CALIFORNIA FISH™ GAME "CONSERVATION OF WILDLIFE THROUGH EDUCATION" California Fish and Game is a journal devoted to the conservation of wild- life. If its contents are reproduced elsewhere, the authors and the California Department of Fish and Game would appreciate being acknowledged. Subscriptions may be obtained at the rate of $5 per year by placing an order with the California Department of Fish and Game, 1416 Ninth Street, Sacramento, California 95814. Money orders and checks should be made out to California Department of Fish and Game. Inquiries regarding paid sub- scriptions should be directed to the Editor. Complimentary subscriptions are granted, on a limited basis, to libraries, scientific and educational institutions, conservation agencies, and on exchange. Complimentary subscriptions must be renewed annually by returning the post- card enclosed with each October issue. Please direct correspondence to: Kenneth A. Hashagen, Jr., Editor California Fish and Game 1416 Ninth Street Sacramento, California 95814 u I] VOLUME 64 JULY 1978 NUMBER 3 Published Quarterly by STATE OF CALIFORNIA THE RESOURCES AGENCY DEPARTMENT OF FISH AND GAME —LDA— CALH IRCES AGENCY LIBRARY Resources Building. Room tl7 1418 9th Street STATE OF CALIFORNIA EDMUND G. BROWN JR., Governor THE RESOURCES AGENCY HUEY D JOHNSON, Secretary for Resources FISH AND GAME COMMISSION BERGER C. BENSON, President San Mateo SHERMAN CHICKERING, Vice President ABEL GALLETTI, Member San Francisco Rancho Palos Verdes RAYMOND DASMANN, Member ELIZABETH L. VENRICK, Member Santa Cruz Cardiff-by-the-Sea DEPARTMENT OF FISH AND GAME E. C. FULLERTON, Director 1416 9th Street Sacramento 95814 CALIFORNIA FISH AND GAME Editorial Staff KENNETH A. HASHAGEN, JR., Editor-in-Chief Sacramento DARLENE A. OSBORNE, Editor tor Inland Fisheries Sacramento RONALD M. JUREK, Editor for Wildlife Sacramento J. R. RAYMOND ALLY, Editor for Marine Resources Long Beach DAVID A. HOOPAUGH, Editor for Salmon and Steelhead Sacramento DONALD E. STEVENS, Editor for Striped Bass, Sturgeon, and Shad Stockton CONTENTS 137 Page Taxonomy and Distribution of the Bull Trout, Salvelinus confluentus (Suckley), from the American Northwest Ted M. Cavender 139 Reproduction and Spawning of the Northern Anchovy, Engraulis mordax, in San Pedro Bay, California Gary D. Brewer 175 Hooking Mortality of Juvenile Largemouth Bass, Micropterus salmoides Ronald J. Pelzman 185 Catch-per-Unit-of-Effort Studies of Northern California Dungeness Crabs, Cancer magister Daniel W. Gotshall 189 Sex Ratios of the Northern Anchovy, Engraulis mordax, Off Southern California Richard A. Klingbeil 200 The Origins of Rainbow Trout, Salmo gairdneri Richardson, in New Zealand. D. Scott, J. Hewitson, and J. C. Fraser 210 NOTES The First Eastern Pacific Records of Bulleye, Cookeolus boops, (Bloch and Schneider, 1801) (Pisces, Priacanthidae) Ronald A. Fritzsche 219 A Hermaphroditic California Halibut, Paralichthys califomicus Jack W. Schott 221 Spinal Column Deformity in a Pile Surfperch, Damalichthys vacca Robert N. Tasto 223 A Note on Production Modeling of Populations with Discon- tinuous Reproduction Alec MacCall 225 Observations of Agonistic Behavior in the Treefish, Sebastes serriceps (Scorpaenidae) Peter L. Haaker 227 REVIEWS ERRATUM Smith, Gary E., 1978. An evaluation of disk-dangler tag shedding by striped bass (Morone saxatilis) in the Sacramento-San Joaquin Estuary. California Fish and Game, 64(2):93-97. Page 94: The equation printed as: log [A+B N, / (A^B Nj + b-aH)] should read: log [A.BN;/ (A.B N, + 8_A N,)] = log (1-AK) + i log (1-AS) Page 95: The last sentence of Figure 1's caption (Numbers in parentheses indicate number of fish from which tags were obtained.) should be deleted. 138 CHANGE OF EDITORSHIP With this issue Kenneth A. Hashagen, Jr. of Inland Fisheries Branch assumes the duties of Editor-in-Chief of California Fish and Game. Mr. Hashagen's assumption of the editorship follows the Depart- ment's policy of rotating the editorship between staff members repre- senting Marine Resources, Inland Fisheries, and Wildlife Management. For 7 years Mr. Hashagen, Senior Fishery Biologist, has served as Inland Fisheries Editor of the Quarterly. Through this service he has gained a knowledge of editorial policies and procedures of the Journal. Under his guidance, the Journal will continue its policy of presenting to the public the results of scientific investigations as they relate to management programs and the conservation of California fish and wild- life resources. Mr. Hashagen will be ably assisted in his duties by five associate editors: Darlene A. Osborne, Inland Fisheries; Ronald M. Jurek, Wildlife Management; J. R. Raymond Ally, Marine Resources; Donald E. Stevens, Anadromous Fisheries; and David A. Hoopaugh, Salmon and Steelhead. To Mr. Collins, Editor-in-Chief the past 4 years, we wish to express our appreciation for a job well done. £ C. Fullerton, Director, California Department of Fish and Game. TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 139 Calif. Fish and Came 64 (3 ) : 1 39-1 74. 1 978 TAXONOMY AND DISTRIBUTION OF THE BULL TROUT, SALVELINUS CONFLUENTUS (SUCKLEY), FROM THE AMERICAN NORTHWEST1 by Ted M. Cavender Museum of Zoology The Ohio State University Morphological and distributional evidence is presented favoring the specific dis- tinction of the bull trout, Salvelinus confluentus (Suckley ), a currently unrecognized form of Salvelinus native to western North America. This species has been confused with the Dolly Varden, Salvelinus malma (Walbaum). Separation of the two is based primarily on characters of the head and cranial skeleton. Diagnosis and description are given for the bull trout along with a history of its early taxonomy. Past and present distribution of the bull trout ranges between lat 41° and 60° N. North of the 49th parallel it is found in most of the major drainages originating on both sides of the Continental Divide. The bull trout is or at one time was sympatric with the Dolly Varden in at least three major river systems, as well as the waters of Puget Sound. INTRODUCTION The salmonid genus Salvelinus has long been recognized as a difficult taxo- nomic group. Without exaggeration, Vladykov (1954) referred to the taxonomy of Salvelinus (along with other salmonid genera) as "extremely involved and time-consuming." This is particularly true for species which are native to areas bordering the North Pacific Ocean. McPhail (1961) referred to these popula- tions of Salvelinus as part of the Salvelinus alpinus (Linnaeus) (arctic char) complex because of their similarity, the unsatisfactory state of their taxonomy, and the incompleteness of representative material. The possible existence of more than one species of Salvelinus in the American northwest has been a controversial subject since the days of the Pacific Railroad surveys in the 1850's. Morton (1970) was the last to deal with the subject. He concluded that only one species, Salvelinus malma (Walbaum) was recogniza- ble, while none of the proposed subspecies was valid. However, in this paper I will present morphometric, meristic, osteological, and distributional evidence to show that there are two widely distributed forms of Salvelinus native to the western United States and Canada; the Dolly Varden, S. malma, and the bull trout, S. confluentus. Although primarily an inland species, collection records show that the bull trout is not strictly an interior, nonanadromous form, but combines coastal and inland, as well as northern and southern aspects to its distribution. Because of the taxonomic difficulties with this species, the bull trout has lacked uniform scientific recognition even though it is a well-known sport fish. The scientific literature, especially compendia of regional fish faunas, has for so long lumped information about their ecology, morphology, and life history, that S. malma and S. confluentus have been largely confused. If there are morphological characters and aspects of their biology that consistently separate S. malma and S. confluen- tus over their distributional ranges, why then has there been a taxonomic prob- 1 Accepted for publication September 1977. 140 CALIFORNIA FISH AND GAME lem? The answer is due primarily to the lack of adequate comparative material available to any one investigator. The findings in this paper are part of a more extensive systematic treatment of the genus Salvelinus I began at the University of Michigan in 1968 and have continued at Ohio State University. Conclusions on the specific distinction of the bull trout reached here are essentially those presented in a preliminary report at the 1969 Annual Meeting of the American Society of Ichthyologists and Herpetologists in New York City (Cavender 1969). MATERIALS AND METHODS Specimens of Salvelinus studied were principally those in the collections of the University of Michigan Museum of Zoology (UMMZ); the National Museum of Natural Sciences, Canada (NMC); and the United States National Museum of Natural History (USNM). Locations given on the distribution map represent collections housed in these institutions ( Figure 1 ) . Other material was examined from the California Academy of Sciences (CAS). Morphometric data were taken only from specimens from which standard length could be recorded. The specimens of S. malma compared morphometri- cally with S. confluentus were taken from the Pacific drainages of the United States, including Alaska, and Canada. Most represent a type which McPhail (1961) has termed the southern form of the Dolly Varden. The National Mu- seum of Natural Sciences, Canada, possesses excellent material both of S. con- fluentus and S. malma. A large part of this material came from the University of British Columbia, Vancouver (UBC). A number of old and valuable specimens of Salvelinus axe at the United States National Museum, including holotypes of Salmo spectabilis and Salmo confluentus; Livingston Stone's collection from the McCloud River, California, in the 1 870's; and specimens taken in the early 1 880's from Puget Sound and coastal waters of British Columbia. In 1974 I collected bull trout from the Flathead River drainage in Montana, which are now housed at the Ohio State University Museum of Zoology (OSUM). Osteological data were obtained from both dry-skeletal and cleared-and- stained material housed at the Ohio State University and the University of Michigan. The number of gill rakers was determined by removing the first gill arch on the right side and counting under a dissecting microscope the individual rakers, including all rudimentary ones. Branchiostegal rays were counted after locating the smallest, most anterior ray by dissection. Mandibular pores were counted by exposing the openings of the mandibular sensory canal with a fine jet of compressed air. The count includes all pores, but not the opening from which the canal exists at the rear of the lower jaw. Pyloric caeca were cleaned of connecting tissue and fat deposits prior to counting. Vertebrae were counted from radiographs, which have been deposited at the University of Michigan Museum of Zoology. Counts were repeated until the same number was obtained twice. To study gill raker morphology, the anterior right gill arch was removed, stained in alizarin red-s, and cleared in glycerin. The most posterior gill raker on the lower limb was then illustrated with the aid of a Wild-M5 Stereoscope with integral camera lucida attachment. Osteological abbreviations are as follows: AO (antorbital); ART (articular- TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 141 • conflunntui ■ malma * hybrid FIGURE 1 . Distribution of Salvelinus confluentus and Salvelinus malma over same latitudinal range in North America; plotted from localities of specimens examined. angular); ATL (atlas vertebra); BOC (basioccipital); BR (branchiostegal ray); BS (basiphenoid); DE (dentary); ECPT (ectopterygoid); ENPT (endop- terygoid); EOC (exoccipital); EPO (epiotic); FR (frontal); HYO (hyomandibu- lar); INT (intercalar); IO (infraorbital); IOP (interopercle); LE (lateral ethmoid); MPT (metapterygoid); MX (maxilla); NA (nasal); ORS (orbito- sphenoid); OP (opercle); P (parasphenoid); PA (parietal); PMX (premaxilla); POP (preopercle); PRO (prootic); PTF (posttemporal fossa); PTO (pterotic); 142 CALIFORNIA FISH AND GAME QU (quadrate); SE (supraethmoid); SO (supraorbital); SOC (supraoccipital); SOP (subopercle); SPO (sphenotic); SPOP (suprapreopercle); SY (symplec- tic); V (vomer). Specimens Examined Salvelinus confluentus (Suckley) Sacramento R. Basin, McCloud R. drainage, Calif.: CAS 25691, (1), 110 mm,2 Shasta Co.; CAS 38787, (1) 136 mm, near Nosoni Cr., Shasta Co.; CAS 19889, (1 ), 147 mm, near Bollibokka Mt., Shasta Co.; CAS 38788, (1 ), 173 mm, Mt. Shasta State Hatchery near Mt. Shasta, Shasta Co.; USNM 26196, (1), skeleton, USNM 10547, (1), 175 mm; USNM 27820, (3), 211- 284 mm, USNM 15549, (2), 101-163 mm; USNM, (1), 339 mm; USNM 22452, (1), 301 mm. Klamath R. Basin, Ore.: UMMZ 188849, (5— two cleared and stained), 140-168 mm, Long Cr., Lake Co.; UMMZ 188851, (1 skeleton), 169 mm, Long Cr., Lake Co.; USNM 16793, (1 skin and skeleton), Linn Creek, Ft. Klamath. Columbia R. Basin, Snake R. Drainage: USNM 125309, (2), 142-154 mm, Meadow Cr. near Sawtooth, Ida.; UMMZ 117879, (1), 216 mm, Stanley Lake, a tributary to the Salmon R., Custer Co., Ida.; UMMZ 162298, (1), 166 mm, Dave Cr., Elko Co., Nev., a tributary to the Jarbidge R.; Univ. Utah No. 1, (1), 103 mm. Dave Cr., Elko Co., Nev.; CAS 38789, (6— one cleared and stained3 ), 90-155 mm, West Fork Jarbidge R., Elko Co., Nev.; UMMZ 127607, (1), 134 mm, Little Lost R. on Custer Co.-Butte Co. line, Ida. Columbia R. Basin, mainstream and minor tributaries: USNM 7078, (1 ), type of Salmo spectabilis, Columbia R. at The Dalles, Ore.; USNM 25273- 25276, (4), 279-313 mm, Walla Walla, Wash.; NMC 66-61, (1 ), Columbia R. at Arrowhead, Brit. Col.; NMC 66-63, (22), Mars Cr., tributary to Co- lumbia R. at Big Bend Highway, Brit. Col.; NMC 66-64, (5) Kinbasket Lake at Tsar Cr., tributary to Columbia R., Brit. Col.; NMC 66-70, (3 — includes one head), Lower Arrow Lake off Deer Cr., tributary to Columbia R., Brit. Col.; NMC 59-169, (20), Luthead Lake, Banff National Park, Brit. Col.; UMMZ 164591, (1 ), 252 mm, Emerald Lake, Yoho National Park, Brit. Col. Columbia R. Basin, Clark Fork Drainage: USNM 38028, (1), 310 mm, Clark Fork River; USNM 44002, (1 ), 158 mm, Rattlesnake Cr. at Missoula, Mont.; NMC 59-148, (11), Shepp Cr., tributary to Flathead R., Brit. Col.; NMC 66-72, (77), Pollock Cr., tributary to Flathead R., Brit. Col.; NMC 66-77, (17), Gumbo Flats Cr., tributary to Flathead R., Brit. Col.; UMMZ 161871, (3), 134-166 mm, Flathead Lake, Lake Co., Mont.; UMMZ 172458, (7 — including three cranial skeletons), 212-423 mm, Flathead Lake, Lake Co., Mont.; UMMZ 161866, (1 ), 350 mm, Flathead Lake, Lake Co., Mont.; UMMZ 188857, (3— one cleared and stained), 130-145 mm, Morrell Cr. near Seeley Lake, tributary to Clearwater R.; UMMZ 188856, (4), 47-90 mm, Big Creek, tributary to North Fork Flathead R. on western boundary of Glacier National Park; OSUM 25212, (2— one cleared and stained), 208-235 mm, Hungry Horse Cr., tributary to Hungry Horse Reservoir on South Fork Flathead R., Mont.; OSUM 25213-5, (1 skeleton), Flathead R. 2 All fish lengths are standard length. 3 The cleared and stained specimen from this lot has since been given a separate number: CAS 38790. TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 143 at Coram below the confluence of North and Middle Forks, Flathead Co., Mont.; UMMZ 102948, (1 ), 257 mm, Pend Oreille Lake, Banner Co., Ida. Columbia R. Basin, Kootenay Drainage: NMC 66-52, (2), Slocan Lake, tributary to Kootenay R., Brit. Col.; NMC 66-57, (4) Duncan Lake near Howser, tributary to Kootenay R., Brit. Col.; NMC 66-59, (1), Kootenay Lake at the mouth of the Duncan R., Brit. Col.; NMC 66-68, (6), Kootenay R. at Canal Flats, Brit. Col.; NMC 66-74, (1), mouth of Wolf Creek at Skookumchuck, tributary to Kootenay R., Brit. Col.; NMC 66-78, (2), mouth of Gold Cr. near Newgate, tributary to Kootenay R., Brit. Col.; NMC (no number), (12), Lardeau R., tributary to Duncan R., Brit. Col. Puget Sound, Washington: USNM 1135, (1), 733 mm, Pullayup R. at Steilacoom, type of Salmo confluentus; USNM 27264, (2), 295-568 mm, D. S. Jordan, 1880; USNM 42044, (1 ), 297 mm, Elliot Bay, O. B. Johnson, 21 May 1889. Fraser R. Basin, Brit. Col.; NMC 55-130, (2), 184-185 mm, Salmon R. at Hart Highway, north of Prince George. Skeena R. Basin, Brit. Col.: UMMZ 159357, (3), 223-358 mm, Lakelse Lake; UMMZ 159345, (2), 181-195 mm, Damshilgwit (Cabin) Lake; U M MZ 1 59333, ( 1 ) , 230 mm, Morrison Lake; U M MZ 1 59337, ( 2 ) , 1 88-203 mm, Morice Lake; UMMZ 159352, (1 ), 368 mm, Slamgeesh Lake; UMMZ 159351, (1 ), 341 mm, Sustut Lake, USNM 86207, (1 ), 265 mm, Bear Lake. Taku R. Basin: NMC 68-896, (5), 150-246 mm, Flannigan Slough, Taku R. at International boundary between British Columbia and Alaska. Upper Yukon R. Basin: NMC 68-1231, (7 of 15), 97-198 mm, Partridge Cr., tributary to the Swift R. near Yukon-British Columbia boundary. MacKenzie R. Basin, Liard Drainage, Brit. Col.; NMC 62-234, (1), 313 mm, Tatsho Cr. near Dense Lake, tributary to Liard R.; NMC 62-235, (2), 340-344 mm, Letain Lake near King Mt.; NMC 68-1230, (4), 129-176 mm, outlet from Little Lake. MacKenzie R. Basin, Peace R. Drainage: NMC 66-435, (1), 375 mm, Chuchi Lake on Nation R., tributary to Parsnip R. N.W.T.; NMC 66-436, (1 of 2), 113 mm, Germansen Lake on Omineca R., Brit. Col.; NMC 66-437, (1 ), 510 mm, Finlay R., about 4 miles upstream from Ft. Grahame, Brit. Col; NMC 66-438, (7), 172-333 mm, mouth of Finlay R. to Manson R., Brit. Col.; NMC 66-440, (1 ), 206 mm, Peace R., 25 miles downstream from Hudson- Hope, Brit. Col.; NMC 68-802, (1) 171 mm, Peace R., 11 miles west of Hudson-Hope, Brit. Col. MacKenzie R. Basin, Athabaska Drainage, Alberta: UMMZ 80837, (1), 253 mm, Jacques Lake, tributary to Rocky R., Jasper National Park; UMMZ 159930, (5), 225-324 mm, Jacques Lake, Jasper National Park; NMC 59-48, (A — including 2 heads), Jacques Lake, Jasper National Park. Saskatchewan R. Basin: USNM 64326, (1 ), 715 mm, headwaters of Bra- zeau R., tributary to North Saskatchewan R., Alberta; UMMZ 164943, (1 ), 164 mm, Banff National Park, Spray R., tributary to Bow R. of South Saskat- chewan drainage, Alberta; UMMZ 164928, (1), 190 mm, Bow Lake, Banff National Park, Alberta; UMMZ 164930, (1 ), 215 mm, Bow R., Alberta; NMC 60-343, (1 ), Red Deer R. drainage at Morrin, 65 miles northeast of Calgary, Alberta; USNM 44444, (1 ), 231 mm, Oldman R., tributary to S. Saskatche- wan R., Alberta; UMMZ 188900, (8), 190-267 mm, Cracker Lake, tributary 144 CALIFORNIA FISH AND GAME to St. Mary's R. of S. Saskatchewan R. drainage, Glacier National Park, Mont. Salvelinus ma I ma (Walbaum) Sacramento R. Basin, McCloud River, Calif.: USNM 20819, (2), 235-241 mm, sent by Livingston Stone, catalog entry made Nov. 24, 1877. Soleduck R., Washington: UMMZ 93829, (14 — three cleared and stained), 100-134 mm, above Soleduck Falls, Olympic Peninsula. Puget Sound, Washington: USNM 34301-34305, (5), 252-274 mm, Port Townsend, James G. Swan, coll. in 1884 or earlier. Skagit R. Drainage, British Columbia: UMMZ 179422, (2), 50-65 mm, Skagit R. near Hope. Skeena R. Basin, British Columbia: UMMZ 159323, (1 ), 200 mm, Alastair Lake; UMMZ 159344, (2), 140-144 mm, Johanson Lake. British Columbia: NMC 59-150, (2 of 6), 126-130 mm, East Fork Seltat Cr. at Haines Rd. and about 1 mile below Snowater Lake, tributary to Klehin R.; NMC 65-213, (1 head), Nass Harbour, Iceberg Bay near mouth of Nass R.; NMC 65-212, (3), 146-159 mm, Nass Harbour just north of Jacques Point, Iceberg Bay; NMC 65-225, (5), 189-242 mm, mouth of stream, cove on south shore steamer passage about % mile west of Khutzeymaten Inlet; NMC 65-159, (2), 197-293 mm, cove on west side of Refuge Bay and at north end of Porcher I., south of Prince Rupert; USNM 31979, (1 ), 285 mm, Port Simpson, Capt. H. E. Nichols, June 1882; USNM 37610, (1 ), 98 mm, taken in fresh water at Port Simpson. Taku R. Basin, Alaska: NMC 58-402, (2), Twin Glacier Lake, tributary Taku R.; NMC (UBC 58-388), (5— part of large series), Canyon I., Taku R. Alaska, Aleutian Islands: UMMZ 106266, (13 — two cleared and stained), 52-127 mm, small stream on Atka I.; UMMZ 106529, (3), 383-497 mm, Unalaska I. Alaska: UMMZ 128983, (4), 257-310 mm, vicinity of King Cove, Belkof- ski Bay, Alaska Peninsula; UMMZ 106260, (1 ), 288 mm, freshwater stream tributary to Three Saints Bay, Kodiak I.; UMMZ 126507, (2), 286-289 mm, KarlukR., Kodiak I.; UMMZ 126476, (11— two cleared and stained), 71-151 mm, Upper Thumb R., tributary to KarlukL., Kodiak I.; UMMZ 159395, (2), 100-125 mm, North Fork of Upper Thumb R. above falls, Kodiak I.; UMMZ 159393, (1 ), 53 mm, Falls Cr., above falls, tributary to Karluk L., Kodiak I.; OSUM 25213, (5), 98-138 mm, mouth of tributary entering Karluk L., Kodiak I.; UMMZ 106267, (1 ), 174 mm, Lake at north end of Sitkalijak I.; U M MZ 1 82299, ( 1 ) , 1 34 mm, Baranof I ., stream on north shore of Port Lucy at west end of Island. Copper River Drainage, Alaska: UMMZ 162600, (1 ), 173 mm, Chitina R.; N MC ( BC 58-227 ) , ( 5 ) , 205-327 mm, South L. of Chenan Lakes to Copper R. Yukon River Basin: UMMZ 144581 4 , (1 ), 85 mm, Grant Cr. about 30 miles west of Tanana; UMMZ 133553 4, (2), 186-194 mm, Riley Cr., McKin- ley National Park; NMC (UBC 58-271), (4), 152-190 mm, Dry Cr. near Alaska Highway, Tanana R. drainage. 4 Specimens that represent the "northern form" of Salvelinus ma/ma (McPhail 1961). TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 145 MacKenzie R. Basin, British Columbia: USNM 147661, (1 ), 102 mm, Hot Springs, 3 miles WNW of junction of Trout R. and Liard R. July 6-8, 1948, J. R. Alcorn. U.S.S.R.: UMMZ 145814, (1), 123 mm, Pogarna R., Kamchatka. Japan and Kuril Islands: NMC 60-154, (6), Shokotsu R., Soya Province, Hokkaido; UMMZ 186872, (6— one cleared and stained), 130-135 mm, Pon-mataochi R., Nemuro Province, Hokkaido; UMMZ 188710, (9 — one cleared and stained), 79-168 mm, Ishikari R. at Sounkei, Ishikari Province, Hokkaido; UMMZ 188711, (9), 65-180 mm, Kuzira Bay, Paramusiro I., Northern Kuril Islands. Korea: UMMZ 188698, (4), 70-187 mm, Upper Tuman R., Mozan; UMMZ 188712, (1) 429, Joshin Bay near Seishin. NOMENCLATURE Salvelinus confluentus (Suckley) Salmo spectabilis — Girard, 1856: 218 (orig. descr.) Salmo spectabilis — Girard, 1858: 307-308 (amended descr.) Salmo confluentus — Suckley, 1858: 8-9 (orig. descr.) Salmo spectabilis — Suckley, 1860: 342-343 (correct locality given) Salmo confluentus— Suckley, 1860: 334-335 (amended descr.) Salmo bairdii— Suckley, 1861: 309 (orig. descr.) Salmo parkei—Suck\ey, 1861: 309-310 (orig. descr.) Salmo campbelli— Suckley, 1861: 313 (subst. name and amended descr.) Salvelinus spectabilis — Jordan, 1879: 79-81 (amended diag.) Salvelinus malma — (in part) Jordan and Gilbert, 1882: 319-320 (synonymy given) For years the bull trout and the Dolly Varden, here considered specifically distinct, have been combined under one name, the Dolly Varden, Salvelinus malma (Walbaum). The latter name is correctly applied to the form which is generally characterized as anadromous. As now understood, the range of S. malma spans the entire arc of the North Pacific from the Sea of Japan and Kuril Islands, across the Aleutian chain to Alaska and south along the North American Pacific coast to the northwestern U. S. (Figure 1 ). Because of its international usage, the common name Dolly Varden is better reserved for S. malma although it may originally have applied to S. confluentus. The latter ranges well inland in the United States and Canada and is generally nonanadromous. Where S. con- fluentus reaches an adult size of several kilograms or more, such as in the Kootenay River of British Columbia and the Flathead River drainage of Montana, fishermen refer to this species as the bull trout (Dymond 1932; Brown 1971 ); a name inspired by its large, broad head, its large mouth and prominent jaws, and its highly piscivorous diet. The original name for the bull trout, Salmo spectabilis Girard, is a secondary homonym (Suckley 1861, Morton 1970), but there are four other scientific names available for the bull trout, all proposed by Suckley (1858, 1861 ): Salmo confluentus, Salmo bairdii, Salmo parkei, and Salmo campbelli. The first bi no- men, Salvelinus confluentus (Suckley), is selected for use because, 1) it has precedence over the other three in date of publication (Suckley 1858), 2) fewer nomenclatural problems, including spelling, are anticipated for the name con- 146 CALIFORNIA FISH AND GAME fluentus, 3) Suckley's type, a dried head and skin (USNM 1135), is still in existence, and 4) S. bairdii and S. parkei lack type specimens. Jordan and Evermann (1896) placed Sa Imo con fluentus Suckley in synonymy with Oncor- hynchus tshawytscha (Walbaum), probably because Suckley's (1858, 1860) descriptions read, in part, like that of a Pacific salmon. However, careful com- parison of the head of the type specimen with Suckley's description has led me to conclude they are the same. TAXONOMIC HISTORY As indicated above, the bull trout 5 was first described as Salmo spectabilis (Girard 1856). Girard gave the locality of the holotype of spectabilis (USNM 7078) as St. Mary's Mission on the Clark Fork of the Columbia River, Montana, but George Suckley (1860) who had collected the specimen in 1854, rede- scribed spectabilis and corrected Girard's locality information, stating that the holotype came from Ft. Dalles on the lower Columbia River. This specimen is now a mutilated, half-rotted individual estimated at 200 mm sl. In another paper, Suckley (1861) realized that the name spectabilis was preoccupied and substituted Salmo campbelli (after Archibald Campbell, Chief of the N.W. Boundary Commission). Morton (1970) has given a detailed ac- count of this name change and the rules that pertain to it. In the same paper, Suckley also described Salmo bairdii and Salmo parkei, which are both con- specific with spectabilis. I searched the USNM collections but failed to find the types of bairdii and parkei. However, the holotype of Suckley's "Salmo con- fluentus" \n&s found (Figure 2A). This specimen consists of a dried head and skin now preserved in alcohol. The head is that of a bull trout, but the description by Suckley (1858) states that the dorsal, adipose, and caudal fins were spotted profusely with dark brown and black (unlike Salvelinus) . This could not be confirmed, for on examination of the type, no dark spots were found. Thus, I believe it possible that the description of Salmo confluentus was based on the remains or observations of two different individuals, one of which was a bull trout and the other a Pacific salmon. The type-locality for Salmo confluentus was given as the Puyallup River near Ft. Steilacoom, Washington Territory. The type was procured September 27, 1856 by Suckley, apparently from Indians who had captured the fish — a large (approximately 700 mm) male in spawning condition. It possesses a kype that fits into a deep notch between the premaxillae. The notch is accentuated be- cause of the dried condition of the specimen. Thus, we see that Suckley not only collected the holotype of S. spectabilis and later redescribed it as Salmo campbelli, but within a few years described the same species three more times under different names {confluentus, bairdii, parkei), using specimens he himself had collected. Jordan (1879), in his key to the species of Salvelinus found within the United States, included characters of the bull trout — "head large, stout, broad and flattened above" — in his diagnosis of Salvelinus spectabilis (Girard). He appar- ently based his key characters on specimens taken from the Clackamas River, Oregon, by Livingston Stone. He also examined Girard's type of Salmo spectabi- Although "char" may be a more appropriate term for members of the genus Salvelinus (Morton 1955), "trout" is used herein as the common name for the bull trout in accordance with the American Fisheries Society's attempt to stabilize fish nomenclature (Bailey 1970). TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 147 lis, noting that the holotype was still preserved in the U.S. National Museum, while that of parkei was lost. He added that parkei was unquestionably the same as 5. spectabilis. In the same paper, Jordan recognized the bull trout from the McCloud River as S. bairdii (Suckley), believing it differed from 5. spectabilis by lacking basibranchial teeth. He did not list S. malma (Walbaum) in his key. Jordan and Gilbert (1882) were responsible for placing the bull trout in synonymy with Salvelinus malma (Walbaum). Jordan and Evermann (1896) followed this precedent, which has been continued to the present. Jordan never FIGURE 2. (A) Holotype of Salmo confluentus Suckley 1858, USNM 1135, head length 173 mm, Puyallup R. near Steilacoom, Washington; (B) lateral and ventral view of head of spawning male Salvelinus confluentus, NMC 66-70, estimated 550 mm standard length, Deer Cr., British Columbia; (C) lateral and ventral view of head of spawning male Salvelinus malma, UMMZ 126507, 295 mm, Karluk R., Kodiak I., Alaska. Photograph by the author. 148 CALIFORNIA FISH AND GAME correctly distinguished the bull trout from Dolly Varden where they occur to- gether in Puget Sound. Under his use of the name, "spectabilis", he lumped the bull trout with the Dolly Varden of coastal areas as far north as Alaska. In later papers (Jordan 1923, Jordan and McGregor 1925) the name Salvelinus spectabi- lis was used for a supposed "southern form" which ranged from northern California northward to Alaska and the name Salvelinus malma applied to a "northern form" known from Unalaska to Kamchatka. The earlier descriptions of " Sal mo spectabilis" mentioned a number of distinc- tive characteristics of the bull trout. Girard (1856, 1858) emphasized the curved maxilla in the type specimen and its elongate head which entered the standard length 3.5 times. Under his description of Salmo confluentus, Suckley (1858) was the first to mention the "projection of the chin anterior to the front teeth"; in fact, he confused this character with the hooked lower jaw of spawning male Oncorhynchus. He also added that the type of confluentus had 1 3 or 1 4 branchi- ostegals. Again, in listing characters for Salmo bairdii, Suckley (1861 ) described the "snout having a deep notch between the extremities of the premaxillaries receiving a conical fleshy protuberance projecting upwards from the chin." In the same paper under Salmo parkei, Suckley noted the large head "about four and a half times in the total length; its top flat; muzzel pointed" and "branchi- ostegals 13-14." Also in parkei, Suckley mentioned "a disposition toward the formation of a fleshy 'tit' projecting upwards at the point of lower jaws with a corresponding notch between the premaxillaries." Despite what Morton (1970) has published to the contrary, there is little doubt in my mind that in his descrip- tion of Salmo bairdii and Salmo parkei, Suckley (1861 ) was writing about the bull trout and not the Dolly Varden. To conclude this discussion of taxonomic history, it is emphasized that the name confluentus is to be substituted for spectabilis in accordance with the rules of zoological nomenclature (Article 53) and the holotype of this species becomes USNM 1135. DIAGNOSIS A large species of North American Salvelinus reaching a greater size (to 18.3 kg, about 40 lb, Hart 1973) than S. malma; distinguished from the latter by its long, broad head which is flat above and sharply tapered through the snout with the eye positioned near the dorsal margin; head measures 3.7 in standard length, averaging 3.6 in juveniles and 3.9 in large adults (5. malma measures about 4.3); jaws and teeth well-developed, with cleft of mouth terminal; maxilla constantly curved downward; branchiostegals typically 13 or 14 on the right side, 24-31 both sides combined; mandibular pores usually 7-9 on a side, 16 combined (5. malma typically has 6 on a side); gill rakers robust with strong teeth projecting from mesial edges toward the branchial cavity (whereas S. malma has relatively long, finely tapered flexible gill rakers that are much compressed dorsoventrally and lack teeth projecting from the mesial edge); gill rakers 14-19, pyloric caeca 21-36, and vertebrae 62-67; basibranchial teeth 0-11, averaging 4, and consist- ently arranged in a single longitudinal row; in all specimens examined, there is a pronounced gap on each side between the palatine and vomerine tooth rows; in anterior view, there is a notch at the lower terminus of the snout (between the premaxillae) that receives a fleshy protuberance on the symphysis of the mandibles, although best developed in the adult (both male and female). This TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 149 character is also seen in larger juveniles and is not to be confused with the kype of spawning male salmonines, even though the same area is similarly modified in large, breeding males of S. confluentus. DESCRIPTION The characters of S. confluentus described below were found to be consistent in samples taken from localities throughout its known range. These samples comprise a total of 332 individuals from eight major river basins draining to the Pacific and Arctic oceans and Hudson Bay: Sacramento, Klamath, Columbia, Skeena, Taku, Yukon, MacKenzie, and the Saskatchewan. Some of the characters which have been employed for many years in Sal- velinus taxonomy, such as numbers of gill rakers and pyloric caeca, will not separate 5. ma/ma and 5. confluentus. This is one reason why the bull trout has not been recognized as a distinct form. Most useful in separating S. confluentus from S. malma are the shape and size of the head. In addition, characteristics of the jaws, teeth, and gill rakers, number of mandibular pores, number of branchiostegal rays, arrangement of basibranchial teeth, neurocranium profile, and configuration of certain bones of the cranium, such as the supraethmoid, frontal, preopercle, and opercle, are very useful. Numbers of gill rakers, pyloric caeca, and vertebrae show considerable overlap between the two taxa. Head Size and Shape In dorsal view (Figure 3A) the head appears very broad and flat on top and is hard to the touch. The frontals slope only slightly in a lateral direction away from the midline. In S. malma (Figure 3B) the head is more compressed and the frontals usually peak at the midline. Unlike S. confluentus, the frontals of 5. malma are usually covered with thick, fatty tissue underlying the skin. This is best observed in anadromous S. malma. In lateral view (Figure 3A), the head of S. confluentus is low and sharply conical (also see Paetzand Nelson 1970). The terminal cleft of the mouth evenly divides the anterior profile of the head, while in S. malma the snout tends to be shorter and deeper and often overhangs the tip of the lower jaw, especially in juveniles (Figure 4B). The eye of confluentus (Figures 2B, 3A) is more dorsal in position than in 5. malma (Figures 2C, 3B). The vertical distance from the center of the eye to the top of the head falls well short of the distance from the center of the eye to the nares, while in S. malma, this distance reaches the nares or nearly so. In anterior view, the greater breadth of the head is again noticeable in S. confluentus (Figure 3). A major characteristic seen in this aspect is the notch dividing the premaxillae, which receives a fleshy protuberance at the teminus of the lower jaw ( Figure 3A) . This character is best developed in adults of both sexes and reaches maximum development in spawning males (Figures 2A, 2B). Juveniles older than 2 years may also have this feature. S. malma may have a well-developed kype in spawning males of anadromous populations, as shown by Morton (1965), but the kype is barely evident in spawning males of the nonanadromous S. confluentus ( Morton 1965). On speci- mens I have examined, the kype was best developed on the largest spawning males (over 500 mm sl). Its existence is probably a function of size. The kype 150 CALIFORNIA FISH AND CAME FIGURE 3. Head form (lateral, dorsal, ventral, and frontal views) in the adult female of (A) Salvelinus confluentus, NMC 66-437, 580 mm, Finlay R. trib., Brit. Col.; (B) Salvelinus malma, UMMZ 126507, 289 mm, Karluk R., Kodiak I., Alaska. Photograph by the author. in S. malma is directed dorsally, whereas in S. confluentus it has a slightly more anterodorsal orientation. The upper jaw of the bull trout in lateral view (Figure 3A) always exhibits a pronounced downward curve as seen in the concavity of the toothed margin and convexity of the dorsal margin, particularly where the supermaxilla is seated. Typically, the maxilla of S. malma is more slender, with the toothed shaft straight or only slightly curved downward. Head length of 5. confluentus (Table 1 ) typically enters the standard length less than 4.0 times, whereas in S. malma it enters the standard length about 4.25 times. The difference is greater when data from dwarf landlocked populations TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 151 FIGURE 4. Head form in juveniles of (A) Salvelinus confluentus, OSUM 25212, 235 mm, female, trib. of S. Fork Flathead R., Montana, (frontal, lateral, dorsal views); (B) Salvelinus malma, OSUM 25213, 138 mm, female, trib. Karluk R., Kodiak I., Alaska, (frontal, lateral, dorsal views); (C) Salvelinus confluentus UMMZ 188857, 134 mm, male, trib. of Clearwater R., Montana (lateral view). Photograph by Cus Spreitzer. of S. malma are excluded. However, the proportion of head length to standard length changes with the size of the fish in both S. malma and S. confluentus (Table 2). While individuals under 100 mm would be difficult to separate using this character, those over 250 mm show a high degree of separation. I found no difference in proportionate head size between adult male and female S. con- fluentus, although I suspect that males over 500 mm will have relatively longer heads. In S. malma, however, males over 200 mm tended to have larger heads than females of the same size. Branchiostegal Rays S. confluentus had the highest branchiostegal ray count of all Salvelinus I investigated. It averaged 14 rays on the left side and 13 on the right side in 120 specimens, with a range from 12 to 16 (Table 3). The range for both sides combined was 24 to 31 with a mean of 27 (Table 4). Among the species of 2 — 77343 152 CALIFORNIA FISH AND GAME TABLE 1. Head Size as the Proportion of Standard Length to Head Length in Salvelinus confluentus and Salvelinus malma. Drainage Sample size Range Mean Salvelinus confluentus McCloud River, California 12 3.2-3.9 3.6 Klamath River Basin, Oregon 5 3.5—4.1 3.8 Columbia River Basin, Snake River drainage 11 3.6-4.7' 3.8 Columbia River Basin (excluding Snake drainage) Puget Sound, Fraser River Basin 23 3.3-4.1 3.7 Skeena River Basin 10 3.5-4.0 3.7 Taku River Basin 5 3.7-4.1 3.9 Yukon River Basin 7 3.6-4.0 3.7 MacKenzie River Basin 25 3.5-4.1 3.7 Saskatchewan River Basin 13 3.5-4.0 3.7 TOTAL 111 3.2-4.1(4.7)' 3.73 Salvelinus malma McCloud River, California 2 4.5-4.7 4.6 Soleduck River, Washington 5 3.8-3.9 3.9 Puget Sound, Washington 4 4.2-4.6 4.4 Skeena River Basin 3 3.9^.5 4.1 Coastal British Columbia 14 3.5-4.5 4.2 Taku River Basin 7 3.5-4.6 4.1 Gulf of Alaska, Kodiak Island, Coastal Streams and Islands 20 3.6-4.7 4.3 TOTAL 55 3.5-4.7 4.23 1 One individual, Univ. Utah No. 1, with abnormally small head. TABLE 2. Head Size as the Proportion of Standard Length to Head Length in Salvelinus confluentus and Salvelinus malma, in Relation to Body Size and Sex. Standard length (mm) Sample size Range Mean Salvelinus confluentus 50-100 5 3.3-3.8 3.63 101-150 20 3.4-3.9 3.62 151-200 27 3.2—4.0 3.67 201-250 24 3.5-^.1 3.74 251-300 15 3.6-1.1 3.82 301-350 11 3.5-3.9 3.70 351-568 10 3.6-4.1 3.86 ; (200-568) 20 3.5^1.0 3.73 - (200-423) 18 3.5-4.0 3.73 Salvelinus malma 50-100 10 3.3-4.1 3.63 101-150 10 3.8-4.3 4.00 151-200 10 3.6-1.5 4.16 201-250 9 3.9-4.7 4.36 251-300 14 4.0-4.7 4.38 301-445 8 3.8-4.7 4.40 J (200-380) 12 4.1-1.7 4.46 r (200-445) 10 3.8-4.4 4.10 Salvelinus, S. namaycush is closest to S. confluentus in number of branchioste- gals. Vladykov (1954) reported S. namaycush To average 13 on the left side with a mean of 25.3 for both sides combined. This character is important in separating S. confluentus from S. malma. The average for 88 specimens of S. malma was 11.61 on the left side, 11.0 on the right side, with a range of 9 to 12 (Table 3). TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 153 I 1 3 ! •a c I I c 8 3 .c I 4 4; O c P3 -C ■o c c o c 3 01 CO < C5 IN o. o o o en en ps ps - ^t CO M (^ ro m tj- rsi v£> oO I I* <3 i— .— PS Is .— PS CT> >— © © in is ^ po © PO ■— PS »— PS 2 cz 01 00 ai > o B — 00 ro OJ u6 .B ai c «i > ■-= 5 Oi PO CO OJj "O . bt 3 -C 1 .9 « U £ U P0 Cj 55 > c: > > -s «2 'i/» .£ co co »— rs po in is pspoco ^ ^ PO •— co T5 c P0 P0 £ . c : c o u 2 c cS '5 ™ N g -f 8 pocS^ £ ^- tinc,ij- 2 in l- c 2 u oc m c > 1c bi > oo.S -= > P0 ui ,9 . -^ po LJ i- in co > -» -T *- ^ - £ a> 0£ co > po To Qi £ £ P > -D 3 u -a O U I/O t— O 3 _^ O ro 1/1 Q- l^) <_) r— a; oo P0 c P0 o u < -a iO 3 ^ O 154 CALIFORNIA FISH AND GAME 3 1 3 c 3 c a (j 3 ! 3 ■o V c 15 E o u (0 'J? ■*- o> -j -o J c 5 "^5 £ 5 m Q£ BO a; i/> o 1c u c c 3 >- U c 3 cr o> CO < CO CO o -T ro CO s? CN 3 CO ■"■ O •— ' o o o ' ■ 1 c ? 1 8 CO rs. m CO o r^l O rr r-» ro rv CO cm rv r-v' cn fN CO Ol iv fN IV (N ■""I § oromcOLnrMrv r^iOrO' — (N K ffi l/l tN (N rn r'i (N tN i — rs| •— m lti tt co -- m in (N 00 fN CM fN K (N m ^t fN (N CM t— fN M O If) (N 1>D»— O M fl t ■- fN O ro CO i— cm I s 3 y— rs| T— tJ- r- C\ 1— cm ro «— l\ »D fN (N 0> •- •^■'^J-CO J S fc ro r— M Tf 1- (N 1 s; CsJ r- fN fN O rn ^D 3 rv cm o o .— 0-1 .— O ro o cm >— cm ■— CM LO *— rvrOLAIvCMrocO ■— i— ro .— CO _o u u o I/} al 00 o. > > OZ OIL c ro ■5 =0 ro c C : c CO £ a c -a .2 £ co "5 ,2 S 5 u T3 O ro Id E ro t_ 1 ^^ *- in > = Qi CO 2 2 I s c ro O0 ro 2 tn ro O u J5 < t/) ^ o|< r- O < r- TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 155 For both sides combined, S. malma had a range from 19 to 25 with a mean of 22.6. Counting branchiostegal rays on the right side separated 90% of the 5. confluentus from S. malma. By combining both sides, the separation was not increased significantly. Mandibular Pores Of 118 5. confluentus examined, 63% had 16 or more mandibular pores on both sides with a range from 12 to 19 (Table 5). In 5. malma there are 10 to 15 total with a mean of 12.1. Elsewhere among the species of Salvelinus, only 5. namaycush has a higher mandibular pore count (Morton and Miller 1954). Basibranchial Teeth Basibranchial teeth of S. confluentus axe characteristically arranged in a single longitudinal row. They usually number from 3 to 5; however, of 40 individuals, 6 (15%) had no basibranchial teeth. In S. malma basibranchial teeth are usually more numerous and arranged in one to three rows on the basibranchial plate. Three of 35 (9%) S. malma that I examined had no teeth, and Morton and Miller ( 1 954 ) found that 4 of 20 ( 20% ) specimens they examined had no basibranchi- al teeth. Gill Raker Morphology After examining large numbers of gill arches from S. confluentus and S. malma, I conclude that the form of the raker and the characteristics of its dentition are more important in separating S. confluentus from S. malma than actual gill raker counts, which have been employed so extensively in salmonid taxonomy. Gill raker morphology provided a high degree of separation (98%) of S. confluentus from S. malma. Features of the raker that are most distinctive are the shape and degree of dorsoventral compression, the relative size of the teeth, and their presence or absence along the mesial edge of the rakers (the edge that faces the branchial cavity). The gill rakers of S. confluentus ate robust and oval in cross section. They have strong teeth projecting well out from the mesial edge (Figure 5A) as well as having smaller teeth on the dorsal and ventral surfaces. The raker is ornamented with prominent ridges along its lateral margin. S. malma, in contrast, has rakers that are strongly compressed dorsoventrally so that the broad surfaces of the rakers are flat and weakly ridged (Figure 5B). Usually the rakers possess long tapered tips that are quite delicate. Although there are usually small teeth on these surfaces in S. malma, the mesial edge lacks the strong projecting teeth entirely. Body Form The trunk of S. confluentus tends to be slender and rounded with only slight lateral compression (Figure 6). In S. malma the trunk, like the head, is more laterally compressed. A proportional difference that helps to separate the two species is found by dividing the head length into the distance from the vent to the base of the tail. The head length nearly always equals or exceeds this distance in S. confluentus, whereas in S. malma it falls short. Spotting Pattern The size and distribution of spots are weak characters, although they have been used for identifying North American Salvelinus, especially by early taxono- 156 CALIFORNIA FISH AND GAME •a c I | 3 ev C 15 E o u ■o «75 -1 a .2? 3 • — ■o c (0 1 (A 5 c V 3 m -, =- I vs. ^p ~! a f oco©Ti\a->cO'— r-^^- L/SrOLnsdvduSi-riLnTj-Lri |v.rN©i-nQrvvOO u->CT%OvDO0V0000 OOOOOOOO COQQOOvO<— O ^-COrNO—;^^ rvi rvi rvj rvi cvi cvi .— ' cvi t— (N ^ fO . — i — Tf vD CM (NNiflrOi-KNlfl I— r— VD ■* t CO IN ^D CM •— "* 1— T}- __^U-1>— CM CM >— u-1 > > OtL Of. O) OJ c r0 <*0 C nj CO CO 'y, CD _ _ 13 ,E .5 ™ co «* -C _Q _Q ro CO c -C gEE?"5^* E 3 3 0) 3 9 ^ U U lO r- >- c o cu > ra E :*: To < u -* I— u 3 o c/-> ra sj *- en ■O c ro CD 00 ro C ro Q o u > 2 'l — CO > ^3 < ro rr ^3 Oi o CI J [SI 3 . t cu o ^ 3 (/) U r- u 3 I— -i o TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 157 FIGURE 5. Most posterior gill raker of first arch, right side (ventral view, left; dorsal view, right) from (A) Salvelinus confluentus, UMMZ 17248, 285 mm, Flathead L., Montana; (B) Salvelinus malma, UMMZ 128983, 312 mm, King Cove, Alaska Peninsula. Photograph by the author. FIGURE 6. Body form and pigmentation in Salvelinus confluentus (A) UMMZ 188857, 134 mm, juvenile, trib. Clearwater R., Montana; (B) OSUM 25212, 235 mm, juvenile, trib. S. Fork Flathead R., Montana. Photograph by the author. 158 CALIFORNIA FISH AND CAME mists. The light-colored spots of 5. confluentus are normally smaller than the diameter of the pupil of the eye, yet large enough to be seen on a fish in clear water. Typically, they cover most of the back and are best developed around the dorsal fin. Suckley (1860) and Campbell (1882) noted these light-colored spots on the back of the bull trout. Other Meristic Characters The range of variation in numbers of vertebrae, gill rakers, and pyloric caeca were about the same in S. confluentus and S. malma, with almost a complete overlap between the two species. The mean numbers of pyloric caeca differed by less than two (Table 6). 5. confluentus has a relatively high number of vertebrae (mean of 64.8) compared to other Salvelinus taxa from the Pacific basin (Table 7). Southern S. malma has a mean of 62.9. The increase in number occurs in the precaudal series (36 in S. malma vs. 39 in S. confluentus) and probably is correlated with the piscivorous habits of S. confluentus. Other fish eaters, such as species of Esox, have a long abdominal cavity to accommodate large prey. The bull trout is characterized by a slightly lower number of gill rakers than other North American Salvelinus. The number was found to range between 14 and 20 in S. confluentus with a mean of 16.6. For malma the range was 14 to 23 with a mean of 18 (Table 8). Osteology The skeleton of Salvelinus confluentus offers an impressive set of characters that fully separate this species from S. malma. It was an examination of an osteocranium of a specimen from Flathead Lake that first led me to suspect that the bull trout was different from the Dolly Varden. The external morphology of the head described previously is related to the features of the cranial skeleton. The distinctive features found are consistent in samples taken from the various drainages where the bull trout occurs. These characters have been carefully checked by dissecting preserved material, radio- graphing nearly 100 preserved specimens, clearing and staining, and by examin- ing dried skeletal preparations. The latter have been used for purposes of illustration. A comparative osteological study of the head skeleton in certain Salvelinus, including S. malma from the western Pacific basin, has been pub- lished by Shaposhnikova (1971). This should be referred to in comparing 5. confluentus with S. malma. The articulated cranium shows: 1 I a flattened skull roof; 2) an elliptical orbit with a longitudinal axis much longer than the vertical; 3) a large cavity behind the orbit which was occupied bv the adductor muscle of the lower jaw, and 4) massive jaws with strong teeth ( Figure 7A). The jaw teeth are best developed on the premaxilla, dentarv, and anterior end of the maxillary alveolar shaft. The latter is curved with its dorsal margin covered posteriorly by a sigmoid-shaped supramaxilla. The large area for attachment of the adductor mandibulae muscle is made possible b\ expansion of the lateral h\ omandibular surface ( Figure 8A ) . Because of this expansion, the hyomandibular is one of the most diagnostic of all the cranial elements. Among the Salvelinus investigated, the lake trout, Salvelinus namaycush, has a similar hyomandibular. TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 159 u c 3 a- a 1 ^o a r~> E <•) lr> 3 ""1 .c ^- 1 s; 5 <3 *n ■o "-i c nj r^ «) **> a v. e ^ Oj T3 ^ a y 55 2 d s **•> 3 5 - .c CV ^ •^- b Q S ^ 2P * *"" < r3 3> c: 8 Oi iP <*> *" nj ^N a u .5 "5; $ 0 <>5 CL * — ^. o c 5? o — 3 ^ « — -O "N «— m rsi .— ov en .— T (VI — .— (vi rsi in *x> .— o-> O rv in vD "J- (vi o (V| (V| . — ov •— co ■— (vi r^ co 1— eo (vi »— rv c o U ._ t/i Q£ (0 CO "§^ V E 2 S OJ > > 5 01 2 CO 00 OO ■- *^ (^ E as n: a; co \n 15 25 9f ™ c: c P ™ co 33 O) =) OO _2 (0 c c n3 ■3! co ro _. co b •3> Jj 5 as rsi OJ co c -C <^ "^ a zi a ^ ^0 < = 5 ^ O c: O c > /. SJ .-ST > 00 > (0 . ^^ >- tn ai > -a oi c ■* o "? OJ Ji 00 0 p tn 0. (0 J3 E c = > c oi co (0 c 0j I 2^0 (0 L_ Q to 1>1 (0 "O O C •si -* -a c O I7i U 13 (0 Ol C 00 n —z ro -x; ■< =) 1— -i O < H- 160 CALIFORNIA FISH AND CAME I i c "5 c li C 8 I w > c c 3 -2 O >- u c - - < a ^ =0 c c ?5 Fo I OOO^DroOO^i-n i--vLn!\Lncoo^l*0 oooooooooo CO CX"> O *— t— rsicT-O-— NLncO'-CO'-sOfO O © O ~ © *- '■ — '■ ^ 8QKO^QNOcOfN Ot-il^-OO-sD^D — CO •*& ,*0 \^ v»D "sO ^O ^.O ^O ^D '-O — o r\ i_n r^ ld in in o> ^.O ^O ^D >s0 ^.O vO \0 \& CM Tj- CM CM TJ" — t— \£> .— LD — t— o-i <— OO rn O CM 1 ~T •— ^D •— i_n Cv| cr 3 5 ■*t ' — CN m r 1 1 •— — > T co o r\ i — u-i r\ ui cm a** .— — ■— cm .— <3-> t— (vimM-'^eooaD I— i — (N < — t\ <5 U o 0) OO > > — _c u c &l 1- 3 —^ a> 0£ - g V C 5 [/I Ll_ c C d J- I/! j~ * f0 - GO CO EC m _ ■35 CO f0 "> .~ a; $ 2 2 co '3; J3 'n £ -C -Q la "^ f f ^ C C ' O r- ^ ±i ~i J5 o o^ig^^ SO ^c V f0 c > ^5 1c J- r*L E rfl — 3 5 rB C -^ o _^ rfl u k— ! to cc a> -C > ■a ^ lyi ^ 3 o > ' w _^ 2^ CO 3 -T R) ~~3 — J. 3 c 1 O 1/1 ^ ^1 u C T3 C ao c □ rg < — ? *— U < ^ TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 161 60 .c E 3 z I I s ! C I I C 2 3 I 01 at c 3 u c 3 c- 0> eo tu —i co < t— .— .— .— ^-' --' O O ^-' »-' T^ ^ m CO rv O in o o o oo .— .— t in c^ O CN CO co \D ^D r^ vO ^O u~) 0O \D ^O ^D ^O CO § 43 o LT) 1^ in En CO in o O o ' 1 ^~ ' • in r CO 8 U1 o co o CO CO CO t— >— T vO i— cn C cn 8 cn cn cn ^ t in a> S iO i- LO ^O cn >— cn CO *3" O Vj cn TJ- CN m ro^t m •*? m I— r- TJ- rsi TJ u rB co CD E _2 o U > l_ ai u_ 3 . o C <-n ca aj c co oo '^ 3 ca 2 O. CD k- co CO a; c _^ CO on h- E _3 o U eg CO c o « co CU $ co N ai " c -c cu u _i *: £3 < .5 c? E 1£ ■£: ro -± co /•— , C* "=■ TJ -* 3 U O 3 i- ) ~° CU oo ,S ro i/i -* f0 ir> co c _ cu 3 ^£ O- cyi KM \ FISH \\M (.\M[ mm. FIGURE 7. (A] Osteocranium oi Sahefotus confluentus, UMMZ i""J458, head length 112 .. Flathead L, Montana; (8) neurocranium of same individual in dorsal, ventral, and lateral views. V : : • •: ■■■i.-z ire •:: y :~e cere a~c :~e rg - 5 character st ca ■ zrenu ated - Rind on the : sterior ma 2 - - ■£ ac- : 5 "" :' _:.:--' - ~^£ a~~ ~a- c : _ ^r in 5. '£: i'r thout nriuc- [xnarnentatiori and •■• - -:ek an e e-. a: c o* :-e ~ c '~e rs-'£~3 "5. ~i -s:-e;':-;:rr'.: :; af - = :r" "•-•' ~~£ ~£a. . :: .£- ~g c" :a-" • "-_- :.-'-t:';":i; "5. ~~a ~a has given rise fee ~any ridges pits ties as • - a- z^sing the pores of :~£ i-cacc :a :a~a :o ce £ £ -• ccr. :-ces ace .e the surface. - ; with the hyo^ a - d oular, the supraethmoid is also highly diagnostic for the bull trout (Figo'e 9 nstead of being sharply divide: " twop a is I head and posterior extension) by a pronounced constriction, as in S. malma. the lateral margins of the bone are near parallel posteriorly and then aper gradually ard the inter or end The head - the bone is marked b . jst a slight lateral £\ca~5 :r ace: ~a f.'. a. a eg :~£ ate'a ~ar2 - 3rr "re ~eac a-c cere' - extension of the f_ca£:hmoid in S. confluentus are more elongated than in $ vufaa and in this respect the) a: z'Dach the condition ir 5 lamay :^sh. The a-:£- :- ~a-g-s c* :~e as:£-c -^ :: i£fS£s ;*' :-e c-£~a- a£ a::ac^ a \- 1 :-e s _ p -a£:hmoid ~ £a c a — a '£ correspondingly lengthened in 5. confluentus. Again. this is a character a sc -ound in tH- a- £ trout !>£-£_ 'ocranium is depressed in comparison to that of other Safveiinus, with the parasphenoid only slightly flexed (Figure "3 5. malma possesses a well- 164 CALIFORNIA FISH AND CAME FIGURE 9 Comparison or supraethmoids in Sa/vethus mj/mj \-C and Sa '■elinus confluentus (D-f). not drawn to same scale MMZ 126507 286 mm, female. karluk R., • • A • aska; \B] L MMZ -0c:5J 383 mm. female Unalaska I. Alaska: (C) UMMZ To>: -nm remale Unalaska I. Alaska- ID) UMMZ 159333 227 mm, male. Morrison I. Brit. Co! - IE) UMMZ 172458 est. 490 mm Flathead L Montana ' ' " ' "."- : ; est 420 mm Flathead L Montana. developed flexure in the parasphenoid which is consistent with its deeper neuro- cranium. The lateral profile of the neurocranium showing these characteristics can usuallv be seen in radiographs taken with the specimens King flat on their sides When observing the palate in 5 confluentus this species, like 5. malma (Morton and Miller 1954K does not show a well-developed toothed platform TAXONOV^- AND DISTRIBUTION OF THE BLLL TROUT 165 on the vomer. The form of vomerine teeth usual - tshallovt \. Sometimes these teeth are arranged in a single trar-- - . ~here is a ell- f loped gap between the palatine and vomerine tooth rows. This is a consist- ent feature of S. confluentus ■■■ riereas in S. malma it is variable. HYBRIDS Two of nine specimens of bull tro~: I v.'Z 188652 I taken from Long Creek. Lake Co.. Oregon, in the upper Klamath basin were identified bv me as hvbrids with brook trout. Salvelinus fontinalis < Figure 10 > . The hvbrids had much darker pigmentation over the head, bodv, and fins than the other seven specimens, the dorsal fins were mottled, and the lower fins were tricolored. Light spots on the flanks were uniformly smaller than those on the flanks of the bull trout. The maxillary bones of the upper ja . re long and straight, a characteristic of the brook trout. Vertebral counts for the hybrids were intermediate • both with t_ "rreas the Long Creek bull trout ranged from 64 to 66. with a mean of 65. According to \ idyko\ 1954 brook trout ha\ e 58 to 62 \ ertebrae with a mean r9.5. The branchiostegal ray count was high. 13 and 14 on the left side, like that of the bull trou' whereas brook trout usualK have 11 on the left side ko\ 195- * of the Long Creek specimens, including the two hybrids, possessed basibranchial teeth. Hvbridization between the bull trout and the brook trout has also been men- tioned b\ Paetz and Nelson I97C to occur in the Clearwater Riser drainage • \iberta. Two possible hvbrids between S. malma and 5. confluentus were identified from lakes in the Skeena River basin. British Columbia. The first individual L \'\'Z 159328 203 mm. ferr is taken from Swan Lake. It resembles 5. confluentus in the number of branchiostega s 27 and \ ertebrae i66> but is more like S. malma in number of gill rakers 2 ind mandibular pores i " Gill raker charateristics are intermediate between S. malma and S. confluentus. In head form and maxillary shape, the hybrid also resembles S. malma. The second specimen L \'\'Z ' 5 ^333. 188 mm, female' from Morrison Lake - similar in appearance to the Swan Lake individual, especiallv in its head mor- phologv . It has 24 branchiostegals. 6~ \ ertebrae. 20 gill rakers, and 1 5 mandibular pores. Like the first specimen, the gill rakers are also intermediate in shape and structure between those of S. malma and those of S. confluentus. DISTRIBUTION Salvelinus confluentus is distributed in a north-south belt along the Rocky Mountain and Cascade ranges of northwestern North America I Figure 1 }. The area stretches from lat 41" \ to lat 60c \ or slightK be\ond. Localities plotted are about equallv distributed on both sides of the Continental Divide between lat 50; and 60= N. Major river drainages involved in the distribution pattern on the Pacific slope are: The McCloud in the upper Sacramento River basin of California; the upper Klamath in Oregon; the Snake in Oregon, Idaho, and Nevada; the Columbia River throughout its length; the Pend Oreille; the Clark- Fork of Idaho and Montana, including the Flathead; the Kootenav Riv er of British Columbia; the Fra- er of British Columbia, plus Puget Sound iri "ington. and the Skeena and Taku rivers of British Columbia. In the Bering Sea drair _ thev are found in the headwaters of the Yukon at the boundarv between British 166 CALIFORNIA FISH AND CAME FIGURE 10. Two hybrid Salvelinus confluentus X S. fontinalis (below) compared with two 5. confluentus (above); all from same catch by W. Seegrist, Long Cr., Lake Co., Oregon. Photograph by the author. Columbia and the Yukon Territory. No specimens were examined from the Nass, Stikine, or Alsek drainages of the Pacific slope of Canada. McPhail's ( 1 961 ) data indicate bull trout are in Bowser Lake of the Nass drainage. On the east side of the Continental Divide, S. confluentus is found in the headwaters of the South and North Saskatchewan rivers of the Hudson Bay drainage in Alberta and in headwater areas of the Athabaska, Peace, and Liard rivers of the MacKenzie system in Alberta and British Columbia, the latter draining to the Arctic Ocean. The distributional pattern of S. confluentus is largely the result of headwater migration and drainage crossover by stream capture. In the northern half of its range this has occurred following the retreat of the last continental ice sheet. Stream capture of the upper Columbia tributaries by those of the Saskatchewan enabled the bull trout to cross to the east side of the Continental Divide along TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 167 with Salmo clarki and Prosopium williamsoni. Lindsey (1964) has explained transfer of fishes from the upper Fraser to the Peace River by a drainage block at the time the continental ice sheet was retreating from northern British Co- lumbia. It is possible that S. confluentus entered the Taku River by means of migration along the coastal waters, but other species entered the Taku from the east through the Liard tributaries or from the Liard via the Yukon headwaters without any coastal connections (McPhail and Lindsey 1970). S. confluentus is known from collections prior to 1900 to have entered coastal waters in Puget Sound, Washington. Because the bull trout is so generally distributed throughout the Columbia River basin and because this basin borders others where the bull trout is found, it is likely that this species originated there. Transfer to drainage basins in the southern part of its range, such as the Klamath and Sacramento, may have occurred during or following the last glaciation, but it is possible this distribution follows an older pattern, as suggested by the isolated occurrence of 5. confluen- tus'\n the upper reaches of the Sacramento, Klamath, and Snake rivers. However, according to R. Behnke (Colorado State Univ., pers. commun.) the occurrence of S. confluentus in the Snake River drainage above Shoshone Falls ( Hubbs and Miller 1948) is most likely the result of headwater transfer from the Salmon River drainage. The distribution of the bull trout corresponds in many ways to that of the mountain whitefish, Prosopium williamsoni. The latter species differs in having reached the headwaters of the Missouri drainage in northwestern Wyoming and western Montana. It also has crossed into the Bonneville and Lahontan basins. The mountain whitefish is not known from the headwater tributaries of the Yukon drainage where the bull trout is present (Scott and Crossman 1973). There is a possibility that S. confluentus was present at an earlier time in the Bear River of the Bonneville basin. Rostlund (1951) indicates a species of Salvelinus was known in these waters before 1850. If this is true, it may also have entered the Lahontan basin, as did Prosopium williamsoni, but has since disap- peared. The retreat of S. confluentus from the southern extremes of its range is occurring today as it probably has in the past. The gradual change in climate since Late Pleistocene with subsequent loss of water once supplied from moun- tain glaciers and snowfields has been a major factor in eliminating habitats where the bull trout can survive. Once a prominent species in the McCloud River of the Sacramento system in California, the bull trout has gradually declined since the late 1800's and is now close to extirpation. Modifications of the river by man may have accelerated the rate of decline. Since Livingston Stone first called attention to the bull trout of the McCloud River in 1872, very few specimens from the McCloud have been deposited in museums. Fifteen are known to me, including one probably taken on Stone's initial tour of the McCloud. The old USNM specimens were originally preserved in alcohol without fixation in formalin; several are in such a poor state of preservation that they cannot be removed from their glass container. Two speci- mens (USNM 15549) labeled Wye-dai-deek-it, the original Indian name for the bull trout in the McCloud Valley, were collected by Stone in 1874. Seven (in- cluding one skeleton ) were sent by Stone to the USNM between 1 875 and 1 881 . Five specimens were taken singly between 1938-1975; one came from the Mt. Shasta Hatchery at Mt. Shasta in 1956; three small individuals, taken on hook 3—77343 168 CALIFORNIA FISH AND GAME and line between 1938 and 1950, are housed in the California Academy of Sciences; a larger adult taken on hook and line July 19, 1975, is in the collection of the California Department of Fish and Game (CDFG 0513). The exact loca- tions of capture of Stone's bull trout are not known, although they were probably taken in the vicinity of the trout hatchery on Green's Creek ( Wales 1 939 ) where set lines were placed to catch rainbow trout. Wales (1939) presented testimony that the bull trout formerly occurred in the upper Sacramento River in the vicinity of Dunsmuir, Siskiyou Co., and in the Pit River near the mouth of Squaw Creek, Shasta Co. Jordan (1907) listed upper Soda Springs on the Sacramento River as a locality, but Evermann and Bryant (1919) stated the McCloud River is the only stream in California in which the bull trout is known to be native. Campbell (1882) reported bull trout occurred in the McCloud River from its mouth upstream to Big Springs. Big Springs is a source of melt water from snowfields on Mt. Shasta, which yields a nearly constant flow of 7 C (45 F) water (Wales 1939). Although Campbell had fished the entire McCloud River, he never took bull trout above Lower Falls (3.2 km above Big Springs) where he reported water temperatures from 15.5 C to 21 C (60 F to 70 F). Downstream at the fish hatchery on Greens Creek (now inundated by Lake Shasta), Campbell reported water temperatures of 13 C to 15.5 C (55 F to 60 F) at midday in the hottest weather, and from the hatchery up to the Big Springs the river got "one degree colder about every 10 to 12 miles for the distance of 65 or 70 miles." The early records from Oregon and Washington indicate the bull trout was once more widely distributed on the Pacific side of the Cascade and Coast ranges than it appears to be today. Specimens taken from Puget Sound by Jordan and others in the 1880's are assignable to either 5. confluentus or S. malma. The head of Suckley's (1858) type of Salmo confluentus is a bull trout taken from a Coast Range drainage (Pullayup River) that empties into Puget Sound. Cope (1879) reported one of the earliest records for Salvelinus confluentus from the Klamath basin; this is the only previously published record based on an actual specimen from that basin. The earliest confirmed record for S. confluentus is the specimen collected in 1854 from The Dalles on the lower Columbia River. In contrast, this species was not collected from the Snake River drainage of Nevada until much later (Miller and Morton 1952). DISCUSSION I have attempted to provide evidence for the separation of S. confluentus from S. malma, particularly in that part of the range of S. malma southward from southern Alaska on the eastern side of the Pacific basin in North America. The systematics of S. malma throughout its range, including Alaska and the North Pacific basin, is lengthy and complicated and will be treated elsewhere. There are valid reasons for considering the bull trout a full biological species: 1 ) The character states investigated which separate S. confluentus from S. mal- ma are constant throughout a broad geographical area occupied by S. confluen- tus. 2) Stability is further indicated by the fact that variation of all characters investigated is minimal when compared to other recognized species of Sal- velinus. 3) Even though the ranges of 5. malma and 5. confluentus overlap in the Pacific slope drainages from northern California to southern Alaska, there is no evidence of introgression in the material studied. TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 169 Sympatry Sympatry of 5. ma/ma and 5. confluentus occurs in the Puget Sound of Wash- ington, the Skeena Basin of British Columbia, and the Taku Basin of southern Alaska and northern British Columbia. Information available at this time on Livingston Stone's incompletely cataloged USNM specimens indicates that sym- patry between S. ma/ma and 5. confluentus may also have occurred in the Sacramento basin of California. Both S. ma/ma and 5. con fluentus were identified from the Yukon Basin; however, S. ma/ma came from the Tanana drainage of central Alaska and 5. confluentus from the extreme headwaters in British Co- lumbia, Canada. A complete understanding of how S. confluentus and S. malma maintain sympatry will probably come from a study of these species in the Skeena Basin of British Columbia. Both appear to be disseminated throughout these waters, where considerable study of the "Dolly Varden" has already taken place. In the Skeena drainage (Figure 1, insert), specimens of S. confluentus were studied from the following lakes: Lakelse, Damschilgwit (Cabin), Morrison, Morice, Slamgeesh, Bear, and Sustut. Specimens of S. malma came from Johan- son and Alastair lakes. Two suspected hybrids came from Swan and Morrison lakes. Although the hybrids indicate both species probably occur in some lakes, it is possible that certain of the lakes may contain only one of the species pair. Larger samples from these lakes may be helpful in determining to what extent hybridization is taking place. In the taku River, S. malma was taken 22 km (14 miles) upstream from the mouth at Canyon Island where Meehan and Siniff (1962) reported downstream migration of juvenile Dolly Varden in 1961. 5. confluentus was taken further inland in Flannigan Slough, a small tributary to the Taku River. The Taku River collections suggest that S. malma and S. confluentus may be ecologically separated, with anadromous S. malma occupying most of the lower drainage, especially Twin Lakes, the lower main channel, and the river mouth, and S. confluentus restricted to certain small tributaries, such as Flanni- gan Slough and perhaps the headwaters. Of nine specimens examined from Puget Sound, five were S. malma taken near Port Townsend, Washington in 1884. Labeling, as well as their silvery pigmentation, indicated they were sea run individuals. The other four specimens are S. confluentus. Two of these were taken by D. S. Jordan in 1880, and labeled "Puget Sound." The third was taken in Elliot Bay at Seattle in 1 889 and the fourth was captured in a freshwater tributary to Puget Sound near Ft. Steilacoom in 1856. Jordan's specimens are two of ten sent to the United States National Museum and may have been collected during his work on the fishes of Puget Sound (Jordan and Starks 1895). S. confluentus appears to be distributed throughout the MacKenzie basin in British Columbia and Alberta. Unfortunately, no material was available from the Nahanni River, part of the Liard drainage in the Northwest Territories, and only one specimen of S. malma was found in samples from the MacKenzie (USNM 147661). McPhail (1961 ) demonstrated the existence of northern and southern forms of 5. malma in North America. The northern form occurs in the coastal drainages of the Bering Sea north to Seward Peninsula. Specimens corresponding to the northern form have been examined for this study from the Yukon basin of central Alaska. These specimens also resemble the S. malma-Wke char described as 170 CALIFORNIA FISH AND CAME Salvelinus anaktuvukensisby Morrow (1973) from the north slope of the Brooks Range, Alaska. Data from these Salvelinus were not used in making comparisons with S. confluentus. McPhail (1961 ) stated that head size would not distinguish the form which Jordan, Evermann, and Clark (1930) called Salvelinus malma spectabilis. However, if McPhail's data (1961 ) on head size are reexamined in light of the present work, it can be seen that the samples shown to have the longest heads (Cottonwood River, Sage Creek, Glacier National Park, and prob- ably also Bowser Lake in the Nass drainage) are from localities where S. con- fluentus is found. Except for the Taku River, where S. confluentus and S. malma may have been present in a mixed sample, the other samples represent S. malma and show an average smaller head size. Much of McPhail's material was reex- amined for this study. The samples of S. confluentus from the Taku River drain- age were found to be smaller-headed with respect to samples of 5. confluentus from other localities, including the adjacent Yukon basin. Taku River basin 5. confluentus live at the extreme end of the northern dispersal route. Transfer to this basin probably took place from the Liard via the Yukon headwaters or directly from the western reaches of the Liard. Variability At the end of the southern dispersal route in the Sacramento basin, the McCloud River S. confluentus possess the largest head size and are also distin- guished in having a greater percentage of individuals (50% of those examined) lacking basibranchial teeth. Only one specimen outside of the McCloud River was found to lack basibranchial teeth. Jordan ( 1 879 ) at one time recognized the McCloud River population as a separate species, {Salvelinus bairdif), on the basis of this character. Two specimens collected sometime prior to 1877 are part of Livingston Sto- ne's material sent to the USNM, most of which came from the McCloud River or Sacramento River drainage. Small head size identifies them as S. malma. In other characters, such as number of gill rakers, compressed body, straight maxil- la and supraethmoid, they resemble S. malma. Silver pigment on their sides indicated they may have been anadromous individuals. In the shape of their gill rakers and head form they are not typical of 5. malma but the deteriorated condition of these specimens prevented further study. Of the coastal collections, S. malma from Puget Sound, Washington and the Soleduck River of Washington's Olympic Peninsula are geographically closest to the McCloud River specimens. Those taken from Puget Sound in the 1880's were anadromous and compared closely with S. malma from further north along the coast of British Columbia. The Soleduck River specimens represent a popula- tion isolated above a high falls. The largest individual measured 135 mm. They differed from anadromous S. malma in their larger heads and fewer vertebrae and gill rakers. It appears likely that prior to 1900, S. malma ranged south along the Pacific Coast to California and may have been fairly common in Washing- ton's coastal waters of Puget Sound (Jordan and Starks 1895; Dymond 1942). Specialization Radiographs of 1 12 5. confluentus showed about 13% with whole fish in their stomachs. Sculpins predominated in the stomach contents, but at least two bull trout had eaten salmonids. Fewer S. malma had fish in their stomachs. Some S. TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 171 malma had eaten gastropods, a food also seen in the stomachs of three S. confluentus. One 380 mm S. confluentus examined from the Finlay River near Ft. Grahame, British Columbia (NMC 66-437), had swallowed a 21 5 mm individ- ual of its own species. Two bull trout from northern British Columbia had eaten small mammals (one a shrew and the other an unidentified rodent). Fish eating habits of the bull trout have been noted by Ricker (1941 ), Godfrey (1955), Jeppson and Platts (1959), Bjornn (1961), and Thompson and Tufts (1967). Armstrong and Morton (1969) summarized other food habit studies of the bull trout. Scott and Crossman (1973) have recently reviewed the food habits of the bull trout under their section on the Dolly Varden. According to Brown (1971 ), the principal food of the adult bull trout is fish, although it will utilize other vertebrates of suitable size, such as frogs, snakes, mice, and duck- lings. Characters of the jaws, teeth, and head of S. confluentus which differ from those of S. malma are best explained as adaptions to a piscivorous habit. The large area for jaw muscle insertion, large jaws and teeth, the curved maxilla, the elongate hyoid bar with numerous branchiostegal rays, the high number of mandibular sensory pores, and the elongated head of the supraethmoid for articulation of the extended premaxillary ascending processes, are features par- alleled in the lake trout, S. namaycush, a species well-known for its piscivorous diet. The direction of evolution in S. confluentus has been away from a diverse predator, such as anadromous S. malma toward a more specialized fish-eating mode of existence entirely in fresh water. The changes in the head and compo- nent bones are the easiest to document, but there are probably marked behav- ioral and ecological differences between 5. confluentus and S. malma that have not yet been studied. Most of the similarities pointed out above between the bull trout and the lake trout are considered as parallel developments and are not considered to be evidence of a close phyletic relationship. SUMMARY AND CONCLUSIONS The bull trout was first described as Salmo spectabilis in 1856 by Girard from a specimen collected at The Dalles, Oregon, on the lower Columbia River. Girard's specific name "spectabilis" does not stand because of the rules of zoological nomenclature pertaining to homonyms. The name campbelli was substituted by Suckley, who described the bull trout three more times under the names "confluentus", "bairdii", and " parkei." In 1882 Jordan and Gilbert included the above names except confluentus in the synonymy of 5. malma, where they have remained to the present. With formal species recognition for the bull trout, its scientific name becomes Salvelinus confluentus (Suckley); the common name, Dolly Varden, applies to Salvelinus malma (Walbaum). Evidence for specific distinction of the bull trout is found in a series of osteo- logical, morphometric, and meristic characters that remain relatively constant throughout its distributional range including areas of sympatry with S. malma. The size and shape of the head and jaws, head length, the number of branchi- ostegal rays, and the morphology of the gill rakers are the easiest characters to use in separating the bull trout from S. malma. The shape of the supraethmoid bone is also highly diagnostic for the bull trout, but it must be exposed through removal of overlying tissue. The shape of the neurocranium and hyomandibular 172 CALIFORNIA FISH AND GAME are also distinctive. Additional osteological characters are found in the maxilla, premaxilla, ceratohyal, opercle, and frontal bones. The cranial structures in the bull trout that differ from those of the Dolly Varden are interpreted as modifications toward a more piscivorous existence in freshwater. The changes are genetic and have occurred through the long process of selection and adaption. By no means can they be construed as environmental- ly-induced somatic changes. The life history of the bull trout differs significantly from that of the Dolly Varden in being almost completely nonanadromous throughout its known range. Ecological and behavioral differences between the two have not been studied in detail. The bull trout is widely distributed in montane lake and stream habitats on both sides of the Continental Divide between lat 50° and 60° N. It appears to have an affinity for cold waters fed by mountain glaciers and snowfields. In the deglaciated part of its range in western Canada, the bull trout has dispersed from the Columbia River basin through headwater transfer and crossover following the retreat of the Cordilleran ice sheet in Late Wisconsin time. The distribution pattern to the south of the ice sheet may be older. Populations of bull trout in the Sacramento, Klamath, and southern Snake River drainages all show signs of depletion. ACKNOWLEDGMENTS I am deeply grateful to the staffs at the University of Michigan Museum of Zoology, National Museum of Natural Sciences, Canada, and the United States National Museum of Natural History, Smithsonian Institution, for their coopera- tion. Specimens examined in this study were made available by R. M. Bailey and R. R. Miller (UMMZ); D. E. McAllister (NMC); E. A. Lachner and W. R. Taylor (USNM); W. I. Follett, (CAS); R. J. Behnke, Colorado State University; C. M. Barbour, (when at the) University of Utah; R. Wasem, U. S. National Park Service; W. Seegrist, U. S. Forest Service. S. J. Nicola, California Department of Fish and Game, provided information on the bull trout captured in 1975 from the McCloud River. R. J. Behnke, R. R. Miller, R. M. Bailey, S. J. Nicola, and J. E. Morrow critically read the manuscript. I wish to thank these individuals for their considerable assistance. I am particularly indebted to R. R. Miller who supported the early aspects of this study between 1968 and 1970 through his National Science Foundation Grant (GB-4854) on Cenozoic Fishes. This work is an outgrowth of previous studies on fossil salmonids at the University of Michigan. REFERENCES Armstrong, R. H., and W. M. Morton. 1969. Revised annotated bibliography on the Dolly Varden char. Alaska Dep. Fish Came Res. Rep. 7, 107 p. Bailey, R. M., chairman. 1970. A list of common and scientific names of fishes from the United States and Canada, 3rd ed. Amer. Fish. Soc. Spec. Publ. 6, 150 p. Bjornn, T. C. 1961 . Harvest, age structure and growth of game fish populations from Priest and Upper Priest lakes. Amer. Fish. Soc, Trans., 90(1): 27-31. Brown, C. J. D. 1971. Fishes of Montana. Montana State University, 207 p. Campbell,). B. 1882. Notes on McCloud River, California and some of its fishes. U. S. Fish. Comm. Bull. 1, (1881): 44-46. Cavender, T. M. 1969. An early Pliocene Salvelinus from Nevada with comparative notes on recent species. Abstracts 49th Meet. Amer. Soc. Ichthy. Herpet., 1969: 29-30. TAXONOMY AND DISTRIBUTION OF THE BULL TROUT 173 Cope, E. D. 1879. The fishes of Klamath Lake, Oregon. Amer. Natural., 13(12): 784-785. Dymond, J. R. 1932. The trout and other game fishes of British Columbia. Canada Dep. Fish., Ottawa, 51 p. . 1942. The occurrence of the Dolly Varden charr in salt water off British Columbia. Can. Field Natural., 14(7): 110-112. Evermann, B. W., and H. C. Bryant. 1919. California trout. Calif. Fish Game, 5(3): 105-135. Cirard, C. F. 1856. Notice upon the species of the genus Salmo of authors, observed chiefly in Oregon and California. Proc. Acad. Nat. Sci. Philadelphia, 8: 217-220. 1858. Fishes, pp. 1-400. /^general report on the zoology of the several Pacific Railroad routes. U. S. Pacific Railroad Survey, 10(4). Godfrey, H. 1955. On the ecology of the Skeena River whitefishes, Coregonus and Prosopium. Canada, Fish. Res. Bd., )., 12(4): 499-542. Hart, ). L. 1973. Pacific fishes of Canada. Canada, Fish. Res. Bd. Bull. 180, 740 p. Hubbs, C. L., and R. R. Miller. 1948. The zoological evidence/correlation between fish distribution and hydro- graphic history in the desert basins of western United States. In The Great Basin, with emphasis on Glacial and Postglacial times. Bull. Univ. Utah, 38(20), Biol. Ser. 10: 17-166. Jeppson, P. W„ and W. S. Platts. 1959. Ecology and control of the Columbia squawfish in northern Idaho lakes. Amer. Fish. Soc, Trans., 88(3): 197-202. Jordan, D. S. 1879. Notes on a collection of fishes from the Clackamas River, Oregon. Proc. U. S. Nat. Mus. 1878, 1 : 69-85. 1 907. The trout and salmon of the Pacific Coast. Rep. State Bd. Fish Comm. Calif., 1 905-1 906. Sacramento, 112 p. 1923. The name of the Dolly Varden trout, Salvelinus spectabilis (Girard). Copeia, 121: 85-86. Jordan, D. S., and B. W. Evermann. 1896. The fishes of North and Middle America. U. S. Nat. Mus. Bull. 47, Pt. 1. 1,240 p. Jordan, D. S., B. W. Evermann, and H. W. Clark. 1930. Checklist of the fishes and fish-like vertebrates of North and Middle America, north of the northern boundary of Venezuela and Columbia. U. S. Comm. Fish, Rep., (1928): 1-670. Jordan, D. S., and C. H. Gilbert. 1882. Synopsis of fishes of North America U. S. Nat. Mus. Bull. 16: 1,018 p. Jordan, D. S., and E. A. McGregor. 1925. Family Salmonidae. In records of fishes obtained by David Starr Jordan in Japan, 1922, by D. S. Jordan and C. L. Hubbs. Memoirs Carnegie Mus., 10(2): 93-346. Jordan, D. S., and E. C. Starks. 1895. The fishes of Puget Sound. Contributions to biology from the Hopkins Laboratory of Biology III, Leland Stanford Jr. Univ. Publ.,: 785-855. Lindsey, C. C. 1964. Problems in zoogeography of the lake trout, Salvelinus namaycush. Canada, Fish. Res. Bd., J., 21(5): 979-994. McPhail, ). D. 1961. A systematic study of the Salvelinus alpinus complex in North America. Canada, Fish. Res. Bd., Jr., 18(5): 793-816. McPhail, ). D., and C. C. Lindsey. 1970. Freshwater fishes of northwestern Canada and Alaska. Canada, Fish, Res. Bd., Bull. 173, Ottawa, 381 p. Meehan, W. R., and D. B. Siniff. 1962. A study of downstream migrations of anadromous fishes in the Taku River, Alaska. Amer. Fish. Soc, Trans., 91 (4): 399^07. Miller, R. R., and W. M. Morton. 1952. First record of the Dolly Varden Salvelinus ma/ma from Nevada. Copeia, 3: 207-208. Morrow, J. E. 1973. A new species of Salvelinus from the Brooks Range, northern Alaska. Univ. Alaska Biol. Pap. 13: 1-8. Morton, W. M. 1955. Charr or char — history of a common name for Salvelinus. Science, 121 (3155): 874-875. 1965. The taxonomic significance of the kype in American salmonids. Copeia, 1965 (1): 14-19. 1970. On the validity of all subspecific descriptions of North American Salvelinus ma/ma (Walbaum). Copeia, 1970 (3): 581-587. Morton, W. M., and R. R. Miller. 1954. Systematic position of the lake trout, Salvelinus namaycush. Copeia, 1954 (2): 116-124. Paetz, M. J., and J. S. Nelson. 1970. The fishes of Alberta. Prov. of Alberta, Edmonton. 281 p. Ricker, W. E. 1941. Consumption of young sockeye salmon by predaceous fish. Canada, Fish. Res. Bd., J., 5(3): 293-313. Rostlund, E. 1951. Three early historical reports of North American freshwater fishes. Copeia, 1951 (4): 295-296. Scott, W. B., and E. J. Crossman. 1973. Freshwater fishes of Canada. Canada, Fish. Res. Bd. Bull. 184, Ottawa, 966 P- Shaposhnikova, G. K. 1971. Comparative morphological description of certain species of the genus Salvelinus (Nilsson) Richardson. Tr. Zool. Inst. Leningrad, Acad. Sci. USSR, 48 : 4-29. 174 CALIFORNIA FISH AND CAME Suckley, C. 1858. Descriptions of several new species of Salmonidae, from the North-West coast of America. Ann. Lye. Nat. Hist., 7 : 1-10. 1860. Report upon Salmonidae, Chapter I. In Pacific Railroad reports, fishes, 12 (5) : 307-349. 1861. Notices of certain new species of North American Salmonidae, chiefly in the collection of the N. W. Boundary Commission in charge: Archibald Campbell, Esq., Commissioner of the United States, collected by Doctor C. B. R. Kennedy, Naturalist of the Commission. Ann. Lye. Nat. Hist. 7 : 306-313. Thompson, R. B., and D. F. Tufts. 1967. Predation by the Dolly Varden and northern squawfish on hatchery-reared sockeye salmon in Lake Wenatchee, Washington. Amer. Fish. Soc, Trans., 96(4) : 424—427. Vladykov, V. D. 1954. Taxonomic characters of the eastern North American chars (Salvelinus and Cristivomer). Canada, Fish. Res. Bd., J., 11(6) : 904-932. Wales, J. H. 1939. General report of investigations on the McCloud River drainage in 1938. Calif. Fish Came, 25(4): 272-309. REPRODUCTION AND SPAWNING OF NORTHERN ANCHOVY 175 Calif. Fish and Came 64 ( 3 ): 1 75-1 84 1 978 REPRODUCTION AND SPAWNING OF THE NORTHERN ANCHOVY, ENGRAULIS MORDAX, IN SAN PEDRO BAY, CALIFORNIA1 GARY D. BREWER Allan Hancock Foundation University of Southern California Los Angeles, California 90007 3 The gonosomatic index (GSI) was computed for samples of adult northern an- chovy collected from the Los Angeles-Long Beach Harbor live-bait fishery during 1973 and 1974. GSI values were highest in February and March of both years and declined to the lowest values in September. The abundance of anchovy eggs and larvae in plankton samples taken from San Pedro Bay during the same period was also highest in February and March and lowest during August and September. Field and laboratory data suggest that within an environmental temperature range of 13 to 18 C (55 to 66 F), northern anchovy have the potential to breed all year, but are constrained to a seasonal reproductive cycle by dietary requirements that exceed available production of zooplankton. INTRODUCTION Environmental investigations of the Los Angeles-Long Beach Harbor and San Pedro Bay by the Allan Hancock Foundation, University of Southern California, recognized the northern anchovy, Engraulis mordax Girard, as a key species, both ecologically and economically, in southern California nearshore waters (Soule and Oguri 1972-1976). This understanding prompted a summary of the biology and fishery of £ mordax (Brewer 1975a) and a study on the thermal tolerance and resistance of the fish (Brewer 1976). This paper is an outgrowth of those studies. Data are given on the reproductive cycle and spawning of the anchovy in nearshore waters off southern California. The results are discussed in relation to environmental factors which may control anchovy spawning throughout the fish's geographic range. The northern anchovy is probably the most abundant fish in the California Current (Mais 1974). A dramatic increase in anchovy biomass during the past decade ( Smith 1 972 ) has led to an intensive effort to understand the biology and population dynamics of this relatively unexploited resource (Calif. Dept. Fish and Game 1971). Comprehensive data on the reproduction of commercial fishes is a prerequi- site for intelligent fishery management. Understanding the environmental factors that influence spawning seasons and regulate fecundity in fishes is necessary if fishery stocks are to be predicted and exploited wisely. Such information on northern anchovy was, in part, detailed by Bolin (1936), Ahlstrom (1959, 1966, 1967), MacGregor (1968), Leong (1971), and Mais (1974). While Ahlstrom's studies have compiled an enormous data bank on the seasonal occurrence and distribution of anchovy spawning in offshore waters, California's bays, harbors, and estuaries have been neglected. In light of man's increasing alteration of shallow marine habitats for various industrial and recreational pursuits, it is essential to understand the role that these areas play as spawning and nursery grounds for fishes. ' Accepted for publication February 1978. 176 CALIFORNIA FISH AND GAME MATERIALS AND METHODS Samples of adult £ mordax were obtained from the Long Beach, California, live-bait fishery between February 1973 and September 1974. The anchovy live-bait fishery was described by Wood and Strachan (1970). Fish were meas- ured to the nearest millimeter of standard length (sl) and weighed to the nearest 0.01 g after being lightly blotted. The left gonad of each fish, which is generally larger than the right, was excised, blotted lightly, and weighed to the nearest 0.01 g. The gonosomatic index (GSI) was computed for each individual according to the following formula: wet gonad weight/ wet specimen weight X 104. Varia- tions of the GSI have been used for a number of species as a measure of reproductive maturity (Moser 1967; Khanna and Pant 1967; Mclnerney and Evans 1970; Haydock 1971; Kaya 1973). By following monthly changes in the weight of the testes and ovaries, as expressed by the GSI, reproductive matura- tion can be estimated without time-consuming histological examination of the gonads. Bimonthly plankton collections for anchovy eggs and larvae were initiated in February 1973 at 15 stations in the outer Los Angeles-Long Beach Harbor. Later, more stations were added within the main channels of the inner harbor as well as adjacent areas outside the harbor. Eventually, 22 stations were monitored (Figure 1, Table 1). Each of the 561 plankton collections was standardized. Samples were taken with a 0.5-m (1.7-ft), 222/x mesh (nylon) conical net, towed at about 2 knots at a depth of 4 m (13 ft) (calculated by wire angle) for Pacific Ocean I I I I I I I I I I I I I FIGURE 1. Los Angeles-Long Beach Harbor and San Pedro Bay with plankton tow station loca- tions. REPRODUCTION AND SPAWNING OF NORTHERN ANCHOVY 177 Station no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 TABLE 1. Plankton Tow Station Data. Bottom depth (m) 12 12 9 6 11 20 16 12 7 11 16 12 7 19 19 11 11 11 7 21 23 36 Inclusive trawl dates Feb. 1973-May 1974 Feb. 1973-Sept. 1974 Feb. 1973-Sept. 1974 Feb. 1973-Sept. 1974 Feb. 1973-May 1974 Feb. 1973-May 1974 Feb. 1973-May 1974 Feb. 1973-May 1974 Feb. 1973-May 1974 Feb. 1973-May 1974 Feb. 1973-May 1974 Feb. 1973-May 1974 Feb. 1973-May 1974 Feb. 1973-May 1974 Feb. 1973-May 1974 Apr. 1973-Aug. 1974 Nov. 1973-Aug. 1974 Jan. 1974-Aug. 1974 Mar. 1974-Aug. 1974 Mar. 1974-Aug. 1974 July 1974-Sept. 1974 July 1974-Sept. 1974 5 min. A calibrated rotometer, located inside the mouth of the net, was used to determine the volume of water filtered during 40 of the standardized tows. The mean volume of water strained was calculated to be 50.5 m3 per tow (range, 43-57; standard deviation, 5.1; standard error of mean, 0.8). All samples were taken during daylight hours, with the exception of one complete series. Samples were preserved in 5% formalin solution and later sorted with the aid of a dissecting microscope. Large samples were aliquoted with a Folsom Splitter. RESULTS Standard lengths, weights, and gonosomatic indices were computed for 359 female and 280 male £ mordax sampled during 1973 and 1974. The female-male ratio of 1.28:1 is close to the estimate by Clark and Phillips (1952) for the 1947-1951 live-bait fishery (1.2:1), and the statewide estimate of 1.27:1 given by Miller etal. (1955). However, Klingbeil (1977) suggests spatial and temporal segregation of the sexes in anchovies which may lead to sampling bias when specimens are obtained from commercial or bait fishermen's nets. The highest individual GSI values (male, 524; female, 751) and the highest mean GSI values occurred in February and March of both years and gradually declined to the lowest values in September (Figure 2). A very rapid increase in the GSI values occurred between January and February. Although fish with seemingly ripe gonads (large, yolked eggs plainly visible in ovaries; testes en- larged and cream-colored) were found throughout the year, the proportion of fish with ripe gonads was highest in February and lowest in late summer. Female anchovies as small as 81 mm (3.2 inches) sl were found with well developed ova, and males as small as 78 mm (3.1 inches) sl were found with enlarged testes. Anchovies within this size range are probably 1 year old (Collins 1969). 178 CALIFORNIA FISH AND CAME 6OO-1 500' 400- x LU o Z u 300- < o m O Z o O 200' 150- 100- 75- 50 25- RANGE FEMALE T .MALE • MEAN M A — r- M N MONTH FIGURE 2. Conosomatic indices for Engraulis mordax sampled from the Los Angeles-Long Beach Harbor live-bait fishery (original data in Brewer 19756 ). Although spawning was recorded in every month, eggs and larvae occurred most frequently and were most numerous in February and March when 61.5% of all eggs and 68.6% of all larvae were taken (Tables 2 and 3; Figure 3). Tows made during the period between February and May captured 80% of all eggs and 81.4% of all larvae. This same period included 50.5% of all trawls taken. The greatest number of anchovy eggs taken during one trawl was 1720 (May 1974, station 20), equivalent to 34,059 eggs per 1000 m3 of water filtered. The greatest number of larvae taken during one trawl was 812 (March 1974, station 14), or 16,079 larvae per 1000 m3 of water filtered. During the entire 20-month REPRODUCTION AND SPAWNING OF NORTHERN ANCHOVY 179 towing period, San Pedro Bay yielded 536 eggs and 128 larvae per 1000 m3 of water filtered. TABLE 2. Eggs and Larvae of Engraulis mordax taken in Los Angeles-Long Beach Harbor and San Pedro Bay, California, February 1973-September 1974, Summarized by Month. No. of Month tows January 36 February 51 March 68 April 72 May 92 June 30 July 37 August 37 September 38 October 32 November 34 December 34 TOTALS 561 No. of No. of Mean No. occurrences specimens per , 1000 m3 Eggs Larvae Eggs Larvae Eggs Larvae 2 2 4 2 2 1 31 24 4054 1098 1574 426 60 60 5189 1471 1511 428 42 37 863 377 237 104 45 35 3214 258 692 56 1 0 1 0 1 0 13 5 326 15 174 8 11 6 980 151 524 81 18 12 403 22 210 11 12 1 20 1 12 1 6 23 129 155 75 90 5 16 7 68 4 40 246 221 15190 3618 TABLE 3. Eggs and Larvae of Engraulis mordax taken in Los Angeles-Long Beach Harbor and San Pedro Bay, California, February 1973-September 1974, Summarized by Station. Station 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 TOTALS No. of tows 32 35 35 35 32 31 32 32 31 32 32 32 32 30 35 29 15 11 6 6 3 _3 561 No. of occurrences Eggs 10 18 14 14 15 15 10 10 10 15 13 19 14 20 19 5 5 4 5 5 3 _3 246 Larvae 12 16 14 10 9 13 14 10 14 14 14 19 14 20 19 5 8 4 4 5 2 _2 221 No. of specimens Eggs Larvae 128 79 466 209 434 229 361 122 267 105 370 175 272 62 736 70 247 105 192 127 417 98 1246 293 846 153 2927 930 3101 614 8 11 42 22 7 10 94 23 2195 130 348 29 486 22 Mean No. per 1000 m3 Eggs 79 264 246 204 165 236 168 455 153 119 258 771 523 1932 1754 5 55 13 310 7244 2297 3208 Larvae 49 118 130 69 65 112 38 43 65 79 61 181 95 614 347 8 29 18 76 429 191 145 15190 3618 Some of the increase in the number of eggs spawned during the summer months resulted from increased effort during July, August, and September 1974, with the initiation of off-shore stations 21 and 22 (Figure 1 ). Some 49% of all eggs collected during the summer months of 1973 and 1974 were taken from a 180 lOOOn 500' 4 ioo-r 90- < 80- 0£ uu °- 70- LU < 60- > > O I 50' U z < 40.05). 188 CALIFORNIA FISH AND CAME TABLE 1. Mortality by Hooking Area and the Control Group Hooking area Control group Roof of Floor of Lip mouth Gills Tongue mouth Esophagus (50 Fish) No. of fish hooked 50 50 50 35 50 50 0 No. of immediate mortalities 0 0 0 0 0 20 0 ( < 24 hours) No. of delayed mortalities.... 2 0 10 1 8 1 ( > 24 hours) Total mortality 2 0 10 1 28 ' 1 Percent mortality 4.0 0.0 2.0 0.0 2.0 56.0 2.0 95% C.I. for percent total mortality 0.0-19.0 0.0-7.0 0.0-14.0 0.0-7.0 0.0-14.0 41.0-70.0 0.0-14.0 ' Does not include one mortality attributed to disease. DISCUSSION While anglers using certain types of small baits (worms, salmon eggs, etc.) hook fish in areas of the mouth unlikely to cause death, they are more likely to deep-hook small bass than anglers using lures or minnows. However, the result- ant mortality is probably minor since anglers using these baits catch compara- tively few small bass. Of the sublegal largemouth bass hooked and released at Merle Collins Reservoir during 1974 and 1975, only 12.0% were caught by anglers using small baits exclusively. Conversely, anglers using lures or minnows released 58.9% of their catch during the same period (unpublished data). The size of the hook used in this study represents the average hook size used by most anglers. It is probably larger and more damaging to fish than hooks normally used by anglers fishing small baits. Recorded losses of esophageally- hooked fish under study conditions may, therefore, be greater than those which normally occur. I did not test the effects of summertime temperatures on hooked fish. Results of this study strongly suggest, however, that hooking mortality is not a factor which materially reduces the value of size limit regulations. ACKNOWLEDGMENTS Robert R. Rawstron prompted the study and was instrumental in its design. Stephen A. Rapp and William C. Robinson ably assisted in the hooking opera- tion. Vida B. Wong prepared the figure. REFERENCES Funk, John L, ed. 1974. Largemouth bass harvest in the midwest, an overview. Symposium on overharvest and management of largemouth bass in small impoundments. North Cent. Div., Amer. Fish. Soc, Spec. Pub. No. 3, July 1974: 116 p. Hashagen, Kenneth A., Jr. 1973. Population structure changes and yields of fishes during the initial eight years of impoundment of a warmwater reservoir. Calif. Fish Game, 59(4): 221-244. Rawstron, Robert R., and Kenneth A. Hashagen, Jr. 1972. Mortality and survival rates of tagged largemouth bass (Micropterus salmoides) at Merle Collins Reservoir. Calif. Fish Came, 58(3): 221-230. Rutledge, William P. 1974. Performance report-hooking mortality study. Fed. Aid Proj. F-23-R-3, Hearts of the Hills Fish. Expt. Res. Sta., Texas Parks and Wildlife Dept., Austin, Texas. 5p. DUNGENESS CRAB CPUE 189 Calif. Fish and Came 64 ( 3 ): 1 89-1 99. 1 978 CATCH-PER-UNIT-OF-EFFORT STUDIES OF NORTHERN CALIFORNIA DUNGENESS CRABS, Cancer magister^ DANIEL W. COTSHALL Operations Research Branch California Department of Fish and Game P. O. Box 98 Avila Beach, California 93424 Northern California commercial Dungeness crab fishermen were interviewed to determine their catch-per-unit-of-effort. The data were used to calculate population size and mortality rates of legal sized crabs and to predict season landings. Generally, the mean catch per trap increased as the number of fishing days increased. As the season progressed, fishermen tended to fish their traps for longer periods. The calculated weight of legal male crabs at the beginning of the season ranged from a low of 1.1 Mg (2.4 million lb) in November 1971 to a high of 7.5 Mg (16.5 million lb) in November 1969. Instantaneous total mortality rates (z) ranged from —0.00490 during the 1966-67 season to —0.18300 during the 1971-72 season. The fishermen probably harvest a greater percentage of the available crabs during seasons of high abundance. INTRODUCTION Accurate and consistent catch-per-unit-of-effort data are vital to population dynamics studies. California Dungeness crab fishermen are not required to submit records of daily catch and effort. In December 1964 we initiated inter- views with commercial crab fishermen in Eureka to determine the number of traps fished and the weight of crabs caught. A similar study was undertaken at Crescent City in January 1966 (Gotshall and Hardy 1969). Data were collected from both areas through the 1971-1972 season. Originally, these data were used to obtain estimates of the abundance of legal sized crabs and to predict season landings. It soon became apparent from the data that population size estimates and mortality rates of legal sized crabs could be calculated. We also distributed log books (voluntary) to cooperative fishermen beginning in 1966. Unfortunate- ly, very few of the fishermen kept consistent data and this experiment was terminated during the 1968-69 season. METHODS Commercial fishermen were interviewed whenever possible as they were unloading the day's catch at the dock. Eureka interviews were conducted by Department of Fish and Came biologists and seasonal aids. Crescent City interviews were conducted by the Pacific Marine Fisheries Cooperative Port Sampling Project biologists (Public Law 88-309) from January 1966 through May 1968. From December 1968 through February 1972 these interviews were conducted by Department biologists from Eureka. The information collected during each interiew included: number of traps pulled that day, number of days traps had fished (1 day equals approximately 24 hr of fishing), location, depth traps were fished, and weight of crabs landed. RESULTS From December 1964 through February 1972, 1,460 interviews were con- ' Accepted for publication March 1978. 190 CALIFORNIA FISH AND GAME ducted at Eureka, and 1,230 at Crescent City (Tables 1 and 2). These numbers do not include interviews where the fishermen had pulled traps fished for two or more different time periods. Generally, the Eureka fishermen fish an area bounded on the south by Cape Mendocino and on the north by Dry Lagoon ( Figure 1 ) . Crescent City fishermen do most of their fishing between Big Lagoon and the Oregon border. Monthly mean catch per trap for one day's fishing at Eureka ranged from a high of 9.8 kg (21.7 lb) per trap in May 1966 (one interview) to a low of 0.4 kg (0.9 lb) per trap in February 1972. These figures represent 14.3 and 0.5 crabs per trap, respectively, based on mean weight data (Table 3). TABLE 1. Number of Crab Fishermen Interviewed and Mean Catch (Kg) of Crabs Per Trap by Days Fished, Cape Mendocino to Big Lagoon, December 1964 - February 1972. Days fished 1 2 3 4 5 No. of No. ot No. of No. of No. of Inter- Kg/ Inter- Kg/ Inter- Kg/ Inter- Kg/ Inter- Kg/ Month views trap views trap views trap views trap views trap 1964- 65 Season December 23 5.3 - - 1 10.1 - - - - January 13 3.8 12 5.6 2 4.6 2 5.1 3 9.6 February 4 3.1 9 4.1 2 3.7 2 2.8 10 4.5 March 13 4.4 4 1965- 2.6 66 Season 2 1.7 — — 1 3.9 December - - 11 10.8 - - - - 1 - January 7 9.5 4 10.7 5 12.8 2 12.7 4 11.8 February 3 7.8 9 9.8 1 9.4 1 7.8 4 11.0 March 3 6.5 7 5.6 6 6.5 6 8.9 1 5.8 April 1 7.9 1 10.9 - - - - 1 19.1 May 1 9.9 4 1966- 12.2 67 Season — — — _ — — December 80 6.9 21 7.8 4 7.0 - - 2 6.6 January 17 4.4 5 6.9 1 10.9 1 8.5 2 4.7 February 7 3.8 2 6.2 1 1.6 1 7.6 1 8.1 March 4 5.9 1 3.2 - - - - - - April 5 3.3 5 2.7 1 8.4 - - 1 6.7 May 1 1.0 3 1967- 3.5 68 Season — ~~ 1 5.4 December 20 7.6 3 6.6 5 8.9 1 6.0 4 8.1 January 31 5.6 6 8.5 2 6.5 - - 3 7.3 23 4.7 4 4.3 3 4.0 - - 3 6.5 March 15 3.4 7 3.1 4 3.7 1 1.7 3 2.8 April 10 2.1 9 5.1 5 5.0 10 5.6 10 3.7 May 5 2.6 7 3.3 4 2.6 1 3.2 8 5.6 1968-69 Season December 27 6.0 15 11.7 4 9.8 - - 2 7.7 January 65 5.9 42 5.5 11 6.0 4 4.4 5 4.5 February 10 3.5 8 3.0 8 2.5 9 2.8 14 3.2 March 19 2.3 20 2.0 22 2.3 4 2.0 5 2.8 April - - 5 3.6 6 4.0 - - 8 2.2 May 8 3.9 2 8.6 1 11.4 - - 6 13.6 June - - 1 3.9 1 1.5 - - - - 1969-70 Season * January 120 8.8 1 9.0 - - - - - - February 71 4.6 48 5.3 7 5.5 4 5.3 2 1.9 March 17 4.3 29 3.2 21 3.1 4 3.2 12 27 April 3 1.5 2 2.9 5 2.7 - - 8 2.5 DUNGENESS CRAB CPUE 1970-71 Season ** February 48 5.5 17 5.7 1 March 25 2.6 16 2.5 10 April 5 1.6 May - - 1 1971-72 Season December 20 3.2 13 3.9 14 lanuary 2 1.0 3 1.3 3 February 2 0.4 1 0.8 * Season opened December 1, fishing did not begin until January. ** Season opened December 1, fishing did not begin until February. 191 9.4 - - - - 2.9 7 2.8 5 2.1 - 1 1.5 4 1.2 2.7 - - - - 5.0 2 5.6 7 3.3 1.7 2 0.7 6 1.5 - - - 1 0.3 TABLE 2. Number of Crab Fishermen Interviewed and Mean Catch (Kg) of Crabs Per Trap by Days Fished, Dry Lagoon to Oregon Border, December 1966-December 1971 Days fished 1 2 3 4 5 No. of No. of No. of No. of No. of inter- Kg/ inter- Kg/ inter- Kg/ inter- Kg/ inter- Kg/ Month views trap views trap views trap views trap views trap 1966-67 Season December 38 6.1 9 6.3 6 8.5 3 6.8 1 10.7 January 9 9.4 26 6.4 10 6.0 10 6.1 12 6.4 February 36 6.5 25 5.5 10 6.0 6 6.0 6 8.9 March 18 5.9 13 4.3 15 6.6 2 6.5 4 11.0 April 1 0.8 1 1.1 1 2.8 1 1.9 1 0.9 May - -- -- -- -- June - 3 2.8 1967-68 Season December 49 9.3 17 11.7 7 10.8 1 25.0 31 12.5 January 40 7.6 24 9.8 9 12.2 8 11.8 11 11.0 February 25 3.6 55 5.0 18 5.3 9 7.9 20 9.2 March 1 4.3 6 4.9 4 3.6 4 3.4 10 5.6 April - 4 2.8 8 4.1 2 3.7 13 4.1 May 1 1.4 1 1.4 3 2.9 9 3.5 1968-69 Season December 41 9.5 9 9.4 1 12.5 2 11.9 6 12.0 January 26 6.3 12 5.8 8 4.8 4 5.8 7 5.4 February 4 2.0 4 3.1 5 2.5 2 3.4 13 4.4 March 3 1.8 14 2.5 16 2.4 5 3.5 18 3.9 April 3 3.3 7 2.9 4 2.4 9 4.0 May 3 2.2 4 7.1 1 1.7 2 7.2 18 7.0 June - -- -- -- -16.4 1969-70 Season * lanuary 46 10.5 16 15.2 10 15.8 2 22.1 8 17.4 February 17 4.9 12 3.4 4 5.5 3 4.7 7 5.3 March 1 1.3 4 3.6 8 1.5 3 1.7 7 3.5 April 4 4.3 9 4.1 5 3.3 9 5.8 May - 1 3.0 - - - 1 5.7 1970-71 Season ** February 43 10.9 2 7.4 3 17.4 2 17.2 1 19.1 March 31 5.7 9 6.1 3 5.1 2 4.3 April 2 4.5 1 5.6 1 4.3 May 1 5.4 June - 1 4.2 1971-72 Season December 16 8.9 15 2.7 6 3.1 2 3.7 3 0.4 * Season opened December 1, fishing did not begin until January. ** Season opened December 1, fishing did not begin until February. 192 CALIFORNIA FISH AND GAME 4: 41. 40°. 39' vPoint St. George .Crescent City ) Stone Lagoon 'Dry Lagoon 'Big Lagoon Albion 30 Nautical Miles Point Arena 125" 124° FIGURE 1. Northern California fishing areas and landmarks. 123 DUNCENESS CRAB CPUE 193 TABLE 3. Monthly and Seasonal Mean Weights (Kg) of Dungeness Crabs Landed at Crescent City and Eureka, 1964-1972 Season Month 1964-65 1965-66 1966-67 1967-68 1968-69 1969-70 1970-71 1971-72 Eureka December 0.91 0.94 0.85 0.91 0.82 January 0.88 0.91 0.94 0.74 0.91 0.73 0.83 February 0.86 0.88 0.98 0.84 0.93 0.80 0.92 March 0.89 0.78 0.90 0.73 0.86 April - - 0.99 0.71 0.94 0.77 0.83 May 0.69 0.91 0.82 0.94 June ______-- Mean for Season 0.87 0.86 0.95 0.78 0.92 0.75 0.86 0.83 Crescent City December 0.94 0.83 0.89 0.77 January 0.94 0.99 0.84 0.92 0.89 0.81 February - 0.77 0.94 0.82 0.91 0.85 0.96 March 0.84 0.93 0.78 0.93 0.84 0.94 April - 0.78 0.89 0.79 0.86 0.84 0.92 May - 0.76 0.84 0.94 June 0.75 0.84 - Mean for Season 0.81 0.94 0.79 0.90 0.86 0.94 0.78 The highest monthly mean recorded by Crescent City fishermen for 1 day's fishing occurred in February 1971 and was 10.8 kg (23.9 lb) or 11.3 crabs per trap (Table 3). Generally, the monthly mean catch per trap increased as the number of fishing days increased, particularly during the first month of the season. During most seasons the catch-per-unit-of-effort declined steadily each month and then in- creased again in March, April, or May. As the season (December 1 to July 15) progressed, fishermen tended to fish their traps for longer periods. Approximately 70% of the fishermen pulled their traps every day during the 1 st month of the season ( Figure 2 ) . However, by the 5th month this figure dropped to 12.3% and increased slightly to 19.1% during the 6th month. Conversely, only 8.1% fished their traps 5 or more days during the 1st month but, by the 7th month, this figure increased to 66.7%. It appears that the fishermen could have caught more crabs had they pulled their traps every day, particularly after the 1st month of the season. For example, in March 1969 Eureka fishermen averaged 2.3 kg (5.1 lb) or 2.6 crabs per trap overnight and 2.3 kg (5.1 lb) per trap after three nights. If they had pulled their traps every day, they could have averaged 6.9 kg ( 1 5.3 lb ) or 7.7 crabs; assuming, of course, that weather conditions permitted fishing every day. Population Size A population size estimate of legal sized crabs at the start of the fishing season was calculated from the combined Eureka and Crescent City catch per trap data for 1 day of fishing, and cumulative landing data (Table 4). The following equations were used, based on Leslie's least square method (Ricker 1975): Y = a + bX and N„ = a/c and c = —b 194 CALIFORNIA FISH AND CAME 3C * c Jl o u in ** It n) O U 1/5 ■o DC c C Ifl IE ^^ £ 0) iZ "O SJ w f 0 #rf CO •*- c o o DC ec c 'E 0 c '5c o o> ** co a c 'Z *- 0 ^- "O ffl s •- £ c S-.2 u ^ w 3 ecjjj I 3 01 O > ** '■5 "o 2 >. ex. $ § 9j > in ■C; So -5 -^~ 1 c -5 U N. C4. $ 2 Qj 1, <*) ^ 6t> -5 .sr 9 "S 1 -5 inom CT^mo^l'mmLnc'Nj coiniri(Ncovflc>eo in in co n "f" cK o> i-C »— co o cr> m t i- «— (N o oo in co c^ »-_ r-^ *— " rs" rn rn o i\ m o^i tJ- o> cn ^O m rs o r\ o^ o> ■— »— rs CO rs O O CO o o> •* ^- ^r in tt Tt ^D in m" co o^" r-v o ^t cOfNroncofnMinrnfnm t o •- n vo k co q >- fN ct^ " tt" tt ^ ^*" ^-" in in in in ^O^COOC^OCOOON^C^OinCTiiAifln cdc^O^^Nco^inin^'inr^rn^^ni-'rn N co O »- m O CO ■*£ Q »— inONroinneon fNfNO^^,^-NC^v^^^OM>i)OCOCr''^,O^N r\ fN co (Ti cr^ co c^ 0> ro rs" rs cK rsT o^ cd" 0^0,:l'^)0>'X)rsjo>cOLnrooOinm'^- m*— i— coiococOTrc^^OC^^^OfN ^ 5 *— ▼— (Nn^itinmocococ^OO rs rs OinNNC^OMN'-roOrMncqvCN n N co ^ ^ d in '^ ^ W 'l' <*n rn r>i f-' ro 8«n Tt- m m n ^o •- co^KfsinfNOO^coo co co ^t cn o*» < — mnic^cO'-(NOCT>cocOLn r^rNroc^C^in^c^»-NmininN'-^,0^0 o" in ^o rs" rC »-^ rC rsT tj^" o> »- ^v cr^ k <*n o" rC i-KmnCTi^NO^'CpKtNr-KmN^ *— o>^— cOCOr^cO'tr^Tf'CT'ifNTtco^r- L*^ e¥\ rsT tj^ in o" o rC co" co" cd" cS" cr^" o> o o o OO^t-Orn^tOOrsi C^C^^CTii/ICONvO' — rsc^cT>^KC^^Orncornc^^;cococ^^;^D din^'i^fn^ nsi^'inMf^rOr^(N(NdfN>C^C^CO,T »-K3cOCO(Nrn^KO>>Tm'tOinC^(NNrninNNN 00^ fN^ o^ *— ^b *$■ o^ <•— N N K C0v fN cK in" o" *— ■ ^ O O rC rs C^ (N N fO LA ^" in" vd" Tt" in" in rsT ^-~ »—* &? in" rC ^~ ^r" *— ro" ^t _ OCT^fNr-'tOnNNKiA'-^'tC^'-NNN'-Lnn m c^ fN r- mfNinfNOmior- inocorooo^,^",^co»— »— " ro" ro ^f" ^* in" ^n ^d" ^d" h^" N CO CO CO Cfi CT> CT> . vd" tj- csi in i— rs -^- o CO o" o" o o" o" o^^j-^- co^tcOfnq^fNfNc^^''- fN co co co cdinincdinin^cr^^in^incdcdininOt— , O^ Q N ^D t 't C^ wincoo^fno^fOt ^ COmfNinfNCTiKv£fN>J) 3 3 8^ mu-i-— N 0O fN _ i- K N CN 1^ »— i — •— fN m rn r^ O N C^ t CMO N ■^ *3-~ ^-" in Ln ^D co •— fNmfinot^coo^O"— fNrol-tnot^ coo^O"— rN n f in .— .— fN IN (N (N IN IN vJ3 r~^ fN fN DUNCENESS CRAB CPUE 195 695 First M onth 8.1 460 Second Month 92 25.8 Third Month 19 5 246 Fourth Month 193 50 28.6 305 UJ 98 61 pili, 23 5 > A 62 en 16 2 1- Z 12 0 8 7 | 1 7 80 U_ o o Fifth Month 32.6 19 1 Sixth Month 447 0 Seventh Month 66 7 1 2 3 4 54 < §» LlI Q. 25- 22? 203 223 12 3 123 74 64 16 7 16.7 0 , 1 1 2 3 4 5 + i 2 3 4 5+ 1 2 3 4 5+ DAYS TRAPS FISHED FIGURE 2. Percentage of crab fishermen interviews and number of days traps were fished, Cres- cent City and Eureka dates combined, 1964 through 1972. Where: Y = mean catch in pounds per trap for 1 day of fishing a = intercept of line on Y axis b = slope of line X = cumulative landings a — productof the original population, N0 and the catchability, c — b='catchability, c N„ = estimated population in pounds of legal sized crabs at the start of the fishing season Population estimates were calculated for the 1966-67 through the 1971-72 season (Table 5). The largest calculated population estimate, 7.6 Mg (16.7 million lb), was present at the beginning of the 1969-70 season, the lowest population estimate, 1.1 Mg (2.4 million lb), was obtained from data collected during the early part of the 1971-72 season. TABLE 5. Calculated Population Estimates of Northern California Dungeness Crabs by Weight (Kg) at the Beginning of the Fishing Season, Cape Mendocino to Oregon Border.* Lower 95% Upper 95% Total Exploitation Intercept Slope Population confidence confidence catch rate Season (a) (b) (mean no.) limit limit Mg (u) 1966-67 16.34318 0.00000102 7,274,318 3,257,893 41,044,989 4.597 63.2 1967-68 20.85992 0.00000140 6,764,575 4,624,158 10,504,761 4.859 71.8 1968-69 25.29165 0.00000185 6,206,707 3,625,982 11,863,158 5.321 85.7 1969-70 22.40918 0.00000134 7,592,363 6)684,943 8,962,693 6.118 80.6 1970-71 17.84373 0.00000217 3,733,205 1,674,335 7,153,291 3.258 87.3 1971-72 12.29002 0.00000505 1,104,886 793,287 3,120,111 0.804 72.8 * Fort Bragg area landings not included. Based on these population estimates, fishermen were harvesting between 63 and 87% of the available crabs during the seasons considered (Table 5). Jow 196 CALIFORNIA FISH AND CAME ( 1 965 ) estimated that fishermen harvested 85% of the available crabs during the 1962-63 season. He also speculated from tag recovery data that fishermen harvested a higher percentage of the population during years of low abundance than during years of high abundance. From my data it appears that fishermen are harvesting a higher proportion of the legal males during years of high abun- dance, and a year or so after the population begins to decline, than during years of low abundance and years when the population is increasing. Fishing effort (numbers of boats) follows a similar pattern, thus supporting catch-per-trap data, assuming that a relationship exists between fishing effort and the percent- age of crabs harvested (Table 6). TABLE 6. Number of Boats Per Season Landing Dungeness Crabs at Ports Between Cape Mendocino and Oregon Border. Mean number of months Season Boats boat made landings 1966-67 198 1967-68 213 1968-69 244 3.3 1969-70 264 3.4 1970-71 257 2.7 1971-72 189 2.8 Mortality Rates In the calculation of weekly instantaneous mortality rates during the fishing season, the following equation was used. (Calculations by Eugene Witeck). In Y = a + b X Where: In Y = natural log of weekly mean catch in pounds per trap (Table 7). a — intercept of line on Y axis. — b= monthly instantaneous total mortality rate (Z) X = week number of fishing season. TABLE 7. Weekly Mean Catch (Lb and Kg) Per Trap of Northern California Dungeness Crabs Used to Calculate Weekly Instantaneous Mortality Rates. 1966-67 1967-68 1968-69 1969-70 1970-71 1971-72 Week number Lb/ Kg/ Lb/ Kg/ Lb/ Kg/ Lb/ Kg/ Lb/ Kg/ Lb/ Kg/ of season trap trap trap trap trap trap trap trap trap trap trap trap 1 19.6 8.9 19.0 8.6 16.8 7.6 22.4 10.2 20.3 9.2 15.9 7.2 2 11.6 5.3 20.9 9.5 19.2 8.7 20.3 9.2 14.9 6.8 3.8 1.7 3 16.6 7.5 18.8 8.5 - 15.2 6.9 14.1 6.4 6.4 2.9 4 12.8 5.8 - - 25.7 11.7 17.9 8.1 8.0 3.6 5 9.4 4.3 17.4 7.9 15.8 7.2 12.1 5.5 11.6 5.3 5.4 2.4 6 14.9 6.8 14.9 6.8 10.9 4.9 10.2 4.6 7.0 3.2 7 11.5 5.2 19.0 8.6 8.7 3.9 9.6 4.4 - - 8 - 11.2 5.1 16.4 7.4 7.6 3.4 - - 9 9.3 4.2 7.5 3.4 8.8 4.0 10.0 4.5 - - 2.6 1.2 10 13.1 5.9 7.7 3.5 7.5 3.4 10.9 4.9 - - 0.9 0.4 11 18.4 8.4 10.3 4.7 10.3 4.7 5.7 2.6 12 17.9 8.1 8.3 3.8 14.1 6.4 4.8 2.2 10.0 4.5 13 11.4 5.2 7.3 3.3 5.3 2.4 10.3 4.7 14 12.7 5.8 8.7 3.9 - - 3.3 1.5 15 5.3 2.4 3.2 1.4 2.3 1.0 16 6.2 2.8 3.1 1.4 6.0 2.7 17 6.2 2.8 10.0 4.5 DUNCENESS CRAB CPUE 197 18 2.0 0.9 10.8 4.9 19 - - 4.5 2.0 20 - 21 5.1 2.3 22 5.5 2.5 23 7.0 3.2 10.0 4.5 25 4.8 2.2 26 8.6 3.9 27 4.2 1.9 Instantaneous rates of fishing mortality (F) and natural mortality (M) were calculated from the following (Notation from Ricker 1975). u= L/N0 Where: u = annual expectation of death from fishery (rate of exploitation from Table 5). L = season landings in pounds (Table 4). N0 = population in pounds at start of season (Table 5). and: A = 1 - e -z Z/A = F/u Thus: F = uZ/A Since: Z = F + M Then: M = Z - F Where: Z = instantaneous total mortality rate. A = total annual expectation of death. F = instantaneous fishing mortality rate. M = instantaneous natural mortality rate. These calculations were based on the assumption that the calculated population estimate at the beginning of each season was truly representative. The data yielded significant correlation coefficient estimates for only three seasons, 1967- 68, 1968-69, and 1969-70, at the 5% level or less. The weekly instantaneous total mortality rates (Z) varied considerably from season to season (Table 8). The highest rate occurred during the 1971-72 season (-0.18300) and the lowest rate during the 1966-67 season (-0.00490). These rates correspond to seasonal survival rates (S) of less than 0.01% and 86%, respectively. TABLE 8. Mortality Rates and 95% Confidence Intervals for Northern California Dunge- ness Crabs * (Cape Mendocino to Oregon) During Fishing Season. — /' Lower Upper Fishing (weekly confidence confidence season Seasonal rates ** Season rate) limit limit (weeks) -Zt A u Ft Mt 1966-67 -0.00490 -0.00943 +0.00037 30 -0.15 0.14 0.63 0.68 1967-68 -0.07710 -0.07794 -0.07626 32 -2.47 0.92 0.72 1.93 0.54 1968-69 -0.04900 -0.04990 -0.04810 32 -1.57 0.79 0.86 1.71 1969-70 -0.11550 -0.11745 -0.11355 32 -3.70 0.98 0.81 2.94 0.76 1970-71 -0.05740 -0.06663 -0.04817 26 -1.49 0.77 0.87 1.68 1971-72 -0.18300 -0.20602 -0.15998 37 -6.77 0.99 0.73 4.99 1.78 * 159 mm carapace width and larger. ** Notation from Ricker 1975. The amount of interview data (40 interviews) used to calculate rates for the 198 CALIFORNIA FISH AND GAME 1971-72 season was small when compared to other seasons (i.e. 149 during 1970-71 ); this, plus the fact that most of the interviews were conducted in the Eureka area thus biasing the data, might explain in part the high mortality rate for that season (Table 8). The calculated seasonal instantaneous fishing mortality rates (F) ranged from 0.68 during the 1966-67 season to 4.99 during the 1971-72 season. The highest calculated instantaneous natural mortality rate (1.78) occurred during the 1971-72 season; the lowest rate (0.54) occurred during the 1967-68 season. If these rates are representative, we might expect the range of survival during a 6-month closed season to be 17 to 58%, disregarding seasons where natural mortality rates could not be calculated because of the low calculated total mortality rate and corresponding high fishing mortality rate. Jow (1965) calculated a bi-monthly survival of 27% from tagged crabs released in Pelican Bay. The preceding calculated population estimate and mortality rates yield valua- ble insights into the condition of crab stocks off northern California and therefore should be helpful in estimating change in yield due to changes in legal sizes. However, the estimates are no better than the data used to calculate them. Unfortunately, the catch-per-trap interviews contain two basic weaknesses: they were neither consistent nor random. These two factors probably account for most of the variation and inconsistencies. A third factor involved is the increase in mean pounds per trap that occurred during most seasons in March, April, or May. At present, it is unclear whether this increase is due to the small number of interviews, a sudden influx of crabs from some other area, or a new molt. There is some evidence that some sublegal crabs undergo ecdysis during the later winter as shown by an increase in the number of soft crabs during this period. The Leslie population estimate relies on three assumptions: 1 ) the population vulnerability must not change during the experiment, 2) the entire population must be available for capture, 3) there should not be an excess of recruitment and immigration, over emi- gration and natural mortality. The crab population estimates were for legal-sized males only, thus, the first assumption was met. The second assumption presents little problem during the first 2 or 3 months of the season; but the increase in the catch per trap in and during early spring might indicate that either a new segment of the population has moved into the area, or a molt of sublegal males has increased the number of legal males. From our tagging studies we speculate that there is very little movement of crabs into or out of the area. We do not have sufficient natural mortality data to determine whether it is excessive or not. The natural mortality rates in Table 8 are based, in part, on the catch-per-unit-of-effort data. Therefore, they are subject to the same weaknesses as the population estimates. One method of obtaining truly representative catch-per-trap data would be a mandatory log book system. Such a system would provide data for the entire fishery throughout the season. At present, it is costly, primarily in terms of man-days, to obtain sufficient interviews to be meaningful during the latter part of the season when only a few boats are operating out of each port. However, DUNCENESS CRAB CPUE 199 a log book system might prove more expensive than the interview system because of the large number of boats in the fishery and the number of man-days that would be required to process the logs. Utilization of port samplers hired only for the season to conduct interviews could be less expensive and they might obtain more accurate data. A log book might prove to be a valuable aid to the fishermen, particularly new, inexperienced fishermen, in following trends in crab behavior from year to year. ACKNOWLEDGMENTS This work could not have been completed without the assistance of many Department employees. Melvin Willis, Steven Taylor, Robert Hardy, Nancy Nelson, John Spann, and Paul Dinnel spent many hours on the docks interview- ing fishermen. Eugene Witeck calculated population estimates and mortality rates. Timothy Farley reviewed the calculations and offered many suggestions for the data analysis. Therese Hoban and Jane Dykzeul designed the figures. Ronald Warner and Richard Heimann provided the data included in Table 3. REFERENCES Cotshall, Daniel W., and Robert Hardy. 1969. Final report of port sampling. January 1966-November 1968. Pac. Mar. Fish Comm., 21st Ann. Rept.: 28-35. Jow, Tom. 1965. California-Oregon cooperative crab tagging study. Pac. Mar. Fish. Comm., 17th Ann. Rept.: 51-52. Ricker, W.E. 1975. Computation and interpretation of biological statistics of fish populations. Can., Fish. Res. Bd., Bull., (191):1-382. 200 CALIFORNIA FISH AND CAME Calif. Fish and Came 64 ( 3 ) : 200-209 1 978 SEX RATIOS OF THE NORTHERN ANCHOVY, ENGRAULIS MORDAX, OFF SOUTHERN CALIFORNIA1 RICHARD A. KLINGBEIL Marine Resources Region California Department of Fish and Game Long Beach, California 90802 Sex ratios for the northern anchovy, Engraulis mordax, were calculated from sam- ples taken from the southern California commercial reduction fishery and California Department of Fish and Game Sea Survey cruises for the period 1966-1975. Sex ratios calculated on a seasonal, monthly, and daily basis for the commercial fishery seldom approached the expected 1:1 ratio. Only a few ratios slightly favored males, whereas females were frequently dominant. Although sex ratios for sea survey cruises were most always close to the expected 1:1 ratio, there were a large number of samples that showed extreme dominance of either females or males for those cruises occurring during the period February to June. Sex ratios calculated by area of capture for this same time period suggested a spatial segregation of the sexes. INTRODUCTION Sex ratio data for the northern anchovy has been recorded routinely since the California Department of Fish and Game began continuous sampling of the southern California commercial reduction fishery in 1966. The female to male sex ratio calculated from samples taken during the 1968-69 season was 1.42:1. For the next six seasons the ratio varied between 1.14:1 and 2.02:1 (Table 1 ). TABLE 1. Calculated Sex Ratios (F:M) of Anchovies Taken at San Pedro, 1968-69 Through 1974-75 Reduction Seasons.* Calculated from Actual Estimated numbers Season numbers sampled sampled * 1968-69 1.45 : 1 1.42 : 1 1969-70 1.14 : 1 1.14:1 1970-71 1 .60 : 1 1 .60 : 1 1971-72 1.52 : 1 1.52 : 1 1972-73 1 .99 : 1 1 .98 : 1 1973-74 2.02 : 1 2.02 : 1 1974-75 1.57:1 * Using correction factors. Whether or not these data reflect the actual sex composition of the anchovy population off southern California is of importance to fisheries management. Anchovy biomass estimates, calculated by the National Marine Fisheries Service from California Cooperative Oceanic Fisheries Investigations (CalCOFI ) egg and larval surveys, assume a 1 : 1 sex ratio (Smith 1972). If, however, the sex ratio is not 1:1, then biomass estimates should be adjusted. The mechanisms of meiosis and fertilization tend to dispel the concept of a pelagic fish population consistently exhibiting a sex ratio other than 1:1. Accept- ing this premise raises the question of why anchovy fishery sex ratios vary so much and seldom approach the 1 : 1 ratio expected. ' Accepted for publication March 1978. NORTHERN ANCHOVY SEX RATIO 201 METHODS Data for this paper were taken from two principal sources, the southern California reduction fishery and California Department of Fish and Game Sea Survey Project cruises. Mais (19746) summarized the scope of these survey cruises. Estimated sex ratios of the commercial catch from 1968-69 to 1972-73 seasons were published in age and length composition reports (Collins 1971; Spratt 19726, 1973a, 19736; Sunada 1975). Computer compilations and individ- ual sample sheets were obtained for these seasons and for the 1973-74 and 1974-75 seasons from Department of Fish and Game files. Sex ratio data from Sea Survey cruises were taken from individual sample sheets for the years 1966 to 1975. Station data for these cruises were available from CalCOFI Data Reports 16-24 (Mais 1969a, 19696, 1971a, 19716, 1971c, 1972, 1973, 1974a, 1975). Sex ratios calculated for an entire commercial season were based on an estimate of total numbers of females and males landed. These ratios involved a correction for any differences in average weight of the sexes. In most cases, the correction factors were small and the sex ratios obtained from the actual numbers of females and males sampled were in close agreement with corrected ratios (Table 1 ). Sex ratios reported by month or area of capture for the com- mercial season and all sex ratios reported from sea survey data were calculated directly from numbers sampled. All sex ratios reported here represent the female to male (F : M) ratio. In some cases, samples contained a varying number of fish whose sex were reported as unknown, usually because of immaturity of the sex organs. These fish were ignored in calculating the sex ratio. The numbers of fish used to calculate the sex ratios represented an extremely small portion of the total population and the standard deviations were high when using these figures to estimate the sex ratio of the entire central anchovy popula- tion (Spratt 1972a; Vrooman and Paloma 1975). In general, sex ratios were compared as they related to both time and space. Date and location of capture were available on most sample sheets. For Sea Survey data, actual coordinates of stations were available. For the commercial fishery, location of capture was reported by the captain of the fishing vessel on logs collected by the Department of Fish and Game. RESULTS Sex ratios calculated from sampling the fishery daily varied considerably. Monthly sex ratios (Figure 1 ) calculated for 1968-69 through 1974-75 seasons twice were as high as 3 : 1 and frequently were higher than 2:1. No month showed a preponderance of males in the samples and only three times did the monthly sex ratio favor males. The combined numbers for all seasons resulted in a sex ratio of 1.60 : 1. In contrast, sex ratios from Sea Survey cruises (Figure 1 ) conducted during the same period varied to a lesser degree and combined numbers for all cruises resulted in a sex ratio of 1.09 : 1. Consistent seasonal or cyclic trends were not apparent from either set of data. The monthly sex ratios for the fishery also were collated by Fish and Game block number (Pinkas 1951 ) or area of capture to determine if the location of capture by the fishing fleet reflected any geographical differences ( Figures 2 and 3). The San Pedro Channel was the most frequent and consistently fished area. Landings often occurred from areas northwest, east, and southeast of the chan- 202 CALIFORNIA FISH AND CAME 3:1' \ 1968-69 1969-70 2:1 i:i< JS. s \ _® _^f ® ONDJ FMAMJ J X ASONDJFMAM UJ —J < • ■ UJ < 3:1 • 2:1 1970-71 • J8L 1971-72 ® *x— •«^#^.x' LU 1 S ONDJ FMAMJ J ® ASONDJ FMAM o < X UJ CO 3:1 • 2:1 1972-73 v v. • ® • ® • *\ 1973-74 I. V s ONDJ FMAMJ J ASONDJ FMAM 3:1 - 1974-75 2:1 ■ C ammercial fishery- x < 20 samples • > 20 samples i:i< x •"""" v ••""* Sea survey cruises-® — A I*. - s ONDJ F M A M FIGURE 1. Sex ratios (F:M) of anchovies by month computed from samples taken from the commercial fishery (1968-69 through 1974-75 seasons) and from California Depart- ment of Fish and Came Sea Survey cruises. nel. In a majority of months there was one area where the sex ratio was high with adjacent areas exhibiting lower values. Notable landings were made from the Santa Barbara Channel only occasionally, but the sex ratio was often high in samples from these landings (6.40: 1, 4.23: 1, 1.92: 1, 1.28: 1, 3.03: 1). In April 1973, landings arrived from a fairly large area ranging from just northwest of Palos Verdes Peninsula to east of San Clemente Island. The sex ratios for this month exhibited a declining trend from north to south. Data from Sea Survey cruises were better suited than commercial fishery data for studying the possibility of spatial differences in the sex ratio. Samples, ob- tained by mid-water trawl at night ( Mais 1 974b) , were taken over a much larger area than those from the commercial fishery. The reported location of capture for these samples was also more reliable. The duration of cruises in the Southern California Bight normally was less than 1 month, reducing the problem of move- ment of portions of the population. The sample size was standardized by number (25 fish ) and, as a rule, all fish were sexed ( samples from the commercial fishery seldom contained more than 17 fish) (Figure 4). The principal disadvantage of these data was the relatively small number of samples from each cruise, which necessitated grouping the data. Since 1966, most cruises took place during five NORTHERN ANCHOVY SEX RATIO 1970-71 SEASON ^^^^^^^^1971-72 SEASON flfl^ \ "^^~ *™n Pedro *o«° ^^^> ^ ^ ^c°honnel f* SEP -1.56 203 Bn.9/ ■■Xzo/ PF 1.4 v-5< C^P ^^Uft J^oX 1 ^ I ^ ^+> « ^^ OCT - 1.84 -x» 1 ** ^^ OCT -1.89 ^ V 1 V ^ * ^^ NOV - 1.48 -s» 1,^ NOV •^ wn*\ IK^A ^^tv VjKp s\ v r ^ \ ^ ■* ^^^ DEC -1.46 ■^ 1 m ^^ DEC X^ wx V ^ \ J ^ *> ^^^ JAN - 1.52 ^, 1 * ^^ JAN - 1.39 ^ V \ -*> ^^v-i-- i.iP^.A • ^^ FEB -^ 1 * ^^ FEB -1.20 «%» hT ^^^^^ n Pedro ^y,onnel -* -» ^^* MAR >^ i g ^* MAR ^ p^ I» \s ^c^P^^ «• ^^ APR «V» 1 * ^^ APR -1.74 ^ m<^> V —* [ V ^ m ^^ MAY - 2.39 -w I ^ ^^ MAY •»*» FIGURE 2. Sex ratios (F:M) of anchovies computed by month and block number (area of cap- ture) from samples obtained from the San Pedro reduction fishery in the 1970-71 and 1971-72 seasons. 204 CALIFORNIA FISH AND CAME 1972-73 SEASON 1973-74 SEASON Son *«*'<> £honnel San Clemente SEP- - 3.03 ^^ 1 18. OCT- 2.65 NOV- 1.75 NOV -2.44 1 * ^^ DEC -2.27 "N* 1 • ^^ DEC -2.15 ») (4) Potential population growth (ABpot) is the increase in population size which would occur if there were no fishery, and is equal to B, — B0 in equation (1 ). However, B0 is also given by equation (3), so ABpot is calculated by (1 ) minus (3), and recruitment terms cancel: ABpo, = B0e" ,M-G,t + R - (B.e- * + R) = B0 (e ~ (M-G" -e -if + M-Gj.) (5) The above equation allows us to compare potential population growth undei 22 6 CALIFORNIA FISH AND CAME conditions of no fishing (5) with equilibrium yield (4). This comparison is best expressed as a ratio (8) which is a function of F, M — G and t: 5 = vB^/Y. =[B0(e-,M-G"- e ~ (F. + M-G" )]/[B„(F./ (F. +M -C) ) = r(Fe + m-C) e-,M-G,,(1 -e -'•')] /[F. (1 - e^'**-01')] (6) Example values of 8 are given in Table 1. The quantity 8 may be thought of as a biological discount factor. When equilibrium yield is not harvested, the population will grow, but by an amount less than the yield which is foregone. The values in Table 1 indicate that the quantity ( M-C ) t is much more influen- tial than Ft in determining the value of 8. Therefore, a single value of 8 may be sufficient for most modeling purposes, and 8 can be treated as a constant for all likely values of F. Since the above calculation is for continuous fishing, seasonal fisheries will give different values of 8. For fisheries occurring just before repro- duction, 8 will be near 1 , and for fisheries occurring just after reproduction 8 will be smaller than the values in Table 1. Table 1. Example Values of 8. Ft (M - G)t _0J 1.0 0.0 100 1.00 0.1 0.95 0.94 0.2 0.90 0.89 0.3 0.86 0.84 0.4 0.81 0.79 0.6 0.73 0.70 0.8 0.65 0.61 1.0 0.58 0.54 1.5 0.43 0.38 Consideration of 8 is particularly important in catch-transition methods of estimating the parameters of production models such as in the generalized production model of Pella and Tomlinson (1969). Continuous reproduction is a basic assumption in production models, but they are very often empirically applied to fisheries wherein reproduction is distinctly seasonal. Under continu- ous reproduction, the relationship AB = Ye - Y (7) where Y is actual harvest, is appropriate. However, when reproduction is discon- tinuous, the relationship AB = 8 (Y. - Y) (8) should be used. Equation (7) is therefore a special case of equation (8), wherein t = 0 and 8 = 1. Since production modeling is often applied when biological information is minimal, a reasonable guess for the quantities (M — G) and t should be sufficient, and will certainly be an improvement over use of equation ( 7 ) . Estimates of the catchability coefficient and the asymptotic maximum popu- lation size which are obtained from the catch-transition approach are often considered to be unreliable (Pella and Tomlinson 1969) and this modification should improve accuracy. Again, the discount factor may be an important consideration in rehabilitating depleted resources. This factor may be a contributory cause of the apparently slow recovery of the Peruvian anchoveta fishery, where 8 = 0.6 is a likely value. NOTES 2 27 Conversely, when the potential rate of growth of a population is known, and equilibrium yield is to be estimated, maximum sustainable yield will be greater than maximum population growth rate by 8"'. Such is the case in the northern anchovy fishery in California, where S-' = 1.54. In the case of marine mammals, the appropriate time interval (t) would not be the gestation period, since mortality would result in the loss of offspring as well as the parent. Rather, t would be the interval between the time when the offspring would survive the death of the mother and the time when the mother would next become pregnant. Since this is a relatively short time, and M — G is relatively small, the discount rate for marine mammals may approach 1.0. REFERENCE Pella, J. )., and P. K. Tomlinson. 1969. A generalized stock production model. Inter-Am. Trop. Tuna Comm., Bull., 13: 419^496. — Alec D. MacCall, California Department of Fish and Game, Operations Re- search Branch, c/o SWFC, Box 271, Lajolla, CA 92038. Accepted for publica- tion March 1978. OBSERVATIONS OF AGONISTIC BEHAVIOR IN THE TREE- FISH, SEBASTES SERRICEPS (SCORPAENIDAE) On March 9, 1976, while scuba diving near the west end of Santa Catalina Island, Los Angeles County, California, I observed a display of agonistic behavior between two treefish, Sebastes serriceps (Jordan and Gilbert). I came upon the treefish at a depth of about 7 m (23 ft) in front of a rocky cave approximately 0.5 m (20 inches) high by 1 m (39 inches) wide, and several meters deep. A 25-cm (10-inch) fish, subsequently identified as the defender of the cave, had the head of a 20-cm (8-inch) treefish to about mid-orbit in its mouth. The two fish slowly moved back and forth in front of the cave for several minutes while interlocked in this manner. After the two separated, the larger treefish went into the cave and the intruder swam several meters away. When the intruder returned to the cave entrance, he was approached by the larger fish with erect dorsal fin and a widely gaping mouth. The intruder responded by erecting its dorsal fin and opening its mouth. Both fish, facing one another, then made a series of short darts in an attempt to grasp the other's jaw. When one fish did grasp the other's jaw, the attacked fish would bite down on the attacker's snout. In this manner, the two fish again moved back and forth. During one dart, the attacker missed the lower jaw and its head went com- pletely into the other fish's mouth. On another dart, one fish bit only one side of the lower jaw and the two hovered at right angles for a short period. The defender did not initiate agonistic behavior until the intruder approached the cave entrance. At no time during any of the described encounters was contact sufficient to cause visible damage to either fish. This grasping behavior, which might inhibit respiratory efficiency, terminated when the fish with the other's head in its mouth released its grip. This release might be stimulated by reduced blood oxygen levels. On February 20, 1970, a similar observation was made at the Los Angeles County Fish and Game Commission's Redondo Canyon Artificial Reef by John Duffy and Robert Hardy (Calif. Dept. Fish and Game, pers. commun.). They 228 CALIFORNIA FISH AND CAME observed and photographed what they presumed to be one sequence in the mating behavior of the treefish. Two animals oriented head-to-head, grasped each others lips, held on for several minutes at a time, and occasionally shook each other violently. Unfortunately, the photographs they took are not of suffi- cient quality to be published. Little information has been published on aspects of treefish life history and behavior. Phillips (1957) and Miller and Lea (1972) gave systematic accounts of the treefish and described the adult color as yellow to olive with five or six black bars along the body. Feder, Turner, and Limbaugh (1974) reported that treefish are aggressive, territorial, and take shelter in rocky crevices. The behavior exhibited by the treefish on both occasions was not observed to completion, nor was either pair of fish collected; thus it cannot be determined if the behavior was territorial or reproductive. In contrast to their normal adult coloration, the treefish I observed were dark green overall; the black bars were indistinct. However, photographs of the treefish observed at Redondo Canyon show the usual adult coloration. Leon Hallacher ( U.C. Berkeley, pers. commun. ) has observed displays similar to those of the treefish in the gopher rockfish, Sebastes carnatus, including the facing and grasping behavior. He considers such behavior rare, having observed it only once in 7 years of diving. Kim McCleneghan and James Houk (Calif. Dept. Fish and Game, pers. commun.) observed similar behavior in the kelp rockfish, S. atrovirens. Comparable displays, manifested by two individuals facing each other with mouths agape and dorsal and pectoral fins erect, have been observed in several Californian fishes, including sheephead, Pimelometopon pulchrum (T. J. Muell- er, Univ. So. Calif, pers. commun.) and the blackeye goby, Coryphopterus nicholsii (Wiley 1973). I have seen such displays in the blue banded goby, Lythrypnus da///. Physical contact accompanying such activities has been only a nip, which caused no apparent damage, and not prolonged grasping bouts as in these treefish. ACKNOWLEDGMENTS I wish to express my gratitude to John Duffy and Robert Hardy for allowing me the use of their data and photographs. I also thank John Fitch and Leon Hallacher for reading the manuscript and their valuable suggestions. These observations were made during work performed under Dingell-Johnson project F-27-D, Sportfish-Kelp Habitat Study. REFERENCES Feder, Howard M, Charles H. Turner, and Conrad Limbaugh. 1974. Observations on fishes associated with kelp beds in southern California. Calif. Dept. Fish and Came, Fish Bull., (160): 1-144. Miller, Daniel )., and Robert N. Lea. 1972. Guide to Coastal Marine Fishes of California. Calif. Dept. Fish and Came, Fish Bull., (157): 1-235. Phillips, Julius B. 1957. A review of the rockfishes of California (family Scorpaenidae) . Calif. Dept. Fish and Came, Fish Bull., (104): 1-158. Wiley, James W. 1973. Life history of the western North American goby, Coryphopterus nicholsii (Bean). San Diego Soc. Nat. Hist., Trans., 17(14): 187-208. — Peter L. Haaker, Calif. Dept. Fish and Game, Marine Resources Region, 350 Golden Shore, Long Beach, California 90802. Accepted for publication March 1978. Reviews 229 BOOK REVIEWS Bright Rivers by Nick Lyons; J. B. Lippincott Co., Philadelphia and New York, 1977; 166p. $8.95. Bright Rivers is an autobiography. It is a reflection on, and recitation of the successes and failures of Nick Lyons as a fisherman. He uses the qualities developed in his profession of English professor to enliven this work in a "literary" style. None of us could write so lucidly about fish lost and found. However, Lyons has a tendency to dwell a little too long and often on his fishing skills, or lack of skills. I think Nick Lyons is a better fisherman than he lets on. Close personal contact with the better fishermen of the United States as he has, must sooner or later rub off. One consolation is that our exposure to his troubles reduces the sting of our own. The book is divided into two parts. The first deals primarily with eastern fishing, focusing especially on a summer vacation spent in the Catskills. The second deals with fishing trips to the west, especially to Montana. He really gets turned on by Montana — it must be heaven to him. It is in this section that his writing seems to be more alive. The western reader may identify better with this section. He may also be thankful there are more waters to fish and larger fish to be found in this part of the country. Apparently 1 0-inch fish are a prize in many heavily fished waters where Lyons usually fishes. I hope the author fully develops his skills as a fisherman. When he does, his literary training, combined with angling skill, should coalesce to really produce fine quality angling works. — Ed Littrell The Fresh and Salt Water Fishes of the World by Edward C. Migdalski and George S. Fichter; Alfred A. Knopf, Inc. N.Y., 1976. 316 pp., illustrated in color. $25.00. This beautifully illustrated book represents another attempt to present to the general reader an up-to-date review of the world's fish families. The very readable text leads off with a short discussion of classification, fossil fishes, fish morphology, anatomy, physiology, age and growth, and migrations. The discussion of the families highlights size, description, and life histories of some of the more well known as well as rare species. Full color illustrations of over 500 species were produced by Norman Weaver and are truly spectacular. Unfortunately, I found the text to contain many errors or omissions. For example, the record size of the white shark, Carcharodon carcharias, is reported to be 36.5 ft. However, a recent paper indicates that this record is invalid; the statement that most chinook salmon, Oncorhynchus tshawyt- scha, are caught as they move upstream to spawn, completely ignores the large ocean catches by sport and commercial fishermen. The statement that the pink salmon, O. gorbuscha, "occurs in greater abundances further south than do other species", I assume is a typographical error. Appar- ently the authors did not refer to Miller and Lea's Fish Bulletin 157 as there are several mistakes in geographical ranges, size records, and names of California fish. On the positive side, some of the species descriptions I found to be very informative as well as entertaining, in particular, the narrative on the carp, Cyprinus carpio. Despite the faults in the text, I feel that the excellent color illustrations will make this book very attractive to professionals and non-professionals who never tire of looking at fish. — Daniel W. Gotshall Inland Fishes of California by Peter B. Moyle; Univ. of Calif. Press, Berkeley, Los Angeles, London, 1976; 405 pp; illustrated $20.00. Inland Fishes of California is a much needed and valuable reference book for the student, biologist, amateur naturalist, or interested angler. Unfortunately, there are errors throughout the text, which undoubtedly will be corrected in subsequent editions. Meanwhile, Dr. Moyle has indicated that any purchaser may get a three-page list of errors and additions to Inland Fishes of California by writing to: Dr. Peter B. Moyle Department of Wildlife and Fisheries Biology University of California, Davis 95616 A significant difference exists in the quality of the drawings and there is a general lack of illustrations of adult fish. In fact, some of the illustrations resemble juvenile specimens recently removed from bottles of formalin or alcohol. When asked why his illustrations were predominately of juvenile specimens, the author replied that it was partly a matter of availability and partly because of his intention to illustrate sizes which might be encountered most often by students using the usual collecting techniques (traps and seines). I did not test the keys, however they seemed quit* simple and probably would be easy to use. 230 CALIFORNIA FISH AND CAME Although I share the author's feelings about the loss or deterioration of populations of native (particularly non-game) species, his preoccupation with this sentiment reduces the objectivity of his presentation. He tends to portray an aura of helplessness and resignation. When discussing the Sacramento-San )oaquin Delta in his chapter entitled "Ecology", he says "Today it hardly seems worthwhile even to devote much time to the interactions among the introduced species since if any stable associations have been established they are likely to be soon upset as the waters of this zone continue to change, and as other new species become established in them." One could argue that in view of gross changes which are occurring, it should be desirable to predict the effect of such changes and to consider the desirability of modifying them. His strong identification with native species tends to make his "Ecology" chapter more of a description of what the system was, or might have been, rather than the ecology of what now exists. Introduced species seem to be regarded as "noise" to the theoretical system. In his discussion of the ecology of Clear Lake, he gives a somewhat detailed description of the ecological relationships of the native species which previously inhabited the lake. Although, as he points out, "at least 13 species are now established and only 4 of the native species are still maintaining large populations", he dismisses recent ecological relationships by saying "too little is known about their ecology to make reasonable speculations about their interactions", and that the species composition is still changing. No one can deny that much change has taken place recently, but on the other hand, many ecological studies also have been made recently in Clear Lake. One wonders if the ability to speculate is more constrained in the face of recent "hard", but conflicting data. Certainly, the dynamics of change, itself, is an important facet of ecology, which is glossed-over lightly. His descriptions of change consistently carry value judgments. In general, changes brought about by man's influence are labeled "bad", while changes which took place without man's influence are considered "good". Knowing the author, one is aware that this is precisely the message that he is intending to convey. While I appreciate his sentiments, empathize with him in these feelings, and recognize his right to express them, I feel that a more objective treatment of these subjects would have improved the text considerably. In summary, however, this is an outstanding book and should be included on the bookshelf of any ecologist, student, or angler interested in California's inland fishes or fisheries. — John Radovich Photoelectronic composition by CALIFORNIA OFFICE OF STATE PRINTINC 7734S— 800 3-78 4,500 LDA TI 0 0) n H > r m 2 a. > 0 3! ^T GO o H Z m > CS 33 Q 3> n I — x "0 •n n > o H 3 -i 3D C 2 2! 71 n z 5 z H -n -p 0 3= 0 H •*1 'Jl *» > o I 35 > a z o 0 c 0 > m > > n z PI 0 0O 7§ 1.1 c In 0° ^ _ r— ■* p 0 K RA POST, AID •O D J* -H *» a O m *o — m -*»