow the recommendations of the Commission on Biochemical Nomenclature ( 1973). For multilocus enzyme systems, loci were given alphabetic designations when appropriate (e.g., Gpi-A) or were simply assigned a number beginning with 1 for the most anodally migrat- ing isozyme. The most common allele at each locus was designated 100, and all other alleles at that locus were numbered according to their elec- trophoretic mobility relative to the 100 allele. Negative numbers refer to alleles with cathodal migration. The putative genotype data were sum- marized as genotype and allele frequency distri- butions. The genotype distributions were exam- ined for internal consistency with the Mendelian inheritance model by chi-square testing of good- ness-of-fit of observed genotype ratios with those expected for a single random mating population in the absence of differential selection among the alleles. The expected ratios were computed from observed allele frequencies using Levene's (1949) unbiased method for small samples. Heterozy- gosity for each locus (h) was calculated as h = 1 -- XX i 2 where Xi is the frequency of the /th allele. Average heterozygosity (H) was calculated as the mean of h over all loci examined. RESULTS Tissue samples from 95 Pacific blue marlin were analyzed. A total of 23 enzyme systems representing 35 gene loci were satisfactorily re- solved using extracts of muscle, liver, and eye (Table 1). Heart and brain tissue did not add sig- nificantly to this total. Eleven loci exhibited de- tectable genetic variation in the sample of 95 fish analyzed. The enzymes adenosine deaminase (Ada), mannosephosphate isomerase (Mpi), and phosphoglucomutase (Pgm) all behaved as mono- mers with two-banded heterozygotes. Aspartate aminotransferase ( Aat-1), alcohol dehydrogenase (Adh), glucosephosphate isomerase (Gpi-A), mus- cle glycerol-3-phosphate dehydrogenase (G-3- Pdh-2), liver isocitrate dehydrogenase (Idh-1), phosphogluconate dehydrogenase (Pgdh), and umbelliferyl esterase (Umb) behaved as dimers exhibiting triple-banded heterozygous patterns. L-iditol dehydrogenase (Iddh), often referred to as sorbitol dehydrogenase in the literature, ap- peared to be a tetramer as heterozygotes exhib- ited a five-banded phenotype. Two of the 1 1 variable loci were represented by only a single heterozygous individual out of the 95 fish screened. The remaining nine loci were TABLE 1.— Electrophoretic analysis of Makaira nigricans from Hawaii. M = muscle, E = eye, L = liver. Enzyme Loci Name (Enzyme Commission number) Abbr. Tissue Invariant Variable aspartate aminotransferase (2.6.1.1) adenosine deaminase (3.5.4.4) alcohol dehydrogenase (1.1.1.1) creatine kinase (2.7.3.2) enolase (4.2.1.11) esterase (3.1.1.—) glyceraldehy de-phosphate dehydrogenase (1.2.1.12) glutamate dehydrogenase (1.4.1.2) glucosephosphate isomerase (5.3.1.9) glycerol-3-phosphate dehydrogenase (1.1.1.8) hexose diphosphatase (3.1.3.11) L-iditol dehydrogenase (1.1.1.14) isocitrate dehydrogenase (1.1.1.42) lactate dehydrogenase (1.1.1.27) malate dehydrogenase (1 .1.1.37) malate dehydrogenase (NADP*) (1.1.1.40) mannosephosphate isomerase (5 3.1.8) peptidase (3.4.11 . — ) phosphogluconate dehydrogenase (1.1.1.44) phosphoglucomutase (2.7.5.1) superoxide dismutase (1.15.1.1) umbelliferyl esterase xanthine dehydrogenase (1.2.1.37) Aat L 1 Ada M Adh L — Ck M + E 2 Eno M 1 Est L 2 Gapdh M 2 Gdh L 1 Gpi M 1 G-3-Pdh M+L 1 Hdp L 1 Iddh L — Idh M+L 1 Ldh M + E 3 Mdh M 3 Mdh(Nadp+) M 1 Mpi M — Pep M 2 Pgdh M — Pgm M Sod L 1 Umb M Xdh L 1 87 FISHERY BULLETIN: VOL. 81, NO. 1 polymorphic by the normal criteria (common allele at a frequency of 0.99 or less) with five loci (Ada, Mpi, Pgdh, Pgm, and Umb) having the most common allele at a frequency of between 0.95 and 0.99 and four loci (Aat-1, Adh, G-3-Pdh- 2, and Iddh) having the most common allele at a frequency of <0.95 (Table 2). All 11 variable loci exhibited two or three alleles except for Aat-1 which had five different alleles, three of which were reasonably common. Heterozygosity values for the individual loci ranged from zero for all of the apparently mono- morphic loci to 0.494 for G-3-Pdh-2. The average heterozygosity (H) across all 35 loci was 0.0605. Where possible, the observed genotype distri- butions were tested for goodness-of-fit to Hardy- Weinberg equilibrium expectations (Gpi-A, Idh- 1, Mpi, and Pgm were not tested because of the very small number of observed variants in the sample). Where necessary, rare alleles were pool- ed prior to the tests. All tests were nonsignificant except that for Adh ( x 2 = 6.97, df = 1; P<0.01) where there was a significant deficiency of het- erozygotes. Further analysis of the Adh data was under- taken to attempt to identify the major contribu- tors) to this significant chi-square value. Since sex linkage of a locus can result in a deficiency of heterozygotes, the sample of blue marlin was subdivided into males and females. A chi-square test of the Adh genotypes of the 81 male fish also revealed a significant deficiency of heterozygotes (x 2 = 9.36, df = 1; P<0.005). Another possible source of the deficiency of heterozygotes could be the pooling of different year classes which actu- ally had different frequencies of the Adh alleles. Indeed, year class fluctuations of allele frequency have been reported in other fishes (Williams et al. 1973; Mitton and Koehn 1975; Smith et al. 1978; Smith 1979). In the absence of growth data for this species, the only subdivision we could make was on the basis of size. The 95 blue marlin were subdivided into two groups: 1) 100-200 lb total weight and 2) 201-450 lb. There were 74 fish in the 100-200 lb group and the statistical analy- sis of this group once again revealed a deficiency of heterozygotes ( x 2 = 9.95, df = 1; P<0.005). The similarity of the results for small fish and for males is not unexpected since all of the female fish (N = 13) were in the large size class (>200 lb). Therefore, the deficiency of heterozygotes seems to characterize the overall sample and cannot be attributed to sexual or gross age (= size) differences. Table 2.— Allele frequencies and heterozygosities for 11 variable loci in Makaira nigricans from Hawaii. Locus' (heterozygosity) Allele Frequency Aat-1 250 0.016 (h = 0.4497) 145 0.145 100 0.720 27 0.102 -30 0.016 Ada 107 0005 (h = 0.0716) 100 0.963 92 0.032 Adh -220 0.344 (h = 0.4556) -100 0.656 Gpi-A 100 0995 (h = 0.0099) 86 0.005 G-3-Pdh-2 100 0.595 (h = 0.4944) 75 0389 60 0.016 Iddh 147 0.006 (h = 0.4219) 100 0.702 22 0.292 ldh-1 100 0.995 (h = 0.0099) 84 0005 Mpi 104 0005 (h = 0.0316) 100 0.984 90 0.011 Pgdh 155 0.016 (h = 0.0529) 100 0.093 67 0.011 Pgm 144 0.011 (h = 0.0316) 100 0.984 33 0005 Umb 100 0.953 (h = 0.0896) 87 0.047 'Sample size = 95 fish for each locus ex- cept for Mpi where N = 94, Aat-1 and Adh where N = 93, and Iddh where N = 84. DISCUSSION In spite of the low levels of genetic variation re- ported in the literature for skipjack tuna(Fujino 1970; Fujino et al. 1981; Lewis 1981; Richardson in press) and suggested by Selander (1976:34) that "...levels of variability are unusually low in large marine vertebrates such as tuna fish and porpoises," the above data clearly indicate that the Pacific blue marlin does not have abnormally low levels of genetic variation. The level of poly- morphism (P.99) observed for marlin in the pres- ent study (P = 0.26; i.e., 9 out of 35 loci), although slightly lower than the average for fish (P = 0.31) reported by Selander (1976), is higher than the averages for reptiles, birds, and mammals (0.23, 0.15, and 0.23, respectively). Furthermore, the average heterozygosity (H) of 0.0605 for the 35 loci screened in the Pacific blue marlin is greater than the average of 0.0494 calculated by Nevo (1978) for 135 species of vertebrates and the aver- age of 0.0478 calculated by Winans (1980) for 82 species of fishes. Although perhaps somewhat un- 88 SHAKLEE ET AL.: BIOCHEMICAL GENETICS OF PACIFIC BLUE MARLIN expected, the relatively high level of genetic variation reported above for blue marlin is not unique among large marine vertebrates as sev- eral other scombroid fishes (e.g., white marlin, southern bluefin tuna, and Spanish mackerel) ex- hibit similar or even higher levels of variation (Edmunds 1972; Smith and Jamieson 1980; Lew- is 1981; Shaklee unpubl. data). The observed pattern of genetic variation can be used to subdivide the 11 variable loci into two general categories. Four loci (Aat, Adh, G-3-Pdh- 2, and Iddh) form one group characterized by high heterozygosities (0.4219-0.4944) due to the presence of at least two relatively common alleles. The second group, which is composed of the re- maining seven variable loci (Ada, Gpi-A, Idh-1, Mpi, Pgdh, Pgm, and Umb), is characterized by low heterozygosity per locus (0.0099-0.0896) and the presence of a single common allele at a fre- quency of at least 0.95. In reality, both groups of loci are of utility in population analyses because the power of the statistical tests for detecting sig- nificant differences in allele frequency between pairs of samples actually increases somewhat as the frequencies approach the extremes (i.e., and 1.0) compared with samples having fre- quency distributions close to 0.5. The close agreement between observed and ex- pected genotypic frequencies for all but one of the variable loci is consistent with all 95 fish analyzed belonging to a single panmictic popula- tion. However, the significant (P<0.01) defi- ciency of heterozygotes observed at the Adh locus is not. Although this observation may be due to selection or simply be an anomaly, another poten- tial explanation is that this heterozygote defi- ciency is due to the mixing of two or more differ- ent stocks of blue marlin which have different frequencies of the two Adh alleles (Wahlund effect). Such potential stock mixing would not be unreasonable given the presumed migratory na- ture of Pacific blue marlin. The fundamental significance of the above ob- served levels of genetic variation is that they are adequate to allow a biochemical genetic analysis of stock structure in Pacific blue marlin. We are now in the process of initiating just such an anal- ysis and are employing an experimental design which should allow us to detect stock heteroge- neity which has either a stable geographical basis or a temporally shifting geographic basis. The former basis for stock heterogeneity, namely the localization of two or more stocks in different re- gions of the species' range is by far the most com- monly observed form of population subdivision among organisms. We will be testing for this type of heterogeneity by analyzing samples from different localities (Hawaii, Guam, Samoa, etc.) throughout the range of the Pacific blue marlin. However, this type of analysis is complicated by limited access to marlin caught in many areas of the range, by the shifting patterns of abundance which characterize this species, and by the vir- tual impossibility of obtaining simultaneous sam- ples of blue marlin from multiple localities throughout the range. Indeed, it is the apparent migratory nature of the species which suggests that the second type of population structuring, that based on both temporal and geographic iso- lation, may be occurring in this species. Our ap- proach to this problem is to sample continuously in the Hawaiian Islands to look for significant seasonal shifts in allele frequency such as might be expected if different stocks of blue marlin mi- grate past the Hawaiian Islands at different times of the year. Given the present potentially overfished nature of billfish stocks and the diffi- culties associated with alternative forms of stock analysis such as tag-recapture studies, this bio- chemical genetic approach may well represent the only practical means of gathering informa- tion on stock structure in a time frame compat- ible with the urgent need for the formulation of meaningful management programs for this spe- cies. ACKNOWLEDGMENTS We wish to gratefully acknowledge the support and encouragement provided by the staff and Board of Governors of the Hawaiian Interna- tional Billfish Association and the cooperation of all the participants in the 1980 Hawaiian In- ternational Billfish Tournament. The valuable help provided by Jerry Kinney and his staff at the Volcano Isle Fish Co. is also sincerely appre- ciated. LITERATURE CITED Allendorf, F. W.. and F. M. Utter. 1979. Population genetics. InW. S Hoar, D. J. Randall, and J. R. Brett (editors). Fish physiology, Vol. 8, p. 407- 454. Academic Press, Inc., N.Y. Ayala, F. J. (editor). 1976. Molecular evolution. Sinauer Assoc, Inc., Sun- derland, Mass., 277 p. Berst, A. H.. and R. C. Simon. 1981. Introduction to the Proceedings of the 1980 Stock Concept International Symposium (STOCS). Can. J. 89 FISHERY BULLETIN: VOL. 81. NO. 1 Fish. Aquat. Sci. 38:1457-1458. BONNELL, M. L., AND R. K. SELANDER. 1974. Elephant seals: Genetic variation and near extinc- tion. Science (Wash.. D.C.) 184:908-909. Commission on Biochemical Nomenclature. 1973. Enzyme nomenclature. American Elsevier Publ. Co., Inc., N.Y., 443 p. De Ligny, W. 1969. Serological and biochemical studies on fish popula- tions. Oceanogr. Mar. Biol. Annu. Rev. 7:411-513. Edmunds, P. H. 1972. Genie polymorphism of blood proteins from white marlin. U.S. Dep. Inter., Fish Wildl. Serv., Bur. Sport Fish. Wildl., Res. Rep. 77:1-15. Fujino, K. 1970. Immunological and biochemical genetics of tunas. Trans. Am. Fish. Soc. 99:152-178. 1976. Subpopulation identification of skipjack tuna speci- mens from the southwestern Pacific Ocean. Bull. Jpn. Soc. Sci. Fish. 42:1229-1235. Fujino, K., K. Sasaki, and S. Okumura. 1981. Genetic diversity of skipjack tuna in the Atlantic, Indian, and Pacific Oceans. Bull. Jpn. Soc. Sci. Fish. 47:215-222. Howard, J. K., and S. Ueyanagi. 1965. Distribution and relative abundance of billfish- es (Istiophoridae) of the Pacific Ocean. Stud. Trop. Oceanogr., Inst. Mar. Sci. Univ. Miami 2, 134 p. Levene, H. 1949. On a matching problem arising in genetics. Annu. Math. Stat. 20:91-94. Lewis, A. D. 1981. Population genetics, ecology and systematics of Indo-Australian scombrid fishes, with particular refer- ence to skipjack tuna (Katsutvonus pelamis). Ph.D. Thesis, Australian National Univ., Canberra, 314 p. Marr, J. C. 1957. The problem of defining and recognizing subpopu- lations of fishes. U.S. Fish Wildl. Serv., Spec. Sci. Rep.— Fish. 208:1-6. Matsumoto, W. M., and T. K. Kazama. 1974. Occurrence of young billf ishes in the central Pacific Ocean. In R. S. Shomura and F. Williams (editors), Proceedings of the International Billfish Symposium, Kailua-Kona, Hawaii, 9-12 August 1972. Part 2, p. 238- 251. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-675. McDermid, E. M., R. Ananthakrishnan, and N. S. Agar. 1972. Electrophoretic investigation of plasma and red cell proteins and enzymes of Macquarie Island elephant seals. Anim. Blood Groups. Biochem. Genet. 3:85-94. Mitton, J. B.. and R. K. Koehn. 1975. Genetic organization and adaptive response of allo- zymes to ecological variables in Fundvlus heteroclitus. Genetics 79:97-111. Nevo, E. 1978. Genetic variation in natural populations: Patterns and theory. Theor. Pop. Biol. 13:121-177. Richardson, B. J. In press. The distribution of protein variation in skipjack tuna (Katsuwonus pelamis) from the central and south- western Pacific. Aust. J. Mar. Freshw. Res. 34. Rivas, L. R. 1975. Synopsis of biological data on blue marlin, Maka ira nigricans Lacepede, 1802. In R. S. Shomura and F. Williams (editors). Proceedings of the International Bill- fish Symposium, Kailua-Kona, Hawaii. 9-12 August 1972. Part 3, p. 1-16. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-675. Ryman, N., C. Reuterwall, K. Nygren, and T. Nygren. 1980. Genetic variation and differentiation in Scandi- navian moose (Alces alces): Are large mammals mono- morphic? Evolution 34:1037-1049. Selander, R. K. 1976. Genie variation in natural populations. In F. J. Ayala (editor), Molecular evolution, p. 21-45. Sinauer Assoc, Inc., Sunderland, Mass. Selander, R. K., and D. W. Kaufman. 1973. Genie variability and strategies of adaptation in animals. Proc. Natl. Acad. Sci. USA 70:1875-1877. Selander, R. K., M. H. Smith, S. Y. Yang, W. E. Johnson, and J. B. Gentry. 1971. Biochemical polymorphism and systematics in the genus Peromyscus. I. Variation in the old field mouse (Peromyscus polionotus). Stud. Genet. VI, Univ. Texas Publ. 7103:49-90. Shaw, C. R., and R. Prasad. 1970. Starch gel electrophoresis of enzymes— a compila- tion of recipes. Biochem. Genet. 4:297-320. Shomura, R. S. (editor). 1980. Summary report of the billfish stock assessment workshop Pacific resources. U.S. Dep. Commer.. NOAA-TM-NMFS-SWFC-5. Siciliano, M. J., and C. R. Shaw. 1976. Separation and visualization of enzymes in gels. In I. Smith and J. W. T. Seakins (editors), Chromato- graphic and electrophoretic techniques Vol. 2, Zone elec- trophoresis, 4th ed., p. 185-209. Halsted Press. Smith, P. J. 1979. Esterase gene frequencies and temperature rela- tionships in the New Zealand snapper, Chrysophrys auratus. Mar. Biol. (Berl.) 53:305-310. Smith, P. J., and A. Jamieson. 1980. Protein variation in the Atlantic mackerel Scomber scombrus. Anim. Blood Groups Biochem. Genet. 11: 207-214. Smith, P. J., R. I. C. C. Francis, and L. J. Paul. 1978. Genetic variation and population structure in the New Zealand snapper. N.Z. J. Mar. Freshw. Res. 12: 343-350. Strasburg, D. W. 1970. A report on the billfishes of the central Pacific Ocean. Bull. Mar. Sci. 20:575-604. Williams, G. C, R. K. Koehn, and J. B. Mitton. 1973. Genetic differentiation without isolation in the American eel, Anguilla rostrata. Evolution 27:192-204. WlNANS, G. A. 1980. Geographic variation in the milkfish Chanos chan- os. I. Biochemical evidence. Evolution 34:558-574. Yoshida, H. O. 1981. Status report: Pacific striped, blue, and black mar- lins. In Status reports on world tuna and billfish stocks, p. 277-300. U.S. Dep. Commer., NOAA-TM-NMFS- SWFC-15. Yuen, H. S. H., and P. Miyake. 1980. Rapporteurs' reports, blue marlin, Makaira nigri- cans. In R. S. Shomura (editor), Summary report of the billfish stock assessment workshop Pacific resources, p. 13-19. U.S. Dep. Commer., NOAA-TM-NMFS- SWFC-5. 90 STOCHASTIC AGE-FREQUENCY ESTIMATION USING THE VON BERTALANFFY GROWTH EQUATION Norman VV. Bartoo 1 and Keith R. Parker 2 ABSTRACT The method of estimating age frequency from length frequency via the von Bertalanffy growth equation is deterministic and yields biased results. Most of the bias can be removed by incorporating a stochastic element in the von Bertalanffy relationship. The stochastic element is based on esti- mated probabilities of lengths by intervals at age, the probabilities being estimated from variances in lengths-at-age. Based on age-length samples from the Pacific bonito fishery the stochastic method gives improved age-frequency estimates over those obtained by the deterministic method. The sto- chastic application may be generalized to all growth models including discontinuous growth such as in crustaceans. Complex population dynamics techniques rely heavily on age-structure information. Frequent- ly, appropriate assessment techniques for a stock require an estimate of the age frequency of that stock. For example yield-per-recruit analysis (Ricker 1958) is computed on the dynamic rela- tionship between growth and mortality: Mortal- ity rates when computed via cohort analysis (Murphy 1965) are based on estimated age fre- quency. For some species accurate aging methods are not available. When feasible, determining the age of fish and consequently computing an age frequency are most accurately accomplished by visual inspection of scales, otoliths, or other struc- tures (Ricker 1958). Such visual inspection is time consuming and often expensive. To reduce the cost and time of estimating the age structure of a fisheries catch, age frequency is usually esti- mated from sampled length frequency, the age- length relationship being described by either an age-length key or a growth curve such as the von Bertalanffy growth curve (Ricker 1958). The growth curve method is used when there are in- sufficient data to construct an age-length key. Age-length keys work on the principle that age can be estimated from length using information contained in a previously or concurrently aged sample from the population. As long as the pro- portion of length-at-age remains the same for all ages, then the age-length key will yield unbiased 'Southwest Fisheries Center La Jolla Laboratory. National Marine Fisheries Service, NOAA, 8604 La Jolla Shores Drive, La Jolla, CA 92038. 2 1837 Puterbaugh Street, San Diego. CA 92103. estimates of age for any sampled lengths from that population. However, since the estimated parameters of an age-length key — proportions of age-at-length— are dependent on the sampled population used to construct the key, the applica- tion of the key to the population with altered age structures can yield inaccurate results. Kimura (1977) and later Westrheim and Ricker (1978) demonstrated that under conditions of varying year-class strength and substantial overlap of lengths between ages, age-length keys can yield nearly useless estimates of numbers-at-age. Clark (1981) effectively removes age-length key bias by first proportioning numbers in length intervals at age over time and then using the ma- trix of these proportions standardized over time to compute least-squares estimates of age fre- quency from the vector of length frequency. Effective applications of many stock assessment growth and mortality based methods require that ages are expressed in fractions of years (Ricker 1958; Lenarz et al. 1974). The large num- ber of aged fish required to construct a sufficient key for a large number of ages is difficult and ex- pensive to attain. Even with Clark's bias correc- tion procedure, the construction of a sufficient key can present difficulties due to data needs. In this paper we deal specifically with the von Bertalanffy growth equation and the application of stochastic methods to reduce or eliminate bi- ases. However, it should be noted that the method presented here may be applied to any growth equation as well as to cases where no growth equation has been fitted or where growth is dis- continuous as in crustaceans. The von Bertalanffy growth equation mathe- Manuscript accepted June 1982. FISHERY BULLETIN: VOL. 81. NO. 1. 1983. 91 FISHKRY BULLETIN: VOL. 81. NO. 1 matically models the relationship between age and length, length being the dependent variable (see Equation (1)). As suggested by Gulland (1973), age is estimated from length by algebrai- cally rearranging the growth equation so that age is the dependent variable (see Equation (2)). Re- gardless of whether length or age is the depen- dent variable, the von Bertalanffy relationship is deterministic: There is a one-to-one correspon- dence between age and length. For the von Bertalanffy growth equation, age frequency is estimated from a length sample as follows: 1) For each length compute the corresponding age. 2) For each age interval, usually the interval between midpoint ages of adjacent ages, sum the number of aged fish falling within the in- terval. 3) The age frequency is then the total number of aged fish falling within each age interval. Use of the von Bertalanffy growth equation for age-frequency estimation results in several types of biases, different from those inherent in age- length keys. This paper documents these biases and proposes a method for their resolution. BIASES When growth is modeled according to the von Bertalanffy age-length relationship (Brody 1945; Ricker 1958), Lt = L x (1 -exp[-k(t - h)]), (1) then age, t, can be converted to length: t = h + \n (\ - L t IU)l(-k) (2) where Lt L x k to length at age t the asymptotic length the rate at which length reaches L hypothetical age at which fish would have zero length. When computing numbers-at-age from Equa- tion (2), estimation bias occurs. One bias is due to L x being a fitted parameter. Thus, all numbers- at-length greater than L x must either be elimi- nated or arbitrarily distributed to older ages. Bias also results when lengths approach L^ and are mathematically allocated to ages above those attainable by fish within the stock. As lengths (L) approach L x , Equation (2) will yield unreason- ably old ages. Additional bias results from the deterministic nature of the von Bertalanffy equation: Back cal- culations of length to age, Equation (2), are on a one-to-one basis. Thus, for any length there is a determined age. In reality, there can be a num- ber of possible ages for any given length, the most probable age-at-length being that with the highest relative contribution of numbers-at- length. Since these back calculations are without probabilistic arguments, the determined age is not necessarily the most probable for the given length. Back calculations of length to age also result in a mathematical estimation bias due to the switch- ing of independent and dependent variables in going from Equation (1) to Equation (2). The de- gree of bias is likely to be a function of the amount of residual error in fitting Equation (1). The bias will probably not be consistent between cases and the degree of bias will have to be con- sidered separately for each case. Consequently this bias is not specifically dealt with in this paper. A computer model can readily demonstrate this bias. For von Bertalanffy parameters: L x = 90.0 units, to = 0.0 units, and k = 0.30, predeter- mined numbers-at-age are arbitrarily distrib- uted normally with a standard deviation equal to 3 units about the von Bertalanffy length-at-age, Equation (1), for ages I through X. A length-fre- quency vector is then generated by 1) multiply- ing the number-at-age times the probability of age occurring within each 0.5 unit length inter- val, thus for each age generating a vector of num- ber-at-length for length intervals between and 100 units, and 2) accumulating numbers-at- length for each length interval over all ages. The numbers-at-age are then deterministically esti- mated from Equation (2) by accumulating num- bers-at-length over the length intervals at age. The bias from this model is illustrated in Table 1. Input and back-calculated numbers-at-age and their differences are listed in columns 2, 3, and 4, respectively. The input numbers-at-age represent a sample age distribution where either catchability, recruitment, mortality, or some combination thereof, are age-class variant. Dif- ferences, column 4, indicate a strong bias which increases with overlap of length distributions at age. One hundred and eleven fish were aged to be 92 BARTOOancl PARKER: STOCHASTIC AOE-FREUUKNCY ESTIMATION Tablk 1.— Input and estimated numbers-at-age for both the deterministic (col. 3) and stochastic (col. 5) models, with the in- put numbers-at-age in column 2. The differences between the input numbers-at-age and the deterministic estimates are given in column 4. Numbers -at-age Age Input Deterministic Difference Stochastic (1) (2) (3) (4) (5) I 200 199 1 200 II 400 399 1 400 III 800 760 40 800 IV 200 267 -67 200 V 600 441 159 600 VI 300 378 78 300 VII 400 320 80 400 VIII 300 258 42 300 IX 100 164 -64 100 X 100 68 32 100 >x — 111 - 111 — Inf. — 35 -35 — greater than the maximum age, 10. Thirty-five had lengths greater than L x and, consequently, were not classifiable. BIAS RESOLUTION With estimated variance of length-at-age a sto- chastic model can be built from the von Berta- lanffy relationship: For any age the probability of a specific length interval is the probability of that interval taken over all length intervals con- taining that age. Thus for all ages a probability matrix ("P"-matrix) of dimension r by c can be computed, where r = the number of rows, or length intervals, and c — the number of columns, or ages, then P (1,1) = (max. length, min. age). If the number-at-age vector is "a" (ai = (min. age)) and the number-at-length vector is L (L\ = (max. length)), then P a = L. (3) And as long as r > c then the numbers-at-age vec- tor can be uniquely solved via least-squares: a = (P'PY'P'L (4) Applying this stochastic method (Equation (4)) to the previous example, the numbers-at-age generated from the number-at-length vector are given in column 5 of Table 1. Since the probabili- ties of the P-matrix are the same as those used to generate the number-at-length vector, it is not surprising that the solution yields unbiased re- sults. This computed example illustrates that the stochastic method yields unbiased estimates of age frequency. PACIFIC BONITO For the Pacific bonito, Sarda chiliensis, of the eastern tropical Pacific, Campbell and Collins (1975), using ages determined from otoliths, esti- mated the von Bertalanffy growth parameters to be L^ = 76 87 cm, t = -0.785 yr. and k = 0.6215. Numbers-at-length for 1 cm intervals for ages I through V are shown in Figure 1 with the corre- sponding length-frequency plot in Figure 2. These numbers represent the 1973 catch from California waters and are a subset of the data used to estimate the von Bertalanffy parameters. This example serves to demonstrate bias and illustrate application of the stochastic method. If desired a variance-covariance matrix can be gen- erated (Draper and Smith 1981) to estimate pre- cision in the resulting age structure. 80 r- 70 - E o h- 60 O z HI 50 - 40 1 2 3 5 2 1 9 11 8 ^S^lV, 5 >*^ 12 1 — 1 6* r 4 2 A 2 1 3/ 9* 15 3 1 /38 /41 8 2 1 J / 25 / 15 ' 3 1 — 1 / 13 / 26/ 44/ 48 71 100=76.870111 T=-0.785yr ft 13 3 K = 0.6215 2 1 1 1 i i IV V AGE (years) Figure 1.— Numbers-at-age by length in 1 cm intervals for the Pacific bonito. Sarda chiliensis. Data from 1973 California landings (Campbell and Collins 1975). The length-frequency information and von Bertalanffy parameters are used to generate both deterministic and stochastic estimates of numbers-at-age. The estimated length-at-age 93 FISHERY BULLETIN: VOL. 81, NO. 1 80 r LENGTH (cm) Figure 2.— Length frequency for the Pacific bonito, Sarda chiliensis, from 1973 California landings (Campbell and Col- lins 1975). and sample standard deviations are Age Length Standard deviation I 51.5 cm 2.7 cm II 63.3 cm 2.1 cm III 69.5 cm 2.3 cm IV 72.9 cm 2.0 cm V 74.8 cm 1.9 cm Deterministic estimates were made on lengths rounded to the nearest 0.1 cm. From Equation (2) the deterministic numbers-at-age are shown in column 3 of Table 2 with the difference between the true and estimated numbers in column 4. While the estimates are reasonably close over the first two ages, they become increasingly dispar- ate for older ages. Thirteen fish had lengths greater than those at the maximum age. Seven had lengths greater than L and consequently were unclassifiable. Quarter centimeter intervals were used to com- pute the stochastic estimates of age frequency. The results are shown in column 6 of Table 2, with the difference between true and estimated numbers in column 7. For all ages, especially the older ages, the stochastic estimates are closer and less biased than those of the deterministic method (column 3). Some insight into the improvement of the sto- chastic age estimates over the deterministic age estimates can be gained by inspection of Figure 1. Lengths of age I fish overlap those of age 1 1 fish and vice versa. Since the deterministic cutoff point for age I fish is 58.7 cm (1.5 yr), all overlap is lost in the deterministic model. In contrast, for the stochastic model, overlaps in lengths-at-age are shared between ages, the degree of sharing being relative to the probabilities of length inter- vals at the respective ages. With increasing age, the extent of relative overlap and, consequently, misaging increases for the deterministic model; allocation of lengths to ages becomes more sensitive. Only if the de- gree of overlap between adjacent ages is equal do accurate estimates of numbers-at-age result from the deterministic model. In the present ex- ample varying year-class strength and random variability in lengths-at-age offset this sensitive compensatory mechanism needed for accurate estimation with the deterministic model. Fish lengths above 75.3 cm, the length at age 5.5 yr, are misclassified either as older or of in- finite ages for the deterministic model. Since for the stochastic model probabilities of length inter- vals at age exist for all ages and lengths, even for lengths above L x , fish at lengths above the 75.3 cm cutoff point are distributed to all ages rela- tive to their respective probabilities for length intervals. DISCUSSION Calculations of age from length via the von Bertalanffy growth equation result in several Table 2.— Deterministic (col. 3) and stochastic (col. 6) estimates of numbers- at-age with their respective differences from the true numbers-at-age in col- umns 4 and 7 for the Pacific bonito. Sarda chiliensis. from 1973 California landings (Campbell and Collins 1975). Numbers-at ■age Deter- Differ- % Dif- Sto- Differ- % Dif- Age True ministic ence ference chastic ence ference (D (2) (3) (4) (5) (6) (7) (8) I 424 411 13 3 1 415 9 2.1 II 158 167 -9 -5.7 162 -4 -2.5 III 54 39 15 27.8 49 5 93 IV 80 71 9 11 3 85 -5 -6.3 V 21 29 -8 38.1 26 -5 -238 >v — 13 13 Inf — — 0.0 Inf. — 7 -7 Inf. — — 00 94 BARTOO and PARKER: STOCHASTIC AGE-FREQUENCY ESTIMATION types of bias. The degree of bias is proportional to overlap in lengths-at-age and changes with weak or strong year classes. When overlap increases with age, age-frequency estimates will generally be more biased for older than younger ages. When overlap occurs, biases will always result, since the numbers-at-length will be allocated to unreasonably old ages. Any numbers-at-length for lengths greater than L x will be undetermined in age estimation, resulting in downward biases for those ages contributing such lengths. Age estimation biases can be effectively re- moved by creating a stochastic model based on a matrix of length interval probabilities at age. The probability matrix (P-matrix) is indepen- dent of year-class strength and will effectively remove all sources of estimation bias except that due to random variation in length-frequency esti- mation. As long as the von Bertalanffy growth parameters remain the same over time, the sto- chastic method based on accurate estimates of variance in length-at-age will always yield un- biased results. A probability model of the distribution of length-at-age with estimated parameters is nec- essary for estimating probabilities of length intervals at age for the P-matrix. If age infor- mation is unavailable then variances can be esti- mated from visually separable length-frequency modes. In the case where modes are separable for the first few ages only, there will be a problem in estimating variances for older ages: A model re- lating the variance in length-at-age with age can be used in estimating variances for these older ages. Ricker (1969) proposed that while distribu- tions in lengths-at-age remain normal, variances increase during the first few years, stabilize, and then decrease over the final years. The trend in variances with age for a similar species might also be substituted in cases where variances are unavailable. The principal strengths of the stochastic meth- od are that few fish are required to be aged to estimate the P-matrix and that existing von Ber- talanffy growth relations can be used. Accurate estimates of variance in length-at-age can prob- ably be achieved with as few as 20 to 30 fish/ age, which is likely to be a much smaller number of fish than needed to estimate accurate propor- tions of age-at-length necessary to construct an age-length key. Von Bertalanffy growth parameters have been estimated for many species. Since most stocks have variant year-class strength, overlaps in lengths-at-age, and lengths exceeding the up- per bound for the last age attainable, conversion to a stochastic model may be necessary, if unbi- ased estimates of age frequency are desired. Re- examination of age-length data used to estimate the von Bertalanffy parameters may be useful in estimating variances in lengths-at-age for the P- matrix. Taking additional age-length samples may be a cost-effective way of improving age-fre- quency estimation. In fishery management, the overestimation of maximum age by the deterministic von Berta- lanffy equation may produce underestimates of mortality rates which may result in overesti- mates of population size and recruitment. Fur- ther, the deterministic method tends to "fill in" weak year classes which results in underesti- mates of year-class variability and overestimates of recruitment stability. In general, all of these affect accuracy of a stock assessment and con- tribute to improper advice for fishery manage- ment. Application of the stochastic method shown here to cover other growth equations and situa- tions, such as discontinuous growth, is handled by simply estimating appropriate elements in the P-matrix for each case. ACKNOWLEDGMENTS We thank Douglas Chapman and Alec MacCall for helping to define the problem and evaluating the solution. Mark Farber, Joseph Powers, Lewis J. Bledsoe, and Gary Sakagawa provided critical review and comment for which the authors are grateful. We are also grateful to R. A. Collins of the California Department of Fish and Game for providing data. LITERATURE CITED Brody, S. 1945. Bioenergetics and growth. Reinhold Publ. N.Y.. 1023 p. Campbell, G., and R. A. Collins. 1975. The age and growth of the Pacific bonito, Sarda chiliensis, in the eastern North Pacific. Calif. Fish Game 61:181-200. Clark. W. G. 1981. Restricted-least squares estimate of age composi- tion from length composition. Can. J. Fish. Aquat. Sci. 38:297-307. Draper, N. R., and H. Smith. 1981. Applied regression analysis. 2d ed. John Wiley and Sons, Inc.. N.Y.. 709 p. Gulland, J. A. 1973. Manual of methods for fish stock assessment. Part 95 FISHERY BULLETIN: VOL. 81, NO. 1 1. Fish population analysis. FAO Man. Fish. Sci. 4. Board Can. 22:191-202. 154 p. RlCKER, W. E. KiMi'RA, D. K. 1958. Handbook of computations for biological statistics 1977. Statistical assessment of the age-length key. J. of fish populations. Fish. Res. Board Can. Bull. 119, Fish. Res. Board Can. 34:317-324. 300 p. Lenarz, W. H., W. W. Fox, Jr., G. T. Sakagawa, and B. J. 1969. Effect of size-selective mortality and sampling Rothschild. bias on estimates of growth, mortality, production, and 1974. An examination of the yield per recruit basis for a yield. J. Fish. Res. Board Can. 26:479-541. minimum size regulation for the Atlantic yellowfin WESTRHEIM, S. J., AND W. E. RlCKER. tuna, Thunnus albacares. Fish. Bull., U.S. 72:37-61. 1978. Bias in using an age-length key to estimate age-fre- Murphy, G. I. quency distributions. J. Fish. Res. Board Can. 36:184- 1965. A solution of the catch equation. J. Fish. Res. 189. 96 AGE, GROWTH, AND MORTALITY OF KING MACKEREL, SCOMBEROMORUS CAVALLA, FROM THE SOUTHEASTERN UNITED STATES 1 Allyn G. Johnson, William A. Fable, Jr., Mark L. Williams, and Lyman E. Barger 2 ABSTRACT Age, growth, and mortality of king mackerel, Scomberomorus en ml In, from the southeastern United States were studied. Otoliths from 1,449 fish were used to estimate age composition, growth rates, and mortality rates of this species. Age composition varied between locations (Texas, Louisiana, Florida, South Carolina, and North Carolina). The majority of older fish were found in Louisiana waters. The oldest females were 14 + years old and the oldest males were 9+ years old. Compensatory growth was found in both sexes. The von Bertalanffy growth equations were as follows: Males (all areas) 1, =965(1 — e' ' 17) ); females from Louisiana 1 , = 1 ,529 (1 — e '); and females (excluding Louisiana) 1 ( = 1,067 ( 1 — e -0 291 1 -0 97) where 1 = fork length (mm) and / = years. The mean annual mortality rate determined by six methods of analysis ranged from 0.32 to 0.42. The length-weight relations of king mackerel were for males: W = 0.8064 X lO^L 29928 ; for females: W = 0.8801 X 10" 5 L 29827 . where W = weight in grams and L = fork length in millimeters. King mackerel, Scomberomorus cavalla, is a major recreational and commercial fisheries re- source in the southeastern United States (Ma- nooch 1979). Age, growth, and mortality informa- tion has been based on small specimens collected from a limited geographical area (Beaumariage 1973). A need has existed to reexamine age, growth, and mortality from broader geographi- cally based samples. King mackerel of Brazil have been studied in- tensively, but the great distance separating these Brazilian fish from those in the United States makes application of their results to king mack- erel in United States waters a questionable prac- tice (see Manooch et al. 1978 for annotated bib- liography on this species). A geographically comprehensive sampling of king mackerel in U.S. waters was initiated by us in 1977. Recreational landings were sampled be- cause the sport fishery is less localized than the commercial fishery. We utilized samples from Texas to North Carolina to meet our objectives of determining the age composition, growth rates, length-weight relationships, and mortality rates of king mackerel from U.S. waters. •Contribution No. 82-29-PC, Southeast Fisheries Center Panama City Laboratory, National Marine Fisheries Service, NOAA, Panama City, Fla. 2 Southeast Fisheries Center Panama City Laboratory, Na- tional Marine Fisheries Service, NOAA, 3500 Delwood Beach Road. Panama City, FL 32407-7499. METHODS AND MATERIALS King mackerel (7,723 fish) were collected from Texas, Louisiana, Florida, South Carolina, and North Carolina from June 1977 through August 1979 (Fig. 1). They were caught by recreational hook and line, except for some small individuals, which were caught in shrimp trawls at Cape Canaveral, Fla., in December 1978. The trawl- caught fish were used in determining the rela- tion between otolith radius and fish length. In 1979, 121 fish samples were taken only in north- west Florida and were used to supplement exist- ing samples for the marginal increment analysis. Processing the fish samples involved several steps. The fish were sexed when possible, mea- sured to the nearest millimeter of fork length (FL), and weighed to the nearest gram. Otoliths were removed from the fish, cleaned, and stored either dry or in 100% glycerin. The otoliths were examined under reflected light in a black-bottomed watch glass containing 100% glycerin with a binocular dissecting micro- scope at 28X. The otolith radius (OR) was mea- sured on the posterior surface from the focus to the distal margin along the axis approximating the extension of the sulcus acousticus. All mea- surements were made in ocular micrometer units (1 om/j. = 0.0363 mm). Marks were counted and measured along the radius to their distal edge. The marks were opaque (light) under re- Manuscript accepted June 1982. FISHERY BULLETIN: VOL. 81. NO. 1. 1983. 97 FISHERY BULLETIN: VOL. 81. NO. 1 NORTH CAROLINA SOUTH CAROLINA CAPE CANAVERAL Figure 1. — Location of king mackerel, Scomberomorus cavalla, sampling sites. fleeted light, while the interspaces were hyaline (dark). Otoliths were classified into age groups accord- ing to the number of opaque nonmarginal marks (following the method of Beaumariage 1973). Each otolith was examined by two readers. If the readers did not agree on the age of a fish, data for that fish were not used. We determined the time of mark formation by comparing frequency per month of otoliths with opaque margins. A high percentage of opaque margins indicated recent mark formation. Comparison of age estimations was made, based on surface (whole) and internal (sectional) examination of 133 otoliths. Three to 10 otoliths from each age (0+ through 14+) were used for the comparison. Three to six sections, each 0.15 mm thick, were made through the focus of each otolith, using a Norton 3 diamond blade (SD519- N50m-l/8) rotating at about 285 rpm on an Iso- met low-speed saw. The otolith was mounted in thermoplastic (quartz) cement (No. 70C Lake- side) and cooled with mineral spirits during sec- tioning. Later the cement was dissolved by soak- ing in 50% isopropanol. The free sections were then mounted on glass slides using Piccolyte ce- 'Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. ment and examined with a binocular dissecting microscope. The relationship of the size of the aging struc- ture (OR) to the size of the fish (FL) was deter- mined by using least-square regressions with both linear and power curves. Once the relation- ship was established, fork lengths at earlier ages were back-calculated from surface otolith mea- surements, using methods adopted from Tesch (1971), Ricker (1975), and Everhart et al. (1975). Otolith measurements were analyzed for im- plications of compensatory growth. A frequency distribution of otolith lengths from the focus to the proximal edge of the first opaque mark was developed. Both slow- and fast-growing fish were separated from those that grow at inter- mediate rates, and lengths at earlier ages were back-calculated for both the slow and fast grow- ers. A computer program by Abramson (1971) was used to fit von Bertalanffy theoretical growth curves. Each age was given equal weight, and mean back-calculated lengths were used in the computations. Length-weight equations were developed for the entire king mackerel collection, and for males and females separately, by a computer program following Ricker's (1975) suggestions. Nonlogarithmic length intervals (50 mm) and 98 JOHNSON ET AL.: AGE. GROWTH. AND MORTALITY OF KING MACKEREL weight intervals (computed by the program) were used. A maximum of 20 length-weight val- ues was randomly selected for the analysis with- in each qualifying length and weight interval. If any length or weight interval contained fewer than 20 values, all were utilized. Estimates of annual mortality rate A (after Ricker 1975) were developed by catch-curve analysis of south Florida length-frequency data. These data were used because they best repre- sented the king mackerel in U.S. waters accord- ing to Trent et al. (1981). Since these data were not separated by sex, two age-length keys were developed, one combining males and females as- suming a 1:1 sex ratio and the other assuming a 1 male:2 female ratio (the approximate ratio in our collection). The length-frequency data were converted to age-frequency distributions {N, = number of fish caught in age-class i ) by applying each of the combined age-length keys. Age classes I through X of the resultant catch curves were analyzed by 1. Heincke's (1913) method; 2. Jackson's (1939) method; 3. Rounsefell and Everhart's (1953) method; 4. Beverton and Holt's (1957) method, using the mean of values computed with their equation 13.4 between successive age groups; 5. Robson and Chapman's (1961) method, un- corrected for possible age-length key bias; and 6. finding the slope (m) of a regression line fitted to In (N,) and i and substituting in the equation A — 1 — e m . RESULTS AND DISCUSSION Age The validity of using otoliths for estimating the age and past growth history depends on these structures being directly correlated with the growth of the fish and on otolith mark formation being periodic. We found the otolith radii to be closely correlated to fork lengths, especially when the data were transformed to represent a "power" function. The "power curve" equation, FL = 1.232 OR 1331 with correlation coefficient r = 0.987, had a better fit than the linear equa- tion, FL = 5.559 OR + 84.818 with r = 0.847. This close correlation of OR and FL satisfied the first criterion for validation of otoliths as an age determination structure. The second criterion, mark formation of known periodicity, needed further investigation. Beaumariage (1973) found king mackerel with opaque margins during 8 mo of the year (February-September); the highest percentage of otoliths with opaque margins oc- curred in May. He concluded, "Most otolith mar- gins become opaque (form annuli) during April. May, and June...." Fish in our collections exhib- ited opaque margins in 11 moof the year with the peak during May (54%); however, few fish were collected during the winter months (November- February). No month had a high percentage (over 75%) of fish with opaque margins, and only one month (March) lacked fish whose otoliths had opaque margins (Table 1). In recent years the use of whole otoliths for estimating the age of fish has been questioned. Beamish ( 1979) indicated that a fish's age may be underestimated using surface examination and that otolith sections are more reliable. However, we found 96.5% agreement between king mack- erel age estimates (number of opaque marks) comparing surface and sectional readings. This indicates that our age estimations for whole oto- liths are similar to those of sectioned ones. The agreement betw r een tw r o readers about the number of marks on king mackerel otoliths was 98%. The number of otoliths found to be usable was 1,449. Age and Size Composition Age composition of king mackerel varied greatly among the areas (Table 2). Younger fish were taken in northwest Florida, while older fish were caught off Louisiana, particularly in 1978. Fish of intermediate age were landed primarily in Texas, South Carolina, and North Carolina. The oldest females in our sample were 14+ yr (over 1,400 mm FL) and the oldest males were 9+ yr (970 mm FL;. Much age variation occurred within a single length group in our data (Tables 3, 4) as it did in Beaumariage's (1973) data. For example, we found females 850-899 mm FL were 1-8 yr old (Table 3). Back-Calculated Growth The weighted means of the back-calculated fork lengths for male and female king mackerel from all areas and years sampled in this study are shown in Tables 5 and 6. Differences in mean 99 FISHERY BULLETIN: VOL. 81. NO. 1 Table 1.— Percentages by month, area, and year of king mackerel otoliths having opaque margins. ( ber of fish. : total num- Area Year Jan. Feb. Mar Apr May June July Aug Sept Oct. Nov Dec. Texas 1977 — — — — — 26.7 (15) 286 0.0 (5) — - — — 1978 ~ 0.0 (5) 0.0 (17) 25 (40) — — Louisiana 1977 — — — — — 00 (4) — — 00 (15) 00 (22) 00 (18) — 1978 0.0 167 0.0 40.6 00 154 65 135 00 00 50 143 (7) (6) (43) (32) (2) (26) (62) (37) (5) (51) (20) (7) NW Florida 1977 — — — — — 18.2 (11) 94 (64) 3 1 (65) 0.0 (73) 4.3 (46) — — 1978 " " ~ 0.0 (15) 00 (160) 00 (97) 00 (107) 11 1 (135) — — 1979 " 61.2 (62) 200 (20) 19.2 (27) 00 (12) — — SE Florida 1978 — — 500 (6) 1979 83 3 (6) ~ — — South Carolina 1978 " " 2.9 (104) — — North Carolina 1978 00 (5) 63.6 (22) 38.5 (13) 26.7 (15) 3.8 (53) 8.9 (313) — — Total 38.5 16.7 0.0 40.6 543 23 7 7.2 44 08 7.2 2.6 333 (13) (6) (43) (32) (70) (118) (364) (271) (253) (671) (38) (13) Table 2.— Percentages of king mackerel by area and year within each age group, developed from age-length keys and length- frequency distributions. Area Age in years Year 10 11 12 13 14 No fish Males Texas 1977 1978 Louisiana 1977 1978 NW Florida 1977 1978 South Carolina 1978 North Carolina 1978 Total males Females Texas 1977 1978 Louisiana 1977 1978 NW Florida 1977 1978 South Carolina 1978 North Carolina 1978 Total females 0.8 2.0 0.6 — 69 24.1 24.1 2.6 1 9 13 5 165 26 9 31.3 167 20 5 93.1 2.7 1 2 04 21.1 88 21 8 136 5.2 5.2 183 35 7 488 86 105 126 _ 27 9 48.8 70 4 1 8.5 5.8 37.3 04 08 12.6 289 — 0.4 13 60 39.6 30.4 12.5 10.0 850 5.9 2 5 2 1 17 3 3.6 26.5 21.7 4.5 37 19 7 20 4 37,9 10.9 99 11 1 27 6 3.5 6 9 6.9 _____ 2 9 20.6 32.5 3.6 3.3 5.8 — — — — 533 100 — _______ 10 20.0 24 36.0 8.0 12 — — — — — 25 2 2.6 — — — 498 0.8 - — 1.107 136 19 7 1 4 — — — — — 147 20.0 8.6 3.5 3.5 — — — — 115 7 7 6.5 17 14 14 ___ 2.507 9.3 4.7 2.3 — — — — — 43 23.6 99 10.8 — — — — — — 780 30 1 10.9 6.7 2.9 6.7 — — — — 239 144 24 4 11.9 7.7 77 109 88 44 13 08 479 5.8 6 0.6 0.4 1 — — 1,393 16 8 0.1 — — 1,463 5.6 112 4 4 5 6 2.4 8 0.9 — 249 19 2 16 4 8.5 4 3.2 — 4 — — 402 86 0.3 4 2 2 1.7 1.7 1.1 07 0.2 1 5,216 length occurred from year to year and from area to area. Only data for Louisiana, however, where five or more individuals were used in computing a mean, showed the range of means within an age group to vary more than 100 mm. In 2 yr of sampling in Louisiana, over 300 fe- males were sampled, but too few males were col- lected to back-calculate size at previous ages. Generally, the Louisiana fish were also much larger than those taken elsewhere, and we con- cluded that this must be an anomalous group of fish. We separated Louisiana females from other females for growth computations, except those dealing with compensatory growth. 100 JOHNSON ET AL.: AGE. GROWTH. AND MORTALITY OF KING MACKEREL Table 3.— Length composition (%) of female king mackerel by age group (locations combined). Length group (mm FL) Age in years Total 1 2 3 4 5 6 7 8 9 10 11 12 13 14 fish 350-399 1000 1 400-449 33.3 667 6 450-499 43 5 565 23 500-549 100.0 48 550-599 100 90 600-649 96 4 36 112 650-699 77.5 197 28 71 700-749 25.3 65 1 7.2 1 2 1.2 83 750-799 30 360 430 160 20 100 800-849 24 11 36.2 31 5 134 39 1 6 127 850-899 1 6 08 18.9 33.6 320 98 25 08 122 900-949 1 11.0 22.0 250 28.0 90 4 100 950-999 2 5 23.4 31 2 260 143 1.3 1 3 77 1,000-1.049 167 23.1 34.6 11 5 64 38 2.6 1 3 78 1,050-1.099 4 1 286 265 102 102 163 4 1 49 1.100-1.149 1.9 11.5 40 4 13.5 192 7.7 5.8 52 1.150-1.199 11 9 21.4 33.3 95 95 7 1 48 2.5 42 1.200-1.249 29 15 2 21 2 21.2 9 1 15.2 6.1 9 1 33 1.250-1.299 125 8 3 42 16 7 333 8.3 16.7 24 1.300-1.349 43 4 3 130 8.7 21.7 26.3 130 87 23 1,350-1,399 50 150 30.0 350 5.0 50 50 20 1,400-1,449 26.7 133 333 200 6 7 15 1,450-1,499 143 57.1 143 143 7 1.500-1.549 1,550-1,599 50.0 500 2 Table 4.- —Length composition (%) of male king mackerel by age group (locations combined). Length Age in years Total no. (mm FL) 1 2 3 4 5 6 7 8 9 10 11 12 fish 400-449 100 450-499 15.2 84.8 500-549 100.0 550-599 983 1.7 600-649 93.0 5.3 1 7 650-699 37 5 37 5 146 104 700-749 11.9 35.7 31.0 16.6 2.4 24 750-799 111 27 8 463 130 1 8 800-849 2.0 15.4 346 21 2 192 38 38 850-899 150 50 350 30.0 100 5.0 900-949 14.2 429 429 950-999 25.0 250 250 25.0 1.000-1.049 25.0 75.0 1.050-1,199 1,200-1,249 100.0 4 33 51 60 57 48 42 54 52 20 7 4 4 1 Table 5. — Weighted means of back-calculated fork lengths (mm) for female king mackerel from all areas, 1977-78. South North Age class Texas Lou siana NW Fl 1977 orida 1978 Carolina 1978 Carolina 1977 1978 1977 1978 1978 1 487 457 504 502 463 443 415 393 II 688 673 718 714 670 687 638 627 III 777 748 824 824 755 764 750 738 IV 847 811 906 909 805 838 809 798 V '805 853 970 983 866 895 864 844 VI '849 937 990 1,045 '897 '934 916 891 VII '932 '885 '1,097 1,096 '963 941 939 VIII '1,203 1,148 996 992 IX '1.361 1,202 1.033 '1,000 X 1.252 '1,034 XI 1.311 XII 1.332 XIII '1.350 XIV '1,399 'Lengths based on less than 5 samples. 101 FISHERY BULLETIN: VOL. 81, NO. 1 Table 6.— Weighted means of back-calculated fork lengths (mm) for male king mackerel from all areas, 1977-78. South North Age class Te xas Lo jisiana NW Fl orida Carolina 1978 Carolina 1977 1978 1977 1978 1977 1978 1978 1 414 413 — 473 407 373 385 II 588 574 635 665 607 614 III 659 658 686 '734 715 702 IV 703 720 736 '746 746 747 V 747 790 '798 '769 781 VI 1 754 829 '850 '821 795 VII '803 '896 '810 VIII '789 '951 IX '943 'Lengths based on less than 5 samples Back-calculations for male king mackerel from all areas combined are shown in Table 7. Growth is rapid until the third year of life, after which time the annular growth increment de- creases and stabilizes at an average 42 mm FL. Females from the combined areas (Table 8), excluding Louisiana, also showed rapid growth in the first 3 yr, after which the annual growth increment decreased to an average 40 mm FL. Females were larger than males for all ages. Fish from Louisiana ( all females) exhibited an impressive growth rate (Table 9). They averaged 69 mm longer than other females at age 1 , and by age 10 were 218 mm longer than their counter- parts. The yearly growth increment was over 60 mm to age 6, an increment not maintained by other females, or males, past age 3 in other loca- tions. Our combined back-calculated data were com- pared with those from Beaumariage (1973) (Table 10). His data were converted to fork lengths from standard lengths (SL) using his equation: FL = 1 .096 SL - 17.143. Disregarding Louisiana females, both male and female mean Table 7.— Average back-calculated fork lengths (mm) at age for male king mackerel from all areas, 1977-78. Mean length Age at capture class (mm FL) Ag e in years N 1 2 3 4 5 6 7 8 9 1 570.3 206 425.0 II 708.6 41 4226 667 3 III 7670 41 408 5 6187 737.6 IV 772.5 44 403.3 5946 677.5 747 9 V 820.4 22 375 1 590 5 6690 733.9 796.1 VI 832.6 16 349.2 559.4 641 9 700 1 7558 808.3 VII 852 3 3 3896 579.3 6486 717 7 752.7 802.1 8383 VIII 9200 2 415.2 5788 649 2 7145 773 1 8173 8625 896.0 IX 9700 1 476.6 5603 623.3 754.1 796.2 830.2 864 7 8994 943.4 Weighted mean 376 414.1 6134 6892 734 777 4 809.3 850 8 897 1 943 4 Annual increment 199 3 75.8 448 434 31.9 41.5 46.3 46.3 Table 8.— Average back-calculated fork lengths ( mm) at age for female k ing mackerel from all areas except Louisiana, 1977-78. Age class Mean length at capture (mm FL) N Ag e in years 1 2 3 4 5 6 7 8 9 10 1 6048 315 456.4 II 741.2 112 427 9 693 8 III 8096 105 435.8 6454 7744 IV 858 7 100 426.2 6489 753 5 830.2 V 897.1 79 4993 635 3 7296 8007 8654 VI 933.7 44 405.1 630 3 727.2 791.6 848.2 9083 VII 9602 21 363 613.3 703 3 760.5 827.4 884.5 937.5 VIII 1,028 8 3924 635.0 732.5 7966 852.2 910.0 9553 1,020.9 IX 1,056.0 6 337.4 609 2 732.0 790.9 847.3 893 9 938.1 987.7 1,034.6 X 1.062 1 2 325.5 557.8 683 1 747 4 796 9 833.9 883 6 9347 9784 1,033.6 Weighted mean 792 433.9 652.0 747.1 806.5 853.5 8994 938.5 997 7 1,020.6 1,033.6 Annual increment 218.1 95.1 59.4 470 459 39.1 59.2 22.9 130 102 JOHNSON ET AL.: ACE. OROWTH. AND MORTALITY OF KINO MACKEREL oo t- 05 o S3 O) s3 3 3 bt s o T3 CD 03 -Q CD bt s3 •-_ ai > I ai w j B5 < If* O) en rf co CO o tn cm a) r-- co o r*- co o CO CO CO o r-* o> c\i t- (D O T O O CO CO 0) CO CD CO CO CO CO o 05 r~ o r~ CO CO CO C\J CO CO CO CO en en CO IT) t a> a> en CO i-~ co CM CO CD CO CO CO 0) CO o CO o CO r\j o O) CO in m r-~ o DO (0 CM CO CO r^- ■-cr o CO CO CO CO o o CO CO 0) o o 0) CD O in o CO 'CO o LO un o CO CO o m 0) o nid •- to in co o •** r- t co i c\i 1- c\i ^- co h- r- o ro n O) m ^ co cd tt cd O O G) O O CD OO O c\jc\jcocdcdcococom-c\j co m i-WNOcnmcD'-T-co t^ c\i --OlCDtDLnCMN^TrOCO CO I s - OCDCOCDOOCRCDCDOOCO CO COOltDOCOCOCOtmr-tfi CO t- NcDco^-coifii-ocDn'T oo m OJCNJCNJCMNCOinOCD^r- o CO coo^cr>cr>cocoaoa>cococo oo OTj-^h-oir-cNJcncMoaico COCOCOCOCONCONCONNN CO i- '-ocnocomoioO'-iomo) m co NlfiinCOS'-COCD^lOi-COr- tj- i-^ co^ojcMCNjT-tDcoN'-comrn i- i- r-h-r-.r*-r-r*-cococOr*-cotoco r- eg mcooO)cDNc\iocO'-oro^o co i-oniONdiococjcoiriaios c\j NOCMCO'-CNJO'tCDNOCDCOCO O mininifiifim^^ir^LnTjcoco m '-^IDOCOCO'JmOIDCMt-t-OJCNJ i- i-co^-r-c\jt-*-c\ji-*- co ocococDCMio^-ifiNcomO'-m mmdmcriQO)^'- co ct> oi -^ a> CO'-OinciMnotDNCDCDO'- CDCOCOCDOO'-CVJCMCNJCOCO'CJ-TJ 03 E — — XX - = > ~' ^ X X 5 < Table 10. — Mean back-calculated fork length (mm) at ages, from Beaumariage (1973) and this study. Beaumariage'sdata were transformed from standard length by his formula FL = 1.096 SL- 17.143. Males Females (t Beau- 'xcept La ) Beau- Johnson Johnson Age manage et al. manage et al. 1 457 414 491 434 2 643 613 703 652 3 705 689 793 747 4 752 734 857 807 5 795 777 928 854 6 822 809 986 899 7 839 851 1,033 939 fork lengths at age were smaller in our study than in his in all eases but one (7-yr old males). Several explanations for the differences seem reasonable. First, our back-calculations em- ployed a power curve, whereas his employed a linear equation. Secondly, our fish were sampled from a wide geographical range, which yielded fish with wide variation in age composition, whereas Beaumariage sampled from a more re- stricted area. Lastly, our sampling occurred almost 10 yr after his, and various changes may have occurred in the population owing to exploi- tation or other influences. Compensatory Growth Compensatory growth (Ricker 1975) appeared to occur in both male and female king mackerel. Length-frequency distributions of otolith mea- surements from the focus to the proximal edge of the first opaque mark in both sexes showed a nor- mal distribution of values. After examination of the distributions, we defined slow-growing fish (both sexes) as those with an incremented 50 om^u or less, fast-growing males as those with an in- crement of 81 om/u or more, and fast-growing fe- males as those with an increment of 86 om/i or more. Back-calculated lengths for these fish are shown in Table 11. While fast-growing males grew 525 mm in year 1, they grew only 135 mm in year 2. The slow-growing males grew 303 mm in their first year, but made up some of their size difference by growing 285 mm in their second year. Females showed a similar trend, with fast- growing fish having a first-year increment of 559 mm and a second-year increment of 184 mm. The slow-growing females grew 282 mm in year 1 and 334 mm in year 2. Beyond age 2, yearly growth increments were similar within each sex. Growth compensation in king mackerel is 103 FISHERY BULLETIN: VOL. 81, NO. 1 Table 11.— Annual fork length in- crements (mm) computed from back- calculations on fast- and slow-grow- ing male and female king mackerel (from all areas combined). Males Femai es Age Fast Slow Fast Slow 1 525 303 559 282 2 135 285 184 334 3 87 85 101 99 4 53 72 89 67 5 104 63 75 63 6 49 64 66 7 46 75 8 52 65 9 53 47 10 44 47 11 51 67 12 34 10 13 67 35 14 11 100 probably the result of an extended spawning sea- son. Long spawning seasons and multiple spawns are discussed by Beaumariage (1973) and would result in great size variation in young-of-the- year king mackerel. Some of that size variation would be decreased as the smaller fish continue to grow at a higher rate in their second year than do larger fish in their second year. Although the slow-growing fish make up some difference in size during year 2, they remain smaller than the fast growers throughout their lives. Theoretical Growth The von Bertalanffy theoretical growth param- eters computed from back-calculated fork lengths are shown in Table 12, along with those reported by other authors. The von Bertalanffy (1938, 1957) growth equation is the following: 1* = L„(l -*<< to)\ where 1; = length at age t, Loo = asymptotic length, k = growth coefficient, and Table 12.— von Bertalanffy growth parameters for king mackerel. k t. fo Author value (mm FL) (yr) Males Johnson et al ., all areas 0.28 965 -1.17 Beaumariage (1973) 0.35 903 -2.50 Nomura and Rodrigues (1967) 0.18 1,160 -0.22 Females Johnson et al.. excl La 0.29 1,067 -0.97 Johnson et al.. La. 0.14 1,529 -2.08 Beaumariage (1973) 0.21 1,243 -2.40 Nomura and Rodrigues (1967) 0.15 1,370 -0.13 to = time when length would theoreti- cally be zero. Our theoretical growth parameters are be- tween those calculated by Beaumariage (1973) and Nomura and Rodrigues (1967). Beaumar- iage 's theoretical growth parameters were calcu- lated by employing observed sizes offish at each age, while Nomura and Rodrigues apparently combined both back-calculated lengths and em- pirical lengths in their calculations. We employed mean back-calculated lengths at age in our com- putations, which may account for some of the dif- ferences between our values and those of the other investigators. Length-Weight Relationship The length-weight values for king mackerel computed for the equation W = a L b , where W is weight in grams and L is fork length in milli- meters, are presented in Table 13. Male length- weight values from our study were within the confidence intervals set by Beaumariage (1973), but for both our female and combined sexes, length-weight values were below his lower con- fidence intervals. Mortality Mortality estimates are presented in Table 14. The mean annual mortality rate (A = 0.37) is low- er than Beaumariage's (1973) estimate (A = 0.54). We feel that our results are more concor- dant with generally accepted techniques of catch- curve analysis, in that our catch-curves were de- veloped from age-frequency data, as opposed to the length-frequency catch-curve used by Beau- mariage. We also feel that our results are less in- fluenced by the effects of gear selectivity than Beaumariage's results, since Trent et al. (1981) stated that commercial hook-and-line gear ex- cludes small and large king mackerel to a great- er extent than does recreational hook-and-line gear. Nevertheless, there are many difficulties in using catch-curve analysis in our study. Spe- cific problems are related to the Beverton and Holt ( 1957 ) and Robson and Chapman ( 1961 ) tech- niques. The first technique involves using sev- eral consecutive years of data, which were un- available in our study. With the second technique, we used age-length keys as the basis for our catch-curves but were unable to make correc- tions for the bias when such keys were used ( Rob- 104 JOHNSON ET AL.: AGE, GROWTH. AND MORTALITY OF KING MACKEREL Table 13.— Summary of length-weight relations of U.S. king mackerel. W - weight ir i gram s; L = tc >rk : le n) *th in mill lmeters. No Range (mm FL) w a L" 95% confidence i interval Correlation Sex fish 3 b Lower Upper ir) Male 701 428- 1.355 08064 X 10 -5 2.9928 2.9572 30284 0.9909 Female 2.023 351- 1,554 08801 • 10 29827 29562 30092 0.9910 Sexes combined 2.821 351- 1,554 8464 X 10 29881 3.0153 30153 0.9899 Table 14.— Estimated annual mortality rate (A) by estimation technique, assuming 1:1 and 1:2 male:female ratios. Estimation technique Male: Female ratio Heincke (1913) Jackson (1939) Rounsefell Beverton & Everharl & Holt (1953) (1957) Robson & Chapman (1961) Regression analysis Mean A 1:1 1:2 0.35 034 034 035 42 0.42 042 042 032 033 0.35 036 037 0.37 son and Chapman 1961). This was a result of the age-length keys being developed for a different fish sample than the one being analyzed for mor- tality rates. The difficulties in applying Robson and Chapman's technique resulted in an implica- tion that king mackerel are not fully recruited into the south Florida recreational fishery until age 7, after which the annual mortality rate is 0.53. This mortality estimate is similar to Beau- mariage's (A = 0.54), but the age at recruitment was found by Beaumariage to be 2-3. His esti- mate was based on a smaller age range (0-7) than was ours. This difference probably influenced the resulting mortality estimates. Many difficulties are also involved in the basic concept of using catch-curve analysis to estimate mortality in king mackerel. Rounsefell and Ever- hart (1953) emphasized that catch-curve analy- sis is based on false assumptions when applied to most pelagic species, including mackerel. Rob- son and Chapman (1961) reiterated this warning, stating, "if year classes.. .vary in strength and survival rates vary from year class to year class and age to age, then the age-frequency distribu- tion in the catch of a single season provides no identifiable information whatsoever regarding [mortality rates]...." These comments force us to state our mortality findings with some wariness. ACKNOWLEDGMENTS We thank Michael Crow, Mark Farber, and Dennis Lee of the National Marine Fisheries Service, Miami Laboratory, Miami, Fla., for their constructive reviews of this manuscript. ADDENDUM Fischer (1980) reported on the length-weight relationship of king mackerel off Louisiana. His length-weight values are similar to ours. LITERATURE CITED Abramson, N. J. 1971. Computer programs for fish stock assessment. FAO Fish. Tech. Pap. 101, 149 p. Beamish. R. J. 1979. Differences in the age of Pacific hake (Merluccius productus) using whole otoliths and sections of otoliths. J. Fish. Res. Board Can. 36:141-151. Beaumariage, D. S. 1973. Age, growth, and reproduction of king mackerel, Scomberomorus cavalla, in Florida. Fla. Mar. Res. Publ. 1. 45 p. Beverton, F. J. H., and S. J. Holt. 1957. On the dynamics of exploited fish populations. Fish. Invest. Minist. Agric. Fish. Food (G.B.), Ser. II. 19, 533 p. Everhart, W. H., A. W. Eipper, and W. D. Youngs. 1975. Principlesof fishery science. Cornell Univ. Press, Ithaca. N.Y., 288 p. Fischer, M. 1980. Size distribution, length-weight relationships, sex ratios, and seasonal occurrence of king mackerel (Scom- beromorus cavcdla) off the southeast Louisiana coast. La. Dep. Wildl. Fish. Tech. Bull. 31:1-21. Heincke, F. 1913. Investigations on the plaice. General report. I. Plaice fishery and protective measures. Preliminary brief summary of the most important points of the re- port. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 16, 67 p. Jackson, C. H. N. 1939. The analysis of an animal population. J. Anim. Ecol. 8:238-246. 105 FISHERY BULLETIN: VOL. 81. NO. 1 MANOOCH. C. S., III. 1979. Recreational and commercial fisheries for king- mackerel, Scomberomorus cavalla, in the South Atlantic Bight and Gulf of Mexico, U.S.A. In E. L. Nakamura and H. R. Bullis, Jr. (editors), Proceedings of the mack- erel colloquium, p. 33-41. Gulf States Mar. Fish. Comm., Brownsville, Tex. Manooch, C. S., Ill, E. L. Nakamura, and A. B. Hall. 1978. Annotated bibliography of four Atlantic scom- brids: Scomberomorus brasiliensis, S. cavalla, S. macu- latus, and S. regalis. U.S. Dep. Commer., NOAA Tech. Rep. NMFSCirc. 418, 166 p. Nomura, H.. and M. S. S. Rodrigues. 1967. Biological notes on king mackerel, Scomberomorus cavalla (Cuvier), from northeastern Brazil. Arquivos do Estocao de Biologia marinha do Universidade Fed- eral do Ceara. 7(l):79-85. RlCKER. W. E. 1975. Computation and interpretation of biological sta- tistics of fish populations. Fish. Res. Board Can., Bull. 191, 382 p. Robson, D. S., and D. G. Chapman. 1961. Catch curves and mortality rates. Trans. Am. Fish. Soc. 90:181-189. ROUNSEFELL, G. A., AND W. H. EVERHART. 1953. Fishery science: its methods and applications. John Wiley and Sons, Inc., N.Y., 444 p. Tesch, F. W. 1971. Age and growth. In W. E. Ricker (editor). Meth- ods for assessment of fish production in fresh waters. 2d ed., p. 98-130. Blackwell Sci. Publ., Oxford. Trent, L., R. 0. Williams, R. G. Taylor, C. H. Saloman, and C. S. Manooch III. 1981. Size and sex ratio of king mackerel. Scomberomor- us cavalla, in the southeastern United States. U.S. Dep. Commer.. NOAA Tech. Memo. NMFS-SEFC-62, 59 p. National Marine Fisheries Service, Panama City, Fla. von Bertalanffy, L. 1938. A quantitative theory of organic growth (inquiries on growth laws. I). Hum. Biol. 10:181-213. 1957. Quantitative laws in metabolism and growth. Q. Rev. Biol. 32:217-231. 106 REVIEW AND ANALYSIS OF THE BLUEFIN TUNA, THUNNUS THYNNUS, FISHERY IN THE EASTERN NORTH PACIFIC OCEAN Doyle A. Hanan 1 ABSTRACT Northern bluefin tuna migrate from waters near Japan to the eastern North Pacific where they are fished primarily by purse seine. While annual catches fluctuate greatly, two major periods are identified. The average annual catch in the second period ( 1950-present) is nearly double that for the first period (1921-50) and is attributed to increased fishing effort by the "high-seas" tuna fleet oper- ating off Baja California. The declining catch per unit effort in the second period and declining catches after 1963 are assumed to indicate declining abundance of bluefin tuna in the eastern North Pacific. Length-frequency analysis reveals 1) significantly smaller bluefin tuna in U.S. waters than in waters off Baja California and 2) significant variation in mean lengths among years. Analysis of tag-recapture data confirms seasonal northward migration and vulnerability to the fishery for as many as three fishing seasons. A catchability coefficient of 1.66 X l(T 4 /boat-day and an annual instantaneous total mortality rate of 2.07, both estimated from the tag-recapture data, are used with summaries of fishing effort to calculate an average annual exploitation rate of 30% for bluefin tuna in the eastern North Pacific. Purse seining for northern bluefin tuna, Thun- nus thynnus Linnaeus, in the eastern North Pa- cific Ocean began about 1914, with the first large commercial landings in 1918 (Whitehead 1931). Prior to the development of this purse seine fish- ery, a sport fishery existed off southern Califor- nia at Santa Catalina Island; and since bluefin tuna are difficult to catch by hook and line, elab- orate fishing methods evolved such as using a kite to make the bait (flying fish) skip across the water (Clemens and Craig 1965). The Tuna Club of Avalon at Santa Catalina Island even awarded "blue buttons" to its members for catching the large and wary prize. Because of this difficulty in hooking bluefin, the commercial "high-seas" fleet did not fish for bluefin until the late 1950's, when most of the fleet had converted from pole- and-line gear to purse seines (Bell 2 ). Currently the bluefin fishery consists of a "wet- fish" fleet, principally out of San Pedro, Calif.; a high-seas fleet mostly out of San Diego, Calif.; and since 1975, an expanding Mexican fleet most- ly out of Ensenada, Baja California. The bluefin 'California Department of Fish and Game, c/o Southwest Fisheries Center, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038. 2 Bell, Robert R. 1970. Bluefin tuna Thunnus thynnus ori- entalis in the northeastern Pacific Ocean. Unpubl. manuscr. Calif. Dep. Fish Game, 350 Golden Shore, Long Beach, CA 90802. Manuscript accepted Julv 1982. FISHERY BULLETIN: VOL. 81, NO. 1. 1983. fishery extends along the coast of North America from Cabo San Lucas, Baja California, to Point Conception, Calif., and occasionally farther north (Table 1). The bluefin catch is composed mainly of 1-, 2-, and 3-yr-old fish, which appear to mi- grate to the eastern North Pacific from the west- ern Pacific near Japan (Schultze and Collins 1977); however, older and much larger bluefin are reported and occasionally caught in the east- ern North Pacific. This paper reviews and analyzes the bluefin tuna fishery in the eastern North Pacific, using data collected by the California Department of Fish and Game (CFG) in cooperation with the Inter-American Tropical Tuna Commission (IATTC), and the National Marine Fisheries Service (NMFS) of the U.S. Department of Com- merce. CATCH AND EFFORT ANALYSIS Although annual bluefin catches have fluctu- ated considerably in the eastern North Pacific (Table 2), two major periods are identified in the catch by a plot of a 10-yr running average (Fig. 1). During the first period, about 1921-50, total landings averaged 5,066 t (metric tons)/yr and were declining toward the end of the period. During this time, bluefin were landed almost ex- 107 FISHERY BULLETIN: VOL. 81, NO. 1 Table 1.— Total number of months in which bluefin tuna catch exceeded 50 t within a 1° area of latitude and longitude for the years 1957-69 and 1974. Each latitude and longitude indicates the southeast corner of the 1° area of considera- tion. Asterisks indicate the coastline. Lati- Long tude tude 125 120 119 118 117 116 115 114 113 112 111 110 40 1* 34 1 * 33 1 3 7 2* 32 7 12 14* 31 3 11 5* 30 8 11 • 29 3 6 2 • 28 7 3 7 * 27 1 8 4- 26 2 10 7 • 25 1 12 6* 24 6 8 1* 23 4 3 1* Table 2.— Total landings of bluefin tuna by commercial (in metric tons (t)) and sport fisheries (no. fish) in the eastern North Pacific Ocean, 1918-81. Asterisks indicate no data available. Land ngs Landings Commer- Sport Commer- Sport Year cial (t) (no. fish) Year cial (t) (no. fish) 1918 2,722 1950 1.242 27 1919 6.800 1951 1,752 7,142 1920 4,776 1952 2,076 145 1921 894 1953 4.433 4.276 1922 1.275 1954 9,537 966 1923 1.460 1955 6,173 8.179 1924 1.470 1956 5,727 34.187 1925 1,725 1957 9,215 6.428 1926 2.960 1958 13,934 884 1927 2.222 1959 6,914 1,330 1928 6.215 1960 5.422 97 1929 3.414 1961 9,603 2,268 1930 9.943 1962 14,651 2,453 1931 1.603 1963 14,189 737 1932 486 1964 10.642 693 1933 254 1965 7,556 92 1934 8,327 1966 16,846 1.998 1935 11.418 1967 6,601 3.166 1936 8.584 2,920 1968 6,063 1,231 1937 5.758 4,020 1969 7,172 1.470 1938 8.041 11,927 1970 4,024 1,833 1939 5,369 9,909 1971 8,415 749 1940 9,058 6,878 1972 13.390 1,470 1941 4,318 1973 10,576 5,347 1942 5,826 1974 5,748 5.765 1943 4,617 1975 9,578 3,348 1944 9,228 1976 10,561 2,040 1945 9.341 1977 5.151 1,838 1946 9,993 528 1978 5.903 479 1947 9,452 2,194 1979 6,743 1.087 1948 2,961 104 1980 3,128 729 1949 1,991 1,841 1981 1,016 clusively by the San Pedro wetfish fleet, which seasonally targets fishing effort on sardines, an- chovies, mackerel, bonito, bluefin tuna, and other fishes, depending on fish availability, mar- ket price, and market demand (cannery orders). During the second period, about 1950-present, annual landings increased to 16,846 t in 1966, then declined to 1,016 1 in 1981, averaging 9,076 1 for the period. At the beginning of this period, many of the high-seas boats that had converted to purse seining began catching large numbers of bluefin off Baja California, although they target- ed their fishing on yellowfin tuna, Thunnus alba- cares, and skipjack tuna, Katsuwonus pela- m is. Two sources of data were used in summarizing total catch by area. The first, landings reported by CFG, separates pounds landed in California for 1918-79 into those caught in California wa- ters and those caught south of California waters. These data reveal an overall decreasing trend in bluefin catch north of the international border; and, until about 1963, there was an overall in- creasing trend in total catches south of the inter- national border (Fig. 2). The second source of catch data also includes effort information and was compiled into a data base for summary and analysis. These data, rep- resenting about 87% of the catch during the peri- ods 1954-69 and 1971-74, came from summaries of skippers' logs, from interviews with skippers and engineers, from CFG landing receipts, and from IATTC summaries of the high-seas fleet. Catch and effort in the data base are recorded by 1° areas of latitude and longitude. For this study, one boat-day or part of a boat-day of effort is as- signed to a seiner for each day or partial day of purse seining or searching for tuna in the bluefin fishing range (north of lat. 22°N) during months in which bluefin were caught. Catch data, sum- marized by areas north and south of lat. 32°N (the parallel nearest the international border), show trends similar to the reported California landings for the same years (Fig. 3). For comparison with the CFG reported Cali- fornia landings and for future consideration of the effects of Mexican regulations concerning 108 HANAN: BLUEFIN TUNA FISHERY 18,000 15,000 BLUEFIN TUNA • • ANNUAL CATCH o o 10-YEAR RUNNING AVERAGE 1925 1945 1950 YEAR 1955 1970 FIGURE 1.— Annual catches of northern bluefin tuna in the eastern North Pacific Ocean for the years 1918-81. I8,000r 1920 1930 1940 1950 1960 1970 1980 YEAR Figure 2.— Annual California landings (metric tons) of north- ern bluefin tuna caught north and south of the United States- Mexico international border, 1918-79. South is represented by solid circles and lines, north by open circles and broken lines. the 200-mi exclusive economic zone, much of the data in this paper are separated into areas north and south of lat. 32°N. A better division from a biological standpoint would be north and south of lat. 29°N (Fig. 4). The increase in California landings of bluefin caught south of the border during the 1957-66 period can be attributed to increased fishing effort, but the decline in catch north of the border cannot be explained by declining effort, since ef- fort remained comparatively level throughout the period (Fig. 5). Because bluefin are valuable ($l,180/short ton in 1981) and because fishing ef- fort north of the 32d parallel remained fairly con- stant, the decline in northern catches is attrib- uted to a decrease in abundance in that area. During this period, increased catches south of the border appear to have offset the decline in catches to the north and to indicate the fish were intercepted before migrating northward. If this is true, the recent catch decline south of the bor- der indicates declining bluefin abundance in the eastern North Pacific. Catch and effort data summarized by latitude show a bimodal distribution centering just north of the 25th and 32d parallels (Table 3). The catch- es are concentrated in the period June-Septem- ber, with the largest catches shifting northward during the fishing season (Fig. 4). Early in the 15,000 12,500 - p 10,000h I 7,500 - o < <-> 5,000 2,500 \ ,\ / \ NORTH v v V __----" _L J_ 1957 1960 1963 1966 1969 1972 1975 YEAR Figure 3.— Logged annual catches (metric tons) of bluefin tuna north and south of lat. 32°N for 1957-69 and 1974. 109 FISHERY BULLETIN: VOL. 81. NO. 1 MONTH JFMAMJJASOND 6,000 <25 25-300 >300 Figure 4. — Mean monthly catches of bluefin tuna (^_) for the 13 period 1957-69 in metric tons per latitude. Totals are for areas between a given parallel and the next higher parallel. Table 3.— Mean catch of bluefin tuna (metric tons (t)) and mean effort (boat-days) per latitude for the years 1957-69 and 1974. Mean catch Mean effort Latitude (t) (boat-days) 36 3 29 1.16 35 000 0.29 34 2306 6.67 33 452.60 75.27 32 2,160.34 460.39 31 1.048.67 214.39 30 62263 155.11 29 544.82 13596 28 561.82 170 37 27 61546 253.89 26 734.33 359.86 25 981.61 62956 24 543.07 348.16 23 234.57 353 10 22 0.44 122.02 season there are catches both in northern and southern parts of the bluefin range, whereas there are relatively few catches late in the season in the southern part of the range, thus indicating northward movement. This shift is also apparent in the number of occurrences of recorded bluefin 1957 1960 1963 1966 1969 1972 1975 YEAR Figure 5. — Logged annual effort (boat-days/year) for bluefin tuna north and south of lat. 32°N for 1957-69 and 1974. catch per month and latitude during the 1957-69 period (Table 4). The northward shift in location of the largest catches does not reflect a shift in fishing effort, since effort remains high in the south throughout the season (Fig. 6). Apparently, bluefin move northward or there is a shift in bluefin vulnerability towards the north during the fishing season. Catch and effort data for 1957-69 summarized by vessel size indicate that seiners of 101-300 ton capacities accounted for more than 70% of blue- fin landings and that smaller vessels tended to be phased out of the fishery and replaced by larger ones (Table 5). CATCH-PER-UNIT-EFFORT ANALYSIS Catch per unit effort (CPUE) is calculated for each year as total catch divided by total effort (Table 6). The relationship between the CFG data and the IATTC data (Bayliff and Calkins 1979) is expressed as a ratio which includes the origin. The ratio estimator 2,y/%x, obtained from years for which both CFG and IATTC measures of CPUE are available (1966-74), yielded a value of 1.01, by which the CFG values (1954-65) were multiplied to obtain IATTC equivalents. These equivalent CPUE values were then plotted, and a regression line fit to them reveals a decline in CPUE with time (Fig. 7). This observed decline is probably conservative because fishing effort, which was not standardized, has most likely be- come more effective with time (Pella and Psaro- pulos 1975). CPUE values were highest in the northern 110 HANAN: BLUEFIN TUNA FISHERY Table 4. — Total occurrences of recorded bluefin tuna catch per latitude and month during 1957-69 and 1974. Totals are for the areas between a given parallel and the next higher par- allel. Lati- M Dnth tude J F M A M J J A S O N D Total 36 1 1 2 35 34 2 2 1 5 33 2 1 5 9 9 4 2 32 32 1 3 2 6 11 13 12 4 2 1 55 31 2 9 11 8 1 31 30 8 10 6 2 26 29 3 1 1 3 9 6 1 2 26 28 7 5 4 5 7 9 11 7 4 1 3 2 65 27 3 9 7 2 1 1 23 26 9 10 1 1 21 25 1 3 11 11 3 1 30 24 3 10 7 20 23 1 3 6 4 14 22 1 1 2 Total 8 8 5 13 19 65 101 70 41 13 6 3 MONTH FMAMJ JASOND LU Q <1 1-3 >3 Figure 6.— Mean monthly effort for bluefin tuna (Jil.) for the 13 period 1957-69 in boat-days per latitude. Totals are for areas between a given parallel and the next higher parallel. part of the bluefin range (Figs. 8-10), and is at- tributed to differences in searching and fishing methods between the wetfish and the high-seas fleets. Because effort data are not available for the Y = 10.5-0.1 14X 1955 1960 1965 1970 1975 1980 YEAR Figure 7.— Annual catch per unit effort (metric tons/boat- day) of bluefin tuna plotted by year ( 1954-78) with a regression line fitted to the curve. MONTH J FMAMJ JASOND 0*5 6-75 >75 Figure 8.— Mean monthly catch per unit effort of bluefin tuna m < i 'per latitude (metric tons/boat-day) for the period 1957-69. 13 Totals are for areas between a given parallel and the next high- er parallel. 1979-81 period, the rapid decline in total catches during those years cannot be explained by direct 111 FISHERY BULLETIN: VOL. 81, NO. 1 Table 5. — Yearly catch (metric tons) of bluefin tuna and effort (in parentheses) by vessel size class for 1957-69 and 1974, from logbook data. Hold capacity in short tons: 0-50 = Class 1; 51-100 = Class 2; 101-200 = Class 3; 201-300 = Class 4; 301-400 = Class 5; over 400 = Class 6. Vessel class Year 1 2 3 4 5 6 Total 1957 205 2,537 4,614 35 — — 7,391 (20) (328) (679) (24) (-) (-) (1,051) 1958 276 1,840 6,767 646 — — 9.529 (37) (433) (1.266) (42) (-) (-) (1.778) 1959 164 1.912 2,468 522 330 — 5,396 (29) (408) (1,352) (267) (117) (-) (2,173) 1960 5 287 2,318 1,067 1.081 69 4,827 (33) (194) (1.495) (730) (341) (39) (2,832) 1961 21 526 5.325 2,331 1.015 4 9,222 (4) (171) (2.222) (925) (352) (12) (3,686) 1962 14 959 7,061 2,840 1.498 — 12,372 (8) (185) (2,447) (1,515) (603) (32) (4,790) 1963 — 544 5,483 4,228 3,055 87 13,397 (-) (85) (1,667) (1.729) (1,051) (56) (4,588) 1964 18 523 3,641 2,937 1.565 — 8,684 (4) (77) (1.367) (1,577) (749) (19) (3,793) 1965 60 294 2,538 2,242 1.312 36 6,482 (12) (51) (1.641) (1.338) (753) (13) (3.808) 1966 21 429 5,576 5,400 3,107 561 15,094 (16) (112) (1,479) (1,299) (580) (38) (3,524) 1967 60 289 1.318 2,530 1,804 51 6.052 (10) (33) (1.103) (1.936) (1.435) (270) (4.787) 1968 — 399 2,038 1,481 1,300 293 5.511 (-) (69) (1.162) (895) (493) (71) (2,690) 1969 32 175 3,370 2,338 605 448 6,968 (7) (40) (1.200) (1,280) (479) (232) (3,238) 1974 60 257 1,712 905 719 677 4,330 (3) (81) (672) (450) (500) (251) (1.957) Total 936 10,971 54,229 29,502 17,391 2,226 115,255 (183) (2,267) (19,752) (14,007) (7,453) (1.033) (44,695) Total 1% 10% 47% 26% 15% 2% 100% (0.4%) (5%) (44%) (31%) (17%) (2%) ( 1 00%) 14 12 10 3 a. o ' NORTH 1957 1960 1963 1966 1969 1972 1975 YEAR Figure 9. —Annual catch per unit effort of bluefin tuna (metric tons/ boat-day) for 1957-69 and 1974 north and south of lat. 32 °N. CPUE evidence. However, if it is assumed that effort remained at about the same levels, CPUE 6.0 4.5 2 3.0 - 1.5 - 0.0 L z> a. O 20° _L 24° 28° 32° LATITUDE 36° 40° Figure 10.— Bluefin tuna catch per unit effort (metric tons/ boat-day) by latitude for 1957-69 and 1974. Latitude area is that lying between a given latitude and the next higher lati- tude. would have declined by an even greater rate than that predicted by the trend in Figure 7. This in- dicates that bluefin abundance in the eastern North Pacific has declined severely. 112 HANAN: BLUEFIN TUNA FISHERY Table 6.— Bluefin tuna CPUE values from this study (CFG) for the years 1954-74 and from IATTC (Bayliff and Calkins 1979) for the years 1954-78. CFG values converted to IATTC equivalent values are in parentheses. CPUE values CPUE values Year CFG IATTC Year CFG IATTC 1954 4 49 (4.55) 1967 1.26 1 63 1955 5.44 (5.52) 1968 2.05 2.35 1956 3.59 (364) 1969 2 15 1.96 1957 7 03 (7 13) 1970 — 1 71 1958 5.36 (5 44) 1971 2 31 2.11 1959 2.48 (2.52) 1972 3.61 3.23 1960 1 70 (1 72) 1973 3.15 289 1961 250 (2.54) 1974 2 21 1 75 1962 258 (2.62) 1975 2.73 1963 292 (2.96) 1976 298 1964 2.29 (2.32) 1977 1 86 1965 1 71 (1.73) 1978 1.62 1966 428 5.40 LENGTH-FREQUENCY ANALYSIS Length-frequency data summaries (Figs. 11- 14) were obtained from two CFG data sets of fork-length samples taken as frozen bluefin were unloaded at Terminal Island, Calif., canneries. Set 1 (1952-65) represents random samples of 50 fish/seiner; set 2 (1963-71 and 1974) represents random samples of 20 fish for every 200 short tons landed from each 1° area of latitude and lon- gitude. Set 2 samples were taken for an age de- termination study. Although a smaller number of bluefin were sampled, they appear to repre- sent the same population as the first data set, z O 4 a LU a 2 1952 N=1,140 60 80 100 120 140 160 180 120 140 160 180 60 80 100 120 140 160 1 180 O 4 1954 N = 3,680 ML T — i 1 60 80 100 120 140 160 180 60 80 100 120 140 160 1! FORK LENGTH (cm) FORK LENGTH (cm) Figure 11.— Bluefin tuna percent length frequencies, 1952-57. 113 FISHERY BULLETIN: VOL. 81. NO. 1 60 80 100 120 140 160 180 60 80 100 120 140 160 180 10 8 - Z 6 HI o a LU 4 1960 N = 3.498 80 100 120 140 FORK LENGTH (cm) FORK LENGTH (cm) FIGURE 12.— Bluefin tuna percent length frequencies. 1958-62. 1 r 160 180 Table 7. — Mean length frequencies of bluefin tuna, north and south of lat. 32°N, 1952-65. Mean length Year South North Combined 1952 737 70.2 732 1953 67.1 63.8 68.1 1954 79.7 66.3 76.3 1955 83.1 72.3 78.8 1956 90.4 65.8 83.1 1957 83.7 71.5 73.0 1958 81.3 77.7 78.6 1959 85.8 90.6 90.3 1960 112.1 968 105.6 1961 72.3 71 2 71 7 1962 73.5 64.0 68.8 1963 80.3 68.9 76.4 1964 70.4 62.8 679 1965 79.8 658 76.0 when overlapping years (1963-65) and composite samples for both data sets are compared (Fig. 15). Analysis of fish lengths from the first data set shows a decrease in mean length with increasing latitude. These data (1952-65) were also sum- marized by year for areas north and south of the 32d parallel (Table 7) for a two-way analysis of variance. The analysis shows significant differ- ences (P<0.01) among years and between areas. These results show that bluefin caught in the north are smaller than those to the south (Fig. 16) and that mean lengths vary considerably, as much as 39.8 cm/yr. 114 HANAN: BLUEFIN TUNA FISHERY 1963 LWA N = 1,533 Ih-Mj ' I 1— 120 140 160 180 14 - 1964 N = 4,812 12 J 10 PERCENT O) 00 4 1 2 A %■■ i i T " T 1 f 60 14 I 1964 LWA N- 1,446 12 10 8 1 6 1 4 1 2 L l if " I i i 80 100 120 140 160 180 60 80 100 120 140 160 180 O 4 100 120 140 160 180 FORK LENGTH (cm) 4 - 1965 LWA N = 1,380 120 140 160 180 FORK LENGTH (cm) FIGURE 13.— Bluefin tuna percent length frequencies, 1963-65. Graphs to the left are based on length-frequency samples only, whereas those to the right are based on length-weight-age frequency samples. TAGGING DATA ANALYSIS From 1953 to 1958, 186 bluefin were tagged and released by CFG and IATTC in the eastern North Pacific incidental to tagging other spe- cies. From 1962 to 1968 a tagging cooperative of CFG, U.S. Bureau of Commercial Fisheries (NMFS), and the Mission Bay Research Founda- tion of San Diego tagged and released 2,836 blue- fin. Of these, 565 (20%) were recaptured in the eastern North Pacific, including 7 by sport fish- ing and 9 in the western Pacific (Clemens and Flittner 1969). Bluefin for tagging were caught by purse seine and tagged with spaghetti-loop tags prior to 1960 and with spaghetti-dart tags since then. Bluefin are caught within about 200 mi of the coast, thus spatial analysis of tag returns is ex- pressed only by latitude. Of the 565 tagged blue- fin caught in the eastern North Pacific, recovery latitude information is available for 540 returns. Data from tagged fish recovered during the sea- son in which they were released (62%) show a general movement northward (Table 8); how- ever, many were caught near the release point and to the south (Table 9). Tagged fish recaptured during the second and third fishing seasons after tagging were well dispersed throughout the fish- 115 FISHERY BULLETIN: VOL. 81. NO. 1 O 4 0. 2 60 80 100 120 140 160 180 60 80 100 120 140 FORK LENGTH (cm) 160 60 80 100 120 140 160 60 80 100 120 140 160 180 60 80 100 120 140 160 FORK LENGTH (cr Figure 14.— Bluefin tuna percent length frequencies for 1966-71 and 1974 taken from length-weight-age samples. 180 Table 8. — Bluefin tuna tags returned during tagging season (1958-68) summarized by latitude of release and of return. Totals are for areas be- tween a given parallel and the next higher parallel. Return latitude latitude 33 32 31 30 29 28 27 26 25 24 Total 33 13 28 41 32 85 28 1 1 115 31 20 16 2 2 40 30 16 19 7 4 2 48 29 1 2 1 3 7 28 2 1 1 5 1 10 27 2 8 5 1 1 17 26 25 24 1 1 2 7 23 2 22_ 58 Total 13 154 72 18 7 16 9 23 2 22 336 116 HANAN: BLUEFIN TUNA FISHERY 4 - 10 60 80 100 120 140 160 180 H 3 Z UJ o DC UJ 0. 2 1963-1971 and 1974 N = 9,980 ,1,LJJ " 1 60 80 100 120 140 160 180 FORK LENGTH (cm) Figure 15.— Composite bluefin tuna percent length frequen- cies. Upper graph summarizes length-frequency samples for 1952-65. and lower graph summarizes length-weight-age sam- ples for 1963-71 and 1974. ing grounds and fishing season, indicating good mixing with the untagged population. Gulland (1963) described a method of estimat- ing fishing mortality from tagging experiments; this method was modified and applied to the z UJ o °- 4 NORTH OF 32° N. N = 30.444 60 80 100 120 140 160 180 10 8 - z UJ o oc UJ 0. 4 SOUTH OF 32° N. N=31,773 60 80 100 120 140 160 180 FORK LENGTH (cm) Figure 16.— Bluefin tuna percent length-frequency compos- ites for 1952-65, north (top) and south (bottom) of lat. 32°N. bluefin data. It was assumed that the number of tags returned per unit of effort is proportional to the CPUE, and no provision was made for immi- gration or emigration. For any period following tagging, an estimate of catchability (q) would be the number of tags returned per unit of effort di- vided by the initial number released. When these Table 9. — Total number of returned bluefin tags summarized by latitude of release and of return. Totals are for areas between a given parallel and the next higher parallel. Release Retu rn latitude latitude 33 32 31 30 29 28 27 26 25 24 23 Total 33 13 28 1 4 3 4 3 9 3 68 32 87 30 5 9 5 7 6 20 7 1 177 31 21 17 3 2 4 1 2 8 1 1 60 30 17 27 14 10 3 2 6 8 3 90 29 3 2 9 2 8 4 28 28 5 2 7 7 8 5 2 36 27 2 8 5 4 1 1 21 26 25 24 1 1 2 7 25 2 22. 60 Total 13 163 87 41 34 37 29 39 57 38 2 540 117 FISHERY BULLETIN: VOL. 81. NO. 1 estimates are plotted against time, the intercept at time zero is an estimate of (/ for bluefin in the eastern North Pacific. As tagged bluefin were not fully dispersed dur- ing the season of tagging, monthly estimates of q were calculated as the monthly mean, per 1° area of latitude and longitude, for 1° areas from which tagged fish were caught. For the second and third seasons, when tagged fish appeared to be fully dispersed, monthly estimates were calcu- lated for the entire bluefin range; then, the nat- ural logarithms of these values and those for the first season were plotted (Fig. 17). Effort and therefore q are expressed in boat-days. The re- gression line fitting these points (Y= —8.7363 — 0. 1725 X, R 2 « 68%) was weighted by the number of tagged fish released each year, since the num- ber of tagged fish varied between 35 and 960/yr. The best estimate of q from the tag-recapture data is the antilogarithm of the regression line intercept, 1.66 X 10~ 4 /boat-day with a 95% con- fidence interval of 0.99 X 1(T 4 to 2.63 X KT 4 / boat-day corrected for geometric mean bias (Beauchamp and Olson 1973). The slope of the re- gression (—0.17, S 2 = 0.02) is an estimate of the monthly instantaneous mortality coefficient (Z), and was expanded to estimate the yearly instan- taneous mortality (Z = 2.07, S 2 = 0.24) including immigration and emigration. This estimate com- pares favorably with Bayliff and Calkins' (1979) and Bayliff s (1980) estimates (Z = 2.08, S 2 = 0.8) for 1962-66. They call these estimates "rates 1 2 8 9 10 11 12 13 RETURN MONTHS 222324 Figure 17.— Natural logarithms of adjusted return rates for tagged bluefin tuna plotted against number of months between tagging and recapture, for the years 1962-64, 1966, and 1968. The predicted catc liability coefficient (q) from straight-line re- gression and the 95% confidence interval around (§) are shown at the zero-month intercept. of attrition," since immigration and emigration are included. The ratio of fishing mortality to instantaneous total mortality is an estimate of the exploitation ratio (Ricker 1975) and was calculated as a mean for the period 1962-70 because q was also calcu- lated for that period. The mean annual fishing effort in that period was 4,215 boat days which, multiplied by (/, estimates a fishing mortality of 0.7/yr. Dividing this value by estimated Z(2.07/ yr) yields an exploitation ratio of 0.34, and then multiplying by the annual mortality or "attri- tion" (0.87) yields a 30% exploitation rate. DISCUSSION The review and analysis of data concerning the bluefin tuna fishery in the eastern North Pacific show large fluctuations in the catch to be a major part of two important phases. The decline in catch near the end of the first phase (1921-50) is offset by the development of a "high seas" purse seine fleet and the resultant increased catch of bluefin off Baja California. The current decline (1963-present) is probably due to a decline in the abundance of bluefin as indicated by CPUE evi- dence. The effect on the resource of Mexico's 200- mi regulations was not assessed at this time; how- ever, the apparent decline in catch and CPUE cannot be attributed to such regulation since it has been enforced only recently. The declines in catch and CPUE in the eastern North Pacific are significant and are reflected by an even greater decline in catch and nomi- nal CPUE in the western Pacific (Figs. 18, 19). 50 O o o I o 40 30 20 o 10 _L 1948 1956 1964 1972 1980 YEAR Figure 18.— Annual Japanese landings of northern Pacific bluefin tuna for the years 1951-59 (metric tons X 1.000) and 1962-79 (thousands of fish). 118 HANAN: BLUEFIN TUNA FISHERY 2.80 2.10 UJ Q. O 1.40 0.70 1960 1964 1968 1972 YEAR 1976 1980 FIGURE 19.— Annual Japanese catch per unit effort (metric tons/boat-day) from long-line catches of northern bluefin tuna for the years 1962-79. Although those data (Anonymous 1981; Yama- naka and Staff 1963) represent only a portion oi the fishing effort in the western Pacific, they in- dicate a need for more extensive and explicit data from that area. With improved data, mathe- matical models for estimating sustainable yields can be used to describe the status of the bluefin resource throughout the North Pacific Ocean. Based on strong evidence of declining stock abundance, the bluefin tuna fisheries in the Pa- cific Ocean should receive an extensive analyti- cal review, and nations fishing bluefin, especially Japan, Mexico, and the United States, should consider needed actions. If management to con- serve this valuable resource is to be taken, it should be soon, so that the resource can return to an optimal level of abundance. ACKNOWLEDGMENTS I sincerely thank Alec MacCall (CFG) for his help and recommendations during this study. Norman Bartoo (NMFS) provided valuable ad- vice in the preparation of the manuscript. The editorial reviewers, especially Harold Clemens (CFG) and William Bayliff (IATTC), are thanked for their comments and suggestions. Funds for the study were provided by contract to the South- west Fisheries Center, NMFS. LITERATURE CITED Anonymous. 1981. Annual report of catch and effort statistics by area on Japanese longline fishery 1979. Fish. Agency Jpn., Res. Dev. Dep., 234 p. Bayliff, W. H. 1980. Synopsis of biological data on the northern bluefin tuna, Thunnus thynnus (Linnaeus, 1758), in the Pacific Ocean. Inter-Am. Trop. Tuna Comm., Spec. Rep. 2: 261-294. Bayliff, W. H., and T. P. Calkins. 1979. Information pertinent to stock assessment of north- ern bluefin tuna, Thunnus thynnus, in the Pacific Ocean. Inter-Am. Trop. Tuna Comm., Intern. Rep. 12, 78 p. Beauchamp, J. J., and J. S. Olson. 1973. Corrections for bias in regression estimates after logarithmic transformation. Ecology 54:1403-1407. Clemens, H. B., and W. L. Craig. 1965. An analysis of California's albacore fishery. Calif. Dep. Fish Game, Fish Bull. 128, 301 p. Clemens, H. B., and G. A. Flittner. 1969. Bluefin tuna migrate across the Pacific Ocean. Calif. Fish Game 55:132-135. GULLAND, J. A. 1963. The estimation of fishing mortality from tagging experiments. Int. Comm. Northwest Atl. Fish., Spec. Publ. 4:218-227. PELLA, J. J., AND C. T. PSAROPULOS. 1975. Measures of tuna abundance from purse-seine oper- ations in the eastern Pacific Ocean, adjusted for fleet- wide evolution of increased fishing power, 1960-1971. [In Engl, and Span.] Inter-Am. Trop. Tuna Comm., Bull. 16:281-400. RlCKER, W. E. 1975. Computation and interpretation of biological sta- tistics of fish populations. Fish. Res. Board Can., Bull. 191, 382 p. SCHULTZE, D. L., AND R. A. COLLINS. 1977. Age composition of California landings of bluefin tuna, Thunnus thynnus, 1963 through 1969. Calif. Dep. Fish Game, Mar. Res. Tech. Rep. 38, 44 p. Whitehead, S. S. 1931. Fishing methods for the bluefin tuna (Thunnus thynnus) and an analysis of the catches. Calif. Dep. Fish Game, Fish Bull. 33, 32 p. Yamanaka, H., and Staff. 1963. Synopsis of biological data on kuromaguro Thun- nus orientalis (Temminck and Schlegel) 1842 (Pacific Ocean). FAO Fish. Rep. 6(2):180-217. 119 INTERACTIONS BETWEEN FUR SEAL POPULATIONS AND FISHERIES IN THE BERING SEA Gordon L. Swartzman and Robert T. Haar 1 ABSTRACT In this paper we consider fur seal-fisheries interaction in the Bering Sea by asking whether the slower than originally predicted recovery of the fur seal stock from female fur seal harvest during 1956-68 might be a result of a reduction in carrying capacity because of the large fishery harvest of walleye pollock and Pacific herring— fish which are important fur seal prey. The changes we found occurring in the fur seal population did not support the hypothesis that fur seal carrying capacity was reduced by the fisheries. In fact the population parameters changed little, or changed in a direction opposite to that proposed by the hypothesis. Study of the fur seal diet data indicated that walleye pollock comprised a larger part of the fur seal diet in the 1970s, after the establishment of the fishery, than earlier, although average pollock size appeared to drop significantly. This trend may have been induced by an increased harvest of older fish. Since walleye pollock are cannibalistic, the removal of the older fish by the fishery could result in lower mortality among the younger pollock stocks, the outcome being an increase in the pollock resource available to both the fishery and the fur seal. In this paper we assess and clarify possible rela- tionships between fur seals and fisheries in the Bering Sea. The event most prominent in focus- ing concern on fur seal-fisheries interactions was the failure of the Pribilof Islands' fur seal herd to recover as predicted from large female harvests during 1956-68. While the present herd appears to have stabilized, it has stabilized at a population 30% below the maximum sustained productivity estimates made in 1955 (York and Hartley 1981). A number of possible explana- tions for this have been presented, including re- duced fur seal carrying capacity. In this paper we 1) briefly summarize and highlight the available fur seal and fish data, in- cluding studies of cases of other known marine mammal-fish interactions, 2) consider the evi- dence about fur seal population dynamics and seal-fish interactions, and 3) suggest analyses of existing data and further field sampling needed to clarify the effect of the Bering Sea fishery on fur seal populations. AVAILABLE DATA The relevant data may be divided into fur seal data, Bering Sea fish stock and fishery data, and anecdotal marine mammal-fish interaction data. 'Center for Quantitative Science, University of Washington, Seattle, WA 98195. The fur seal data consist of 1) annual fur seal col- lections at sea during 1958-74 in the eastern North Pacific Ocean and the eastern Bering Sea conducted jointly by the United States and Can- ada under terms of the Fur Seal Interim Conven- tion (Kajimura et al. 1979, 2 1980 3 ); 2) harvests from 1950 to 1978 on the Pribilof Islands of sub- adult males (Lander 1981) and counts of harem and nonharem bulls from 1905 to 1978 on other island rookeries; 3) estimates of pup production on the Pribilof Islands from 1912 to 1924 and from 1951 to 1979 (Johnson 1975; Lander 1981), and counts of dead pups from 1950 to 1979 (Lan- der 1981); and 4) studies of fur seal rookery be- havior (Bartholomew and Hoel 1953; Gentry 4 ), food habits (Spalding 1964; May 1937; Wilke and 194-1**- Manuscript accepted July 1982. FISHERY BULLETIN: VOL. 81. NO. 1. 1983. 2 Kajimura, H.. R. H. Lander, M. A. Perez. A. E. York, and M. A. Bigg. 1979. Preliminary analysis of pelagic fur seal data collected by the United States and Canada during 1958- 74. Report submitted to the 22d Annual Meetingof the Stand- ing Scientific Committee, North Pacific Fur Seal Commission, 247 p. Northwest and Alaska Fisheries Center National Ma- rine Mammal Laboratory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way NE., Seattle, WA 98115. 3 Kajimura, H., R H. Lander, M. A. Perez, A. E. York, and M. A. Bigg. 1980. Further analysis of pelagic fur seal data collected by the United States and Canada during 1958-74. Part 1. Submitted to the 23d Annual Meetingof the Standing Scientific Committee, North Pacific Fur Seal Commission, 94 p. Northwest and Alaska Fisheries Center National Marine Mammal Laboratory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way NE., Seattle, WA 98115. 4 R. Gentry, Northwest and Alaska Fisheries Center National Marine Mammal Laboratory, National Marine Fisheries Ser- vice. NOAA, 7600 Sand Point Way NE., Seattle. WA 98115, pers. commun. May 1980. 121 FISHERY BULLETIN: VOL. 81, NO. 1 Kenyon 1957; Fiscus 1979; Kajimura et al. foot- notes 2, 3), and fertility (Abegglen and Roppel 1959). Bering Sea groundfish and pelagic fisheries data, which give estimates of relative abundance, life history parameters, and migratory patterns of important fish stocks, are contained in a num- ber of Northwest and Alaska Fisheries Center (NWAFC) reports (Pereyra et al. 1976 5 ; Favorite et al. 1979 6 ; Pruter 1973; Bakkala et al. 1979 7 ). These data cover the period of development of the large foreign groundfish fishery in the eastern Bering Sea (1954-78) and include catch, catch per unit effort (CPUE), mortality, seasonal mi- gration patterns, and diets for a number of com- mercially important fish, including walleye pol- lock and Pacific herring, important food sources for the fur seal in the eastern Bering Sea. Fur Seal Data Synopsis Seal Data Collected at Sea Fur seal migration patterns were deduced from fur seals sampled at sea from 1958 to 1974. Adult males remain year-round in the Bering Sea and Gulf of Alaska, while females migrate south in winter, with smaller (younger) females tending to migrate the farthest south. Many sub- adult males also migrate south, but not nearly so far as the females. Females begin returning to the rookeries of the Pribilof Islands in June, and the rookeries are almost completely established by the end of July (Kajimura et al. footnotes 2, 3). Pelagic data were also used to construct a fur seal life table (Lander 1981) which, along with a pup production estimate, gave an overall fur seal biomass estimate for the Pribilof Islands stock of 29,000 t or 1.25 million animals. Seasonal pat- terns of growth were also computed from the 5 Pereyra, W. T., J. E. Reeves, and R. G. Bakkala. 1976. Demersal fish and shellfish resources of the eastern Bering Sea in the baseline year 1975. Proc. Rep., 619 p. Northwest and Alaska Fisheries Center Seattle Laboratory, National Marine Fisheries Service, NOAA, 2725 Montlake Blvd. E., Seattle, WA 98112. "Favorite, F., W. J. Ingraham, Jr., K. D. Waldron, E. A. Best, V. G. Wespestad, L. H. Barton, G. B. Smith, R. G. Bakkala, R. R. Straty, and T. Laevastu. 1979. Fisheries oceanog- raphy — eastern Bering Sea Shelf. Proc. Rep. 79-20, 481 p. Northwest and Alaska Fisheries Center Seattle Laboratory, National Marine Fisheries Service, NOAA, 2725 Montlake Blvd. E., Seattle, WA 98112. 7 Bakkala, R., L. Low, and V. Wespestad. 1979. Condition of groundfish resources in the Bering Sea and Aleutian area. NMFS Northwest and Alaska Fisheries Center report sub- mitted to the International North Pacific Fisheries Commis- sion, 106 p. pelagic survey data (Lander 1981). Stomach con- tent data were pooled over years by region and by month, and were presented as the frequency of occurrence (proportion of stomachs containing a particular food item), the volume and the percent of total food volume comprised by each prey type, and the number of specimens of each prey type and their percent of the total diet. Diet compo- sition of fur seal stomachs by percent volume (which we consider to be the most reliable mea- sure of prey abundance in predator stomachs) in the eastern Bering Sea is given in Table 1 (modi- fied from Kajimura et al. footnote 3) pooled by month over all years of data collection. Table 1.— Major species in fur seal diets in the east- ern Bering Sea (percent volume), June-September. (Kajimura et al. footnote 3). Species June July August September Herring — 0.2 13.2 0.2 Capelin 699 16.4 170 15 2 Pollock 4.1 50.9 26.1 383 Deepsea smelt — 4.0 3.5 8.6 Atka mackerel 19 4 1.5 1.7 18 Squid 4.9 22.0 29.4 17.5 Fur seals are pelagic feeders and are highly opportunistic (Kajimura 1981 8 ), feeding on a wide variety of species. Of their major prey only pollock and herring are target species for a fish- ery. Data on fur seal diets outside the eastern Bering Sea corroborate the pattern of fur seals feeding primarily on schooling fish. South of British Columbia, hake replaces pollock in seal stomachs and herring and sand lance are increas- ingly important, while capelin decreases in im- portance. Anchovy is the most important fur seal food off California. Since fur seals and fisheries both tend to exploit schooling species, a possible competitive relationship may exist between fur seals and fisheries. Most fur seal feeding in the Bering Sea is done by lactating females during the summer pupping period, so the importance of food during this period cannot be overempha- sized. Since this is the period of rapid pup growth and is also the period of maximum growth for nonpregnant females and subadult males (Fig. 1), food limitation during this period could have drastic consequences to pup survival, especially after they leave the rookeries. 8 Kajimura, H. 1981. The opportunistic feeding of north- ern fur seals off California. Unpubl. manuscr., 46 p. North- west and Alaska Fisheries Center National Marine Mammal Laboratory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way NE., Seattle, WA 98115. 122 SWARTZMAN and HAAR: INTERACTIONS BETWEEN FUR SEALS AND FISHERIES 130 120 110 100 - 90 80 ?0 7 years 6 years 5 years _1_ Jan Feb Mar Apr May June July Aug See Oct Nov Dec Figure 1.— Seasonal pattern of growth in mean length (cm) of nonpregnant female fur seals of age 1-7. Curves are drawn by inspection with the restriction of no downward curvature. An x designates <10 seals. From Lander (1981). Sampling on the Fur Seal Rookeries The herds on the Pribilof Islands (St. Paul and St. George Islands and Sea Lion Rock) are esti- mated to comprise 80% of the total world fur seal population. Every year from 1912 to 1924 and since 1950 some census of pup births has been made. Dead pup counts have also been made. Harvests of subadult males on the island hauling grounds have yielded information on weights, lengths, and age composition of these animals as well as limited food data from stomach samples. An estimate has also been made annually of num- bers of harem bulls. From 1956 to 1968 almost 300,000 females were harvested from St. Paul and St. George Islands, presumably to increase the sustained productivity of the herds. The herd subsequently failed to achieve a higher sustained productivity as was postulated from higher pregnancy and survival rates predicted from population projec- tions (Abegglen et al. 1956 9 ). From 1912 to 1924, pup populations were esti- mated from direct counts. Fur seal populations increased steadily over this period at an 8% an- nual rate, as they recovered from heavy losses due to pelagic sealing in the late 19th and early 20th centuries. Direct counts were discontinued from 1924 to 1948, but an 8% annual population increase was assumed. However, estimates of pups in 1948 showed that the 8% increase had not continued. In 1947, tagging studies were set up to estimate pups and were continued until 1961. In 1960 an estimation procedure involving pup shearing and direct counts was initiated to re- place the tagging method. Estimates of the num- ber of pups born were computed by adding live pup estimates to dead pup counts. The 1951-61 tagging studies are presently thought to have greatly overestimated actual pup abundance because of procedural difficul- ties and lost tags (Chapman 1973). The pup shear- ing procedure, although shown to be unbiased by comparing pup estimates with direct counts on small rookeries (Chapman and Johnson 1968), may be biased for large rookeries in such a way as to underestimate actual pup numbers (Fowler 10 ). Age-specific survival and weight at age were estimated from the weighing and aging of the preadult males harvested annually on the rook- eries. Male harvest was discontinued on St. George Island in 1972 to study the effect of the male population density on seal population dy- namics. Recent pup survival on St. George Island appeared lower than on St. Paul Island (Lander 1981), and this has been linked to the increased abundance of idle males on the rookeries (Fowler footnote 10). Bering Sea Fish Data Data on commercially important Bering Sea fish stocks by species have been compiled by the NWAFC. Catch data from Japanese, Russian, Korean, Polish, United States, and Canadian fishing operations have been included. The ma- jor species (in order of magnitude of catch) are walleye pollock, Theragra chalcogramma; yel- lowfin sole, Limanda aspera; Pacific herring, Clupea harengus pallasi; Pacific salmon, Onco- rhynchus spp; Pacific cod, Gadusmacrocephalus; sablefish, Anoplopoma fimbria; Pacific halibut, Hippoglossus stenolepis; other flatfish (rock sole, flathead sole, Alaska plaice, Greenland turbot, 9 Abegglen, C. F., A. Y. Roppel, and F. Wilke. 1956. Alas- ka fur seal investigations, Pribilof Islands, Alaska. Manuscr. rep., 143 p. Northwest and Alaska Fisheries Center National Marine Mammal Laboratory, National Marine Fisheries Ser- vice, NOAA, 7600 Sand Point Way NE., Seattle. WA 98115. 10 C. W. Fowler, Head, Fur seal investigations group, North- west and Alaska Fisheries Center National Marine Mammal Laboratory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way NE., Seattle, WA 98115, pers. commun. June 1980. 123 FISHERY BULLETIN: VOL. 81. NO. 1 and arrowtooth flounder); and Pacific ocean perch, Sebastes alutus. Pacific herring and wall- eye pollock (hereafter referred to as herring and pollock) are the most important of these species in the diet of fur seals in the Bering Sea, and have been heavily fished (as have yellowfin sole, hali- but, and Pacific ocean perch). The intensity of fishing on herring and pollock suggests the pos- sibility of fur seal stock depletion due to de- creased food abundance, although stock deple- tions can also have other causes. Figure 2, adapted from Pereyraetal. (footnote 5) and Favorite et al. (footnote 6), gives the total catch for pollock and herring as well as an index of relative abundance (CPUE) based on research trawl surveys conducted by the International Pa- cific Halibut Commission, the National Marine Fisheries Service (NMFS), and the Japanese Fishery Agency. Pollock stocks have been heavily fished since 1964, with peak yields coming in the early 1970's. A steady increase in CPUE between 1964 and 1968 may have been due in part to improvements in fishing gear and tactics, but must also have been due to higher levels of recruitment of young fish (Pruter 1973), possibly because of reduced cannibalism. Pruter ( 1973) pointed out that, since only a few age groups of pollock are utilized in 1-8 to c o sz 10 -r- CO 3 o 160- herring — catch 120- '/ i '/ i '/ X '/ i '/ i '/ i 1 / * — CPUE 80- i / * / I / 1 / 1 I 1 I 40- \ \ 1 1 i i «T 20 c o o CD E 3 « 8 3 o sz "O CD &_ "O c 3 sz A s\ 15- 10- / / / / 1 1 1 § 1 J \f\ r * \ / * \ / % / \ y \ 5- 1 1 1 pollock* — catch -—CPUE 1 1 1 -4 -2 -30 20 3 O JZ CD LU Q. Z> en D. c O 2 o "55 E 10 3 o si CD LU q. 3 ; CD < 5.5 5.0 j i i l i i i i i i 54 56 58 60 62 64 66 Year class Figure 3.— Estimated average age at first reproduction of fe- male northern fur seals based on females pregnant at least once for the 1954-64 year classes. From Kajimura et al. (text foot- note 3). 126 SWARTZMAN and HAAR: INTERACTIONS BETWEEN FUR SEALS AND FISHERIES entire population would leave the nonrookery population with a higher proportion of imma- tures which would then affect the samples taken at sea. Another difficulty with these data is that only 2 yr of pre- 1956 age class data were avail- able from the pelagic cruises, and the other pre- 1956 data reported by Kajimura et al. (footnote 3) may not have used the same index of maturity as Kajimura et al. (footnote 3). Other possible sources of bias in the age at maturity estimate were the tendency of the pelagic fur seal samples to contain a higher number of older individuals than expected, and the underlying assumption that survival rates of pregnant and nonpregnant females are the same (Kajimura et al. footnote 3). Growth With Age Preliminary analysis by the National Marine Mammal Laboratory (NMML) (Fowler footnote 10) of the data from 3-yr-old males harvested on the Pribilof Islands showed a statistically signifi- cant increase in weight over time from 1964 to 1970 in contrast to growth rate reductions to be expected under a reduced fur seal carrying ca- pacity. Kajimura et al. (footnote 2) plotted the average length of pregnant females against age for the time periods 1958-62, 1963-68, and 1969-74. Their results (Fig. 4) indicate that growth rates were greater from 1963 to 1974 than from 1958 to 130 125 120 1963-68 \ ^ ,-10 seals. From Kajimura et al. (text footnote 2) 15 J. Berdine, Judson Hall, Room 621, 53 Washington Square South, New York, NY 10012, pers. commun. August 1980. 127 FISHERY BULLETIN: VOL. 81, NO. 1 on St. Paul Island. Gentry (footnote 4) made simi- lar observations on nursing fur seals in the late 1970's and found no significant change in time at sea from those of Bartholomew's study. Pup Survival to Age 2 Lander (1981) calculated early survival rates to age 2 for male fur seals from the 1950-70 year classes. York and Hartley (1981) analyzed these estimates, using Mann-Whitney and Student's t tests, and found pre- 1956 rates to be significantly lower than post-1956 rates (0.32 vs. 0.40 aver- age). This does not appear to support the hypothe- sis of reduced carrying capacity. Time Trends in Fur Seal Diets Fur seal stomach contents taken in 1960, 1962- 64, 1968, and 1973-74 cruises were used to inves- tigate trends in fur seal diets to see whether these might have changed after development of the pollock fishery. These data were summarized by month. Figure 5 indicates that the age composition in catch in the pollock fishery shifted from a mode of 4 yr in 1964 to 3 yr in 1974 with the 2-yr-old catch also being strongly represented. H. Kaji- mura, who was present on the cruises, suggested that the size of pollock in fur seal stomach sam- ples decreased from 1964 to 1974. Examination of average volume per pollock specimen in fur seal stomachs (Unpubl. data 16 ; Table 2) corrobo- rates this observation, with average specimen size decreasing significantly between 1968 and 1973-74. We also note that the percentage volume of the total stomach content comprised of pollock was consistently high in 1973-74 (>48%), while earlier, especially before 1968, pollock comprised a variable and usually low percentage of the diet (<20% in 8 of the 11 mo sampled). These data indicate that there may have been an interaction between fur seal diets and the pol- lock fishery. As fishing pressure on pollock in- creased, fishing out of older age classes reduced the average size of the fish and increased the average growth rate of the pollock. Furthermore, young pollock survival may have been increased through reduced cannibalism. These increased en a on =3 o H z to -J a > a z i— i O 8 4 O 6 4 8 4 O 8 4 8 4 8 O 8 W 4 O 8 O 8 4 O 8 O 8 4 O 1964 CW 8.53 CN 15,214 1965 CW 9.25 CN 15,462 1966 CW 6.83 CN 12,009 1967 CW 8.23 CN 14,764 968 CW I 1.84 CN 21,574 1969 CW 10.41 CN 18,965 19 70 CW 8.99 CN 19,453 1971 CW 5.79 CN 12,392 1972 CW 9.33 CN 20,493 973 CW 6.25 CN 16,347 1974 CW 5.38 CN 15,51 23456789 AGE Figure 5.— Age composition in catch per unit effort (CPUE) of walleye pollock from the Japanese trawl fishery in the eastern Bering Sea. Japanese trawl fishery includes the mothership fishery and the North Pacific trawl fishery, but not land-based- dragnet fishery. From Salveson and Alton (text footnote 12). C W = CPUE in weight in metric tons; CN = CPUE in number. 16 Data obtained from Dr. M. Tillman, Director, Northwest and Alaska Fisheries Center National Marine Mammal Lab- oratory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way NE., Seattle, WA 98115. stocks of smaller fish were reflected by the in- crease in abundance of pollock in fur seal diets after 1968 and by a marked decrease in the aver- age size of fish taken by the fur seals. This in- 128 SWARTZMAN and HAAR: INTERACTIONS BETWEEN FUR SEALS AND FISHERIES Table 2. — Fur seal dietof walleye pollock from pelagic samples in theeastern Bering Sea. (Unpubl. data (text footnote 16).) Number of Volume of pollock diet Number of Percent of Pollock volume/ stomachs i n pollock in diet total specimen (cm 3 ) Date with food cm 3 Percent numbers June 1960 4 385 12.3 19 5.2 20.26 July 1960 152 39.807 61 403 9.8 98.7 Aug 1960 61 37.124 75 148 10 251 June 1962 53 295 24 2 0.16 147.5 July 1962 137 4.343 126 45 1.1 96.5 Aug. 1962 277 17,266 183 323 3.1 53.45 Sept. 1962 111 10,342 28 235 5.4 44.0 July 1963 256 11.188 14.16 62 0.56 180.45 Aug 1963 536 9,758 5 163 0.59 59.9 Sept. 1963 17 700 11.06 1 0.11 700 July 1964 97 2.354 95 7 0.27 336 Aug 1964 213 29,296 15.4 792 98 37 July 1968 78 31,901 76.9 384 14.3 83 Aug. 1968 53 11.206 37.4 30 1.21 3735 July 1973 148 72,427 907 1.418 33.0 51.07 Aug 1973 191 36,564 60.7 1,305 15.1 43.34 Sept. 1973 178 32,511 48.5 2,172 23.7 14.9 July 1974 52 13,658 87 4 244 58.6 36.0 Aug 1974 110 15,198 63.2 390 20.2 38.9 crease in total stock biomass, mostly in the youn- ger age classes, can account both for the increased fur seal diets on (mostly smaller) pollock and the continued high yield of the fishery after over 10 yr of heavy fishing pressure. Table 2 indicates that both fur seals and the fishery may have exploited the same pollock re- source, since both show a drop in size of "catch" over time. We suspect that the trend toward greatly increased abundance of pollock juveniles in the Bering Sea has also resulted in larger schools (patches) of juvenile pollock, which has made them an easier target for the fur seals and also the fishery, than previously. One possible, dangerous consequence of future increased fish- ing pressure on pollock, however, is that most of the catch will be of premature individuals. With continued heavy fishing pressure, this might re- sult in inadequate recruitment to maintain the stock. A possible alternative explanation for why pol- lock were so consistently taken by fur seals in 1973-74 is that these were relatively cold years with pollock aggregating more on the outer shelf than in warmer years (Pereyra et al. footnote 5). Another possible explanation is that the Pribilof area, where the bulk of the 1973 and 1974 stom- ach samples were taken (unlike the earlier sam- ples which did not focus as heavily on this area), is a nursery area for young-of-the-year pollock, which may account for the reduced average size and increased abundance of pollock in fur seal stomachs during 1973 and 1974. Despite these possible alternatives, the most plausible hypothe- sis is that pollock has increased in importance in fur seal diets since the initiation of the pollock fishery. Energetics Approach to Fur Seal Food Consumption The total amount of food consumed by fur seals (and other marine mammals as well) in the east- ern Bering Sea has been estimated by a num- ber of individuals (Laevastu and Larkins 1981; McAlister and Perez 1976 17 ; Anonymous 1979 18 ). McAlister and Perez (footnote 17) estimated that fur seals eat 378,000 t of fish and squid every year. They used an estimated feeding rate of 7.5% body weight daily while Miller ( 1978) 19 suggested that 14% body weight daily may be more appro- priate to support seals at 7°C, the average sum- mer temperature in the Bering Sea. Miller based his arguments on metabolic studies in which he recorded oxygen consumption at different tem- peratures in the laboratory for a number of juve- nile seals and also conducted feeding studies using food most commonly found in the diet of "McAlister, W. B., and M. A. Perez. 1976. Preliminary estimates of pinniped-finfish relationships in the Bering Sea. Background paper for the 19th meeting of the North Pacific Fur Seal Commission, 29 p. Northwest and Alaska Fisheries Center National Marine Mammal Laboratory, National Ma- rine Fisheries Service, NOAA, 7600 Sand Point Way NE., Seattle, WA 98115. 18 Anonymous. 1979. Draft environmental impact state- ment of the Interim Convention on Conservation of North Pa- cific Fur Seals. U.S. Dep. Commer., NOAA, Natl. Mar. Fish. Serv., Seattle, Wash., 39 p. 19 Miller, L. K. 1978. Energeticsof the northern fur seal in relation to climate and food resources of the Bering Sea. U.S. Marine Mammal Commission Report MMC-75/08, 27 p. 129 FISHERY BULLETIN: VOL. 81. NO. 1 fur seals in the Bering Sea. Using Miller's esti- mate for consumption would give an estimate of 705,000 t eaten annually by fur seals. Laevastu and Larkins ( 1981 ) gave an estimate of 513,000 t taken by fur seals annually in the eastern Bering Sea, with an additional 368,000 t taken in the Aleutian region. The latter estimates were based on runs of the PROBUB (prognostic bulk bio- mass) model. Estimates of fur seal populations of the Bering Sea and the Aleutian Islands and their mean consumption rates, given in Table 3 (Anonymous footnote 18), were used to compute a total fur seal consumption of 219,000 t. These estimates can be compared with annual fish catches in the eastern Bering Sea and Aleu- tian Islands (North Pacific Fishery Management Council 20 ). Between 1968 and 1976, annual fish catches varied between 750,000 and 2,100,000 t in the eastern Bering Sea and between 40,000 and 80,000 t near the Aleutian Islands. These fig- ures indicate that fish harvests by marine mammals and by man in the Bering Sea are com- parable and that the marine mammals' harvest exceeds man's in the Aleutian Islands' area. It is important to note, however, that fur seals prey on a larger number of species than man, and thus a part of their harvest is not in direct competition with man's. As a consequence of the fur seals' greater ability to switch prey when abundances of preferred prey species are low, total fur seal consumption is probably fairly steady from year to year, while man's is highly variable. It has been estimated (Anonymous footnote 18: table 12) that 9.8% of fish standing stock in the eastern Bering Sea and the Aleutian Islands is consumed annually by marine mammals, 5% by man, and 1.8% by birds (1.9% by fur seals). Lae- vastu and Larkins (1981) estimated a total com- mercial fish standing stock of 24,880,000 1 in the Bering Sea and Gulf of Alaska, which implies that 3.5% of all commercial fish stocks are taken by fur seals annually and 10.7% by all marine mammals. The fur seal figures are deceptive, since fur seal impact on fish stocks is relatively localized. Thus, fur seals near the Pribilof Islands are probably consuming considerably more fish than man is, though man may be harvesting some different species than fur seals. This ener- getics computation is inconclusive with respect to fur seal-fishery interaction, except to show that competition between the two is possible. DISCUSSION Suggested Analyses of Existing Data Population Indices -"Data available from North Pacific Fishery Management Council, 333 W. Fourth Ave., Suite 32, P.O. Box 3136DT, An- chorage. AK 96813. Following Eberhardt and Siniffs (1977) sug- gestion that a population's response to impact may be reflected by various indices, we suggest Table 3. — Fur seal population estimates at sea (June-November) in the eastern Bering Sea and Aleutian area (Anonymous text footnote 18). June-Nov. Estimated Estimated eastern percent of population Mean 2 Mean 3 daily Population 1 Bering Sea time at sea at sea weight consumption Age class total and Aleutian (June-Nov.) (June-Nov ) (kg) rate (%) Pups 349.000 '321,000 10 32,100 1000 14.00 M+F, age 1 174,000 67,000 5 90 78,300 9.54 13.76 M+F, age 2 122,000 61.000 5 75 45,750 1669 12 32 F. age 3 55,000 23.000 5 80 22,400 1880 12.53 F, >age 4 582,000 46,000 6 79 368.140 35.64 11.76 M. age 3-7 101,000 71,000 7 10 7.100 3260 7.60 M, >age 7 11,000 9,000 7 10 900 105.25 701 Total 1,394,000 1,043,000 "(754,100) 554.690 2992 11.71 'Average 1969 to 1974. 2 Based on National Marine Mammal Division, NMFS pelagic research data, 1958-74, N = 13,772, except average weight for pups (10 kg) based on observations in the Pribilof Islands during September; total mean weight based on an effective fishery population 754,000, on time spent on land and at sea for each class during June and September. 'Weighed by mean animal weight of estimated body weight for animals weighing <10 kg or <45 kg in waters colder than 15°C; 7% for >10 kg on land or >45 kg at sea. "Based on the ratio of males to females (0.085) in the eastern Bering Sea during June-November from National Marine Mammal Division, NMFS pelagic research data, 1958-74 (N = 4,451). 5 8% mortality, pups estimated to feed at sea only 18 d (10% of time) during September-November 6 These percentages represent proportions of the total population of the respective age class not on the rookeries during the breeding season 7 Based on percent of time out of 130 d not on rookery. 8 Effective fishery population (June-November) 130 SWARTZMAN and HAAR: INTERACTIONS BETWEEN FUR SEALS AND FISHERIES that the available data from which these indices are computed be also studied for trends. Indices that are most easily obtained for the fur seal are pup birth estimates, dead pup counts, male sur- vival to age 3 (from male harvest data), and length at age for preadult males (from harvested males). Fur Seal Diet Trend We have suggested a relationship between fur seals and the fishery via greatly increased abun- dance of juvenile pollock (Table 2). The data used, however, were already combined in such a way that we were unable to separate the data by region where the data were collected and the de- gree of digestion of the prey. We suggest that the original data be used to conduct a complete sta- tistical analysis with corrections made for the area in which the sample was taken and, if pos- sible, the time of day the samples were taken (assuming that the correlations found between the proportion of the stomachs empty and time of day the samples were taken also applies to the percentage of food digested). Variance estimates can also be computed and used to make statisti- cal tests for time trends both in the average size of pollock in fur seal stomachs and in the percent- age of the total diet comprised of pollock. Role of Patchiness in Seal Feeding Although we suspect from survey data on pol- lock (Smith 1979) that pollock are quite patchily distributed in the eastern Bering Sea, the survey data need to be reexamined for an indication of the size of patches or degree of aggregation. An . attempt should be made to represent this patchi- ness stochastically (in terms of probability). One important question to be considered with these data is whether or not there has been a trend in pollock school size from 1963 to 1974 in the east- ern Bering Sea. Another approach to consider patchiness is to use the abundance of pollock in fur seal stomachs collected at different locations as an index to the spatial separation and size of pollock schools. Suggested Future Data Collection We suggest that a fish trawl survey targeting on pollock be conducted between the Pribilof Islands and Unimak Pass from June to Septem- ber with study designed to focus on areas of high pollock density to determine the size distribution of pollock, the size of the schools, and, if possible, to observe fur seal feeding intensity around the schools. The pollock and fur seals might be tracked by using multibeam sonar techniques. Additional sto