Fishery Bulletin National Oceanic and Atmospheric Administration • National Marine Fisheries Service ^^ATES 0^ ^ ORARV '"'1 ^■ Vol. 74, No. 1 I January 1976 I BROTHERS, EDWARD B., CHRISTOPHER Pf MAX-HEWS /and REUBEN LASKERI Daily growth increments in otoliths from larval and adult fishes -.-*?-^?4=l. STRUHSAKER, PAUL, and JAMES H. UCHIYAMA. Age and growth of the nehu! Stole- phorus purpureas (Pisces: Engraulidae), from the Hawaiian Islands as indicated by daily growth increments of sagittae MERRINER, JOHN V. Aspects of the reproductive biology of the weakfish, Cynoscion regalis (Sciaenidae), in North Carolina MacGREGOR, JOHN S. DDT and its metabolites in the sediments off southern California . SHARP, GARY D., and ROBERT C. FRANCIS. An energetics model for the exploited yellowfin tuna, Thunnus albacares, population in the eastern Pacific Ocean ROGERS, CAROLYN A. Effects of temperature and saUnity on the survival of winter flounder embryos UCHIDA, RICHARD N. Reevaluation of fishing effort and apparent abundance in the Hawaiian fishery for skipjack tuna, Katsuwonus pelamis, 1948-70 PEARCY, WILLIAM G. Seasonal and inshore-offshore variations in the standing stocks of micronekton and macrozooplankton off Oregon HUNTER, JOHN R. Culture and grovii;h of northern anchovy, Engraulis mordux, larvae CRAWFORD, L., and M. J. KRETSCH. Effects of cooking in air or in nitrogen on the develop- ment of fishy flavor in the breast meat of turkeys fed tuna oil with and without a-tocopherol supplement or injection LEWIS, THOMAS C, and RALPH W. YERGER. Biology of five species of searobins (Pisces^ Triglidae) from the northeastern Gulf of Mexico LORD, GARY, WILLIAM C. ACKER, ALLAN C. HARTT, and BRIAN J. ROTHSCHILD. acoustic method for the high-seas assessment of migrating salmon PRISTAS, PAUL J., ELDON J. LEVI, and ROBERT L. DRYFOOS. Analysis of returns of tagged Gulf menhaden TILLMAN, MICHAEL F , and DONALD STADELMAN. Development and example appUca- tion of a simulation model of the northern anchovy fishery MASON, J. C, and S. MACHIDORI. Populations of sympatric sculpins, Cottus aleuticus and Cottus asper, in four adjacent salmon-producing coastal streams on Vancouver Island, B.C BUTLER, JOHN L., and ELBERT H. AHLSTROM. Review of the deep-sea fish genus Scopelengys (Neoscopehdae) with a description of a new species, Scopelengys clarkei, from the central Pacific CHITTENDEN, MARK E., JR. Weight loss, mortaUty, feeding and duration of residence of adult American shad, Alosa sapidissima, in fresh water An 9 18 27 36 52 59 70 81 89 93 104 112 118 131 142 151 (Continued on back cover) Seattle, Washington / U.S. DEPARTMENT OF COMMERCE Rogers C. B. Morton, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION Robert M. White, Administrator NATIONAL MARINE FISHERIES SERVICE Robert W. Schoning, Director Fishery Bulletin The Fishery Bulletin carries original research reports and technical notes on investigations in fishery science, engineering, a economics. The Bulletin of the United States Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates were issued as documents through volume 46; the li document was No. 1103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate appeared 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 instead of being issued individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a periodic issued quarterly. In this form, it is available by subscription from the Superintendent of Documents, U.S. Government Printing Offi Washington, D.C. 20402. It is also available free in limited numbers to libraries, research institutions, State and Federal agencies, a in exchange for other scientific publications. EDITOR Dr. Bruce B. Collette Scientific Editor, Fishery Bulletin National Marine Fisheries Service Systematics Laboratory National Museum of Natural History Washington, DC 20560 Editorial Committee Dr. Elbert H. Ahlstrom National Marine Fisheries Service Dr. William H. Bayliff Inter-American Tropical Tuna Commission Dr. Roger F. Cressey, Jr. U.S. National Museum Mr. John E. Fitch California Department of Fish and Game Dr. William W. Fox, Jr. National Marine Fisheries Service Dr. Marvin D. Grosslein National Marine Fisheries Service Dr. Edward D. Houde University of Miami Dr. Merton C. Ingham National Marine Fisheries Service Dr. Reuben Lasker National Marine Fisheries Service Dr. Jay C. Quast National Marine Fisheries Service Dr. Paul J. Struhsaker National Marine Fisheries Service Dr. Austin Williams National Marine Fisheries Service Kiyoshi G. Fukano, Managing Editor The Fishery Bulletin is published quarterly by Scientific Publications Staff, National Marine Fisheries Service, NCAA, Room 450, 1107 NE 45th Street, Seattle, WA 98105. Controlled circulation postage paid at Seattle, Wash. The Secretary of Commerce has determined that the publication of this periodical is necessary in the transaction of the public business required by law of this Department. Use of funds for printing of this periodical has been approved by the Director of the Office of Management and Budget through 31 May 1977. Fishery Bulletin CONTENTS Vol. 74, No. 1 January 1976 BROTHERS, EDWARD B., CHRISTOPHER R MATHEWS, and REUBEN LASKER. Daily growth increments in otoliths from larval and adult fishes 1 STRUHSAKER, PAUL, and JAMES H. UCHIYAMA. Age and growth of the nehu, Stole- phorus purpureas (Pisces: Engraulidae), from the Hawaiian Islands as indicated by daily growth increments of sagittae 9 MERRINER, JOHN V. Aspects of the reproductive biology of the weakfish, Cynoscion regalis (Sciaenidae), in North Carolina 18 MacGREGOR, JOHN S. DDT and its metabolites in the sediments off southern California . 27 SHARP, GARY D., and ROBERT C. FRANCIS. An energetics model for the exploited yellowfin tuna, Thunnus albacares, population in the eastern Pacific Ocean 36 ROGERS, CAROLYN A. Effects of temperature and salinity on the survival of winter flounder embryos 52 UCHIDA, RICHARD N. Reevaluation of fishing effort and apparent abundance in the Hawaiian fishery for skipjack tuna, Katsuwonus pelamis, 1948-70 59 PEARCY, WILLIAM G. Seasonal and inshore-offshore variations in the standing stocks of micronekton and macrozooplankton off Oregon 70 HUNTER, JOHN R. Culture and growth of northern anchovy, Engraulis mordax, larvae . . 81 CRAWFORD, L., and M. J. KRETSCH. Effects of cooking in air or in nitrogen on the develop- ment of fishy flavor in the breast meat of turkeys fed tuna oil with and without a-tocopherol supplement or injection 89 LEWIS, THOMAS C, and RALPH W YERGER. Biology of five species of searobins (Pisces, Triglidae) from the northeastern Gulf of Mexico 93 LORD, GARY, WILLIAM C. ACKER, ALLAN C. HARTT, and BRIAN J. ROTHSCHILD. An acoustic method for the high-seas assessment of migrating salmon 104 PRIST AS, PAUL J., ELDON J. LEVI, and ROBERT L. DRYFOOS. Analysis of returns of tagged Gulf menhaden 112 TILLMAN, MICHAEL F, and DONALD STADELMAN. Development and example applica- tion of a simulation model of the northern anchovy fishery 118 MASON, J. C, and S. MACHIDORI. Populations of sympatric sculpins, Cottus aleuticus and Cottus asper, in four adjacent salmon-producing coastal streams on Vancouver Island, B.C. 131 BUTLER, JOHN L., and ELBERT H. AHLSTROM. Review of the deep-sea fish genus Scopelengys (Neoscopelidae) with a description of a new species, Scopelengys clarkei, from the central Pacific 142 CHITTENDEN, MARK E., JR. Weight loss, mortaUty, feeding, and duration of residence of adult American shad, Alosa sapidissima , in fresh water 151 (Continued on next page) Seattle, Washington For sale by the Superintendent of Documents, U.S. Government Printing Office, Washing- ton, D.C. 20402 — Subscription price: $11.80 per year ($2.95 additional for foreign mail- ing). Cost per single issue - $2.95. Contents — continued BRUSHER, HAROLD A., and LARRY H. OGREN. Distribution, abundance, and size of penaeid shrimps in the St. Andrew Bay system, Florida 158 MASON, J. C. Some features of coho salmon, Oncorhynchus kisutch, fry emerging from simu- lated redds and concurrent changes in photobehavior 167 HURLEY, ANN C. Feeding behavior, food consumption, growth, and respiration of the squid Loligo opalescens raised in the laboratory 176 GARRISON, DAVID L. Contribution of the net plankton and nannoplankton to the standing stocks and primary productivity in Monterey Bay, California during the upwelling season . 183 TRENT, LEE, EDWARD J. PULLEN, and RAPHAEL PROCTOR. Abundance of macrocrusta- ceans in a natural marsh and a marsh altered by dredging, bulkheading, and filling 195 Notes FISHER, WILLL\M S., and DANIEL W. WICKHAM. Mortalities and epibiotic fouling of eggs from wild populations of the Dungeness crab. Cancer magister 201 MATSUMOTO, WALTER M. Second record of black skipjack, Euthynniis lineatus, from the Hawaiian Islands 207 WEIS, JUDITH S., and PEDDRICK WEIS. Optical malformations induced by insecticides in embryos of the Atlantic silverside, Menidia menidia 208 CHENG, LANNA, and RALPH A. LEWIN. Goose barnacles (Cirripedia: Thoracica) on flotsam beached at La Jolla, California 212 LAURENCE, GEOFFREY C. Caloric values of some North Atlantic calanoid copepods 218 HAURY, LOREN R. Method for restraining living planktonic crustaceans 220 STILLWELL, CHARLES E., and JOHN G. CASEY. Observation on the bigeye thresher shark, Alopias superciliosus, in the western North Atlantic 221 LEWIS, ELIZABETH G. Epizoites associated with Bathynectes superbus (Decapoda: Portunidae) 225 Vol. 73, No. 4 was published on 11 December 1975. The National Marine Fisheries Service (NMFS) does not approve, recommend or endorse any proprietary product or proprietary material mentioned in this publica- tion. No reference shall be made to MNFS, or to this publication furnished by NMFS, in any advertising or sales promotion which would indicate or imply that NMFS approves, recommends or endorses any proprietary product 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 purchased because of this NMFS publication. DAILY GROWTH INCREMENTS IN OTOLITHS FROM LARVAL AND ADULT FISHES Edward B. Brothers,^ Christopher P. Mathews,^ and Reuben Lasker^ ABSTRACT Daily growth increments have been found in otoliths offish larvae. The daily nature of these layers was verified by examining larval fish of known age reared in the laboratory. A simple technique for observing these marks is described and can be used on otoliths from larvae and adults. This provides a convenient method for determining early growth in fishes and is particularly useful for fishes which do not lay down annual or seasonal rings. The use of otoliths in age determination (by means of annual marks) is well known. The techniques used have been described by Williams and Bedford (1974) and Blacker (1974). Recently Pannella (1971) has suggested that daily marks may be formed in the sagittae (the otoliths used almost universally in age determinations) of some temperate species, while in 1974 Pannella claimed to have detected them in a number of tropical species. He also studied the temperate species — silver hake, Merluccius bilivoaris; red hake, Urophycis chuss; Atlantic cod, Gadus mor- hua; and winter flounder, Pseudopleuronectes americanus — in greater detail in this latter pa- per. For some of these temperate species, particu- larly for the first, Pannella was able to show that there were fortnightly, monthly, and annual pat- terns. The annual marks detected in the conven- tional way were shown to contain about 365 daily units. Pannella used acetate replicas of ground otoliths which had been previously etched with HCl. Pannella's work appears to justify the fol- lowing conclusions: 1 . Daily increments^ occur in certain temperate fish, e.g., M. bilinearis. 2. Periodic variations in increment thickness occur with fortnightly, monthly, and annual frequencies in this species. 'Scripps Institution of Oceanography, La Jolla, CA 92038; present address: Langmuir Laboratory, Section of Ecology and Systematics, Cornell University, Ithaca, NY 14853. ^Department of Fisheries, Escuela Superior de Ciencias Marinas, University of Baja California A.P. 453, Ensenada, B.C., Mexico. ^Southwest Fisheries Center, National Marine Fisheries Ser- vice, NOAA, P.O. Box 271, La Jolla, CA 92038. ■•The smallest visible concentric layers seen in an otolith. 3. Structural units that are similar to those shown to be daily in their occurrence in tem- perate species are also found in some tropical species. Pannella (1974) was careful to explain that the marks present in otoliths of tropical fish that ap- peared to be annual on the basis of conventional criteria could be deceptive. He suggested that by analogy with temperate species, certain struc- tures found in otoliths of tropical fish were also daily in occurrence. Although he found spawning marks, he did not find any seasonal or winter growth checks in the otoliths of tropical fish. In view of Pannella's expressed skepticism about the formation of annual marks and his tentative con- clusions, further evidence is needed that daily increments occur in tropical fish. Furthermore, no one appears so far to have attempted to apply this method of age determination to larval fish, yet it is in this last area that the most accurate and useful results might be expected. Pannella (1974) com- mented on the great regularity of the presumably daily marks near the center of the otoliths of both tropical and temperate fish. In these portions of the otoliths, no superposition of more complex patterns (e.g., 14 day, 28 day) were found. It is the object of this paper to show that 1) true daily increments are found in the otoliths of the larvae of several species, and that daily marks may be used to determine the ages of larval fish with great accuracy and precision, at least for approximately the first 100 days of life; and 2) in adults offish from a variety of habitats, including tropical waters, daily increments may be proven to exist, and so to confirm Pannella's work. Manuscript accepted July 1975. FISHERY BULLETIN: VOL. 74, NO. 1. 1976. FISHERY BULLETIN: VOL. 74, NO. 1 METHODS Some material was examined with a Stereo- scan^ S4 scanning electron microscope (Cam- bridge Scientific Instruments Ltd.). These otoliths were prepared for viewing by embedding them in polyester resin, grinding and polishing them to the vertical mid-sagittal plane with a graded series of silicon carbide or aluminum oxide com- pounds (400, 600, and 900 grit), and finishing with 1-yum diamond paste. The polished surface was then etched with 0.1 N HCl before being rotary coated in a vacuum evaporator with 150 A of gold-palladium alloy. Both this technique and that of Pannella (1974) involve the use of equipment and materials that may be inaccessible in many countries. This is particularly true for those countries in which daily growth increments might prove to be espe- cially helpful in stock assessment of commercial fish, so that an alternative practical method with minimal equipment was also used here and found to be successful. Otoliths of adult fish were ground by hand on a glass plate covered with a water-silicon carbide powder mixture (400-600 grit). The final polish may be administered with diamond paste, but this step is not essential. The ground otolith was then examined in immersion oil. The grinding was done in the same plane as described by Pannella (1974). It is possible that storage in oil over a long period of time may reduce the resolution obtained when an otolith is examined. This appears to be particularly true for larval otoliths. The above technique is simple and requires only a good com- pound microscope. Magnifications used in this work ranged to 1,800 x; at least 600 x is required for general viewing. Otoliths from larvae were removed by teasing them from fresh specimens. Oven-dried material needed only to be moistened with a drop of water before otolith removal. The otoliths were manipu- lated and transferred to clean slides by picking them up on the end of a fine dissecting needle wetted with immersion oil. No additional prep- aration was necessary, and the otoliths were examined in immersion oil or after being perma- nently mounted under a cover slip in a quick- drying, neutral mounting medium. Ground sec- tions from juveniles and adults may be similarly ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. mounted with no apparent loss in clarity. Larval otoliths are thin enough that only optical section- ing (i.e., carefully focusing to the plane of maximum clarity) is necessary to make total increment counts. Material from a variety of species was ex- amined and larval material of known age was obtained by rearing eggs that had been fertilized in the laboratory (Lasker et al. 1970; Leong 1971). The chronological age from these fish was known and could be compared with the number of growth increments observed in their otoliths. Larvae of northern anchovy, Engraulis mordax, were kindly made available to us by John R. Hunter of the Southwest Fisheries Center, National Marine Fisheries Service, NOAA, at La Jolla, Calif RESULTS Otoliths of 15 E. mordax, aged 6 days, were examined. The mean total length of the fish was 4.5 mm. The yolk-sac had been absorbed by the fifth day after hatching. Figure la shows the ap- pearance of the otolith of one of these fish. Only one or two daily increments were present, suggesting that daily growth increments ap- peared in the otoliths of E. mordax only after completion of yolk-sac absorption. In the labora- tory, anchovy larvae were maintained in 14 h of light when feeding took place and 10 h of darkness when no feeding occurred (Lasker et al. 1970). Table 1 shows the relation between chronologi- cal age and number of apparently daily incre- ments for larvae of E. mordax aged 6 to 100 days. It is clear that there is an extremely close corre- spondence between the chronological age in days and the number of increments. Figure lb is a micrograph showing the daily increments in an anchovy otolith from a larva 18 days old. Additional data presently being collected on laboratory and wild-caught larvae indicates that there is some interaction between the rate of larval growth and the rate of increment formation which may complicate the interpretation of oto- lith age estimates. Figure 2 shows the structure of adult anchovy otoliths with successively greater magnification of the scanning electron microscope. The darker areas in the photographs represent areas of the otolith that were more heavily etched because they contained a higher proportion of CaCOs, while the lighter areas have relatively more organic material, probably otolin (see Degens et al. 1969). It is seen from Figure 2 that the smallest BROTHERS ET AL.: DAILY GROWTH INCREMENTS IN OTOLITHS Figure l. — Light microscope photographs of otoliths from laboratory-reared northern an- chovy: a) 8-day-old larval otolith showing two daily growth rings; b) 18-day-old larval otolith showing 12 daily growth rings. Table l. — Chronological age (days from hatching) and numbers of growth increments in otoliths of northern anchovy. Number Cfironological Chronological Me an number of fish age in days age less 5 days of increments Range 15 6 1 1 0- 2 10 8 3 3 2- 4 10 12 7 7 4- 8 10 15 10 10 8- 11 7 16 11 10 9- 11 5 18 13 13 12- 15 7 20 15 15 14-16 8 24 19 18 16- 19 9 25 20 20 18- 21 3 26 21 21 18- 23 4 94 89 97 95-100 cyclical units are 1 to 2 ^tm thick in this part of the anchovy otolith and that they do not appear to contain any smaller units. It is these units that are counted and appear in the data in Table 1. The daily increment would therefore appear to be the smallest unit of growth that is formed at the supra-molecular level and, as such, is in principle the most natural unit to use for age estimation. Fertilized eggs of the California grunion, Leuresthes tenuis, were obtained and reared in the laboratory. The larvae were maintained in a natural light cycle at 17° to 20°C with food { Artemia nauplii) continuously available. Larvae were sacrificed at intervals and their otoliths were examined. Table 2 shows the results obtained and Figure 3 shows a photograph of a grunion otolith. Table 2 shows that there is a close relation be- tween the number of growth increments and the chronological age of the larvae. Although the agreement between age and daily increments is not as good as it is for the anchovy, the results are still very good. Table 2 also shows that in L. tenuis, daily increments appear at hatching, rather than at yolk absorption. Prehatching marks also occur, although they were not tallied in Table 2. Clearly the exact timing of the initi- ation of daily increment formation varies from FISHERY BULLETIN: VOL. 74, NO. 1 E a. O k. E o m ..^,. 4...4^aH>- aL.'.,.i ^.LWjk'. BROTHERS ET AL.: DAILY GROWTH INCREMENTS IN OTOUTHS Table 2. — Chronological age and number of growth increments in the otoliths of the California grunion. Number Chronological Mean number of fish age In days of increments Range 2 2 1- 2 3 7 9 8-10 2 16 11 10-12 3 18 17 16-18 5 26 24 20-26 species to species and must be independently determined for each one. Young striped bass, Morone saxatilis, were col- lected on 2 July 1974 in the Sacramento River delta (Tracy Pumping Station), Calif. These five fish measured 29 to 37 mm SL (standard length) and their otoliths had 62 to 120 observable incre- ments; i.e., a sample of striped bass which should have been 2 to 4 mo old according to their known spawning season (Scofield 1931) were 2 to 4 mo old according to the presence of growth layers found in their otoliths. The spread in the age calculated from daily increments probably corresponds to a considerable spread in the dates when the fish examined were hatched. Otoliths from two striped bass 135 and 142 mm SL were also examined. Published information on the growth rate of this species (Scofield 1931) indicates that striped bass of this size taken in July should be 14 to 16 mo old. The ages obtained by counting the presumed daily growth marks were 419 and 445 days respectively, i.e., 14 to 15 mo old. Figure 4 shows the daily marks in an otolith of striped bass. Daily increments were fairly thick near the center, thinner in an intermediate area corresponding to the hyaline zone, and wider again near the edge. In one specimen the central f» FIGURE 3. — Daily growth rings in an otolith of a California grunion larva. The larva was approxi- mately 26 days old. ^ Jf Figure 4. — ^Daily growth rings in a striped bass otolith. This fish was approximately 15 mo old. Differential growth can be seen in rings grown in adjacent seasons. F = fall; W = winter. FISHERY BULLETIN: VOL. 74, NO. 1 area contained 231 daily increments, the mar- ginal area contained 120, and there were 94 thinner marks in the middle zone. Working back- wards from the 2 July collection date, this indi- cated the slow growth zone occurred in December, January, and February. These figures correspond well with the known life cycle (Scofield 1931) which suggests a fast growth period in spring, summer, and fall (230 days, ~ 8 mo), a short winter of slow growth (~ 3 mo), and a spring and early summer (~ 4 mo) of faster growth prior to capture. Otoliths of postlarvae of the gobies Clevelandia ios, Ilypnus gilherti, and Quietula y-cauda were also examined. The fish were collected in Mission Bay, San Diego. The 2-mo larval period indicated in the otoliths agree with several independent estimates of the length of time between hatching and settlement (Brothers 1975). Otoliths of two species of hake obtained from the Gulf of California were studied. Mathews ( 1975) has shown that annual marks (annuli) may be detected by means of the usual discrimination of hyaline and opaque zones in Merluccius an- gustimanus while in Merluccius sp. (Mathews in press) the same techniques have also been applied successfully. The ages of hake determined by means of annuli may be compared with age determined from counting the number of daily increments; these are identified by analogy with the structures shown to be daily in their incidence in anchovy, grunion, striped bass, and other fish and which appear to be the same as those shown by Pannella (1971) to be daily in M. bilinearis (Figures 5, 6). In most cases, direct total counts were not possible because increments were not equally visible over a complete nucleus to margin radius. For these otoliths measurements of incre- ^ 50jjm I I ^ O I Figure 5. — a) Nucleus of an otolith from aMer- luccius sp., 7 yr old; b) daily growth increments shown from near the center of the otolith. BROTHERS ET AL.: DAILY GROWTH INCREMENTS IN OTOUTHS lOjjm Figure 6. — Daily growth increments from the otolith of Merluccius angustimanus . Note radial fibers crossing the growth layers. ment width were made at five or more locations along a radius and then total counts were calcu- lated by extrapolation. No larval or very young hake were available for examination. For Merluccius sp., data were available for 22 specimens aged 1 to 7 yr from the annuli present in their otoliths. Figure 7 shows the graph of age by annuli against age by daily increments for this species. The correlation coefficient was 0.91 (20 df, P » 0.001). The slope of the regression line was 1.14 (99% confidence limits [C.L.], 0.81-1.46). This is not significantly different from the value 7i- 1 2 3 4 5 6 7 AGE BY DAILY GROWTH INCREMENTS (years) Figure 7. — Graph of age-by-annuli against age-by-daily- growth-rings in the otoliths of Merluccius sp. The encircled point represents two points at the same position. of 1.00 expected if age by years and by days were to yield identical values. Data from seven specimens of M. angusti- manus were available and they varied in age from only 1 to 2 yr. Given the much narrower ranges and the smaller sample, the results obtained were acceptable: r = 0.74 (0.05 > P > 0.01) and the slope of the line was 1.25 (99% C.L., 0.24-2.25); i.e., the slope was significantly different from zero, but not from 1.0. The precision of estimates of age obtained for M. angustimanus was not very good, with deviations of up to 0.5 yr being obtained; however, for Merluccius sp. a somewhat narrower range was usual, with some values differing by 0. 1 yr or less. Extreme variations occurred with fish aged 7 to 13 yr, where errors of up to 2 to 3 yr could be obtained where daily counts were made. The average widths of the daily bands found in the hake otoliths were 3 to 4 /xm, with wider and narrower bands appearing sometimes in appar- ently weekly, fortnightly, and monthly units. The incidence of these units has not been examined in detail and requires further study, but preliminary work suggests that the basic unit used in age esti- mates should be the daily unit; the higher order units may be of great ecological interest, but should probably not be used in aging these hake: Only daily increments occur with the necessary consistency and regularity. In addition to the species mentioned above, apparently daily marks have been found in a wide variety of other fish, e.g., in Tilapia zilli, T. nilotica, and Clarias mossambicus from Lake Victoria (examined by E. B. B. and C. P. M.; specimens kindly collected by John Rinne and Dr. FISHERY BULLETIN: VOL. 74. NO. 1 Peretti of the East African Freshwater Fisheries Research Organization, Jinja, East Africa), and the following species examined by one of the authors (E.B.B.): in the deep living Pacific rattail Coryphaenoides acrolepis (58 cm SL; 10 to 11 yr); in the myctophids Stenobrachius leucopsarus, Tarletonbeania crenularis, and Triphoturus mexi- canus; in the freshwater fish Cottus asper and Salmo gairdneri; in the tropical marine fish Chromis atrilobata and Apogon retrosella; in adults of the gobies Cleuelandia ios and Gil- lichthys mirabilis, where clear growth checks also occur, so that daily marks alone would lead to distinct underestimates of age; and in four species of rapidly growing tropical and temperate tunas. Statoliths from the squid Loligo opalescens (both wild caught adults and laboratory-reared juve- niles) also show what appear to be growth layers analogous to those in fish otoliths. The appear- ance of growth interruptions in a number of species, e.g., the rockfish (genus Sebastes), either as winter checks, spawning checks, or apparently dispersed more evenly throughout the year, may impose a severe limitation upon the use of daily marks to age these fish. The technique seems best suited to larvae, juveniles, fast-growing species, and tropical species. It is clear from our work that some difficulties must be overcome before age estimation by means of daily rings can become a standard tool in fish- eries biology. However, it is also clear that 1. Daily rings may be used to estimate the ages of larvae of some species up to 100 days old with very great precision and that they prob- ably can be used for fish up to 1 yr of age, perhaps with a smaller degree of precision. Struhsaker and Uchiyama (1976) show simi- lar results with the tropical engraulid Stole- phorus purpureus. 2. Daily marks may be used as a means of ac- curate age determination for at least some species of fish up to 6 yr old. 3. Daily marks may be used for age determina- tion of at least some tropical fish. Pannella's (1974) suggestion that daily increments might be used in tropical fish as a means of age estimation is almost certainly true, and should be applicable to most species. LITERATURE CITED Blacker, R. W. 1974. Recent advances in otolith studies. In F. R. Harden Jones (editor), Sea fisheries research, p. 67-90. John Wiley and Sons, N.Y. BROTHERS, E. B. 1975. Comparative ecology and behavior of three sjonpatric California gobies. Ph.D. Thesis, Univ. California, San Diego, 370 p. DEGENS, E. T, W. G. DEUSER, AND R. L. HAEDRICH. 1969. Molecular structure and composition offish otoliths. Mar Biol. (Berl.) 2:105-113. Lasker, R., H. M. Feder, G. H. Theilacker, AND R. C. May. 1970. Feeding, growth, and survival of Engraulis mordax larvae reared in the laboratory. Mar. Biol. (Berl.) 5:345-353. LEONG, R. 1971. Induced spawning of the northern anchovy, En- graulis mordax Girard. Fish. Bull., U.S. 69:357-360. MATHEWS, C. P. 1975. Some observations on the ecology and the population dynamics of Merluccius angustimanus in the south Gulf of California. J. Fish. Biol. 7:83-94. In press. The biology, ecology and population dynamics of the large Gulf of California hake. Symposium in Fisheries Biology, Ensenada, B.C., Mexico. Ciencas Marinas, Spec. Suppl. PANNELLA, G. 1971. Fish otoliths: daily growth layers and periodical patterns. Science (Wash., D.C.) 173:1124-1127. 1974. Otolith growth patterns: An aid in age determina- tion in temperate and tropical fishes. In T B. Bagenal (editor), The ageing of fish, p. 28-39. Unwin Brothers, Ltd., Surrey. SCOFIELD, E. C. 1931. The striped bass of California (Roccus lineatus). Calif. Dep. Fish Game, Fish Bull. 29, 84 p. STRUHSAKER, P., AND J. H. UCHIYAMA. 1976. Age and growth of the nehu, Stolephorus purpureus (Pisces: Engraulidae), from the Hawaiian Islands as indi- cated by daily growth increments of sagittae. Fish. Bull., U.S. 74:9-17. WlLLUMS, T., AND B. C. BEDFORD. 1974. The use of otoliths for age determination. In T. B. Bagenal (editor). The ageing offish, p. 114-123. Unwin Brothers, Ltd., Surrey. 8 AGE AND GROWTH OF THE NEHU, STOLEPHORUS PURPUREUS (PISCES: ENGRAULIDAE), FROM THE HAWAIIAN ISLANDS AS INDICATED BY DAILY GROWTH INCREMENTS OF SAGITTAE Paul Struhsaker and James H. Uchiyamai ABSTRACT Direct evidence is presented that the sagittae of nehu, Stolephorus purpureas, grow by discernible daily increments. Aging by daily growth increments provides the means to establish a general growth curve for the first 6 mo of life for this species. Adult nehu exhibit nearly linear growth between 30 and 60 mm standard length. Preliminary evidence is presented that the nehu population of Pearl Harbor may grow more rapidly than that of Kaneohe Bay. Attempts to age tropical fishes by conventional methods have generally been thwarted by the absence of well-defined annuli in calcarious structures and protracted spawning periods which make length-frequency mode progression analyses difficult. Recognizing that exceptions to the above statement exist, Pannella's work (1971) providing indirect evidence of the pres- ence of daily growth layers and periodical deposition patterns in the sagittae (otoliths) of three species of boreal fishes from the western North Atlantic suggested a means for conducting age and growth studies of tropical species. He concluded in that report: "Preliminary observa- tion of growth patterns in sagittae of other species, living at various depths and different climates, appears to support the idea that daily growth may be a universal feature of fish oto- liths." Pannella's (1974) later work in Puerto Rico provided circumstantial evidence of daily growth layers in sagittae of several species of tropical fishes. To gain direct evidence that daily growth incre- ments exist in tropical fishes we studied the nehu, Stolephorus purpureas Fowler, a small engraulid endemic to the Hawaiian Islands. The nehu is the basis of a live-bait fishery producing about 4,000 metric tons annually of skipjack tuna, Katsu- wonus pelamis (Linnaeus), from the vicinity of the Hawaiian Islands. Stolephorus purpureus is a short-lived species (less than 1 yr) and has been the subject of relatively numerous studies: Naka- mura (1970) has summarized the biological 'Southwest Fisheries Center, Honolulu Laboratory, National Marine Fisheries Service, NOAA, Honolulu, HI 96812. knowledge of this species available through 1965. Our work provides evidence of the presence of daily growth increments in the sagittae of nehu and permits the assembly of a growth curve for the first 6 mo of life for this species. Brothers et al. (1976) have recently demon- strated the presence of daily growth increments in larval Engraulis mordax Girard and Leuresthes tenuis (Ayres) and presented evidence that the phenomenon occurs in several other species of California fishes. METHODS AND MATERIALS The nehu samples were taken with three types of gear in Pearl Harbor and the southeastern end of Kaneohe Bay, Oahu, Hawaiian Islands. Adults and juveniles (> about 30 mm standard length (SL) ) were sampled with commercial bait seines (square mesh measuring 3.2 mm to a bar) in Pearl Harbor. Postlarvae (about ^ 20 mm SL), juveniles, and adults were obtained in Kaneohe Bay by a similar seine having a bar mesh measurement of 1.6 mm. Larvae (< 20 mm SL) were obtained near Coconut Island by personnel of the Hawaii Insti- tute of Marine Biology with 0.5-m ring nets with mesh sizes of 550 /um. Three separate holding experiments were con- ducted to test the hypothesis that the sagittae of nehu grow by discernible daily increments. All animals for these experiments were collected in Pearl Harbor and held in tanks of 38-kl capacity at the National Marine Fisheries Service (NMFS) Kewalo Basin Facility. The tanks were supplied with well sea water of 23°-24°C and 33-35%o salin- ity at a rate of about 300 liters/min. The nehu Manuscript accepted August 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 1976. FISHERY BULLETIN: VOL. 74, NO. 1 were fed with frozen and live brine shrimp, Artemia sp., under variable regimes as described below. Each experimental population of nehu was sampled during placement in holding tanks, and then subsampled at various time intervals as described for each experiment. Otoliths were ex- tracted from most specimens within a few hours of sampling. The remaining samples were frozen in seawater or preserved in 75% solution of iso- propanol until extraction of otoliths (removal of tissue from otoliths of alcohol preserved speci- mens is difficult). The first holding experiment was begun 5 April 1972. A 16-day sample (21 April) and a 34-day sample (9 May) were obtained from this population. The animals were fed once a day With frozen and/or live brine shrimp. The second holding experiment was begun 15 Decem- ber 1972. This population was initially fed once a day. A high mortality was observed during the first 2 wk, after which food was provided twice daily. Samples were collected weekly after 1 mo of captivity. We examined sagittae from animals collected on 19 January and 26 January 1973. The third holding experiment was begun 4 May 1973. This population was fed two or three times daily with frozen brine shrimp. Samples were obtained weekly between 4 May and 6 July. We examined sagittae from animals collected 25 May and 8 June 1973. Wild populations of larval, juvenile, and adult nehu were sampled 13 times in Kaneohe Bay be- tween 19 March 1972 and 13 July 1973 to obtain estimates of growth rates at various seasons. Al- though a second species of Stolephorus (S. buc- caneeri Strasburg) occurs in Hawaii, larvae of this species have not yet been collected in the south- eastern end of Kaneohe Bay (Watson and Leis 1974; W. Watson pers. commun.). After extraction, the sagittae were cleaned and etched for up to 3 min in a 1% solution of HCl, then washed and mounted whole on glass slides with the mounting medium EuparaP and covered with glass cover slips. Short lengths of monofilament line were used to prevent the contact of the specimen by the cover slide. Although the small- est growth increments are microscopically dis- cernible immediately after extraction their detec- tion was enhanced after about 30 days of clearing in the mounting medium. Sagittae used in the ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. first holding experiment and those collected from Kaneohe Bay and Pearl Harbor during spring 1972 were placed in glycerine on slides and covered. Some erosion of the sagittae edges was noted after about 5 mo, and this practice was discontinued after the first experiment. Slides were either labeled with date of collection and length of fish or assigned a five digit random number for identification. Our initial counts were taken from thin sections of sagittae taken on the frontal plane. After mounting the sagittae in epoxy resin, the initial plane of polishing was made with rough sand- paper. As the surface approached the desired section, fine wet silicon carbide sandpaper (400 grit) was used. Final polishing of the surface was done with suspensions of aluminum oxide parti- cles having diameters of 15, 5, and 0.3 /^m. The section was thinned on the opposite side to a practical thickness and etched in a 1% solution of HCl for variable periods up to 3 min. A few attempts to make acetate peels of the small nehu sagittae sections as described by Pannella (1971) and Pannella and MacClintock (1968) were un- successful. We eventually abandoned the section- ing of sagittae because of the time required and the difficulty in obtaining a precise section from the nucleus to the posterior edge of the sagitta. Sagittae were obtained from larvae less than about 20 mm SL by placing the specimen on a slide and gently teasing the otoliths from the head re- gion. The sagittae were then mounted in Euparal and read immediately. These otoliths tended to clear completely within a few hours, and photo- graphs are the only permanent record of these specimens. The smallest growth increments of the mounted sagittae were counted with a compound micro- scope at magnifications of 400-800 x . The smallest growth increment in all fish otoliths consists of both an organic and an inorganic layer (Degens et al. 1969). These two layers in the nehu otolith to- gether measure about 1-4 ^im thick. A zoom fea- ture of the microscope was found to be extremely useful. Counts were maintained on a hand tally. Enumeration of the smallest growth increment layers in whole sagittae is tedious, and reliable counts can be obtained only after a moderate amount of experience has been acquired. Enu- meration is, obviously, much easier in sagittae from smaller fishes (Figure 1). Usually, readings cannot be made in a direct line from the nucleus to the selected point on the edge of the sagitta; 10 STRUHSAKER and UCHIYAMA: AGE AND GROWTH OF STOLEPHORUS PURPUREUS B 4 ^ * I 4 41 * .4 m A' * , N ,1 * ^ % v^ D *, < %■ 7, e.^ ■^ ^ Figure l.— Sagittae of larval Stolephorus purpureas. A: Portion of sagitta from a 28.8-mm SL individual with about 65 growth increments. B: 12.6 mm SL, 14 increments. C: 7.3 mm SL, 7 increments. D: 3.9 mm SL, 1 increment. 11 FISHERY BULLETIN: VOL. 74, NO. 1 rather, a somewhat circuitous route must usually be taken from one area of the sagitta to another by following a prominent growth increment. Each sagitta was counted several times in succession, the number of counts (up to 10) being proportional to the size of the sagitta. Counts were made from the nucleus to the antirostrum, ros- trum, and postrostrum (terminology of Messieh 1972). A consistent count for the number of lamel- lae was then obtained. Verification counts were then made by the same reader at a later time. Ver- ification counts were made by a second reader on 167 otoliths from the second and third holding ex- periments, as well as randomly selected sagittae representing the wild populations: 26.3% of these counts agreed with the original count; 48.5% dif- fered by less than 1%; 72.5% differed by less than 2%; 86.9% differed by less than 3%; 92.9% differed by less than 4%; and 95.9% differed by less than 5%. Errors of less than 5% were considered accept- able, and the median values of the two readers were then utilized in the analyses. In cases where the results differed by more than 5% , the sagittae were reexamined and either a consensus of opin- ion reached or the data discarded. Standard lengths were taken to the nearest 0.01 mm with dial calipers. Sagittae were measured with a micrometer eyepiece. RESULTS Holding Experiments The holding experiments were undertaken as one means to determine if the smallest growth increments observable in the sagittae of nehu rep- resent daily growth increments. We examined sagittae of specimens from samples taken at vari- ous time periods after the initial collection to determine if there was an increase in mean number of increments approximating the num- bers of days between sampling. (Length data collected from all samples indicate that the length-frequency distributions of most of the captive populations studied were normally distributed.) The data obtained for each holding experiment were subjected to analysis of covariance and the results are summarized in Table 1 and Figures 2- 4. There was homogeneous variance within the samples for each of the three experiments as indi- cated by Bartlett's test of homogeneity (chi-square values = 0.56, 3.59, and 0.59, respectively). In the first experiment there were no significant differences between the means of the independent variable (standard length) for each of the three samples at the P< 0.05 level. There were signifi- cant differences between the regression coeffi- cients and the ielevation of the regression curves for each sample at the P<0.01 level (Table 1, Figure 2). The significant differences between regression coefficients seems best explained by the effects of captivity. Hypothetically, the regression coeffi- cient of the initial sample of 5 April represents the relationship between number of growth incre- ments and standard length in the wild population. The smaller regression coefficient value of the 21 April sample indicates a slower growth rate of the captive population during the 16-day interval between sampling. This is probably due to less than optimal food supply and/or other effects of captivity. The intermediate regression coefficient value of the 9 May sample indicates that the Table l. — Summary of analysis of covariance for three holding experiments. Sampling date Dependent variable (Increments) F ratios Unadjusted y Adjusted 7 Independent variable (standard lengthi) Regression coefficient Elevation 5 Apr 1972 21 Apr 1972 9 May 1972 First experiment 19 Jan. 1973 26 Jan. 1973 Second experiment 25 May 1973 8 June 1973 Ttiird experiment 84.9 101.1 118.1 114.8 120.8 124.9 140.0 86.7 0.77 30 38.2 100.6 0.76 24 10.8 116.4 0.74 24 20.6 114.0 0.95 25 14.7 121.6 0.85 24 31.1 132.1 0.97 23 13.8 133.4 0.95 24 6 1 1.2- 0.1 34— 5.4- 1.3 1.1 206*" 31* 1.1 "P sO.01. '"P 5^0.001. 12 STRUHSAKER and UCfflYAMA: AGE AND GROWTH OF STOLEPHORUS PURPUREUS 150 140 130 120 110 -1 \ r- -T 1 1 1 T MAY 9, 1972 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 STANDARD LENGTH (mm ) Figure 2. — Stolephorus purpureus: First holding experiment. 170 160 ISO 140 \2 130 z UJ S 120 CO u S no 100 90 80 JANUARY 26, 1973 JANUARY 19, 1973 _l 1 I 1 1 ] 1 I I L. 34 36 38 40 42 44 46 48 50 52 54 56 58 STANDARD LENGTH (mm) Figure 3. — Stolephorus purpureas: Second holding experiment. 170 160 ISO I- z UJ S 130 UJ (£ O ? 120 MO 100 -I 1 1 1 T "1 1 ^ r MAY 25,1973 >xA( JUNE 8,1973 46 47 48 49 50 51 52 53 54 55 56 57 58 STANDARD LENGTH (mm) Figure 4. — Stolephorus purpureus: Third holding experiment. growth rate has increased in the captive popula- tion after 34 days in captivity, but has not reached the value of the wild population from which it was taken. In the first holding experiment, the second and third samples were collected 16 and 34 days, re- spectively, after the initial sample. For unadjust- edy values, these samples differed from the initial sample by 16.2 and 33.2 increments, whereas for the adjusted y values, they differed from the initial sample by 13.9 and 29.7 increments (Table 1). The results of the two samples (collected after more than 30 days in captivity) collected 19 and 26 January 1973, and compared in the second hold- ing experiment, are summarized in Table 1 and Figure 3. There were no significant diff'erences between the means of the independent variables or the regression coefficients at theP ssO.OS level. The elevations of the two regression curves are significantly different at theP ^0.001 level. The differences in number of increments between unadjusted y values (6.0) and adjusted y values (7.6) again closely approximate the expected dif- ference of 7 days between samples. The results of the samples of 25 May and 8 June 1973 compared from the third holding experiment are given in Table 1 and Figure 4. In this experi- ment there was a significant difference between the means of the independent variable (P <0.001), but no differences between the regression coef- ficients and elevations of the two regression curves at theP «0.05 level. The significant differ- ence in mean length between the two samples is probably attributable to the increased amount of food provided to the captive population and the resulting high growth rate exhibited throughout the duration of the experiment. Because the treatment significantly affected the independent variable, further examination of the regression statistics is unwarranted. However, if the two samples are subjected to a two-group comparison, there is a significant difference between the mean number of increments for each sample (P <0.05). The difference between the means for each sample (25 May,y = 124.9; 8 June, J = 140.0) closely ap- proximates the expected difference of 14 days be- tween samples. We conclude from the relatively good agree- ment between the increase in mean number of growth increments and the number of days be- tween collection of samples, that these data from the holding experiments provide direct evidence of the presence of daily growth increments in the sagittae of nehu. Growth of Sagittae The total lengths of sagittae from the 5 April and 9 May 1972 nehu samples (the initial sample from the wild population and the 34-day sample) 13 FISHERY BULLETIN: VOL. 74, NO. 1 of the first holding experiment were taken in order to examine the effects of captivity on sagit- tal growth. Four measurements for the 5 April sample were arbitrarily deleted because their values were well below the distribution of the majority of the sample. All 24 measurements from the 9 May sample were utilized. There are signifi- cant relationships between sagitta length and fish length for the two samples (P <0.001, r^ values: 5 April, 0.82; 9 May, 0.70) (Figure 5). The first experiment demonstrated that there was a signifi- cant increase in the mean number of increments between the two samples. Analysis of covariance of sagittae lengths indicated that there were no significant differences between the means of the independent variables, regression coefficients, or elevations of the regression curves for the two samples (respective F ratios: 2.5, 1.0, 1.2) presum- ably because of intrinsic variation, limited preci- sion of measurements, and the relatively short time period between samples. Although there were no statistically significant differences found in the comparison of the two curves, the two regression coefficients exhibit perhaps expectable trends. The lesser regression coefficient and r^ value for the 9 May sample may be indicative of a decreased growth rate and more variable re- sponses of individuals in the population to the highly variable, and probably less than optimal, conditions of the holding facility. In addition, the differences between the unadjusted and adjusted means of sagittal lengths between the 5 April (1.094 mm; 1.070 mm, respectively) and 9 May (1.176 mm; 1.201 mm, respectively) samples of 0.082 mm and 0.131 mm are to be expected with daily growth increments of about 3-4 /u.m. We have noted one apparent example of pro- visioning rates affecting the growth rates of sagit- tae of captive nehu. Sagittae from the 19 January sample of the second holding experiment usually exhibited 23-24 distinctive, more widely spaced increments on the edge of the otolith. The num- bers of distinctive increments approximately cor- respond to the number of days during which the daily amount of food provided the sample popula- tion was double the initial ration. As might be ex- pected otoliths collected 7 days later in the 26 January sample exhibited 30-31 distinctive incre- ments. Indeed, the wider increments observed after provisioning rates were doubled were much more effective in "labeling" the sagitta than our attempts to accomplish the same objective with Tetracyclene. Possibly, controlled experiments MAY 9,1972 APRIL 5,1972 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 STANDARD LENGTH (mm) Figure 5. — Stolephorus purpureas: Growth of sagittae during first holding experiment. with rapidly growing fish species incorporating this treatment would be a much more expeditious test of the daily growth increment hypothesis. Age and Growth in Wild Populations We examined larval, juvenile, and adult nehu collected in Kaneohe Bay to obtain an estimate of age and growth of a wild population based on the assumption that the smallest observable growth layers in the sagittae represent daily growth in- crements. We examined 213 specimens from 13 collections made during most seasons between spring 1972 and summer 1973 (no collections were made in the months November through January). The growth curves obtained from the individual collections are given in Figure 6. Because all individuals in a sample have been exposed to the vagaries of the environment during their ob- served lifespan, a composite growth curve for all collections is presented in Figure 6F. Although some variation between samples is apparent, the composite scattergram serves as a first estimate of the growth pattern of nehu in Kaneohe Bay. There are two well-defined segments to the composite growth curve (Figure 6F). Young lar- vae exhibit exponential growth to a length of about 15-17 mm. At about 20 mm the population enters an almost linear growth phase to about 60 mm. The composite scattergram obscures another, lesser inflection at about 20-30 mm exhibited by the spring 1972 collections (Figure 6A). Yama- shita (1951) has demonstrated that nehu have completed larval metamorphosis at about 30 mm. The major inflection at a length of about 17 mm appears to reflect the fact that nehu begin to exhibit exponential growth in body depth at this 14 STRUHSAKER and UCHIYAMA: AGE AND GROWTH OF STOLEPHORUS PURPUREUS E E o z o tr < o z o o Q 1 1 ! I 1 1 I 60 ^ (A) o 50 ~ o o " 40 - <* - 30 ~ . a, °° * # - 20 .^ - 10 1 1 1 1 - 60 50 40 30 a tr < o z 20 240 Total Standard length' (mm) Female Male Pamlico Sound Morehead City Pamlico Sound Morehead City Number % mature Number % mature Number % mature Number % mature 42 4 4 3 5 20 15 7 18 22 15 7 12 42 20 15 26 23 24 21 19 26 27 48 28 57 45 69 16 81 27 93 18 78 20 85 10 80 10 100 6 100 9 100 5 100 83 99 469 4 25 11 24 6 33 21 67 29 76 31 87 20 90 23 100 26 85 24 100 24 96 29 90 23 96 27 96 16 100 17 100 11 100 12 100 8 100 3 100 16 100 383 100 5 20 4 9 33 12 33 9 11 13 61 15 47 18 56 24 83 19 58 15 87 31 77 20 85 34 77 36 94 25 100 32 97 14 100 13 100 15 100 7 100 6 100 4 100 5 100 25 100 2 7 14 2 2 1 100 5 60 3 67 9 56 9 67 15 80 25 96 32 100 40 95 33 97 24 100 28 100 22 100 40 100 28 100 35 100 19 100 21 100 10 100 4 100 7 100 4 100 3 100 1 100 411 473 'Midpoint of length interval (102.6 to 107.5 = 105, etc.] Table 6. — Relationship of age-group and standard length to percent sexually mature by sex for weakfish from the vicinity of Morehead City, N.C. ( 1968-69). slOO 110 115 120 125 130 135 140 145 150 155 160 165 170 175 180 185 190 195 200 >200 Total Mean Standard length' Age-group Age-group 1 Age-group II Male Female Male Female Male Female (mm) Number % mature Number % mature Number % mature Number % mature Number % mature Number % mature 42 2 7 2 2 1 4 2 7 7 13 22 31 35 14 7 4 203 14 100 50 50 43 71 77 96 100 97 100 100 100 100 67 4 25 9 44 5 40 19 63 29 76 26 85 9 78 13 100 5 100 1 100 1 100 1 100 2 100 2 50 2 100 3 100 1 100 5 80 19 98 16 100 24 100 22 100 38 100 26 100 82 100 2 100 1 2 100 5 100 11 100 10 100 21 81 22 100 23 96 29 90 22 96 81 99 100 1 100 2 100 22 100 122 244 229 26 1 1 1 29 32 73 99 95 100 100 100 100 100 100 'Midpoint of length interval. the female weakfish of age-group I less than 175 mm SL (n = 21). All females in age-group II were mature (n = 32). Males of age-group from Pam- lico Sound were sexually mature at 150 mm SL while males in age-group I reached maturity at 135 mm SL (Table 7). All age-group II males examined from this area were sexually mature {n = 89). Females of age-group in the Pamlico Sound area reached sexual maturity at 175 mm SL (n = 3, 1 immature) while age- group I females 22 MERRINER: REPRODUCTIVE BIOLOGY OF THE WEAKFISH Table 7. — Relationship of age-group and standard length to percent sexually mature by sex for weakfish from Pamlico Sound, N.C. (1967-69). Standard length' (mm) Age-group Age-gro up 1 Age-groL P II Male Female Male Female Male Female Number % mature N umber % mature Number % mature Number % mature Number % mature Number % mature 105 1 100 125 4 25 3 1 1 130 2 3 2 1 135 6 17 3 3 67 140 9 22 3 3 67 2 50 145 5 12 4 25 3 33 150 12 58 10 10 1 100 8 38 155 7 43 9 11 8 50 6 160 4 50 5 40 14 57 7 43 165 4 100 5 20 20 80 14 7 1 100 170 1 100 6 17 18 55 20 25 175 2 100 3 67 13 85 21 14 180 3 100 31 77 16 12 185 19 84 27 48 1 100 190 1 33 76 25 52 1 100 195 1 100 1 100 31 94 44 68 4 100 200 1 100 17 100 13 85 7 100 3 67 205 19 95 25 92 12 100 2 100 210 7 100 15 73 7 100 3 100 215 4 100 14 93 9 100 6 67 220 5 100 5 60 10 100 5 100 >220 9 100 33 100 38 100 80 99 Total 59 67 262 300 89 100 Mean 44 18 80 56 100 96 'Midpoint of lengtfi interval. were mature at a length of 190 mm (n = 25, 52% mature) (Table 7). Average estimated fecundity increased with age from 45,000 eggs for age females to 1,726,000 eggs for age IV females. The increases in fecundity with age were significant (F = 15.64, df = 17.4; P < 0.01; Table 8). Variation within individual age groups was great with the stan- dard deviation approaching one-third of the mean estimated fecundity. Relative fecundity, the number of eggs per gram of ovary, decreased from 37,650 at age to 14,867 at age IV. Regression analysis indicated significant rela- tionships between fecundity and fish length and weight. The equations describing the relation- ships and coefficients of determination are: F -- F -- F -- Table 8. 0.116SL2"55^^2 = 0.85; 0.152 TL2-64i8^ ^2 = 0.86 (Figure 4); 21,198 + 1,279 W, r2 = 0.88. DISCUSSION Weakfish spawn in or near the various inlets along the coast of North Carolina (Welsh and Breder 1923; Higgins and Pearson 1928; Hilde- brand and Cable 1934) and also in Pamlico Sound. Earlier authors did not include sounds and bays as probable spawning sites since no female weakfish in spawning condition had been taken from inshore waters of North Carolina (Roelofs 1951). Higgins and Pearson (1928) re- ported a few weakfish with "free running ripe eggs" in Pamlico Sound. Twenty-four female weakfish in the ripe running condition were ob- tained from Pamlico Sound, and this indicates weakfish may also spawn in sounds and bays. These areas may be at the edge of the spawning zone, however. Spawning activity in coastal waters north of North Carolina is cited by Hildebrand and -Fecundity estimates and relative fecundity for 22 weakfish from North Carolina and analysis of variance results for age versus fecundity. Age- group Number examined Mean fecundity estimates Standard deviation Mean no. of eggs per gram of ovary Standard length range (mm) Anova Source df Sum of squares Mean square F 1 II III IV Total 2 8 7 2 3 22 44,880 285.740 579,660 491,700 1,725,920 10,693 105,600 302,700 186,900 614,300 37,650 21.225 19,400 15,150 14,867 145-160 190-268 245-308 292-335 395-480 Age Error Total 4 17 21 5.219 X 10'2 1.418 X 10'2 6.637 X 10'2 1.305 X 10'2 8.341 X 10'" 15.64*' "Probability less than 0.01. 23 FISHERY BULLETIN: VOL. 74, NO. 1 O O UJ 26-1 24- 22- 20- 18- 16- 14- 12- 10- 8- 6- 4- 2- 0- FECUNDITY = 0.152 TL 0.86 2.6418 FECUNDITY = 0.116 SL ^ - -'']^ '2 = 0.85 T" T 100 200 300 400 500 FISH LENGTH (mm) 600 FIGURE 4.- - Relationship of weakfish fecundity to fish length based upon data from 22 females. Schroeder (1927), Pearson (1941), and Massman (1963) for Chesapeake Bay; by Parr (1933), Daiber (1954), Harmic (1958), and Thomas (1971) for Delaware Bay; by Nesbit (1954) and Perlmut- ter et al. (1956) for New York and New Jersey waters; and by Bigelow and Schroeder (1953) for the Gulf of Maine. However, the magnitude of spawning in northern areas is unknown. Progeny from spawning activity north of Chesapeake Bay are considered insufficient to maintain the north- ern stock (Harmic 1958), and young from the Carolinas and Chesapeake Bay are thought to be recruited to the northern population as age III or older fish (Pearson 1941; Nesbit 1954; Perlmutter et al. 1956; Harmic 1958). The validity of this supposition remains to be documented. Mature weakfish enter the inshore waters, sounds, and bays of North Carolina in early spring (Hildebrand and Schroeder 1927; Hilde- brand and Cable 1934; Roelofs 1951). Fertilized eggs have been taken in Delaware Bay when water temperatures ranged from 17° to 26.5°C and at salinities from 12.1 to 31.3'L (Harmic 1958). Weakfish apparently have an extended spawn- ing season in North Carolina waters as reported by Welsh and Breder (1923), Higgins and Pearson (1928), Hildebrand and Cable (1934), and Pear- son (1941). Distributional data for weakfish eggs and larvae are lacking in North Carolina waters. Peak spawning activity occurs from late April through June as indicated by gonad condition and gonadal index. Females appear to spawn the major portion of their eggs in May or June with a second spawn of smaller magnitude possibly oc- curring in late July or August. Thus, weakfish of a given year class may vary considerably in size due to their extended spawning season and mul- tiple spawning by females. Weakfish males and females probably attain sexual maturity as 1-yr-old fish throughout their geographic range, though some of the small- er members of a year class may not mature until their second year of life. Weakfish in North Carolina waters were previously reported to reach sexual maturity at age II for males and age III for females (Taylor 1916; Welsh and Breder 1923; Higgins and Pearson 1928), and subsequent papers have reiterated these ages without ver- ification. Higgins and Pearson (1928) reported no mature females less than 200 mm fork length (approximately 170 mm SL) and that a fork length of 230 mm was attained before 50% of the female weakfish mature in Pamlico Sound. This size group was allocated to age-group III without examining scales for annuli. I consider their allo- cation of age-classes to be in error on the basis of data presented here and in Merriner (1973). I found 21 mature female weakfish 170 mm SL in samples from Pamlico Sound and 90 mature female weakfish of the same size from the vicinity of Morehead City. Over one-half of the female weakfish were mature at 190 mm SL in samples from Pamlico Sound, and male weakfish become sexually mature at a smaller size than females. Weakfish spawned in May or June would be ma- ture the following May or June. Those fish spawned in late July or August probably would not be sexually mature until late summer of the year following their hatch or the following spring. Scrap samples from pound nets in Chesapeake Bay contained mature female weakfish measur- ing 170 to 250 mm TL during late spring and summer months (McHugh 1960). Maturation at a small size is also likely for fish from more north- erly areas (Daiber 1954; Thomas 1971). No evidence of alternate year spawning was found even in the oldest specimens examined. All of the females of age III or older were either in spawning condition or mature during early sum- mer. However, some of the older weakfish in the population may not migrate inshore during spring and summer. Weakfish are characterized by high fecundity. 24 MERRI>fER: REPRODUCTIVE BIOLOGY OF THE WEAKFISH In Delaware Bay a female weakfish, 190 mm SL, contained a total of 267,500 eggs and would re- lease approximately 52,000 eggs at one spawning (Daiber 1954). My estimates of fecundity for females of a similar size are equivalent to the total egg production figure for Delaware Bay. Fecundity increases by approximately 106,000 eggs for each 100 g of body weight for weakfish in Delaware Bay, while my data indicate an in- crease of 127,900 eggs per 100 g of body weight. The variation in fecundity per age-group is best explained by the size range present in the sam- ples of each age-group. Regression analysis showed a significant relationship between fecun- dity and fish length (coefficient of determination = r^ = 0.85) and between fecundity and fish weight (r^ = 0.88). The average range of standard length for all females in age-groups to IV was 57 mm. High variability in fecundity estimates for age-groups is expected due to the range in fish size and variation in gonad size among fish of the same size (Bagenal 1967). It is highly unlikely that weakfish experienced a synchronous failure or severe depression of em- bryonic or larval survival in all spawning areas. Harmic (1958) analyzed the early life history of weakfish in Delaware Bay. Fertilized eggs are pelagic and measure from 0.87 to 0.99 mm in diameter. Weakfish larvae emerge after about 40 h at water temperatures of 68° to 70° F and aver- age 1.8 mm SL. Soon after hatching, the demersal larvae disperse into the nursery areas. Through- out the coastal waters from North Carolina to at least New York, anomalous water conditions (such as rapid changes in salinity, temperature, or dissolved oxygen) may occur in small areas due to local weather phenomena or industrial- domestic development. Hurricanes, however, may affect the entire eastern seaboard (tropical storm Agnes — 1972) or portions of it (Hurricane Camille — 1969) with the greatest impact occur- ring in the estuarine areas (i.e., weakfish nur- sery). The extended spawning season of weakfish would tend to minimize any effect of a short-term calamity upon a local population. Tolerance of weakfish eggs and larvae to tem- perature, salinity, dissolved oxygen, etc., remains poorly known. According to data compiled by Harmic (1958), natural fluctuations in the es- tuary approach the ranges that are detrimental to weakfish survival. For Delaware Bay and pre- sumably throughout its range, the variation in water parameters due to natural phenomena alone may largely explain fluctuations in the weakfish population abundance and year class strength. LITERATURE CITED Bagenal, T. B. 1967. A short review of fish fecundity. In S. D. Gerking (editor), The biological basis of freshwater fish produc- tion, p. 89-111. John Wiley and Sons Inc., N.Y. BIGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53, 577 p. BRowTj, J., AND E. McCoy. 1969. A review of the North Carolina scrap fishery. N.C. Dep. Conserv. Dev., Div., Conuner. Sport Fish., Mimeo. Rep., 13 p. Daiber, F. C. 1954. Fisheries research program. Mar. Lab. Dep. Biol. Sci., Univ. Del. Biennial Rep. 1953 and 1954. Publ. 2:50-64. fahy, W. E. 1965a. Report of trash-fish study in North Carolina in 1962. Div. Commer. Fish., N.C. Dep. Conserv. Dev., Spec. Sci. Rep. 5, Mimeo., 20 p. 1965b. Report of trash-fish study in North Carolina in 1964. Div. Commer. Fish., N.C. Dep. Conserv. Dev., Spec. Sci. Rep. 7, Mimeo., 13 p. Harmic, J. L. 1958. Some aspects of the development and ecology of the pelagic phase of the gray squeteague, Cynoscion regalis (Bloch and Schneider), in the Delaware estuary. Thesis, Univ. Delaware, Newark, 84 p. HIGGINS, E., AND J. C. PEARSON. 1928. Examination of the summer fisheries of Pamlico and Core sounds, N.C, with special reference to the de- struction of undersized fish and the protection of the gray trout Cynoscion regalis (Bloch and Schneider). Rep. U.S. Comm. Fish., 1927 append. 2:29-65. HILDEBRAND, S. F., AND L. E. CABLE. 1934. Reproduction and development of whitings or king- fishes, drums, spot, croaker, and weakfishes or sea- trouts, family Sciaenidae, of the Atlantic Coast of the United States. U.S. Bur. Fish., Bull. 48:41-117. HILDEBRAND, S. F., AND W. C. SCHROEDER. 1927. Fishes of Chesapeake Bay. U.S. Bur. Fish., Bull. 43:1-366. JOSEPH, E. B. 1972. The status of the sciaenid stocks of the middle Atlantic Coast. Chesapeake Sci. 13:87-100. KESTEVEN, G. L. (editor). 1960. Manual of field methods in fisheries biology. FAO Man. Fish. Sci. 1, 152 p. MCHUGH, J. L. 1960. The pound-net fishery in Virginia. Part 2 - Species composition of landings reported as menhaden. Com- mer. Fish. Rev. 22(2):1-16. MASSMANN, W. H. 1963. Age and size composition of weakfish, Cynoscion regalis, from pound nets in Chesapeake Bay, Virginia 1954-1958. Chesapeake Sci. 4:43-51. MASSMANN, W. H., J. P. WHITCOMB, AND A. L. PACHECO. 1958. Distribution and abundance of gray weakfish in the 25 FISHERY BULLETIN: VOL. 74, NO. 1 York River system, Virginia. Trans. 22nd North Aro. Wildl. Conf. , p. 361-369. MERRINER, J. V. 1973. Assessment of the weakfish resource, a suggested management plan, and aspects of life history in North Carolina. Ph.D. Thesis, North Carolina State Univ., Raleigh, 201 p. NESBIT, r. a. 1954. Weakfish migration in relation to its conservation. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 115, 81 p. PARR, A. E. 1933. A geographical-ecological analysis of the seasonal changes in temperature conditions in shallow water along the Atlantic Coast of the United States. Bull. Bingham Oceanogr. Collect. Yale Univ. 4(3): 1-90. PEARSON, J. C. 1941. The young of some marine fishes taken in lower Chesapeake Bay, Virginia, with special reference to the gray sea trout, Cynoscion regalis (Bloch). U.S. Bur. Fish., Bull. 50:79-102. Perlmutter, a. 1959. Changes in the populations of fishes and in their fisheries in the Middle Atlantic and Chesapeake regions, 1930 to 1955. Trans. N.Y. Acad. Sci., Ser. H, 21:484- 496. Perlmutter, a., W. S. miller, AND J. C. Poole. 1956. The weakfish {Cynoscion regalis) in New York waters. N.Y. Fish Game J. 3:1-43. ROELOFS, E. W. 1951. The edible finfishes of North Carolina. In H. F. Taylor (editor), Survey of marine fisheries of North Carolina, p. 109-139. Univ. N.C. Press, Chapel Hill. Steel, R. G. d., and J. H. Torrie. I960. Principles and procedures of statistics with special reference to the Biological Sciences. McGraw-Hill Book Co., N.Y., 481 p. Taylor, H. F. 1916. The structure and growth of the scales of the squeteague and the pigfish as indicative of life history. U.S. Bur. Fish., Bull. 34:285-330. THOMAS, D. L. 1971. The early life history and ecology of six species of drum (Sciaenidae) in the lower Delaware River, a brack- ish tidal estuary. Ichthyol. Assoc. Bull. 3. An ecological study of the Delaware River in the vicinity of artificial island. Delaware Progress Report for the period January- December 1970. Part III, 247 p. Welsh, w. W., and C. M. Breder, Jr. 1923. Contributions to life histories of Sciaenidae of the eastern United States coast. U.S. Bur. Fish., Bull. 39:141-201. 26 DDT AND ITS METABOLITES IN THE SEDIMENTS OFF SOUTHERN CALIFORNIA John S. MacGregori ABSTRACT To assess the degree of DDT contamination in the marine sediments off Los Angeles, 103 stations in the Pacific Ocean off southern Cahfomia were sampled in July and August 1971 for DDT and its metabolites, DDD and DDE. Heavy contamination of bottom sediments in this area was expected because of large amounts of DDT that have entered the ocean through the Los Angeles County sewer system as waste from a DDT manufacturing plant. From the data acquired, it was estimated that there were about 200 metric tons of DDT, DDD, and DDE in the sediments in an area of 14 square nautical miles near the sewer outfalls and 300 metric tons in the entire 911 square nautical mile area sampled. The heaviest concentrations of total DDT were distributed in the relatively shallow- water area on the Palos Verdes shelf to the northwest of the sewer outfalls in the general direction of the current flow. Metabolism of DDT was inhibited in deepwater sediments. Ratios of DDE to DDT were low, and DDT was more abundant than DDE at some stations. In sediments from shallow- water stations, DDE exceeded DDT by more than 10 times. The bottom of the ocean off Los Angeles, CaHf, has been very heavily contaminated with the pesticide DDT owing to the discharge of wastes from a DDT manufacturing plant into the Los Angeles County sewer system over a period of about 20 yr ending in 1970 (MacGregor 1974). The amount of DDT which entered the ocean through the Los Angeles County sewer system was estimated at 250 kg/day. Following the ces- sation of DDT discharges by the manufacturer, the amount entering the ocean dropped to 45 kg/day in December 1970 and to 11 kg/day in October 1971. Most of these later discharges re- sulted from sewer cleaning operations which stirred up old deposits of DDT in the sewer lines. The discharges resulting from the cleaning oper- ations were primarily DDD and DDE, metabo- lites of DDT, while the earlier discharges were primarily DDT. Because there has been a great deal of specula- tion about the fate of DDT and other toxic chem- icals released into the environment by man (Woodwell et. al. 1971; National Academy of Sciences 1971), this investigation was under- taken to determine the areal distribution and fate of these chemicals in the bottom sediments in the ocean off Los Angeles. ^Southwest Fisheries Center, La Jolla Laboratory, National Marine Fisheries Service, NOAA, La Jolla, CA 92038. MATERIALS AND METHODS The bottom sediments were sampled from a grid of 103 stations between lat. 33°30' and 33°58'N and long. 118°00' and 118°44'W (Figure 1). The stations were designated by four-digit numbers, the first two indicating minutes north Manuscript accepted June 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 1976. Figure l. — Distribution of total DDT in milligrams per square meter of bottom in the sediments of southern California. Total DDT ranged from 6,600 mg/m^ of bottom at station 43-22 to 0.12 mg/m^ at station 30-08. 27 FISHERY BULLETIN: VOL. 74, NO. 1 of lat. 33°N and the second two indicating min- utes west of long. 118°W. The samples were taken aboard the National Marine Fisheries Service RV David Starr Jordan between 26 July and 3 August 1971. The Shipek bottom sampler was used to obtain the samples of sediment. This device obtains a block of material equal to 400 cm^ of bottom sediment to a depth of about 10 cm, or slightly more, in soft mud or to a depth of half as much or less in coarse sand. Two samples were taken at each station in order to obtain an estimate of sampling error. The vessel was allowed to drift while the samples were being taken, so the sample pairs were taken in only approximately the same location. How- ever, agreement in the various parameters be- tween samples from the same station was good. The samples were placed in aluminum foil- lined containers of approximately the same size as the sampling bucket and were quick-frozen. They were stored in a freezer until removed for analysis. In most samples, DDT was confined to the top 2 or 3 cm of the sediment. At most of the stations where the sampler sampled to 10 cm, and at all of the stations where it sampled to a lesser depth, it appeared that all of the DDT under the 400 cm^ had been sampled. In this study, therefore, DDT concentrations are given as the weight of DDT per unit area of bottom to a depth of 10 cm. In a few areas of rapid sedimentation, where the sam- pler sampled to about 10 cm depth, there were still significant amounts of DDT below 10 cm. Estimates for the amounts of DDT below 10 cm are based on core samples taken by other investi- gators in this area. The bottom sediment samples were thawed and blended in a 1-gallon Waring^ commercial blender. Before blending, small stones were re- moved from the few samples that contained them. Some samples contained a few small mol- luscs or brittle stars, but these were not removed. Measured amounts of distilled water were added to some of the drier (sandy) samples to facilitate blending. A sample of 15 to 20 g of blended sediment was weighed onto a watch glass, dried to constant weight, and reweighed to obtain percent water in the sediment. This gave an index of bottom type ^Reference to trade name does not imply endorsement by the National Marine Fisheries Service, NOAA. ranging from 30 to 40% water for coarse sand to 60 to 70% water for fine silt. A second sample weighing about 30 g was weighed into a 1-pint Mason jar for DDT deter- mination. About four or five times as much Na2S04 was weighed into the jar as a drying agent. The sediment and Na2S04 were mixed using a stainless steel spatula, and the mixture was frozen. A cutting assembly was fitted to the jar, and the frozen mixture was thoroughly blended to a powder using an Osterizer blender. About 5 g of the powder was weighed into a tared, large disposable pipet (Matheson super pipet) plugged with glass wool. The powder was extracted into a 15-ml graduated centrifuge tube with 5 ml of hexane and 5 ml of acetone. The extract was evaporated to dryness and redis- solved in 1 ml of hexane. This sample was eluted through a super pipet filled with activated alu- mina (McClure 1972) using enough hexane to obtain a 6-ml sample. This sample was reduced or increased in vol- ume as required and injected into a model 402 Hewlett Packard gas chromatograph (GLC) with a Ni^^ electron capture detector. The 6-foot glass column contained 4% SE-30/6% QF-1 on 100/120 mesh Supelcoport. There was evidence of a polychlorinated bi- phenyl, Aroclor 1254, in all samples, but the DDT peaks were so dominant in the chromato- grams that they generally obliterated any traces of other chlorinated hydrocarbons within their range. Only the six peaks representing the ortho- para and para-para forms of DDE, DDD, and DDT were quantified. "Total DDT" is used to designate the sum of these six analogs. RESULTS AND DISCUSSION Fifty-five correlations were obtained for 11 parameters to determine various DDT relation- ships. The 55 correlations were obtained for all 103 stations (Table 1, values above 1.000 correla- tion diagonal) and for 76 stations leaving out those 27 stations having total DDT readings greater than 100 mg/m^ (Table 1, values below 1.000 correlation diagonal). For 100 observations a correlation coefficient of 0.254 indicates a probability of 0.01. Logarithms were used for total DDT and distance from outfall, arithmetic values for the other nine measurements. There is a very high negative correlation be- tween log total DDT and log distance from the 28 MacGREGOR: DDT OFF SOUTHERN CALIFORNIA Table l. — Correlation coefficients for 11 parameters relating to DDT and its metabolites in bottom sediments off southern California. Values above 1.000 correlation diagonal are for 103 stations. Values below diagonal are for 76 stations leaving out those 27 stations having total DDT readings greater than 100 mg/m^. For 100 observations, a correlation coefficient of 0.254 indicates a probability of 0.01. Log total Log distance from Sample % H2O in p,p DDD p,p' DDE p,p' DDE o,p' DDE o,p' DDD o,p' DDT Parameter DDT outfall Depth weight sample p,p' DDT p,p' DDT p.pDDD P.P'DDE p,p' DDD p,p' DDT Log total DDT 1.000 -0.871 -0.253 0.221 0.157 0.142 0.144 0.281 -0,040 -0,272 -0,334 Log distance from outfall -0.604 1.000 0.228 -0.095 -0.147 -0.032 -0.043 -0.238 0,016 0.332 0,334 Depth 0.078 -0.036 1.000 0.643 0.771 -0.443 -0.512 -0.572 0,315 0,168 0.095 Sample weight 0.245 0.041 0.761 1.000 0.743 -0.265 -0.341 -0.332 0.190 0,036 -0 151 % H2O in sample 0.095 0.002 0.921 0.756 1.000 -0.315 -0.390 -0,396 0.228 0,019 -0.013 p.p'DDD/p.p'DDT 0.123 0.066 -0,418 -0.268 -0.325 1.000 0.909 0,297 -0.168 -0.194 0.556 p,pDDE/p,pDDT 0.100 0.040 -0.492 -0.366 -0.416 0.907 1.000 0,446 -0,208 -0.171 0,463 p,p 'DDE/p,p'DDD 0.060 -0.075 -0.535 -0.418 -0.487 0.268 0.408 1.000 -0,289 -0,043 -0,011 o,p'DDE/p,pDDE 0.190 -0.234 0.296 0.225 0.248 -0.151 -0.196 -0.273 1,000 0.021 0,113 o,p 'DDD/p,pDDD -0.129 0.275 0.106 0090 0.059 -0.177 -0.151 0.023 -0.024 1.000 0,130 o,p'DDT/p,p'DDT -0.154 0.167 0.006 -0.127 0.025 0.630 0.537 0,084 0.068 0.060 1,000 Los Angeles County sewer outfalls (r = —0.871). Values ranged from 6,600 mg of total DDT/m^ of bottom near the sewer outfalls to about 1 mg/m^ at more distant stations. The distribution of DDT was modified some- what by currents which tended to deposit the DDT along the coast and to the northwest more than to the east (Figure 1). The apparent relation between total DDT and depth results from the fact that the sewers discharge into relatively shallow coastal waters and the sludge tends to remain there. The deeper waters are merely farther from the sewer outfalls and the areas along the coast favored by the currents. McDermott et. al. (1974) took sediment sam- ples from the Palos Verdes shallow-water shelf area in the vicinity of the sewer outfalls only. Their tables A-1 and A-4 give total DDT in parts per million dry weight from gravity core samples taken in 1972. I have contoured their data (Figure 2B) for the top 10 cm of sediment to compare with the 1971 data (Figure 2A) which has been converted to parts per million dry weight. Their 1973 data in their table A-5 repre- sents parts per million dry weight of total DDT in the top 5 cm of Shipek samples taken in the same area (Figure 2C). In each of the 3 yr the patch of sediment representing more than 100 ppm. total DDT tends to retain its integrity fairly well as an oblong area stretching to the northwest of the sewer outfalls. The contours representing 10 to 100 ppm. seem to be expanding somewhat to the northwest and in 1973 to the southeast also. At the Los Angeles County sewage disposal plant, most of the solids are removed by centri- fuging, but the supernatant is pumped into the ocean along with the water from the settling tanks. This reduces the amount of particulate matter being discharged into the ocean. Never- theless, quantities of relatively DDT-free partic- ulate matter have been deposited on the Palos Verdes shelf since dumping of DDT into the sew- er system was stopped. In time this could cause a change in the DDT profile of the sediments. On the other hand, most of the shallow inshore areas along this section of coast tend to have sandy bottoms, and the silt bottoms in the vicin- ity of the sewer outfalls would appear to be unstable artifacts. Storms, tides, and currents could remove or deposit layers of bottom silts in this shallow-water area and further change the DDT profiles. Based on the paired samples taken in 1971, the variation within a sampling area for one sample would be roughly plus 100% minus 50% at a one standard deviation level. For an average for two samples it would be plus 70% minus 40%. This could account for the differences in the distribu- ion of total DDT for the 3 yr. However, the similarities are much more striking than the differences. High sample weight and high water content both indicate samples containing more silt, while lower weights and lower water content indicate samples containing more sand. Both of these measurements are related to depth, with the bot- tom in deep basins tending to be fine silt while shallow areas tend to be sandy. This tendency is masked in shallow-water areas where there are sewer outfalls which deposit large quantities of fine material in the shallower waters. This is in- dicated by the improvement of the correlation coefficients from 0.643 to 0.761 for sample weight and from 0.771 to 0.921 for percent water with 29 FISHERY BULLETIN: VOL. 74, NO. 1 33° 40 B 33° 40' FIGURE 2.— Total DDT (parts per million dry weight) in the bottom sediments off Palos Verdes in the vicinity of the Los Angeles County sewer out- falls. A. Shipek samples (pres- ent paper); B. top 10 cm of gravity cores (McDermott et al. 1974); C. top 5 cm of Shipek samples (McDermott et al. 1974). •n8° 20- 118° 16- 30 MacGREGOR: DDT OFF SOUTHERN CALIFORNIA depth when the 27 stations of heavy sewer deposi- tion in shallower waters are omitted. The very high correlation coefficient (0.909) be- tween p,p'DDD/p,p'DDT and p,p'DDE/p,p'DDT shows that when metabolism of DDT to DDD is high, metabolism of DDT to DDE is high also. These high rates of metabolism are negatively correlated with depth. Actually, they are more probably associated with some of the conditions prevailing at depth in the ocean off Los Angeles. The deep areas sampled tend to be anaerobic, and it is probably the lack of oxygen and colder water that determines the low rate of metabolism. The high correlations of the ratios with sample weight and percent water are secondary effects of the correlations of these two factors with depth. The high negative correlation between p,p'DDE/p,p'DDD and depth indicates that metabolism of DDT to DDE is favored over me- tabolism to DDD in shallower waters. However, the positive correlation ofp,p ' DDE /p,p' DDD with p,p'DDD/p,p'DDT (0.297) as well as with p,p'DDE/p,p'DDT (0.446) supports the conclusion that metabolism to both metabolites is much greater in shallow aerobic waters than in deep anaerobic waters. Actually, much more DDT is probably metabolized to DDD than to DDE under all circumstances prevailing in the study area, but the DDE is much more persistent than the DDD and accumulates to a greater degree while DDD is further metabolized to DDMU and other metabolites. There was at least 10 times as much DDE as DDT in the bottom sediments from stations along the coast of the study area, while 10 stations in deeper waters north of Santa Catalina Island had less DDE than DDT in the bottom samples (Fig- ures 3, 4). DDD tended to follow somewhat the same pattern (Figure 5). At the 10 stations the average total DDT was 19.9 mg/m2, of which 60% was DDT, 19% DDD, and 2V/c DDE. Mean depth was 341 fathoms (623 m) and the total area represented by the 10 sta- tions was 111 sq nautical miles containing an es- timated 5.74 metric tons of total DDT. It appears that most of the pesticide discharged from the Los Angeles County sewer outfalls has been DDT with the exception of the period of sewer cleaning operations in 1970-71 when DDD and DDE predominated (MacGregor 1974). Most of the DDT settles on the bottom close to the out- falls in shallow waters. Once the DDT becomes part of the bottom sediment it tends to stay there Figure 3.— Distribution of ratios ofp.p'DDE top,p'DDT. In the shallow waters near shore the ratios exceed 10:1, while in the deeper waters north of Santa Catalina Island the ratios are less than 1:1. B Figure 4. — Chromatograms of: A — a deepwater sample 1274 m), station 30-40, showing highp,p'DDT peak, and D— a shallow-water sample (36.5 m), station 40-16, showing a high DDE peak. B and C are standards of the DDT analogs. and metabolize in place, rapidly in shallower waters and more slowly in deeper waters. 31 FISHERY BULLETIN: VOL. 74, NO. 1 Figure 5. —Distribution of ratios ofp,p'DDD to p,p' DDT. In the shallow nearshore areas the ratios exceed 2:1, while in the deeper waters the ratios are less than 1:1. The higher ratios were probably enhanced by sewer cleaning operations in 1970-71. The DDT deposits in the deeper waters must have been transported there directly from the sewer outlets before much metabolism could take place. If they had originated from bottom sedi- ments closer to the sewer outfalls and in shal- lower waters, the DDE content would be much higher. DDE averages about 85% of total DDT in biological material in this area; therefore, most of the total DDT in bottom sediments in the deeper water could not have originated from this source. For the time series for total DDT accumulation in myctophid fish (MacGregor 1974), DDE was less than DDT from 1949 to 1956, but in the subsequent years DDE became much higher. If the deep water with relatively low DDE had re- sulted from biological fallout as represented by the myctophids for 22 yr (1949-70), and if there had been no metabolism at depth, the DDE would have been twice as high as the DDT rather than one-third as high. There is either very little metabolism in deep- water sediments, or there is no metabolism, and the small amounts of DDD and DDE found there are the result of fallout from material metabo- lized in the better-oxygenated surface and inter- mediate depths. In commercial DDT, the ratio of p,p' DDT to o,p'DDT is about 4:1 (i.e., o,p' DDT is about 25% ofp,p'DDT). The distribution of these latter val- ues for the sediment samples indicate that o,p 'DDT is about what might be expected, while o,p' DDD is higher and o,p' DDE is lower (Table 2). In the case of DDT this may mean that o,p ' DDT metabolizes as readily as p,p' DDT. The two high positive correlations with the parameters indicat- ing high metabolism, p,p'DDD/p,p'DDT and p,p'DDE/p,p'DDT, may indicate that o,p' DDT metabolizes more readily than p,p' DDT under conditions of low metabolism of DDT to DDD and DDE. Both the ratios of o,p' DDT to p,p' DDT and o,p ' DDD to p,p ' DDD tend to be high in the bottom sediments north of Santa Catalina Island and in Santa Monica Bay, while ratios tend to be low just south of Palos Verdes Peninsula and in the sandy shallower waters to the east of this area (Figures 6, 7). The association of greater distance from the sewer outfalls and lower total DDT values with high ratios is undoubtedly fortuitous, although the few very high ratios are associated with very low DDT values and probably result from poorer resulting measurements and interfering sub- stances that are no longer completely dominated by DDT at these very low values. The ratios of o,p' DDE to p,p' DDE are greater than 1.00:1.00 for 19 stations. Unlike the other two ratios these high ratios are associated with depth. They also tend to be concentrated in the deeper waters just off the Palos Verdes shelf where the sewer outfalls are located (Figure 8). These apparent high relative values of o,p' DDE are probably caused by interfering substances, probably DDMU, a metabolite of DDD, which is not being further metabolized under the condi- tions prevailing at these stations. Table 2. — Frequency distributions of ortho-para isomer as a percent of para-para isomer of DDT, DDD, and DDE in bottom sediments. Percent DDT DDD DDE 0.0- 5.0 4 1 5.1- 10.0 5 2 10 1- 15.0 15 11 15.1- 20.0 13 7 20 20.1- 25.0 10 8 19 25.1- 30.0 12 16 10 30.1- 35.0 11 24 6 35.1- 40.0 7 16 3 40.1- 45.0 6 9 3 45.1- 50.0 6 9 3 50.1- 55.0 1 1 3 55.1- 60.0 4 4 60.1- 65.0 1 1 1 65.1- 70.0 1 70.1- 75.0 1 1 75.1- 80.0 2 80.1- 85.0 2 85.1- 90.0 1 90.1- 95.0 1 1 1 95.1- 100.0 2 >100 1 4 19 32 MacGREGOR: DDT OFF SOUTHERN CALIFORNIA ►'«•.. «,.«•'•• O o o o FIGURE 6.— Stations at which the ratio of o,p' DDT top.p'DDT was greater than 0.40:1.00. Figure 7. — Stations at which the ratio of o,p' DDD to p,p 'DDD was greater than 0.40:1.00. To estimate the amount of DDT stored in the bottom sediments in the approximately 911 sq nautical miles between long. 117°58' and 118°46'W and lat. 33°18'N and the California coast, represented by the 103 stations, we must assume that each station is representative of its surrounding area. Each pair of samples from each station showed a high correlation for all pairs of parameters. The correlation coefficient for the logarithms of total DDT for paired samples from 94 stations from which two samples were ob- tained was 0.964 and the standard error of esti- mate ±0.321. The Shipek sampler took bottom silts only to a depth of about 10 cm and sandy bottoms or shal- low sediment deposits to a lesser depth. At all stations except those where bottom deposition was very rapid, as near sewer outfalls, all DDT in the sediments was sampled. Near the sewer out- falls the sample represents only DDT deposits in the top 10 cm of sediment. The total amount of DDT determined for the 911 sq nautical mile sampling area was 217 metric tons in the top 10 cm of bottom sediment. Of this total, 179 metric tons (82%) was DDE, 22 metric tons (10%) was DDD, and 16 metric tons (8%) was DDT. McDer- mott and Heesen (1974) found that the total DDT in the top 5 cm of sediment consisted of 86% DDE, 11% DDD, and 3% DDT in the area of the Palos Verdes shelf These somewhat different percent- ages may have resulted from further metabolism of DDT without replenishment. In addition, the DDE percentages tend to be higher in this area, and the DDD was increased in 1970-71 because of sewer cleaning operations. The total DDT ranged from an estimated 0.42 kg per sq nautical mile at station 30-08 repre- senting 13.3 sq nautical miles to 28.6 metric tons per sq nautical mile at station 43-22 representing 1.25 sq nautical miles. Five stations representing 6.24 sq nautical miles or 0.7% of the 911 sq nautical mile area represented by the 103 stations contained 47.3% Figure 8.— Stations at which the ratio of o,p' DDE to p,p' DDE was greater than 1.00:1.00. The high apparent o.p'DDE values probably were caused by DDMU which has the same retention time as o,p 'DDE on the column used. 33 FISHERY BULLETIN: VOL. 74, NO. 1 of the total DDT (102.7 metric tons). Sixteen sta- tions representing 18.1 sq nautical miles (2.0% of total area) contained 64.0% (193 metric tons) of the total DDT. Subsamples taken from the tops and bottoms of the blocks of sediment obtained with the Shipek sampler indicated that most of the pesticide was concentrated in the top strata of the samples ex- cept for samples taken in the vicinity of the sewer outfalls where deposition was very rapid. Cores were taken from one sample taken near the sewer outfalls and from a second taken at a greater dis- tance from the outfalls to determine more about vertical distribution of DDT in the sediments (Table 3). At station 42-36 only p,p' DDE was measured because DDT and DDD were not readily measur- able in the deeper sediment sections. Half of the DDE was found in the top 2 cm, 81% in the top 4 cm, and 95% in the top 6 cm. At station 42-20, close to the sewer outfall where sewer sediment deposition was heavy, there was very little change in the chlorinated hydrocarbon concen- trations at all five depths. Vance McClure (pers. commun.) has provided me with a plot of the depth distribution of DDT, DDE, DDD, and DDMU found in a box core sam- ple taken about 1 nautical mile west-northwest of the sewer outfall. Subsamples were taken from the core at 3-cm intervals from to 12 cm and at 6-cm intervals from 12 to 36 cm. The pesticide values remained high through 12 cm depth and dropped off rapidly between 12 and 18 cm. DDMU had a deeper distribution than the other three components and increased to a maximum at 9 cm and was still present at 36 cm. DDE was last measured at 24 cm, and DDD and DDT at 18 cm. Excluding DDMU, 72% of the pesticide was found in the column corresponding to the top 10 cm and 28% below that depth. Including DDMU, 67% was in the top 10 cm and 33% below. If the box core sample is typical of the stations near the sewer undergoing rapid sedimentation, about 30% of the pesticide was missed by sam- pling only to a depth of 10 cm at these stations. Because these stations near the sewer outfalls contain most of the pesticide, the 217 metric tons of pesticide estimated for the entire area in the top 10 cm could be increased to roughly 300 met- ric tons as a maximum estimate of total DDT in the area. In the area of the Palos Verdes shelf only, McDermott and Heesen (1974) estimated that Table 3. — Vertical distribution of DDT in the sediments as determined from core samples taken at stations 42-20 and 42-36. Stn. 42-20 Sfn. 42-36 Core o,p' DDD o,p' DDT Aroclor depth o.p' DDE p.p'DDE p.p'DDD p,p' DDT 1254 p.p'DDE (cm) (ppm.) (ppm.) (ppm.) (ppm.) (ppm.) (ppm.) 0-2 14.6 67.4 10.1 3.1 6.2 0.0233 2-4 19.4 90.7 11.4 3.5 6.0 0.0149 4-6 16.2 84.2 10.1 2.9 4.8 0.0063 6-8 348 64.4 13.2 3.5 6.2 0.00180 8-10 34.0 79.1 8.2 2.8 4.8 0.00065 there were 218 tons of total DDT under 62 km^ of bottom. They calculated that 85% of the total DDT was in the top 12 cm of sediment. If the pesticide is fairly equally distributed in the top 12 cm, about 14% would be in the 10- to 12-cm layer, and the Shipek sampler would sample about 71% of the total DDT. Sixteen contiguous stations on the Palos Ver- des shelf sampled by us in 1971 represented an area of 18.1 sq nautical miles (62.0 km^) and a total DDT load of 139 metric tons. If this was only 71% of the total DDT in the area (the load of the top 10 cm only), then the corrected esti- mate including DDT below 10 cm would be 196 metric tons. McDermott et al. (1974) using a reduced sam- pling area of 48 km^ determined that there were 156 tons of total DDT in their revised sampling area. In this present study the area can be ad- justed to 48 km^ by omitting the effect of ^Vi peripheral stations. Estimated total DDT then would be 132 metric tons. However, McDermott et al. (1974, table 5) give estimates of total DDT in the area in 2-cm increments down to a depth of 30 cm of sediments. This table indicates that only about 59% of the total DDT is in the top 10 cm in this area. This would increase my estimate of total DDT to 224 metric tons for the 48 km^ area. The available data indicate that there is con- siderable variation in the depth distribution of total DDT in the sediments on the Palos Verdes shelf However, the general conclusion that can be drawn from the samples is that there are about 200 metric tons of total DDT in the bottom sedi- ments in the 14 sq nautical mile area (48 km^) in the vicinity of the sewer outfalls and another 100 metric tons in the remaining 897 sq nautical miles of the 1971 survey area. On 27-28 June 1972, 11 mo after the first sam- ples were taken, additional samples were ob- tained from seven of the original stations. Four of these stations were in deeper water, between 600 34 MacGREGOR: DDT OFF SOUTHERN CALIFORNIA and 890 m deep, and 5 to 11 nautical miles from the sewer outfalls. Total DDT remained low in these stations averaging about 30 mg/m^ of bottom, and the composition was essentially unchanged. The remaining three stations, in areas of much higher pollution within 1.3 nautical miles of the sewer outfalls and in shallower water, showed some apparent changes in grams per square meter of bottom (Table 4). Table 4. — Changes in composition (in grams per square meter of bottom) at stations 42-21, 43-21, and 42-19 in 11 mo. Station Depth o,p DDE o.p'DDD o.pDDT Total year (m) DDMU p,p' DDE p.p'DDD p,p' DDT DDT 42-21 1971 119 0.54 3.45 0.53 0.19 4.71 1972 2.09 3.36 0.61 0.23 6.29 43-21 1971 33 0.46 1.80 0.33 0.14 2.73 1972 0.99 0.80 0.11 0.05 1.95 42-19 1971 37 0.38 1.78 0.32 0.14 2.62 1972 0.82 0.71 0.16 0.13 1.82 Totals 1971 1.38 7.03 1.18 0.47 10.06 1972 3.90 4.87 0.88 0.41 10.06 At station 42-21, DDT, DDD, andp,p'DDE re- mained relatively unchanged with a total of 4.2 g/m^ of bottom in both years, while the o,p'DDE-DDMU peak increased by almost four times. At the two shallower stations, 43-21 and 42-19, DDT, DDD, and p,p'DDE decreased in 1972 to less than half its value in 1971, while the o,p'DDE-DDMU peak more than doubled. These changes could be caused by metabolism, by the addition of sewage deposits that were relatively free of DDT combined with metabolism, or even by the removal of a few centimeters of the deposits in the shallow-water areas without metabolism. CONCLUSIONS Total DDT in the bottom sediments in the ocean off southern California in an area of 911 sq nautical miles was estimated to be between 200 and 300 metric tons. Most of the total DDT was concentrated in a relatively small area with- in a few miles of the Los Angeles County sewer outfalls. Total DDT in the top 10 cm of sediment ranged from 6,600 mg/m^ of bottom near the sewer out- falls to about 1 mg/m^ of bottom at the more dis- tant stations. Eighty-two percent of the total DDT was DDE; 10%, DDD; and 8%, DDT. Metabolism of DDT to DDD and DDE was more rapid in shallow waters and apparently very slow or lacking in deep, cold waters that were low in oxygen. Seven samples taken 11 mo later tended to confirm these findings. ACKNOWLEDGMENTS I am indebted to W. Rommel, G. Boehlert, and V. McClure for their advice and help in process- ing the samples; to G. Stauffer for programming the data for the computer; to R. Lasker for valu- able criticism and guidance; to the personnel of the RV David Starr Jordan for their cooperation, assistance, and interest; and to K. Raymond for preparing the figures. This work was supported in part by NOAA, Office of Sea Grant, under grant #UCSD 2-35208 with the Institute of Marine Re- sources, University of California. LITERATURE CITED MacGregor, J. S. 1974. Changes in the amount and proportions of DDT and its metaboHtes, DDE and DDD, in the marine en- vironment off southern California, 1949-72. Fish. Bull., U.S. 72:275-293. McCLURE, V. E. 1972. Precisely deactivated adsorbents applied to the separation of chlorinated hydrocarbons. J. Chro- matogr. 70:168-170. MCDERMOTT, D. J., AND T. C. HEESEN. 1974. Inventory of DDT in sediments. Annual report for the year ended 30 June 1974. Southern California Coastal Water Research Project, p. 123-127. MCDERMOTT, D. J., T. C. HEESEN, AND D. R. YOUNG. 1974. DDT in bottom sediments around five southern California outfall systems. TM 217. Southern California Coastal Water Research Project, 54 p. NATIONAL ACADEMY OF SCIENCES. 1971. Chlorinated hydrocarbons in the marine environ- ment. Wash., D.C., 42 p. WooDWELL, G. M., P. P. Craig, and H. A. Johnson. 1971. DDT in the biosphere: Where does it go? Science (Wash., D.C.) 174:1101-1107. 35 AN ENERGETICS MODEL FOR THE EXPLOITED YELLOWFIN TUNA, THUNNUS ALB AC ARES, POPULATION IN THE EASTERN PACIFIC OCEAN Gary D. Sharp and Robert C. Francis^ ABSTRACT An energetics model (ENSIM) for the exploited yellowfin tuna, Thunnus albacares, population in the eastern Pacific Ocean is developed. Hydrodynamic properties and respiration-swimming work theory are combined to describe the energy expenditure due to swimming as a function of length for tunas. Growth and maintenance energetics are estimated and incorporated into a simplistic three process model. This model is interfaced with a population simulator (TUNP0P) and minimal energy requirements for the exploited yellowfin tuna population are derived for the simulated fishing years 1964-72. A theoretical unexploited population simulation is made, and the energy requirements by this population are compared with primary productivity rates and minimum micronekton (forage) standing stock availability. No obvious food limitation is indicated for yellowfin tunas greater than 40 cm, particularly since the exploited population is at a level of, at most, 50% of the unexploited biomass estimates. Population limitation processes are examined and indications that the recruit- ment rates are independent of exploited biomass are discussed. The intent of studies of the population dynamics of exploited populations is the determination of the numbers, biomass, age structure, and poten- tial yield from a population in order that rational management decisions can be made about the manner and rate of exploitation in order to insure efficient utilization of the resource. The validity of the resulting estimates of numbers, biomass, and potential yield is of concern to all those involved with the resource. Underestimations generally result in conservative efforts which are "safe" but not necessarily efficient. Overestima- tions can result in reduced profit margins or, in the extreme case, decimation of the resource. Since the implementation of the program for conservation of yellowfin tuna, Thunnus alba- cares, in the eastern tropical Pacific in 1966, a series of complex changes in the fishery have occurred which make production model results less and less comparable between years (Inter- American Tropical Tuna Commission Annual Re- ports). Attempts to account for multiple changes in the effort variables and corresponding but independent changes in the exploited population have resulted in serious interpretation problems as to the relative status of the exploited stock. 'Inter-American Tropical Tuna Commission, c/o Scripps In- stitution of Oceanography, La Jolla, CA 92037. The economic and temporal problems inherent in the collection and analysis of biological data and the difficulties in representation of the biological processes in a useful mathematical manner has served to hinder utilization in the management procedures of what sparse physi- ological and ecological information is available. In this report, an energy budget model is de- veloped for the exploited yellowfin tuna popula- tion in the eastern Pacific Ocean within the Inter- American Tropical Tuna Commission's Yel- lowfin Regulatory Area (CYRA). The model will be used to assess the energy flow through the exploited yellowfin tuna population and also to compare the estimated utilization of energy by yellowfin tuna with the estimated primary pro- ductivity in the CYRA. Comparisons will be made using simulations of the population under both exploited and unexploited conditions. The energy budget estimates are interfaced with an age dependent population simulation model (TUNP0P) (Francis 1974) resulting in a model of the energy utilization by semiannual recruitment cohorts. This model is referred to as ENSIM. The model incorporates the population parameter estimates and variables of TUNP0P and the empirical and estimated size dependent relationships for the major energy consuming processes, resulting in estimates of energy utili- zation rates. The development of the empirical Manuscript accepted June 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 1976. 36 SHARP and FRANCIS: ENERGETICS MODEL FOR YELLOWFIN TUNA POPULATION relationships and the resulting formulations are presented so as to encourage research in the area so that improvements on this crude model can be made in the future. THE MODEL Population Dynamics In an attempt to produce a new, more detailed method for evaluating the population or stock status it was decided that the development of TUNP0P, a biologically oriented population simulator, would be appropriate. The only avail- able population data which are collected on a routine basis from within the fishery are length- frequency information from commercial catches. These data are collected according to criteria which require that the several time-area strata be sampled regularly and multiply, whenever possible (Hennemuth 1961). Data from the period 1963-72 have been analyzed and processed in the following manner. The 12 existing sampling areas in the CYRA were reassembled into three major areas: N — North of lat. 10°N except east of long. 95°W; 5 — North of lat. 5°N to the boundary of area N; S — all the CYRA south of the boundary of area 5 (see Figure 1). The areas N and S tend to have separable length-frequency distributions during any given time interval. Area 5 tends to have unique components as compared to N and S, but also contributions from both the other areas can be observed in the data from area 5. (This phe- nomenon is tjqjically nonseasonal or noncyclic with respect to the fishing year and is probably related to population and environmental pres- sures within the separate areas.) In all three areas, recruitment components of a semestral nature are evidenced. The apparent relative abundance of these components within the areas changes seasonally and also between years (Table 1). Analysis of this phenomenon has made the separation of the semestral cohorts seem the first logical step when the available genetic, mor- phometric, and length-frequency data are con- sidered. The catch data associated with each length- frequency sample were obtained. The individual sample sets were then given relative values pro- portional to the contributions of the catches (in weight) from which they were drawn. From this basic processing of all the length-frequency data, 1 JO* 1^0° iiO" 00° 90° 80" "-"tPN - AU-. V*k c^ -ssn? vx -4- iiistT*!^ -a-X j: 5^j\ I 1 it: i ± ... S J—X it -- -I : i: qii :" I M ^ ^^ J *fc^ ; -»^ 1 TV / 3" V / - - ;^x j^V / -- K + ±Js.V^--^ / - -II ? """"w^S^ S "d> - - - --\^-- - .r ,' iit- --- - - - - " A 1 - W- t !i >» »«- -^ - ^ 1 ± — J X - ^ -1 -1 \ 0' ± „.:..:...r+ -.^^ Figure l. — The study area CYRA (Commission Yellowfin Regulatory Area) used in the simulations is enclosed in the dark outline. Three subareas were used in the preliminary population dynamics work in estimating cohort strength from the length-frequency and catch and effort data appropriate to these areas. N = North of lat. 10°N except inside of long. 95°W; 5 = North of lat. 5°N to boundary of N; S = all CYRA South of boundary of subarea 5. estimates of the catch composition with respect to size-age for each fishing area were made and a growth curve was determined for each of two semestral cohorts. The two curves were essen- tially identical and warrant no further discussion here other than to say that from 40 to 145 cm fork length it is possible to give relative monthly ages to all individuals, given a length and correspond- ing date of capture. The labeling problem was handled such that any fish that was 40 cm from 1 January to 30 June is labeled S^ and correspond- ingly 40-cm recruits from 1 July to 31 December are labeled 83. The cohorts are identified in relation to their recruitment year when they are 40 cm, not their spawned year. For example, a 40-cm fish caught in February 1969 is attributed to the cohort labeled S^, 1969; and a 40-cm fish caught in October 1968 is attributed to the semester cohort labeled Sg, 1968. The two semestral groups can be treated as independent units in the population and provide a biological basis in assessment of population size with re- spect to size-age classes within the fishing year. The annual growth increment in the most often encountered cohort classes (40-140 cm) in the fishery appears to be about 32 cm/yr; therefore, 37 FISHERY BULLETIN: VOL. 74, NO. 1 Table l. — For the years 1964-71 the data are presented for the catch in short tons by semestral cohort in the three areas (N, 5, S) within the CYRA. Also given are the percent of the total catch (Sp^ + Sg + Big) by cohort within the areas. The category, Big, represents the fish of length / greater than 145 cm which we feel are not ageable under the present system. The percent of the individual semestral cohorts (S^ or Sg) caught in the three areas is also given. Note the erratic shifting of the cohort dominance (S^ or Sg) in the catch as well as the distribution of the cohorts between areas. Year North A 5 A South A Total A North B 5 B South B Total B Total A + B Big 1964 % total A + B % total A or B 1965 1966 1967 1968 1969 1970 1971 27,452 9,401 5,209 42,062 33,561 5,881 17,515 56,957 26.9 9.2 5.1 41.2 32.9 5.8 17.2 55.9 65.3 22.4 12.4 58.9 10.3 30.8 18,967 13,512 6,406 38,885 24,064 14,164 8,386 46,614 21.1 15.0 7.1 43.2 26.7 15.7 9.3 51.8 48.8 34.7 16.5 51.6 30.4 18.0 7,769 23,128 20,176 51,073 10,292 11,394 14,771 36,457 8.5 25.4 22.1 56.0 11.3 12.5 16.2 40.0 15.2 45.3 39.5 28.2 31.3 40.5 20,699 9,564 7,664 37,927 29,482 8,572 11,867 49,921 23.1 10.7 8.5 42.3 32.9 9.6 13.2 55.7 54.6 25.2 20.2 59.1 17.2 23.8 16,361 23,921 13,552 53,834 33,917 22,132 3,128 59,177 14.3 20.9 11.8 47.0 29.6 19.3 2.7 51.6 30.4 44.4 25.2 57.3 37.4 5.3 22,437 20,034 9,030 51,501 34,887 29,587 5,648 70,122 17.7 15.8 7.1 40.7 27.6 23.4 4.5 55.4 43.6 38.9 17.5 49.8 42.2 8.1 39,197 15,942 10,529 65,668 43,476 13,257 11,125 67,858 27.5 11.2 7.3 46.0 30.5 9.3 7.8 47.6 59.7 24.3 16.0 64.1 19.5 16.4 12,372 18,719 14,453 45,544 17,357 25,283 15,712 58,352 10.9 16.5 12.8 40.2 15.3 22.3 13.9 51.6 27.2 41.1 31.7 29.7 43.3 26.9 99,019 85,499 87,530 87,848 113,011 121,623 133,526 103,896 2,921 2.9 4,543 5.0 3,626 4.0 1,802 2.0 1,602 1.4 4,888 3.9 9,176 6.4 9,277 8.2 the mean lengths and modes of the two semes- teral cohorts are separated by approximately 16 cm (Tomlinson and Sharp work in progress). A significant number of animals may shift from the leading edge of one labeled distribution into the trailing edge of the other, but we are assuming that countershifts are equally as probable and both are irreversible. An effect of shortening the sampling "season," since the implementation of regulations, has been to distort the apparent abundance of the two groups and merge the modal distributions into a single amorphous dis- tribution (Figure 2). The cohorts are treated independently by the model. Each cohort is considered to have a unique effect in the analysis of the net biomass and numbers estimates for a given fishing year. Dif- ferential exploitation of these cohorts can be determined from the catch-effort length- frequency data and as such warrants this disin- tegration technique as opposed to treating the year class as a single unit. We have, however, decided not to present in this report the area breakdown results in the simulations. When the cohorts are separated, it is possible to construct a catch table for each from the length-frequency sample data from the fishery. With this catch table and the catch data (yield) it is possible to determine the relative mortality (F) attributable to fishing, by assuming a constant natural mor- tality (M), a necessary, but perhaps poor assump- tion in the case of tunas due to the inherent rapid changes in ecological status as they grow. The Murphy cohort analysis procedure (Murphy 1965; Tomlinson 1970) was used for estimation of re- cruitment at first availability to the fishery (A'40). Using this approach we have generated the un- derlying population structure for the historical series we wish to represent. Energetics The energetics parameters for free-swimming predatory species such as the tunas must be size-related functions due to the broad range of sizes commonly encountered in the fishery; 1.3 kg to greater than 62 kg, or 40 cm to greater than 145 cm. In no case for fish has anyone measured physiological parameters from such a range of sizes. Magnuson (1973) discussed the effect of gas bladders and lift surfaces on the velocity of ob- ligatory swimmers such as the tunas. He deter- mined the relationships between size and mini- mum velocity for maintenance of hydrostatic equilibrium for several scombrid species, includ- ing skipjack tuna, Katsuwonus pelamis, and Thunnus albacares. This work has provided a 38 SHARP and FRANCIS: ENERGETICS MODEL FOR YELLOWFIN TUNA POPULATION 10 RE (AS OF 20 30 40 50 LATIVE MONTHS OF AGE JANUARY OF INDICATED YEAR) basis for determining some of the relative energy expenditures in the population simulation study. The energy utilization which is simulated in ENSIM is that attributable to 1) maintenance of metabolic stasis, 2) growth, and 3) swimming. Each is calculated independently and summed with the others to give an estimate of the total minimum energy utilized on a daily basis. No attempt has been made to evaluate energy ex- penditures due to gonad maturation or migratory behavior beyond the daily forage or flight be- havior levels because of the inherent void in our knowledge of these processes in tunas. Metabolic maintenance of stasis energy re- quirements (£',„) are difficult to assess under optimum conditions and are typically derived from extrapolation of O2 consumption versus activity relationships to a zero activity level. The magnitude varies greatly between species and in general is a tenuous function of size and physio- logical state. It is essentially impossible to di- rectly measure the stasis energy requirements of tunas due to their continuous swimming be- havior. Estimates of E,,, should not include the energy expenditures due to even minimum swim- ming activity if it is to be useful in the deter- mination of energy expenditures due solely to swimming work. The respiration rate attributable to tissue stasis can be estimated from the metabolic weight (Wn^gt) of fish of length / from the equation: £„, = 24 ^ W^et (modified from Winberg 1960) where W^^^ = iMf)''-^ and Mf = 1.858 X 10-2 (/)3.o2i (grams) (Chatwin 1959) and where k is estimated to equal 1 cal/g h from data and estimates for other highly active fishes (Fry 1957; Winberg 1960). Therefore E^ = 4.46 X 10-1 (/)302i cal/day. Figure 2.— The numbers offish caught in the fishing years 1966, 1968, 1970, as a function of their recruitment month, and age, relative to the fish of the year are graphically represented. Se- mestral (A, B) and annual cohort labeling is as indicated. Note the central tendency of the peaks within the semestral limits in the years 1966 and 1968. In these years the fishing "season" was quite long (>6 mo) as compared to 1970 (<3 mo), which combined with cyclic migratory behavior and subsequent availability of cohorts probably results in the drastic change from multimodality to the amorphous distribution seen in the 1970 data. 39 FISHERY BULLETIN: VOL. 74, NO. 1 When estimates of the true stasis energy rela- tions are finally available, they can be easily incorporated into the model. Probably the most difficult process to define, estimate, and measure is that of growth. The energy requisite to growth (Eg) can be esti- mated minimally as the biomass gain per time period as converted to calories. This is a highly unsatisfactory method because of the many energy requiring steps between ingestion of a food organism and the consequential deposition of the materials assimilated into the living bio- mass of the growing organism (Phillips 1969). One slight change in the accepted method- ology of bioenergetic accounting which we will make is in our definition of specific dynamic action (SDA). If one is willing to accept that the SDA contributed little other than heat to the feed- ing organism, then it can be defined as the loss of energy due to the inefficiency of the digestive processes, including cost of transport, deamina- tion, biosynthesis, and related processes. The rate of inefficiency (percent of SDA energy with respect to total ingested energy) is variable in most animals studied as a function of feeding level (Warren and Davis 1967) and environmental conditions (Warren 1971). In our definition of SDA we do not include the unavailable portion of foodstuffs. For our purposes we will assume that growth of yellowfin tuna in the CYRA is relatively con- tinuous with respect to season or environmental state. There are several assumptions involved in this basic tenet which require some discussion. Tunas are highly endothermic animals, and Carey and Teal (1966) have shown the presence of a relatively high efficiency heat exchange (conservation) mechanism in tunas. This sug- gests that tunas are likely to be somewhat inde- pendent of ambient temperatures in that the temperature variability encountered within the core of these fishes is likely less than the ambient variability. Their large mass (>1 kg) would con- tribute to thermal stability over a wide ambient change (Neill and Stevens 1974). Observations of temperature dependent activ- ity indicate a lower activity as temperature de- creases in small yellowfin tuna (<50 cm, <2.5 kg) at a Qio of near 2 (Neill, pers. commun.). This size of yellowfin tuna is rarely encountered in the CYRA at temperatures below 23°C and is found aggregated on the warm side of the north-south surface temperature cline including this tempera- ture, indicating some preference for tempera- tures near 23°C. Preliminary studies of effects of the environmental characteristics on the abun- dance and availability of 40- to 70-cm yellowfin tuna in the CYRA indicate a direct relationship between the 23°C isotherm depth of the av- erage number of fish per school, and the overall availability of these fish to surface fishing gear (Inter-American Tropical Tuna Commission 1975). All this is emphasized to indicate the limited range of temperatures likely to be affecting the metabolic rates of yellowfin tuna as compared to that affecting smaller species without the complex stabilization mechanisms (heat ex- changers, etc.) as is the typical situation in fishes. The relative activity, mobility, and distribution with respect to temperature of yellowfin tuna can be used as supportive background for as- suming a relatively stable growth energy avail- ability as they developed, bringing us to the con- clusion that a first approximation of the SDA can be made with respect to the energy equiva- lent to the biomass change on a daily basis. From studies discussed by Paloheimo and Dickie (1966) and Warren and Davis (1967) on several species and estimates by Kitchel et al.^ for K. pelamis, it appears that SDA probably accounts for 30-40% of the total consumed calories which could be part of the growth process. We have, therefore, assumed that Eg is going to equal the equivalent caloric value of the tissues plus the SDA which will be given by the relation SDA (Biomass change in grams per day) where, if 1 g is calorically equivalent to 1.46 kcal (Kitchell et al. see footnote 2) then 3 Eg = — Biomass change (grams) (1.46 kcal/g) = 2,190 kcal /kg growth. Smit (1965) has provided the mathematical basis for our determinations of energy output and caloric requirements due to swimming. He shows that: (Meg S) (143 X 103) gcm2 (1) Power 3,600 ^Kitchell, J. F., W. H. Neill, and J. J. Magnuson. Bio- energetics of skipjack tuna, Katsuwonus pelamis. Manuscr. 40 SHARP and FRANCIS: ENERGETICS MODEL FOR YELLOWFIN TUNA POPULATION where Mg is the efficiency of the muscle tissue when converting chemical energy to mechanical work; S is the respiration due to activity in mg 02/h; and^ is the acceleration due to gravity (981 cm s'2). The propulsion efficiency is as- sumed to be 0.90 (Lighthill 1970) and is included in the resulting muscle efficiency figure. For our purposes we assume M^ to be 0.18. Therefore from Equation (1) (Power) (3,600 s/h) S = (0.18) (143 X 103 g cm) (981 cm/s^) mg Oa/h. (lA) From the hydrodynamics theory (Streeter 1962) Power =-g-Ay3Crf^^ where p = the density of seawater (1.025 g/cm^) A= 0.4(Z)2 from Bainbridge (1961) (cm^) V'= is derived from Magnuson's empirical relationships between / and species velocity V (cm/s) Cd = the coefficient of total drag of the fish, which is derived from an empirical re- lation including the results of studies by Pyatetskiy (1971). We can therefore rewrite the equation so that respiration due to swimming is equal to pAV^Cd ' 2 (7,017.66) 2.59 X 10-5 (/)2 (y)3 Crfmg Oz/h. (2) We now have an Equation (2) of three elements for which we have solutions for two (V and Cd) as functions of the third (I) given below. V Determination From Magnuson (1970), the relation for the minimum velocity (Vioo) for sustained hydro- static equilibrium by tunas is given as 100 1/2 (3) where Cj^ = the coefficient of lift for the pectoral fins Af^ = the total lifting area of the pectoral fins (cm2), log Af^ = -1.2154 + 1.87 log J Ci = the coefficient of lift of the keel A^ = the lifting area of the keel (cm^), logA* = -2.7033 + 2.26 log /(cm2) Lf = the total weight of the fish in sea- water (dynes). (L, values are ob- tained by multiplying Mf values by appropriate constants as pro- vided by Magnuson (1973) by species and weight class.) Mf= mass of the fish = 1.858 x lO'^ (/)3.02i (grams). Determination of the Coefficient of Total Drag C^ The relation between the total drag coefficient iCd) and the Reynolds number (Re) for Atlantic bonito, Sarda sarda , reported by Pyatetskiy (1971) is taken to be representative in form for scombri- / V form fishes. Re = , where v is the kinematic V viscosity of seawater or 0.01 cm^/s; / is the fish fork length in centimeters; and V is the fish velocity in centimeters per second. An analytical expression was derived for esti- mating the Cd values in the following manner: R. Gooding (Gooding et al. 1973) of the National Marine Fisheries Service Honolulu Laboratory, Honolulu, Hawaii reported respiration rates for unfed K. pelamis from 32 to 36 cm fork length, swimming at or near minimum velocities (Vioo)- From these data it was possible to calculate Cd given the observed respiration rate (St<,tai) was 431.5 mg 02/kg h and / = 35 cm. The minimum velocity (Vioo) = 59.1 cm/s and Re = 2.07 x 10^ at this velocity. For skipjack tuna of Z = 35 cm, W^et = 200.5 g, so that Sm = 60.0 mg Oa/h Stotai - S^ = S, = 371.5 mg Oa/h. From Equation (2) it is now possible to deter- mine that Cd- 371.5 2.59 X 10-5 (35)2 (59.1)3 = 0.057. This value of Cd was related to the values graphically displayed by Pyatetskiy (1971) and what was assumed to be a good approximation 41 FISHERY BULLETIN: VOL. 74, NO. 1 of the total drag on the test animals was derived relative to his graphed observations as a function of Re. From Re, one can determine the approxi- mate coefficient of total drag (C^) from the relation: Cd= 0.262 e -4-805 X 10- i?. (4) Gooding also reported respiration data for skipjack, ranging from 45 to 53 cm, swimming at or near Vioo where S^otai = 1.403 mg Oa/h. These test animals had also been deprived of food for 24 h. Assuming / = 50 cm: Wmet= 523.5 g; Vioo = 70.5 cm/s; Re = 3.525 x 10^ S„ = 156 mg Oa/h; Cd = 0.262 e-^-^^^ " "^'^ '^-^^^ ^ '°' = 0.048. /.S, = 2.59 X 10-5 (50)2(70.5)3(0.048) = 1,233 mg Oa/h. Ss+Sn, -S,,,,i = {1,233 + 156} mg O2 = 1,389 mg Oa/h, (expected) where S total = 1,403 mg 02/h, (observed) leaving 14 mg 02/h, (difference). The Relation (4) we have used for Q as a function of Re appears to be adequate for our purposes. Within the factors M^ and C^ there are an in- separable pair of modifying effects which must be accounted for, but which are essentially in- determinate at the present state of the art. One is the mechanical propulsion efficiency, and the other is the effect of the short-term flux of the rates of acceleration due to caudal fin position and velocity wdthin a single tail beat cycle on the "average" calculations of M^ and C^. The Me and Q values are continuous variables within the tail beat cycle and are inextricably bound together. Where in the integration and estima- tion of these two values the trade off is made is inconsequential due to the equal and direct effect of the estimate of one on the other value. Until either value is measured and fixed, the other coefficient is relative and therefore not necessarily realistic. The effect of velocity on propulsion efficiency is probably great in tunas (and other large organisms) due to several processes, including local heating phenomena and subsequent con- traction rate increases of the muscle fibers (Walters 1962; Sharp and Vlymen^). The graded increase in utilization of white muscle fibers as velocity is increased should result in generalized heating and increased overall efficiency of the energy conversion processes in the muscles. This and other effects may indeed account for the con- siderable efficiency changes in work done as com- pared to respiration rate when extended periods of white muscle utilization are monitored (Kutty 1968). The higher scombrids {Auxis, Euthynnus, Katsuwonus, and Thunnus) have incorporated, in various designs, a subcutaneous vascular system which is the distribution mechanism for transport of arterial and venous blood to and from the warm swimming musculature (Kishi- nouye 1923). The direct transport of "warm" venous blood to the fish's surface probably affects the hydrodynamics of the fish and con- tributes to the dynamic flux of the Cd value. Since no data are available for these phenomena, they have to be ignored in this treatment of the swim- ming energetics, but future laboratory studies should not ignore or delete these potential effectors. Considering the range of possible error in estimating both muscle efficiency and/or the co- efficient of total drag, the close agreement be- tween observed and expected respiration rates indicates that we have useful estimates of energy requirements. The only available respiration-activity data from tunas is for K. pelamis. Assuming that Magnuson's (1973) empirical relations and density multipliers are representative of the relative hydrodynamic status of the several species, these relations should give a similarly good approximation of energy consuming proc- esses in T. albacares as they appear to give for K. pelamis. The three continuous energy consuming pro- cesses are, therefore, roughly accountable using the previously described relations. The conver- sion of oxygen consumption to caloric utiliza- tion is made on the basis that 3.359 cal are avail- able from 1 mg O2. Apparently the major energy consumption process is swimming, including feeding and flight behavior. The_energy ex- pended is a function of the velocity Vjyp which is ^Sharp, G. D., and W. J. Vlymen III. The relation between heat generation, conservation and the swimming energetics of tunas. Manuscr. 42 SHARP and FRANCIS: ENERGETICS MODEL FOR YELLOWFIN TUNA POPULATION in turn a function of the length of the individuals (see Figure 3). In Magnuson's (1973) relation- ships the variables necessary for a solution for the minimum velocity are I and the density of the fish. Magnuson (1973) provided data for fish density (in the form of empirically derived multi- pliers) by weight class for several species in- cluding yellowfin tuna. We have extrapolated his data to fit our size distribution with an asymp- totic lower limit of fish density at 1.06 g/cm^ reached by 120-cm fish. We are assuming that the animals have their pectoral fins 75% extended all of the time that they are in nonfeeding-flight behavior, hence Ci = 0.75, and that the keel surface is 85% ef- fective so that Cj , = 0.85. This results in a fish that is swimming somewhat faster on the average than its Vjoo or minimum velocity. These values are "best guess" estimates and as such, repre- sent only minor changes in the appropriate di- rection as opposed to using absolute minimum energy utilization in the population simulation. Magnuson's Vioo for a 50-cm yellowfin tuna is 50.91 cm/s. Solving for the "typical" velocity under our "best guess" conditions results in a V^y^ of 58.29 cm/s. We have set a "typical" feeding-flight speed at 3 m/s. This is an integrated average that in- cludes all velocities above V^^^ and includes the burst speed forays. Since the energy required for different speeds is proportional to a cubic function of the velocities, it should be noted that the most probable velocity is less than 2 m/s, since the energy requirements for a few short bursts of up to 10 body lengths/s rapidly increase the overall energy utilization. With this in mind. FORK LENGTH ICMI Figure 3. — The energy utilization (in kcal/day) for growth (Eg), maintenance (£„,), and the total (Eg + E„ + E^ = ^total* energy utilization are portrayed as functions of length 7. we have at^ibuted 95% of the day or 22.8 h of the day to V^yp requirements and 5% or 1.2 h to Vfeed behavior. This is not to say that the fish are limited to 1.2 h/day of feeding but that on the average the increased velocity due to external stimuli are exhibited for this period. One sus- pects that the feeding of large and small tuna is entirely different in nature, but for simplicity and since no data are available, it is not unreason- able to assume that the relative effectiveness of feeding is somewhat similar over the life history of the animals. Based on these estimates we hope to have contrived a "reasonable" fiction for use in our model. The need for better estimates is obvious. MODELING RESULTS The model ENSIM computes the caloric re- quirement of each semestral cohort in the ex- ploited population, by quarter of the fishing year. Summary data are listed after each quarterly out- put which differentiate the semester A cohort caloric expenditure from that of the semester B cohort, and a composite total expenditure is listed (see Table 2). An annual summary for 1972 is also generated and an example is presented in Table 3. Initial biomass and numbers, yield in weight and numbers, gross growth, and average bio- mass are tabulated for each quarter, and sum- mary tables are generated for the individual semestral cohorts as well as composite values. The biomass of food ingested per day is gen- erated for each cohort, assuming 1.00 kcal (Paloheimo and Dickie 1966) are available per gram food ingested. The minimum percent bio- mass ingested per day with respect to the cohort biomass is also calculated for each cohort (see Figure 4). The caloric requirements for mainte- nance, swimming (at V^yp, Vfeed); ^^^ growth are tabulated by size of the average animal in each cohort in the simulation by quarter (see Table 4). We have simulated the fishing years 1964-72 and included the best available estimates for cohort strength, fishing effort, and availability parameters. We have also simulated a nonex- ploited population which was recruited at the average level for the data from the last 5 yr which includes all the population indicated or expected from inside our study area (see Figure 5). From Figure 5, the plot of the average annual biomass estimate, one can readily see the effect of fishery 43 FISHERY BULLETIN: VOL. 74, NO. 1 Table 2.— ENSIM output for quarter three of the 1972 simulation is presented. The calculated kilocalories expended by each cohort (age-class) in the exploited population is given. The appropriate averages (N/-, weight (kg) and I) are also listed for each cohort. Summary data are given by cohort and for both cohorts summed together. Age Maintenance Swimming V^p Swimming V^^^^ 'g ^ total N-l Weight (kg) / 1 .634079E+11 .640848E+11 .280680E+12 .377884E+11 .445961 E+ 12 .186707E + 08 .176060E+01 .444182E+02 2 0. 0. 0. 0. 0. 0. 0. 0. 3 .889373E+11 .776321E+11 .354145E+12 .431443E+11 .563858E+12 .141776E + 08 .379126E + 01 .572565E + 02 4 0. 0. 0. 0. 0. 0. 0. 0. 5 .501222E+11 .307991 E + 11 .185411E+12 .264195E+11 .292751E+12 .521301E + 07 .646561E+01 .683217E + 02 6 0. 0. 0. 0. 0. 0. 0. 0. 7 .463850E+11 .196178E+11 .157393E+12 .237458E+11 .247142E+12 .292458E + 07 .120872E + 02 .840421 E+02 8 0. 0. 0. 0. 0. 0. 0. 0. 9 .352487E+11 .120881E+11 .111902E+12 .138239E+11 .173063E + 12 .151084E + 07 .195812E + 02 .985929E + 02 10 0. 0. 0. 0. 0. 0. 0. 0. 11 .176499E+11 .514907E+10 .528273E+11 .601945E+10 .816457E+11 .537688E + 06 .300053E + 02 .113553E + 03 12 0. 0. 0. 0. 0. 0. 0. 0. 13 .928307E+10 .302828E+10 .263712E+11 .297797E+10 .416605E+11 .208937E+06 .438055E + 02 128704E + 03 14 0. 0. 0. 0. 0. 0. 0. 0. 15 .465814E + 09 .166772E + 09 .128095E+10 .796327E + 08 .199317E+10 .868343E + 04 .55441 4E+ 02 .139141E + 03 16 0. 0. 0. 0. 0. 0. 0. 0. 17 .695863E + 09 .256764E + 09 .189348E + 10 .249005E + 08 .287100E+10 .122020E + 05 .598477E + 02 .142709E + 03 18 0. 0. 0. 0. 0. 0. 0. 0. Total A .153438E+12 .102566E+12 .565646E + 12 .729892E+11 .894639E+12 Total B .158758E+12 .110257E+12 .606257E+12 .810347E+11 .956307E+12 Total .312196E+12 .212823E+12 .117190E+13 .154024E+12 .185095E+13 Table 3. — The 1972 annual summary data are hsted which give the yield in number and weight for each of the semestral cohorts as well as the kilocalories utilized in the year by the cohorts and the combined sum. Yield weight Kilocalories Yield numbers (metric ton) utilized Total S^ Total S Total S/i B 0.491282E + 7 0,534418 E + 7 0. 1 02570 E + 8 0.653748 E + 5 0,640257 E + 5 0.129427 E + 6 3.85 E + 12 2.28 E + 12 7.13 E + 12 ;,.o 1 1 1 1 1 1 1 1 1 IL N. 2 w \^ Z ^^^-^^^ - so ^"^^^^,_^ t — _______^ 1 M ~ — - ^• FORK LENGTH (CM ) Figure 4. — The amount of food required per day is given in percent body weight of the individual yellowfin tuna of length /. Table 4. — The estimates of the daily energy utilization (in kcal/day) for maintenance, swimming at V^yp and Vfggj, growth, and the total daily energy utilized due to all these activities is provided for the average individual of length L and weight W for each cohort in the population during each quarterly time period. The average number of individuals present in each cohort is given in the column headed N. The semestral cohorts are separated (Total A or Total B) and the energy utilization estimates summed and listed for each. The composite estimates (S^ + Sg) are also listed (Total). Age Maintenance Swimming Vjyp Swimming V(gg|j ^g ^total N W L 1 .377346E + 02 .381374E + 02 .167035E + 03 .224882E + 02 .265395E + 03 .192145E + 08 .176060E + 01 .444182E + 02 2 0. 0. 0. 0. 0. 0. 0. 0. 3 .667851E + 02 .604781 E + 02 .267903E + 03 .225852E + 02 .417752E-^03 .125166E + 08 .359406E + 01 .562530E + 02 4 0. 0. 0. 0. 0. 0. 0. 0. 5 .116280E + 03 .663364E + 02 .423896E + 03 .451705E^02 .651683E + 03 .863385E + 07 .718813E+01 .707599E + 02 6 0. 0. 0. 0. 0. 0. 0. 0. 7 .181619E + 03 .751903E + 02 .613072E + 03 .902863E + 02 .960167E + 03 . 562388 E + 07 .125513E + 02 .850967E + 02 8 0. 0. 0. 0. 0. 0. 0. 0. 9 .271052E + 03 .911480E + 02 .853897E + 03 ,112988E + 03 .132909E + 04 .387916E + 07 .207038E + 02 ,100429E + 03 10 0. 0. 0. 0. 0. 0. 0. 0. 11 .381489E + 03 .112386E + 03 ,113300E + 04 .169496E + 03 .179638E + 04 .252689E+07 .317387E+02 .115684E + 03 12 0. 0. 0. 0. 0. 0. 0. 0. 13 .497444E + 03 .162886E + 03 ,141127E + 04 ,180946E + 03 .225255E + 04 .174290E + 07 .442248E + 02 .129111E + 03 14 0. 0. 0. 0. 0. 0. 0. 0. 15 .596227E + 03 .213496E + 03 .163949E + 04 .101935E + 03 .255115E + 04 .113565E + 07 .554626E + 02 ,139159E + 03 16 0. 0, 0. 0. 0. 0. 0. 17 .639329E + 03 .236943E + 03 .173697E + 04 .340073E + 02 .264725E + 04 .783083E + 06 .605190E + 02 .143236E + 03 18 0. 0. 0. 0. 0. 0. 0. 0. Total A .156184E + 04 .595451E + 03 .459307E + 04 .395600E + 03 .714596E + 04 Total B ,122612E + 04 .461550E + 03 .365347E + 04 .384302E + 03 .572544E + 04 Total .278796E + 04 .105700E + 04 .824654E+04 .779903E + 03 .128714E + 05 44 SHARP and FRANCIS; ENERGETICS MODEL FOR YELLOWFIN TUNA POPULATION Figure 5. — The average biomass estimate of the exploited yellowfin tuna population in the CYRA is shown. The historical fishery label indicates the coastal fishery which operated prior to 1965; the expanded fishery indicates the process of seaward areal expansion which dramatically changed the estimates of exploited biomass from 1966 until approximately 1968. Fishery regulation was implemented in September 1966. The simulation of the unexploited populations yielded estimates of the average biomass for the two cohorts to be S;^ = 282,400 metric tons; Sg = 272,700 metric tons; S^ + Sg = 555,100 metric tons. Recruitment was assumed to be consistent with recent levels. NONREGULATED FISHERY REGULATED FISHERY HISTORICfiL FISHERY ^ EXPftNOEO FISHERY - O " . G \ / ' O / ^O^ "" • SEMESTER A COHORTS ' ' X SEMESTER B COHORTS O TOTAL [SEMESTER A + SEMESTER B) 1964 1965 1966 1967 1968 1969 1970 1971 1972 Figure 6. — The catch in metric tons of yellowfin tuna from the CYRA is shown for the study period. The cohorts and tottd catch are indicated by symbols as in Figure 5. -• SEMESTER fl COMOfiTS -X SEMESTER B COHORTS -o TOTAL (SEMESTER A * SEMESTER B) FIGURE 7. — Estimates from ENSIM of the kilocalories used per year by yellowfin tuna in the exploited CYRA population for the 1964-72 period. growth (areal expansion) on population size esti- mates. From Figures 6 and 7 it is obvious that the catch has great fluctuations (e.g., 1971) but the energy flow seems to have stabilized in the ex- ploited population estimates. This may be artifac- tual but we think it may be significant to attempt interpretation. The ratio of yield in weight to gross growth is another interesting indicator (Figure 8). Note the differential rate of exploitation of the semestral cohorts through time prior to 1967. The S^ and Sg cohorts became approximately equally ex- ploited in this respect about 1967 or at about the end of the changes in fishery strategy and when Rtg loted Filoricol Fishery 1 X-- - X Semes'*' B coho"s -^. 1 Semes'er B ) "^^ ^^^q\ A 1 ^ ^^ ^ ^ X Z® ^ \^^ 1 1967 (968 YEAR Figure 8. — The ratio of the yield in weight (catch) to gross growth for the years 1964-72. Note the relative similarity of the levels of the cohorts respective ratios in the regulated years as compared to the preregulated years. 45 FISHERY BULLETIN: VOL. 74, NO. 1 regulation occurred. The indication is that since approximately 1969, the biomass and exploita- tion levels on the semestral cohorts have some- how paralleled a somewhat uniform energy utili- zation by the two cohorts, whereas from 1966 until 1969 a larger semester A biomass was under exploitation compared to the semester B cohort. The large discrepancies in biomass caught as compared to gross growth in the early data (1964-65) compared to the recent data (1969-72) may be an indicator of the relative health of the stocks under exploitation in recent years in contrast to the preregulatory years. SPECULATIONS The utility of simulation studies lies in the process of linking together observations, using generalized principles where possible, to gen- erate testable hypotheses which ultimately lead to resolution of cause and effect relationships. As examples, from the results of the simulation model ENSIM, hypotheses were conceived con- cerning the relative importance of forage or- ganisms, primary productivity and the size of the animals with respect to recruitment limitations. Food as a Population Regulator The availability of food is classically attributed the role of limiting population size. We do not intend to assail this premise, but intend only to show that the most probable source of limitations is at very early ages in tunas (<40 cm), and not on the late juvenile or adult population. Forage for tunas is generally considered to be in the micronekton size range (1-10 cm). It probably extends upwards to 30 cm or more in length for larger sizes of tunas (Magnuson and Heitz 1971; Perrin et al. 1973). Tunas eat largely crustaceans, fishes, and cephalopods in most regions (Alverson 1963; Magnuson and Heitz 1971; Perrin et al. 1973). These organisms are poorly sampled by micronekton sampling devices. The EASTROPAC cruises sampled from our study area over the year 1967 and early 1968. Productivity, micronekton, and most physical and chemical properties which are linked to biological productivity were sampled. EASTRO- PAC data (Blackburn et al. 1970) indicate that the average minimum micronekton night haul contained 5 ml of micronekton per 10^ m^ of water sampled. The samples represent a 200-m water column. The surface area of the CYRA is estimated to be 5,012,643 sq nautical miles or 1.696 x lO^^ m2. The minimum available forage is therefore (1.696 X 1013 m2) (200 m) 5 ml forage 103 m^ ) = 1.696 X 10^3 cc. If 1 cm^ forage has approximately 1 g or 1.25 kcal caloric equivalency, then one should expect that there is a minimum forage availability of 1.25 kcal/m^ or assuming 80% utilization efficiency of these calories by predators (Winberg 1960), 1.0 kcal/m^ are present for metabolic utilization. Owen and Zeitzschell (1970) in their analysis of EASTROPAC data also show that the primary productivity averages 169 mg carbon m'^ day^ over long. 119°- 112° W, 219 mg carbon mr^ day^ at long. 105°W, and 282 mg carbon m'^ day^ along long. 98°W. They also indicate coastal effects as being the probable cause of the east- ward increase in productivity. The average pro- ductivity over the entire study area was 205 mg carbon m'^ day^. The energetic equivalent value for 1 mg carbon fixation is 11.4 cal (Piatt and Erwin 1973), so that the average caloric productivity is 2,340 cal/m^ day (or 2.34 kcal/m^ day). We have seen that the minimum estimate of the micronekton standing stocks caloric value is 1,250 cal/m^, indicating that the probable daily turnover rate is less than 125 cal/m^ so that maintenance of this stock is not unreasonable if the primary production is 2,340 cal/m^ day. The yellowfin tuna population simulation pro- cedure based on average Murphy recruitment estimates of the 1966-71 S^ and Sg cohorts indi- cates that an unfished population (exhibiting a stable age structure) would have the biomass of 600,000 metric tons (6.0 x lO^^ g). Assuming that the yellowrfin tuna (YF) are distributed pro- portionally over the forage: 6.0 X 1011 g YF 1.696 X 1013 ni = 3.54 X 10-2 g YF/m2 = 35.4 mg YF/m2; 35.4 mg YF/m2 x 1.2 cal/mg YF = 42.5 cal/m^. 46 SHARP and FRANCIS: ENERGETICS MODEL FOR YELLOWFIN TUNA POPULATION Assuming the average caloric consumption by the yellowfin tuna population per day to be 10% of its caloric biomass, a somewhat higher than realistic estimate, daily utilization in calories would be 4.25 cal/m^ day. The results of the ENSIM estimates of the total calories utilized per year for the unexploited population was 14.96 x 10^^ cal/annum, so that the resulting utilization per square meter day is given by: 14.96 X 10^5 cal/annum _ 2.5 cal (365 day /annum) (1.696 x lO^^ m2) ~ m^ day The results of the simulations of the exploited fishery for the years 1964-72 yield estimates of less than 50% of this figure as the energy utili- zation by the yellowfin tuna population. One would expect the true values of caloric utiliza- tion to lie somewhere in the range from approxi- mately 1.5 cal/m^ day to the upper value of 4.25 cal/m^ day. With the primary productivity estimated to be at an average level of 2.34 kcal/m^ day and forage standing stock utilizable caloric values averaging at a minimum of 1.00 kcal/m^, it seems hardly likely that yellowfin tuna are food limited from the 40-cm recruitment size. This brings up the problem of how the east- ern tropical Pacific yellowfin tuna population is limited. This, of course, is best taken in per- spective. Population limitation examples are typically taken from terrestrial populations and extrapolations made to ecosimilar strategies in closed systems such as lakes and estuaries where primary productivity is greatly affected by season, and indeed can be determined to be the limiting factor in population numbers and biomass. In those marine animals where density de- pendent growth functions are evidenced there is generally a two-dimensional limitation imposed such that crowding is likely to affect each indi- vidual. For filter-feeding organisms, such as herring and menhaden, the density dependent function is easily conceptualized. One needs only to examine the relative abun- dance of food available to highly mobile preda- tory species which feed opportunistically on organisms ranging in size from 1 to 30 cm, which are available on a relatively continuous basis in a tropical system, to see that dogma general to terrestrial, estuarine, limnetic, two-dimen- sional substrate tied, or filter-feeding animal ecology does not generally apply to the 40- to 140-cm yellowfin tuna. There are, however, several possibilities con- cerning the survival of yellowfin tuna from larvae to 40 cm which would certainly fit into the schemes which typically limit species. Since they are probably particulate feeders (e.g., do not undergo ecometamorphoses at early ages from filter feeders to predators), it can easily be seen that they are victims of the availability of con- centrations of food at smaller sizes because of their relative lack of mobility. If a 40-cm tuna requires 10-20% of its body weight per day to maintain, as compared to 3-5% in large yellowfin tuna, then one can hypothesize that the smaller predators must consume even greater amounts due to the pressures of very rapid growth, feed- ing activity, and competition with peers, indicat- ing that they are more likely severely affected by density of both conspecifics and food than are the larger sized fish. Another consideration is the size distribution of the forage organisms. It is obvious that there are considerably larger amounts of the smaller food organisms than the bigger sizes, which would perhaps indicate that the real density competition pressures are on the intermediate sizes (vis. 10-40 cm) as compared to the post- larval sizes. This brings us to the next important process, larval survival. Spawning Survival Versus Population Biomass For our hypothesized unexploited population of 600,000 metric tons of individuals from 40 to 140 cm fork length, we can calculate the requisite number of postlarval survivors which must be generated each year to maintain this stock at equilibrium. Assuming 40-cm yellowfin tuna are approximately 7 mo of age and that the survival rate is constant for all ages after postlarval trans- formation and is approximately equal to e"-^ on an annual basis (Hennemuth 1961), the number of postlarval survivors each year is given by the relation A^, = A^4oe 0.8(1) If A^4o is approximately 2.12 x 10'^ individuals per year in cohort S^, and 2.06 x 10^ in cohort Sq, then there are approximately 6.67 x 10'' sur- vivors/yr. If we assume that they are aggregated 47 FISHERY BULLETIN: VOL. 74, NO. 1 spatially but not temporally (there are two co- horts of 3.33 X 10'' postlarvae spread approxi- mately evenly over the year), approximately 9.13 X 10"* postlarvae enter the system daily. (This is the equivalent of nearly 1% reproductive suc- cess of either one 155-cm female or five 87- cm females.) The relative fecundity of yellowfin tuna is given by Joseph (1963) to the following: Number of eggs = 8.955 x lO'^ Z2.791 where / is the fork length of the fish in mm. If we assume the average spawning female to weigh 25 kg and we estimate the presence of 175,000 metric tons of females of reproductive age in our unexploited population, then the equiv- alent number of reproductive females is ap- proximately equal to 7 x 10^. These females would be an average of 107 cm in length and therefore: (8.955 X IQrH (l,0702-''9i) (7 x 10« females) = 1.79 X 10^^ eggs produced. So if 6.67 X 10"'' postlarvae start the process we need invoke only 3.72 postlarval survivors per million eggs spawned. This estimate is conserva- tive due to the assumption that females only spawTi once per year, whereas they could spawn more often. (No evidence for or against multiple spawnings is in existence for yellowfin tuna.) It does, however, seem likely that spawning suc- cess (survival to postlarvae) is greater than 3.72 individuals per million eggs produced (Sette 1943; Farris 1961). It is also important to mention that all attempts at relating spawning biomass to recruitment estimates for yellowfin tuna in the CYRA have been futile. This could be due to error in either, or both, estimates of spawning biomass and recruitment and/or the possibility that environmental conditions indeed override any obvious relationships. These comments are presented to point up the likelihood that the density dependent factors for limiting yellowfin tuna abundance are probably more effective on the egg to larvae to juvenile stages than at 40 cm or more. The larvae to 40-cm fish are likely very narrowly distributed in the water column (approximating a two-dimensional distribution) due to thermal and energetic re- quirements. The recruitment at 40 cm in the highly productive regions such as the periphery of the Costa Rica Dome and the Panama Bight- Ecuador coastal regions can perhaps be best explained by the high productivity levels in these regions which ranges from 500 to 700 mg carbon m'2 day"^ as compared to the 205 mg carbon m"^ day^ average CYRA carbon fixation rate, in con- junction with the relatively shallow oxygen mini- mum and thermal optima which probably act to compress the available habitat toward the sur- face. If one could invoke the ability of yellowfin tuna to climb a food gradient, a simple volume change in the preferred thermal-oxygen regime combined with a negatively correlated food gradient could result in the observed coastal "emergence" of recruits, which "grow out" of their previous thermal-oxygen limitations as they develop, and exploit a significantly wider niche than they could as relatively poikilothermal enti- ties at sizes below 40 cm. To summarize, larval tunas are relatively im- mobile and for survival are probably dependent on aggregations of food resources. The ability of tunas, particularly postlarval sizes, to detect food gradients is unknown, but may indeed ac- count for the easterly trend in abundance of recruits. The wider distributions of larger fish (postrecruits) probably is a response to competi- tive feeding problems and changing physiologi- cal capabilities. These larger fish are increasing their daily demands but are gaining in adaptive physiological and morphological characteristics which widen their niche as compared to smaller sizes. Their mass and mobility insure their ability to move rapidly from low to high availabilities of food resources, in response to seasonal and areal fluctuations in productivity, perhaps ac- counting for the cyclic migratory behavior ob- served in their first few years in the fishery. The relative offshore surface distribution of the larger fish (>40 cm) may be roughly correlated vdth the depth distribution of the 22°-23°C isotherms, a relationship which we are now starting to study. As the larger fish grow in mass, they can afford deeper and longer forays into colder than optimal zones with low O2 availability to obtain larger and more calorific food sources; and by thus increas- ing the maximum excursion depth, competition is likely to be less severe. The disaggregation of larger sized fish into smaller schools (number of individuals) may be accounted for by these effects. The large yellowfin tuna in the offshore areas are certainly concentrated at the surface 48 SHARP and FRANCIS: ENERGETICS MODEL FOR YELLOWFIN TUNA POPULATION over highly productive regions where their main sources of competition are probably porpoise and bigeye tuna, Thunnus obesus. The porpoise- tuna composite likely indicates the optimum availability offish and squid in the eastern tropi- cal Pacific. It is obvious from the Perrin et al. (1973) studies that the two Stenella species and tunas coexist but tend to feed differentially. The tuna diet shares most of the organisms found in both species indicating that they are less selective and/or feed throughout the water column. No data support the concept of food limitation for population size in yellowfin tuna in post- recruit sizes and in most cases the arguments tend toward the opposite conclusion. Since no stable relationship can be found to exist be- tween recruitment and spawning biomass, it is unlikely that reproductive success is affected by spawning biomass at the population levels we are experiencing. More probable is that the environ- mental parameters are more important in regulat- ing the absolute numbers of surviving larval or juvenile yellowfin tuna which are recruited to the fishery. In the future, we plan to incorporate the avail- able productivity and environmental data (tem- perature, oxygen, etc.) with a more complete version of this model. We hope to determine the environmental correlates with the fluctuations in the catch, effort, and length-frequency data generated from the fishery on yellowfin tuna. Pre- liminary studies have been encouraging (Inter- American Tropical Tuna Commission 1975) and point up the need for data on the thermal pref- erences (perhaps indicating energetic optima) and the levels of environmental variability which can be sensed and therefore compensated for by the several tuna species at the various develop- mental stages in their life cycles. Also obvious is the need to work with smaller areas and corre- sponding population segments rather than as- suming "average" conditions in environmental and population parameters. The ultimate goal of these studies is the development of predictive tools for use in assessing likely catch conditions as well as the basic distributional properties of the tunas. The use of unsupported guesses based on overviews which integrate vast areas with sig- nificant oceanographic and population structure differences may do little more than obscure the existing relationships which are important to this goal. The application of the crude model we have described in this study will depend upon the development of better estimates of the physiological parameters and appropriate use of the areal breakdown in the population simu- lator. Studies of trophic dynamics and competi- tion interactions would help complete the pic- ture necessary to "efficiently" manage a dynamic resource. We hope to generalize, where possible, the relationships which arise fi-om these analyses in order to provide a useful descriptive tool as well as a hypothesis testing device for studying the occurrence, abundance, and availability of tunas in the world ocean. LITERATURE CITED Alverson, F. G. 1963. The food of yellowfin and skipjack tunas in the eastern tropical Pacific Ocean. [In Engl, and Span.] Inter-Am. Trop. Tuna Comm. Bull. 7:293-396. BAINBRIDGE, R. 1961. Problems of fish locomotion. Symp. Zool. Soc. Lond. 5:13-32. BLACKBURN, M., R. M. LAURS, R. W. OWEN, AND B. ZEITZSCHEL. 1970. Seasonal and areal changes in standing stocks of phytoplankton, zooplankton and micronekton in the eastern tropical Pacific. Mar. Biol. (Berl.) 7:14-31. Carey, F. G., and J. M. Teal. 1966. Heat conservation in tuna fish muscle. Zoology 56:1464-1469. CHATWIN, B. M. 1959. The relationships between length and weight of yellowfin tuna (Neothunnus macropterus) and skipjack tuna (Katsuwonus pelamis) fi-om the Eastern Tropical Pacific Ocean. [In Engl, and Span.] Inter-Am. Trop. Tuna Comm. Bull. 3:305-352. farris, d. a. 1961. Abundance and distribution of eggs and larvae and survival of larvae of jack mackerel (Trachurus sym- metricus). U.S. Fish Wildl. Serv., Fish. Bull. 61:247-279. Francis, R. C. 1974. TUNP0P, a computer simulation model of the yellowfin tuna population and the surface tuna fishery of the eastern Pacific Ocean. [In Engl, and Span.] Inter-Am. Trop. Tuna Comm. Bull. 16:235-279. FRY, F. E. J. 1957. The aquatic respiration of fish. In M. E. Brown (editor), The physiology of fishes, Vol. 1, p. 1-63. Aca- demic Press Inc., N.Y. Gooding, R., E. Poe, and C. Nagamine. 1973. Tuna newsletter No. 9 July 1973. Natl. Mar. Fish. Serv., Southwest Fisheries Center, La Jolla, Calif hennemuth, r. c. 1961. Size and year class composition of catch, age and growth of yellowfin tuna in the eastern tropical Pacific Ocean for the years 1954-1958. [In Engl, and Span.] Inter-Am. Trop. Tuna Comm. Bull. 5:1-112. Inter-American Tropical Tuna commission. 1975. Aimual Report of the Inter-American Tropical Tuna Commission, 1974. [in Engl, and Span.] 169 p. 49 FISHERY BULLETIN: VOL. 74, NO. 1 JOSEPH, J. 1963. Fecundity of yellowfin tuna (Thunnus albacares) and skipjack {Katsuwonus pelamis) from the Pacific Ocean. [In EngL and Span.] Inter-Am. Trop. Tuna Comm. Bull. 7:257-292. KISHINOUYE, K. 1923. Contributions to the comparative study of the so- called scombroid fishes. J. Coll. Agric, Imp. Univ. Tokyo 8:293-475. KUTTY, M. N. 1968. Respiratory quotients in goldfish and rainbow trout. J. Fish. Res. Board Can. 25:1689-1728. LIGHTHILL, M. J. 1970. Aquatic animal propulsion of high hydromechani- cal efficiency. J. Fluid Mech. 44:265-301. MAGNUSON, J. J. 1966. Continuous locomotion in scombroid fishes. (Abstr.) Am. Zool. 6:503-504. 1970. Hydrostatic equilibrium of Euthynnus affinis, a pelagic teleost without a gas bladder. Copeia 1970:56-85. 1973. Comparative study of adaptations for continuous swimming and hydrostatic equilibrium of scombroid and xiphoid fishes. Fish. Bull., U.S. 71:337-356. MAGNUSON, J. J., AND J. G. HEITZ. 1971. Gill raker apparatus and food selectivity among mackerels, tunas, and dolphins. Fish. Bull., U.S. 69:361-370. Murphy, G. I. 1965. A solution of the catch equation. J. Fish. Res. Board Can. 22:191-202. NEILL, W. H., AND E. D. STEVENS. 1974. Thermal inertia versus thermoregulation in "Warm" turtles and tunas. Science (Wash., D.C.) 184:1008-1010. OWEN, R. W., AND B. ZEITZSCHELL. 1970. Phytoplankton production: Seasonal change in the oceanic eastern tropical Pacific. Mar. Biol. (Berl.) 7:32-36. PALOHEIMO, J. E., AND L. M. DICKIE. 1966. Food and growth of fishes. 11. Effects of food and temperature on the relation between metabolism and body weight. J. Fish. Res. Board Can. 23:869-908. Perrin, W. F., R. R. Warner, C. H. Fiscus, and D. B. Holts. 1973. Stomach contents of porpoise, Stenella spp., and yellowfin tuna, Thunnus albacares, in mixed-species aggregations. Fish. Bull., U.S. 71:1077-1092. PHILLIPS, A. M., JR. 1969. Nutrition, digestion, and energy utilization. In W. S. Hoar and D. J. Randall (editors). Fish physiology. Vol. 1, p. 391-432. Academic Press, N.Y. PLATT, T., AND B. ERWIN. 1973. Caloric content of phytoplankton. Limnol. Oceanogr. 18:306-310. Pyatetskiy, V. Ye. 1971. Hydrodynamic swimming characteristics of some fast marine fish, from translation of monograph; Kiev, Bionika, Russian, No. 4; 1970, p. 3-120. Hydrodynamic Problems of Bionics. JPRS 52605, p. 24-31. Sette, O. E. 1943. Biology of the Atlantic mackerel (Scomber scom- brus) of North America. Part I: Early life history, includ- ing the growth, drift, and mortality of the egg and lar- val populations. U.S. Fish Wildl. Serv., Fish. Bull. 50:149-237. SMIT, H. 1965. Some experiments on the oxygen consumption of goldfish iCarassius auratus L.) in relation to swimming speed. Can. J. Zool. 43:623-633. STREETER, V. L. 1962. Fluid mechanics. 3rd ed. McGraw Hill Book Co., N.Y., 555 p. TOMLINSON, P. K. 1970. A generalization of the Murphy catch equation. J. Fish. Res. Board Can. 27:821-825. Walters, V. 1962. Body form and swimming performance in the scombroid fishes. Am. Zool. 2:143-149. 1966. The "problematic" hydrodynamic performance of Gero's great barracuda. Nature (Lond.) 212:215-216. WARREN, C. E. 1971. Biology and water pollution control. W. B. Saunders Co., Philadelphia, 434 p. WARREN, C, E., AND G. E. DAVIS. 1967. Laboratory studies on the feeding, bioenergetics, and growth of fish. In S. D. Gerking (editor). The biological basis for freshwater fish production, p. 175- 214. John Wiley & Sons Inc., N.Y. WINBERG, G. G. 1956. [Rate of metabolism and food requirements of fishes.] Nauchnye Tr. Belorussk. Gos. Univ. Minsk, 253 p. (Transl. 1960. Fish. Res. Board Can., Transl. Ser. 194.) 50 SHARP and FRANCIS: ENERGETICS MODEL FOR YELLOWFIN TUNA POPULATION APPENDIX.— GLOSSARY OF TERMS A A, E„ E^ = E. = F = g k I T L, = M = M, = Mf = Nr N. = N^ = Re = wetted surface area of the fish. the total lifting area of the pectoral fins. the total lifting area of the keel. the coefficient of lift of the pectoral fins. the coefficient of lift of the keel. coefficient of total drag of fish of length J which includes an inseparable effi- ciency term involving acceleration pro- cesses during continuous swimming. the daily caloric expenditure of fish of length J attributable to growth in the form of positive changes in mass. the daily caloric expenditure of fish of length 7 to maintain metabolic stasis. the daily caloric energy expenditure of fish of length I utilized by swimming work, a function of swimming velocity (V.eal)- the instantaneous mortality rate due to fishing. acceleration due to the force of gravity. the rate of oxygen consumption due to met- abolic stasis of 1 g of respiring tissue, not doing external work. the length of a fish from snout to fork of tail in millimeters. the fork length of a fish in centimeters. the total weight of a fish in seawater of density p, in dynes. the instantaneous natural mortality rate. the efficiency of muscle when converting chemical energy to mechanical work. mass of the fish in grams where for yellow- fin tuna: M f 1.858 X 10-2 (/) 3.021 (Chatwin 1959). the estimated number of individuals of length I. the number of postlarval survivors from a spawning, the number of recruits at 40 cm. the Reynolds number. Sa Sr 'total = the density of seawater, in this work p = 1.025 g/cm2. = the rate of oxygen consumption due to swimming activity, from the power equation of Smit (1965). = recruitment cohort label for all individuals that attain 40 cm fork length from 1 January to 30 June of each year. = recruitment cohort label for all individuals that attain 40 cm fork length from 1 July to 31 December of each year. = the oxygen consumption rate of fish of length J attributable to metabolic stasis. = the oxygen consumption rate of a fish of length 7 attributable to swimming en- ergy expenditures. S + S = respiration rate attributable Mf X 10-3 V V V V 100 to swimming and metabolic stasis en- ergy expenditures. = the kinematic viscosity of seawater. = the constant velocity of a fish, in centi- meters per second. = the estimated integrated velocity of a fish of length 7 used in determining Re and C78%) occurred from 15 to 35%>, but survival decreased sharply at all other temperatures for salinities above 25%. Table l. — Number of winter flounder eggs at each of 67 temperature-salinity combinations. Number of replicates shown in parentheses Salinity Temperature (°C) C^) 3 5 7 10 12 14 0.5 100(2) 100(2) 100(2) 100 (2) 100 (2) 100(2) 5.0 100(2) 300(6) 400(8) 300(6) 300(6) 300 (6) 7.5 100 (2) 100(2) 100(2) 50(1) 100 (2) 10.0 100 (2) 300(6) 400(8) 300(6) 300(6) 300(6) 15.0 300(6) 400(8) 200(4) 300(6) 300(6) 20.0 100(2) 300(6) 400(8) 300(6) 300 (6) 300(6) 25.0 200(4) 300(6) 300(6) 300(6) 200(4) 30.0 100(2) 300(6) 400(8) 300(6) 300(6) 300(6) 35.0 100(2) 300(6) 400(8) 300 (6) 300(6) 300(6) 37.5 100 (2) 100(2) 100(2) 100 (2) 100 (2) 40.0 200(4) 300 (6) 200(4) 200(4) 200 (4) 45.0 100 (2) 100(2) 100 (2) 100 (2) 100 (2) 100(2) Influence of Temperature and Salinity on Total and Viable Hatch The influence of temperature and salinity is shown in the percentages of mean total hatch and mean viable hatch (Table 3). There is a sharp de- crease in mean total hatch and mean viable hatch at temperatures over 3°C, while these means ap- proximate a normal distribution at the salinities tested. The mean percentage of abnormal larvae calculated from total and viable hatch data shows 53 FISHERY BULLETIN: VOL. 74, NO. 1 0.5 5 7.6 10 15 20 25 SALINITY (%.) 30 35 Figure l. — The effects of temperature and salinity on the percent viable hatch of winter flounder embryos. Table 2. — Mean percent total and viable ( ) hatch at the various temperature- salinity combinations. Salinity Temperature (°C) (y„) 3 5 7 10 12 14 0.5 0(0) 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) 5.0 26(0) 6.3 (0) 14.5 (6.5) 23.0 (0) 0.0 (0) 0.0 (0) 7.5 — 58.0 (26.0) 48.0 (26.0) 49.0 (21.0) 46.0 (17.0) 23.0 (7.0) 10.0 88 (61) 79,7 (65.7) 71.5 (57.8) 59.0 (32.0) 82.7 (65,3) 55.3 (32.7) 15.0 92 (84) 79.3(75.7) 76.5 (71.0) 71.0 (57.0) 77.0 (69.0) 69.3 (57.3) 20.0 100(99) 82.3 (79.3) 83.8 (82.0) 70.3(61.0) 78.3 (68.0) 61.3 (48.7) 25.0 — 75.5 (74,0) 69.3 (66.7) 74.0 (66,5) 62.0 (56 5) 57 5 (42.0) 30.0 74 (64) 54.7 (45.7) 59.8 (51.3) 50.0 (43.7) 63.0 (54.7) 48.7 (32.3) 35.0 84 (67) 31.3 (27.3) 42.8 (37.5) 31.0 (24.0) 47,0 (34,5) 21.0 (7.7) 37.5 — 40.0 (37.0) 86.0 (78.0) 57.0 (52.0) 38.0 (16.0) 5.0 (0) 40.0 — 19.0 (9.5) 63.0 (15.7) 34.0 (17.0) 26.0 (3.5) 0.0 (0) 45.0 0(0) 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) Table 3. — Means and ranges for percent total and viable hatches and mean abnormal hatches for each salinity at all temperatures, and each temperature at all salinities. Mean % total hatch Mean % viable hatch Mean abnormal Item (Range) (Range) hatch' (%) 0.5% No hatch No hatch 5.07oo 12,8(2.3-26.0) 1.6(0-6.5) 11.2 7.57.. 44.8(23.0-58.0) 19.4(7.0-26.0) 25.4 io.oy„ 72.7(55.3-88.0) 52.4(32.0-65.7) 20.3 15.07.. 77.5(69.3-92.0) 72.3(57.0-84.0) 5.2 20.07.O 79.3(61.3-100) 73.0(48.7-99.0) 6.3 25.07„ 67.7(57,5-74.0) 61.1(42.0-74.0) 6.6 30.07„ 58.4(48.7-74.0) 48.6(32.3-64.0) 9.8 35.07„ 42.9(21,0-84.0) 33.0(7.7-67.0) 9.9 37.57.0 45.2(5.0-86.0) 36.6(0-78.0) 8.6 40.07.. 15.1(9.5-21.0) 11.4(3.5-17.0) 3.7 45.07„ No hatch No hatch 3°C 77.3(26.0-100) 62.5(0-99.0) 14.8 5°C 51.7(6.3-82.3) 44.0(9.5-79.3) 7.7 TC 57,3(14,5-86.0) 49.3(6.5-82.0) 8.0 10°C 48.1(2.3-74.0) 37.4(0-66.5) 10.7 12''C 56.3(13.0-82.7) 42.7(3.5-69.0) 13.6 14°C 42.6(5.0-69.3) 28.5(0-57.3) 14.1 'Mean abnormal hatch viable hatches. mean percent total hatches - mean percent no trend with temperature, but a high percentage of abnormal larvae for salinities of lO'L and be- low. Lowest percentages for abnormal larvae were for salinities between 15 and 35%. The low percentage for 40.0% reflects low hatching rates and mortality during embryonic stages and does not reflect values which can be compared with salinities of 37.5%o and below. Analysis of variance performed on the survival data indicate that salinity and temperature are both significant factors (Table 4). Because of miss- ing data (Table 1), it was not possible to test for interaction between the two factors; however, by examining the data, especially as it is expressed in Figure 1, it is reasonable to conclude that an interaction does occur. The multiple comparison of means indicates significant differences be- tween hatch means at various temperatures and 54 ROGERS: EFFECTS OF TEMPERATURE AND SALINITY ON WINTER FLOUNDER Table 4. — Analysis of variance for the effects of tempera- ture and salinity on the survival and hatching of winter flounder embryos. Source of variation Total Salinity Temperature Residual "significant at P = 0.005. Table 5. — Duncan's multiple comparison of means for temperature-salinity studies of winter flounder embryos. (Means with similar symbols denote similar mean survi- val percentages.)' Degrees of Sum of Mean freedom squares square F 71 69.248.75 11 51,935.36 4,721.39 31.5 5 9,078.78 1,815.76 12.2 55 8,234.61 149.72 Temperature Mean survival (°C) (%) 3 56. IV 5 36.2* 7 41. rV 10 31.4*x 12 32.4-^ 14 18.9^ Salinities Mean survival (%) 0.5 5.0 7.5 10.0 15.0 20.0 25.0 30.0 35.0 37.5 40.0 45.0 0.0 V 1 . 1 v 21.6" 53.7° 69.9t 74.3t 67.4t 52.6° 35.6^ 40.3x 15.3* 0.0 V 'P = 0.05. salinities and allows a grouping of each in order of its significance (Table 5). The grouping of the hatch means for variations in both temperature and salinity coincides closely with viable hatch curves illustrated in Figure 1. Incubation Time and Duration of Hatching Interval The time to 50% hatch and the total range of hatching time for each temperature and salinity combination are recorded in Table 6. Figure 2 il- lustrates the time to 50% hatch and the mean incubation time for each temperature and salin- ity respectively. The mean hatching interval 26 > 22 < o 18 < X 14 10 2 - 5. 00 4.0 C % 10 15 20 25 30 SALINITY (%.) 35 40 FIGURE 2. -The effects of salinity on the time to 50% hatch of winter flounder embryos. ranges from 25 days at 3°C (10%) to 7 days at 12° and 14°C (37.5 and 35% respectively). Individual eggs hatched in as few as 5 days in most salinities at 12° and 14°C, but took as long as 31 days at 3°C (10%). An inverse relationship for tempera- ture with respect to the duration of hatching time is evident. There is also a trend toward the same inverse relationship with respect to salinity as can be seen in Figure 2 where the time to mean 50% hatch at all temperatures decreased slightly with increasing salinities. This phenomenon of greater hatching time at low salinities was noted in Pacific cod eggs by Forrester and Alderdice (1966). When salinity means versus incubation time is considered by least squares regression, there is a low correlation coefficient and a regres- sion relationship is not applicable (Figure 3). However, temperature means have a high corre- lation coefficient and there is a strong regression relationship present. Table 6. — Time in days to 50% hatch. Range of hatching interval in days shown in parentheses. NH denotes no hatch. Temperature Salinity (%) (°C) 5.0 7.5 10.0 15.0 20.0 25.0 30.0 35.0 37.5 40.0 3 24 25 (19-31) 22 (19-27) 20 (19-25) 20 (17-25) 19 (16-25) 5 21 20 20 19 19 19 18 17 16 16 (16-20) (17-29) (17-25) (17-29) (22-24) (13-25) (11-25) (14-16) (14-16) 7 22 13 15 15 15 13 14 13 12 12 (10-16) (12-23) (12-23) (12-25) (8-17) (8-21) (11-19) (12-14) (8-14) 10 15 12 12 11 10 9 9 9 9 9 (10-14) (7-15) (9-14) (7-16) (7-17) (5-13) (8-10) (7-10) (7-10) 12 NH 10 9 9 9 9 8 8 7 8 (7-12) (5-10) (7-12) (7-10) (5-10) (5-10) (5-10) (5-10) (5-10) 14 NH 8 8 8 8 8 8 7 NH NH (5-10) (5-10) (6-10) (5-10) (5-10) (5-10) (5-10) 55 FISHERY BULLETIN: VOL. 74, NO. 1 30 26 CO V 22 < o UJ 16 t- z o 14 K < m o 10 SALINITY %. 15 20 25 30 35 40 45 T 1 1 r o SALINITY • y 16.0903-0.06824, /• = -0.54509 f- 2.959 • TEMPERATURE y- 28.7585-1.5441.. •S. r -- 0.96095 /^= 48.342 O >s. _ _ • \v O o .^ ° ° SALINITY N. o • \. ^V • - TEMPERATURE 1 J 1 1 6 8 10 12 14 16 16 TEMPERATURE "C Figure 3. — The mean hatching time of winter flounder embryos for each temperature and sahnity. Effects of Temperature and Salinity on Embryonic Development In each of the three experiments, general ob- servations were made on the eggs, embryos, and larvae (Figure 4). No development occurred in a salinity of 0.5%; however, the eggs swelled ap- proximately 20% before death occurred. A diame- ter increase of 8 to 10% was also observed in eggs held at 5%. Below 10°C, embryos held in 5% ap- peared to develop normally, then died just prior to hatching. At 10°C and above, most of the embryos died during gastrulation. Embryos held in a sa- linity of 10% had the highest mortalities just prior to hatching and at hatching; many larvae were observed dead partly emerged from the chorion. Mortality occurred throughout develop- ment at 12° and 14°C. In salinities between 15 and 30%, most mor- talities occurred just prior to hatching, although MO bEVELOPMEWr 14 i 12- < UJ UJ 5- 3- >. CO L.\_AP5 1 X £a^ftfcK30KJfV\.(VU bfeVELOPMEKir ' I OF EMMlyO ) t —I 1 1 1 1 1 1 1 1 1 1 '_ 0.5 5 7.5 10 15 20 25 30 35 37.5 40 45 SALINITY (%•) Figure 4.— The qualitative effects of temperature and salinity on the development and hatching of wdnter flounder embryos. 56 ROGERS: EFFECTS OF TEMPERATURE AND SALINITY ON WINTER FLOUNDER at temperatures of 10°C and above some mor- talities usually occurred during gastrulation. At salinities of 35 to 40%, abnormal development of the embryos was observed. The embryos were shorter and thicker than normal and died just prior to hatching. Collapsing eggs were noted at 37.5'Ii and above. Embryos incubated at 40% died during gastrulation and throughout development at all temperatures while all embryos held at 45% died during gastrulation. At both 40 and 45% em- bryos exhibited shrinkage and often collapsed. DISCUSSION The results indicate that although temperature and salinity are both significant, the major effect of increased temperature is to decrease the incu- bation period, whereas salinity is the factor which has more effect on the successful hatching and survival of winter flounder embryos and lar- vae (Figure 4, Table 4). It is apparent however, that an interaction between the two does occur since, at the optimum experimental temperature (3°C), the salinity range over which high percen- tages of viable hatches occurred was extended by 10% (Figure 1). At higher than optimal experi- mental temperatures, the survival curves appear to be dictated primarily by salinity; however, survival occurs over a broad enough range that the embryos and larvae can be described as euryhaline with regard to the natural environ- ment in which they are normally spawned. At all temperatures tested, there was a decrease in in- cubation time at higher salinities, a phenomenon which was also reported in studies done on Clupea harengus (Holliday and Blaxter 1960) and Pacific cod (Forrester and Alderdice 1966). Those authors speculated that the relationships of temperature and salinity with hatching are de- pendent on conditions that minimize the energy required of the embryos in maintaining osmotic equilibrium with their environment. Salinity also appears to influence the time of embryo mortal- ity. Observations on eggs indicated that mortality usually occurred either at gastrulation, in salinities of 40 and 45% at all temperatures, or just prior to hatching in the lower salinities. Bat- tle (1930) noted increased mortality of the four bearded rockling, Enchelyopus cimbrius, at hatching in low salinities and she attributed this to poorly developed tail musculature. McMynn and Hoar (1953), working with embryos of the Pacific herring, Clupea harengus pallasi, ob- served that with the closing of the blastopore at the end of grastrulation, embryos had a greater ability to tolerate low salinities. However, many embryos died just prior to hatching or when partly emerged. Holliday (1965, 1969) observed a similar occurrence in cod, Gadus callarius, and plaice, Pleuronectes platessa. He felt that the low specific gravity of such salinities made it difficult for larvae to free themselves from the chorion so that they died partly emerged. He also main- tained that chorions did not rupture as easily at low salinities. This phenomenon is also clearly demonstrated for winter flounder in Table 3. The highest percentages of abnormalities which were aborted or partially hatched occurred at salinities below 15%. Results of these laboratory experiments indi- cate that successful incubation of embryos oc- curred over a temperature range which exceeded normal spawning season temperatures by as much as 10°C, but coincide quite closely with natural observations for salinity, although there is a shift in survival toward slightly higher salinities than would have been expected. It is possible that the adults, while being held in the laboratory, were conditioned to slightly higher salinities than would have been encountered in a spawning migration into estuaries. This might explain the differences between natural popula- tions and results of laboratory experiments. Most winter flounder populations move to in- shore and estuarine waters to spawn (Perlmutter 1947; Bigelow and Schroeder 1953; Saila 1961), but there are also spawning populations that re- main in offshore shoals (Bigelow and Schroeder 1953; Marak et al. 1962). Field observations in two estuaries of Narragansett Bay and in the Bay itself indicate that spawning occurs at salinities ranging from 11 to 32%. Plankton tows taken in upper Chesapeake Bay produced one egg in 20% with maximum numbers of larvae occurring be- tween 6 and 14% (Dovel 1971). Salinities in sus- pected offshore shoal spawning areas range from 32 to 35.5%, at the bottom (Bumpus 1973), so an overall spawning range from 5 or 6 to 35.5% is indicated for natural populations. The normal temperature range for spawning is 0° to 3.3°C with maximum temperatures for any appreciable egg production and spawning being 4.2° to 5.6°C (Bigelow and Schroeder 1953). Since the eggs are demersal and adhesive, they are not subject to transport into areas of unsuitable temperatures; being estuarine, they are subjected instead to 57 FISHERY BULLETIN: VOL. 74, NO, 1 changes in salinity. However, the euryhaline properties of the eggs insure successful incuba- tion and larval development in a constantly vary- ing salinity environment. ACKNOWLEDGMENTS I thank John Green and Geoffrey C. Laurence for their review of the manuscript, and Geoffrey C. Laurence for his assistance on the statistical analyses. Technical assistance given by Thomas Halavik and Alphonse Smigielski is gratefully acknowledged. The illustrations were prepared by Merrie Marsh and Lianne Armstrong. LITERATURE CITED Alderdice, d. F., and C. R. Forrester. 1968. Some effects of salinity and temperature on early development and survival of the English sole iParo- phrys vetulus). J. Fish. Res. Board Can. 25:495-521. 1971a. Effects of salinity and temperature on embryonic development of the petrale sole Eopsetta jordani. J. Fish. Res. Board Can. 28:727-744. 1971b. Effects of salinity, temperature, and dissolved oxygen on early development of the Pacific cod (Gadus macrocephalus). J. Fish. Res. Board Can. 28:883-902. Battle, H. I. 1930. Effects of extreme temperatures and salinities on the development of Enchelyopus cimbrius (L.). Con- trib. Can. Biol. Fish., New Ser., 5:107-192. BIGELOW, H. B., AND W. C. SCHROEDER 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53, 577 p. BUMPUS, D. F. 1973. Continental Shelf, arms of the sea. In Coastal and offshore environmental inventory. Cape Hatteras to Nantucket Shoals. Saul B. Saila, co-ordinator. Univ. R.I., Mar Publ. Ser. 2, p. 1-1 - 1-46. COLTON, J. B., Jr., and R. R. STODDARD. 1973. Bottom-water temperatures on the Continental Shelf, Nova Scotia to New Jersey. U.S. Dep. Commer., NOAA Tech. Rep. NMFS CIRC-376, 55 p. DOVEL, W. L. 1971. Fish eggs and larvae of the upper Chesapeake Bay. Univ. Md. Nat. Res. Inst., Spec. Rep. 4, 71 p. FORRESTER, C. R., AND D. F. ALDERDICE. 1966. Effects of salinity and temperature on embryonic development of the Pacific cod (Gadus macrocephalus). J. Fish. Res. Board Can. 23:319-340. HOLLIDAY, F. G. T. 1965. Osmoregulation in marine teleost eggs and larvae. Calif Coop. Oceanic Fish. Invest. Rep. 10:89-95. 1969. The effects of salinity on the eggs and larvae of teleosts. In W. S. Hoar and D. J. Randall (editors), Fish physiology. Vol. l,p. 293-311. Academic Press, N.Y. HOLLIDAY, F. G. T., AND J. H. S. BLAXTER. 1960. The effects of salinity on the developing eggs and larvae of the herring. Mar. Biol. Assoc. U.K. 39:591-603. MARAK, R. R., J. B. COLTON, JR., AND D. B. FOSTER. 1962. Distribution of fish eggs and larvae, temperature, and salinity in the Georges Bank-Gulf of Maine area, 1955. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 411, 66 p. MCMYNN, R. G., AND W. H. HOAR 1953. Effects of salinity on the development of the Pacific herring. Can. J. Zool. 31:417-432. PEARCY, W. G. 1962. Ecology of an estuarine population of winter flounder Pseudopleuronectes americanus (Walbaum). Bull. Bingham Oceanogr. Collect., Yale Univ. 18(l):5-78. perlmutter, a. 1947. The blackback flounder and its fishery in New England and New York. Bull. Bingham Oceanogr. Collect., Yale Univ. ll(2):l-92. Saila, S. B. 1961. The contribution of estuaries to the offshore winter flounder fishery in Rhode Island. Proc. Gulf Caribb. Inst., 14th Annu. Sess., p. 95-109. Scott, w. c. m. 1929. A note on the effect of temperature and salinity on the hatching of eggs of the winter flounder (Pseudo- pleuronectes americanus, Walbaum). Contrib. Can. BioL 4(11):137-141. Smigielski, A. S., and C. R. Arnold. 1972. Separating and incubating winter flounder eggs. Prog. Fish-Cult. 34:113. Steel, R. G. D., and J. H. Torrie. I960. Principles and procedures of statistics with special reference to the biological sciences. McGraw-Hill Co., N.Y., 481 p. 58 REEVALUATION OF FISHING EFFORT AND APPARENT ABUNDANCE IN THE HAWAIIAN FISHERY FOR SKIPJACK TUNA, KATSUWONUS PELAMIS, 1948-70 Richard N. Uchidai ABSTRACT Catch per effective trip, used in 1948-64 as an index of apparent abundance of skipjack tuna, Kat- suwonus pelamis , in Hawaiian waters, is biased because effective trip, defined as one on which fish were caught, underestimates effort. Catch per day fished, calculated from data collected in 1965-70, is a refined index because effort includes days with or without catches. This paper describes the existence of a linear relationship between catch per effective trip and catch per day fished in 1965-70, and a method of estimating the latter from the former in 1948-64 based on this relationship. Fishing intensity, which was measured by standard effective trips in past studies, is calculated in standard days fished. Changes in catch per standard day fished are not associated with changes in relative fishing intensity. Skipjack tuna abundance in Hawaiian waters, therefore, is fishery independent and is probably influenced by availability and strength of year classes. In the study of the dynamics of any exploited fish population, data on commercial catch and fishing effort can be interpreted in a number of ways, giving various estimates of apparent abundance. The ultimate objective, however, is to obtain the best possible estimate of apparent abundance. Prior to 1965, studies on catch and effort statis- tics in the Hawaiian pole-and-line fishery for skipjack tuna, Katsuwonus pelamis, defined fishing effort as a "productive" or "effective" trip, that is, one in which skipjack tuna were caught (Yamashita 1958; Shippen 1961; Uchida 1966, 1967). Effective trip underestimated the actual amount of fishing pressure, but it was used be- cause catch report forms used by the fishermen in 1948-65 provided no spaces for recording zero- catch trips. Zero-catch trips should be considered as effort expended to catch fish because they include time spent searching for schools of fish. But the rela- tive importance of search and fishing time de- pends on type of gear used. Gulland (1969) used whaling as an example of a fishery where the im- portant measure was time spent searching, the gear being operational only for a few minutes. The other extreme was bottom trawling, where the important measure was time spent catching fish with the gear on the bottom and searching ^Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, Honolulu, HI 96812. Manuscript accepted May 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 1976. was minimal. Beverton and Parrish (1956) suggested that where searching time is impor- tant, the gear may have to be regarded as being engaged in searching for fish but giving no catch until a school is encountered. For pole-and-line fishing, where much time is devoted to searching for schools of fish, Shimada and Schaefer (1956) used the day spent on the grounds as the basic unit of fishing time. Catch reports of 1965-70 were used to obtain two indices of skipjack tuna apparent abundance: catch per effective trip (C/ET), calculated from data on trips with catches, and catch per day fished (C/DF), calculated from total days fished including zero-catch fishing days. The purpose of this study is to determine whether a relationship exists between C/ET and C/DF. The importance of the relationship is that it affords a means of converting C/ET to C/DF for 1948-64, those years for which no data on C/DF exist but for which good C/ET information is available. A corrected measure of apparent abundance, derived from standard days fished instead of standard effective trip, is used to estimate the relative fishing inten- sity in 1948-70. COLLECTION OF DATA Data on skipjack tuna catch and fishing effort were obtained from the Hawaii State Division of Fish and Game, which collects fish catch statis- tics in the Hawaiian Islands. In addition, catch 59 FISHERY BULLETIN: VOL. 74, NO. 1 and effort data were also collected routinely at the cannery by personnel of the Honolulu Laboratory, National Marine Fisheries Service. The cannery records, however, were deficient in that they did not provide information on vessels not returning to Kewalo Basin, where the can- nery is located, on vessels based on neighboring islands, or on the area of operation. Catch Reports of 1948-64 The forms for reporting skipjack tuna catch have been revised several times over the years. Essentially, all the different versions used in 1948-64 had spaces for recording the date of land- ings, the amount of skipjack tuna landed, and the area fished. The date of landing represented an effective trip that may have lasted from one to several days. Because Hawaiian vessels have limited cruising range, a trip usually lasts 1 day. Studies of interview data collected in 1960 showed that of 329 effective trips, 315 or 96% lasted 1 day (Uchida 1967). Catch Reports of 1965-70 The catch report forms of 1965-70 provided spaces for recording not only the amount of skip- jack tuna caught and the area fished, but also the date of each day spent on the fishing ground, a zero catch when no fish was caught, and the number of men aboard per trip. Each entry repre- sented 1 day's fishing. In using data for these years, therefore, days with catches were assumed to be equivalent to effective trips. The sum of days with and without catches was taken as the total number of days fished. Reporting of Zero-Catch Trips Review of catch reports and cannery records for 1965-70 showed that some vessels occasionally failed to report zero-catch fishing days. When the number of zero-catch trips recorded in the can- nery records exceeded that reported in the catch reports, the difference was assumed to be the number of unreported zero catches. Most vessels reported more zero catches in the catch reports than were recorded in the cannery records; pre- sumably, trips were not recorded at the cannery when a vessel did not return to home port. These catch reports were assumed to be accurate. Not all unreported zero-catch days were ac- counted for. In a few cases, vessels failed to indi- cate a zero catch in the catch report after an un- successful day of fishing and also failed to return to Kewalo Basin, site of the cannery and home port of the Honolulu-based fleet. Then, neither the catch report nor the cannery record showed the effort expended. For Honolulu-based vessels, unreported zero- catch days in 1965-70 varied between 0.5 and 3.8% of the estimated annual number of days fished (Table 1). Differences between reported and estimated number of days fished were not significant it = 1.020; df = 5;P = 0.36); therefore the few zero-catch days that went unreported should not seriously affect the data in this study. Table l. — Total days fished as reported, estimated number and percentage of zero-catch days not reported, and esti- mated total days fished by Honolulu-based Hawaiian skipjack tuna fishing vessels, 1965-70. Total days fished Estimated zero-catch Estimated total as reported days not reponea days fished Year (Number) Number Percent (Number) 1965 1,938 10 0.5 1.948 1966 1,773 39 2.2 1,812 1967 1,678 67 3.8 1,745 1968 1.923 42 2.1 1,965 1969 1,469 54 3.5 1,523 1970 1,605 51 3.1 1,656 SOURCES OF VARIABILITY IN FISHING POWER AMONG VESSELS Fishing power is usually calculated on the basis of a physical feature of the vessel such as gross tonnage or engine horsepower. Differences in fishing power, however, are certainly more complicated than a comparison of these physical attributes. Rothschild (1972) stated that "A con- siderable portion of the variability in fishing power among fishing units can be attributed to variability in skill of the fishing skipper." Fishing skill cannot be measured easily, but its influ- ence on the fishing power of the vessels should be understood. Variability in crew size from trip to trip also complicates the comparison of fishing power among the vessels. For example, catch reports showed that crew size in 1970 varied between 5 and 11 men per trip. Frequently, small vessels were fully crewed while large vessels operated shorthanded. The result was that some of the small vessels were outperforming the larger ones in some years. 60 UCHIDA: REEVALUATION OF FISfflNG EFFORT ANALYTICAL PROCEDURES In the sections that follow, the procedures used in grouping vessels and fishing areas and in treating the data are discussed. Classes of Vessels The difficulties that arise from differences in fishing power among the vessels may be reduced by separating them into relatively homogeneous classes, using physical features such as gross tonnage. It is convenient, therefore, to determine which of the physical features of the vessels is, on the average, proportional to fishing power, and to use it to group the vessels into classes. In a study covering the period 1952-62, the ves- sels were grouped into two size classes according to their bait-carrying capacities. Class 1 vessels had capacities up to 3,000 liters per baitwell whereas class 2 vessels had capacities greater than that (Uchida 1967). But the ability of class 2 vessels to catch more fish than class 1 vessels is not necessarily a permanent characteristic. Al- though baitwell capacity was a good measure of fishing power in the 1952-62 study, it did not reflect fishing power of the vessels satisfactorily after 1962. In 1963-70, some vessels with small bait capacities had catch rates as high as or higher than those with larger capacities. Reevaluation of the data showed that gross ton- nage provided a better approximation of vessel performance. CIET and bait capacity were corre- lated significantly in 8 out of 11 yr in 1952-62, but only in 2 out of 8 yr in 1963-70 (Table 2). Correla- tion between CIET and gross tonnage, on the other hand, was significant not only in 8 yr in 1952-62, but also in 6 yr in 1963-70. For this study, therefore, vessels of 27 to 44 gross tons were called class 1 and those of 45 to 77 gross tons were called class 2. The selection of the division point between class 1 and class 2 vessels was based on the tendency of CIET, when plotted against gross tonnage, to be closely grouped among class 1 vessels for almost all the years examined. In contrast, CIET of class 2 vessels varied widely in most years. The relationship of fishing power to vessel age and to bait usage cannot be overlooked. Among 8 class 1 vessels fishing in 1963-70, only 1 was built after World War II whereas 9 out of 12 class 2 vessels fishing in 1963-70 were built after the war. The relative comfort and reliability of most Table 2. — Correlation coefficients of CIET on baitwell capacity and on gross tonnage of Hawaiian skipjack tuna fishing vessels, 1952-70. A single asterisk denotes probabilities be- tween 0.05 and 0.01; two asterisks denote probabilities equal to or less than 0.01. Correlation coefficient of Correlation coefficient Year df CIET on baitwell capacity of CIET on gross tonnage 1952 23 0.326 0.387 1953 23 0,306 0.275 1954 24 0.602" 0.463* 1955 26 0.498" 0.490- 1956 24 0.390* 0.318 1957 23 0.461- 0.457* 1958 21 0.625" 0.678" 1959 18 0.721" 0.669" 1960 19 0.477- 0.464* 1961 19 0.462- 0.499* 1962 17 0.356 0.528* 1963 18 0.703" 0.757** 1964 18 0.403 0.596** 1965 17 0.368 0.327 1966 15 0.400 0.531* 1967 15 0.593- 0.521* 1968 14 0.434 0.529* 1969 13 0.382 0.516* 1970 13 0.510 0.447 class 2 vessels undoubtedly accentuated the rela- tion between fishing power and tonnage by at- tracting better captains and fishermen. Also, the difference between vessel classes in the amount of bait used was pronounced. Whereas class 1 ves- sels used an average of 8.3 buckets of bait per day fished, class 2 vessels averaged 12.3 buckets. Each year in the Hawaiian fishery the same few vessel captains vie for the distinction of being captain of the "top boat." Variability in skill among captains, therefore, complicated the com- parison of fishing power among vessels. Further- more, captains and crew frequently shifted from one vessel to another, taking their fishing skills with them. In 1965-70, for example, a minimum of nine vessels changed captains and the transfer of a highly regarded captain usually involved the transfer of part of his former crew. The shifting of personnel caused some high-producing vessels to become low- or marginal-producers. Fishing Areas After the establishment of the vessel classes, the data within each size class were then grouped into inshore and offshore fishing areas. In the Hawaiian fishery, the deployment of fishing effort and the resulting catches are recorded according to a statistical area system that was established for Hawaiian waters by the Hawaii State Divi- sion of Fish and Game in 1947 (Uchida 1970). Basically, three general areas are recognized. The 61 FISHERY BULLETIN: VOL. 74, NO. 1 first extends from the coastline to just outside the reef, a distance of about 4 km, and the second extends from 4 to 37 km. Combined and called inshore for this study, these two areas are made up of relatively small statistical areas of unequal sizes. It has been estimated that about 80% of the effort and 75% of the skipjack tuna catch are con- centrated within these areas (Uchida 1967). Beyond 37 km is the third area, called offshore here; the statistical divisions within it are large and nearly equal in size. The inshore fishing ground, restricted to waters within 37 km of the coastline, covered roughly 69,000 km^. The offshore ground, on the other hand, was restricted only by the range of the ves- sels, and varied from year to year. In 1948-65, the vessels covered 111,000 km^ in their offshore fishing, but many distant offshore areas were vis- ited in only 1 or 2 yr over this period. The offshore areas visited most frequently totaled roughly 69,600 km2. Comparison of Catch Per Effective Trip and Catch Per Day Fished The monthly catches of skipjack tuna in 1965- 70, separated into inshore and offshore areas within each vessel size class, were divided by two different units of effort. One was the number of days with catches, which was assumed to be equiv- alent to effective trips; and the index derived was CIET. The other v/as the total number of days fished, which included days of fishing with and without catches; and the index was CIDF. The assumption that days with catches was equiv- alent to effective trips appears justified; Uchida (1967) showed that 96% of the effective trips lasted 1 day. Figure 1 illustrates the relationship of the monthly CIDF (Y) against CIET (X) calculated for class 1 and class 2 vessels fishing the inshore and offshore areas in 1965. The least squares re- gression of y on X resulted in a close linear fit with the regression line having an angle of 45°. A good fit between CIET and CIDF can be ex- pected because both indexes are small when fishing is poor and large when fishing is good. In Hawaiian waters, periods of high tuna apparent abundance are characterized by the presence of larger schools and more frequent encounters be- tween vessels and fish schools (Uchida and Sumida 1971). o a. ^ 3 o 5 2 1 : 'o™o're CLASS , VESSELS :;3Tp"s°ho'reC^*SS 2 VESSELS V /k 1 + 0.994 X / /o * / 2 3 4 5 6 CATCH /EFFECTIVE TRIP ( METRIC TONS ) Figure l. — Relationship between catch per effective trip and catch per day fished of Hawaiian skipjack tuna vessels, by areas fished, January-December 1965. Homogeneity of Data At the outset of the study, it was decided that one regression equation should be calculated for each area within the size classes. The resulting equations could then be used to estimate CIDF from CIET for 1948-64. The decision to calculate one equation for each area by pooling the data for 1965-70 is appropriate, because the data included those years for which skipjack tuna catches from Hawaiian waters were the lowest (1969) and highest (1965) on record. Including data from these 2 yr should provide sufficient low and high values to determine accurately the slope and level of each regression line. Pooling is appropriate when the samples are homogeneous; therefore, it was necessary to test the hypothesis of homogeneity. Statistical testing of the data, discussed in the following sections, was confined to only one index, CIET, because of the close association between CIET and CIDF. The tests for homogeneity showed that yearly variances of inshore CIET among class 2 vessels differed significantly (x^ = 11.92; df = 5;P<0.05). A plot of the yearly means and standard devia- tions, shown in Figure 2A, indicated that they were significantly correlated (r = 0.883; df = 22; P<0.01). Furthermore, the distribution of CIET was skewed because of many low and few high 62 UCHIDA: REEVALUATION OF FISHING EFFORT 2.0 1 : ' z o 1.5 < > O 1.0 o IT < o ^ 05 (A) UNTRANSFORMED C/ET r = 0883 1 at =22. c UJ o Q 01 K < o INSHORE • OFFSHORE i INSHORE » OFFSHORE CLASS I VESSELS CLASS 2 VESSELS o g (B) LOG-TRANSFORMED DATA BEFORE ELIMINATION r ! -0 458 . dt = 22, p<0 05 o 03 0.2 0.1 -0.1 • A A A a ' • fig ° ° o . ' A o H A (C) LOG -TRANSFORMED DATA AFTER ELIMINATION " f ^ 085. d( = 22 , p>0 05 0.1 0.2 LOG MEAN 03 04 05 Figure 2. — Relationship between mean and standard devia- tion of catch per effective trip, before and after logarithmic transformation and elimination, by vessel size classes and areas, 1965-70. values. Because the application of routine statis- tical procedures requires a normal distribution and independence of the mean and standard de- viation, a transformation of the data was re- quired. A logarithmic transformation was selected because the standard deviations tended to be proportional to their means (Figure 2A). Transformation of the Data A logarithmic transformation has several theoretical advantages in analyzing catch data (Murphy and ElHott 1954; Gulland 1956). Usually the transformation tends to stabilize the var- iances and make them independent of the mean. Furthermore, the random components tend to be independently and normally distributed about zero mean and with a common variance. After the transformation, the means and stan- dard deviations continued to be significantly but negatively correlated (r = -0.458; df = 22; P<0.05). Examination of the transformed data revealed that there were two points (Figure 2B) that were aberrant and diverged from the cluster of other points. These points represented data for class 1 vessels fishing offshore in 1969 and in- shore in 1970. The original monthly catch data showed that the catch rates were affected by very low C/ET, all of which were 0.15 MT (metric ton) or less. These catch rates fell close to or beyond IJL±3cr and their elimination from subsequent analysis reduced the correlation between the means and standard deviation (Figure 2C) and stabilized the variances (r = 0.058; df = 22; P>0.05). Tests for homogeneity of variances also indicated that the transformed data for all years could now be grouped by areas within size classes. Figure 3 shows the frequency distribution and fitted normal curve of the deviations from the mean of log C/ET for each area within the size classes. None of the histograms departed sig- nificantly from normality when chi-square tests were applied. Therefore, the fit of the normal curve is as good as can be expected (x^ ranged from 2.18 to 7.59; P<0.05). 30 > z UJ o UJ oz u. 20- 10 INSHORE — ' — • — I — I — t — r OFFSHORE X'= = 2I8 DF=3 P = 054 1 1 — I — r k CLASS I -0.8 -04 +04 +08 >- u z UJ o UJ IT CLASS 2 +04 +0 8 -08 -04 +04 +08 DEVIATION FROM MEAN LOG C/ET Figure 3. — Frequency distribution and fitted normal curve of the deviations from the mean of log C/ET. 63 FISHERY BULLETIN: VOL. 74, NO. 1 Differences in Log Catch Per Effective Trip Between Vessel Classes, Between Areas, and Among Years A factorial analysis of variance in a ran- domized complete-block design was used to test whether significant differences occurred in log CIET between vessel classes (blocks), and be- tween areas and among years (main treatment effects). The analysis showed that log CIET with respect to the two vessel classes differed sig- nificantly {F = 12.34; df = 1 and 265; P<0.01). Significant differences in log CIET also occurred with respect to inshore and offshore areas fished {F = 9.38; df = 1 and 5;P<0.05). Furthermore, the results showed significant differences occurred among years fished {F = 9.45; df = 5 and 5; P<0.05). A Duncan multiple-range test (Steel and Torrie 1960), wath Kramer's (1956) extension of the test, determined that a significant differ- ence in the means occurred primarily between 1965 and 1969, years in which there were consid- erable differences in fishing conditions. Relation Between Log Catch Per Day Fished and Log Catch Per Effective Trip Log CIDF increased linearly with log CIET in each of the areas within the size classes. Regres- sion lines, fitted to the data pooled for 1965-70, showed that the scatter about the regression lines was relatively narrow; there were, however, a few observations in each set of data that appeared to have large residuals. To assess the validity or ap- propriateness of the least-squares fitting of log CIDF on log CIET, these residuals were analyzed. Figure 4 shows the scatter diagrams in which the residuals were plotted against log CIET for the four sets of data. With the exception of a few outliers which can be seen as isolated points with extreme negative ordinates, there were no noticeable peculiarities in the distribution of the residuals. The outliers were rejected at a multiple of the standard deviation using a premium of 2.5% (see Anscombe and Tukey 1963). The overall distribution of the residuals after the rejection procedure appeared in the form of a horizontal band, which indicated that the least- squares analysis of the log transformed data was satisfactory. After the rejection of large residuals, regres- sion lines were fitted to the data as shown in Fig- ure 5. The dashed lines on either side of the re- +0 3 +0.2 +0.1 -0.1 -02 -03 +0 3 +0 2 +0 1 3 -0 2 I CLASS I (INSHORE) REJECTED I I I I I I I L >s I 1 1 I e o o o o « s 1 ) I I 1 I I I CLASS 1 (OFFSHORE) o o "■rejected 111 5 ^^y'°' o 1 1 [ 1 1 1 1 1 1 I 1 1 1 I 1 I 1 I I I CLASS 2 (INSHORE) o • °<>°oo f «^°*o °%°a° r. o o - e o o ao o ft « ** 1 *^R?JECTED I 1 I 1 1 1 1 CLASS 2 (OFFSHORE) - - o o °8 ° °°°»>i;V;*°% „= „ ~ o o rf. o"*" „ ° - o ° oo o o - »■> o e - 1 1 1 1 1 .•-REJECTED 1,1, -03 -0 4 g +03 - +0 2 < +0.1 o Ui "= -0 I h -02 -03 +03 +0 2 +0 I -0.1 -0.2 -0.3 -QA -0.4 -02 +0 2 +0 4 +0 6 +0 8 LOG C/ET FIGURE 4.— Plots of residuals (log C/DF - log CrDF) against log C/ET for class 1 and class 2 vessels fishing inshore and offshore in 1965-70. gression lines indicate the 95% confidence limits for the estimates of log CIDF. The values of the regression equation and correlation coefficient of log CIDF on log CIET are given in Table 3. Substitution of values of log a and b into the logarithmic equation logioC/DF = logioa + blogioC lET and solution of the equation provided estimates of CIDF from CIET, by month, for Table 3. — Data on the regression and correlation of \ogioC/DF on logioC/£r in the Hawaiian skipjack tuna fishery, by vessel size classes and areas, 1965-70. Two asterisks denote probabilities equal to or less than 0.01. Vessel size class Area Log „a b r df 1 Inshore -0.11566 1.13915 0.963" 68 Offshore -0.12549 1 .08370 0.954" 64 2 Inshore -0.10342 1.13340 0.976" 69 Offshore -0.12268 1.13120 0.968" 66 64 UCHIDA: REEVALUATION OF FISHING EFFORT INSHORE 09 OFFSHORE 08- 07- 0.6- 0.5- 04- 03- 02- u. 1- a ■V <-> 0- o o -I -0 I - -0.2- -0.3- -0.4- -0.5- -0 6- -0.7- T 1 1 r CLASS I NOT INCLUDED IN REGRESSION -08 09 0.8 0.7 0.6 05 0.4 0.3 0.2 u. 1 o o o o -" -0.1 -0 2 -0.3- -0.4 -0.5- -0 6 -0.7- -0 8 J L ^NOT INCLUDED IN REGRESSION -0.4 -0.3 -02 -0 1 1 02 03 04 05 06 07 08 09 -0.4 -0.3 -02 -0 1 1 0.2 3 04 0.5 0.6 07 0.8 09 CLASS 2 NOT INCLUDED 'IN REGRESSION J 1 L NOT INCLUDED IN REGRESSION J I I I I I I L- I I -0 4 -0 3 -0.2 -0 1 1 2 0.3 0.4 0.5 0.6 0.7 8 0.9 -0.4 -0.3 -0.2 -0.1 1 2 3 04 5 6 07 0.8 LOG C/ET LOG C/ET Figure 5. — Regression of log C/DF on log C/ET for class 1 and class 2 vessels fishing inshore and oflfshore in 1965-70. 0.9 65 FISHERY BULLETIN: VOL. 74, NO. 1 Table 4. — Estimating the number of days fished among class 1 vessels fishing in the inshore area, January-December 1948. Effective Calculated Estimated Catch trips CIET CIDF days fished Month (MT) (No.) (MT) Log,oC/£7 Log,oC/Df (MT) (No.) January 205.48 77 2.66857 0.42627 0.36993 2.34388 88 February 108.87 73 1.49137 0.17358 0.08207 1.20803 90 March 59.33 72 0.82403 -0.08405 -0.21141 0.61458 96 April 76.91 99 0,77687 -0.10965 -0.24057 0.57468 134 May 133.94 119 1.12555 0.05136 -0.05714 0,87669 153 June 285.80 154 1.85584 0.26854 0.19024 1.54970 184 July 352.30 147 2.39660 0.37959 0.31675 2.07374 170 August 239.72 120 1 99767 0.30052 0.22668 1.68531 142 September 191.07 104 1.83721 0.26415 0.18525 1.53199 125 October 101.31 81 1.25074 0.09716 -0.00497 0.98861 102 November 49 59 44 1 1 2704 0.05194 -0,05649 0,87802 56 December 19 26 25 0.77040 -0.11328 -0.24470 0.56923 34 Total 1,823.58 1,115 1,374 1948-64. For example, Table 4 shows the data used in the computations and the results obtained among class 1 vessels fishing the inshore area in 1948. CIET was derived from the equation, CIET (col. 3) Monthly catch (col. 1) Number of effective trips (col. 2) and converted to logarithms (col. 4). Log CIDF (col. 5) was derived from the equation, log CIDF = log a + 6 log CIET and converted to CIDF (col. 6). Days fished were estimated from the equation. Days fished (col. 7) = Monthly catch (col. 1) CIDF (col. 6) Standardization of Catch Per Day Fished A method of standardizing effort of different size classes of vessel has been discussed by Shimada and Schaefer (1956) for the eastern Pacific yellowfin and skipjack tuna fishery. I used a similar method to estimate relative fishing power of class 1 vessels in the Hawaiian fishery so that their unit of effort was comparable to that of class 2 vessels, which were selected as the stan- dard size class (Uchida 1966, 1967). Briefly, the method involves the use of correction or efficiency factors that are calculated from CIDF of the ves- sel size classes. Efficiency factors adjust the fishing effort of one size class to that of a standard class. For example, under conditions of equal abundance, the class 1 vessels can be expected to produce a smaller catch than the class 2 vessels. From the catches of the two classes, the fishing power of class 1 vessels can be determined rela- tive to class 2, the standard class, for a given fishing area. To illustrate the calculation of efficiency factors and the standard unit of effort, the annual CIDF given in Table 5 by vessel size classes and areas were used. In 1948, the efficiency factor for class 1 vessels fishing inshore was 1.33/1.78 = 0.747 and for offshore was 2.07/3.46 = 0.598. The efficiency factors for class 2 vessels were fixed at 1.000 for all years. The mean efficiency factor, 0.668, is the geometric mean of the inshore and offshore val- ues. The geometric mean is appropriate for av- eraging ratios. Varying from 0.59 to 0.82 (rounded) and av- eraging 0.71 in 1948-70, the efficiency factors demonstrated not only the greater capability of class 2 vessels, but also the wide variability of the factors from year to year. There was no evidence that the efficiency of class 1 vessels increased or decreased relative to class 2 vessels. Therefore, neither the efficiency of the standard class nor that of class 1 vessels has been altered by the loss of the less efficient or marginal vessels. MEASURES OF APPARENT ABUNDANCE AND FISHING INTENSITY Estimate of the apparent abundance of skipjack tuna on the fishing grounds, expressed as catch per standard day fished (CISDF), can be calcu- lated from efficiency factors and the total number of days fished for each of the two classes of ves- sels. For example, in 1948 there were an esti- mated 1,444 fishing days among class 1 vessels and 829 days among class 2 vessels. The standard days fished is the sum of the products of the mean efficiency factor and the total number of fishing days of the size classes. CISDF is found by, 66 UCHIDA: REEVALUATION OF FISHING EFFORT Table 5. — Catch per day fished inshore and offshore among class 1 and class 2 vessels, class 1 efficiency factors, and their geometric mean, 1948-70. Inshore Offsfiore Efficiency Efficiency Geometric Year Class 1 Class 2 factors Class 1 Class 2 factors mean 1948 1.33 1.78 0.747 2.07 3.46 0.598 0.668 1949 1.56 2.24 0.696 2.54 4.12 0.616 0.655 1950 1.34 1.74 0.770 2.10 3.38 0.621 0.692 1951 1.64 2.59 0.633 2.60 3.58 0.726 0.678 1952 1.31 1.66 0.789 1.31 2 19 0.598 0.687 1953 1.53 1.98 0.773 2.37 2.69 0.881 0.825 1954 1.36 2.54 0.535 2.89 3.80 0.760 0.638 1955 1.39 1,99 0.698 2.08 2.32 0.896 0.791 1956 1.90 2.36 0.805 2.30 3.27 0.703 0.752 1957 1.18 1 63 0.724 1.28 1.61 0.795 0.759 1958 1.17 1.87 0.626 1.79 2.36 0.758 0.689 1959 1.97 3.03 0.650 2.37 2.91 0.814 0.728 1960 1.32 2.02 0.653 1.94 2.40 0.803 0.727 1961 1.82 2.37 0.768 2.42 4.05 0.598 0.677 1962 1.49 2,45 0.608 2.22 3.43 0.647 0.627 1963 1.17 1.77 0.661 1.87 3.55 0.527 0.590 1964 1.40 1.69 828 2.07 2.90 0.714 0.769 1965 2.39 2.90 0.824 3.32 4.01 0.828 0.826 1966 1.54 1.82 0.846 1.93 2.91 0.663 0.749 1967 1.47 1.84 0.799 1.65 2.31 0.714 0.755 1968 1.57 1.68 0.934 2.04 2.93 0.696 0.807 1969 1.12 1.43 0.783 1.58 2.26 0.699 0.740 1970 1.32 1.74 0.759 1.30 2.36 0.551 0.646 r'/Qnj? TCi + TC 2 fished an av( erage of 86.1 da ^'^^^ - (EF) iDF,) + DF2 1948-58 when their numbers d( where TCi = total catch of class 1 vessels, TC2 = total catch of class 2 vessels, EF = efficiency factor, DFi = days fished among class 1 vessels, and DF2 = days fished among class 2 vessels. In 1948-70, C/SDF of skipjack tuna in Ha- waiian waters ranged from a low of 1.61 MT in 1957 to a high of 3.29 MT in 1965, but no trend with time was discernible (Table 6; Figure 6). Relative fishing intensity is estimated from C/SDF and the total state catch, which includes catches of part-time as well as full-time vessels: Relative fishing intensity — C/SDF where TC^ = total state catch. When examined over the 23-yr period, fishing intensity did not decrease appreciably despite a gradual decrease in the number of vessels fishing from a maximum of 28 in 1951 to 15 in 1970. With a reduction in the fleet, which occurred primarily among the older class 1 vessels, fishing intensity would be expected to decline, but it did not. The reason was that the average days fished per vessel per year increased. Class 1 vessels 10 vessels and 121.2 days in 1959-70 when their numbers further decreased from 8 to 4 vessels (Figure 7). Class 2 vessels have not decreased in number drastically, declining from 14 in 1955 to 11 in 1970. Averaging 86.9 days fished prior to 1964, class 2 vessels subsequently averaged 119.8 days per year. INTERRELATION OF TOTAL CATCH, FISHING INTENSITY, AND APPARENT ABUNDANCE The total catch of skipjack tuna, given in Table 6 and shown in Figure 6, fluctuated with C/SDF in a similar fashion in 1948-70 (r = 0.902; df = 21; P<0.01). For the years studied, then, total catch may be satisfactory as a gross index of changing apparent abundance but may not be suitable in future years because it is obviously sensitive to changes in demand or fishing effort, competition from other fisheries, and economic constraints upon the fishery. Changes in C/SDF are not associated with changes in fishing intensity (r = 0.302; df = 21; P>0.05); therefore, the apparent abundance of skipjack tuna in Hawaiian waters is not influenced by changes in the amount of fishing effort expended, but by fishery-independent fac- tors such as variations in availability, which in turn is related to changes in the fishes' habits or 67 FISHERY BULLETIN: VOL. 74, NO. 1 Table 6. — Total landings in metric tons (MT) of skipjack tuna in Hawaii, catch per stan- dard day fished, relative fishing intensity, catch per standard effective trip, and relative effective fishing intensity, 1948-70. Catch per Relative Catch per Relative standard fishing standard effective fishing Total catch day fished intensity effective trip intensity Year (MT) (MT) (Class 2 days) (MT) (Class 2 trips) 1948 3,802.96 2.01 1,891 2.30 1,653 1949 4.488.23 2.53 1,773 2.85 1.575 1950 4,314.38 1.99 2,161 2.31 1,868 1951 5,863.37 2.93 2,001 3.28 1,788 1952 3,307.58 1.83 1,806 2.15 1,538 1953 5.470.15 2.14 2,552 2.46 2,224 1954 6,360.13 2.81 2,256 3.16 2.013 1955 4,397,43 1.95 2,248 2.26 1,946 1956 5,049.58 2.59 1,946 2.91 1,735 1957 2,780.66 1.61 1,726 1.90 1,464 1958 3,100.15 1.87 1,652 2.18 1,422 1959 5,630 65 2.93 1,919 3.26 1,727 1960 3,338.46 1.99 1.673 2.30 1,452 1961 4,941.66 2.69 1,835 3.01 1,642 1962 4,270.81 2.56 1,665 2.88 1,483 1963 3,67386 2.15 1,712 2.48 1,481 1964 4,093.10 1.98 2,065 2.29 1,787 1965 7,328.96 3.29 2,221 3.54 2,070 1966 4,256.82 2.24 1,896 2.52 1,689 1967 3,646.80 1.99 1,832 2.30 1,586 1968 4,227.41 2.04 2,067 2.32 1,822 1969 2,704,94 1.63 1,658 2.02 1,339 1970 3.334.46 1.89 1,760 2.19 1,523 z o 2 t- o £ to Q. in Q =) ir X o O o 2 2 u> -^ — I — I — I — i — I — 1 — I — I — 1 — I — \ — r TOTAL CATCH CATCH/STANDARD DAY FISHED RELATIVE FISHING INTENSITY 28 2 6 10 Q. (O to < 24 CO o z < o 22 f W z I- z -20 I 8 16 J I 1 1 L. 1950 1955 I960 YEAR 1965 1970 o z I (O UJ > < UJ Figure 6. — Total catch, catch per standard day fished, and the relative fishing intensity for skipjack tuna in Hawaii, 1948-70. in the environment, and to the strength of the year classes. Catch per standard effective trip {CI SET) and relative effective fishing intensity, the two indi- ces used in previous studies (Uchida 1966, 1967, 1970), are also given in Table 6. As expected, both CISDF and CISET fluctuated similarly in 1948- 70 (r = 0.998; df = 21; P<0.01). Likewise the correlation between relative fishing intensity and relative effective fishing intensity was sig- nificant, indicating that changes in one paral- leled changes in the other (r = 0.982; df = 21; P<0.01). It can be concluded that although the use of effective trips in previous studies produced biased results, which deviated from more precise estimates calculated from days fished, its use did 160 n — \ — \ — \ — I — I — \ — r CLASS 2 VESSELS _J \ 1 L_ 1950 1952 1954 1956 1958 I960 1962 1964 1966 1968 1970 YEAR Figure 7. — Average number of days fished per vessel per year among class 1 and class 2 Hawaiian skipjack tuna vessels, 1948-70. 68 UCHIDA: REEVALUATION OF FISHING EFFORT not lead to faulty conclusions about the status of the Hawaiian skipjack tuna fishery. The only serious bias appears to be that fluctuations in the CI SET were slightly exaggerated and those in ef- fective fishing intensity were dampened. SUMMARY The existence of a linear relationship between catch per effective trip and catch per day fished in 1965-70 was described. Based on this relation- ship, catch per day fished was estimated from catch per effective trip for 1948-64. Efficiency factors were used to standardize fishing effort of class 1 vessels to that of class 2. The data showed that in 1948-70, efficiency fac- tors for class 1 vessels remained constant relative to class 2 vessels. Fishing intensity, calculated in standard days fished, did not decline over the 23-yr period despite the gradual decrease in the number of vessels fishing. Data from the catch reports showed that in the face of this decline in fleet size, the remaining vessels increased effort by fishing more frequently. Total catch correlated significantly with C/SDF; therefore, it was a good gross indicator of skipjack tuna apparent abundance. Evidence supported the conclusion that in Hawaiian wa- ters, skipjack tuna apparent abundance was not influenced by changes in the amount of fishing effort expended but by fishery-independent fac- tors. And although effective trips as a measure of fishing pressure in previous studies underesti- mated effort and, therefore, provided a biased estimate of skipjack tuna apparent abundance in the Hawaiian fishery, its use did not lead to faulty conclusions. ACKNOWLEDGMENTS I am indebted to Kenji Ego and Tamotsu Shimizu of the Hawaii State Division of Fish and Game for their time and effort in designing and issuing the revised catch report forms of 1964 from which the basic data for this study were ob- tained. Thanks go also to William H. Lenarz, Gene R. Huntsman, and William Nicholson for reading the manuscript and offering valuable suggestions for its improvement. LITERATURE CITED ANSCOMBE, F. J., AND J. W. TUKEY. 1963. The examination and analysis of residuals. Tech- nometrics 5:141-160. BEVERTON, R. J. H., AND B. B. PARRISH. 1956. Commercial statistics in fish population studies. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 140 (Part I): 58-66. GULLAND, J. A. 1956. On the fishing effort in English demersal fisheries. Fish. Invest. Minist. Agric. Fish. Food (G.B.), Ser. II, 20(5), 41 p. 1969. Manual of methods for fish stock assessment. Part 1. Fish population analysis. FAO (Food Agric. Organ. U.N.) Man. Fish. Sci. 4, 154 p. KRAMER, C. Y. 1956. Extension of multiple range tests to group means with unequal numbers of replications. Biometrics 12:307-310. MURPHY, G. I., AND K. C. ELLIOTT. 1954. Variability of longline catches of yellowfin tuna. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 119, 30 p. Rothschild, B. J. 1972. An exposition on the definition of fishing effort. Fish. Bull., U.S. 70:671-679. Shimada, b. M., and M. B. SCHAEFER 1956. A study of changes in fishing effort, abundance, and yield for yellowfin and skipjack tuna in the eastern tropical Pacific Ocean. Inter-Am. Trop. Tima Comm. Bull. 1:351-469. Shippen, H. H. 1961. Distribution and abundance of skipjack in the Hawaiian fishery, 1952-53. U.S. Fish Wildl. Serv., Fish. Bull. 61:281-300. STEEL, R. G. D., and J. H. TORRIE. 1960. Principles and procedures of statistics: With spe- cial reference to the biological sciences. McGraw-Hill, N.Y., 481 p. UCHIDA, R. N. 1966. The skipjack tuna fishery in Hawaii. In T. A. Manar (editor). Proceedings, Governor's Conference on Central Pacific Fishery Resources, State of Hawaii, p. 147-159. 1967. Catch and estimates of fishing effort and apparent abundance in the fishery for skipjack tuna (Katsuwonus pelamis) in Hawaiian waters, 1952-62. U.S. Fish Wildl. Serv., Fish. Bull. 66:181-194. 1970. Distribution of fishing effort and catches of skip- jack tuna, Katsuwonus pelamis, in Hawaiian waters, by quarters of the year, 1948-65. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 615, 37 p. UCHIDA, R. N., AND R. F. SUMIDA. 1971. Analysis of the operations of seven Hawaiian skipjack tuna fishing vessels, June-August 1967. U.S. Dep. Commer., Natl. Mar. Fish. Serv., Spec. Sci. Rep. Fish. 629, 25 p. Yamashita, D. T. 1958. Analysis of catch statistics of the Hawaiian skip- jack fishery. U.S. Fish Wildl. Serv., Fish. Bull. 58:253-278. 69 SEASONAL AND INSHORE-OFFSHORE VARIATIONS IN THE STANDING STOCKS OF MICRONEKTON AND MACROZOOPLANKTON OFF OREGON William G. Pearcy^ ABSTRACT Dry weights of pelagic animals captured along an inshore-offshore station line with Isaacs-Kidd mid-water trawls and 1-m diameter plankton nets during a 5-yr period provided evidence for seasonal changes in the standing stocks of carnivores. Micronekton catches (fishes, shrimps, and squids) were largest inshore (28 and 46 km offshore) in the winter (November-April), and offshore (84 and 120 km) during the summer (May-October), the season of coastal upwelling. No seasonal difference was detected in the biomass of herbivores, or in its primary components, the copepods and euphausiids. Increased biomass of medusae during the summer resulted in significant seasonal differences in the planktonic carnivores at the inshore stations. The average biomass (grams per square meter) of small nektonic and planktonic carnivores, averaged over the year, peaked at the 84-km station. The biomass of fishes was greater than shrimps and the biomass of shrimps was greater than that of squids at all stations, except 46 km where shrimps predominated. Herbivore biomass was maximal at 46 km, over the inner continental slope, largely because of the high catches of euphausiids at this station. The occurrence of largest average catches at intermediate distances from shore, and inshore-offshore shifts in peak biomass with seasons, may result from seasonal changes in upwelling and downwelling and exclusion of vertical migrants from shoal waters on the shelf Herbivore: carnivore biomass ratios differed significantly between inshore and offshore stations. Standing stocks of herbivores were several times larger than those of carnivores in nearshore waters, but the ratio was about 1.0 in offshore waters. Coefficients of variation (s/x) of herbivore and plank- tonic carnivore stocks for the entire sampling period were highest inshore, indicating high variabil- ity, and decreased markedly in offshore waters. These trends suggest that, compared to offshore or oceanic communities, the pelagic inshore-upwelling ecosystem may be less predictable and have a lower ecological efficiency. This research was designed to answer two ecologi- cal questions about intermediate consumers in the pelagic food chain off Oregon: (1) Are seasonal variations obvious in the standing stocks of small nekton and macrozoo- plankton off Oregon, perhaps in response to up- welling along the coast during the summer? (2) Are there trends in the standing stocks of these animals from oceanic waters into neritic waters and, if so, do they reflect basic ecological differences in these pelagic communities? Pelagic animals such as fishes, squids, shrimps, and euphausiids are ubiquitous in the open oceans and are important intermediates in the food chain between small plankton and large pelagic carni- vores. Yet little is known about their seasonal variations, inshore-offshore distributions, or gen- 'School of Oceanography, Oregon State University, Corval- lis, OR 97331 eral ecology. The life span and generation time of many of these intermediate consumers are 1 yr or greater, limiting short-term changes in population sizes. Moreover, many of these ani- mals reside below the depth of seasonal tempera- ture change much of the time. They may under- take diel vertical migrations, and some species may migrate through the thermocline at night. In any event, seasonal changes in physical environ- ment are expected to be less pronounced than those experienced by inhabitants of surface waters. Thus, seasonal variations in population size of these animals are expected to be less than those of small planktonic organisms. Movements of water may also affect seasonal changes in the abundance of animals at one locality, or spatial distributions within a general region. In areas where water masses and as- sociated pelagic fauna overlap and mix, species structure may be complicated, primarily a result Manuscript accepted April 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 1976. 70 y-^ - ^^ PEARCY: MICRONEKTON AND MACROZOOPLANKTON OFF OREGON of advective processes rather than biological in- teractions (McGowan 1971). In the headwater region of the California Current off Oregon, how- ever, the water type is predominantly Subarctic and common species of some taxonomic groups of pelagic animals are the same within and among years (Pearcy 1972). In addition to in situ population changes and changes affected by advection, small nektonic animals may be able to swim or to migrate horizontally. Though migrations of large nektonic animals such as tuna, salmon, hake, etc., are known to result in large seasonal changes in the abundance of these animals off Oregon, little evidence exists for horizontal movements of mi- cronekton, even on a reduced scale. This is another reason to expect temporal stability of their populations. Basic differences in the structure and energy pathways of neritic and oceanic ecosystems in the northeastern Pacific have been inferred by differ- ences in the seasonal production cycle, seasonal variations in chlorophyll a concentrations, and the size of individual phytoplankton and mi- crozooplankton (McAllister et al. 1960; Anderson 1965; Parsons and LeBrasseur 1970; LeBrasseur and Kennedy 1972). Inshore-offshore differences in the standing stocks of pelagic herbivores and carnivores, which have not been studied, are therefore to be expected. METHODS Micronekton and macrozooplankton were col- lected at night with 1.8-m Isaacs-Kidd mid-water trawls (IKMT) and with 1-m diameter plankton nets (MN) along stations west of Newport, Oreg. (lat. 44°39.1'N). The stations were located 28, 46, 84, 120, and >120 km, respectively, offshore (Fig- ure 1). Collections, made about every month, totalled 243 IKMT tows between August 1962 and July 1967, and 179 MN collections between June 1963 and July 1967. The IKMT had a 5-mm (bar measure) nylon liner throughout. Oblique tows were made to a depth of approximately 200 m, except at inshore stations where about one-half the depth of the water column was sampled (40 m and 130 m at the 28- and 46-km stations, respectively). Tow speed was 6 knots. The trawl was lowered at 50 m wire/min until a 4:1 scope was attained. The trawl was then retrieved at 30 m wire/min to the surface. Volume of water filtered and depth of Figure l. — Location of the sampling stations off Newport, Oreg. Stations are designed in kilometers from the coast. Depth contours are in fathoms (100 fathoms = 183 m, 500 fathoms = 914 m, 1,000 fathoms = 1,829 m, 1,500 fathoms = 2,743 m). trawling was estimated from TSK^ depth- distance recorders and flowmeters. The meter nets, which were made of 0.571-mm Nitex, were towed immediately before or after each IKMT tow. From June to November 1963 oblique tows were made to approximately the same depths as the IKMT tows, but because of difficulties resulting from preferential sampling of near-surface waters, oblique tows were aban- doned in favor of vertical tows in December 1963. Vertical tows were from 200 m to the surface, or from 60 or 150 m to the surface at the two inshore stations. After a vertical wire angle was obtained, they were retrieved at 50 m wire/min. Flow- meters mounted in the mouth of MN's provided estimates of volumes filtered. In a few instances flowmeters malfunctioned. Volumes were then estimated from the distance towed and 85% IKMT filtration efficiency (Pearcy and Laurs 1966) or from the average volume of other MN tows to the same depth. ^Tsurumi-Seiki Kosakusho Co. Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 71 FISHERY BULLETIN: VOL. 74, NO. 1 Samples were preserved with Formalin at sea and sorted into taxonomic groups ashore. Wet (drained) weights were obtained for micronekton (fishes, shrimps, and squids). Micronekton from 32 different IKMT collections were dried to a con- stant weight in a drying oven at 65°C. The mean dry weight: wet weight ratios were then used to convert wet weights of other collections to dry weights. The means and standard deviations of the dry:wet weight ratios were 0.23 ± 0.06 for fishes, 0. 15 ± 0.02 for shrimps, and 0. 1 1 ± 0.04 for squids. Dry weights were obtained for all major taxa sorted from MN samples: euphausiids, copepods, chaetognaths, medusae, amphipods, salps-dolio- lids, and shrimps. These taxa generally com- prised over 95% of the total collection weights. The remainder usually consisted of annelids, pteropods, and heteropods. Ctenophores usually disintegrated in the samples, but when fragments were identifiable they were weighed with the medusae. In this paper, dry weights are used as a measure of standing stock, which is considered to be synonymous with biomass. Sampling Variability Several series of IKMT's at a single station during a single night were taken to assess sam- pling variability. The variability of total micro- nektonic dry weight per 1,000 m^ (Table 1) indi- cates that the variance for these series was ap- preciably less than the mean. These data on total biomass of micronekton, which are not in dis- agreement with the high variability encountered for individual species of micronekton captured in repeated tows at one station (e.g., Pearcy 1964; Ebeling et al. 1970), suggest that most of the temporal fluctuations of biomass illustrated in Table l. — Sampling variability oftotal biomass of micronekton and macroplankton (grams dry weight per 1,000 m^) collected during repeated tows during separate nights. Distance offshore No Average Variance Gear Date (km) tows W (s') Mid-water Dec 1964 84 5 2.7 0.6 trawl Nov. 1966 120 3 4.7 0.9 Feb. 1967 120 5 1.8 0.2 Feb. 1967 120 3 1.5 0.02 June 1967 306 6 1.9 0.01 June 1967 120 6 2.2 0.4 Meter net June 1964 93 6 5.0 3.1 June 1966 93 5 20.3 99.0 Nov. 1966 111 7 9.6 2.4 Feb. 1967 46 3 4.6 1.1 Mar. 1967 787 6 10.0 101.8 Figure 2 are independent of short-term sampling variability. Variances of macrozooplankton biomass from repeated MN tows, on the other hand, were much larger than those for the IKMT (Table 1). In two out of the five series, variance surpassed the mean. Hence, a larger portion of the temporal variability of zooplankton can be ascribed to sampling varia- bility. RESULTS Micronekton Variations of the dry weights of micronekton (fishes, shrimps, and squids) captured per 1,000 m^ are shown in Figure 2 for four stations, 1962-67. Several trends are apparent. Seasonal peaks in the biomass occur inshore at the 28- and 46-km stations during the winter months, with very low values during intervening months. A reversed trend, though less pronounced, is found offshore at the 84- and 120-km stations where maximum catches generally were made during the summer or fall months. Average biomass values appear to be lowest inshore, highest at 84 km, and lower again at 120 km where total variability is the lowest. The spatial peak of micronekton biomass at 84 km is more obvious in Figure 3, where dry weight is plotted per square meter instead of per cubic meter (to compensate for different depths of sam- pling at inshore stations). The standing stocks of fishes were greater than shrimps, and shrimp stocks were greater than squids at all stations except at 46 km where shrimps predominated. The neritic, benthopelagic shrimp, Pandalus jordani, occasionally made up the bulk of the biomass of collections at both 28 and 46 km (Pearcy 1970). However, mesopelagic animals comprised most of the nighttime IKMT catches: mainly the fishes Stenobrachius leucopsarus, Diaphus theta, Tar- letonbeania crenularis, and Tactostoma macropus (Pearcy 1964, 1972; Pearcy and Laurs 1966; Pearcy and Mesecar 1971); the shrimp Sergestes similis (Pearcy and Forss 1966, 1969); and the squids Gonatus spp. and Abraliopsis felis (Pearcy 1965, 1972). Seasonal variations in the total biomass (grams/10 m^) of micronekton are illustrated in Figure 4 for two general seasons: May-October, which includes the upwelling season; and November-April, when surface currents are usu- 72 PEARCY: MICRONEKTON AND MACROZOOPLANKTON OFF OREGON A SONOIJF MAMJ JASONDIJFMAMJ JAS ONOI 1962 1963 I lae^ JFMAMJJASONOIJFMAHJ JASONDIJFMAMJ J 1965 1966 I 196/ Figure 2. — Biomass of micronekton captured in Isaacks-Kidd mid-water trawl collections at four sta- tions, 1962-1967. Each point represents one collection. Average depth of tows was 40 m for 28-km station, 130 m for 46-km station, and 200 m for 84- and 120-km stations. TOTAL DISTANCE OFFSHORE (km) >I20 Figure 3. — Inshore-offshore variations in the average total micronekton biomass (grams per 10 m^ ± 1 SE) and in its component fishes, shrimps, and squids. ally reversed, downwelling occurs, and the David- son Current is often present along the coast (Wyatt et al. 1972; Bakun 1973). The means and medians of the biomass of total micronekton per 10 m^, and of its constituents — fishes, shrimps, and squids — are given in Table 2 for these two sea- sons, along with the probabilities that the two e o >- I20 Figure 4. — inshore-offshore variations in the biomass of mi- cronekton during two seasons, May-October and November- April. Shaded areas included means ± 1 SE. seasonal values are the same. Seasonal differences of total biomass are significant (P<0.05) at 46 73 FISHERY BULLETIN: VOL. 74, NO. 1 Table 2. — The mean and median biomass (grams dry weight per 10 m^) for micronekton and macro- plankton during summer (S = May-October) and winter (W = November- April) at five stations (28, 46, 84, 120 and >120 km) off the Oregon Coast. Probabilities resulting from Mann- Whitney U and t tests of seasonal differences are given. Stn 1. 28 km Stn. 46 km Stn. 84 km Stn. 120 km Stn. > 120 km Item S W S W S W S W S W Total micronekton Mean 0.19 0.32 0.51 4.30 8.20 5.24 6.20 3.26 350 4.30 Median 0.004 0.03 0.38 2.75 7.84 4.04 5.04 2.76 3.28 3.18 Probabilities U NS SW P = 0.08 S>W P- =0.04 NS t NS SW P<0.05 S>W"P<0.01 NS Probabilities U test Fishes t SW"P=0.01 S>W* P- =0.02 NS Shrimps t SWP<0.01 NS NS Total macroplankton Mean 24.9 19.3 31.3 38.7 37.0 15.6 27,4 26.6 11.8 15.7 Median 12.6 12.1 9.4 8.1 12.2 6.5 80 8.6 4.9 5.0 Probabilities U NS NS S>W P<0.04 NS NS t NS NS S >W" P<0.01 NS NS Probabilities ftest Copepods NS NS NS NS NS Euphausllds NS NS NS NS NS Salps t t t t NS Medusae S -W •• P- ;0.01 S>W"P<0.01 S>W P = 0.06 NS NS Chaetognaths NS NS NS NS NS Amphipods NS NS NS NS NS Shrimps t t S W P<0.05 NS NS NS - not significant. t - too many zeros for valid tests. and 120 km using the non-parametric Mann- Whitney U test (Tate and Clelland 1957 ) and at 46, 84, and 120 km using the parametric t test. Mann- Whitney U tests for the three taxa of micronekton indicated significant seasonal differences for standing stocks of fishes at 46, 84, and 120 km, for shrimps at 46 km and for squids at 84 km. Macrozooplankton Values for the biomass of macrozooplankton collected at four stations during 1963-67 are shown in Figure 5 and Table 3. Inshore-offshore and seasonal trends are less apparent than for micronekton. The total MN biomass per 10 m^ is lowest at the 28-km stations, greater at the 120-km stations, and highest at the 46-, 84-, and 120-km stations (Table 3). Of the taxonomic groups composing the MN samples, copepods were most important on an average dry weight basis at all stations except at 46 km where euphausiids were very abundant (Table 3). The standing stock of medusae ranked second after copepods at all stations except at 46 km where it ranked third after copepods. Even though the maximum biomass of all groups oc- curred at 46, 84, or 120 km on a square meter basis, the maximum weights of copepods and Table 3. — Biomass of zooplankton per 10 m^ collected with 1-m diameter nets at the stations off Newport, Oreg. Stn. Stn. Stn. Stn. Stn. Item 28 km 46 km 84 km 120 km : >120km Total biomass Mean 21 36 26 27 14 Median 8.0 15 16 15 10 SD 34 58 27 33 10 No. collections 36 40 41 37 25 Ave. sampling depth 60 152 200 200 200 Copepods Mean 11.9 12.2 7.9 11.7 4.3 Median 2.4 2.0 2 1 2.5 1.5 Euphausllds Mean 2.6 20.0 6.2 2.4 2.5 Median 0.6 3.7 2.2 1.5 1.1 Salps Mean 0.04 0.03 3.2 4.1 1.4 Median 0.002 0.002 Medusae Mean 5.8 2.3 6.8 6.4 3.8 Median 1.2 1,1 3.2 2.5 2.0 Chaetognaths Mean 5 0.9 1.6 1.7 0.9 Median 0.07 0.6 1.0 1.1 0.7 Amphipods Mean 0.07 0.1 0.2 03 0.3 Median 0.02 0.06 0.2 0.2 0.2 Shrimps Mean 0.02 0.6 0.5 0.6 0.8 Median 0.2 0.2 0.1 medusae on a cubic meter basis were found at 28 km, nearest the coast. Differences in the biomass of macrozooplankton between the two seasons were only significant at one station, 84 km offshore (Table 2), although distinct peaks occurred during the summers of 2 yr at 120 km (Figure 5). Surprisingly, most of the taxonomic groups of zooplankton, including copepods and euphausiids, evidenced no seasonal changes at any stations. The only significant 74 PEARCY: MICRONEKTON AND MACROZOOPLANKTON OFF OREGON 100 I I I I I 1 I I I I T 1 1 i I i I l; I I I rr 1 1 1 I ! I I I . . I I I I 1 1 1 v y I •;• I I I I I I I I jjasond'jf' mamj jasond 1963 1964 JFMAMJJASONDJFMAMJJASOND 1965 ' 1966 J F M A M J J 1967 Figure 5. — Biomass of macrozooplankton captured in 1-m diameter plankton nets at four stations, 1963-1967. Each point represents one collection. differences were for medusae, whose standing stocks in the summer exceeded those in the winter at 28 and 46 km (Mann-Whitney U, P<0.01) and perhaps at 84 km (P = 0.06), and for shrimps at 84 km, where again biomass was larger during sum- mer than winter (Table 2). Trophic Groups To estimate seasonal and inshore-offshore vari- ations in the standing stocks of the lower trophic levels of oceanic consumers, the dry weights of the various taxa were combined. Herbivores were assumed to include copepods, euphausiids, and salps-doliolids. Planktonic carnivores included chaetognaths, medusae, amphipods, and shrimps. Nektonic carnivores included fishes, squids, and shrimps. Although it is recognized that some euphausiids and copepods may be carnivorous, the main species captured off Oregon, Euphausia pacifica, Thysanoessa spinifera, and Calanus spp., are considered to be largely herbivorous. Inshore-offshore variations in standing stocks are illustrated in Figure 6. On the average, the biomass of herbivores was greater than planktonic 40 35 ^"£ 30 O ^25 20 X y 15 g 10 h^ PLANKTONIC j CARNIVORE§-.| J._ \ NEKTONIC CARNIVORES — O'- 28 46 84 120 DISTANCE OFFSHORE (km! >I20 Figure 6. — Inshore-offshore variations in the average biomass (± 1 SE) of herbivores, planktonic carnivores, and nektonic carnivores at five stations. carnivores, and the biomass of these organisms was greater than that of micronektonic carni- 75 FISHERY BULLETIN: VOL. 74, NO. 1 vores at all stations. The high catches of herbivores at 46 km were due to abundant concentrations of euphausiids. Both groups of carnivores, on the other hand, had lowest bio- mass at the inshore stations and attained maxima farther offshore. Seasonal variations in the standing stocks of herbivores and planktonic carnivores are illus- trated in Figure 7. Mann-Whitney U tests of differences between the two seasons were not significant (all P>0.1) for any station, providing no evidence for seasonal changes in the biomass of herbivores. The biomass of planktonic carnivores increased wdth distance offshore during the winter and tended to decrease during summer. The biomass at 28 km was higher in summer than winter (P<0.01), largely due to high catches of medusae during the summer. At 84 km, 50 40 30 30 E Q 20 Q. 5 10 a: Q 20- 15- 10- ^ PLANKTONIC CARNIVORES MAY- OCT ■-^NOV-APR >i^ 28 46 84 120 >I20 DISTANCE OFFSHORE (km) Figure 7. — Seasonal variations in the average biomass (± 1 SE) of herbivores (upper) and planktonic carnivores (lower). ^ 20 > O y 10- o PLANKTONIC NEKTONIC ■•■. CARNIVORES'^ 28 46 84 120 DISTANCE OFFSHORE (km) >I20 Figure 8. — Variability in the catches of herbivore, planktonic carnivores, and nektonic carnivores vs. distance offshore. Variability is expressed as coefficients of variation based on dry weights per 1,000 m^. planktonic carnivores also appeared to be more abundant during the summer (P = 0.08), again because of higher catches of medusae. No seasonal differences were apparent at other stations (P>0.1). The ratio of herbivore:carnivore biomass, as expected from the data shown in Figure 6, aver- ages about 2.0 at 28 km and 4.0 at 46 km, but only about 1.0 at the oceanic stations 84, 120, and >120 km. These ratios were ranked among stations for individual cruises. The sum of the ranks for stations were significantly different (P<0.01, Friedman two-way ANOVA by ranks, Tate and Clelland 1957). Thus herbivores pre- dominated over carnivores in inshore waters, whereas the standing stocks of herbivores and carnivores were about equal in oceanic waters 84 km offshore and beyond. No seasonal differences in herbivore:carnivore ratios were found (P> 0.05, Mann- Whitney U tests). As a measure of variability of the standing stocks of trophic groups over the sampling period, coefficients of variation (six) of the catches are plotted for each station in Figure 8. A marked decline in the variability of both herbivores and carnivores takes place from inshore into offshore waters. DISCUSSION Regional Comparisons of Zooplankton Standing Stocks Values for the standing stocks of zooplankton in 76 PEARCY: MICRONEKTON AND MACROZOOPLANKTON OFF OREGON the upper 140 to 300 m are summarized by Gushing (1971) for upwelling regions of the world. The average biomass of zooplankton col- lected within 120 km of the Oregon coast (Table 4) is within the range of values given by Gush- ing, after conversion to displacement volume per 1,000 m^ and to grams carbon per square meter. Zooplankton standing stocks off Oregon can also be compared with those reported by the Galifornia Gooperative Oceanic Fisheries Investigations (GALCOFI) which used 0.25-0.55-mm mesh in nets towed obliquely from 140 m to the surface. Zooplankton displacement volumes near the Ore- gon coast accord with values of Reid et al. (1958) and Reid (1962) greater than 400 cm3/l,000 m^ for July and August 1955 from Point Gonception, Galif , to northern Washington, and with Thrail- kill's (1956) values of 100-900 cm3/l,000 m^ for 1949 and 1950 off Oregon and northern Galifor- nia. Smith's (1971) median displacement vol- umes for pooled areas within 100 miles of shore between Point Gonception and San Francisco Bay, Galif., are 200-400 cm3/l,000 m^ during April-July 1951-60, with decreased volumes south of Point Gonception. Median displacement volumes for Oregon (either on an annual or a summer basis, Tables 2 and 4) are appreciably lower than Smith's values for northern Galifor- nia. This difference may be ascribed to differ- ences between vertical and oblique tows, mesh size, or annual differences in standing stocks. Or, a real trend may exist for the nearshore zoo- plankton standing stocks to increase in the Galifornia Gurrent system between Oregon and northern Galifornia, a trend that may be attrib- uted to the more intense upwelling — and hence higher productivity — that occurs off northern Galifornia (Bakun 1973). Zooplankton volumes within 120 km of Oregon are several times those given by McAllister Table 4. — Dry weight of Oregon zooplankton converted to dis- placement volumes and grams carbon. Stn. Stn. Stn. Stn. stn. Item 28 km 46 km 84 km 120 km >120km Mean cm^/i.OOOm^* 552 450 228 274 85 Median cm^/i.ooo m^ 160 157 140 121 83 Mean gC/m^t 1.1 1.8 1-3 1.3 0.7 Mean gC/m^J 2.3 3.9 2.8 2,9 1.5 "Conversion based on data of Ahlstrom and Thrailkill (1963, Table 7): wet weight plus interstitial water ("displacement volume) « 0.06 = dry weight. tC was estimated to be 50% of the dry weight (see Omori 1969, Table 5). tCalculated using Gushing s (1971) conversion of 0.065 x displacement volume = gC. This conversion assumes that displacement volumes do not in- clude interstitial water, but according to the data of Ahlstrom and Thrailkill (1963, Table 7) an average of 42% of the wet weight of mixed zooplankton is interstitial water. (1961) and LeBrasseur (1965) for oceanic areas of the Gulf of Alaska (0-150 m vertical tows with a 0.45-cm diameter net, 0.35-mesh), even after their catches are adjusted for the relatively low catching power of their net (McAllister 1969; LeBrasseur and Kennedy 1972). Average vol- umes at weather station "P" (lat. 50'^N, long. 145°W) were more similar to those at the station >120 km off the Oregon coast. Increased produc- tivity associated with coastal upwelling along Oregon, therefore, enhances the average zoo- plankton standing stocks out to about 120 km from shore several times above the stocks farther offshore or upstream in the North Pacific Drift (see also Reid 1962). The width of this zone of high zooplankton standing stocks appears to be considerably less than the 200-500 km reported by Gushing (1971) for the region off northern Galifornia. Seasonality of Standing Stocks Seasonality in the biomass of zooplankton, with maxima in the summer and minima in the winter, has been reported in the Galifornia Gur- rent system off central Galifornia (Lasker 1970; Smith 1971) and in waters off the Oregon- Wash- ington coast (Peterson 1972). Yet there was lim- ited evidence for differences in macrozooplankton standing stocks between the two seasons in Ore- gon waters. Thus seasonality of standing stocks appeared to be more pronounced for micronekton than macrozooplankton, or for carnivores than herbivores. This may be because the high vari- ability of macrozooplankton catches (Figure 8) makes important seasonal changes difficult to detect. Also the months selected for the two seasons may not match the periodicity of natural cycles. Another possible explanation is that the seasonality in catches of common animals such as Euphausia pacifica and Calanus spp. may be less than that in small herbivores with shorter life spans and generation times. Small copepods such as Pseudocalanus, Oithona, and Acartia, which were not sampled adequately with my nets, are known to be abundant in Oregon-Washington waters in the summer, especially in upwelled waters along the coast (Frolander 1962; Gross 1964; Peterson 1972; Peterson and Miller 1975). Inshore-Offshore Variations Largest standing stocks of macrozooplankton and micronekton (grams per square meter but not 77 FISHERY BULLETIN: VOL. 74, NO. 1 grams/1,000 m^, Tables 3 and 4, Figure 6) were found intermediate distances off the Oregon coast, namely over the continental slope at sta- tions 46 and 84 km offshore. A trend for maxima at intermediate distances offshore has been re- ported for other regions. Standing stocks of zoo- plankton were highest at the edge of the shelf or over the inner slope off New York (Grice and Hart 1962), intermediate distances from shore off California (Smith 1971), and near mid-shelf in the Florida Current off Cape Hatteras, N.C. (St. John 1958). Macrozooplankton and micronekton collected with a 0.9-m IKMT off Vancouver Is- land, Canada and Washington were maximal over the outer edge of the shelf (Day 1971). The reduced feeding activity of pink, chum, and sock- eye salmon as they approach the coast is pur- portedly explained by the low macroplankton concentrations in neritic waters and higher con- centrations in offshore waters of the northwest- ern Pacific off Kamchatka (Andrievskaya 1957; Mednikov 1958). All of these studies indicate that small intermediate consumers may achieve maximum importance in the pelagic food chain in deep waters beyond the inner shelf (see also Wilhams et al. 1968). The reason why catches of micronekton and macrozooplankton were higher offshore than nearshore may be related to their vertical migra- tions. Most of the species of micronekton and euphausiids caught in upper waters at night undertake diel vertical migrations (Pearcy and Laurs 1966; Pearcy and Forss 1966; Brinton 1967; Pearcy and Mesecar 1971); hence they may be most abundant in waters deep enough to permit vertical movements but where productiv- ity is enhanced near land (Pearcy 1964). If they drift over the shelf, they may be eaten by large benthic or pelagic predators (Isaacs and Schwartzlose 1965; Pereyra et al. 1969). The inshore-offshore changes in standing stocks of micronekton for the two seasons (Figure 4) suggest that these distributions are interre- lated. Movement of animals may be correlated with seasonal oceanographic changes. During the summer, when the biomass increases greatly from 46 km to a peak at 84 km, large inshore- offshore gradients also occur in physical proper- ties because of upwelling, and there is an offshore component of nearshore surface waters (Pillsbury 1972). During the winter, when biomass from 46 to >120 km is relatively uniform, inshore- offshore gradients are weak, surface currents are onshore, and downwelling occurs (Hebard 1966; Laurs 1967). The significant increase in biomass at 46 km in the winter may be caused by inshore advection of surface water and animals and the concentrating effect of shallow water near the edge of the shelf on vertical migrants. The peak at 84 km in the summer, though far from the coast, may be related to upwelling. Sometimes Laurs (1967) found maximum biomass of carni- vores at 65-84 km and maxima of lower trophic levels closer inshore off Brookings, Oreg., during the summer, suggesting a succession of trophic level maxima such as reported by Sette (1955), King ( 1958), and Vinogradov and Voronina ( 1962) in areas of oceanic upwelling in equatorial waters. Herbivore: Carnivore Ratios Others have also found that the herbivore: carnivore biomass ratios decrease from shal- low, eutrophic waters to oceanic waters. Grice and Hart (1962) reported that well over one-half of the zooplankton by volume in shelf waters off New York herbivorous, while in the Sargasso Sea only about one-half belonged to this trophic level. The percentage of herbivores in the zooplankton catches decreased from inshore waters that were affected by upwelling into offshore waters of the California Current off Baja California (Longhurst 1967). Greze ( 1970) reported that the biomass and production of herbivores and carnivores was a larger percentage of that of primary producers in the Equatorial Atlantic or Ionian Sea than the shallow waters of the Black Sea or Sevastopol Bay. These trends suggest that (a) a smaller fraction of the herbivorous biomass is captured in oceanic than neritic waters because of escape- ment through coarse mesh or avoidance, (b) pro- duction per unit biomass of herbivores is higher relative to that of carnivores in offshore waters, or (c) that ecological efficiences (food consumed by tropic level n + 1 to food consumed by trophic level n ) are higher in oceanic than neritic waters. ACKNOWLEDGMENTS This research was supported by the National Science Foundation (Grant GB-1588) and the Office of Naval Research (Contract NOOO 14-67- A-0369-0007 under project NR 083-102). I am grateful to Harriet Lorz, Henry Donaldson, Lyle Hubbard and others who helped with the field 78 PEARCY: MICRONEKTON AND MACROZOOPLANKTON OFF OREGON and laboratory work. Charles B. Miller and Law- rence F. Small made helpful comments on the manuscript. LITERATURE CITED AHLSTROM, E. H., AND J. R. THRAILKILL. 1963. Plankton volume loss with time of preservation. Calif. Coop. Oceanic Fish. Invest. Rep. 9:57-73. Anderson, G. C. 1965. Fractionation of phytoplankton communities off the Washington and Oregon coasts. Limnol. Oceanogr. 10:477-480. ANDRIEVSKAYA. L. D. 1957. The food of Pacific salmon in the northwestern Pacific Ocean. Vses. Nauchno-Issled. Inst. Morsk. Rybn. Khoz. Okeanogr. (VNIRO), Mosc, p. 64-75. (Fish. Res. Board Can., Transl. Ser. 182, 12 p., 1958). Bakun, a. 1973. Coastal upwelling indices, west coast of North America, 1946-71. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-671, 103 p. Brinton, E. 1967. Vertical migration and avoidance capability of euphausiids in the California Current. Limnol. Ocean- ogr. 12:451-483. Cross, f. a. 1964. Seasonal and geographic distribution of pelagic copepods in Oregon waters. M.S. Thesis, Oregon State Univ., Corvallis, 71 p. CUSHING, D. H. 1971. Upwelling and the production of fish. Adv. Mar. Biol. 9:255-334. DAY, D. S. 1971. Macrozooplankton and small nekton in the coastal waters off Vancouver Island (Canada) and Washington, spring and fall of 1963. U.S. Dep. Commer, Natl. Mar. Fish. Serv., Spec. Sci. Rep. Fish. 619, 94 p. EBELING. a. W, R. M. IBARA, R. J. LAVENBURG, and F. J. ROHLF 1970. Ecological groups of deep-sea animals off Southern California. Bull. Los Ang. Cty. Mus. Nat. Hist. Sci. 6, 43 p. FROLANDER H. F. 1962. Quantitative estimations of temporal variations of zooplankton off the coast of Washington and British Columbia. J. Fish. Res. Board Can. 19:657-675. GREZE, V. N. 1970. The biomass and production of different trophic levels in the pelagic communities of south seas. In J. H. Steele (editor). Marine food chains, p. 458-467. Univ. Calif. Press, Berkley and Los Ang. Grice, G. d., and a. d. Hart. 1962. The abundance, seasonal occurrence and distribu- tion of epizooplankton between New York and Bermuda. Ecol. Monogr. 32:287-309 HEBARD, J. F. 1966. Distribution of Euphausiacea and Copepoda off Oregon in relation to oceanographic conditions. Ph.D. Thesis, Oregon State Univ., Corvallis, 85 p. ISAACS, J. D., AND R. A. SCHWARTZLOSE. 1965. Migrant sound scatterers: Interaction with the sea floor Science (Wash., D.C.) 150:1810-1813. KING, J. E. 1958. Variation in abundance of zooplankton and forage organisms in the central Pacific in respect to the equa- torial upwelling. Proc. Ninth Pac. Sci. Congr. 16:98-107. LASKER, R. 1970. Utilization of zooplankton energy by a Pacific sardine population in the California current. In J. H. Steele (editor), Marine food chains, p. 265-284. Univ. Calif. Press, Berkley and Los Ang. Laurs, R. M. 1967. Coastal upwelling and the ecology of lower trophic levels. Ph.D. Thesis, Oregon State Univ., Corvallis, 103 p. LeBRASSEUR, R. J. 1965. Seasonal and annual variations of net zooplankton at Ocean Station P. 1956-1964. Fish. Res. Board Can., Manuscr Rep. Ser. 202, 33 p. LeBRASSEUR, R. J., AND O. D. KENNEDY 1972. Microzooplankton in coastal and oceanic areas on the Pacific Subarctic Water mass: A preliminary report. In A. Y. Takenouti (editor). Biological oceanography of the northern North Pacific Ocean, p. 355-365. Idemitsu Shoten, Tokyo. LONGHURST, A. R. 1967. Diversity and trophic structure of zooplankton communities in the California Current. Deep-Sea Res. 14:393-408. McAllister, C. D. 1961. Zooplankton studies at ocean weather station "P" in the northeast Pacific Ocean. J. Fish. Res. Board Can. 18:1-29. 1969. Aspects of estimating zooplankton production from phytoplankton production. J. Fish. Res. Board Can. 26:199-220. McAllister, C. D., T R. Parsons, and J. D. H. Strickland. I960. Primary productivity and fertility at station "P" in the Northeast Pacific Ocean. J. Cons. 25:240-259 McGowan, J. A. 1971. Oceanic biogeography of the Pacific. In B. M. Funnell and W. R. Riedel (editors), The micropaleontol- ogy of oceans, p. 3-74. Cambridge Univ. Press, Camb. MEDNIKOV, B. M. 1958. On the plankton of the northwestern part of the Pacific Ocean. Vses. Nauchno-Issled. Inst. Morsk. Rybn. Khoz. Okeanogr (VNIRO), Mosc, p. 76-86. (Fish. Res. Board Can., Transl. Ser 183, 7 p., 1958). OMORI, M. 1969. Weight and chemical composition of some impor- tant oceanic zooplankton in the North Pacific Ocean. Mar Biol. (Berl.) 3:4-10. Parsons. T. R., and R. J. LeBrasseur. 1970. The availability of food to different trophic levels in the marine food chain. In J. H. Steele (editor), Marine food chains, p. 325-343. Univ. Calif Press, Berkley and Los Ang. PEARCY W G. 1964. Some distributional features of mesopelagic fishes off Oregon. J. Mar. Res. 22:83-102. 1965. Species composition and distribution of pelagic cephalopods from the Pacific Ocean off Oregon. Pac. Sci. 19:261-266. 1970. Vertical migration of the ocean shrimp, Pandalus jordani: A feeding and dispersal mechanism. Calif. Fish Game 56:125-129 79 FISHERY BULLETIN: VOL. 74, NO. 1 1972. Distribution and ecology of oceanic animals off Oregon. In A. T. Pruter and D. L. Alverson (editors), The Columbia River estuary and adjacent ocean waters, p. 351-377. Univ. Wash. Press, Seattle. PEARCY, W. G., AND C. A. FORSS. 1966. Depth distribution of oceanic shrimps (Decapoda: Natantia) off Oregon. J. Fish. Res. Board Can. 23: 1135-1143. 1969. The oceanic shrimp Sergestes similis off the Ore- gon coast. Limnol. Oceanogr. 14:755-765. PEARCY, W. G., AND R. M. LAURS. 1966. Vertical migration and distribution of mesopelagic fishes off Oregon. Deep-Sea Res. 13:153-165. PEARCY, W. G., AND R. S. MESECAR. 1971. Scattering layers and vertical distribution of oceanic animals off Oregon. In G. B. Farquhar (editor), Proceedings of an international symposium on biologi- cal sound scattering in the ocean, p. 381-394. Maury Cent. Ocean Sci., Wash,, D.C. PEREYRA, W. T., W. G. PEARCY, AND F. E. CARVEY, JR. 1969. Sebastodes flavidus, a shelf rockfish feeding on mesopelagic fauna, with consideration of the ecological implications. J. Fish. Res. Board Can. 26:2211-2215. PETERSON, W. K. 1972. Distribution of pelagic Copepoda off the coasts of Washington and Oregon during 1961 and 1962. In A. T. Pruter and D. L. Alverson (editors), The Columbia River estuary and adjacent ocean waters, p. 313-343. Univ. Wash. Press, Seattle. PETERSON, W. T., AND C. B. MILLER. 1975. Year-to-year variations in the planktology of the Oregon upwelling zone. Fish. Bull., U.S. 73:642-653. PILLSBURY, R. D. 1972. A description of hydrography, winds, and currents during the upwelling season near Newport, Oregon. Ph.D. Thesis, Oregon State Univ., Corvallis, 163 p. Reid, J. L., Jr. 1962. On circulation, phosphate-phosphorous content, and zooplankton volumes in the upper part of the Pacific Ocean. Limnol. Oceanogr 7:287-306. REID, J. L., JR., G. I. RODEN, AND J. G. WYLLIE. 1958. Studies of the California Current system. Calif Coop. Oceanic Fish. Invest. Progr. Rep. 1 July 1956 to 1 January 1958, p. 25-57. SETTE, O. E. 1955. Consideration of midocean fish production as re- lated to oceanic circulatory systems. J. Mar. Res. 14:398-414. Smith, P. E. 1971. Distributional atlas of zooplankton volume in the California Current region, 1951 through 1966. Calif. Coop. Oceanic Fish. Invest. Atlas 13, 144 p. St JOHN, P. A. 1958. A volumetric study of zooplankton distribution in the Cape Hatteras area. Limnol. Oceanogr. 3:387-397. Tate, M. W, and R. C. Clelland. 1957. Nonparametric and shortcut statistics in the social, biological, and medical sciences. Interstate Printers and Publishers, Inc., Danville, 111., 171 p. THRAILKILL, J. R. 1956. Relative areal zooplankton abundance off the Pacific Coast. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 188, 85 p. Vinogradov, M., and N. Voronina. 1962. The distribution of different groups of plankton in accordance with their trophic level in the Indian Equa- torial Current area. Rapp. P.-V. Reun., Cons. Perm. Int. Explor Mer 153:200-204. WILLIAMS, R. B., M. B. MURDOCH, AND L. K. THOMAS. 1968. Standing crop and importance of zooplankton in a system of shallow estuaries. Chesapeake Sci. 9:42-51. Wyatt, B., W. V. Burt, and J. G. Pattullo. 1972. Surface currents off Oregon as determined fi-om drift bottle returns. J. Phys. Oceanogr. 2:286-293. 80 CULTURE AND GROWTH OF NORTHERN ANCHOVY, ENGRAULIS MORDAX, LARVAE John R. Hunter^ ABSTRACT Culture techniques used to rear larval anchovy through metamorphosis using laboratory cultured foods are described. Anchovy larvae fed dinoflagellates Gymnodinium splendens, rotifers Brachionus plicatilis, harpacticoid copepods Tisbe furcata, and brine shrimp nauplii Artemia salina, completed metamorphosis (35 mm) in 74 days at 16°C with a minimum survival of 12.5*^. Growth in length and weight were recorded over this interval and an excellent fit to the Laird-Gompertz growth equation was obtained. Growth was comparable to that on a wild plankton diet. In a starvation experiment, most of the fish that completed metamorphosis withstood a starvation period of 12- 15 days, whereas those that had not completed metamorphosis did not. Knowledge of the growth rate of northern an- chovy, Engraulis mordax Girard, is essential for estimating year class success or larval survival. Another important element in estimating sur- vival is the time fish or larvae can withstand starvation. In this report I describe the growth rate of larval anchovy to metamorphosis and present data on the ability of newly metamor- phosed juveniles to withstand starvation. Special attention is also given to culture techniques be- cause this is the first time northern anchovy have been reared through metamorphosis entirely on cultured foods. Kramer and Zweifel (1970) recorded the growth of anchovy larvae at 17° and 22°C for periods of up to 34 days. In their experiments larvae at- tained an average length of 17 mm but did not reach metamorphosis, which is complete at about 35 mm standard length. Their larvae were fed wild plankton supplemented by Artemia salina nauplii. In the ensuing years, rearing techniques using cultured foods have gradually been de- veloped: Gymnodinium splendens for 3- to 5-day- old larvae (Lasker et al. 1970), and Brachionus plicatilis for 5- to 20-day-old larvae (Theilacker and McMaster 1971). This paper describes the use of the harpacticoid copepod Tisbe furcata which are the proper size food for larvae older than 20 days (10 mm). All previous attempts to rear anchovy larvae beyond 35 days on cultured foods have failed. In all attempts Artemia nauplii were used after 20 days. 'Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, La Jolla, CA 92038. METHODS Five rearing experiments were done, four at 16°C and one at 17° to 18°C (Table 1). Eggs for all experiments were obtained from a captive popu- lation of anchovy which were maintained in breeding condition continuously at the South- west Fisheries Center La Jolla Laboratory (Leong 1971). Rearing tanks were cylindrical, black fiber- glass, 122 cm diameter, 36 cm deep, covered with a transparent acrylic plastic top, and im- mersed in a water bath regulated by a refrigera- tion unit. Temperature was maintained near 16°C in all but one experiment, and the salinity was 35%. Fluorescent lamps suspended directly over each tank provided about 2,000 Ix at the water surface. The volume of water in the tanks gradually increased from an initial volume of 200 liters of filtered seawater to 400 liters by about 20 days because of additions of seawater contain- ing algae and food organisms. Thereafter, the volume was maintained at about 400 liters by siphoning water from the bottom from time to time which also cleaned the tank. Records were kept of the quantity of food or algae added to tanks and on alternate days 16, 0.20-ml aliquots were taken to measure the den- sity of Brachionus plicatilis, Gymnodinium splendens, and Artemia salina nauplii in the tanks. Concentrations of Tisbe furcata in the tanks were not recorded because they were con- centrated on or near the walls and bottom of the tank, but records were kept of the numbers added to the tank. Details regarding the feeding of Manuscript accepted June 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 1976. 81 FISHERY BULLETIN: VOL. 74, NO. 1 o o in o +1 CO c\j CD 8° ■* o +1 iri 8' o CO 00 in o o o CM O +1 00 CO CO o CD CD Cvj 8 6 CO '" "-i +1 ' — ' co'r^ o ir> r~. a, o-uj u, c S 0) _OT- - X 3 c ? *^ T3 ■— C E ^ J3 2"; != ? H? CO (D "O E a- r^ 4 -A I CM CM CM CO I I I CM CM CD r^ CO O CM O CM CM Tt CM Tt in T^ in ^ CM CM ■.- CO CM r^ CO CD ■r- C) CD I O CD O CM ^ CM ^2 I I ^ CO in h- CO r) o o 1- Tt o o -^ O -'^ -<;J- CVJ CD CM CO CO LO -A IT) T^ Tt CSJ C\J OJ I I I o o o c> cb CM CO O '- CM CO O 00 tT CM ■<)• r-. CD T^ CO CM CD 2 111 III o o oo o >- CM O CD CM CO I CM -A I ■■- CM I I I I III I III I ° III I I I I I I I I I I I I I I CO 3 _ c; o -9 o 2 S DD O ,^ S _ a. 3 .._ CD o o -o O CD "D "D IB ^^ ^ f— r- — c:> p c c m C "1 fc CD H) CD c _ (D O Q) (/I O en " ~ 3 = 5 = ^ o ic o o CD _ C CD o " — "D >--o CD "D "D CD 0) -3 fatty acids. Crawford et al. (1975) entertained the possibility that such oxidation and subsequent fishy flavor development occur mostly in vivo. At first glance, the effects of dietary a-tocopherol acetate on prevention of fishy flavor seem to support this hypothesis. However, the effective- ness of injecting a-tocopherol only a few days before slaughter casts some doubt on this reason- 'Western Regional Research Laboratory, Agriculture Re- search Service, U.S. Department of Agriculture, Berkeley, CA 94710. Manuscript accepted June 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 1976. ing since in vivo oxidation prior to injection should have had ample time to occur. Whereas this doubt does not call for total apostasy, it does suggest that postmortem oxidation and sub- sequent development of fishy flavor is indeed a possibility and deserves consideration. The exact nature and origin of flshy flavor in turkeys is not known, but it is known that the development of such flavor requires the uptake of oi-S fatty acids from dietary oils rich in these fatty acids. Most fish oils are rich sources of long chained w-3 fatty acids which are readily taken up into the carcass of turkeys when included in their diet. Linseed oil contains more than 50% linolenic acid and when incorporated into turkey diets, the linolenic acid is taken up and elongated to the longer chained homologues thereby caus- ing fishy flavor to develop (-Klose et al. 1951; Miller et al. 1967a, b; Crawford et al. 1974). If postmortem oxidation plays a major role in the development of fishy flavor, it is likely that the development would occur largely during cook- ing. Pippen and Nonaka (1963) found that the amount of volatiles from raw chicken was small and the aroma rather insipid when compared to the relatively large amount of highly odoriferous volatiles from cooked chicken. They also reported that chicken boiled in air yielded a more complex and larger volatile fraction than chicken boiled in nitrogen. Crawford (1972) reported that replace- ment of air in the headspace with nitrogen gave some protection against scorch during the retort- ing of 4-pound cans of tuna. This suggests that less carbonyls (volatiles) were formed under ni- trogen since volatile carbonyls, sugars, and 89 FISHERY BULLETIN: VOL. 74, NO. 1 amino compounds (Fujimoto et al. 1968) have been implicated in such nonenzymatic browning (Tarr 1954; Jones 1962). It is clear that the development of the normal aroma of poultry is time-temperature dependent and that air or nitrogen cooking atmospheres have profound effects on the development of this aroma. Therefore, it is likely that control of the cooking atmosphere may affect the development of fishy flavor in poultry meat if this flavor requires air and/or heat for its development. This paper explores the effects of cooking in different atmospheres on the flavor of breast meat from turkeys fed diets containing tuna oil w^ith and without dietary a-tocopherol acetate or a-tocopherol injection. Diced breast meat was cooked in air as well as under nitrogen with a slight vacuum. EXPERIMENTAL Turkey Diets and Feeding The turkeys used in this experiment were taken from groups of turkeys raised experi- mentally for other work. Their diets and feeding are described in some detail by Crawford et al. (1975). Briefly, there were 50 White Broad Breast poults in experiment C that were divided into five groups of 10 each and they were fed as follows: chick starter (6.75% fish meal) was fed to 3 wk of age, then a 50:50 mixture of chick starter and a 50% soybean meal basal diet for a few days, followed by the 50% soybean meal diet supple- mented with 2% soybean oil and 2% beef fat to 8 wk of age. At 8 wk of age, the following fat and oil supplements replaced the previous ones and they were fed from 8 to 14 wk of age: Oil Supplement to Basal Diet^ 4% BF Group 1 C 2C 3 C 4C 5C 2% BF + 2% TO 2% BF + 2% TO 2% BF + 2% TO 2% BF + 2% TO iBF = Beef fat; TO = Tuna fish oil. At 14 wk of age, the above groups of turkeys were fed a 30% soymeal basal diet plus the following oil supplement to 16 wk of age: Group Oil Supplement to Basal Diet^ 1 C Keep on 4% BF 2 C Change to 4% BF Group 1 B 2B 3 B 4B 5B 3 C Change to 4% BF + 100 mg Vit. E/kg 4 C Change to 4% BF + 200 mg Vit E/kg 5 C Keep on 2% BF + 2% TO ^BF = Beef fat; Vit. E = dl a-tocopherol acetate; TO = Tuna fish oil In experiment B, 50 poults were obtained and handled as above. On day 3, they were fed a basal diet plus 4% beef fat to 14 wk of age. From 14 to 16 wk of age, they were fed as follows: Oil Supplement to Basal Diet 4% BF 2% BF + 2% TO 2% BF + 2% TO (+ injection of 170 mg a-tocopherol into thigh at 72, 48, 24 h before sacrifice) 2% BF + 2% TO + 100 mg Vit. E/kg 2% BF + 2% TO + 500 mg Vit. E/kg Sampling, Canning, and Analysis All turkeys were sacrificed at 16 wk of age then handled and stored at -30°C as described by Crawford et al. (1974). Two turkeys from each group were randomly selected and thawed over- night in a 2°C cold room. The breasts were excised and diced in the cold after the skin had been removed. Breast meat from turkeys of the same group were mixed together and appropriately identified. The diced breast meat was canned immediately as follows: breast meat from each group was hand packed into 307 x 113 cans (eight cans per group) leaving a headspace of about V2 inch. All cans from each group were alternately evacuated and flushed with nitrogen several times. On the final nitrogen flush, the lids were sealed when the vacuum dropped to 5 inches. Four of the cans from each group were frozen at -30°C until used and the other four cans were cooked immediately at 116°C (15 psi) for 80 min to an internal temperature of ca. 112°-115°C, cooled, and stored at 2°C until used. The four uncooked cans from each group were removed from -30°C storage, thawed to about 2°C, opened, and the contents cooked in aluminum trays (with loose covers) at about 117°C for 30 min (internal tem- perature ca. 70°C) before serving. Those cans that were cooked at 116°C were warmed in boiling water for 10 min before opening and serving. Organoleptic analysis was performed by a panel of eight judges using a balanced incomplete block design (t = 5, r = 4). Only one panel per day was 90 CRAWFORD and KRETSCH: FISHY FLAVOR IN TURKEY convened and the air and nitrogen packs w^ere randomly offered from day to day. Duncan's mul- tiple range test (a = 0.05) was used to compare the adjusted mean of the taste panel scores. The scoring was: 1 = no fishy flavor, 5 = very fishy flavor RESULTS AND DISCUSSION The results reported in Table 2 are to be inter- preted with some caution because of the low level of fishiness in the meat from turkeys fed 2% fish oil for only 2 wk. Therefore, only trends are indicated for the results in Table 2 where statisti- cal significance could not be achieved. Tables 1 and 2 report Duncan's multiple range test of the mean taste panel scores of breast meat cooked in air or nitrogen from turkeys fed various diets containing tuna oil and/or beef fat with and without dietary a-tocopherol acetate or a-tocoph- erol injection. All meats that contained «-tocopherol gave taste panel scores that were comparable to the scores for the control for all methods of cooking. When breast meat is cooked under nitrogen with a slight vacuum no appreci- able difference in flavor is caused by any of the diets. However, the breast meat from turkeys fed diets containing 2% tuna oil (treatments 5C and 2B) did have slightly higher scores, although not statistically different from the control (treat- ments IC or IB, 47c beef fat). The breast meat cooked in air from turkeys fed diets containing 2% tuna oil (treatments 5C and 2B) showed more off flavor than those cooked in nitrogen when each is compared to its control (treatments IC or IB, 4% beef fat). Furthermore, the order and rank of the scores for the air-cooked meat were very similar to those of breast meat from whole roasted turkeys previously reported by Crawford et al. (1975). These turkeys were randomly selected from the same groups of turkeys used in this experiment and were roasted at 177°C to center breast temperature of about 70°C. From the results of this experiment, it may be concluded that cooking breast meat of potentially fishy flavored turkeys under nitrogen is nearly as effective in preventing fishy flavor development as feeding a-tocopherol acetate (in the diets with the tuna oil) and roasting in the normal manner This implies that fishy flavor develops postmor- tem and requires air for its development. Alter- nately, it could be concluded that cooking under nitrogen per se had practically no effect in pre- TABLE 1. — Duncan's multiple range test of mean' taste panel scores^ for breast meat cooked in air or nitrogen from turkeys fed various diets containing tuna oil and/or beef with and without a-tocopherol acetate. Cooked in nitrogen Cooked in air Roasted no 5 Treatment^ rmally" Treatment^ Scores Treatment^ Scores Scores 5C 2% TO 2.05 5C 2% TO 3.23 5C 2% TO 3.14 4C 4% BF + 1 .80 3C 4°o BF + 1.66 2C 4% BF 2.43 200 E 100 E 3C 4% BF + 1.77 2C 4% BF, 1.63 3C 4°o BF + 1.31 100 E 100 E 2C4%BF, 1.71 1 C 4°o BF 1.24 IC 4% BF 1.29 1C4%BF 1.65 4C 4°o BF 200 E ' 1.08 4C 4% BF + 200 E 0.99 'Mean taste panel scores connected by a common line are not signifi- cantly different at tfie 0.05 probability level. ^Taste panel scoring: 1 = no fishy flavor, 5 = very fishy flavor. Abbrevia- tions: TO = tuna oil; E = mg d/ a-tocopherol acetate per kilogram of diet; BF = beef fat; BF, = beef fat substituted for 2% TO +2% BF. ^AII groups (except group IC, the control which was maintained on diet with 4°o BF for all 1 6 wk) were fed a basal diet with 2% TO plus 2% BF from 8 to 14 wk of age and from 14 to 16 wk of age, they were fed a basal diet with: group IC =4°b BF, group 2C = change to 4*^0 BF, group 3C = change to 4°o BF -r 100 mg/kg a-tocopherol acetate, group 4C = change to 4''o BF * 200 mg a-tocopherol acetate, group 5C = kept on 2% TO + 2% BF. ■■These results for the breast meat of normally roasted whole turkeys were previously reported by Crawford et al. (1975). Table 2. — Duncan's multiple range test of mean^ taste panel scores^ for breast meat cooked in air or nitrogen from turkeys fed various diets containing tuna oil and/or beef fat with and without a-tocopherol acetate supplement or injection. Cooked in n itrogen Scores Cooked n air Roasted nc Treatment^ rmally" Treatment^ ; Treatment^ Scores Scores 2B 2% TO 1.82 2B 2% TO 2.16 2B 2% TO 2.23 48 2% TO + 1.71 5B 2% TO + 1.74 5B 2% TO + 2.18 100 E 500 E 500 E 5B 2% TO + 1.61 4B 2% TO + 1.41 4B 2% TO + 1.86 500 E 100 E 100 E 3B 2% TO - n 1.59 3B 2% TO + In 1.35 3B 2% TO + In 1.32 IB 4°o BF 1.43 1 B 4% BF 1.22 1B 4% BF 1.19 'Mean taste panel scores connected by a common line are not signifi- cantly different at the 0.05 probability level. ^Taste panel scoring: 1 ^ no fishy flavor, 5 = very fishy flavor. Abbrevia- tions: TO = tuna oil: E = milligrams d/ a-tocopherol acetate per kilogram of diet; In = inject ^»fl^ ' " "■ I 5 6 7 8 9 10 II TOTAL LENGTH (cm) Lymn Creek contained older fish of both species but C. aleuticus lived longer than did C. asper. Distribution of Biomass The autumnal distribution of biomass by stream zone was derived from population esti- mates and length-weight data for both species of sculpin in Lymn, Waterloo, and Cabin creeks (Table 4). Density of sculpin biomass (grams per square meter) was lowest in the estuaries and increased upstream. Cottus aleuticus showed the greatest increase in biomass density with in- creased distance upstream, particularly when proceeding from the estuary upstream into the lower stream zone. About 69-94% of sculpin biomass in the estuaries was C. asper, whereas about 60-100% of sculpin biomass were C. aleuti- cus in the upper zones whose downstream boundaries were marked by the first significant streambed obstruction. Species biomass in the Figure 7. — Length-frequency histograms for sculpin popu- lations in Chef and Waterloo creeks from collections made in September and October. Sampling stations are identified as distances upstream or downstream (-) from a high tide mark (0) in meters. Table 2. — Age distributions of Cottus aleuticus in successive 5-mm intervals of total length, sexes combined. Number in parentheses indicates total number of fish when not all fish in the length interval were aged. 14.5-19.4 19.5-244 24.5-294 29.5-344 34.5-39.4 39.5-44.4 44.5-494 49.5-54.4 54.5-594 59.5-64.4 64.5-69.4 69.5-74.4 74.5-79.4 79.5-84.4 84.5-894 89.5-944 94.5-99.4 99.5-104.4 104.5-109.4 109.5-114.4 114.5-119.4 119.5-124.4 124.5-129.4 129.5-134.4 134.5-139.4 139.5-144.4 144.5-149.4 Total length Lymn Creek Chef Creek Waterloo Creek August September October September-October October (mm) 1 II III IV V VI VII 1 II III IV 1 II III IV 1 II III IV 1 II III IV (7) 12(83) 5(73) 3(50) 8(15) 9 8 40 37 25 6 (4) 3(65) 6(75) 10(80) 6(40) 21 1 1 2 4 10 6(24) 10(154) 14(169) 3(50) 9 16 4 Total fish 240 116 25 16 4 5 11 288 25 12 6 426 2 2 9 9 12 3 3 (8) 3(137) 18(132) 9(64) 2(13) 3(6) 22(42) 39 27 13 6 5 10 9 12 3 1 2 5 13 17 13 2 1 1 5 14 15 15 6 3 40 19 7 2 347 140 40 14 1 52 60 17 12 4 137 FISHERY BULLETIN: VOL. 74, NO, 1 Table 3.- -Age distributions ofCottus asper in successive 5-mm intervals of total length, sexes combined. Number in parentheses indicates total number of fish when not all fish in the length interval were aged. 14.5-19.4 19.5-24.4 24.5-29.4 29.5-34.4 34.5-39 4 39.5-44.4 44.5-49.4 49.5-544 54.5-59.4 59.5-64.4 64.5-69.4 69.5-74.4 74.5-79.4 79.5-84.4 84.5-89.4 89 5-944 94.5-99.4 99.5-104.4 104.5-109.4 109.5-114.4 114.5-119.4 119.5-124.4 124.5-129 4 129.5-134.4 134.5-139.4 139.5-144.4 Total fish Lymn Creek Chef Creek Waterloo Creek Total length (mm) August September October September-October 1 II III October 1 II III IV V 1 II III 1 II III IV 1 II III IV V (3) (25) 3(51) 6(55) 1(41) 5(33) 8(13) 1 1 5 8 8 11 8 2 1 7 15 8 2 5 1 (4) (16) 1(39) 9(46) 18(49) 33(38) 26 11 1(3) 2(19) 8(30) 11(27) 24 10 3 (2) 1(3) 8(21) 7(17) 9 5 4 (8) 1(9) (1) 7 2 1 1 1 2 2 6 3 2 1 1 1 222 43 39 7 1 229 16 116 13 22 61 28 4 14 Table 4. — The autumnal distribution of sculpin (Coitus) biomass in three streams. Stream Zone Lymn Estuary Lower Upper Total area Waterloo Estuary Lower Upper Total area Cabin Estuary Lower Upper Total area Sculpin biomass (kg) (g/m2) 2.727 4.345 4.772 0.98 1.72 3.38 11 844 0.310 9.052 16.454 1.76 0.69 3.49 3.45 25.816 0.037 1.493 0.508 3.31 033 4.87 2.85 C. aleuticus C. asper Table 5. — The autumnal distribution of sculpin iCottus) biomass by age-class in three streams, expressed as a per- centage of species biomass. (g/m2) (%) (g/m^j (%) Stream 1 II Ill IV V VI VII 0.31 31.6 0.67 68 8 Lymn Creek: 0.88 51.2 0.84 48.8 C. aleuticus 6.7 12.5 10.5 19.6 225 17.6 4.4 6.2 3.38 100.0 — — C asper Waterloo CreSR- 17.2 14.0 35.2 19.9 10.8 2.7 1,17 663 0.75 39.0 C. aleuticus 29 286 20.8 32.3 15.3 0.04 5.8 0.65 94.2 C asper 4.0 54 47.2 249 18.5 2.95 84.5 0.54 15.5 Cabin Creek: 3.45 100.0 — — C. aleuticus 14.9 42.5 20.5 17.8 4.2 3.09 93.5 12.1 0.56 0.29 6.5 87.9 C asper 17.1 15.4 39.6 27.8 0.04 4.42 90.8 0.45 59.2 1.72 60.4 1.13 39.6 Annu a\ (; row th. Vfor tahl V. ai nd 2.038 3.41 2.79 81.8 0.62 18.2 lower stream zone was nearly equal in Lymn Creek but was predominantly C. aleuticus (85%) in Cabin and Waterloo creeks. The two sculpins differed in relative distribu- tion of biomass by age group within their popula- tions (Table 5). Whereas C. asper in their third growth season (age II) constituted 35-47% of population biomass, the biomass of C. aleuticus populations in Lymn and Waterloo creeks was more evenly distributed in older age groups. The contribution of age I to population biomass of C. aleuticus was considerably higher in the two smaller streams than in Lymn Creek and 3-5 times higher than for C. asper in these two streams. Length-Weight Relations The annual growth of both sculpins showed a consistent ranking in three streams. Growth was most rapid in Lymn Creek, intermediate in Waterloo Creek, and slowest in Chef Creek (Fig- ure 8) although the growth of C. asper in Lymn and Waterloo was not statistically different. Dis- similarities in rate of grovvi:h were greatest for C. aleuticus, possibly reflecting its greater re- liance on the productivity of the freshwater stream than in the case of C. asper, which spends considerably more time in the estuary through- out its life history. C. asper grew more rapidly than did C. aleuti- cus, the age-specific disparity in weight gain increasing with age. Growth of the Lymn Creek 138 MASON and MACfflDORI: POPULATIONS OF SYMPATRIC SCULPINS AGE-CLASS Figure 8. — Annual growth rates (weight) of Coitus aleuticus and C. asper in Lymn, Chef, and Waterloo creeks. population, which was first sampled in early April, was most rapid during the spring and early summer and nearly completed by mid-August. The largest coastrange sculpin captured was 145 mm in length and 8 yr old while the largest prickly sculpin was 144 mm in length and 6 yr old. Length-weight linear regressions based on logged data were calculated for both species in the three largest streams (Table 6) and compared by analysis of variance. The length-weight rela- tion was similar for both species in all three streams except for the coastrange sculpin in Chef Creek, which was considerably lighter per unit length than in the other two systems (F2 2737 = 77.5). Slow annual growth and a lower slope (6) may reflect poorer feeding conditions or as- sociated population stress during the late sum- mer when the flow in a 500- to 600-m section of this stream becomes intermittent. Estimates of average annual mortality for both species of sculpins in Lymn, Cabin, and Chef creeks ranged between 58 and 75%, the differ- ences between species and streams depicted in Figure 9 being statistically non-significant. Al- though similar for both sculpins, mortality in Waterloo Creek was considerably lower than in the other three streams 38-40%. No estimate of Table 6. — Length-weight regression parameters (log y = a + bx) for Cottus aleuticus and C. asper in three streams on Vancouver Island, B.C. Param- C aleuticus C asper eter Lymn Chef Waterloo Lymn Chef 73 Waterloo N 1,565 767 397 1,225 49 a -5.312 -5.001 -5.297 -5.268 -5.143 -5.363 b 3.237 3.041 3.224 3.203 3.122 3.259 r 0.993 0994 0,996 0.992 0.997 0,998 Syx 00096 00122 0.0136 00115 0.0308 0.0268 annual mortality was attempted for C asper in Chef Creek due to the small population present. Despite close agreement to the linear function of the majority of point estimates, some points for young and old age-classes deviated considerably and are taken to indicate poor survival, low re- cruitment of subyearlings from the estuary in some years, or inadequate sampling. For exam- ple, poor survival of age I of C. asper is indicated for Lymn, Waterloo, and Cabin creeks (Figure 9). Similarly, age of both species were poorly repre- sented in Waterloo Creek, as were age in Chef Creek, despite intensive sampling in the down- stream areas in which they were distributed. In Chef Creek, age IV of C. aleuticus was very poorly represented, suggesting either a sudden exten- sive mortality or inadequate sampling effort in the larger pools upstream where these fish reside. DISCUSSION The ecological importance of cottid fishes in the simple fish communities of these coastal streams remains essentially unknown but the present findings appear to be timely in view of the resurg- ing interest in enhancing the natural production of anadromous stream salmonids. Previous 10.000 1000 I to O q: 100 m s < jr 10 C.ALEUTICUS \ 8 A ■••\ y \ \ ^"'■' °'^ \ V \ '.ox \ \ \ \ \ \ \ \« \ \ \ \ • C ASPER \ • \ \ \ \ \ a -•- LYMN CREEK -0- CHEF CREEK ■••A- WATERLOO CREEK --D- CABIN CREEK \ • 1 1 1 y Vr yl AGE-CLASS TV Figure 9. — Declining numbers with increasing age within sympatric populations of Cottus aleuticus and C. asper in four streams. Straight lines describe least-square regressions of best fit. 139 FISHERY BULLETIN: VOL. 74, NO. 1 studies on C. aleuticus and C asper, which are widely distributed and commonly abundant in coastal streams from California to Alaska, have emphasized their potentially destructive role as predators on the eggs and fry of salmon and trout (Shapovalov and Taft 1954; Hunter 1959; McLar- ney 1967). Conversely, it has been generally shown that sculpins in streams of the North Temperate Zone prey incidentally on salmon and trout, but sculpins do share a common source of food — the benthic invertebrate community. The probable importance of interspecific com- petition in general, and for food in particular, in such streams where the several species of fishes consume in common a wide variety of food or- ganisms has been readily acknowledged (Hartley 1948; Maitland 1965; Mann and Orr 1969) but continues to defy quantitative analysis. The over- lapping summer foods of juvenile coho salmon, cutthroat trout, and coastrange sculpins in Cabin Creek (Table 7) clearly show the possibility of competition for food in the present study streams. Numerically, Ephemeroptera and Diptera were important in all three diets but most important in the coho salmon diet, while Trichoptera were most important in the trout and sculpin diets. The sculpins showed the least varied diet as Ephemeroptera, Diptera, and Trichoptera com- posed nearly 95% of the food items consumed. Dietary differences can be related to behavioral differences in feeding and habitat response. The Table 7. — The percentage composition by frequency of occurrence (0) and number (N) of the midsummer (June-July) foods eaten by juvenile coho salmon, cutthroat trout, and coastrange sculpins in Cabin Creek. Based on 30 fish of each species collected simultaneously. Coho Troul Sculp <7cO in Food category %0 %N %0 %W %N Oligochaeta 10.0 12 Diplopoda — — 40.0 6.2 — — Collembola 23.3 41 — — — — Ephemeroptera 46.7 17.4 30.0 9.2 60.0 30.3 Plecoptera 16.7 4.1 20.0 3.1 3.3 <1 Hemiptera 20.0 3.3 10.0 1.5 — — Coleoptera Adults 30.0 6.2 30,0 46 33 <1 Larvae 13.3 1.7 — — — — Trichoptera 10.0 2.1 400 262 56.6 44.9 Lepidoptera' 3.3 <1 50.0 92 6.7 2.8 Diptera Adults 70.0 285 20,0 7.7 — — Larvae 26.7 20.7 43,4 154 33.3 18.3 Hymenoptera' 10.0 1.2 20,0 3.1 — — Araneida 36,7 6.6 10,0 1.5 — — Acarina 6.7 1.7 23.3 10.8 3.3 1.8 Gasteropoda 3.3 <1 — — — — 'Refers to adult stage, all categories of Insecta are larval stages unless noted otherwise. sculpins were abundant in all habitats but ate few foods of surface origin, being crepuscular grazers on the benthos. The trout were princi- pally riffle-dwellers and grazed the benthos (both trout and sculpins ate large numbers of Trichop- tera larvae) but exploited the invertebrate drift to a lesser extent than did the coho salmon, which preferred the pool and glide habitats of low cur- rent velocity. Despite this behavioral diversity, niche differentiation remains poorly developed in the Eltonian sense discussed by Weatherley (1963) who proposed that the niche be defined as "...the nutritional role of the animal in its ecosystem... ." Recent experiments have clearly illustrated that populations of juvenile coho salmon in these streams are limited by their food supply during the summer months (Mason 1974b, 1974c). Rates of growrth, survival and emigration were amena- ble to manipulation by varying population den- sity and food availability. Thus, in that young coho salmon share a common food supply with both trout and sculpins, the likelihood of food competition is strongly suspected. Since direct documentation of competition among stream fishes in natural environments continues to elude us, the inferential definition of competition proposed by Maitland (1965) appears to have greater utility than the modus operandi definition of Larkin (1956), ". . .the demand, typi- cally at the same time, of more than one or- ganism, for the same resources of the environ- ment in excess of immediate supply." Maitland (1965) suggested that competition occurs "... when the presence of more than one species causes the average total biomass (standing crop) of one of them to be less than it would be if that species were existing alone — species which are directly parasitic or predatory on one another being excepted." Fish biomass in small coastal streams of Van- couver Island usually ranges between 7 and 10 g/m^ in midsummer (unpubl. data). Of this 3-6 g/m^ (50-80%) consists of sculpins (C. asper and C. aleuticus) in the first several kilometers above the estuarine zone. Studies by Brocksen et al. (1968) have shown that, within the carrying capacity of laboratory streams producing natural drift foods, production of cutthroat trout was de- termined by the biomass ratio of trout and scul- pin, C. perplexus, at time of stocking, whereas sculpin production remained independent of trout biomass. These results were obtained over a 140 MASON and MACfflDORI: POPULATIONS OF SYMPATRIC SCULPINS range of species biomass levels commensurate with those encountered in nature and suggest that the availability of drift foods for the trout was determined by the intensity of grazing by sculpins on the stream benthos. From the present study, the restricted ability of both species of sculpins to surmount obstacles in the streambed, coupled with the life history features of planktonic young and downstream spawning migrations, lend themselves to the po- tential development of a management strategy for enhancing the production of salmonid smolts to the sea. If the findings of Brocksen et al. (1968) can be corroborated in stream simulator systems more closely approximating the natural envi- ronment, studies on the locomotory ability of these sculpins relative to the performance of their communal salmonids could provide the design criteria for physical barriers to be located on test streams at suitable sites above the influence of high tide. LITERATURE CITED BROCKSEN, R. W., G. E. DAVIS, AND C. E. WARREN. 1968. Competition, food consumption, and production of sculpins and trout in laboratory stream communi- ties. J. Wildl. Manage. 32:51-75. Hartley, p. H. T. 1948. Food and feeding relationships in a community of fresh-water fishes. J. Anim. Ecol. 17:1-14. Hunter, J. G. 1959. Survival and production of pink and chum salmon in a coastal stream. J. Fish. Res. Board Can. 16:835-886. KREJSA, R. J. 1967. The systematics of the prickly sculpin, Cottus asper Richardson, a polytypic species. Part H. Studies on the life history, with especial reference to migra- tion. Pac. Sci. 21:414-422. Larkin, p. a. 1956. Interspecific competition and population control in freshwater fish. J. Fish. Res. Board Can. 13:327-342. Maitland, p. S. 1965. The feeding relationships of salmon, trout, min- nows, stone loach and three-spined sticklebacks in the River Endrick, Scotland. J. Anim. Ecol. 34:109-133. MANN, R. H. K., AND D. R. O. ORR. 1969. A preliminary study of the feeding relationships of fish in a hard-water and a soft-water stream in southern England. J. Fish. Biol. 1:31-44. Mason, J. C. 1974a. Movements of fish populations in Lymn Creek, Vancouver Island: A summary from weir operations diu-ing 1971 and 1972, including comments on species life histories. [Can.] Dep. Environ., Fish. Mar. Serv. Tech. Rep. 483, 35 p. 1974b. A first appraisal of the response of juvenile coho salmon (O. kisutch) to supplemental feeding in an experimental rearing stream. [Can.] Dep. Environ., Fish. Mar. Serv. Tech. Rep. 469, 21 p. 1974c. A further appraisal of the response to supple- mental feeding of juvenile coho (O. kisutch) in an experimental stream. [Can.] Dep. Environ., Fish. Mar. Serv. Tech. Rep. 470, 26 p. McAllister, D. E., and C. C. Lindsey. I960. Systematics of the freshwater sculpins (Cottus) of British Columbia. Natl. Mus. Can. Contrib. Zool., Bull. 172:66-89. McLARNEY, W. O. 1967. Intra-stream movement and food habits of a popu- lation of coastrange sculpins, Cottus aleuticus, in rela- tion to a spawning run of the pink salmon, Oncorhyn- chus gorbuscha. Ph.D. Thesis, Univ. Michigan, Ann Arbor, 154 p. 1968. Spawning habits and morphological variation in the coastrange sculpin, Cottus aleuticus, and the prickly sculpin, Cottus asper. Trans. Am. Fish. Soc. 97:46-48. Shapovalov, l., and a. C. TAFT. 1954. The life histories of the steelhead rainbow trout (Salmo gairdneri gairdneri) and silver salmon (Oncorhyn- chus kisutch). Calif Dep. Fish Game, Fish Bull. 98:1-375. Weatherley, a. H. 1963. Notions of niche and competition among animals, with special reference to freshwater fish. Nature (Lond.) 197:14-17. 141 REVIEW OF THE DEEP-SEA FISH GENUS SCOPELENGYS (NEOSCOPELIDAE) WITH A DESCRIPTION OF A NEW SPECIES, SCOPELENGYS CLARKEI, FROM THE CENTRAL PACIFIC John L. Butler^ and Elbert H. Ahlstrom^ ABSTRACT Scopelengys has been known previously from a few widely scattered collections. Recent collections by the Scripps Institution of Oceanography in the Pacific, the RV Walther Herwig in the Atlantic, and the International Indian Ocean Expedition have made possible a critical study of this genus. No significant differences were found in either morphometric characters or meristic counts between specimens of S. tristis Alcock from the eastern North Pacific (la 1. 16 ° to 33°N, long. 117° to 126°W) and those from the eastern South Pacific (lat. 5° to 16°S, long. 77° to 90°W). When Pacific Ocean specimens were compared with those from the Atlantic and Indian oceans, no significant differences were found in morphometric characters, and although differences in average meristic counts were some- what larger between oceans than among Pacific specimens, such differences exceed one for only one meristic character (gill rakers), and the ranges for all counts from all oceans almost completely overlapped. Scopelengys clarkei is described from the central North Pacific. It differs from S. tristis mainly in pectoral ray count (2.5 average difference), average counts of vertebrae (3.3 average difference), deeper caudal peduncle, narrower maxillary, and in a differently pigmented larva. In 1890, Alcock described a new genus and species, Scopelengys tristis, from a single denuded specimen collected in the Arabian Sea. Although there was no evidence of photophores, Alcock placed his new genus in the family Scopelidae ( = Myctophidae) allowing that the "exact position among the Scopelidae cannot be accurately de- fined at present." Garman (1899) described S. dispar from two specimens collected in the Gulf of Panama. Garman distinguished S. dispar from S. tristis by its lower dorsal- and anal-fin ray counts. Scopelengys dispar was considered a junior synonym by Parr (1928), Bolin (1939), and Nor- man (1939). Until 1963, Scopelengys was known only from the Indian and Pacific oceans. Its dis- covery in the Caribbean Sea by Mead (1963) re- sulted in the description of a third species, S. whoi Mead. A recent survey of mid-water fishes conducted by the California Cooperative Oceanic Fisheries Investigations (CalCOFI) provided us with specimens which indicated that two species of Scopelengys were present in the Pacific Ocean. Additional specimens made available to us by Thomas A. Clarke of the Hawaiian Institute of 'Smithsonian Institution, Southwest Fisheries Center, Na- tional Marine Fisheries Service, NOAA, La Jolla, CA 92038. ^Southwest Fisheries Center, National Marine Fisheries Ser- vice, NOAA, La Jolla, CA 92038. Marine Biology (see in this regard Clarke 1973), confirmed that the second form was an unde- scribed species. Study of Scopelengys from the At- lantic, Pacific, and Indian oceans indicates thatS. dispar andS. whoi Mead are synonyms ofS. tris- tis Alcock. METHODS AND MATERIALS Measurements were made following Hubbs and Lagler (1958). Measurements are given in percent of standard length (SL), unless indicated other- wise. Only lath-shaped gill rakers on the first gill arch are included in gill raker counts. Vertebral counts were determined from radiographs; the urostyle was included as one vertebra. Morphometric and meristic data were obtained from 211 specimens from the Atlantic, Pacific, and Indian oceans. Subsamples equal to the smallest N (32 in the Atlantic) were randomly taken from the Indian Ocean, the eastern North Pacific be- tween lat. 16° and 33°N and long. 117° to 126°W, and the eastern tropical Pacific between lat. 5° and 16°S and long. 77° to 90° W. Morphometric data were compared by analysis of covariance. Meristic data were compared by Tukey's multiple comparison procedure at the 5% level (Rothschild 1963). Material was examined from the following col- lections: Scripps Institution of Oceanography Manuscript accepted June 1975. FISHERY BULLETIN: VOL. 74, NO. 1. 1976. 142 BUTLER and AHLSTROM: NEW SPECIES, SCOPELENGYS CLARKEI (SIO); University of Southern California (USC); Institut fiir Seefischerei, Hamburg (ISH); Museum of Comparative Zoology (MCZ); U.S. Na- tional Museum (USNM); International Indian Ocean Expedition (IIOE); and Field Museum of Natural History (FMNH). GENUS SCOPELENGYS ALCOCK 1890 Type-species Scopelengys tristis Alcock, by monotypy. Description. — Head and body laterally com- pressed, eyes small, mouth large. Premaxillary, dentary, and palatines with bands of villiform teeth. Teeth absent at symphysis of upper and lower jaw. Vomer indented at head with teeth in two patches. Teeth on basihyal and on gill rakers. Anterior gill rakers reduced to toothed knobs. Maxillary extending past eye, expanded pos- teriorly. Supramaxillary present. Head and body covered with large deciduous, cycloid scales. Pec- toral fins lateral, extending beyond bases of pel- vic fins. Pelvic fins abdominal. Origin of dorsal fin about over base of pelvic fin. Anal fin com- pletely behind dorsal. Base of adipose fin over posterior half of anal fin. No photophores. No swim bladder in adults. D 11-13; A 12-14; P 12-17; V 8; Br 8; C principal 19 (1 + 17 + 1); procurrent C 6-9 dorsal and 7-8 ventral, hypurals (including parhypural) 4-1-3; epurals 3; uroneurals 2. Urostyle with two centra. As in all myctophiform fishes retaining two ural centra (personal observation reenforced by Rosen and Patterson 1969), the anterior ural cen- trum (labelled PUi -I- Ui in Rosen and Patterson) supports both the parhypural and the 2 inferior hypurals, whereas the posterior ural centrum (U2 in Rosen and Patterson) is associated exclu- sively with the 4 superior hypurals. Scopelengys tristis Alcock Scopelengys tristis Alcock 1890:302. Scopelengys dispar Garman 1899:254, plate 54, fig. 2-2d. Scopelengys lugubris Garman 1899:400, (syn- onym Scopelengys dispar). Scopelengys whoi Mead 1963:255, fig. 1. Description of Adult Body moderately slender, maximum body depth at nape, tapering to a narrow caudal peduncle (Figures lA, 2A); body depth at dorsal origin 11.7-19.8 (15.4); least depth at caudal peduncle 5.6-8.3 (6.8). Dorsal profile of head slightly con- cave; head length 24.4-33.9 (29.4); head depth 16.7-25.5 (20.2); eye small, orbit 3.1-4.2 (3.5); snout 7.5-10.1 (8.8). Width of maxillary as per- centage of its length 29.9-36.7 (32.2). Snout to: dorsal fin origin 36.1-47.0 (41.9); anal fin origin 56.4-72.6 (66.4); ventral fin origin 34.7-48.0 (41.8). Meristic Data.— D.11-13 (11.5); A 12-14 (13.0); P 14-17 (15.4); vertebrae 29-32 (30.8); total gill rakers 7-11 (8.5). Larvae Twenty-five specimens 3.5-10.3 mm were avail- able from the eastern Pacific. Measurements and counts were given for two eastern Pacific (EAS- TROPAC) specimens (6.2 and 6.4 mm SL) by Okiyama (1974) and the smaller specimen illus- trated. The larvae have a small round eye with- out choroid tissue, a snout as long proportionately as in adults, a gut terminating just forward of the anal fin, and a gas bladder, best seen on late preflection and flexion specimens, becoming obscured by overlying musculature in larger postflexion specimens. Rays form early in the pectoral fins; a 3.5-mm specimen has large pectorals extending posteriad to the anus; caudal fin forms and notochord flex- ion occurs between ca. 5 and 7 mm; dorsal and anal fins form during flexion; pelvic buds appear between 6.5 and 7.0 mm; fin formation, including procurrent caudal rays, complete by about 10.0 mm. Pigmentation is scanty; pigment develops on dorsal margin of peritoneal cavity, spreading lat- erally on prefiexion and flexion stage specimens but becoming obscured on postflexion larvae; preflexion larvae have a series of 6 or 7 small, inconspicuous spots along the ventral margin of the tail which are later obscured by the anal fin formation and lacking on late postflexion larvae; head pigment, best developed on postflexion specimens, consists of a striking horizontal bar extending from snout to eye and continuing be- hind the eye onto the operculum (Figure 3A). Distribution Records are from the tropical Atlantic, Pacific, and Indian oceans (Figure 4). The range is ex- panded poleward in the eastern part of the Pacific 143 FISHERY BULLETIN: VOL. 74, NO. 1 Figure l. — A: Scopelengys tristis, 126 mm, Velero IV, cruise 1238, stn. 18762/10. B: S. clarkei, 176 mm. SIO 73-160, holotype. Figure 2.— A: Scopelengys tristis, 126 mm, Velero IV, cruise 1238, stn. 18762/10. B: S. clarkei, 176 mm. SIO 73-160, holotype. 144 BUTLER and AHLSTROM: NEW SPECIES, SCOPELENGYS CLARKEl Figure 3. — A: Scopelengys tristis, 13.9 mm, from the western Indian Ocean. B: S. clarkei, 15.4 mm, from off Hawaii. — — — — — — — — — — — "" ~ ' % ^1 ^^ .^^ r 4-r^-^^ 4- M \ ^ V ^ Tyi/^/^rn-N-MQ^ \ UW 1 lA V A W Hj77-/^uTrH~L^ ^ a:*^ <^ * \ UTT XtUA \ iw w/74sZyA7WT^;ai^«^ ^ V J XU- VA\. .r^^Vi un-i~un~H4^ ^ \ /^^'■^~ld//rT~r~l~L j TTTT xti IX^ (- rm. S ?^ 1 . [ ^ 1 i i 1 ^\ ''^rvX'Vh\\. rr 1 1 jj -tft+J .- I AT I - 4^ -U-iu j Tttf+^ \\\t^ VTITr •1 11 -rh+ruji vrr r rni_L 1 .^ jTTTT ~j~f~'~~-LJj TTT 1 IlttH-UlN \ kWL^ \T 44-4- XWuJJJt /~l--^~*~~i~[/ ■1 Xl^^-^ 1 JJfT r tf-wZ/yTTV- A AW\r\ TTLL V /njJTFf-hh r-t-L^/" -TT^ i -LL 1 / /X ; rti~N~LrTf~U V WUrVr^rrr XU-r /^\ L Liliihr- -///"'"TV ■"^ ? TOX mffSM^ \^^^ '\\\\ llLr tj//7~/~/-/J'yy~TC/ U\ J JJTn DXtt 1 1/ ,rT7 T/VVVtCV/Ts// r\ 1 Th IT Vr^ -^m /~ T~r7~/~/yyry0// rr / t ^^^^^^^ \^^^^^^ -4ra i~f~f~^~/~i/yyyL/y^ \X>cS<<"^\wvV\'^ , / ^ LLL Ji//77 /rr-Zj/yy^/y^// N^^^^^vXIliL / W^^^^^^^ ^""^^^^^^ \II_LLZ7777 /Ty^y/yy^/y^/ V \l /// WvmiLWZZ=^ '7%///yCy'?C^ *^ ^ h ^^^ ^^^^ __ ^^" ^^•«*TS UM/THM (Ouiu. Mf * MK UfC ■o ^ I ? 1 Figure 4. — Distribution of Scopelengys tristis (circles) and S. clarkei (triangles). Small symbols indicate larvae. 145 FISHERY BULLETIN: VOL. 74, NO. 1 and Atlantic oceans and is narrowed along the equator to the west. In the western part of the Pacific and Atlantic, the species appears to be rare. Records of larvae from the Indian Ocean were presented by Nellen (1973). Geographic Variation Most of the specimens studied from each area were in poor condition, which added to the vari- ability of body proportions (Table 1). No significant difference was found in any morphometric character between regions. Meristic characters of 32 specimens each from four areas are presented in Table 2. Samples from the two eastern Pacific areas showed no significant differences between means of any meristic character Indian Ocean specimens differed from Pacific material in mean vertebral counts (30.4 vs. 30.9), pectoral-fin ray counts (15.2 vs. 15.7), and in gill raker counts (9.1 vs. 7.9). Atlantic material differed from Pacific material in dorsal-fin ray counts (12.0 vs. 11.4), in anal-fin ray coimts (13.4 vs. 12.8), in gill raker counts (9.2 vs. 7.9), and in pectoral-fin ray counts (15.0 vs. 15.7). Atlantic material differed from In- dian Ocean material in dorsal-fin ray counts (12.0 vs. 11.1), anal-fin ray counts (13.4 vs. 12.9), and vertebral counts (31.1 vs. 30.4). Although these differences are small, they are as marked be- tween Indian and Atlantic ocean specimens as be- TABLE 1. — Comparison of morphometric characters of Scopelengys tristis from four geographic areas {N = 32 for each area). Eastern Eastern North Pacific tropi cal Pacific Indian Ocean Atlantic Ocean Character Mean Range Mean Range Mean Range Mean Range Standard length (mm) 124.0 73.8-154.3 130.2 74.5-172.0 104.6 28.8-185.9 133.1 69.5-185.8 Head length 30.9 28.7- 33.9 28.5 28 2- 33.2 30.5 27.8- 33.5 27.6 24.4- 32.2 Head depth 20.2 18.6- 25.5 20.0 16.7- 22.5 20.9 17.2- 24.6 19.6 17.5- 21.7 Snout to origin of dorsal fin 42.7 39.6- 45.5 42.5 395- 46.2 41.9 36.1- 47.0 40.6 37.1- 43.6 Snout to ongin of pelvic fin 41.7 39.4- 45.3 428 39.4- 48.0 41.2 34.7- 46.1 41.6 37.8- 46.9 Snout to ongin of anal fin 663 63.0- 69.5 675 62.5- 72.1 65.2 56.4- 71.2 66.6 61.3- 72.6 Least depth of caudal peduncle 6.8 5.7- 8.1 7.0 6.1- 7.9 6.7 56- 8.3 6.9 5.9- 8.0 Body depth at origin of dorsal fin 16.0 13.3- 19.4 15.6 12.6- 19.3 14.5 11.7- 17.2 15.7 12.6- 19.8 Table 2.— Meristic data for Scopelengys tristis from the eastern North Pacific, eastern tropical Pacific, the Indian Ocean, and the Atlantic Ocean. Meristic character Area Numbers of character and frequency N Mean Overall mean Dorsal rays 10 11 12 13 32 32 32 32 32 32 32 32 64 64 64 64 32 32 32 32 63 63 63 63 11.44 11 34 11.12 11.97 12.75 12.81 12.94 13.41 15.59 15.88 15.19 1503 30 88 31.03 30.38 31.06 7,86 7.89 9.08 9.22 Eastern North Pacific Eastern tropical Pacific Indian Ocean Atlantic Ocean 18 21 28 2 14 11 4 29 1 11.47 Anal rays 11 12 13 14 Eastern North Pacific Eastern tropical Pacific Indian Ocean Atlantic Ocean 10 8 5 20 22 24 19 2 2 3 13 12.98 Pectoral rays 14 15 16 17 Eastern North Pacific Eastern tropical Pacific Indian Ocean Atlantic Ocean 2 33 20 48 62 24 32 14 2 7 12 15.42 Vertebrae 29 30 31 32 Eastern North Paafic Eastern tropical Pacific Indian Ocean Atlantic Ocean 1 1 5 3 20 3 28 25 9 21 4 2 7 30.84 Gill rakers 7 8 9 10 11 Eastern North Pacific Eastern tropical Pacific Indian Ocean Atlantic Ocean 10 17 1 52 39 4 2 1 4 49 47 3 7 2 12 2 8.51 146 BUTLER and AHLSTROM: NEW SPECIES, SCOPELENGYS CLARKEI tween specimens from these areas and from the Pacific. Because there is no clinal pattern in the variation and because of extensive overlap in all counts, no taxonomic importance was placed on the small meristic differences. Garman distinguished S. dispar from S. tristis by the lower dorsal and anal-fin ray counts: D 11 vs. 12 and A 12-11 vs. 13 (Garman 1899). The types of S. dispar are in poor condition but the anal fins appear to have 12 or 13 rays (Robert Schoknecht pers. commun.). The counts of S. dispar are within the range of S. tristis. Scopelengys dispar has been correctly considered a junior synonym by Parr (1928), Bohn (1939), and Norman (1939). Scopelengys lugubris Garman 1899:400, the specific name regarded as a lapsus calami by Bolin (1939), is a synonym ofS. dispar, hence ofS. tristis. Scopelengys whoi was described from the Carribbean Sea (Mead 1963). The diagnosis was based on a shorter head, higher number of anal fin rays (14 vs. 12-13), and the insertion of the pelvic fin in advance of the origin of the dorsal. According to Mead (1963), however, the head length is ". . .a poor measurement because of the condition of the opercular flap." The anal-fin ray count is within the range of S. tristis (Table 3). The insertion of the pelvic fin is a variable character in S. tristis. In most specimens the fin is inserted below the origin of the dorsal fin but in- sertion in advance of the dorsal is not uncommon. Based on this study, we conclude that S. whoi is a junior synonym of S. tristis. Study Material PACIFIC OCEAN ADULTS.— SIO 51-186 1 (134); SIO 64-21 6(78-148); SIO 65-243 2(122-134); SIO 64-997 1(122); SIO 65-244 1(75); SIO 55-229 9(31-113); SIO 65-206 1(92); SIO 60-212 4(20-133); SIO 52-309 2(36-56); SIO 73-170 1(49); SIO 73-171 1(30); SIO 55-265 1(54); SIO 65-620 1(139); SIO 65-606 4(92-151); SIO 65-220 5(14-138); SIO 65-611 17(85-176); SIO 51-84 3(74-123); SIO 69-497 6(92-170); SIO 72-186 8(73-179); SIO 65-215 1(121); SIO 54-124 1(147); SIO 52-367 1(145); SIO 60-232 1(168); SIO 65-213 3(88-158); SIO 60-219 2(42-170); SIO 55-246 4(65-140); SIO 68-579 1(140); SIO 53-235 1(154); SIO 51-146 3(127-144); SIO 65-603 17(62-160); SIO 55-244 2(159-167); SIO 72-195 17(88-175); SIO 65-608 14(43-200); SIO 72-193 2(106-169); SIO 72-192 18(10.2-177); SIO 60-216 2(42-76); SIO 60-218 1(48); SIO 66-355 1(135); SIO 69-19 1(24); SIO 72-182 1(90); SIO 66-407 1(42); SIO 64-24 1(116); SIO 60-234 1(69); SIO 64-13 1(113); SIO 52-409 1(65); SIO 59-202 1(83); SIO 52-90 1(113); SIO 64-15 1(85); SIO 63-444 1(103); SIO 60-243 4(18- 44); SIO 68-534 1(28); SIO 65-443 1(142); SIO 68-104 1(97); SIO 60-209 1(78); SIO 52-363 2(56- 115); SIO 64-28 3(95-144); SIO 57-43 1(126); SIO 65-237 1(128); SIO 61-32 2(105-106); SIO 63-42 1(109); SIO 66-30 1(113); SIO 51-45 1(132); SIO 60-215 7(19-94); SIO 52-32 1(150); SIO 50-270 2(110-115); SIO 51-77 1(110); SIO 51-189 1(120); SIO 54-82 1(107); SIO 54-102 2(116-147); USC Vel- ero IV, cruise 1238, stn. 18762/10; MCZ 41695 2(121-141); USNM 135842 1 (X-ray); MCZ 28058 1 (X-ray) (lectotype S. dispar Garman). PACIFIC OCEAN LARVAE^.— Larvae taken at 17 EASTROPAC stations and 2 CalCOFI sta- tions as follows: EASTROPAC stations 11.282 1(4.8); 13.105 1(5.5); 13.172 2(6.4, 6.8); 20.018 1(5.5); 30.114 2(4.0, 4.5); 45.032 1(8.1); 45.073 1(6.0); 45.078 1(10.3); 45.293 1(6.6); 45.316 1(6.9) 46.034 1(6.2); 46.096 2(6.7, 6.9); 47.001 1(5.2) 47.005 4(3.5-4.3); 47.035 1(7.0); 47.040 1(5.3) 47.065 1(9.2); CalCOFI 7205-20.127 1(5.0); 4907- 112 1(9.1). ATLANTIC OCEAN.— MCZ 41638 l(X-ray) ^Station data in EASTROPAC Information Paper 6 and Ahl- strom (1972). Table 3. — Means and differences among means of meristic counts of Scopelengys tristis from four areas (eastern North Pacific, ENP; eastern tropical Pacific, ETP; Indian Ocean, 10; and Atlantic Ocean, AO) and S. clarkei. S. tristis S. c/a rkei Difference in counts between S. clarkei and S. tristis Average Least Overall Mean Greatest differences among regions character Range mean ENP ETP 10 AO Range Mean difference difference Dorsal rays 11-13 11.47 11.4 11.3 11.1 12.0 0.9 13 13.0 1.5 1 .0-AO Anal rays 12-14 12.98 12.8 12.8 12.9 13.4 0.6 14 14.0 1.0 0.6-AO Pectoral rays 14-17 15.42 15.6 15.9 15.2 15.0 0.9 12-13 12.9 2.5 2.1-AO Vertebrae 29-32 30.84 30.9 31.0 30.4 31.1 0.7 34-35 34.1 3.3 30-AO Gill raker 7-11 8.51 7.9 7.9 9.1 9.2 1.3 7-10 8.2 0.3 0.3-EP 147 FISHERY BULLETIN: VOL, 74, NO. 1 (type S. whoi Mead); USNM 20678, 5(152-164), eastern tropical Atlantic, lat. 07°32'N, long. 20°54'W, 1813-2125, 12 April 1971, 1,300 m, 1,600-mesh Engels trawl, RV Walther Herwig; ISH 623/68, 7(73-162), eastern tropical Atlantic, lat. 12°07'N, long. 23°08'W, 30 January 1968, 2,000 m, 1,600-mesh Engels trawl, RV Walther Herwig; ISH 2095/71, 1(167), eastern tropical At- lantic, lat. 05°30'S, long. 16°28'W, 9 April 1971, 1,950 m, 1,600-mesh Engels trawl, RV Walther Herwig; ISH 2447/71, 12(86-160), eastern tropical Atlantic, lat. 04°38'N, long. 19°21'W, 13 April 1971, 756 m, 1,600-mesh Engels trawl, RV Walther Herwig; ISH 3099/71, 5(132-160), eastern tropical Atlantic, lat. 07°32'N, long. 20°54'W, 14 April 1971, 1,300 m, 1,600-mesh Engels trawl, RV Walther Herwig; ISH 2942/71, 2(134-155), eastern tropical Atlantic, lat. 23°47'N, long. 20°59"W, 19 April 1971, 2,100 m, 1,600-mesh Engels trawl, RV Walther Herwig. INDIAN OCEANl— IIOE 7001 Anton Bruun III, 16 (25-94); IIOE 7004 Anton Bruun III, 7 (32-120); IIOE 7012 Anton Bruun III, 2 (23-25); IIOE 7022 Anton Bruun III, 1 (113); IIOE 7027 Anton Bruun III, 1(138); IIOE 7037 Anton Bruun III, 2 (40-87); IIOE 7046 Anton Bruun III, 3 (66- 179); IIOE 7143 Anton Bruun VI, 1 (131); IIOE 7147 Anton Bruun VI, 28 (28-142); IIOE 7153 Anton Bruun VI, 4 (42-161); IIOE 7154 Anton Bruun VI, 12 (48-114); IIOE 7163 Anton Bruun VI, 12 (28-152); IIOE 7165 Anton Bruun VI, 3 (22-27); IIOE 7206 Anton Bruun VI, 1 (27); IIOE 7277 Anton Bruun VI, 2 (40-87). Scopelengys clarkei n.sp. Holotype SIO 73-160, female (176 mm), central Pacific, lat. 29°56.0'N, long. 144°56.6'W, 0224-0556 h; 14 February 1973, 10-foot IKMT, 0-1,000 m, RV Alexander Agassiz. Paratypes USNM 210707, male (160 mm), central Pacific, lat. 21°20-30'N, long. 158°20-30'W, 1204-1637 h; 15 September 1970, 10-foot IKMT, 0-1,000 m, RV El Pescadero I; USNM 210706, male (156 mm). "Station data in Nafpaktitis and Nafpaktitis (1969). 148 central Pacific, lat. 24°N, long. 139°W, 0049-0149 h; 29 November 1972, 50-foot Universal trawl, 0-494 m, RV David Starr Jordan; FMNH 76366, female (154 mm), central Pacific, lat. 22°N, long. 158°W, 1240-1645 h; 13 November 1969, 10-foot IKMT, 0-800 m. Other Materials Studied SIO 51-76, female (109 mm), southeast of Guadalupe Island, 17 March 1951, 10-foot IKMT, 0-549 m; FMNH 76367, juvenile (65 mm), central Pacific, lat. 21°20-30'N, long. 158°20-30'W, 0421- 0600 h; 27 February 1971, % Cobb trawl, 0-150 m, RV Townsend Cromwell; FMNH 76368, juvenile (42 mm), central Pacific, lat. 21"20-30'N, long. 158°20-30'W, 2236-0105 h; 16-17 November 1969, 10-foot IKMT, 0-250 m, RV Teritu; T. Clarke, 71-3-9, larva (15 mm), central Pacific, lat. 21°20- 30 'N, long. 158°20-30'W, 1252-1645 h; 2 March 1971, 10-foot IKMT, 800-900 m, Y(N El Pescadero I, retained at the Southwest Fisheries Center. Adult Morphology Body proportions of the holotype are given first, followed, in parentheses, by range of values for holotype and three paratypes. Body slender; greatest body depth at origin of dorsal fin, 19.0 (18.4-19.0), tapering to a moderately deep caudal peduncle (Figures IB, 2B), less than three in length of head, 9.4 (9.4-10.2). Head slightly con- cave in dorsal profile, head length 25.4 (24.5- 26.4); head depth 17.6 (16.7-17.9); eye small, orbit 3.0 (2.9-3.6); interorbital width 8.7 (7.6-8.7); snout about one-third of head length, 8.3 (7.7- 8.8); length of maxillary 11.3 (11.3-12.6), greatest width of maxillary 3.1 (2.8-3.6). Snout to: dorsal fin origin 43.5 (39.0-43.5); anal fin origin 68.6 (65.1-69.6); pelvic fin origin 40.2 (40.2-43.4). Length of dorsal fin base 17.3 (17.0-19.4); length of anal fin base 16.0 (16.0-17.9). Color dark brown, preserved in alcohol. Meristic Data Counts are based on all seven specimens. D 13 (7); A 14 (6), ? (1); P 13/13 (6), 13/12 (1); V 8/8 (7); principal C 10 + 9 (7), procurrent C 7-8/6-9; bran- chiostegal rays 8/8 (7); vertebrae 15 + 19 = 34 (6), 15 + 20 = 35 (1); gill rakers 1-2 + (6-8) = 7-10 (mean 8.2). BUTLER and AHLSTROM: NEW SPECIES, SCOPELENGYS CLARKEl Larvae A single specimen was available, 15.4 mm SL (Figure 3B). Body shape similar to that of adults but with a relatively larger head — length 35.7 and depth 25.0; eye 5.5; snout 12.8; body depth 23.5; least depth of caudal peduncle, 14.6. Fin origins farther back on body than in adults. Snout to: dorsal fin origin 50.0; anal fin origin 72.8; pelvic base 53.6. Pigment confined to head and nape, extensively developed on the operculum and lower jaw; a small pigment patch on upper jaw behind eye; several melanophores on mid-brain; body pigment confined to nape and to a patch an- terior to pectoral base. Name This species from the central North Pacific is named in honor of Dr. Thomas A. Clarke of the Hawaii Institute of Marine Biology. COMPARISON OF SCOPELENGYS CLARKEl AND SCOPELENGYS TRISTIS Scopelengys clarkei differs from S. tristis in meristic counts, in some morphometric charac- ters, and in larval pigmentation. For differences in meristic characters, refer to Table 3. Most marked differences are in average number of vertebrae — 34.1 (S. clarkei) vs. 30.8 = 3.3; average pectoral-fin ray count — 12.9 {S. clarkei) vs. 15.4 =2.5; and average dorsal-fin ray count — 13.0 vs. 11.5 =1.5. As regards morphomet- ric characters, S. clarkei has a deeper caudal peduncle, a narrower maxillary, and a more fusiform body. Several distinctive adult charac- ters also can be recognized in larger larvae of the two species, i.e., differences in meristic characters and depth of caudal peduncle. The most striking differences between larvae of the two species are found in the head pigment which is restricted to an eye-bar in S. tristis, as compared with the scat- tered pigment on the operculum, lower jaw, etc. of S. clarkei. The two species are similar in general body shape, head size, eye size, length of snout, and position of fins on the body. Scopelengys clarkei has its greatest body depth at the dorsal origin, whereas S. tristis has its greatest body depth at the nape. When an analysis of covariance was performed on the morphometric characters of 7 S. clarkei and 32 S. tristis from the eastern North Pacific, eastern tropical Pacific, Indian, and Atlantic oceans, only the least depth of caudal peduncle E 6 z o liJ < O < 20p 18- 16- 14 12 10 8 6 4 2 Scopelen gys clorkei Scopelen gys tristis INDIAN OCEAN EASTERN TROR PAG. EASTERN NORTH PAG EASTERN TROP. ATL. ID 20 30 40 50 60 70 80 90 100 no 120 STANDARD LENGTH (mm) 30 140 150 160 Figure 5.— Regression of least depth of caudal peduncle on standard length of Scopelengys tristis and S. clarkei. 149 FISHERY BULLETIN: VOL. 74, NO. 1 showed a significant difference at the 1% level, F = 3.72, between the two species. The Atlantic specimens of S. tristis had counts for four characters that were closer to those of S. clarkei than were counts of these characters from other geographic areas. These differences in counts between Atlantic S. tristis and S. clarkei were as follows: dorsal fin rays 1.0 (12.0 vs. 13.0), anal fin rays 0.6 (13.4 vs. 14.0), pectoral fin rays 2.1 (15.0 vs. 12.9), and vertebrae 3.1 (31.1 vs. 34.2). Differences of two in pectoral-fin ray counts and three for vertebrae are much greater than the regional variability found among specimens of S. tristis. ACKNOWLEDGMENTS We are grateful to the following individuals and institutions for the loan of specimens: T. Clarke, Hawaiian Institute of Marine Biology, Kaneohe, Hawaii; G. Kreftt, Institut fiir Seefischerei, Hamburg, Germany; R. H. Rosen- blatt, M. A. Barnett, J. Copp, and D. Dockins, Scripps Institution of Oceanography, La Jolla, Calif; B. G. Nafpaktitis, University of South- ern California, Los Angeles, Calif; R. K. Johnson, Field Museum of Natural History, Chicago, 111.; W. Nellen, Institut fiir Meereskunde, Kiel, Germany; R. Schoknecht and M. M. Dick, Museum of Comparative Zoology, Cambridge, Mass.; R. H. Gibbs and W. R. Taylor, Division of Fishes, National Museum of Natural His- tory, Washington, D.C. We give special thanks to H. G. Moser, Southwest Fisheries Center, Na- tional Marine Fisheries Service, La Jolla, and to C. L. Hubbs, Scripps Institution of Oceanography, for their advice and criticisms and to R. Schok- necht for examining the types of S. whoi Mead and S. dispar Garman. LITERATURE CITED Ahlstrom. E. H. 1972. Kinds and abundance of fish larvae in the eastern tropical Pacific on the second multivessel EASTROPAC survey, and observations on the annual cycle of larval abundance. Fish. Bull, U.S. 70:U53-1242. Alcock, a. 1890. On the bathybial fishes of the Arabian Sea, obtained during the season 1889-90. Ann. Mag. Nat. Hist., Ser. 6, 6:295-311. BOLIN, R. L. 1939. A review of the myctophid fishes of the Pacific coast of the United States and of lower California. Stanford Ichthyol. Bull. 1:89-156. CLARKE, T. A. 1973. Some aspects of the ecology of lanternfishes (Myc- tophidae) in the Pacific Ocean near Hawaii. Fish. Bull., U.S. 71:401-434. Garman, S. 1899. Reports on an exploration off the west coasts of Mexico, Central and South America, and off the Gala- pagos Islands, in charge of Alexander Agassiz, by the U.S. Fish Commission steamer "Albatross," during 1891, Lieut. Commander Z. L. Tanner, U. S. N., commanding. XXVI. The fishes. Mem. Mus. Comp. Zool., Harvard Coll. 24:1-431. HUBBS, C. L., AND K. F. LAGLER. 1958. Fishes of the Great Lakes region. Revised ed. Cran- brook Inst. Sci. Bull. 26, 213 p. Mead, G. W. 1963. Observations on fishes caught over the anoxic waters of the Cariaco Trench, Venezuela. Deep-Sea Res. 10:251-257. Nafpaktitis, B. G., and M. Nafpaktitis 1969. Lanternfishes (family Myctophidae) collected during cruises 3 and 6 of the RA^ Anton Bruun in the Indian Ocean. Bull. Los Ang. Cty. Mus. Nat. Hist. Sci. 5, 79 p. Nellen, W. 1973. Fischlarven des Indischen Ozeans. "Meteor" Forsh-Ergebnisse, Ser. D, 14:1-66. Norman, J. R. 1939. Fishes. John Murray Exped. 1933-1934, Sci. Rep. 2(1):1-116. OKIYAMA, M. 1974. The larval taxonomy of the primitive Myctophiform fishes. In J. H. S. Blaxter (editor). The early life history of fish, p. 609-621. Proc. Int. Symp. Dunstaffnage Mar. Res. Lab., Scott. Mar. Biol. Assoc, Oban, Scotl. May 17-23, 1973. Springer- Verlag, N.Y. Parr, a. e. 1928. Deepsea fishes of the order Iniomi from the waters around the Bahama and Bermuda islands. With anno- tated keys to the Sudidae, Myctophidae, Scopelarchidae, Evermannellidae, Omosudidae, Cetomimidae and Ron- deletidae of the world. Bull. Bingham Oceanogr. Col- lect., Yale Univ. 3(3):1-193. Rosen, D. E., and C. Patterson. 1969. The structure and relationships of the Paracanthop- terygian fishes. Bull. Am. Mus. Nat. Hist. 141(3 1:357-474. Rothschild, B. j. 1963. Graphic comparisons of meristic data. Copeia 1963:601-603. 150 WEIGHT LOSS, MORTALITY, FEEDING, AND DURATION OF RESIDENCE OF ADULT AMERICAN SHAD, ALOSA SAPIDISSIMA, IN FRESH WATERS Mark E. Chittenden, Jr.^ ABSTRACT Linear regression equations are given for each sex for the regressions of total weight, somatic weight, and gonad weight on length prior to spawning, and for total weight on length after prolonged stay in fresh water Most shad began to return seaward by late June and probably had spent a maximum of about 2 mo in fresh water Many fish, however, remained near the spawning grounds well into summer; and many died near the spawning grounds, probably from starvation. Opportunistic feeding oc- curred on "planktonic" items, but adult shad do not regularly obtain energy sufficient to maintain their weight in fresh water. Weight loss was related to sex and increased with increasing size. Mean length males and females averaged 45 and ^T7c total weight loss, respectively. Daily somatic weight loss was at least 5.75 g for males of average size and 12.47 g for females. The anadromous American shad, Alosa sapidls- sima, an important commercial and sport fish, ranges widely on the Atlantic and Pacific coasts of North America. There is much literature on this fish, but little of it pertains to adults in fresh wa- ter, except for aspects of their spawning and popu- lation dynamics. In the course of other studies on the Delaware River from 1960 to 1968, I made many opportunistic observations on weight loss, mortality, feeding behavior, and duration of resi- dence of adult shad on their spawning grounds in fi'esh water. This paper summarizes those obser- vations and presents data on total-fork length conversion, regressions of total weight, somatic weight and gonad weight on length prior to spawning, and regressions of total weight on length after spawning. MATERIALS AND METHODS Adult shad were collected during their spawn- ing runs at Lambertville, N.J., 22.5 km above tidal water (but far downstream of the present- day spawning grounds) using a 76-mm stretch- mesh, 107 m long and 3.6 m deep haul seine that was paid out from a boat and landed about 400 m downstream. Sampling occurred at 3- or 4-day in- tervals from 5 April to 19 May 1963, from 20 'Based on part of a dissertation submitted in partial fulfill- ment of the requirements for a Ph.D. degree, Rutgers Univer- sity, New Brunswick, N.J. ^Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, TX 77840. March to 18 May 1964, from 26 March to 7 May 1965, and from 27 March to 19 May 1966. Data for the period 1959-62 were obtained from rotenone surveys (hereinafter referred to as the Tri-State Surveys) during July and August by the States of New Jersey, New York, and Pennsylvania in cooperation with the U.S. Fish and Wildlife Service. I examined grossly the stomach contents of many adults captured during the Tri-State Sur- veys in mid-July 1961, most of the 526 fish col- lected at Lambertville and many fish captured on the spawning grounds after 1962. Length and total weight were determined on most fish in 1961 and 1962 and on all fish thereaf- ter. Gonad weight was measured after 1962. Length, always taken in inches, was measured as fork length during 1961 and 1962 and as total length thereafter. To develop conversion factors, both measurements were taken on 490 adults col- lected at Lambertville during 1963 and 1964 and on 100 young captured in summer 1966. Total weight was measured in pounds (to the closest 0.1 lb) during 1961-63 but in grams thereafter Gonad weights were always taken in grams (to the closest 0.1 g). All measurements were converted to grams, millimeters, and fork lengths for pre- sentation herein. Regression analyses and related statistics were calculated using computer program BMD-03R (Dixon 1967). All regressions presented herein were significant at a = 0.005. The coefficient of Manuscript accepted July 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 1976. 151 FISHERY BULLETIN: VOL. 74, NO. 1 determination (Steel and Torrie 1960) was used to estimate the amount of variation in y associated with variation in x. Residuals were used to examine the data for differences due to categories of classification such as year of collection. Size ranges are given within which regressions were linear. When first referred to, locations are followed in parentheses by their approximate distances in kilometers upstream from Marcus Hook, Pa., which is situated about 90 km downstream from the fall line at Trenton, N.J. and near the transi- tion between brackish and fresh water. RESULTS AND DISCUSSION Total and Fork Length Conversion The relationship between total length (TL) and fork length (FL) for 590 young and adult fish was linear, and 99.96*7^ of the variation in one mea- surement was explained by variation in the other. Regression equations were FL = 1.28 + 0.88 TL and TL = 1.00 + 1.13 FL. Extreme deviations from regression were about ±7.6 mm for adults and less for young. The slope of the regression of fork length on total length coincides with La Pointe's (1958) factor of 0.894 to convert total length to fork length. Total Weight- Length Relationships Prior to Spawning The relationships between total weight (TW) and length determined for fish captured at Lam- bertville were TW = 1,106.77 + 8.09 (FL - 427.98) for 268 males and TW = 1,737.26 + 11.54 (FL - 476.71) for 244 females. About 81^f (males) and 78% (females) of the variation in total weight was associated with variation in length. Valid ranges for linear interpolation were about 330- 520 mm for males and 410-550 mm for females. The observed arithmetic mean weights with 95% confidence limits were 1,107 ± 36 g for males and 1,737 ± 45 g for females. The smallest males were 272 and 680 g and the smallest female was 1,089 g. The heaviest male and female fish were 1,905 and 2,585 g, respectively. Somatic Weight-Length Relationships Prior to Spawning The relationships between log somatic weight 152 (SW) and length determined for 85 males and 130 females captured at Lambertville in 1964 and 1965 were logio SW - 3.0047 + 0.0036 (FL - 428.20) for males and logio SW = 3.1807 + 0.0029 (FL - 480.73) for females. About 91% (males) and 81% (females) of the variation in log somatic weight was associated with variation in length. Valid ranges for linear interpolation were about 360-500 mm for males and 410-540 mm for females. Mean somatic weights with 95% confidence limits were 1,011 ± 56 g for males and 1,516 ± 38 g for females. Gonad Weight- Length Relationships Prior to Spawning The relationships between log total gonad weight (TGW) and length determined for 267 males and 244 females captured at Lambertville were logio TGW = 1.8633 + 0.0033 (FL - 428.43) for males and logio TGW = 2.3892 + 0.0024 (FL - 476.93) for females. Valid ranges for linear in- terpolation were about 330-520 mm for males and 410-550 mm for females. About 45% (males) and 26% (females) of the variation in log total gonad weight was associated with length variation. Much variation in gonad weight, especially for females, is not explained by the regression equa- tions. Much gonad development occurs during the spawning run (Chittenden 1969), and residual plots suggested that gonad weights were heavier in 1963 than in 1964. These factors account for some unexplained variation in gonad weight. Mean total gonad weights with 95% confidence limits were 73 ± 7 g for males and 245 ± 22 g for females. Duration of the Freshwater Residence Most fish begin to return seaward by about late June. I observed hundreds of adults near Han- cock, N.Y. (403) until 17 June 1964, but very few were present on 14 July. Most fish had died or migrated seawards during the interim period. De- laware River shad runs begin in early April at Lambertville and the peak occurs about 1 May, depending upon the degree of pollution near Philadelphia (34) (Chittenden 1969). This suggests most fish probably spend a maximum of 2 mo in fi^esh water before returning seaward, in agreement wdth Bean's (1892, 1903) observations. Many fish remain near the spawning grounds well into summer. The Tri-State Surveys cap- CHITTENDEN: ADULT AMERICAN SHAD IN FRESH WATER tured many adults during midsummer between Skinners Falls, N.Y. (348) and Minisink Island, N.J. (266): 538 fish were captured at three sta- tions in mid-July 1961; 237 fish were captured at two stations in mid-July 1962; 30 adults were captured near Milford, Pa. (269) on 7 August 1959, and 13 were captured there on 1 August 1961. Upstream Mortality There was a large mortality of shad upstream near the spawning grounds about the end of the spawning period. In 1963, I observed many dead fish along the banks or in shallow water on 5 July; and a surface gill net set overnight at Mil- ford, Pa. on 22 June captured 15 fish that ap- peared to have been dead for several days. In 1964, dead shad first appeared in the East Branch near Hancock about 14 June; on that date, I walked the bank for about 0.8 km and observed 26 dead fish within 10 m of the shoreline. I ob- served hundreds of dead shad on 8 July 1964 dur- ing a 19-km float from Matamoras, Pa. (274) to Dingmans Ferry, Pa. (258). I frequently saw dead fish in shallow water during August. Shad may die before being completely spent. Some dead fish examined near Hancock had ovaries about a fourth the size of those in fish captured at Lambertville. The ovaries of these dead fish contained many translucent eggs, a criterion (Milner 1874; Brice 1898; Leach 1925) indicating that the fish is ripe. shad, 2) 6 darters and 17 shad, 3) 46 shad, and 4) 15 shad. Young shad were the first fish to react to rotenone, and the adults probably foraged on dis- tressed and dying young. Weight Loss in Fresh Water Much weight was lost while the adult shad were in fresh water Fish captured near Hancock had noticeably lost weight by late May, and they became more emaciated the longer they remained in fresh water. Tri-State Survey data obtained 10-13 July 1961 from Belvidere, N.J. (197) to Han- cock, N.Y. and 16-17 July 1962 at Minisink Island and Skinners Falls were used to estimate the changed weight-length relationship for each sex. The relationships between total weight and length of these fish were TW = 536.34 + 3.24(FL - 407.34) for 296 males and TW = 661.29 + 3.01(FL - 451.18) for 19 females. Valid ranges for linear interpolation were about 265-450 mm for males and 340-475 mm for females. About 66% (males) and 63% (females) of the variation in total weight was associated with variation in length. These regressions explain less variation in total weight than the 80% explained for fish taken at Lam- bertville. The average percentages of total weight loss in fresh water were estimated by comparing Lam- bertville and Tri-State Survey regression means at different lengths for each sex (Figure 1). The Feeding Behavior in Fresh Water Feeding did occur in freshwater, at least near the upstream spawning grounds. The stomachs of most shad captured at Lambertville were empty, but a few contained a slight amount of amorphous material. Stomachs of fish collected upstream from Port Jervis, N.Y. (295) in late May and June frequently contained a few insects. I observed a large mayfly hatch in late May 1964 near Han- cock: hundreds of adult shad were rising to the surface, apparently to feed, and the stomachs of many fish (about 50) captured by angling were packed with mayflies. Similar surface feeding be- havior was observed on several other occasions, although fish were not collected to confirm feed- ing. Many adults captured during the Tri-State Surveys contained recently eaten young shad and shield darters, Percina peltata.For example, four stomachs contained: 1) 2 darters and 9 young I en (/) O z UJ 5 70 60 50 40 30 20 10 9 300 400 500 FORK LENGTH (mm) 600 FIGURE 1. — Minimum average total weight loss of American shad in fresh water. 153 FISHERY BULLETIN: VOL. 74, NO, 1 average percent weight loss depended upon length. Large fish lost a greater percentage than small fish. Average total weight loss was from 30 to 50% for 359-493 mm FL males and from 48 to 62% for 421-531 mm FL females, sizes which closely approximate the observed size range of fish in the 1963 and 1964 runs (Chittenden 1969). The observed mean fork lengths of fish captured at Lambertville were 428 mm for males and 477 mm for females, based upon the regression equa- tions, and these sizes averaged 45 and 57% total weight loss, respectively. Somatic weight loss, a better measure of the toll taken by the spawning migration, was esti- mated by subtracting the predicted total gonad weight from the predicted total weight at Lam- bertville before making a comparison with the Tri-State Survey total weight regressions. No correction was made for the gonads of fish cap- tured during the Tri-State Surveys; however, these were a negligible fraction of the total weight. The total testes weights of 15 males col- lected near Hancock on 14 July 1964 and on 21, 24 June and 1 July 1965 ranged from 3.7 to 27 g and averaged 15.9 g while the total ovary weights of 3 females collected then varied from 18.2 to 35 g and averaged 27.1 g. The average percentage of somatic weight loss in males was 24% at 359 mm, 46% at 493 mm, and 42% for the mean-sized male of 428 mm. For females, somatic weight loss was 38% at 421 mm, 56% at 531 mm, and 50% for the mean-sized female of 477 mm. Absolute daily weight loss was estimated from the duration of the freshwater residency. Fish captured during the Tri-State Surveys had prob- ably been upstream about 75 days. This approxi- mates their maximum stay in fresh water because the peak of the run at Lambertville is about 1 May (Chittenden 1969), and most fish move sea- ward from the Hancock area by late June. There- fore, the average daily loss in somatic weight of males was 1.63 g at 359 mm, 9.37 g at 493 mm, and 5.75 g for mean-sized males of 428 mm. For females the average daily loss in somatic weight was 5.75 g at 421 mm, 18.87 g at 531 mm, and 12.47 g for mean-sized females of 477 mm. Daily weight loss can be used to suggest how long fish of different sizes can remain in freshwa- ter before death. The amount of weight loss which results in death of shad is not known, but death occurs in many animals when weight loss exceeds 40% (Curtis 1949). Assume 50% for simplicity in calculation, this may not be quite correct, but it may be conservative and the size pattern, at least, remains the same if the percentage is a con- stant. From this, males could remain 154 days at 359 mm, 81 days at 493 mm, and the average sized male (428 mm) could remain 90 days. Females could remain 100 days at 421 mm but only 68 days at 531 mm, and the mean-sized female of 477 mm could remain 75 days. There is apparently little difference in the amount of time an average to maximum-sized fish can spend in fresh water before death, but small fish can sur- vive much longer. GENERAL DISCUSSION Weight loss data presented herein agrees reasonably with those of Leggett (1972) who noted that his figures were probably underesti- mates. The present figures ignore weight loss in the 100-km migration between Marcus Hook and Lambertville and may be based on a longer than average stay in fresh water. Both factors tend to underestimate weight loss which affects related estimates. Many shad apparently remain upstream near the spawning grounds well into the summer. However, the percentage they comprise of the run is unknown. A few fish remain far upstream until late fall. Bishop (1936) captured emaciated indi- viduals 305-330 mm long near Hancock in November These fish must have migrated up- stream during the previous spring, because low dissolved oxygen water near Philadelphia pre- sents a virtually impassable barrier through summer and fall (Ellis et al. 1947; Sykes and Lehman 1957; Chittenden 1969). Nichols (1959) captured an emaciated male during October in the Connecticut River and estimated it had been in freshwater at least 120 days. I captured an emaciated male (287 mm FL, 194 g) in fresh water in the James River, Va. on 7 October 1969. The finding of little or no food in adults col- lected at Lambertville is similar to the reports of Bean (1903), Leim (1924), Leach (1925), Hilde- brand and Schroeder (1928), Moss (1946), and Hildebrand (1963) that adults take little or no food while ascending rivers. My observations of instances of intensive feeding while upstream are exceptional, although Atkinson (1951) reported an artificial instance of feeding in freshwater ponds. Adult shad at sea feed largely on planktonic forms such as copepods and mysids (Leim 1924; Hildebrand and Schroeder 1928; 154 CHITTENDEN: ADULT AMERICAN SHAD IN FRESH WATER Bigelow and Schroeder 1953; Hildebrand 1963; Leim and Scott 1966), although Holland and Yel- verton (1973) reported that they occasionally take large amounts offish. Atkinson (1951) attributed the general absence of food in the stomachs of adults to their planktonic feeding habit and the absence of suitably large plankton in fresh water. My observations suggest that adult shad would opportunistically feed in freshwater if suitably large "planktonic" forms were readily available. Although adults feed opportunistically in fresh water, they do not regularly obtain energy sufficient to maintain their weight and must use energy reserves accumulated during their life at sea to support migration in fresh water, final de- velopment of the gonads, and spawning. Adults use up their somatic substance at a size and sex dependent rate of at least about 1.6-18.9 g/day. Their physical activity deteriorates greatly as Fowler (1908) and Walburg (1960) noted. Death by starvation may occur when weight loss ex- ceeds 40% (Curtis 1949), and this is probably the main cause of the mortality I observed on the spawning grounds. Further work is needed to quantitatively describe upstream mortality, but its magnitude would appear large as Bean (1892, 1903) and Anonymous (1902) also observed in the Delaware River and Walburg (1960) observed in the St. Johns River, Fla. Weight loss was related to sex and size in agreement with Leggett (1972). The apparent re- lationship between weight loss and sex, however, may not be direct. Metabolic rate, in general, in- creases with temperature within limits. Leggett (1972) noted that females tend to migrate later and at a higher temperature than males and suggested that temperature was responsible for the apparent sex difference in weight loss. The relationship between size and total metabolism in a wide variety of organisms can be expressed as: log M = \oga + b \ogW where M is total metabolism and W is weight (Paloheimo and Dickie 1966; Prosser 1973). The relationship between metabolic rate and size can be expressed (Prosser 1973) as: log M/W = log a + (6 - 1) log W. From the latter expression it follows that a b value less than 1.0 implies that the metabolic rate decreases with increasing size, while a b value greater than 1.0 indicates that the metabolic rate increases with size. The value generally found for b is about 0.8 (Paloheimo and Dickie 1966; Pros- ser 1973), although Fry (1971) cautions that this value should not yet be accepted as dogma. Pres- ent findings on the relationship between size and weight loss in shad on their spawning migration are consistent with a b value greater than 1.0. Calculations made herein obviously assume that adult fish of all sizes are in fresh water the same length of time. If 6 is not greater than 1.0, we must conclude that: 1) small adults enter fresh water later than large fish and thus are in fresh water for a shorter period of time, or 2) small fish make better use of available freshwater food resources. Estimates of the time that adults can remain in fresh water suggest that only small fish can sur- vive upstream into the fall. The small fish I cap- tured in the James River in October apparently had lost only about 33-39% of its weight in com- parison with the Delaware River somatic weight regression at Lambertville and an unusually small fish (285 mm FL, 288 g) captured at Lam- bertville. It is noteworthy that, except for Nichols' (1959) report of a 430 mm FL male, the adult shad reported in fresh water during the fall have all been males about 305 mm long. Fish this small, however, are rare in the age compositions reported from many Atlantic Coast rivers (Talbot 1954; Fredin 1954; Walburg 1956, 1957, 1960, 1961; Sykes 1956; Sykes and Lehman 1957; Wal- burg and Sykes 1957; La Pointe 1958; Nichols and Tagatz 1960; Nichols and Massmann 1963; God- win 1968; Leggett 1969; Chittenden 1975). ACKNOWLEDGMENTS For assisting in field collections, I am deeply grateful to J. Westman, J. Hoff, J. Harakal, D. Riemer, J. Barker, F. Bolton, R. Coluntuno, K. Compton, R. Gross, C. Masser, R. Stewart, J. Miletich, S. Hoyt, L. Schulman, H. Dinje, H. Buckley, J. Musick, M. Bender, J. Gift, C. Townsend, R. Bogaczk, and K. Marcellus of or formerly of Rutgers University, Harvard Univer- sity, New Jersey Division of Fish and Game, and New York Department of Environmental Con- servation. Fred and William Lewis, Jr. generously gave permission to collect shad at their fishery at Lambertville and frequently assisted in seining. J. D. McEachran and W H. Neill of Texas A&M 155 FISHERY BULLETIN: VOL. 74, NO. 1 University reviewed the manuscript. The U.S. Bureau of Sport Fisheries and Wildhfe, New Jer- sey Division of Fish and Game, Pennsylvania Fish Commission, and New York Department of Environmental Conservation permitted use of data collected during the Tri-State Surveys of the Delaware River. Financial support was provided, in part, by Rutgers University, The Sport Fishing Institute, Delaware River Basin Commission, and U.S. Public Health Service. One collection was made while the author was employed at the Virginia Institute of Marine Science. LITERATURE CITED ANONYMOUS. 1902. Rep. Pa. State Comm. Fish. 1902. ATKINSON. C. E. 1951. Feeding habits of adult shad [Alosa sapidissima) in fresh water. Ecology 32:556-557. Bean, T. H. 1892. The fishes of Pennsylvania. Rep. Pa. Comm. Fish. 1889-90-91, p. 1-149 1903. Catalogue of the fishes of New York. N.Y. State Mus. Bull. 60, 784 p. BIGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53, 577 p. BISHOP, S.C. 1936. Fisheries investigations in the Delaware and Sus- quehanna Rivers. In A biological survey of the Delaware and Susquehanna watersheds, p. 122-139. N.Y. Conserv. Dep. Suppl. 25th Annu. Rep., 1935. BRICE, J. J. 1898. A manual of fish-culture, based on the methods of the United States Commission of Fish and Fisher- ies. U.S. Comm. Fish Fish. Part 23, Rep. Comm. 1897, append., p. 1-261. Chittenden, M. E., Jr. 1969. Life history and ecology of the American shad, Alosa sapidissima, in the Delaware River. Ph.D. Thesis, Rut- gers Univ., New Brunswick, N.J., 458 p. 1975. Dynamics of American shad, Alosa sapidissima, runs in the Delaware River. Fish. Bull., U.S. 73:487-494. Curtis, B. 1949. The life story of the fish. Scientific Book Club, Lond., 276 p. DiXON, W. J. (editor). 1967. Biomedical computer programs. 2nd ed. Univ. Calif. Press, Berkeley, 600 p. ELLIS, M. M., B. A. WESTFALL, D. K. MEYER, AND W. S. PLATNER. 1947. Water quality studies of the Delaware River with reference to shad migration. U.S. Fish Wildl. Serv., Spec. Sci. Rep. 38, 19 p. FOWLER, H. W 1908. The king crab fisheries in Delaware Bay, and further note on New Jersey fishes, amphibians and rep- tiles. Annu. Rep. N.J. State Mus. 1907 Part 3, p. 349-432. FREDIN, R. A. 1954. Causes of fluctuations in abundance of Connecticut River shad. U.S. Fish Wildl. Serv., Fish. Bull. 54:247- 259. FRY, F. E. J. 1971. The effect of environmental factors on the physiology of fish. In W. S. Hoar and D. J. Randall (editors), Fish Physiology, Vol. 6, p. 1-98. Academic Press, N.Y. GODWIN, W. F. 1968. The shad fishery of the Altamaha River, Georgia. Ga. Game Fish Comm., Mar. Fish. Div., Con- trib. Ser., 8, 39 p. HILDEBRAND, S. F. 1963. Family Clupeidae. In Fishes of the western North Atlantic, Part 3, p. 257-454. Mem. Sears Found. Mar. Res. Yale Univ. 1. HILDEBRAND, S. F., AND W. C. SCHROEDER. 1928. Fishes of Chesapeake Bay. U.S. Bur Fish., Bull. 43, Part 1, 366 p. HOLLAND, B. F., JR., AND G. F. YELVERTON. 1973. Distribution and biological studies of anadromous fishes offshore North Carolina. N.C. Dep. Nat. Econ. Re- sour., Div. Commer. Sport Fish., Spec. Sci. Rep. 24, 132 p. LaPOINTE, D. F. 1958. Age and growth of the American shad, from three Atlantic Coast Rivers. Trans. Am. Fish See. 87:139-150. LEACH. G. C. 1925. Artificial propagation of shad. Rep. U.S. Comm. Fish. 1924, append. 8, p. 459-486. Doc. 981. LEGGETT, W. C. 1969. A study of the reproductive potential of the Ameri- can shad (Alosa sapidissima) in the Connecticut River, and of the possible effects of natural or man induced changes in the population structure of the species on its reproductive success. Essex Mar. Lab., Essex, Conn., 72 p. 1972. Weight loss in American shad (Alosa sapidissima, Wilson) during the freshwater migration. Trans. Am. Fish. Soc. 101:549-552. leim, a. H. 1924. The life history of the shad (Alosa sapidissima (Wil- son) ) with special reference to the factors limiting its abundance. Contrib. Can. Biol., New Ser. 2:163-284. LEIM, A. H., and W. B. Scott. 1966. Fishes of the Atlantic Coast of Canada. Fish. Res. Board Can., Bull. 155, 485 p. MILNER, J. W 1874. Report on the propagation of the shad (Alosa sapidissima) and its introduction into new waters by the U.S. Fish Commissioner in 1873. U.S. Comm. Fish Fish., Part 2, Rep. Comm. 1872-73, p. 419-451. MOSS, D. D. 1946. Preliminary studies of the shad (Alosa sapidissima) catch in the lower Connecticut River, 1944. Trans. North Am. Wildl. Conf. 11:230-239. NICHOLS, P. R. 1959. Extreme loss in body weight of an American shad (Alosa sapidissima). Copeia 1959:343-344. NICHOLS, P. R., AND W. H. MASSMANN. 1963. Abundance, age, and fecundity of shad, York River, Va., 1953-59. U.S. Fish Wildl. Serv., Fish. Bull. 63:179- 187. NICHOLS, P. R., AND M. E. TAGATZ. 1960. Creel census Connecticut River shad sport fishery, 1957-58, and estimate of catch, 1941-56. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 351, 12 p. 156 CHITTENDEN: ADULT AMERICAN SHAD IN FRESH WATER PALOHEIMO, J. E., AND L. M. DICKIE. 1966. Food and growth of fishes. II. Effects of food and temperature on the relation between metabolism and body weight. J. Fish. Res. Board Can. 23:869-908. PROSSER. C. L. 1973. Oxygen: respiration and metabolism. In C. L. Pros- ser (editor). Comparative animal physiology, 3rd ed., p. 165-211. Saunders, Phila. Steel, R. G. D., and J. H. Torrie. I960. Principles and procedures of statistics with special reference to the biological sciences. McGraw Hill, N.Y., 481 p. SYKES, J. E. 1965. Shad fishery of the Ogeechee River, Georgia, in 1954. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 191, lip. Sykes, J. E., AND B. A. Lehman. 1957. Past and present Delaware River shad fishery and considerations for its future. U.S. Fish Wildl. Serv., Res. Rep. 46, 25 p. TALBOT. G. B. 1954. Factors associated wdth fluctuations in abundance of Hudson River Shad. U.S. Fish Wildl. Serv., Fish. Bull. 56:373-413. WALBURG, C. H. 1956. Commercial and sport shad fisheries of the Edisto River South Carolina, 1955. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 187, 9 p. 1957. Neuse River shad investigations, 1953. U.S. Fish Wildl. Serv, Spec. Sci. Rep. Fish. 206, 13 p. 1960. Abundance and life history of shad, St. Johns River, Florida. U.S. Fish Wildl. Serv., Fish. Bull. 60:487-501. 1961. Natural mortality of American shad. Trans. Am. Fish. Soc. 90:228-230. WALBURG, C. H., and J. E. SYKES. 1957. Shad fishery of Chesapeake Bay with special em- phasis on the fishery of Virginia. U.S. Fish Wildl. Serv., Res. Rep. 48, 26 p. 157 DISTRIBUTION, ABUNDANCE, AND SIZE OF PENAEID SHRIMPS IN THE ST. ANDREW BAY SYSTEM, FLORIDA Harold A. Brusher and Larry H. Ogren^ ABSTRACT Shrimp collections were made every 2 weeks at 12 stations in varying depths (1.5-12.2 m) of the St. Andrew Bay system, Fla., from September 1972 through August 1973. The eight species of penaeid shrimps caught in 312 trawl hauls were, in decreasing order of abundance: pink shrimp, Penaeus duorarum; broken-neck shrimp, Trachypenaeus similis; rock shrimp, Sicyonia brevirostris; rock shrimp, S. dorsalis; broken-neck shrimp, T. constrictus; brown shrimp, P. aztecus; white shrimp, P. setiferus; and rock shrimp, S. typica. Of the total catch of penaeids, 57.7% were of the genus Penaeas, 22.6% oi Sicyonia , and 19.7% oi Trachypenaeus . Penaeids were more abundant in the sections of the bay system close to the Gulf of Mexico. Seasonal abundance varied for each species. Shrimps of the genus Penaeus were larger in deeper sections of the bay. The hydrological characteristics of the St. Andrew Bay system are much more similair to the waters of the Gulf of Mexico than are those of other estuaries of the northern gulf This similarity probably accounts for the relatively high abundance of shrimps of the genera Trachypenaeus and Sicyonia in the bay system. Also, this similarity probably delays the gulfward migration of shrimps of the genus Penaeus and accounts for their large sizes in the system. Personal observations made on exploratory col- lecting trips and on cruises aboard shrimp trawl- ers within the St. Andrew Bay system in north- west Florida had led us to believe that some species of marine organisms normally found in offshore waters of the Gulf of Mexico occurred commonly within the system. For example, penaeid shrimps of the genera Trachypenaeus and Sicyonia, which are rare in bay systems of the northern gulf, were observed frequently. Also, shrimps of the genus Penaeus appeared to be much larger within the St. Andrew Bay system than other estuarine areas. It thus appeared to us that the penaeid shrimps of the St. Andrew Bay system were unusual in terms of species composi- tion and size. Although utilization of estuarine waters by populations of shrimps of the genus Penaeus is well known (Lindner and Cook 1970; Cook and Lindner 1970; Costello and Allen 1970), the abundance, distribution, and size are not com- pletely described for all penaeid species within many estuarine waters. This information is espe- cially lacking along the northwest Florida coast. The objectives of our study were to estimate these parameters for penaeid shrimps in the St. An- drew Bay system. 'Gulf Coastal Fisheries Center, Panama City Laboratory, Na- tional Marine Fisheries Service NOAA, Panama City, FL 32401. STUDY AREA The St. Andrew Bay estuarine system is lo- cated on the northwest coast of Florida between lat. 30°00' and 30°20'N and long. 85°23' and 85°53'W. The system consists of four bays — North, West, East, and St. Andrew (Figure 1) — with mean depths of 1.8, 2.1, 2.1, and 5.2 m, re- spectively, and covers an area of 280 km^ (McNulty et al. 1972). Various aspects of the physical and biological characteristics of the St. Andrew Bay system have been presented by Ichiye and Jones (1961), Waller (1961), Vick (1964), Hopkins (1966), Salsman et al. (1966), Cosper (1972), and McNulty et al. (1972). GUIF OF MEXICO Figure l. — Location of sampling stations in the St. Andrew Bay System, Fla. Manuscript accepted July 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 158 /S^^-/4(a. 1976. BRUSHER and OGREN: PENAEID SHRIMPS IN ST. ANDREW BAY SYSTEM Waters in the St. Andrew Bay system are rela- tively high in transparency. This high transpar- ency results in part from the porosity of the soils of the watershed, the low freshwater inflow, and the proximity of the system to the clear waters of the northeastern Gulf of Mexico. In terms of ex- tinction coefficients, the transparency of gulf waters adjacent to St. Andrew Bay are typical of clear oceanic waters (Tolbert and Austin 1959). The bottom of the bay system is composed of distinct sediment regimes. The sand regime (>80% sand) is generally restricted to areas near the passes and in depths less than 6 m. The silt- clay regime (>50% clay, <50% silt, and <20% sand) is located in the deeper waters of the sys- tem, but not in the passes (Waller 1961). The bay system also contains areas covered by rooted submerged vegetation. The submerged vegetation includes turtle grass, Thalassia tes- tudinum; manatee grass, Syringodium filiforme; and shoal grass, Diplanthera wrightii. These grasses cover an area of about 3,200 hectares. METHODS Sampling was conducted every 2 wk from 6 September 1972 through 21 August 1973 at 12 stations (Figure 1, Table 1). Two consecutive nights were required to sample at all stations with samples taken between sunset and 0200 h. On 23-24 August 1973 additional sampling was conducted between 1000 and 1400 h at the 12 sta- tions to compare day catches with the night catches of 20-21 August 1973. Biological samples were obtained at each sta- tion with an 11.5-m wing trawl with stretched meshes of 7.6 cm in the wings, 3.8 cm in the body, and 2.5 cm in the cod end. The trawl was towed at about 3.5 knots for 10 min. The entire catch at each station was placed on ice and transported to Table l. — Locations and depth ranges of sampling stations in the St. Andrew Bay system, Fla. Identifying Depth range station Lat.' Long.' landmark (m) 1 30050N 85°31.0'W Goose Point 4.6- 6.1 2 30 06,3N 8535.0W Shoal Point 7.6- 9.1 3 30=07.6N 85='37.7'W Palmetto Point 7.6- 9.1 4 30 09.0'N 85=40. 8W Redflsh Point 10.7-12.2 5 30'09,5'N 85'41.6W Baker Bayou 6.1- 7.6 6 30'06.2N 85°41.3'W Stiell Island 6.1- 7.6 7 3009.4'N 85°42.8'W Courtney Point 76- 9.1 8 30M0.4'N 85=43. 0'W Lake Huntington 6.1- 7.6 9 30'10,5'N 85°44.2'W Dyers Point 10.7-12.2 10 30=1 4. VN 85'44.3'W Shell Point 6.1- 7.6 11 30°15.7'N 85=46. 6'W Breakfast Point 3.1- 4.6 12 30^'15.4'N 85=40.0'W Haven Point 1.5- 3.1 the laboratory and frozen. Catches were thawed and processed usually within 1 wk of collection. Penaeid shrimps from each sample were enumer- ated by species, and 30 individuals, or all if less than 30, were measured to the nearest 0.5 cm total length (tip of rostrum to tip of telson). Environmental data were also obtained at each station. A water sample for determining dis- solved oxygen and turbidity was taken 0.5 m above the bottom at each station with a 3-liter water sampler Salinity and temperature were determined in situ with a Beckman^ RS5-3 por- table salinometer (accuracy ±0.5°C and ±0.31,) at the above mentioned depth. Turbidity was determined with a Hach turbidimeter (Formazin turbidity units — accuracy ±0.02 FTU), and dis- solved oxygen determined by the modified Wink- ler method (accuracy ±0.05 ml/liter). For each species, differences in catch per unit effort (average catch per tow), and in size (aver- age length by date) between subareas were tested statistically with Tukey's a;-procedure (Steel and Torrie 1960). For length comparisons, data were used for only those dates when shrimps of a species were caught in all subareas. For compari- sons of distribution and abundance, the data were grouped into the following subareas: East Bay (stations 1, 2); North Bay (station 12); West Bay (stations 10, 11); St. Andrew Bay (stations 3-5, 7-9); and East Pass (station 6). Mean catches per tow and mean total lengths were also compared between upper and lower bay areas. The upper area included all stations in East Bay, North Bay, and West Bay, and the lower area included all stations in St. Andrew Bay and East Pass. ENVIRONMENTAL FACTORS Mean values of environmental factors near the bottom were determined for subareas. Salinities and dissolved oxygen were higher in St. Andrew Bay and East Pass than in the other subareas (Table 2). Turbidities in North Bay, East Bay, and West Bay were greater than in St. Andrew Bay and East Pass. Bottom temperatures, however, were similar among subareas. When subarea data were combined into the re- spective upper and lower areas, the average val- ues were: salinity— 29.2, 33.2%; turbidity— 3.0, 'United States Department of Commerce, Nautical Chart 868-SC. ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 159 FISHERY BULLETIN: VOL. 74, NO. 1 Table 2. — Annual means and ranges of environmental factors measured in 1972-73 in five subareas of the St. Andrew Bay system, Fla. Environmental factor North Bay West Bay East Bay East Pass St. Andrew Bay Salinity (%.) Mean 27.20 29.08 30.34 32.97 33.27 Range 13.1-32.5 20.5-34.1 25.3-33.9 30.3-35.2 30.6-35.6 Turbidity (FTU)' Mean 2.69 3.40 2.63 1.09 1.75 Range 0.50-13.00 1 53-7.55 1.50-5.20 0.60-2.15 0.87-4.09 Temp (°C) Mean 21.74 21.82 21.79 22 13 21.74 Range 13.1-31.1 13.6-30.2 13.8-29.9 13.0-30.2 13.2-30.0 Dissolved 02 (ml/liter) Mean 3.87 3.77 3.27 4.43 4.01 Range 1.33-5.37 2.06-4.70 1.64-5.58 3.47-5.13 3.13-4.80 No. of samples 26 52 52 26 182 'Formazin turbidity units. 1.7 FTU; temperature— 21.8°, 21.8°C; dissolved oxygen — 3.6, 4.1 ml/liter. Generally, salinity and dissolved oxygen values were higher in the lower area, turbidity values were higher in the upper area, and temperatures were similar between areas (Figure 2). The only noteworthy variation Figure 2. — Mean values of salinity, turbid- ity, temperature, and dissolved oxygen in the upper and lower areas of the St. Andrew Bay system, Fla., 1972-73. "4 i« 3 17 30 U 27 13 26 8 22 5 20 5 19 Stp OCT. NOV. DEC. JAN. FEB MAR, 1972 3 11 30 14 29 13 2S 9 23 APR MAT JUN. nil. 1973 7 20 AUG. 160 BRUSHER and OGREN: PENAEID SHRIMPS IN ST. ANDREW BAY SYSTEM in these values occurred in the salinity of the upper area where heavy spring rains accounted for an exceptional drop in salinity in early April. Compared to hydrological data from other northern gulf estuaries (Gunter 1950; Swingle 1971; Dunham 1972; Stokes 1974), the values within the St. Andrew Bay system are more oceanic than estuarine (Waller 1961; Hopkins 1966). CATCHES Eight species of penaeids were taken during the study: three species ofPenaeus (P. duorarum, P. aztecus, and P. setiferus), two species of Trachypenaeus (T. similis and T. constrictus) , and three species of Sicyonia iS. breuirostris, S. dor- salis, and S. typica). Catches of each species at each of the 12 stations are shown in Table 3. The greatest number of individual shrimps (species combined) was taken at station 4 (St. Andrew Bay), the least at station 2 (East Bay). Penaeus duorarum was the most abundant species, S. typica the least. Since only 25 S. typica (ranging in size from 3.5 to 5.5 cm) were caught, this species will not be discussed in the following sections. Although methods were similar, a striking dif- ference was apparent between our catches and those from other estuarine systems in the north- ern Gulf of Mexico. In our study, 57.6% of the total penaeid catch consisted of members of the genus Penaeus, 22.6% of the genus Sicyonia, and 19.7% of Trachypenaeus. In contrast, studies in other estuarine systems in Alabama (Swingle 1971), Louisiana (Dunham 1972), and Texas (Gunter 1950; Moffett 1968; Stokes 1974) showed that the genus Penaeus represented 99 to 100% of the total trawl catch of penaeids. DISTRIBUTION AND ABUNDANCE To determine where shrimp were more abun- dant in the St. Andrew Bay system, relative abundances were compared by subarea (Table 4). Significant differences were found for four of the seven species: T. similis, S. breuirostris, S. dor- salis, and T. constrictus. Either St. Andrew Bay or East Pass or both had significantly greater abun- dance of these species than the other subareas. When subarea data were combined for each species and apportioned into upper and lower areas, the relative abundances were greater in the upper area for P. aztecus and P. setiferus and were greater in the lower area for the other penaeids. Average catches per tow for the upper and lower areas, respectively, were: P. duorarum, 110.8, 129.3; T. similis, 12.8, 49.4; S. breuirostris, 6.0, 51.3; S. dorsalis, 2.9, 32.9; T. constrictus, 3.1, 14.8; P. aztecus, 10.1, 4.6; P. setiferus, 2.7, 0.3. To determine seasonal distribution and abun- dance, the catches per tow were calculated by area and by date for each species. The results, shown in Figure 3, indicate summer and fall abundances for the three species of Penaeus, al- though not necessarily in both areas. For Trachy- penaeus and Sicyonia, seasonal abundances were evident only in the lower area, with T similis and S. dorsalis more abundant during spring and summer, S. breuirostris more abundant during winter and early spring, and T constrictus during spring. Table 3. — Total numbers of penaeid shrimps caught in 312 trawl hauls within the St. Andrew Bay system, Fla., from September 1972 through August 1973. Station Species 1 2 3 4 5 6 7 8 9 10 11 12 Total Pink shrimp, Penaeus duorarum 3,485 1,613 2,724 1,879 5,097 3,115 3,348 4,767 2,382 3,062 3,371 2,737 37,580 Broken-neck shrimp, Trachypenaeus similis 79 1,140 1,553 2,724 101 418 1,095 1,218 1,878 383 7 3 10,599 Rock shrimp. Sicyonia brevirostris 12 19 147 984 1,758 3,812 1,552 717 198 17 9 9 9,234 Rock shrimp, Sicyonia dorsalis 3 273 632 3,433 66 247 434 226 993 80 6,387 Broken-neck shrimp, Trachypenaeus constnctus 56 53 150 207 704 907 275 248 208 41 93 122 3,064 Brown shrimp, Penaeus aztecus 125 81 144 119 19 146 187 165 85 197 342 279 1,889 White shrimp. Penaeus setiferus 42 22 18 5 14 13 21 52 166 71 424 Rock shnmp. Sicyonia typica 4 4 12 2 3 25 Total 3,802 3,201 5,368 9,355 7,749 8,657 6,905 7,356 5,768 3,832 3,988 3,221 69,202 Rank 10 12 7 1 3 2 5 4 6 9 8 11 161 FISHERY BULLETIN: VOL. 74, NO. 1 Table 4. — Comparisons of mean catch per tow of penaeid shrimps between subareas (Tukey's w-procedure with 125 df) of the St. Andrew Bay system, Fla., from September 1972 through August 1973. Species Subarea, mean catch in parentheses and significance lines' Penaeus duorarum Trachypenaeus similis Sicyonia brevirostris East North East West St. Andrew Bay Bay Pass Bay Bay (100.4) (105.3) (119.8) (124.4) (130.9) North West East East St. Andrew Bay Bay Pass Bay Bay (0.1) (7.7) (15.9) (24.3) (55.0) North East West St. Andrew East Bay Bay Bay Bay Pass (0.4) (0.6) (0.7) (35.4) (146.6) North West East East St. Andrew Bay Bay Bay Pass Bay S. dorsalis (0.0) (1.8) (5.5) (9.5) (36.8) East West North St. Andrew East Bay Bay Bay Bay Pass T. constrictus (2.2) (3.1) (4.7) (11.4) (34.9) East St. Andrew East West North Bay Bay Pass Bay Bay P. aztecus (4.0) (4.4) (5.6) (10.5) (10.7) East St. Andrew East North West Pass Bay Bay Bay Bay P. setiferus (0.0) (0.4) (1.4) (2.6) (4.1) 'Any two means not underscored by the same line are significantly different at the 5% level. 2 o < 9 > a z Z Z < Figure 3. — Mean catch per tow of seven penaeid shrimp species in the upper and lower areas of the St. Andrew Bay system, Fla., 1972-73. 6 IS 3 17 30 14 27 13 26 8 23 5 20 5 19 SEP OCT NOV DEC JAN FEB MAR 1972 3 18 30 14 29 13 25 9 23 7 20 APR MAY JUN JUL AUG 1973 162 BRUSHER and OGREN: PENAEID SHRIMPS IN ST. ANDREW BAY SYSTEM Penaeid shrimps taken from the St. Andrew Bay system showed definite habitat preference by genera when abundance was related to depth. As shown in Table 5, the higher mean catches per tow for Penaeus occurred in the shallower waters, while those for Trachypenaeus and Sicyonia oc- curred in the intermediate and deeper waters of the sampled area. Ninety -two percent of all Tra- chypenaeus and Sicyonia were taken from the lower area where the average station depth was 8.6 m. Day and night comparisons showed mean catch per tow to be greater at night for all seven species (Table 6). Table 5. — Comparisons of mean catch per tow and mean length (cm) of penaeid shrimps in relation to depth and species within the St. Andrew Bay system, Fla., from September 1972 through August 1973. 1.5-4.6 m 4.6-7.6 m 7.6-12.2 m Species Stn. 11, 12 1, 5, 6, 8, 10 2. 3, 4, 7, 9 Penaeus duorarum 117.5 150,2 91,9 (9.1) (9.5) (10.0) Trachypenaeus similis 0.2 16.9 64.6 (6.3) (6.3) (6.8) Sicyonia brevirostris 0.3 48.6 22.3 (6.1) (6.0) (6.2) S. dorsalis 0.0 4.8 38.8 (-) (5.3) (5.5) T. constrictus 4.2 15.1 6.9 (4.7) (4.8) (4.9) P. aztecus 12.0 5.0 4.7 (11.1) (12.4) (12.7) P. setiferus 4.6 0.8 0.6 (11.5) (12.9) (14.1) Table 6. — Comparisons of mean catch per tow and mean total length (cm) between day and night catches of penaeid shrimps taken from the St. Andrews Bay system, Fla., in August 1973. Species Day Night Penaeus duorarum Tracliypenaeus similis Sicyonia brevirostns S. dorsalis T. constrictus P. aztecus P. setiferus No of tows 34.8 172.4 (8.2) (8,4) 0.3 8.0 (6.4) (6.2) 1.7 (-) (7.6) 1.2 5.2 (5.6) (5.4) 0.3 (-) (4.1) 3.0 9.5 (13.1) (13.2) 0.8 1.5 (11.1) (10.2) 12 12 SIZE Shrimps of the genus Penaeus were larger than shrimps of the other two genera. Penaeus seti- ferus had the largest mean length, while S. dor- salis had the smallest. Mean total lengths in cen- timeters and length ranges in centimeters for each species in the St. Andrew Bay system were: P. duorarum, 9.5, 4.0-18.5; T. similis,6.6, 3.0-10.0; S. brevirostris, 5.7, 2.8-9.5; S. dorsalis, 5.5, 2.0- 7.8; T. constrictus, 4.5, 2.5-8.0; P. aztecus, 12.4, 4.5-18.5; and P. setiferus, 13.3, 7.0-16.0. Differences in lengths of shrimps associated with water depth were examined (Table 5); nota- ble differences were discernible only for the genus Penaeus, the larger specimens of which generally were found in deeper waters. This relation has also been reported by others (Lindner and Cook 1970; Cook and Lindner 1970; Costello and Allen 1970). Species of Trachypenaeus and Sicyonia showed little difference in mean lengths with water depths, although the largest mean sizes were found in the deeper zone. Examination for differences in lengths associ- ated with sampling at night and during the day revealed clearly that hour of sampling had no ef- fect on size of captured shrimps (Table 6). Comparisons of mean total lengths for the seven species between those subareas from which sufficient data were available showed that the largest shrimps were in either St. Andrew Bay or East Pass (Table 7). However, statistically sig- nificant differences were found for only three species: P duorarum, T. similis, and P setiferus. For five of the seven species, larger specimens were caught in the lower area more often than in the upper area. The situation was reversed for S. brevirostris, whereas, for T constrictus the mean sizes for the two areas were the same. Mean lengths in centimeters by species between upper and lower bay areas, respectively, were: P. duorarum, 9.1, 9.9; T. similis, 6.4, 6.7; S. bre- virostris, 6.3, 5.7; S. dorsalis, 5.4, 5.6; T. constric- tus, 4.5, 4.5; P. aztecus, 11.9, 12.8; and P. setiferus, 11.7, 14.7. Shrimps of the genus Penaeus were almost con- sistently larger in the lower area throughout the year (Figure 4). As shrimps of this genus grow larger, they tend to move into deeper, more saline, and less turbid waters. When present in both areas at the same time, the two species of Trachypenaeus were larger in the lower area more often than in the upper, whereas the reverse was true of the two species of Sicyonia. DISCUSSION AND CONCLUSIONS In general, water depths and salinities are greater, and turbidities, temperature fluctua- 163 FISHERY BULLETIN: VOL. 74, NO. 1 Table 7. — Comparisons of mean total length (cm) of penaeid shrimps between subareas (Tukey's u;-procedure) of the St. Andrew Bay system, Fla., from September 1972 through August 1973. Species Subareas, mean total length in parentheses, and significance lines' df Penaeus duorarum North Bay (8.89) West Bay (9.12) East Bay (9.19) East Pass (9.77) St. Andrew Bay (682) East Pass (6.30) North Bay (4.77) St. Andrew Bay (9.81) East Pass (4.90) 120 Trachypenaeus similis East Pass (5.86) West Bay (6.20) East Bay (6.67) 72 Sicyonia brevirostns S. dorsalis T. constrictus St. Andrew Bay (5.66) East Bay (5.37) East Bay (4.23) East Pass (5.81) West Bay (5.44) St. Andrew Bay (4.43) St. Andrew Bay (5.50) West Bay (4.67) 24 36 10 P. aztecus North Bay (11.41) West Bay (11.53) East Bay (12.50) St. Andrew Bay (12.79) East Pass (12.96) 30 P. setiferus East Bay (11.03) West Bay (11.68) North Bay (12.90) St Andrew Bay (14.68) 12 'Any two means not underscored by the same line are significantly different at the 5% level. tions, and river discharges are lower in the St. Andrew Bay system than in other northern gulf estuaries (Apalachicola Bay to the Rio Grande River). The dominant group of spermatophytes in the lower area are the submerged sea grasses, whereas in most other northern gulf estuaries the dominant groups are the emergent grasses in the intertidal zone (Kutkuhn 1966). This unusual es- tuarine environment in the St. Andrew Bay sys- tem may induce shrimps of the genus Penaeus to remain within the system for longer periods of time, especially in the lower areas where oceanic conditions often prevail. Such environmental differences probably ac- count for the differences observed in composition, abundance, and size of penaeid shrimps between the St. Andrew Bay system and other estuarine systems in the northern Gulf of Mexico. For example: 1) large adult (total length ranges of 16.5 to 18.5 cm) P. duorarum and P. aztecus usu- ally occur only in offshore waters, but we caught many of these large specimens throughout the St. Andrew Bay system; 2) in low salinity waters characteristic of other bay systems subadult P. setiferus and P. aztecus are more abundant than P. duorarum, whereas in the St. Andrew Bay system we found subadult P. duorarum more abundant than P. setiferus and P. aztecus; and 3) previous reports indicated that T. similis, S. brevirostris, and .S. dorsalis do not ordinarily enter estuaries (Eldred 1959; Joyce 1965; Kut- kuhn 1966; Cobb et al. 1973), but we caught many individuals of these species within the St. Andrew Bay system. The abundance of shrimps of Trachypenaeus and Sicyonia in the St. Andrew Bay system con- trasts sharply with those reported from other estuarine areas of the Gulf of Mexico. Other in- vestigators have included catches made adjacent to barrier islands or tidal passes and reported abundances of less than 1 shrimp per tow. (Dun- ham 1972; Gunter 1950; Saloman 1964, 1965; Swingle 1971). In our study, average catch per tow (excluding Station 6, which is adjacent to a bar- rier island) for each species was: T. similis, 36; T. constrictus, 8; S. brevirostris, 19; S. dorsalis, 21. Periods of greatest abundance of S. brevirostris in offshore waters of the northwestern and south- eastern gulf occur in summer and early fall (Brusher et al. 1972; Cobb et al. 1973). In the St. Andrew Bay system, this species was almost ab- sent during this period. We believe that this shrimp migrates from inshore to offshore gulf waters during spring months. Means and ranges of total lengths of species of Trachypenaeus or Sicyonia taken in other es- tuarine areas were usually less (Swingle 1971; Dunham 1972) than those taken in offshore areas 164 BRUSHER and OGREN: PENAEID SHRIMPS IN ST. ANDREW BAY SYSTEM o > a z < ' O 4 ^• z < \6 UJ S u n 10 e 16 14 13 10 8 Penaeus duorarum Trachypenaeus iimilis Sicyonia brevirosfris Sicyonio dorsalis Trachypenaeus constrictui LOWER AREA UPPER AREA Penaeus aziecus Penaeus setilerus '''''' ''''''' Figure 4. — Mean total lengths of seven penaeid shrimp species in the upper and lower areas of the St. Andrew Bay system, Fla., 1972-73. 6 IS 3 17 30 14 27 13 26 8 22 5 20 5 19 3 18 30 14 29 13 25 9 23 7 20 SEP OCI NOV 1972 DEC JAN FEB MAR APR 1973 MAY JUN JUL. AUG of the Gulf of Mexico (Brusher et al. 1972). The mean total lengths of the penaeids with the ex- ception of T. constrictus (Table 7) were similar to those reported by Brusher et al. (1972) for speci- mens caught in the Gulf of Mexico. We believe that species of Trachypenaeus and Sicyonia utilize St. Andrew Bay as a nursery area owing to the similarity of the bay to offshore oceanic habitats. Of the three species oi Penaeus caught in this study, P. duorarum was the most abundant. High abundance of P. duorarum was expected, because the highest concentration of this species in the Gulf of Mexico occurs in the eastern areas (Cos- tello and Allen 1970). Costello and Allen as- sociated P. duorarum with grass beds; grass beds are abundant in St. Andrew Bay. Low abundance of P. aztecus and P. setiferus was expected also, as these are found most abundantly in the north- western (Texas coast) and north central (Louisi- ana coast) portions of the Gulf of Mexico, respec- tively (Cook and Lindner 1970; Lindner and Cook 1970). Although similar gear and trawling methods were used, mean total lengths and length ranges of P. aztecus and P. duorarum caught in the St. Andrew Bay system differed greatly from those caught in other gulf estuaries (Saloman 1965; Trent et al. 1969; Dimham 1972). Our catches in- cluded many specimens over 13.0 cm total length which, according to Joyce (1965), is well above the size at which shrimps of the genus Penaeus are believed to leave estuarine areas. Shrimps of this genus greater than 10 cm total length are usually found in offshore waters (Lindner and Cook 1970; Cook and Lindner 1970; Costello and Allen 1970). We conclude that the St. Andrew Bay system is unusual among estuaries of the northern Gulf of 165 FISHERY BULLETIN: VOL. 74, NO. 1 Mexico; its environmental qualities which are much more similar to those in the gulf account for the common occurrence in the bay of penaeid shrimps of the genera Trachypenaeus and Si- cyonia normally found in the offshore waters of the open gulf; the unusual environmental factors within the system also delay the migration of penaeid shrimps of the genus Penaeus into the open gulf, thereby allowing them to grow larger within the St. Andrew Bay system. ACKNOWLEDGMENTS We thank Maxwell Miller and Leslie Touart for their help in collecting and processing the sam- ples and David Muenzel, Captain of the RV Rachel Carson, for his assistance in keeping us on schedule. We gratefully acknowledge the critical reviews of this manuscript by David Aldrich (Texas A&M University), Donald Allen (National Marine Fisheries Service, NOAA), and William Lyons (Florida Department of Natural Re- sources). LITERATURE CITED brusher, h. a., w. c. renfro, and r. a. neal. 1972. Notes on distribution, size, and ovarian development of some penaeid shrimps in the northwestern Gulf of Mexico, 1961-62. Contrib. Mar Sci. 16:75-87. Cobb, S. p., C. R. Futch, and D. K. Camp. 1973. The rock shrimp, Sicyonia brevirostris Stimpson, 1871 (Decapoda, Penaeidae). Mem. Hourglass Cruises 3(l):l-38. Cook, H. L., and M. J. Lindner. 1970. Synopsis of biological data on the browTi shrimp Penaeus aztecus aztecus Ives, 1891. FAO (Food Agric. Organ. U.N.) Fish. Rep. 57:1471-1497. COSPER, T. C. 1972. The identification of tintinnids (Protozoa: Ciliata: Tintinnida) of the St. Andrew Bay system, Florida. Bull. Mar Sci. 22:391-418. COSTELLO, T J., AND D. M. ALLEN. 1970. Synopsis of biological data on the pink shrimp Penaeus duorarum duorarum Burkenroad, 1939. FAO (Food Agric. Organ. U.N.) Fish. Rep. 57:1499-1537. DUNHAM, F. 1972. A study of commercially important estuarine- dependent industrial fishes. La. Wildl. Fish. Comm., Tech. Bull. 4, 63 p. ELDRED, B. 1959. A report on the shrimps (Penaeidae) collected from the Tortugas controlled area. Fla. State Board Conserv. Mar. Lab., Spec. Sci. Rep. 2, 6 p. GUNTER, G. 1950. Seasonal population changes and distributions as re- lated to salinity, of certain invertebrates of the Texas coast, including the commercial shrimp. Publ. Inst. Mar Sci., Univ. Tex. 1(2):7-51. Hopkins, T. L. 1966. The plankton of the St. Andrew Bay system, Flor- ida. Publ. Inst. Mar Sci., Univ. Tex. 11:12-64. ICHIYE, T, AND M. L. Jones. 1961. On the hydrography of the St. Andrew Bay system, Florida. Limnol. Oceanogr. 6:302-311. JOYCE, E. A., JR. 1965. The commercial shrimps of the northeast coast of Florida. Fla. State Board Conserv. Mar. Lab., Prof Pap. Ser. 6, 224 p. KUTKUHN, J. H. 1966. The role of estuaries in the development and per- petuation of commercial shrimp resources. Am. Fish. Soc, Spec. Publ. 3:16-36. LINDNER, M. J., AND H. L. COOK. 1970. Synopsis of biological data on the white shrimp Penaeus setiferus (Linnaeus) 1797. FAO (Food Agric. Organ. U.N.) Fish. Rep. 57:1439-1469. MCNULTY, J. K., W. N. LINDALL, JR.. AND J. E. SYKES. 1972. Cooperative Gulf of Mexico estuarine inventory and study, Florida: Phase 1, area description. U.S. Dep. Commer, NOAA Tech. Rep. NMFS CIRC-368, 126 p. MOFFETT, A. 1968. A study of Texas shrimp populations, 1968. Tex. Parks Wildl. Dep., Coastal Fish. Proj. Rep., p. 67-93. SALOMAN. C. H. 1964. The shrimp Trachypeneus similis in Tampa Bay. Q. J. Fla. Acad. Sci. 27:160-164. 1965. Bait shrimp {Penaeus duorarum) in Tampa Bay, Florida — biology, fishery economics, and changing habitat. U.S. Fish Wildl. Serv, Spec. Sci. Rep. Fish. 520, 16 p. SALSMAN, G. G., W. H. TOLBERT, AND R. G. VILLARS. 1966. Sand-ridge migration in St. Andrew Bay, Flori- da. Mar. Geol. 4:11-19. Steel. R. G. D., and J. H. Torrie. 1960. Principles and procedures of statistics, with special reference to the biological sciences. McGraw-Hill Book Co., N.Y., 481 p. Stokes, G. M. 1974. The distribution and abundance of penaeid shrimp in the lower Laguna Madre of Texas, with a description of the live bait shrimp fishery. Tex. Parks Wildl. Dep., Tech. Ser. 15, 32 p. SWINGLE, H. A. 1971. Biology of Alabama estuarine areas — cooperative Gulf of Mexico estuarine inventory. Ala. Mar Resour. Bull. 5, 123 p. TOLBERT, W. H., AND G. B. AUSTIN 1959. On the nearshore marine environment of the Gulf of Mexico at Panama City, Florida. U.S. Navy Mine Def Lab., Tech. Pap. TP-161, 104 p. TRENT, W. L., E. J. PULLEN, C. R. MOCK, AND D. MOORE. 1969. Ecology of western Gulf estuaries. In Report of the Bureau of Commercial Fisheries Biological Laboratory, Galveston, Texas, fiscal year 1968, p. 18-24. U.S. Fish Wildl. Serv., Circ. 325. VICK, N. G. 1964. The marine ichthyofauna of St. Andrew Bay, Florida, and nearshore habitats of the northeastern Gulf of Mexico. Tex. A&M Univ. Dep. Oceanogr. Meteorol., Proj. 286-D, Ref 64-19T, 77 p. WALLER, R. A. 1961. Ostracods of the St. Andrew Bay system. M.S. Thesis, Florida State Univ., Tallahassee, 46 p. 166 SOME FEATURES OF COHO SALMON, ONCORHYNCHUS KISUTCH, FRY EMERGING FROM SIMULATED REDDS AND CONCURRENT CHANGES IN PHOTOBEHAVIOR J. C. Mason^ ABSTRACT The emergence of sibling coho fry from simulated redds lasted 20-23 days during which 97-98% of the fry emerged. Average size of emerging fry increased with time but the largest fry emerged during the peak of emergence. No clear preference was shown for nocturnal or daylight emergence but the latter increased with time. Fry showed a positive current response, 69-82% moving upstream following emergence. Most fry emerged when yolk reserve was reduced to less than 10% of total dry weight. Later-emerging fry did not have lower yolk reserves, but fry moving downstream had slightly more yolk reserve than did fry moving upstream. Fry which were captured shortly after emergence had fed actively but had not yet filled their air bladders. Chironomids composed 70% of their diet. Photoresponse of sibling fry denied the redd experience was studied in light-dark choice boxes with reference to the timing of emergence of fry from the simulated redds. The pronounced photonegative behavior of the denied fry was suddenly lessened at time of emergence but remained photonegative. Weakening of the negative photoresponse was not the outcome of starvation or recent light experi- ence, and was not modified by repeated testing. Retention of the photonegative response is referred to hiding behavior and use of the gravel bed as a refuge. The anadromous female Pacific salmon, On- corhynchus, usually buries her eggs in several adjacent pockets in streambed or lakeshore ma- terials and these egg pockets collectively consti- tute a redd. The eggs hatch after several months and the larvae may spend several weeks or months using up their extensive yolk stores prior to emerging from the redd area into open water. Mortality during this extended period of sub- terranean life may be considerable (Royce 1959) and probably routinely exceeds 70% for most species of salmonids in natural habitats. Adap- tion to suboptimal conditions includes physiologic and behavioral responses in the embryo and larva which were reviewed, especially for sockeye salmon, O. nerka, by Bams (1969). Because destructive influences on the egg and alevin stages are amenable to amelioration through manipulation of substrate structure and flow regime, spawning channels pioneered by Wickett (1952) at Nile Creek have become a major component of salmon enhancement strategy. Despite these advances, we have yet to define optimal redd conditions, biotic and abiotic, which maximize preemergence survival of any 'Department of the Environment, Fisheries and Marine Ser- vice, Research and Development Directorate, Biological Sta- tion, Nanaimo, B.C. V9R 5K6, Canada. salmonid. Furthermore, fry surviving to emergence may face extended ecological conse- quences of suboptimal conditions in the redd which alter timing of, or size at, emergence (Ma- son and Chapman 1965; Mason 1969). Neither can we yet define for the emerging fry physiologic and behavioral states which optimize survival in open waters. Thus, premature emergence, imply- ing underdevelopment and reduced ability to re- spond adaptively is not referrable to a defined state of normality. Alevins of Oncorhynchus , as are those ofSalmo and Saluelinus (White 1915; Stuart 1953; Woodhead 1957), are initially negatively photo- tactic and respond to light by hiding (Hoar 1958). They become positively phototactic and rheotac- tic as emerged fry, orientation to current preced- ing the shift from negative to positive phototaxis (Dill 1969) as in Salmo (Grey 1929a; Stuart 1953) but the timing of this photobehavioral change in relation to emergence and remaining yolk re- serve remains unknown in Oncorhynchus and disagreement has arisen as to its timing in Salmo (Woodhead 1957). Histophysiological studies by Ali (1959) showed that only emerged fry and older stages of Oncorhynchus are capable of full retinomotor responses; however, partially devel- oped responses have obvious survival value. In this paper, some features of sibling coho fry Manuscript accepted September 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 1976. 167 FISHERY BULLETIN: VOL. 74, NO. 1 emerging from simulated stream redds are de- scribed. Light and current responses; length, weight, and condition; remaining yolk reserves at emergence; and changes in photoresponse were investigated. The possible effects on photore- sponse of repeated testing, previous exposure to light, and feeding experience were also examined. MATERIALS AND METHODS Emergence from Simulated Redds The emergence of coho salmon fry of known parentage (two males x one female) from four simulated redds was investigated in two pairs of wooden channels (Figure 1) located outdoors. Each channel was divided into three equal-sized compartments, and to simulate a redd, each center compartment was filled to a depth of 27 cm with stream pebbles 2-5 cm in diameter. A stand- pipe terminating at its lower end in a 10 cm x 10 cm platform on 10 cm stilts so as to enclose a chamber of 100 cm^ volume was buried in each redd at this time. In each redd the gravel surface was entirely underwater, but a shallow median depression served to concentrate the surface flow issuing through the V-notch openings. The frames of the inner partitions were covered with a double layer of fine plastic screen to allow for circulation through the redds. Water flow through each channel was 12 liters/min, about 30% of which passed through the redds. Ten days after hatching, 150 alevins from eggs incubated and hatched in standard baskets and previously unexposed to light were introduced into each redd at night via its standpipe and al- lowed to emerge spontaneously. Each standpipe was cleared of fry 1 h after stocking the redd by inserting a wire rod capped with rubber stoppers at either end and leaving the rod in place. Emerg- ed fry could enter either the upstream or down- stream compartments by way of the V-notch openings and were collected there daily at dawn and dusk. Emerging fry were anesthetized with MS-222,2 fork length was measured to the nearest 0.1 mm using a dissecting microscope, weight determined to the nearest 0. 1 mg on a Mettler Grammamatic balance after blotting, and the fry then preserved in 5% Formalin. For each redd, samples of 20 fry were extracted from each quartile of the emerg- ing population (total of 80 fry per redd) divided between fry moving upstream or downstream fol- lowing emergence. Yolk reserve at emergence was determined by dissecting out the yolk ma- terial, drying both yolk and fry to constant weight at 80°C, and expressing yolk reserve as a percentage of total dry weight. The resulting data were processed by regres- sion and analysis of variance techniques to ex- pose possible correlations between length, weight, condition {K) and yolk reserve with time, directional movement in current, and emergence during the daylight or darkness. Photoresponse Tests Ten days after hatching, sibling alevins from the same experimental stock as those used for the emergence study but denied the redd experience were separated into five groups of 50 fish each and held indoors in wire baskets except during testing. Two groups were held in complete dark- ness. One of these groups was tested frequently (dark experimental, DE); the other was tested once then not retested until 15 days later (dark control, DC). The three remaining groups were held in baskets partly exposed to daylight of about 200 ft-candles peak intensity from an ad- jacent window and were given three different treatments. One group was tested frequently (light experimental, LE); one was tested once then not retested until 15 days later (light con- trol, LC). The remaining group was not tested until the 18th day and, in contrast to the other groups, was fed frozen ground beef liver three times daily from day 9 onward (light control plus food). Figure l. — Compartmentalized wooden channels. Center compartments contained the simulated redds. Dotted areas signify screens. ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 168 MASON: FEATURES OF EMERGING COHO SALMON FRY Photoresponse tests were conducted in four choice boxes placed in an uncompartmentalized replicate of the emergence channels and located adjacent to them. The choice boxes were con- structed of fine plastic screen on a wire frame- work (Figure 2) and divided equally into two compartments by a vertical partition that al- lowed a passage height of 1.5 cm beneath it. Both hinged top and the partition were made of black polyethylene sheeting. The wooden channel was covered with the same material, except in the areas taken up by the boxes, so that the com- partments not covered by the hinged tops re- ceived most of the illumination in the boxes. Each box presented a choice between sharply contrast- ing light conditions rather than between "light" and "no light," because some light leaked under the partitions. A series of mirrors was mounted 1 m above the water surface, allowing observation from a blind. Water flow in the channel was 10 liters/min and velocity less than 10 cm/min. Water depth in the choice boxes was 10 cm providing an air space of 3 cm between the water surface and the ceiling of the covered compartment. Average fish density was set so as to allow about twice as much water volume and 2.4 times as much bottom area as in the holding baskets. Temperature of the water supply (stream) ranged from 7.8° to 11.7°C during Figure 2. — Light-dark choice box showing the reversible opaque lid. the experimental period. Light intensities at the exposed water surface ranged from 700 to 4,000 ft-candles during photoresponse tests. The procedure for a photoresponse test was as follows. The appropriate group of fry was trans- ferred to the test site in a covered pail, 40 fry were netted out and 10 fry put in each of the four choice boxes with the lids in an upright position. The lids were then closed in a common direction, and the remaining fry were returned to their holding basket. In the choice boxes, all fry swam into the dark compartments when the lids were dropped. After 30 min, the number of fry observed in the light compartments were recorded every 10 min for 40 min (5 observations in each of 4 compart- ments = 20 observations). A fish was considered to be in a light compartment when its head was visible. The lids were then reversed and, after 10 min, five additional observations were made at 10-min intervals. Thus, for each test, 40 counts were recorded on 40 fry, which spent 2 h in the choice boxes per test and about 10 min in the transfer process. The photoresponse tests were initiated 1 day before fry began emerging from the simulated redds and continued until the 22nd day of emergence. Length and weight measure- ments were taken for all fry groups on the follow- ing day. Data were tested for homogeneity using chi-square. There was no significant difference (P<0.01) between the first and second runs of five observations each made in individual choice boxes, x^ values ranging from 0.0 to 2.8 in 132 pairs of runs. Similarly, the data from individual choice boxes proved homogeneous within each test in 29 of the 33 tests performed (P<0.01) x^ values ranging from 1.1 to 7.8 with 3 degrees of freedom. The remaining four tests contained heterogeneous data, x^ values ranging from 14.7 to 36.5 and were excluded from further analysis. With homogeneity assured within most tests, the data within tests were pooled and processed. RESULTS Emergence from the Simulated Redds Fry began emerging 25 days after hatching and 15 days following introduction to the redds. Emergence proceeded for 20-23 days during which 97-98% of the fry emerged. All four redds showed a similar pattern of emergence, peaking at the same time, 74 to 94% of the fry emerging during the median 10 days (Figure 3). 169 FISHERY BULLETIN: VOL. 74, NO. 1 ® ® © ®J \. r 1 1^ Jir" nnl \ J A^ J^ ^ DAY OF EMERGENCE Figure 3. — Timing of coho salmon fry emergence from the simulated redds. The average size of fry increased significantly (P<0.01) as emergence proceeded but the largest fry emerged during the peak of emergence from day 10 to day 15 (Figure 4). More fry emerged at night than during the day in redds 2 and 3 (57% and 60%, respectively), but more fry emerged during the day in redd 1 (Table 1). No preference was shown by fry in redd 4 which emerged in equal numbers. Dividing the data into two time intervals, days 1 through 11, and days 12 through 24, revealed that emergence during the day increased some 30% in all four redds in the latter period. Emerging fry showed a strong positive current response, the majority (69-82%) moving upstream subsequent to emergence, upstream movement increasing but slightly as emergence proceeded. There were no significant differences in aver- age length and weight (P<0.01) of fry moving up- stream or downstream following emergence, but fry emerging during the day were, on the aver- age, larger than those emerging at night (Table 1), significantly so in two redds (redd 3, P<0.05; I I- o z UJ o ® 39 38 37 36 35 41 40 39 38 37 © ® * 35^ ® ••:? ^^■. 12 15 21 24 6 9 12 15 DAY OF EMERGENCE Figure 4. — Size of coho salmon fry at emergence including the regression lines. 170 MASON: FEATURES OF EMERGING COHO SALMON FRY Table l. — Average lengths of sibling coho fry emerging from four simulated redds, stratified as to night and day timing and direction of movement. Values in parentheses are percentages. Upstream Downstream Nigfit Day Redd movement movement emergence emergence 1 Number of fry 94(69.1) 42(30.9) 53(39.0) 83(61.0) Mean fork lengtfi (mm) ± SE 38.39 ± 0.11 38 12 ± 0.18 38.13 ± 18 0.15 ± 0.15 2 Number of fry 115(80.4) 28(19.6) 82(57,3) 61(42.7) Mean fork lengtfi (mm) ± SE 38,08 ± 0.11 37.91 ± 0.21 38.00 ± 0.12 38,11 ± 0,15 3 Number of fry 113(79.0) 30(21.0) 86(60,1) 57(39,9) Mean fork lengtfi (mm) ± SE 38.25 ± 0.11 38.48 ± 0.18 38.14 ± 0.11 38,54 ±0,15 4 Number of fry 116(82,3) 25(17.7) 70(49.6) 71(50.4) Mean fork length (mm) ± SE 38.10 ± 0,11 38.00 ± 0.38 37.68 ± 0.15 38.48 ± 13 Total emerging fry 438(77.8) 125(22.2) 291(51.7) 272(48,3) Pooled mean fork lengtfi (mm) 38.19 38.14 38.12 38.38 redd 4, P<0.01). This is the outcome of the ten- dencies for both increased emergence during the day and increased size at emergence as time progressed. Of the 584 fry that emerged from the simulated redds, 14 rather small fry emerged 5 or more days prior to the onset of general emergence. Twelve of these fry emerged at night and went downstream. They, and seven additional fry which also moved downstream and were designated as cripples due to truncated vertebral columns, were deleted from the analyses. Most fry emerged when their yolk reserve was reduced to less than 10% of total dry weight (Fig- ure 5), average reserve being 5-7% of total dry weight. The three rather high points (days 9-10) for fry moving downstream represent small sam- ples whose means were inflated by premature fry. Yolk reserve in these samples was either less than 8% or ranged between 26 and 60% for indi- vidual fry. The large standard errors shown in Figure 5 are all associated with mean values inflated by premature fry. Yolk reserve did not diminish with time, indicating that the majority of fry were in a similar nutritional state at emergence. Although there were no significant differences in length and weight between fry moving up or downstream, fry moving down- stream had more yolk reserve (9.2%) than did fry moving upstream (7.4%), this difference being significant at the 1% level. Similarly, in 13 of 16 possible pairs of samples from the four redds, the downstream fry contained more yolk reserve. To discover if fry were feeding within a short time of emergence, the digestive tracts of 75 fry emerging from redds 1 and 2 were examined. These fry were representative with regard to night or day emergence and upstream or down- stream movement following emergence, through- out the period of emergence. No dietary differ- ences were found in the 74 fry that had fed shortly before their capture. As both stomach and intestine contained food particles, fry emerging at night probably fed that night or during the preceding hours of daylight while in the redd. Chironomids constituted nearly 65% of their diet (Table 2), mites and Collembola made up 17%, and together these three items were consumed by 83% of the fry. At time of capture in the upstream and down- stream compartments, no fry had yet reached neutral buoyancy although some had partially filled air bladders. T3 O O 50 45 40 • 25 o 20 15 > ^ in o >- O redd 1 A redd 2 i i D redd 3 D redd 4 1 » J i f T b 1 c . \ I '1 i D I > 4 6 8 10 12 14 16 IB DAY OF EMERGENCE Figure 5. — Yolk reserve of coho salmon fry at emergence. Solid symbols indicate downstream movement following emergence; open symbols indicate upstream movement. Ver- tical bars indicate ±2 SE. For the remaining points, the range in SE was 0.1-1.0, and 90% ranged from 0.1 to 0.6. 171 FISHERY BULLETIN: VOL. 74, NO. 1 Table 2. — Diet of 75 coho salmon fry emerging from two of four simulated redds supplied with river water. Number of % of total % Food item items items incidence Chironomidae; Larvae 43 34.4 32.0 Pupae 20 16.0 21 4 Imagines 16 12.8 14.7 Total 79 63,2 68.1 Hydracarina 17 13.6 2.7 Collembola 13 10.4 12.0 Ephemeroptera nymphs 5 4.0 6.7 Arachnida 4 3.2 5.3 Trichoptera larvae 2 1.6 2.7 Plecoptera nymphs 1 0.8 1.3 Coleoptera imagines 1 0.8 1.3 Hymenoptera 1 08 1.3 Plant fragments 2 1.6 2.7 Concurrent Changes in Photoresponse Photoresponse testing of fry denied the redd experience began on day 1, 1 day before their counterparts in the simulated redds began emerging. Their photoresponse remained essen- tially negative throughout the time period when, normally, they would have emerged. Until the eighth day of emergence (day 9), less than 3% of the denied fry were seen in the light compart- ments (Figure 6) and they remained strongly photonegative although nearly 13% of their sibs had emerged from the redds. By day 12, the col- lective negative photoresponse had weakened considerably, and nearly 15% of the denied fry were recorded then in the light compartments. By the 16th day of emergence, when 90% of their sibs had emerged, the percent of the fry recorded in the light compartments reached a plateau. From day 17 onward, 20-30% of the fry were seen in the light compartments (15 of 19 tests), but the re- sponse was more variable during the last day of testing, two of the tests (LC and DC) providing heterogeneous data. Interaction stemming from territorial behavior was the most likely source of variability, the light compartments being sporad- ically defended by single fry attempting to drive the others away. Despite a decidedly negative photoresponse during the first 10 tests (Figure 6, Table 3), in 8 of these tests more fry held in darkness between tests were recorded in the light compartments than were those exposed to illumination between tests (P<0.01). Because there was no significant difference attributable to light history in sub- sequent tests, novelty due to limited light experi- ence may have stimulated exploratory behavior during testing in fry held in darkness between tests. When tested on days 15 and 17, the control DAY OF EMERGENCE Figure 6. — Change in photoresponse of coho salmon fry held in baskets between tests. The histogram depicts the concurrent rate of emergence of 584 sibling fry from the four simulated redds. TABLE 3. — Fry sightings in the light compartment of each of four choice boxes containing 10 fry during photoresponse tests, expressed as a percen- tage of possible sightings (400/test). Bracketed values are standard errors. Light Dark Light Dark Light control experimental experimental control control plus food Day (LE) (DE) (LC) (DC) (LC+F) 1 0.3(0.3) 2.5(0.6) 0.5(0.3) 1.8(0.5) 3 03(0.6) 2.0(0 8) 6 0.3(0.1) 0,5(0.3) 9 0.0 3.0(0.8) 13 13.3(1.6) 13.8(1.5) 15 32.2(2.4) 28.0(2.3) •17 26.3(1.8) 22.3(1.9) 36.3(2.9) 30.0(2.3) 18 26.8(2.8) '29.3(2.9) 26.5(2.3) 19 21.5(1,8) 23.3(1.9) 24.3(2.0) 24.8(1.4) 220 25.0(1.6) 23.8(1.9) 19.8(1.6) '16.3(1.9) 23 27.0(1.8) 27.5(2.3) '41.0(3.6) '10.8(1.3) 'Heterogeneous data. 2Fed In previous evening and 1 h prior to testing. 172 MASON: FEATURES OF EMERGING COHO SALMON FRY groups LC and DC showed higher counts (P<0.01) than did their experimental counter- parts LE and DE tested on day 17 (Table 3). Nonsignificant differences in subsequent tests suggested that frequency of testing may have de- pressed the magnitude of photoresponse change. Fry receiving supplemental food (LC-l-F) made scores similar to DC and LE groups (P<0.01) when tested on day 18, but lack of homogeneity in the data from one of the three tests performed precluded further evaluation. Light history and recent feeding did not sig- nificantly affect response level when the four pre- viously unfed groups were tested on day 20 {t = 1.3 with 158 df, P<0.020). Differences in average length among the four unfed groups of fry 1 day after the last tests were not significant (Table 4, F = 0.33 with 3, 96 df) but fish in the LE and DE groups weighed sig- nificantly more and therefore had higher K val- ues. Their heavier weight is attributed to feeding on natural drift foods available only in the choice boxes. The control group given supplemental food from day 9 onward were significantly longer than the other four groups of fry in average length {F = 11.4 with 4, 122 df, P<0.01) and weighed con- siderably more. Table 4. — Average lengths, weights, and condition factors {K) of samples of 25 coho fry used in the photoresponse experiment, measured 1 day after final testing. Fork length Live weight Treatment (mm) SE (mg) K' Light experimental 38.38 ± 0.23 442.2 0.783 Dark experimental 38.29 ± 0.21 432.4 0.771 Light control 38.33 ± 0.19 391.6 0.695 Darl< control 38.17 ± 0.23 3996 0.719 Light control with food supplement 39.60 ± 0.24 473.6 0.763 'K = W X 10^//-^ where W is weight in milligrams and/, is length in millimeters. The average length of fry emerging from the redds (Table 1) did not differ significantly from that of the unfed siblings used in the photore- sponse tests (Table 4). However, the emerging fry weighed somewhat less than fry of the experi- mental groups but more than those of the control groups (X = 425.7 mg) and were in similar condi- tion to the experimental groups (K = 0.766). DISCUSSION Emergence from these simulated redds in- volved several differences from that reported by Koski (1966) for natural redds of coho salmon. Fry from individual natural redds took from 10 to 47 days {X = 35 days) to complete emergence which peaked 8-10 days after first emergence, and size of fry decreased as emergence proceeded. In the simulated redds, duration of emergence was 20-23 days peaking at 12-13 days and size increased with time although yolk reserve re- mained nearly constant. The physical structure of the natural redd, particularly the proportion of smaller particle sizes, restricted permeability and impeded emergence. Low permeability re- duced size of fry and increased mortality, later- emerging fry and those failing to emerge that were excavated from redds were emaciated, weight loss indicating exhaustion of yolk prior to emergence. As yolk reserves remained fairly constant throughout emergence from the simulated redds, the larger, later-emerging fry probably developed from larger eggs. Koski (1966) found that large female spawners produced large fry at emer- gence, but large size of progeny did not alleviate physical hindrance to emergence, typifying the majority of redds, leading to decreasing size of fry as emergence progressed. The strong upstream response shown by fry emerging from the simulated redds is charac- teristic of coho fry emerging in natural streams. Apart from counteracting downstream transport, upstream movement provides for the seeding of upstream rearing areas unavailable to, or not used by, spawners. The small but significant dif- ference in yolk reserve between fry moving up- stream or downstream may reflect, rather than a minor difference in swimming ability, behavioral differences associated with rising aggression, onset of territoriality, and commencement of feeding on the invertebrate drift. The lack of preference for nocturnal emergence is in contrast to findings for sockeye salmon; pink salmon, O. gorbuscha; and chum salmon, O. keta, fry which emerge primarily at night (Neave 1955; Heard 1964). But like these other species, the coho salmon fry retained a photonegative re- sponse at emergence of potential survival value, e.g. escape from predators. Stuart (1953) also re- ported that fry of brown trout, Salmo trutta, re- mained photonegative during their ascent in simulated redds, even upon reaching positions only 1 or 2 inches from the gravel surface. For several days after emerging, fry of coho salmon and cutthroat trout, S. clarki, will bolt back into the gravel bed when disturbed (pers. obs.) and 173 FISHERY BULLETIN: VOL. 74, NO. 1 similar observations led Neave (1955) to com- ment that migrating chum and pink salmon fry, failing to reach the ocean in a single night, hide during the day and resume migration at night- fall. Hiding behavior disappears in coho salmon fry at time of complete yolk absorption but is re- tained for several days at high light intensities (Hoar 1958); this suggests a threshold intensity for the avoidance response which increases as the alevin stage proceeds. Concurrence between change in numbers of fry observed in the choice chambers, a collective re- sponse, and the accumulated number of emergent sibs could reflect either a sudden shift in photo- response of individual fry or gradual erosion of the negative response occurring simultaneously in all fry. The sudden shift alternative is best supported by three patterns of behavior noted in the choice chambers. Individual fry were ob- served to spend considerable time in the light compartment upon entering it, alternately swimming about slowly and remaining locally quiescent. Positions were commonly adopted with the head projecting into the light compartment (Figure 1), or entrance, and departure was rapid, irrespective of the presence or absence there of other fry until the last few days of testing when aggression was observed (Figure 6). Despite near depletion of vitellus at time of emergence, the shift in photoresponse did not ap- pear to be due to starvation because the response was not altered significantly by feeding. This is of interest as Smith (1952) reported marked meta- bolic changes in rainbow trout, S. gairdneri, ale- vins a few days prior to emergence, suggesting that these physiological events signified the onset of starvation. The change in photobehavior ap- pears to be an ontogenetic behavioral change normally associated with emergence from the redd rather than one instigated by nutritional deficiency, premature feeding, or light experi- ence. It remains unclear as to whether or under what conditions such stimuli can modify this change significantly; however, under hatchery conditions, Harvey (1966) found that sockeye salmon fry took food 2 wk after hatching but that emergence of fry from a simulated redd coincided with complete yolk absorption some 3 wk later. Heard (1964) noted that most emerging sockeye salmon fry trapped from natural redds in an Alaskan stream contained little or no yolk, re- mained photonegative, and emerged primarily during hours of darkness. The timing of the photoresponse change rela- tive to emergence and yolk reserves may vary within common limits for most stream salmonids and differences may reflect species-specific adap- tions of value to fishery biologists. As in the fry emerging from the simulated redds, the yolk re- serve of coho salmon fry emerging from natural redds averaged 7% (unpubl. data). Stuart (1953) observed a definite change in photoresponse of S. trutta when yolk neared depletion, and the photo- response change was employed by Gray (1929b) to denote the conclusion of incubation when measuring the effect of temperature on alevin size at time of yolk depletion. Woodhead (1957) disagreed with Stuart as to the timing of the photoresponse change in S. trutta, and asserted that it occurred coincident with maximum activ- ity of the alevin 15 days after hatching when yolk reserve constituted 70% of the dry weight of the fry. This considerable difference in timing re- mains unresolved. Denying the photoresponse fry streambed ex- perience during the last few weeks of the alevin stage had no apparent effect on the final size of the fry, probably due to their advanced stage of development prior to application of treatment dif- ferences. Marr (1963, 1965) has shown that de- velopmental efficiency is reduced by exposure to natural light or lack of substrate contour which stimulate locomotor activity at the expense of growth. However, marked effects on locomotor ac- tivity were only measurable until development was 75-80% complete. The weight disparity be- tween experimental and control groups of fry (Table 4) which was the outcome of weight loss or reduced weight gain is presumed to be an out- come of reduced feeding opportunity. In summary, the present results show that coho salmon fry underwent a definite shift (sudden or otherwise) from a strong to a weak negative photoresponse. This shift was accompanied by a positive response to water current leading to pre- ferred movement upstream. The emerging fry was an actively feeding animal yet to fill, or in the process of filling, its air bladder, fed in the gravel prior to emergence, and emerged when av- erage yolk reserves declined to 7% of total dry weight. In contrast to fry emerging from natural redds (Koski 1966), later-emerging fry were larger than those emerging earlier and may have derived from larger eggs. Because first-emerging fry held ecological advantage over later-emerging fry in stream aquaria (Mason and Chapman 174 MASON: FEATURES OF EMERGING COHO SALMON FRY 1965), the timing of emergence and environmen- tal conditions which modify it and the ecological state of fry at emergence should be fruitful con- siderations in future research. ACKNOWLEDGMENTS I am grateful to R. A. Bams and W. Percy Wickett (Pacific Biological Station) for con- structive criticism, and to D. W. Rimmer for tech- nical assistance. LITERATURE CITED Ali, m. a. 1959. The ocular structure, retinomotor and photobe- havioral responses of juvenile Pacific salmon. Can. J. Zool. 37:965-996. BAMS, R. A. 1969. Adaptations of sockeye salmon associated with incubation in stream gravels. In T. G. Northcote (editor), Salmon and trout in streams, p. 71-87. H. R. MacMillan Lectures in Fisheries. Univ. British Columbia, Vancou- ver, B.C. DILL, L. M. 1969. The sub-gravel behavior of Pacific salmon larvae. In T. G. Northcote (editor), Salmon and trout in streams, p. 89-99. H. R. MacMillan Lectures in Fisheries. Univ. British Columbia, Vancouver, B.C. Gray, J. 1929a. The growth of fish. IL The growth-rate of the embryo oi Salmo fario. J. Exp. Biol. 6:110-124. 1929b. The growth of fish. HI. The effect of temperature on the development of the eggs of Salmo fario. J. Exp. Biol. 6:125-130. Harvey, H. H. 1966. Commencement of feeding in the sockeye salmon iOncorhynchus nerka). Verb. Int. Ver. Theor. Angew. Limnol. 16:1044-1055. Heard, W. R. 1964. Phototactic behaviour of emerging sockeye salmon fry. Anim. Behav. 12:382-388. HOAR, W. S. 1958. The evolution of migratory behaviour among juve- nile salmon of the genus Oncorhynchus. J. Fish. Res. Board Can. 15:391-428. KOSKI, K. V. 1966. The survival of coho salmon (Oncorhynchus kisutch) from egg deposition to emergence in three Oregon coastal streams. M.S. Thesis, Oregon State Univ., Corvallis, 84 p. MARR, D. h. a. 1963. The influence of surface contour on the behaviour of trout alevins S. trutta L. Anim. Behav. 11:412. 1965. Factors affecting the growth of salmon alevins and their survival and growth during the fry stage. Assoc. River Auth. Yearb., 1965:1-9. Mason, J. C. 1969. Hypoxial stress prior to emergence and competi- tion among coho salmon Iry. J. Fish. Res. Board Can. 26:63-91. Mason, J. C, and D. W. Chapman. 1965. Significance of early emergence, environmental rearing capacity, and behavioral ecology of juvenile coho salmon in stream channels. J. Fish. Res. Board Can. 22:173-190. NEAVE, F. 1955. Notes on the seaward migration of pink and chum salmon fiy. J. Fish. Res. Board Can. 12:369-374. ROYCE, W. F. 1959. On the possibilities of improving salmon spawning areas. Trans. North Am. Wildl. Conf 24:356-366. Smith, S. 1952. Studies in the development of the rainbow trout (Salmo irrideus). H. The metabolism of carbohydrates and fats. J. Exp. Biol. 29:650-666. STUART, T. A. 1953. Spawning migration, reproduction and younger stages of loch trout (Salmo trutta L.). Freshwater Salmon Fish. Res. 5:1-39. WHITE, G. M. 1915. The behaviour of brook trout embryos from the time of hatching to the absorption of the yolk sac. J. Anim. Behav. 5:44-60. WICKETT, W. P. 1952. Production of chum and pink salmon in a controlled stream. Prog. Rep. Pac. Coast Stn., Fish. Res. Board Can. 93:7-9. WOODHEAD, P. M. J. 1957. Reactions of salmonid larvae to light. J. Exp. Biol. 34:402-416. 175 FEEDING BEHAVIOR, FOOD CONSUMPTION, GROWTH, AND RESPIRATION OF THE SQUID LOLIGO OPALESCENS RAISED IN THE LABORATORY Ann C. Hurley^ ABSTRACT The squid Loligo opalescens was raised in the laboratory to a maximum age of 100 days on a diet of Artemia nauplii and adults. Newly hatched squid (2.7 mm mantle length) readily attacked Artemia nauplii (length 0.7 mm), Artemia adults (length 5 mm), copepods (length 1 mm), and larval fish (length 4 mm). Feeding rates varied between 35 and 80% of squid body weight per day. Growth rate was highly variable in different individuals, ranging from 0.5 to nearly 4.5 mm mantle length per month. Respiration rates were obtained at 15°C for squid of three different ages and at 10°, 15°, and 20°C for 1-day-old squid. The squid Loligo opalescens Berry is a common pelagic predator off the west coast of North America from British Columbia to Baja Califor- nia. Because a fishery exists for this species, con- siderable information is available concerning adults in the spawning schools (Fields 1965), but little is known about the early life stages. In a paper on larval squid abundance off California, Okutani and McGowan (1969) found few L. opalescens in their samples; and McGowan (1954) reported that despite considerable effort he could not catch newly hatched L. opalescens over the spawning grounds. To obtain information on the early life history, I reared L. opalescens in the laboratory. Several workers have succeeded in rearing decapod cephalopods, but all of the species they used tend to be closely associated with the bottom (Choe 1966, three species of Sepia, the squid Sepioteu- this lessoniana, the sepiolid Euprymna berry i; LaRoe 1971, S. sepioidea; Boletzky et al. 1971, four species ofSepiola and two species of Sepietta; Arnold et al. 1972, the sepiolid E. scolopes). At- tempts to raise pelagic species such as Loligo opalescens have met with little success (Fields 1965; Arnold et al. 1974). Workers have attrib- uted their failure to lack of food and to infec- tions. I describe here a simple technique for rear- ing early stages of L. opalescens and present data on the growth, respiration, and food requirements of L. opalescens reared for 100 days in the laboratory. 'Scripps Institution of Oceanography, University of Califor- nia, La Jolla, CA 92093. MATERIALS AND METHODS Five groups (referred to as groups 1 through 5) of squid have been reared, three ( 1 through 3) of which will be described in detail in this report. Eggs were collected from a water depth of 20 m off La Jolla, Calif, and were maintained in circulat- ing seawater at about 13°C. The young squid were transferred to the rearing tanks after they had hatched. Fields (1965) and McGowan (1954) have described the methods of egg deposition and structure of the egg masses in detail. The rearing tanks were cylindrical (122 cm diameter, 36 cm deep) and made of black fiber glass. Tanks were illuminated by fluorescent lights which had a cycle of 18 h light, 6 h dark. During the dark period, lights in other rooms of the aquarium building provided a source of dim light. The tanks were immersed in water baths which kept the temperature within the tanks be- tween 15° and 17°C. Squid were transferred to the rearing tanks with a beaker. Squid in groups 2 and 3 were counted during transfer. In group 1, the number of squid was estimated after the squid were in the tank. Groups 1 and 2 began with 300 squid; group 3 began with 250. The water in the tanks was noncirculating. Each tank was aerated by a gently bubbling air supply. The squid in group 1 were transferred to a holding tank on day 62 and on day 76, and on each day their tank was drained, cleaned, and refilled. Tanks 2 and 3 were both similarly cleaned on day 49. Dead food was removed from the bottom of all tanks with a siphon, and small amounts of seawa- ter were added to maintain a constant volume. Manuscript accepted September 1975. FISHERY BULLETIN: VOL. 74. NO. 1, 1976. 176 HURLEY: LOLIGO OPALESCENS RAISED IN THE LABORATORY During the first 4 wk the squid (groups 1 through 3) were fed newly hatched brine shrimp, Artemia salina, nauplii which were kept at densi- ties ranging from 1 to 20 nauplii/ml. After this time, small adult brine shrimp were added (aver- age length 5.4 mm; range 2.5 to 8.0 mm) and were the major source of nourishment for the remainder of the rearing period. In groups 4 and 5, small adult Artemm as well as nauplii were used as food during the first 4 wk. Squid were measured using an optical microm- eter on a dissecting microscope. Measurements are of dorsal mantle length (measured dorsally from the tip of the tail to the farthest anter- ior point on the mantle). Mantle length is less variable than a measurement of total length, which depends upon the degree of stretch of the arms and tentacles. To make possible conversions to total length, measurements were made of both dorsal mantle length (ML) and total length (to tips of arms, not tentacles) (TL) on 35 juvenile animals, and the average ratio ML/TL was 0.62 ± 0.014 ( ±2 SE). Measurements are all on freshly dead unpreserved animals. For weight measure- ments, squid were rinsed in distilled water and oven dried at 60°C to a constant weight. Respiration measurements were made using a Warburg constant volume respirometer with respiration vessels kept at constant temperature in a water bath. The respiration vessels contained from 2 to 30 squid and were kept in constant mo- tion by gentle shaking. Estimates were made of the number of squid surviving at intervals throughout the study. The number of squid alive on any day was the average of three counts taken of live animals in the tank. Daily observations were made of the feeding behavior of the squid. At various times through- out the day, a squid was selected and observed for about 5 min. The number of feeding attempts and successful captures of prey were recorded. RESULTS Survival Mortality in all of the tanks was initially high (Figure 1). This is similar to what LaRoe (1971) found in rearing Sepioteuthis sepioidea. LaRoe speculated that the high initial mortality was due to insufficient quantities of food. This probably was not the case in my studies, as a large amount lOOi 90 80 70 ^ 60 < > ff 50 to ^ 40 30- 20- 10 D GROUP I A GROUP 2 O GROUP 3 \ V& ''*^'^^'** ®-o-ocp-oo^ ^-. 2°=^ 10 20 30 40 50 DAYS 60 70 80 90 100 Figure l. — Estimated percent survival of Loligo opalescens in the rearing tanks. Group 1 started with 300 squid; group 2, with 300 squid; and group 3, with 250 squid. of food was continually available at this stage. Some of this mortality could have been caused by squid which did not initiate feeding. Fields (1965) found that L. opalescens which did not appear to be feeding lived up to 10 days and still had some internal yolk reserves left at the end of this time. From 30 to 60 days mortality was low, but after 60 to 70 days mortality again increased. It is pos- sible that the brine shrimp did not provide an adequate diet for squid older than 60 days. Feeding Behavior Attack The attack of a young L. opalescens is similar to that described for adult Loligo (Fields 1965), Sepioteuthis (LaRoe 1970), and Sepia (Messenger 1968). Messenger divided the Sepia attack into three motor patterns: attention, positioning, and seizure. These three patterns may also be used to describe the attack of yoimg L. opalescens. Dur- ing attention, the squid orients toward a particu- lar prey. The arms and tentacles are extended in front of the squid and form a tight cone which is pointed toward the prey. Color changes such as those noted for Sepioteuthis (LaRoe 1970) and Sepia (Messenger 1968) were not observed. 177 FISHERY BULLETIN: VOL. 74, NO. 1 After the squid oriented toward a particular prey, it approached the prey until it was within attacking distance (positioning). This distance was not constant. At times there was no clear separation between the attention and positioning patterns. LaRoe (1970) suggested that the posi- tioning approach is an example of an aggression- fear conflict. This appears to be the case in Loligo. The young squid would sometimes flee rapidly after closely approaching a large prey. The prey was usually captured with the tenta- cles (seizure), although occasionally the arms alone were used. The arms were used to maneu- ver the food toward the mouth. At times a new attack began while the squid was holding other prey in the arms. LaRoe (1970) reported that for Sepioteuthis sepioidea physical fights over food were rare. This was not true for young L. opalescens. Fighting between squid was never observed when prey was small (brine shrimp nauplii), but if the prey was large and could not be completely enclosed within the arms, other squid would often chase the one which caught the food and try to take the food away from it. Often several (in one case, four) squid held on to the captured prey and all fed on it. The prey would be tugged about until one squid pulled it away from the others. This be- havior occurred even when there was an abun- dance of prey in the tank. This attack on captured prey at times allowed small squid to eat larger prey organisms than they could normally subdue alone. Prey Selection Unlike Sepioteuthis (LaRoe 1971), young L. opalescens were not extremely selective as to the type and size of prey they would attack. Within a few days after hatching, the young Loligo (2.7 mm ML) readily attacked Artemia nauplii (0.7 mm long), Artemia adults (5 mm long), copepods (1 mm long), and larval fish (4 mm long). Occa- sionally, squid attacked and ate dead prey (e.g., 6.ea.6. Artemia dropped into the tank), but usually the food had to move before it was attacked. An exception to this was that the squid attacked fish larvae which appeared to be motionless in the water. When the squid were 17 days old, nine squid from group 2 were placed in a small cylindrical container (8 liters of water) to determine whether a food size preference existed in Loligo. The food used was Artemia nauplii (0.6 to 0.8 mm long) at 10/ml and small adult Artemia (2 to 4 mm long) at 0.2/ml. After the squid were added, I recorded the number of attacks until a prey was captured and the type of prey being attacked. If no prey was captured in 20 min, I selected another squid. At this age, the squid attacked both large and small prey. During the 164 min of observation, 23 nauplii were attacked (9 actually captured) and 30 adults were attacked (8 actually captured). These results are different from those given for Sepioteuthis sepioides (LaRoe 1971). That squid only attacked food species in a very limited size range. Within several days, Sepioteuthis would cease to attack the prey it had previously eaten and would only attack larger prey. This seemed to occur when the squid were 1 to IV2 times as large as their prey. Although Loligo captured both large and small prey with about equal frequency, a preference may exist for larger prey as their density in the container was much lower. An experiment was run with group 1 when the L. opalescens were 49 days old. In this case the choice was between two different prey species of approximately the same size. Two thousand 2-day-old chub mackerel, iScom6erJaponicus, lar- vae were added to one of the rearing tanks where the squid had been feeding on Artemia adults. There were approximately 2,000 Artemia in the tank. The same method was used to record feeding as in the previous experiment. Observation time in this case was 69 min. The squid showed a high incidence of attacks on fish larvae (52 attacks, 6 captures) even though the success rate was much lower than when attacking Artemia (4 attacks, 3 captures). This may indicate a preference for fish larvae, but without further experiments it is im- possible to say whether this is true. Feeding Success The ability of the squid to successfully complete an attack sequence depended on the size and species of prey and the age and experience of the young squid. Figure 2 is a record of the percent of successful attacks on Artemia nauplii as a func- tion of the age of the squid. Each point is an aver- age from the squid observed during that day. The number of squid observed per day ranged from 5 to 11, with the total daily observation time rang- ing from 25 to 55 min. The attack efficiency in- creased with the age of the squid, but a number of prey were still being lost even after 3 wk. LaRoe 178 HURLEY: LOLIGO OPALESCENS RAISED IN THE LABORATORY 100. 90- 80 70 60 50 40- 30 20- 10 • • • • — I 1 10 IB DAYS AFTER HATCHING 25 —I 30 Figure 2. — Percent of attacks on Artemia nauplii which were successful as a function of the age of the squid. (1970) found that for Sepioteuthis , the majority of the prey were lost because the squid were unable to judge the attack distance. In my experiments, most unsuccessful attacks occurred because the prey managed to escape after being initially struck. Some of the variability in success rates may have been due to different motivational states of the squid. Feeding Rates Several methods were used to determine the food ration of the developing squid. When the squid fed on nauplii, feeding rates were deter- mined at irregular intervals by choosing a squid and watching it for 5 min to determine the number oi^ Artemia nauplii consumed during this period. All of the observations accumulated dur- ing a given week were combined. For each week, I calculated the food eaten over a 24-h and 18-h feeding period. The squid captured prey when the Table l. — Estimated feeding rates (percent body weight eaten {)er day) of squid in rearing tanks. Each value is average for all values for a given week. Values through week 4 are based upon observed short-term feeding rates on Artemia nauplii and are given for assumed 18- and 24-h feeding periods. Subsequent values are based on counts of Artemia adults consumed in tanks 1 and 2. Na jplil Adults Week 18 h 24 h Tank 1 Tank 2 1 46 60 — __ 2 46 61 — 3 47 63 — — 4 5 6 37 50 — — 36 45 7 — — 67 80 8 — — 48 51 overhead lights were off, but it was not possible to establish how much was eaten. When adult Ar- temia was the primary source of nourishment, record was kept of the approximate number of food organisms introduced to the tank and their average weight. There is some error introduced here because some of the brine shrimp died and were not consumed. The average weight of the squid during each week was obtained from the growth data and length-weight relationships pre- sented in the next section. Average weight of Artemia adults was 0.3 mg (obtained from six random samples of 10 to 20 individuals each) and average weight of nauplii was 0.002 mg (John R. Hunter pers. commun.). Food consumption is shown in Table 1. One short-term experiment was performed to examine the feeding rate of 36-day-old squid on yolk-sac larval anchovies. Five squid were placed with 100 anchovy larvae in 8 liters of water and were left for 285 min. At the end of this period 58 larvae had been eaten. This gives a feeding rate of 2.4 larvae/squid • hour Theilacker and Lasker (1974) gave the average weight of a larva of this size as 0.022 mg. Using this information and the average weight of the squid, a feeding rate of 0.028 mg anchovy/mg squid h is obtained. Growth Since the number of squid being reared was small, specimens were not sacrificed for growth measurements alone. Every time a squid died, it was immediately measured. These measure- ments constitute the majority of the points on the growth curve shown in Figure 3. The points indi- cated by the x's are measurements which were made on squid that had been selected while alive 179 FISHERY BULLETIN: VOL. 74, NO. 1 Figure 3. — Size data ior Loligo opalescens. A dot denotes mea- surement made on squid which had died, and x denotes mea- surement made on squid that had been selected while alive to give an indication of the size range of individuals in the tanks. For days 1, 17, and 22, the numbers of squid measured, means, and ranges are given. The upper solid line gives a constant growth rate of 4.5 mm/mo. The lower one gives a rate of 0.5 mm/mo. to give an indication of the full size range of squid in the tank. Since the squid were not randomly sampled during this time, Figure 3 cannot be taken to give an average growth rate for the population, but it does give an indication of the range of growth rates. There was a large differ- ence in the rates of growth of individuals. Maxi- mum growth rates were nearly 4.5 mm/mo (upper line in Figure 3). Minimum growth rates were 0.5 mnVmo (lower line in Figure 3). The linear regression equation for the log length-log weight relationship for the developing squid is log weight (mg) = -1.22 + 2.37 log length (mm) with little scatter around the re- gression line. Respiration Measurements were taken of the oxygen con- sumption of young L. opalescens using a Warburg respirometer and a constant temperature water bath. Measurements were taken at 15°C for squid of three different ages and at 10°, 15°, and 20°C for 1-day-old squid (Table 2). Average oxygen consumption values are as follows: 1 day, 10°C, 1.5 /ul 02/mg squid h; 1 day, 15°C, 2.5 /aI 02/mg squid -h; 1 day, 20°C, 3.5 /ul 02/mg squid h; 3 wk, 15°C, 3.5 )ul 02/mg squid h; 8 wk, 15°C, 3.7 /xl 02/mg squid h. These measurements may be ar- tificially high because of the crowding which oc- curred in the small respiration vessels. It was ob- served, however, that the oxygen consumption tended to decrease (at a given temperature) with increasing number of animals present in the same vessel. It is possible that these lower rates occurred because some of the animals became moribund in the crowded conditions. But this is not likely, since the respiration rates remained constant over the course of the 2-h experiments. To compare these measurements to those made by other investigators, conversion factors had to be obtained to transform dry weight to wet weight. The ratio wet/dry was calculated for nine juvenile squid and gave a mean of 5.4 ± 0.21 ( ±2 SE). Wet weights were calculated by placing the squid on the weighing pan, blotting it with filter paper, weighing it at measured time intervals, and extrapolating the line obtained to zero time. The previous rates expressed in terms of wet weight are: 0.28, 0.46, 0.65, 0.65, and 0.69 ix\ 02/mg squid h. These values are similar to those obtained by LaRoe (1971) for 2- and 6-day-old Sepioteuthis sepioidea (0.64 ixVmgh at 23°C) and with the figure of 0.60 fiVmgh for adult L. pealei, calculated from data in Redfield and Goodkind (1929). Table 2. — Oxygen consumption rates for Loligo opalescens. Respiration vessels had a volume of 18 ml and contained approx- imately 5 ml seawater. The duration of the experiments was 2 h. Temp. (°C) N Age of squid (days) Number of squid/vessel Range of oxygen consumption (/jl Oj/mg squid (dry wt) h) 10 15 20 15 15 3 3 3 1 2 1 1 1 21 56 10-30 8-25 10-21 10 2-3 1.4-1.6 2.1-3.6 3.2-3.8 3.5 3.5-3.9 DISCUSSION It is extremely difficult to assess the role which an animal such as L. opalescens plays in the California Current ecosystem. Estimates of popu- lation size of adults are very poor because of the difficulties involved in sampling large active 180 HURLEY: LOLIGO OPALESCENS RAISED IN THE LABORATORY animals. Fisheries statistics are not particularly helpful because the catches come mainly from a few locations. It has been possible to get some field information on the diet of the adult squid (Fields 1965) but these data are completely lacking on such necessary information as feed- ing rates. It appears to be equally difficult to obtain in- formation on young L. opalescens from field sam- ples. The young squid have well-developed eyes and are very sensitive to vibrations. Therefore, even the young are likely to be able to avoid many nets. Okutani and McGowan (1969) pub- lished data on the abundance of young L. opales- cens (size range 3.5 to 7 mm dorsal ML) taken in net tows during the California Cooperative Oceanic Fisheries Investigations cruises in 1954 to 1957. In their report, however, they emphasized the problems involved in sampling the young squid and stressed that the abundances given probably should only be used to compare relative abundances of different species. They found that L. opalescens was the third most abundant species of larval squid present in their samples, but that its abundance was quite low when com- pared to the most common fish larvae present (e.g., 0.008 times the abundance of northern an- chovy, ^n^raw/is mordax). If the role of a young L. opalescens as a predator is to be evaluated, it is necessary to know the type of prey which it eats. Fields (1965) has deter- mined the diet of the adult squid from an exami- nation of stomach contents, but to my knowledge no one has done a similar study on the very small squid. From the laboratory results presented in this paper, it appears that young L. opalescens must be considered as predators on a wide range of prey types and prey sizes. They are capable of preying on species ranging in size from 0.7 to 7 mm and they readily attack prey species ranging from brine shrimp adults and nauplii to copepods and larval fish. McGowan (pers. commun.) has found that they also successfully attack the mysid Metamysidopsis elongata. It is also possible to use the data presented here to estimate a feeding rate for the young squid. The respiration data can be used to calculate the amount of food a young squid would need to sus- tain itself. The respiration rate of the squid in the rearing tanks can be taken as 3 ix\ 02/mg dry wth. An average value for the caloric value of oxygen consumed is 5 x 10"^ cal//il of O2. There- fore, a newly hatched squid (2.7 mm ML, weigh- ing 0.625 mg) would use 0.22 cal for respiration alone in 24 h. It is possible to determine how many prey items of different types of prey would satisfy this re- quirement. A newly hatched Ar^emia nauplius is the equivalent of 0.0096 cal (John Hunter pers. commun.). Therefore, a newly hatched squid would need 23 Artemia nauplii per day. If the squid were instead feeding on newly hatched northern anchovies, it would need a total of 2 an- chovy larvae per day (using a value of 5 cal/mg, weight of larva = 0.022 mg; Theilacker and Lasker 1974). Similar calculations can be made for older squid. A squid 7 mm ML (~2 mo old, 6 mg) would consume 225 nauplii or 20 anchovy larvae simply to meet its metabolic needs. The actual amount of food consumed per day was ap- preciably more than this, averaging about 50% of body weight per day. At this rate, a newly hatched squid would consume 150 nauplii or 14 anchovy larvae per day, while a 7-mm squid would consume 1,500 nauplii or 135 anchovy lar- vae per day. Data on feeding rates and abundance could be used to calculate the impact that young squid might have on populations of potential prey items, but before such calculations can be mean- ingful, more information must be known about the ability of the squid to locate sources of food. Loligo opalescens was only one hundredth as abundant as the most common fish larvae (Oku- tani and McGowan 1969). But with feeding rates of 15 to 135 larvae per day, young squid could potentially have a large impact on such popula- tions if they concentrate on this type of food and if they have effective means of finding such prey. Laboratory observations indicate that larval fish may be a preferred food, and the squid do occur in areas where larval fish are common. Okutani and McGowan found that L. opalescens was most common in the upper 40 m, and this is the stratum where the highest abundance of north- ern anchovy larvae occur (Ahlstrom 1959). ACKNOWLEDGMENTS I thank J. Hunter, R. Lasker, and D. Lange for help during this work. This study was done while I was on a NOAA Associateship at the Southwest Fisheries Center, National Marine Fisheries Ser- vice, NOAA, La Jolla, Calif. 181 LITERATURE CITED Ahlstrom, E. H. 1959. Vertical distribution of pelagic fish eggs and larvae off California and Baja California. U.S. Fish. Wildl. Serv., Fish. Bull. 60:107-146. Arnold, J., W. Summers, D. Gilbert, R. Manalis, N. daw, AND R. Lasek. 1974. A guide to laboratory use of the squid Loligo pealei. Mar Biol. Lab., Woods Hole, 74 p. Arnold, J. M., C. T. Singley, and L. D. Williams-Arnold. 1972. Embryonic development and post-hatching survival of the sepiolid squid Euprymna scolopes under laboratory conditions. Veliger 14:361-364. BOLETZKY, S. VON, M. V. VON BOLETZKY, D. FROSCH, AND V. GATZL 1971. Laboratory rearing of Sepiolinae (Mollusca: Cephalopoda). Mar Biol. (Berl.) 8:82-87. CHOE, S. 1966. On the eggs, rearing, habits of the fry, and growth of some Cephalopoda. Bull. Mar. Sci. 16:330-348. FIELDS, W. G. 1965. The structure, development, food relations, repro- duction, and life history of the squid Loligo opalescens Berry. Calif. Dep. Fish Game, Fish. Bull. 131, 108 p. LaROE, E. T. 1970. The rearing and maintenance of squid in confinement FISHERY BULLETIN: VOL. 74, NO. 1 with observations on their behavior in the laboratory. Ph.D. Thesis, Univ. Miami, Coral Gables, Fla., 136 p. 1971. The culture and maintenance of the loliginid squids Sepioteuthis sepioidea and Doryteuthis plei. Mar Biol. (Berl.) 9:9-25. MCGOWAN, J. A. 1954. Observations on the sexual behavior and spawning of the squid, Loligo opalescens, at La Jolla, Califor- nia. Calif. Fish Game 40:47-54. Messenger, J. B. 1968. The visual attack of the cuttlefish, Sepia officinalis. Anim. Behav. 16:342-357. OKUTANI, T., AND J. A. MCGOWAN. 1969. Systematics, distribution, and abundance of the epiplanktonic squid (Cephalopoda, Decapoda) larvae of the California current, April, 1954-March, 1957. Bull. Scripps Inst. Oceanogr., Univ. Calif. 14, 90 p. Redfield, a. C, and R. GOODKIND. 1929. The significance of the Bohr effect in the respiration and asphyxiation of the squid, Loligo pealei. J. Exp. Biol. 6:340-349. Theilacker, G. H., and R. Lasker. 1974. Laboratory studies of predation by euphausiid shrimps on fish larvae. In J. H. S. Blaxter (editor), The early life history of fish, p. 287-299. Springer- Verlag, N.Y. 182 CONTRIBUTION OF THE NET PLANKTON AND NANNOPLANKTON TO THE STANDING STOCKS AND PRIMARY PRODUCTIVITY IN MONTEREY BAY, CALIFORNIA DURING THE UPWELLING SEASON David L. Garrison^ ABSTRACT Net plankton and nannoplankton standing stocks and primary production were measured in Mon- terey Bay, Calif, from January through August 1972. Throughout the period of seasonal upwelling, the phytoplankton stocks were dominated by net plankton. Both fractions showed seasonal changes: the net plankton concentrations increased dramatically during upwelling, while nannoplankton concentrations were decreased. Nannoplankton growth rates exceeded net plankton rates at in- cubator light levels; however, at higher in situ light levels near the surface, this relationship ap- peared to be reversed. Nannoplankton decreases may have been related to their selective removal from the area of up- welling by horizontal advection or selective grazing on the nannoplankton fraction. Net plankton dominance during upwelling has been related to their higher growth rates when populations are retained in shallow nutrient-rich nearshore waters. Frequently, phytoplankton are divided into two size classes, depending on whether they are re- tained by fine mesh nets (net plankton) or pass through the mesh (nannoplankton). The inade- quacy of net collections for estimating standing stocks or production is clear. The standing stocks of the two fractions and their relative contribu- tions to primary productivity, however, are less well-known. The size distribution, which may be environmentally controlled (Semina 1972; Par- sons and Takahashi 1973), is an important fea- ture of the phytoplankton populations because the size of the primary producers may affect the length and efficiency of pelagic food chains (Ryther 1969; Parsons and LeBrasseur 1970). The purpose of this study was to determine the rela- tive importance of the two fractions during the upwelling season in Monterey Bay, a neritic envi- ronment of the California Current system. Most previous studies reported that the nan- noplankton fraction usually exceeds the net plankton fraction, often accounting for 80 to 100% of the standing stocks and primary produc- tion (e.g., Steeman Nielsen and Jensen 1957; Holmes 1958; Yentsch and Ryther 1959; Kawa- mura 1961; Holmes and Anderson 1963; Teixeira 1963; Gilmartin 1964; Saijo 1964; Anderson 1965; 'Moss Landing Marine Laboratories, Moss Landing, CA 95039; present address: Coastal Marine Laboratory, University of California, Santa Cruz, CA 95064. Manuscript accepted September 1975. FISHERY BULLETIN: VOL. 74. NO. 1. 1976. Saijo and Takesue 1965; Malone 1971a, c; Parsons 1972; McCarthy et al. 1974). Only a few authors reported net plankton dominated communities (Digby 1953; Subrahmanyan and Sarma 1965). It is difficult to compare these studies, however, be- cause mesh sizes of 22 to 110 ^im have been vari- ously used to separate the net plankton and nan- noplankton fractions. The nannoplankton fraction may show little seasonal fluctuation, while the net plankton shows pronounced seasonal trends with periods of abundance corresponding to increased water temperatures (Yentsch and Ryther 1959), peak periods of primary production (Subrahmanyan and Sarma 1965), or seasonal upwelling (Malone 1971c). Malone (1971a) reported higher net: nanno ratios for standing stocks and production in neritic environments as compared with oceanic areas and pronounced onshore to offshore lower- ing of the ratio in the California Current region during upwelling (Malone 1971c). The growth rate (as indicated by the assimilation ratio = mg C mg Chi a"^ h"^) of the nannoplank- ton fraction is greater than that of the net plankton fraction (Yentsch and Ryther 1959; Saijo and Takesue 1965; Malone 1971a, c). Arguments presented for the predominance of net plankton or nannoplankton in a given envi- ronment relate cell area to volume ratios (Malone 1971a, c; Eppley 1972; Parsons and Takahashi 1973). There is a general relationship between 183 FISHERY BULLETIN: VOL. 74, NO. 1 cell size and the ability to take up nutrients (Dugdale 1967; Eppley et al. 1969; Eppley and Thomas 1969). Large species generally have higher half saturation constants (Kg) and may have higher maximum uptake rates (Vmax)' whereas small species have low^er Kg and V^ax (Dugdale 1967). Maximum net plankton grovi'th rates are favored at higher ambient nutrient concentrations while nannoplankton reach their maximum growth rates at lower ambient nutrient levels. There is also a direct relationship of in- creasing cell size (or chain length) with increas- ing sinking rates (Smayda 1970), and larger cells and chain formers tend to be aggregated in areas of upward advection, while motile or posi- tively buoyant cells tend to be concentrated in areas of downward advection (Stommel 1949). Net plankton will have a longer residence time in the euphotic zone and concentrate in areas of upwelling, while the nannoplankton (if the population is primarily motile flagellates) will be concentrated in areas of downwelling. Parsons and Takahashi (1973) related the growth rate (/u,) to physiological characteristics of the cell (maximum grovvi:h rate, half saturation constants for nutrients and light, and sinking rates) and environmental conditions (incident radiation, extinction coefficients, mixed layer depth, and upwelling rates) and used the rela- tionship to explain characteristic phytoplankton cell size in a number of environments. Recently, Laws (1975) expanded the Parsons and Taka- hashi model and showed that under certain light conditions the decreasing respiration rate with increasing cell size may regulate the growth rate of large versus small cells. The effect of grazing on the net:nanno ratios and, conversely, the size of the primary producers on food chains have not been well documented. Grazing may ultimately control net plankton stocks (Malone 1971c; Ryther et al. 1971) and de- termine the lower net:nanno standing stock ra- tios in oceanic as opposed to neritic areas (Malone 1971a). Grazing has been suggested as the pri- mary cause for failure of phytoplankton stocks to develop in otherwise favorable waters (Mc Allister et al. 1960; Strickland et al. 1969). Shorter food chains have been shown for some clupeid fishes which feed directly on the large phytoplankton species (e.g., Bayliff 1963; Rojas de Mendiola 1969; Dhulkhed 1972) and for her- bivorous euphausids in the diatom-rich antarctic region (Marr 1962). The general argument for larger phytoplankton cells resulting in shorter, more efficient food chains may not always apply to the smaller grazers, as Parsons and LeBras- seur (1970) have reported on selective feeding re- lated to cell shape. Previous studies have been made on the hydro- graphic seasons in Monterey Bay and their rela- tionship to the seasonal phytoplankton blooms (Bolin and Abbott 1963; Abbott and Albee 1967). Malone (1971c) reported the seasonal variability of the net plankton and nannoplankton in the California Current, which included one deep sta- tion on the edge of Monterey Bay. The present study was part of a monthly sampling program conducted by Moss Landing Marine Laboratories to provide information on the hydrographic con- ditions and plankton populations in Monterey Bay, particularly from the extensive shallow areas of the bay. Although it was not possible to carry this study through a complete seasonal cy- cle, information is presented for the upwelling period, when seasonal blooms of phytoplankton appear in Monterey Bay. MATERIALS AND METHODS Measurements of primary productivity and phytoplankton standing stocks were made at sta- tions 3 and 8 for the period January through Au- gust 1972 and at station 15 for the period June through August 1972 (Figure 1). The stations were located over the Monterey Submarine Can- yon at depths of 110, 240, and 718 m, respectively. Samples were taken monthly during hydrograph- ic and plankton cruises conducted by Moss Land- ing Marine Laboratories and, occasionally, be- tween these periods on instructional cruises. Sampling times varied between cruises but fell between 0700 and 1100 h. Samples were collected with 5-liter Niskin water sampling bottles from depths correspond- ing to 100, 50, 25, 10, and 1% light penetration levels as measured with a submarine photometer or calculated using the relationship: depth of 1% light = 3.5 X Secchi disk (Silver and Hansen 1971a). Hydrographic parameters (salinity, °L; temperature, °C; O2) and nutrients (PO4, NO3, NO2, NH3, Si02) were samples at standard depths (Broenkow and Benz 1973). Primary productivity was measured using the carbon-14 method (Steeman Nielsen 1952). For each depth two light and one dark bottles were innoculated with 5 or 10 fxCi of Naa^^COg. The 184 GARRISON: NET PLANKTON AND NANNOPLANKTON IN MONTEREY BAY MONTEREY BAY 37* N Vf 40' 36'3r Figure l. — Location of stations in Monterey Bay. Broken lines indicate the position of the 100-fathom (183-m) contour line. samples were incubated immediately after collec- tion for 3 to 4 h in a shipboard incubator (Doty and Oguri 1958) using Luxor Magnalux fluores- cent lamps^ (approx. 0.06 langley min"^). Neutral density filters of 50, 25, 10, and 1% transmittance were used on subsurface samples. The net plankton and nannoplankton fractions were separated by passing the samples through a 22-/xm Nitex-net filter (net plankton) and then a Gelman, type A glass-fiber filter having 0.3-/>tm pore size (nannoplankton). Both filters were washed with approximately 20 ml of freshly filtered seawater and placed directly in scintilla- tion fluor for counting at a later time. All samples were counted for at least 10 min with a Nuclear Chicago (Unilux II) scintillation counter. Carbon uptake was calculated as out- lined in Strickland and Parsons (1968). Since Malone (1971b) reported no diurnal periodicity in assimilation ratios in the California Current re- gions, daily production was estimated by using ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. the sunrise to sunset interval as the day length and multiplying by the hourly production rates that were determined during the first part of the day. Phytoplankton standing stocks were measured as chlorophyll a by using the fluorometric method of Holm-Hansen et al. (1965). The Turner fluoro- meter (model 111) was calibrated using the spectro- photometric method for chlorophyll a as outlined by Strickland and Parsons (1968). The two size fractions were separated by taking two replicate samples from each depth and passing one through a Gelman glass-fiber filter (total chloro- phyll) while the other sample was filtered through 22 ;um Nitex-net filter and then a glass-fiber filter (nannoplankton). Both filters were immediately frozen, stored in the dark, and analyzed within a month after collection. Net plankton was calcu- lated as the difference between total chlorophyll and nannoplankton chlorophyll. Productivity and chlorophyll a values deter- mined for the discrete samples were integrated to the depth of the 1% light level by trapezoidal ap- proximation. Carbonxhlorophyll a ratios vary widely and depend on light and nutrient condi- tions. For most of the study, nutrient levels were high and a C:Chl a ratio of 40 was used to convert chlorophyll a to carbon biomass (Lorenzen 1968; Eppley et al. 1970; Eppley et al. 1971). Phyto- plankton growth rate and standing stock dou- bling time were calculated using exponen- tial growth expression. RESULTS In January, the weak thermal gradient in the upper 50 m (Figure 2) is indicative of the David- son Current period, when the subsurface counter- current extends to the surface and flows north- westward on the inshore side of the California Current (Reid et al. 1958; Bolin and Abbott 1963; Smethie 1973). Rising isotherms and nitrate iso- pleths from February through May indicate up- welling over the Monterey Submarine Canyon. After May there was a slacking or an end to up- welling, and the isotherms and isopleths are found progressively deeper as denser upwelling waters subside. In July and August, conditions of the oceanic period were evident with low nutrient levels, higher surface temperatures, and lower salinities; however, upward movement of the isotherms and isopleths in August may indicate a developing upwelling pulse. 185 FISHERY BULLETIN: VOL. 74, NO, 1 300 Jan. Feb. Mar. Apr. May June July Aug. Q. 0.1 0.6 29 Aug. 3 30 M.4 20.2 0.094 0.507 1.4 0.7 8 30 14.6 32.4 0309 0.612 0.6 0.6 15 29 '3.2 18.7 0.488 0.252 2.3 0.4 'Value appears low, corresponding growth rate [fJi) may be too high. seasonal effect during upwelling seems to be a reduction of the average concentration of nan- noplankton and an increase in the average con- centration of net plankton. ro I E oi o o> 2 E 9.26 mg Chi a M'' 4 6.88 1^ . 4 + + Jan. Feb Mar. Apr. May June July Aug. Figure 5. — Seasonal changes in the concentration of net plankton chlorophyll (heavy line) and nannoplankton chloro- phyll (thin line). (Davidson Current period — January; upwell- ing period — February through June; oceanic period — July, August.) Average and range of concentrations in the euphotic zone are shown. The number of samples for each month is given in Table 1. Standing Stock Growth Rate The growth rate, /x (doublings day"^), and as- similation ratio (mg C mg Chi a'^ h"^), of the nan- noplankton fraction was greater than the corre- sponding value for the net plankton during all three seasons, and both fractions showed their highest growth rate during the upwelling period; however, assimilation ratios of the surface sam- ples for both fractions were higher in the oceanic period than during upwelling (Table 2). There is no correlation {P > 0.10) between the growth rates of either phytoplankton fraction and aver- age nutrients (NO3, Si02) in the upper 10 m on individual sampling days for the three hydro- graphic periods. Net plankton growth rates exceeded nanno- plankton growth rates in only two of the samples; Table 2. — Growth rates of the standing stocks in the euphotic zone and assimilation ratios of surface samples.' Hydro- graphic period Growth rate, fj. (doublings day-') Assimilation ratio (mg C mg Chi a-' h-') Net Nanno Net Nanno Davidson Current Upwelling Oceanic 0.1 0.4 0.2 ± 0.0(2) 0.3 ± 0.1(2) ±0.1(9) 0.9 ± 0.2(9) ± 0.3(4) 0.7 ± 0.2(6) 0.4 ± 0.2(2) 2.2 ± 0.5(2) 2.7 ± 1.5(9) 5.2 ± 2.2(9) 3.0 ± 1.6(3) 10.3 ± 1.2(4) 'Growth rates were calculated from daily productivity and standing stock estimates integrated to the depth of 1% light penetration, while assimilation ratios are for surface samples incubated at 0.06 langley min*'. 7 ± SD(W); questionable data indicated in Table 1 have been excluded. 187 FISHERY BULLETIN: VOL. 74, NO. 1 however, the growth rates were determined at in- cubator light levels which were not representa- tive of in situ conditions. The regression of light level on the ratio of the growth rates (/u net:/u, nanno) is significant (P < 0.01) during the up- welling months (Figure 6). Light levels approxi- mately equivalent to full incubator light are found at depths of 8 to 15 m during the upwelling period, and the upper one-fourth to one-third of the euphotic zone receives light which is in excess of incubator light levels. o c c o c c CO 0) o .06 0.6 lO'^langley min' Figure 6. — Regression of incubator light levels on the net'.nanno growth rates. Distribution in the Water Column Since nannoplankton concentrations were rela- tively homogeneous in the water column, max- ima were often not well defined. Net plankton maxima, however, were usually apparent and cor- responded to the depth of the seasonal pycnocline. There was no regularly observed depth relation- ship between nannoplankton and net plankton maxima, and they often were at the same depth. Phaeophytin peaks appeared at the surface and in conjunction with, or just below, the chlorophyll maxima. High NH3 concentrations in the deeper phaeophytin maxima may be indicative of grazing on the phytoplankton stocks in the chlorophyll maxima (see Figures 7-10). During the Davidson Current period there is little vertical stability in the water column, and the net plankton stocks are poorly developed (Figure 7). With the onset of upwelling net plankton stocks develop above the strong, shal- low pycnocline (Figures 8, 9) and the nanno- NOj ;ug atomt liter'' NH, IO"'>jg atoms liter"' 10 20 a. 0) Q 100 25.00 26.0O «^ Figure 7. — Vertical distribution of phytoplankton standing stocks, phaeophytin, and hydrographic parameters during the Davidson Current period. NOj^jg atoms liter'' NH3 IO''*jg atoms liter'' 10 20 Ot x: 50 o. Q 100 25.00 26.00 O-f Figure 8. — Vertical distribution of phytoplankton standing stocks, phaeophytin, and hydrographic parameters during upwelling period. Station was sampled during a flowing tide. plankton stocks decline. With strong or persistent upwelling, the pycnocline may intersect the sur- face and the phytoplankton stocks are concen- trated in a relatively shallow layer (Figure 9). After a slacking of upwelling the denser waters subside and the pycnocline depths become pro- gressively deeper. The surface layer can be strongly stratified by the onshore movement of warmer, low salinity oceanic water, and nutrient concentrations in the near surface waters are low during the oceanic period. The net plankton 188 GARRISON: NET PLANKTON AND NANNOPLANKTON IN MONTEREY BAY NO3 xig atoms lifer'' NHj IO-'>ug atoms liter"' 10 20 r- -c 50 a Q 100 Figure 9. — Vertical distribution of phytoplankton standing stocks, phaeophytin, and hydrographic parameters during upwelling period. Station was sampled during an ebbing tide. NOj iug atoms liter'' NH3 IO''4Jg atoms liter'' jz 50 Q. (U Q 100- STATION 8 18 JUL. 72 25.00 26.00 Figure 10. — Vertical distribution of phytoplankton standing stocks, phaeophytin, and hydrographic parameters during the oceanic period. maximum remains associated with the sinking pycnocline and, although nutrients do not reach limiting concentrations in the pycnocline, light levels are below optimal intensity for maximum growth rates (Figure 10). Broenkow and McKain (1972) demonstrated that tidal effects have a marked influence on the distribution of hydrographic parameters over the canyon: during a flow tide there is a down-canyon current and isotherms and isopleths over the canyon are depressed; conversely, during an ebb tide the flow is up the canyon and isotherms and isopleths are nearer the surface. The source wa- ters for the down-canyon flow are subsurface wa- ters from the shallow areas adjacent to the can- yon. These tidal effects can be identified in the distribution of the phytoplankton stocks (Silver and Hansen 1971b), but their importance is un- known. The chlorophyll a maximum at station 8 (in Figure 8) appears to be an intrusion of stocks developed in shallower areas and carried to depth by the down canyon flow during the flow tide. Station 3 was sampled earlier during an ebb tide, and the sigma-t surface at 50 m (crt = 26.14) was found deeper than 100 m at station 8 (see Figure 8). At a full ebb tide the pycnocline and the stand- ing stocks may be located very near the surface (Figure 9). DISCUSSION The net plankton-dominated blooms that de- veloped during this study were similar to those described by Bolin and Abbott (1963) and Abbott and Albee (1967) in their close association with seasonal upwelling and in their composition (i.e., the net plankton was dominated by colonial diatoms — M. Silver unpubl. data^). Malone (1971c) noted an increase in net plankton fraction during the upwelling season; however, he re- ported net plankton dominated stocks only dur- ing strong upwelling pulses. Malone also reported a marked decrease in net plankton chlorophyll and productivity between inshore and offshore stations near the end of the upwelling season. Al- though these studies cannot be directly com- pared, they suggest phytoplankton blooms which develop during upwelling are mostly net plank- ton forms, and higher standing stocks may develop inshore. There seems to be a fundamental contradiction in the measured growth rates of the two fractions and the observed standing stocks. The growth rates of the nannoplankton were consistently higher than those of the net plankton, whereas the standing stocks of nannoplankton decrease and the stocks of net plankton increase during the upwelling season. The observed development of the stocks could result theoretically from one or a combination of the following conditions: 1) ^The unpublished data supplied by M. Silver can be found in a data report filed in 1971-72 at Oceanographic Services, Inc., 135 East Ortega Street, Santa Barbara, CA 93101. 189 FISHERY BULLETIN: VOL. 74, NO. 1 the nannoplankton fraction may be selectively removed from the area by horizontal advection because of their low sinking rates; 2) nanno- plankton may be selectively grazed; 3) environ- mental conditions may favor higher net plankton growth rates. Malone (1971c) discussed the argument for selective removal of nannoplankton from upwell- ing areas by horizontal advection. Briefly re- stated, nannoplankton cells tend to have slower sinking rates than net plankton cells (or they are motile) and in convection cells they will tend to be removed from the areas of upward movement and concentrated in areas of downward movement (Stommel 1949). In upwelling areas then, nan- noplankton may be selectively removed by mass transport of surface waters offshore. There is lit- tle direct evidence to show that this takes place; however, the advection hj^othesis is supported by the observed decrease in nannoplankton stocks between the Davidson Current period and the upwelling period. During the Davidson Cur- rent period there is a general onshore movement of surface waters with water sinking along the coast, while during the upwelling period the cir- culation is reversed and water moved upward along the coast, and the surface waters are trans- ported offshore (Skogsberg 1936; Bolin and Ab- bott 1963). Malone (1971c) found the level of the nannoplankton stocks remained relatively con- stant throughout the year; however, he reported that during periods of onshore water movement there was an enhancement which could be attrib- uted to concentrating the nannoplankton in an area of downward water movement. The decrease in nannoplankton stocks reported in the present study may have been the result of selective grazing by microzooplankton and planktotrophic larvae (Thorsen 1950; Beers and Stewart 1969; Parsons and LeBrasseur 1970). In this area many of the benthic invertebrates have their reproductive season during the spring (M. Houk pers. commun.)"*; increased grazing pres- sure by these larvae may have caused the de- crease in nannoplankton stocks. However, the extent of grazing on either fraction of the phy- toplankton in Monterey Bay is not known. Zooplankton samples were collected as part of the routine sampling program, but gelatinous ■•M. Houk, Department of Natural Science, University of California, Santa Cruz, CA 95064. and colonial phytoplankton could not be sepa- rated from the zooplankton for biomass estimates. Throughout the period of upwelling, nitrate levels in the upper 10 m remained high (> 5 ^ig atoms liter"^) and the chlorophyll maximum was frequently located near the surface. At these shallow depths light levels were in excess of in- cubator light levels (0.06 langley min"M. Eppley et al. (1969) have shown that the diatoms Skeletonema costatum and Ditylum brightwellii grow faster than Coccolithus huxleyi at high light levels (0.1 langley min"^) when nitrate levels are in excess of 0.8 /xg atoms liter"^, while at lower light levels (0.02 langley min"^), the situation is reversed and C. huxleyi will grow faster at any nitrate concentration. In situ nutrient and light conditions near the surface during the upwelling period should favor net plankton growth. In the present study and in that of Malone (1971c), growth rates of the net plankton were lower than the growth rates of the nannoplank- ton; however, the two fractions responded differ- ently to increasing light as showai by the ratio of the growth rates (/x net.fx nanno) increasing with higher light levels (Figure 6). The regression pre- dicts that net plankton growth rates would ex- ceed the nannoplankton growth rates at light levels similar to those where Eppley et al. (1969) showed a reversal of growth rate relationships. Estimated light levels in the upper part of the euphotic zone are higher than the incubator light levels which have been used in this study and that of Malone. Since the net plankton growth rates show greater enhancement with increasing light than the nannoplankton, light levels in the upper water column may favor the growth of the net plankton fraction and lead to net plankton domination of the standing stocks. Laws (1975) suggested that, under certain en- vironmental conditions, large cells may realize a higher net growth rate because of a decreasing respiration rate with increasing cell size. In Laws' model, when surface light levels are low or the product of the attenuation coefficient and mixed layer depth is large, integral productivity efficiency is low and respiration losses become more important. During the present study, how- ever, under low light levels, the net growth rates of the smaller cells (nannoplankton) exceeded larger cells, and the phytoplankton populations were net plankton dominated at a time when the mixed layer was extremely shallow. Notwithstanding the possible effects of selec- 190 GARRISON: NET PLANKTON AND NANNOPLANKTON IN MONTEREY BAY tive grazing on the nannoplankton or their selec- tive removal by horizontal advection, the de- velopment of the upwelling bloom in Monterey Bay is largely a result of the increase in the net plankton fraction and may be explained in terms of conditions which are favorable for net plankton growth. High nutrient concentrations can be maintained in the euphotic zone by downward mixing from the surface which extends below the pycnocline or by a continual input of nutrients to the surface waters by upwelling. Optimal light levels, however, are found only in the upper part of the euphotic zone. The combination of these conditions that constitute optimal growth condi- tions for the net plankton fraction occur when the phytoplankton stocks are restricted to a shallow mixed layer above the pycnocline which has been "pushed up" by upwelling water. Optimal growth conditions vary spatially and seasonally and may be primarily responsible for the net plankton and nannoplankton relationship observed in Mon- terey Bay. Nutrients do not appear to limit the growth rates of either fraction as correlation coefficients of nutrient levels with growth rates were not sig- nificant and, although nutrient levels change seasonally, Malone (1971c) reported little sea- sonal variation in assimilation rates. Light levels, however, are potentially limiting a short distance from the surface and can influence the ratio of net:nanno growth rates. An increase in the depth of the mixed layer results in a decrease in the average light expo- sure for phytoplankton cells in the mixed layer (Parsons and Takahashi 1973). The net plankton fraction will be more strongly influenced than the nannoplankton because their optimal growth rates occur at light levels near the surface, and their vertical distribution is strongly controlled by water movement. Upwelling water move- ments result in a shallow pycnocline and shallow mixed layer; with a slack in the upwelling rate, the pycnocline sinks and there is a deeper mixed layer. In the present study, net plankton maxima were concentrated above the pycnocline, whereas no particularly strong relationship between the nannoplankton maxima were observed (the nan- noplankton maxima were often not well defined). Malone (1971c) showed that the net plankton maxima were located below the nannoplankton maxima during periods when upwelling was slack or that both were located at the surface dur- ing periods of upwelling, and he emphasized the role of upward movement in controlling the verti- cal distribution of the net plankton fraction. Malone (1971c) showed an onshore to offshore decrease in the ratio of net:nanno standing stocks. Yoshida (1967) showed the potential for a narrow zone of stronger upwelling associated with the edge of the continental shelf where the effects of upwelling are maximal at the edge of the shelf and decrease exponentially shoreward and seaward. A decrease in the upwelling rate away from the continental shelf would result in reduced suspension of sinking cells, a deeper mixed layer, and lower average light levels for phytoplankton cells in the mixed layer and could reduce the net:nanno growth rate ratio. Malone's data showed shallow mixed layers during periods of strong upwelling at inshore stations and a trend for an increasing mixed layer depth offshore. In Monterey Bay during the upwelling season, the mixed layer is frequently shallow or the pycnocline intersects the surface. There are considerable amounts of hydrographic data which show this characteristic distribution (Broenkow and Benz 1973) and corresponding phytoplankton standing stock data which show significant strat- ification of the phytoplankton standing stocks above the shallow pycnocline (Silver see footnote 3). The depth of the pycnocline and mixed layer vary seasonally in response to the upward move- ment of isotherms during upwelling and the sink- ing of isotherms when upwelling ceases. Upwell- ing, however, is not a continuous process and may be particularly sporadic near the end of the upwelling season (Bolin and Abbott 1963; Smethie 1973). Malone (1971c) reported net plankton dominated stocks only during periods of strong upwelling, which suggests that in deep water continual upwelling is necessary to main- tain optimal growth conditions for the net plankton fraction. During the present study the net plankton fraction dominated the phytoplank- ton populations in shallow water throughout the upwelling season. This evidence and previous evidence for an offshore decrease in the netrnanno ratios (Malone 1971c) suggest that physical pro- cesses in shallow water are sufficient to maintain net plankton populations and mitigate the lack of continual upwelling. The physical processes in shallow water that could serve to maintain favorable growth condi- tions for the net plankton fraction or maintain the population between periods of favorable con- 191 FISHERY BULLETIN: VOL. 74, NO, 1 ditions are poorly known. Tidal mixing and in- creased turbulence in shallow water could facili- tate cell suspension of sinking populations or resting spores, and increase nutrient input to the surface waters. Over Monterey Canyon and, to a lesser extent, in the shallow areas of the bay, the vertical distribution of nutrients (Broenkow and McKain 1972; Smethie 1973) and phytoplankton stocks (Silver and Hansen 1971b; Silver see foot- note 3) are strongly influenced by tidal effects. Turbulence and mixing in deep water results in a decrease in the average amount of light to which a phytoplankton cell is exposed; however, in shal- low water the depth of mixing is limited by the bottom and mixing here may result in resuspen- sion of sinking cells. Many of the neritic diatoms form resting spores which sink to the bottom and may be an important source of innoculum to ini- tiate blooms if they are resuspended by turbu- lence during favorable growth conditions. The decline in the net plankton populations during this study corresponded to the influx of oceanic waters in July. The end of net plankton domination of the population appears to have been the result of the low nutrient concentrations in the oceanic surface waters and subsidence of previously upwelled waters and its entrained net plankton populations. During oceanic conditions, nutrient levels in the surface waters favor the growth of nannoplankton and the light levels in the sinking net plankton maxima are not optimal for growth. Malone (1971c) suggested, however, that the net plankton are ultimately limited by grazers as the grazing index (phaeo:Chl a) in- creased and the netplankton concentrations de- creased even before the end of the upwelling period. Direct evidence for the extent of grazing in Monterey Bay is not available; however, when upwelling becomes sporadic and periodic influxes of oceanic water occur, the stage is set for a de- cline in the net plankton fraction without the need for an increase in grazing pressure. ACKNOWLEDGMENTS I am grateful for the help of David Seielstad, Sara Tanner, and many others who participated in the sampling cruises. I thank W. W. Broenkow, Scott McKain, and Sandra Benz for providing the hydrographic data. I am particularly indebted to Mary Silver for her encouragement, support, and advice throughout the study and during the prep- aration of this manuscript. Greg Cailliet re- viewed the manuscript and offered suggestions for its improvement. This research was supported by Grant 2-35137 from the office of Sea Grant Programs, National Oceanographic and Atmospheric Administration, Department of Commerce; the Association of Monterey Bay Area Governments; and the Soci- ety of the Sigma Xi and was based on a thesis submitted as a partial requirement for a M.A. degree at San Francisco State University, Calif LITERATURE CITED ABBOTT, D. P., AND R. ALBEE. 1967. Summary of thermal conditions and phytoplankton volumes measured in Monterey Bay, California 1961- 1966. Calif Coop. Oceanic Fish. Invest. Rep. 11:155-156. ANDERSON, G. C. 1965. Fractionation of phytoplankton communities off the Washington and Oregon coasts. Limnol. Oceanogr. 10:477-480. BAYLIFF, W. H. 1963. The food and feeding habits of the anchoveta, Centengraulis mysticetus, in the Gulf of Panama. Inter- Am. Trop. Tuna Comm. Bull. 7:399-432. BEERS, J. R., AND G. L. STEWART. 1969. The vertical distribution of microzooplankton and some ecological observations. J. Cons. 33:30-44. BOLIN, R. L., AND D. P. ABBOTT. 1963. Studies on the marine climate and phytoplankton of the central coastal area of California, 1954-1960. Calif Coop. Oceanic Fish. Invest. Rep. 9:23-45. BROENKOW, W. W., AND S. BENZ. 1973. Oceanographic observations in Monterey Bay, California January 1972 to April 1973. Moss Landing Mar. Lab., TP73-3, 336 p. BROENKOW, W. W., AND S. J. MCKAIN. 1972. Tidal oscillations at the head of Monterey Sub- marine Canyon and their relation to oceanographic sampling and circulation of water in Monterey Bay. Moss Landing Mar. Lab., TP72-5, 42 p. Dhulkhed, M. H. 1972. Food of the oil sardine taken by bottom nets and surface gill net in the Mungalore area. Indian J. Fish. 19:163-166, DiGBY, P. S. B. 1953. Plankton production in Scoresby Sound, East Greenland. J. Anim. Ecol. 22:289-322. DOTY, M. S., AND M. OGURI. 1958. Selected features of the isotopic carbon primary productivity technique. Rapp. P.-V. Reun., Cons. Perm. Int. Explor. Mer 144:47-55. DUGDALE. R. C. 1967. Nutrient limitation in the sea: dynamics, identifica- tion, and significance. Limnol. Oceanogr. 12:685-695. Eppley, R. W. 1972. Temperature and phytoplankton growth in the sea. Fish. Bull., U.S. 70:1063-1085. Eppley, R. W., a. F. Carlucci, O. holm-Hansen, D. Keifer, J. J. McCarthy, E. Venrick, and P. M. Williams. 1971. Phytoplankton growth and composition in ship- board cultures supplied with nitrate, ammonium, or 192 GARRISON: NET PLANKTON AND NANNOPLANKTON IN MONTEREY BAY urea as the nitrogen source. Limnol. Oceanogr. 16:741-751. EPPLEY, R. W,, F. M. H, REiD, AND J. D. H. Strickland. 1970. The ecology of the plankton off La Jolla, California, in the period April through September, 1967. Part III. Estimates of phytoplankton crop size, growth rate, and primary production. Bull. Scripps Inst. Oceanogr., Univ. Calif 17:33-42. EPPLEY, R. W., J. N. ROGERS, AND J. J. MCCARTHY. 1969. Half-saturation constants for uptake of nitrate and ammonium by marine phytoplankton. Limnol. Oceanogr. 14:912-920. EPPLEY, R. W., AND W. H. Thomas. 1969. Comparison of half-saturation constants for growth and nitrate uptake of marine phytoplankton. J. Phycol. 5:375-379. GILMARTIN, M. 1964. The primary production of a British Columbia fjord. J. Fish. Res. Board Can. 21:505-538. Holmes, R. W. 1958. Size fractionation of photosynthesizing phytoplank- ton. In Physical, chemical, and biological oceano- graphic observations obtained on expedition SCOPE in the Eastern Tropical Pacific, November - December 1956, p. 69-71. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 279, 117 p. Holmes, R. W., and G. C. Anderson. 1963. Size fractionation of C'^-labeled natural phyto- plankton communities. In C. H. Oppenheimer (editor). Symposium on marine microbiology, p. 241-250. Charles C. Thomas Publisher, Springfield, 111. HOLM-HANSEN, O., C. J. LORENZEN, R. W. HOLMES, AND J. D. H. Strickland. 1965. Fluorometric determination of chlorophyll. J. Cons. 30:3-15. KAWAMURA, T. 1961. Data on phytoplankton photosynthetic activity. Data Rec. Oceanogr. Obs. Expl. Fish (Hokkaido) 5:142-165. LAWS, E. A. 1975. The importance of respiration losses in controlling the size distribution of marine phytoplankton. Ecology 56:419-426. LORENZEN, C. J. 1968. Carbon/chlorophyll relationships in an upwelling area. Limnol. Oceanogr. 13:202-204. Malone, T. C. 1971a. The relative importance of nannoplankton and net plankton as primary producers in tropical oceanic and neritic phytoplankton communities. Limnol. Oceanogr. 16:633-639. 1971b. Diurnal rhythms in net plankton and nannoplank- ton assimilation ratios. Mar. Biol. (Berl.) 10:285-289. 1791c. The relative importance of nannoplankton and net plankton as primary producers in the California Current system. Fish. Bull., U.S. 69:799-820. MARR, J. 1962. The natural history and geography of the Antarctic krill (Euphausia superba Dana). Discovery Rep. 32:33-463. McAllister, C. D., T R. Parsons, and J. D. H. Strickland. I960. Primary productivity and fertility at station "P" in the north-east Pacific Ocean. J. Cons. 25:240-259. McCarthy, J. J., w. R. Taylor, and M. E. loftus. 1974. Significance of nannoplankton in the Chesapeake Bay estuary and problems associated with the mea- surement of nannoplankton productivity. Mar. Biol. (Berl.) 24:7-16. Parsons, T R. 1972. Size fractionation of primary producers in the Sub- arctic Pacific Ocean. In A. Y. Takenouti (editor), Bio- logical oceanography of the northern North Pacific Ocean, p. 275-278. Idemitsu Shoten, Tokyo. Parsons, T. R., and R. J. LeBrasseur. 1970. The availability of food to different trophic levels in the marine food chain. In J. H. Steele (eidtor). Marine food chains, p. 325-343. Oliver & Boyd, Edinb., Scotl. Parsons, T. R., and M. Takahashi. 1973. Environmental control of phytoplankton cell size. Limnol. Oceanogr. 18:511-515. Reid, J. L., JR., G. I. Roden, and J. G. Wyllie. 1958. Studies of the California Ciu-rent system. Calif. Coop. Oceanic Fish. Invest., Prog. Rep., 1 July 1956 to 1 Jan. 1958, p. 27-56. ROJAS DE MENDIOLA, B. 1969. The food on the Peruvian anchovy. J. Cons. 32:433-434. RYTHER, J. H. 1969. Photosynthesis and fish production in the sea. Science (Wash., D.C.) 166:72-76. RYTHER, J. H., D. W. MENZEL, E. M. HULBURT, C. J. LORENZEN, AND N. CORWIN. 1971. The production and utilization of organic matter in the Peru coastal current. Invest. Pesq. 35:43-59. SAIJO, Y. 1964. Size distribution of photosynthesizing phytoplank- ton in the Indian Ocean. J. Oceanogr. Soc. Jap. 19:187-189. SAIJO, Y., AND K. TAKESUE. 1965. Further studies on the size of photosynthesizing photoplankton in the Indian Ocean. J. Oceanogr. Soc. Jap. 20:264-271. SEMINA, H. J. 1972. The size of phytoplankton cells in the Pacific Ocean. Int. Rev. Gesamten Hydrobiol. 57:177-205. Silver, M., and J. Hansen. 1971a. Plankton studies. In Environmental studies of Monterey Bay and the central California coastal zone. Annual Report July 1971, p. 116-131. Moss landing Mar. Lab., Moss Landing, Calif 1971b. Plankton studies in nearshore Monterey Bay. In Abstract volume, Second National Coastal and Shallow Water Research Conference. Geog. Programs, Off Nav. Res. Skogsberg, T. 1936. Hydrography of Monterey Bay, California. Thermal conditions, 1929-1933. Trans. Am. Philos. Soc, New Ser. 29:1-152. SMAYDA, T. J. 1970. The suspension and sinking of phytoplankton in the sea. Oceanogr. Mar. Biol., Annu. Rev. 8:353-414. SMETHIE, W. M. 1973. Some aspects of the temperature, oxygen and nutrient distributions in Monterey Bay, CaUfomia. Moss Landing Mar, Lab. TP73-1, 97 p. Steeman Nielsen, E. 1952. The use of radio-active carbon (C") for measuring organic production in the sea. J. Cons. 18:117-140. Steeman Nielsen, E., and E. a. Jensen. 1957. Primary oceanic production. The autotrophic pro- 193 FISHERY BULLETIN: VOL. 74, NO. 1 duction of organic matter in the ocean. Galathea Rep. 1:49-136. STOMMEL, H. 1949. Trajectories of small bodies sinking slowly through convection cells. J. Mar. Res. 8:24-29. Strickland, J. D. H., R. w. eppley, and B. Rojas de Mendiola. 1969. Phytoplankton populations, nutrients, and photo- synthesis in Peruvian coastal waters. Inst. Mar. Peru (Callao), Bol. 2:4-45. Strickland, J. D. H., and T. R. Parsons. 1968. A practical handbook of seawater analysis. Fish. Res. Board Can., Bull. 167, 311 p. Subrahmanyan, R., and a. H. Viswanatha Sarma. 1965. Studies on the phytoplankton of the west coast of India. Part IV. Magnitude of the standing crop for 1955-1962, vdth observations on nanoplankton and its significance to fisheries. J. Mar. Biol. Assoc. India 7:406-419. TEIXEIRA, C. 1963. Relative rates of photosynthesis and standing stock of net phytoplankton and nannoplankton. Bol. Inst. Oceanogr, (Sao Paulo) 13:53-60. THORSEN. G. 1950. Reproductive and larval ecology of marine bottom invertebrates. Biol. R«v. (Camb.) 25:1-45. YENTSCH, C. S., AND J. H. RYTHER. 1959. Relative significance of the net phytoplankton and nannoplankton in the waters of Vineyard Sound. J. Cons. 24:231:238. YOSfflDA, K. 1967. Circulation in the eastern tropical oceans with special reference to upwelling and undercurrents. Jap. J. Geophys. 4(2):l-75. 194 ABUNDANCE OF MACROCRUSTACEANS IN A NATURAL MARSH AND A MARSH ALTERED BY DREDGING, BULKHEADING, AND FILLING^ Lee Trent,2 Edward J. Pullen,^ and Raphael Proctor^ ABSTRACT Indices of abundance of macrocrustaceans during March-October 1969 in West Bay, Tex., were determined for day and night and statistically compared between 1) a natural marsh area, 2) upland and bayward canal areas of a housing development, and 3 ) an open bay area. Significance levels of 5% or 1% were used in the statistical comparisons. Catches of brown shrimp, Penaeus aztecus; white shrimp, P. setiferus;h\ue crab, Callinectes sapidus;and pink shrimp, P. duorarum, were significantly greater at night than during the day at one or more stations in the marsh. More grass shrimp, Palaemonetes sp., were caught at night than during the day, but the differences were not statistically significant. Individuals of each species appeared to migrate into the more shallow areas of the marsh at night. At night, brown shrimp and blue crabs were significantly more abundant in the marsh and bayward canal areas than in the upland canal and bay areas, white shrimp were significantly more abundant in the marsh area than in the other three areas, and pink shrimp were significantly more abundant in the marsh than in the upland and bayward canal areas. During the day, brown shrimp were significantly more abundant in the bayward canal area than in the upland canal and bay areas, while pink shrimp were significantly more abundant in the marsh area than in the upland canal area. The generally lower catches of each species in the open bay and upland canal areas when compared with the marsh and bayward canal areas were attributed to: 1 ) permanent loss of intertidal vegetation in the housing development; 2) low abundance of detrital material and benthic macroinvertebrates in the open bay and upland canal areas; and 3) eutrophic conditions in the upland canal area. Development of bayshore property into housing sites by dredging, bulkheading, and filling is oc- curring in many estuaries. When this property is developed, shallow bay and tidal marsh areas are often dredged or filled with spoil, thus changing the environment for marine organisms. Informa- tion is available on some of the environmental changes that are critical, but the effects of these changes on the abundance of macrocrustaceans in Gulf coast estuaries are poorly known. Ecological studies conducted by personnel of the National Marine Fisheries Service in the Jamaica Beach housing development in West Bay, Tex., during 1969 were reported by Trent et al. (1972). That report described the study area and included summary information on phytoplankton produc- tion, oyster production, benthic organisms, sedi- ments, hydrology, and the abundance of macro- crustaceans and fishes. Detailed analyses were reported by Corliss and Trent (1971) on phyto- iContribution No. 398, Gulf Coastal Fisheries Center, Na- tional Marine Fisheries Service, NOAA, Galveston, TX 77550. ^Present Address: Gulf Coastal Fisheries Center Panama City Beach Laboratory, NMFS, NOAA, Panama City, FL 32401. ^Present address: Department of the Army, U.S. Corps of En- gineers, Galveston, TX 77550. Manuscript accepted April 1975. FISHERY BULLETIN: VOL. 74, NO. 1, 1976. plankton production, Moore and Trent (1971) on oyster production, and Gilmore and Trent (1974) on benthic organisms and sediments. Mock (1966) studied the abundance of brown shrimp, Penaeus aztecus, and white shrimp, P. setiferus, in Galveston Bay, Tex., after the bayshore was altered by bulkheading. He stated that catches of brown shrimp were 2.5 times greater, and catches of white shrimp were 14 times greater in a natural habitat than in a bulk- head area. The objectives of this study in the Jamaica Beach area during 1969 were to evaluate relative abundance of selected macrocrustaceans be- tween: 1) day and night; 2) housing development canals, natural marsh areas, and open bay areas; and 3) areas with different concentrations of dis- solved oxygen. STUDY AREA AND METHODS The study area, located in West Bay, included a natural marsh area, an open bay area, and a canal area. The canal area was similar to the natural marsh before it was altered by channeli- zation, bulkheading, and filling (Figure 1). The altered area, which included, prior to alteration, 195 FISHERY BULLETIN: VOL. 74, NO. 1 Figure l. — Study area and sampling locations in the Jamaica Beach area of West Bay, Tex. Winkler method (Carritt and Carpenter 1966). Crustaceans were collected in a trawl that had a mouth opening of 0.6 m by 3.0 m and a stretched mesh of 28.0 mm in the body and 2.5 mm in the cod end. At each station the trawl was towed 200 m at about 2 knots. "Abundance" and "catch" are used synonymously in this report as our index of relative abundance. These terms refer to either the number or average number of animals caught per 200-m tow with the trawl. Data were treated differently than those re- ported by Trent et al. (1972) in that stations 1-5 in the altered area were subclassified into upland canal area (stations 1-3) and bay ward canal area (stations 4, 5); classification of stations 6-9 in the marsh and station 10 in the bay remained the same. The data were treated statistically as follows for the five species caught in greatest abundance (Table 1): differences in catches between day and night were tested with a paired-comparison ^test using individual catches at a station as observa- tions; differences between areas were tested with Tukey's a;-procedure (Steel and Torrie 1960) using the average catch by area, date, and time of day as observations. COMPARISONS OF CATCH BETWEEN DAY AND NIGHT Eight genera and at least 11 species were rep- resented in the catches (Table 1). Four species and members of the genus Palaemonetes were about 45 hectares of emergent marsh vegetation (predominantly Spartina alterniftora) , intertidal mud flats, and subtidal water area was reduced to about 32 hectares of subtidal water area by dredging and filling; water volume (mean low tide level) was increased from about 184,000 m^ to about 394,000 m^. Ten sampling stations were established in the study area. Average water depths (mean low tide level) at stations 1 through 10 were 1.6, 2.6, 2.2, 1.4, 1.3, 0.5, 0.2, 0.4, 0.5, and 1.0 m, respectively. Samples of water and crustaceans were col- lected during the day between 1000 and 1400 h and at night between 2200 and 0200 h at 2-wk intervals from 25 March to 21 October 1969 at each station. Water samples for determining dis- solved oxygen were taken 30 cm above the bot- tom. Oxygen was measured using a modified Table l. — Species or genera and total numbers of crustaceans caught by area during the study. Species Brown shrimp, Penaeus aztecus White shrimp, P. setiferus Grass shnmp, Palaemonetes sp. Blue crab. Callinectes sapidus Pinl< shrimp. Penaeus duorarum Mantis shrimp, Squilla sp. Brokenback shrimp, Trachypenaeus sp. Stone crab, Menippe mercenaha Mud crab, Eurypanopeus sp. Swimming crab, Callinectes similis Pistol shrimp, Alpheus sp. Upland Bayward canal canal Marsh Bay 6.112 16,195 27,063 2,505 1,150 2,738 10,961 172 54 23 8,336 21 181 583 1,149 59 78 80 636 61 2 70 7 7 8 1 9 2 2 1 1 1 196 TRENT ET AL.: ABUNDANCE OF MACROCRUSTACEANS IN MARSHES caught in sufficient numbers for detailed analyses. Brown shrimp was caught in greater numbers during the day in the canal and bay areas and in greater numbers at night in the marsh area ex- cept at station 6 (Table 2). In the canals, day catches were much greater than night catches at the upland canal stations but were only slightly greater than night catches at the bayward canal stations. In the marsh, night catches were sig- nificantly greater than day catches at stations 8 and 9, slightly greater than day catches at station 7, and less than day catches at station 6. White shrimp was caught in greater numbers at night than during the day at all stations except station 5. The differences were statistically sig- nificant at stations 7-9. Grass shrimp, Palaemonetes sp., was caught in greater numbers during the day at two of the canal stations and in greater numbers at night at the remaining stations; the differences were not statistically significant, however. Blue crab, Callinectes sapidus, was caught in greater numbers during the day at the upland canal stations (significant at station 3) and in greater numbers during the night at the remain- ing stations (statistically significant at stations 5-8). Pink shrimp, Penae us duorarum, was caught in greater numbers at night than during the day at all stations except station 6. Differences were statistically significant at stations 5 and 8. COMPARISONS OF CATCH BETWEEN AREAS Statistically significant differences in night catches between areas were observed for four of the five species; day catches were significantly different between areas only for brown and pink shrimps (Table 3). Abundance of brown shrimp during the day was significantly greater in the bayward canal area than in the upland canal and bay areas, whereas at night, brown shrimp were significantly more abundant in the marsh and bayward canal areas than in the other two areas. Catches of white shrimp at night were sig- nificantly greater in the marsh area than in the other three areas. Blue crabs were significantly more abundant at night in the marsh and bay- ward canal than in the bay and upland canal areas. Catches of pink shrimp were significantly greater in the marsh than in the upland canal area during the day and significantly greater in the marsh than in both canal areas at night. CATCH RELATED TO DISSOLVED OXYGEN Mean dissolved oxygen values and mean catch of each species by date and area are shown in Figure 2. Mean oxygen values in the bajrward canal, marsh, and bay areas were above 3.0 ml/ liter throughout the study except on 1 July in the bayward canal and on 23 September in the Table 2. — Comparisons between day and night catches (mean number caught per tow) by species and station (paired comparison f-test with 15 df ). Upland Bayward Bay Species and canal stations canal stations 4 5 Marsh stations station time of day 1 2 3 6 7 8 9 10 Brown shrimp; Day 194.4 27.6 77.3 222 9 298.1 210.7 137.1 212.5 93.5 81.4 Night 47.1 14.5 21.1 203.4 287.8 167.6 177.3 481.9 210.8 75.2 f-value -1.90 -1.02 -1.24 -0.42 -0.18 -1.04 0.98 3.29" 4.43" -0.20 White shrimp: Day 5.8 1.7 12.5 30.1 73.4 76.0 16.1 4.4 8.1 2.9 Night 11.9 3.3 36.7 35.0 32.6 127.6 188.6 178.4 85.8 7.9 f-value 0.79 1.42 1.23 0.75 -0.89 1.18 3.25" 2.93- 2.55- 2.00 Grass shnmp: Day 0.1 1.5 0.0 0.4 0.4 31.8 37.7 22.4 2.2 0.4 Night 1.0 0.4 0.4 0.1 0.6 320.4 43.0 61.0 2.5 0.9 f-value 1.45 -0.94 1.60 -1.23 1.00 1.03 0.21 1.40 0.18 1.09 Blue crab: Day 3.9 1.5 2.4 6.6 7.6 8.8 2.3 8.0 2.6 1.3 Night 1.6 0.8 1.1 7.4 14.8 18.8 10.7 16.2 4.4 2.4 f-value -1.61 -1.74 -2.77* 0.46 2.74- 1.93 2.87* 2.85- 1.04 1.28 Pink shrimp: Day 0.1 0.1 0.1 0.5 0.2 4.8 2.1 1.1 0.2 1.0 Night 1.6 1.9 1.1 1.2 3.2 0.6 4.2 12.2 14.6 2.8 f-value 1,67 1.24 1.52 0.90 2.35- -1.79 1.20 2.12- 2.04 1.93 'Significant at 5% level. "Significant at 1% level. 197 FISHERY BULLETIN: VOL. 74, NO. 1 Table 3. — Comparisons of catches between areas (bay; bay- ward canal, BC; marsh; upland canal, UC) by species and time of day (Tukey's w-procedure with 60 df). Species and time of day Area, mean catch ( ), and significance lines' Brown shrimp: Day Night White shrimp: Day Night Grass shrimp: Day Night Blue crabs: Day Night Pink shrimp: Day Night Bay (81.4) UC (99.7) Marsh (163.4) BC (260.5) UC (27.6) Bay (2.9) Bay (75.2) UC (6.7) BC (245.7) Marsh (26.1) Marsh (259.4) BC (51.7) Bay (7.9) UC (17,3) BC (33.8) Marsh (145.1) Bay (0.4) BC (0.4) UC (0.5) Marsh (23.5) BC (0.3) UC (0.6) Bay (0.9) Marsh (106.7) Bay (1.3) UC (2.8) Marsh (5.4) BC (7.1) UC (1.2) Bay (2.4) BC (11.2) Marsh (12.5) UC (0.1) BC (0.3) Bay (1.0) Marsh (2.0) UC (1.5) BC (2.2) Bay (2.8) Marsh (7.9) 'Any two means not underscored by the same line are significantly different at the 5% level. marsh. In contrast, mean oxygen values observed in the upland canal area remained below 3.0 ml/ liter from 20 May to 12 August and were below 2.0 ml/liter on three occasions. From 20 May to 12 August, about 24% of the individual observations of oxygen values from the upland canal stations were below 1.0 ml/liter, whereas all individual observations from the other three areas were above 1.5 ml/liter. The normal patterns of seasonal abundance were reflected for brown shrimp, white shrimp, and blue crabs by catches in the bayward canal, marsh, and bay areas (Figure 2). Immigration and emigration in Galveston Bay by brown and white shrimps occur during different seasons (Baxter and Renfro 1966; Trent 1967; Pullen and Trent 1969). Brown shrimp postlarvae immigrate in late winter and early spring and most of the juveniles emigrate in late spring and early sum- mer. White shrimp postlarvae immigrate in the summer, and the juveniles emigrate in the fall or early winter depending on water temperature. Blue crabs are abundant throughout the year in Galveston Bay (Chapman 1965). N. 3 E BROWN SHRIMP 20 ■*'''''"'"~'>^f,,„X«t""""*'''''i^ PINK SHRIMP ^Hfe»«,, „.„ ,„„ 25 8 22 6 20 3 17 I IS 29 12 26 9 23 7 21 MAR APR MAY JUNE JULY AUG SEPT OCT Figure 2. — Mean dissolved oxygen values, and mean catch of each species by area and time of year. Patterns of seasonal abundance for grass and pink shrimps are not documented for the Galves- ton Bay system. In Redfish Bay, Tex. (about 150 miles southwest of our study area), Hoese and 198 TRENT ET AL.: ABUNDANCE OF MACROCRUSTACEANS IN MARSHES Jones (1963) caught grass shrimp in greatest numbers during late winter and early spring and pink shrimp in greatest numbers during spring and early fall. Seasonal abundance patterns reflected by catches in this study were similar to those reported in Redfish Bay: for grass shrimp in the marsh area; and for pink shrimp in the bay- ward canal, marsh and bay areas during late summer and early fall. Seasonal abundance of brown shrimp, white shrimp, blue crabs, and pink shrimp deviated from what we expected in the upland canal area. These deviations were probably caused by low dis- solved oxygen. During the period of low dissolved oxygen (below 3.0 ml/liter; from 20 May to 12 August) in the upland canal area, mean catches of brown shrimp dropped and remained below the mean catches of brown shrimp in the other three areas; mean catches of white shrimp and blue crabs remained below mean catches of white shrimp and blue crabs in the bayward canal and marsh areas after 3 June. The abundance of pink shrimp increased on 29 July in all areas except the upland canal area and remained higher than in the upland canal area until 7 September Grass shrimp were not caught in large numbers in any area except the marsh and therefore were not used to evaluate the effects of low dissolved oxygen. DISCUSSION AND SUMMARY Indices of abundance revealed differences in day-night distribution of brown shrimp, white shrimp, blue crabs, and pink shrimp in the study area. Assuming that our catch per unit effort data provided an index which unbiasedly represented density, migration of individuals of all four species into the more shallow areas of the marsh at night best explains these distributional differ- ences. Inherent in the assumption that catch per unit effort unbiasedly estimates density is the equal vulnerability of the animals to capture dur- ing both day and night. Factors which could make this assumption invalid include: 1) burrowing or swimming above the trawl by the animals during one but not the other time period, and 2) avoid- ance of the trawl during the day or night. Re- gardless of the correctness of our assumption, the importance of sampling during both day and night to determine differences in abundance be- tween areas was clearly shown. All five species were more abundant in the marsh than in the upland canal area during both day and night. Brown shrimp, white shrimp, blue crabs, and pink shrimp were more abundant in the bayward canal area than in the upland canal area. The distributional patterns of pink shrimp and blue crabs in this study were similar to those reported by Lindall et al. (1975), who provided data showing that catches of blue crabs and pink shrimp were highest in the bayward portion of an upland canal in a housing development in Tampa Bay, Fla. Four factors probably account for most of the differences observed in abundance of shrimps be- tween areas. Intertidal vegetation was perma- nently eliminated by alteration of the natural area for the housing development. Detrital mate- rials and abundance of benthic macroinverte- brates were lowest in the open bay area, low in the upland canal area, and highest in the bay- ward canal and marsh areas (Gilmore and Trent 1974). Eutrophic conditions observed represent the fourth factor. Eutrophic conditions, indicated by the observed low values of dissolved oxygen in the upland ca- nals of the housing development during the summer, probably account for the comparatively low catches of brown shrimp, white shrimp, pink shrimp, and blue crabs during that period. Further evidence of eutrophication in this area was provided by studies on: the American oyster, Crassostrea virginica, in which setting, survival, and growth rates were less in the upland canal area than in the marsh area (Moore and Trent 1971); phytoplankton in which production was higher in the upland canal area than in the marsh or bay areas (Corliss and Trent 1971); and benthic macroinvertebrates in which the abun- dance of the organisms declined drastically dur- ing the summer months in the upland canal area (Gilmore and Trent 1974). Symptoms of eutrophic conditions in the upland canals of the housing development include inadequate water exchange and high nutrient levels. These factors were dis- cussed in detail by Moore and Trent (1971). Alteration of estuaries by dredging and filling for housing developments and boat basins results in an environment highly susceptible to recur- rent low dissolved oxygen levels. This probelm of low oxygen has been shown also in Forida (Taylor and Saloman 1968; Lindall et al. 1973) and California (Reish 1961). Stresses resulting from low dissolved oxygen reduce the abundance of crustaceans and other animals in the stressed 199 FISHERY BULLETIN: VOL. 74, NO. 1 areas and may produce mass mortalities. Flow dynamics and sedimentation patterns should be carefully evaluated when new developments in estuaries are being considered in order to prevent areas of stagnant water from being created. ACKNOWLEDGMENTS Sincere appreciation is extended to Edwin A. Joyce, Jr., and his staff, Florida Department of Natural Resources, for reviewing this manuscript and for helpful suggestions. LITERATURE CITED Baxter, K. N., and W. C. Renfro. 1966. Seasonal occurrence and size distribution of postlar- val brown and white shrimp near Galveston, Texas, with notes on species identification. U.S. Fish Wildl. Serv., Fish. Bull. 66:149-158. CARRITT, D. E., AND J. H. CARPENTER. 1966. Comparison and evaluation of currently employed modifications of the Winkler method for determining dis- solved oxygen in seawater; a NASCO report. J. Mar. Res. 24:286-318. CHAPMAN, C. R. 1965. Estuarine program. In Biological Laboratory, Gal- veston, Tex. fishery research for the year ending June 30, 1964, p. 60-75. U.S. Fish Wildl. Serv., Circ. 230. Corliss, J., and L. Trent. 1971. Comparison of phytoplankton production between natural and altered areas in West Bay, Texas. Fish. Bull., U.S. 69:829-832. Gilmore, G., and L. Trent. 1974. Abundance of benthic macroinvertebrates in natural and altered estuarine areas. U.S. Dep. Commer., NCAA Tech. Rep. NMFS SSRF-677, 13 p. HOESE, H. D., AND R. S. JONES. 1963. Seasonality of larger animals in a Texas turtle grass community. Publ. Inst. Mar. Sci., Univ. Tex. 9:37-47. LINDALL, W. N., JR., W. A. FABLE, JR., AND L. A. COLLINS. 1975. Additional studies of the fishes, macroinvertebrates, and hydrological conditions of upland canals in Tampa Bay, Florida. Fish Bull., U.S. 73:81-85. LINDALL, W. N., Jr., J. R. Hall, and C. H. Saloman. 1973. Fishes, macroinvertebrates, and hydrological condi- tions of upland canals in Tampa Bay, Florida. Fish. Bull., U.S. 71:155-163. Mock, C. R. 1966. Natural and altered estuarine habitats of penaeid shrimp. Proc. Gulf Caribb. Fish. Inst., 19th Annu. Sess., p. 86-98. Moore, D., and L. Trent. 1971. Setting, growth, and mortality of Crassostrea vir- ginica in a natural miarsh and a marsh altered by a hous- ing development. Proc. Natl. Shellfish. Assoc. 61:51-58. Pullen, E. J., AND W. L. Trent. 1969. White shrimp emigration in relation to size, sex, temperature and salinity. FAQ Fish. Rep. 57:1001-1014. REISH, D. J. 1961. A study of benthic fauna in a recently constructed boat harbor in southern California. Ecology 42:84-91. Steel, r. g. d., and j. r. torrie. I960. Principles and procedures of statistics, with special reference to the biological sciences. McGraw-Hill Book Co.,N.Y., 481 p. Taylor, J. L., and C. H. Saloman. 1968. Some effects of hydraulic dredging and coastal de- velopment in Boca Ciega Bay, Florida. U.S. Fish Wildl. Serv., Fish. Bull. 67:213-242. TRENT, W. L. 1967. Size of brown shrimp and time of emigration from the Galveston Bay system, Texas. Proc. Gulf Caribb. Fish. Inst., 19th Annu. Sess., p. 7-16. Trent, w. l., e. j. pullen, and d. moore. 1972. Waterfront housing developments: Their effect on the ecology of a Texas estuarine area. In M. Ruivo (editor). Marine pollution and sea life, p. 411-417. Fishing News (Books) Ltd., West Byfleet, Surrey. 200 NOTES MORTALITIES AND EPIBIOTIC FOULING OF EGGS FROM WILD POPULATIONS OF THE DUNGENESS CRAB, CANCER MAGISTER''' Cultured crustaceans have been found to be sus- ceptible to fouling by a variety of epibionts. Nilson et al. (1975) recently described mortalities attrib- uted to epibiotic fouling in the eggs and larvae of the American lobster, Homarus americanus, the larvae of the prawn, Pandalus platyceros, and lar- vae of the Dungeness crab, Cancer magister Dana. This same type of fouling has also been found on juveniles of Penaeid shrimp, where it causes death in rearing ponds with low oxygen content by in- habiting the gill filaments and suffocating the animal (Johnson et al. 1974; Lightner et al. 1975). The organisms most commonly encountered have been filamentous bacteria and algae. Work on the larval cultivation of the Dungeness crab at the Bodega Marine Laboratory, Bodega Bay, Calif., revealed heavy fouling on the eggs of oviposited female crabs held in rearing tanks. Further investigation showed that the condition also existed on eggs of crabs obtained from local fishermen. Egg masses with extensive fouling also showed a large number of empty egg cases, al- though eyespot development on the remaining embryos showed the time until hatching to be dis- tant. Similar fouling of the eggs of wild caught Atlantic blue crabs, Callinectes sapidus, has been observed and well documented (Sandoz et al. 1944; Rogers-Talbert 1948). With Callinectes, however, the predominant fouling organism appears to be the fungus Lagenidium callinecti. These observations of fouling and mortality in the natural population suggest a possible explana- tion for the decline in Dungeness crab catches recorded in the San Francisco Bay region since 1960 (Biostatistical Section 1961, 1963, 1964, 1965; Greenhood and Mackett 1965, 1967; Heimann and Frey 1968a,b; Heimann and Carlisle 1970; Pinkas 1970; Bell 1971; Oliphant 1973). In order to inves- tigate this possibility, a distributional study was undertaken, comparing mortalities and epibiotic fouling of crab eggs from various locations along the coast of northern California. Materials and Methods Egg samples of C. magister were obtained from fishermen along the northern California coast during the period from 27 November 1974 to 30 January 1975. A total of 105 samples of eggs from individual crabs were obtained from six regions which included the following localities (Figure 1): region I — Pacifica (4 samples); region II — Drake's Bay (18 samples); region III — Point Reyes (39 samples); region IV — Bodega Bay, Russian River, and Gualala (10 samples); region V — Fort Bragg (20 samples); region VI — Eureka (14 sam- ples). In the field, a portion of eggs were removed from the Dungeness crab egg masses and placed in vials Froncisco Boy 'This work was supported by California State Legislature Aquaculture funds. ^This work was done at the University of California, Bodega Marine Laboratory at Bodega Bay, CA 94923. Figure l.— The coast of northern California showing the Dungeness crab collection sites: I - Pacifica; II - Drake's Bay; III - Point Reyes; IV - Bodega Bay, Russian River, and Gualala; V - Fort Bragg; VI - Eureka. 201 containing 10% Formalin^ in seawater. The sam- ple size was variable — all exceeded 100 eggs, usu- ally several hundred. The vials were then shipped to the laboratory for examination with the aid of a dissecting microscope. The epibiotic organisms were clearly visible using transmitted light for illumination (Fisher et al. 1975). Closer examina- tion of the egg cases was carried out with a phase microscope to aid in the characterization of the fouling organisms. Portions of the samples were categorized as to the comparative developmental state of the eggs, extent of epibiotic fouling, and egg mortality by the following methods: 1. The following observations of the eyespots which develop as the embryos develop were used to give a comparative estimate of the time the eggs had been carried externally on the female: Dl. No visible eyespot. D2. Emerging eyespot. D3. Full eyespot. Any samples which showed evidence of hatch- ing were not used. Occasionally, there was varia- tion in the degree of development of the eggs from a single sample, in which case the eggs that had developed furthest were used for observation. 2. The extent of epibiotic fouling was deter- mined by the following observations of the exter- nal egg membrane: Fl. None — no evidence of epibionts at 100 x (Figure 2A). F2. Light — occasional short filaments. F3. Moderate — the majority of the surface covered with short filaments and occa- sional long filaments (Figure 2B). F4. Heavy — the surface extensively covered with short and long filaments (Figure 2C). F5. Very heavy — the surface extensively covered with short filaments, long fila- ments, and detrital material. 3. The number of empty egg cases was used as an estimate of mortality. Ml. <10% mortality. M2. 10-25% mortality. ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. M3. 26-50% mortality. M4. 51-75% mortality. M5. 76-100% mortality. Only empty egg cases (Figure 3) were consid- ered mortalities. Other abnormal conditions, such as discolored eggs which might have eventu- ally led to mortalities, were observed but not used in the estimates. All developmental stage D3 samples were checked for emerging embryos to ensure that the empty egg cases were not due to hatching. In addition to the field samples, seven ovigerous females from the Point Reyes area were examined before being placed into flow-through seawater tanks at the laboratory. After 25 days the eggs were reexamined to determine the progress of the infestation. In addition, one complete egg mass from an ovigerous female was examined to deter- mine the homogeneity of the fouling condition throughout the egg mass. Results Observation of eyespot development placed 10.5% of the samples into category Dl, 35.2% into D2, and 54.3% into D3. Fouling was observed in all developmental categories, but mortalities were generally higher in the more developed eggs. The histograms presented in Figure 4 show the percent of samples from each region placed in each mortal- ity category (M1-M5) and fouling category (F1-F5) after combining the developmental categories. The eggs of the seven females held in the labo- ratory for 25 days showed an average increase in their development, fouling and mortality of one level in each category. The greatest observed change was on an egg mass in developmental stage 2 which originally showed light fouling (cat- egory 2) and were in mortality category M2. After 25 days it was in developmental stage 3 and showed very heavy fouling (category 5) and had advanced to mortality category, M5. Another showed no increase in fouling as it matured from developmental stages 1 to 3, but the egg mortality category advanced from Ml to M3. Examination of the entire egg mass of one specimen showed that the extent of the fouling was variable and concentrated mostly on the periphery of the mass and on the inner eggs near the fold of the abdomen. This raises the possibility of sampling error; however, it would probably be insignificant since the field samples came primar- ily from the exterior of the egg masses. 202 A ^ ^ n** 4 ■'■f ^•^%lfc^ ^. \ •'4:^ -'■ r' Figure 2. — Dungeness crab egg samples showing (A) no epibio- tic folding on the egg membrane (78x), (B) moderate epibiotic fouling on the egg membrane (96x ), and (C) very heavy epibiotic fouling on the egg membrane (57 x). The epibiotic fouling organisms found were similar to those noted on other crustaceans by Nilson et al. ( 1975). Particularly prominent were the long filamentous cyanophytes which resem- bled Oscillatoria and bacterial filaments similar to Leucothrix. In heavily fouled samples stalked pro- tozoans (vorticellids) were also observed. These and the filamentous organisms trapped detrital material, which added to the overall contamina- tion of the eggs. Fouling on the egg stalk was often more extensive than fouling on the egg membrane proper. Empty egg cases also showed heavier foul- ing than those containing embryos. In many cases where fouling was observed, worms were found, and the population of worms was generally larger on egg samples with heavier fouling. The worms were identified as the nemertean egg predator Carcinonemertes epialti as described by Kuris (1973). 203 Figure 3. — Dungeness crab egg sample showing empty egg cases representing egg mortalities alongside viable eggs from the same egg mass (24. 5x). Discussion Various workers have attributed mortalities (Johnson et al. 1974; Lightner et al. 1975; Nilson et al. 1975; Fisher et al. 1975) in cultured crusta- ceans to epibiotic fouling. These reports suggest that death may be caused either by mechanical interference in larval molting or restriction of gaseous exchange across the egg or gill mem- brane. The fouling organisms may also consume a great deal of the available oxygen from the envi- ronment. The dramatic effect of this condition may be seen in Figure 5 where the moderately fouled egg case is entirely intact, yet the embryo is at- rophied and nonviable. Infestation with fouling organisms presumably does not begin until the eggs are oviposited. Al- though heavy fouling may occur, few mortalities are observed in the early developmental periods. Fouling on the eggs held in rearing tanks pro- gressed as the eggs developed. The progression was an increase in the number or filament length of any one type of the organisms or the addition of other types of organisms. By the second and third developmental categories, mortalities were regu- larly encountered where fouling occurred. The samples obtained from regions II and III showed the heaviest epibiotic fouling, as well as the highest levels of mortality. In comparison, re- gion V showed the least extensive fouling and the fewest mortalities. This suggests that there is a relationship between epibiotic fouling and egg mortality. Closer examination of the histograms in Figure 4 reveals a possible trend of mortalities and foul- ing progressively decreasing from region II to re- gion V. Although the number of samples obtained from region I may not be conclusive evidence, they suggest that the trend may not continue south of San Francisco Bay. The region VI data show a slight reversal of the trend although mortalities and fouling are still comparatively low. The mortalities observed in regions II and III are particularly relevant when the coastal crab catch over the last 25 yr is considered. Figure 6 shows a general coast-wide decline in Dungeness crab catch commencing in 1958. In 1965, the northern fishery areas began a strong recovery, whereas the San Francisco area remained at low level. During this decline, the catch of the San Francisco fishery dropped from 8y2 million pounds to less than 1 million pounds where it has re- mained. Several studies have investigated the potential impact of overfishing on the Dungeness crab popu- lation. Poole (1962) and Cordier (1966) showed 204 Ml I 2 ' 3 ' 3 I 5 ' Morfolity levels » Fouling levels > Figure 4. — Histograms representing the percent of Dungeness crab samples from each region found in the mortality and fouling categories. The arrows represent increasing mortalities and in- creasing fouling. Sample sizes from each region are shown. that 99% and 98%, respectively, of the adult female population had been inseminated, indicat- ing that the fishing industry (which only legally catches males greater than 6^^ inches across the carapace) is not significantly reducing the repro- ductive capabilities of the crab population. Also, tagging studies have shown that an estimated 90 to 100% of the legal-size males in fishing areas of the California coast have been caught each year since 1929 (Pacific Marine Fisheries Commission 1965). Cleaver (1949) and Peterson (1973) stated that the fishing pressure has been similar in Washington and Oregon. It therefore appears that fisheries along the coast are capable of maintain- ing production despite the virtually maximum fishing pressures. Poole and Gotshall ( 1965) con- cluded that the fishing regulations at that time were sufficient to protect the crab from depletion through overfishing. Physical factors may be responsible for periodic fluctuations in crab abundance. The Pacific Marine Fisheries Commission (1965) suggested that shifting currents played a role in these fluc- FlGURE 5. — A single Dungeness crab egg showing an intact membrane, an atrophied and nonviable embryo (168 x). - Eureka- Ft Bragg -Oregon San Frartciico Crab 5«oson Ycori (1948-19/21 Figure 6. — A graph comparing the Dungeness crab catches reported from 1948 to 1972 in three areas. Note that the San Francisco crab catch did not increase from the 1961-62 level. tuations by disturbing larval settlement. Lough (1974) found a correlation between rainfall during salinity-sensitive larval stages and crab catch 4 yr later when those larvae were to enter the fishery. Peterson (1973) and Botsford and Wickham (1975) have found a positive correlation between upwel- ling intensity and crab catch. 205 Our observations indicate that disease is a fac- tor to be considered in evaluating the decline of the San Francisco area crab population. The re- productive capacity of the population must be af- fected by this epibiotic fouling condition especially if it can also infest the larval stages as indicated by the studies on other crustaceans (Fisher et al. 1975). The variety of fouling organisms and the geo- graphical trends observed in this disease situation suggest a complex relationship with external en- vironmental factors. In view of the saprophytic nature of the fouling organisms, their major source of nutrients is probably external. As such, the growth of the contaminants are affected by the nutrient level in the seawater. It appears that the external factors involved may originate in the San Francisco Bay effluent. This is suggested by the decreasing trend of mor- talities and fouling heading north from this area, presumably reflecting the dilution of the effluent waters. The normal water currents in this area flow in a southerly direction; however, during the period from November through February, the pre- vailing inshore flow is the northerly Davidson Current (Reid et al. 1958). During the egg-bearing season, the effluent from San Francisco Bay is carried northward. The observations of this study were limited by the collection of samples during only the 1974-75 crab season. Because of the potential relationship of these findings to a valuable natural resource, we felt that it was important to communicate the available information. It is clear that further stud- ies during the next season will enhance our under- standing of the situation. Acknowledgments We thank Robert Shleser for his support under California State Legislature funds and in the preparation of the manuscript; Edgar Nilson for his assistance and suggestions; Louis Cavellini, Earl Carpenter, Tom Burke, and the crab fisher- men of northern California who obtained egg samples. Literature Cited Bell, R. R. 1971. California marine fish landings 1970. Calif. Fish Game, Fish Bull. 154, 50 p. BIOSTATISTICAL SECTION, MARINE RESOURCES OPERATIONS. 1961. The marine catch of California for the year 1960. Calif Fish Game, Fish Bull. 117, 45 p. 1963. The California marine fish catch for 1961. Calif. Fish Game, Fish Bull. 121: 1-47. 1964. The California marine fish catch for 1962. Calif. Fish Game, Fish Bull. 125, 45 p. 1965. The California marine fish catch for 1963. Calif. Fish Game, Fish Bull. 129, 45 p. BOTSFORD, L. W., AND D. E. WiCKHAM. 1975. Correlation of upwelling index and Dungeness crab catch. Fish. Bull., U.S. 73:901-907. Cleaver, F. C. 1949. Preliminary results of the coastal crab (Cancer magister) investigation. Wash. Dep. Fish., Biol. Rep. 49A:47-82. CORDIER, P. 1966. Cruise report 66-N-9 crab. Calif. Dep. Fish Game, Mar. Resour. Oper. FISHER, W. S., E. H. NiLSON, AND R. A. SHLESER. 1975. Diagnostic procedures for diseases found in egg, lar- val, and juvenile cultured American lobsters [Homarus americanus). Proc. 6th Annu. Workshop World Maricult. Soc, Seattle, Wash., 1975. Greenhood, E. C, and D. J. MACKETT. 1965. The California marine fish catch for 1964. Calif. Fish Game, Fish Bull. 132, 45 p. 1967. The California marine fish catch for 1965. Calif. Fish Game, Fish Bull. 135:1-42. Heimann, R. F. G., and H. W. Frey. 1968a. The California marine fish catch for 1966. Calif. Fish Game, Fish Bull. 138:1-48. 1968b. The California marine fish catch for 1967. Calif Fish Game, Fish Bull. 144, 47 p. Heimann, R. F. G., and J. G. Carlisle, jr. 1970. The California marine fish catch for 1968 and histori- cal review 1916-1968. Calif. Fish Game, Fish Bull. 149, 70 p. JOHNSON, S. K., J. C. Parker, and H. Holcomb. 1974. Control of Zoothamnium sp. on penaeid shrimp. Proc. 4th Annu. Workshop World Maricult. Soc, Monterrey, Mexico, 1973. KURIS, A. 1973. Population interactions between a shore crab and two symbionts. Ph.D. Thesis, Univ. California, Berkeley. LIGHTNER. D. v., C. T. FONTAINE, AND K. HANKS. 1975. Some forms of gill disease in penaeid shrimp. Proc. 6th Annu. Workshop World Maricult. Soc, Seattle, Wash., 1975. LOUGH, R. G. 1975. Dynamics of crab larvae (Anomura, Brachyura) off the central Oregon coast, 1969-1971. Ph.D. Thesis, Ore- gon State Univ., Corvallis, 229 p. NILSON, E. H., W. S. FISHER, AND R. A. SHLESER. 1975. Filamentous infestations observed on eggs and lar- vae of cultured crustaceans. Proc. 6th Annu. Workshop World Maricult. Soc, Seattle, Wash., 1975. OLIPHANT, M. S. 1973. California marine fish landings for 1971. Calif Fish Game, Fish Bull. 159, 49 p. Pacific Marine fisheries Commission. 1965. Discussion that followed the report on Dungeness crabs. 16th and 17th Annu. Rep. Pac. Mar. Fish. Comm., p. 38-39. Peterson, W. T. 1973. Upwelling indices and annual catches of Dungeness crab, Cancer magister, along the west coast of the United States. Fish. Bull., U.S. 71:902-910. 206 PINKAS, L. 1970. The California marine fish catch for 1969. Calif Fish Game, Fish Bull. 153, 47 p. POOLE, R. 1962. Cruise report 62-N-2g, h, i and 1 crab. Calif. Dep. Fish Game, Mar. Resour. Oper. POOLE, R., AND D. GOTSHALL. 1965. Regulations and the market crab fishery. Outdoor Calif. 26(9):7-8. REID, J. L., JR., G. I. RODEN, AND J. G. WYLLIE. 1958. Studies of the California Current System. Calif. Coop. Oceanic Fish. Invest. Rep. 1 July 1956 - 1 January 1958, p. 27-56. ROGERS-TALBERT, R. 1948. The fungus Lagenidium callinectes Couch (1942) on eggs of the blue crab in Chesapeake Bay. Biol. Bull. (Woods Hole) 94:214-228. Sandoz, M. D., R. Rogers, and C. L. Newcombe. 1944 Fungus infection of eggs of the blue crab Callinectes sapidus Rathbun. Science (Wash., D.C.) 99:124-125. WILLIAM S. FISHER Department of Food Science and Technology University of California Davis, CA 95616 Daniel E. Wickham Department of Zoology University of California Berkeley, CA 94620 SECOND RECORD OF BLACK SKIPJACK, EUTHYNNUS LINEATUS, FROM THE HAWAIIAN ISLANDS Matsumoto and Kang (1967) reported the first capture of the black skipiack, Euthynnus lineatus Kishinouye, in the Hawaiian Islands. Recently (14 July 1975), a second black skipjack was taken in these waters by a Hawaiian pole-and-line skip- jack tuna fishing vessel, the Mar/m, skippered by Walter Asari. The fish was noticed by a fish re- ceiver at Hawaiian Tuna Packers, Richard How- ell, who contacted Robert T. B. Iversen, South- west Region Representative stationed at the Southwest Fisheries Center Honolulu Laboratory. Iversen brought the fish to me for identification. The specimen, 454 mm fork length, and weigh- ing 1.53 kg, was caught from a school of small skipjack tuna, Katsuwonus pelamis, at the ex- treme tip of Penguin Banks, about 40 km south of the eastern end of Oahu. The specimen is de- posited in the U.S. National Museum collection (USNM 214683). Measurements in millimeters taken according to the methods described by Grodsil and Byers (1944) are as follows: Fork length - 454; head length - 126; 1st dorsal insertion - 144; 2d dorsal insertion - 271; anal fin insertion - 306; ventral fin insertion - 144; greatest body depth - 112; greatest body width - 73; dorsal-ventral distance - 108; dorsal-anal distance - 188; ventral insertion to vent - 160; length 1st dorsal base - 130; length 2d dorsal base - 29; length anal base - 25; length pectoral - 70; height 1st dorsal - 61; height 2d dorsal - 28; height anal - 28; diameter of iris - 19; maxillary length - 50; snout to posterior margin of eye - 54. Counts: 1st dorsal spines - 14, plus 1 imbedded; 2d dorsal rays - 12; dorsal finlets - 8; anal rays - 12; anal finlets - 7; pectoral rays - 26; gill rakers - left side 9 -H 1 -h 24 = 34, right side 9 + 1 + 25 = 35. The external characters agree with that of the previous capture (Matsumoto and Kang 1967) and with Godsil's (1954) description of the species. Five black unbranched stripes run paral- lel to the longitudinal axis of the body on the back fi'om the corselet to the caudal fin, and five or six faint unbranched stripes run horizontally on the belly. Two black thoracic spots are located on each side at the indentation of the corselet near the ventral margin of the body. The vertebral count is 20 + 17 = 37. As in the previous capture, four large protuberances are present on the 31st vertebra, a characteristic of this species (Godsil 1954). Although this is only the second specimen re- corded, an interview with the skipper of the ves- sel disclosed that fish similar to this are often caught but are not reported. The question posed in 1967 as to whether this is a chance migrant from the eastern Pacific Ocean still stands. Literature Cited GODSIL, H. C. 1954. A descriptive study of certain tuna-like fishes. Calif. Dep. Fish Game, Fish Bull. 97, 185 p. Godsil, H. C, and R. D. Byers. 1944. A systematic study of the Pacific tunas. Calif Dep. Fish Game, Fish Bull. 60, 131 p. Matsumoto, W. M., and T. Kang. 1967. The first record of black skipjack, Euthynnus lineatus, from the Hawaiian Islands. Copeia 1967:837- 838. Walter M. matsumoto Southwest Fisheries Center Honolulu Laboratory National Marine Fisheries Service, NOAA Honolulu, HI 96812 207 OPTICAL MALFORMATIONS INDUCED BY INSECTICIDES IN EMBRYOS OF THE ATLANTIC SILVERSIDE, MENIDIA MENIDIA Since the banning of DDT from use in the United States, other insecticides such as malathion, parathion, and Sevin^ (carbaryl) have come into greater use. Though not persistent like DDT, these insecticides, Uke DDT, find their way into aquatic ecosystems and thus into the spawning grounds of aquatic organisms. Various insec- ticides have been shown to cause developmental abnormalities. Malathion, for example, has been shown to cause skeletal malformations in birds (McLaughlin et al. 1963; Walker 1967; Greenberg and LaHam 1969), mammals (Tanimura et al. 1967), and reptiles (Mitchell and Yntema 1973). The experiments described herein were de- signed to study the effects of DDT, malathion, and Sevin on the development of the Atlantic silver- side, Menidia menidia. Since previous studies had all indicated that sensitivity decreases with em- bryonic age, we initiated our treatment early in development. Materials and Methods Adult M. menidia, from the vicinity of Mon- tauk, N.Y., were collected by a seine during June and July. Eggs and sperm were obtained by stripping the fish, as described by Costello et al. (1957:228-233). The fertilized eggs were sep- arated into small clumps and, after being washed, were placed randomly in glass finger bowls in 100 ml of Millipore-filtered seawater (salinity 30%) and incubated at 20°C. The insecticides malath- ion (95% analytical reagent, Supelco Inc., Belle- fonte, Pa.), DDT (p,p'-DDT, 729c technical grade, Montrose Chemical Co., Torrance, Calif., recrys- tallized from ethanol to yield 98% p,p'-DDT), and Sevin (99.2% carbaryl. Union Carbide Corp., New York, N.Y.) were introduced as acetone solutions into experimental dishes during either early cleavage (2-4 cell stage) or late cleavage (about 100 cells— see Costello et al. 1957, fig. 104), at concentrations of 10 to 500 parts per billion (ppb). Control dishes received an equivalent amount of acetone (10 /xl). The solutions were not changed; thus we were studying the effect of a single appli- cation of the chemicals (the concentration of 'Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. which undoubtedly decreased over time due to adsorption). Development was followed with ref- erence to the descriptions of Costello et al. (1957). At appropriate times, eggs were examined to see the percentage which had successfully completed gastrulation and, later, the percentage which had successfully initiated heartbeat. In the first two experiments hatching rates were noted and only the newly hatched fry were examined for mal- formations. Since they appeared normal, in the subsequent experiments embryos were examined for malformations with considerably more suc- cess. Some embryos were preserved in glutaral- dehyde, dechorionated, sectioned, and stained with hematoxylin and eosin. A repeat experiment was performed in the fol- lowing summer using the same procedures. Results In the first experiment, eggs were treated at the late cleavage stage with malathion at 10 and 100 ppb and Sevin at 25 and 100 ppb. There were over 200 eggs in each dish. Percents of successful axis formation and heartbeat initiation were lower than controls in most treated groups (Table 1) but did not always show a dose-related effect. Hatching commenced 14 days after fertilization and continued for 6 days, at which time the ex- periment was terminated. No difference was noted in hatching times in the various groups and no abnormalities were observed in the fry, al- though some dead ones were seen in each group. In the second experiment, eggs at the 2-4 cell stage were exposed to DDT at 25 and 100 ppb and to malathion at 10 and 100 ppb. There were again about 200 eggs in each dish. As in the previous experiment (Table 1) treated groups had lower rates of axis formation and of heartbeat initiation than controls. Hatching commenced 14 days after fertilization and continued for 6 days, at which time the experiment was terminated. No differ- ence was noted in hatching times in the various groups and no abnormalities were noted in the fry, although, as before, some dead ones were noted in each group. In the third experiment, eggs at the late cleav- age stage were exposed to DDT at 10, 25, and 100 ppb, malathion at 10, 100, and 500 ppb, and Sevin at 25, 100, and 500 ppb. There were about 50 eggs in each dish. When checked for axis formation and heartbeat initiation, the treated eggs were again lower than controls. Embryos were care- 208 Table l. — Insecticide effects on percentage of axis formation, heartbeat, optic abnormalities, and hatching. Concentrations in parts per billion (ppb). Control 10 DDT (ppb) 25 100 Malathi ion (ppb) Sevin (ppb) Item 10 25 100 500 1,000 2,500 10 25 100 500 Experiment 1 (late cleavage): Axis formation 54 23 19 36 48 Heartbeat 46 22 13 35 48 Hatch 21 19 6 21 27 Experiment 2 (2-4 cell stage): Axis formation 41 27 9 28 21 Heartbeat 41 25 5 23 21 Hatch 28 14 2 11 7 Experiment 3 (late cleavage): Axis formation 17 12 13 10 17 13 15 16 7 13 Heartbeat 17 12 11 6 13 9 13 16 7 9 Optic anomalies 50 50 60 40 60 33 40 57 25 Experiment 4 (late cleavage): Axis formation 53 45 30 21 37 13 10 30 24 Heartbeat 53 43 30 20 34 13 6 30 20 Optic anomalies 1 11 9 15 9 22 30 17 11 Experiment 5 (late cleavage): Axis formation 96 81 82 29 83 65 65 50 50 Heartbeat 96 70 82 6 83 32 62 50 50 Optic anomalies 1 50 4 12 6 12 30 fully examined for developmental abnormalities, and various optic malformations were discovered in the insecticide-treated embryos. These took the form of unilateral and bilateral microphthalmia (reduced size of eyes), unilateral and bilateral anophthalmia (absence of eyes), and cyclopia (a single median eye) (Figure 1). Severely retarded embryos were also noted. Percentages of those with successful axis formation which showed op- tical abnormalities were quite high in all treated groups, while none were observed in the control group. Abnormal embryos were fixed prior to hatching. (It was assumed that they would die prior to hatching since no abnormal fry had been found in the previous experiments.) At hatching, which commenced 15 days after fertilization and continued for 7 days, one fish with scoliosis was noted in 10 ppb malathion. In the fourth experiment, eggs were again ex- posed at the late cleavage stage to DDT at 10, 25, and 100 ppb, malathion at 10, 25, and 100 ppb, and Sevin at 10 and 25 ppb. There were about 200 eggs in each dish. When checked for axis forma- tion and heartbeat, treated groups were lower t |k B FIGURE 1.— Photomicrographs of whole, fixed, 2-wk-o\d Menidia menidia embryos at approximately 20 x. A is a control embryo, while B is a 10 ppb Sevin-treated embryo with unilateral anophthalmia (the site of the undeveloped eye is marked by X), and C is a cyclopic embryo from a 10 ppb malathion-treated batch (transmitted light illuminates the single lens at L). 209 than controls. At this time, and for several days after, abnormal embryos were noted. These in- cluded the severely retarded embryos and the optical abnormalities noted earlier. Only one control embryo showed slight microphthalmia. Hatching commenced after 11 days and continued for 9 days, at which time the experiment was terminated. After hatching, lordotic fry were seen in the 10 ppb malathion, 10 ppb Sevin, and 25 ppb DDT groups. These skeletal abnormalities were quite rare, however. Eye diameters of hatched fry were measured with an ocular micrometer to see if there were slight reductions in optic size in the apparently normal specimens, but no difference between ex- perimental and control fry was seen. The fifth experiment was performed the follow- ing summer using about 100 eggs per dish. Eggs were exposed at late cleavage to DDT at 10, 25, and 100 ppb, Sevin at 10, 25, and 100 ppb, and malathion at 1 and 2.5 ppm. Treated groups were again lower than controls in rate of axis forma- tion and heartbeat initiation. Abnormal embryos were seen in most treated groups (Table 1) and all embryos which exhibited optic malformations also showed retardation, stunting of growth, sparse body pigment, and abnormal cardiac de- velopment in which the heart remained a very thin, feebly beating tube without differentiation of the chambers. There were also embryos with this syndrome in which the eyes appeared nor- mal. Hatching commenced after 12 days, and sev- eral fry with scoliosis were seen in the mala- thion dishes. Discussion The three insecticides reduced survival of Menidia embryos, although this reduction was not always correlated with the dose and varied in different batches of eggs. The main embryotoxic effect was at early stages, preventing successful axis formation. Of those which formed axes, most went on to establishment of heartbeat. Notable optic malformations were observed in embryos exposed to DDT, malathion, and Sevin. These three insecticides are quite different from each other chemically, and the fact that they all produced similar malformations may indicate that this species has a propensity toward this type of malformation and various agents can in- voke them. This propensity is supported by the presence of one control embryo with slight mi- crophthalmia in one eye. McEwan et al. (1949) likewise concluded that the jewelfish, Hemi- chromis bimaculata, had a tendency to vary ab- normally in certain directions and that an ab- normal environment accentuated this tendency. The most common optic abnormalities seen in our fish were unilateral anophthalmia and microph- thalmia. True cases of cyclopia were rare, though several embryos showed partial convergence of the eye cups, with optic cups directed somewhat ventrally rather than laterally. Stockard (1907) produced cyclopia in Fundulus embryos by treatment with MgCl2. In another study (1910) he produced cyclopean, anophthal- mic, and monophthalmic Fundulus embryos after treatment with alcohol, results similar to those in the present study. Histological examination of our material re- vealed a case in which the optic cup had partly formed, but appeared to be facing inward rather than outward and had lost its connection to the brain. No lens was present in this specimen. Smithberg (1962) found that tolbutamide caused eye malformations in the medaka, Oryzias latipes. However, these malformations involved degeneration of the eye cup after the lens had been formed, and lenses were present in all the abnormal embryos. These malformations were accompanied by circulatory defects, which he considered responsible for the eye defects. Retardation of development was seen by Battle and Hisaoka ( 1952) in their studies of effects of ethyl carbamate (urethan) on embryos of the zebrafish, Brachydanio rerio. Some of their em- bryos also exhibited optical malformations in- cluding anophthalmia, microphthalmia, and cyclopia. In Hisaoka's subsequent study (1958) of 2-acetylaminofluorene on zebrafish embryos, microphthalmia was one abnormality produced by this carcinogen. The antibiotic chloram- phenicol was found by Anderson and Battle (1967) to cause a variety of teratogenic effects in zebrafish, including cyclopia and intermediate stages leading to this condition. Colchicine was likewise found by Waterman (1940) to cause a variety of anomalies in the medaka, including cyclopia. Aside from general retardation, the optic mal- formations were the major teratological effect of the insecticides on Menidia the first year. Skeletal malformations were also noted but they were relatively rare. In the following year, a vari- ety of malformations in addition to the optic ab- 210 normalities were produced. This difference is perplexing, and is probably due to genetic dif- ferences among individuals of this species in susceptibility to the chemicals. This is more understandable when it is realized that rela- tively few females can supply all the eggs needed for an entire experiment. Such variability in response makes this species a poor one to use in teratological studies. Effects were seen at dosages as low as 10 ppb. These are levels far lower than those which pro- duced noticeable effects in Fundulus heteroclitus embryos (Weis and Weis 1974) in which it was necessary to increase the dosage to parts per mil- lion. This may be due to differential permeability of the chorions of the two species and/or to a higher general resistance of Fundulus which is generally considered a hardier fish than Menidia. The dose levels which affected Menidia are levels near those which have been found temporarily, in solution or suspension, in natural areas (Finley et al. 1970; Kennedy and Walsh 1970). Therefore these adverse effects could occur during the de- velopment of fish embryos in nature. Acknowledgments We thank John C. Baiardi, Director of the New York Ocean Science Laboratory for making facilities available to us. We are especially grate- ful to Eugene Premuzic and John P. Wourms for sharing their laboratory and advice. Thanks are also extended to S. H. Gilani for reviewing our manuscript. Literature Cited Anderson, p. d., and h. I. Battle. 1967. Effects of chloramphenicol on the development of the zebrafish, Brachydanio rerio. Can. J. Zool. 45:191-205. BATTLE, H. I., AND K. K. HISAOKA. 1952. Effects of ethyl carbamate (urethan) on the early development of the teleost Brachydanio rerio. Cancer Res. 12:334-340. COSTELLO, D. P., M. E. DAVIDSON, A. EGGERS, M. H. FOX, AND C. HENLEY. 1957. Methods for obtaining and handling marine eggs and embryos. Mar. Biol. Lab., Woods Hole, Mass., 247 p. FINLEY, M. T., D. E. FERGUSON, AND J. L. LUDKE. 1970. Possible selective mechanisms in the development of insecticide-resistant fish. Pestic. Monit. J. 3:212-218. GREENBERG, J., AND Q. N. LAHAM. 1969. Malathion-induced teratisms in the developing chick. Can. J. Zool. 47:539-542. HISAOKA, K. K. 1958. The effects of 2-acetylaminofluorene on the em- bryonic development of the zebrafish. I. Morphological studies. Cancer Res. 18:527-535. KENNEDY, H. D., AND D. F. WALSH. 1970. Effects of malathion on twfo warmwater fishes and aquatic invertebrates in ponds. U.S. Bur. Sport Fish. Wild!., Tech. Pap. 55, 13 p. MCEWAN, R. S., J. B. BRIGGS, AND M. S. GILBERT. 1949. The effects of various abnormal agents applied to early developmental stages of Hemichromis bimaculata, and their theoretical significance. Anat. Rec. 105:491 (abstr.) MCLAUGHLIN, J., Jr., J.-P MARLIAC, M. J. VERRETT, M. K. MUTCHLER, AND O. G. FiTZHUGH. 1963. The injection of chemicals into the yolk sac of fertile eggs prior to incubation as a toxicity test. Toxicol. Appl. Pharmacol. 5:760-771. MITCHELL, J. T, AND C. L. YNTEMA. 1973. Teratogenic effect of malathion and captan in the embryo of the common snapping turtle, Chelydra serpen- tina. Anat. Rec. 175:390. Smithberg, M. 1962. Teratogenic effects of tolbutamide on the early de- velopment of the fish, Oryzias latipes. Am. J. Anat. 111:205-213. STOCKARD, C. R. 1907. The artificial production of a single median Cyclo- pean eye in the fish embryo by means of sea water solu- tions of magnesium chloride. Arch. Entwicklungsmech. Org. (Wilhelm Roux) 23:249-258. 1910. The influence of alcohol and other anaesthetics on embryonic development. Am. J. Anat. 10:369-392. TANIMURA, T, T KATSUYA, and H. NISHIMURA. 1967. Embryotoxicity of acute exposure to methyl para- thion in rats and mice. Arch. Environ. Health 15:609-613. WALKER, N. E. 1967. Distribution of chemicals injected into fertile eggs and its effect upon apparent toxicity. Toxicol. Appl. Pharmacol. 10:290-299 Waterman, A. J. 1940. Effect of colchicine on the development of the fish embryo, Oryzias latipes. Biol. Bull. (Woods Hole) 78:29-34. WEIS, P., AND J. S. WEIS. 1974. Cardiac malformations and other effects due to in- secticides in embryos of the killifish, Fundulus hetero- clitus. Teratology 10:263-267. JUDITH S. WEIS Department of Zoology and Physiology Rutgers University Newark, NJ 07102 peddrick Weis Department of Anatomy College of Medicine and Dentistry of New Jersey Newark, NJ 07103 211 GOOSE BARNACLES (CIRRIPEDIA: THORACICA) ON FLOTSAM BEACHED AT LA JOLLA, CALIFORNIA The macroscopic floating biota of the ocean surface — the pleuston — has been comparatively httle studied (see review by Cheng 1975). It com- prises a few species of insects, which skim over the surface; a few species of siphonophores equipped with floats; a few species of barnacles; etc. These organisms can be collected by the use of special nets towed at the level of the ocean surface, but the numbers of such tows made on oceanographic ex- peditions have been comparatively few compared with the much larger numbers of plankton tows made below the sea surface. Under exceptional circumstances, when an onshore wind blows for an extended period, pleustonic organisms are cast ashore in appreciable numbers, presenting un- usual opportunities to study numbers of individu- als of this little known community. Such mass beachings of the siphonophores Physalia and Ve- lella have been reported in several parts of the world (Bingham and Albertson 1974; Cheng 1975). This paper presents some data on a mass beaching of pleustonic goose barnacles, mostly at- tached to floating objects and mostly still living, found washed ashore between 5 and 9 July 1974, in front of the Scripps Institution of Oceanog- raphy, La Jolla, Calif. Methods A stretch of beach approximately 1 km long and 5 m wide was searched systematically for five successive days, around the time of the low tide in daylight, and every barnacle or piece of flotsam bearing barnacles was collected, taken to the laboratory in plastic bags, and there kept in tanks with running seawater. Some observations were made on the living animals, which remained alive, feeding actively, for several days, and specimens were photographed (Figure lA-F). They were sorted according to substrate, the species were identified, and the lengths of the capitula were measured from base of scutum to apex of tergum (peduncle lengths being variable). Observations In all, some 329 substrate objects were collected and examined; they bore a total of 2,555 individual barnacles. The data, for all collections, are sum- marized in Tables 1 and 2, and the size distribu- tions of each species on each of the major substrate types are shown in Figure 2A-L. The following generalizations were made on the basis of this material. Table l. — Numbers and percentages of substrates bearing barnacles: Lepas (Dosima) fascicularis and Lepas (Lepas) pad fica. Total Number Dosima + Lepas Dosima No. % Lepas % of total Dosima specimens Substrates No. % Lepas Feathers 878 657 75 221 25 34 34 Sea grass leaves: Phyllospadix Zostera 537 373 Subtotal 910 835 92 75 8 44 12 Brown algae; Macrocystis 202 117 58 85 42 6 13 Colpomenia Egregia Hall dry s Sargassum Scytosiphon 18 3 55 6 2 Subtotal 84 83 99 1 1 4 Terrestrial debris: Wood 69 Peanut shells 2 Plastic straws 9 Cigarette filters 5 Subtotal 85 47 55 38 45 3 6 Tar lumps None 322 74 113 61 35 82 209 13 65 18 6 3 33 2 Total 2,555 1,913 75 642 25 100 100 212 wmmmt^ >',:o- -tiSl, .B Figure l. — Lcpas (Dosima) fascicularis and Lepas (Lepas) pacifica, living specimens photographed in aquarium. A. Specimens of Dosima supported by their own floats at the water surface (note young barnacles attached to specimen on left); B. Right-hand specimen, from Figure lA, showing cirri withdrawn; C. Dosima on detached float of Macrocystis; D. Dosima on feather; E. Goose barnacles, mostly Lepas, on piece of Macrocystis; F. Small specimens of Lepas on flat lump of tar. 213 Table 2. — Numbers and percentages of barnacles [Lepas (Dosima) fascicularis and Lepas (Lepas) pacifica] on various substrates. Dosima only Lepas only Dosima + Lepas Total of substrates Substrates No. % No. % No. % No. % Feathers 41 45.5 7 8 42 46.5 90 27 Sea grass leaves: Phyllospadix 76 2 8 86 Zostera 22 1 13 36 Subtotal 98 80 3 3 21 17 122 37 Brown algae: Macrocystis 11 55 9 45 20 6 Colpomenia 5 5 Egregia 1 1 2 Halidrys 9 9 Sargassum 1 1 Scytosiphon 2 2 Subtotal 18 95 1 5 19 6 Terrestrial debris: Wood 8 3 6 17 Peanut shells 2 2 Plastic straws 1 1 Cigarette filters 1 1 Subtotal 11 52 4 19 6 29 21 7 Tar lumps 14 25 24 42 19 33 57 17 Total 193 59 39 12 97 29 329 100 The most common barnacle-bearing substrate was foimd to be bird feathers (90 items). The next most common were leaves of the surfgrass Phyl- lospadix (86 pieces) and tar (57 lumps). Other sub- strates included bits of brown algae Colpomenia, Egregia, Halidrys, Macrocystis, Sargassum, Scytosiphon; leaves of the sea grass Zostera; pieces of wood; cigarette filters; peanut shells; and plastic drinking straws. (Pieces of other debris without barnacles, such as polystyrene cups and plastic bottles and caps — many clearly of local origin — were not collected and are not further discussed here.) Most of the barnacles belonged to two species: Lepas (Dosima) fascicularis Ellis and Solander, the soft blue barnacle (about 75% of the individu- als); and Lepas (Lepas) pacifica Henry, a common Pacific goose barnacle (about 25%). Two other species of barnacle were also found: three speci- mens of Tetraclita squamosa on pieces of Mac- rocystis stipe, and one young specimen oi Lepas (Lepas) anatifera on a piece of tar. These have not been included in the data of Tables 1 and 2, and will not be considered further. Unattached (Figure lA, B) An appreciable number of the Dosima speci- mens (61) were found unattached to flotsam, either occurring singly, each with its own float, or else with several specimens sharing a communal float. Whether these had previously been attached to any substrate was not determined. The 13 unat- tached Lepas specimens found in our collections had probably become detached from substrates after they were collected. Feathers (Figures ID, 2 A, B) The feathers bearing barnacles were mostly large, more than 10 cm long, and were relatively intact with both quill and vanes. Most were white or grey; the species of seabirds from which they originated were not identified. Though a few of the barnacles were attached singly along the shaft, most occurred in clusters, generally near the dis- tal end of the feather. Such clusters comprised as many as 20 individuals of different sizes, many or all of which must have contributed to the com- munal bubble floats which in some specimens reached a diameter of almost 20 mm. The largest Dosima specimen found on a feather was 20 mm long; the largest Lepas, only 13 mm. About 50% of the feathers bearing barnacles had only Dosima specimens; only seven (7.8%) were found carrying Lepas alone, and on all of these the barnacles were rather small and few. On the feathers that carried a mixture of both species, the majority of the ani- mals were Dosima; in fact, some 18 of the Lepas specimens (all less than 10 mm) were found at- tached to the larger individuals of Dosima. The highest cluster numbers found on single feathers 214 Dosima FEATHERS Lepos ° lOOn a. SEA GRASS LEAVES [l 20 MACROCYSTIS Figure 2. — Size-frequency distributions of Lepas (Dosima) fascicularis and Lepas (Lepas) pacifica on various substrates as indicated. were 34 for Dosima and 15 for Lepas. Thirty-six of the Dosima clusters consisted of more than 10 individuals, whereas only six of the Lepas clumps on feathers comprised more than 10 animals. From these data it appears that on feathers Dosima is much commoner than Lepas and can occur more densely and in larger clumps, presum- ably because of its ability to produce its own float. Sea Grass Leaves (Figure 2C, D) Many of the Phyllospadix and Zostera leaves bearing barnacles had been completely bleached; possibly they had become detached from the par- ent plants and had drifted out to sea before being colonized. The majority of the leaf sections col- lected were found to carry one or more specimens of Dosima. Almost 809^ carried only Dostma; only 3% bore Lepas alone; the rest had both. As in the case of the feathers, the Dosima specimens at- tached to leaves had produced their own floats, as many as 23 individuals being found in one cluster. The largest specimens of Dosima found on Phyl- lospadix andZostera were 22 mm and 19 mm long, respectively. In contrast, on these substrates the Lepas individuals generally occurred either singly or in pairs, and the majority of these animals did not measure more than 5-6 mm in length, though a few of those which occurred together with Z)ostma exceeded 10 mm. Presumably, larger specimens of Lepas cannot be supported by a floating leaf sec- tion unless additional buoyancy is supplied by floats of Dosima. Brown Algae (Figures IC, E, 2E-H) It is significant that the only algae found bear- ing barnacles are parts of brown algae (Phaeophyta), which either produce well- differentiated gas-filled floats or, as in the cases of Colpomenia and Scytosiphon, have hollow thalli usually filled with air. The majority of the barna- cles were found on float-bearing segments of Mac- rocystis, and in Tables 1 and 2 the data for this alga, which occurs in offshore waters, are presented separately from those of other brown algae, which are more or less intertidal. Since none of these algae normally carry goose barnacles while grow- ing in their natural habitats, it appears probable that the pieces of thallus were colonized by barna- cles after they had been detached. They must have floated for some time, however, since the barnacles had reached appreciable sizes: up to 21 mm in length for Dosima and up to 12 mm in length for Lepas. With the exception of one piece of Egregia bearing a small 2-mm Lepas, the littoral brown algae bore only Dosima (83 specimens in all), whereas a large proportion of the Macrocystis pieces bore mixed populations. Terrestrial Debris (Figure 21, J) The majority of the fragments grouped in this category were pieces of wood, which may be con- sidered a "natural" substrate since fallen branches are a normal component of the flotsam carried 215 by rivers out to sea. So far, plastics — in pieces sufficiently large and buoyant to support goose barnacles — evidently constitute a substrate of only minor importance for this kind of animal. Tar (Figures IF, 2K, L) The 57 pieces of barnacle-bearing tar, presuma- bly originating from natural seepage or oil bun- kers, were mostly flattened 2-3 mm thick, 10-60 mm in diameter. This substrate, unlike those de- scribed hitherto, appeared to be preferred by Lepas. More than 42% of the lumps collected bore only this species, and many of the pieces had more than 10 animals attached. About 65% of the bar- nacles found on tar were of this species. Some were more than 15 mm long. They were generally not clumped, but occurred scattered over the surface of the substrate, often on both upper and under sur- faces, suggesting that the lump had repeatedly turned over while afloat on the ocean. Compara- tively fewer of the tar lumps bore only specimens ofDosima, and only 10 of these had more than 10 animals each. Per unit of surface area, the indi- viduals of Dosima appeared to be more sparsely distributed on tar than on feathers or grass leaves. Discussion Lepas (Dosima) fascicularis is the most specialized pleustonic goose barnacle, with an al- most uncalcified shell and a gas-filled bubble float. The larval stages were described on the basis of material collected and reared during the Chal- lenger Expedition (Willemoes-Suhm 1876). Since there were several errors and omissions in that paper, all the stages were redescribed by Bain- bridge and Roskell (1966). Boetius (1952-53) reported that all of the speci- mens of Dosima, which he found on the Danish North Sea coast in September 1952, had floats roughly proportional in diameter to the length of the animal. These barnacles are able to support themselves in the adult stage by their own float, but the cyprid larvae must settle on some sub- strate before they can metamorphose. The larvae of Dosima have been shown to settle preferen- tially on small floating objects; only later do they produce a bubble float which enables them to stay at the sea surface even when detached from such a support (Boetius 1952-53; Newman 1974). In our collections, all of the Dosima specimens, but none of the Lepas specimens, were attached to bubble floats of their own making. Some 27 individuals of Lepas (1-10 mm), the smaller of the two species, were found attached to larger specimens of Dosima, but, despite their larger absolute num- bers, only 8 Dosima specimens (1-14 mm) were found on other animals of this species. Evidently floating barnacle colonies do not normally grow by accretion in this way. The blue pigment of Dosima was studied by Fox et al. (1967), who reported that it is a conjugated carotenoid. Although many of the blue barnacles which they studied (washed ashore in the same location) were found attached to the floats of Ve- lella, and although we have found large numbers of these siphonophores stranded at various other times in recent years, we found no Velella floats among the barnacle substrates in this study. In fact, although hundreds of pleustonic barnacles were stranded on our beach during the period studied, we found no specimen of Physalia, Velella , or lanthina, which are all common components of the pleuston community in the open ocean. We found only one Glaucus (a pelagic nudibranch), a few specimens of Fiona (another nudibranch, normally associated with Macrocystis), and sev- eral polychaete worms. This probably indicates the relatively nearshore rather than oceanic ori- gin of the barnacle colonies. Although, when brought back to a laboratory aquarium and given fresh running seawater, many of the specimens remained alive and apparently healthy for more than 1 wk, such stranded animals are normally unable to return to the sea. When exposed to the sun on the beach they would probably be eaten by gulls or dry up within a few hours. We have not attempted to study the gut contents of our animals but assume that, like other barna- cles in nature, they probably feed mainly on microorganisms and small zooplankton (Howard and Scott 1959; Crisp and Southward 1961). We noted that in the laboratory, when supplied with a suspension of the unicellular alga Platymonas , many individuals of Dosima extended their cirri, apparently moving them towards the food source, directing it towards the mouth. Goose barnacles are hermaphrodites. Adults develop both male and female organs at the same time and can cross-fertilize each other. The eggs are brooded in the mantle cavities, and hatch as larvae which live in the plankton before settling. They attach themselves to a solid substrate by an adhesive secreted by the cement glands; the com- position of the cement of Lepas fascicularis has 216 been analyzed by Barnes and Blackstock (1974). We do not know how long it takes for them to reach the adult stage after metamorphosis. Horn et al. (1970), who collected 150 specimens ofLepas pec- tinata, 2-8 mm long, attached to four lumps of tar found floating on the sea surface, noted that, in the laboratory, these animals increased in length by about 1 mm per week. The larger specimens in our collections (20 mm for Dosima, 15 mm for Lepas) contained mature eggs. We have no information on the numbers of generations in the year; our size distribution data (Figure 2) show no evidence for separate generations (which might have been in- dicated by distinguishable size-class modes). Lepas species are known to be widely distributed from tropical to polar seas. Our specimens proba- bly came from populations floating in the eastern Pacific Ocean, which is the most likely area af- fected by the anomalous meterological conditions occuring during June and July 1974 (J. Namias, pers. commun.). Summary A total of 1,913 specimens of Lepas (Dosima) fascicularis and 642 specimens of L. (Lepas) paciftca, many still alive, were collected on a 1,000-m stretch of beach at La Jolla between 5 and 9 July 1974. They were attached to various sub- strates which had enabled them to float at the sea surface before being cast ashore. The predominant substrates were feathers (90 pieces, bearing 657 Dosima, 221 Lepas), sea grass leaves (122 pieces: 835 Dosima, 75 Lepas), brown algae (39 pieces: 200 Dosima, 86 Lepas), and tar (57 pieces: 113 Dosima, 209 Lepas). Dosima is the predominant species on most of the substrates whereas tar lumps appeared to be preferentially settled by Lepas. The size distributions (Dosima, 1-22 mm; Lepas, 1-16 mm) provided no indications of gen- erational discontinuities. The beaching of these normally pleustonic animals should be considered in relation to preceding and prevail- ing wind conditions. Acknowledgments We thank Connie L. Fey for her patient help in sorting and measuring the barnacles, James E. Rupert and James R. Lance for taking the photo- graphs, William A. Newman for specific iden- tification of the barnacles and valuable comments on the manuscript, and Jerome Namias for discus- sion of relevent meteorological data. Financial support from the Marine Life Research Group of Scripps Institution of Oceanography and from the Foundation for Ocean Research to Lanna Cheng is also grate fill ly acknowledged. Literature Cited BAINBRIDGE, V., AND J. ROSKELL. 1966. A re-description of the larvae of Lepas fascicularis Ellis and Solander with observations on the distribution of Lepas nauplii in the north-eastern Atlantic. In H. Barnes (editor), Some contemporary studies in marine science, p. 67-81. George Allen & Unwin Ltd., Lond. Barnes, H., and J. Blackstock. 1974. The biochemical composition of the cement of a pedunculate cirripede. J. Exp. Mar. Biol. Ecol. 16:87-91. Bingham, F. O., and H. D. Albertson. 1974. Observations on beach strandings of the Physalia (Portuguese-man-of-war) community. Veliger 17:220- 224. BOETIUS, J. 1952-53. Some notes on the relation to the substratum of Lepas anatifera L. and Lepas fascicularis E. et S. Oikos 4:112-117. Cheng, l. 1975. Marine pleuston — animals at the sea-air interface. Oceanogr. Mar. Biol., Annu. Rev. 13:181-212. Crisp, D. J., and a. J. southward. 1961. Different types of cirral activity of barnacles. Philos. Trans. R. Soc. Lond., Ser. B, Biol. Sci. 243:271-307. Fox, D. L., V. E. SMITH, AND A. A. WOLFSON. 1967. Disposition of carotenoids in the blue goose barnacle Lepas fascicularis. Experientia 23:965-967. Horn, M. H., J. M. Teal, and R. H. Backus. 1970. Petroleimi lumps on the surface of the sea. Science (Wash., D.C.) 168:245-246. Howard, G. K., and H. C. Scott. 1959. FVedaceous feeding in two common gooseneck barna- cles. Science (Wash., D.C.) 129:717-718. Newman, W. a. 1974. Cirripedia. Encyclopaedia Britannica, 15th ed. 4:641-643. WILLEMOES-SUHM, R. VON. 1876. On the development of Lepas fascicularis and the Archizoea of Cirripedia. Philos. Trans. R. Soc. Lond., Ser. B, Biol. Sci. 166:131-154. LANNA CHENG RALPH A. LEWIN Scripps Institution of Oceanography University of California La Jolla, CA 92093 217 CALORIC VALUES OF SOME NORTH ATLANTIC CALANOID COPEPODS Evaluation of the dynamics of energy exchange of a marine ecosystem necessitates a knowledge of the caloric equivalents of its living constituents. This information, in combination with informa- tion on growth, metabolism, and assimilation rates can lead to predictions of energy conver- sion between trophic levels and estimates of pro- duction. Researchers have accumulated a considerable quantity of data concerning the caloric value of marine organisms (Cummins 1967; Thayer et al. 1973; Tyler 1973); however, values recorded for marine, planktonic copepod species have been few (Slobodkin and Richman 1961; Comita et al. 1966; Cummins 1967). My research reports the caloric values for seven species of marine copepods, six of which apparently have not been previously re- corded. These studies are part of an overall inves- tigation of the bioenergetics of the early life stages of some North Atlantic fish species. Materials and Methods Plankton samples were collected in July and August 1972 off Narragansett Bay, R.I. except for samples of Pseudocalanus minutus which were collected in April 1971 off the coast of Delaware. All samples were preserved in 5% Formalin^ and were prepared and combusted in July and August 1972. Laboratory preparation included rinsing the samples in distilled water for 1 h, sieving through a coarse mesh screen to remove large detritus, and hand sorting adults of the various copepod species under a dissecting microscope. Pure copepod species samples were dried for 24 h at 90°C and desiccated in a silica gel desiccator after which they were made into pellets for combustion. All combustion was done in a Parr 1241 automatic, adiabatic calorimeter adapted for a microbomb. Combustion samples for each copepod species were done in triplicate. Percent ash for each copepod species was determined by ashing uncombusted pellets in triplicate at 500°C for 4 h in a muffle furnace. Results Mean values for the caloric determinations of the seven species of copepods (Table 1) were as follows: 5,251.9 cal/g dry weight, 5,626.3 cal/g ash-free dry weight, and 6.70% ash. Statistical analysis of the means of caloric values for each species (Duncan's New Multiple Range Test, Steel and Torrie 1960) indicated that Calanus ftninar- chicus had significantly higher values of both calories per gram dry weight and calories per gram ash-free dry weight than all other species, that Temora longicornis had significantly lower values for calories per gram ash-free dry weight than all species except Centropages hamatus, and that the differences between Acartia tonsa, Tor- tanus discaudatus, P. minutus, Centropages typicus, and C. hamatus were minimal (Table 1). Temora longicornis had the highest percent ash. Acartia tonsa and P. minutus also had relatively high ash values in comparison with the other species, while Calanus finmarchicus was inter- mediate and higher than the three remaining species (Table 1). Table l. — Caloric and ash values for some North Atlantic copepods. Species are recorded in order from largest to smallest mean value under each category. Those species side-scored have similar means (Duncan's New Multiple Range Test, P = 0.05). Standard Species Mean deviation cal/g dry weight {Calanus finmarchicus 6,425.1 ±187.0 Tortanus discaudatus Centropages typicus Acartia tonsa Pseudocalanus minutus Centropages hamatus 5,398.3 5,244.7 5,160.0 5,070.9 4,998.6 ±14.6 ±183.3 ±78.8 ±181.7 ±246.3 Temora longicornis 4,466.3 ±92.8 cal/g Calanus finmarchicus ash-free dry weight 6,835.2 ±191.2 Acartia tonsa Tortanus discaudatus Pseudocalanus minutus Centropages typicus 5,664.1 5,642.0 5,541.9 5,503.4 ±86.6 ±15.3 ±198.6 ±192.3 Centropages hamatus Temora longicornis 5,212.3 4,984.7 ±256.9 ±103.6 Temora longicornis % ash 10.40 ±0.16 Acartia tonsa Pseudocalanus minutus 8.90 8.50 ±0.16 ±0.11 Calanus finmarchicus 6.00 ±1.82 Centropages typicus Tortanus discaudatus Centropages hamatus 4.70 4.32 4.10 tO.28 b0.07 t0.13 ^Reference to trade names does not imply endorsement by National Marine Fisheries Service, NOAA. Discussion Since the species in this study were preserved in Formalin for short periods of time and rinsed in distilled water to remove the Formalin before pro- cessing, the estimates of caloric and ash content 218 and dry weight may have been slightly affected due to an unknown loss of chemical constituents. Methods of preservation of animals before com- busting or determining chemical composition and weights have been a subject of debate. Omori (1970) showed there was considerable variation with no apparent trend of chemical composition and weight ofCalanus cristatus that were frozen, dried, or preserved in Formalin. Except for dry weight, which was lowest in Formalin-preserved specimens, he found no clear relationship between percent ash, carbon, nitrogen, and hydrogen com- position and the methods of preservation. Faustov and Zotin ( 1965) determined that fixing by drying or in 4% Formalin had no significant effect on the caloric value of fish embryos and, consequently, results obtained with fresh or fixed material could be directly compared. In the present study, sam- ples of fresh and preserved (5% Formalin) C. finmarchicus were compared. Calories per gram dry weight and percent ash were less for the pre- served sample, however, the differences were min- imal (274.8 cal/g dry weight and 3.78% ash which corresponds to 275.0 cal/g ash-free dry weight) and only slightly greater than one standard deviation (Table 1). In view of the apparent lack of specific effects of preservation method on chemical composition, weights, and caloric values reported in the litera- ture and the results with C finmarchicus in this research, it may be concluded that the values pre- sented in this paper are only slightly underesti- mated, if at all. Also, since all samples in this study were treated the same way, relative com- parisons between them should be valid. Attempts to explain the differences in caloric values on the basis of phylogeny proved in- adequate. All species are calanoid copepods and, although C. finmarchicus and P. minutus are members of a different, more primitive taxonomic subdivision under the Calanoida than the other species (Sars 1903), the values for P. minutus were statistically more similar to the lower values for the other species than to C. finmarchicus. There is a lack of information on the specific chemical composition of the species tested in this research with the exception of C finmarchicus. Calanus finmarchicus is known to have a reasona- bly high fat content. Comita et al. (1966) noted that, upon fixation, globules of fat were extruded from living specimens and that a layer of oil formed on the surface of the fixed sample. They determined the caloric value of the fat of C. finmarchicus to be 9,500 cal/g. Fisher (1962) de- termined the lipid content for a number of marine Crustacea and found the concentrations in C finmarchicus to be consistently among the higher values recorded. Although there are no fat content values for the six other species tested in this re- search to compare with C. finmarchicus, the im- plication is that the lipid content in C. finmarchi- cus may be the cause of its higher caloric value. The caloric determinations of C. finmarchicus recorded in this research (Table 1) compare closely with the results of other workers (Slobodkin 1962; Comita and Schindler 1963; Comita et al. 1966). In fact, the caloric values of C. finmarchicus have been some of the highest recorded for copepods. Temora longicornis had lower caloric values than the other species and the highest percentage of ash (Table 1). This may be the result of its morphology which is somewhat different com- pared to the other species. It has a proportionately rounder and deeper cephalothorax that may con- tribute to a higher percentage of inorganic exo- skeleton. The overall means for the caloric values of all the species (5,251.9 cal/g dry weight and 5,626.3 cal/g ash-free dry weight) are similar to composite sample caloric values recorded by other inves- tigators. A calculation based on the data of Os- tapenya et al. (1967) using their values of calories per gram dry weight and percent organic matter for Gulf of Mexico plankton samples, which were predominantly copepods including Acartia sp., Centropages sp., and Temora sp. (separate values for each of these genera were not reported), pro- duced a mean value of 5,187 cal/g ash-free dry weight. A similar confirming value of 5,016 cal/g dry weight was obtained using the percent organic matter in the dry material in my research (calcu- lated by subtracting the mean percent ash, 6.70%, from 100) and the regression relationship between that and ash-fi-ee dry weight devised by Piatt et al. (1969). Seasonal changes in the caloric value of zoo- plankton have been verified in several studies (Comita et al. 1966; Conover 1968; Siefken and Armitage 1968). The species in this study undoubt- edly undergo seasonal variations also, and this is a subject for future investigation. However, all the species used in this research, with the exception of P. minutus, were collected at approximately the same time in the same general area and can be used for a comparison of the potential energy avail- able to predators at a particular time and place. 219 Examination of data on the abundance of adult and nauplii stages in the Narragansett Bay and Block Island Sound areas (Deevey 1952; Faber 1966) for the time of year samples for this research were collected (July-August) showed that, al- though all seven species were present, only A. tonsa, T. longicornis, andC. hamatus were availa- ble in sufficient quantity to be considered major prey organisms. They represented 24.6, 10.8, and 10.4%, respectively, of the total copepods availa- ble, while the other four species were less than 3%. The results of this study in calories per gram ash- free dry weight (Table 1) show that A. tonsa had the second highest value while C. hamatus and T. longicornis had the two lowest values. In fact, the difference between A. tonsa and T. longicornis is 680 cal/g. This indicates, assuming equivalent as- similation rates, that predators utilizing the copepods like A. tonsa with higher caloric values may have an advantage in acquiring energy for growth and metabolic processes. Predators feed- ing on copepods with lower values, especially T. longicornis, would have to consume more prey or- ganisms for an equivalent energy intake and, given the same density of plankton, would spend more energy searching for their prey. Acknowledgments I thank John B. Colton, Jr. for his critical re- view of the manuscript and Stephen Hale for his technical assistance. Literature Cited COMITA, G. W., AND D. W. SCHINDLER. 1963. Calorific values of microcrustacea. Science (Wash., D.C.) 140:1394-1396. COMITA, G. W., S. M. MARSHALL, AND A. P. ORR. 1966. On the biology of Calanus finmarchicus . XIII. Sea- sonal change in weight, calorific value and organic mat- ter. J. Mar. Biol. Assoc. U.K. 46:1-17. CONOVER, R. J. 1968. Zooplankton — Life in a nutritionally dilute environ- ment. Am. Zool. 8:107-118. Cummins, K. W. 1967. Calorific equivalents for studies in ecological ener- getics. 2nd ed. Univ. Pittsburg, Pittsburg, 52 p. DEEVY. G. B. 1952. Quantity and composition of the zooplankton of Block Island Sound, 1949. Bull. Bingham Oceanogr. Collect., Yale Univ. 13:120-164. Faber, D. J. 1966. Seasonal occurrence and abundance of free-swim- ming copepod nauplii in Narragansett Bay. J. Fish. Res. Board Can. 23:415-422. FAUSTOV, V. S., AND A. I. ZOTIN. 1965. Changes in the heat of combustion of the eggs of fishes and amphibians during development. Akad. Nauk SSSR (Doklady) Biol. Sci. 162:965-968. FISHER, L. R. 1962. The total lipid material in some species of marine zooplankton. Rapp. P.-V. Reun., Cons. Perm. Int. Explor. Mer 153:129-136. OMORI, M. 1970. Variations of length, weight, respiratory rate and chemical composition of Calanus cristatus in relation to its food and feeding. In J. H. Steele (editor), Marine food chains, p. 113-126. Univ. Calif. Press, Berkeley. OSTAPENYA, A. P., L. M. SUSHCHENYA, AND N. N. KHMELEVA. 1967. Caloricity of plankton from the tropical zone of the ocean. [In Russ., Engl, abstr.] Okeanol. Keanologiza 6:1100-1107. PLATT, T., V. M. BRAWN, AND B. IRWIN. 1969. Caloric and carbon equivalents of zooplankton biomass. J. Fish. Res. Board Can. 26:2345-2349. SARS, G. O. 1903. An account of the Crustacea of Norway, Vol. 4, Cope- poda Calanoida. Bergen Museum, Christiana, 171 p. SIEFKEN, M., AND K. B. ARMITAGE. 1968. Seasonal variation in metabolism and organic nu- trients in three Diaptomus (Crustacea: Copepoda). Comp. Biochem. Physiol. 24:591-609. SLOBODKIN, L. B. 1962. Energy in animal ecology. Adv. Ecol. Res. 1:69-101. SLOBODKIN, L. B., and S. RICHMAN. 1961. Calories/gm in species of animals. Nature (Lond.) 191:299. STEEL, R. G. D., and J. H. TORRIE. 1960. Principles and procedures of statistics. McGraw-Hill Book Co., Inc., N.Y., 481 p. thayer, g. w., w. e. schaaf, j. w. angelovic, and m. w. LaCroix. 1973. Caloric measurements of some estuarine organisms. Fish. Bull., U.S. 71:289-296. TYLER, A. V. 1973. Caloric values of some North Atlantic invertebrates. Mar. Biol. (Beri.) 19:258-261. Geoffrey C. Laurence Northeast Fisheries Center Narrangansett Laboratory National Marine Fisheries Service, NOAA Narragansett, RI 02882 METHOD FOR RESTRAINING LIVING PLANKTONIC CRUSTACEANS* Studies of the feeding and swimming mecha- nisms of small, active planktonic crustaceans re- quire restraining the organisms so that water flow and limb movements can be observed under the microscope. The usual technique is to place the organism in a watch glass or cavity slide (Cannon 1928; Gauld 1966) or to secure the dorsal side of the animal to a drop of stopcock grease in 'Contribution No. 3488 from the Woods Hole Oceanographic Institution. This work was supported by NSF Grant GA-41188. 220 some type of water chamber (McMahon and Rigler 1963). For many studies, these methods are undesirable because of the confinement of the animal to a small volume of medium or because of the solid boundaries nearby, both of which affect the flow of water and possibly the movement of limbs or other behavior by the animal (Lowndes 1935). Whenever the animal must be placed within a relatively large volume of water, other methods must be used. In a study of mate-seeking behavior, Katona (1973) tethered female copepods by means of fine stainless steel wires looped about their bodies. While this method allows the subsequent release of the animals unharmed, the restraining wire can interfere with limb movements. I have found a relatively simple method for re- straining small crustaceans in large volumes of water for extended periods of microscopic exami- nation. A short segment (1-2 cm) of nylon mono- filament fishing line of small diameter relative to the organism is mounted in a dissecting needle holder or pin vise. The free tip of the mono- filament is then cut off square with a razor blade. The animal is placed dorsal side up in a small drop of water on a microscope slide or watch glass. The tip of the monofilament is dipped in a fresh droplet of "instant" drying polymer glue (such as Dixon Duradix)^ and quickly applied and held to the center line of the dorsal surface of the animal for about 5 s. The organism can then be lifted from the slide and placed in the test vessel, with the dissecting needle holder mounted in a micromanipulator or other type of clamping de- vice. The rapid filming over of the glue and its tendency to spread when placed on the wet ani- mal sometimes makes a neat attachment difficult and several attempts may be needed before a satisfactory mount is achieved. Organisms restrained in this way appear to carry out swimming movements in a natural manner and live for several days on the mount. Removal of the animal from the monofilament usually results in its death. To make limb move- ments easier to observe, organisms can be vitally stained with neutral red prior to mounting (Dres- sel et al. 1972). I have since found a description of this mounting technique given by Scourfield (1900) in which he regrets that no satisfactory cement could be found. The polymer glues appear to solve the problem. ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Literature Cited Cannon, H. G. 1928. On the feeding mechanism of the copepods, Calanus finrrmrchicus and Diaptomus gracilis. Br. J. Exp. Biol. 6:131-144. DRESSEL, D. M., D. R. HEINLE, AND M. C. GROTE. 1972. Vital staining to sort dead and live copepods. Chesa- peake Sci. 13:156-159. Gauld, D. T. 1966. The swimming and feeding of planktonic cope- pods. In H. Barnes (editor), Some contemporary studies in marine science, p. 313-334. George Allen and Unwin, Ltd., Lond. KATONA, S. K. 1973. Evidence for sex pheromones in planktonic cope- pods. Limnol. Oceanogr. 18:574-583. LOWNDES, A. G. 1935. The swimming and feeding of certain calanoid cope- pods. Proc. Zool. Soc. Lond. 1935:687-715. MCMAHON, J. W., AND F. H. RiGLER. 1963. Mechanisms regiilating the feeding rate of Daphnia magna Straus. Can. J. Zool. 41:321-332. SCOURFIELD. D. J. 1900. The swimming peculiarities of Daphnia and its al- lies, with an account of a new method of examining living Entomostraca and similar organisms. J. Quekett Microsc. Club 7:395-404. LOREN R. HAURY Woods Hole Oceanographic Institution Woods Hole, MA 02543 OBSERVATIONS ON THE BIGEYE THRESHER SHARK, ALOPIAS SUPERCILIOSUS, IN THE WESTERN NORTH ATLANTIC Thresher sharks of the genus Alopias are distrib- uted throughout the tropical and warm temper- ate zones of the world's oceans. Of the two species reported from the western North Atlantic, the thresher shark, A. vulpinus, is commonly found in coastal waters of the middle Atlantic states (Bigelow and Schroeder 1948). The second member of the genus, the bigeye thresher, A. superciliosus , is a little known offshore resident of the continental slope and open sea. Lowe first described the bigeye thresher in 1840 from a specimen taken off the island of Madeira (Bigelow and Schroeder 1948). The species was not reported again until 1941 when Springer (1943) documented the occurrence of a gravid female taken near Salerno, Fla. Records of other bigeye threshers from the Atlantic include a gravid female, two embryos, a juvenile male, and an 18-foot specimen all taken from the north 221 coast of Cuba in the late 1940's (Bigelow and Schroeder 1948); an adult female from Nassau in 1962 and an adult male from Cape Hatteras, N.C., in 1963 (Fitch and Craig 1964). Bigelow and Schroeder (1948) reported proportional mea- surements from two individuals taken off Cuba; Strasburg (1958) and Fitch and Craig (1964) re- ported similar data from two Pacific specimens. We report observations of A. superciliosus taken on pelagic longlines aboard the commercial fishing vessel Cap'n Bill III, in 1962, the RY Dol- phin of the Sandy Hook Laboratory in 1966-69, and the RV Gosnold of the Woods Hole Oceano- graphic Institution in 1971. All previous evidence suggests A. superciliosus is not abundant any- where in its range. However, our data, together with anecdotal information from experienced commercial longliners, show that concentrations of bigeye threshers occur during April-June off Cape Hatteras. Other sharks and teleosts occur- ring in the area with A. superciliosus included blue shark, Prionace glauca; short fin mako shark, Isurus oxyrinchus; scalloped hammerhead, Sphyrna lewini; bignose shark, Carcharhinus al- timus; night shark, Hypoprion signatus; dusky shark, C. obscurus; and silky shark, C falciformis , along with swordfish, Xiphias gladius; and yel- lowfin tuna, Thunnus albacares. Additional 40* 35* 30* 65« Number taken at each station Cruise Sta. Position (Start Haul) Depth No. Vessel No. Date No. Lat 1 tude liClO' LongI tude Heters 110 Caught 1 Capt. Bill III 11-12-62 R/V Oolpfiin 0-66-'. S-Oft-Sfi 2 35°30' 7'<°'.7' 91'. 6 R/V Dolphin D-66-6 6-08-66 8 35°'<2' 7'<°36' 1829 6 R/V Dolphin D-68-5 6-06-68 9 35°12' 7'.°56' 1280 II R/V Dolphin D-68-5 6-07-68 10 35°I8' 7'<°57' 220 22 R/V Dolphin D-69-7 '(-03-69 6 27°I9' 63°00' 5'.86 1 R/V Dolphin D-69-1 1 5-17-69 i< 36°I2' Tk'sr 82 1 R/V Dolphin 0-69-11 5-18-69 5 35°23' 7'.°5I' 768 11 R/V Dolphin D-69-n 5-19-69 6 35°'.5' 7'.°'(6' 886 1. R/V Gosnold 175 it-lB-ZI 6 35°33' 7'.°36' 1829 1 R/V Gosnold 175 2-20-71 d 3'.°35' 75°25' 1829 1 65 z® _l L. 40' 35' -30' 80* 75* 70' 65' Figure l. — Location o{Alopias superciliosus longline catches in the western North Altantic. 222 species taken occasionally, included sandbar shark, C. milberti; oceanic white tip, C lon- gimanus; and porbeagle, Lamna nasus; bluefin tuna, Thunnus thynnus; white marlin, Tetrap- turus alhidus; sailfish, Istiophorus platypterus; dolphin, Coryphaena hippurus; and lancetfish, Alepisaurus sp. All longline sets resulting in catches of bigeye threshers were made between 0000 and 0300 with gear retrieval beginning after 0700. The depth at which the gear was fished ranged from near surface to a maximum of 65 m and was controlled by float lines of varying length. Tem- perature-depth profiles obtained from bathyther- mograph casts were routinely used to determine the optimum depth for the gear. The best catches of bigeye threshers were made in areas where the water regime ranged from 16° to 25°C at the sur- face to a minimum of 14°C at 75 m. A total of 65 A. superciliosus were hooked at 11 longline stations (Figure 1); of these, 7 broke free as they were being held alongside the vessel, 23 were tagged and released, and 35 ( 15 females and 20 males) were brought aboard for examination. Length measurements and internal examination of stomachs and reproductive organs were made on all sharks brought aboard. Total lengths (TL) for the 15 females ranged from 233 to 399 cm {x — 312 cm); the 20 males ranged from 155 to 352 cm (x = 307 cm). Morphometric measurements from eight males and four females, summarized in Table 1 as per- cents of fork length, were collected following the methods of Bigelow and Schroeder (1948). Fork length (FL) measurements were used as a pri- mary growth parameter in the morphometric re- lationships in order to discern more accurately any changes occurring in body proportions with increasing size. The same accuracy could not be expected if total lengths were used because of the difficulty in obtaining precise length measure- ments due to the extreme size and shape of the caudal fin. Proportional data from Table 1 shows that al- lometric growth is reflected in several characters. The most obvious change associated with increas- ing fork length is a proportionately shorter head length resulting in a decrease in the ratios of snout to: eye, nostrils, mouth, first gill, and pec- toral fin. The relative size of the eye and mouth also decrease as the body lengthens. Characters that increase allometrically with growth include height of first dorsal, length of claspers in males, and interspaces between fins except in females Table l. — Proportional dimensions of body parts in percent of fork length for 12 Alopias superciliosus. Male Female Body part 1 2 3 4 5 6 7 8 1 2 3 4 Total length (cm) 155.0 307.0 315.0 331.0 332.0 342.4 351.7 339.0 257.5 340.0 355.0 399.0 Fork length (cm) 100.0 188.0 192.5 197,0 1970 207.0 212.5 217.0 167.0 207.0 210.0 221.0 % of total length 64.5 61.2 61.1 59,5 59.3 60.4 60.4 64.0 64.9 60.8 59.1 55.3 Distance from snout to: eyes 9.0 6.4 7.4 6.3 6.3 7.5 6.7 7.4 6.7 6.8 62 7.2 nostrils 6.5 5.3 6.0 5.6 5.2 6.0 6.0 5.5 5.7 5.5 5.0 5.2 mouth 9.5 7.8 8.2 7.9 7.7 7.9 7.8 8.3 8.0 7.5 7.4 7.9 first gill (base) 25.5 21.2 23.6 23.4 20.6 22.7 22.3 226 22.4 22.9 22.1 21.7 pectoral 29.0 24.1 282 27.2 24.4 26.1 25.7 24.0 25.6 26.3 24.3 25.8 first dorsal 57.0 55.4 52.5 51 8 55.1 55.5 53.2 53.0 51.2 52.2 51.0 52.3 second dorsal 82.0 82.2 79.2 80.5 81.7 82.1 80.5 802 79.9 82.6 79.5 80.5 pelvic 66.0 65.2 66.2 66.0 64.5 66.9 65.2 64,1 65.3 64.4 64.3 66.0 anal 87.0 88.0 87.6 87.3 863 87.9 87.8 85.7 83.5 85.0 83.3 86.0 upper caudal pit 90.5 90.8 89.7 89.3 90.4 90.6 89.7 88.8 91.8 90.0 90.5 Interspace between: 1st & 2nd dorsal 16 8 17.8 16.4 17.0 198 16.6 17.2 18.0 18.9 18.4 17.9 18.1 2nd dorsal & caudal 7.2 7.8 86 8.8 8.8 8.1 8.3 9.2 7.7 7.9 8.2 pelvic & anal 9.5 12.2 13.0 12.9 11,3 11.8 13.2 12.9 7.2 8.2 7.4 8.1 anal & caudal 3.2 3.6 3.4 4.3 4.6 3.6 3.3 3.2 6.0 4.8 4.3 5.0 nostrils (proximal) 4.5 2.5 2.7 2.7 2.5 2.9 2.6 2.8 2.4 2.7 2.4 2.5 Height of: first dorsal 10.0 11.5 13.0 11,5 11.9 11.7 11.6 12.4 11.8 13.5 12.8 14.0 free tip 1.5 1.6 1.7 2.0 2.0 1.9 1.8 1.8 1.8 1.9 1.3 second dorsal .8 .9 .8 8 .8 .8 .8 .7 1.1 1.3 1.0 1.8 free tip .2 2.1 29 2.5 2,5 2.4 2.6 2.8 2.7 3.4 2.5 4.3 Diameter of eye' horizontal 3.5 2.5 2.6 2.4 2.4 2.9 2.5 3.5 2.8 2.8 2.8 3.2 vertical 4.0 4.2 4.2 4.2 4.5 4.4 4.4 3.8 3.8 Right clasper 3.0 12.4 13.0 12.9 11.9 12.4 10.8 12.0 Left clasper 3.1 12.4 11.4 12.9 11.7 12.1 11.6 12.0 Width of mouth 9.0 6.2 7.3 7.0 7.0 7.7 7.7 8.3 7.6 7.5 8.1 Height of mouth 5.0 4.5 4.7 4.8 4.4 4.3 5.0 3.6 4.7 4.3 4.5 Max length pectoral fin 32.3 31.2 33.6 32.0 32.1 31.6 31.8 31.8 32.3 35.5 32.4 33.5 'Orbit. 223 where the distance between anal and caudal fin decreases. The length-weight relationship for this species (Figure 2) was derived using data from 5 females and 11 males. To determine the regression line, the equation, log Y = 11.1204 + 2.99269 log X was calculated using the nonlinear least squares method of Pienaar and Thomson (1969). Clark and von Schmidt (1965) noted that adult and juvenile males of several species of sharks can be distinguished by the differences in the rel- ative size and rigidity of the claspers. This characteristic applies to A. superciliosus. Of the males examined, the claspers of all but five indi- viduals were large (10.8-13.0% of their FL), heav- ily calcified, and quite obviously mature. Internal examinations of the larger males revealed the presence of sperm in the epididymis and sper- (0 < o o UJ 200- / / / 150- / 100 / 90 r SO- / TO / SO- / SO V 40 O/ o 30 46 mm) supported a variety of epizoites while those ^35 mm were colonized by Perigonimus only. This may be attributable to the greater sur- face available for epizoite set on larger crabs and the lower frequency of molt for these crabs. The epizoites are inhabitants of the shelf-edge upper slope habitat within the bathymetric range of Bathynectes. Trilasmis (Poecilasma) has a known range along the western Atlantic from Martha's Vineyard, Mass. to Key West, Fla., hav- ing been recorded at depths from 21.6 to 1,733 m, chiefly on the carapace of the brachyurans Geryon 225 CO _J < 9 > 00 ID o Peridonimus Anomia Trilasmis I I SteQo poma Worm tubes ^ No fouling Figure l. — Occurrence ofepizoiteson male and female Ba^/i^'/ieciessMperfcus atanecdysis (C1-C4), proecdysis (D1-D4), and postecdysis (A2-B2) for four modal size (short carapace width) groups («35 mm, 36-45 mm, 46-57 mm, s=58 mm). quinquedens Smith (Pilsbry 1907) and Cancer borealis Stimpson, collected from the same cruises from yjhich Bathynectes were obtained. Trilasmis (Poecilasma) was also observed on mature lobsters, i/omarus americanus H. Milne-Edwards. These decapods are bathymetric associates of 5. superbus (Lewis 1975). Trilasmis (Poecilasma) has also been found on Hyposophrys noar, a brachyuran from the Straits of Florida (Williams 1974). Anomia aculeata has been recorded from the Arctic Ocean to Cape Hatteras, N.C. within a bathymetric range of 1.8 to 144 m (Smith 1937). The stations at which this pelecypod occurred on Bathynectes were in depths greater than 200 m. The hydroid, Stegopoma plicatile, is common along the east coast of the United States from Hudson Bay to Cape Hatteras with a bathymetric range of 45 to 1,733 m (Fraser 1944). Acknowledgments I thank Frank Holland (North Carolina Divi- sion of Commercial and Sports Fisheries) for col- lection of specimens (RV Dan Moore) and Charles Wenner (Virginia Institute of Marine Science; RV Albatross IV). Ship time on RV Columbus Iselin and RV Eastward was provided by J. A. Musick through NSF grants GS-37561 and GS-27725, re- spectively. Dale Calder (South Carolina Marine Resources Research Institute) identified the hy- droids found onB. superbus, and Mariana Doyle (U.S. National Museum) confirmed identification of Trilasmis (Poecilasma) kaempferi in- aequilaterale . I also thank George Grant, P. A. Haefner, Jr., Fred Jacobs, and W. A. Van Engel for their criticism of the manuscript. Literature Cited Capart, a. 1951. Crustaces decapodes, Brachyures. Exped. oceanogr. Beige dans les Eaux cotieres afr. Atl. Sud. (1948-1949) 3(l):ll-205. Drach, p., and C. TCHERNIGOVTZEFF. 1967. Sur la methode de determination des stades d'inter- mue et son application generale aux crustaces. Vie Milieu 18:595-609. 226 FRASER, C. M. 1944. Hydroids of the Atlantic coast of North America. Univ. Toronto Press, Toronto, 451 p. Lewis, E. G. 1975. Contributions to the biology of Bathynectes superbus (Costa) (Decapoda: Portunidae) from the Chesapeake Bight of the western North Atlantic. M.A. Thesis, Col- lege of William and Mary, Williamsburg. PILSBRY, H. A. 1907. The barnacles (Cirripedia) contained in the collec- tions of the U.S. National Museum. Smithson. Bull. 60:1-122. Smith, M. 1937. East coast marine shells. Edwards Brothers, Inc., Ann Arbor, 308 p. WILLIAMS, A. B. 1974. A new species oiHypsophrys (Decapoda: Homolidae) from the Straits of Florida, with notes on related crabs. Proc. Biol. Soc. Wash. 87:485-492. ELIZABETH G. Lewis Virginia Institute of Marine Science Gloucester Point, VA 23062 227 INFORMATION FOR CONTRIBUTORS TO THE FISHERY BULLETIN Manuscripts submitted to the Fishery Bulletin will reach print faster if they conform to the following instructions. These are not absolute requirements, of course, but desiderata. CONTENT OF MANUSCRIPT The title page should give only the title of the paper, the author's name, his affiliation, and mailing address, including Zip code. The abstract should not exceed one double- spaced page. In the text. Fishery Bulletin style, for the most part, follows that of the Style Manual for Biologi- cal Journals. Fish names follow the style of the American Fisheries Society Special Publication No. 6, A List of Common and Scientific Names of Fishes from the United States and Canada, Third Edition, 1970. The Merriam-Webster Third New International Dictionary is used as the authority for correct spelling and word division. Text footnotes should be typed separately fi'om the text. Figures and tables, with their legends and headings, should be self-explanatory, not requir- ing reference to the text. Their placement should be indicated in the right-hand margin of the manuscript. Preferably figures should be reduced by photo- graphy to 5% inches (for single-column figures, allowing for 50% reduction in printing), or to 12 inches (for double-column figures). The maximum height, for either width, is 14 inches. Photo- graphs should be printed on glossy paper. Do not send original drawings to the Scientific Editor; if they, rather than the photographic re- ductions, are needed by the printer, the Scientific Publications Staff will request them. Each table should start on a separate page. Consistency in headings and format is desirable. Vertical rules should be avoided, as they make the tables more expensive to print. Footnotes in tables should be numbered sequentially in arable numerals. To avoid confusion with powers, they should be placed to the left of numerals. Acknowledgments, if included, are placed at the end of the text. Literature is cited in the text as: Lynn and Reid (1968) or (Lynn and Reid 1968). All papers re- ferred to in the text should be listed alphabetically by the senior author's surname under the heading "Literature Cited." Only the author's surname and initials are required in the literature cited. The accuracy of the literature cited is the re- sponsibility of the author. Abbreviations of names of periodicals and serials should conform to Biological Abstracts List of Serials with Title Abbreviations. (Chemical Abstracts also uses this system, which was developed by the American Standards Association.) Common abbreviations and symbols, such as mm, m, g, ml, mg, °C (for Celsius), %, %» and so forth, should be used. Abbreviate units of mea- sure only when used with numerals. Periods are only rarely used with abbreviations. We prefer that measurements be given in metric units; other equivalent units may be given in parentheses. FORM OF THE MANUSCRIPT The original of the manuscript should be typed, double-spaced, on white bond paper. Please triple space above headings. We would rather receive good duplicated copies of manuscripts than car- bon copies. The sequence of the material should be: TITLE PAGE ABSTRACT TEXT LITERATURE CITED APPENDIX TEXT FOOTNOTES TABLES (Each table should be numbered with an arable numeral and heading provided) LIST OF FIGURES (entire figure legends) FIGURES (Each figure should be numbered with an arable numeral; legends are desired) ADDITIONAL INFORMATION Send the ribbon copy and two duplicated or carbon copies of the manuscript to: Dr Bruce B. Collette, Scientific Editor Fishery Bulletin National Marine Fisheries Service Systematics Laboratory National Museum of Natural History Washington, DC 20560 Fifty separates will be supplied to an author free of charge and 100 supplied to his organiza- tion. No covers will be supplied. I Contents — continued BRUSHER, HAROLD A., and LARRY H. OGREN. Distribution, abundance, and size of penaeid shrimps in the St. Andrew Bay system, Florida 158 MASON, J. C. Some features of coho salmon, Oncorhynchus kisuich, fry emerging from simu- lated redds and concurrent changes in photobehavior 167 HURLEY, ANN C. Feeding behavior, food consumption, growth, and respiration of the squid Loligo opalescens raised in the laboratory 176 GARRISON, DAVID L. Contribution of the net plankton and nannoplankton to the standing stocks and primary productivity in Monterey Bay, California during the upwelling season . 183 TRENT, LEE, EDWARD J. PULLEN, and RAPHAEL PROCTOR. Abundance of macrocrusta- ceans in a natural marsh and a marsh altered by dredging, bulkheading, and filling 195 Notes FISHER, WILLIAM S., and DANIEL W. WICKHAM. Mortalities and epibiotic fouling of eggs from wild populations of the Dungeness crab. Cancer magister 201 ■^ MATSUMOTO, WALTER M. Second record of black skipjack, Euthynnus lineatus, from the Hawaiian Islands 207 WEIS, JUDITH S., and PEDDRICK WEIS. Optical malformations induced by insecticides in embryos of the Atlantic silverside, Menidia menidia 208 CHENG, LANNA, and RALPH A. LEWIN. Goose barnacles (Cirripedia: Thoracica) on flotsam beached at La Jolla, California 212 ■*' LAURENCE, GEOFFREY C. Caloric values of some North Atlantic calanoid copepods .... 218 — HAURY, LOREN R. Method for restraining living planktonic crustaceans 220 " STILLWELL, CHARLES E., and JOHN G. CASEY. Observation on the bigeye thresher shark, Alopias superciliosus, in the western North Atlantic 221 LEWIS, ELIZABETH G. Epizoites associated with Bathynectes superbus (Decapoda: Portunidae) 225 5 ^'^^ ^ AMERICAS ,> ^^ FIRST INDUSTRY W rxv»r\ ftQC-O'a'a Fishery Bulletin ^ National Oceanic and Atmospheric Administration • National Marine Fisheries Service Mm Biological laiioralorii j LIBRARY AUG^ »7S Vol. 74, No. 2 I Woods Hole, M^ss^ j ^p^H .,975 PERRIN, WILLIAM R, JAMES M. COE, and JAMES R. ZWEIFEL. Growth and reproduction of the spotted porpoise, Stenella attenuata, in the offshore eastern tropical Pacific 229 SAKAGAWA, GARY T., and MAKOTO KIMURA. Growth of laboratory-reared northern anchovy, Engraulis mordax, from southern California 271 HEWITT, ROGER R, PAUL E. SMITH, and JOHN C. BROWN. Development and use of sonar mapping for pelagic stock assessment in the California Current area . . . 281 GRIFFIN, WADE L., NEWTON J. WARDLAW, and JOHN R NICHOLS. Economic and financial analysis of increasing costs in the Gulf shrimp fleet 301 V LIVINGSTON, ROBERT J., GERARD J. KOBYLINSKI, FRANK G. LEWIS, III, and PETER F. SHERIDAN. Long-term fluctuations of epibenthic fish and invertebrate l populations in Apalachicola Bay, Florida 311^T HAYNES, EVAN. Description of zoeae of coonstripe shrimp, Pandalus hypsinotus, reared in the laboratory 323 V CHITTENDEN, MARK E., JR. Present and historical spawning grounds and nurser- ies of American shad, Alosa sapidissima, in the Delaware River 343 LOUGH, R. GREGORY. Larval dynamics of the Dungeness crab, Cancer magister, off the central Oregon coast, 1970-71 353 HANSON, CHARLES H., and JONATHAN BELL. Subtidal and intertidal marine fouling on artificial substrata in northern Puget Sound, Washington 377 HASTINGS, ROBERT W, LARRY H. OGREN, and MICHAEL T MABRY. Observa- tions on the fish fauna associated with offshore platforms in the northeastern Gulf of Mexico 387 CRADDOCK, DONOVAN R. Effects of increased water temperature on Daphnia pulex 403 ^ ABLE, K. W., and J. A. MUSICK. Life history, ecology, and behavior oi Liparis inquilinus (Pisces: Cyclopteridae) associated with the sea scallop, PZacopecien ma- gelanicus 409 KJELSON, MARTIN A., and GEORGE N. JOHNSON. Further observations of the 1 feeding ecology of postlarval pinfish, Lagodon rhomboides, and spot, Leiostomus xanthurus 423 BREWER, GARY D. Thermal tolerance and resistance of the northern anchovy, I Engraulis mordax 433 (Continued on back cover) V. V Seattle, Washington U.S. DEPARTMENT OF COMMERCE Elliot L. Richardson, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION Robert M. White, Administrator NATIONAL MARINE FISHERIES SERVICE Robert W. Schoning, Director 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 WildUfe Service in 1941 . Separates were issued as documents through volume 46; the last document was No. 1103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate appeared as a numbered bulletin. A new system began in 1963 with volume 63 in which papers are bound together in a single issue of the bulletin instead of being issued individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued quarterly. In this form, it is available by subscription from the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 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. Bruce B. Collette Scientific Editor, Fishery Bulletin National Marine Fisheries Service Systematics Laboratory National Museum of Natural History Washington, DC 20560 Editorial Committee Dr. Elbert H. Ahlstrom National Marine Fisheries Service Dr. WilHam H. Bayliff Inter-American Tropical Tuna Commission Dr. Roger F. Cressey, Jr. U.S. National Museum Mr. John E. Fitch California Department of Fish and Game Dr. William W. Fox, Jr. National Marine Fisheries Service Dr. Marvin D. Grosslein National Marine Fisheries Service Dr. Edward D. Houde University of Miami Dr. Merton C. Ingham National Marine Fisheries Service Dr. Reuben Lasker National Marine Fisheries Service Dr. Sally L. Richardson Oregon State University Dr. Paul J. Struhsaker National Marine Fisheries Service Dr. Austin Williams National Marine Fisheries Service Kiyoshi G. Fukano, Managing Editor The Fishery Bulletin is published quarterly by Scientific Publications Staft National Marine Fisheries Service, NOAA, Room 450, 1107 NE 45th Street, Seattle, WA 98105. Controlled circulation postage paid at Seattle, Wash. The Secretary of Commerce has determined that the publication of this periodical is necessary in the transaction of the public business required by law of this Department. Use of funds for printing of this periodical has been approved by the Director of the Office of Management and Budget through 31 May 1977. Fishery Bulletin CONTENTS Vol. 74, No. 2 April 1976 PERRIN, WILLIAM R, JAMES M. COE, and JAMES R. ZWEIFEL. Growth and reproduction of the spotted porpoise, Stenella attenuata, in the offshore eastern tropical Pacific 229 SAKAGAWA, GARY T., and MAKOTO KIMURA. Growth of laboratory-reared northern anchovy, Engraulis mordax, from southern California 271 HEWITT, ROGER R, RAUL E. SMITH, and JOHN C. BROWN. Development and use of sonar mapping for pelagic stock assessment in the California Current area . . . 281 GRIFFIN, WADE L., NEWTON J. WARDLAW, and JOHN R NICHOLS. Economic and financial analysis of increasing costs in the Gulf shrimp fleet 301 LIVINGSTON, ROBERT J., GERARD J. KOBYLINSKI, FRANK G. LEWIS, III, and PETER F. SHERIDAN. Long-term fluctuations of epibenthic fish and invertebrate populations in Apalachicola Bay, Florida 311 HAYNES, EVAN. Description of zoeae of coonstripe shrimp, Panc?aZr/s hypsinotus, reared in the laboratory 323 CHITTENDEN, MARK E., JR. Present and historical spawning grounds and nurser- ies of American shad, Alosa sapidissima , in the Delaware River 343 LOUGH, R. GREGORY. Larval dynamics of the Dungeness crab, Cancer magister, off" the central Oregon coast, 1970-71 353 HANSON, CHARLES H., and JONATHAN BELL. Subtidal and intertidal marine fouling on artificial substrata in northern Puget Sound, Washington 377 HASTINGS, ROBERT W., LARRY H. OGREN, and MICHAEL T. MABRY Observa- tions on the fish fauna associated with offshore platforms in the northeastern Gulf of Mexico 387 CRADDOCK, DONOVAN R. Effects of increased water temperature on Daphnia pulex 403 ABLE, K. W., and J. A. MUSICK. Life history, ecology, and behavior of Liparis inquilinus (Pisces: Cyclopteridae) associated with the sea scallop, P/acopecten ma- gelanicus 409 KJELSON, MARTIN A., and GEORGE N. JOHNSON. Further observations of the feeding ecology of postlarval pinfish, Lagodon rhomboides, and spot, Leiostomus xanthurus 423 BREWER, GARY D. Thermal tolerance and resistance of the northern anchovy, Engraulis mordax 433 (Continued on next page) Seattle, Washington For sale by the Superintendent of Documents, U.S. Government Prmtmg Office, Washing- ton, D.C. 20402 — Subscnption price: $11.80 per year ($2.95 additional for foreign mail- ing). Cost per single issue - $2.95. Contents — continued Notes HARRELL, LEE W., ANTHONY J. NOVOTNY, MICHAEL H. SCHIEWE, and HAROLD O. HODGINS. Isolation and description of two vibrios pathogenic to Pacific salmon in Puget Sound, Washington 447 MAY, NELSON, LEE TRENT, and PAUL J. PRISTAS. Relation offish catches in gill nets to frontal periods 449 LANSFORD, LAWRENCE M., CHARLES W. CAILLOUET, and KENNETH T MARVIN. Phosphoglucomutase polymorphism in two penaeid shrimps, Penaeus brasiliensis and Penaeus aztecus subtilis 453 PERRIN, WILLIAM F. First record of the melon-headed whale, Peponocephala electra, in the eastern Pacific, with a summary of world distribution 457 CARLSON, H. RICHARD. Foods of juvenile sockeye salmon, Oncorhynchus nerka, in the inshore coastal waters of Bristol Bay, Alaska, 1966-67 458 LAIRD, CHAE E. , ELIZABETH G. LEWIS, and PAUL A. HAEFNER, JR. Occurrence of two galatheid crustaceans, Munida forceps and Munidopsis bermudezi, in the Chesapeake Bight of the western North Atlantic Ocean 462 WEIS, JUDITH S. Effects of mercury, cadmium, and lead salts on regeneration and ecdysis in the fiddler crab, Uca pugilator 464 FUIMAN, LEE A. Notes on the early development of the sea raven, Hemitripterus americanus 467 Vol. 74, No. 1 was published on 8 April 1976. The National Marine Fisheries Service (NMFS) does not approve, recommend or endorse any proprietary product or proprietary material mentioned in this publica- tion. No reference shall be made to MNFS, or to this publication furnished by NMFS, in any advertising or sales promotion which would indicate or imply that NMFS approves, recommends or endorses any proprietary product 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 purchased because of this NMFS publication. GROWTH AND REPRODUCTION OF THE SPOTTED PORPOISE, STENELLA ATTENUATA, IN THE OFFSHORE EASTERN TROPICAL PACIFIC William F. Perrin, James M. Coe, and James R. Zweifel^ ABSTRACT This study is based on data from several thousand specimens of spotted porpoise, Stenella attenuata, incidentally killed in the purse seine fishery for yellowfin tuna, Thunnus albacares. Average length at birth is 82.5 cm. Gestation is 11.5 mo. Average length at 1 yr is 138 cm. Length-weight equations are given for fetuses and postnatal males and females. Age was estimated from dentinal layers in thin sections of teeth. A two-phase Laird-Gompertz growth model was fitted to the layer-length data. Direct calibration of the dentinal layers beyond the first year (two layers) was not possible, and three alternative hypotheses were considered: 1) two layers per year, until pulp cavity occluded, 2) two layers per year in first year, and one per year thereafter, and 3 1 two layers per year until puberty, and one per year thereafter The second alternative is most probably the correct one, but reproductive parameters were estimated in terms of layers. Breeding is diffusely seasonal, with prolonged calving seasons in spring and fall and a pronounced low in winter A third calving season may exist in the summer. Average age at attainment of sexual maturity of males is approximately 12 layers (average length about 195 cm and averjige weight about 75 kg). Females attain sexual maturity on the average at about 9 layers and 181 cm. Ovarian changes in adult females are described. Apparently postreproductive females were encountered in the samples. It is concluded that corpora albicantia of ovulation and pregnancy persist indefinitely in the ovaries. It was not possible to distinguish between the two types of corpora. Ovulation rate changes with age, from about four per layer in very young adult females, to about one per layer in older females. The average calving interval is 26 mo long and consists of 11.5 mo of pregnancy, 11.2 mo of lactation, and 3.3 mo of resting and/or estrus. About 9.6% of lactating females are also pregnant. Pregnancy rate decreases with age, from about 0.6 per year at 8 to 10 layers, to about 0.3 at 16 layers. The overall sample contained 44.9% males and 55.1% females. Sex ratio changes with age, from near parity at birth, indicating higher mortality rates for males. Gross annual production of calves, based on age and sex structures of the sample and the estimated pregnancy rate, is 14.4% of the papulation per year. No evidence was found of age or sex segregation in schooling. The estimated parameters differ in a consistent way from those estimated for a population of Stenella attenuata in the western Pacific, possibly reflecting the exploitation in the eastern Pacific. Porpoises of the genera Stenella and Delphinus are killed incidentally in the tuna seine fishery in the eastern tropical Pacific (Perrin 1969, 1970a; National Oceanic and Atmospheric Admin- istration^). Since 1968, the National Ma- rine Fisheries Service (NMFS) has conducted a program of research into the population biology of the major porpoise species to assess the impact of this fishing mortality on the porpoise stocks. The purpose of this paper is to describe the life history of the spotted porpoise, Stenella attenuata ^Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92039. ^National Oceanic and Atmospheric Administration. 1972. Report of the NOAA Tuna-Porpoise Review Committee, Sep- tember 8, 1972. Unpubl. rep. U.S. Dep. Commer, Wash., D.C., 63 p. (Gray),^ the animal most frequently killed in the fishery. Little information on life history of the spotted porpoise has been available until very recently. Harrison et al. (1972) examined the gonads of 6 specimens from Japan (5 males and 1 female) and 45 specimens of S. attenuata fi-om the eastern tropical Pacific (19 males and 26 females), but did not separate their results and conclusions fi-om Manuscript accepted December 1975. FISHERY BULLETIN: VOL. 74, NO. 2, 1976. ^The taxonomy of the spotted porpoise has long been con- fused. Recent morphological studies (Perrin in press) have shown that the spotted porpoise in the tuna fishery is conspecific with the spotted porpoise occurring around Hawaii. The name S. attenuata (Gray 1846, holotype from un- known locality) applied by True (1903) to the Hawaiian form is used here for the eastern Pacific form, taking priority over S. graffmani (Lonnberg 1934). This usage is strictly provisional, pending the completion of current taxonomic studies, when a different name, such as S. dubla (G. Cuvier 1812) or S. frontalis (G. Cuvier 1829) may take priority. 229 FISHERY BULLETIN: VOL. 74, NO. 2 those for S. longirostris . Preliminary unpublished results of our studies indicate that these two species are probably disparate in such growth parameters as length at birth, length at maturity, and asymptotic length. Harrison et al. (1972) stated that lengths of the fetuses examined indi- cate that parturition occurs both in the spring and in the autumn. They described in detail the gross and microscopic histological appearances of several pairs of ovaries. A maximum of nine cor- pora albicantia were encountered. They con- cluded that if all the corpora albicantia in ovaries of specimens of this species do not represent past pregnancies, either the fertility is very low or the corpora are not permanent. Nishiwaki et al. (1965) published length- frequency distributions of 34 fetuses (up to 106 cm long) and 194 postnatal animals (104 to 208 cm) from a school driven ashore in Japan. They esti- mated that gestation lasts 1 yr, length at birth is about 105 cm, juveniles reach 150 cm in 6 mo, and adult size (180 cm for females and 190 cm for males) is reached in 1 yr. They concluded that there are two seasons for mating and parturition, in the spring and in the autumn, and that there are fewer males than females among adults. On- togenetic changes in coloration, external propor- tions, organ weights, the skeleton, parasite load, and feeding habits have been described (Perrin 1970b, in press; Perrin and Roberts 1972; Dailey and Perrin 1973; Perrin et al. 1973). Kasuya et al. (1974) recently published results of a study of several hundred specimens caught in the Japanese fishery for S. attenuata. Their re- sults are discussed and compared with ours in the body of this paper. METHODS AND MATERIALS Observer Program Beginning in 1968, NMFS placed observers aboard U.S. tuna seiners to collect information on the incidental take of cetaceans in the eastern tropical Pacific. Observers were placed on 1 cruise in 1968, 5 in 1971, 12 in 1972, and 22 in 1973. Most of the cruises were 30 to 60 days long. In addition, biological data were collected during chartered cruises of commercial seiners: one in 1971, one in 1972, and two in 1973. The data collecting had to be carried out in such a way as to not interfere with the fishing operation. Hence, the amount of information col- lected on the animals killed in a net set varied vddely, depending on the amount of time that was available before the next set was made. Following is the hierarchy of types of data that were col- lected (sample sizes were largest for the first and smallest for the last): Animals killed were: 1. Counted (estimates were made in cases where counts were not possible), usually on the deck or in the net, 2. Identified to species (and race when possible), 3. (S. attenuata only) identified to develop- mental color pattern phase (Perrin 1970b), and sexed, 4. Measured (to nearest centimeter with 2-m calipers), and 5. Dissected to collect information on repro- ductive condition (for females, mammaries were examined and reproductive tract col- lected; for males, the right testis was col- lected) and age (a section of the left lower jaw at midlength was collected). The gonad- al material and jaw sections were pre- served in 10% Formalin.'* Small fetuses (^ 30 cm) were preserved in the uterus. Larger fetuses were removed from the uterus and frozen. For each specimen that was at least measured (step 4 above), a field serial number was assigned, and a specimen data sheet was filled out. Data for specimens that were not at least measured were collected on a running tally. The Study Area One of us has described the distribution ofS. attenuata in the eastern tropical Pacific (Perrin 1975). The known occurrence of mixed aggrega- tions of cetaceans and tuna is strongly correlated with certain oceanographic conditions peculiar to that region. The porpoise-tuna association is known only in the eastern tropical portion of the Pacific. That area, which has been called the North Pacific Equatorial water mass (Seckel 1972), has an unusual oxygen-salinity- temperature structure. The reason for this is not ••Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 230 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE fully understood but certainly has to do with the latitude of the area, its position relative to the rest of the Pacific and to the American continents, and the shapes of the adjacent land masses. These factors interact with general global oceanic and atmospheric circulation to produce a water mass with relatively high surface temperature, low surface salinity, a strongly developed, shallow thermocline (usually within 100 m of the surface), and a pronounced, thick oxygen minimum layer just below the thermocline. The effect is to create a very extensive but shallow warm habitat with a sharp oxythermal floor. To the west, these condi- tions tail off along a divergence centered on lat. 10°N (Wyrtki 1964). The conditions must be caus- ally interrelated, but one of the more striking correlations with the occurence of the mixed- species aggregation is in the thickness of the oxy- gen minimum layer (Figure 1). The occurrence of the aggregation is not tightly correlated with the geographic distributions of the major prey species of the participating pred- ators. Major shared prey items are the omma- strephid squid Dosidicus gigas, an unidentified ommastrephid (probably Symplectoteuthis sp.), a scombrid fish Aj/x/s sp. (A. thazard or A. rochei), and the exocoetid fish Oxyporhamphus microp- terus (Perrin et al. 1973). Dosidicus gigas is primarily equatorial but migrates sporadically as far as California and southern Chile, far beyond the limits of the distribution of the mixed-species aggregation (Clarke 1966; Young 1972). Species of Symplectoteuthis occur widely in the tropical Pacific and Indian oceans (Clarke 1966). Auxis thazard occurs in "tropical and subtropical wa- ters of the Indo-Pacific and Atlantic oceans," and A. rochei in "tropical and subtropical waters of the Indo-Pacific and Atlantic oceans, including the Mediterranean Sea" (Richards and Klawe 1972). The genus Oxyporhamphus is also pantropi- cal (Bruun 1935). At least some of the several myctophid fishes in the aggregate apparently are a mainstay of the diet of the spinner porpoise in mixed schools (Perrin et al. 1973) and are not re- stricted to the tropics but occur also in temperate waters of the eastern Pacific (Moser and Ahlstrom 1970) and elsewhere. These facts, combined with the pantropical distributions of the cetaceans, tunas, and birds, suggest that the multispecies aggregation does not have its roots in the dis- tribution of the component species or their prey but rather in the peculiarities of the physical oceanography of the region. The Sample In 1971 and early 1972, when more specimens were decked than could be processed in the time available (the limit per net set was usually about 35 to 40 specimens), adult females were selected for measuring and dissection. The intention was to insure that sample sizes would be large enough to allow estimation of pregnancy rate with adequate precision. The information on age struc- ture of the catch for that period is limited to the coloration phase data. The observer program sub- sequently expanded, and beginning in October 1972 no selection was practiced in determining which animals were to be dissected; the first 35 to 40 specimens of both sexes and all ages that came to hand were set aside for measuring and dissec- tion and the remainder discarded. The length data for 1968 and for October 1972-December 1973 are presumably cross-sectional with respect to the kill. The sample of animals at least measured in- cluded 3,504 postnatal animals and associated fetuses fi'om known localities and 23 from impre- cisely known localities (Figure 2). Coloration phase and sex data were collected for another 6,150 specimens. In addition, some data were available for 45 other specimens collected by other research agencies, museums, and private individuals. Because of the seasonal nature of the tuna fishery, the sample is heavily biased toward the early months of the year, with minimal cov- erage of the latter part of the year and practically no specimens from the summer months (Table 1). Two races of S. attenuata exist in the eastern tropical Pacific — a large coastal form and a small offshore form (Perrin 1975, in press). This paper deals only with the offshore form. The es- timates of life history parameters cannot be as- sumed to apply also to the coastal form. Table L— Samples of postnatal spotted porpoise by month for all years. Month Males Females Total January 748 443 1,191 February 263 209 472 March 298 147 445 April 216 155 371 May 181 97 278 June 69 58 127 July 1 1 August 6 5 1 1 September October 222 158 380 November 110 87 197 December 30 24 54 Total 2,144 1,383 3,527 231 FISHERY BULLETIN; VOL. 74, NO. 2 30' 20' 10' 0°- 160= 140"= 140* 120 100* Figure l. — CompEirison of the known occurrence of spotted porpoise in the eastern Pacific (above) with average thickness of the subsurface layer of water (contours in meters) in which the dissolved oxygen is less than 0.25 ml/liter (below, after Knauss 1963). The entire layer lies above 1,000 m. 232 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE 135° 130" 125" 120" 0° 105° 100° 95" 90° 85° 60" 76 123 12 13 63 32 Galopagos is c^ 4 1 8 -■ 32 135° (30" 125" 120" 115° 110° 105° 100" 95° 90° 85° 80° Figure 2. — Samples of spotted por- poise used in life history studies by 5° square. Does not include speci- mens that were not at least measured. Because the field program is a continuing one, the sample sizes for the various analyses were different and depended on how much material was available at the time each analysis com- menced. Restrictions on sample size are set out in the text below. Laboratory Procedures Fetuses were measured with dial calipers or with calipers mounted on a 1-m stick. Postnatal animals were weighed to the nearest pound on platform scales. Fetuses were weighed to the nearest gram on a triple beam balance. Testes were weighed to the nearest gram on a platform balance. A 1-cm^ cube from the center of each testis^ and a similarly sized sample of the epididymis from midlength of the testis were sec- tioned and stained with hematoxylin and eosin. *Some early samples were taken near the dorsal surface of the testis. Tubule diameter in these was subsequently found not to differ relative to length, weight, and age of the animal from that in those taken at the center of the testis, and the lots were therefore combined for analysis. The mounted sections were subsequently examined under a compound microscope. Ovaries were weighed to the nearest 0.1 g on a platform balance. They were then cut into trans- verse sections approximately 1 mm thick with a scalpel and the sections examined under a dissect- ing microscope. The corpora albicantia in each ovary were scored to eight categories based on size, color, vascularization, and gross appearance (categories described below). If a corpus luteum was present, it was measured with dial calipers to the nearest millimeter in its three largest dimen- sions. The diameter of the largest follicle was measured to the nearest 0.1 mm. Age was estimated for 442 animals by exami- nation of dentinal layers in the teeth. Three or four teeth were extracted from the lower right tooth row at approximately midlength and mounted on wooden blocks in dental wax or plas- tic resin. A longitudinal section 0.012 inch (0.31 mm) thick was cut from each tooth with a diamond saw. The sections were cleared for sev- eral days in a 1:1 mixture of glycerine and 95% ethanol, mounted under cover slips in balsam, and examined with transmitted light under a 233 FISHERY BULLETIN: VOL. 74, NO. 2 compound microscope at approximately 30 diameters. One postnatal layer was considered to consist of an opaque subunit and a translucent subunit (Figure 3). The layers in most of the teeth examined were not as well-defined or as regular in thickness as those illustrated by Kasuya (1972) for Stenella coeruleoalba or by Klevezal' and Kleinenberg (1969) for Delphinus delphis. Teeth from 39 of the 442 animals were completely un- scorable, being heavily worn or showing no dis- crete layers in the sections examined. All the teeth were scored several times, over a period of several months, without referring to specimen numbers or to values obtained previously, until the scorer felt confident of the results. The values used in the analyses are those obtained in the final round of scoring. The teeth were scored to the nearest postnatal layer when possible, or a range, e.g., "8 to 10 layers," was estimated. Aver- age accuracy is estimated at ±1 layer for teeth with 5 layers or less and ±2 layers for teeth with 5 to 12 layers. Convoluted secondary dentine was present in most of the teeth vdth more than 12 layers, making counts very difficult and of dubi- ous reliability. We feel that the counts for many of these teeth are probably underestimates. Teeth in which the pulp cavity was entirely closed in all sections examined were scored as "occluded." The NORMSEP computer program was used to define modes in the length-frequency distri- butions for fetuses. The program was written by Hasselblad (1966) and modified by Patrick K. Tomlinson, Inter-American Tropical Tuna Com- mission. The program separates the mixture of normal length distribution into its components, assuming that the lengths of individuals within age groups are normally distributed and that an unbiased sample of the length distribution was obtained. GROWTH Length at Birth Average length at birth of 82.5 cm was obtained from a linear regression line based on 3-cm group- ings of fetuses and neonatals (Figure 4). The largest fetus of the 461 examined was 904 mm long. The smallest neonatal animal was 780 mm long. Eighty-six calves and fetuses between 73 and 94 cm were measured in random samples. As- sumptions inherent in the method used to arrive at this estimate are that pregnant females and calves are 1) equally vulnerable to capture in the purse seine, 2) equally likely to die once captured, and 3) equally represented in the sample of dead animals measured. For example, if neonates were less likely to be included in the samples than were pregnant females, average length at birth would be overestimated. Other potential sources of error are differential rates of prenatal and postnatal natural mortality and premature births caused by stresses imposed by pursuit and by capture in the purse seine. Gestation Period and Fetal Growth The most commonly used method for estimating the gestation time of cetaceans is that of Huggett and Widdas ( 195 1). They showed that for a variety of mammals of widely different orders, a plot of the cube root of fetal weight on age is linear except during the first part of pregnancy, when growth is exponential. Their model can be expressed in the general formula W" = a(t - to), where W = weight, t = age, a = the "Specific Fetal Growth Velocity," andto = "the intercept where the linear part of the plot, if produced backwards, cuts the time axis." Laws (1959) applied the method of Huggett and Widdas to fetal length/time data for three odontocetes [Physeter catodon, Delphinap- terus leucas , and Phocoena phocoena) and obtained estimates of gestation periods (15, 14, and 11 mo, respectively). He assumed that weight is propor- tional to the cube of length and used the form L = aitg - to),-whereL = length. This assumption is not entirely correct (see length-weight results below), but is a close enough approximation of the real relationship between length and weight to allow its use in estimating gestation period. Laws' esti- mates corresponded closely with other estimates obtained by more direct methods. Laws' version of Huggett and Widdas' method is used here. A gestation period of 11.5 mo was obtained from an analysis based on 281 fetal and postpartum specimens collected in January, February, March, April, May, and October 1972 (Figure 5). The January-May samples comprised all of the fetuses of all of the females examined. The postpartum samples in these months were not random and are therefore not included. The October samples were random over all age-classes in the catch; therefore, all specimens less than 160 cm long, approxi- mately the length at onset of puberty (Harrison et al. 1972), are included in the plot. Obvious modes are present in the length distributions (seasonal- 234 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE \r\ \ % S I) '* f ^i :-?•<. \ ■v^. ^^ Figure 3. — Longitudinal thin sections of teeth from two specimens of Stenella attenuata from the offshore eastern tropical Pacific. (Left) field number CV300 male, 144 cm, with two postnatal dentinal layers; (right) number LR55 female, 191 cm, with 13 layers. 235 < I— < CO o CL 100 80 60 40 20 Average length at birtti I I I I I j^ 74 77 80 83 86 89 92 (14) (9) (17) (II) (II) (17) (7) LENGTH (cm) Figure 4. — Linear regression analysis of percent postnatality on body length for 86 fetuses and calves of Stenella attenuata from the offshore eastern Pacific grouped in 3-cm intervals. Sample size for each 3-cm interval in parentheses. ity is discussed below). Apparent progression of the smaller mode in the January 1972 sample is consistent with a gestation period of roughly 1 yr. Sample sizes for the other apparent modes are not large enough for similar analysis. Linear regres- sion analysis of the modal lengths plotted on month (Figure 6) yields an estimate of the slope to use in Laws' equation: L = 8.283 {t - to), or length at birth = 8.283 {t^ - to) with (tg -^(j) (using months of 30. 4 days) = 9.96 mo or 303 days, where tg = total gestation period. Laws (1959) proposed that to for length data is slightly less than for weight data and assumed ^OLn = 0-9 ^0^^- Roughly interpolating between Huggett and Widdas' values for tf^/tgOf 0.1 for tg > 400 days and 0.2 for tg = 100 to 400 days (using provisional tg = 330 days to enter the itera- tion) (Figure 7) and applying Laws' correction, ^OLn of - 0.135 tg is obtained. This value yields an estimate of gestation time of 11.5 mo (349 days). The estimate of ^^ (47 days) is crude, but the true FISHERY BULLETIN: VOL. 74, NO. 2 AVERAGE LENGTH AT BIRTH 20 10 20 10 ii* 20 -g 10 I I I I '~i p JAN 1972 FEB. 0-*— ^ ^ 20i CO ID -FF| n I fp MAR. 20 10 r-n APR. 04 201 10 _a_ P , Fh n n ■T 1 \ r- MAY OCT. 20 40 60 80 100 120 140 160 LENGTH (cm ) Figure 5. — Length-frequency distributions by month for 281 fetal (open) and young (hatched) postnatal specimens of Stenella attenuata captured by tuna seiners in the offshore eastern tropical Pacific in 1972. value probably lies between 0.12 tg and 0. 15 ^g . We therefore estimate the gestation period to be 11.5 ± 0.2 mo (interval between estimates using ^o = 0.12 tg and 0.15 tg), or 11.3 to 11.7 mo. Postnatal Growth The same cohort used for analysis of fetal growth can be followed through the samples until approximate length of 125 cm (Figure 8) at the age of approximately 8.5 mo. In order to optimize res- olution, 4-cm intervals were used, and the sam- ples for April, May, and June were combined. Modes were estimated with NORMSEP. A linear regression line through the modal lengths yields the postnatal growth equation 236 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE IbU ' 140 _ e CJ l?0 . ri c> > 100 z AVERAGE LENGTH AT BIRTH ^^-^* BO . ^^ 60 o ill ;V^'' (') o < 40 ^^9^ liJ > < 20 ^ ^ o 1 1 1 2 •1971 I 1972-^ MONTH Figure 6. — Linear regression analysis of modal lengths of fetal and neonatal specimens of Stenella attenuata (from Figure 4). Open circles are modes not included in the analysis. Modal lengths defined with computer program NORMSEP (see Materials and Methods). 05 CJ> 0.4 0.3 0.2 0.0 \ \ \ \ \ 100 200 300 400 500 GESTATION PERIOD (days)-tg Figure 7. — Ratio toltg interpolated between empirical esti- mates of Huggett and Widdas (1951) — ". . .for gestation times up to 50 days ^o ~ 0.4 x (gestation time), from 50-100 days 6 - / ' Globlcepholo 2 c / o a> — Phocoeno / ^2 /v4 - r * 2 coeruleoQiba »- trt a> a; 5 " 3 - 1 2 25 - a> — 2 ■n o O 3 *" CT -— ai 15 o a> Li. 1 1 3 20 30 40 60 100 160 200 300 400 600 800 1000 S attenuoto Length at birth (cm ) Figure lO. — Relationship of difference between fetal growth rate during linear phase and average growth rate during post- natal period equal to gestation period to length at birth in five odontocete cetaceans. Line is linear regression line of log difference on log length. Data from Figure 11. Y is predicted difference for Stenella attenuata from the offshore eastern tropical Pacific. 4.5 for Globicephala, and 4. 5 for P. phocoe?ia)^; and yields a predicted length at 1 yr of 138 cm. Length Relative to Tooth Layers Total length was plotted on number of postnatal layers for 115 males and 306 females (Figure 11). The teeth of five males and three females had completely filled pulp cavities. These are included in the plots in a separate category "occluded." The plots of means for 2-layer intervals (the points in Figure 12; the curves were fitted as ex- plained below) very closely resemble the growth curve obtained by Sergeant (1962) for Globicephala. Asymptotic length (L ^) for females is approximately 190 cm and for males approxi- mately 200 cm. There appears to be a secondary surge in growth at about 6 layers. With the restric- tion that the curves must pass through birth length of 82.5 cm and asymptotic lengths of 190 and 200 cm, it is not possible to fit any continuous equation to the data satisfactorily. Continuous curves that fit well at the upper and lower ranges of layer count seriously underestimate length at 5 *Fisher and Harrison (1970) stated that their data suggest that Phocoena in Canadian waters grows approximately 30 cm during the first year of life, or at an average rate of about 2.5 cm/mo, as opposed to the 4.5 cm/mo hypothesized by M0hl- Hansen (1954). However, they also suggested, and their figure 2 showed, an average rate of at least 5 cm/mo during the first 4 mo. It seems unlikely that the rate would drop to an average of — 1.25 cm/mo in the remaining 8 mo of the first year. to 7 layers. Kasuya (1972) also encountered difficulty in attempting to fit a continuous model to growth of a delphinid, S. coeruleoalba . Good fits can be obtained, however, by assuming a dynamic growth function. A two-phase version of Laird's (1969) growth model was fitted to the 2-cm means for all males and females, using an iterative least-squares method. The occluded specimens were assigned to the 16+ interval. Laird's model is Lit) exp exp (- at) where L(t) = length at time t Lq = length at birth ( 82. 5 cm in this case) t = time (layers in this case) a = specific rate of exponential growth a = rate ofdecay of exponential growth. This model assumes that an organism's growth pattern is determined at conception. The fitted parameters a and a express the premise that "growth is fundamentally exponential (implied by the normal binary fission of cells), and it also un- dergoes exponential retardation by some as yet unknown physiological mechanism" (Laird 1969). In the two-phase approach, separate equations were simultaneously fitted to the upper and lower range of means. The assumptions were made that juvenile growth is the same for males and females (supported by the data) and that the growth dis- continuity comes at about the same age for males and females. The only fixed point was 82.5 cm at layers (birth). The convergence point (inflection in the growth curve) was allowed to float to the posi- tion that gave the best fit, with males and females considered jointly for lesser ages. The equations converged at 5.59 layers (rounded off to 6 below) at which predicted length is 159.9 cm. The fit is excel- lent for females (Figures 11, 12). Asymptotic length is 190 cm at predicted age of 18 layers. Average length of adult females (those with ovar- ian scars) is 187.3 cm, based on a sample of 555 (Perrin 1975). The largest female of 2,138 mea- sured was 220 cm long. The equation for juvenile growth to less than 6 layers is L = 82.5 exp 0.4817 0.7172 [l - exp (-0.7172^], where L = length, in centimeters t = age in layers. 239 FISHERY BULLETIN: VOL. 74. NO, 2 200 % 150 o X I- 21 LU < I- O 100 -/ 50 !i . ! • • • •• N = 120 Average_lenqt^h_cit_bM^tlni , s I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 OCCLUDED POSTNATAL DENTINAL LAYERS (number) - — • • • • • 200 — • • I f jti^M* * 1 »■ i • • • 10 % E - t — ■^ ^ QJ 150 — ^ O • %/* • : X 1- ./ 99 LiJ _J -J N -309 _l 100 < 1- I U 1- 50 . 1 1 1 1 1 Average lengt_h_at_b[rth III 1 1 1 1 I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 OCCLUDED POSTNATAL DENTINAL LAYERS (number) Figure ll. — Scatterplots of body length on number of postnatal dentinal layers from males (top) and females (bottom) otStenella attenuata from the offshore eastern Pacific. Lines are fit to the growth model (see text). 240 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE 220 200 180 160 100-/ 60 p ^ _^^i^^-— - -- ^ ^ ^^^ ^^ If i u - / Average length at birtti - 8 14 5 5 21 29 23 8 7 N (Cf) - 7 6 5 15 31 63 91 40 12 i 1 1 1 1 i N(J) 1 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 POSTNATAL DENTINAL LAYERS (number) Figure 12. — Fit of the double Laird growth model (see text) to 2-cm mean values of body length on number of postnatal dentinal layers for males and females of Stenella attenuata from the offshore eastern Pacific. For samples greater than 30, ± standard errors indicated as vertical line. For females with 6 or more layers, the growth equation is ' 0.0657 L = 159 exp 0.3707 [l - exp (- 0.3707U - 5.588))] In this case, both the growth rate and the rate of decay of growth are sharply lower than for juveniles. The fit for males is not as good (Figures 1 1, 12) as it is for females, probably due to greater variabil- ity and to inadequate sample sizes for the two oldest strata (the tooth-reading effort was concen- trated on females because of their importance in population dynamics). Another possible explana- tion for the relatively poorer fit for males is that growth (real, or as inferred from tooth layers) in adult males is more complex than in adult females, and a model more complex than the Laird model is called for. Inferred asymptotic length is 206 cm, achieved at predicted age of 26 layers. Average length of adult males (defined as those having testes weighing 200 g or more) is 200.7 cm, based on a sample of 253 (Perrin 1975). The largest male of 1,083 measured was 226 cm long. The growth equation for males with 6 or more layers is L = 159.5 exp 0.0524 , 0.2032 [l - exp (- 0.2032(^ - 5.588))] The secondary growth rate (a, 0.0524) is very slightly smaller than for females, but the rate of decay {a, 0.2032) is sharply smaller, reflecting the attainment of greater size in males. The equations rearranged and reduced for estimating age (in terms of layers) from length are ^(M and F <160 cm) = -1.394 In (7.531 - 1.48 In L) = 5.588 - 2.698 In (29.606 - 5.64 In L) = 5.588 -4.921 In (20.669 -3.878 In L). ^(F ^160 cm) ?(M^160cm) Note: These equations should not be used to esti- mate age from length except for grouped samples of smaller animals (about 180 cm or less), for which growth rate is still large compared to indi- vidual variation in length. The juvenile growth curve based on tooth layers can be calibrated for the first year by comparison with the growth curve derived from analysis of modal progression (above) and by deduction from what is known about juvenile growth of other odontocetes (the fetal-postnatal growth argument above). Estimated average length at 8 mo based on analysis of modal progression is 125.5 cm. The predicted number of layers at that length (Figure 12) is 1.53. If the average growth rate during the first year is assumed to be the same as the average during the first 8 mo, the predicted number of layers at 1 yr (1.53 • 12 - 8) is 2.3. This extrapola- tion, however, is a slight overestimate, because while growth during the first year in delphinids is approximately linear, there is some decay of rate. The predicted number of tooth layers (using Fig- ure 12) at 138 cm, the above-predicted length at 1 yr based on camparison with other odontocetes, is 2.0. It seems safe to assume that about 2 layers are laid down during the first year of life. Calibration of the remainder of the tooth-layer curve is more difficult. Kasuya et al. (1974) examined the innermost layer in teeth of S. at- tenuata and related type and thickness of layer to season of capture. They concluded that one layer (one transparent plus one opaque subunit) repre- sents 1 yr of growth. We found no correlation between thickness of the innermost layer and season of capture. Almost all of the samples for which teeth were sectioned, however, were col- lected in the first few months of the year. Lacking such direct calibration, several alternative pos- sibilities can be examined. The results, however. 241 must remain tentative and inconclusive until growth has been monitored directly in one or more captive or free-ranging, tagged individuals. Some alternatives that can be considered are: 1. Two layers per year until the teeth are occluded. 2. Two layers in the first year and one per year thereafter until the teeth are occluded. 3. Two layers per year until puberty (about nine layers in males and seven in females; see section below on age at puberty), and one per year thereafter. This list of alternatives can be extended to great length by making assumptions such as that layers are laid down at irregular intervals, males and females lay down layers at different rates, layers disappear with age, etc., but the above are proba- bly the main possibilities that should be con- sidered. All references below to age are in terms of layers, with the above alternative possibilities considered or implied. None of the alternatives can be eliminated with certainty. One tooth layer deposited per year has been inferred for the west- ern Pacific population of S . attenuata by Kasuya et al. (1974). One layer per year has also been sug- gested for other closely related delphinids, includ- ing S. coeruleoalba (Kasuya 1972) and Tursiops truncatus (Sergeant et al. 1973). Two tooth layers per year have been found in Delphinapterus leucas (Sergeant 1973), but this form is less closely re- lated to Stenella. Thus, there is more support in the literature for the one-layer-per-year model (number 2 above) than for the others. Length-Weight Relationships Length-weight relationships were determined for 218 fetuses, 66 postnatal males, and 33 nonpregnant, postnatal females by using linear regressions of log weight on log length. Fetuses The fetuses ranged from 20 to 897 mm long and weighed from 2 to 7,588 g. Ten fetuses less than 20 mm long were not included. The regression equa- tion is log W = 3.5532 + 2.501 logL, where W = weight in grams L = length in millimeters. 242 FISHERY BULLETIN: VOL. 74, NO. 2 In exponential form, the relationship is W = 2.79 X 10-4 L2-501. Females The females ranged from 100 to 200 cm and weighed from 12.0 to 69.1 kg. The regression equa- tion is log W = -4.1576 + 2.6120 log L, where W = weight in kilograms L = length in centimeters, or in exponen- tial form, W = 6.95 x 10"^ L^^'^. Males The males ranged from 86 to 218 cm and weighed from 6.8 to 90.0 kg. The regression equa- tion is log W = -4.7135 + 2.873 logL, where W = weight in kilograms L = length in centimeters, or in exponen- tial form, W = 1.93 x lO'^ L^-^^s The slopes of the regression equations are statistically different (^-test ata = 0.05) for males and females. Males are lighter for their length at birth, and heavier for their length after about 135 cm has been attained. Color Pattern Perrin (1970b) has previously described the de- velopment of the color pattern of S. attenuata in the offshore eastern Pacific. The animal begins life unspotted, develops dark spots ventrally that later coalesce, as light spots develop dorsally. The ontogenetic continuum can be divided into five stages as defined below and as shown in Figures 13 and 14: 1. Newborn stage. Dark purplish-gray dorsal surfaces and lateral blazes, with white ven- tral surfaces and no spots; about 80 to 160 cm. 2. Two-tone stage. General two-tone pattern with dark-gray surfaces above, lighter gray lower surfaces, and a well-defined pattern in varying shades of gray about the head and flippers; no spots; about 95 to 175 cm. PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE The division between this and the pre- vious category is somewhat subjective and arbitrary. 3. Speckled stage. Same as two- tone but with discrete, very dark-gray spots on the ven- tral surfaces; discrete light-gray spots on the upper, darker surfaces present on some animals but lacking on others; about 140 to 190 cm. 120- 110- 100 - 90 - 80- 70- 60- 50- 40- 30- 20- 10- o z 1x1 O dd FUSED (604) AVERAGE 30 20- 10 - : 50- 40- 30- 20- 10- 30 20 10 MOTTLED (138) AVERAGE _tj_ SPECKLED (255) 20 10 AVERAGE ' i . i>-i .1:1 2- TONE (288) NEONATAL (72) 1 I i 70 60 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 TOTAL LENGTH (cm) Figure 13. — Length-frequency distributions of males of Stenella attenuata from the offshore eastern Pacific, by color pattern phase. Mottled stage. Ventral spots converging and overlapping in places, but patches of the lighter gray background still visible, yield- ing a mottled effect; discrete or merging light-gray spots present on the upper sur- faces; about 155 to 210 cm. Fused stage. Ventral spots completely convergent, to give the effect of a uniform, medium-gray to dark-gray surface; on close inspection, the individual overlapping spots still discernible; about 160 to 230 cm. REPRODUCTION Seasonality Nishiwaki et al. (1965) suggested that S. at- tenuata in Japanese waters breeds in the spring and in the autumn. Harrison et al. (1972) stated that lengths of fetuses indicate that parturition in the eastern tropical Pacific (of S. graffmani = S. attenuata ) also occurs both in the autumn and in the spring. The postnatal length-frequency data for large samples (Figures 15, 16; April 1968 and October 1972, for example) support the thesis of major reproductive seasons in spring and au- tumn but also suggest that there is a reproduc- tive peak in summer as well. There is year-to- year variation in the timing of reproductive peaks, and there is some reproduction occurring throughout most of the year. It is difficult to define the reproductive seasons with precision be- cause most of the sampling effort was in the early (January- April) and late (October-December) parts of the calendar year. The sampling inter- sected obvious calving seasons in April 1968, January 1972, October 1972, January 1973, and June 1973 (Figures 15, 16). Calving peaks were probably also present in some of the other sam- pling months, but the samples were too small to detect them or were biased in some fashion. A summary of predicted birth dates for 373 fetuses more than 15 cm long collected in 1971, 1972, and 1973, however, indicates that there may have been three calving peaks in each of the 3 yr (Figure 17). In each year there was a definite calving low in winter. The synchrony was diffuse, and some peaks were much sharper than others. The statis- tical evidence for three annual peaks in calving is weak, and when the data for all years are com- bined, all that can be said with certainty is that the calving season is prolonged, with a low point in winter and a tendency for high points in spring and fall. The Male Sexual development of the male was examined under three criteria: 1) weight of testes, 2) aver- age diameter of seminiferous tubules, and 3) amoimt of sperm in the epididymis. Each of these was examined relative to total length, weight, and age (number of postnatal dentinal layers). Weight of the testes (Figure 18) increases pre- cipitously at body length of about 175 to 190 cm, 243 FISHERY BULLETIN: VOL. 74, NO. 2 3 3IOp 3CX) 290 280 270 260- 250- 240- 230- 220- 210- 200- 190- 180- 170- 160- 150- 140- 130- 120- 110- 100- 90- 80- 70- 60- 50- 40- 30 20- I 60- 50- 40 30- 20 10 60 50- 40- 30- 20- ?9 FUSED (g83) ' I ' ' I I I I I AVERAQE MOTTLED (194) '''!''■ . I SPECKLED (210) a^ m/BMBE I ''■ ' 1-^ I J Figure 14. — Length-frequency distributions of fe- males of Stenella attenuata from the offshore east- ern Pacific, by color pattern phases. On cruises be- tween January 1971 and October 1972, adult females (3^160 cm) were selected for measuring and dissection. Earlier and later samples were non- selective. Average lengths for neonatal, two-tone, and fused are based on all the samples (no length bias), and averages for speckled and mottled are based on the nonselective samples (shaded). The analyses of coloration transition are based on all the samples for neonatal-to-two-tone and on the nonselective samples for the remaining three transitions. Size of sample used for calculation of average is given in parentheses. tb m/fjmt . i i I t R-t-< 80 90 K)0 110 120 130 140 ISO 160 170 180 190 200 210 220 230 TOTAL LENGTH (cm) 244 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE AVERAGE LENGTH AT BIRTH AVERAGE LENGTH AT BIRTH > o 2 5 UJ O cr 5 5 5 5 5 15 10 5 NOVEMBER 1971 DECEMBER 1971 ' ' ' ^~i ''"' I JAN 1972 'r^!;^^^^ MARCH 1972 APRIL 1972 -^ — .n MAY 1972 r^l I I OCTOBER 1972 LiT_, 10 5 5 10 5 5 5 5 5- j I I I i_i ,, n, MAY 1973 OCTOBER 1973 r-l . . f—i I I 1 ) ) f— 1 -i 1 \ I I I NOVEMBER 1973 i~i i r-t ' i i I 1 L. DECEMBER 1973 EL -I l_l I L_i L. I I I I I""" I I I I I 75 100 125 150 175 200 225 250 75 LENGTH (cm) 100 125 150 175 200 225 250 LENGTH (cm) Figure 15. — Length-frequency distributions of postnatal male specimens ofStenella attenuata, 1968-73, by month. but in animals larger than about 200 cm, there is little correlation with length. The largest testes encountered weighed 2,400 g and were possessed by a male 196 cm long. However, some males more than 210 cm long had testes weighing less than 300 g. Testes weight begins to increase sharply at 50- to 55-kg body weight and is strongly correlated with weight in larger ani- mals. Males in the sample that weighed more than 70 kg (eight animals) had testes weighing more than a kilogram. The male with the third heaviest testes (2,017 g — heaviest testes for which body weight also available) weighed 80 kg; the heaviest male in the sample weighed 91 kg and had testes weighing 1,348 g. A rapid increase in testes weight (Figure 19) occurs at age 7 to 13 layers, with maximum size increasing until 12 to 16 layers. All animals with more than 14 layers had testes weighing 500 g or more. Again, there is wide variation in testes size relative to age. Part of the variation is ascribable to the consider- able error in the estimate of number of dentinal layers (±2 layers for animals with more than 5 to 12 layers, more for older), but it must be con- cluded that there is probably about a 5-layer period during which the onset of puberty may 245 FISHERY BULLETIN: VOL. 74, NO. 2 20 15 10 5 20 15 10 5 5 5 5 5 5 5 75 bj 65 O 60 UJ t!»-».*if» 1 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 LAYERS (number) Figure 19. — Scatterplot of testes weight on number of postnatal dentinal layers in Stenella attenuata. ~i. Z50 . • in - • UJ _i • . 13 CD 200 1 • • • rj _ 1- - O 150 '_ ••. *•► .• . • a: _ • •• n • Ul •• Ll • z • 2 100 _ • • UJ CO - • • . u. o - • • .• • • • •• • tr LU 1- 50 • • 1 .. .• ^w.* i • UJ S . < n 1 1 L- 1 1 1 1 1 1 100 120 140 160 ISO BODY LENGTH (cm) 220 Figure 20. — Scatterplot of average diameter of seminiferous tubules on body length in Stenella attenuata. 3. 250 ~ to UJ _J D m 200 _ t • • • 1 t- to • • • • • • t • 1 • • • • 3 O 150 L •> • • • • CE • • UJ • • Lj. • • 2 S 100 _ UJ • • • • U- o 1 • • • • 1 • • • t UJ 50 • t • • ■T •.• • • • • • • • • • • • • 1 t UJ S < 1 1 1 1 1 1 J- -L -1. _L J_ 1 1 1 1 1 Q I 2 3 4 5 6 7 8 9 10 M 12 13 14 15 16 17 IB 19 LAYERS (number) Figure 21. — Scatterplot of average diameter of seminiferous tubules on number of postnatal dentinal layers in Stenella attenuata. 248 Sperm in the epididymis were scored as "ab- sent," "present in small numbers," or "copious" (easily seen in the histological sample without searching). The shortest individual with large numbers of sperm in the epididymis was 179 cm long. This animal weighed 62 kg. In animals larger than 180 cm and heavier than 58 kg, pres- ence or absence of sperm in the epididymis bears little relationship to total length. Thirty-six large adults (>200 cm) were equally distributed among the three categories of no sperm, some sperm, and copious sperm. The smallest testes bearing epididymis with sperm weighed 200 g, and the smallest testes with copious sperm weighed about twice as much. Some animals with testes heavier than 1.5 kg, however, had no sperm in the epididymis. The same pattern of wide variation is apparent in the relationship between epididymis code and layers. The youngest male with sperm in the epididymis had 9 layers. The youngest animal with copious sperm had 10 layers. After about 10 layers, there appears to be no relationship between age and presence or absence of large numbers of sperm. In summary, the onset of puberty, as indicated by a rapid increase in diameter of seminiferous tubules and increase in testes weight, is at 7 to 12 layers (average ~-9 layers; an estimate of ages at puberty) and at lengths of 155 to 170 cm and weights of 40 to 50 kg. Sexual maturity is at- tained about 2 to 4 layers later, at 10 to 14 layers, ^180 cm, and &58 kg. The midpoint of the range of 10 to 14 layers, or 12 layers, may be taken as an approximation of average age at attainment of sexual maturity. Whether or not males at this point are "socially mature" (sense of Best 1969) can be determined only through behavioral studies. Average length of males 12 layers old is about 195 cm, and average weight is about 75 kg. The Female Attainment of Sexual Maturity Harrison et al. (1972) described and figured the ovaries of S. attenuata (as S. graffmani). The ovaries weigh less than 0.5 g each at birth. Weight increases gradually to about 1.5 g at about age 6 to 8 layers (average ~7 layers; an estimate of age at puberty), when there is a sud- den increase in average ovary size and weight due to presence of corpora of ovulation and/or pregnancy (Figure 22). This change comes at an PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE 16- X 10 C2 LlJ 8 to UJ cr % ^ o - o - o o " o - «. o°° ° ° o ° o ° ° ° o ° - o ° 8 = • ° c^." 1 8 T° o ° • o 8 8 °o8 ° too • • • o • • • • ° - • .'*?.i h:.* : • • • •• • _ • r • f. -th . • • , i ' • • . ► • 1 i , 1 , 1 , i — 1 — . 1 i i 1 i 1 1 2 4 6 8 10 12 14 16 18 LAYERS (no) Figure 22. — Relationship between weight of ovaries and number of postnatal dentinal layers in Stenella attenuata. Open dots are females with a corpus luteum. average total length of about 170 to 180 cm and weight of 50 to 60 kg (Figure 23). Analysis of lengths of females with and without ovarian scars yields a more precise estimate of length at attainment of sexual maturity. The smallest of 1,410 specimens (160 cm long or longer) that possessed scars were two that were 167 cm long (one with a corpus luteum only and one wdth a corpus luteum and four corpora al- bicantia). The largest female with no scars was 193 cm long. The length-maturity curve is slightly asymmetrical, but a linear regression line through the nearly linear central portion (M = 5.76L - 960.95) estimates that average length at which scarring is first evident is 175.4 cm. This analysis probably underestimates length at at- tainment of maturity, because some of the small adults (170 to 180 cm) with many scars are those that have stopped growing at a shorter-than- average length. In other words, the left-hand por- tion of the frequency distribution of physically mature adults to an unknown extent artificially elevates the central portion of the length- maturity curve, making it asymmetrical. An estimate of age and length at attainment of sexual maturity can also be derived directly from the smaller sample of females for which the number of tooth layers was determined. The youngest specimen exhibiting ovarian scarring had 7.5 layers. The oldest with no scarring had 12 layers. The estimated age at which 50% have scars is 9.14 layers (M = 19.5^ - 128.25). Pre- dicted length at this age is 181.6 cm (based on growth equation above). This estimate is less biased than the others above but based on much smaller samples, especially at the lower end of the layer-maturity curve. Another estimate of length and age at first ovu- lation can be made by back extrapolation of a relationship between body length and number of corpora (including corpora lutea) in the ovaries (Figure 24). Length increases with corpora count until at least six to eight corpora have been ac- cumulated, at about 183 to 190 cm. A fit of the data to the Laird growth model (above) yields the equation L = 180.17 cm exp{0.0541[l - exp(-0.2815C)]}, where L = length in centimeters C = number of corpora. Back extrapolation of the curve to zero corpora yields an estimate of 180.2 cm. Predicted age from the growth equation is 8.74 layers. An estimate of length at first conception can be made by calculating the average length of preg- nant females with a corpus luteum only (indicat- ing first pregnancy) and subtracting the growth that they can be assumed to have undergone dur- ing pregnancy. Fifty-four primiparous females averaged 181.7 cm in length (range 167 to 193 cm). Predicted age at that length is 9.17 layers. The average length of their fetuses was 372 mm. This length is attained by about the beginning of the sixth month of gestation. Using the growth equa- tions above to predict growth during 6 mo for the various tooth-layering models and substracting the growth increment from 181.7 cm yields esti- mates of length at first conception ranging from 177.7 to 180.0 cm (number 4 in Table 2). The primiparous females in this sample, however, are only those that became pregnant at the first ovu- lation. This may cause the estimate to be an underestimate, because many females ovulate several times, and presumably continue to grow, before becoming pregnant the first time (see Ovarian Changes below). The various methods of estimating age and length at attainment of sexual maturity yield es- timates of varying accuracy (Table 2). The esti- mates based on tooth layers and length at first 249 24 22 20- — 16 en E 1^ LJ ^ 12 CO LU CC 10 < > o 8 6- 4- 2 - FISHERY BULLETIN: VOL. 74, NO. 2 126 •I 90 100 110 120 130 140 150 160 170 BODY LENGTH (cm) 80 190 200 210 Figure 23. — Scatterplot of weight of both ovaries on body length in Stenella attenuata. Open dots represent females with a corpus luteum. For animals 160 cm or longer, where sample for 5-cm interval is 10 or more, means (circled symbols) and ranges are graphed and points are not plotted. Where the sample is &30, ± two standard errors are indicated by bars. conception are the best of the four and probably bracket the true values. Under method number 3, age hypotheses numbers II and III are more prob- ably correct than number I. Accordingly, we esti- mate that sexual maturity is, on the average, at- tained at 181 ± 1 cm and 9.0 (8.6 to 9.3) layers (5.1 to 8.3 yr, depending on the alternative layering hypothesis used). An increase in size of Graafian follicles is another criterion of approaching sexual maturity. Diameter of the largest follicle also shows a sharp increase after 160 cm total length (Figure 25), con- 250 current wdth the increase in ovary weight (Figure 23). The largest follicle in immature females usu- ally is less than 1 mm in diameter. The largest follicles in most ovaries containing scars are be- tween 1 and 8 mm in diameter, but a few follicles (possibly cystic) as large as 10 to 16 mm in diame- ter were encountered. Ovarian Changes in Adults The analyses of ovarian changes are based on material collected through 1972. The corpus PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE 2IOr E e 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 CORPORA (number) FIGURE 24. — Relationship between body length and number of corpora in Stenella attenuata. Average (bar), ± two standard errors (box), range (vertical line), and sample size shown. Table 2. — Results of analyses of length and age at attainment of sexual maturity in Stenella attenuata, with comments (in parentheses) on pros and cons of the methods. Lengths and layer counts predicted with the growth equations are in parentheses. Analysis Length (cm) Layers (no.) Age (yr) under hypoth 1 II esis III 1 Length at which 50% have cor- pora (probable underestimate). 175.4 (7.66) 38 6.7 43 2. Number ot tooth layers at which 50% have corpora (interpolation, but small sample sizes). (181.6) 9.14 4.6 8.1 5.6 3 Back-extrapolation of corpora- length cun/e (large samples, but extrapolation). 180.2 (8.74) 4.4 7.7 5.2 4. Length at first conception under hypothesis: 1 II III (includes only those that become pregnant at first ovulation; prob- able underestimate). 177.7 180.0 180.0 (8.17) (8.57) (8.57) (4.1) 7.6 5.1 luteum of pregnancy arises from the ruptured fol- licle and has an important secretory function in maintaining early pregnancy in all mammals and full gestation in most (Amoroso and Finn 1962). The gross and microscopic structures of corpora lutea in various delphinids, including S. attenuata, have been described by Harrison et al. (1972). The corpus luteum decreases in size during ges- tation (Figure 26). Of 242 females with corpora lutea, 229 were pregnant. Eleven with fetuses less than 20 mm long (range 1 to 20 mm) had UJ _J o o LJ O < Ll_ O cc LJ H LlI < 140 150 160 170 ISO 190 BODY LENGTH (cm] 200 210 Figure 25. — Relationship between body length and diameter of the largest Graafian follicle in Stenella attenuata. Open dots represent females with corpus luteum. For length ^160 and n 5=10, means (circled symbols) and ranges shown. Forn 3=30, ± two standard errors are shown. Not included are 27 "senile" specimens with follicles <0, 1 nmi and five juveniles 88 to 122 cm with 0- to 1-mm follicles. corpora with diameters of 23 to 29 mm (average 26.0 mm, SD 2.90). The mean diameter dropped sharply to 23.6 mm (range 21 to 27 mm, SD 2.27) in 17 females with fetuses between 20 and 100 mm (using Student's t, means are significantly different at a = 0.01). This amounts to about a 32% decrease in luteal volume. Size of the corpus luteum continues to decrease at a slower rate, to 22.2 mm (range 19 to 28 mm, SD 1.79) in females with fetuses 700 to 825 mm (average length at birth is 825 mm) long, a further decrease in vol- ume of about 15%. Luteal volume in females with near-term fetuses is only about half of that shortly after conception. Mean diameter in 10 females with fetuses longer than average birth length (825 mm) was 24.0 mm (range 20 to 26 mm, SD 2.21, greater than mean for 700 to 825 mm at a = 0.01), a volume difference of about 38% more than for fetuses 700 to 825 mm long. Delayed re- gression (or re-enlargement) of the corpus luteum is apparently correlated with greater-than- average length at birth. 251 FISHERY BULLETIN: VOL. 74, NO. 2 o ro 1 1 1 1 1 1 1 1 1 1 1 1 1 1 !£2 Q in 1 1 1 1 1 1 1 1 1 1 1 1 1 t c 5 COS h5 to II < 33 OQ-LlI •"ogoSS^g^oSScxft^, o (VJ II • iijt/5 ^^l- OZ) : • t UJ< liJ Q. o_i oo o o o 1- II O / ^ o I- e 1 UTERI NOT DISTEN < Q !-5 • o < > < — o UJ h- . t .••• 3 _J • • * • 1 , . • • • — o -p O c < a: • • • • ' • ® ® • : o => o • • — o ^- a. o • « 1 • • • ' • • lO UJ u. o t . _ O O o O 2 • t ® ® •: i ^ X 1- •® l-n. z: < O UJ > -J — O _l -! Q ro 2 z o < . •! 1- t- Z 2 • r^ < < 9 2 Z — O o o OJ LlI LiJ ^^ cr q: Q. CL ® • • . r- II II o • ® — g • * t ® • : ® • t N / r?^-«t-' o- 1 rTii|riii|iii 1 1 1 1 1 1 1 1 1 1 1 1 1 1 —^ o ro in C\J o CM lO o CJ lO (ujuj) sndaoo dO a3i3kMvia 3 o G 01 o T3 o o u .4-) C CO u 0) 13 C o o c til 1^ ca E C T3 C CO M C CO C CO & 01 o. _3 tn 3 P. !-• o 5 ■5 0) CO u V > < CD w D O 252 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE Nine obviously postpartum females had cor- pora lutea 13 to 25 mm in diameter (Figure 26). Four lactating females with uteri not obviously distended had corpora lutea in the same size range. Average luteal volume in these lactating animals was less than half of that in animals at parturition. Some of these 13 cases may represent miscarriages. The corpus luteum of pregnancy shrinks still further during the suckling period, losing its glandular appearance and becoming a corpus al- bicans. Nine of 197 lactating females without cor- pora lutea each had a single corpus albicans, which must represent the regressed corpus luteum of the first pregnancy. These corpora (Fig- ure 26) were approximately spherical and 5.9 to 10.6 mm in diameter (average 8.5 mm). The lower end of this range — about 6 mm — must approx- imate the limit of regression during the suckling period (about 11.2 mo; see below). The small number of lactating females with corpora lutea (13) compared to the number with only copora al- bicantia (197) indicates that initial regression fol- lowing parturition must be very rapid, perhaps occurring in less than 15 days. Still further re- gression in size and histological structure of the corpus albicans of pregnancy probably occurs. Many adult females have a large corpus albicans (in most cases, one of several) between 3 and 6 mm in diameter (Figure 26) with greatly degen- erated structure. Unless these corpora all repre- sent ovarian events not resulting in pregnancy, i.e., the females are all completely barren, the corpus albicans of pregnancy must decrease in diameter during a resting period following a pregnancy, to possibly as little as 3 mm. Multiple corpora lutea are uncommon in S. at- tenuata. They were encountered in only 2 out of 258 females with corpora lutea. One of these was pregnant with twin fetuses (males, 83 and 86 mm) in the left horn of the uterus. The left ovary contained two corpora lutea of approximately equal size, each possessing a surface scar of ovu- lation, together with seven corpora albicantia vis- ible on the surface. The right ovary was devoid of scars. Another female with two corpora lutea had a 592-mm fetus (male) in the left horn of the uterus. The left ovary looked very much like that of the specimen with twin fetuses, having two corpora lutea of approximately equal size and eight corpora albicantia on the surface. Neither corpus luteum bore a discernible surface scar. The right ovary was unscarred. There are two possible explanations for the presence of two corpora lutea in this specimen: 1) one of them was an accessory corpus, or 2) one of a pair of twin fetuses was aborted during early pregnancy. In any case, the incidence of multiple corpora lutea is very low in S. attenuata, less than 1% in the sample examined. This is in sharp contrast to some other cetaceans, in which rates of presence of accessory corpora range to 15.6% iDelphinapterus leucas — Brodie 1972). The contribution of double and accessory corpora lutea to the accumulation of corpora albicantia can be considered to be neg- ligible in S. attenuata. Corpora albicantia in S. attenuata represent both regressed corpora lutea of pregnancy and re- gressed corpora of ovulations that do not result in pregnancy. This conclusion is based on the ac- cumulation rate of corpora albicantia and on the estimate of the mean length of the calving inter- val (see below). We were not, however, able to differentiate between small regressed corpora lutea and regressed corpora of ovulation. This impasse, also encountered by workers dealing with other cetaceans (Harrison et al. 1972) is caused by the wide and largely discordant varia- tion in size, shape, surface texture, and internal structure and color of the corpora albicantia. If one looks at enough corpora, it is possible to find corpora with these characters in almost any com- bination of expressions. Harrison et al. (1972) found no more than six corpora albicantia in the ovaries of any Stenella female. In the present sample, however, nearly half (44%) of the females had more than six cor- pora, including the corpus luteum. Fifty-five females of 1,131 had 15 or more corpora; one had 28 (Figure 27). Three thousand five hundred and two corpora from ovaries of 530 females were scored to six categories. These categories are somewhat arbitrary in view of the continuity of regression and the wide variation discussed above, but, nonetheless, they are useful in analyzing the course of regression. The numbers and proportion of total corpora complement rep- resented by each of these categories varies with the total number of corpora (Table 3, Figure 28). The categories were defined as follows: Type 1. Surface raised, smooth or slightly wrinkled. Looks externally like a small corpus luteum. Cortex white or yellow, with obvious remnants of vascularization. Center solid or loosely constructed, consisting mainly of white 253 FISHERY BULLETIN: VOL. 74, NO. 2 120 r- 100 >- o 3 O a: 80 60 40 20 ±1. Th^ I 5 10 15 20 TOTAL CORPORA (number] 25 30 Figure 27. — Frequency distribution of corpora count in 1,131 females ofStenella attenuata. connective tissue, 3.5 to 15.5 mm in diameter, av- erage 7 mm. These corpora almost certainly are nearly all regressed corpora lutea. Four hundred fifty-six were encountered (13.29f ). Females w^ith two or more corpora have, on the average, about one Type 1 corpus (Figure 28), although as many as five may be present (Table 3). Table 3. — Types of corpora present in ovaries ofS. attenuata in relation to total corpora. Averages in Figure 28. Total number Range of number of each tvoe of of corpora (including corpora lutea) Sample size corpus albica ns— Type; (no.) 1 2 3 4 5 6 1 35 0-1 0-1 2 42 0-2 0-2 0-1 0-1 0-1 3 48 0-3 0-3 0-2 0-1 4 53 0-3 0-4 0-3 0-1 0-2 5 49 0-5 0-4 0-5 0-2 0-2 6 49 0-5 0-4 0-6 0-3 0-1 7 50 0-5 0-5 0-6 0-3 0-2 0-1 8 46 0-5 0-6 0-7 0-2 0-4 9 36 0-2 0-5 2-9 0-4 0-3 0-1 10-11 48 0-4 0-5 3-10 0-2 0-2 0-1 (average 10,4) 12-14 32 0-2 0-5 3-14 0-3 0-4 0-3 (average 12.9) 15-27 25 0-4 0-5 7-19 0-2 0-9 0-1 (average 17.2) Total 513 0-5 0-6 0-19 0-4 0-9 0-3 Type 2. Surface raised and wrinkled. Interior white to yellow, often v^dth traces of luteal cortex and vascularization. Center solid or loosely con- structed, consisting mainly of white connective tissue. Definitely less integrated in structure than Type 1 (above). Diameter 3.0 to 12.0 mm, average 6 mm. The evidence on accumulation rate (below) suggests that these corpora are prob- ably a mixture of regressed corpora lutea and corpora of ovulation. We found 787 of this type (22.5%). The number of Type 2 corpora is rela- tively constant in females with three or more cor- pora, at about one and one-half (Figure 28) vdth a ^ 12 c • • • • • • • 1- • •• t. • • • • =C 6 • • •o • O • • • • UJ 5 _ • • •• • h ? T J. : • • • • t • i • • • • • • •• • • • • • (f) • I • M • • UJ r • • • > o 3 • _ • • - : • • •• t •• Z • • 1 • • • • • • • • •o • • • • • • • : • • • t 1 1 •• • 1 • • • • • • • • • •• •» • • o • o • o • • • < • • • • • • • • •• • • 8. o • o o • - • 2 1 n 1 1 1 1 L. • 1 1 —L- o o _L_ l_ • o 1 o o 1 1 I 1 1 1 1 1 1 1 1 1 1 I 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 CORPORA (number) Figure 30. — Scatterplot of ovaries weight on number of corpora in Stenella attenuata. Females with corpus luteum not included. Open dots are females with no Type 1 or 2 corpora albicantia. 256 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE I7r- 14 ^ 12 E E UJ o II 10 •- 9 CO UJ O (T < 8 tr UJ I- UJ 6 < 4- 2- <05 • X •X o o XX X X J I I I ± ± o oo J L •o I • cX •• J I I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 CORPORA (number) Figure 31. — Scatterplot of diameter of largest follicle on number of corpora inStenella attenuata. Females with corpus luteum in- dicated with X . Open dots are females with nor corpus luteum or Type 1 or 2 corpora albicantia. 257 FISHERY BULLETIN: VOL. 74, NO. 2 albicantia. They also typically had very small Graafian follicles (Figure 31). A consideration of these females bears on the question of persistence of corpora albicantia. Sergeant (1962) encoun- tered similar females in Globicephala. They com- prised about 5% of adult females. He called them "senile" and concluded that the ovarian scars in these animals represent some residual subset of the maximum complement of corpora of preg- nancy, ovulation, and other events. He implied that they probably are the corpora of pregnancy, since those corpora are larger at the outset and presumably less likely to regress to the point of macroscopic disappearance. The ovarian data for S. attenuata do not support this hypothesis of dis- appearance of some corpora in regressed ovaries. The regressed ovaries have 10 to 15 corpora (Fig- ures 30 and 31). The ovaries of other, still repro- ductive females are larger and have 16 to 27 cor- pora, although follicles are typically smaller than in reproductive females w^ith fewer corpora (Fig- ure 31). Three alternative hypotheses explain this apparent dichotomy in females with 10 or more corpora: 1. The usual maximum number of corpora is about 21, and some corpora disappear in postreproductive females, i.e., the "senile" group in Figure 30 properly belongs at the far right side of the plot at the end of a downward trend in ovary weight (the hy- pothesis of Sergeant 1962). 2. Corpora are laid down at about the same rate in all individuals, but some become postreproductive at about 10 to 15 corpora while others continue to accumulate corpora (16 to 27) until a greater age, i.e., the corpo- ra scale in Figure 30 is effectively an age scale. Under this hjrpothesis, corpora do not disappear. 3. Corpora are accumulated at rates varying widely among individuals, but the typical maximum complement is 10 to 15 corpora, i.e., the reproductive females with more than 15 corpora in Figure 30 properly belong in the body of the distribution in the left two-thirds of the plot. A possible explana- tion for widely varying rates of accumula- tion is that some females are more fecund and the senile period is reached with some maximum number of pregnancies, so that the varying ratios of corpora of pregnancy to corpora of ovulation may produce the appar- ent dichotomy. Sergeant (1973) found greatly varying individual rates of ovulation in the white whale, Delphinapterus leucas. In order to examine these alternative hypoth- eses, the females in Figure 30 and 31 with 10 or more corpora were examined in three groups — A, B, and C: A. 10 to 15 corpora, reproductively active (corpus luteum and/or Types 1 and 2 corpora albicantia). B. 16 or more corpora, reproductively active. C. 10 to 15 corpora, postreproductive (ovaries regressed, no corpus luteum or Types 1 or 2 corpora albicantia). The three groups were compared in terms of corpora count, weight of ovaries, size of largest follicle, number of dentinal layers, total length, and relative corpora counts in left and right ovaries (Table 5). Only nonpregnant females were included in the sample for ovary weight. Follicle size was examined separately for pregnant and nonpregnant animals. Ovary weight and follicle size for nonpregnant animals decline progressively from A to C. This is Table 5. — Characteristics of females of Stenella attenuata in groups A, B, and C (see text). Item A B C Corpora (no.) Sample size 67 24 15 Average 11.2 189 12.9 Range 10-15 16-27 10-15 SD 1.41 256 — Ovary weight (g) (nonpregnant) Sample size 44 13 15 Average 4.4 3.6 2.2 Range 2.0-8.5 2.2-48 1,0-3,1 SD 1.61 0.81 0,59 Largest follicle (mm) (nonpregnant) Sample size 27 13 <14 Average 2.9 1.5 <0.5 Range <0.5-10.3 <0. 5-8.0 0.5-4.3 SD 2.53 2.35 — Layers (no) Sample size 30 18 7 Average 13.1 13.1 13.2 Range 10.0-16.0 11.0-15,0 11.5-16.0 SD 1.39 1.31 1.52 Length (cm) Sample size 67 24 15 Average 190,1 190 3 187.0 Range 172-202 1 77-204 179-192 SD 6.43 6,78 3,54 Left/right ovary Sample size 65 23 15 Average in right (%) 24 33 29 Lett/nght (no./no.) 548/178 291/144 — 258 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE a requirement of hypothesis 1, above, but does not eliminate hypotheses 2 and 3. The three groups do not differ in average esti- mated number of tooth layers. This may, in part, be due to the difficulty of accurately counting the innermost layers in teeth with more than 12 layers (the number of layers is probably undere- stimated by as much as one-third in teeth with large amounts of convoluted secondary dentine), but careful comparison of the teeth of the three groups in terms of other features presumably cor- related with age, such as tip wear, degree of clo- sure of the pulp cavity, and amount of secondary dentine does not indicate that any group is older than any other. This evidence is against hypothesis 1, which requires that group C be older than A, and hypothesis 2, which requires that B be older than A and C. Groups A and B have reached asymptotic length (-190 cm). The animals in group C aver- aged about 3 cm less. A statistical comparison of A with B using Student's t indicates that the dif- ference is significant at a = 0.05. These results do not eliminate or support directly any of the hypotheses. Since A, B, and C are about the same age, the length data indicate that asymptotic length may be less for females that become senile with 10 to 15 corpora. This indirectly supports the idea of considerable individual variation in life history. The most convincing evidence against hy- pothesis 1 has to do with number of corpora in right versus left ovaries. If the emphasis in cor- pus deposition shifts from left to right at about 10 corpora, and if group C regresses from group B (animals with about 20 corpora) losing about 6 corpora in the process, then group C should have about equal numbers of corpora in the right and left ovaries. If most corpora of ovulation come in early reproductive life (as data analyzed below indicate) and, as suggested by Sergeant (1962), are more likely to disappear than corpora of preg- nancy because of smaller initial size, then the regressed group C should have, on the average, more corpora on the right than on the left, be- cause most of the corpora of preg- nancy would be in the right ovary. Forty-one per- cent of the corpora in 14 individuals having 18 to 22 corpora (average 19) were on the right. Only 2^c of the corpora in group C were on the right. The difference between C and A (29 and 24% ) can be accounted for simply by the difference in aver- age total corpora count (12.9 and 11.2). These re- sults eliminate the hypothesis (number 1 above) of loss of corpora with regression of ovaries. The various lines of evidence largely speak against hypotheses 1 and 2 and support hy- pothesis 3, that of great individual variation in life history and of persistence of corpora albican- tia. This is in line with findings by some other workers in small cetaceans (Sergeant 1962, 1973; Brodiel971). The data on the relationship of percent occur- rence of corpora lutea to number of corpora (Fig- ure 28) also support the hypothesis of widely varying rate of accumulation of corpora albican- tia. After stabilization at about 50% at 3 to 4 cor- pora, the rate declines after 8 to 9 corpora to 20% at 13 corpora; but the rate for females with 17 to 27 corpora is again 50%. Assuming that fecundity is inversely related to age, this pattern suggests that the females in the 17 to 27 group are about the same age as those in the 3 to 9 group. Ovulation Rate Even assuming that corpora albicantia persist and represent various ovarian events, estimating average rates of accumulation is difficult because of 1) the above-mentioned unreliability of age estimates based on more than 12 tooth layers, 2) the evident individual variation in accumulation rate, and 3) change in ovulation rate during the reproductive span. All of these factors must con- tribute to the scatter in a plot of corpora number (including corpus luteum) on estimated age (Fig- ure 32). Several workers have pointed out that cetacean ovaries often contain two or more cor- pora of the same size and same stage of regres- sion. It has been suggested that these are the result of multiple infertile ovulations or luteali- zation of atretic follicles in newly mature ani- mals (Harrison et al. 1972). Many in the present series of ovaries had two or more corpora (of Type 1 or 2) that were very similar in size and struc- ture and must have resulted from nearly contem- poraneous events. One probable multiple ovula- tion is apparent in Figure 32. This female, field number CW0R8, possessed 7 or 8 well-defined layers in its teeth. In spite of its extreme youth, it had a small corpus luteum, three Type 1 corpora, two Type 2 corpora, one Type 3 corpus, and one Type 4 corpus. The uterus was empty, and there was no milk in the mammaries. The animal could not have been reproductively active for more than about a year, but had already experienced eight 259 e 3 C FISHERY BULLETIN; VOL. 74, NO. 2 27 25 20 15- < cr o Q- cr O 10 o ® im ® ®» «) • • • ®M ••• • • ® 5 - ••• • A •• • • ••••• 10 -! \ r- I 12 13 14 LAYERS (number) 15 T 16 17 OCCLUDED Figure 32. — Scatterplot of number of corpora on number of postnatal dentinal layers in Stenella attenuata. Circled s3rmbols are senile females [shriveled oveiries with no corpus luteum or Type 1 or 2 corpora albicantia]. apparently nonreproductive ovarian events that resulted in corpora belonging to all of the t5^es through which a corpus luteum must pass during regression to a small corpus albicans. Calculation of average ovulation rates from the data in Figure 32 must take into account indi- vidual variation in age at first ovulation. The fe- males in Figure 32 were grouped into 10 one- layer intervals beginning with 7.5 layers (Table 6). The average reproductive age in interval p was calculated as A = [ 2 a,&< J ^ c, where a, = % maturing in iih. interval (% mature in i minus % mature in i - 1) b, = average reproductive age in interval p of females mature in i c = % mature in interval p. Average reproductive age in the ith interval of 260 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE TABLE 6. — Average reproductive ages and corpora counts of females oiStenella attenuata used in estimating ovulation rate based on corpora and tooth layers. Average Proportion reprodu ctive Average Layers Sample mature age of mature corpora (no.) (no.) (°o) (laye rs) (no. ) 7.5- 8 13 46.2 0.50 1 4.50 8.5- 9 18 44.4 1.56 \ 1.43 3.25 4.21 9.5-10 24 79.2 1.67 ) 4.53 ) 10.5-11 25 84.0 2.56 6.42 11.5-12 52 94.2 3.25 8.35 12.5-13 36 100.0 4.05 8.92 13.5-14 31 100.0 5.06 8.71 14.5-15 15 100.0 6.07 ) 10.87 1 15.5-16 7 100.0 7.08 \ 6.60 9.86 \ 10.45 >16 3 100.0 8.09 ) 9.75 ) Total 224 Note: Teeth of all available females with more than 12 corpora were sec- tioned, while only a nonselective subsample of females with fewer corpora were included. The effect on estimate of average reproductive age is negligible, since nearly all had 11 or more layers. females maturing in i was set at 0.50 layers. Be- cause of small sample sizes, the first three inter- vals and the last three were pooled. The results show an increase in average corpora count (number of ovulations) with reproductive age (Figure 33). A curvilinear fit to the interval means, using a power model forced through the origin, fits well and indicates that ovulation rate is higher in animals of reproductive age 0-2 layers than in older animals. The breaking point seems to come at about 12 layers, when about 6 corpora have been accumulated and rate appears to become nearly constant. Average ovulation rates estimated from the curve are about four during the first layer, two during the second, and about one per layer thereafter. 12 3 4 5 6 7 AVERAGE REPRODUCTIVE AGE (layers) Figure 33. — Relationship between average number of corpora and average reproductive age (in layers) in Stenella attenuata. Calving Interval The pattern of reproduction definable with the methods used here consists of three phases: preg- nancy, lactation, and a period of inactivity and/or estrus called here "resting/estrus." The length of pregnancy was estimated above as 11.5 ± 0.2 mo. We estimated length of lactation in three ways, based on 1) stomach contents of calves, 2) num- bers of lactating females and calves, and 3) ratio between numbers of lactating and pregnant females. The forestomachs of 45 calves less than 150 cm long were opened and examined by eye for pres- ence of milk. Twenty-one were empty. The stomachs of four calves 120 to 130 cm long con- tained both milk and solid food (fish and/or squid). Stomachs of 8 smaller calves (80 to 115 cm) contained only milk, and 12 of the larger calves (130 to 150 cm) apparently contained only solid food. About 130 cm appears to be the length at which effective weaning occurs. The estimated time required to grow to 130 cm is 9.4 mo (based on growth curve above). This estimate is not very reliable for two reasons: the sample is small, and small amounts of milk could be present and undetectable by eye, i.e., suckling could continue at a low level after the effective shift to solid food. The estimate can, however, be considered to be a probable lower bound on length of lactation. A second estimate is based on the assumptions that 1) a suckling calf exists for each lactating female and 2) the samples of specimens are un- biased with respect to suckling calves and lactat- ing females. Given these assumptions, the length at which the cumulative frequency of calves in a sample equals the number of lactating females should be the average length at weaning. This length in eight variously sized, 1-mo "random" samples of calves and females ranged from 125 to 145 cm (Table 7). The aggregate estimate for the eight samples pooled (320 lactating females) is 137 cm. Average age at 137 cm is estimated at 1.94 tooth layers, or (assuming two layers ac- cumulated during first year) 11.6 mo. If calves were overrepresented in the samples, this would be an underestimate. If they were underrepre- sented, it would be an overestimate. It would be an overestimate if the assumption that the number of lactating females equals the number of nursing calves were not valid. The assumption is not valid if the mortality of nursing calves is 261 FISHERY BULLETIN: VOL. 74, NO, 2 Table 7. — Length at which cumulative frequency of calves equals the number of lactating females in eight 1-mo samples of Stenella attenuata. Lactating Length (cm) at which cumulative Sample females frequency of calves = no (mo) (no.) lactating females Oct. 1972 51 132 Jan. 1973 65 125 Feb. 1973 50 144 Mar. 1973 48 136 Apr. 1973 13 142 May 1973 32 142 June 1973 18 145 Nov. 1973 43 142 (Oct. 28-Dec. 1 1 ) Total 320 137 higher than that of lactating females and lacta- tion continues after death of a nursing calf. A third estimate of length of lactation was de- rived from the ratio of lactating to pregnant females. This analysis included all the material from 1971 and 1972, when only adult females were sampled, as well as the material included in the calf-lactating female analysis above. Females both lactating and pregnant were included in both categories. The assumption is made that samples were unbiased with respect to relative representativeness for lactating and pregnant females. The ratio was 0.95 in the 1971 sample (86 adult females), 1.00 in 1972 (455), 0.96 in 1973 (573), and 0.97 for the pooled samples {n = 1,114; Table 8). The ratio of lactating to pregnant should equal the ratio of the lactation period to the ges- tation period. Gestation is 11.5 mo, therefore lac- tation is by this method estimated at 0.97 times 11.5 mo, or 11.2 mo. Estimated length at this age is 135.5 cm. The three estimates of 9.6, 11.6, and 11.2 mo are based on largely independent assumptions and are close enough to each other to indicate that length of lactation is almost certainly between 9 and 12 mo. Of the three, the central estimate, 11.2 mo, is best in terms of sample size and probable validity of assumptions and is used below in es- timating length of the calving interval. The basic data used for estimating average length of calving interval were the relative fre- quencies of adult females in several reproductive conditions (Table 8). Adult females were defined as those wath at least one corpus luteum or corpus albicans. Senile females were those with 10 or more corpora albicantia, no corpus luteum or Type 1 or 2 corpora albicantia and ovaries weigh- ing less than 3.5 g. Resting/estrus females were those nonsenile adults that were neither preg- nant nor lactating. Many of these (16 to 31%) had a corpus luteum. The corpus luteum may have represented an undetected very early pregnancy, a very recently aborted pregnancy, loss of a calf shortly after birth (resulting in cessation of lacta- tion), or may have been a corpus luteum of ovula- tion. All of these alternatives may be represented in the samples. In calculating the proportions of females in the three phases of pregnant, lactating, and resting (Table 9), senile females were excluded. One-half of the animals simultaneously pregnant and lac- tating were assigned to the pregnant category and one-half to the lactating category. The average length of calving interval was es- timated by two methods — 1) using the estimates of gestation and lactation periods and 2) using the percentage of females pregnant. The data for the 3 yr are comparable (Table 9), with the exception of possible existence of a trend in proportion rest- ing; therefore, length of calving interval was es- timated from the pooled data. Eighty-four and one-half percent of reproductive females were pregnant or lactating. Pregnancy (11.5 mo) plus lactation (11.2 mo) total 22.7 mo. If the proportion in a phase is equal to the proportion of the total Table 8. — Reproductive condition of 1,114 adult female specimens o{ Stenella attenuata, collected 1973.' 1971 1972 1973 Total No, % No, % No. % No. % Pregnant only (P) 31 36.0 180 39.7 233 40.7 444 39.6 Lactating only (L) 29 33.7 180 39.7 223 38.9 432 38.8 Pregnant and lactating (PL) 13 15.1 16 3.5 17 3.0 46 4.1 Resting/estrus ( „ ) (R) ^^' "( 1) '- «( 54 ; 14.1 -fi) 16.4 '«» (J) 15.2 Senile^ 2 2.3 15 3.3 6 1,0 23 2.1 Total 86 100 455 100 573 100 1,114 100 'In the resting/estrus category, subcategones A and B (in parentheses) are specimens with and without a corpus luteum, respectively. ^3^10 corpora, no Type 1 or 2 corpora, and ovaries s 3.5 g. 262 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE Table 9. — Proportions of 1,091 adult reproductive females of Stenella attenuata in pregnant, lactating, and resting/estrus phases. 1971 1972 1973 Total No. % No. % No. % No. % Pregnant (P + V2PL in Table 8) Lactating (L+ V2PL) "Resting" (R) Total reproductive females 37.5 35.5 11 84 44.6 42.3 13.1 100 188 188 64 440 42.8 42.8 14.4 100 241.5 231.5 94 567 42.6 40.8 16.6 100 467 455 169 1,091 42.8 41.7 15.5 100 calving interval spent in that phase, then total length of the interval cycle is 22.7 mo divided by 0.845, or 26.9 mo. A second estimate was obtained directly from the proportion of females pregnant. In calculating this proportion, all pregnant animals were in- cluded (P + PL in Table 8): 490 of 1,091 reproduc- tive females were pregnant, or 44.9% . Division by length of pregnancy, 0.958 yr (11.5 mo), yields an estimate of annual pregnancy rate, 0.469. The re- ciprocal of pregnancy rate, 2.133 yr, or 25.6 mo, is an estimate of average length of calving interval. Both estimates of length of calving interval, 26.9 and 25.6 mo, are overestimates to the extent that the "resting" females with corpora lutea rep- resented uncounted pregnancies, but the effect can be at most very minor For example, if all these females represented undetected pregnan- cies or pregnancies aborted during capture, the unlikely extreme case, the estimates would be 25.7 and 24.7 mo respectively, an average differ- ence of about 1 mo. Since the "resting" females with corpora lutea probably represent a mixture of causes and conditions, including nonfertile ovulations, the probable effect on the estimates is less than 1 mo. Considering this factor and the closeness of the two estimates to each other, it seems certain that the true length of the interval is between 24 and 27 mo. The lower of the two estimates, which is based on fewer assumptions and calculations, was rounded off to 26 mo and is used below in further analysis of life history. The average pattern of events then, consists of 11.5 mo of pregnancy, 11.2 mo of lactation, and 3.3 mo of resting and/or estrus. Overlapping Lactation and Pregnancy About 9.6% of lactating females were also preg- nant (Table 8). Most had fetuses less than 35 to 40 cm long (Figure 26), about halfway through the gestation period. This suggests that overlap when it occurs is usually about 5 to 6 mo long, i.e., conception occurs about halfway through the lac- tation period of about 11 mo, making the calving interval about 20 mo long instead of 26. The very few lactating females with near-term fetuses may have conceived during postpartum estrus or may have begun to lactate shortly before parturition. The data on Graafian follicles are consistent with the theory that postpartum estrus occurs during lactation (Figure 34). The largest follicle in the ovaries of resting/estrus females (including those presumably about to ovulate) is on the av- erage 3 to 4 mm in diameter. After ovulation and conception, the remaining large follicles regress rapidly to about 2 mm (or become lutealized or atretic). There is a further net decline during ges- tation to about 1.5 mm, and during lactation the main modal diameter is about 1.0 mm. During both pregnancy and lactation, however, about 10% of the females (excluding senile individuals, as defined above) have follicles that are within the size range (^3.0 mm) of the presumably ripe follicles present during the resting/estrus phase. This is most clear-cut during lactation. Most of the larger follicles during pregnancy occur in females having fetuses 400 to 500 mm long, or about halfway through the gestation period (Fig- ure 34). Decrease in Reproductive Rate with Age Reproductive rate decreases with age. Age- specific estimates of pregnancy rates and lactation rate were calculated from a random sample of the data for specimens for which teeth were sectioned (stratified to insure representation of corpora- number strata in about the proportions as in the entire sample). The analysis shows decline of preg- nancy rate from about 0.6 at 8 to 9 layers to about 0.3 at 16 to 17 layers (Figure 35). The weighted rate for the pooled sample of 138 used in the calcu- lation was 0.51, comparable to the rate of 0.47 obtained for 1,091 animals (above). The specimens for which teeth were sectioned were about one- third from 1971 and two-thirds from 1972, with a few specimens from earlier years. Lactation rate 263 £ E Ld _l O _l _l o li. H cn LU o cr < o cc UJ t- LlI ll- 10 5 3 2 - I - PREGNANT LACTATING NON-PREGNANT r-N-F I FISHERY BULLETIN; VOL. 74, NO. 2 -ill 10 RESTING/ ESTRUS 11% ♦aar ave - 7 - 6 - 5 -AVE. ? 62 % 100 200 300 400 500 600 700 800 900 LENGTH OF FETUS (mm) Figure 34. — Diameter of largest follicle in pregnant, lactating, and resting females of Stenella attenuata. (Figure 35) increases from about 0.1 at 8 layers to about 0.6 at 12 layers and then again decreases to about 0.5. The initial very low lactation rate compared to pregnancy rate, of course, reflects the fact that a very high percentage of the young females are pregnant for the first time and thus cannot be lactating. The lactation rate climbs rapidly to a level about equal to the pregnancy rate (at about 12 layers) and behaves like the pregnancy rate thereafter. The apparent decline of reproductive rates in older females may be related to the physiological or social mechanisms that cause the appearance of postreproductive females in this age group (see above; not included here). Sex Ratios The overall sex ratio was 44.9% males and 55.1% females (Figure 36). Many large samples examined were predominantly female. Fourteen of 32 single-school samples of 50 or more speci- mens were more than 60% female, whereas none was more than 60% male. The largest single- school sample examined (342) was almost half and half males and females. Sex ratio changes with age (Table 10). This is, of course, making the assumption that the samples examined were representative of the population. Neonates and two-tone animals were almost equally divided between the sexes, but only about 264 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE PREGNflNCr RATE LACTATION RATE II 12 13 LAYERS (number) Figure 35. — Age-related changes in pregnancy (solid line) and lactation rates (dashed line) in Stenella attenuata, based on tooth layer data. Postreproductive females not included. PARITY OVERALL RATIO I I I I I I I I ' I I I I I ' ' ' ' I ' I ' I I 100 150 200 250 SAMPLE SIZE (no) 300 350 Figure 36. — Scatterplot of sex ratio (percent males) on sam- ple size in single-school samples of five or more specimens of Stenella attenuata. Overall ratio (dashed line) from Table 12. Table lO. — Sex ratio, by color pattern stage, in 9,371 speci- mens ot Stenella attenuata, 1971-73. Color pattern Males Females stage No. % No. % Total Neonate Two-tone Speckled Mottled Fused Total 205 666 609 569 2,154 4,203 49.8 487 47.8 43.8 429 44.9 207 701 666 729 2,865 5,168 50.2 51.3 52.2 56.2 57.1 55.1 412 1,367 1,275 1,298 5,019 9,371 43% of the adults examined were males. The greatest change in ratio, from 48.0 to 43.5% male, comes about during the transition to mottled col- oration between 7 and 8 layers of age. Assuming random sampling of the population, male and female mortality rates must diverge sharply at this point. Gross Annual Production An estimate of average gross annual production of calves for 1971 to 1973 was calculated based on the estimate of annual pregnancy rate, the color pattern phase data, and the proportions of mottled and fused females found to be sexually mature (Table 11). Seven hundred and twenty-nine of 9,371 ani- mals were mottled females (7.8%) and 2,865 were fused females (30.6%). Of 127 mottled and 1,141 fused females, 47.4 and 88.4% were sexually mature, respectively (Table 11). Average preg- nancy rate was 0.469. Production = [(0.078 x 0.474) + (0.306 X 0.884)] 0.469 = 0.144 of the popu- lation per year. Table ll. — Sexual maturity (presence of ovarian corpora) in mottled and fused females of Stenella attenuata, 1971-73. Mottled Fused^ N Mat J re N Meture No. ' % Year No. % 1971 1972 1973 Total 6 92 170 268 5 37 85 127 (- ) (40.2) (50.0) (47,4) 99 473 569 1,141 82 (82.8) 417 (88.2) 510 (89.6) 1,009 (88.4) Schooling in Relation to Reproduction Kasuya (1972) reported changes in structure and size of schools of .S. coeruleoalba correlated with breeding condition and breeding activities. Kasuya et al. (1974) proposed a complex hypothet- ical system of school formation and breakdown determined by reproductive activities in the Japanese population of S. attenuata. They suggested that juveniles of S. attenuata in Japa- nese waters leave breeding schools and school sep- arately, rejoining the breeding schools at puberty. There is nothing to indicate that this happens in the eastern Pacific. We examined the coloration structure (= age structure) of single-school sam- ples. Of 324 single-school samples of seven or more animals, only 1 (of 17 animals) contained no adults (or neonatal calves, which would indicate presence of adult lactating females in the school). This sam- ple (8 two-tone, 2 speckled, and 7 mottled) was from a school of about 600 spotted porpoise, S. attenuata, congregated with about 600 spinner porpoise, iS. longirostris. Given that about half the animals examined were adults, the probability of 265 FISHERY BULLETIN: VOL. 74, NO. 2 a single-school sample of seven containing no "fused" individuals is aboutO.Ol (= 0.5''). If schools consisting only of juveniles were common, many more all-juvenile samples would have been en- countered. Conversely, juveniles (two-tone, speck- led, and/or mottled) occurred in all but 3 of the 324 samples. It must be concluded that juveniles prob- ably do not school separately in the eastern Pacific. Another possibility, albeit unlikely, is that all-juvenile schools exist but are not captured by tuna seiners. COMPARISON WITH THE JAPANESE POPULATION Many of the estimates of life history parameters presented here differ from those published by Kasuya et al. (1974) for the relatively unexploited population of S. attenuata in Japanese waters (Ta- ble 12). The differences could be caused by 1 ) differ- ential procedures or analytical methods, 2) in- trinsic racial differences between the populations, or 3) differential population status, e.g., exploited versus unexploited. The comparisons below of similarly calculated average estimates, of course, rest on the assumption that the overall samples in both cases were not biased with respect to age or sex. The major sampling differences between the two studies is that the Japanese sample consisted mostly of large samples from a few schools. whereas our sample consisted mainly of aggre- gated, small samples from many schools. Both studies assume no sampling bias. Comparison of large, single-month samples in the present study with large, single-school samples in the Japanese study (e.g., the October 1972 sample in Figures 15 and 16 with sample number 2 in Figure 2 of Kasuya et al. 1974) indicate very similar length- frequency distributions and support the idea that the aggregated samples are probably not biased, or, if biased, are biased in the same way. This inference is, of course, based on the assumption that the underlying population structures are about the same in the two populations. The estimate of Kasuya et al. ( 1974) of length at birth was based on only 5 full-term fetuses and newborn calves versus 86 in the present study. Our estimate can, therefore, be considered more reliable, although the possibility does exist that length at birth is greater in the Japanese popula- tion. The difference between the estimated lengths at 1 yr for the two populations is about the same as the difference between the estimates of length at birth. Estimated length at attainment of sexual maturity and maximum length (for males) are also greater for the Japanese samples. The estimate of length at maturity of males is greater in spite of the fact that Kasuya et al. used a lower testis-weight criterion than we did (68 versus 100 g). The average lengths of both adult males and Table 12. — Comparison of estimates of average life history parameters of Stenella attenuata by Kasuya et al. (1974) and in present paper. Parameter (average) Kasuya et al. Perrin et al. 1 . Length at birth 89 cm 82.5 cm 2. Growth rate in 1st year 4.5 cm per mo 4.6 cm per mo 3. Length at 1 yr 143 cm 138 cm 4. Length at onset of sexual maturity: Males 197 cm -195 cm Females 187 cm 181 ±1 cm 5. Age at onset of sexual maturity: Males 10.3 layers (10 3 yr) 12 layers (6-11 yr) Females 8.2 layers (8.2 yr) 9 layers (4.5-8 yr) 6. Average length of sexually mature adults: Males 204-207 cm 200.7 cm Females 192-195 cm 187.3 cm 7. Maximum length: Males 234 cm 226 cm Females 220 cm 220 cm 8. Maximum number of consistently readable tooth layers -13 12-13 9. Average ovulation rate (based on layers) 0.8 per layer -1 per layer (0.8 per yr) (1 or 2 per yr) in fully mature, more in younger 10. Pregnancy rate (overall) 0.27 per yr 0.47 per yr 11. Breeding seasons 3 per yr multiple 12. Gestation 11.2 mo 11.5 ±0.2 mo 13. Lactation 29.3 mo 11.2 mo 14. Resting 9.8 mo 3.3 mo 15. Length of calving interval 4,19 yr 2.17 yr 16. Sex ratio: Overall 0.76 male:1 female 0.81 male:1 female At birth 1.3-1.5:1 1.00:1 Adults 0.58:1 0.75:1 266 PERRIN ET AL : GROWTH AND REPRODUCTION OF SPOTTED PORPOISE females are also greater in the Japanese popula- tion, and in this case, all four of the estimates are based on large and certainly adequate samples. These differences all suggest that the Japanese form is about 6 to 8 cm larger than the eastern Pacific form. The estimates of Kasuya et al. (1974) of age at attainment of sexual maturity are based on their conclusion that one tooth layer corresponds to 1 yr of growth. It appears from comparisons of their first-year growth curve with ours (note rate in first year and length at 1 yr) that our first two layers correspond to their first layer. Kasuya (1972) in his paper on growth of S. coeruleoalba mentioned observing "one or two faint translucent layers in the thick opaque layer accumulated just after the birth" that were "not used for age determination because it was not expected to show the annual accumulation cycle," and Kasuya et al. (1974) stated that the "dentinal growth layers of this species [S. attenuata] does not differ so much from that of S. coeruleoalba.'' After the first year, our hypothesis 2 corresponds to the assumption of Kasuya et al . of one layer per year, e.g., nine layers of Perrin et al. (1973) = eight layers of Kasuya et al. = 8 yr. The average length of calving interval in both studies was estimated by several methods that converged on the respective central estimates. One minor difference between the two analyses is that Kasuya et al. (1974) did not exclude postre- productive females from the "resting/estrus" group. Thus, their estimate of the average resting/estrus period of 9.8 mo may be a slight overestimate. The probable effect of this on the estimate of length of total calving interval is very small, however, and it therefore seems that the estimates are analyti- cally comparable and that the difference between them is real. Kasuya et al. estimated that indi- vidual intervals in the Japanese population vary from 23 to 60 mo, with modes at 28 to 30, 36 to 38, and 54 to 56 mo. The potential thus probably exists for a shift in average length from 50 mo (4.17 yr) to 26 mo (2.17 yr) under exploitation. Kasuya et al. (1974) used the same methods used here to estimate length of the lactation period and arrived at a "best" estimate of 29.3 mo, some 18 mo longer than our estimate of 11.2 mo. They found that the major shift from milk to solid food occurs at body length of about 133 cm, about the same as in our sample, but that some suckling and lactation of the mother continues for an average additional 20 mo. The prolonged suckling is prob- ably nutritionally a largely nonfunctional aspect of general prolonged parental care. It has been suggested on the basis of comparison of the life histories and behavior of mysticetes and odonto- cetes that this period in odontocetes may allow for "sophisticated" communicational-navigational training (Brodie 1969). Thus the apparent shorter lactation period in the eastern Pacific, and the concomitant shorter calving interval and higher pregnancy rate, does not necessarily mean earlier effective weaning, but may reflect a truncated pa- rental care period. The apparent overall sex ratios are almost the same for the two populations, but the proportion of males was higher at birth and lower at maturity in the Japanese samples than in the eastern Pacific samples. A lower proportion of males at birth could be a response to exploitation. Kasuya et al. (1974) suggested that the very low proportions of males in mature age-classes in the Japanese catches could be partially caused by segregation of adult males or by differential catchability but are largely due to differential mortality rates. If the decrease in proportion of males with age is caused by differential mortality, the apparent faster de- crease in the Japanese population must mean that the disparity in mortality rates between the sexes is greater there than in the eastern Pacific. In summary, the two sets of estimates differ in a consistent way, and the differences are real. It seems possible that the differences in some way reflect exploitation in the eastern Pacific. ACKNOWLEDGMENTS This study would not have been possible without the generous cooperation and assistance of the owners, masters, and crews of the tuna seiners Conte Bianco, Carol Virginia {now Carol S), Larry Roe, Nautilus, Mary Antoinette, San Juan, Con- cho, Kerri M, Queen Mary, Eastern Pacific, John F. Kennedy, Sea Preme, Westport, Anne M, Pacific Queen, J. M. Martinac, Lois Seaver, Marietta, Independence, Sea Quest, Bold Contender, Jac- queline A, Frances Ann, Elsie A, Sea Royal, Jac- queline Marie, Trinidad, Mermaid, Bettie M, An- tonina C, Day Island, Connie Jean, and Denise Marie. Scientists and technicians who collected data and specimens aboard the vessels include C. E. Bowlby, R. W. Cunningham, W. E. Evans, R. S. Garvie, J. M. Greene, D. B. Holts, J. La Grange, J. S. Leatherwood, R. E. Loghry, R. L. McNeely, C. W. Oliver, R. J. Olson, C. J. Orange, D. J. Otis, J. W. 267 FISHERY BULLETIN: VOL. 74, NO. 2 Ploeger, A. Poshkus, F. M. Ralston, S. B. Reiley, J. M. Rosen, C. R. Ryan, K. D. Sexton, G. M. Treinen, J. A. Young, and D. B. Zantiny. R. L. Brownell, Jr., G. D. Fitzgerald, D. W. Rice, W. A. Walker, and D. W. Waller contributed unpublished data. D. B. Holts sectioned the teeth, and D. W. Rice assisted with the readings. R. B. Miller processed and examined the ovaries. T. D. Smith and N. K. Wiley provided advice and assistance in data processing and analysis. I. Barrett, P. J. H. van Bree, P. F. Brodie, R. L. Brownell, Jr., W. Clark, W. E. Evans, C. L. Hubbs, T. Kasuya, W. H. Lenarz, J. G. Mead, D. W. Rice, D. E. Sergeant, T. D. Smith, and G. Stauffer read the manuscript. We thank these per- sons and others not mentioned for their invalu- able assistance. LITERATURE CITED Amoroso, E. C, and C. a. Finn 1962. Ovarian activity during gestation, ovum transport and implantation. In S. Zuckerman (editor), The ovary. Vol. 1, p. 451-537. Academic Press, N.Y. Best, p. B. 1968. The sperm whale (Physeter catodon) off the west coast of South Africa. 2. Reproduction in the female. S. Afr. Div. Sea Fish., Invest. Rep. 66, 32 p. 1969. The sperm whale (Physeter catodon) off the west coast of South Africa. 3. Reproduction in the male. S. Afr. Div. Sea Fish., Invest. Rep. 72, 20 p. 1970. The sperm whale (Physeter catodon) off the west coast of South Africa. 5. Age, growth and mortality. S. Afr. Div. Sea Fish., Invest. Rep. 79, 27 p. Brodie, p. F. 1969. Duration of lactation in Cetacea: An indicator of required learning? Am. Midi. Nat. 82:312-314. 1971. A reconsideration of aspects of growth, reproduc- tion, and behavior of the white whale (Delphinapterus leucas), with reference to the Cumberland Sound, Baffin Island, population. J. Fish. Res. Board Can. 28:1309-1318. 1972. Significance of accessory corpora lutea in odonto- cetes with reference to Delphinapterus leucas. J. Mammal. 53:614-616. Brunn, a. 1935. Flying-fishes (Exocoetidae) of the Atlantic. Sys- tematic and biological studies. Dana Rep. 2(6), 106 p. Clarke. M. R. 1966. A review of the systematics and ecology of oceanic squids. Adv. Mar. Biol. 4:91-300. BAILEY, M. D., AND W. F. PERRIN. 1973. Helminth parasites of porpoises of the genus Stenella in the eastern tropical Pacific, with descrip- tions of two new species: Mastigonema stenellae gen. et sp. n. (Nematoda: Spiruroidea) and Zalophotrema pacifkum sp. n. (Trematoda: Digenea). Fish. Bull., U.S. 71:455-471. Fisher, H. d., and R. J. Harrison. 1970. Reproduction in the common porpose (Phocoena phocoena) of the North Atlantic. J. Zool. (Lond.) 161:471- 486. Harrison, r. J., R. L. Brownell, Jr., and r. C. Boice. 1972. Reproduction and gonadal appearances in some odontocetes. In R. J. Harrison (editor). Functional anat- omy of marine mammals. Vol. 1, p. 361-429. Academic Press, N.Y. Hasselblad, v. 1966. Estimation of parameters for a mixture of normal distributions. Technometrics 8:431-444. HUGGETT, A. St. G., and W. F. WIDDAS. 1951. The relationship between mammalian foetal weight and conception age. J. Physiol. 114:306-317. Kasuya. T. 1972. Growth and reproduction oi Stennella caeruleoalba based on the age determination by means of dentinal growth layers. Sci. Rep. Whales Res. Inst. (Tokyo) 24:57-79. 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. KLEVEZAL', G. A., AND S. E. KLEINENBERG. 1969. Age determination of mammals from annual layers in teeth and bones. Izdatel'stvo "Nauka," Mosc, 1967. (Translated by Israel Program Sci. Transl., Jerus., 128 p.) KNAUSS, J. A. 1963. Equatorial current systems. In M. N. Hill (editor), The sea. Vol. 2, p. 235-252. Interscience Publ., N.Y. LAIRD, A. K. 1969. The dynamics of growth. Res./Dev. 20(8):28-31. Laws, R. M. 1959. The foetal growth rates of whales with special reference to the fin whale, Balaenoptera physalus Linn. Discovery Rep. 29:281-307. M0HL-HANSEN, U. 1954. Investigations on reproduction and growth of the porpoise (Phocaena phocaena (L.)) from the Baltic. Vidensk. Medd. Dan. Naturhist. Foren. Kbh. 116:369-396. MOSER, H. G., AND E. H. AHLSTROM. 1970. Development of lanternfishes (family Myctophidae) in the California Current. Part I. Species with narrow- eyed larvae. Los Ang. Cty. Mus. Nat. Hist. Sci. 7, 145 p. NISHIWAKI, M., M. NAKAJIMA, AND T KAMIYA. 1965. A rare species of dolphin (Stenella attenuata) ft-om Arari, Japan. Sci. Rep. Whales Res. Inst. (Tokyo) 19:53-64. PERRIN, W. F. 1969. Using porpoise to catch tuna. World Fishing 18(6): 42-45. 1970a. The problem of porpoise mortality in the U.S. tropi- cal tuna fishery. Proc. 6th Annu. Conf. Biol. Sonar Diving Mammals, Stanford Res. Inst., Menlo Park, Calif., p. 45-48. 1970b. Color pattern of the eastern Pacific spotted por- poise Stenella graffmani Ldnnberg (Cetacea, Delphin- idae). Zoologica (N.Y.) 54:135-149. 1975. Distribution and differentiation of populations of dolphins of the genus Stenella in the eastern tropical Pacific. J. Fish. Res. Board Can. 32:1059-1067. In press. Variation of spotted and spinner porpoises (genus Stenella) of the eastern tropical Pacific and Hawaii. Bull. Scripps Inst. Oceanogr., Univ. Calif. 268 PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE PERRIN, W. F., AND E. L. ROBERTS. 1972. Organ weights of non-captive porpoise (Stenella spp.). Bull. South. Calif. Acad. Sci. 71:19-32. PERRIN, W. F., R. R. WARNER, C. H. FiSCUS, AND D. B. HOLTS. 1973. Stomach contents of porpoise, Stenella spp., and yellowfin tuna, Thunnus albacares, in mixed-species aggregations. Fish. Bull., U.S. 71:1077-1092. RICHARDS, W. J., AND W. L. KLAWE. 1972. Indexed bibliography of the eggs and young of tunas and other scombrids (Pisces, Scombridae), 1880- 1970. U.S. Dep. Commer., NOAA Tech. Rep., NMFS SSRF-652, 107 p. SECKEL, G. R. 1972. Hawaiian-caught skipjack tuna and their physical environment. Fish. Bull., U.S. 70:763-787. SERGEANT. D. E. 1962. The biology of the pilot or pothead whale Globi- cephala melaena (Traill) in Newfoundland waters. Fish. Res. Board Can. Bull. 132, 84 p. 1973. Biology of white whales (Delphinapterus leucas) in western Hudson Bay. J. Fish. Res. Board Can. 30:1065-1090. SERGEANT, D. E., D. K. CALDWELL, AND M. C. CALDWELL. 1973. Age, growth, and maturity of bottlenosed dolphin (Tursiops truncatus) from northeast Florida. J. Fish. Res. Board Can. 30:1009-1011. TRUE, F. W. 1906. Notes on a porpoise of the genus Prodelphinus from the Hawaiian Islands. Bull. U.S. Fish Comm. 23:811-815. Wyrtki, K. 1964. The thermal structure of the eastern Pacific Ocean. Erganzungsh. Dtsch. Hydrogr. Z. Reihe A 8(6), 84 p. YOUNG, R. E. 1972. The systematics and areal distribution of pelagic cephalopods from the seas off southern California. Smithson. Contrib. Zool. 97, 159 p. 269 GROWTH OF LABORATORY-REARED NORTHERN ANCHOVY, ENGRAULIS MORDAX, FROM SOUTHERN CALIFORNIA Gary T. Sakagawa and Makoto Kimura^ ABSTRACT The northern anchovy, Engraulis mordax, was experimentally reared in the laboratory at the South- west Fisheries Center, La Jolla, Calif. Data from three experiments were used to empirically fit a two-phase Gompertz growth model. The model describes growth from hatching to about 20 mo of age. It was estimated that the average length of laboratory-reared anchovies is 102 mm at 1 yr old and 119 mm at 2 jT old. Growth of laboratory-reared anchovies was comparable to that of anchovies in the wild. Attempts to rear the northern anchovy, En- graulis mordax, at the Southwest Fisheries Center (SWFC), La Jolla, Calif., were begun in 1966 when G. O. Schumann collected anchovy larvae in the ocean off La Jolla and successfully reared them using wild plankton as food in the laboratory (Bardach 1968). Schumann's success was followed by other laboratory experiments in which anchovies were reared from eggs, larvae, and juveniles that were caught in the ocean (Ta- ble 1). In 1970, Leong (1971) developed a method for artificially inducing anchovies to spawn by controlling the photoperiod and injecting hor- mones. This technique is currently used at the SWFC to produce eggs and to rear anchovies for experimental purposes. One of the purposes of the rearing experiments at the SWFC has been to obtain physiological and biochemical information needed for describing the energy budget of the northern anchovy, and to relate the results to the feeding dynamics of the anchovy population in the California Current, which consists of primarily young fish less than 3 yr old. Growth data are needed for analysis of the budget, and various attempts have been made to measure growth in the laboratory. Kramer and Zweifel (1970) and Lasker et al. (1970) reported growth rates of anchovy larvae. In this report we extend their analyses to include growth from hatching to about 20 mo old. We also present a mathematical model that describes this growth and compare our results with those of other investigators. ^Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, La Jolla, CA 92038. SOURCES OF DATA Data primarily from experiments of G. O. Schumann (Schumann-I; Schumann-ID, G. O. Schumann and A. Saraspe (Schumann-Ill), and R. Leong (pers. commun., SWFC) were used in our study (Table 1). Schumann-II successfully reared larval an- chovies for 22 days at about 22°C water tempera- ture, which is higher than the temperature (15° to 16°C) at which anchovy larvae are frequently found in large numbers in the California Current (pers. commun., P. Smith, SWFC). The larvae were fed wild plankton and samples were taken for length measurement approximately daily. Schumann-Ill reared anchovies from the egg stage through the juvenile stage in aquaria for 83 days on a diet of wild plankton, Artemia salina, and commercial trout food. The experiment was conducted from March to June and the water temperatures in the aquaria were not recorded. However, during March to June the average water temperature in rearing aquaria at the SWFC is generally about 18° to 22°C. Leong (pers. commun.) obtained juvenile an- chovies from a live-bait dealer and reared the fish to maturity in a 4.6-m diameter pool (13.2 kl) with circulating seawater. The water tempera- ture in the pool was a few degrees higher than the prevailing water temperature oflFScripps Pier, La Jolla, site of the water intake for the experimen- tal pool (Lasker and Vlymen 1969). Leong fed the fish a diet o( Artemia salina, ground squid and anchovies, and commercial trout food. Once a month about 25 fish were sacrificed for length and weight measurements. Manuscript accepted October 1975. FISHERY BULLETIN: VOL. 74, NO. 2, 1976. 271 FISHERY BULLETIN: VOL. 74, NO. 2 Table l. — Laboratory experiments of rearing the northern anchovy at the Southwest Fisheries Center, La JoUa, Calif. Source Hunter (1976) Kramer and Zweifel (1970) Lasker et al. (1970) Leong (unpubl. data)' Paloma (see text footnote 3) Schumann-I (G. O. Schumann unpubl. data)^ Sctiumann-ll (Kramer and Zweifel 1970) Scfiumann-lll (G. O. Schumann and A. Saraspe unpubl. data)^ Theilacker and McMaster (1971) Rearing Average length (mm^ Life stage at start start of duration ' rearing (days) Stan Finish Food Eggs April 74 4.0 35.0 Gymnodinium splendens, Brachionus plicatilis, Tisbe furcata, and Anemia saline Eggs August and September 35 3.2 17.4 Wild plankton and A salina Eggs February 50 3.4 21.0 Bulla gouldiana. G. splendens, and A. salina Juveniles Apnl 474 88.3 117.7 Squid, anchovy, A salina, and trout food Juveniles November 624 75.0 106.2 Anemia salina and trout food Larvae Inarch 97 18.0 81.9 Wild plankton Eggs March 22 2.9 16.2 Wild plankton Eggs March 83 3.5 67.1 Wild plankton, A. salina, and trout food Eggs 19 12.0 Gymnodinium splendens, B. plicatilis, and A. salina 'Pers. commun.. Southwest Fisheries Center, La Jolla, Calif. ^Data are on file at the Southwest Fisheries Center, La Jolla, Calif. In all of these experiments the fish were from the southern California stock (Vrooman and Smith 1971), reared at laboratory ambient water temperature, and not subjected to experimental treatment or excessive handling. All fish sampled for measurements were sacrificed. The length measurement is standard length. TREATMENT OF DATA The age of anchovies reared by Schumann-II and Schumann-Ill were known because the an- chovies were hatched from eggs at the start of the rearing experiments. In Leong's {pers. commun.) experiment, the exact age of his fish was not known because juvenile fish of average length of 88.3 mm were used at the start of the experiment. We estimated the age of Leong's fish from data from Schumann-I in which anchovies were reared for 97 days from an average length of 18.0 to 81.9 mm (Table 1), and data from Schumann-Ill which indicated that an 18.0 mm fish, raised from eggs, was about 30 days old. Our age estimate is 4 mo. Several mathematical models describing growth of organisms are available (e.g., Parker and Larkin 1959; Richards 1959; Laird 1969). The commonly used models in fisheries are the exponential, the von Bertalanffy, and the Gom- pertz models (Beverton and Holt 1957; Silliman 1969). The Gompertz model was selected for our study because it was shown by Kramer and Zweifel (1970) to be better than the exponential model for describing growth of laboratory-reared anchovy larvae and because it generally de- scribes the growth of fishes fairly well. Also, pre- liminary analysis of our data indicated that the von Bertalanffy model poorly described the growth of young fish. The Laird version of the Gompertz growth model (Laird 1969) describes an asymmetric sig- moid curve of the form. L, = L exp {C [ 1 - exp i-at )] } where Lq = length at zero age or hatching C = a constant a = rate of decay of exponential growth t = age in months. This model was fitted to our data using an itera- tive least squares procedure (Conway et al. 1970). Our goal was to describe growth on a coarse time scale, i.e., monthly rather than on a fine time scale, i.e., daily. GROWTH FROM HATCHING TO JUVENILE STAGE The Gompertz growth model and an exponen- tial growth model were applied to data of Schu- mann-II by Kramer and Zweifel (1970). Both models described the data from Schumann-II reasonably well, although the Gompertz model de- 272 SAKAGAWA and KIMURA: GROWTH OF LABORATORY-REARED ANCHOVY scribed the data better. In the Kramer-Zweifel analysis the length at zero age,Lo, was fixed at 2.5 mm, the average size at hatching. We also applied the Gompertz grovd:h model to data of Schumann- II. Kramer and Zw^eifel (1970) used data only for 17 days of growth. We used all of the data of Schu- mann-II, which included sampling through 22 days of growth, and fitted the model first with Lq fixed at 2.5 mm and again without this constraint, i.e.,Lo was estimated. The results (Figure 1) indi- cate that there is not much difference in the curves withLo fixed or estimated within the range of the data. Outside the range of the data, the curves diverge considerably and there is a substantial difference; the curve withLo estimated has a lower asymptotic length (61 mm) than the curve withLo fixed at 2.5 mm (asymptotic length of about 696 mm). Zweifel and Lasker^ showed that a two-phase Gompertz curve described the data from Schu- mann-II better than a single-phase Gompertz curve. The separation of the phases occurred at about 6 days of age, the onset of feeding in an- chovy larvae. Schumann-Ill reared anchovies for a longer period than Schumann-II. Fish reared by Schu- ^Zweifel, J. R. and R. Lasker. 1974. Prenatal and postnatal growth of fishes — a general model. Unpubl. manuscr. Southwest Fisheries Center, La JoUa, CA 92038. mann-II, however, were larger than those reared by Schumann-Ill at similar ages. For example, at 0.5 mo of age fish reared by Schumann-II aver- aged 12.1 mm long and fish reared by Schu- mann-Ill, 8.2 mm long. Although the sample size is small, this difference is statistically significant at the 1% probability level. Differences in rearing procedures, i.e., diet and temperature of water, probably produced the difference in growth (Kramer and Zweifel 1970; Lasker et al. 1970). The Gompertz growth model was applied to data from Schumann-Ill first withLo fixed at 2.5 mm and then with Lo estimated (Figure 2). As in the case with data from Schumann-II, this model describes the grovd;h data reasonably well, and the curve with Lo estimated has a lower asymp- totic length (81 mm) than the curve with Lq fixed (asymptotic length of about 93 mm). GROWTH DURING JUVENILE TO ADULT STAGE Anchovies reared by Leong (pers. commun.) were juveniles at the start of the experiment and grew to an average size of 117.7 mm in 474 days (Table 2). Growth was in steplike stages charac- terized by rapid growth followed by a leveling off. The first stage was between 4 and 12 mo of age and the second was between 12 and about 20 mo of age. 25 E ?0 E X 1- 15 z UJ _l Q 10 < Q z