aiB WKasBmuKKm SH l! A^-Sfs3 Fidr U.S. Department of Commerce Volume 96 Number 1 January 1998 Fishery Bulletin U.S. Department of Commerce William M. Daley Secretary National Oceanic and Atmospheric Administration D. James Baker Under Secretary for Oceans and Atmosphere National Marine Fisheries Service Rolland Schmitten Assistant Administrator for Fisheries The Fishery Bulletin (ISSN 0090-0656) is published quarterly by the Scientific Publications Office, National Marine Fish- eries Service, NOAA, 7600 Sand Point Way NE, BIN C15700, Seattle, WA 98115- 0070. Periodicals postage is paid at Se- attle, WA, and at additional mailing of- fices. POSTMASTER: Send address changes for subscriptions to Fishery Bul- letin, Superintendent of Documents, Attn.: Chief, Mail List Branch, Mail Stop SSOM, Washington, DC 20402-9373. Although the contents of this publica- tion have not been copyrighted and may be reprinted entirely, reference to source is appreciated. The Secretary of Commerce has deter- mined that the publication of this peri- odical is necessary according to law for the transaction of public business of this De- partment. Use of funds for printing of this periodical has been approved by the Di- rector of the Office of Management and Budget. For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. Subscrip- tion price per year: $34.00 domestic and $42.50 foreign. Cost per single issue: $13.00 doemestic and $16.25 foreign. See back for order form. Scientific Editor Dr. John B. Pearce Editorial Assistant Laura Gamer Northeast Fisheries Science Center National Marine Fisheries Service, NOAA 1 66 Water Street Woods Ftole, Massachusetts 02543-1097 Editorial Committee Dr. Andrew E. Dizon National Marine Fisheries Service Dr. Harlyn O. Halvorson University of Massachusetts, Boston Dr. Ronald W. Hardy University of Idaho, Hagerman Dr. Richard D. Methot National Marine Fisheries Service Dr. Theodore W. Pietsch University of Washington, Seattle Dr. Joseph E. Powers National Marine Fisheries Service Dr. Harald Rosenthal Universitat Kiel, Germany Dr. Fredric M. Serchuk National Marine Fisheries Service Managing Editor Sharyn Matriotti National Marine Fisheries Service Scientific Publications Office 7600 Sand Point Way NE, BIN C 1 5700 Seattle, Washington 98115-0070 The Fishery Bulletin carries original research reports and technical notes on investiga- tions 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. Begin- ning 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. 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. i U.S. Department of Commerce Seattle, Washington Volume 96 Number 1 January 1998 Fisheiy Bulletin The National Marine Fisheries Service (NMFS) does not approve, recommend, or endorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to NMFS, or to this publication furnished by NMFS, in any advertising or sales promotion which would indicate or imply that NMFS approves, recommends, or endorses any proprietary product or proprietary material mentioned herein, or which has as its purpose an intent to cause directly or indirectly the adver- tised product to be used or purchased because of this NMFS publication. Contents 1-1 82 Munroe, Thomas A. Systematics and ecology of tonguefishes of the genus Symphurus (Cynoglossidae: Pleuronectiformes) from the western Atlantic Ocean 3 Taxonomic history 5 Materials and methods 2 1 Artificial key to western Atlantic Symphurus Species accounts 26 Symphurus nebulosus (Goode and Bean, 1 883) 30 Symphurus arawak Robins and Randall, 1 965 34 Symphurus rhytisma Bohlke, 1 96 1 38 Symphurus billykrietei, new species 45 Symphurus stigmosus, new species 50 Symphurus ginsburgi Menezes and Benvegnu, 1976 54 Symphurus pelicanus Ginsburg, 1 95 1 58 Symphurus pusillus (Goode and Bean, 1 885) 62 Symphurus marginatus (Goode and Bean, 1 886) 66 Symphurus piger (Goode and Bean, 1 886) 7 1 Symphurus trewavasae Chabanaud, 1 948 75 Symphurus kyaropterygium Menezes and Benvegnu, 1976 78 Symphurus minor Ginsburg, 1 95 1 82 Symphurus parvus Ginsburg, 1951 87 Symphurus ommaspilus Bohlke, 1961 91 Symphurus diomedeanus (Goode and Bean, 1885) Fishery Bulletin 96(1 ), 1998 99 Symphurus jenynsi Evermann and Kendall, 1907 104 Symphurus plagiusa (Linnaeus, 1 766) 1 16 Symphurus urospilus Ginsburg, 1 95 1 120 Symphurus caribbeanus Munroe, 1991 125 Symphurus civitatium Ginsburg, 1 95 1 131 Symphurus oculellus Munroe, 1991 136 Symphurus plagusia (Schneider, in Bloch and Schneider, 1801 141 Symphurus tessellatus (Quoy and Gaimard, 1824) 1 46 Discussion 1 57 Acknowledgments 1 58 Literature cited 1 68 Appendix 1 84 Fishery Bulletin subscription form Systematics and ecology of tonguefishes of the genus Symphurus (Cynoglossidae: Pleuronectiformes) from the western Atlantic Ocean* Thomas A. JVSumroe National Marine Fisheries Service Systematics Laboratory, MRC-1 53 National Museum of Natural History Washington, D.C. 20560-0153 E-mail address: munroet@nmnh. si.edu Abstract.— The 24 species of the cynoglossid genus Symphurus Rafinesque, 1810 occurring in the western Atlantic Ocean are revised. Symphurus species are found from the southern Scotian Shelf (ca. 45°N ) southward to central Argentina (ca. 45°S). These small to medium-size, left- sided flatfishes inhabit diverse substrates ranging from shallow estuarine habitats to deepwater substrates on the outer con- tinental shelf and upper continental slope. Thirty-four nominal species of symphurine tonguefishes have been described previ- ously from this area. Twenty-four, includ- ing two new species, are considered valid: S. arawak Robins and Randall, 1965, in shallow sandy habitats adjacent to coral reefs from Alligator Reef, Florida, through the Caribbean Sea to Colombia; S. billy- krietei, new species, on mud bottoms of the outer continental shelf from the southern Scotian Shelf (ca. 45°N ) to the central Gulf of Mexico, differing from other species in meristic and morphometric characters, black peritoneum, relatively small eye without pupillary operculum, dark brown stripe covering fin rays and connecting membranes on basal one-third of dorsal and anal fins, and small, darkly pigmented spot on scaly portion of caudal fin; S. caribbeanus Munroe, 1991, on sandy and silty substrates in estuarine and neritic waters in the Caribbean, including the Greater Antilles and coastal waters off Central America to Colombia; S. civitatium Ginsburg, 1951, on sand substrates in nearshore and neritic waters from Cape Hatteras, North Carolina, to the Yucatan Peninsula, and rarely at Bermuda; S. diomedeanus (Goode and Bean, 1885), widespread on continental shelf calcare- ous muds and sands from Cape Hatteras, North Carolina, through the Gulf of Mexico and Caribbean Sea, south to Uruguay; S. ginsburgi Menezes and Benvegnu, 1976, on outer continental shelf mud bottoms from about Cabo Frio, Brazil (ca. 23°S), to Maldonado, Uruguay (ca. 35°S )\S. jenynsi Evermann and Kendall, 1907, on mud bot- toms in neritic waters from near Cabo Frio, Brazil (ca. 22°S), to northern Argentina; S. kyaropterygium Menezes and Ben- vegnu, 1976, on the inner continental shelf from Baia da Ilha Grande (ca. 23°S) to Rio Grande do Sul (ca. 31°S); S. marginatus (Goode and Bean, 1886), on outer continen- tal shelf and upper continental slope mud bottoms from southern New Jersey (ca. 40°N) to central Brazil (21°34'S); S. minor Ginsburg, 1951, primarily on live-bottom habitats off the southeastern United States and eastern Gulf of Mexico west- ward to about DeSoto Submarine Canyon Manuscript accepted 8 September 1997. Fishery Bulletin 96:1-182 (1998). (ca. 29°87'W), rarely off southern Scotian Shelf to ca. 44°N ; S. nebulosus ( Goode and Bean, 1883), on the outer continental shelf and upper continental slope from near Long Island, New York ( ca. 40°48'N ), to the Blake Plateau off Fort Lauderdale, Florida (ca. 26°28'N); S. oculellus Munroe, 1991, on the inner continental shelf on mud bot- toms from Guyana (57°W) to northeastern Brazil (2°S, 40°W); S. ommaspilus Bohlke, 1961, on shallow-water sandy substrates in the Caribbean Sea from the Bahamas, Lesser Antilles, and Belize; S. parvus Ginsburg, 1951, on inner continental shelf mud bottoms from off Cape Lookout, North Carolina (ca. 34°23'N), through the Gulf of Mexico and Caribbean Sea to Venezu- ela; S. pelicanus Ginsburg, 1951, on conti- nental shelf soft mud bottoms from the western and central Gulf of Mexico and Caribbean Sea to Trinidad; S. piger (Goode and Bean, 1886), on the outer continental shelf and upper continental slope from southern Florida (ca. 30°N), the Gulf of Mexico, and throughout the Caribbean Sea to Suriname (ca. 7°N, 53'W); S. plagiusa (Linnaeus, 1766), on soft mud and silt sub- strates in estuarine and neritic habitats from Long Island Sound to the Campeche Peninsula, also the Bahamas and Cuba; S. plagusia (Schneider, in Bloch and Schneider, 1801), on sand and silt sub- strates in estuarine and neritic habitats from the Greater Antilles and Central America to about Rio de Janeiro, Brazil; S. pusillus (Goode and Bean, 1885), on the outer continental shelf from off Long Is- land, New York (ca. 40°N), to DeSoto Sub- marine Canyon in the eastern Gulf of Mexico (ca. 29°87'W); S. rhytisma Bohlke, 1961, from the Bahamas, Belize, Curasao, and perhaps Brazil; S. stigmosus, new spe- cies, in deepwater areas of the Straits of Florida and Gulf Stream off southern Florida and in the Caribbean off Yucatan, Serrana Bank, and Dominica, differing from other species in its combination of meristic and morphometric features, black peritoneum, relatively large rounded con- tiguous eyes without pupillary operculum, and dorsal and anal fins with combination of 1) dark brown longitudinal stripe on basal one-third which covers fin rays and intervening membranes and 2) usually a series of distinct, darkly pigmented blotches alternating with unpigmented areas of somewhat larger size on posterior two-thirds of fins; S. tessellatus ( Quoy and Gaimard, 1824), on sandy and silty sub- strates in estuarine and neritic habitats from the Greater Antilles and Central America to northern Argentina; S. trewava- sae Chabanaud, 1948, on continental shelf mud bottoms from about Cabo Frio, Brazil (ca. 22°53'S), to central Argentina (ca. 45°S); and S. urospilus Ginsburg, 1951, on live-bottom habitats on the inner continen- tal shelf from about Cape Hatteras, North Carolina, to Yucatan Peninsula, and Cuba. Seven species are synonymized. Achirus ornata Lacepede, 1802, and Plagusia brasiliensis Agassiz, in Spix and Agassiz, 1831, are synonyms of Symphurus tessel- latus (Quoy and Gaimard, 1824); Plagusia fasciata DeKay, 1842, is a synonym of Symphurus plagiusa (Linnaeus, 1766); Symphurus bergi Thompson, 1916, is a syn- onym of S. jenynsi Evermann and Kendall; Symphurus sumptuosus Chabanaud, 1948, and S. pterospilotus Ginsburg, 1951, are syn- onyms ofS. diomedeanus (Goode and Bean, 1885); and Symphurus meridionalis Lema and Oliveira, 1977, is a synonym of S. jenynsi Evermann and Kendall. Descriptions, differ- ential diagnoses, an artificial key, and sum- maries of ecological information are provided for 24 species of western Atlantic sym- phurine tonguefishes. * Contribution number 2081 of the Virginia Institute of Marine Science, Gloucester Point, Virginia 23062. 2 Fishery Bulletin 96( 1 ), 1998 Symphurine tonguefishes belong to one genus ( Symphurus Rafinesque, 1810) of approximately 75 species of small to medium-size, left-sided flatfishes (Munroe, 1992). Superficially, these flatfishes are recognized in having a small mouth with strongly curved and toothed jaws on the blind side, in having the caudal, dorsal, and anal fins united, and in hav- ing lost pectoral fins, lateral line, and left-side pel- vic fin. The right pelvic fin has a reduced number of fin rays compared with that for other flatfishes, com- prising only four rays, and it is situated along the midline of the body. Symphurine tonguefishes are the most speciose and widely distributed members of the Cynogloss- idae, which comprises approximately three genera and some 125 species. Within the Cynoglossidae, Symphurus forms a monophyletic taxon that is the sister group of the Cynoglossus-Paraplagusia lineage (Chapleau, 1988). Synapomorphies diagnosing this genus (Chapleau, 1988) include a greatly reduced, ocular-side, lateral ethmoid lacking an osseous at- tachment either to the interorbital complex dorsally or to the vomer ventrally, the lateral ethmoid with a long posterodorsal arm in contact with the anterior process of the supraoccipital, fusion between ocular- and blind-side anterior arms of the frontals, replace- ment of the anterior portion of the supraoccipital bone by a cranial fontanelle (a character perhaps unique among flatfishes), all proximal radials anterior to the first hemal spine of the caudal region of the body equally long and in contact with this spine, lack of a lateral line canal on the ocular side (unique among flatfishes), and only a single pterygiophore inserted in the first interneural space (Munroe, 1992). Species of Symphurus have been reported from all temperate and tropical oceans (Chabanaud, 1955a, 1955b, 1956; Mahadeva, 1956; Ginsburg, 1951; Menezes and Benvegmi, 1976; Munroe, 1992) but are the only cynoglossids found in the New World. In fact, it is in these waters where the greatest diversity of species of Symphurus is found, with approximately 30 nominal species recorded from both coasts of the Americas (Ginsburg, 1951; Menezes and Benvegnu, 1976; Munroe, 1992). Compared with other flatfishes, Symphurus is the most diverse genus of flatfish oc- curring in the New World, and its species occupy the greatest variety of habitats within this region. In the western Atlantic Ocean, tonguefishes occur from the Scotian Shelf (ca. 45°N; Scott and Scott, 1988) southward to central Argentina (ca. 45°S, Evermann and Kendall, 1907; Menezes and Ben- vegnu, 1976; Lazzaro, 1973, 1977; Lema et al., 1980; this study). Throughout this region, Symphurus spe- cies occur in diverse habitats including such shal- low-water areas as muddy and silty substrates in turbid estuaries, sandy patches in seagrass beds in clear tropical waters, and sand substrates on, or ad- jacent to, coral reefs. In addition, species of Symphurus also inhabit a variety of different sub- strates in moderate depths on the continental shelf, and some species live even on deepwater substrates located on the outer continental shelf and upper con- tinental slope. In fact, S. nebulosus and S. margi- natus, collected as deep as 810 m and 750 m, respec- tively (see below), are among the deepest-dwelling flatfishes in the western Atlantic. In some demersal fish communities, especially those on soft-bottom habitats in the western Atlan- tic Ocean (Wenner and Sedberry, 1989), symphurine tonguefishes can be abundant and probably account for a significant portion of the fish biomass. Although not usually targeted commercially, some tongue- fishes, other small-size flatfishes ( Etropus , Citha- richthys), and juveniles of larger species of flatfishes (i.e. Syacium spp., etc.) may represent a significant proportion of bycatch in trawl fisheries for shrimps and commercially important demersal finfishes (Roithmayr, 1965; Anderson, 1968; Furnell, 1982; Pellegrin, 1982; Maharaj, 1989; Maharaj and Recksiek, 1991). Irrespective of limited commercial importance, these relatively small flatfishes, because of their abundance and diversity, play significant ecological roles as both predator and prey in trophic interactions within benthic communities of the west- ern Atlantic (Yanez-Arancibia and Sanchez-Gil, 1986;. Accurate identification of fauna in bycatch of com- mercial fisheries is important to determine environ- mental impacts of commercial fishing on both target and nontarget species (Villegas and Dragovich, 1984; Sheridan et al., 1984; Rothschild and Brunenmeister, 1984; Maharaj and Recksiek, 1991; Andrew and Pepperell, 1992; Murray et al., 1992; Murawski, 1994). Such impacts, along with other large-scale anthropogenic changes on the biosphere, have high- lighted the urgency for careful evaluations of oce- anic biodiversity in order to provide baseline infor- mation for researchers tasked with monitoring ef- fects of such changes on the biotas. Meaningful esti- mates of biodiversity, as well as accurate estimates of the faunal composition of noncommercial bycatch in commercial fisheries, depend upon accurate iden- tifications of the taxa involved. Accurate identifications, however, require detailed systematic studies of the fauna. Uncertainties re- garding the taxonomic status of several species of western Atlantic tonguefishes, concomitant with in- adequate diagnoses, and until recently, the relative scarcity of representative size series for many deep- sea species, have precluded accurate identifications Munroe: Systematics of western Atlantic Symphurus 3 for many western Atlantic Symphurus. Inherent dif- ficulties with identifications have also prevented detailed comparative study of ecologies and life his- tories for many tonguefishes occurring in this region. Consequently, despite the fact that Symphurus is the most speciose western Atlantic flatfish genus, it has remained, both systematically and ecologically, one of the least known western hemisphere groups of flat- fishes, particularly with respect to those species oc- curring in bathyal regions. Only in the last three decades has intensified study of fish communities inhabiting bathyal regions yielded larger samples of deepwater tonguefishes. However, this material has largely remained unidentified owing to a lack of ad- equate descriptions and identification keys for most Atlantic Symphurus species. Consequently, ecologi- cal information associated with these specimens has been minimally assessed. Objectives of this study are to revise the species of symphurine tonguefishes occurring in the western Atlantic Ocean, including evaluation of all nominal species described previously; to present detailed de- scriptions, diagnoses, and an identification key for the 24 species herein considered valid; and to sum- marize available distributional and other ecological data for each species. Early life history stages of Symphurus are abundant in ichthyoplankton collec- tions throughout the western Atlantic, but larval series of only a few species have thus far been iden- tified (Olney and Grant, 1976; Kurtz and Matsuura, 1994). Meristic data in this paper, coupled with geo- graphic information for the species, should facilitate identification of larval series for more species. This work complements earlier revisionary stud- ies on symphurine tonguefishes occurring in the At- lantic Ocean (Ginsburg, 1951; Menezes and Ben- vegnu, 1976; Munroe, 1990, 1991). Revisionary stud- ies are being presented regionally because of the large number of species in the genus and need for analysis of each species before a phylogenetic study can be accomplished. No phylogenetic hypotheses of rela- tionships for species of Symphurus have been pro- posed, precluding interpretation of geographical or ecological information for this taxon within an his- torical context. Munroe (1992) recognized nine spe- cies groups within Symphurus primarily on the ba- sis of shared similarities in interdigitation (ID) pat- terns. Although some species groups are perhaps not monophyletic, tonguefishes possessing similar ID patterns were found to have additional shared fea- tures, thus supporting the hypothesis that species with the same ID pattern are more closely related than those possessing different ID patterns. Species descriptions and discussions of distributional and size-related life history information for the western Atlantic tonguefishes are presented below within the context of these species groups. Taxonomic history At least 34 nominal species of western Atlantic symphurine tonguefishes (Table 1) have been de- scribed, commencing with the earliest descriptions of Pleuronectes plagiusa Linnaeus, 1766 (=S. plagiusa, this study), and Pleuronectes plagusia Schneider, in Bloch and Schneider, 1801 (=S. plagusia). From the early 1800’s until collections were made in deepwater habitats during oceano- graphic surveys in the 1880’s and early 1900’s, lit- erature dealing with western Atlantic Symphurus consisted almost entirely of nomenclatural re- arrangements of previously described taxa with little new information. However, in the mid-1880’s, explo- ration of New World deep-sea environments began in earnest, and major oceanographic expeditions re- covered many new species of fishes. Among these were five tonguefishes: Aphoristia ( -Symphurus ) nebulosa, A. diomedeana, A. pusilla, A. marginata, and A. pigra, described in a series of papers by Goode and Bean (1883, 1885b, 1886) and which still repre- sent the majority of deep-sea western North Atlan- tic tonguefish species. In 1889, Jordan and Goss evaluated the validity of pleuronectiform species of Europe and the Ameri- cas. No new species of tonguefishes were described, but these authors proposed that Aphoristia nebulosa (=S. nebulosus) differed significantly enough from other tonguefishes to be placed in a separate genus or subgenus (Acedia ). Also apparent in this and ear- lier works (Jordan, 1886a, 1886b) is that subtle varia- tions in meristic and morphometric features, char- acteristic of members of this taxon, were not fully appreciated by these authors. Consequently, charac- ters useful in properly diagnosing the species were not identified. Jordan and Goss, for example, con- cluded that the western Atlantic S. pusillus (Goode and Bean) and S. diomedeanus (Goode and Bean) were probably not distinct species but represented geographically variable populations of the common, abundant inshore species S. plagiusa (Linnaeus). Likewise, they also considered the eastern Pacific species S. elongatus (Gunther) and S. atricaudus (Jor- dan and Gilbert) as probably being geographic vari- ants of the tropical western Atlantic S. plagusia (Schneider, in Bloch and Schneider). These taxa are all now regarded as distinct species (Munroe, 1992). Later, Jordan and Evermann ( 1898) reviewed pub- lished information and evaluated the status of tonguefishes occurring in northern and central re- 4 Fishery Bulletin 96(1 ), 1 998 Table 1 Status of specific and subspecific names and new combinations assigned to western Atlantic species of Symphurus in chronologi- cal order. (Original authorship, generic placement, and spelling are maintained in the table.) Taxon Status Plagusia Browne, 1756 nonbinomial (rejected) Pleuronectes plagiusa Linnaeus, 1766 S. plagiusa Pleuronectes plagusia Browne, 1789 (rejected) Pleuronectes plagusia Schneider, in Bloch and Schneider, 1801 (after Browne) S. plagusia Achirus ornatus Lacepede, 1802 nomen dubium Plagusia ornata Cuvier, 1816 IS. tessellatus Plagusia tessellata Quoy and Gaimard, 1824 S. tessellatus Plagusia brasiliensis Agassiz, 1829 S. tessellatus Plagusia fasciata DeKay, 1842 S. plagiusa Aphoristia ornata Kaup, 1858 (n. comb.) IS. tessellatus Glossichthys plagiusa Gill, 1861 (n. comb.) S. plagiusa Plagusia plagiusa Gill, 1864 (n. comb.) S. plagiusa Aphoristia nebulosa Goode and Bean, 1883 S. nebulosus Aphoristia diomedeana Goode and Bean, 1885 S. diomedeanus Aphoristia pusilla Goode and Bean, 1885 S. pusillus Aphoristia marginata Goode and Bean, 1886 S. marginatus Aphoristia pigra Goode and Bean, 1886 S. piger Acedia nebulosa Jordan and Goss, 1889 (n. comb.) S. nebulosus Aphoristia fasciata (not DeKay) Goode and Bean, 1895 S. tessellatus Symphurus jenynsi Evermann and Kendall, 1907 S. jenynsi Symphurus bergi Thompson, 1916 S. jenynsi Symphurus trewavasae Chabanaud, 1948 S. trewavasae Symphurus sumptuosus Chabanaud, 1948 S. diomedeanus Symphurus minor Ginsburg, 1951 S. minor Symphurus parvus Ginsburg, 1951 S. parvus Symphurus pelicanus Ginsburg, 1951 S. pelicanus Symphurus pterospilotus Ginsburg, 1951 S. diomedeanus Symphurus civitatum Ginsburg, 1951 S. civitatium Symphurus urospilus Ginsburg, 1951 S. urospilus Symphurus ommaspilus Bohlke, 1961 S. ommaspilus Symphurus rhytisma Bohlke, 1961 S. rhytisma Symphurus arawak Robins and Randall, 1965 S. arawak Symphurus kyaropterygium Menezes and Benvegnu, 1976 S. kyaropterygium Symphurus ginsburgi Menezes and Benvegnu, 1976 S. ginsburgi Symphurus meridionalis Lema and Oliveira, 1977 S. jenynsi Symphurus oculellus Munroe, 1991 S. oculellus Symphurus caribbeanus Munroe, 1991 S. caribbeanus Symphurus billykrietei n. sp. S. billykrietei Symphurus stigmosus n. sp. S. stigmosus gions of the New World. They described no new spe- cies, but contrary to Jordan and Goss (1889), S. pusillus, S. diomedeanus, S. elongatus, and S. atricaudus were recognized as distinct species. It is evident, however, that these authors were still in- fluenced by earlier conclusions presented in Jordan and Goss (1889) because Jordan and Evermann hy- pothesized that S. pusillus and S. diomedeanus were closely related to S. plagiusa (Linnaeus). A more re- cent hypothesis (Munroe, 1992) indicates that S. pusillus belongs to a species group distinct from that including S. diomedeanus and S. plagiusa. Four additional nominal species of Symphurus from western South Atlantic localities were described during the first half of this century. In 1907, Evermann and Kendall described S. jenynsi from Argentina, and Thompson (1916) described S. bergi (=S. jenynsi) from the same geographic area. Later, Chabanaud (1948a) described S. trewavasae and S. sumptuosus ( =S . diomedeanus ) from off Brazil and Uruguay, respectively. In 1951, Ginsburg published the first revision of western Atlantic tonguefishes since that of Jordan and Evermann (1898). He recognized 15 nominal species of western Atlantic tonguefishes, including six previously undescribed. Five of the new species ( S . civitatum, S. minor , S. parvus, S. pelicanus, and S. urospilus) were described on the basis of material Munroe: Systematics of western Atlantic Symphurus 5 from the Caribbean Sea and more northern areas, whereas S. pterospilotus (=S. diomedeanus\ see Menezes and Benvegnu, 1976; this study) was de- scribed from a single specimen taken off Uruguay. In addition to describing new species, Ginsburg di- agnosed the genus, evaluated taxonomic characters considered important for identifying tonguefishes, and updated information on distributions and diag- nostic features for all 15 nominal species that he recognized. Despite these important contributions, Ginsburg’s study was limited because his treatment (particu- larly of deep-sea, southern Caribbean Sea, and South Atlantic species) was constrained by insufficient material. In addition, Ginsburg relied almost exclu- sively on external characters (primarily fin-ray counts) to identify and diagnose his specimens and therefore was unable to resolve problems involving externally phenetically similar species that differ unambiguously in internal characters, such as in- terdigitation patterns and vertebral numbers (Munroe, 1987). Soon after Ginsburg’s revision, two species ( S . ommaspilus and S. rhytisma) of shallow-water, dwarf tonguefishes were discovered in the Caribbean on patches of sand adjacent to coral reefs (Bohlke, 1961). In 1965, Robins and Randall described S. arawak, a third species of dwarf tonguefish collected in similar habitats. The first significant revision of western South At- lantic tonguefishes was by Menezes and Benvegnu ( 1976), who studied primarily tonguefishes occurring along the eastern coast of South America, although including comparative material from elsewhere in the western Atlantic whenever possible. Two new species (S. kyaropterygium and S. ginsburgi ), col- lected from moderate depths on the continental shelf off southern Brazil, were described in their work, and S. pterospilotus Ginsburg was placed in the syn- onymy ofS. diomedeanus (Goode and Bean). Menezes and Benvegnu’s study complemented that of Ginsburg (1951), but the regional nature and lim- ited study material of this revision prevented reso- lution of the status of several nominal species of western Atlantic tonguefishes. In 1977, Lema and Oliveira published a key to western Atlantic species of Symphurus , based almost entirely on information (primarily counts of fin rays) gathered from published literature accounts. In ad- dition to their identification key, these authors dis- cussed the distribution of symphurine tonguefishes in southern Brazilian waters and described S'. meridionalis ( =S . jenynsi, see below) from shallow waters on the inner continental shelf off southern Brazil. The most recent systematic treatment of Atlantic symphurine tonguefishes is that of an unpublished dissertation (Munroe, 1987), in which 23 species, including three new ones, were recognized in the western Atlantic. Five of these 23, including two undescribed species (now S. oculellus and S. caribbeanus), represent the Atlantic members of the S. plagusia complex and were documented earlier (Munroe, 1991). The present study expands upon earlier research on this group of fishes. It includes one previously undescribed species in addition to those reported in Munroe (1987), and it provides additional information on Atlantic members of the S. plagusia complex. Western Atlantic tonguefishes belong to the same species groups as eastern Atlantic and eastern Pa- cific tonguefishes, and several species pairs, compris- ing a western Atlantic species and another species from these areas, are hypothesized. No western At- lantic Symphurus species occur in these other areas; therefore, detailed comparisons distinguishing west- ern Atlantic tonguefishes from species occurring in other geographic regions are provided only among phenetically similar species or hypothesized species pairs. Comparative information on eastern Atlantic tonguefishes is found in Munroe (1990, 1992), whereas Munroe and Mahadeva (1989), Mahadeva and Munroe (1990), Munroe and Nizinski (1990), Munroe et al. ( 1991), Munroe ( 1992), and Munroe et al. (1995) provide data for eastern Pacific tongue- fishes. Among western Atlantic tonguefishes, only S. nebulosus has a pterygiophore interdigitation pat- tern (1-2-2) commonly found in Indo-West Pacific tonguefishes (Munroe, 1992). Therefore, all western Atlantic tonguefishes, except S. nebulosus, are readily distinguished from those of the Indo-West Pacific region by differences in ID patterns (Munroe, 1992), and no further comparisons between species from western Atlantic and Indo-Pacific regions are neces- sary to distinguish the species. Methods and materials Counts and measurements (Figs. 1-5J Descriptions of pigmentation are based on formalin- fixed fishes stored in alcohol. In text and tables (whenever possible), species that share a common interdigitation pattern (Fig. 1, A-E) are grouped to- gether and arranged alphabetically within this grouping. The order of presentation begins with S. nebulosus , the only western Atlantic species with a 1-2-2 interdigitation pattern; then follow species with the 1-3-2 pattern, unpigmented peritoneum, and four 6 Fishery Bulletin 96( 1 ), 1998 hypurals; species with 1-3-2 pattern, pigmented peri- toneum, and four hypurals, with newly described species presented first; those with 1-3-2 pattern, pig- mented peritoneum, and five hypurals; S. trewavasae with the 1-3-3 pattern; four species with a 1-4-2 pat- tern (three with shared pigmentation pattern fol- lowed by S. ommaspilus)-, and nine species with the 1-4-3 pattern, arranged alphabetically in subgroups of increasing numbers ( 10, 11, 12) of caudal-fin rays. In the species accounts, only total ranges for mer- istic features are presented; modal counts can be found in the tables. Variation in meristic features of widely distributed species were examined for speci- mens collected throughout the geographic range of the species. Although statistically significant in- traspecific differences were not apparent in features examined, meristic data, partitioned by geographic region, were tabulated to facilitate identifications. Some sympatric species have nearly complete over- lap in meristic features when data are summarized for specimens collected throughout their entire ranges. However, when meristic data for specimens within smaller geographic regions are examined more closely, the amount of overlap in counts between some pairs of co-occurring species was found to be less, and thus counts of meristic features were more in- formative, thereby facilitating identifications. Material examined for western Atlantic species is listed in the Appendix. Specimens of species occur- ring in other regions and used in comparative analy- ses were listed in Munroe (1990, 1992). Interdigitation pattern (ID) Patterns of interdigi- tation of proximal dorsal pterygiophores and neural spines (Fig. 1, A-E) were counted and recorded (Table 2) according to the methods of Munroe (1992) for the Munroe: Systematics of western Atlantic Symphurus 7 Table 2 Number of specimens with predominant interdigitation patterns (ID pattern) of dorsal pterygiophores and neural spines in western Atlantic species of Symphurus. ID Pattern Species 1-2-2 1-3-2 1-3-3 1-4-2 1-4-3 1-5-2 1-5-3 Other nebulosus 22 — — — — — 5 arawak 1 35 1 — — — — 4 rhytisma — 9 - — — — — 0 ginsburgi — 56 1 1 — — — 2 billykrietei — 89 3 — — — — 0 stigmosus — 12 — — — — — 0 pusillus — 22 1 — — — — 2 pelicanus — 56 1 — — — — 3 marginatus 2 77 11 — — — — 9 piger 2 137 1 — — — — 1 trewavasae — 4 49 11 3 — — 6 kyaropterygium — — 1 13 — — — 0 minor — 2 — 74 1 — — 1 parvus — 1 — 33 6 35 — 5 ommaspilus — — — 28 — — — 0 diomedeanus — — 9 13 160 6 13 17 jenynsi — — 4 2 62 2 6 10 plagiusa — — 22 24 85 4 1 5 urospilus — — 4 8 74 15 4 5 plagusia — — 5 2 33 — 1 3 civitatium — — 8 19 128 7 3 6 tessellatus — — 10 11 209 3 15 30 oculellus — — 2 2 55 — 3 2 caribbeanus — — 8 — 69 2 — 5 first three, or in unusual cases, the first five inter- neural spaces. Only data for occurrence of predomi- nant patterns are reported here (additional informa- tion on variation in this character was presented in Munroe, 1992). The number of dorsal pterygiophores inserted into interneural spaces 1-3 was found to be diagnostic for species or groups of species of Symphurus (Munroe, 1987; 1992). Interdigitation patterns are recorded as a formula, such as 1-3-2 (Fig. IB), indicating that one pterygiophore inserts into interneural space one, three into interneural space two, and two into interneural space three. The first neural spine abuts directly against the cranium so that there is no obvious space between it and the cranium. Therefore, the first interneural space re- flected in the formula is that between the first and second neural spines. Caudal-fin rays (Table 3) Previous authors (Gins- burg, 1951; Mahadeva, 1956; Menezes and Benvegnu, 1976; Munroe, 1987, 1990, 1991) have found that the caudal-fin ray count is extremely conservative within species of this genus. Previous studies have included the ultimate dorsal- and anal-fin rays, which lie on the same vertical plane as the caudal-fin rays (Fig. 4, B-C), in the caudal-fin ray counts. This method is followed in the present study. Counts are usually even numbers (10, 12, 14) and rarely odd numbers (11 in 5. urospilus). Dorsal (Table 4J and anal-fin rays (Table 5J These include all rays except the ultimate ray. The thick, muscular gonadal duct preceding the first anal-fin ray is not counted. Vertebral counts (Table 6| All western Atlantic Symphurus consistently have nine abdominal verte- brae, three without and six with haemopophyses; abdominal vertebral counts are thus reported as (3+6). Counts of total vertebrae include the urostylar centrum. Hypural counts (Fig. 4, B-CJ These include all separate hypurals without any implied interpreta- tion of the fate (fused or lost during ontogeny) of the fifth hypural, which may not always be present as a separate element. Scale counts (Fig. 2D) Accurate, repeatable scale counts are difficult to make on species of Symphurus, 8 Fishery Bulletin 96(1 ), 1998 Table 3 Frequency distribution of numbers of caudal-fin rays for western Atlantic Symphurus species. Number of caudal-fin rays Species 8 9 10 11 12 13 14 15 16 nebulosus 2 24 1 arawak — — — 3 38 1 1 — — rhytisma — — — — 9 — — — — ginsburgi - 1 2 2 49 1 — — — billykrietei — — — 6 84 1 — — — stigmosus — — — 1 11 - — — — pusillus — — — — 26 — — — — pelicanus — — 1 4 50 1 — — — marginatus — — — 3 98 1 — — — piger — — 2 2 134 1 — — — trewavasae — — 73 — — — — — — kya ropterygi u m — — 13 1 — — — — — minor 1 1 74 3 — — — — — parvus - 3 70 3 — — - — — ommaspilus — — 28 — — — — — — diomedeanus — 9 202 1 1 — — — — jenynsi - 3 78 - — — — — — plagiusa — 5 132 2 — — — — — urospilus — 1 4 108 — — — — — plagusia — — — 2 41 1 — — — civitatium — — — 8 164 — — — — tessellatus — — 4 17 249 3 — — — oculellus — — — 4 59 — — — — caribbeanus — — 1 2 81 1 — — — especially for those specimens collected by trawls at considerable depths, because scales are often abraded and lost. For specimens missing scales, approximate counts were based on partial scale counts, counts of scale pockets, or on a combination of the two when- ever possible. Longitudinal scale count (Table 7) in- cludes the total number of complete diagonal rows of scales along a hypothetical line starting immedi- ately above the opercular angle and continuing pos- teriorly along the middle of the body to the end of the hypural plate (Fig. 2D, number 1); the few rows of scales along the caudal-fin base are not included, and the last scale to be included in the count must be at least half way in front of the hypural plate. The head scale count (Table 8) includes all oblique rows of scales on the head counted posteriorly from the first complete row of scales immediately behind the posterior border of the lower eye (Fig. 2D, num- ber 2); it includes the last complete row of scales immediately anterior to the midpoint emargination on the posterior border of the operculum, but the few small rows of scales present on either the dorsal or ventral fleshy lobes of the operculum are not in- cluded. The transverse scale count (Table 9) is the number of scales in a diagonal row from the base of the dorsal fin at a point directly above the posterior margin of the operculum to the base of the anal fin (Fig. 2D, number 3). Scales extending out onto the dorsal- and anal-fin rays are not included. Measurements were made either on ocular- or blind-side surfaces (Fig. 2, A-C). All measurements in the text refer to standard length, unless other- wise noted. Measurements less than 150 mm were taken to the nearest 0.1 mm with dial calipers or an ocular micrometer. Measurements over 150 mm were taken to the nearest mm with a steel ruler. Morpho- metric features are expressed either as measure- ments in thousandths of standard length (SL) or thousandths of head length (HL) and are defined as follows: Standard length (SL): distance from tip of fleshy snout to posterior end of hypural plate. Trunk length (TKL; not measured on all species): longitu- dinal distance from posterior angle of operculum to caudal-fin base. Body depth (BD): distance across body at the anus, exclusive of fins; measured on blind side. Preanal length (PAL): tip of fleshy snout to ori- gin of anal fin; measured on blind side. Dorsal-fin length (DBL): base of anteriormost dorsal-fin ray to posterior end of hypural plate. Predorsal length (PDL): tip of fleshy snout to base of first dorsal-fin ray. Anal-fin length ( ABL): base of anteriormost anal- fin ray to posterior end of hypural plate. Caudal-fin Munroe: Systematics of western Atlantic Symphurus 9 10 Fishery Bulletin 96(1 ), 1998 Munroe: Systematics of western Atlantic Symphurus Table 6 Frequency distributions of numbers of total vertebrae for western Atlantic Symphurus species. Number of vertebrae Species 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 nebulosus — — — — — — — — — — — — — — — — — — 1 9 15 2 arawak 117 21 6 — — — — — — — — — — — — — — — — — — rhytisma — — — — — — — 1 6 2 — — — — — — — — — — — — billykrietei — — — — — — — — — — — 9 49 33 2 — — — — — — — stigmosus — — — — — — — — — — — — 5 7 — — — — — — — — ginsburgi — — — — — — — — — — — 7 38 13 — — — — — — — — pelicanus — — — — 1 7 25 27 — — — — — — — — — — — — — — pusillus — — — — — — — — 4 15 10 — — — — — — — — — — — marginatus — — — — — — — — — — — — 1 17 35 41 2 2 — — — — piger — — — — — — 1 4 48 80 11 — — — — — — — — — — — trewavasae — — — — — — — — 3 25 31 8 3 — — — — — — — — — kyaropterygium — — — — — — — 1 9 3 1 — — — — — — — — — — — minor — — 11 45 19 4 — — — — — — — — — — — — — — — — parvus — — — — 1 13 23 34 7 — — — — — — — — — — — — — ommaspilus — — — — 10 16 — — — — — — — — — — — — — — — — diomedeanus — — — — — — — — 4 50 131 34 — — — — — — — — — — jenynsi — — — — — — — — — — — — — — — — — — 4 37 36 11 plagiusa — — — — — 1 16 30 69 26 1 — — — — — — — — — — — urospilus — — — — — 3 54 49 7 1 — — — — — — — — — — — — caribbeanus — — — — — — — — — 1 43 33 5 — — — — — — — — — civitatium — — — — — — — 2 34 109 28 2 — — — — — — — — — — oculellus — — — — — — — — — — — — — 6 35 20 2 — — — — — plagusia — — — — — — — — 1 4 21 13 5 — — — — — — — — — tessellatus — — — — — — — — — 1 3 52 97 89 37 3 — — — — — — length (CFL): base of articulations of middle caudal- fin rays to tip of longest middle rays. Pelvic-fin length (PL); (only blind-side pelvic fin present in adults): basal articulation to distal tip of longest ray. Pelvic to anal length (PA): shortest horizontal distance from base of most posterior pelvic-fin ray to anal-fin ori- gin. Head length (HL): tip of fleshy snout to most posterior extension of upper fleshy lobe of opercu- lum. Head width (HW): greatest distance across head at posterior portion of operculum. Postorbital head length (POL): posterior margin of lower eye to poste- rior extent of upper fleshy lobe of operculum. Upper head lobe width (UHL): distance at operculum from dorsal margin of body to dorsal origin of operculum. Lower head lobe width (LHL): distance at opercu- lum from dorsal origin of operculum to most ventral part of operculum. Snout length (SNL): anterior rim of lower eye to tip of snout. Upper jaw length (UJL): shortest horizontal distance from bony tip of premax- illa to angle of mouth. Eye diameter (ED): greatest horizontal diameter of the cornea of the lower eye; does not include fleshy tissue surrounding eye. Chin depth (CD): vertical distance from angle of mouth to most ventral aspect of head. Upper opercular lobe (OPUL): vertical distance from midpoint of opercu- lar indentation to dorsal origin of operculum. Lower opercular lobe (OPLL): vertical distance from mid- point of opercular indentation to ventral margin of operculum. Qualitative characters The following qualitative characters are also impor- tant in identifying Symphurus, especially when used in combination with meristic and morphometric data. Pupillary operculum (Fig. 3AJ A triangular or rounded, pigmented structure on the upper part of the cornea. A presumed function of the pupillary oper- culum is to shade the retina from direct exposure to light. Chabanaud (1948a), Ginsburg (1951), and Menezes and Benvegnu (1976) did not use this char- acter in their studies on Atlantic tonguefishes. Mahadeva (1956) and Munroe ( 1987), however, found this character useful for diagnosing some eastern Pacific and Atlantic species. For example, a pupil- lary operculum is not found in any eastern Atlantic tonguefish nor in any western Atlantic deepwater species, and its absence is useful in distinguishing these Symphurus from other western Atlantic species with similar meristic and morphometric features but which possess a well-developed pupillary operculum. 12 Fishery Bulletin 96(1 ), 1998 Jaw position (Fig. 3, A-EJ Relative position of the posteriormost point of the jaws with respect to the lower eye is useful in diagnosing some species. Five DBL Figure 2 Body and head locations where measurements and scale counts (defined in text) were taken. Abbreviations are de- fined in “Counts and measurements” section. (A) Measure- ments made on ocular side of body. ( B ) Measurements made on blind side of body. (C) Measurements made on ocular side of head. (D) Scale count locations: 1 = longitudinal scale count; 2 = head scale count; 3 = lateral scale count. different positions of the posterior margin of the jaws were evident among the species. Species with short jaws were those with the posterior margin of the jaws at the anterior margin of the eye. Those with moder- ately long jaws have the posterior margin at the mideye region or at the posterior margin of the pu- pil. Species with long jaws are those with the poste- rior margin of the jaws at, or beyond, the vertical through the posterior margin of the eye. Dentition on ocular-side jaws Degree of develop- ment of dentition on ocular-side jaws is useful in di- agnosing some species. Some species have teeth along the entire margin of both jaws, others have only a partial row of teeth along the margin, and some lack teeth on the ocular-side jaws. Fleshy ridge on ocular-side lower jaw (Fig. 3D) Presence or absence of a fleshy ridge on the ocular- side lower jaw is diagnostic for some species. Dorsal-fin origin (Fig. 3, A-F) Relative position of the dorsal-fin origin with respect to the migrating (upper) eye is useful in identifying some species. Squamation on dorsal- and anal-fin rays (Fig. 4A) Presence and approximate number of scales on dorsal- and anal-fin rays, especially on blind sides of the fin rays, is useful for identifying some species. Membrane ostia in dorsal and anal fins Presence or absence of membrane ostia (small pores) in the basal part of the membranes of the dorsal and anal fins is useful for identifying some species. Body pigmentation Numbers in parentheses refer to numbers on Figure 5, A and B. Ocular surface col- oration is unique for some species. Frequently ob- served pigmentation patterns consist of uniform col- oration with or without a bold caudal blotch (7), or patterns featuring a variable number of bold cross- bands (2). Some species have the blind side of the body pigmented with a pepper-dot pattern (8) or with median dermal spots (i.e. those located internally along the vertebral column and visible externally (9), but most species usually have uniformly creamy- white or slightly yellowish coloration on the blind side. Fin pigmentation Numbers in parentheses refer to Figure 5A. Pigment patterns on dorsal, anal, and caudal fins distinguish some species. Fin pigmenta- tion showed the following variation: fins uniformly pigmented; fins with blotches (6); fins with rounded (5) or ocellated spots (4); and fins with a longitudi- nal stripe (3). Munroe: Systematics of western Atlantic Symphurus 13 Table 7 Frequency distributions of the numbers of longitudinal scale rows for western Atlantic Symphurus species. Number of scale rows Species 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 arawak 1 — 1 3 4 6 3 7 7 3 1 — — — — — — — — — — — — — — — — — — — — — — ___ pelicanus — — — — — — — 1 1 1 2 1 1 2 1 1 — — — — — — — — — — — — — — — — — — — — piger 1 - 3 1 9 9 10 13 13 9 10 6 1 1 trewavasae — — — — — — — — — — — — 1 1 1 2 4 3 5 12 1 5 3 — — — — — — — — — — — — — minor 2 1 3 3 6 9 8 8 6 3 1 1 2 — — — — — — — — — — ______ — — — — _ — — ommaspilus — — — 4 1 2 1 8 4 2 — — — — — — ____ — — — — — — — _ — — — — — — — — parvus -2 2 3 1 5 4 45 2 12 3 2 - 1 4 1 1 kyaropterygium — — —— — — — — — —— —— —— — — — 2 — 1 1 3 — — — 1 — — — — — —— _ — urospilus 1 22 3 5 6 3 5 5 11 5 4 5 4 2 1 civitatium — — — — — — — — — — — 1 — — — 2 — — 1 4 8 11 5 14 8 7 3 3 1 — — — — — — — plagiusa — — — — — — — — — — — — — — — — — — — — — 1 2 1 4 2 6 4 2 7 — 1 — — — — plagusia — — — — — — — — — — — — — — — — — — — — — — — — 1 3 2 3 2 2 1 3 — — 2 — caribbeanus — — — — — — — — — — — — — — — — ____ — — _ 3 — 411 4 3 6 2 2 1 1 3 — Number of scale rows Species 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 billykrietei — 1 — — — — 3 — — 4 11 — 412 — — — 2 1 2 ginsburgi ________ 2 2— 2 1 — 2 2 — — — — — — stigmosus __________ _______ 2 1 3 marginatus — — — — — — — 335 566 85653 1 1 2 — rhytisma — — — — — — — — — — — — 1 1 1 — 1 1 1 — — — diomedeanus 1 1 1 4 5 7 11 12 13 11 8 13 9 3 1 1 tessellatus 2 2 8 10 10 11 21 31 21 21 14 5 4 2 - 1 oculellus 1 3 1 7 5 26 4 2 - 1 1 Number of scale rows Species 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 jenynsi 1-32331213322521-1 Number of scale rows Species 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 nebulosus 1 — — 11 2 2—4 1 4 — 1—1 1 Peritoneum pigmentation Relative intensity and coverage of pigmentation on the peritoneum is diag- nostic for species especially when used in combina- tion with other characters. Some species have an un- pigmented peritoneum; some have a spotted perito- neum; and others have a black peritoneum. Opercular pigmentation Degree of pigmentation on outer and inner surfaces of the opercle and isth- mus, in combination with other characters, is useful for identifying some species. Some species have a prominent opercular blotch (Fig. 5A, number 1). Ecological assessment Maturity was estimated by macroscopic examination of extent of posterior elongation of the ovary and pres- ence of developing ova in the ovaries (ovaries of ma- ture females are sometimes conspicuous through the body wall in transmitted light; in immature and large females, ovaries are best observed by dissection). Be- cause no obvious differences in sizes of testes between immature and mature males were apparent, esti- mates of maturity were based entirely on females (see Figs. 6-9). Immature females were those with 14 Fishery Bulletin 96( 1 ), 1 998 Frequency distributions of the number of scales on Table 8 the posterior head region for western Atlantic Symphurus species. Species Number of scale rows 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 nebulosus 1 2 4 8 3 1 arawak 2 6 14 16 rhytisma — — — — — — 4 1 1 billykrietei - — — — 1 2 3 7 6 3 1 — — — — — — stigmosus — — — — — — — 3 2 2 1 — — — — — — ginsburgi — — — 1 2 2 4 5 3 pelicanus — — 4 2 1 1 pusillus — — — — — 4 3 2 marginatus — — — — 6 32 25 5 piger — — — — 3 28 26 13 1 1 — — — — — trewavasae — — — 4 23 10 3 — 1 kyaropterygium — - - — 4 5 1 minor 2 13 23 12 parvus — 1 5 20 8 4 1 ommaspilus — — 5 9 6 3 diomedeanus — — — — 11 25 32 29 3 2 — — — — — — — jenynsi 2 15 12 5 1 — — — plagiusa — — — 1 2 17 8 3 urospilus — 2 11 25 26 3 caribbeanus — — — — — 1 1 9 16 8 3 — — — — — — civitatium — — — — 3 18 32 16 3 oculellus — — — — — — — 3 8 1 1 2 — — — — — plagusia — — — — — — 8 4 8 2 1 — — — — — — tessellatus — — — — — — 1 — 8 8 4 1 — — — Table 9 Frequency distributions of the number of lateral rows of scales for western Atlantic Symphurus species. Number of lateral scale rows Species 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 nebulosus — — — — — — — • — — — — — — — — — — — — 1 1 1 1 1 — 3 1 — arawak — — — 1 — 1 — 1 9 9 5 4 8 — — — — — — — — — — — — — — — rhytisma — — — — — — — — — — — — — — — — — — — 2 2 1 — — — — — — ginsburgi — — — — — — — — 1 — — 1 — 2 1 2 1 4 1 — 1 — - — — — — — — pelicanus 2 1 5 — 1 1 — — — — — — — — — — — — — — — — — — — — — — pusillus — — — — — — — — — 1 1 — — — — — — — — — — — — — — — — — marginatus — — — — — — 3 1 1 — 2 3 — 1 — — — — — — — — — — — — — — piger — — — — — — — — - 1 2 2 5 3 2 2 — — — — — — — — — — — - — — trewavasae — — — — — — — - 1 2 8 11 9 — 2 — — — — — — — — — - — — — — — kyaropterygium — — — — — — — — — — — — 4 3 — 2 — — — — — — — — — — — — minor 1 2 2 1 3 6 7 7 — — — — — — — — — — — — — — — — — — — — parvus — — 2 — 1 2 5 4 5 2 2 1 — — — — — — — — — — — — — — — — ommaspilus — — — 1 — 6 5 5 2 1 — — — — — — — — — — — — — — — — — - — diomedeanus — — 2 2 3 4 12 15 16 7 3 4 4 1 1 1 — — — — jenynsi — _______ 1 1 1 3 6 6 7 5 2 — 1 plagiusa — — — — — — 1 — — 2 2 5 4 6 6 4 — — — — — — — — — — — — urospilus — — — 1 — 1 2 8 15 11 13 3 3 1 8 7 8 — — — — — — - — — — — — caribbeanus — — — — — — — — — — — — 2 3 6 10 3 6 2 1 1 — — — — — — — civitatium — — 2 1 2 — 2 5 4 6 2 11 5 6 6 1 — — — — — — — — — — — — oculellus — — — — — — — — — — — — 1 1 6 2 2 1 1 — — — — — — — — — plagusia — — — — — — — — — — — 1 — — 2 3 7 3 3 1 — — — — — — — — tessellatus — — — — — — — — — — — — — — 1 2 4 5 2 1 3 2 — — — — — — Munroe: Systematics of western Atlantic Symphurus 15 Pupillary operculum Figure 3 Eye structures, lower jaw positions , dorsal-fin positions, and jaw struc- tures observed in western Atlantic Symphurus. (A) Eyes with well-devel- oped pupillary operculum. (B) eyes without pupillary operculum. (A-E) Position of lower jaw relative to nonmigrating (lower) eye: (A) mideye; (B) anterior margin of eye; (C) posterior margin of eye; (D) posterior margin of pupil; and (E) posterior to eye. (A-F) Position of dorsal-fin origin in rela- tion to eye: (A and C) at mideye; (B) at anterior margin of eye; (D) anterior to eye; and (E) posterior to eye. (A-E) Jaw structures: (A-C, E) lower jaw without fleshy ridge; (D) lower jaw with fleshy ridge. nonelongate or only partially elongate ovaries. Ma- ture females had fully elongate ovaries. Gravid fe- males were those with enlarged ovaries filled with large, macroscopically visible ova. When available, depth-of-capture information (con- verted and rounded to the nearest meter) was re- corded and summarized (Table 10) for specimens listed in the Appendix. If depth of capture comprised a range of depths over which the nets were towed, a mean depth for that particular trawl was calculated and used in analyses. Evaluations of bathymetric distribution were not based on random-stratified sampling but on information from available mate- rial, supplemented with depth information from the literature. There exists, therefore, a potential for bias with respect to depth-of-capture information. Statistical analyses (SPSS, 1975; SPSS-X, 1986) were conducted on the Primos computer system at the Virginia Institute of Marine Science. All statisti- cal analyses of morphometric and meristic data were conducted on log-transformed data. Synonymies appearing in accounts for S. plagusia, S. tessellatus, and S. civitatium (see below) are se- lective and abbreviated from the more detailed list compiled by Munroe (1991) for these species. The synonymy for S. plagiusa complements the compila- tion of detailed literature for S. plagiusa presented in Topp and Hoff (1972). Abbreviations for institutions providing study material, or in which type material is deposited, fol- low Leviton et al. (1985). Additional collections are as follows: CAS-SU: California Academy of Sciences, 16 Fishery Bulletin 96( 1 ), 1998 B C Figure 4 Degree of development of scales on blind sides of dorsal- and anal-fin rays and method of counting caudal-fin rays and hypurals for western Atlantic species of Symphurus. (A) Dorsal- and anal-fin rays with small ctenoid scales along their lengths. (B) Species with 12 caudal-fin rays and five hypurals (in black). (C) Species with 10 caudal- fin rays and four hypurals (in black). San Francisco (collections formerly at Stanford Uni- versity); IBUNAM: Institute de Biologia, Universidad Nacional Autonoma de Mexico, Mexico, DF; IMS: Marine Sciences Institute, University of Texas at Austin, Port Aransas; INIDEP: Institute Nacional de Investigacion y Desarrollo Pesquero, Mar del Plata; UFPB: Departamento de Sistematica e Ecologia, Universidade Federal da Parafba, Joao Pessoa; UMML: Rosenstiel School of Marine and At- mospheric Sciences, University of Miami, Miami (now part of University of Florida (UF) collections; add 200,000 to original UMML catalog number); USU: Universidade Santa Ursula, Rio de Janeiro; ZMA: Zoologisch Museum, Universiteit van Amsterdam, Amsterdam (now Institute for Systematics and Popu- lation Biology, Zoologisch Museum). Munroe: Systematics of western Atlantic Symphurus 17 Figure 5 Pigment characteristics found on body and fins of western Atlantic Sym- phurus. (A) Ocular-side pigment patterns occurring on body and fins: 1 = opercular blotch; 2 = bold crossband; 3 = longitudinal fin stripe; 4 = ocellated fin spot; 5 = rounded fin spot; 6 = fin blotch; and 7 = caudal blotch. (B) Blind-side pigmentation: 8 = pepper-dot pattern; 9 = median dermal spots. A S. a raw ak B S. rhytisma 1. 15- IQ- S' 0- C/J JU £ o -O E 3 z C S. pusillus immature mature D S. pe lie anus E S. minor 20-i 21-30 31-40 41-50 51-60 61-70 71-80 F 5. ommaspilus 20 1 15- 10- ■ 21-30 31-40 41-50 51-60 61-70 71-80 Size (mm) Figure 6 Frequency histograms indicating relative sizes (mm standard length) of immature and mature females, and sizes at sexual maturity for females of western Atlantic dwarf species (defined in text) of Symphurus. 18 Fishery Bulletin 96( 1 ), 1998 A S. parvus B S. ginsburgi 5 c 3 z C S. nebulosus I I immature 1 • B mature Size (mm) Figure 7 Frequency histograms indicating relative sizes (mm standard length) of immature and mature females, and sizes at sexual maturity for females of diminutive species ( defined in text) of western Atlantic Symphurus. A S. billykrietei B S. stigmosus E S. trewavasae F S. kyaroplerygium 20 1 20- 15- 15- 10- 10- 5- 5- (H G S. plagusia 20 40 50 60 70 80 90 100 110 120 130 140 150 H S. caribbeanus 40 50 60 70 80 90 100 110 120 130 140 150 Size (mm) Figure 8 Frequency histograms indicating relative sizes (mm standard length) of immature and mature females, and sizes at sexual maturity for females of medium-size species (defined in text) of western Atlantic Symphurus. Munroe: Systematics of western Atlantic Symphurus 19 Table 10 Summary of bathymetric data (depth in meters) for western Atlantic Symphurus species. Abbreviations: plagi = plagiusa\ di = diomedeanus; je = jenynsi; ur = urospilus; ca = caribbeanus; plagu = plagusia; ci = ciuitatium; te = tessellatus; oc = oculellus ; om = ommaspilus\ mi = minor, pa = parvus ; ky = kyaropterygium; tr = trewavasae; ar = arawak; rh = rhytisma\ pe = pelicanus\ pu = pusillus ; gi = ginsburgi; bi = billykrietei; st = stigmosus; pi = piger\ ma = marginatus; ne = nebulosus. Species Depth plagi di je ur ca plagu ci te OC om mi pa ky tr ar rh pe pu gi bi st Pi ma ne 1-10 116 i 5 ii 49 21 130 49 2 13 1 9 3 1 11-20 173 2 22 32 30 — 56 91 12 10 8 1 — 3 22 2 21-30 20 25 12 57 16 — 77 13 12 1 14 4 — 7 7 1 1 31-40 5 93 — 22 — 4 58 87 17 — 24 5 2 — 3 — 5 — — — — — i — 41-50 5 65 12 1 — — 13 67 6 — 21 18 1 14 — — 6 — — — — — 3 — 51-60 1 54 — 1 — 1 8 17 2 — 6 39 3 12 — — 22 — — — — — — — 61-70 2 24 — — — — 1 28 27 — 3 18 6 17 — — 16 — — - — — 1 — 71-80 81-90 91-100 101-110 111-120 — 23 10 13 2 — — — — 1 21 1 2 I = 6 12 — 17 I = 1 1 I = - - 1 - 1 — — — — — - 1 — - 6 — — — 1 7 2 4 3 2 i - 4 — — 121-130 — 1 1 131-140 141-150 151-160 161-170 171-180 181-190 191-200 — 2 — — — — — — — — — 1 — — — — 3 3 1 i — 4 12 — — — — — — — — — — 3 — - — — — — 2 11 2 — 13 36 — — 2 15 1 1 201-210 2 1 2 1 — — 211-220 221-230 3 6 24 231-240 241-250 251-260 1 4 4 11 3 4 2 261-270 271-280 — — — — — — — — — — — — — — — — — — 1 8 7 1 2 36 — — 281-290 291-300 1 5 3 3 1 6 2 1 — 301-310 311-320 321-330 331-340 341-350 351-360 2 5 7 4 1 1 1 1 10 1 10 2 — — — — — — — — — — — — — — — — — — — 1 4 3 2 1 361-370 1 1 2 7 — 371-380 4 1 2 4 — 381-390 1 1 15 1 391-400 1 2 401-410 1 7 — 411-420 1 — 421-430 431-440 2 1 441-450 1 1 451-460 2 3 — 461-470 4 — 471-480 5 — 481-490 1 491-500 10 501-510 1 — 511-520 2 1 continued 20 Fishery Bulletin 96(1 ), 1998 Table 10 (continued) Species Depth plagi di je ur ca plagu ci te oc om mi pa ky tr ar rh pe pu gi bi st pi ma ne 521-530 531-540 541-550 551-560 561-570 571-580 581-590 591-600 601-610 611-620 621-630 631-640 641-650 651-660 661-670 671-680 681-690 691-700 701-810 - — 1 - 2 — 4 10 - 2 2 1 1 2 4 3 A 5. urospilus B S. plagiusa C S. civitatiwn I I immature ■■ mature D S. diomedeanus 25 -I 25 -I 20- 3 J 20- JB 15- ■ f ! 15- 10- 10- '■L 5- 0- 5- 0- E S. oculellus F S. tessellatus ■ 1 T ■ 1 1 1 G S. jenynsi Size (mm) Figure 9 Frequency histograms indicating relative sizes (mm standard length) of immature and mature females, and sizes at sexual maturity for females of large-size species (defined in text) of west- ern Atlantic Symphurus. Munroe: Systematics of western Atlantic Symphurus 21 Artificial key to western Atlantic Symphurus la Caudal-fin rays 14 (rarely 13); dorsal-fin rays more than 104; anal-fin rays 91-98; peritoneum black, usually visible through abdominal wall on both sides of body; body elongate, of nearly uniform width throughout most of its length (Fig. 10); total vertebrae 57-60; ID pattern usually 1-2-2 (Fig. 1A) So nehulosus lb Caudal-fin rays less than 13 (rarely 13); dorsal-fin rays usually less than 104; anal-fin rays less than 91; peritoneum black or unpigmented; body usually deeper in anterior one-third of length and tapering noticeably posteriorly; total vertebrae less than 57; ID pattern with 3 or 4 pterygiophores inserted into second interneural space (usually 1-3-2, 1-3-3, 1-4-2, or 1-4-3) (Fig. 1, B-E) 2 2a Peritoneum black, usually visible through abdominal wall on both sides of body; caudal-fin rays usually 12; pupillary operculum absent (Fig. 3B); teeth present on entire margins of ocular-side jaws; ID pattern usually 1-3-2 (see Fig. IB) 3 2b Peritoneum unpigmented; caudal-fin rays 10-12; pupillary operculum present (Fig. 3A) or absent (Fig 3B); teeth present or absent over entire margins of both ocular-side jaws; ID pattern usually 1-3-2 (Fig. IB), 1-3-3 (Fig. 1C), 1-4-2 (Fig. ID), or 1-4-3 (Fig. IE) 9 3a Blind side of body with pepper-dot pattern of melanophores (usually heaviest along bases of dorsal and anal fins) (Fig. 5B, number 8); dorsal-fin rays 77-85; anal-fin rays 64-70; 70 or fewer scales in longitudinal series; dorsal-fin origin in posterior position, usually only reaching vertical through posterior margin of upper eye or occasionally not reaching that point; dorsal and anal fins without pigmented blotches or stripes; total vertebrae 43—46 S. pelicanus 3b Blind side of body without pepper-dot pattern of melanophores; dorsal-fin rays usually greater than 85; anal-fin rays usually greater than 69; usually more than 70 scales in longitudinal series; dorsal-fin origin in more anterior position, usually at point between verticals through middle of pupil and anterior margin of upper eye; dorsal and anal fins either uniformly pig- mented, or with series of alternating pigmented blotches and unpigmented regions, or with stripe along basal margin of fin; total vertebrae usually greater than 46 4 4a Dorsal-fin rays 93-104; anal-fin rays 80-89; ocular-surface usually with a large, dark brown dia- mond-shaped blotch on caudal region of body (see Fig. 32), but otherwise uniformly pigmented and without pattern of distinct crossbands; basal margins of dorsal and anal fins with dark brown stripe, but without blotches; total vertebrae 51-56, usually 52-54 S. marginatus 4b Dorsal-fin rays usually less than 95; anal-fin rays usually 84 or fewer; ocular-surface of body without dark brown, diamond-shaped blotch on caudal region, with or without distinct pattern of crossbands; dorsal and anal fins with or without pigmented blotches; total vertebrae 47-53, usually 52 or less 5 22 Fishery Bulletin 96( 1 ), 1998 5a Scales fewer, 62-75 in a longitudinal series; 5 hypurals; anal-fin rays 68-74; inner opercular linings and both sides of isthmus usually lightly pigmented; total vertebrae 45-49, usually 47-49 S. piger 5b Scales more numerous, usually 77-100 in a longitudinal series; 4 hypurals; anal-fin rays 71- 84; inner opercular linings and isthmus unpigmented; total vertebrae 47-52, usually greater than 48 6 6a Dorsal-fin rays 83—88; anal-fin rays 71-75; total vertebrae 47-49; scales in longitudinal series 77-87; ocular surface usually yellowish or lightly straw-colored, with one to two prominent, complete crossbands immediately posterior to opercular opening (Fig. 28); dorsal and anal fins without stripe along basal margin; (adult size relatively small, usually not exceeding 80 mm SL) S. pusillus 6b Dorsal-fin rays 87-95; anal-fin rays 74-84; total vertebrae 50-53; scales in longitudinal series 85-99; ocular surface usually dark brown, straw-colored or yellowish, with series of mostly incomplete crossbands posterior to opercular opening, or ocular surface uniformly pigmented without crossbands; dorsal and anal fins frequently with dark brown stripe along basal mar- gins, sometimes in combination with series of large, pigmented blotches alternating with un pigmented areas on dorsal and anal fins 7 7a Dorsal and anal fins usually with alternating series of prominent, darkly pigmented blotches (see Fig. 22); blotches usually wider than intervening unpigmented areas; no pigmented spot on scaly base of caudal fin; eyeballs round, usually contiguous, or nearly contiguous, within fleshy orbital sac (see Fig. 21A) S. stigmosus n. sp. 7b Dorsal and anal fins usually without alternating series of prominent, darkly pigmented blotches; if blotches present, then as wide as, or only slightly narrower than, width of intervening un- pigmented areas; pigmented spot present on scaly base of caudal fin; eyeballs longer than wide, separated by small space within fleshy orbital sac (see Fig. 2 IB) 8 8a Eye diameter relatively large (see Fig. 20), ratio of ED to TKL [ED:TKL] = 3. 2-4. 7 (usually exceeding 3.5% of trunk length) S. ginsburgi 8b Eye diameter relatively small (see Fig. 20), ED:TKL = 2. 5-4.0 (usually less than 3.4% of trunk length) S. billykrietei n. sp. 9a Caudal-fin rays usually 12; pupillary operculum absent (Fig. 3B); ID pattern usually 1-3-2 (Fig. IB) or 1-4-3 (Fig. IE) 10 9b Caudal-fin rays usually 10 or 11; pupillary operculum present (Fig. 3A) or absent (Fig. 3B); ID pattern usually 1-3-3 (Fig. 1C), 1-4-2 (Fig. ID), or 1-4-3 (Fig. IE) 16 Munroe: Systematics of western Atlantic Symphurus 23 10a Dorsal-fin rays 70-76; anal-fin rays 55-61; 55-65 scales in longitudinal series; pattern of pep- per-dots (Fig. 5B, number 8) on blind side of body (usually); some specimens with darkly pig- mented, triangularly shaped caudal blotch; total vertebrae 39-42; ID pattern usually 1-3-2 (Fig. IB); adult sizes usually less than 50 mm SL S . arawak 10b Dorsal-fin rays usually more than 80; anal-fin rays 68 or more; 66-97 scales in longitudinal series; no pepper-dots on blind side of body; caudal blotch present or absent; total vertebrae 46 or more; ID patterns usually 1-3-2 (Fig. IB) or 1-4-3 (Fig. IE); small (<45 mm SL) or large (>70 mm SL) adult sizes 11 11a Body whitish or pallid, occasionally with faint crossbands; a darkly pigmented blotch on cau- dal region of ocular side of body in some specimens; dorsal-fin rays 83-87; anal-fin rays 68-71; total vertebrae 46-48; teeth well-developed along margins of both ocular-side jaws; inner oper- cular linings and isthmus on both sides of body unpigmented; eye relatively large, ED 11.6- 15.8% of HL; ocular-side lower jaw without fleshy ridge (Fig. 3E); ID pattern usually 1-3-2 (Fig. IB); adults usually less than 45 mm SL S. rhytisma lib Body usually darkly pigmented, straw-colored to dark brown, with prominent crossbands or uniformly pigmented; no darkly pigmented caudal blotch on ocular side of body; dorsal-fin rays 86-107; anal-fin rays 70-89; total vertebrae 47-55; teeth usually absent or only poorly developed on margins of ocular-side jaws (especially upper jaw); inner opercular lining and isthmus on ocular side of body heavily pigmented; eye relatively small, ED 6.4-11.4% of HL; fleshy ridge present or absent on ocular-side lower jaw; ID patterns usually with 4 or more pterygiophores inserted into interneural space two; adults exceeding 70 mm SL 12 12a Large black spot on outer surface of ocular-side operculum; dorsal-fin rays 91-106; anal-fin rays 74-89; total vertebrae 48-54 13 12b Ocular-side operculum without obvious black spot; dorsal-fin rays 86-97; anal-fin rays 70-81; total vertebrae 46-51 14 13a From 4 to 8 small ctenoid scales on blind sides of posterior rays of dorsal and anal fins (Fig. 4A); ocular-side lower jaw without fleshy ridge on posterior portion (Fig. 3E); posterior exten- sion of ocular-side jaws reaching only to point between verticals through posterior margin of pupil and posterior margin of eye; ocular surface usually with nine, or fewer, wide crossbands; posterior one-third of dorsal and anal fins becoming progressively darker (black in mature males); dorsal and anal fins without blotches; dorsal-fin rays 91-102; anal-fin rays 74-86; total vertebrae 48-54, usually 50-53 S. tessellatus 13b No ctenoid scales on blind sides of posterior rays of dorsal and anal fins; ocular-side lower jaw with pronounced fleshy ridge on posterior portion (Fig. 3D); posterior extension of ocular-side jaws reaching vertical at posterior margin of lower eye or reaching vertical slightly posterior to posterior margin of lower eye; ocular surface with 10-14 narrow crossbands; posterior one- third of dorsal and anal fins usually without progressive posterior darkening, but with alter- nating series of blotches and unpigmented areas; dorsal-fin rays 97-106; anal-fin rays 81-89; total vertebrae 52-55, usually 53-54 S. oculellus 24 Fishery Bulletin 96 ( 1 ), 1998 14a Dorsal and anal fins with alternating series of pigmented blotches and unpigmented areas; lower jaw on ocular-side without fleshy ridge (Fig. 3E); snout pointed; distance between upper eye and dorsal-fin base only slightly greater than eye diameter; ocular surface usually with 9-15, promi- nent, narrow crossbands; eye relatively large, usually 9.0-10.0% of HL S. caribbeanus 14b Dorsal and anal fins without alternating series of pigmented blotches and unpigmented areas; lower jaw on ocular-side with fleshy ridge (Fig. 3D); snout squarish; distance from upper eye to dorsal-fin base much greater than eye diameter; ocular surface uniformly pigmented or with faint crossbands occasionally present; eye relatively small, usually only 6. 4-9. 4% of HL 15 15a Total vertebrae 47-51, usually 49-51; dorsal-fin rays 89-97; anal-fin rays 73-81; 79-89 scales in longitudinal series; eye relatively small, usually only 6. 4-9. 4% of HL S. plagusia 15b Total vertebrae 46-50, usually 47-49; dorsal-fin rays 86-93; anal-fin rays 70-78; 66-83 scales in longitudinal series; eye relatively large (7.0-11% of HL) S. civitatium 16a Caudal-fin rays usually 11; large ocellated spot on caudal fin; dorsal and anal fins without spots; pupillary operculum well developed (Fig. 3A) S. urospilus 16b Caudal-fin rays usually 10; no ocellated spot on caudal fin; if spot present on caudal fin (occa- sionally in S. diomedeanus), then spots also present on posterior dorsal and anal fins; pupillary operculum present or absent 17 17a Dark brown blotch on caudal region of ocular surface of body (see Figs. 38, 40, and 42) or single ocellated spot on posterior dorsal and anal fins (see Fig. 44); pupillary operculum present (Fig. 3A); no fleshy ridge on ocular-side lower jaw (Fig. 3E); ostia present in bases of membranes of dorsal and anal fins; ID patterns usually 1-4-2 (Fig. ID) or 1-5-2 18 17b No dark brown blotch on caudal region of ocular surface of body; no ocellated spots on posterior dorsal and anal fins; pupillary operculum and fleshy ridge on ocular-side lower jaw present or absent; no ostia in membranes at bases of dorsal and anal fins; ID patterns usually 1-3-3 (Fig. 1C) or 1-4-3 (Fig. IE) 21 18a Single ocellated spot on posterior region of dorsal and anal fins (see Fig. 44); ocular surface whitish or yellowish-white without dark brown blotch in caudal region S. ommaspilus 18b No ocellated spots on dorsal and anal fins; ocular surface straw-colored to dark brown with dark brown blotch on caudal region (see Figs. 38, 40, 42) 19 19a Dorsal-fin rays 80-87; anal-fin rays 67-72; total vertebrae 46-49, usually 47-48; scales in longitudinal series 73-81; (continental shelf off southern Brazil) S. kyaropterygium 19b Dorsal-fin rays 69-86, usually less than 83; anal-fin rays usually less than 68; total vertebrae usually less than 47; scales in longitudinal series usually less than 78; (continental shelf off eastern United States, Gulf of Mexico, and Caribbean Sea) 20 Munroe: Systematics of western Atlantic Symphurus 25 20a Dorsal-fin rays 69-81, usually 72-77; anal-fin rays 55-64, usually 56-64; total vertebrae 41-44, usually 41-43; 55-67 scales in a longitudinal series S. minor 20b Dorsal-fin rays 75-86, usually 77-84; anal-fin rays 60-70, usually 62-67; total vertebrae 43-47, usually 44-46; 59-78 scales in longitudinal series S. parvus 21a Posterior dorsal and anal fins spotted (usually); pupillary operculum present (Fig. 3A) S. diomedeanus 21b Dorsal and anal fins without spots; pupillary operculum absent or only weakly developed (Fig. 3B) 22 22a Dorsal-fin rays 107-115; anal-fin rays 91-99; scales on head posterior to lower eye 21-25; scales in longitudinal series 102-119; total vertebrae 57-60 S. jenynsi 22b Dorsal-fin rays less than 95; anal-fin rays 79 or fewer; scales on head posterior to lower eye 20 or fewer; scales in longitudinal series 86 or fewer; total vertebrae 51 or fewer 23 23a From 4 to 8 small ctenoid scales on blind sides of posterior rays of dorsal and anal fins (Fig. 4A); fleshy ridge usually present on ocular-side lower jaw (Fig. 3D); inner opercular linings and both sides of isthmus heavily pigmented; prominent black spot usually present on outer surface of ocular-side opercle; dorsal-fin rays 81-91; anal-fin rays 66-75; 76-86 scales in lon- gitudinal series; total vertebrae 44-49, usually 45-48; eye relatively small (8.3-12.6% of HL); ID pattern usually 1-4-3 (Fig. IE) S. plagiusa 23b No ctenoid scales on blind sides of posterior rays of dorsal and anal fins; ocular-side lower jaw without fleshy ridge (Fig. 3E); inner opercular lining and isthmus on ocular side of body lightly pigmented, those of blind side unpigmented; no black spot on outer surface of ocular-side opercle; dorsal-fin rays 88-94; anal-fin rays 73-79; 67-77 scales in longitudinal series; total vertebrae 47-51, usually 48-49; eye relatively large (11.4-16.2% of HL); ID pattern usually 1-3-3 (Fig. 1C) S. trewavasae 26 Fishery Bulletin 96(1 ), 1998 Species accounts Figure 10 Symphurus nebulosus (Goode and Bean), UMO 311.8, 76.7 mm SL, North Atlantic 40°48'N, 66°36'-38'W. Symphurus nebulosus (Goode and Bean, 1883| f Figs. 1A, 7, 10C, 1 1; Tables 1-11} Freckled tonguefish Aphoristia nebulosa Goode and Bean, 1883:192 (origi- nal description, counts, measurements; 421 m; Gulf Stream, 32°07'N, 78°37'30"W). Gunther, 1887:167 (based on Goode and Bean, 1883). Goode and Bean, 1896:458 (redescription, figure of holotype). Aphoristia marginata (not of Goode and Bean, 1886). Goode and Bean, 1886:154 (in part; specimen from Fish Hawk Station 1154 is S. nebulosus). Symphurus (Acedia) nebulosus. Jordan and Goss, 1889:321-323, 326-327 (new subgenus; redescrip- tion based on Goode and Bean, 1883; in key). Jor- dan and Evermann, 1898:2712 (in key; redescrip- tion of holotype; counts, measurements, color de- scription). Evermann and Marsh, 1900:332 (in key). Symphurus nebulosus . Chabanaud, 1939:26 (Atlan- tic coast of Carolinas). Chabanaud, 1952:5 (brief comparison with S. ligulatus). Ginsburg, 1951:200 (redescription, counts, measurements for three specimens; off Long Island, New York, to northern Florida). Munroe, 1990:475 (comparison with S. ligulatus). Munroe, 1992:367, 374 (ID pattern; geo- graphic, bathymetric distributions). Diagnosis Symphurus nebulosus is the only west- ern Atlantic tonguefish with the combination of a 1-2-2 ID pattern and 14 caudal-fin rays. The follow- ing combination of characteristics distinguishes S. nebulosus from all congeners, except the eastern At- lantic S. ligulatus : predominant 1-2-2 ID pattern; 14 caudal-fin rays; 5 hypurals; 57-60 (modally 58-59) total vertebrae; 105-113 dorsal-fin rays; 91-98 anal- fin rays; 120-135 scales in longitudinal series; black peritoneum; absence of pupillary operculum; absence of fleshy ridge on ocular-side lower jaw; complete row of teeth along margins of ocular-side jaws; absence of scales on blind sides of dorsal- and anal-fin rays; elongate body with relatively uniform and narrow depth (165-282 SL, usually 225-240 SL); uniformly pigmented ocular surface without caudal blotch; and uniformly pigmented dorsal, anal, and caudal fins lacking spots or blotches. Symphurus nebulosus is distinguished from S. ligulatus by modal differences in total vertebral counts, relatively longer head and wider lower head lobe, longer postorbital length, somewhat deeper body, and ocular-side upper lip usually lacking a prominent pigmented band (lightly pigmented band occasionally present). Description A diminutive species attaining maxi- mum sizes of ca. 87 mm SL. ID pattern usually 1-2-2 Munroe: Systematics of western Atlantic Symphurus 27 (22/27 specimens), rarely 2-2-2 or 1-2-3 (Table 2). Caudal-fin rays 14 (24/27), infrequently 13 or 16 (Table 3). Dorsal-fin rays 105-113, usually 107-111 (Table 4). Anal-fin rays 91-98, usually 93-98 (Table 5). Total vertebrae 57-60 (25/27), usually 58-59 (Table 6). Hypurals usually 5 (19/20 individuals), rarely 4 (1/20). Longitudinal scale rows 120-135, usually 125-130 (Table 7). Scale rows on head poste- rior to lower orbit 19-24, usually 21-22 (Table 8). Transverse scales 43-50 (Table 9). Proportions of morphometric features presented in Table 11. Body notably slender; of nearly uniform width for most of length with gradual taper in poste- rior one-fourth of body. Preanal length slightly smaller than body depth. Head long and narrow; usually slightly narrower than body depth. Head length slightly shorter than head width (HW: HL=1.00-1.28, x=1.20). Upper head lobe notably narrow and usually slightly smaller than postorbital length. Lower head lobe wide, slightly less than post- orbital length; usually projecting posteriorly beyond upper head lobe. Snout moderately short, somewhat rounded; with scaleless area on dorsal portion. Der- mal papillae well developed on blind-side snout and chin; papillae usually not extending posteriorly to vertical equal with dorsal-fin origin. Anterior nostril Table 1 1 Morphometries for holotype (MCZ 27966) and 22 other specimens of Symphurus nebulosus. (Abbreviations defined in methods section; SL is expressed in mm; characters 2 to 14 are expressed in thousandths of SL; 15 to 21 in thou- sandths of HL; n = no. of specimens measured.) Character Holotype n Range Mean SD 1. SL 76.6 23 45.0-86.2 72.6 10.18 2. BD 243 23 165-282 233.5 23.92 3. PDL 54 23 33-69 50.3 9.41 4. PAL 236 23 163-246 223.2 17.73 5. DBL 946 23 931-967 949.7 9.41 6. ABL 739 23 708-790 757.0 19.44 7. PL 57 20 46-81 65.4 8.30 8. PA 63 21 37-67 49.3 9.06 9. CFL — 21 80-116 102.4 10.42 10. HL 187 23 159-208 186.7 12.40 11. HW 215 23 186-239 216.4 13.45 12. POL 127 23 110-133 124.0 6.44 13. UHL 124 23 64-144 122.4 16.20 14. LHL 91 23 82-129 103.6 10.85 15. POL 678 23 620-711 665.6 26.87 16. SNL 189 23 160-248 208.7 22.91 17. UJL 182 22 169-248 206.9 20.23 18. ED 133 23 94-133 114.2 11.50 19. CD 147 23 141-308 210.0 38.56 20. OPLL 329 23 233-479 358.4 49.47 21. OPUL 126 23 126-253 188.7 35.64 on ocular side short when depressed posteriorly, not reaching anterior margin of lower eye. Jaws short, slightly arched; maxilla usually extending posteri- orly to vertical through anterior margin of pupil of lower eye; less frequently, extending posteriorly to point between verticals through middle and anterior margin of lower eye. Ocular-side lower jaw without fleshy ridge. Teeth well developed on jaws. Margins of ocular-side dentary and premaxilla usually with complete row of slender teeth; less frequently, with teeth only on anterior two-thirds of premaxillary margin. Chin depth usually just slightly larger than snout length. Lower eye small; subelliptical; eyes usually equal in position or upper eye slightly in advance of lower eye. Anterior and medial surfaces of eyes usually without scales; 1-3 small scales usu- ally in narrow interorbital space. Pupillary opercu- lum absent. Dorsal-fin origin posteriorly placed, usu- ally almost equal with vertical through middle of upper eye; less frequently reaching vertical through anterior margin of pupil, or occasionally reaching vertical through posterior margin of upper eye; predorsal length moderate. Anteriormost dorsal-fin rays shorter and with wider separation between bases than more posterior fin rays. Scales absent on blind sides of dorsal- and anal-fin rays. Pelvic fin moderately long; longest pelvic-fin ray, when ex- tended posteriorly, usually reaching base of first, sometimes second, anal-fin ray. Posteriormost pel- vic-fin ray connected to body by delicate membrane terminating immediately anterior to anus, or occa- sionally extending posteriorly almost to anal-fin ori- gin (membrane torn in specimens examined). Cau- dal fin short. Scales small, numerous, strongly ctenoid on both sides of body. Pigmentation (Fig. 10J Body coloration generally similar for both sexes. Ocular surface of head and body almost always uniformly straw-colored to dark brown, sometimes with overlying pattern of ill-de- fined dark brown cloudy areas, but otherwise with- out distinctive markings. Abdominal area immedi- ately posterior to opercular opening, sometimes darker than general body color. Individual scales on ocular side have underlying longitudinal streak of black pigment; about 40-60 longitudinally continu- ous pigmented streaks along ocular-side length. Outer surface of ocular-side opercle generally with similar background coloration as on body; inner lin- ings of opercles and isthmus on both sides of body yellowish-white. Slight pigment band occasionally present on ocular-side upper lip (not in most speci- mens examined). Blind side off-white; usually with median line of internal black spots showing through skin along axis of vertebral column. Smaller speci- 28 Fishery Bulletin 96(1 ), 1998 mens typically with single longitudinal series of dark, internal spots on blind side of body at proximal ends of dorsal- and anal-fin pterygiophores. Peritoneum dark black, showing through abdominal wall on both sides, especially prominent in lightly pigmented and smaller specimens. Fin rays of dorsal and anal fins uniformly light brown along length of fins with little, if any, pigment on fin membranes and without obvious pigmented blotches or spots. Proximal one-third of caudal fin with diffuse brown pigment similar to that on body; distal portion of caudal-fin rays usually unpigmented. Goode and Bean (1883) described the holotype as grayish, everywhere mottled with brown. Ginsburg (1951) described three specimens he examined as partly faded and having almost uniformly reddish or yellowish-brown coloration. Size and sexual maturity {Fig. 7CJ Symphurus rtebu- losus is a diminutive species that attains a maximum size of ca. 87 mm. Ten males (64.3-86.2 mm), 12 fe- males (54.9-83.8 mm), and two immature fish (36.4- 45.0 mm) of indeterminate sex were examined for life history information. No size differences between sexes were apparent. Sexual maturity, based on fe- males, occurs at ca. 60-65 mm (Fig. 7C). Females 63.8-83.8 mm were gravid; two others (54.9 and 68.1 mm) had elongate ovaries without evidence of ripen- ing ova. Of 10 specimens taken in a single trawl (now divided between UNC 4951 and USNM 285758), six were gravid females (69.5-83.8 mm) and four were adult males (78.3-86.2 mm), indicating that adults of both sexes occupy similar habitats. Gravid females were present in collections from lat. 40°N to 28°N and from 239 to 800 m depth, indi- cating spawning probably occurs throughout the en- tire geographic and bathymetric ranges of the spe- cies. The two smallest specimens examined (36.4 and 45.0 mm) were taken near the northern end of the geographic range (39°40'N and 39°55’N, respectively). The only other specimen (UMO 311.8; 76.7 mm) from this region (at 40°48'N) is a gravid female. Capture of both juveniles and adults at the north of the range indicates that this region has an adult population of S. nebulosus and is likely not populated only by lar- vae that have been transported by the Gulf Stream from more southern regions, as reported for S. minor (Markle et al., 1980) and other Symphurus species (Evseenko, 1982; Scott and Scott, 1988). Geographic distribution fFig. 1 1 1 Western North Atlantic in outer continental shelf and upper conti- nental slope deep waters from the region just south of Long Island, New York (40°48'N), to the Blake Pla- teau off Fort Lauderdale, Florida (26°28'N). Figure 1 1 Geographic distribution of Symphurus nebulosus based on mate- rial examined (discussion of geographic distribution appears in species account). Bathymetric distribution This species has rarely been collected (only 27 specimens located). Specimens were captured on soft mud substrates at depths rang- ing from 239 (USNM 291326, two individuals) to 810 m (UMML 20746, two specimens); most between 400- 600 m. Six of 27 specimens were taken deeper than 530 m (Table 10). Because of its rarity in collections, little is known about the biology of S. nebulosus. Most captures (8/12) were of single specimens. The larg- est collection of 10 specimens occurred east of Cape Fear, North Carolina, at 495 m. Whether rarity in collections reflects natural low abundance or diffi- culties in sampling such relatively small fishes at depths where this species occurs is unknown. Remarks Historically, comparisons of S. nebulosus have been made with S. ligulatus (Chabanaud, 1952; Munroe, 1987, 1990), a deepwater species of the Mediterranean Sea and eastern North Atlantic, which, as first noted by Chabanaud (1952) in a brief footnote, has meristic features similar to those of S. nebulosus. On the basis of similarities in meristic and morphological features (Munroe 1987, 1990), shared osteological characters (Munroe, unpubl. data), and similarities in pigmentation, it is hypoth- esized that S. nebulosus and S. ligulatus represent a closely related species pair with distributions in bathyal depths of temperate waters on opposite sides of the northern Atlantic. Both belong to a larger spe- cies group characterized by several features includ- ing a 1-2-2 ID pattern (Munroe, 1992). Members of this group occur primarily at bathyal depths through- out the Indo-Pacific Ocean. Munroe: Systematics of western Atlantic Symphurus 29 Comparisons Among western Atlantic tongue- fishes, S. nebulosus overlaps only S.jenynsi in some meristic features but can easily be distinguished from the latter by its 14 caudal-fin rays (vs. 10 in S. jenynsi), black peritoneum, and ocular-side pigmen- tation without pattern of crossbanding (vs. unpig- mented peritoneum and ocular surface usually with pattern of crossbanding in S.jenynsi), in having com- plete dentition on both ocular-side jaws (reduced dentition on ocular-side premaxilla and teeth absent on ocular-side dentary in S. jenynsi), more numer- ous scales in longitudinal series (120-135 vs. 102- 119 in S. jenynsi), its different ID pattern (1-2-2 vs. 1-4-3 in S. jenynsi), its larger eye (94-113 vs. 74-95 in S. jenynsi), in having 5 hypurals (vs. 4 in S. jenynsi), and by its much narrower (BD 165-282 SL in S. nebulosus vs. 231-328 SL in S. jenynsi) and smaller body (maximum sizes ca. 87 mm in S. nebulosus vs. >300 mm in S. jenynsi). Body depth measurements in S. nebulosus approxi- mate those in juvenile and small adultS. marginatus, another deepwater western Atlantic species (see be- low). However, S. nebulosus lacks the conspicuous dark brown blotch on the ocular-side caudal region of the body and posterior dorsal and anal fins and also lacks the stripe along basal margins of the dor- sal and anal fins (both features present in S. marginatus). In addition, S. marginatus differs in having 12 caudal-fin rays, fewer total vertebrae (51- 56), dorsal-fin rays (93-104), anal-fin rays (80-89), and fewer scales in a longitudinal series (86-99), its larger eye (125-248 HL), and ID pattern (1-3-2 in S. marginatus). Of tonguefishes occurring outside the western At- lantic Ocean, S. nebulosus most closely resembles two deepwater species: S. variegatus, known only from two specimens taken in the western Indian Ocean off South Africa; and S. ligulatus, a species in the eastern Atlantic and Mediterranean Sea that possi- bly forms an amphi-Atlantic species pair with S. nebulosus. All three species have similar numbers of caudal-fin rays, slender bodies, and possess a 1-2-2 ID pattern. Symphurus nebulosus is distinguished from S. variegatus in having more vertebrae (57-60, usually 58-59 vs. 56 in S. variegatus). Within the genus, S. nebulosus is most similar to S. ligulatus. Comparisons of S. ligulatus and S. nebulosus re- vealed subtle but significant differences between these two species in 8 of 14 morphometric charac- ters examined (Munroe, 1990). According to results from a discriminant function analysis (DFA), notable differences occur in postorbital and head lengths (both longer in S. nebulosus), lower head lobe width (wider in S. nebulosus), and body depth, which is somewhat deeper in S. nebulosus (BD 16.5-28.2% SL, but usually 22.5-24.0% SL vs. 19.4-23.8%, but usually 21.0-22.0% SL in S. ligulatus). Symphurus nebulosus , despite having a more darkly pigmented body in general, has only a slight band of pigment, if any, on the ocular-side upper lip. In contrast, S. ligulatus specimens generally have a well-developed band of pigment on both ocular-side lips. The spe- cies also differ in modal counts of total vertebrae (58 — 59 in S. nebulosus vs. 59-60 in S. ligulatus). Symphurus nebulosus is readily distinguished from S. vanmelleae, a deepwater tropical eastern Atlantic species, in caudal-fin rays (14 vs. 12 in S. vanmelleae), vertebrae (9 abdominal vertebrae, 57-60, modally 58-59, total vertebrae vs. 10-11 abdominal and 56- 58 total vertebrae in S. vanmelleae), modally higher meristic features (dorsal-fin rays 105-113 in S. nebulosus vs. 101-108 in S. vanmelleae', anal-fin rays 91-98 vs. 86-93), and ID patterns (1-2-2-2-2 vs. 1-2- 2-1-2 or 1-2-2-2-1 in S. vanmelleae). 30 Fishery Bulletin 96(1 ), 1998 Figure II 2 Symphurus arawak Robins and Randall, USNM 267784, 29.3 mm SL, off Belize 16°48'N, 88°04'W. Symphurus arawak Robins and Randall, 1 965 (Figs. 6A, 12, 1 3; Tables 1-10, 12J Caribbean tonguefish Symphurus arawak Robins and Randall, 1965:331 (original description with photograph; Curasao). Bohlke and Chaplin, 1968:223 (Bahamas; diagno- sis, counts, figure, distribution). Randall, 1968:166 (Caribbean distribution). Starck, 1968:31 (Alliga- tor Reef, Florida). Topp and Hoff, 1972:107 (distri- bution). Garzon and Acero, 1983:106 (Caribbean Sea, Colombia; counts, measurements, photo- graph). Munroe, 1990:509 (maximum size and size at maturity). Munroe, 1992:368, 377 (IB pattern; geographic, bathymetric distributions). Cervigon et al., 1993:306 (Venezuela). Diagnosis Symphurus arawak is a distinctive spe- cies characterized by a predominant 1-3-2 ID pat- tern; 12 caudal-fin rays; 39-42 total vertebrae; 70- 76 dorsal-fin rays; 55-61 anal-fin rays (the lowest vertebral and fin-ray counts of any species in the genus); 55-65 scales in longitudinal series; unpig- mented peritoneum; teeth along entire margin of ocular-side jaws; without pupillary operculum; with- out scales on blind sides of dorsal- and anal-fin rays; without fleshy ridge on ocular-side lower jaw; with- out membrane ostia in dorsal and anal fins; ocular surface pigmentation consisting of crossbands and, in some adults, of a darkly pigmented caudal blotch; with pepper-dots (Fig. 5B, number 8) on blind side of body (especially prominent on larger adults); dorsal, anal, and caudal fins without spots, but some adults with darkly pigmented blotch on posterior dorsal and anal fins; and small adult size (usually <50 mm). Description Symphurus arawak is a dwarf species reaching maximum sizes of only ca. 50 mm SL. ID pattern usually 1-3-2 (35/41 specimens), rarely 1-2-2 or 1-3-3 (Table 2). Caudal-fin rays usually 12 (38/ 43), rarely 11, 13, or 14 (Table 3). Dorsal-fin rays 70- 76 (Table 4). Anal-fin rays 55-61 (Table 5). Total ver- tebrae 39-42, usually 40-41 (38/45) (Table 6). Hypurals 4 (45/45). Longitudinal scale rows 55-65, usually 58-64 (Table 7). Scale rows on head poste- rior to lower orbit 12-15, usually 14-15 (Table 8). Transverse scales 27-36, usually 32-36 (Table 9). Proportions of morphometric features presented in Table 12. Body relatively deep, of stocky build with greatest depth in anterior one-third of body; body depth tapering rapidly posterior to midpoint. Prea- nal length slightly smaller than body depth. Head long and wide; somewhat shorter than body depth. Head length usually slightly shorter than head width (HW:HL 0.9-1. 3, 3c = 1 . 1 ). Postorbital length consid- erably smaller than body depth. Lower head lobe width nearly equalling postorbital length; slightly smaller than width of upper head lobe. Lower oper- cular lobe of ocular side considerably wider than upper opercular lobe. Snout long and pointed; cov- Munroe: Systematics of western Atlantic Symphurus 31 Table 1 2 Morphometries for holotype (ANSP 101985) and 39 other specimens of Symphurus arawak. (Abbreviations defined in methods section; SL is expressed in mm; characters 2 to 14 are expressed in thousandths of SL; 15 to 21 in thou- sandths of HL; n = no. of specimens measured.) Character Holotype n Range Mean SD 1. SL 33.1 40 11.7-49.3 28.2 8.18 2. BD 254 40 254-377 317.2 24.98 3. PDL 94 39 49-145 93.1 15.12 4. PAL 302 40 188-350 300.2 30.96 5. DBL 909 40 649-941 903.3 44.33 6. ABL 695 40 495-776 687.0 39.75 7. PL 76 36 60-100 83.6 9.88 8. PA 66 40 43-99 66.4 10.53 9. CFL 160 34 120-204 169.6 17.99 10. HL 275 40 182-299 265.8 20.58 11. HW 299 40 225-316 282.0 17.04 12. POL 169 40 114-188 156.6 15.96 13. UHL 184 40 109-192 153.3 15.23 14. LHL 130 40 110-174 140.0 14.73 15. POL 615 40 500-660 588.7 34.46 16. SNL 198 40 163-307 224.2 32.90 17. UJL 231 40 200-333 235.6 23.42 18. ED 132 40 97-200 144.1 19.71 19. CD 165 40 132-347 204.8 36.14 20. OPLL 253 39 229-379 288.5 34.64 21. OPUL 220 39 159-303 230.8 30.17 ered to tip with small ctenoid scales; scales not em- bedded, but rather deciduous. Dermal papillae well developed on snout and chin of blind side. Anterior nostril on ocular side moderately long, almost reach- ing anterior margin of lower eye when depressed posteriorly. Jaws moderately long; maxilla usually extending posteriorly to vertical through middle, or sometimes only to vertical through anterior margin of pupil of lower eye, rarely only reaching vertical through anterior margin of lower eye. Ocular-side lower jaw without fleshy ridge. Teeth well developed on all jaws. Ocular-side dentary usually with com- plete row of slender teeth; less frequently teeth present only on anterior three-fourths of dentary. Ocular-side premaxilla with single row of teeth on anterior three-fourths of margin of upper jaw. Chin depth nearly equal to length of snout. Lower eye large; eyes usually equal in position. Anterior and medial surfaces of eyes not covered with scales; usu- ally with 1 or 2, occasionally 3, small scales in nar- row interorbital region. Pupillary operculum absent. Dorsal-fin origin usually reaching point between ver- ticals through anterior margin and midpoint of up- per eye; predorsal length long. No scales on blind sides of dorsal- and anal-fin rays. Pelvic fin long; long- est pelvic-fin ray, when extended posteriorly, usually reaching base of first anal-fin ray. Posteriormost pel- vic-fin ray connected to body by delicate membrane terminating immediately anterior to anal-fin origin (membrane tom in many specimens). Caudal fin long. Scales large, strongly ctenoid on both sides of body. Pigmentation (Fig. 1 2} Coloration similar for both sexes. Ocular surface usually off-white or pale yel- lowish, with about one-half of individuals with 2-7 (usually 4-5), conspicuous, dark brown, complete crossbands on body that sometimes extend onto fin rays. Other specimens with short, incomplete crossbands forming 6-10 large, and variably posi- tioned, dark brown blotches on body; blotches best developed on caudal one-third of body. Blotches (rarely large ovoid spots) in midbody region some- what offset, best developed on body in dorsal and ventral regions at bases of dorsal and anal fins. Each scale on body and head with numerous small mel- anophores, but background coloration pale in con- trast to dark crossbands or blotches on body. Mel- anophores more heavily concentrated on scales in caudal one-third of body, forming dark caudal patch in some specimens. Crossbands on body, relatively wide, usually 4-8 scale rows wide, and beginning immediately posterior to opercular opening and con- tinuing to base of caudal fin. Posteriormost pair of crossbands usually conjoined, forming darkly pig- mented, M- or Y-shaped mark near point approxi- mately one-third distance between caudal-fin base and opercular opening. Most specimens with narrow, dark, vertical bar extending from upper eye to dor- sal profile, otherwise head with same background coloration as found on body. Dorsal margin of outer surface of ocular-side opercle with small dark spot near opercular opening, but otherwise with same general background pigmentation as body. Inner lin- ings of both opercles and isthmus on both sides of body unpigmented. Band of pigmentation of variable intensity usually developed on ocular-side upper lip; lower lip on ocular side without pigmented band. Dark spot usually present in posterior angle of ocu- lar-side jaws. Blind side of body in approximately one-half of specimens (mostly those larger than ca. 20 mm) with small pepper-dots extending variable distances along trunk, but usually best developed and most heavily concentrated in region overlying proxi- mal pterygiophores of dorsal- and anal-fin rays and covering entire caudal one-third of body (Fig. 5B). Peritoneum unpigmented. Anteriormost dorsal and anal fins without obvi- ous pattern of spots or blotches, but with dark brown melanophores along length of each finray; melano- phores becoming increasingly darker, almost dark brown or black, and more heavily concentrated in 32 Fishery Bulletin 96(1 ), 1998 posterior one-third of body. Posteriorly progressive in- crease in density of melanophores on fin rays through- out length of dorsal and anal fins, sometimes forming small blotches in area of fins proximate to body regi ons with blotches or crossbands. Melanophores particularly dense on dorsal- and anal-fin rays in caudal one-third of body, making posterior fin rays usually strikingly darker than those in anterior regions of fins. Caudal fin dark brown or black throughout entire length. Size and sexual maturity (Fig. 6AJ Symphurus arawak attains a maximum size of only ca. 50 mm (Fig. 6A) and is among the smallest of flatfishes (see “Discussion” section below). Most individuals were 25-40 mm, and only 3/42 exceeded 40 mm. Maximum size of females (49.3 mm) was somewhat larger than that of males (38.0 mm). Diminutive size is reflected in this species’ ontogeny, as juveniles 11.7, 13.2, and 13.9 mm had already metamorphosed and assumed a benthic existence. Specimens examined included 7 immature fish (sex undetermined, 11.7-28.0 mm), 10 males (23.3-38.0 mm), and 29 females (24.0-49.3 mm). Sexual matu- rity of females occurs at ca. 25-30 mm. Most females 24-30 mm had elongate ovaries, but only some had developing ova. Females larger than 30 mm had elon- gate ovaries that were either gravid or contained developing ova. The smallest gravid female was 30 mm. Nine females between 21-30 mm were imma- ture with ovaries undergoing posterior elongation without visible ova. Geographic distribution (Fig. 13) Throughout the Caribbean Sea, extending south and west to Isla de Tierra Bomba, Colombia; and a single capture at Al- ligator Reef, Florida (Starck, 1968). The majority of specimens were collected in the Bahamas and insu- lar Caribbean regions including Curasao, Dominica, Haiti, Jamaica, Puerto Rico, Providencia Island, and Cayman Islands. Robins and Randall (1965) reported this species at St. John, Virgin Islands. In addition to capture at Alligator Reef, other specimens were collected at continental reef areas at Belize and Cabo de la Aguja (one specimen) and Bahia de Gayraca (three specimens), Colombia (Garzon and Acero, 1983). Bathymetric distribution The majority of S. arawak have been captured on sandy bottoms in clear wa- 90° 85° 80° 75° 70° 65° 60° Figure 1 3 Geographic distribution of Symphurus arawak based on material examined (discussion of geographic distribution appears in species account). Munroe: Systematics of western Atlantic Symphurus 33 ters adjacent to coral reefs at depths of 6-39 m, most between 6 and 30 m (Table 10). Only three speci- mens were taken deeper than 30 m (one each at 31, 34, and 39 m). Little else is known concerning the life history of this diminutive flatfish. Remarks Robins and Randall (1965) tentatively proposed that S. arawak and S. minor are a north- south species pair occurring in tropical and warm temperate regions of the western Atlantic that may have differentiated because of repeated latitudinal fluctuations in the fish fauna during glacial and in- terglacial periods in the western Atlantic. This hy- pothesis is not supported by information presented in Munroe (1987; 1992), who has hypothesized that S. arawak and S. minor belong to different species groups within the genus. Symphurus arawak is a member of the species group characterized by a 1-3-2 ID pattern, 12 caudal-fin rays, lack of a pupillary operculum, and lack of ostia in basal portions of the dorsal- and anal-fin membranes. Symphurus minor , in turn, belongs to the species group characterized by a 1-4-2 ID pattern, 10 caudal-fin rays, and pos- sessing a well-developed pupillary operculum and membrane ostia. Similarities between S. arawak and S. m inor in meristic features and in small adult sizes do not appear to be synapomorphies that reflect a common ancestry but, rather, are probably con- vergently evolved traits. Comparisons Symphurus arawak has the lowest counts for any species in the genus (Munroe, 1992). Among tonguefishes, only the western Atlantic S. minor and the sympatrically occurring S. ommas- pilus approach the ranges in meristic features and small size of S. arawak. However, S. arawak differs significantly from these species in caudal-fin ray counts (12 vs. 10 in the others), absence of a pupil- lary operculum and membrane ostia (both present in S. minor and S. ommaspilus), a pepper-dot pat- tern of melanophores on the blind side of the body (absent in S. minor and S. ommaspilus ), and S. arawak has modally fewer vertebrae (40-42 vs. 41- 43 in S. minor and 43-44 in S. ommaspilus), and a different ID pattern (1-3-2 vs. 1-4-2 in S. minor and S. ommaspilus, respectively). Symphurus arawak differs further from S. ommaspilus in having body crossbands or a darkly pigmented blotch on the ocu- lar side and no spots on the dorsal and anal fins (vs. uniformly whitish body without crossbands or pig- mented blotch but having a single ocellated spot on dorsal and anal fins of S. ommaspilus). Symphurus arawak is similar to three other At- lantic dwarf species, the sympatric S. rhytisma and two dwarf species, S. lubbocki Munroe and S. reticulatus Munroe, occurring on shallow-water sub- strates at midocean and eastern Atlantic islands. Symphurus arawak is easily separated from these others because of its lower counts (dorsal-fin rays 70-76 vs. 82-89; anal-fin rays 55-61 vs. 68-75, and total vertebrae 39-42 vs. 46-49 in these others). Symphurus arawak also has considerably larger and fewer (55-65) scales in a longitudinal series com- pared with those of S. rhytisma (91-97), S. lubbocki ( 107-109), and S. reticulatus ( 101-109). None of the other dwarf Symphurus have pepper-dot pigmenta- tion on the blind side, and for each of these other species the pattern of pigmentation on the ocular surface also differs significantly from that of S. arawak. In S. rhytisma, ocular-side pigmentation consists of crossbands on the trunk, with usually the two posteriormost crossbands coalesced and forming a heavily pigmented caudal patch. Symphurus lubbocki has a cream-colored ocular surface with sev- eral, mostly incomplete, crossbands, whereas the ocular-side pattern of S. reticulatus is dark, choco- late-brown with alternating X- and Y-shaped mark- ings, and in this species the dorsal and anal fins also have an alternating series of blotches and unpigmented areas (vs. uniformly pigmented areas and intensified pigmentation in caudal region in S. arawak). Symphurus pelicanus is a diminutive ( usually <70 mm), relatively deepwater, western Atlantic tongue- fish that, reminiscent of S. arawak, also has a pep- per-dot pattern of melanophores on the blind side of the body, 12 caudal-fin rays, and a 1-3-2 ID pattern. Symphurus arawak is easily distinguished from S. pelicanus by its unpigmented peritoneum (vs. black in S. pelicanus), lower meristic features (vs. dorsal- fin rays 77-85, anal-fin rays 64-70, and total verte- brae 43-46 in S. pelicanus), in having unpigmented inner opercular linings and isthmus (vs. inner oper- cular linings and isthmus sprinkled with melano- phores in S. pelicanus), and S. pelicanus lacks the pigmented blotch on the ocular side of the body that is present on some adult S. arawak. 34 Fishery Bulletin 96(1 ), 1998 Figure 14 Symphurus rhytisma Bohlke, FMNH 94821, 21.7 mm SL, Belize at Glovers Reef. Symphurus rhytisma Bohlke, 1 96 1 (Figs. 6B, 14, 15; Tables 1-10, 13) Patchtai! tonguefish Symphurus rhytisma Bohlke, 1961:3 (original de- scription with photograph; Bahamas). Robins and Randall, 1965:334 (Curasao, Lagoen; counts, mea- surements). Bohlke and Chaplin, 1968:224 (distri- bution, redescription, counts, figure). Munroe, 1990:485, 488 (comparison with eastern and mid- Atlantic species of Symphurus ). Munroe, 1992:368, 377 (IB pattern; geographic, bathymetric distri- butions). Symphurus plagusia (not of Schneider, in Bloch and Schneider, 1801). Andreata and Seret, 1995:590 (two specimens; inner continental shelf Brazil). Diagnosis Symphurus rhytisma is readily distin- guished from all congeners by the combination of 1-3-2 ID pattern; 12 caudal-fin rays; 83-87 dorsal- fin rays; 68-71 anal-fin rays; 46-48 total vertebrae; 91-97 scales in longitudinal series; unpigmented peritoneum; absence of pupillary operculum; absence of fleshy ridge on ocular-side lower jaw; teeth along entire margin of ocular-side jaws; absence of scales on blind sides of dorsal- and anal-fin rays; absence of membrane ostia in dorsal and anal fins; unpig- mented blind side; ocular surface pigmentation gen- erally consisting of several incomplete, dark brown crossbands contrasting against pallid background and of darkly pigmented blotch covering posterior one-third of ocular side of body (in all but largest specimens); with dorsal and anal fins unpigmented anteriorly, with pigmented blotches in midbody re- gion, and heavy pigmentation on both sides of fins in caudal region; and small adult size (<50 mm). Description Symphurus rhytisma is a dwarf spe- cies, attaining maximum size of only ca. 45 mm SL. ID pattern 1-3-2 (Table 2). Caudal-fin rays 12 (Table 3). Dorsal-fin rays 83-87 (Table 4). Anal-fin rays 68- 71 (Table 5). Total vertebrae 46-48, usually 47 (Table 6). Hypurals 4 (9/9). Longitudinal scale rows 91-97 (Table 7). Scale rows on head posterior to lower orbit 18-20, usually 18 (Table 8). Transverse scales 43-45 (Table 9). Proportions of morphometric features of specimens from Caribbean and Brazilian localities appear sepa- rately in Table 13 (description is based on Caribbean material only because the Brazilian specimens were in poor condition and measurements were imprecise and unreliable). Body moderately deep, maximum depth in anterior one-third of body; body depth ta- pering fairly moderately posterior to anus. Preanal length slightly shorter than body depth. Head long and narrow; head width considerably narrower than body depth. Head length usually just slightly shorter than head width (HW:HL 1.0-1. 3, ic=1.09). Postor- bital length considerably shorter than body depth. Lower head lobe moderately wide, less than postor- bital length; narrower than upper head lobe. Lower opercular lobe of ocular side considerably wider than Munroe: Systematics of western Atlantic Symphurus 35 Table 13 Morphometries for holotype ( ANSP 93812)and seven other specimens of Symphurus rhytisrna. Data for Caribbean specimens {n =6) listed above those for Brazilian specimens (n= 2). (Abbreviations defined in methods section; SL is expressed in mm; characters 2 to 14 are expressed in thou- sandths of SL; 15 to 21 in thousandths of HL; n = no. of specimens measured.) Character Holotype n Range Mean SD 1. SL 25.6 6 21.7-45.1 31.5 8.71 2 26.7-27.0 26.8 — 2. BD 297 6 295-333 302.4 17.46 2 243-266 254.5 — 3. PDL 78 6 69-78 72.0 3.44 2 64-67 65.5 — 4. PAL 262 6 262-295 274.1 13.25 2 262-307 284.5 — 5. DBL 938 6 929-960 937.2 11.66 2 933-936 934.5 — 6. ABL 738 6 719-767 737.0 17.70 2 730-772 751.0 — 7. PL 78 6 70-83 78.2 4.60 1 82 — — 8. PA 66 6 35-66 54.2 12.19 2 Not available 9. CFL 113 6 104-143 119.4 13.32 1 139 — — 10. HL 262 6 203-262 235.1 21.75 2 243-262 252.5 — 11. HW 262 6 241-264 256.3 11.00 2 236-277 256.5 — 12. POL 160 6 126-160 148.0 11.92 2 150-157 153.5 — 13. UHL 152 6 134-166 154.0 11.04 2 161-176 168.5 — 14. LHL 125 6 108-131 117.2 8.78 2 112-120 116.0 — 15. POL 612 6 588-773 633.2 69.29 2 600-615 607.5 — 16. SNL 224 6 206-250 228.0 15.92 2 200-231 215.5 — 17. UJL 224 6 175-250 214.8 27.30 2 231-243 237.0 — 18. ED 119 6 108-165 136.2 22.47 2 123-157 140.0 — 19. CD 239 6 213-277 238.3 22.82 2 171-200 185.5 — 20. OPLL 269 6 238-325 280.0 34.76 2 229-262 245.5 — 21. OPUL 194 6 159-268 209.2 38.35 2 154-171 162.5 — upper opercular lobe. Snout moderately long and pointed; mostly naked. Scales, when present on snout, deciduous and only in areas where dermal papillae absent. Dermal papillae dense, large and obvious on snout regions of both sides of body; ex- tending onto chin region of blind side. Anterior nos- tril on ocular side long, usually reaching anterior margin of lower eye when depressed posteriorly. Jaws moderately long; maxilla usually extending posteri- orly to vertical through anterior margin of pupil of lower eye; occasionally reaching vertical through midpoint of lower eye. Ocular-side lower jaw with- out fleshy ridge. Teeth well developed on all jaws. Blind-side jaws with small band of teeth on both upper and lower jaws. Ocular-side jaws usually with single row of teeth along complete margin of jaw; occasionally with teeth present only on anterior three-fourths of ocular-side premaxilla and dentary. Chin depth slightly shorter than snout length. Lower eye moderate in size; eyes usually equal in position, or with upper eye slightly in advance of lower eye. Anterior and medial surfaces of eyes usually not cov- ered with scales; 0-4 small ctenoid scales in narrow interorbital region. Pupillary operculum absent. Dorsal-fin origin usually equal with vertical through midpoint of upper eye; predorsal length long. Scales absent on blind sides of dorsal- and anal-fin rays. Pelvic fin moderately long; longest pelvic-fin ray, when extended posteriorly, usually reaching base of first anal-fin ray. Posteriormost pelvic-fin ray con- nected to body by delicate membrane terminating immediately anterior to anus, or occasionally extend- ing posteriorly almost to origin of anal fin (membrane torn in some specimens). Caudal fin moderately long. Scales ctenoid, relatively small. Pigmentation (Fig. 14) Coloration similar for both sexes. Ocular surface generally pallid, usually with traces of 2-8 (usually 8) incomplete, narrow, brown crossbands on head and body. Smaller individuals with conspicuous dark blotch on caudal region of body. Larger individuals with more diffuse and less well-defined caudal blotch; therefore this pigmenta- tion feature may be better developed and more char- acteristic of juveniles. Crossbands usually beginning on head about at level equal with fifth dorsal-fin ray, and continuing at irregular intervals to base of cau- dal fin. Two anteriormost crossbands often incom- plete and faintly pigmented, barely perceptible with- out magnification. Third crossband at, or slightly posterior to, anal-fin origin, most often across entire body and usually darkest of anteriormost crossbands. Posteriormost two crossbands on trunk usually con- joined, forming caudal blotch. Only crossbands in middle of body (usually fifth, sixth, and eighth), if any, extending onto fin rays of dorsal and anal fins. Head usually with two faintly pigmented crossbands about 3-4 scale rows wide; anteriormost crossband immediately posterior to eyes; posterior pigment band crossing distal margin of operculum. Ocular- side outer opercle usually with same background coloration as body. Inner linings of opercles and isth- 36 Fishery Bulletin 96( I ), 1 998 mus on both sides of body unpigmented. No pigment evident on ocular-side lips. Blind side uniformly pale, off-white. Some specimens with single, median line of black dermal spots showing through skin along axis of vertebral column on blind side (Fig. 5B, num- ber 9). Peritoneum unpigmented. Dorsal and anal fins unpigmented anteriorly, fins in midregion of body with pigmented blotches (ex- tensions of body crossbands onto fins); dorsal and anal fins becoming increasingly pigmented in cau- dal region. In smaller specimens, especially, fin rays and membranes on both sides of vertical fins in re- gion of caudal blotch heavily pigmented. Posterior- most dorsal- and anal-fin rays with pigment concen- trated on proximal one-half of finrays forming a dif- fuse dark blotch on fins. Proximal one-third of cau- dal fin usually covered with dark melanophores; pos- terior two-thirds of fin unpigmented. Size and sexual maturity (Fig. 6B) Of nine speci- mens studied, five were males (25.7-36.6 mm), two females (32.7, 45.1 mm), and two juveniles (21.7, 25.6 mm) of indeterminate sex. Males and females are somewhat similar in size. The largest female was gravid with obvious ova present throughout the elon- gate ovary. A second female (32.7 mm) was mature, or was approaching maturity, because its ovaries were elongate, but without any obvious ova. The small sizes (ca. 33-45 mm) at which this species reaches sexual maturity indicate that S. rhytisma is a dwarf spe- cies (Fig. 6B). Geographic distribution (Fig. 15) Known mostly from Caribbean region of western North Atlantic with two specimens tentatively identified as S. rhytisma (see “Remarks” section below) collected off Espirito Santo, Brazil (between 20°S and 21°S latitude). This species has not been collected very frequently (only nine' specimens located for this study). Of six Carib- bean collections, three were made in the Bahamas, and the others at Glovers Reef, Belize, and Curafao. Bathymetric distribution All but one collection of S. rhytisma from the Caribbean region were made at stations treated with rotenone on sandy substrates adjacent to coral reefs. For six specimens with depth of capture information, two were from 6 m, and single specimens were collected at each of the following depths: 3, 14, 16, and 25 m (Table 10). Two speci- mens captured off Brazil (reported as S. plagusia ) were taken by trawling at 37 and 97 m (Andreata and Seret, 1995). Remarks Two specimens (USU 1054 and 1079) col- lected off Brazil are tentatively identified as S. Figure 1 5 Geographic distribution of Symphurus rhytisma based on mate- rial examined (discussion of geographic distribution appears in species account). rhytisma. Both are small (26.7 and 27.0 mm) males. Meristic features, including ID pattern (both 1-3-2), total vertebral counts (47; 47), dorsal-fin rays (83; 83), and anal-fin rays (70; 71) lie within ranges re- ported for S. rhytisma (see Tables 4-9), as do most morphometric features (Table 13). Scale counts or counts of scale pockets could not be taken from ei- ther specimen because scales were missing and the skin was abraded in many places. Coloration of both specimens is uniformly yellowish-whitish on the ocu- lar surface, in agreement with most other specimens studied. However, neither has any trace of the pig- mented caudal patch found on some other S. rhytisma specimens. One specimen (USU 1079) has some dark pigment on the posteriormost regions of the dorsal and anal fins that may be the remnants of a caudal blotch. Both specimens have a series of dermal melanophores along the vertebral column in the body midregion that is also found in other S. rhytisma specimens. A 25.7-mm specimen (USNM 324677, from 15°42'N, 63°38'W near Isla de Aves in the Lesser Antilles) included in the material examined listed for S. pelicanus in Munroe (1992:402) is actually S. rhytisma. Munroe: Systematics of western Atlantic Symphurus 37 Comparisons In the western Atlantic and Carib- bean region, only S. rhytisma and S. arawak possess the combination of a 1-3-2 ID pattern, 12 caudal-fin rays, unpigmented peritoneum, and small adult size. Symphurus rhytisma is readily diagnosed from S. arawak by differences outlined in the “Comparisons” section of the account for S. arawak. Symphurus rhytisma has some meristic features that overlap those of two deepwater, western Atlan- tic species, S. pusillus and S. piger. Although both have a 1-3-2 ID pattern and 12 caudal-fin rays as does S. rhytisma , they are easily separated from S. rhytisma in having a black peritoneum (unpigmented in S. rhytisma) and by lacking a pigmented blotch on the caudal region of the ocular side of the body (present in S. rhytisma). Symphurus rhytisma has much smaller and more numerous scales in a longi- tudinal series than does either of the other species (91-97 in S. rhytisma vs. 62-75 in S. piger and 77- 87 in S. pusillus). Symphurus rhytisma also has fewer anal-fin rays (68—7 1 ) than does S. pusillus (71-75). Symphurus rhytisma differs further from S. piger in having an unpigmented isthmus and inner opercu- lar linings (vs. lightly pigmented in S. piger), 4 hypurals (vs. 5 in S. piger), and in its much smaller size (45 mm vs. ca. 130 mm in S. piger). Other Atlantic species most closely resembling S. rhytisma are S. lubbocki and S. reticulatus, dwarf species of tonguefishes collected in shallow-water habitats in insular locations in the central and east- ern Atlantic. Symphurus rhytisma may be distin- guished from both species by its lower meristic fea- tures (dorsal-fin rays 83-87 vs. 87-89 in S. lubbocki and S. reticulatus', anal-fin rays 68-71 vs. 74-75; total vertebrae 46-48 vs. 48-49 in these other species), somewhat larger scales (91-97 scales in a longitudi- nal series vs. 101-109 in the other two species), and by differences in pigmentation. In S. rhytisma, all but the largest specimens have a dark blotch across the posterior one-third of the body (vs. no blotch in these other species), and the ocular surface of the body of S. rhytisma generally has a series of incom- plete, dark brown crossbands contrasting against a pallid background. This combination of features con- trasts with that observed in S. reticulatus, which has a dark, chocolate-brown body with X- and Y-mark- ings and a series of alternating blotches and unpig- mented areas in the dorsal and anal fins, and that observed in ■ 15 \ 90i 75i i60 i45 Figure 29 Geographic distribution of Symphurus pusillus based on mate- rial examined (discussion of geographic distribution appears in species account). represented in collections. Most samples consist of solitary individuals, undoubtedly reflecting difficul- ties in collecting this small species at the relatively great depths it inhabits. Baughman (1950:138) reported two specimens purportedly of this species from the western Gulf of Mexico near Corpus Christi, Texas. However, these specimens (USNM 93584 and not USNM 93854 as listed in Baughman’s paper) are actually S. plagiusa. Bathymetric distribution Symphurus pusillus in- habits mud substrates in moderate depths (102- 233 m) on the continental shelf (Table 10). Most of 21 specimens with depth of capture information were collected at 115-233 m. Only two specimens (USNM 153099) occurred shallower than 110 m (at 102 m). Remarks Goode and Bean ( 1885b:590) based their description of Aphoristia pusilla on three specimens collected from off Long Island, New York. Of the three syn types, one female (USNM 28778) in the best over- all condition is designated as the lectotype. The other two syntypes in USNM 28730 and USNM 325958 (formerly 28778, in part) now become paralectotypes. The lectotype (54.5 mm) was collected in 139 m at 40°01'N, 69°56'W on 4 Aug 1881. It has a 1-3-2 ID pattern, 12 caudal-fin rays, 87 dorsal- and 73 anal- fin rays, 48 total vertebrae, and ca. 83 scales in lon- gitudinal series. Comparisons In some meristic and other features, S. pusillus resembles two deep-water species, the western Atlantic S. piger and the eastern Atlantic S. nigrescens. Fin-ray counts of S. pusillus overlap al- most completely those of S. piger. However, this spe- cies differs from S. piger in having more longitudi- nal scales (77-87 vs. 62-75 in S. piger), an unpig- mented isthmus and unpigmented inner opercular linings (both structures lightly pigmented in S. piger), in four hypurals (vs. 5 in S. piger), in its different morphometries (Fig. 30, A-B), and S. pusillus is a much smaller species (see Figs. 6C and 8D), attain- ing maximum lengths of only about 77 mm, whereas, S. piger reaches lengths nearly double that size (ca. 130 mm). Symphurus pusillus and the eastern Atlantic <8. nigrescens are possibly a closely related species pair with distributions on the continental shelf on either side of the Atlantic. Although only slight differences in meristic features are noted between S. pusillus and S. nigrescens, these species are distinct in that S. pusillus has a longer caudal fin (11.5-15.4% SL vs. 7.6-12.2% SL in S. nigrescens-, Fig. 31), and the dorsal and anal fins in S. pusillus are pigmented basally, but not distally, and these fins usually lack Munroe: Systematics of western Atlantic Symphurus 61 ■5 Cl. TD 73 O CO 100 120 140 B 350 300 =1 I 250 200 20 40 60 80 100 120 140 Standard length Figure 30 Comparisons of selected morphometric features for Symphurus pusillus and S. piger. (A) Body depth (thou- sandths of standard length) versus standard length (mm). (B) Head width (thousandths of standard length) versus standard length (mm). darkly pigmented blotches (if blotches are present, they are small and only lightly pigmented) or lack any noticeably dark pigment streaks on the fin rays. In contrast, S. nigrescens usually has quite colorful dorsal and anal fins featuring a series of alternating dark blotches and unpigmented areas entirely throughout these fins, or if pigmented blotches are not present, then the individual fin rays throughout the entire dorsal and anal fins are streaked with dark 16 CTJ T3 3 03 U + 14 - + + 12 - + + + + + + 10 - 8 - + "1 6 1 L 30 50 70 B nigrescens + pusillus _J l 90 110 Standard length Figure 3 1 Comparison of caudal-fin length (expressed in percent stan- dard length) versus standard length for Symphurus pusillus (western Atlantic) and S. nigrescens (eastern Atlantic). pigment over their entire lengths. Symphurus pusillus is also a smaller species reaching maximum sizes of only ca. 77 mm and maturing at sizes as small as 45 mm, whereas, S. nigrescens attains larger sizes (to 117 mm and not reaching maturity until 70 mm or larger). Symphurus pusillus has some morphological fea- tures similar to those observed in the western At- lantic S. pelicanus and S. rhytisma. Differences be- tween these species are substantial and discussed in the “Comparisons” sections in accounts for S. rhytisma and S. pusillus. Some meristic features and ocular-side coloration of S. pusillus are similar to those found in S. billykrietei, S. ginsburgi, and S. stigmosus. In fact, throughout its range S. pusillus co-occurs with, but is not usually syntopic with, S. billykrietei (only one lot in VIMS collection (1905) taken at 40°N and 233 m, and coincidentally, the deepest known cap- ture for S. pusillus, contained both species) and S. stigmosus. Symphurus pusillus differs from all three species in its generally lower, mostly nonoverlapping, meristic features (dorsal-fin rays 83-88 vs. 87-95, anal-fin rays 71-75 vs. 74-84, and total vertebrae 47-49 vs. 50-53 in the others). Other differences be- tween S. pusillus and these species are discussed in the “Comparisons” sections for S. ginsburgi, S. billykrietei, and S. stigmosus, respectively. 62 Fishery Bulletin 96( 1 ), I 998 Figure 32 Symphurus marginatus (Goode and Bean), USNM 236609, 115 mm SL, Florida 29°39'N, 80°11'W. Symphurus marginatus (Goode and Bean, 1 886| (Figs. 8C, 32-33; Tables 1-10, 20} Margined tonguefish Aphoristia marginata Goode and Bean, 1886:154 (in part); (original description; Gulf of Mexico, off Mississippi; nontype specimen from Fish Hawk Station 1154 belongs to S. nebulosus). Goode and Bean, 1896:459 (in part); (redescription with fig- ure; based on Goode and Bean, 1886). Symphurus marginatus . Jordan and Goss, 1889:323 (after Goode and Bean). Jordan and Evermann, 1898:2706 (after Goode and Bean). Evermann and Marsh, 1900:332 (in key). Chabanaud, 1939:26 (American Atlantic). Ginsburg, 1951:198 (counts, measurements, distribution, in key). Fowler, 1952:143 (New Jersey, offshore record based on Goode and Bean). Bright, 1968:58 (four specimens, central Gulf of Mexico; 585-732 m). Topp and Hoff, 1972:107 (geographical distribution). Potts and Ramsey, 1987:88 (Gulf of Mexico; color description; 333-832 m). Seret and Andreata, 1992:94 (one specimen; southern Brazil; 600 m). Munroe, 1992:368, 377 (ID pattern; geographic, bathymet- ric distributions). Symphurus diomedianus (not of Goode and Bean, 1885). Longley and Hildebrand, 1941:49 (Tortugas, Florida). Symphurus plagusia (not of Schneider, in Bloch and Schneider, 1801). Seret and Andreata, 1992:94 (southern Brazil; five specimens; 640 m). Diagnosis Symphurus marginatus is a deepwater species that can be distinguished from all congeners by the following combination of characters: predomi- nant 1-3-2 ID pattern; 12 caudal-fin rays; 4, or less frequently, 5 hypurals; 93-104 dorsal-fin rays; SO- SO anal-fin rays; 51-56 total vertebrae; 86-99 scales in longitudinal series; absence of pupillary opercu- lum; black peritoneum; teeth along entire margin of ocular-side jaws; absence of fleshy ridge on ocular- side lower jaw; absence of scales on blind sides of dorsal- and anal-fin rays; elongate, somewhat slen- der body of nearly uniform depth along anterior two- thirds; ocular surface pigmentation featuring dark brown caudal blotch; posterior one-third of dorsal and anal fins with large caudal blotch, but without spots; and caudal fin without spots or blotches. Description A medium-size species attaining a maximum length of ca. 146 mm SL. ID pattern usu- ally 1-3-2 (77/98 specimens), less frequently 1-3-3 (11/ 98) and 1-2-3, rarely 1-2-2 (Table 2). Caudal-fin rays 12 (97/101), rarely 11 or 13 (Table 3). Dorsal-fin rays 93-104, usually 95-101 (Table 4). Anal-fin rays 80- 89 (Table 5). Total vertebrae 51-56, usually 52-54 (92/97) (Table 6). Hypurals 4 (41/57 specimens), less frequently 5 (16/57 specimens). Longitudinal scale rows 86-99, usually 88-96 (Table 7). Scale rows on head posterior to lower orbit 16-19, usually 17-18 (Table 8). Transverse scales 30-37 (Table 9). Proportions of morphometric features presented in Table 20. Body relatively elongate, of nearly uniform width along anterior two-thirds, with gradual taper Munroe: Systematics of western Atlantic Symphurus 63 Table 20 Morphometries for holotype (MCZ 27967) and 29 additional specimens of Symphurus marginatus. (Abbreviations de- fined in methods section; SL is expressed in mm; charac- ters 2 to 14 are expressed in thousandths of SL; 15 to 21 in thousandths of HL; n = no. of specimens measured.) Character Holotype n Range Mean SD 1. SL 90.1 30 56.9-146.1 106.5 18.08 2. BD 218 30 200-315 250.3 27.14 3. PDL 58 30 44-81 56.0 7.33 4. PAL 203 30 182-256 219.1 17.40 5. DBL 942 30 919-956 944.0 7.34 6. ABL 741 30 616-846 768.2 34.22 7. PL — 24 42-74 58.0 8.24 8. PA 50 30 27-74 52.9 11.50 9. CFL — 25 80-125 105.9 11.46 10. HL 192 30 127-221 182.1 15.56 11. HW 196 30 147-227 191.2 16.74 12. POL 119 30 99-144 112.0 8.88 13. UHL 90 30 90-133 110.3 11.86 14. LHL 115 30 84-129 97.1 11.08 15. POL 618 30 571-802 617.0 41.52 16. SNL 185 30 168-331 207.2 31.91 17. UJL 231 30 180-331 213.6 25.77 18. ED 150 30 125-248 148.6 22.98 19. CD 144 30 144-256 197.4 30.83 20. OPLL 300 30 208-372 290.4 36.30 21. OPUL 156 30 144-331 218.9 38.03 posteriorly beyond this point. Body depth increasing with size, juveniles with narrower body, usually pro- portionately less than 280 SL; adults with body depth ranging from 280-315 SL. Preanal length slightly shorter than body depth. Head moderately long and relatively narrow, slightly shorter than body depth. Head usually just slightly wider than long (HW:HL 0.84-1.25, x =1.05). Lower head lobe narrow, slightly less than postorbital length; slightly narrower than upper head lobe. Lower opercular lobe of ocular side considerably wider than upper lobe. Snout short, somewhat pointed; covered with small ctenoid scales. Poorly developed dermal papillae occasionally present on blind-side snout. Anterior nostril on ocu- lar side long, when depressed posteriorly, usually falling just short of anterior border of lower eye (about two-thirds of specimens), or just reaching to ante- rior border of lower eye in about one-third of speci- mens. Jaws moderately long; maxilla extending pos- teriorly to vertical through anterior margin of lower eye. Ocular-side lower jaw without fleshy ridge. Teeth well developed on blind-side jaws. Ocular-side dentary with row of teeth along complete margin of jaw; ocular-side premaxilla usually with single row of teeth along anterior four- fifths of margin of jaw, occasionally with complete tooth row. Chin depth slightly smaller than snout length. Lower eye large; eyes usually equal in position, with large and obvi- ous lens. Anterior and medial surfaces of eyes par- tially covered with 4-6 small ctenoid scales; 4-6 small ctenoid scales in narrow interorbital region. Pupil- lary operculum absent. Dorsal-fin origin usually equal with vertical through midpoint of upper eye, occasionally located more posteriorly, only reaching vertical through posterior margin of upper eye. Scales absent on blind sides of dorsal- and anal-fin rays. Pelvic fin short; longest pelvic-fin ray, when extended posteriorly, usually reaching base of first anal-fin ray. Posteriormost pelvic-fin ray connected to body by short delicate membrane terminating anterior to anus, or occasionally reaching posteriorly to anal-fin origin (membrane torn in most specimens). Caudal fin short. Relatively small ctenoid scales on both sides of body. Pigmentation (Fig. 32J Coloration similar for both sexes. Ocular surface usually uniformly dark brown, sometimes with yellowish tint, without crossbands. The most consistent and obvious pigmentation in preserved specimens are longitudinal black stripes along bases of the dorsal and anal fins, and a dark brown blotch, roughly circular in outline, usually covering the entire ocular-side caudal region. Cau- dal blotch usually extending over ca. 10 scale rows and 13-14 posteriormost fin rays of dorsal and anal fins; occasionally caudal blotch extended onto cau- dal-fin base. Ocular-side outer opercle with back- ground coloration of body. Inner linings of opercles and isthmus on both sides of body usually unpig- mented. Ocular-side upper lip with variably pig- mented band; ocular-side lower lip occasionally spot- ted, but without prominent pigment band. Small patch of pigment of variable intensity occasionally at base of anterior nostril. Blind side off-white, or yellowish. Peritoneum black, usually visible through abdominal wall on both sides. Anal pore white. Basal one-half of dorsal- and anal-fin rays in ante- rior two-thirds of body uniformly pigmented with dark brown or black pigment forming longitudinal stripe along fin-ray bases; distal one-half of those fin rays unpigmented or only lightly pigmented with diffuse melanophores. Caudal region of body, espe- cially proximate to caudal blotch, with fin rays of dor- sal and anal fins heavily pigmented. Caudal fin usu- ally heavily pigmented on proximal one-half; distal one-half with diffuse pattern of light melanophores of similar coloration to anterior two-thirds of ocular- side of body, or occasionally unpigmented. Size and sexua! maturity (Fig. 8C} Symphurus marginatus is a medium-size tonguefish attaining maximum lengths of about 146 mm. Most specimens were much smaller; nearly one-half of 93 specimens 64 Fishery Bulletin 96(1), 1998 Figure 33 Geographic distribution of Symphurus marginatus based on material exam- ined (discussion of geographic distribution appears in species account). examined for size-related life history in- formation were 100-120 mm, whereas another 22% were 80-100 mm. Females attain somewhat larger sizes. The largest S. marginatus examined in this study was a female of 146.1 mm; the largest male measured 130.5 mm. Specimens <80 mm are generally rare in collections; only 10 in this size range were available for study. Of specimens examined, 43 were males (56.9-130.5 mm) and 51 females (58.7- 146.1 mm), with 49 females being mature (79-146 mm). Sexual maturity in females occurs at a relatively large size (ca. 79-90 mm). The smallest female with elongate ovaries was 78.6 mm, and all but one fe- male larger than 80 mm had elongate ova- ries. Most females 85-105 mm, although having elongate ovaries, lacked evident mature ova. The smallest gravid female was 87.5 mm, but this size is apparently unusual, because only three of 20 other females smaller than 105 mm were gravid. Two fe- males, 58.7 and 80.1 mm, were immature with ovaries scarcely elongate. Geographic distribution (Fig. 33 J Prima- rily in deepwater outer continental shelf habitats from off New Jersey (39°55'N) southward along the eastern United States, in eastern and central regions of the Gulf of Mexico (to Louisiana, 91°18'W), off the Ba- hamas, the Greater Antilles at Puerto Rico, widespread throughout the southern Caribbean Sea from Honduras to Venezu- ela, and from Trinidad and Tobago to southeastern Brazil (21°34'S) (Seret and Andreata, 1992). Although S. marginatus has occasionally been collected as far north as southern New Jersey (39°N) and Virginia (36°N), the majority of specimens were taken farther south, primarily off southern Florida, in eastern and central regions of the Gulf of Mexico, and throughout the southern Caribbean Sea. Southernmost records for this species (Seret and Andreata, 1992) are for speci- mens from off southeastern Brazil (ca. 21°S). Bathymetric distribution This species usually in- habits deepwater soft mud substrates on the outer continental shelf and upper continental slope. Symphurus marginatus has been collected at depths of 37-750 m (Table 10), but its center of abundance occurs between 320 and 550 m, where the majority of specimens (88/108 or 81%) were collected. Of 108 specimens with available depth information, only nine were collected at depths shallower than 300 m. Single specimens were collected at 37 m (UMML 17440, east coast of Florida), 66 m (UMML 35237, Nicaragua), and 72 m (MCZ 58657, Nicaragua); whereas three specimens (UMML 30106) collected at 10-1 1°N off Costa Rica were taken at 45 m. Two specimens were collected between 280 and 290 m (UMML 35240; Colombia) and one (UMML 35231; Florida) was taken at 293 m. Only 11 specimens were taken deeper than 550 m. The two deepest captures (three specimens at 713 m, USU 1371; and one speci- men at 750 m, FMNH 47908) were taken off Brazil and in the Gulf of Mexico, respectively. Potts and Ramsey (1987) reported a depth range of 333-832 m for this species in the Gulf of Mexico. Little is known concerning life history of this species. Remarks In the original description, Goode and Bean ( 1886: 154) mistakenly identified a specimen of S. nebulosus as their new species, Aphoristia marginata. This specimen, however, was not designated Munroe: Systematics of western Atlantic Symphurus 65 as part of the type series and therefore does not com- promise the original concept of Aphoristia marginata. Metzelaar’s reference (1919:134) to S. marginatus from Saint Eustatius is based on a misidentified specimen (ZMA 119.422) of S. ommaspilus. Comparisons Symphurus marginatus has some- what similar geographic and bathymetric distribu- tions throughout the Caribbean Sea and Gulf of Mexico as those reported for S. piger (see below), and these are the only western Atlantic species with a combination of a 1-3-2 ID pattern, black peritoneum, and five hypurals (although five hypurals occur much less frequently in S. marginatus — only 28% of 57 specimens of this species had this count vs. 99% of 136 specimens of S. piger). However, these distinctive species are not usually collected syntopically and the two species can be readily identified (compare Figs. 32 and 34). Symphurus marginatus lacks crossbands on the ocular surface and has a prominent blotch on the ocular-side caudal region, whereas S. piger usually has prominent crossbands and lacks any blotch on the ocu- lar-side caudal region. The isthmus and inner opercu- lar linings of S. marginatus are unpigmented, whereas those of S. piger are lightly sprinkled with melano- phores, and basal margins of the dorsal and anal fins in S. marginatus have a dark brown stripe that is ab- sent in S. piger. Symphurus marginatus also differs from S. piger by its much higher and nonoverlapping meristic features (93-104 dorsal-fin rays vs. 80-88 in S. piger; 80-89 anal-fin rays vs. 68-74; 86-99 scales in a longitudinal series vs. 62-75; and 51-56 total verte- brae vs. 45^9 in S. piger). In addition, S. marginatus has a more elongate body (BD 200-315, x =250) with a relatively narrow head ( 147-227 SL, x =191) compared with that of S. piger (wide body 244—350 SL, x=322 and wide head 242-313 SL, x=277). The relatively elongate body of S. marginatus is reminiscent of other Atlantic slender-bodied, deepwater species, namely the western Atlantic S. nebulosus, and S. ligulatus and S. uanmelleae from the eastern Atlantic. Symphurus marginatus is readily distinguished from S. nebulosus and S. ligulatus by its fewer caudal-fin rays (12 vs. 14 in these others), from all three species by differences in ID pattern (1-3-2-2-2 in S. marginatus vs. 1-2-2- 2-2 in S. nebulosus and S. ligulatus and 1-2-2-1-2 in S. uanmelleae), and its generally lower meristic fea- tures (93-104 dorsal-fin rays in S. marginatus vs. 105-113 in S. nebulosus , 101-108 in S. uanmelleae, and 102-113 in S. ligulatus', 80-89 anal-fin rays in S. marginatus vs. 91-98 in S. nebulosus , 86-93 in S. uanmelleae, and 90-102 in S. ligulatus', 51-56 total vertebrae in S. marginatus vs. 57-60 in S. nebulosus, 55-59 in S. uanmelleae, and 56-61 in S. ligulatus', 86-99 scales in a longitudinal series in S. marginatus vs. 120-135 in S. nebulosus, 107-124 in S. uan- melleae, and 115-135 in S. ligulatus). In addition, all three species lack the dark brown ocular-side cau- dal blotch present in S. marginatus . From S. uanmelleae, S. marginatus differs further in having only nine abdominal vertebrae (vs. 10 or 11). Differences between S. marginatus and S. billykrietei, S. ginsburgi, and S. stigmosus, three other western Atlantic members of this species group, are discussed in the “Comparisons” section of each species account, respectively. Symphurus marginatus is similar in some meris- tic features to the western Atlantic, shallow-water species, S. tessellatus, S. oculellus, and S. carib- beanus, but is easily recognized from all three by its black peritoneum (unpigmented in these other spe- cies), well-developed dentition on ocular-side jaws (vs. absent or reduced dentition on ocular-side jaws), and differences in pigmentation patterns. The ocular sur- face in S. marginatus is uniformly pigmented with a single dark brown blotch in the caudal region, the isth- mus and inner opercular linings are unpigmented, and spots or blotches are lacking on the ocular-side opercle. In contrast, the ocular surface of these other species usually has well-developed crossbands, the ocular-side isthmus and inner opercular linings are heavily pig- mented, and all lack the dark brown caudal blotch char- acteristic of S. marginatus. Both S. tessellatus and S. oculellus also differ in having a dark blotch on the outer surface of the ocular-side opercle. Symphurus margi- natus has a different ID pattern (1-3-2) than that in these others (1-4-3 ID pattern). Approximately 28% (16/57) of the S. marginatus ex- amined had five hypurals (the remainder had four). Four other species in the genus, the western Atlantic S. piger and eastern Pacific S. microlepis, S. diabolicus, and S. oligomerus, also have the combination of a 1-3- 2 ID pattern, 12 caudal-fin rays, black peritoneum, and five hypurals. Differences between S. marginatus and S. piger were discussed above. Symphurus marginatus is easily distinguished from the eastern Pacific species because they lack the large dark brown blotch on the ocular-side caudal region. From S. oligomerus, S. marginatus differs further in having dorsal and anal fins with a dark brown stripe along the basal margins (vs. an alternating series of rectilinear pigmented blotches and unpigmented regions on the dorsal and anal fins and no basal stripe in S. oligomerus). Meris- tic features of S. marginatus are distinctly lower than those of S. microlepis and S. diabolicus (less than 105 dorsal-fin rays, 90 or fewer anal-fin rays, and 56 or fewer vertebrae in S. marginatus vs. more than 105 dorsal-fin rays and 91 anal-fin rays, and more than 56 total vertebrae in these other species). 66 Fishery Bulletin 96( 1 ), 1998 Figure 34 Symphurus piger (Goode and Bean), USNM 159211, 106 mm SL, French Guiana 7°18'N, 53°32'W. Symphurus piger (Goode and Bean, 1886J (Figs. 8D, 30, 34-35; Tables 1-10, 21} Deepwater tonguefish Aphoristia pigra Goode and Bean, 1886:154 (in part) (original description; more than one species in ac- count). Goode and Bean, 1896:460 (in part) (rede- scription, figure; based on specimens in previous citation). Cockerell, 1912:172 (brief discussion and figure of scales). Symphurus piger. Jordan and Goss, 1889:326 (after Goode and Bean, 1886). Jordan and Evermann, 1898:2705 (after Goode and Bean, 1886). Ever- mann and Marsh, 1900:332 (in key). Chabanaud, 1939:26 (Caribbean Sea, 457 m). Ginsburg, 1951:197 (in part; more than one species in rede- scription; designation of lectotype). Topp and Hoff, 1972:108 (distribution). Guitart, 1978:727 (Cuba; figure, counts, in key). Potts and Ramsey, 1987:89 (Gulf of Mexico; color description; 92-194 m). Munroe, 1992:368, 377 (ID pattern; geographic, bathymetric distributions). Symphurus pusillus (not of Goode and Bean, 1885). Longley and Hildebrand, 1941:50 (Tortugas, Florida). Diagnosis Symphurus piger is a deepwater species distinguished from all congeners by the combination of predominant 1-3-2 ID pattern; 12 caudal-fin rays; 5 hypurals; 80-88 dorsal-fin rays; 68-74 anal-fin rays; 45-49 total vertebrae; 62-75 scales in longitu- dinal series; black peritoneum; absence of pupillary operculum; teeth along entire margin of ocular-side jaws; absence of fleshy ridge on ocular-side lower jaw; absence of scales on blind side of dorsal- and anal- fin rays; wide body; wide head; ocular surface pig- mentation with strong crossbanding pattern with- out caudal blotch; and dorsal, anal, and caudal fins without spots or blotches. Description A medium-size Symphurus attaining maximum lengths of ca. 130 mm SL. ID pattern usu- ally 1-3-2 (137/141 specimens), rarely 1-2-2 or 1-3-3 (Table 2). Caudal-fin rays usually 12 ( 134/139), rarely 10, 11, or 13 (Table 3). Dorsal-fin rays 80-90, usu- ally 83-88 (Table 4). Anal-fin rays 68-74 (Table 5). Total vertebrae 45-49, usually 47^19 (137/142) (Table 6). Hypurals 5 (138/139 specimens; 1/139 with 4 hypurals). Longitudinal scale rows 62-75, usually 66-73 (Table 7). Scale rows on head posterior to lower orbit 16-21, usually 17-19 (Table 8). Transverse scales 32-38 (Table 9). Proportions of morphometric features presented in Table 21. Body relatively deep, maximum depth in anterior one-third of body; body tapering relatively rapidly posterior to midpoint. Preanal length shorter Munroe: Systematics of western Atlantic Symphurus 67 TafaJe 2 1 Morphometries for holotype (MCZ 27965) and 32 additional specimens of Symphurus piger. (Abbreviations defined in methods section; SL is expressed in mm; characters 2 to 14 are expressed in thousandths of SL; 15 to 21 in thou- sandths of HL; n = no. of specimens measured.) Character Holotype n Range Mean SD 1. SL 84.6 33 58.6-120.5 85.0 14.71 2. BD 314 33 244-350 322.5 23.09 3. PDL 53 33 46-90 58.4 7.98 4. PAL 260 33 176-327 255.4 26.40 5. DBL 947 33 910-954 941.6 7.98 6. ABL 742 33 730-778 748.5 12.65 7. PL 58 33 58-86 73.3 7.13 8. PA 32 33 28-67 46.1 9.09 9. CFL 142 33 103-168 147.9 13.92 10. HL 234 33 182-256 236.8 12.58 11. HW 266 33 242-313 276.8 13.63 12. POL 155 33 149-226 160.6 13.22 13. UHL 147 33 124-198 171.3 13.08 14. LHL 136 33 103-142 126.5 9.76 15. POL 662 33 633-910 680.0 61.58 16. SNL 182 33 172-282 203.7 19.10 17. UJL 202 33 188-338 228.2 29.60 18. ED 146 32 101-167 118.1 13.84 19. CD 222 33 190-317 232.5 26.91 20. OPLL 328 33 269-401 321.8 29.94 21. OPUL 242 33 170-375 226.9 37.52 than body depth. Head long and wide; head length shorter than body depth. Head much shorter than wide (HW:HL 1.03-1.44, x=1.21). Lower head lobe width considerably less than postorbital length; nar- rower than upper head lobe. Lower opercular lobe of ocular side considerably wider than upper opercular lobe. Snout short, somewhat rounded; covered with small ctenoid scales. Dermal papillae usually well developed on blind side of snout. Anterior nostril on ocular side long, when depressed posteriorly, reach- ing anterior border of lower eye in about one-half of specimens, falling just short of anterior margin of lower eye in remaining specimens. Jaws long; max- illa usually extending posteriorly to vertical through mid-point of lower eye; less frequently reaching ver- tical through posterior margin of pupil of lower eye. Ocular-side lower jaw without fleshy ridge. Teeth well developed on blind-side jaws. Teeth along entire margin of ocular-side dentary. Anterior three-fourths of margin of ocular-side premaxilla usually with teeth; occasionally teeth over entire marginal sur- face of premaxilla. Chin depth slightly larger than snout length. Lower eye relatively small; eyes usu- ally equal in position. Anterior and medial surfaces of eyes usually covered with 4-5 short rows of small ctenoid scales; 3-7 small ctenoid scales in narrow interorbital region . Pupillary operculum absent. Dor- sal-fin origin usually equal with vertical through posterior margin of pupil of upper eye, occasionally reaching vertical through anterior margin of upper eye. Scales absent on blind sides of dorsal- and anal- fin rays. Pelvic fin moderately long; longest pelvic- fin ray, when extended posteriorly, usually reaching base of first anal-fin ray. Posteriormost pelvic-fin ray connected to body by delicate membrane terminat- ing immediately anterior to anus, or occasionally extending posteriorly almost to anal-fin origin (mem- brane torn in most specimens). Caudal fin long. Scales large, ctenoid; with cteni about equally devel- oped on both sides of body. Pigmentation (Fig. 34J Coloration generally simi- lar in both sexes. Ocular surface usually dark brown with 3-10 (usually 5-8) well-developed, darker brown, sharply-contrasting, rather narrow cross- bands on head and body. Crossbands continued onto dorsal and anal fins as small, elongate or irregularly shaped, somewhat diffuse, blotches. Occasionally, crossbands scarcely evident against exceptionally dark background coloration. Individuals from sev- eral locations presumably collected on light-colored sandy substrates yellowish, with faint, almost im- perceptible crossbands. Older specimens mostly faded with little evidence of crossbanding. First crossband on body immediately posterior to opercu- lum. Second and third crossbands, usually darkest; crossing body immediately posterior to operculum and almost at midpoint, respectively. Number and degree of completeness of crossbands variable in posterior one- half of body. Posteriormost crossband just anterior to caudal-fin base. Ocular-side outer opercle with back- ground coloration as body. Inner linings of both opercles and isthmus on both sides of body lightly pigmented. Ocular-side lips usually with dark band of pigment, occasional specimens with only light spotting on lips. Blind side uniformly yellowish-white. Peritoneum black, visible through abdominal wall on both sides. Dorsal and anal fins generally lightly pigmented anteriorly, usually becoming increasingly darker brown, but not black, on posterior one-third to one- half of body; dorsal and anal fins without definite spots. Fin rays usually evenly pigmented along their lengths. Dorsal and anal fins more heavily sprinkled with melanophores, or with melanophores coalesced into irregular elongate blotches, in regions proximate to body crossbands. Caudal-fin rays and membranes uniformly darkly pigmented throughout length of fin; without pigmented spot at caudal-fin base. Size and sexual maturity (Fig. 8D) Symphurus piger is a medium-size species attaining a maximum 68 Fishery Bulletin 96(1 ), 1998 size of ca. 130 mm; however, most specimens were much smaller (80-105 mm). Only 15/161 (10.7%) fish examined for size-related life history information were larger than 110 mm. Males and females attain similar sizes. The largest S. piger was a female mea- suring 127 mm; the largest male was 118 mm. Among 161 specimens examined for life history information, were 88 males (51.6-118.3 mm), 70 females (57.9- 127.1 mm), and three immature fish (27.3-51.2 mm) of indeterminate sex. Sexual maturity of females occurs at ca. 69 mm. Mature females (n=66) were 69.1-127.1 mm, and all females larger than 73 mm were mature with fully elongate ovaries. The small- est gravid female was 69.1 mm. Four females, 57.9- 66.6 mm, were immature with ovaries just undergo- ing elongation. Geographic distribution (Fig. 35) On outer conti- nental shelf and upper continental slope from off southern Florida (off St. Augustine, ca. 30°N), Florida Straits, and Bahamas, infrequently in the Gulf of Mexico, and south through the Caribbean Sea, in- cluding waters off the Greater and Lesser Antilles, as well as off Mexico (Yucatan Peninsula), Central America, and northern South America to about French Guiana (7°N, 53°W). According to material available, S. piger is primarily a tropical species wide- spread in relatively deepwater areas through the Caribbean Sea and tropical Atlantic Ocean. Only a few lots containing this species were taken in the Gulf of Mexico. Of these, one (UF 44356, con- taining one large adult of 127 mm) was from the east- ern Gulf, off the Mississippi Delta of Louisiana (29°12'N, 88°25'W), whereas three others are from southern Florida (FDNR 12566) and the Tortugas region (USNM 117176; USNM 117287). Three lots (TCWC 4468.11; 6097.14; and 6207.17), totalling 21 specimens, were collected on the continental shelf in the western Gulf off Yucatan ( 18.5-20.3°N). Baughman’s (1950) report of S. piger from Freeport, Texas, is based on a specimen (CAS-SU 40556) of S. civitatium. Relative scarcity of specimens from the Gulf of Mexico indicates that S. piger is probably not a regular component of the resident deep-sea fauna of this region, particularly in northcentral and north- western areas of the Gulf. Bathymetric distribution Symphurus piger occurs on relatively deep soft mud bottoms on the outer con- tinental shelf and upper continental slope. Depth- of-capture information for 175/178 specimens (Table 10) reveals that this species has been collected over a wide depth range (92-549 m). However, the center of abundance occurs between 141 and 300 m where Figure 35 Geographic distribution of Symphurus piger based on material examined (discussion of geographic distribution appears in species account). Munroe: Systematics of western Atlantic Symphurus 69 143/175 (82%) of the individuals were captured. Of interest is the capture of a 27.3 mm specimen at 329 m, and a 38.5 mm specimen at 238 m, indicating that small juveniles also occur at depths inhabited by adults. Only 5/175 (3%) S. piger have been collected at depths shallower than 110 m, the shallowest cap- ture (92 m) being that of three specimens (UMML 35255) taken off the Netherlands Antilles. Only 21/ 175 (12%) fish were taken deeper than 300 m, with the deepest record (UMML 7124) being that of four specimens taken at 549 m off Puerto Rico. Remarks In his revision of western Atlantic tongue- fishes, Ginsburg (1951) stated that he was unsure which, if any, of specimens on which the original de- scription was based, was the holotype of A. pigra. He selected ( 1951:197) the specimen from Blake Sta- tion XXIII (now MCZ 27965) as the lectotype. This action was invalid, because in the original descrip- tion Goode and Bean (1886:154) designated the speci- men from Station XXIII as “the type” of A. pigra and listed specimens from Albatross Stations 2318 and 2405 as collateral types (^paratypes in current ter- minology), not cotypes as stated in Ginsburg ( 1951). This distinction is important because of all presently available specimens believed to form the basis for the description of A. pigra, the only specimen that is actually this species is the holotype collected at Blake Station XXIII (see below). Among seven type specimens included in the origi- nal description of A. pigra (Goode and Bean, 1886), all still extant, with the exception of the holotype, are S. parvus Ginsburg, 1951. In addition, among other, nontype material listed in the original descrip- tion, Goode and Bean included specimens of a third species, S. minor Ginsburg, 1951. The status or ex- istence of this material, other than the holotype, is presently somewhat confusing. Ginsburg ( 1951) and I have tried unsuccessfully to locate all specimens included in the original description of A. pigra. Diffi- culty with tracing this material results from the fact that specimens in the original description were listed only by Albatross or Blake station numbers. Goode and Bean, furthermore, did not always provide the number of specimens examined from each station. Although some tonguefishes from these stations were located, it is uncertain whether these are the specimens used by Goode and Bean. For example, Ginsburg (1951) was unsure that specimens he ex- amined from Albatross Station 2318 were the ones used by Goode and Bean in the original treatment of A. pigra. Ginsburg was also unable to locate two specimens from Albatross Station 2405 and an un- known number of specimens from Station 2425 listed in the original description of this species. It is note- worthy that Ginsburg discussed not four specimens from Albatross Station 2318, as in Goode and Bean’s original description, but rather six specimens. Ap- parently, Ginsburg was unaware of the discrepancy. During the course of the present study, attempts were made to locate the six paratypes of A. pigra and to discover the reason for the discrepancy in specimen number. Field data now included with these six USNM specimens reveal that all jar labels and museum registers list data only for Albatross Sta- tion 2318. If station data for all six specimens are correct, Goode and Bean either erred in listing only four specimens from this station (not likely), or two additional specimens collected at this station were not included in the original description (possible). If, however, the four specimens from Albatross Station 2318, as reported in Goode and Bean’s account, are the correct number, and these authors used all avail- able material, then station information for two of the six specimens now listed from Alba tross Station 2318 have been lost or transposed prior to examination by Ginsburg. Efforts to retrace the history of these specimens prior to their inclusion in USNM holdings, including a check of museum registers and catalogue of the Bureau of Fisheries Collection at the USNM, uncovered no additional information. A search through accession files at the USNM regarding trans- fer of these specimens from the Bureau of Fisheries also was unsuccessful because no accession num- bers) were listed for the specimens. Therefore, I can- not unequivocally demonstrate that the discussion below regarding the history of paratype material for A. pigra is fully correct. The six specimens listed in Ginsburg (1951) sup- posedly collected from Alba tross Station 2318 are now assigned the following museum numbers: four are USNM 74330 (three were designated as S. parvus paratypes by Ginsburg; the fourth, not listed in Ginsburg, was on loan at the time to R Chabanaud ( MNHN ) is also S. parvus but was not designated as a paratype); the two other specimens from Albatross Station 2318 also are part of the type series of S. parvus. One specimen is now USNM 84491, the ho- lotype of S. parvus and the other (USNM 152733) is a paratype of S. parvus. Both specimens originally were contained in the same jar (indicated both in mu- seum catalogue records and on jar labels). These (USNM 84491 and 152733) may be the two specimens from Albatross Station 2405 listed in Goode and Bean’s account of A. pigra. If this assumption is correct, then these specimens together with the four in USNM 74330 would account for the six paratypes of A. pigra. The fate and status of the remaining material listed in the original description of A. pigra is as follows. Specimens from Albatross Station 2374 (nontype sta- 70 Fishery Bulletin 96( 1 ), 1 998 tus for A. pigra ) are now paratypes (USNM 131590, 131591) of S. minor Ginsburg. The whereabouts of specimens (unknown number) from Albatross Sta- tion 2425 (nontype status for A. pigra ) are unknown. Both Ginsburg and I were unsuccessful in locating these specimens. Comparisons Symphurus piger is only one of five species in the genus (the western Atlantic S. marginatus and eastern Pacific S. microlepis, S. diabolicus, and S. oligomerus are the others) with the combination of a 1-3-2 ID pattern, 12 caudal-fin rays, black peritoneum, and five hypurals. Sym- phurus piger is readily distinguished from S. marginatus by differences in pigmentation and by its lower and nonoverlapping meristic features. Fur- ther comparisons between S. piger and S. marginatus are listed in the “Comparisons” section of the species account for S. marginatus. A discussion of charac- teristics distinguishing S. piger from S. oligomerus, S. microlepis, and S. diabolicus follows after the com- parisons with other western Atlantic species. Among Atlantic species with a 1-3-2 ID pattern, S. piger is most similar in some meristic features to the western Atlantic S. pusillus, the eastern Atlantic S. nigrescens, and the shallow-water western Atlantic S. rhytisma. Symphurus piger is readily distin- guished from all three by its much larger scales (62 — 75 longitudinal scales vs. 72-97 in these other spe- cies) and by its hypural count (5 vs. 4 in the others). Symphurus piger differs further from S. rhytisma in its black peritoneum (unpigmented in S. rhytisma) and by its much larger size (to at least 130 mm vs. 45 mm in S. rhytisma). Features further distinguish- ing S. piger from S. pusillus (see Fig. 30) and S. rhytisma are discussed in the “Comparisons” sections in the accounts for each of those species. Differences between S. piger and other western At- lantic species (S. ginsburgi, S. billykrietei, S. stigmosus, and S. pelicanus ) possessing a 1-3-2 ID pattern, 12 cau- dal-fin rays, and a black peritoneum are discussed in the “Comparisons” sections of each species account. Symphurus piger is quite distinctive from S. microlepis and S. diabolicus, having a much stockier and deeper body with the greatest depth occurring in the anterior one-third, compared with S. microlepis or S. diabolicus which have a more elongate shape (BD 244-350 in S. piger vs. 203 and 279, respectively, for S. microlepis and S. diabolicus ) with body depth nearly equal over the anterior two-thirds. Symphurus piger has much lower meristic features in compari- son with those of the other species (dorsal-fin rays 80-88 in S. piger vs. 106-109 in the others; anal-fin rays 68-74 vs. 92-94; 62-75 longitudinal scales vs. 126-135; and 45-49 total vertebrae vs. 57-58 in S. microlepis and S. diabolicus). Symphurus piger is distinguished from S. oligo- merus by its relatively uniformly pigmented dorsal and anal fins (vs. dorsal and anal fins of S. oligomerus with alternating series of boldly pigmented blotches and unpigmented areas), and differences in many meristic features (those of S. piger listed first): 80- 88 dorsal-fin rays vs. 87-97; 68-74 anal-fin rays vs. 72-83; 62-75 longitudinal scales vs. 86-96; and 45- 49 vs. 48-52 total vertebrae). Munroe: Systematics of western Atlantic Symphurus 71 Symphurus trewavasae Chabamaud, 1 948 (Figs. 8E, 36-37; Tables 1-10, 22} Trewavas's tonguefish Symphurus plagiusa (not of Linnaeus, 1766). Regan, 1914:23 (Cabo Frio, Brazil). Lazzaro, 1973:245 (Puerto Quequen, Argentina; in key). Roux, 1973:176 (southern Brazil). Menni et al., 1984:202 (reidentifications; previous citations by Roux (1973) for Brazil, and Lazzaro (1973) for Argen- tina correspond to S. trewavasae). Symphurus trewavasae Chabanaud, 1948:508 (origi- nal description; Cabo Frio, Brazil). Ginsburg, 1951:185 (brief comparison with S. plagiusa). Menezes and Benvegnu, 1976:144 (redescription, photograph; diagnosed from S. plagiusa ; ecologi- cal notes). Menni et al., 1984:202 (southern Brazil to Argentina). Munroe, 1992:369, 377 (ID pattern; geographic, bathymetric distributions). Andreata and Seret, 1995:590 (continental shelf, Brazil). Symphurus plagusia (not of Schneider, in Bloch and Schneider, 1801). Andreata and Seret, 1995:590 (in part; specimen from 23°07'S, 42°03'W is S. trewavasae). Diagnosis Symphurus trewavasae is readily distin- guished from all congeners by the combination of predominant 1-3-3 ID pattern; 10 caudal-fin rays; 4 hypurals; 88-94 dorsal-fin rays; 73-79 anal-fin rays; 47-51, usually 48-49, total vertebrae; 67-77 scales in longitudinal series; unpigmented peritoneum; absence of pupillary operculum; absence of scales on blind sides of dorsal- and anal-fin rays; teeth along entire margin of ocular-side dentary; anterior one-half or less of margin of ocular-side premaxilla with teeth; lack of fleshy ridge on ocular-side lower jaw and membrane ostia; ocular surface with bold pattern of crossbands without caudal blotch; lack of pepper-dot pigment on blind side of body; and dor- sal, anal, and caudal fins without spots or blotches. Description A medium-size tonguefish attaining maximum size of ca. 139 mm SL. ID pattern usually 1-3-3 (49/73 specimens), less frequently 1-4-2 (11/73) or 1-3-2 (4/73) (Table 2). Caudal-fin rays 10 (Table 3). Dorsal-fin rays 88-94 (Table 4). Anal-fin rays 73- 79 (Table 5). Total vertebrae 47-51, usually 48-49 (56/70) (Table 6). Hypurals 4 (70/70). Longitudinal scale rows 67-77, usually 71-77 (Table 7). Scale rows on head posterior to lower orbit 15-20, usually lb- 17 (Table 8). Transverse scales 31-37, usually 32-35 (Table 9). Proportions of morphometric features presented in Table 22. Body moderately deep; greatest depth from approximately region of tenth anal-fin ray to mid- point of body; body tapering fairly rapidly posterior to midpoint. Head moderately short and moderately wide, narrower than body depth. Head shorter than wide (HW:HL 1.04-1.49, x=1.2). Lower head lobe width less than postorbital length; narrower than 72 Fishery Bulletin 96(1 ), 1998 Table 22 Morphometries for holotype (BMNH 1913.12.4:264) and 19 additional specimens of Symphurus trewauasae. (Abbre- viations defined in methods section; SL is expressed in mm; characters 2 to 14 are expressed in thousandths of SL; 15 to 21 in thousandths ofHL; n = no. of specimens measured.) Character Holotype n Range Mean SD 1 SL 117.0 20 69.1-131.0 107.3 17.17 2. BD 310 19 267-419 307.9 31.01 3. PDL 41 20 36-50 41.4 4.25 4. PAL 197 20 184-243 217.1 15.05 5. DBL 959 20 952-969 959.6 4.35 6. ABL 791 20 757-812 783.1 15.56 7. PL 70 18 57-80 68.8 6.29 8. PA 64 20 37-83 60.8 12.76 9. CFL 116 19 83-129 114.2 13.70 10. HL 193 20 154-197 182.0 10.62 11. HW 260 20 206-260 227.2 15.92 12. POL 128 20 106-132 118.7 8.41 13. UHL 170 19 122-199 151.3 20.28 14. LHL 103 19 73-123 100.8 13.89 15. POL 664 20 613-696 651.7 25.95 16. SNL 181 20 181-263 206.0 22.84 17. UJL 186 20 186-250 218.8 17.72 18. ED 124 20 114—162 130.6 12.67 19. CD 159 20 159-264 227.2 26.08 20. OPLL 283 19 216-419 293.6 46.52 21. OPUL 186 19 159-297 227.3 38.54 upper lobe. Lower opercular lobe on ocular side usu- ally considerably wider than upper lobe. Snout short and rounded; covered with small ctenoid scales. Der- mal papillae evident, but not highly developed, on blind-side snout. Anterior nostril on ocular side short, not reaching anterior border of lower eye when de- pressed posteriorly. Jaws moderately long; maxilla usually extending posteriorly to vertical through anterior margin of pupil of lower eye. Ocular-side lower jaw without fleshy ridge. Teeth well developed on blind-side jaws. Dentary on ocular side with slen- der teeth along entire margin; a small number of slen- der teeth on anterior one-half to one-third of margin of ocular-side premaxilla. Chin depth usually slightly larger than snout length. Lower eye relatively large; eyes usually equal in position, occasionally eyes slightly subequal with upper in advance of lower eye. Anterior and medial surfaces of eyes and narrow in- terorbital space partially covered with 4-6 small ctenoid scales. Pupillary operculum absent. Dorsal- fin origin usually equal with, or occasionally slightly anterior to, vertical through anterior margin of up- per eye; predorsal length long. Scales absent from blind sides of dorsal- and anal-fin rays. Pelvic fin moderately long; longest pelvic-fin ray, when ex- tended posteriorly, usually reaching base of first anal- fin ray. Posteriormost pelvic-fin ray connected to body by delicate membrane terminating immediately an- terior to anus, or occasionally extending posteriorly nearly to anal-fin origin (membrane torn in most specimens examined). Caudal fin moderately long. Scales large, ctenoid; with cteni about equally devel- oped on both sides of body. Pigmentation (Fig. 36) Body coloration similar for both sexes. Ocular surface usually medium to light brown or straw-colored, with 3-7 (usually 3-5) com- plete, sharply contrasting dark brown crossbands on head and body. Crossbands not continued on dorsal and anal fins. Crossbands on head and posteriormost body usually faint and sometimes incomplete, but otherwise usually visible without magnification. Usually three conspicuous crossbands on body be- tween posterior margin of head and base of caudal fin. Anteriormost crossband on body at, or slightly posterior to, opercular opening. Ocular-side outer opercle with small cluster of brown speckles near ventral margin (remnants of incomplete band?). In- ner linings of opercles on both sides of body occa- sionally lightly pigmented. Isthmus unpigmented on both sides of body. Ocular-side upper lip with slight band of pigment; lower lip usually only lightly spot- ted, without definite pigment band. Blind side uni- formly creamy white. Peritoneum unpigmented. Dorsal- and anal-fin rays, along entire fins, with faint light brown pigment, heaviest on proximal one- half of fin rays; fins without blotches or spots. Cau- dal fin usually pale throughout entire length, occa- sionally scale-covered base of caudal fin darker brown than distal, scaleless portion of caudal-fin rays, but without well-developed spot. Size and sexual maturity (Fig. 8E) Symphurus trewavasae is a medium-size species reported to at- tain a maximum size of ca. 139 mm (Menezes and Benvegnu, 1976). The largest specimens examined in this study were males measuring 124, 125, and 131 mm, with the largest female (123 mm) only slightly smaller. Of 48 specimens for which size-re- lated life history information was available, 19 were males (52.5-131.0 mm) and 29 females (63.7-123.8 mm). Based on reproductive stages of females, sexual maturity in the species occurs at ca. 70-80 mm. All but three females larger than 80 mm had fully elon- gate ovaries and were either gravid or spent. The smallest gravid female was 74 mm. Seven immature females ranged from 69.1-122.9 mm. The smallest of these (69 and 78 mm, respectively) had ovaries just undergoing elongation, whereas ovaries of other immature females were partially elongate without indications of developing ova. Munroe: Systematics of western Atlantic Symphurus 73 Geographic distribution {Fig. 37 J Western South Atlantic inner continental shelf from southeastern Brazil to central Argentina. The northernmost record for this species (22°53'S) roughly corresponds to the region off Cabo Frio, Brazil (Menezes and Benvegnu, 1976). The specimen (INIDEP 476) from 45°S repre- sents the southernmost point of capture for this spe- cies. The specimen identified as S. plagiusa by Lazzaro (1973) from Puerto Quequen, Argentina (ca. 38°S), is also this species. Bathymetric distribution Symphurus trewavasae has been collected over a depth range from 7 (this study) to 179 m (Menezes and Benvegnu, 1976), with the majority (60/80, 75%) of captures at moderate depths (40-80 m) on the continental shelf (Table 10). Depth-of-capture information summarized from Menezes and Benvegnu (1976) revealed that ca. 85% of examined specimens were taken between 50 and 100 m, and only three at depths shallower than 40 m. Menezes and Benvegnu noted that all size classes were found at these depths, indicating it is unlikely that this species uses shallow inshore habitats as nursery grounds. The deepest capture reported for the species was for one specimen collected at 179 m (Menezes and Benvegnu, 1976). Shallow-water cap- tures of S. trewavasae listed in this study include a specimen (MNHN 1992-1411) taken at 7 m and three others collected between 14-19 m. Ecology Nothing else is known of the biology of this species. Figure 37 Geographic distribution of Symphurus trewavasae based on material examined (discussion of geographic distribu- tion appears in species account). Remarks Symphurus trewavasae was described by Chabanaud (1948a:508) from specimens taken in coastal waters off Cabo Frio, Brazil. In his revision of western Atlantic Symphurus, Ginsburg ( 1951:185) compared briefly the account of S. trewavasae from the literature with data for the North Atlantic S. plagiusa and suggested that S. trewavasae was pos- sibly not distinct from S. plagiusa. He noted that purported eye-size differences between the species, considered diagnostic by Chabanaud ( 1948a), did not always successfully separate them. Furthermore, Ginsburg pointed out that although there were modal differences in fin-ray counts between the two spe- cies, there was sufficient overlap in most features examined to necessitate direct comparison of the two nominal species, an analysis that he did not perform in his study. Menezes and Benvegnu (1976:145) studied both nominal species in detail and concluded that indeed they were distinct. They noted among several dis- tinct differences that S. trewavasae had more dor- sal- and anal-fin rays, more scales in a longitudinal series, a shorter gape, and a larger eye. Furthermore, they noted that S. plagiusa characteristically has a large, black spot on the upper part of the ocular-side opercle that is absent in 100 mm) occur regularly in mod- erate depths (10-30 m) on the inner continental shelf (Struhsaker, 1969; Franks et al., 1972; Topp and Hoff, 1972; Yanez-Arancibia and Sanchez-Gil, 1986; Wenner and Sedberry, 1989; this study). Of the S. plagiusa examined in the present study, most, col- lected in deeper waters on the continental shelf, were large individuals of over 120 mm. Larger fish (100- 190 mm TL; x=ca. 140 mm TL) have also been re- ported in the shrimp bycatch taken in coastal wa- ters off South Carolina (Reiser, 1976). Samples ofS. plagiusa collected in coastal waters in the South At- lantic Bight at ca. 10 m depth (Wenner and Sedberry, 1989) were almost entirely large individuals (140- 210 mm TL, x =140-150 mm TL), whereas smaller specimens (<100 mm TL) were unusual in trawls made at these locations. Wenner and Sedberry con- cluded that juveniles were not collected at deeper stations either because they occupied different habi- tats or because they did not recruit to the fishing gear until reaching a larger size. Although most records of S. plagiusa are from waters shallower than ca. 40 m, several studies re- corded this species from much gr eater depths. Bullis and Thompson (1965) and Franks et al. (1972) re- ported collecting S. plagiusa at 92 m, Staiger (1970) listed one specimen (62 mm) taken at 132 m on the Pourtales Terrace (specimen not located during this study), and Chittenden and Moore (1977) listed this species among those occurring at the 110-m bathy- metric contour off Louisiana and northern Texas (specimens not examined in this study). The deepest records for S. plagiusa examined in this study are those of one specimen (USNM 265181) taken at 183 m off the southern Bahamas (20°54'N, 73°33'W), and 25 specimens (USNM 159616) purportedly collected between 186 and 189 m (101 and 102 fm). The depth of capture for this lot is questionable, however, because coordinates for this station (RV Silver Bay 156, 29°04'N, 85°49 W) place it in a region where depths range be- tween 70 and 93 m (38 and 50 fm). Because of this un- certainty, depth-of-capture data for these specimens were not included in Table 10. Ecology More ecological information is available for S. plagiusa than for any other tongueflsh. Occur- rences of all postsettlement stages of S. plagiusa, even those taken in neritic waters, generally corre- spond with the distribution of substrates with a high percentage of silt or fine sand (Reid, 1954; Springer and Woodburn, 1960; Reichert and Van der Veer, 1991; Baltz et al., 1993). Preference for soft mud sub- strates is reflected in the almost universal occurrence of S. plagiusa in studied fish communities in pro- tected waters throughout its range. In protected ar- eas where soft substrates abound, this species is sometimes found in great abundance. Conversely, S. plagiusa occurs only sporadically and generally in much lower abundance in coarse sand habitats, such Munroe: Systematics of western Atlantic Symphurus as those in high energy surfzone areas (Modde and Ross, 1981), or on live-bottom areas (Topp and Hoff, 1972; McCaffrey, 1981; Darcy and Gutherz, 1984) with calcareous sands. In saltmarshes, juveniles oc- cur where stem density of Spartina is low ( Rakocinski et al., 1992, Baltz et al., 1993). Symphurus plagiusa is the most euryhaline of North American tongueflshes, and has been recorded at salinities of 0.0-42.9%e (Gunter, 1945; Springer and Woodburn, 1960; Tagatz,1968; Roessler, 1970; Swingle, 1971; Topp and Hoff, 1972; Shealy et al., 1974; Schwartz et al., 1981). Simmons (1957) indi- cated that S. plagiusa did not tolerate salinity much above 35 %c. Symphurus plagiusa undergoes an on- togenetic shift in habitats in relation to salinity, with smaller juveniles occupying lower salinity regions of the estuary and larger juveniles and adults moving into higher salinity areas (Gunter, 1945; Springer and Woodburn, 1960; Swingle, 1971; Ogren and Brusher, 1977). Baltz et al. (1993) indicated that an apparent ontogenetic shift in estuarine occurrence of the species may not be a primary response to a salinity gradient (see Rakocinski et al., 1992), but rather an ontogenetic shift to higher velocity microhabitats. From Chesapeake Bay and south through its range in the southern Gulf of Mexico, S. plagiusa is the most common tonguefish occurring on soft-bottom substrates and is a year-round resident in nearshore marine and estuarine waters. In fact, as Topp and Hoff (1972) noted, this species represents a signifi- cant proportion of the demersal fish community of nearly every major estuarine system through much of its range. In lower Chesapeake Bay and tributary rivers, Bonzek et al. (1993) recorded S. plagiusa as the sixth most abundant fish species overall in trawl surveys of primarily juvenile fishes. Here, tongue- fishes, 34-204 mm TL (mean size 127 mm), were taken every month of the year and were widespread at stations along lower segments of tributary rivers and in lower Chesapeake Bay. Average catch of the species was 6.5 fish/trawl, which was second in abun- dance only to the hogchoker ( Trinectes maculatus) among flatfishes. In coastal habitats along the southeast United States, S. plagiusa is also very abundant (Webster et al.3). This species was present in all major estua- rine areas and coastal regions of South Carolina sampled during a 12-month period (Shealy et al., 1974), where it ranked 13th in numerical abundance and 14th in biomass among the fishes collected. Shealy et al. (1974) considered it the most abundant tonguefish in South Carolina inshore waters, where various life history stages of S. plagiusa were present during every month of the year. In coastal waters of South Carolina, tongueflshes (primarily S. plagiusa) were reported as abundant in bycatch of penaeid shrimp fisheries from December to April and least abundant May through August and represented 2% by numbers and ranked 12th in weight of the 105 species in the fish bycatch (Reiser, 1976). At 10 m or less in nearshore environments in the South Atlan- tic Bight from Cape Fear, North Carolina, to St. John’s River, Florida, Wenner and Sedberry (1989) recorded S. plagiusa as the 11th most numerous and the fifth most abundant species in biomass of all spe- cies taken during trawling. This species was wide- spread through the area, with occurrences at 78% of stations sampled and with fewer fish being taken in winter than for other seasons. Symphurus plagiusa is also listed as a year-round resident species in Geor- gia estuaries (Dahlberg and Odum, 1970; Hoese, 1973). In Gulf of Mexico estuaries, S. plagiusa is also a common and abundant species (Springer and Woodburn, 1960; Swingle, 1971; Ogren and Brusher, 1977; Naughton and Saloman, 1978). In inshore ar- eas of the eastern Gulf where live-bottom substrates predominate, it is less abundant. McCaffrey (1981) considered this species to be only an occasional com- ponent of the fish fauna on the inner continental shelf of the northeastern Gulf of Mexico. In contrast, S. plagiusa are considered very abundant in soft mud substrate estuarine environments in the northern Gulf ( Gunter, 1945 ). At offshore locations in the west- ern Gulf, Hildebrand (1955) noted that this species was particularly common in 14-16 m off Punta Morros, Mexico, but was less abundant on the brown shrimp grounds off Texas (Hildebrand, 1954). It was considered abundant in demersal fish communities of the southern Gulf of Mexico (Yanez-Arancibia and Sanchez-Gil, 1986), where it ranked 26th among the 30 most abundant species occurring on the inner con- tinental shelf of Campeche Sound. Although catches of this species off the south At- lantic states and in the Gulf of Mexico are large enough to constitute a small percentage of industrial fisheries (Siebenhaler, 1952; Roithmayr, 1965; Ander- son, 1968), most fishermen regard tonguefish as a nuisance because these small, and relatively narrow flatfishes clog fishing nets and interfere with effi- ciency of the gear (Topp and Hoff, 1972). Symphurus plagiusa is a nondiscriminate, benthic omnivore. Throughout its life history, except for post larvae, which consume a variety of planktonic dia- toms (Strickney and Shumway, 1974), blackcheek tonguefish feed close to the substrate. Lists of dietary items include a variety of benthic prey items, algae, and sand grains (Linton, 1905; Hildebrand and Schroedei’, 1928; Reid, 1954; Springer and Woodburn, 1960; Stickney, 1976). Food items recorded for this Fishery Bulletin 96(1 ), 1998 1 12 species include copepods, amphipods, ostracods, cumaceans, brachiopods, crabs, polychaetes, and pele- cypods. The most extensive study of food habits of this species (Stickney, 1976) reveals that S. plagiusa of all sizes fed on over 40 different plant and animal taxa but primarily consumed benthic molluscs, small crustaceans, and organic matter. Plant detritus was only a minor component of the diet, indicating this material was not actively sought as a nutrient source. Possibly, 100 mm were mature with either fully elongate or gravid ovaries. The smallest gravid female was 88.8 mm, an appar- ently unusual specimen because the other 16 gravid females were 101-149 mm. Of interest is the general absence of small juve- niles of this species in the collections examined. Only six small fish (25.0-56.8 mm) were found. Because juveniles and adults of this species inhabit live-bot- tom substrates (see below), relative scarcity of small juveniles in collections may reflect the limited suc- cess of conventional trawling gear generally in cap- turing small flatfishes in this habitat, or it could also indicate that juveniles occur in habitats other than those usually sampled by trawling. Geographic distribution (Fig. 53 J A warm-temper- ate species with a fairly restricted and somewhat discontinuous distribution in the western North At- lantic from just south of Cape Hatteras, North Caro- lina, through the Gulf of Mexico to the Campeche Bank region off the Yucatan Peninsula, Mexico. There is also a single citation of this species from Cuba (Vergara Rodriguez, 1976). The occurrence of S. urospilus corresponds with the general distribution of live-bottom habitats in the region. Along the southeastern Atlantic coast of the United States, S. urospilus ranges from south of Cape Hatteras, North Carolina, to southern Florida. In the Gulf of Mexico, it has been taken at the southern tip of Florida, including the Florida Keys and Tortugas regions, and is common in the eastern Gulf along the west Florida shelf, as far north and west as Apalachee Bay (Topp and Hoff, 1972). I did not ex- amine specimens collected in the central Gulf of Figure 53 Geographic distribution of Symphurus urospilus based on mate- rial examined (discussion of geographic distribution appears in species account). Mexico, but several lots taken on the inner continen- tal shelf in the western Gulf off western Louisiana and Texas contained this species. Symphurus urospilus also occurs on live-bottom substrates in the Campeche Bank region of the Yucatan Peninsula, Mexico (Hildebrand, 1955; Topp and Hoff, 1972). Bathymetric distribution Symphurus urospilus examined in this study were taken at depths of 5— 324 m (Table 10). The center of abundance for this species, where 122/126 (96%) of the specimens ex- amined in this study were captured, occurs on live- bottom substrates in the relatively narrow depth zone between 5 and 40 m. Symphurus urospilus has not been reported from estuarine areas, and all juveniles examined, including the two smallest specimens (25.0 and 26.2 mm), were collected on live-bottom sub- strates on the inner continental shelf at depths oc- cupied by adults. Only four S. urospilus examined in this study were taken deeper than 40 m (one each at 42 and 64 m, and two at 324 m). The capture depth (324 m) for two specimens in TU 14789 is very unusual for this species because it is 260 m deeper than any other recorded for the species. Depth of capture for this station appears legitimate because other fishes col- lected in the trawl with these tonguefish.es include chlorophthalmids and macrourids, typical deep-sea species. Ecology Little is known regarding the life history of S. urospilus. From their small samples collected on the west Florida shelf, Topp and Hoff ( 1972) noted that S. urospilus were taken at bottom temperatures Munroe: Systematics of western Atlantic Symphurus 1 19 of 16.4-30.0°C and salinities of 32.8-36.2%e. Three specimens had fed on crustaceans, and one had in- gested a gastropod. Topp and Hoff (1972) also noted that specimens of S. urospilus caught off the West Florida Shelf in September had ripe and ripening gonads, and a specimen collected in late August had developing ova. Their smallest specimen (27 mm) was collected in November, further suggesting a late sum- mer-early fall spawning period for this species in the eastern Gulf of Mexico. Webster et al.3 collected 457 individuals (12.8 kg) by trawling in the South Atlantic Bight. In their study, this species ranked 70th in abundance of 244 taxa collected. Geographic variation Only slight variation was evident in meristic features (Table 33) examined in specimens from three different regions of the geo- graphic range. Symphurus urospilus from off the southeastern United States and western Gulf of Mexico had similar counts for dorsal- and anal-fin rays and total vertebrae. Counts for these features were consistently higher in specimens from these Table 33 Geographic variation in selected meristic features of Sym- phurus urospilus. Abbreviations: SEUS = southeastern United States; EGMX = eastern Gulf of Mexico; WGMX = western Gulf of Mexico including Yucatan shelf region; n = no. of specimens measured. Character n Area Mean Range SD Dorsal rays 20 SEUS 86.6 84-90 1.35 70 EGMX 85.0 82-88 1.24 26 WGMX 87.3 84-89 1.25 Anal rays 20 SEUS 70.4 67-72 1.14 70 EGMX 69.0 64-71 1.27 26 WGMX 71.0 68-74 1.37 Vertebrae 20 SEUS 45.8 45-47 0.52 70 EGMX 45.2 44-46 0.51 26 WGMX 46.3 45-48 0.60 regions than were those for S. urospilus from the eastern Gulf of Mexico. Comparisons Symphurus urospilus is one of the most distinctive species in the genus. Its unique com- bination of 11 caudal-fin rays, pupillary operculum, spotted caudal fin, and ID pattern distinguishes this species from all congeners. Other western Atlantic species with spotted fins differ in caudal-fin ray count (S. diomedeanus and S. ommaspilus have 10 cau- dal-fin rays) and either lack a caudal spot altogether, or if a caudal spot is present (occasionally in S. diomedeanus ), there are spots present also on the dorsal and anal fins. Other differences between S. urospilus and these species are discussed in the “Comparisons” sections in accounts for S. ommaspilus and S. diomedeanus. The eastern Pacific S. melasmatotheca and S. undecimplerus, the only other congeners with 1 1 cau- dal-fin rays, also have a pupillary operculum. Symphurus urospilus differs from both in peritoneal color (unpigmented vs. black or spotted in these other species), its spot on the caudal fin (absent in these others), in its mostly nonoverlapping fin-ray and ver- tebral counts (82-90 dorsal-fin rays vs. 90-98 in S. melasmatotheca and 97-105 in S. undecimplerus ; 64- 74 anal-fin rays vs. 74-80 in S. melasmatotheca and 80-87 in S. undecimplerus ; and 44-48 total verte- brae vs. 49-52 in S. melasmatotheca and 52-56 in S. undecimplerus), and ID pattern (1-4-3 vs. 1-5-3 in these others). The eastern Pacific species, HL). Symphurus tessellatus, especially juveniles and small adults (to about 150 mm), are superficially similar in overall body shape, relative eye size, and body pigmentation (crossbanding) to S. caribbeanus. However, S. caribbeanus is easily distinguished from S. tessellatus in lacking the black spot on the outer surface of the ocular-side opercle and scales on blind- side dorsal and anal fins (both present in S. tessellatus), and S. tessellatus has the posterior dor- sal and anal fins, as well as the caudal fin, uniformly darkly pigmented without alternating blotches and unpigmented areas and often has black pigment patches on the blind side of the body. In contrast, the posterior regions of the vertical fins of S. caribbean us have alternating dark blotches and unpigmented areas without a progressive darkening in coloration posteriorly in these fins, and the blind side of the body lacks black pigment patches. Symphurus caribbeanus also has modally lower counts than S. tessellatus (total vertebrae 49-50 vs. 50-53 in S. tessellatus; dorsal-fin rays 89-96 vs. 91-102; anal- fin rays 74-80 vs. 77-86; 78-89 vs. 81-96 longitudi- nal scales). Symphurus caribbeanus differs considerably from S. civitatium and S. oculellus. There is almost com- plete overlap in several meristic features between S. civitatium and S. caribbeanus , however, S. civitatium has a fleshy ridge on the ocular-side lower jaw (ab- sent in S. caribbeanus; see Fig. 3, D and E) and has lower modal counts for total vertebrae (47-49 vs. 49- 50 in 130 mm), whereas the S. oculellus were a mixture of sizes, with juveniles as small as 78 and 82 mm. Remarks Discussion of the synonymy for this spe- cies was provided in Munroe (1991:276). Comparisons Symphurus oculellus most closely resembles and is largely sympatric with S. tessel- latus, S. diomedeanus, and S. plagusia. Differences between S. oculellus and S. diomedeanus are pre- sented in the “Comparisons” section of the account for S. diomedeanus. Symphurus oculellus differs from S. tessellatus in lacking the 4-8 small, but well-de- veloped, scales on blind-side dorsal- and anal-fin rays characteristic of S. tessellatus (especially evident in specimens larger than 70 mm), a smaller eye (68- 104, x =84 HL vs. 79-114, x =95 HL in S. tessellatus ), and higher meristic values (dorsal-fin rays 97-106 vs. 91-102 in S. oculellus-, anal-fin rays 81-89 vs. 74-86; total vertebrae usually 53-54 vs. 50-53). Sym- phurus tessellatus also lacks the fleshy ridge on the ocular-side lower jaw that is usually present and well developed in S. oculellus (compare Fig. 3, D and E). And, the posterior extension of the jaws is slightly less extensive in S. tessellatus, reaching only to about the vertical through the posterior margin of the pu- pil or posterior margin of the lower eye. In S. oculellus, the jaws extend farther backwards reach- ing a vertical through the posterior margin of the eye, and in many specimens the jaws extend slightly posterior to the vertical through the posterior mar- gin of the lower eye. Symphurus oculellus has 10 to 14 (usually 10-12), narrower, crossbands; S. tessellatus generally has about nine, wide, dark-brown crossbands. In S. oculellus, the dorsal and anal fins are not uniformly dark brown or black but, instead, in the posterior two-thirds of the dorsal and anal fins there is an al- ternating series of blotches and unpigmented areas, and the blind-side inner opercular lining and isth- mus are much more lightly pigmented than corre- sponding structures on the ocular surface of the body. In S. tessellatus, the caudal fin and the posterior third of the dorsal and anal fins are usually dark brown or black and without alternating series of blotches and unpigmented areas, and the isthmus and inner opercular lining on the blind side are heavily pigmented, similar to those on the ocular side of the body. Symphurus oculellus is also similar to S. plagusia with respect to small eye size and presence of a fleshy ridge on the ocular-side lower jaw. It differs from this species, however, in its much higher counts (52-55 total vertebrae vs. 47-51 in S. plagusia; dorsal-fin rays 97-106 vs. 89-97; anal-fin rays 81-89 vs. 73- 81) and pigmentation pattern. Symphurus oculellus has sharply contrasting crossbands, pigmented blotches alternating with unpigmented areas in the dorsal and anal fins, and a black spot on outer opercle (vs. a relatively uniform body coloration with faint crossbands, uniformly pigmented fins without blotches and no pigment spot on the outer opercle in S. plagusia). Furthermore, in S. plagusia, the first, and occasionally the second, rays of the dorsal fin are usually located along a vertical line anterior to the upper eye, whereas in S. oculellus, the dorsal-fin origin usually extends anteriorly only to the vertical through the anterior margin or mideye region of the upper eye. Differences in morphometries between the two species are that S. oculellus has a narrower body (231-297 SL vs. 278-319 SL in S. plagusia) and at- tains larger sizes (up to 190 mm vs. largest of only 131 mm in S. plagusia). Symphurus oculellus is similar to S. civitatium with respect to small eye size and presence of a fleshy ridge on the ocular-side lower jaw. Differences be- tween these species are discussed in the “Compari- sons” section under the species account for S. civitatium. There are seven eastern Pacific Symphurus with similar ID patterns, comparable fin-ray counts, or pigment patterns reminiscent of those observed in S. oculellus . Of these seven species, only S. chabanaudi and S. elongatus are similar to S. oculellus in lacking a pupillary operculum. Many meristic features of S. oculellus completely overlap those of the eastern Pacific S. chabanaudi. Symphurus oculellus differs from S. chabanaudi, however, in lack- ing the 4-8 small, but well-developed scales on blind- side dorsal- and anal-fin rays prominent in S. chabanaudi, especially those larger than 60 mm; in having a somewhat smaller eye (1.2-1. 9, x = 1.5 SL in S. oculellus vs. 1.7-2. 3, x =1.9 SL), and S. oculellus has a well-developed fleshy ridge on the ocular-side lower jaw (absent in S. chabanaudi). The jaws in S. oculellus extend posteriorly to the vertical through the posterior margin of the lower eye, and in many specimens the jaws actually extend slightly beyond the posterior margin of the eyes, whereas in S. chabanaudi the posterior extension of the jaws reaches only to a vertical through the posterior mar- gin of the pupil or the posterior margin of the lower eye. Symphurus oculellus also differs from S. chabanaudi in the relative frequencies of specimens Munroe: Systematics of western Atlantic Symphurus 135 possessing 1-5-3 and 1-4-3 ID patterns. Symphurus chabanaudi has a much higher frequency of occur- rence of the 1-5-3 ID pattern (50% of individuals ex- amined) compared with only 30% with a 1-4-3 pat- tern. In contrast, 55 of 64 (86%) of the S', oculellus examined had a 1-4-3 pattern and only three speci- mens possessed a 1-5-3 pattern. Symphurus chabanaudi also differs from S. oculellus in that this species generally has about nine, wide, dark-brown crossbands compared with the more numerous ( 10- 14, usually 10-12), narrower bands in S. oculellus. In addition, in S. oculellus the posterior two-thirds of the dorsal and anal fins usually have alternating series of blotches and unpigmented areas, whereas in S. chabanaudi the posterior third of the dorsal and anal fins, and the caudal fin, are usually uniformly dark brown or black without alternating blotches and unpigmented areas. There is almost complete overlap in fin-ray and vertebral counts between those of S. oculellus and S. elongatus, however, these species are otherwise dis- tinct. Symphurus oculellus has prominent crossbands on the body and a dark blotch on the ocular-side opercle, whereas in S. elongatus the body is uniformly pigmented without crossbands and a prominent blotch on the ocular-side opercle is wanting. 136 Fishery Bulletin 96( 1 ), 1998 ' ®***«sSS8S Figure 60 Symphurus plagusia (Schneider), Neotype, ANSP 132030, female, 103.2 mm SL, Puerto Yabucoa, Puerto Rico. Symphurus plagusia (Schneider, in Bloch and Schneider, 1801) (Figs. 8G, 60-61; Tables 1-10, 38) Plagusia Browne, 1756 (Jamaica; nonbinomial; sup- pressed (Opinion 89 [Hemming and Noakes, 1958:9], Plenary Powers for nomenclatorial pur- poses, Direction 32. Published 17 May 1956). Pleuronectes plagusia Browne, 1789:445 (Jamaica; nonbinomial; suppressed (Opinion 89 LHemming and Noakes, 1958:9], Plenary Powers for nomen- clatorial purposes, Direction 32. Published 17 May 1956). Cuvier, 1816:224 (listed). Cuvier, 1829:344 (listed). Pleuronectes plagusia Schneider, in Bloch and Schneider, 1801:162 (original description based on Browne, 1789). lAchirus ornata (nomen dubium) Lacepede, 1802: 659, 663 (original description of tonguefish donated to France by Holland, but of uncertain identity and geographic origin). Aphoristia ornata. Kaup, 1858:107 (in part) (new combination; synonymized with Plagusia tessellata Quoy and Gaimard, 1824). Gunther, 1862:490 (in part) (synonymy; meristics; synonymized with Plagusia tessellata Quoy and Gaimard, 1824). Poey, 1868:409 (in part) (Cuba; synonymy). Poey, 1875- 1876:182 (in part) (Cuba; synonymy). Goode and Bean, 1885a:196 (in part; substitute name for Pleuronectes plagiusa Linnaeus, 1766). Jordan 1885:395 (in part; possible synonymy of A. ornata Lacepede, 1802, with Pleuronectes plagiusa Linnaeus, 1766 \ Aphoristia ornata Lacepede, 1802 from Jamaica distinct from A. fasciata [ =Plagusia fasciata] Holbrook in DeKay, 1842). Aphoristia plagiusa (not of Linnaeus, 1766). Jordan, 1886a:31 (Cuba; equals A. ornata of Poey). Jordan, 1886b:603 (in part) (West Indies; equals A. ornata of Poey). Symphurus plagusia. Jordan and Goss, 1889:100 (in part) (West Indies to Brazil; synonymy, nomencla- ture; comparison with S. plagiusa ; synonymized with Plagusia tessellata Quoy and Gaimard, 1824). Eigenmann and Eigenmann, 1891:73 (in part; east coast South America, West Indies; synonymy). Jor- dan and Evermann, 1898:2709 (in part; synonymy, counts, measurements, redescription; after Jordan and Goss). Evermann and Marsh, 1900:332 (in part; in key). Meek and Hildebrand, 1928:1005 (in part; Panama; synonymy; counts, measurements, redescription; distribution records). Chabanaud, 1939:26 (Antilles). Chabanaud, 1940:182 (descrip- tive osteology). Fowler, 1941:146 (in part; Brazil- ian localities). Chabanaud, 1949:82 (mouth of Amazon River; synonymy; redescription, counts, measurements, scales; figures; radiograph). Boeseman, 1956:197 (Suriname). Duarte-Bello and Buesa, 1973:234 (in part; Cuba; synonymy). Menezes and Benvegnii, 1976:142 (in part; Brazil; recommended re-examination of Ginsburg’s diag- noses of two subspecies). Guitart, 1978:728 (in part; Cuba; in key, figure, meristic features, color de- scription; finray counts probably include another species). Rosa, 1980:222 (in part; Paraiba, Brazil; nearshore, estuarine habitats). Lema et al., 1980:44 (in part; southern Brazil; synonymy). Munroe: Systematics of western Atlantic Symphurus 137 Correa et al., 1986:37 (Brazilian localities; com- mon names; figure). Garzon-F., 1989:158 (Bahia de Portete, Colombia; abundance). Munroe, 1991:256 (Greater Antilles and Central America to southern Brazil; redescription, diagnosis, des- ignation of neotype; nomenclature; synonymy; counts, measurements, photograph; in key; bathy- metric distribution; size and sexual maturity). Munroe, 1992:371, 382 (ID pattern; geographic, bathymetric distributions). Symphurus plagusia plagusia. Ginsburg, 1951:199 (in part; synonymized with Plagusia tessellata Quoy and Gaimard, 1824; description and diag- noses of subspecies; four species included in mate- rial studied). Cervigon, 1961:42 (Venezuela). Carvalho et al., 1968:22 (in part; Antilles, Central America to Brazil; brief description; in key). Palacio, 1974:87 (in part; Colombia; counts; suggested re- examination of subspecies status). Lema and Oliveira, 1977:6 (Brazil; in key; suggested synonymy of Pleuronectes plagusia , Plagusia tessellata, and Symphurus civitatium). Soares, 1978:23 (in part; northern Brazil). ISymphurus plagusia (Linnaeus, 1766). Valdez and Aguilera, 1987:175 (in part; Gulf of Venezuela; description, figure). Misidentification Seret and Andreata, 1992:94 (southern Brazil; 640 m; five specimens actually S. marginatus). Diagnosis Symphurus plagusia is distinguished from all congeners by the following combination of characters: predominant 1-4-3 ID pattern; 12 cau- dal-fin rays; 4 hypurals; 89-97 dorsal-fin rays; 73- 81 anal-fin rays; 47-51, usually 49-51, total verte- brae; 79-89 scales in longitudinal series; absence of pupillary operculum; unpigmented peritoneum; fleshy ridge on ocular-side lower jaw; ocular-side dentary without teeth, or with short row of small teeth developed only on anterior one-half to one-third of jaw margin; anterior region of ocular-side premax- illa usually with small, mostly incomplete row of teeth along margin; relatively small, spherical eye (64-95 HL, x=82); moderately long jaws, usually extending posteriorly to vertical line through poste- rior margin of lower eye, less frequently to vertical through posterior margin of pupil or slightly poste- rior to posterior margin of lower eye; dorsal-fin ori- gin far forward, usually at vertical through anterior margin of upper eye, or with first and sometimes second rays inserting anterior to vertical through anterior margin of upper eye; scales absent on blind sides of dorsal- and anal-fin rays; ocular surface pig- mentation usually uniformly light brown or yellow- ish, occasionally with 8-14, narrow, faint crossbands, but without blotch on caudal region; outer surface of ocular-side opercle without black blotch, pigmenta- tion usually same as on body (some specimens with dusky blotch on upper opercular lobe as a conse- quence of pigment on inner lining of ocular-side opercle showing through to outer surface); inner lin- ing of ocular-side opercle and isthmus dusky to dark brown, that of blind side usually unpigmented or occasionally with small patch of pepper-dot pigmen- tation on ventral margin; blind side without pepper- dot pigmentation; dorsal and anal fins uniformly pig- mented, without spots or blotches and without pro- gressive darkening or alternating series of pigmented blotches and unpigmented areas posteriorly; caudal fin without spots or blotches. Description A medium-size species attaining maxi- mum sizes of ca. 130 mm SL. ID pattern usually 1-4-3 (33/44 individuals), less frequently 1-3-3 (5), 1-3-4 (3), or 1-4-2 (2) (Table 2). Caudal-fin rays usually 12 (41/44), infrequently 10, 11, or 13 (Table 3). Dorsal- fin rays 89-97, usually 91-96 (Table 4). Anal-fin rays 73-81, usually 75-79 (Table 5). Total vertebrae 47- 51, usually 49-51 (39/44 specimens) (Table 6). Hypurals 4 (43/43). Longitudinal scale rows 79-89 (Table 7). Scale rows on head posterior to lower orbit 18-22, usually 18-20 (Table 8). Transverse scales 35- 43 (Table 9). Proportions of morphometric features presented in Table 38. Body relatively deep, with greatest depth in anterior one-third of body; body depth tapering fairly gradually posterior to midpoint. Preanal length shorter than body depth. Head wide, somewhat nar- rower than body depth. Head length usually much shorter than head width (HW:HL=1.2-1.3, x=1.2). Lower head lobe narrow, its width nearly equal to postorbital length; considerably narrower than up- per head lobe. Lower opercular lobe of ocular side considerably wider than upper opercular lobe. Snout moderately long, somewhat square (Fig. 60), covered with small ctenoid scales. Dermal papillae well de- veloped on snout and chin regions on blind side of body. Anterior nostril on ocular side short, when de- pressed posteriorly, usually falling just short of an- terior margin of lower eye. Jaws long; maxilla usu- ally reaching posteriorly to vertical through poste- rior margin of lower eye, less frequently only reach- ing to vertical through posterior margin of pupil or vertical slightly posterior to posterior margin of lower eye. Ocular-side lower jaw with distinct, fleshy ridge near posterior margin (Fig. 3D). Teeth well devel- oped on blind-side jaws. Ocular-side dentary with- out teeth or with short row of small teeth developed only on anterior one-half to one-third of margin; pre- 138 Fishery Bulletin 96(1 ), 1998 Table 38 Morphometries for neotype (ANSP 132030) and 14 non- type specimens of Symphurus plagusia. (Abbreviations defined in methods section; SL is expressed in mm; char- acters 2 to 14 are expressed in thousandths of SL; 15 to 21 in thousandths of HL; n = no. of specimens measured.) Character Neotype n Range Mean SD 1. SL 103.2 14 57.4-130.3 98.8 22.29 2. BD 304 14 278-319 292.1 13.05 3. PDL 30 14 23-50 32.9 6.87 4. PAL 222 14 166-244 209.3 18.60 5. DBL 970 14 950-977 967.1 6.87 6. ABL 776 14 758-802 785.6 15.38 7. PL 64 14 51-73 63.6 5.76 8. PA 48 14 38-60 50.0 7.00 9. CFL 98 14 88-111 100.3 7.13 10. HL 196 14 174-216 189.6 11.98 11. HW 239 14 218-256 236.4 13.25 12. POL 130 14 110-143 125.9 9.26 13. UHL 142 14 125-186 160.1 15.93 14. LHL 107 14 81-115 96.8 10.22 15. POL 663 14 630-714 665.8 25.18 16. SNL 228 14 205-250 229.1 15.62 17. UJL 213 14 200-250 227.6 14.99 18. ED 79 14 64-95 81.9 9.57 19. CD 213 14 222-374 275.1 40.32 20. OPLL 272 14 250-346 296.9 29.10 21. OPUL 223 14 169-272 211.9 27.24 maxilla on ocular side usually with small, single, mostly incomplete row of slender teeth on margin anterior to vertical equal with anterior nostril. Chin depth slightly larger than snout length. Lower eye small, spherical; eyes slightly subequal in position with upper usually slightly in advance of lower eye. Anterior and medial surfaces of eyes not covered with scales; usually 1-2 small ctenoid scales in narrow interorbital region. Pupillary operculum absent. Dorsal-fin origin far forward (Fig. 3D), usually at vertical through anterior margin of upper eye, or with first and sometimes second dorsal-fin rays inserting anterior to vertical through anterior margin of up- per eye. Scales absent on blind sides of dorsal- and anal-fin rays. Pelvic-fin short; longest pelvic-fin ray reaching base of first, or occasionally second, anal- fin ray. Posteriormost pelvic-fin ray connected to body by delicate membrane terminating immediately an- terior to anus or occasionally extending posteriorly almost to anal-fin origin (membrane torn in many specimens). Caudal-fin length moderate. Scales large, ctenoid on both sides of body. sal and anal fins; mostly complete in anterior trunk region; on rest of body obvious only as vertical mark- ings at body margin along dorsal- and anal-fin bases. Blind side creamy white. Peritoneum unpigmented. Pigmentation of outer surface of ocular-side opercle usually same as that of body; occasionally with dusky blotch on upper opercular lobe due to pigment on inner lining of ocular-side opercle showing through to outer surface. Inner lining of opercle and isthmus on ocular side usually dusky; some specimens with dark brown pigmentation on inner opercular lining; inner opercle and isthmus on blind side usually un- pigmented or occasionally with small patch of pepper-dot pigmentation on ventral margin. Usually with slight pigment band on ocular-side upper lip and diffuse pattern of melanophores on lower lip. Dorsal and anal fins dusky throughout their lengths; fin rays streaked with pigment darker brown than that of connecting membrane, thereby clearly outlining each fin ray; sometimes with alternating series of darker pigmented rays (usually 2-3 in suc- cession) separated by about 4-5 successive, lighter pigmented rays. Basal half (scale-covered) of caudal fin dark brown; fin rays in distal one-half of caudal fin streaked with darker pigment than connecting membrane. Size and sexual maturity (Fig. 8GJ Symphurus plagusia is a medium-size species attaining sizes of about 130 mm. Males and females attain similar sizes. The largest of five males examined in this study was 130 mm; the largest of 24 females was only slightly smaller (127 mm). Sexual maturity occurs at a relatively large size in this species. All females larger than 80 mm were mature. All but two females smaller than 80 mm were immature with gonads undergoing elongation without ripening ova or with ovaries barely elongating. Geographic distribution (Fig. 61 J Widely distrib- uted in shallow waters of the tropical western At- lantic. In the northern portion of its range, this spe- cies occurs in Puerto Rico, Cuba, and Hispaniola but is unknown from the Bahamas (Bdhlke and Chaplin, 1968). Along the continental margin of Central America, S. plagusia has been collected at Belize, Nicaragua, Costa Rica, and Panama, whereas far- ther south it ranges along the Atlantic coast of Co- lombia, and coastal regions of Guyana, Suriname, Tobago, and Brazil as far south as Rio de Janeiro. Pigmentation (Fig. 60J Body coloration similar for Bathymetric distribution Symphurus plagusia is a both sexes. Ocular surface usually uniformly light shallow-water species (1-51 m) most commonly in- brown or yellowish, occasionally with 8-14, narrow, habiting mud substrates between the shoreline and faint crossbands. Crossbands not continued onto dor- 10 m (Table 10), where 21/26 (81%) of specimens ex- Munroe: Systematics of western Atlantic Symphurus 139 Figure 61 Geographic distribution of Symphurus plagusia based on mate- rial examined (discussion of geographic distribution appears in species account). amined were taken. All life history stages occur in these shallow areas and only occasionally were indi- viduals taken at deeper locations (one specimen at 51m, three specimens at 40 m, and one specimen at 37 m). Little is known concerning the biology of S. plagusia. Its general rarity in collections indicates that it occurs in rarely sampled habitats. Remarks Discussion of nomenclature, synonymy, and designation of a neotype was provided in Munroe (1991). UMML 34347 was incorrectly listed in Munroe as having been collected off Panama. The correct locality information for this specimen is off Guyana at 7°42’N, 57°32'W. Comparisons Of western Atlantic tonguefishes, S. plagusia most closely resembles S. civitatium. Dif- ferences between these species are discussed in the “Comparisons” section in the account for S. civita- tium. Differences between 60 mm) S. plagiusa, there are 4-8 ctenoid scales on blind-sides of the dorsal- and anal- fin rays (scales usually absent altogether, or occa- sionally 1-2 scales along bases of fin rays in S. plagusia). Other differences between these species are discussed in the “Comparisons” section in the account of S. plagiusa. Meristic values of S. plagusia overlap with those of six eastern Pacific species possessing either a 1-4-3 or 1-5-3 ID pattern (Munroe, 1992). Of these, S. plagusia is most similar to S. melanurus in that both possess a fleshy ridge on the ocular-side lower jaw, and both have the first dorsal-fin ray reaching a ver- tical equal with, or anterior to, the anterior margin of the upper eye. The two species are distinguished in that S. plagusia lacks a pupillary operculum (vs. a weakly developed pupillary operculum usually present in S. melanurus), has fewer scales in longi- tudinal series (79-89 vs. 89-108 in S. melanurus), has a lightly pigmented inner lining on the blind- side opercle (vs. darkly pigmented inner lining on the blind-side opercle in S. melanurus), and in S. plagusia the posterior dorsal and anal fins and the cau- dal fin are not darker than the anterior regions (vs. progressive darkening in posterior dorsal and anal fins and darkly pigmented caudal fin in S. melanurus). Munroe: Systematics of western Atlantic Symphurus 141 Symphurus tessellatus (Quoy and Gaimard, 1824| (Figs. 9F, 62-63; Tables 1-10, 39-40J Plagusia tessellata Quoy and Gaimard, 1824:240 (original description; Rio de Janeiro Bay [ = Guanabara Bay], Brazil; counts, color description). Plagusia brasiliensis Agassiz in Spix and Agassiz, 1831:89 (original description; Bahia, Brazil; counts, color figure). Castelnau, 1855:79 (brief description, figure). Whitehead and Myers, 1971:495 (nomen- clature and dating of Spix and Agassiz’s Brazilian Fishes). Kottelat, 1984:150 (in type catalogue, MHNN). Kottelat, 1988:79 (nomenclature, type status of species described in Spix and Agassiz’s Brazilian Fishes). Aphoristia ornata. Kaup, 1858:106 (in part; South America; synonymy). Gunther, 1862:490 (in part; Atlantic coasts of tropical America; synonymized with S. plagusia Schneider, in Bloch and Schneider, 1801; brief description; counts). Kner, 1865-67:292 (Rio de Janeiro, Brazil). Symphurus plagusia (not of Schneider, in Bloch and Schneider, 1801). Jordan and Goss, 1889:324 (in part; West Indies to Rio de Janeiro, Brazil; syn- onymy; in key; brief redescription; nomenclature). Berg, 1895:79 (in part; Mar del Plata-Montevideo; counts include those for S. jenynsi). Jordan and Evermann, 1898:2709 (in part; after Jordan and Goss, 1889). Evermann and Marsh, 1900:332 (West Indies to Brazil; common; in key; synonymy; rede- scription, counts, measurements; comparison with S. plagiusa). Thompson, 1916:416 (in part; after Jordan and Goss; counts, measurements, brief color description). Devincenzi, 1920:135 (Rio de la Plata, Uruguay; counts, measurements; distin- guished from S. jenynsi). Devincenzi, 1924-26:281 (Uruguay; counts). Meek and Hildebrand, 1928: 1005 (in part; Panama; color description; counts). Beebe and Tee-Van, 1928:77 (Haiti; color descrip- tion with figure; size). Puyo, 1949:178 (in part; French Guyana; figure, counts, color description). Lowe-McConnell, 1962:694 (in part; British Guiana). Caldwell, 1966:84 (offshore localities, Jamaica). Cervigon, 1966:816 (Venezuela; probably S. tessellatus based on high meristic features, color description, and large sizes reported). Palacio, 1974:87 (in part; Colombia; specimens misidenti- fied as S. p. plagusia). Menezes and Benvegnu, 1976:142 (Brazil; synonymized with S. plagusia). Soares, 1978:23 (Rio Grande do Norte, Brazil; counts, color description, figure). Lema et al., 1980:44 ( Rio de la Plata region, Rio Grande do Sul, Brazil; synonymy). Rosa, 1980:222 (in part; nearshore and estuarine habitats, Paraiba, Bra- zil). Lucena and Lucena, 1982:56 (southern Bra- zil). Matsuura, 1983:463 (French Guiana, Suri- name; counts, measurements, color photograph). Aphoristia fasciata (not of DeKay, 1842). Goode and Bean, 1896:458 (Jamaica; in key; figured). Symphurus plagusia tessellata. Ginsburg, 1951:199 (diagnosis and description of subspecies; Brazil- Uruguay). Ringuelet and Aramburu, 1960:91 (Ar- gentina; in key; figure; synonymy). Carvalho et al., 1968:22 (in part; northern Brazil; synonymy; in key; brief description). Lazzaro, 1973:247 (south- ern Brazil and Uruguay; in key). Palacio 1974:87 (north of Puerto, Colombia). Lazzaro, 1977:70 (Uru- guay; in key). Lema and Oliveira, 1977:7 (Santa Catarina, Brazil; in key). Menni et al., 1984:201 142 Fishery Bulletin 96( 1 ), 1998 (Uruguay and Argentina; partial synonymy; com- mon names). Symphurus pterospilotus (not of Ginsburg, 1951). Lema and Oliveira, 1977:7 (in part; southern Brazil). Symphurus tessellatus (Quoy and Gaimard). Munroe, 1991:269 (Greater Antilles and Central America to Uruguay; removed from synonymy of S. plagusia (Schneider, in Bloch and Schneider); redescription and diagnosis; nomenclature; synonymy; counts, measurements, photograph; in key; bathymetric distribution; size and sexual maturity). Munroe, 1992:371, 382 (ID pattern; geographic, bathymet- ric distributions). Cervigon et al., 1993:305-306 (Venezuela; descriptive characters; distribution; figure). Diagnosis Symphurus tessellatus is distinguished from all congeners by the following combination of characters: predominant 1-4-3 ID pattern; 12 cau- dal-fin rays; 4 hypurals; 91-102 dorsal-fin rays; 74- 86, usually 78-84, anal-fin rays; 48-54, usually 50- 53 total vertebrae; 81-96, usually 83-93 scales in longitudinal series; unpigmented peritoneum; mod- erately large eye (79-114 HL, ic=95) without pupil- lary operculum; 4-8 small ctenoid scales on blind sides of dorsal- and anal-fin rays (best developed on fin rays in posterior one-third of body in specimens larger than 70 mm); lacking fleshy ridge on ocular- side lower jaw; moderately long jaws usually extend- ing to vertical through middle or posterior margin of pupil of lower eye; margin of ocular-side dentary usu- ally with single, mostly incomplete row of teeth; pre- maxilla on ocular side either lacking teeth or with very short row of teeth on anterior margin; dorsal- fin origin reaching vertical through anterior margin of upper eye, or occasionally only reaching vertical through middle of upper eye; ocular-surface pigmen- tation dark to light brown, with 5-9 well-developed, sharply contrasting, relatively wide, dark brown crossbands on head and trunk, but without pig- mented blotch on caudal region of body; distinct, dark brown or black, almost spherical blotch on outer sur- face of ocular-side opercle; inner lining of opercle and isthmus heavily pigmented on both sides of body; dorsal and anal fins without an alternating series of pigmented blotches and unpigmented areas and with- out spots; anterior dorsal- and anal-fin rays usually streaked with brown pigment; dorsal- and anal-fin rays and membranes on posterior two-thirds of body becoming progressively darker posteriorly; males with posteriormost regions of fins almost uniformly black, whereas in females, posterior portions of fins, although darker than anterior regions, usually dark brown and not as intensively pigmented as in ma- ture males; caudal fin without spots or blotches. Description A large species attaining maximum sizes to 220 mm SL. ID pattern (Table 2) usually 1-4-3 (209/278 specimens), less frequently 1-5-3 (15), 1-4- 2 (11), or 1-3-3 (10). Caudal-fin rays usually 12 (249/ 273), less frequently 10, 11, or 13 (Table 3). Dorsal- fin rays 91-102, usually 93-101 (Table 4). Anal-fin rays 74-86, usually 78-84 (Table 5). Total vertebrae 48-54, usually 50-53 (275/282) (Table 6). Hypurals 4 (273/273). Longitudinal scale rows 81-96, usually 83-93 (Table 7). Scale rows on head posterior to lower orbit 18-23, usually 20-22 (Table 8). Transverse scales 38-45 (Table 9). Proportions of morphometric features presented in Table 39. Body relatively elongate, only moderately deep; with greatest depth usually occurring in ante- rior one-third of body; body depth tapering fairly gradually posterior to midpoint. Preanal length con- siderably shorter than body depth. Head wide, some- what narrower than body depth. Head length shorter than head width (HW:HL=1. 1-1.4, 5c =1.2). Lower head lobe width somewhat less than postorbital length; narrower than upper head lobe. Lower oper- cular lobe on ocular side wider than upper opercular lobe. Snout moderately long and somewhat pointed; covered with small ctenoid scales. Dermal papillae well developed, but not particularly dense, on snout Table 39 Morphometries for 22 specimens of Symphurus tessellatus. (Abbreviations defined in methods section; SL is expressed in mm; characters 2 to 14 are expressed in thousandths of SL; 15 to 21 in thousandths of HL; n = no. of specimens measured). Character n Range Mean SD 1. SL 22 97.9-203 145.0 27.66 2. BD 22 247-312 280.2 18.82 3. PDL 22 32-48 41.7 4.48 4. PAL 22 181-227 204.7 10.58 5. DBL 22 952-968 958.3 4.48 6. ABL 22 771-876 798.0 22.90 7. PL 22 44-73 59.0 6.47 8. PA 22 27-56 41.5 6.01 9. CFL 22 72-118 90.9 10.36 10. HL 22 170-199 186.6 7.37 11. HW 22 193-247 218.6 15.58 12. POL 22 117-135 125.9 5.38 13. UHL 22 113-163 143.3 12.03 14. LHL 22 80-114 97.8 10.56 15. POL 22 593-723 674.9 25.07 16. SNL 22 196-231 215.7 9.25 17. UJL 22 222-278 248.1 15.58 18. ED 22 79-114 95.2 10.06 19. CD 22 173-322 245.0 31.85 20. OPLL 22 243-359 306.8 31.68 21. OPUL 22 161-252 205.7 24.03 Munroe: Systematics of western Atlantic Symphurus 143 and chin regions on blind side of body, occasionally extending onto ocular-side snout. Anterior nostril, when depressed posteriorly, not reaching anterior margin of lower eye. Jaws long; maxilla usually reaching posteriorly to point between verticals through middle and posterior margin of pupil of lower eye. Ocular-side lower jaw lacking fleshy ridge (Fig. 3E). Teeth well developed on blind-side jaws. Mar- gin of ocular-side dentary usually with single, mostly incomplete row of slender teeth; margin of ocular- side premaxilla either with very short row of teeth anterior to vertical through base of anterior nostril or lacking teeth altogether. Chin depth slightly larger than snout length. Lower eye moderately small; eyes slightly subequal in position with upper usually slightly in advance of lower eye. Anterior and me- dial surfaces of eyes not covered with scales; usually 1-3 small ctenoid scales in narrow interorbital re- gion. Pupillary operculum absent. Dorsal-fin origin usually reaching vertical through anterior margin of upper eye, or occasionally only reaching vertical line through middle of upper eye; predorsal length short. Four to eight scales present on blind sides (Fig. 4A) of dorsal- and anal-fin rays (best developed on fin rays in posterior one-third of fin of specimens larger than 70 mm). Pelvic fin short; longest pelvic- fin ray, when extended posteriorly, usually reaching base of first anal-fin ray, or occasionally falling short of that point. Posteriormost pelvic-fin ray connected to body by delicate membrane terminating immedi- ately anterior to anus, or occasionally extending pos- teriorly almost to anal-fin origin (membrane torn in most specimens). Caudal fin short. Scales large, strongly ctenoid on both sides of body. Pigmentation (Fig. 62J General pattern of body pigmentation similar in both sexes at all sizes but usually more intense in sexually mature males. Males, especially those in breeding condition (col- lected with gravid females), usually with more in- tense banding, dark black fins, dark black spot on ocular-side opercle, and some specimens with irregu- larly shaped, black pigment patches on posterior one- half of blind side of body. In contrast, mature females also with crossbands, but less conspicuous than in males and with posterior portions of fins dark brown but usually not black. Females lack black pigment patches on blind side observed in males. Ocular-surface background pigmentation ranging from dark to light brown. Body usually with 5-9 (usu- ally 5-7) well-developed, sharply contrasting, rela- tively wide, dark brown crossbands on head and trunk. First two bands relatively consistent in posi- tion; first crossing head immediately posterior to eyes; second crossing body immediately behind oper- cular opening. Crossbands on trunk variable in num- ber and degree of completeness, especially those be- tween opercular opening and point about equal to two-thirds of trunk length. Males usually with 3-4 well-developed and lesser number of incomplete bands along trunk. Two posteriormost bands, just anterior to caudal-fin base, slightly arched and usu- ally darker than others on body. Blind side usually uniformly creamy white; some mature males with irregular patches of black pigment on caudal one- third of blind side. Peritoneum unpigmented. Outer surface of ocular-side opercle usually with distinct, dark brown or black spot on ventral margin slightly anterior to posterior margin of opercle. Opercular spot ranging from almost spherical to dorsoventrally elongate black blotch covering most of lower opercle. Intensity of pigmentation in spot maximally devel- oped in sexually mature adults. Inner linings of opercles and isthmus on both sides of body heavily pigmented. Pigment band well developed on ocular- side upper lip; ocular-side lower lip frequently spot- ted, but without well-defined band. Anterior dorsal- and anal-fin rays usually streaked with brown pigment, more heavily pigmented than connecting membranes. Fin rays and membranes of dorsal and anal fins on posterior two-thirds of body becoming increasingly darker posteriorly. Males with posteriormost regions of fins almost uniformly black, whereas in females, posterior portions of fins, although darker than anterior regions, usually dark brown and not as intensively pigmented as in mature males. Cau- dal fin dark brown or black throughout its length. Size and sexual maturity (Fig. 9FJ Symphurus tessellatus is one of the largest species in the genus and is the second largest species of Atlantic tonguefish after S.jenynsi (Ginsburg, 1951; Menezes and Benvegnu, 1976; Munroe, 1987, 1991). Size-re- lated life history information is based on data from 385 fish. Males and females attain nearly similar sizes, but females are somewhat larger. The largest fish measured in this study was a female of 220 mm; the largest male measured 205 mm. There were 214 males (51.5-205 mm), 155 females (49.5-220 mm), and 16 immature fish (13.4-65.8 mm) of indetermi- nate sex among material examined. Mature females (n=124) ranged in size from 104 to 220 mm. Based on reproductive stages for females, sexual maturity in this species occurs at sizes of 104-120 mm, but usually larger than 115 mm. Most mature females exceeded 140 mm, with only nine smaller than 125 mm and two smaller than 1 10 mm among fish exam- ined. Thirty-one females of 49.5-119 mm were im- mature. The smallest of these, measuring 49.5 and 62.8 mm, had scarcely elongate ovaries. Other imma- 144 Fishery Bulletin 96(1 ), 1998 ture females (68.6-119 mm) had only partially elon- gate ovaries without indications of developing ova. Geographic distribution (Fig. 63) A widespread tropical species ranging from the larger Caribbean Islands such as Cuba, Hispaniola, and Puerto Rico, south to Uruguay. In the West Indies, adults and ju- veniles have frequently been taken in abundance at several localities but appear to be limited to soft silt and mud sediments which are more common on the larger islands with riverine and estuarine habitats. Symphurus tessallatus has been taken at several inshore locations in Puerto Rico, Cuba, and Haiti, and a large number of adults were collected by the RV Oregon on the shelf area southwest of Jamaica (Caldwell, 1966). Juveniles have been taken from several inshore areas in Jamaica as well. Along the continental margin S. tessellatus has been frequently captured on muddy bottoms from Belize ( 17°12'N) south to Uruguay (ca. 37°S). Absence of this species in the Yucatan region may be explained by upwelling (Rivas, 1968) or by different sediments in this region. The Yucatan Shelf is a broad lime- stone plateau with a minimum of land-derived de- Figure 63 Geographic distribution of Symphurus tessellatus based on mate- rial examined (discussion of geographic distribution appears in species account) trital sediments (Harding, 1964;Topp and Hoff, 1972). Sediments on the inner shelf off the Yucatan Penin- sula are firm, consisting of skeletal remains of vari- ous planktonic and benthonic organisms, ooids, cal- careous pellets, lithic fragments, and grapestone ag- gregates, instead of soft silt and mud typical of more southern locations. This dramatic change in sub- strates to firmer sediments in the Yucatan region may account for the absence of S. tessellatus in the wa- ters off southern Mexico. Symphurus tessellatus is one of the most abundant and frequently collected tonguefish species, especially in trawls, from Belize and Honduras south to Ven- ezuela and along the entire coastline of northern South America from the Guianas to about southern Brazil (Meek and Hildebrand, 1928; Cervigon, 1966; Palacio, 1974; Carvalho et al., 1968; Menezes and Benvegnu, 1976). Menezes and Benvegnu (1976) de- scribed S. tessellatus as the most abundant tongue- fish collected along the Brazilian coast from about 26°49'S to 4°S in northern Brazil. South of 28°S, it appears to be much rarer, and all specimens I exam- ined from Rio Grande do Sul and southwards were juveniles. This suggests that adult S. tessellatus are not regular components of the ichthyofauna of Uru- guay and northern Argentina, but that juveniles ei- ther seasonally migrate into, or are passively trans- ported into, the waters off Uruguay and northern coastal Argentina. Thus it appears that the region south of Rio Grande do Sul, which comes under peri- odic influence from the cold Falkland Current, does not harbor large populations of this essentially tropical species so common in warmer waters farther north. The specimen from the inner continental shelf of Argentina identified by Lazzaro ( 1973) as S. plagiusa and listed in the distribution section for S. plagusia ( -S . tessellatus in the present study) by Menezes and Benvegnu ( 1976) is probably not S. tessellatus. From the counts and figure provided by Lazzaro, it more closely matches S. trewavasae in meristic features and general body shape. Bathymetric distribution Throughout its range, juvenile S. tessellatus are commonly taken by beach seine in nearshore habitats, and larger adults are frequently captured by trawl in deeper waters. Indi- viduals have been collected from depths of 1 to 86 m (Table 10). There is an ontogenetic migration off- shore. Juveniles occur commonly in medium- to high- salinity regions of estuaries and in high-salinity, soft- bottom habitats in nearshore mudflats. Adults gen- erally range into deeper water, although a few large fishes that I examined were taken in relatively shal- low water. Most (352/374, 94%) of the S. tessellatus examined in this study were collected between 1 and Munroe: Systematics of western Atlantic Symphurus 145 70 m (Table 10), but the majority of captures, and the center of abundance for this species, occurs in depths between 1 and 50 m (82% of the individuals in this study). The deepest captures are for a single specimen taken at 86 m and 21 individuals at 73 m. The majority of shallow water captures were speci- mens smaller than 130 mm. Interestingly, Menezes and Benvegnu (1976) re- ported that in southern Brazil, 50 m) environments. Symphurus nebulosus, S. marginatus, and S. billykrietei have been collected at depths ranging from 500 to 810 m, and are among the deepest dwelling of western At- lantic flatfishes. Other deep-dwelling flatfishes in this region include Reinhardtius hippoglossoides and Hippoglossus hippoglossus, reportedly from depths reaching 2,000 m (Nielsen, 1986), Glyptocephalus cyno- glossus, collected at about 1,570 m (Scott and Scott, 1988), and Chascanopsetta lugubris danae, ranging from 160 to 460 m (Amaoka and Yamamoto, 1984). Previous researchers (Ginsburg, 1951; Topp and Hoff, 1972; Menezes and Benvegnu, 1976; Munroe, 1990, 1991) noted that Atlantic tonguefish species generally inhabit rather discrete depth zones. Data Munroe: Systematics of western Atlantic Symphurus 151 summarized herein indicate that for the majority of western Atlantic species, the center of abundance of the adult population is usually concentrated within a relatively narrow depth range. Among western At- lantic species, S. nebulosus is unique in that its bathymetric center of occurrence (500-810 m) is nearly completely allotopic from that of its western Atlantic congeners. Nor does any other western At- lantic Symphurus have as wide a bathymetric range as S. nebulosus. The approximately 600-m-wide depth range noted for this species is typical for most other deep-water tonguefishes with the 1-2-2 ID pat- tern in other geographic regions (Munroe, 1992). Presumably, in contrast to the more dynamic condi- tions on the inner continental shelf, environmental parameters in deep-sea environments are more uni- form over a broader depth range, with this unifor- mity over depth reflected in the distinctively broader depth range of deepwater tonguefishes. Topp and Hoff ( 1972 ) examined distributional pat- terns of 18 species of flatfishes, including those of five tonguefishes, inhabiting the inner continental shelf off west Florida. They concluded that through resource partitioning these flatfishes co-exist sym- patrically without competing for resources. However, any hypothesis that invokes ecological co-existence of multiple species as resulting only from reduced competition through resource partitioning completely ignores historical information about the species, es- pecially evolutionary information directly related to the distributional ecology of that species. Distributional patterns of western Atlantic tonguefishes have not been examined within the con- text of the evolutionary history of the genus. Although this information would best be analyzed within the framework of hypothesized relationships of the spe- cies, such an hypothesis of intrageneric relationships of species of Symphurus is unavailable. Information (Munroe, 1992) used to define species groups within Symphurus provides a preliminary framework to serve as a basis for proposing testable hypotheses regarding distributional ecology of these fishes. For example, members of each species group as defined by ID pattern have a bathymetric distribution some- what different from that of most other groups (Munroe, 1992). Species with the 1-4-3 ID pattern are primarily shallow-water inhabitants with most member species commonly inhabiting depths shal- lower than 100 m ( S . oculellus to 110 m). Several species with this ID pattern, in fact, occur within estuarine and extremely shallow (<1 m) coastal en- vironments. Species with a 1-4-2 ID pattern occur predominantly on the inner continental shelf. West- ern Atlantic species with the 1-3-2 ID pattern inhabit a bathymetric range from ca. 1 to 750 m, but most occur in deeper waters (between 30 and 200 m) on the continental shelf. There are two ecological groups of species with this ID pattern that have quite dif- ferent bathymetric distributions. Species possessing the 1-3-2 ID pattern and a black peritoneum live on the continental shelf: S. pelicanus (30-150 m); S. pusillus (100-233 m); S. ginsburgi (110-300 m); S. billykrietei (48-650 m); S. stigmosus (192-373 m), S. piger (92-549 m), and 110 mm SL). The majority of S. plagusia and S. caribbeanus (80% and 83%, respectively), including all juveniles examined, were collected in waters shallower than 20 m, with most taken by beach seine and small otter trawls in less than 10 m on nearshore mudflats, in mangrove habitats, and other estuarine locations. Symphurus oculellus, although occurring sympat- rically with S. plagusia and S. tessellatus (see Figs. 59, 61, and 63), apparently has a different life his- tory than these other species. Symphurus oculellus 154 Fishery Bulletin 96( 1 ), 1998 inhabits deeper waters than the others (Table 10), spanning an overall bathymetric range from 7 to 110 m, but is captured most frequently in waters deeper than 20 m (83% collected deeper than 20 m). Symphurus oculellus, including juveniles as small as 76 mm SL, have been collected in neritic waters deeper than 7 m, and none have been collected from estuarine habitats contrary to the capture depths of S. plagusia and S. tessellatus. However, estuarine environments in the geographic range of S. oculellus along northeastern South America have not been as thoroughly sampled as have the nearshore habitats occupied by S. plagusia and juvenile S. tessellatus in the northern Caribbean and southern Brazilian ar- eas. Symphurus civitatium, the northernmost-occur- ring species in this group, is the only Atlantic spe- cies with a distribution that is allopatric in compari- son with that of other members of this species group. Depth distribution and substrate requirements of this species were discussed above. The majority of the S. jenynsi population is allo- patric in comparison with populations of other tonguefishes. This species lives on mud bottoms on the inner continental shelf in the South Atlantic. The bathymetric distribution spans depths ranging from about 17 m to 190 m (Menezes and Benvegnu, 1976), but most specimens have been collected between 12 and 25 m. In this region, S. jenynsi is sometimes col- lected with juvenile S. tessellatus. Factors influencing ecological distributions ob- served for western Atlantic tonguefishes are rather complex. On a geographic scale, the species’ range corresponds mostly with the limits of previously iden- tified faunal regions. The ecological, or local, distri- bution of a species is reflected in the spatial and bathymetric distribution of particular substrates (Topp and Hoff, 1972) within the broader geographic range. Association of most flatfishes with sediments rather than hard substrata indicates that the struc- ture of the sea bed is an important factor controlling their distribution (Gibson, 1994, and references therein). Topp and Hoff (1972) suggested that strong interrelationships between apparent substrate re- quirements of individual species of flatfish could pos- sibly explain patterns of geographical and bathymet- ric distributions observed for these fishes. Other stud- ies (Pearcy, 1978) cautioned that direct examination only of substrate types without examination of depth of occurrence may be incomplete because influences of depth and sediments on the distribution of benthic organisms are usually closely correlated and are dif- ficult to separate. Sediment texture generally de- creases with increasing depth of water, with small particles usually transported from regions of high energy waves and currents into deep, low-energy sedi- mentary environments, and coarse sediments, such as sands, generally are deposited in shallow water close to their continental source. According to Thorson (1957), physical and chemical compositions of sediments may be the main factor in determining the general patterns of distributions of infaunal and epifaunal invertebrates on the level sea floor. Fau- nal changes in both benthic invertebrates and verte- brates on the continental shelf and slope have also been thought to result from depth-related changes in physicochemical properties (Sanders and Hessler 1969; Haedrich et al., 1975). However, the emerging paradigm is that the relations between organism dis- tributions and the dynamic sedimentary and hydro- dynamic environment are complex (Snelgrove and Butman, 1994), especially when considering that grain size covaries with other factors including sedi- mentary organic matter content, pore-water chem- istry, and microbial abundance and composition, all of which are influenced by near-bed flow regime. Ecological distributional patterns observed for western Atlantic tonguefishes may result directly through an active process in which tonguefishes se- lect particular substrates based on physical charac- teristics of the substrate (i.e. particle size and com- position of the sediments, ease of burying, coloration of the sediments). Conversely, tonguefishes may in- directly occupy particular substrates because the suite of physicochemical characteristics (temperature regime, current strength, water depth, salinity, oxy- gen concentration, ambient light levels, etc.) required by that species may occur coincident with deposi- tional environments for particular sediment types. Reichert and Van der Veer (1991) provided descrip- tive information regarding substrate preference of settling S. plagiusa juveniles, but no experimental work on substrate selection or preference has been done for any species of western Atlantic tonguefish, and hypotheses concerning active selection of sub- strates by tonguefishes remain untested. Yet another factor potentially contributing to sub- strate selection, and indirectly to the ecological dis- tributional patterns observed for tonguefishes, may be related to substrate selection by biotic associates of tonguefishes. Diet studies (Mahadeva, 1956; Aus- tin and Austin, 1971; Topp and Hoff, 1972; Stickney, 1976; Kawakami, 1976; MacPherson, 1978; Toepfer and Fleeger, 1995) indicate strong preferences by symphurine tonguefishes for small epibenthic and infaunal invertebrates as food sources. Morphologi- cal characteristics of tonguefishes, such as the rela- tively small size of the mouth and reduced or absent dentition on ocular-side jaws in many shallow-water species, may also reflect adaptations (or constraints) for specialized feeding that would determine the size Munroe: Systematics of western Atlantic Symphurus 155 spectrum and variety of organisms that these fishes include in their diets (Stickney, 1976). Some inver- tebrates exhibit strong substrate preferences and are differentially distributed with respect to substrate types (Hedgpeth, 1953; 1954; Williams, 1958; Butman, 1987). Hence, substrate selectivity (either by passive or active means) exhibited by inverte- brates that are preferred food sources of tonguefishes would in turn be mirrored by tonguefishes through their selective foraging activities. This hypothesis is also untested, because, beyond descriptive studies for those few species where diets have been examined, vir- tually nothing is known concerning the relative degree (if any, see Toepfer and Fleeger, 1995) of dietary selec- tion or preference exercised by these flatfishes. Size-related life history information Adult western Atlantic symphurine tonguefishes span a size continuum from ca. 25 to 320 mm SL, encompassing the entire size range within the ge- nus. In these waters, both the largest (S. jenynsi, at- taining maximum sizes ca. 320 mm and maturing at sizes of ca. 120 mm or more) and smallest (S. arawak, maximum sizes ca. 50 mm and maturing at sizes as small as 25 mm) members of the genus occur. Most western Atlantic tonguefishes, however, are interme- diate in size, usually reaching maximum sizes smaller than 200 mm (usually between 80-180 mm) and maturing at sizes between 50 and 110 mm. Al- though no western Atlantic tonguefish attains sizes large enough to support directed commercial fisher- ies, several species have been used when taken in industrial fisheries (Roithmayr, 1965). Species of Symphurus are the smallest members of the Cynoglossidae. For example, adults of five spe- cies of Paraplagusia range in size from 184 to 334 mm SL (Chapleau and Renaud, 1993), whereas adult sizes of 49 species of Cynoglossus range between 99 and 530 mm SL (Menon, 1977). Despite superficial similarities in overall external morphology and general body plan, western Atlantic Symphurus display striking differences in body sizes and the relative sizes at which maturity is attained in individual species (Figs. 6-9). Although species of western Atlantic Symphurus more or less form a con- tinuum along a size gradient, marked differences in size-related features between species are apparent when overall maximum sizes (based on the largest specimen observed for the species) and minimum sizes at sexual maturity (based only on females) are compared (Figs. 6-9). Based on these parameters, species were assigned to one of four size categories. Dwarf tonguefishes are those with adult sizes rang- ing up to ca. 80 mm and attaining sexual maturity at sizes usually of 40 mm or less. Six western Atlan- tic tonguefishes categorized as dwarf species are (in order of increasing maximum size) 0. The yj loglinear model for this two-way table can be written as log mii=ii + ^+xyJ + ^t where p = Z -Z- log mtj / IJ; \x = Z; log mij/'J- p; = Z- log mi . 7 1 - p; and \xy = log mij - X* - x/ + p. This model perfectly describes any set of positive expected frequencies and is referred to as the satu- rated model. The right-hand side of this equation resembles the formula for the cell-means ANOVA. The parameters [Xx\ and {Ayv} are deviations about a mean and I, X-xy = Z, X,xy = Z, Xx = Z, Xf = 0. This model can also be described in the notation form as [XY]. A saturated loglinear model always expresses a given table of categorical data perfectly. This model has the maximum achievable log likelihood because it is the most general model, with as many para- meters as observations. However, it is possible that a simpler model may provide a fit as statistically good as that of the saturated model. How well this model fits is represented by the scaled deviance, a function of twice the difference in the log likelihoods of the saturated model and the simpler model. In addition to testing the fit of a model, one can use the deviance to diagnose lack of fit through residual analysis. For example, consider a three-dimensional satu- rated model with variables X, Y, and Z. For this model [ XYZ ], log = p + Xf + Xf + X/f + XtJxy + X,ff" + Xjkyz+ Xijkxyz. When Xl}}^yz - 0, there is no three-factor interaction, and the association between two vari- ables is identical at each level of the third variable and reduces to the loglinear model [XY XZ YZ ]. Fur- ther, if Xijkxyz - 0 and Xjkyz = 0, then for any given level of X, Y and Z are conditionally independent [AY XZ\. Similarly if XlJ^xyz = 0 , Xj^yz = 0 and X^2 - 0, then Z is jointly independent of X and Y [ XY Z]. Finally if hjkxyz = 0, Xjkyz = 0, Xikxz = 0, and XtJxy = 0, then X, Y and Z are mutually independent [XYZ]. With these criteria and beginning with a saturated model, we used a stepwise model selection procedure with de- viance in the form of the G2 test statistic to find a simpler model that fits as well as the saturated model. This simpler model would enable one to ex- plore multidimensional tables to find simpler repre- sentations of the information contained therein. Another advantage of loglinear models is that when one of the variables can be modeled as a response, and the others as explanatory variables, certain loglinear models are equivalent to logit models with categorical explanatory variables. Such logit models enable us to study the problem of interest in a man- ner analogous to ANOVA. Many categorical response variables have only two categories. The response can be classified either as a success or a failure. The Bernoulli distribution, which belongs to the natural exponential family, forms the basis of modeling the logit model. For such a dichoto- mous variable, the probability of observing response 0 can be defined as P(Y-0) - k, and the probability of observing response 1 , as P( Y= 1 ) = 1 - 7i. The link func- tion for this model, log ki / ( l-7t-), known as the logit, is equivalent to the log odds. Consider the following example, where we exam- ine the presence or absence of bycatch in two areas. In this example (Table 1), the 2x2 table has rows ix(area 1), and i2 ( area 2) and columns j1 ( presence of bycatch) and j2 (absence of by catch). The counts in the cells of the table are the number of units of effort (individual sets) observed in each category. In this case, the odds, Q.iv of observing^ ( presence of by catch) given you are in category i1 (area 1) is computed as the ratio of the conditional probabili- ties of observing a set with bycatch to that of observ- ing a set with no bycatch in area 1: {^m/^2ia} is 0.4/0.6 = 0.67. Similarly, the odds, Oj2, of observing j ^presence of by catch) given you are in category i2 ( area 2) is com- puted as the ratio of the conditional probabilities of 196 Fishery Bulletin 96(2), 1998 observing a set with bycatch to that of observing a set with no bycatch in area 2: is 0.33/0.67 - 0.5. The odds ratio, 0, is computed as Qfl/Qf2 is 0.67/0.5= 1.34. Thus, the odds of observing response j1 (presence of bycatch ) is 1.34 times more likely for row i, ( area 1) than for row i9 (area 2). An odds ratio of 1 indicates that you are equally likely to observe response j ^presence of bycatch) for row i1 (area 1) and row i2 (area 2) and thus indicates independence between the rows and columns of the table. The logit model has two forms. One form occurs where the explanatory variables are continuous and is the logistic regression model. The second occurs where the explanatory variables are categorical. The logistic regression model is analogous to a regres- sion model, whereas the second type is analogous to an ANOVA model. For the previous example, a model with a single categorical explanatory factor (area), the logit form of the model is \og(njVl/nj2U) = a + [3 frea, where a = the mean of the logits; and pArea _ deviation from the mean for row i. (5, describes the effects of the factor on the response. For this model the higher /J, becomes, the higher the logit in row i, and the higher the value of Kjm The constraints on this model are I /3, = 0. In this case the right-hand side of the equation resembles the cell- means model of a one-way ANOVA. This logit model would be equivalent to log ( myl) - log ( mlj2) = 2XjArea + 2XjiBycatch + 2 Xij1AreaBycatch in loglinear form. Bycatch sampling and data set description Bycatch from the gulf menhaden fishery was sampled April through October 1995 by two to three onboard samplers on a total of twenty-seven week-long trips aboard vessels operating from menhaden processing plants in the U.S. Gulf of Mexico. To maximize cover- age of the Gulf, samplers boarded vessels from ports in the western, central, and eastern regions in a given week as often as possible. During each sampling trip, all sets made by the vessel were alternatively sampled, either for releasable bycatch or automati- cally retained bycatch. For all sets sampled, the pres- ence of dolphins in the vicinity was also noted by the observers. In addition, the boat captains visually es- timated catch in standard menhaden (1,000 standard menhaden [-305 kg]) and recorded the latitude and longitude of a set location. The location was used to identify in which National Marine Fisheries Service (NMFS) statistical zone (Fig. 1) the set was made (after Kutkuhn, 1962). To collect releasable bycatch data, samplers ob- served the purse seine from the time it was brought alongside the carrier ship and throughout the pump- ing procedure, until the net was emptied and cleaned. During this time, the species, number, and fate of the releasable bycatch were recorded. The seven cat- egories of bycatch fate were as follows: gilled in the net (gilled); kept by the crew for consumption (kept); released with no apparent harm (released healthy ); released seriously injured or dead (released dead); released after being bruised or after being kept in the set for a long time (released disoriented); collected by the crew from the net or deck and put into the hold (caught and put in hold); and observed in the net but fate unknown (unknown). Statistical analysis Preliminary analysis For the variables bycatch num- ber, bycatch percentage, and estimated catch, we cal- culated a series of commonly used statistical descrip- tors, namely the mean, standard deviation, 95% con- fidence intervals, median, skewness, and kurtosis. In addition, we also calculated the winsorized mean and its standard deviation. We initially attempted to examine spatial and tem- poral patterns in the bycatch with a two-way ANOVA model. For the analysis, data were classified into sea- son (S) consisting of three groups: 1) spring (April through June), 2) summer (July through August), and 3) fall (September through October). Adjacent NMFS zones (Fig. 1) were combined to form four area (A) groups: 11-12, 13-14, 15-16, and 17-18. Bycatch patterns were examined with two response variables: 1) bycatch numbers; and 2) bycatch per- centage ([bycatch number/total catch] x 100). For each of these two response variables, spatial and tempo- ral patterns were examined by using the ANOVA model with season, area, and their interaction term as independent variables. Because we anticipated that neither model would satisfy the model assump- tions of normality of residuals and homogeneous vari- ances, we also examined the models by using the log and square-root transformations for both response variables. In addition we also used 2 arcsin V (by- catch/menhaden catch) suggested by Neter et al. (1990) for transformation of proportions. All seven de Silva and Condrey: Patterns in patchy data discerned from Brevoortia patronus bycatch 197 models were examined to determine if o o model assumptions were met. § S Spatial and temporal patterns in by- catch For our analysis using loglinear and logit models with categorical explana- tory variables, we used a four-way con- tingency table with a unit of effort (the set) as the count. Our main interest was 1) to examine the spatial and temporal patterns in bycatch and 2) to determine if the presence of dolphins in the vicinity when the set was made might be an indi- cator of bycatch patterns. Exploratory analysis with loglinear and logit models To examine bycatch as a response of interest with categorical mod- els, a new dichotomous categorical vari- able, bycatch, based on the median bycatch percentage, was created. Each set was clas- sified either as high bycatch if the bycatch rate of the set was greater than the me- dian value of all sets or as low bycatch if the bycatch rate of the set was less than or equal to the median bycatch of all the sets. We used the median rate because it is a robust measure of central tendency. In de- ciding on possible criteria for defining this variable, more extreme conditions, such as bycatch rates greater than the 75th percen- tile, were considered. However, by choosing more extreme values, we increased the number of sparse cells and thus affected the validity of the G2 test statistic. In analyzing contingency tables, it is necessary that the number of cell counts with zero frequencies be low (a minimum expected value of 1 is satisfactory as long as <20% of cells have counts of 5 or less) for the test statistic to be valid (Agresti, 1990). To reduce the number of cells with zero frequencies, months and zones were combined, generating two new variables, season and area, corresponding to those used in the AN OVA. The presence of dol- phins was used as a dichotomous variable, dolphins (D). To identify the most appropriate and simplest loglinear model for the data us- ing the variables season, area, bycatch, and dolphins, we employed a stepwise backward solution procedure commenc- ing with the saturated loglinear model (Agresti, 1990). Here the saturated model Figure 1 Map encompassing the extent of the U.S. gulf menhaden fishery from the Texas to Alabama coasts. The eight fishing zones numbered 11-18 are from NMFS (after Kutkuhn, 1962). 198 Fishery Bulletin 96(2), 1 998 is denoted as [SARD] were S stands for season, A for area, B for bycatch, and D for dolphins, and the model includes all possible interactions up to and includ- ing the four-way interaction. Because the saturated model would naturally provide the best fit, we were interested in determining if a simpler model could be found that would also satisfy the criteria of a logit model with bycatch as the response variable. The stan- dardized residuals of the resulting model were then examined to ensure that lack of fit was not a problem. Contrasts We anticipated that we would have sig- nificant interaction terms in our analyses that would require detailed examination of interactions. For the logit form of the selected model, we constructed a series of contrasts that might help to explain the nature of these potential interactions. The contrasts of interest had two general forms: 1 Given a specific area, are the odds of observing a set with high bycatch the same between any two different seasons. This results in three unique con- trasts for each area (spring vs. summer, spring vs. fall, summer vs. fall) and a total of 12 contrasts. Let Ftj be the logit of high bycatch for season i and area j and let Fh/ be the logit of high by catch for season h and area j. The hypotheses being tested were Ho '■ Fy - Fhj = 0, where Fy = a + (3;s + (3 A + (3(/SA; FJhj = a + (3/ + (3';' + p'/A; and h and i = spring, summer, and fall, such that h * i for each j, and j = area 11-12, area 13-14, area 15-16, and area 17-18. 2 Given a specific season, are the odds of observing a set with high bycatch the same between any two different areas. This results in six unique contrasts for a given season and a total of 18 contrasts (11— 12 vs. 13-14, 11-12 vs. 15-16, 11-12 vs. 17-18, ,15-16 vs. 17-18). Let F(/ be the logit of high bycatch for season i and area j and let Fjk be the logit of high bycatch for season i and area k. The hypotheses being tested were Ho'Fij ~ Fik ~ 0’ where F-- = a + (3 s + (3 A + [3 SA ; Fik= a + (3;s+(3/ + (3;/A;and j and k = area 11-12, area 13-14, area 15-16, and area 17-18, such that j * k for each i, where i = spring, summer, and fall. Because 30 contrasts were performed, the type-I error level of 0.10 was adjusted by using the Bonferroni technique, and only P-values less than 0.0033 were considered significant. The estimated odds ratios for the conditions associated with the hypotheses were calculated from the parameter es- timates given by the analysis. Bycatch species associations To examine the as- sociation between species and fates of the releasable bycatch, we used correspondence analysis on a spe- cies-by-fate table for all seasons and areas combined. Area and species associations of the releasable by- catch were also examined for each of the three seasons with correspondence analysis on species-by-area tables. For all correspondence analyses we defined two groups of species. The first group, consisting of those species that were common in terms of number and oc- currence, was used in the main table. Releasable bycatch species falling into this group had a minimum of 230 individuals and were found in at least 30% of the sets. The second group of species consisted of re- leasable bycatch that were less common; these were species for which a minimum of 30 individuals were observed, which occurred in at least 4% of the sets, and which did not meet our criteria for well represented species. These species were included as supplementary variables in our analysis (Greenacre, 1984). Supplemen- tary variables are represented as points in the joint row and column space but are not used in determining the locations of the active rows and columns of the table. Species included in the main and supplementary table accounted for 97% of the total number of organisms observed during the study period. Species that did not meet these criteria were not used in the analyses. Results Preliminary analysis A total of 15,579 bycatch organisms representing 62 species or species groups were observed as releas- able bycatch in 257 sets. The estimated catch of stan- dard menhaden per set ranged from 5,000 to 500,000 with a median, mean, and standard deviation of 50,000, 67,000, and 61,000 respectively. Skewness and kurtosis values of 2.5 and 10.7 indicated that the distribution of the estimated menhaden catch was positively skewed (Fig. 2). de Silva and Condrey: Patterns in patchy data discerned from Brevoortia patronus bycatch 199 Menhaden catch (std menhaden) Figure 2 Distribution of menhaden catch sampled during the 1995 gulf menhaden fishing season, std = standard (1,000 standard menhaden ~ 305 kg). The number of by catch observed in each set ranged from 0 to 1,600 organisms, with a median, mean, and standard deviation of 15, 61, and 153 respectively. The winsorized mean and its standard deviation val- ues were 53 and 6.9 respectively. The 95% confidence interval of the mean was between 41.8 and 79.3. The distribution of bycatch organisms was strongly posi- tively skewed (5.9) and peaked sharply with a Kur- tosis value of 47.2 (Fig. 3). The bycatch percentage ranged from 0% to 4% with a median, mean, and standard deviation of 0.033%, 0.168%, and 0.48 respectively. The winsorized mean bycatch percentage and its standard deviation were 0.14% and 0.02, respectively. The 95% confidence in- terval of the mean was between 0.11% and 0.22%. The distribution of the bycatch percentage was also found to be positively skewed and strongly peaked, with skewness and kurtosis values of 5.4 and 32.7 respectively (Fig. 4). Analysis of variance with bycatch, bycatch percent- age, and their respective transformations, together with the arcsine transformation, did not meet model assumptions. In all cases the modified Levene’s test indicated that the variances were nonhomogeneous, and both the residual plots and Shapiro-Wilk test indicated that the assumption of normality of residu- als were not met. For example, for the response log (bycatch percentage +1), the residuals of the model were not normally distributed (Shapiro-Wilk W=0.678, PF=0.0001). These characteristics suggest that the model assumptions were grossly violated and that ANOVA may not be an appropriate form of analysis in this case. Spatial and temporal patterns in bycatch Exploratory analysis using Sogls near models With the backward selection procedure, loglinear models [SAB SAD DB] and [SAB SAD] (as defined in Table 2 ) satisfied the criteria for a logit model and had good fit (Table 2). The simplest of these models, [SAB SAD], was selected; this loglinear model corresponds to the logit model with categorical explanatory variables of the form l0g^*^ = a + AA+As+Af, U) ^ low\ik 200 Fishery Bulletin 96(2), 1998 0.033%) for the sea- son-area combinations. Table 4 Estimated odds ratios and Wald chi-square (%2) values for contrasts (in parentheses) for observing high bycatch between areas, given a set was made in a certain season. For example, the odds of observing high bycatch are 1.628 times greater in area 11-12 than in area 13-14. * indicates a ratio significantly different from 1. Wald y2 values are in parenthesis. For all contrasts, y2 df=l. Season Area Area 11-12 13-14 15-16 Spring 13-14 1.628 (0.33) 15-16 3.877 (6.64) 2.380 (1.23) 17-18 10.857* (9.29) 6.666(3.79) 2.799 (2.13) Summer 13-14 0.791 (0.04) 15-16 1.407 (0.42) 1.777 (0.23) 17-18 2.671 (2.33) 3.374(0.94) 1.898 (1.22) Fall 13-14 0.615 (0.27) 15-16 0.230 (3.63) 0.375 (1.50) 17-18 0.288 (2.43) 0.468 (0.84) 1.250 (0.12) cant differences between these two sea- sons existed for area 11-12 (Wald %2=10.03, df=l,P>%2=0.0015). When the summer and fall seasons were con- trasted, a significant difference between seasons was observed only for area 11— 12 (Wald %2=9.25, df=l, P>%2=0.0024). For contrasts of the odds of high bycatch between areas for a given sea- son, significant differences between ar- eas 11-12 and 17—18 in the spring sea- son were observed (Wald %2=9.29, df=l, P>^2=0.0023). All other contrasts be- tween areas for the spring, summer, and fall seasons were not significant. Wald values for these contrasts are pre- sented in Table 4. The estimated odds ratios for the con- ditions associated with the hypotheses are also given in Tables 3 and 4. Only estimated odds ratios of contrasts for the rejected hypotheses of no difference are examined in detail. If a fishing boat was in area 11-12, the odds of observing a set with high bycatch are about 11.8 times higher if a boat was fishing in the spring rather than the fall. Also, if a fishing boat was in area 11-12, the odds of observing a set with high bycatch are about 0.1 times (or approximately 10.3 times lower) in the fall than in the summer. It appears that for area 11-12, the odds of observing a set with high bycatch in the fall are significantly lower than in the spring or summer. The third significant contrast indi- cated that for a vessel fishing in the spring, the odds of observing a set with high bycatch are about 10.8 times higher in area 11-12 than in area 17-18. Refining the final model We were in- terested, on the basis of our contrasts, in determining if a model with simpler dichotomous classes for areas and sea- sons would provide as good a statistical fit as this “full model” (Eq. 1). We compared three potential models that had one or both of these variables with reduced classes against the full model (Table 5). For these models we classified area into two groups: 1 ) east of the Mississippi River; and 2) west of the Mississippi River. Season was also classified into two groups; 1) early — sets sampled April through August; and 2) late — sets sampled September through October. All three reduced models provided a fit as good as that of our full model (Table 5); therefore we chose the model that had reduced classes for season and area because it was the simplest. Four contrasts of interest were examined: 1 Test for seasonal differences in the odds of ob- serving a set with high bycatch given the set was sampled east of the Mississippi River; de Silva and Condrey: Patterns in patchy data discerned from Brevoortia patronus bycatch 203 Table 5 Comparison of reduced logit models with the full model. Diff full model and reduced model. = difference between Likelihood ratio statistics Model comparisons df G2 P>G 2 Diff (df) Diff (G2) P- Value Full model Season= Apr-Jun, Jul-Aug, Sep-Oct Area = 11-12, 13-14, 15-16, 17-18 12 17.82 0.1213 Reduced model I Season=spring, summer, fall Area = 11-12, 13-18 6 9.76 0.1352 6 8.06 0.237 Reduced model II Season=Apr-Aug, Sep-Oct Area = 11-12, 13-14, 15-16, 17-18 8 9.99 0.2659 4 7.83 0.098 Reduced model III Season= Apr-Aug, Sep-Oct Area = 11-12, 13-18 4 7.54 0.1100 8 10.28 0.246 2 Test for seasonal differences in the odds of observing a set with high bycatch given the set was sampled west of the Mississippi River; 3 Test for area differences in the odds of observing a set with high bycatch given the set was sampled in the early season; and 4 Test for area differences in the odds of observing a set with high bycatch given the set was sampled in the late season. The contrasts were written similarly to those for the “full model.” For the four contrasts, we used an a level of 0.025 as significant. Of the four contrasts, two were significant. The first indi- cated that in areas east of the river, the odds of observing a set with high bycatch was significantly different between sets sampled in the early season and sets sampled in the late season (Wald %2=11.41, df=l, P>X2- 0.0007 ). The odds of observing a set with high bycatch east of the river in the early season was 11 times greater than the odds of observing a set with high bycatch in the late season. The second significant contrast indicated that for sets sampled in the early season, the odds of observing a set with high bycatch east of the river was significantly different from the odds of observing a set with high bycatch west of the river (Wald x2=7.99, df=l, P>y2=0.QQ47). In the early season, the odds of observing a set with high by catch east of the river was 2.7 times greater than observ- ing a set with high bycatch west of the river during the same period. Bycatch species associations Of the 62 species groups observed, 20 occurred in two or fewer sets. The most frequently occurring species were Atlantic cutlassfish, Trichiurus lepturus (44% of sets), Atlantic croaker, Micropogonias undulatus (38% of sets), Spanish mackerel, Scomberomorus maculatus (36% of sets), sand seatrout, Cynoscion arenarius (35% of sets), and gafftopsail catfish, Bagre marinus (34% of sets). In terms of total abundance (Table 6), Atlantic croaker, sand seatrout, and Atlan- tic bumper, Chloroscombrus chrysurus, accounted for 71% of the total releasable bycatch. Species included in the main table were Atlantic croaker, sand seatrout, crevalle jack, Caranx hippos, gafftopsail catfish, Spanish mackerel, and Atlantic cutlassfish (Table 6 ). Species included as supplemen- tary variables were striped mullet Mugil cephalus, unidentified requiem sharks, gulf butterfish, Peprilus burti, cownose ray, Rhinoptera bonasus, spotted seatrout, Cynoscion nebulosus, Atlantic bumper, blacktip shark, Carcharhinus limbatus, red drum, Sciaenops ocellatus, unidentified penaeid shrimp, hardhead catfish, Arius felis, brown shrimp, Panaeus aztecus, cabbage head jellyfish, Stomolophus meleagris , bull shark, Carcharhinus leucas, and uni- dentified tonguefish (Soleidae) (Table 6). The fate of releasable bycatch Correspondence analy- sis on the fate-by-species table for the entire fishing season indicated that the first two axes explained 97% of the total inertia (conceptually similar to variance) and offered a good representation of the fate-species associations. From the two-dimensional plot (Fig. 6) we discerned three major and one minor groupings. Species primarily associated with being released dead or disoriented were unidentified requiem sharks, red drum, crevalle jack, and bull sharks. Spe- cies secondarily associated with being released dead or disoriented were cownose rays and blacktip sharks. These last two species were primarily asso- ciated with being released healthy and appeared to form their own minor grouping. 204 Fishery Bulletin 96(2), 1 998 Table 6 Species used in correspondence analyses. (M) signifies species used in main table and (S) in supplementary table (see “Materials and methods” section. Areas ( 11-12, 13-14, etc. are shown under each season. Counts are number of organisms observed. Unid. = unidentified. Spring Summer Fall 11-12 13-14 15-16 17-18 11-12 13-14 15-16 17-18 11-12 13-14 15-16 17-18 Total Atlantic croaker (M) 132 20 50 1 554 1,604 1,732 212 40 97 51 612 5,105 sand seatrout (M) 104 26 17 0 813 1,500 572 2 91 91 197 53 3,466 gafftopsail catfish (M) 255 70 72 24 161 0 96 48 8 28 33 1 796 Atlantic cutlassfish (M) 41 28 209 22 5 1 53 55 7 0 13 36 470 crevalle jack (M) 71 12 133 17 5 0 31 9 22 41 8 0 349 Spanish mackerel (M) 22 11 34 33 29 0 56 33 0 7 6 10 241 Atlantic bumper (S) 0 0 0 0 0 0 2,166 330 0 0 1 0 2,497 striped mullet (S) 344 31 0 0 511 0 0 0 0 9 0 0 895 red drum (S) 21 0 34 0 23 0 24 6 0 12 9 116 245 hardhead catfish (S) 36 5 3 0 100 3 38 3 12 1 3 2 206 tonguefish spp. (S) 0 0 2 0 0 0 1 0 0 0 200 1 204 blacktip shark (S) 37 0 54 1 20 0 19 52 0 0 0 1 184 cownose ray (S) 27 2 26 0 3 0 4 7 0 0 1 0 70 brown shrimp (S) 0 0 1 2 0 0 39 0 0 0 10 13 65 cabbbagehead jellyfish (S) 0 0 5 12 0 0 33 0 1 0 4 6 61 unid. requiem sharks (S) 0 0 23 2 3 0 16 0 0 6 1 6 57 spotted seatrout (S) 2 8 0 0 31 0 0 0 0 0 0 0 41 gulf butterfish (S) 2 0 3 0 0 0 6 3 0 0 0 26 40 bull shark (S) 5 0 0 0 31 0 2 0 0 0 0 1 39 unid. penaeid shrimp (S) 0 0 1 9 5 0 0 0 1 0 0 16 32 Others (not used in CA) 80 5 86 11 147 5 56 24 13 8 72 9 516 Column Total 1,179 218 753 134 2,441 3,113 4,944 784 195 300 609 909 15,579 The second group, species primarily associated with being gilled, were Atlantic croaker, sand sea- trout, and unidentified tonguefish. Other species that were associated with being gilled were unidentified penaeid shrimp, Atlantic cutlassfish, gulf butterfish, and Atlantic bumper. These four species were also associated with the third group, species kept by the crew, and those whose fate was unknown. Other spe- cies associated with this third group were hardhead catfish, brown shrimp, Spanish mackerel, gafftopsail catfish, striped mullet, and the cabbage head jellyfish. Temporal and spats'a! patterns of bycatch species Spring Correspondence analysis of area by species for spring indicated that the first two axes explained 97% of the inertia and offered a good representation of species-area associations. From the two-dimen- sional plot (Fig. 7), we discerned three major group- ings. The first axis separates the eastern areas of the fishery (zone groups 11-12 and 13-14) from the western areas (zone groups 15-16 and 17-18). The second axis also separates zone group 15-16 from zone group 17-18. The eastern areas are associated with Atlantic croaker, gafftopsail catfish, sand seatrout, hardhead catfish, bull shark, striped mullet, and spotted seatrout. Zone group 15-16 is primarily associated with unidentified tonguefish, red drum, crevalle jacks, blacktip shark, unidentified requiem sharks, gulf butterfish, and Atlantic cutlassfish. Zone group 17- 18 was associated with Spanish mackerel, brown shrimp, cabbage head jellyfish, and unidentified shrimp. Summer For the summer, correspondence analysis indicated that the two axes accounted for 94% of to- tal inertia. As in spring, three major species area groupings were observed (Fig. 8). Notable differences in these grouping were that zone group 15-16 ap- peared to be closer to the eastern groups ( 13-14 and 11-12). Furthermore, group 13-14 was separated fur- ther from zone group 11-12. de Silva and Condrey: Patterns in patchy data discerned from Brevoortia patronus bycatch 205 5 I 4 - 3 - 3 J o .? £ U- 3 « <2 Jh d d o bo 3 d q; 2 -2 J3 H 44) ^ d x, .2 05 jh 2 w *3 “ .c & CO C o d Eh . pH d ‘S £ w g *H rd (1) CO ground in the northeast Indian Ocean; and Japanese longine vessels operating in the South- ern Ocean (on feeding grounds off Tasmania, New Zealand, and South Africa and on staging grounds in the southeast Indian Ocean) (Fig 1). The ovaries of 475 females were collected from the com- plete size range of southern bluefin tuna caught on the spawning ground. Ovaries were removed at sea, labelled, stored on ice, and matched with the corresponding tuna carcass at Benoa in the Sunda Islands. Dressed weight (where fish was gilled, gutted, its fins removed, and its tail stock left intact) was measured to the nearest kg, and fork lengths of most fish to the nearest cm. A subsample was removed from 200 ovaries and fixed in 10% buffered formalin. The ovaries were frozen and flown by airfreight, along with the subsample, to Australia. In addi- tion, up to 30% of all longline catches landed at Benoa were monitored and the individual weights and lengths of southern bluefin tuna were recorded. Ovaries from the Southern Ocean were collected at sea from the fishing grounds around Tas- mania, New Zealand, South Af- rica, and the southeast Indian Ocean. The areas around Tas- mania, New Zealand, and South Africa are considered to be feed- ing grounds for immature and adult southern bluefin tuna, whereas the southeast Indian Ocean is the area where pre- and postspawning fish are caught during the spawning season. In total, ovaries were collected from 2,340 females from the feeding grounds in the Southern Ocean and from 393 females from the southeast Indian Ocean. Ovaries were frozen immediately after collection. Fork lengths were measured to the nearest cm. Farley and Davis: Reproductive dynamics of Thunnus maccoyii 225 In the laboratory, a core subsample was taken from each ovary while frozen. Subsamples were fixed in 10% buffered formalin, embedded in paraffin, and standard sections were prepared for histological ex- amination (cut to 6 pm and stained with Harris’s haematoxylin and eosin). All ovaries collected on the spawning grounds and 53 of the ovaries collected from the Southern Ocean were processed in this way. Ovaries were thawed, trimmed of fat, blotted dry, and weighed to the nearest g. The mean diameter (random axis to the nearest pm) of five oocytes from the most advanced group of oocytes (MAGO) was determined for each ovary with a video coordinate digitizer connected to an Ikegami video camera mounted on a stereomicroscope at 50x magnification. Gonad index was calculated as GI = W/L 3 x IQ4, where W = gonad weight in g; and L = length to caudal fork in cm (Kikawa, 1964a; Shingu, 1970). All ovaries were examined for residual hydrated oo- cytes as evidence of spawning activity. Histological classification and spawning frequency Because many of the ovaries were frozen before a subsample could be removed, it was necessary to as- sess whether histological sections prepared from fro- zen ovarian material could be used to determine go- nad stage and spawning activity. A comparison was made between histological sections prepared from 200 ovaries which were subsampled both before and after freezing. Although considerable cell destruction was observed in the histological sections from frozen tissue, oocytes, atretic oocytes, and postovulatory fol- licles were still distinguishable and could be classified. Histological sections were classified with criteria similar to those developed for northern anchovy, Engraulis mordax (Hunter and Goldberg, 1980; Hunter and Macewicz, 1980, 1985a, 1985b), skipjack tuna, Katsuwonus pelamis (Hunter et al., 1986) and yellowfin tuna, Thunnus albacares (Schaefer, 1996). Each ovary was staged by the most advanced group of oocytes present into one of five classes: unyolked, early yolked, advanced yolked (Fig. 2A), migratory nucleus (Fig. 2B), or hydrated (Fig. 2C). In order to determine the relation between atresia (resorption of oocytes) and spawning, ovaries were also classi- fied by the level of a and (3 stage of atresia in ad- vanced yolked oocytes. During the a stage of atre- sia, yolk resorption takes place (Fig. 2D). During the P stage of atresia, the remaining granulosa and the- cal cells are reorganized and resorbed leaving a com- pact structure containing several intercellular vacu- oles. Ovaries were classified into one of the following five atretic states: 0 no a atresia present, but advanced yolked oocytes are; 1 <10% of advanced yolked oocytes are in the a stage of atresia; 2 10-50% of advanced yolked oocytes are in the a stage of atresia; 3 >50% of advanced yolked oocytes are in the a stage of atresia; 4 100% of advanced yolked oocytes are in the a stage of atresia, or no advanced yolked oocytes are present but oocytes in the (3 stage of atresia are present. Spawning frequency was determined by the post- ovulatory follicle method of Hunter and Macewicz ( 1985a). This method uses the incidence of females with postovulatory follicles less than 24 hours old to define the fraction of the population spawning. Be- cause the time of capture was not available, we could not assign ages to postovulatory follicles based on the estimated time of death relative to the estimated time of spawning. Postovulatory follicles were, there- fore, aged according to their state of degeneration with criteria developed for skipjack tuna, yellowfin tuna, and bigeye tuna, Thunnus obesus (Hunter et al., 1986; McPherson, 1988; Nikaido et al., 1991; Schaefer, 1996), all of which spawn in water tempera- tures above 24°C and resorb their postovulatory fol- licles within 24 hours of spawning. We assumed that southern bluefin tuna resorb postovulatory follicles at the same rate as other tropical spawning tuna because water temperature appears to be the domi- nant factor governing resorption rates ( Fitzhugh and Hettler, 1995). We recorded postovulatory follicles in histological sections according to methods of Hunter and Macewicz (1985a), Hunter et al. (1986) and Schaefer ( 1996). Postovulatory follicles were classified as either 0 (absent), 1 (new) (Fig. 2E), 2 (less than 12 hours old), 3 (12 to 24 hours old), or 4 (indistinguish- able owing to tissue decay). The incidence of females with postovulatory follicles of any age was used to de- termine spawning frequency. Females were classified into one of four spawning states depending on the oocytes, atretic state, and postovulatory follicle class present. 1 Immature: Ovaries contain no advanced yolked oocytes or advanced yolked oocytes in the a stage of atresia. No residual hydrated oocytes present. 226 Fishery Bulletin 96(2), 1998 2 Spawning: Ovary contains advanced yolked oo- cytes and evidence of spawning activity (migra- tory nucleus or hydrated oocytes or postovulatory follicles). Less than 100% of advanced yolked oo- cytes are in the a stage of atresia. If >50% of ad- vanced yolked oocytes are atretic, early yolked oocytes are considered nonatretic. Figure 2 Photomicrographs of southern bluefin tuna, Thunnus maccoyii, ovarian tissue collected from the spawning ground (northeast Indian Ocean) between October 1992 and June 1995. (A-E) Transverse sections stained with Harris’s haematoxylin and eosin. (A) advanced yolked-stage oocyte; (B) migratory-nucleus-stage oocyte; (C) oocyte in the early stages of hydration with some yolk plates still visible; (D) fully yolked oocyte in a stage of a atresia; (E) new postovulatory follicle; (F) hydrated-stage whole oocyte. Bar = 0.1 mm. Farley and Davis: Reproductive dynamics of Thunnus maccoyii 227 3 Nonspawning (mature): Ovary contains advanced yolked oocytes but no evidence of spawning ac- tivity (migratory nucleus or hydrated oocytes or postovulatory follicles). Less than 100% of ad- vanced yolked oocytes are in the a stage of atre- sia. If >50% of advanced yolked oocytes are atretic, early yolked oocytes are considered nonatretic. 4 Postspawning: Ovaries contain either: 1) >50% of both early and advanced yolked oocytes in the a stage of atresia; 2) 100% of advanced yolked oocytes in the a stage of atresia; or 3) no yolked oocytes are present but oocytes in the P stage of atresia are, and residual hydrated oocytes may or may not be present. Fecundity To find out if the annual fecundity of southern blue- fin tuna could be determined before spawning be- gan, we measured the distribution of oocyte sizes within ovaries of four females at various stages of maturity. Ovarian subsamples containing at least 1,000 oocytes were removed from each ovary, teased apart, and each oocyte (greater than 100 pm in di- ameter) was measured in a random orientation to the nearest pm under a stereomicroscope. We estimated batch fecundity (the number of hy- drated oocytes released per spawning) by the gravi- metric method (Hunter et al., 1985) for 21 females with unovulated hydrated oocytes (Fig. 2F). A split- plot analysis of variance (ANOVA) was used to de- termine the appropriate locations to subsample the ovary. The data were structured with 6 fish exam- ined as blocks, ovarian lobe (left or right) as main plots and twelve subsamples as subplot effects. Six subsamples were taken from each ovarian lobe in the anterior, middle, and poste- rior regions of both lateral sides (Table 1). A significant differ- ence between ovarian lobes was found in the number of hy- drated oocytes per gram of to- tal ovary weight and in the lobe x lateral side interaction. Con- sequently, a subsample of less than 1 g was taken from both sides of each ovarian lobe. Each subsample, consisting of a core from the periphery to the lu- men, was weighed to the near- est 0.01 mg and fixed in 10% buffered formalin. Each sub- sample was teased apart and washed through two sieves, similar to those of Lowerre-Barbieri and Barbieri ( 1993) to separate out the hydrated oocytes, which were counted under a stereomicroscope. The number of hydrated oocytes per gram of ovary was raised to the weight of both ovaries to give an estimate of batch fecundity for each of the four subsamples. Results Ovary maturation Ovaries obtained from fish on the spawning ground (northeast Indian Ocean), the staging ground (south- east Indian Ocean), and the feeding grounds (South- ern Ocean) show clear differences in development based on ovary weight (Fig. 3). Females less than 140 cm showed no or minor ovary development; there- fore it appears that they would not spawn in the com- ing season. The majority of ovaries collected from the feeding grounds weighed less than 1 kg and had a gonad index (GI) of <3.2, whereas ovaries from the southeast Indian Ocean weighed up to 2.8 kg and had GI values up to 4.9. Some of the ovaries of fe- males caught between August and December on the feeding grounds showed signs of maturity, because their MAGO’s were in advanced yolked stage with diameters greater than 400 pm. Females from the spawning ground had larger ovaries, weighing up to 7.4 kg. All ovaries from fish collected on the spawn- ing ground had a MAGO diameter greater than 400 pm, except three ovaries with MAGO’s between Table 1 Split-plot ANOVA of the effect of location of tissue sample from southern bluefin tuna, Thunnus maccoyii, ovaries on the number of hydrated oocytes per gram of ovary. Source df Sum of squares Mean square F-value P-value Blocks Fish 5 17,201,897 3,440,380 279.721 <0.001 Main plots Lobe 1 168,326 168,326 13.686 0.014 Main plot error 5 61,497 12,299 0.764 0.580 Subplot effect Lateral side 1 711 711 0.044 0.834 Region 2 34209 17,104 1.062 0.353 Lateral side x region 2 82,599 41,300 2.565 0.087 Lobe x lateral side 1 152,159 152,159 9.450 0.003 Lobe x region 2 12,944 6,472 0.402 0.671 Lobe x region x lateral side Subplot error 2 50 58,207 805,044 29,103 16,101 1.808 0.175 228 Fishery Bulletin 96(2), I 998 153 and 207 |im. These three female had recently finished spawning because their ovaries contained either 100% of advanced yolked oocytes in an a atretic state or residual hydrated oocytes but no healthy advanced yolked oocytes. Spawning Season Southern bluefin tuna were caught on the spawning ground during every month, except July. Catch per unit of effort (CPUE), expressed as the number of fish caught per vessel fishing day, showed that abun- dance was low from May to August (Fig. 4). In the 1993-94 season, abundance peaked in February, and in the following season, abundance peaked in Octo- ber and February. Females with high GI values were caught on the spawning ground during all months that ovaries were collected. Gonad index values were variable, however, ranging from 1.3 to 13.1. There were no peaks or trends during this time. These observations suggest that spawning activity is constant throughout the season. Histological classification All ovaries examined histologically from the South- ern Ocean were classified as immature, nonspawning (mature), or postspawning. Females greater than 140 cm, with advanced yolked oocytes, collected between August and December were probably prespawners preparing to migrate to the spawning ground. Sur- prisingly, 78% of these prespawning females had a atresia of advanced yolked oocytes. The level of atre- sia present was related to ovary weight or GI (Fig. 5). The ovaries of all females with GI’s<2 had no or <10% a atresia (atretic states 0 or 1), whereas 80% of fe- males with GI’s>2 had 10 to 50% a atresia (atretic state 2). This finding suggests that females begin resorbing yolked oocytes as their ovaries mature. Atresia of unyolked or early yolked oocytes was not observed in these females. Postspawning females were found in the Southern Ocean in all months that ovaries were collected, including both the early and late months of the spawning season (September to December). Ovaries, however, were not collected be- tween January and March from the Southern Ocean. The ovaries of all females sampled on the spawn- ing ground were classified as mature because they contained either advanced yolked oocytes or oocytes in either an a or P stage of atresia. Of the females, 69.2% were classed as spawning, 30.2% as non- spawning, and the remaining 0.6% as postspawning on the basis of oocyte stage, atretic stage, and postovulatory follicle class present. The mean GI 8000 n 7000 6000- 5000- | 4000 6 3000- 2000 1000- o Spawning ground X SE Indian Ocean • Tas./N.Z./S.Afr. o-1 80 100 120 140 160 180 200 220 Length (cm) Figure 3 The relation between ovary weight and body length for southern bluefin tuna, Thunnus maccoyii, caught on the spawning ground in the northeast Indian Ocean, on the staging grounds in the southeast Indian Ocean, and on feeding grounds off Tasmania, New Zealand, and South Africa. The line represents a gonad index of 2. (±95% Cl) of spawning and nonspawning females was similar at 5.4 (0.371 and 5.6 (0.29) but GI was not calculated for postspawning females because the sample size was too small (n=3). Spawning and nonspawning females were found on the spawning ground throughout the spawning season. All postspawning females were collected in October. The majority of females identified as spawning (86.1%) had ovaries containing no or less than 10% a atre- sia. The remaining 13.9% of females identified as spawning contained more than 10% a atresia. Nearly 90% of nonspawning females, however, contained 10- 50% of advanced yolked oocytes in an a atretic state. We have used this criterion of <10% a atresia to group fish we considered to be at the height of spawning, which we refer to as “prime spawning condition.” There did not appear to be any seasonal trends in the proportions of each atretic state throughout the spawning season (Fig. 6), suggesting that spawning in southern bluefin tuna is not synchronized. This is confirmed by an absence of an increase in the inci- dence of postspawning females at the end of the spawning season. Farley and Davis. Reproductive dynamics of Thunnus maccoyit 229 0.6 0.5 0.4 0.3 0.2 0.1 0.0- 1992-93 nd nd nd 1 1 Figure 4 CPUE of southern bluefin tuna, Thunnus maccoyii, (number caught per vessel per fishing day) by month for three consecutive spawning seasons. From December 1993, additional companies with higher catch rates were monitored causing an increase in CPUE from that date. Spawning frequency The ovaries of 68.7% of females had evidence of re- cent or imminent spawning activity. The absence of a peak in the percentage of females spawning dur- ing the season (Fig. 7) suggests that spawning in- tensity was constant. The ovaries of 120 females gave evidence of two spawning events, that is, they con- tained maturing oocytes (either migratory nucleus or hydrated) and postovulatory follicles. The major- ity of ovaries contained progressively more developed oocytes with progressively older postovulatory fol- licles. If postovulatory follicles are resorbed within 24 hours of spawning, then southern bluefin tuna are capable of spawning daily. The fraction of females that spawned per day, mea- sured by the fraction of females with postovulatory fol- licles, was 0.62 (Table 2). This gave a weighted mean spawning interval of about 1.6 days. If we examine only those females in “prime spawning condition” (<10% a atresia), then the spawning frac- tion was 0.90 giving a weighted mean inter- val of 1. 1 days. This value suggests that once females start spawning, they spawn daily. Batch fecundity The ovaries of nonspawning females col- lected from the southeast Indian Ocean contained a large number of unyolked oo- cytes less than 150 pm in diameter (Fig. 8). This number was reduced as females matured during the season and began spawning. Oocytes of all stages were present in the ovary of spawning females; therefore it appears that southern bluefin tuna have an asynchronous pattern of oo- cyte development. Unlike species with de- terminate annual fecundity, a gap in the oocyte size-frequency distribution did not appear between unyolked (<200 pm dia- meter) and early yolked (200 to 400 pm diameter) oocytes (Fig. 8). This finding in- dicates that advanced yolked oocytes are continually maturing during the spawn- ing season, from the pool of unyolked oo- cytes, and are spawned. The ovaries of 86% of females identified as spawning contained no or <10% a atresia, indicating that the majority of advanced yolked oocytes that are matured are spawned. Residual hydrated oocytes were found in the ovaries of only 5.6% of females collected from the spawning ground, which suggests that all hydrated oocytes are released at each spawning. The mean relative batch fecundity and 95% Cl’s estimated for 20 females of known weight was 56.5 (±16.1) oocytes per gram of body weight. The rela- tion between length (cm) and batch fecundity (BF) was best described by the equation BF = (4.78242 x 10“17) x L7-530 (F= 23.9, df=l,18, P<0.001) (Fig. 9). However, the relation was highly variable with only 57% of the variance explained by the regression. Discussion Spawning ground and season All female southern bluefin tuna sampled on the spawning ground were mature, and the smallest was 230 Fishery Bulletin 96(2), 1998 147 cm long. Davis2 showed that the mean length at which 50% of southern bluefin tuna were mature was around 152 cm based on oocyte diameters>400 pm, and 162 cm based on GFs>2. Warashina and Hisada ( 1970) considered that southern bluefin tuna caught on the “Oka” grounds (Fig. 1) in the 1960’s reached maturity at 130 cm, although reanalysis of these data 2 Davis, T. L. O. 1995. Size at first maturity of southern blue- fin tuna. Council for the conservation of southern bluefin tuna scientific meeting; 10-19 July 1995, Shimizu, Japan. Far Seas Fisheries Res. Lab., Shimizu, Japan, Rep. CCSBT/95/9, 9 p. 3000- a no atresia o <10% atresia • 1 0-50% atresia — Gl=2 ■g 2000- 5 £■ ns 1000- 0J 130 150 170 190 210 Length (cm) Figure 5 Levels of atresia by length of southern bluefin tuna, Thunnus maccoyii, caught off the spawning ground in the northeast In- dian Ocean. showed that this was the smallest size at which they matured and that 50% maturity was not reached until 146 cm (Anonymous3). Size at maturity appears to have increased progres- sively since then, being 154 cm in the period 1985-89. The increase in length at maturity can be attributed, in part, to an increase in the growth rate of southern bluefin tuna between the 1960’s and 1980’s reported by Hearn,4 be- cause maturity appears to be determined by age rather than length. The abundance of southern bluefin tuna on the spawning ground was not constant through- out the spawning season. Catch-per-unit-of-ef- fort data indicated that a peak in catches oc- curred in October for the 1993-94 season and in October and February for the 1994-95 sea- son. Japanese CPUE data also indicate there were two peaks in abundance on the spawning ground in the early years of the fishery; the first in September and October and the second in February and March (Davis and Farley5). The reason for two peaks is unknown but could be linked to the widespread distribution of south- ern bluefin tuna along the feeding grounds of the West Wind Drift. It is possible that the cues for migration to the spawning grounds take place at different times depending on where the fish are prior to spawning. Fish may reach spawning condition earlier in some areas than in others. Also the time needed to migrate to the spawning grounds would differ between areas. Spawning duration o tr o Cl O CL 1 993 1 994 1 995 Atretic stages H=4 a =3 0=2 □ =1 □ =0 Figure 6 The proportion of each atretic stage by month recorded in southern bluefin tuna, Thunnus maccoyii , caught on the spawning ground. This study extends the known du- ration of the spawning season of southern bluefin tuna. Females were caught on the spawning ground in 3 Anonymous. 1994b. Report of the south- ern bluefin tuna trilateral workshop, 17 January-4 February 1994, CSIRO, Hobart, Australia, 172 p. 4 Hearn, W. S. 1994. Models for estimat- ing SBT age at length during the transi- tion period. 13th SBT trilaterial scientific meeting; 19-29 April 1994, Wellington, New Zealand. Ministry of Agriculture and Fisheries, Wellington, New Zealand, Rep. SBFWS/94913, 16 p. 5 Davis, T. L. O., and J. H. Farley. 1995. Catch monitoring of fresh tuna caught by the Bali-based Indonesian/Taiwanese longline fishery. Council for the conserva- tion of southern bluefin tuna scientific meeting; 10-19 July 1995, Shimizu, Japan. Far Seas Fisheries Res. Lab., Shimizu, Japan, Rep. CCSBT/95/2, 18 p. Farley and Davis: Reproductive dynamics of Thunnus maccoyii 231 Table 2 Spawning fraction and spawning interval of southern bluefin tuna, Thunnus maccoyii, by month for all females and females in prime spawning condition collected from the spawning ground between October 1992 and March 1995. POF = postovulatory follicles, n = no. of fish in sample. Females in prime spawning All females condition (<10% atresia) n with Spawning Spawning n with Spawning Spawning n POF’s fraction interval (days) n POF’s fraction interval (days) 1992 Oct 3 1 0.3 3.0 1 1 1.0 1.0 Nov 1 1 1.0 1.0 1 1 1.0 1.0 Dec 1 1 1.0 1.0 1 1 1.0 1.0 1993 Jan 9 8 0.9 1.1 7 7 1.0 1.0 Feb 3 1 0.3 3.0 2 1 0.5 2.0 Mar 4 4 1.0 1.0 4 4 1.0 1.0 Apr 2 0 0.0 Aug 3 3 1.0 1.0 1 1 1.0 1.0 Sep 9 4 0.4 2.3 3 3 1.0 1.0 Oct 11 8 0.7 1.4 9 8 0.9 1.1 Nov 7 4 0.6 1.8 4 4 1.0 1.0 Dec 26 20 0.9 1.1 19 19 1.0 1.0 1994 Jan 11 6 0.5 1.8 6 6 1.0 1.0 Feb 53 36 0.7 1.5 34 31 0.9 1.1 Mar 48 31 0.6 1.5 31 28 0.9 1.1 Apr 16 13 0.8 1.2 11 11 1.0 1.0 May 1 0 1 0 Aug 1 1 1.0 1.0 Sep 32 12 0.4 2.7 16 11 0.7 1.5 Oct 43 26 0.6 1.7 26 22 0.8 1.2 Nov 40 22 0.6 1.8 21 19 0.9 1.1 Dec 40 23 0.6 1.7 26 22 0.8 1.2 1995 Jan 47 30 0.6 1.6 31 28 0.9 1.1 Feb 32 20 0.6 1.6 19 19 1.0 1.0 Mar 8 3 0.4 2.7 4 3 0.8 1.3 Unknown 16 11 0.7 1.5 8 8 1.0 1.0 Total 467 289 0.62 1.62 287 259 0.90 1.11 all months except July, although the main spawning season appeared to be from Sep- tember to April when CPUE was highest. Pre- viously, the spawning period was reported to be limited to the months of September to March (Mimura, 1958; Kikawa, 1964a). Our study included fish from areas north of that traditionally fished by Japanese vessels, which may explain the extended spawning season. An increase in the length of the spawning season towards equatorial waters has been suggested for many species includ- ing black skipjack, Euthynnus lineatus (Schaefer, 1987), yellowfin tuna, Thunnus albacares (McPherson, 1991), blue marlin, Makaira nigricans (Hopper, 1990), and other multiple spawning fish (Qasim, 1955). The actual duration of spawning in indi- vidual southern bluefin tuna could not be de- 100 80- cc 60- o. o t: o o. 40- 20- 7 26 32 43 47 _ 40 i i i i i i — i i m i i i — i — i — i — i — i — i — i A S O N D J FMAMJ JASOND'j FM 1993 1994 1995 Figure 7 The proportion of southern bluefin tuna, Thunnus maccoyii , spawn- ing by month. The vertical bars indicate 95% confidence limits of the mean for samples sizes >5. The numbers indicate sample sizes. 232 Fishery Bulletin 96(2), 1 998 200 400 600 800 1000 1200 Oocyte diameter (pm) Figure 8 Size-frequency distribution of oocytes representing the advanced stages of matu- ration, by 50-pm intervals, in southern bluefin tuna, Thunnus maccoyii, ovaries. Only oocytes >100 pm were measured. termined. However, the prevalence of females in different phases of spawning activity (nonspawning, spawning, or postspawning) on the spawning ground should be in di- rect proportion to the duration of that phase, if females in each phase are equally catchable. The ratio of nonspa wners : spawners :postspawn- ers for females caught on the spawn- ing ground was estimated to be 0.30:0.69:0.01 or 0.44:1:0.01 if spawners are set as 1. In other words, spawning females were 2.3 times more prevalent than non- spawning females and 108 times more prevalent than postspawning females on the spawning ground. This finding indicates that the du- ration of the nonspawning phase is less than half, and the postspawning phase one onehundredth, of the du- ration of the spawning phase. Spawning frequency and fecundity The potential annual fecundity of southern bluefin tuna is indeter- minate (not fixed prior to spawn- ing) because unyolked oocytes are continually matured and spawned during the season. Annual fecun- dity is indeterminate in many tuna species such as black skipjack (Schaefer, 1987), skipjack tuna (Hunter et al., 1986), and yellow- fin tuna (Schaefer, 1996). The fe- cundity of southern bluefin tuna was estimated by Kikawa (1964b) as the number of advanced yolked oocytes in the ovary, and by Thoro- good (1986) as the number of oo- cytes >300 pm in diameter in the ovary. These esti- mations of fecundity were not potential annual fe- cundity because southern bluefin tuna are capable of continuously maturing and spawning oocytes from a pool of unyolked oocytes (<300 pm). Southern bluefin tuna can spawn many times dur- ing a season. The ovaries of 25% of the females col- lected from the spawning ground contained evidence that they had all recently spawned (postovulatory follicles) and were about to do so again (migratory nucleus or hydrated oocytes). We assume that postovulatory follicles persist in the ovaries of south- ern bluefin tuna for about 24 hours as has been found in other tropical spawning tunas (Hunter et al., 1986; McPherson, 1988; Nikaido et al., 1991; Schaefer, 1996). The average interval between spawning in southern bluefin tuna, estimated from the propor- tion of ovaries containing postovulatory follicles, was 1.1 days for females in “prime spawning condition.” Similar spawning rates have been reported in other tuna species that spawn in tropical waters: 1.54 and 1.27 days for yellowfin tuna (McPherson, 1991; Schaefer, 1996); 1.1 days for bigeye tuna (Nikaido et al, 1991) and 1.18 days for skipjack tuna (Hunter et Farley and Davis: Reproductive dynamics of Thunnus maccoyii 233 al., 1986). The mean spawning interval of reproduc- tively active yellowfin tuna (those whose ovaries con- tain advanced yolked oocytes and may or may not contain postovulatory follicles and contain no or less than 50% a atresia) was 1.14 days (Schaefer, 1996). Histological sections indicated that slight variations in the lengths of individual spawning intervals exist in southern bluefin tuna, which may be normal or the result of external stresses such as decreased food avail- ability or the stress of capture (Hunter et al., 1986). Our estimate of mean relative batch fecundity for southern bluefin tuna (57 oocytes per gram of body weight) is similar to that found in yellowfin tuna (68 oocytes per gram of body weight) ( Schaefer, 1996) but less than that for black skipjack tuna (81 to 153 oo- cytes per gram of body weight) (Schaefer, 1987) and skipjack tuna (40 to 130 oocytes per gram of body weight) (Stequert and Ramcharrum, 1995). Hunter et al. (1985) reported that the minimum number of females needed for a reliable batch fecundity esti- mate in northern anchovy was 50. We, however, found only 21 southern bluefin tuna ovaries with unovu- lated hydrated oocytes from the 475 ovaries collected from the spawning ground. There could be several reasons for this. Oocytes in the hydrated stage may be very short-lived in southern bluefin tuna, which would reduce the chance of sampling females with hydrated ovaries. Spawning may occur during a spe- cific time of the day, and if sampling was not con- ducted just prior to this, fewer females with hydrated ovaries would be collected. Spawning in many tuna species is known to occur in the late evening or early morning (Hunter, et al., 1986; Schaefer, 1987; McPherson, 1991; Nikaido et al., 1991; Schaefer, 1996). Because our ovaries were sampled from fe- males caught on Indonesian-based longlining vessels, which generally operate during daylight hours (Ishida et al., 1994), the chances of catching females that were about to spawn would be reduced. Further, if spontaneous spawning occurred while southern bluefin tuna were on the longlines, then fewer fe- males with unovulated hydrated oocytes would be sampled. Spontaneous spawning has been observed in skipjack tuna soon after capture ( Kaya et al., 1982). Batch fecundity in southern bluefin tuna increased with body length. The variation in estimates for fe- males of similar size may be normal because females were collected in different years and at different times in their spawning cycles. Hunter et al. (1985) suggested that the relation between batch fecundity and fish weight should be estimated annually in northern anchovy because batch fecundity can vary by a factor of 2 between years. Batch fecundity is also known to vary significantly during the spawn- ing season in many fish species (Conover, 1985). A decrease in batch fecundity during the spawning sea- son was found in Atlantic mackerel, Scomber scombrus (Watson et al., 1992), as individuals moved northwards. Batch fecundity can also peak during the middle of the spawning season if conditions are suitable, or it can remain constant if conditions are unpredictable (Conover, 1985). Spawning strategies We found that many southern bluefin tuna ovaries, collected both on and off the spawning ground, con- tained moderate levels of a atresia (10-50% of ad- vanced yolked oocytes). High levels of a atresia are thought to indicate a decline in the spawning rate (Hunter and Macewicz, 1985a) and would normally occur towards the end of an individual’s spawning season. Hunter and Macewicz ( 1980) classified north- ern anchovy as early postspawning if their ovaries contained less than 50% of advanced yolked oocytes in an a atretic state and postspawning if their ova- ries contained more that 50% a atresia. Because we found that ovarian atresia increased with gonad in- dex in the ovaries of prespawning females from the Southern Oceans, it appears that increased a atre- sia will not always mark the cessation of spawning in southern bluefin tuna. This ovarian atresia found in prespawning females could be a normal hormonal process that occurs during ovary maturation, as sug- gested by Macer ( 1974) for horse mackerel ( Traehurus 234 Fishery Bulletin 96(2), 1 998 trachurus). Atresia can also be caused by starvation as Scott (1962) found in maturing rainbow trout ( Salmo gairdneri). It is not known if prespawning southern bluefin tuna are resorbing developing oo- cytes to gain the energy required for migration to the spawning grounds or because the ovary can only accommodate a certain volume of oocytes. Many ovaries from nonspawning southern bluefin tuna ovaries collected from the spawning ground con- tained 10-50% atresia of their advanced yolked oo- cytes. The precise reproductive stage of these females is unclear. The similarity in atretic levels between these females and the prespawning females from the southern oceans, and in their high mean GI values, suggests that these females may have only just ar- rived on the spawning ground or were in the early stages of their spawning cycle. If this is the case, the relatively large numbers of nonspawning females (30% of females sampled) indicates that southern bluefin tuna may delay the onset of spawning possi- bly to recover from the energetic costs of migration. Their presence on the spawning ground throughout the spawning season suggests that there is a con- tinual supply of new spawners onto the ground. Al- ternatively, southern bluefin tuna may not spawn continuously while on the spawning ground, but in pulses. The nonspawning females may be experienc- ing a lull in spawning activity between spawning episodes. The presence of many nonatretic yolked oocytes in their ovaries suggests that they could re- commence spawning in the current season. Lunar spawning cycles have been documented in many tropical spawning fishes (see reviews by Johannes, 1978; Taylor, 1984; Robertson et al., 1990), however, we detected no evidence of a lunar cycle. Spawning in southern bluefin tuna is not synchro- nized for the stock as a whole. There are several lines of evidence to support this statement: the presence of prespawning females both on and off the spawn- ing ground throughout the spawning season; the absence of a peak in GI during the spawning season; the constant level of spawning intensity (percentage of females spawning) during the spawning season; the absence of any increase in the incidence of oo- cyte atresia towards the end of the spawning season; and the presence of postspawning females off the spawning ground both early and late in the spawn- ing season. Thus it appears that there is a turnover of new spawners replacing old spawners on the spawning ground throughout the season. A similar turnover of pre- and postspawning southern bluefin tuna has been reported on the “Oki” fishing ground (Fig. 1) south of the spawning ground (Kikawa, 1964b). Nonsynchronized spawning has been found in other multiple spawning species such as skipjack tuna (Cayre and Farrugio, 1986), chub mackerel, Scomber japonicus (Dickerson et al., 1992), and At- lantic croaker, Micropogonias undulatus (Barbieri et al., 1994). Cayre and Farrugio (1986) reported that spawning in skipjack tuna in the Atlantic is synchro- nized within schools. Individuals in a school can mature rapidly and spawn batches of oocytes simul- taneously when conditions become favourable. It is unclear if southern bluefin tuna can do this, or to what extent the long spawning season is the result of individual fish or schools arriving on the spawn- ing ground and maturing at different times. It is also unclear if the spawning period is constant for all in- dividuals. In many species, including jack mackerel, Trachurus symmetricus, and chub mackerel, Scomber japonicus , older spawners are reported to have a longer spawning period than younger spawners (Knaggs and Parrish, 1973; MacCall et al., 1980). The low number of postspawning female southern bluefin tuna (3) found on the spawning ground sug- gests that as soon as individuals have finished spawn- ing they quickly move off the ground. The reasons for this departure are uncertain, but it may be due to decreased food availability through increased com- petition because many fish are gathering to spawn in a relatively small area, or to an inability to with- stand the warmer water temperatures on the spawn- ing ground for extended periods of time. Adult blue- fin tuna are unique among the tunas because they live predominantly in cold water (as low as 5°C) and only move into warmer waters to spawn (Olson, 1980). Their ability to maintain their body tempera- ture above ambient water temperature, through the development of an increased lateral blood supply and heat exchangers, has enabled them to occupy higher latitudes than many tuna species can tolerate. This adaptation to cold water, however, may preclude them from extended stays in warm water, resulting in a rapid migration off the spawning ground after spawning. After spawning, southern bluefin tuna migrate south from the spawning ground into the West Wind Drift (Mimura, 1962) to feed and gain condition over the southern winter months. The minimum time for an individual to travel to the Southern Ocean how- ever, is unknown. The maximum sustained swimming speeds of small yellowfin and skipjack tunas are pre- dicted to be between 2 and 4 body lengths/s (Brill, 1996), and these values are thought to be similar in other active fish species. Southern bluefin tuna are unlikely to travel at their maximum sustained swim- ming speed in a straight line from the spawning grounds to the southern oceans because they will be feeding on the way south. If a 180-cm fish travelled at between 1 and 2 body lengths/s, from the spawn- ing ground to Tasmania (6,000 km), it would take Farley and Davis: Reproductive dynamics of Thunnus maccoyii 235 between 19 and 39 days. Warashina and Hisada (1970) reported that lean fish with brownish meat have been caught in the Tasman Sea as early in the spawning season as November. These lean fish are believed to be postspawning females who have re- cently migrated from the spawning ground. The ear- liest that postspawning females were observed on the spawning ground was in October of both the 1993-94 and 1994-95 spawning seasons. The appearance of these postspawning fish on the spawning ground in October and again around Tasmania in November sup- ports the idea that southern bluefin tuna are capable of travelling that distance in approximately one month. It also supports the idea that individuals spawn for a relatively short time in contrast to the whole season. Acknowledgments This research was facilitated through a joint moni- toring program set up by CSIRO and the Central Research Institute for Fisheries of Indonesia. We thank the Director of the Research Institute of Ma- rine Fisheries, N. Naamin, and S. Bahar, the project manager in Indonesia, for their continuing support. We especially thank S. Simorangkir, the Director of PT (perseroan terbatas) Perikanan Samodra Besar, for facilitating biological sampling at the processing plant; our field staff, L. Siregar and M. Machmud, for collecting data and biological samples; and the masters of fishing vessels that voluntarily collected ovaries of southern bluefin tuna. We thank D. Le for laboratory assistance, K. Haskard and A. Cowling for statistical advice, and AFZ and RTMP observers for obtaining samples outside Indonesian waters. We thank A. Koslow, K. Schaefer, and G. West for re- viewing the manuscript and V. Mawson for editorial suggestions. This research was supported by Fish- eries Resources Research Fund grants from the Aus- tralian Fisheries Management Authority. Literature cited Barbieri, L. R., M. E., Chittenden Jr., and S. K. Lowerre-Barbieri. 1994. Maturity, spawning, and ovarian cycle of Atlantic croaker, Micropogonias undulatus, in the Chesapeake Bay and adjacent coastal waters. Fish. Bull. 92:671-685. Brill, R. W. 1996. Selective advantages conferred by the high perfor- mance physiology of tunas, billfish, and dolphin fish. Comp. Biochem. Physiol. 113:5-15. Caton, A., K. McLoughlin, and M. J. Williams. 1990. Southern bluefin tuna; scientific background to the debate. Bureau of Rural Resources Bulletin 3, 41 p. Aus- tralian Government Publishing Service, Canberra. Carey, F. G. 1973. Fishes with warm bodies. Sci. Am. 228:36-44. Cayre, P., and H. Farrugio. 1986. Biologie de la reproduction du Listao ( Katsuwonus pelamis) de l’Ocean Atlantique. In P. E. K. Symons et al. (eds), Proceedings of the ICCAT conference on the inter- national skipjack year program, Madrid, p. 252-272. Int. Comm. Conserv. Atl. Tunas. Conover, D. A. 1985. Field and laboratory assessment of patterns in fe- cundity of a multiple spawning fish: the Atlantic silver- side Menidia menidia. Fish. Bull. 83:331-341. Davis, T. L. O., G. P. Jenkins, and J. W. Young. 1990. Patterns of horizontal distribution of the larvae of southern bluefin ( Thunnus maccoyii ), and other tuna in the Indian Ocean. J. Plank. Res. 12:1295-1314. Dickerson, T. L., B. J. Macewicz, and J. R. Hunter. 1992. Spawning frequency and batch fecundity of chub mackerel, Scomber japonicus, during 1985. Calif. Coop. Oceanic Fish. Invest. Rep. 33:130-140. Fitzhugh, G. R., and W. F. Hettler. 1995. Temperature influence on postovulatory follicle de- generation in Atlantic menhaden, Brevoortia tryannus. Fish. Bull. 93:568-572. Hopper, C. N. 1990. Patterns of Pacific blue marlin reproduction in Ha- waiian waters. In R. H. Stroud (ed), Planning the future of billfishes, part 2: research and management in the 90s and beyond. Proc. Sec. Int. Billfish Symp., Hawaii, USA, 1988. Nat. Coal. Mar. Conserv., Savannah, Georgia, p. 29-39. Hunter, J. R., and S. R. Goldberg. 1980. Spawning incidence and batch fecundity in northern anchovy, Engraulis mordax. Fish. Bull. 77:641-652. Hunter, J. R., and B. J. Macewicz. 1980. Sexual maturity, batch fecundity, spawning frequency, and temporal pattern of spawning for the northern anchovy, Engraulis mordax, during the 1979 spawning season. Calif. Coop. Oceanic Fish. Invest. Rep. 21:139-149. 1985a. Measurement of spawning frequency in multiple spawning fishes. In R. Lasker (ed.), An egg production method for estimating spawning biomass of pelagic fish: application to the northern anchovy, Engraulis mordax. U.S. Dep. Commer., NOAATech. Rep. NMFS 36:79-94. 1985b. Rates of atresia in the ovary of captive and wild north- ern anchovy, Engraulis mordax. Fish. Bull. 83:119-136. Hunter, J. R., N. C. H. Lo, and R. J. H. Leong. 1985. Batch fecundity in multiple spawning fishes. In R. Lasker (ed.), An egg production method for estimating spawning biomass of pelagic fish: application to the north- ern anchovy, Engraulis mordax. U.S. Dep. Commer., NOAATech' Rep. NMFS 36:67-77. Hunter, J. R., B. J. Macewicz, and J. R. Sibert. 1986. The spawning frequency of skipjack tuna, Katsuwonus pelamis , from the south Pacific. Fish. Bull. 84:895—903. Ishida, K., T. Yamamoto, and B. Gafa. 1994. Development of fisheries for tuna and tuna-like fish in Indonesia with particular reference to the Jakarta-based tuna longline fishery. Indo-Pac. Tuna Develop. Manage. Programme, Colombo, IPTP/92/WP/26, 37 p. Johannes, R. E. 1978. Reproductive strategies of coastal marine fishes in the tropics. Env. Biol. Fish. 3:65-84. Kaya, C. M., A. E. Dizon, S. D. Hendrix, T. K. Kazama, and M. K. K. Queemth. 1982. Rapid and spontaneous maturation, ovulation, and 236 Fishery Bulletin 96(2), 1998 spawning of ova by newly captured skipjack tuna, Katsu- wonus pelamis. Fish. Bull. 80:393-396. Kikawa, S. 1964a. A study of the indomaguro ( Thunnus thynnus maccoyii) in the area south of 20° S lat. from the examina- tion of ovary weight and maturity. Rep. Nankai Reg. Lab. 20:37-57. 1964b. Estimated number of eggs spawned out of the ova- ries of an indomaguro (Thunnus thynnus maccoyii?). Rep. Nankai Reg. Fish. Res. Lab. 20:27-35. Knaggs, E. H., and R. H. Parrish. 1973. Maturation and growth of Pacific mackerel Scomber japonicus Houttuyn. Calif. Dep. Fish Game, Mar. Res. Tech. Rep. 3, 19 p. Lowerre-Barbieri, S. K., and L. R. Barbieri. 1993. A new method of oocyte separation and preservation for fish reproduction studies. Fish. Bull. 9 1( 1): 165—170. Macer, C. T. 1974. The reproductive biology of the horse mackerel Trachurus trachurus (L.) in the North Sea and English Channel. J. Fish. Biol. 6:415-438. MacCall, A. D., H. W. Frey, D. D. Huppert, E. H. Knaggs, J. A. McMillan, and G. D. Stauffer. 1980. Biology and economics of the fishery for jack mack- erel in the northeastern Pacific. U.S. Dep. Commer., NOAATech. Memo. NMFS-SWFC-4, 79 p. Maxwell, J. G. H., and G. R. Cresswell. 1981 . Dispersal of tropical marine fauna to the Great Aus- tralian Bight by the Leeuwin Current. Aust. J. Mar. Freshwater Res. 32:493-500. McPherson, G. R. 1988. Verification of the postovulatory follicle method for establishing the spawning frequency of yellowfin, bigeye and skipjack tuna in the Coral Sea. Queensland Dep. Primary Industries, Northern Fisheries Res. Center, Cairns, Australia, Tech. Rep. FRB 88/9, 42 p. 1991. Reproductive biology of yellowfin tuna in the east- ern Australian fishing zone, with special reference to the North-western Coral Sea. Aust. J. Mar. Freshwater Res. 42:465-477. Mimura, K. 1958. Fishing condition in the so-called Indo-magro ( Thunnus maccoyii ?) in the eastern seas of the Indian Ocean. Rep. Nankai Reg. Lab. 7:49-58. 1962. Studies on indomaguro, Thunnus maccoyii ? (prelimi- nary report). Occas. Rep. Nankai Reg. Fish. Res. Lab. 1:1522. Nikaido, H., N. Miyabe, and S. Ueyanagi. 1991. Spawning time and frequency of bigeye tuna, Thunnus obesus. Bull. Nat. Res. Inst. Far Seas Fish. 28:47-73. Nishikawa, Y., M. Honma, S. Ueyanagi, and S. Kikawa. 1985. Average distribution of larvae of oceanic species of scombrid fishes, 1956-1981. Far. Seas Fish. Res. Lab. S Ser. 12:1-99. Olson, R. J. 1980. Synopsis of biological data on the southern bluefin tuna, Thunnus maccoyii (Castlenau, 1872). Int.-Am. Trop. Tuna Comm. Spec. Rep. 2:151-212. Qasim, S. Z. 1955. Time and duration of spawning season of teleosts in relation to their distribution. J. Cons. Int. Explor. Mer. 21:144-155. Robertson, D. R., C. W. Petersen, and J. D. Brawn. 1990. Lunar reproductive cycles of benthic -brooding reef fishes: reflections of larval biology or adult biology? Ecol. Mono. 60:311-329. Schaefer, K. M. 1987. Reproductive biology of black skipjack, Euthynnus lineatus, an eastern pacific tuna. Int.-Am. Trop. Tuna Comm. Bull. 19:169-260. 1996. Spawning time, frequency, and batch fecundity of yellowfin tuna, Thunnus albacares , near Clipperton Atoll in the eastern Pacific Ocean. Fish. Bull. 94:98-112. Scott, D. P. 1962. Effects of food quality on fecundity in rainbow trout, Salmo gairdneri . J. Fish. Res. Board Canada 19:715-731. Shingu, C. 1967. Distribution and migration of the southern bluefin tuna. Rep. Nankai Reg. Fish. Res. Lab. 25:19-36. 1970. Studies relevant to distribution and migration of southern bluefin tuna. Bull. Far Seas Res. Lab. 3:57-113. Stequert, B., and B. Ramcharrum. 1995. La fecondite du listao (Katsuwonus pelamis) de l’ouest de 1’ocean Indien. Aquat. Living Resour. 8:79-89. Taylor, M. H. 1984. Lunar synchronization of fish reproduction. Trans. Am. Fish. Soc. 113:484-493. Thorogood, J. 1986. Aspects of the reproductive biology of the southern bluefin tuna ( Thynnus maccoyii). Fish. Res. 4:297-315. Ueyanagi, S. 1969. Observations on the distribution of tuna larvae in the Indo-Pacific Ocean with emphasis on the delineation of the spawning areas of the albacore, Thunnus alalunga. Bull. Far. Seas Fish. Res. Lab. (Shimizu) 2:177-254. Warashina, I., and K. Hisada. 1970. Spawning activity and discoloration of meat and loss of weight in the southern bluefin tuna. Bull. Far Seas Fish. Res. Lab. 3:147-165. Watson, J. J., I. G. Priede, P. R. Witthames, and A. Owori-Wadune. 1992. Batch fecundity of Atlantic mackerel, Scomber scombrus L. J. Fish Biol. 40:591-598. Yonemori, T., and J. Morita. 1978. Report on 1977 research cruise of the R/V Shoyo Maru. Distribution of tuna and billfishes larvae and oceano- graphic observation in the eastern Indian Ocean, October- December, 1977. Rep. Res. Div., Fish. Agency Jpn. 52: 1-48. Yukinawa, M. 1987. Report on 1986 research cruise of the R/V Shoyo Maru . Distribution of tuna and billfishes larvae and oceano- graphic observation in the eastern Indian Ocean January- March, 1987. Rep. Res. Div., Fish. Agency Jpn. 61:1-100. Yukinawa, M., and T. Koido. 1985. Report on 1984 research cruise of the R/V Shoyo Maru . Distribution of tuna and billfishes larvae and oceano- graphic observation in the eastern Indian Ocean January- March, 1985. Rep. Res. Div., Fish. Agency Jpn. 59:1-108. Yukinawa, M., and N. Miyabe. 1984. Report on 1983 research cruise of the R/V Shoyo Maru. Distribution of tuna and billfishes larvae and oceano- graphic observation in the eastern Indian Ocean October- December, 1983. Rep. Res. Div., Fish. Agency Jpn. 58: 1-103. 237 Abstract .—Acoustic backscatter data from a 38-kHz echo sounder and a 150-kHz acoustic Doppler current profiler (ADCP) were collected during Southwest Fisheries Science Center marine mammal surveys in the east- ern Pacific aboard the NOAA ship David Starr Jordan in 1992 and 1993. These data were processed to give pro- files of volume scattering strength. A deep scattering layer dominated the time-depth patterns of backscatter. This layer migrated from a depth of 300-400 m during the day to 0-100 m at night, when it was located just above the thermocline. The source of backscatter was primarily small fish and squid, which are dolphin prey. Dolphin sighting rates were correlated positively with nighttime volume scat- tering strength above the thermocline. Spatial and temporal variability of prey biomass in the surface layer, as indexed by volume scattering strength, may have important consequences for dolphin feeding, as well as for distri- butional responses to habitat variabil- ity by dolphins. Manuscript accepted 19 June 1997. Fishery Bulletin 96: 237-247 (1998). Dolphin prey abundance determined from acoustic backscatter data in eastern Pacific surveys Paul C. Fiedler Jay Barlow Tim Gerrodette Southwest Fisheries Science Center National Marine Fisheries Service, NOAA PO. Box 271, La Jolla, California 92038 E-mail address (for R Fiedler): PFiedler@noaa.gov Acoustic data, consisting of the re- turns of pulses of sound from tar- gets in the water column, are rou- tinely used to assess fish distribu- tion and abundance (Forbes and Nakken, 1972; Johannesson and Mitson, 1983; MacLennan and Simmonds, 1992). Such data are now also being used to quantify the biomass and distribution of zoo- plankton and micronekton (Greene and Wiebe, 1990; Hewitt and Demer, 1993). Innovative techniques using multifrequency amplitude and phase information can give esti- mates of size, movement, and acous- tic properties of individual plank- tonic organisms (Farmer and Hus- ton, 1988; Holliday et al., 1989). Another approach in acoustic stud- ies of zooplankton and micronekton has been to employ instrumentation that is used routinely, notably the acoustic Doppler current profiler (ADCP, Roe and Griffiths, 1993). Several studies have used uncali- brated ADCP echo intensity data, with various levels of processing, to estimate relative patterns of plank- ton biomass distribution in space and time (Flagg and Smith, 1989; Plueddemann and Pinkel, 1989; Smith et ah, 1989; Heywood et al., 1991). We have been collecting echo in- tensity data from a 38-kHz echo sounder since 1986 and from a 150- kHz ADCP since 1992 on Southwest Fisheries Science Center marine mammal surveys. This report de- scribes data collection and process- ing and presents results from two cruises during which nearly com- plete data sets were obtained (popu- lation of Delphinus stocks 1992 [PODS92] and 1993 [PODS93]) (Fig. 1). Other data collected on these surveys provide a context for these results in terms of predators and habitat. Line-transect data were collected for dolphin abun- dance estimates (Mangels and Gerrodette, 1994a, 1994b). We use sighting rate here as an index of dolphin abundance. Physical and biological oceanographic data were also collected to characterize habitat variability (Philbrick et al., 1993). Materials and methods A 38-kHz Simrad EK-400 scientific sounder was used on NOAA ship David Starr Jordan beginning in 1986. Rather than a commercial echo integrator, we used an acous- tic data acquisition (ADA) system, consisting of an analog-to-digital converter and a personal computer. A 150-kHz ADCP was installed on the Jordan in 1991. We collected nearly complete sets of 38- and 150- kHz acoustic data in 1992 and 1993. A sonar equation relating volume backscattering strength per m3 ( Sv , 238 Fishery Bulletin 96(2), 1998 Figure 1 Cruise tracks of NOAA ship David Starr Jordan during Population of Delphinus Stocks (PODS) cruises (28 July- 2 November 1992) and (28 July-6 November 1993). dB // l\xPa at 1 m) to the echo level (EL), integrated and recorded after 201ogR TVG amplification in the transceiver, was formulated as follows (Simmonds, 1990): Sv = EL - ( SL + RR) - 10 log \\i - 10 log ct/2, (1) where EL = 20 log Vo (dB) (Vo=“gain-controlled en- velope” signal, volts); SL = source level (dB); RR - receiver response (including receiver gain and transducer on-axis voltage re- sponse), dB; 10 log \| / = equivalent beam angle; c = speed of sound (m/s); t = pulse length ( 1 x 1/1 ,000 per millisecond ). Transmission loss does not appear in this equa- tion because it was corrected by a 201ogR time-var- ied gain (TVG) function applied to Vo in the trans- ceiver. The TVG was checked with a transceiver test mode that input a constant voltage signal. The re- corded 201ogR TVG output did not differ significantly from the theoretical function (TVG = 20 log/? + 2a R, where a is the nominal sound absorption coefficient of 0.008 dB/m and R is range in meters). The range (R) was based on a sound velocity entered into the control unit at the start of the cruise. On-axis sensitivity ( SL+RR ) was estimated by measuring the system response (EL) to a standard calibration sphere (38.1-mm diameter tungsten car- bide with 6% cobalt binder, target strength TS = - 42.3 dB; Foote, 1990) suspended 7 m below the trans- ducer. In this case, Equation 1 reduces to EL = TS + (SL + RR), (2) where EL = 20 log Vo + 20 log R. The “gain-controlled envelope” signal (Vo) represents a peak voltage and was converted to a root-mean- square (rms) voltage by multiplying by 1/V2. The additional 201ogR term was added to the system’s 201ogR TVG, used during the echo-integration sur- veys, to correct for the spherical spreading of the re- turn from the point calibration target. Solving Equa- Fiedler et a I.: Dolphin prey abundance determined from acoustic backscatter data 239 tion 2 for the on-axis sensitivity gave ( SL+RR ) = 74.2 dB. Operationally, the ADA program squared VQ, which is proportional to pressure at the face of the trans- ducer, to give an echo intensity value (Eq. 1) propor- tional to power or energy flux per unit of area. The intensity values were then binned in 50 10-m bins between 10-m and 5 10-m depth. The binned intensi- ties were averaged at 10-sec intervals over ten min- utes (60 pings), equivalent to an interval of 3 km at the typical ship speed of 18 km/h. Recorded intensi- ties were converted to units of decibels (dB=10 log// I , where Ir represents the standard reference of IpPa at a 1-m distance from the face of the transducer). Then, a correction to the TVG function was formu- lated by using depth-dependent sound velocities and attenuation coefficients calculated from CTD data taken once or twice daily with the algorithms of Francois and Garrison (1982). Finally, a constant 53.4 dB was subtracted to account for the combined source level, transducer response, receiver gain, and beam pattern terms (-SL - RR - 10 log Vj/ - 10 log t(/ 2) in Equation 1. The equivalent beam angle is very difficult to measure directly without special appara- tus; therefore we used the manufacturer’s nominal value (10 log vj/=-19.6 dB; Simrad EK400 Scientific Sounder Instruction Manual). Acoustic Doppler current profilers have been used since the early 1980’s to estimate current velocity from the Doppler shift of acoustic signals backscattered from suspended matter (plankton and sediments) in the water column. Echo intensity is calculated and recorded during data-processing. RD Instruments has developed instrument modifications and an al- gorithm to calibrate these values and to estimate vol- ume backscattering strength (Sv, dB). The algorithm is based on the following working version of a sonar equation (RD Instruments, 1990): Sv = 10 log 4.47 x 10~20 K2Ks(Tt +273) (10 k(E—Er)/ 10 -j\^2 / (cPK1 10 ■2aR/10\ (3) where K0 = system noise factor = 4.3; Ks = system constant = 4.17 x 105; Tf = transducer temperature (°C); k = conversion factor = 0.435 dB/count; E = echo intensity (counts); Er = reference level for echo intensity (counts); R = slant range (m); c = sound velocity (m/s); P = transmit pulse length = 8 m; ATj = power into the water, W; a - sound absorption coefficient (dB/m). Although this equation was formulated to be ap- plied to individual beams of the ADCP, we used av- erage amplitude data for E because we were using an empirical reference level (Er). The reference level for echo intensity, Er, represents thermal noise in the system electronics plus ship noise. Er was assumed equal to the minimum value of E (<40 counts) ob- served in each average profile or ensemble (-740 pings in 10 minutes). The manufacturer states that this in-situ method of estimating Er is preferable to calculating Er from measured system temperatures. Because ship temperature was relatively constant and water temperature ranged over only 4°C in 1992 and 10°C in 1993, Er varied by only (±7-8 counts. Values of E < 10 counts were not used (RD Instru- ments, 1990) and appear as blank areas at depth in 150 kHz time-depth sections (Fig. 2). Values for c and a were calculated from CTD data taken once or twice daily according to Francois and Garrison (1982). Transducer temperature, Tt, was assumed equal to observed surface temperature. Slant range, R, was corrected for estimated sound velocity in relation to the value assumed by the ADCP software (1,475 m/s). The conversion factor, k (dB/ count), depends on the temperature of the system electronics. However, the deck unit was located in a temperature-controlled room (18-21°C) so that the range of k was 0.433 to 0.437 (<1 dB). Power into the water, Kv was assumed constant and estimated to be 53.4 W (RD Instruments, 1990). The ADCP was set up with 50 bins of length 8 m, starting at 4 m from the transducers. In both the 38- and 150-kHz data sets, the depth bin representing the bottom was identified initially by an algorithm searching for strong gradients of Sv(z). These results were modified by visual inspec- tion of time-depth sections. Data from the bottom depth bin and below were excluded from subsequent analysis. Some 1992 and 1993 profiles of S appeared to be biased low on account of sound attenuation due to bubbles beneath the ship during stationkeeping and during very rough weather (cf. Fig. 2, 20 August, at the first tick marking the morning CTD station). We identified biased profiles from anomalously low Sy in bins below the top depth bin. The problem was more prevalent in the 150-kHz data (2.5% of 1992 profiles and 1.0% of 1993 profiles) than in the 38- kHz data (0.5% of 1992 profiles and 0.1% of 1993 profiles). We excluded the biased profiles from sta- tistical analyses. Dolphin sightings were recorded during daylight hours from the flying bridge of the vessel by a team 240 Fishery Bulletin 96(2), 1998 920820 920821 920822 920823 920824 Figure 2 Time-depth sections of 38-kHz and 150-kHz volume scattering strength (S , dB) and temperature during five days of PODS92 surveys. Contour lines are isotherms (°C), from XBT (expendable bathythermograph) profiles at tick marks along top of section. of three observers using 25-power binoculars (Man- gels and Gerrodette, 1994a, 1994b). Local dolphin abundance was indexed by the number of sightings of spotted ( Stenella attenuata), spinner (S. longi- rostris), striped ( S . coeruleoalba), common (Delphi- nus delphis and D. capensis), bottlenose ( Tursiops truncatus), Risso’s (Grampus griseus), rough-toothed ( Steno bredanensis ), Fraser’s ( Lagenodelphis hosei), Pacific white-sided (Lagenorhynchus obliquidens), and unidentified dolphins. Daily numbers of sightings were standardized to sighting rates of schools or in- dividuals per 100 km in optimum sea-state condi- tions (Gerrodette1 ). Sighting rates were log-trans- formed to normalize distributions before testing cor- relation with various time and depth means of Sv(z,t). We report correlations with 150 kHz Sv here because data were collected on about 25% more days than those for 38kHz Sv . Acoustic and other data were gridded by kriging with the software SURFER (Golden Software, 1995). Variogram models were fitted to the data and the best fit was selected by using the original code based 1 Gerrodette, T. 1996. On estimating relative abundance. Unpubl. manuscript. on Pannatier (1996). Contour maps were also gener- ated with SURFER . Results The five-day time-depth sections in Figure 2 illus- trate features observed in all sections: 1) data at one or both frequencies were occasionally missing; 2) the 150-kHz signal was lost in the noise at depths >200 m; and 3) bottom depth was <400 m as on 24 August after about 1700 hours. All time-depth sections of Sv were dominated by a deep-scattering layer (DSL) at 300^100 m depth during daylight hours and 0-100 m during the night. The DSL migrated about 200 m in 1-2 hours near dawn and dusk. Morning descent was generally more rapid than evening ascent. During some days, the DSL was split into two distinct lay- ers at depth, separated by up to 100 m (cf. Fig. 2, 38 kHz section on 22 and 23 August). The DSL was detected at both 38 and 150 kHz. However, the 150-kHz return from the DSL at its daytime depth was barely above background noise levels. The smaller effective depth range of the ADCP was due to 1) greater attenuation of sound at the higher frequency (a=0.039 dB/m at 150 kHz com- Fiedler et al.: Dolphin prey abundance determined from acoustic backscatter data 241 PODS93 PODS92 38 kHz Sv (dB) Figure 3 Relationships between 38-kHz and 150-kHz volume scattering strength (Sv, dB) in corresponding 10-and 8-m depth bins (cen- tered at 25, 65, 105, 145, 185, 225, 265, and 305 m) at all times of day or night during PODS92 (rc=24,193) and PODS93 (n=26,963) surveys. -100 -100 -100 -100 Table 1 Mean day (0800-1600) and night (2000-0400) volume scattering strengths (S,,,dB) in the surface layer (0-100 m) in 1992 and 1993. Values are cruise means, with range in parentheses. PODS92 = Population of Delphi nus Stocks 1992 cruise; PODS93 = Population of Delphinus Stocks 1993 cruise. Day Night PODS92 38 kHz -75.1 (-82.2 to -70.7) 150 kHz -73.9 (-83.7 to -69.0) -68.6 (-77.1 to -62.0) -67.5 (-76.8 to -57.3) PODS93 38 kHz -76.3 (-85.8 to -69.4) 150 kHz -79.2 (-90.2 to -70.1) -65.3 (-73.5 to -60.2) -68.2 (-80.3 to -59.7 ) pared with 0.008 dB/m at 38 kHz) and 2) transducer orientation 30° off the vertical, resulting in a slant range 15% greater than depth. Volume scattering strengths at 38 and 150 kHz were significantly cor- related (Fig. 3, r= 0.58 in 1992 and r=0.77 in 1993). Some error in this relation is due to the different sizes and inexact depth matching of bins between frequencies. The ADCP samples echo intensity only in the last quarter of each depth bin, so that scattering is averaged over only 2 m, compared with 10 m for the 38-kHz echo sounder. Cor- relation decreased at depth (e.g. in 1993, r=0.83 at 25 m and r=0.53 at 185 m). This could have been due sim- ply to the low signal to noise ratio at depth in the 150- kHz data. However, it probably also reflects changes in species and size composition, and thus spectral re- sponse, of scatterers at depth. The mean difference be- tween 38- and 150-kHz volume scattering strength in Figure 3 was -2.9 dB in 1992 and 0.0 dB in 1993. Spatial patterns of mean day and night volume scattering strength ( Sv ) in the surface layer (0-100 m) were very similar at the two frequencies; therefore only the 38-kHz patterns are illustrated here (Figs. 4 and 5). On average, night values exceeded day val- ues by 6.4 dB in 1992 and 11.0 dB in 1993 (Table 1). Thus, near-surface Sv was 4 and 13 times greater at night than during the day, indicating that 77-92% of the scatterersjeft the surface layer during the day. Mean surface Sv values (daily or nightly, 38 or 150 kHz) varied by 11.5 to 20.6 dB during the cruises, representing ranges of 14 to 115 times. In 1992, low Sv was_pbserved along 8-9°N both night and day (Fig. 4). Sv increased to the south to- wards the equator and to the north and east towards 242 Fishery Bulletin 96(2), 1998 the coast. The highest Sv values were observed along the coast of Costa Rica and the coast of Mexico south of the Gulf of Tehuantepec (16°N, 95°W). In 1993, 1, “geometric” scattering), early models predicted TS to be approximately con- stant (Love, 1977). Most other models of geometric scattering are highly nonlinear, with deep nulls in individual TS versus ( ka ) curves. However, Chu et al. (1992) suggested that averaging over pings and individuals in field studies of volume backscattering should “smear out” these nonlinearities. The wavelengths of 38-and 150-kHz sound corre- spond to target dimensions of approximately 4 and 1 cm, respectively. This is the effective size of macro- zooplankton and micronekton organisms, such as euphausiids, siphonophores, small fish, and squid. Targets <4 cm, such as copepods, will scatter sound strongly at 150 kHz. Scattering of 38-kHz sound will be relatively weak Rayleigh scattering and may be undetectable over system noise. These smaller tar- gets may have weakened the correlation between backscattering at the two frequencies (Fig. 3). We were unable to make net tows to identify acous- tic targets, owing to time constraints on the cetacean survey cruises. Therefore, we can make only an edu- cated guess about the composition of the observed deep-scattering layers. Roe et al. ( 1984), in a uniquely comprehensive study of a community of vertically migrating organisms in the temperate northeast At- lantic, considered fish, decapod crustaceans, mysids, euphausiids, amphipods, copepods, ostracods, siphonophores, medusae, ctenophores, and chaetog- naths. Ignoring small or weak scatterers, we must consider fish, crustaceans (decapods, mysids, and euphausiids), and siphonophores, as well as cepha- lopods, as possible components of the observed DSL’s. Average biomass densities of mesopelagic fish (pri- marily myctophids) and cephalopods are equal to, or slightly greater than, those of crustaceans in the east- ern tropical Pacific (Blackburn, 1968; Blackburn et al., 1970) and subtropical Pacific (Maynard et al., 1975). Target strength depends on the composition, Table 3 Correlations between dolphin sighting rates (schools and individuals per 100 nautical miles of effort, log-transformed) and mean 150-kHz volume scattering strength (S , dB) above the thermocline at night (2000-0400), and 0 to 400 m night and day (0800-1600). *** = P<0.001, ** = P<0.01, * = P<0.05. PODS92 = Population of Delphinus Stocks 1992 cruise; PODS93 = Population of Delphinus Stocks 1993 cruise. Dolphin sighting rates Schools Individuals PODS92 Night Sv above thermocline +0.35* +0.38** Night Sv 0-400m +0.05 +0.02 Day Sv 0-400m -0.01 -0.02 PODS93 Night Sy above thermocline +0.27 +0.15 Night Sv 0-400m +0.36** +0.20 Day Sv 0-400m +0.38** +0.18 shape, and orientation of the target, as well as on relative size. At 38-120 kHz, small fish with gas-filled swim bladders (e.g. clupeoids) and squid have target strengths 5-10 dB greater than those for crustaceans of the same size (Tables 6.3 and 6.4 in MacLennan and Simmonds, 1992; Marchal et al., 1993). Siphono- phores have a gas bladder that scatters sound and have been observed migrating with myctophids in a DSL off Baja California (Barham, 1966). However, myctophids, flying fish, and squid were much more abundant than siphonophores at night dipnet sta- tions on our cruises.2 Therefore, mesopelagic fish and 2 Pitman, R. 1996. Southwest Fisheries Science Center, Natl. Mar. Fish. Serv., NOAA, Box 271, La Jolla, CA92038. Personal commun. Fiedler et a I.: Dolphin prey abundance determined from acoustic backscatter data 245 PODS92 PODS93 Thermocline depth (m), average nighttime backscatter above the thermocline (dB), and dolphin sighting rate observed in 1992 and 1993. 246 Fishery Bulletin 96(2), 1 998 squid were most likely the major component of the observed DSL’s, although a variety of crustaceans were undoubtedly present in the survey areas. Nonmigrating, epipelagic fish and squid are likely represented by near-surface backscattering during the day. The small fish and squid that we argue were de- tected by our acoustic systems are dolphin prey (Fitch and Brownell, 1968; Miyazaki et al., 1973; Perrin et al., 1973; Robison and Craddock, 1983; Robertson and Chivers, 1997). The ranges of volume scattering strength we observed — more than three orders of magnitude among bins (Fig. 2), up to lOx variation between night and day and 500x variation among daily means — must be important factors in dolphin feeding strategies and habitat choices. Thermocline depth, which influences prey distribution, is an im- portant component in dolphin habitat variability (Reilly and Fiedler, 1994). We observed a positive correlation between the abundance of dolphins (observed during daylight) and the abundance of prey at night, when spotted and spinner dolphins are known to feed on mesopelagic prey near the surface (Perrin et al., 1973; Robertson and Chivers, 1997). Over 90% of the dolphins identi- fied in both surveys were common, spotted, spinner, and striped dolphins. In the eastern tropical Pacific (PODS92), a region characterized by a strong and shallow thermocline (Wyrtki, 1967), nighttime prey abundance was more closely related to thermocline depth. Off Baja California (PODS93), the thermocline had less influence on distribution of prey. Dolphin community structure was also different: common dolphins represented 61% of the identified dolphins in this region, compared with only 37% in the PODS92 survey area. Our results indicate that dolphin distribution can be related to prey abundance measured acoustically and that the physical environment influences the abundance of dolphin prey. Marchal et al. (1993) ob- served a similar DSL in the eastern tropical Atlantic that was related both to thermal structure and to tuna abundance. Simple and readily available sonar systems can yield useful results when survey time or resources are limited. Monitoring long-term changes in zooplankton or micronekton biomass, as Roemmich and McGowan ( 1995) did with net samples off southern California, could be facilitated by the application of such acoustic methods. However, in present and future studies of cetacean feeding and cetacean responses to habitat variability, we will use multispectral acoustic systems and validation by net samples to quantify availability and distribution of specific prey types. Acknowledgments We thank Valerie Philbrick for collecting these data; Denny Sutton and Jim Anthony of NOAA’s Pacific Marine Center for support in maintaining the acous- tic systems; Roger Hewitt and Dave Demer for assis- tance in the calibration of the EK-400 echo sounder; Ken Richter, Dave Demer, and Paul Smith for com- ments on an earlier draft of this manuscript; and Steve Reilly for continued support of environmental studies during marine mammal surveys. Literature cited Barham, E. G. 1966. Deep scattering layer migration and composition: observations from a diving saucer. Science (Wash. D.C.) 151:1399-1403. Blackburn, M. R. 1968. Micronekton of the eastern tropical Pacific Ocean: family composition, distribution, abundance, and relations to tuna. Fish. Bull. 67( 1 ):71— 115. Blackburn, M. R., M. Laurs, R. W. Owen, and B. Zeitschel. 1970. Seasonal and areal changes in standing stocks of phytoplankton, zooplankton and micronekton in the east- ern tropical Pacific. Mar. Biol. 7:14-31. Brooks, A. L. 1977. A study of the swimbladders of selected mesopelagic fish species. In N. R. Andersen and B. J. Zahuranec (eds.), Oceanic sound scattering prediction, p. 565-590. Plenum Press, New York, NY. Chapman, W. M. 1947. The wealth of the ocean. Sci. Monthly (NY ). 64:192- 197. Chu, D., T. K. Stanton, and P. H. Wiebe. 1992. Frequency dependence of sound backscattering from live individual zooplankton. ICES J. Mar. Sci. 49:97-106. Demer, D. A., and L. V. Martin. 1995. Zooplankton target strength: volumetric or areal dependence? J. Acoust. Soc. Am. 98(2 1:1111—1118 Farmer, D. M., and R. D. Huston. 1988. Novel applications of acoustic backscatter to biologi- cal measurements. In B. J. Rothschild (ed.), Toward a theory on biological-physical interactions in the world oceans, p. 599-614. Kluwer, Dordrecht. Fitch, J. E., and R. L. Brownell Jr. 1968. Fish otoliths in cetacean stomachs and their impor- tance in interpreting feeding habits. J. Fish. Res. Board Canada 25(121:2561-2574. Flagg, C. N., and S. L. Smith. 1989. On the use of the acoustic Doppler current profiler to measure zooplankton abundance. Deep-Sea Res. 36:455- 474. Foote, K. G. 1990. Spheres for calibrating an eleven-frequency acoustic measurement system. J. Cons. Int. Explor. Mer 46:284—286. Forbes, S. T., and O. Nakken (eds.). 1972. Manual of methods for fisheries resource survey and appraisal. Part 2: The use of acoustic instruments for fish detection and abundance estimation. FAO Man. Fish. Sci. 5:138 p. Fiedler et a I.: Dolphin prey abundance determined from acoustic backscatter data 247 Francois, R. E., and G. R. Garrison. 1982. Sound absorption in sea water: Part II: Boric acid contribution and equation for total absorption. J. Acoust. Soc.Am. 72:1879-1890. Golden Software. 1995. SURFER for Windows, version 6. Golden Software, Inc., Golden, CO, 483 p. Greene, C. H., and P. H. Wiebe. 1990. Bioacoustical oceanography: new tools for zooplank- ton and micronekton research in the 1990’s. Ocean- ography 3:12-17. Hersey. J. B., and H. B. Moore. 1948. Progress report on scattering layer observations in the Atlantic Ocean. Trans. Am. Geophys. Union 29:341- 354. Hewitt, R. P., and D. A. Demer. 1993. Dispersion and abundance of Antarctic krill in the vicinity of Elephant Island in the 1992 austral summer. Mar. Ecol. Prog. Ser. 99:29-39. Heywood, K. J., S. Scrope-Howe, and E. D. Barton. 1991. Estimation of zooplankton abundance from shipborne ADCP backscatter. Deep-Sea Res. 38:677-691. Holliday, D. V., R. E. Pieper, and G. S. Kleppel. 1989. Determination of zooplankton size and distribution with multifrequency acoustic technology. J. Cons. Int. Explor. Mer 46:52-61. Johannesson, K. A., and R. B. Mitson. 1983. Fisheries acoustics: a practical manual for aquatic biomass estimation. FAO Fish. Tech. Paper 240, 249 p. Love, R. H. 1977. Target srength of an individual fish at any aspect. J. Acoust. Soc.Am. 62:1397-1403. Lyman, J. 1948. The sea’s phantom bottom. Sci. Monthly (NY) 66: 87-88 MacLennan, D. N., and E. J. Simmonds. 1992. Fisheries acoustics. Chapman and Hall, London, 325 p. Mangels, K. F., and T. Gerrodette. 1994a. Report on cetacean sightings during a marine mam- mal survey in the eastern tropical Pacific Ocean aboard the NOAA ships McArthur and David Starr Jordan , July 28-November 2, 1992. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-SWFSC-200, 74 p. 1994b. Report on cetacean sightings during a marine mam- mal survey in the eastern Pacific Ocean and the Gulf of California aboard the NOAA ships McArthur and David Starr Jordan , July 28-November 6, 1993. U.S. Dep. Commer., NOAA Tech. Memo. NMFS- SWFSC-211, 88 p. Marchal, E., F. Gerlotto, and B. Stequert. 1993. On the relationship between scattering layer, ther- mal structure and tuna abundance in the Eastern Atlan- tic equatorial current system. Oceanol. Acta 16(3):261- 273. Marshall, N. B. 1951. Bathypelagic fishes as sound scatterers in the ocean. J. Mar. Res. 11:1—17. Maynard, S. D., F. V. Riggs, and J. F. Walters. 1975. Mesopelagic micronekton in Hawaiian waters: fau- nal composition, standing stock, and diel vertical migration. Fish. Bull. 73:726-736. Miyazaki, N., T. Kusaka, and M. Nishiwaki. 1973. Food of Stenella coeruleoalba. Sci. Rep. Whales Res. Inst., Tokyo, 25:265-275. Pannatier, Y. 1996. VARIOWIN: software for spatial data analysis in 2D. Springer- Verlag, New York, NY, 91 p. Perrin, W. F., R. R. Warner, C. H. Fiscus, and D. B. Holts. 1973. Stomach contents of porpoise, Stenella spp. , and yel- lowfin tuna, Thunnus albacares, in mixed-species aggre- gations. Fish. Bull. 71:1077-1091. Philbrick, V. A., P. C. Fiedler, S. B. Reilly, R. L. Pitman, and L. T. Ballance. 1993. Report of ecosystem studies conducted during the 1992 eastern tropical Pacific common dolphin survey on research vessels David Starr Jordan and McArthur. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-SWFSC-180, 74 p. Plueddemann, A. J., and R. Pinkel. 1989. Characterisation of the patterns of diel migration using a Doppler sonar. Deep-Sea Res. 36:509-530. RD Instruments. 1990. Calculating absolute backscatter. Tech. Bull. ADCP- 90-04, San Diego, CA, 24 p. Reilly, S. B., and P. C. Fiedler. 1994. Interannual variability of dolphin habitats in the eastern tropical Pacific. I: Research vessel surveys, 1986- 1990. Fish. Bull. 92:434-450. Robertson, K., and S. J. Chivers. 1997. Prey occurrence in pantropical spotted dolphins, Stenella attenuata , from the eastern tropical Pacific. Fish. Bull. 95(2):334-348. Robison, B. H., and J. E. Craddock. 1983. Mesopelagic fishes eaten by Fraser’s dolphin, Lageno- delphis hosei. Fish. Bull. 81(2):283-289. Roe, H. S. J., M. V. Angel, J. Badcock, P. Domanski, P. T. James, P. R. Pugh, and M. H. Thurston. 1984. The diel migrations and distributions within a me- sopelagic community in the northeast Atlantic. 1. Intro- duction and sampling procedures. Prog. Oceanogr. 13: 245-268. Roe, H. S. J., and G. Griffiths. 1993. Biological information from an Acoustic Doppler Cur- rent Profiler. Mar. Biol. 115:339-346. Roemmich, D., and J. McGowan. 1995. Climatic warming and the decline of zooplankton in the California Current. Science (Wash. D.C.) 267:1324- 1326. Simmonds, E. J. 1990. Very accurate calibration of a vertical echo sounder: a five-year assessment of performance and accuracy. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 189:183-191. Smith, P. E., M. D. Ohman, and L. E. Eber. 1989. Analysis of the patterns of distribution of zooplank- ton aggregations from an Acoustic Doppler Current Profiler. Rep. Calif. Coop. Ocean. Fish. Invest. 30:88-103. Tont, S. A. 1976. Deep scattering layers: patterns in the Pacific. Rep. Calif. Coop. Ocean. Fish. Invest. 18:112-117. Wyrtki, K. 1967. Circulation and water masses in the eastern equato- rial Pacific Ocean. Int. J. Oceanol. Limnol. 1:117-147. 248 Aibstracf.-Peiagic larval summer flounder, Paralichthys dentatus, were collected in the NW Atlantic Ocean from 1985 to 1993, and their feeding habits were examined in relation to lar- val stage. Collections included pre- flexion, flexion, premetamorphic, and metamorphic larvae, although pre- flexion larvae composed the bulk of the specimens. Incidence of feeding and gut-fullness data indicated that larvae began feeding near sunrise and contin- ued feeding throughout daylight hours. Incidence of feeding reached its lowest point, 8.3%, at 0400-0559 hours, then dramatically increased to 54.6% at 0600-0759. Maximum gut fullness was seen between 1200 and 1559. The only time during which all larvae contained prey in their guts was 0800-0959. Im- mature copepodites were the primary prey for all oceanic larval stages. In addition, small prey, such as tintinnids and copepod nauplii, made major con- tributions to the preflexion diet, and large prey, such as calanoid copepods and appendicularians, made major con- tributions to the diets of premetamor- phic and metamorphic larvae. Meta- morphic larvae were also collected as they entered a New Jersey estuary, at night, and their diet was examined. At 1800-1959 and 2000-2159 hours, the incidence of feeding in estuarine larvae was significantly lower than in oceanic larvae. The estuarine diet was domi- nated by the calanoid copepod, Temora longicornis. Incidence of feeding was observed to decline as metamorphosis progressed, from 19.1% at stage G to 2.9% at stage H. However, this appar- ent cessation in feeding, when the right eye was at the midpoint of migration, may not bring about undue ecological peril to summer flounder larvae. Manuscript accepted 10 June 1997. Fishery Bulletin 96:248-257 (1998). Feeding habits of pelagic summer flounder, Paralichthys dentatus, larvae in oceanic and estuarine habitats Jill J. Grover College of Oceanic and Atmospheric Sciences Hatfield Marine Science Center Oregon State University Newport, Oregon 97365 E-mail address: jj.grover@hmsc.orst.edu The summer flounder, Paralichthys dentatus , an important component of recreational and commercial fish- eries over the continental shelf and in estuaries of the Middle Atlantic Bight (United States), is currently overexploited. Evidence of this in- cludes a drastic reduction in both commercial and recreational land- ings, in relation to peak years, and compressed age structure (Able and Kaiser, 1994). Both the fishery’s de- cline and efforts to culture summer flounder (Bengtson et al., 1994; Bengtson and Nardi, 1995) are re- viving interest in the basic biology of this species. Although much is known about the early life history of summer flounder, beginning with a description of eggs and larvae (Smith and Fahay, 1970) and includ- ing the timing of spawning, offshore distribution of eggs and larvae (Smith, 1973; Able et al., 1990), in- shore occurrence of larvae and ju- veniles (Able et al., 1990), patterns of metamorphosis (Keefe and Able, 1993), and settlement processes (Burke et al., 1991; Keefe and Able, 1994; Norcross and Wyanski, 1994), many gaps still exist in our collec- tive knowledge of larval and early- juvenile ecology. Because survival beyond the early life history stages is based on an “eat and not be eaten” strategy (Keenleyside, 1979; Olla and Davis, 1988; Olla et al., 1994), feeding hab- its clearly define much of larval and early-juvenile ecology. Recently, lar- val summer flounder studies have focused on the effects of delayed feeding on survival and growth (Bisbal and Bengtson, 1995a), de- scribing the starving condition (Bisbal and Bengtson, 1995b) and the development of the digestive tract (Bisbal and Bengtson, 1995c). Although these studies examined natural rates and processes, they were all conducted on larvae that were reared in a laboratory. Most of what is known about the early feed- ing ecology of wild summer floun- der has been derived from metamor- phosing larvae and early juveniles that were captured with benthic trawls in estuarine nursery habi- tats (Burke, 1995). Beyond the patent inference that planktonic larvae ingest plankton (e.g. Morse, 1981), very little is known regard- ing the early pelagic feeding ecol- ogy of this species (Rogers and Van Den Avyle, 1983; Grimes et al., 1989). This study considers feeding ecol- ogy of larval summer flounder col- lected in pelagic habitats in the NW Atlantic Ocean from 1985 to 1993. Feeding habits were examined in relation to larval stage. Metamor- phic larvae were also collected as they entered an estuarine habitat in New Jersey, and their diet was examined. Grover: Feeding habits of pelagic summer flounder larvae 249 Materials and methods During fall and winter, pelagic larval summer flounder were collected in the NW Atlantic Ocean from 1985 to 1993 by the National Ma- rine Fisheries Service (NMFS) Marine Re- sources Monitoring, Assessment, and Prediction (MARMAP) surveys (Able et. al., 1990). These monthly to bimonthly surveys were conducted over the continental shelf from Cape Lookout, North Carolina, to Cape Sable, Nova Scotia, by the RV Albatross IV and the RV Delaware II. The vast majority (96.4%) of oceanic specimens were collected during October and November, and a few were collected in December, January, and March. Surveys used 61-cm diameter bongo frames fitted with 0.333- and 0.505-mm mesh nets (Sherman, 1980). The nets were towed ob- liquely through the water from the surface to depths of 22-75 m and back to the surface at a speed between 1.5 and 3.5 knots. Samples from 50 stations were the focus of this study (Fig. 1). Collections were chosen for analysis on the ba- sis of number of summer flounder larvae that were present. All collections with 10 or more sum- mer flounder larvae were examined. Additional stations with fewer specimens were examined to balance time blocks so that each 2-h block included at least 10 specimens and at least 2 collections, and to include the full size range of larvae. Specimens were preserved in 10% formalin at sea and remained in formalin for approximately 12 months. They were then transferred to 70% ethanol. Specimen shrinkage clearly occurred as a result of fixation. Within 24 hours of examination specimens were soaked in glycerin where they remained throughout their dissection (see Arthur, 1976; Gadomski and Boehlert, 1984). After standard length ( SL) (snout to notochord tip until full flexion, then to posterior edge of hypurals, Gadomski and Boehlert, 1984) of each larva was measured and morphological stage was determined, the digestive tract was removed. Contents of the en- tire digestive tract were evaluated. Gut contents were teased out and prey items were identified to the low- est possible taxon. Diet was analyzed in terms of numerical percent- age composition (%N), volumetric percentage com- position (%Vol), and percent frequency of occurrence (%FO). Prey that comprised <1% of the diet by num- ber and by volume were pooled into the “other” prey category. Prey volumes were estimated, generally by assuming a spheroidical geometry, from prey dimen- sions. The three analyses (%N, %Vol, and %FO) were combined to yield a more comprehensive assessment Figure 1 Location of the oceanic sites in the NW Atlantic Ocean, off the NE United States, where pelagic larval summer flounder, P. dentatus, were collected. of prey importance, the index of relative importance (IRI = (%N + %Vol) x %FO) (Pinkas et al., 1971). A comparison of larval sizes with established defi- nitions of length at yolk absorption and length at flexion (Smith and Fahay, 1970; Martin and Drewry, 1978) revealed that considerable shrinkage had oc- curred prior to the examination of these specimens. As a result, data were pooled across stations for analysis by morphological stage, rather than by size class. Morphological stages were defined as preflexion (PF): straight notochord and no indication of caudal- fin ray development; flexion (FLX): beginning of cau- dal-fin ray development (i.e. ossification), accompa- nied or not by an upturn of the notochord tip and ossification of ural bones (hypurals, epurals, para- hypurals); postflexion, premetamorph (PM): comple- tion of caudal-fin ray ossification, upturn of noto- chord, and ossification of the ural bones, accompa- nied by resorption of the notochord tip, such that it no longer extends beyond the edge of the hypural bones (Fahay1); and metamorph (M): metamorphic 1 Fahay, M. 1995. Northeast Fisheries Science Center, Howard Marine Sciences Laboratory, Natl. Mar. Fish. Serv., NOAA, High- lands, NJ 07732. Personal commun. 250 Fishery Bulletin 96(2), 1998 stages F- through I (Keefe and Able, 1993), based on degree of eye migration. Incidence of feeding and gut fullness were also examined, as a function of 2-h time blocks. Gut full- ness (F) was recorded as 0 = empty, 1 = 1-25%, 2 = 26-50%, 3 = 51-75%, and 4 = 76-100% full. Incidence of feeding was recorded as the percent frequency of lar- vae that had prey in their guts (i.e. F > 0), in relation to the total number of specimens examined in a time block. A total of 550 oceanic larvae were examined. Of these, 18 were excluded from analysis because large portions of their guts were missing, and 11 were ex- cluded because their primitive condition likely impaired their ability to ingest prey (Blaxter, 1986; Gadomski2). Primitive characteristics included presence of yolk sac, absence of eye pigment, and undeveloped mouth. Evi- dence of prey ingestion was lacking in the 29 excluded larvae. This study was based on 521 oceanic larvae that were deemed to be functionally intact. A second series of collections were examined to determine the feeding ecology of summer flounder larvae as they enter an estuarine nursery habitat. Metamorphic larvae were collected from plankton entering the Great Bay-Little Egg Harbor (New Jer- sey) estuary (Fig. 2) during fall, winter, and spring from 1989 to 1995. Most estuarine specimens (71.4%) were collected between November and January. The remainder were collected between February and June, and in October. Stationary plankton nets (1-m diameter, 1.0-mm mesh) were set for 30 min, at the surface and just off the bottom, on nocturnal flood tides, from a bridge across Little Sheepshead Creek (Szedlmayer et al., 1992; Keefe and Able, 1993). This site is characterized by Atlantic Ocean water on flood tides (Charlesworth, 1968, cited in Szedlmayer et al., 1992); thus larvae were likely captured as they en- tered the estuary from the ocean, or soon after. Estuarine specimens were preserved in 95% etha- nol, and they remained in ethanol until their examina- tion. Specimens were processed as above. Because all estuarine specimens were metamorphic ( ME ), morpho- logical stages were recorded in terms of Keefe and Abie’s (1993) metamorphic stages: F- through I. The estua- rine portion of this study was based on 119 specimens. Results Preflexion (PF) larvae were the dominant morpho- logical stage in oceanic collections, accounting for 84.3% of the specimens (Table 1). Later morphologi- 2 Gadomski, D. 1995. Columbia River Research Laboratory, National Biological Service, 5501 A Cook-Underwood Road, Cook, WA 98605. Personal commun. Figure 2 Location of the estuarine site in Little Sheepshead Creek, New Jersey, where metamorphic summer flounder, P. dentatus, were collected. (Figure by K. W. Able, Rutgers University.) cal stages were progressively rarer, with flexion (FLX) larvae accounting for 12.5%, postflexion, premeta- morphic (PM) larvae accounting for 1.7%, and meta- morphic (M) larvae accounting for 1.5% of the oce- anic specimens in this study. Although greater than 33% of the larvae in each morphological stage con- tained prey in their guts, stages PM and M were com- bined for dietary analyses because of their small rep- resentation in these collections and similarity in diet. Larval lengths that were recorded at the time samples were sorted (usually within 12 months of collection) were compared with lengths at the time of examination in order to estimate shrinkage. Com- parisons based on 329 specimens defined mean shrinkage as 13.7% (SD=5.83). Diet of preflexion (PFJ larvae The dominant stage in this study, PF, was represented by a total of 439 specimens: 56.9% contained recog- nizable prey, and 43.1% had empty guts. Lengths at the time of examination ranged from 1.9 to 6.9 mm SL (Table 1). Grover: Feeding habits of pelagic summer flounder larvae 251 Table 1 Feeding incidence and number of prey ingested by summer flounder, P. dentatus, larvae by time block, habitat, and larval stage. OC = oceanic, ES = estuarine, PF = preflexion, FLX = flexion, PM = premetamorphic, and M = metamorphic (oceanic). Estuarine metamorphic stages are as defined by Keefe and Able (1993). Numeric data exclude copepod eggs that appeared to have been ingested incidentally by metamorphic larvae. Standard length (mm) Habitat type No. larvae examined mean SD range No. larvae with food Percentage feeding Total no. food items No. food items per feeding larvae Time block (h) 0000-0159 OC 67 3.8 1.22 2. 3-7. 5 11 16.4 40 3.6 ES 15 11.4 1.29 8.6-13.4 1 6.7 1 1.0 0200-0359 OC 18 4.5 1.04 3. 4-6. 7 2 11.1 3 1.5 ES 11 12.3 1.91 8.8-14.6 1 9.1 1 1.0 0400-0559 OC 60 4.8 1.35 2. 2-7. 2 5 8.3 7 1.4 ES 11 11.3 0.97 9.5-12.9 2 18.2 3 1.5 0600-0759 OC 44 4.3 1.39 2.6-7. 6 24 54.5 104 4.3 ES 2 10.3 0.44 10.0-10.6 0 0.0 0 0.0 0800-0959 OC 10 4.1 0.76 3. 2-5. 3 10 100.0 46 4.6 1000-1159 OC 33 3.8 0.96 2. 7-5.8 27 81.8 107 4.0 1200-1359 OC 39 3.7 1.23 2. 1-8.0 38 97.4 259 6.8 1400-1559 OC 41 3.6 1.16 2. 1-8.3 36 87.8 256 7.1 1600-1759 OC 67 3.6 0.70 2. 6-6. 2 47 70.1 168 3.6 1800-1959 OC 32 3.8 1.12 1.9-7. 9 27 84.4 106 3.9 ES 36 12.4 1.62 8.8-14.5 5 13.9 13 2.6 2000-2159 OC 60 3.9 1.37 2. 3-9.0 42 70.0 258 6.1 ES 39 11.9 1.48 8.1-13.9 5 12.8 12 2.4 2200-2359 OC 50 4.1 1.25 2. 1-8.3 11 22.0 31 2.8 ES 5 11.3 1.76 8.8-13.5 1 20.0 2 2.0 Larval stage PF OC 439 3.6 0.85 1.9-6. 9 250 56.9 1276 5.1 FLX OC 65 5.6 0.75 3. 7-7. 2 22 33.8 86 3.9 PM OC 9 6.6 0.92 4.8-7. 6 4 44.4 17 4.3 M OC 8 7.7 0.97 5. 8-9.0 3 37.5 6 2.0 G ES 47 11.9 1.74 8.1-14.4 9 19.1 19 2.1 H- ES 34 12.0 1.51 8.8-14.6 3 8.8 7 2.3 H ES 35 11.8 1.35 8.8-14.5 1 2.9 2 2.0 H+ ES 1 12.6 12.6 1 100.0 1 1.0 I ES 2 13.3 0.71 12.8-13.8 1 50.0 3 3.0 Immature copepodites composed the bulk of the diet (61.4% Vol, 37.3% IRI; Table 2) of PF larvae. Copepod nauplii, the second most important prey, composed 20.0% (N and IRI) of the diet. Tintinnids, despite being the most abundantly ingested prey (28.7% N), ranked third in importance at 19.3% (IRI). Bivalve larvae and copepod eggs were the only other prey that accounted for >1% of the diet, and together they composed 21.7% (IRI). The size of ingested prey was directly related to larval size. Only PF larvae, the smallest larvae, abun- dantly ingested small prey, i.e. tintinnids, copepod nauplii, copepod eggs, and bivalve larvae. Diatoms and dinoflagellates were also occasionally ingested (4.4% FO), although both were of such minor impor- tance in the total diet that they were pooled into the “other” prey category. The mean number of prey found in the guts of PF larvae was 5.1 (SD=4.47, range: 1-25). For PF larvae, the ingestion of cope- pod eggs and invertebrate eggs was clearly indepen- dent of the ingestion of berried females. Visual cues of prey may be important for feeding larvae because grains of sand that visually “mimicked” bivalve lar- vae, in size, shape, and color, were ingested along with bivalve larvae by 4.4% of PF larvae. Diet of flexion fFLX) larvae A total of 65 FLX specimens were represented in this study: 33.8% contained recognizable prey, and 66.2% had empty guts. Lengths at the time of examination ranged from 3.7 to 7.2 mm SL (Table 1). Immature copepodites dominated the diet of FLX larvae, regardless of method of analysis (Table 3). 252 Fishery Bulletin 96(2), 1998 Table 2 Diet of pelagic preflexion (PF ) summer flounder, P. dentatus, larvae collected from the NW Atlantic, in terms of numeri- cal percentage composition (%N), volumetric percentage composition (%Vol), percentage frequency of occurrence (%FO), and percentage index of relative importance (IRI = (%N + %Vol ) x %FO). All values are based on a sample size, with prey, of 250 larvae. Prey %N %Vol %FO %IRI Tintinnids 28.7 3.3 37.6 19.3 Copepod nauplii 20.0 10.2 41.2 20.0 Copepodites 16.0 61.4 30.0 37.3 Calanoids 0.6 4.9 2.0 0.2 Cyclopoids 0.6 2.0 2.4 0.1 Copepod eggs 16.0 1.2 34.8 9.6 Bivalve larvae 12.1 14.8 28.0 12.1 Invertebrate eggs 3.7 0.9 11.6 0.9 Other 2.3 1.3 9.2 0.5 Table 3 Diet of pelagic flexion (FLX) summer flounder, P. dentatus , larvae collected from the NW Atlantic, in terms of numeri- cal percentage composition (%N), volumetric percentage composition (%Vol), percentage frequency of occurrence (%FO), and percentage index of relative importance (IRI = ( %N + %Vol) x %FO). All values size, with prey, of 22 larvae. are based on a sample Prey %N %Vol %FO %IRI Polychaete larvae 1.2 0.5 4.5 0.1 Copepod nauplii 2.3 0.2 9.1 0.2 Copepodites 68.6 66.7 86.4 92.8 Calanoids 13.9 27.5 18.2 6.0 Cyclopoids 3.5 1.9 9.1 0.4 Copepod eggs 3.5 <0.1 9.1 0.3 Bivalve larvae 5.8 2.8 4.5 0.3 Invertebrate eggs 1.2 0.4 4.5 0.1 The well-digested condition of copepodites precluded specific identifications, but most appeared to be calanoids. The only other prey category to represent more than 10% of the diet by any analysis was adult calanoid copepods, which despite accounting for 27.5% of dietary volume, composed only 6.0% IRI. Tintinnids were not observed in the FLX larval diet, and other small prey items (copepod nauplii, copepod eggs, bivalve larvae, and invertebrate eggs) contributed little to the diet. The mean number of prey ingested by FLX larvae was 3.9 (SB=2.33, range=l-9). Diet of oceanic premetamorphic and rrsetamorphic fPIWI+IWSJ larvae A total of 9 PM and 8 M specimens were represented in this study: 44.4% of PM larvae and 37.5% of M larvae contained recognizable prey, and 55.6% of PM larvae and 62.5% of M larvae had empty guts. Lengths at the time of examination ranged from 4.8 to 7.6 mm SL for PM larvae and 5.8 to 9.0 mm SL for M larvae (Table 1). Combining the two stages (PM+M): 41.2% contained prey, and 58.8% had empty guts. As with earlier morphological stages, immature copepodites were the primary prey for PM and M larvae, accounting for 50.1% of the diet (IRI). How- ever, adult calanoid copepods and appendicularians, the secondary and tertiary prey (in terms of IRI), together contributed more than twice as much as copepodites to dietary volume (Table 4). Although it was not always possible to identify copepods to spe- cies, the most frequently ingested calanoid copepod Table 4 Diet of pelagic postflexion, premetamorphic (PM), and metamorphic (M) summer flounder, P. dentatus , larvae col- lected from the NW Atlantic, in terms of numerical per- centage composition (%N), volumetric percentage compo- sition (%Vol), percentage frequency of occurrence (%FO), and percentage index of relative importance (IRI = ( %N + %Vol) x %FO). All values are based on a sample size, with prey, of 7 larvae. Data exclude copepod eggs that appeared to have been ingested incidentally. Prey %N %Vol %FO %IRI Copepod nauplii 8.7 0.5 28.6 1.7 Copepodites 60.9 27.1 42.9 50.1 Calanoids 13.0 34.6 42.9 27.2 Appendicularians 17.4 37.8 28.6 21.0 appeared to be Centropages typicus. Tetnora longi- cornis and Pseudocalanus spp. also contributed to the diet. Appendicularians were only observed in the diet of PM and M larvae. The mean number of prey ingested by PM and M larvae was 3.3 (SD=2.14, range=l-6). Copepod eggs that were abundant in the gut of a single M-stage larva were excluded because they were probably in- gested incidentally. Incidence of feeding An examination of incidence of feeding (defined as the percentage frequency of larvae with prey in their guts) in relation to 2-h time blocks over the course of Grover: Feeding habits of pelagic summer flounder larvae 253 0000 0400 0800 1200 1600 2000 0200 0600 1000 1400 1800 2200 Onset of time block Figure 3 Gut fullness (F) of pelagic larval summer flounder P. dentatus, in the NW Atlantic Ocean, as a function of time of day (EST), where 0 = empty, 1 = 1— 25%, 2 = 26-50%, 3 = 51-75%, and 4 = 76-100% full. 24 h (Table 1) revealed that pelagic larvae began feeding near sunrise in oceanic habitats. The presence of prey in larval guts reached its lowest point, 8.3%, at 0400-0559 hours, then dra- matically increased to 54.5% at 0600- 0759 hours, the time block of sunrise for all collections. The only time dur- ing which incidence of feeding was 100% was 0800-0959. Throughout hours of full daylight, incidence of feed- ing remained high, >70%. Late after- noon or early-evening feeding may have buoyed the feeding incidence at 2000-2159, but evidence of recent feeding decreased sharply from 2000-2159 to 2200-2359 hours. Dur- ing late night hours, the percentage of larvae that had prey in their guts gradually decreased from 22.0% at 2200-2359 to 8.3% at 0400-0559. Gut fullness Because incidence of feeding data were generated from gut-fullness data (i.e. F>0), they provide a qualitative measure of gut fullness; how- ever, gut-fullness data also provide direct estimates of the volume of prey in the gut over the course of 24 h (Fig. 3). Although larvae appeared to begin feeding at 0600-0759, full guts were not observed until 1200- 1359. Maximum gut fullness (F=A) was only seen at 1200-1559 and 2000-2159. The only time block dur- ing which all larvae contained prey in their guts was 0800-0959. Diet of metamorphic (JV1EJ larvae in estuaries The estuarine portion of this study was based on 119 metamorphic larvae (ME), representing metamorphic stages G through I (Keefe and Able, 1993) (with n- 47 at stage G, n- 34 at stage H-, n= 35 at stage H, n=l at stage H+, and n= 2 at stage I, Table 1). The guts of 15 (12.6%) ME larvae contained recognizable prey, and 87.4% had empty guts. At the time of examina- tion, larval lengths ranged from 8.1 to 14.6 mm SL. Pelagic feeding was clearly demonstrated by > 85% of ME larvae. The primary prey of pelagic ME larvae was the calanoid copepod Temora longicornis, regard- less of the method of analysis (Table 5). The mean number of prey found in the guts of ME larvae was 2.1 (SD=0.99, range=l-4). The relatively good condition of some prey items in the guts of fish collected late at night suggests that some prey were likely ingested after dark. Table 5 Diet of metamorphic (ME) summer flounder, P. dentatus, larvae collected from a New Jersey estuary, in terms of numerical percentage composition (%N), volumetric per- centage composition (%Vol), percentage frequency of occur- rence (%FO), and percentage index of relative importance (IRI = (%N + %Vol)x %FO). All values are based on a sample size, with prey, of 15 larvae. Prey were pelagic unless oth- erwise indicated. Prey %N %Vol %FO %IRI Polychaete larvae 3.1 0.7 6.7 0.4 Polychaete tentacles (benthic) 3.1 12.0 6.7 1.6 Cirripede larvae 3.1 0.2 6.7 0.3 Paracalanus parvus 3.1 2.0 6.7 0.5 Centropages typicus 6.3 13.6 6.7 2.1 Temora longicornis 46.8 55.3 53.3 86.2 Acartia sp. 3.1 2.9 6.7 0.6 Unidentified calanoids 6.3 3.7 13.3 2.1 Harpacticoids (benthic) 6.3 1.5 6.7 0.8 Mysids (diurnal migrants) 9.4 6.1 13.3 3.3 Oikopleura sp. 6.3 1.8 13.3 1.7 Unidentified prey remnants 3.1 0.2 6.7 0.4 Evidence of benthic feeding was observed only in late-stage metamorphic larvae (H+ and I). Of two late-stage larvae that had prey in their guts, one 254 Fishery Bulletin 96(2), 1998 single larva (stage I, 13.8 mm SL) clearly demon- strated benthic feeding, having ingested polychaete ten- tacles and harpacticoid copepods, and the other larva had ingested a mysid, a taxon noted for diurnal migra- tions (Newell and Newell, 1977). All late-stage larvae in estuarine collections displayed noticeably darker pig- mentation than was observed in earlier stages. incidence of feeding in estuaries A comparison of the incidence of feeding of oceanic and estuarine summer flounder larvae was limited to nocturnal collections and revealed strikingly dif- ferent patterns (Table 1; Fig. 4). Between 1800 and 0759 hours the incidence of feeding of estuarine lar- vae never exceeded 20.0%, whereas values for oce- anic larvae ranged from 8.3% to 84.4%. The greatest differences between oceanic and estuarine larvae were observed early in the night: at 1800-1959 and 2000-2159 the incidence of feeding in estuarine col- lections was significantly lower than in oceanic col- lections (P<0.01, x2=33.78, 1 df, at 1800-1959; P<0.01, %2=30.99, 1 df, at 2000-2159) (Fig. 4). Be- cause most (63.0%) larvae in estuarine samples were collected between 1800 and 2159 (Table 1), these com- parisons were not biased by low sample size. Within estuarine habitats, incidence of feeding declined with metamorphic stage, from 19.1% at stage G to 2.9% at stage H (Table 1). Time was not a factor in this decline because the distribution of lar- vae of stages G, H-, and H was the same across early- (1800-2159), mid- (2200-0159), and late-night (0200- 0559) collections (P=0.18, %2=6.22, 4 df). Discussion The diet of all stages of oceanic summer flounder lar- vae was dominated by immature copepodites in the NW Atlantic Ocean. The size of other ingested prey was directly related to larval size. Small prey items such as copepod nauplii, tintinnids, and bivalve lar- vae, were important only in the diet of early (PF) stage larvae. Large prey items, such as adult calanoid copepods and appendicularians were important only in the diet of later (PM and M) stage larvae. The diet of metamorphosing (ME) larvae that were collected in a New Jersey estuary was dominated by the calanoid copepod Temora longicornis. These observations differ sharply from the diet of metamorphosing larval summer flounder in a North Carolina estuary (Burke, 1995). Burke (1995) re- ported that the larval diet was dominated by poly- chaetes and mysids. However, all larvae in the North Carolina study (Burke, 1995) were caught in estua- rine nursery areas with benthic trawls, whereas in the present study oceanic larvae were caught with bongo nets, and estuarine larvae were caught with stationary plankton nets. Clearly these two studies captured metamorphosing larvae during different ecological phases of the transition from pelagic to benthic ecology. The diet of late-stage metamorphic larvae that demonstrated benthic feeding in an estuarine habitat in the present study was similar to the diet of comparable size larvae in Burke’s (1995) study. The darker pigmenta- tion of benthic-feeding, late-stage lar- vae suggests that this pigmentation may be a valid marker of adaptation to the benthic habitat in summer flounder, as has been observed in other species (Grover et al., in press). The pelagic diet of a congener, the olive flounder,3 Paraliehthys olivaceus, may represent a more cogent com- parison. In Wakasa Bay in the Japan Sea, Minami (1982) reported that 3 The common and scientific names of fish spe- cies used in Fishery Bulletin are those recom- mended by the American Fisheries Society (Am. Fish. Soc. Spec. Publ. 20, 5th ed., 1991). The following names have been changed ac- cordingly: Japanese flounder to olive flounder; Parophrys vetulus to Pleuronectes vet ulus', Isopsetta isolepis to Pleuronectes isolepis\ and Linxanda ferruginea to Pleuronectes ferrugineus. Estuarine Oceanic Onset of time block Figure 4 Incidence of feeding of larval summer flounder, P. dentatus, in the NW Atlantic Ocean and in a New Jersey estuary, as a function of time of day (EST) through the night. Data are based on a total of 331 oceanic and 119 estuarine specimens. Grover: Feeding habits of pelagic summer flounder larvae 255 Summer flounder Plaice 1953a Plaice 1953b -3- Plaice 1964 — E3— Plaice 1978 Onset of time block Figure 5 Incidence of feeding of pelagic larval summer flounder, P. dentatus, in the NW Atlantic Ocean and of pelagic larval plaice, P. platessa, in the Southern Bight of the North Sea and the English Channel, as a function of time of day. Sources of plaice data are as follows: 1953a = 48-h station, Shelbourne, 1953; 1953b = other stations (Shelbourne, 1953); 1964 = Ryland, 1964; 1978 = Last, 1978. copepod nauplii were the most impor- tant prey in the diet of all premeta- morphic stages of Japanese flounder larvae. However, Ikewaki and Tanaka (1993) later reported that olive floun- der diet, in Wakasa Bay, was domi- nated by copepod nauplii only for first- feeding larvae and by appendic- ularians, Oikopleura spp., for later stages through early-metamorphic- phase larvae. Minami (1982) re- ported that appendicularians, cope- pods, and mysids were the dominant prey of early-, mid-, and late-meta- morphic larvae, respectively. The siz- able contribution of copepod nauplii to the diet of p reflexion (PF) summer flounder larvae resembled the con- tribution observed in the diet of first- feeding (Ikewaki and Tanaka, 1993) and early stage olive flounder larvae (Minami, 1982). However, appen- dicularians, the dominant prey item for early-metamorphic larvae (Min- ami, 1982), and for all larval stages beyond first-feeding in olive flounder (Ikewaki and Tanaka, 1993), occurred only in the diet of premetamorphic and metamorphic (PM+M) sum- mer flounder larvae. Pelagic summer flounder larvae that were collected in the NW Atlantic Ocean displayed a diurnal feed- ing pattern similar to that reported for several other flatfish larvae (e.g. plaic e,Pleuronectes platessa : Shel- bourne, 1953; Ryland, 1964; Last, 1978; dab, Limanda limanda : Last, 1978; English sole, Pleuro- nectes vetulus, and butter sol e, Pleuronectes isolepis ; Gadomski and Boehlert, 1984). Both incidence of feeding and gut-fullness data (Figs. 3 and 4) appear to confirm the visual nature of larval summer floun- der feeding in oceanic collections. This is not sur- prising, because marine fish larvae are mostly vi- sual feeders (Hunter, 1981; Blaxter, 1986; Huse, 1994). However, the optimal illumination level for feeding varies with species (Huse, 1994). For ex- ample, Atlantic cod, Gadus morhua, larvae feed pref- erentially at very low light levels, and turbot, Scoph- thalmus maximus, larvae feed preferentially at high levels of illumination, whereas plaice larvae feed over a wide range of illumination levels (Huse, 1994). Within the flatfishes, additional relationships be- tween illumination and larval feeding have been dem- onstrated. For example, sole, Solea solea, larvae can feed in the dark from the early posthatching stage (Blaxter, 1969). Oceanic collections of yellowtail floun- der, Pleuronectes ferrugineus, larvae have shown that the highest incidence of feeding occurred between 1600 and 0100 (Smith et ah, 1978). The near absence of feeding between 0700 and 1300 suggests that on- set of feeding in yellowtail flounder larvae is trig- gered by something other than, or in addition to, il- lumination (Smith et al., 1978). Data from oceanic collections in the present study suggest that the relationship between illumination and feeding of summer flounder larvae is much like that observed in one of the two flatfish species stud- ied by Huse ( 1994). Laboratory observations are re- quired to ascertain whether summer flounder lar- vae feed preferentially at high illumination levels or over a wide illumination range. However, an exami- nation of incidence of feeding of pelagic plaice larvae (Shelbourne, 1953; Ryland, 1964; Last, 1978) in rela- tion to comparable summer flounder data (Fig. 5) re- veals similar patterns of feeding periodicity. From this, a similarity in optimal illumination levels for feeding for summer flounder and plaice larvae is suggested. To wit, pelagic summer flounder larvae would be expected to feed well over a wide range of illumination levels. If the ecological analogy between plaice larvae and summer flounder larvae extends through metamor- phosis, then a dramatic increase in light sensitivity (and a lower threshold light intensity for feeding) at metamorphosis (Blaxter, 1968) would be predicted for summer flounder. Limited field evidence supports this. A comparison of the incidence of feeding of oce- 256 Fishery Bulletin 96(2), 1998 anic and estuarine summer flounder larvae revealed strikingly different patterns (Fig. 4). Early in the night, incidence of feeding was much lower in estua- rine collections. Fasting during metamorphosis, as has been noted in plaice (Riley, 1966; Lockwood 1984; Hamerlynck et al., 1989), may have contributed to this pattern. Plaice and sole have shown marked decreases in food-searching behaviors (distance cov- ered per minute and time spent in feeding activity) at metamorphosis (Blaxter and Staines, 1971). Late at night, at 0400-0559, incidence of feeding in es- tuarine collections was double the rate seen in oce- anic summer flounder collections (Fig. 4). A lower threshold light intensity for feeding at metamorpho- sis may have contributed to this, although small sample sizes of late-night estuarine collections pre- clude overreaching conclusions. Resolution of the threshold of light intensity for feeding of metamor- phic summer flounder in a controlled laboratory set- ting could result in determination of optimal illumi- nation levels at metamorphosis in aquaculture. Metamorphosis has been considered a critical inter- val in the early life of some marine fishes (Thorisson, 1994). In the present study, the incidence of feeding of summer flounder larvae was observed to decline with metamorphic development to stage H. This sug- gests that the midpoint in the migration of the right eye (Keefe and Able, 1993) may be a critical period for summer flounder. In a laboratory study, a cessa- tion of feeding was observed at stage G (Keefe and Able, 1993). However, cessation of feeding at meta- morphosis may not place flatfish in any real danger of starvation (Lockwood, 1984). Metamorphic plaice larvae are capable of surviving without food for 7-25 d, without reaching the “point of no return” (Wyatt, 1972). Blaxter and Hempel (1963) have defined the “point of no return” as the point at which starved larvae become too weak to feed. Midmetamorphic summer flounder larvae are capable of surviving 56 d without reaching the “point of no return” (Keefe and Able, 1993). At meta- morphosis and shortly thereafter flatfish may be more vulnerable to predation (van der Veer and Bergman, 1987; Witting and Able, 1993, 1995) than to starvation (Thorisson, 1994). Acknowledgments I would like to thank Mike Fahay, of the NOAA’s NMFS James Howard Marine Sciences Laboratory, who provided access to oceanic collections and data, and Ken Able and Stacy Hagan, Rutgers University Marine Field Station, who provided access to estua- rine collections and data. A portion of this study was supported by the University of Connecticut. Literature cited Able, K. W., and S. C. Kaiser. 1994. Synthesis of summer flounder habitat para- meters. NOAA Coastal Ocean Program Decision Analy- sis Series 1, NOAA Coastal Ocean Office, Silver Spring, MD. 68 p. + biblio. + 3 append. Able, K. W., R. E. Matheson, W. W. Morse, M. P. Fahay, and G. Shepherd. 1990. Patterns of summer flounder Paralichthys dentatus early life history in the Mid-Atlantic Bight and in New Jersey estuaries. Fish. Bull. 88:1-12. 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. 74:517-530. Bengtson, D. A., G. Bisbal, H. Iken, and R. P. Athanas. 1994. Culture of summer flounder Paralichthys dentatus : research on hatchery and growout phases. J. Shellfish Res. 13:312. Bengtson, D., and G. Nardi. 1995. Summer flounder culture: an update on research and the development of an industry. J. Shellfish Res. 14:239. Bisbal, G. A., and D. A. Bengtson. 1995a. Effects of delayed feeding on survival and growth of summer flounder Paralichthys dentatus larvae. Mar. Ecol. Prog. Ser. 121:301-306. 1995b. Description of the starving condition in summer flounder, Paralichthys dentatus , early life history stages. Fish. Bull. 93:217-230. 1995c. Development of the digestive tract in larval sum- mer flounder. J. Fish. Biol. 47:277-291. Blaxter, J. H. S. 1968. Light intensity, vision, and feeding in young plaice. J. Exp. Mar. Biol. Ecol. 2:293-307. 1969. Visual thresholds and spectral sensitivity of flatfish larvae. J. Exp. Biol. 51:221-230. 1986. Development of sense organs and behaviour of te- leost larvae with special reference to feeding and predator avoidance. Trans. Am. Fish. Soc. 115:98-114. Blaxter, J. H. S., and G. Hempel. 1963. The influence of egg size on herring larvae. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 28:211-240. Blaxter, J. H. S., and M. E. Staines. 1971. Food searching potential in marine fish larvae. In D. J. Crisp ( ed. ), Fourth European marine biology sympo- sium, p. 467-485. Cambridge Univ. Press., London. Burke, J. S. 1995. Role of feeding and prey distribution of summer and southern flounder in selection of estuarine nursery habitats. J. Fish Biol. 47:355-366. Burke, J. S., J. M. Miller, and D. E. Hoss. 1991. Immigration and settlement pattern of Paralichthys dentatus and P. lethostigma in an estuarine nursery ground, North Carolina, U.S.A. Neth. J. Sea Res. 27:393-405. Charlesworth, L. J., Jr. 1968. Bay, inlet and nearshore marine sedimentation: Beach Haven-Little Egg Inlet region, New Jersey. Ph D. diss., Univ. Michigan, Ann Arbor, MI, 614 p. Gadomski, D. M., and G. W. Boehlert. 1984. Feeding ecology of pelagic larvae of English sole Parophrys vetulus and butter sole Isopsetta isolepis off the Oregon coast. Mar. Ecol. Prog. Ser. 20:1-12. Grimes, B. H., M. T. Huish, J. H. Kerby, and D. Moran. 1989. Species profiles: life histories and environmental re- quirements of coastal fishes and invertebrates (Mid-Atlan- Grover: Feeding habits of pelagic summer flounder larvae 257 tic). Summer and winter flounder. U.S. Fish Wildl. Serv. Biol. Rep. 82(11.112). U.S. Army Corps of Engineers TR EL-82-4, 18 p. Grover, J. J., D. B. Eggleston, and J. M. Shenker. In press. Transition from pelagic to demersal phase in early- juvenile Nassau grouper, Epinephelus striatus: pigmentation, squamation, and ontogeny of diet. Bull. Mar. Sci. Hamerlynck, O., C. R. Janssen, and E. Van Landtschoote. 1989. Fasting and feeding in late larval and early post-lar- val plaice (Pleuronectes platessa L.) (Abstract). In J. H. S. Blaxter, J. C. Gamble, and H. v. Westernhagen (eds.), The early life history of fish, The third ICES symposium, Bergen, 3-5 October 1988. Rapp. R-V. Reun. Cons. Int. Explor. Mer 191:465. Hunter, J. R. 1981. Feeding ecology and predation of marine fish larvae. In R. Lasker (ed.), Marine fish larvae: morphol- ogy, ecology, and relation to fisheries, p. 34-77. Washing- ton Sea Grant Program, Univ. Wash., Seattle. Huse, I. 1994. Feeding at different illumination levels in larvae of three marine teleost species: cod, Gadus morhua L., pla- ice, Pleuronectes platessa L., and turbot, Scophthalmus maximus (L). Aquacult. Fish. Manage. 25:687-695. Ikewaki, Y., and M. Tanaka. 1993. Feeding habits of Japanese flounder (Paralichthys olivaceus ) larvae in the western part of Wakasa Bay, the Japan Sea. Nippon Suisan Gakkaishi 59:951-956. Keefe, M., and K. W. Able. 1993. Patterns of metamorphosis in summer flounder, Paralichthys dentatus. J. Fish. Biol. 42:713-728. 1994. Contributions of abiotic and biotic factors to settle- ment in summer flounder, Paralichthys dentatus. Copeia 1994:458-465. Keenleyside, M. H. A. 1979. Diversity and adaptation in fish behavior. Springer- Verlag. Berlin, 208 p. Last, J. M. 1978. The food of four species of pleuronectiform larvae in the eastern English Channel and southern North Sea. Mar. Biol. 45:359-368. Lockwood, S. J. 1984. The daily food intake of 0-group plaice ( Pleuronectes platessa L.) under natural conditions: changes with size and season. J. Cons. Int. Explor. Mer 41:181-193. Martin, F. D., and F. E. Drewry. 1978. Development of fishes of the Mid-Atlantic Bight. Vol- ume VI: Stromateidae through Ogcocephalidae. U.S. Fish and Wild. Serv., Biological Service Program FWS/OBS-78/ 12, 416 p. Minami, T. 1982. The early life history of a flounder Paralichthys olivaceus. Bull. Jpn. Soc. Sci. Fish. 48:1581-1588. Morse, W. W. 1981. Reproduction of the summer flounder Paralichthys dentatus ( L. ). J. Fish Biol. 19:189-203. Newell, G. E., and R. C. Newell. 1977. Marine plankton: a practical guide. Hutchinson & Co. Ltd. London, 244 p. Norcross, B. L., and D. M. Wyamski. 1994. Interannual variation in the recruitment pattern and abundance of age-0 summer flounder, Paralichthys dentatus, in Virginia estuaries. Fish. Bull. 92:591-598. Olla, B. L., and M. Davis. 1988. To eat or not be eaten: do hatchery-reared salmon need to learn survival skills? Underwater Nat. 17(3): 1€> — 18. Olla, B. L., M. W. Davis, and C. H. Ryer. 1994. Behavioural deficits in hatchery-reared fish: poten- tial effects on survival following release. Aquacult. Fish. Manage. 25(suppl. 1): 19-34. Pinkas, L., M. S. Oliphant, and I. L. K. Iverson. 1971. Food habits of albacore, bluefin tuna, and bonito in Cali- fornia waters. Calif. Dep. Fish Game Fish Bull. 152, 105 p. Riley, J. D. 1966. Marine fish culture in Britain. VII. Plaice (Pleuro- nectes platessa L. ) post-larval feeding on Artemia salina L. nauplii and the effects of varying feeding levels. J. Cons. Int. Explor. Mer 30:204-221. Rogers, S. G., and M. J. Van Den Avyle. 1983. Species profiles: life histories and environmental re- quirements of coastal fishes and invertebrates (South At- lantic). Summer flounder. United States Fish and Wild- life Service FWS/OBS-82/11.15. United States Army Corps of Engineers Report TR EL-82-4, 14 p. Ryland, J. S. 1964. The feeding of plaice and sand-eel larvae in the south- ern North Sea. J. Mar. Biol. Assoc. U.K. 44:343-364. Shelbourne, J. E. 1953. The feeding habits of plaice post-larvae in the South- ern Bight. J. Mar. Biol. Assoc. U.K. 32:149-159. Sherman, K. 1980. MARMAP, a fisheries ecosystem study in the North- west Atlantic:fluctuations in ichthyoplankton-zooplankton components and their potential for impact on the system. In F. P. Diemer, F. J. Vernberg, and D. R. Mirkes (eds.), Advanced concepts in ocean measurements for ma- rine biology, p. 9-37. Belle W. Baruch Library in Marine Science 10, Univ. South Carolina Press, Columbia, SC. Smith, W. G. 1973. The distribution of summer flounder, Paralichthys den - tatus, eggs and larvae on the continental shelf between Cape Cod and Cape Lookout, 1965-66. Fish. Bull. 71:527-548. Smith, W. G., and M. P. Fahay. 1970. Description of eggs and larvae of the summer floun- der, Paralichthys dentatus. U.S. Fish and Wildl. Serv. Res. Rep. 75, 21 p. Smith, W. G., J. D. Sibunka, and A. Wells. 1978. Diel movements of larval yellowtail flounder, Limanda ferruginea , determined from discrete depth sampling. Fish. Bull. 76:167-178. Szedlmayer, S. T., K. W. Able, and R. A. Rountree. 1992. Growth and temperature induced mortality of young- of-the-year summer flounder (Paralichthys dentatus) in southern New Jersey. Copeia 1992:120-128. Thorisson, K. 1994. Is metamorphosis a critical interval in the early life of marine fishes? Environ. Biol. Fish. 40:23-36. van der Veer, H. W., and M. J. N. Bergman. 1987. Predation by crustaceans on a newly settled 0-group plaice Pleuronectes platessa population in the western Wadden Sea. Mar. Ecol. Prog. Ser. 35:203-215. Witting, D. A., and K. W. Able. 1993. Effects of body size on probability of predation for juvenile summer and winter flounder based on laboratory experiments. Fish. Bull. 91:577-581. 1995. Predation by sevenspine bay shrimp Crangon septem- spinosa on winter flounder Pleuronectes americanus dur- ing settlement: laboratory observations. Mar. Ecol. Prog. Ser. 123:23-31. Wyatt, T. 1972. Some effects of food density on the growth and behaviour of plaice larvae. Mar. Biol. 14:210-216. 258 Age composition, growth, reproductive biology, and recruitment of King George whiting, Sillaginodes punctata, in coastal waters of southwestern Australia Glenn A. Hyndes Margaret E. Platell lan C. Potter Biological Sciences, Murdoch University Murdoch, Western Australia, 6 1 50, Australia E-mail address (for G. A. Hyndes): hyndes@central.murdoch.edu.au Rodney C. J. Lenanton Bernard Bowen Fisheries Research Institute Western Australian Marine Research Laboratories RO. Box 20, North Beach Western Australia, 6020, Australia Abstract , —The age structure, growth and reproductive biology have been determined for the recreationally and commercially important King George whiting, Sillaginodes punctata, off southwestern Australia. The maxi- mum lengths and ages, asymptotic lengths (LJ, and growth coefficients (K) were 596 mm, 14 years, 532 mm, and 0.47, respectively, for females, and 555 mm, 13 years, 500 mm and 0.53, respectively, for males. Sexual maturity is attained by 50% of female S. punctata by ca. 410 mm in length, and by the ma- jority of both female and male fish at the end of their fourth year of life. The monthly trends in the proportions of mature gonads and the prevalence of different oocyte stages and postovu- latory follicles indicated that, in south- western Australia, S. punctata spawns from June to September. Spawning is thus initiated when water tempera- tures are declining from their maxima. During the spawning period, many of the ovaries of large fish contained yolk vesicle and yolk granule oocytes, as well as hydrated oocytes or postovulatory follicles (or both), indicating that S. punctata is a multiple spawner. Fur- thermore, because hydrated oocytes or postovulatory follicles were often found together with large numbers of yolk granule oocytes, S. punctata presum- ably releases eggs in batches during the spawning period. Recruitment of S. punctata into sheltered nearshore wa- ters (<1.5 m) commences in late Sep- tember, three months after the onset of spawning, and continues until early November. When juvenile S. punctata reach ca. 1.5 years of age and ca. 250 mm, the legal minimum length for cap- ture, they move out into slightly deeper waters (2-6 m) in marine embayments and estuaries. After attaining ages of ca. 4 years and lengths of ca. 370 mm, they then migrate from these waters, where fishing pressure is greatest, into regions near or around reefs at depths of 6-50 m, where spawning occurs. In contrast to S. punctata , the five other whiting species in southwestern Aus- tralian waters, which all belong to the genus Sillago, spawn between late spring and early autumn. In the case of the three Sillago species that undergo an offshore migration, this movement occurs at a relatively small size and young age and leads to their occupying open sandy areas. The implications of S. punctata habitat and biological data for fishery management are discussed. Manuscript accepted 18 July 1997. Fishery Bulletin 96:258-270 (1998). The King George whiting, Sillagi- nodes punctata , which is the larg- est of the 31 species belonging to the Sillaginidae, occurs along the lower west and southern coasts of Austra- lia (McKay, 1992), where it is a very important recreational and com- mercial species (see Kailola et al., 1993). Although a number of stud- ies have been carried out on the bi- ology of S. punctata (Scott, 1954; Thomson, 1957a; Gilmour, 1969; Robertson, 1977; Caton1; Cockrum and Jones2), these have concen- trated mainly on fish caught in shal- low waters, in which mature fish tend not to be found. Although Cockrum and Jones2 obtained S. punctata from deeper waters, where spawning presumably occurs, data on the pattern of gonadal develop- ment, which is required to deter- mine precisely the peak time and duration of spawning of this species, are very limited. Indeed, estimates of the timing and duration of spawn- ing of S. punctata have been derived predominantly from backcalcula- tions of daily growth increments of recently settled juveniles (Bruce, 1989; Bruce and Short, 1992; Jen- kins and May, 1994; Fowler and Short, 1996). These latter studies, which were carried out off south- eastern Australia, indicate that S. punctata spawns between early au- tumn and early winter. However, S. punctata is recruited into nearshore waters far earlier off South Austra- lia than off southwestern Australia, which suggests that this species spawns later in the latter region (cf. Fowler and Short, 1996; Hyndes et al., 1996a). Although age structures and growth parameters have been esti- mated for populations of 300 mm, were obtained from recreational anglers, who were fishing between February 1993 and August 1996 in waters <50 m in depth between 31°55'S and 32°45'S. These anglers kindly supplied frozen car- casses, including gonads, after they had been filleted, together with a record of the date, location, and depth at which each fish had been captured. The total length (TL) of each fish was measured to the nearest 1 mm. When a gonad could be identified 3 Jones, G. K„ D. A. Hall, K. L. Hill, and A. J. Stamford. 1990. The South Australian marine scalefish fishery stock assessment; economics; management. South Australian Department of Fisheries Green Paper, January 1990, 186 p. as either an ovary or a testis, it was assigned to one of eight developmental stages, according to the cri- teria of Laevastu (1965): I - virgin; II = maturing virgin; III - developing; IV = maturing; V = mature; VI - spawning; VII = spent; and VIII = recovering or spent. However, multiple spawning by representa- tives of this species in each spawning season (see “Results” section) meant that it was often difficult to ascertain whether certain gonads should be recorded as stages V, VI, or VII. The data for these three stages, subsequently referred to collectively as mature go- nads, were therefore pooled. It was not possible to recognize stage VIII gonads in males. Ovaries of the large female S. punctata collected in each month of the study were placed for 24 h in Bouin’s fixative and then dehydrated in a series of ethanols. The midregion of each of these ovaries was embedded in paraffin wax, cut transversely (6 pm) and stained with Mallory’s trichrome. The circumferences of 30 oocytes, where the section passed through the nucleus, were recorded to the nearest 5 pm by using the OPTIMAS (OPTIMAS Corp., 1994) computer imaging package. This enabled the diameters of these oocytes to be calculated. Because hydrated oocytes had collapsed as a result of freezing, their circumferences could not be measured. The oocyte measurements for fish in each calendar month during the three years of this study were pooled in order to provide histological data for up to 10 ovaries from each of those months. The terminology for the oocyte stages was adapted from that given in Khoo ( 1979). The length at which 50% of female S. punctata first attained maturity (L50) was calculated by fitting a logistic function to the proportion of mature fish in each 50-mm length interval in the spawning period of June to September (see “Results” section) by a nonlinear technique (Saila et al., 1988) with a non- linear subroutine in SPSS (SPSS Inc., 1994). The lo- gistic equation is PL =[l + e(a+bL)]~1 where PL - the proportion of fish with mature go- nads at length interval L; and a and b = constants. The L50 was then derived from the equation L50 = —ab Preliminary examination of scales and sagittal otoliths collected from large S. punctata caught dur- ing the initial stages of the study showed that growth zones were often difficult to detect in scales, whereas at least some growth zones could be clearly discerned 260 Fishery Bulletin 96(2), 1 998 in whole otoliths. For this reason, sagittal otoliths were subsequently removed from each fish, cleaned, dried, stored, and used to estimate the ages of fish. In order to determine whether it was necessary to section otoliths to reveal all translucent zones, the number of translucent zones in the otoliths of 100 randomly selected 500 mm (Fig. 4A). The length at which 50% of female S. punctata were mature, represented by the L50, was 413 mm. Maturity was first attained by fe- males at the end of their third year of life, when ca. 15% offish possessed mature ovaries (Fig. 4B). How- ever, by the end of the fourth year of life, the propor- tion of mature females had increased markedly to 72% and in subsequent years to 100%. A similar pat- tern was shown by the maturity stages for male fish (data not shown). Gonadal and oocyte development Because the vast majority ofS. punctata did not reach maturity until the end of their fourth year of life, the following monthly trends shown by gonadal stages were derived from fish that were >3.4 years old and thus these fish were expected to reach maturity in the following spawning period. Between January and May, the ovaries of all female S. punctata >3.4 years old were at either stages III or IV (Fig. 5). Mature ovaries were found in over 70% of females by June and in 90 and 100% by July and August, respectively. The proportion of mature ovaries declined markedly to 43% in September, whereas the contribution of 262 Fishery Bulletin 96(2), 1 998 recovering or spent ovaries (stage VIII) increased from 9 to 48% between August and September (Fig. 5). Female fish with mature ovaries were virtually absent by November, and those with stage-VIII ova- ries represented only ca. 8% of the fish caught both in this month and in December. The gonadal devel- opment of males followed a similar trend to that of females (Fig. 5). In each month, the oocyte diameters for S. punctata exhibited a well-defined mode between 50 and 80 pm (Fig. 6), which represents the perinuclear oo- cytes. Yolk vesicle and yolk granule oo- cytes first appeared in ovaries in April, when the diameters of these larger oo- cytes ranged between 255 and 430 pm. However, many other yolk vesicle and yolk granule oocytes were undergoing atresia. In the following month, the maximum oocyte diameter declined to 150 pm, reflecting the fact that no yolk vesicle or yolk granule oocytes were present. In June, the maximum dia- meter increased to 465 pm (Fig. 6), as a result of the development of yolk vesicle and yolk granule oocytes. These large oocytes were abundant between June and August, and many ovaries were dominated by yolk granule oocytes. The proportion of yolk vesicle and yolk gran- ule oocytes declined in September, and the few remaining yolk vesicle and yolk gran- ule oocytes that were present in October were at an advanced stage of atresia. Hydrated oocytes that had collapsed during sectioning were found in large numbers in some ovaries between June and August. Because the ovaries in those months sometimes contained large numbers of postovulatory follicles, they had already discharged hydrated oocytes. Furthermore, hydrated oocytes and postovulatory follicles were occa- sionally found in the same ovary. Juvenile recruitment and depth distribution A 30 - I I I I I I I 0 100 200 300 400 500 600 B ; 1 Nearshore waters Embayments & estuaries I — ^ — I Offshore waters I — 1 i i i i i i i 0 100 200 300 400 500 600 Total length (mm) Figure 2 (A) Frequency histograms for total lengths of the different age classes of Sillaginodes punctata, from data obtained from samples collected by the 21.5-m seine net in nearshore marine waters and from anglers in deeper waters. Sample sizes are given in parentheses. (B) Median, 90 percen- tiles, and range of lengths for fish caught in shallow nearshore marine waters, deeper waters in marine embayments and estuaries, and in deeper and more offshore marine waters around reefs. November-January IL L3 (132) SSSSSk* ►», 40 1" February-April (169) Age class H 0+ □ i + 2+ ■ 3+ m 4+ Hi 5+ El 6-14+ The new 0+ recruits of S. punctata were first caught in late September 1994. The minimum standard length of these fish remained at ca. 14 mm between 23 Sep- tember and 3 November, before increas- ing to ca. 30 mm between mid-November and mid-December (Fig. 7). The maximum length of these new recruits increased progressively from 26 mm in late Septem- ber to 64 mm by mid-December (Fig. 7). Hyndes et al.: Age composition, growth, reproductive biology, and recruitment of Sillaginodes punctata 263 Age (years) Figure 3 Von Bertalanffy growth curves fitted to total length-at-age data de- rived from sagittal otoliths of female and male Sillaginodes punctata. Although the vast majority of S. punctata caught by anglers in embayments and estu- aries at depths of 2-6 m were <370 mm in length and <4 years of age, those that were collected outside embayments and estuaries and in the vicinity of reefs at depths of 6-50 m were predominantly greater than this length and age (Fig. 8, A and B). Furthermore, S. punctata that contained mature (stages V- VII) or recovering or spent (stage VIII) go- nads were caught predominantly in and around reefs and at the edges of seagrass beds adjacent to these areas, where water depths exceeded 6 m (Fig. 8). Discussion Sexual maturity and spawning period Gonadal data indicate that, in southwestern Australia, female S. punctata reach matu- rity at the end of their fourth year of life or when they have attained a length of ca. 410 mm. This length far exceeds the 350 mm at which 50% of the members of this species reach maturity in southeastern Australian waters (see Cockrum and Jones2). However, the data of Cockrum and Jones2 suggest that the length at maturity of S. punctata in southeastern Australia has declined since the 1950’s as a result of fishing pressure, a trend that has been observed with increased fishing pressure in other species of teleosts (Wootton, 1990). Such a conclusion would be consistent with the fact that in southwest- ern Australia, where S. punctata is not at present as heavily exploited, the length at maturity (ca. 410 mm) is similar to that re- corded in South Australia in the 1950’s (Scott, 1954; Cockrum and Jones2). Because large numbers of yolk granule and hydrated oocytes and postovulatory follicles were first found in the ovaries of large female S. punctata in June and were prevalent in ovaries through Septem- ber, we conclude that this species spawns during this four-month period between early winter and early spring. Because the advanced oocytes present in ova- ries in October and November were usually under- going atresia, the spawning period of S. punctata rarely extends beyond September. The spawning of (LJ Wm) Spawning Species F M F M F M F M F M Nursery grounds grounds Sillaginodes punctata 532 500 0.47 0.53 14 13 400 400 4 4 Sheltered nearshore Reefs Sillago bassensis 329 307 0.26 0.29 7 9 200 200 3 3 Exposed nearshore Deep offshore Sillago burrus 188 179 2.37 2.44 4 4 130 120 1 1 Sheltered nearshore Shallow offshore Sillago robusta 169 172 1.03 0.98 6 5 150 140 2 2 Offshore Deep offshore Sillago schomburgkii 333 325 0.52 0.53 7 7 200 180 2 2 Sheltered nearshore Nearshore Sillago vittata 331 312 0.43 0.45 7 6 140 130 1 1 Sheltered nearshore Shallow offshore Females 25 4 29 II 5 II 10 II 21 13 34 12 Gonad stage ii □ ill m rv EE v-vn IB vm Month Figure 5 Monthly percent frequencies of occurrence of sequential gonadal development stages in female and male Sillaginodes punctata >3.4 years old. Numbers indicate sample sizes. that this size range of S. punctata was consis- tently observed in its specimens from near- shore waters between late September and early November (midspring), indicates that this species settles in its nearshore nursery grounds predominantly during this period. Thus, the first recruits of S. punctata enter those nearshore waters approximately three months after spawning is initiated, a time that corresponds to the time that the larvae of this species take to recruit into their nursery ar- eas in South Australia (Bruce, 1989; Bruce and Short, 1992; Fowler and Short, 1996). Recruit- ment of juvenile S. punctata into the shallows commences far earlier in South Australian than in southwestern Australian waters, i.e. June vs. September, reflecting an earlier start to the spawning period, i.e. March vs. June (cf. Bruce, 1989; Fowler and Short, 1996). However, because the juveniles that are re- cruited into nursery habitats much farther east in Victoria are derived from a spawning that occurs in South Australia (Jenkins and Black, 1994; Jenkins and May, 1994), their larvae have to travel a far longer distance to- wards their nursery grounds than those re- cruited into nearshore waters of South Austra- lia; thus recruitment commences far later in Victoria, i.e. September vs. June (Brace, 1989; Jenkins and May, 1994; Fowler and Short, 1996). After settlement, large numbers of S. punctata remain in the sheltered nearshore marine waters of the lower west coast for about 1.5 years (Hyndes et al., 1996a). The fact that the densities of juveniles were far higher in these waters than in nearby exposed waters and even the shallows of estuaries (Hyndes et al., 1996a) indicates that the juveniles of this species prefer sheltered nearshore marine habitats. This conclusion is sup- 266 Fishery Bulletin 96(2), 1 998 ported by the fact that juveniles of S. punctata are absent in nearshore waters along the southern coast- line of southwestern Australia (Lenanton, 1982; Ayvazian and Hyndes, 1995), where the shoreline is more exposed to rough sea conditions (Hegge et al., 1996). The relative paucity of sheltered nearshore marine habitats on this coastline would account for the relatively high densities of juvenile S. punctata that are found in the relatively protected waters of estuaries in this region (Potter et al., 1993; Potter and Hyndes, 1994). Estuaries along the southern coast of southwest- ern Australia, where the adjacent marine waters are exposed to wave and swell activity, may thus pro- vide particularly important nursery habitats for S. punctata. Many estuaries in this region, however, become closed off from the sea during the summer and autumn, when, as a result of low freshwater dis- charge, a sand bar forms at their mouths (Lenanton and Hodgkin, 1985). It is thus relevant that S. punc- tata spawns during winter, because this would en- able juveniles to enter those estuaries before their mouths become closed. This event parallels the situ- ation with the mugilids Mugil cephalus and xi S 3 Z 23 September 7 October 20 * (100) 10 0 20 ■ (100) 10 0 20 ' (100) 10 0 21 October 20 r 10 - 0 - 20 r 10 - 0 - (100) (52) 0 10 20 5 December _i i i i i 30 40 50 60 70 Standard length (mm) Figure 7 Frequency histograms for standard lengths of Sillaginodes punctata caught at two weekly intervals by the 5.5-m seine net in sheltered nearshore marine waters between Sep- tember and December 1994. The numbers in parentheses represent the number of fish measured. Hyndes et a I.: Age composition, growth, reproductive biology, and recruitment of Sillaginodes punctata 267 -S 00 c o H 650 600 550 500 450 400 350 300 - 250 - 200 ■ u a • . p □ □ • □ I . oh "□ b . p ° ■.s'j !”» i< lli'i j I if': ■ i ;'S Gonad stage ■ II -IV □ v-vni Figure 8 (A) Total lengths and (B) ages at capture, of female and male Sillaginodes punctata with immature (stages II— IV ) and mature or recovering or spent (or both) (stages V-VIII) gonads collected at different depths by recreational fishers in the inner-shelf waters of the lower west coast of Australia. Aldrichetta forsteri, which also both spawn in winter in southwestern Australia and whose juveniles soon after enter estuaries in large numbers (Thomson, 1957b; Chubb et al., 1981). Sillaginodes punctata typically remains in sheltered nearshore marine and estuarine waters until it reaches ca. 250 mm in length and ca. 1.5 years of age (Robertson, 1977; Pot- ter et al., 1983; Loneragan et al., 1989; Potter et al., 1993; Potter and Hyndes, 1994; Hyndes et al., 1996a). Thus, unlike smaller species of whiting, such as Sillago burrus and Sillago vittata, which typically move offshore at lengths of ca. 100 mm and at ages of three to nine months (Hyndes et al., 1996a, 1996b), S. punctata remains in its nursery habitats for a far longer period. Because S. burrus and S. vittata reach maturity at the end of their first year of life, whereas most S. punctata first at- tain maturity at the end of their fourth year of life, this last species takes advantage of the pro- ductive waters of marine embayments or estu- aries for a longer period before migrating out into spawning grounds in deeper marine waters. Because the nursery areas of S. punctata are located initially in nearshore waters and subsequently in the deeper waters of marine embayments and estuaries, whereas adults occupy areas in and around reefs where wa- ter depths are generally greater, this species migrates offshore from shallow nursery grounds as it approaches the size and age at which it will become mature. A movement off- shore as body size increases parallels that re- corded for populations of S. punctata in south- eastern Australia and for other whiting spe- cies (Scott, 1954; Gilmour, 1969; Robertson, 1977; Weng, 1986; Burchmore et al., 1988; Hyndes et al., 1996a; Caton1). Although an- glers often caught S. punctata in deeper wa- ters near reefs or at the edges of adjacent seagrass beds, they rarely caught them during ex- tensive trawling over the expansive open, sandy ar- eas of the same region where large numbers of other whiting species were caught (Hyndes et al., 1996a). The presence of mature and recovering or spent go- nads in the larger members of S. punctata caught in areas around reefs also suggests that, unlike the three whiting species Sillago burrus, S. vittata, and S. bassensis, which migrate out into the more open and sandy areas of the inner shelf to spawn (Table 2; Hyndes et al., 1996b; Hyndes and Potter, 1996), S. punctata spawns in areas in and around reefs or at the edges of adjacent seagrass beds. Furthermore, whereas S. burrus and S. vittata move into spawn- ing grounds ranging from 5-15 m in depth, and S. bassensis migrates into more offshore waters where the depth is 20-35 m deep (Hyndes et al., 1996a, 1996b; Hyndes and Potter, 1996), S. punctata spawns at depths that range from six to at least 50 m. Implications for management Because S. punctata exhibits size-related offshore movements, that involve firstly a movement from nearshore nursery areas to deeper waters in marine embayments and estuaries and subsequently a move- 268 Fishery Bulletin 96(2), 1 998 ment into areas around reefs farther offshore (Fig. 2B), this species essentially occupies three different habitats during the course of its life cycle. Any man- agement plans for S. puncata must therefore take into account the need both to protect these habitats and to ensure that fishing pressure in those habi- tats where fishing occurs does not have a deleteri- ous effect on the stock of this species. The postlarvae of S. punctata settle predominantly in very sheltered nearshore waters of marine embay- ments and estuaries (Jenkins and May, 1994; Fowler and Short, 1996; Hyndes et al., 1996a; Hyndes, unpubl. data). However, S. punctata moves out into slightly deeper waters (2-10 m) in marine embay - ments and estuaries at ca. 1.5 years in age and at a total length of 250 mm (Fig. 2B). Because this latter length corresponds to the legal minimum length for capture (LML) of S. punctata on the lower west coast and approaches the LML of 280 mm on the south coast, this species does not become commercially and recreationally exploited until it has left its very shal- low nursery areas and has entered deeper waters. This species remains in these slightly deeper waters until it reaches 350-400 mm (Fig. 2B) and is thus available for capture in marine embayments and estuaries when it is predominantly between 1.5 and 2.5 years in age. Because this species is most heavily exploited when it is in the deeper waters of marine embayments and estuaries, it is fished mainly dur- ing this relatively restricted period of its life cycle. From a management point of view, it is also relevant that the fishery in marine embayments and estuar- ies is based on fish that have not yet reached 410 mm, the length at which they typically first become ma- ture. Subsequently, those S. punctata that have run the “gauntlet” of numerous fishermen in marine embayments and estuaries move farther out into areas in and around reefs in deeper waters where they attain maturity and where the reduced num- ber of fishermen targeting this species makes it less susceptible to capture. Furthermore, the catches of S. punctata in the more offshore waters are further reduced in winter, when sea conditions are far less favorable for fishing and when, according to fisher- men, S. punctata are less likely to take bait. Thus, because S. punctata spawns during winter, fishing pressure on this species is relatively low during the spawning period. Because catch and effort statistics for the commer- cial fishery in southwestern Australia do not sug- gest that the catch rate for S. punctata is declining, this species would not appear currently to be over- exploited in this region. However, the number of shel- tered nearshore areas that act as nursery areas for S. punctata are limited and are often located in the type of region where marinas and other developments are likely to be proposed. Furthermore, recreational fishing effort is rising markedly, and the increasing use of larger vessels and more sophisticated equip- ment, such as global positioning systems (GPS), means that, particularly in deeper waters, fish are now beginning to be exploited to a greater extent. The resultant advances in fishing efficiency also mean that more sophisticated measures of effort are required to obtain reliable and comparable catch-per- unit-of-effort data for S. punctata from the recre- ational sector, which is ultimately expected to be- come the main exploiter of this resource. Acknowledgments We thank the large number of people who helped with sampling and the numerous recreational anglers who provided fish carcasses, particularly Ron Robinson and Doug Clegg. Gratitude is also expressed to Gor- don Thomson for sectioning ovaries, to Alex Hesp for sectioning otoliths, and also to Keith Jones, Tony Fowler, Norm Hall, and Greg Jenkins for construc- tive comments on the manuscript. Financial support was provided by the Western Australian Fisheries Department, Australian Fisheries Research and De- velopment Corporation, and Murdoch University. Literature cited Ayvazian, S. G., and G. A. Hyndes. 1995. Surf-zone fish assemblages in south-western Austra- lia: do adjacent nearshore habitats and the warm Leeuwin Current influence the characteristics of the fish fauna? Mar. Biol. 122: 527-536. Beamish, R. J., and G. A. McFarlane. 1987. Current trends in age determination metho- dology. In R. C. Summerfelt and G. E. Hall (eds), The age and growth of fish, p. 15-42. Iowa State Univ. Press, IA. Beverton, R. J. H. 1992. Patterns of reproductive strategy parameters in some marine teleost fishes. J. Fish Biol, (suppl. B.) 41:137-160. Bruce, B. D. 1989. Studying larval fish ecology. Safish 13:4-9. 1995. Larval development of King George whiting, Sillagi- nodes punctata, school whiting, Sillago bassensis, and yellow fin whiting, Sillago schomburgkii (Percoidei: Sillaginidae), from South Australian waters. Fish. Bull. 93:27-43. Bruce, B. D., and D. A. Short. 1992. Observations on the distribution of larval fish in re- lation to a frontal zone at the mouth of Spencer Gulf, South Australia. In D. A. Hancock (ed.), Larval biology, Austra- lian Society for Fish Biology workshop, Hobart, 20 August 1991, p. 124-137. Bureau of Rural Resources Proceedings 15, Bureau of Rural Resources, Canberra. Burchmore, J. J., D. A. Pollard, M. J. Middleton, J. D. Bell, and B. C. Pease. 1988. Biology offour species of whiting (Pisces: Sillaginidae) Hyndes et a\. : Age composition, growth, reproductive biology, and recruitment of Sillaginodes punctata 269 in Botany Bay, New South Wales. Aust. J. Mar. Freshwa- ter Res. 39:709-727. Campana, S. E. 1984. Comparison of age determination methods for the starry flounder. Trans. Am. Fish. Soc. 113:365-369. Casselman, J. M. 1987. Determination of age and growth. In A. H. Weatherly and H. S. Gill (eds), The biology of fish growth, p. 209- 242. Academic Press, London. Chubb, C. F., I. C. Potter, C. J. Grant, R. C. J. Lenanton, and J. Wallace. 1981. The age structure, growth rates and movements of sea mullet, Mugil cephalus L., and yellow-eye mullet, Aldrichetta forsteri (Valenciennes), in the Swan-Avon River system, Western Australia. Aust. J. Mar. Freshwater Res. 32:605-628. Conover, D. O. 1992. Seasonality and the scheduling of life history at dif- ferent latitudes. J. Fish Biol. 41:161-178. de Vlaming, V. L. 1983. Oocyte development patterns and hormonal involve- ments among teleosts. In J. C. Rankin, T. J. Pitcher, and R. T. Duggan (eds), Control processes in fish physiology, p. 176-199. Croom Helm, Beckenham. Fowler, A. J., and D. A. Short. 1996. Temporal variation in the early life-history charac- teristics of King George whiting ( Sillaginodes punctata) from analysis of otolith microstructure. Mar. Freshwater Res. 47:809-818. Gallucci, V. F., and T. J. Quinn. 1979. Reparameterizing, fitting and testing a simple growth model. Trans. Am. Fish. Soc. 108:14-25. Gilmour, A. J. 1969. The ecology of the King George Whiting Sillaginodes punctatus (Cuvier and Valenciennes) in Westernport Bay, Victoria. Ph.D. diss., Monash Univ., Victoria, 372 p. Hegge, B., I. Eliot, and J. Hsu. 1996. Sheltered sandy beaches of southwestern Australia. J. Coastal Res. 12:748-760. Hyndes, G. A., and I. C. Potter. 1996. Comparisons between the age structures, growth and reproductive biology of two co-occurring sillaginids, Sillago robusta and S. bassensis, in temperate coastal waters of Australia. J. Fish Biol. 49:14-32. 1997. Age, growth and reproduction of Sillago schomburgkii in south-western Australian, nearshore waters and com- parisons of life history styles of a suite of Sillago species. Environ. Biol. Fishes 49:425-447. Hyndes, G. A., N. R. Loneragan, and I. C. Potter. 1992a. Influence of sectioning otoliths on marginal incre- ment trends and age and growth estimates for the flat- head Platycephalus speculator. Fish. Bull. 90:276-284. Hyndes, G. A., F. J. Neira, and I. C. Potter. 1992b. Reproductive biology and early life history of the marine teleost Platycephalus speculator Klunzinger (Platycephalidae) in a temperate Australian estuary. J. Fish Biol. 40:859-874. Hyndes, G. A., I. C. Potter and R. C. J. Lenanton. 1996a. Habitat partitioning by whiting species (Sillag- inidae) in coastal waters. Environ. Biol. Fish. 45: 21-40. Hyndes, G. A., I. C. Potter, and S. A. Hesp. 1996b. Relationships between the movements, growth, age structures and reproductive biology of the teleosts Sillago burrus and Sillago vittata in temperate marine waters. Mar. Biol. 126:549-558. Jenkins, G. P., and K. P. Black. 1994. Temporal variability in settlement of a coastal fish (Sillaginodes punctata) determined by low-frequency hydrodynamics. Limnol. Oceanogr. 39:1744-1754. Jenkins, G. P., and H. M. A. May. 1994. Variation in larval duration and settlement of King George whiting, Sillaginodes punctata , in Swan Bay, Victoria. Bull. Mar. Sci. 54:281-296. Jones, K. 1980. Research on the biology of spotted (King George) whiting in South Australian waters. Safish 4:3-7. Kailola, P. J., M. J. Williams, P. C. Stewart, R. E. Reichelt, A. McNee, and C. Grieve. 1993. Australian fisheries resources. Bureau of Resource Sciences, Canberra, 422 p. Khoo, K. H. 1979. The histochemistry and endocrine control of vitello- genesis in goldfish ovaries. Can. J. Zool. 57:617-626. Kimura, D. K. 1980. Likelihood methods for the von Bertalanffy growth curve. Fish. Bull. 77:765-776. Laevastu, T. 1965. Manual of methods in fisheries biology. FAO, Rome, 51 p. Lenanton, R. C. J. 1982. Alternative non-estuarine nursery habitats for some commercially and recreationally important fish of south- western Australia. Aust. J. Mar. Freshwater Res. 33:881- 900. Lenanton, R. C. J., and E. P. Hodgkin. 1 985. Life history strategies of fish in some temperate Aus- tralian estuaries. In A. Yanez-Arancibia (ed.), Fish com- munity ecology in estuaries and coastal lagoons: towards an ecosystem integration, p. 267-283. Universidad Nacional Autonoma de Mexico, Mexico. Loneragan, N. R., I. C. Potter, and R. C. J. Lenanton. 1989. Influence of site, season and year on contributions made by marine, estuarine, diadromous and freshwater species to the fish fauna of a temperate Australian estuary. Mar. Biol. 103:461-479. McKay, R. J. 1992. Sillaginid fishes of the world: an annotated and il- lustrated catalogue of the Sillago, smelt or Indo-Pacific whiting species known to date. FAO Fisheries Synopsis 125, FAO, Rome, 87 p. Morton, R. M. 1985. The reproductive biology of summer whiting, Sillago ciliata C. and V., in northern Moreton Bay, Queensland. Aust. Zool. 121:491-502. Optimas Corp. 1994. Optimas 5 user guide, version 5.1a. Optimas Corp., Seattle, WA, 340 p. Potter, I. C., and G. A. Hyndes. 1994. The composition of the fish fauna of a permanently open estuary on the southern coast of Australia, and com- parisons with a nearby seasonally closed estuary. Mar. Biol. 121:199-209. Potter, I. C., G. A. Hyndes, and F. M. Baronie. 1993. The fish fauna of a seasonally closed Australian es- tuary. Is the prevalence of estuarine-spawning high? Mar. Biol. 116:19-30. Potter, I. C., N. R. Loneragan, R. C. J. Lenanton, P. J. Chrystal, and C. J. Grant. 1983. Abundance, distribution and age structure of fish populations in a Western Australian estuary. J. Zool. (Lond.) 200:21-50. 270 Fishery Bulletin 96(2), 1998 Robertson, A. I. 1977. Ecology of juvenile King George whiting Sillaginodes punctatus (Cuvier and Valenciennes) (Pisces: Perciformes) in Western Port, Victoria. Aust. J. Mar. Freshwater Res. 28:35-43. Saila, S. B., C. W. Recksieck, and M. H. Prager. 1988. Basic fisheries science programs. Elsevier, New York, NY, 230 p. Sarre, G. A., G. A. Hyndes, and I. C. Potter. 1997. Habitat, reproductive biology and size composition of Parequula melbournensis, a gerreid with a temperate distribution. J. Fish Biol. 50:341—357. Scott, T. D. 1954. The life history of the spotted whiting, Sillaginodes punctatus (Cuv. and Val.) in South Australia. M.S. the- sis, Univ. Adelaide, Adelaide, 78 p. SPSS Inc. 1994. SPSS for Windows™. SPSS Inc., Chicago, IL, 828 p. Thomson, J. M. 1957a. The size at maturity and spawning times of some Western Australian estuarine fish. Fish. Bull. W. Aust. 7:3-8. 1957b. Biological studies of economic significance of the yellow-eye mullet, Aldrichetta forsteri (Cuvier and Valen- ciennes) (Mugilidae). Aust. J. Mar. Freshwater Res. 8: 1-13. Vouglitois, J. J., K. W. Able, R. J. Kurtz, and K. A. Tighe. 1987. Life history and population dynamics of the bay an- chovy in New Jersey. Trans. Am. Fish. Soc. 116:141-153. Weng, H. T. 1986. Spatial and temporal distribution of whiting (Sillaginidae) in Moreton Bay, Queensland. J. Fish Biol. 29:755-764. Wootton, R. J. 1990. Ecology of teleost fishes. Chapman and Hall, Lon- don, 404 p. 271 Estimates of marine mammal, turtle, and seabird mortality for two California gillnet fisheries: 1 990-1 995 Abstract .—Incidental kills of ma- rine mammals, turtles, and seabirds are estimated for the California drift gillnet fishery for broadbill swordfish, Xiphias gladius, common thresher shark, Alopias vulpinus, and shortfin mako shark, Isurus oxyrinchus, and the set gillnet fishery for California halibut, Paralichthys californicus, and Pacific angel shark, Squatina californica, for the period July 1990 through Decem- ber 1995. Estimates were based on ob- servations made by National Marine Fisheries Service observers placed aboard commercial fishing vessels. Yearly observer coverage varied be- tween 4% and 18% of estimated total effort. Total fishing effort-days per Cali- fornia Department of Fish and Game fishing block was used as the measure of effort for the drift and set gillnet fish- eries. Incidental kill was estimated from observed data and estimates of total effort by using mean-per-unit and ratio estimators. Additional bycatch data collected by NMFS observers were used to derive kill estimates of marine turtles and seabirds. In the drift gillnet fishery, seven out of 387 mammals observed entangled were released alive. In the set gillnet fishery, five out of 1,263 mammals ob- served entangled were released alive. Estimates of incidental kill are pre- sented along with estimates of en- tanglement for species that were ob- served to be released alive. For the pe- riod under consideration, the estimated mortality for the drift gillnet fishery was over 450 marine mammals each year. A total of 20 turtles and 3 seabirds were observed entangled during the entire period. The most frequently en- tangled species in this fishery were common dolphins, Delphinus spp., and northern elephant seals, Mirounga angustirostris. Estimated cetacean mortality in the driftnet fishery de- creased from 650 in 1991 to 417 in 1995; pinniped mortality decreased from 173 in 1991 to 116 in 1995. Estimated ceta- cean mortality in the set gillnet fish- ery ranged from a high of 38 in 1991 to a low 14 in 1993; pinniped mortality rose to a high of 4,777 in 1992 and then decreased to 1,016 in 1995. We postu- late that there has been a decline in the number of pinnipeds and cetaceans in the setnet fishery owing to area closure. No similar proposal can be made for the driftnet fishery. The most frequently entangled mammals in the setnet fish- ery were California sea lions, Zalophus californianus, and harbor seals, Phoca vitulina. Six turtles and 1,018 seabirds were estimated entangled in this fishery during the NMFS Observer Program from July 1990 to December 1995. Manuscript accepted 25 June 1997. Fishery Bulletin 96:271-284 (1998). Fred Julian 1 42 Sierra Way Chula Vista, California 91911 E-mail address: fjulian@osiris.ucsd.edu Marilyn Beeson California Department of Fish and Game 330 Golden Shore, Suite 50 Long Beach, California 90802 Two major gillnet fisheries in Cali- fornia are known to kill marine mammals, turtles, and seabirds in- cidentally: the drift gillnet fishery for broadbill swordfish, Xiphias gladius, common thresher shark, Alopias vulpinus, and shortfin mako shark, Isurus oxyrinchus, and the set gillnet fishery for California halibut, Paralichthys californicus, and Pacific angel shark, Squatina californica. Historically, concern was focused on incidental kill of sea- birds, sea otters, Enhydra lutris, harbor porpoise, Phocoena phocoena, harbor seals, Phoca vitulina, and California sea lions, Zalophus californianus, in the setnet fishery (Salzman, 1989; Jefferson et al., 1994; Diamond and Hanan1 2 3; Hanan et al.2,3; Hanan and Diamond4). In recent years the driftnet fishery has received more attention because it interacts with more cetaceans (Bar- low et al., 1994; Lennert et al., 1994). Estimation for the driftnet fishery has been possible because, in July 1990, the National Marine Fisheries Service (NMFS) imple- mented an observer program to monitor the marine mammal by- catch. Complementing the observer program was a project by the Cali- fornia Department of Fish and Game (CDFG) to develop estimates of total effort in the drift and set gillnet fisheries. Results from these programs were used to estimate incidental kill with stratified ratio and mean-per-unit estimation 1 Diamond, S. L., and D. A. Hanan. 1986. An estimate of harbor porpoise mortality in California set net fisheries, April 1. 1983 through March 31, 1984. NOAA/NMFS SWR Admin. Rep. SWR-86-15, 40 p. [Avail- able from Southwest Fisheries Science Center, National Marine Fisheries Service, PO. Box 271, La Jolla, CA 92038.] 2 Hanan, D. A., S. L. Diamond, and J. P. Scholl. 1986. An estimate of harbor por- poise mortality in California set net fish- eries, April 1, 1984 through March 31, 1985. NOAA/NMFS SWR Admin. Rep. SWR-86-16, 38p. [Available from South- west Fisheries Science Center, National Marine Fisheries Service, PO. Box 271, La Jolla, CA 92038.] 3 Hanan, D. A., S. L. Diamond, and J. P. Scholl. 1988. Estimates of sea lion and harbor seal mortalities in California set net fisheries for 1983, 1984, and 1985. Fi- nal Rep. NA-86-ABH00018 submitted to NOAA Fisheries, SWR, 10 p. [Available from Southwest Fisheries Science Center, National Marine Fisheries Service, PO. Box 271, La Jolla, CA 92038 ] 4 Hanan, D. A., and S. L. Diamond. 1989. Estimates of sea lion, harbor seal, and har- bor porpoise mortalities in California set net fisheries for the 1986-87 fishing year. Final Rep. NA-86-ABH00018 sub- mitted to NOAA Fisheries, SWR, 10 p. [Available from Southwest Fisheries Sci- ence Center, National Marine Fisheries Service, PO. Box 271, La Jolla, CA 92038.] 272 Fishery Bulletin 96(2), I 998 methods (Lennert et al., 1994; Perkins et al.5 6 7 8; Julian6, 7> 8). Separate estimates of entanglement are provided for species that had individuals released alive. This paper documents incidental marine mam- mal, turtle, and seabird kill estimates in these two California gillnet fisheries, based on data from the NMFS observer program and CDFG effort estimates for the period July 1990 through December 1994, and documents the process and methods leading to these estimates. Methods Data collection National Marine Fisheries Service observer data, daily logbooks of commercial gillnet fishermen, and receipts of landed fish sales were used in marine mammal mortality estimation. NMFS observer data were collected by trained technicians aboard com- mercial gillnet fishing boats that had a Marine Mam- mal Protection Act (MMPA) Exemption Permit and that met minimum U.S. Coast Guard safety stan- dards. There were two general observation catego- ries: observation of randomly selected trips and ob- servation of approximately every fifth vessel trip. In the setnet fishery, systematically selected trips were further divided. Notification prior to the setting of nets resulted in a preset, systematic observation rather than a postset, systematic observation where notification to the vessel was given after the nets were set (see “Discussion” section). NMFS observers re- corded data on location, date, marine mammal en- tanglements, including location of mammals in the 5 Perkins, P., J. Barlow, and M. Beeson. 1992. Pinniped and cetacean mortality in California gillnet fisheries: 1991. Inter- national Whaling Commission Scientific Committee working paper SC/44/SM14. [Available from Southwest Fisheries Sci- ence Center, National Marine Fisheries Service, PO. Box 271, La Jolla, CA 92038.] 6 Julian, F. 1993. Pinniped and cetacean mortality in Califor- nia gillnet fisheries: preliminary estimates for 1992, rev. 2/ 94. International Whaling Commission Scientific Committee working paper SC/45/022. [Available from Southwest Fisher- ies Science Center, National Marine Fisheries Service, P.O. Box 271, La Jolla, CA 92038.] 7 Julian, F. 1994. Pinniped and cetacean mortality in Califor- nia gillnet fisheries: preliminary estimates for 1993. Interna- tional Whaling Commission Scientific Committee working pa- per SC/46/011. [Available from Southwest Fisheries Science Center, National Marine Fisheries Service, P.O. Box 271, La Jolla, CA 92038.] 8 Julian, F. 1995. Cetacean and pinniped mortality in Califor- nia gillnet fisheries: preliminary estimates for 1994. Interna- tional Whaling Commission Scientific Committee working pa- per SC/47/05. [Available from Southwest Fisheries Science Center, National Marine Fisheries Service, P.O. Box 271, La Jolla, CA 92038.] net (by thirds of the net — vertically and horizontally), gear, bycatch, and target species catch for each net pull observed during a trip (Lennert et al., 1994). Observers recorded twelve net-related parameters for drift and set nets. They were net type (set, drift, float, or trammel), net material (monofilament, multi- filament, or a combination), net strength (pounds test or twine size depending on strength code), strength code, net length (fathoms), net depth (number of meshes), stretched mesh size (inches), extender length (feet, float, and drift nets only), hanging line material (synthetic or natural fiber), percent slack in net, number of meshes hanging (between knots to the cork line), and hanging length (distance between knots on the cork line in inches). Not infrequently, a drift or set net will consist of panels of varying char- acteristics. In this case, observers would record char- acteristics on up to 5 different panels. Net charac- teristics for both fisheries are summarized in Table 1. Although the variability in these characteristics con- tributes somewhat to the variability in mortality es- timates, the significant factors for mortality estima- tion are the amount of effort and the general loca- tion of the effort. For some species, e.g. pinnipeds, quarter of the year is also significant (Perkins et al.5). Collected data were entered into a database file, checked for accuracy, and tabulated for mortality estimation (Tables 2-5). After an initial six month period, this observation method continued unchanged in both fisheries. Realized observation rates varied between 4.4% and 17.9% yearly, but observation rates were more variable if stratified by area and quarter. Observation in the driftnet fishery continued through December 1995, whereas observation in the setnet fishery terminated by July 1994 because of a signifi- cant decrease in fishing effort in that fishery (due to regulations that restricted areas open to gillnet fish- ing). Observer data were complemented by informa- tion from vessel logbooks and landing receipts (i.e. receipts from sales of landed fish). Vessel logbooks were submitted monthly to CDFG and constituted the major source of information for estimation of total effort. Data for each logbook en- try included date, vessel and permit identification, area fished by CDFG block number (Lennert et al. 1994), gear, number of sets made, and number and species of fish caught (CDFG blocks are typically 10' square; larger blocks are defined for areas further from shore). Logbook information was entered into a database by technicians and checked for accuracy by biologists. Fish species targeted for catch by the fish- ermen were determined and assigned to each data entry by CDFG personnel according to fish caught, gear used, and other pertinent factors. Purchases of landed fish by commercial fish buyers were recorded Julian and Beeson: Estimates of marine mammal, turtle, and seabird mortality 273 Table 1 Observed net characteristics for the driftnet and setnet fisheries taken from nets characterized by one set of characteristics only (some nets may consist of two or more panels with differing characteristics), n = number of sets observed. Characteristic Driftnet fishery (n=2,932) Setnet fishery (n= 7,994) Net type All 2,932 were drift nets. 1,592 set nets and 6,278 1-panel trammel nets. Net material 2,838 (97%) were multifilament nets. 7,520 (94%) were monofilament and 439 were multifilament. Net strength and strength code Twine size of 24 was used for 25% of the nets; size 27 was used for 36%, and size 30 was used for 22% of the nets. 3,203 (40%) nets with twine size 66, 10% with twine size 55, and 19.4% nets with unrecorded data. Net length (m) mean=l,784.9; SD=55.4; mode=l,828.8 mean=468.9, SD=164.7; mode=457.2 Net depth (meshes) mean=128; SD=24; mode=130 mean=23.7; SD=8.9; mode=20 (n=7,880) Mesh size (cm) mean=52.1; SD=3.9; mode=53.34 mean=21.2 ; SD=2.2; mode=21.6 (n=7,968) Extender length (m) mean=11.48; SD=4.37; mode=11.30 Extenders not typically used; 98.4% of the nets did not use them. Hanging line material 2,809 (95.8%) nets were of synthetic fiber. 7,628 (95.4%) nets were of synthetic fiber. Percent slack mean=45%; SD=5.4%; mode=50% (n=2,609). No slack indicated for 38% of the nets. For nets with slack, mean=57%; SD=11%; mode=50% (n =4,986). Meshes hanging 60% had 2 and 35% had 1 mesh hanging between knots tied to the cork line. 42% had 6, 26% had 4, and 17% had 6 meshes hanging between knots tied to the cork line. Hanging length (cm) mean=50.7; SD=14.6; mode=60.0 for distance between knots on the cork line. mean = 38.4; SD=6.5; mode=38.1 for distance between knots on the cork line. and those records were submitted to CDFG twice monthly. Landing receipts included information on species landed, weight by species, price, gear, area fished, and vessel and permit identification. Land- ing information was entered into a database and checked for accuracy. A target fish species was as- signed to each entry. Logbook data, landing informa- tion, and NMFS observer data (date, set position, gear, and catch) were subsequently used in estimat- ing effort. Estimation of total effort Fishing effort in both fisheries is an unknown quan- tity and absolute determination is impractical. Con- sequently, estimates of total fishing effort in each fish- ery were used to estimate incidental kill. These esti- mates were based on the combination of observer records, logbook data, and landing receipts. Effort was measured in “effort-days” which was defined as one day of fishing for one vessel. In the driftnet fishery, one effort-day was considered equivalent to setting and retrieving one net, generally 1828.8 m (1,000 fm) in length. (One vessel, targeting thresher shark, made two sets per day.) In the setnet fishery, typi- cally two to four net settings, each of about 457.2 m (250 fm) in length, made up one effort-day. Days ac- tually fished was used as the measure of total effort in each fishery because previous exploratory analy- sis determined that number of days of effort and gen- eral location of effort were significant factors in esti- mation of mortality (Perkins, et al.5). For some spe- cies, quarter of the year was also determined to be significant. These factors were available for all ef- fort through the California Fish and Game Depart- ment and although other approaches to estimation (e.g. a modeling approach) can be developed, the cur- rent analysis is based on these factors. Data on other variables, such as total number of nets fished, total length of nets fished, or tons of target fish caught, were not readily available or contained additional variability due to nonsampling errors. Nonsampling 274 Fishery Bulletin 96(2), 1 998 Table 2 Observed (obs) and (est) estimated cetacean, pinniped, turtle, and seabird mortality, stratified by year, in the California sword- fish and shark drift gillnet fishery during the NMFS Observer Program, July 1990-December 1995. Estimates of total mortality are reported to the nearest individual. Estimated coefficients of variation (CV) are included in parentheses; ( — ) indicates CV was undefined. Effort and estimates for 1990 pertain to the third and fourth quarters only. Year 1990 1991 1992 1993 1994 1995 Estimated days effort 4,078 4,778 4,379 5,442 4,248 3,673 Observed days effort 178 470 596 728 759 572 Percent observer coverage 4.4% 9.8% 13.6% 13.4% 17.9% 15.6% Observed trips effort 54 88 97 107 134 97 obs est CV obs est CV obs est CV obs est CV obs est CV obs est CV Dali’s porpoise 1 23 (0.95) 2 20 (0.67) 1 7 (0.92) 9 67 (0.44) 2 11 (0.64) 1 6 (0.92) Pacific white-sided dolphin 3 69 (0.56) 5 51 (0.63) 3 22 (0.70) 2 15 (0.66) 3 17 (0.67) 1 6 (0.92) Risso’s dolphin 0 0 (— ) 5 51 (0.50) 5 37 (0.48) 7 52 (0.51) 1 6 (0.91) 6 39 (0.57) Bottlenose dolphin 0 0 (— ) 0 0 (— ) 3 22 (0.93) 0 0 (— ) 0 0 (— ) 0 0 (— ) Striped dolphin 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 6 (0.90) 0 0 (— ) Common dolphin (unknown stock) 4 92 (0.79) 7 71 (0.70) 5 37 (0.40) 4 30 (0.57) 1 6 (0.91) 0 0 (— ) Common dolphin (long beak) 0 0 (— ) 0 0 (— ) 2 15 (0.92) 0 0 (— ) 1 6 (0.91) 6 39 (0.65) Common dolphin (short beak) 4 92 (0.47) 37 376 (0.21) 39 287 (0.21) 24 179 (0.26) 25 140 (0.18) 36 231 (0.29) Northern right whale dolphin 0 0 (— ) 7 71 (0.41) 2 15 (0.65) 7 52 (0.39) 7 39 (0.42) 9 58 (0.59) Killer whale 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 6 (0.92) Short-finned pilot whale 1 23 (0.95) 0 0 (— ) 1 7 (0.92) 8 60 (0.54) 0 0 (— ) 0 0 (— ) Baird’s beaked whale 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 6 (0.90) 0 0 (— ) Stejneger’s beaked whale 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 6 (0.91) 0 0 (— ) Hubbs’ beaked whale 0 0 (— ) 0 0 (— ) 3 22 (0.53) 0 0 (— ) 2 11 (0.64) 0 0 (— ) Mesoplodont beaked whale 1 23 (0.97) 0 0 (— ) 1 7 (0.93) 0 0 (— ) 0 0 (— ) 0 0 (— ) Cuvier’s beaked whale 0 0 (— ) 0 0 (— ) 6 44 (0.36) 3 22 (0.53) 6 34 (0.36) 5 32 (0.40) Unidentified beaked whale 0 0 (— ) 0 0 (— ) 2 15 (0.65) 0 0 (— ) 1 6 (0.90) 0 0 (— ) Sperm whale 0 0 (— ) 0 0 (— ) 1 7 (0.94) 2 15 (0.66) 0 0 (— ) 0 0 (— ) Pygmy sperm whale 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 7 (0.93) 0 0 (— ) 0 0 (— ) Unidentified Kogia 0 0 (— ) 0 0 (— ) 1 7 (0.92) 0 0 (— ) 0 0 (— ) 0 0 (— ) Minke whale 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 6 (0.91) 0 0 (— ) Unidentified cetacean 0 0 (— ) 1 10 (0.95) 1 7 (0.93) 0 0 (— ) 0 0 (— ) 0 0 (— ) Unidentified dolphin 0 0 (— ) 0 0 (— ) 1 7 (0.93) 0 0 (— ) 0 0 (— ) 0 0 (— ) Unidentified whale 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 7 (0.93) 0 0 (— ) 0 0 (— ) Steller sea lion 0 0 (— ) 0 0 (— ) 1 7 (0.92) 0 0 (— ) 1 6 (0.91) 0 0 (— ) California sea lion 2 46 (0.99) 4 41 (0.58) 9 66 (0.34) 11 82 (0.42) 5 28 (0.40) 4 26 (0.45) Unidentified sea lion 2 46 (0.97) 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) Harbor seal 1 23 (0.95) 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) Northern elephant seal 5 115 (0.44) 13 132 (0.25) 15 110 (0.24) 14 105 (0.26) 22 123 (0.23) 14 90 (0.25) Loggerhead turtle 0 0 (— ) 0 0 (— ) 1 7 (0.93) 0 0 (— ) 0 0 (— ) 0 0 (— ) Leatherback turtle 1 23 (0.97) 0 0 (— ) 2 15 (0.65) 2 15 (0.66) 0 0 (— ) 4 26 (0.55) Unidentified turtle 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 7 (0.93) 0 0 (— ) 0 0 (— ) Seabirds (all unidentified) 1 23 (0.98) 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 6 (0.90) 0 0 (— ) errors, those not due to the sampling design, include but are not limited to made. The occurrence of this type of error implies that estimates of mortality are biased lower than actual mortality levels. 1 Effort that is not recorded. This type of error may occur for several reasons, namely the times when no marketable target fish are caught during the entire day of effort and when a log entry is not 2 Incorrect reporting of effort location by fishermen. This type of error may bias estimated mortality either higher or lower. Julian and Beeson: Estimates of marine mammal, turtle, and seabird mortality 275 Table 3 Observed (obs) and estimated (est) cetacean, pinniped, turtle, and seabird entanglement, stratified by year, in the California swordfish and shark drift gillnet fishery during the NMFS Observer Program, July 1990-December 1995. Estimates of entanglement are reported to the nearest individual. Estimated coefficients of variation (CV) are included in parentheses; ( — ) indicates CV was undefined. Effort and estimates for 1990 pertain to the third and fourth quarters only. Year 1990 1991 1992 1993 1994 1995 obs est CV obs est CV obs est CV obs est CV obs est CV obs est CV Common dolphin (unknown stock) 4 92 (0.79) 7 71 (0.70) 6 44 (0.36) 4 30 (0.57) 1 6 (0.91) 0 0 (— ) Cuvier’s beaked whale 0 0 (— ) 0 0 (— ) 6 44 (0.36) 3 22 (0.53) 6 34 (0.36) 6 39 (0.36) Sperm whale 0 0 (— ) 0 0 (— ) 3 22 (0.94) 3 22 (0.69) 0 0 (— ) 0 0 (— ) Humpback whale 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 6 (0.91) 0 0 (— ) California sea lion 2 46 (0.99) 4 41 (0.58) 9 66 (0.34) 12 90 (0.39) 5 28 (0.40) 5 32 (0.40) Unidentified sea lion 2 46 (0.97) 1 10 (0.95) 0 0 (— ) 0 0 (— ) 0 0 (—) 0 0 (— ) Loggerhead turtle 0 0 (— ) 0 0 (— ) 2 15 (0.66) 5 37 (0.49) 0 0 (— ) 0 0 (— ) Leatherback turtle 1 23 (0.97) 1 10 (0.94) 4 29 (0.46) 3 22 (0.53) 1 6 (0.91) 5 32 (0.47) Unidentified turtle 0 0 (— ) 0 0 (— ) 0 0 (— ) 3 22 (0.93) 0 0 (— ) 0 0 (— ) Seabirds (all unidentified) 1 23 (0.98) 0 0 (— ) 1 7 (0.93) 0 0 (— ) 1 6 (0.90) 0 0 (— ) 3 Underestimation of effort from landing receipts. This type of error may occur when a landing re- ceipt, in absence of additional information, is as- sumed to represent one effort-day (default- value). A reliable determination of the magnitude of bias these errors cause in estimates of mortality rates has not been made; however, their characteristics indi- cate that estimates of mortality may be lower than actual values. Once quarterly effort data were collected, computer programs developed by CDFG were used to assign target species to landing receipts on the basis of in- formation provided by logbooks and observations (Beeson and Hanan9). Landing data were then con- firmed or modified on the basis of logbook and ob- server data for the same target species, date, and vessel number. After all three data sources were com- pared, a day of effort was tallied for each record with a logbook entry or observer record. Landing receipts without corresponding logbook or observer entries three days before and after the receipt date were assumed to represent one day of effort. The num- bers of days fished in each CDFG block were then tallied and the resultant data represented estimated total effort. Total effort was estimated quarterly and yearly. Delayed submission of data to CDFG was the primary reason that estimates of yearly effort differed from the sum of the quarterly estimates of effort. 9 Beeson, M., and D. Hanan. 1996. Manuscript submitted to California Fish and Game. [Available from the authors at Cali- fornia Department of Fish and Game, 330 Golden Shore, Suite 50, Long Beach, CA 90802.] Mortality estimation Sampling design The NMFS observer program began in July 1990. Initially, the plan was to sample every fifth trip made by a drift gillnet vessel accord- ing to an assignment schedule. The order in which vessels with MMPA exemption certificates were scheduled to be observed was randomly selected. It became evident during the 1990 season that this scheme would not work because of logistical difficul- ties in adhering to the sampling plan. Beginning in January 1991, gillnet vessel trips were selected according to the targeted coverage rate (20%), the availability of observer personnel, call-ins (fishermen called in prior to departure), and the abil- ity to notify fishermen of their obligation to carry an observer. (Occasionally fishermen did not call the Observer Program administrator for possible ob- server assignment.) In addition, the NMFS Fisher- ies Observer Branch began monitoring vessel activ- ity (arrivals and departures) to estimate observer coverage for placement purposes. If estimated ob- server coverage dropped below 20% for a vessel, the owner was notified of the obligation to carry an ob- server. In the setnet fishery, most fishermen were notified after their nets were set whether they would be required to carry an observer. As the program evolved, setnet fishermen began to expect an observer about every fifth trip. Mortality estimation in the drift gillnet fishery In the swordfish and shark drift gillnet fishery, vessels made trips lasting from one to about 15 days so that 276 Fishery Bulletin 96(2), 1998 Table 4 Observed (obs) and estimated (est) cetacean, pinniped, turtle, and seabird mortality, stratified by year, in the California halibut and angel shark set gillnet fishery during the NMFS Observer Program, July 1990-December 1995. Estimates of total mortality are reported to the nearest individual. Estimated coefficients of variation (CV) are included in parentheses; ( — ) indicates CV was undefined. Effort and estimates for 1990 pertain to the third and fourth quarters only. Year 1990 1991 1992 1993 1994 1995' Estimated days effort Observed days effort Percent observer coverage Observed trips effort 3,041 158 5.2% 406 7,171 706 9.8% 2,233 5,577 698 12.5% 2,123 5,680 875 15.4% 2,642 1,943 150 7.7% 547 2,257 0 0% 0 obs est CV obs est CV obs est CV obs est CV obs > est CV obs est CV Harbor porpoise 4 37 (0.56) 5 38 (0.47) 6 48 (0.46) 2 13 (0.64) 1 14 (0.96) — 14 (0.64) Common dolphin (unknown stock) 0 0 (— ) 0 0 (— ) 2 15 (0.65) 0 0 (— ) 0 0 (— ) (— ) Unidentified cetacean 0 0 (— ) 0 0 (— ) 1 8 (0.92) 0 0 (— ) 0 0 (— ) — — (— ) California sea lion 67 867 (0.22) 142 1,842 (0.16) 338 3,418 (0.28) 237 1,942 (0.13) 109 905 (0.15) — 724 (0.08) Unidentified sea lion 1 23 (0.96) 6 109 (0.53) 7 54 (0.34) 0 0 (— ) 0 0 (— ) — — (— ) Harbor seal 30 411 (0.23) 42 601 (0.23) 90 1,204 (0.47) 71 475 (0.13) 23 227 (0.33) — 228 (0.13) Northern elephant seal 13 119 (0.40) 3 30 (0.55) 7 51 (0.35) 11 70 (0.27) 2 16 (0.66) — 47 (0.29) Unidentified pinniped 2 42 (0.79) 3 30 (0.55) 7 50 (0.39) 7 32 (0.90) 1 8 (0.94) — 17 (0.83) Sea otter 3 27 (0.53) 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) — — (— ) Green/black turtle 0 0 (— ) 0 0 (— ) 1 8 (0.92) 1 6 (0.90) 0 0 (— ) — 2 (0.61) Loggerhead turtle 0 0 (— ) 0 0 (— ) 1 8 (0.92) 0 0 (— ) 0 0 (— ) — — (— ) Leatherback turtle 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 8 (0.94) — — (— ) Unidentified turtle 0 0 (— ) 0 0 (— ) 0 0 (— ) 1 6 (0.90) 0 0 (— ) — 2 (0.61) Pacific loon 0 0 (— ) 1 13 (0.94) 0 0 (— ) 0 0 (— ) 0 0 (— ) — — (— ) Common loon 0 0 (— ) 2 22 (0.68) 1 7 (0.92) 0 0 (— ) 0 0 (— ) — — (— ) Unidentified loon 1 23 (0.96) 4 48 (0.48) 0 0 (— ) 0 0 (— ) 0 0 (— ) — — (— ) Western grebe 0 0 (— ) 1 8 (0.92) 3 23 (0.70) 1 6 (0.90) 0 0 (— ) — 2 (0.61) Unidentified grebe 0 0 (— ) 0 0 (— ) 4 31 (0.92) 1 6 (0.90) 0 0 (— ) — 3 (0.83) Double-crested cormorant 2 18 (0.93) 0 0 (— ) 1 7 (0.92) 0 0 (— ) 1 8 (0.94) — — (— ) Brandt’s cormorant 2 41 (0.78) 36 409 (0.44) 14 279 (0.67) 3 13 (0.64) 2 16 (0.66) — 3 (0.43) Pelagic cormorant 1 33 (0.98) 1 8 (0.92) 0 0 (— ) 0 0 (— ) 0 0 (— ) — — (— ) Unidentified cormorant 9 132 (0.45) 15 450 (0.92) 9 68 (0.30) 5 32 (0.40) 0 0 (— ) — 10 (0.35) Common murre 142 1,300 (0.21) 289 2,201 (0.27) 292 2,333 (0.28) 137 879 (0.32) 20 284 (0.29) — 967 (0.32) Unidentified alcid 1 9 (0.93) 0 0 (— ) 0 0 (— ) 0 0 (— ) 0 0 (— ) — — (— ) Unid. seabird 0 0 (— ) 2 22 (0.68) 3 23 (0.53) 1 6 (0.90) 0 0 (— ) — 3 (0.83) 1 Estimates for 1995 were based on stratified rates from 1993 results. when a trip was chosen to be observed, the NMFS technician observed all net pulls during the trip. A single net per day was set at dusk and retrieved be- fore dawn. Net pulls and effort-days were equiva- lent units for this fishery. For estimation of inciden- tal kill, the collection of observed trips during a year was treated as a random sample (an approximation) and a ratio estimator was used. Trips were treated as sampling units and the number of days per trip was treated as an auxiliary variable. Stratification by quarter of year or set location was not used for yearly estimates because previous exploratory analysis had not found this type of stratification to be significantly related to incidental kill. Yearly estimates, 1991-95, were calculated for each species observed entangled (“Results” section). Estimates for 1990 correspond only to the last two quarters of that year. Formulae from Cochran (1977) were used for estimating kill rate, r , total incidental kill, m , and variances: Julian and Beeson: Estimates of marine mammal, turtle, and seabird mortality 277 Table 5 Observed and estimated cetacean, pinniped, turtle, and seabird entanglement, stratified by year, in the California halibut and angel shark set gillnet fishery during the NMFS Observer Program, July 1990-December 1995. Estimates of entanglement are reported to the nearest individual. Estimated coefficients of variation (CV) are included in parentheses; ( — ) indicates CV was undefined. Effort and estimates for 1990 pertain to the third and fourth quarters only. Year 1990 1991 1992 1993 1994 1995 1 Unidentified sea lion 67 867 (0.22) 143 1,850 (0.16) 341 3,438 (0.28) 239 1,977 (0.13) 109 905 (0.15) — 729 (0.08) Harbor seal 30 411 (0.23) 43 615 (0.23) 90 1,204 (0.47) 71 475 (0.13) 23 227 (0.33) — 228 (0.13) Unidentified turtle 0 0 (— ) 0 0 (— ) 0 0 (— ) 2 13 (0.64) 0 0 (— ) — 5 (0.59) Common loon 0 0 (— ) 4 48 (0.60) 1 7 (0.92) 0 0 (— ) 0 0 (— ) — — (— ) Western grebe 0 0 (— ) 2 22 (0.68) 3 23 (0.70) 1 6 (0.90) 0 0 (— ) — 2 (0.61) Brandt’s cormorant 2 41 (0.78) 41 494 (0.37) 20 321 (0.58) 5 25 (0.45) 2 16 (0.66) — 9 (0.40) Unidentified seabird 0 0 (— ) 2 22 (0.68) 5 37 (0.41) 1 6 (0.90) 0 0 (— ) — 3 (0.83) 1 Estimates for 1995 were based on stratified rates from 1993 results. m = Dr , (3) ~ 2 n2^2 = U or (4) Variables kf and di represent the observed kill and number of days for the ith trip; davg is the sampled mean number of days per trip; ms = Dsrs , (5) (6) (7) 278 Fishery Bulletin 96(2), 1998 o: n, 2-2 = O s ^ r,s * (8) The variable kt s represents ob- served kill for the ith observed day in stratum s, and a l s is the sample variance of the ob- served kill. Variables a? and D , are observed and total number of days of effort in the stratum, respectively. Estimates of over- all kill rate, r , and total inciden- tal kill, rh , across all strata, and variances, are then weighted averages: ^sDsK D y T~\ 2 ~ 2 -2 .9 -s ° r,s , O = 5 ’ r D2 rh = Dr , -2 n2-2 - D 0.60) between pink and sockeye salmon, pink and chum salmon, and chum and sockeye salmon. Coho salmon diet overlap was <0.60 in all paired comparisons. Nearly all (98.6%) of the 2,210 stomachs examined were at least half full. Although, in general, prey con- sumed were not very similar to prey observed in the environment, the com- position of salmon diets was more simi- lar to neuston collections than to zoo- plankton collections. Manuscript accepted 11 August 1997. Fishery Bulletin 96:285-302 (1998). Feeding habits of juvenile Pacific salmon in marine waters of southeastern Alaska and northern British Columbia Joyce H. Landingham Molly V. Sturdevant Auke Bay Laboratory, Alaska Fisheries Science Center National Marine Fisheries Service, NOAA I 1305 Glacier Highway, Juneau, Alaska 99801-8626 E-mail address (for Molly Sturdevant, contact author): molly.sturdevant@noaa.gov Richard D. Brodeur Alaska Fisheries Science Center National Marine Fisheries Service, NOAA 7600 Sand Point Way NE, Seattle, Washington 981 1 5-0070 All Pacific salmon ( Oncorhynchus spp.) migrate as juveniles from shal- low waters near shore to coastal and oceanic feeding areas of the North Pacific Ocean, where species, stocks, and age classes mix (Hartt and Dell, 1986; Ware and McFarlane, 1989; Pearcy, 1992). Substantial evidence suggests that salmon production around the Gulf of Alaska increased dramatically in the late 1970’s, pos- sibly owing to enhanced early ocean survival (Pearcy, 1992; Beamish and Bouillon, 1993; Brodeur and Ware, 1995). Increased densities are evi- dent in the commercial harvest of salmon in Alaska; for example, salmon catch reached record num- bers in 1993, nearly doubling over 25 years to approach 200 million fish (Wertheimer, 1997). This in- crease is attributed to several fac- tors, including growing enhance- ment efforts and environmental conditions that favor survival of both wild and hatchery salmon stocks. The increase in numbers of sub- adult and adult salmon feeding in marine waters has prompted inter- est in possible density-dependent effects on survival, growth, and pro- duction of salmon stocks around the Pacific rim (e.g. Helle, 1989; Kaeri- yama, 1989; Ishida et al., 1993; Helle and Hoffman, 1995). Fishery managers are concerned about the effects of increasing interactions between salmon populations — both wild and enhanced — in the various marine habitats where they mature (McNeil and Himsworth, 1980; Peterson et al., 1982; Brodeur, 1989; Brodeur and Pearcy, 1990; Pearcy, 1992). The feeding ecology of juve- nile salmon is a pertinent, but not well known, topic. The objective of this study was to describe the feeding habits of juve- nile salmon during their first sum- mer in coastal waters of the east- ern Gulf of Alaska. Although Hartt and Dell (1986) summarized the dis- tribution, migration, and growth of juvenile salmon in the North Pacific Ocean and the Bering Sea, diet and feeding were addressed superfi- cially. Other references concerning juvenile salmon feeding habits dur- ing early marine residence, espe- cially along the open coast, are lim- ited (Pearcy, 1992; Brodeur1), and juvenile salmon diets are often 1 See next page for footnote. 286 Fishery Bulletin 96(2), 1 998 treated as a subset of adult diets (e.g. Groot and Margolis, 1991). Therefore, as a first step to un- derstanding trophic interactions during their first marine summer, our research focused on the feeding habits and diet overlap of juvenile chinook (O. tshawytscha), coho (O. kisutch), sock- eye (O. nerka), chum (O. keta ), and pink (O. gorbuscha) salmon. We compared diet overlap between species, habitats, months, and years and also examined the similarity of salmon diets with samples of prey collected concurrently. Methods Fish coliection Juvenile salmon were collected in 120 purse-seine sets made along transects with two vessels and two seines (Table 1; see also Jaenicke and Celewycz, 1994). The 28-m NOAA RV John N. Cobb fished a table seine in the marine waters of southeastern Alaska during August 1983 and July and August 1984 (Fig. 1). The 24-m FV Bering Sea fished a drum seine in the open coastal waters of northern British Columbia dur- ing July 1984. Purse-seine sets at all locations were standardized to compensate for different sizes, meshes, and areas enclosed; both nets had 25-mm mesh in the bunt (Jaenicke and Celewycz, 1994). Each station was fished only once during a period, except on a few occasions when a set was repeated following an empty haul. Fishing was conducted almost exclusively between 0600 and 1800 hours. All sets were round hauls (i.e. the seine was not towed or held open to increase catches). The distribution, size, and abundance of juvenile salmon examined for our study have been summarized in a companion paper by Jaenicke and Celewycz (1994). The waters sampled were partitioned into discrete habitats: 1) outside waters: the open waters adjacent to the Gulf of Alaska; 2) inside waters: the enclosed marine waters of the Alexander Archipelago; and 3) protected outer-coast inlets (Fig. 1). Outside waters were further partitioned into nearshore (0-37 km offshore) and offshore (46-74 km offshore) during August 1984; seining was restricted to within 37 km of shore during other sampling periods. Inside wa- ters were partitioned into bays and passages (con- necting to the Gulf of Alaska). Sampling in outside 1 Brodeur, R. D. 1990. A synthesis of the food habits and feed- ing ecology of salmonids in marine waters of the North Pacific (INPFC Doc.) FRI-UW-9016. Fish. Res. Inst., Univ. Washing- ton, Seattle, 38 p. Figure 1 Location of purse-seine sets in southeastern Alaska in 1983 and 1984 and in British Columbia in 1984. waters was along transects about 72 km apart in southeastern Alaska and 108 km apart in British Columbia. Transects began as close to shore as sub- marine topography and seine depth permitted and continued up to 74 km offshore, depending on fish abundance and weather. Fish from each purse-seine haul were anesthetized with MS 222 (tricaine methanesulfonate), sorted, identified, and measured on board the vessel. A subsample of each salmon species (n< 25) was pre- served in 10% formalin-seawater solution for later stomach analysis (Table 1). Nonsalmonids caught incidentally were identified and enumerated, catch per unit of effort (CPUE) was estimated, and a subsample was measured for length as part of prey assemblage assessments (see below). Sea surface temperature (SST) to the nearest 0.2°C was recorded at every purse-seine site. Landingham et a I.: Feeding habits of juvenile Pacific salmon 287 Table 1 Number of purse-seine sets by time period and habitat and subsample size by species of juvenile salmon used in diet analyses. BC = British Columbia; AK = Alaska. Number of sets catching Number of fish Habitat Number of sets juvenile salmon Pink Chum Sockeye Coho Chinook August 1983 Inside inlet 15 3 21 23 2 14 1 Inside passage 39 23 142 112 44 88 2 Outer coast inlet 27 8 60 47 0 21 0 Outside waters 8 4 15 2 9 26 0 Total 89 38 238 184 55 149 3 July 1984 Inside inlet 13 5 14 17 12 63 10 Inside passage 5 2 20 3 9 44 0 Outer coast inlet 14 5 8 3 0 28 0 Outside waters (BC) 21 11 69 30 55 7 3 Outside waters (AK) 33 15 94 40 83 27 1 Total 86 38 205 93 159 169 14 August 1984 Inside inlet 18 6 35 16 10 82 6 Inside passage 19 7 61 10 5 73 10 Outer coast inlet 4 1 0 12 0 3 0 Outside nearshore 26 20 183 97 80 55 5 Outside offshore 11 10 93 41 52 12 0 Total 78 44 372 176 147 225 21 All periods combined Inside inlet 46 14 70 56 24 159 17 Inside passage 63 32 223 125 58 205 12 Outer coast inlet 45 14 68 62 0 52 0 Outside waters 99 60 454 210 279 127 9 Total 253 120 815 453 361 543 38 Stomach analysis Each fish was weighed to the nearest milligram and measured to the nearest millimeter fork length (FL) in the laboratory. Stomachs were excised and placed in 70% isopropyl alcohol. During analysis, stomach fullness on a scale of 0-6 (0=empty, 6=distended) and digestion on a scale of 1-4 (l=fresh prey items, 4=completely digested) were visually estimated. Stomach contents were weighed, and prey items were separated, identified to the lowest convenient taxon, and counted. Up to 100 individuals of each prey cat- egory that had been removed in good condition from the stomach were used to measure initial wet weights. Prey fish in an advanced state of digestion were as- signed to discrete weight categories based on the most complete specimens encountered: small (estimated 6.0 mg), medium (184.4 mg), and large (580.5 mg). Dry weights were obtained by drying the samples in an oven at 60°C until constant weights were obtained. Sampling and analysis of prey assemblages Prey assemblages were sampled with neuston and plankton nets in the areas fished in 1984. Neuston collections were made with a rectangular 100 x 35- cm-opening neuston-net frame containing a conical 505-pm-mesh net; tows were made at 45 of 54 out- side-water locations, 7 of 14 outer-coast-inlet loca- tions, and 5 of 18 inside-water locations. The neus- ton net was towed horizontally, half-submerged, so that it sampled the water column from the surface to approximately 17 cm depth (Brodeur, 1989; Brodeur2). The plankton collections were made with a 70-cm diameter conical plankton net of 303-pm mesh; tows were made along purse-seine transects at four 4-km locations and six 16-km locations in 2 Brodeur, R. D., W. G. Pearcy, B. C. Mundy, and R. W. Wisseman. 1987. The neustonic fauna in coastal waters of the northeast Pacific: abundance, distribution, and utilization by juvenile salmonids. Oregon State Univ. Publ. ORESU-T-87-001, 61 p. 288 Fishery Bulletin 96(2), 1998 outside waters, and at six locations in inside waters. All plankton tows were oblique from the surface to a depth of 50 m. Samples were preserved in a 10%- formalin-seawater solution after debris was removed. Plankton and neuston samples were sorted to re- move large organisms, such as gelatinous zooplank- ton. We then split the sample with a Folsom splitter until a subsample of about 500 organisms remained. Plankton organisms were identified to the lowest convenient taxon and counted. Detailed composition of many of the neuston samples has been presented elsewhere (Brodeur2). Data analysis Stomach data were partitioned into subsets accord- ing to salmon species, geographic area, habitat, dis- tance offshore, month, and year. The index of rela- tive importance (IRI; Pinkas etal., 1971) was calcu- lated for each data subset. The modified IRI (dry weight rather than volume) was used to character- ize the diet of each species and to rank prey taxa: IRI = (N + W)F, where N is numerical percentage, W is weight per- centage, and F is frequency of occurrence (FO) per- centage. In all comparisons, the IRI is expressed as a percentage of total IRI for each data subset. Morisita’s index of overlap as modified by Horn (1966) was used to calculate overlap between spe- cies pairs; values range from 0 (no overlap) to 1 (com- plete overlap): s 2 i=l y, i= i i= 1 where x; and y ■ are proportions of the numbers of individuals of prey species i found in the predator species x andy, respectively. The percent similarity index (PSI; Whittaker, 1975) was used to compare stomach samples to plankton samples: PSI = 2^min(pg or pb), where pa is percentage number for a given species in sample A, and pb is percentage number for the same species in sample B. A PSI value of 1.00 shows com- plete similarity; a value of 0 indicates no similarity. We considered values >0.60 to be significant for both overlap indices. Chinook salmon were not included in the analysis of overlap because of small sample sizes (Table 1). Following Brodeur and Pearcy ( 1990), we tested for differences in the occurrence of princi- pal prey between years, months, areas, and distance offshore using ^2. We examined data for neuston and plankton prey samples collected at locations where stomachs of at least five specimens of one salmon species were avail- able and included a taxon in the prey collections. To measure prey selection, we used Strauss’s linear food selection index (Strauss, 1979): L = ri-Pl, where r and pt are the proportional abundances of prey item i in the gut and habitat, respectively. Se- lection values range from -1, indicating avoidance or negative accessibility, to +1, indicating preference or positive selection; 0 indicates random feeding. Values are extreme only when the prey item is pro- portionately abundant but rarely consumed (-1), or is proportionately rare but consumed almost exclu- sively ( + 1). We tabulated selection values >0.10 or <-0.10 for an indication of the positive or negative selection of a particular taxon. To compare the number of stomachs required to characterize the breadth of diet for each species of salmon, we pooled the stomachs over all periods and habitats. Stomachs were selected randomly, and the cumulative number of taxa were plotted versus the number of stomachs until the asymptote was reached (Hurtubia, 1973). Results Description of diet All salmon species The prey spectrum for juveniles of five Pacific salmon species comprised at least 30 taxa (Table 2). The six taxonomic groups of greatest importance (IRI) were calanoid copepods, hyperiid amphipods, euphausiids, decapods, tunicates, and fishes (Table 3). In pooled samples, fish were the most important prey for coho and chinook salmon (IRI=63.8% and 76.4%) but were only moderately important for the other species (IRI 28.3-40.3%). Hyperiid amphipods, most commonly Themisto spp., were also important prey for pink, chum, and sock- eye salmon (IRI 28.0-39.6%). However, the biomass of teleost prey made up more than 75% of the total biomass consumed by each of the juvenile salmon species in pooled samples. The full breadth of the prey spectrum for juvenile salmon species was obtained by randomly selecting Landingham et a I.: Feeding habits of juvenile Pacific salmon 289 16-95% of the actual number of stomachs analyzed (Fig. 2). The highest cumulative numbers of prey taxa in- 25 and 26, respec- tively) were observed after we had ran- domly subsampled 86 coho and 396 pink salmon stomachs (Fig. 3). Virtually all chinook salmon analyzed (39) were needed to reach the 14 cumulative prey taxa ob- served. The curves of cumulative number of taxa for pink and chum salmon were more similar than for any other pair of species (Fig. 3). Pink salmon The prey of juvenile pink salmon (85-222 mm FL; x =142 mm; SD=22.9) encompassed 26 taxa and sev- eral life-history stages. Hyperiid amphi- pods, especially the genus Themisto, had the highest total IRI (Fig. 4) and highest FO and mean abundance over all periods and habitats (Table 3). Juvenile fish had the second-highest IRI and the greatest biomass — 76% of pooled weight of stom- ach contents. In 1984, the IRI for fish prey did not rank first, but it was more than twice the 1983 value; fish prey ranked higher in Alaska than in British Colum- bia. Tunicates, primarily the larvacean Oikopleura dioica, were the third most important prey in pooled samples. Chum salmon The prey spectrum for ju- venile chum salmon (80-276 mm FL; x=151 mm; SB=28.4) included 22 taxo- nomic groups and several life-history stages. Juvenile fish, tunicates (salps and the larvacean O. dioica ), and hyperiid amphipods ( Themisto spp.) had the high- est IRI’s overall (Fig. 5). Sockeye salmon The prey spectrum for juvenile sockeye salmon (91-202 mm FL; v = 1 5 1 mm; SD=18.9) encompassed 18 taxonomic groups and several life-history stages. Fish prey had the highest IRI (40.3%) and the greatest average weight (87.2%) pooled over all habitats and peri- ods (Table 3; Fig. 6), although no fish were sampled from outer-coast inlets (Table 1). Fish prey were more important in 1984 than in 1983, especially in the outside waters of southeastern Alaska in July 1984 and inside inlets and outside waters (>37 km) in August 1984 (Fig. 6). Hyperiid am- phipods ( Themisto spp.) were the second Table 2 Prey from juvenile salmon stomachs as index of relative importance (IRI). IRI = (N + W)FO, where N= numerical percentage, W = weight percentage, and FO - frequency of occurrence percentage. Numbers in parentheses are totals of taxonomic groups for which more than one taxon is listed. IRI (%) Prey Pink Chum Sockeye Coho Chinook Polychaeta Unidentified <0.01 <0.01 <0.01 <0.01 Mollusca (0.90) (0.16) (0.17) (0.02) (— ) Gastropoda Limacina helicina 0.82 0.16 0.15 0.02 Bivalvia Unidentified <0.01 <0.01 Cephalopoda Unidentified 0.08 <0.01 0.02 <0.01 0.18 Copepoda (3.71) (2.93) (4.15) (0.02) (<0.01 ) Neocalanus cristatus <0.01 Epilabidocera longipedata <0.01 <0.01 Unidentified 3.70 2.93 4.14 0.02 <0.01 Cumacea Unidentified <0.01 <0.01 Amphipoda Hyperiidea (39.61) (28.03) (36.48) (6.84) (0.10) Hyperia sp. 0.01 0.01 0.01 0.02 0.01 Themisto spp. 38.75 27.44 36.11 6.73 0.08 Primno macropa 0.83 0.58 0.36 0.09 0.01 Vibilia sp. 0.02 <0.01 0.01 <0.01 — Euphausiacea (5.95) (6.18) (12.62) (3.88) (18.34) Euphausia paeifica <0.01 — — <0.01 — Thysanoessa spinifera <0.01 0.01 — 0.01 0.08 Unidentified 5.94 6.17 12.62 3.86 18.26 Decapoda Unidentified 4.67 3.16 3.64 25.26 4.81 Insecta Unidentified 0.01 <0.01 <0.01 0.01 Chaetognatha Unidentified 0.01 0.01 <0.01 <0.01 Urochordata (tunicates) (13.56) (30.32) (2.26) (0.03) Salpidae Unidentified <0.01 0.21 <0.01 Larvacea Oikopleura dioica 13.56 30.11 2.26 0.03 Osteichthyes (31.11) (29.88) (40.27) (63.78) (76.42) Clupea pallasi — — — 0.01 0.26 Osmeridae <0.01 0.02 — <0.01 0.61 Myctophidae 0.02 0.02 <0.01 — — Theragra chalcogramma 0.01 Sebastes spp. 0.04 — 0.03 0.04 — Cottidae — — — <0.01 — Stichaeidae <0.01 — <0.01 <0.01 0.02 Ammodytes hexapterus <0.01 0.03 0.30 0.47 Pleuronectidae — <0.01 <0.01 0.02 — Unidentified 31.04 29.80 40.22 63.39 75.06 290 Fishery Bulletin 96(2), 1998 Table 3 Stomach contents of juvenile salmon in marine waters of southeastern Alaska and northern British Columbia in 1983 and 1984. N is numerical percentage, W is dry weight percentage, FO is frequency of occurrence percentage of fish with prey item i , and IRI is percent of total IRI for all prey taxa. IRI = (N + W)FO. Taxa are omitted if IRI is <1% for all salmon. Pink salmon Chum salmon Sockeye salmon Coho salmon Chinook salmon N W FO IRI N W FO IRI N W FO IRI N W FO IRI N W FO IRI (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) All periods combined Calanoids 5.6 0.9 47.9 3.7 5.4 0.8 32.9 3.0 9.0 0.5 41.5 4.2 Decapods 3.8 4.0 48.6 4.7 2.4 3.6 36.6 3.1 5.4 2.6 43.2 3.6 52.0 0.1 56.3 25.3 11.2 0.2 38.5 4.8 Euphausiids 5.2 5.5 46.2 6.0 6.0 5.9 39.8 6.1 16.5 4.3 57.4 12.6 12.8 0.1 35.8 3.9 52.6 0.2 33.3 18.3 Hyperiids 45.2 5.8 90.3 39.6 29.5 5.7 73.6 28.0 46.2 3.7 90.1 36.5 22.9 0.1 46.4 6.8 1.5 0.2 7.7 0.1 Tunicates 32.2 0.1 35.2 13.6 52.5 0.1 46.7 30.1 16.4 <0.1 13.1 2.3 Teleosts 1.3 76.0 39.1 31.2 0.9 82.7 28.6 28.3 2.9 87.2 47.0 40.3 8.4 99.5 73.2 63.8 30.2 99.0 71.8 76.4 Total 93.3 92.3 98.8 96.7 98.7 98.6 96.4 98.3 99.5 96.1 99.8 99.8 95.5 99.5 99.6 August 1983 Calanoids 6.4 2.9 62.6 6.6 2.0 4.7 39.7 3.1 10.9 4.6 52.7 6.0 0.4 <0.1 10.1 <0.1 Decapods 4.3 8.6 60.9 8.9 0.9 12.1 32.6 4.8 1.1 8.3 30.9 2.2 46.4 0.5 63.1 26.4 Euphausiids 4.9 14.1 52.9 11.4 5.3 32.7 46.2 18.4 17.0 39.0 76.4 31.5 5.8 0.6 36.9 2.0 Hyperiids 37.0 12.4 71.0 33.6 31.5 28.0 64.7 38.5 58.5 30.7 94.5 55.1 41.5 0.6 59.7 21.7 Tunicates 38.7 0.2 48.7 20.3 56.8 0.5 60.9 28.8 8.7 0.1 14.5 0.9 0.1 <0.1 6.0 <0.1 Teleosts 0.4 56.7 29.9 16.6 0.2 20.2 19.6 4.6 0.3 16.7 21.8 2.7 2.7 96.4 61.1 49.1 Total 91.7 94.9 97.4 96.7 98.2 98.2 96.5 99.4 98.4 96.9 98.1 99.2 July 1984 Calanoids 15.0 0.8 39.0 7.3 23.4 1.4 28.0 11.0 12.1 0.4 4.3 5.6 0.7 <0.1 7.1 <0.1 Decapods 4.7 1.9 42.0 3.3 1.1 2.1 26.9 1.4 8.8 0.8 32.1 3.2 54.3 0.1 65.7 27.6 Euphausiids 7.0 5.0 46.3 6.6 4.8 8.1 37.6 7.7 23.0 3.4 62.3 17.0 16.6 0.1 33.7 3.8 Hyperiids 45.2 4.3 68.3 39.4 26.0 6.3 53.8 27.1 37.8 2.0 70.4 28.9 6.1 0.1 25.4 0.8 Tunicates 11.9 <0.1 17.6 2.5 34.3 0.1 29.0 15.8 9.1 <0.1 7.5 0.7 4.1 <0.1 5.3 <0.1 Teleosts 3.6 84.2 39.0 38.4 3.9 77.4 32.3 35.0 4.2 92.0 45.9 43.9 15.5 99.7 84.0 67.6 Total 87.4 96.2 97.5 93.5 95.4 98.0 95.0 98.6 99.3 97.3 100 99.8 August 1984 Calanoids 2.3 0.7 37.9 1.3 3.8 0.3 25.1 1.5 4.7 0.5 32.7 1.8 0.2 <0.1 2.7 <0.1 Decapods 3.2 3.9 43.4 3.7 5.0 2.9 42.9 5.0 4.5 5.3 61.2 6.4 58.3 0.1 42.7 24.7 Euphausiids 4.8 3.7 39.3 3.9 7.3 2.4 29.7 4.2 9.6 3.8 41.5 5.9 19.7 0.1 32.4 6.3 Hyperiids 50.1 5.0 65.3 40.8 27.9 3.0 53.1 22.6 47.5 5.1 68.7 36.5 10.3 0.1 24.0 2.1 Tunicates 34.4 <0.1 34.7 14.4 52.8 <0.1 37.1 28.9 28.3 <0.1 19.7 5.9 <0.1 <0.1 0.1 <0.1 Teleosts 1.1 85.1 38.5 35.5 0.8 90.8 28.0 37.3 3.1 83.1 47.6 43.4 10.7 99.6 72.9 66.8 Total 95.9 98.4 99.6 97.6 99.4 99.5 97.7 97.8 99.9 99.2 99.9 99.9 most important prey group (IRI=36.5%) and had the highest FO and average count (Table 3). Coho salmon The prey spectrum of juvenile coho salmon (112-309 mm FL; x =222 mm; SD=35.1) en- compassed 25 taxonomic groups, including eight iden- tifiable fish families (Table 2). Teleosts made up 99.5% of the total prey weight, with a mean number of 8.4 fish consumed by 73.2% of the coho predators (Table 3). Juvenile fishes were the principal compo- nent of coho salmon diet in all habitats except in the outside waters during August 1983 (Fig. 7), when decapod larvae were more important (IRI=66%). The IRI for prey fish ranged from 51% to 81% in the other habitats, with Pacific sand lance ( Ammodytes hexapterus) the most prevalent species. Other iden- tifiable prey fish (in order of importance) included walleye pollock (Theragra chalcogramma ), rockfish ( Sebastes spp.), flatfish (Pleuronectidae), Pacific her- ring ( Clupea pallasi), smelt (Osmeridae), prickle- backs (Stichaeidae), and sculpins (Cottidae). Fish were more important in 1984 than in 1983 and higher in Alaska than in British Columbia. Chinook salmon The prey spectrum of juvenile chinook salmon (139-324 mm FL; x =224 mm; SD=58.4) encompassed 14 taxonomic categories and four identifiable fish families (Table 2). Fish were the most important prey: IRI’s ranged from 28% to 100% among habitats (data not shown). Fish made up 99% (biomass) of chinook salmon stomach contents in pooled samples, with a mean of 30.2 individuals con- Landingham et al.: Feeding habits of juvenile Pacific salmon 291 sumed by approximately 72% of the predators (Table 3). Identifiable fish (in order of importance) included smelt, Pacific sand lance. Pacific her- ring, and pricklebacks. Euphausiids were an important prey principally in inside inlets during July 1984. Incidental prey included decapod lar- vae, squid, and hyperiid amphipods. Variation in principal prey Diets of all four species varied greatly (Table 4). Variation in the FO of the six principal prey (calanoids, decapod larvae, euphausiids, fish, hyperiid amphipods, and tunicates) was great- est when compared between years: the FO of most invertebrate taxa decreased significantly (P<0.05) in the diets from August 1983 to August 1984, whereas the FO of teleost prey increased (Tables 3 and 4). Except for coho salmon, variation in the FO of fish was greatest for inshore-offshore comparisons; significantly more fish were consumed by the other species offshore (>37 km from the coast) than inshore (0-37 km from the coast) (Table 4; 12.690.60) was found between pink and sockeye salmon, pink and chum salmon, and chum and sockeye salmon when all samples were pooled. Coho salmon diet overlap was less than 0.60 in all comparisons. Diet overlap was not calculated for chinook salmon because of small sample sizes (Table 1). The greatest diet overlap was between pink and sockeye salmon juveniles ( CA =0.93; Table 5); both species ate fish and hyperiid amphipods. The sec- ond-greatest diet overlap ( Cx =0.91) was between pink 292 Fishery Bulletin 96(2), 1998 Inside water (IRI) Outside waters and (partitioned into inlets and passages) outer coast inlets (IRI) ■■ Calanoid copepods Euphausiids r l Hyperiid amphipods CFJ Decapod larvae ESS Fishes F13 Tunicates Figure 4 Index of relative importance (IRI) of principal prey of juvenile pink salmon in inside and outside waters and outer coast inlets of southeastern Alaska in 1983 and 1984 and outside waters and outer coast inlets of northern British Columbia in 1984. and chum salmon; both ate hyperiid amphipods, fish, and Oikopleura dioica. The third-highest diet overlap in all pooled samples was between chum and sockeye salmon ( CA =0.73); both species consumed fish, hyperiid amphipods, and euphausiids. Temporal diet overlap The degree of diet overlap among salmon species pairs varied by time period when habitat was not considered. In August 1983, diet over- lap, based on the consumption of hyperiid amphipods, was greatest for pink and chum salmon juveniles ( Q =0.95; Table 5), O. dioica, and euphausiids. Diets of juvenile pink and sockeye salmon overlapped signifi- cantly ( CA- 0.72) on the basis of hyperiid amphipods and euphausiids. Alesser, but still significant, diet over- lap (C?= 0.62) occurred between sockeye and coho salmon on the basis of hyperiid amphipods, decapod larvae, and fish. Diet overlap was lower in July 1984, when samples included waters outside of British Columbia, than in August 1984. Diet overlap was significant for combi- nations of pink, chum, and sockeye salmon, but not for combinations including coho salmon (Table 5). The greatest overlap occurred between pink and chum salmon ( CA =0.80), pink and sockeye salmon ( CA =0.79), and chum and sockeye salmon ( CA=0.73), which was principally due to similar proportions of hyperiid am- phipods and juvenile fish. The relative importance of larvaceans accounted for the greatest difference in the diets of these three species: O. dioica made up nearly 20% of the IRI for chum salmon, compared with less than 3% for pink and sockeye salmon. In August 1984, the diets of pink, sockeye, and chum salmon overlapped greatly, whereas coho salmon diet did not overlap with that of any other species (Table 5). Diets of pink and sockeye salmon overlapped almost Landingham et a I.: Feeding habits of juvenile Pacific salmon 293 Inside water (IRI) (partitioned into inlets and passages) Outside waters and outer coast inlets (IRI) 80 60 40 20 0 20 40 60 80 Calanoid copepods Decapod larvae FTTT1 Euphausiids CZU Hyperiid amphipods E23 Fishes I" 1 "3 Tunicates Figure 5 Index of relative importance (IRI) of principal prey of juvenile chum salmon in inside and outside waters and outer coast inlets of southeastern Alaska in 1983 and 1984 and outside waters and outer coast inlets of northern British Columbia in 1984. completely ( (A =0.99); prey composition and IRI for each prey category were highly similar (Table 2). Pink and chum salmon diets ( (^=0.88) and sockeye and chum salmon diets ( Cx= 0.85) had almost identical prey spe- cies compositions, although proportions of prey catego- ries differed (Table 2). Spatial diet overlap Diet overlap was more com- mon for species comparisons in outside waters than in inside waters. Of all possible habitat comparisons, diet overlap was significant ( CA>0.60) in 42% of the outside-waters comparisons, 39% of inside-passage comparisons, and 29% of inside-inlet comparisons (Table 5). Within each habitat, mean overlap (all pe- riods) was significant in inside inlets for pink and chum salmon ( CA=0.66), inside passages for pink and chum salmon (CA=0.91), and chum and sockeye salmon ( CA=0.73), and outside waters for pink and sockeye salmon ( CA =0.83), chum and sockeye salmon ( CA=0.68), and pink and chum salmon ( CA =0.67). Fullness and digestion Most salmon stomachs were evaluated to be at least half full (fullness index of 3); coho salmon stomachs were more full than those of other species (Table 6). Stomach fullness was always greater for fish from the inside waters than from outside waters; stom- achs of fish caught 0-37 km offshore were less full than those of fish caught >37 km offshore. Only 32 of 2,210 (1.4%) stomachs sampled were empty: 72% in outside waters, 3% in coastal inlets, and 25% in in- side waters. The contents of most stomachs were “partly digested” (digestion index of 2). Pink, chum, and coho salmon from inside waters had stomach contents in an earlier stage of digestion than fish 294 Fishery Bulletin 96(2), 1 998 Inside water (IRI) (partitioned into inlets and passages) Outside waters and outer coast inlets (IRI) 80 60 40 20 0 20 40 60 80 i i INLET AUGUST 1983 1 " " ll NO DATA I NO DATA — 1 1 | INLET AUGUST 1983 AUGUST 1983 AUGUST 1983 ■ r n = 9 n = 44 tS».'U:'v ...Tl t INLET JULY 1984 \T. NO DATA INLET JULY 1984 n= 12 ■ N. BRITISH COLUMBIA n = 55 l nOOnuL JULY 1984 JULY 1984 n = 9 1 J u- n = 83 INLET AUGUST 1984 NO DATA INLET AUGUST 1984 n« 10 i 1 PASSAGE ■ AUGUST 1984 AUGUST 1984 1 1 n-80 n= 5 C 0-37 KM POOLED HABITATS r AUGUST 1984 AND YEARS frTTTT 1 n- 52 n = 361 1...... t: ''T :T:V 1f ‘ ” ' " 1 EM ■■ Calanoid copepods HTTP i 1 Decapod larvae ► — J Euphausiids Fishes L—J Hyperiid amphipods E3 Tunicates Figure 6 Index of relative importance (IRI) of principal prey of juvenile sockeye salmon in inside and outside waters and outer coast inlets of southeastern Alaska in 1983 and 1984 and outside waters and outer coast inlets of northern British Columbia in 1984. from outside waters; the reverse was observed for sockeye salmon. Salmon diet and the prey assemblages Neuston samples from the outside waters of British Columbia and southeastern Alaska included 13 and 14 major taxa, respectively (Table 7). In British Co- lumbia, decapod larvae made up 83% of the number of prey, hyperiid amphipods 12%, calanoid copepods 4%, and all other taxa <1%. Prey diversity of neus- ton samples was more even for southeastern Alaska than for British Columbia; half of the 14 taxa repre- sented >2% of the total abundance. In Alaska, calanoid copepods were the most abundant organ- ism (59% of the total), decapod larvae were second (21%), and gammarid amphipods were third (9%). The density of neustonic organisms was about five times greater in samples from British Columbia than in those from southeastern Alaska. In zooplankton samples, the number of individuals and dominant taxa differed with time and habitat (Table 7). Zooplankton abundance in outside waters was about twice that of inside waters of southeastern Alaska in July 1984. Calanoid copepods were the dominant organisms in the samples (90% of total abundance). The PSI indicated little relationship between prey consumed by salmon and prey available in the envi- ronment (Table 8). Neuston samples were generally more similar to salmon diet than were zooplankton samples. Similarity for the neuston samples ranged from 0% to 37.8% and averaged from 9.1% for sock- eye salmon to 16.7% for coho salmon (Table 8). The PSI values for the plankton samples ranged from 0.2% to 24.9% and averaged from 2.2% for chum salmon to 9.3% for sockeye salmon. Landingham et al.: Feeding habits of juvenile Pacific salmon 295 Inside water (IRI) (partitioned into inlets and passages) Outside waters and outer coast inlets (IRI) I J Decapod larvae I — I Euphausiids E23 Fishes t— J Hyperiid amphipods i * Tunicates Figure 7 Index of relative importance (IRI) of principal prey of juvenile coho salmon in inside and outside waters and outer coast inlets of southeastern Alaska in 1983 and 1984 and outside waters and outer coast inlets of northern British Columbia in 1984. We examined patterns of prey selectivity by ap- plying Strauss’s linear index of food selection to the same sets of predator-prey samples that had been analyzed for similarity. Pink and sockeye salmon selected (L>0.10) neustonic prey more often than planktonic prey, and neustonic hyperiid amphipods were the most frequently selected organism. Chum and coho salmon selected neustonic and planktonic prey in nearly equal frequencies (Table 9). All salmon species avoided (L <— 0. 10 ) neustonic decapod larvae. Selection patterns for planktonic prey varied more than for neustonic prey; however, salmon generally selected planktonic decapod larvae, hyperiid amphi- pods, euphausiids, and fishes and avoided the pro- portionately more abundant calanoid copepods. Discussion This study of the food habits of five sympatric spe- cies of Pacific salmon during their first summer in the marine waters of southeastern Alaska and north- ern British Columbia is the first detailed study for this geographic area. Feeding patterns were dynamic, with shifts in the important prey categories between salmon species, years, and areas. We analyzed the importance of two major prey cat- egories, zooplankton and teleosts, and examined data for temporal and spatial shifts in feeding patterns among salmon species. First, in pooled samples (all periods and habitats), zooplankton were much more important in pink, chum, and sockeye salmon diets (60-70% IRI) than in coho and chinook salmon diets (36% and 24% IRI; Table 3). When the two years of data were analyzed separately, however, contrasting patterns emerged. In 1983 (August), pink, chum, and sockeye salmon were mostly planktivorous (83-97% IRI ), and zooplankton IRI was over 50% even for coho salmon. In 1984 (July and August), the proportional number, weight, and FO of teleost prey increased in diets of all salmon species; however, whereas zoo- plankton remained most important in the diets of pink, 296 Fishery Bulletin 96(2), 1 998 chum, and sockeye salmon (56-65% IRI), teleosts were most important in coho salmon diets (68%IRI). Spatial factors that influence both predator and prey, especially latitude and proximity to shore, may account for similarities or differences in principal prey (Andrievskaya, 1970; Brodeur and Pearcy, 1990). Most species of salmon are opportunistic and feed on a wide variety of prey (Beacham, 1986; Brodeur1), but their diets are commonly composed of a few taxa readily available at a given time and location. The importance of zooplankton in the diets of juvenile pink, sockeye, and chum salmon in our study is simi- lar to results reported in other studies from the east- ern Gulf of Alaska (Manzer, 1969; Jaenicke et al., Table 4 Chi-square (^2) values for variation in frequency of occur- rence of principal prey in four species of juvenile salmon. Comparisons were made between years (August 1983 and August 1984), months (July 1984 and August 1984), area (northern British Columbia and southeastern Alaska), and distance offshore (0-37km vs. 46-76km). No asterisk indi- cates P > 0.05. * = P<0.05; ** = P<0.01; and *** = P<0.001. X2 Between inshore Between Between Between and Taxon years months areas offshore Pink salmon Calanoids 35.29*** 0.07 2.47 0.35 Decapods 18.41*** 0.07 1.82 13.04*** Euphausiids 11.35*** 2.90 17.13*** 3.86* Hyperiids 2.34 0.62 1.18 8.60*** Tunicates 7.13** 19.48*** 6.08* 13.83*** Teleosts 6.39* 0.00 7.33** 47 72*** Chum salmon Calanoids 8.62*** 0.25 6.08* 0.37 Decapods 4.02* 6.63* 2.04 5.94* Euphausiids 9.06*** 1.74 5.66* 5.15* Hyperiids 4.93* 0.01 0.12 15.30*** Tunicates 3.78 1.77 3.43 1.36 Teleosts 3.53 0.53 6.94** 26.46*** Sockeye salmon Calanoids 6.84** 3.73 3.72 0.08 Decapods 14.77*** 26.12*** 1.74 1.36 Euphausiids 19 47*** 13.21*** 14.81*** 0.68 Hyperiids 73 17*** 57.10*** 0.34 0.04 Tunicates 0.72 0 77*** 5.45* 4.51* Teleosts 11.05*** 0.09 13.42*** 12.69*** Coho salmon Calanoids 9.26*** 4.35* — — Decapods 14.96*** 20.50*** 0.00 1.46 Euphausiids 0.80 0.07 7.20** 3.98* Hyperiids 48.46*** 0.11 3.86* — Tunicates 0.86 7.00** 0.33 — Teleosts 5.77* 6.90** 1.84 0.59 1984; Hartt and Bell, 1986), the Sea of Okhotsk (Andrievskaya, 1968, 1970), and other regions of the northeastern Pacific Ocean (Peterson et al., 1982; Brodeur, 1989; Brodeur and Pearcy, 1990). We found that hyperiid amphipods were especially important in pink and sockeye salmon diets, and less so in chum salmon diets. The principal hyperiid amphipod found in diets of these species in the northeastern Pacific Ocean, Iincluding our study, was Themisto pacifica. In the more southern latitudes of coastal Washing- ton and Oregon, another hyperiid amphipod ( Hyperoche medusarum ) is an important component in chum, coho, chinook, and sockeye salmon diets (Brodeur and Pearcy, 1990). In that area, however, euphausiids are more important in juvenile chum diets and, to a lesser extent in juvenile coho and chinook salmon diets, than are hyperiid amphipods (Peterson et al., 1982; Brodeur and Pearcy, 1990). Similarly, we found that euphausiids occurred less frequently and were less important in diets of all juvenile salmon located off southeastern Alaska, a downwelling region, compared to diets of salmon from southern British Columbia to Oregon, where up- welling is more prevalent (Ware and McFarlane, 1989). Although our study showed that chum salmon are primarily planktivorous, like pink and sockeye salmon, we observed another difference in their most important prey. Tunicates (larvacea and salps) were important only in the diets of chum salmon. The IRI for tunicates, especially the larvacean Oikopleura dioica, was greater than for any other taxon in chum salmon diets (30. 1%). Tunicates were most prevalent in samples from outer coast inlets and outside wa- ters (0-37 km). This finding is not unusual. Oikopleura was the dominant prey item in two other studies of the diet of juvenile chum salmon from northern British Columbia: 62% FO in fish 32-106 mm FL from Chatham Sound (Manzer, 1969) and 70- 76% FO in fish 105-158 mm FL from Hecate Strait (Healey, 1991). In addition to interspecific differences in the most important prey, shifts in diet differed among the four salmon species both spatially and temporally. We observed spatial differences in July 1984, when sam- pling included waters of northern British Columbia, and in August 1984, when sampling was extended to 76 km offshore. Pink, chum, and sockeye salmon con- sumed fish prey about twice as often offshore as they did inshore, and more often in British Colombia than in Alaska. The more piscivorous coho salmon, how- ever, consumed fish just as frequently inshore as off- shore and just as frequently in marine waters of Alaska and British Columbia. Diet varied by season and distance offshore for the highly similar pink, Landingham et al.: Feeding habits of juvenile Pacific salmon 297 Table 5 Diet overlap values ( CA, Morisita’s index for prey numbers; Horn, 1966) for inside and outside waters of British Columbia (BC) and Alaska by sampling period and pooled over time. Nearshore = 0-37 km; offshore = >37 km. An asterisk indicates a significant value. Diet overlap ( CA ) July 1984 August 1984 Pooled Comparison Aug 1983 BC Alaska Nearshore Offshore over time Inside inlets Pink-chum 0.55 0.87* 0.56 0.66* Pink-coho 0.89* 0.04 0.29 0.41 Pink-sockeye — 0.21 0.15 0.37 Chum-coho 0.47 0.11 0.14 0.24 Chum-sockeye — 0.25 0.85* 0.51 Coho-sockeye Inside passages — 0.64* 0.01 0.55 Pink-chum 0.94* — 0.91* 0.91* Pink-coho 0.54 0.11 0.09 0.25 Pink-sockeye 0.85* 0.49 — 0.58 Chum-coho 0.45 — 0.20 0.25 Chum-sockeye 0.74* — — 0.73* Coho-sockeye Outside waters 0.65* 0.46 — 0.49 Pink-chum — 0.44 0.82* 1.00* 0.85* 0.67* Pink-coho 0.26 0.24 0.10 0.12 0.11 0.20 Pink-sockeye 0.77* 0.85* 0.97* 0.67* 0.79* 0.83* Chum-coho — 0.96* 0.10 0.10 0.09 0.32 Chum-sockeye — 0.39 0.82* 0.53 0.97* 0.68* Coho-sockeye All habitats 0.16 0.20 0.17 0.13 0.09 0.17 Pink-chum 0.95* 0.80* 0.88* 0.91* Pink-coho 0.43 0.21 0.18 0.38 Pink-sockeye 0.72* 0.79* 0.99* 0.93* Chum-coho 0.31 0.15 0.25 0.24 Chum-sockeye 0.58 0.73* 0.85* 0.73* Coho-sockeye 0.62* 0.37 0.18 0.49 chum, and sockeye salmon in the Sea of Okhotsk and the Bering Sea (Andrievskaya, 1968). In our study, temporal patterns based on seasonal (monthly) dif- ferences in the FO of fish prey were rare; the diets of pink, chum, or sockeye salmon included fish as fre- quently in July 1984 (when transects extended only to 37 km) as in August 1984. However, interannual differences were observed: all species, except chum salmon, ate fish more frequently in August 1984 than in August 1983. These observations suggest that in- creased frequency of fish prey in salmon diets is re- lated more to annual variations in teleost prey abun- dance than to distance offshore. Juvenile fishes are often identified as important prey for juvenile salmon, although seldom to the ex- tent that we observed in 1984 in northern southeast- ern Alaska. Comparison of our results with other studies of juvenile salmon diet in Alaska suggests that, although pink, chum, and sockeye salmon feed principally on planktonic taxa, they readily switch to teleost prey when available. For example, crusta- ceans were the principal prey of pink, chum, and sock- eye salmon in the outside waters of southeastern Alaska in 1982, whereas fish were the principal prey of coho salmon ( Jaenicke etal., 1984). In our study in southeastern Alaska in 1983, prey fish made up only half as much biomass as that observed in juvenile sockeye salmon stomachs collected from the Gulf of Alaska and Bering Sea in 1967 and 1968 (Hartt and Bell, 1986); in 1984, however, we recorded substan- tially higher teleost prey biomass than that observed by Hartt and Dell (TablelO). Although we found that sockeye salmon diet varied more than that of the other species, increases in the FO of predation on teleosts were consistent among pink, chum, and sock- eye salmon. 298 Fishery Bulletin 96(2), 1998 Some of the interspecific differences in the utiliza- tion of fish prey may also be attributed to differences in predator size. The mean size of juvenile coho and chinook salmon was greater than that of pink, chum, and sockeye salmon. Although the importance of size- related variability in prey consumption of juvenile salmon is recognized (e.g. Brodeur, 1991), we consid- ered that a detailed analysis by predator size was beyond the scope of this study. Increased fish prey in the diets of juvenile salmon in 1984 may reflect higher abundances of larval and juvenile stages of certain prey species compared with the previous year. Unfortunately, most fish prey spe- cies found in the salmon stomachs in our study could not be identified. However, we noted opposite trends in the CPUE for two potential teleost prey species from 1983 to 1984: the CPUE of juvenile walleye pol- lock in our seine catches increased 25-fold whereas that of juvenile herring declined 40-fold. In 1984, ju- venile walleye pollock were important in the diet of adult coho salmon (Fisk3); the conditions that pro- longed offshore feeding by adult coho salmon on these fish in 1984 may also have favored increased piscivory by juveniles. An increase in available fish prey between years may also correlate with an increase in environmen- tal temperature. The average SST in 1984 ( x =14.5°C) was higher than in 1983 ( x =13.7°C). Increased SST in northern waters may have beneficial effects on the early life history of fish preyed upon by juvenile salmon (Bailey and Incze, 1985). Temperatures in outside waters in 1984 followed a long-term warm- ing trend related to the 1982-83 El Nino event. Tem- perature increased at depth as well as at the sur- face, and positive temperature anomalies persisted in northern waters beyond 1983 (Cannon et al., 1985; Royer, 1985). In coastal waters off Oregon and Wash- ington, oceanographic conditions varied greatly ow- ing to the 1982-83 El Nino event and affected prey species composition, but the proportion of fish bio- mass in juvenile salmon diets generally did not vary for the same months between years (Brodeur and Pearcy, 1990). Changes in salmonid diet patterns may reflect den- sity-dependent species interactions. Other workers have noted such changes with respect to an increase in the density of pink salmon, typically the most abundant species: 1) the diet of other salmonids be- came more diverse, particularly in chum salmon; and 2) diet overlap among pink, chum, and sockeye salmon decreased (Birman, 1969; Andrievskaya, 3 Fisk, G. 1985. Final report 1984 troll logbook program. Alaska Trollers Assoc., 130 Seward St., No. 213, Juneau, AK 99801, 41 p. Table 6 Estimated mean fullness (0=empty, 6=distended), degree of digestion ( l=fresh, 4=completely digested), and percent- age of empty stomachs for pooled habitats and time peri- ods for juvenile salmon collected in marine waters of south- eastern Alaska and northern British Columbia in 1983 and 1984; n = sample size. Habitat n Stomach fullness (0-6) Degree of digestion (1-4) Empty stomachs (%) Pink salmon Inside inlet 70 4.0 1.9 0 Inside passage 223 3.9 2.3 0.4 Outer coast inlet 68 3.5 2.7 1.5 Outside (0-37 km) 361 3.0 2.2 3.3 Outside (>37 km) 93 4.0 2.3 0 Chum salmon Inside inlet 56 4.2 2.3 0 Inside passage 125 4.7 2.4 0 Outer coast inlet 62 4.6 2.2 0 Outside (0-37 km) 169 2.6 2.7 5.3 Outside (>37 km) 41 3.3 2.4 0 Sockeye salmon Inside inlet 24 4.3 2.5 0 Inside passage 58 4.5 3.0 1.7 Outer coast inlet 0 — — — Outside (0-37 km) 227 3.4 2.3 0.9 Outside (>37 km) 52 3.6 2.5 0 Coho salmon Inside inlet 159 4.2 2.3 1.3 Inside passage 205 4.4 2.3 0.5 Outer coast inlet 52 4.2 2.5 0 Outside (0-37 km) 115 4.0 2.4 0 Outside (>37 km) 12 4.3 3.0 0 1970; Tadokoro et al., 1996). In the Sea of Okhotsk, diet overlap was lower in the coastal zone, where salmon density was greatest (Andrievskaya, 1970). We did not observe such density effects. We found significant diet overlap even in the four cases out of five where density effects could have been demon- strated among these species (see Jaenicke and Celewycz, 1994). The exception, when chum salmon diet did not overlap significantly with either pink or sockeye salmon diets, occurred during July 1984 in the outside waters of British Columbia. Density-dependent effects on diet may be reflected in a third dietary attribute, the amount of food con- sumed. In the eastern coastal zone of the Sea of Okhotsk, for example, 30% of juvenile pink salmon and 66% of juvenile chum salmon sampled with gill nets (time of day not presented) had empty stom- achs, a condition not found offshore, where juvenile salmon density was lower (Andrievskaya, 1970). Feeding conditions appeared to be much better for Landingham et a I.: Feeding habits of juvenile Pacific salmon 299 Table 7 Neuston and zooplankton organisms (number per 100 m3) from outside- and inside-water habitats in British Columbia (BC) and Alaska, collected in July 1984. Neuston Zooplankton Outside Outside BC Combined Alaska BC Alaska Inside Alaska Number of samples 20 16 3 9 5 Invertebrate eggs — — 343.00 1,449,78 304.20 Polychaeta 0.13 0.13 — — — Gastropoda Limacina helicina 124.67 243.67 62.80 Bivalvia — — — — 26.60 Cephalopoda 0.05 0.13 31.00 — — Cladocera — — 145.33 524.00 234.20 Calanoida 39.77 114.56 48,672.34 60,074.67 25,956.00 Harpacticoida 0.50 4.31 0 34.33 5.20 Cirripedia 1.90 0.75 187.00 3,203.44 543.40 Mysidacea — 0.13 0 579.89 — Gammariidea 0.92 18.17 0 2.22 — Hyperiidea 126.12 5.21 436.33 566.33 1,142.00 Euphausiacea 8.86 0.17 4,072.70 126.78 164.80 Decapoda 873.11 40.18 363.67 510.33 230.00 Insecta 0.21 0.21 — - 3.00 Chaetognatha 0.34 5.63 249.33 171.67 42.20 Larvacea 0.15 5.00 62.33 262.78 204.00 Teleosts2 1.46 0.79 31.00 6.11 — 1 Eggs, larvae, and juveniles combined. salmon in our study (seined during the day); of 2,216 stomachs examined, only 32 (1.4%) were empty. Al- though we found few empty stomachs, we did find evidence of decreased feeding in the 0-37 km region of outside waters, which includes the region of peak catches of pink and chum salmon (Jaenicke and Celewycz, 1994); fullness was lower for all species, and empty stomachs were most common for pink and chum salmon collected in these outside waters. Juvenile coho salmon diet rarely showed signifi- cant overlap with pink, chum, and sockeye salmon diets. Coho salmon differed in distribution and size from the other three species (Jaenicke and Celewycz, 1994) and were more piscivorous. Curves of cumula- tive number of taxa (Fig. 3) sloped more steeply for coho salmon than for the other species; this rapid rate of increase in prey types may reflect opportu- nistic feeding of juvenile coho salmon. Diets are even more diverse among coho salmon individuals than among pink salmon individuals, which consumed a similarly broad array of prey taxa. Less aggregation and lower densities (CPUE) than those for other salmon (Jaenicke and Celewycz, 1994) allow coho salmon to exploit fully all locations within a habitat type. The other three species tended to occur in fewer Table 8 Mean percent similarity index (PSI) values indicating amount of overlap between prey fields and juvenile salmon diet. Stations where prey were collected were included in the analysis for a particular salmon species if at least five stomachs were collected at the station. Similarity values Number of stations Mean Range Neuston collection Pink 8 10.8 0-24.6 Chum 2 9.3 0-18.5 Sockeye 8 9.1 0-37.8 Coho 2 16.7 0-33.3 Zooplankton collection Pink 8 8.0 1.0-22.7 Chum 5 2.2 0.2-6. 8 Sockeye 4 9.3 1.0-24.9 Coho 6 6.2 0.7-11.7 locations (although they were well represented within habitat types) and tended to be more highly aggregated, increasing the tendency toward more 300 Fishery Bulletin 96(2), 1998 Table 9 Prey selection frequencies (Strauss’s linear index of food selection (L); Strauss 1979) in juvenile salmon diet for neuston and zooplankton collections, July and August 1984. A positive (pos. ) value indicates that a prey was more abundant in the diet than in the environment at a level of >0.10. A negative (neg.) value indicates that a prey was less abundant in the diet than in the environment at a level of <0.10. A random (ran.) value indicates that a prey was about equally abundant in the diet and in the environment. Absent indicates that a prey item was absent from the diet and from the environment in a particular location. N = number of groups tested, where group size was 5 or more of a species. Prey selection frequency Pink salmon Chum salmon Sockeye salmon Coho salmon Taxon Pos. Neg. Ran. Absent Pos. Neg. Ran. Absent Pos. Neg. Ran. Absent Pos. Neg. Ran. Absent Neuston Polychaetes 0 0 1 7 0 0 0 2 0 0 0 8 0 0 0 2 Pteropods 1 0 2 5 0 0 2 2 0 0 0 8 0 0 0 2 Squid 0 1 0 7 0 0 0 2 0 0 0 8 0 0 0 2 Calanoid copepods 1 1 4 2 0 0 0 2 4 2 1 1 0 0 0 2 Barnacle larvae 0 0 0 8 0 0 0 2 0 1 0 7 0 0 0 2 Gammarid amphipods 0 2 2 4 0 1 0 1 0 1 1 6 0 1 0 1 Hyperiid amphipods 5 1 1 1 1 0 0 1 6 2 0 0 0 1 0 1 Euphausiids 4 1 1 2 0 0 1 1 3 1 1 3 0 0 1 1 Decapod larvae 2 4 1 1 0 2 0 0 2 3 1 2 1 1 0 0 Insects 0 0 1 7 0 0 0 2 0 0 0 8 0 0 0 2 Chaetognaths 0 0 2 6 1 0 0 2 0 0 1 7 0 0 0 2 Salps 0 0 0 8 0 0 0 2 0 0 0 8 0 0 1 1 Larvaceans 1 0 3 4 1 0 0 1 1 0 0 7 1 0 1 0 Teleosts 1 0 5 2 0 0 1 1 0 0 5 3 1 0 0 1 Percent selection 13.5 9.0 20.7 57.4 7.2 10.7 14.3 74.8 14.4 9.0 9.0 67.9 10.8 10.8 10.8 67.7 Plankton Cnidaria 0 0 3 5 0 0 2 3 0 0 2 2 0 0 1 5 Polychaetes 0 0 0 8 0 0 1 4 0 0 0 4 0 0 1 5 Pteropods 1 0 1 6 0 0 0 5 0 0 1 3 0 0 1 5 Bivalve larvae 0 0 0 8 0 0 0 5 0 0 0 4 0 1 1 4 Squid 0 0 0 8 0 0 0 5 0 0 1 3 0 0 0 6 Invertebrate eggs 0 2 1 5 0 1 0 4 0 1 0 3 0 0 1 5 Calanoid copepods 0 7 0 1 0 4 0 1 0 4 0 0 0 6 0 0 Barnacle larvae 0 0 2 6 0 0 1 4 0 1 1 2 0 1 1 4 Mysids 0 0 0 8 0 0 0 5 0 0 0 4 0 0 0 6 Hyperiid amphipods 2 0 2 4 2 0 1 2 2 0 1 1 1 0 4 1 Euphausiids 2 0 2 4 1 0 1 3 1 0 1 2 2 0 3 1 Decapod larvae 4 0 1 3 1 0 2 2 2 0 1 1 5 0 1 0 Insects 0 0 0 8 0 0 0 5 0 0 0 4 0 0 2 4 Cladocera 0 0 0 8 0 0 0 5 0 0 1 3 0 0 2 4 Chaetognaths 0 0 2 6 0 0 0 5 0 0 2 2 0 0 1 5 Salps 0 0 1 7 0 0 1 4 0 0 0 4 0 0 1 5 Larvaceans 2 0 1 5 2 0 1 2 0 0 2 2 1 0 3 2 Teleosts 1 0 2 5 1 0 1 3 1 0 3 0 3 0 2 1 Percent selection 8.4 6.3 12.6 73.5 7.7 5.5 12.1 73.7 8.4 8.4 21.0 61.6 11.1 7.4 23.1 58.2 similar diets. Many seine sets caught only coho salmon, whereas catches of the more highly aggre- gated species often also contained coho salmon. In laboratory experiments, coho salmon smolts in sea- water demonstrated agonistic behavior which, if oc- curring in the wild, would maintain discrete feeding territories and a dispersed population (Paszkowski and Olla, 1985). Differences in diet, distribution, and size indicate that juvenile coho salmon have a distinct feeding ecology in comparison with these other, more plank- tivorous, juvenile salmon co-occurring in the south- ern Gulf of Alaska. In both our study and that of Brodeur and Pearcy (1990), fish were more impor- tant in the diets of juvenile coho salmon than in other species. During the second year of our study — when Landingham et al.: Feeding habits of juvenile Pacific salmon 301 Table 10 Principal prey of juvenile sockeye salmon in southeastern Alaska and northern British Columbia in 1983 and 1984 and in the Gulf of Alaska and Bering Sea in 1967 and 1968 (Hartt and Dell, 1986). Values are percent biomass of total diet. The number of fish sampled is shown in parantheses, followed by fork length ranges (mm). Prey Pooled 1967-68 (996) 130-180 Pooled 1983-84 (361) 100-300 August 1983 (55) 130-249 July 1984 (159) 100-209 August 1984 (147) 110-300 Copepods 5.1 0.5 4.6 0.4 0.5 Decapod larvae 2.6 8.3 0.8 5.3 Euphausiids 42.0 4.3 39.0 3.4 3.8 Amphipods 1.6 3.7 30.7 2.0 5.1 Pteropods 6.2 — — — — Larval fish 30.8 87.2 16.7 92.0 83.1 teleost prey increased in all diets — pink, chum, and sockeye salmon readily switched from small zoo- plankton to larger teleost prey in response to an ap- parent increase of available larval fish prey. The abil- ity of salmon species to maintain plasticity in their diets may be an adaptation to changing ocean condi- tions— one that may improve marine survival. On the basis of our comparisons between the prey composition of juvenile salmonids and the taxa found in neuston and plankton tows, we conclude that these juveniles are selecting a limited subset of available prey. Although other factors besides feeding prefer- ences (e.g. prey patchiness, gear selectivity, differ- ential digestion rates of prey) could lead to low se- lectivity values, certain taxa do appear to be con- sumed in high proportions in relation to their abun- dance. Our results suggest that juvenile salmon are visual predators and select prey on the basis of prey size and visibility and not on local abundance. For example, relatively rare hyperiid amphipods were se- lected by most salmon, whereas slightly smaller but much more numerous copepods were ignored. As ob- served by Peterson et al. (1982), this prey selection may be due to the heavy pigmentation and unusual swimming motion of hyperiids in contrast to the light pigmentation and fast swimming motion of copepods. Although the diets of both chum and coho salmon appear to be more similar to the neuston than the zooplankton catches, a reliance on neustonic fauna, as suggested by Brodeur (1989) for coho salmon off Washington and Oregon, is not conclusively demon- strated because of the small number of comparisons that we were able to make. More detailed field and laboratory studies are required to determine whether juvenile salmon show a reliance on certain prey or an inability to switch to alternate prey when pre- ferred prey resources are depleted. Under these con- ditions, the availability of the right kinds of prey may have more important implications for the survival of juvenile salmon in coastal waters than the overall production of prey. Acknowledgments We thank the biologists and technicians who helped in the field and laboratory, particularly P. D. Mothershead. We also thank the crews on the NOAA RV John N. Cobb and FV Bering Sea for their coop- eration during seining operations. The FV Bering Sea cruise was part of a cooperative coastwide survey from California to southeastern Alaska with W. Pearcy, Oregon State University. We especially ac- knowledge the review of the manuscript by A. Wer- theimer and significant editorial assistance from G. Duker. Literature cited Andrievskaya, L. D. 1968. Feeding of Pacific salmon fry in the sea. Izv. Tikhookean. Nauchno-Issled. Inst. Rybn. Khoz. Okeanogr. (TINRO) 64:73-80. [In Russ.; Engl, transl. 1970, Fish. Res. Board Can. Transl. Ser. 1423, 16 p.] 1970. Feeding of Pacific salmon juveniles in the Sea of Okhotsk. Izv. Tikhookean. Nauchno-Issled. Inst. Rybn. Khoz. Okeanogr. (TINRO) 78:105-115. [In Russ.; Engl, transl. 1973, Fish. Res. Board Can. Transl. Ser.2441, 20 p. I Bailey, K. M., and L. S. Incze. 1985. El Nino and the early life history and recruitment of fishes in temperate marine waters. In W. S. Wooster and D. L. Fluharty (eds.), El Nino North: Nino effects in the eastern subarctic Pacific Ocean, p.143-165. Washington Sea Grant WSG-WO-85-30, Univ. Washington, Seattle, WA, 312 p. Beacham, T. D. 1986. Type, quantity, and size of food of Pacific salmon ( Oncorhynchus ) in the Strait of Juan de Fuca, British Columbia. Fish. Bull. 84:77-89. Beamish, R. J., and D. R. Bouillon. 1993. Pacific salmon production trends in relation to climate. Can. J. Fish. Aquat. Sci. 50:1002-1016. Birman, I. B. 1969. Distribution and growth of young Pacific salmon of the genus Oncorhynchus in the sea. Probl. Ichthyol. 9:651-666. Brodeur, R. D. 1989. Neustonic feeding by juvenile salmonids in coastal waters of the Northeast Pacific. Can. J. Zool. 67:1995- 2007. 1991. Ontogenetic variations in the type and size of prey consumed by juvenile coho, Oncorhynchus kisutch, and chinook, O. tshawytscha, salmon. Env. Biol. Fishes 30:303-315. 302 Fishery Bulletin 96(2), 1998 Brodeur, R. D., and W. G. Pearcy. 1990. Trophic relations of juvenile Pacific salmon off the Oregon and Washington coast. Fish. Bull. 88:617-636. Brodeur, R. D., and B. M. Ware. 1995. Interdecadal variability in distribution and catch rates of epipelagic nekton in the Northeast Pacific Ocean. In R. J. Beamish (ed. ), Climate change and north- ern fish populations. Can. Spec. Publ. Fish. Aquat. Sci. 121, p. 329-356. Cannon, G. A., R. K. Reed, and P. E. Pullen. 1985. Comparison of El Nino events off the Pacific Northwest. In W. S. Wooster and D. L. Fluharty (eds.), El Nino North: Nino effects in the eastern subarctic Pacific Ocean, p. 75-84. Washington Sea Grant WSG-WO-85-3, Univ. Washington, Seattle, 312 p. Groot, C., and L. Margolis. 1991. Pacific salmon life histories. Univ. British Colum- bia, Vancouver, 564 p. Hartt, A. C., and M. B. Dell. 1986. Early oceanic migrations and growth of juvenile Pa- cific salmon and steelhead trout. Int. North Pac. Fish. Comm. Bull. 46, 105 p. Healey, M. C. 1991. Diets and feeding rates of juvenile pink, chum, and sockeye salmon in Hecate Strait, British Columbia. Trans. Am. Fish. Soc. 120:303-318. Helle, J. H. 1989. Relation between size-at-maturity and survival of progeny in chum salmon, Oncorhynchus keta (Walbaum). J. Fish Biol. 35(suppl. A):99-107. Helle, J. H., and M. S. Hoffman. 1995. Size decline and older age at maturity of two chum salmon ( Oncorhynchus keta) stocks in western North America, 1972-92. In R. J. Beamish (ed.), Climate change and northern fish populations. Can. Spec. Publ. Fish. Aquat. Sci. 121, p. 245-260. Horn, H. S. 1966. Measurement of “overlap” in comparative ecological studies. Am. Naturalist 100:419-424. Hurtubia, J. 1973. Trophic diversity measurement in sympatric preda- tory species. Ecology 54:885-890. Ishida, Y., S. Ito, M. Kaeriyama, S. McKinnell, and K. Nagasawa. 1993. Recent changes in age and size of chum salmon ( Oncorhynchus keta) in the North Pacific Ocean and pos- sible causes. Can. J. Fish. Aquat. Sci. 50:290-295. Jaenicke, H. W., and A. G. Celewycz. 1994. Marine distribution and size of juvenile Pacific salmon in Southeast Alaska and northern British Columbia. Fish Bull. 92:79-90. Jaenicke, H. W., R. D. Brodeur, and T. Fujii. 1984. Exploratory gillnetting from the Oshoro-maru for juvenile salmonids off southeastern Alaska, 24-25 July 1982. Bull. Fac. Fish. Hokkaido Univ. 35(3):154-160. Kaeriyama, M. 1989. Aspects of salmon ranching in Japan. Physiol. Ecol. Japan, spec. vol. 1:625-638. Manzer, J. I. 1969. Stomach contents of juvenile Pacific salmon in Chatham Sound and adjacent waters. J. Fish. Res. Board Can. 26:2219-2223. McNeil, W. J., and D. C. Himsworth (eds.) 1980. Salmonid ecosystems of the North Pacific. Oregon State Univ. Press, Corvallis, OR, 331 p. Paszkowski, C. A., and B. L. Olla. 1985. Social interactions of coho salmon ( Oncorhynchus kisutch ) smolts in seawater. Can. J. Zool. 63:2401-2407. Pearcy, W. G. 1992. Ocean ecology of North Pacific salmonids. Univ. Washington Sea Grant, Seattle, WA, 179 p. Peterson, W. T., R. D. Brodeur, and W. G. Pearcy. 1982. Food habits of juvenile salmon in the Oregon coastal zone, June 1979. Fish. Bull. 80:841-851. Pinkas, L., M. S. Oliphant, and I. L. K. Iverson. 1971. Food habits of albacore, bluefin tuna, and bonito in California waters. Calif. Dep. Fish Game, Fish. Bull. 152, 105 p. Royer, T. C. 1985. Coastal temperature and salinity anomalies in the northern Gulf of Alaska, 1970-84. In W. S. Wooster and D. L. Fluharty (eds.), El Nino North: Nino effects in the eastern subarctic Pacific Ocean, p. 107-115. Washington Sea Grant WSG-WO-85-3, Univ. Washington, Seattle, WA, 312 p. Strauss, R. E. 1979. Reliability estimates for Ivlev’s electivity index, the forage ratio, and a proposed linear index of food selection. Trans. Am. Fish. Soc. 108:344—352. Tadokoro, K., Y. Ishida, N. D. Davis, S. Ueyanagi, and T. Sugimoto. 1996. Change in chum salmon (Oncorhynchus keta) stom- ach contents associated with fluctuation of pink salmon (O. gorbuscha) abundance in the central subarctic Pacific and Bering Sea. Fish. Oceanog. 5(2):89-99. Ware, D. M., and G. A. McFarlane. 1989. Fisheries production domains in the Northeast Pa- cific Ocean. In R. J. Beamish and G. A. McFarlane (eds.), Effects of ocean variability on recruitment and an evalua- tion of parameters used in stock assessment models, p. 359- 379. Can. Spec. Publ. Fish. Aquat. Sci. 108. Wertheimer, A. C. 1997. Status of Alaska salmon. In D. J. Stouder, P. A. Bisson, and R. J. Naiman (eds.). Pacific salmon and their ecosystems, p. 179-197. Chapman and Hall, New York, NY, 685 p. Whittaker, R. H. 1975. Communities and ecosystems, 2nd ed. MacMillan Co., New York, NY, 385 p. 303 Abstract .—Northern searobins, Prionotus carolinus, and striped sea- robins, P. evolans, are commonly taken in bottom trawl, pound-net, and hook- and-line fisheries of the temperate, western North Atlantic. Seasonal move- ments, size structure, and interannual variations in population size of both species were examined with data from three fishery-independent trawl sur- veys. Geographic distributions of both species overlapped year-round, but on average, northern searobins were found in colder, deeper water than were striped searobins. Northern searobins were found north and east of Cape Cod during the warmest months and north- east of Hudson Canyon during the cold- est months, whereas striped searobins were rarely found in these regions dur- ing these seasons. Furthermore, north- ern searobins moved north and near- shore earlier in spring and left these habitats earlier in autumn. Modal size of adult northern searobins was gener- ally between 17 and 21 cm total length, about 5 cm shorter than that of adult striped searobins. Overall, northern searobins were collected more fre- quently and were more numerous than striped searobins. Because they were smaller and occurred in cooler waters, however, they were not dominant by weight in coastal habitats during au- tumn. During the past 25-30 years, the annual population size of both species has varied by an order of magnitude, fluctuating without any clear trend. Although seasonally abundant in coastal and continental shelf waters, searobins, because of their small size in relation to other species, are usually discarded and contribute little to land- ings from the mid-Atlantic states. Manuscript accepted 13 August 1997. Fishery Bulletin 96:303-314 ( 1998). Interspecific comparisons of searobin [Prionotus spp.) movements, size structure, and abundance in the temperate western North Atlantic Richard S. McBride Marine Field Station, Institute of Marine and Coastal Sciences Rutgers University 132 Great Bay Blvd., Tuckerton, New Jersey 08087 Present address: Florida Marine Research Institute 1 00 Eighth Avenue S.E. St. Petersburg, Florida 33701-5095 E-mail address: mcbride_r@harpo.dep. state. fl. us Joseph B. O'Gorman Northeast Fisheries Science Center National Marine Fisheries Service, NOAA Woods Hole, Massachusetts 02543 Kenneth W. Able Marine Field Station, Institute of Marine and Coastal Sciences Rutgers University 132 Great Bay Blvd., Tuckerton, New Jersey 08087 Searobins (family Triglidae) are commonly taken in pound-net and bottom trawl fisheries along the United States east coast but are only occasionally sold as foodfish, lobster bait, or livestock feed (e.g. Goode, 1888; Smith, 1894a). Com- mercial landings of 100-200 metric tons annually were reported from the mid-Atlantic states during the 1880’s (Collins, 1892), 1930’s (Mar- shall, 1946), and 1950’s (McHugh, 1977). Millions of searobins are also taken incidentally by hook-and-line recreational anglers (Table 1). Re- searchers have long recognized the potential for increasing U.S. land- ings by developing a searobin fish- ery (Smith, 1894b; Marshall, 1946; Merriman and Warfel, 1948; Perl- mutter, 1959-60), but the market potential for searobins is limited by their relatively small size and a false perception that they are poi- sonous or difficult to handle (John- son et al., 1987; Murray et al., 1987). Two searobin species are common in the temperate, western North At- lantic: northern searobin, Prionotus carolinus (Linnaeus, 1771), and striped searobin, P evolans (Lin- naeus, 1766). They range from Canada to northern Florida (Bige- low and Schroeder, 1953; Gilmore, 1977; Scott and Scott, 1988; Russell et al., 1992) but are most common year-round on the continental shelf from Cape Cod to Cape Hatteras (Nichols and Breder, 1926; Hilde- brand and Schroeder, 1928; Ed- wards et al., 1962). Life history in- formation is provided in Marshall (1946), Wong (1968), McEachran and Davis (1970), Richards et al. Contribution 98-03 of the Institute of Ma- rine and Coastal Sciences, Rutgers Univ., New Brunswick, New Jersey 08901-8521. 304 Fishery Bulletin 96(2), 1998 (1979), Yuschak and Lund (1984), Yuschak (1985), Keirans et al. (1986), and McBride and Able (1994). This study describes and compares the seasonal movements, size structure, and interannual varia- Table 1 Estimated numbers of searobins (all triglid species com- bined) caught by marine recreational anglers from 1979 to 1991 in the Mid-Atlantic, as well as in all subregions of the U.S. east coast combined: Gulf of Maine, Mid-Atlantic, South Atlantic, and Gulf of Mexico. Data prior to 1979 are not available. Numbers ( in millions) Mid- U.S. Year Atlantic east coast Source 1979 3.548 5.145 Holliday, 1984 1980 7.102 7.957 Holliday, 1984 1981 1.624 2.613 NMFS, 1985a 1982 2.795 5.074 NMFS, 1985a 1983 8.750 10.058 NMFS, 1985b 1984 6.189 6.752 NMFS, 1985b 1985 3.562 4.361 NMFS, 1986 1986 10.908 11.858 NMFS, 1987 1987 4.824 5.071 Essig et al., 1991 1988 5.182 5,899 Essig et al., 1991 1989 2.631 2.947 Essig et al., 1991 1990 4.896 5.518 Van Voorhees, et al.,1992 1991 7.254 7.799 Van Voorhees, et al.,1992 Mean (peryr) 5.328 6.235 Total 69.265 81.052 tions in population size of northern and striped searobins in the Mid-Atlantic Bight. No previous study has synthesized searobin data over such a large geographic scale. The present and potential roles of searobins in fisheries of the temperate western North Atlantic are also discussed. Materials and methods Bata were derived from three fishery-independent bottom trawl surveys (Table 2). One survey, con- ducted by the National Marine Fisheries Service, covered the entire Mid-Atlantic Bight (i.e. the coastal region from Cape Cod to Cape Hatteras), as well as Georges Bank and the Gulf of Maine (Fig. 1). Bata from two regional surveys were also examined: one at the northern extent of the Mid-Atlantic Bight near Cape Cod, the other within a central portion of the Mid-Atlantic Bight offshore of New Jersey. In each survey a stratified, random design was used for allo- cating trawl tows. Strata were established by depth categories and other physiographic boundaries. Num- bers of tows within each stratum were proportional to stratum area; within each stratum, tow locations were randomly assigned. Fishes were counted and weighed; measurements were reported to the near- est centimeter total length (TL). In preliminary analyses geographic and length-distribution data were plotted for all years available, but herein only data from selected cruises (1991-92) are graphed. Preliminary calculations of abundance showed that Table 2 Sources and details of trawl data examined for this study. Data for the period 1982-91 were the focus of most analyses. Data regarding distribution and size structure for the years 1991-92 were plotted (Figs. 2-7) as representative years. Data source Trawl headrope (m) (mesh [mm]) Tow time (min) Season Month Sampling depth (m) Years Throughout the Mid-Atlantic Bight NEFSC, NMFS7 18.3-19.8 30 Spring Mar-May 9-366 1968-95 (12.7) 30 Summer Jun-Aug 9-200 1963-65,69,77-81 30 Autumn Sep-Nov 9-366 1963-95 30 Winter Dec-Feb 9-200 1964-6,78,81,91-2 Offshore of Massachusetts and New Jersey MDMF2 11.9 20 Spring May 9-55 1978-95 (12.7) 20 Autumn Sep 9-55 1978-95 NJBMF3 24.4 204 5-6/year Jan-Dec 6-27 1988-92 (6.3) 1 Northeast Fisheries Science Center, National Marine Fisheries Service. 2 Massachusetts Division of Marine Fisheries. 3 New Jersey Bureau of Marine Fisheries. 4 Tows in 1988 were for 30 minutes. McBride et al.: Interspecific comparisons of searobin, Prionotus spp., movements, size structure and abundance 305 fish numbers and weight had similar trends (McBride, 1994); therefore only weight indices are reported here. Some descriptive statistics were also based on subsets of data (1982-91), as noted below. Statistical significance was evaluated at P<0.05. The above comments pertain to all data examined, and comments below out- line procedures specific to each sampling program. N.VfFS sampling Survey data from the National Marine Fisheries Service (NMFS), Northeast Fish- eries Science Center (NEFSC), were exam- ined. The survey area was divided into 76 strata from Cape Fear, North Carolina, to Cape Sable, Canada, in waters 9-366 m deep. Spring and autumn cruises sampled a total of about 350 stations during a 6-8 week period, whereas summer and winter cruises sampled about half as many sta- tions (Table 2). The primary gear used was a no. 36 Yankee trawl. In spring, summer, and autumn cruises a roller-rigged footrope was used; however, in winter cruises a cookie sweep and ground cable designed to target flatfish were used. Latitude, longi- tude, and depth were recorded at the start of each tow. Bottom temperature was usu- ally recorded by using an XBT cast or a CTD probe. See Grosslein (1969, 1976), Azarovitz (1981), and Bespres-Patanjo et al. (1988) for further sampling details. Exploratory data analyses considered all 76 strata, but interannual comparisons of abundance considered only latitudinal strata within the Mid-Atlantic Bight be- cause both species were uncommon or ab- sent in the Gulf of Maine and on Georges Bank and because sampling south of Cape Hatteras was not broad in coverage or con- sistent between years. Because the distri- bution of both species varied seasonally, largely as an onshore-offshore pattern, dif- ferent depth strata sets were used for cal- culating spring and autumn abundances; the autumn strata selected were generally <27 m and the spring strata were >27 m deep. Because the purpose here was to com- pare interspecific abundances, the final selection of strata included the broadest possible latitudinal range within the Mid- Atlantic Bight where both species occurred. 71 • 39' 62° 58' Figure 1 Study area of the U.S. east coast, with some landmarks iden- tified in the text (additional landmarks are depicted in Fig. 3). The Mid-Atlantic Bight is defined as the coastal region be- tween Cape Cod and Cape Hatteras. Sampling strata for the Mid-Atlantic Bight survey (NEFSC, NMFS, Table 2) extended in total from Cape Fear to Cape Sable. Sampling regions for the Massachusetts and New Jersey surveys (Mass. Div. Mar. Fisheries and New Jersey Bureau of Mar. Fisheries; Table 2) are depicted in Figures 3 and 4. 306 Fishery Bulletin 96(2), 1998 The autumn strata selected were NMFS inshore strata identification numbers 01— 61, and the spring strata were NMFS off- shore strata 01-12, 25, and 61-72. The summer and winter seasons were not sampled consistently enough to be in- cluded in these comparisons of abun- dance. Abundance of each species was estimated as a stratified mean weight per tow with the methods of Finney (1941) and Pennington (1983). These values are log-transformed (ln[x+l] ) means, weighted by stratum area. Regional sampling The Massachusetts Division of Marine Fisheries (MDMF) sampled 23 strata along the Massachusetts shoreline (<55 m). Approximately 80-90 stations were sampled in May and again in Sep- tember (Table 2) with a 3/4 North Atlan- tic type-2 seam trawl at random locations, except where precluded by untowable bottom and extensive fixed commercial gear (Howe, 1989). Abundance indices were calculated in the same manner as those calculated for the sampling by NMFS (see above). Strata were selected to encompass an area where both species occurred in each season (MDMF strata no. 10-19 for spring; 11-12 for autumn). The New Jersey Bureau of Marine Fisheries (NJBMF) sampled 15 strata along the New Jersey coast (<27 m). Cruises occurred every 6-10 weeks dur- ing 1988-92 (Table 2). Each cruise sampled 25-39 stations during a 1-4 week period using a 3-in-l trawl at ran- dom locations (Byrne, 1989). Mean weight per tow (ln[x+lp was calculated without weighting for strata areas, and tows from all strata were used because both species occurred throughout the NJBMF sampling area during most of the year. Results and discussion Movements Northern searobins, but not striped searobins, moved Bank (Figs. 2 and 3). Striped searobins occurred seasonally between the Mid-Atlantic Bight and the rarely north or east of Cape Cod in any year, but Gulf of Maine, and they used habitats on Georges northern searobins, although not common there, oc- Figure 2 Geographic distribution of northern searobins (open symbols, left column) and striped searobins (filled symbols, right column) in the Mid-Atlantic Bight during autumn, 1991 (top), winter, 1992 (middle), and spring, 1992 (bottom) based on Northeast Fish. Sci. Center, NMFS, resource surveys (Table 2). Much of the Gulf of Maine and Georges Bank are not depicted here because sampling there collected only a single northern searobin (at a station immediately north of Cape Ann; see Fig. 3 for landmark reference). McBride et al.: Interspecific comparisons of searobin, Prionotus spp., movements, size structure and abundance 307 Sound Rhode Island Sound May 1 992 Cape Ann Massachusetts Bay Cape Cod Bay Cape Cod Figure 3 Geographic distribution of northern searobins (open symbols) and striped searobins (filled symbols) during two consecutive sampling seasons for the Massachusetts coastal survey (Table 2). The dotted line represents the off- shore limit of sampling (55 m). curred annually on parts of Georges Bank and in the southern portion of the Gulf of Maine during sum- mer and autumn. In general, both species were broadly distributed within the central Mid-Atlantic Bight but moved southward during winter and spring (Figs. 2 and 4). Striped searobins moved well south and west of Hudson Canyon, which appears as an approximate northern limit for striped searobins during the overwintering season, whereas northern searobins had a broader latitudinal distribution in the Mid-Atlantic Bight while overwintering (see also McBride and Able, 1994). These data suggest that by spring large striped searobins moved either far- ther offshore into continental slope waters (>200 m) or migrated south of the zoogeographic boundary represented by Cape Hatteras (i.e. between the Mid- Atlantic and the South Atlantic Bight) because sizes > 25 cm TL are uncommon during spring cruises (Fig. 5). We cannot prove this hypothesis, however, because these southern and offshore regions were not well sampled during the overwintering season (see Table 2 for sampling area limits). Both searobin species migrated seasonally in an onshore-offshore direction as well as in a north- south direction (Fig. 2). Northern searobins were, however, consistently found in significantly deeper waters than were striped searobins during both spring (78.3 m ± 5.27 vs. 60.4 m ± 9.30; mean depth ± 95% confidence limits [CL] ) and autumn (30.7 m ± 1.73 vs. 20.0 m ± 1.44) during 1982-91. Northern searobins were also found in significantly colder bot- tom temperatures than were striped searobins in spring (9.2 ± 0.3°C vs. 11.8 ± 0.56°C; mean ± 95% CL) and autumn (17.6 ± 0.40°C vs. 19.6 ± 0.40°C), although both species were collected in a wide range of temperatures (4-28°C) during 1982-91. These re- sults support McBride and Abie’s (1994) postulation that temperature is the principal factor producing distinctive geographic distributions for each species. Other western North Atlantic searobins (Prionotus and Bellator species) that are distributed in subtropi- cal and tropical waters reside within species-specific depth regions but show little tendency to migrate seasonally (Lewis and Yerger, 1976; Ross, 1977; Floyd, 1980; Hoff, 1992). South of Cape Hatteras, where temperatures fluctuate less between seasons than in the Mid-Atlantic Bight, even northern searobins show little tendency to migrate seasonally (Floyd, 1980). Size structure Modal size of northern searobins was consistent be- tween seasons (17-21 cm TL), but striped searobins were both larger and more variable in size ( 18-28 cm TL; Fig. 5). The larger size ofP. evolans is largely accounted for by its faster growth rate and greater maximum age (Richards et al., 1979). An earlier analysis of NMFS survey data has reported north- ern searobins as large as 34 cm and striped searobins as large as 41 cm (Wilk et al., 1978). The maxima observed in our study were 5-7 cm larger, but these 308 Fishery Bulletin 96(2), 1 998 are most likely the result of much larger sample sizes (total for combined species was 1,314 fish measured in 1974-75 LWilk et al., 1978] vs. 69,072 in 1982- 91[our study]). Briggs ( 1977) reported a single record of a striped searobin measuring 485 mm TL, which equaled the size of our largest striped searobin. Floyd (1980) reported the largest northern searobin as 26 cm TL (n=1795) and the largest striped searobin as 35 cm TL (n- 47) based on trawl collections south of Cape Hatteras. These find- ings support Ginsburg’s ( 1950) observation that searobin size increases with higher latitudes. Young-of-the-year (YOY) of both species were evident in all surveys (Figs. 5-7), par- ticularly during autumn months, when a size of 15 cm TL was used to separate YOY from older year classes (Richards et al., 1979; McBride and Able, 1994). Collections from Massachusett waters did not indicate over- wintering by YOY searobins near Cape Cod, a finding similar to that reported by Richards et al. (1979). In New Jersey nearshore wa- ters, YOY northern searobins were evident in April, suggesting that they had migrated back inshore in this region much sooner than had been observed farther north. Abundance Northern searobins were more abundant than striped searobins by numbers and fre- quency of occurrence. A total of 66,064 north- ern searobins were collected in 1,125 tows, whereas only 3,008 striped searobins were collected in 366 tows out of a total of 7,369 NMFS tows during spring and autumn, 1982-91. Both species were caught together in 202 of these 7,369 tows. In Massachusetts waters, northern searobins were also more numerous than striped searobins in spring (88,565 vs. 75 fish) and autumn (10,966 vs. 374 fish) during 1982-91. And in New Jer- sey waters, northern searobins were more numerous than striped searobins (35,471 vs. 5,258 fish) in 700 tows during 1989-92. Northern searobin aggregate weight was not, however, always heavier in coastal waters during autumn (Figs. 8 and 9) because striped searobins weigh more on average and because striped searobins occupy inshore habitats more frequently in the warmest season. In prox- imity to New Jersey, northern searobins ap- pear to migrate inshore earlier, peak in abun- dance earlier, and move offshore earlier than striped searobins, a behavior consistent with their postulated temperature preference. The Mid- Atlantic Bight (NMFS) time se- ries of stratified mean weight per tow was striped searobins (filled symbols) during five consecutive cruises for the New Jersey coastal survey (Table 2). McBride et at: Interspecific comparisons of searobin, Prionotus spp., movements, size structure and abundance 309 highly variable (for northern searobins, CV=48 in spring and 49 in autumn; for striped searobins, CV= 68 in spring and 101 in autumn; Fig. 8) but showed no clear trend over time. There was one notable period of decline for northern searobins during spring surveys throughout the 1970’s, but there was also a modest rebound in abundance during the 1980’s. The low values in the early 1990’s were similar to values measured in the late 1960’s. The abundance of northern searobins during autumn cruises suggested no trend but was punctuated with sporadic peaks. Striped searobins also appeared to be more abundant during the 1970’s than in the 1960’s or 1990’s, according to spring indices. The autumn time series for striped searobins declined overall, until the last year (1995), which showed a sudden increase. Abundance was compared between species to establish if each species’ population size followed simi- lar seasonal and interannual trends. Corre- lation analyses detected a significant rela- tion between northern and striped searobin annual indices measured in the spring sur- vey (Spearman rank correlation, rs=0.47, P=0.01, n=28), but the correlation between species abundance in the autumn surveys was not significant (rs=0.006, P= 0.97, n=24). The spring and autumn indices for a given year were not significantly correlated for northern searobins (rs=0.14, P=0.51, n- 24) but were significantly correlated for striped searobins (rs=0.65, P=0.0006, n= 24). These latter results suggested that the selected spring and autumn strata measured similar trends of abundance for striped searobins but not for northern searobins. Earlier estimates of searobin population size also noted considerable yearly variations in abundance. Clark and Brown (1977) reported a dramatic 97% decline in abundance (by weight) of searobins in the southern Mid-Atlantic region (also depicted in Grosslein [1976]) based on the same database dur- ing the period 1967-74. But they also reported a 143% increase in abundance of searobins in the southern New England region. It is difficult to com- pare their results with our Figure 8 because they grouped northern and striped searobins together. They also used autumn collections from offshore strata sets, but our autecological analyses revealed these strata as only marginally appropriate, at least for striped searobins that are still inshore during autumn. Their analysis of abundance reviewed years during which fairly rapid change was occurring for both searobins and the groundfish assemblage in general, but our analyses demonstrate that searobin abundance has fluctuated up and down since the early 1970’s. Northern searobin annual abundance in the Mas- sachusetts coastal survey has also shown large long- term variations (Fig. 8). Modest increases for north- ern searobin abundance in the early 1990’s barely offset dramatic declines during the 1980’s. In con- trast, there was no trend in abundance of striped searobins during this 20-year period. These indices could be biased with respect to interannual varia- tions in coastal temperatures because Murawski (1993) demonstrated that the latitudinal range of northern searobins varied between years with respect to sea-temperature anomalies. Thus, one might ex- pect that searobins are more abundant in coastal Massachusetts waters during warmer years. Such does not appear to be the case, however, because 310 Fishery Bulletin 96(2), 1 998 Figure 6 Length-frequency distributions of northern searobins (open bars) and striped searobins (filled bars) for the Massachu- setts coastal survey. Sampling periods are the same as those shown in Figure 3. rc=number of fish. Murawski reported that after an anomalously warm period in the 1970’s (when northern searobin abun- dance did increase) temperatures in the 1980’s fluc- tuated without a trend (but northern searobin abun- dance declined dramatically; see Fig. 8). The indices of stratified mean weight per tow be- tween the Mid-Atlantic Bight (NMFS) and the Mas- sachusetts (MDMF) surveys were compared to ex- amine how robust these measures of abundance were. Correlation analyses revealed a significant relation between northern searobin abundance in spring (rs=0.51, P=0.032, n=18) between the two surveys, but autumn abundances in similar years were nega- tively correlated between the two surveys (rs=-0.60; P=0.QQ8, n = 18). There was a significant relation be- tween these surveys for striped searobin abundance in spring (r, =0.51, P=0. 029, n=18) but not for autumn (rs=0.077, P= 0.76, n = 18). These significant, positive relations between the two surveys support our use of spring but not autumn indices for examining north- ern and striped searobin population trends. implications for fishery ecology Northern and striped searobins belong to a “warm- temperate” group (Musick et ah, 1989), an assem- blage that migrates onshore and north as tempera- tures increase, offshore and south as temperatures Figure 7 Length-frequency distributions of northern searobins (open bars) and striped searobins (filled bars) during five consecutive cruises for the New Jersey coastal survey (Table 2). Sam- pling periods are the same as those shown in Figure 4. n=number of fish. decline. Our descriptive analysis supports this char- acterization of both species but reveals interspecific differences in the timing and extent of seasonal move- ments, as well as in size and abundance. Selection of sampling strata for calculating abundance indices was constrained by our goal to compare the ecology of both species. Nevertheless, indices from spring cruises in the Mid-Atlantic Bight and along the Mas- sachusetts coast were significantly and positively McBride et al.: Interspecific comparisons of searobin, Prionotus spp., movements, size structure and abundance 31 I Figure 8 Annual biomass indices for northern searobins (open symbols) and striped searobins (filled symbols). Mid- Atlantic Bight indices (left) are based on sampling by Northeast Fish. Sci. Center, NMFS, resource surveys (Table 2). Massachusetts indices (right) are based on the Massachusetts (Mass. Div. Mar. Fisheries [MDMF]) coastal surveys (note the scale of the ordinate is doubled for MDMF Spring). All values are log-transformed and plotted with ±1 standard error bar. See text for details of strata selection and statistical calculations (e.g. spring and autumn strata are different). correlated, so that they were espe- cially useful in evaluating popula- tion trends. On the other hand, the inshore strata selected for autumn cruises in the Mid-Atlantic Bight were adequate for sampling striped searobins but were too shallow for northern searobins because the lat- ter occupied a mean depth of 31 m during that season. Searobins are not important fish- ery species in the United States but are components of bycatch in com- mercial and sport fisheries; their population sizes could be affected by fishing or other factors. However, their population size varied without a specific directional trend during three decades. The long-term trend of annual biomass indices for sea- robins is similar to other demersal finfish such as white hake ( Urophycis 312 Fishery Bulletin 96(2), 1 998 tenuis), cusk ( Brosme brosme), scup ( Stenotomus chrysops), and ocean pout ( Macrozoarces americanus) (U.S. Dep. Commerce, 1993). In contrast, population declines have been characteristic of traditional groundfish and flounder species, and aggregate bio- mass of principal pelagic species, skates, and spiny dogfish have increased in the western North Atlan- tic (U.S. Dep. Commerce, 1993). Fishery landings, value, and catch rates for mid- Atlantic states have all declined over the past sev- eral decades (McHugh and Conover, 1986), and underused species such as searobins have limited potential to counter these trends. Our survey results, for New Jersey in particular, show that searobins are seasonally abundant in areas amenable to in- shore commercial and recreational fisheries during summer and autumn. Trawlers can also find large aggregations offshore during winter and spring. Searobins are also a large proportion of pound-net landings during summer, although this fishing method is now reduced in comparison with histori- cal levels (McHugh and Conover, 1986). Striped searobins may be a more marketable species because they are larger on average. Efforts to expand the marketability of searobins, such as has occurred re- cently by selling them live in restaurants (Lynch1), may lead to increased total landings and a more di- verse fishery resource for the mid-Atlantic region. Acknowledgments We thank the crew and scientists aboard the RV’s Albatross IV, ARGO Maine, Delaware II, Gloria Michelle, and the FV Amy-Oiane as well as T. Azarovitz (NEFSC, NMFS), P. D. Byrne (NJBMF), and A. Howe (MDMF) who provided data for analy- sis. R. Cowen, M. Fahay, J. P. Grassle, R. Loveland, and C. L. Smith made helpful comments in discussions and on earlier drafts. J. Bexley produced Figure 1. This study, part of a doctoral dissertation (R.S.M. ), was sup- ported by the Institute of Marine and Coastal Sci- ences (IMCS-Rutgers), Rutgers University Marine Field Station, Manasquan Marlin and Tuna Club, NOAA National Undersea Research Program for the New York Bight, NOAA Sea Grant Program, and the Fisheries and Aquaculture TEX Center (Rutgers). Final revisions were made while the senior author was at the Florida Marine Research Institute, Florida Department of Environmental Protection. 1 Lynch, T. R. 1993. Rhode Island Department of Environmen- tal Management, 150 Fowler Street, Wickford, Rhode Island 02852. Personal commun. Literature cited Azarovitz, T. R. 1981. A brief historical review of the Woods Hole Labora- tory trawl survey time series. In W. G. Doubleday and D. Rivard (eds.), Bottom trawl surveys/Releves au chalut de fond. Canadian Spec. Publ. Fish, and Aquat. Sci. 58: 62-67. Bigelow, H. B., and W. C. Schroeder. 1953. Fishes of the Gulf of Maine. Fish. Bull. 53:1-557. Briggs, P. T. 1977. An unusually large striped searobin from Long Is- land Sound, New York. N.Y. Fish Game J. 24 ( 1 ):96— 97. Byrne, D. M. 1989. New Jersey trawl surveys. In T. R. Azarovitz, J. McGurrin, and R. Seagraves (eds.), Proceedings of a work- shop on bottom trawl surveys. Atl. States Mar. Fish. Comm. Washington, D.C- Spec. Rep. 17, p. 46-48. Clark, S. H., and B. E. Brown. 1977. Changes in biomass of finfishes and squids from the Gulf of Maine to Cape Hatteras, 1963-1974, as determined from research vessel survey data. Fish. Bull. 75 ( 1 ): 1—22. Collins, J. W. 1892. Statistical review of the coast fisheries of the United States. Rep. U.S. Comm. Fish Fish for 1888, part 16, p. 271-378. Despres-Patanjo, L. I., T. R. Azarovitz, and C. J. Byrne. 1988. Twenty-five years of fish surveys in the northwest Atlantic: the NMFS Northeast Fisheries Center’s bottom trawl survey program. Mar. Fish. Rev. 50 (4):69-71. Edwards, R. L., R. Livingstone Jr., and P. E. Hamer. 1962. Winter water temperatures and an annotated list of fishes-Nantucket Shoals to Cape Hatteras. Albatross III Cruise 126. U.S. Fish Wildl. Ser. Spec. Sci. Rep. Fish. 397 (Washington D.C.), 31 p. Essig, R. A., J. F. Witzig, and M. C. Holliday. 1991. Marine recreational fishery statistics survey, Atlan- tic and Gulf coasts, 1987-89. U.S. Dep. Commer., NOAA, NMFS, Current Fisheries Statistics 8904, Silver Spring, MD, p. 25, 95, 165. Finney, D. J. 1941. On the distribution of a variate where logarithm is normally distributed. R. Stat. Soc., suppl. 7(1):55-61. Floyd, P. B. 1 980. Aspects of the distribution and ecology of searobins ( Pi- sces: Triglidae) off the southeast coast of the United States. M.S. thesis, College of Charleston, Charleston, SC, 104 p. Gilmore, R. G. 1977. Fishes of the Indian River Lagoon and adjacent wa- ter, Florida. Bull. Florida State Mus. 22 (3):101-148. Ginsburg, I. 1950. Review of the western Atlantic Triglidae (fishes). Texas J. Sci. 1950 (4):489-527 Goode, G. B. 1888. American fishes. W.A. Houghton, New York, NY, p. 301-306. Grosslein, M. D. 1969. Groundfish survey program of BCF Woods Hole. Comm. Fish. Rev. 31 (8 — 9):22 — 30. 1976. Some results of fish surveys in the mid-Atlantic im- portant for assessing environmental impacts. In M. G. Gross (ed.), Middle Atlantic Continental Shelf and the New York Bight: proceedings of the symposium; 3-5 November 1975, Lawrence, KS, p 312-328. Am. Soc. Limnol. Oceanog. special symposium, vol. 2. McBride et a I.: Interspecific comparisons of searobin, Prionotus spp.. movements, size structure and abundance 313 Hildebrand, S. F. and W. C. Schroeder. 1928. Fishes of the Chesapeake Bay. Bull. U.S. Bur. Fish. 43(part 1), 388 p. Hoff, J. G. 1992. Comparative biology and population dynamics of searobins (Genus Prionotus) with emphasis on populations in the northwestern Gulf of Mexico. Ph.D. diss., College of William and Mary, Gloucester Point, VA, 229 p. Holliday, M. C. 1984. Marine recreational fishery statistics survey, Atlan- tic and Gulf coasts, 1979(revised)-1980. U.S. Dep. Commer., NOAA, NMFS, Current Fisheries Statistics 8322, Washington, D.C., p. 24, 131. Howe, A. B. 1989. State of Massachusetts inshore bottom trawl survey. In T. R. Azarovitz, J. McGurrin, and R. Seagraves (eds. ), Proceedings of a workshop on bottom trawl surveys. Atl. States Mar. Fish. Comm. Washington, D.C. Spec. Rep. 17, p. 33-8. Johnson, J. C., D. C. Griffith, and J. D. Murray. 1987. Encouraging the use of underutilized marine fishes by southeastern U.S. anglers, Part I. The research. Mar. Fish. Rev. 49(21:122-137. Keirans, W. J., S. S. Herman, and R. G. Malsberger. 1986. Differentiation of Prionotus carolinus and Prionotus evolans eggs in Hereford Inlet estuary, southern New Jer- sey, using immunodiffusion. Fish. Bull. 84(11:63-68. Lewis, T. C., and R. W. Yerger. 1976. Biology of five species of searobins (Pisces, Triglidae) from the northeastern Gulf of Mexico. Fish. Bull. 74(1): 93-103. Marshall, N. 1946. Observations of the comparative ecology and life his- tory of two searobins, Prionotus carolinus and Prionotus evolans strigatus. Copeia. 1946(31:118—144. McBride, R. S. 1994. Comparative ecology and life history of two temper- ate, northwestern Atlantic searobins, Prionotus carolinus and P. evolans (Pisces: Triglidae). Ph D. diss., Rutgers, The State University of New Jersey, Tuckerton, NJ, 176 p. McBride, R. S„ and K. W. Able. 1994. Reproductive seasonality, distribution, and abun- dance of Prionotus carolinus and P. evolans (Pisces: Triglidae) in the New York Bight. Estuarine Coastal Shelf Sci. 38: 173-188. McEaehran, J. D., and J. Davis. 1970. Age and growth of the striped searobin. Trans. Am. Fish. Soc. 99(21:343-352. McHugh, J. L. 1977. Fisheries and fishery resources of the New York Bight. U.S. Dep. Commer, NOAA Tech. Rep NMFS Circ. 401, 50 p. McHugh, J. L., and D. Conover. 1986. History and condition of food finfisheries in the Middle Atlantic region compared with other sections of the coast. Fisheries (Bethesda) 11(51:8-13. Merriman, D., and H. E. Warfel. 1948. Studies on the marine resources of southern New England. VII. Analysis of a fish population. Bull. Bingham Oceanogr. Collect. Yale Univ. 11(41:131-164. Murawski, S. A. 1993. Climate change and marine fish distributions: fore- casting from historical analogy. Trans. Am. Fish. Soc. 122(51:647-658. Murray, J. D., J. C. Johnson, and D. C. Griffith. 1987. Encouraging the use of underutilized marine fishes by southeastern U.S. anglers. Part II. Educational objec- tives and strategy. Mar. Fish. Rev. 49(2): 138-142. Musick, J. A., J. A. Colvocoresses, and E. J. Foell. 1989. Seasonality and the distribution, availability and composition offish assemblages in Chesapeake Bight: pro- ceedings of the international symposium on utilization of coastal ecosystems: planning, pollution and productivity. 21-27 November 1982, Rio Grande, Brazil, vol. 2:337-376. Nichols, J. T., and C. M. Breder Jr. 1926. The marine fishes of New York and southern New England. Zoologica (N.Y.) 9 (11:1-192. NMFS. 1987. Marine recreational fishery statistics survey, Atlan- tic and Gulf Coasts, 1986. U.S. Dep. Commer., NOAA, NMFS, Current Fisheries Statistics 8392, Washington, D.C., 21 p. 1986. Marine recreational fishery statistics survey, Atlan- tic and Gulf coasts, 1985. U.S. Dep. Commer., NOAA, NMFS, Current Fisheries Statistics 8327, Washington, D.C., 21 p. 1985b. Marine recreational fishery statistics survey, Atlan- tic and Gulf coasts, 1983-1984. U.S. Dep. Commer., NOAA, NMFS, Current Fisheries Statistics 8326, Wash- ington, D.C., 23, 117 p.. 1985a. Marine recreational fishery statistics survey, Atlan- tic and Gulf coasts, 1981-1982. U.S. Dep. Commer., NOAA, NMFS, Current Fisheries Statistics 8324, Wash- ington, D.C., 23, 113 p. Pennington, M. 1983. Efficient estimators of abundance, for fish and plank- ton surveys. Biometrics. 39(11:281-286. Perl mutter, A. 1959-60. The searobin: a neglected food and sport fish. The N.Y. State Conservationist, December-January, p. 22-23. Richards, S. W., J. M. Mane, and J. A. Walker. 1979. Comparison of spawning seasons, age, growth rates, and food of two sympatric species of searobins, Prionotus carolinus and Prionotus evolans, from Long Island Sound. Estuaries 2(41:255-268. Ross, S. T. 1977. Patterns of resource partitioning in searobins (Pisces: Triglidae). Copeia 1977(31:561-571. Russell, M., M. Grace, and E. J. Gutherz. 1992. Field guide to the searobins ( Prionotus and Bellator) in the western north Atlantic. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 107, 26 p. Scott, W. B., and M. G. Scott. 1988. Atlantic fishes of Canada. Can. Bull. Fish. Aquat. Sci. No. 219, 731 p. Smith, H. M. 1894a. Economic and natural history notes on fishes of the northern coast of New Jersey. Bull. U. S. Fish. Comm. 12: 365-380. 1894b. Remarks on the maintenance and improvement of the American fisheries. Bull. U.S. Fish. Comm. 13:287-292. U.S. Department of Commerce. 1993. Status of fishery resources off the northeastern United States for 1993. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-F/NEC-101, 140 p. Van Voorhees, D. A., J. F. Witzig, M. F. Osborn, M. C. Holliday, and R. J. Essig. 1992. Marine recreational fishery statistics survey, Atlan- tic and Gulf coasts, 1990-1991. U.S. Dep. Commer., NOAA, NMFS, Current Fisheries Statistics 9204, Silver Spring, MD, p. 27, 116. 314 Fishery Bulletin 96(2), 1998 Wilk, S. J., W, W. Morse, and D. E. Ralph. 1978. Length-weight relationships of fishes collected in the New York Bight. Bull. N.J. Acad. Sci. 23(21:58-64. Wong, R. S. P. 1968. Age and growth of the northern searobin, Prionotus carolinus (Linnaeus). M.A. thesis. College of William and Mary, Gloucester Point, VA, 48 p. Yuschak, P. 1985. Fecundity, eggs, larvae and osteological development of the striped searobin, Prionotus evolans (Pisces, Trig- lidae). J. Northwest Atl. Fish. Sci. 6:65-85. Yuschak, P., and W. A. Lund. 1984. Eggs, larvae and osteological development of the northern searobin, Prionotus carolinus (Pisces, Trig- lidae). J. Northwest Atl. Fish. Sci. 5:1-15. 315 Abundance, growth, and mortality of young-of-the-year pinfish, Lagodon rhomboides, in three estuaries along the gulf coast of Florida Abstract .—Fixed-station and ran- dom-sampling data from 1989-94 were used to examine spatial and temporal patterns in abundance and size struc- ture of young-of-the-year (YOY) pinfish, Lagodon rhomboides , in three Florida estuaries. Young-of-the-year pinfish first appeared at shallow- water (<1.4 m) seine stations in November in Choctawhat- chee Bay (Florida Panhandle), and in December in Tampa Bay and Charlotte Harbor, both along the southwest Florida peninsula. Pinfish were caught at deep- water (>1.6 m) trawl stations within one month after their initial appearance at shallow-water (<1.4 m) sites in Choc- tawhatchee and Tampa bays. However, YOY were absent in the deep water of Charlotte Harbor until 1-3 months af- ter their first appearance in shallow water. Most YOY pinfish were caught in waters <3.5 m. Young-of-the-year pinfish in shallow-water areas were associated with bottom vegetation, mostly seagrasses, in all bays. Annual variation in YOY abundance was cor- related with variations in adult abun- dance in Tampa Bay and with tempera- ture in Charlotte Harbor. Instanta- neous growth rates were rapid (0.10 to 0.26/month) and were similar to pub- lished rates for other Florida and gulf coast populations. Similar rates of to- tal instantaneous mortality (0.021 to 0.023/day) were estimated for all bay populations. Manuscript accepted 3 December 1998. Fishery Bulletin 96:315-328 (1998). Gary A. Nelson Florida Marine Research Institute Department of Environmental Protection 1 00 Eighth Avenue SE St. Petersburg, Florida 33701-5095 E-mail address: nelson_ga@sellers. dep.state.fi. us Young-of-the-year pinfish, Lagodon rhomboides, play important ecologi- cal roles in the northeastern Gulf of Mexico as prey for fish (Carr and Adams, 1973; Seaman and Collins, 1983) and as predators on a range of invertebrates, often to a degree where entire assemblages of macro- benthic fauna are affected (Young et al., 1976; Young and Young, 1977; Nelson, 1978). In addition, YOY pin- fish are an important link between primary and secondary production because they consume seagrasses (Stoner, 1982; Weinstein et al., 1982; Montgomery and Targett, 1992). Despite the ecological importance of pinfish, their population dynam- ics have been inadequately exam- ined. It is unknown if seasonal changes in abundance or move- ments occur throughout entire es- tuaries, if growth rates are similar among populations, or if abun- dances fluctuate annually because past studies have had limited spa- tial coverage (usually 1-4 seagrass sites were sampled in waters <2 m) and short sampling durations (<2 yr)(Reid, 1954; Caldwell, 1957; Hellier, 1962; Hoese and Jones, 1963; Cameron, 1969; Hansen, 1970; Stoner, 1983). In addition, fac- tors that may influence year-class strength have not been examined, and mortality rates have not been estimated. In this study, I use two to six years of data to document seasonal changes in abundance, distribution, and movements within shallow- and deepwater areas to identify factors that may influence spatial and an- nual abundance and to estimate and compare growth and mortality rates among three estuarine popu- lations of YOY pinfish along the gulf coast of Florida, USA. Methods Young-of-the-year pinfish were studied in 1) Choctawhatchee Bay and Santa Rosa Sound (surface area: ca. 450 km2), located in the western Florida Panhandle, 2) Tampa Bay (ca. 886 km2), and 3) Charlotte Harbor (ca. 575 km2), the latter two located on the gulf side of the Florida peninsula (Fig. 1). All three bay systems are characterized by average depths of <5 m, salini- ties of 0-36 ppt, freshwater inflow from rivers, and expanses of bottom vegetation, primarily seagrasses ( Halodule wrightii and Thalassia testudinum), in shallow areas. Sea- sonal mean water temperatures range from 10 to 29°C in Choc- tawhatchee Bay and Santa Rosa 316 Fishery Bulletin 96(2), 1998 Sound, from 15 to 30°C in Tampa Bay, and from 18 to 32°C in Charlotte Harbor. Pinfish were sampled monthly from 1989 to 1994 at fixed seine and trawl stations. Fixed stations were approximately evenly distributed throughout shal- low- and deepwater areas and included sites in ma- jor rivers in Tampa Bay and Charlotte Harbor. Monthly sampling began in 1992 in Choctawhatchee Bay, in 1989 in Tampa Bay, and in 1991 in Charlotte Harbor. Samples were collected with a 21.3-m x 1.8- m, 3.2-mm stretched-mesh seine, or a 6.1-m, 38-mm stretched-mesh otter trawl containing a 3.2-mm stretched-mesh codend liner. At beach stations, seines were set adjacent to the shoreline and hauled on- shore; at offshore stations (<1.4 m), seines were set in open-water habitats away from the shoreline and retrieved offshore. In rivers, seines were set from the shoreline in a semicircular pattern from a boat. In deep water (>1.6 m), trawls were towed 1 knot for 5 min at river sites and for 10 min at bay sites. Three hauls or tows were made at each fixed station dur- ing daylight hours. Sampling occurred during the first two weeks of each month. Pinfish were also sampled in spring (March to June) at randomly selected sites to provide more ac- curate estimates of YOY abundance. To coordinate sampling logistics, each bay was subdivided into 5-6 arbitrarily lettered, permanent zones (bay: zones A- E; rivers: zone F). All bay zones encompassed about equal area. Within each zone, 1' latitude x 1' longi- tude microgrids, representing the sites to be sampled, were randomly selected within randomly selected 1° latitude x 1° longitude grids. Sampling entailed ran- domly selecting a zone and then sequentially sam- pling all sites within each zone. At each site, three hauls were made with the same gears and deployment techniques used at fixed stations. Random sampling began in Choctawhatchee Bay in 1993 and in Tampa Bay and Charlotte Harbor in 1989 and occurred over eight, twelve, and ten weeks, re- spectively. For all hauls, total numbers of pinfishes were counted, standard lengths were measured (±1 mm) for 20 randomly selected individuals per sample, and all fish were released. When large numbers of indi- viduals (>1,000) were captured, the total number was estimated by fractional expansion of subsampled portions of the total catch split with a modified Motoda splitter (Motoda, 1959). Salinity (ppt), tem- perature (°C), depth (m), and bottom types were also recorded at all sites. Dominant vegetation types were recorded at seine sites only. Seasonal changes in YOY abundance and size structure To examine seasonal changes in YOY abundance in shallow- and deepwater areas, comparable monthly mean number of individuals per 100 m2 were calcu- lated from fixed station data by year. I separated YOY data used in all analyses from data on older individuals by using maximum size limits se- lected from monthly length-frequency plots. Monthly length frequencies based on propor- tions of fish found in each length class were combined over years to describe within-year trends rather than year-to-year variability. Length maximum size limits used for Choctawhatchee Bay data were in agreement with the maximum lengths of scale-aged YOY pinfish from Pensacola Bay (Hansen, 1970). Depth distribution To determine whether YOY were restricted to depth ranges during the period surround- ing peak abundance, the cumulative fre- quency distributions of trawl and YOY pin- fish depth occurrences were compared by using the Kolmogorov-Smirnov test (Perry and Smith, 1994). The cumulative frequency distributions for YOY pinfish were con- structed by weighting depth at each random site by the number of YOY pinfish captured at that site. Only spring data from the first Figure 1 Map of Florida showing the locations of Choctawhatchee Bay, Tampa Bay, and Charlotte Harbor. Nelson: Abundance, growth, and mortality of young-of-the-year Lagodon rhomboides 317 trawl made at each random site were used to ensure independence of observations. Data from 1989-94 were combined to examine within-year trends rather than year-to-year variability. Factors influencing spatial YOY abundance I examined variation in YOY pinfish abundance in shallow-water areas for year, deployment technique, month, zone, sediment type (mud or sand), absence or presence of bottom vegetation (mostly seagrasses), temperature, and salinity effects by bay. Spring catch data (transformed by using ln(x+l) prior to analy- sis) from the first seine haul at each randomly se- lected site were analyzed with general linear models (GLM; Hilborn and Walters, 1992) and PROC GLM (SAS Institute, 1988). Year, month, deployment tech- nique, zone, sediment type, and bottom vegetation were treated as main effects, and temperature and salinity (transformed by using In (x+ 1 ) prior to analy- sis) as covariates. All first-order interactions of the main effects were also tested. Any variable or first order interaction not significant at a = 0.05 with type- III (partial) sum of squares was dropped from the initial GLM model and the analysis was repeated. In addition, least squares means and their 95% con- fidence intervals (Searle et al., 1980; SAS Institute, 1988) from the GLM’s were back-transformed (Sokal and Rohlf, 1981) to examine significant abundance and main effect relationships. Initial analyses revealed that the only significant first-order interactions were related to the random selection of zones for sampling. Because these inter- actions were not considered relevant to this study, they were absorbed in the error term, and the main effects and covariates were retested. Factors influencing YOY annual abundance To determine if annual variations in YOY abundance were correlated with variations in temperature, I compared annual relative abundance indices (least squares means for the effect of year) to monthly means of sea-surface temperature before and dur- ing the first appearance of YOY pinfish, using Pearson product moment correlation (Sokal and Rohlf, 1981; Tyler, 1992). Temperature data were obtained from the National Oceanic and Atmospheric Administration’s (NOAA) oceanographic monthly summary series. I also used Pearson product moment correlation to determine if annual variations in YOY relative abundance were correlated with variations in adult abundance. Adult (>80 mm SL) pinfish abundance indices were derived from data collected in the Ma- rine Recreational Fisheries Statistics Survey (MRFSS on Florida’s west coast in 1988-93 )(U.S. Dep. Commer., 1990; 1992). The GLM approach was used only to derive annual least-squares mean catch- per-intercept estimates (relative abundance) by ad- justing the total number of fish caught per intercept for the classification variables of two-month sampling wave, fishing mode (party or charter boat, private or rental boat, or shore-based fishing boat), area fished, counties where interviews were conducted, and for the covariates of number of anglers per intercept and hours fished by anglers. All variables were signifi- cant contributors to the overall variation in catch rates in Tampa Bay (model: F |28 3853]=11.55, P<0.001, r2=0.08), but only year, sampling wave, county, and hours fished by anglers were significant for Char- lotte Harbor (Model: F[20 10601=5.74,P<0.001, r2=0.10). Growth I examined annual growth of YOY by estimating in- stantaneous growth rates ( G ) using mean lengths for each bay and year. Growth was estimated with the following model: In L, = In L0+Gxt , where G = the instantaneous growth rate (per month); Lf = monthly mean length (mm); L0 = the theoretical length at which pinfishes recruit to each bay; and t = time in months (Ricker, 1975; DeAngelis et al., 1980). Mortality Daily instantaneous total mortality rates were esti- mated for each bay population of pinfish by means of the relationship where Z = the daily instantaneous total mortality; and N = the index of relative abundance at months t and t+ 1 (Ricker, 1975). Monthly indices of relative abundance from fixed seine stations were used in the equation. Although immigration and emigration in the shallow-water areas can bias the rate of decline in abundance used 318 Fishery Bulletin 96(2), 1 998 to estimate mortality, I used data only from months over which these processes appeared low. Results Seasonal changes in YOY abundance and size structure Young-of-the-year pinfish appeared first as post- larvae (9-12 mm) at shallow-water fixed seine sta- tions during early November in Choctawhatchee Bay and early December in Tampa Bay and Charlotte Harbor, one to two months before mean temperatures were lowest (Fig. 2). In Choctawhatchee Bay and Tampa Bay, pinfish were first collected at deepwater trawl stations in the same month or one month after their initial appearance at seine stations (Fig. 2). In Charlotte Harbor, however, YOY were absent at trawl stations until January-March, one to three months after their first appearance at seine stations (Fig. 2). Nelson: Abundance, growth, and mortality of young-of-the-year Lagodon rhomboides 319 Relative abundance at fixed seine stations peaked in January and May in Choctawhatchee Bay, gener- ally during March-April in Tampa Bay and Char- lotte Harbor, and declined thereafter (Fig. 2). In Choctawhatchee Bay and Tampa Bay, relative abun- dance at trawl stations generally followed fluctua- tions in seine catches, but usually peaked one to two months before the peak at seine stations (Fig. 2). A second peak in trawl abundance was observed in Tampa Bay from August to October (Fig. 2). In Char- lotte Harbor, relative abundance at trawl stations peaked during June-September, two to five months after the peak in seine abundance (Fig. 2). In all bays, smaller pinfish were generally captured at fixed trawl stations during November to March than at fixed seine stations (Figs. 3-5). Progression of the smallest fish size beyond the minimum size measured during the months of first capture indi- cated that settlement of postlarvae to fixed stations ended by March-April in all bays (Figs. 3-5). In Tampa Bay, catch proportions of YOY <40 mm decreased at trawl stations in March. In July, YOY >60 mm were captured in higher proportions at trawl stations in all bays than during the preceding months (Fig. 3-5). Pinfish overwintering at shallow-water seine stations in Choctawhatchee Bay tended to be smaller than those remaining at seine stations in Tampa Bay and Charlotte Harbor (Figs. 3-5). Depth distribution About 80% of the trawl catches ofYOY pinfish in spring occurred in waters <3.1 m, <3.5 m, and <2.8 m in Choctawhatchee Bay, Tampa Bay, and Charlotte Harbor, respectively. Few fish (<1% of total catches) were captured in waters >5 m. The Kolmogorov- Smirnov test showed that cumulative frequency dis- tributions for YOY pinfish depth were significantly different from those for trawl depth in all bays (Choctawhatchee Bay: Z)=0.51, ndepth= 80, n^ish= 34, P <0.001; Tampa Bay: D=0.44, nd th= 373, nfish= 76, P<0.001; Charlotte Harbor: D- 0.48, n . ,=268, nfis/=59, P<0.001). Factors influencing YOY spatial abundance The final GLM’s accounted for proportions of 0.33- 0.44 of the total variation in spring catches, depend- ing on bay system (Table 1). Pinfish abundance was associated with the presence of bottom vegetation in all bays (Table 1; Fig. 6A), with rivers (zone F) and zones near bay mouths in Tampa Bay (D and E) and Table 1 Final results of the general linear model analyses of pinfish catches for Choctawhatchee Bay, Tampa Bay, and Charlotte Harbor in spring. Partial (type-III) mean squares are shown. * = P<0.05, **= P<0.01, and *** = PcO.001. Location Source df Mean square P-value Choctawhatchee Bay Model 3 71.120 23.70*** Year 1 0.097 0.03 Deployment 1 42.637 14 21*** Bottom vegetation 1 179.747 59.89*** Error 92 3.001 Corrected total 95 5.152 Tampa Bay Model 11 67.312 19.84*** Year 5 14.041 4.14** Zone 5 65.287 19.25*** Bottom vegetation 1 193.316 56.99*** Error 425 3.392 Corrected total 436 5.005 Charlotte Harbor Model 13 59.461 15.85*** Year 5 24.823 6.62*** Zone 4 33.193 8.85*** Bottom vegetation 1 24.406 6.51* Bottom sediment 1 43.380 11.56*** Salinity 1 62.795 16.74*** Temperature 1 22.778 6.07* Error 283 3.752 Corrected total 296 6.198 r 2 0.436 0.334 0.421 320 Fishery Bulletin 96(2), 1998 Charlotte Harbor (B and C) (Table 1; Fig. 6B), with beach seine sets in Choctawhatchee Bay (Table 1; Fig. 60, and with mud bottom in Charlotte Harbor (Table 1; Fig. 6D). Young-of-the-year pinfish catches were related to salinity and temperature in Char- lotte Harbor only (Table 1). 0> -Q E 3 c iS o o c o tr o Q. O CL Standard length (mm) Figure 3 Monthly length-frequency distributions of pinfish captured at fixed seine ( — ) and trawl (• • •) stations in Choctawhatchee Bay from 1989 to 1994. Vertical arrows indicate the monthly maximum size limits for young-of-the-year data, n is the number of pinfish measured. Nelson: Abundance, growth, and mortality of young-of-the-year Lagodon rhomboides 321 Young-of-the-year pinfish abundance varied signifi- cantly between years in Tampa Bay and Charlotte Harbor (although not significant, year was included in the model for Choctawhatchee Bay to compute least squares meansXTable 1; Fig. 7). Pinfish abun- dance was also generally highest in Choctawhatchee Standard length (mm) Figure 4 Monthly length-frequency distributions of pinfish captured at fixed seine( — ) and trawl (• • •) stations in Tampa Bay from 1989 to 1994. Vertical arrows indicate the monthly maximum size limits for young-of-the- year data, n is the number of pinfish measured. 322 Fishery Bulletin 96(2), 1998 Bay (range: 55.0-59.8 fish/haul), followed by Char- lotte Harbor (5.5-50.4 fish/haul) and Tampa Bay (7.8-27.8 fish/haul )( Fig. 7). Factors affecting YOY annual abundance Relative abundance of YOY pinfish was significantly Standard length (mm) Figure 5 Monthly length-frequency distributions of pinfish captured at fixed seine! — ) and trawl (■ ■ ■) stations in Charlotte Harbor from 1989 to 1994. Vertical arrows indicate the monthly maximum size limits for young- of-the-year data, n is the number of pinfish measured. Nelson: Abundance, growth, and mortality of young-of-the-year Lagodon rhomboides 323 and positively correlated with mean sea-surface tem- peratures in November (the month before first appear- ance) in Charlotte Harbor and with indices of adult abundance from 1988 to 1993 in Tampa Bay (Table 2). Growth Growth rates were estimated with mean length data from fixed seine stations only from April through July to minimize biases associated with potential move- ments ofYOY pinfish. Year-class growth rates were similar among bays. Instantaneous growth coeffi- cients ranged from 0.18 to 0.26/month for Choctaw- hatchee Bay, 0.06-0.21/month in Tampa Bay, and 0.14-0.26/month for Charlotte Harbor, indicating that monthly growth ofYOY pinfishes is rapid dur- ing late spring and mid-summer months (Table 3). Comparisons among bays revealed no consistent interannual patterns in growth except from 1993 to 1994 when rates declined in all bays (Table 3). E l — l Choctawhatchee Bay l wwa Tampa Bay V / / I Charlotte Harbor Figure 6 Least-squares means and 95% confidence intervals for main effects of (A) presence or absence of bottom vegetation, (Bl zones, (C) deployment techniques, and (Dl sediment type. Zones A-E en- compass bay areas, whereas zone F encompasses rivers. 324 Fishery Bulletin 96(2), I 998 Mortality The estimates of Z were calculated from the decline in indices of relative abundance at shallow-water fixed stations from May (month of secondary peak in abundance) to June for Choctawhatchee Bay, and April (month of peak abundance) to May for Tampa Bay and Charlotte Harbor. Relative abundance in- dices were averaged over all years prior to calcula- tion in order to reduce the effects of interannual vari- ability. Estimates of Z were 0.022/d, 0.021/d, and 0.023/d for Choctawhatchee Bay, Tampa Bay, and Charlotte Harbor, respectively (Table 4). Discussion Seasonal changes in YOY abundance and size structure Pelagic pinfish larvae are transported from oceanic spawning areas and dis- persed into estuaries via near-surface water currents (Darcy, 1985). After metamorphosis, postlarvae settle more or less near the bottom in estuaries (Hildebrand and Cable, 1938; Caldwell, 1957). Initial appearance of postlarvae (9-12 mm) and larger YOY (<28 mm) in shallow and deep water within a one- month period suggests pinfish settle to both areas in Choctawhatchee Bay and Tampa Bay. In Charlotte Harbor, settle- ment appears to occur first to shallow water, and then to deep water, because YOY pinfish were absent in trawls un- til 1-3 months after their first appear- ance at fixed seine stations. One expla- nation for these different patterns of settlement may lie with the bathymetry of each bay: more shallow-water area with seagrass beds (YOY pinfish are dependent on seagrass for protection from predators and food [Stoner, 1980; 1983]) is available in Charlotte Harbor (seagrass area=202 km2) than in Choc- tawhatchee Bay (33 km2) or Tampa Bay (83 km2)( Sargent et al., 1995). The occurrence of YOY pinfish mostly in water <3.5 m suggests that pinfish distribution is depth-restricted. The pro- pensity of YOY pinfish to limit their depth is probably due to their depen- dence on seagrasses for cover (sea- grasses are generally restricted to wa- ters <2.3 m in Choctawhatchee Bay, Tampa Bay, and Charlotte Harbor [Durako1]), the distribution of pelagic and mobile epibenthic prey (Stoner, 1980), and light intensity (Gulbrandsen, 1996) (YOY pinfish must see their prey 100 80 - 60 40 20 -I 0 Choctawhatchee Bay 1989 1990 1991 1992 1993 1994 Figure 7 Annual indices of young-of-the-year abundance and 95% confidence inter- vals for Choctawhatchee Bay, Tampa Bay, and Charlotte Harbor from 1989 to 1994. 1 Durako, M. 1996. Florida Marine Research Institute, 100 Eighth Avenue SE, St. Peters- burg, FL 33701-5095. Personal commun. Nelson: Abundance, growth, and mortality of young-of-the-year Lagodon rhomboides 325 Table 2 Results of the Pearson product moment correlation tests for indices of recruit abundance from 1989 to 1994 versus mean sea- surface temperature during the months before and during initial recruitment from 1988 to 1993 and for indices of adult abun- dance from 1988 to 1993 for Tampa Bay and Charlotte Harbor, r is the Pearson product moment correlation coefficient, P is the significance probability, and n is the sample size. MRFSS = Marine Recreational Fisheries Statistical Survey. Tampa Bay Charlotte Harbor Month r P r P n YOY indices-temperature October -0.05 0.92 0.67 0.15 6 November -0.78 0.11 0.84 0.03 6 December 0.15 0.78 -0.21 0.67 6 YOY indices-MRFSS adult indices 0.86 0.02 -0.15 0.78 6 Tab!e 3 Regression statistics for mean length (ML) versus month (m) for young-of-the-year pinfish captured at fixed seine stations. The regression takes the form: Ln (ML) = ln(L0) + G x m where ln(L0) is the intercept and G is the instanta- neous growth rate or slope. Only data from April to July were used. All slopes and intercepts were significantly dif- ferent from zero. Year ln(L0) SE G-value SE r2 n Choctawhatchee Bay 1993 2.03 0.217 0.26 0.038 0.956 4 1994 2.61 0.162 0.18 0.029 0.954 4 Tampa Bay 1989 2.77 0.078 0.18 0.014 0.988 4 1990 3.53 0.116 0.06 0.021 0.818 4 1991 3.17 0.030 0.10 0.005 0.994 4 1992 2.64 0.072 0.20 0.013 0.992 4 1993 2.49 0.274 0.21 0.049 0.904 4 1994 2.98 0.175 0.14 0.031 0.905 4 Charlotte Harbor 1991 2.38 0.095 0.26 0.017 0.991 4 1992 2.96 0.123 0.14 0.022 0.954 4 1993 2.46 0.243 0.21 0.043 0.920 4 1994 2.81 0.266 0.16 0.047 0.857 4 and therefore feed little at night [Kjelson and John- son, 1976]). Young-of-the-year pinfish that settle to deep wa- ter move to shallow-water areas in early spring. Evi- dence for this is the peak in trawl abundance ofYOY pinfish in all years, followed by a rapid decline be- fore the peaks in abundance at seine stations in Choctawhatchee Bay and Tampa Bay (Fig. 2). This movement to shallow water may be due to YOY seek- ing food and refuge from predators because it coin- cides with seasonal increases in seagrass biomass and prey abundance in shallow waters (Thoemke, Table 4 Relative abundance (no. of fish/100 m2) for April to June averaged over all years and estimates of total instanta- neous mortality (Z) for young-of-the-year pinfish ( Lagodon rhomboides) in Choctawhatchee Bay, Tampa Bay, and Char- lotte Harbor. Data used to estimate Z are boldface. Month Choctawhatchee Bay Tampa Bay Charlotte Harbor April 163.6 164.2 97.2 May 228.2 88.5 48.9 June 119.2 39.1 18.7 -Z 0.022 0.021 0.023 1979; Lewis et al., 1985) and because the shallow water and structural complexity of seagrasses may provide protection from predation (Savino and Stein, 1982; Stoner, 1983; Ruiz et al., 1993). Young-of-the-year pinfish move from shallow-water to deepwater areas in mid- to late summer, and this movement appears to be related to YOY size. Evi- dence for this is the shift to larger sizes in trawls in July (Figs. 3-5) concurrent with increasing catches in trawls in Tampa Bay and Charlotte Harbor (Fig. 2). This movement may represent the initiation of their fall spawning migration (Caldwell, 1957; Han- sen, 1970) because gonadal recrudescence begins as early as July (Cody and Bortone, 1992) and because the modal lengths observed in trawls in July were about 80 mm, which is the minimum size observed for YOY pinfish with developing gonads (Hansen, 1970). Factors affecting YOY spatial abundance Without meaningful, significant first-order interac- tions in the GLM analyses, the YOY abundance and main effect (zones, deployment technique, and bot- 326 Fishery Bulletin 96(2), 1998 tom type) relationships were difficult to assess. How- ever, because pinfish abundance was associated with vegetation (seagrasses) in this and other studies (Stoner, 1980; Stoner, 1983), the significance of the main effects can be proposed in relation to seagrass distribution. Higher YOY pinfish abundances may have occurred in zones located near bay mouths be- cause the largest areas of seagrass are located within these zones (Sargent et al., 1995). The absence of large expanses of seagrasses in Choctawhatchee Bay (Sargent et al, 1995) may have caused YOY pinfish to populate shallow-water beach areas abundantly to avoid predation (Ruiz et al., 1993). Because mud is commonly associated with seagrass beds in shallow- water areas of Charlotte Harbor (Mitchell2), higher abundances of YOY over mud may be expected. It was surprising that more YOY pinfish were cap- tured in Choctawhatchee Bay because there is less seagrass in this bay than in Tampa Bay or Charlotte Harbor (Sargent et al., 1995). Higher abundances of YOY pinfish may occur in Choctawhatchee Bay be- cause Halodule wrightii, a thin-blade seagrass pre- ferred by YOY pinfish for refuge and amphipod for- aging ( Stoner, 1982; 1983), occurred more frequently ( 62%) at vegetated sites in Choctawhatchee Bay than at vegetated sites in Tampa Bay (46%) and in Char- lotte Harbor (40%). 3 The patchily distributed sea- grass beds in Choctawhatchee Bay may also support higher abundances of pinfish than the continuously distributed seagrass beds in Tampa Bay and Char- lotte Harbor because the ecotone between seagrasses and unvegetated areas may provide greater habitat complexity, offering protection from predators while providing close access to alternative feeding areas (Holt et al., 1983). Factors influencing YOY annual abundance Although this study is an exploratory analysis, the positive correlation between YOY abundance and sea- surface temperatures in Charlotte Harbor suggests that oceanographic or biological events that occur before settlement may be important factors contrib- uting to the annual variability in pinfish abundance. Higher temperatures may favor increased hatching success (Postuma, 1971) or increased growth of lar- vae, or both (Hunter, 1981; Miller et al., 1985; Pepin, 1991), or they may affect transport mechanisms (Lasker, 1984; Rothschild, 1986). For pinfish, both 2 Mitchell, M. E. 1997. Florida Marine Research Institute, Charlotte Harbor Field Laboratory, 1481-A Market Circle, Port Charlotte, FL 33953. Personal commun. 3 Percent occurrence for seagrasses was estimated from random sampling data in spring. explanations are plausible given that adults spawn in offshore Gulf waters. Direct spawning stock-recruitment relationships are often masked by variability in recruitment (Fogarty et al., 1991). The lack of correlation between YOY and adult abundances in Charlotte Harbor sug- gests that temperature may be a more influential factor for this bay. The significant correlation be- tween YOY and adult indices at such a low sample size does suggest that for Tampa Bay, the relation- ship may not be markedly masked, and identifica- tion of the actual spawning stock-recruitment pat- terns may be possible with additional years of data. Growth To compare growth rates for YOY pinfish from this study with those found for pinfish from Cedar Key, FL, Redfish Bay, TX, and the Laguna Madre, TX, I fitted the same growth equation to mean length data of YOY pinfish estimated from graphical plots shown in Caldwell (1957) for Cedar Key, Cameron (1969) for Redfish Bay, and Hellier ( 1962) for Laguna Madre, for April-July. Instantaneous growth rates were 0.10/ month for YOY pinfish from Redfish Bay, TX, 0.17/ month from the Laguna Madre, TX, and 0.25/month from Cedar Key, FL, indicating that growth in these bays was similar to growth of YOY pinfish in the three bays studied (Table 4). Similar growth rates were expected given that temperatures experienced by YOY pinfish among the Gulf coast estuaries were alike during the April to July growth period (Caldwell, 1957; Cameron, 1969; Hellier, 1962). Mortality Daily mortality of YOY pinfish in shallow- water ar- eas of the three Florida estuaries was low. My esti- mates of mortality (0.021-0.023) were similar to those made for other estuarine-dependent species such as juvenile gulf menhaden ( Brevoortia patronus ) in Fourleague Bay, Louisiana (0. 017-0. 021)(Deegan, 1990), spot (Leiostomus xanthurus) in York River, Virginia (0.017) (Weinstein, 1983), and Atlantic croaker (Micropogonias undulatus ) in Rose Bay, North Carolina (0.023) (Currin et al., 1984). Unfor- tunately, I could not estimate mortality of YOY pin- fish in deep water because emigration and immigra- tion appeared to occur continuously over the spring- summer period at fixed trawl stations. In summary, YOY pinfish first appeared in shal- low-water areas during November in Choctawhatchee Bay and during December in Tampa Bay and Char- lotte Harbor. In Choctawhatchee Bay and Tampa Bay, they were captured in deep water within one month Nelson: Abundance, growth, and mortality of young-of-the-year Lagodon rhomboides 327 after their shallow-water appearance. In Charlotte Harbor, YOY pinfish were absent in deep water un- til 1-3 months after they first appeared in shallow water. Young-of-the-year pinfish were generally re- stricted to depths <3.5 m in all bays. Two general movements were evident: from deep water to shal- low water in spring, and from shallow water to deep water in mid- to late summer, the latter movement being size-dependent. High abundances of pinfish were commonly associated with the presence of seagrasses. Despite differences in abundance among bay populations, growth and mortality rates of young- of-the-year pinfish were similar in all bays. Acknowledgments Funding for this study was provided in part by the State of Florida Recreational Fishing License and in part by the Department of the Interior, U.S. Fish and Wildlife Service, Federal Aid for Sportfish Restora- tion, Project Number F-43 to the Florida Department of Environmental Protection. The author wishes to thank R. Muller, who provided statistical advice and generated the adult indices from the MRFSS data- base, A. Veri-Bodie, who provided encouragement throughout the study, and the staff of the Fisheries- Independent Monitoring Program at the Florida Marine Research Institute for their dedication to sampling. Comments by K. Guindon-Tisdel, J. Lieby, M. Murphy, R. McMichael, J. Wallin, and D. Winkel- man, and three anonymous reviewers improved the quality of the manuscript. Literature cited Caldwell, D. K. 1957. The biology and systematics of the pinfish, Lagodon rhomboides (Linnaeus). Bull. Fla. State Mus. Biol. Sci. 2:77-173. Cameron, J. N. 1969. Growth, respiratory metabolism and seasonal distri- bution of juvenile pinfish ( Lagodon rhomboides Linnaeus) in Redfish Bay, Texas. Contrib. Mar. Sci. 14:19-36. Carr, W. E. S., and C. A. Adams. 1973. Food habits of juvenile marine fishes occupying seagrass beds in the estuarine zone near Crystal River, Florida. Trans. Am. Fish. Soc. 102(3):511-540. Cody, R. P., and S. A. Bortone. 1992. An investigation of the reproductive mode of the pin- fish, Lagodon rhomboides Linnaeus ( Osteichthys: Sparidae). Northeast Gulf Sci. 12(2):99-110. Currin, B. M., J. P. Reed, and J. M. Miller. 1984. Growth, production, food consumption, and mortal- ity of juvenile spot and croaker: a comparison of tidal and nontidal nursery areas. Estuaries 7(4A):451-459. Darcy, G. H. 1985. Synopsis of biological data on the pinfish, Lagodon rhomboides (Pisces: Sparidae). U.S. Dep. Commer., NOAA Tech. Rep. NMFS, 32 p. DeAngelis, D. L., P. A. Hackney, and J. C. Webb. 1980. A partial differential equation model of changing sizes and numbers in a cohort of juvenile fish. Environ. Biol. Fish. 5(3):261-266. Deegan, L. A. 1990. Effects of estuarine environmental conditions on population dynamics of young-of-the-year gulf menhaden. Mar. Ecol. Prog. Ser. 68:195-205. Fogarty, M. J., M. P. Sissenwine, and E. B. Cohen. 1991. Recruitment variability and the dynamics of exploited marine populations. Trends Evol. 6:241-246. Gulbrandsen, J. 1996. Effects of spatial distribution of light on prey inges- tion of Atlantic halibut larvae. J. Fish. Biol. 48:478-483. Hansen, D. J. 1970. Food, growth, migration, reproduction and abundance of pinfish, Lagodon rhomboides , and Atlantic croaker, Micropogonias undulatus, near Pensacola, Florida, 1963- 65. Fish. Bull. 68(1): 135-146. Hellier, T. R. 1962. Fish production and biomass studies in relation to photosynthesis in the Laguna Madre of Texas. Publ. Inst. Mar. Sci. Univ. Tex. 8:1-22. Hilborn, R., and C. J. Waiters. 1992. Quantitative fisheries stock assessment: choice dy- namics and uncertainty. Routledge, Chapman and Hall, Inc., New York, NY, 570 p. Hildebrand, S. F., and L. E. Cable. 1938. Further notes on the development and life history of some teleosts at Beaufort, N. C. Bull. Bur. Fish. 48:505-642. Hoese, H. D., and R. S. Jones. 1963. Seasonality of larger animals in a Texas turtle grass community. Publ. Inst. Mar Sci. Univ. Tex. 9:37-46. Holt, S. A., C. L. Kitting, and C. R. Arnold. 1983. Distribution of young red drums among different sea- grass meadows. Trans. Am. Fish. Soc. 112:267-271. Hunter, J. R. 1981 . Feeding ecology and predation of marine fish larvae. In R. Lasker, (ed.), Marine fish larvae: morphology, ecol- ogy, and relation to fisheries. Washington Sea Grant Pro- gram Proj. E/F-4, p. 33-71. Kjelson, M. A., and G. N. Johnson. 1976. Further observations of the feeding ecology of post- larval pinfish, Lagodon rhomboides, and spot, Leiostomus xanthurus. Fish. Bull. 74(2):423-432. Lasker, R. 1984. The role of a stable ocean in larval fish survival and subsequent recruitment. In R. Lasker, (ed.), Marine fish larvae: morphology, ecology, and relation to fisheries. Washington Sea Grant Program Proj. E/F-4, p. 80-85. Lewis, R. R., M. J. Durako, M. D. Moffler, and R. C. Phillips. 1985. Seagrass meadows of Tampa Bay: a review. In S. F. Treat, J. L. Simon, R. R. Lewis III, and R. L. Whitman Jr. (eds.), Proceedings of the Tampa Bay area scientific infor- mation symposium, p. 210-247. Bellwether Press, Tampa, FL. Miller, J. M., L. B. Crowder, and M. L. Moser. 1985. Migration and utilization of estuarine nurseries by juvenile fishes, an evolutionary perspective. In M. Rankin (ed.), Migration: mechanisms and adaptive significance, p. 338-342. Contrib. Mar. Sci. 27 (suppl ). Montgomery, J. L. M., and T. E. Targett. 1992. The nutritional role of seagrass in the diet of the 328 Fishery Bulletin 96(2), 1998 omnivorous pinfish Lagodon rhomboides (L.). J. Exp. Mar. Biol. Ecol. 158:37-57. Motoda, S. 1959. Devices of simple plankton apparatus. Mem. Fac. Fish. Hokkaido Univ. 7:73-94. Nelson, W. G. 1978. Organization of a subtidal seagrass amphipod guild: the roles of predation, competition, and physical stress. Ph.D. diss., Duke Univ., Durham, NC, 223 p. Pepin, P. 1991. Effect of temperature and size on development, mortal- ity, and survival rates of the pelagic early life history stages of marine fish. Can. J. Fish. Aquat. Sci. 48:503-518. Perry, R. I., and S. J. Smith. 1994. Identifying habitat associations of marine fishes us- ing survey data: an application to the Northwest Atlantic. Can. J. Fish. Aquat. Sci. 51:589-602. Posturaa, K. H. 1971. The effect of temperature in the spawning and nurs- ery areas on recruitment of autumn-spawning herring in the North Sea. Rapp. P. V. Reun. Cons. Int. Explor. Mer 160:175-183. Reid, G. K., Jr. 1954. An ecological study of the Gulf of Mexico fishes in the vicinity of Cedar Key, Florida. Bull. Mar. Sci. Gulf Caribb. 4:1-94. Ricker, W. E. 1975. Computation and interpretation of biological statis- tics of fish populations. Bull. Fish. Res. Board Can. 191, 382 p. Rothschild, B. J. 1986. Dynamics of marine fish populations. Harvard Univ. Press, Cambridge, MA, 277 p. Ruiz, G. M., A. H. Hines, and M. H. Posey. 1993. Shallow water as a refuge habitat for fish and crus- taceans in non-vegetated estuaries: an example from Chesapeake Bay. Mar. Ecol. Prog. Ser. 99:1-6. SAS Institute. 1988. Proc GLM procedures. SAS/Stat user’s guide, vol. 2, release 6.03 ed. SAS Inst., Inc., Cary, NC, p. 893-993. Sargent, F. J., T. J. Leary, D. W. Crewz, and C. R. Kruer. 1995. Scarring of Florida’s seagrasses: assessment and management options. Fla. Mar. Res. Inst. Tech. Rep. TR- 1, 71 p. Savino, J., and R. A. Stein. 1982. Predator-prey interaction between largemouth bass and bluegills as influenced by simulated, submerged vegetation. Trans. Am. Fish. Soc. 111:255-266. Seaman, W., Jr., and M. Collins. 1983. Species profiles: life histories and environmental re- quirements of coastal fishes and invertebrates (South Florida— snook. U.S. Fish Wildl. Serv. FWS/OBS-82/11.16. Searle, S. R., F. M. Speed, G. A. Milliken. 1980. Population marginal means in the linear model: an alternative to least squares means. Am. Stat. 34:216-221. Sokal, R. R., and F. J. Rohlf. 1981. Biometry. W.H. Freeman, New York, NY, 859 p. Stoner, A. W. 1980. Feeding ecology of Lagodon rhomboides (Pisces: Sparidae): variation and functional responses. Fish. Bull. 78(21:337-352. 1982. The influence of benthic macrophytes on the forag- ing behavior of pinfish, Lagodon rhomboides (Linnaeus). J. Exp. Mar. Biol. Ecol. 58:271-284. 1983. Distribution of fishes in seagrass meadows: role of macrophyte biomass and species composition. Fish. Bull. 81(4):837-846. Thoemke, K. W. 1979. The life histories and population dynamics of four subtidal amphipods from Tampa Bay, Florida. Ph. D. diss., Univ. South Florida, FL, 139 p. Tyler, A. V. 1992. A context for recruitment correlations: why marine fisheries biologists should still look for them. Fish. Oceanogr. 1(1):97-107. U.S. Department of Commerce. 1990. Marine recreational fishery statistics survey, Atlan- tic and Gulf coasts, 1987-1990. U.S. Dep. Commer., NOAA, NMFS, Current Fishery Statistics 8904, Washing- ton, D.C., 363 p. U.S. Department of Commerce. 1992. Marine recreational fishery statistics survey, Atlan- tic and Gulf coasts, 1990-1991. U.S. Dep. Commer., NOAA, NMFS, Current Fishery Statistics 9204, Washing- ton, D.C., 275 p. Weinstein, M. P. 1983. Population dynamics of an estuarine-dependent fish, the spot ( Leiostomus xanthurus), along a tidal-creek- seagrass meadow coenocline. Can. J. Fish. Aquat. Sci. 40:1633-1638. Weinstein, M. P., K. L. Keck Jr., P. E. Giebel, and J. E. Gates. 1982. The role of herbivory in pinfish ( Lagodon rhomboides):a preliminary investigation. Bull. Mar. Sci. 32(3):791-795. Young, D. K., M. A. Buzas, and M. W. Young. 1976. Species densities of macrobenthos associated with seagrass: a field experimental study of predation. J. Mar. Res. 34:577-592. Young, D. K., and M. W. Young. 1977. Community structure of the macrobenthos associated with seagrass of the Indian River estuary, Florida. In B. C. Coull (ed.), Ecology of marine benthos, p. 359-381. Belle W. Baruch Library in Marine Science 6, Univ. South Caro- lina Press, Columbia, SC. 329 Abstract .—Low-frequency volume scattering measurements were con- ducted by the Naval Research Labora- tory (NRL) on the upper slope, slope base, and abyssal plain along the U.S. west coast between the Strait of Juan de Fuca, Washington, and Cape Mendo- cino, California. Comparisons of swim- bladder radii estimated from resonances with those estimated from fish lengths obtained from quasisynoptic National Marine Fisheries Service (NMFS) trawl catches of Pacific hake, Merluccius pro- ductus, strongly suggest that the major source of the low-frequency scattering were Pacific hake. Estimates of hake den- sity from the NRL low-frequency mea- surements and the NMFS acoustic, midwater trawl survey on the slope gave comparable values of 195-439 kg/ha and 91-369 kg/ha respectively. However, NRL measurements, up to 50 km offshore of the outer limit of the NMFS survey, found densities of 270-300 kg/ha in layers peaking at 225 to 450 m depth. This finding suggests that high densities of hake may occur farther offshore of the traditional limit of NMFS surveys and at depths which may be difficult to sur- vey with conventional fisheries sounders. Manuscript accepted 2 October 1997. Fishery Bulletin 96:329-343 ( 1998). Low-frequency acoustic measurements of Pacific hake, Merluccius productus, off the west coast of the United States Redwood W. Nero Charles H. Thompson Richard H. Love Naval Research Laboratory, Code 7174 Stenms Space Center, Mississippi 39529-5004 E-mail address for R.W. Nero:woody.nero@nrlssc.navy.mil Pacific hake, Merluccius productus, are the most abundant large mid- water fish on the continental shelf and slope of the west coast of the United States during summer (Ware and McFarlane, 1989). Hake winter between Point Conception, CA, and Baja California. In spring they migrate northward to feed in the productive waters along the con- tinental shelf from northern Cali- fornia to Vancouver Island. They remain in this area from May through October. The migration be- tween the winter spawning grounds and the summer feeding grounds is size structured, i.e. related to fish length and optimum swimming ve- locity (Francis, 1983; Smith et al., 1992), with the larger fish swim- ming faster and farther. Large fish migrate farthest north to near Vancouver Island, smaller fish reach only as far as southern Oregon and northern California (Stauffer, 1985). A midwater trawl fishery targets hake occurring in dense feeding ag- gregations along the shelf break in midsummer ( Stauffer, 1985). Since 1977, the National Marine Fisher- ies Service (NMFS) has conducted a triannual acoustic, midwater trawl survey to determine hake dis- tribution and abundance (Dorn et al., 1994). During the 1992 survey, NMFS integrated volume scattering with a high-frequency (38-kHz) echo sounder and corroborated fish iden- tity and size with midwater trawls. Volume scattering was converted to fish biomass with a volume scatter- ing to biomass regression (Dorn et al., 1994). Coincident with the 1992 fishery survey, the Naval Research Labo- ratory (NRL) conducted an acous- tic survey of volume scattering us- ing a broad-band low-frequency measurement system (Thompson and Love, 1996). Because the NRL survey was conducted quasisynop- tically with the fisheries survey, comparisons were possible between the two surveys. Although echo sounders at rela- tively high frequencies (>20 kHz) are a standard tool in fisheries sur- veys (MacLennan and Simmonds, 1992), a broad-band low-frequency sound at the natural resonance of the population of fish swimbladders at depth have been used as an alter- native tool (Holliday, 1972; 1977a). A variety of acoustic models has been developed to help predict and understand various aspects of this resonance (Anderson, 1950; An- dreeva, 1964; Love, 1978; Stanton, 1989; Clay, 1991, Feuillade and Werby, 1994). With these models, the acoustic resonance of a popula- tion of fish swimbladders at depth can provide quantitative informa- tion about the fish population. The spectrum of the resonance is depen- 330 Fishery Bulletin 96(2), 1 998 dent on swimbladder size, depth, and fish behavior (Sand and Hawkins, 1973; Love, 1993). The scattered sound at resonance is almost completely omnidirec- tional, independent of fish aspect (Feuillade and Werby, 1994). As a result, resonance measurements are useful for inferring the size and abundance of the swimbladders. An inverse solution of measured scattering levels can be used to generate a size-fre- quency distribution of swimbladders (Holliday, 1977b), from which fish size can then be inferred. On the U.S. west coast, Pacific hake, measuring 45 cm in fork length and occupying depths of 50 to 500 m, are expected to be resonant near 500 to 1,500 Hz (Love, 1978). Midwater trawls show that no other large pelagic fish are as abundant as hake (Dorn et al., 1994). Other fishes expected to contribute to scat- tering are anchovy and mesopelagic fishes (Kalish et al., 1986). Both are much smaller than hake and at depths of 50 to 500 m should only contribute to scattering above 2 kHz. This study reports on low-frequency volume scat- tering measurements from what are assumed to be Pacific hake at depths of 50 to 500 m over the conti- nental slope off the Strait of Juan de Fuca and the Oregon-California coasts. Bioacoustic models are used to estimate the number and size of swim- bladders at depth. These results are compared with estimates of swimbladder size and abundance ob- 1 1992. Alaska Fisheries Science Center, Natl. Mar. Fish. Serv., NOAA, 7600 Sand Point Way N.E., Seattle, WA 98115- 0070. Unpubl. data. tained from quasisynoptic acoustic data and mid- water trawl catches collected by the National Ma- rine Fisheries Service.1 A comparison of swimbladder size is problematic. Fish can either let their swim- bladders compress according to Boyle’s law during descent, or they can actively add gas to the swim- bladder to maintain neutral buoyancy. In addition, the relation between fish size and swimbladder vol- ume is also highly variable within a group of fish and is affected by feeding and other aspects of fish behavior (Ona, 1990). We consider these possibili- ties and suggest several hypotheses about the swimbladder behavior of Pacific hake. Materials and methods Measurements The NRL low-frequency acoustic measurements were made from 14 to 29 August 1992 aboard the USNS Wilkes (T-AGS-33) in conjunction with a Naval Oceanographic Office (NAVOCEANO) survey. The NRL conducted measurements at 9 stations located on the continental slope and abyssal plain between the Strait of Juan de Fuca and Cape Mendocino (Table 1; Fig. 1). Stations 2 and 3 off the Washington coast were missed owing to heavy seas. Because we expected a north-south trend in the size composi- tion of the hake, stations at similar latitude were grouped and named according to their proximity to geographic features: station 1 — La Perouse; 4 and Table 1 Volume reverberation stations conducted by the Naval Research Laboratory. Station no. Date Time (local [h] ) Day or night Lat. N Long. W Bottom depth (m) i 14 Aug 1836-1920 day 48°26' 126°20' 900 4 18 Aug 0120-0223 night 44°44' 125°04' 1,100 5 19 Aug 1325-1420 day 44°45' 125°42' 2,800 5 20 Aug 0145-0240 night 44°38' 125°43' 2,900 6 21 Aug 0000-0105 night 43°28' 125°40' 3,100 6 21 Aug 0925-1032 day 43°27' 125°38' 3,100 7 23 Aug 0025-0109 night 43°26' 125°04' 1,070 7 23 Aug 0858-0935 day 43°30' 125°06’ 1,245 8 24 Aug 0905-0936 day 42°18' 124°54' 1,025 9 25 Aug 2212-2248 night 42°19' 125°40' 2,800 9 26 Aug 0856-0927 day 42°23' 125°39' 2,800 10 27 Aug 0925-1001 day 40°15' 125°35' 2,400 11 28 Aug 0117-0200 night 40°04' 125°09' 1,465 11 29 Aug 0931-1006 day 40°03' 125°08' 1,500 Nero et al.: Low-frequency acoustic measurements of Merluccius productus 331 5 — Heceta North; 6 and 7 — Heceta South; 8 and 9 — Sebastian; and 10 and 11 — Mendocino. Just prior to the NRL survey, NMFS conducted a high-frequency acoustic-midwater trawl survey of Pacific hake. The survey began just south of San Francisco and ended off Cape Scott, Vancouver Is- land (Dorn et al., 1994). Generally, trawls were lo- cated inshore of the NRL stations. The NRL and NMFS surveys were separated by 43 days in the south (station 11, trawl 8), but only 5 days in the north (sta- tion 1, trawl 45)(Fig. 1). The NMFS survey provided information on fish size and abundance that could be compared with the results of the low-frequency acous- tic analysis. Comparisons were restricted to midwater trawls located close to the NRL stations within each of the above named latitudinal regions. A north-south trend in fish size was evident, although a gap occurred in the size distributions between 38 and 40 cm (Fig. 2). Trawls appeared to catch fish on either side of this gap; only a few trawls contained both size classes. This find- ing suggests that, at sea, different size classes remain spatially separate (Fig. 1). The NRL sound scattering measurements used an explosive sound source and a directional acoustic receiver. A 0.23-kg TNT charge detonated at 0.5 m depth provided a high source level over a wide fre- quency range. The shallow detonation depth allowed the gas bubble created by the explosion to vent to the surface and prevented the multiple sound pulses caused by bubble oscillations characteristic of charges detonated at greater depth. Measurement sequences consisted of 4 to 6 shots over a 30-min period. Because the ship typically drifted at 1-2 knots, the distance over which data were collected was on the order of a mile. The acoustic receiver used for these mea- surements consists of a thirty-two element line hydrophone, 0.9 m long, mounted along the axis of a conical reflector that had a height of 0.9 m and an open base with a diameter of 1.8 m. The purpose of the reflector was to map an annular area of the opening to each hydrophone element for sound entering the reflector parallel to the axis and to decrease sensitivity to sound entering the reflector from other directions. The receiver was originally de- signed for use at frequencies between 2.5 and 20 kHz. By varying the number of ac- tive hydrophone elements, the receiver’s 3-dB beam width can be maintained be- tween 10° and 20° in that frequency range. The eight elements nearest the vertex of the cone form a 45-cm aperture and are used for frequencies from 10 to 20 kHz. By doubling the active length with eight ad- ditional elements, a 90-cm aperture is formed that can be used for frequencies between 5 and 10 kHz, and all 32 elements form a 180-cm aperture that can be used below 5 kHz. Between 2.5 and 5 kHz, mea- sured beam patterns of the receiver show a main lobe with shape and 3-dB beam width similar to the main lobe of theoreti- cal beam patterns (Urick, 1983) for a plane circular array 1.8 m in diameter, whereas the side lobes are lower than those for a plane circular array. Below 2.5 kHz, the main beam widens beyond 20°, reaching 38° at 1,600 Hz, 58° at 800 Hz, and 74° at 500 Hz. Received signals from each TNT shot were amplified, high- and low-pass filtered NRL Stations + NMFS Trawls by fish size • Hake<40 o Hake>40 ❖ Mixed Longitude Figure 1 Map of the U.S. west coast showing Naval Research Laboratory (NRL) stations and locations where pertinent National Marine Fisheries Ser- vice (NMFS) midwater trawls were undertaken. 332 Fishery Bulletin 96(2), 1998 at 400 and 6,000 Hz, respectively, digitized at a 20- kHz sampling rate, and stored. Digitally stored sig- nals were subsequently filtered into 21 1/6 octave bands from 500 to 5,000 Hz and amplitude versus time envelopes calculated for each band. There was little variation between shots in a sequence. Enve- lopes of all shots in a sequence were averaged to ob- tain a single set that was representative of scatter- ing during that sequence, and this averaged data was used to calculate volume scattering strength, S , the scattering strength of a unit volume of water, as a function of depth. Data were displayed as 2-dimen- sional images showing Sv in dB on a color scale as a function of frequency and depth. Scatterer depths were determined from the Sv images; integration over these depths produced a series of layer scattering strength (SL) versus frequency curves. On the basis of the distinct nature of the SL curves, scatterers were assumed to be swimbladder-bearing fish. Hence, SL curves were used to determine swimbladder size with a swimbladder scattering model. Fish swimbladders are not spherical, nonetheless their size may be con- veniently expressed in terms of equivalent spherical radii (ESR) (i.e. the radii of spheres of equivalent vol- ume). Thus ESR can be derived independently from acoustic data and trawl data and compared to evalu- ate each method for assessing population statistics. Inverse solution The ESR of the scatterers were obtained from the SL curves of the acoustic data with the non- negative least squares solution of S = an, (1) ( Holliday, 1977b) for n, a column vector contain- ing the number of scatterers in each of p size classes of swimbladder radii. S is a column vec- tor of measured layer backscattering cross sec- tions a6 (SL = 10 log abSL ) for the q frequencies at which measurements were made. The ma- trix a is of dimension q x p and contains indi- vidual backscattering cross sections obs. that are calculated over each of the q = 21 measure- ment frequencies ( 1/6 octaves from 500 to 5,000 Hz) and p = 18 semilogarithmically spaced ra- dii (r= 0.25, 0.3, 0.35, 0.4, 0.45, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.2 1.4, 1.6, 1.8, 2.0, 2.25, and 2.5 cm) with a model of swimbladder resonance (Eq. 3 below). In solving Equation 1, the number of measurement frequencies varied from station to station depending on data quality, with from 4 to 6 of the lower 1/6 octaves excluded. This variation was usually due to lower-frequency data that were low level and affected by noise. The inverse solution was also restricted to ex- clude radii smaller than the radius expected to be resonant at 5,000 Hz at the depth of each layer. These radii were calculated with the sub- sequent Equation 4 for the resonance frequency, f(y The quality of the inverse solution of Equa- tion 1 for n was examined by comparing the original data, S, to an estimate of the layer strength obtained from the forward calculation 5 = an- Solution of Equation 1 above assumes that the abSL are the backscattering cross sections of a unit horizontal area of a layer of dispersed acoustically noninteracting scatterers, such that Nero et al.: Low-frequency acoustic measurements of Merluccius productus 333 i'=i where crfes. is the backscattering cross section of the ith scatterer and 0 is the number of individuals within a 1-m2 vertical column that extends over the depth of the layer. At some stations, usually at night, sev- eral layers were evident. In these cases, Equation 1 was solved for each individual layer, and the subse- quent distributions of ESR were summed. Swimbladder model The matrix c of individual backscattering cross sec- tions was calculated with Love’s model (1978) which models a fish swimbladder as an air-filled vis- cous spherical shell. The acoustic cross section in the back-scattered direction, 40 cm) and small hake (modal size <40 cm). These two groups rarely occurred in the same trawls. Where the two size groups overlapped (between 38°N and 43.5°N), two trawls caught large numbers of both size groups, two trawls caught large fish, and six trawls caught small fish. Some of the trawls with small fish were made near NRL stations 7 and 8. We hypothesize that at the time of the NRL survey (35 to 39 days after the NMFS survey), these small hake may have moved slightly inshore or south of the NRL stations and were replaced by larger fish. Our interpretation of resonance and calculation of swimbladder size and fish size required assumptions of the relation of swimbladder radius, r, to fish length, L. We also assumed that this relation was constant with depth. As gadoids migrate they do adjust the gas vol- ume of the swimbladder (Sand and Hawkins, 1973). Unfortunately our examination of swimbladder behav- ior in hake is inconclusive. The results of examining radius as a function of depth suggest that hake do compensate somewhat for pressure changes by add- ing and removing gas; however, they do not appear to compensate completely. We caution that this con- clusion is conjectural and could be an artifact of size- dependent stratification of hake in the water column. Overall, bias introduced in our estimates by not ac- counting for the possible addition and removal of gas to the swimbladders of fish descending and ascend- ing could have resulted in, assuming the worst case, an underestimate of length and biomass by 10% and 35%, respectively. This bias would not have affected our estimates of fish numbers because they were dependent on the number of radii independent of associated fish size. Comparison of the NMFS survey estimates of sur- face density with the NRL estimates of surface den- sity produced several important findings. First, al- though there were spatial and temporal differences in the surveys and although different acoustic tech- niques were used, the similarities between the sur- veys were encouraging and suggest that good accu- racy is obtained with both techniques. Second, an NRL measurement at one station (one-mile drift) may be influenced by local minima or maxima in hake density. This hypothesis is suggested by the high values at station 1 where only a few days previously, in NMFS samples, much lower values were obtained. By chance at station 1, NRL may have made a mea- surement over an isolated concentration of fish. Third, relatively high densities of fish were seen at offshore stations in NRL sampling. Some NMFS acoustic transects suggest a high abundance of fish near the offshore limit of several transects taken near 43°N (Dorn et al., 1994). This high abundance would indicate that, at times, high densities of hake occur offshore, probably as a result of a particular set of as-yet-unknown oceanographic conditions. This study has two important implications for fish- eries surveys. First, hake may occur in high num- bers, up to 300 kg/ha, approximately 50 km offshore of the outer edge of traditional fisheries acoustic sur- veys. Second, at these offshore sites the peak in the main scattering layer of hake may at times occur at depths of 400 to 450 m. At these depths hake could be difficult to assess by conventional 38-kHz fisher- ies sonars; therefore underestimates of the abun- dance of hake over deep water are possible. Acknowledgments We are indebted to J. J. Traynor, Alaska Fisheries Science Center, NMFS, for helping coordinate this incidental experiment and for providing a large amount of unpublished data. We thank N. V. Lombard, the Captain and crew of the USNS Wilkes, and E. Beeson, senior NAVOCEANO representative, for providing logistic support and assistance at sea. We also thank several anonymous reviewers who pro- vided critical reviews of the manuscript. This research was supported by the Office of Naval Research. Nero et al : Low-frequency acoustic measurements of Merluccius productus 343 Literature cited Anderson, V. C. 1950. Sound scattering from a fluid sphere. J. Acoust. Soc. Am. 22:426-431. Andreeva, I. B. 1964. Scattering of sound by air bladders of fish in deep sound- scattering ocean layers. Sov. Phys. Acoust. 10:17-20. Clay, C. S. 1991 . Low-resolution acoustic scattering models: fluid-filled cylinders and fish with swim bladders. J. Acoust. Soc. Am. 89:2168-2179. Dorn, M. W., E. P. Nunnallee, C. D. Wilson, and M. E. Wilkins. 1994. Status of the coastal Pacific whiting resource in 1993. U.S. Dep. Commer., NOAA Tech. Memo. NMFS- AFSC-47, 101 p. Feuillade, C., and M. F. Werby. 1994. Resonances of deformed gas bubbles in liquids. J. Acoust. Soc. Am. 96:3684-3692. Foote, K. G. 1985. Rather high frequency sound scattering by swim- bladdered fish. J. Acoust. Soc. Am. 78:688-700. Francis, R. C. 1983. Population and trophic dynamics of Pacific hake (Merluccius productus). Can. J. Fish. Aquat. Sci. 40:1925- 1943. Holliday, D. V. 1972. Resonance structure in echoes from schooled pelagic fish. J. Acoust. Soc. Am. 51:1322-1332. 1977a. The use of swimbladder resonance in the sizing of schooled pelagic fish. Rapp P.-V. R. Cons. Int. Explor. Mer 170:130-135. 1977b. Extracting bio-physical information from the acous- tic signatures of marine organisms. In N. R. Andersen and B. J. Zahuranec ( eds. ), Oceanic sound scattering pre- diction, p. 619-624. Plenum Publishing Corp., New York, NY. Jones, F. R. H., and N. B, Marshall. 1953. The structure and functions of the teleostean swim- bladder. Biol. Rev. 28:16-83. Jones, F. R. H., and P. Schole 1985. Gas secretion and resorption in the swimbladder of the cod Gadus morhua. J. Comp. Physiol. B 155:319-331. Kalish, J. M., C. F. Greenlaw, W. G. Pearcy, and D. V. Holliday. 1986. The biological and acoustical structure of sound scat- tering layers off Oregon. Deep-Sea Res. 33:631-653. Livingston, P. A. 1983. Food habits of Pacific whiting, Merluccius productus, off the west coast of North America, 1967 and 1980. Fish. Bull. 81:629-636. Love, R. H. 1978. Resonant acoustic scattering by swimbladder-bear- ing fish. J. Acoust. Soc. Am. 64:571-580. 1993. A comparison of volume scattering strength data with model calculations based on quasisynoptically collected fishery data. J. Acoust. Soc. Am. 94:2255-2268. MacLennan, D. N., and E. J. Simmonds. 1992. Fisheries acoustics. Chapman and Hall, London, 325 p. Nero, R. W., C. H. Thompson, and R. H. Love. 1997. Abyssopelagic grenadiers: the probable cause of low frequency sound scattering at great depths off the Oregon and California coasts. Deep-Sea Res. 44:627-645. Ona, E. 1990. Physiological factors causing natural variations in acoustic target strength of fish. J. Mar. Biol. Assoc. U.K. 70:107-127. Sand, O., and A. D. Hawkins. 1973. Acoustic properties of the cod swimbladder. J. Exp. Biol. 58:797-820. Smith, B. D., G. A. McFarlane, and M. W. Saunders. 1992. Inferring the summer distribution of migratory Pa- cific hake ( Merluccius productus ) from latitudinal varia- tion in mean lengths-at-age and length frequency distribu- tions. Can. J. Fish. Aquat. Sci. 49:708-721. Stanton, T. K. 1989. Simple approximate formulas for backscattering of sound by spherical and elongated objects. J. Acoust. Soc. Am. 86:1499-1510. Stauffer, G. 1985. Biology and life history of the coastal stock of Pacific whiting, Merluccius productus. Mar. Fish. Rev. 47:2-7. Thompson, C. H., and R. H. Love. 1996. Determination of fish size distributions and areal densities using broadband low-frequency measurements. ICES J. Mar. Sci. 53:197-201. Urick, R. J. 1983. Principles of underwater sound, 3rd ed. McGraw-Hill Book Co., New York, NY, 423 p. Ware, D. M., and G. A. McFarlane. 1989. Fisheries production domains in the northeast pa- cific ocean. In R. J. Beamish, and G. A. McFarlane (eds.), Effects of ocean variability on recruitment and an evalua- tion of parameters used in stock assessment models, p. 359- 379. Can. Spec. Publ. Fish. Aquat. Sci. 108. Weston, D. 1967. Sound propagation in the presence of bladder fish. In V. M. Albers (ed.), Underwater acoustics, vol 2, p. 55- 88. Plenum Press, New York, NY. Ye, Z. 1977. Low-frequency acoustic scattering by gas-filled prolate spheroids in liquids. J. Acoust. Soc. Am. 101:1045-1052. 344 Abstract .“Annual fluctuation in the number of newly settled juveniles and in the stock size of the sunray surf clam, Mactra chinensis, were examined from 1987 to 1994 off the coast of Tomakomai, southwest Hokkaido. Stock size was estimated with a model based on two known population parameters, juvenile density and age composition. Juvenile density and the stock size (or mass) ranges were 1.3 to 157.0 ind./m2 and 249.8 to 1,127.4 metric tons, re- spectively; a time lag between the rela- tive fluctuations of both agreed with age-at-recruitment to the stock. Our results suggest that population dynam- ics are directly influenced by the num- ber of juveniles. The predicted stock size approximated the measured value and, consequently, a long-term predic- tion of stock size was deemed possible, provided age composition and juvenile density are determined. Manuscript accepted 16 June 1997. Fishery Bulletin 96:344-351 (1998). Population dynamics and stock size prediction for the sunray surfclam, Mactra chinensis, at Tomakomai, southwest Hokkaido, Japan Izumi Sakurai Hokkaido Central Fisheries Experimental Station 238, Hamanaka, Yoichi, Hokkaido 046, Japan E-mail address, kb6i-skri@asahi-net.or.jp Takashi Horii Muroran Branch, Hakodate Fisheries Experimental Station 1-133-31, Funami. Muroran, Hokkaido 05 1 , Japan Osamu Murakami Hokkaido Abashiri Fisheries Experimental Station 3 1 , Masuura, Abashiri, Hokkaido 099-3 1 , Japan Shigeru Nakao Faculty of Fisheries, Hokkaido University 3-1-1, Minato, Hakodate, Hokkaido 041, Japan The sunray surfclam, Mactra chi- nensis, is a commercially important bivalve belonging to the family Mactridae and is widely distributed on the upper subtidal sandy bottom off the coast of Japan, China, Ko- rea, Sakhalin, and the maritime province of Siberia (Habe, 1977). This clam has been harvested from marine waters off Hokkaido as in- cidental catch in the fishery for Japanese surfclam, Pseudocardium sachalinensis, which is the most commercially important bivalve in northern Japan (Kinoshita and Terai, 1954). Recently, the demand for M. chin- ensis caught off Hokkaido has in- creased owing to decreasing catches on Honshu Island (Kurata, 1991). However, from 1985 to 1994, the annual Hokkaido catch fluctuated between 705 and 1,310 metric tons. In response to this wide fluctuation, stock management plans are now required. Although there are some studies of the breeding season (Miyazaki, 1957; Tomita, 1974), growth (Hanaoka and Shimazu, 1949), spatial distribution (Hayashi et al., 1965, 1967), increase in the number of juveniles (Inoue and Ozu, 1960; Yoshimatsu, 1977), and catch fluctuation (Saito et al., 1982), these studies have provided only frag- mentary information. Of necessity, several temporary stock manage- ment measures, similar to those used forP. sachalinensis, have been implemented by several fisheries co- operative associations in Hokkaido to regulate the fishery for M. chinensis . These measures include limiting allowable catch and impos- ing a minimum harvestable shell length and period or area for fish- ing. We have systematically stud- ied the life history of M. chinensis and have already reported on its re- productive cycle (Sakurai et al., 1992), age at first maturation (Sa- kurai et al., 1992), relation between Sakurai et a I.: Population dynamics and stock size of Mactra chinensis 345 age and growth (Sakurai, 1993), spatial distribution pattern (Sakurai, 1994), and annual mortality rate (Sakurai, 1996) off the coast of Tomakomai, south- west Hokkaido, Japan. The aim of this study is first to examine the fluc- tuation in density of newly settled and harvestable- size M. chinensis at Tomakomai and to predict the stock size according to the population parameters that we have already reported. Materials and methods Study area The study was conducted on a subtidal sandy bot- tom off Tomakomai (42°36'N, 141°32'E), where M. chinensis with shells longer than 60 mm are har- vested commercially. Figure 1 shows the survey area in relation to the entire fishing ground, which is about 32 km2 in area. The survey area is situated on the fishing grounds within 6 km of shore and covers depth ranges from 3 to 11 m. The sediment in the area is fine or very fine sand with low organic car- bon content, less than 0.3% of dry weight (Sakurai et al., 1991). The bottom water temperature at 10-m depth ranges between 3.0°C in March and 20.2°C in September (Sakurai, 1993). Sampling and processing of clams The breeding season of M. chinensis occurs between July and September at Tomakomai (Sakurai et al., 1992), the planktonic larval period lasts for 14-21 days (Tsurui, 1980; Kobayashi and Ujima, 1983), and shells of newly settled juveniles grow to 1-3 mm long between September and November (Sakurai, 1994). The density of newly settled juveniles was therefore Figure 1 Location of the survey area off the coast of Tomakomai, southwest Hokkaido, Japan. (A) Shaded and dotted portions represent the entire fishing ground of M. Chinensis', dotted portion represents the survey area. (B) Solid and open circles indicate stations sampled by a Smith-Mclntyre grab sampler and Japanese surfclam hydraulic jet dredge, respectively. 346 investigated during September and November in 1987-94 as tabulated in Table 1. Thirty sampling stations used for the collection of juveniles were ar- ranged as shown in Figure 1. Each of the stations was set up at six transects including depths of 3, 5, 7, 9, and 11 m for evaluating the average density of juveniles in the fishing ground. The distribution of juveniles had been previously observed at 9 m, par- allel to the shoreline in Tomakomai (Sakurai, 1994). Two samples were collected at each station with a Smith-Mclntyre grab (sampling area: 0.05 m2), sieved through a 1-mm mesh, and preserved immediately in 5% buffered formalin in the field. Sorting, identi- fication based on morphological characteristic of the shell (Habe, 1977), and counting of juveniles were done in the laboratory. Harvestable-size clams with shells >60 mm long were collected at 18 sampling stations in April, 1987- 94 (Fig. 1; Table 1). Samples were collected with a hydraulic jet dredge (width=1.2 m, mesh size of net bag=70 mm, spacing of teeth=36 mm, and angle of tooth=60°) normally used for collecting P sachali- nensis. The dredge was towed parallel to shore for 50-100 m at each station. Distance of each tow was estimated by measuring the length of the dredge- connected rope wrapped by a winch. Sorting was con- ducted in the field. Shell lengths were measured to the nearest 0.1 mm with a sliding caliper in the labo- ratory. Clams with shells >60 mm long were grouped into 5-mm intervals and counted at each station. These counts were subsequently adjusted on the ba- sis of catch efficiencies of the dredge for the different 5-mm-long intervals (60-65 mm to 0.75, 65-70 mm to 0.82, 70-75 mm to 0.88, 75-80 mm to 0.95, and >80 mm to 1.00), which were based on the selectiv- ity curve of M. chinensis (Nashimoto, 1984), before density (ind./sample area) was estimated. The den- sity was then converted to biomass (g/sample area) Table 1 Sampling dates of newly settled juveniles and harvestable- size clams of M. chinensis. Newly settled Harvest-size Year juveniles clams 1987 20 Nov 2-5 Apr 1988 15 Nov 1-4 Apr 1989 21 Nov 2-5 Apr 1990 4 Sep 11, 12 Apr 1991 26 Nov 10, 11 Apr 1992 15 Oct 16, 18 Apr 1993 20 Sep 12, 13 Apr 1994 29 Nov 20, 21 Apr Fishery Bulletin 96(2), 1 998 by using the weight (g) equivalent for the median shell length of each size interval. The length-weight equivalent was determined from 107 clams with shells 57.8-83.4 mm long sampled at station N (Fig. 1) in April 1991 as follows: log W = 3. 16 log L - 4.071, (r = 0.9728, P < 0.01) (1) where W = the weight (g); and L = the shell length (mm). Next, the stock size (Ng; unit: t [age in years]) in the survey area was estimated with the following equation: 18 Ne = 18-1 A^bjr'w-1 x 1(T6 i= 1 where A = survey area (72 x 105 m2); bt = biomass (g/sample area) at each station; l. = tow distance (m) at each station; and w = width of the dredge (1.2 m). In M. chinensis, an external growth ring with a light penetrable band, as shown in Figure 2, is formed annually on the shell during declining water tem- perature, November-January, and is able to be dis- tinguished from other rings (Sakurai, 1993). For the stock size prediction, therefore, the age of clams was determined by counting these rings. Clams with shells >60 mm long sampled at station N (Fig. 1) in April 1991 were used for age determination because the distribution of harvestable-size M. chinensis was observed at 9 m parallel to the shoreline in Toma- komai (Sakurai, 1996) and because specimens were considered typical of age composition in the survey area. Age determination was conducted by first wash- ing out the shell periostracum of the specimens with a chlorine bleaching agent and by drying it in the shade. Next, the specimens were put on a light box, and external growth rings with light penetrable bands were counted. Age-determined clams were grouped into 5-mm intervals at each age and counted. These counts (ind./ sample area) were adjusted by using the above effi- ciencies, and then density (ind./m2) was estimated at each age by multiplying the counts by tow dis- tance and width of the dredge. Stock size prediction The predicted stock size (Npi) [=1992,1993, . . . , 1997; unit: /), based on the age composition in April 1991, was estimated for each year class as follows: (2) Sakurai et al. : Population dynamics and stock size of Mactra chinensis 347 Np' = A| X W'+i + no,i-3 So-,( W,c I x 10 6 (3) where n( = density (ind./m2) at age t; t = age at first capture; and tx = longevity. Although M. chinensis attains a commercial size at age 2.3 yr (i.e. in November of the second year) at Tomakomai (Sakurai, 1993), we assumed t to be 2.7 yr (i.e. in April of the second year) because the real stock size and the age composition in this study were inves- tigated during April. St represents the survival rate at age t and was calculated from the following equation: St = l-100-1m(, (4) where mf is the annual mortality rate at age t (age 0 yr: 94.0%; 1 yr: 53.1%; 2 yr: 44.8%; 3-9 yr: 42.7%) calculated with data from Sakurai (1996) with the assumption that annual mortality is constant. In April 1992-97, nt was estimated in turn by multi- plying the previous year by each Sf and by adding density of recruitment which was estimated by mul- tiplying each n0,._3 by SQ tc. Density at age 10 yr and older (^>10) was not considered in the estimated age composition because the longevity (tx) of M. chinensis is 10 yr (Sakurai, 1993). The weight at age [f+0.7] yr, Wt, was calculated according to Equation 1 and the following equation from Sakurai (1993) : L = 78.3ljl-e'067u~034)}, (5) where L = shell length (mm); and t = age (years) of M, chinensis. 348 Fishery Bulletin 96(2), 1998 1987 1988 1989 1990 1991 1992 1993 1994 Year Figure 3 Annual fluctuation of the mean density of newly settled juvenile sunray surfclams. Vertical bars represent standard deviations. 2000 H 1500 1 1000 500 0 3 (D CD Year Figure 4 Annual fluctuation of the density (open circles) and the measured stock size (solid circles) of sunray surfclams. Results The densities of newly settled juveniles, as shown in Figure 3, ranged from 1.3 to 157.0 ind./m2 in 1987-94. The densities and the measured stock size of harvestable-size clams ranged from 0.7 to 3.1 ind./m2 and from 249.8 to 1127.4 tons in 1987- 94, respectively (Fig. 4). The age composition of sunray surfclam in April 1991 is shown in Figure 5. The 2-year age group predominated with a share of 33.9%; the 6-, 7- and 3-year groups accounted for 17.3%, 12.2%, and 11.1%, respectively, and the other age groups rep- resented less than 10% of the whole composition. Clams older than 9 years were not found in the survey area. The predicted stock size at each age in April 1992-97 calculated from Equation 3 is shown in Figure 6. It was predicted that the age groups spawned in 1988 and 1991 would predominate from age 2 to 5 in the har- vestable-size clams. Although bivalve fishing was prohib- ited in 1985-91 at depths of 5 to 10 m off the coast of Tomakomai, including the survey area, in order to protect P. sack alinensis juveniles, fishing resumed there in July 1992. Therefore, the catch of M. chinensis from 1992 to 1994 in the survey area was estimated by multiply- ing the annual catch at Tomakomai (Hokkaido, 1994, 1995, 1996) by the ra- tio of the survey area to the whole fish- ery area (Table 2). Subsequently, stock size was estimated by subtracting the value of the survey area in Table 2 from the sum of the age composition in Figure 6. The results are shown in Figure 7 to- gether with the real stock sizes from Fig- ure 4. The predicted values were within about 10% of the real values (Table 3). Discussion It is well known that marine invertebrates with planktonic larval stages show considerable fluctua- tions in numbers of recruits; sudden increases in numbers of juveniles occur occasionally (Hanaoka, 1972). Nakaoka (1993) noted that it is necessary for these populations to experience such sudden in- creases in order to maintain their populations. Sud- den increases of M. chinensis have been reported from many of the fishing grounds in Japan. For example, a maximum density of about 980 ind./m2 for clams with 60-mm-long shells was reported off the coast of Miura, Kanagawa Prefecture (Inoue and Ozu, 1960). In the present study, densities of newly settled juve- niles ranged from 1.3 to 157.0 ind/m2 (Fig. 3); during our study we did not find such a remarkable fluctua- tion in recruitment off the coast of Tomakomai com- pared with that observed for the population off Miura. Additionally, there were no dominant or vacant age groups in the age composition (Fig. 5); thus, it ap- pears that the population of M. chinensis is main- tained by comparatively stable recruitment occurring off the coast of Tomakomai. Mactra chinensis grows to harvestable size at age 2.3 yr at Tomakomai (Sakurai, 1993); consequently newly settled juveniles spawned in the years 1987 through 1991 are presumed to have recruited to the Sakurai et al.: Population dynamics and stock size of Mactra chinensis 349 stock in April 1990-94. As Figures 3 and 4 show, juvenile densities in 1988, 1989, and 1991 were relatively high, causing the real stock sizes in 1991, 1992, and 1994 to increase in comparison with each previous year. In contrast, the lower juvenile densities in 1987 and 1990 caused the real stock size in 1990 and 1993 to decrease in comparison with each previous year. Therefore, it is suggested that the population dynamics at Tomakomai are directly influenced by the num- ber of newly settled juveniles. The number of juveniles is determined by the settlement and mortality of early benthic stages (Gunther, 1992). We think that it is necessary to examine the fac- tors that cause fluctuations in the number of ju- veniles to gain a full understanding of the popu- lation dynamics of M. chinensis. In the present study, we define a model using two known population parameters, juvenile den- sity and age compositions, to predict the stock size of M. chinensis. It has been shown that the age groups that have relatively higher densities at the juvenile stage predominate for four years from age 2 to 5 in harvestable-size clams. Saito et al. (1982) reported that a large catch of M. chinensis contin- ued for a few years off the coast of Ishikari, Hokkaido, because the dominant age group was maintained. This was the case in our study. Therefore, we con- sider that our model is appropriate for predicting the population dynamics of M. chinensis at Tomakomai and other sites. Furthermore, the predicted stock size of our model was close to the measured value (Fig. 7), and consequently accurate long-term prediction of stock size is possible with this model, provided age composition and juvenile density are determined. On the other hand, analysis of age and growth based on external growth rings has been conducted for various bivalves including Placopecten magel- lanicus (e.g. Stevenson and Dickie, 1954; Claereboudt and Himmelman, 1996), Pecten maximus (e.g. Dillon and Clark, 1980), Megangulus venulosus (Goshima et al., 1991), Mesodesma mactroides (Defeo et al., 1992) , Clinocardium californiense (Goshima and Noda, 1992), and Abra tenuis (Dekker and Beukema, 1993) . However, it is suggested that age determina- tion based on external growth rings of bivalves would be unreliable because it is difficult to distinguish the growth rings from other rings owing to several stimuli (Dillon and Clark, 1980). Nevertheless, the growth rings of M. chinensis have been distinguished clearly from other innumerable fine rings by check- ing the rings with light penetrable bands (Sakurai, 1993). These bands are regarded as nacreous layers. Therefore, age determination in the present study would be as reliable as that based on microgrowth Table 2 Estimated catch of M. chinensis in the survey area and whole fishery ground off the coast of Tomakomai. Year Catch (metric ton) Survey area Whole fishery ground7 1992 7.2 36.0 1993 14.6 73.0 1994 22.0 108.0 1 Based on data from Hokkaido Fisheries Statistics in 1994, 1995, and 1996. Table 3 Comparison of the real [R) and predictive (P) stock size for M. chinensis. Year R (metric ton) P (metric ton) ( R-P ) x 100/R (%) 1992 728.3 651.8 10.5 1993 550.3 586.0 -6.5 1994 1,127.4 1,015.8 9.9 analysis with acetate peels or thin sections (Lutz and Rhoads, 1980), and we consider that our model would be able to predict the stock size without over- or un- derestimates. In our model, however, annual mor- tality is assumed to be constant; therefore, age com- 350 Fishery Bulletin 96(2), 1998 position would need to be reinvestigated if there was an unexpected change in mortality, such as that re- sulting from clams being washed ashore by a large typhoon or other storms (Sakurai et al., 1996). 23456789 Age (year) Figure 6 Predicted age compositions for the harvestable-size sunray surfclams in April from 1991 to 1997. Year Figure 7 Annual fluctuation of the measured (open circles) and predicted (solid circles) stock size of sunray surfclams. Acknowledgments We wish to express our appreciation to Y. Kanno and S. Goshima, Faculty of Fisheries, Hokkaido University, for their helpful advice and discussion. We are also grateful to the staff of Tomakomai Fisheries Cooperative Associa- tion for their kindness during this work. Literature cited Claereboudt, M. R., and J. H. Himmelman. 1996. Recruitment, growth and production of giant scallops ( Placopecten magellanicus) along an envi- ronmental gradient in Baie des Chaleurs, eastern Canada. Mar. Biol. 124:661-670. Defeo, O., E. Ortiz, and J. C. Castilla. 1992. Growth, mortality and recruitment of the yellow clam Mesodesma mactroides on Uruguayan beachs. Mar. Biol. 114:429-437. Dekker, R., and J. J. Beukema. 1993. Dynamics and growth of a bivalve, Abra tenuis, at the northern edge of its distribution. J. Mar. Biol. Assoc. U. K. 73:497-511. Dillon, J. F., and G. R. Clark II. 1980. Growth-line analysis as a test for contem- poraneity in populations. In D. C. Rhoads and R. A. Lutz (eds. ), Skeletal growth of aquatic organ- isms, p. 395-415. Plenum Press, New York, NY. Goshima, S., K. Nagamoto, K. Kawai, and S. Nakao. 1991. Reproductive cycle and growth of the north- ern great tellin, Megangulus venulosus, in Shiri- uchi, Hokkaido. Benthos Res. 40:23-33. [In Japa- nese with English abstract.] Goshima, S., and T. Noda. 1992. Shell growth of the North Pacific cockle Clinocardium californiense in Hakodate Bay, Hok- kaido. Benthos Res. 42:39-48. [In Japanese with English abstract.] Gunther, C. P. 1992. Settlement and recruitment of Mya arenaria L. in the Wadden Sea. J. Exp. Mar. Biol. Ecol. 159:203-215. Habe, T. 1977. Systematics of Mollusca in Japan. Bivalvia and Scaphopoda. Hokuryukan, Tokyo, p. 178-180. [In Japanese.] Hanaoka, T. 1972. Increase in number of animals. In S. Iwao and T. Hanaoka (eds.), Kyoritsu Shuppan, Tokyo, p. 88-91. [In Japanese.] Hanaoka, T., and T. Shimazu. 1949. Studies on the morphometry and rate of growth in clam, Mactra suleataria Reeve, in Tokyo Bay. Nippon Suisan Gakkaishi 15:311-317. [In Japanese.] Sakurai et a\.: Population dynamics and stock size of Mactra chinensis 351 Hayashi, T., K. Kawamura, Y. Yokoyama, and T. Hanada. 1965. Survey report on sunray surf clam Mactra chinensis in Ishikari, Hokkaido. J. Hokkaido Fish. Exp. Sta. 22:21- 43. [In Japanese.] Hayashi, T., K. Terai, and K. Arima. 1967. Studies on the planktonic larvae and young shells of Japanese surf clam, Spisula sachalinensis on the coast of Yakumo, Oshima Prov., Hokkaido. Sci. Rep. Hokkaido Fish. Exp. Sta. 7:8-71. [In Japanese with English abstract. 1 Hokkaido. 1994. Hokkaido annual fishery statistics in 1992. Hokkai- do Government, Sapporo, Hokkaido, p. 38-39. [In Japanese.] 1995. Hokkaido annual fishery statistics in 1993. Hokkai- do Government, Sapporo, Hokkaido, p. 38-39. [In Japanese.] 1996. Hokkaido annual fishery statistics in 1994. Hokkai- do Government, Sapporo, Hokkaido, p. 38-39. [In Japanese.] Inoue, M., and J. Ozu. 1960. Unusual increase in the surf clam ( Mactra chinensis Reeve) population off Kamimiyata in Tokyo Bay. Suisan- zosyoku 7:1-6. [In Japanese.] Kinoshita, T., and K. Terai. 1954. Utilization of sunray surf clam Mactra chinensis. J. Hokkaido Fish. Exp. Sta. 11:29-32. [In Japanese.] Kobayashi, M., and J. Ujima. 1983. Annual report of 1981 fiscal year of Fukuoka Pref. Buzen Fish. Exp. Sta., p. 87-91. [In Japanese.] Kurata, M. 1991. Fishes and marine invertebrates of Hokkaido. In K. Nagasawa and M. Torisawa (eds.). Kita-nihon Kaiyo Center, Sapporo, p. 252-253. [In Japanese.] Lutz, R. A., and D. C. Rhoads. 1980. Growth patterns within the molluscan shell: an over- view. In D. C. Rhoads and R. A. Lutz (eds.). Skeletal growth of aquatic organisms, p. 203-254. Plenum Press, New York, NY. Miyazaki, I. 1957. Aquaculture for bivalve. Isana-syobo, Tokyo, p.77- 79. [In Japanese.] Nakaoka, M. 1993. Effect of recruitment fluctuation on population dy- namics of marine benthos: analysis using projection ma- trix model. Gekkan Kaiyo 25:264-268. [In Japanese.] Nashimnoto, K. 1984. The selectivity of the sunray surf clam dredge. Nip- pon Suisan Gakkaishi 50:1145-1155. [In Japanese with English abstract.] Saito, K., T. Miyamoto, and Y. Nagai. 1982. Some ecological notes of Japanese surf clam Pseudo- cardium sybillae, Chinese surf clam Mactra chinensis , and great northern tellin Peronidia venulosa at Ishikari, Hok- kaido. J. Hokkaido Fish. Exp. Sta. 39:211-229. [In Japanese.] Sakurai, I. 1993. Age and growth ofthe sunray surf clam Mactra chinensis in Tomakomai, southwest Hokkaido. Nippon Suisan Gakkaishi 59:469-472. [in Japanese with English abstract.] 1994. Distribution and mortality of the sunray surf clam Mactra chinensis in young stages in Tomakomai, south- west Hokkaido. Nippon Suisan Gakkaishi 60:585-591. [In Japanese with English abstract.] 1996. Population ecological studies on the resource man- agement for the sunray surf clam Mactra chinensis. Ph.D. diss., Hokkaido Univ., Hokkaido, 196 p. [In Japanese.] Sakurai, I., M. Kurata, and E. Abe. 1996. Age structure and mortality of the sunray surf clam Mactra chinensis off Tomakomai, southwest Hokkaido. Fisheries Sci. 62:168-172. Sakurai, I., M. Kurata, and T. Miyamoto. 1992. Breeding season of the sunray surf clam Mactra chinensis in Tomakomai, southwest Hokkaido. Nippon Suisan Gakkaishi 58:1279-1283. [In Japanese with En- glish abstract.] Sakurai, I., T. Miyamoto, and K. Takahashi. 1991. Environmental characteristics and distribution of young bivalves in Japanese surf clam bed off Tomakomai coast, Hokkaido. Sci. Rep. Hokkaido Fish. Exp. Sta. 36:39- 59. [In Japanese with English abstract.] Stevenson, J. A., and L. M. Dickie. 1954. Annual growth rings and rate of growth of the giant scallop, Placopecten magellanicus (Gmelin) in the Digby area ofthe Bay of Fundy. J. Fish. Res. Board Can. 2: 660-671. Tomita, K. 1974. Some studies about the sunray surf clam Mactra chin- ensis. J. Hokkaido Fish. Exp. Sta., 31:8-14. [In Japanese.] Tsurui, K. 1980. Artificial egg collection and initial breeding for Mactra chinensis. Saibai Giken 9: 13-20. [In Japanese.] Yoshimatsu, S. 1977. Survey report on outbreak of bivalves in Kagawa Prefecture. Saibai Giken 6:1-12. [In Japanese.] 352 Reproductive biology, growth, and natural mortality of Puget Sound rockfish, Sebastes emphaeus (Starks, 1911) Andreas T. Beckmann Prof. Hofmann Laboratory Department of Zoology and Parasitology Ruhr-University Bochum, ND 05-334, 44801 Bochum, Germany Donald R. Gunderson Bruce S. fVHSIer School of Fisheries University of Washington, Seattle, Washington 98195 E-mail address (for Donald R. Gunderson, contact author) dgun@fish.washington.edu Raymond M. Buckley Washington Department of Fish and Wildlife University of Washington, Seattle, Washington 98 1 95 Betty Goetz Alaska Fisheries Science Center National Marine Fisheries Service, NOAA 7600 Sandpoint Way N.E., Seattle, Washington 98115 In spite of being one of the most common rockfish in the rocky near- shore region of the Strait of Geor- gia, relatively little is known about the early life history, ecology, re- cruitment, and reproductive biology of Puget Sound rockfish, Sebastes emphaeus. In rockfish, fecundity at 50% maturity ranges from 2,000 (S. dalli ) to 417,000 (S. paucispinis) eggs per female (Haldorson and Love, 1991). Parental care in this genus is essentially lecithotrophic, characterized by primitive, unspec- ialized viviparity at an evolution- ary stage when eggs, embryos, and early larval stages are protected inside the female body rather than born live as fairly advanced young (Wourms, 1991). Although most other northeastern Pacific Sebastes release their young between Janu- ary and July (Westrheim, 1975), parturition occurs in August and September for Puget Sound rockfish (Moulton, 1975; Beckmann, 1995). The goals of this study were to delineate the spawning season more closely and to determine the age and length at first maturity, the growth and mortality rates, and the length-fecundity relationship for Puget Sound rockfish. Materials and methods A total of 362 Puget Sound rockfish older than age 0 were collected dur- ing 23 sampling trips in coastal waters of the San Juan Islands, from 26 June 1994 to 12 March 1995 (Table 1). The fish were col- lected either by anglers using hook- and-line gear or by SCUBA divers using spears and hand nets. The average collection depth for these rockfish was 15 m to 21 m, with a maximum depth of 76 m. In addition, 19 collection trips were made with hand nets for young-of-the-year (YOY) speci- mens. These trips extended from 26 February through 18 August 1994, and 424 YOY were obtained. The spawning period was defined as occurring from the time when the first females with spent ovaries were caught to the time when no females with embryos were ob- served. A total of 253 females older than age 0, collected between June 1994 and March 1995, were classi- fied by maturity stage according to the criteria described in Finckh McDermott ( 1994). Nonlinear least square regression was used to fit the data to the logistic model: p - 1 , L l+e~(a+pL) where PL = proportion mature at length L (mm FL); a. P = constants; and a ~~n = length at which 50% of fish are mature ( =L() 5). Sagittal otoliths were removed from the fish and stored in 50% ethanol. Readings prepared by the break-and-burn method were taken either from left or right otoliths, following the international conven- tion of considering 1 January as the birth date. All final age determina- tions were made by the coauthor with extensive age-determination experience (B. Goetz), after exam- ining the results obtained by the senior author and then examining edge type and annulus formation for the entire collection. A sample of 59 otoliths was read twice (with six months between readings) and showed 66% agreement (to the ex- Manuscript accepted 7 August 1997. Fishery Bulletin 96: 352-356 (1998). NOTE Beckmann et ai : Reproductive biology, growth, and natural mortality of Sebastes emphaeus 353 Table 1 Date and sampling method for collections of Puget Sound rockfish older than age 0. Date Number collected Method Date Number collected Method 26 Jun 1994 6 spear 3 Aug 1994 41 hook and line 28 Jun 1994 4 spear 8 Aug 1994 19 hook and line 29 Jun 1994 12 spear or hand net 21 Aug 1994 11 spear or hook and line 5 Jul 1994 5 hook and line 26 Aug 1994 8 hook and line 6 Jul 1994 22 spear 29 Aug 1994 8 hook and line 7 Jul 1994 11 spear 30 Aug 1994 10 hook and line 7 Jul 1994 20 hook and line 2 Sep 1994 21 hook and line 13 Jul 1994 35 hook and line 6 Sep 1994 7 hook and line 22 Jul 1994 15 hook and line 10 Sep 1994 13 hook and line 24 Jul 1994 3 hand net 30 Dec 1994 21 hook and line 26 Jul 1994 6 hand net 12 Mar 1995 37 hook and line 28 Jul 1994 27 hook and line act year) between independent readings, with no obvious bias. Age-length relationships were estimated from the mean length at age for age groups 1-13, derived from the otolith readings. The mean length of age-0 fish was estimated from the mean length of the YOY cap- tured with nets during February- August 1994. Lengths for YOY were originally recorded as stan- dard length, and were converted to fork length by using the relationship for shortbelly rockfish (S. jordani) reported by Echeverria and Lenarz (1984). The mean age of YOY (0.7 years) was estimated by assuming that parturition occurred on 1 September. Mean length at age for age groups 1-13 was esti- mated from fish collected during June-December, 1994. Data were fitted to the von Bertalanfly growth model (Ricker, 1975) with nonlinear least squares re- gression. Male and female fish were treated separately owing to the difference in sex-specific growth rates. Mortality was estimated from the age composition of catches (281 fish aged successfully, sexes combined) made during June-December. Because Puget Sound rockfish are not the target of any fishery, the catch curve (Ricker, 1975) was used to estimate natural mortality (M) with linear regression of log (frequency) on age. Only fish 3 years and older were used to esti- mate M, because it was apparent (Fig. 1) that 1- and 2-year-olds were not fully available to our sampling gear. Fork length (FL) to the nearest millimeter and wet gonad weight (GW) and wet somatic weight (SW) to the nearest milligram were measured for all adult females. Somatic weight was defined as the weight of the fish with the stomach emptied and the gonads removed. The gonadosomatic index ( GSI = GW/SW ) was used to determine relative reproductive effort for a mature female of average size. Allometric rela- tionships between GW-body length and SW-body length were determined by using log-log linear re- gression (Gunderson and Dygert, 1988). Only fully mature females in the later stages of vitellogenesis were used in estimating GSI. Histological examina- tion of a subsample of 12 individuals, representing a broad range of lengths and relative GSI values, showed that all of them were in the migratory nucleus stage of egg development just prior to ovulation (McDermott1 ). Fecundity was estimated by counting all nonatretic oocytes in advanced vitellogenesis with subsamples from both ovaries of 16 fish ranging from 112 to 178 mm FL. From preliminary measurements it could be seen that vitellogenetic oocytes were at least twice the size of the resting and immature primary oocytes. All oocytes with the longest axis exceeding 250 pm were counted. A gravimetric subsampling method (Nichol and Pikitch, 1994) was chosen to estimate oocyte numbers. Because of the nonuniform distribution of oocytes and connective tissue within the ovaries, ovarian sacs and ovarian stroma were removed prior to subsamp- ling. Thin sections of the remaining egg mass were taken from five different transects equally spaced along the longitudinal axis of each ovary. Each subsample was weighed, pipetted into a plastic grid dish, and the eggs of each subsample were counted. Final fecundity estimates were made by multiplying 1 McDermott, S. 1995. School of Fisheries, Univ. Washington, Seattle, WA 98195. Personal commun. 354 Fishery Bulletin 96(2), 1998 the mean oocytes/mg, averaged across all sub- samples, by total egg mass weight. Fecundity data were fit to an allometric relationship with log length- log fecundity linear regression (Haldorson and Love, 1991). Results Females with unfertilized eggs were caught from 26 June to 8 August 1994. Embryos without pigmented eyes were observed from 22 July to 29 August, and embryos with pigmented eyes from 24 July to 29 August. The first spent female was observed on 21 August, and all mature females caught from 31 Au- gust to 10 September showed spent ovaries. Only resting females were captured on 30 December. Ma- ture females captured on 12 March 1995, were either resting or in the initial stages of vitellogen- esis. Spawning occurred over a relatively short time period, with only 9 days between the time when the first spent female and the last gravid female were ob- served. Ovaries did not contain developing embryos and cleaving eggs at the same time, indicating that individual Puget Sound rock- fish spawn only one brood per season. The oldest observed fish was a female 13 years old, followed by a 12-year-old male. The largest male caught was 145 mm FL and the largest female 178 mm FL (Fig. 2). The age- length relation (von Bertalanffy growth model) for male Puget Sound rockfish was Lt = 137.39 (1 - e-°-7042<'+0 3232>) and for fe- males Lf = 170.72 (1 - e-0.5353 tt+0.4603))_ Based on the regression result of ln(fre- quency) = 5.554-0.44 (age), the instanta- neous annual rate of natural mortality (M) for fish 3 years and older was estimated to be 0.44 (Fig. 1). Female Puget Sound rockfish reached sexual maturity at a median length of L0 5 = 121.71 mm FL (Fig. 3) corresponding to a predicted age of 1.87 years with the von Bertalanffy growth model. The length- maturity parameters derived from the lo- gistic model were a - -39.678 and = 0.326. The average length of all mature females larger than L0 5 was LAV = 150.93 mm. The somatic weight-length relationship was SW = 5.882 x 10-5 L2 687 with the pre- dicted somatic weight for LAV = 42.06 g. The gonad weight-length relationship was GW = 5.869 x 10“7 L3151 with the predicted gonad weight for LAV = 4.30 g. From these two weights the GSI for Puget Sound rock- fish was computed as 0.10. The lowest fecundity was 3,295 eggs for a female of 112 mm FL and the highest was 57,787 for an individual 156 mm FL (Fig. 4). The length-fecundity (F) relation- ship was F = 0.052 L2,679, and the fecun- dity for a mature female of average length (150.93 mm) was estimated to be 35,723 eggs. The estimated fecundity per gram for a female L0 5 cm long (Haldorson and Love, 1991) was 779 eggs. Age (yr) Figure 2 Comparison of female and male age-length (FL) relationships for Puget Sound rockfish. Solid line and filled circles show predicted relationship (von Bertalanffy growth model) and data for females; dashed line and open triangles show predicted relationship and data for males. NOTE Beckmann et at: Reproductive biology, growth, and natural mortality of Sebastes emphaeus 355 Although egg size was variable, the ma- jority of the eggs within an individual fe- male were about the same size and at the same developmental stage. The time span between the capture of the first fertilized female and the first spent female yields a rough estimate of 30 days for the gesta- tion period. Discussion Puget Sound rockfish show high fecundity per gram, early age at maturation, small body size, and a relatively short life ex- pectancy compared with other rockfish of the genus. No other rockfish studied to date has a higher fecundity per gram at L05 (Haldorson and Love, 1991). A high natural mortality would be expected to ac- company these life history characteristics, and the estimated M of 0.44 is high compared with that of other rockfish . Puget Sound rockfish have a GSI (0.10) close to that of other early maturing rock- fish (e.g. shortbelly rockfish, GSI=0.07, M=0.26, maturation age=2 years; Gunderson, 1997). However, a comparison of the catch- curve estimate of M with empirically derived estimates suggests some possibility that the catch curve is based on an unrepresentative sample and has been overestimated. The maximum age observed in our study (13 years) corresponds to an M of 0.32 with Hoenig’s (1983) empirical estimation tech- nique, whereas the GSI (0.10) gives an esti- mate of M = 0.18 with Gunderson’s (1997) estimator. A constant recruitment is assumed if the catch curve is used, yet the data in Figure 2 suggest that the 1988 year class (age 6 in 1994) appears to have recruited at significantly lower levels than the others. Our catch-curve estimate of M should probably be regarded as provisional, and additional collections of age-composition data or tag- ging studies should be undertaken to validate it. The fecundity of 3,300-57,800 eggs observed for Puget Sound rockfish is in the range of fecundities of similar-size rockfish. Shortbelly rockfish ( 18-31 cm TL) has a fecundity from 7,000 to 50,000 (Hart 1973). Our fecundity estimates for Puget Sound rockfish are comparable to those obtained by Moulton (1975). The estimated gestation period for Puget Sound rockfish is similar to the 29-day gestation period re- ported for yellowtail rockfish (S. flavidus), a species which has been shown to be lecithotrophic, provid- ing only a negligible amount of maternal energy dur- ing gestation (Hopkins et al. 1995). The high fecun- dity per gram and relatively brief gestation period suggest that parental care is limited in comparison to other rockfish. Acknowledgments This project was supported by stipends provided to the senior author by University of Washington’s Fri- day Harbor Laboratories. The paper is based in part on a diploma thesis submitted to Bochum Univer- 356 Fishery Bulletin 96(2), 1 998 sity, Zoology Department, Embryology workgroup. The authors wish to thank K.-D. Hofmann for coop- eration and support. We also express our gratitude to J. Moreland, K. Lawrence, S. Maclean, M. Woodburry and others who served as divers and boat tenders. Literature cited Beckmann, A. T. 1995. Recruitment ecology and reproductive biology of the Puget Sound rockfish, Sebastes emphaeus (Starks, 1911). Diploma thesis, Bochum Univ., Bochum, Germany, 99 p. Echeverria, T. W., and W. H. Lenarz. 1984. Conversions between total, fork and standard length in 35 species of Sebastes from California. Fish. Bull. 82:249-251. Finckh McDermott, S. 1994. Reproductive biology of rougheye and shortraker rockfish, Sebastes aleutianus and Sebastes borealis. M.S. thesis, Univ. Washington, Seattle, WA, 76 p. Gunderson, D. R. 1997. The trade-off between reproductive effort and adult survival in oviparous and viviparous fishes. Can. J. Fish. Aquat. Sci. 54:990-998. Gunderson, D. R., and P. H. Dygert. 1988. Reproductive effort as a predictor of natural mortal- ity rate. J. Cons. Int. Explor. Mer 44:200-209. Haldorson, L., and M. Love. 1991. Maturity and fecundity in the rockfishes Sebastes spp., a review. Mar. Fish. Rev. 53:25-31. Hart, J. L. 1973. Pacific fishes of Canada. Bull. Fish. Res. Board Can. 180, 740 p. Hoenig, J. M. 1983. Empirical use of longevity data to estimate mortal- ity rates. Fish. Bull. 82:898-903. Hopkins, T. E., M. B. Eldridge, and J. J. Cech Jr. 1995. Metabolic cost of viviparity in yellowtail rockfish, Sebastes flavidus. Env. Biol. Fish. 43:77-84. Moulton, L. L. 1975. Life history observations on the Puget Sound rock- fish, Sebastes emphaeus (Starks, 1911). J. Fish. Res. Board Can. 32:1439-1442. Nichol, D. G., and E. K. Pikitch. 1994. Reproduction of darkblotched rockfish off the Oregon Coast. Trans. Am. Fish. Soc. 123:469-481. Ricker, W. E. 1975. Computation and interpretation of biological statis- tics of fish populations. Bull. Fish. Res. Board Can. 190, 382 p. Westrheim, S. J. 1975. Reproduction, maturation, and identification of larvae of some Sebastes (Scorpaenidae) species in the northeast Pa- cific Ocean. J. Fish. Res. Board Can. 32:2399-2411. Wourms, J. P. 1991. Reproduction and development of Sebastes in the context of the evolution of piscine viviparity. Env. Biol. Fish. 30:111-126. 357 Age, growth, and calving season of bottlenose dolphins, Tursiops truncatus, off coastal Texas Stephanie Fernandez Institute Tecnologico y de Estudios Superiores de Monterrey Campus Guaymas, Apdo. Postal 484, Guaymas, Sonora 85400, Mexico Aleta A. Hohn Office of Protected Resources National Marine Fisheries Service, NOAA 1335 East-West Highway, Silver Spring, Maryland 20910 Present address: National Marine Fisheries Service 1 0 1 Pivers Island Road Beaufort, North Carolina 28516 E-mail address (for A. A. Hohn, contact author), aleta.hohn@noaa.gov Life history studies of bottlenose dolphins, Tursiops truncatus , in North America began over a cen- tury ago with observations by True (1890) on dolphins captured in a directed fishery along the coast of Cape Hatteras, North Carolina. He detected an approximately equal sex ratio but variability in size among individuals in almost every catch. He noted tooth development (eruption ) and state of fissure of the umbilical cord in young dolphins, as well as reproductive state in fe- males (i.e. lactating or pregnant). Furthermore, he suggested that calving took place primarily in the spring. Since then, much of the knowledge of life history of bottle- nose dolphins in the western Atlan- tic has been obtained from dolphins captured incidentally during fish- ing operations or directly for research or display, and trom stranded ani- mals (Sergeant et al., 1973; Hohn, 1980; Hersh, 1987; Barros and Odell, 1990; Mead and Potter, 1990). More recent and extensive information on age, growth, and population biology has been ob- tained from a long-term study of free-ranging bottlenose dolphins in Sarasota Bay, Florida (e.g. Hohn et al., 1989; Scott et al., 1990; Wells and Scott, 1990; Read et al., 1993). Equivalent life history studies have not been conducted on bottle- nose dolphins from the north or western Gulf of Mexico, although this is the most common cetacean inhabiting the coastal waters of the northwestern Gulf (Barham et al., 1980; Fritts et al., 1983; Leather- wood and Reeves, 1983; Mullin et al., 1990). It also is the most com- monly stranded species along the Gulf coast, particularly along the coast of Texas. Strandings of bottle- nose dolphins have been recorded routinely in Texas since 1974 ( Jones, 1987; Schmidly and Shane1). Since late 1980, the Texas Marine Mam- mal Stranding Network (TMMSN) has recovered over 800 carcasses. In the spring of 1990 and 1992, the TMMSN documented the highest frequency of bottlenose dolphin strandings since the program be- gan (Hansen2). Unusually high mortalities among bottlenose dol- phins along the Texas coast have emphasized the need for baseline information on their life history to allow better interpretation of the impacts of these mortalities on the populations. The available source of data and samples to begin such studies is the strandings them- selves. This study of stranded speci- mens was conducted to estimate 1 ) age structure of stranded speci- mens, 2) growth from length-at-age data, 3) mean length at birth, 4) calving season, and, where pos- sible, 5) age and length at sexual and physical maturation. These data provided the material for a preliminary examination of the life history of bottlenose dolphins that inhabit the coastal waters of Texas. Materials and methods From January 1981 through De- cember 1990, 898 stranded bottle- nose dolphins (373 males, 292 fe- males, and 233 specimens of un- known sex) were recovered by mem- bers of TMMSN along the 715-km Texas coastline. Tooth sections were prepared for 205 specimens (Table 1), which comprised the entire sample of specimens from which teeth were collected. The teeth were fixed in 10% buffered formalin and pro- cessed following the decalcification, sectioning, and staining protocol described by Hohn et al. ( 1989 ). Age was estimated by counting the number of GLG’s (growth layer groups, Perrin and Myrick, 1980) in dentine and cement without ref- erence to data on the sex and length 1 Schmidly and Shane. 1978. A biologi- cal assessment of the cetacean fauna of the Texas coast. Rep. U.S. Mar. Mamm. Comm., contract report MMC-74/05, 38 p. 2 Hansen, L. J. 1992. Stranding rate and trends. In L. J. Hansen (ed.), Report on investigation of 1990 Gulf of Mexico bottle- nose dolphin strandings, p. 15-20. Na- tional Marine Fisheries Service, South- east Fisheries Center, Contribution MIA- 92-93-21. Manuscript accepted 26 June 1997. Fishery Bulletin 96:357-365 ( 1998). 358 Fishery Bulletin 96(2), 1 998 labile 1 Sample sizes and sex ratios of bottlenose dolphins, Thrsiops truncatus, stranded along the Texas coast from January 1981 through December 1990 and of the subsample for which age was estimated. Sample Males Females Sex unknown Sex ratio (M:F) Total 373 292 233 1.3 : 1 Aged 78 81 36 1 : 1.04 <1.0 yr old 22 10 7 2.2 : 1 1-20 yr old 46 39 22 1.2 : 1 >20 yr old 10 32 7 1 : 3.2 of the animals. The age structure of males and fe- males was compared by using the nonparametric Kolmogorov-Smirnov (K-S) test (Sokal and Rohlf, 1981). Growth was determined by fitting the nonlinear, least-squares Gompertz model to length-at-age data (SAS Institute, 1985): Reproductive data were available for 57 females from the total sample of stranded females from which some data were collected (n= 292). Teeth were avail- able for 25 of these. Sexual maturity of females was determined by noting 1) the presence or absence of a corpus of ovulation from external examination of ovaries or 2) the presence of a fetus or an extended uterus, indicating a pregnancy. Sexual maturity of males could not be evaluated because few data or samples from testes were collected. Vertebrae from 24 males, 25 females, and 12 dol- phins of unknown sex were collected. All were exam- ined for state of physical maturity. From this sample, teeth were available for only 11 males, 12 females, and 8 specimens of unknown sex. Physical maturity was determined by noting the degree of fusion of the epiphysis to the centrum of thoracic and lumbar ver- tebrae. Three categories were used to classify the specimens: a) not fused (immature), b) fusing (ma- turing), and c) fused (mature). Results Sit) = A(exp(-6exp(-£))), (1) Age and sex composition where S A b k t a measure of size (cm); asymptotic length; the constant of integration; the rate of growth constant; and age (yr) (Laird, 1966). Total body length was measured to the nearest cen- timeter in a straight line from the tip of the upper jaw to the notch of the fluke (Norris, 1961). Predicted asymptotic lengths for males and females were com- pared by using the approximate Gtest (Sokal and Rohlf, 1981). These lengths were then compared with the corresponding asymptotic lengths for male and female bottlenose dolphins from Sarasota Bay, Florida (Read et al., 1993). Length at birth was estimated in three ways: 1) as the mean length of 21 specimens from the entire data set (n= 898) determined to be neonates on the basis of the presence of remains of the umbilical cord or folded dorsal or caudal fins; 2) as the mean length of 21 specimens (no overlap with the previous sample of 21) estimated to be <0.1 year old because the neo- natal line in tooth sections had not yet formed or had just started to form (Hohn and Hammond, 1985); 3) as the predicted value from the Gompertz model. Calving season was also identified in three ways: as the date of stranding of 1) neonates, 2) specimens <0.1 year old, and 3) specimens measuring 90-120 cm in length. Estimates of age were obtained for 78 males, 81 fe- males, and 36 specimens of unknown sex (Table 1; Fig. 1). For an additional 10 dolphins (5 males, 3 fe- males, and 2 dolphins of unknown sex), only mini- mum ages, ranging between 12+ to 20+ years, were obtained because poor-quality tooth sections pre- cluded accurate age estimates. These 10 specimens were excluded from all analyses. Occlusion of the pulp cavity was not seen in any of the tooth sections. The oldest male and the oldest specimen of un- known sex were 33 years old and the oldest female was 41 years old (Fig. 2). Age distributions for males and females were significantly different (K-S test, P<0.01). Specimens less than 1 year old accounted for a relatively large part (20%) of the aged sample and were skewed towards males, but with increas- ing age, the proportion of females in the sample in- creased significantly (chi-square (%2) P- 0.003, for data stratified by age group 0-0.9, 1.0-9. 9, 10.0-19.9, and >20 years; Table 1). Because of the relatively large number of males less than one year of age, the K-S test was rerun with that age class excluded for both males and females. The difference was still sig- nificant (K-S test, P<0.01). Growth The Gompertz model gave predicted asymptotic lengths of 263.5 cm for males and 244.7 cm for fe- NOTE Fernandez and Hohn: Age, growth, and calving season of Tursiops truncatus 359 males. Trends in the residuals, however, showed that the fit was poor for a number of age classes (Fig. 3). To obtain a better predictor, the model was fitted to a data set that excluded all specimens less than 1 year of age, analogous to the procedure in Read et al. (1993) where the Gompertz model was found to describe well the growth of bottlenose dolphins from Sarasota Bay, Florida. With this subset of data, as- ymptotic length was slightly higher for males (268.0 cm) and for females (246.7 cm) but neither were sig- nificantly different from the predicted values from the total sample (approximate Gtest, P> 0.05, Table 2). Trends in the residuals showed that predicted length at age was a better fit for ages 1-9 years and was slightly overestimated at asymptote. In an attempt to find a fit with evenly dispersed residuals, the Gompertz model was modified to fix length at birth at 109.5 cm, as calculated below by independent means, rather than by solving for the asymptotic value, A: S(t)= 109.5^exp(6(l- 6exp(-£)))j. (2) Predicted asymptotic length was 261.7 cm for males and 244.4 cm for females. Predicted length from birth through age 2 years was very similar to that esti- mated from the total sample with the Equation 1 ver- sion of the Gompertz model. Length at age was still underestimated and this derivation of the model did not improve the fit. Although none of the iterations produced a fit that represented all age classes well, the best overall pre- dicted values occurred with Equation 1) when speci- mens <1 year of age were excluded (Fig. 3) and, for comparative purposes, predicted lengths were taken from this model (268 cm for males and 246.7 cm for females). Males were significantly larger than fe- males at asymptotic length (approximate Gtest, P<0.05). Asymptotic lengths of males and females were not significantly different (approximate Gtest, P>0.05) between dolphins from Texas and Sarasota Bay, Florida (Table 2). Length at birth and calving seasonality There was no difference in length at birth between male and female neonates or between males and fe- males <0.1 year old U-test, P>0.05). Therefore, data were combined to produce a single estimate with each method (Table 3). Mean length at birth from com- bined data was not significantly different between these two methods (Gtest, P>0.05), therefore the to- tal data set was combined (males, females, and sex the Gompertz curve was greater than length at birth unknown, n=42) to give an estimated length at birth estimated from the more direct measures. In addi- of 109.4 cm (SD =8. 5). Length at age 0 estimated from tion, because the best-fit version of the Gompertz Figure 1 Midlongitudinal stained thin sections from the tooth of a 29- year-old, 261-cm female bottlenose dolphin, Tursiops trun- catus, stranded in Texas. Growth layer group (GLG) bound- aries were considered to be the relatively thin dark layers; most of these are easily visible. Boundaries between the first few layers are marked with a small black line on the upper left side of the section. The last few, very fine GLG’s near the pulp cavity are difficult to see at this low magnification. 360 Fishery Bulletin 96(2), 1998 Males Females Sex unknown Age class (yr) Figure 2 Age distribution of bottlenose dolphins, Thrsiops truncatus, stranded along the coast of Texas from 1981 to 1990. Table 2 Parameter values and their asymptotic standard errors ( SE ) from the Gompertz growth model fit to two sets of length-at-age data from bottlenose dolphins, Tursiops truncatus, stranded along the coast of Texas. The first set included all the available data; the second set was a subset of the first set that excluded specimens <1.0 yr of age. A = asymptotic length, b = the constant of integration, and k = the rate of growth constant, n = number of dolphins in each sample. n A± SE b ± SE K± SE Female length All data 78 244.7 ± 1.90 0.755 ± 0.0423 0.497 ± 0.0490 Specimens of age <1.0 yr excluded 68 246.7 ± 2.22 0.482 ± 0.0873 0.294 + 0.0655 Male length All data 75 263.5 ± 2.93 0.785 ± 0.0308 0.351 ± 0.0313 Specimens of age <1.0 yr excluded 54 268.0 ± 3.64 0.592 ± 0.0679 0.2341 0.0388 model excluded specimens less than 1-yr-old extrapo- lation to age 0 to estimate length at birth was inap- propriate. The ratio of mean length at birth to as- ymptotic length was 40.8% for males and 44.4% for females. The mode in strandings of neonates (n=12 of 21) and specimens <0.1 year of age (n=ll of21) occurred in March (Fig. 4). In both data sets, 86% of the speci- mens stranded during March and April. Except for five specimens found in November and December, all neonates stranded from February through June. Sexual and physical maturation Because reproductive data were collected from only 57 females (Table 4), the sample size was too small to estimate the mean age and length at sexual matu- ration. Only seven pregnant females were noted; the smallest was 233 cm and the longest was 335 cm. The latter, an extremely large female, also had the largest fetus (122-cm male) recorded in the TMMSN stranding database. Teeth were available for three of the seven pregnant females, and they were esti- NOTE Fernandez and Hohn: Age, growth, and calving season of Tursiops truncatus 361 mated to be 9.8, 23, and 27 years of age. A 4.8-year-old pregnant female, 245 cm long, was excluded from this sample owing to possible mislabeling of samples. No reproductive data were avail- able for males. Mean age and length at physical maturation could not be calculated owing to the small sample sizes (Table 4). Two specimens, a 280-cm male and a 253-cm female, were notable because they were relatively large yet phy- sically immature. Discussion Despite the widespread dis- tribution and high abun- dance of bottlenose dolphins along the Atlantic and Gulf of Mexico coasts of the United States, the majority of bio- logical information for this species has come from a few locations, a relatively small number of studies, and small sample sizes. Basic questions still remain throughout much of the range. The current data provided the opportu- nity to begin to evaluate simi- larities and differences in bottlenose dolphin life his- tory in relation to areas where previous studies had been conducted. Results from this study support the hypothesis that in bottlenose dolphins sexual dimorphism in body length is not exhibited at birth (Ser- geant et al., 1973; Hohn, 1980; Kasuya et al., 1986; Cockcroft and Ross, 1990) but does occur in adults (Read et al., 1993). Although a number of studies have concluded that sexual dimorphism in adult size does not exist in bottlenose dolphins from the Atlantic and Gulf of Mexico coasts of the United States (for ex- ample, Hohn, 1980; Hersh, 1987; Hersh et al., 1990; Mead and Potter, 1990), the contradictory results could be explained partly because sample sizes in □ m ■ □i D 1 □ D |s j—- — V u ‘ □ Females Mature ■ Immature ♦ Unknown □ j i i i i i_ _i I i i i i I i i i i I j 20 25 30 35 40 45 £ a> c Age (yr) Figure 3 Scatterplots of length at age for female and male bottlenose dolphins, Tursiops truncatus , stranded along the coast of Texas. The solid lines represent the predicted growth tra- jectory from the Gompertz model when specimens less than 1 yr old are excluded; growth between birth and age 1 was determined by using the independently estimated length at birth and the predicted length at age 1. The dashed lines represent the predicted growth trajectory from the Gompertz model with the entire data set. Although the asymptotic lengths are similar, the growth trajectories are different, particularly for females. earlier studies were small but also because these other studies did not fit a growth curve to obtain pre- dicted asymptotic lengths that could be directly com- pared (Read et al., 1993). The dimorphism at asymp- totic length may result from females diverting en- ergy to reproduction at the same ages that males are gaining girth and mass (Read et al., 1993) and thus would explain why sexual dimorphism occurs in adults, but not at birth. 362 Fishery Bulletin 96(2), 1 998 100 90 ■1 80 ||||||| 90-120 cm 70 ||» M , 60 O c a) g- 50 Q) H Neonates 111 ^li LL 40 30 ■i iii 20 - !■ In 10 - 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month Fsgure 4 Calving seasonality illustrated as the monthly distribution of strandings of neonates, specimens <0.1 years old, and specimens 90-120 cm in length. Table 3 Estimated length at birth (cm) for bottlenose dolphins, Tiirsiops truncatus, stranded along the coast of Texas. The two methods of defining a newborn animal (“neonates” and “specimens <0.1 yr”) are described in the text. The Gompertz values were the predicted lengths at age 0 (L0) from the Gompertz curve fitted to age-at-length data for males and females. The samples for neonates and speci- mens <0.1 yr are independent, n = number of dolphins in each sample. Sex Males Females unknown Total Neonates mean 109.5 111.2 SD 9.3 13.6 n 12 5 range 94-124 91-128 Specimens < 0.1 yr mean 109.7 107.4 SD 7.7 9.7 n 13 5 range 93-121 97-122 Gompertz mean 128.2 115.1 SD 3.6 4.9 107.3 109.5 5.0 9.3 4 21 102-112 91-128 113.7 109.7 3.5 7.5 3 21 110-117 93-122 n range Bottlenose dolphins from the coast of Texas are not significantly different in adult size from those in Sarasota Bay, Florida (Read et al., 1993), the only other western Atlantic or Gulf of Mexico area for which growth curves have been fitted to length-at- age data. Contrary to the analysis of Read et al. (1993), however, we found that the Gompertz curve did not adequately describe growth across all age classes. The curve did not respond readily enough to simulate rapid growth during the first couple of years. The fit was improved by excluding specimens less than 1 year of age, the phase of growth where the rate and absolute incremental increase are greatest. Although this approach provided a data set comparable in range of ages to that used in Read et al. (1993), the Gompertz model proved a better predictor for the Sarasota Bay, Florida, sample where the predicted length of young animals was larger. The large and un- biased sample of live animals that represented most of the individuals in Sarasota Bay probably accounts for these differences. The fit might be further improved if the sample was large enough to allow for a two-stage Gompertz fit (e.g. see Perrin et al., 1976) that would accommodate the growth spurt occurring at matu- ration (Cheal and Gales, 1992; Read et al., 1993). Estimates of length at birth are not available for the Sarasota Bay animals. Along the mid-Atlantic NOTE Fernandez and Hohn: Age, growth, and calving season of Tursiops truncatus 363 coast of the United States, however, Mead and Pot- ter (1990) estimated length at birth (117 cm) as the mean length of 13 specimens classified as neonates on the basis of a folded or flaccid dorsal fin. Using the equivalent data set from the Texas sample (neo- nates), we found that the length at birth of mid- Atlan- tic coastal bottlenose dolphins was significantly differ- ent (7-test, P-0.02) from the estimate of 109.4 cm for Texas animals, but this result should be considered preliminary until larger sample sizes are available. Mean length at birth may best be estimated by using morphological characteristics. Tooth-layer deposition patterns proved to be accurate in compari- son to results from morphological data and can be valuable when morphological characteristics are not recorded. The method, however, is more time con- suming. In contrast, use of predicted lengths of 0-yr- old animals from a fitted growth curve may be inac- curate, as was the case with this sample. Strandings of neonates along the Texas coast are highly seasonal, with a mode in March. Using strand- ing data, Urian et al. (1996) found a significant dif- ference in calving season between bottlenose dolphins from the coast of Texas (citing this data set) and those from the central-west coast of Florida, including Sarasota Bay. Stranding patterns may not accurately reflect actual calving, however, because neonates that did not survive may be those that were born earlier or later than calves that did survive. In the Sarasota Bay area, Urian et al. (1996) found that the mean date of birth estimated from stranded neonates was 16 days earlier than that estimated from sighting records of females with new calves, although this difference was not significant. They suggested that the photo-identification data may also be biased ow- ing to unequal and discontinuous distribution of sur- vey effort, a finding also made by Caughley and Caughley (1974). Further error may arise from lack of detection of neonates that stay very close to the female. Photo-identification studies of calving sea- son along the Texas coast identified a peak in num- ber of neonates in May (Shane, 1977). This later de- tection of neonates, in comparison with stranded animals, is consistent with the results from the cen- tral-west coast of Florida and limitations of the photo- identification method. The increase in proportion of females with age in the Texas sample is similar to a general pattern iden- tified in delphinids, with a slight bias towards young males and old females (Ralls et al., 1980; Perrin and Reilly, 1984; Wells and Scott, 1990). Some authors have attributed this shift to higher natural mortal- ity rates in males (Kasuya, 1976; Miyazaki, 1977; Wells and Scott, 1990). No direct or unbiased esti- mates of mortality are available for bottlenose dol- TabUe 4 Sample sizes and descriptions of sexual and physical matu- rity in bottlenose dolphins, Tursiops truncatus, stranded along the coast of Texas. Reproductive data were not collected for males. Sex Sample Females Males unknown Total stranded with at least minimum data collected 292 373 233 With reproductive data 57 Number immature 17 Number mature 40 Largest immature (cm) 235 Shortest mature (cm) 225 With reproductive and age data 25 Number immature 6 Number mature 19 Oldest immature (yr) 7.8 Youngest mature (yr) 9.8 Youngest pregnant (yr) 9.8 With vertebrae examined for physical maturity 25 25 12 Number immature 10 16 8 Number maturing 2 2 0 Number mature 13 7 4 Largest immature (cm) 253 280 253 Shortest mature (cm) 219 237 231 With physical maturation and age data 12 11 8 Number immature 2 7 4 Number maturing 1 1 0 Number mature 9 4 4 Oldest immature (yr) 15 16 9.9 Youngest mature (yr) 13 18 20 phins from Texas, and patterns from stranded ani- mals may not accurately reflect the population. The proportion of specimens less than 1 year of age in the sample of stranded animals is higher than expected. In bottlenose dolphins from Sarasota Bay, Florida, Wells and Scott (1990) calculated that ani- mals in the same age class represented 3.4% of the population, and that the crude birth rate was 5.5% annually (range of 1.1-10.4% over eight years). The higher percentage ( 20%) of calves in the Texas sample most likely is due to the high rate of mortality in neonates and young calves. It is also interesting to speculate that cows with young calves may occupy more protected waters and, therefore, are more likely to wash ashore when they die, contributing to a strand- ing sample not representative of the population. Some of the specimens in the sample were notably large, e.g. a 300-cm male, a 335-cm female, a 280-cm 364 Fishery Bulletin 96(2), 1998 physically immature male, and a 253-cm physically immature female. Although the sample is assumed to consist predominantly of the coastal form of bottle- nose dolphins, it is possible that specimens from the larger offshore form (Duffield et al., 1983; Hersh and Duffield, 1990; Hersh et al., 1990) were included. These specimens remained in the sample because they were not specifically identified as offshore indi- viduals from independent criteria. The inclusion of offshore individuals of various lengths may be re- sponsible for some of the variability in length at age seen in this study. Results from this study show that bottlenose dol- phins from the coast of Texas have some similarities and some differences with bottlenose dolphins from the west coast of Florida and the mid-Atlantic coast of the United States. However, samples size are small, precluding, for example, any substantial re- productive analyses. We strongly encourage the rou- tine collection of reproductive and other samples from stranded specimens. Such basic information as the average age at sexual maturation and calving inter- val remains unknown even though large numbers of bottlenose dolphins are found stranded each year. Acknowledgments This research was funded by National Marine Fish- eries Service, Southeast Fisheries Science Center, Miami Laboratory (SEFSC) and was conducted pri- marily at the SEFSC, Galveston Laboratory. We thank Raymond Tarpley (Founder and first Presi- dent) and Graham Worthy (current President) of the Texas Marine Mammal Stranding Network (TMMSN) for providing access to samples and data. We are par- ticularly grateful to Elsa Haubold for facilitating access to database records and assisting with clarifi- cation of questions about the data. Efforts by regional coordinators and volunteers from Texas A&M Uni- versity at Galveston who collected data on these specimens are greatly appreciated. Literature cited Barham, E. G., J. C. Sweeney, S. Leatherwood, R. K. Beggs, and C. L. Barham. 1980. Aerial census of the bottlenose dolphin, Tursiops truncatus , in a region of the Texas coast. Fish. Bull. 77(31:585-595. Barros, N. B., and D. K. Odell. 1990. Food habits of bottlenose dolphins in the southeast- ern United States. In S. Leatherwood and R. R. Reeves (eds.), The bottlenose dolphin, p. 309-328. Academic Press, New York, NY. Caughley, G. and J. Caughley. 1974. Estimating median date at birth. J. Wildl. Manage. 38:552-556. Cheal, A. J., and N. J. Gales. 1992. Growth, sexual maturity and food intake of Austra- lian Indian Ocean bottlenose dolphins, Tkrsiops truncatus , in captivity. Aust. J. Zool. 40:215-223. Cockcroft, V. G., and G. J. B. Ross. 1990. Age, growth, and reproduction of bottlenose dolphins Tursiops truncatus from the east coast of southern Africa. Fish. Bull. 88:289-302. Duffield, D. A., S. M. Ridgway, ad L. H. Cornell. 1983. Hematology distinguishes coastal and offshore forms of dolphins (Tursiops). Can. J. Zool. 61:930-933. Fritts, T. H., A. B. Irvine, R. D. Jennings, L. A. Collum, W. Hoffman, and M. A. McGehee. 1983. Turtles, birds, and mammals in the northern Gulf of Mexico and nearby Atlantic waters. U.S. Fish and Wild- life Service, Office of Biological Services, FWS/OBS/82/65, Washington, D.C., 455 p. Hersh, S. L. 1987. Characterization and differentiation of bottlenose dolphin populations (genus Ttirsiops) in the southeastern U.S. based on mortality patterns and morphometries. Ph.D. diss., Univ. Miami, Miami, FL, 213 p. Hersh, S. L., and D. A. Duffield. 1990. Distinction between northwest Atlantic offshore and coastal bottlenose dolphins based on hemoglobin profile and morphometry. In S. Leatherwood and R. R. Reeves (eds.), The bottlenose dolphin, p. 129-139. Academic Press, New York, NY. Hersh, S. L., D. K. Odell, and E. Asper. 1990. Sexual dimorphism in bottlenose dolphins from the east coast of Florida. Mar. Mammal Sci. 6:305-315. Hohn, A. A. 1980. Age determination and age related factors in the teeth of Western North Atlantic bottlenose dolphins. Sci. Rep. Whales Res. Inst. 32:39-66. Hohn, A. A., and P. S. Hammond. 1985. Early postnatal growth of the spotted dolphin, Sten- ella attenuata, in the offshore eastern Tropical Pacific. Fish. Bull. 83(4):553-566. Hohn, A. A., M. D. Scott, R. S. Wells, J. C. Sweeney, and A. B. Irvine. 1989. Growth layers in teeth from known-age, free-rang- ing bottlenose dolphins. Mar. Mammal Sci. 5(41:315-342. Jones, S. C. 1987. Patterns of recent marine mammal strandings along the upper Texas coast. Cetus 7(2): 10-14. Kasuya, T. 1976. Reconsideration of life history parameters of the spot- ted and striped dolphins based on cemental layers. Sci. Rep. Whales Res. Inst. 28:73-106. Kasuya, T., T. Tobayama, T. Saiga, and T. Kataoka. 1986. Perinatal growth of delphinoids: information from aquarium reared bottlenose dolphins and finless porpoises. Sci. Rep. Whales Res. Inst. 36:85-97. Laird, A. K. 1966. Post-natal growth of birds and mammals. Growth 30:349-363. Leatherwood, S., and R. S. Reeves. 1983. Abundance of bottlenose dolphins in Corpus Christi Bay and coastal southern Texas. Contrib. Mar. Sci. 26:179-199. Mead, J. G., and C. W. Potter. 1990. Natural history of bottlenose dolphins along the cen- tral Atlantic coast of the United States. In S. Leather- NOTE Fernandez and Hohn: Age, growth, and calving season of Tursiops truncatus 365 wood and R. R. Reeves (eds.), The bottlenose dolphin, p. 65-195, Academic Press, New York, NY. Miyazaki, N. 1977. Growth and reproduction in the striped dolphin Stenella coeruleoalba off the coast of Japan. Sci. Rep. Whales Res. Inst. 2:21-48. Mullin, K., R. R. Lohoefener, W. Hoggard, C. L. Roden, and C. Rogers. 1990. Abundance of bottlenose dolphins, Tursiops trun- catus, in the coastal Gulf of Mexico. Northeast Gulf Sci. 11(21:113-122. Norris, K. S. (ed.). 1961. Standardized methods for measuring and recording data on the smaller cetaceans. J. Mammal. 42:471-476. Perrin, W. F., J. M. Coe, and J. R. Zweifel. 1976. Growth and reproduction of the spotted porpoise, Stenella attenuata, in the offshore eastern tropical Pacific. Fish. Bull. 74:229-269. Perrin, W. F, and A. C. Myrick Jr. (eds.). 1980. Report of the workshop. Rep. Int. Whal. Comm., spec, issue 3:1-50. Perrin, W. F., and S. B. Reilly. 1984. Reproductive parameters of dolphins and small whales of the Family Delphinidae. In W. F. Perrin, R. L. Brownell, and D. DeMaster (eds.), Reproduction of whales, dolphins and porpoises. Rep. Int. Whaling Comm., spec, issue 6, p.7-134. Ralls, K., R. L. Brownell Jr., and J. Ballou. 1980. Differential mortality by sex and age in mammals, with specific reference to the sperm whale. In Sperm whales: special issue, p. 233-243. Rep. Int. Whaling Comm., spec, issue 2. Read, A. J., R. S. Wells, A. A. Hohn, and M. D. Scott. 1993. Patterns of growth in wild bottlenose dolphins, Tursiops truncatus. J. Zool. Soc., Lond. 231:107-123. SAS Institute. 1985. SAS/STAT user’s guide, version 6. SAS Institute, Cary, NC, 1848 p. Scott, M. D., R. S. Wells, and A. B. Irvine. 1990. A long-term study of bottlenose dolphins on the west coast of Florida. In S. Leatherwood, and R. R. Reeves (eds.), The bottlenose dolphin, p. 235-244. Academic Press, New York, NY. Sergeant, D., D. K. Caldwell, and M. C. Caldwell. 1973. Age, growth, and maturity of bottlenosed dolphin ( Tursiops truncatus ) from northeast Florida. J. Fish. Res. Board Can. 30(7):1009-1011. Shane, S. H. 1977. The population biology of the Atlantic bottlenose dol- phin, Tursiops truncatus , in the Aransas Pass area of Texas. M S. thesis, Texas A&M Univ., College Station, TX, 239 p. Sokal, R. R., and F. J. Rohlf. 1981. Biometry. W. H. Freeman and Company, San Fran- cisco, CA, 859 p. True, F. W. 1890. Observations on the life history of the bottlenose porpoise. Proc. U.S. Nat. Mus. 13:197-203. Urian, K. W., D. A. Duffield, A. J. Read, R. S. Wells, and E. D. Shell. 1996. Seasonality of reproduction in bottlenose dolphins, Tursiops truncatus. J. Mammal. 7(2):394-403. Wells, R. W., and M. D. Scott. 1990. Estimating bottlenose dolphin population parameters from individual identification and capture-release tech- niques. In P. S. Hammond, S. A. Mizroch, and G. P. Donovan (eds), Individual recognition of cetaceans: use of photo-identification and other techniques to estimate popu- lation parameters, p. 407-415. Rep. Int. Whaling Comm., spec, issue 12. 366 Diet of dusky dolphins, Lagenorhynchus obscurus, in waters off Patagonia, Argentina Mariano Koen Alonso Enrique Alberto Crespo Nestor Anibal Garcia Susana Noemi Pedraza Mariano Alberto Coscarella Centro Nacional Patagonico (CONICET) Universidad Nacional de la Patagonia and Fundacion Patagonia Natural Boulevard Brown s/n, (9120) Puerto Madryn, Chubut, Argentina E-mail address (for M. Koen. Alonso): koen@cenpat.edu. ar The dusky dolphin, Lagenorhyn- chus obscurus, is probably the best known species of the genus Lagen- orhynchus in the Southern Hemi- sphere. It is common in temperate waters of New Zealand, South Af- rica, Peru, northern Chile, and Ar- gentina (Crespo, 1991; Van Waere- beek, 1992; Crespo et al., in press). Incidental catches have been re- corded in coastal and offshore fish- eries of Argentina (Corcuera et al., 1994; Crespo et al., 1994a, 1994b; Dans et al., in press), whereas di- rected captures have occurred off Peru (Read et al., 1988; Van Waere- beek and Reyes, 1990; Reyes, 1992). The biology of dusky dolphins has been studied in Argentina since the 1970’s (Wiirsig and Wiirsig, 1980; Wiirsig and Bastida, 1986; Dans et al., 1993). A detailed behavioral study was carried out at Peninsula Valdes (Wiirsig and Wiirsig, 1980), and preliminary information about the reproductive biology of females was obtained by Dans et al. ( 1993, 1997). Feeding habits have been briefly described for dusky dolphins from New Zealand (Gaskin, 1972), Pata- gonia (Dans et al., 1993; Crespo et al., 1994c), and Antarctic and sub- Antarctic waters (Goodall and Ga- leazzi, 1985). In Patagonia, the be- havior of dusky dolphins feeding on Argentine anchovy, Engraulis anchoita, schools has been de- scribed by Wiirsig and Wiirsig (1980). A more recent and detailed study of their feeding habits by stomach content analysis was car- ried out in coastal waters of central Peru (McKinnon, 1994). The Peru- vian sample consisted of animals caught mostly in gill nets (McKin- non, 1994). Peruvian anchoveta, Engraulis ringens, was the most important prey species for Peru- vian dusky dolphins, accounting for 92.5% in number and 83.8% in weight of the total sample, and 97.8% frequency of occurrence (McKinnon, 1994). Differences in diet between lactating and nonlac- tating mature females were not found, and the anchoveta was al- most the exclusive prey species in both sexes (McKinnon, 1994). The objective of the present study is to describe the diet of the dusky dolphin in Patagonia in terms of both species composition and prey size. Seasonal and annual compo- sition were not investigated be- cause the sample size was small when subdivided along these lines. Materials and methods The sample consisted of 25 dolphins (6 males and 19 females) caught incidentally between November 1989 and April 1994 by the trawl- ing fishery that operates over the Patagonian Continental Shelf. All dolphins were caught in the area between 43°S and 46°30'S and be- tween the coastline and the 100-m isobath. Most individuals were fro- zen onboard to -20°C when they were caught. The total weight, standard length, and other body measurements (Nor- ris, 1961) were recorded for each dolphin before dissections were made. Stomach contents were col- lected and then stored, either fro- zen or in 70% alcohol. They were separated with several sieves with a mesh size between 25 mm and 0.5 mm. Flesh material from the sieved contents was placed in a deep, wa- ter-filled plastic tray, enabling re- moval of the remaining flesh mate- rial. Otoliths, fish bones, and ceph- alopod beaks were identified with the aid of published catalogues (Clarke, 1962, 1980, 1986; Torno, 1976; Menni et al., 1984; Roper et al., 1984; Boschi et al., 1992; Gosz- tonyi and Kuba1) and a reference collection of otoliths, cephalopod beaks, and crustaceans belonging to the Marine Mammals Labora- tory, Centro Nacional Patagonico, CONICET. The number of individuals of a particular fish species (NF) found in a stomach was calculated as 1 Gosztonyi, A., and L. Kuba. 1996. Atlas de huesos craneales y de la cintura escapular de peces costeros patagonicos. Plan de Manejo Integrado de la Zona Costera Patagonica (PMIZCP), Global En- vironmental Facility (GEF), Programa de las Naciones Unidas para el Desarrollo (PNUD), Wildlife Conservation Soc. (WCS), and Fundacion Patagonia Natu- ral (FPN) Informe Tecnico 4, 29 p. Manuscript accepted 6 August 1997. Fishery Bulletin 96:366-374 (1998). NOTE Koen Alonso et al. : Diet of Lagenorhynchus obscurus in waters off Patagonia 367 NF - LO + 0.5NO, if LO >RO, or as NF = RO + 0.5NO, if RO>LO, where LO, RO, and NO = the number of left, right, and not assigned otoliths respectively. Minimal number of cephalopods was obtained with maximum count of upper and lower beaks for each species. All complete and nondigested elements (otoliths and beaks) were measured with digital calipers. Some otoliths or beaks were digested, or broken, and could not be measured. Any otolith was considered too di- gested to be measured if it presented rounded bor- ders and rostrum, and an ill-defined sulcus. Around 42% of the prey items were represented by broken or digested elements. Size values for these pieces were randomly assigned from sizes of items taken from the stomach and measured. When only broken or digested elements were found in the stomach, the average individual weight of the prey species in the whole sample was used to estimate the individual weight of the items. The length of prey items was estimated by using regressions between total length (TL) and otolith length (OL) for fishes, and between dorsal mantle length (DML) and lower rostral length (LRL), or lower hood length (LHL), for squids. Wet weight (W) was estimated from regression-estimated length of prey by using regressions between W and TL for fishes, and between W and DML for squids. Regres- sions for most prey species were developed during this study on the basis of materials collected from commercial hauls. All regressions were calculated as simple linear regressions. Those variables that did not have linear relationships were made linear with natural logarithms (Table 1). The TL and W of Notothenia sp. was estimated with Notothenia angustifrons regressions (Table 1). The W of Octo- pus tehuelchus was estimated with Octopus vulgaris regression between W and LHL (Table 1). The W of each Semirossia tenera was assumed to be 3 g, as was the average individual weight recorded for this species in a sample taken onboard a commercial trawler by one of the authors (N.A.G.). In one stom- ach (that of L022 dolphin), only fish eye lenses and a few beaks were found. In this case, eye lenses were similar in size to those of anchovy; therefore the an- Table 1 Equation, sample size (n), and determination coefficient ( r 2) of the regressions used to estimate size and weight of prey of dusky dolphins in Patagonia. The variables used in the regressions were otolith length (OL), lower hood length (LHL), lower rostral length (LRL), total length (TL), dorsal mantle length (DML) and wet weight (W). OL, LHL, and LRL are in mm, TL and DML are in cm, and W is in g. Scientific name Common name Equation n r2 Source Engraulis anchoita Argentine anchovy TL = 2.368 + 3.56 LO 79 0.70 This study W= 2.5 10-3 TL3 353 81 0.93 Merluccius hubbsi Argentine hake TL = 1.823 LO1 072 if OL<15 447 0.93 W = 4.76 HP3 TL3 061 if OL< 15 469 0.92 This study TL = 1.984 LO1 05 if OL>15 693 0.91 W = 9.72 HP3 TL2 886 if OL>15 742 0.96 Stromateus brasiliensis “pampanito” TL = 3.042 LO1 159 51 0.98 This study W = 6.418 10-4 TL3 917 63 0.98 Notothenia angustifrons Southern cod TL = 4.142 LO0 768 22 0.91 Hecht, 1987 W = 3.73 10-6 (10 TL)3 16 24 0.98 Illex argentinus Argentine shortfin squid DML = -3.178 + 5.617 LHL 27 0.93 This study DML = 8.257 10~2 + 6.009 LRL 63 0.87 W = 9.82 10-3 LDM3 238 66 0.98 Loligo gahi Patagonian squid DML = -0.712 + 4.622 LHL 98 0.76 This study W = 2.6 10-2 LDM 2 753 102 0.93 Octopus vulgaris octopus _ gl.82 + 3.03 Ln (LHL) 108 1 Clarke, 1986 ' The determination coefficient of this regression was not available in the original source. 368 Fishery Bulletin 96(2), 1998 chovy average individual weight was used to esti- mate the W of these unidentified fishes. This last estimation was made to evaluate the relative contri- bution by weight of these unidentified fishes. The “percent frequency of occurrence” (%FO), the “percent total number” (%N), and the “percent total wet weight” (%W) were calculated (Hyslop, 1980; Castley et al., 1991; McKinnon, 1994). The relative importance of each prey species was evaluated with the index of relative importance (IRI) (Pinkas et al., 197 1 ), where the volumetric percentage was replaced by the %W (. IRI=(%N+%W)%FO ) (Castley et al., 1991). Results Dolphins used in this study (Table 2) were caught mostly at night in midwater hauls for the shrimp Pleoticus muelleri, with the exception of LO10 and L023 dolphins that were caught in diurnal bottom hauls for shrimp, L022 dolphin that was captured in a diurnal bottom haul for hake, and L026 dolphin that was caught by an unknown fishing vessel. The locations of the catches were principally in the Golfo San Jorge (approximately between 45 and 46°30'S), except L022 dolphin that was captured in the vicin- ity of Isla Escondida (43°S). There is no information Table 2 Biological information (sex, size, age, and reproductive status), number of prey species eaten, and most important prey species by weight in the stomach of each dusky dolphin considered in this study. The biological information of females (F) were taken from Dans et al. (1997) and those of males (M) were provided by S.L. Dans.1 The date of death is approximate and indicates which dolphins were caught in the same fishing trip. Field number Date of death Sex Length (cm) Age (years) Reproductive status No. of prey eaten Important prey species in the stomach LOOl 19 Nov 89 F 158 5 immature 3 Argentine anchovy and Argentine shortfin squid LO02 19 Nov 89 F 157 4 immature 4 Argentine shortfin squid and Argentine anchovy LO04 27 Apr 90 F 159 6 immature 3 Argentine shortfin squid and Argentine anchovy LO05 01 Jun 90 F 166.5 6 mature (lactating) 1 Argentine anchovy LO06 01 Jun 90 F 174 8 mature (resting) 4 Argentine anchovy LO07 Jul 90 M 169 6 immature 5 Argentine anchovy LO08 Apr 92 F 161 7 mature (pregnant) 2 Argentine anchovy LO09 Apr 92 F 174 11 mature (pregnant) 4 Argentine anchovy LO10 08 Sep 92 F 172 3+ immature 4 Argentine shortfin squid and Argentine hake LOll Sep 92 F 172 7+ mature (resting) 4 Argentine anchovy L012 Sep 92 F 162 7+ mature (pregnant) 6 Argentine hake L013 Sep 92 F 170 8+ mature (resting) 5 Argentine anchovy and hake L014 Sep 92 F 167 7+ immature 3 Argentine hake L015 11 Mar 93 F 166 5 immature 2 Argentine anchovy L016 11 Mar 93 F 174 6 mature (pregnant) 3 Argentine anchovy L017 16 Mar 93 F 164 4 immature 2 Argentine anchovy L018 06 Apr 94 F 171 5 immature 4 Argentine anchovy L019 11 May 93 M 164 8 mature 3 Argentine anchovy LO20 Sep 93 M 161 9 mature 2 Argentine hake L021 02 Sep 93 F 158 3 immature 4 Argentine hake L022 30 Oct 93 F 164 3 immature 2 Unidentified fishes L023 15 Oct 93 M 169 9 mature 3 Argentine shortfin squid L024 Mar 94 M 173 10 mature 4 Argentine anchovy L025 Mar 94 M 162 8+ unknown 3 Argentine anchovy L026 94 F 161 8 mature 5 Argentine anchovy and hake 1 Dans, S. L. Laboratorio de Mamiferos Marinos, CENPAT (CONICET), Boulevard 3600, (9120) Puerto Madryn, Chubut, Argentina NOTE Koen Alonso et al.: Diet of Lagenorhynchus obscurus in waters off Patagonia 369 about the location of collection of two dolphins (LO07 and L026). The entire sample was biased toward females and it was composed of immature and ma- ture dolphins (Table 2) (Dans et al., 1997). Some dolphins in the sample may have been caught in the same haul because fishermen reported that occasionally more than one dolphin are caught. How- ever, when fishermen capture several dolphins in a fishing trip (between 30 and 60 days), they may dis- card some of them. In these cases it is uncertain whether those dolphins caught during the same fish- ing trip (same date of death, see Table 2) were caught together in the same haul. With the exception of one stomach (L022) which contained fish eye lenses and a few beaks, all others were half full or full. On average (± standard devia- tion), 3.40 ± 1. 19 prey species per stomach were found, but only one or two of them were important by weight in the stomach (Table 2). The mean number of prey items per stomach was 148.08 ±137.91, and the mean regression-estimated ingested biomass was 2,675.63 ±1,848.44 g per stomach. A total of 3,702 prey items belonging to eight spe- cies were found in the 25 stomach contents analyzed, including four fish and four cephalopod species. The fish species were Argentine anchovy, Engraulis anchoita; Argentine hake, Merluccius hubbsi; “pampanito, ” Stromateus brasiliensis, and the south- ern cod Notothenia sp. The cephalopods were Argen- tine shortfin squid, Illex argentinus; Patagonian squid, Loligo gahi, the sepiolid Semirossia tenera; and common octopus, Octopus tehuelchus. Argentine anchovy was the most important prey species, representing 39% of prey by number, 46% by weight (Table 3). However, the most frequent prey was the Patagonian squid which was present in 84% of stomachs (Table 3). According to the IRI, the sec- ond species in importance was the Argentine shortfin squid, followed by the Patagonian squid and the Ar- gentine hake, whereas by weight the second species was hake, followed by Argentine shortfin and Patagonian squid (Table 3). The “pampanito,” octo- pus, and southern cod were of little importance in dusky dolphin diet (Table 3). The Argentine anchovy had a unimodal length-fre- quency distribution with a mode at 16 cm (Fig. 1). Individual anchovies eaten by dusky dolphins were mostly of mature sizes (Hansen, 1994). Both Argentine shortfin and Patagonian squids were consumed at small sizes. Patagonian squids had a modal DML of 3 cm (Fig. 2), and most of them were smaller than 8 cm, a size that corresponded to that of immature squid (Hatfield et al., 1990). The DML frequency distribution of shortfin squid showed two important peaks at 1-2 cm, and 7-8 cm of DML (Fig. 3) which corresponded to juvenile sizes ( Brunetti and Ivanovic, 1992). Argentine hake showed the widest size range (Fig. 4). Its TL frequency distribution showed three peaks at 8 cm, 22-24 cm, and 36 cm in order of importance (Fig. 4). These sizes corresponded to age-0 group, age 1-2 group, and age 2-5 group, respectively (Gaggiotti and Renzi, 1990). Average individual length of fish eaten by dusky dolphins ranged from 7.54 cm (southern cod) to 19.81 cm (Argentine hake). The DML range for cephalo- pod preys was estimated to be between 2 cm (ob- served DML of S', tenera) and 6.35 cm (shortfin squid). Relative Table 3 importance of prey species of dusky dolphins in the Patagonian continental shelf. Percent total number (%N) Percent total wet weight (%W) Percent frequency of occurrence (%FO) Index of relative importance (IRI) Cephalopods Illex argentinus 30.71 21.05 68.00 3,520.11 Loligo gahi 20.64 4.58 84.00 2,118.09 Semirossia tenera 3.40 0.57 36.00 142.87 Octopus tehuelchus 0.05 0.00 4.00 0.22 Fishes Engraulis anchoita 39.25 46.36 80.00 6,848.60 Merluccius hubbsi 5.19 26.46 48.00 1,518.82 Stromateus brasiliensis 0.14 0.29 12.00 5.05 Notothenia sp. 0.03 0.00 4.00 0.11 Unidentified fishes 0.59 0.70 4.00 5.18 Total 100.00 100.01 370 Fishery Bulletin 96(2), 1 998 For all prey species, average individual length was smaller than 20 cm. Average individual wet weights of Argentine shortfm squid, Patagonian squid, Argentine hake, and Argentine anchovy ranged from 4 to 92 g. The Argentine anchovy, Argentine hake, and Argentine shortfm squid had an average individual wet weight (± standard deviation) of 21.34 ±7.49, 92.17 ±121.62, and 12.39 ±33.62 g, respectively. Discussion Patagonian dusky dolphins fed mostly on pelagic species or pelagic stages of dem- ersal species. Argentine anchovy is typically a pelagic species (Brandhorst et al., 1974; Ange- lescu, 1982; Angelescu and Anga- nuzzi, 1986; UNESCO, 1990), and Argentine shortfin squid, Patagonian squid, Argentine hake, and “pampanito” are dem- ersal-pelagic species (FAO, 1983; Angelescu and Prenski, 1987; Nigmatullin, 1989; Hatfield et al., 1990; Rodhouse and Hatfield, 1990; Brunetti and Ivanovic, 1992). The sepiolid and southern cod are demersal-benthic species, whereas common octopus is ben- thic (Roper et al., 1984; Angelescu and Prenski, 1987). Anchovies are the main food item for dusky dolphins in Argen- tine and Peruvian waters. Engrau- lis ringens was found to be the principal prey species for dusky dolphins in waters off Peru (Mc- Kinnon, 1994), and this study shows that E. anchoita is the most important prey species in Argentine waters. Furthermore, squids were the second prey by numbers (I. argentinus and L. gahi in Argentine waters, and L. gahi and Dosidicas gigas in Pe- ruvian waters) and hakes were rated second by weight (M. hub- bsi and Merluccius gayi in Argen- tine and Peruvian waters respec- tively) (McKinnon, 1994). More- over, the number of commonly eaten and important prey species (by weight) was quite similar be- tween Argentine and Peruvian dusky dolphins (4 and 6 prey spe- cies respectively), considering the difference in the sample size in this work (n= 25) and that in McKinnon’s (1994) study (n=136). Both populations of the dusky NOTE Koen Alonso et a I.: Diet of Lagenorhynchus obscurus in waters off Patagonia 371 Total length (cm) Figure 4 Length-frequency distribution of Argentine hake, Merluccius hubbsi, eaten by dusky dolphins in Patagonia. dolphin feed on ecologically simi- lar species, mostly schooling fishes. However, whereas ancho- veta represented 80% in weight of the diet of Peruvian dolphins, Argentine anchovy represented between 40% and 50% in weight of the diet of Argentine dolphins. Sample size in this study is relatively small and possibly bi- ased by age and sex (Dans et al., 1997; in press). However, our sample included mature and im- mature dolphins of both sexes, and all the stomach contents were very similar in composition (Table 2). Moreover, dusky dol- phins had never been reported to be feeding around fishing vessels on discards and disturbed fish, even when the dolphins in this sample were caught by shrimp trawlers. Furthermore, the diet composition of Peruvian dusky dolphins did not show any differ- ence between sexes or reproduc- tive status (McKinnon, 1994). At Peninsula Valdes, dusky dolphin feeding groups consisted of indi- viduals of both sexes and of al- most all age classes (calves and small young could be excluded from the feeding group) (Wiirsig and Wiirsig, 1980). Therefore, considering the general similari- ties in diet composition between Argentine and Peruvian popula- tions, the absence of differences in the diet between sexes and reproductive status in the Peru- vian population, the lack of evi- dence to suggest that dolphins feed on discards and disturbed fishes, and the gregarious feed- ing behavior of patagonian dusky dolphins described by Wiirsig and Wiirsig ( 1980), we conclude that the sample analyzed in this study is a reasonable indicator of diet composition of dusky dolphins in Argentine waters. At least two populations of Argentine anchovy have been reported over the Argentine Continental Shelf: the bonaerensis (northern) and the Patagonian (southern) stocks (Brandhorst et al., 1974; Angelescu, 1982; Angelescu and Anganuzzi, 1986; UNESCO, 1990; Hansen, 1994). Preadults and adults of the bonaerensis stock migrate northwest from coastal spawning areas to offshore feeding grounds; avail- able information indicates that the Patagonian stock, however, does not show a seasonal migration (Brandhorst et al., 1974; Angelescu, 1982; Angelescu and Anganuzzi, 1986; UNESCO, 1990; Hansen, 372 Fishery Bulletin 96(2), 1998 1994). Dolphins in this study were caught off central Patagonia throughout the year between 43°S and 46°30'S; thus dusky dolphins appear to feed on Ar- gentine anchovies from the Patagonian stock. The absence of a well-defined seasonal migration pattern in Argentine dusky dolphins (Wtirsig and Wiirsig, 1980) could be related to predation on the Patagonian stock of Argentine anchovy. However, long-distance movements of two dusky dolphins have been recorded after several years in Argentine waters (Wiirsig and Bastida, 1986). The Argentine shortfin squid is a neritic-oceanic species, distributed from 30°S to 54°S (Rodhouse and Hatfield, 1990). At least three spawning stocks, 1) the bonaerensis to north Patagonian (early winter), 2) the south Patagonian (winter), and 3) the sum- mer spawning stock (Nigmatullin, 1989; Brunetti et al., 1991; Brunetti and Ivanovic, 1992) have been described over the Argentine Continental Shelf. The latter spends its entire life cycle over the continen- tal shelf, the former two have their spawning grounds on the shelf break and in oceanic waters. Their lar- vae and juveniles grow and mature over the conti- nental shelf (Nigmatullin, 1989; Rodhouse and Hatfield, 1990; Brunetti and Ivanovic, 1992). The shortfin squids captured by dolphins could belong to any of the spawning stocks because their distribu- tions overlap in the area where dolphins were caught. In addition, the size of squids eaten by dolphins cor- responds with that for juvenile squids from all spawn- ing stocks. These juvenile squids are pelagic (Brunetti and Ivanovic, 1992; Ivanovic and Brunetti, 1994) and can be found mostly between the 50- and 100-m isobaths (Brunetti and Ivanovic, 1992). The Patagonian squid is a neritic species and its life cycle is associated with the Malvinas Current over the continental shelf and slope (Hatfield et al., 1990). Its spawning grounds are in shallow and coastal waters, and it migrates to deeper waters to grow, reaching maturity at sizes larger than 10 cm of DML (Hatfield et al., 1990). Patagonian squids eaten by dusky dolphins were mostly of small size; thus feeding on this species could occur as juveniles migrate from the coastal area to deeper feeding grounds beyond the shelf break. The main TL peak for Argentine hake eaten by dolphins corresponds to the age-0 group. This group exhibits pelagic schooling behavior (Angelescu and Prenski, 1987). Moreover, juveniles do not follow the adult hake migration pattern but remain in their nursery grounds (south of the Golfo San Jorge and Isla Escondida area) (Angelescu and Prenski, 1987). In sum, these results indicate that dusky dolphins of Patagonia feed mostly on species whose stages exhibit pelagic and schooling behavior. Finally, Argentine hake, Argentive anchovy, and squids are key species in the trophic web of the Pata- gonian Continental Shelf ecosystem and seem to present some degree of association (Angelescu and Prenski, 1987). Peneid crustaceans, like the shrimp, appear to be associated with these species in the study area even when their abundance is comparatively less (Angelescu and Prenski, 1987). Shrimp were not found in dolphins stomachs, even when they were caught in- cidentally in shrimp hauls. There are no data about the behavior of the dolphins at the moment of the catch, but the absence of shrimp in the diet suggests that dol- phins avoid shrimp as a prey. Acknowledgments The authors are indebted to Silvana L. Dans for her critical readings and helpful comments on an earlier version of this paper; Pablo Yorio for his critical read- ing, editorial suggestions, and help with translations; Maria del Carmen Falcon for help with English; Pablo Mariotti, Laura Reyes, and Barbara Beron Vera for assistance in the field and laboratory; and to all the fishermen for their help at sea and con- stant interest in our work. Kimberly Murray, John B. Pearce, and three anonymous reviewers made in- teresting and useful comments that enhanced the validity of results. Sharyn Matriotti gave us impor- tant help with this article. This work was carried out with the financial support of the National Geo- graphic Society (grant 4249/90 to E.A. Crespo and A.C.M. Schiavini), the Whale and Dolphin Conser- vation Society, the Patagonian Coastal Zone Man- agement Plan (GEF/PNUD), and the Programa de Cooperacion Cientifica con Iberoamerica (BOE 29- III-96). The University of Patagonia, CENPAT- CONICET, and Fundacion Patagonia Natural pro- vided logistical support. An earlier version of this paper was presented by E.A. Crespo as SC/48/SM19 in the 48th Annual Meeting of the International Whal- ing Commission, Aberdeen, Scotland, 5-17 June 1996; he attended the meeting with support from the IWC and Cetacean Society International. Thanks are also given to Tony Martin and Kate O’Connell. Literature cited Angelescu, V. 1982. Ecologia trofica de la anchoita del Mar Argentine) (En- graulidae, Engraulis anchoita). Parte II. Alimentacion, comportamiento y relaciones troficas en el ecosistema. Contrib. Inst. Nac. Invest. Des. Pesq. (INIDEP) 409, 83 p. Angelescu, V., and A. Anganuzzi. 1986. Ecologia trofica de la anchoita ( Engraulidae, NOTE Koen Alonso et a I.: Diet of Lagenorhynchus obscurus in waters off Patagonia 373 Engraulis anchoita) del Mar Argentine). Parte III: Requerimiento trofico individual en relacion con el crecimiento, ciclo sexual y las migraciones estacion- ales. Rev. Invest. Des. Pesq. (INIDEP) 5:194-223. Angelescu, V., and L. B. Prenski. 1987. Ecologia trofica de la merluza comun del Mar Argentino (Merlucciidae, Merluccius hubbsi). Parte 2: Dinamica de la alimentacion analizada sobre la base de las condiciones ambientales, la estructura y las evalua- ciones de los efectivos en su area de distribucion. Contrib. Inst. Nac. Invest. Des. Pesq. (INIDEP) 561, 205 p. Boschi, E. E., C. E. Fischbach, and M. I. Iorio. 1992. Catalogo ilustrado de los crustaceos estomatopodos y decapodos marinos de Argentina. Frente Maritimo, vol.10, sec. A:7-94. Brandhorst, W., J. P. Castello, M. B. Cousseau, and D. A. Capezzani. 1974. Evaluacion de los recursos de anchoita ( Engraulis anchoita) frente a la Argentina y Uruguay. VIII. Desove, crecimiento, mortalidad y estructura de la poblacion. Physis (sec. A) 33(861:37-58. Brunetti, N. E., and M. L. Ivanovic. 1992. Distribution and abundance of early life stages of squid (Illex argentinus ) in the south-west Atlantic. ICES J. Mar. Sci. 49:175-183. Brunetti, N. E., M. L. Ivanovic, E. Louge, and H. E. Christiansen. 1991. Estudio de la biologia reproductiva y de la fecundidad en dos subpoblaciones del calamar ( Illex argentinus). Frente Maritimo, vol. 8, sec. A:73-84. Castley, J. G., V. G. Cockcroft, and G. I. H. Kerley. 1991. A note on the stomach contents of fur seals Arctocephalus pusillus beached on the south-east coast of South Africa. S. Afr. J. Mar. Sci. 11:573-577. Clarke, M. R. 1962. The identification of cephalopods beaks and the re- lationship between beak size and total body weight. Bull. Br. Mus. Nat. Hist. 8( 10):419-480. 1980. Cephalopoda in the diet of sperm whales of the south- ern hemisphere and their bearing on sperm whale biology. Discovery Rep. 37, 324 p. 1986. A handbook for the identification of cephalopods beaks. Clarendon Press, Oxford, 273 p. Corcuera, J., F. Monzon, E. A. Crespo, A. Aguilar, and J. A. Raga. 1994. Interactions between marine mammals and the coastal fisheries of Necochea and Claromeco (Buenos Aires province, Argentina). In W. F. Perrin, G. P. Donovan, and J. Barlow (eds.), Gillnets and cetaceans, p 283-290. Rep. Int. Whal. Comm, (special issue 15). Crespo, E. A. 1991. Lagenorhynchus obscurus, delfi'n oscuro. In H. L. Capozzo and M. Junin (eds.), Estado de conservacion de los mamiferos marinos del Atlantico Sudoccidental, p 42- 45. Informes y estudios del Programa de Mares Regionales del PNUMA 138. Crespo, E. A., L. Reyes, M. Koen, and N. Garcia. 1994a. Interacciones entre mamiferos marinos y pesquerias en el litoral norpatagonico. In J. A. Oporto (ed.), Anales IV Reu. Trab. Esp. Mamif. Acuat. America del Sur, p 89- 96. Imprenta Universitaria S.A., Universidad Austral de Chile (UACH). Crespo, E. A., J. F. Corcuera, and A. Lopez Cazorla. 1994b. Interactions between marine mammals and fisher- ies in some coastal fishing areas of Argentina. In W. F. Perrin, G. P. Donovan, and J. Barlow (eds.) Gillnets and cetaceans, p 269-281. Rep. Int. Whal. Comm, (special issue 15). Crespo, E. A., P. Nepomnaschy, M. Koen Alonso, and N. Garcia. 1994c. Analisis preliminar de la dieta de mamiferos marinos sobre la base de contenidos estomacales y heces. In J. A. Oporto (ed.), Anales IV Reu. Trab. Esp. Mamif. Acuat. America del Sur, p 75-88. Imprenta Universitaria S.A., Universidad Austral de Chile, Valdivia. Crespo, E. A., S. N. Pedraza, M. Coscarella, N. A. Garcia, S. L. Dans, M. Iniguez, L. M. Reyes, M. Koen Alonso, A. C. M. Schiavini, and R. Gonzalez. In press. Distribution and school size of dusky dolphins Lagenorhynchus obscurus (Gray, 1821) in the southwest- ern South Atlantic Ocean. Rep. Int. Whal. Comm. 47, 22 p. Dans, S., E. A. Crespo, M. Koen Alonso, L. M. Reyes, and N. A. Garcia. 1993. Biologia y ecologia del delfi'n oscuro ( Lagenorhynchus obscurus) en el litoral patagonico. Aspectos prelimin- ares. In G. Caille, E. A. Crespo and J. L. Esteves (eds.), Actas de las J. Nac. de Cs. del Mar ’91, p 142-148. Uni- versidad Nacional de la Patagonia and Centro Nacional Patagonico (CONICET), Puerto Madryn.. Dans, S. L., E. A. Crespo, S. N. Pedraza, and M. Koen Alonso. 1997. Notes on the reproductive biology of female dusky dolphins ( Lagenorhynchus obscurus) off the patagonian coast. Mar. Mamm. Sci. 13(2):303-307. Dans, S. L., E. A. Crespo, N. A. Garcia, L. M. Reyes, S. N. Pedraza, and M. Koen Alonso. In press. Incidental mortality of patagonian dusky dol- phins in mid-water trawling: retrospective effects from the early 80’s. Rep. Int. Whal. Comm. 47, 12 p. FAO. 1983. Informe del Grupo Ad Hoc de Trabajo sobre los recursos pesqueros de la Plataforma Continental Pata- gonica, Roma, 7-11 de febrero 1983. Una reunion preparatoria para la Conferencia Mundial de la FAO sobre Ordenacion y Desarrollo. FAO Informe de Pesca 297, 83 p. Gaggiotti, O. E., and M. A. Renzi. 1990. El crecimiento alometrico en la merluza comun ( Merluccius hubbsi) y eleccion de un modelo de crecimiento apropiado. Frente Maritimo, vol. 6, sec. A:7-24. Gaskin, D. E. 1972. Whales, dolphins and seals with special reference to the New Zealand region. St. Martin’s Press, New York, NY, 200 p. Goodall , R. N. P., and A. R. Galeazzi. 1985. A review of the food habits of the small cetaceans of the Antarctic and the sub-Antarctic. In W. R. Siegfried, P. R. Condy, and R. M. Laws (eds.), Antarctic nutrient cycles and food webs, p 566-572. Springer- Verlag, Berlin. Hansen, J. E. 1994. Diferencias entre parametros vitales de las poblacio- nes bonaerense y patagonica de anchoita argentina. Rev. Invest. Des. Pesq. (INIDEP) 9:11-24. Hatfield, E. M. C., P. G. Rodhouse, and J. Porebski. 1990. Demography and distribution of the Patagonian squid (Loligo gahi d’Orbigny) during the austral winter. J. Cons. Int. Explor. Mer 46:306-312. Hecht, T. 1987. A guide to the otoliths of southern ocean fishes. S. Afr. J. Antarct. Res. Vol. 17(1), 87 p. Hyslop, E. J. 1980. Stomach content analysis. A review of methods and their aplication. J. Fish. Biol. 17:411-429. 374 Fishery Bulletin 96(2), 1998 Ivanovic, M. L., and N. E. Brunetti. 1994. Food and feeding of Illex argentinus. Antarctic Sci- ence 6(2): 185—193. McKinnon, J. 1994. Feeding habits of the dusky dolphin, Lagenorhynchus obscurus, in the coastal waters of central Peru. Fish. Bull. 92:569-578. Menni, R. C., R. A. Ringuelet, and R. A. Aramburu. 1984. Pecesmarinosde la Argentina y Uruguay. Editorial Hemisferio Sur S.A., Buenos Aires, 169 p. Nigimatullin, Ch. M. 1 989. Las especies de calamar mas abundantes del Atlantico Sudoeste y sinopsis sobre la ecologia del calamar ( Illex argentinus). Frente Maritimo, vol. 5, sec. A:71-81. Norris, K. S. 1961. Standarized methods for measuring and recording data on the smaller cetaceans. J. Mammal. 42(41:471^176. Pinkas, L., M. S. Oliphant, and I. L. K. Iverson. 1971. Food habits of albacore, bluefin tuna and bonito in California waters. Fish Bull. 152:1-105. Read, A. J. , K. Van Waerebeek, J. C. Reyes, J. S. McKinnon, and L. C. Lehman. 1988. The exploitation of small cetaceans in coastal Peru. Biol. Conserv. 46:53-70. Reyes, J. C. 1992. Informe nacional sobre la situacion de los mamiferos marinos en Peru. Informes y estudios del Programa de Mares Regionales del PNUMA 145, 21 p. Rodhouse, P. G., and E. M. C. Hatfield. 1990. Dynamics of growth and maturation in the cephalo- pod Illex argentinus de Castellanos, 1960 (Teuthoidea: Ommastrephidae). Phil. Trans. R. Soc. Lond. 329:229- 241. Roper, C. F. E., M. J. Sweeney, and C. E. Nauen, 1984. FAO species catalogue. Vol. 3. Cephalopods of the world. An annotated and illustrated catalogue of species of interest to fisheries. FAO Fish. Synop. 125(31:1-227. Tor no, A. E. 1976. Descripcion y comparacion de los otolitos de algunas familias de peces de la Plataforma Argentina. Rev. Mus. Arg. Cs Nat. “Bernardino Rivadavia” (Zool.) 12(41:2-20. UNESCO (United Nations Educational, Scientific, and Cultural Organization). 1990. Synopsis on the reproductive biology and early life history of Engraulis anchoita, and related enviromental conditions in argentine waters. IOC Workshop Report 65, annex V, 49 p. Van Waerebeek, K. 1992. Records of dusky dolphins, Lagenorhynchus obscurus (Gray, 1828) in the Eastern South Pacific. Beaufortia 43(41:45-61. Van Waerebeek, K., and J. C. Reyes. 1990. Catch of small cetaceans at Pucusana port, central Peru, during 1987. Biol. Conserv. 51:15-22. Wiirsig, B., and R. Bastida. 1986. Long-range movement and individual associations of two dusky dolphins ( Lagenorhynchus obscurus) off Argentina. J. Mammal. 67(41:773-774. Wiirsig, B., and M. Wiirsig. 1980. Behavior and ecology of the dusky dolphin, Lagen- orhynchus obscurus, in the South Atlantic. Fish. Bull. 77:871-890. 375 Preliminary estimate of spawning frequency and batch fecundity of striped weakfish, Cynoscion striatus, in coastal waters off Buenos Aires province* Gustavo J. JVIacchif Consejo Nacional de Investigaciones Cientificas y Tecnicas (CONICET) Instituto Nacional de Investigacion y Desarrollo Pesquero (INIDEP) CC. 175, Mar del Plata (7600), Argentina E-mail address: gmacchi@inidep.edu. ar The striped weakfish, Cynoscion striatus, is a demersal species in- habiting coastal waters of the Southwest Atlantic. It ranges from Brazil (22°35'S) to the south of Buenos Aires province (Argentina) (42°S) (Cordo, 1986). The commer- cial catch takes place mainly off Buenos Aires province, between 35° and 40°S, to 50 m depths. The an- nual catch of this species is about 20,000 metric tons,* 1 corresponding to 20% of the total commercial catch of coastal species in Argentina. Despite a variety of studies de- scribing C. striatus life history, such as feeding (Ciechomski and Ehr- lich, 1977; Cordo, 1986), growth (Ciechomski and Cassia, 1978), embryonic and larval development (Ciechomski and Cassia, 1982), length distributions, and sex ratios (Cousseau et al., 1986), studies on reproductive biology are rare and largely incomplete. Only one pre- vious report has included histologi- cal analysis of the ovaries and fe- cundity estimates (Cassia, 1986), providing evidence that striped weakfish is a multiple spawner. In the present study, ovaries of C. striatus from the coastal area of the Buenos Aires province (Argen- tina, Southwest Atlantic) were ex- amined histologically, and particu- lar attention was paid to evidence for multiple spawning and to esti- mates of spawning frequency and batch fecundity. Material and methods Cynoscion striatus was colit \ ted in the zone known as “El Rincon” ( Buenos Aires province, Argentina, 38°50'-40°30'S) during two re- search trawl cruises carried out in early November 1994 (n=179) and September 1995 (n= 224) (Fig. 1). Mature females caught near “Punta del Este” (Uruguayan coast, 35°10'S- 55°W) in November 1994 were also used to estimate fecundity (n= 26). Individuals were measured to the nearest centimeter (total length, TL) and weighed (total weight, TW). Length range of sampled fe- males was 31-58 cm TL, which cor- responds to 3-15 yr old individu- als (Cioffi, 1992). The decision to sample only females of C. striatus greater than about 30 cm was based on the fact that this length is the size at maturity (Macchi and Acha2). Immature ovaries therefore were underrepresented in the samples. The ovaries of individuals sampled were removed immediately after capture and fixed in 10% neutral- buffered formalin or in Davidson’s fluid for two weeks. Later, fixed gonads were weighed (GW), and a portion of tissue (about 2.0 g) was removed from the center of each ovary, dehydrated in methanol, cleared in benzol, and embedded in paraffin. Tissues were sectioned at approximately 5 pm thickness, stained with Harris’s hematoxylin followed by eosin counterstain. His- tological classification of ovaries was based on the occurrence and relative abundance of five stages of oocyte development (Table 1) as well as on the occurrence and in- tensity of atresia. The terminology used in the description of oocyte stages was adapted from Forberg (1982), and the classification of atresia followed that given by Hunter and Macewicz ( 1985). Ova- ries were classified into eight stages: 1) immature, 2) developing early, 3) developing late, 4) fully developed, 5) gravid, 6) partially spent, 7) spent, and 8) resting. This classification is a modification of that given by Mayer et al. ( 1988). Oocyte measurements were taken of the ova- ries in different maturity stages (3 ovaries for each stage). Oocytes with visible nuclei showing were mea- sured along the longest axes with an ocular micrometer. About 10 fields of view per histological section were chosen at random (Fig. 2). To estimate the spawning frac- tion, only the samples collected during November 1994 were used. The percentage of spawning fe- * Contribution 1013 of the Instituto Nacio- nal de Investigacion y Desarrollo Pes- quero, Mar del Plata, Argentina. 1 Secretaria de Agriculture, Pesca y Alimen- tacion. 1996. Flota pesquera Argentina (capturas maritimas totales). [Available from SAPyA., Paseo Colon 982, Buenos Aires, Argentina.] 2 Macchi, G. J., and M. E. Acha. In press. Chapter 4: Aspectos reproductivos de las principales especies de peces. In Lasta (ed.), Resultados de la campana H-13/94, Serie Informes Tecnicos INIDEP. Manuscript accepted 9 July 1997. Fishery Bulletin 96:375-381 (1998). 376 Fishery Bulletin 96(2), 1 998 males was determined by estimating the incidence of fish with postovulatory follicles (POF) (Hunter and Goldberg, 1980). These structures were classified as day-0, day-1, or day-2 postovulatory follicles, following the description by Goldberg et al. (1984) for Sardinops sagax and by Melo (1994) for Engraulis capensis. 38°S 39° 40° 41 ° W Figure 1 Location where samples of female striped weakfish were collected off “El Rincon” area. The proportion of spawning females was estimated from the different stages of POF, but the spawning frequency was determinated by taking the average of the percentages of day-0 and day-1 spawning fe- males (Fitzhugh et al., 1993). Batch fecundity (BF) (number of oocytes released per spawning) was estimated gravimetrically with the hydrated oocyte method (Hunter et al., 1985) on fixed ovarian samples. Hydrated ovaries that con- tained new postovulatory follicles were not used for this estimation. Batch fecundity was determined for 41 females (17 from “El Rincon” and 24 from the Uruguayan coast). Three 0.1-g tissue subsamples were collected from the anterior, middle, and poste- rior sections of one ovary of each pair. These subsamples were weighed to the nearest 0.00001 g. All hydrated oocytes in each subsample were counted under stereoscopic microscope (magnification of 6x). Batch fecundity for each female was the product of the mean number of hydrated oocytes per unit weight and the total weight of the ovaries. Relative fecun- dity (RF) (hydrated oocytes per gram of body weight) was calculated as the batch fecundity divided by fe- male weight (without ovary). To compare the “El Rincon” data with those from the Uruguayan coast, the length ranges coincident with the two zones were truncated between 40 and 50 cm TL (“El Rincon,” n= 15; Uruguayan coast, n= 17). The fecundity val- ues and the coefficients of the batch fecundity to Table 1 Morphological changes observed in the different oocyte growth phases. Adapted from Forberg (1982). Oocyte growth stages Description First growth phase Cromatin nucleolus stage Early nucleolus stage Late nucleolus stage Previtellogenic oocytes with diameters smaller than 100 pm. Cytoplasm is basophilic and the nucleus shows a number of nucleoli situated peripherally. Follicular cells are flattened and become visible on the outer surface of oocyte. Second growth phase Yolk vesicle stage Diameter ranged between 100-250 pm. Cytoplasm is basophilic and shows small uncolored vacuoles (yolk vesicles). Granulosa and follicular thecae are distinguished and the zona radiata becomes visible around the periphery of oocyte. Primary yolk stage Diameter ranged between 250 and 350 pm. Numerous eosinophilic yolk globules appear between the yolk vesicles. These yolk globules are composed mainly of proteins. The zona radiata and the follicle epithelium are more prominent. Secondary yolk stage Diameter between 400 and 600 pm. The yolk globules have multiplied and increased in size, occupying all the cytoplasm. The nucleus exhibits an irregular shape, and the zona radiata increases in thickness. Tertiary yolk stage or final oocyte maturation Oocyte diameter ranged between 600 and 700 pm. In the first instance, the nucleus is displaced to the animal pole (migratory nucleus stage). The nuclear membrane disintegrates and the yolk globules tend to coalesce. In the section, these oocytes appear with a cytoplasm weakly eosinophilic and an irregular shape. NOTE Macchi. Estimate of spawning frequency and batch fecundity of Cynoscion striatus 377 total length relationships obtained for both areas were compared with a test of equality of means and a test of equality of coefficients (Draper and Smith, 1981 ). The relations of batch fecundity to total length and total weight (without ovary) were assessed with regression analysis (Draper and Smith, 1981). Sig- nificance of these relations were evaluated by test- ing if the slope of the regression was significantly different from zero. Results Oocyte diameter distributions of different maturity stages (immature to gravid) of C. striatus show that 40 -i n=520 B 25 100 175 250 325 400 475 550 625 700 >* o c . o Q) ex 0.56). A test of equality of coefficients indicated that no significant difference existed be- tween the equations for BF vs. TL for the two areas (P>0.05). The general equations for the total length and total weight were determined by combining all data (Fig. 4, A and B). No statistical differences were observed between the mean values of relative fecun- dity obtained for both zones (P>0.25). This para- meter ranged between 50 and 420 hydrated oocytes per gram of female (ovary free); an increase in the number of oocytes with size was observed (Fig. 4, C and D). Discussion The results obtained indicate that striped weakfish is a multiple spawner with indeterminate annual fecundity. Hence, the potential annual fecundity is not fixed prior to the onset of spawning, and unyolked oocytes continue to mature and to be spawned dur- ing the reproductive season ( Hunter et al., 1992). This assumption is based on two lines of evidence: 1) the presence of maturing ovaries with postovulatory fol- licles (partially spent stage), which indicates that after spawning one batch of eggs, a new batch devel- ops and is released; and 2) oocyte size frequency, which shows a continuous distribution of growing vitello- genic oocytes. Reproductive activity of striped weakfish in the coastal waters of Buenos Aires province ranges from October to March, with a main peak in November (Cassia, 1986). The histological examination of ova- ries obtained during September did not show evi- dence of spawning, but those taken in early Novem- ber indicated considerable reproductive activity, with a high proportion of postovulatory follicles and the appearance of hydrated oocytes. This corroborates the hypothesis that the spawning of C. striatus in “El Rincon” area generally begins in October. Unfor- tunately, it was not possible to sample this species during the rest of the spawning season, which lasts about 6 mo. according to Cassia (1986). The long re- productive season of this species has also been ob- served in other sciaenids inhabiting temperate wa- ters, such as Seriphus politus (7 mo.) (De Martini and Fountain, 1981), Cynoscion nebulosus (7 mo.) ( Brown-Peterson et al., 1988), Micropogonias undulatus (7 mo.) (White and Chittenden, 1977; Barbieri et al., 1994), Genyonemus lineatus (7 mo.) (Love et al., 1984), Cynoscion nothus (6 mo.) (De Vries and Chittenden, 1982) and Cynoscion regalis (5 mo.) (Merriner, 1976). In Argentinian coastal waters, the main sciaenid associated with C. striatus is the white croaker ( Micropogonias furnieri), which has repro- ductive activity from November to March (Macchi and Christiansen, 1996). NOTE Macchi: Estimate of spawning frequency and batch fecundity of Cynoscion striatus 379 Because vitellogenic oocytes are recruited continu- ally during the reproductive season, it is necessary to determine the batch fecundity and the number of spawnings in the season in order to estimate total egg production (Hunter et al., 1985). Spawning fre- quency estimated from the percentage of hydrated ovaries (12%) was similar to those calculated with postovulatory follicles (11-13%). The average be- tween day-0 and day-1 POF’s ( 12 ±6%) indicates that spawning occurred once every 8 days during the main peak of the reproductive season (November). Daily spawning fraction of C. striatus was similar to that reported for three other scienids: Seriphus politus (De Martini and Fountain, 1981), Genyonemus linea- tus (Love et al., 1984) and Cynoscion nebulosus (Brown- Peterson et al., 1988), but was lower than estimates for Pogonias cromis (31%) (Fitzhugh et al., 1993). Annual spawning frequency estimated for Sciaenops ocellatus (Wilson and Nieland, 1994) and Cynoscion regalis (Lowere-Barbieri et al., 1996) varied widely, 3 to 80 days and 2 to 13 days, respectively. Daily spawning percentage of white croaker ( Micropogo - nias furnieri) in the Southwest Atlantic was 8.83% (Macchi et al., in press), indicating that the average interval between spawnings for this sciaenid is about 12 days . Cassia (1986) estimated total fecundity for C. striatus by counting the number of growing oocytes, which is inappropriate for a multiple spawning spe- cies (Hunter and Goldberg, 1980). This author esti- mated a total fecundity of 600,000 oocytes for one 40-cm-TL female, when the batch fecundity for that length is about 100,000 hydrated oocytes. In the present study, analysis of variance applied to the fe- cundity data from “El Rincon” and the Uruguayan coast indicated no statistical differences between these areas. Batch fecundity was a power function of length and a linear function of ovary-free body weight and ranged between 50,000 (34 cm TL) and 450,000 (53 cm TL) hydrated oocytes. Batch fecun- dity values estimated for striped weakfish were higher than those obtained for smaller sciaenids, such as Genyonemus lineatus (800-37,200 oocytes) (Love et al., 1984) and Seriphus politus (5,000-90,000 oocytes) (De Martini and Fountain, 1981). Sciaenids with length ranges similar to C. striatus, such as Cynoscion regalis (Lowere-Barbieri et al., 1996) and Cynoscion nebulosus (Brown-Peterson et al., 1988), show slightly higher batch fecundity values. The re- lations between batch fecundity vs. length and batch fecundity vs. weight for these species had relatively low coefficients of determination, similar to those for striped weakfish. The mean relative fecundity for C. striatus (210 ±35 oocytes/g ovary-free body weight) was less than that obtained for Cynoscion nebulosus (451 ±43 oocytes) (Brown-Peterson et al., 1988) and 380 Fishery Bulletin 96(2), 1998 Cy noscion regalis (200-750 oocytes) ( Lowere-Barbieri et al., 1996). Accurate annual fecundity estimations are diffi- cult to determine for multiple spawning fishes with an extended reproductive season (Brown-Peterson et al., 1988). During November-March, a female C. striatus would spawn 22 times with a spawning fre- quency of 8 days. Consequently, annual fecundity estimates for a 40.0-cm-TL female would be about 2.0 million eggs. However, annual egg production depends on spawning frequency, and it is possible that this parameter varies during the reproductive season and between years (Lowerre-Barbieri et al., 1996). The annual fecundity estimate for C. striatus was similar to that obtained for white croaker of the Rio de la Plata estuary ( Micropogonias furnieri) (1.8 million eggs for a 40.0 cm TL female) (Macchi et al., in press). Both species are among the most heavily ex- ploited coastal resources in Argentina and Uruguay. Acknowledgments This research was part of INIDEP’s Coastal Project. I thank Teresa Carle and Virginia Habegger for the preparations of histological sections. Literature cited Barbieri, L. R., M. E. Chittenden Jr., and S. K. Lowerre-Barbieri. 1994. Maturity, spawning, and ovarian cycle of Atlantic croaker, Micropogonias undulatus, in the Chesapeake Bay and adjacent coastal waters. Fish. Bull. 92:671-685. Brown-Peterson, N., P. Thomas, and C. R. Arnold. 1988. Reproductive biology of the spotted seatrout, Cy noscion nebulosus, in south Texas. Fish. Bull. 86( 2 ):373— 388. Cassia, M.C. 1986. Reproduccion y fecundidad de la pescadilla de red (Cynoscion striatus ). Publ. Com. Tec. Mix. Fr. Mar. 1( 1 ): 191—203. Ciechomski, J. D. de, and M. C. Cassia. 1978. Studies on the growth of juveniles of Cynoscion striatus in the sea and in aquaria. J. Fish Biol. 13:521-526. 1982. Observaciones sobre embriones, larvas y juveniles de la pescadilla Cynoscion striatus. Rev. Invest. Des. Pesq. INIDEP, Mar del Plata 3:5-13. Ciechomski, J. D. de, and M. D. Ehrlich. 1977. Alimentacion de juveniles de pescadilla Cynoscion striatus (Cuvier, 1829) Jordan & Evermann, 1889 en el mar y en condiciones experimentales. Pisces, Sciaenidae. Physis (sec. A) 37(93):1-12. Cioffi, C. A. 1992. Evaluacion de la pescadilla de red ( Cynoscion striatus) en la Zona Comun de Pesca Argentino-Uruguaya, en el otono de 1989. M.S. thesis, Universidad Nacional de Mar del Plata, Mar del Plata, Argentina, 40 p. Cordo, H. D. 1986. Estudios biologicos sobre peces costeros con datos de dos campanas de investigacion realizadas en 1981. Ill: La pescadilla de red ( Cynoscion striatus). Publ. Com. Tec. Mix. Fr. Mar. l(l):15-27. Cousseau, M. B., C. P. Cotrina, H. D. Cordo, and G. E. Burgos. 1986. Analisis de datos biologicos de corvina rubia ( Micro- pogonias furnieri) y pescadilla de red ( Cynoscion striatus) obtenidos en dos campanas del ano 1983. Publ. Com. Tec. Mix. Fr. Mar. l(2):319-332. DeMartini, E. E., and R. K. Fountain. 1981. Ovarian cycling frequency and batch fecundity in the queenfish, Seriphus politus: attributes representative of serial spawning fishes. Fish. Bull. 79(31:547-560. DeVries, D. A., and M. E. Chittenden Jr. 1982. Spawning, age determination, longevity, and mortal- ity of the silver seatrout, Cynoscion nothus , in the Gulf of Mexico. Fish. Bull. 80(3):487-500. Draper, N., and H. Smith. 1981. Applied regression analysis, 2nd ed. J. Wiley & Sons, New York, NY, 709 p. Fitzhugh, G. R., B. A. Thompson, and T. G. Snider III. 1993. Ovarian development, fecundity, and spawning fre- quency of black drum Pogonia cromis in Louisiana. Fish. Bull. 91:244-253. Forberg, K. G. 1982. A histological study of development of oocytes in cape- lin, Mallotus villosus villosus (Muller). J. Fish Biol. 20:143-154. Goldberg, S. R., V. H. Alarcon, and J. Alheith. 1984. Postovulatory follicle histology of the Pacific sardine, Sardinops sagax, from Peru. Fish. Bull. 82(2):443-445. Hunter, J. R., and S. R. Goldberg. 1980. Spawning incidence and batch fecundity in northern anchovy, Engraulis mordax. Fish. Bull. 77(3):641-652. Hunter, J. R., and B. J. Macewicz. 1985. Rates of atresia in the ovary of captive and wild north- ern anchovy, Engraulis mordax. Fish. Bull. 83:119-136. Hunter, J. R., N. C. H. Lo, and R. J. H. Leong. 1985. Batch fecundity in multiple spawning fishes. In R. M. Lasker (ed.), An egg production method for estimating spawning biomass of pelagics fish: application to the north- ern anchovy, Engraulis mordax, p. 67-77. U.S. Dep. Commer., NOAATech. Rep. NMFS 36. Hunter, J. R., B. J. Macewicz, N. C. H. Lo, and C. A Kimbrell. 1992. Fecundity, spawning, and maturity of female Dover sole Microstomus pad ficus, with an evaluation of assump- tions and precision. Fish. Bull. 90:101-128. Love, M. S., G. E. McGowen, W. Westphal, R. J. Lavenberg, and L. Martin. 1984. Aspects of the life history and fishery of the white croaker, Genyonemus lineatus (Sciaenidae), off California. Fish. Bull. 82:179-198. Lowerre-Barbieri, S. K., M. E. Chittenden Jr., and L. R. Barbieri. 1996. Variable spawning activity and annual fecundity of weakfish in Chesapeake Bay. Trans. Am. Fish. Soc. 125:532-545. Macchi, G. J., and H. E. Christiansen. 1996. Analisis temporal del proceso de maduracion y determinacion de la incidencia de atresias en la corvina rubia (Micropogonias furnieri). Frente Maritimo 16: 93-101. Macchi, G. J., E. M. Acha, and C. A. Lasta. In press. Desove y fecundidad de la corvina rubia (Micropogonias furnieri Desmarest, 1823) del estuario del Rio de la Plata, Argentina. Bol. Inst. Esp. Ocean. NOTE Macchi: Estimate of spawning frequency and batch fecundity of Cynoscion striatus 381 Mayer, I., S. E. Shackley, and J. S. Ryland. 1988. Aspects of the reproductive biology of the bass, Dicentrarchus labrax L.I. An histological and histochemi- cal study of oocyte development. J. Fish Biol. 33:609-622. Melo, Y. C. 1994. Spawning frequency of the anchovy Engraulis capensis. S. Afr. J. Mar. Sci. 14:321-331. Merriner, J. V. 1976. Aspects of the reproductive biology of the weakfish, Cynoscion regalis (Sciaenidae), in North Carolina. Fish. Bull. 74(1): 18-26. White, M. L., and M. E. Chittenden Jr. 1977. Age determination, reproduction, and population dynamics of the Atlantic croaker, Micropogonias undu- latus. Fish. Bull. 75(1):109-124. Wilson, C. A., and D. L. Nieland. 1994. Reproductive biology of red drum, Scienops ocellatus, from the neritic waters of the northern Gulf of Mexico. Fish. Bull. 92(4):841-850. Direct validation of ages determined for adult black drum, Pogonias cromis, in east-central Florida, with notes on black drum migration Michael D. Murphy Florida Marine Research Institute, Department of Environmental Protection 1 00 Eighth Avenue St- Petersburg, Florida 33701-5095 E-mail address: murphy_m@harpo. dep.state.fi. us Douglas H. Adams Derek SVL Tremain Florida Marine Research Institute, Department of Environmental Protection I 220 Prospect Avenue, Suite 285 Melbourne, Florida 32901 Brent L Winner Florida Marine Research Institute, Department of Environmental Protection 1 00 Eighth Avenue St. Petersburg, Florida 33701-5095 Findings from indirect age valida- tion studies, in which marginal in- crements in black drum, Pogonias cromis, otolith sections were exam- ined, indicate that the number of opaque bands observed in otoliths is a measure of the true age in years of black drum (Murphy and Taylor, 1989; Beckman et al., 1990; Beck- man et al.1). However, validation studies of scale-determined ages have indicated that more than one mark is deposited on black drum scales each year after they reach about age four (Richards, 1973; Cornelius, 1984). In addition, evi- dence from an analysis of the mar- ginal increments for a related sciaenid, the croaker Micropogonias opercularis, has indicated that more than one opaque band was depos- ited on its otoliths each year (Hai- movici, 1977). Unlike indirect validation stud- ies which provide estimates of the periodicity of annulus formation for a sample of fish, mark-recapture studies provide proof of the accu- racy of an age determination tech- nique for an individual fish during the period of time between mark- ing and recapture (Beamish and McFarlane, 1983). This direct tech- nique has been used to validate ages for a variety of species, includ- ing red drum, Sciaenops ocellatus (Murphy and Taylor, 1991; Ross et al., 1995); snapper, Pagrus auratus (Francis et al., 1992); sablefish, Anoplopoma fimbria (McFarlane and Beamish, 1995); and the tropi- cal parrotfish Scarus schlegeli (Lou, 1992). Because mark-recapture studies provide information on the validity of age marks formed since the mark was applied, they cannot be extrapolated to imply the valid- ity of younger ages (Francis, 1995). In assessing the status of black drum stocks, accurate age determi- nation is especially critical. These fish are suspected to live to about 60 years old, and the large fisher- ies that they have supported appar- ently exploited infrequent, large year classes (Jones et al., in press; Beckman et al.1). The purpose of this study was to validate directly the ages of adult black drum by using mark-recapture methods. The sample of recaptured fish also provided insight into the popula- tion dynamics of black drum in east-central Florida. Materials and methods We used a 549-m nylon trammel net (124- and 356-mm mesh, inside and outside stretch, respectively) to capture adult black drum in the southern Mosquito Lagoon and northern Banana River areas of the Indian River Lagoon system along Florida’s Atlantic coast (Fig. 1). Black drum were captured for tag- ging in February, June, and Sep- tember 1992, and in February, June, and August 1993. Each fish was measured for total length (mm), tagged with a 100-mm Hall- print plastic dart tag, injected in- tramuscularly with approximately 25 mg of oxytetracycline (OTC) per kilogram of body weight, and re- leased to the capture area. The dart tag was placed in the side of the fish about 25 mm ventral to the ante- rior insertion of the second dorsal fin. This placement allowed the barb to lodge behind a ptery- giophore while leaving the external message (streamer) visible. This external streamer bore an identifi- cation number and a message in- 1 Beckman, D. W., C. A. Wilson, D. L. Nieland, and A. L. Stanley. 1990. Age structure, growth rates, and reproductive biology of black drum in the northern Gulf of Mexico. Final report, U.S. Dep. Com- merce Cooperative Agreement NA89WC- H-MF017, Marine Fisheries Initiative (MARFIN) Program, 77 p. Manuscript accepted 7 August 1997. Fishery Bulletin 96:382-387 (1998) NOTE Murphy et al.: Direct validation of ages determined for Pogonias cromis 383 Total length (mm) Figure 2 Total length-frequency distribution (mm) for black drum that were tagged and released (line, n=707) and for those that were recap- tured (bars, n=26). of the sulcus acousticus with the aid of an Optimus digital-image processing system. Figure 1 Sampling area for black drum along the Atlantic coast of Florida. Most sam- pling was done in the area shown by diagonal lines just west of Cape Can- averal. Some sampling was also done in the southern portion of Mosquito Lagoon. structing anglers to contact us. The OTC provided a reference mark defining the otolith margin at the time of initial capture; all recaptures of fish whose otoliths were recovered were made by biologists dur- ing sampling trips conducted in August 1993 and January, October, and December 1995. Recaptured fish were sacrificed and returned to the laboratory, where they were measured for total, fork, and standard lengths; weighed to the nearest tenth of a kilogram; and sexed macroscopically. Sag- ittal otoliths were excised and stored dry. An Isomet low-speed saw was used to make transverse cuts near the core of the whole otolith. Otolith sections were about 0.5-mm thick and were mounted on glass slides with coverbond mounting media. Opaque bands and OTC marks were examined with a compound dissect- ing microscope (4x magnification) equipped with an ultraviolet light source. Distances from the otolith core to the proximal edge of the OTC mark and to each opaque band distal to the OTC mark were mea- sured along the axis from the core to the ventral edge Results and discussion A total of 707 adult black drum, from 515 to 1,237 mm TL, were captured, injected, and tagged during this study. Twenty-six of these fish were recaptured, and their otoliths were excised. Lengths of recaptured black drum ranged from 656 to 1,011 mm TL (Fig. 2). Most of the large (>1,050 mm TL) adult black drum found in our sample came from a school of very large fish caught in southern Mosquito Lagoon; according to two tag returns (see below), this school had apparently left the area shortly after tagging and release. Ages of recaptured black drum ranged from 3 to 13 years (Table 1); 16 were age 8 and older and were likely to have been mature adults at time of recap- ture (Murphy and Taylor, 1989). Time-at-large for recaptured fish was from just over a month to just less than three years, i.e. 42-1,031 days. Observations of the number of opaque bands de- posited since the fish had been marked supported our hypothesis that one band formed each year from the late winter to early spring. We observed that five fish tagged during February or June 1993 and recap- tured 42-176 days later in August 1993 had not depos- ited an opaque band since being marked (Table 1 ). One fish at large for 555 days between June 1993 and January 1995 had deposited one band since release. All other fish recaptured had been at large for 805 to 1,031 days, during which two late-winter to early- 384 Fishery Bulletin 96(2), 1 998 Table 1 Tag, recapture, and otolith-measurement data for OTC-marked black drum. Otolith measurements were made from the core to the proximal edge of the OTC band and all subsequent opaque bands to the otolith edge along the ventral margin of the sulcus acousticus. Tagged Recaptured Distance from otolith core (mm) Date Total length (mm) Date Total length (mm) Days free Age (yr) OTC mark Annulus Annulus Otolith edge Jun 93 660 Aug 93 656 42 3 2.33 2.35 Feb 93 836 Aug 93 838 176 5 2.75 2.86 Feb 93 975 Aug 93 945 176 11 3.99 4.08 Feb 93 925 Aug 93 921 176 11 3.85 3.90 Feb 93 940 Aug 93 926 176 11 3.79 3.86 Jun 93 732 Jan 95 790 555 4 2.39 2.65 2.81 Aug 93 832 Oct 95 887 805 13 3.56 3.62 3.74 3.93 Aug 93 899 Oct 95 968 805 9 3.44 3.50 3.68 3.87 Aug 93 922 Oct 95 961 805 13 3.75 3.81 3.95 4.11 Jun 93 711 Oct 95 826 847 5 1.74 1.83 2.04 2.31 Feb 93 975 Oct 95 1,011 991 13 3.61 3.75 3.94 4.13 Aug 93 821 Oct 95 859 805 13 3.73 3.77 3.93 4.09 Feb 93 873 Oct 95 886 981 9 3.00 3.16 3.29 3.51 Aug 93 925 Oct 95 952 805 13 4.03 4.08 4.23 4.40 Feb 93 673 Oct 95 842 981 6 2.50 2.63 2.86 3.10 Jun 93 Oct 95 903 847 5 2.44 2.55 2.81 3.03 Aug 93 832 Oct 95 843 805 13 3.77 3.80 3.96 4.11 Aug 93 941 Oct 95 943 805 13 3.95 4.00 4.17 4.33 Aug 93 852 Oct 95 886 805 13 3.98 4.02 4.15 4.29 Jun 93 620 Oct 95 771 847 5 2.37 2.55 2.83 3.06 Feb 93 880 Dec 95 965 1,031 9 3.30 3.45 3.67 3.92 Feb 93 892 Dec 95 936 1,031 13 3.78 3.90 4.03 4.20 Aug 93 782 Dec 95 899 855 6 2.99 3.08 3.40 3.67 Feb 93 790 Dec 95 911 1,031 7 2.79 2.96 3.18 3.44 Aug 93 789 Dec 95 917 855 8 3.24 3.28 3.53 3.76 Aug 93 690 Oct 95 830 806 5 2.55 2.64 2.91 3.24 spring periods of opaque band formation had oc- curred. All of the latter had deposited two bands on their otoliths (Fig. 3). Our direct validation for most ages from 3 years to 13 years indicates that for young adult black drum, at least, age can be determined accurately by count- ing saggital otolith annuli. Our results extend the findings of an indirect validation study that had used the periodicity of marginal increments as evidence for validation of ages for black drum up to age four in northeast Florida (Murphy and Taylor, 1989). In addition, the seasonal appearance of opaque margins on the otolith sections in older black drum between 20 and 37 years old in the northern Gulf of Mexico (Beckman et ah, 1990) is evidence for the validity of this aging technique in older fish. In contrast, the formation of two growth increments or marks per year on scales appears to begin at the time of matu- rity; formation of the second mark has been attrib- uted to spawning (Richards, 1973). All black drum larger than 700 mm FL (717 mm TL) and more than four years old that were sampled in the northern Gulf of Mexico (Nieland and Wilson, 1993), and all black drum larger than 650 mm TL and more than 5 years old sampled off northeast Florida ( Murphy and Tay- lor, 1989) were mature. Because our samples included adult fish of these sizes and ages, they were likely mature. Therefore black drum in our sample had clearly deposited only one opaque band on the sagittae each year for a number of years after they first matured. It seems likely then that, unless black drum begin to deposit multiple annuli each year af- ter they attain some age older than that found in our samples, their maximum age would be at least equal to the highest reported opaque band count, about 60 for black drum in U.S. Atlantic coastal wa- ters (Murphy and Taylor, 1989; Jones et al., in press). Our relatively small sample of validated ages for black drum lends support to previous findings of highly variable year-class strength for this species, and for the generally rapid growth of some fishes inhabiting the Indian River and Mosquito Lagoons. Eleven of the 26 black drum we recaptured were members of the 1982 year class, the oldest year class NOTE Murphy et al.: Direct validation of ages determined for Pogonias cromis 385 Figure 3 A black drum otolith section shown under reflected light only (top) and under reflected light and ultraviolet light ( bottom ). Opaque bands are num- bered consecutively from the core, and the fluorescing oxytetracycline ( OTC ) band is indicated. This fish was tagged and injected in August 1993 and recaptured 941 days later in October 1995. in our sample. Beckman et al.1 also noted the persistence of several large year classes (the 1979, 1974, 1970, and 1966 year classes) in the purse-seine fishery catch in the northern Gulf of Mexico, during 1986-89, and Jones et al. (in press) noted that the exception- ally large 1942 and 1934 year classes were still prominent in 1990-92 samples collected from Chesapeake Bay fisher- ies. Observed sizes at age of black drum in our sample, when released, were 620-732 mm TL for 3 year olds, 782- 836 mm for 4 year olds, and 789-880 mm for 6 year olds. These sizes at age were much larger than those predicted for northeast Florida black drum: 477, 555, 625, and 687 mm TL for ages 3-6 (Murphy and Taylor, 1989). Young red drum, Sciaenops ocellatus, and spotted seatrout, Cynoscion nebulosus, are also larger at age in the Indian River La- goon than in other areas of Florida (Murphy and Taylor, 1990, 1994). Limited tag-recapture information gathered during this study indicates that some larger black drum make ex- tensive migrations from the Indian River Lagoon system. Seven of the 9 angler-recaptured, tagged black drum (none of whose otoliths were recovered) were caught in the northern Indian River Lagoon within 10 km of their re- lease site. Two fish (1,025 and 1,163 mm TL), however, tagged and released in February 1992 travelled about 1,370 km north to Chesapeake Bay, where they were captured by anglers in late May and early June. Although the reproduc- tive states of these two tagged fish were not recorded at release, many of the males tagged along with them were run- ning ripe. The fact that these fish moved to Chesapeake Bay during the time pe- riod when peak spawning progresses up the Atlantic coast may imply that black drum con- tinue spawning as they move northward during the spring. This migration would enable them to broad- cast their average annual fecundity of 32—45 million eggs (Fitzhugh et al., 1993; Nieland and Wilson, 1993) over a wide range of suitable egg and larval habitats. Extensive tagging studies in Florida during the early and mid- 1960’s showed that, except for some of those released on beaches, black drum hardly moved from a release site (Topp, 1963; Beaumariage, 1964, 1969; Beaumariage and Wittich, 1966); even the greatest movement of fish along the beaches was less than 145 km. The total lengths of fish tagged during these studies were generally less than 500 mm. In Texas, Osburn and Matlock ( 1984) found substan- tial intrabay movement of small immature black drum, but little movement between bays. Music and Pafford (1984) also found that most black drum tagged in Georgia did not move far from the area of release. However, in Georgia 13% of all returned fish 386 Fishery Bulletin 96(2), 1998 had moved more than 100 km, reaching as far south as West Palm Beach, Florida (619 km), and as far north as Murrells Inlet, North Carolina (437 km). Surprisingly, the two black drum that had travelled the farthest from their release sites in Georgia were less than 350 mm TL. Because black drum are long-lived and capable of extensive movements, knowledge about the magni- tude and frequency of their mixing during their po- tential 60-year life span is critical for understand- ing how fishing pressure within one geographic area affects fish populations in other areas. In Florida, strict regulations on the harvest of black drum were implemented in 1989 following unvalidated evidence that they were long-lived fish (Murphy and Taylor, 1989). Because many current stock assessment tech- niques often depend on an accurate estimate of maxi- mum life span so that a natural mortality rate for a population can be determined, validation studies are especially critical in helping to prevent mismanagement of long-live species (Beamish and McFarlane, 1983). Acknowledgments We thank R. Muller, P. Hood, D. Leffler, and J. Leiby for comments made on earlier drafts of this report, and L. Brant for the map. This work was supported in part by funding from the Department of Com- merce, National Oceanographic and Atmospheric Administration, Marine Fisheries Initiative Award NA90AA-H-MF734. We wish to thank the Depart- ment of the Interior U.S. Fish and Wildlife Service Merritt Island National Wildlife Refuge and the National Aeronautics and Space Adminstration Kennedy Space Center for permission to work in their restricted areas. Literature cited Beamish, R. J., and G. A. McFarlane. 1983. The forgotten requirement for age validation in fish- eries biology. Trans. Am. Fish. Soc. 112:735-743. Beaumariage, D. S. 1964. Returns from the 1963 Schlitz tagging program. Fla. Board Conserv. Mar. Lab. Tech. Ser. 43, 34 p. 1969. Returns from the 1965 Schlitz tagging program in- cluding a cumulative analysis of previous results. Fla. Board Conserv. Mar. Lab. Tech. Ser. 59, 38 p. Beaumariage, D. S., and A. C. Wittich. 1966. Returns from the 1964 Schlitz tagging program. Fla. Board Conserv. Mar. Lab. Tech. Ser. 47, 50 p. Beckman, D. W., A. L. Stanley, J. H. Render, and C. A. Wilson. 1990. Age and growth of black drum in Louisiana waters of the Gulf of Mexico. Trans. Am. Fish. Soc. 119:537-544. Cornelius, S. E. 1984. Contribution to the life history of black drum and analysis of the commercial fishery of Baffin Bay. Vol. 2. Caesar Kleburg Wildlife Res. Institute, Tech. Bull. 6, 53 p. Fitzhugh, G. R., B. A. Thompson, and T. G. Snider III. 1993. Ovarian development, fecundity, and spawning fre- quency of black drum Pogonias cromis in Louisiana. Fish. Bull. 91:244-253. Francis, R. I. C. C. 1995. The analysis of otolith data — a mathematician’s per- spective (what precisely is your model?). In D. H, Secor, J. M. Dean, and S. E. Campana (eds.). Recent developments in fish otolith research, p. 81-95. Belle W. Baruch Library in Marine Science Number 19, Univ. South Carolina Press, Columbia, SC. Francis, R. I. C. C., L. J. Paul, and K. P. Mulligin. 1992. Ageing of adult snapper (Pagrus auratus) from otolith annual ring counts: validation by tagging and oxytetracy- cline injection. Aust. J. Mar. Freshwater Res. 43:1069- 1089. Haimovici, M. 1977. Age, growth and general aspects of the biology of the “white croaker,” Micropogon opercularis (Quoy et Gaimard, 1824) (Pisces, Sciaenidae). Atlantica, Rio Grande 2:21-49. Jones, C. M., M. E. Chittenden Jr., and B. Wells. In press. Age, growth, and mortality of black drum, Pogon- ias cromis, in the Chesapeake Bay. Fish. Bull. 96. Lou, D. C. 1992. Validation of annual growth bands in the otolith of tropical parrotfishes (Scams schlegeli Bleeker). J. Fish. Biol. 41:775-790. McFarlane, G. A., and R. J. Beamish. 1995. Validation of the otolith cross-section method of age determination for sablefish (Anoplopoma fimbria ) using oxytetracycline. In D. H, Secor, J. M. Dean, and S. E. Campana (eds.). Recent developments in fish otolith re- search, p. 319-329. Belle W. Baruch Library in Marine Science Number 19, Univ. South Carolina Press, Colum- bia, SC. Murphy, M. O., and R. G. Taylor. 1989. Reproduction and growth of black drum, Pogonias cromis, in northeast Florida. Northeast Gulf Sci. 10: 127-137. 1990. Reproduction, growth, and mortality of red drum, Sciaenops ocellatus, in Florida. Fish. Bull. 88:531-542. 1991. Direct validation of ages determined for adult red drums from otolith sections. Trans. Am. Fish. Soc. 120:267-269. 1994. Age, growth, and mortality of spotted seatrout in Florida waters. Trans. Am. Fish. Soc. 123:482-497. Music, J. L., and J. M. Pafford. 1984. Population dynamics and life history aspects of major marine sportfishes in Georgia’s coastal waters. Georgia Department of Natural Resources, Contribution Series 38, 382 p. Nieland, D. L., and C. A. Wilson. 1993. Reproductive biology and annual variation of repro- ductive variables of black drum in the northern Gulf of Mexico. Trans. Am. Fish. Soc. 122:318-327. Osburn, H. R., and G. C. Matlock. 1984. Black drum movement in Texas Bays. N. Am. J. Fish. Manage. 4:523-530. Richards, C. E. 1973. Age, growth, and distribution of the black drum (Pogo- nias cromis) in Virginia. Trans. Am. Fish. Soc. 102:584—590. NOTE Murphy et al.: Direct validation of ages determined for Pogonias cromis 387 Ross, J. L., T. M. Stevens, and D. S. Vaughan. Topp, R. W. 1995. Age, growth, mortaliity, and reproductive biology of 1963. Returns from the 1962 Schlitz tagging program. Fla. red drums in North Carolina waters. Trans. Am. Fish. Board Conserv. Mar. Lab. Prof. Pap. Ser. 5, 76 p. Soc. 124:37-54. 388 The influence of spear fishing on species composition and size of groupers on patch reefs in the upper Florida Keys Robert D„ Sluka* Kathleen M. Sullivan Department of Biology, University of Miami Coral Gables, Florida 33 1 24 and Florida and Caribbean Marine Conservation Science Center The Nature Conservancy PO. Box 249 1 1 8, Coral Gables, Florida 33 1 24 *Present address: Oceanographic Society of Maldives PO. Box 2075, Male, Republic of Maldives E-mail address (for K. M. Sullivan, contact author): sullivan@benthos.cox.mlami.edu Groupers are an important fishery resource throughout tropical and subtropical regions of the world (Heemstra and Randall, 1993). De- pendent and independent fishery surveys of grouper populations in south Florida and the Caribbean have shown drastic declines in populations, most likely due to in- tense fishing pressure (Sadovy, 1994; Bohnsack et ah, 1994). Com- bined grouper species landings by weight have declined in the Florida Keys by more than half since the mid 198Q’s (Bohnsack et al., 1994). Fishermen tend to target the larger fish in a population, with the result that a decrease in the den- sity, average size, and relative abundance of exploited species is inevitable (Bohnsack, 1982; Russ, 1985; Plan Development Team (PDT), 1990; Roberts and Polunin, 1991). Grouper populations are es- pecially sensitive to fishing pres- sure, exhibiting reductions in den- sity and average size, as well as shifts in species composition be- tween sites that are fished and those that are unfished (Russ, 1985). For example, the mean weight of three species of grouper was found to be significantly greater in unfished than in fished sites in the Red Sea (Roberts and Polunin, 1993). Similarly, the density and biomass of groupers was signifi- cantly greater in sites protected from fishing than in sites unpro- tected in the Philippines (Russ and Alcala, 1989). Craik (1981) found that a commercially important grouper ( Pleetropomus leopardus ) on the Great Barrier Reef had a higher mean size at an unfished reef than at a fished reef. Russ (1985) and Craik (1981) observed that larger individuals of a grouper species were abundant only at unfished reefs. Sluka et al. (1997) showed that the biomass, average size, and reproduction (the total number of eggs produced per hect- are) of Nassau grouper, Epinephelus striatus, was significantly greater inside a marine fishery reserve than outside. Many studies have attrib- uted a change in the relative abun- dance of grouper species to fishing pressure (Goeden, 1982; Bohnsack, 1982; Russ, 1985; Watson and Or- mond, 1994; Sluka, 1995). In the Florida Keys there has been a history of management mea- sures that have affected fishing pressure on groupers. In 1980, the state of Florida banned fish traps in its waters (<3 nautical miles), and in 1992 the United States gov- ernment banned fish traps in fed- eral waters of the Florida Keys (>3 nautical miles to the 150 fathom depth contour). During this study, there was a bag limit of five fish and a minimum size limit of 51 cm (20 in.) for six grouper species (red grouper, Epinephelus morio, black grouper, Mycteroperca bonaci, yellowmouth grouper, M. inter- stitialis, gag, M. microlepis, scamp, M. phenax, and yellowfin grouper, M. venenosa). Harvest of two spe- cies (jewfish, E. itajara, in 1990 and Nassau grouper, E. striatus, in 1991) has been prohibited in the south Atlantic waters of the United States. Reefs within Key Largo Na- tional Marine Sanctuary (KLNMS) and John Pennekamp Coral Reef State Park (JPCRSP) have been protected from spear fishing since 1960. Since the time of this study, the entire Florida Keys has come under the management of the Florida Keys National Marine Sanctuary. One of the results of the previ- ous management scheme is that the upper Florida Keys can be divided into two areas: one area protected from spear fishing and the other unprotected from spear fishing. It is assumed that the intensity of hook-and-line fishing is similar in- side and outside of KLNMS and JPCRSP. Thus the goal of this study was to examine the influence of spear fishing on the size and spe- cies composition of groupers in the upper Florida Keys and to discuss potential implications for the man- agement of these populations. These population parameters were Manuscript accepted 19 August 1997. Fishery Bulletin 96:388-392 (1998). NOTE Sluka and Sullivan. Spear fishing on species composition and size of groupers on reefs 389 expected to differ between sites under differing fish- ing intensities owing to the biology of the species, their site-attached nature, and the susceptibility of these species to fishing pressure. Materials and methods Sampling on four patch (shallow-water, small) reefs occurred in February 1992, April 1993, September 1993, January 1994, April 1994, and September 1994. Nine additional patch reefs were sampled in Septem- ber 1994. A total of 13 patch reef sites were sampled (Fig. 1). There were no significant seasonal differences in the size distribution of all grouper species combined at these sites (Sluka and Sullivan, 1996). Thus, data from all seasons were combined for analyses. Researchers were trained to estimate lengths of fish consistently and accurately using methods out- lined in Bell et al. (1985). Observers had approxi- mately five minutes to estimate the length of a series of fish models of varying lengths. The length of each model was recorded in one of five categories: <5 cm, 5-15 cm, 16-25 cm, 26-35 cm, and >35 cm. The fre- quency distribution of estimated model lengths was compared to the known distribution by using a chi- square test. The bias of each observer was determined as either consistently underestimating or overestimat- ing the size of the fish models. The information on bi- ases was given to each observer. The observer then re- peated the length estimation procedure until there was no significant difference between the observed and ex- pected distributions (P>0.05). Observers were found to be competent for length estimation after 2-3 trials. At four of the sites (MPR, TS1, TS2, and HOP), transect lines of 20 m or 25 m in length were used to sample the number, species, and length of groupers. The transect lines were laid in representative por- tions of each patch reef. The transect line was searched 6 m out from each side for a total width of 12 m (width by visual estimation). Within each transect all groupers were enumerated and their length category and species recorded on underwater paper. Observers using SCUBA searched through- out the width of the transect, examining all crevices, caves, and holes. At the other nine sites, observers sampled the entire patch reef. Patch reef size was not quantified at these sites; thus density could not 390 Fishery Bulletin 96(2), 1 998 be compared between protected sites and those un- protected from spear fishing. It was hypothesized that spear fishermen were targeting some species more than others owing to their larger size. Targeted species were assumed to be red grouper, black grouper, scamp, and gag. The latter two species were rare, constituting <14% of individuals observed by site. Nassau grouper were assumed to be a nontargeted species because of the ban on harvest. Differences in the relative abundance of targeted grouper were assessed with a £-test on the arcsine-transformed percentage of the total num- ber of grouper observed at a site. Individual sites were the replicates for the analysis. It was also hypothesized that the average size of targeted species would be significantly different be- tween protected sites and those unprotected from spear fishing. A nested ANOVA was used to test this hypothesis, with the main factor being protection level (spear fished or protected) and sites nested within protection level. The mid points of size cat- egories were used as an estimate of the individual sizes of sampled fish. Thus, for this analysis, indi- vidual fish sizes were the replicates. A value of 40 cm was used for individuals in the >35 cm category. This results in a conservative analysis because many of the fish in this category were much larger than 40 cm (R. Sluka, personal observation). Results There was no significant influence of spear fishing on the relative abundance of targeted grouper spe- cies (£=0.658, P>0.55). Targeted species constituted an average of 65% (rc=9) and 71% (n=4) on protected and unprotected sites, respectively. Nontargeted spe- cies (graysby, Epinephelus cruentatus, coney, E. fulvus, rock hind, E. adscensionis, red hind, E. guttatus, and Nassau grouper, E. striatus) constituted on average 35% and 29% of the total number of indi- viduals on protected and unprotected sites, respec- tively. Nassau grouper constituted a greater propor- tion of the individuals observed at protected (15%) than at unprotected (2%) sites. These values were statistically different (£=2.46, P<0.05). Patch reefs protected from spear fishing had a sig- nificantly greater size of targeted individuals than did unprotected patch reefs (F, ,)n=3.874, P= 0.05). The mean size (+/- 1 SE) of targeted grouper species on protected patch reefs was 29.3 (0.7) cm, whereas the mean size of targeted species on unprotected sites was 26.0 (0.8) cm. There was also a significant dif- ference among sites nested within protection levels (Fn 2n=5- 151, P cO.OOD. Discussion In this study, spear fishing appears to primarily in- fluence the average size of groupers. Sites in which spear fishing was not allowed had grouper assem- blages that were characterized by larger-size indi- viduals. This result is similar to those found in stud- ies examining the effects of hook-and-line fishing, where fishermen target the larger individuals in a population and thus decrease the average size of a fish species (Roberts and Polunin, 1991). Overall, grouper species composition was not significantly influenced by the presence or absence of spear fish- ing; targeted species were similarly abundant on protected and unprotected patch reefs. Sluka and Sullivan (1996) have shown that grouper species, such as black grouper, red grouper, and Nassau grou- per, are more abundant, but smaller, on inshore patch reefs than on offshore bank reefs of the upper Florida Keys. It is likely that species such as these are re- cruiting inshore before they move offshore (Ross and Moser, 1995). Nassau grouper, however, were more abundant on patch reefs protected from spear fish- ing than on unprotected patch reefs. Although there is a ban on harvesting these species, spear fisher- men may still collect individuals. It is important to examine how effective the ban on harvesting this species is in the Florida Keys. Offshore bank reef sites in the upper Florida Keys protected from spear fishing had snappers (Lutjani- dae) and grunts (Haemulidae) of larger size and greater abundance than did a lower Keys site sub- jected to spear fishing (Bohnsack, 1982). Clark et al. (1989) found similar results when sites inside Looe Key National Marine Sanctuary (lower Florida Keys) were compared before and after protection from spear fishing. The present study did not examine differ- ences in abundance between protection levels. How- ever, it is expected that there would be no signifi- cant differences in abundance between these sites; this result is due to the nature of both spear fishing and grouper growth and reproductive characteristics. Spear fishing targets the largest fish in an assem- blage but only accounts for a small percentage of the total fishing effort (PDT, 1990). For example, spear fishing accounted for 10.5% of the total recreational fishing catch in Biscayne National Park.1 Thus the magnitude of the selection pressure is likely much less than that from hook-and-line fishing because of 1 Tilmant, J. T., and R. Stone. 1984. Reef fish harvest trends, Biscayne National Park, Dade County Florida. Unpublished report presented to the 1984 stock assessment workshop. South- east Fisheries Center, National Marine Fisheries Service, Mi- ami, Florida, 26 p. NOTE Sluka and Sullivan: Spear fishing on species composition and size of groupers on reefs 391 the smaller number of spear fishermen. However, the lower number of spear fishermen may exert a more directed selectional effect on grouper species because they can be highly selective for the largest individuals. The growth and reproductive characteristics of groupers render these species especially susceptible to overfishing (Bannerot et al., 1987; Shapiro, 1987; Huntsman and Schaaf, 1994). Groupers that are tar- geted by fishing grow slowly to a large maximum size (Manooch, 1987). The removal of larger individuals leaves behind smaller individuals to spawn. Over many generations, this can result in a decrease in the size and age at sexual maturity (Ricker, 1981) and also decrease the average size of the population (Roberts and Polunin, 1991). Many grouper species are protogynous hermaphrodites, changing sex from females to males later in life (Shapiro, 1987). Larger groupers are generally males, and at intensive fish- ing levels, the number of males in the population can be drastically reduced. If too many males are re- moved, sperm are reduced for reproduction (Bannerot et al., 1987). If sperm are reduced, protogynous stocks are more vulnerable to overfishing than are gono- choristic stocks (Huntsman and Schaaf, 1994). Spe- cies that are protogynous may experience a drastic reduction in reproductive capacity, even at moder- ate levels of fishing (Huntsman and Schaaf, 1994). However, there may be mechanisms by which the population can compensate for a changing sex ratio in the presence of overfishing (Claro et al., 1990; Huntsman and Schaaf, 1994). Huntsman and Schaaf (1994) showed that these types of compensation mechanisms can reduce the detrimental impacts of fishing pressure on protogynous species. In addition to protogyny, the reproductive behavior of groupers may increase their susceptibility to overfishing. Many species of grouper aggregate to spawn during one or two months of the year (Smith, 1972; Shapiro, 1987; Claro et al., 1990). These spawning aggregations are subject to intense fishing pressure (Olsen and LaPlace, 1978, Claro et al., 1990; Sadovy, 1994). In many parts of the Caribbean, aggregations have dis- appeared as a result of overfishing (Sadovy, 1994). Sluka et al. (1997) showed that reproduction (the total number of eggs produced per hectare) was six times greater in a marine fishery reserve in the cen- tral Bahamas than in the surrounding unprotected region. It is concluded that the ban on spear fishing in the upper Florida Keys has significantly benefitted the size distribution of groupers. However, it appears that a ban on spear fishing alone has not resulted in re- covering population levels of grouper in this region. Bohnsack et al. (1994) has clearly shown a decline in commercially and recreationally targeted grouper landings throughout the Florida Keys. Sluka and Sullivan (1996) have shown that the offshore grou- per assemblage over shallow bank reefs of the upper Florida Keys is dominated numerically by graysby (88%), a small, nontargeted species. Densities of tar- geted species are low compared with regions where fishing was prohibited or had not yet taken place (Sluka and Reichenbach, 1996; Sluka et al., 1997). It is recommended that marine fishery reserves be considered as a management measure because of their success in other regions and because of a strong theoretical basis (PDT, 1990). Grouper assemblages inside marine fishery reserves are more dense, of greater average size, and produce greater numbers of eggs per hectare than in similar, unfished sites (Bohnsack, 1982; Russ and Alcala, 1989, Roberts and Polunin, 1993; Watson and Ormond, 1994; Sluka et al.,1997). Although management measures, such as bans on spear fishing, have some beneficial effects, the available evidence suggests that the establishment of marine fishery reserves is the most successful method for restoring and conserving grouper assemblages. Acknowledgments The authors would like to thank the staff of the Na- tional Undersea Research Program Florida Keys Program for logistical support. Fieldwork was as- sisted by M. Chiappone, T. Potts, G. Meester, and J. Levy. Figure 1 was prepared by R. Wright. This re- search was funded by NOAA’s National Undersea Research Program under NURC/UNCW grant UNCW-9420 to K.M. Sullivan and was conducted in the Florida Keys under National Marine Sanctuary Permit 93-27. Support was also obtained from Uni- versity of Miami (Department of Biology) and The Nature Conservancy’s Florida Keys Initiative and Caribbean program. Research was conducted as part of a doctoral dissertation completed by R. Sluka at the University of Miami. This manuscript benefit- ted significantly from readings by J. Bohnsack, S. Bolden, N. Ehrhardt, J. Prince, C. R. Robins, and three anonymous reviewers. Literature cited Bannerot, S., W. W. Fox Jr., and J. E. Powers. 1987. Reproductive strategies and the management of snap- pers and groupers. In J. J. Polovina and S. Ralston (eds.), Tropical snappers and groupers: biology and fisheries man- agement, p. 561-603. Westview Press Inc., Boulder, CO. Bell, J. O., G. J. S. Craik, D. A. Pollard, and B. C. Russel. 1985. Estimating length frequency distributions of large reef fish underwater. Coral Reefs 4:41—44. 392 Fishery Bulletin 96(2), 1 998 Bohnsack, J. A. 1982. Effects of piscivorous predator removal on coral reef fish community structure. In G. M. Cailliet and C. A. Simenstad (eds.), Gutshop ’81: fish food habits studies, p. 258-267. Wash. Sea Grant Publ., Seattle, WA. Bohnsack, J. A., D. E. Harper, and D. B. McClellan. 1994. Fisheries trends from Monroe County, Florida. Bull. Mar. Sci. 54:982-1018. Clark, J. R., B. Causey, and J. A. Bohnsack. 1989. Benefits from coral reef protection: Looe Key reef, Florida. Coastal zone ’89. Proc. 6th Symp. Coastal Ocean Manage. 4:3076-3086. Claro, R., A. Garcia Cagide, L. M. Sierra, and J. P Garia Arteaga. 1990. Caracteristicas biologico pesqueras de la cherna criolla, Epinephelus striatus (Bloch) (Pisces: Serranidae) en la plataforma cubana. Biologia Marina 23:23-43. Craik, G. J. S. 1981. Underwater survey of coral trout Plectropomus leopardus (Serranidae) populations in the Capricorn sec- tion of the Great Barrier Reef Marine Park. Proc. 4th Int. Coral Reef Symp. 1:53-58. Goedem, G. B. 1982. Intensive fishing and a ‘keystone’ predator species: ingredients for community instability. Biol. Conserv. 22:273-281. Heemstra, P. C., and J. E. Randall. 1993. FAO species catalogue. Vol. 16: Groupers of the world (family Serranidae, subfamily Epinephelinae): an anno- tated and illustrated catalogue of grouper, rockcod, hind, coral grouper and lyretail species known to date. FAO Fisheries Synopsis 125, FAO, Rome, 383 p. Huntsman, G. R., and W. E. Schaaf. 1994. Simulation of the impact of fishing on reproduction of a protogynous grouper, the graysby. N. Am. J. Fish. Manage. 14:41-52. Mamooch, C. S., III. 1987. Age and growth of snappers and groupers. In J. J. Polovina and S. Ralston (eds.), Tropical snappers and grou- pers: biology and fisheries management, p. 329-373. Westview Press Inc., Boulder, CO. Olsen, B. A., and J. A. LaPlace. 1978. A study of a Virgin Island grouper fishery based on a breeding aggregation. Proc. Gulf. Caribb. Fish. Inst. 31:130-144. Plan Development Team (PDT). 1990. The potential of marine fishery reserves for reef fish management in the U.S. Southern Atlantic. U.S. Dep. Commer., NOAATech. Memo. NMFS SEFC 261, 40 p. Ricker, W. E. 1981. Changes in the average size and average age of Pa- cific salmon. Can. J. Fish. Aquat. Sci. 38:1636-1656. Roberts, C. M., and N. V. C. Polumin. 1991. Are marine reserves effective in management of reef fisheries? Reviews in Fish Biology and Fisheries 1:65-91. 1993. Marine reserves: simple solutions to managing com- plex fisheries? Ambio 22:363-368. Ross, S. W., and M. L. Moser. 1995. Life history of juvenile gag, Mycteroperca microlepis, in N.C. estuaries. Bull. Mar. Sci. 56:222-237. Russ, G. 1985. Effects of protective management on coral reef fishes in the central Philippines. Proc. 5th Int. Coral Reef Symp. 4:219-224 Russ, G. R., and A. C. Alcala. 1989. Effects of intense fishing pressure on an assemblage of coral reef fishes. Mar. Ecol. Prog. Ser. 56:13-27. Sadovy, Y. 1994. Grouper stocks of the Western Central Atlantic: The need for management and management needs. Proc. Gulf Caribb. Fish. Inst. 43:43-64. Shapiro, B. Y. 1987. Reproduction in groupers. In J. J. Polovina and S. Ralston (eds.), Tropical snappers and groupers: biology and fisheries management, p. 295-327. Westview Press Inc., Boulder, CO. Sluka, R. D. 1995. The influence of habitat on density, species richness and size distribution of groupers in the upper Florida Keys, USA and central Bahamas. Ph.D. diss., Univ. Miami, Coral Gables, FL, 229 p. Sluka, R. D., and N. Reichenbaeh. 1996. Grouper density and diversity at two sites in the Republic of Maldives. Atoll Res. Bull. 438:1-16. Sluka, R. D., and K. M. Sullivan. 1996. The influence of habitat on the size distribution of groupers in the upper Florida Keys. Env. Biol. Fishes 47:177-189. Sluka, R. B., M. Chiappone, K. M. Sullivan, and R. Wright. 1997. The benefits of marine fishery reserve status for Nassau grouper Epinephelus striatus in the central Bahamas. Proc. 8th Int. Coral Reef Symp. 2:1961-1964. Smith, C. L. 1972. A spawning aggregation of Nassau grouper Epin- ephelus striatus. Trans. Am. Fish. Soc. 2:257-261. Watson, M., and R. F. G. Ormond. 1994, Effect of an artisanal fishery on the fish and urchin populations of a Kenyan coral reef. Mar. Ecol. Prog. Ser. 109:115-129. 394 Fishery Bulletin 96(2), 1 998 Superintendent of Documents Publications Order Form *5178 □yes, please send me the following publications: subscriptions to Fishery Bulletin for $34.00 per year ($42.50 foreign) The total cost of my order is $ . Prices include regular domestic postage and handling and are subject to change. (Company or Personal Name) (Please type or print) (Additional address/attention line) (Street address) (City, State, ZIP Code) Charge your order. It’s Easy! (Daytime phone including area code) (Purchase Order No.) Please Choose Method of Payment: I | Check Payable to the Superintendent of Documents | ] GPO Deposit Account | | | [ | | [ ~| — Q j I ] VISA or MasterCard Account To fax your orders (202) 512-2250 (Credit card expiration date) (Authorizing Signature) Mail To: Superintendent of Documents P.O. Box 371954, Pittsburgh, PA 15250-7954 Thank you for your order! Fishery Bulletin Guide for Contributors Content Articles published in Fishery Bulletin de- scribe original research in fishery marine science, engineering and economics, and the environmental and ecological sciences, in- "cluding modeling. Articles may range from relatively short to extensive; notes are re- ports of 5 to 10 pages without an abstract and describing methods or results not sup- ported by a large body of data. Although all contributions are subject to peer review, responsibility for the contents of papers rests upon the authors and not upon the editor or the publisher. It is therefore im- portant that the contents of the manuscript are carefully considered by the authors. Submission of an article is understood to imply that the article is original and is not being considered for publication elsewhere. Manuscripts should be written in English. Authors whose native language is not En- glish are strongly advised to have their manuscripts checked by English-speaking colleagues prior to submission. Preparation Title page should include authors’ full names and mailing addresses, the corre- sponding author’s telephone, FAX number, and E-mail address, and a list of key words to describe the contents of the manuscript. Abstract should not exceed one double- spaced typed page. It should state the main scope of the research but emphasize its con- clusions and relevant findings. Because ab- stracts are circulated by abstracting agencies, it is important that they represent the research clearly and concisely. Text must be typed double-spaced through- out. A brief introduction should portray the broad significance of the paper; the remain- der of the paper should be divided into the following sections: Materials and meth- ods, Results, Discussion (or Conclusions), and Acknowledgments. Headings within each section must be short, reflect a logical sequence, and follow the rules of multiple subdivision (i.e. there can be no subdivision without at least two items). The entire text should be intelligible to interdisciplinary readers; therefore, all acronyms, abbrevia- tions, and technical terms should be spelled out the first time they are mentioned. The scientific names of species must be written out the first time they are mentioned; sub- sequent mention of scientific names may be abbreviated. Follow the U.S. Government Printing Office Style Manual (1984 ed.) and the CBE Scientific Style and Format (6th ed.) for editorial style, and the most current issue of the American Fisheries Society’s Common and Scientific Names of Fishes from the United States and Canada for fish nomenclature. Dates should be written as follows: 11 November 1991. Measurements should be expressed in metric units, e.g., metric tons as (t); if other units of measure- ment are used, please make this fact explicit to the reader. The numeral one (1) should be typed as a one, not as a lower-case el (1). Use of appendices is discouraged. Text footnotes should be numbered with Arabic numerals. Footnote all personal communications, unpublished data, and un- published manuscripts with full address of the communicator or author, or, as in the case of unpublished data, where the data are on file. Authors are advised to avoid ref- erences to nonstandard (gray) literature, such as internal, project, processed, or ad- ministrative reports. Where these refer- ences are used, please include whether they are available from NTIS (National Techni- cal Information Service) or from some other public depository. Literature cited comprises published works and those accepted for publication in peer- reviewed literature (in press). Follow the name and year system for citation format. In the text, cite as follows: Smith and Jones (1977) or (Smith and Jones, 1977). If there is a sequence of citations, list by year: Smith, 1932; Smith and Jones, 1985; Smith and Allen, 1986. Abbreviations of serials should conform to abbreviations given in Serial Sources for the BIOSIS Previews Database. Authors are responsible for the accuracy and completeness of all citations. Tables should not be excessive in size and must be cited in numerical order in the text. Headings should be short but ample enough to allow the table to be intelligible on its own. Adi unusual symbols must be ex- plained in the table legend. Other inciden- tal comments may be footnoted with italic numerals. Use the asterisk only to indicate probability in statistical data. Because ta- bles are typeset, they need only be submit- ted typed and formatted, with double-spaced legends. Zeros should precede all decimal points for values less than one. Figures must be cited in numerical order in the text. The senior author’s last name and the figure number should be written on the back of each one. Hand-drawn illustrations should be submitted as originals and not as photocopies. Submit photographs as glossy prints or slides that show good contrast, otherwise we cannot guarantee a good final printed copy. Graphic illustrations should be submitted as laser-printed copies, not as photocopies. Label all figures with Helve- tica typeface and capitalize the first letter of the first word in axis labels. Do not use boldface. Italicize species name and vari- ables in equations. Use zeros before all deci- mal points. Use uppercase Times Roman bold typeface to label the parts of a figure, e.g. A, B, C, etc. Send original figures to the Scientific Editor when the manuscript has been accepted for publication. Each figure legend should explain all symbols and ab- breviations in the figure and should be dou- ble-spaced and placed at the end of the manuscript. Copyright law does not cover government publications; they fall within the public do- main. If an author reproduces any part of a government publication in his work, refer- ence to source is appreciated. Submission Send printed copies (original and three cop- ies without staples) to the Scientific Editor: Dr. John B. Pearce, Scientific Editor Northeast Fisheries Science Center National Marine Fisheries Service 166 Water Street Woods Hole, MA 02543-1097 Once the manuscript has been accepted for publication, you will be asked to submit a software copy of your manuscript to the Man- aging Editor. The software copy should be submitted in WordPerfect or Microsoft Word text format and placed on a 3.5-inch disk that is double-sided, double or high density, and that is compatible with either DOS or Apple Macintosh systems. A copy of page proofs will be sent to the au- thor for final approval prior to publication. Copies of published articles and notes are available free of charge to the senior author (50 copies) and to his or her laboratory (50 copies). Additional copies may be purchased in lots of 100 when the author receives page proofs. ☆ U.S. Gov. Printing Office 1998 689015/80000 Heckman BINDERY, INC. Bound-To-Please® FEB 00 N. MANCHESTER, INDIANA 46962