Fishery Bulletin ^^ATES O^ ^ r Marine Bioiogicai Laburasor- LIBRARY Vol. 78, No. 1 Woods Hole, Mass. ADAMS, PETER B. Life history patterns in marine fishes and their consequences ' for fisheries management 1 AMBLER, JULIE W. Species of Munidopsis (Crvistacea, Galatheidae) occurring off Oregon and in adjacent waters 13 MENDELSSOHN, ROY. Using Markov decision models and related techniques for purposes other than simple optimization: analyzing the consequences of policy alternatives on the management of salmon runs 35 STOUT, VIRGINIA F. Organochlorine residues in fishes from the northwest Atlan- tic Ocean and Gulf of Mexico 51 PIETSCH, THEODORE W., and JOHN P. VAN DUZER. Systematics and distribu- tion of ceratioid anglerfishes of the family Melanocetidae with the description of a new species from the eastern North Pacific Ocean 59 HUNTER, JOHN R., and CAROL M. KIMBRELL. Early life history of Pacific mack- erel, Scomber japonicus o9 MORSE, WALLACE W. Spawning and fecundity of Atlantic mackerel. Scomber scombrus, in the Middle Atlantic Bight 103 WEIHS, DANIEL. Respiration and depth control as possible reasons for swimming of northern anchovy, Engraulis mordax, yolk-sac larvae 109 POWLES, HOWARD. Descriptions of larval silver perch, Bairdiella chrysoura. banded drum, Larimus fasciatus, and star drum, Stellifer lanceolatus (Sciaenidae) 119 GRIMES, CHURCHILL B., and GENE R. HUNTSMAN. Reproductive biologj' of vermilion snapper, Rhomboplites aurorubens, from North Carolina and South Carolina 137 PETERSON, R. H., P. H. JOHANSEN, and J. L. METCALFE. Observations on ear- ly life stages of Atlantic tomcod, Microgadus tomcod 147 Notes HINES, ANSON H., and THOMAS R. LOUGHLIN. Observations of sea otters dig- ging for clams at Monterey Harbor, California 159 WEIS, PEDDRICK, and JUDITH SHULMAN WEIS. Effect of zinc on fin regener- ation in the mummichog, Fundulus heteroclitus , and its interaction with methyl- mercury 163 (Continued on back cover) J Seattle, Washington U.S. DEPARTMENT OF COMMERCE Philip M. Klutznick, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION Richard A. Frank, Administrator Terry L. Leitzell, Assistant Administrator for Fisheries NATIONAL MARINE FISHERIES SERVICE 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 1 904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941 . Separates were issued as documents through volume 46; the last document was No. 1103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate appeared as a numbered bulletin. A new system began in 1963 with volume 63 in which papers are bound together in a single issue of the bulletin instead of being issued individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued quarterly. In this form, it is available by subscription from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. It is also available free in limited numbers to libraries, research institutions. State and Federal agencies, and in exchange for other scientific publications. EDITOR Dr. Jay C. Quast Scientific Editor, Fishery Bulletin Northwest and Alaska Fisheries Center Auke Bay Laboratory National Marine Fisheries Service, NOAA P.O. Box 155, Auke Bay, AK 99821 Editorial Committee Dr. Elbert H. Ahlstrom Dr. Merton C. Ingham National Marine Fisheries Service National Marine Fisheries Service Dr. Bruce B. Collette Dr. Reuben Lasker National Marine Fisheries Service National Marine Fisheries Service Dr. Edward D. Houde Dr. Jerome J. Pella University of Miami National Marine Fisheries Service Dr. Sally L. Richardson Gulf Coast Research Laboratory Kiyoshi G. Fukano, Managing Editor The Fishery Bulletin (USPS 090-870) is published quarterly by Scientific Publications Office, National Marine Fisheries Service, NOAA, Room 336, 1700 Westlake Avenue North, Seattle, WA 98109. Controlled circulation paid to Finance Department, USPS, Washington, DC 20260. Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated. The Secretary of Commerce has determined that the publication of this periodical 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 March 1982. Fishery Bulletin CONTENTS Vol. 78, No. 1 ADAMS, PETER B. Life history patterns in marine fishes and their consequences for fisheries management 1 AMBLER, JULIE W. Species of Munidopsis (Crustacea, Galatheidae) occurring off Oregon and in adjacent waters 13 MENDELSSOHN, ROY. Using Markov decision models and related techniques for purposes other than simple optimization: analyzing the consequences of policy alternatives on the management of salmon runs 35 STOUT, VIRGINIA F. Organochlorine residues in fishes from the northwest Atlan- tic Ocean and Gulf of Mexico 51 PIETSCH, THEODORE W., and JOHN P. VAN DUZER. Systematics and distribu- tion of ceratioid anglerfishes of the family Melanocetidae with the description of a new species from the eastern North Pacific Ocean 59 HUNTER, JOHN R., and CAROL M. KIMBRELL. Early life history of Pacific mack- erel. Scomber japonicus 89 MORSE, WALLACE W. Spawning and fecundity^ of Atlantic mackerel, Scomber scombrus, in the Middle Atlantic Bight 103 WEIHS, DANIEL. Respiration and depth control as possible reasons for swimming of northern anchovy, Engraulis mordax, yolk-sac larvae 109 POWLES, HOWARD. Descriptions of larval silver perch, Bairdiella chrysoura, banded drum, Larimus fasciatus, and star drum, Stellifer lanceolatus (Sciaenidae) 119 GRIMES, CHURCHILL B., and GENE R. HUNTSMAN. Reproductive biology of vermilion snapper, Rhomboplites aurorubens, from North Carolina and South Carolina 137 PETERSON, R. H., P. H. JOHANSEN, and J. L. METCALFE. Observations on ear- ly life stages of Atlantic tomcod, Microgadus tomcod 147 Notes HINES, ANSON H., and THOMAS R. LOUGHLIN. Observations of sea otters dig- ging for clams at Monterey Harbor, California 159 WEIS, PEDDRICK, and JUDITH SHULMAN WEIS. Effect of zinc on fin regener- ation in the m\iTam\c\\og, Fundulus heteroclitus , and its interaction with methyl- mercury 163 (Continued on next page) Seattle, Washington 1980 For sale by the Superintendent of Documents. U.S. Government Printing Office. Washington, DC 20402 — Subscription price per year $12 00 domestic and $15 00 foreign Cost per single issue $3 00 domestic and $3 75 foreign Contents-continued TESTAVERDE, SALVATORE A., and JAMES G. MEAD. Southern distribution of the Atlantic whitesided dolphin, Lagenorhynchus acutus, in the western North Atlantic 167 MATARESE, ANN C, and DAVID L. STEIN. Additional records of the sculpin Psychrolutes phrictus in the eastern Bering Sea and off Oregon 169 ODELL, DANIEL K., EDWARD D. ASPER, JOE BAUCOM, and LANNY H. COR- NELL. A recurrent mass stranding of the false killer whale, Pseudorca crassi- dens, in Florida 171 BRANSTETTER, STEVEN, and ROBERT L. SHIPP. Occurrence of the finetooth shark, Carcharhinus isodon, off Dauphin Island, Alabama 177 BAGLIN, RAYMOND E., JR., MARK I. FARBER, WILLIAM H. LENARZ, and JOHN M. MASON, JR. Shedding rates of plastic and metal dart tags from Atlantic blue- fin tuna, Thunnus thynnus 179 HAYNES, JAMES M., and ROBERT H. GRAY. Influence of Little Goose Dam on upstream movements of adult chinook salmon, Oncorhynchus tshawytscha 185 MORSE, WALLACE W. Maturity, spawning, and fecundity of Atlantic croaker, Micropogonias undulatus , occurring north of Cape Hatteras, North Carolina 190 MILLER, ROBERT E., DOUGLAS W. CAMPBELL, and PAMELA J. LUNSFORD. Comparison of sampling devices for the juvenile blue crab, Callinectes sapidus . . . 196 Notices NOAA Technical Reports NMFS published during the last 6 mo of 1979 199 Vol. 77, No. 4 was published on 23 July 1980. The National Marine Fisheries Service (NMFS) does not approve, rec- ommend or endorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to NMFS, or to this publication furnished by NMFS, in any advertising or sales pro- motion 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. LIFE HISTORY PATTERNS IN MARINE FISHES AND THEIR CONSEQUENCES FOR FISHERIES MANAGEMENT Peter B. Adams* ABSTRACT Natural selection operates at the life history level to maximize the number of surviving offspring. Life history characteristics will vary in consistent patterns to meet this constraint. When theoretical patterns in life histories were investigated in terms of r and K selection and compared with actual trends in life history characteristics of fishes, the agreement between observed and predicted trends was significant. The effects of harvesting on stocks with these life history trends were investigated and it was found that K selected type species would be highly sensitive to overfishing and, once depleted, recovery would require a long time. The ecological and genetic properties of a species are intimately linked. The morphological and re- productive characteristics, population sizes, and genetic frequencies of species are adjusted to their environments by natural selection. Species in- habiting different environments show different patterns of life history characteristics. The rela- tionship among habitat, ecological strategies, and population parameters has been termed r and K selection (Mac Arthur and Wilson 1967) and/or op- timal life histories ( Gadgil and Bossert 1970). This body of theory is based on the assumption that natural selection operates on these characteristics in order to maximize the number of surviving offspring produced. Under an environmental re- gime with a large component of unpredictable, nonselective, mortality an organism will allocate a larger portion of its resources to reproductive activities (an r strategist). Conversely the optimal allocation of resources for a population subjected to a high proportion of predictable, selective mor- tality will be toward increasing individual fitness, frequently through competitive ability (a K strategist). With the number and variability of factors operating on any particular species, no species is going to be an r or X strategist in an absolute sense. A species will only occupy a rela- tive position on the r and K continuum. In fisheries biology, the value of comparative studies of life history parameters has long been recognized (Holt 1962; Beverton 1963; Gushing 1971; Alverson and Carney 1975). These life his- ' Southwest Fisheries Center Tiburon Laboratory, National Marine Fisheries Service, NOAA, Tiburon, CA 94920. tory parameters should vary in a consistent pat- tern which can be predicted from the theory of r and K selection. In this paper, these predictions are tested with life history parameters from major groups of marine fishes. The theory has implica- tions for management, particularly when fisheries are in the initial stages of development. THEORY OF r AND K SELECTION The theory of r andX selection is based on two assumptions about the allocation of a population's resources between competitive and reproductive functions (Pianka 1974; Gadgil and Bossert 1970; Schaffer and Gadgil 1975). The first is that there is a positive relationship between the amount of re- sources spent on an offspring and the fitness of that offspring. The second assumption is that any species only has a fixed amount of resources avail- able. This results in an inverse relationship be- tween the number of offspring produced and their average fitness. The criterion for success in natural selection is the number of surviving offspring that a parent produces (Crow and Kimura 1970). Therefore, the best reproductive strategy is a compromise between two conflictmg demands: production of the largest possible total number of offspring {r selection), and production of offspring with the highest possible fitness ^ selec- tion). The particular point of compromise for any species will be a function of the selection factors operating on that species and would be that species' position on the r and K continuum. The second part of the theory concerns the rela- tionship between these life history strategies and 1^ Manuscript accepted September 1979. FISHERY BULLETIN; VOL. 78, NO. 1. 1980. FISHERY BULLETIN: VOL. 78, NO. 1 the habitat the species occupies (Southwood et al. 1974; Southwood and Comins 1976). If mortality factors in an environment are variable and/or un- predictable, then their effects are likely to be less selective in terms of population size or of the phenotype involved. Under these circumstances, individual competitive fitness is of relatively less importance. The best strategy would be to place maximal resources into reproduction and produce as many offspring as possible (r selection). The contrasting situation is an environment in which mortality factors are stable and/or predict- able. Mortality under these circumstances will re- sult in strong selection for individual fitness and there will be pronounced differences between their effects on different phenotypes. In these stable environments, the optimal strategy would be to produce offspring with substantial competitive ability {K selection). Due to the previously as- sumed relationship between fitness per offspring and the number of offspring produced, this also means the production of fewer offspring. The two situations described above are end points of a spectrum. Species will always have a number of different selective pressures operating on them, both spatially and temporally. This is particularly evident in aquatic organisms which characteristically go through several life history stages. This again emphasizes that the concept of r and K selection should be applied only in a comparative sense. Finally, comparisons must be made between species of a similar ecological na- ture. Comparisons between species of different ecological types is meaningless since fundamen- tally different types of selective factors will be operating in those cases. r AND K SELECTION IN MARINE FISHES Natural selection will favor nonreproductive ac- tivities at the expense of reproductive activities only when they enhance reproduction at later stages in the life history and thereby maximize overall survival (Crow and Kimura 1970). Changes in allocation of a species' resources from reproductive to competitive activities will only occur in habitats where competitive activities en- hance the survival of future offspring. The result of this is that organisms under different selection pressures will have characteristic life history pat- terns. An r selected species will have life history strategies which tend toward productivity. Thei^ selected species will have life strategies which tend toward efficient exploitation of a specific limiting resource (Pianka 1974). Therefore, specific combinations of population parameters can be identified as being characteristic of an r strategist, while the opposing combination would be characteristic of aK strategist. A species which is exposed to a large component of nonselective or catastrophic mortality (i.e., an r strategist) would be selected for characteristics that would increase productivity. Increasing pro- ductivity through reproductive activity generally implies: 1) early maturity, 2) rapid growth rates, 3) production of larger numbers of offspring at a given parental size, and 4) maximum production of offspring at early age (Gadgil and Bossert 1970). Other characteristics which are results of the allo- cation of large portions of resources to reproduc- tive activity are: 1) small body size, 2) high rates of mortality, and 3) shorter life span (Pianka 1974; Gadgil and Solbrig 1972). In terms of commonly measured population parameters in fishery biol- ogy, an r selected species would have: 1) a low age at first maturity, 2) a high value ofk from the von Bertalanffy growth equation, 3) a small Lx from the von Bertalanffy growth equation, 4) high rates of instantaneous natural mortality (M), and 5) low maximum age. Even in environments with predictable mortal- ity sources, increased allocation of resources to competitive activities will only occur when two prerequisites are met (Schaffer and Gadgil 1975). The first is that reproductive potential increases with some function of age. The second is that there is some additional mortality risk associated with reproduction. Under these assumptions, the attri- butes associated with aK strategist would be: 1) delayed maturity, 2) reduced growth rates, 3) low mortality rates, 4) large body size, and 5) longer life span. Again in terms measured in fishery biol- ogy, aK selected species would have: 1) a high age at first maturity, 2) a low k from the von Ber- talanffy growth equation, 3) a large L-x from the von Bertalanffy growth equation, 4) lowM, and 5) a high maximum age. Using these life history correlates of r and K selection (summarized in Table 1), it is possible to predict the signs of a correlation matrix between life history parameters (Table 2). The predicted matrix can be compared with actual matrices cal- culated using Spearman's rank correlation coefficient. This coefficient only assumes that the observed data are mutually independent and come ADAMS: LIFE HISTORY PATTERNS IN MARINE FISHES Table l .—Summary of hypothetical r and K correlates in life history parameters of fishes. Characteristics r selected K selected Body size. L x' Maximum age Age at first maturity Natural mortality, M Growtti rate./t' Small Low Low High High Large High High Low Low 'The parameter from the von Bertalanffy growth equation was used to represent the actual characteristic. Table 2. — Predicted signs of correlation matrix of life history parameters in fishes. Characteristics Maximum L rJ age Age at first maturity M Body size, L J Maximum age Age at first maturity Natural mortality, M Growth rate, h' 1.0 1.0 + + 1.0 1.0 + 1.0 'The parameter from the von Bertalanffy growth equation was used to represent the actual characteristic. from a continuous bivariate population (Hollan- der and Wolfe 1973). RESULTS Life history parameters were gathered from the literature for several major groups of marine fishes. Often there were multiple sets of data for the same species from different locations. Each set of values was used as a separate data case. The literature citations for the actual parameters are listed by group in Appendix I. Correlation ma- trices were calculated for the following groups of fish: 1) herring and anchovies, Clupeidae and En- graulidae (Table 3), 2) salmons, Salmonidae (Ta- ble 4), 3) cods, Gadidae (Table 5), 4) rockfishes, Table 3. — Correlation coefficients between life-history parameters for herring and anchovies (families Clupeidae and Engraulidae). For sources of data see Appendix I. The number in parentheses represents the significance value for that particular coefficient since the number of data cases was different for each correlation. Age at Table 4. — Correlation matrix between life-history parameters for salmons (family Salmonidae). For sources of data see Appen- dix I. The number in parentheses represents the significance value for that particular coefficient since the number of data cases was different for each correlation. Characteristics L-x' Maximum age Age at first maturity M Body size.Lx' Maximum age Age at first maturity Natural mortality, M Growth rate,/(' 10 0.765 0728 -0785 -0.730 (0,001) (0032) (0 001) (0002) 10 776 -0 737 -0.674 (0.020) (0003) (0.004) 10 -0644 -0812 (0.084) (0013) 1.0 896 (0 001) 1.0 'The parameter from the von Bertalanffy growth equation was used to represent the actual characteristic. Table 5. — Correlation matrix between life-history parameters for cods ( family Gadidae) . For sources of data see Appendix I. The number in parentheses represents the significance value for that particular coefficient since the number of data cases was differ- ent for each correlation. Characteristics L-x' Maximum age Age at first maturity M k-' Body size, L-x' 1.0 0.795 0.833 -0.647 -0.666 Maximum age (0.002) 1.0 (0.001) 0.737 (0.022) -0.654 (0.001) -0.702 Age at first maturity Natural mortality, M Growth rate, k' (0.014) 1.0 (0.028) -0.715 (0.035) 1.0 (0.008) -0.658 (0.008) 0.950 (0.001) 1.0 'The parameter from the von Bertalanffy growth equation was used to represent the actual characteristic. Scorpaenidae, genus Sebastodes (Table 6), and 5) flatfishes, Pleuronectiformes (Table 7). All of the observed correlations agree with the predicted correlations in sign (Table 8). Of the observed correlations, 40 of a possible 46 (or 87%) were significantly different from zero at a 5% probability level. If the observed correlation agreement of coefficients were distributed ran- TABLE 6. — Correlation matrix between life-history parameters for rockfishes (family Scorpaenidae, genus Sebastodes). For sources of data see Appendix I. The number in parentheses represents the significance value for that particular coefficient since the number of data cases was different for each correlation. Characteristics Ly-: age maturity M /c' Characteristics Lx' Maximum age Age at first maturity Body size, Lx' ity M 1.0 0.846 (0.001) 1.0 0.816 (0.001) 0.904 (0.001) 1.0 -0.746 (0.001) -0.797 (0.001) -0.702 (0.001) 1.0 -0.720 (0.001) -0.763 (0.001) -0.732 (0.001) 0.876 (0.001) 1.0 /(' Maximum age Age at first matur Natural mortality. Growth rate./c' Body size, L-x' Maximum age Age at first maturity Growth rate, fc' 1.0 0.662 (0.019) 1.0 0.456 (0.088) 0.612 (0.030) 1.0 -0.490 (0.075) -0.567 (0.040) -0.651 (0.021) 1.0 'The parameter from the von Bertalanffy growth equation was used to represent the actual characteristic. 'The parameter from the von Bertalanffy growth equation was used to represent the actual characteristic. FISHERY BULLETIN. VOL. 78, NO. 1 Table 7. — Correlation matrix between the life-history parameters for flatfishes (order Pleuronectiformes). For sources of data see Appendix I. The number in parentheses represents the significance value for that particular coefficient since the number of data cases was different for each correlation. Characteristics i-x' Maximum age Age at first maturity M k' Body size, Lx' 1.0 0.755 0.956 -0.291 -0.619 Maximum age (0.001) 1.0 (0.001) 0.824 (0.156) -0.355 (0.005) -0.808 Age at first maturity (0.001) 1.0 (0.142) -0.630 (0.001) -0.732 Natural mortality. M (0.014) 1.0 (0.001) 0.367 Growttirate,/(' (0.098) 1.0 'Thie parameter from the von Bertalanffy growth equation was used to represent the actual characteristic. Table 8. — Summary of the number of agreements between pre- dicted and observed correlation coefficients among life-history parameters within selected taxonomic groups. Level of agreement Number in agreement Number possible Percent in agreement Sign 5% probability level 1% probability level 46 40 31 46 46 46 100 87 67 domly (i.e., p = probability of agreement = 0.5, and q = probability of disagreement = 0.5), then the number of agreements would follow a binomial distribution. The binomial test (Hollander and Wolfe 1973 ) can be used to test the hypothesis that the number of agreements between the predicted and observed correlations differs from the number that would have occurred randomly. The number of agreements is significantly different than would have occurred randomly iz = 4.86, P<0.001), when only correlations that were significant at the 5% level were used. maximum age of a fish. From r andK selection, we can predict how these parameters will vary. Con- sider a situation with three hypothetical species: one species will be more r selected, another species will be moreK selected, and another will be inter- mediate between the first two. The biological parameters will vary as shown in Table 9. Bever- ton and Holt yield per recruit curves were calcu- lated for a constant age at first capture {t^ = 4.2 yr) with varying fishing mortality (Figure 1), and for a constant fishing mortality (F = 0.25) with a varying age at first capture (Figure 2). The yield per recruit analysis points up that there are specific differences in fisheries based on r or K selected species. In fisheries based on K selected species, the maximum yield per recruit would occur at a lower level of fishing mortality and at a later age at first entry than in fisheries based on r selected species. The curves also indi- cate that K selected species would be much more sensitive to overfishing both in terms of fishing mortality and age at first entry. The surplus production model of Schaefer com- bines reproductive and mortality functions into one parameter (Ricker 1975). The biological parameters in this model are 5x, the maximum stock size (or carrying capacity in weight), and /e, the instantaneous rate of increase of the stock at densities approaching zero. Again these parame- ters can be predicted for the three hypothetical species from r and if selection (Table 10). In the surplus production model analysis (Figure 3), ther selected species have the highest productivity. As in the yield per recruit analysis, the maximum yield occurs at a lower fishing mortality for the A" RESPONSE OF r AND K SELECTED SPECIES TO HARVESTING The interaction of life history characteristics will have a strong affect on the response of a species to fishing pressure. The Beverton and Holt yield per recruit equation estimates the yield that can be harvested from the growi:h of a cohort. The model assumes that fish grovvi:h is described by the von Bertalanffy growth curve and that mortality processes are exponential (Beverton and Holt 1957; Ricker 1975). The biological parameters in the model are: 1) M, the instantaneous rate of natural mortality, 2) Wy-, the mean asymptotic weight which corresponds to Ly-_, 3) k, the von Bertalanffy growth coefficient, and 4) t^, the Table 9. — Biological parameters for use in yield per recruit analysis for three hypothetical r andK selected species. r selected Intermediate K selected Biological parameters species species species Natural mortality, M 0.30 0.20 0.10 Mean asymptotic weight, w-.. 641 g 1,141 g 1,641 g von Bertalanffy growth coefficient, fc 0.22 0.14 007 Maximum age, (^ 13yr 20 yr 35 yr Table lO. — Biological parameters for surplus production model analysis for three hypothetical r and K selected species. Biological parameters r selected species Intermediate species K selected species Maximum stock size (8 -^J Rate of I.54xi08g 2.04- 10«g 2.54y-[Cfig increase [k) 0.912 0.612 0.312 ADAMS: LIFE HISTORY PATTERNS IN MARINE FISHES 100 .0 FIGURE 1.— The effect of different levels of fishing mortality with constant age of recruitment (4.2 yr) on yield per recruit of three hypothetical fish species demonstrating the range of r and K selection. FIGURE 2.— The effect of different mean ages of recruitment at constant fishing mortality (F = 0.25) on yield per recruit of three hypothetical fish species dem- onstrating the range of r and K selec- tion. ° SPECIES a ' SPECIES B 90.0 « SPECIES C BO .0 70.0 60 .0 50.0 40.0 30.0 • 1 20.0 "^"^■^^"^ — 10.0 ' 0.0, >-^- — , — ^^ — , — , — , . ._ 0.00 .25 .50 .75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 FISHING HDRTPLITY (Fl 200 .0 175 .0 — 150.0 125.0 ° SPECIES ' SPECIES B o SPECIES C 4.0 B.O 12.0 16.0 20.0 24.0 2B.0 32.0 36.0 OGE AT FIRST RECRUITMENT selected species than for the r selected species. The K selected species is reduced to levels lower than the maximum sustainable yield by overfishing much more rapidly than the r selected species. FISHERY BULLETIN: VOL. 78. NO. 1 Figure 3. — Maximum equilibrium yields (x 10* g) from Schaefer surplus production curves as a function of fishing mortality for three hypothetical fish species demonstrating the range of r and K selection. 40 .00 36.10 32 .20 . 2G .30 U3 X; 24.40 X o ■^ 20.50 16 .60 > 12 .70 B.BO 4 .90 1 .00 ° SPECIES « ' SPECIES B ° SPECIES C 0.00 .10 .20 .30 .40 .50 .60 .70 .BO .90 1.00 FISHING MDRTPLITY (F) DISCUSSION Life history parameters vary in consistent pat- terns. These patterns are explainable and predict- able by the theoretical constructs of r and K selec- tion. This is not a particularly new or unique idea in fisheries biology. Beverton and Holt (1959) in- vestigated a positive relationship between body size and life span and between mortality and growth rates. Gushing (1971) suggested that there is a negative relationship between degree of den- sity dependent regulation and fecundity. Alverson and Carney ( 1975) have suggested a positive rela- tionship between body size and the time when a cohort maximizes its biomass. In population ecol- ogy, similar relationships have been investigated for zooplankton (Allan 1976), plants (Gadgil and Solbrig 1972; MacNaughton 1975), and animals (Smith 1954; Bonner 1965). All these empirical observed trends in life history parameters, along with the trends described here, are consistent with r and K selection. It is important to reemphasize here the com- parative nature of r andi^ selection. The r and K continuum is a model and as such occurs only in an idealized sense. The idealized r selected species occurs in an ecological vacuum with no density effects and no competition. The idealized K selected species occurs in a completely saturated ecosystem where densities are high compared with carrying capacities and competition for re- sources is intense. The problem of applying this model to any real situation is not a trivial one. Species are not simply subjected to a single selec- tive pressure, or even to a single set of selective pressures. Because of this, r and /C concepts should only be applied in a comparative sense between groups of species that have some degree of func- tional similarity. No species is r selected or K selected in an absolute sense; it is only relatively more r selected or K selected than some other reference species. This theory will only have value in a situation where the population dynamics of one member of a species group are fairly well un- derstood. The results of the model analysis give several indications about the reaction to harvesting pres- sure of species which are more or less r or K selected. Fisheries based on more r selected species will be more productive. They can be fished at younger ages and at higher levels of fishing mortality. Given a minimum population size, these fisheries should also have a quicker recovery from overfishing. Species which are more r ADAMS: LIFE HISTORY PATTERNS IN MARINE FISHES selected are likely to be strongly influenced by physical forces in the environment (Pianka 1974). Relationships of this type, e.g., between anchovies and upwelling, should be important considera- tions in management plans for these species. Fisheries based on more K selected species will have a high maximum yield per recruit, but there will be fewer fish. Maximum equilibrium yield will occur at later ages of entry into the fishery and at lower levels of fishing mortality. These fisheries would be more susceptible to overfishing and stock depletion. Besides these species' sensitivity to overfishing, more K selected species are much more likely to have sophisticated life history mechanisms (Pianka 1974) which would have to be recognized in a management plan. These mechanisms might include parental care systems such as nesting or live births, mating systems, or territoriality. The more K selected species are much more likely to have strong interspecific rela- tionships, usually competitive ones. The relation- ship between competition and harvesting has been dealt with by Larkin (1963) and Tanner (1975). Additional density independent mortality (fishing mortality) increases the advantage for the popula- tion with a higher population growth rate (i.e., more r selected). Therefore, even low levels of fishing pressure can destabilize a previously sta- ble competitive pair and result in decline of the harvested species. Interestingly, the opposite re- sult is also possible; harvesting pressure can stabilize a previously unstable species pair as Slobodkin ( 1962) found with experimental popula- tions of hydra. Fisheries based on more r selected species are likely to be of a boom and bust nature. Although in some years catches in these fisheries will be very large, they will be characterized by erratic produc- tion levels. The most efficient form of harvesting these fisheries will be fleets which are capable of switching between a number of target species rel- atively quickly. Fisheries based on more K selected species, in contrast to the boom and bust nature of r selected fisheries, will be characterized by relatively stable population sizes and therefore catch levels. Given some initial measure of year class strength, possi- bly through larval or prerecruitment surveys, the prediction of future catches from that fishery could be made with a fair degree of accuracy. However, once fisheries based on these species become over- fished, it would require a long period for the stock to rebuild to levels which can support economical profitable fisheries. An extremely K selected species would only be suitable for trophy fisheries. Fisheries based on r and K selected species have been discussed in a comparative sense, but preda- tion (in the case of a fishery , human predation) will also have effects on an individual species. The gene pool of any species is going to contain within it some range of variation of both r and/C selected traits. The effects of increasing fishing mortality, which is assumed to be density independent (Gushing 1975), on life history characteristics has been theoretically analyzed by Roughgarden (1971). The general effect is an increase in selec- tive advantage for the r selected proportions of the gene pool. This would mean an increase in growth rates, reduced age at first maturity, and greater fecundity at age. These trends will be more con- spicuous in species that are relatively more K selected. Species that are more strongly r selected are likely to have less range of variation in this direction. One example of these effects of preda- tion pressure is a comparison of lake trout, Sal- velinus namaycush , populations under heavy pre- dation pressure from the freshwater harbor seal, Phoca uitulina, to populations in nearby lakes without seals (Power and Gregoire 1978). The lake trout populations which were preyed upon by seals had faster growth rates, small maximum body size, reduced maximum age, lower age at sexual maturity, and greater individual fecundity com- pared with populations in lakes without seals. Growth and maturation rates of certain seal species have also increased where populations have been reduced by fisheries (Sergeant 1973). These affects can be attributed to changes in selec- tion pressure resulting from sustained harvesting. In summary, r and K selection seems to have been an important evolutionary trend on marine fish populations. The basic hypotheses are con- firmed by the data presented here. The result of patterns in population parameters which arise from r and K selection is that different manage- ment strategies would be appropriate. The value of this approach is likely to be in initial stages of development of a fishery. As a fishery becomes more developed and more specific information be- comes available, a more refined management strategy would become possible. ACKNOWLEDGMENTS This paper benefited from readings by M. E. Adams, E. O. Garton, E. S. Hobson, W. H. Lenarz, FISHERY BULLETIN: VOL. 78, NO. 1 and H. Li. Naturally any errors in the paper are the sole responsibility of the author. LITERATURE CITED Alander, H. 1950. Baltic herring. Ann. Biol. 6:191-192. ALLAN, J. D. 1976. Life history patterns in zooplankton. Am. Nat. 110:165-180. Alverson, D. L., and M. J. Carney. 1975. A graphic review of the growth and decay of popula- tion cohorts. J. Cons. 36:133-143. ARORA, H. L. 1951. An investigation of the California sand dab, Citharichthys sordidus (Girard). Calif. Fish Game 37:3-42. AYUSHIN, B. N. 1963. Abundance dynamics of herring populations in the seas of the Far East, and reasons for the introduction of fishery regulations. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 154:262-269. B.arrett, I., AND G. V. Howard. 1961. Studies of the age, growth, sexual maturity and spawning of populations of anchoveta (Cetengraulis mys- ticetus) of the coast of the eastern tropical Pacific Ocean. Inter-Am. Trop. Tuna Comm. Bull. 5:113-165. BEN-TUVIA, A. 1960. Synopsis of biological data on Sardinella aurita of the Mediterranean Sea and other waters. In H. Rosa, Jr. and G. Murphy (editors), Proc. World Sci. Meet. Biol. Sardines Relat. Species 2:287-312. FAO, Rome. BEVERTON, R. J. H. 1963. Maturation, growth and mortality of clupeid and engraulid stocks in relation to fishing. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 154:44-67. BEVERTON, R. J. H., AND S. J. HOLT. 1957. On the dynamics of exploited fish popula- tions. Fish. Invest. Minist. Agric, Fish. Food (G.B.), Ser. II, 19, 533 p. 1959. A review of the life spans and mortality rates in nature, and their relation to growth and other physiologi- cal characteristics. Ciba Found. Colloq. Ageing 5:142- 177. Blackburn, M. 1950. The Tasmanian whitebait, Lovettia seali (Johnson), and the whitebait fishery. Aust. J. Mar. Freshwater Res. 1:155-198. Bonner, J. T. 1965. Size and cycle, an essay on the structure of biolo- gy. Princeton Univ. Press, Princeton, N.J., 219 p. Bough, p. 1952. La cruissance des poissons mediterrancius. Vie Milieu, Suppl. 2, 2:118-146. Bowers, A. B. 1963. Recent changes in the Manx herring fishery. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 154:220-226. Burd, a. C. 1962. Growth and recruitment in the herring of the south- em North Sea. Fish. Invest. Minist. Agric, Fish. Food (G.B.),Ser. II,23(5):l-42. Clark, F. N., and J. C. Marr. 1955. Populationdynamicsof the Pacific sardine. Calif. Ocean. Fish. Invest. Prog. Rep. 1 July 1953 to 31 March 1955, p. 11-48. Clark, F. N., and J. B. Phillips. 1952. The northern anchovy (Engraulis mordax mordax) in the California fishery. Calif. Fish Game 38:189-207. CLAYDEN, A.C. 1972. Simulation of changes in abundance of the cod (Gadus morhua L.) and the distribution of fishing in the North Atlantic. Fish. Invest. Minist. Agric, Fish. Food (G.B.), Ser. II, 27(1), 58 p. Crow, J. F., and M. Kimura. 1970. An introduction to population genetics theory. Harper and Row, N.Y., 591 p. CULLEY, M. 1971. The pilchard, biology and exploitation. Pergamon Press, N.Y.,241p. Gushing, d. H. 1959. On the effect of fishing on the herring of the southern North Sea. J. Cons. 24:283-307. 1961. On the failure of the Plymouth herring fishery. J.' Mar. Biol. Assoc. U.K. 41:799-816. 1971. The dependence of recruitment on parent stock in different groups of fishes. J. Cons. 33:340-362. 1975. Marine ecology and fisheries. Camb. Univ. Press, N.Y.,278p. Da VIES, D. H. 1958. The South African pilchard (Sardinops ocellata), preliminary report on the age composition of the commer- cial catches, 1950-55. Union S. Afr., Dep. Commer. In- dust., Div. Fish. Invest. Rep. 33, 20 p. Day, L. R. 1957. Populations of herring in the southern Gulf of St. Lawrence. In A. H. Leim, S. N. Tibbo, L. R. Day, L. Lauzier, R. W. Trites, H. B. Hachey, and W. B. Bailey, Report of the Atlantic herring investigation committee, p. 121-137. Fish. Res. Board Can., Bull. 111. DEASON, H. J., AND R. HILE. 1947. Age and growth of the Kiyi, Leucichthys kiyi Koelz, in Lake Michigan. Trans. Am. Fish. Soc. 74:88-142. DEJAGER, B. V. D. 1960. Synopsis on the biology of the South African pil- chard, Sardinops ocellata (Pappe). In H. Rosa, Jr. and G. Murphy (editors), Proc. World Sci. Meet. Biol. Sardines Relat. Species 2:97-114. FAO, Rome. DICKIE, L. M., AND F. D. MCCRACKEN. 1955. Isopleth diagrams to predict equilibrium yields of a small flounder fishery. J. Fish. Res. Board Can. 12:187- 209. DRUCKER, B. 1972. Some life history characteristics of coho salmon of the Karluk River system, Kodiak Island, Alaska. Fish. Bull, U.S. 70:79-94. ELWERTOWSKI, J. 1957. Polish investigations on the sprat from the southern Baltic. Ann. Biol. 12:216. 1958. The sprat in the southern Baltic. Ann. Biol. 13:229-230. 1959. Polish investigations on Baltic sprat in 1957. Ann. Biol. 14:202-203. 1960. Polish investigations on sprat from the southern part ofthe Central Baltic -1958. Ann. Biol. 15:192-193. Face, L. 1920. "Engraulidae, Clupeidae". Rep. Danish oceanogr. Exped. Medit., 2(A.9). (Cited in Beverton 1963.) 8 ADAMS: LIFE HISTORY PATTERNS IN MARINE FISHES FAURE, L. 1950. Le sprat de la Bale de Douamez peche, biometric et biologie. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 126:96-102. FLEMING, A. M. 1960. Age, growth and sexual maturity of cod (Gadus morhua L.) in the Newfoundland area, 1947-1950. J. Fish. Res. Board Can. 17:775-809. FOERSTER, R. E. 1968. The sockeye salmon, Oncorhynchus nerka. Fish. Res. Board Can., Bull. 162, 422 p. Fridriksson, a. 1950. On the herring of the north coast of Iceland during the summer of 1949. Ann. Biol. 6:162-167. 1951 . On the herring and the herring fishery of the north coast of Iceland during the summer of 1950. Ann. Biol. 7:122-125. 1952. The Icelandic north coast herring in 1951. Ann. Biol. 8:136-140. 1953. The Icelandic north coast herring in 1952. Ann. Biol. 9:164-167. 1954. The Icelandic north coast herring in 1953. Ann. Biol. 10:143-147. 1956. The Icelandic north coast herring in 1954. Ann. Biol. 11:114-118. 1957. The Icelandic north coast herring in 1955. Ann. Biol. 12:158-162. 1958. The Icelandic north coast herring in 1956. Ann. Biol. 13:173-176. 1959. The Icelandic north coast" herring in 1957. Ann. Biol. 14:149-154. 1960. The Icelandic north coast herring in 1958. Ann. Biol. 15:126-131. 1961. The Icelandic north coast herring in 1959. Ann. Biol. 16:163-167. FURNESTIN, J. 1945. "Note preliminaire sur I'anchois (Engraulis encras- sicholus L.) du Giolfe de Gascogne". Rev. Trav. Inst. Pech. Marit. 13:197-209. (Cited in Beverton 1963.) GADGIL, M., AND W. H. BOSSERT. 1970. Life historical consequences of natural selec- tion. Am. Nat. 104:1-24. GADGIL, M., AND O. T. SOLBRIG. 1972. The concept of r- and iiT- selection: evidence from wild flowers and some theoretical considerations. Am. Nat. 106:14-31. GILIS, C. 1957. Fall-herring concentrations exploited by the Bel- gian herring trawlers in 1955. Ann. Biol. 12:204-208. 1958. The Belgian herring fisheries in 1956—1957 (30. July 1956 - 19. January 1957), the North Sea - English Channel. Ann. Biol. 13:205-209. 1959. The Belgian herring fisheries in 1957—1958, the North Sea - English Channel. Ann. Biol. 14:182-188. 1960. TheBelgianherringfisheriesin 1958—1959. Ann. Biol. 15:176-181. 1961. TheBelgianherringfisheriesin 1959—1960. Ann. Biol. 16:212-215. Hanamura, N. 1953 . On the herring resources of Hokkaido and the South Saghalen. Bull. Jpn. Soc. Sci. Fish. 19:283-291. HANNERZ, L. 1956. Preliminary results of the herring investigations in the Bothnian Sea 1954. Ann. Biol. 11:156-158. Hart, J. L. 1931. The growth of the whitefish Coregonus clupeaformis (Mitchell). Contrib. Can. Biol. 6:427-444. 1948. Age and growth rate in the butter sole, Isopsetta isolepis. Trans. R. Soc. Can., Ser. 3, Sect. 5, 42:65-72. HAYASHI, S. 1961. Fishery biology of the Japanese anchovy, Engraulis japonica (Houttuyn). Bull. Tokai Reg. Fish. Res. Lab. 31:145-268. HAYASHI, S., AND K. KONDO. 1957 . Growth of the Japanese anchovy — IV. Age determi- nation with use of scales. Bull. Tokai Reg. Fish. Res. Lab. 17:31-64. HILE, R. 1936. Age and growth of the cisco, Leucichthys artedi (Le Sueur), in the lakes of the northeastern highlands, Wis- consin. Bull. U.S. Bur. Fish. 48:211-317. HILE, R., AND H. J. DEASON. 1934. Growth of the whitefish Coregonus clupeaformis (Mitchell), in Trout Lake, northeastern highlands, Wis- consin. Trans. Am. Fish. Soc. 64:231-237. Hodgson, W.C. 1957. The herring and its fishing. Routledge & Kegan Paul, Lond., 197 p. Hodgson, W. C, and I. D. Richardson. 1949. The experiments on the Cornish pilchard fishery in 1947-8. Fish. Invest. Minist. Agric, Fish. Food (G.B.), Ser. II, 27(2), 21 p. HOLLANDER, M., AND D. A. WOLFE. 1973. Nonparametric statistical methods. Wiley, N.Y., 503 p. HOLT, S. J. 1962. The application of comparative population studies to fisheries biology — an exploration. In E. D. Le Cren and M. W. Holdgate (editors). The exploitation of natural animal populations, p. 51-69. Br. Ecol. Soc. Symp. 2. JENSEN, A. J. C. 1947. The herring at Bomholm. Ann. Biol. 2:155-158. KENNEDY, W. A. 1943. The whitefish, Coregonus clupeaformis {Mitchell ), of Lake Opeongo, Algonquin Park, Ontario. Publ. Ont. Fish. Lab. 62:21-66. 1953. Growth, maturity, fecundity and mortality in the relatively unexploited whitefish, Coregonus clupeafor- mis, of Great Slave Lake. J. Fish. Res. Board Can. 10:413-441. 1954. Growth, maturity and mortality in the relatively unexploited lake trout, Cristivomer namaycush , of Great Slave Lake. J. Fish. Res. Board Can. 11:827-852. KETCHEN, K. S. 1964. Preliminary results of studies on growth and mortal- ity of Pacific cod [Gadus macrocephalus) in Hecate Strait, British Columbia. J. Fish. Res. Board Can. 21:1051- 1067. KETCHEN, K. S., AND C. R. FORRESTER. 1966. Population dynamics of the petrale sole, Eopsetta jordani , in waters off western Canada. Fish. Res. Board Can., Bull. 153, 195 p. Larkin, p. a. 1963. Interspecific competition and exploitation. J. Fish Res. Board Can. 20:647-678. Larraneta, M. G. 1960. Synopsis of biological data on Sardina pilchardus of the Mediterranean and adjacent seas. In H. Rosa, Jr. FISHERY BULLETIN: VOL. 78, NO. 1 and G. Murphy (editors), Proc. World Sci. Meet. Biol. Sardines Relat. Species 2:137-173. FAO, Rome. Lea, E. 1919. Age and growth of herring in Canadian waters. In J. Hjort (director), Canadian fisheries expedition, 1914- 1915, investigations in the Gulf of St. Lawrence and At- lantic waters of Canada, p. 75-164. J. de Labroquerie Tache, Ottawa. MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography. Princeton Univ. Press, Princeton, N.J., 203 p. MacKinnon, J. C. 1973. Analysis of energy flow and production in an un- exploited marine flatfish population. J. Fish. Res. Board Can. 30:1717-1728. MACNAUGHTON, S. J. 1975. r- and iC-selection in Typha. Am. Nat. 109:251- 261. MEN0N,M. D. 1950. Bionomics of the poor-cod (Gadus minutus L.) in the Plymouth area. J. Mar. Biol. Assoc. U.K. 29:185-240. Miller, D. J., A. E. Daugherty, F. E. Felin, and J. Mac- Gregor. 1955. Age and length composition of the northern anchovy catch off the coast of California in 1952-53 and 1953- 54. Calif. Dep. Fish Game, Fish Bull. 101:37-66. Miller, D. J., and r. S. wolf. 1958. Age and length composition of the northern anchovy catch off the coast of California in 1954-55, 1955-56, and 1956-57. Calif. Dep. Fish Game, Fish Bull. 106:27-72. MOLANDER, A. R. 1943. Sprat and milieu-conditions. Ann. Biol. 1:165-174. MOSHER, K. H., AND H. H. ECKLES. 1954. Age determination of Pacific sardines from oto- liths. U.S. Fish Wildl. Serv., Res. Rep. 37, 40 p. MOTODA, S., AND Y. HiRANO. 1963. Review of Japanese herring investigations. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 154:249-261. MURPHY, G. L 1966. Population biology of the Pacific sardine (Sardinops caerulea ). Proc. Calif. Acad. Sci., Ser. 4, 34:1-84. NAIR, R. V. 1960. Synopsis on the biology and fishery of the Indian sardines. In H. Rosa, Jr. and G. Murphy (editors), Proc. World Meet. Biol. Sardines Relat. Species 2:329-414. Nielsen, J. 1960. Preliminary results of tagging experiments with herring (Clupea harengus L.) in Greenland. J. Cons. 26:73-79. Parrish, B. B., and R. E. Craig. 1963. The herring of the north-western North Sea, post> war changes in the stock fished by Scottish drifters. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 154:139-158. Phillips, J. B. 1948. Growth of the sardine, Sardinops caerulea , 1941-42 through 1946-47. Calif. Div. Fish Game, Fish Bull. 71, 33 p. 1964. Life history studies on ten species of rockfish (genus Sebastodes). Calif. Dep. Fish Game, Fish Bull. 126,70 p. PlANKA, E. R. 1974. Evolutionary ecology. Harper & Row, N.Y., 356 p. Pinhorn, a. T, 1969. Fishery and biology of Atlantic cod (Gadus morhua ) off the southwest coast of Newfoundland. J. Fish. Res. Board Can. 26:3133-3164. 10 postel, e. 1955. Resume des connaissances acquises sur les clupeides de rOuest^Africain. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 137:14-16. POSTUMA, K. H. 1963. The catch per unit effort and mortality rates in the Southern Bight and Channel fisheries. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 154:190-197. Power, G., and J. Gregoire. 1978. Predation by freshwater seals on the fish community of lower Seal Lake, Quebec. J. Fish. Res. Board Can. 35:844-850. POWLES, P. M. 1965. Life history and ecology of American plaice (Hippo- glossoides platessoides F.) in the Magdalen Shallows. J. Fish. Res. Boaid Can. 22:565-598. 1969. Size changes, mortality, and equilibrium yields in an exploited stock of American plaice (Hippoglossoides platessoides). J. Fish. Res. Board Can. 26:1205-1235. Raitt, D. S. 1939. The rate of mortality of the haddock of the North Sea stock, 1919-1938. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 110:65-79. Richardson, I., G. Vazzoler, A. De Faria, and M. De MORAES. I960. Report on sardine investigations in Brazil. In H. Rosa, Jr. and G. Murphy (editors), Proc. World Sci. Meet. Biol. Sardines Relat. Species 3:1051-1079. FAO, Rome. RICKER, W. E. 1949. Mortality rates in some little-exploited populations of freshwater fishes. Trans. Am. Fish. Soc. 77:114-128. 1958. Handbook of computations for biological statistics of fish populations. Fish. Res. Board Can., Bull. 119, 300 p. 1975. Computation and interpretation of biological statis- tics offish populations. Fish. Res. Board Can., Bull. 191, 382 p. ROBERTSON, J. A. 1938. The sprat and the sprat fishery of England. Fish. Invest. Minist. Agric, Fish. Food (G.B.), Ser. II, 16(2), 103 p. ROSSI GNOL, M. 1955. Premieres observations sur la biologie des Sardinel- les dans la region de Pointe-Noire. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 137:16-20. ROUGHGARDEN, J. 1971. Density-dependent natural selection. Ecology 52:453-468. SCHAFFER, W. M., AND M. D. GADGIL. 1975. Selection for optimal life histories in plants. In M. L. Cody and J. M. Diamond (editors). Ecology and evolu- tion of communities, p. 142-157. Harv. Univ. Press, Camb., Mass. SERGEANT, D. E. 1973. Environment and reproduction in seals. J. Rep rod. Fertil. Suppl. 19:555-561. SHAPOVALOV, L., AND A. C. TAFT. 1954. The life histories of the steelhead rainbow trout iSalmo gairdneri gairdneri) and silver salmon iOn- corhynchus kisutch) with special reference to Waddell Creek, California, and recommendations regarding their management. Calif. Dep. Fish Game, Fish Bull. 98, 375 p. SlLLIMAN,R. P. 1943. Studies on the Pacific pilchard or sardine {Sardinops caerulea). 5. — A method of computing mortalities and re- ADAMS: LIFE HISTORY PATTERNS IN MARINE FISHES placements. U.S. Fish Wildl. Serv., Spec. Sci. Rep. 24, 10 p. Slobodkin, L. B. 1962. Growth and regulation of animal populations. Holt, Rinehart and Winston, N.Y., 184 p. SMITH, F. E. 1954. Quantitative aspects of population growth. InE.J. Boell (editor). Dynamics of growth process, p. 277-294. Princeton Univ. Press, Princeton, N.J. Smith, W.C. 1957. Gonad condition, age and size of Manx herrings, 1945-53, and comparison with earlier years. Annu. Rep. Mar. Biol. Stn. Pt. Erin. 69:21-28. SOUTHWOOD, T. R. E., AND H. N. COMINS. 1976. A synoptic population model. J. Anim. Ecol. 45:949-965. SOUTHWOOD, T. R. E., R. M. May, M. P. Hassell, and G. R. Conway. 1974. Ecological strategies and population parameters. Am. Nat. 108:791-804. SUND, O. 1943a. The age-composition of the Norwegian spawn her- ring observed during 36 years. Ann. Biol. 1:45-49. 1943b. The size of the Norwegian spawn herring. Ann. Biol. 1:50-51. TANAKA, S. 1960. Studies on the djTiamics and the management offish populations. [In Jpn., Engl, summ.] Bull. Tokai Reg. Fish. Res. Lab. 28:1-200. Tanner, J. T. 1975. The stability and the intrinsic growth rates of prey and predator populations. Ecology 56:855-867. Taylor, C. C. 1958. Cod growth and temperature. J. Cons. 23:366-370. Tester, a. L. 1955. Estimation of recruitment and natural mortality rate from age-composition and catch data in British Co- lumbia herring populations. J. Fish. Res. Board Can. 12:649-681. TIBBO, S. N. 1956. Populations of herring (Clupea harengus L.) in New- foundland waters. J. Fish. Res. Board Can. 13:449-466. 1957a. Herring of the Chaleur Bay area, /n A. H. Leim, S. N. Tibbo, L. R. Day, L. Lauzier, R. W. Trites, H. B. Hachey, and W. B. Bailey, Report of the Atlantic herring investigation committee, p. 85-102. Fish. Res. Board Can., Bull. 111. 1957b. Herring populations on the south and west coasts of Newfoundland. In A. H. Leim, S. N. Tibbo, L. R. Day, L. Lauzier, R. W. Trites, H. B. Hachey, and W. B. Bailey, Report of the Atlantic herring investigation committee, p. 153-164. Fish. Res. Board Can., Bull. 111. TOKAI Regional Fisheries Research Laboratory. I960. Synopsis on the biology oi Sardinops melanosticta (Temminck and Schlegel). In H. Rosa, Jr. and G. Mur- phy (editors), Proc. World Sci. Meet. Biol. Sardines Relat. Species 2:213-244. FAO, Rome. Van Cleve, R., and D. E. Bevan. 1973. Evaluation of causes for the decline of the Karluk sockeye salmon runs and recommendations for rehabilita- tion. Fish. Bull., U.S. 71:627-649. Wat AN ABE, K. 1958. Growth of the anchovy in the Japan Sea. Annu. Rep. Jpn. Sea Fish. Res. Lab. 4:147-152. APPENDIX I: LITERATURE CITATIONS FOR POPULATION PARAMETERS BY SPECIES Herring and Anchovies, Families Clupeidae and Engraulidae Clupea harengus — Lea 1919; Sund 1943a, b; Jen- sen 1947; Fridriksson 1950, 1951-61; Alander 1950; Tibbo 1956, 1957a, b; Hannerz 1956; Gilis 1957-61; Smith 1957; Day 1957; Gushing 1959; Nielsen 1960; Burd 1962; Parrish and Craig 1963; Postuma 1963; Bowers 1963. C. pallasii — Hanamura 1953; Tester 1955; Kicker 1958; Tanaka 1960; Ayushin 1963; Motoda and Hirano 1963. Sprattus sprattus — Robertson 1938; Molander 1943; Faure 1950; Elwertowski 1957-60. Sardinops caerulea — Silliman 1943; Phillips 1948; Mosher and Eckles 1954; Clark and Marr 1955; Murphy 1966; Culley 1971. S. melanosticta — Tanaka 1960; Tokai Regional Fisheries Research Laboratory 1960. S. neopilchardus — Blackburn 1950. S. ocellata — Davies 1958; De Jager 1960; Culley 1971. Sardina pilchardus — Hodgson and Richardson 1949; Bough 1952; Hodgson 1957; Larraneta 1960; Cushing 1961; Culley 1971. Sardinella aurita — Postel 1955; Rossignol 1955; Richardson et al. 1960; Ben-Tuvia 1960; Bever- ton 1963. S. longiceps — Nair 1960. Engraulis encrasicholus — Fage 1920; Fumestin 1945. E. japonicus — Hayashi and Kondo 1957; Watanabe 1958; Tanaka 1960; Hayashi 1961. E. mordax mordax — Clark and Phillips 1952; Mil- ler et al. 1955; Miller and Wolf 1958; Culley 1971. Cetengraulis mysticetus — Barrett and Howard 1961. 11 Salmons, Family Salmonidae Coregonus clupeaformis — Hart 1931; Hile and Deason 1934; Kennedy 1943, 1953; Ricker 1949. Cristivomer namaycush — Kennedy 1954. Leucichthys artedii — Hile 1936. L. kiyi — Deason and Hile 1947. Oncorhynchus kisutch — Shapovalov and Taft 1954; Drucker 1972. O. nerka — Foerster 1968; Van Cleve and Bevan 1973. Cods, Family Gadidae Boreogadus saida — Beverton and Holt 1959. Gadus callarias — Beverton and Holt 1957; Taylor 1958. G. macrocephalus — Ketchen 1964. G. minutus — Menon 1950. G. morAaa— Fleming 1960; Pinhorn 1969; Clayden 1972. G. virens — Beverton and Holt 1959. Melanogrammus aeglefinus — Raitt 1939; Bever- ton and Holt 1957. Merluccius merluccius — Beverton and Holt 1959. FISHERY BULLETIN: VOL. 78, NO. 1 Rockfishes, Family Scorpaenidae, Genus Sebastodes Sebastodes crameri — Phillips 1964. S. diploproa — Phillips 1964. S. entomelas — Phillips 1964. S. /Zaf^fdus— Phillips 1964. S. ^oodei— Phillips 1964. S.jordani — Phillips 1964. S. miniatus — Phillips 1964. S. paucispinis — Phillips 1964. S. pinniger — Phillips 1964. S. saxicola — Phillips 1964. Flatfishes, Order Pleuronectiformes Citharichthys sordidus — Arora 1951. Eopsetta jordani — Ketchen and Forrester 1966. Hippoglossus platessoides — Powles 1965, 1969; MacKinnon 1973. H. vulgaris — Beverton and Holt 1959. Isopsetta isolepis — Hart 1948. Pleuronectes platessa — Beverton and Holt 1959. Pseudopleuronectes americanus — Dickie and McCracken 1955. Solea vulgaris — Beverton and Holt 1957. 12 SPECIES OF MUNIDOPSIS (CRUSTACEA, GALATHEIDAE) OCCURRING OFF OREGON AND IN ADJACENT WATERS Julie W. Ambler' ABSTRACT Twelve species of Munidopsis (Decapoda: Crustacea: Galatheidae) were collected from the continental slope, Cascadia Basin, and Tufts Abyssal Plain off Oregon and in adjacent waters. Three new species are described: Munidopsis cascadia, M. tuftsi, and M. yaquinensis . One specimen, Munidopsis sp., closely related to M. bairdii, is described but unnamed, pending capture of more specimens. Munidop- sis chacei is synonymized with M. bairdii and M. geyeri is synonymized with M. subsquamosa . The ranges of seven previously described species are now extended to Oregon and Washington: M. aries, M. bairdii, M. beringana, M. ciliata, M. latirostris, M. subsquamosa, and M. verrucosus. The 12 species occurred between 950 and 4,194 m; 3 species were found on the continental slope (950-2,189 m); 9 species were found on Cascadia Basin (1,900-3,025 m); and 3 species were found on Tvifts Plain (3,390-4,194 m). Species composition on Cascadia Basin differed from east to west. The highest densities (number of specimens per trawl) occurred at the base of the continental slope and 40 miles farther west. One species, M. latirostris, contributed 73.0% of the total number of specimens, and three other species (M. bairdii, M. ciliata , and M. subsquamosa ) contributed an additional 20 .2% . The species collected also occur in the Atlantic (M. bairdii, M. aries), tropical Pacific and Indian (M. ciliata), tropical Pacific (A/, latirostris), Arctic (M. beringana), southern temperate Pacific (M. verrucosus), or are cosmopolitan (Af. subsquamosa), or are endemic on Cascadia Basin (M. cascadia, M. yaquinensis), and on Tufts Plain (M. tuftsi). Species of Munidopsis are found from intertidal waters to the abyssal plains of the deep sea. Munidopsis polymorpha is found in saltwater lakes in caverns connected to the sea in the Ca- nary Islands (Dinkins 1969). Munidopsis crassa, the deepest known species, was found at 4,700 m in the Bay of Biscay (Sivertson and Holthuis 1956). Recently, an unidentified Munidopsis sp. has been found near submarine hot springs near the Galapagos (Corliss and Ballard 1977). In general, the genus is found in the deep sea with about half of the known species occurring deeper than 800 m (Doflein and Balss 1913). In the eastern Pacific Ocean, the first Munidop- sis species were collected off Chile by the Chal- lenger (Henderson 1888) and in the eastern tropi- cal Pacificby the A /6a^ross (Faxon 1895). Benedict (1902) described additional new species collected by the Albatross off southern California and the Galapagos, and in the Bering Sea. Since then, Bahamonde (1964) and Khodkina (1973) have found new species off Chile, and Pequegnat and Pequegnat ( 1973) described a new species off Baja California and Costa Rica from the Albatross and 'School of Oceanography, Oregon State University, Corvallis, Oreg.; present address: Department of Oceanography, Texas A&M University, College Station, TX 77843. Manuscript accepted August 1979. FISHERY BULLETIN; VOL. 78, NO. 1, 1980. Galathea collections. Little work has been done on Munidopsis occurring off the west coast of the United States. Schmitt ( 1921), in a key to Muni- dopsis species found off California, included M. verrilli, M. hystrix, M. aspera, and M. quadrata. Haig (1956) modified this key to include M. de- pressa. This paper discusses the taxonomy and distribu- tion of 12 Munidopsis species collected mainly off Oregon from 950 to 4,194 m. These depths include the lower slope and the abyssal plains of Cascadia Basin and Tufts Plain (Figure 1). Among species found off Oregon, only M. quadrata has previously been collected from the west coast of the United States. Three new species are described: M. cas- cadia, M. tuftsi, and M. yaquinensis. One species, Munidopsis sp., is described but left unnamed until more specimens become available to eluci- date its relationship to M. bairdii. The ranges of seven previously described species are extended to Oregon: M. bairdii, M. beringana, M. ciliata, M. aries, M. verrucosus, M. latirostris, and M. sub- squamosa. METHODS A total of 803 specimens were collected from 146 13 ^3 FISHERY BULLETIN: VOL. 78, NO. 1 TUFTS PLAIN TP-C TP-9 TP-A Figure l. — Stations sampled in Cascadia Basin and Tufts Plain, off Oregon and Washington. benthic otter trawls (0TB, OT) and beam trawls (BMT) during 13 yr of sampling by Andrew G. Carey, Jr., School of Oceanography, Oregon State University. On Cascadia Basin, samples were col- lected on three north-south lines ranging from the CP-1 line (Figure 1) at the base of the continental slope to the CP-3 line 80 mi farther offshore. The base of the continental slope varies from 1,900 m on the Astoria Fan at CP-l-A to 2,816 m at CP-l-E. At the base of the continental slope farther south, between lat. 43° and 44° N, a small trench occurs in which depths reach 3,000 m. Stations become deeper both to the south and to the west in Cas- cadia Basin. One sample was taken from Gorda Ridge off California, south of Cascadia Basin. Ten tows were made in northern Cascadia Basin on Nitinat Fan off Washington, during a study of deepwater dumpsites (Carey et al. 1973). Three areas (TP-C, TP-B, TP-A) were sampled on Tufts Plain. Station abbreviations are as follows: NAD = Newport hydrographic line, mainly slope sta- tions; CP = Cascadia Basin, off Oregon; TP = Tufts Plain; and DWD = Deepwater dumpsite, northern Cascadia Basin, off Washington. The beam trawl is a semiquantitative sampler (Carey and Heyamoto 1972), with a rigid frame of steel skids connected by a 3 m aluminum pipe, with a collecting net of 4.1 cm stretch mesh lined with 1.3 cm mesh net. The otter trawl is a 7 m semiballoon Gulf of Mexico shrimp trawl with 4.1 cm stretch mesh with a 1.3 cm mesh liner. Samples were preserved at sea in 10% formaldehyde and sorted in the laboratory. The specimens were examined through a dis- secting microscope and measured with vernier calipers usually to the nearest millimeter. The following measurements were used (Figure 2): Carapace length (CL) = tip of rostrum to middle of posterior margin of carapace. Anterior width of carapace = width between an- terolateral spines. Posterior width of carapace = width at posterior margin of carapace. Rostrum length = tip of rostrum to rostrum base, which lies on an imaginary line between the bases of the ocular peduncles. Cheliped length = tip of chela to articulation between ischium and sternum. Chela length = tip of chela to articulation be- tween chela and carpus, on the ventral side. Eyespine length = tip of eyespine to proximal end of cornea. Incomplete synonymies are given for each species. References include original description, first redescription if the original description was very short, first figure, and all synonyms. The specimens from Oregon State University (OSU) were compared with those borrowed from the U.S. National Museum (USNM), the Museum of Comparative Zoology at Harvard (MCZ), and from Texas A&M University (TAMU). Specimens of each species are cataloged in the Oregon State University Benthic Invertebrate Museum (OSUBI). The holot5^es and a few paratypes of the new species were deposited at the U.S. National Museum. MUNIDOPSIS WHITE AVES 1874 Munidopsis Whiteaves 1874:212 (original de- scription); Smith 1885:493 (synonomy with Galacantha); Milne-Edwards and Bouvier 1894:271, 1897:63 (redescriptions); Faxon 1895:81-83 (synonomy with Orophorhynchus, Elasmonotus, Galathodes, and Anoplonotus); Chace 1942:69 (synonomy with Galacantha). Galathodes Milne-Edwards 1880:53 (original de- scription); Milne-Edwards and Bouvier 1894:276, 1897:94 (redescriptions). Orophorhynchus Milne-Edwards 1880:58 (origi- nal description); Milne-Edwards and Bouvier 1894:283, 1897:110 (redescriptions). Elasmonotus Milne-Edwards 1880:60 (original description); Milne-Edwards and Bouvier 1894:279, 1897:98 (redescriptions); Henderson 1888:165 (synonomy with GaZaf/iopsis). 14 AMBLER: SPECIES OF A/ l/MDOPS/S OFF OREGON CHELIPED SECOND ANTENNAE INNER EYESPINE OUTER EYESPINE ANTEROLATERAL SPINE "^^^^^NAL SPINE ANTERIOR BRANCH OF CERVICAL GROOVE FIRST LATERAL SPINE POSTERIOR BRANCH OF CERVICAL GROOVE AMBULATORY LEGS ^ ABDOMINAL SEGMENTS | In Xi — In a a P(d2^ Z <2i Amended to Equation (1.2), this would imply a one-period net benefit of pix^ - y^) - y[z - (x^ - yi)V - e[(x^ - y^) -zV where (a)^ denotes the positive part of a. An alter- nate form is to let e = (y-e)/2andc = (y+e)/2.Then the one-period return is: pix^ -y^) + e(x^-y^)-c\{Xf-yi)-z\ - ez. One advantage to the smoothing cost approach over other approaches is that p, e, and c can be normalized so as to be interpreted as relative prices. That is, the normalized values p - 1, elp, and dp can be interpreted as the value of having the between period harvest "smoothed" by one unit relative to the value of one unit of additional har- vest. Actual relative values are often difficult to determine. But by parameterizing on e and c, it is possible to present a decision maker not only a range of possible "optimal" policies and their con- sequences, but also some feeling for the relative ■•Mendelssohn, R. 1976. Harvesting with smoothing costs. SWFC Admin. Rep. 9H, 26 p. Southwest Fish. Cent. Honolulu Lab.. Natl. Mar. Fish, Serv., NOAA, Honolulu, HI 96812. trade off between total income and the smoothness of the received income stream. For the Wood and Branch Rivers, two sets of computations were performed. The first set as- sumes that y = e, i.e., there is an equal concern for increases in allowable harvest as well as for de- creases. This is equivalent to e = 0.0 and c = y (or equivalently e). The motivation for this cost struc- ture is that fishermen typically resist any decrease in the allowed harvest, hence y>0. However, al- lowing increases in the harvest size often signals fishermen to gear up and invest in equipment, thereby making it even more difficult to decrease the allowable harvest later on. Therefore this cost should be equal to a cost due to a decrease in the harvest. As a counterbalance to this, a second set of com- putations were performed with y>0 but e = 0, i.e., a cost only if the harvest is decreased. This is equivalent toe = e = y/2. For the first set of computations, with e = 0.0 andp = 1.0, values of c of 0.25, 0.50, 0.75, 1.00, 1.25, 1.50, 1.75, and 2.00 were used. These are equivalent to relative values of Vs, V4, %, ¥2, %, %, ■%, and 1. For the second set of runs, withe = e, and p = 1.0, values of 0.25, 0.50, 0.75, 1.00 and 1.25 were used. These are equivalent to a ratio of yip equal to Va, V2, %, 1, VA. The results are sum- marized in Figure 2(a)-(m) and Figure 3(a)-(m), which show an optimal policy for each river for each of these cases. All computations were per- formed on 26-point grids. The figures are read as follows. Suppose 2 was harvested last year and x is the observed popula- tion size this period. Find the point {x, z) on the graph and follow the arrow in that zone to the appropriate boundary as indicated. Then read off the 2 value of this point; this is the optimal amount to harvest during this period. For example, if e = 0.50, e = 0.00, x^ = 0.84, and the harvest last period was 0.28, Figure 2(b) shows that the optimal policy for the Wood River is to harvest 0.28 this period. Note that the dashed line is the equivalent base stock harvest with no smoothing costs. While the policies in Figures 2 and 3 are optimal for the given relative values of p, e, and c, they are complex in nature and would be difficult for a layperson to understand. Practical management often implies determining simpler, good but sub- optimal policies that achieve the same objectives. These policies are often more desirable since they 40 MENDELSSOHN: USING MARKOV DECISION MODELS are easier to implement and easier to explain the rationale to the public. As an example of suboptimal, approximate policies, the following nine modified base stock policies were examined: 3) Policy of base stock size of 0.56 till 1.40, then a base stock size of 0.84. 4) Harvest till 0.28, harvest 0.28 till 0.84, a base stock size of 0.56 till 2.52, then a base stock size of 0.84. Wood River Branch River 1) Base stock policy, base stock size = 0.84. 2) Policy of base stock size of 0.56 till 2.52, then a base stock size of 0.84. 5) Base stock policy, base stock size of 0.40. 6) Base stock size of 0.4 till 1.6, then a base stock size of 0.6. Figure 2(a-m). — Optimal policy functions for the Wood River for various assumptions about the relative value of smoothing costs. (See text for details). 41 FISHERY BULLETIN: VOL. 78, NO. 1 7) Base stock size of 0.2 till 0.6, then a base stock size of 0.4. 8) Base stock size of 0.2 till 1.0, then a base stock size of 0.4. 9) Base stock size of 0.2 till 0.4, base stock size of 0.4 till 1.2, base stock size of 0.6 after that. These nine approximate policies were devised by examining the functions that define the three regions in Figures 2 and 3. These approximate the boundaries of the three regions where the smooth- ing costs are one-fourth to one-half the per unit value of the harvest. The mean per period harvest, variance, standard deviation, median per period harvest, etc. for these nine policies are given in Table 2. Policies 3 and 4 for the Wood River and 8 and 9 for the Branch River demonstrate how these ap- proximate policies tend toward smoothing policies. For example, policy 4 has the same me- dian harvest as the optimal base stock harvest, almost never closes the fishery, significantly de- ,56 1.12 1.68 2.24 2.80 3.36 3.92 4.48 5.04 5.60 6.16 6.72 .56 1.12 1.68 2 24 2.80 3.36 3.92 4.48 5.04 5.60 6.16 6.72 Figure 2.— Continued. 42 MENDELSSOHN. USING MARKOV DECISION MODELS creases the percent of time there are low catches, and only reduces the mean per period harvest by 33,800 fish. In order to achieve a smoother catch, "potlatch" harvests from time to time have been sacrificed. When looked at closely, these policies are actu- ally very intuitive and represent an interesting variant of a base stock policy. These policies re- place a single base stock size by a dual base stock size policy. The first base stock size is lower than the original one, while the second base stock size is greater than or equal to the original base stock size. This means that there are fewer states where there is no harvesting, but also lowers the likeli- hood of the really big harvests. The mean per period harvest tends to be very sensitive to these big harvests, while the median is not, particularly since the very large harvests are not too frequent. It is curious that the population dynamics are so sensitive to such fine tuning, for the difference between policy 1 and policy 3, say, is quite mar- ginal. It would be an interesting area of future .56 112 1.68 224 2.80 336 3.92 4.48 5.04 5.60 6.16 6.72 .56 1.12 1.68 2.24 2.80 3.36 3.92 4.48 504 5.60 6.16 6.72 X X Figure 2.— Continued. 43 FISHERY BULLETIN: VOL. 78, NO. 1 672 - 1 1 1 1 I 1 ' 1 1 1 1 6.16 - (m) C»E« 1.25 7 / / 5.60 / 5.04 " / / 4.48 - ' / / 3.92 - / / o 3 36 - I / / ? 2.80 - / c n / - 2.24 - X 1.68 _ // / - 1.12 [— / ■' ^ / / ^' \ m 56 /Sr ^.Z. ■•'1 1 1 : 1 1 1 1 1 1 I .56 1.12 168 2.24 2.80 3 36 3.92 4 48 5.04 5,60 616 672 X Figure 2.— Continued. research to determine guidelines for when fine tuning would be expected to produce such "trim- ming" of the tails of the ergodic (long-run probabil- ity) distribution. Including smoothing costs also tells us a great deal about traditional concepts of fisheries man- agement, such as MSY. It is clear from Figures 2 and 3 that anything close to an MSY policy is optimal only if the smoothing costs exceed the per unit value of the harvest. As whole systems of laws for regulating fisheries have been constructed around the idea of smooth, constant harvests, it is clear that this imputes lower average catches, and a significant preference for constancy of the har- vest over total amount harvested. The analysis has assumed that Equation (1.1) or similar equations are available, and that the parameter estimates are accurate (in this case, estimates of R^, R^, and cr^). In the latter case, management measures would seem more reason- able if they were known to be robust against mis- specifying the parameters. This involves knowing how an optimal policy and total expected value would vary if the true underlying parameter val- ues differ from those specified, and also how the estimate of the long-run probability distribution differs from the true one. Walters and Hilbom (1976) have examined a similar question of trying to solve the Bayes model of this problem, i.e., where there is an original prior probability given to each value of the parameter, and this probability is updated each period using Bayes theorem and the observed val- ues during the period. However, they could not obtain a solution, and Walters and Hilborn ( 1978) raised questions as to the validity of some of their numerical approximations. Fortunately, qualitative results are possible for this particular class of Bayes problems. Let 6 be the parameter (or vector of parameters) under consideration. Let q^iQ) be the initial prior dis- tribution on 6, and let q^ ( B) be the updated prior distribution after n period has elapsed. Let Cl be the set of all possible prior distributions. Then it is proven in van Hee (1977a) that if the state of the system is expanded to {Xf,q^), the resulting optimi- zation problem is Markovian. Following argu- ments similar to those in Scarf (1959) and van Hee (1977a) it follows that an optimal Bayes policy takes the form: For each element q e Ci, there is an x{q) such that: do not harvest ifx^^xiq) harvest x^ - x{q) ifXf>x(q). For example, if cr^ in the distribution of c? is itself a random variable, then each possible probability distribution of cr^ yields a possibly unique base stock size policy. Table 2. — Vital statistics for the nine policies approximating the smoothing cost policies for Wood and Branch Rivers. Mean per Variance of % time % time period per period Standard % time less than greater than Median Relative value: River Policy harvest harvest deviation no catch 25% of mean mean catch catch smoothingprice Wood 1 1.1357 0.8468 09202 5.6 16.8 39 0.98 0/1 2 1.0993 05460 0.7389 1.7 10.7 39.8 0.98 1/8 3 1.1203 0.6506 08066 1.1 7.7 43.2 0.91 1/4 4 1.1019 0.5758 07588 002 10.47 40 0.98 1/2 Branch 5 0.6528 0.3982 0.6310 9.2 21.8 40 0.500 0/1 6 0.6290 02532 0.5032 9.1 21.5 37.2 0.500 1/4 7 0.6272 0.3077 0.5547 1.2 27.7 31.3 0.400 1/2 8 0.5920 0.2202 0.4693 1.9 35.7 26.3 0.500 3/8 9 0.5995 0.3038 0.5512 72 22.83 39.3 0.500 3/4 44 MENDELSSOHN: USING MARKOV DECISION MODELS Van Hee (1977a) defined a set of policies that he terms Bayes equivalent policies. For problems such as the salmon models under discussion, a Bayes equivalent policy would be found as follows: 1) At the start of the period, the prior probability distribution is q^( B). 2) The expected transition function (expectation with respect to O) is calculated, i.e., p(d,q) = jp(d\e)q(de) (4.1) where p( • | • ) describes the dependence of the ran- dom variable d on O. 3 )p(d,q) is used to solve a non-Bayesian Markov decision process, with p{d, q) as the transition function. 4)The optimal policy from step 3 above is used for one period. 5) q-l O) is updated using Bayes theorem and the observations from the last period, and the updated (?( • ) is used in step 1 at the next time period. It is worth noting that a Bayes eqivalent policy 8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 44 4.8 X Figure 3(a-m). — Optimal policy functions for the Branch River for various assumptions about the relative value of smoothing costs. (See text for details.) 45 FISHERY BULLETIN: VOL. 78, NO. 1 is adaptive, as the prior distribution is updated each period. Moreover, it is not the same as fixing 9 at its estimated value, and using a fixed value of 6 in step 3. The difference can be seen in the integral in Equation (4.1). The reason for consider- ing Bayes equivalent policies is that van Hee (1977a, theorem 3.1) proved that for the models under discussion, when the objective is given by Equation (1.2a) or (1.2b), then the Bayes equiva- lent policy is optimal for the full Bayes model. For example, in Walters and Hilborn (1976), the parameter 6 is a scalar, i.e., flg ^^ our notation. Their problem, for which an optimal policy was not found, can be solved by following a policy outlined in the five steps above. Many models will not have the necessary struc- ture for a Bayes equivalent policy to be optimal for the full Bayes model, and unlike salmon manage- ment, estimates of the population size may not be available every year. A legitimate question is: suppose the present best estimate of B were to be used from hereafter. What would be the loss in MENDELSSOHN: USING MARKOV DECISION MODELS expected value? Van Hee ( 1977b) gave bounds on this expected loss that are easy to compute. To obtain a feel for these bounds, both (r'^ and R^ are assumed to be random variables. For the Wood River, /?2 could take on the values -0.6, -0.8 and -1.0, and for the Branch River R^ could take on the values -1.5, -1.85, and -2.00. For the Wood River, 0-2 could assume the values of 0.35, 0.45, and 0.55, and for the Branch River a^ could assume the values 0.48, 0.58, and 0.68. Three probability dis- tributions were used as the present prior probabil- ity of the parameter values. These were (%, Vs, Va,), (1/4, V2, 1/4), (Vs, %, Vs). The results of the optimiza- tion using the parameters at each fixed value (which are needed to calculate the bounds) are given in Table 3. Table 4 gives the bounds on the expected loss of value from using the present esti- mates of the parameters as in Equation (1.1). Table 3 suggests that as cr^ varies for fixed val- ues of /?j , /?2 ' the mean per period harvest varies little, but the variance of the long-term harvest size distribution increases significantly. As i?2 4.8 - 1 I 1 1 1 1 1 1 1 1 I 44 - (i) C = E = 0.25 /: / 4.0 - / / / 3.6 - / / 3.2 - / / - 2.8 _ ' / _ / 24 - I / / - 2.0 - / . 1.6 - / / - 1.2 _ / / ffl / .8 - / / - 4 — / / / / 1 1 I i ' ! 1 4.8 - 1 11:11 1 ; 1 1 1 44 - (J) C = E = 0.50 /_ / / / / 4.0 / / - / / / / 3.6 - / / / / 3.2 - / / ' 2.8 - / / / / - ?4 ■A / / ?0 I / '/ ^ / '&/ 1.6 - . // - / / 1.2 - /-—^ m / .8 - A / / - .4 " ^' - n _ ^ Av i 1 1 1 1 1 1 1 1 1 "1 r 1 r (k) C = E = 0.75 J I L "I r "1 r "1 r (I) C=E = I.OO J I I L 4 .8 12 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 4 8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 40 44 4.8 X X Figure 3.— Continued. 47 FISHERY BULLETIN: VOL. 78, NO. 1 4 .8 1.2 1.6 2.0 2 4 2.8 3.2 3.6 4.0 4.1 4.8 X Figure 3.— Continued. varies for fixed values ofi?i , cr^, both the mean and the variance vary significantly. Table 4 reinforces this impression to a degree. If the mean per period harvest does not vary significantly with changes in the value of cr^, it might be expected that the present estimate of a^ will suffice. This is borne out by Table 4, where the bounds on the maximum expected total loss is <0.01, which is <1% of the optimal Bayes expected value. Some significant expected loss in value wheni?2 varies is seen, but the loss is less than might be expected from Table 3. The values in Table 4 when i?2 varies are all <4% of the true value. These results suggest that if Equation (1.1) is the correct form of the model, and the present parameter es- timates have relatively small variance, then little is gained in expected value if the more complicated policy is used. The same may not be true if the population size is unobserved. All of these results suggest a model that is fairly robust to our lack of understanding of nature. A possible explanation for this can be made from the discussion on the effect of grid size. As long as there is some cutoff population size below which no harvesting is allowed, and this cutoff assures that the absorbing state cannot be reached with proba- bility one, then our management can only damage the stocks to a degree. All of the policies examined in this paper have such a minimum cutoff. The rest of the policy will determine the relative mean and variance of the harvest, and techniques are presented to examine these features in detail. Uncertainty about the values of the parameters will affect the total re- turn, but present estimates often can give a satis- factory approximation. The truly risk adverse de- cision maker can use present estimates of the parameters that are weighted to be on the cautious side. SUMMARY Uncertainty in fisheries management can be faced head on. Techniques exist that allow us to gain much insight on managing randomly varying populations. Optimization procedures allow us to reduce our attention to the few best policies, and to analyze their properties, rather than to pick policies ad hoc that meet no special criteria. Optimization under uncertainty can also lead to a reconsideration of what is valued in managing a Table 3. — ^Trials with varied parameters. Rrver Value CJf fl2 Value of (7 Optimal policy Mean per penod harvest Variance % time no harvest Wood Wood Wood Wood Branch Branch Branch Branch 0.800 0.35 min (X(, 0.7) 1.0680 0390976 0.79 0.800 0.55 min (Xf, 0.77) 1.2267 1 .2422 7.8 0.600 0.458 mm (x,, 0.980) 1.5108 1.3136 3.8 1.000 0.458 mm (x,. 0.560) 0.9225 0.4839 3.29 1.845 048 mm (Xf. 0.35) 06122 02253 3.54 1.845 0.68 mm (Xf. 0.35) — — — 1.500 0.579 mm (X,, 0-40) 1.989 0.5254 582 2.000 0.579 mm (xf, 0.30) 0.9075 0.3068 5.82 Table 4. -Largest possible deviation in value of the approximate policy compared with the true Bayes policy. Probability distribution When Ri is uncertain When (T is uncertain 1/3, V3, Va y4,V2,'/4 V8,%,y8 Vj, V3, V3 Vo. V2. Va Ve, %, Va Wood River Branch River 1.4 1.04 0.51 0.5 0.47 0.38 0.04 0.01 0.03 «0.01 0.03 S0.01 48 MENDELSSOHN: USING MARKOV DECISION MODELS fishery — in the examples considered, some consis- tency in the amount harvested is a desirable al- ternative to high year-to-year fluctuations in the harvest size. But this reduced the average per period catch. Only in extreme situations, where the cost of smoothing out the catch is greater than the unit value of the catch, does any policy re- sembling MSY become optimal. Finally, it is possible to obtain an understand- ing of how robust the management measures are to misspecifications of the underlying model. This is important, since the model is only a guide to our decision making, not the answer. In the models considered, the "best" policies are robust in view of this uncertainty. A question not examined is the assumption that the population size is observed at the start of each period. This too is usually costly, and inexact. Re- cently, I and E. J. Sondik developed an efficient algorithm that addresses the relative merits of different sampling intervals for obtaining popula- tion estimates.^ Together, all of these techniques allow for an integrated, realistic approach to man- agement under uncertainty. ACKNOWLEDGMENTS Debra Chow provided invaluable assistance in programming the computer runs. David Stoutemeyer of the University of Hawaii gave much useful advice on maximizing the efficiency of the optimization algorithm used. Lee Anderson, George Fishman, and Adi Ben-Israel gave impor- tant comments for improving an earlier version of this paper. The paper also benefited greatly from the comments of one of the referees and from C. Walters who tempered some remarks in a previous version. LITERATURE CITED ANDERSON, L. G. 1977. The economics of fisheries management. The Johns Hopkins Univ. Press, Baltimore, 214 p. Bertsekas, D. p. 1976. Dynamic programming and stochastic con- trol Acad. Press, NY., 396 p. Clark, C. W. 1976. Mathematical bioeconomics: The optimal manage- ^Mendelssohn, R., and E. J. Sondik. 1979. The cost of in- formation seeking in the optimal management of random re- newable resources. SWFC Admin. Rep. H-79-12, 15 p. Southwest Fish. Cent. Honolulu Lab., Natl. Mar. Fish. Serv., NOAA, Honolulu, HI 96812. ment of renewable resources. John Wiley and Sons, N.Y.,352p. FOX, B. L. 1973. Discretizing dynamic programs. J. Optim. Theory Appl. 11:228-234. Fox, W. W.,JR. 1970. An exponential surplus-yield model for optimizing exploited fish populations. Trans. Am. Fish. Soc. 99:80- 88. 1971. Random variability and parameter estimation for the generalized production model. Fish. Bull., U.S. 69:569-580. 1975. Fitting the generalized stock production model by least-squares and equilibrium approximation. Fish. Bull., U.S. 73:23-37. Hastings, N. a. J., and J. A. E. E. van Nunen. 1977. The action elimination algorithm for Markov deci- sion processes. In H. C. Tijms and J. Wessels (editors), Markov decision theory, p. 161-170. Proceedings of the Advanced Seminar on Markov Decision Theory held at Amsterdam, the Netherlands, September 13-17, 1976. Mathematical Centre Tract 93. Mathematisch Centrum, Amsterdam. HINDERER, K. 1978. On approximate solution of finite-stage dynamic program. Universitat Karlsruhe, Fukultat fiir Mathematik, Bericht 8, 42 p. Mathews, S. B. 1967. The economic consequences of forecasting sockeye salmon (Oncorhynchus nerka, Walbaum), runs to Bristol Bay, Alaska: A computer simulation study of the potential benefits to a salmon canning industry from accurate fore- casts of the nms. Ph.D. Thesis, Univ. Washington, Seatr tie, 225 p. MENDELSSOHN, R., AND M. J. SOBEL. In press. Capital accumulation and the optimization of renewable resource models. J. Econ. Theory. PELLA, J. J., AND P. K. TOMLINSON. 1969. A generalized stock production model. [In Engl, and Span.] Inter-Am. Trop. Tuna Comm., Bull. 13:419- 496. PORTEUS, E. L. 1971 . Some bounds for discounted sequential decision pro- cesses. Manage. Sci. 18:7-11. REED, W.J. 1974. A stochastic model for the economic management of a renewable animal resource. Math. Biosci. 22:313-337. KICKER, W. E. 1958. Maximum sustained yields from fluctuating envi- ronments and mixed stocks. J. Fish. Res. Board Can. 15:991-1006. Scarf, H. 1959. Bayes solutions of the statistical inventory prob- lem. Ann. Math. Stat. 30:490-508. SCHAEFER, M. B. 1954. Some aspects of the dynamics of populations impor- tant to the management of the commercial marine fisheries. Inter-Am. Trop. Tuna Comm., Bull. 1:25-56. VAN HEE, K. M. 1977a. Adaptive control of specially structured Markov chains. In M. Schal (editor), Dynamische optimierung, p. 99-116. Bonner Mathematische Schriften 98, Bonn. 1977b. Approximations in Bayesian controlled Markov chains. In H. C. Tijms and J. Wessels (editors), Markov 49 decision theory, p. 171-182. Proceedings of the Advanced Seminar on Markov Decision Theory held at Amsterdam, the Netherlands, September 13-17, 1976. Mathematical Centre Tract 93. Mathematisch Centrum, Amsterdam. Waldmann, K.-H. 1978. On approximation of dynamic programs. Preprint No. 439, Fachberech Mathematik, Technische Hochschule Darmstadt, 17 p. Walters, C.J. 1975. Optimal harvest strategies for salmon in relation to FISHERY BULLETIN: VOL. 78, NO. 1 environmental variability and uncertain production parameters. J. Fish. Res. Board Can. 32:1777-1784. Walters, C. J., and r. Hilborn. 1976. Adaptive control of fishing systems. J. Fish. Res. Board. Can. 33:145-159. 1978. Ecological optimization and adaptive manage- ment. Annu. Rev. Ecol. Syst. 9:157-188. WHITT, W. 1978. Approximations of dynamic programs, I. Math. Oper. Res. 3:231-243. 50 ORGANOCHLORINE RESIDUES IN FISHES FROM THE NORTHWEST ATLANTIC OCEAN AND GULF OF MEXICO Virginia F. Stouts ABSTRACT Residues of SDDT (p,p -DDT and its metabolites p,p -TDE and p,p -DDE), PCB (polychlorinated biphenyls), dieldrin, and endrin were determined in the flesh of 700 specimens of fishes caught between 1973 and 1975 in the northwestern Atlantic Ocean and northern Gulf of Mexico off the coasts of the southeastern United States. Species with lowest oil content (<3%) — gag, Mycteroperca microlepis; black grouper, M. bonaci; red grouper, Epinephelus morio; and red snapper, Lutjanus campechanus — contained the least amounts of chlorinated hydrocarbon residues. Species with higher oil content — king mackerel, Scomberomorus cavalla, (3.5%) and Spanish mackerel, S. maculatus, (4.6%) — more consistently contained residues, but still at low levels. Significant correlations (P<0.05) were found in red snapper and king mackerel between lipid and size and between lipid and chlorinated hydrocarbon content. The correlations between lipid and PCB in gag and between lipid and DDT in black grouper were also significant. The highest mean values for any species were 0.18 ppm SDDT, 0.32 ppm PCB, 0.007 ppm dieldrin, and 0.008 ppm endrin. The highest level in any one composite sample of 10 fish was 1.0 ppm IDDT, 1.8 ppm PCB, 0.026 ppm dieldrin, and 0.026 ppm endrin. Haifa century ago manufacturers of surface coat- ings and of electrical equipment found a common interest in a newly introduced group of or- ganochlorine chemicals, the PCB.^ These com- pounds dissolve the inks in carbonless carbon paper, which duplicates without the use of carbon paper. They plasticize the waterproof coatings for dairy silos and fish tanks, and marine antifouling paint. The thermal and electrical properties of PCB are highly desirable in dielectric fluids, the electrical insulators in transformers and capacitors. The PCB are also highly resistant to chemical and biological degradation, and these properties, too, are valued by industrial users. Immediately following World War II, another organ ochlorine chemical, DDT, became the magic tool of the medical profession, deeply concerned with preventing outbreaks of infectious, insect- borne diseases. Enormous quantities of DDT were used to prevent epidemics of typhus and plague in war-torn Europe. DDT was dramatically effective in controlling lice and fleas, which carried these 'Northwest and Alaska Fisheries Center Utilization Research Division, National Marine Fisheries Service, NOAA, 2725 Montlake Boulevard East, Seattle, WA 98112. ^Abbreviations used in this manuscript: DDE = p,p'-di- chlorodiphenyldichloroethylene; DDMU = p,p'-dichlorodi- phenylchloroethylene; DDT = p,p'-dichlorodiphenyl- trichloroethane; PCB = polychlorinated biphenyls; I DDT = DDT and its metabolites DDE and TDE; TDE = p,p'- dichlorodiphenyldichloroethane. Manuscript accepted August 1979. FISHERY BULLETIN: VOL. 78, NO. 1, 1980. diseases. The use of this and related compounds spread rapidly for mosquito and agricultural pest control. Before insect resistance developed, DDT reduced the incidence of malaria to a very low level. Since PCB and the organochlorine pesticides are not only stable but also easily dispersed in the air and through the water, it is not surprising in retrospect that they are now found even in the polar regions (Sladen et al. 1966; Risebrough et al. 1976; Bowes and Jonkel 1975) and that chlori- nated hydrocarbon pollution has become a worldwide problem. Not until the 1960's did ap- preciation of the adverse effects of these chlori- nated hydrocarbons begin to develop. "Silent Spring" (Carson 1962) described the effects on the environment of the accumulation of DDT, and Jensen (Anonymous 1966; Jensen et al. 1969) found PCB in marine animals. Burdick et al. (1964) noted reproductive failure in lake trout when the eggs contained a high level of SDDT. Aulerich et al. (1973) traced the reproductive fail- ure and mortality in ranch-grown mink back to coho salmon, Oncorhynchus kisutch, from Lake Michigan and ultimately to the PCB they con- tained. Only recently fin erosion, a disease associated with municipal and industrial dis- charges, was related to iDDT (Mearns and Sher- wood 1977). Montrose Chemical Company re- leased wastes from DDT manufacture directly into 51 the Los Angeles County sewer system which flows via sewer outfalls into the marine environment of southern California. From this unanticipated source of iDDT pollution, high levels of IDDT developed throughout the southern California marine environment from mollusks (Young et al. 1976), crustaceans (Burnett 1971), and fishes (McDermott-Ehrlich et al. 1978) to marine birds (Anderson et al. 1975) and cetaceans (Le Boeuf and Bonnell 1971). Henderson et al. (1969, 1971) found organochlorine residues in freshwater fishes throughout the United States. Some species of freshwater fishes, especially those from the Great Lakes, contained levels of chlorinated hy- drocarbons that exceeded the U.S. Food and Drug Administration guidelines (Reinert 1970; Veith 1975). In recent years, the use of chlorinated hy- drocarbons has been drastically curtailed, but concern remains about the continuing occurrence of these toxic compounds in the marine environ- ment. Fishes and shellfishes are excellent organisms for monitoring chlorinated hydrocarbons. Shellfishes have been used as indicators of short- term pollution (Butler 1973; Goldberg et al. 1978) because they accumulate and depurate these sub- stances readily. Fishes, on the other hand, reflect long-term exposure since they lose chlorinated hydrocarbons slowly, if at all (Lieb et al. 1974). Butler and Schutzmann (1978) have reported on pesticides and PCB in yearling estuarine fishes of the United States. Outside of their study, how- ever, few data on fishes for human consumption from the western Atlantic Ocean have been pub- lished except on fishes from Canadian waters (Sims et al. 1977). The study reported here was undertaken to obtain information about levels of SDDT, PCB, dieldrin, and endrin in fillets from fishes caught in the northwestern Atlantic Ocean and northern Gulf of Mexico. Six marine species of both commercial and sport value have been examined. FISHERY BULLETIN: VOL. 78. NO. 1 METHODS Samples of gag, Mycteroperca microlepis; black grouper, M. bonaci; red grouper, Epinephelus morio; red snapper, Lutjanus campechanus; king mackerel, S comber omor us cavalla; and Spanish mackerel, S. maculatus, were collected from the northwestern Atlantic Ocean and northern Gulf of Mexico, from Beaufort, N.C., south to the Florida Keys, and west to Aransas Pass, Tex. Sampling occurred between October 1973 and October 1975, but mainly in 1975. Specimens were frozen and held at -18° C. They were thawed for filleting, grinding, compositing, and refrozen until analysis. In the aggregate, 70 samples each con- taining 10 fish of the same species and of similar size were obtained. Each sample was a composite of equal weights of ground skinless fillets from the 10 individual fish. At most sites, two or three sam- ples from different size fish were selected. The collection sites, size, and lipid content of the specimens are listed in Table 1. Extracts were prepared by the procedures of Reinert (1970). Samples for SDDT and PCB analysis were extracted with propanol-2/benzene (1:1), and the extracted materials transferred to hexane by repeated codistillation of the propanol-2, benzene, and water with hexane. One aliquot of the hexane extract was evaporated to minimum weight for determination of the lipid content. Another aliquot was cleaned up on Florisil.^ PCB were separated from TDE, DDT, and most of the DDE on activated silica gel (Snyder and Reinert 1971), which also separates the interfering hydrocarbons, phenanthrene, fluoranthrene, and pyrene from the PCB (Zitko 1978). At least 90% of the PCB was contained in the pentane fraction, which also contained part of ^Mention of specific products or companies in this paper does not imply endorsement by the National Marine Fisheries Ser- vice, NOAA. Table l. — Collection sites, size, and lipid content of fishes from the northwest Atlantic Ocean and Gulf of Mexico. Species Gag Black grouper Red grouper Red snapper King mackerel Spanish mackerel No.2 Fork length (cm) Mean Range^ Weight (kg) Lipid ci Mean ontent (%) Sites' Mean Range^ Range^ 1,2,3 8 86 64-108 9.77 3.79-18.32 2.9 0.2-5.9 5,6 6 77 54- 96 7.01 1.98-12.70 0.6 0.2-1.2 1,4, 5, 6 10 68 52- 82 608 2.42-11.50 0.5 0.1-1.0 1,2,3,4.5, 6. 7,8 18 65 45- 83 5.94 1.65-11 35 1.5 0.4-3.9 1,2,3,4,5, 6, 9 18 87 55-119 6.01 1.20-12.91 3.5 0.4-7.5 3,4, 5,6, 7 10 54 45- 64 1.41 0.48- 2.34 4.6 1.7-9.4 ' 1 - Beaufort, N.C.; 2 Savannah, Ga.; 3 - Florida, East Coast; 4 - Florida Keys; 5 - Tampa/St. Petersburg, Fla.; 6 - Panama City, Fla.; 7 - Mobile, Ala.; 8 - Pascagoula, Miss.; 9 - Aransas Pass, Tex. ^Number of composites, each consisting of 10 fish. ^Range of means of individual composites. 52 STOUT: ORGANOCHLORINE RESIDUES IN FISHES the DDE. All of the TDE and DDT eluted into the benzene fraction. For dieldrin and endrin analysis, tissues were saponified with KOH in aqueous ethanol, extracted with hexane, and cleaned up on Florisil. Since the specimens, origi- nally collected for trace-metal analysis, were stored in polypropylene containers (Falcon No. 4014) with polyethylene lids (Falcon No. 4017), the containers and lids were also analyzed to as- sure freedom from interfering substances. The de- tails of our procedures were published previously (Stout and Beezhold 1979). Blanks carried through the whole procedure for either PCB and SDDT or dieldrin and endrin showed no chromatographic peaks which interfered with quantitation of the chlorinated hydrocarbons. The extracts of the fishes were quantitated by electron-capture gas chromatography. A Varian 600 C gas chromatograph with a titanium tritide detector was fitted with a 1.5 m (5 ft) x 0.32 cm (0.125 in) o.d. glass column containing a mixture of equal parts of 15% QF-1 on 80-100 mesh Gas Chrom Q and 10% DC-200 on the same support (Burke and Holswade 1966). Reference standards of p,p'-DDE, p,p'-DDMU, p,p'-TDE, p,p'-DDT, dieldrin, and endrin were obtained from the U.S. Environmental Protection Agency, Health Effects Research Laboratory, Research Triangle Park, N.C. Aroclor 1254 obtained from the Monsanto Company, St. Louis, Mo., was the standard for the PCB because the residues in the fishes matched this Aroclor most closely. Standard curves of peak height versus concentration were used to deter- mine the concentrations of components in the ex- tracts. The sensitivity throughout each run was assured by frequent injections of standard solu- tions. The quantifiable limit was about 0.003 ppm for DDE, TDE, DDT, dieldrin, and endrin, and about 0.04 ppm for PCB depending on daily sen- sitivity. The mean relative standard deviation for samples analyzed in duplicate was 11%. The aver- age recovery of samples spiked with standards was 85%. The values reported were not corrected for recovery. Residue values were calculated on the basis of micrograms chlorinated hydrocarbon per gram wet tissue or parts per million (ppm). The PCB were quantitated by summing the peak heights corresponding to the five major peaks in Aroclor 1254 after omitting the twin peak with a retention time similar to that of p,p'-DDE. As Veith ( 1975) also concluded, use of five peaks in- creased the accuracy of PCB measurement by reducing the effect of minor variations in concen- tration of individual components in the samples (Figure 1). Since part of the DDE eluted from silica gel with the PCB fraction and one of the major peaks in Aroclor 1254 overlapped the DDE peak in the gas-chromatographic traces, a special technique was needed to quantitate the DDE in the PCB fraction. First the gross DDE concentration was determined in the usual way from the peak height versus DDE concentration curve. Next, the con- tribution of PCB to that overlapping peak was calculated based on the assumption that the height of Aroclor peak 3, the peak which over- lapped DDE, was proportional to the heights of the five PCB peaks used to calculate the concentration of PCB. To make this calculation, the sum of the peak heights of the five major PCB peaks exclud- ing the "DDE" peak was plotted against the peak height of the "DDE" peak in PCB standards of increasing concentrations. From the sum of the five PCB peaks in the sample, the peak height of the PCB portion of peak 3 in that sample was determined. This peak height was converted to concentration of DDE via the peak height versus concentration curve for DDE. The apparent con- centration of DDE from PCB peak 3 was sub- tracted from the gross DDE concentration in the "DDE" peak to obtain the net concentration of DDE in the PCB fraction. The elctron-capture de- tector is so much more sensitive to DDE than PCB that this method of calculation affected the accu- racy of DDE determination only to a small extent (Figure 1). Use of a minicomputer expedited these calculations. Confirmation of DDT and its metabolites and PCB was effected by saponifying separate portions of tissue (Reinert 1970). DDT is dehydrochlori- nated by base to DDE, and TDE similarly to DDMU. PCB are stable to base. The levels of diel- drin and endrin were too low to warrant confirma- tion studies. RESULTS AND DISCUSSION The marine fishes from the northwestern Atlan- tic Ocean and northern Gulf of Mexico analyzed in this study contained relatively low levels of SDDT and PCB. Of the 70 composite samples, only 29 contained more than 0.05 ppm SDDT and 0. 1 ppm PCB in the edible portion (skinless fillets), and only one as much as 1 ppm SDDT and PCB. Red grouper contained the lowest levels of both chlori- nated hydrocarbons. The mean IDDT content for 53 FISHERY BULLETIN: VOL. 78, NO. 1 AROCLOR 1254 (1100 pg) TDE (40 pg) DDT (40 pg) Figure l. — Gas chromatographic curves of Aroclor 1254, the PCB fraction of an extract from Spanish mackerel, and DDT and its metabolites DDE and TDE; pg = picograms or 10 54 10 sets of red grouper was 0.008 ppm; for most samples, PCB were not detectable. Black grouper and gag contained slightly higher levels of iDDT and PCB. Red snapper also contained little iDDT (mean 0.039 ppm), but the PCB level in six sam- ples exceeded 0.1 ppm. Only the two species of mackerel consistently contained quantifiable amounts of both SDDT and PCB. The mean levels of SDDT were 0.144 ppm in Spanish mackerel and 0.177 ppm in king mackerel. The mean PCB level in both species of mackerel was 0.32 ppm. The highest levels of both chlorinated hydrocarbons were found in one composite sample of king mack- erel from the Florida Keys, 0.996 ppm iDDT and 1.8 ppm PCB. The data are summarized in Table 2. The limited data in the literature convey a simi- lar picture. Groupers of the genera Epinephelus and Mycteroperca from the Gulf of Mexico and the Grand Bahamas contained 0-0.10 ppm SDDT and 0.003-0.22 ppm PCB in muscle (Giam et al. 1974). Red snapper from Mobile Bay, Ala., contained 0.086 ppm SDDT and 0.14 ppm PCB in the whole animal; gray snapper, Lutjanus griseus, from Jacksonville, Fla., 0.007 ppm SDDT and no PCB in the whole animal (Markin et al. 1974). Small Spanish mackerel (306 g) from the Savannah River estuary in Georgia contained neither SDDT nor PCB in muscle. (Butler^) Although somewhat larger than those fish, the smallest fish in the current study (475 g) contained barely detectable amounts of these substances (0.008 ppm SDDT and 0.034 ppm PCB). Markin et al. (1974) found 0.04-0.16 ppm SDDT and <0.01-0.18 ppm PCB in seven whole Spanish mackerel from the south- eastern United States. A single sample of king mackerel muscle from the Gulf of Mexico off Mexico contained 0.024 ppm SDDT and 0.034 ppm PCB (Giam et al. 1972). Atlantic mackerel, Scomber scombrus, collected in 1971-72 in Cana- dian waters (Sims et al. 1977) contained more SDDT (0.26 ppm) and PCB (0.41 ppm) in the "edi- ble portion" (which may have contained skin and/or bones) than did skinless fillets of either species of mackerel examined in my study. On the other hand, muscle from Atlantic mackerel from the Bay of Fundy-Gulf of Maine contained the same level of PCB (0.35 ppm) (Zitko et al. 1972). The proportions of p,p '-DDT and its metabolites were very similar in the king and Spanish mack- erel examined in our study. p,p'-DDE is the -12 g- "Butler, P. A. 1978. EPA-NOAA Cooperative Estuarine Monitoring Program, Final Report, Gulf Breeze, Fla., 108 p. STOUT: ORGANOCHLORINE RESIDUES IN FISHES Table 2.— iDDT and PCB in fishes from the northwest Atlantic Ocean and Gulf of Mexico, nq = not queintifiable; nd = none detected. No.' iDDT (ppm) PCB (ppm) Species Mean + SD Range Mean = SD Range Gag 8 0.036^0.013 0.017-0050 0087 = 0.023 nq-0 129 Black grouper 6 0.009r0.007 0.003-0.020 nq nq-0.059 Red grouper 10 0.008 = 007 n.d,-0,025 nd nd-nq Red snapper 18 0.039 = 0076 n.d.-0.322 0.121=0.108 nd-0464 King mackerel 18 0.177 = 0.239 0.009-0.996 0322 = 0.414 0,060-1.78 Spanish mackerel 10 0.144 = 0.097 0.008-0.319 0319 = 0.263 0.034-0.821 'Number of composites, each consisting of 10 fish. major component (—65%) accompanied by about 25% of the parent compound and 10% p,p'-TDE. Although the composite picture for red snapper looked markedly different (Table 3), in fact the DDT residues in that species actually fell into two categories. In the first group, p,p -TDE and p,p '- DDT were not quantifiable, and the maximum content of p,p -DDE was 0.029 ppm. In the second group, five samples containing >0.029 ppm p,p'- DDE, both p,p'-TDE and p,p -DDT were quanti- fiable. In those five samples, the proportions of the three components were the same as in the mack- erel. Finfish from the Atlantic coast of Canada (Sims et al. 1977) contained proportionately much lessp,p -DDE (45% ) and morep,p -DDT (40% ) and p,p '-TDE (15%). The increase in the proportion of p,p -DDE in the present samples may reflect deg- radation of the parent compound in the environ- ment before it accumulated in the fishes. Samples for the Canadian study were collected in 1971 and 1972, soon after usage of DDT had been drastically curtailed (around 1970) as the result of increasing insect resistance, problems with indirect con- tamination of foodstuffs, and concern about effects of DDT on nontarget species. Several years elapsed before the samples for the present study were collected, mainly in 1975. In the interval, DDT was degrading aerobically to DDE and anaerobically in the marine environment to TDE. Alternatively, DDT may metabolize more rapidly to DDE in the more temperate climate of the re- gion studied and somewhat less rapidly to TDE. Table 3. — Mean proportions of 5^DDT present as p,p '-DDE, p,p '- TDE, and p,p -DDT in fishes from the northwest Atlantic Ocean and Gulf of Mexico. p.p DDE (%) Mean = SO Range p.p TDE (%) p.p -DDT (%) Species IV1ean=SD Range Mean = SD Range Red snapper, all 88=18 58-100 3=5 0-15 9 = 13 0-30 Red snapper' 62= 6 58- 72 10 = 3 7-15 28-4 21-32 King mackerel 64= 8 48- 77 8 = 6 0-17 28= 6 16-40 Spanish mackerel 66 = 19 48-100 12 = 8 0-24 21=12 0-37 'Five samples which contained >0.029 ppm DDE. The concentration of PCB, when present, was higher than that of IDDT in all samples but one. The mean ratio of PCB to SDDT was 1.8 for gag, 2.2 for king and Spanish mackerel, and 2.6 for six sets of red snapper. In the two other samples of red snapper with quantifiable levels of PCB, the PCB/SDDT ratios were 22.7 and 24.3. The con- centrations of SDDT were below 0.01 ppm in both cases. The one sample of red snapper in which the PCB/IDDT ratio was below 1 contained a rela- tively large amount of SDDT, 0.096 ppm, the sec- ond highest IDDT value in the 18 samples of red snapper. In contrast, the PCB concentration was low both in absolute amount, <0.06 ppm, and in rank, 14th out of 18 samples. The chlorinated-hydrocarbon content of the specimens in this study was, in general terms, directly related to the lipid content. The groupers, which contained <1% lipid, contained the least iDDT and PCB. King and Spanish mackerel had the highest lipid contents, 3.5 and 4.6%, respec- tively, and the highest levels of both IDDT and PCB. In three of the six species, the correlation between lipid and IDDT was significant, i.e., P<0.05. Similarly, in three of the four species for which PCB were quantifiable the correlation be- tween lipid and PCB was significant. Possible rela- tionships between size and chlorinated hydrocar- bon content were also examined. Although length, weight, iDDT, and PCB were all studied, in no case was a significant correlation found in any of the six species (Table 4). In two of the six species, the correlations between length and lipid and also between weight and lipid were significant. In both species, red snapper and king mackerel, the P values for lipid versus chlorinated hydrocarbon, were below 0.01. Giam et al. (1974) noted a rela- tionship between size and concentration of pollu- tants in groupers from the Gulf of Mexico, but only in the area with the highest contamination, i.e. , up to 0.1 ppm. SDDT and PCB levels within single species were compared at the various sites. Fish from the 55 FISHERY BULLETIN: VOL. 78. NO. 1 o c HI ■^ -r- CM CO CM CD CD CO eg >- >- in •- r^ o o CD ■.- o ■» o O CD 6 in (71 in CD eg in in CX5 eg CD r-~ -^ odd o o O C3) CNJ CJ> CD ■- 00 00 CO CD t^ CO in ^ "- o 1- <>i d d d o o d oj CO '- 00 CO r»- CO in T- CO CO T- CM CO in ^ CO ^ d o d d d d CO o eg >- CD CO in o 00 c:) -d- r^ CD -^ o •- ■- >- o d o o d d o^ CD CD CO r^ CO 00 eg t^ eg in CD t- Tt in CO CO -^ o o o o o d «!■ r-- o CO O) r- ^ '- r- o o ■» ■"3- C3) o o o r^ o o o o o o r^ CD in CO 00 h- .- in CJ) CD (3) 1- co o in CD in >- d d d d d d CO r^ o CO CO CO CD -^ >-•--- eg CO CO »- o o CJl d d d d d d C71 eg CO »- in in ID o CO r^ r^ CO CO »- in in in o o o o o o o CO CO O 00 CO o -So 2 (D ra d (J ^ (0 03 .c E « CO .ii' cr "" (0 oi CJ." So en I- o^ _^ C31 {/) oi o -o -a Pro (0 iO 0) Qj c Q_ O m DC Ct ^ t/5 (J C= ro ro uiQ trt <1> O EUJcD ZQQ- Florida Keys contained the greatest amounts of both substances, indicating that contamination from agricultural and domestic effluents was greatest in south Florida, an area of intense ag- riculture and relatively dense population as well. Fish caught in the vicinity of Beaufort and Mobile contained slightly less !iDDT and PCB. Fish from the other sites contained lower levels of chlori- nated hydrocarbons, which were not distinguish- able by site, except that only DDE was quanti- fiable in the one sample from Pascagoula, Miss. The Mississippi River, which receives the runoff from 40% of the land mass of the conterminous United States, including the corn belt and the cotton belt, did not seem to be the main source of either SDDT or PCB. Low levels of dieldrin and endrin, two highly toxic organochlorine pesticides, were found in the fish included in this study. The highest concentra- tion of either compound was 0.026 ppm. The diel- drin and endrin content of only the three species with the highest levels of SDDT and PCB were determined. Nonetheless, dieldrin and endrin were quantifiable only in about one-third of the samples. Two of 18 red snapper, 7 of 18 king mack- erel, and 6 of 10 Spanish mackerel samples con- tained quantifiable amounts of both compounds. The mean levels for Spanish mackerel, the species with the highest mean concentrations, were 0.007 ppm dieldrin and 0.008 ppm endrin (Table 5). King mackerel from Aransas Pass contained the high- est concentration of dieldrin, 0.026 ppm; Spanish mackerel from Panama City, Fla., the highest concentration of endrin, also 0.026 ppm. The diel- drin and endrin concentrations in the three species followed the same distribution patterns with relation to species and lipid content as was observed for SDDT and PCB. Butler (see footnote 4) found no dieldrin in the muscle of small Spanish mackerel from the Savannah River estuary. CONCLUSIONS Residues of SDDT, PCB, dieldrin, and endrin, although generally low, were found in all species and all locations except Pascagoula, Miss., where only DDE was quantifiable. The highest levels of SDDT and PCB, and the only ones to reach 1 ppm, were in one composite sample of king mackerel from the Florida Keys, 0.996 ppm SDDT and 1.8 ppm PCB. The highest level of dieldrin, 0.026 ppm, was in king mackerel from Aransas Pass, Tex., and of endrin, also 0.026 ppm, in Spanish mack- 56 STOUT: ORGANOCHLORINE RESIDUES IN FISHES Table 5. — Dieldrin and endrin in fishes from the northwest Atlantic Ocean and Gulf of Mexico, nd = not detected; nq == not quantifiable. No.' Dieldrin i Ippm) Endnn (ppm) Species Mean=:SD Range Mean*SD Range Red snapper King mackerel Spanish mackerel 18 18 10 nd 0.005 i:0.006 0.007^0.004 nd-nq nd-0.026 nd-0.014 nd 0.004*0.004 0.008 ±0.010 nd-0.003 nd-0.014 nd-0.026 'Number of composites, each consisting of 1 fish. erel from Panama City, Fla. The species with the highest Hpid contents contained the highest con- centrations of chlorinated hydrocarbons. Sig- nificant correlations (P<0.05) were found between lipid and size and between lipid and chlorinated hydrocarbon content in two of the six species. In two other species the correlation between lipid and either SDDT or PCB was significant. In all cases but one, the chlorinated hydrocarbon levels were substantially below the U.S. Food and Drug Ad- ministration guidelines: 5 ppm SDDT, 2 ppm PCB, and 0.3 ppm dieldrin and endrin. ACKNOWLEDGMENTS I thank the Southeast Fisheries Center Charles- ton Laboratory, NMFS, NOAA, for providing the samples and Laura G. Lewis for technical support throughout this research. Russell F. Kappenman, Northwest and Alaska Fisheries Center Fisheries Data & Management Systems Division, provided guidance for statistical analyses. LITERATURE CITED anderson, d. w., j. r. jehl, jr., r. w. risebrough, l. a. Woods, Jr., L. R. Deweese, and W. G. Edgecomb. 1975. Brown pelicans: Improved reproduction off the southern California coast. Science (Wash., D.C..) 190:806-808. ANONYMOUS. 1966. Report of a new chemical hazard. New Sci. 32:612. AULERICH, R. J., R. K. RINGER, AND S. IWAMOTO. 1973. Reproductive failure and mortality in mink fed on Great Lakes fish. J. Reprod. Fertil., Suppl. 19:365-376. BOWES, G. W., AND C. J. JONKEL. 1975. Presence and distribution of polychlorinated biphenyls (PCB) in arctic and subarctic marine food chains. J. Fish. Res. Board Can. 32:2111-2123. BuRDiCK, G. E., E. J. Harris, H. J. Dean, T. M. Walker, J. Skea, AND D. Colby. 1964. The accumulation of DDT in lake trout and the effect on reproduction. Trans. Am. Fish. Sec. 93:127-136. Burke, j. a., and W. holswade. 1966. A gas chromatographic column for pesticide residue analysis: Retention times and response data. J. Assoc. Off. Anal. Chem. 49:374-388. Burnett, R. 1971. DDT residues: Distribution of concentrations in Emerita analoga (Stimpson) along coastal California. Science (Wash., D. C.) 174:606-608. Butler, p. a. 1973. Organochlorine residues in estuarine mollusks, 1965-1972 — National pesticide monitoring pro- gram. Pestic. Monit. J. 6:238-362. Butler, P. A., and R. L. Schutzmann. 1978. Residues of pesticides and PCBs in estuarine fish, 1972-1976 — National pesticide monitoring program. Pestic. Monit. J. 12:51-59. Carson, R. 1962. Silent spring. Houghton-Miflin Co., Boston, Mass., 368 p. GiAM, C. S., A. R. Hanks, R. L. Richardson, W. m. SACKErr, andM. K.Wong. 1972. DDT, DDE, and polychlorinated biphenyls in biota from the Gulf of Mexico and Caribbean Sea — 1971. Pestic. Monit. J. 6:139-143. GIAM, C. S., R. L. RICHARDSON, D. TAYLOR, AND M. K. WONG. 1974. DDT, DDE and PCBs in the tissues of reef dwelling groupers (Serranidae) in the Gulf of Mexico and the Grand Bahamas. Bull. Environ. Contam. Toxicol. 11:189-192. Goldberg, E. D., V. T. Bowen, J. W. F.arrington, G. Har- vey, J. H. Martin, P. L. Parker, R. W. risebrough, W. Robertson, E. Schneider, and E. Gamble. 1978. The mussel watch. Environ. Conserv. 5:101-125. Henderson, C, A. Inglis, and w. l. Johnson. 1971. Organochlorine insecticide residues in fish — Fall 1969 National pesticide monitoring program. Pestic. Monit. J. 5:1-11. HENDERSON, C, W. L. JOHNSON, AND A. INGLIS. 1969. Organochlorine insecticide residues in fish (Na- tional pesticide monitoring program). Pestic. Monit. J. 3:145-171. JENSEN, S., A. G. JOHNELS, M. OLSSON, AND G. OTTERLIND. 1969. DDT and PCB in marine animals from Swedish wa- ters. Nature (Lond.) 224:247-250. LeBOEUF, B. J., AND M. L. BONNELL. 1971. DDT in California sea lions. Nature (Lond.) 234:108-110. LlEB, A. J., D. D. BILLS, AND R. O. SINNHUBER. 1974. Accumulation of dietary polychlorinated biphenyls (Aroclor 1254) by rainbow trout (Salmo gairdneri). J. Agric. Food Chem. 22:638-642. MARKIN, G. p., J. C. HAWTHORNE, H. L. COLLINS, AND J. H. Ford. 1974. Levels of mirex and some other organochlorine resi- dues in seafood from Atlantic and Gulf coastal states. Pestic. Monit. J. 7:139-143. McDermott-Ehrlich, D., D. R. Young, and T. C. Heesen. 1978 . DDT and PCB in flatfish around southern California municipal outfalls. Chemosphere 6, p. 453-461. Mearns, a. J., and M. J. Sherwood. 1977. Distribution of neoplasms and other diseases in marine fishes relative to the discharge of waste wa- ter. Ann. N.Y. Acad. Sci. 298:210-224. Reinert, R. E. 1970. Pesticide concentrations in Great Lakes fish. Pes- tic. Monit. J. 3:233-240. Risebrough, R. W., W. Walker II, T. T. Schmidt, B. W. de Lappe, and C. W. Connors. 1976. Transfer of chlorinated biphenyls to Antarcti- ca. Nature (Lond.) 264:738-739. 57 FISHERY BULLETIN: VOL. 78, NO. 1 Sims, G. G., J. R. Campbell, F. Zemlyak, and J. M. Graham. 1977. Organochlorine residues in fish and fishery products from the Northwest Atlantic. Bull. Environ. Contam. Toxicol. 18:697-705. Sladen, W. J. L., C. M. Menzie, and W. L. REICHEL. 1966. DDT residues in Adelie penguins and a crabeater seal from Antarctica. Nature (Lond.) 210:670-673. Snyder, D., and r. reinert. 1971. Rapid separation of polychlorinated biphenyls from DDT and its analogues on silica gel. Bull. Environ. Con- tam. Toxicol. 6:385-390. Stout, V. F., and F. L. beezhold. 1979. Analysis of chlorinated hydrocarbon pollutants: A simplified extraction and cleanup procedure for fishery products. Fish. Bull., U.S. 76:880-886. VEITH, G. D. 1975. Baseline concentrations of polychlorinated biphenyls and DDT in Lake Michigan fish, 1971. Monit. J. 9:21-29. Pestle. Young, D. R., d. j. mcDermott, and T. C. Heesen. 1976. DDT in sediments and organisms around southern California outfalls. J. Water Pollut. Control Fed. 48:1919-1928. ZITKO, V. 1978. The interference of aromatic hydrocarbons in the determination of PCB's. Proceedings, 4th Joint Confer- ence on Sensing of Environmental Pollutants, American Chemical Society, Pap. 201, p. 757-760. ZiTKO, v., O. HUTZINGER, and P. M. K. CHOI. 1972. Contamination of the Bay of Fundy — Gulf of Maine area with polychlorinated biphenyls, polychlorinated ter- phenyls, chlorinated dibenzodioxins, and dibenzofiirans. Environ. Health Perspect. 1:47-50. 58 SYSTEMATICS AND DISTRIBUTION OF CERATIOID ANGLERFISHES OF THE FAMILY MELANOCETIDAE WITH THE DESCRIPTION OF A NEW SPECIES FROM THE EASTERN NORTH PACIFIC OCEAN* Theodore W. Pietsch and John P. Van Duzer^ ABSTRACT The ceratioid anglerfish family Melanocetidae is revised on the basis of a study of approximately 600 specimens collected from all oceans. Of the 11 nominal species of Melanocetus based on females, 4 are recognized: M.johnsoni, with M. krechi, M. rotundatus, M. ferox, M. cirrifer, and M. megalodontis as synonyms; M. polyactis; M. niger; and M. murrayi , with M. vorax and M. tumidus as synonyms. A fifth species is newly described from a single female collected from the eastern Pacific Ocean off Mazatlan, Sinaloa, Mexico. The new form differs most strikingly from its allies in having a larger escal bulb and shorter jaw teeth. Melanocetus is widely distributed throughout all the major oceans of the world between about 250 m and some unknown lower depth limit that exceeds 3,000 m. Melanocetus johnsoni and M. murrayi are wide ranging forms, whereas M. polyactis and M. niger are apparently restricted to the eastern tropical Pacific. Melanocetus murrayi appears to be the most phylogenetically derived member of the family. The four remaining species are much more closely related to each other than any is to M. murrayi. Melanocetus johnsoni is perhaps derived in having a relatively long illicium, and in having fewer, but longer jaw teeth. Melanocetus polyactis and M. niger are similar in having relatively short jaw teeth, a similar escal morphology, and a sympatric geographic distribution that is limited to the eastern tropical Pacific. The newly described form is derived in having an extremely large escal bulb, comparable with no other known ceratioid. The Melanocetidae include globose, bathypelagic anglerfishes, easily separated from members of allied families by having 12 or more dorsal fin rays, 3 or 4 anal rays, and large, fanglike jaw teeth (Bertelsen 1951; Pietsch 1972a). The only recog- nized genus of the family was established by Giinther (1864) with the description of Mel- anocetus johnsoni, based on a single female specimen collected in the Atlantic Ocean, off Madeira. Since that time, 10 additional species based on females have been described (Table 1). From a comparison of the characters used to dis- tinguish these nominal forms, Bertelsen (1951, table 4) doubted that M. krechi and M. cirrifer could be maintained and that M. ferox and M. niger might be synonyms. Melanocetus murrayi and M. johnsoni were recognized as the only species known from the Atlantic; M. niger, M. ferox, and M. polyactis were considered forms re- stricted to the eastern tropical Pacific. Six larval 'Contribution No. 516 from the College of Fisheries, Univer- sity of Washington, Seattle, WA 98195. J'College of Fisheries, University of Washington, Seattle, WA joiyo. Manuscri rio'Ji'4'^"P' accepted September 1979. nSHERV BULLETIN: Vol. 78, NO. 1, 1980. specimens from the Gulf of Panama were assigned to M. polyactis. The remaining larvae (approxi- mately 600 individuals) were separated into two groups, representing M. murrayi and M.johnsoni, on the basis of geographic distribution, fin ray counts, and a comparison of larval and adolescent female pigmentation. Despite these allocations, Bertelsen (1951) made it clear that ". . . the sep- aration of the species is still very uncertain and future investigations and material will probably make it necessary to revise this synopsis." At the time of Bertelsen's (1951) monograph on the Ceratioidei, 19 metamorphosed melanocetid males were known. Of these, 14 had been set up as tjrpes of 12 separate species, and 5 were uncer- tainly placed. On the basis of subdermal pigment, fin ray counts, and geographic distribution, Ber- telsen (1951) synonymized 6 of these 12 nominal forms with M. johnsoni and 4 with M. murrayi. The remaining two species based on males, M. longirostris and M. nudus, each differing slightly from the rest of the material, were tentatively retained (Table 1). With the vast increase in the amount of mate- rial of Melanocetus made available in the last 25 59 FISHERY BULLETIN: VOL. 78, NO. 1 Table l. — Reallocation of nominal forms ofMelanocetus. Valid names on right. Synonymy for species based on males after Bertelsen 1951. Females: Melanocetus johnsoni Gunther 1864 Melanocetus krechi Brauer 1 902 Melanocetus rotundatus Gilchrist 1903 Melanocetus ferox Regan 1926 Melanocetus cirrifer Regan and Trewavas 1932 Melanocetus megalodontis Beebe and Crane 1 947 Melanocetus polyactis Regan 1925 Melanocetus niger Regan 1925 (in part) Melanocetus niger Regan 1925 (in part) Melanocetus murrayi Gunther 1887 Melanocetus vorax Brauer 1 902 Melanocetus tumidus Parr 1 927 Males: Centrocetus spinulosus Regan and Trewavas 1 932 Xenoceratias micracanthus Regan and Trewavas 1 932 Xenoceratias heterorhynchus Regan and Trewavas 1932 Xenoceratias laevis Regan and Trewavas 1932 Xenoceratias brevirostris Regan and Trewavas 1 932 Xenoceratias braueri Koefoed 1 944 Rhynchoceratias rostratus Regan 1926 (in part) Rhynchoceratias leucorhinus Regan 1926 (in part) Rhynchoceratias acanthlrostris Parr 1927 Rhynchoceratias latirhinus Parr 1927 Rhynchoceratias longipinnis Parr 1 930 Xenoceratias regani Koefoed 1 944 Melanocetus johnsoni Gunther 1 864 Melanocetus polyactis Regan 1 925 Melanocetus niger 1 925 Melanocetus murrayi Gunther 1 887 Melanocetus johnsoni Gunther 1 864 Melanocetus polyactis Regan 1 925 Melanocetus murrayi Gunther 1887 yr, we are able to recognize five species based on females. Four of these are previously described forms: M. johnsoni, represented by 346 specimens collected from all three major oceans of the world; M. polyactis and M. niger, known from 15 and 6 specimens both restricted to the eastern tropical Pacific; and M. murrayi, 140 specimens of worldwide distribution. The fifth is a new species recently collected by the Velero IV of the Univer- sity of Southern California in the eastern Pacific off Mazatlan, Sinaloa, Mexico. It differs strikingly from its allies in having a considerably larger escal bulb and shorter jaw teeth. Although the number of known male specimens has increased nearly fourfold since Bertelsen's (1951) work, no new diagnostic data are available. We have examined 73 individuals (11.5-24 mm standard length), none of which can be satisfactor- ily identified to species based on females. As pre- dicted by Bertelsen (1951), the variation in the number of denticular teeth is greater than previ- ously thought and values given in his key overlap to a much greater extent than is indicated. An attempt to utilize differences in larval pigmenta- tion, thought to be more or less retained, at least in the younger metamorphosed males, failed to sepa- rate the material into groups that could be as- sociated with species based on females. Although Bertelsen's ( 1951 ) synonymies for nominal species based on males are retained here, additional male specimens are listed as Melanocetus sp. METHODS AND MATERIALS Standard lengths (SL) are used throughout un- less otherwise stated. Measurements were taken from the left side whenever possible and rounded to the nearest 0.5 mm. To ensure accurate fin ray counts, skin was removed from the pectoral fins and incisions were made to reveal the rays of the dorsal and anal fins. Sockets, indicating missing teeth in the jaws and on the vomer, were included in total tooth counts. Jaw tooth counts are the sum of both right and left sides. Head depth is the distance from the tip of the sphenotic spine to the base of the quadrate spine. Head width is the dis- tance between the anterolateralmost margins of the sphenotic bones. Lower jaw length is the dis- tance from the symphysial spine to the posterior- most margin of the articular. Illicium length is the distance from the articulation of the pterygio- phore of the illicium and the illicial bone to the distal surface of the esca, excluding escal append- ages. The width of the pectoral fin lobe is the distance between the point of articulation of the uppermost fin ray to the articulation of the lower- most fin ray. Terminology used in describing the various parts of the angling apparatus follows Bradbury (1967). Definitions of terms used for the different stages of development follow Bertelsen (1951). Complete locality data are given for pri- mary type material only. The comparative osteological investigation was 60 PIETSCH and VAN DUZER: SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES based primarily on five female specimens (two M. murrayi, 75 and 84 mm SL, and three M.johnsoni, 44.5, 60, and 75 mm SL; material representing the remaining species of the genus was unavailable) cleared and stained with Alizarin red S following the trypsin digestion technique (Taylor 1967). Bone terminology follows Pietsch (1974). Unless otherwise indicated, all diagnoses and de- scriptions are based on female specimens >20 mm SL. For males and larvae see Bertelsen (1951). Material is catalogued in the following institu- tions: Australian Museum, Sydney ( AMS); British Museum (Natural History), London (BMNH); Bingham Oceanographic Collections, Peabody Museum of Natural History, Yale University (BOC); California Academy of Sciences, San Fran- cisco (CAS); Florida State Museum, University of Florida, Gainsville (FSM); Institute of Oceanol- ogy, Academy of Sciences, U.S.S.R., Moscow (lOAN); Institute of Oceanographic Sciences, Sur- rey, England (lOS); Institut fiir Seefischerei, Hamburg (ISH); Natural History Museum of Los Angeles County (LACM); Museum of Compara- tive Zoology, Harvard University (MCZ); National Museum of New Zealand, Wellington (NMNZ); Royal Ontario Museum, Toronto (ROM); South African Museum, Cape Town (SAM); University of Bergen, Zoological Museum (UBZM);University of Miami Marine laboratory, (UMML); National Museum of Natural History, Washington, D.C. (USNM); Virginia Institute of Marine Science, Gloucester Point (VIMS); Zoological Museum, Humboldt University, Berlin (ZMHU); Zoological Museum, University of Copenhagen (ZMUC). OSTEOLOGY OF FEMALES The osteology of Melanocetus was partially de- scribed by Regan (1926, fig. 10), Parr (1930a, fig. 2-5, male only), Regan and Trewavas (1932, fig. 19-28), and Bertelsen ( 1951, fig. 13, 14). In the following account, only those comparative aspects that need amending or that have not previously appeared in the literature are discussed. Cranium (Figures 1-9). — The anterior portion of the cranium of Melanocetus is considerably wider, relative to the posterior portion, than in other ceratioids; the distance between the lateral mar- gins of the ethmoid cartilage is nearly equal to the Sphenotic Posttemporal Supraethmoid Vomer Exoccipital EpiotJc Lateral ethmoid Pterotic Pterosphenoid Supraoccipital Figure l.— Dorsal view of cranium of Melanocetus johnsoni, LACM 32786-1, 75 mm SL. Cartilage stippled, open spaces solid black. 61 FISHERY BULLETIN: VOL 78, NO. 1 distance between the tips of the sphenotic bones (Figures 1, 2) The head of the vomer, bearing as many as 10 recurved teeth, is also unusually wide (Figures 1, 2, 5-7). The frontals are triradiate in shape and widely separated from each other along their dorsal margins, approaching one another on the midline only at their ventromedial extensions. A semicircular-shaped pterosphenoid is present under the posterior extension of each frontal. The parasphenoid is well separated from the ven- tromedial extensions of the frontals. Posterome- dially, the parasphenoid underlies the anterior Sphenotic Pterosphenoid Parietal Supraethmoid Vomer Lateral ethmoid Posttemporal Exoccipital Epiotic Frontal Pterotic Supraoccipital Figure 2.— Dorsal view of cranium oi Melanocetus murrayi, LACM 31501-3, 84 mm SL. Cartilage stippled, open spaces solid black. Pterosphenoid Sphenotic /Parietal Posttemporal Lateral ethmoid Vomer Supraethmoid Exoccipital 20th pre-ural centrum Basioccipital Parasphenoid Pterotic Supraoccipital Prootic Figure 3.— Lateral view of cranium of Melanocetus johnsoni , LACM 32786-1, 75 mm SL. Cartilage stippled, open spaces solid black. 62 PIETSCH and VAN DUZER: SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES projection of the supraoccipital; posteriorly di- rected dorsolateral extensions of the parasphenoid make contact laterally with the respective prootic (Figures 3, 4). The large prootics are separated from each other anteriorly by the anterior process of the supra- occipital. Ventrally, each prootic forms a rela- tively large, anterolaterally directed, conical pro- jection not found in other ceratioids (Figures 3-5, 8). The supraoccipital is the largest element of the cranium, making up a considerable portion of the roof of the cranium. Together with the frontals, the supraoccipital forms the floor of a deep, V-shaped illicial trough (Figure 8). An anteriorly directed extension of the supraoccipital that sep- Pterosphenoid Sphenotic Parietal Frontal Supraethmoid Lateral ethmoid Vomer Posttemporal Exoccipital 20th pre-ural centrum Basloccipital Parasphenoid Supraoccipital Pterotic Prootic FIGURE 4.— Lateral view of cranium of Melanocetus murrayi, LACM 31501-3, 84 mm SL. Cartilage stippled, open spaces solid black. Pterosphenoid Prootic Supraethmoid Pterotic Parasphenoid Exoccipital Vomer 20th pre-ural centrum Basioccipital Lateral ethmoid Posttemporal Frontal Sphenotic Figure 5.— Ventral view of cranium of Melanocetus murrayi, LACM 31501-3, 84 mm SL. Cartilage stippled, open spaces solid black. 63 Supraethmold Frontal Lateral ethmoid Vomer Figure 6.— Anterior view of anterior half of cranium of Melanocetus johnsoni, LACM 32786-1, 75 mm SL. Cartilage stippled. Supraethmold Lateral / ethmoid Frontal Vomer Figure 7. — Anterior view of anterior half of cranium of Melanocetus murrayi, LACM 31501-3, 84 mm SL. Cartilage stippled. arates the prootics on the midline, is narrowly separated by cartilage from the ends of the ventromedial extensions of the frontals. The supraoccipital is not overlapped by the parietals (Figures 1-5). The cranium of M. murrayi is considerably more elongate and compressed compared with that of its congeners (Figures 3, 4). As a consequence, the anterior margin of the vomer of M. murrayi is deeply concave (nearly straight in other forms), the frontals are more elongate and considerably lighter in construction, the sphenotics are much less produced forward, and the parietals do not extend posteriorly to overlap the posttemporals as they do in other Melanocetus species (Figures 1, 2). Mandibular arch (Figures 10, 11). — The anterior portion of the premaxilla bears a short ascending process and a slightly longer articular process. A small (compared with those of oneirodids, Pietsch 1974) symphysial cartilage lies just behind the FISHERY BULLETIN; VOL. 78, NO. 1 Supraoccipital Cut surface of frontal Epiotic Pterosplienoid Cut surface of supraoccipital Splienotic Prootic Cut surface of parasplienoid FIGURE 8. — Anterior view of posterior half of cranium of Melanocetus murrayi, LACM 31501-3, 84 mm SL, anterior por- tion removed. Cartilage stippled. Epiotic Parietal Sphenotic Posttemporal Pterotic Exoccipital 20th pre-ural centrum Figure 9. — Posterior view of cranium of Melanocetus murrayi, LACM 31501-3, 84 mm SL. Cartilage stippled. posteriorly notched symphysis of the premaxillae. There is no postmaxillary process of the pre- maxilla. The elongate portion of each premaxilla may bear up to 89 recurved, depressible teeth of mixed sizes (Figure 10). On each side, the posterior ends of the pre- maxilla and maxilla are united by connective tis- sue to each other and to the lateral surface of the ascending process of the articular of the lower jaw, preventing any forward rotation of the upper jaw bones to close off the corners of the mouth. There is no elongate, anterior maxillomandibular liga- ment originating on the dentary as in oneirodids (labial cartilage of Le Danois 1964; Pietsch 1972a, 1974). The dentaries meet on the midline to form a strong symphysial spine. Each dentary may bear up to 71 recurved, depressible teeth of mixed sizes (Figure 11). Palatine and hyoid arches (Figures 11, 12). — The distal portion of the palatine arch (including the mesopterygoid, ectopterygoid, and palatine) is elongate and slender throughout (Figure 11). The small mesopterygoid is in contact with the metap- 64 PIETSCH and VAN DUZER: SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES terygoid. The suspensorium is unusually narrow and elongate (due largely to the extremely narrow and elongate quadrate), and directed obliquely backward. The posterior head of the hyomandibu- lar is the larger of the two heads forming a broad articulation with the pterotic. The interhyal is Maxilla Premaxilla Symphyslal cartilage FIGURE 10. — Elements of upper jaw of Melanocetus murrayi, LACM 31501-3, 84 mm SL: A. Maxilla and premaxilla, left lateral view; B. Symphysis of premaxillae, dorsal view. short and relatively thick (compared with that of oneirodids, Pietsch 1974). The hyoid apparatus (including epihyal, ceratohyal, and upper and lower hypohyals) is re- latively short and thick (Figure 12). The lower hypohyal extends down beyond the ventral mar- gin of the ceratohyal. In other ways this portion of the hyoid arch does not differ substantially from that described for oneirodids (Pietsch 1974). Opercular apparatus (Figure 11). — The opercu- lar apparatus is somewhat reduced (compared with that of oneirodids, Pietsch 1974). The opercle Dorsal hypohyal Ceratohyal Epihyal Ventral hypohyal ' Branchiostegal rays FIGURE 12. — Lateral view of hyoid apparatus of Melanocetus murrayi, LACM 31501-3, 84 mm SL, interhyal not shown. Car- tilage stippled. Hyomandibular Symplectic Interhyal Palatine Ectopterygoid Figure ll. — Lateral view of lower jaw, suspensorium, and opercular apparatus of Melanocetus murrayi, LACM 31501- 3, 84 mm SL. Cartilage stippled, open spaces solid black. Subopercle Preopercle Interopercle Quadrate Dentary Articular 65 is notched posteriorly, but the upper fork of this bone is considerably shorter than the lower fork and sometimes absent. The subopercle is narrow and elongate, the upper part tapering to a fine point, the lower part rounded with a well- developed anterior spine or projection. The in- teropercle is unusually long and slender. The preopercle is more or less straight. Branchial arches (Figure 13). — Pharyngobran- chials I and IV are absent; those of the second and third arches are well developed, bearing four to nine recurved and depressible fangs. Epibranchial FISHERY BULLETIN: VOL. 78, NO. 1 I is reduced lying free in the connective tissue matrix. Ceratobranchial V is also reduced but tightly connected to the medial-proximal margin of ceratobranchial IV. There are three hypo- branchials, and a single basibranchial ossification surrounded by a triangular-shaped cartilage. Vertebrae and caudal skeleton (Figure 14). — The vertebral column forms a sigmoid curve, dipping down behind the region of the gut and sloping up again to support the tail. In the five cleared and stained specimens examined there are 20 verte- bral centra (including the half-centrum to which is Hypobranchial I Ceratobranchial I Epibranchial I Pharyngobranchial II Basibranchial Ceratobranchial V Epibranchial IV Figure 13.— Branchial arches of Melanocetus murrayi, LACM 31501-3, 84 mm SL. The ventral portion of the branchial basket is shown in dorsal view, the dorsal portion (epibranchials and pharyngobranchials) is folded back and shown in ventral view. Cartilage stippled. Pharyngobranchial III Figure 14. — Lateral view of vertebral col- umn, dorsal and anal fins, and caudal skele- ton oi Melanocetus murrayi, LACM 31501- 3, 84 mm SL. Cartilage stippled. 66 PIETSCH and VAN DUZER: SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES fused the hypural plate, Pietsch 1972a) of which 13-15 are caudal vertebrae (those bearing com- plete haemal arches). Epurals are absent. The hypural plate is unnotched posteriorly and bears the overlapping bases of nine principal caudal rays, the uppermost of which is exceptionally large. The uppermost and two lowermost caudal rays are simple, the central rays are bifurcated distally. Median fins and illicial apparatus (Figures 14, 15). — There are 13-19 biserial, segmented, and unbranched dorsal fin rays, the number varying somewhat among species. The rays are supported by elongate, closely associated radials, usually one less than the number of rays. All species of the genus have 4 anal fin rays (of 353 specimens counted, only 2 had 3 anal rays, and only 2 had 5 rays) that are like those of the dorsal fin, invari- ably supported by 3 similar, closely associated ra- dials (Table 2, Figure 14). The pterygiophore of the illicium is strongly compressed with a thin, bladelike ventral expan- sion. The length of the pterygiophore varies from 17*7^ SL in M. murrayi to 33% SL in M.johnsoni. The remnant of the second cephalic ray is a minute ossification lying on the pterygiophore just behind the articulation with the illicial bone (Figure 15). The length of the illicial bone varies slightly among Melanocetus species, becoming longer pro- portionately with growth. Pectoral girdle, pectoral fin, and pelvic bone (Fig- ure 16). — Each posttemporal overlaps the re- spective pterotic, epiotic, and exoccipital. It is in turn overlapped by the parietal in M.johnsoni, hut widely separated from this bone in M. murrayi. An ossified posteroventral process of the coracoid is absent but perhaps represented by a posteroventral cartilaginous extension. There are four pectoral radials, the lower two of which be- come completely fused with each other giving the appearance of only three radials (Regan and Tre- wavas 1932, fig. 22; Pietsch 1972a). The pectoral fin lobe of M. murrayi is consid- erably smaller than that of other Melanocetus species (Figure 16 A). In other ways the elements of the pectoral girdle, pectoral fin, and pelvic bone do not differ substantially from those of oneirodids (Pietsch 1974). Skin spines. — Minute dermal spines (approxi- mately 0.03-0.1 1 mm long) are present in the skin Illicial bone Remnant of 2nd cephalic ray Pterygiophore of illicium Figure 15. — Bones of illicial apparatus of Melanocetus murrayi, LACM 31501-3, 84 mm SL, left lateral view. Cartilage stippled. of the two species examined osteologically. In M. johnsoni they are most numerous on the side of the trunk under the dorsal fin ( where there are about 6 spines/mm2)but become progressively more widely scattered anteriorly and finally disappear in the area of the upper and lower jaws (Struhsaker 1962). In the two specimens of M. murrayi examined osteologically the spines are confined to the caudal peduncle. SYSTEMATICS Family Melanocetidae Regan 1912 Type genus Melanocetus Giinther 1864 Diagnosis. — The metamorphosed females of the Melanocetidae are distinguished from those of all other ceratioid families by having the following combination of characters: jaws equal anteriorly; supraethmoid present; parietals present; ptero- sphenoid present; anterior maxillomandibular lig- ament absent (Pietsch 1972a); hyomandibular with a double head; hypohyals 2; branchiostegal rays 6; operculum bifurcate, upper fork reduced; suboperculum slender, as long as lower fork of operculum, with strong anterior spine; pharyn- gobranchials I and IV absent; epibranchial I reduced; a single ossified basibranchial; epibranchial and ceratobranchial teeth absent; epurals absent; only an ossified remnant of second cephalic ray present; dorsal fin rays 13-17, anal fin 67 FISHERY BULLETIN: VOL 78, NO. 1 Table 2. Fin ray frequencies for females of Melanocetus species. Species Melanocetus johnsoni Melanocetus polyactis Melanocetus niger Melanocetus eustalus Melanocetus murrayi Total 32 38 Dorsal 12 13 14 15 16 17 3 70 4 3 29 106 64 4 2 1 71 Anal 1 136 13 5 1 62 1 217 Pectoral (both sides) 1 5 15 16 17 18 19 20 21 22 23 1 13 34 69 36 12 2 1 2 3 12 4 2 1 2 2 5 1 2 5 17 37 8 5 1 2 5 19 52 47 88 46 15 3 Supracleithrum B ^ Radials Coracoid Cleithrum Postcleithrum Figure 16. — Lateral view of pectoral girdle, pectoral radials, and pelvic bone: A. Melanocetus murrayi, LACM 31501-3, 84 mm SL; B. M. johnsoni, LACM 32786-1, 75 mm SL. Cartilage stippled. Pelvic bone rays 4 (rarely 3 or 5), caudal fin rays 9 (1-6-2); ossified posteroventral process of coracoid absent; pectoral radials 4, fusing to 3 with growth; pelvic bones expanded distally; esca without denticles; minute, widely spaced skin spines present in at least some species. The metamorphosed males of the Melanocetidae are distinguished from those of all other ceratioid families in having the following combination of characters: free-living; jaw teeth absent; upper denticular with 2-3 semicircular series of strong, recurved denticles, fused with a median series of 3-9 enlarged dermal spines that articulate with the pterygiophore of the illicium; lower denticular with 10-23 recurved denticles, fused into a median and two lateral groups; eyes directed laterally, elliptical in shape, pupil larger than lens; olfac- tory organs large, nostrils lateral, nasal area un- pigmented, inflated; dorsal fin rays 12-16, anal fin rays 4, caudal fin rays 9 (1-6-2); skin spinulose or naked. Description. — Females relatively short and deep, globular (but often appearing highly compressed apparently due to deformation upon capture, com- pare Figures 17, 18); head short; mouth large, nearly vertical, cleft not extending past eye; lower jaw with a well-developed symphysial spine; oral valve weakly developed; two nostrils on each side on distal surface of a rounded papilla; eye small, subcutaneous, appearing through a circular, translucent area of integument within a shallow, orbital pit formed between sphenotic and frontal bones; gill opening oval in shape, situated posteri- 68 PIETSCH and VAN DUZER: SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES Figure 17. — Anterior view of Melanocetus johnsoni, LACM 31484-1, 85 mm SL. Drawn by Elizabeth Anne Hoxie. or to pectoral lobe; all four epibranchials closely bound together by connective tissue; anterior half of ceratobranchial I bound to medial surface of ceratohyal by connective tissue, posterior half free; gill filaments present on anteriormost tip of ceratobranchial I and full length of cerato- branchials II through IV; pseudobranch absent; no opening behind fourth gill arch; ovaries paired; pyloric caeca absent. Illicial length 23.1-60.8^f SL; anteriormost tip of pterygiophore of illicium exposed, emerging on snout between eyes, its posterior end concealed under skin; escal bulb simple, usually with a rounded or conical, distal prolongation, and often with posterior and anterior crests; elongate ap- pendages and filaments absent. Jaw teeth slender, recurved, and depressible, some slightly hooked distally, those in lower jaw less numerous, but slightly longer than those in upper jaw; number of teeth in lower jaw 32-142, in upper jaw 29-178; longest tooth in lower jaw 6.9- FlGURE 18.— Holotype of Melanocetus eustalus, LACM 30037- 12, 111 mm SL, anterior view. Drawn by Elizabeth Anne Hoxie. 69 FISHERY BULLETIN: VOL. 78, NO. 1 25.0'7r SL; vomer with 0-12 teeth; pharyngobran- chials II and III heavily toothed. Color in preservative dark brown to black over entire surface of body (except for distal portion of escal bulb) and oral cavity; all fins colorless in specimens less than about 40 mm SL (except for caudal rays in adolescent M. murrayi, Bertelsen 1951, fig. 161). Pectoral fin rays 15-23 (Table 2); pelvic fins ab- sent. The following measurements, in percent of standard length, are summarized for females (20- 120 mm SL) of all species: head depth 42.5-82.0; least outside width between frontals 9.1-28.6; head width 22.6-45.0; premaxillary length 36.3- 76.0; lower jaw length 36.7-78.0; width of pectoral fin lobe 6.1-17.8; escal bulb width 1.9-11.3 For description of males see Diagnosis above and Bertelsen (1951). Genus Melanocetus Giinther 1864 Females Melanocetus Giinther 1864:301-302, pi. 25 (type species Melanocetus johnsoni Giinther 1864, by monotypy). Melanocetus (subgenus Liocetus) Giinther 1887:56, pi. 11, fig. A (type species Melanocetus murrayi Giinther 1887, by monotypy). Liocetus Goode and Bean 1896: 495-496, fig. 407 (type species Melanocetus murrayi Giinther 1887, by monotypy). Melanocoetus Smith 1949:429 (erroneous spelling of Melanocetus, therefore taking the same type species, Melanocetus johnsoni Giinther 1864). Linocetus Bertelsen 1951:40, 44 (erroneous spell- ing of Liocetus, therefore taking the same type species, Melanocetus murrayi Giinther 1887). Males Rhynchoceratias Parr 1927:30-33, fig. 11-12 (in part; type species Rhynchoceratias brevirostris Regan 1925, by subsequent designation of Fowler 1936). Centrocetus Regan and Trewavas 1932:53, fig. 79 (type species Centrocetus spinulosus Regan and Trewavas 1932, by monotypy). Xenoceratias Regan and Trewavas 1932:54-57, fig. 80-84 (type species Xenoceratias longirostris Regan and Trewavas 1932, by subsequent des- ignation of Fowler 1936). Diagnosis and description same as for family. Key to Species Based on Females The following key will differentiate female specimens >20 mm SL (for males and larvae see Bertelsen 1951). The key should be used in conjunction with Figures 19-24. lA. IB. 2A. 2B. 3A. 3B. 4A. Escal bulb width 11.3% SL in 111 mm specimen (Figures 18, 28); longest lower jaw tooth 5.9% SL in 111 mm specimen Melanocetus eustalus n. sp. (single known female) Escal bulb width <10% SL (Figure 17); longest lower jaw tooth 6.9-25.0% SL 2 Anterior margin of vomer deeply concave (Figure 2); least outside width between frontals 9.1-17.8% SL (Figure 19); number of lower jaw teeth 46-142 (>60 in specimens 25 mm and larger) (Figure 20); escal bulb width 1.9-5.1% SL (<3% SL in specimens >50 mm SL) Melanocetus murrayi Giinther 1887 Anterior margin of vomer nearly straight (Figure 1); least outside width between frontals 13.5-28.6% SL (Figure 19); number of lower jaw teeth 32-90 (Figure 20); escal bulb width 3.8-8.6% SL (>4% SL in specimens >50 mm SL) 3 Longest lower jaw tooth 8.4-25.0% SL (Figure 21); esca with compressed posterior and (usually) anterior crests (Figure 25); distribution nearly cosmopolitan Melanocetus johnsoni Giinther 1864 Longest lower jaw tooth 6.9-13.1% SL (Figure 21); esca without posterior or anterior crests (Figures 26, 27); distribution restricted to eastern tropical Pacific 4 Number of lower jaw teeth 58-90 (Figure 22); escal bulb width 5.2-8.5% SL (Figure 23); illicium length 34.6-56.0% SL (Figure 24); escal bulb with a conical, distal prolongation occasionally pigmented on tip (Figure 26) Melanocetus polyactis Regan 1925 70 PIETSCH and VAN DUZER: SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES 4B. Number of lower jaw teeth 37-57 (Figure 22); escal bulb width 3.8-5.0% SL (Figure 23); illicium length 29.8-38.8% SL (Figure 24); escal bulb with a low, rounded distal prolon- gation usually darkly pigmented on tip (Figure 27) Melanocetus niger Regan 1925 20 • M. johnsoni n = 107 I I I 1 T -1— r 1 — F °M. murrayi n = 61 c 17 5 • ' o 15 % • • • • • •• • - o 12 5 100 • • • •••• • • • :/oB o o o o o ■a 3 O c/> 75 50 2 5 . • • • •t • , •• • • o • o8 o o 'm ••° o °o^oo° Co Q r^ O O Ooo o ^ ° o OOo o o o o o° o . . 1 1 I 1 I 1 1 1 1 1 1 1 , 1 ' 1 1 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 Standard Length in mm Figure 19. — Relationship between least outside width of frontals and standard length for two species oi Melanocetus. Melanocetus johnsoni Giinther 1864 Figures 1, 3, 6, 16B, 17, 19-21, 25, 30, 31 Females Melanocetus johnsoni Giinther 1864:301-303, pi. 25 (original description, single specimen, holotype BMNH 1864.7.18.6, 64 mm, Madeira, 24 December 1863); Liitken 1871:64, 74 (com- parison with Oneirodes eschrichtii); Liitken 1872:329-340, 343 (after Liitken 1871); Giinther 1880:473, fig. 211 (after Giinther 1864); Giinther 1887:56-57 (after Giinther 1864, com- parison with M. murrayi); Vaillant 1888:346 (after Giinther 1864); Goode and Bean 1896:494, fig. 406 (description after Gunther 1864); Gill 1909:582, 584, 585, fig. 20 (after Gunther 1864, Goode and Bean 1896); Regan 1912:286, fig. 60 (cranial osteology); Regan 1913:1096 (descrip- tion of additional specimen, natural history); Regan 1926:18, 32, 33, fig. 10 (description of additional material, cranial osteology; M. krechi and M. rotundatus synonyms); Parr 1927:29 (description of additional specimen); Norman 1930:354 (additional record); Regan and Trewavas 1932:27-29, 49-52, fig. 19-21, 22A, B, 72, 73 (description of additional mate- rial, osteology, and esca figured, in key); Fowler 1936:1143, 1144, 1346, 1363 (description after Giinther 1864, Regan 1926, Norman 1930); Norman 1939:114 (additional material); Koe- foed 1944:3-5, pi. 1, fig. 1 (description of addi- tional specimen, comparison with M. murrayi; Beebe and Crane 1947:152 (description of addi- tional specimen, color); Fowler 1949:158-159 (listed); Bertelsen 1951:7, 40-41, 43-46, 48-53, fig. 13, 15, 17-19, tables 4, 6 (description of females, males, larvae, comparison with all known material, in key); Grey 1956:235-236 (synonymy; distribution); Monod 1960:687, fig. 80 (pectoral radials); Maul 1961:91-92, fig. 1 (de- scription of additional material); Maul 1962a:6-7 (description, additional material); Struhsaker 1962:841-842 (description, addi- tional specimen, skin spines); Bussing 1965:222 (additional specimen); Fitch and Lavenberg 1968:127, fig. 70 (distinguishing characters, natural history); Pietsch 1972a:29, 35, 36, 38, 45 (osteological comments); Maul 1973:667 (synonymy, after Bertelsen 1951). Melanocetus krechi Brauer 1902:293-294 (original description, single specimen, holotype ZMHU 17688, 45 mm, Valdivia stn. 239, Indian Ocean, 5°42' S, 43°36' E, 0-2,500 m); Brauer 1906:319- 320, pi. 15, fig. 1, 2 (description after Brauer 1902); Gill 1909:583, 584, fig. 21 (after Brauer 71 FISHERY BULLETIN: VOL. 78, NO. 1 1 -» « -I T 1 1 1 1 1 1— T 1 r 1 1 I 1 1 I 1 1 — 140 • M. johnspni n = 116 0° 130 o M. murrayi n = 66 80 120 - - 110 r°-> - 100 ° 0^0 ° 90 80 • °o • • - 70 .0° ° • •• • • • • • • • . % - 60 - • • •• * •^* • • • • - .0 -»^„ • • 1 • • •• • «• • • SO 40 • • • • • • • • • • • • • - 30 »n J 1 1 — 1 , 1 1 1 1 1 1 1 1 1 1 1 . 10 IS 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 Standard length in mm Figure 20. — Relationship between number of lower jaw teeth and standard length for two species of Melanocetus. 1902, 1906); Murray and Hjort 1912:87, 614 (in part, additional specimen, misidentification); Borodin 1931:84 (additional specimen, misiden- tification); Regan and Trewavas 1932:49, 52, fig. 74 (misidentification, in key); Fowler 1936: 1143, 1144 (description after Brauer 1902, 1906, in key); Bertelsen 1951:40, table 4 (com- parison with all known material). Melanocetus rotundatus Gilchrist 1903:206-208, pi. 15 (original description, two specimens, both lost [see Comments, p. 76], the largest about 28 mm, off Cape Point and Natal coast. South Af- rica, 1,098 m); Gilchrist and Thompson 1917:417 (after Gilchrist 1903); Barnard 1927:1007, pi. 37, fig. 5 (after Gilchrist 1903); Bertelsen 1951:48 (in synonymy of M. john- soni). Melanocoetus rotundatus, Smith 1949:429, fig. 1232 (after Gilchrist 1903); Penrith 1967:187, 188 (type material lost; a synonym of M. johnsoni). Melanocetus ferox Regan 1926:33, pi. 9, fig. 1 (orig- inal description, single specimen, holotype ZMUC P9257, 78 mm, Dana stn. 1208(14), Gulf of Panama, 6°48' N, 80°33' W, 3,100 m wire, 1715 h, 16 January 1922); Regan and Trewavas 1932:49, 52, fig. 75 (in part, only holotype, addi- 72 tional material here referred to M. polyactis, in key); Beebe and Crane 1947:152 (in part, only holotype); Bertelsen 1951:44, 53, table 4 (in part, only holotype; comparison with all known material, in key); Grey 1956:237 (synonymy, distribution). Melanocetus cirrifer Regan and Trewavas 1932:52-53, fig. 76A, 77, pi. 2, fig. 1 (original description, two females, lectotype ZMUC P9258, 25.5 mm, Dana stn. 3678(2), Banda Sea, 4°05' S, 128°16' E, 4,000 m wire, bottom depth 4,700 m, 1840 h, 24 March 1929, in key); Ber- telsen 1951:44, 53, table 4 (description, com- parison with all known material, in key); Grey 1956:237 (synonymy, distribution). Melanocetus niger, Gregory 1933:400, fig. 272 (misidentification, osteology). Melanocetus megalodontis Beebe and Crane 1947:152, fig. 1 (original description, single specimen, holotype CAS-SU 46488 [originally NYZS 25791], 25.5 mm, Templeton Crocker Ex- pedition stn. 165 T-3, eastern tropical Pacific, 20°36' N, 115°07' W, 0-915 m, 17 May 1936); Bertelsen 1951:43, 48, table 4 (description, com- parison with all known material, in key); Grey 1956:235 (synonymy, distribution); Mead 1958:133 (holotype passed to CAS). PIETSCH and VAN DUZER: SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES 12 5 1 1 T — • M. johnsoni ▼ M. polyactis 1 1 1 1 II n =139 n = i3 , 1 1 1 1 1 1 1 • T 1 r 1 1— 1 1— 12 A M. niger n =6 • - 115 - • • 110 - • •• • - 10 5 - « - 100 " . • . • • • • - 95 - • • • - 90 " . « . • - 85 ■" • . • .. • - 80 - • - 75 - •• • • - 70 ■" . . . • • - 65 • • ••• • ^ - 60 • • T - 55 • •• •• a - 50 • • . • . ▼ - 45 • • • - 40 • .»•: " 35 •• • . : - - 30 2 5 • ... • IV.. T - 2 -.. .▼ A - 15 1 n T 1 1 1 1 1 1 J 1 1 1 1 I . ' 1 1 1 1 1 ' 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 Standard length in mm Figure 21. — Relationship between length of longest lower jaw tooth £ind standard length for three species oi Melanocetus. Melanocetus sp. Roule and Angel 1930:121, pi. 6, fig. 159 (additional material). Males Centrocetus spinulosus Regan and Trewavas 1932:53, 54, fig. 79 (original description, two specimens, lectotype ZMUC P9246, 15.5 mm, Dana stn. 3847(2), Indian Ocean, 12°02' S, 96°43' E, 3,000 m wire, bottom depth 2,825 m, 2100 h, 11 October 1929). Xenoceratias macracanthus Regan and Trewavas 1932:11, 12 (erroneous spelling of specific name, listed). Xenoceratias micracanthus Regan and Trewavas 1932:54, 55, fig. 81 (original description, single 73 FISHERY BULLETIN: VOL. 78, NO. 1 Figure 22.— Relationship between number of lower jaw teeth and standard length in two species oi Melanocetus . 100 90 .c 'a> 80 a> .2. ''° a> S 60 - o ° 50 40 30 20 ▼ M. pojyactis n = i5 A M. niger n=6 »y ▼ I I I 1 L. I I I 1- 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 Standard length in mm Figure 23.— Relationship between escal bulb width and standard length in two species oi Melanocetus . E E c: 4.5 1 -r- 1 1 1 1 1 1 1 1 1 I 1 1 4.0 - A M. M. polyactis n = niger n=5 11 A 3.5 - T ▼ " 3.0 - ▼ A 2.5 - ▼ - 2.0 - T W ▼ ▼ A A - 1.5 - T A - 1.0 - ▼ .5 n - 1 1 r 1 1 1 1 1 1 1 1 J J 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 Standard length in mm specimen, holotype ZMUC P9250, 28 mm, Dana stn. 4000(8), eastern tropical Atlantic, 0°31' S, 11°02' W, 4,000 m wire, bottom depth 3,750 m, 0630 h, 4 March 1930, in key); Fowler 1936:1364 (listed). Xenoceratias heterorhynchus Regan and Tre- wavas 1932:54, 56, fig. 82 (original description, single specimen, holotype ZMUC P9248, 27 mm, Dana stn. 3716(2), South China Sea, 19°18.5' N, 120°13' E, 3,000 m wire, bottom depth 3,225 m, 1400 h, 22 May 1929, in key); Grey 1956:236 (synonymy, distribution). Xenoceratias laeuis Regan and Trewavas 1932:54, 57, fig. 83 (original description, single specimen, holotype ZMUC P9249, 23 mm, Dana stn. 3731( 13), South China Sea, 14°37 ' N, 1 19°52 ' E, 2,000 m wire, bottom depth 2,300 m, 0200 h, 17 June 1929, in key). Xenoceratias brevirostris Regan and Trewavas 1932:54, 57, fig. 84 (original description, single specimen, holotype ZMUC P9247, 19 mm, Dana stn. 3739(8), Celebes Sea, 3°20' N, 123°50' E, 3,000 m wire, bottom depth 4,475 m, 0700 h, 2 July 1929, in key). Xenoceratias braueri Koefoed 1944:6, fig. 2 (origi- nal description, single specimen, holotype UBZM 4309, 18.5 mm, Michael Sars North At- lantic Deep-Sea Expedition stn. 53, central North Atlantic, 34°59' N, 33°01' W, 2,600 m wire, bottom depth 2,615-2,865 m, 8-9 June 1910). Melanocetus johnsoni, Bertelsen 1951:44, 48-53, fig. 17C, D, F-H, table 6 (synonymy, distribu- tion, comparison with all known material, in key); Grey 1956:236 (synonymy, distribution); Maul 1962b:36-37, fig. 2 (description of addi- 74 PIETSCH and VAN DUZER; SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES FIGURE 24.— Relationship between Ulicium length and standard length in two species of Melanocetus. E E e: 32 5 — r T — r- M. pjolyacLis n = i2 I r r - -» 1 1 A 30.0 - A M. niger n = 6 275 - T A - 25.0 - - 225 - ▼ - 20.0 - ▼ - 175 - ▼ A - 15.0 - T ▼▼ A 12.5 - T ▼ T A 10.0 - T T A - 75 - 5.0 9 «i - — 1_ I 1 1 _l 1 1 1 1 - to 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 Standard length in mm tional specimen); Maul 1973:667 (synonymy, after Bertelsen 1951). Material. — Metamorphosed females, 346 (10-119 mm): AMS, 33 (11-88 mm); BMNH, 16 (13-64 mm); CAS, 1 (25.5 mm); FSM, 10 (13-76 mm); lOAN, 35 (12-75 mm); lOS, 7 (35-91 mm); ISH, 82 (15-119 mm); LACM, 65 (13.5-83 mm); MCZ, 25 (12-75 mm); NMNZ, 7 (12-55 mm); ROM, 4 (15-56 mm); SAM, 6 (10-13 mm); UMML, 3 (27-82 mm); USNM, 15 (12-85 mm); ZMUC, 37 (11.5-89 mm). Diagnosis. — A species of Melanocetus unique in having the following combination of characters: anterior margin of vomer nearly straight (Figure 1); least outside width between frontals 13.5- 28.67c SL (Figure 19); number of lower jaw teeth 32-78 (Figure 20), length of longest lower jaw tooth 8.4-25.0% SL (Figure 21); width of pectoral fin lobe 10.7-17.87f SL; escal bulb width 4.3-8.6% SL; illicium length 32.4-60.8'y^ SL; esca with pos- terior and (usually) anterior crests (Figure 25); minute skin spines present over most of body; integument relatively thick (1.55 mm). Description. — Escal bulb slightly compressed with a low, rounded or conical distal prolongation nearly always darkly pigmented on tip; a com- pressed posterior crest usually darkly pigmented, becoming larger and more conspicuous with growi;h; and a considerably smaller, compressed, anterior crest present in some specimens (Figure 25); integument relatively thick (cross sections measure 1.55 mm in thickness), not easily torn, usually retaining heavy pigmentation during fixation and preservation. Number of upper jaw teeth 48-134; dorsal fin rays 13-15 (rarely 16), pectoral fin rays 17-22 (rarely 23) (Table 2). Distribution. — Melanocetus johnsoni has a wide horizontal distribution in tropical and subtropical waters of all three major oceans of the world (see Distribution, p. 83). Compared with M. murrayi, it appears to occupy relatively shallow depths: about 62% of the material (for which data was available) was captured by open nets fished at maximum depths of 1,000 m; 827c of the material can be accounted for by gear fished above 1,500 m, and 98% by gear fished above 2,100 m (see Distribu- tion, p. 83). Comments. — Melanocetus krechi Brauer (1902) was synonymized with M. johnsoni by Regan (1926), resurrected by Regan and Trewavas (1932), and tentatively synonymized again with M. johnsoni by Bertelsen (1951). From the de- scription and figure given by Brauer (1902, 1906) 75 FISHERY BULLETIN: VOL. 78, NO. 1 I .5 w , 2 mm FIGURE 25.— Escae ofMelanocetusjohnsoni: A. ISH 1261/71, 21 mm SL; B. ISH 753/71, 38 mm SL; C. MCZ 49849, 75 mm SL; D. ISH 1534/71, 78 mm SL. and based on a much greater knowledge of varia- tion within the genus, there can be little doubt that this nominal form has been correctly placed within the synonymy of M. johnsoni. Melanocetus ferox was described from a single specimen (78 mm) collected in the Gulf of Panama (Regan 1926). Two additional specimens of this nominal form were listed by Regan and Trewavas (1932). A thorough comparison of all known mate- rial led Bertelsen ( 1951 , table 4) to suspect that M. ferox might represent individual variation of M. niger. The holotype of M. ferox, however, has rela- tively long lower jaw teeth (longest, 12.0% SL; Figure 21). In this, and in all other morphometric and meristic characters used here, it fits well within the material here recognized as M. johnsoni. Although the esca of the holotype is in poor condition, traces of a posterior crest remain. For these reasons M. ferox is synonymized with M. johnsoni. The two additional specimens identified as M. ferox by Regan and Trewavas (1932) (ZMUC P92210, 30.5 mm; BMNH 1932.5.3.6, 42 mm) have short jaw teeth; in this and in other ways they fit well within the material of M. polyactls (see p. 77). Melanocetus cirrifer Regan and Trewavas (1932), described on the basis of two small females, was tentatively maintained by Bertelsen (1951) because of supposed differences in escal morphol- ogy and pigmentation which now can easily be 76 shown to be part of the variation found within M. johnsoni. Melanocetus megalodontis Beebe and Crane (1947), based on a single specimen, was distinguished from all other species of the genus by ". . . the character of the illicium; in the great length and robustness of the fangs . . . and in the shortness of the lower jaw. . . ." However, speci- mens of M. johnsoni may have longer teeth and individuals of several species of Melanocetus may have as short a lower jaw (Bertelsen 1951, table 4). Further (as predicted by Bertelsen 1951), the "peculiar minute distal flaps" of the esca are arti- facts. In all ways the holotype of M. megalodontis fits well within the variation now known to occur within M. johnsoni. Thus these nominal forms, M. cirrifer and M. megalodontis , are placed within the synonymy of M. johnsoni. Finally, the holotype and paratype of M. rotun- datus Gilchrist (1903) have been lost. The cir- cumstances of their demise are the same as for the holotype of Dolopichthys cornutus described elsewhere (Pietsch 1972b; see also Barnard 1927, Penrith 1967). Although Gilchrist's (1903) origi- nal description is poor, the figure provided by him shows rather long jaw teeth, a long illicium bear- ing a relatively large escal bulb, and a large pec- toral fin lobe. This combination of characters makes it nearly certain that M. rotundatus is a synonym of M. johnsoni (Penrith 1967). PIETSCH and VAN DUZER: SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES Melanocetus polyactis Regan 1925 Figures 21-24, 26, 30 Females Melanocetus polyactis Regan 1925:565 (original description, 3 specimens, lectotype designated by Bertelsen 1951, ZMUC P9260, 61 mm, Dana stn. 1206(3), Gulf of Panama, 6°40' N, 80°47' W, 3,500 m wire, 1845 h, 14 January 1922); Regan 1926:34, pi. 8, fig. 2 (description after Regan 1925); Regan and Trewavas 1932:53, fig. 78 (listed, after Regan 1925, 1926); Bertelsen 1951:44, 54-55, tables 4, 7, 8 (description, addi- tional material, 2 males, 6 larvae, comparison with all known material, in key); Grey 1956:238 (synonymy, distribution). Melanocetus niger Regan 1925:565 (in part, 3 of 11 cotypes [see Comments, p. 79], all Gulf of Panama); Regan 1926:33, pi. 8, fig. 1 (in part, description, 4 additional specimens); Regan and Trewavas 1932:53, fig. 76B (in part, listed, after Regan 1925, 1926); Bertelsen 1951:44, 53, table 4 (in part, description, comparison with all known material, in key); Grey 1956:237 (in part, synonymy, distribution). Melanocetus ferox, Regan and Trewavas 1932:49, 52, fig. 75 (in part, nontype material only, in key); Bertelsen 1951:44, 53, table 4 (in part, nontype material only, comparison with all known material, in key); Grey 1956:237 (in part, after Bertelsen 1951, synonymy, distribution). Males Rhynchoceratias rostratus, Regan 1926:44 (in part, misidentification). Rhynchoceratias leucorhinus, Regan 1926:44 (in part, misidentification). Material. — Metamorphosed females, 15 (16.5-61 mm): BMNH 1925.8.11.30, 26 mm; BMNH 1925.8.11.32, 42 mm (paralectotype); BMNH 1932.5.3.6, 42 mm; lOAN uncatalogued, 33 mm; LACM 33603-4, 2 (16.5 and 30 mm); LACM 33574-5, 17 mm; LACM 33624-1, 33 mm; LACM 33629-3, 35 mm; ZMUC P92155, 25 mm (paralec- totype); ZMUC P921974, 26 mm; ZMUC P9251, 29 mm; ZMUC P92210, 30.5 mm; ZMUC P9253, 47 mm; ZMUC P9260, 61 mm (lectotype). The following adolescent females, all collected from the eastern tropical Pacific, are only tenta- tively referred to M. polyactis: LACM 33618-2, 16 mm; LACM 31119-2, 2 (17 and 18 mm); LACM 31109-2, 18 mm; LACM 31120-20, 2 (18 and 19 mm); LACM 31126-29, 2 (19.5 and 20 mm). Metamorphosed males, 2: ZMUC P92460, 16 mm (22 mm total length (TL)); ZMUC P92459, 19.5 mm (30 mm TL). Larvae, 6 (2 males, 4 females, 3-9 mm TL): ZMUC P92461; ZMUC P92462. Diagnosis. — A species of Melanocetus unique in having the following combination of characters: anterior margin of vomer nearly straight; least outside width between frontals 18.0-26.0% SL; number of lower jaw teeth 58-90 (Figure 22); length of longest lower jaw tooth 9.3-13.1% SL (Figure 21); width of pectoral fin lobe 10.9-16.0% SL; escal bulb width 5.2-8.5% SL (Figure 23); illicium length 34.6-56.0% SL (Figure 24); esca with a conical, distal prolongation, crests absent (Figure 26); integument relatively thick. Description. — Escal bulb not compressed, with conical distal prolongation nearly always slightly constricted at base, and usually as long as or longer than length of escal bulb, pigmented on tip in some specimens, posterior and anterior crests absent (Figure 26); integument as in M.johnsoni. Number of upper jaw teeth 42-120; dorsal fin rays 14-17; pectoral fin rays 17-21 (rarely 22 and 23) (Table 2). Distribution. — Melanocetus polyactis appears to be restricted to the eastern tropical Pacific Ocean where 15 specimens have been collected between lat. 10° N and 13° S as far west as long. 88° W (see Distribution, p. 83). Approximately 67% of the material was captured by open nets fished at max- imum depths of 1,000 m or below. Comments. — Melanocetus polyactis is most easily confused with M. niger. Both forms are similar in having exceptionally short lower jaw teeth (Fig- ure 21). They differ significantly, however, in the number of lower jaw teeth, escal bulb width, and illicial length (see Key, Figures 22-24). Part of the material originally listed as M. niger has been reallocated to M. polyactis (see Com- ments, p. 79). Also included with the material of M. polyactis are two specimens (ZMUC P92210, 30.5 mm; BMNH 1932.5.3.6, 42 mm) previously iden- tified as M. ferox by Regan and Trewavas (1932). 77 FISHERY BULLETIN; VOL. 78, NO. 1 B 1 mm Figure 26.— Escae ofMelanocetuspolyacUs: A. LACM 33603-4, 16.5 mm SL; B. Paralectotype, ZMUC P92155, 25 mm SL (lack of distal pigmentation probably due to abrasion); C. ZMUC P921974, 26 mm SL; D. ZMUC P9251, 29 mm SL (a cotype of M. niger); E. LACM 33603-4, 30 mm SL; F. ZMUC P92210, 30.5 mm SL; G. LACM 33629-3, 35 mm SL; H. Lectotype, ZMUC P9260, 61 mm SL. Melanocetus niger Regan 1925 Figures 21-24, 27, 30 Melanocetus niger Regan 1925:565 (original de- scription, in part, 4 of 1 1 cotypes [see Comments, p. 79] all Gulf of Panama, lectotype hereby des- ignated, ZMUC P9252, 80 mm, Dana stn. 1208(4), 6°48' N, 80°33 ' W, 3,500 m wire, 0810 h, 16 January 1922); Regan 1926:33, pi. 8, fig. 1 (in part, description, 4 additional females); Regan and Trewavas 1932:53, fig. 76B (in part, listed after Regan 1925, 1926); Beebe and Crane 1947:153-154 (in part, description of 4 addi- tional females not seen by us); Bertelsen 1951:44, 53, table 4 (in part, description, com- parison with all known material, in key); Grey 1956:237 (in part, synonymy, distribution). Ma^erm/. — Metamorphosed females, 6 (22-80 mm): BMNH 1925.8.11.29, 47 mm (paralec- totype); lOAN uncatalogued, 77 mm; ZMUC 78 P9254, 22 mm (paralectotype); ZMUC P9256, 37 mm (paralectotype); ZMUC P921973, 42 mm (Galathea stn. 727); ZMUC P9252, 80 mm (lec- totype). Males and larvae unknown. Diagnosis. — A species of Melanocetus unique in having the following combination of characters: anterior margin of vomer nearly straight; least outside width between frontals 14.3-24.3% SL number of lower jaw teeth 37-57 (Figure 22) longest lower jaw tooth 6.9-10.5% SL (Figure 21) width of pectoral fin lobe 9.1-13.5% SL; escal bulb width 3.8-5.0% SL (Figure 23); illicium length 29.8-38.8% SL (Figure 24); esca without crests (Figure 27); integument relatively thick. Description. — Escal bulb not compressed, with a low, rounded or conical distal prolongation nearly always pigmented on tip; anterior and posterior PIETSCH and VAN DUZER: SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES crests absent (Figure 27); integument as in M. johnsoni. Number of upper jaw teeth 29-86; dorsal fin rays 14-15; pectoral fin rays 18-21 (Table 2). Distribution. — All six known specimens of M. niger were collected in the Gulf of Panama and adjacent waters of the eastern tropical Pacific Ocean as far west as approximately long. 90" W (see Distribution, p. 83). Eighty-three percent of the material was captured by open nets fished at maximum depths of 1,500 m and below. Comments. — Melanocetus niger was briefly de- scribed by Regan ( 1925) from seven specimens col- lected in the Gulf of Panama without type desig- nation and without a listing of individual sizes, station numbers, or other means of identification. Regan (1926) added four more specimens without providing means of separating the original seven. All 11 specimens bear labels indicating cotype status and all are treated here as part of the origi- nal type material. One of these is designated the lectotype (ZMUC P9252, 80 mm), three are re- ferred to M. polyactis (BMNH 1925.8.11.30, 26 mm; ZMUC P9251, 29 mm; ZMUC P9253, 47 mm), three unidentifiable specimens are listed below as Melanocetus sp. (ZMUC P9255, 13.5 mm; BMNH 1925.8.11.31, 14 mm; BMNH 1925.8.11.28, 43 mm), and one is unaccounted for and presumed lost [Dana stn. 1209(3), 37 mm total length). The remaining three specimens are recognized as paralectotypes of M. niger. Melanocetus eustalus n. sp. Figures 18, 28, 30 Melanocetus ferox, Pietsch 1972b: 10 (misiden- tification, luminescence); Brewer 1973:25 (after Pietsch 1972b, distribution). Melanocetus sp. Pietsch 1976:782, 783 (reproduc- tion). Material. — A single female, the holotype, LACM 30037-12, 111 mm, Velero IV stn. 11748, eastern Pacific off Mazatlan, Sinaloa, Mexico, 21°39' N, 106°58' W, 3 m IKMT, 0-1,675 m, bottom depth 2,820 m, 1320-2136 h, 11 November 1967. Diagnosis. — A species of Melanocetus unique in having the following combination of characters: anterior margin of vomer nearly straight; least outside width between frontals 18.0% SL; number of lower jaw teeth 60; longest lower jaw tooth 5.9% SL; illicium length 30.6% SL; width of pectoral fin lobe 9.9% SL; escal bulb width 11.3% SL; esca without crests (Figures 18, 28); integument rela- tively thick. Description of holotype. — Escal bulb large (length 14.4% SL), slightly compressed, with a low conical distal prolongation, pigment absent; pos- , 1 mm Figure 27.— Escae of Melanocetus niger: A. Paralectotype, ZMUC P9256, 37 mm SL; B. Paralectotype, BMNH 1925.8.11.29, 47 mm SL; lOAN uncatalogued, 77 mm SL; D. Lectotype, ZMUC P9252, 80 mm SL. 79 FISHERY BULLETIN: VOL. 78, NO. 1 terior and anterior crests absent (Figure 28); integument as in M.johnsoni. Gill opening exceptionally large, greatest diameter 23.47f SL; number of upper jaw teeth 91; vomerine teeth 8; dorsal fin rays 15; pectoral fin rays 16 (Table 2). Etymology. — The name eustalus is derived from the Greek eustales, an adjective meaning well equipped, in reference to the enormous esca of this ceratioid. Luminescence. — Upon capture, the holotype of Melanocetus eustalus was maintained alive for several minutes during which the bulb of the esca glowed continuously with a bright, golden-orange light. The amount of light actually emitted, how- ever, appeared to be controlled by an up and down 80 FIGURE 28.-Holotype of Melanocetus eustalus, LACM 30037-12, 111 mm SL, lateral view. Drawn by Elizabeth Anne Hoxie. PIETSCH and VAN DUZER: SYSTEMATICS AND DISTRIBUTION OF ANGLERFISHES movement of the darkly pigmented, inner wall of the photophore of the esca. The glowing bulb was almost entirely covered and uncovered four or five times within a period of at least 1 min (B. G. Nafpaktitis^; see Pietsch 1972b). A mechanism of this kind would provide a rapid means of extin- guishing light that may not only attract potential mates and prey, but also predators. Melanocetiis murrayi Gunther 1887 Figures 2, 4, 5, 7-15, 16 A, 19, 20, 29-31 Females Melanocetus murrayi Gunther 1887:57, pi. 11, fig. A (original description, two specimens, lec- totype BMNH 1887.12.7.17, 71 mm. Challenger stn. 106, central Atlantic, 1°47' N, 24=26' W, 0-3,386 m); Regan 1926:32 (description, addi- tional material, in key); Parr 1927:27 (descrip- tion, additional material); Regan and Trewavas 1932:27, 49-50, fig. 22C, 23, 71 (description, ad- ditional material; pectoral radials, pelvic bone, escae figured; in key); Beebe 1932:99-102, fig. 29, 30 (description of postlarvae); Parr 1934:7 (listed); Fowler 1936:1143, 1144-1145, 1346, 1363, fig. 483 (after Gunther 1887; Regan 1926; in key); Koefoed 1944:3, 5 (description, compari- son, additional material); Fowler 1949:158 (listed); Bertelsen 1951:40-48, fig. 16, tables 4, 5 (description of females, males, larvae, compari- son with all known material, in key); Grey 1955:299 (additional material, color); Grey 1956:234 (synonymy; distribution); Monod 1960:687, fig. 80 (pectoral radials); Pietsch 1972a:34, 38 (osteological comments); Maul 1973:667 (synonymy, after Bertelsen 1951). Melanocetus bispinosus Gunther 1880:473 (name only); Goode and Bean 1896:495 (in synonymy). Melanocetus (Liocetus) murrayi Gunther 1887:56 (original description, a distinct subgenus). Liocetus murrayi, Goode and Bean 1896:495, fig. 407 (new combination, after Gunther 1887); Gill 1909:583, 584, fig. 22 (after Gunther 1887; Goode and Bean 1896). Melanocetus vorax Brauer 1902:294 (original de- scription, single specimen, holotype ZMHU 17710, 85 mm, Valdiuia stn. 63, Gulf of Guinea, 'B. G. Nafpaktitis, Professor, Department of Biological Sci- ences, University of Southern California, Los Angeles, CA 90007, pers. commun. November 1967. 2°00' N, 8°04' W, 0-2,492 m); Brauer 1906:320- 321, pi. 15, fig. 4 (description after Brauer 1902). Fowler 1936:1143, 1144 (description after Brauer 1902, 1906; in key). Melanocetus johnsoni, Brauer 1906:319 (misiden- tification); Regan 1926:33 (in part, misiden- tification); Murray and Hjort 1912:609, 614, 618, fig. 469 (misidentification); Fowler 1936, fig. 482 (figure after Brauer 1906). Melanocetus krechi, Murray and Hjort 1912:614, 618 (in part, misidentification). Melanocetus tumidus Parr 1927:28-29, fig. 10 (original description, single juvenile, holotype BOC 2022, 15 mm,Pawnee Third Oceanograph- ic Expedition stn. 11, western North Atlantic, 23°58' N, 77°26' W, 2,135 m wire, 2 March 1927); Regan and Trewavas 1932:49 (men- tioned); Grey 1956:239 (synonymy, distribution, a young female M. murrayi). Melanocetus niger, Parr 1927:29 (misidentifica- tion); Beebe 1929:18 (misidentification). Males Rhynchoceratias acanthirostris Parr 1927:31, fig. 11 (original description, single specimen, holotype BOC 2011, 20 mm, Pawnee Third Oceanographic Expedition stn. 22, western North Atlantic, 23°37' N, 77°15' W, 2,135 m wire, 12 March 1927); Parr 1930b: 130, 134 (anatomy, life history). Rhynchoceratias latirhinus Parr 1927:32, 33, fig. 12 (original description, single specimen, holotype BOC 2012, 15 mm, Pawnee Third Oceanographic Expedition stn. 33, western North Atlantic, 24^11' N, 75°37' W, 2,440 m wire, 22 March 1927). Rhynchoceratias longipinnis Parr 1930a:7, fig. 2-5 (original description, single specimen, holotype BOC 2592, 16 mm. Pawnee Third Oceano- graphic Expedition stn. 59, Bermuda, 32=19' N, 64=32' W, 2,440 m wire, 21 April 1927, osteol- ogy); Parr 1930b: 129, fig. 1-3, 6, 7 (anatomy, life history). Xenoceratias acanthirostris , Regan and Trewavas 1932:54, 55 (new combination; description after Parr 1927, in key). Xenoceratias longipinnis, Regan and Trewavas 1932:54, 56 (new combination; description after Parr 1927, in key). Xenoceratias latirhinus, Regan and Trewavas 1932:54, 57 (new combination; description after Parr 1927, in key). 81 FISHERY BULLETIN: VOL. 78, NO. 1 Xenoceratias regani Koefoed 1944:4, 6, pi. 1, fig. 6 (original description, single specimen, holotype UBNM 4311, 20 mm, Michael Sars North At- lantic Deep-Sea Expedition stn. 53, central North Atlantic, 34°59' N, 33°01' W, 2,600 m wire, bottom depth 2,615-2,865 m, 8-9 June 1910). Melanocetus murrayi, Bertelsen 1951:44-48, fig. 16A, D, F, H, table 5 (synonymy, description, comparison with all known material, in key); Grey 1956:235 (synonymy, distribution); Maul 1962b:37-38, fig. 3 (description of additional specimen); Maul 1973:667 (synonymy, after Bertelsen 1951). Material. — Metamorphosed females, 140 (13.5- 120 mm): BMNH, 8 (21-57 mm); BOC, 1 (15 mm); CAS, 3 (14.5-51 mm); FSM, 6 ( 13.5-54 mm); lOAN, 5 (14-56 mm); lOS, 6 ( 17-68 mm); ISH, 33 ( 15-120 mm); LACM, 14 ( 13-84 mm); MCZ, 14 (13-84 mm); UMML, 28 (17-99 mm); USNM, 7 (15-78 mm); VIMS, 1 (33 mm); ZMUC, 14 (14-80 mm). Diagnosis. — A species of Melanocetus unique in having the following combination of characters: anterior margin of vomer deeply concave (Figure 2); least outside width between frontals 9.1-17.8% SL (Figure 19); number of lower jaw teeth 46-142 (Figure 20); longest lower jaw tooth 7.7-16.7% SL; width of pectoral fin lobe 6.1-8.9% SL; escal bulb width 1.9-5.1% SL; illicium length 23.1-37.2% SL; esca with crests minute or absent (Figure 29); mi- nute skin spines restricted to caudal peduncle; integument relatively thin (0.48 mm). Description. — Escal bulb not compressed, with a low, rounded distal prolongation usually unpig- mented on tip; posterior and anterior crests mi- nute or absent (Figure 29); integument thin, easily torn (cross sections measure 0.48 mm in thickness), pigment readily lost during fixation and preservation, often transparent, especially in gill region and over branchiostegal rays. Number of upper jaw teeth 34-178; dorsal fin rays 12-14, pectoral fin rays 15-19 (rarely 20) (Ta- ble 2). Distribution. — Melanocetus murrayi has a wide horizontal distribution in the Atlantic and Pacific, but is apparently absent from the Indian Ocean (see Distribution, p. 83). Compared with M. johnsoni, it is a much deeper living form: only 10% of the material (for which data was available) was captured in open nets fished at maximum depths of 19 mm SL) were measured by timing the fish as they swam a measured distance (35-201 cm) around the perimeter of the rearing tank; in this case mean speeds were for individual fish and observation times ranged from 3 to 26 s. The size at which Pacific mackerel larvae were capable of ingesting various prey was evaluated by placing them in a 110 1 container with the prey and estimating the number that fed by examina- tion of stomach contents. The type of prey, mean prey size, prey density, and duration of feeding respectively were: yolk-sac anchovy larvae, 2.7 mm SL, 8/1, 2 h; A. salina nauplii, 0.2 mm wide, 11/1, 4 h; and anchovy eggs, 0.67 mm wide, 10/1, 2 h. The mouth width, prey width, and standard length of the larvae were measured and percent- age feeding success was estimated for size classes of larval length and mouth width. The number of fish per size class was >9. Size thresholds for 50% feeding success and 95% success were estimated by probit analysis (Finney 1952) and expressed as a function of mean larval length, or mean prey width/mean mouth width. The sizes of food items eaten by Pacific mackerel larvae in the sea was determined by examination of the stomachs of 86 larvae taken in routine ichthyoplankton surveys along the California coast. We recorded the length of each larva and the number and maximum width of all identifiable food items (Arthur 1976; Shirota 1970). Food requirements were estimated by feeding the rotifer Brachionus plicatilis to 3-5 d old Pacific mackerel larvae. Seven to eight samples of 9-16 larvae each were taken over each of three 12-h feeding days, the number of rotifers in the guts of each larva were counted, and the counts converted to equivalent dry weight using the conversion factor of 0.16 /^g/rotifer (Theilacker and McMaster 1971). Daily changes in larval weight were esti- mated from mean standard lengths using a length-dry weight conversion given in the results. The rate of gastric evacuation for 4 mm SL larvae was measured. They were allowed to feed for 4 h and then transferred to a tank without food; sam- ples of 13-16 larvae were taken at about hourly intervals until the stomachs were empty. The number of rotifers in stomachs were counted and converted to dry weight, and the rate of gastric evacuation was estimated in terms of dry weight. The daily ration was estimated from the mean stomach contents and the rate of gastric evacua- tion. Gross growth efficiency was estimated in terms of dry weight from the daily ration and weight gain over 24 h. Metabolic requirements of Pacific mackerel lar- vae were estimated using a Warburg respirometer and standard manometric techniques (Umbreit et al. 1964) to measure oxygen consumption. One or more Pacific mackerel larvae were added to an 18 ml Warburg flask filled with 4.4-8.7 ml of filtered seawater (salinity 33.58-33.93%o). Larvae >0.06 mg dry weight were tested individually. Twenty- one tests were made at 18.0°C of larvae or groups of larvae ranging in length from 3.7 to 17.9 mm SL (0.038-12.74 mg) and 14 at 22.0° C, of larvae 3.2- 10.5 mm SL (0.025-2.86 mg). Flasks were shaken at 102 times/min for 5 out of every 30 min; read- ings were taken after the first 2 h and continued for 150-360 min. At the end of a test, larvae in each flask were measured, rinsed in distilled water, oven dried to a constant weight, and weighed. Mean weight was obtained for fish tested in groups. All runs were made under normal room illumination, about 700 Ix. Logj^ oxygen con- sumption in microliters Og per hour was regressed on logio body weight for the 18.0° and 22.0° C experiments. As the slopes were close to unity, oxygen consumption was expressed in microli- 90 HUNTER and KIMBRELL: EARLY LIFE HISTORY OF PACIFIC MACKEREL ters per milligram per hour. Metabolic require- ments were compared with daily ration by con- verting oxygen consumption and daily ration to calories (1 /i.1 Oj = 0.005 cal: one Brachionus plicatilis ter 1971). 0.00085 cal, Theilacker and McMas- Culture of Larvae Seven groups of Pacific mackerel were reared to metamorphosis to determine growth rates and ef- fects of temperature on growth. The rearing con- tainers were black fiber glass, cylindrical tanks (122 cm in diameter x 36 cm deep). Culture vol- ume increased during the rearing period from 200 to 400 1 because of the addition of seawater con- taining food and algae. Tank temperature was controlled by a regulated water bath, and groups were reared at temperatures ranging from 16.8° to 22.1° C. Illumination at the water surface during the 12-h day was about 2,000 Ix. Tanks were started with 3,000 eggs/group. Initially, larvae were fed laboratory-cultured Brachionus plicatilis. At age 5 d, laboratory-cultured copepodids and adult copepods (Tishe sp.) were added; 200,000 copepods were added daily until metamorphosis. Initially 30 or more rotifers/ml were added for the first few days of feeding; there- after the density of rotifers was allowed to decline. On a diet of rotifers alone, growth slowed after larvae reached 5 mm SL and few larvae survived longer than 15 d. Newly metamorphosed juveniles were fed live and frozen adult A. salina and minced squid {Loligo opalescens ) and northern an- chovy. From 1 to 6 1 of algal culture, Tetraselmis sp. (300,000-500,000 cells/ml), were added daily to provide food for rotifers and copepods. Samples of 10 or more larvae were taken on alternate days for length measurements. Some samples were washed in distilled water, dried, and subsequently weighed to obtain a relation between length and dry weight. RESULTS Hatching, Onset of Feeding and Starvation Eggs of Pacific mackerel are transparent spheres, ranging in diameter from 1.06 to 1.14 mm (Kramer 1960). Incubation times ranged from 33 h at 23° C to 117 h at 14° C (Figure 1); eggs did not hatch below 14° C. The curve for hatching time as a function of temperature for the western Pacific population (data from Watanabe 1970) appears to be the same as the one for the eastern Pacific population. I20|- 110- 100- o 90 < X ^ 80 O in o 70 ^ 60 O X 50 40 H = 3580e-6.50(l-e-00527T) 30^ oV' ' I _L J \ I I I I I I I 12 14 16 18 20 TEMPERATURE °C 22 24 FIGURE 1. — Incubation time (fertilization to 50% hatch) of Scomber japonicus eggs. Solid line is for present data, points are estimated time to 50% hatch of eggs in five test tubes per tem- perature, dashed line is for data of Watanabe ( 1970). The general equation was developed by Zweifel and Lasker (1976) and fit to the 50% values. At hatching, larvae averaged 3.1 mm SL (Fig- ure 2B) and weighed 0.040 mg dry weight, of which 50% was yolk. At 19° C, first feeding oc- curred 46 h after hatching; by 60 h after hatching, all larvae had ingested one or more rotifers in 4 h (Figure 2D). Thus the 50% threshold for onset of feeding at 19° C occurred at about 50 h (2 d) after hatching. At this time larvae were 3.6 mm SL, the eyes were fully pigmented and 10% of the yolk remained, principally the remnants of the oil drop- let (Figure 2A, B). Over the threshold for the onset of feeding, the mean number of rotifers in Pacific mackerel stomachs increased from 2 at 46 h to 14 at 68 h (Figure 2E). The larvae in each group had no previous feeding exposure, hence the increase in feeding activity with time could not be attrib- uted to experience. 91 FISHERY BULLETIN: VOL. 78, NO. 1 -2 0.0' I I I I I I I I I I I ' I I I I I ' I I I 40 80 120 160 200 I 1 r T- I 1^ 2 3 T" 5 7 8 100 T3 O O o - 5 60- > 3 (/) c o Q- O'l I I I I I I I I I I I I ' I ' VT~f I I 40 80 120 160 200 1 1 1 3 T" 4 1^ 7 8 o o e o c/5 e 3 ' M I I I I M I I I I I 40 80 120 160 200 1 1 r 2 3 4 6 8 38 r E E I I I 120 160 200 r ' I I I ' I I I Hours 40 80 1 1 1 1 I f I I I ' 120 160 Days I 1^ 6 T-1 200 -r 8 Figure 2. — Yolk absorption, onset of feeding, starvation, and point of irreversible starvation in Pacific mackerel larvae at 19° C: A. Rate of yolk absorption — open circles mean area of yolk-sac, solid circles mean area of oil droplet (mm^). B. Mean length of larvae from hatching through yolk-absorption. C. Percent survival of larvae without food. D. Percent of larvae tested at various times after hatching that had ingested one or more Brachionus plicatilis in a 4-h test period — arrow indicates the 50% threshold for the onset of first feeding. E. Mean number of S. plicatilis per positive stomach of larvae tested at various times after hatching. F. Percentage survival at age 8 d for larvae fed for the first time at age 2.5, 3.5, 4.5, and 5.5 d — percentages are plotted at the time food was first added. Bars in A, B, and E represent ±2 SE of mean. 92 HUNTER and KIMBRELL: EARLY LIFE HISTORY OF PACIFIC MACKEREL If larvae were not fed, most died between ages 4 and 7 d and none survived longer than 7 d (Figure 2C). Highest survival (on the eighth day after hatching) occurred when food was added for the first time at age 2.5 d; survival was somewhat lower if food was added at 3.5 d and negligible if added at 4.5 d (Figure 2F). Thus at 19' C starva- tion appeared irreversible if food was not provided before 4.5 d. Pacific mackerel larvae, unlike herring or an- chovy (Blaxter and Hempel 1963; Lasker et al. 1970), did not cease swimming or feeding at the time of irreversible starvation. At age 5 d, the incidence of larvae with rotifers in their stomachs was relatively high (80%) (Figure 2D), but the average number of rotifers per positive stomach was much less than in larvae fed first at age 2 or 3 d (Figure 2E). Thus at age 5 d, most larvae were still able to feed, but owing to their weakened condi- tion, none were able to capture enough prey to survive. Vulnerability of larvae to starvation persisted through metamorphosis. All juvenile Pacific mackerel appeared emaciated and swam slowly by the fourth day of starvation. Mortality of 10% oc- curred in the group starved 4 d; 50% mortality occurred in those fish starved 5 d. All juveniles surviving 4 and 5 d of starvation recovered when food was added; no mortality occurred in the con- trols. Thus, newly metamorphosed Pacific mack- erel were able to withstand 1 or 2 d more of starva- tion than first-feeding larvae, but they were better able to recover from food deprivation. Growth Growth in length of Pacific mackerel larvae was slow and almost linear over the first 10-15 d until larvae reached about 6-7 mm SL; there followed a rapid acceleration through metamorphosis. We did not fit equations to these data because none of the standard growth equations gave a good fit to the entire growth curve. The effect of temperature on growth was not distinguishable over the initial growth period, but became obvious during the period of rapid growth (Figure 3, Table 1). To pro- vide an index of the effect of temperature on growth, we expressed the duration of the larvae period (hatching to metamorphosis, 15 mm) as a function of temperature (inset in Figure 3). The Qio was 3.0 when calculated from the equation in Figure 3 for the temperature range of our observa- tions ( 16. 8'-22.1° C). The length-weight relation for Pacific mackerel larvae and juveniles is shown in Figure 4. The form of this equation was developed by James Zweifel and used by Hunter ( 1976) to express the length-weight relation for northern anchovy lar- vae. The curvilinear nature of the length-weight relation, still evident in the log-log plot ( Figure 4), indicates that if a linear regression of logj^ weight Table l. — Growth data (millimeters SL) for seven groups of Scomber japonicus larvae reared at different mean temperatures from hatching through metamorphosis. Age 22.1=0' 20.4= 19.6=0 19.5=0 19.2=0 18.9°0 16.8" (days) n X SD n X SD n X SD n X SD n X SD n X SD n X SD 1 10 3.1 0.24 31 3.5 0.12 2 14 3.3 0.21 16 38 0.06 16 3.7 0.10 15 39 009 15 3.8 006 3 14 3.5 0.30 4 5 10 36 022 18 3.8 026 33 42 0.23 10 4.0 0.17 15 4.0 0.20 10 3.7 0.20 6 10 3.9 0.21 15 4.2 0.41 15 48 0.44 13 4.2 0.14 15 4.4 0.37 10 4.1 0.39 15 4.3 042 7 27 4.5 0.40 8 12 4.5 0.55 15 5.7 0.77 15 6.0 0.39 25 4.7 0.56 15 5.0 0.40 15 5.1 0.43 9 5 5.9 13 10 12 6.0 1 13 15 63 091 15 6,6 070 15 4,8 0,43 19 56 0.77 11 5.2 0.45 11 15 6.5 0.70 12 13 22 8.4 1 88 16 82 1.31 15 6,5 0-75 11 5,9 1.08 25 6.4 0.68 12 6.1 0,60 15 7.0 0.50 14 10 89 1,46 10 11.5 1.54 15 8,4 1,12 30 6,4 1,09 15 6.6 0.80 17 7.7 1.40 15 11 8.1 1.08 16 10 14.9 1.45 15 14.1 2.01 15 8.9 1,76 16 8.5 2,21 15 7.1 081 13 7.6 1.50 17 10 10.0 2.71 12 9.0 1.33 15 10.3 2.11 18 15 17.8 1,70 15 10,3 2,48 15 10.3 3.21 13 9.4 1.61 17 10.8 2,84 19 15 24.1 5.81 20 15 12.5 3.18 17 17.7 4.21 15 11.7 2.88 9 11.5 2.81 22 15 17.5 4.51 15 14.6 462 23 15 13.7 1.26 24 24 20.4 5.10 16 18.5 5.38 10 19.8 2.36 25 17 17.1 5.07 'Juvenile growth (age. n.x, and (SD)): 26 d. 13,34.4 mm (4.37); 29 d, 9, 43.9 mm (3.60); 39 d. 5, 55.0 mm (8.74); and 47 d. 9. 67.3 mm (10.30). 93 FISHERY BULLETIN: VOL. 78, NO. 1 24 22 20 18 16 X 14 1- o z 1? Ixl _J _l 10 < > tr R < _j 6 4 - < t- - O >- < o 22.1 E E If) O I Q. cr o 24 22 20 \ .\ y • - \ • - \ • __ • V — D = 51186- 1 593 T \ - r2 = 706 • \ — 1 1 1 1 1 1 f 15 16 17 18 19 20 21 TEMPERATURE CO TEMPERATURE (°C) 22.1 +.2 20. 4 + . 5 19.6 +.4 19 . 5 + . 8 19.2 + .1 18-9 + .1 16.8 +.2 8 10 12 14 16 18 20 AGE (DAYS) 22 24 26 Figure 3. — Growth of seven groups of Pacific mackerel larvae reared in the laboratory from hatching (age d) through metamorphosis (15 mm SL). Lines connect means given in Table 1 ; rearing temperatures ( ±2 SE) given on right side of figure and at end of lines. Inset at top: elapsed time (days) from hatching to metamorphosis (15 mm), as a function of rearing temperature. on logjj, length were used, it would produce inac- curate estimates. Swimming Behavior At typical cruising speeds, larval Pacific mack- erel (3-5 mm SL) have a high tail beat frequency of about 30 beats/s and a low tail beat amplitude of 0.16 standard length. At slow speeds, tail beat frequency remained relatively constant but the amplitude of the tail beat changed. At higher speeds, both amplitude and frequency changed but the relative increase in amplitude was much greater than that of frequency (Table 2). Thus larval Pacific mackerel, unlike the adults (Hunter and Zweifel 1971), predominantly modulate tail beat amplitude to effect changes in speed. Cruising speeds of Pacific mackerel increased markedly over the larval period from 0.46 cm/s (1.3 standard body lengths/s) for first-feeding lar- vae (3.6 mm SL) to 5.6 cm/s (3.8 standard body lengths/s) for fish at metamorphosis (Figure 5). This differs from the pattern in adult fishes where speed relative to size decreases with an increase in fish size (Webb 1975). Feeding Behavior Upon sighting a prey (rotifer or copepod), a Pa- cific mackerel larva advanced toward the prey. 94 HUNTER and KIMBRELL: EARLY LIFE HISTORY OF PACIFIC MACKEREL 2000 - 1000 r 500 100 50 ^ 10 £ X C2 UJ >- cr Q 05 01 0.05 001 Ln W= -4.660 + (-672l + 6.333 Ln l)°-S'° I I h I I I 2 4 6 8 10 20 40 6080100 STANDARD LENGTH (mm) Figure 4. — Relation between dry weight (W) of larval and juvenile Pacific mackerel in milligrams and standard length (L) in millimeters. Points are observed values for individuals >18 mm and for larvae <18 mm; points are means for groups of 15 larvae. stopped, drew back the tail, and held it in a slightly recurved, high amplitude position while the rest of the body remained relatively straight. Table 2. — Tail beat frequency and amplitude and speed of 3-5 mm larval Pacific mackerel ix = 4.23 ±0.09 mm SL) expressed as a function of standard length. Tail beat Swimming speed (SUs) Class interval 18 30 9 3 3 1 0.01- 1.01- 2.01- 3.1 10.1 1.0 2.0 3.0 5.0 15.0 Frequency Mean (beats/s) Amplitude/SL 0.58 33.2 0.12 1.48 30.2 0.16 2.48 30.2 0.17 4.00 37.5 0.18 12.18 39.4 0.29 37.04 38.9 0.33 E o Q UJ UJ Q. Ul CO 30 20 - S = 2.780L r2 = 0.966 753 / 10 _ 4 / 8.0 - / 6.0 - J 4.0 - • 7 I 2.0 - i / 1.0 — I 1 0.8 - *f 0.6 - ? 0.4 - * 111! Mil .3 .4 .5.6.7.8 1.0 2.0 3.0 STANDARD LENGTH (cm) Figure 5. — Relation between swimming speed and standard length of Pacific mackerel larvae (log^g scales) at 19° C. Each point <2.0 cm is the mean of 15-25 observations. For ^2.0 cm, individual fish were measured. Feeding involved driving the tail posteriorly and opening the mouth. Larvae often attacked the same prey two or more times if the previous strike was unsuccessful, and repositioned for subsequent strikes by moving backward. Handling times were negligible because the prey was engulfed instan- taneously. Older Pacific mackerel larvae de- veloped a set of motor patterns for feeding on fish larvae; larvae were seized from the side and car- ried crosswise in the mouth. Larger prey were repeatedly released and grasped until they ceased 95 FISHERY BULLETIN: VOL. 78, NO. 1 struggling, then released and ingested, usually head first. Handling times increased with prey size. The length at which 50% of Pacific mackerel larvae were capable of capturing and ingesting anchovy yolk-sac larvae (LD50, Finney 1952) was 8.1 mm SL (95% confidence interval, 7.2-9.5 mm) (Figure 6). Sibling cannibalism began when the mean length of the group was about 8 mm SL. At this size, the mean length of six cannibals was 10.8 mm SL (range 9.9-12.0 mm) and that of their prey was 6.2 mm SL (range 5.9-6.5 mm). Cannibalism in rearing containers ended as Pacific mackerel approached metamorphosis (15 mm SL) and schooling began. Rearing at higher temperatures increased the growth rate and thereby decreased the period over which sibling cannibalism oc- curred. Consequently, survival at metamorphosis was higher in groups reared at 20°-22° C (5-6%) than it was at 19° C or lower temperatures (1-2% ). Near metamorphosis. Pacific mackerel were able to eat relatively large fish larvae. Three Pacific mackerel, 15.4-16.0 mm SL, placed in a rearing tank with northern anchovy larvae (12.0-20.6 mm SL) captured and began to ingest larvae of 11.7-13.5 mm SL, within 6 min. Thus, as Pacific mackerel larvae grew from 8 mm SL to metamorphosis, the size of anchovy larvae, they were able to eat increased from about 3 to 13 mm SL. This increase in prey size was not closely re- lated to mouth size of the Pacific mackerel because the mouth can be greatly expanded when ingest- ing a larval fish. Mouth size probably was in- versely related to handling time as in the case for adult fishes (Kislalioglu and Gibson 1976). When prey are engulfed rather than seized, mouth size may give a good indication of the size of prey a larvae is capable of ingesting. The relation between mouth width and length in Pacific mack- erel larvae was slightly curvilinear, and mouth width increased from 0.216 mm for first-feeding larvae (3.6 mm SL) to 0.987 mm at metamorphosis (15mmSL) (Figure 7). l-4i- _ LnM= -49419 + (15755 + 5.4815 LnL) E X I- Q 3 O 8 - 6 - W/- J \ \ \ \ L J \ \ I \ I I I \ I 18 20 6 8 10 12 14 16 STANDARD LENGTH {mm) en to UJ u o CO UJ CD < UJ O PREY = YOLK-SAC ANCHOVY LARVAE - S=-2976 + 87ll Log|Q L y^ - S = 7o IN PROBITS - yo - 1 1 1 1 1 1 1 1 1 1 1 1 1 95 90 80 70 60 50 40 30 20 10 6 7 8 9 10 II 12 13 STANDARD LENGTH (mm) Figure 6.— Percentage of Pacific mackerel larvae (probit scale) that captured one or more yolk-sac anchovy larvae in relation to standard length of the mackerel (log,j, scale). The length class was variable. Larvae were ranked by length and classes set at 10 observation intervals. The LD^^^ was 8.1 mm (95% confidence interval 7.2-9.5 mm). Figure 7. — Mouth width as a function of standard length of Pacific mackerel larvae. Points represent single larva. The threshold, in terms of length for feeding on A. salina nauplii, was distinctly different from that for feeding on anchovy eggs. The 50% threshold for nauplii was 4.5 mm SL (95% con- fidence interval, 4.1-4.8 mm) and that for eggs was 12.2 mm SL (11.3-13.1 mm). This could be ex- pected because anchovy eggs are nearly three times as large as A. salina nauplii. On the other hand, when feeding success was expressed as a function of the ratio, mean prey width/mean mouth width, the percentage feeding success of Pacific mackerel fed A. salina was similar to that of larvae fed eggs (Figure 8). At first feeding, rela- tive prey size (prey width/mouth width) was near unity for larvae fed either A. salina or eggs, indi- cating the width of the mouth established the upper size limit of prey. Since the 50% threshold for relative prey size for the combined data given in Figure 8 was 0.85 (95% confidence interval. 96 HUNTER and KIMBRELL: EARLY LIFE HISTORY OF PACIFIC MACKEREL 95 90 - 80 V. 70 UJ 8 60 ^ 50 UJ ^ 40 I- g 30 ^ 20 • • PREY O Enq raulis mordox eggs (minor axis = 0674 mm) # Artemio solino nauplii (0238 mm) 5 = 4 322-9580 Log S= % in probits 10 _L _L 4 5 6 7 8 9 1.0 1.5 PREY WIDTH /MOUTH WIDTH (R) Figure 8. — Relation between average feeding success (probit scale) and average relative prey size (prey width/mouth width), for larval groups fed Artemia salina nauplii (closed circles) and northern anchovy eggs (open circles). Each point is the percent- age success of fish within a mouth width/prey size class, where n>9. Line is the regression of percentage success probit on log^^ of the prey-width to mouth-width ratio, for A. salina nauplii and anchovy egg data combined. The LD^p for the combined data was 0.85 (95<7c confidence interval 0.79-0.91). 0.79-0.91) and the 95% threshhold was 0.57 (0.47-0.70), nearly all Pacific mackerel larvae (95% ) were able to ingest a prey when it was 57% of the width of the mouth and 50% were able to do so when it was 85% of the mouth width. Nearly all prey eaten by Pacific mackerel larvae in the sea fell within the range of sizes predicted from the laboratory work; few prey exceeded the width of the mouth (Figure 9). Fifty-nine percent of all identifiable food items in the stomachs of sea-caught larvae were stages of copepods; other items included cladocerans, oikopleurans, gas- tropods, invertebrate eggs, diatoms, fecal pellets, and one fish larvae. Although laboratory data indicated that 50% of Pacific mackerel larvae were able to ingest prey having a width of 85% of the mouth width, the mean diameter of prey eaten in the sea was 38±2% (2 SE) of the mouth width. Thus, a sub- stantial number of prey eaten by larvae in the sea was much smaller than the maximum size of prey they were capable of ingesting. This may reflect a shortage of larger prey in the sea. Larger prey probably are important nutritionally. If one as- sumes the prey given in Figure 8 to be spherical, then 50% of the prey items accounted for about 85-90% of the total volume of food, depending on larval size. Conversely, the small prey items that contributed 50% by number, contributed only 10-15% of the total volume of prey eaten. This calculation underestimates the volume of the larger prey because they are more elongate or less spherical than smaller ones. Nevertheless, it indi- cates that prey less than the mean size eaten con- tributed relatively little nutritionally to the diet of Pacific mackerel larvae, and that the relatively large, but more rare, prey probably made the major contribution to growth. Ration, Growth Efficiency and Metabolism Pacific mackerel larvae (age 3-5 d) fed actively throughout the day; the gut was filled within the first hour of feeding and it remained full through- out the remainder of the 12-h feeding day, despite a high rate of gastric evacuation. Evacuated Brachionus plicatilis were well digested; only the lorica remained after digestion. Our measure- ments of evacuation rates indicated that about half the gut contents was evacuated in 2 h (Figure 10). Growth of larvae used for ration estimates was about the same as that for other groups reared at 19° C (Figure 3). To grow at this rate in the laboratory, Pacific mackerel larvae (age 3-5 d) consumed an average of about 87% of their dry body weight per day, or about 165-538 rotifers/day (Table 3). This estimate of ration was based on the dry weight of the mean number of rotifers in stomachs, adjusted for the rate of evacuation (Stauffer 1973). The mean gross growth efficiency in dry weight was 33%, which falls within the range of estimates for fish larvae and young fishes (Pandian 1967; Stepien 1976). Our respiration experiments indicated that Pacific mackerel larvae at 18.0° C consumed 6.1 ± 1.4 (2 SE) Ml Oa/mg per h (n = 24) and at 22.0° C they consumed 11.4±3.0 ^tl Og/mg per h (n = 14). By interpolation, the rate at 19° C, the tem- perature of the ration experiments, is estimated as 7.4 ^x\ Oj/mg per h. This metabolic expenditure, converted to calories per day (footnote 6, Table 3) was, on the average, about 18% of the mean daily ration for larvae given in the table. This is proba- bly an underestimate of their metabolic require- ment because the activity of larvae confined in 97 .8 .7 £.6 E Q O O .5 O X Q .4 .3 .2 FISHERY BULLETIN: VOL. 78, NO. 1 / / / / / / / / / / / / •• •• • W •• ..4/ / / / oVv / / / • / / / / /-/ / / / • • / 'A. . ... 4PX fi.0. - - • - •4, #9" _4 _ll mm % »^ m A • • «• .^ . Mi w • ^ lb* A" • « 1^ • « • • • m ^o: A A y *^S4* 49 10 12 13 14 15 16 17 STANDARD LENGTH (mm) Figure 9.— Width of foods eaten in the sea by Pacific mackerel larvae of various standard lengths. Each small point is the width of a single prey; larger points represent multiple points for prey of the same size and number observations. Dashed lines indicate the prey width equal to 20-80% of the mouth width , or equal to the mouth width ( 100% ) , for Pacific mackerel larvae of 3-16 mm (calculated from data given in Figure 7). 98 HUNTER and KIMBRELL: EARLY LIFE mSTORY OF PACIFIC MACKEREL 005- LnW = -5 442-0 337 T r^ =0 970 e OD I ^ 0005 - 2 < liJ 5 001 - .000 2 3 4 5 ELAPSED TIME (h) FIGURE 10.— Rate of gastric evacuation of 4.01+0.03 mm SL Pacific mackerel larvae fed Brachionus plicatilis. Each point represents the mean dry weight of fi. plicatilis in guts of 13-16 larvae. Dry weight estimated by counting numbers of B. plicatilis in stomachs and multiplying by the mean dry weight of one B. plicatilis iO.lSfig) iTheilacker and McMaster 1971). Warburg flasks was probably less than that of free-swimming larvae. These respiration mea- surements do establish a lower limit to food ration, because the ration would have to exceed the metabolic requirement just to meet maintenance costs. DISCUSSION The characteristics of the embryonic period (du- ration of incubation and yolk-sac periods, extent of yolk reserves, size at first feeding, and ability to withstand starvation) were similar to other tem- perate fishes with small pelagic eggs (Lasker et al. 1970; Zweifel and Lasker 1976) and did not differ greatly from some subtropical species (Houde 1974). Small differences in these characteristics may be of importance (Houde 1974) but growth, metabolism, feeding, and swimming behavior are of more value in characterizing the early life his- tory of Pacific mackerel. Pacific mackerel larvae grew rapidly, complet- ing metamorphosis (15 mm SL) in 2-3 wk. Fast growth appears to be characteristic of scombroid larvae and is even more rapid in tropical scom- broids: Auxis thazard grew to 64 mm SL in 17 d (Harada, Murata, and Furutani 1973) and A. tapeinosoma grew to 49 mm SL in 18 d (Harada, Murata, and Miyashita 1973). Fast growth re- quires a large food ration; we found that Pacific mackerel larvae consumed about 877f of their dry Table 3. — Estimate of ration, metabolism, and growth efficiency of 3-5 d old Pacific mackerel larvae fed Brachionus plicatilis. Tempera- ture V 0) Larval SL±2SE(mm) Experimental conditions Larval weight' Food density (no. /ml) Mean No. samp weight in stomachs^ Larval age(d) On day of ration estimation (;ug) Gain 1 d after estimation (^g) of xr2SE es= (^g) 3 4 5 X 18.7 19.0 19.4 19.0 3.56-0.03 3.76±0.02 4.38±0.08 3.93 37.8 43.0 84.6 55.1 5.2 14.0 37.5 18.9 157 47 198 134 7 8 7 4.8-0.8 6.9±1.0 15.6-4.5 9.1 Ration" Ration, metabolism and growth efficiencies Metabolic rate^ (cal/d) Weight gain' (cal) Gros Larval age (d) /ug/d Percent body welghl'd 5cal;d ;s growth efficiency^ (percentage) 3 4 5 X 26.5 38.1 86.2 50.3 70 89 102 87 0.141 0.203 0.460 0.268 0.0338 0.0384 0.0756 0.0493 0.026 0070 0.188 0.094 20 37 44 33 'Calculated from mean larval length using relation given in Figure 4. Mean counts of 8. plicatilis in stomach converted to weight using one B. plicatilis = 0.16 ^g (Theilacker and McMaster 1971). Each sample consisted of 13-16 larvae; sampling began after first hour of feeding. Ration = (/•■/(• f) - r, where r is mean stomach contents, k is rate of gastric evacuation (0.377), and t is duration of feeding period (12 h). (From G. Stauffer, unpubl. manuscr. Southwest Fisheries Center. La Jolla. Calif ) Caloric value of B. plicatilis = 5.335 cal/g (Theilacker and McMaster 1971). Maintenance requirement from; 7.45 /nl 02,mg per h; 1 ^1 O2 = 0.005 cal; time = 24 h; and dry weight of larvae on day ration estimated. Caloric value of weight gained assumed to equal 5.000 cal/g. Gross efficiency (dry weight) = weight gam/ration. 99 FISHERY BULLETIN: VOL. 78, NO. 1 weight per day and weight increased from 0.034 mg to 7.5 mg over the larval period. To capture sufficient numbers of prey to support such rapid grovvi;h requires that the size of the prey and the size of the mouth increase rapidly. Our analysis of sea-caught Pacific mackerel larvae showed that the maximum size of prey did increase rapidly, more or less, in proportion to mouth size. The mean and minimum size of prey eaten by Pacific mackerel changed more slowly but the smaller prey, those less than the average size, may consti- tute <15% of the volume of food eaten. A similar pattern of rapidly increasing prey size with length also has been documented for Scomber japonicus larvae by Shirota (1970) and Yokota et al. (1961). A dependency on larger prey and fast growth requires faster swimming to increase the volume of water searched for prey because abundance de- clines with increased prey size (Sheldon et al. 1972). The swimming behavior of Pacific mackerel larvae appeared consistent with this argument. Cruising speeds increased rapidly with length, roughly to the 1.8 power, and speeds of the larger larvae were at the upper end of the range, typical of larval fishes (3 SL/s) (Blaxter 1969). Higher speeds require a greater metabolic expenditure. The rate of oxygen consumption for Pacific mack- erel (6-11 fx\ Og/mg per h) was above that for other marine fish larvae (Blaxter 1969) indicating a higher-than-average metabolic expenditure de- spite the fact that the rates probably do not reflect the entire cost of high speed swimming. Piscivorous feeding was an import".:.;. e- havioral trait in the early life history of Pacific mackerel because larvae were no longer limited to prey sizes equal to or less than the size of an open mouth. In piscivorous feeding, prey were seized, manipulated and the mouth greatly expanded during ingestion, permitting consumption of much larger diameter foods. In our samples of sea-caught larvae, only one stomach contained a larval fish, but the actual incidence may be higher because larvae are digested rapidly. Cannibalism, a correlate of piscivorous feeding, was common in laboratory groups after the larvae reached 8 mm SL. This also has been observed from stomach contents of the Atlantic mackerel, S. scombrus (Lett 1978). Cannibalism appears to be a common feature of scombroid life history; Mayo (1973) re- marked that Euthynnus alletteratus , Scom- beromorus cavalla, S. regalis, and Auxis sp. be- came cannibalistic at about 5 mm SL. He also noted that cannibalism ceased as the fish became 100 juveniles which agrees with our observation that cannibalism ended as Pacific mackerel ap- proached metamorphosis and began to school. The extent that cannibalism affected the form of our laboratory growth curves is unknown. Although cannibalism was high in all groups, survival was higher in groups reared at high temperatures be- cause of the faster growth rate, which meant faster transit through cannibalistic sizes. In summary, traits that characterize the early life history of Pacific mackerel are the interrelated characteristics of fast growth, fast swimming, high metabolism, a dependence on increasingly larger prey, and cannibalism. The high food re- quirements of the larvae, and the fact that in the sea they feed upon many prey substantially small- er than they are capable of eating, indicates that growth or survival in the sea might be limited by the availability of larger prey. LITERATURE CITED Arthur, D. K. 1976. Food and feeding of larvae of three fishes occurring in the California Current, Sardinops sagax, Engraulis mordax, and Trachurus symmetricus. Fish. Bull., U.S. 74:517-530. Blaxter, J. H. S. 1969. Development: eggs and larvae. In W. S. Hoar and D.J.Randall (editors), Fish physiology. Vol. 3, p. 177-252. Acad. Press, N.Y. Blaxter, J. H. S., and G. Hempel. 1963. The influence of egg size on herring larvae. J. Cons. 28:211-240. Finney, D. J. 1952. Probit analysis: a statistical treatment of the sig- moid response curve. Univ. Press, Camb., 318 p. Harada, T., O. Murata, and H. FURUTANI. 1973. On the artificial fertilization and rearing of larvae in Marusoda, Auxis tapeinosoma. [In Jpn., Engl, abstr.] J. Fac. Agric, Kinki Univ. 6:113-116. Harada, T., O. Murata, and S. Miyashita. 1973. On the artificial fertilization and rearing of larvae in Hirasoda, Aitxjs thazard. [In Jpn., Engl, abstr.] J. Fac. Agric, Kinki Univ. 6:109-112. HOUDE, E. D. 1974. Effects of temperature and delayed feeding on growth and survival of larvae of three species of subtropi- cal marine fishes. Mar. Biol. (Berl.) 26:271-285. Hunter, J. R. 1972. Swimming and feeding behavior of larval anchovy Engraulis mordax. Fish. Bull., U.S. 70:821-838. 1976. Culture and growth of northern anchovy, Engraulis mordax, larvae. Fish. Bull., U.S. 74:81-88. Hunter, J. R., and j. r. Zweifel. 1971. Swimming speed, tail beat frequency, tail beat amplitude, and size in jack mackerel, Trachurus symmet- ricus, and other fishes. Fish. Bull., U.S. 69:253-266. Kislalioglu, M., and r. N. Gibson. 1976. Prey 'handling time' and its importance in food HUNTER and KIMBRELL: EARLY LIFE HISTORY OF PACIFIC MACKEREL selection by the 15-spined stickleback, Spinachia spinachia (L.) J. Exp. Mar. Biol. Ecol. 25:151-158. KRAMER, D. 1960. Development of eggs and larvae of Pacific mackerel and distribution and abundance of larvae 1952- 1956. U.S. Fish. Wildl. Serv., Fish. Bull. 60:393-438. Kramer, D., and p. E. Smith. 1970. Seasonal and geographic characteristics of fishery resources, California Current region — IV. Pacific mack- eral. Commer. Fish. Rev. 32(10:47-49. LASKER, R. 1964. An experimental study of the effect of temperature on the incubation time, development, and growth of Pacific SEirdine embryos and larvae. Copeia 1964:399- 405. 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. 1977. Maturation and induced spawning of captive Pacific mackerel, Scomber japonicus. Fish. Bull., U.S. 75:205- 211. LETT, P. F. 1978. A comparative study of the recruitment mechanisms of cod and mackerel, their interaction and its implication for dual stock management. Ph.D. Thesis, Dalhousie Univ., Halifax, 125 p. MAYO, C. A. 1973. Rearing, growth, and development of the eggs and larvae of seven scombrid fishes from the Straits of Flori- da. Ph.D. Thesis, Univ. Miami, 127 p. Pandian, T. J. 1967. Intake, digestion, absorption, and conversion of food in the fishes Megalops cyprinoides and Ophiocephalus striatus. Mar. Biol. (Berl.) 1:16-32. SHELDON, R. W., A. PRAKASH, AND W. H. SUTCLIFFE, jR. 1972. The size distribution of particles in the ocean. Limnol. Oceanogr. 17:327-340. Shirota, a. 1970. Studies on the mouth size of fish larvae. [In Jpn., Engl, summ.] Bull. Jpn. Soc. Sci. Fish. 36:353-368. (Transl. by Fish. Res. Board Can., Transl. Ser. 1978.) Stauffer, G. 1973. A growth model for salmonids reared in hatchery environments. Ph.D. Thesis, Univ. Washington, Seat- tle, 213 p. Stepien, W. P., JR. 1976. Feeding of laboratory-reared larvae of the sea bream Archosargus rhomboidalis (Sparidae). Mar. Biol. (Berl.) 38:1-16. THEILACKER, G. H., AND M. F. MCMASTER. 1971. Mass culture of the rotifer Brachionus plicatilis and its evaluation as a food for larval anchovies. Mar. Biol. (Berl.) 10:183-188. Umbreit, w. w., r. h. Burris, and j. f. stauffer. 1964. Manometric techniques. A manual describing methods applicable to the study of tissue metabolism. Burgess, Minneap., 305 p. Watanabe, T. 1970. Morphology and ecology of early stages of life in Japanese common mackerel. Scomber japonicus Hout- tuyn, with special reference to fluctuation of popula- tion. [InJpn.,Engl. abstr.] Bull. Tokai Reg. Fish. Res. Lab. 62:1-283. WEBB, P. W. 1975. Hydrodynamics and energetics of fish propul- sion. Fish. Res. Board Can., Bull. 190, 158 p. wolfson, f. h. 1965. The optical comparator as a tool in plankton re- search. Limnol. Oceanogr. 10:156-157. YOKOTA, T., M. TORIYAMA, F. KANAI, AND S. NOMURA. 1961. Studies on the feeding habit of fishes. [In Jpn., Engl, summ.] Rep. Nankai Reg. Fish. Res. Lab. 14, 234 p. ZWEIFEL, J. R., AND R. LASKER. 1976. Prehatch and posthatch grow^th of fishes — a general model. Fish. Bull., U.S. 74:609-621. 101 SPAWNING AND FECUNDITY OF ATLANTIC MACKEREL, SCOMBER SCOMBRUS, IN THE MIDDLE ATLANTIC BIGHT Wallace W. Morse ' ABSTRACT Collections of Atlantic mackerel, Scomfcerscomirus, were made during spring 1977 from Maryland to Rhode Island. Length-weight relationships were determined for total and fork lengths and total and gutted weights. Spawning time was determined from gonad somatic indices and peak spawning occurred between 21 April and 4 May. Egg diameter frequencies from running ripe ovaries indicated five to seven egg batches are spawned by each female during the spawning season. Fecundity was estimated and ranged from 285,000 to 1,980,000 for fish between 307 and 438 mm fork length. Fecundity was related to fork length, gutted weight, and age. The Atlantic mackerel, Scomber scombrus Lin- naeus, is a schooling, pelagic species ranging from the Gulf of St. Lawrence to North Carolina in the northwest Atlantic and from Norway to Spain in the northeast Atlantic. The northwest Atlantic population has been separated into northern and southern contingents on the basis of size composi- tion, spawning times, summer distributions, and tagging studies (Sette 1950; Moores et al. 1975; MacKay2).The northern contingent spawns in the southern Gulf of St. Lawrence from about the end of May to mid-August (Ware 1977). The southern contingent spawns from mid-April to June from North Carolina to Massachusetts (Berrien 1978). Fecundity estimates of northwest Atlantic mackerel are limited to a few observations rang- ing from about 500,000 to 1,000,000 eggs (Brice 1898: 208-213; Sette 1943). Fecundity of northeast Atlantic mackerel ranged from approximately 130,000 to 1,100,000 eggs for fish 28.5-46.0 cm total length (Macer^; Lockwood'*). This paper pre- sents the results of a fecundity and spawning time investigation of the southern contingent. METHODS Atlantic mackerel were collected between 9 April and 21 May 1977 from recreational and 'Northeast Fisheries Center Sandy Hook Laboratory, Na- tional Marine Fisheries Service, NOAA, Highlands, NJ 07732. ^MacKay, K, T. 1973. Aspects of the biology of Atlantic mackerel in ICNAF Subarea 4. Int. Comm. Northwest Atl. Fish., Res. Doc. 73/70, 11 p. ^Macer, C. T. 1976. Observations on the maturity and fecundity of mackerel {Scomber scombrus L.) Int. Counc. Explor. Sea, CM 1976/H:6, 7 p. ■•Lockwood, S. J. 1978. The fecundity of mackerel. Scomber scombrus L. Int. Counc. Explor. Sea, CM 1978/H:9, 5 p. commercial catches from Maryland to Rhode Is- land (Table 1). Length frequencies of males and females are shown in Figure 1. All fish were mea- sured to the nearest millimeter fork length (FL) and total length (TL), and weighed to the nearest gram total weight (TW) and gutted or somatic weight (GW). Otoliths were extracted for age de- termination. Ovaries of all mature females were exsected, weighed to the nearest 0.01 g, and pre- served in 10% Formalin.^ Preliminary observations of eggs from ovaries in the spawning condition revealed that three egg types were present: 1) small, translucent eggs; 2) ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Table l. — Catch data of Atlantic mackerel sampled in 1977. NumI bers of fish examined Capture Date Port Female Male method 9 April Ocean City, Md. 16 8 Otter trawl 16 Cape May. N.J. 15 10 Hook and line 20 Ocean City 26 26 Otter Trawl 25 Barnegat, N.J. 20 25 Hook and line 27 Greenport. N.Y. 64 36 Pound net 28 Belford, N.J. 10 15 Otter trawl 28 Sheepshead Bay, N.Y. 6 16 Hook and line 30 Sheepshead Bay 9 12 Hook and line 1 May Sheepshead Bay 6 17 Hook and line 4 Barnegat 16 36 Hook and line 5 Barnegat 39 66 Hook and line 7 Sheepshead Bay 11 19 Hook and line 8 Point Pleasant. N.J. 17 8 Hook and line 9 Point Judith. R.I 42 43 Otter trawl 11 Belmar, N.J. 5 20 Hook and line 14 Sheepshead Bay 12 38 Hook and line 15 Sheepshead Bay 4 33 Hook and line 17 Sandy Hook, NJ. 22 20 Hook and line 18 Point Judith 32 48 Otter trawl 22 Sheepshead Bay 77 21 Hook and line Totals 449 517 Manuscript accepted August 1979 FISHERV BULLETIN: VOL 103 78, NO. 1, 1980. FISHERY BULLETIN: VOL. 78, NO. 1 z UJ Table 2. — Maturity stages of Atlantic mackerel ovaries. 30 Figure l. 32 34 36 38 40 42 44 FORK LENGTH 10.5 cm midpoints I -Length frequencies of male and female Atlantic mackerel used in this study. larger opaque, yolked eggs; and 3) large translu- cent eggs. There appeared to be no clear size sep- aration between egg types, which is indicative of serial spawners (Hickling and Rutenberg 1936). Therefore, the method described by Hislop and Hall (1974) for whiting, Merlangus merlangus, was used to determine which eggs would be shed during the current spawning season. Since yolk deposition indicates eggs are ripening for spawn- ing, random samples of 300 eggs were measured from ovaries at successive maturity stages to de- termine the average minimum size of yolked eggs. Eggs 0.20 mm and larger contained yolk and were included for fecundity estimation. Ovaries were classified into four maturity stages based upon macroscopic examination and the occurrence of mature eggs (Table 2). Egg diameter frequencies of yolked eggs from ovaries in the developing, ripe, running ripe, and partially spent condition are shown in Figure 2. Ovaries in the ripe condition (Figure 2b) were used for fecundity estimations. If large translu- cent eggs (1.00-1.35 mm) were present in the lumen of the ovary, which is indicative of the run- ning ripe condition (Figure 2c), the ovary was not utilized for fecundity because some eggs may have been shed and fecundity would be underestimated. Stage Description 1 . Developing 2. Ripe 3. Running ripe 4. Partially spent Ovary enlarged, usually orange colored with a granular appearance No translucent eggs, maximum egg diameters 0,8-0.9 mm. Ovary fills most of gut cavity, yellow colored, in advanced stage some translucent eggs are visible ttirough wall. Maximum egg diameter 1 0-1.2 mm. Similar in appearance to stage 2, eggs are ex- truded with pressure on abdomen of fish, t^aximum egg diameters 1.2-1.4 mm. Ovary is flaccid, often hemorrhaging is evi- dent at anterior portion of ovary, some residual mature eggs (1.1-1.4 mm) present. ^^^ 25 20 15 10- 1 InJ IL N . nrJI 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 12 1.3 1.4 EGG DIAMETER I mm I Figure 2. — Egg diameter frequencies of Atlantic mackerel ova- ries in stages: developing (a), ripe (b), running ripe (c), and par- tially spent (d). Each graph based on 300 egg measurements. Suitable ovaries were removed from the Formalin solution, placed in glacial acetic acid for 5 min, and washed, and the eggs were separated with a gentle 104 MORSE: SPAWNING AND FECUNDITY OF ATLANTIC MACKEREL stream of water and agitated in a 0.20 mm mesh sieve. Following removal of the ovarian tissue, the eggs were air dried on blotter paper for 2-3 min and weighed (±0.01 g), and two subsamples were removed and weighed (±0.01 mg). All eggs in each subsample were counted and the mean used to calculate total egg numbers based on the weight of all eggs in the ovary. If the two subsample counts differed by 10*^ or more, additional samples were taken until two counts differed by <10%. Ages were determined from otoliths as de- scribed by Steven (1952). RESULTS The allometric relationships of length-weight were expressed by the power function: Y =aX^ (1) where X is length, Y is weight, and a and h are constants. Equation (1) was converted to the linear form by a logarithmic (base 10) transforma- tion to: log Y = log a + 61og X (2) The interrelationships between length mea- surements and between weight measurements were expressed by the linear function: Y = a + bX (3) where Y andX are both length or both weight. All data were fitted using least-squares regression techniques. Predictive regression equations were calculated using all observations for males and females and an analysis of covariance applied to determine possible sex related differences. No significant dif- ferences iP = 0.05) were indicated between sexes and sexes were therefore pooled. The pooled re- gression equations and associated statistics are presented in Table 3. To determine the peak spawning time the mean gonad somatic index (GSI = percent ovary weight of the gutted weight) was calculated for each week of the sampling period (Figure 3). It appears that individual fish attain their maximum GSI just prior to spawning the first egg batch and a decline in GSI occurs as successive batches are spawned. This was shown by comparing the mean GSI of each maturity stage (Table 4) which showed an Table 3. — Length and weight relationships of Atlantic mack- erel collected in the Middle Atlantic Bight, 1977. TW = total weight (grams); GW = gutted weight I grams); TL = total length (millimeters); and FL = fork length (millimeters). Symbols refer to the equation Y = a + bX; n = sample size; r = correlation coefficient; S = standard deviation about the line. y a b X n r Sy^ Curvilinear relationships between transformed variates log TW log GW logTW log GW TL TW -5.767 -5.420 -5.780 -5.374 3.275 3.106 3.334 3.140 log TL log TL log FL logFL 966 966 966 966 0.905 0.924 0905 0.924 Linear relationships between untransformed variates 1.793 1.098 FL 966 0.986 -20.410 1.282 GW 966 0.979 0.036 0.030 0.036 030 3.594 22.397 19-25 APRIL Figure 3. — Mean gonad somatic index (ovary weight as a per- cent of gutted weight) plotted by week for Atlantic mackerel sampled in 1977. Numbers in parentheses are sample sizes. Table 4. — Mean gonad somatic index (GSD and standard devia- tion for each maturity stage of Atlantic mackerel. stage Mean GSI so 1 . Developing 2. Ripe 3. Running ripe 4. Partially spent 10.4 15.0 24.9 8.6 2.9 68 4.2 247 4.7 41 2.2 93 increase from stage 1 to 3 and a rapid decrease at stage 4. Similar results were reported by Kaiser (1973 ) for horse mackerel, Trachurus murphyi. He found that gonad somatic indices reflected mat- uration changes of the ovaries and a sharp decline in the mean GSI coincided with the appearance of the earliest spawning females. In this study the weekly mean GSI increased during the first 3 wk of sampling, peaked between 21 April and 4 May, and then declined steadily through the end of the sampling (22 May). All females examined from the last sampling week were partially spent and indi- cated spawning was nearly completed within the study area. 105 FISHERY BULLETIN: VOL. 78, NO. 1 The egg diameter frequencies shown in Figure 2 indicate Atlantic mackerel are serial spawners, i.e., several batches of eggs are shed by individuals throughout the spawning season. The presence of multiple modes in the egg diameter frequencies (Figure 2a-c) and ripening eggs in partially spent ovaries (Figure 2d) are indicators of serial spawn- ing (Clark 1935; MacGregor 1957). A cytological study by Bara ( 1960) has shown that eggs are not shed continuously as stated by Cunningham (1889) but are shed in several batches during the 2-mo spawning period. The potential number of batches spawned was estimated by determining the ratio of ripe eggs to all yolked eggs in six running ripe ovaries. Atlan- tic mackerel eggs, from plankton samples, ranged from 1.01 to 1.29 mm diameter (Berrien 1975; Ware 1977); therefore, in this study, eggs 1.05 mm and larger were assumed to constitute the next egg batch to be spawned. The ratios ranged from 13.7 to 21.7% and averaged 17.0%. Thus the po- tential number of batches spawned per individual was five to seven and averaged six batches. Fecundity estimates ranged from 285,000 to 1,980,000 eggs for fish between 307 and 438 mm FL. Preliminary plots indicated a curvilinear rela- tionship existed for fecundity-length and a linear relationship for fecundity-weight and fecundity- age. However, correlation coefficients (r) were higher for the logarithmic relationships of fecundity-weight and fecundity-age, therefore, all variables were transformed and linear regression equations of the form log Y = a + b(\og X) were calculated. Data plots and the equations relating fecundity to fork length, gutted weight, and age are shown in Figures 4-6. DISCUSSION Spavming by the southern contingent of Atlan- tic mackerel apparently peaked during the 2-wk period between 21 April and 4 May 1977. This 2-wk period represents the mean peak spawning time within the study area (Maryland to Rhode Island) since there is a north and eastward pro- gression of spawning during the spring migration (Bigelow and Schroeder 1953; Berrien 1978). Ber- rien et al.^ observed the north and east progression *Berrien,P. L., A. Naplin, and M.R.Pennington. 1979. At- lantic mackerel, Scomber scombrus, egg production and spawn- ing population estimates for 1977 in the Gulf of Maine, Georges Bank, and Middle Atlantic Bight. Int. Counc. Explor. Sea ICES/ELH Symp./DS:9, 17 p. 1.70 1.50 30 log F 8.346 + 5.544llog FLI N 218 r 88 Si X 0.066 360 380 FORK LENGTH(mm) Figure 4. — Relationship of fecundity and length and the pre- dictive logarithmic (base 10) regression for Atlantic mackerel in 1977. 3 90 O log F 1721 + 1.547 Hog GWI N 218 r 81 Sy X 081 .'•>::'Vi' '■ ■• 300 400 500 600 GUTTED WEIGHT I g) 700 800 900 Figure 5. — Relationship of fecundity and weight and the pre- dictive logarithmic (base 10) regression for Atlantic mackerel in 1977. in plankton mackerel egg densities. They found spawming intensity in the Middle Atlantic Bight was low during mid-April and increased rapidly by late April, and maximum egg densities oc- curred about 25 April. Spawning continued at a reduced rate throughout May and then decreased steadily during June. Very similar results are in- dicated from my analysis of gonad somatic indices during the 1977 spawning season. 106 MORSE: SPAWNING AND FECUNDITY OF ATLANTIC MACKEREL 1.90 log F 5 264+0 840llogAI N 197 r 76 Sy X 084 FIGURE 6. — Relationship of fecundity and age and the predic- tive logarithmic (base 10) regression for Atiantic mackerel in 1977. Observations of spawning times of various temperate-water fish have indicated peak spawn- ing dates may be relatively fixed. Gushing (1969) postulated an indirect link between the fixity of spawning season and the primary production cy- cle. Ware (1977) investigated the relationship of spawning time of Atlantic mackerel at St. Georges Bay, Nova Scotia, to the size and abundance of 80 /xm plankton. He found the mean peak egg produc- tion date was 1 July ± 1 wk and coincided with the maximum abundance of summer plankton. It would appear, at least for the southern contingent, that the time of peak spawning is more variable than that indicated for St. Georges Bay. Sette (1943) determined maximum spawning occurred during mid-May (1928-32) off Middle Atlantic and southern New England States. Ichthyoplankton surveys during the mackerel spawning season in 1966 and 1975-77 (Berrien 1978; Berrien et al. see footnote 6; Berrien and Anderson') within the Middle Atlantic Bight indicated spawning peaked during May in 1966 and 1975 and during April in 1976 and 1977. In fact, eggs were collected as early as 13 April in 1977. Berrien and Anderson (see footnote 7) attribute the April 1976 spawning Bemen, p. L., and E. D. Anderson. 1976. Scomber scom- brus spawning stock estimates in ICNAF Subarea 5 and Statisti- 1Q7C T ^' ^sed on egg catches during 1966, 1975, and ly/b. Int. Comm. Northwest Ati. Fish., Res. Doc. 76/XII/140 i-U p. peak to increased water temperatures within the study area. The factors controlling the spawning time of Atlantic mackerel are unclear. The regularity shown by Ware (1977) would indicate internal control or a constant external stimulus such as photoperiod. Sette (1943) presented evidence indi- cating water temperature is a limiting factor con- trolling migration and in turn the timing of spawning in a fixed location. Gushing ( 1967, 1969) suggested that some fish spawn at a relatively fixed date that is linked to planktonic productivity and that changes in plankton production would cause dramatic changes in year-class success. It appears that a variable spawning date, as shown by the southern contingent— linked to the factors affecting plankton productivity, e.g., tempera- ture, photoperiod, nutrient content— would in- crease the chances for larval survival. The fecundity estimates presented here must be considered as maximum potential egg production because, as reported by Macer (see footnote 3), resorption may significantly reduce the number of eggs spawned. Preliminary observations by Macer indicated an average of 11.4% ofyolked eggs were being resorbed. Bara (1960) observed degc lera- tion in a "few" mature eggs though no quantita- tive data were presented. Studies are needed to define the extent and possible annual changes of resorption rates and their relationship to fecun- dity. ACKNOWLEDGMENTS I wish to thank Darryl Ghristensen and others who provided samples from the recreational catches throughout the course of this study. I ap- preciate the critical reviews and comments by E. Anderson and S. J. Wilk. Special thanks to M. Montone for typing this manuscript and M. Gox for preparation of the figures. LITERATURE CITED Bara, G. I960. Histological and cytological changes in the ovaries of the mackerel, Scomber scombrus L., during the annual cycle. Rev. Fac. Sci. Univ. Istanbul, Ser. B, 25:49-86. BERRIEN, P. L. 1975. A description of Atiantic mackerel. Scomber scom- brus, eggs and early larvae. Fish. Bull., U.S. 73:186-192. 1978. Eggs and larvae of Scomber scombrus and Scomber japonicus in continental shelf waters between Mas- sachusetts and Florida. Fish. Bull., U.S. 76:95-115. 107 FISHERY BULLETIN: VOL. 78, NO. 1 BIGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53, 577 p. Brice.J. J. 1898. A manual of fish-culture, based on the methods of the United States Commission of Fish and Fisheries. Rep. U.S. Comm. Fish Fish. 23:1-261. CLARK, F. N. 1935. Maturity of the California sardine (Sardina caerulea), determined by ova diameter measure- ments. Calif. Dep. Fish Game, Fish Bull. 42:5-49. Cunningham, J. T. 1889. Studies of the reproduction and development of tele- ostean fishes occurring in the neighborhood of Plymouth. J. Mar. Biol. Assoc. U.K. 1:10-54. Gushing, D. H. 1967. The grouping of herring populations. J. Mar. Biol. Assoc. U.K. 47:193-208. 1969. The regularity of the spawning season of some fishes. J. Cons. 33:81-92. HICKLING, C. F., AND E. RUTENBERG. 1936. The ovary as an indicator of the spawning period in fishes. J. Mar. Biol. Assoc. U.K. 21:311-316. HISLOP, J, R. G., AND W. B. HALL. 1974. The fecundity of whiting, Merlangus merlangus(L.), in the North Sea, the Minch and at Iceland. J. Cons. 36:42-49. Kaiser, C. E. 1973. Gonadal maturation and fecundity of horse mack- erel Trachurus murphyi (Nichols) off the coast of Chile. Trans. Am. Fish. Soc. 102:101-108. MacGregor,J. S. 1957. Fecimdity of the Pacific sardine (Sardinops caeru- lea). U.S. Fish Wildl. Serv., Fish. Bull. ?>1A21-AAQ. Moores, J. A., G. H. Winters, and L. S. Parsons. 1975. Migrations and biological characteristics of Atlantic mackerel (Scomber scombrus) occurring in Newfound- land waters. J. Fish. Res. Board Can. 32:1347-1357. SETTE, O. E. 1943. Biology of the Atlantic mackerel (Scomber scom- brus) of North America. Part I: Early life history, includ- ing growth, drift, and mortality of the egg and larval populations. U.S. Fish Wildl. Serv., Fish. Bull. 50:149- 237. 1950. Biology of the Atlantic mackerel (Scomber scom- brus) of North America. Part II: Migrations and habits. U.S. Fish Wildl. Serv., Fish. Bull. 51:251-358. Steven, G. a. 1952. Contributions to the biology of the mackerel, Scomber scombrus L. III. Age and growth. J. Mar. Biol. Assoc. U.K. 30:549-568. WARE, D. M. 1977. Spawning time and egg size of Atlantic mackerel, Scomber scombrus, in relation to the plankton. J. Fish. Res. Board Can. 34:2308-2315. 108 RESPIRATION AND DEPTH CONTROL AS POSSIBLE REASONS FOR SWIMMING OF NORTHERN ANCHOVY, ENGRAULIS MORDAX, YOLK-SAC LARVAE Daniel Weihs' ABSTRACT Larval northern anchovy in the yolk-sac (nonfeeding) stage exhibit regular bursts of continuous swimming during the first 3 days after hatching. It has been suggested that this behavior may have a respiratory function. A different possibility is depth control, countering the tendency of the larvae to sink when motionless. This paper includes a theoretical and experimental investigation of the possible functions of these swimming bouts. The theoretical approach was to define a model and calculate the oxygen available to the larva when resting and while moving, and experiments were jjerformed as a check of the theoretical results. The experiments were conducted on yolk-sac larvae in sealed tanks with varying dissolved oxygen concentrations to determine the effects of reducing the available oxygen on the frequency and duration of the swimming bursts. Results of the experiments confirmed the theoretical model. They indicate that the swimming bouts both help the larva stay at a constant depth and have a respiratory function when the oxygen concentration in seawater is less than 60% of saturation. Newly hatched northern anchovy, Engraulis mordax, larvae exhibit a pattern of regular short bouts of continuous swimming interspersed with periods of resting. These larvae are still in the yolk-sac stage and are not feeding so that the locomotory behavior must have some other pur- pose, as these motions are energy consuming and also endanger the animal by attracting predators (Lillelund and Lasker 1971). Hunter (1972) suggested that these swimming bouts might have a respiratory function. Respiration has to be by cutaneous diffusion through the 2-3 /xm thick skin (Lillelund and Lasker 1971) of the larvae as the gills develop only at a later stage. The purpose of this paper is to test this hypothesis and another possibility, depth control, to counter sinking due to the negative buoyancy, using theoretical and experimental methods. First, I develop a theoretical model for oxygen transport to motionless and swimming yolk-sac larvae and estimate the possible oxygen uptake. Next, I describe the experiments to test the predic- tion of the theory for both proposed mechanisms and compare their results. 'Southwest Fisheries Center La Jolla Laboratory, National Manne Fisheries Service, NOAA, La Jolla, Calif.; present ad- dress: Department of Aeronautical Engineering, Technion, Hidfa, Israel. METHODS Analytical Model A mathematical model is now introduced to con- sider the possible respiratory function of the bouts of continuous swimming of yolk-sac anchovy lar- vae. First, we calculate the oxygen transport to a motionless larva. This transport is then compared with the metabolic requirements. If the metabolic requirements are not met, larval motion (and the resulting convective diffusion) is required. The size of yolk-sac larvae (2.7-4.0 mm total length) and their swimming speeds (Hunter 1972) lead to tj^ical Reynolds numbers, based on larval length (Weihs 1980) of <20. (The Reynolds number is a nondimensional factor indicating the relative importance of pressure and viscous effects on a body moving in a fluid under given circum- stances — the higher the Reynolds number, the smaller the influence of the viscosity.) The larvae, as a direct result of their small size, are in a highly viscous laminar flow situation in which turbulent effects can be neglected. Thus, the larvae and their immediately surrounding water would be transported together in oceanic turbulent eddies, which are of the order of tens of centimeters in diameter. As a result, a nonswimming larva would stay for a relatively long period in the same mass Manuscript accepted; July 1979. FISHERf BULLETIN: VOL. 78, NO. 1, 1980. 109 FISHERY BULLETIN: VOL. 78, NO. 1 of water, even though that mass is convected on a much larger scale. The motionless larva and its surrounding water mass may therefore be analyzed separately, as a distinct system in the thermodynamic sense. Within this system, oxygen transport to the larva is controlled by molecular diffusion because the gill system is not developed at this stage. This process is time dependent, beginning when the larva arrives in a certain location (by swimming) and rests, ending when swimming begins again. Initially the oxygen concentration in the water mass surrounding the larva is uniform, but the larva now starts acting as an oxygen sink, gradu- ally depleting the oxygen content of the water surrounding it. This concept of the larva as an oxygen sink simplifies the calculations, as knowl- edge of the exact distribution of oxygen diffusivity on the animal's surface is not required. The sink model also is useful here as it averages out the direction of local transport and the body of the larva into which the oxygen diffuses can be taken as an equivalent sphere of equal surface area (Figure 1). Diffusion into a sphere is most conveniently analyzed in the spherical coordinate system. The governing conservation of mass equation can be written (Crank 1975) as dt = D c 2 dc + dr (1) where c is the mass fraction of oxygen (a function of the distance and time); r is the radial distance, measured for the center of the equivalent spheri- cal body (the sink); t is the time; and D is the diffusion coefficient of oxygen in seawater. The temporal boundary condition is the initial, uniform state c(r,0) = c^ while the spatial conditions are c(^,0 =Co (2) (3) which states that far from the animal the oxygen concentration stays unchanged at all times. Strictly, the condition should be defined at some finite distance but as that distance is much larger than the animal equivalent radius, it can be ap- proximated by ^. Next, the oxygen concentration boundary condition at the surface of the equiva- lent sphere r = a is obtained. Muscles and vas- cularized tissues have much higher oxygen trans- port rates than seawater, due to internal uptake augmented by active transport. Thus, oxygen will be absorbed at the surface of the larva as fast as it arrives by diffusion from the surrounding water. The oxygen concentration c at the larva's surface (r = a) is thus constant, and very low, i.e.. c (a, t) = Cj where Cj->0 and ^>0. (4) Equations (2)-(4) enable solving equation (1) analytically, by classical methods. The solution can be written in nondimensional form for the concentration as C-Cq ci -Co ± erfc ^~° '' 2Vd7 (5) where the complementary error function, erfc, is defined as erfc(2) 2 " — 22 f e dz \/lT (6) Numerical values of the complementary error function are found in most mathematical tables (e.g., Abramowitz and Stegun 1965). The rate of mass transfer (flux) J to the animal is now obtained from A O^ dA (surface A) (7) where p is the density and A the surface area of the body. For the equivalent sphere of radius a,A- Aira"^. dc/dr is assumed spherically symmetric so that J = -pDA dc_ dr (8) where the concentration derivative is obtained from Equation (5). Substituting this, and the value for the surface area, and setting c^ = as in Equation (4), the total mass flux per unit time J^ is 110 WEIHS: RESPIRATION AND DEPTH CONTROL IN ENGRAUUS MORDAX Figure l . — Schematic description of model and spherical coordinate system centered on the center of mass of a northern anchovy larva: a is the radius of an equivalent sphere of equal surface area (not to scale), I is the larval length, and b and I are the average tail strip depth and length. Jd = -pDcQ Anr^— erfc ^2y/Dt (9) and when r = a, this simplifies to 2/1 1 Jd = AnpDcQa'^ (— + ° 2VDt^ (10) Equation (10) consists of two terms in the brack- ets, multiplied by a constant factor. The first term in the brackets is the constant, time-independent contribution while the second describes the initial 111 FISHERY BULLETIN: VOL. 78, NO. 1 transient. The latter drops rapidly, proportionally to the square root of time elapsed since arrival of the larva. Substituting numerical values into Equation (10), the oxygen flux can be compared with oxygen requirements of larval anchovy to see if the swimming motions are required for respira- tion. The equivalent radius, a, of the larval anchovy is found by equating the surface area of the larva and the equivalent sphere. The larva, at this yolk-sac stage, is described for diffusion purposes as a sphere of radius q (the yolk sac) attached to an almost flat ribbon of length Zj and average breadth b. The combined surface area of the sphere and ribbon is then taken to be equal to the area of the equivalent sphere appearing in Equation (10). Thus 47ra2 = 2l^b + Anq^. (11) Using typical values for these parameters for newly hatched larvae we obtain /j = 1.4 mm, b = 0.3 mm, and q = 0.3 mm (from drawings by E. H. Ahlstrom, Senior Scientist, Southwest Fisheries Center, NMFS, NOAA, La Jolla, CA 92038), i.e., a = 0.0395 cm. The mass content of oxygen in sea- water at 20° C is Co =7.8 x IQ-^ g/cm^^ (Prosser 1973), and the mass fraction is obtained by divid- ing by the density of sea water, which then cancels out in Equation (10). Finally, the diffusion coefficient of oxygen is approximately equal for freshwater and seawater (Riley and Skirrow 1965) so that a reasonable value for 20° C is 1.8 x 10'^ cm2/s (O'Brien et al. 1978), or D = 1.08 x lO'^ cm^/min. Substituting all these values into Equa- tion (10) we obtain J= (4.18 + 2.51 r'/2) 10-^ g/min (12) when the water is 100% saturated. Reducing the oxygen content of the water causes the value of the oxygen flux, J, to go down proportionally, i.e., by multiplying J from Equation ( 12) by the fraction of saturation. Some typical values of J appear in Figure 2 with the percent of saturation as the parameter. When the larva starts swimming, two changes in the oxygen supply occur. First, the animal's motion produces a convective local flow relative to the body, thus bringing new, oxygen-rich water closer and removing the respiratory waste prod- ucts. Secondly, the absolute motion will bring the larva to an area where the oxygen concentration is oo mm Figure 2.— Rate of oxygen transport (J) to motionless northern anchovy larva by diffusion versus time (t). Broken part of curve shows asymptotic value, after the initial transient has disap- peared. Parameter is oxygen concentration in percentage of sat- uration. still at the initial ambient value, starting the pro- cess described by Equations (l)-(4) again. Following this reasoning, even relatively slight motions causing just a local flow around the ani- mal's body would suffice for respiratory functions. Thus, actual swimming would not be required. However, the yolk-sac larvae are increasingly negatively buoyant with age (Hunter and Sanchez 1976), which causes them to sink. Therefore, ac- tive absolute motion is necessary for the larva to stay at a given depth for feeding and future school- ing. When the larva is swimming, the process of transport of oxygen changes to convective diffu- sion, and as such is described by a different model (Daykin 1965). Daykin's work dealt with station- ary eggs in a moving river environment, but for mass transfer purposes this is equivalent to a larva (or egg) moving at constant speed relative to the water. In the convective diffusion process (Levich 1962; Daykin 1965), the mass transfer to the larva can be roughly described by '^con = 47TaHc„-C^)k (13) 112 WEIHS; RESPIRATION AND DEPTH CONTROL IN ENGRAULIS MORDAX where ^ is a diffusion coefficient obtained from experimental correlations of the diffusional flux with the Reynolds (Re) and Schmidt numbers (Sc). (The Schmidt number is the ratio of the kinematic viscosity to the diffusivity and nondimensionally indicates the relative importance of these two ef- fects in a given flow situation.) For the present circumstances D k = — {2 + 0.6 Re '/2 Sc'/3) 2a (14) for average swimming speeds of approximately 5 cm/s ( Hunter 1972) and a Schmidt number of 600 we have J,„„ = 1.27 X 10"^ g/min. (15) Hence, oxygen transport due to convective diffu- sion is over 20 times higher than for the motion- less larva (Equation ( 12)). The calculation leading to Equation (15) is approximate, as the larva's shape will influence the coefficient 0.6 in k (Equa- tion (14)) and also change the form of Equation ( 13). It is, however, accurate to at least an order of magnitude (Levich 1962). Thus, once the larva starts swimming, the mass transfer of oxygen to its surface increases by at least an order of mag- nitude. Recently, an additional mechanism for oxygen transport to stationary eggs was identified by O'Brien et al. (1978) who showed that under certain riverbed conditions natural convection, due to the oxygen and metabolite gradients, may contribute to the oxygen transfer. This effect may play a supplementary role in the present (pelagic) case as the natural convection effects are much smaller than the forced convection. Tests with Larvae Egg batches were obtained once a week from groups of adult northern anchovy maintained in the laboratory and induced to spawn. Measure- ments were made each week during a 6-wk period to minimize bias due to a single cohort group. Water temperature ranged from 19° to 21° C, and overhead fluorescent lighting was used. The 50% hatching point was determined and defined as "day 0" for each batch. Experiments were carried out on age day larvae every week (six times). A set of five 2,000 ml graduated cylinders filled with filtered seawater was used for the environ- mental tests. Oxygen concentrations of 100, 80, 60, 40, and 20% of saturation at the measured temperature were produced by bubbling nitrogen through each of the cylinders. After the larvae were added (about 25 individuals/cylinder), the cylinders were sealed off with rubber stoppers. Oxygen concentrations were measured periodi- cally during the experiments with a Beckman In- strument Model 160 Physiological Gas Analyzer^ to check on initial values and possible drift;. Individual fish were monitored for a 5-min period, and duration and number of swimming bursts were recorded on a Esterline- Angus Opera- tion Recorder Model AW. Records were also made of approximate swimming direction (measured from horizontal) as well as the change in orienta- tion of motionless larvae while they were sinking during the resting periods. Five active larvae were monitored in each container every week, for both day and day 1 tests. After the day experiments were finished each week, the equipment was reset and the day 1 tests conducted 24 h later with additional larvae from the same batch. The latter larvae were kept in oxygen-saturated water from hatching to minimize stress due to oxygen starvation. RESULTS No appreciable change in the proportion of time spent in burst swimming was observed when the measurement at 100% of saturation concentration of oxygen (which is the oxygen level in the natural state in the sea because of turbulent interchange with the atmosphere) was compared with the time spent in motion at the 80 and 60% oxygen levels (Figures). When oxygen levels were <60% of saturation, large increases in the time spent swimming were observed. The rate of increase of swimming time in both ages (day and day 1) were similar. Various attempts at describing all five data points for each age-group by means of a single empirical exponen- tial function were not successful (low coefficients of determination). Thus, it seems that a different behavioral mechanism is triggered when oxygen levels fall below 60%^^ of saturation at the given temperatures, i.e., much lower than expected oxy- gen concentrations in the upper layers of the sea, where the anchovy larvae are usually found. ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 113 FISHERY BULLETIN: VOL. 78, NO. 1 015 0.10 O05 S(0. Figure 3. — Fraction of time spent actively swimming, ^, versus oxygen concentration in percentage of saturation. S[02], for newly hatched (day 0) and 24-h-old (day 1) northern anchovy larvae. Each point on the full curves is an average of between 20 and 25 individual 5-min observations. Bounds are standard er- rors. Dashed lines indicate idealized model of constant fraction, I,, at high S[02l and linearly increasing i at low SLOj]. The duration of bursts increased monotonically as the oxygen levels decreased, while the number of bursts dropped significantly to a minimum of 1/min at 60-80%, increasing sharply after that (Figure 4). No satisfactory explanation has been found for the drop in the number of bursts at 80% of saturation. The main result illustrated by Fig- ure 4 is that both the frequency and duration of bursts increase markedly at low oxygen concen- trations, both contributing to the increase in time spent swimming. The center of gravity of anchovy larvae is in the vicinity of the head and therefore they tend to be oriented in an oblique head-dowm configuration after swimming ceases. More mature larvae, which have converted significant amounts of yolk into denser tissue are negatively buoyant (Hunter and Sanchez 1976) and tend to sink head down- ward at rates of approximately 1-2 mm/s. To check the vertical station-keeping hypothesis, the direc- tion of swimming was recorded, as well as the body 40% 60% sLo^] Figure 4. — Average duration (t) of swimming bouts (full lines) and number of bouts (n) in 5 min (dashed lines) for day (solid triangles) and day 1 (solid circles) larvae, versus oxygen in per- centage of saturation concentration (SIOjl). orientation, when swimming started. The results of over 1 ,400 recorded swimming periods appear in Table 1, which lists average values of the body angle at the onset of swimming and the direction of swimming. No significant variation in swimming direction with oxygen concentration was found for either day or day 1 larvae (Table 1). The spread in results was large, as is noticeable from the stan- dard errors. The total possible spread of data is ±90°, which suggests that swimming direction is actually a random phenomenon for day larvae. At age 1 day, a positive bias was observed in the swimming direction, still with large variation. The body inclination at the beginning of the swimming periods was consistent, at around -65° with the exception of the day 0, 20% oxygen data, which is influenced by additional factors, dis- cussed below. DISCUSSION The analytical model predicted that a motion- less larva would be able to pick up oxygen at a decreasing rate at any given spot (Equation 12). Table l. — Initial orientation of the body and duration of swimming during bouts of continuous sv«mming by newly hatched and 1-d-old northern anchovy larvae. Error bounds are standard error. Angles are measured from the horizontal. Positive values indicate upward motion. Averages of day do not include 20% O^ values as these (indicated by question marks) include different phenomena. Oxygen concentration N. number of observed events Day larvae Day 1 larvae Orientation of body at start of swimming period (degrees) Direction of svifimming relative to horizontal (degrees) (% of saturation) Day larvae Day 1 larvae Day larvae Day 1 larvae 100 80 60 40 20 Weighted average 106 103 109 161 210 157 113 112 140 204 -68.9 = 23.7 -66.3 = 29.4 -72.1=15.6 -61.0=30.3 -45.6 = 36.7(7) -66.4=25.2 -74.6 = 11.8 -80.0± 9.2 -66.7=17.6 -69.2 = 22.0 -63.8=27.8 -70.2 = 16.9 -4.7 = 56.1 -15.0 = 59.0 -3.9 = 73.6 12.3 = 540 38.7 = 43.3(7) -2.8 = 58.3 44.7 = 45.8 28,2 = 58.1 51.9=31.9 33.2=45.9 37.4 = 52.3 39.0=48.1 114 WEIHS: RESPIRATION AND DEPTH CONTROL W ENGRAUUS MORDAX This prediction has now to be compared with the requirements of the organisms to determine if ad- ditional oxygen is needed. Data for oxygen con- sumption at 17° C of late-stage anchovy eggs and larvae of differing ages as a function of time since spawning has been obtained by Theilacker,^ by measurement in a respirometer. These data were adjusted to 20° C, the temperature at which most of my experiments were conducted (Figure 5), by means of a temperature-growth correlation for larval northern anchovy (Zweifel and Hunter^). This adjustment was made by calculating the size of the larvae at 17° C at the ages recorded by Theilacker, then translating these into age for the same size at the new temperature, which gave a smaller size because growth rates increase with temperature. Thus, an estimate for the oxygen requirements at 20° C of size-defined larvae was obtained as a function of their age ( the 20° C line in Figure 5). The value of the ratio of oxygen consumption 1 d after hatching to that at hatching was about 1.6 (Figure 5). Returning now to Figure 3 we see that the ratio for the average percent of time spent swimming of the two age-groups is approximately 1.66. I conclude, therefore, that the distance be- tween the day and day 1 curves in Figure 3 is an indication of the increased general activity of the larvae as they grow. This correlation indicated in Figures 3 and 5 serves as an additional verifica- tion of both Theilacker's respirometer data and the present swimming data. To compare experimental values of oxygen con- sumption to the prediction of the model we plot the data as Figure 6, where the horizontal lines show the range of oxygen requirements (from Theilack- er's data) at hatching and 24 h later. The steady- state oxygen available by steady-state diffusion only (after the initial transient) obtained from the time-dependent first term in Equation ( 12) is now superimposed. Figure 6 indicates that pure diffu- sion supplies all the oxygen required for the day larvae only when <42±47f of the O2 saturation concentration is available. This changes to 63 ± 4% of saturation for the day 1 larvae. The sharp dis- continuity in the swimming data, occurring be- 0.61- ^G. Theilacker, Fishery Biologist, Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, La Jolla, CA 92030, pers. commun. November 1978. "Zweifel, J. R., and J. R. Hunter. 1978. Temperature specific equations for grovrth and development of anchovy (En- graulis mordax) during embryonic and larval stages. Unpubl. manuscr. , 37 p. Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, La Jolla, CA 92038. 0.5 0.4 0.3 0.2 0.1 / / / 20»C / / / / // / / 17 'C / / / / / /X"" Hatching ).ol \ L 2 4 6 8 10 12 DAYS 14 Figure 5. — Oxygen consumption by northern anchovy eggs and larvae versus time elapsed since spawning. Open circles indicate experimental data for 17° C. The line for 20" C is extrapolated from the 17^ C data with the aid of the Zweifel and Himter model (see text). s[o,] Figure 6. — Estimated oxygen requirements (J^) of northern anchovy larvae at day (hatching) and day 1, and steady-state oxygen supply by diffusion versus oxygen in percentage of sat- uration concentration (S[02]). The triangle and circle denote concentrations at which observed swimming behavior changes at day and day 1, respectively (see Figure 3). 115 FISHERY BULLETIN: VOL. 78, NO. 1 tween 60 and 407f oxygen concentration ( Figure 3 ) can now be understood in terms of the theoretical results above. When the oxygen concentration in this water is 60% or higher, diffusion alone can satisfy the respiratory requirements of both day and day 1 motionless larvae. Thus, the swimming activity at the higher concentrations is due to other factors, such as depth control. The measured activity level (Figure 3) does not change between 60 and lOO'/f oxygen concentration, as expected from the theoretical model's predictions. The increased swimming activity observed when the concentration drops below the 40-60% level must therefore be a respiratory reaction. Ac- tive swimming causes convective diffusion (which, as shown in the Analytical Model section, leads to much higher oxygen transport rates) and moves the larva to a new, nondepleted position. As ex- pected from this mechanism, activity increases with decreasing ambient oxygen concentration, as oxygen transport rates drop below the required level faster at low ambient concentration, initiat- ing motion more often. The dashed lines in Figure 3 verify this theoretical reasoning and the ob- tained values of 59% concentration (day 1) and 55% concentration (day 0) for the beginning of respiration-driven swimming are in very good agreement with Figure 6, especially considering the experimental errors involved in the various data sources. Next, I consider the significance of swimming activity at higher oxygen concentration. The most plausible reason for the swimming behavior is to keep the larvae, which are negatively buoyant, from sinking out of the preferred depth zone in the sea. Day 1 larvae swim at an average angle of 39° upwards from the horizontal with no significant variation with oxygen concentration (Table 1). The large standard error is an indication of the wide spread of observed directions. The average swimming speed at this stage is 5.2 ±4.1 cm/s (Hunter 1972) and the average duration of a swimming bout (at oxygen concentration of 60- 100%) is about 2.1s (Figure 4). The average verti- cal component of the distance moved during a single bout is therefore /i„p = Vt sin a = 5.2 2.1 sin 39° 6.9 cm. (16) The uncertainty in this value is large due to the standard errors in both the swimming angle and 116 the average swimming speed, but it is probably accurate at least to an order of magnitude. Be- tween swimming periods, the larvae sink at a speed of 0.12 ±0.03 cm/s (Hunter and Sanchez 1976). The average number of swimming bouts per 5-min period was found to be about 7 (Figure 4), i.e., giving an average sinking time of 43 s. This leads to a vertical distance of 5.2 cm, which is close enough to the value of 6.9 cm of Equation (16) to show that the swimming of day 1 larvae at high oxygen levels most probably is a depth-control mechanism. The newly hatched (day 0) larvae present a dif- ferent situation. Pelagic eggs are slightly posi- tively buoyant (Blaxter 1969) while the chorion, which is shed during hatching, is somewhat nega- tively buoyant. Thus, while no measurements in- dependent of the present ones exist, it is reason- able to assume that these newly hatched larva are approximately neutrally buoyant due to their large yolk sac. As the yolk is consumed, the specific gravity increases and the sinking rates for day 1 are obtained. The larvae are approximately neutrally buoyant during the first hours after hatching so that no net sinking or upward swim- ming is expected. Table 1 shows that this is actu- ally the case at day 0, where the average direction is very close to horizontal and the large error indi- cates almost random swimming direction. Some upward swimming may be discerned at the very low O2 concentration experiments (20% O2). This may be a result of an inadvertent oxygen gradient in the tank or a phototactic response induced by the low oxygen concentration. Phototaxis is prob- ably the means by which the older larvae choose swimming direction, and is directed upwards as light in the present experiments comes from the surface. ACKNOWLEDGMENTS This paper was written while I was a NRC- NOAA Senior Research Associate, on leave from the Department of Aeronautical Engineering, Technion, Haifa, Israel. I would like to thank John R. Hunter and Reuben Lasker for reading the manuscript and various discussions, E. H. Ahlstrom and Gail Theilacker for generously al- lowing me to use their unpublished data, and Bob Millman and Steve Lucas for help with the exper- iments. WEIHS: RESPIRATION AND DEPTH CONTROL IN ENGRAULIS MORDAX LITERATURE CITED ABRAMOWITZ, M., AND I. A. STEGUN (editors). 1965. Handbook of mathematical functions with formulas, graphs, and mathematical tables. Dover, N.Y., 1046 p. BLAXTER, J. H. S. 1969. Development: eggs and larvae. In W. S. Hoar and D. J. Randall (editors), Fish physiology. Vol. 3, p. 177-252. Acad. Press, N.Y. Crank, J. 1975. The mathematics of diffusion. 2d ed. Clarendon Press, Oxf.,414 p. DA'iKIN, P. N. 1965. Application of mass transfer theory to the problem of respiration of fish eggs. J. Fish. Res. Board Can. 22:159-171. HUNTER, J. R. 1972. Swimming and feeding behavior of larval anchovy Engraulis mordax. Fish. Bull., U.S. 70:821-838. HUNTER, J. R., AND C. SANCHEZ. 1976. Die! changes in swim bladder inflation of the larvae of the northern anchovy, En^rau/is mordax. Fish. Bull., U.S. 74:847-855. LEVICH, V. G. 1962. Physiochemical hydrodynamics. Prentice-Hall, Inc., Englewood Cliffs, N.J., 700 p. LILLELUND, K., AND R. LASKER. 1971. Laboratory studies of predation by marine copepods on fish larvae. Fish. Bull., U.S. 69:655-667. O'Brien, r. n., S. visaisouk, r. Raine, and d. f. Alderdice. 1978. Natural convection: a mechanism for transporting oxygen to incubating salmon eggs. J. Fish. Res. Board Can. 35:1316-1321. Prosser, C. 1973. Comparative animal physiology. W. B. Saunders Co.,N.Y.,966p. Riley, J. P., and G. Skirrow. 1965. Chemical oceanography. Academic Press, N.Y., Vol. 1., 712 p. weihs, D. 1980. Energetic significance of changes in swimming modes during growth of larval anchovy, Engraulis mor- dax. Fish. Bull., U.S. 77:597-604. 117 DESCRIPTIONS OF LARVAL SILVER PERCH, BAIRDIELLA CHRYSOURA, BANDED DRUM, LARIMUS FASCIATUS, AND STAR DRUM, STELLIFER LANCEOLATUS (SCIAENIDAE)'^ Howard Powles^ ABSTRACT This paper presents descriptions and illustrations of larval Bairdiella chrysoura (3.1-8.8 mm standard length), Larimus fasciatus (3.0-5.9 mm standard length), and Stellifer lanceolatus (2.8-15.1 mm standard length). Larimus fasciatus larvae are characterized by brain pigment, pectoral fin pigment, and early-developing pectoral fin rays. Larval B. chrysoura resemble S. lanceolatus, but B. chrysoura have a swath of expanded melanophores from nape to cleithral symphysis. These two species also can be differentiated by the sequence of melanophores in the midventral line posterior to the anus. Off the southeastern United States, L. fasciatus spawn in continental shelf waters from May to October, andS. chrysoura and S. lanceolatus spawn in coastal and estuarine waters during late spring and summer. The perciform family Sciaenidae is represented by 18 species off the southeastern United States (Ta- ble 1). Taxonomy of adult Sciaenidae of the west- em North Atlantic has recently been revised by Chao (1978); nomenclature in the present paper follows Chao (1978) rather than Bailey et al. (1970). Studies of larval sciaenids of the east coast of the United States have been numerous; these have recently been summarized in several publi- cations (Scotton et al. 1973; Johnson 1978; Powles and Stender 1978; Lippson and Moran"*.) Despite the number of larval studies, their quality has been uneven; for example, larval series now known to consist of more than one species have been described as single species (Menticirrhus americanus and Stellifer lanceolatus of Hilde- brand and Cable 1934), damaged or distorted specimens have been illustrated and described (Sciaenops ocellata of Pearson 1929; Leiostomus xanthurus of Hildebrand and Cable 1930), and illustrations have differed from descriptions of larvae of the same species in the same publication (early Stellifer lanceolatus of Hildebrand and Cable 1934). Further, few detailed developmental 'South Carolina Marine Resources Center Contribution No. 94. ^MARMAP Contribution No. 164. 'Marine Resources Research Institute, Charleston, S.C; pres- ent address: Gouvemement du Canada, Peches et Oceans, Divi- sion des Sciences halieutiques.C.P. 15500, Quebec, Canada GIK "Lippson, A. J., and R. L. Moran. 1974. Manual for iden- tification of early developmental stages of fishes of the Potomac River estuary. Md. Dep. Nat. Resour., Power Plant Siting Pro- gram, PPSP-MP- 13: 1-282. series of morphometric, meristic, and pigmenta- tion data have been published, making separa- tion of larvae to species impossible in the early stages before complete development of fin ele- ments. Thus, both description of undescribed or incompletely described larvae and redescription of larvae which have been poorly described in the literature are necessary to specific identification of sciaenid larvae. The three species whose larvae are treated in this paper are generally similar in habitat and probably in ecology. They are small fishes (maximum total lengths 20-23 cm) of coastal and estuarine waters (Hildebrand and Schroeder 1928; Hoese and Moore 1977). None are important commercial or sport fish, but all are abundant in estuaries (Dahlberg 1972; Shealy et al. 1974) and on coastal shrimp grounds (Anderson 1968; Reiser 1976). Because of their abundance and small size, all may be important prey items for larger, pre- dacious fishes. Descriptions of larvae of all three species have been published. Kuntz (1915) described eggs and yolk-sac larvae o{ Bairdiella chrysoura from eggs obtained from a ripe female and further described larvae and early juveniles from plankton collec- tions. Since he examined live or fresh material rather than Formalin-preserved^ material, it is to be expected that body proportions and pigment characters of his series might differ from those in Manuscnpt accepted August 1979 FISHERY BULLETIN; VOL. 78, NO. 1, 1980. ^Reference to trade name does not imply endorsement by the National Marine Fisheries Service, NOAA. 119 larvae from preserved series. His description was cited by later authors (Hildebrand and Cable FISHERY BULLETIN: VOL. 78, NO. 1 1934; Scotton et al. 1973; Johnson 1978; Lippson and Moran see footnote 4) in compilations of larval Table l. — Reported meristics of South Atlantic Bight Sciaenidae. Counts in parentheses occur infrequently and semicolon indicates separate dorsal fins. Dorsal Anal Gill rakers Species Source' Spines Rays Spines Rays Caudal procurrent Vertebrae^ Upper Lower Bairdiella chrysoura Cynoscion nebulosus C nothus C regalis Equetus acuminatus E. lanceolatus E. punctatus E. umbrosus Larimus fasciatus Leiostomus xanthurus Menticirrhus americanus M. littoralis M. saxatilis Micropogonias undulatus Pogonias cromis Sciaenops ocellata Stelliler lanceolatus 1 4 5 7 1 3 4 5 7 9 1 2 3 4 5 7 9 1 4 5 6 7 1 5 7 8 1 5 7 8 1 5 7 8 1 5 7 1 4 5 7 1 4 5 7 1 4 5 7 1 4 5 7 1 4 5 7 1 4 5 7 1 4 5 7 1 4 5 7 1 5 7 X;l XI-XII Xl;l XII IX-X X(XI);I X:l XI-XII X;l X;l X;l XI X:i X;l X:i X-XI;I XI VIII-IX;I X;l X-XI IX-X;I XII-XIII;I XIV-XVI;I XIII-XIV XIII-XIV XI-XII;I XI-XII;I XIII XI-XII;I X-XI;I X;l IX-X I X:l X;l X;l XI-XII IX;I X;l X;l XI-XII X;l X;l X;l XI X-XI X;l X;l XI X;l X;l X;l XI X;l X;l X;l XI X;l X;l X:l XI X;l X;l X;l XI XI-XII;I Xl;l XII-XIII 19-23 19-21 22 19-22 25-28 25 24-26 25-27 24-27 25-27 26-30 27-30(31) 24-28 28-29 27-29 28-30 26-29 26-29 25-28 26-29 24-29 24-28 37-41 38-40 36-40 37-40 47-55 53 46-50 49-55 45-47 46 44-49 45-47 38-40 40 38-39 24-27 24-27 24-26 25-27 29-35 30-34 31 29-32 20-26 24-27 24-25 24-26 19-26 24-26 23-25 24-25 22-27 24-26 26-27 23-25 27-30 28-29 28-29 28-29 19-21 20-22 21 21-23 23-25 23-25 24 23-25 20-24 20-23 21-24 8-10 9-10 10 8-10 10-11 10-12 10-11 10 10-11 9-10 8-10 8-9(10) 8-10 9 9-10 9 9-11 11-13 11-12 11-13 10-12 10-12 7-8 7 6-8 6 5 6 6-8 6-7 7-8 7 7 7 6-8 5-6 6 12-13 12-13 12 12-13 6-8 7-8 7 7-8 7 7 7 7-9 8-9 8 7-8 8-9 8 7 8 5-6 6-7 5-6 6 8-9 8 8 7-8 7-8 7-9 8-9, 5-8 6-9, 5-7 7-8, 6-8 7-9, 5-7 7-8, 6-7 7-8, 6-7 7,5-7 7-8,7 6-7, 4-7 6-8, 6-8 8-9,7 7-8.6 6-8,6 8-9,8 8-9,7 8-10,7-9 7-9, 6-9 12 + 13 11 + 14 (12)13 + 12(13) 25 13 + 12 15 + 12 27(26) 14 + 13 (12)13+12(13) 14-15+10 13 + 12 10 + 15 10 + 15 10+15 10 + 15 10+15 10+15 10 + 15 10 + 15 11 + 14 10+15 10+15 10 + 15 10+15 10 + 15 10+15 10+15 10 + 15 10+15 10 + 15 10+15 10+14 10 + 14 10 + 15 7-8 2-3 4 3-4 4-5 5-6 6 5-6 6 10+15 11 + 14 10+15 4-6 11-13 12 8-12 8 2-3 3-5 3-5 8-10 7 4-6 4 4-5 5 10-13 13 14-16 14-16 16 7-9 8 8 7 6-8 8-10 12-14(15) 12-14 9- 9 10-13 11-13 5 11 9-14 9 10-13 9 10-13 11 10-12 22-25 23-25 24 20-23 22-23 22 0-7 6 0-8 7-8 7 0-7 6 14-18 14-16 16 12-16 14-16 12 7-9 8-9 7 22-23 22 120 POWLES: DESCMPTIONS OF LARVAL SILVER PERCH, BANDED DRUM, AND STAR DRUM T.\BLE 1. — Continued. Source' Dorsal Anal Caudal procurrent Vertebrae Gill rakers Species Spines Rays Spines Rays Upper Lower Umbhna coroides 1 4 5 X;l X;l X;l 26-31 29 27-28 II 6 II 6 II 6-7 11-1-14 5-7 7-10 11 5 9 'Sources: 1. Chao 1978 6. Lippson and Moran (see text footnote 4) 2. Ginsburg 1929 7 Miller and Jorgensen 1973 3. Hildebrand and Cable 1934 8. Randall 1968 4. Hildebrand and Schroeder 1928 9 Welsh and Breder 1923. 5. Jordan and Evermann 1896 'Includes urostyle. sciaenids. Jannke^ illustrated 5. chrysoura of 2.0 and 5.0 mm SL (standard length). Hildebrand and Cable (1934) described a series identified as Larimus fasciatus. Although there is some dis- agreement between illustrations and descriptions of early larvae in this work, the series appears to represent a single species and to be correctly identified. Hildebrand and Cable also described larvae and juveniles identified as Stellifer lanceolatus . Their early larvae represent a mixed series; larvae <4 mm long had pectoral fin pig- ment and developed pectoral rays, but larvae >4.5 mm long had no pectoral pigment and pectoral rays that developed at ^5.6 mm. Body proportions also changed between 4.0 and 4.5 mm. Their series appears coherent and correctly identified at lengths 5=5.6 mm. There were also some dis- crepancies between drawings and descriptions of early stages in their paper. The purpose of the present paper is to redescribe larvae of these three species and to summarize characters for differentiating between the three species. In addition, notes are given on time and place of larval collections and on separation of larvae of these three species from those of other marine sciaenids of the southeastern United States. METHODS of North Carolina. Those from continental shelf waters were collected with Boothbay neuston nets (mouth 1 m high x 2 m wide, mesh size 0.947 mm, tow velocity 2.6 m/s), MARMAP neuston nets (mouth 0.5 X 1.0 m, mesh size 0.505 mm, tow velocity 1 .0 m/s), and 60 cm bongo nets ( mesh size 0.505 mm, towing velocity 0.8 m/s) towed in a double oblique pattern between surface and bot- tom or 200 m depth. Specimens from South Carolina estuaries were collected with 0.5 m diameter conical nets (mesh size 0.571 mm, towing velocity 1.3-1.5 m/s) towed at surface or bottom. South Carolina tidal passes were sampled with 1.0 m mouth diameter plankton nets (mesh size 0.571 mm) moored to bridges or piers and fished near bottom for 1 h at early or middle flood tide. Speci- mens from the Cape Fear River estuary were col- lected with 1.0 m mouth diameter conical nets (mesh size 0.760 mm) towed at surface at 0.5 m/s. The number of samples available from each area except the Cape Fear River estuary by month (Ta- ble 2) provides an estimate of seasonal and areal effort distribution for comparison with data on time and place of capture of larvae. Tidal pass sampling was not carried out from August to January, and estuarine samples from August to December were not available. All specimens were preserved in 2% formaldehyde buffered by saturating with borax. Specimens from continen- Approximately 50 specimens of each species were examined. Descriptions are based on the fol- lowing numbers of specimens: silver perch, Bair- diella chrysoura, 21; banded drum, Larimus fas- ciatus, 21; star drum, Stellifer lanceolatus, 26. Specimens on which descriptions were based were collected from continental shelf waters of the South Atlantic Bight, estuaries and tidal passes of South Carolina, and the Cape Fear River estuary ^Jannke, T. E. 1971. Abundance of young sciaenid fishes in Everglades National Park, Fla., in relation to season and other variables. Univ. Miami Sea Grant Tech. Bull. 11:1-128. Table 2. — Numbers of plankton samples from South Carolina estuaries (1974), South Carolina tidal passes (1976), and the South Atlantic Bight continental shelf (1973-76) that were sorted for larval Sciaenidae. Estuar les Tidal passes Continen Neuston ital shelf Month Surface Bottom Bongo Jan. 33 30 30 30 Feb. 17 14 2 30 30 Mar. 17 14 1 47 47 Apr. 33 28 1 52 38 May 19 17 2 48 48 June 17 17 5 — — July 16 16 8 — — Aug. — — — 40 37 Sept. — — — 39 1 Oct. — — 10 11 Nov. — — — 31 16 121 FISHERY BULLETIN: VOL. 78, NO. 1 tal shelf waters were initially fixed by immersing net cod ends in 8% formaldehyde for 2 min im- mediately following net washdown. Measurements were made on the left side of the body, by ocular micrometer on a dissecting micro- scope. All measurements were made along or per- pendicular to the body midline. Measurements are defined as follows: Notochord length (NL) — symphysis of upper jaw to tip of notochord (measured in preflexion larvae). Standard length (SL) — symphysis of upper jaw to posterior edge of hypurals (measured in larvae undergoing notochord flexion and in postflexion larvae). Snout length — symphysis of upper jaw to an- terior margin of eye. Eye diameter — horizontal diameter of eye. Head length — symphysis of upper jaw to pos- terior margin of opercular membrane. Preanus length — symphysis of upper jaw to posterior margin of anus. Snout to origin of spinous dorsal fin — sym- physis of upper jaw to anterior margin of first developed dorsal spine base. Snout to origin of soft dorsal fin — symphysis of upper jaw to anterior margin of first developed dorsal ray base. Snout to dorsal fin termination — symphysis of upper jaw to posterior margin of last developed dorsal ray base. Snout to anal fin origin — symphysis of upper jaw to anterior margin of first developed anal ele- ment base. Snout to anal fin termination — symphysis of upper jaw to posterior margin of last developed anal ray base. Anus to anal fin — posterior margin of anus to first developed anal element base. Snout to pelvic fin insertion — symphysis of upper jaw to anterior margin of base of pelvic fin. Depth at cleithral symphysis — vertical dis- tance between dorsal margin of body and ventral symphysis of cleithra. Depth at caudal peduncle — least vertical dis- tance between dorsal and ventral margins of body in the area posterior to the terminal dorsal and anal fin rays and anterior to the hypural bones. Fin counts include all elements of which any part (including pterygiophore) was developed. Counts were made in unstained specimens since the primary purpose of the study was to permit identification of specimens from field collections. Unless otherwise stated, lengths referred to in this paper are standard lengths. Data on occurrences of larval Larimus fasciatus in plankton tows from continental shelf waters between Martha's Vineyard, Mass., and Palm Beach, Fla., were provided by Peter Berrien (Fisheries Biologist, Northeast Fisheries Center Sandy Hook Laboratory, National Marine Fisheries Service, NOAA, Highlands, NJ 07732). Collection methods and station distribution are given in Clark et al. (1969, 1970). RESULTS Bairdiella chrysoura Morphology. Body proportions change gradually during larval development (Table 3). Body depth at the cleithral symphysis increases slightly with growth and is >30% SL in all specimens examined. Caudal peduncle depth remains con- stant through development. Preanus length in- creases from 40-45% SL in preflexion and flexion larvae to >50% SL at ^5.7 mm. Positions of the dorsal, anal, and pelvic fins remain quite constant as the fins develop, whereas the decrease in length of the anus-anal fin gap corresponds to the in- crease in preanus length. Snout length and eye diameter change little during development, whereas head length increases from 27-31% NL or SL in preflexion and flexion larvae to 35% SL in larvae 4.9 mm. Lateral and marginal preopercular spines are present throughout the series, becoming more numerous with growth until a maximum of five lateral and four marginal spines are present at 7.0-8.8 mm. A single posttemporal spine is present at 5.0-7.7 mm, and two such spines are present at 8.8 mm. Fin development. The pectoral fin is present in all specimens examined; ray development begins at 5.7 mm and 16 rays are present by 8.8 mm (Table 4). Notochord flexion occurs at 4.1-4.4 mm SL. Development of caudal rays begins at the same time as notochord flexion. The full complement of principal rays is developed soon after completion of notochord flexion. Procurrent caudal rays begin to form at 5.7 mm and an incomplete procurrent ray count is present at 8.8 mm. The soft dorsal and anal fins begin ray development at the start of 122 POWLES: DESCRIPTIONS OF LARVAL SILVER PERCH, BANDED DRUM, AND STAR DRUM Table 3. — Body proportions (percentage of NL or SL) of larval Bairdiella chrysoura. Specimens between dashed lines undergoing notochord flexion; lengths are NL above upper dashed line, SL below. Snout to Snout Snout Anus Snout Snout to Snout to NLor Eye spinous to soft to soft to to anal anal fin pelvic Body Caudal SL Snout dia- Head Preanus dorsal dorsal dorsal anal fin termina- fin depth at peduncle (mm) length meter length length origin origin termination fin origin tion insertion cleithrum depth 3 1 35 36 37 37 38 87 6.8 74 74 63 7.4 100 10.3 9.8 106 10.5 95 30 31 276 308 273 28.7 425 41 3 404 41 4 400 446 350 31 308 31 9 32 1 329 4 1 72 11.7 31.5 468 _ 54.1 77,5 15.4 622 74,8 351 8.1 44 106 12.4 372 487 42,5 61 9 83.2 177 664 81 4 372 97 44 89 11 1 30 3 44 6 40 1 57 1 76.7 21.7 663 803 33 9 7 1 48 98 11,4 31 1 45.9 39.3 57,3 78.6 16.3 62.2 73.7 - 336 6.5 49 7.9 103 373 484 38.9 57,1 825 16.7 651 802 357 365 10.3 50 8.6 10.9 352 49.2 39,9 57,8 836 179 67 1 805 359 367 10 1 57 86 11 4 35.7 48.6 386 600 829 17,1 657 800 35.7 37,1 114 57 86 114 35.7 51.4 37 1 58,6 828 114 628 81,4 37.1 37.1 114 70 70 11.6 372 57.0 372 593 825 8,1 65 1 80 2 38 4 37 2 116 75 109 10.9 358 55.4 38 1 59,8 81.5 7.6 63 78 3 358 358 10.9 7.5 87 109 369 53.3 39,1 57,6 83.7 11.9 652 79,3 35.8 358 8.7 77 106 10.6 38 3 58.5 43,6 596 829 96 68 1 808 40,4 372 106 88 93 11.2 36.4 56.1 39.3 59,8 85.0 8.4 645 79,5 39,3 36.4 10.3 Table 4. — Fin element counts of larval Bairdiella chrysoura. Specimens between dashed lines undergoing notochord flexion; lengths are NL above upper dashed line, SL below. Caudal Caudal NLor Spinous Soft Pec- prin- procur- SL(mm) dorsal dorsal Anal toral' Pelvic cipal rent 3.1 - _ _ + - - - 3.5 - - - + - - - 3.6 - - - + - - - 3.7 - - - + - - - 3.7 - - - + - - - 38 - - - + - - - 4.1 _ 14 6 + - 7^6 _ 4.4 - 15 8 + - 7-7 - 4.4 - 18 6 + - 7^7 - 4.8 _ 20 10 + - 8-7 _ 4.9 - 19 1.9 + - 9*8 - 5.0 - 1.21 1,9 * - 9-8 - 5.7 XI 21 11,9 + 3 9^8 3,1 5.7 XI 1,21 11.9 6 1,2 9-8 - 7.0 XI 1.21 11,9 11 1,5 9-8 4,3 7.5 XI 1,21 11.9 12 1,5 9-8 5,4 7.5 XI 1,22 11.9 8 1,5 9-8 4,4 7.7 XI 1,21 11,9 12 1,5 9-8 5,4 8.8 XI 1,22 11.9 16 1,5 9-8 6.5 ' - = fin present, no developed elements notochortd flexion and attain adult complements at >4.8 mm. The spinous dorsal begins development between 5.0 and 5.7 mm; spine development is rapid, with the adult complement present at 5.7 mm. Pelvic fins are first present at 5.7 mm and adult element complements are present at 2^7.0 mm. Pigmentation. Larvae are characterized by an oblique swath of internal and external pigment, paralleling the cleithrum. from nape to cleithral symphysis (Figure 1). Melanophores of several areas constitute this swath: in the musculature of the nape, on the anterior and dorsal surfaces of the visceral mass, ventral to the brain, and on the ventral body surface. In small larvae (<5.0 mm), melanophores in these areas are usually ex- panded, so that a continuous swath of pigment is formed. Occasionally melanophores may be con- tracted, but are always present in the areas listed. In large larvae (&5.0 mm), melanophores of these areas are more frequently contracted than in smaller larvae, and thickening of the body wall begins to obscure some of the swath pigment. Pigment of the ventral midline of the tail begins as a continuous row of small melanophores in the smallest larvae and develops into a characteristic sequence of melanophores with growth. About 10 melanophores are present at 3.1-3.8 mm; one of these (two-thirds of the distance from anus to notochord tip) is larger than the others. In the dorsal midline of the tail, a few specimens =£3.5 mm NL have a small melanophore dorsal to the large melanophore of the ventral midline. At 5^4. 1 mm, melanophores of the ventral row are placed as follows: one or two anterior to the anal base, one at the origin of the anal fin, one at its termination, and three or four posterior to the anal fin. In most specimens 5=7.0 mm, no pigment is present an- terior to the anal base, but the rest of the sequence remains, and small melanophores begin to appear at the bases of individual rays. Other head and visceral mass pigmentation characterizes these larvae. A melanophore is present at the angle of the lower jaw throughout the series. Pigment is present at the tip of the 123 FISHERY BULLETIN: VOL. 78, NO. 1 3.5mmNL 4.1mmSL 8.8mmSL 124 Figure l. — Larval Bairdiella chrysoura. Scale equals 1 mm. POWLES: DESCRIPTIONS OF LARVAL SILVER PERCH, BANDED DRUM, AND STAR DRUM premaxillary at ^5.7 mm and at the tip of the lower jaw at ssT.S mm. A melanophore is usually present on the medial surface of each dentary at 3.1-7.0 mm, while at 5^7.5 mm, one or several melanophores consistently occur in this position. Melanophores are present above the anterior part of the midbrain, and above the eye, at 2^7.0 mm. Melanophores are present on the surface of the midbrain at its junction with the hindbrain at 5^5.7 mm, and on the dorsal surface of the fore- brain at ^7.7 mm. Two melanophores occur in the midventral line on the ventral surface of the vis- ceral mass: one midway between the cleithral symphysis and the anus (between the pelvic fin bases when these are developed), present at 3=3.3 mm, and another on the anterioventral surface of the anus, present at 3.1-5.0 mm. A melanophore midway between these is occasionally present at 3.5-4.7 mm and always present at 224.8 mm. On the posterior surface of the visceral mass, above the anus, a melanophore is present at 3=4.9 mm; this melanophore becomes increasingly branched and dark at &5.7 mm. Body surface pigment increases in extent in late larvae (^7.0 mm). This includes a cluster of melanophores in the dorsal midline anterior to the spinous dorsal fin, a group of melanophores ventral to this cluster, a group on the dorsal sur- face of the head, and a group on the lateral sur- face of the visceral mass. larval S. lanceolatus. Late larvae of this series have the broadly rounded caudal fin characteristic of 5. chrysoura (Hildebrand and Schroeder 1928; Dahlberg 1975) while late larvae of S. lanceolatus have the lanceolate caudal fin characteristic of the adult. Finally, larvae of my series are similar in major characters (the swath of pigment between head and visceral mass, midventral pigment pos- terior to the anus) to the larvae described by Kuntz (1915), which were apparently correctly identified. Spawning season and area. Larval B. chrysoura occurred in six surface and six bottom tows in May and in five surface and five bottom tows in June 1974 in South Carolina estuaries. None occurred between January and April or in July. In South Carolina tidal passes, larvae occurred in two May samples, one June sample, and one July sample, and did not occur between February and April. A single specimen was taken in continental shelf waters, in a bongo net tow made in 31 m on 8 April 1974 (Figure 2). Thus spawning appears to occur primarily in coastal and estuarine waters of the southeastern United States, at least from April through July. Spawning may occur later in the year, but no samples after July from coastal and estuarine waters were examined. Larimus fasciatus Identification of the series. This series was iden- tified as fi. chrysoura by fin ray counts, pigmenta- tion, caudal fin shape, and similarity to a pub- lished description. Fin ray counts (dorsal 21-22, anal 9) of late larvae in the series could have been attributed to Menticirrhus americanus, M. saxatilis, or Stellifer lanceolatus as well as B. chrysoura. A series o^ Menticirrhus larvae (iden- tified by presence of a single mental barbel at &9.2 mm, tentatively as M. americanus), which I have examined, is characterized by heavy and exten- sive body pigment, and the absence of such pig- ment in larvae of the series described here indi- cated that it was B. chrysoura rather than M. americanus. Heavy body pigmentation has been described for M. americanus (Hildebrand and Cable 1934) and M. saxatilis (Scotton et al. 1973). Although species identifications in those descrip- tions may not be accurate, heavy body pigmenta- tion is probably characteristic of larvae of the genus Menticirrhus. Caudal fin shape distin- guished larvae of the series here described from Morphology. Body proportions change little dur- ing development (Table 5). The larvae are deep bodied (depth at cleithral symphysis >35% SL, except for a 3.8 mm specimen). Preanus length is >50% SL in all specimens but one. Positions of the fins change little during development. Anus to anal fin distance is variable in length, <6% SL in most larvae but with a maximum value of 10.2% SL. Caudal peduncle depth increases with de- velopment, from <9% SL at «4.2 mm to >9% SL in most larger specimens. Preopercular spines are present in all larvae. Lateral spines are smaller than marginal spines, and numbers in both series increase with growrth. One or two small posttemporal spines and a low, spinous supraorbital ridge are present at ^5.5 mm. Fin development. The pectoral fins are present throughout development (Table 6). Pectoral ele- ments are first present at 3^4.0 mm; elements are incomplete at 5.9 mm, the largest larva available 125 FISHERY BULLETIN: VOL. 78, NO. 1 Table 5. — Body proportions (percentage of NL or SL) of larval Larimus fasciatus. Specimens between dashed lines are undergoing notochord flexion. Lengths are NL above upper dashed line, SL below. NLor SL (mm) Snout length Eye dia- meter Head length Preanus length Snout to spinous dorsal origin Snout to soft dorsal origin Snout to soft dorsal termination Anus to anal fin Snout to anal fin origin Snout to anal fin termina- tion Snout to pelvic fin insertion Body depth at cleithrum Caudal peduncle depth 3.0 3.2 10.4 9.6 10.4 12.0 33.8 36.1 53.2 53.0 - 50.6 68.8 - - - - 50.6 39.8 6.5 7.2 3.6 3.8 4.0 10.9 9.2 10.8 13.0 12.2 12.8 35.9 34.7 38.8 55.4 49.0 56 8 41.3 53.3 52.0 76.1 76.8 4.4 10.2 4.0 59.8 59.2 60.8 71.7 70.4 74.4 39.1 35.6 39.1 33.7 39.6 87 6.1 9.2 4.2 7.3 11.9 36.7 54.1 36.7 50.5 81.7 5,5 59.6 71.6 37.6 37,6 8.3 4.3 10.1 13.8 40.4 57.8 39.4 55.0 81.7 5,5 63.3 74.3 35.8 40.4 92 4.3 8.2 12.7 38.2 57.3 41.8 54.5 85.5 3,6 60.9 71,8 39.1 40.9 82 4.4 9.7 13.3 38.1 58.4 43.4 58.4 89.4 8,9 67,3 75.2 36.3 42.5 10,6 4.5 9.6 13.2 36.8 54,4 40.4 49.1 86.0 96 64,0 76.3 36.0 37.7 88 4.8 8.1 13.8 41.5 65,0 41.5 58.5 87.0 4 1 69,1 78.9 40.7 41 5 10,6 4.9 10.3 14.3 37.3 61.9 40.5 57.1 87.3 0.8 62.7 77,0 39.7 44,4 10,3 5.0 9.4 15.0 40.2 58,3 39.4 55.1 87.4 4,7 63.0 75.6 39.4 46,5 110 5.5 7.5 13.4 35.8 59.7 41,8 56.7 89.6 4,5 64,2 77.6 373 44,8 11,9 5.7 10.3 14.5 37.9 60.0 38.6 57.9 89.0 5,5 65,5 79.3 37.2 44.1 11,0 58 99 127 38.0 60.6 42.3 59.2 87.3 1,6 62.0 74,6 39.4 45.1 11,3 5.9 12.5 13.9 40.3 59.7 44.4 59.7 87.5 5.6 65.3 76.4 37.5 43.1 9.7 • 3rf 2i jK (^ Larimus fasciatus Bairdiella chrysoura 3i zi 31 30 2tf 28 79" Table 6. — Fin element counts in larval Larimus fasciatus. Specimens between dashed lines are undergoing notochord flex- ion. Lengths are above upper dashed line, SL below. NLor SL(mm) Spinous dorsal Soft dorsal Anal Pec- toral' Pelvic' Caudal prin- cipal Caudal procur- rent 3.0 32 - - - + + - - - 3.6 3.8 4.0 - 16 18 19 5 7 + + 10 -1- + 6-1-5 3-1-3 9-1-8 - 4.2 - 16 7 -1- -1- 8-1-6 - 4.3 - 15 6 7 + 4-1-5 - 4.3 - 20 7 6 + 8-1-7 - 4.4 IX 27 11,6 11 1.4 9 + 8 - 4,5 III 22 7 10 + 8+6 - 4,8 X 25 6 10 + 9 + 7 - 4,9 X 27 11,6 12 1,3 9+8 - 5,0 X 26 11,6 10 1,1 9 + 8 - 5.5 X 27 11,6 15 1,5 9 + 8 0,1 5,7 X 27 11,6 14 1,4 9 + 8 0,1 5.8 XI 26 11.6 16 1,5 9 + 8 1,2 5.9 IX 27 11,6 15 1,5 9 + 8 2,2 ' + = fin present, no developed elements. (adult complement 17 in nine adults, 16 in one, all from South Carolina waters). Caudal flexion is occurring in specimens of 3.6-4.0 mm. Principal caudal rays are first seen in flexion specimens and are usually complete after 4.9 mm. Procurrent caudal rays appear at 5.5 mm and are incomplete in the largest specimen available. The soft dorsal fin base is present in the smallest larva, with no discernible elements; pterygiophores are count- able at 3.6 mm and rays are consistently complete at ^4.8 mm. Dorsal spines first appear at 4.4 mm FIGURE 2. — Occurrence of larval Bairdiella chrysoura and Larimus fasciatus in South Carolina-MARMAP plankton tows in continental shelf waters of the South Atlantic Bight. Num- bers indicate numbers of Isirvae at stations. 126 POWLES: DESCRIPTIONS OF LARVAL SILVER PERCH, BANDED DRUM, AND STAR DRUM and are complete in one specimen at 5.8 mm. The anal fin is first present at 3.6 mm; the complete anal ray complement is present at ^^4.0 mm and anal spines are complete consistently at 4.9 mm. The pelvic fin bud is first present at 3.6 mm, and the complete element complement consistently present at 2=5.8 mm. Pigmentation. Characteristic pigment patterns of the brain and pectoral fin are useful for iden- tifying larval L. fasciatus (Figure 3). Melano- phores are present on the anterior surface of the forebrain, the anterior and posterior surfaces of the midbrain, the posterodorsal surface of the hindbrain, and the ventral surface of the brain posterior to the eye, throughout the available series. The midbrain pigment appears to ring the midbrain when viewed from dorsally. The pec- toral fin base and membrane are heavily pig- mented throughout the series. Pigment in the membrane, diffuse in small larvae, is present be- tween the rays when these are developed (2=4.3 mm). An expanded melanophore is present on the visceral mass just ventral to the pectoral fin base throughout the series; two or more melanophores may occur here at 2*4.2 mm. Other head pigment includes two to four melanophores on the gular isthmus between the lower jaw rami, melanophores on the preoper- culum posterior to the eye, a melanophore at the angle of the lower jaw, and one anterior to the cleithral symphysis. In the ventral midline of the visceral mass, early larvae have three melanophores: one posterior to the cleithral symphysis (between pelvic fin bases when present), one midway between cleithral symphysis and anus, and one on the antero ventral surface of the anus. At 2=3.6 mm, the anus melanophore is absent, and at ^4.5 mm, two or three melanophores may occur at the other two ventral midline locations. The anterior, dorsal, and posterior surfaces of the visceral mass are pigmented throughout the series, and at ^5.0 mm, melanophores appear and increase in numbers on the lateral surface of the visceral mass. In the ventral midline posterior to the anus, a row of six melanophores is present in the smallest larva (3.0 mm), the fifth of which, midway between the anus and notochord tip, is larger than the others. At ^3.2 mm, two melanophores occur in the ventral midline, one at the position of the large melanophore of the original series (at the posterior end of the anal base when developed) and one anterior to this (just posterior to the anterior end of the anal base when developed). In the dorsal midline, a melanophore is present anterior to the origin of the finfold or spinous dor- sal at ^ 3 .8 mm; two or three melanophores may be present here at ^4.5 mm. Two melanophores, one on either side of the midline, are present midway along the spinous dorsal base at >4.8 mm, and a similar pair of melanophores is present two-thirds of the distance along the soft dorsal base at >5.9 mm. On the lateral surface of the body, between the spinous dorsal base and the visceral mass, melanophores appear at 4.4 mm and increase in number with growth. Identification of the series. This larval series was identified as L. fasciatus by dorsal and anal fin ray counts, pigmentation, and by correspondence vdth a published description of late larval and early juvenile stages. Fin ray counts (dorsal 26-27, anal 6) observed in late larvae of this series could only have been of L. fasciatus or M. americanus (Table 1). The absence of heavy, extensive body pigmen- tation characteristic of Menticirrhus larvae indi- cated that the series described here was L. fas- ciatus rather than M. americanus. The pectoral fin pigment of the series here described is similar to that of L. fasciatus late larvae and early juveniles described by Hildebrand and Cable (1934). Al- though descriptions of early larvae in that paper are inadequate, late larvae (2=10.5 mm) and juveniles represent a coherent series apparently correctly identified. Spawning season and area. No larval L. fas- ciatus were present in samples from South Carolina estuaries or tidal passes throughout the months sampled, January-July. In MARMAP tows in shelf waters, larvae were taken in April- May 1974, August-September 1974, and Sep- tember 1975; larvae were most frequently taken on the inner two-thirds of the continental shelf and occurred from Cape Canaveral to Cape Fear (Figure 2). Information from plankton collections made by personnel of Northeast Fisheries Center Sandy Hook Laboratory (Figure 4) shows larval L. fasciatus to have been distributed across the width of the continental shelf and as far south as lat. 27°43' N. Large collections of larvae (6-18 speci- mens) were common off northern Florida and southern Georgia. Larval L. fasciatus were taken in cruises made during May, July, and October off 127 FISHERY BULLETIN: VOL. 78, NO. 1 ^C^. 5.8mmSL Figure 3. — Larval Larimus fasciatus. Scale equals 1 mm. 128 POWLES: DESCRIPTIONS OF LARVAL SILVER PERCH. BANDED DRUM, AND STAR DRUM 80" 34" 33' 32" 31 30 29" K 77 x: M O M J / J M / CHARLESTON/ ^ ^ IM M O M - MAY ■NJ J - JULY O - OCTOBER J - 1- 5 LARVAE ®- 6-18 LARVAE CAPE CANAVERAL 33f 32' 31 30 29 28' 2/ 79' t Figure 4. — Distribution of captures of larval Larimus fasciatus during plankton tows conducted by personnel of Northeast Fisheries Center Sandy Hook Laboratory off the southeastern United States (data supplied by Peter Berrien, Fishery Biologist, Northeast Fisheries Center Sandy Hook Laboratory, NMFS, NOAA, Highlands, NJ 07732). the southeast United States; larvae were absent from the cruises made in January-February. On cruises made north of Cape Lookout, N.C., by Northeast Fisheries Center Sandy Hook Labora- tory personnel, larval L. fasciatus were found as far north as lat. 36°22' N (just south of the mouth of Chesapeake Bay); larvae were approximately as widely distributed and as abundant in continental shelf waters between Cape Lookout and Chesa- peake Bay as off the southeastern United States (Berrien'). Larval L. fasciatus were collected in April to June and August to October on these "northern section" cruises. Stellifer lanceolatus Morphology. Body proportions change little dur- ing larval development (Table 7). The body is fairly deep (depth at cleithral symphysis 34-41% SL in most specimens). Preanus length, 40-50% SL through most of the series, increases to 55% SL in most late larvae ( ^ 10.2 mm SL). Fins develop at the adult positions. The anus-anal fin gap, 12-20% SL in most specimens <8 mm, decreases with an increase in preanus length in larvae >10 mm. Head length increases slightly with development to the late larval stages; snout length and eye diameter change little over the size range avail- able. Depth of the caudal peduncle increases slightly before and during notochord flexion and remains constant after flexion is complete. Small lateral and large marginal preopercular spines are present throughout the series, as are premaxillary and dentary teeth. A posttemporal spine is present at 5.1-7.8 mm; at ^10.2 mm a "scale bone" with four spinous projections is pres- ent in the posttemporal region. Fin development. The pectoral fin, present throughout the series, first has elements at 6.9 mm and has the complete ray complement consis- tently at ^14.0 mm, although the complete ray complement may be present in smaller larvae (Table 8). Notochord flexion occurs between 3.3 and 4.3 mm. Principal caudal rays are present in one preflexion larva and are consistently present during and after flexion; the adult complement is present at 5=5.5 mm. Procurrent caudal rays first appear at 5.1 mm and are complete at ^10.2 mm. Bases of the soft dorsal and anal fins are present, with no discernible elements, in two preflexion specimens, and are consistently present with de- veloped pterygiophores in flexion and postflexion specimens. Complete anal ray counts occur at s^3.3 mm, and complete anal spine complements at &5.5 mm. Dorsal ray complements are consis- tently complete at 2=5.5 mm although complete counts may occur at 4.5 mm. Dorsal spines are occasionally seen at 4.5-5.8 mm and are consis- ^Peter L. Berrien, Fishery Biologist, Northeast Fisheries Center Sandy Hook Laboratory, National Marine Fisheries Ser- vice, NOAA, Highlands, NJ 07732, pers. commun. June 1979. 129 FISHERY BULLETIN: VOL. 78, NO. 1 Table 7. — Body proportions (percentage of NL or SL) of larval and juvenile Stellifer lanceolatus . Specimens between dashed lines are undergoing notochord flexion. Lengths are NL above upper dashed line, SL below. Snout to Snout Snout Anus Snout Snout to Snout to NLor Eye spinous to soft to soft to to anal anal fin pelvic Body Caudal SL Snout dia- Head Preanus dorsal dorsal dorsal anal tin termina- fin depth at peduncle (mm) length meter length length origin origin termination fin origin tion insertion cleithrum depth 2.8 8.2 11.0 30.1 45.2 - - - - - - - 37.0 6.8 2.9 9.3 10.7 33.3 453 - 506 66.7 13.3 58.6 69.3 - 37.3 6.7 3.1 88 11.4 32.9 41.7 - 48.0 69.6 21.5 63.2 74.6 - 38.0 76 3.1 8.9 11.4 35.4 41.8 - - - - - - - 35.4 76 3.5 6.6 11.0 27.4 39.5 - - - - - - - 34.1 6.6 3.3 5.9 12.9 34.1 47.1 — 61.2 90.5 187 658 84.7 - 41.2 94 3.4 6.9 11.5 36.8 46.0 - 51.7 77.0 13.8 59.8 74.7 - 40.3 8.1 3.8 7.2 9.3 27.8 41.2 - 49.4 70.1 17.6 58.8 73.2 - 35.0 6.2 4.1 7.5 11.3 31.1 45.3 39.6 54.7 86.8 15.1 60.4 79.2 - 39.6 10,4 4.3 8.2 11.8 31.8 45.5 - 53.6 81.8 16.3 61.8 78.2 - 38.2 10.9 4.5 8.7 11.3 34.7 48.7 35.7 57.4 86.1 12.1 60.8 77.3 37.4 40.8 11.3 4.9 8.7 9.5 32.5 42.9 36.5 55.6 83.3 19.0 61.9 80.2 30.2 36.5 8,7 5.1 9.7 12.9 35.5 48.4 35.5 56.4 87.1 12.9 61.3 77.4 35.5 38.7 9,7 5.5 7.5 10.4 32.8 43.3 32.8 49.3 80.6 14.9 58.2 76.1 - 35.8 90 5.5 9.0 10.4 31 3 47.8 34.3 52.2 90.0 11.9 59.7 79.1 - 37.3 9,0 5.8 7.0 14.1 31.0 43.7 35.2 50.7 84.5 15.5 59.2 76.1 32.4 32.4 8.5. 6.2 8.0 8.0 32.0 46.7 34.7 54.7 82.7 12.0 58.7 76.0 32.0 36.0 9.3 6.9 8.3 9.5 34.5 45.2 36.9 54.7 82.1 14.3 59.5 76.2 29.7 33.3 9.5 7.4 10.0 10.0 37.8 54.4 41.1 56.7 84.4 8.9 63.3 77.8 41.1 37.8 10.0 7.6 8.6 8.6 35.4 47.3 37.6 55.9 86.0 11.8 59.1 76.3 31.1 34.4 9-7 7.8 7.4 9.5 32.6 45.3 358 53.6 83.2 12.6 57.9 79,0 30.5 35.8 9,5 10.2 8.8 9.6 38.4 55.2 40.0 59.2 85.6 88 64.0 79.2 38.4 36.0 56 13.1 10.3 9.0 38.5 57.6 38.5 58.9 85.9 6.5 64.1 78.2 39.7 35.8 9,0 13.9 9.6 7.2 39.7 54.2 38.5 59.0 85.5 97 63.9 783 36.1 33.7 9,6 14.0 9.0 9.0 37.3 55.4 36.2 54.1 85.5 8.4 63.8 78.3 385 33.7 9.6 15.1 10.0 8.9 38.9 57.8 36.7 55.5 84.5 7.7 65.5 77.7 38.9 33.3 8.9 tently present at 6.2 mm; the adult complement is consistently present at S3l0.2 mm. The pelvic fin bud is first present at 4.9 mm and is consistently present at ^6.2 mm; adult element complements are present at ^6.9 mm. Pigmentation. Pigmentation of the body pos- terior to the anus is of particular value in iden- tification of larval S. lanceolatus (Figure 5). In the ventral midline, small larvae (^=3.1 mm) have a row of five or six melanophores between the anus and the notochord tip; one or two of these, two- thirds of the distance from anus to notochord tip, are larger than the others. In larger specimens, an expanded melanophore is present two-thirds of the distance from anus to notochord tip (at the pos- terior end of the anal base when developed); this melanophore branches dorsally , often as far as the midlateral line. In some specimens (as shown. Figure 5) two expanded, branching melanophores are present at the posterior end of the anal fin base. In most specimens s^3.1 mm, a melanophore is present (at the anterior end of the anal base when developed) anterior to this expanded melanophore. One to three small melanophores are present posterior to the anal base in most specimens 3.3-6.2 mm; none are present at 6.9- 10.2 mm, and at >10.2 mm three or four melano- phores are present here. A small, faint pigment Table 8. — Fin element counts in larval and juvenile Stellifer lanceolatus . Specimens between dashed lines are undergoing notochord flexion. Lengths are NL above upper dashed line, SL below. NLor SL (mm) Spinous Soft dorsal dorsal Anal Pec- toral' Pelvic' Caudal Caudal prin- procur- cipal rent 2.8 - - - + - - - 2.9 - - - + - - - 3.1 - - - + - - - 3.1 - - - + - 2-1-2 - 3,5 - - - + - - - 3,3 _ 15 8 + _ 6-f6 _ 3.4 - 19 7 + - 4-1-5 - 3.8 - 18 8 -1- - 8-f7 - 4.1 - 17 1.8 -1- - 8-^7 - 4.3 - 19 8 -t- - 8-1-6 - 4,5 II 21 1 ,8 -1- - 9+8 - 4.9 - 20 ,8 -1- + 9-f7 - 5.1 V 19 1 ,8 + - 9-1-8 1,1 5.5 - 19 ,8 + - 8-1-7 - 5.5 - 22 1 ,9 + - 9-1-8 1,1 5.8 - 22 1 ,8 + - 9-1-8 - 6.2 VII 22 1 ,8 + + 9-^8 2,2 6.9 XI 1,22 1 ,8 7 1,4 9-f8 4,3 7.4 XI 1,23 1 ,8 13 1,5 9 + 8 4,4 7.6 XI 1,22 1 .8 7 1,3 9 + 8 5,4 7.8 XI 22 1 ,8 13 1,5 9 + 8 6,4 10.2 XI 1,23 1 .8 19 1,5 9 + 8 3,8 13.1 XI 1,22 1 .8 19 1,5 9 + 8 9,8 13.9 XI 1,22 1 ,8 15 1,5 9 + 8 9,8 14.0 XI 1.23 1 ,8 19 1,5 9+8 9,8 15.1 XI 1,21 1 ,8 20 1,5 9 + 8 9,8 fin present, no developed elements. spot is present in the midlateral line above the melanophore at the posterior end of the anal base in some specimens 3.1-5.5 mm, often connected to 130 POWLES: DESCRIPTIONS OF LARVAL SILVER PERCH, BANDED DRUM, AND STAR DRUM 2.9mm NL 4.3mmSL 6.2 mm SL FIGURE 5. — Larval Stellifer lanceolatus. Scale equals 1 mm. 131 FISHERY BULLETIN: VOL. 78, NO. 1 the dorsal branches of the expanded melanophore. A melanophore is present in the dorsal midline dorsal to the melanophore at the anal fin termina- tion in most specimens 2.9-6.2 mm. Head and visceral mass pigment is also useful in identifying larval S. lanceolatus. A large melano- phore is present on the anterior surface of the visceral mass, between the cleithra, throughout development. A similar melanophore appears on the posterior surface of the visceral mass at ^4.1 mm; this melanophore becomes extensively branched at >6.9 mm, and additional expanded melanophores appear dorsal and ventral to this one at 2^10.2 mm. In the ventral midline of the visceral mass a melanophore is present midway from cleithral symphysis to anus at 2.9-6.2 mm (between pelvic fin bases when present), and a second melanophore occurs on the anteroventral surface of the anus at 2.9-5.8 mm. In small larvae (^3.8 mm), pigment is present in the dorsal mid- line and internally, on both sides of the notochord, above the visceral mass. A characteristic pigment area at the dorsal end of the operculum, which appears to roof a cavity in this area, is present at &7.4 mm. Pigment occurs at the angle of the lower jaw at <6.2 mm and anterior to the cleithral sym- physis throughout the development. Further pigment develops in late larvae (>10.2 mm). On the body surface, this includes a scatter- ing of melanophores between the spinous dorsal and the visceral mass, four clusters of small melanophores in the dorsal midline along the dor- sal fin base, and a few internal melanophores in the midlateral line above the anal base. Small melanophores appear in the spinous dorsal mem- brane and at the tip of the caudal fin at 13.1 mm, and in the soft dorsal membrane at 15. 1 mm. Even in late larvae, pigmentation is not particularly heavy. Identification of the series. The series was iden- tified as S. lanceolatus by fin ray counts, pigmen- tation, caudal fin shape, and similarity to a pub- lished description of late larvae and juveniles. Fin ray counts of late larvae in this series (dorsal 21-23, anal 8) could be those of 5. chrysoura, M. americanus, M. saxatilis, or S. lanceolatus (Table 1). Lack of heavy extensive body pigment indi- cates that the series is not Menticirrhus . The late larvae of the series have a lanceolate caudal fin, characteristic of S. lanceolatus but not of B. chrysoura (Hildebrand and Schroeder 1928; Dalhlberg 1975). Late larvae (s=9 mm) and early juveniles of S. lanceolatus described by Hilde- brand and Cable (1934) represent a coherent series leading to a correctly identified young adult; late larvae of the series described here are similar to the late larvae of Hildebrand and Cable (1934), notably in the presence of an area of pigment at the upper end of the operculum, which appears to roof a cavity in this area. Spawning season and area. Larval S. lanceo- latus occurred in two South Carolina estuary sam- ples in June and in one in July 1974; all three samples came from bottom rather than surface tows. Tidal pass sampling yielded five samples containing larvae in June and five with larvae in July; no larvae were taken from February to May. No S. lanceolatus larvae were taken in continental shelf tows. Thus, spawning appears to occur in estuarine and coastal waters, and not in shelf wa- ters. Spawning occurs in early summer and may continue later into the year. DISCUSSION Comparisons with earlier descriptions Bairdiella chrysoura My material is in agreement with the descrip- tion of Kuntz (1915), except that fin development occurs at larger sizes in Kuntz's description than in my material, and some pigment details are dif- ferent. Kuntz's 5 mm specimen had a flexing noto- chord, 13 caudal rays, a developing dorsal fin base, and no anal base; these characters are found in my larvae of 4.3-4.4 mm. His 7.5 mm specimen is equivalent to my specimens of about 5.0 mm, hav- ing about 25 dorsal fin elements, 11 anal elements, no pelvic fin buds, and the full complement of caudal rays. Kuntz's larvae of ^7.5 mm had a melanophore in the dorsal midline above the large melanophore of the ventral midline, present in only a few specimens «3.5 mm SL in my series, and a melanophore anterior to the dorsal fin ori- gin, present in none of my specimens. These dis- crepancies may be due to Kuntz's use of fresh material while I used formaldehyde-preserved material. Shrinkage of formaldehyde-preserved larvae could account for the developmental differ- ences, and melanophores may be contracted dur- ing preservation or degraded with storage in formaldehyde. Kuntz probably used total lengths rather than standard lengths which might par- 132 POWLES: DESCRIPTIONS OF LARVAL SILVER PERCH. BANDED DRUM, AND STAR DRUM tially account for development rate discrepancies. Jannke (see footnote 6) illustrated 2.0 mm and 5.0 mm larvae identified as B. chrysoura. The 2.0 mm specimen, smaller than my smallest larva, does not resemble Kuntz's larvae at similar sizes nor my earliest specimen in pigmentation. Jann- ke's 5.0 mm specimen is probably correctly iden- tified: it has the fin counts and the characteristic cleithral pigment swath of S. chrysoura but lacks characteristic pigment of the ventral midline. Larimus fasciatus My material agrees fairly well with the only published description, that of Hildebrand and Cable (1934). The pigmented pectoral fin, em- phasized by Hildebrand and Cable (1934), would appear to confirm identification of all specimens of their series. The drawings are not in good agree- ment with the description in their text; for exam- ple, their drawing of a 4.5 mm specimen shows none of the pigment described. The pigment of the brain which I have found characteristic of L. fas- ciatus larvae was not mentioned by Hildebrand and Cable; perhaps fading due to preservation was responsible. The few body proportions given by Hildebrand and Cable (particularly the position of the anus at >50% SL) are characteristic of L. fasciatus. Stellifer lanceolatus Hildebrand and Cable's (1934) description of a series identified as S. lanceolatus was based on a mixture of that species and L. fasciatus. Early larvae ( =s3.5 mm SL) had pigmented pectoral fins and developing pectoral rays characteristic of L. fasciatus. Later larvae (^4.5 mm SL) lacked pec- toral fin and brain pigment and showed pectoral ray development only at >5.6 mm. Body depth and preanus length values of the early larvae are closer to those of L. fasciatus than of S. lan- ceolatus. Characters given by Hildebrand and Cable (1934) for separating early L. fasciatus and S. lanceolatus were preopercular spination, mouth shape, maxillary length, and amount of space around the brain. However, my observations indi- cate that none of these characteristics are suitable for separating these species. As with L. fasciatus, there are discrepancies between text and illustrations in the description of Hildebrand and Cable (1934), particularly in the early stages. My material agrees fairly well with that of Hildebrand and Cable (1934) at ^4.5 mm SL, although I observed a dorsal midline melano- phore dorsal to the termination of the anal fin in specimens s=7.0 mm SL, which was not indicated by those authors. Spawning Seasons and Areas Bairdiella chrysoura Spawning is reported by various authors to occur in late spring and summer on the east coast of the United States and the Gulf of Mexico, and year-round in South Florida. The season appears to begin later and to be shorter at higher latitudes: June to August off New Jersey (Welsh and Breder 1923), May to July in Delaware Bay (Thomas 1971) and Chesapeake Bay (Hildebrand and Schroeder 1928; Joseph et al. 1964), April to Au- gust at Beaufort, N.C. (Kuntz 1915), April to May off Georgia (Dahlberg 1972), and April to Sep- tember off Louisiana (Sabins 1973). Year-round spawning with peaks in January to February and April to June is reported in South Florida (Jannke see footnote 6). Spawning occurs at least April to July in South Carolina waters, according to data presented in the present study. Spawning reportedly occurs primarily in es- tuarine and coastal waters, and this is indicated by my data also. Hildebrand and Cable (1930) reported captures of eggs and early larvae in es- tuaries and to 19-24 km offshore off Beaufort, N.C, but the reliability of their identifications is uncertain. Their descriptions of eggs and early larvae of 5. chrysoura are insufficiently detailed to ensure separation from those of other sciaenid and perciform fishes. Jannke (see footnote 6) and Sabins (1973) judged from the small size of larvae caught in tidal passes that spawning must have occurred nearby in estuarine or coastal waters. Specimens I examined were mostly taken in South Carolina estuaries and tidal passes, with only one specimen coming from continental shelf waters. Larimus fasciatus The information I have presented and the lim- ited literature reports available indicate a long spawning period, extending at least from May to October, for L. fasciatus. Although larvae were more abundant in MARMAP tows made from Au- gust to September than tows made from April to May, larvae were abundant in NMFS Sandy Hook 133 FISHERY BULLETIN: VOL. 78, NO. 1 Laboratory plankton tows made in May and in July. Hildebrand and Cable (1934) took specimens of length < 5 mm from July to October off Beaufort, but presence of small juveniles in these months indicated spawning began in May. Larvae were taken from April to October in plankton tows be- tween Chesapeake Bay and Cape Lookout, N.C. (Berrien et al. 1978). Spawning apparently occurs in continental shelf waters. I obtained no larvae from South Carolina estuaries, but larvae were abundant in continental shelf plankton tows. Hildebrand and Cable ( 1934) took larvae from the coast to 22 km offshore. Larvae were common in plankton tows in continental shelf waters between Chesapeake Bay and Cape Lookout (Berrien et al. 1978). United States, L. fasciatus larvae are easily sepa- rated from all others by pigmentation, fin de- velopment sequence, and preanus distance. Fore- brain pigment and pectoral fin pigment are not present in early larvae (larvae with dorsal and anal fin rays undeveloped or incompletely de- veloped) of other sciaenids of the area, and pig- ment on the anterior surface of the midbrain ap- pears earlier than in other sciaenids of the area. Pectoral fin rays begin development earlier than in other sciaenids of the area, in fact earlier than in larvae of most known teleosts. The preanus distance of >50% SL is greater than in other sci- aenid larvae of the area. Bairdiella chrysoura and Stellifer lanceolatus Stellifer lanceolatus My data point to spawning at least in June and July in South Carolina waters; later spawning may occur but no samples were available from the second half of the year. Hildebrand and Cable ( 1934) reported presence of small larvae from July to September, but their small "S. lanceolatus" were probably L. fasciatus so this report may not be accurate. Welsh and Breder (1923) reported spawning in late spring and early summer on the U.S. east coast, and Dahlberg (1972) reported May to September spawning off Georgia. Fahay (1975) reported a 28.2 mm SL specimen taken in October off Florida. Spawning in coastal and estuarine waters rather than continental shelf waters was indi- cated by my observations. Hildebrand and Cable (1934) reported small larvae from the coast to 22 km offshore, but again these larvae may have been misidentified. Larvae have been taken in Georgia estuaries (Berrien^). Fahay's (1975) single speci- men was from inshore 7.5 km south of Cape Canaveral; being relatively large, this specimen could have originated in another spawning area. Comparisons With Other Larval Sciaenidae Larimus fasciatus Although superficially similar to larvae of sev- eral other marine sciaenids, of the southeast These species are treated together because they are quite similar as larvae, and resemble larvae of other species, notably Cynoscion regalis (Pearson 1941) and Cynoscion nothus (Stender^). Typical larvae of B. chrysoura have a well-developed swath of pigment from nape to cleithral sym- physis, which is not found in larvae of the other species; however, melanophores of the swath may be contracted or faded by preservation. Pigment of the ventral midline posterior to the anus is the most reliable character for separation of B. chrysoura from S. lanceolatus larvae. Both have a melanophore at the posterior end of the anal fin base; however, B. chrysoura has a melanophore anterior to the anal fin base (at 4. 1-7.0 mm SL) and a melanophore at the anterior end of the anal base, while S. lanceolatus has no melanophore anterior to the anal base but has a melanophore just pos- terior to the anterior end of the anal base (at the base of the second anal spine when this is developed). Larvae of the two Cynoscion species mentioned can be separated from larval B. chrysoura and S. lanceolatus by careful attention to pigment of the midventral line (Stender see footnote 9; pers. ob- serv.). Identification of small larvae with unde- veloped anal fin bases may be difficult, since the characteristic pigment sequences develop (from a row of small melanophores) at about the same time as anal-base development. Presence of a melanophore in the dorsal midline, above the posterior end of the anal base, in most S. lan- 8Peter L. Berrien, Fishery Biologist, Northeast Fisheries Center Sandy Hook Laboratory, National Marine Fisheries Ser- vice, NOAA, Highlands, NJ 07732, pers. commun. May 1975. 134 9Bruce W. Stender, Biologist, South Carolina Wildlife and Marine Resource Division, P.O. Box 12559, Charleston, SC 29412, pers. commun. February 1978. POWLES: DESCRIPTIONS OF LARVAL SILVER PERCH, BANDED DRUM, AND STAR DRUM ceolatus 2.9-6.9 mm SL may also assist in separat- ing the species; such a melanophore is present in only a few B. chrysoura at ^3.5 mm SL. Cynoscion regalis has a single melanophore in the dorsal midline above the anal fin throughout development. ACKNOWLEDGMENTS I am grateful to Malcolm H. Shealy, Jr., South Carolina Estuarine Survey Program, Charles Farmer and Charles Boardman, South Carolina Crustacean Management Program, and Ronald Hodson, North Carolina State University, for permission to use sciaenid specimens from their collections, and to Peter L. Berrien, Northeast Fisheries Center Sandy Hook Laboratory, for data on Larimus fasciatus collections off the south- eastern United States. Bruce W. Stender has been a stimulating collaborator on studies of larval sci- aenids and carefully reviewed the manuscript. Paul A. Sandifer, Charles A. Barans, and Peter Berrien also reviewed the manuscript. Hope Mix- son sorted larvae from samples. Kathleen Meuli and Lyne Imbeau typed the manuscript. To all many thanks. This research was supported by the National Marine Fisheries Service MARMAP (Marine Resources Monitoring, Assessment and Prediction) Program under contract number 6-35147. LITERATURE CITED ANDERSON, W. W. 1968. Fishes taken during shrimp trawling along the South Atlantic Coast of the United States, 193 1-35 . U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 570, 60 p. Bailey, R. M., J. E. Fitch, E. S. Herald, E. a. Lachner, C. C. LINDSEY, C. R. ROBINS, AND W. B. SCOTT. 1970. A list of common and scientific names of fishes from the United States and Canada. 3d ed. Am. Fish. Soc. Spec. Publ. 6, 149 p. BERRIEN, P. L., M. p. FAHAY, A. W. KENDALL, JR., AND W. G. Smith. 1978. Ichthyoplankton from the RV Dolphin survey of con- tinental shelf waters between Martha's Vineyard, Mas- sachusetts and Cape Lookout, North Carolina, 1965- 66. U.S. Dep. Commer., NOAA, Sandy Hook Lab. Tech. Serv. Rep. 15, 152 p. CHAO, L. N. 1978. A basis for classifying western Atlantic Sciaenidae (Teleostei: Perciformes). U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circ 415, 64 p. CLARK, J., W.G.Smith, A. W.KENDALL, Jr., andM. P. Fahay. 1969. Studies of estuarine dependence of Atlantic coastal fishes. Data Report I: Northern section. Cape Cod to Cape Lookout. R.V. Dolphin cruises 1965-66: Zooplankton vol- umes, mid-water trawl collections, temperatures and salinities. U.S. Fish Wildl. Serv. Tech. Pap. 28, 132 p. 1970. Studies of estuarine dependence of Atlantic coastal fishes. Data Report \\: Southern section, New River Inlet, N.C., to Palm Beach, Fla. R.V. Dolphin cruises 1967-68: Zooplankton volumes, surface-meter net collections, temperatures, and salinities. U.S. Bur. Sport Fish. Wildl. Tech. Pap. 59, 97 p. Dahlberg, M. D. 1972. An ecological study of Georgia coastal fishes. Fish. Bull, U.S. 70:323-353. 1975. Guide to coastal fishes of Georgia and nearby states. Univ. Ga. Press, Athens, 186 p. Fahay, M. P. 1975. An annoted list of larval and juvenile fishes captured with surface-towed meter net in the South Atlantic Bight during four RV Dolphin cruises between May 1967 and February 1968. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-685, 39 p. GINSBURG, \. 1929. Review of the weakfishes (Cynoscion) of the Atlantic and Gulf coasts of the United States, with a description of a new species. Bull. U.S. Bur. Fish. 45:71-85. Hildebrand, S. F., and L. E. Cable. 1930. Development and life history of fourteen teleosteem fishes at Beaufort, N.C. Bull. U.S. Bur. Fish. 46:383-488. 1934. Reproduction and development of whitings or kingfishes, 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. 1928. Fishes of Chesapeake Bay. Bull. U.S. Bur. Fish. 43 (part 1), 388 p. HOESE, H. D., AND R. H. Moore. 1977. Fishes of the Gulf of Mexico, Texas, Louisiana, and adjacent waters. Texas A&M Univ. Press, College Sta- tion, 327 p. Johnson, G. D. 1978. Development of fishes of the mid- Atlantic Bight, an atlas of egg, larval and juvenile stages. Volume IV. Carangidae through Ephippidae. U.S. Dep. Inter., Fish Wildl. Serv., Biol. Serv. Program, FWS/OBS-78/12, 314 p. JORDAN, D. S., AND B. W. EVERMANN. 1896. The fishes of north and middle America: a descrip- tive catalogue of the species offish-like vertebrates found in the waters of North America, north of the Isthmus of Panama. U.S. Natl. Mus. Bull. 47, 3313 p. (4 vols.) JOSEPH, E. B., W. H. MASSMANN, AND J. J. NORCROSS. 1964. The pelagic eggs and early larval stages of the black drum from Chesapeake Bay. Copeia 1964:425-434. Keiser, R. K., Jr. 1976. Species composition, magnitude and utilization of the incidental catch of the South Carolina shrimp fishery. S.C. Mar. Resour. Cent. Tech. Rep. 16, 95 p. KUNTZ, A. 1915. The embryology and larval development of Bair- diella chrysura and Anchovia mitchilli. Bull. U.S. Bur. Fish. 33:1-19. Miller, G. L., and S. C. Jorgenson. 1973. Meristic characters of some marine fishes of the Western Atlantic Ocean. Fish. Bull., U.S. 71:301-312. PEARSON, J. C. 1929. Natural history and conservation of redfish and 135 FISHERY BULLETIN: VOL 78, NO. 1 other commercial sciaenids on the Texas coast. Bull. U.S. Bur. Fish. 44:129-214 (Doc. 1046). 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. Fish Wildl. Serv., Fish. Bull. 50:79-102. POWLES, H., AND B. W. STENDER. 1978. Taxonomic data on the early life history stages of Sciaenidae of the South Atlantic Bight of the United States. S.C. Mar. Resour. Cent. Tech. Rep. 31, 64 p. Randall, J. E. 1968. Caribbean reef fishes. T.F.H. Publications Inc., Hong Kong, 318 p. SABINS, D. S. 1973. Diel studies of larval and juvenile fishes of the Caminada Pass area, Louisiana. M.S. Thesis, Louisiana State Univ., Baton Rouge, 163 p. ScoTTON, L. N., R. E. Smith, N. S. Smith, K. S. Price, and D. p. DE Sylva. 1973. Pictoral guide to fish larvae of Delaware Bay. Del. Bay. Rep. Ser. Coll. Mar. Stud. Univ. Del. 7, 206 p. SHEALY, M. H., Jr., J, V. MIGLARESE, AND E. B. JOSEPH. 1974. Bottom fishes of South Carolina estuaries — relative abundance, seasonal distribution, and length-frequency relationships. S.C. Mar. Resour. Cent. Tech. Rep. Ser. 6, 189 p. THOMAS, D. L. 1971. An ecological study of the Delaware River in the vicinity of Artificial Island. Part III. Six species of drum (Sciaenidae) in the lower Delaware River, a brackish tidal estuary. Ichthyol. Assoc. Bull. 3:1-247. Welsh, W. W., and C. M. breder, jr. 1923. Contributions to life histories of Sciaenidae of the eastern United States coast. Bull. U.S. Bur. Fish. 39:141-201. 136 REPRODUCTIVE BIOLOGY OF THE VERMILION SNAPPER, RHOMBOPLITES AURORUBENS, FROM NORTH CAROLINA AND SOUTH CAROLINA Churchill B. Grimes* and Gene R. Huntsman* ABSTRACT The vermilion snapper, Rhomboplites aurorubens, a species often associated with Caribbean reefs and banks, is an important recreational fish of the outer continental shelf of North Carolina amd South Carolina. Serial spawning occurs from late April through September off the Carolinas at depths ranging from 31 to 91 m. Most females spawn in the third or fourth year at about 205-275 mm total length. Larger, older females (age 5-10; up to 530 mm total length) appear to spawn longer each reproductive season, which may be an optimal strategy for maximizing reproductive biomass (balanc- ing the physiological costs of somatic and gonadal growth). Overall sex ratio is unequal in favor of females ( approximately 60% ) , but the ratio is 1 : 1 for small fish (less than 150 mm total length) and heavily in favor of large females (69-100?J^ for fish greater than 500 mm total length) because they live longer than males. Fecundity of first spawners is estimated at 17-42 thousand eggs, and large females produce 1.5 million eggs. The vermilion snapper, i?/iom6op/iYesai/rora6ens, is a small lutjanid which grows to 600 mm total length (TL) and 2,600 g (illustrated in Bohlke and Chaplin 1968). It occurs from Cape Hatteras, N.C., to Bermuda, southward throughout the Gulf of Mexico and Caribbean Sea to southeastern Brazil. The species is abundant, ranking second or third in weight and numbers in the Carolina headboat^ fishery (which landed between 590 and 730 metric tons of demersal fishes annually) between 1972 and 1975 (Huntsman 1976). Vermilion snapper and other reef fishes nor- mally associated with deep (>70-90 m) Caribbean reefs and banks occur in two habitats of the outer continental shelf of the Carolinas (Figure 1). The most spectacular of the habitats, the shelf break zone (Struhsaker 1969), occurs at the edge of the continental shelf (55-180 m) where the gently sloping bottom plunges abruptly downward as the continental slope. It is an area of jagged peaks, precipitous cliffs and rocky ledges associated with drowned Pleistocene reefs (Maclntyre and Milli- ' Department of Environmental Resources, Cook College and New Jersey Agricultural Experiment Station, Rutgers Univer- sity, New Brunswick, NJ 08903. ^Southeast Fisheries Center Beaufort Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 570, Beaufort, NC 28516. ^Headboats are recreational fishing vessels which charge anglers for a day's fishing on an individual, thus "per head," basis. Manuscript accepted July 1979. FISHERIES BULLETIN: VOL. 78, NO. 1, 1980. man 1970). The second habitat (inshore live bot- tom) occurs at 25-55 m and consists of broken reefs and rock outcroppings, rocky ledges, and coral patches dispersed over the continental shelf shoreward of the shelf break zone. Knowledge concerning reproduction of the ver- milion snapper is lacking. Longley and Hilde- brand (1941) reported gravid specimens about the Tortugas, Fla., in July and concluded that spawn- ing takes place in midsummer. Breder (1929) wTote that vermilion snapper probably spawn in early spring along the South Atlantic coast of the United States, and Walker ( 1950) reported spav^oi- ing off North Carolina in February. Monroe et al. (1973) collected a ripe female off Jamaica in November, and Fahay (1975) and Laroche (1977) recorded larvae off Georgia in July and August. Erdman (1976) found ripe fish from February through June in the northeastern Caribbean. In this paper we describe the seasonality, spawning frequency, sex ratio, age and size at maturity, and fecundity of the vermilion snapper and discuss possible adaptive strategies for its re- production. The study area (Cape Hatteras, N.C., to Char- leston, S.C.) was stratified by depth (i.e., inshore and offshore, the dividing depth being 55 m), and specimens were collected throughout. Most fish were obtained from the recreational fisheries throughout the Carolinas; however, some speci- mens were collected by hook and line or trawl from 137 FISHERY BULLETIN: VOL. 78, NO. 1 N.C. Figure l.— Continental shelf off North Carolina and South Carolina and important ba thy metric features that relate to the ver- milion snapper study. l)^"^ CAPE LOOKOUT / REA 1 Isobath 1 100 km the RV Onslow Bay and the RV Eastward; most juveniles were trawled from RV Dolphin. Temperature was taken by expendable bathythermograph, and photoperiod was obtained from the National Ocean Survey tide tables (U.S. Department of Commerce 1971, 1972, 1973). Specimens were weighed (nearest gram) and mea- sured (nearest millimeter). Gonads were removed, preserved in 10% Formalin"* for at least 1 wk, washed in tap water for several days, and then placed in 70% isopropyl alcohol. Frequency dis- tributions of ovum diameters were plotted by month to determine seasonality, frequency, and duration of spawning (Hickling and Rutenberg 1936; Fahay 1954). The diameters of approxi- mately 100 randomly selected ova from each of two females per month were measured to the nearest 0.05 mm by dissecting binocular microscope at 25 X . To validate measuring ova from any portion of an ovary, we determined by analysis of variance that ova sizes were distributed uniformly (indicat- ing uniform development) throughout the ovaries (Table 1). ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 138 Later, the gonads were removed from preserva- tive and weighed (nearest 0.1 g) after surface moisture was absorbed by blotting. A gonosomatic index was used as a measure of reproductive development (Finkelstein 1969) for determining spavvTiing seasonality and maturity. The index was calculated according to the formula KG =W/TL^ where KG = gonad index, W = pre- served (blotted dry) gonad weight in grams, TL = total length in millimeters. We realize that as- suming the cubic relationship is arbitrary. Quast (1968) has showm for kelp bass, Paralabrax clath- ratus, that the percentage of body weight con- tained in gonads increases with fish length. There- fore, the true exponent is undoubtedly >3, but data limitations preclude more accurate formulation. Ovaries used for fecundity studies were pre- served in modified Gilson's fixative (Bagenal and Braum 1968). The ovarian tunic was removed and washed free of adhering ova. Additional washings separated developing ova from undifferentiated oocytes and follicular material. A small subsam- ple of ova (about 1,000 or less) was stored wet for later counting in a gridded Petri dish under a binocular dissecting scope. Subsamples and origi- GRIMES and HUNTSMAN: REPRODUCTIVE BIOLOGY OF THE VERMILION SNAPPER Table l. — Analysis of variance testing the hypothesis that there is no difference in ovum diameters between anterior, pos- terior, and center sections of ovaries from three vermilion snap- per. NS = not significant. Fish no. Source of variation df MS F 1 Between sections Within sections 2 309 1 .0769 2.341 0.46 NS 2 Between sections Within sections 2 309 0.5417 0.5770 0.97 NS 3 Between sections Within sections 2 309 29.907 27.3512 1.09 NS nal ova samples (total sample minus counting subsample) were drained and dried for 24 h at 40° C. Subsample and original ova sample dry weights were determined to 0.001 g on a beam balance, and the sum of these two weights provided the total ova sample dry weight. Fecundity was determined by proportionality where: fecundity total ova dry weight number of ova in subsample subsample dry weight Fecundity models were fitted by semilog transfor- mation (log fecundity = a + 6 x length, weight, or age) and regressions are the functional regres- sions of Ricker (1973). The semilog formulation of fecundity models was used instead of more tradi- tional log-log models because they fit the data best. RESULTS Seasonality, Frequency, and Duration of Spawning Several lines of evidence indicate that spawning occurs from late spring through early fall. Males and females with ripe-appearing gonads were col- lected from late April through September, but few females were collected with ova loose within the ovarian tunic. Microscopic examination of pre- served ovaries showed three types of maturing ova present during this period (in addition to matur- ing ova, undifferentiated transparent oocytes were present and were by far the most numerous): the smallest developing ova (0.11-0.2 mm in diameter) were translucent and were the most numerous developing type; the next largest ova (0.33-0.43 mm in diameter) were nearly opaque throughout and less abundant than the preceding; the largest developing ova (0.46-0.71 mm in diameter) were typical mature teleost eggs with opaque cytoplasm occupying one pole of the egg which also contained transparent to translucent yolk material and oil globules. We observed these most mature ova only in ovaries collected from May to September, although what appeared to be ripe ovaries were also seen in April; furthermore, these mature ova occurred in only 7 of 149 ripe- appearing ovaries examined. Frequency distributions of maturing ovum diameters (Table 2) show at least two size modes of ova present from April to October (three were pre- sent in one of two June samples), while only one smaller size mode or undifferentiated oocytes were present in November, December, and March. No ova collected in January or February were examined. These data indicate spawning begins in late April or May and continues through Sep- tember or perhaps early October. Monthly mean gonad index values for 101 sexu- ally mature females, sampled from May 1972 to April 1974, also denote late spring to fall spawn- ing (Figure 2). No fish were collected in January or February 1973 or 1974; however, adult females collected in February 1975 had gonad index values of 0.51 and 0.40 which are consistent with gonad index trends indicated in Figure 2. Increasing Table 2. — Developing ovum diameter-frequency distributions (percent) from two female vermilion snapper that were examined during various months of the study. 0.08 mm interval 1972 1973 1974 (midpoint) May June July Aug. Sept. Oct. Nov. Mar. May June July Aug. Sept. Oct. Nov. Dec. Apr. 0.06 1 0.14 40 45 39 29 35 35 16 17 28 19 80 100 26 0.22 27 13 32 15 8 21 49 22 36 38 12 17 0.30 11 7 7 2 1 10 5 5 11 16 14 0.38 11 10 6 9 12 5 9 17 10 13 6 13 0.46 4 25 2 8 41 28 15 36 9 13 2 27 0.54 4 1 13 3 1 3 6 1 3 0.62 3 5 23 2 0.70 8 4 0.78 2 Total 194 198 172 214 227 179 199 209 200 197 100 100 200 Mean, mm 0.25 0.27 0.27 0.37 0.32 — — — 0.28 0.29 0.34 0.27 0.27 0.17 0.14 — 0.3 139 FISHERY BULLETIN: VOL. 78. NO. 1 < at I? s o Q o 25 ^ "^ ? 2o O I z \J I I I I I I I I \ \ \ \ / / / \. \ \ / \ / / ./ I I I I I I I I I I I -T— 1 — r I I I I 2 4 a Z < z O 1° o 5 '0 N = Wl n=34 . n=18 I I I I I I I I I I I I I I I I I I I I I I I I MJJASONOJ fMAMJJASONDJ FMA MONTHS Figure 2. — Monthly mean gonad index of female vermilion snapper collected from May 1972 to April 1974, mean bottom temperatures at collecting sites, and photoperiod (U.S. Depart- ment of Commerce 1971, 1972, 1973). monthly mean gonad index is well correlated with lengthening photoperiod and increasing bottom temperature. The seasonal occurrence of juveniles and the large size variation within the youngest age-group provides additional evidence of an extended sum- mer spawning season. During October and November 1973, several hundred juveniles rang- ing from 53 to 227 mm TL were trawled in Long Bay, N.C. and S.C., and also off Charleston. Aging of these fish from scales showed the sample to contain mostly age-groups and 1. Using the growth rate for the first year of life (Grimes 1978) for extrapolating backwards, we determined that the age fish collected in October and November 1973 were spawned throughout the summer months. Although actual spawning was never observed, it probably occurs around rough bottom from 31 to 91 m but may be more concentrated in deeper areas (55-91 m). Ripe fish were taken over rough bottom at depths of 31-91 m when bottom temper- atures were 20.6°-24.8° C. In Raleigh Bay and northern Onslow Bay, ripe fish were more abun- dant from 55 to 91 m; however, in the southern portion of the study area (southern Onslow Bay and Long Bay) ripe individuals were more equally distributed with depth. Reproductive synchrony within schools may be indicated by hook-and-line sampling. Fish were usually caught in sudden bursts of fishing activity; seldom were single individuals encountered. Gonad indices for fish of similar size caught over a short time interval (probably from the same school) were nearly identical, indicating that re- production within schools may be highly syn- chronized. Multiple spawnings each season are indicated by the relative abundance of ova types (described earlier) at different times during the spawning season (Table 2). Maturing ova were present April through October and spawning apparently takes place during this period. Early in the spawning season (May) all three developing ova types were present in considerable abundance. When ripe ova were present later in the season (June or July), fewer smaller developing ova occurred. In August and September (late in the spawning season) when ripe ova were present, smaller developing ova were absent. The total of the developing ova types may represent all that will be spawned that sea- son, and at each spawning a female develops only as many ova as her abdominal capacity will allow. This process could be repeated a number of times during the season until all eggs are spawned. Variation in gonad index during the spawning period for similar size fish may also indicate frac- tional spawning. This was evident during the spawning months of 1972 and 1973 when the gonad index of both males and females of similar size varied by as much as a factor of 12 (Figure 3). The small size of ripe gonads combined with high fecundity (see subsequent discussion) is additional evidence for fractional spawning. The mean per- cent of body weight (observed) for ovaries of ma- ture females collected during spawning months was 2.4% (0.6-5.8%, n = 40). Mature males had testes averaging 1.1% of body weight (0.4-2.4%, n = 15) during the same months. Also, we fre- quently observed semiflacid ovaries with no loose ova (perhaps partially spawned) in large adult females from June through September. Maturation Age and growth data of Grimes (1978) and our reproductive data indicate that most fish attain sexual maturity during their third or fourth years of life (186-256 and 256-324 mm TL), but a few 140 GRIMES and HUNTSMAN: REPRODUCTIVE BIOLOGY OF THE VERMILION SNAPPER o z , Males I Females 8 6- 4 2 O 4 « 2 May 1972 " ii . '\ •°. June t 10.3 July ° 3 O O I ° . August September 300 400 500 600 300 TOTAL LENGTH (mm) April 1973 • a m May • • • • . ° • . 1 _ June • • • • • • . July 1 • o - August D a D • o • 1 1 1 _ September a . , ° 3 • • 400 500 600 Figure 3. — Individual gonad index (gonad weight/( total length)^) values for vermilion snapper over the spawTiing months of 1972 and 1973 plotted against total length. precocious individuals may mature in their second year (100-186 mm TL) at about 150 mm TL. We determined age and size at maturity by examining a plot of monthly mean gonad index of females collected in the spawning season (June- September) by total length (Figure 4). Further- more, spawning season (April-September) gonad index values for males and females (Figure 3) and monthly mean gonad index for each age-group of fish (Figure 5) show that fish age 4-9 0324 mm TL) ripen earlier (April or May vs. June) and re- mained in reproductive condition longer (April to September vs. June to August) than younger spawning fish. Sex Ratio Sex ratio varies significantly from 1:1. From 1972 to 1974 we sexed 874 fish; 546 (62.5%) were females and 328 (37.5%) were males (1 df; x^ = 54.4; P<0. 001). The total sample was analyzed by YEARS OF LIFE -4—1 5 1- -7*8 — 1-9-|-10-| 6 - A 5 . / \ / \y z S 3 Z o o 2 1 / n J -109 138 163 186 213 238 263 288 313 338 363 388 413 438 463 488 513 538 563 588 25mm TOTAL LENGTH INTERVAL Figure 4.— Mean gonad index plotted by 25 mm TL intervals for female vermilion snapper during the spawning season (June- August). Approximate size at age was determined from Grimes (1978). year of collection and sex ratios were judged sig- nificantly different from 1:1 in all years (Table 3). Higher proportions of females (62.5%) were col- 141 FISHERY BULLETIN; VOL. 78, NO. 1 AGE GROUP 7 I I I I I I I I I AGE GROUP 8 I I I I I I I I I I •— "■ AGE GROUP 9 I I I I I I I AGE GROUP 10 MAR'aPr'mAy'jUn' JUL'aUg' SEP 'oCt'nOv'dEc' mar' APR 'MAY' JUN' JUL 'AUG' SEP 'OCT-NOVDEC MONTHS n = 220 Figure 5. — Monthly mean gonad index (gonad weight/(total length)^') for female vermilion snapper by age-group. Table 3. — Test of the hypothesis that sex ratios of vermilion snapper did not vary significantly from 1:1 within years of collec- tion. Item 1972 1973 1974 Percentage of females 71.9 59.4 62.5 Sample size 135 424 315 Chi-square value 25.8' 15.1* 20.8* •P 01. lected at shelf break habitats than inshore Hve bottom (60.5%), but the hypothesis that sex ratio and capture depth were independent was not re- jected (x2P>0.05, n = 852). There is a significant difference in sex ratio throughout the Hfe of the fish (as estimated by total length); however, differences within some size intervals were not judged significant (Table 4). Contingency-table analysis showed that sex ratio and size were dependent (x^P>0.05, df = 9) (i.e., with growth sex ratios were different from 1:1 and changed significantly). In small fish (101-150 mm TL) the number of males and females was nearly equal, but at 151-200 mm TL the percent- 142 Table 4. — Tests of the hypothesis that sex ratio of vermilion snapper did not vary significantly from 1:1 within 50 mm TL intervals. Total length (mm) Females Chi- (%) square Total length (mm) Females Chi- (%) square 101-150 105 49.5 00096 401-450 102 638 7.19- 151-200 117 63.4 883- 451-500 60 61 7 3.27 201-250 68 49.3 0014 501-550 58 69.2 7.69- 251-300 99 61 2 4.94- 551-600 32 893 17.28- 301-350 90 59.4 3.38 601-650 1 100 351-400 142 65.9 12.7- •P<0.05. age of females increased to about 60% , where it remained somewhat stable until 501-550 mm TL, when percentage of females began to steadily in- crease (Table 4). Only one fish >600 mm TL was collected (618 mm) and it was a female. Fecundity Estimates of fecundity ranged from 8,168 to 1,789,998 ova for 41 females ranging from 229 to 557 mm TL (3-8 yr old and 136-2,293 g). Because GRIMES and HUNTSMAN: REPRODUCTIVE BIOLOGY OF THE VERMILION SNAPPER females may spawn several times per season, fecundity estimates were from ovaries collected early in the spawning season (May and June) and all classes of maturing ova were counted. In Table 5, fecundity was separately regressed on total length (millimeters), weight (grams), and age (years) and, as expected, fecundity increases as a function of all three correlates. Fecundity increases so markedly in larger (older) fish (Figure 6) that semilog models were needed to adequately describe the relationship between fecundity and length, weight, and age. Length and weight are approximately equally good predictors of fecun- dity (r = 0.864 and 0.863, respectively). First spawners probably are about 205-275 mm TL and produce between 16,800 and 41,700 eggs. This es- timate assumes that spawning extends from late June through September for young fish (age 2-4), that first spawning occurs in the third or fourth year (186-256 or 256-324 mm TL), that scale an- nuli form in March (Grimes 1978), and that ap- proximately 25% of annual growth occurs from annulus formation to late June. Table 5. — Functional equations for fecundity in vermilion snapper. Age (A ) determined from scales. Predictof Equation Total length (nnm) Age (yr) Weight (g) F = exp(7.07 + 0.01 ari) F = exp (7 57 * 87aA) F = exp (10 21 + 0.002VV) 0863 41 0853 41 0.864 41 DISCUSSION Spawning Seasonality The conclusion that spawning occurs from late April through September is corroborated by Longley and Hildebrand's (1941) statement that spawning took place in summer around the Tor- tugas. Powles^ and Fahay (1975) reported that larvae were collected at the surface off South Carolina and Georgia in June and July, and Laroche (1977) described a larval series collected off Georgia in August. Walker (1950), however, reports collecting vermilion snapper in spawning condition off North Carolina in February. Erdman (1976) sampled 400 vermilion snapper in the northeastern Caribbean and found fish in spawn- ing condition January through June. Monroe et al. (1973) collected a ripe female in November off ^H. Powles, Fishery Biologist, South Carolina Marine Re- source Center, Charleston, S.C., pers. commun. May 1975. 20 Fecundity - e 15 o X o z < > o o z 10 7.065 + .013 TL n=41 r = .863 • • _L J_ 200 300 400 500 600 TOTAL LENGTH (mm) Figure 6. — The relationship of fecundity to total length of female vermilion snapper. Jamaica and suggested, on the basis of these and more extensive data for other reef species, that spawning probably occurs year-round in the Caribbean, but that peak spawning is in winter where surface temperature is about 26.5° C. The larvae reported on by Fahay (1975) and Laroche (1977) were collected at 27° and 26.5° C, respec- tively, and we collected ripe fish off North Carolina when sea surface temperature was be- tween 26° and 27° C. It appears that spawning of vermilion snapper off North Carolina and South Carolina is re- stricted to warm months (late April or May- September), yet spawning may occur almost year-round in the Caribbean (Monroe et al. 1973; Erdman 1976). Similar life history variations in response to local environmental conditions are re- ported by Leggett and Carscadden (1978) for American shad. 143 FISHERY BULLETIN: VOL. 78. NO. 1 Several authors report fractional spawning in marine fishes (Starck and Schroeder 1971; Beaumarriage 1973; de Silva 1973; Macer 1974). Our inference that variation in gonad index dur- ing a spawTiing month for similar size fish suggests fractional spawning is supported by Starck and Schroeder's (1971) findings on a re- lated species, Lutjanus griseus, the gray snapper. They concluded, from variation of ovary lengths and weights from fish of similar size, that spawn- ing probably occurs more than once in the same season in south Florida waters. In the results, we described three types of maturing ova and concluded that they indicate fractional spawning, yet the most mature ova type was found only in a few ripe-appearing females. Evidently final ova maturation occurs nearly simultaneously with spawTiing so that the proba- bility of catching a completely ripe fish is low. Maturation There are no published reports on maturation in vermilion snapper, but Starck and Schroeder's (1971) results on gray snapper agree closely. They wrote that females are mature at age 3 and 190- 200 mm SL. Results for vermilion snapper also agree with Starck and Schroeder's findings thatL. griseus females >375-400 mm SL probably spawn more times each year than smaller ones, and Mos- ley (1966) observed (from a sample offish 223-456 mm SL) that early in the spawning season, small- er red snapper, L. campechanus, showed less gonad development than larger ones, perhaps in- dicating earlier spawning by larger fish. Also similar to our results, Quast ( 1968) showed earlier and longer seasonal gonad maturation with growth in kelp bass. Earlier spawning by older fish can probably be explained via the interplay between somatic and gonad growth and maintenance. Sexual maturity marks diminished growth in many fishes (Hubbs 1926; Magnuson and Smith 1963; lies 1974). Female vermilion snapper older than 5 yr (390 mm TL) are beyond the years of most rapid somat- ic grov^fth (Grimes 1978) and undoubtedly can af- ford to put more energy into gonad development, even though the energy costs of maintenance are greater for ^arger fish as well. Cohen (1&76), using a theoretical mathematical model, predicts that if reproductive success de- pends upon maximizing reproductive biomass, the change in the fraction of reproductive growth (di- 144 minished somatic growth and beginning repro- ductive growth) will occur at a time and mass just prior to maximum growth rate. We used annual length increments and a length-weight relation (Grimes 1978) to derive annual increments in mass, so that we could evaluate how well vermil- ion snapper fit the optimal timing of reproduction model. The greatest annual growth increment (weight) occurs between age 6 and 7. Age 5, then, is the year of life the model predicts the growth change, and Figure 5 shows that fish age 5 and older reflect the growth change by maturing ear- lier and being mature for a longer time each re- productive season. Sex Ratio The literature on other lutjanids provides little help in interpreting our findings that sex ratios of vermilion snapper vary significantly from 1:1 overall, and throughout life (as measured by length). Camber's (1955) data on red snapper showed a greater proportion of males when small (200-400 mm TL) but a higher percentage of females among larger fish (400 mm TL). Mosley (1966) reported 56% males and 44% females among red snapper (200-400 mm TL). Bradley and Bryan (1974), however, reported a 1:1 ratio for 1,129 adult red snapper (no size range reported), and Starck and Schroeder (1971) gave a 1:1 ratio for 772 gray snapper (including small juveniles to adults). Wenner (1972) suggested several possibilities to account for unequal sex ratios (i.e., differential mortality, growth, and longevity; sex reversal; sex difference in activity; and in or out migiration from sampling area by one sex). There is no evi- dence to support any of these explanations in ver- milion snapper, except differential mortality and longevity. Our results show conclusively that rel- ative numbers of females begin to increase (to about 60%) at about 250-300 mm TL, further in- crease to about 70% at 500-550 mm TL, and even- tually reach 90% above 550 mm TL (Table 2). These results indicate that males experience great- er mortalities above 250-300 mm TL, and Grimes (1978) demonstrated greater longevity for females (no male was older than 8 yr, but females reach at least age 10). It is interesting to note that differen- tial mortality commences approximately coinci- den tally with the onset of sexual maturity. Our fecundity estimates agree reasonably well with published results for other lutjanids. Starck GRIMES and HUNTSMAN: REPRODUCTIVE BIOLOGY OF THE VERMILION SNAPPER and Schroeder ( 1971) estimated fecundity for gray snapper and found a 354 mm SL female to contain about 550,000 ova. Using length conversion equa- tions for vermilion snapper (Grimes 1978), the 354 mm SL is approximately equivalent to a 450 mm TL vermilion snapper which would contain about 410,000 ova. ACKNOWLEDGMENTS We wish to thank the staff members of the National Marine Fisheries Service (NMFS), Beaufort, N.C., and Jeflfery Ross, Virginia Insti- tute of Marine Sciences, for their assistance in collecting fish and taking of data from specimens. C. A. Barans, South Carolina Marine Resources Center, provided juvenile specimens. D. W. Ahrenholz, D. R. Colby, C. S. Manooch, NMFS, and W. E. Fahy, University of North Carolina, Institute of Marine Sciences, contributed much through helpful discussions and advice. LITERATURE CITED BAGENAL, T. B., AND E. BRAUM. 1968. Eggs and early life history. In W. E. Ricker (editor), Methods for assessment offish production in fresh waters, p. 166-198. Blackwell Sci. Publ., Oxford. BEAUMARRIAGE, D. S. 1973. Age, growth, and reproduction of king mackerel, Scomberomorus cavalla, in Florida. Fla. Dep. Nat. Re- sour., Mar. Res. Rep. 1, 45 p. Bohlke, J. E., AND C. C. G. Chaplin. 1968. Fishes of the Bahamas and adjacent tropical waters. Livingstone Publ. Co., Wynnewood, Pa., 771 p. Bradley, E., and C. E. Bryan. 1975. Life history and fishery of the red snapper (Lutjanus campechanus) in the northwestern Gulf of Mexico: 1970- 1974. Proc. 27th Annu. Sess. Gulf Caribb. Fish. Inst., p. 77-106. BREDER, CM., Jr. 1929. Field book of marine fishes of the Atlantic coast from Labrador to Texas. G. P. Putnam's Sons, N.Y., 332 p. Camber, C. L 1955. A survey of the red snapper fishery of the Gulf of Mexico with special reference to the Campeche Banks. Fla. State Board Conserv., Tech. Ser. 12, 64 p. Cohen, D. 1976. The optimal timing of reproduction. Am. Nat. 110:801-807. DE SILVA, S. S. 1973. Aspects of the reproductive biology of the sprat, Spmttus sprattus ( L.) in inshore waters of the west coast of Scotland. J. Fish Biol. 5:689-705. ERDMAN, D. S. 1976. Spawning patterns of fishes from the northeastern Caribbean. Dep. Agric, P. R., Cont. Serv. Esp. 3, 36 p. Fahay, M. p. 1975. An annotated list of larval and juvenile fishes cap- tured with surface- towed meter net in the South Atlantic Bight during four RV Dolphin cruises between May 1967 and February 1968. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-685, 39 p. Fahy, W. E. 1954. The life history of the northern greenside darter, Etheostoma belennioides blennioides Rafmesque. J. Elisha Mitchell Sci. Soc. 70:139-205. FINKELSTEIN, S. L. 1969. Age at maturity of scup from New York waters. N.Y. Fish Game J. 16:224-237. Grimes, C. B. 1978. Age, growth, and length-weight relationship of vermilion snapper, Rhomboplites aurorubens, from North Carolina and South Carolina waters. Trans. Am. Fish. Soc. 107:454-456. HICKLING, C. F., AND E. RUTENBERG. 1936. The ovary as an indicator of the spawning period in fishes. J. Mar. Biol. Assoc. U.K. 21:311-316. HUBBS, C. L. 1926. The structural consequences of modifications of the developmental rate in fishes, considered in reference to certain problems in evolution. Am. Nat. 60:57-81. Huntsman, G. R. 1976. Offshore headboat fishing in North Carolina and South Carolina. Mar. Fish. Rev. 38(3):13-23. ILES, T. D. 1974. The tactics and strategy of growth in fishes. InF. R. Harden Jones (editor). Sea fisheries research, p. 331- 345. Wiley, N.Y. Laroche, W. a. 1977. Description of larval and early juvenile vermilion snapper, Rhomboplites aurorubens. Fish. Bull., U.S. 75:547-554. LEGGETT, W. C, AND J. E. CARSCADDEN. 1978. Latitudinal variation in reproductive characteris- tics of American shad [Alosa sapidissima): evidence for population specific life history strategies in fish. J. Fish. Res. Board Can. 35:1469-1478. Longley, W. H., and S. F. HILDEBRAND. 1941. Systematic catalogue of the fishes of Tortugas, Fla., with observations on color, habits, and local distribu- tion. Pap.TortugasLab.Camegielnst. Wash. 34,331 p. Macer, C. T. 1974. The reproductive biology of the horse mackerel, Trachurus trachurus (L.) in the North Sea and English Channell. J. Fish Biol. 6:415-438. MACINTYRE, J. G., AND J. D. MILLIMAN. 1970. Physiographic features on the outer shelf and upper slope, Atlantic continental margin, southeastern United States. Geol. Soc. Am. Bull. 81:2577-2598. Magnuson, J. J., AND L. L. Smith, Jr. 1963. Some phases of the life history of the trout- perch. Ecology 44:83-95. MOSLEY, F. 1966. Biology of the red snapp)er, Lutjanus aya Bloch, of the northwestern Gulf of Mexico. Publ. Inst. Mar. Sci., Univ. Tex. 11:90-101. Munro, J. L., V. C, Gaut, R. Thompson, and p. H. reeson. 1973. The spawning seasons of Caribbean reef fishes. J. Fish Biol. 5:69-84. QUAST, J. C. 1968. Observations on the food and biology of the kelp hass, Paralabrax dathratus with notes on its sportfishery 145 at San Diego, California. Calif. Fish Game, Fish Bull. 139:81-108. RICKER, W. E. 1973. Linear regressions in fishery research. J. Fish. Res. Board Can. 30:409-434. STARCK, W. A., 11, AND R. E. SCHROEDER. 1971. Investigations of the gray snapper, Lutjanus griseus. Stud. Trop. Oceanogr. (Miami) 10, 224 p. Struhsaker, p. 1969. Demersal fish resources: composition, distribution, and commercial potential of the continental shelf stocks off southeastern United States. Fish. Ind. Res. 4:261- 300. U.S. DEPARTMENT OF COMMERCE. 1971. Tide tables, high and low water predictions, 1972: east coast of North and South America, including Green- FISHERY BULLETIN: VOL. 78, NO. 1 land. U.S. Dep. Commer., NOAA, Natl. Ocean Survey, 290 p. 1972. Tide tables, high and low water predictions, 1973: east coast of North and South America, including Green- land. U.S. Dep. Commer., NOAA, Natl. Ocean Survey, 288 p. 1973. Tide tables, high and low water predictions, 1974: east coast of North and South America, including Green- land. U.S. Dep. Commer., NOAA, Natl. Ocean Survey, 288 p. Walker, E. T. 1950. Spawning records of fishes seldom reported from North Carolina waters. Copeia 1950:319. wenner, a. 1972. Sex ratio as a function of size in marine Crus- tacea. Am. Nat. 106:321-350. 146 OBSERVATIONS ON EARLY LIFE STAGES OF ATLANTIC TOMCOD, MICROGADUS TOMCOD R. H. Peterson,^ P. H. Johansen,^ and J. L. Metcalfe' ABSTRACT In southern New Brunswick, tomcod spawn in streams from late December to mid-January. The benthic eggs hatch and newly hatched larvae drift to sea in mid-March to mid-April at which time ocean temperatures are beginning to increase. Larval migration to sea is probably aided by active swimming of larvae to the surface to fill the swim bladder, which must be filled within 24 hours of hatching. Photopositivity of the larvae may assist in guiding larvae to the surface. Water content and specific gravity of eggs reared in 0%o were 2.8 mg and 1.030. Eggs reared at 10-30%o had about 2.3 mg water per egg. Specific gravity of eggs incubated in 10%o was constant for 27 days (at 2°-4° C) at 1.038, then decreased to 1.033. This decrease is associated with water uptake of 0.5-0.6 mg per egg and elimination of salt. The specific gravity of eggs incubated in 20%« declined linearly from 1.044 to 1.037, associated with accumulation of 0.2 mg of water and elimination of a greater salt load. The specific gravity of eggs incubated at 30%o declined linearly from 1.049 to 1.045, associated with 0.1 mg water uptake and apparently insufficient salt elimination. Water uptake and salt excretion problems are minimized for eggs reared in freshwater, and under the experimental conditions described here. Normal development could not occur in continued exposure to 30%o. In natural spawning areas, the incubation medium is freshwater for most of the total cycle, with seawater invading the area only at extreme high tide. The salinity tolerance of tomcod eggs is compared with that of freshwater and marine fish eggs in general. Calculation of specific gravity of egg solids may prove a useful indirect way to investigate salt regulation in fish eggs. The Atlantic tomcod, Microgadus tomcod (Wal- baum), is an anadromous species of coastal streams from Newfoundland to Virginia. Adults ascend the lower reaches of southern New Brunswick streams in December and January. These spawning migrations form the basis for a recreational ice fishery in some larger rivers. An annual commercial catch of about 200 t is said to be taken from inshore waters of the northwest Atlantic (Scott and Grossman 1973). Local dip net fishermen take numbers of spawners for both human and animal consumption. Details of the life history of the early stages (e.g., time of hatching, time of descent into saltwa- ter) have been little studied. Leim (1924) observed that eggs would hatch in freshwater or saline wa- ter, but larvae would survive only in saline water. Booth ( 1967) found sperm motility to be maximal in low salinities, and that salinities of 0-15%o per- mitted the highest percentages of eggs to develop 'Fisheries and Environmental Sciences, Fisheries and Oceans Canada, Biological Station, St. Andrews, NB, E0G2X0, Canada. ^Fisheries and Environmental Sciences, Fisheries and Oceans Canada, Biological Station, St. Andrews, New Brunswick; pre- sent address: Department of Biology, Queen's University, Kingston, ON K7L 3N6, Canada. Manuscript accepted September 1979. FISHERY BULLETIN: VOL. 78, NO. 1, 1980. to the blastula stage. Howe (1971) described the food habits and growth rates of young tomcod in the Weweantic River estuary, Mass. The early stages of tomcod development have not been studied extensively; therefore, field studies were performed to obtain information on spawning habitat, rates of egg development, and timing of larval descent to saltwater. Tomcod eggs are deposited in areas subject to variable salinities, so the embryonic development and water balance of tomcod eggs reared in several salinities were also investigated to see how the responses of this species compare with those of freshwater and marine species. METHODS Field Studies The mouth and estuary of Frost Fish Creek (frost fish is a local name for tomcod) were chosen as a study area because the stream hosts a large and regular spawning migration of tomcod which is undisturbed except for some local dip net fishing. It is a small stream (2-4 m wide) forming a common estuary with the Digdeguash River in 147 FISHERY BULLETIN: VOL. 78, NO. 1 southwestern New Brunswick, with a midsummer discharge ca. 80 1/s and a drainage area of 570 ha (Symons and Martin 1978; Symons and Harding^). The drainage basin is typical spruce-fir boreal forest with no human habitation. Portions of it farther upstream have been recently logged. Tomcod spawn in a 10 to 1 5 m stretch at the head of tide in Frost Fish Creek (Figure 1). This area is freshwater for most of the tidal cycle, but has a variable bottom salinity (depending upon the height of the particular tide) during high tide. Extreme neap tides do not invade the spawning area. The stream substrate in the spawning area varies from ledge to boulders and cobbles. Most of the eggs settle in substrate interstices. ^Symons, P. E. K., and G. D. Harding. 1974. Biomass changes of stream fishes after forest spraying with the insec- ticide fenitrothion. Fish. Res. Board Can. Tech. Rep. 432, 47 p. Fisheries and Environmental Sciences, Fisheries and Oceans Canada, Biological Station, St. Andrews, NB EOG 2X0. 7oo -* I .0%o 13.5 7oo 23.5 7oo 25.0 %o;r^ 26.5 7oo-^ \\S\\\\ v\ \ Edge of Trees =p^epth 12 cm Marsh ;^v'v Spawning Area ;^ Depth = 60 cm 3 Metres Rood Culvert Depth Gauge 1977-78 Drift Sampler 977-78 1976-77 rSolt^: Marsh^ .\\\S\V\S Figure l. — Diagram of tomcod spawning area in Frost Fish Creek in the Digdeguash River estuary. New Brunswick. Depths and salinities are for a "typical" high-tide situation. Salinities were measured at the stream bottom. Hatched area indicates spawning area. 148 Drift samples were installed downstream of the area of egg deposition (Figure 1) near cessation of spawning (26 Dec. to 2 Jan.) to sample egg and larval drift. The samplers consisted of a galvanized-metal funnel, the narrow opening (5 x 20 cm) facing upstream, with a cloth bag attached to the downstream end (10x20 cm). The eggs and larvae accumulated in a 250 ml plastic beaker, with a screened, 2.5 cm diameter hole in one side, clamped to the bag. The sampler was threaded onto an iron rod driven into the stream bottom. A meter stick was installed to measure stream water levels, and stream salinities were measured with a salinity meter. Stream temperatures and drift samples were taken twice weekly at low tide, with the numbers of eggs collected averaged on a per- day basis. Eggs and larvae sampled were pre- served in 10% Formalin^ or 70% ethanol, those preserved in Formalin being cleared later (Galat 1972) to determine degree of development. One sample of eggs was taken from the area of egg deposition with a Surber sampler in January 1977 to see if development of drifting eggs was the same as those that were not. Egg Collection Adults (1 female:2 males) anaesthetized in MS-222 were stripped of eggs and milt in the field. Immediately, the eggs were fertilized by the "dry" method and were washed with stream water 30 s after mixing (temperature at fertilization near 0° C). This water was fresh and was taken from a part of the stream where tomcod were spawning at the time (although spawning may continue into high tide conditions when the water would be of vari- able salinity). After 1 min the water was changed, and the bottle of eggs was packed in ice and trans- ported to the laboratory. The eggs were trans- ferred to the various incubation salinities 30 min after fertilization. The eggs are weakly adhesive initially, but this adhesiveness disappears if the eggs are separated. Laboratory Studies Eggs were incubated in columns of PVC pipe and fittings holding 190 ml of water (Figure 2). Screened floors and lids retained eggs and larvae. Water flowed through the columns at 100 ml/min "Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. PETERSON ET AL.; EARLY UFE STAGES OF ATLANTIC TOMCOD Water Level in Tank Water Level in Compartment I I 5 cm Figure 2. — Diagram of incubation chambers used to rear tom- cod eggs. .9!^ 4- 2- . - 30 7oo o- 10 %o • - Fresh Water 30 %. ■^«efi FW. 10 20 ?.*Ssf°o "o*^?*"- .^.^?i ■■•^;^«»*'"°o _L 10 20 30 40 50 Time ( Days after fertilizotion ) 60 (range, 91-111). The columns were immersed in a freshwater bath cooled (2°-4° C) by recirculation of water to refrigerated header tanks. Water flowing to the columns passed through titanium coils in the bath. Water temperature decreased from 4.5° C at the beginning of January to 2.0° C in mid- February (40-d postfertilization), then increased to 2.5° C by the end of February (Figure 3). Eggs were incubated in salinities of (2 columns), 10.1 ±0.3 (1 column), 20.2 ±0.6 (1 column), and 30%o (2 columns). About 250-300 eggs were incu- bated in each column. Temperature and salinities, by conversion of specific gravity of water with Knudsen's (1962) hydrographical tables, were measured daily. Columns were checked for egg and larval mor- talities every 2-3 d. Every third day, three eggs were removed from each salinity and preserved in 10% Formalin for subsequent study of degree of development. About 100 newly hatched larvae were measured ( ±0.1 mm) from each salinity and the percentage of deformed larvae noted. Water content of 10 eggs (combined) from each salinity was measured every fifth day by measur- ing loss of weight after drying for 16 h at 40° C under vacuum. Specific gravities (sp. gr.) of eggs were measured by glycerol flotation at 10° C as described by Peterson and Metcalfe (1977). Specific gravity of egg solids was calculated from Figure 3. — Developmental stages and incubation temperatures for laboratory experiments on tomcod development. Upper panel: Appearance of developmental stages of tomcod eggs incu- bated in freshwater. Lower panel: Incubation temperatures for tomcod eggs at various incubation salinities. Arrows indicate median hatching dates at each salinity. total sp. gr. and water content where solid sp. gr. = egg dry vvrt/(egg vol. - vol. HgO), egg vol. = wet wt/sp. gr., and vol. H^O = water content/sp. gr. of HgO. Egg diameters were measured microscopi- cally to the nearest 10 ^tm. Buoyancies of newly hatched larvae, due to air content of the swim bladder, were measured by a Cartesian diver technique (Saunders 1965). Photic responses of larvae were observed by placing groups of five larvae in a Petri dish, half of which was painted black, on a finger bowl full of ice. Uniform overhead illumination was used. Statistical Procedures Differences in water content and dry weights among eggs incubated in the various salinities were tested by one way ANOVA with individual differences detected by means of Duncan's Multi- ple Range Test. Changes in water content and dry weights during larval development were analyzed by linear regression methods. 149 FISHERY BULLETIN: VOL. 78, NO. 1 RESULTS AND DISCUSSION Developmental Stages To assess development of eggs under natural conditions, a series of embryological stages was constructed (Table 1; Figure 3, lower) based on systematically sampled, laboratory-reared eggs. We attempted to make them consistent with those published previously for other species (e.g., Bon- net 1939 for Atlantic cod), although comparisons were difficult in more advanced embryos. For example, Atlantic cod lack a well-differentiated lower jaw at hatching, but it is well developed in tomcod. The stages are also referred to comparable figures in Hardy (1978:278-289) where possible (Table 1). Sampling eggs more frequently would have been useful in some instances; e.g., many anatomical features appeared between days 10 and 13, and are grouped into stage 11. The earliest stages were missed by taking the first sample at 24 h. Stages 3-6 were observed from field samples. Table 1. — Summary of development stages and day of first appearance of anatomical features for tomcod eggs incubated at different salinities. For temperature regime see Figure 3. The stages are for eggs developing in freshwater. Stages 3-5 and 9 were observed in field-collected material only. Day of first appearance at Corresponding stage designa- tion by Hardy (1978) stage Description 0%. 10%o 20%c 30%o 1 Prior to first cleavage <1 <1 <1 <1 — 2 2 cells <1 <1 <1 <1 168B 3 4 cells <1 <1 <1 <1 — 4 8 cells <1 <1 <1 <1 168C 5 16 cells <1 <1 <1 <1 — 3 Large celled morula <1 <1 <1 <1 168E 7 Small celled morula 4 4 4 4 168H 8 Embryonic axis 7 7 7 7 168J 9 Kupfer's vesicle and first somite — — — — 168E 10 Notochord 10 10 10 10 — Optic vesicle 10 10 10 10 169G 11 Eye lens 13 13 13 16 169K Ear placode 13 13 13 16 — Pericardium 13 13 13 19 — Brain lobes differ- entiating 13 13 13 22 — Fin fold 13 13 13 16 — 12 Pectoral fin buds. axial pigmentation 16 16 16 28 170E 13 Eye faintly pigmented 19 19 19 19 170G 14 Gill slit 22 22 22 22 — Swim bladder 22 22 25 28 — 15 Nasal placodes 25 25 31 none — 16 Beginning of pro- nounced snout 31 31 31 none 171A 17 Lower jaw, moutti not opened 34 34 34 34 171B 18 Moutti can be or is open 37 37 37 — 1710 19 Pigment on lower jaw 45 46 — — 172A 20 Hatching — — — — 172A Irregularities in development were seen in the later stages of development at 30%o. The snout failed to develop normally, and the development of pectoral fin buds and brain lobes was delayed. Field Observations Largest numbers of tomcod eggs were sampled by the drift samplers (Figures 4, 5) in the 15- to 20-d period after spawTiing. The numbers corre- lated fairly well with stream water level for 1977-78 when water levels were measured ( Figure 5). Largest numbers of drifting eggs may also be related to spawning activity rather than stream water levels per se. Typical numbers of eggs col- *3 0- o. tl -10 •^""""'y Febfaoty MofCh Sea Surfoce Temp Fresh Water Temp 1000 100 Eqqs ond Larvae Larvae 30 45 60 Doys from Dec 29 Figure 4. — Movements of tomcod eggs and larvae out of Frost Fish Creek. Upper: Environmental conditions and numbers of sampled tomcod eggs and larvae are shown for the 3 mo of egg and larval stream residence in 1976-77. Freshwater tempera- tures (solid line) and sea surface temperatures (dashed line) for January- April 1976-77. Lower: Histogram of numbers of tom- cod eggs and larvae caught in drift samplers. Open bars, eggs; solid bars, eggs and larvae; hatched bars, larvae. 150 PETERSON ET AL.: EARLY LIFE STAGES OF ATLANTIC TOMCOD Joouor y Febiuory Match Figure 5.— Similar to Figure 4 but for 1977-78. Upper panel includes water levels (vertical bars). Asterisk indicates occur- rence of a winter freshet, at which time drift samplers were washed out. 20 10 - 5 - C- lected ranged from a few hundred to a few thousand per 24 h during these first 15-20 d. In mid-February, stream flow rates decreased as the precipitation accumulated in snow and ice. Typi- cal numbers of eggs sampled/24 h during this period were 10-100. Hatching occurred in March and April (Figures 4, 5). Larvae began to be cap- tured somewhat earlier in 1976-77 and were taken in greater numbers than in 1977-78. This latter phenomenon is thought to be because the sampler was totally submerged in 1977-78, whereas part of it was emergent in 1976-77. Larvae probably emigrate into saltwater near the surface im- mediately after hatching, as will be discussed in a later section and may have passed over the sub- merged sampler in 1977-78. Some of the earliest larvae may have hatched in the samplers as a result of warming on the return to the laboratory. These larvae appeared normal and viable. The hatching period in nature corresponded to rising stream water levels in late March in 1977-78. Sea surface temperatures had also risen to 3.0°-4.0° C during fry emigration. Catches of larvae termi- nated in early to mid- April of both years. The earliest stages of development obtained in the drift samplers were stage 3 and 4 eggs (Figure 6), owing to lag from spawning to sampling. By the third week of January the embryonic axis was discernible in most eggs sampled. By mid- February, eyes and body axis had become pig- mented, nasal placodes and fin fold were appear- ing, and the tail had curled past the posterior Notching 1976-77 1977 -78 16) (10) (2) (I) (6) (I2)(4) (25) (4) (12) (III (2) (4) T (20)(3I (3) I (9) (7) (2) I (3) (3) (2) I (3) T (5) (21 I (5) I (3) I 20 30 January 10 20 February 10 20 March 30 Figure 6. — Stages of embryonic development for eggs sampled from Frost Fish Creek with drift samples in 1976-77 and 1977-78. Vertical bars indicate ranges of development observed. Numbers of eggs inspected are in parentheses. Small arrows indicate samples where all eggs were at the same stage. 151 FISHERY BULLETIN: VOL. 78, NO. 1 margin of the eye. Hatching began in late Feb- ruary to mid-March. A series of relatively early stage eggs was also obtained in late February to mid-March 1978 (Figure 6), indicating a possible second spawning of tomcod in late January to early February. Laboratory Observations Survival to Hatch and Length at Hatching Tomcod egg survival to hatching was highest in freshwater (Table 2). Fifty-eight percent of the freshwater eggs hatched, compared with 50, 37, and 13% at 10, 20, and 30%o salinities, respec- tively. About half of the mortality at and 10%o occurred at about day 30 (stage 15). Above 10%o high mortality also occurred at earlier stages of development. Table 2.— Percentage survival to hatching, total larval length at hatching, and median time to hatch for tomcod eggs incubated at four salinities. Standard deviations are given for larval lengths. Item 0%o 10%o 20%« 30%o Percentage egg survival to stage 19 70 73 48 21 Percentage hatched 58 50 37 13 Mean length at hatching (mm) 7.56±0.69 7.25±0.31 6.31 ±0.44 — Number of larvae measured 165 104 85 Time to median hatch (d) 54 51 51 38 Larvae hatched in freshwater were significantly longer than those from higher salinities (7.54 mm for freshwater vs. 7.25 and 6.31 mm at 10 and 20%o, respectively). Larvae at 30%o had severe spi- nal curvature and could not be measured accu- rately. Hatching was earlier at the higher salinities. The developmental success of tomcod eggs var- ied with salinity, so various parameters associated with water balance were measured on eggs reared at 0, 10, 20, and 30%o. These parameters are all interdependent so that changes in one may result in concomitant changes in others. Specific Gravity The sp. gr. of freshwater (FW) eggs was constant throughout development (Figure 7) at 1.030, im- plying that weight and volume were not changing or that they were changing in such a way that the sp. gr. was constant. In contrast, eggs incubated at 20 and 30%o decreased in sp. gr. throughout de- 152 050 045 1.040 1035 1030 ,025 , 30 7oo 20 %o 10 7oo F W _L 20 ~ Time (Doys after 40 fertilization ) 60 Figure 7 .—Specific gravity of tomcod eggs incubated at various times from fertilization at various incubation salinities. Each point is based on the mean of measurements made on 10 eggs. Lines fitted by eye. FW = freshwater. velopment. Specific gravity at 10%o was constant for the first 25 d of incubation, then decreased linearly. The sp. gr. of water at 10, 20, and 30%o (10° C) are 1.009, 1.017, and 1.024, so that eggs were denser than the incubation medium at all salinities and by approximately the same amount. For example, the sp. gr. of FW eggs is 1.030 com- pared with 1.000 for freshwater, a difference of 0.03 sp. gr. units. The sp. gr. of 20%o eggs (extrapo- lated to time at 20%o from Figure 7) is 1.045 compared with 1.017 for the incubation medium, again a difference of 0.03 sp. gr. units. Changes in sp. gr. may be associated with changes in water content, loss of solids through metabolism, change of salt content of eggs, or a combination of these factors. Water Content Mean water content of FW eggs was 2.83 mg/egg (Table 3) with no trends throughout development. The water content and percentage water content of FW eggs for the first 27 d of development were significantly higher than the values obtained at the other incubation salinities (P<0.05, ANOVA, Duncan's Multiple Range Test). The percentage water content increased from 86.4 to 89%o over the final 25 d of egg development, attributable to de- creases in egg dry weight (Tables 4, 5). There were no significant differences among the percentages of water content of eggs reared at the three higher incubation salinities for the first 27 d PETERSON ET AL.: EARLY LIFE STAGES OF ATLANTIC TOMCOD Table 3. — Water content (milligrams per egg) and percentage water (parentheses) in tomcod eggs at various incubation salinities and days from fertilization. Each value represents 10 eggs. Sampling periods were fewer at 20 and 30%o due to earlier hatch. Days of incubation 9 12 17 22 27 27-d mean 32 37 42 47 52 Newly hatched larvae 0%o 10%o 20%« 2.86(86.7) 2.87(87.2) 2.72(86.6) 2.94(85.5) 2.78(86.1) 2.83(86.4) 2.78(86.8) 2.87(87.5) 2.86(89.4) 2.88(87.8) 2.76(89.0) 1.41(85.5) 2.19(82.3) 2.14(84.3) 2.36(81.9) 2.22(81.5) 2.34(83.9) 2.25(82.8) 240(84.5) 2.44(85.3) 266(86.0) 258(85.4) 2.91(88.2) 250(84.7) 2.48(83.5) 2.47(83.4) 2.39(83.9) 2.30(82.7) 2.43(83.6) 241(84.0) 2.44(84.5) 2.60(86.1) 2.65(85.8) 30% I 28 26 30% «-« — « — ft — ft _L J_ 10 20 30 40 50 Time from Fertilization (Doys) 60 Figure 9. — The specific gravity of egg solids (as calculated from water content and egg specific gravity) at various times from fertilization for various incubation salinities. Lines fitted by eye. 154 PETERSON ET AL.: EARLY LIFE STAGES OF ATLANTIC TOMCOD decrease with time until a minimum is attained at 27 d postfertilization. The sp. gr. then rose again, as for eggs reared at and 10%o. The decrease in egg solids' sp. gr. from fertilization to day 27 at 20 and 30%o may be due to more efficient salt elimina- tion as the embryo grows. After 27 d the decrease due to salt elimination is more than balanced by increases due to factors postulated above for FW eggs. Apparently, eggs reared at 20%o were suc- cessful in eliminating salt, because the sp. gr. of their egg solids tends to converge with that for eggs reared at lower salinities. At 30%o, however, the embryos appeared to be unable to eliminate excess salt successfully because the solids' sp. gr. never approached those for eggs reared at lower incubation salinities. The 30%o eggs hatch earlier than those incubated at lower salinities, perhaps in response to high salt concentrations. As men- tioned previously, they were abnormally de- veloped. The higher sp. gr. of egg solids at 30%o would require that 15% of the solids be excess salt (using the formula 1.27a + 1.8(1 -a) = 1.35; where a = proportion of egg solids that is not salt; 1 .8 = sp. gr. of salt, 1.27 = sp.gr. of FW egg solids, 1.35 = sp.gr. of 30%o egg solids). This amounts to about 0.08 mg. This is reasonably close to the increase in dry weight of 30%o eggs over FW eggs (0.07 mg). The decrease in sp. gr. of 30%o egg solids to its minimum of 1.33 would require the loss of 0.02 mg of salt. Newly hatched larvae in freshwater that did not have access to the water surface had an sp. gr. of 1.032, which is nearly identical to that of eggs incubated in freshwater. If larvae were permitted access to air at the water surface, the sp. gr. de- clined within 7 h to 1.01, coincident with ingestion of air into the swim bladder. Newly hatched larvae were observed swallowing air at the surface. The filling of the swim bladder was investigated further with a Cartesian diver technique. Larvae that had access to air floated at a flotation pressure (Saunders 1965) of 154 mm Hg (130-170, n = 6) (0.8 atm). Those that had been denied access to air for 24 h failed to float at 675 mm Hg (645-685, n = 15), corresponding to 0.1 atm (the greatest vac- uum attainable with the apparatus). No air was observed in the swim bladder of these larvae. These larvae were then allowed access to the sur- face overnight. When tested subsequently, none floated at 675 mm Hg, nor was air observed in the swim bladder. These latter larvae, unlike larvae with air in the swim bladder (that spend most of their time near the surface), stayed on the bottom of the container. Newly hatched larvae were photopositive as tested in a half-blackened Petri dish. In two trials, 78% (39/50) and 64% ( 16/25) larvae were observed in the lighted (unpainted) half of the Petri dish. When the dish was kept in darkness, 46% (23/50) and 36% (9/25) larvae were observed in the unpainted half. Larvae were commonly observed to aggregate near the lighted sides of rearing containers. GENERAL DISCUSSION The changes that occurred in eggs reared in various salinities will first be summarized: Eggs reared in freshwater consisted of 2.8 mg water, sufficient for the embryo's needs, being con- stant throughout development. The egg sp. gr. was also constant despite decrease in solid materials (ca. 0.1 mg) — the egg diameter should therefore decrease slightly (about a 1.7% decrease is re- quired), although this was not observed, as it is within experimental error. Eggs reared in 107oo salinity have about 2.2 mg water for the first month of development, but take up an additional 0.5-0.6 mg in the later stages of development, due to the greater water require- ments of embryonic tissue. Some of this uptake may also be associated with formation of fluid filled body cavities (Zotin 1965). This water up- take was associated with a decreased egg sp. gr. The 10%o eggs may have a slight salt load which is probably eliminated in the later de- velopmental stages. The egg dry weight declined by only 0.07 mg, and newly hatched 10%o larvae may have larger yolks than do those in 0%o (Figure 9). Eggs reared in 20%o salinity had a water content equal to or greater than that of 10%o eggs in the early stages of development, but had to tolerate a higher salt load (ca. 0.04 mg/egg) as a result. Egg sp. gr. declined throughout development due to salt elimination as the embryo developed and to accumulation of about 0.2 mg water during the later developmental stages. Advanced embryos eliminated much of the initial salt load as the egg solid sp. gr. of late stage eggs is nearly identical to that of eggs reared at lower salinities. The concept of salt elimination seems reasonable, but is subject to some uncertainty in these experiments because the solids associated with the chorion and perivitelline fluid are included in the estimates of 155 FISHERY BULLETIN: VOL. 78, NO. 1 solids sp. gr. These compartments of the egg obvi- ously would have no capacity for elimination of salts. Egg dry weight declined by only 0.04 mg at hatching as newly hatched 20%o larvae appear to have even more yolk than 10%o larvae (Figure 8). The 20%o curve in Figure 9 suggests that salt elimination began very early and increased as the embryo grew. It is probable that at least the early salt elimination had a cellular rather than organ basis. Eggs reared in 30%o salinity also had a water content similar to those reared at 10 and 20%o, but the salt load was high. Water accumulation dur- ing the later developmental stages was low (ca. 0. 1 mg). The sp. gr. of egg solids goes down over the first 27 d of development, indicating some elimina- tion of salt. The pattern during the later stages of development is strikingly different from that at 20%o in that the solids' sp. gr. again rose to about 1.37 at 37 d of incubation at which point the larva hatched. Problems with salt balance and os- moregulation may have led to the deformities and early hatching. The dry weight of 30%o eggs de- creased only slightly during development. It has been shown, for the eggs and larvae of some marine organisms, that the salinity in which fertilization and the earliest stage of development occur may influence development and growth of subsequent stages in the life history (Kinne 1962). It is therefore possible that eggs fertilized in water of higher salinity might have responded different- ly to the various experimental salinities. Booth (1967) obtained data suggesting that fertilization could occur in salinities as high as 15%o. It is nota- ble, however, that the eggs of Cyprinodon macularius in Kinne's experiments were allowed to develop 3-6 h in the spawning salinity, and at a higher temperature (27° C) than was the case for the tomcod experiments. It is probable that the eggs of C. macularius had developed further be- fore experimentation. The tomcod's early life history seems adapted to the hydrodynamics of streams in which it spawois. Spawning migrations occur in late December to early January while water levels are still high from the fall freshets. The eggs develop through- out midwinter when water levels are low, thus minimizing loss of eggs from the stream, then hatch when water levels are rising coincident with the early snow melt. The higher water levels dur- ing hatching would ensure rapid flushing of larvae into the estuary. Newly hatched larvae probably rise to the stream surface soon after hatching and 156 ingest air into the swim bladder, with possibly the positive response to light facilitating surfacing. This behavior of newly hatched larvae would also ensure rapid flushing into the estuary. The continuous drift of eggs out of the stream is somewhat puzzling. Most eggs taken in the drift samples were alive and apparently developing normally. These perhaps are eggs which had been deposited where they were likely to be taken up into suspension. In support of this suggestion, egg drift was positively correlated with stream level. Whether these eggs continue to develop would de- pend in part upon the salinity conditions where they finally settle and the ambient salinity during earlier development. Laboratory results indicate that less than full salinities are required for nor- . mal development from fertilization, but the eff"ects of variable salinities on tomcod egg development were not investigated. The tomcod egg resembled those of freshwater species (rather than marine species) in regard to salt tolerance, assuming that the responses re- ported here are typical. Eggs of brook trout exhibit increased mortality above 6%o salinity with total mortality at 12%o (Sutterlin et al.^). Species such as Abramis will hatch in salinities up to 20%o, although 2.5-5%o is optimal (Holliday 1969). With the tomcod, between 20 and 30%o salinity appears to be the upper limit for production of normal larvae. By way of contrast, eggs of several marine species have been hatched in salinities up to 60%o (cod, herring, plaice), although optima are usually in the 25-30%o range (Holliday and Blaxter 1960; Holliday 1965). Eggs of marine species tend to swell at low salinities (usually <15%o); above this salinity egg diameter is constant (Holliday 1965; Solemdal 1967). Tomcod eggs require salinities of <10%o for noticeable swelling to occur. Several parameters measured (water content, dry weight, solids' sp. gr., and egg sp. gr. for 10%o incubation) begin to change dramatically at about 27 d of incubation. In relation to embryonic de- velopment it seems probable the embryonic mass is beginning to increase dramatically at this point, resulting in the noted physiological changes. Perhaps these changes are linked to the high mor- tality occurring at this stage of development. ^Sutterlin, A. M.,P. Harmon, andH. Barchard. 1976. The culture of brook trout in salt water. Fish. Mar. Serv. Res. Dev. Tech. Rep. 636, 21 p. Fisheries and Environmental Sciences, Fisheries and Oceans Canada, Biological Station, St. Andrews, NB EOG 2X0. PETERSON ET AL.: EARLY LIFE STAGES OF ATLANTIC TOMCOD Zotin (1965) reported that eggs of freshwater teleosts (e.g., loach, zander (Lucioperca)) took up no water after water hardening until the chorion began to stretch due to weakening by the hatching enzyme. The mullet egg took up water during the second half of development during which time the perivitelline space first appeared. With the tom- cod, water uptake occurred in the latter stages of development in the three higher salinities. It is not known where this water was distributed within the egg, but it was probably incorporated into embryonic tissue. It is inferred, from calculated specific gravity of egg solids, that tomcod embryos osmoregulated to some degree, becoming more proficient as de- velopment proceeded. This may be simply a func- tion of embryonic size, resulting in more osmoregulating tissue. It has been suggested by Holliday ( 1965) that plaice embryos can regulate osmotic concentration after gastrulation, which occurs in 9 d or less in these tomcod eggs. Holliday (1969) also showed that flounder eggs could regu- late yolk sodium from fertilization. Unfortu- nately, we did not make measurements here be- fore 9 d of incubation. Holliday and Blaxter (1960) and Forrester and Alderdice (1966) observed development to proceed faster at higher salinities for herring and Pacific cod, respectively. While tomcod hatched earlier at higher salinities, there is little suggestion that development occurred more rapidly. Rather, it ap- peared that the freshwater larvae grew larger prior to hatching. Some structures were delayed, or never appeared in 30%o embryos, but this is due to abnormal development at this salinity. Abnor- mal development has frequently been recorded at abnormally high salinities. Usually the defor- mities are skeletal as are observed for tomcod, or involved body cavity deformities (Holliday 1965; Alderdice and Forrester 1971). Although the tomcod is a physoclist species, the pneumatic duct is apparently functional in the newly hatched larva. In <24 h the duct is closed, and the larva can no longer fill the swim bladder by air ingestion. Larval loss of the pneumatic duct has been implied for physoclists generally (Har- den Jones 1957). Whether or not the duct is utilized in initial filling of the bladder is apparent- ly quite variable (Johnston 1953; Schwarz 1971). ACKNOWLEDGMENTS We wish to acknowledge P. Harmon for assist- ing in some aspects of field studies; P. W. G. McMullon and F. Cunningham for performing the photography and drafting; A. Sreedharan for pro- viding statistical analyses of the data; and D. W. McLeese and M. J. Dadswell for reviewing the manuscript. LITERATURE CITED ALDERDICE, D. F., AND C. R. FORRESTER. 1971. Effects of salinity, temperature, and dissolved oxy- gen on early development of the Pacific cod (Gadus ma- crocephalus). J. Fish. Res. Board Can. 28:883-902. BONNET, D. D. 1939. Mortality of the cod egg in relation to tempera- ture. Biol. Bull. (Woods Hole) 76:428-441. BOOTH, R. A. 1967. A description of the larval stages of the tomcod, Microgadus tomcod, with comments on its spawning ecol- ogy. Ph.D. Thesis, Univ. Connecticut, Storrs, 53 p. FORRESTER, C. R., AND D. F. ALDERDICE. 1966. Effects of salinity and temperature on embryonic development of the Pacific cod (Gadus ma- crocephalus). J. Fish. Res. Board Can. 23:319-340. Galat, D. L. 1972. Preparing teleost embryos for study. Prog. Fish- Cult. 34:43-48. Harden Jones, F. R. 1957. The swimbladder. In M. E. Brown (editor 1, The physiology of fishes, Vol. II, p. 305-322. Acad. Press, N.Y. Hardy, J. D., Jr. 1978. Development of fishes of the mid-Atlantic Bight, an atlas of egg, larval, and juvenile stages. Vol. II, Anguil- lidae through Syngnathidae. U.S. Fish Wildl. Serv., Biol. Serv. Program Chesapeake Biol. Lab. FWS-OBS- 78-12, 458 p. HOLLIDAY, F. G. T. 1965. Osmoregulation in marine teleost eggs and lar- vae. 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. I, p. 293-311. Acad. 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. J. Mar. Biol. Assoc. U.K. 39:591- 603. HOWE, A. B. 1971. Biological investigations of Atlantic tomcod, Micro- gadus tomcod (Walbaum), in the Weweantic River es- tuary, Massachusetts, 1967. M.S. Thesis, Univ. Mas- sachusetts, Amherst, 82 p. JOHNSTON, P. M. 1953. The embryonic development of the swim bladder of the largemouth black bass Micropterus salmonides sal- monides (Lacepede). J. Morphol. 93:45-47. KINNE, O. 1962. Irreversible nongenetic adaptation. Comp. Biochem. Physiol. 5:265-282. KNUDSEN, M. 1962. The determination of chlorinity by the Knudsen method. G. M. Mfg. Co., N.Y., 63 p. 157 FISHERY BULLETIN: VOL. 78, NO. 1 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. PETERSON. R. H., .AND J. L. METCALFE. 1977. Changes in specific gravity of Atlantic salmon (Salmo salar) alevins. J. Fish. Res. Board Can. 34:2388-2395. Saunders, R. L. 1965. Adjustment of buoyancy in young Atlantic salmon and brook trout by changes in swimbladder volume. J. Fish. Res. Board Can. 22:335-352. Schwarz, a. 197 1 . Swimbladder development and function in the had- dock, Melanogrammus aeglefinus L. Biol. Bull. (Woods Hole) 141:176-188. SCOTT, W. B., AND E. J. CROSSMAN. 1973. Freshwater fishes of Canada. Fish. Res. Board Can., Bull. 184, 966 p. Solemdal, p. 1967. The effect of salinity on buoyancy, size and develop- ment of flounder eggs. Sarsia 29:43 1-442 . Symons, p. E. K., and J. D. Martin. 1978. Discovery of juvenile Pacific salmon (coho) in a small coastal stream of New Brunswick. Fish. Bull., U.S. 76:487-489. ZOTIN, A. I. • 1965. The uptake and movement of water in em- bryos. Symp. Soc. Exp. Biol. 19:365-384. 158 NOTES OBSERVATIONS OF SEA OTTERS DIGGING FOR CLAMS AT MONTEREY HARBOR, CALIFORNIA Although the feeding behavior of the sea otter, Enhydra lutris, is frequently observed from the surface, few underwater observations of foraging sea otters have been published. Faro (1969) and Houk and Geibel (1974) described the underwater behavior and tool use of sea otters when they re- moved abalone from rock substrates. Shimek (1977) observed a sea otter foraging for snails and presumably other invertebrates by patting the surface of rocks and feeling into cracks. Shimek also described a sea otter digging up the echiuroid worm, Urechis caupo, from a silt and cobble sub- strate. Further deductions about underwater foraging behavior have been made from collec- tions of abalone shells with the characteristic "ot- ter break" hole in the middle (Wild and Ames^) and from observations of aluminum beverage cans bitten by otters to remove octopus (McCleneghan and Ames 1976). Some sea otters can be enticed underwater to take food offered them (pers. obs). However, these latter observations of underwater food manipulation are of limited value because the otters also take items unpalatable to them (e.g., the holothuroid Stichopus californicus) , and be- cause the otters were clearly interacting with the diver observer. These accounts of underwater foraging indicate that sea otters use primarily tac- tile sensitivity of the forelimbs to locate and cap- ture prey, whereas all other marine mammals (pinnipeds and cetaceans) use their jaws to cap- ture prey. Radinsky (1968) hypothesized that the sea otter evolved forelimb tactile sensitivity sepa- rately from the aonychoid otters. The large impact of sea otters on Pismo clam, Tiuela stultorum, populations in California has been documented (Stephenson 1977; Miller et al.2), and in Prince William Sound, Alaska, 81% of the food items taken by sea otters were bivalves, especially Saxidomus gigantea (Calkins 1978). •Wild, P. W., and J. A. Ames. 1974. A report on the sea otter, Enhydra lutris L., in California. Calif. Dep. Fish and Game Mar. Resour. Tech. Rep. 20, 93 p. =*Miller, D. J., J. E. Hardwick, and W. A. Dahl- strom. 1975. Pismo clams and sea otters. Calif. Dep. Fish Game Mar. Resour. Tech. Rep. 31, 49 p. FISHERY BULLETIN: VOL. 78, NO. 1, 1980. The Alaskan otters "dug out clams with their forepaws while maintaining a head downward position" in intertidal and shallow subtidal water. However, Shimek's (1977) description is the only detailed underwater observation of sea otters foraging on soft substrate. Detailed observations of sea otters taking prey from soft substrates are more difficult than those on rock, because the ot- ter's disturbance of the bottom often results in clouds of sediment obscuring further vision. In the present account, we describe underwater ob- servations of sea otters digging clams in a silty sand substrate and present information about the impact of this foraging on the distribution and abundance of subtidal clams at Monterey Harbor, Calif. Observations In 1976-77 we observed sea otters eating large numbers of the Washington c\a,va, Saxidomus nut- talli, primarily in two specific areas of Monterey Harbor (A and B of Figure 1 ). From vantage points along the floating boat slips and elevated wharves, we observed sea otters at the surface feeding on 211 prey items: S. nuttalli (88.6%); the crabs Pugettia producta (4.2%) and Cancer sp. (3.3% — probably C. antennarius or C. productus, but not C. magister); the rock jingle hiwalve Pododesmus cepio (1.4%); and unidentified items (2.4%). Dur- ing spring 1976, as many as four sea otters were foraging at one time in the harbor vicinity, but an average of about one sea otter was observed on 38 counting trips made to the area. The underwater path of foraging sea otters could often be observed from the surface by follow- ing the trail of air bubbles escaping from their compressed fur. The paths of foraging dives made in the inner harbor were often contorted, 50-60 m or more long, and lasted 45-80 s. These dives usu- ally produced no prey, but the prey taken were mostly crabs and rarely clams. On those dives that resulted in the capture of kelp crabs, sea otters usually (eight out of nine dives) finished their search with a swim under 10-20 m of the floating docks in the inner area of the harbor. During scuba dives in this area, we repeatedly observed kelp crabs on the undersides of these floats and rarely elsewhere. It was difficult to observe the paths of 159 ^ 1 » 1 » 300 m N 1 • • • • • • • • • • "^^=^. i^ "" ^^ • • • • • Outer • • > * *i Harbor • • • , f^^ ^3 • • • • wv,. ^tULu^ ''F ^s ^ Inner B • * • • • Harbor • • • « ^^~'^^^~r~-~^ Figure l. — Monterey Harbor, Calif. High densities of clams were foraged by sea otters in areas A and B. 1 = Fisherman's Wharf; 2 = Wharf No. 2; 3 = north and east sea wails of the inner harbor; 4 = breakwater for the outer harbor. (Traced from an aerial photograph in Haderlie and Donat 1978.) dives made in the middle of the outer harbor, but sea otters usually surfaced without prey 50 m or more from the start of the dive. However, the paths of feeding dives in the two locations where clams were taken in abundance were usually quite short, only 10 m or less. The usual sequence of dives in the harbor region began with otters making one to three 50-90 s dives that produced no prey. After about 10-20 s on the surface and a little grooming, the otters usu- ally dove again to the same spot. A series of short (25-40 s) dives followed the initial dives, and each of these invariably resulted in a single S. nuttalli about 10 cm long. The otters took 30-90 s to open and eat the clams before diving to the same spot. Sometimes they pounded the clams on a rock anvil on their chest; other times they simply twisted or pried the clams open with their teeth. An average of 6 and as many as 19 clams were taken in a single series of these dives. Following such a series, ot- ters usually spent up to 30 min grooming before they swam away, sometimes to forage in a new location. In spring 1976, we conducted an underwater survey of most of the bottom of the inner and outer harbor, noting variations in the substrate and counting protruding clam siphons in haphazardly tossed y^ m^ quadrats. Depths in the harbor ranged from 2 to 8 m, with area A being 4.5 m and area B 2-3 m (Figure 1). The substrate in much of the enclosed inner harbor was black mud and silt, and most of the rest of the harbor (including areas A and B) was silty sand. The two areas where otters fed extensively on clams had high densities of clam siphons: for area A,x = 13.5/m2, SD = 8.9, n = 19; for area B,x = 9.3/m2, SD = 7.2, n = 16. The area under Wharf No. 2 adjacent to area A had even higher densities of siphons: x = ITA/m.^, SD = 11.9, n= 18. However, other areas of the harbor had siphon densities < l.O/m^, and the black mud of the inner harbor had densities <0.04/m2. ^g inserted a slender rod down siphon holes in the substrate until the rod contacted the clam shell, and, with considerable difficulty, we used a stream of freshwater from a garden hose to obtain a few 8-14 cm long clams from area A. We used these specimens to distinguish the two species present by the morphology of the protruding siphons. The species composition in areas A, B, and under Wharf No. 2 were the same: 95% wereS. nuttalli, 5% were the gaper clam, Tresus nuttallii. In this way we also determined that the clams were lo- cated 10-50 cm into the substrate and that larger individuals of both species tended to occur at the deeper end of this range in the sediment. We re- corded the densities of clam siphons in area A and also in the adjacent area of highest density under Wharf No. 2 at approximately bimonthly intervals from February 1976 to March 1977. The densities and proportions of the two species of clams did not change (ANOVA,P>0.05). The bottom in the two areas where sea otters took large numbers of clams was littered with hundreds of shells, both on the surface and mixed into the sediments. About 58% of the shells did not have pairs of connected valves, and about one- third of the valves were broken. Of 89 shells sam- pled, 99% wereS. nuttalli and 1% wereT. nuttallii. The bottom topography was hummocky in these areas, and there were many craters 0.5-1.0 m across and 10-15 cm deep. The bottom under Wharf No. 2, where the density of clam siphons was highest, was mixed with debris consisting of chunks of asphalt apparently from resurfacing of the road on the wharf and of clumps of large bar- nacle, Balanus nubiluS, tests which had fallen 160 from the massive barnacles encrusting the pihngs. There were considerably fewer craters in this area compared with the adjacent area A, and our at- tempts to dig into the substrate under the wharf proved difficult as a result of the debris embedded in the sediment. Sea otters were in the process of foraging on clams during several of the scuba dives in area A. Although these otters were not bothered by our presence under water, attempts to observe pre- cisely how they were capturing clams usually failed because they stirred up large clouds of sedi- ment that obscured all of their activity. When the otters stopped foraging and the clouds of sediment dispersed, a large hole up to 1.0-1.5 m across and 0.5 m deep had obviously resulted from their dig- ging. The sides of these holes were initially nearly vertical, but collapsed within minutes. Details of a sea otter digging for clams were observed by the first author on a single occasion on 30 March 1977, when a strong current rapidly dispersed the clouds of sediment. Upon observing a young male otter begin a typical sequence of foraging dives in area A, the observer moved along the bottom and approached the digging site from an upstream direction. The otter was clearly visi- ble at a distance of 5 m and was just leaving the bottom after completing the second longer dive of the series. He returned to the bottom within 20 s but abandoned the initial digging site, leaving a small hole about 0.5 m across and 25 cm deep. Instead, on this third dive, he moved immediately to a new spot about 4 m away and began to dig rapidly with his front paws in a fashion very much like a dog, producing a large conical cloud of sedi- ment extending downstream. Digging lasted about 45 s, followed by a 20 s surface interval. On the fourth dive the otter resumed digging in the same spot, and as during all digging periods, he faced into the current. The observer was able to approach < 1 m from the sea otter by creeping up in a prone position on the bottom while the otter substantially enlarged the hole to a short trench about 1 m long, 0.5 m across, and 25 cm deep by digging rapidly with both front paws. His back flippers were moving at a slower rate, which prob- ably helped maintain his position and also ap- peared to assist in digging. Toward the end of the digging on this dive the otter began to roll re- peatedly from side to side to enlarge the front end of the trench laterally, until he apparently en- countered a clam and suddenly surfaced for 45 s. On the fifth dive this rapid process of rolling and lateral digging with the front paws continued again for about 30 s until another clam was caught and the activity suddenly stopped. The hole at this time was over 0.5 m deep and the otter's body was entirely below the level of the substrate surface while digging. The otter used this process of lat- eral digging on three more dives lasting about 30 s each with 40-60 s surface intervals, before the observer ran out of air and surfaced. The trench at that time was over 1.5 m long and remained about 0.5 m wide and deep. The otter terminated the series of feeding dives with one additional dive while the observer was at the surface. It paid no apparent attention to the observer's close presence during the entire series. Simultaneous observa- tions by the second author from the surface indi- cated that none of the first three dives (including two dives at the first spot) produced a clam, but that each of the six subsequent dives resulted in a single clam. The otter did not use a rock to open the clams. Discussion In 1966, prior to the return of sea otters to Mon- terey Harbor, Calif., Department of Fish and Game divers made qualitative surveys of the bot- tom and used a garden hose to remove several clams from the substrate for identification. The bottom topography was smooth, clams were abun- dant, and T. nuttallii was the dominant species removed from as deep as 50 cm in the substrate (Ebert^). Follow-up survey dives soon after the return of sea otters indicated that clams were less abundant and the bottom topography was hum- mocky (Ebert, see footnote 3). Although definitive quantitative data are not available for that period, and although construction and dredging opera- tions in the inner marina portion of the harbor may have had important impact on clam popula- tions, information in the present report indicates that sea otters may have limited the abundance and distribution of S. nuttalli and T. nuttallii and that T. nuttallii is now only a minor species. The cause of this apparent shift in dominance from T. nuttallii toS. nuttalli is unclear. Our limited mea- surements of the depths of these clams in the sub- strate indicated that larger individuals were found deeper (to about 50 cm), but that neither ^E. E. Ebert, Director, Marine Culture Laboratory, California Department of Fish and Game, Granite Canyon, Coast Route, Monterey, CA 93940, pers. commun. June 1979. 161 species had a depth refuge from predation by sea otters, which excavated deeper than 50 cm. Tresus nuttallii attains larger size than S. nuttalli (pers. obs.), and if sea otters prefer larger clams, they may have preyed preferentially upon T. nuttallii. However, clams remained abundant in small areas of the harbor in spite of heavy predation by sea otters. Densities under Wharf No. 2 averaged about 17 clams/m^; and in this area they appear to have a partial refuge from sea otters, which may have found it too difficult to dig through the debris of chunks of asphalt and clumps of barnacle tests embedded in the sediment. No such impediment to digging exists in areas A and B, where clams have persisted in somewhat lower densities of about 14 and 9/m^, respectively. However, the species com- position of clams was the same under Wharf No. 2 and in areas A and B, regardless of predation in- tensity. By following tagged animals, Loughlin (1977) showed that certain sea otters made daily foraging trips to Monterey Harbor from rafting locations as far as 2 km away. In the present descriptions of their dive paths, sea otters feeding on items other than clams apparently located prey in a random manner similar to Shimek's (1977) description of an otter patting the surface of rocks and feeling the cracks. Observations of the bubble paths of otters taking clams in areas A and B of the harbor, however, indicated that they usually did not spend time searching for a suitable place to dig, nor did visual selection of a patch of clams appear to occur. If the density of clams in area A averaged 14/m2, and if an average spot dug up by an otter was 0.5 x 1.5 m (0.75 m^) as observed in this report, then random digging in area A would produce about 10 clams. This is greater than the average number of six clams taken by otters on a series of dives. Perhaps the otters had learned the location of the clam patches, and because sediment clouds nor- mally prevented visual cues as soon as the sub- strate was disturbed, they simply dug haphazard- ly within the patch. Indeed, Gentry and Peterson (1967) compared the underwater visual acuity of sea otters with the sea lion, Zalophus califor- nianus, and harbor seal, Phoca vitulina, and pro- posed that vision in otters may be better adapted for aerial situations of predator detection rather than for underwater prey location. The strategy of repeatedly enlarging the hole to capture clams is a good one, because it makes efficient use of the labor to start the hole on initial dives. Anyone who has dug in sand at the seas' --e knows that it is relatively easy to enlarge a hole, and it would be advantageous to do this rather than dig straight down for each individual clam. The behavior of digging like a dog has also been reported by Shimek (1977) for a sea otter taking subtidal echiuroid worms and is apparently simi- lar to the behavior of sea otters taking clams in shallow subtidal and intertidal waters in Alaska (Calkins 1978). The holes reported by these au- thors were only half the size of freshly dug holes at Monterey Harbor, however. The first author has observed similar (1.5 m across and 0.5 m deep) holes dug by otters in the sand channels in 12 m of water off kelp forests at Pacific Grove, Calif. In areas such as Prince William Sound and Monterey Harbor, where otters forage heavily on clams, their digging must cause a major disturbance of the infaunal community. Sea otters have been termed "keystone pred- ators" (Estes and Palmisano 1974; Estes et al. 1978), because they regulate populations of epibenthic invertebrates, perhaps through a pro- cess of switching between prey species. At Mon- terey Harbor there is circumstantial evidence that sea otters have had a major impact on two other prey items. Surveys by the California Department of Fish and Game showed C. antennarius and C productus were taken in abundance by fishermen from the Monterey wharves prior to the return of sea otters, but they were rarely taken at Monterey in 1972-74, while still caught in abundance at piers north of the range of sea otters (California Department of Fish and Game"*). Observations on the scuba dives reported here for 1976-77 confirm that cancer crabs are rare in the harbor. Mytilus edulis and M. californianus formed dense clumps on wharf pilings prior to the return of sea otters (Haderlie^), but mussels are small and uncommon there now (Haderlie and Donat 1978). Curiously, large specimens ofB. nubilus are still abundant on the pilings and were not taken in appreciable numbers by sea otters, even though these barna- cles were taken frequently by otters at other loca- tions in the Monterey area (pers. obs.). The factors which regulate prey selection by sea otters remain poorly understood. ■"California Department of Fish and Game. 1976. A pro- posal for sea otter protection and research and request for the return of management to the State of California. Calif. Dep. Fish Game, Oper. Res. Branch, Vol. 1: Text and summaries, 271 p. ^E. C. Haderlie, Professor, Naval Postgraduate School, Mon- terey, CA 93940, pers. commun. May 1976. 162 Acknowledgments John S. Pearse, Ronald Jameson, and an anonymous reviewer provided advice and critical discussion for this study. Christopher Harrold gave technical and diving help. We thank the staff at Hopkins Marine Station of Stanford University for their assistance. This work was partially funded by the U.S. Fish and Wildlife Service. Literature Cited Calkins, D. G. 1978. Feeding behavior and major prey species of the sea otter, Enhydra lutris, in Montague Strait, Prince William Sound, Alaska. Fish. Bull., U.S.76:125-131. ESTES, J. A., AND J. F. PALMISANO. 1974. Sea otters: their role in structuring nearshore com- munities. Science (Wash., D.C.) 185:1058-1060. ESTES, J. A., N. S. Smith, and J. F. Palmisano. 1978. Sea otter predation and community organization in the western Aleutian Islands, Alaska. Ecology 59:822- 833. Faro, J. B. 1969. A survey of subtidal sea otter habitat off Point Pinos, California. M.A. Thesis, Humboldt State Univ., Areata, Calif., 278 p. Gentry, R. L., and R. S. Peterson. 1967 . Underwater vision of the sea otter. Nature (Lend.) 216:435-436. haderlie, E. C, and W. donat ni. 1978. Wharf piling fauna and flora in Monterey Harbor, California. Veliger 21:45-69. HOUK, J. L., AND J. J. GEIBEL. 1974. Observation of underwater tool use by the sea otter, Enhydra lutris Linnaeus. Calif. Fish Game 60:207-208. LOUGHLIN, T. R. 1977. Activity patterns, habitat partitioning, and groom- ing behavior of the sea otter, Enhydra lutris, in Califor- nia. Ph.D. Thesis, Univ. California, Los Ang., 110 p. MCCLENEGHAN, K., .A.ND J. A. AMES. 1976. A unique method of prey capture by a sea otter, Enhydra lutris. J. Mammal. 57:410-412. Radinsky, L. B 1968. Evolution of somatic sensory specialization in otter brains. J. Comp. Neurology 134:495-505. SHIMEK,S. J. 1977. The underwater foraging habits of the sea otter, Enhydra lutris. Calif. Fish Game 63:120-122. Stephenson, M. D. 1977. Sea otter predation on Pismo clams in Monterey Bay. Calif. Fish Game 63:117-120. ANSON H. HINES Chesapeake Bay Center for Environmental Studies Smithsonian Institution P.O. Box 28, Edgewater, MD 21037 Thomas r. loughlin Office of Marine Mammals and Endangered Species National Marine Fisheries Service, NOAA Washington, D C 20235 EFFECT OF ZINC ON FIN REGENERATION IN THE MUMMICHOG, FUP^DULUS HETEROCLITVS, AND ITS INTERACTION WITH METHYLMERCURY Methylmercury has been found to retard fin re- generation in the marsh killifish, Fundulus confluentus, and striped mullet, Mugil cephalus (Weis and Weis 1978). In F. confluentus the re- tarding effect of methylmercury was masked in water of reduced salinity (9%o). Cadmium, which also retarded fin regeneration in killifish (Weis and Weis 1976), interacted antagonistically with methylmercury so that fish exposed simultane- ously to the two metals exhibited growth rates comparable to controls (Weis and Weis 1978). This paper reports on the effects of zinc on re- generation in the mummichog, F. heteroclitus , and the effects of combinations of methylmercury and zinc on this process. Methods Fish were collected by seining in the vicinity of Montauk, N.Y. The lower portion of each caudal fin was amputated with a scalpel, and approxi- mately 15 fish were placed in each of several all- glass aquaria with 10 1 of 30%o salinity water. The temperature was 20°-22° C and the photoperiod was 14 h light/10 h darkness. Fish were fed com- merical fish food and live grass shrimp, Palaemo- netes pugio. Tanks were dosed with methylmer- curic chloride (I.C.N. Pharmaceuticals, Plainview, N.Y.i) from a 0.1 mg/ml stock solution in 0.2% NaHCOg to yield a final calculated concen- tration of 0.050 or 0.025 ppm depending on the experiment, and/or with ZnClg (Reagent Grade, Fisher Scientific) from a 1.0 mg/ml stock solution to yield calculated concentrations of 1.0, 3.0, or 10.0 ppm. Aquaria were washed, refilled, and re- dosed after 2, 4, 7, 9, and 11 days. Regenerating fins were measured with a calibrated ocular mi- crometer in a stereomicroscope at 7, 9, 11, and 14 days. Experiments were terminated at 2 wk be- cause after that time it became difficult to ascer- tain the point at which the amputation had been made. The amputation plane can be seen clearly in Figure 1, a control fin 1 wk after amputation. Three experiments were performed. Experi- ment I involved exposure offish 3.5-4.2 cm stan- » Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. FISHERY BULLETIN: VOL. 78, NO. 1, 1980. 163 Figure l. — Photograph of regener- ating caudal fin, 1 wk after amputa- tion. Measurements were made at A toB. mm _ dard length (SL) to 0.05 ppm methylmercury, 1.0, 3.0, or 10.0 ppm zinc or combinations of 0.05 ppm methylmercury with 1.0, 3.0, or 10.0 ppm zinc. Experiment II was similar, but used 0.025 ppm methylmercury and fish 4.1-5.2 cm SL. Experi- ment III used 0.05 ppm methylmercury and the same concentrations of zinc, but was performed in water of reduced salinity (10%o) on fish 4.3-5.1 cm SL. Fish were frozen at the end of some experi- ments and later analyzed for metal uptake by atomic absorption spectrophotometry (cold vapor technique for mercury, flameless atomic absorp- tion spectrophotometry for zinc). These analyses can be considered accurate within 10%. Results Table l. — Analysis of variance on effects of methylmercury and zinc on fin regeneration in Fundulus heteroclitus. Source of variation df SS MS Hg Zn Hg X Zn 24.778 2.200 0.411 24.778 0.733 0.137 173.671 5.139 0.960 0.001 0.003 0.416 Table 2. — Growth of tail regenerates in Fundulus heteroclitus e5q)osed to methylmercury and zinc for 14 days in Experiment II and 11 days in Experiment III. Experiment II E: xperiment III Exposure n mm ±SE n mm ±SE Controls 12 3.62±0.078 12 2. 28^0.069 meHg 15 3.31 ±0.128* 8 1,67±0.125- meHg + 1 ppmZn 12 3.59±0.099 7 1. 82 ± 0.094* meHg + 3 ppm Zn 13 3.49±0.116 2 2. 14 ±0.060 meHg + 10 ppmZn 12 3.50^0.084 1 2.24±0.00 1 ppm Zn 13 3.73±0,144 11 2. 43 ±0.098 3 ppm Zn 14 3.86i0.142 9 2.44±0.111 lOppmZn 10 4.18±0.159* 2 2.46±0.020* *Significantly different from controls (P<0.05) by (-test. In Experiment I, caudal fin regeneration was retarded by methylmercury and was accelerated by zinc in a dose-dependent fashion. The retarda- tion produced by the mercury could be partially counteracted by the zinc (Figure 2). Analysis of variance of day 14 (Table 1) showed significant effects of mercury, and of zinc, but not of interac- tion. Experiment II, using 0.025 ppm methylmer- cury, produced similar results (Table 2). It can be seen that zinc again accelerated growth in a dose-dependent manner and counteracted the methylmercury-caused depression of growth. Only the group in methylmercury alone and the group in 10 ppm zinc were significantly different from controls (P^0.05) as determined by the ^-test. In Experiment III (10%o salinity) a similar pat- tern was seen (Table 2). High mortality due to interruption in air supply caused the experiment to be terminated early. The groups in methylmer- cury, methylmercury + 1 ppm zinc, and in 10 ppm zinc were significantly different from con- trols, as seen by the ^-test. Analysis of mercury uptake revealed consider- able variation (Table 3). However, it seems likely that the tissue residues are dose dependent and that zinc does not change the uptake of mercury 164 4r E E c a> 11 14 Days Figure 2. — Regenerative growth of tail fin of Fundulus heter- oclitus exposed to methylmercury and zinc in seawater, Exper- iment I. Key: A 10.0 ppm Zn (n = 15), V 3.0 ppm Zn(n = 15), D 1.0 ppm Zn (« = 15), C Control in = 14), A 0.05 ppm meHg + 10.0 ppm Zn {n = 11), ▼ 0.05 ppm meHg + 3.0 ppm Zn {n = 12), 0.05 ppm meHg + 1.0 ppm Zn in = 9), • 0.05 ppm meHg (n = 12). Table 3. — Average mercury uptake (ppm Hg/wet weight ± SE) by Fundulus heteroclitus. Control Hg Hg + 10 ppm Zn Item Uptake n Uptake n Uptake n Experiment 1: Carcass n.d.' 3 32-5.0 3 37 = 7.2 3 Brain n.d. 3 11=0.9 3 19 = 4.8 4 Experiment II: Carcass n.d. 3 7.4=0.7 3 15.4 = 3.1 5 Brain n.d. 3 4.8=26 3 9.7 = 1.6 5 Experiment III: Carcass n.d. 3 33 = 12 3 25=1.4 3 Brain n.d. 3 28 = 6.0 3 25±2.5 3 'n.d. = not detectable, <0.03ppm. into the brain or the rest of the body. Accumula- tion of zinc was not altered by methylmercury. Animals in 10 ppm Zn accumulated 246±1.41 ppm; those in 10 ppm Zn + 0.05 ppm meHg ac- cumulated 250±3.54 ppm. Those in 1 and 3 ppm Zn accumulated 221±25.2 and 250±4.95 ppm, showing no clear dose-dependent relationship. Discussion The data indicate that in F. heteroclitus, zinc can accelerate regenerative growth, and, by so doing, can counteract the retarding effects of methylmercury. In this species, the regeneration rate of controls was similar in 30%o and 10%o sa- linity, and the methylmercury retarded growth at both salinities. This is in contrast to F. confluen- tus in which decreased salinities depressed the regeneration rate, thus masking the effects of methylmercury in water of 9%o salinity (Weis and Weis 1978). Methylmercury has previously been observed to retard regeneration (Chang et al. 1976; Weis and Weis 1978) and other developmental processes (Chang et al. 1974). Its action as an inhibitor of mitosis (Ramel 1969) could be the cause of these effects on growth processes. As a potent nerve poison it could further inhibit regenerative growth by interfering with the neurotrophic influence necessary for regeneration. Previous studies on the effects of zinc on grov^dh include the work of Hirsch and Hurley (1978) in which zinc was found to counteract the teratogenic effects of 6-mercaptopurine in the rat. They felt that the drug lowered DNA synthesis and that the zinc counteracted this. Swenerton et al. (1969) correlated zinc deficiency with reduced DNA synthesis in rat embryos, and Falchuck et al. (1975) have associated zinc with promoting cell division in Euglena gracilis. Thus, if zinc can promote DNA synthesis and cell division in fish also, that would account for the observed acceler- ation of regenerative growi:h. However, previous studies on fish have not indicated such an effect. Crandall and Goodnight (1962) reported that 1.15 ppm zinc retarded the growth of newborn gup- pies. Rachlin and Perlmutter (1969) found that 18 ppm Zn reduced the mitotic index of cultured rainbow trout, Salmo gairdneri, cells, but that 1.8-10.0 ppm had no effect on the mitotic index. On the other hand, zinc has often been found to counteract toxic effects of other heavy metals. Dixon and Compher (1977) found that zinc could reverse a cadmium-caused inhibition of regenera- tion in the nevft. Zinc has been found to coun- teract the toxic effects of mercury in rats (Yamane et al. 1977) and to counteract the teratological effects of methylmercury in killifish embryos (Weis et al. in press). In view of reports of fin rot of unknovvTi etiology in flatfish from polluted environments (Ziskowski 165 and Murchelano 1975), the retardation of growth by heavy metals may be of significance in inhibit- ing regeneration of fins eroded by the benthic substrate. Addendum It has recently been demonstrated that, in cer- tain poeciliid fishes, some environmental vari- ables which affect general growth rate do not af- fect the rate of fin regeneration. Factors which do cause differences in length of regenerated fin generally affect the time needed for wound heal- ing and blastema formation, rather than rates of regeneration per se (E. Zimmerer, Ph.D. disserta- tion, Rutgers University, 1980). We tested the data represented in Figure 1 for this possibility. In regression analysis, the slope equals regenera- tive rate per se and the elevation (^'-intercept) represents the time needed for wound healing and blastema formation. Analysis of covariance indicates that when Hg treated fish are compared with control fish, both the slopes and j -intercepts are significantly different {F = 10.23 and 80.76, respectively). Similarly, when 10 ppm Zn treated and control fish are compared, the slopes and ^/-intercepts are significantly different {F = 6.83 and 41.29, respectively). Therefore, it appears that these heavy metals affect both the initial wound healing and blastema formation and the rate of regeneration per se. Acknowledgments This work is a result of research sponsored by NOAA Office of Sea Grant, U.S. Department of Commerce, under grant No. 04-7-158-44042. Thanks are extended to J. C. Baiardi of the New York Ocean Science Laboratory for the use of facilities for this study, and to J. Seebode Jr., J. Ricci, and G. Millinger for technical assistance. We thank S. L. Cheng of the New Jersey Institute of Technology for the atomic absorption spec- trophotometry, and J. Chou of the College of Medicine and Dentistry of New Jersey for statis- tical assistance. Literature Cited Chang, L. W., l. M. Mak, and a. h. Martin. 1976. Dose-dependent effects of methylmercury on limb regeneration of newts (Triturus viridescens). Environ. Res. 11:305-309. Chang, L. W., K. R. Reuhl, and a. W. Dudley, Jr. 1974. Effects of methylmercury chloride on Rana pipiens tadpoles. Environ. Res. 8:82-91. Crandall, C. a., and C. J. Goodnight. 1962. Effects of sublethal concentrations of several toxi- cants on growth of the common guppy, Lebistes re- ticulatus. Limnol. Oceanogr. 7:233-239. DIXON, C, AND K. COMPHER. 1977. The protective action of zinc against the deleterious effects of cadmium in the regenerating forelimb of the adult newt, Notophthalmus viridescens. Growth 41:95-103. Falchuck, K. H., D. w. Fawcett, and B. L. Vallee. 1975. Role of zinc in cell division of £ug/en) 26 73.31 12.43 23 71.83 1082 318 61 72 42 Lymphocytes (%) 26 13.50 6.59 23 16 96 7.30 318 21 72 29 Eosinophils (°o) 26 13.38 9.65 23 9 17 765 318 14- 72 22 Monocytes (%) 26 0.58 0.86 23 0.43 0.73 318 3 72 5 Basophils(°o) 26 23 — — — — Blood urea nitrogen (mg 100 ml) 29 24.40 4.91 20 46.40" 14.82 232 51- 62 37 Calcium (mg 100 ml) 29 10.67 2.01 17 8.67" 056 166 10 29 10 Creatinine phosphokinase (lU liter) 28 2293 16.09 13 86.77" 134.82 — — — — Total cholesterol (mg, 100 ml) 29 218.17 47.30 20 305 00" 87.38 301 221 92 154 Lactic acid dehydrogenase (lU liter) Females 17 566.24 80.65 17 364 65" 76,52 100 113 11 179 Males 12 508.83 56.83 — — — 80 130 18 244 Alkaline phosphatase (lU liter) 29 98.72 56.51 20 149.15" 113 80 71 241- 6 256- Serum glutamic oxaloacetic transaminase (lU/liter) 29 217.52 88 46 20 382 15" 1 66.82 172 98- 34 110- Serum glutamic pyruvic transaminase (lU/liter) 29 43.03 63.58 20 31 89" 2843 88 19 14 45 Glucose (mg 100 ml) 29 80.76 28.76 20 167.10" 52.34 231 129 52 117 Total protein (g 100 ml) 29 7.60 0.56 20 7.32 085 133 8.0 10 9.0 Albumin (g 100 ml) 29 3.55 0.34 20 3.71 0.40 109 34 10 3.9 Globulin (g 100 ml) 29 3.91 0.73 20 3.55 0.75 — — — — 'Animals from Loggerhead Key; each animal sampled once ^The four females held at Sea World; each animal sampled several times. ^Data from Ridgway et al, (1970), •■Numtjer of determinations made 5Values calculated using mean values for hematocrit, hemoglobin and red cell count. 'Significant difference between males and females, f-test. 0.01 level. cm female consumed an average (±SD) of 20 ±7.6 kg of mackerel and herring/day, in a ratio of 1.5:1, from 25 July through 9 August. Food consumption decreased significantly on 10 August and the ani- mal died on 13 August. Similarly, the 297 cm female consumed 15.3 ±4.0 kg of mackerel and herring (2.2:1) between 24 July and 3 August. Food consumption dropped to 1.8 kg on 4 August, rose to 18.6 kg on 8 August when smelt was added to the diet, and then decreased to 5.4 kg on 13 August. Overall food consumption between 24 July and 13 August was 11.7 ±5.7 kg/day. The animal died on 14 August. The 358 cm female consumed 15.1±8.5 kg of mackerel and herring/ day (1.2:1) between 24 July and 7 August. Food consumption decreased on 4 August and remained stable through 7 August {x = 9.0 ±2.5 kg/day). Consumption between 24 July and 3 August was 17.1 ±8.9 kg/day. Smelt was introduced on 8 Au- gust in place of mackerel and total consumption was 22.7 kg. Squid was also added on 9 August. The animal died on 10 August. The 475 cm female had an erratic food consumption (mackerel and herring, 19.4 ±16.2 kg/day) between 24 July and 29 July when it died. The individual blood chemis- try analyses for these four animals reflected their deteriorating condition (Odell et al. see footnote 8) and their combined values were significantly different from the Loggerhead Key animals (Table 1). Relationships Among Strandings It is clear, based on photographs, that some of the false killer whales that left Pine Island Sound were the same individuals that stranded on Loggerhead Key. Low altitude (helicopter) aerial photographs were taken of the animals leaving Captiva Pass (Larson see footnote 3). Comparison of these photographs with photographs of dorsal fins of the Tortugas animals provided positive identification of several individuals. Dorsal fin shapes have been used to identify specific indi- vidual dolphins over periods of several months 175 (Wiirsig and Wiirsig 1977). Assuming that the animals left Captiva Pass at about 1200 h on July 22. and that they travelled in a straight line, they had to travel about 80 km/day to reach Loggerhead Key at 1300 h on 25 July. When these animals were escorted away from Loggerhead Key on 27 July, they apparently headed northeast (Schimpft"''*). The dead animals we found on Cape Sable were too decomposed to tell if they were the Captiva-Tortugas animals. Three large black whales were seen by a National Park Service pilot several kilometers east of the Dry Tortugas when the other animals were stranded. These may be the first three animals found floating off Cape Sable on 2 August by Park Service personnel, but the evidence is only circumstantial. The sequence of strandings described herein roughly parallels a series of pilot whale, Globicephala macrorhynchus , strandings that oc- curred in the same vicinity on 19-20 August 1971 (Fehring and Wells 1976). Forty-four pilot whales stranded on Manasota Key and on Gasparilla Is- land a few kilometers to the south (Figure 1). On 25 August 1971, 12 or 13 pilot whales were found stranded on the Marquesas Keys east of Key West (Figure 1). At least one of these was positively identified to be from the previous stranding. Fehring and Wells (1976) reported that the pilot whales observed stranding on Gasparilla Island made "a deliberate shoreward movement" as op- posed to "disoriented panic." Eugene Shinn (see footnote 5) unknowingly photographed the false killer whales minutes before they beached on Loggerhead Key while taking aerial photographs of the reef formations. The photographs show two close-knit pods of whales heading towards the beach. Fehring and Wells also reported that the behavior of the stranded animals changed after the two largest pilot whales were towed offshore and held there with ropes around their caudal peduncles. The remainder of the animals then showed less tendency to return to shore when pushed off. Several of the larger Loggerhead Key whales were forced offshore (headfirst, without ropes around their tails) in hope that the others would follow. The animals herded offshore re- turned to the beach when released. The operation was only successful when all of the animals were forced offshore simultaneously and herded to "Robert Schimpff, Department of Pathology, College of Medicine, University of Florida, Gainesville, FL 32610, pers commun. 1977. 176 deeper water, using swimmers and two boats. While on the Loggerhead Key beach, the whales were docile, as Fehring and Wells (1976) reported for the pilot whales. Coast Guard personnel who followed the animals offshore reported that the herd split into two groups (one of 17-18 and one of 10 or 11 animals) (Schimpff see footnote 14). Conclusions From the veterinary medical standpoint, we would doubt the ability of those animals that were necropsied to function normally with the heavy parasite load in the pterygoid sinus complexes. The benefits of forcing live stranded animals back out to sea must be carefully weighed against the benefits of bringing them into captivity, where they can be observed closely and thoroughly ne- cropsied should death occur. If stranded whales are returned to sea, they should be given permanent, individual identification marks (e.g., freeze brands) and, ideally, outfitted for radio tracking. Acknowledgments The data presented in this paper could not have been collected without the generous assistance of many organizations and individuals, including the Florida Marine Patrol, U.S. Coast Guard, U.S. National Park Service, National Marine Fisheries Service, Wometco Miami Seaquarium, Gary Davis, Gary Hendrix, Deke Buesse, Ralph Miele, John Reynolds, and others. Material from the four captive animals at Sea World was examined as follows: ovaries - Richard J. Harrison; parasites - Robin M. Overstreet; histopathology - Armed Forces Institute of Pathology. Gordon Hubbell kindly provided the information on the 1972 stranding. Donald Forrester, Robert Schimpff, and Nicholas Hall commented on an early draft of this paper. William F. Perrin, James G. Mead, and Edward D. Houde provided useful criticisms on a later draft. Literature Cited ASH, C. E. 1952. The body weights of whales. [In Engl, and Norw.] Nor. Hvalfangst-Tidende 41:364-374. 1953. Weights of Antarctic humpback whales. Nor. Hvalfangst-Tidende 42:387-391. BEAUCHAMP, J. J., AND J. S. OLrfON. 1973. Corrections for bias in regression estimates after logarithmic transformation. Ecology 54:1603-1607. Brown, D. H., D. K. Caldwell, and M. C. Caldwell. 1966. Observations on the behavior of wild and captive false killer whales, with notes on associated behavior of other genera of captive delphinids. Bull. Los. Ang. Cty. Mus. Nat. Hist. Sci. 95, 32 p. BRYDEN, M. M. 1972. Growth and development of marine mammals. In R. J. Harrison (editor), Functional anatomy of marine mammals. Vol. I, p. 1-79. Acad. Press, N.Y. Caldwell, D. K., M. C. Caldwell, and C. M. Walker, Jr. 1970. Mass and individual strandings of false killer whales, Pseudorca crassidens , in Florida. J. Mammal. 51:634-636. Comrie, L. C, and a. B. Adams. 1938. The female reproductive system eind corpora lutea of the false killer whale, Pseudorca crassidens Owen. Trans. R. Soc. Edinb. 59(2):521-531. DUDOK VAN Heel, W. H. 1962. Sound and Cetacea. Neth. J. Sea. Res. 1:407-507. Fehring, W. K., and R. S. Wells. 1976. A series of strandings by a single herd of pilot whales on the west coast of Florida. J. Mammeil. 57:191-194. FUJINO, K. 1955. On the body weight of the sei whales located in the adjacent waters of Japan (II). Whales Res. Inst. Sci. Rep. 10:133-139. Marelu, C. A. 1953 . Documentos iconigraficos sobre cetaceos de las costa Argentinas. Ann. Nahuel Huapi 3:133-143. Mitchell, E. 1975a. Porpoise, dolphin, and small whale fisheries of the world; status and problems. Int. Union Conserv. Nat. Nat. Resour., Monogr. 3, 129 p. Mitchell, E. D. (editor). 1975b. Review of biology and fisheries for smaller ceta- ceans. Report of the meeting on smaller cetaceans, Montreal April 1-11, 1974, subcommittee on small ceta- ceans, scientific committee, International Whaling Com- mission. J. Fish. Res. Board Can. 32:889-983. Norman, J. R., and F. C. Eraser. 1948. Giant fishes, whales, and dolphins. Putnam, Lond.,376p. Omura, H. 1950. On the body weight of sperm and sei whales located in the adjacent waters of Japan. Whales Res. Inst. Sci. Rep. 4:1-13. Porter, J. W. 1977. Pseudorca stranding. Oceans 10(4):8-16. PURVES, P. E., AND G. PILLERI. 1978. The functional anatomy and general biology of Pseudorca crassidens (Owen) with a review of hydro- dynamics and acoustics in Cetacea. Invest. Cetacea 9:67-227. REIGER, G. 1975. Dolphin sacred, porpoise profane. Audubon 77(l):2-29. RIDGWAY, S. H. 1972. Homeostasis in the aquatic environment. In S. H. Ridgway (editor). Mammals of the sea: biology and medicine, p. 590-747. Charles C. Thomas, Springfield, 111. RIDGWAY, S. H., J. G. SIMPSON, G. S. PATTON, AND W. G. GILMARTIN. 1970. Hematologic findings in certain small ceta- ceans. J. Am. Vet. Med. Assoc. 157:566-575. Sergeant, D. E., and P. F. Brodie. 1969. Body size in white whales Delphinapterus leucas. J. Fish. Res. Board Can. 26:2561-2580. TOMILIN, A. G. 1957. Mammals of the U.S.S.R. and adjacent countries. Vol. 9. Cetacea. |In Russ.) Izd. Akad. Nauk. SSSR, Mosk., 756 p. (Translated by Isr. Program Sci. Transl., Jerusalem, 1967, 717 p.) WiJRSIG, B., AND M. WURSIG. 1977. The photographic determination of group size, com- position, and stability of coastal porpoises (Tursiops trun- catus). Science (Wash., D.C.) 198:755-756. DANIEL K. ODELL Division of Biology and Living Resources Rosenstiel School of Marine and Atmospheric Science University of Miami 4600 Rickenbacker Causeway, Miami, FL 33149 EDWARD D. ASPER Joe Baucom Sea World, Inc. 7007 Sea World Drive Orlando, FL 32809 Sea World, Inc. 1720 South Shores Road San Diego, CA 92109 Lanny H. Cornell OCCURRENCE OF THE FINETOOTH SHARK, CARCHARHISLS ISODO^\ OFF DAUPHIN ISLAND, ALABAMA' Carcharhinus isodon (Valenciennes) is an in- frequently encountered species with a poorly known life history. The literature on this species covering the western North Atlantic contains much information on juveniles, but very little on adults. All lengths discussed herein are total lengths. Radcliffe (1916) reported a single specimen 50.8 cm in the Bureau of Fisheries collection at Beaufort, N.C. Burton's (1940) record of an im- mature male, 74.4 cm, was the first from South Carolina waters. Specimens examined by Bigelow and Schroeder (1948:304-308) ranged from 46 to 56.7 cm. Springer (1950) examined 20 adult females 147-155 cm collected in December off Salerno, Fla. Thirteen had from one to six em- bryos 43-48 cm; the remaining seven had en- larged flaccid uteri and medium-sized ovarian ^Contribution No. 028, Dauphin Island Sea Lab. fishery BULLETIN: VOL. 78, NO. 1, 1980. 177 eggs. He suggested a winter pupping period. Clark and von Schmidt (1965) recorded only a single female (76 cm) in 9 yr of shark research off Sarasota, Fla. Dahlberg and Heard (1969) re- ported the capture of 30 individuals from July through September 1968 off Georgia. Of these, 29 were between 52 and 94 cm. The other specimen (144 cm) was probably the only mature indi- vidual, although there was no mention of sex or reproductive development. Hoese and Moore (1977, appendix 5) listed C. isodon as a spring through fall spawner based on collections of juveniles at Port Aransas, Texas. Compagno (1978), in his review of the species, assigned this species to the genus Carcharhinus. During longlining operations in the northern Gulf of Mexico in summer 1978, a gravid female and two males were collected off Dauphin Island, Ala. On 2 July 1979 one male and one female were collected in the mouth of Mobile Bay. With so few reports of mature C. isodon, these captures will serve to better define the reproduc- tive life history of this species. On 5 June 1978 the gravid female ( 139 cm) was collected by longline in water about 5 m deep, approximately 1 km southwest of Sand Island, a small barrier island approximately 5 km south of the east end of Dauphin Island. The shark carried four embryos ranging from 49 to 51 cm. These appeared to be near-term pups. There were two pups in each uterus, each positioned with the head toward the anterior end of the uterus. Each pup was enveloped by a membrane which was filled anteriorly with a translucent yellow fluid. Each had a highly vascularized placenta attached to the posterior portion of the uterus, and the connecting umbilical cords measured 20.6-30.0 cm. Where an umbilical cord attached to a placenta there were three saclike extensions con- taining a small amount of clear fluid. In earlier embryonic stages of other species of carcharhinid sharks these sacs contain the remaining uncon- sumed portion of the yolk (Gilbert and Schlernit- zauer 1966). The left uterus contained two males; the right uterus one male and one female. The pups and jaws of the female were deposited in the Table i.- Item -Measurements (centimeters; methods after Bass et al. 1973) of the gravid female Carcharhinus isodon and the four pups. Gravid female Pup no. 1 male Pup no. 2 male Pup no. 3 male Pup no. 4 female Total length Fork length! Standard length Snout to: Dorsal 1 Dorsal 2 Pectoral fin Pelvic fin Anal fin (^outh Mouth breadth Between nostrils Eye diameter Gill lengths: No. No. No. No. No, Dorsal 1 height Dorsal 1 base Dorsal 1 free margin Dorsal 2 height Dorsal 2 base Dorsal 2 free margin Anal height Anal base Anal free margin Pectoral height Pectoral base Pectoral free margin Pelvic anterior margin Pelvic distal margin Uppper caudal length Lower caudal length Interspace base dorsal 1 to origin dorsal 2 Interspace base dorsal 2 to caudal pit Origin of pectoral to origin of pelvic Origin of pelvic to origin of anal Weight (g) 139 49 51 50.5 50.5 118 38.5 40.5 40 40.5 106 35 37 36 36.5 46 15.7 16.5 16.5 16.5 90 30.2 31.5 31 32.5 31.5 11 12.2 11.9 12.2 74 22.8 25.4 23.8 24.2 88 28.9 31.9 31.3 30.4 9 3.9 3.9 3.8 3.8 13 4.1 4.4 4.2 4.1 7 2.7 2.7 2.7 2.8 1.8 .8 .8 .8 .7 8 2.5 2.3 2.2 2.3 8.6 2.6 2.5 2.4 2.5 9 2.7 2.7 2.6 2.6 8.5 2.5 2.4 2.3 2.4 7.5 2.1 1.8 1.8 1.8 14.7 3.9 3.8 3.5 3.8 14 4.3 4.6 4.6 5.1 5.5 2.0 2.2 1.8 2.0 4.0 1.2 1.4 1.4 1.1 6.5 1.9 2.2 2.2 2.1 5.5 2.0 2.2 1.9 1.9 4.3 1.4 1.5 1.5 1.3 7.4 2.2 2.4 2.4 2.1 4.8 1.9 1.9 1.8 1.8 22 6.0 6.9 6.3 6.5 8 2.5 2.6 2.5 2.5 6.5 2.2 2.4 2.4 2.4 6.5 2.4 2.6 2.7 2.8 8.8 2.8 2.9 2.7 2.8 39 13.8 14.9 14.7 14.1 17.5 5.1 5.5 5.1 5.4 33 10.1 10.2 9.9 10.3 10.5 3.8 3.9 3.4 3.4 42 11.8 13.2 11.9 12.0 16 6.1 6.5 7.5 6.2 — 704 810 737 758 178 University of South Alabama Ichthyological Col- lection (USAIC 6278). Measurements and weights are found in Table 1. Since most records of C. isodon are of juveniles, there is little information on the reproductive biology of the species. Based on the cited litera- ture and these data, pups appear to be 45-55 cm at birth. However, seasonality is uncertain as the records of Springer (1950) are not in accord with those of either Hoese and Moore (1977) or this report. Length at maturity can be closely estimated. One male (112 cm) collected 13 July 1978 was immature — based on incomplete calcification of the claspers and incompletely developed siphon sacs, each sac being 7.5 cm long and 1.0 cm wide. The other two males ( 120 and 127 cm) collected 2 July 1979 and 28 June 1978 had well-calcified claspers and fully developed siphon sacs. The only literature on mature males (Springer 1950) listed lengths of 140-152 cm. Males apparently mature between 115 and 120 cm. Maturity in females must be reached at a larger size. The female col- lected in July 1979 was 127 cm, yet was imma- ture with only small undeveloped ovarian eggs. The gravid female reported here was 139 cm, and those reported by Springer (1950) were 147-155 cm. Carcharhinus isodon was only collected when similarly sized specimens of blacktip shark, C. limbatus, were caught: 3 C. limbatus ( 126-166 cm) with the gravid female, 12 C limbatus (102-117 cm) with the 112 cm male, 2 C limbatus (111 and 124 cm) with the 127 cm male, and 12 C. limbatus (100-130 cm) with the two specimens caught in 1979. If C. isodon is an uncommon straggler into the northern Gulf of Mexico it may be schooling with other sharks of like size. Sharks that school have been noted to do so by sex or size (Ford 1921). Literature Cited Bass, a. J., J. D. D' Aubrey, and N. kistnasamy. 1973. Sharks of the east coast of Southern Africa. I. The genus Carcharhinus (Carcharhinidae). Oceanogr. Res. Inst. (Durban), Invest. Rep. 33, 168 p. BIGELOW, H. B., AND W. C. SCHROEDER. 1948. Sharks. In Fishes of the western North Atlantic. Part one, p. 59-546. Mem. Sears Found. Mar. Res., Yale Univ. 1. Burton, E. M. 1940. Aprionodon isodon from South Carolina. Copeia 1940:140. Clark, E., and K. Von Schmidt. 1965. Sharks of the central Gulf Coast of Florida. Bull. Mar. Sci. 15:13-83. COMPAGNO, L. J. V. 1978. Sharks. In W. Fischer (editor), FAO species iden- tification sheets for fishery purposes: western central At- lantic. Vol. 5, unpaginated. Dahlberg, M. C, and R. W. Heard, III. 1969. Observations on elasmobranchs from Georgia. Q. J. Fla. Acad. Sci. 32:21-25. FORD, E. 192 1 . A contribution to our knowledge of the life-histories of the dogfishes landed at Plymouth. J. Mar. Biol. As- soc. U.K. 12:468-505. Gilbert, P. W., and D. A. Schlernitzauer. 1966. The placenta and gravid uterus of Carcharhinus falciformis. Copeia 1966:451-457. Hoese, H. D., and r. H. Moore. 1977. Fishes of the Gulf of Mexico, Texas, Louisiana, and adjacent waters. Tex. A&M Univ. Press, College Sta- tion, 327 p. Radcliffe, L. 1916. The sharks and rays of Beaufort, North Caroli- na. Bull. U.S. Bur. Fish. 34:239-284. Springer, S. 1950. A revision of North American sharks allied to the genus Carcharhinus. Am. Mus. Novit. 1451, 13 p. Steven Branstetter ROBERT L. SHIPP University of South Alabama Dauphin Island Sea Lab P.O. Box 386 Dauphin Island, AL 36528 SHEDDING RATES OF PLASTIC AND METAL DART TAGS FROM ATLANTIC BLUEFIN TUNA, THUNNUS THYNMUS^ In 1971, the International Commission for the Conservation of Atlantic Tunas (ICCAT) recom- mended that a double-tagging experiment be con- ducted on Atlantic bluefin tuna, Thunnus thyn- nus, to determine whether plastic or metal dart tags were more efficient and to estimate im- mediate and instantaneous tag shedding rates. A knowledge of shedding rates is necessary so that appropriate adjustments can be made when es- timating mortality rates from tag return data. This study was begun in 1971 by the National Marine Fisheries Service (NMFS), the Woods Hole Oceanographic Institution (WHOI), and the Fisheries Research Board of Canada (FRBC). The ^Southeast Fisheries Center Contribution Number 80-14M. FISHERY BULLETIN: VOL. 78, NO. 1, 1980. 179 results obtained through 1972, for 580 double- tagged bluefin tuna released during 1971 off the east coast of the United States, were reported by Lenarz et al. ( 1973). Their results were partially based on tags supplied by the FRBC, some of which had longer streamers than the tags supplied by WHOI. For our present analysis, we used only data from the WHOI tags. In this paper we present the overall findings obtained through 1978 for 3,121 double-tagged bluefin tuna. These fish were released primarily from U.S. purse seine vessels fishing off the east coast of the United States from Virginia to Mas- sachusetts from 1971 through 1977. Methods The U.S. double-tagging program for Atlantic bluefin tuna was conducted jointly by the NMFS and WHOI. Tags and tagging procedures were those described by the Food and Agriculture Or- ganization (1972). All fish were tagged and re- leased from U.S. purse seine vessels (98% of all releases) and from a few sport fishing vessels. Tagging occurred throughout the purse seine fishing season during 1971, 1973, and 1974, and at the end of the season during 1972, 1975, 1976, and 1977. The double-tagging operation was con- ducted entirely by John Mason during each year except 1974, when two assistants aided in the dou- ble tagging. Precise release dates were available for all of the fish. In a few instances only the month and year were known for the recapture data. In these cases, the 15th of the month was arbitrarily selected to represent the recapture date. The vast majority of returns fall into an annual cycle dur- ing which the recapture periods are approxi- mately 2-3 summer months. The interval mid- points of the time intervals can be considered to be on a yearly cycle. Therefore, we grouped returns into "first year returns," "second year returns," etc., and calculated average days out from the individual days out for each return. Tag shedding rates were estimated using the notation and methodology of Bayliff and Mobrand (1972) for yellowfin tuna, which Lenarz et al. (1973) used for bluefin tuna and Laurs et al. (1976) used for North Pacific albacore. Chapman et al. (1965) developed the original model with the assumption of only one type of shedding which occurs at a constant in- stantaneous rate. Bayliff and Mobrand (1972) as- sumed that there are two types. Type I which oc- curs immediately after the fish are released and 180 Type II, the type described by Chapman et al. (1965). Bayliff and Mobrand's modifications^ of the Chapman et al. (1965) approximate equations for tag returns of double-tagged fish are: nddk^FrNj)Trp^e^p-iF + X + 2L)tk (1) n^sk = 2FtNd 7rp(l - pexp{-Ltk)) expi-{F + X + L)tk) (2) where n^^ = number of returns of double-tagged fish retaining both tags caught dur- ing the recapture period tk, n^gi^ = number of returns of double-tagged fish retaining only one tag caught during the period ^^ , F = instantaneous rate of fishing mortal- ity, A^^ = number of double-tagged fish re- leased, 77 = proportion of tagged fish which re- main alive after the Type-I mortality (immediate) has taken place, p = proportion of the tags which are re- tained after Type-I shedding (im- mediate) has taken place, X = instantaneous rate of mortality due to natural causes, Type-II tagging mortality (long term), and emigra- tion from the fishing grounds, L = instantaneous rate of tag shedding (Type II), and tf^ = time at the middle of the /sth recap- ture period of length T(k = 1, 2, 3). From Equations (1) and (2) it follows that ^dsk 2(1 - pexp(-L^fe ))exp(L^fe ) n ddk and therefore ridsk expiLtk)-p exp(L^fe) 2nddk ^^ddk ^^ddk Rearranging terms yields ^As pointed out by Laurs et al. ( 1976), there was typographical error in both Bayliffand Mobrand (1972) and Lenarz etal (1973) in Equation (2). 2n ddk and hence In f^dsk + ^f^ddk ^^ddk ndsk ■•" 2A2ddfe = pexp(-Lffe) = Inp — Ltfj = y. where Y^ is an estimate of the natural logarithm of the proportion of tags retained up to time t/,. Given n^idk , ^^s* > and t^ , then L and p can be esti- mated using linear regression. We first estimated these parameters using the usual least-squares linear regression which assumes homoscedastic- ity. We also believed that it would be appropriate to consider that variability may increase as a func- tion of time as the number of recoveries decreases. To accomplish this, a weighting factor was intro- duced and a weighted least-squares linear regres- sion model was fitted to calculate values of Inp and L, as was done by Bayliff and Mobrand (1972). The weights for each time interval k (k = 1, 2, 3) were equated to the ratio of the number of returns of double-tagged fish during interval k to the total number of returns of double-tagged fish during all /j -periods. This can be simply expressed as: GJfe ^ddk + l^dsk 3 2 i^ddi + nasi) i=l While we consider this a reasonable first approxi- mation of the correct weight, further investiga- tions of the statistical properties of Yf^ to formally determine the correct weighting procedure are de- sirable. Estimates of Inp and L were then made using weighted linear regression. Results and Discussion The double-tag releases during 1971 through 1977 and returns in 1971 through 1978 are shown by tag type (Table 1). A sufficient number of tag returns existed to allow examination of three separate recapture periods. Only a few returns existed from beyond the third recapture period. There were approximately equal numbers of each tag type released each year. Table 1 constitutes the basic data used throughout this study. Using the basic data, we estimated values of immediate (Type I) and instantaneous (Type II) shedding rates for each tag type. Further, we tested several hypotheses including: 1) equality of return rates for same year recaptures; 2) equality of return rates by estimated age; and 3) differences in re- turns and nonreturns over 2 or 3 yr time periods for various time intervals. Using the double-tagging release data for all years combined (1971-77) the return rate for plas- tic tags was 5.1% the first year, 8.6% the second year, and 1.6% the third year. The return rate for metal dart tags was 5.5% the first year, 9.1% the second year, and 2.9% the third year. Therefore, for both types of tags the return rates increased the second year and decreased the third year. This should be expected since tagging occurred at the end of the purse seine season for several of the release years studied. Chi-square tests (not cor- Table 1. — Tag releases and returns from northwestern Atlantic bluefin tuna double-tag study. For each of /z = 1, 2, or 3 recapture periods the number of returns of double-tagged fish retaining both tags is n^^/^ and those retaining only one tags is n^^/^ . The average number of days-at-large for each period is tf;. [ Double-tagged releases First-year returns Second-year returns Third-year returns Tag type Year Number "dd1 "ds^ t^ (days) "dd2 "ds2 f2 (days) "dd3 "ds3 fg (days) Plastic dart 1971 150 4 7.25 20 9 349.07 3 1 724.00 (D-tag) 1972 75 6 12.83 17 4 340.52 1 1 726.50 1973 134 18 2 18.45 6 4 354.20 1 708.00 1974 629 25 4 12.07 18 12 352.17 4 7 727.82 1975 50 1 40.00 1 1 384.50 1976 267 12 2 16.36 2 2 341.00 1 2 707.33 1977 223 3 1 47.50 25 4 361.83 — — — Total 1,528 68 10 16.46 89 36 352.06 9 12 723.10 Metal dart 1971 162 4 1 18.60 10 9 358.63 2 3 724.80 (H-tag) 1972 77 1 11.00 9 11 343.55 1 740.00 1973 131 12 5 16.88 1 3 373.25 2 720.00 1974 666 28 2 10.97 40 13 358.57 15 11 703.19 1975 58 1 43.00 4 5 339 11 2 687.50 1976 271 23 3 23.08 6 311.00 4 759.25 1977 228 8 36.00 24 2 365.19 — — — Total 1,593 76 12 18.76 94 43 354.71 19 21 712.48 Grand total 3.121 144 22 17.68 183 79 353.44 28 33 716.13 181 rected for continuity) showed that there were no significant differences at the 0.01 level, with 1 degree of freedom, in return and nonreturn rates between tag types for fish at liberty for 1, 2, or 3 yr (Table 2). However, returns were significantly better at the 0.05 level for metal tags in the third year. Further, there was no significant difference at the 0.01 level in the first-year return and non- return rates between the two types of dart tag, whether comparing each year individually or comparing all years combined (Table 3). We also tested for differences in return and non- return rates between age-groups. Fish were aged from unpublished length-age tables (Rivas^). Chi-square values for fish tagged at ages 1, 2, 3, ^L. R. Rivas, Southeast Fisheries Center Miami Laboratory, Natl. Mar. Fish. Serv., NOAA, 75 Virginia Beach Drive, Miami, FL 33149. and 4 + , were not significant at the 0.01 level (Ta- ble 4). The results of the chi-square test indicated that tag types and ages could be combined. Unweighted and weighted linear regression models were used to estimate immediate tag shed- ding rate (1 - p) and instantaneous shedding rate (L) (Table 5). The unweighted model for both tags combined yielded an estimate of immediate tag shedding ( 1 - p) to be 0.040 (0.042 for the weighted model). The overall estimate of the instantaneous rate of tag shedding (L) on an annual basis using the model was 0.205 (0.186 for the weighted model). (The annual rate analog for L from the unweighted model is 0.19.) Therefore, the results from each model were similar. We chose to use the unweighted results, which give a slightly higher L value. While results of the chi-square test indi- cated that tag types could be combined, estimates were also made for each tag type separately to Table 2. — Chi-square tests (df = 1) of equality of yearly return and nonreturn rates between double-tagged releases for 1971-77 combined, for A = 1 , 2, or 3 yr at liberty, using plastic or metal dart tags on bluefin tuna in the northwestern Atlantic Ocean. The number of returns of double- tagged fish retaining both tags is n^^/^ and those retaining only one tag is n^g/i- Plastic dart tags Metal dart tags Return year (k) Double-tagged releases Total returns k\h year Return rate Double-tagged releases Total returns /tth year ^"ddk^"dsk^ Return rate Chi-square value 1 2 3 1.528 1,450 1,325 78 125 21 Average 0.05105 0.08621 0.01585 0.05104 1,593 1,505 1,368 88 137 40 0.05524 0.09103 0.02924 0.05850 0.272 0.213 5.452- •P«0.05. T.ABLE 3. — Chi-square tests (df = 1) of equality of return and nonreturn rates between double-tagged releases recaptured the same year using plastic or metal dart tags on bluefin tuna in the northwestern Atlantic Ocean. The number of returns of double-tagged fish retaining both tags during the first year after release is n^^j , and those retaining only one tag is n^jgi. Plastic dart tags Metal dart tags Double-tagged Total returns same year Return Double-tagged Total returns same year Return Chi-square Year releases <"ddl ^"dsl> rate releases <"dd1 ^"dsl' rate value 1971 150 4 0.02667 162 5 0.03086 0.049 1972 75 6 08000 77 1 0.01299 3.884- 1973 134 20 0,14925 131 17 0.12977 0.209 1974 629 29 0.04610 666 30 0.04505 0.008 1975 50 1 0.02000 58 1 0.01724 0.011 1976 267 14 0.05243 271 26 0.09594 3.699 1977 223 4 0.01794 228 8 0.03509 1.280 Total 1,528 78 0.05105 1,593 88 0.05524 0.272 •PsO.05 Table 4. — Chi-square tests (df = 1) of equality of return and nonreturn rates by estimated age between double- tagged releases for all years 1971-77 combined, recaptured the same year, using plastic or metal dart tags on bluefin tuna in the northwestern Atlantic Ocean. The number of returns of double- tagged fish retaining both tags during the first year after release is n^j and those retaining only one tagis«dsl- Plastic dart tags Metal dart tags Estimated age at release Double-tagged releases Total returns same ("ddl+^dsl' year Return rate Double- tagged releases Total returns same ("ddl ^"dsl' year Return rate Chi-square value' 1 2 3 4-I- 641 631 212 44 1.528 29 43 4 _2 78 0.04524 0.0681 5 0.01887 0.04545 647 656 226 64 1,593 31 43 12 _2 88 0,04791 0.06555 0.05310 0.03125 0.052 0.035 3.642 0.148 182 Table 5. — Estimates of immediate (1 - p) and annual instan- taneous it) tag shedding rates for northwestern Atlantic bluefin tuna double-tagging study for all years combined (1971-77) based on a 3-yr return period using unweighted and weighted linear regression models. (The weights used in the weighted model were equated to the ratio of the number of returns of double- tagged fish during each return period to the total number of returns of double-tagged fish during all periods.) Model and tag type 1 - P L (annual) Linear regression: Plastic dart Metal dart Combined Weighted linear regression: Plastic dart Metal dart Combined 0.027 0.049 0.040 0.033 0.049 0.042 22886 0.19201 0.20452 0.19200 0.18213 0.18596 indicate the magnitude of the variances of the estimates. Our estimate of (1 - p) is slightly greater than the overall estimate of 0.027 given for bluefin tuna from the northwest Atlantic by Lenarz et al. (1973). The difference is small relative to the pre- cision of the estimates. Our estimate of (1 - p) for northwest Atlantic bluefin tuna is less than the value of 0.10 reported for Pacific yellowfin tuna by Bayliff and Mobrand (1972) and the value of 0.12 reported for North Pacific albacore by Laurs et al. (1976). Our estimate of L is less than the overall esti- mate of 0.31 reported by Lenarz et al. (1973) for bluefin tuna and the L estimate of 0.278 reported for yellowfin tuna by Bayliff and Mobrand (1972). Our L estimate is greater than the estimates of between 0.086 and 0.098 reported for albacore by Laurs et al. (1976). As previously noted, there was no significant difference in return rates found for the two types of dart tags for 1971-77. However, from examination of the data presented in Table 1, there appeared to be changes occurring in the shedding rates of each type of tag and a difference between the 1971-73 and 1974-77 time intervals. Therefore, we calcu- lated (1 - p) andL for each time interval and con- ducted chi-square tests (df = 6) for differences in returns over three recapture periods ik = 3) be- tween time intervals and between tag types (Table 6). We found significant differences between time intervals for each of the tag types and significant differences between tag types for each of the time intervals. The plastic dart tags became less efficient, i.e., L increased over the time intervals, and the metal dart tags improved, i.e., L decreased over the time intervals. The model of Chapman et al. (1965), which was Table 6. — Estimates of immediate (1 - p) and einnual instan- taneous (L) tag shedding rates for northwestern Atlantic bluefin tuna double-tagging study for time intervals 1971-73 and 1974- 77 based on a * = 3-yr return period. (A contingency table, 7x2, was constructed containing the number of double and single returns for each of the three recapture periods plus the number of nonreturns for each tag type and each time interval.) Results of chi-square tests (df = 6) for differences in double and single tag returns and total nonreturns between time intervals and tag types over a 3-yr recapture period are given. Tag type and time interval L (annual) Chi-square value Plastic dart: 1971-73 1 974-77 Metal dart: 1971-73 1974-77 1971-73: Plastic dart Metal dart 1974-77: Plastic dart Metal dart 029 0.023 0.140 0.007 029 0.140 0.023 0.007 0.14838 0.28455 0.37163 0.17242 0.14838 0.37163 0.28455 0.17242 64.286" 33.489" 18.924" 18.135" "P«0.01. modified by Bayliff and Mobrand (1972), assumes constant L over recapture periods. We decided to examine values of L over the two pairs of recap- ture periods k = (1, 2) and k = (2, 3) to determine how well our data fit the model. Since only two recapture periods were used, L and ( 1 - p) were estimated by solving two simultaneous equations. For the tag types and time intervals examined, there is an indication that L is not constant (Table 7). In fact, L increased in three out of four cases. The sequence of events could have happened due to chance alone, for if the changes in L came from a binomial distribution with P =0.5, then the prob- ability of L decreasing in three of the four cases or L increasing in three of the four cases is ^0.25. However, L during the second time period is more than 60% >L in the first time period in three cases and only 16% 2.33, P^O.Ol) differences were found between sexes within years for 1975 and 1976 and between years for each sex except 1973 and 1976 males and 1974 and 1975 females. This indicates that L^q for males and females was greater in 1974 and 1975 than in 1973 and 1976. The greatest difference was found between 1975 and 1976 when L^„ decreased approximately 4.8 cm for females and 3.6 cm for males. Significant long-term changes in L^^ have oc- curred since Wallace's (1940) studies of Chesapeake Bay croakers. The smallest mature female he observed during 1938-40 was 27.5 cm, indicating a L^^ of at least 30 cm. His collections were made during July and August; therefore, due to additional growth during early fall, 30 cm is an underestimation of Lg^ for comparison with my results obtained from September-November. Spawning The percentage frequencies of maturity stages indicate spawning commenced at least as early as the beginning of September, peaked during Oc- tober, and ended by late December. The maturity stages and sample years were combined for analysis (Table 4). The percentage of ripe ovaries remained high during September and October, then dropped to a low level in November. No ripe females were found in December. As would be expected the percentage of spawned fish (partially spent, spent, and resting) increased during the sampling period and indicated spawning was nearly completed by mid-December. Because of difficulty in assigning a specific maturity stage to testes and since ovarian development was the best indicator of spawning, males were not analyzed. The beginning of the spawning season was not sampled; however, an examination of Wallace's (1940) maturity stage data for July and August showed that over 50% of the ovaries were develop- ing (stages II and III) and <10% were ripe (stage IV). The remainder was classified as resting (stage I). Wallace made additional collections in November which showed that ovaries were either partially spent (stage VI) or spent (stage VII). His findings support this study, indicating that spawning commenced about mid-August and was completed by the end of December. The presence of small juveniles (20-40 mm TL) during April and May have led to speculations of different spawning populations and a spring spawning peak. Chao and Musick (1977) appar- ently detected a modal group "entering" the York River in May and suggested they may represent Table 3. — Matrix of 2-values (Natrella 1966) and significance for differences in L,„ (length at which 50% of specimens were mature) of male (M) and female (F) Atlantic croaker collected in 1973-76. 1973 1974 1975 1976 M F M F M F M F 1973 M 0.196 7.269" 16.214" 9.632- 17.546** 6.937** 5.779** F 9.058** 7.544** 3.196** 1974M 4.467** 2.903* 6.367** 6329** 12.546" F 1.400 12.122** 1975 M 3.400** 8.055** 15.507** F 9.972** 1976 M 0.729 •PsO.01; "PsO.OOI. Table 4. — Percentage frequency of maturity stages of female Atlantic croaker collected between Cape May, N.J., and Cape Hatteras, N.C., during 1973-76. Maturity stage Developing Ripe Partially spent Spent Resting Total Sampling interval Wallace's 16-18 29 Sept. 5-20 30 Oct- 14-17 stages Sept. 1 Oct. Oct. 1973, 6 Nov. Dec. (1940) 1975 1974 1976 1975 1975 Hand III 51 29 23 IV and V 46 51 41 12 VI 3 7 13 31 10 VII 11 17 32 28 1 2 6 25 62 42 286 448 196 51 192 progeny from a different spawning population. Haven (1957) found 20-30 mm fish during April and concluded the spawning season extended over almost the entire year with a possible spring peak. The apparent 9- or 10-mo spawning season may result from little or no overwinter growth or sam- pling bias due to differential size distribution or trawl avoidance (Haven 1957; White and Chitten- den 1977; Chao and Musick 1977). Maturity ob- servations made during this study showed essen- tially all adult fish spawned during August through December and it is unlikely a spring spawning peak would occur from the Atlantic croaker population north of Cape Hatteras. Fecundity Fecundity ranged from 100,800 to 1,742,000 for fish from 196 to 390 mm TL. Preliminary plots of fish length versus fecundity indicated a curvilinear relationship and plots of fish weight and ovary weight versus fecundity appeared linearly re- lated. Therefore, fish length and fecundity were transformed to logarithms (base 10) and least squares regression lines fitted to the data by year using the equation log fecundity = logo + b (log length). Fish weight and ovary weight versus fecundity were related by the linear regression equation Y =a + bX where Y is fecundity andX is either fish weight or ovary weight. Analysis of variance indicated no significant (P«0.05) differ- ences in variance about the regression between years for each of fecundity versus length, weight or ovary weight. Analysis of covariance was used to test for between years differences in fecundity relationship. No significant (P = 0.01) difference was indicated; therefore, regression equations were calculated for pooled data. Scatter diagrams and fitted lines are shown in Figures 1-3. 18 17 16 15 14 13 o o 6 11 o > Q Z o Figure l. — Relationship between fecun- dity and total length for Atlantic croaker collected in 1973 and 1974. 10 9 8 7 6 5 4 3 2 1 - • • - log F^-2.586 + 3.361 Hog LI - r = 0.86 n=113 - Syx- 0.1394 - • - • / - f - •y - - • • jC • • " • _ • . *. • • ^^ Vt-"^** • . • ^^^-f^ • • ••• • ^ — • • • •• • J 1 1 1 1 1 ! i : _L • • • 1 \ ^ 1 1 ^ \ 190 210 230 250 270 290 310 TOTAL LENGTH imml 330 350 370 193 o o o o o Q z O ID Figure 2. — Relationship between fecun- dity and fisii weight for Atlantic croaker collected in 1973 and 1974. 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 F = -29,175 + 1,624W r =0.89 n = 113 Sy X =143,755 75 150 225 300 375 450 525 600 675 750 825 900 975 1050 FISH WEIGHT Igl The correlation coefficients for the relationships of fecundity to length, weight, and ovary weight show ovary weight was most closely associated with the variation of fecundity. Unless the ovaries are selected, however, ovary weight is the least reliable predictor of fecundity. It is the most vari- able parameter and, unless ovaries are collected at the penultimate development stage, the relation- ship of ovary weight and fecundity will vary sea- sonally. Fish weight will also vary seasonally and, when ovary weight is included with fish weight, some autocorrelation is present. For general pre- diction of fecundity, length appears to be the most reliable measure. Acknowledgments I wish to thank the many people who helped collect samples during National Marine Fisheries Service cruises and S. J. Wilk and W. G. Smith for 194 their critical reviews of early drafts. Special thanks to M. Montone for her assistance and typ- ing of this manuscript and to M. Cox for prepara- tion of the figures. Literature Cited Bagenal, T. B. 1957. The breeding and fecundity of the long rough dab Hippoglossoides platessoides (Fabr.) and the associated cycle in condition. J. Mar. Biol. Assoc. U.K. 36:339-375. Chad, L. N. 1978. A basis for classifying western Atlantic Sciaenidae (Teleostei: Perciformes). U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circ. 415, 64 p. CHAO, L. N., and J. A. MUSICK. 1977. Life history, feeding habits, and functional mor- phology of juvenile sciaenid fishes in the York River es- tuary, Virginia. Fish. Bull., U.S. 75:657-702. FINNEY, D. J. 1971. Probit analysis. 3ded. Camb. Univ. Press, Lond., 333 p. o o o d o Q Z u UJ Figure 3. — Relationship between fecun- dity and ovary weight for Atlantic croaker collected in 1973 and 1974. 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 F =8.603 + 19.966V r 0.98 n =113 Sy x = 67.958 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 OVARY WEIGHT Igl FISHER, R. A., AND F. Yates. 1964. Statistical tables for biological, agricultural and medical research. 6th ed. Oliver and Boyd, Edinb., 146 p. GROSSLEIN, M. D. 1969. Groundfish survey program of BCF Woods Hole. Commer. Fish. Rev. 31(8-9):22-30. GUTHERZ, E. J., G. M. RUSSELL, A. R. SERRA, AND B. A. ROHR. 1975. Synopsis of the northern Gulf of Mexico industrial and food fish industries. Mar. Fish. Rev. 37(7):1-11. Haven, D. S. 1957. Distribution, growth, and availability of juvenile croaker, Micropogon undulatus, in Virginia. Ecology 38:88-97. Hildebrand, S. F., and W. C. SCHROEDER. 1928. Fishes of Chesapeake Bay. Bull U.S. Bur. Fish. 43(1), 366 p. Joseph, E. B. 1972. The status of the sciaenid stocks of the middle Atlan- tic coast. Chesapeake Sci. 13:87-100. MCHUGH, J. L. 1977. Fisheries and fishery resources of New York Bight. U.S. Dep. Commer., NCAA Tech. Rep. NMFS Circ. 401, 50 p. NATRELLA, M. G. 1963. Experimental statistics. U.S. Dep. Commer., Natl. Bur. Stand. Handb. 91, 528 p. Simpson, a. C. 1951. The fecundity of the plaice. Fish. Invest. Minist. Agric. Fish. Food (G.B.) Ser. II, 17(5), 27 p. Wallace, D. H 1940. Sexual development of the croaker, Micropogon un- dulatus, and distribution of the early stages in Chesapeake Bay. Trans. Am. Fish. Soc. 70:475-482. WELSH, W. W., and C. M. BREDER, JR. 1923. Contributions to life histories of Sciaenidae of the eastern United States coast. Bull. U.S. Bur. Fish. 39:141-201. White, M. L., and M. E. Chittenden, Jr. 1977. Age determination, reproduction, and population dynamics of the Atlantic croaker, Micropogonias un- dulatus. Fish. Bull., U.S. 75:109-123. Wallace W. Morse Northeast Fisheries Center Sandy Hook Laboratory National Marine Fisheries Service, NOAA Highlands, NJ 07732 195 COMPARISON OF SAMPLING DEVICES FOR THE JUVENILE BLUE CRAB, CALLINECTES SAPIDUS^ The behavior of the blue crab, Callinectes sapidus Rathbun, in the Chesapeake Bay varies consider- ably with age, temperature, and molting cycle. These behavioral differences make efforts difficult to sample effectively the population densities in the Chesapeake Bay and its tributaries. No single gear type appears to sample effectively the blue crab during winter and summer at all depths and types of bottom. During winter blue crabs burrow in the mud in the deeper channels of Chesapeake Bay (Churchill 1917). This pattern is the basis for an active winter dredge fishery in the lower portion of the bay (Van Engel 1962). During a 3-yr survey of blue crabs, Lippson^ found that juveniles were also present in deeper waters in winter. Comparative effectiveness of two dredges for winter sampling of juvenile and adult blue crabs was reported by Sulkin and Miller (1975). Blue crabs move about in relatively shallow water in warm weather presumably because of the abundance of food here and for protection among submerged aquatics while in the soft shell condition. During the summer 7.3 m otter trawls have been found to be an effective gear to sample the adult population of blue crabs (Lippson see footnote 2). The otter trawl, with a small stretch mesh (0.6 cm) liner in the cod end, is also effective for catching juveniles in deeper water; however, juveniles spend much of their time in shallow waters during the warmer months. The push net (Figure 1), beach seine, and small otter trawls have all been used with some degree of success in this shallow region. It is the purpose of this study to compare the effectiveness of the push net, otter trawl, and crab scrape (Figure 2) in catching juvenile blue crabs in shallow water. Methods and Results Smith Island in the Chesapeake Bay has extensive grassy (Zostera marina) beds which are ideal habitats for juvenile crabs (Stevenson and Confer 1978). This region was chosen to compare 'Contribution No. 992HPEL from the Center for Environmen- tal and Estuarine Studies, University of Maryland. ^Lippson, R. L. 1969. Blue crab study in Chesapeake Bay-Maryland. Nat. Resour. Inst. Q. Prog. Rep. 3, Ref No 69-33B:l-13. the catch effectiveness among a 3.7 m otter trawl, 81.3 cm push net, and a 96.5 cm modified crab scrape during summer 1975. The otter trawl opened to a working width of 3.6 m. The gear was towed by the RV Chelae in depths of 1-2 m for 0.7 km. The cod end was lined with 0.6 cm stretch mesh netting. The trawl door size was 30.5 cm X 61.0 cm and the length of the bridle was 45.7 m. The push net had a steel frame 81.3 cm wide and 60.9 cm high fitted with a 0.6 cm stretch mesh bag. The leading edge had a 7 .6 cm diameter pipe which Figure l. — Push net used for blue crab fishing with the roller bar on the leading edge. 196 FISHERY BULLETIN: VOL. 78, NO. 1, 1980. 'S*****! Figure 2. — The crab scrape used for blue crab fishing. rolled over the bottom. The handle was 1.5 m long. The net was manually pushed along a 0.7 km course, waist to chest deep, and parallel to the shoreline. The crab scrape, used commercially for catching shedding crabs, had a metal frame 96.5 cm wide and 38.1 cm high. The lead bar on the crab scrape has no teeth, a basic difference between it and a dredge. A 3.8 cm twine net 182.9 cm long was fitted to this frame. The crab scrape is towed from a shallow-draft boat over grassy beds. The crab scrape used in this study was modified by fitting it with a 0.6 cm stretch mesh net to retain small (>3 mm) crabs. The otter trawl and crab scrape were towed simultaneously beside each other from two small outboard motorboats for 6 min at an engine speed of 2,000 r/min. The push net was then pushed parallel to the trawl and crab scrape tows over the same distance but closer to shore. The depths for the trawl and crab scrape tows ranged from 1 to 2 m whereas the push net sampled in depths of 0.6-1.1 m. Eighteen samples were collected for each gear type. The sex and size class of crabs were determined after each tow. Crab size was determined using carapace width from one lateral spine tip to the other. Crabs >60 mm wide were excluded from consideration in this study because they were not in the most recent year class. Three size classes were used: class I measured 1 to 20.0 mm; class II, 20.1 to 40.0 mm; and class III, 40.1 to 60.0 mm. The mean number of crabs per square meter is shown in Figure 3. It is apparent that the trawl is comparatively ineffective for classes I and II. The trawl and push net are about equally as effective for class III although neither is as effective as the modified crab scrape for classes I, II, or III. The TOTAL AREA FISHED A1552Mi_)08 3 6M^ I2870m2 to CO < u O LU CO ID z 24 20 16 12 8 Trawl Push Net Crab Scrape I il ill I II III SIZE CLASS I II Figure 3. — Mean number of blue crabs per square meter for each haul and total area fished by each gear. Size class (carapace width) I = 1-20.0 mm, II = 20. 1-40.0 mm, and III = 40. 1-60.0 mm. crab scrape is the most effective gear for sampling juvenile blue crabs. 197 Discussion Acknowledgments In evaluating various gear for sampling juvenile blue crabs, a variety of factors should be considered, such as catch effectiveness, gear cost, ease of handling, and person hours. It generally requires two persons to efficiently operate an otter trawl from a small outboard motorboat. Handling an otter trawl from a small outboard motorboat is not only difficult but dangerous as the net can become fouled in the propeller. If the net fills with mud or too much debris, it is impossible to bring the gear on board and the catch must be sacrificed. The push net is operable by one person and snags are infrequent. Mud, as well as high rooted aquatics, makes pushing the net difficult. The push net is effective in shallow water (Strawn 1954). Clear shallow water, however, decreases the effectiveness as many small crabs see the net approaching and swim out of its path (pers. obs.). The crab scrape can be easily handled by one person and seldom becomes snagged (pers. obs. and observations of commercial crabbers). The cost of a 3.7 m otter trawl is about $150.00. The push net cost varies. They are not available commercially and must be constructed, usually by a local blacksmith. The bag may be cut from a ripped beach seine net. The approximate cost of the crab scrape is $55.00. Although gear cost, ease of handling, and hours involved are considered in gear selection, the most important factor is catch effectiveness. The push net was more effective catching small blue crabs than the trawl but the modified crab scrape was more effective than either the push net or the trawl when sampling in shallow water. Considering all pertinent factors, it would seem that the crab scrape is the preferred gear for quantitative studies of juvenile crab abundance. We thank Fred Dobbs and Marion Ross for their valuable assistance during the field sampling, Frances Younger for her technical help in the preparation of the graphs, and Leo Minasian and Donna L. Smawley for their photographic assistance. Our thanks are also extended to Nancy Robbins and Ida Marbury for their help in the preparation of the manuscript. This research has been supported by the National Marine Fisheries Service and the State of Maryland Fisheries Administration (Project No. 3-186-R-2, Contract No. 04-5-043-43). Literature Cited Churchill, E. p., Jr. 1917. Life-history of the blue crab of the Chesapeake Bay. Off. Bull., Conserv. Comm. Md. 2:11-18. STRAWN, K. 1954. The pushnet, a one-man net for collecting in attached vegatation. Copeia 1954:195-197. SULKIN, S. D., AND R. E. MILLER. 1975. Modified commercial crab and oyster dredges as sampling devices for the blue crab Callinectes sapidus Rathbun. Chesapeake Sci. 16:137-139. STEVENSON, J. C, AND N. M. CONFER. 1978. Summary of available information on Chesapeake Bay submerged vegatation. U.S. Fish Wildl. Serv., Off. Biol. Serv. FWS/OBS-78/66, 335 p. Van engel, w. a. 1962. The blue crab and its fishery in Chesapeake Bay. Part 2 - Types of gear for hard crab fishing. Commer. Fish. Rev. 24(9):1-10. ROBERT E. Miller Horn Point Environmental Laboratories Center for Environmental and Estuarine Studies University of Maryland P.O. Box 775, Cambridge, MD 21613 DOUGLAS w. Campbell Pamela J. Lunsford Fisheries Administration Maryland Department of Natural Resources Tawes State Office Building Annapolis, MD 21401 198 NOTICES NOAA Technical Reports NMFS published during the last 6 mo of 1979. Circular 424. Guide to the leptocephali (Elopiformes, Anguil- liformes, and Notacanthiformes). By David G. Smith. July 1979, iv + 39 p., 54 fig., 1 app. 425. Marine flora and fauna of the northeastern United States. Arthropoda: Cirripedia. By Victor A. Zullo. April 1979, iii + 29 p., 40 fig., 1 table. For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402 - Stock No. 003-017-00453-8. 426. Synopsis of biological data on the rock crab. Cancer irroratus Say. By Thomas E. Bigford. May 1979, V + 26 p., 11 fig., 21 tables. Also FAO Fisheries Synopsis No. 123. 427. Ocean variability in the U.S. Fishery Conserva- tion Zone, 1976. By JulienR. Goulet, Jr. and Elizabeth D. Haynes, editors. July 1979, iv + 362 p. 428. Morphological comparisons of North American sea bass larvae (Pisces: Serranidae). By Arthur W. Kendall, Jr. August 1979, iv -f 50 p., 43 fig., 9 tables, 1 app. table. 429. Synopsis of biological data on tunas of the genus Euthynnus . By Howard O. Yoshida. October 1979, iv -l- 57 p., 40 fig., 30 tables. Also FAO Fisheries Synopsis No. 122. Special Scientific Report — Fisheries 735. History of the fishery and summary statistics of the sockeye salmon, Oncorhynchus nerka, runs to the Chignik Lakes, Alaska, 1888-1966. By Michael L. Dahlberg. August 1979, iv + 16 p., 15 fig., 11 tables. For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402 - Stock No. 003-017-00459-7. 736. A historical and descriptive account of Pacific coast anadromous salmonid rearing facilities and a summary of their releases by region, 1960-76. By Roy J. Wahle and Robert Z. Smith. September 1979, iv + 40 p., 15 fig., 25 tables, 2 app. tables. For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402 - Stock No. 003-017-00460-1. 737. Movements of pelagic dolphins (Stenella spp.) in the eastern tropical Pacific as indicated by results of tagging, with summary of tagging operations, 1969- 76. By W. F. Perrin, W. E. Evans, and D. B. Holts. September 1979, iii + 14 p., 9 fig., 8 tables. For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402 - Stock No. 003-017-00462-7. 738. Environmental baselines in Long Island Sound, 1972-73. By R. N. Reid, A. B. Frame, and A. F. Draxler. December 1979, iv -i- 31 p., 40 fig., 6 tables. For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402 - Stock No. 003-017-00466-0. 739. Bottom-water temperature trends in the Middle Atlantic Bight during spring and autumn, 1964-76. By Clarence W. Davis. December 1979, iii + 13 p., 10 fig., 9 tables. For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402 - Stock No. 003-017-00467-8. Most NOAA publications are available by purchase from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. Individual copies of NOAA Technical Reports (in limited numbers) are available free to Federal and State government agencies and may be obtained by writing to User Services Branch (D822), Environmental Science Information Center, NOAA, Rockville, MD 20852. 199 INFORMATION FOR CONTRIBUTORS TO THE FISHERY BULLETIN Manuscripts submitted to the F/.s'/ierviBi///£'^/;? 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 U.S. Government Printing Office Style Manual. Fish names follow the style of the American Fisheries Society Special Publi- cation No. 6, A List of Common and Scientific Names of Fishes from the United States and Canada. Third Edition. 1970. Text footnotes should be typed separately from 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 photog- raphy to 5''4 inches (for single-column figures, allowing for 5CKy 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 Office will request them. Each table should start on a separate page. Consistency in headings and format is desirable. Vertical rules should be avoided, as they make the tables more expensive to print. Footnotes in tables should be numbered sequentially in arable numerals. To avoid confusion with powers, they should be placed to the lefi of numerals. Acknowledgments, if included, are placed at the end of the text. Literature is cited in the text as: Lynn and Reid 11968) 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 Bio- logical Abstracts List of Serials with Title Abbrevi- ations. (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), %, %o,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. Jay C. Quasi, Scientific Editor Fishery Bulletin Northwest and Alaska Fisheries Center Auke Bay Laboratory, National Marine Fisheries Service, NOAA P.O. Box 155, Auke Bay, AK 99821 Fifty separates will be supplied to an author free of charge and 100 supplied to his organiza- tion. No covers will be supplied. Contents-Continued .4-.. fi- . ;-^' ilv TESTA VERDE, SALVATORE A., and JAWfes G. MEAD: Southern distribution of the Atlantic whitesided dolphin, Lagenorhynchus acutus, in the western North Atlantic , .- 167 MATARESE. ANN C, and DAVID L. STEIN. Additional records of the sculpin Psychrolutes phrictus in the eastern Bering Sea and off Oregon 169 ODELL, DANIEL K., EDWARD D. ASPER, JOE BAUCOM, and LANNY H. COR- NELL. A recurrent mass stranding of the false killer whale, Pseudorca crassi- dens, in Florida 171 BRANSTETTER, STEVEN, and ROBERT L. SHIPP. Occurrence of the finetooth shark, Carcharhinus isodon, off Dauphin Island, Alabama 177 BAGLIN, RAYMOND E., JR., MARK I. FARBER, WILLIAM H. LENARZ, and JOHN M. MASON, JR. Shedding rates of plastic and metal dart tags from Atlantic blue- fin tuna, Thunnus thynnus 179 HAYNES, JAMES M., and ROBERT H. GRAY. Influence of Little Goose Dam on upstream movements of adult chinook salmon, Oncorhynchus tshawytscha 185 MORSE, WALLACE W. Maturity, spawning, and fecundity of Atlantic croaker, Micropogonias undulatus, occurring north of Cape Hatteras, North Carolina 190 MILLER, ROBERT E., DOUGLAS W. CAMPBELL, and PAMELA J. LUNSFORD. Comparison of sampling devices for the juvenile blue crab, Callinectes sapidus . . . 196 Notices NOAA Technical Reports NMFS published during the last 6 mo of 1979 199 ft GPO 696-404 '^'-^TES O^ ^ Bulletin MarineBioiogical Laboratory LIBRARY DEC 2 198) Wood*; Hu i t!, Ma^ii. r Vol. 78, No. 2 April 1980 RANDALL, JOHN E. A survey of ciguatera at Enewetak and Bikini, Marshall Islands, with notes on the systematics and food habits of ciguatoxic fishes 201 SMYTH, PETER 0. Callinectes (Decapoda: Portunidae) larvae in the Middle Atlan- tic Bight, 1975-77 251 HUPPERT, D. D. An analysis of the United States demand for fish meal 267 POTTHOFF, THOMAS. Development and structure of fins and fin supports in dol- phin fishes Coryphaena hippurus and Coryphaena equiselis (Coryphaenidae) 277 KNIGHT, MARGARET D. Larval development of Euphausia eximia (Crustacea: Euphausiacea) with notes on its vertical distribution and morphological divergence between populations 313 STONER, ALLAN W. Feeding ecology of Lagodon rhomboides (Pisces: Sparidae): variation and functional responses 337 MEAD, JAMES G., DANIEL K. ODELL, RANDALL S. WELLS, and MICHAEL D. SCOTT. Observations on a mass stranding of spinner dolphin, Stenella longiros- tris , from the west coast of Florida 353 EBELING, ALFRED W., RALPH J. LARSON, WILLIAM S. ALEVIZON, and RICHARD N. BRAY. Annual variability of reef-fish assemblages in kelp forests off Santa Barbara, California 361 PIETSCH, THEODORE W, and JEFFREY A. SEIGEL. Ceratioid anglerfishes of the Philippine Archipelago, with descriptions of five new species 379 RICHARDSON, SALLY L., JEAN R. DUNN, and NANCY ANNE NAPLIN. Eggs and larvae of butter sole, Isopsetta isolepis (Pleuronectidae), off Oregon and Washington 401 WEINSTEIN, MICHAEL R, SIDNEY L. WEISS, RONALD G. HODSON, and LAWRENCE R. GERRY. Retention of three taxa of postlarval fishes in an inten- sively flushed tidal estuary, Cape Fear River, North Carolina 419 OLIVER, JOHN S., PETER N. SLATTERY, LARRY W. HULBERG, and JAMES W NYBAKKEN. Relationships between wave disturbance and zonation of benthic invertebrate communities along a subtidal high-energy beach in Monterey Bay, California 437 FERRARO, STEVEN R Daily time of spawning of 12 fishes in the Peconic Bays, New York 455 BULLARD, FERN A., and JEFF COLLINS. An improved method to analyze tri- methylamine in fish and the interference of ammonia and dimethylamine 465 (Continued on back cover) U.S. DEPARTMENT OF COMMERCE Philip M. Klutznick, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION Richard A. Frank, Administrator Terry L. Leitzell, Assistant Administrator for Fistieries NATIONAL MARINE FISHERIES SERVICE Fishery Bulletin The Fishery Bulletin carries original research reports and technical notes on investigations in fishery science, engineering, and economics. The Bulletin of the United States Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries in 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941 . Separates were issued as documents through volume 46; the last document was No. 1103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate appeared as a numbered bulletin. A new system began in 1963 with volume 63 in which papers are bound together in a single issue of the bulletin instead of being issued individually. Beginning with volume 70, number 1, Januairy 1972, the Fishery Bulletin became a periodical, issued quarterly. In this form, it is available by subscription from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. It is also available free in limited numbers to libraries, research institutions. State and Federal agencies, and in exchange for other scientific publications. EDITOR Dr. Jay C. Quast Scientific Editor, Fishery Bulletin Northwest and Alaska Fisheries Center Auke Bay Laboratory National Marine Fisheries Service, NOAA P.O. Box 155, Auke Bay, AK 99821 Editorial Committee Dr. Elbert H. Ahlstrom Dr. Merton C. Ingham National Marine Fisheries Service National Marine Fisheries Service Dr. Bruce B. Collette Dr. Reuben Lasker National Marine Fisheries Service National Marine Fisheries Service Dr. Edward D. Houde Dr. Jerome J. Pella University of Miami National Marine Fisheries Service Dr. Sally L. Richardson Gulf Coast Research Laboratory Kiyoshi G. Fukano, Managing Editor The Fishery Bulletin (USPS 090-870) is published quarterly by Scientific Publications Office, National Marine Fisheries Service, NOAA, Room 336, 1700 Westtake Avenue North, Seattle, WA 98109. Controlled circulation paid to Finance Department, USPS, Washington, DC 20260. Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated. The Secretary of Commerce has determined that the publication of this periodical 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 March 1982. Fishery Bulletin CONTENTS Vol. 78, No. 2 April1980 RANDALL, JOHN E. A survey of ciguatera at Enewetak and Bikini, Marshall Islands, with notes on the systematics and food habits of ciguatoxic fishes 201 SMYTH, PETER O. Callinectes (Decapoda: Portunidae) larvae in the Middle Atlan- tic Bight, 1975-77 251 HUPPERT, D. D. An analysis of the United States demand for fish meal 267 POTTHOFF, THOMAS. Development and structure of fins and fin supports in dol- phin fishes Coryphaena hippurus and Coryphaena equiselis (Coryphaenidae) 277 KNIGHT, MARGARET D. Larval development of Euphausia eximia (Crustacea: Euphausiacea) with notes on its vertical distribution and morphological divergence between populations 313 STONER, ALLAN W. Feeding ecology of Lagodon rhomboides (Pisces: Sparidae): variation and functional responses 337 MEAD, JAMES G., DANIEL K. ODELL, RANDALL S. WELLS, and MICHAEL D. SCOTT. Observations on a mass stranding of spinner dolphin, Stenella longiros- tris, from the west coast of Florida 353 EBELING, ALFRED W., RALPH J. LARSON, WILLIAM S. ALEVIZON, and RICHARD N. BRAY. Annual variability of reef-fish assemblages in kelp forests off Santa Barbara, California 361 PIETSCH, THEODORE W, and JEFFREY A. SEIGEL. Ceratioid anglerfishes of the Philippine Archipelago, with descriptions of five new species 379 RICHARDSON, SALLY L., JEAN R. DUNN, and NANCY ANNE NAPLIN. Eggs and larvae of butter sole, Isopsetta isolepis (Pleuronectidae), off Oregon and Washington 401 WEINSTEIN, MICHAEL R, SIDNEY L. WEISS, RONALD G. HODSON, and LAWRENCE R. GERRY. Retention of three taxa of postlarval fishes in an inten- sively flushed tidal estuary. Cape Fear River, North Carolina 419 OLIVER, JOHN S., PETER N. SLATTERY, LARRY W. HULBERG, and JAMES W. NYBAKKEN. Relationships between wave disturbance and zonation of benthic invertebrate communities along a subtidal high-energy beach in Monterey Bay, California 437 FERRARO, STEVEN R Daily time of spawning of 12 fishes in the Peconic Bays, New York 455 BULLARD, FERN A., and JEFF COLLINS. An improved method to analyze tri- methylamine in fish and the interference of ammonia and dimethylamine 465 (Continued on next page) Seattle, Washington 1980 For sale by the Superintendent of Documents, US. Government Printing Office, Washington, DC 20402— Subscription price per year: $12.00 domestic and $15.00 foreign. Cost per single issue: $3.00 domestic and $3.75 foreign. Contents-continued O'CONNELL, CHARLES P. Percentage of starving northern anchovy, Engraulis mordax, larvae in the sea as estimated by histological methods 475 EBEL, WESLEY J. Transportation of chinook salmon, Oncorhynchus tshawytscha, and stee\h.eaid,Salmo gairdneri, smolts in the Columbia River and effects on adult returns 491 BENIRSCHKE, K., MARY L. JOHNSON, and ROLF J. BENIRSCHKE. Is ovula- tion in dolphins, Stenella longirostris and Stenella attenuata, always copulation- induced? 507 Notes PEARCY, WILLIAM G. A large, opening-closing midwater trawl for sampling oceanic nekton, and comparison of catches with an Isaacs-Kidd midwater trawl . . 529 COE, JAMES M., and WARREN E. STUNTZ. Passive behavior by the spotted dolphin, Stenella attenuata, in tuna purse seine nets 535 KRAEUTER, JOHN N., and MICHAEL CASTAGNA. Effects of large predators on the field culture of the hard clam, Mercenaria mercenaria 538 PARKER, KEITH. A direct method for estimating northern anchovy, Engraulis mordax, spawning biomass 541 PITCHER, KENNETH W Food of the harbor seal, Phoca vitulina richardsi, in the Gulf of Alaska 544 JOHNSON, JAMES H. Production and growth of subyearling coho salmon, On- corhynchus kisutch, chinook salmon, Oncorhynchus tshawytscha, and steelhead, Salmo gairdneri , in Orwell Brook, tributary of Salmon River, New York 549 Vol. 78, No. 1 was published on 28 August 1980. The National Marine Fisheries Service (NMFS) does not approve, rec- ommend or endorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to NMFS, or to this publication furnished by NMFS, in any advertising or sales pro- motion 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. A SURVEY OF CIGUATERA AT ENEWETAK AND BIKINI, MARSHALL ISLANDS, WITH NOTES ON THE SYSTEMATICS AND FOOD HABITS OF CIGUATOXIC FISHES^ John E. Randall^ ABSTRACT A total of 551 specimens of 48 species of potentially ciguatoxic fishes from Enewetak and 256 specimens of 23 species from Bikini, Marshall Islands, were tested for ciguatoxin by feeding liver or liver and viscera from these fishes to mongooses at 10% body weight (except for sharks, when only muscle tissue was used.) The fishes are representatives of the following families: Orectolobidae, Carcharhinidae, Dasyatidae, Muraenidae, Holocentridae, Sphyraenidae, Mugilidae, Serranidae, Lutjanidae, Leth- rinidae, Carangidae, Scombridae, Labridae, Scaridae, Acanthuridae, and Balistidae. The species selected were all ones for which toxicity can be expected, including the worst offenders from reports of ciguatera throughout Oceania; only moderate to large-sized adults were tested. In all, 37.3% of the fishes fi-om Enewetak and 19. 7"%- from Bikini gave a positive reaction for ciguatoxin. Because liver and other viscera are more toxic than muscle, the percentage of positive reactions at the level which might cause illness in humans eating only the flesh of these fishes collectively would drop to 16.2 for Enewetak and 1 .4 for Bikini. This level of toxicity is not regarded as high for Pacific islands, in general . Because ciguatoxin is acquired through feeding, the food habits of these fishes were investigated. Most of the highly toxic species, including seven of the eight causing severe illness or death in the test animals (Lycodontis javanicus , Cephalopholis argus, Epinephelus hoedtii, E. microdon , Plectropomus leopardus, Aprion virescens, and Lutjanus bohar) are primarily piscivorous. Some such as Lethrinus kallopterus (which also produced a mongoose death) feed mainly on echinoids and mollusks. Among the larger herbivorous fishes that were tested, only one individual ofKyphosus and two ofScarus caused a weak reaction in the test animals. In view of the importance of correct identification of the ciguatoxic fishes, diagnostic remarks and an illustration are provided for each of the species tested. Some alteration in scientific names was necessary for a few of the fishes. The Marshall Islands are the easternmost islands of Micronesia and of the Trust Territory of the Pacific Islands. They consist of 34 low islands, most of which are atolls, and numerous reefs which occur between lat. 4° 30' and 15° N and long. 161° and 173° E. They lie in two parallel groups in a northeast-southwest direction, the easternmost being the Ratak ("Sunrise") Chain and the west- ernmost the Ralik ("Sunset") Chain. Two tj^jes offish poisoning are known from the Marshall Islands: tetraodontid (puffer) poisoning (Hiyama 1943; Yudkin 1944; Halstead 1967) and ciguatera. This paper is restricted to the latter toxemia. It results from the ingestion of a great variety of tropical reef and semipelagic fishes. Ciguatoxin is thermostable, hence unaffected by cooking or freezing of the fishes. It is not the result of decomposition but is present to a varying degree 'Contribution of the Mid-Pacific Research Laboratory. ^Bemice P. Bishop Museum, Box 19000-A, Honolulu, HI 96819. Manuscript accepted October 1979. FISHERY BULLETIN: VOL. 78, NO. 2, 1980. in the different tissues of fishes when entirely fresh. The severity of the illness and the symptomatology depend upon the concentration of the toxin and the amount of fish eaten; fatalities are rare. Symptoms appear about 1-10 h after a toxic fish is consumed; those most commonly given are: weakness or prostration; diarrhea; tingling or numbness of the lips, hands, and feet; confusion of the sensations heat and cold; nausea; joint and muscular pain; inability to coordinate voluntary muscular movements; difficulty in breathing; burning urination; and itching. Probably the most common diagnostic symptoms are unpleasant tingling sensations of the palms of the hands and soles of the feet on contact with cool materials and the feeling of heat when cold objects are touched or cold liquids taken into the mouth. Light cases may not exhibit these sensations, however. The Marshall Islands have long been known to harbor ciguatoxic fishes. The earliest report from these islands seems to be that of Steinbach ( 1895) who wrote of fishes being toxic on the west side of 201- FISHERY BULLETIN: VOL. 78, NO. 2 the lagoon at Jaluit. Becke (1901) recorded poisonous fishes from Rahk (Ebon), the south- ernmost atoll in the Marshalls (reference from Halstead 1967). With the takeover of the Mar- shalls by Japan at the start of World War I ( 1914), the documentation of ciguatera at these islands shifted to the Japanese. Hishikari ( 1921), Matsuo (1934), and Hiyama (1943) published on poisonous fishes at Jaluit. Some fishes in the vicinity of Utirik Island, Utirik Atoll, have been reported as poisonous (Hydrographic Office, U.S. Navy 1945). Historically, Jaluit was the principal atoll of the Marshalls. It was the center of government, had the greatest shipping activity, and the highest population (1,683 in 1933). As a result of military activity during World War II, Kwajalein (the largest atoll in the world) and Majuro became more important. Majuro is the District Adminis- trative Center of the islands. By 1958 the popula- tion was 3,336 (compared with 783 in 1935), whereas the population at Jaluit had declined to 1,112 in 1958 (Robson 1959). Concurrent with the buildup in population and commerce at Majuro and Kwajalein was the ap- pearance of ciguatera (or at least the first records in the literature of its incidence). Halstead and Lively (1954) reported one death and five persons seriously ill from the consumption of a moray eel at Kwajalein. Bartsch et al. (1959, table 2) documented the marked increase in cases of fish poisoning at the hospital at Majuro; there were 22 in 1955 (all in the last half of the year), 100 in 1956, and 211 in 1957. Banner and Helfrich ( 1964) stated that the atolls in the Marshall Islands where poisonous fishes are most commonly found are Kwajalein, Mille, Ailinglaplap, Jaluit, and Majuro. They tested numerous fishes of many species from Enewetak (formerly spelled Eniwetok) collected in 1958, but none were found to be toxic. Balaz,^ on the other hand, interviewed Chief Johannes, the last remaining traditional chief of the Enewetak people, at Majuro on 15 March 1974. Johannes stated that poisonous fishes were known at the atoll at the time of his departure in 1946 from the islands of the eastern side between the deep passage and the northern end. It should be pointed out, however, that a short-term field survey of ciguatera at an atoll, such as that carried out by Banner and Helfrich, is difficult to equate to the continuous human bioas- 'George H. Balaz, Research Associate, Hawaii Institute of Marine Biology, pers. commun. 1974. 202 say of a population of native people dependent on fishes as their principal source of protein. One should also emphasize that even in highly toxic sectors, the percentage of poisonous fishes that will cause ciguatera when eaten is small. Never- theless, only a few cases in an area may be needed to prevent residents from fishing in that area. The atolls of Enewetak and Bikini are located at the northern end of the Ratak Chain 165 mi apart between lat. 11° and 12° N. The native people of Bikini were moved from their island to Rongerik Atoll and later to Kill Island when a series of nuclear explosion tests were carried out by the United States beginning in 1946. The people of Enewetak were transferrred to Ujelang Atoll in 1947 for the same reason. When repatriation of these Micronesian people was contemplated, a question arose as to the current level of toxicity of the food fishes of Bikini and Enewetak. Fluctuation in the toxicity of fishes in reef ecosystems has long been recognized (Banner and Helfrich 1964; Cooper 1964; Halstead 1967; and Helfrich and Banner 1968). Furthermore, Randall (1958) hypothesized that disruptions of the marine environment resulting in the creation of new sur- faces (particularly the repetitive formation of new surfaces) in potentially ciguatoxic areas may be linked to outbreaks of the toxemia. This hypothesis has received support from Cooper (1964) who related toxic sectors in the Gilbert Is- lands to the locations of wrecks and anchorages, by Helfrich et al. (1968) who documented the first outbreak of ciguatera at Washington Island, Line Islands, following the wreck of the MS Southbank in late 1964, and by Bagnis (1969) who reported numerous cases of ciguatera at the previously nontoxic atoll of Hao in the Tuamotu Archipelago after the atoll was altered as a staging area for nuclear testing at Mururoa. de Sylva (1963) misinterpreted this hypothesis. He stated that Randall found poisonous fishes in estuarine areas. On the contrary, Randall re- ported toxic fishes in the Society Islands from cer- tain areas of slight or intermittent freshwater drainage which are ordinarily flushed with clear water from the open sea. During periods of heavy rain the freshwater runoff to a normally marine habitat may cause death of stenohaline sessile marine animals, thus forming a new surface for benthic growth. After stating that the basic toxic organism must be benthic, Randall (1958) wrote, "Since obli- gately herbivorous fishes and detritus-feeding RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS fishes may be poisonous, the toxic organism would most likely be an alga, a fungus, a protozoan, or a bacterium." He added that if it were an alga it must be fine because certain potentially toxic sur- geonfishes are unable to feed on coarse types. Of the algae, he wrote that blue-greens were the most probable source of ciguatoxin. Yasumoto et al. (1977), however, have shown that the "likely culprit" of ciguatera is a di- noflagellate which lives attached to dead coral or benthic algae. Though identified initially as a new species of Diplopsalis, it was later shown (Taylor 1979) to be a new genus as well. Subsequently, Adachi and Fukuyo (1979) named it Gambierdis- cus toxicus. Although a fat-soluble toxin, later identified as ciguatoxin, was isolated from wild dinoflagellates of this species, this organism pro- duced "... only meager amounts of ciguatoxin, if any,. . ." under culture conditions (Yasumoto et al. 1979). The author first visited Enewetak in 1967, then in use as the terminus of a missile range by the United States. The resident personnel had been informed of the hazard of ciguatera, and local reef fishes were not served in the mess. In spite of the warning, some cases of ciguatera still occur, espe- cially with the crews on supply ships to the island who sometimes catch and eat fishes, particularly red snapper, Lutjanus bohar, from the vessels be- fore they could be informed of the danger. The most recent case was reported by Roth.'* In 1968 six residents of the atoll ate a large reddish brown grouper with small blue spots (probably Plectropomus leopardus) that one of them had caught off the garbage pier at the southwest end of Enewetak Island. They had asked a cook in the mess hall to prepare the fish for a meal. The cook refused, explaining that the species was one which could make them sick. Dis- believing, the men took the fish to their quarters and cooked it themselves. They all contracted ciguatera and were hospitalized (Spillman). These cases offish poisoning and the knowledge that the marine environments of both Enewetak and Bikini have indeed been disrupted underlined the need for a survey of ciguatera at the two atolls. The survey was supported by the U.S. Energy "Robert M. Roth, Capt. USAF, MC, Command Surgeon, Joint Task Group, Enewetak, described the illness (symptoms typical of ciguatera) of Francisco Romolor, age 28, a civilian deckhand on the cargo ship Muskingum, following ingestion of a red snap- per caught in the lagoon (pers. commun. 2 June 1978). ^Louis C. Spillman, Jr., Chief Medical Officer, Enewetak, pers. commun. 1968. Research and Development Administration (now Department of Energy). The field work at Enewetak was based at the Mid-Pacific Marine Laboratory, and the fishing at Bikini was carried out from the RV Liktahur.^ The testing of fishes for ciguatoxin was done at the Hawaii Institute of Marine Biology, University of Hawaii, under the supervision of A. H. Banner. Six fishing expeditions of 2 to 4 wk duration were dispatched from Hawaii to Enewetak within the period September 1974-May 1978. There were four fishing cruises to Bikini (fishing periods of 3-7 days at the atoll) from December 1974 to July 1976. In addition, 12 potentially toxic L. bohar were caught from the Liktanur at the atoll of Rongelap in November 1975. Fishes were collected by spearing, hook and line, trolling lures, explosives, and the ichthyocide rotenone. The specimens were held in chests of crushed ice until they were returned either to the Mid-Pacific Marine Laboratory or the Liktanur. They were then measured and weighed, tagged with a metal tag, and a sample taken for testing which included the liver, other viscera, and mus- cle. A data sheet was filled out for each specimen; the upper half of each sheet was used for field data and the lower half to record the testing for toxicity. A chart of the atoll was printed on the back of each data sheet (separate sets of sheets were main- tained for Enewetak and Bikini) so that the local- ity of capture could be recorded. At Enewetak the entire fish was frozen after the sample was taken for testing. Aboard the Liktanur only the samples were retained. The Enewetak specimens, which proved to be highly toxic (rated 4 or 5, see below), were transported frozen to the University of Hawaii for use in biochemical and pharmacologi- cal research on ciguatoxin; the remaining fishes were either discarded or used as bait or chum. For the testing, the samples of fish were fed to mongooses (Herpestes mungo) in single meals at 10% body weight. The mongoose is a good animal for the bioassay of ciguatoxin (Banner et al. 1960) because it has a symptomology similar to humans suffering from ciguatera, it retains a meal of toxic fishes (in contrast to cats which are prone to re- gurgitate fish when it is very poisonous), and be- cause of its availability in Hawaii (where it is regarded as a pest). The mongoose symptoms were «The RV Liktanur is a converted U.S. Navy LCM, 115 ft in length, operated then by the U.S. Energy Research and De- velopment Administration as a research and supply vessel. 203 FISHERY BULLETIN: VOL. 78, NO. 2 divided into five progressive categories from 1 (diarrhea, slight weakness, and flexion of the forelimbs) to 5 (death within 48 h). The lack of symptoms was recorded as 0. The tests which re- sulted in a reaction of 4 or 5 were repeated on other mongooses if sufficient material was available. Also any questionable or unexpected tests were repeated. Two tests were run on most of the fishes, one based on the feeding of liver or liver and viscera to the mongooses and one on muscle tissue. The liver of a toxic fish invariably gave a stronger reaction than muscle. A reaction of 3 to liver feeding would generally elicit a reaction of 1 with flesh. Helfrich et al. (1968) found liver more than 50 times as toxic as the muscle tissue of L. bohar. The remain- ing viscera are also more toxic than somatic mus- cle. Since the liver alone was often insufficient in mass to provide a meal to a mongoose at 10% body weight, it was usually necessary to combine it with other viscera (generally alimentary tract tissue). The level of toxicity reported herein is from the liver-viscera feeding, with the exception of sharks. Shark liver may cause a toxemia from the high level of vitamin A. Furthermore, it was noted that mongooses will either not eat shark liver or will not consume enough to equal 10% of their body weight. Thus the toxicity data on sharks are based on muscle tissue alone. Once a mongoose exhibited symptoms of cigua- tera, it was not used again for testing. If it showed no symptoms at all, it was used a second time, but only after a period of at least 1 wk had elapsed. No mongooses were fed potentially toxic fish more than three times even when no symptoms were elicited. The reason for this is the known tendency for ciguatoxin to accumulate in a test animal. Though a fish may cause no illness when eaten, it may still have some toxin at the subsymptomatic level. Eating several such fishes in succession might result in a positive test for the last one, even though there was insufficient toxin to produce ill- ness in a mongoose consuming such a fish for the first time. The results of our first sampling of potentially toxic fishes at Enewetak and Bikini did not in- dicate a high level of toxicity. Only the larger individuals of the usual offending species' were poisonous. Most of these species are carnivores, in particular those feeding heavily on fishes (Randall 1958). Therefore, subsequent fishing was concentrated on the larger fishes of these species. Because of this selectivity, both for spe- cies and size, more fishing effort was spent per fish caught; however, this meant less effort expended later in useless testing. Prior to the present study, information on the food habits of ciguatoxic fishes was insufficient for most species. When a trained marine biologist familiar with the Marshallese marine biota was present, an analysis was made of the stomach con- tents of the fishes that were caught. Since cigua- toxin is known to pass through food chains to the larger fishes, where it is concentrated, analyses of the stomach contents of these fishes are needed for an understanding of the feeding inter- relationships. Some previously unpublished stomach-content Table l. — Summary of mongoose feeding tests (liver- viscera, sharks excepted) of fishes collected at Enewetak (0 = nontoxic; 5 = death of test animal). Intensity of reaction Species Nebrius ferrugineus Carcharhinus albimarginatus C. amblyrhynchos C galapagensis C. Iimbatus Galeocerdo cuvier Triaenodon obesus Taeniura melanospilos Lycodontis javanicus Adioryx spinifer Sphyraena barracuda Cephalophalis argus Epinephelus fuscoguttatus E. hoedtii E. maculatus E. microdon £. socialis E. tauvina Plectropomus leopardus P. melanoleucus P. truncatus Variola louti Aprion virescens Lutjanus bohar L. fulvus L. gibbus L. monostigmus Macolor niger Lethrinus amboinensis L kallopterus L. miniatus L. xanthochilus Monotaxis grandoculis Kyphosus cinerascens Caranx ignobilis C. lugubris C. melampygus C. sexfasciatus Gymnosarda unicolor Cheilinus undulatus Coris aygula Epibulus insldiator Hipposcarus harid Scarus gibbus S. rubroviolaceus Acanthurus xanthopterus Balistoides viridescens 2 4 11 1 2 8 1 6 3 2 2 6 8 8 2 4 12 13 8 56 2 21 1 23 24 14 6 2 4 1 3 11 24 2 7 6 4 5 3 18 3 2 1 4 1 22 2 1 3 4 1 1 8 13 1 8 4 1 1 2 1 11 5 5 2 1 1 2 Totals 346 74 53 37 11 204 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS analyses obtained by the author from localities other than the Marshall Islands have been in- cluded in this report. The food habit data are pre- sented in the species accounts following the assay of toxicity. The length measurement most often used for bony fishes was standard length (SL). This was taken from the front of the snout to the end of the hypural plate (hence base of caudal fin). When the method of measuring length is not specified, stan- dard length is intended. Fork length (FL) was used for carangids and scombrids because scutes or keels laterally on the caudal peduncle prevent the accurate external determination of the base of the caudal fin. Total length (TL) was employed for eels, nurse sharks, and some proportional mea- surements. The length usually given for requiem sharks is precaudal length (PCL). Schultz and collaborators (1953-66) was the primary reference for the identification of ciguatoxic fishes and the fishes from their stomachs. Enewetak and Bikini were among the islands from which large collections of fishes were made for this systematic work. When names other than those given by Schultz and collaborators are used, an explantion is given in the species accounts. The species of ciguatoxic fishes which were studied are discussed in approximate phylogenetic sequence below. The results of the mongoose feed- ing tests are summarized in Table 1 for Enewetak and Table 2 for Bikini. RESULTS Orectolobidae (Nurse Sharks) Nebrius ferrugineus (Lesson) (Figure 1): Like Table 2. — Summary of mongoose feeding tests (liver-viscera, sharks excepted) of fishes collected at Bikini (0 = nontoxic; 5 = death of test animal). Intensify of reaction Species 1 2 3 4 5 Carcharhinus amblyrhynchos 1 Galeocerdo cuvier 1 Sphyraena barracuda 1 1 1 S. forsteri 2 Crenimugil crenilabis 3 Epinephelus hoedtii 1 E maculatus 2 E. microdon 2 1 4 1 1 E. tauvina 1 1 Plectropomus leopardus 1 Variola louti 2 Aprion virescens 7 1 Lutjanus bohar 112 15 8 6 2 L gibbus 32 3 L monostigmus 4 1 Lethrinus amboinensis 9 L kallopterus 2 L miniatus 12 L xanthochilus 2 Caranx igriobilis 2 C. melampygus 6 Gymnosarda unicolor 3 Pseudobalistes flavimarginatus 2 Totals 208 22 16 7 2 1 other orectolobids, this shark has a prominent nasal barbel, relatively small mouth, the fourth and fifth gill openings over the pectoral base, and the two dorsal fins set posteriorly on the body. The teeth, which are small and in numerous rows (the first three or four functional), have a large central cusp with four to six smaller cusps on each side (the teeth of the related genus Ginglymostoma have an even larger central cusp and only two small cusps on each side). The two dorsal fins are of nearly equal size, the first originating slightly an- terior to the origin of the pelvic fins and the second distinctly anterior to the origin of the anal fin; the caudal fin is about 2>Q% TL. Nebrius concolor Riippell appears to be a junior synonym. Bass et al. (1975b) employed this name, Figure l.— Nebrius ferrugineus. 1,080 mm PCL, 1,496 mmTL, 18.1 kg, Enewetak, Marshall Islands. 205 FISHERY BULLETIN: VOL. 78, NO. 2 but admitted that Lesson's ferrugineus might have priority. This shark is a shallow-water species, usually seen at rest on the bottom during daylight hours. It is not common in the Marshall Islands. Two specimens were obtained from Enewetak, 1,400 and 1,487 mm TL, weighing 11.7 and 14.1 kg. The flesh of neither was toxic. Fourmanoir (1961) stated that the principal food of this shark consists of octopuses and xanthid crabs. Gohar and Mazhar (1964) reported cephalopods, fishes, and parts of corals (Stylophora) from the stomachs of Red Sea speci- mens (the corals were probably accidentally in- gested). Hiatt and Strasburg (1960) found a rab- bitfish, Siganus sp., in the stomach of one of three specimens collected at Enewetak. The stomach of the smaller of the two specimens taken during the present study contained a sur- geonfish, Acanthurus glaucopareius, 95 mm SL. Three other Enewetak specimens and one from the Tuamotu Archipelago (to 1,615 mm TL, 20.9 kg) had empty stomachs. Carcharhinidae (Requiem Sharks) Carcharhinus albimarginatus (Riippell) (Figure 2): The silvertip shark is one of three carcharhinid sharks with white on the tips of its fins; the others are the oceanic whitetip shark, C. maou (Lesson) (C. longimanus a junior synonym), and the whitetip reef shark, Triaenodon obesus (Riippell). The name silvertip has been adopted by Kato et al. (1967) and others to avoid confusion with the other two species with white-tipped fins. The white on the silvertip's fins is not confined to the distal ends but continues along the posterior margins. The apex of the first dorsal fin is somewhat pointed (broadly rounded on C. maou); the origin of this fin is over the inner edge of the pectoral fin. The pec- torals are about 18% TL (about 28% on C. maou). A median interdorsal ridge is present. There are usually 26 upper and 24 lower teeth. The pre- caudal vertebrae vary from 115 to 125. Carcharhinus albimarginatus has not been re- ported as poisonous [Halstead (1967, pi. VI, fig. 3) illustrated it but misidentified the figure as Triaenodon obesus], but it would seem to have the potential for causing ciguatera because it preys in part on reef fishes. In the Marshall Islands it was usually seen on exposed outer reefs in water >30 m, though one individual was observed in the Enewetak lagoon in water only 2 m deep. Bass et al. (1973) summarized the depth distribution, not- ing records to 400 m. This species has attacked man. The flesh of four silvertips, 933-1,650 mm PCL (15.0-73.5 kg), from Enewetak was nontoxic. Fourmanoir (1961) reported the following wide variety of fishes from the stomachs of silvertips from Madagascar: Promethichthys prometheus, Pristipomoides typus, Seriola songoro, Coris gaimard, Caesio coerulaureus , Acanthocybium solandri , Euthynnus pelamis , and Neothunnus al- FIGURE 2.— Carcharhinus albimarginatus, 1,650 mm PCL, 2,154 mm TL, 73.5 kg, Enewetak, Marshall Islands. 206 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS bacora (= Thunnus albacares). Bass et al. (1973) examined the stomach contents of 10 specimens of C. albimarginatus from the western Indian Ocean. They found teleost fishes (exocoetid, myctophid, several soleids) in seven sharks; a spotted eagle ray, Aetobatis narinari, in one; and an octopus in one. The stomach of one of the Enewetak silvertips (1,240 mm PCL, 1,650 mm TL) contained a gray reef shark, Carcharhinus amblyrhynchos , 483 mm PCL and 616 mm TL, as well as the dental plates and pharyngeal mills of three parrotfishes (Scarus). The stomachs of the other three sharks were empty. Carcharhinus amblyrhynchos (Bleeker) Figure 3): This shark, now popularly known as the gray reef shark, was referred to by Schultz in Schultz and collaborators (1953) as Carcharhinus menisorrah (Miiller and Henle). Bass et al. (1973) and Garrick (in press) are followed in the use of the name C. amblyrhynchos herein. The gray reef shark lacks dark pigment distally on the first dorsal fin, but the tips of the other fins are broadly blackish, and there is a broad black margin posteriorly on the caudal fin. The dark markings on the fins are more evident on live than on dead specimens. The origin of the first dorsal fin is over the pectoral axil or anterior part of the inner edge of the pectoral fin. A short interdorsal ridge is present or absent. There are 26-28 upper teeth and 24-26 lower teeth; the precaudal verte- brae vary from 110 to 117 (two Enewetak speci- mens had 117). This shark is abundant in the Marshall Islands. It occurs in many habitats from lagoons to ocean reefs, but it is most commonly encountered in deep channels and outer reef areas. It does not pene- trate the shallows as readily as C. melanopterus. The flesh of 11 specimens from Enewetak, 1,017-1,190 mm PCL (17.2-26.3 kg), and 1 from Bikini (3.6 kg, length not taken) was tested. All gave a zero reaction for ciguatoxin. The viscera of one of these, 1,158 mm PCL, from Enewetak pro- duced a reaction of 2, however. The stomachs of 74 individuals, 520-1,230 mm PCL (2.7-32.4 kg), from Enewetak, Fanning and Palmyra in the Line Islands, Marcus Island, Johnston Island, Palau Islands, and Ducie and Henderson in the Pitcairn Group, were examined for food. Forty-nine stomachs were empty or con- tained only bait. Three had eaten cephalopods (two octopus, one squid), and the rest contained the remains of fishes (in some cases only the lens of an eye or a few remnants of spines or bones). The fishes that could be identified to family or genus were the following: muraenid, belonid, exocoetid, Fistularia, Decapterus , Trachinotus, Acanthurus, and another acanthurid (either Acanthurus or Ctenochaetus). In spite of its relatively small size, the gray reef shark constituted a hazard to the personnel of the survey program, particularly when divers were spearfishing or collecting with rotenone. Several Figure 3.— Carcharhinus amblyrhynchos, 1,288 mm PCL, 1,640 mm TL, Tahiti, Society Islands. 207 FISHERY BULLETIN; VOL. 78, NO. 2 of these sharks were killed by powerheads when their aggressive behavior and proximity war- ranted. On 12 July 1975 the companion diver of the author, Russell E. Miller, sustained 7 gashes in his head requiring 25 stitches as the result of an attack by a C. amhlyrhynchos of about 1,500 mm TL. The shark first exhibited threat posturing (see Johnson and Nelson 1973) at the back of the au- thor. Miller sounded a warning by rapping on his scuba tank with his powerhead handle. The shark immediately turned and swam toward him, re- peating the exaggerated sinuous swimming of its threat behavior as it approached. Miller struck the shark with his powerhead but the shell misfired. The shark came on to slash his head (and cut the rubber strap of his face mask) with its upper jaw. On another occasion Gordon W. Tribble had the end of his speargun seized by a gray reef shark and vigorously shaken. Richard C. Wass ( 1971) has made a comparative study of the biology of the gray reef shark and the sandbar shark in the Hawaiian Islands. Carcharhinus galapagensis (Snodgrass and Hel- ler) (Figure 4): This shark is circumtropical in distribution, but as noted by Garrick (1967), it shows a preference for the sea around oceanic islands. The Galapagos shark has no distinctive mark- ings; it is dark gray dorsally, pale ventrally. The origin of the first dorsal fin is anterior to the inner free corner of the pectoral fin. The second dorsal fin is relatively large for a Carcharhinus, its height 2.4-2.8% TL. A distinctive median interdorsal ridge is present on the back. There are 26-30 upper teeth (the anterior upper teeth broadly triangular) and 26-29 lower teeth. The precaudal vertebrae range from 103 to 109. A single specimen, 1,831 mm PCL, 2,426 mm TL, 41.2 kg, was collected at Enewetak. Its flesh was nontoxic. Tester'^ examined the stomach contents of 41 Galapagos sharks caught from the Hawaiian Is- lands; 51% were empty. Sixty percent of the sharks had eaten bony fishes, 35% cephalopods, 20% sharks and rays, and 10% crustaceans. He commented that the larger individuals (maximum length estimated as 10 ft or 3,048 mm) fed mostly on larger fishes which were torn into chunks. He regarded it as a dangerous species; Randall ( 1963) documented a fatal attack. Bass et al. ( 1973) found food in 18 of 22 individu- als of this species; 12 of the stomachs contained teleost fishes (serranid, Platycephalus , and a flatfish) and 10 had squids or octopuses (plus the shell of a bivalve mollusk). The Enewetak specimen was empty as were three others 1,460-1,690 mm PCL from the Pit- cairn Group. One of 1,580 mm PCL from Rapa contained the head of an unidentified eel. Carcharhinus limbatus (Valenciennes) (Figure 5): The shark occurs in the Atlantic as well as the Indo-West Pacific. In the Atlantic it bears the common name of blacktip shark, a name which is poor for the species in the Pacific for two reasons. ■'Tester, A. L. 1969. Final Report, Cooperative Shark Re- search and Control Program, University of Hawaii. Mimeogr. Rep., 47 + 36 append, p. Figure a.— Carcharhinus galapagensis, 1,831 mm PCL, 2,426 mm TL, 87 kg, Enewetak, Marshall Islands. 208 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Figure 5. — Carcharhinus limbatus, 1,415 mm PCL, 1,910 mm TL, 38.2 kg, Enewetak, Marshall Islands. The tips of the fins of Pacific individuals, particu- larly adults, are only slightly tipped or edged in black. Also the common Indo-West Pacific C. melanopterus has very pronounced black tips on its fins (see Randall and Helfman 1973, fig. 1, 2). To avoid confusion in common names, C. melanop- terus has been referred to by many recent authors as the blacktip reef shark (though Bass et al. 1973, call it the blackfin reef shark). Carcharhinus limbatus is distinctive in lacking a median ridge on the back between the dorsal fins, having a relatively long snout, and the cusp of its teeth notably narrow and erect. It has 29-32 upper, 28-32 lower teeth, and 88-102 precaudal vertebrae. The color is gray to bronze on the back, white below, with a long band of the dark dorsal color extending posteriorly from the last gill open- ing into the pale ventral color as far as the pelvic fins. Two individuals of C. limbatus were caught at Enewetak during the ciguatera survey; these con- stitute the first records of the species from the Marshall Islands. The head of the illustrated specimen (which was 1,415 mm PCL, 1,910 mm TL, and weighed 38 kg) has been preserved in the Bernice P. Bishop Museum under catalog number 18074. Only the second specimen, which was about 1,700 mm PCL (original data sheet with mea- surements was lost), was tested for toxicity. The viscera gave a reaction of 2 when fed to a mongoose. Bassetal. (1973) reported on 55 of 101 sharks of this species with food in their stomachs. Fifty-one of the sharks had eaten teleost fishes, including: Scomberomorus commerson, S. leopardus, Pomadasys sp., Sarpa salpa, Johnius hololepidosus , Leiognathus equula, Flops saurus, Tilapia mossambica, and a soleid. Six contained elasmobranchs, including a small C. obscurus and aRhinobatus annulatus. Two had eaten Sepia sp., and one a spiny lobster, Panulirus homarus. The two Enewetak specimens had empty stomachs. Galeocerdo cuvier (Peron and Lesueur) (Figure 6): The circumtropical tiger shark is readily iden- tified by its broad bluntly rounded snout, distinc- tive teeth (heavily serrate, convex on the medial margin, and deeply notched on the lateral), low longitudinal keel on the side of the caudal pedun- cle, and dark bars (though these tend to fade with age). The flesh of two tiger sharks from Enewetak, 1,770 and 2,410 PCL (72 and 174 kg), and one from Bikini, 1,498 mm PCL, was tested. None of these sharks were toxic. The Bikini specimen was caught at 4:30 a.m. in only 1.7 m of water. Bigelow and Schroeder ( 1948) summarized the literature on food habits, danger to man, etc., of this shark. Other authors such as Clark and von Schmidt ( 1965), Bass et al. (1975a), and Tester (see footnote 7), have added to the list of the great variety of marine animals (mainly fishes) that this species will take as food, as well as sundry items of garbage and refuse discarded into the sea by man. The stomach of the largest of the Marshall Is- lands specimens contained the scutes of a green turtle, Chelonia mydas, estimated to be 500 mm carapace length and the bait (a gray reef shark). 209 FISHERY BULLETIN: VOL. 78, NO, 2 Figure 6. — Galeocerdo cuvier, 2,410 mm PCL, 3,055 mm TL, 175 kg, Enewetak, Marshall Islands. The stomachs of the other two sharks were empty. Three other tiger sharks from Enewetak had food in their stomachs. One of 3,150 mm TL contained shark vertebrae. The second of 3,581 mm TL had the scutes of a green turtle and bird feathers. The third, 3,048 mm TL, was filled with pieces of a porpoise and the digested remains of shark fins. A tiger shark of 3,327 mm TL from Ua Huka, Marquesas Islands, was empty, as was one of 2,895 mm TL from Oahu. Another from Oahu of 3,048 mm had an extremly distended stomach filled with heads of skipjack tuna (neatly cut by a knife, hence probably discarded from a fishing boat), plastic bags of garbage and aluminum foil, a cat, and two small reef fishes (one a balistid). It also contained the bait (the head of a calf). A 3,100 mm specimem weighing 174.6 kg taken by a set line at night at Rapa had eaten parts of a tiger shark larger than itself (probably from an individual caught on another hook), as well as a seabird. Triaenodon obesus (Riippell) (Figure 7): The whitetip reef shark, once classified by most ichthyologists in the family Triakidae, is now re- corded as a carcharhinid (Compagno 1973). In spite of its scientific name, it is rather slender compared with most species of the family. Apart from its slim form and white-tipped first dorsal fin and upper caudal lobe, T. obesus is distinctive in its very blunt snout and teeth which bear a small cusp on each side of the main central one. It is widespread throughout the tropical and subtropi- cal Indo-West Pacific region and ranges to the eastern Pacific as well. Banner and Helfrich (1964) and Brock et al. (1965) have reported this species as poisonous from Johnston Island. Figure T .—Triaenodon obesus, 1,218 mm PCL, 1,520 mm TL, 23.5 kg, Tahiti, Society Islands. 210 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS The toxicity of eight whitetip reef sharks, 1,003-1,183 mm PCL (8.4-22.4 kg), from Enewetak was tested. The flesh of all was negative for ciguatoxin by feeding to mongooses, but the viscera of two gave positive reactions of 1. Randall (1977) studied the biology of this species. He opened the stomachs of 56 specimens (24 of which were from Enewetak): 33 were empty, 6 had eaten octopuses (2 of these also contained fishes), and the rest had the remains of reef fishes, especially scarids and acanthurids. Dasyatidae (Sting Rays) Taeniura melanospilos Bleeker (Figure 8): The specimens collected in the Marshall Islands were initially called Taeniura brocki Schultz. However, it now seems more likely that they should be iden- tified as T. melanospilos Bleeker. Schultz in Schultz and collaborators (1953) differentiated T. brocki by its having the venomous spine inserted ". . . at about half length of tail . . ." in contrast to a little behind the first third on T. melanospilos, in having the snout contained five times in the greatest width of the disc (given as six by Bleeker for T. melanospilos), and in having ". . .very numer- ous irregularly shaped small brownish to blackish spots and blotches speckling dorsal surface of disk ...," as opposed to "... numerous rounded black spots . . ." for T. melanospilos. After noting the measurements of the position of the spine and the length of the tail of his only specimens of T. brocki, Schultz wrote, ". . . end of the tail may have been bitten off " On the specimen illustrated herein, the spine is inserted at a point 41% the length of the tail from the base. From Schultz' measure- ments, the snout of T. brocki is contained 5.14 times in the width of the disc. The snout of the specimen illustrated herein is contained about 5.3 times in the disc width. Without knowledge of variation of this character and perhaps propor- tional differences with growth, the differentiation of species on this magnitude of snout length is questionable. Furthermore, Bleeker's (1853) de- FIGURE 8.— Taeniura melanospilos, 1,255 mm disc length, 2,008 mm TL, 69 kg, Enewetak, Marshall Islands. 211 FISHERY BULLETIN: VOL. 78, NO. 2 scription of the dark spots of T. melanospilos was not simply round as Schultz stated, but round and oblong and variably small, medium, and large. One stingray of this species from Enewetak with a disc length of 870 mm (disc width 950 mm; tail length 950 mm), weighing 19.05 kg, was tested. It was not poisonous. The stomach of this ray was empty. Another of 1,255 mm disc length weighing 68.9 kg collected by the author at Enewetak in 1968 had eaten two labrid fishes (Xyrichtys) and a parrotfish, Scarus sp. It was a female with seven embryos. Muraenidae (Morays) Lycodontis javanicus (Bleeker) (Figure 9): This moray is brown with large dark blotches and numerous small dark spots; the gill opening is in a large black spot; there are no pale margins on the fins. It is probably the species of eel reported by Khlentzos (1950) which poisoned 57 Filipino laborers at Saipan, Mariana Islands. In spite of prompt gastric lavage, 14 of these men became comatose and 2 died. The severity of illness from the consumption of moray eels led Halstead and Lively (1954) to re- gard this as a distinct category of fish toxemia which they termed "Gymnothorax poisoning." However, it appears to be principally an acute form of ciguatera, though there is a possibility of involvement of one or more other toxins. Lycodontis javanicus is not common in the Mar- shall Islands, but it is abundant (for a large carni- vore) at Johnston Island; in recent years it has served as the primary source of ciguatoxin for biochemical and pharmacological study at the University of Hawaii by a team of scientists headed by A. H. Banner. Nine specimens from Enewetak measuring 1,086-1,540 mm TL and weighing 3.6-13.0 kg were tested. All were toxic, two at the 2 level, one at 3, three at 4, and three at 5 from the feeding of liver and viscera to mongooses. The flesh of two of these eels with a mongoose reaction of 4 was tested; one was a 1 and the other a 2. One of the eels with a mongoose test of 5 for liver-viscera gave a reaction of 3 with flesh. Brock ( 1972) studied some aspects of the biology oiL. javanicus, including an analysis of the toxic- ity at Johnston Island. Of 1,074 specimens, only 158 (14.7%) contained food; 88.8% of the stomach- content material consisted of fishes (representing 17 different families, the Scaridae predominat- ing). Among the more interesting prey species was Figure Q.— Lycodontis javanicus, 732 mm TL, Enewetak, Marshall Islands. 212 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS a whitetip reef shark 465 mm PCL, taken from the stomach of a 1,422 mm moray. Octopus and spiny lobster were also eaten. This is the largest species of moray in the Indo- Pacific region. Schultz ( 1949) attributed an attack on the late Vernon E. Brock at Johnston Island by a moray of about 10 ft (3,048 mm) in length to be Enchelynassa canina. Following a later interview with Brock, Randall (1969) reported that the eel was actually L. javanicus and the length 7-8 ft (2,134-2,438 mm). Stephens (1963) noted that the largest moray measured at Johnston during his stay at the island to be 7 ft 10 in (2,388 mm). The author tended to disbelieve occasional reports by divers of individual L. javanicus of 10-12 ft (3,048- 3,658 mm) until he observed one of an estimated 3,000 mm long off Mafia Island, Tanzania, which was flushed from a cave with rotenone (the eel recovered from the affect of the rotenone and re- turned to its cave). The stomach contents of 1 1 specimens 417-1 ,905 mm TL (the largest weighed 24.5 kg) were examined during the present study. Six of these eels were from Enewetak, the rest from Oeno, Pit- cairn, Johnston, and Truk. Four had empty stomachs. The smallest contained a crab chela. The others had eaten fishes (two contained Scarus sp., one Diodon sp., and another Thalassoma pur- pureum). The stomach of a 1,540 mm, 13 kg speci- men was distended with Kyphosus cinerascens, Acanthurus nigricauda (identified as A. nigricans by Schultz and Woods in Schultz and collaborators 1953, and as A. gahhm by Randall 1956), and A. nigroris, all of which totalled 1.5 kg. These fishes must be discounted as normal prey, however, as they were undoubtedly eaten as a result of a dynamite station at the Enewetak garbage pier. The eel was collected with a powerhead blast im- mediately after the dynamite explosion when it was discovered within the area in which many other fishes had just been killed. Hoiocentridae (Squirrelfishes) Adioryx spinifer (Forsskal) (Figure 10): The largest of the squirrelfishes, this species exceeds 300 mm SL. It has a deep body, the depth about 2.5-2.7 in SL, projecting lower jaw, 40-44 lateral line scales, and a well-developed venomous spine at the corner of the preopercle. The color is red and silvery with a deep red spot behind the eye and another on the pectoral axil; the fins are yellow except the spinous dorsal which is deep red. De- scribed from the Red Sea, the species has since Figure lO. — Adioryx spinifer, 260 mm SL, Enewetak, Marshall Islands. 213 FISHERY BULLETIN: VOL. 78, NO. 2 been recorded from throughout the tropical Indo-West Pacific. It is often found in caves. Randall (1958) reported this species as occa- sionally toxic in Tahiti. In his review of ciguatoxic fishes, Halstead (1967) cited this and three other references. Six specimens, 232-280 mm SL (0.35-0.66 kg), were collected for assay of toxicity from Enewetak. All were nontoxic. Randall (1958) found fishes in the stomachs of two adults from Tahiti. Hiatt and Strasburg (1960) examined the stomachs of nine from Enewetak, one of which was empty. Crabs (espe- cially xanthids) dominated the stomach contents of the other specimens: 12% contained stoma topods and 12% fishes. The stomachs of five specimens from Hawaii reported by Hobson ( 1974) contained crustaceans, mainly caridean shrimps, and xanthid crabs. For this food-habit study a total of 31 specimens ranging from 182 to 280 mm SL were obtained, principally from Enewetak, but a few from the Red Sea, Society Islands, Hawaiian Islands, and American Samoa. Because this species is noctur- nal, like other holocentrids in general, most specimens were collected in early morning hours. Twenty-eight of these fish had food in their stomachs; 82% by volume consisted of crabs, mostly xanthids, 5% fishes (including Lycodontis rueppelliae and a prejuvenile acanthurid), and the rest shrimps, a hermit crab, unidentified crusta- ceans (mostly larval), larval moUusks, and an un- identified worm. Sphyraenidae (Barracudas) Sphyraena harracudaa (Walbaum) (Figure 11): The great barracuda is distinctive in having a few blackish blotches on the side, especially posterior- ly and ventrally, and the lowest lateral line scale count of the genus (76-85). It is the worst offender for causing ciguatera in the West Indies, due not only to the high level of toxicity of occasional indi- viduals but also to its relative abundance there. The species is far less common in the Indo-West Pacific. Seven specimens 563-1,182 mm SL (1.5-13.6 kg) from Enewetak were tested, and three from Bi- kini, 640-1,143 mm SL, the largest 15.0 kg. Three from Enewetak were nontoxic, one was toxic at the level of 1; two gave a reaction of 2; and one (1,050 mm, 12.7 kg) was a 3. The three from Bikini were tested at 0, 1, and 2. de Sylva (1963) made a studv of thesystematics and life history of the great barracuda, principally from material from the western Atlantic. He re- viewed previous papers which presented limited data on the food habits of this species. Among them was the report by Ommanney in Wheeler and Ommanney (1953) on the fishes found in the stomachs of 5 of 12 specimens of S. commersoni (now known to be a junior synonym of S. bar- racuda) from the Seychelles. One of the five con- tained an unidentified eel and another Lethrinus ramak. de Sylva mistakenly reported Omman- ney's five barracuda as all having eaten L. ramak. de Sylva opened the stomachs of 901 great bar- racuda, including juveniles, from various localities in the tropical western Atlantic. Of these, 529 (58.7%) contained food. Fishes were found in 82.2% of the stomachs, plant material in 2.8% (probably accidentally ingested with prey), invertebrates (notably squids and shrimps) in 2.6%, and unidentifiable material in 12.2%. The Hemiramphidae was the most common family of fishes found in the stomachs of 446 adult bar- racuda from Florida, whereas tetraodontid fishes predominated in the stomach contents of 132 adults from Bimini, Bahamas. The stomachs of 13 specimens, 560-1,182 mm Figure ll. — Sphyraena barracuda, 524 mm SL, Palmyra, Line Islands. 214 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS SL, from the Marshall Islands and Line Islands were examined during the present study. Nine of these fishes had empty stomachs. The rest con- tained the remains of fishes. Sphyraena forsteri Cuvier (Figure 12): This bar- racuda is readily distinguished from all others by its large eye, black blotch in the axil of the pectoral fins, and spiniferous plates on its first gill arch instead of gill rakers or no trace of rakers at all. de Sylva (1973:18) wrote that this species of barracuda has been implicated in poisoning but added that the examples appear to be misiden- tifications of S. barracuda. Hiyama (1943, pi. 3, fig. 9), however, did not confuse his specimens with ■S. barracuda. He reported S. forsteri from the Marshall Islands as slightly poisonous from feed- ing flesh to cats and mice. Only two individuals of this species were caught during the survey, both from Bikini and both the same size (610 mm SL, 1.8 kg). Each produced a toxic reaction of 2. The stomach of one of these fish contained fish remains; the other was empty. Mugilidae (Mullets) Crenimugil crenilabis (Forsskal) (Figure 13): This large mullet has a deeply emarginate caudal fin, a black spot at upper pectoral base, and 37-39 rows of scales between the gill opening and the caudal base. Widespread in the tropical Indo-West Pacific region, it is usually seen in small aggrega- tions in the shallows of lagoons and on outer reef flats. It appears to feed on fine algae and detritus from the substratum. After feeding on a sandy bottom it has been observed to expel sand from its gill openings. The spawning by a large school at the surface at night in the Enewetak lagoon was described by Helfrich and Allen (1975). Cre- nimugil crenilabis has been reported as poisonous by Randall (1958) from the Society Islands. It is probably the species of mullet that Ross (1947) Figure 12.— Sphyraena forsteri, 540 mm SL, 1.2 kg, Enewetak, Marshall Islands. Figure 13. — Crenimugil crenilabis, 337 mm SL, Enewetak, Marshall Islands. 215 Fanning, Line Islands (Randall 1 in ee specimens, 248-406 mm obtained for testing from Bikini. found toxic at 1958). Three specimens, 248-406 mm (0.4-0.7 kg), were n "■'■■ None were toxic. Serranidae (Groupers) Cephalopholis argus Bloch and Schneider (Fig- ure 14): This common brown blue-spotted grouper has 9 dorsal spines ( in contrast to 1 1 for the group- ers of the genus Epinephelus). It does not reach large size, but is has occasionally been implicated in ciguatera. Although it is most abundant in outer reef areas, it also occurs on lagoon reefs. A total of nine specimens from Enewetak were tested; these ranged from 278 to 390 SL (0.45-1.6 kg). Two were nontoxic, three gave a reaction of 1, three were recorded as 2, and the largest was a 4. Randall (1955a) found 8 of 10 individuals of this grouper from the Gilbert Islands with empty stomachs; 1 had eaten a fish (probably from rotenone), and 1 a penaeid shrimp. Randall and Brock (1960) obtained 280 specimens for food- habit study, of which 182 were empty; 77.5% con- tained fishes and the rest crustaceans. Hiatt and Strasburg (1960) reported on food in five of eight stomachs from the Marshall Islands as crusta- ceans, fishes, and polychaetes. Helfrich et al. (1968) examined the stomachs of 51 from Palmyra; they found fishes in 89% of the stomachs and crus- taceans. Harmelin-Vivien and Bouchon (1976) caught 43 C. argus for stomach-content analyses in Madagascar. They found fishes 95.7% by FISHERY BULLETIN: VOL. 78, NO. 2 weight, shrimps 3.9%, and stomatopods 0.4% in the stomachs. For the present study the stomachs of 39 speci- mens, 145-392 mm SL, from Enewetak, Society Islands, Samoa Islands, Palmyra, Marcus Island, and Pitcairn, were examined. Twenty-six were empty, one had eaten a stomatopod, and the rest contained fishes (two of these were the acronurus stage of Acanthuridae, one a labrid, one an anten- nariid, and one Apogon kallopterus). Epinephelus fuscoguttatus (Forsskal) (Figure 15): The name E. fuscoguttatus was used by Schultz in Schultz and collaborators ( 1953) for the more common related species properly called E. microdon (Bleeker) (systematic clarification by Randall 1964). The two are similar in morphology and color. Epinephelus fuscoguttatus has higher pectoral ray counts (18 or 19, compared with 16 or 17 for E. microdon) and more lower limb gill rak- ers ( 18-21, including rudiments, compared with 16 or 17 for.E. microdon). This grouper is a large species; it is not common. Furthermore, it is the most wary of Marshall Is- lands groupers. Seven specimens (335-780 mm SL, 3.1-15.4 kg) were taken at Enewetak for testing, and none at Bikini. Four of the seven were toxic at the 2 level, and one (710 mm SL) was a 3. Harmelin-Vivien and Bouchon (1976) caught four individuals of this species for food-habit study in Madagascar. The stomachs contained 94.2% fishes by weight and 5.8% brachyuran crabs. The stomachs of seven specimens from Figure U.— Cephalopholis argus, 232 mm SL, Tahiti, Society Islands. 216 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS -Mfc(^ "^ Figure 15. — Epinephelus fuscoguttatus, 551 mm SL, 5.9 kg, Gulf of Aqaba, Red Sea. Enewetak were examined. Four were empty, one had an octopus, one had fish remains, and the last contained unidentified tissue which appeared to be cephalopod in origin. A specimen 408 mm SL from the Red Sea had crab remains in its stomach. Epinephelus hoedtii (Bleeker) (Figure 16): Schultz in Schultz and collaborators (1953) de- scribed this fish as new from the Marshall Islands, naming it E. kohleri. It is relatively deep bodied and has a slightly emarginate to truncate caudal fin. He differentiated it from "... all of the 'var- ieties' of flavocaeruleus described by Boulenger in having the body spotted with dark blotches in ad- dition to tiny dark specks." Although more study is needed of the complex of forms which Boulenger Figure 16. — Epinephelus hoedtii, 319 mm SL, Enewetak, Marshall Islands. 217 FISHERY BULLETIN: VOL. 78, NO. 2 (1895) regarded as varieties of flauocaeruleus, I believe that the variety called £^. hoedtii (Bleeker) is a valid species and that Schultz' E. kohleri is a junior synonym of it. Adult specimens, such as the type ofE. kohleri, have the dark blotches, whereas smaller individuals, such as Bleeker's specimens, lack them. Hiyama (1943:81, pi. 18, fig. 49) iden- tified it as Serranus flavocaeruleus (Lacepede). This grouper was found around isolated coral heads in the lagoon of Enewetak. Eleven speci- mens, 348-429 mm SL, 1.36-2.72 kg, were tested. Six were nontoxic, three gave reactions of 1, one was a 2, and one (400 mm SL) a 4. A single speci- men (2 kg) from Bikini was nontoxic. Hiatt and Strasburg (1960) found fish fragments in the stomach of one of two specimens from the Marshall Islands. The stomachs of the 11 Enewetak specimens were opened. Five were empty, one (429 mm SL) contained a 520 mm snake eel, Leiuranus semicinctus; two had eaten calappid crabs; and the remaining three contained the digested remains of fishes. Epinephelus maculatus (Bloch) (Figure 17): This dark-spotted grouper was identified as E. medurensis (Giinther) by Schultz in Schultz and collaborators (1953). It has also been called E. fario (Thunberg) by some authors. The oldest valid name, however, is E. maculatus (Bloch). Though the author ascertained that the type-specimen is no longer extant, Bloch's description and illustra- tion match that of the juvenile of this species, particularly with reference to the large pale mark- ings. Adults are distinctive in the rather elevated third and fourth dorsal spines; also there are two large dark areas on the dorsal fin and adjacent back which are separated by a pale area (both dark and light areas still have the profusion of small dark spots). Like E. hoedtii, this species is found mainly around coral knolls in sandy stretches of atoll la- goons. Eleven specimens from Enewetak, 270-334 mm SL, 0.45-0.9 kg, were tested. Eight were non- toxic and three gave reactions of 1. Two from Bi- kini, 343 and 356 mm SL, were nontoxic. The stomachs of 13 specimens from Enewetak and 2 from Bikini, 270-380 mm SL, were examined. One of 334 mm contained a portunid crab and unidentified fish remains; another of 345 mm had eaten a calappid crab (15% by volume), two microdesmid fish 78 and 86 mm SL (identified as Gunnellichthys monostigma by C. E. Dawson), and a digested fish; a third (308 mm SL) also con- tained G. monostigma; a fourth (288 mm SL) an octopus; and two others fish remains. The remain- ing nine stomachs were empty. Epinephelus microdon (Bleeker) (Figure 18): This is a common species in the Marshall Islands for a grouper of moderate size. It is found on both Figure 17. — Epinephelus maculatus, 280 mm SL, Enewetak, Marshall Islands. 218 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Figure 18.— Epinephelus microdon, 282 mm SL, Enewetak, Marshall Islands. ocean and lagoon reefs. As mentioned in the ac- count of E. fuscoguttatus above, it has been con- fused with this species by Schultz and other authors. Thirty-nine specimens from Enewetak were tested for ciguatera toxin. These ranged from 310 to 470 mm SL and weighed 1.4-3.1 kg. Eight were nontoxic, 3 were toxic at the 1 level, 8 were 2, 13 were 3, 5 were 4, and 2 were 5 (caused death of test animals). Nine specimens from Bikini, 279-508 mm SL, 0.9-4.1 kg, were tested. Two (279 and 342 mm) were nontoxic, one was poisonous at the 1 level, four were 2, one was 3, and one was 5 (460 mm SL). Randall and Brock (1960) reported on the food habits of this species (as E. fuscoguttatus) from 33 specimens taken in the Society Islands and Tuamotu Archipelago. Of 10 with food in their stomachs, 5 had eaten crustaceans (mainly crabs) and 5 of them fishes. The eight specimens recorded by Hiatt and Strasburg (1960) as E. fuscoguttatus were probably E. microdon. Three fish had empty stomachs and the remaining five contained fishes and crustaceans. Helfrich et al. ( 1968) examined the stomachs of 150 specimens from the Line Islands of which 81 contained food, mainly fishes and crustaceans. The latter accounted for 64% of the total by volume (portunid crabs and scyllarid lobsters were the most frequently recorded). A few of the groupers had eaten gastropods and cephalopods. For the present food-habit study 44 specimens (210-610 mm SL) were examined, of which 28 were from Enewetak. The remaining 15 were from Palmyra, Tutuila, and Rapa (where the largest specimen was taken). Thirty of the 44 groupers had empty stomachs. Eight contained crabs (mainly porcellanids and portunids; one had eaten the xanthid Carpilius conuexus), three contained fishes (one identified as Scarus), two had eaten octopus, and one a spiny lobster, Panulirus. Epinephelus socialis (Giinther) (Figure 19): This grouper has numerous small dark brown spots which tend to coalesce to form irregular longitudi- nal bands, especially posteriorly. The caudal fin and soft portions of the dorsal and anal fins have narrow pale margins and broad blackish submar- ginal zones. It is found mainly on the outer reef fiat of the atoll environment, sometimes in surpris- ingly shallow and often turbulent water. Al- though fishes living entirely in this habitat would not be expected to be ciguatoxic, Halstead and Schall (1958) reported one specimen of this species as weakly toxic from Maiden Island. Two specimens, 354 and 360 mm SL, 1.1 and 1.6 kg, from Enewetak were tested. Both proved to be nontoxic. The stomach contents of seven specimens from Enewetak, 235-360 mm SL, and one from Ducie Atoll, Pitcaim Group (420 mm SL, 2.3 kg) were examined. Three had eaten crabs (grapsids in two, 219 FISHERY BULLETIN: VOL. 78, NO. 2 Figure 19. — Epinephelus socialis, 222 mm SL, Po- ape, Caroline Islands. including Percnon, and the xanthid crab Eriphia sebana was found in the third); one of 330 mm contained an octopus (60% by volume) and a pre- juvenile acanthurid fish; one of 354 mm contained an acanthurid 165 mm SL. The remaining stomachs were empty. Epinephelus tauvina (Forsskal) (Figure 20): The name E. tauvina has often been applied to a huge grouper for whicli the name E. lanceolatus seems correct. Though the true E. tauvina can attain moderately large size (to perhaps 800 mm SL or more), it is not a giant species. Schultz in Schultz and collaborators (1953) described this fish as E. elongatus from the Marshall Islands, Mariana Is- lands, Phoenix Islands, and Samoa Islands, and Smith and Smith (1963) named it E. salonotus from the Seychelles. Katayama (1960) and Ran- dall (1964) showed that E. to«i;ma, described from the Red Sea, is the senior synonym. This species may be confused with other dark-spotted groupers such as E. merra Bloch and E. hexagonus (Bloch and Schneider), particularly when it is small. It is differentiated from them in having 15 instead of Figure 20. — Epinephelus tauvina, 252 mm SL, Enewetak, Marshall Islands. 220 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS 16 soft rays in the dorsal fin, a total of 27-30 gill rakers, instead of 20-27, and a more elongate body. Epinephelus tauvina is not very common in the Marshall Islands. It may be found in both lagoon and outer reef environments. Six specimens from Enewetak, 324-434 mm SL, 0.45-2.54 kg, were tested. Four were nontoxic, and one each was poisonous at the 1 and 2 levels. One specimen from Bikini, 400 mm SL, was nontoxic, while a second, 450 mm SL (2.3 kg), gave a mon- goose test of 3. Randall and Brock {I960) found food in the stomachs of 3 of 12 specimens (identified as E. elongatus) collected in the Society Islands. All had eaten fishes; in addition, one stomach contained a crab chela. Thirty-four specimens, 204-500 mm SL, from Enewetak, Society Islands, Line Islands, Cook Is- lands, Rapa, and the Red Sea were examined for food. Nineteen had empty stomachs, one contained a crab, and the rest had eaten fishes, of which one could be identified to species {Adioryx lacteogut- tatus) and three to family (Pomacentridae, Holocentridae, and MuUidae). Plectropomus leopardus (Lacepede) (Figure 21): This is the largest and most common of four group- ers of the genus Plectropomus in the Marshall Islands. The genus is readily distinguished from other Micronesian serranid genera in having eight dorsal spines and large canine teeth in the jaws; also the body is more elongate than most other groupers. Plectropomus leopardus is reddish with small dark-edged blue spots and an emargin- ate caudal fin. This species is among the worst offenders in Oceania for causing ciguatera. Halstead (1967) listed it as poisonous from Jaluit in the Marshall Islands and in the Tuamotus. He cited 10 papers that have reported on its toxicity in the Pacific, among them Randall (1958) who noted it as the most toxic of the groupers in Tahiti and reported his own poisoning from the Tuamotus. Thirty-one specimens were collected at Enewetak for ciguatoxin content, mainly by spearing. The fish ranged from 426 to 790 mm SL and weighed from 1.8 to 11.8 kg. Twelve were nontoxic, six were poisonous at the 1 level, eight at 2, four at 3, and one (520 mm SL) was a 5. One specimen (8.2 kg) from Bikini was nontoxic. Randall and Brock (1960) recorded the food of seven specimens from the Society Islands: four had empty stomachs and the rest had eaten fishes. Thirty-seven specimens 426-790 mm SL were collected for food-habit study from Enewetak, So- ciety Islands, and Okinawa. Fifteen had empty stomachs, and the rest contained fishes. Five had eaten parrotfishes (one grouper, 702 mm SL, con- tained a Scarus gibbus 3 13 mm SL). A 643 mm fish contained two acanthurids, one of which was a Ctenochaetus striatus 153 mm SL. A 705 mm grouper had eaten a labrid, Cheilinus undulatus, 270 mm SL. Two others had groupers in their stomachs, a 659 mm fish contained E. tauvina 267 mm SL, and a 790 mm fish contained a half- digested Epinephelus sp. Plectropomus melanoleucus (Lacepede) (Figure 22): This distinctively colored grouper, white with Figure 2l.— Plectropomus leopardus, 513 mm SL, 3.4 kg, Enewetak, Marshall Islands. 221 FISHERY BULLETIN: VOL. 78, NO. 2 Figure 22. — Plectropomus melanoleucus, 492 mm SL, Enewetak, Marshall Islands. black saddlelike bars, scattered small blue spots, and yellow fins has been regarded as a color phase of P. maculatus (Bloch) by a number of authors from Boulenger (1895) to Smith and Smith (1963). Plectropomus melanoleucus, however, is a valid species. In addition to color, it differs from P. maculatus (and P. leopardus and truncatus) by usually having 17 instead of 16 pectoral rays. This species is rare in Oceania. Only a single specimen, 506 mm SL, 2.95 kg, was taken at Enewetak during the ciguatera survey. Its viscera produced a reaction of 2 when fed to a mongoose. Its stomach was empty. Plectropomus truncatus Fowler (Figure 23): Like P. leopardus, this grouper has dark-edged blue spots, but the spots are larger in fishes of about the same size. The best field character to distinguish this species is its truncate caudal fin. One specimen (384 mm SL, 1.45 kg) from Enewetak produced a ciguatoxic reaction of 2; its stomach was empty. Hiatt and Strasburg(1960) reported one of three specimens of this grouper collected at Enewetak with a holocentrid fish in its stomach. Variola louti (Forsskal) (Figure 24): This color- ful grouper is yellowish brown to orange, profusely spotted with blue or pink (blue from shallow wa- ter, pink in deeper water), with broad zones of yellow posteriorly on the median and pectoral fins. Apart from color, it is readily distinguished by its deeply concave caudal fin. It is usually found on outer reefs at depths >15 or 20 m. Figure 23. — Plectropomus truncatus, 350 mm SL, Enewetak, Marshall Islands. 222 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Figure 24.— Variola louti, 273 mmSL, Enewetak, Marshall Islands. Variola louti is well known as a cause of cigua- tera. In Mauritius it is prohibited from sale in fish markets. Toxicity there has been reported by Wheeler in Wheeler and Ommanney (1953). Nineteen specimens were obtained for testing at Enewetak. They ranged from 352 to 418 mm SL and weighed from 1.1 to 2.2 kg. Thirteen were nontoxic, four gave a mongoose test of 1, and two a test of 3. Two from Bikini, 292 and 420 mm SL, were nontoxic. Randall and Brock (1960) opened seven stomachs of the species from the Society Islands. Five were empty and two contained digested fishes. Hiatt and Strasburg ( 1960) found a juvenile unicornfish, Naso sp., in the stomach of one of two specimens from Bikini. Helfrich et al. (1968) examined the stomach contents of 44 specimens from the Line Islands. They found fishes, includ- ing acanthurids, balistids, and muraenids, in 80% of the stomachs, and crustaceans in 11%. For the present food-habit study 44 specimens were examined. These ranged from 180 to 560 mm SL (largest weighed 5.45 kg). They were caught at Enewetak, Bikini, Line Islands, Tahiti, Rarotonga, Pitcairn Group, Rapa, and Tutuila. Twenty had empty stomachs. One contained a crab, one a spiny lobster, and the rest had eaten fishes, of which the following were identified at least to family: Adioryx sp. ( 120 mm specimen in a 295 mm grouper), Scorpaena sp. ( 28 mm specimen in a 235 mm grouper), Parupeneus trifasciatus (a 53 mm transforming specimen in a 470 mm grouper), juvenile Chaetodon sp., Anampses caeruleopunctatus (identified from scales), Scarus sordidus ( 140 mm specimen in a 375 mm grouper), and a pomacentrid. Lutjanidae (Snappers) Aprion uirescens Valenciennes (Figure 25): This elongate snapper has been reported as poisonous from a number of Pacific localities, including the leeward Hawaiian Islands (Halstead 1967). Halstead did not list any Indian Ocean localities. It may therefore be worthy of note that the author incurred a mild case of ciguatera from eating a fish of this species at Mauritius. Also he was informed that poisoning is known from nearby Reunion. Aprion uirescens is a roving carnivore of open water but often found within or near reef areas, both in atoll lagoons and in outer reef zones. It is difficult to approach underwater; most of the specimens were obtained by hook and line, often while trolling. Eleven specimens, 435-622 mm SL (1.6-5.2 kg), from Enewetak were tested. Eight were nontoxic, one (589 mm) gave a reaction of 1, one (622 mm) was a 3, and one (620 mm) a 4. Eight were taken at Bikini ranging from 406 to 685 mm SL (1.8-5.4 kg). All, exept one of 457 mm SL which produced a mongoose reaction of 1, were nontoxic. Ommanney in Wheeler and Ommanney (1953) reported on the analysis of stomach contents of 80 A. uirescens from the Mauritius-Seychelles region. Forty-four of these were empty. The stomachs of 223 FISHERY BULLETIN: VOL. 78, NO. 2 Figure 25. — Aprion virescens, 503 mm SL, 2.5 kg, Nuku Hiva, Marquesas Islands. 21 contained fishes; 6 had fishes and macro- plankton; 5 had only macroplankton; and 4 con- tained cephalopods. Among the fishes taken from the stomachs were scarids, ostraciids, siganids, a bothid, and Caesio coerulaureus. Talbot (1960) reported on 259 specimens caught by handline and surface lure on the east African coast which ranged from 202 to 800 mm SL (weight to 11.3 kg). He presented a diagram of the relative abundance of food organisms for this fish as follows: fishes 49%, plankton 17%, cephalopods 14%, and crustaceans exclusive of plankton (mainly portunid crabs) 12%. He did not indicate how many specimens had empty stomachs. The stomachs of 15 specimens from the Mar- shall Islands and 1 from Hawaii were examined. Ten were empty. Four contained fishes; (one prey fish identified as Scarus sp.); one A. virescens (481 mm SL) had also eaten an octopus (one-third stomach volume). A 457 mm fish contained a 10 mm calappid crab, and one of 650 mm a stomatopod. Lutjanus hohar (Forsskal) (Figure 26): This red snapper has been implicated more frequently in ciguatera than any fish of the Indo-Pacific region. It is probably the species which sickened the crew of Captain Cook in the New Hebrides in 1774 (Banner 1965). Its toxicity has also been reported under the junior synonym Lutjanus coatesi Whit- ley. This species occurs along seaward reefs and in passes. It is more common around atolls and low coral islands than high islands (Randall and Brock 1960). It is especially abundant in the Line Is- lands. Reef fishes became highly toxic there dur- ing and immediately after World War II; the toxic- ity declined in the early 1960's (Banner and Helfrich 1964 ). When the toxicity was high, L. bohar from these islands was used for the chemical and pharmacological work on ciguatoxin at the Uni- versity of Hawaii (replaced by Lycodontis javanicus from Johnston Island in later years). It was the species used by Banner et al. (1966) to demonstrate the long periods of retention of ciguatoxin in the tissues of poisonous fishes when removed from the source of the toxin. The toxicity of 95 specimens from Enewetak from 430 to 635 mm SL (2.5-7.5 kg) was tested. Fifty-six were nontoxic; 22 gave a mongoose test of 1; 11 were 2, 5 were 3, and 1 (533 mm) was a 5. From Bikini 143 specimens which ranged from 330 to 760 mm SL were tested. Of these, 112 were nontoxic, 15 were 1, 8 were 2, 6 were 3, and 2 gave a mongoose test of 4. From the atoll of Rongelap (lat. 11° N, long. 167° E) in the Marshall Islands we obtained 12 specimens of L. bohar which weighed from 3.2 to 9.1 kg. Eight of these were nontoxic, two gave a reaction of 2, one was a 3, and one a 5. Hiatt and Strasburg (1960) found fragments of fish in one of two specimens of L. bohar from Bikini. Talbot (1960) examined 854 specimens from the East African coast; 58% had empty stomachs. Fishes composed 62% of the food mate- rial, crustaceans 24%, and mollusks 8%. Helfrich et al. (1968) determined the diet of 2,276 speci- mens from Palmyra and Christmas Islands in the Line Islands; 21.4% of these were empty. Fishes dominated the stomach contents (48.7% by volume at Palmyra and 65.4% at Christmas), of which acanthurids were the most common among those identified. Mollusks represented 19.1% by volume 224 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Figure 26.— Lutjanus bohar, 520 mm SL, 4.9 kg, Ulithi, Caroline Islands. of the total food material at Palmyra and 18*7?^ at Christmas. Crustaceans (principally decapod megalops) composed 15.4% of the food among Palmyra fish and 13.3% of Christmas Island specimens. The stomachs of 121 adult specimens of L. bohar from the Marshall Islands, 330-635 mm SL, most of which were taken by hook and line, were examined. Eighty-six were empty. Of those with identifiable food, 76.2% contained fishes (includ- ing Lycodontis sp., Cephalopholis urodelus, Ar- chamia sp., Lethrinus variegatus , Scarus sp., and Ostracion sp.), 10.8% had eaten crabs (including portunids), 8.7% contained octopus, and 4.3% shrimps. Lutjanus fuluus (Schneider) (Figure 27): The names L. vaigiensis (Quoy and Gaimard) and L. marginatus (Cuvier) are junior synonyms that have often been used for this snapper. It is yel- lowish on the body, the head gray, the caudal fin reddish black with a narrow white posterior bor- der; the dorsal fin is reddish and the anal and pelvic fins yellow. It is a small inshore species, abundant throughout the Indo-West Pacific. It is found more often in sheltered than exposed envi- ronments. Hiyama (1943:48-49, pi. 6, fig. 17) re- ported that Marshallese natives informed him that this fish (which he identified as L. flavipes Valenciennes), rarely causes ciguatera; when it does, the cases are light. Halstead (1967:98, pi. 68, fig. 4) listed it among the ciguatoxic fishes [misidentified as L.janthinuropterus (Bleeker)]. Two specimens from Enewetak, 207 and 217 mm SL, were nontoxic. Randall (1955a) examined the stomachs of six specimens taken with rotenone at Tarawa, Gilbert Islands. One had eaten a small holothurian, one a brachyuran crab, and two contained fishes that were probably prior victims of the ichthyocide. Hiatt and Strasburg (1960) analyzed the stomach contents of six juveniles from Amo, Marshall Is- lands; they reported the following food items: crabs, fishes, amphipods, shrimps, and stomatopods. Randall and Brock (1960) examined 50 specimens which had food in their stomachs; 54.3% of these contained crustaceans (mainly crabs) and 42.4% fishes. Helfrich et al. (1968) col- lected 51 specimens from Palmyra for food-habit study. The dominant food items were mugilid, mullid, and pomacentrid fishes; crustaceans made up the next most frequent organisms of the diet. For the present study 44 specimens 182-250 mm SL were collected in the Marshall Islands, Mariana Islands, and Caroline Islands. Thirty-one of these had empty stomachs. Of those with food, 68.4% had eaten crustaceans (nearly all crabs, mainly calappids) and 31.6% fishes. Lutjanus gibbus (Forsskal) (Figure 28): This snapper is also reddish like L. bohar, but it does not attain such large size. The dorsal profile of adults, beginning with the nape, is highly convex, which is the basis for the specific name. Schultz in Schultz and collaborators (1953) stated, "This species was taken only in moderately deep water. 225 FISHERY BULLETIN: VOL. 78, NO. 2 Figure 27. — Lutjanus fulvus, 215 mm SL, Enewetak, Marshall Islands. Figure 28. — Lutjanus gibbus, 291 mm SL, Palmyra, Line Islands. It did not occur over the shallow parts of the reefs." Talbot (1960), on the other hand, wrote in refer- ence to L. gibbus in east Africa, "It was only found 226 in shallow water of from 3 to 8 fathoms." Actually, the species may occur either in the shallows or at moderate depths, but in the Marshall Islands, at RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS least, it is usually encountered in water of 20 m or more. Though mainly found on the outside of sea reefs and in passes, it may also occur in lagoons. It is often observed in large aggregations. Thirty-one specimens of L. gibbus from Enewetak 302-375 mm SL (largest 1.8 kg) were collected. Twenty-one of these fishes were non- poisonous; five were rated as 1, two were 2, and three ranked 3 by the mongoose test. Thirty-five specimens from Bikini, 279-385 mm SL, were tested. All but three were nontoxic; the three toxic fish produced a mongoose reaction of only 1. Randall and Brock (1960) collected 23 speci- mens in the Society Islands of which only 9 had food in their stomachs (5 of these were juveniles). The four adults contained fishes, crabs, and un- identified crustaceans. Hiatt and Strasburg( 1960) examined 43 specimens (175-260 mm SL) from the Marshall Islands of which 10 had empty stomachs. Crustaceans were the main food, especially crabs (60% contained xanthids and 17% portunids); Amphineura were found in 13% of the stomachs. Octopus, Natica,Ptychodera, small holothurians, polychaetes, sipunculids, and ^sh. Apogon were all found in 4% of the stomachs. Talbot (1960) re- ported on the capture of 121 specimens. He wrote, "Foods eaten were mainly crustaceans, including crabs and Penaeid prawnis. Small coral fishes were also occasionally taken." Helfrich et al. (1968) found food in 36 of 45 stomachs of adults from the Line Islands; fishes were the main item of diet, with crustaceans the second most abundant. [Fishes included unidentified eels, acanthurids, and Pomacentrus nigricans (= Stegastes nigri- cans).] Most crustaceans were brachyuran crabs, but there were also alpheid shrimps and slipper lobster. Mollusk remains were mainly proso- branchs, but opisthobranchs and cephalopods were also found. Sea urchins were the most com- mon of the miscellaneous invertebrates composing the rest of the stomach contents. During the present study the stomachs of 51 specimens from the Marshall Islands, 260-419 mm SL, were examined. Twenty-seven were empty. Of those with food, 40% had eaten crabs, 26% fishes (including Pseudocheilinus sp. and Adioryx mi- crostomus), 17% echinoids {inclxiding Eucidar is sp. and Heterocentrotus mamillatus), 12% ophiuroids (including Ophiocoma erinaceus), 2.1% alpheid shrimps, 2.1% octopus, and 0.3% gastropods. Lutjanus monostigmus (Cuvier) (Figure 29): This species, named from the blackish spot usually present on its side (on lateral line), is capable of causing severe cases of ciguatera. Belotte (1955) gave the case history of an American who was in a coma 3 days after eating this snapper in Tahiti; the author also interviewed this man. The sale of this species in Tahiti, where it is called "taivaiva," (Randall 1972) is forbidden. It is found in reef environments from shallow water to moderate Figure 29.— Lutjanus monostigmus, 249 mm SL, Florida Island, Solomon Islands. 227 FISHERY BULLETIN; VOL. 78, NO. 2 depths, especially where there is deep shelter. Not infrequently it is encountered in small aggrega- tions. Adults are wary, hence difficult to spear. Only three specimens were obtained from Enewetak, 310-420 mm SL (0.8-1.6 kg), for testing of toxicity. One was nontoxic, one was 1, and one a 2. Five specimens, 400-445 mm SL, were collected in Bikini. Four were nontoxic; the largest gave a mongoose reaction of 1. Randall and Brock (1960) opened 32 stomachs of adults of this species, of which 18 were empty. Those with full stomachs all contained fishes, among them Decapterus pinnulatus, Selar crumenophthalmus, and Ctenochaetus striatus. Hiatt and Strasburg ( 1960) found a goatfish in the stomach of one of three specimens from Enewetak; the other two were empty. Talbot (1960) collected 18 specimens off east Africa. He reported fish re- mains (including a mullid and a labrid) in most stomachs; penaeid prawn remains were also found. Helfrich et al. (1968) examined 29 speci- mens from the Line Islands. They found fishes in 92% of the stomachs and crustaceans (stomatopod larvae and one slipper lobster) in 23%. For the present food-habit study 41 specimens of L. monostigmus were examined from the Marshall Islands, Society Islands, Line Islands, and Samoa Islands. Twenty-three had empty stomachs. Of those with food, 92% by volume had eaten fishes (including the holocentrid Adioryx microstomas, acanthurids, and a balistid), and 8% crabs (includ- ing a portunid). Macolor niger (Forsskal) (Figure 30): Although not previously reported as poisonous, this lutjanid fish attains moderate size, is a reef-dweller, and carnivorous; this would seem to have the potential for causing ciguatera. A total of 25 adults, 403-445 mm SL (2.3-2.95 kg), were taken, all from Enewetak, and mainly from explosive stations in the lagoon. Twenty-three were nontoxic and two gave a reaction of 1 on the feeding of liver and viscera to mongooses. The stomachs of eight adult specimens taken at 9:00 a.m. at Enewetak were examined. All were empty. The large eyes of this species is suggestive of nocturnal habits, and the numerous (about 72) long gill rakers would seem to indicate at least some feeding on zooplankton (perhaps more im- portant in smaller individuals than large adults). Some of the specimens were caught by hook and line baited with fish. Hiatt and Strasburg (1960) also reported that this species can be caught on a baited hook. Lethrinidae (Emperors) Lethrinus amboinensis Bleeker (Figure 31): Fol- lowing Sato (1978), this emperor is identified as L. amboinensis. It lacks characteristic color mark- ings, being light brownish to greenish dorsally Figure 30.— Macolomiger, 378 mm SL, Enewetak, Marshall Islands. 228 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Figure 31. — Lethrinus amboinensis, 294 mm SL, Enewetak, Marshall Islands. with small dark brown spots and blotches, shading to white ventrally. It is somewhat elongate, the head length greater than the depth; the snout is moderate, its length in adults 1.8-1.0 in head length; the maxilla reaches a vertical a little pos- terior to the anterior nostril. The teeth along the sides of the jaws are conical. Twenty-four specimens were taken at Enewetak and nine at Bikini, the largest 310 mm SL. All were nontoxic. Helfrich et al. (1968) determined the food of 14 specimens of L. amboinensis from Palmyra, Line Islands. Fishes were found in 75% of the stomachs, mollusks in 25%, and crustaceans in 17%; all specimens had some sea urchin fragments. Fish remains were found in one of two stomachs examined at Bikini. Lethrinus kallopterus Bleeker (Figure 32): This Lethrinus is distinctive in having orange fins and blackish spots over occasional scales; the snout is short, the maxilla reaching a vertical at anterior edge of eye. The teeth at the sides of the jaws are nodular (i.e., neither conical nor well-developed molars). It was most often seen in the deeper parts of the atoll lagoons. A total of 19 specimens were collected at Enewetak for the testing of toxicity. These ranged from 337 to 443 mm SL (1.1-2.7 kg). Fourteen were nontoxic, two produced a reaction of 1, two were 2, and one (368 mm SL) was a 5. Two specimens, 330 and 457 mm SL, were pro- cured from Bikini; neither was toxic. The stomachs and intestines of 13 specimens, 330-443 mm SL, from the Marshall Islands were opened. Five of the fish were empty. Four had eaten only echinoids (including Echinometra mathaei); one contained mostly echinoids but also the cowrie Cypraea carneola; another (the largest) had eaten just the cowrie C. vitella; still another had a cowrie in its gut (20% by volume of the food material), and the rest of the food material con- sisted of crinoids; one specimen contained only a starfish arm. Lethrinus miniatus (Forster in Bloch and Schneider) (Figure 33): This emperor has an espe- cially long snout (1.6-1.8 in head length of adults). It is primarily gray in color, but can alter its pat- tern, like many other Lethrinus, to one of dark irregular bars and blotches. Often there are two or three bluish streaks on the snout passing an- teriorly and diagonally downward from the eye. The teeth on the sides of the jaws are conical. This species was seen in both lagoon and outer reef environments, but mainly in lagoons. It is among the largest of the emperors, reported to attain Im. Of nine adults, 435-530 mm SL (1.8-3.6 kg), which were caught at Enewetak, six were non- toxic, and three gave a reaction of 3. Twelve specimens from Bikini, 381-635 mm SL, 1.4-7.3 kg, were nontoxic. Eight of 14 specimens from Enewetak and Bi- kini had food in their stomachs. Three contained fish remains, one of which (456 mm SL) included a 229 FISHERY BULLETIN: VOL. 78, NO. 2 Figure 32. — Lethrinus kalloptenis, 345 mm SL, Enewetak, Marshall Islands. Figure 33. — Lethrinus miniatus, 430 mm SL, Enewetak, Marshall Islands. Lethrinus 115 mm SL; the remaining three had eaten crustaceans (stomatopod, crab, and alpheid shrimp). Lethrinus xanthochilus Klunzinger (Figure 34): This emperor is one of the more slender species of Lethrinus (depth 3.1-3.3 in SL). The interorbital space is nearly flat. The teeth on the sides of the jaws are conical. The upper lip is orange-yellow, and there is a red spot at the upper pectoral base. It is found more in lagoons than exposed reef habitats; it will venture into shallow water. Two specimens, 445 and 550 mm SL (smallest 1.7 kg, largest not weighed), were collected at 230 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Figure 34. — Lethrinus xanthochilus, 395 mm SL, Fanning Island, Line Islands. ^ ^^^^.^.M'- .^ , \ Figure 35. — Monotaxis grandocuHs, 220 mm SL, Enewetak, Marshall Islands. Enewetak; two, 305 and 432 mm SL, were taken at Bikini. All were nontoxic. The stomach and gut contents of these four specimens and one of 466 mm from the Society Islands were examined. One fish contained crushed echinoids, one the remains of a calappid crab, one a digested fish, another both crab and fish remains, and one (the largest) a freshly in- gested fish (probably from a rotenone station). Monotaxis grandoculis (Forsskal) (Figure 35): M. grandocuHs has been classified in the past princi- pally in the Sparidae or Lutjanidae, but is now recognized as a lethrinid. It is readily distin- guished by its large eyes, short blunt snout, and single row of well-developed molariform teeth along the side of the jaws. It occurs in a wide variety of reef habitats. Adults are difficult to ap- proach underwater. This fish feeds mainly on in- 231 FISHERY BULLETIN: VOL. 78, NO. 2 vertebrates with calcareous or chitinous hard parts. Although rarely implicated in serious cases of poisoning, it is capable of being ciguatoxic. Halstead (1967) listed 11 references attesting to its toxicity. Five specimens, 277-362 mm SL (0.73-1.6 kg), from Enewetak were tested for toxicity. Four were nontoxic and one gave a reaction of 1 from the mongoose feeding. Randall (1955a) reported on the gut contents of two specimens, 158 and 160 mm SL, from the Gil- bert Islands; these consisted mainly of crushed shells of small mollusks and sea urchins. Hiatt and Strasburg (1960) examined the contents of the digestive tracts of eight specimens, 195-220 mm SL, from the Marshall Islands. One fish was empty. Crushed gastropods (including Atys sp. and Cerithium sp.) were found in all stomachs, pelecypods in 71%, crabs in 42%, hermit crabs in 28% , and spatangids and polychaetes each in 14% . Hobson ( 1974) collected five specimens in Hawaii. He found the principal prey, in order of importance in the diet, to be prosobranch gastropods, ophiuroids, echinoids, opisthobranch gastropods, and pagurid crabs. Forty-eight specimens of M. grandoculis, 155- 440 mm SL, were collected from the Marshall Is- lands, Line Islands, Cook Islands, Society Islands, Pitcaim, Hawaiian Islands, New Guinea, and the Red Sea for the study of food habits. Unless fish of this species are captured during the night or very early morning hours, their stomachs are nearly always empty. Occasional feeding by M. grand- oculis does occur during the day, as indicated by Hiatt and Strasburg's (1960) observation of its "blowing" away sand to expose fossorial forms. Also one specimen taken at 2 p.m. during the au- thor's survey had the remains of a freshly ingested crab in its stomach. Five of the fish collected in late afternoon hours had completely empty digestive tracts. The remaining fishes contained, on a vol- ume basis, 39.4% gastropods, 18.9% crabs, 16.8% pelecypods, 13.9% echinoids (principally Echinometra and spatangoids such as Clypeaster), 6.1% pagurid crabs, 1.7% ophiuroids, 1.2% polychaetes, 1.0% unidentified worms, 0.7% fishes, and 0.3% foraminifera. Kyphosidae (Sea Chubs) Kyphosus cinerascens (Forsskal) (Figure 36): This chub may be distinguished from the other Kyphosus by the high soft portion of the dorsal fin (longest dorsal spine contained about 1.8 times in longest soft ray). It occurs in lagoon or outer reef areas and is often seen in loose aggregations. It is associated with hard substratum for its algal food, generally in the vicinity of crevices or caves with more than one entrance. Bartsch et al. (1959) re- ported this species as toxic from Majuro, Marshall Islands, but their data and the few other records of Figure 36. — Kyphosus cinerascens, 234 mm SL, Enewetak, Marshall Islands. 232 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS toxicity would seem to indicate that it is only mar- ginally a cause of ciguatera. Of two specimens from Enewetak, 356 and 364 mm SL, 1.8 kg, one was nonpoisonous and the other toxic at level 1. Hiatt and Strasburg ( 1960) found benthic algae in the stomachs of three of four specimens examined in the Marshall Islands. Hobson (1974) reported algae in three stomachs from Hawaii; however, K. cinerascens apparently does not occur in the Hawaiian Islands (though two other species are present). Three specimens, 330-364 mm SL, from Enewetak were opened. The stomach of one was empty, and the other two contained benthic algae. Q The algae of one were identified by Tsuda as the reds Gelidium pusillum, Champia parvula, and Leveillea jungermannoids (90%) and the brown Sphacelaria tribuloides. Carangidae (Jacks) Caranx ignobilis (Forsskal) (Figure 37): This steep-headed jack is the largest species of the genus. Bagnis et al. (1972) stated that it can attain a length of 2 m and a weight of 80 kg. It can be differentiated from other Marshall Islands species *Roy T. Tsuda, Marine Laboratory, University of Guam, Box EK, Agana, Guam 96910, pers. commun. 1972. by the absence of scales on the thorax except for a small median patch. Like other large carangids, it is a roving carnivore; it may be encountered any- where in the atoll environment including water surprisingly shallow for such a large fish. The author interviewed a man and wife in Moorea who were poisoned from eating the liver of a large individual of this species (estimated 1.5 m) which overturned their canoe in the long struggle to catch it. Both were very ill with ciguatera, the man comatose for several hours. Five specimens, 573-920 mm FL, 3.6-16.3 kg, were obtained at Enewetak for the testing of toxic- ity . Three gave a reaction and two a reaction of 1 . Two specimens from Bikini, 635 and 1,105 mm FL, 4.5 and 27.3 kg, were nontoxic. A total of 14 specimens were collected for food- habit study from the Marshall Islands, Line Is- lands, Hawaiian Islands, Pitcairn Group, and the Marquesas. Seven stomachs were empty, and the rest contained the digested remains of fishes, of which only one could be identified to species, the surgeonfish Zebrasoma flauescens. One stomach-content fish was a scorpaenid, and another (from a jack of 1,217 mm FL, 37.5 kg) a scar id. Caranx lugubris Poey (Figure 38): The black jack is a circumtropical species with a well-earned reputation for causing ciguatera. Although the Figure 31 .—Camnx ignobilis, 378 mm FL, Fanning Island, Line Islands. 233 FISHERY BULLETIN: VOL. 78, NO. 2 (^ r T *¥ Figure SS.—Camnx lugubris, 557 mm FL, 3.1 kg, St. John, U.S. Virgin Islands. dark color (especially on the scutes) and distinc- tive configuration generally permit identification, the fully scaled breast will provide separation from C. ignobilis, the low number of scutes on straight portion of lateral line (26-33) from C. melampygus, and the high gill raker count (18-20 on lower limb) from C. sexfasciatus. This species is found mainly around oceanic islands and is nearly always encountered in the clear water of outer reef environments. Fifteen specimens were obtained at Enewetak for testing. These ranged from 488 to 910 mm FL and weighed from 2.5 to 15.5 kg. Eleven were nonpoisonous, two gave reactions of 1, one was a 2, and one a 3. No specimens were collected at Bikini. Randall (1955a) reported a fish in one of two specimens collected in the Gilbert Islands and Randall ( 1967) found fishes in two of six specimens from the Caribbean Sea. For the present food-habit study, 10 specimens were obtained from Enewetak and Henderson Is- land in the Pitcairn Group. Four had empty stomachs, and the remaining six contained the remains of fishes, one of which was a labrid. Caranx melampygus Cuvier and Valenciennes (Figure 39): This is the most abundant jack of the genus in Oceania; it is widespread in the tropical 234 and subtropical Indo-West Pacific and ranges to the eastern Pacific as well. It is irridescent blue along the back and median fins in life with a scat- tering of small blackish spots on the head and body except ventrally. The chest is completely scaled, and there are 38-44 scutes in the straight portion of the lateral line. Thirty specimens were collected at Enewetak, 417-722 mm FL, 1.4-6.8 kg, for the assay of cigua- tera. Twenty-four were nontoxic, four gave a reac- tion of 1, one was a 2, and one a 3. Six specimens from Bikini, 394-686 mm FL, 1.8-6.6 kg, were nontoxic. Randall ( 1955a) examined the stomach contents of four specimens from the Gilbert Islands. Two contained many small freshly ingested fishes which were probably the result of a rotenone sta- tion kill. Of the other two which were speared, one contained the anthiine fish Mirolabrichthys tuka (= Anthias pascalus) . Hiatt and Strasburg (1960) found fish in the stomachs of two from the Mar- shall Islands, one of which was identified as Trachurops {— Selar) crumenophthalmus . Hobson (1974) examined the stomach contents of six specimens from Hawaii. One contained larval fishes and mysids, a second had fish and shrimp remains, and three contained well-digested frag- ments at least one of which was fish. RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Figure 39.— Camnx melampygus, 498 mm FL, 1.9 kg, Sanganeb Atoll, Red Sea. Sixty-one specimens, 298-722 mm FL, from the Marshall Islands, Hawaiian Islands, Line Islands, Marcus Island, Solomon Islands, and the Red Sea were collected for stomach-content study. Seven- teen stomachs were empty. All the others con- tained the digested remains of fishes, though one had, in addition, a squid pen. The following fishes were identified from the stomach material: eel, Anthias thompsoni ,Caranx sp. (90 mm FL in a 520 mm C. melampygus), Priacanthus cruentatus, Cirrhitops fasciatus ,Caesio sp.,Parupeneus sp.,P. trifasciatus,Pomacentrus pavo, Chromis caerulea, labrid, Thalassoma purpureum, Ptereleotris mi- crolepis, Caracanthus unipinnus , Acanthurus tri- ostegus, acronurus stage of acanthurids (in two stomachs), and a subadult acanthurid. Caranx sexfasciatus Quoy and Gaimard (Figure 40): This jack, which ranges from the Red Sea to eastern Oceania, is closely related to C. hippos of the Atlantic. It is usually seen in small schools, but is not common in the Marshall Islands. It is more elongate than the Caranx spp. discussed above, and it has a larger eye. The lower-limb gill raker count is 15-17. The scutes are blackish, there is a small black spot at the upper end of the gill opening, and the soft dorsal and anal fins are tipped with white. The dark bars of the young are the basis for the specific name. Only two specimens were caught at Enewetak, 496 and 700 mm FL, 2.3 and 4.6 kg. Both were nontoxic. No specimens were obtained from Bikini. Ommanney in Wheeler and Ommanney (1953) reported on the stomach contents of specimens caught during a survey of the Mauritius- Seychelles region. Eight specimens contained fish remains, one had squid remains, one had megalops larvae, and nine were empty. A par- rotfish and two eels were noted among the stomach contents. The stomachs of six specimens of C. sexfasciatus, 385-700 mm FL, from Enewetak and Tahiti were opened. Five were empty, and one contained well-digested fish remains. (This jack was speared at 11:30 a.m.) Bagnis et al. ( 1972) stated that C. sexfasciatus is nocturnal. Wheeler and Ommanney (1953) on the other hand, wrote, "It often takes a lure..."; pre- sumably he meant one trolled by day. Scombridae (Tunas) Gymnosarda unicolor (Riippell) (Figure 41): The dogtooth tuna is named for its large conical teeth; it is also unique in having two patches of villiform teeth on the tongue. It lacks dark stripes or spots on the body; the second dorsal and anal fins are 235 FISHERY BULLETIN: VOL. 78, NO. 2 Figure 40. — Camnx sexfasciatus, 472 mm FL, 1.8 kg, Enewetak, Marshall Islands. Figure 41. — Gymnosarda unicolor, 645 mm FL, 3.6 kg, Enewetak, Marshall Islands. white tipped. A large species, Masuda et al. (1975) recorded it to a length of 2.4 m. Unlike other large tunas, in general, it occurs in relatively shallow coastal water, often around coral reefs, and it read- ily penetrates the deeper lagoons of atolls. Thirteen individuals were collected from Enewetak which ranged from 550 to 1,350 mm FL (3.2-35.4 kg). Seven caused no symptoms when liver tissue was fed to mongooses; four produced a reaction of 1, one was a 2, and one a 3. Three from Bikini, 737-940 mm FL, 6.4-11.8 kg, were nontoxic. Hiatt and Brock (1948, after unpublished data of J. Marr and O. Smith) stated that the scad, 236 Decapterus sanctaehelenae, was most frequently encountered in the stomachs of dogtooth tunas in the Marshall Islands. Schultz in Schultz and col- laborators (1953) reported that D. muroadsi and Caesio xanthonotus were regurgitated by Gym- nosarda nuda ( = G. unicolor) which were caught at Bikini. Five of 17 specimens from the Marshall Islands taken during the survey had empty stomachs. The others contained fishes, five of which were iden- tified as: Naso brevirostris, N. vlamingii, Cir- rhilahrus sp., Caesio sp., Pterocaesio sp. The two prey specimens of Naso were large adults. The A'^. vlamingii, taken from the largest G. unicolor, measured 370 mm SL. RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Labridae (Wrasses) Cheilinus undulatus Riippell (Figure 42): The giant humphead wrasse is one of the largest of bony fishes. It has been recorded to a length of 2.29 m and a weight of 190.5 kg (Marshall 1964). The hump on the forehead develops only on larger in- dividuals. Two dark lines which extend pos- teriorly from the eye are useful in identifying juveniles and subadults of this species. It is usu- ally found on outer reef slopes or in deep channels, but also occurs in lagoons. It is difficult to ap- proach underwater. According to Bagnis et al. (1972) individual fish have a home cave to which they retreat when threatened and to which they retire at night. Randall (1958) reported this species as capable of being moderately to strongly toxic in Tahiti, where it is called "mara" (Randall 1972). It is one of nine species of fishes which are banned from sale in the Papeete market (Bagnis 1968). Seven specimens, 515-995 mm SL, the largest weighing 34.5 kg, were procured at Enewetak for testing. The largest gave a reaction of 2 on feeding to mongooses; the others were 0. Randall et al. (1978) reported on the food habits of the giant humphead wrasse based on the examination of 72 specimens from the Red Sea and islands of Oceania. The diet is highly varied, the dominant groups of food organisms being mol- lusks (gastropods a little more numerous than pelecypods), crustaceans (especially crabs), echinoids, and fishes. The hard parts of the inver- tebrates are crushed to fragments by the powerful pharyngeal dentition. Coris aygula Lacepede (Figure 43): This is one of the two largest species oi Coris (the other an unde- scribed endemic from Lord Howe Island). The largest collected, from the Red Sea, measured 465 mm SL and 583 mm TL. Adult males develop a gibbosity on the forehead similar to that of Cheilinus undulatus, but these two wrasses could hardly be confused; the Coris is more elongate (depth about 3.2 in SL) and has small scales (60-65 lateral line scales for C. aygula, compared with about 25 for Cheilinus undulatus); also, the lateral line of Cheilinus is interrupted. Coris aygula has apparently not been reported as causing ciguatera but because of its large size and similar food habits it would seem to be at least as suspect as C. gaimard which is known to be poisonous at times. The latter is more colorful, displaying bright blue spots and a yellow caudal fin. Five adults of C. aygula, 329-377 mm SL (0.9-1.9 kg), were obtained from Enewetak for testing. One of 368 mm SL ( 1 .4 kg) produced a toxic reaction of 1 when its liver and viscera were fed to a mongoose; the others were nontoxic. Figure 42.— Cheilinus undulatus, 915 mm SL, 25.8 kg, Enewetak, Marshall Islands. 237 FISHERY BULLETIN: VOL. 78, NO. 2 Al-Hussaini (1947) listed the food of the species as gastropods {Turbo, Trochus), Dentalium, and hermit crabs. Hiatt and Strasburg (1960) found the crushed remains of sand-dwelling pelecypods and gastropods in a single specimen (identified as C. angulata) from Enewetak. Randall, G. J. Vermeij, and H. A. Rehder (man- uscript in progress) will report in detail on the food habits of this wrasse. The principal food animals are gastropods, pelecypods, pagurid crabs, echinoids, and brachyuran crabs. Epibulus insidiator (Pallas) (Figure 44): This unmistakable labrid, popularly known as the slingjaw wrasse because of its ability to enor- mously protrude its mouth, occurs from the Red Sea and east Africa to French Polynesia. Halstead (1967) listed nine references citing it as ciguatoxic. Five specimens from Enewetak, 175-228 mm SL, 0.34-0.55 kg, were tested for toxicity. None caused any symptoms in the mongooses. Hiatt and Strasburg (1960) collected one speci- men from Enewetak and one from Bikini for food- habit study; both fish had eaten alpheid shrimps. They wrote, "This wrasse habitually feeds in ramose corals by extending its exceedingly pro- tractile snout into the interstices to capture small alpheid shrimps and xanthid crabs living there." For the present food-habit study 16 specimens, 183-240 mm SL, were collected from the Marshall Figure 43. — Cons aygula, 380 mm SL, Marcus Island. Islands, Johnston Island, American Samoa, and the Society Islands. Two had empty stomachs; six had eaten only fishes and four only crabs. Other food items were shrimps, unidentified crusta- ceans, polychaetes, bryozoans, and unidentified eggs. Scaridae (Parrotfishes) Hipposcarus harid (Forsskal) (Figure 45): Smith (1956) created a new genus, Hipposcarus, for this species on the basis of the triangular patch of scales on the cheek with three or four rows behind, pointed snout, and minute nostrils. Al- though Schultz (1958, 1969) did not recognize this genus, it will be considered valid by Nelson and Randall. Smith (1959) described a Philippine form of this species as new, naming it H. schultzi. Schultz (1969) preferred to regard this form, for which he gave the range central and western Pacific Ocean, as a subspecies, Hipposcarus harid longiceps (Cuvier and Valenciennes). Halstead (1967) has listed four references re- porting the occasional toxicity of this parrotfish. ®G. J. Nelson, Department of Ichthyology, American Museum of Natural History, New York, and the author conferred in Oc- tober 1977 on the generic limits of the Scaridae. Eventual publi- cation is planned. 238 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Figure 44. — Epibulus insidiator, 185 mm SL, Enewetak, Marshall Islands. Figure 45. — Hipposcarus harid, 282 mm SL, Rangiroa, Tuamotu Archipelago. Of four specimens, 340-412 mm SL, 1.3-1.8 kg, obtained at Enewetak, three were nontoxic, and one produced a reaction of 1. Reporting on the stomach contents of Ceto- scarus bicolor, Scarus sordidus, and seven uniden- tified species of Scarus in the Marshall Islands, Hiatt and Strasburg (1960) concluded that they fed mainly on live coral. This is contradictory to the investigation of scarid food habits by Wood- Jones (1910), Choat (1966), Randall (1967, 1974), Rosenblatt and Hobson ( 1969), and Hobson (1974). Randall (1974), however, presented evidence that the largest of the parrotfishes, Bolbometopon muricatus, feed heavily on living coral. Also Glynn et al. (1972) listed three scarids as coral predators off the Pacific coast of Panama. 239 FISHERY BULLETIN: VOL. 78, NO. 2 Scarus gibbus Riippell (Figure 46): The name Scarus microrhinos Bleeker has generally been used in the Pacific for this species (Schultz 1958), and it is under this name that its toxicity has been reported (Halstead 1967). Smith (1959) resur- rected the name Scams j^t66«s Riippell for the Red Sea form of this species, though he still recognized S. microrhinos. Schultz (1969) placed four nomi- nal species, including S. microrhinos, under the one name S. gibbus. The large males are readily distinguished by the near-vertical anterior profile of the head. Other useful characters for distin- guishing the species are four median predorsal scales, three row^s of scales on the cheek, and 16 or 17 pectoral rays. Of 19 specimens, 326-414 mm SL, 1.05-2.8 kg, speared from Enewetak, only 1 of 410 mm (2.3 kg) was slightly toxic (mongoose reaction of 1). Scarus rubroviolaceus Bleeker (Figure 47): This parrotfish w^as selected as the type-species of a new genus, Scarops, by Schultz (1958) principally on the basis of its having a single enlarged row of teeth on each upper pharyngeal bone. This genus, however, was not recognized by Rosenblatt and Hobson (1969). The primary phase of S. rubro- violaceus is reddish with small blackish spots and short streaks on the scales; the terminal male phase (the nominal Pseudoscarusjordani Jenkins and Callyodon africanus Smith were based on this form) is complexly colored, but mainly purplish on the anterior part of the body and abruptly green posterior to about the base of the seventh dorsal spine (this bicolored effect more evident in live than on freshly dead specimens); the head is mainly blue-green, shading to orange-yellow on the opercle, with transverse bands of turquoise and salmon on the lips and chin. There are gener- ally 6 median predorsal scales, 3 rows of scales on the cheek, and 15 pectoral rays. Like the other species of Scarus, S. rubro- violaceus is closely tied to coral reefs. It ranges from east Africa to the tropical eastern Pacific. Although this species has not been reported as poisonous, it would seem to have the same poten- tiality of causing ciguatera as other parrotfishes which may be toxic. Three specimens, 355-370 mm SL, 1.4-1.6 kg, were obtained from Enewetak in order to test for possible toxicity. None were toxic. Rosenblatt and Hobson ( 1969) wrote, "All of the eastern Pacific species of Scarus feed by scraping algae from the surface of rocks. We did not see evidence that they bit off pieces of coral " Scarus rubroviolaceus is one of the four species they studied. Glynn et al. ( 1972), on the other hand, included S. rubroviolaceus among the three scarids they regarded as coral predators from ob- servations off Panama. Acanthuridae (Surgeonfishes) Acanthurus xanthopterus Cuvier and Valen- ciennes (Figure 48): This is the largest member of FIGURE 46.— Scarus gibbus, 417 mm SL, 2.8 kg, Tahiti, Society Islands. 240 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Figure Al.—Scarus rubroviolaceus, 355 mm SL, 1.35 kg, Enewetak, Marshall Islands. FIGURE 48.— Acanthurus xanthopterus, 400 mm SL, 2.3 kg, Enewetak, Marshall Islands. the genus Acanthurus. It is one of a complex of species with a gizzardlike stomach. In the Mar- shall Islands it could only be confused with A. mata, also a large species. The outer third of the pectoral fins of A. xanthopterus are yellowish (fins uniform brown on A. mata), and there are about 4 lengthwise bands in the dorsal fin (about 8 in the fin of A. mata); there are fewer gill rakers (16-22 for A. xanthopterus, compared with 21-25 for A. mata). This species is distributed from east Africa to the eastern Pacific. It occurs more in lagoons and bays than exposed outer reef areas, and it ranges into deeper water than other Acanthurus in general. Also it ventures farther from the cover of coral reefs than other species. Schultz in Schultz and collaborators (1953) used the name Acan- thurus fuliginosus Lesson for this fish, but there is no basis for equating it to Lesson's illustration and description, as explained by Randall (1956). The junior synonym Teuthis crestonis Jordan and 241 FISHERY BULLETIN: VOL. 78, NO. 2 Starks was created for the species from Mexico. Two specimens, 423 and 425 mm SL, 2.7 kg, were obtained from Enewetak for the assay of tox- icity. Neither were toxic. Hiatt and Strasburg (1960) examined the stomachs of four specimens from Enewetak, two of which were empty. The other two contained short filaments of algae with much sand, hydroid hy- drocaulus, and wood splinters (probably from grazing on pilings). Jones (1968) classified A. xanthopterus as a grazer on diatoms and detritus in sand patches. That it will take animal food when the opportunity arises was aptly shown by Helfrich and Banner (1963) who used this species to induce ciguatera toxicity by feeding the poison- ous flesh of Lutjanus bohar. Ctenochaetus striatus (Quoy and Gaimard) (Figure 49): This surgeonfish is much the most common of the four species of the genus that occur in the Marshall Islands. It is, in fact, one of the most abundant reef fishes throughout the Indo- West Pacific region (though not Hawaii). The genus is named for its comblike teeth which are numerous, slender with expanded incurved tips, and flexible in the jaws. Randall (1955b) has dif- ferentiated C. striatus from the other species by having 5-7 denticulations on the expanded distal tips of the upper teeth, the highest average number of dorsal and anal soft rays (modally 29 dorsal rays and 26 or 27 anal rays), and a lunate caudal fin. Bagnis et al. ( 1968) reported that surgeonfishes ( particularly C. striatus ) are responsible for 65% of the cases of ciguatera in Tahiti. There are three reasons for this: 1) the abundance of C. striatus, 2) its good-eating quality, and 3) the knowledge that the symptoms will be mild if ciguatera is incurred. Bagnis (1968) documented the great variation in the symptoms of ciguatera in French Polynesia. He noted that digestive and neurologic symptoms predominated among those patients who had in- gested surgeonfishes. Yasumoto et al. (1971) determined that there are two principal toxins in C. striatus, one of which is fat soluble and chromatographically identical with ciguatoxin, and the other is water soluble. The latter was found only in the liver and gut contents. In a few specimens from Tahiti a differ- ent fat-soluble toxin and a different water-soluble toxin were detected. In order to determine if more than one toxin is present in C. striatus in the Marshall Islands, 22 adult specimens were speared on lagoon reefs of Figure 49. — Ctenochaetus striatus, 94 mm SL, Enewetak, Marshall Islands. 242 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Enewetak and sent frozen to the Laboratory of Marine Biochemistry of the University of Tokyo. The specimens, which weighed from 130 to 230 g each, were pooled in groups of three (four for one group). The flesh and the viscera were separated for each group, and the gut contents of all groups were pooled. Fat-soluble and water-soluble frac- tions were prepared and injected intraperitoneally into mice at doses of 4,000, 2,000, and 1,000 ^g/g- Two mice were used for each dose. If the mice were not killed by a dose of 4,000 ;Ltg/g within the obser- vation period of 48 h, the preparation was re- garded as nontoxic. If they were killed by a dose of 1,000 iJ-g/g, it was classified as strongly toxic, at 2,000 ixg/g moderately toxic, and at 4,000 /xg/g weakly toxic. The results were reported in a letter by the late Yoshiro Hashimoto, then the Director of the Laboratory. None of the preparations were strongly toxic. All the preparations from the flesh were nontoxic. Four of the seven fat-soluble prep- arations of the viscera were moderately toxic, one was weakly toxic, and one nontoxic. Five of the seven water-soluble fractions from the viscera were weakly toxic and the remaining two non- toxic. Both the fat-soluble and the water-soluble preparations of the pooled gut contents were mod- erately toxic. The food habits and mode of feeding of C. strigosus from the Hawaiian Islands were investi- gated by Randall (1955b); underwater observa- tions of C. striatus indicate that its feeding is es- sentially the same. These fishes are detritus feeders. From a near- vertical position (if the bot- tom is horizontal) about 15 mm above the sub- stratum, the fish move abruptly downward with mouth open. The lips and teeth scrape over the surface at the same time that suction is initiated. The soft detrital material and fine inorganic sedi- ment are ingested. If coarse particles of sand are picked up, they are forcefully ejected. The stomach contents of seven adults of C. strigosus from Hawaii consisted of inorganic sediment (up to 90^ by volume); fragments of red, green, and blue- green algae; diatoms; and unidentified soft or- ganic material. In an aquarium experiment C. strigosus was unable to feed on an intact thallus of the filamentous alga Polysiphonia sp. When the same algae was finely fragmented and placed on the bottom, it was readily consumed. Baiistidae (Triggerfishes) Pseudobalistes flavimarginatus (Riippell) (Fig- ure 50): This is one of three large species of trig- gerfishes that occur in the Marshall Islands. It may be distinguished from other balistids by the following characters collectively: the second dor- sal and anal fins elevated anteriorly, five or six rows of spines on the caudal peduncle, no scales on the cheek (of adults), caudal fin of adults emargin- ate, and yellowish margins on the median fins. Woods in Schultz and collaborators (1966) failed to list this species from the Marshall Islands, but the color plate in Hiyama ( 1943, pi. 22, fig. 61) and the study of Hiatt and Strasburg (1960) clearly indicate its presence there. Hiatt and Strasburg stated that it is solitary, uncommon, and occurs on lagoon and interisland reefs in quiet water of 10-30 ft (3.1-9.1 m) deep. Although Hiyama wrote that this fish was not regarded as poisonous in the Marshalls, other records (Halstead 1967) dem- onstrate its capacity for causing ciguatera. It is one of the nine species of fishes forbidden to be sold in the fish market in Papeete, Tahiti (Bagnis 1968). Two specimens, 465 and 535 mm SL, weight not taken, were collected in Bikini. Neither was poisonous. Clark and Gohar (1953) reported pieces of branched coral (Stylophora) 2-3 cm long in the stomach of a specimen 440 mm SL from the Red Sea. Hiatt and Strasburg (1960) examined two stomachs from the Marshall Islands. They found the crustacean Lydia annulipes and isopods, crushed gastropods including Oliva sp., foraminif- era, and colonial tunicate fragments. The stomach and gut contents of only two specimens were obtained for the present study. One of 254 mm SL from the Red Sea was empty. The second of 390 mm SL from Tahiti had eaten Diadema. Balistoides uiridescens (Bloch and Schneider) (Figure 51): This is another large triggerfish for which there have been a few records of toxicity. It shares the elevated anterior part of the second dorsal and anal fins and the rows of spines on the caudal peduncle with P. flavimarginatus, but is differentiated by having its cheek totally scaled and its caudal fin rounded to slightly double emar- ginate as an adult; also the margins of its median fins are broadly blackish. It ranges from the Red Sea to eastern Oceania. It occurs in both lagoons and outer reef slopes. Like other triggerfishes, it has a favorite hiding place in the reef into which it wedges itself when threatened. 243 FISHERY BULLETIN: VOL. 78, NO. 2 Figure 50. — Pseudobalistes flavimarginatus , 265 mm SL, Palmyra, Line Islands. A single specimen, 456 mm SL, 4.5 kg, from Enewetak was nontoxic. The stomach and gut contents of four specimens, 278-525 mm SL, from the Society Islands and the Red Sea were examined. Echinoids, including Diadema, Echinometra, and spatangoids, were the main items of diet, but pelecypods, crabs, polychaete tube worms, gastropods, chitons, foraminifera, and algae and detritus were also present. DISCUSSION AND SUMMARY As mentioned in introductory remarks, the ini- tial testing for level of ciguatera at both Enewetak and Bikini revealed only an occasional toxic fish among the species responsible for most cases of this type of poisoning in the Pacific. As expected, the toxic individuals were invariably adults of moderate to large size for the species. A decision was then made to concentrate the fishing effort on the larger individuals of the species most often implicated in ciguatera. These dangerous species are, in general, not common. They are at or near the peak of the well-known "pyramid of numbers," i.e., the reduction in number of individuals one encounters analyzing the populations in succes- sive steps up the food chain. Consequently, much more effort was expended in catching not only these fishes but just the larger individuals of these species. Also, it is for this reason that some rela- tively common species such as Lutjanus fulvus and Adioryx spinifer are represented by few individuals in this report and others such as the smaller species of groupers of the genera Epine- phelus and Cephalopholis were not collected. In highly toxic sectors these species can be poisonous, though even there the incidence is low. A total of 551 specimens of 48 species were tested from Enewetak and 256 specimens of 23 species from Bikini. In addition, 12 adult speci- mens of Lutjanus bohar from Rongelap were tested, one of which was toxic at the 5 level. The results of the testing of fishes from Enewetak are sum- marized in Table 1, and for Bikini in Table 2; 37.3% of the fishes from Enewetak gave a positive reaction for ciguatoxin, and 19.7% of those from Bikini. 244 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS "% Figure 51. — Balistoides viridescens, 456 mm SL, 4.5 kg, Tetiaroa, Society Islands. It must be emphasized that liver and viscera of the suspect fishes were used in the mongoose feed- ing tests (except for sharks) and not flesh. Because of the much higher level of ciguatoxin in the inter- nal organs than in muscle tissue, low-level toxic- ity (indicated by mongoose reactions of 1 or 2) from liver and viscera would probably not result in a detectable level of toxin if flesh from these fishes had been used in the tests. When the percentage of toxicity is computed for the reactions 3-5 (it is this level at which a human eating the flesh of these species might be expected to fall ill with cigua- tera), the percentage of toxic fishes drops to 16.2 for Enewetak and 1.4 for Bikini. When one considers the effort directed almost entirely to the worst offenders in ciguatera, the level of toxicity at Enewetak must be regarded as relatively low and that of Bikini decidedly so. Most of these fishes are avoided as adults by islanders in Oceania regardless of the area of capture. There- fore it is concluded that the returning residents to Enewetak and Bikini need not fear at this time any unexpected threat of ciguatera at their atolls. Only eight species of fishes produced reactions of 4 or 5 in the test animals; that is, severe illness or death: Lycodontis javanicus, Cephalopholis argus, Epinephelus hoedtii, E. microdon, Plec- tropomus leopardus, Aprion virescens, Lutjanus bohar, and Lethrinus kallopterus. Had more specimens of Sphyraena barracuda, Caranx ig- nobilis, and Cheilinus undulatus, particularly of large size (none of the specimens taken during this survey approached the maximum size), been col- lected, then they may be expected to be included in the above list (in view of their reputations for causing ciguatera in other areas). The moray Lycodontis javanicus was clearly the most toxic of all the species tested, with all indi- viduals producing a reaction of 2 or more in mon- gooses and one-third of them the lethal 5. Randall (1958) analyzed the kinds of fishes which have caused ciguatera in terms of habitat, mode of life, and food habits. These species are shore fishes associated with reefs. Usually they are bottom-dwelling generally in < 60 m, but they may be semipelagic open-water forms that range into the reef habitat to feed. They may be car- nivorous or they may feed on benthic algae or detritus. Of the carnivores, those that prey heavily on reef fishes are the most prone to be poisonous. 245 FISHERY BULLETIN: VOL. 78, NO. 2 whereas those that eat mainly benthic crusta- ceans the least. Fishes that feed wholly or primar- ily on plankton are not apt to be toxic. Some mol- lusk and echinoid feeders may cause severe cases of ciguatera. The level of toxicity among benthic herbivores and detritus feeders is consistently low. The food-habit studies of this survey support these generalizations. Seven of the eight most toxic species are piscivorous. The one other, Leth- rinus kallopterus, appears to feed mainly on echinoids and mollusks. No specimens ofLutjanus fulvus,Epinephelus socialis, and Adioryx spinifer were found to be toxic (although relatively few specimens were collected); these feed more on crustaceans than fishes. Among the herbivores tested, only two individuals of Scarus and one of Kyphosus gave a reaction of 1. A water-soluble toxin as well as ciguatoxin were found in the detritus-feeding surgeonfish Ctenochaetus striatus, but in small amounts. The relatively low level of ciguatoxin in sharks is surprising. Because they feed heavily on fishes and are believed to be long-lived, one might expect them to be as ciguatoxic as the larger moray eels. The tropical species of sharks are not as widely eaten as bony fishes. If they were, no doubt more cases of ciguatera would be attributed to them. The species of Carcharhinus appear to prey to a significant degree on pelagic fishes, and when they do feed on reef-dwelling species, they seem to take many plankton-feeding forms. This may in part explain their apparent relatively low level of ciguatoxin. Still another possibility is that sharks may not accumulate as much ciguatoxin in their tissues as bony fishes. Because ciguatera can be highly localized to cer- tain reefs or even sectors of reefs, the fish collect- ing was carried out at many different locations at the atolls. No one area was detected as having a notably higher level of toxicity. Many of the most dangerous ciguatoxic species are roving predators. Examples are the bar- racudas, jacks, dogtooth tuna, emperors, and, to a lesser extent, the snappers. They can be caught at a different area from which they acquired most of their toxicity. The strong localization of ciguatera occurs more where the level of toxicity is high and the smaller more resident species are poisonous. Our fishing has not been sufficiently extensive to demonstrate minor differences in the incidence of ciguatera with locality. At Enewetak most of the fishing was un- dertaken on the southern part of the atoll, par- ticularly in the vicinity of Enewetak Island, the largest of the atoll . This island had the largest pop- ulation of Marshallese people before they were evacuated from the atoll, and it is expected that it will have the largest number when all have been repatriated. Also it is in this area that most of the long-term disturbances of the marine environment, such as the dumping of unwanted material, have taken place. It is fortunate that ciguatera, though more in evidence at Enewetak than at Bikini, is not a major problem as might have been predicted from the impact of western man on the atoll. ACKNOWLEDGMENTS The author and associates acknowledge with gratitude the support of the U.S. Energy Research and Development Administration through con- tract E(26-l)-641 with the University of Hawaii and Bemice P. Bishop Museum, Honolulu. The following individuals participated in the fishing program in the Marshall Islands: Gerald S. Akiyama, Bruce A. Carlson, the late David Erlen- kotter. Glen H. Fredholm, James W. Fry, Gregory Gahagan, Gerald Gulden, Walter C. Gutjahr, Guy S. Haywood, George MacGuire, Oliver K. McCaus- land, Rhett M. McNair, Robert F. Meyers, Takeo Okamura, John E. Randall, Robert P. H. Ruther- ford, Arnold Y. Suzumoto, and Gordon W. Tribble. In addition, Phillip B. Lamberson, formerly of the Mid-Pacific Marine Laboratory, Harry J. Miller and Russell E. Miller, former resident personnel of the atoll, assisted in the collecting at Enewetak. The testing of the toxicity was carried out by James Murphy and Lambert Yamashita at the Hawaii Institute of Marine Biology of the Univer- sity of Hawaii, under the direction of A. H. Ban- ner, except for the last sampling in May 1978 for which an initial screening was made by Yoshi- tsugi Hokama of the Department of Pathology, University of Hawaii, by radioimmunoassay (Hokama et al. 1977). The higher reactions were then confirmed by mongoose feeding at the Ber- nice P. Bishop Museum by Arnold Y. Suzumoto. A. H. Banner and Helen A. Randall provided helpful suggestions and reviewed the manuscript. LITERATURE CITED ADACHI, R., AND Y. FUKUYO. 1979. The thecal structure of a marine toxic dinoflagellate 246 RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS Gambierdiscus toxicus gen. et sp. nov. collected in a ciguatera-endemic area. Bull. Jpn. Soc. Sci. Fish. 45:67-71. AL-HUSSAINI, A. H. 1947. The feeding habits and the morphology of the alimentary tract of some teleosts living in the neighbour- hood of the Marine Biological Station, Ghardaqa, Red Sea. Publ. Mar. Biol. Stn. Ghardaqa (Red Sea) 5:1-61. BaGNIS, R. 1968. Clinical aspects of ciguatera (fish poisoning) in French Polynesia. Hawaii Med. J. 28:25-28. 1969. Naissance et developpement d'une flambee de ciguatera dans un atoll des Tuamotu. Rev. Corps Sante 10(6):783-795. BAGNIS, R., J. Bennett, V. Joutain, and F. Nanai. 1968. Ciguatera in Tahiti: an endemic disease. South Pac. Comm. Sem. Ichthyosarcotoxism, Fr. Polynesia, 15 p. Bagnis, R., p. Mazellier, J. Bennett, and E. Christian. 1972. Poissons de Polynesie. Les Editions du Pacifique, Papeete, 368 p. Banner, A. H. 1965. Ciguatera fish poisoning: a symposium. Ciguatera in the Pacific. Hawaii Med. J. 24:353-361. Banner, a. H., and p. Helfrich. 1964. The distribution of ciguatera in the tropical Pa- cific. Hawaii Mar. Lab. Tech. Rep. 3, 48 p. Banner, a. H., P. Helfrich, and T. Piyakarnchana. 1966. Retention of ciguatera toxin by the red snapper, Lutjanus bohnr. Copeia 1966:297-301. Banner, a. H., P. J. Scheuer, S. Sasaki, P. Helfrich, and C. B. Alender. 1960. Observations on ciguatera- type toxin in fish. Ann. N.Y. Acad. Sci. 90:770-787. Bartsch, a. F., R. H. drachman, and E. F. McFarren. 1959. Report of a survey of the fish poisoning problem in the Marshall Islands. U.S. Dep. Health, Educ, Welfare, Public Health Serv., 117 p. B.ASS, A. J., J. D. D'Aubrey, and N. KISTNASAMY. 1973. Sharks of the east coast of southern Africa. I. The genus Carcharhinus (Carcharhinidael. Oceanogr. Res. Inst. (Durban) Invest. Rep. 33, 168 p. 1975a. Sharks of the east coast of southern Africa. III. The families Carcharhinidae (excluding Mustelus and Car- charhinus) and Sphymidae. Oceanogr. Res. Inst. (Dur- ban) Invest. Rep. 38, 100 p. 1975b. Sharks of the east coast of southern Africa. IV. The families Odontaspididae, Scapanorhynchidae, Isuridae, Cetorhinidae, Alopiidae, Orectolobidae and Rhiniodon- tidae. Oceanogr. Res. Inst. (Durban) Invest. Rep. 39, 102 p. Becke, L. 1901. Poisonous fish of the Pacific Islands. In L. Becke, Yorke the adventurer, p. 246-254, J. B. Lippincott Col., Lond. Belotte, J. 1955. Les Poissons "empoisormes" du lagon Tahitien. Med. Trop. 15:232-236. BIGELOW, H. B., AND W. C. SCHROEDER. 1948. Sharks, /n A. E. Parr (editor), Fishes of the west- em North Atlantic, Part one, p. 59-546. Mem. Sears Foimd. Mar. Res., Yale Univ. 1. BOULENGER, G. A. 1895. Catalogue of the perciform fishes in the British Museum. 2d ed., Vol. 1. Br. Mus. (Nat. Hist.), Lond., 394 p. BROCK, R. E. 1972. A contribution to the biology of Gymnothorax javanicus (Bleeker). M.S. Thesis, Univ. Hawaii, 121 p. BROCK, V. E., R. S. JONES, AND P. HELFRICH. 1965. An ecological reconnaissance of Johnston Island and the effects of dredging. Hawaii Mar. Lab. Tech. Rep. 5, 90 p. CHOAT, J. H. 1966. Parrot fish. Aust. Nat. Hist. 15:265-268. CLARK, E., AND H. A. F. GOHAR. 1953. The fishes of the Red Sea: Order Plectog- nathi. Publ. Mar. Biol. Stn. Al Ghardaqa (Red Sea) 8, 80 p. Clark, e., and K. Von Schmidt. 1965. Sharks of the central Gulf coast of Florida. Bull. Mar. Sci. 15:13-83. COMPAGNO, L. J. V. 1973. Gogolia filewoodi, a new genus and species of shark from New Guinea (Carcharhiniformes: Triakidae), with a redefinition of the family Triakidae and a key to triakid genera. Proc. Calif. Acad. Sci., Ser. 4, 39:383-410. Cooper, M. J. 1964. Ciguatera and other marine poisoning in the Gilbert Islands. Pac. Sci. 18:411-440. DE Sylva, D. p. 1963. Systematics and life history of the great barracuda Sphyraena barracuda (Walbaum). Stud. Trop. Oceanogr. (Miami) 1, 179 p. 1973. Barracudas (Pisces: Sphyraenidae) of the Indian Ocean and adjacent seas - a preliminary review of their systematics and ecology. J. Mar. Biol. Assoc. India 15:74-94. FOURMANOIR, P. 1961. Requins de la cote ouest de Madagascar. Mem. Inst. Sci. Madagascar, ser. F, 4:1-81. Garrick, J. A. F. 1967. A broad overview of Carcharhinus species, their sys- tematics and distribution. In P. W. Gilbert, R. F. Mathewson, and D. R. Rail (editors), Sharks, skates, and rays, p. 85-91. Johns Hopkins Press, Baltimore. In press. Sharks of the genus Carcharhinus. U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circ. Glynn, P. W., R. H. Stewart, and J. E. McCosker. 1972. Pacific coral reefs of Panama: structure, distribution and predators. Geol. Rundsch. 61:483-519. GOHAR, H. A. F., AND F. M. MAZHAR. 1964. The elasmobranchs of the north-western Red Sea. Publ. Mar. Biol. Stn. Al-Ghardaqa (Red Sea) 13:1- 144. Halstead, b. w. 1967. Poisonous and venomous marine animals of the world. Vol. 2. U.S. Gov. Print. Off., Wash., D.C., 1070 p. Halstead, b. W., and W. M. Lively, Jr. 1954. Poisonous fishes and ichthyosarcotoxism. Their re- lationship to the Armed Forces. U.S. Armed Forces Med. J. 5:157-175. Halstead, B. W., and D. W. Schall. 1958. A report on the poisonous fishes of the Line Islands. Acta Trop. 15:193-233. HARMELIN-VIMEN, M. L., .\ND C. BOUCHON. 1976. Feeding behavior of some carnivorous fishes (Ser- ranidae and Scorpaenidae) from Tulear (Madagas- car). Mar. Biol. (Beri.) 37:329-340. Helfrich, P., and P. A. Allen. 1975 . Observations on the spawning of mullet, Crenimugil 247 FISHERY BULLETIN: VOL. 78. NO. 2 crenilabis (Forskal), at Enewetak, Marshall Islands. Micronesica 11:219-225. HELFRICH, P., AND A. H. BANNER. 1963. Experimental induction of ciguatera toxicity in fish through diet. Nature (Lond.) 197:1025-1026. 1968. Ciguateric fish poisoning. II. General patterns of development in the Pacific. Occas. Pap. Bemice P. Bishop Mus. 23:371-382. HELFRICH, P., T. PIYAK.ARNCHANA, AND P. S. MILES. 1968. Ciguatera fish poisoning. 1. The ecology of ciguateric reef fishes in the Line Islands. Occas. Pap. Bemice P. Bishop Mus. 23:305-369. HIATT, R. W., AND V. E. BROCK. 1948. On the herding of prey and the schooling of the black skipjack, Euthynnus yaito Kishinouye. Pac. Sci. 2:297- 298. HiATT, R. W., AND D. W. STRASBURG. 1960. Ecological relationships of the fish fauna on coral reefs of the Marshall Islands. Ecol. Monogr. 30:65-127. HISHIKARI, J. 1921. On poisonous fishes in South Sea Islands. [In Jpn.] Nippon Beseibut. Zasshi 16:1533-1543. HIYAMA, Y. 1943. Report of an investigation of poisonous fishes of South Seas. [In Jpn.] Nissan Fish. Exp. Stn. Odawara Br., 137 p. (Translated by W. G. Van Campen, 1950, Poisonous fishes of the South Seas, p. 1-188. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 25.) HOBSON, E. S. 1974. Feeding relationships of teleostean fishes on coral reefs in Kona, Hawaii. Fish. Bull., U.S. 72:915-1031. HOKAMA, Y., A. H. BANNER, AND D. B. BOYLAN. 1977. A radioimmunoassay for the detection of cigua- toxin. Toxicon 15:317-325. Hydrographic Office, U.S. Navy. 1945. Sailing directions for the Pacific Islands. Vol. I. Western groups, ed. 5. H. O. Publ. 165, 775 p. Johnson, r. h., and D. R. Nelson. 1973. Agonistic display in the gray reef shark, Car- charhinus menisorrah, and its relationship to attacks on man. Copeia 1973:76-84. Jones, R. S. 1968. Ecological relationships in Hawaiian and Johnston Island Acanthuridae (surgeonfishes). Micronesica 4:309-361. K ATA YAM A, M. 1960. Fauna Japonica. Serranidae (Pisces). Tokyo News Serv., Tokyo, 189 p. Kato, S., S. Springer, and M. H. Wagner. 1967. Field guide to eastern Pacific and Hawaiian sharks. U.S. Fish Wildl. Serv., Circ. 271, 47 p. KHLENTZOS, C. T. 1950. Seventeen cases of poisoning due to ingestion of an eel, Gymnothorax flavimarginatus. Am. J. Trop. Med. 30:785-793. Marshall, T. C. 1964. Fishes of the Great Barrier Reef and coast waters of Queensland. Angus and Robertson, Sydney, 566 p. MASUDA, H., C. ARAGA, AND T. YOSHINO. 1975. Coastal fishes of Southern Japan. Tokai Univ. Press, Tokyo, 379 p. Matsuo, R. 1934. Investigation of the poisonous fishes of Jaluit Is- land. [In Jpn.] Nanyo Gunto Chihobyo Chosa Ronbun- shu 2:309-326. 248 Randall, J. E. 1955a. Fishes ofthe Gilbert Islands. Atoll Res. Bull. 47, 243 p. 1955b. A revision of the surgeon fish genus Ctenochaetus, family Acanthuridae, with descriptions of five new species. Zoologica (N.Y.) 40:149-166. 1956. A revision of the surgeon fish genus Acan- thurus. Pac. Sci. 10:159-235. 1958. A review of ciguatera, tropical fish poisoning, with a tentative explanation of its cause. Bull. Mar. Sci. Gulf Caribb. 8:236-267. 1963. A fatal attack by the shark Carcharhinus galapagensis at St. Thomas, Virgin Islands. Caribb. J. Sci. 3(4):201-205. 1964. Notes on the groupers of Tahiti, with description of a new serranid fish genus. Pac. Sci. 18:281-296. 1967. Food habits of reef fishes of the West Indies. Stud. Trop. Oceanogr. (Miami) 5:665-847. 1969. How dangerous is the moray eel? Aust. Nat. Hist. 16(6):177-182. 1972. Tahitian fish names and a preliminary checklist of the fishes ofthe Society Islands. Occas. Pap. Bemice P. Bishop Mus. 24:167-214. 1974. The effect of fishes on coral reefs. Proc. 2d Int. Coral Reef. Symp. 1:159-166. 1977. Contribution to the biology ofthe whitetip reef shark (Triaenodon obesus). Pac. Sci. 31:143-164. Randall, J. E., and V. E. Brock. I960. Observations on the ecology of epinepheline and lut- janid fishes ofthe Society Islands, with emphasis on food habits. Trans. Am. Fish. Soc. 89:9-16. RANDALL, J. E., S. M. HEAD, AND A. P. L. SANDERS. 1978. Food habits of the giant humphead wrasse, Cheilinus undulatus (Labridae). Environ. Biol. Fish. 3(2):235-238. RANDALL, J. E., AND G. S. HELFMAN. 1973. Attacks on humans by the blacktip reef shark (Car- charhinus melanopterus) . Pac. Sci. 27:226-238. R0BS0N,R.W.( editor). 1959. Pacific Islands Year Book, 8th ed. Pac. Publ., Sydney, 479 p. ROSENBLATT, R. H., AND E. S. HOBSON. 1969. Parrotfishes (Scaridae) ofthe eastern Pacific, with a generic rearrangement of the Scaridae. Copeia 1969:434-453. ROSS, S. G. 1947. Preliminary report on fish poisoning at Fanning Is- land (Central Pacific). Med. J. Aust. 34(2):617-621. Sato, T. 1978. A synopsis ofthe sparoid fish genus Lethrinus, wdth the description of a new species. Univ. Mus., Univ. To- kyo, Bull. 15, 70 p. SCHULTZ, L. P. 1949. At grips with a giant moray eel. Nat. Hist. 58:42- 43. 1958. Review of the parrotfishes family Scaridae. U.S. Natl.Mus. Bull.214, 143p. 1969. The taxonomic status of the controversial genera and species of parrotfishes with a descriptive list (family Scaridae). Smithson. Contrib. Zool. 17, 49 p. SCHULTZ, L. P., AND COLLABORATORS. 1953-66. Fishes of the Marshall and Marianas Islands. U.S. Natl. Mus. Bull. 202, 3 vol. Smith, j. l. B. 1956. The parrotfishes of the family Callyodontidae of the RANDALL: SURVEY OF CIGUATERA AT MARSHALL ISLANDS western Indian Ocean. Rhodes Univ., Grahamstown, Ichthyol. Bull. 1:1-23. 1959. The identity of Scarus gibbus Ruppell, 1828 and of other parrotfishes of the family Callyodontidae from the Red Sea and the western Indian Ocean. Rhodes Univ., Grahamstown, Ichthyol. Bull. 16:265-282. SMITH, J. L. B., AND M. M. Smith. 1963. The fishes of Seychelles. Dep. Ichthyol., Rhodes Univ., Grahamstown, 215 p. STEINBACH, E. 1895. Bericht liber die Gesundheitsverhaltnisse der Eingeborenen der Marshallinseln im Jahre 1893/94 und Bemerhung liber Fischgift. Mitt. Forschung. Gelehrt dtsch. Schutzgeb. 8(2):157-171. Stephens, W. M. 1963. Stay away from the moray. Sea Front. 9:152-159. Talbot, F. H. 1960. Noteson the biology of the Lutjanidae (Pisces) of the East African coast, with special reference to L. bohar (Forskal). Ann. S. Afr. Mus. 45:549-573. Taylor, F. J. R. 1979. A description of the benthic dinoflagellate associated with ciguatoxin, including observations on Hawaiian material. In D. L. Taylor and H. H. Seliger (editors). Toxic dinoglagellate blooms, p. 71-76. Elsevier, North Holland. wass, R. C. 1971 . A comparative study of the life history, distribution, and ecology of the sandbar shark and the gray reef shark in Hawaii. Ph.D. Thesis, Univ. Hawaii, 219 p. Wheeler, J. F. G., and F. D. Ommanney. 1953. Report on the Mauritius-Seychelles Fisheries Sur- vey 1948-1949. Colon. Off. Fish. Publ. 1(3):1-145. Wood-Jones, F. 1910. Coral and atolls. Reeve and Co., Lond., 392 p. Yasumoto, T. , Y. Hashimoto, R. Bagnis, J. E. Randall, and A. H. Banner. 1971. Toxicity of the surgeonfishes. Bull. Jpn. Soc. Sci. Fish. 37:724-734. Yasumoto, T., I. Nakajima, R. Bagnis, and R. Adachi. 1977. Finding of a dinoflagellate as a likely culprit of ciguatera. Bull. Jpn. Soc. Sci. Fish. 43:1021-1026. Yasumoto, T., I. Nakajima, Y. Oshima, and R. Bagnis. 1979. A new toxic dinoflagellate found in association with ciguatera. In D. L. Taylor and H. H. Seliger (editors). Toxic dinoflagellate blooms, p. 65-70. Elsevier, North Holland. Yudkin,W. H. 1944. Tetrodon poisoning. Bull. Bingham Oceanogr. Col- lect. Yale Univ. 9(1), 18 p. 249 CALLINECTES (DECAPODA: PORTUNIDAE) LARVAE IN THE MIDDLE ATLANTIC BIGHT, 1975-77^ Peter O. Smyth^ ABSTRACT Distribution of Callinectes larvae in surface (neuston) and subsurface shelf waters in the Middle Atlantic Bight was determined from quarterly zooplankton collections taken during a 2-year study. Observations confirmed the presence in offshore waters of a large larval pool from which recruitment may take place. Larvae were predominantly late zoeae and megalope, with peak abundances in late summer collections reaching 16,000 per 100 m^ in neuston collections. During summer, crab larvae were distributed across the shelf with the majority at 10-80 km offshore. Abundances were sig- nificantly greater in neuston than subsurface zooplankton collections and generally greater in neuston collections taken at night. Water temperature and distance from shore were factors most closely correlated with abundance of larvae in the neuston. Megalopae of Callinectes were present at outer shelf stations in winter and spring and together with megalopae oi Portunus and other forms were of southern origin. Based on experimentally determined temperature-salinity preferences reported in the literature for Callinectes larvae, metamorphosis may be delayed in cooler offshore waters, thus increasing chances of long-range transport. The community of organisms of the surface layer (the neuston^ ) has received increasing attention in terms of samphng problems and possible ecologi- cal significance. Zaitsev ( 1970 ) described the neus- ton as consisting chiefly of early developmental stages of fishes and invertebrates. Berkowitz (1976) and Morris (1975), however, found oceanic neuston faunistically impoverished in comparison with zooplankton of the immediate subsurface. Few studies of the neuston of shelf and shallow waters exist; preliminary indications are that the zooplankton of the surface waters of the continen- tal shelf are at least quantitatively enriched (Grant^). Callinectes , euryhaline members of the predom- inantly marine Portunidae, spawn along the shore of open oceans and in mouths of inlets and es- tuaries. Larval development occurs in shelf wa- 'Contribution No. 952 from the Virginia Institute of Marine Science. From part of a dissertation to be submitted in partial fulfillment of requirements for the degree of Doctor of Philosophy, College of William and Mary. ^Virginia Institute of Marine Science and School of Marine Science, the College of William and Mary, Gloucester Point, VA 23062. 'Neuston has generally been defined operationally as the community of organisms sampled by gear specifically designed to sample the surface layer. The term is used in that sense in this paper. For a review of numerous terms associated with the sur- face layers, see Banse (1975). "Grant, G. C. 1977. Middle Atlantic Bight zooplankton: seasonal bongo and neuston collections along a transect off southern New Jersey. Spec. Rep. Appl. Mar. Sci. Ocean Eng., Va. Inst. Mar. Sci. 173, 138 p. Manuscript accepted November 1979. FISHERY BULLETIN; VOL. 78, NO. 2, 1980. ters, with probable return inshore by megalopae andjuveniles (Williams 1965, 1971, 1974; Costlow 1967; Tagatz 1968). Callinectes megalopae have been reported offshore in shelf waters (Nichols and Keney 1963; Dudley and Judy 1971); retention in shelf waters and subsequent transport of megalopae have been proposed as mechanisms in dispersal, widespread distribution, and mainte- nance of genetic continuity in the species (Costlow 1967; Williams 1971, 1974; Cole and Morgan 1978). Callinectes larvae, at least zoeae, have surface affinities (Tagatz 1968; Dudley and Judy 1971; Sandifer 1972), but megalopae have generally been less numerous in collections than zoeae and limited to bottom samples (Tagatz 1968; Sandifer 1972; Goy 1976). Williams (1971), however, re- ported Callinectes megalopae to be active in es- tuarine surface waters at night. With the widespread distribution and known abundance of Callinectes adults and the accepted migratory sequence of developmental stages (inshore-offshore-inshore), the reported abun- dance of late stage larvae is surprisingly low. Furthermore, the existence in shelf waters of a Callinectes larval pool from which recruitment to estuaries may occur is based on relatively few studies and limited sampling. This paper reports the identification, distribu- tion, and abundance oi Callinectes larvae in neus- 251- FISHERY BULLETIN: VOL. 78, NO. 2 ton and subsurface water column collections from shelf waters in the Middle Atlantic Bight. My ob- jectives specifically were to: 1) determine whether a reservoir of Callinectes larvae, particularly megalopae, exists in shelf waters; 2) determine abundance relationships between Callinectes lar- vae in neuston and water column samples; 3) examine the role of certain environmental factors (e.g., temperature, salinity, location) in the dis- tribution and abundance of these larvae; 4) assess the role of Callinectes megalopae in larval re- cruitment and dispersal in view of my findings and results of laboratory studies of temperature- salinity tolerances of larvae; and 5) examine in- teraction of the developmental migratory se- quence, biogeography, and evolutionary history of Callinectes. METHODS Zooplankton collections were made as part of a 2-yr survey (Table 1) conducted by the Virginia Institute of Marine Science (VIMS) for the Bureau of Land Management (1975-77). This study was designed to provide ecological information prior to drilling for oil on the Middle Atlantic Bight conti- nental shelf. In addition to zooplankton studies the survey included studies of benthic and epiben- thic communities and the physical, chemical, and geographical oceanography of the shelf and over- lying waters. During the first year, six stations were occupied seasonally (quarterly) on a transect across the shelf off Atlantic City, N.J. (Figure 1; Table 2: CI, Dl, N3, E3, F2, Jl). Zooplankton in the water column was sampled at night by paired, double oblique tows with 60 cm diameter, opening-closing bongo nets (McGowan and Brown^) (505 fim and 202 ixm mesh). Bongo nets were metered (General Oceanics, Inc. flowmeters^) and were closed during passage through the surface layer. Neuston was sampled every 3 h over a 24-h period with a neus- ton net designed at Woods Hole Oceanographic Institution. This sampler consisted of two hydrodynamically-shaped, foam-filled floats con- nected by an endless fiber glass band (Grant'). The Table l. — Dates for cruises in the Middle Atlantic Bight, 1975- 77, over which Callinectes larvae were sampled. *McGowan, J. A., and D. M. Brown. 1966. A new opening- closing paired zooplankton net. Univ. Calif., Scripps Inst. Oceanogr. Ref. 66-23, 56 p. ^Reference to trade names dpes not imply endorsement by the National Marine Fisheries Service, NOAA. 'Grant, G. C. 1979. Middle Atlantic Bight zoo- plankton. Spec. Rep. Appl. Mar. Sci. Ocean Eng., Va. Inst. Mar. Sci. 192, 236 p. 252 First year Second year Season Cruise Date Cruise Date Fall Winter spring Summer 01W 02W 03W 04W 23-30 Oct. 1975 5-16 Feb. 1976 8-16 June 1976 1-9 Sept. 1976 05W 06W 07W 08W 5-28 Nov. 1976' 20Feb.-6 Mar. 1977 18-28 May 1977 19-29 Aug. 1977 'Cruise split into two legs. \ > )>r CI • / .['2000 ^^^^\ \ B5>\ f:^ N J |\ % \ [Iff ^ T CI \ • /^ • .-■ E3 t \ • '■ ^ DEL. Js\ \ -. / ^ M D bi LI • i ■^ \ G) ^•A^ i '^p^ ■■> i-\ Figure l. — Study area and sampling stations for surface and subsurface zooplankton in the Middle Atlantic Bight, 1975-77. Stations LI, L2, L4, L6, B5, A2 were sampled only during the second year of the study; CI, Dl, N3, E3, F2, Jl were sampled both years. mouth of the net was 1.0 m wide, and in calm water the gear sampled approximately the upper 12 cm of the water column. However, the net appeared to sample, on the average, less than the upper 12 cm due to sea conditions and towing characteristics of the ship and sampler. Calculated volumes were based on a 12 cm sample depth and were thus overestimated, resulting in underestimation of SMYTH: CALLINECTES LARVAE IN MIDDLE ATLANTIC BIGHT Table 2. — Station data forzooplankton collections in the Middle Atlantic Bight, 1975-77. Location Distance from Bottom Station Lat. N Long. W shore (km) depth (m) A2 39°21.8' 72=31.8' 149 131 B5 39^28. 3' 73°02.r 93.6 62.6 CI 39=22.2' 74°14.9' 10.2 16.8 01 3904.7- 73°53,2' 56.5 37.2 N3 38''51.4' 73°44.8' 83.4 44.7 E3 38 41.2' 73°32.5' 112 59.5 F2 38"44.4' 73°09.2' 132 108 J1 38 44.2' 73°00.7' 141 355 L1 37°31.1' 75°18.3' 31.5 22.3 12 3r'20.r 74°58.6' 65.8 41.3 L4 37°08.1' 74°36.8' 105 94.6 L6 3r04.4' 74°33.1' 113 322 larval densities. Tows were of 20-min duration except when large abundances of neuston required premature termination of a tow. The net was me- tered beginning with the June 1976 cruise; before that cruise, sample volumes were estimated on the basis of a standard 20-min tow. Tows were made from an extended boom alongside the ship at speeds of 1.5-2.5 kn. During the second year two stations to the north and a transect to the south of the original transect were added. On each cruise neuston samples were taken over 24 h at nine stations (Figure 1; Table 2: A2, B5, CI, E3, Jl, LI, L2, L4, L6). A single neus- ton tow was made at stations Dl, N3, and F2 as a companion to bongo tows. Bongo tows were made at all 12 stations following the procedure used during the first year. In addition, replicate tows were made at stations A2, B5, and E3 (repeated tows of two bongo nets with paired 202 yum and 505 yum mesh nets). Three such replicate tows were made at night at each designated station. Samples were preserved in a 4% solution of borax-buffered formaldehyde and seawater (Steedman 1976). In the laboratory, major taxonomic groups were quantitatively sorted from whole or split samples (Burrell et al. 1974). Deca- pods were sorted to species and identified (when possible ) on the basis of published descriptions and taxonomic keys. Megalopae of several taxa, including Cal- linectes , were reared aboard ship to juveniles. Sev- eral megalopae were removed from a sample and tentatively identified or identification characters noted. Megalopae were placed in plastic tackle boxes with 505 ixm mesh bottoms and the boxes were floated in an aquarium filled with seawater taken in situ. Megalopae were fed Artemia salina nauplii and bits of fresh fish meat. Megalopae with the same characteristics as the megalopae used for rearing were fixed and preserved. Abundance was expressed as number per 100 m^; for graphical presentation and certain statis- tical procedures abundance was compressed by the transformation logio(Z + 1). Most statistical pro- cedures were based on station means, with eight neuston collections per station. The distribution of sample means tends to normality as the sample size increases (Snedecor and Cochran 1967), and the logarithmic transformation tends to make var- iance independent of the mean (Sokal and Rohlf 1969). Based on the i^-max test (Sokal and Rohlf 1969), untransformed abundances within stations were very heteroscedastic, while log-transformed abundances at stations with Callinectes larvae in at least six samples did not have unequal var- iances at P <0.05. Coefficients of variation for each station were reduced considerably by the log transformation, and abundances appeared better centered about the median based on "box and whisker" diagrams (Tukey 1977). The assumption of a multivariate normal distribution could not be tested for the data set. Significance levels for mulitvariate data are often difficult to interpret; therefore, significance levels, where indicated for parametric procedures, should be taken as a guide. A larval stage index (LSI) similar to that of Manzi and Maddox ( 1976) was calculated for sev- eral larval types. The LSI is a point along the continuum of development from hatching (first zoea) to juvenile; the LSI characterizes the stage of an average individual of a given species in a sam- ple. It is calculated as a weighted average, i.e., LSI = 2 iSi f=i where i n number of the developmental stage, number of developmental stages, first zoeae through adult, abundance of the ith stage. The LSI is standardized and constrained in the interval 0.0-1.0 by the assignment of a stage number (1> the megalopa) to the adult stage. Thus, an LSI = 0.67 characterizes animals that have completed, on the average, about two-thirds of the developmental sequence from hatching to first crab. The LSI is, however, a measure of cen- tral tendency and does not indicate statistical dis- 253 FISHERY BULLETIN: VOL. 78, NO. 2 persion. Based on Costlow and Bookhout ( 1959), n was set at 10 for eight zoeal stages, a megalopa, and an adult. Comparisons between Callinectes abundance in neuston and bongo (surface vs. subsurface) collec- tions at each station were made for: 1) maximum abundance for each gear type; 2) mean abundance of the consecutive pair of tows with the largest collective abundance; and 3) mean abundance for each gear type. Significance of differences for these means was determined by the Wilcoxon signed rank test (Wilcoxon 1945), a distribution- free method (Hollander and Wolfe 1973). Comparisons between neuston and bongo collec- tions are comparisons between abundances in a single "layer" and abundances integrated over the water column. Therefore, abundances in bongo collections represent mean abundances in the water column (excepting the surface) and do not indicate vertical distribution of the animals. Diel patterns in neuston abundance during each cruise were represented by total numbers per 100 m^ for each sampling time interval (3 h) summed over the stations in a cruise. To weight frequency as well as abundance during a single cruise, ranks were assigned to abundance during each time in- terval (lowest to highest) at each station. The rank sum of each time interval was calculated as the sum of the ranks during that time interval over all stations during a single cruise. For neuston collections the relationship be- tween mean abundance per station and environ- mental factors (temperature, salinity, station depth, and distance from shore) was examined. Data were analyzed using subprograms ( multiple) Regression and Partial Corr (partial correlation) of the Statistical Package for the Social Sciences (SPSS, Nie et al. 1975). Relationships between abundance and factors were examined in terms of bivariate as well as multivariate distributions. RESULTS Identification Callinectes zoeae were identified and staged on the basis of Sandifer's ( 1972) key and descriptions of laboratory -reared zoeae of C. sapidus (Costlow and Bookhout 1959) and C. similis (Bookhout and Costlow 1977). Key characters include: 1) relative length of the antennal exopodite (<'/3 protopodite length) and the presence of two unequal terminal setae on the antennal exopodite; 2) the presence of lateral projections on abdominal somites 2 and 3; 3) the presence of relatively long, sharply pointed posterolateral spines on abdominal somites 3-5; and 4) the presence of one dorsal and one lateral spine in each telson furca. Structure and setation of mouthparts and appendages were compared with published descriptions for further confirma- tion. The above characters effectively separated Callinectes zoeae from all other zoeal types in my collections. The planktonic material appeared to include seven or eight distinct zoeal stages after allowance for individual variation in certain structures, setal counts, relative lengths, etc. (e.g., the antennal endopodite "bud" denoting stage 5, which varied from little more than a swelling to a definite projection). Identification of Portunidae megalopae was based on Kurata's (1975) list of familial and sub- familial (Portuninae) characters, which include the presence of sternal cornua (paired spines pro- jecting posteriorly from the fourth sternal seg- ment beyond the base of the fifth leg) (Figure 2), and the presence of paddlelike dactyls with long, hooked setae on the fifth pereopods. Callinectes and Portunus megalopae were sepa- rated on the basis of the characters listed by Bookhout and Costlow (1974), which include the absence in Callinectes and the presence in Por- tunus of a ventral spine on the coxa of the second pereopod ( Figure 2), and carpal spine(s) on the first pereopod. My collections included numerous megalopae attributable to Portunus; all had a coxal spine on the second pereopod and a carpal spine on the first pereopod. The basischiopodite hook reported for Abdominal somife Portunus Coxal / Spine Sternal cornua Callinectes Figure 2. — Lateral profile including the abdomen ofCalUnectes and Portunus megalopae. Distinguishing characters are indi- cated. Sizes are not relative. 254 SMYTH: CALLINECTES LARVAE IN MIDDLE ATLANTIC BIGHT Portunus (Bookhout and Costlow 1974; Kurata 1975) and Callinectes (Costlow and Bookhout 1959; Bookhout and Costlow 1977) was present for all Callinectes and most, but not all, Portunus specimens. The profile of abdominal somites was a more subjective, yet reliable, criterion for the separa- tion of Portunus and Callinectes megalopae. In Portunus the dorsal surface of each somite, par- ticularly the first, was noticeably raised, creating a "bumpy" profile; in Callinectes the profile was noticeably smoother (Figure 2). (See also Book- hout and Costlow 1974, fig. 11; 1977, fig. 11.) Al- though identifications were based on numerous characters, this particular criterion was consis- tent and reliable. Carapace lengths in = 418, X = 1.56, SE = 0.004 mm) of Callinectes megalopae (measured dorsally from the base of the rostrum to the pos- terior edge of the carapace) were slightly less than carapace lengths reported for C. sapidus (X = 1.65 mm) but considerably greater than lengths re- ported for laboratory-reared C. similis (X = 1.30 mm) (Bookhout and Costlow 1977). I recognized no specific differences among Cal- linectes megalopae or zoeae; therefore, larvae re- ferred to as Callinectes may represent more than one species. Abundance and known distribution of adults (Williams 1974) indicated that most, if not all, specimens were C. sapidus. Several small adult C. similis, however, were taken in neuston collections at station Cl in late October 1975. The above characteristics used to separate Cal- linectes and Portunus megalopae were confirmed by specimens reared to the juvenile stage. Por- tunus juveniles were too small ( <10 mm carapace width) for specific identification. One Callinectes megalopa developed to a juvenile stage tentatively identified as C sapidus. Distribution Callinectes larvae were collected on six of eight cruises and were most abundant in late summer (Figure 3). Mean abundance in neuston collections (n = 8) at a single station reached 3,100/100 m^ at L2 in August 1977; at this station abundance in a single neuston collection reached a peak of 16,000/100 m^. Abundance generally decreased offshore of the 50 m isobath during the summer-fall cruises (sta- tion Jl in August 1977 was an exception). During the second year, with additional stations, larvae were generally more abundant at stations on the most southern transect than at more northerly stations. Peak abundance often coincided with de- pressed LSI's inshore, evidence that reproductive activity inshore closely preceded the sampling periods. Except during the summer, larval popula- tions consisted almost exclusively of late zoeae and megalopae, particularly in central and outer shelf waters. Collections of Callinectes during the fall of 1975 and winter and spring of 1977 com- prised only megalopae. Mean and maximum abundance (Figure 3; Table 3) was greater in neuston than in bongo collections except at three stations during summer 1977 (Figure 3). During winter 1977, occurrences were too few to be tested at the 0.05 confidence level by the signed rank test. On all other cruises during which Callinectes larvae occurred, abun- dance was significantly greater in surface than subsurface collections (Table 3). Diel patterns in neuston abundance of Cal- linectes were not consistent over all cruises (Fig- ure 4). A dawn peak in abundance was evident in summer 1976. Dusk peaks appeared in fall 1975 and possibly spring 1977. Total abundance was greatest during darkness (between sunset and Table 3. — Comparison of surface and subsurface (neuston vs. bongo) abundance of Callinectes larvae, based on the signed rank test (Wilcoxon 1945). N denotes greater abundance in neuston, significance level indicated by asterisks (* = 0.05, ** = 0.01); P is the probability of a rank sum equal or greater under the null hypothesis of equal surface and subsurface abundance; fraction in parentheses: numerator is the number of occurrences in particular abundance category and denominator is the number of possible occurrences in abundance category. Season and year of collection Comparison Fain 975 Summer 1976 Fall 1976 Winter 1977 Spring 1977 Summer 1977 Maximum neuston vs. N-P = 0.016 N-P = 0.016 N"P = 0.004 N P = 0.062 N-P = 0.016 N"P = 0.008 maximum bongo (6/6) (6/6) (8/12) (4/12) (6/12) (12/12) Maximum consecutive pair of neuston tows vs. N-P = 0.016 N'P = 0.016 N"P = 0.008 NP = 0.125 N-P =0.016 N"P= 0.004 bongos, mean (6/6) (6/6) (7/9) (3/9) (6/9) (9/9) '-euston vs. N-P = 0.016 N-P =0.016 N" P = 0.008 NP = 0.125 N-P = 0.016 N"P = 00.10 bongos, mean (6/6) (6/6) (7/9) (3/9) (6/9) (9/9) 255 FISHERY BULLETIN: VOL. 78, NO. 2 3 2 0- I 30 5 17 A 23-30 OCT 1975 33 6 32 3 32 8 163 16 9 32 8 166 34 7 20 2 I 00 80 60 40 20 * LSI 30.2 Solinily (%.) 19.2 Temperolure(°C) r-; MAXIMUM ABUNDANCE •MEAN ABUNDANCE i Dl N3 E3 F2 Jl + X o o _i E O o q: UJ Q. 40 tr - 20 I 00 80 60 40 20 257 FISHERY BULLETIN; VOL. 78, NO. 2 4 3.0- 2 23-30 OCT 1975 T — r*-i — I — \ — ! — I — I — H — I — r 1-9 SEPT 1976 r 50 40 - 30 20 - 10 C 3.0 o o E O O q: LU Q. Oi OH \li CD 2.0- 1.0- 4.0 30- 2 I 0- 5-28 NOV 1976 fL ~\ 1^ I I I I I I i 1 r 18- 28 MAY 1977 RANK SUM AT TIME INTERVAL ABUNDANCE AT TIME INTERVAL 20 FEB - 6 MAR 1977 19-29 AUG 1977 - 30 20 13 V) < cc T I — I — I — I — I — I — I — I — T" — r — r ooooooooooooo ooooooooooooo (\J'tiDooOoj^t\l'J"(X)aDO(\| — — — — CM(M(MOOOO — — r 50 - 40 - 30 - 20 - 10 T — I r^T 1 1 1 1 T 1 — r ooooooooooooo ooooooooooooo OJ^tDCDOOJ'J-CVI'J'lOCDOCM ___ — CJf^c^OOOO — — TIME OF DAY (EST) Figure 4. — Die! changes in abundance ofCallinectes Icirvae in neuston collections in the Middle Atlantic Bight, 1975-77. Abundance (dashed line) for each time interval was averaged over the stations of a cruise. The rank sum for each cruise (solid line) was calculated by summing over all stations the ranks (lowest to highest) of the abundances at each time interval. (The rank sum weights the frequency of occurrence at each time interval.) sunrise) on all cruises except fall 1976, when abundance was greatest during daylight hours. 258 The rank sum, which weights both abundance and frequency of occurrence, indicated patterns of SMYTH: CALUNECTES LARVAE IN MIDDLE ATLANTIC BIGHT diel change similar to diel patterns of total abundance — except during fall 1976. As distance from shore increased and abundance decreased, however, Callinectes larvae (late zoeae and megalopae) were generally collected at the surface only at night. Ten of 15, and 10 of 12 occurrences (megalopae) during winter and spring 1977 were at night. Larvae were taken at salinities ranging from 30.5 to 35.8%o and temperatures from 1 1 .0° to 25.7° C (surface temperature and salinity); peak abun- dance occurred in the ranges 31.6-34.9%o and 20.6°-25.7° C. Mean temperature, salinity, and distance from shore, weighted for abundance, for all neuston collections of Callinectes larvae were 22.9° C, 31.9%o, and 55.9 km. Plots of temperature and salinity vs. abundance indicated no clear rela- tionships among these variables. For the independent variables — temperature, salinity, distance from shore, and depth — simple (bivariate) correlation analysis indicated strongest correlation between mean neuston abundance per station and salinity and weakest correlation of abundance with bottom depth (Table 4). Table 4. — Simple correlation matrix for surface abundance of Callinectes larvae and selected environmental variables. Table 5. — Partial correlation coefficients for surface abundance log,(,[X-i-l]) oi Callinectes larvae with selected environmental variables. Variable Abundance (log,o[X+1]) Temperature (°C) Salinity Distance from shore (km) Temperature Salinity Distance from shore Bottom depth 0.6260— -0.7086— -0.5812"- -0.4024" -0.5133" -0.1695 -0.1218 0.6259"- 0.5621" 0.6261 — -•p< o.oi,---p c ■a c 3 CO (I + X ) 6o| ' ^uj 001 ■'•d sjaquj«N en C 8 C8 s >> ^ g 8% fish meal, the poultry tends to pick up a "fishy" flavor. With <2% fish meal, further substitution of vegetable protein meals for fish meal will result in slower growth because the fixed quantity of feed eaten per day per chick cannot contain the ideal mix of amino acids. When fish meal is extremely high priced or 'Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, La Jolla, CA 92038. ^Public Law 94-265, 94th Congress, 2d Session 13 April 1976, 16 use 1801 et seq. (Suppl. 1977). Hereafter, FCMA. Manuscript accepted November 1979. FISHERY BULLETIN: VOL. 78, NO. 2, 1980. 3J. Vondruska. 1979. Postwar production, consumption, and prices of fish meal. Unpubl. manuscr., 66 p. National Marine Fisheries Service, 3300 Whitehaven St., N.W., Washington, DC 20235. 267 unavailable, the lysine and methionine content of the feed can be augmented with synthetic pro- teins. Kolhonen (1974) described the develop- ment of synthetic methionine and lysine for use in feed formulas. Linear programming has been widely adopted by formula feed manufacturers in the United States and western Europe (Kolhonen 1974). Least-cost combinations of feed constituents needed for adequate nutrition are quickly and ac- curately computed for any vector of constituent prices. Thus, the demand for feed ingredients is expected to exhibit great sensitivity to relative prices. In a recent examination of demand for ag- ricultural feed ingredients, Meilke (1974) re- ported that price elasticities are generally >2 in absolute value. It is expected that the demand for fish meal will be elastic also, at least when avail- able quantities allow the feed formula manufac- turers to include between 2 and 8% fish meal in poultry rations. When the supply of fish meal is low enough to jeopardize the maintenance of at least 2% fish meal, the demand may become in- elastic. Thus, one hypothesis to be tested is that the own price elasticity of demand for fish meal falls with increasing price and decreasing quan- tity. Markets for fish meal in the United States are, for obvious reasons, concentrated in the poultry- producing regions — California, Arkansas, and states in the Deep South. Domestic production of fish meal occurs mainly in California, the Gulf Coast States, and the South Atlantic States. In some years, however, much of the domestic sup- ply is imported from major foreign producers such as Peru. Foreign meal is a perfect substitute for the domestic product, but the supply of foreign meal has undergone tremendous fluctuations due to variations in fish stocks (especially the Peru- vian anchoveta, Engraulis ringens). Domestic supplies have also been strongly influenced by uncontrolled variations in domestic stocks (espe- cially menhaden Brevoortia tyannus and B. pat- ronus) and by administrative decisions of fishery management agencies (California's anchovy, £;/i- graulis mordax, fishery, e.g., see Pacific Fishery Management Council 1978: 31660-31664). On the supply side of the domestic market, therefore, the major fluctuations are not price induced, but are due to exogeneous factors. On the demand side the poultry industry experienced a steady expan- sion starting in the early 1950's and continuing until about 1970. FISHERY BULLETIN: VOL. 78, NO. 2 DEVELOPMENT OF DEMAND MODEL Demand and price analysis has been a cor- nerstone of applied economic research since the 1930's (Working 1927; Schultz 1938; Wold and Jureen 1953). Agricultural economists have been particularly active in developing demand models for commodities. Research on demand for fish is of more recent vintage but differs in few im- portant respects from that for agricultural com- modities. For an excellent review of the historical development of demand analysis, see Waugh and Norton (1969). Among the methodological issues addressed in applied demand studies are: 1) spec- ification of the demand model, 2) development of appropriate functional forms, 3) treatment of simultaneity bias in market demand and supply function estimates, and 4) incorporation of dynamic response mechanisms in the demand model. These issues are discussed seriatim. Specification The specification of a demand model consists of the choice of dependent and independent vari- ables. Annual quantity demanded, as measured by quantity purchased, should be the dependent variable. Purchased quantities are difficult to ob- tain, however, while production, import, and ex- port statistics are well documented. Also, meals derived from different sources differ in protein content and sell at different prices. Both the quantities and the prices must be aggregated such that they represent a reasonably homoge- neous commodity. Fish meal quantities (Table 1, columns 1-6) are converted to a protein equiva- lent basis by multiplying the quantity of each type of meal by the prevailing percentage of pro- tein content. The total available domestic quan- tity, computed by summation of protein equiva- lent fish meals and subtraction of exports, is listed in Table 1, column 7. Similarly, since the prices of the various fish meal types (Table 2, col- umns 1-4) are based on protein content, each price is converted to a protein basis. The aggre- gate price of fish meal introduced as an indepen- dent variable in the demand model is the average price per unit protein for all meal supplied to the U.S. market (Table 2, column 5). Some specifica- tion error may enter the model because domestic supply rather than quantity purchased is used for the dependent variable, but this problem is un- avoidable with available information. 268 HUPPERT: ANALYSIS OF UNITED STATES DEMAND FOR FISH MEAL Table l. — United States fish meal supplies, 1955-76 (thousands of metric tons). (National Marine Fisheries Service 1975, 1977.) (1) (2) (3) (4) (5) (6) (7) Total supply Men- An- Im- Ex- protein Year haden Tuna chovy Other' ports ports basis^ Table 2. — Annual average prices for various fish meals and average price per unit of protein in fish meal in the United States. (National Marine Fisheries Service 1975.) 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 172.9 191 1 156.4 143.4 203 1 198 1 2246 217.5 167.1 145.4 159,7 122.5 108-1 129.9 144.7 171.1 200.4 175.6 171.3 185.0 1736 192.9 21.2 23.9 23.3 23.0 23.0 24.0 192 24 1 24.5 19.1 23.0 23.0 23.1 26.1 24.4 24.2 26.6 39.2 39.6 43.7 33.7 36.4 0.0 0.0 00 00 0.0 0.0 0.0 00 0.0 0.0 0.0 4.1 5.1 2-5 10.3 14.7 6.9 10.1 20.0 12.8 25.1 19.9 37.6 44.1 51.3 502 43.1 33 1 28 8 31 5 33.4 39.6 37.3 42.9 46.5 47.4 42 1 232 22,7 238 22.4 23.0 20 9 220 88.9 820 73.7 91.1 120.6 119.4 1976 2289 341.4 398.3 2455 406.2 591.0 775.9 325,1 227,8 256,9 355,6 62 1 62.0 107.4 127.4 na,-' n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a, 43 92 9.5 33.3 50.3 10.7 30.0 195.8 2075 185.3 188.0 238.8 2296 291.0 311.4 355.7 380.4 290.4 378.6 492.9 626.7 343.6 284.9 314.4 373.2 171.4 167.2 215.0 226.7 'Primarily from offal, waste, and scrap from groundfish and herring ^Converted to protein as follows: menhaden, exports, and other meal as- sumed to be 60°o protein: anchovy and imports assumed to be 65% protein: tuna meal assumed to be 55% protein. Total supply is production plus imports minus exports, ^n.a, means data not available. Year 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 (1) Men- haden (2) Tuna (3) (4) Domestic Peruvian anchovy anchovy (5) Average price per unit protein in fish meal Actual' Deflated' -dollars per metric ton of meal- 123.4 121.7 117.7 125.0 116.2 84.4 106.4 112.6 114.1 119.3 152.9 146.1 123,8 131,8 158,1 167.4 143.3 168.3 433.8 250.5 216.9 314.3 130.7 121.7 114.8 124.8 117,1 86,0 99,7 109.5 106.2 115.8 143.2 134.4 117.6 114.2 132.6 155.4 128.0 141.4 359.6 245.5 206.3 347.8 137.3 117.5 110.7 137.9 156.0 140.4 154.1 3655 270.3 214.8 247.5 121.5 128.2 131.9 86.1 100.0 111.2 109.7 119.7 140.3 141.9 118.1 118.8 142.5 176-4 150.7 162.6 409.8 260.6 226.5 309 9 1.99 1.97 1.95 201 1.95 1.39 1.69 1.81 1.76 1.88 2.30 2.23 1.86 1.88 2.31 272 2.33 259 675 4.15 3.54 4.92 4.34 4.16 4.01 4.19 3.94 2.80 3.40 364 3.58 3.81 4.63 4.33 3.57 3.53 424 4.72 3.92 4.15 9.70 499 388 5.15 ' For each meal, price per unit protein equals price per ton divided by percent protein. Average price computed by weighting the price per unit protein for each meal by the proportion of U.S. fish meal protein supplied by that meal. ^Deflated by Wholesale Price Index, all commodities (January 1977 = 100). In addition to the price offish meal, the demand model should contain independent variables rep- resenting 1) the prices of close substitute prod- ucts, 2) prices of complementary products, and 3) the level of production activity that governs the demand for fish meal. Several high protein meals (e.g., soybean, cottonseed, meat, and bone meals) are potential substitutes for fish meal in poultry rations. Soybean meal is the most com- mon substitute, and its price is used as an inde- pendent variable in the demand model. The price of corn meal (Table 3, column 2) is introduced as a complementary product price. Demand for fish meal is expected to increase when the price of a substitute product increases, and is expected to decrease when the price of a complementary prod- uct increases. Finally, the overall production of poultry products would cause shifts in the level of demand for fish meal independently of the prices. The poultry and egg production index (Table 3, column 3) is adopted as the appropriate measure of this factor. In summary, the fish meal demand model is specified as follows: 1) Quantity demanded, the independent vari- able, is represented by annual production plus net imports of protein-equivalent meal. 2) Annual price of fish meal is measured as the weighted average of the prices per unit protein for all domestically supplied meals. Table 3. — Exogenous variables in the fish meal demand model. Price of Price of Poultry and egg domestic domestic production index^ Year soybean meaP corn^ (1976 = 100) dollars per metric ton 1955 51.6 2.24 58 1956 46.5 2.30 63 1957 42.7 2.06 64 1958 50.8 1.99 68 1959 51.3 1.94 70 1960 48.2 1.84 70 1961 57.3 1.80 75 1962 60.3 18C 75 1963 65.8 2.00 77 1964 62.8 1.99 80 1965 64.9 2.07 83 1966 76.0 2.18 88 1967 694 2.06 92 1968 70.3 1.80 90 1969 67.6 1.96 92 1970 71.8 2.19 97 1971 70.7 2.15 98 1972 95.2 2.10 100 1973 216.3 3.57 97 1974 127.8 5.20 97 1975 112.6 4.71 94 1976 147.5 4.37 100 'Forty-four percent protein. Simple tional Marine Fisheries Service (1977) 2Price of No. 2 yellow corn, Chicago PES-294, 1965-77. 3From Schultze et al. (1979). average price at Decatur, III., from Na- USDA, ERS, Poultry and Egg Situation, 269 FISHERY BULLETIN: VOL. 78, NO. 2 3) Annual domestic price of corn and annual domestic price of soybean meal are introduced as complementary and substitute product prices. 4) The trend in aggregate demand over time is accounted for by the aggregate poultry and egg production in the United States. All of the variables expressed in dollars are deflated by the Wholesale Price Index to elimi- nate spurious correlations caused by the inflation- ary trend. Functional Form Demand studies typically utilize least squares regression methodology with either a linear or a log-linear equation. As noted by Chang (1977), however, there is no a priori reason to choose one of these forms. Each form imposes some fairly strict conditions upon the characteristics of the demand function which may contradict theoreti- cal considerations or actual experience. Linear equations imply that the elasticity of demand with respect to any independent variable is a de- creasing function of that variable; a log-linear equation implies constant elasticities. Chang suggests that the income elasticity of demand for meat should fall with rising income. A similar consideration applies to fish meal demand. At low prices, feed manufacturers would use near maximum amounts of fish meal allowable and could easily substitute soybean meal for fish meal. With relatively high fish meal prices, feed manufacturers would use a smaller proportion of fish meal, but as price rises further it would be increasingly difficult to maintain desired quan- tities of lysine and methionine by substitution of soybean meal. Thus it is clearly unwarranted to rule out increasing price elasticity through a priori choice of functional form. The function to be fitted by regression analysis can be chosen by determining the appropriate transformation of variables for the linear least squares procedure. The log-linear transformation is a special case of a parametric family of trans- formations introduced by Box and Cox (1964). The parameter defines the transformation function is expressed as r* = = ix' - l)/\. (1) Equation (1) is linear for A = 1, and becomes logarithmic as \ approaches zero. The demand 270 ,* — 6o + ^1 •''^1* + . . . + bf^Xk* + u (2) where q is the quantity demanded, thex's are the independent variables affecting demand, u, is a stochastic error term, and the 6, and k are parameters to be determined. The superscript * indicates that the variable has been transformed as in Equation (1). Price elasticity of demand is defined as the ab- solute value of the ratio of percentage change in quantity demanded to percentage change in price. Assuming that the first independent vari- able is the price, E tion (2) we get dq (t) E = \bi\(q/x^: From Equa- ls) The elasticity defined in Equation (3) is an in- creasing function of x^ when A>0, and is a de- creasing function of x^ when \<0. Thus the esti- mate of the transformation parameter X provides a test of whether the price elasticity increases, decreases, or remains fixed along the demand curve. Simultaneity Bias In economic theory, the supply and demand curves interact to determine the market price. Over a period of time, shifts in both supply and demand factors cause the market price and ob- served quantities of products to vary. Without these shifts, only one price and quantity would be observed, making it impossible to estimate a de- mand or supply curve. When the demand curve remains stable, the observed price-quantity pairs "trace out" the demand curve with, of course, some stochastic error, and a regression analysis will result in a demand curve estimate. When the supply curve remains stable, the observed data will fall along the supply curve, and a regression analysis of the price-quantity relationship results in a supply curve estimate. If shifts in both de- mand and supply occur, the resulting data will not unambiguously identify either of these two curves, and an ordinary least squares regression will generally result in a set of parameters re- flecting neither the supply curve nor the demand HUPPERT: ANALYSIS OF UNITED STATES DEMAND FOR FISH MEAL curve. In this case the estimated parameters are said to suffer from simultaneity bias. The general statistical problems associated with estimation of individual structural relation- ships in a simultaneous equation system were first examined by Haavelmo (1943). Development of appropriate statistical methods for estimating simultaneous equation systems has been a major area of research for econometricians over the last two decades (Kmenta 1971). In estimating the demand curve for fish meal, however, direct re- gression estimates seem appropriate, because most of the observed variations in annual fish meal supplies are due to exogeneous shifts rather than price-induced movements along a stable supply curve. Production offish meal is subject to wide fluctuations due to uncontrolled variations in the fish stocks exploited (Kolhonen 1974). At the same time, formula feed and poultry indus- tries have remained relatively stable during the last 20 yr except for the secular growth accounted for in the analysis. Under conditions in which the random shifts in supply are much greater than the corresponding shifts in demand, the ordinary least squares procedure results in no significant simultaneity bias (Rao and Miller 1971). effect of a price change may be drawn out over several periods of time. A fairly simple model for representing a lagged response is the "partial ad- justment model" originally developed by Nerlove (1958). Corresponding to any given level of the independent variable, p, there is an optimal or desired level of the dependent variable q. For a demand function with one independent variable, the level of demand fully adjusted to input prices by formula manufacturers represents the desired level offish meal usage: qf=bp^ +u, (4) where the superscript d signifies desired level. Because purchasers of meal cannot im- mediately adjust to this desired level of usage, the demand Equation (4) is not directly observable. By assuming a simple structure to the adjust- ment process, however, an estimable equation is obtained. The partial model assumes that a fixed percentage of the adjustment to desired level is made each year. This introduces the difference equation Qt -9m = yiqf-qt.i) (5) Lagged Response Mechanisms The use of annual price and quantity data for estimating the demand function requires that the response to a change in price occurs rather rapidly, at least within a period of time much shorter than a year. Since most domestic formula feed manufacturers employ professional nutri- tionists and cost-minimizing computer routines in calculating formulas, the response to changes in the vector of prices is probably rapid. If so, each annual quantity consumed may be assumed to represent at least approximately an equilibrium demand response to the set of independent vari- ables. The assumption of rapid response and equilibrium approximation, however, has not been directly verified. In the interests of rigor it is useful, therefore, to consider alternative assump- tions. A lagged response to a change in price may occur due to rigidities in mixing procedures or personnel, inventory management problems, or time lags in renegotiating contracts for supply of input or sales of products. If any of these factors results in a sluggish response in the substitution between fish meal and other protein meals, the Solving this for q^ and substituting from Equation (4) yields q=b yp, ^ (1 - y)(7,^i + yu,. (6) The adjustment parameter, y, must be a positive number <1. Larger values of y imply more rapid adjustment to changes in the independent vari- able. The impact of a unit change in p, is distrib- uted over time in an exponentially decaying fash- ion with successive annual changes in q being equal to by, by (1 - y), by [1 - y)^, and so forth. The ultimate change in q due to a change in p is Iq =blp lyil y)J = blp (7) where j = lag. The elements in the sequence under the summation sign are all positive frac- tions, and sum to one, so that the sequence can be treated like a probability distribution. Each ele- ment represents the percentage of the total effect occuring in year t, and the mean of the distribu- tion, ( 1 - y)/y, represents the mean lag in the adjustment process. Distributed lag models like that in Equation (4) result from other conceptual models such as models of expectations formation 271 or habit formation. And the exponentially distri- buted lag is but one of a large class of more com- plex lag models (Griliches 1967; Kmenta 1971; Rao and Miller 1971). Application of the partial adjustment model to the demand Equation (2) results in the following: 4 Q*t = ^0 + ^a.x* + a,ql^ + u, (8) where the coefficients a, can be interpreted in terms of the coefficients of Equation (2) as follows: at = yb,;i = 1, . .4 a. (1- y). STATISTICAL PROCEDURES For a given value of the transformation parameter, k, the coefficients of either the equilibrium model [Equation (2)] or the partial Adjustment model [Equation (8)] can be esti- mated by the ordinary least squares method. Two statistical issues requiring further development, however, are the selection of the "best" value for A, and the test for significance of the lagged ad- justment parameter. An appropriate procedure for estimation of K was first suggested by Box and Cox (1964). The procedure is more clearly explained in the linear regression context by Kmenta (1971) and is reviewed by Chang (1977). For a fixed value of A, the linear regression proce- dure yields an estimate of the error variance &^. Box and Cox showed that the maximized log likelihood is, except for a constant, ^max (^) = -(N/2) log d-2 (A) + (A - 1)S log 9,. (9) A maximum likelihood estimate of A can, there- fore, be found by searching through successive values of A to maximize Equation (9). The use of this likelihood function implies, of course, that the error terms conform to full normal theory as- sumptions, i.e., that the w, are independently normally distributed with zero mean and con- stant variance. An approximate 100% (1 - a) confidence region for A is defined by ■^max *^^) ~ L (A) < 1/2 x,Ha) (10) where XiHa) represents the value of the chi- square distribution with 1 df (Box and Cox 1964). 272 FISHERY BULLETIN: VOL. 78, NO. 2 Serial correlation in the errors of the regression model raises problems in the interpretation of the test statistics for the nonlagged variables and the lagged adjustment parameter, and contradicts the assumptions of the log likelihood function. Careful examination of the hypotheses and statistics regarding the residuals of the regres- sion equation is clearly necessary. Existence of serial correlation in the errors of the static de- mand model can be tested with the Durbin- Watson statistic. If no serial correlation is appar- ent in the residuals, then neither the distributed lag model nor the serial correlation model need be considered. If serial correlation is present in the residuals of the static model, then the problem is to distinguish between the distributed lag model and the serial correlation model. Griliches (1967) showed that the serial correla- tion and lagged adjustment models cannot be dis- tinguished by a simple ^test on the adjustment parameter. For example, if errors generated by a first order Markov process, i.e., e^ = se,_i + Uj, occur in a regression equation, the coefficients of the lagged variables may be judged significant by the usual ^-test even though there is no real lagged response in the underlying structural relationship. Similarly, it can be shown that se- rially correlated residuals will occur if a non- lagged model is mistakenly fit to data from an inherently dynamic process. Although there is no fully satisfactory method for determining which model is the truth, Griliches (1967) developed a provisional test. Briefly, the serial correlation model is 9, =«o +}^^xu +^t (11a) (lib) where s is a positive fraction and u, is a nonse- rially correlated error term. From Equation (lla),e,.i =Qt-i -o-Q- S,a,:<;,,.i;sothate^ = s(9m - Uq - S(a,x„_i) + "r Substituting this into Equation (11a) yields % =<1 -s)ao + S(a^A:,, - 6,x,,.,)+ s g^ ^ + u,. (12) When Equation ( 12) is computed, the serial corre- lation model implies that a^s = -6, for each i. Griliches suggested that the first-order serial cor- relation model be rejected if these four equalities do not appear to hold. Thus, there are four h3^otheses of the following form: HUPPERT: ANALYSIS OF UNITED STATES DEMAND FOR FISH MEAL Ho:(6, +sa,) = 0. (13) An approximate sample variance for ib, + sa^) is computed by the "delta method" described by Saber (1973). The expression for approximate variance of a function of a vector of random vari- ables, G(x), is v[G(x)] = Zy(x,)(|^y + 221: COY (jc i23 LENGTH, mm NL or SL Figure l. — Number of dorsal fin rays in relation to length in 161 Coryphaena equiselis (6.5-230 mm NL or SL) and 211 C. hippurus (5.0-172 mm NL or SL). Range (vertical line), mean (horizontal line), and 2 standard errors about the mean (white and black bars) are indicated. Number of specimens for each length interval is given above the range and is in italics for C. equiselis. 279 FISHERY BULLETIN; VOL. 78, NO 2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 21 23 24 25 26 1 I I I I I I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 < ' ' ' ' 6mm \ \ \ ^VM^\\\(\M^\ \ \ \ \ \ \ //////////////// \\\\\\\\\\\\\\\\v /////// //////////////// ////////// ////// // //- m\ VA\A\A\\\A\A\\\\\\t ,,^ ^v\\\ wwwwwwvwwwv ,r////////////7/7////////////////////////////////F //////////////////////////////////^^^^ r////m///////////////////////7/ // //////////7////////r m^ 1,11 a\vv\vv\ \ 36 mm ,p\\\\\\\ \\\V\\\V\\\\\V\VNN\ I 1 — I — I — I — I — I I I I I I I I I I I I I I 1 2 3 4 5 6 7 8 9 10 11 12 13 14 IS 16 17 18 19 20 21 22 23 24 25 26 280 POTTHOFF: DEVELOPMENT AND STRUCTURE OF FINS IN CORYPHAENA pterygiophores had to be added from place of origin anteriorly than posteriorly. Small C. hip- purus (5 mm NL-11 mm SL) usually had fewer fin rays compared to equal-sized C equiselis (Figure 1). Between 12 and 14 mm SL both species had about equal dorsal fin ray numbers. Spec- imens of C. hippurus 15 mm SL and longer usually had more dorsal fin rays than equal-sized C. equiselis. The developmental sequence of dorsal fin rays in Coryphaena spp. is similar to that observed in Trachurus symmetricus ( Ahlstrom and Ball 1954), Haemulon plumieri (Saksena and Richards 1975), and Archosargus rhomboidalis (Houde and Pott- hoff 1976). It is as though Coryphaena spp. is developing two dorsal fins in the same pattern of the above examples, e.g., first the second dorsal fin followed by the first dorsal. It is of interest to note that most scombroids do not follow this pattern and develop the first dorsal fin first (Voss 1954; Potthoffl975). Dorsal Fin Pterygiophores Counts There was a supporting pterygiophore in both species of Coryphaena in a jointed series for each dorsal fin ray, except for the first two or three anteriormost rays. Each pterygiophore had a proximal and a distal radial. The distal radial was located between the bifurcate base of the fin ray. Proximal and distal radial and fin ray formed a series, hence, a serial association. Each fin ray also closely approximated the following posterior pterygiophore in a secondary association. Thus, each pterygiophore supported a ray in a serial association and an immediately anterior ray in a secondary association. The exceptions were found at the beginning and the end of the fin. The anteriormost pterygiophore supported from one to three rays, but most often two rays (Table 2). Also, in 2 out of 70 specimens of both species, no rays were associated with the anteriormost pterygio- phore, and the pterygiophore was very small and almost a vestige. The posteriormost ray in the dorsal fin was a double ray which was serially Figure 2. — Schematic representation of dorsal and anal.fin and pterygiophore development in Coryphaena hippurus in relation to the vertebral column and head. Oval -shaped representation of pterygiophores are cartilaginous when white and ossifying when black. Scale represents interneural and interhaemal spaces and points align with midpoint of vertebral centra. Table 2. — Number (adult count) of anteriormost dorsal fin rays without distal radials and number of dorsal fin rays associated with the anteriormost dorsal fin pterygiophore for 28 Cory- phaena hippurus (78.8-176 mm SL) and 35 C. equiselis (74.1-172, 314mmSL). Item Species Number of anterior- most dorsal fin rays 12 3 Without distal radials Associated with the ante- riormost pterygiophore C. hippurus C. equiselis C. hippurus C. equiselis 12 1 6 3 1 12 4 25 4 24 1 17 19 associated with the posteriormost pterygiophore. This was the only ray in the dorsal fin which lacked a secondary association. Total dorsal fin ray count in both species was either one less than the pterygiophore count, equal to the pterygio- phore count, or one or two greater than the pterygiophore count. Thus, the two species dif- fered in their pterygiophore number as they differed in their fin ray counts. In larvae, juveniles and small-sized adults of Coryphaena spp. the proximal radials of the dorsal fin were inserted in interneural spaces. The first interneural space was bounded anteriorly by the head and posteriorly by the first neural spine, followed posteriorly by the remaining interneural spaces which were bounded by all other neural spines (Figure 3). Fully developed specimens of the two species of Coryphaena differed by the number of pterygio- phores that occupied the interneural spaces. The number of pterygiophores found in the first inter- neural space separated the species most of the time, with 10-14 (x = 11.0) for C hippurus and 7-11 ix = 8.0) for C. equiselis (Figures 3, 4). The species also differed in the number of pterygio- phores associated with the remainder of the inter- neural spaces. Although individual variability within each interneural space was too great to allow this character to be used to separate the species, the mean number of pterygiophores in each interneural space was always greater for C hippurus. The species also differed in the number of interneural spaces that were occupied by the dorsal fin pterygiophores (Figure 3; Tables 3, 4). In C. hippurus the dorsal fin pterygiophores extended to the 26th interneural space and seldom to the 27th, whereas in C. equiselis they extended to the 28th and seldom to the 27th or 29th space. There was some overlap for the two species in this character, but if the termination of the anal fin pterygiophores was taken into account, together 281 FISHERY BULLETIN: VOL. 78, NO. 2 '^iiilJIIJIJIJUIMWIJIJIJIJIJIJIJIJLUUJJJJJIJJI. 14 15 1 16117 I IB I IVI TO! ^1 1^^ I '•> I^'tl'OI -lo I .43 LENGTH, mm SL Figure 4. — Number of pterygiophores in the first intemeural space in relation to length in 192 Coryphaena equiselis (8.6-173, 314 mm SL) and 193 C. hippurus (8.6-176 mm SL). For explanation of symbols, see Figure 1. with the termination of the dorsal fin pterygio- phores, complete separation for the two species resulted (Figure 3, Table 3). 282 Dorsal fin pterygiophores had the same pattern of appearance in both species of Coryphaena as the dorsal fin rays. Cartilaginous pterygiophores POTTHOFF: DEVELOPMENT AND STRUCTURE OF FINS IN CORYPHAENA Table 3. — Adult and juvenile position of posteriormost dorsal and anal fin pterygiophores in their intemeural and inter- haemal spaces for 193 Coryphaena hippurus (9.0-176 mm SL) and 186 C. equiselis 18.9-172, 314 mm SL). For numbering of vertebrae and spaces, see Figure 3, Intemeural space numbers Interhaemal space numbers Species 26 25 26 26 26 27 27 26 27 27 27 28 28 27 C. hippurus C. equiselis 2 172 3 11 5 2 2 28 28 29 29 28 29 28 29 172 without rays were first seen in the 22d-24th myomeres at 5.9 mm NL in C. hippurus (Figure 2) and with some rays in the 18th-27th myomeres at 6.5 mm NL in the smallest available but more advanced C. equiselis. In both species of Cory- phaena the pterygiophores appeared shortly be- fore the fin rays developed. As pterygiophores were added, anteriorly and posteriorly rays were lacking for one or two anteriormost and posterior- most additions (Figure 2). In both species, the posteriormost intemeural spaces (numbers 26-28) were occupied with ptery- giophores between 7 and 8.5 mm SL (Table 4). The anteriormost intemeural space started to fill with pterygiophores at 9.3 mm SL in C. equiselis and at 13.1 mm SL in C. hippurus (Figure 4, Table 4). Adult counts in the anteriormost intemeural space of 7-11 pterygiophores were obtained for C. equiselis between 12.3 and 23.2 mm SL and for C. hippurus of 10-14 pterygiophores between 18.7 and 30.8 mm SL (Figure 4, Table 4). Ossification of the pterygiophores started in the same area and proceeded in the same directions as the cartilaginous development (Figure 2). Ossifi- cation of pterygiophores occurred first at 8.8 mm SL in C. equiselis and at 9.7 mm SL in C. hippurus (Table 4). The posteriormost intemeural space number 28 of C. equiselis had ossifying pterygio- phores at 8.9 mm SL, and posteriormost inter- neural space number 26 of C. hippurus had ossifying pterygiophores at 10.2 mm SL (Table 4). Specimens 10.3-11.6 mm SL of C. equiselis, and 16.2-19.2 mm SL specimens of C. hippurus had one or more ossifying pterygiophores in the first inter- neural space (Figure 4, Table 5). All dorsal fin pterygiophores were ossifying in both species at about 45 mm SL when the count of ossifying pterygiophores was in the adult range and the first intemeural space did not have anterior cartilaginous pterygiophores (Figures 4, 5). CO c bo C °C Q .J OQ OJ CJ CO r^ CO GO ^^CDOOCOOCOCOOOCDOTcOCD I C\JCMCMC\JCMC\JCJCMCvJC\JCvJ I O0a> ^'-Tf^ y- f- iriiT) cbcb(b(£> ^ to C^CVJ^ CyCMCJCVJ CVJOU ,totDcDr^r^^^rjcvj(£> yDcb ^^CJCMCM CNJCMCslCVJ C\JC\J C\JCNJCNJCMC\JC\JCgrvJ0slC\JCNiC\iCMOJC\J CNJCT)") to to to to ti> to OJCMCNJ CvJCgCvJCM C\JC\J O to I (D in o cr> in y^ c3 t--: CO in ,^ , og ,- t-- C\J CD ■.- .CO'-'-'-COC^ i-t- ^CMCNicS'focDr^cgcvjcvj C\J OJ ■- T- ci)cD(X)uSc\jcvj T^T^ §2 uoioininLninmioininminioinmir) uSybtDtot6tDc6c6cD;bcc)c6cD(DCDtD ■^iO(Dr^cboio»-c\Jco^ir)cor^Qb<7> E > CO 0) en 13 o o o CO. o o £ B Q.Q. o o zz 283 FISHERY BUIXETIN: VOL. 78, NO. 2 Table 5. — Sum ladult count) of anal fin pterygiophores in the two anteriormost interhaemal spaces, numbers 14 and 15, in 35 Coryphaena hippurus (49.9-176 mm SL) and 32 C. equiselis (74.1-172, 314 mm SLj. For numbering interhaemal spaces, see Figure 3. Number of pterygiophores Species 4 5 6 7 8 9 10 C. hippurus C. equiselis 5 20 14 18 6 1 2 1 Morphology' and Development The pterygiophores in the center area of the dorsal fin developed first in both species. A ptery- giophore (proximal and distal radial) appeared as one elongate piece of cartilage (Figure 6). Ossifica- tion was first observed at the middle part of the pterygiophore cartilage (Figure 6) and proceeded distally and proximally along the cartilage until only cartilage tips were present at the extremities. At this point, the sagittal and lateral keels began to develop (Figure 4). Further development of the pterygiophore consisted of growth of the keels, growth of bone around the locus of secondary fin ray association, and segregation and ossification of the distal radial. The distal radial developed from the distal tip of the pterygiophore cartilage late during ontogeny (Figure 6), and ossified into two pieces of bone (Figure 7). The pterygiophores in the posteriormost area of the dorsal fin developed similarly to those of the center area. The posteriormost pterygiophore sup- ported one ray in series. This ray developed from two rays but was counted as one according to Hubbs and Lagler (1958). In adults, the base of the anterior ray fitted closely over the base of the posterior ray and the base of the posterior ray articulated with the distal radial of the posterior- most pterygiophore (Figures 8, 9). The supports of the anterior portion of the dorsal fin developed last. In C. equiselis the first interneural space was almost filled with cartilag- inous pterygiophores, but in equal-sized C hip- purus the first interneural space was empty and the second interneural space had only one carti- laginous pterygiophore (Figures 10, 11). The ante- riormost cartilaginous pterygiophores always had (/) 14[ 13 12f n 10 9 8 7 6 5 4 3 2 C. equiselis^ C. hippurus ^ ^ 10 11 12 ]3 14 15 16 17 18 19 20 21 22 23 24 25^8 33 38 43 LENGTH, mm NL or SL Figure 5.— Number of ossifying pterygiophores in the first interneural space in relation to length in 126 Coryphaena equiselis and 88 C. hippurus. For explanation of symbols see Figure 1. 284 POTTHOFF: DEVELOPMENT AND STRUCTURE OF FINS IN CORYPHAENA Figure 6. — Ontogenetic development of a dorsal fin pterygiophore with its serially associated ray in the 14th intemeural space of Coryphaena equiselis (left lateral view). Pterygiophores drawn in the natural attitude relative to horizontal body axis. Starting from the left specimen length (millimeters SL) was: top, 7.6, 9, 9.8, 11.5, 12.2, 14.2; bottom, 17.4, 22.5, 45.4, 31. Symbols: D, distal radial; P, proximal radial; R, serially associated ray. Stippled, cartilage; darkened, bone. a ray developing concurrently (Figures 2, 10, 11). In specimens of both species, which had the full count of pterygiophores in the first intemeural space, it was common to have a ray develop in front of the cartilaginous pterygiophore (Fig- ure 2). The pterygiophores of the first intemeural space in large juveniles and adults of both species were vertical to the body axis near the first neural spine and slightly anteriorly inclined dorsad near the head (Figure 12). The anteriormost pterygio- phore in the adults was either of normal size (not figured), very small (Figures 12, 13), or just a vestige (not figured). In a few instances, in both species, the anteriormost pterygiophore was com- pletely or partially fused to the second pterygio- phore. The anteriormost pterygiophore of both species had either one, two, or three associated rays (Table 2). For the two species the anterior- most dorsal fin ray was either normal in size or a vestige (Figures 12, 13). In both species three types of first fin ray vestiges were observed: a paired vestige (Figure 13), a single right vestige, and a single left vestige. Distal radials were present between the bases of each fin ray for almost the entire dorsal fin. Distal radials were last to ossify from the distal portion of the pterygiophore cartilage. Only the anterior- most three fin rays of both species sometimes 285 FISHERY BULLETIN: VOL. 78, NO. 2 1.0 mm I 1 Figure 7. — Pterygiophore from 14th intemeural space with its secondarily and serially associated rays from a 230 mm SL (Joryphaena equiselis. Left: anterodorsal view, secondarily associated ray moved to the right of the proximal radial; right: left lateral view. Symbols: D,, distal radial of secondarily associated ray; D, distal radial of serially associated ray; P, proximal radial; R,, secondarily associated ray; R, serially associated ray. Stippled, cartilage; darkened, bone. lacked distal radials. The absence or presence of distal radials was not related to the number of fin rays associated with the anteriormost pterygio- phore (Table 2). The first three or four (anterior- most) distal radials of both species differed in structure from the remainder. These radials con- sisted of one piece of bone (Figure 14) whereas all other radials were of two pieces (Figures 7, 8). The dorsal pterygiophores of Coryphaena spp. differed in several ways from other perciform fishes. Predorsal bones reported in Apogonidae (Fraser 1972), Serranidae and Grammistidae (Kendall 1976), Sparidae (Houde and Potthoff 286 1976), and for all the stromateoid families (Ahl- strom et al. 1976) were lacking. Also lacking was the terminal bone in the dorsal fin support series called a "stay" by Weitzman (1962). Stays have been reported for such families as Characidae (Weitzman 1962), Scombridae (Kramer 1960; Pott- hoff 1975), Sparidae (Houde and Potthoff 1976), Nomeidae and Centrolophidae (Ahlstrom et al. 1976), and Centropomidae, Kyphosidae, Lutjan- idae, Percichthyidae, and Scorpidae (Johnson 1978). A stay was observed in the Scombrolabrac- idae and a double stay in the Gempylidae (Potthoff et al. 1980). POTTHOFF: DEVELOPMENT AND STRUCTURE OF FINS IN CORYPHAENA Figure 8. — Posteriormost dorsal fin pterygiophore with its secondarily and serially associated rays from a 230 mm SL Coryphaena equiselis. Left: anterodorsal view, secondarily associated ray has been moved to the right of the proximal radial; right: left lateral view, pterygiophore has been tilted 30° from the horizontal toward the vertical. Symbols: Dj, distal radial of secondarily associated ray; D, distal radial of serially associated double ray; P, proximal radial; Rj, secondarily associated ray; R, serially associated double ray. Stippled, cartilage; darkened, bone. 1.0 mm Figure 9, — Anterior and right lateral views of right side of a disarticulated posteriormost double dorsal fin ray with its distal radial from a 230 mm SL Coryphaena equiselis. a, right half of anterior ray, anterior view; b, right half of posterior ray and right half of its distal radial, anterior view; c, right half of anterior ray, lateral view; d, right half of posterior ray and right half of its distal radial, lateral view. Symbol; D, distal radial. 287 FISHERY BULLETIN: VOL. 78, NO. 2 HEAD Figure lO. — Left lateral view of anteriormost part of the dorsal fin and pterygiophores for a 11 mm SL Coryphaena equiselis, showing relationship of pterygiophores to head, intemeural spaces, and centra. Symbols: C, first centrum; ENT, second intemeural space; NS, first neural spine; P, proximal radial. Stippled, cartilage; darkened, bone. HEAD 0.5 mm Figure ll. — Left lateral view of anteriormost part of the dorsal fin and pterygiophores for a 11 mm SL Coryphaena hippurus, showing the relationship of pterygiophores to head, intemeural spaces, and centra. Symbols: C, third centrum; R, dorsal fin ray. For explanation of other symbols, see Figure 10. Stippled, cartilage; darkened, bone. 288 POTTHOFF: DEVELOPMENT AND STRUCTURE OF FINS IN CORYPHAENA Figure 12. — Left lateral view of four anteriormost dorsal fin pterygiophores with secondarily and serially associated rays from a 230 mm SL Coryphaena equlselis. Symbols: R, dorsal fin ray; P, proximal radial; D, distal radial. The proximal and distal radials (except the anteriormost three or four) of Coryphaena spp. were similar along the entire fin and were located between the bifurcate bases of the fin rays. Middle radials were absent in the posterior portion of the fin. In most other perciform fishes, distal radials differ between the first and second dorsal fins. The first dorsal fin distal radials are anterior to the bases of the fin spines, and the second dorsal fin distal radials are between the bifurcate bases of the fin rays, and middle radials are pres- ent posteriorly. Anatomically different distal radials for the first and second dorsal fins and the presence of middle radials posteriorly have been reported in the Carangidae (Berry 1969), Scombridae (Kramer 1960; Potthoff 1974, 1975), 289 FISHERY BULLETIN: VOL. 78, NO. 2 k 0.5 mm ^ Figure 13. — Anteriormost dorsal fin pterygiophore with secondarily associated vestigial ray from a 230 mm SL Coryphaena equiselis. Left: anterodorsal view, right: left lateral view. Symbols: P, proximal radial; R, vestigial ray. Sparidae (Houde and Potthoff 1976), Centro- pomidae, Kyphosidae, Lutjanidae, Percichthy- idae, and Scorpidae (Johnson 1978), and Gem- pylidae and Scombrolabracidae (Potthoff et al. 1980). Anal Fin The fully developed anal fin of Coryphaena hippurus has 25-31 rays (N = 147, x = 28, SE = 0.01, 16-172 mm SL) and that of C. equiselis 23-29 (N - 118, X - 26, SE = 0.01, 16-230 mm SL). The anal fin ray counts, in contrast to the dorsal fin ray counts, differ only slightly from those reported by Gibbs and Collette (1959), Rothschild (1964), and Shcherbachev (1973). Both species have adult anal fin ray counts at smaller sizes than dorsal fin ray counts (C. hippurus at 8-11 mm SL, C. equiselis at 8-9 mm SL). Anal fin rays v^^ere first seen in some C. hip- purus at 6 mm NL, just before the onset of dorsal 290 fin ray development and all C. hippurus had anal rays at 7 mm NL (Figure 15). The smallest available (6.5 mm NL) C. equiselis had 14 anal rays. Development of the anal fin of both species began in the finfold at the approximate center of the fin, below the 22d or 23d myomere. Addition of rays was in an anterior and posterior direction for both species (Figure 2). As in the dorsal fin, the posterior portion of the anal fin was completed first and the anteriormost rays developed last. From 6 mm NL to 9 mm SL, C. hippurus had fewer anal fin rays than C. equiselis; at 10 and 11 mm SL, both species had about equal numbers of rays; at 12 mm SL and longer, C. hippurus tended to have more anal rays than C equiselis (Figure 15). Appearance and additional sequence of anal fin rays in Coryphaena spp. are similar to Scomber Japonicus iPneumatophorus diego) (Kramer 1960), Thunnus atlanticus (Potthoff 1975), Haemulon plumieri (Saksena and Richards 1975), and Archo- sargus rhomboidalis (Houde and Potthoff 1976). POTTHOFF: DEVELOPMENT AND STRUCTURE OF FINS IN CORYPHAENA Figure 14. — Second anteriormost dorsal fin pterygiophore with secondarily and serially associated rays from a 230 mm SL Coryphaena equiselis. Left: anterodorsal view, secondarily asso- ciated ray has been moved to right of proximal radial; right: left lateral view, secondarily associated ray has been moved dorsally. For explanation of symbols, see Figure 7. 1.0 mm For Trachurus symmetricus, Ahlstrom and Ball (1954) reported an anterior to posterior anal fin development. Anal Fin Pter^giophores Q)unts The description for dorsal fin pterygiophores in the foregoing section may be applied to anal fin pterygiophores because of the similarities between the two fins and their supports. Pterygio- phores of the anal fin are inserted in the in- terhaemal spaces. The anteriormost (first) in- terhaemal space is bounded anteriorly by the stomach, intestine, and anus and posteriorly by the first haemal spine. The first haemal spine was of variable length, and in many cases did not reach the anal fin pterygiophores. The anal fin pterygio- phores in the two anteriormost interhaemal spaces were therefore summed (Table 5, Figure 3). Fully developed specimens of Coryphaena spp. differed in their numbers and arrangement of anal fin pterygiophores. The total number of pterygio- phores closely approximated the anal fin ray count. For both species the pterygiophore count was equal to or one to two less than the anal fin ray count. The sum of the pterygiophores found in the 291 FISHERY BULLETIN: VOL. 78, NO. 2 40 30 20 :}■ w e. equiselis ^ C. hippurus ^- 165 7 7 "' „ If -i ** % f 6 7 8 9 10 11 >11 LENGTH, mm NL or SL Figure 15. — Number of anal fin rays in relation to length in 159 Coryphaena equiselis (6.5-230 mm NL or SL) and 210 C. hippurus (5.0-230 mm NL or SL). For explanation of symbols, see Figure 1. first two anteriormost interhaemal spaces (14 and 15) separated the species most of the time, with 7-10 (X = 8.0) for C. hippurus and 4-7 (x = 5.0) for C. equiselis (Table 5). The two species also differed in the number of pterygiophores found in the remainder of the interhaemal spaces. Individual variability, however, was too great for each inter- haemal space to serve as a separating character. The mean number of pterygiophores for each interhaemal space was always greater for C. hippurus. Coryphaena equiselis had more in- terhaemal spaces with one pterygiophore and C. hippurus more with two pterygiophores. In C. hippurus the anal fin pterygiophores extended to the 25th, 26th, or 27th interhaemal space, but most often to the 26th, whereas in C. equiselis they extended to the 27th, 28th, or 29th inter- haemal space, but most often to the 28th. There was some overlap for the two species in this character, but if the termination of anal and dorsal fin pterygiophores is considered together, complete separation for the two species results (Table 3). The dorsal and anal fin pterygiophores most often terminated in opposing interneural and interhaemal spaces (Table 3). Morphology and Development Cartilaginous anal fin pterygiophores without fin rays were first observed in the 18th-24th myomeres (which approximately correspond to the 18th-24th interhaemal spaces) in a 5.9 mm NL C. hippurus (Figure 2, Table 6), but rays were developing in a 6 mm NL specimen. The smallest available C. equiselis of 6.5 mm NL had carti- laginous pterygiophores in myomeres 18-27. Fin 292 rays were developing, but a few anteriormost and posteriormost pterygiophores lacked rays. Addi- tion of cartilaginous pterygiophores proceeded in both species anteriorly and posteriorly (Figure 2, Table 6). Rays developed after the addition of the cartilaginous pterygiophores. The posteriormost interhaemal space number 26 of C. hippurus had one to three cartilaginous pterygiophores at 7.3-8.3 mm SL (Table 6). The posteriormost inter- haemal space number 28 of C. equiselis had cartilaginous pterygiophores at 7 mm SL (Table 6). All specimens of C. hippurus > 9.5 mm SL and all those of C. equiselis > 8.5 mm SL had some pterygiophores in their anteriormost interhaemal space number 14. The size at which adult counts were reached for the first interhaemal space was not determined. Ossification of the cartilaginous pterygiophores first occurred in the area where the cartilaginous pterygiophores first appeared and proceeded in the same direction as cartilage development (Fig- ure 2). Ossifying anal fin pterygiophores were first seen at 8.8 mm SL in C. equiselis and at 9.7 mm SL in C. hippurus in the 16th-19th and 16th-25th interhaemal spaces (Table 6), and concurrently with ossifying dorsal fin pterygiophores. The posteriormost interhaemal space number 28 of C. equiselis had ossifying pterygiophores at 8.9 mm SL and space number 26 of C. hippurus had them at 10.2 mm SL (Table 6). All specimens of C. equiselis > 9.4 mm SL and all specimens of C. hippurus > 11 mm SL had some ossifying pterygiophores in the anteriormost interhaemal space number 14, or rarely space number 15 (Table 6). The anteriormost anal fin pterygiophore was ossifying in C. equiselis at 14.9-22 mm SL and in POTTHOFF: DEVELOPMENT AND STRUCTURE OF FINS IN CORYPHAENA in 1 CO Is £ c/l 0) tn 3 C E 3 1 ^ ^ Tt r^ r^ cJi u*) £ o <3J d 0) Q. 0) cy] 02 O CO 0) O CO 2 o X3 OS E 3 a. a --m(y>mof~- cs X hi JJ- C be ^^ « o c >< "§ CD *r^ .CO 1 ^^^ CO CO CD C3> 1 c\l CM C\J CN &> 00 3 t/) o 3 cr CD £ CD Q. O CM C O tn ^ o. "tn ^o o O b ? E to 3 CO 55" o i; uS in S 3 -^ ^- ^.^ ^^ C\i CM CO Q. B CL-^tL,^^CD'cC> CD .CO 3 "tn OJ CM C\J uS en c/l o ^ J)* D- (/) ft) o o i s 01 ^^ to CO ~ ^ T3 o >< a5 ■^ ^ -^ m , tn 5 Q. en o QJ „ CJ CM CM c Q. r ^ CD CD CD CD CD CD o C CM C\J CM CM CM CM CM 01 c q: CO CO uS O) d CM CM CM *3 (T3 C ^ (0 tn o V £ £ O i CO LO ^ ^ J3 tn •K }■ 01 tn ro E o CD uS 4 4 -*-> Q. o. tn tT3 tn c ^ 05 o E o cc 2 CO 1 fCto CD 4^ cj) h- 4 4 § B .1 ::^CMOtDio-- ^ ^ -- en v 0) CO 6 a ■o o CO to (U 0) o t- 1- O O "S CC > Q.Q. 0) O O Q 1 C3)cn 1 CO B B «d ^5 minininininunm Q. Q. O O Cd iricbh-ccieDO.-CM zz J I < . T V T BQ -1 != CDCDCDCDCDCDCDCD ^ E 4iricbr^cocnci'- C. hippurus at 17.2-30 mm SL. The development of individual anal fin pterygiophores was similar to that of the dorsal fin pterygiophores. Each anal fin pterygiophore of both species had two rays; one ray was in a serial association and the preceding ray was in a secondary association. The posteriormost anal ray lacked a secondary association with a pterygiophore and the anterior- most anal ray lacked a serial association (Figure 16). Exceptions were common with the anterior- most pterygiophore and rays. Many specimens of both species had very small first pterygiophores or even vestiges. In a few instances in both species the anteriormost first pterygiophore was com- pletely or partially fused to the second pterygio- phore. The normal number of anal rays associated with the first pterygiophore was two, but for both species one or three rays also were found. The anteriormost anal ray was either normal as in Figure 16, very small, or a vestige. As in the dorsal fin, the vestige was either single left or right, or paired. A distal radial was present between the base of each fin ray almost for the entire anal fin. It developed and ossified from the pterygiophore cartilage. Only the anteriormost anal fin ray sometimes did not have a distal radial between its base (Table 7). Only 1 C. hippurus out of 49 had two anteriormost rays without distal radials. When the anteriormost ray had a distal radial, it was either serially or secondarily associated with the first pterygiophore. When the association was serial, the anteriormost pterygiophore had only one ray; when it was secondary, it had two rays. It is possible that, when the association was sec- ondary, the distal radial of the first fin ray was in actuality a vestigial pterygiophore. The specimen in Figure 16 did not have a distal radial for the anteriormost ray. The absence or presence of distal radials for the anteriormost anal fin ray was not related to the number of fin rays that were Table 7. — Number (adult count) of anteriormost anal fin rays without distal radials and number of anal fin rays associated secondarily and serially with the anteriormost anal fin pterygio- phore in 49 Coryphaena hippurus (41.0-176 mm SL) and 33 C. equiselis (74.1-172, 314 mm SL). Number of anterior- most anal fin rays Item Species 1 2 3 Without distal C. hippurus 24 24 1 radials C. equiselis 10 23 Associated witfi the C. hippurus 3 40 6 anteriormost anal C. equiselis 3 29 1 pterygiophore 293 FISHERY BULLETIN: VOL. 78, NO. 2 Figure 16. — Anteriormost anal fin pterygiophore with secondarily and serially associated rays from a 230 mm SL Coryphaena equiselis. Left: left lateral view; right: anterior view, serially associated ray has been moved to the left of the proximal radial. » Symbols: D, distal radial of serially associated ray; P, proximal radial; Rj, secondarily associated ray; R, serially associated ray. associated with the anteriormost anal pterygio- phore (Table 7). In both species either one, two, or three rays were associated with the first pterygio- phore. In both species the anteriormost distal radial (which was either between the base of the first or second anal fin ray) was a single piece of bone (Figure 16). The second distal radial (which was either between the base of the second or third anal fin ray) consisted of two pieces of bone, as shown for the dorsal pterygiophores in Figures 7 294 POTTHOFF: DEVELOPMENT AND STRUCTURE OF FINS IN CORYPHAENA and 8. The two pieces of the second distal radial were sometimes partially fused; and in a few rare the second distal radial was one piece of cases, Table 8. — Adult caudal fin ray counts for 117 Coryphaena hippurus (19.6-172 mm SL) and 97 C. equiselis (19.6-230 mm SL). Symbols: USCR, upper secondary caudal rays; LSCR, lower secondary caudal rays; PCR, principal caudal rays. Number of specimens Total fin ray count USCR * PCR - LSCR C hippurus C equiselis 37 38 39 39 40 40 41 41 42 43 44 45 10+17-10 10+17*11 10+17+12 11+17+11 11+17+12 12+17+11 11+17+13 12+17+12 12+17+13 13+17+13 13*17+14 14+17*14 2 10 12 3 36 23 26 2 3 1 1 2 28 6 17 32 8 2 bone. All following distal radials of the anal fin were two pieces of bone in both species. The posteriormost anal fin ray consisted of two closely approximated rays with one distal radial. Caudal Fin The two species differed little in ray counts on fully developed caudal fins. Coryphaena hippurus had 38-45 (x = 41.4) caudal rays and C. equiselis had 37-44 (x = 41.1) (Table 8). Coryphaena hip- purus tended to have an equal number of upper and lower secondary caudal rays whereas C equiselis tended to have one or two more lower than upper secondary caudal rays. Adult caudal ray counts for C. hippurus were obtained between 15.6 and 19.6 mm SL and for C. equiselis between 11.6 and 12.5 mm SL (Tables 9, 10). A procur- TaBLE 9. — Caudal fin ray development in 201 Coryphaena hippurus (5.0 mm NL-172 mm SL) and 138 C. equiselis (6.5 mm NL- 230 mm SL). Symbols: SCR, secondary caudal rays; PCR, principal caudal rays. Length mm NL or SL 5.5 6.5 7.5 8.5 9.5 4.6- 5.6- 6.6- 7.6- 8.6- 96-10.5 10.6-11 5 11,6-12.5 12-6-13.5 13.6-14.5 14.6-15.5 15.6-16.5 16.6-17.5 17.6-18.5 18.6-19.5 >19.5 Coryphaena hippurus Coryphaena equiselis Upper Lower Total fl Range n ray count N Upper Lower Total fii Range 1 ray count X Sx SCR PCR PCR SCR SCR PCR PCR SCR N 2-4 2-5 4- 9 7.0 1,5 3 — — — — — — — 4-8 5-8 9-16 11.3 16 4 6 6 12 — — 1 1-8 2-8 0- 1 3-17 12,9 1,3 10 9 8 1- 2 18-19 18.5 0.5 2 9 8 1- 2 18-19 18.8 0,2 6 0- 3 9 8 2- 4 19-24 21.2 12 5 0- 3 9 8 2- 4 19-24 19.8 1,1 5 3- 8 9 8 4- 8 24-33 29.0 1.9 4 1- 4 9 8 2- 5 21-26 233 0,7 7 5- 9 9 8 6- 9 28-35 31.0 1.3 5 0- 4 9 8 2- 5 19-26 24.7 1.0 7 7- 9 9 8 8- 9 32-35 33.6 0.4 7 4- 5 9 8 5- 7 26-29 27.7 0.9 3 10-11 9 8 11-12 38-40 38.5 0.5 4 5- 6 9 8 6 28-29 286 0.2 5 11 9 8 12 40 — — 1 7- 8 9 8 8- 9 32-34 33.3 0.3 6 11 9 8 12 40 — — 1 7- 9 9 8 8-10 32-36 347 0.6 6 12 9 8 12 41 — — 1 9-11 9 8 10-12 36-40 37.5 0.7 6 11 9 8 11-13 39-41 40.0 1.0 2 9-11 9 8 9-11 35-39 366 0.7 5 11 9 8 11-13 39-41 40.0 0,6 3 9-11 9 8 9-11 35-39 38.2 0.8 5 11-12 9 8 12-13 40-42 40.5 0.5 4 8-12 9 8 12 37-41 395 0.6 6 11 9 8 12 40 — — 1 10-14 9 8 11-14 38-45 41.4 0.1 117 10-13 9 8 10-14 37-44 41 1 0.1 97 Table lO. — Length (in millimeters NL or SL) at which parts of the caudal complex first appear in cartilage and then ossify in 41 Coryphaena hippurus (5.0 mm NL-110 mm SL) and 39 C. equiselis (6.5-85 mm SL). "First appearance in cartilage" does not pertaiin to all specimens of that size but only indicates a first appearance. Symbol: Pu, preural centrum. Part Neural spine. PUj Specialized neu- ral arch, PUj Large uroneural Small uroneural Epurals Urostyle Pu J centrum PU3 centrum Haemal spine. PUj Haemal spine, Pu^ Parhypural Hypural 1 Hypural 2 Hypural 3 Hypural 4 Hypural 5 Coryphaena hippurus First appearance in cartilage First evidence of ossification 7.4 7.4 7.4-8.0 6.0 5.0 <5.0 <5.0 <5.0 <5.0 6.0 8.1-9.5 9.5 8.0- Ossifying in all specimens Completely fused 119 11 9 11.9 10.6 11.9 11,9 11.9 14,6 14.6 80 11.9 9,5 11.9 95 11.9 9,5 11.9 9,5 11.9 9,5 11.9 9.5 11.9 9.5 11.9 95 11.9 95 11.9 11.9 11.9 750- 40.0- 850 47.0 106 106.0 Coryphaena equiselis First appearance in cartilage First evidence of ossification Ossifying in all specimens > 6.5 but 7.6 ■6.5 but < 7.6 •6.5but 76 •:^6.5 <6.5 <6.5 <6.5 <6.5 <6.5 65 7.6 7.6 9.5 ■6.5but< 7.6 (9.5?) 10.8 10.8 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6 9.5 8 1 9 5 7.6 10.8 10.8 7.6 9.5 9,5 9.5 9.5 7.6 7,6 7.6 7.6 7.6 9.5 Completely fused 75.0-800 34.0-39.0 690 69.0 295 FISHERY BULLETIN: VOL. 78, NO. 2 rent spur (Johnson 1975) was not observed in either species. The caudal rays first developed in both species from the midline between hypurals 2 and 3 in preflexion larvae (Figure 17). Rays were added in a posterior and anterior direction (Figure 18). After complete notochord flexure the secondary caudal rays were added in an anterior direction. For equal-sized specimens from 6.5 mm NL to 19.5 mm SL, C. hippurus had fewer caudal fin rays than C. equiselis (Table 9). Fnfid I- 0.5 mm Figure 17. — Caudal complex of Coryphaena hippurus, 5.0 mm NL. Symbols: Fnfld, finfold; Hs, haemal spine; Hy, hypural; Nc, notochord; PCR, principal caudal ray; Ph, parhypural. Stippled, cartilage; darkened, ossifying bones or rays. Caudal Fin Supports The caudal fin rays of Coryphaena spp. were supported by some of the bones of the caudal com- plex. Three posteriormost centra were involved in this support. In 2 out of 97 C. equiselis the caudal fin rays were also supported by a fourth centrum. This variation was not observed in C. hippurus. Supporting bones of the caudal complex con- sisted of three centra (urostyle and preural centra numbers 2 and 3), one neural spine, one special- ized neural arch, two autogenous haemal spines, one autogenous parhypural bone, five autogenous hypural bones, two paired uroneural bones, and two epural bones. These parts were seen during 0.5 mm Figure 18. — Caudal complex of Coryphaena hippurus, 6.0 mm NL. For explanation of symbols, see Figure 17. Stippled, cartilage; darkened, ossifying bones or rays. development, but not all the parts are readily discerned in the adults due to ontogenetic fusion. The species did not differ in the anatomy of the caudal complex, but they differed in the size at which parts appeared and ossified. The 6.5 mm NL C. equiselis was at the same stage of caudal development as a 6.5 mm NL C. hippurus. From 7.6 to 16 mm SL, C. equiselis was more advanced. Specimens >16 mm SL of both species had the caudal complex equally ossified for the same lengths, but epural, uroneural, and hypural fu- sions occurred at shorter lengths in C. equiselis. Development of the caudal complex of C. hip- purus is described here rather than C. equiselis because small specimens were not available for C. equiselis. Mostof the illustrations of the caudal complex are of C. equiselis because they were drawn before it was apparent that C. equiselis < 7.6 mm were not available. Because both species had identical caudal complex anatomy, no draw- ings of C. hippurus' caudal complex were made for specimens >7.6 mm SL. At 5 mm NL, C. hippurus had a straight notochord. Hypurals 1 to 3, the parhypural, and the haemal spine of the future preural centrum 2 were present in cartilage and 2+3 principal caudal rays were counted (Figure 17). At 6 mm NL, hypural 4 and an additional cartilaginous haemal spine of the future preural centrum 3 were present (Figure 18). Notochord flexion in C. hip- 296 POTTHOFF: DEVELOPMENT AND STRUCTURE OF FINS IN CORYPHAENA purus was between 7 mm NL and 7.5 mm SL, and in C. equiselis between 6.5 mm NL and 7.6 mm SL (Figure 19). During the flexion stage of some C. hippurus the neural spine of preural cen- trum 3, the speciaHzed neural arch of preural centrum 2, and the two epurals began to develop from cartilage (Figure 19). Hypural 5 was first seen in cartilage at 8.1 mm SL. The two paired uroneurals did not develop from cartilage — in C. hippurus the larger, more ventrally and ante- ognized only one epural for adult C. hippurus. In Coryphaena spp., hypurals 1 and 2 and hypurals 3 and 4 fused to a dorsal and ventral hypural plate (Figures 20-23, Table 10). During fusion paired bony ventrolateral and dorsolateral articular pro- jections formed on the ventral edge of hypural 3 and on the dorsal edge of hypural 2. These projections became the articulatory surfaces be- tween the dorsal and ventral hypural plates (Fig- ures 20-23). The two hypural plates of Cory- EpS-.— Un Figure 19. — Supporting bones of the caudal complex of Coryphaena hippurus (right) 7.6 mm SL and C. equiselis (left) 7.6 mm SL. Symbols: Eps, epurals; Hs, haemal spine; Hy, hypural; "Na", specialized neural arch; Ns, neural spine; Ph, parhypural; Pu, preural centrum; Un, uroneural; Ur, urostyle. Stippled, cartilage; darkened, ossifying bones; stippled darkened areas are cartilage just beginning to ossify. riorly located pair was seen at 8-10.6 mm SL, and the smaller, more dorsally and posteriorly located pair was seen at 11.9 mm SL. Development of the two paired uroneurals occurred at a smaller size in C. equiselis (Figure 19 left. Table 10). The smaller uroneural pair gradually fused to the outside of the larger uroneural pair in both species. This fusion was completed between 75 and 85 mm SL for C. hippurus and between 75 and 80 mm SL for C. equiselis (Table 10). Monod (1968) recog- nized only one uroneural (stegural) pair in adult C. hippurus. Ossification of the cartilage bones in the caudal complex of C hippurus began with the urostyle at 8 mm SL. Last to ossify at 14.6 mm SL were the two epurals. The ossification sequence of all hypural bones is shown in Table 10. The epurals of C. hippurus developed and fused in the same manner as those of C. equiselis, although development and fusion were always at a smaller size for C. equi- selis (Figures 19-23, Table 10). Monod (1968) rec- Eps 0.5 mm FIGURE 20.— Supporting bones of the caudal complex of Cory- phaena equiselis, 11.0 mm SL. Symbols: Uns, uroneurals. For explanation of other symbols, see Figure 19. Stippled, articular cartilage; darkened, bone. 297 FISHERY BULLETIN: VOL. 78, NO. 2 scR PCR Figure 21. — Caudal complex of Coryphaena equiselis, 15.9 mm SL. Symbols: PCR, principal caudal rays; SCR, secondary caudal rays. Stippled, articular cartilage; darkened, bone. autogenous (Figures 23, 24). These were two haemal spines, a parhypural, a ventral and dorsal hypural plate, hypural 5, a uroneural pair (fused from two pairs), and an epural (fused from two). Nonautogenous bones were the specialized neural arch and one neural spine. The relationship of the urostyle with the uroneural pair and hypural 5 is shown in Figure 24. Articular cartilage was pres- ent on all distal parts of the hypural complex posterior to preural centrum 4 (Figure 22). The parhypural and hypurals 1-5 supported the principal caudal rays. The distribution of principal caudal rays on the various hypural bones can only be seen in larvae and smaller juveniles of both species before hypural fusion (Table 11). There was no difference in distribution of principal caudal rays between the two species. Uns Hy3&4 Art Hyi*2 Figure 22. — Supporting bones of the caudal complex of Coryphaena equiselis, 55.5 mm SL. Symbols: Art, articular projection. For explanation of other symbols, see Figures 19, 20. Stippled, articular cartilage; darkened, bone. phaena spp. remained autogenous in the adults, but were closely articulated with the ventroposte- rior edge of the urostyle. During development of the hypural complex bones a small hypurapophysis (Lundberg and Baskin 1969) was observed on hypural 1 in both species. It appeared before hypural fusion, but could not be illustrated in the lateral view. Dis- articulation of adult caudal skeletons of both species of Coryphaena revealed the presence of the hypurapophysis. The hypurapophysis articulated with the urostyle just dorsad of the parhypur- apophysis (Nursall 1963). In the adults of Coryphaena spp., most bones of the hypural complex were closely articulated, but 298 The anatomy and development of the caudal complex of Coryphaena spp. had similarities and dissimilarities with other fish. The hypurapophy- sis observed in Coryphaena spp. was noted in such fish as siluriform catfish (Lundberg and Baskin 1969) and adult sea bream, Archosargus rhom- boidalis (Houde and Potthoff 1976). The hypura- pophysis was not observed in the blackfin tuna, Thunnus atlanticus (Potthoff 1975). In the Coryphaenidae and other percoid fishes such as Apogonidae (Fraser 1972), A. rhomboi- dalis (Houde and Potthoff 1976), Carangidae ( Ahl- strom and Ball 1954; Berry 1969), Haemulon plumieri (Saksena and Richards 1975), and some Scombridae (Conrad 1938; Mago Leccia 1958), the POTTHOFF: DEVELOPMENT AND STRUCTURE OF FINS IN CORYPHAENA Figure 23. — Caudal complex of a Coryphaena equiselis, 230 mm SL. For explanation of symbols, see Figure 21. Stippled, articular cartilage; darkened, bone. Table U. — Distribution of principal caudal rays on the hypurals in 136 Coryphaena hippurus (8.0-53 mm SL) (C. h.) and 75 C. equiselis (7.0-52 mm SL) (C. e.). Number of principal caudal rays 1 2 3 4 5 Part C.h. C. a C.h. C. e. C.h. C. e C.h. C. e. C. h. C. e. Parhypural 51 24 85 51 Hypural 1 52 22 80 51 4 2 Hypural 2 39 32 97 43 Hypural 3 28 27 97 45 11 3 Hypural 4 6 72 36 58 39 Hypural 5 37 16 98 58 1 1 epurals were autogenous. In part of the Scom- bridae (Fierstine and Walters 1968; Monod 1968; Patterson 1968; Collette and Chao 1975; Potthoff 1975) the anteriormost epural is secondarily fused to the specialized neural arch of preural centrum 2. Based on the epurals, Coryphaena spp. is advanced because epural numbers are reduced from 3 to 2 and fused to 1 (Patterson 1968; Fraser 1972). The haemal spines of preural centrum 2 and 3 were autogenous in Coryphaena spp. This state is considered basic because advanced percoids have these spines secondarily fused to the centra ( Fraser 1972). Fusion of these haemal spines occurs in T. atlanticus (Potthoff 1975), and some apogonids (Fraser 1972). The two prezygapophyses of the urostyle (Fig- ure 24) oi Coryphaena spp. are true prezygapophy- ses; whereas in T. atlanticus and other Thunnini and Sardini (Collette and Chao 1975; Potthoff 1975) the prezygapophyses of the urostyle repre- sent the pair of uroneurals which have fused to the urostyle during development. Articular cartilage was present in Coryphaena spp. on the caudal complex on all parts distally inclusive of preural centrum 3. No articular car- tilage was observed anterior to this centrum. Articular cartilage was observed in scombrids by Fierstine and Walters (1968), in T. atlanticus by Potthoff (1975), and in A. rhomboidalis by Houde and Potthoff (1975). The absence of artic- ular cartilage in the caudal complex drawings of 299 FISHERY BULLETIN: VOL. 78, NO. 2 5 mm FIGURE 24.— Urostyle of a Coryphaena equiselis, 330 mm SL, with disarticulated uroneurals and hypural 5. Dashed lines with arrows point towards place of articulation. Symbols: AStr, anterior strut of urostyle; Hy, hypural; Pr, prezygapophysis; PStr, posterior strut of urostyle; Uns, uroneurals; Ur, urostyle. The articular cartilage is not shown on hypural 5 because of the boiling and drying method of preparation. apogonids (Fraser 1972) is probably an oversight by the author since he used cleared and stained material. The lack of articular cartilage in most of the drawings of caudal complexes by Monod (1968) can probably be attributed to the method of skeletal preparation, e.g., boiling and subse- quent drying. 300 Autogenous dorsal and ventral hypural plates were observed in adult Coryphaena spp. The fusion of individual hypural bones was considered advanced by Fraser (1972). Even more advanced is the fusion of all hypural bones to one hypural plate and the fusion of this plate to the urostyle as in scombrids (Fierstine and Walters 1968; Pott- hoffl975). The formation of articulatory projections of membranous origin during ontogeny at the mid- line of the caudal complex between the dorsal and ventral hypural plates was observed in Cory- phaena spp. (Figures 20-22) as well as in Scom- brolabrax heterolepis (Potthoff et al. 1980), but not in T. atlanticus (Potthoff 1975). Both species of Coryphaena had two pairs of uroneurals. The smaller posterior pair gradually moved anteriorly during development and fused to the outsides of the larger anterior pair, until only one pair could be recognized in adults. Fraser (1972) contended that the loss of the posterior pair of uroneurals constituted an evolutionary advance. He did not completely rule out fusion, although he had no evidence for it. There are fishes such as the scombrids which only develop one pair of uroneurals (Potthoff 1975). Loss or fusion of uroneurals can be ascertained through the examination of developmental series. Pectoral Fin and Supports The following description is based upon juve- niles > 13 mm SL of both Coryphaena species with adult counts of 19-21 rays. These counts were obtained between 19 and 13 mm SL in C. equiselis and between 11 and 13 mm SL in C. hippurus. Individual differences in count