r*:,. II ^f mM- ITEO STATES ■ ."ITMENT OF ■ AMERCE ■ LIGATION ■ i(- "^ ^H <:.:..■ >*^ ^^^1 ■ Fishery Bulletin U.S. DEPARTMENT OF COMMERCE National Oceanic and Atmospheric Administration National Marine Fisheries Service J.i^-fn.^ Boicgical Uborcitory LIBRARY FEB 14 1372 Woods Hole, Mass. Vol. 70, No. 1 January 1972 MATSUMOTO, WALTER M., ELBERT H. AHLSTROM, S. JONES, WITOLD L. KLAWE, WILLIAM J. RICHARDS, and SHOJI UEYANAGI. On the clarification of larval tuna identification particularly in the genus Thunnus 1 PARSONS, T. R., K. STEPHENS, and M. TAKAHASHI. The fertilization of Great Central Lake. I. Effect of primary production 13 LeBRASSEUR, R. J., and O. D. KENNEDY. The fertilization of Great Central Lake. II. Zooplankton standing stock 25 BARRACLOUGH, W. E., and D. ROBINSON. The fertilization of Great Central Lake. III. Effect on juvenile sockeye salmon , 37 PERRIN, WILLIAM F., and JOHN R. HUNTER. Escape behavior of the Hawaiian spinner porpoise (Stenella of. S. longirostris) 49 EVANS, W. E., J. D. HALL, A. B. IRVINE, and J. S. LEATHERWOOD. Methods for tagging small cetaceans 61 DAVY, BRENT. A review of the lantemfish genus Taaningichthys (Family Myctophi- dae) with the description of a new species 67 DRUCKER, BENSON. Some life history characteristics of coho salmon of the Karluk River system, Kodiak Island, Alaska 79 PERKINS, HERBERT C. Developmental rates at variou-s temperatures of embryos of the northern lobster (Homarus americanus Milne-Edwards) 95 SICK, LOWELL V., JAMES W. ANDREWS, and DAVID B. WHITE. Preliminary studies of selected environmental and nutritional requirements for the culture of penaeid shrimp 101 KELLEY, CAROLYN E., and ANTHONY W. HARMON. Method of determining car- otenoid contents of Alaska pink shrimp and representative values for several shrimp products Ill LEWIS, ROBERT M., E. PETER H. WILKENS, and HERBERT R. GORDY. A de- scription of young Atlantic menhaden, Brevoortia tyrannus, in the White Oak River estuary. North Carolina Ill KEPSHIRE, BERNARD M., JR., and WILLIAM J. McNEIL. Growth of premigratory Chinook salmon in seawater 11£ MAJOR, RICHARD L., and GERALD J. PAULIK. Effect of encroachment of Wanapum Dam Reservoir on fish passage over Rock Island Dam, Columbia River 12S MOSHER, KENNETH H. Scale features of sockeye salmon from Asian and North American coastal regrions 14j HOUDE, EDWARD D. Development and early life history of the northern sennet, Sphy- raena borealis DeKay (Pisces: Sphyraenidae) reared in the laboratory 18{ CAHN, PHYLLIS H. Sensory factors in the side-to-side spacing and positional orienta- tion of the tuna, Euthynnus af finis, during schooling 19' ROSENTHAL, RICHARD J., and JAMES R. CHESS. A predator-prey relationship between the leather star, Dermasterias imbricata, and the purple urchin, Strongyloeeri' trotus purpuratus 20J JUDKINS, DAVID C, and ABRAHAM FLEMINGER. Comparison and foregut con- tents of Sergestes similis obtained from net collections and albacore stomachs 211 Note HOOPES, DAVID T., and JOHN F. KARINEN. Longevity and growth of tagged king crabs in the eastern Bering Sea . . 22! U.S. DEPARTMENT OF COMMERCE Maurice H. Stans, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION Robert M. White, Administrator NATIONAL MARINE FISHERIES SERVICE Philip M. Roedel, Director Fishery Bulletin The Fishery Bulletin carries original research reports and technical notes on investigations in fishery science, engineering, and economics. The Bulletin of the United States Fish Commission was begun in 1881 ; it became the Bulletin of the Bureau of Fisheries in 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Sep- arates were issued as documents through volume 46; the last document was No. 1103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate appeared as a numbered bulletin. A new system began in 1963 with volume 63 in which papers are bound together in a single issue of the bulletin instead of being issued individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued quarterly. In this form, it is available by subscription from the Superintendent of Documents, U.S. Gov- ernment Printing Office, Washington, D.C. 20402. It is also available free in limited numbers to libraries, research institutions, State and Federal agencies, and in exchange for other scientific publications. EDITOR Dr. Reuben Lasker Scientific Editor, Fishery Bulletin National Marine Fisheries Service Southwest Fisheries Center La JoUa, California 92037 Editorial Committee Dr. Elbert H. Ahlstrom National Marine Fisheries Service Dr. William H. Bayliflf Inter-American Tropical Tuna Commission Dr. Daniel M. Cohen National Marine Fisheries Service Dr. Howard M. Feder University of Alaska Mr. John E. Fitch California Department of Fish and Game Dr. Marvin D. Grosslein National Marine Fisheries Service Dr. J. Frank Hebard National Marine Fisheries Service Dr. John R. Hunter National Marine Fisheries Service Mr. John C. Marr Food and Agriculture Organization of the United Nations Dr. Arthur S. Merrill National Marine Fisheries Service Dr. Virgil J. Norton University of Rhode Island Mr. Alonzo T. Pruter National Marine Fisheries Service Dr. Theodore R. Rice National Marine Fisheries Service Dr. Brian J. Rothschild National Marine Fisheries Service Dr. Oscar E. Sette National Marine Fisheries Service Mr. Maurice E. Stansby National Marine Fisheries Service Dr. Maynard A. Steinberg National Marine Fisheries Service Dr. Roland L. Wigley National Marine Fisheries Service ON THE CLARIFICATION OF LARVAL TUNA IDENTIFICATION PARTICULARLY IN THE GENUS Thunnus ^ r.A v Walter M. Matsumoto,^ Elbert H. Ahlstrom,'' S. Jones,^ WiTOLD L. Klawe,* William J. Richards/ and Shoji Ueyanagi" ABSTRACT A Larval Tuna Identification Workshop was held at the Bureau of Commercial Fisheries Biological Laboratory (now the National Marine Fisheries Service, Southwest Fisheries Center), Honolulu, Hawaii, on March 2-6, 1970, to resolve conflicting views on the identification of larvae of Thunnus alalunga and T. albacares and to clarify the status of larval identification of other Thunnus species. The identification of T. alalunga (Yabe and Ueyanagi, 1962), T. albacares (Matsumoto, 1958), T. obesus (Matsumoto, 1962), and T. thynnus (Yabe, Ueyanagi, and Watanabe, 1966) was agreed upon as correct, except that the description of T. albacares should include the appearance of black pigmen- tation at the tip of the lower jaw when the larva attains a length of about 4.5 mm SL and that the lower size limit of reliable identification of T. alalunga be set at about 4.5 mm SL until further studies indicate more precisely whether the black pigmentation at the tip of the lower jaw in T. albacares appears earlier. There was no difference in appearance of T. thynnus larvae from the Atlantic and Indo-Pacific Oceans. The identification of T. tonggol and T. maccoyii larvae was not conclusive. The larvae of T. atlanticus required further study. The workshop further concurred that juveniles (13-200 mm SL) of several species of Thunnus may be separated by internal and external characters: T. atlanticus by vertebral count, T. alalunga by shape of first elongate haemal spine and arrangement of pterygiophores of the second dorsal fin rel- ative to two adjacent neural spines, and T. thynnus by configuration of lateral line and arrangement of pterygiophores of the second dorsal fin; and that juveniles of T. obesus and T. albacares may be sep- arated from the previous three species by arrangement of pterygiophores of the second dorsal fin, but not from each other. The proper identification of larval tunas has been a perplexing and difficult problem for many years. Although progress in the past two dec- ades has resulted in agreement on the identifica- tion of larvae of a number of species {Katsu- womis pelamis, Euthyymus alletteratus, E. af- finis, E. lineatus, Thimnus thynnus, T. obesus, and Auxis spp.) , there is still some disagreement ^ National Marine Fisheries Service, Southwest Fish- eries Center, Honolulu, HI 96812. ^ National Marine Fisheries Service, Southwest Fish- eries Center, La Jolla, CA 92037. ^ Department of Zoology, University College, Trivan- dum-1, India. (Formerly: Central Marine Fisheries Research Institute, Mandapam Camp, South India.) * Inter-American Tropical Tuna Commission, La Jolla, CA 92037. ^ National Marine Fisheries Service, Southeast Fish- eries Center, Miami, FL 33149. * Far Seas Fisheries Research Laboratory, Shimizu, Japan. and confusion on the identity and description of T. alalunga and T. albacares. At the present time there are two diflferent descriptions given for T. alalunga (Matsumoto, 1962; Yabe and Ueyanagi, 1962). The identity of other tunas, such as T. tonggol, T. maccoyii, and T. atlanticus, has yet to be confirmed or resolved. One of the chief problems in larval tuna iden- tification is the difficulty in obtaining good series of larvae for study. Tuna larvae are seldom taken in sufficient numbers by the usual collect- ing methods, and individuals over 10 mm stan- dard length (SL) are taken rather infrequently. Additionally, although the young of a number of tuna species are found together in many parts of the ocean, some species are localized in certain areas. Consequently, it is extremely difficult for workers in diflferent parts of the world to have Manuscript accepted September 1971. FISHERY BULLETIN; VOL. 70, NO. I, 1972. FISHERY BULLETIN: VOL. 70, NO. 1 at hand a complete series of larvae of more than two to four species. In an attempt to resolve the conflicting views on T. alalunga and to review the identification of larvae of other species of tunas, a Larval Tuna Identification Workshop was held at the Bureau of Commercial Fisheries (BCF) Biological Lab- oratory (now the National Marine Fisheries Service, Southwest Fisheries Center) , Honolulu, Hawaii, on March 2-6, 1970. The workshop also afforded an opportunity to workers specializing on larval tuna identification to assemble speci- mens of the various species of tunas and to ex- amine these together. The procedure followed at the workshop was (1) to summarize the status of larval tuna identification to date by species and (2) to evaluate the identifying char- acters by examining larval specimens. As time permitted, the status of juvenile tuna identifi- cation was also examined. The participants included: Mr. Walter M. Matsumoto, Convenor Dr. Elbert H. Ahlstrom, Advisor Dr. Santhappan Jones Mr. Witold L. Klawe Dr. William J. Richards Dr. Shoji Ueyanagi Dr. Jean-Yves Le Gall of the Centre Ocean- ologique de Bretagne, Brest, France, attended the workshop as an observer. The sessions were conducted informally with a summary of the present status of larval tuna identification, including recent developments, followed by evaluation of the various characters that could be relied upon for positive identifica- tion. Most of the sessions were devoted to direct examination of larval specimens of the various species and discussions of unpublished data of- fered by participants. This report summarizes the proceedings and results of the workshop. RECENT DEVELOPMENTS IN THE IDENTITY OF Thunnus alalunga Two diff'ering versions of the identity and de- scription of T. alalunga had arisen from reliance on black pigmentation in diff"erent parts of the body. Matsumoto (1962) relied upon black pig- mentation on the dorsal and ventral edges of the trunk forward of the caudal fin base, where- as Yabe and Ueyanagi (1962) relied upon black pigmentation on the tips of the upper and lower jaws and the absence of pigmentation on the trunk. The lack of sufficient larvae fitting Matsu- moto's description from areas presumed to be spawning grounds on the basis of gonad studies casts some doubt on his identification. On the other hand, good correspondence in the occur- rence of larvae fitting Yabe and Ueyanagi's de- scription with catches of adult T. alalunga in various areas in the Pacific seemed to support the latter identification. A study of red pigment patterns in larvae prior to preservation (Ueya- nagi, 1966) reinforced Yabe and Ueyanagi's identification and description. Additional ob- servations on red pigmentation by Matsumoto (see later discussion) confirmed Ueyanagi's re- sults and also provided more data to enhance the reliability of red pigmentation as a supple- mentary character for identifying T. alalunga. IDENTIFICATION OF TUNA LARVAE With the problem of differences in the identity and description of T. alalunga larvae fairly well settled at the outset, there remained the tasks of evaluating the various identifying characters, not only for this species but for other tunas as well, and of describing the species at various size categories. DEFINITION OF LARVA In tunas, as in many other fish, it is difficult to clearly separate the larval from the juvenile stages because there is no marked metamor- phosis and the usual adult characters used for species identification develop gradually and sep- arately. It is generally accepted among workers in larval tunas that the larval stage ends when the larva has developed the full complement of spines and rays in all the fins, all the vertebrae have ossified, and the anal opening has moved back near the origin of the anal fin. For nearly all tuna species, these developments occur when the larva has attained 10 to 13 mm SL. We use this as our definition, also. MATSUMOTO ET AL.: LARVAL TUNA IDENTIFICATION EVALUATION OF CHARACTERS In identifying fish larvae collected in plankton nets, the easiest and perhaps the only recourse is to identify the largest stage and work down to the smallest. Unfortunately, very few tuna larvae above 9 mm SL are taken in plankton .j\et tows so that this process cannot be followed W all times and identification, therefore, must depend upon those nonadult characters that are the most distinctive and consistent throughout the size range. Characters that have been used in the past were reviewed and evaluated. A resume of the usefulness of the various characters follows. Meristic The number of myomeres is useful in sepa- rating Katsutvomcs pelamis (42-43) and Euthi/n- nus lineatus (38-39) from other tunas, including other species of Euthynnus, all of which have similar numbers of myomeres (40-41). The number of fin rays and spines are not useful for separation of Thunmis because all species are similar in this respect. Morphological Shape of first dorsal fin, when completely formed, is useful to distinguish late larval stages of K. pelamis, Euthynnus, and Auxis from those of Thunnus. Preopercular spines are unreliable . because they undergo rapid growth changes and position of eye relative to longitudinal axis of body needs to be determined more accurately. Distribution (number and position) of pterygio- phores in the second dorsal fin in relation to neural spines is useful in separating several Thunnus species, but only after these bones have ossified in larvae longer than 10 mm SL. Other characters of the axial skeleton useful in identification, such as the position of the first haemal arch and the position of the zygapophyses on the vertebrae, also form late. Measurements Morphometries have not been used extensively to date, although there may be some with good possibilities, such- as the relations of body depth to standard length, snout length to head length, and snout length to orbit diameter. Some of the reasons for not using measurement data are that the larvae not only shrink in preservatives, but the degree of shrinkage varies in different preservatives and with duration of preservation; the distortion of the body at the time of fixing cannot be controlled; and, more important, there are too few larvae in undistorted condition for reliable measurements. Added to these are other sources of variability such as rapid changes in body parts due to growth, changes which often occur in spurts, and distension of the abdomen, as well as stretching of the body at each feeding. Pigmentation For the most part black pigment patterns have been the most widely used and accepted character in identifying tuna larvae. There are variations and changes in black pigment patterns on tuna larvae due to growth, but in certain areas of the body these patterns have been found to be consistent enough for identification pur- poses. This is particularly true of pigment pat- terns on the first dorsal fin, posterior half of the trunk, forebrain, and tips of both jaws. The larval size at which black pigment cells appear in certain areas of the body, especially at the upper and lower jaw tips, may be useful in sep- arating T. albacares from T. alahinga. Red pig- ment patterns, although not species specific, have been useful in confirming the identification of T. alalunga when used in conjunction with black pigment patterns. Of all the characters reviewed and examined, pigment patterns, both black and red, were con- sidered to be the most reliable for identification of the larval stages, despite their known varia- bility, when supplemented by the use of certain morphological characters such as the distribu- tion of pterygiophores in the second dorsal fin and characteristics of the vertebral column, whenever these are developed. VERIFICATION OF RED PIGMENTATION Ueyanagi (1966) reported on the usefulness of red pigmentation in identifying tuna larvae. Up to then identification of tuna larvae by pig- mentation had been based on black pigment only. 3 FISHERY BULLETIN: VOL. 70. NO. 1 Ueyanagi examined 350 larvae and concluded that T. albacares and T. alalunga, which are difficult to identify by the usual diagnostic char- acters, could be distinguished by differences in red pigment patterns: larvae of T. alalunga consistently had more red pigment spots on the dorsal and ventral edges of the body and along the mid-lateral line forward of the caudal pe- duncle than larvae of T. albacares; red pigment patterns in larvae of T. thynnus and T. obesus were intermediate between those of T. alalunga and T. albacares; the red pigment pattern in Allothunnus fallal was similar to that in Thun- nus; and the pattern in K. pelaniis resembled that in Auxis spp. and E. affinis but differed from that in Thunnus. To confirm these results and to provide addi- tional information on red pigmentation in tuna larvae, the results of observations made on 432 larvae taken in Hawaiian waters during August and September 1967 were presented. Tables 1 and 2 give the number of red pigment cells along the dorsal, ventral, and lateral lines on the pos- terior half of the trunk and a summary of the number of larvae examined, the number of larvae observed with red pigmentation, and the mean numbers of red pigment cells at the three sites. In Table 2, the number of red pigment cells ob- served most frequently are given in bold face type and those observed occasionally or seldom are enclosed in parentheses. The pigment patterns agreed generally with those reported by Ueyanagi for the species listed in the tables. Differences in the patterns were noticeable mainly in the dorsal edge of the trunk and, to a lesser extent, in the mid-lateral line. There was no significant difference in the num- ber of pigment cells t)etween the left and right sides of the body. The appearance and extent of red pigment cells varied in larvae taken in night and day tows. In larvae taken at night the pigment cells were numerous, distinct, and brightly colored, whereas in larvae taken during the day the pig- ment cells were faintly colored, often not visible, and in many instances the pigment spots were united, forming single continuous lines. Of the species taken in both day and night tows {T. al- bacares, T. obesus, and K. pelamis) , red pigmen- tation was not visible in 41.5% of the larvae taken during the day, compared with only 3.6% of the larvae taken at night. Thus, observations of red pigment cells must be made largely on larvae taken at night to reduce the effects of diel variations. Despite the variations, red pigmentation is a useful supplementary character to either sepa- rate certain species or verify the identification made on the basis of other characters. That the red pigment pattern is not species specific is clearly seen in the similarity among K. pelamis, Auxis spp., and E. affinis and between T. obesus and T. albacares; however, it is useful in sepa- rating T. alalunga from the other kinds of Thimnus. EXAMINATION AND DISCUSSION OF SPECIES Thunnus alalunga and T. albacares Larvae of these two species were examined together because they are the only species lack- ing black pigment cells on the trunk, exclusive of the caudal fin and abdomen (Yabe and Ueya- nagi, 1962), Characters, including some that have not been used in the past, for separating the two species are summarized in Table 3. The larval stage was divided into two size categories, small larvae less than 10 mm SL and larger lar- vae 10 to 13 mm SL, because the characters used in differentiating small larvae became ineffective or obscured with growth. As mentioned earlier, pigmentation, particularly the presence of black pigment cells at the tips of the upper and lower jaws and the amount of red pigment cells on the trunk, was the most reliable character in sepa- rating larvae of the two species. In small larvae, black pigment cells on the lower jaw tip were first observed in larvae of T. albacares about 4.5 to 6.0 mm SL, and often as small as 3.8 mm SL; those on the upper jaw tip were first observed in larvae about 7.0 mm SL (Figures 1 and 2, reproduced from Matsumoto, 1958') . In T. alalunga these pigment cells were ' The difference in developmental stages per given size in the figures by Matsumoto (1958) and Ueyanagi (1969) is due to method of preservation: Matsumoto's figures are of larvae preserved in 10% Formalin; Ueya- nagi's figures are of larvae preserved in 70% alcohol. 4 MATSUMOTO ET AL.: LARVAL TUNA IDENTIFICATION US ft W I -t-> m •i-H w s 2 CO c3 C 01 o (V 13 cr BQ H «, t^ S3S i> o "' tX 11 V 0)12 ^:5 5 i; 0) S3^ ,2 oS; 0)3 PI o n 1> 0)73 0>3 J;^ U C - o r>T3 O 0 t/i o »o to 1 I r-. r-^ -^ "1 CO o ^ »— o> lo CN CN O- CN •— •O CO ^ CM O ■* I lO -* — I O •— CN CO ■^ o "O so ' »o 'O o >o — ■» tx ^ — -O I I I i< d - d ' ' ' IT) — — I I I I CO I I I I r I tv -if — rx r ' ' K 00 so K ' CO ■>»• >o O; CO ■* -o — "S K d CO CM ■— CS CS — 00 CO CO CO LO ■— ' — -O — O K CO cs CO r-. <) ■^ "o vd — 00 SD CO CO d — CO cs — o — CN CO -^ "1 -o c 3 -6 is r rs. N. o hv o CO K r 1 1 N. -^ CO "^ rx N. 1 -^ 1 I 1 1 1 1 ' ' K lo d ^o is! i< "— CO •— > iri ' ' ' ' ' ' — CN — ' C(i ■>)■' K CN •- CO — — CM CN — I CM I I I I I I I I I I I I I I I ICM — COt^CSCO'O'* I I I I I I I I I I I I I •— I I I I I I I I I I I I lo I'O'Oixo-p'O'^rs.cssO'^'O'opio i ' lo d 'd->trv-10 mm SL): Array of ^Da pterygiophores between two adjacent neural spines Position of first haemal arch (vertebra number) 3-12 [mean = 7.0] 5-12 [mean 1.0] 1, 2, 1, 1 nth 2, 2, 3, 2, 2, 1 10th 1, 2, 2, 2, 3, 1 D2 refers to second dorsal fin. Other Thunnus species These species, which include T. thynnus {T. thynnus thynnus of Atlantic and T. thynnus or- ientalis of Pacific) , T. tonggol, T. maccoyii, and T. obesus, have been identified mainly by black pigmentation on the trunk other than that over the abdominal w^all. In small T. thynnus of both Atlantic and Pa- cific Oceans (larvae between 3 and 10 mm SL), one or two large black pigment cells are present on the dorsal edge of the trunk between the sec- ond dorsal and caudal fins (Table 4, Fig-ures 5 and 6) , the anterior one usually being the larger. There may also be one to four black pigment cells on the ventral edge of the trunk between the anus and the caudal fin. Black pigmentation in T. thynnus from both oceans agrees quite well, except that in 5 out of 10 Atlantic specimens one or two tiny black pigment cells were noted along the mid-lateral line of the body near the pectoral fin, and in two instances a single tiny black pigment cell was found on the mid-lateral line beneath the posterior end of the second dor- sal fin. These pigment cells were not considered reliable for identification purposes. Observation of red pigmentation on larvae of Atlantic and Pacific T. thynnus is incomplete. Only one Atlantic T. thynnus larvae was exam- ined for this character, but unfortunately the specimen was taken in a day tow so that the pigmentation appeared as a continuous streak on both the dorsal and ventral edges of the trunk as well as on the ventral surface of the lower jaw. In Pacific T. thynnus there were one to five red pigment cells, usually three, on the dorsal edge of the trunk. The number of red pigment cells on the mid-lateral line and ventral edge of the trunk has not been recorded, but according to the Illustration by Ueyanagi (1966), the pig- ment pattern may be similar to that of T. obesus. On the basis of black and red pigmentation, the Atlantic and Pacific T. thynnus were not sep- arable. The identification of T. tonggol, based on size series of 4.2 to 7.3 mm, has yet to be confirmed. Following the description of the species by Ma- tsumoto (1962), larvae similar to these having the anteriormost black pigment cell on the dorsal edge of the body ahead of the second dorsal fin origin have been found in 1963 in the mid-South Atlantic Ocean near Ascension Island. Confir- mation of the species description requires the finding of adults within this area and the finding of additional larvae to extend the identified size range. The identification of T. maccoyii, which was first described as having black pigment pattern similar to that of T. thynnus (Yabe, Ueyanagi, and Watanabe, 1966) and later as having the black pigment cells on the dorsal edge of the trunk reduced to pinpoints (Ueyanagi, 1969), also needs verification (see discussion on T. thyn- nus) . The correspondence of published descrip- tions based on eight specimens and observations of larvae identified as this species were not con- clusive. T. obesus was easily separated from T. thyn- nus by the absence of black pigmentation on the bases of the anterior dorsal finlets. Sometimes a single small black pigment cell was present along the ventral edge of the trunk near the caudal peduncle, but more often one to three pig- 8 MATSUMOTO ET AL.: LARVAL TUNA IDENTIFICATION Table 4. — Characters used to separate larvae of Thunnus species having black pigmentation on trunk. Characters Thunnus thynnus (Atlantic) Thunnus thynnus (Pacific) Thunnus tonggot Thunnus maccoyii Thunnus ohesus Small larvae (3-10 mm SL): Number of black pigment cells: Upper jaw tip Lower jaw tip Dorsal edge trunk Lateral lino Ventral edge trunk Number of red pigment cells: Dorsal edge trunk Lateral line Ventral edge trunk Lower jaw ventral view Large larvae (>10 mm SL): Array of ^D^ pterygio- phores between two adjacent neural spines No observation 2 on inner edge 1 or 2 0-2 near mid-trunk M Streak on caudal peduncle'^ Indistinct! Streak anus to caudal peduncle^ Streak along margin anterior half of jaw and midline^ 1, 2, 2, 3, 2, 2, 1, 1 Appears above 6 mm SL 2 on inner edge above 4 mm SL 1 or 2 None 2 or more 1-5, mostly 3 Number not available Number not available 2 well spaced on anterior half No observation No observation No observation 1, 2, or more None 2 or more No observation No observation No observation No observation No observation Appears above 5 mm Few spots above 5 mm SL SL Appears above 4 mm 0-2 on inner edge SL below 4 mm SL 1 or 2, very small None 0 or 1 near mid-trunk None 1-3 1 or more No observation No observation No observation No observation No observation 0, 1, (2) 0, 1, 2, 3, 4 1-8 [mean = 5.3] 1 on each side near tip 1, 2, 2, 2, 3, 2, 1, 1 • Only one larva taken in a day tow was examined. 2 D2 refers to second dorsal fin. ment cells were present along the base of the posterior half of the anal fin. Red pigmentation did not differ from that in T. albacares. In larger larvae (10-13 mm SL) the array of pterygiophores of the second dorsal fin between two adjacent neural spines was sufficient to sep- arate T. thynnus from T. ohesus and both species from T. alalunga (Tables 2 and 3). In T. thyn- nus the greatest number of pterygiophores (3) between two adjacent neural spines appeared in the fourth position in the array, whereas in T. ohesus and T. alalunga it appeared in the fifth and sixth positions, respectively. T. ohesus was not distinguishable from T. alhacares by this character. The identification of T. atlanticus was not re- solved. No larvae from the distributional range of this species (tropical western Atlantic) have been found which are distinguishable from any of the species considered above. One of us (Richards) suspects that T. atlanticus larvae are very similar to larvae of T. ohesus. This suspicion is based on the great abundance of larvae resembling those of T. ohesus in this area, particularly at times and places where T. ohesus adults are rarely found or absent, studies are needed. Further SUMMARY OF LARVAL IDENTIFICATION On the basis of the examination and discussion above, the workshop agreed that: 1. The description of T. albacares by Matsu- moto (1958) was correct (see Figures 1 and 2), but that the "appearance of black pigmentation at the tip of the lower jaw at about 4.5 mm SL" should be included. 2. The description of T. alalunga by Yabe and Ueyanagi (1962) and illustrations by Ueyanagi (1969) were correct (see Figures 3 and 4), but that the lower size limit should be set at about 4.5 mm SL until further studies indicate more precisely the earlier appearance of black pig- mentation at the tip of the lower jaw in T. al- hacares. 3. It is not possible to separate larvae of T. alhacares from T. alalunga below 4.5 mm SL, prior to the appearance of black pigment cells at the tip of the lower jaw in T. alhacares. 4. The description of T. thynnus by Yabe, Ueyanagi, and Watanabe (1966) was correct FISHERY BULLETIN: VOL. 70, NO. 1 3.6 mm 6.6mm 9.6mm 4 4mm 5.3 mm 6.1mm Figure 5. — Larval stages of Thunnus thynnus, I. (From Yabe, Ueyanagi, and Watanabe, 1966. Lengths have been converted from total to standard.) (see Figures 5 and 6), and that there was no difference in T. thynnus from the Atlantic and Pacific Oceans. 5. The identification of T. tonggol was not substantiated by an adequate size series. 6. The description of T. maccoyii, based on tiny melanophores on the dorsal edge of the trunk, was not conclusive. 7. The description of T. ohesus by Matsumoto (1962) was correct, though it needed to be aug- mented by illustrations of a complete size series. 12.2 mm 16.8mm Figure 6. — Larval stages of Thunnus thynnus, IL (From Yabe, Ueyanagi, and Watanabe, 1966. Lengths have been converted from total to standard.) 8. The identity of T. atlanticus larvae is un- resolved. IDENTIFICATION OF JUVENILES In spite of the intention of the workshop to assemble as many specimens of juvenile tunas as possible, only a few juveniles of T. albacares and T. ohesus, not nearly enough to warrant their detailed examination, were available for study. The discussion on juvenile tuna identifi- cation, therefore, dealt mainly with published reports and with contributed data, resulting in a summary of identifying characters which the workshop considered useful and reliable. 10 MATSUMOTO ET AL. : LARVAL TUNA IDENTIFICATION Once the young tuna has acquired the full complement of spines and rays in all the fins, complete ossification of all the vertebrae, and the relocation of the anus near the origin of the anal fin, it is generally considered a juvenile of the species. Certain characters such as the full num- ber of gill rakers, however, develop much later, when the juvenile has attained a length of 40 or 45 mm SL. If we consider juveniles to in- clude all sizes up to the time of full gonad de- velopment signified by initial spawning, the size range of the juvenile stage would extend from about 13 mm SL to 700 mm FL (fork length) in T. albacares (Yuen and June, 1957) and to 860 mm FL in T. alalunga (Otsu and Hansen, 1962) . For the purpose of clarifying species identifi- cation of the young, however, individuals beyond 200 mm SL need not be included. The term ju- venile, as used here, thus refers to tunas between 13 and 200 mm SL. EVALUATION OF CHARACTERS The greatest difficulty in identifying juveniles of Thunnus is that the most useful characters are located internally. Except for the flattened first elongate haemal spine in T. alalunga, there is no single character that is peculiar to each of the species; but by using a combination of characters it should be possible to identify most of the other species. A summary of the most useful characters discussed is listed in Table 5. The size of juvenile at which each of the char- acters can be observed is listed also. Those char- acters whose usefulness in the early juvenile stages has not been shown conclusively are indi- cated by a question mark (?). The general formula of distribution of pterygiophores of the second dorsal fin has not been used before. The counts and descriptions given for those characters listed with a question mark gener- ally are those of the adults. These have not yet been substantiated for juveniles as well. Changes in the position of the first haemal arch with growth, for example, have been known to exist in other closely related fish such as the wahoo, Acanthocyhium solandri (Matsumoto, 1967). This could be true of the tunas also. Comparisons of body parts, particularly of or- bit diameter, body depth at origins of the first dorsal and anal fins, preanal and postanal dis- tances, and snout length, have not been investi- gated sufficiently in the past. The unavailability of specimens in sufficient numbers as well as the nonuniformity of body lengths (fork and stan- dard) used have contributed greatly to this ne- glect. Acceptance of standard length as the standard measure of body length and publishing of actual measurements in the future should help in the accumulation of sufficient data for anal- yses. This has to be done by all workers in this field of study, since the juveniles are not easily taken in large numbers. Table 5. — Characters for separating juveniles of Thnnnus species. Character Useful on juveniles obove Thunnus thynnus Thunnus alalunga Thunnus atlanticus Thunnus obesus Thunnus albacares First haemal arch ? 10 10 11 11 11 Ceratobranchial including angle 40 mm SL 17-20 15-16 12-13 15-16 15-16 Vertebrae 13 mm SL 18 + 21 18 + 21 19 -f 20 18 + 21 18 -f 21 Array of 1D2 pterygio- phores between two adjacent neural spines 10 mm SL 1, 2, 2, 3, 2, 2, 1, 1 1, 1, 2, 2, 2, 3, 2, 1 1, 2, 2, 2, 3, 2, 1, 1 1, 2, 2, 2, 3, 2, 1, 1 1, 2, 2, 2, 3, 2, 1, 1 First prezygapophysis and position on hoennal arch ? 15, 16, 17, high 15, 16, high 16, 17, low 15, 16, high 13, 14, low Postzygapophysis near first prezygapophysis ? Short, directed posterior Short, directed posterior Long, directed vertical or slightly anterior Short, directed posterior Long, directed vertical, some slightly anterior First haemal spine 30 mm SL Winglike at some stages Extremely wing- like Winglike at some stages — ~ Lateral line above base of pectoral fin 25 mm SL Acute, nearly 90° Obtuse Obtuse Obtuse Obtuse ^ Da refers to second dorsal fin. 11 FISHERY BULLETIN: VOL. 70. NO. 1 DISCUSSION AND SUMMARY T. thynnus below 25 mm SL can be separated from the other Thunnus species by the array of pterygiophores of the second dorsal fin, the last four positions containing 2, 2, 1, 1 pterygio- phores; in T. alabmga the sequence is 2, 3, 2, 1, and in T. atlanticus, T. obesus, and T. albacares it is 3, 2, 1, 1, T. thynnus above 25 mm SL can be separated from all other Thunnus by the sharp angle (nearly 90°) which the lateral line follows near the base of the pectoral fin; in all other species this angle is obtuse. In juveniles above 40 to 45 mm SL, T. thynnus has the highest number of gill rakers on the ceratobranchial, in- cluding that at the angle (Potthoff and Richards, 1970). T. alalunga below 30 mm SL can be separated from other Thunnus species by the distribution of pterygiophores of the second dorsal fin. Above 30 mm SL, T. alalunga is the only species whose first elongated haemal spine is flattened laterally and appears extremely winglike. T. atlanticus as small as 13 mm SL can be sep- arated from other Thunnus species by its dis- tinctive precaudal and caudal vertebral counts. It is the only species having 19 precaudal and 20 caudal vertebrae. Above 40 to 45 mm SL, this species can be separated from the others by the low (12-13) gill raker count on the cerato- branchial (Potthoff and Richards, 1970), in ad- dition to the vertebral formula. T. obesus and T. albacares are the only two species that cannot be distinguished from each other on the basis of internal characters. Com- parisons of body parts, i.e., orbit diameter, body depth or preanal and postanal distances, may have to be used. ACKNOWLEDGMENTS We thank John C. Marr, former Area Director, BCF Biological Laboratory, Honolulu, who orig- inated the idea of the workshop; Richard S. Shomura, former Acting Area Director of the same Laboratory, who continued with the orig- inal idea and organized the workshop ; and the BCF Biological Laboratory, Honolulu, Hawaii, for providing laboratory space and facilities. The workshop was supported entirely by the Bureau of Commercial Fisheries (now National Marine Fisheries Service). LITERATURE CITED Matsumoto, W. M. 1958. Description and distribution of larvae of four species of tuna in central Pacific waters. U.S. Fish Wildl. Serv., Fish. Bull. 58: 31-72. 1962. Identification of larvae of four species of tuna from the Indo-Pacific region I. Dana Rep. Carlsberg Found. 55, 16 p. 1967. Morphology and distribution of larval wahoo Acanthocybimn solandri (Cuvier) in the central Pacific Ocean. U.S. Fish Wildl. Serv., Fish. Bull. 66: 299-322. Otsu, T., and R. J. Hansen. 1962. Sexual maturity and spawning of the alba- core in the central South Pacific Ocean. U.S. Fish Wildl. Serv., Fish. Bull. 62: 151-161. Potthoff, T., and W. J. Richards. 1970. Juvenile bluefin tuna, Thunnus thynnus (Lin- naeus), and other scombrids taken by terns in the Dry Tortugas, Florida. Bull. Mar. Sci. 20: 389-413. Ueyanagi, S. 1966. On the red pigmentation of larval tuna and its usefulness in species identification. [In Jap- anese, English summary.] Rep. Nankai Reg. Fish. Lab. 24: 41-48. 1969. Observations on the distribution of tuna lar- vae in the Indo-Pacific Ocean with emphasis on the delineation of the spawning areas of albacore, Thunnus alalunga. [In Japanese, English sum- mary.] Bull. Far Seas Fish. Res. Lab. 2: 177-256. Yabe, H., and S. Ueyanagi. 1962. Contributions to the study of the early life history of the tunas. Occas. Pap. Nankai Reg. Fish. Res. Lab. 1: 57-72. Yabe, H., S. Ueyanagi, and H. Watanabe. 1966. Studies on the early life history of bluefin tuna Thunnus thynnus and on the larva of the southern bluefin tuna T. maccoyii. [In Japanese, English summary.] Rep. Nankai Reg. Fish. Res. Lab. 23: 95-129. Yuen, H. S. H., and F. C. June. 1957. Yellowfin tuna spawning in the central equa- torial Pacific. U.S. Fish Wildl. Serv., Fish. Bull. 57: 251-264. 12 THE FERTILIZATION OF GREAT CENTRAL LAKE. I. EFFECT OF PRIMARY PRODUCTION T. R. Parsons/ K. Stephens,- and M. Takahashi^ ABSTRACT Commercial fertilizer was added at a rate of 5 tons per week to a lake (51 km^, mean depth 200 m) over a period of 5 months from May to October 1970. As a result of these additions, surface primary produc- tion was increased approximately tenfold while the primary production of the euphotic zone was doubled. The standing stock of primary producers and water clarity were substantially the same as in the pre- vious year when no fertilizer was added. The productive index (mg C/mg Chi a/hr) was increased, especially in the immediate area of nutrient enrichment. However, the principal phytoplankton species were very similar at locations near and distant from the area of fertilization. In conclusion, it appears that as a result of adding nutrients at a low but sustained level, primary productivity was increased without substantially changing the nature of the food chain at the primary level of production. In the Pacific northwest, an earlier study (Nel- son and Edmondson, 1955) on the fertilization of a small salmon-producing lake in Alaska showed that the addition of phosphate and nitrate fer- tilizer increased the production of sockeye salm- on (Oncorhynchus nerka); in more recent studies by Donaldson et al. (1968), an increase in the production of steelhead trout (Salmo gairdneri) was demonstrated in a small lake in the state of Washington. The natural fertiliza- tion of lakes from decomposing salmon carcasses has been discussed by Krokhin (1967), who has suggested that the potential deficit from salmon removed by the fishery should be replaced by ar- tificial fertilization. In the report presented here we have carried out a fertilization experi- ment which differs from the two previous reports (Nelson and Edmondson, 1955; Donaldson et al., 1968) in several respects. These include the size scale of the experiment which was very much larger than any previous experiments, the application of fertilizer as a solution, control of the N:P ratio, and, finally, sustained weekly nu- trient additions over a period of 5 months. ^ Fisheries Research Board of Canada, Biological Sta- tion, Nanaimo, B.C.; present address: Institute of Oceanography, University of British Columbia, Van- couver, B.C., Canada. ^ Fisheries Research Board of Canada, Biological Sta- tion, Nanaimo, B.C., Canada. Manuscript accepted September 1971. FISHERY BULLETIN: VOL. 70, NO. 1, 1972. Preliminary results of our experiment have been reported (Parsons et al., in press) together with our conclusion that lake production was increased by the addition of fertilizer and that this was achieved without causing a condition of eutrophication. The following account deals specifically with the effect of nutrient enrich- ment on the primary level of production. Inten- sive studies on the effect of nutrient additions were carried out during the period May to August 1970 while a more general monitor pro- gram has been maintained from 1969 to the present (March 1971). The first sustained nu- trient additions were made during the period June to October 1970 and further additions are planned for the next 5 years. The primary purpose in this study is to in- crease levels of production in an oligotrophic lake, but not to change the trophic relationships which lead to the production of young sockeye salmon. In this respect the ultimate desideratum of the experiment is to produce larger sockeye smolts at their time of seaward migration; earlier reports have demonstrated that there is a close positive relationship between smolt size and survival (Ruggles, 1965; Johnson, 1965). Since previous studies (Parsons et al., in press) have shown that the migrant smolts from Great Central Lake are small (63 ± 1 mm) and that the primary productivity is very low (ca. 5 g 13 FISHERY BULLETIN: VOL. 70, NO. I C/m^/year) , application of nutrients at a level that would increase primary and secondary pro- duction seemed reasonable. Data used in this presentation have been ob- tained from Stephens et al, (1969') and Kennedy et al. (1971^). METHODS ANALYTICAL PROCEDURES Chlorophyll a, nutrients, oxygen, and total CO2 were all measured as described previously (Strickland and Parsons, 1968); bacteria were enumerated from plate counts after 24 hr in- cubation at room temperature on Millipore uni- versal medium; major phytoplankton species were enumerated after settling preserved sam- ples; conductivity was measured using a Beck- man Solu Bridge (Cedar Grove, N.J.). Primary productivity was measured as the difference in uptake of '^COo in light and dark bottles; how- ever, on a few days the dark-bottle uptake was exceptionally high and this requires further in- vestigation. For the purpose of this presenta- tion, data have been used only for days when the dark-bottle uptake was less than 20 9f of the maximum light-bottle uptake. Radiation was measured with an Epply pyra- nometer and corrected to give photosynthetically available radiation (PAR) as described previ- ously (Parsons and Anderson, 1970). Light attenuation was routinely measured with a Secchi disc (SD), and an empirical relationship between SD depth (m) and the vertical extinc- tion coefficient was established using a Schiiler meter (maximum response at 430 nm). This relationship for light at 430 nm was: K4.30 _ in — 2.1 SD depth The (total) extinction coefficient for the water column was then found from Jerlov's (1957) ' Stephens, K., R. Neuman, and S. Sheehan. 1969. Chemical and physical limnological observations, Babine Lake, British Columbia, 1963 and 1969, and Great Cen- tral Lake, British Columbia, 1969. Fish. Res. Board Can., Manuscr. Rep. 1065: 41-52. * Kennedy, 0. D., K. Stephens, R. J. LeBrasseur, T. R. Parsons, and M. Takahashi. 1971. Primary and sec- ondary production data for Great Central Lake, B.C., 1970. Fish. Res. Board Can., Manuscr. Rep. 1127. 379p. light attenuation curves. Mean radiation (Im) for the water column of depth (dm) was deter- mined from the expression im — h (1- u^nK -e-^^m) where h was the surface radiation and k was the attenuation coefficient for light below the surface meter. The expression was also used to determine the light at various depths in re- lation to the photosynthetic activity at those depths. NUTRIENT ADDITIONS The choice of a suitable fertilizer for the waters of Great Central Lake has been discussed previously (Parsons et al, in press). The nu- trient addition consisted of a commercial grade of ammonium phosphate and ammonium nitrate which contained trace quantities of other ele- ments essential for plant growth. The mixture is known commercially as 27-14-0 (27% N; 14% P2O5; 0% K2O) and has an N:P ratio of 10:1. The ammonium nitrate and ammonium phos- phate were dissolved separately in 5-ton amounts (total) and the concentrated solutions mixed be- fore distribution. A small quantity of organic material was added to the dissolved inorganic fertilizer at a dilution of 6 liters of fish solubles (obtained from B.C. Packers Ltd.) for every 2 tons of nutrient solution. The dissolved ferti- lizer was distributed at 10 gal/min (38 liters/ min) in the wake of a vessel travelling at approx- imately 8 knots. The area of nutrient additions is shown in Figure 1, together with sampling stations 1, 2, and 3. Station 1 was sampled dur- ing 1969 and 1970, Station 2 was sampled during 1970, and Station 3 was sampled sporadically during 1970 in order to check on the flow of nutrients in a westerly direction; in addition, areal surveys for chlorophyll a, transparency, and bacteria were carried out over the whole lake in order to determine within-lake variation. The area over which nutrients were added represented ca. 3 sq mi (8 km-) of lake surface; however, from studies on lake circulation it was apparent that the material was transported east 14 PARSONS, STEPHENS, and TAKAHASHI: LAKE FERTILIZATION. I. Figure 1. — Great Central Lake showing principal sam- pling stations (1, 2, and 3) and area of nutrient en- richment ( crosshatched ) . and west at rates of up to 6 miles per day or ca. 10 km/day (Parsons et al., in press). Thus while Station 2 was generally under the most immediate influence of the nutrient additions, Stations 1 and 3 also received an accumulative enrichment. Fertilizer was added at the rate of 5 tons per week from June through to October 1970. During May 1970 approximately 2 tons of fertilizer were added in experiments to deter- mine the rate of mixing and distribution of nu- trients in the vessel's wake. RESULTS LAKE MORPHOMETRY Great Central Lake is located on Vancouver Island, B.C., at lat 49°20' N on an east/ west axis between long 125°00' W and 125°25' W (Figure 1). It is a long narrow lake (ca. 33 X 1.5 km) with steep sides and a mean depth of 200 m. The yearly mean discharge is approximately 6 X lO^mVday with a range from 0.4 x 10^ to 32 X lO^mVday. TEMPERATURE The temperature structure at Station 1 is shown in Figure 2. The results are representa- tive for the open waters of the whole lake, and it is apparent that the lake was isothermal dur- ing January and February; a thermocline started to form during March and was well established by May. Maximum surface tem- perature during July was 21.2° C; surface cool- ing started in September but a thermocline of Figure 2.— Temperature (°C) stratification at Station 1. 10° persisted through October and the lake did not become isothermal until January of the fol- lowing year. RADIATION Changes in photosynthetically active radiation (PAR) at the lake surface are shown in Figure 3 together with the mean radiation for the water column 0 to 20 m, calculated -on a 24-hr basis. Figure 3. — (A) Surface photosynthetically active ra- diation (PAR) and (B) 15-day mean PAR in the first 20 m. From the latter results it is apparent that radi- ation in the water column increased by 50% from the beginning of May until the middle of June; the decrease in radiation during the second part of June was due to a combination of higher ex- tinction coefficients and lower surface radiation. The average radiation remained virtually con- stant during July and decreased by 20% during the latter half of August. CHLOROPHYLL A Surface chlorophyll a concentrations are pre- sented in Figure 5 in the same way and for the same stations and years as SD data in Figure 4. The two figures have some mirrored similarities ; 15 FISHERY BULLETIN: VOL. 70, NO. 1 14 i t- Q. UJ O 10 I o o 8 1 ii A lo o !,° i 1 ll'" ■ I ■ ,* B I ■ STATION I A 1970 2 ■ 1970 STATION I o 1969 JUNE JULY AUGUST Figure 4. — Secchi disc depth at Stations 1 and 2, 1970, and Station 1, 1969. (Mean and standard deviation of values from areal surveys shown as bars.) ^ 0.6 0.4 - 0.2 0.0 STATION I A 1970 2 ■ 1970 STATION I o 1969 m I i ■'A fc i ' 1 A o JUNE ± JULY AUGUST Figure 5. — Surface chlorophyll a data at Stations 1 and 2, 1970, and Station 1, 1969. (Mean and standard de- viation of values from areal surveys shown as bars.) thus Station 2 chlorophyll a values from June to August were generally higher than either Sta- tion 1 data for 1969 or 1970; minimum mean SD data (8 to 10 m) occurred between June and July during a maximum in the mean chlorophyll a concentration. However, 1969 chlorophyll a data at Station 1 do not appear to be significantly different from 1970 chlorophyll a data at the same station. The depth distribution of chlorophyll a gen- erally showed a maximum between 10 and 20 m following stratification and nutrient depletion in the surface layers. pH, CALCIUM, TOTAL CO,, AND CONDUCTIVITY pH values were generally between 7,1 and 8.3 with some indication of a seasonal cycle towards higher pH values in summer. Several assays for calcium showed a concentration of 5 mg/liter while specific conductivity was very consistent at 33 ix mhos/cm, except in the immediate vi- cinity of small streams entering the lake; total carbon dioxide varied over a range from about 2.2 to 4.2 mg C/liter. OXYGEN Oxygen profiles to 200 m showed that surface oxygen concentrations were between 80 and 90% saturation during winter and up to 110% sat- uration during summer. Deepwater oxygen concentrations appeared constant at around 10 mg/liter or about 80% saturation. An oxy- gen maximum occurred at ca. 20 m during the summer. NITRATE, AMMONIA, PHOSPHATE, AND SILICATE Nitrate depth profiles at Station 1 during May to October, 1969 and 1970, are shown in Figure 6. The general form of the two profiles is similar; thus a depletion in the winter level of nitrate (1.0 to 2.0 fjig at./liter) becomes apparent to- wards the end of May and by the end of June about 1 iJLg at. NOnN/liter has been removed from the water column, 0 to 10 m. During July and August nitrate in the first 10 to 15 m is close to the limit of detection, but there is a partial return to winter levels during September and October. Some difl["erence in the form of these events is apparent between 1969 and 1970; the utilization of nitrate was more rapid and apparently more complete during 1970; in addition, surface ni- trates did not increase in September-October 1970 as they did in 1969. Starting from a winter level of 2 yug at. NO.tN/ liter, the total utilization of nitrate in the water column has been determined for the periods Feb- ruary to May, June, and July-August using data shown in Figure 7. The accumulative amount 16 PARSONS. STEPHENS, and TAKAHASHI : LAKE FERTILIZATION. I. MAY I JUNE I JULY ' AUG I SEPT IQCT 197 0 Figure 6. — Nitrate (^g at./liter) profiles at Station 1, May to October 1969 and 1970. 1200 ^, 1000 - E °- 800 ■o 0) '£ 600 Z. 400 E 200 • N utilized in the woter column A N added as fertilizer O total N utilized F'MA'M'J'JASO Figure 7. — N utilization at Station 1. of inorganic nitrogen added as fertilizer (ex- pressed per m- for the entire 51 km- lake sur- face) is also shown; since this was utilized with- in hours following each addition, the total ni- trogen budget is represented as the sum of the natural and added inorganic nitrogen. Some mixing occurred during September and October, and the utilization of inorganic N during this period is shown as an indefinite extrapolation of the nitrogen utilized by the end of August. From these curves and Figure 5 it may be seen that the fertilizer was the principal source of new nitrogen during the period July-August when the lake nitrate was practically exhausted in the euphotic zone. Ammonia values tended to show sporadic in- creases during 1970, and at times ammonia may have been the principal inorganic form of ni- trogen in the lake, probably through being re- cycled as excretory products of the zooplankton (Beers, 1962). However, due to analytical dif- ficulties with this radical, further investigation of its seasonal behavior is required, especially with reference to the verification of high values. Phosphate showed similar variations to nitrate although the depletion of phosphate was less reg- ular. Seasonal concentrations ranged from <0.01 to 0.04 fxg at. P/liter with about 3% of the values falling in a much higher range of 0.1 to 0.6 /Jig at./liter. A determination of phos- phate utilized and phosphate added (similar to the inorganic N budget shown in Figure 7) was difficult to describe because of the unpredictable occurrence of phosphate throughout the summer; this may have been due to phosphate regener- ation. As an overall assessment, however, if a winter level of 0.03 fig at. P/liter were complete- ly utilized in the water column 0 to 30 m, the addition of 100 tons of 27-14-0 would increase the supply of phosphate over the whole lake by a factor of about 450% compared with the increase in the inorganic nitrogen budget of ap- proximately 100% (Figure 7). From winter to summer, silicate concentra- tions ranged from about 1.8 to 3.0 mg silica/ liter. According to Lund (1965) silicate be- comes rate limiting for diatoms at about 0.5 mg/ liter, which is considerably lower than the sea- sonal range for Great Central Lake. BACTERIA Plate counts of bacterial colonies per 100 ml are shown in Figure 8, together with the range of counts obtained on several days when areal surveys were made. During May, the total num- ber of colonies per 100 ml was generally below the mean value of ca. 9,000 reported by Henrici (1940) for oligotrophic lakes; however, there is 17 FISHERY BULLETIN: VOL. 70, NO. 1 >30,000 - 10,000 - o o ac UJ Q. (O UJ o o -J < a: < ffi 1,000 < 100 I I (6) (10) O-r f •O I (3) (1 0) " (9) * • ' 4 • • • o • < ' (? )) o -' ) • - o o o o lOOA 1 1 MAY JUNE JULY AUG Figure 8. — Bacterial colonies per 100 ml surface lake water (O) Station 2; (•) Station 1; (5) number of samples in areal survey and range, I. some indication in the data that bacterial num- bers increased by one or two orders of magnitude during the latter part of June through to August. Summer increases in bacterial flora have been widely observed in lakes (e.g., Snow and Fred, 1926; Nauwerck, 1963) , and while nutrient level could have affected this increase (e.g., see Bosset, 1965), we have no previous data on which to judge the effect. PHYTOPLANKTON SPECIES Principal phytoplankton species from surface samples at Station 1 and 2 during 1970 are shown in Figure 9 on a relative scale. From these re- sults it is apparent that the predominant algae during May and early June were Dinobryon, Rhizosolenia, and Nitzschia. During June and July Gymnodinium, Cyclotella, and the euglenoid Phncus reached maximum numbers but tended to decline by August. Predominant algae of late summer and autumn were the chlorophyte Nan- nochloris and the cyanophyte Chroococcus. Sec- ond maxima in Dinobryon, Rhizosolenia, and Cyclotella occurred during the winter together with a maximum in Tabellaria. Two studies (May and June) on the depth distribution of the principal species showed that maxima in Rhizosolenia, Tabellaria, and Phaciis were found at the bottom of the thermocline (ca. 20 m) ; Cyclotella and Gymnodinium maxima occurred at the top of the thermocline (ca. 10 m) while Nannochloris, Dinobryon, Nitzschia, and Chroococcus showed maxima within the top 0 to 10 m. PRIMARY PRODUCTION Surface primary production values at Station 1 during 1969 and 1970 and at Station 2 during 1970 are shown in Figure 10; the mean and co- efficient of variation of surface primary pro- duction for the months of June to August are also shown on each figure. The total average pri- mary production in the water column 0 to 30 m at Stations 1 and 2 during 1970 was approxi- mately 12 g C/m^/year compared with approx- imately 6 g C/mVyear at Station 1 during 1969. Primary production per unit of chlorophyll a at different depths for Station 1, 1969 and 1970, and Station 2, 1970, are shown plotted against the light intensity at the same depths in Figure 11. A considerable amount of scatter is appar- ent in the data which is partly due to differences in environmental factors as well as to the lack of precision in attempting to establish photo- synthesis versus light intensity relationships on the basis of ecological rather than experimental data. Polynomial curves were fitted to each set of data using an IBM computer. The shape of these curves is consistent with P vs. / relation- ships obtained by physiologists under experi- mental laboratory conditions and differences in asymptotic values reflect differences in the nu- trient supply (Ichimura and Aruga, 1964). 18 PARSONS, STEPHENS, and TAKAHASHI: LAKE FERTILIZATION. I. Dinobryon (100%= 20 4 cells /ml) Cy dote II a (100% = 8 80 cells/ml) Nannochloris (100% =3880 cells/ml) Months M onths Figure 9.— Principal phytoplankton species at Station 1 (•), Station 2 (A), Station 3 (O) during 1970. 19 4.0 35 3.0 2.5 "£2.0 E 1.5- 1.0 0.5- Station I 1969 N 10 P 0.18 CF 72% Jl Station I 1970 N_ II P0.49 CF 69% Station 2 1970 N 10 PI.63 CF 76% T LJ, M ' J ' J ' A ■ M ' J 'J' A I M ' j ' J ' A MONTHS Figure 10. — Surface primary production, May to August. (N = number of samples, P ^= mean surface produc- tion, and CF ^ coefficient of variation — all values for the period, June to August.) However the degree of scatter in the ecological data requires some expression of confidence limits. At Station 1 (1969), which was located at a considerable distance from the area of fertil- ization, 95 /f confidence limits for the asymptotic value of 1.03 mg C/mg Chi a/hr were 0.84 and 1.22; for 1970 at the same station the 95% con- fidence limits for the asymptotic value of 1.55 Figure 11. — Productivity indices plotted against light intensity at Station 1, 1969 and 1970, and Station 2, 1970 (O O computed best fitting polynomial curve). 2.0 1.5 1.0 0.5 FISHERY BULLETIN: VOL. 70, NO. 1 ST. I, 1969 3.0 2.5 -c 2.0 o M .»: » J L 1.5 o £ 1.0 05 14.0 12.0 10.0 8.0 6.0 4.0 2.0 ST. I, 1970 ST. 2,1970 J_ 0.1 0.2 0.3 0.4 I y / mifi L L_ a5 06^ 20 PARSONS, STEPHENS, and TAKAHASHI: LAKE FERTILIZATION. I. mg C/mg Chi a/hr were 1.12 and 1.98. At Sta- tion 2 in 1970, however, the scatter of points is so great that 95% confidence hmits become very large. The probable reason for this is that the station was sometimes in the area to which nu- trients were first added, and sometimes the move- ment of water containing freshly added nutrients was away from Station 2 (Figure 1) . If in fact it is assumed that there were only two alter- natives in such a narrow lake (i.e., movement of nutrients towards or away from Station 2) then the 50% confidence limits for the asymp- totic value of 4.17 mg C/mg Chi a/hr were 2.26 and 6.07. DISCUSSION The principal purpose of this report is to establish the effect of inorganic nutrient enrich- ment on the primary production of Great Central Lake. From data in Figure 10 it is quite ap- parent that primary productivity was increased in surface samples during 1970 compared with 1969, both at Station 1 and particularly at Sta- tion 2, which was very close to the area of re- peated enrichment. However, while the effect of nutrient enrichment was apparent to the extent of a tenfold increase in surface primary pro- ductivity, the integrated productivity for the water column only showed an approximate dou- bling in primary productivity during the first 3 months of nutrient enrichment (see Parsons et al, in press, for primary production depth pro- files) . This result is in keeping with the fact that the total inorganic nitrogen addition to the lake (Figure 7) was only sufficient to approx- imately double the natural reservoir of inorganic nitrogen in the upper 10 m, based on winter ni- trate levels. However, it does not take into ac- count nitrogen fixation by the blue-green alga, Chroococcus, which may have taken arvantage of the increased supply of phosphate to become one of the predominant summer plankters. The question is, whether some factor other than fertilization could have accounted for the increased primary productivity? Firstly, it is apparent that since the largest increase in pri- mary productivity occurred at the surface, it can- not be argued that the increased primary produc- tivity was due to greater enrichment of the water column from the hypolimnion, especially in view of the high degree of stratification (Figure 3) and apparent nitrate depletion in the epilimnion (Figure 6). It might be argued that the in- creased productivity was due to an increase in standing stock of primary producers and in- creased radiation. Data in Figure 5 indicate that the standing stock of primary producers at Station 2 was generally higher than at Station 1 during 1969, although the effect is within a 95% probability of being due to within-lake variations in standing stock of chlorophyll a. However, in order to examine this question in more detail, primary productivity data for Sta- tions 1 and 2 in 1970 and Station 1 in 1969 have been expressed as the production per unit chlor- ophyll ft and plotted against the calculated light intensity at various depths (Figure 11). This presentation of data has been used by Ichimura and Aruga (1964) to compare the productivity of oligotrophic, mesotrophic, and eutrophic lakes under conditions of different standing stocks of primary producers, light conditions, and photo- synthesis. From their findings it was concluded that oligotrophic lakes had a productive index of between 0.1 and 1.0 mg C/mg Chi a/hr, which is very similar to the range of values computed from the data in Figure 11 for Station 1 during 1969. The computed range for Station 1 during 1970 was appreciably higher, however, and en- ters the classification for mesotrophic lakes which have a photosynthetic index of up to 2 mg C/mg Chi a/hr; finally the asymptotic value (4.17) from Station 2 in 1970 is within Ichi- mura's and Aruga's (1964) range for eutrophic lakes, which the authors report as having photo- sjnithetic indices of up to 6 mg C/mg Chi a/hr. Since the only basis for this classification is the effect of nutrient enrichment in enhancing the photosynthetic response, it may be concluded that our observed increase in primary produc- tivity was determined by the artificial addition of fertilizer. Secondary effects of nutrient enrichment may also have influenced the primary formation of particulate material through a heterotrophic cycle. Unfortunately, our evidence for this is not substantial and rests mainly on the increase 21 FISHERY BULLETIN: VOL. 70, NO. I in bacterial numbers (Figure 8) and the fact that very high dark uptake of ^^C-bicarbonate (up to 50^; of the light bottle uptake) were encountered during the summer at some stations following fertilization. We are at present not sure of the accuracy of this result, however, and it will be reinvestigated during 1971. Nauwerck (1963) has concluded that the heterotrophic for- mation of particulate material is a principal mechanism for supplying food to particle feeders in some lakes and one might expect this mech- anism to be enhanced by the additional avail- ability of nitrogen and phosphorus. The most interesting aspect of changes in the species composition of the principal primary producers is that in spite of differences in sur- face primary productivity at Stations 1 and 2 during 1970 (Figure 10) the relative abundance of principal species at these two stations (and on several occasions at Station 3) was substan- tially the same (Figure 9) . This was important because it was intended that there should be no change in the species composition of organisms leading up the food chain to young salmon, but only an increase in their productivity. In ad- dition, the occurrence of the Cyclotella-Chroococ- cus association is characteristic of oligotrophic lakes (Hutchinson, 1967) which indicates that the general classification of the lake (based on species association) had not been changed by fertilization. However, some eutrophic species of phytoplankton, such as Ceratium, Peridwium, and Scenedesmtis, were also observed as minor constituents of the plankton, especially during the summer. In conclusion, it appears that the fertilization of Great Central Lake resulted in an increased primary production but did not substantially change the standing stock of primary producers, water clarity, or the principal phytoplankton species at locations near and distant from the site of nutrient enrichment. The effect of zoo- plankton on the primary producers was essen- tially to suppress the increase in standing stock of phytoplankton while the standing stock of zooplankton itself increased by almost an order of magnitude. Zooplankton growth and distri- bution are described in the second paper in this series (LeBrasseur and Kennedy, 1972). ACKNOWLEDGMENT The authors wish to acknowledge the assist- ance of S. Sheehan in carrying out analyses of water samples. LITERATURE CITED Beers, J. 1962. Ammonia and inorganic phosphorus excre- tion by the planktonic chaetognath Sagitta lispida Conant. In J. H. Ryther and D. W. Menzel (edi- tors), The biochemical circulation of elements in the Sargasso Sea. Append., Prog. Rep. 1, Sept. 1961 to 31 Mar. 1962, Bermuda Biol. Stn., 6 p. BOSSET, E. 1965. Incidences hygieniques de la vaccination des eaux de boisson au moyen de polyphosphates. MonatsbuU. Schweiz. Ver. Gas-u. Wasserfachm. 45: 146-148. Donaldson, L. R., S. M. Olsen, P. R. Olson, Z. F. Short, J. C. Olsen, H. E. Klassen, and R. W. Kiser. 1968. Fern Lake program. In Research in fish- eries . . . 1967, p. 38-39. Coll. Fish. Fish. Res. Inst., Univ. Wash. Contrib. 280. Henrici, a. T. 1940. The distribution of bacteria in lakes. Am.. Assoc. Adv. Sci., Publ. 10: 39-64. Hutchinson, G. E. 1967. A treatise on limnology. II. Introduction to lake biology and the limnoplankton. John Wiley and Sons Inc., p. 355-397. ICHIMURA, S., AND Y. ArUGA. 1964. Photosynthetic natures of natural algal com- munities in Japanese waters. In Y. Miyake (edi- tor) , Recent researches in the fields of the hydro- sphere, atmosphere and nuclear geochemistry, p. 13-37. Maruzen Co., Ltd., Tokyo, Jerlov, N. G. 1957. Optical studies of ocean waters. Rep. Sweden Deep Sea Exped. 3: 1-59. Johnson, W. E. 1965. On mechanisms of self-regulation of popu- lation abundance in Oncorhynchus nerka. Mitt. Int. Ver. Limnol. 13: 66-87. Krokhin, E. M. 1967. Influence of the intensity of passage of the sockeye salmon Oncorhynchus nerka (Wald.) on the phosphate content of spawning lakes. Izd. "Nauka", Leningrad 15(18): 26-31. (Fish. Res. Board Can. Trans. Ser. 1273.) LeBrasseur, R. J., and 0. D. Kennedy, 1972. The fertilization of Great Central Lake. II. Zooplankton standing stock. Fish. Bull., U.S. 70: 25-36, 22 PARSONS, STEPHENS, and TAKAHASHI: LAK£ FERTILIZATION. I. Lund, J. W. G. 1965. The ecology of the freshwater phytoplankton. Biol. Rev. (Cambridge) 40: 231-293. Nauwerck, a. 1963. Die Beziehungen zwischen Zooplankton und Phytoplankton im Zee Erken. Symb. Bot. Ups. 8(5) : 1-163. Nelson, P. R., and W. T. Edmondson. 1955. Limnological effects of fertilizing Bare Lake, Alaska. U.S. Fish Wildl. Serv., Fish. Bull. 56: 414-436. Parsons, T. R., and G. C. Anderson. 1970. Large scale studies of primary production in the North Pacific Ocean. Deep-Sea Res. 17: 765-776. Parsons, T. R., C. D. McAllister, R. J. LeBrasseur, AND W. E. BARRACLOUGH. In press. The use of nutrients in the enrichment of sockeye salmon nursery lakes - a preliminary re- port. FAO Technical Conference on Marine Pollution, Rome, Dec. 9-18, 1970. RUGGLES, C. p. 1965. Juvenile sockeye studies in Owikeno Lake, British Columbia. Can. Fish. Cult. 36: 3-21. Snow, L. M., and E. B. Fred. 1926. Some characteristics of the bacteria of Lake Mendota. Trans. Wis. Acad. Sci. Arts Lett. 22: 143-154. Strickland, J. D. H., and T.R. Parsons. 1968. A practical handbook of seawater analysis. Fish. Res. Board Can., Bull. 167, 311 p. 23 THE FERTILIZATION OF GREAT CENTRAL LAKE 11. ZOOPLANKTON STANDING STOCK R. J. LeBrasseur and 0. D. Kennedy^ ABSTRACT The regional, vertical, and seasonal abundance of the dominant zooplankton species were studied in con- junction with a series of nutrient additions to Great Central Lake. Two rotifer species, Kellicottia spp., Conochilus unicornis, three cladocera species, Bosmina coregoni, Holopedium gibherum, and Daphnia longiremis, and three copepod species, Cyclops bicuspidutns thomasi, Epischura nevadensis, and Diap- tomus oregonensis were the most numerically abundant zooplankton species. The introduction of the fertilizer and the consequent higher rate of primary production produced no changes in the species com- position. The zooplankton exhibited a relatively uniform horizontal distribution within the upper 20 m along the lake, a factor which was attributed to the lake circulation. All eight species were concentrated in the euphotic zone (upper 40 m), and five were most abundant in the upper 10 m. The center of abun- dance for the remaining three species was between 20 and 30 m depth. The respective depths of maximum abundance for the various species showed little variation between daylight and darkness. Seasonally, there were two periods, June to July and September to October, of maximum abundance for most spe- cies. The cause for somewhat lower levels of abundance in August is not known. The average zoo- plankton biomass showed a similar seasonal pattern with a maximum weight in July which exceeded 8 g/m2. The average biomass over a 6-month period, May through October, exceeded 5 g/m^ (more than 10 times greater than for the comparable period prior to fertilization in 1969). In contrast to the high standing stock of zooplankton, the estimated growth rate for underyearling sockeye salmon, the principal predator species in the lake, was only slightly improved over 1969 (1.2 vs. 0.99^ /day). In comparison with other lakes producing young salmon the growth rates appear low with respect to the zooplankton stock. It was suggested that the temperature structure of the lake, 14° to 23°C above the thermocline and 4° to 6°C below the thermocline, may reduce availability and prevent the efficient utilization of the zooplankton by the underyearling sockeye salmon. The following account is the second in a series of papers which report on the effects of sustained nutrient additions to an oligotrophic lake. In the first report, Parsons et al. (1972) showed that an increased primary productivity resulted from nutrient additions made to Great Central Lake, B.C.; the objective of this report is to de- termine if nutrient additions affected the stand- ing stock and diversity of secondary produc- ers. The overall purpose of these studies has been to determine if nutrient additions will increase sockeye salmon (Oncorhynchus nerka) produc- tion; zooplankton, as the principal food of under- yearling sockeye salmon, occupy a central posi- tion in the food chain of young sockeye during lake residence. Previous studies (Ivlev, 1961; ^ Fisheries Research Board of Canada, Biological Sta- tion, Nanaimo, B.C., Canada. Johnson, 1965; Brocksen et al., 1970) have sug- gested that prey density and availability may limit the predator biomass. The latter authors compiled data for several sockeye nursery lakes with which they were able to demonstrate a direct relationship between mean zooplankton biomass (prey) and the mean growth rate and biomass of underyearling sockeye salmon (pred- ator). Other studies (Ricker, 1962) have indi- cated that the ocean survival, i.e. the return to coastal waters of adult sockeye salmon, can be directly correlated in many instances with the size at which the sockeye as year-old migrants leave the nursery lakes to enter the ocean for 2 or more years. While the above studies rely heavily upon circumstantial data, as well as data which were collected for other purposes, they serve as a rational basis for attempting to increase the available zooplankton biomass for the enhancement of salmon growth. Manuscript accepted September 1971. FISHERY BULLETIN: VOL. 70, NO. 1, 1972. 25 FISHERY BULLETIN: VOL. 70, NO. 1 METHODS Sampling of zooplankton was initiated in mid- 1969, using a 0.25 m- mouth area cylinder-cone net with 100 micron mesh aperture, hauled ver- tically from 20 m or 50 m. The samples were collected at infrequent intervals during 1969 and the first 14 weeks in 1970; thereafter vertical net hauls were made at two locations at least once every 4 days until the first week of No- vember. Additional vertical net hauls were made once or twice each month from the lake bottom, 200 m, to the surface. Miller nets (Mil- ler, 1961) with 0.01 m- mouth area and 100 mi- cron mesh aperture were used at weekly intervals during the period June through August and thereafter at monthly intervals to determine the area] and vertical distribution of zooplankton. The areal sampling at 18 locations along the lake consisted of 5 min oblique tows from 20 m to the surface while underway at 2 m/sec. The daylight vertical distribution of zooplankton was monitored at 18 depths between the surface and 65 m by making three consecutive tows each with six Miller nets at 2 m/sec at one location. Ad- ditional tows were made to sample other depths and also at other periods of the day. Details of the sampling and sampling locations are re- ported elsewhere (Kennedy et al., 1971"). ' Kennedy, 0. D., K. Stephens, R. J. LeBrasseur, T. R. Parsons, and M. Takahashi. 1971. Primary and sec- ondary productivity data for Great Central Lake, B.C., 1970. Fish. Res. Board Can., Manuscr. Rep. No. 1127, 379 p. In the analyses of samples special effort was made to maintain up to date species counts and measurements for comparison with other events as they were occurring in the lake. The common zooplankton constituents were identified, mea- sured into size categories, and counted from an aliquot of the total sample ; fractions of 1/50 or 1/100 using a Stempel pipette were used de- pending upon the sample size. The size cate- gories (length in microns) reported in Table 1 were based upon individual measurements for different stages of development of the respective species. It is to be noted that the lengths refer to mean sizes of organisms during the spring and summer growing period. Individual length measurements for the different species and for diflferent times of the year may be found in the MS data report (see footnote 2). Species counts in vertical net hauls are re- ported as number per m^, in oblique and hori- zontal tows as number per m^ In this report, unless otherwise indicated, counts all refer to numbers of individuals which fall within the size range occupied by mature (Stage VI) copepods and egg-bearing cladocera; these were usually the two largest size groups for the species re- ported in Table 1. RESULTS SPECIES Table 2 lists the species of zooplankton which have been found in Great Central Lake. Addi- tional species may be present as minor constitu- Species Table 1. — Zooplankton size ranges for species sorting in Great Central Lake. Group III IV VI VII Cyclops bicuspidatui Egg 125-275 C. vernalii Egg 275-375 Epischura nevadensis Egg 275-450 Diaplomus orfgoneniis Egg 125-275 D. kenai Egg Bosmina Egg 125-225 llolopedium Egg 375-450 Daphnia longirtmii Egg 450-750 D. pultx Egg 450-650 Kellicottia Egg ca. 80 Conochilus Egg ca. 80 Keratella Egg ca. 80 Size range y. 375-550 550- 750 750- 850 850- 950 950-1,100 750- 900 900-1,100 1.100-1,350 450-650 650- 900 900-1,100 1,100-1,350 1,350-2,250 450-750 750- 900 900-1,100 1,100-1,350 650-1,100 1,100-1,750 1,750-2,500 225-325 325- 450 450- 650 650- 600 450-750 750-1,100 1,100-1,750 450-750 750-1,100 1,100-1,750 450-750 750-1,100 1,100-1,750 1,750-2,500 26 LeBRASSEUR and KENNEDY; LAKE FERTILIZATION. II. Table 2. — Zooplankton species found in Great Central Lake, 1970. Rotifera *Kellicottia spp. Ktratella cochlearis K. quadrata *Conockilus unicornis Cladocera *Bosmina coregoni *Holopedium gibberum "Daphnia longiremis D. pulex Scapholeberis kingi Polyphemus pediculus Alona affinis Copepoda *Cyclops bicuspidatus thomasi C. vernalis *Epischura nevadensis *Diaptomus oregonensis D. kenai Unknown Actinopoda Pollen Egg clusters Arachnoidea (mites— 2 spp.) Chironomid larvae Fish larvae (cottid) * Indicates the most common species. ents of the zooplankton and it is also possible that new species are being introduced into the lake through a hydroelectric installation which discharges water from an adjacent watershed into the lake. It will be noted from Table 2 that the common zooplankton constituents con- sisted of two rotifer species, three species of cladocera, and three species of copepods; these species are identified throughout the text by their generic names. There has been no change in the species composition during the course of the experiment, i.e. the common species have re- mained numerically abundant while the rare spe- cies have continued to occupy a minor role. PATCHINESS It was anticipated that the zooplankton would exhibit contagious distributions reflecting local circulation patterns, species preferences, and predation. Accordingly, oblique samples from 20 m were collected at weekly intervals at 18 positions along the lake, both near the shore and in midlake. In general, with the exception of the area near the inlet and outlet of the lake where the abundance of organisms was some- times low, there was greater variability found with respect to the date of sampling than the lo- cation of sampling. Weekly means and standard deviations computed for each species showed that Cyclops was the only species in which the standard deviation exceeded the weekly count for more than half the surveys (11 out of 17). The apparent variability in Cyclops abundance might be due contagion or, more likely, to the fact that sampling was limited to depths (20 m) where Cyclops were seldom abundant (see sec- tion on vertical distribution) . The major source of variability in weekly mean counts appears to be associated with the number of organisms counted, i.e. the number of organisms in a sample and the size of the aliquot counted. The weekly mean number of organisms for each species were grouped together with their respective standard deviations as follows: 1-50, 51-250, 251-500, 501-1,000, 1,001-2,000, 2,001-5,000. The mean coefficient of variation (C.V.), the range, the number of means present in each group, and the number of times a standard deviation ex- ceeded its respective mean are shown in Figure 1 (e.g. for 50 or fewer organisms counted, the standard deviation in 17 out of 23 samples ex- ceeded the mean). The magnitude of C.V., or the relative variation about a mean, is closely associated with the number of organisms counted. The high degree of variability about a mean of 50 or fewer organisms reflects counting errors due to the subsampling technique used in the initial analyses of the samples. However, counts of organisms of 250 or more per m^ tend 5100 c 50 o o 1 1 r 2 000 Meon (No/m3) Figure 1. — Coefficient of variation computed for mean counts of species sampled in oblique (20 m to surface) tows, where N is the number of weekly means, m and cr is the standard deviation. 27 FISHERY BULLETIN: VOL. 70, NO. 1 to be relatively uniform, suggesting that hori- zontal patchiness or local contagion is not a ty- pical feature of Great Central Lake zooplankton. The observations of lake circulation (McAllister, personal communication) (Parsons et al., in press), and the chlorophyll a distribution (Par- sons et al, 1972) confirm that the epilimnion is well mixed, thus assuring a nearly uniform dis- persal of planktonic organisms along the lake. The 50-m vertical hauls made at Stations 1 and 2 provide further opportunity for examining the variability with respect to different sampling lo- cations. Here, the comparisons of mean counts indicated a high degree of similarity between the two locations with respect to species comj^osition, stage of development, and abundance. However, examination of samples collected on the same day indicated a high degree of variability. For- ty-nine samples were collected from Stations 1 and 2 during May through December; species counts for Station 1 were plotted against the re- spective count for Station 2. Values which fell outside of a mean ±- half the expected mean (where the expected mean equals half the counts for Stations 1 and 2 combined) are tabulated in Table 3. On half the sampling dates the counts for a particular species tended to be si- milar at both locations (e.g. in the first column of Table 3, the number of samples with a mean ± m/2 is generally greater than half the total number of samples, N/2). Greater numbers of four species were found at Station 2 than at Station 1. It is noteworthy that three of the four species, Kellicottia, Cyclops, and Daphnia, have their greatest abundance below the epilimnion at depths greater than that sampled on the areal surveys. However, there was no apparent cor- Table 3. — Comparison of counts of zooplankton from 50-m vertical hauls at Stations 1 and 2, N = 49. Species Counts = (m : 2 Station I >(m + m) Station 2 <(m — m) 2 Cyclops 33 5 11 Epischura 31 10 8 Diaptomus 24 10 15 Bosmina 26 17 6 Ilolopfdium 25 15 9 Daphnia 22 6 21 Kellicottia 24 9 16 Conochilus 29 14 6 relation in the relative abundance of any of these three species with respect to each other or to other species at either station. There were, however, periods when three or more species would be more abundant at one po- sition than at the other. For example, from July 3 to July 21 (six sets of samples) three to six species were most abundant at Station 1 while during the period August 21 to September 8 (six sets of samples) three to seven species were most numerous at Station 2. Similarly, for other periods of 4 to 12 days, one or another species was in greater abundance at one station than the other. These data have not been examined further to show if variation in species abundance be- tween sampling positions can be correlated with variations in the lake circulation or other envi- ronmental factors such as fertilization or pre- dation by underyearling sockeye. However, it is apparent that all seasonal changes in species composition and abundance were reflected throughout the near-surface waters of the lake and that no local area of high or low zooplankton concentration could be clearly defined within the main body of the lake. VERTICAL DISTRIBUTION Horizontal tows made within the upper 60 m revealed marked differences in species compo- sition and abundance with depth during the pe- riod of thermal stratification. As an example the weekly tows made during July were com- bined and the average concentration of each spe- cies at each of 17 depths sampled during daylight is shown in Figure 2. The inset associated with each species distribution shows the relative dis- tribution (25% quartile intervals) of the respec- tive populations sampled during a 24-hr period in August. Five of the eight species shown in Figure 2 have their maximum concentration within the upper 10 m, while the maximum concentration of the other three species was below 20 m. Thus the species maxima fall either above or below the thermocline. However, it should be noted that the number of organisms per m^ decreased from a maximum of greater than 7,000/m^ be- 28 LeBRASSEUR and KENNEDY: LAKE FERTILIZATION. II. CofiOcMji colonel 400 'III CfClops (.9001 No/m^ 200 400 D''omus ()900) No/m' Epischura (i 900 1 No/m^ 300 600 ■ J \ 1_ Boiiofw al thetmoclnw *6*C Figure 2. — Vertical distribution of common zooplankton species in Great Central Lake mean no./m^ for July 1970. (Horizontal lines indicates the top and bottom of the thermocline, McAllister (personal communication). Inset shows the vertical distribution at Ih'/V quartile intervals over a 24-hr period. Note: the scale indicating the quantity of organisms varies for each species.) tween 3 m and 5 m to a minimum of approxi- mately SOO/m'' at depths below 40 m. The max- imum concentration of individual species ranged from 180/m^ for Daphnia to greater than 3,000/m^ for Holopedium. The tendency for some species to show an increase in abundance in deep samples was likely due to contamination from shallower depths since in the process of setting and hauling with nonclosing nets the deeper nets actually sample for a slightly longer time than the shallower nets. Variations in abun- dance with respect to time of sampling was noted for all species (Figure 2 inset) . The effect was generally most pronounced just after sunset when the maximum concentration per m^ of a species might be increased by 30' Rotifers, which were presumably the least motile of the zooplankton, exhibited the largest shift in abundance towards the surface with the onset of darkness. Some species, notably Holopedium and Eplschura, returned to their daylight depth of maximum abundance within 2.5 hr after sun- set. Other species, such as rotifers, Bosm'ma, and Daphhia exhibited relatively little movement during darkness. It is apparent from Figure 2 that the shift in species abundance were all with- in the daylight range occupied by the bulk of the respective populations. Furthermore, more than 75 Sr of the zooplankton populations were at all times within the euphotic zone (i.e. surface to 30-40 m). SEASONAL ABUNDANCE In Figure 3 the mean monthly numbers of zooplankton are shown for the 50-m vertical haul samples. Three species, Cyclops, Bosmina, and KelUcottia, were relatively abundant through- out the year, whereas the other species were present in numbers which exceeded 1,000/m' for periods of 4 to 5 months. Co7iochilus were the only species present during 1970 to appear subsequent to the initiation of nutrient addition. (They were present in 1969 samples.) Cyclops ranged from a winter minimum of 2,000/m' to a maximum in September and Oc- tober of 30,000/m-. Eplschura were never nu- merically dominant but ranged in numbers from 2,000 to 4,000/m- from May through September. Counts of Diaptomus did not exceed l.OOO/m^ until August, but by September there was a 29 FISHERY BULLETIN: VOL. 70. NO. I No/m ahousoKlsl 0 10 20 30 40 SO La_l I I 1 I 'mMui&ii^^ J4N FEB MSH iPR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR Figure 3.-Monthly mean zooplankton abundance (no./m2 in the upper 50 m data from Stations 1 and 2 combined). 7-fold increase which in turn doubled to ca. 26,000/m2 by October. The November-December catches of Diaptomus exceeded lO.OOO/m^, which was approximately two orders of magnitude greater than their standing stock 12 months earlier. Substantial numbers of Diaptomus, 3,000 to 4,000/m-, were carried through into 1971. Bosmina were the most abundant species collected throughout the year. Their numbers ranged from ca. 8,000/m2 in January to ca. 60,000/m- in June and again in October. The December concentrations of Bosmina were twice that of the preceding January. However, by January of 1971 Bosmina had virtually disap- peared from the water column, 0 to 50 m. Holo- pedium attained their maximum abundance in July, approximately 3 months after they began appearing in the samples in significant quan- tities, i.e. greater than l.OOO/m-. Following a secondary maximum in October, Holopedium were virtually absent from samples collected from December through March. Daphnia were the least numerous of the zooplankton spe- cies routinely sampled. They occurred in num- bers of 1,000 to 3,000/m2 from June through September. Kellicottia exceeded 10,000/m2 from May through August and again in October and November. Nearly twice as many Kellicot- tia were present in December as were present at the beginning of 1970. The maximum abun- dance of Conochilus colonies (2,000/m") was during July; no colonies were found prior to June and by December the number of colonies had declined to approximately 500/m-. In toto there were two to three times more zooplankton present in December of 1970 than there were the preceding January. It is of in- terest to note that the greater abundance of zoo- plankton at the end of 1970 was not maintained through the first 3 months of 1971 and further, that Bosmina had been apparently supplanted by Diaptomus in 1971. On a monthly basis there were fewer than 22,000 organisms/m^ , in January while in October, where the max- imum concentration was observed, there were nearly 10 times as many organisms present. Zooplankton counts exceeded 100,000/m- in June, July, September, and October. The de- crease in zooplankton abundance in August was approximately 15 Sr lower than that in either July or September; this decline was attributable mainly to fewer numbers of Bosmina. Individual species counts (4-day running mean number/m-) in vertical hauls have been pre- sented in Figure 4 in order to show the seasonal variations in abundance in greater detail than is shown in Figure 3. The general features of both figures are the same but in Figure 4 the rapid increase and decrease in numbers of some species are shown more clearly, e.g. Holopedium and Epischura. From Figure 4 it is possible to infer some relationship between the addition of nutrients and the appearance of Conochilus or the sustained increase in the abundance of Diap- tomus. It is noteworthy that all species, with the exception of Epischura and possibly Daphnia, went through a secondary maximum in October which was nearly as great as or greater than their level of abundance earlier in the summer. SEX RATIO Adult stages of Cyclops and Epischura showed marked imbalances from an expected 50: 50 ratio of females to males through the year (Table 4). Males of these two species were clearly pre- dominant during the late winter and early spring months. Cyclops females were predominant among the adults taken in June through August whereas Epischura females were never numer- ically dominant for more than two or three sam- pling periods, i.e. July 21 to 31, August 28 to 30 LeBRASSEUR and KENNEDY: LAKE FERTILIZATION. II. H 1 1- -I 1 \ 1 r- r'^^ -4 1 h H ♦- ■MT JUNC JULY AUft SEPT OCT OCC HAT JUNC SePT OCT MOV 0€e •**» JUNE JUL* »U0 SEPT OCT NOV Qt? AUfi «PT OCT MOV OCC Figure 4.— Species counts (no./m^ in 50-m vertical hauls at Stations 1 and 2. (Points represent a 4-day running mean, solid triangle indicates monthly mean, circle with dot indicates more than one sample with the same count. Note: the numbers of organisms are shown on a logarithmic scale.) Table 4. — Copepod sex ratios. Species Month Cycl ops Diapt 7mus Episch F/M ura F/M F2/F1 F/M F2/F1 F2/F1 Jan. -April 0.6 1.2 0.5 May 0.6 2.8 1.5 0.7 0.7 1.3 June 2.6 1.1 1.1 1.9 0.7 1.4 July 1.9 I.O 1.5 1.1 1.0 1.1 Aug. 1.9 1.6 1,3 1.0 1.1 0.8 Sept. 0.8 2.2 1.2 0.9 1.4 0.9 Oct. -Dec. 1.0 2.0 1.3 0.9 1.0 0.5 F/M Number of adult females/number adult males. Fi'/Fi Number of adult females at Station 2/number of adult females at Station I. September 14. In contrast there was a tendency for Diaptomus females to be slightly more abun- dant than males throughout the year. The only period for which Diaptomus males were con- sistently more numerous than females was from June 22 to July 10. Included in Table 4 is the ratio of the number of female copepods at Sta- tion 2 to the corresponding number at Station 1. Cyclops was the only species in which the fe- males were as numerous or more numerous at Station 2 than at Station 1. EGG PRODUCTION Counts were made of all readily identifiable eggs; these consisted of eggs in the brood pouch of cladocera and the egg sacks of Cyclops and Diaptomus. Rotifer species were not examined for eggs, while Epischura eggs were positively identified on only one occasion from a horizontal tow made at 1 m depth in August. It is possible that Epischura eggs develop close to the surface at depths of less than 1 m since they were not found at other standard depths sampled between 1 m and 65 m. Also other data, not presented here, indicate that the smaller size groups of Epischura were found closer to the surface than the adult stages. The data presented in Figure 5 show the ratio of eggs per female for vertical samples collected at Station 1 and Station 2. It was noted in Table 4 that maximum numbers of copepod females oc- curred during the summer, June through Sep- tember; Cyclops females were more numerous at Station 2 than Station 1 and Diaptomus 31 FISHERY BULLETIN: VOL. 70, NO. 1 o E a; UJ O 0} o E JAN ' FEB ' MAR' APR ' MAY ' JUN JUL ' AUG' ' SEP 'OCT ' NOV ' DEC 32 LeBRASSEUR and KENNEDY: LAKE FERTILIZATION. II. females were in about equal numbers at both stations. In Figure 5 the eggs per Cyclops were about equally numerous at both stations. There were three and possibly four periods of maximum egg production for Cyclops females, i.e. June, mid-July through to the third week in August, and the last week of September through to the first week of November; the lat- ter time interval could possibly be interpreted as consisting of two separate periods of egg pro- duction (late September and late October). Diaptomus females at Station 1 had two major periods of egg production, mid-June through mid-July and mid-August through the first week of September, with a period of relatively low egg production from mid-July to the end of August. At Station 2 there was no clear cessation of Diaptomiis egg production from mid-June through to the first week of September. For a major part of this period there were more than 10 eggs per female being produced. There was also a brief period of Diaptomus egg production in mid-May. The production of Bosmlna eggs ranged be- tween 0 and 0.5 per individual. From May through mid-August more eggs were produced at Station 2 than at Station 1 and thereafter the egg production was nearly equal at both sta- tions. The summer minimum which occurred in the first 2 weeks of August was followed by a rise in the number of eggs in the first week of September continuing until the end of the third week of September. The summer max- imum of adult Bosmina shown in Figure 4 oc- curred approximately 1 week after that of the eggs while the maximum standing stock of Bos- mina (which occurred in mid-October) was pre- ceded by the production of eggs 3 to 5 weeks earlier. Holopedium exhibited two clearly de- fined peaks in the production of eggs, from the first to the third week of June and again from the first to the third week of September. The corresponding maximum in the standing stock of Holopedium shown in Figure 4 occurred Figure 5. — Ratio of the number of eggs to the number of adult females. The data from 50-m vertical samples at Stations 1 and 2 have been averaged to give a 4-day running mean ratio. Note: scale changes for different species. from the second week of July through to August 25 and from September 27 to about October 20; the summer minimum occurred between the two peaks. The production of Daphnia eggs took place from June to mid-September with a second brief rise in egg production during mid-October at Station 2. The numbers of eggs produced per female at Station 2 by all species of clado- cera was generally greater or equal to that at Station 1. It should be noted that the latter was found for both the prefertilization period in May as well as during the period of nutrient additions. ZOOPLANKTON BIOMASS The wet weights for 1970 50-m vertical hauls at Stations 1 and 2 were combined and expressed as a monthly mean wet weight (g/m-) together with the range about the mean weight (Figure 6) . Included in Figure 6 (below) are individual weights for the 1969 sampling. The maximum wet weight in 1969 never exceeded 1 g/m- where- as in 1970 the weights ranged as high as 15 g/m^. The average wet weight of zooplankton during the period May through October was approxi- mately 0.5 g during 1969; for the same period in 1970 the average weight was 10 times larger, i.e. 5.3 g. The sample weights increased at a rate of 3'^r per day May through July to a max- imum average wet weight of 8.6 g/m-; there- JON FEB M4HCH IPBIL MAY JUNE JULY 4UG SEPT OCT NOV DEC Figure 6. — Zooplankton wet weight (g/m^) for 50-m vertical hauls. In the lower part of the figure, the points marked "x" indicate individual weights (g/m2) for 1969 samples. 33 FISHERY BULLETIN; VOL. 70, NO. 1 after the weights declined at an average rate of 1.5^'c per day to the December biomass of 0.9 g/m-. The actual rate at which the mean weight of zooplankton declined in any one month fol- lowing July was greater than the average com- puted above due to the increase in biomass in October. Inspection of Figures 3 and 4 indicates that the decline in biomass seen in August was a result of fewer numbers of Epischura, Bosmi- na, and Holopedium. Epischura never reached its earlier level of abundance after August and was virtually absent from the samples by Octo- ber whereas most other species, Daphnia excep- ted, showed an increase in abundance in October which gave rise to the October increase in bio- mass. Dry weights of Great Central Lake zooplank- ton (determined by the freeze-dry method) ranged from 14 '/f to 26 % of the wet weight M'ith a mean of 19 '^r . The variation in the percentage dry weight was directly attributable to the spe- cies composition of a sample. For example, the dry weight of Holopedium was 14 '"r of their wet weight, whereas the dry weight of Cyclops was approximately 26 ^V of the wet weight. The av- erage dry weights of the summer zooplankton (May through October) integrated over a 25-m column, i.e. the depth range in which most zoo- plankton were concentrated, for 1969 and 1970 was 4 mg and 40 mg/m^ respectively (from Fig- ures 2 and 6). Table 5. Species -Length- weight measurements of adult crustaceans. Mean length (Microns) Wet weight (Micrograms) Cyclops Diaptomus Epischura Bosmina Holopedium Daphnia 960 6 1,100 11 1,500 66 300 4 900 \7 900 10 Length-wet weight determinations were made for different sizes and stages of the common crustacean species and the data are summarized in Table 5. The data in Table 6 were obtained by multiplying the maximum concentration (no./m^) of a species within particular depth intervals (Figure 2) by their respective weight from Table 5, thereby providing a measure of biomass with depth. Included in Table 6 are the mean July temperatures within the respective depth intervals. Nearly 60 ^r of the total bio- mass occurs in the upper 10 m where the mean temperature was about 18°C. In the thermo- cline, from 10 to 20 m, with a temperature range from 12° to 6°C (mean temperature, 9°C) the biomass was about 50 mg/m^ or approxi- mately 30 Sr of the total. From 20 to 30 m depth the biomass was about 89r of the total. The re- maining 3 to A*^ of the total biomass occurred below 30 m (30 to 60 m). While these data were derived from July sampling it should be noted that the general distribution of the biomass with depth was similar throughout the period of thermal stratification, i.e. June to October. DISCUSSION The zooplankton standing stock in 1970 shows a phenomenal increase over 1969. This can be largely attributed to the affect of the nutrient additions upon the rate of primary production. The results of Parsons et al. (1972) demonstrate a marked increase in the rate of primary pro- duction within the upper 5 m; at the same time there was little or no change in the standing stock of primary producers. While experiments and observations of a direct relationship between particular species of primary and secondary pro- ducers have not been attempted, the obvious in- ference is that the zooplankton through increas- Table 6.- — Relative biomass of July crustacean zooplankton in various depth interval 3 (from Figure 2). Depth range (m) mT °C Mean maximum b iomass (mg/m' ) Cyclops Diaptomus Epischura Bosmina Holopedium Daphnia Total 0-10 10-20 20-30 >30 18 9 6 <5 1.2 3.1 .6 1.3 .3 .3 39.6 19.8 4.9 2.3 7.2 2.8 1.2 .7 51.0 28.9 3.4 1.7 .1 .5 1.8 .2 99.2 53.5 14.7 5.6 34 LeBRASSEUR and KENNEDY: LAKE FERTILIZATION. II. ing stock size were able to utilize the higher rates of primary production. It is also apparent that the higher biomass in 1970 cannot be entirely attributed to fertiliza- tion since the biomass in May (prefertilization) was also higher than any of the 1969 values. However, the nearly continuous production of eggs by most species and the maintenance of an increased standing stock over a 6-month period are indicative of a direct relation between zoo- plankton and nutrients. It is also noteworthy that there was no change in species diversity. The techniques employed for wet weight de- terminations in this study have produced weights which are apparently lighter than would be ob- tained by other investigators. Wet to dry ratios in the literature suggest that the dry weight is 5% to 10% of the wet weight. Schindler and Noven (1971) employed a ratio of 6%, although their reason for using this particular value is not given; the present results indicate that the dry weight is 19% of the wet weight. Conse- quently, the present weights could be increased approximately three times for comparison with other studies. Thus in the lakes which range from oligotrophic to eutrophic, listed by the above authors. Great Central Lake has, in terms of its mean summer zooplankton biomass, changed from oligotrophic to oligotrophic-meso- trophic, i.e. 12 mg in 1969 to 120 mg dry weight/m^ in 1970. In lakes producing sockeye salmon the mean abundance of zooplankton ranges from values which are less than 5 mg dry weight/m^ to greater than 1 g dry weight/m^ (Johnson, 1965). The mean concentrations in Great Central Lake have increased from the very low end of the range to values which are com- monly reported for some of the larger sockeye producing lakes, e.g., Babine Lake. Johnson (1965) concluded that there was a general relationship between the rate of growth of underyearling sockeye and zooplankton abundance. However, he also suggested that with increasing fish density food abundance was supplanted by a space effect as a limiting factor. In Great Central Lake the underyearling sock- eye in October of 1970 were ca. 30% heavier than fish caught in October of 1969 (Parsons et al., in press; Barraclough and Robinson, 1972). In addition to the increase in weight these authors report (on the basis of the number of adult salm- on spawning) that the number of sockeye fry in the lake were from two to five times more numerous than in the previous year. Assuming an initial weight of 120 mg for individual fry of each year the respective rate of growth over their first 200 days of lake residence was 0.9% and 1.2% per day for 1969 and 1970 respec- tively. The increased growth rate of sockeye in 1970 is less than might be anticipated from the 10-fold increase in zooplankton abundance. Johnson's data (1965) indicated that a pop- ulation density of 1 fish per m- might be the point at which space becomes a factor limiting growth. The maximum estimate of 1 x 10'' sockeye in Great Central Lake during 1970 is approximately 1 fish in every 5 m^. Conse- quently it appears unlikely that the density of the fish population in Great Central Lake limited their growth. Among other factors which limit growth of sockeye, Foerster (1968, Figure 45) indicates that temperature has a major affect upon growth and the efficiency with which food is utilized. The optimum temperature for food conversion for sockeye lies between 10° and 15°C. At higher or lower temperatures the efficiency of food conversion decreases, especially at temper- atures in excess of 20°C or less than 6°C. The laboratory studies of Brett et al. (1969) with fingerling sockeye support the findings reported above. In their experiments 15°C was found to be the optimum temperature for growth at high rations; however, maximum efficiencies with which a ration was utilized occurred at lower temperatures, e.g. the maximum food conversion efliciency of 40% with a 0.2% increase in fish weight per day occurred at a temperature range of 8° to 10°C and a ration of 1.5% of the fish weight/day. Temperatures between 5° and 17°C were found to provide the laboratory fish the optimum conditions for conversion efficien- cies and growth. In Great Central Lake, during their first 200 days of lake residence, the under- yearling sockeye concentrate at depths of 50 m or greater during daylight; with the approach of sunset the fish move to shallower depths and by nightfall the major portion of the population 35 FISHERY BULLETIN: VOL. 70, NO. 1 is at depths between 10 and 20 m while some fraction of the population occur in the upper 10 m. This pattern of vertical migrations ap- pears to be repeated daily (Barraclough and Robinson, 1972). Narver (1970) has reported similar vertical movements for sockeye popula- tions in Babine Lake. For the greater part of the day the salmon in Great Central Lake are at temperatures of 4° to 5°C, a somewhat shorter period (ca. 6 hr) is spent at temperatures of 6° to 12°C (10 to 20 m depth) while a rel- atively brief period (ca. 1 hr) may be spent at temperatures ranging from 14° to 23 °C (0- 10 m depth) . Details of the time actually spent at different depths by the sockeye are reported by Barraclough and Robinson (1972). It is apparent that the fish are utilizing the maximum concentrations of prey which occur at above op- timum temperatures in the upper 10 m for very short intervals. Consequently in assessing the relationship between the increased abundance of prey brought about through fertilization and the sockeye it should be noted that possibly 60 Sr of the total biomass, i.e. the portion in the upper 10 m, may be only partially available to the fish (Table 6) . Furthermore, some prey species be- cause of their size (rotifers) or structure (Holo- pedium) may not be a particularly useful food source for the salmon. Holopedium, for ex- ample, was among the largest and most numer- ous species of crustaceans in the lake; however, a large fraction of their biomass is comprised of a gelatinous material of dubious food value. The difference in the wet to dry weight ratio between Holopedium and other zooplankton (14% to ca. 26^/ -respectively) attests to the water composition of Holopedium. The quality of prey together with the observations of Foers- ter (1968) and Brett et al. (1969) empha- size the need for caution in interpreting preda- tor-prey relations. In the present instance, the benefits of the fertilization appear to have been only partially transferred to the sockeye salmon. Since the thermal structure of the lake is a factor beyond immediate control, it would be in- teresting to consider possible benefits from the addition or deletion of some prey species and to attempt to shift the level of primary and sec- ondary production to depths and temperatures favoring sockeye salmon growth. LITERATURE CITED Barraclough, W. E., and D. G. Robinson. 1972. The fertilization of Great Central Lake. III. Effect on sockeye salmon. Fish. Bull., U.S. 70: 37-48. Brett, J. R., J. E. Shelbourn, and C. T. Shoop. 1969. Growth rate and body composition of finger- ling sockeye salmon, Oncorhynchiis nerka, in re- lation to temperature and ration size. J. Fish. Res. Board Can., 26: 2363-2394. Brocksen, R. W., G. E. Davis, and C. E. Warren. 1970. Analysis of trophic processes on the basis of density-dependent functions. In J. A. Steele (editor), Marine food chains, p. 468-498. Oliver and Boyd, Edinburgh. Foerster, R. E. 1968. The sockeye salmon, Ovcorhynchus nerka. Fish. Res. Board Can. Bull. 162, 422 p. Ivlev, V. S. 1961. Experimental ecology of feeding of fishes (Transl. by D. Scott). Yale Univ. Press, New Haven, 302 p. Johnson, W. E. 1965. On mechanisms of self-regulation of popula- tion abundance in Oncorhynchiis nerka. Mitt. Int. Ver. Limnol. 13: 66-87. Miller, D. 1961. A modification of the small Hardy Plankton Indicator for simultaneous high speed plankton hauls. Bull. Mar. Ecol. 5: 165-172, Narver, D. 1970. Diel vertical movements and feeding of underyearling sockeye salmon and the limnetic zooplankton in Babine Lake, British Columbia. J. Fish. Res. Board Can. 27: 281-316. Parsons, T. R., C. D. McAllister, R. J.LeBrasseur, AND W. E. Barraclough. In press. The use of nutrients in the enrichment of sockeye salmon nursery lakes — a preliminary report. FAO Technical Conference on Marine Pollution, Rome, 9-18 December 1970. Parsons, T. R., K. Stephens, and M. Takahashi. 1972. The fertilization of Great Central Lake. I. Effect on primary production. Fish. Bull. U.S. 70: 13-23. Richer, W. E. 1962. Comparison of ocean growth and mortality of sockeye salmon during their last two years. J. Fish. Res. Board Can. 19: 531-560. Schindler, D. W., and B. Noven. 1971. Vertical distribution and seasonal abundance of zooplankton in two shallow lakes of the exper- imental lakes area, northwestern Ontario. J. Fish. Res. Board Can. 28: 245-256. 36 THE FERTILIZATION OF GREAT CENTRAL LAKE III. EFFECT ON JUVENILE SOCKEYE SALMON W. E. Barraclough and D. Robinson^ ABSTRACT Nutrient levels and rates of primary production in nursery lakes are factors which may limit production of sockeye salmon. This paper describes the effect of artificial fertilization on feeding behavior and growth of juvenile sockeye salmon in Great Central Lake, Vancouver Island, British Columbia. Under- yearling sockeye salmon grew 30% larger in 1970 than in 1969 as a result of adding 100 tons of fertilizer to Great Central Lake. The growth pattern for the whole population was complex, however, and the increase in size of juvenile sockeye was not as much as had been expected from the increase in quantity of their food organisms. The fact that the sockeye did not appear to appreciably crop the high epilem- netic concentrations of zooplankton during July and August 1970 may have been partly due to avoid- ance of high temperatures by the fish. Decomposing carcasses of anadromous fish, such as the sockeye salmon {Oncorhynchiis nerka) , contribute to the fertilization of nursery lakes following spawning in the lake. In most instan- ces the extent of this fertilization is not known but the removal of maturing sockeye by a com- mercial fishery may deny lake waters of their essential nutrients and contribute to lowered productivity. Particular attention has been fo- cussed on the imbalance of phosphate in the na- tural fertilization of lakes from decomposing salmon carcasses (Krokhin, 1959) and the sug- gestion has been made (Krokhin, 1967) that a positive balance should be maintained by the ar- tificial replacement of the phosphate with inor- ganic fertilizers. Early studies carried out in a small unstrati- fied lake in Alaska (Nelson and Edmondson, 1955; Nelson, 1958) showed that the addition of phosphate and nitrate fertilizer resulted in in- creased length and weight of sockeye smolts leaving the lake. The potential role of a natural imbalance of phosphate in nursery lakes on sock- eye salmon is emphasized in the following quo- tation from Foerster (1968): One wonders whether sufficient significance has been given to this feature of the phosphate balance. With ^ Fisheries Research Board of Canada, Biological Station, Nanaimo, B.C., Canada. Manuscript accepted September 1971. FISHERY BULLETIN: VOL. 70, NO. 1, 1972. sockeye populations in all areas showing such evident declines, despite legislation on regulation and limita- tion of fishing, it might well be that some basic factor such as this may be having a much more limiting effect on productivity than seems apparent. In addi- tion to the smaller amounts of phosphorus introduced into a lake in the carcasses of fewer sockeye spawners, there may also be occurring a steady decline in the phosphate content of the runoff waters as the phos- phates of the soil and rock become leached out over the years. Future studies of the phosphate balance of sockeye-producing waters and the direction of its trend may prove most enlightening. Addition of suit- able fertilizers may be found advantageous. In recent years it has become more evident that suitable fertilizers should not only include phosphates but also other nutrients, including trace elements, in order to increase aquatic pro- ductivity (Goldman, 1960, 1964). The theory and application of adding natural fertilizers to aquatic environments has been practiced in fish farming for many centuries. Parsons et al. (in press; 1972) have presented data on various aspects of lake fertilization stud- ies carried out by others. In summary of these findings, there is much evidence to show that the larger the sockeye smolts at the time of seaward migration, the higher the percentage return from the sea (Burgner, 1962; Ricker, 1962). Since food supply is one of the important factors 37 FISHERY BULLETIN: VOL. 70, NO. I governing growth, the effect of increasing the food supply to underyearling sockeye salmon through artificial fertilization of Great Central Lake, B.C., is presented here. It has already been established (Parsons et al., in press; 1972) that the waters of Great Central Lake are relatively unproductive of sockeye salmon, the average size of yearling smolts at the time of seaward migra- tion being much smaller than in Babine Lake, B.C. (63 mm versus 79 mm) (McDonald, 1969). The average size of yearling smolts from 14 other lakes in Washington, British Columbia, Alaska, and Kamchatka is larger than the yearling smolts from Great Central Lake (Foerster, 1968). In the following account particular attention is given to changes in size of juvenile sockeye salm- on in Great Central Lake associated with changes in their food supply prior to and after the addi- tion of inorganic nutrients (see Parsons et al, 1972; LeBrasseur and Kennedy, 1972). STUDY AREA Great Central Lake (Figure 1) is located in central Vancouver Island, British Columbia. The lake is about 33 km long and varies between 1 and 2.5 km in width. The shoreline length is 72 km and the surface area is ca. 51 km^. Ele- vation of the lake surface is 83 m above sea level and the mean depth is 200 m, with a maximum depth of about 280 m. The outlet of the lake runs into the Stamp River. Most of the shore- line slopes very abruptly into deep water. This feature is an important factor in regulating hor- izontal distribution of juvenile salmon in the lake, by providing a maximum amount of the lake surface available to juvenile sockeye. LAKE SPAWNING A brief account of the spawning sites of the sockeye salmon is presented here because the location of the in-lake spawning grounds is an important factor in the emergence of the alevin and dispersal of the fry at the time of their initial intake of food. Little or no published informa- tion is available on the migration and spawning of adults in the lake. Mr. F. C. Boyd of the De- partment of the Environment has kindly granted permission to refer to his internal manuscript reports on the subject. Adult sockeye salmon bound for Great Central Lake first enter the Stamp River as early as the first week in June. This migration up the Stamp River continues through June, peaks in July, and in most years, ends in early August. The peak I2^°25 20' 49''25 49° i^ 05 1250OO' I25°25' GREAT CENTRAL LAKE FISHING STATIONS 20' 15' 10' 49°25 49°20' 05' I25''00 Figure 1. — Great Central Lake showing the six fishing stations and depth contours in meters. 38 BARRACLOUGH and ROBINSON: LAKE FERTILIZATION. III. of migration may not occur until the first week in September. It takes between 2 and 5 days for the sockeye to migrate up the Stamp River into Great Central Lake, depending on the water levels in the river. The fish remain in the lake but do not commence to spawn until the latter part of September. Great Central Lake is one of the few lakes in British Columbia where most of the spawning occurs in areas along the lake- shore rather than in tributary streams. Only a few hundred sockeye spawn in tributary streams a short distance away from the lake. Drinkwater, Lindsay, and Fawn Creeks (Fig- ure 1) receive most of the stream spawners. Lakeshore spawning commences in the last week of September, reaches a series of peaks in three principal locations during October, and ends in November. About 50% of the spawning occurs along 4.63 km of lakeshore between Lind- say Creek and Forestry Creek, 30 '^r along 1.1 km of shoreline west of Fawn Point, and 20 '^'r along 1.6 to 4.8 km of lakeshore off" North Creek. Redds were found at depths between 0.6 and 24 m but most were between 12 to 15 m. Spawnings were observed by scuba divers at depths as great as 41 m. It is now realized that the location of separate major in-lake spawning areas is im- portant in providing the potential basis for the immediate and rapid distribution of juvenile sockeye throughout the lake, shortly after the fry emerge from the gravel and commence to feed. Two spawning areas are adjacent to the lake area where fertilizer was applied (see Par- sons et al., 1972). METHODS LOCATION AND DISTRIBUTION OF JUVENILE SOCKEYE SALMON A high frequency (200 kHz) moist paper re- cording echo sounder (Furuno model No. FNV- 3000)' was used to locate the young sockeye in ^ Reference to trade names in the publication does not imply endorsement of commercial products by the National Marine Fisheries Service. the lake and monitor their horizontal and vertical distribution. During the day, young sockeye are generally distributed throughout the lake at depths between 45 and 90 m, but are most abund- ant at about 65 m. They commence to migrate toward the surface about half an hour before sunset. In the summer months at civil twilight, when the sun is 96° from the zenith (or 6° be- low the horizon) they are distributed irregularly in density between 5 and 30 m. At nautical twi- light when the sun is 102° from the zenith (or 12° below the horizon) the juvenile sockeye form a layer between the depths of 10 and 20 m, with a maximum density of about 14 m. At night during the winter months they are distributed more uniformly between 20 and 60 m. In sum- mer the downward migration commences shortly before sunrise and is usually complete 15 to 30 min after sunrise. The young sockeye were sampled with mid- water trawls. Sampling commenced at night when the fish were in a layer between 10 and 20 m. Samples were also collected during day- light at diff"erent depths throughout the depth range of the young sockeye. The depth of trawling was adjusted to coincide with the depth of maximum fish concentration as shown by the echo sounder traces. FISHING GEAR A trawl net with a mouth opening 3 m wide, 6.1 m deep, and 17.7 m long was towed at 2.7 to 3.2 km/hr by a single vessel, the Decihar, to sample the sockeye between the depths 5 and 25 m. Three mesh sizes of knotless nylon netting were used in the construction of the net: 5 cm and 2.5 cm stretched mesh in the body and 1.3 cm in the cod end. The cod end measured 1.2 m wide by 1.8 m deep at the mouth and it tapered to a blunt end about 76 cm in diameter. A standard Henson plankton net (350/x mesh) 76 cm in diameter at the mouth, was secured to the blunt aft end of the cod end to retain the smallest juvenile sockeye and minimize the loss 39 FISHERY BULLETIN: VOL. 70, NO. 1 of their minute scales by abrasion against net- ting in the main cod end of the trawl. An Isaacs-Kidd midwater trawl was towed from the Decibar to samjile the juvenile sockeye at depths greater than 25 m and to evaluate the fishing capabilities of the large midwater trawl towed at the same depths. The mouth opening of this trawl was 1.9 m^ and the net was con- structed of 6.3 mm stretched mesh knotless netting. FISHING STATIONS Juvenile sockeye were sampled with trawls taken at intervals of about 3 weeks at 6 different stations (Figure 1). Most of the tows were of 30 min duration but some tows wei-e shorter, when the echo sounder traces indicated that young sockeye were especially abundant between 12 and 14 m at night. ANALYSES OF SAMPLES The length of all fish was measured to the near- est millimeter from the snout to the end of the central rays of the caudal fin. This measurement is referred to as the fork length. Lengths of smaller fish were measured in a graduate tray under a binocular microscope; calipers were used for larger individuals. All fish were weighed by fork length groups using a center-loading milligram balance (KERN Model No. T1226-1) . Weights recorded are from "blot-dried" specimens. Moisture was blotted from the exterior of the fish, and gentle pressure was applied to the buccal cavity and branchial chamber to remove moisture from these spaces. Age was determined from scales using y 254 projections of thermoplastic impres- sions. Stomach analyses for food were done on fish selected to represent proportionally as complete a size range as possible. The food weight was measured by subtracting weight of stomach shell from weight of stomach plus food. The number of all species of food organisms were counted according to size and state of condition of each stomach examined. RESULTS FOOD OF UNDERYEARLING (AGE 0) SOCKEYE During the latter part of March and up to mid-April, 1970, pre-mature fry" (24 to 28 mm fork length) with a small portion of the yolk sac remaining were caught at night at a depth of 14 m in midlake positions off the 3 major spawning areas. A few fry (28-30 mm) with empty stomachs were caught during the day at depths between 35 and 100 m in late March, and the first actively feeding fry (28 to 33 mm) were caught at depths between 12 and 55 m at night during the latter part of April. The number of fry caught at midwater depths increased in May at Stations 3, 4, 5, and 6 and reached a maximum in June at all stations. Fry continued to be caught in July and were still being caught in trawl nets at night in late August and early September. The fry and larger underyearling sockeye ate the same food organisms throughout the year, but the larger juvenile fish had more food in their stomachs. Figure 2 shows the number and weight of all species of food organisms per underyearling sockeye (Age 0) from August, 1969, when in- lake sampling began to April, 1970, when about 85 '^r of the fish migrating were yearling smolts.' The percentage of the total number of the six major food categories from all the fish sampled for stomach contents through the same period is shown in Figure 3. A list of the different gen- era of food organisms found in the stomachs of juvenile sockeye from 1969 to 1971 is given in Table 1; the smallest is listed at the top of the column and the largest at the bottom. Epischura was the predominant form (60%) in the stomachs in August, 1969 but was almost replaced by /fo/o?9edmm (60-80%) from Septem- ber to December (Figure 3). The incidence of ^ "Embryo" is defined as a larva minus its yolk-sac. An "alevin" is a larva of an age following hatching but prior to yolk absorption. Following this stage the fish becomes a "fry" (cf. Bams, 1969). ' In 1969 ca. 86% of migrant smolts were yearling, 10% were 2 year old, and 4% were 3 year old. In 1970 ca. 85% of smolts were yearling and 15% were 2 year old. 40 BARRACLOUGH and ROBINSON: LAKE FERTILIZATION. III. 500,- -i50 AUG SEPT ■ OCT 1969 NOV DEC JAN FEB MAR 1970 APR Figure 2. — Average number and weight of all food or- ganisms (all species combined) per fish for underyearling sockeye salmon in Great Central Lake from August, 1969 to April, 1970. lOO AUG I SEP I OCT I NOV DEC JAN FEB 'MAR APR 1969 1970 Figure 3. — Food of underyearling sockeye salmon ex- pressed as a percentage of the total number of organisms from August, 1969 to April, 1970. Table 1. — List of organisms found in juvenile sockeye stomachs. Size range Organism mm 0.3-0.6 Bosmina, usually B. coregoni 0.6-1.1 Cyclops, usually C. biscuipidatus thomasi and C. vernalis 0.9-1.2 Holopedium gibberum 0.9-1.5 Daphnia, usually D. longiremis 0.8-1.3 Diaptomus, usually D. oregonensis 2.0-2.5 D. kenai 1.1-1.9 Epischura nevaiemii 3 Insects of the order Diptera (other than Chironomidae) 3-5 Insects of the family Chironomidae— larvae 8-11 Insects of the family Chironomidae— larvae 6-11 Larvae of the sculpin, Cottus asper Bosmina and Cyclops increased gradually from less than 5 Sf in August to a peak of 30 to 50 % in January-February, 1970. Chironomid larvae were the only organisms eaten from February to early March and in turn were replaced by Bos- mvna (SO^r) in late March and April. The per- centages of Epischura and Holopedium in the stomachs by number (Figure 3) and by weight (Figure 4) were similar from August to De- cember. There was a pronounced difference between the percentages by number and by weight of Cyclops and Bosmina per fish. Al- though the percentage by numbers of both organisms per fish increased markedly between December, 1969 and February, 1970, the per- centage weight per fish remained less than 7% for Bosmina and never exceeded 20% for Cy- clops. The importance of the chironomid larvae in the diet of underyearling sockeye from Feb- ruary through March to early April, 1970 is more indicative when expressed as a percentage by weight (Figure 4) than by number of or- ganisms (Figure 3). AUG I SEP ' OCT I NOV ' DEC ' JAN ' FEB ' MAR 1969 197 0 Figure 4. — Food of underyearling sockeye salmon ex- pressed as a percentage of the total weight of organisms from August, 1969 to April, 1970. The fry which emerged in late March and April, 1970 commenced to feed actively by mid- April and, as juvenile sockeye, they continued to increase their intake in number and weight of all food organisms throughout the summer, 41 FISHERY BULLETIN: VOL. 70, NO. I reaching a peak in September-October (Figure 5). In August, 1970, 3 months after lake fertilization began, the underyearling sockeye had twice the number of organisms per stomach as in August, 1969 and contained about 60% more food by weight. The high consumption in September-October, 1970 represents an increase of about 45^ r in number of organisms, and 40% by weight, compared to the stomach contents per fish in the same period in 1969. A slight decline in number and a significant decline in weight of food organisms per fish was shown from No- vember, 1970 to February, 1971; an abrupt in- crease occurred to a second high in March, which represented an increase manyfold over March, 1970. This increased food consumption occurred I to 2 months prior to their emigration from the lake as yearling smolts. Five species of food organisms contributed chiefly to the diet of underyearling sockeye in Great Central Lake in 1970. In April and May Bosviina contributed about 50 '^r of both the total number (Figure 6) and total weight (Figure 7) of all organisms found in their stomachs. The numbers of Bosmhia consumed were insignificant throughout the rest of 1970 and the first 3 months of 1971. Epischura was the most im- portant food organism from May to July (Fig- ures 6 and 7) and was probably the principal source of energy for the rajiid growth of the underyearlings during this period (see Figures II and 12). There was a transition in late July and August when Cyclops, Holopedium, and Daphnia gradually became more abundant in the stomach samples. In the 3 months which followed, September to November, Cyclops and Holopedium were the predominant genera. Cyclops continued in importance and formed about 50% of the number of food organisms to the end of January, 1971. However, during this period of 6 months Cyclops formed only 15 to 30% of the food by weight whereas Holopedium constituted 30 to 80% by weight. Diaptomus was first observed in the stomachs in the latter part of October, increased markedly in Decem- ber and January, and was the predominant food organism by number in February and March, 1971. Thus Diaptomus was the most numerous food organism in the stomach samples just before SOOr 200 I APR I MAY I JUN I JUL I AUG I SEP IQCT I NOV I DEC I JAN 1 FEB I MAR I 1970 19 7 1 Figure 5. — Number and weight of all food organisms per fish, for underyearling sockeye salmon from April, 1970 to March, 1971. Figure 6. — Food of underyearling sockeye salmon ex- pressed as a percentage of the total number of organisms from April, 1970 to March, 1971. 'APR 'MAY 'JUN ' JUL ' AUG ' SEP ' OCT ' NOV 'DEC ' JAN ' FEB ' mar' 1970 1971 Figure 7. — Food of underyearling sockeye salmon ex- pressed as a percentage of the total weight, from April, 1969 to March, 1971. 42 BARRACLOUGH and ROBINSON: LAKE FERTILIZATION. III. smelt migration in 1971. Bosmina was the most numerous in the previous year. The few chiro- nomid larvae (Figure 6) in the diet of juvenile soekeye from December, 1970 to March, 1971, formed 30 to 70 7^- by weight of all the food or- ganisms (Figure 7). The importance of chiro- nomid larvae during the winter months was ob- served also in the previous winter (Figures 3 and 4). FOOD OF YEARLING (AGE 1) SOCKEYE Those underyearling soekeye in Great Central Lake in 1969 which did not migrate to sea as yearling smolts in 1970, but remained in the lake for a second year, attained a mean length of only 51 mm and weight of 1.1 g between the latter part of April and May, 1970, whereas the migrating smolts had a mean length of 70 mm and weighed 3.5 g. The yearling soekeye which remained in the lake were collected from samples taken at all six stations and not from the end of the lake where smolts were schooling and heading seaward. Reference will be made later to the fact that the smallest size smolt, caught in the Robertson Creek weir (Figure 1) or in the nets set to capture smolts in the Stamp River, measured 55 mm and weighed 1.5 g. Food organisms found in the stomachs of the yearling soekeye were similar to those eaten by the underyearlings during most of the year in 1970, but the yearling soekeye were more se- lective in cropping the larger forms of zooplank- ton (Figure 9). Both the underyearling and yearling soekeye fed heavily upon Epischura from May to July (Figures 6 and 9) , but it was evident from the large numbers and weight of food organisms per fish (Figure 8) that the year- ling soekeye elected to feed or were able to prey more heavily upon Epischura (Figure 9) than the underyearlings during September and Oc- tober. Few yearling soekeye were caught in the trawls during the winter months of 1970-1971 prior to their migration as 2-year-old smolts. Diaptomus, Holopedium, and Cyclops were in the stomachs of these fish. ° 500- 400 100 TBO -70 60 A-A, 50 30 ^ 20 * -10 °l APR : MAY I JUN I JUL UUG I SEP I OCT I NOV I DEC I JAN I FEB I MAR P 1970 1971 Figure 8. — Number and weight of food organisms of yearling soekeye salmon from April, 1970 to February, 1971. APR MAY JUN JUL I AUG I SEP I OCT I NOV I DEC I JAN I FEB 1970 1971 Figure 9. — Food of yearling soekeye salmon expressed as a percentage of the total number of organisms from April, 1970 to February, 1971. DIEL FEEDING OF JUVENILE SOCKEYE From midafternoon on June 17 to midday on June 18 a series of 11 tows, each of 15 min dur- ation, were made with an Isaacs-Kidd midwater trawl. The trawl was towed through the middle of the densest portion of the stock during their 43 FISHERY BULLETIN: VOL. 70. NO. I diel migration. The tows were made to deter- mine what portion of different food organisms contributed to the salmon's ration during the day and night, as well as during the period of their diel migration. Data on depth and time of each tow, the number, size range, and mean length of sockeye caught, together with those sampled for the number and species of food organisms, and the weight of the food as a per- centage of the body weight are given in Table 2. The degree of freshness of food organisms was arbitrarily determined as fresh, fragmented, or largely unidentifiable. Fresh food was des- ignated when no indication of digestion had oc- curred. The percentage of fresh zooplankton in the stomachs is given in Table 2. Depth of the densest portion of the layer of juvenile sockeye at different times of the day and night is indi- cated in Figure 10a by a broken line. The depth and time of each trawl tow relative to the depth of the fish is also indicated in Figure 10a. The 24-hr data collection shows that in the day the densest poi'tion of the layer of juvenile sockeye was formed at 75 m where the temper- ature of the water was 4° C; of the fish exam- ined for stomach contents (Table 2) from tow No. 1, few food organisms per fish (Figure 10b) were noted and only 5Vf of the species were in fresh condition (Bosmina) . The remaining spe- cies, Epischura, Cyclops, and Daphnia, were digested. Tow No. 2, through a less dense sec- ondary layer at 105 m, indicated the same feed- ing pattern. Young sockeye commenced to mi- grate upward from 75 m between 1700 and 1800 hr. No differentiation in migration between underyearling or yearling sockeye could be de- tected at any level in the layer, either by net sampling or from high frequency echograms. A tow just after sunset at a depth of 35 m revealed that the fish were eating Bosmina and Cyclops (Figure 10c) as they moved upward and 22% of the contents were in fresh condition. At 2200 hr the sockeye had passed 25 m where the heaviest concentration of Cyclops and Daphnia was located (LeBrasseur and Kennedy, 1972) ; in passing they had eaten Cyclops (Fig- urelOc) . It should be recognized, however, that there is a natural time lag between feeding at any depth and the time the fish was captured by the trawl at a shallower depth, as they migrated toward the surface. At nautical twilight, most of the fish had com- pleted their upward migration and were distrib- uted in a layer between 10 and 20 m where tem- peratures ranged from 6° to 12°C (Figure 10a), Echograms indicated many of the juvenile sock- eye salmon appeared to spend brief periods be- tween 0 and 10 m at temperatures ranging from 14° to 23° C, during which time the young fish fed heavily upon Epischura (Figure 10c). In the 4 hr between the beginning and end of nau- tical twilight no feeding occurred (Table 2). Table 2.— Die! feeding of juvenile sockeye; tows made with an Isaacs -Kidd midwater trawl, each of 15 min dur- ation, over a 24-hr period from June 17 to 18, 1970 at Station 4 in Great Central Lake. Tow No. Depth of tow Time (PST) at start of 15 min Number fish caught Size range underyearling Mean length Mean weight Size range sampled for food Mean length No. samp for food le Total no. food organisms No. of organisms per fish % fresh Weighl food as % body weight (m) mm mm mg mm mm 1 75 1449 47 23-40 32 292 28-40 33 10 190 19 5.5 1.1 2 105 1538 17 27-36 30 212 27-36 31 9 112 12 1.0 3 55 1820 21 27-38 32 272 27-38 33 10(1)1 67 7 1.0 4 35 2018 11 27-36 31 263 27-36 32 10 164 16 22 I.O 5 18 2142 6 26-36 31 282 26-36 31 6(2) 132 22 46 1.6 6 14 0023 71 26-41 32 315 26-41 33 11 537 49 6 1.7 7 19 0254 50 26-39 32 296 28-39 34 10(1) 257 26 1.5 8 62 0517 14 28-39 32 271 28-39 32 14(4) 586 41 64 2.1 9 68 0738 10 29-37 33 320 29-37 33 9 447 50 84 1.6 10 70 0945 9 29-39 33 278 29-39 33 9 543 60 34 1.6 n 75 1159 6 31-39 35 353 31-39 35 6 324 54 2 2.1 Time of sunset 2011 Time of sunrise 0350 Time of nautical twilight 2201 Time of nautica 1 sunrise 0200 1 Number in parentheses is number of items In sample which contained no food in stomachs. 44 BARRACLOUGH and ROBINSON: LAKE FERTILIZATION. III. 1400 1600 1800 2000 2200 2400 0200 0400 0600 OBOO lOOO '2°° _, (500 1700 SOO 2100 2300 OiOO 0300 0500 0700 0900 "OO 'JTO Figure 10. — (a) Depth of the densest portion of the layer of juvenile sockeye salmon at different times of the day and night from June 17 to June 18, 1970 is indicated by a broken line. A secondary layer is shown at 105 m for a 2-hr period. The depth of each tow with a midwater trawl is shown relative to the depth of the fish, (b) Number of all food organisms of underyearling sockeye. (c) Food species of underyearling sockeye salmon expressed as a percentage of the total number of organisms. A second feeding period was noted at the time of the diel migration downward. Stomach sam- ples from juvenile sockeye collected during this period contained many fresh Daphnia and Cy- clops in tows 8 to 10. Only 2^/r of the zoo- plankton in the stomachs of sockeye caught at midday were in a fresh condition (Table 2), which indicates a marked reduction in feeding activity. GROWTH OF UNDERYEARLING SOCKEYE The average size of underyearling sockeye (Age 0) in 1969 and 1970 in Great Central Lake is shown in Figures 11 and 12. A total of 1,760 underyearling sockeye were caught in 1969 and 20,783 fish in 1970 from all six stations. A com- plete record of all data on which this analysis is based has been reported by Barraclough and 70 60 E t 50 ,o 40 30 Or 1969 Underyeorlings a -. 1970 Underyeorlings ^^IwARlflPRlMflYljUNljUL I AUG I SEP'OCTInOV I DECl JAN I FEBI MAR Figure 11. — Average length of underyearling sockeye salmon in each month, 1969 and 1970. 2 50 - 2 00 E 1 50 100 050 o = 1969 Underyeorlings O = 1970 Underyeorlings D — a-o-16°C, and this reduced their feeding effi- ciency (Foerster, 1968) . Thus increases in tem- perature of the epilimnion through the long period of fry emergence decreased the apparent benefit to late-hatching fish. However, in spite of this, most of the fry hatching later in the year achieved a length greater than 55 mm and mi- grated from the lake in June 1971; under normal conditions it is believed that these fry would not have reached 55 mm and would have migrated the following year as 2-year-olds. Prelimi- nary examination of the scales from juvenile sockeye in 1971 reveals an absence of a winter check on many scales which suggests that the high concentrations of zooplankton persisting through the winter enabled many fish to smoltify and leave the lake. The combination of an early run of very large 1-year-old smolts, combined with this later run of much smaller smolts, tended to reduce the overall apparent effective- ness of lake fertilization. Thus the real effects of fertilization seem likely to be greater than would be judged from considering only changes in overall mean size of smolts. LITERATURE CITED Bams, R. A. 1969. Adaptations of sockeye salmon associated with incubation in stream gravels. In Symposium on salmon and trout in streams, p. 71-87. H. R. MacMillan Lectures in Fisheries, Univ. B. C, Inst. Fish., Vancouver, B.C. Brett, J. R., J. E. Shelbourn, and C. T. Shoop. 1969. Growth rate and body composition of finger- ling sockeye salmon, Oncorhynchus nerka, in re- lation to temperature and ration size. J. Fish. Res. Board Can. 26: 2363-2394. Burgner, R. L. 1962. Studies of red salmon smolts from the Wood River Lakes, Alaska. Univ. Wash. Publ. Fish., New Ser. 1: 247-314. Foerster, R. E. 1968. The sockeye salmon, Oncorhynchus nerka. Fish. Res. Board Can. Bull. 162, 422 p. Goldman, C. R. 1960. Primary productivity and limiting factors in three lakes of the Alaska Peninsula. Ecol. Monogr. 30: 207-230. 1964. Primary productivity and micro-nutrient lim- iting factors in some North American and New Zealand Lakes. Int. Ver. Theor. Angew. Limnol., Verhandl. 15:365-374. Johnson, W. E. ^ 1961. Aspects of the ecology of a pelagic, zooplank- ton-eating fish. Verh. int. Ver. Limnol. 14:727- 731. Krokhin, E. M. 1959. (On the effect of the number of spawned-out sockeye salmon (Oncorhynchus nerka) in a lake on its supply of biogenic elements.) Dokl. Akad. Nauk SSST 128(3) :626-627. (Fish. Res. Board Can. Transl. Ser. 417.) 1967. Influence on the intensity of passage of the sockeye salmon (Oncorhynchus nerka Walb.) on the phosphate content of spawning lakes. Izdatel'- stvo "Nauka" Leningrad 15:26-31. (Fish. Res. Board Can. Transl. Ser. 1273.) LeBrasseur, R. J., W. E. Barraclough, 0. D. Kennedy AND T. R. Parsons. 1969. Production studies in the Strait of Georgia. Part III. Observations on the food of larval and juvenile fish in the Eraser River plume, February to May, 1967. J. Exp. Mar. Biol. Ecol. 3:51-61. 47 FISHERY BULLETIN: VOL. 70. NO. 1 LeBrassei'r, R. J., AND 0. D. Kennedy. 1972. The fertilization of Great Central Lake. II. Zooplankton standing stock. Fish. Bull., U.S. 70: 25-36. McDonald, J. G. 1969. Distribution, groN^-th and survival of sockeye fry (OncorhyncliHs nerka) produced in natural and artificial stream environments. J. Fish. Res. Board Can. 26:229-267. Nelson, P. R. 1958. Relationship between rate of photosjTithesis and growth of juvenile red salmon. Science 128: 205-206. Nelson, P. R., and W. T. Edmondson. 1955. Limnological effects of fertilizing Bare Lake, Alaska. U.S. Fish Wildl. Serv., Fish. Bull. 56: 414-436. Parsons, T. R., C. D. McAllister, R. J. LeBrasseur, and W. E. Barraclough. In press. The use of nutrients in the enrichment of sockeye salmon nursery lakes — a preliminary report. FAO Technical Conference on Marine Pollution, Rome, December 9-18, 1970. Parsons, T. R., K. Stephens, and M. Takahashl 1972. The fertilization of Great Central Lake. I. Effect on primary production. Fish. Bull., U.S. 70:13-23. RiCKER, W. E. 1962. Comparison of ocean growth and mortality of sockeye salmon during their last two years. J. Fish. Res. Board Can. 19:531-560 48 ESCAPE BEHAVIOR OF THE HAWAIIAN SPINNER PORPOISE {Stenella cf. 5". longirostris) William F. Perrin and John R. Hunter^ ABSTRACT Incidental mortality of porpoise (Cetacea, Delphinidae) occurs in the tropical tuna seine fishery. Ex- periments were carried out in a crowding chamber to determine behavioral responses of trained and naive Hawaiian spinner porpoise (Stenella cf. S. longirostris) to barriers of purse-seine netting, monofilament webbing, pol>winyl sheeting, rows of floats, and openings of various dimensions in a net wall. The object of the experiments was to generate information to be used in development of rescue gear and methods for the fishery. Openings of less than 1.5 m in width and/or 1 m in depth markedly inhibited escape. Negative effect of a line of floats aci'oss an opening at the surface was pronounced. Barriers of visually and acoustically relatively transparent monofilament webbing and polyvinyl sheeting were not appar- ently detected by porpoise prior to physical contact. Recommendations pertaining to potential design of rescue gear are presented. Incidental mortality of porpoise occurs in the American purse-seine fishery for tropical tunas (Perrin, 1970). In 1970, the National Marine Fisheries Service began a program of research to develop improved gear and methods to reduce the porpoise mortality due to tuna seining. This paper reports the results of experiments on the responses to netting and other barriers by the Hawaiian spinner porpoise {Stenella cf. S. longi- rostris), a form closely related to one of the species involved in the tuna fishery." We studied the response of the spinner porpoise to barriers of net, transparent monofilament nylon webbing, transparent polyvinyl sheeting, rows of floats, ' National Marine P^isheries Service, Southwest Fish- eries Center, La Jolla, CA 92037. ^ Taxonomic note: The spinner porpoise of Hawaii has been variously referred to Stenella longirostris Grav 1828, by Nishiwaki (1967) and Tomich (1970), and to S. roseivcntris Wagner 1846, by Fraser (in Morris and Mowbray, 1966) and Rice and Scheff"er (1968). The spinner porpoise of the tuna grounds of the far eastern Pacific has been referred to .S'. microps Gray 1846 (Miller and Kellogg, 1955; Handlev, in Hester, Hunter, and Whitney, 1963; Nishiwaki, 'l967; Pilson and Waller, 1970) and to .S. longiros^tris (Rice and Scheffer, 1968; Harrison, Boice, and Brownell, 1969). No critical re- view of the genus has been accomplished since True's work on the Delphinidae in 1889. The usage here of S. longirostris for the Hawaiian spinner is provisional pending the I'esults of taxonomic studies underway at the Southwest Fisheries Center and elsewhere. and to openings of different dimensions in a net wall. The results of these studies will be ap- plied in the design of an escape opening in the tuna purse seine. The experiments were carried out at Oceanic Institute, Oahu, Hawaii, in May, June, and July 1970. METHODS AND MATERIALS THE ANIMALS Three of the five porpoise (Table 1) used in the experiments had been in captivity at Oceanic Institute and Sea Life Park for various lengths of time and are referred to below as the "trained porpoise"; the remaining two, referred to below Table 1. — Hawaiian spinner porpoise (Stenella cf. S. longirostris) used in behavioral experiments. Name Date of capture Sex Weight at time of capture ke Trained porpoise Waimea Mar. 6, 1969 Male 50.0 Nani Dec. 4, 1969 Female 59.2 Nohea Dec. 4, 1969 Male 65.9 Naive porpoise Westward June 11, 1970 Female 72.7 Moana July 9, 1970 Female 51.3 Manuscript accepted September 1971. FISHERY BULLETIN: VOL. 70, NO. I, 1972. 49 FISHERY BULLETIN: VOL. 70, NO. 1 as the "naive porpoise," were freshly captured, were not exposed to any training: procedures prior to the experiments, and were tested imme- diately upon arrival at the Institute. THE APPARATUS The crowdinjr chamber (Fig-ure 1) was con- structed in a large pool at the Oceanic Institute. The pool, known as "Bateson's Bay," is roughly circular, 24.7 m across at its greatest diameter, and approximately 1 m deep at its center. A smaller holding tank communicates with the jwol through a wooden gate. Three hemisjiher- ical underwater viewing ports allow surveillance of the entire pool. Net barriers were placed at various points along the pool wall to construct a circular en- closure or crowding chamber about 20 m in di- ameter in which j^orpoise were tested. The crowding chamber had two radial walls of net- ting that extended from the outer edge of the chamber to a central aluminum mast. One of the walls was stationary and was provided with escape openings of various dimensions. The other wall was movable and was used to drive the animals through the opening in the sta- tionary wall. The movable wall pivoted on the central mast and was supported along the leading edge by an aluminum beam and on the distal end by a plastic float. The edge of the pool was marked at 1° intervals. The walls were made of tuna purse-seine web- bing (4Vi.-iiich stretched mesh [10.8 cm] *42 thread knotted nylon). Flotation was provided by purse-seine-type corkline constructed of 6-inch diameter x 3V^-inch (15 >( 9 cm) sponge- plastic floats. The basic escape opening was 18 ft (5.5 m) wide and 6 ft (1.8 m) deep. Flaps of purse- seine webbing were laced in, to variously de- crease width to 10, 5, or 2V2 ft (approximately 3.0, 1.5, or 0.8 m) and/or depth to 31/2, 3, 2, 1, or 1/2 ft (approximately 1.1, 0.9, 0.6, or 0.2 m) . For tests of response to a barrier across the opening at the water surface, a corkline constructed of hollow plastic floats (5 y 9 inch [13 X 23 cm], 4 per m) was strung across the top of the open- ing. In tests of response to barriers of acous- tically low-reflective materials, a panel of 3%- inch (stretched) mesh (8.6 cm) *12 monofila- ment webbing, a panel of 0.38-mm-thick poly- vinyl sheeting, or a panel of 1.04-mm-thick poly- vinyl sheeting, was laced into the opening. STATIONARY CONTAINING NET WALL P STATIONARY CONTAINING NET WALL Figure 1. — Crowding chamber. Largest diameter of pool is 80 feet, to scale. Sketch not drawn 50 PERRIN and HUNTER: ESCAPE BEHAVIOR OF PORPOISE Acoustical tests carried out by the Naval Under- sea Research and Development Center, San Di- ego, Calif., on plastic sheeting of similar thick- nesses indicated effective acoustic transparency in the range of porpoise emanations (personal communication from W. E. Evans). PROCEDURES The moving wall was rotated around the cen- ter mast at a rate sufficient to completely close the chamber in about 4 min. An attempt was made to maintain a constant rate of rotation. Time required for an animal to escape from the chamber was recorded in seconds with a stop- watch, and position of the moving wall at time of escape was recorded in degrees. The reading in degrees was later used to calculate surface area remaining in the crowding chamber at time of escape. The wall was rotated alternately in clockwise and counterclockwise directions. After escape of an animal, the moving net wall was rotated until it was against the stationary wall, and the two radial nets remained together until the beginning of the next trial. Trails were spaced initially at 15-min intervals, to allow time for changing the escape opening. After our pro- ficiency in altering the opening increased, the in- terval was decreased to 10 min. Two major types of experimental design were used : ( 1 ) a long series of trials alternating two treatments and (2) a series of blocks of consec- utive trials of various treatments. In some ex- periments, the two approaches were combined to yield a factorial design testing simultaneously the effects of variation in two or three of the factors of width, depth, and presence or absence of corkline, monofilament, or polyvinyl barriers. In some tests of the monofilament and polyvinyl panels, the animal was subjected to a single trial with the panel after a series of learning trials without the jianel or at the beginning or conclu- sion of an experiment involving other variables. The design of these experiments is referred to below as "single trial." The results of the first series of experiments (Waimea I, II, and III; see Table 2) using the alternating trials design indicated a probable influence by the direction of rotation of the net wall or by stage of practice effect. The small number of trials in each ex- ]-»eriment precluded complete randomization, but the treatments in subsequent experiments were staggered to offset the effect of direction of ro- tation. A typical sequence of trials was: a, h, a, b, a, a, h, a, b, b; where a and b were different treatments, and rotation in the first trial was clockwise, in the second counterclockwise, and so on in alternating fashion. In this manner, an equal number of clockwise and counterclockwise trials was assured for each treatment. Table 2. — Preliminary experiments with trained porpoise. Porpoise Experiment Variables tested Design Number of trials Waimea 1 Width Alternating trials 20 II Width Alternating trials 20 III Depth, corkline Factorial 26 IV Monofilament panel Single trial 14 Nan! 1 Depth Alternating trials 20 11: trialsl-19 trial 20 Width, depth Corkline Factorial Single trial 19 1 III: trial 1 Corkline Single trial 1 trials 2-16 Depth Block 16 trials 17-30 Depth, monofilament panel Factorial 14 Nohea 1 II Depth Depth Alternating Block trials 22 36 III: trials 1-36 Width Block 36 trials 37-44 Depth Alternating trials 8 trial 45 Monofilament panel Single trial 1 IV Width, depth, corkline Factorial 32 V: trials 1-8 trial 9 Depth Thin polyvinyl panel Alternating Single trial trials 8 1 VI: trials 1-8 trial 9 Depth Thick polyvinyl panel Alternating Single trial trials 8 1 51 The remaining surface area between the ad- vancing net wall and the stationary wall at the time of escape was used as a criterion of the animals' readiness to esca])e. This index was the inverse of latency, since the smaller the area that remained when the animal escaped the longer would be the latency. Although we measured latency in seconds, we felt the net po- sition was the preferable measurement because the rate of net movement was imi)recise, where- as the actual stimulus for escaj^e, the reduction in the swimming area, could be measured rel- atively accurately. In presentation of the data, the logarithm (to base 10) of the remaining area at the time of escape is plotted on trial number. Because procedures and plans were modified during the course of the experiments, results and interpretation are combined in the presen- tation of the results. PRELIMINARY EXPERIMENTS WITH TRAINED PORPOISE We anticipated that the behavior of the por- poise would change rapidly during the course of the experiments; thus, to avoid wastage of the naivete of the limited and expensive supply of untrained animals, we conducted a series of pre- liminary exi:)eriments with three trained por- poise (Table 2, Figures 2-4). WIDTH OF OPENING The effect of the width of the opening on the escape l)ehavior of the three trained porpoise was first tested by presenting on alternate trials an escape route of standard width, 5.5 m, and one either 3.1 or 3.8 m wide. In Waimea (Figures 2-1 and II) and Nohea (Figure 4-1 V) there was some evidence that the porpoise escai)ed sooner when the wider e-scajie route was used but not in Nani (Figure 3-II). We felt this was i)robably an artifact of experimental design as described above and consequently we considered only the data from the block experiments for evaluating effects of width on the trained pori)oise. To de- termine the width of opening that would influ- ence performance, Nohea was tested over six FISHERY BULLETIN: VOL. 70, NO. I WAIMEA FAILED TO ESCAPE 092m DEEP, WIDTH VARIED 092m DEEP. WIDTH VARIED TRIAL NUMBER Figure 2. — Results of experiments with trained porpoise Waimea. Each plot summarizes one day's continuous experimentation, as follows: I. May 21, II. May 22, III. May 23, IV. May 25, 1970. NANI 55m WIDE, DEPTH VARIED / I Im DEEP, WIDTH VARIED / / o 55m WIDE, „ DEPTH VARIED OI5m DEEP O O O FAILED TO ESCAPE 55m WIDE. DEPTH VARIED TRIAL NUMBER Figure 3. — Results of experiments with trained porpoise Nani. I. May 26, II. May 27, III. May 28, 1970. 52 PERRIN and HUNTER: ESCAPE BEHAVIOR OF PORPOISE I • < /^ • / OlSmOEE 1 5.5. n WIDE. DEPTH VAH EO ZO S5 30 35 Vv^ .^V*^^ l-Jm *iOt ; Wm WIDE / 1 DEEP, WIDTH VARIED FAILED TO ESCAPE O 55m WIDE; DEPTH VARIED Sim WIDE; DEPTH VARIED TRIAL NUMBER Figure 4. — Results of experiments with trained porpoise Nohea. I. May 30, II. May 31, III. June 1, IV. June 21, V. June 28, VI. June 29, 1970. blocks of six trials each (Figure 4-III). The width of the escape route was 3.1 m in the sec- ond, 1.5 m in the fourth, and 0.76 m in the sixth block of experimental trials, with interspersed blocks of trials at 5.5 m. A significant decrease in performance occurred only when the escape opening was narrowed to 0.76 m. In one trial at 0.76 m the animal refused to leave the crowd- ing chamber and had to be extricated from the webbing. In others, the porpoise exhaled air and sank passively to the bottom of the pool and did not move even when the chamber was com- pletely closed. Exhalation of air and sinking to the bottom was a pattern that appeared in other porpoise in other experiments and was accom- panied by failure to escape. Our tentative conclusion was that the width of the opening w^as not a significant variable in the block experiments if it exceeded about 1.5 m. DEPTH OF OPENING We determined the effect of the depth of the escape route by varying depth of the hole from 1.8 m to 0.15 m while maintaining the standard hole width of 5.5 m. We will describe the results for each porpoise separately since the experi- ments w^ere diflFerent for each animal. Waimea failed in the first eight trials to escape through an opening 0.92 m deep (Figure 2-1). Performance improved thereafter to a plateau that was maintained throughout a subsequent identical experiment the next day, throughout a series of alternating trials with 0.92-m- and 0.61-m-deep openings on the third day of the ex- periments, and in a fourth experiment (Waimea IV) 2 days later with a 0.92-m-deep opening. Nani showed no difference in response after the first two trials with openings 1.8 m and 1.1m deep (Figure 3-1). High performance contin- ued through a series of trials with a 1.1-m-deep opening, but dropped in blocks of trials of 0.61-m- and 0.30-m-deep openings (II). In sub- sequent experiments Nani failed to escape twice when openings 0.15 m and 0.30 m deep were used and performed erratically in blocks of trials with openings 0.61 m and 0.30 m deep (III). After seven trials with the 0.30-m opening, no failures was experienced in seven trials with a 0.15-m opening, but the animal escaped consistently earlier (larger remaining area) when the 1.1-m- deep opening was used. Nohea escaped earlier when the hole was 1.1m than when it was 0.61 m deep in 14 of the first 16 trials of an alternating series (Figure 4-1), In a subsequent series of blocks of trials at decreas- ing depths (II), a pronounced drop in perform- ance occurred at depths of 1 ft and 5 ft. The following day's performance remained at a high level except when a corkline was strung across the opening (IV). Our tentative interpretation of the above re- sults was that a critical depth of opening lay near 1 m: 11 failures to escape occurred at 0.92 m or shallower; none occurred with openings 1.1 m deep or deeper; and performance was even more adversely aflfected by further decreasing the depth of the opening. We also concluded that the results of the first few trials for each ani- 53 FISHERY BULLETIN: VOL. 70, NO. 1 mal were of most importance in predicting the probable response of naive wild porpoise, as the animals were able quickly to achieve high levels of performance even at very shallow depths. CORKLINE A corkline across the toj) of the ojiening caused Waimea (III) and Nani (II) to fail on initial trials, and greatly affected the performance of Nohea (IV) . Waimea, after four failures, over- came reluctance to pass through an oi:)ennig with a corkline at the surface and reachieved a high level of performance. An interaction between the corkline and depth of opening was apparent in the factorial experiment with Nohea. Initial trials with the corkline (second block) produced a temporary drop in performance with a 1.1-m- deep opening. In the fourth block, the corkline was again inserted, and performance dropped at 0.61 m depth but not at 1.1 m. MONOFILAMENT PANEL When the panel of nylon monofilament web- bing was inserted into the opening (1.1 m deep) after a series of trials in which performance was consistently high, Waimea (IV), Nani (III), and Nohea (III) swam into the webbing as if it did not exist. Performance in subsequent trials without the i^anel was not affected (Nani III). Upon hitting the webbing, the porpoise became entangled and had to be extricated by a diver. POLYVINYL PANEL In the two single trials with a panel of clear polyvinyl sheeting inserted in the opening, Nohea (V and VI) hit the ])anel and slid over the top as it buckled. No difference was noted in be- havior in these trials from that in trials in which the panel was absent. During these experiments Nohea in several trials passed back and forth through the opening two or three times after the initial escape, while the net wall was being closed. The values for the surface area index shown in the figure are for the first passage. The incidence of such be- havior throughout the course of all the experi- ments occurred only after considerable expe- rience with a particular net configuration. In most cases, only one or two double "escapes" occurred during an experiment. EXPERIMENTS WITH NAIVE PORPOISE Eleven experiments were conducted with the naive porpoise (Table 3, Figures 5 and 6). The first naive animal. Westward, was captured on June 12, 1970, and after a relatively short han- dling period was placed in Bateson's Bay. Her swimming behavior during the first 5 days of captivity was unlike that of the trained porpoise. The trained porpoise continually swam about the tank during and between experiments, diving and "porpoising," and spinning. Westward, on Table 3. — Experiments with naive porpoise. Porpoise Experiment Variables tested Design Number of trials Westward 1 Depth Alternating trials 20 II Depth Alternating trials 20 III Depth Block 36 IV: triols 1-40 Depth, width, corkline Factorial 40 trial 41 Monofilament panel Single trial 41 V V\'idth Block 36 VI: trials I-IO Depth, thin polyvinyl panel Factorial 10 trials 11-15 Thick polyvinyl panel Block 5 Moana 1 Depth Alternating trials 11 II Depth Alternating trials 20 III Width Block 36 IV Depth Block 24 V: trials 1-12 Depth Block 12 trials 13-25 Width, corkline Factorial 13 54 PERRIN and HUNTER: ESCAPE BEHAVIOR OF PORPOISE WESTWARD MOANA 55m WIDE: DEPTH VARIED N v\// fAILEO ,T9 ESCAPE, 5.5m WIDE; DEPTH VARIED n DEEP :a6lmOEEP : l.lm DEEP | Q30m DEEP 55m WIDE, DEPTH VARIED ■t] r\ < 5 10 IS 20 25 30 35 TRIAL NUMBER 500 f m 1 Mm DEEP, , w V^^i . S^^ llmOEEP^ . VNmDEEP H- lOO ■'^^im DEEP ;'!'" deep/ \ ,/^lm WtOE / /06lm DEEP m \ '\ 5.5m WIDEi DEPTH VARIED FAILED >l \ TO ESCAPE© a DEPTH VARIE, OI5m DEEP ,V^ O TO ESCAPE /55m WiDf Urn DEEP, WIDTH VARIED : &5m WiOE I DEEP; WIDTH VARIED FAILED TO ESCAPE O 20 25 50 35 TRIAL NUMBER Figure 6. — Results of experiments with naive porpoise Moana. I. July 10, II. July 11, III. July 12, IV. July 13, V. July 15, 1970. carried out the day after capture. Subsequently Westward's behavior slowly changed, until 5 days later on June 17 it was indistinguishable from that of the trained porpoise. The re- mainder of the Westward experiments were car- ried out after June 17. Moana, the second naive porpoise, was cap- tured on July 9, 1970. When placed in Bateson's Bay, she exhibited the same behavior as West- ward, but to a lesser extent. Periods of surface swimming in a semiupright position, but without head-bobbing, were interspersed with periods of normal porpoising and diving. During the first experiment she swam slowly at the surface in the diagonal posture but during the second and subsequent experiments, her behavior was sim- ilar to that of the trained animals. 55 FISHERY BULLETIN: VOL. 70, NO. 1 DEPTH OF OPENING In the first few trials (Fioure 5-1) a marked difference existed in the response of Westward to an opening: 1.1 m deep and one 0.61 m deep. The animal swam slowly at the surface, circling or moving back and forth in the chamber. When the opening was 1.1 m deep, she moved slowly through the opening .iust as the moving wall closed. When the opening was 0.61 m deep, she moved past or circled slowly in front of the open- ing and then dove and entangled herself in the webbing of the moving wall. In the sixth trial, her behavior became more varied; she swam in tight circles beneath the surface and attempted to pass between the end of the moving wall and the periphery of the chamber before passing through the opening, after which she slapped her tail against the water surface. Behavior in subsequent trials became increasingly erratic. In trial 13 she darted through the opening rather than moving through slowly as in the previous trials. In trial 14, she tried again to squeeze past the moving wall and became lodged in the narrow opening. In trial 15, she assumed a position across the corkline of the moving wall, half in and half out of the chamber, and remained there until removed. In the remaining trials, she moved rapidly through the opening, and in the last two, she assumed a horizontal attitude sim- ilar to that usually taken by the trained ])or- poi.se and .stop])ed bobbing her head but still kept her blowhole above the surface. An identical experiment (II) of alternating trials was carried out 5 days later, after all traces of the slow surface-swimming and head-bobbing l)ehavior had disappeared. Performance was consistently higher with the 1.1-m opening. The effect of depth is clearly seen in the results of a block-design experiment for Westward (III). The second naive porpoise, Moana, a smaller and presumably younger animal than Westward, achieved a higher rate of successful passage in the first depth experiment (Figure 6-1). She failed only once, with the 1.1-m-deep opening. In the .second depth experiment (II) her per- foi-mance was extremely variable compared to that of Westward, and no relation between depth and success rate existed. In two trials (18 and 20) while swimming in tight circles near the apex of the chamber, she snagged her flipper in the webbing and had to be extricated. In later block-design experiments (V and IV) the effect of depth was evident as it was for Westward. WIDTH OF OPENING Results of block-design experiments testing the effect of width of opening for Westward (Figure 5-V) and for Moana (III) were similar to those for the trained porpoise (Nohea), but an effect was discernible at widths of 1.5 m. As with the other experiments, performance of Westward was higher and more stable than that of Moana. Westward began to pass through the opening two or three times during a single trial. The frequency of multiple "escapes" was higher for the 5.5-m-wide opening than for the narrower openings (Table 4). CORKLINE Insertion of a corkline at the surface across the top of the opening sharply affected the per- formance of Westward (IV) and Moana (V). The performance of Moana showed the greatest effect. After a series of preparatory trials, Moana failed to pass through the opening in five straight trials with the corkline. In each trial she laid the anterior part of her body across the corkline and remained there until removed. In the block-design experiments with Westward (IV), the second block of trials with a corkline produced a smaller drop in performance than did the first, with the 0.61-m-deep opening only, dem- onstrating as for the trained porpoise (Nohea IV, Figure 4) an interaction between depth and presence or absence of a barrier at the surface. Table 4. — Multiple escapes of Westward. Width of open in block of six trials ng Number of double escapes Number of triple escapes m 5.5 0 0 3.1 1 0 55 2 2 1,5 1 0 5.5 2 0 0.76 0 0 56 PERRIN and HUNTER: ESCAPE BEHAVIOR OF PORPOISE MONOFILAMENT AND POLYVINYL PANELS When the monofilament panel was inserted in- to the 1.1-m-deep openino- at the end of a series of depth trials, Westward (IV) "s'ot u]i a full head of steam and plowed into the monofilament" (extracted from field notes of W. Wasden) and became entangled. Insertion of polyvinyl panels produced similar results in multiple trials; West- ward (VI) in each trial hit the panel and slid over it and out of the chamber. There was nothing in the behavior of the porpoise to indi- cate that they recognized the presence of the panels. DISCUSSION AND CONCLUSIONS The swimming behavior of the naive porpoise Westward and, to a lesser extent, of Moana, the first few days after capture was very similar to that of porpoise {Stenella spp.) in tuna purse seines as ol)served by one of us (Perrin) off Cen- tral America. A typical "failure to escape" epi- sode is illustrated for Moana in Figure 7. Im- mediately after a purse-seine net has been set, when the diameter of the encircled area is great- est (approximately 250 m), the porpoise swim about quite rapidly in small tight groups of a dozen or so individuals, the mem])ers of a group diving and surfacing together (Figure 8). As the net is hauled and the area enclosed becomes smaller, especially after the backing down oper- ation (see Perrin, 1969) , the porpoise congregate and raft near the center of the enclosure and mill very slowly, holding their bodies in a semiup- right position with blowhole exposed and ros- trum at or slightly below the surface (Figure 9) . At this point, individual animals can be seen to leave the grou]) and dive. When the net has been completely hauled, animals are often found with their snouts entangled in the webbing several meters below the corkline. Although the head bobbing exhibited by West- ward was not observed in the purse-seine situ- ation, the similarities in Ijehavior between freshly captured animals and those captured in a purse seine were striking. In both cases the animals did not display normal motor patterns; they rested or swam at abnormally slow speeds, and this behavior was often ended l^y a rapid dive beneath the surface with no noticeable change in behavior preceding the act. The prin- cipal characteristics of this behavior, the inhi- bition of activity in a fear-inducing environ- ment, resembled fear responses described for many other vertebrates and frequently classified as an immobility or freezing response (Ratner and Thompson, 1960; Hinde, 1970). Hogan (1965, 1966) suggested that withdrawal and im- mobility are separate, mutually inhibitory sys- tems. If this view is correct, then driving por- poise through an escape route in the luirse seine would not be successful once the animals began to show the immobility response, because with- drawal would be inhibited. Under these circum- stances the additional fear stimulus associated with driving might be the catalyst for the rapid dive to escape, which results in entanglement. Driving may have to be carried out before im- mobility begins. Once the animals became im- mobile the only strategy may be to ])ull the net out from beneath them as is currently done dur- ing the "backing down operation" (Perrin, 1969). That the behavior of Westward and Moana evolved into more typical behavior during the course of a single ex])eriment also supports the notion that their unusual behavior was caused by the circumstance of captivity rather than ill health. Our conclusions with respect to projected de- sign of a rescue gate for removing porpoise from a purse seine during fishing operations were: 1. The gate should be sufficiently wide so that when the perimeter of the net circle buckles after pursing, the width does not become less than 1.5 m. Considering the equivocal results of the ex])eriments for o])enings wider than 1.5 m, the opening should be as wide as prac- tically possible. 2. Depth of the opening should be not less than 1 m and as deep as it is possible to make it with- out causing loss of the fish in the net. 3. There should be no line, corkline, or other barrier across the oi)ening at the surface. 4. A self-actuating release port that will open when struck by a porpoise swimming into it 57 FISHERY BULLETIN: VOL. 70, NO. I 58 PERRIN and HUNTER: ESCAPE BEHAVIOR OF PORPOISE Figure 7. — Typical "failure to escape" episode. Moana (1) pa- trols moving wall at beginning of trial, then (2) takes up position at apex of chamber and remains there for most of trial, in vertical attitude. As chamber nears clo- sure Moana dives (3) , orients to- ward opening (4) , and turns and swims into moving wall (5), be- coming entangled (6). %fe - -^*^ Figure 8. — Porpoise (Stevella grnffmnni) in tuna purse seine at beginning of set, when net is at near-maximum diameter. An- imals are circling and diving in groups of a dozen or so individu- als. Figure 9. — Porpoise in purse seine, after most of net has been taken aboard. Animals are "rafting" in compact group, each maintaining approximately vertical attitude, with blowhole exposed and dorsal fin submerged. Large fish underwater in foreground are yellowfin tuna. 59 FISHl.RY BULLETIN: VOL. 70, NO. 1 miorht be feasiV)le if constructed of acoustically transi)arent materials, providing: that it were so constructed that the fish in the net would not also use it. 5. It is to be exjiected that grreat difficulty will be encountered in inducing: wild porpoise to pass throuo:h an opening in the perimeter of a purse- seine enclosure. ACKNOWLEDGMENTS We thank the Oceanic Institute and its Di- rector. Dr. Kenneth Norris, for providino- the facilities and the porpoise used in this study. William Wasden, National Marine Fisheries Service, Honolulu, Hawaii, assisted in all phases of the study, and por]wise trainers Tng-rid Kano-, Sea Life Park, and Scott Rutherford. Oceanic Institute, provided as.sistance and advice during the course of the experiments. LITERATURE CITED Harrison, R. J., R. C. Roice, .wd R. I.. Rrownrll, Jr. 1969. Rpproduction in wild and captive dolphins. Nature (London) 222: 1143-1147. HESTKR, F. J., J. R. Ht'NTKR, AND R. R. WHITNEY. 196.'5. Jumping and spinning hohavior in the spin- ner porpoi.se. J. Mammal. 44: 586-588. HiNDE, R. A. 1970. Animal behaviour. A synthesis of ethology and comparative psychology. McGraw-Hill Book Company, 876 p. HOG.AN, J. A. 1965. An experimental study of conflict and fear: An analysis of behavior of young chicks toward a mealworm. Part I. The behavior of chicks which do not eat the mealworm. Behaviour 25: 45-97. 1966. An experimental study of conflict and fear: An analysis of behavior of young chicks toward a mealworm. Part K. The behavior of chicks which eat the mealworm. Behaviour 27: 273-289. Miller, G. S., Jr., and R. Kellogg. 1955. List of North American recent mammals. U.S. Natl. Mus. Bull. 205: 1-954. Morris, R. A., and L. S. Mowbray. 1966. An unusual barnacle attachment on the teeth of the Hawaiian spinning dolphin. Nor. Hval- fangsttid. 55: 15-16. NlSHIWAKI, M. 1967. Distribution and migration of marine mam- mals in the North Pacific area. Bull. Ocean Res. Tnst., Univ. Tokyo 1: 1-64. Perrin, W. F. 1969. Using porpoise to catch tuna. World Fish- ing 18(6): 42-45. 1970. The prolilem of porpoise mortality in the U.S. tropical tuna fishery. Proc. Sixth Annu. Conf. Biol. Sonar and Diving Mammals, Stanford Res. Inst., Menlo Park, Calif., p. 45-48. Pilson, M. E. Q., AND D. W. Waller. 1970. Composition of milk from spotted and spin- ner porpoises. J. Mammal. 51: 74-79. Ratner, S. C, AND R. W. Thompson. 1960. Immobility reactions (fear) of domestic fowl as a function of age and prior experience. Be- haviour 8: 186-191. Rice, D. W., and V. B. Scheffer. 1968. A list of the marine mammals of the world. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 579, 16 p. ToMicii, P. Q. 1969. Mammals in Hawaii. Bernice P. Bishop Mus. Spec. Publ. 57, 238 p. True, F. W. 1889. Contributions to the natural history of the cetaceans. A review of the family Delphinidae. U.S. Natl. Mus. Bull. 36, 191 p. + pi. I-XLVIL 60 METHODS FOR TAGGING SMALL CETACEANS W. E. Evans,'- ^ J. D. Hall,- A. B. Irvine,^ and J. S. Leatherwood^ ABSTRACT Four types of tags have been used on four species of delphinids. These include a circular plastic button tag that is attached to the dorsal fin by a nylon bolt, a highly visible dart-type spaghetti tag that is placed near the base of the dorsal fin, a radio transmitter tag, and a freeze brand. Use of button tags has been discontinued due to high shedding rate. The dart-type spaghetti tag has proved best for tagging large numbers of animals without capturing them. The radio tag provides very detailed information on behavior and movements, while freeze branding provides a permanent mark, though both require capturing the animal. The importance of marking commercially valu- able species of whales (primarily the larger baleen whales and the sperm whale) has long been recognized. Since their development in the mid-1920's, "Discovery-type" tags have been used to mark large numbers of these animals (Rayner, 1940; Brown, 1962; Clark 1962). Re- turns from these tags have provided valuable in- formation on the species' distribution, migration, and abundance and on such basic aspects of their biology as relative growth rates and the timing of the events in their lives (Mackintosh, 1965). The relationship of several small delphinid species to commercial fish populations and the potential of these cetaceans as a major economic resource has renewed interest in their stocks during the last decade (Perrin, 1970). Early attempts to study these populations in the wild have been hampered by the difficulty of posi- tively identifying an animal or a population from one encounter to the next. Therefore, develop- ment of a reasonable method for marking these animals for identification would facilitate studies of their life histories. Although several investigators have tried tag- ging small cetaceans, only three have had even moderate success. In a program conducted by ^ Authors are listed in alphabetical order. '^ Marine Life Sciences Laboratory, Naval Undersea Research and Development Center, San Diego, CA 92132. ' Mote Marine Laboratory, Sarasota, FL. 33581. the Oceanic Institute, Oahu, Hawaii, plastic cat- tle eartags were placed on two Steno hredanensis and one Stenella attennata (Evans, 1967) . This program was continued by Norris and Pryor (1970), and at least one of the tags was still on a Stenella attennata when it was resighted after 3i/4 years. Sergeant and Brodie (1969) tagged 812 be- lugas, Delphi napteriis leucas, in Hudson Bay, Canada, over a 2-year period. Six hundred and ninety-four of these animals were tagged with a spaghetti tag originally designed by Mather ( 1963) for use in tagging pelagic fishes and man- ufactured by Floy Tag Company,' Seattle, Wash. The remaining 118 belugas were tagged with Petersen disc tags, similar to the button tags we used. Of the 812 animals tagged, 2 with spaghetti tags were recovered by the beluga fish- ery. A third spaghetti tag was observed in a live animal temporarily stranded by the ebbing tide 1 year after the original tagging. Perrin and Orange (1971) tagged 21S, Ste7iella spp. in 1969 and approximately 1,000 in 1970 in the eastern tropical Pacific with spaghetti-type dart tags. Five tags have been recovered; max- imum time at liberty was 138 days (916 km net movement) , Since 1968, personnel of the Naval Undersea Research and Development Center's Marine Bio- science Division at San Diego, Calif., have been * Reference to commercial products does not imply endorsement by the National Marine Fisheries Service. Manuscript accepted September 1971. FISHERY BULLETIN: VOL. 70, NO. 1, 1972. 61 FISHERY BULLETIN: VOL. 70, NO. 1 investigating the distribution and biology of sev- eral odontocete cetaceans off the southern Cali- fornia coast. In order to delineate migration routes and to keep track of local herds of the com- mon dolphin, Delphinns delphis auctt., a tagging program was initiated. During the same period, a tagging program was also initiated for Tur- siops tnoicatus on the west coast of Florida. The special problems associated with tagging odontocete cetaceans required the modification of old and the development of new tagging tech- niques. This paper discusses the relative merits of the four marking methods used by our labora- tory. In addition, it presents some preliminary results of the program in order to substantiate the utilitv of the various methods. METHODS AND RESULTS We have used modified dart-type vinyl spa- ghetti tags (Floy Manufacturing Company) on four species of Eastern Pacific delphinids in an area from Point Conception, Calif., to Cabo San Lucas, Baja California, Mexico, and throughout the Gulf of California. Our original spaghetti tags were 5 mm in diameter by 17 cm long. In order to increase visibility and flow character- istics of the tag, we increased the length to 30 cm (Figure 1). Using the modified tag, w^e have marked 240 D. delphis, 10 Lagenorhynchm obliquidens, 8 Tursiops pilli auctt., and 13 Stenel- la gmffmani to date (July 1971). The animals were all tagged at the anterior insertion of the dorsal fin while they were surfing on the bow pressure wave. Several dolphins were observed to continue riding the bow pressure wave after being tagged, so the tagging process apparently did not affect their normal behavior. A T. gilli auctt., tagged on 27 October 1970, off Magdalena Bay, Baja California, was recov- ered by an American tuna boat off Manzainillo, Mexico, on 22 January 1971. The animal had covered at least 816 km between the time of tag- ging and the time of capture, a period of just less than 3 months. Three D. delphis bearing spaghetti tags have been observed swimming in the vicinity of the Coronado Islands near San Diego, Calif., and at least one spaghetti-tagged D. delphis has been sighted off Magdalena Bay, Baja California. Each of these animals was known to have been carrying the tag for from 2 weeks to several months. Circular plastic "button" tags (10 cm diam) (Figure 2) were through-bolted to the dorsal fins of 46 D. delphis and 6 L. obliquidens between 1967 and 1970. These tags are similar to those employed by Norris and Pryor (1970) in Hawaii, but are larger to make them more easily spotted. Button tags were attached to animals captured off the southern California coast, or near Cedros ' V 0 0 1 'J r ■■™ffTraeAi.«i»««u»«i! Figure 1. — The dart-type spaghetti tag in place on the tagging apparatus. 62 EVANS ET AL. ; TAGGING SMALL CETACEANS To meet this need, a new lightweight radio tag (170 g) with a 9-12 month transmitter life was developed. This tag combines the advantages of a radio beacon and a button tag in that it contin- ues to serve as a color coded marked even after it no longer transmits (Figure 3), Further- more, the new radio tag is available commer- cially at less than 10 ^r of the cost of the 900/gm transmitters. Figure 2. — Three Lagenorhynchus obliqiddens with the plastic button tag, just prior to release. Island, Baja California. Two of the L. obliqui- dens tagged in 1969 were resighted almost 1 year later, and a D. delphis tagged in 1968 was re- sighted 21 months later. Twenty-four T. truncatus were tagged with the button tags near Sarasota, Fla., from August 1970 through September 1971. Animals bearing tags have been resighted several times. The third and most successful short-term tag is the radio transmitter tag with which at least four species of small cetaceans have been success- fully marked to date (Evans, in press, Martin, Evans, and Bowers, 1971) . The original package used in these studies was a 27 mHz (11m) trans- mitter and antenna housed in a waterproof envelope which is attached to the dorsal fin of a dolphin or a small whale by means of a spring- loaded corrosible link. The link dissolves and releases in 30 days, allowing the package to slip off the animal. These early radio beacons, designed for short- term transmission (30-60 days), weighed up to 900 g, and though they proved especially useful in studying the detailed movements of D. delphis in the waters off San Diego, Calif., their size, cost, and relatively short transmission time made them unacceptable for long-term monitoring of herd movements. Figure 3. — The lightweight (170 g) radio tag. 63 FISHERY BULLETIN: VOL. 70. NO. 1 The fourth method, freeze branding, consists of applying a supercooled branding iron, usually copper, to the epidermal surface of the dolphin for 5-30 sec. Evidence from freeze branding cattle indicates that the branding process is pain- less to the animal and has no lasting effect other than leaving a permanent mark (Farrell, Lais- ner. and Russell, 1969) . Though evidence of the branding usually becomes indistinct shortly after application, after about 2 months the animal will display a highly legible brand (Figure 4). We have used this method on eight wild T. ti-uncatus near Sarasota in conjunction with either a but- ton tag or a spaghetti tag. The number on a freeze branded animal was clearly visible, from a distance of 40 yards, when the animal was resighted 10 weeks after tagging. FiGi.RE 4. — A Tumiops truncatus with the freeze brand on the dorsal fin. DISCUSSION We have discontinued use of the button tag in favor of the spaghetti and radio tags. In- cidence of loss of button tags from animals has been exceptionally high among the T. truncatus around Sarasota, and the few resightings of but- ton-tagged dolphins off southern California lead us to believe that button tag loss is high in this area also. A major disadvantage of the button tag is that the animal must be captured in order to be tagged. The spaghetti tag, on the other hand, is normally placed in the animal while it is free swimming and thus does not require cap- ture. Using this method we have placed over 50 spaghetti tags in one herd of D. delphis in less than 2 hr. When spaghetti tags are placed in the fibrous tissue at the insertion of the dorsal fin, incidence of tag loss appears to be lower for spaghetti tags than for the button tags (Nishi- waki, Nakajima, and Tobayama, 1966). In either case, the numbered information on the tag is so small that it cannot be read on a moving animal at sea. Unless the spaghetti tags are color-coded, resighting at sea can give no in- formation on the original tagging location. Spa- ghetti tags may also be placed in an animal that has been captured. The radio tags can be placed only on captured animals but provide very detailed information concerning exact movement and diving patterns of the animal. While freeze branding involves capture of the animal, it appears to provide permanent and highly legible identification of cetaceans. Tom- ilin (1962) reported taking a Black Sea D. delphis in 1953 which bore a brand posterior to the eye. The brand was quite legible and contained numbered information. The source and nature of the brand were not known. In the future, we plan to freeze brand all the dolphins we capture for radio tagging and to continue to use the spaghetti tags for free-swim- ming delphinids. An advertisement was placed in the July issue of National Fisherman requesting that any in- formation on sightings of tagged delphinids in the Eastern Pacific be forwarded to the Marine Bioscience Division of the Naval Undersea R&D Center, San Diego, Calif. (Evans, Leatherwood, and Hall, 1971). Copies of this advertisement have been placed at sportfish landings and com- mercial docks from Santa Barbara to San Diego, Calif. LITERATURE CITED Brown, S. G. 1962. A note on migration in fin whales. Nor. Hvalfangst-Tid. (Norw. Whaling Gaz.) 51(1): 13-16. 64 EVANS ET AL.: TAGGING SMALL CETACEANS Clarke, R. 1962. Whale observation and whale marking off the coast of Chile in 1958, and from Ecuador to- wards and beyond the Galapagos Islands in 1959. Nor. Hvalfangst-Tid. (Norw. Whaling Gaz.) 51(7): 265-287. Evans, W. E. 1967. Vocalization among marine mammals. In W. B. Tavolga (editor), Marine Bio-Acoustics, Vol. 2, p. 159-186. Pergamon Press, New York. In press. Orientation behavior of delphinids: radio telemetric studies, presented at the Conference of Animal Orientation : Sensory Basis, sponsored by New York Academy of Sciences, Feb. 8-10, 1971. Ann. N.Y. Acad. Sci. Evans, W. E., J. S. Leatherwood, and J. D. Hall. 1971. Request for information on tagged porpoises on the eastern Pacific. Natl. Fisherman 52(3): 15A. Farrell, R. K., G. a. Laisner, and T. S. Russell. 1969. An international freeze-mark animal identi- fication system. J. Am. Vet. Med. Assoc. 154 : 1561-1572. Mackintosh, N. A. 1965. The stocks of whales. Fishing News (Books) Ltd., London, 232 p. Martin, H., W. E. Evans, and C. A. Bowers. 1971. Methods for radio tracking marine mammals in the open sea. Transactions of the IEEE Con- Terence on Engineering in the Ocean Environment, September 1971, San Diego, Calif. Mather, F. J., III. 1963. Tags and tagging techniques for large pe- lagic fishes. Int. Comm. Northwest Atl. Fish., Spec. Publ. 4: 2. NiSHiwAKi, :m., M. Nakajima, and T. Tobayama. 1966. Preliminary experiments for dolphin mark- ing. Sci. Rep. Whales Res. Inst. 20: 101-107. NORRis, K. S., ANT) K. W. Pryor. 1970. A tagging method for small cetaceans. J. Mammal. 51: 609-610. Perrin, W. F. 1970. The problem of porpoise mortality in the U.S. tropical tuna fishery. Proc. 6th Annu. Conf. Biol. Sonar Diving Mammals. Stanford Res. Inst, Menlo Park, Calif., p. 45-48. Perrin, W. F., ant) C. J. Orange. 1971. Porpoise tagging in the eastern tropical Pa- cific. Proc. 21st Tuna Conf., Lake Arrowhead, Calif., October 1970, p. 5. Rayner, G. W. 1940. W^hale marking, progress and results to De- cember 1939. Discovery Rep. 19: 245-284. Sergeant, D. E., and P. F. Brodie. 1969. Tagging white whales in the Canadian Arctic. J. Fish. Res. Board Can. 25: 2201-2205. TOMILIN, A. G. 1962. The migrations, geographical races, the thermo-regulation and the effect of the tempera- ture of the environment upon the distribution of the cetaceans. Fish. Res. Board Can., Transl. Ser. 385: 1-24. 65 A REVIEW OF THE LANTERNFISH GENUS Taaningkhthys (FAMILY MYCTOPHIDAE) WITH THE DESCRIPTION OF A NEW SPECIES Brent Davy^ ABSTRACT The genus Taaningichthys includes three known species, one of which is here described as new. The species of the genus Taaningichthys do not appear to perform daily vertical migrations. Evidence indicates vertical stratification of juveniles and adults. Although photophores and lateral line are reduced, the species of Taaningichthys possess very large eyes which may be related to capture of luminescent prey. Otoliths of all three species have been examined and found to be taxonomically important. Bolin (1959) erected the genus Taaningichthys to include two species, T. hathyphilus and T. mi- nimus, previously placed in the genus Lampade- na by Taning (1928). The main characters which distinguish Taaningichthys from Lampa- dena are: (1) the origin of the dorsal fin in Taaningichthys is clearly behind the base of the pelvic fins; (2) the development in Taaningich- thys of a crescent of white tissue' on the pos- terior half of the iris, although a similar white (luminous?) crescent is present on the dorsal portion of the iris in Lampadena chavesi (Naf- paktitis and Paxton, 1968); (3) the presence of a single SAO, or none, in Taaningichthys (always three SAO in Lampadena) ; (4) re- duced dentition and lateral line in Taaningich- thys. Taaningichthys may be distinguished from all other myctophid genera by the combination of the white crescent of tissue on the posterior half of the iris, the undivided luminescent caudal glands, and the single or altogether absent SAO. Berry and Perkins (1966) reported what they thought to be a third form of Taaningichthys apparently without photophores. Following the capture of a number of specimens of this form ^ Department of Biological Sciences, Allan Hancock Foundation, University of Southern California, Los An- geles, Calif. 90007. "This tissue is not visible until some time after preser- vation and is hardly distinguishable in specimens initially frozen and then preserved. by the RV Velero IV of the University of Southern California and the examination of con- siderable material made available to me by nu- merous institutions around the world, I felt that a review of the genus was appropriate. MATERIALS AND METHODS Members of the genus Taaningichthys are deep-dwelling, fragile myctophids, easily dam- aged by the net. Scales are readily lost, and damage to the bones of the snout, upper jaw, and operculum is very common. Consequently, measurement of jaw, head, and snout length is often very difficult if at all possible. The follow- ing measurements were taken on the best pre- served specimens: Eye diameter (ED) — hori- zontal distance across the orbit; jaw length ( JL) — length of premaxillary; predorsal (Pre D) — anterior tip of premaxillary to base of anterior- most ray of dorsal fin ; pre ventral (Pre V) — an- terior tip of premaxillary to base of anteriormost ray of ventral fin; preanal (Pre A) — anterior tip of premaxillary to base of anteriormost ray of anal fin; prepectoral (Pre P) — anterior tip of premaxillary to base of anteriormost ray of pec- toral fin; preadipose (Pre Ad) — anterior tip of premaxillary to posterior end of base of adipose fin; length of supra- and infracaudal luminous glands — length of exposed luminous tissue only; anal-infracaudal distance — anterior tip of Manuscript accepted July 1971. FISHERY BULLETIN: VOL. 70, NO. 1, 1972. 67 FISHERY BULLETIN': VOL. 70, NO. I infracaudal gland to end of base of anal fin. Sizes of specimens are given in standard lengths (SL) only. Terminology of body photophores follows that of Bolin (1939). Unless otherwise specified, the term photophore refers to the primary body l^hotophore. Otoliths were measured with an eyepiece mi- crometer as follows: Length (OL)— the great- est length parallel to the sulcus; height (OH) — greatest height iierpendicular to the sulcus. Fol- lowing measurements, otoliths were lightly smeared with graphite to bring out detail and then photographed. Otolith terminology fol- lows that of Frizzell and Dante (1965). Female specimens were considered gravid when eggs included oil globules and completely filled the oviduct. Most specimens examined were captured with open nets and depth sampled is here considered as the maximum depth reached l\v the net (appendix). Counts of procurrent caudal rays are given as dorsal + ventral. KEY TO THE SPECIES OF THE GENUS Taatihigichthys la. VO 8-10; AO 5-7 + 4-6, total 9-13; Pol directly below or anterior to base of adipose fin; Prci-Prc2 interspace equal to or greater than two photophore diameters; as many as five pairs of broad-based, hooklike teeth on dentary near symphysis T. minimus (Taning, 1928) lb. VO, if present, 3-5; AO, if present, 1-4 + 1-2, total 2-5; Pol, if present, clearly behind base of adipose fin; Prci-Prc2, if present, interspace equal to, or less than, one photophore diameter; no broad-based, hooklike teeth on dentary near sympiiysis 2 2a. Photophores present as in lb above; anal-infracaudal distance half as long as length of infracaudal gland, or longer T. hathyphilns (Taning, 1928) 2b. Photophores absent; anal-infracaudal distance less than half length of infracaudal gland T. j)au)'olychnus n. sp. ' /y. ■^^:.- .iAS W^^;- Figure 1. — Taaningichthys minimus (Taning) ; 46 mm, Ocean Acre stn. 7-21. 68 DAVY; REVIEW OF LANTERNFISH GENUS TaaningUhthys GENUS Taaningichthys Bolin Taaningichthys minimus (Taning, 1928) Figures 1 and 2 Lampadena minima Taning, 1928: 63; Parr, 1928: 154, Figure 37. Lampadena (Lampadena) minima Fraser-Brun- ner, 1949: 1078, Figure. Taaningichthys miriimus Bolin, 19.59: 26. D 11-13; A 12-13 (11-14); P 15-17; V 8; gill rakers 4-5 + 1 + 11 (10-13), total 16-17 (15- 18); VO 8-10; AO 5-7 + 4-6, total 9-13; Pre 2 + 1; vertebrae 40-41; procurrent caudal rays 8-10 + 8. Mouth terminal, moderately large, JL about 1.5 in Pre P; maxillary slightly expanded pos- teriorly. Eye large, ED 2.2-3.4 in Pre P. Pter- otic spine long and directed posteriorly. Oper- cular margin concave posterodorsally, slightly convex posteriorly. Pectoral fin long, reaching VOe or VO7; its base about midway between ventral body margin and horizontal septum. Pre V 2-2.4 in SL. Pre D 1.9-2.3 in SL; end of base of dorsal fin clearly in advance of vent. Pre A 1.4-1.6 in SL. Anterior end of base of adipose fin on vertical through posterior end of base of anal fin; Pre Ad 1.2-1.3 in SL. A band of dark pigment along anteroventral margin of orbit containing a series of light gray, triangular patches of tissue not present in the other two species, Dn absent; Vn present between anterior mar- gin of orbit and posteroventral margin of nasal rosette. PVOi on or behind vertical through upper end of base of pectoral fin and about mid- way between it and ventral margin of body; PVO2 in front of middle of base of pectoral fin; a straight line through PVOi and PVO2 passing in front of PLO. PLO about halfway between upper end of base of pectoral fin and horizontal septum. Five PO. VLO above base of pelvic fin, usually closer to horizontal septum than to ventral margin of body. Last VO usually slightly elevated. SAO 1-2 photophore diameters be- low horizontal septum, directly above vent. AO level. AO series overlaps anterior end of in- fracaudal gland. Pol directly below or in ad- vance of base of adipose fin, 1-2 photophore ^ '"^j B «::>'"'"'# X«« Figure 2. — Taayiingichthys minimus. A. Side view, sex- ually dimorphic supracaudal gland of male, 53 mm. B. Top view, sexually dimorphic supracaudal gland of same male. C. Side view, sexually dimorphic supracaudal gland of female, 54 mm. D. Top view, sexually dimorphic supracaudal gland of same female. 69 FISHERY BULLETIN: VOL. 70. NO. 1 diameters loelow horizontal septum. Prci and PrC'j level, behind infracaudal g-land; Prc:i at horizontal septum. Caudal luminous glands undivided, the infra- caudal larg-er than the sujjracaudal and both covered by scales. Sexual dimorphism is evi- dent in supracaudal gland of adults (specimens about 40 mm and larger) ; in males this gland is about twice as large as in females (Figure 2) . Mesopterygoid teeth in narrow oval patches. Narrow band of needlelike teeth on palatine. No vomerine teeth. Both jaws with needlelike teeth which bend medially. A single row of broad-based, anteiiorly hooked teeth occupying posterior two-thirds of media! surface of den- tary. As many as five pairs of similar teeth, most often directed posteriorly, on medial sur- face of dentary near symphysis, and another two to three pairs projecting forward and lat- erally on symphysial area of iiremaxillary; be- low these, on anterior part of premaxillary, sev- eral posteriorly curved teeth (longer than rest of premaxillary teeth) . Tann'myichthys minimus is the shallowest dwelling, most firm-bodied, and smallest of the three sj^ecies, the largest examined specimen measuring 65 mm. Gravid females (about 40 mm and larger) were captured in August-September. Horizontal distribution — T. minimus occurs circumglo))ally between about lat 35° N and 30° S (Figure 5). It has been taken less frequently than T. bathyphilus. Vertical distribution — Closing-net data from the Project "Ocean Acre" in the north Atlantic Ocean suggest vertical stratification of juveniles and adults. Juveniles appear to inhabit depths of 140 to 250 m, the smallest specimen (21 mm) having been captured at 140 m. Adults occur ]-»redominantly in depths between 450 and 500 m. T. minimus does not appear to perform daily vertical migrations. Taanhigichthys bathyphilus (Taning, 1928) Figure 3 Lampadena bafhyphila Taning, 192S: 63; Parr, 1928: 151, Figure 36. Lampadena (Lampadena) bathyphila Fraser- Brunner, 1949: 1078, Figure. Taaningichthys bathyphilus Bolin, 1959: 26, Figure 6. D 12-13 (11-14); A 13 (12-14); P 12-14; V 8; gill rakers 3 + 1+7-8 (5-9), total 11-12 (9-13); VO 4 (3-5); AO 3 (1-4) +1(2), total 4(2-5); Pre 2 + 1; vertebrae 34-36; procurrent caudal rays 7 + 6. Mouth terminal, moderately large, JL about 1.5 in Pre P; maxillary slightly expanded poster- iorly. Eye large, ED about 2.5 in Pre P. Pterotic spine inconsi)icuous. Opercular margin as in T. mininuis. Pectoral rays reaching VOi; base of pectoral fin nearer to horizontal septum than to ventral margin of body. Pre V 2.1-2.5 in SL. Pre D 1.9-2.2 in SL; end of base of dorsal fin on, or slightly in advance of, vertical through SAO. Pre A 1.5-1.7 in SL. Base of adipose fin above end of base of anal fin; Pre Ad 1.2-1.4 in SL. V -O-jU^-a--** ' Fic.rRK ^.—Tamiingiehthys bathyphili(S (Taning) ; 02 mm, RV Velcro stn. 11733, LACM 30034-1. 70 DAVY: REVIEW OF LANTERNFISH GENUS Taamnguhthys i t ■ ■1/ ^^ ' :'^. Figure 4. — Taaningichthys pmirolychnus, new species, holotype, 67 mm, SIO 70-19. Dn absent; a very small oval Vn, visible in young individuals and masked by darkly pig- mented tissue in adults. Position of PVOi far- ther forward than in T. minimus, a line through PVOi and PVO2 passing behind PLO; PVO, mid- way between upper end of base of pectoral fin and ventral margin of body; PVO2 midway between PVOi and upper end of base of pectoral fin. PLO varying in position, usually closer to hor- izontal septum than upper end of base of pec- toral fin. PO 5-6. VLO above base of pelvic fins, closer to horizontal septum than ventral margin of body. VO level. SAO 1-2 photophore dia- meters below horizontal septum, directly above or slightly behind urogenital papilla. AO level; AOp over anterior end of infracaudal gland. Pol position variable, generally midway be- tween anterior end of infracaudal gland and end of base of anal fin (always well behind base of adipose fin), and one photophore diameter or less below horizontal septum. Prci and Prc2 level; Prcs at horizontal septum. Secondary photophores present on snout and rays of caudal fin. Length of supracaudal luminous gland 1.5-2 in length of infracaudal; sexual dimorphism not apparent ; both glands undivided and surround- ed by dark pigment. Mesopterygoid teeth rather sparsely distrib- uted. Single row of needlelike teeth on palatines. No vomerine teeth. Both jaws with needlelike teeth which bend medially (those on the anterior- most premaxillary somewhat longer). Several broad-based, anteriorly hooked teeth on posterior medial surface of dentary (not as many as in T. minimus). Two to three pairs of similar teeth projecting forward and laterally on sym- physial area of premaxillary. Taaningichthys bathyphilus is the intermedi- ate of the three species in terms of depth of occurrence, photophore development and size. It does not seem to grow larger than about 80 mm. Of the specimens examined, only one gravid female (57 mm) was found which had been cap- tured in late June. Horizontal distribution — T. bathyphilus oc- curs circumglobally within a broad zone between lat 41° N and 67°31' S (Figure 5). It appears to be more common or, perhaps, more easily cap- tured than its two congeners. Vertical distribution — The shallowest depth of capture for T. bathyphilus is 580 m (a ju- venile male, 32 mm). An adult female, 65-mm long, was captured at a depth of 675 m. Mem- bers of this species have not been taken above these depths. The maximum depth of occur- rence is not yet known. T. bathyphilus does not appear to perform daily vertical migrations. Taaningichthys paurolychnus, NEW SPECIES Figure 4 Holotype: 1 (67 mm), 17 Dec. 1969, 31° N, 119° W, Scripps Institution of Oceanography. Paratypes: 1 (68 mm), 22 Nov. 1969, 17° 47' N, 25°22' W, National Institute of Ocean- ography; 1 (87 mm), 13 Sept. 1968, 17° S, 71 FISHERY BULLETIN: VOL. 70, NO. I 86° W. Institute of Oceanolog-y, Academy of Sci- ences of the USSR, Moscow; 1 (49 mm), 20 Sept. 1961, 33° N, 17° W, Museu Municipal do Fun- chal; 1 (57 mm), 29 Jan. 1922, 19° N, 79° W, Zooloofical :\Iuseum, University of Copenhagen; 2 (79-95 mm), 17 Dec. 1969, 31° N, 119° W, 2 (65-71 mm), 10 June 1967, 35° N, 123° W, Scripps Institution of Oceanography; 1 (75 mm) , 17 Sept. 1966, 1° N, 81° W, 1 (80 mm), 15 Jan. 1969. 32° X, 120° W, Smithsonian Oceanographic Sorting Center; 1 (82 mm) , 13 Apr. 1962, 30° N, 120° W, 1 (77 mm) , 29 Mar. 1962, 35° N, 129° W, National Marine Fisheries Service. D 12-13 (11); A 13 (11-14); P 14 (13-15); V 8; gill rakers 3-4 + 1 -^ 9-10 (8-11) . total 13-15 (12-16); vertebrae 35-36; procurrent caudal rays 7 + 6-7. Mouth terminal, moderately large, JL about 1.5 in Pre P. Eye large, ED 2.2-3.2 in Pre P. A short pterotic spine directed posterolaterally. Oi)ercular margin slightly concave posterodor- sally to a level above upper end of base of pectoral fin, slightly convex posteriorly. Pectoral rays reaching base of pelvic fins; base of pectoral fin midway between ventral margin of body and hor- izontal septum. Pre V 2.1-2.3 in SL. Pre D 1.9-2.1 in SL; end of base of dorsal fin in ad- vance of origin of anal fin. Pre A 1.5-1.7 in SL. Base of adipose fin directly above, or somewhat behind end of base of anal fin; Pre Ad 1.2-1.4 in SL. Vn apparently absent. Head and body photo- phores absent. Secondary photophores present on snout and interradial membranes of caudal fin. Length of supracaudal luminous gland 1.5-2 in length of infracaudal gland; sexual dimor- phism not apparent; both glands undivided and surrounded by dark pigment. Mesopterygoid teeth rather sparsely distrib- uted. Single row of needlelike teeth on palatine. No vomerine teeth. Both jaws with needlelike teeth which bend medially (those on the anterior part of premaxillary somewhat longer). Se\^- eral broad-based, anteriorly hooked teeth on pos- terior medial surface of dentary as in T. bathy- philus. Two to three pairs of similar teeth pro- jecting forw^ard and laterally on symphysial area of premaxillary. Taaningichthys paurolychnus is the largest of the three species, the longest specimen examined measuring 95 mm. It has apparently lost its °T. minimus »T. bathvphilus *T. paurolychnus W Figure 5. — Catch localities of Tanningichihys viinimuft, T. bathyphilns, and T. paurolychnus. Capture locality for specimen taken from 67°31' S/9()°26' W not .shown. 72 DAVY: REVIEW OF LANTERNFISH GENUS Taaningichthys primary photophores, retaining only the caudal glands and the simple, presumably secondary, photophores, on the head and caudal fin. It is also the deepest dwelling of the three species. Gravid females have been captured in March- April and August-September. The smallest gravid female examined was 65 mm. Horizontal distribution — T. pam'ohjchnus is distributed circumglobally between about lat 40° N and 20° S (Figure 5). It does not appear to be as common as T. bathyphilus. Vertical distribution — T. paurolychnus has not been taken above 900 m (an adult female, 77 mm). The lower limits of its distribution are not yet known. T. paurolychnus does not appear to perform daily vertical migrations. Etymology — The name paurolychnus refers to the absence of primary photophores and the presence of limited, presumably secondary pho- tophores. It is derived from the Greek pauros meaning few, small, and lychnus meaning light. OTOLITHS OL as a percentage of SL ranges from 4.4 to 5.5% in T. minimus, 3.9 to 4.6% in T. bathy- philus, and 2.8 to 3.6% in T. paurolychnus. OH as a percentage of OL ranges from 66.7 to 77.7% in T. minimus, 72.1 to 77.9% in T. bathyphilus, and 78.3 to 91.7% in T. paurolychnus. The sulcus is more pronounced in the otolith of T. bathyphilus than it is that of T. paurolych- nus, but less so than in that of T. minimus (Fig- ure 6). The otolith of T. paurolychnus has al- most no antirostrum, and the antirostrum in T. bathyphilus is less pronounced than that in T. minimus. The posterior margin of the otolith in T. paurolychnus is nearly straight vertically, making the general outline almost square, where- as the otoliths of its two congeners are smoothly rounded posteroventrally, so that the general outline is oval. The otolith (Figure 6) of a single specimen of T. bathyphilus from the north Atlantic is differently shaped and very large for a specimen of its size (60 mm). However, no other differences in the fish were found and more material from the north Atlantic must be examined before anything further can be stated. DISCUSSION The various hypotheses and ideas regarding the function, or functions, of luminous organs of midwater fish are reviewed by Nicol (1969). The photophores within the genus Taaningich- thys show drastic reduction in terms of numbers and development. T. minimus has, relatively, the best developed photophores as well as the greatest number; these organs are seldom rubbed ofl[" unless the specimen is damaged. T. bathyphilus has fewer and less well-developed photophores which are easily rubbed ofl". T. paurolychnus has lost all primary photophores but retains simple, presumably secondary, photo- phores on the snout and caudal fin. As already mentioned, there are no indications that the members of the genus Taaningichthys undertake diel vertical migrations as most myctophids do. It may therefore be that i:)hotophores of the myc- tophid type are not selected for in a deep-water, nonmigratory fish, which would account for the reduction of these organs and, eventually, their loss. Unlike photophores, eyes are very well de- veloped in Taaningichthys, regardless of depth of occurrence. Even the deepest of the species has large, nearly binocular eyes. This may be corre- lated with the food habits of these fish. Mycto- phids, in general, feed on zooplankters, many, if not most, of which are bioluminescent. It is pos- sible therefore that Taaningichthys strongly de- pends on large, presumably highly eflFective eyes for locating and capturing its prey, which is probably not very abundant in those dark mid- water depths. Furthermore, retention of this energetically expensive visual equipment may account for the very poorly developed lateral line system. ACKNOWLEDGMENTS I thank Basil G. Nafpaktitis of the University of Southern California, Robert J. Lavenberg of the Los Angeles County Museum of Natural History, and Theodore W. Pietsch of the Uni- versity of Southern California for their critical review of the manuscript and helpful sugges- tions. Thanks are also due to John E. Fitch 73 FISHERY BULLETIN: VOL. 70, NO. I Figure 6. — Medial views of right otoliths, anterior end to the left. A. Tarnihif/iclithys miniimts, otolith 2.0 mm long, specimen 49 mm. B. T. bathyphilidi, otolith 1.8 mm long, specimen 64 mm. C. T. ixtiirolycIiriKS, otolith 1.7 mm long, specimen 87 mm. D. T. batluiphUiis, otolith 2.8 mm long, specimen (from north Atlantic) 60 mm. of the State of California Department of Fish and Game for comments concerning otoliths. I am indebted to several jieople and their in- stitutions for making si)ecimens available: Rob- ert L. Wisner, Scripps Institution of Oceanogra- phy (SIO); Julian Badcock, National Institute of Oceanography, Surrey, England; Roljert J. Lavenberg; Jorgen Niel.sen, Zoological Museum, University of Copenhagen; V. E. Becker, Insti- tute of Oceanology, Academy of Sciences of the USSR, Moscow ; Robert H. Gibbs, Jr., U.S. Na- tional Museum; Leslie W. Knapp, Smithsonian Oceanographic Sorting Center; G. E. Maul, Museu Municipal do Funchal, Madeira; Thomas Clarke, University of Hawaii; E. H. Ahlstrom and II. Geoffrey Moser, National Marine Fish- eries Service; Richard H. Backus and James E. Craddock, Woods Hole Oceanographic Institu- 74 DAVY: REVIEW OF LANTERNFISH GENUS Taamnguhthys tion; G. Palmer, British Museum (Natural History); Michel Le Gand, Office de la Re- cherche Scientifique et Technique Outre-Mer, Noumea, New Caledonia (material presented to the Los Angeles County Museum of Natural History) . Illustrations were made by Sharon Calloway, Los Angeles County Museum of Natural History, and otoliths were photographed by the Photogra- phy Department, Los Angeles County Museum of Natural History. I am grateful to my friend Lu Duffy for having typed the manuscript. LITERATURE CITED Berry, F. H., and H. C. Perkins. 1966. Survey of pelagic fishes of the California Current area. U.S. Fish Wildl. Serv., Fish. Bull. 65: 625-682. BOLIN, R. L. 1939. A review of the myctophid fishes of the Pa- cific coast of the United States and of lower Cal- ifornia. Stanford Ichthyol. Bull. 1: 89-156. 1959. Iniomi. Myctophidae from the "Michael Sars" North Atl. Deep-Sea Expedition 1910. Rep. Sci. Results "Michael Sars" North Atl. Deep-Sea Exped. 1910 Vol. 4, Part 2, No. 7, 45 p. Fraser-Brunner, a. 1949. A classification of the fishes of the family Myctophidae. Proc. Zool. Soc. London 118: 1019- 1106. Frizzell, D. L., and J. H. Dante. 1965. Otoliths of some early Cenozoic fishes of the Gulf Coast. J. Paleontol. 39: 687-718. Nafpaktitis, B., and J. R. Paxton. 1968. Review of the lanternfish genus Lampadena with the description of a new species. Los An- geles County Mus. Nat. Hist. Contrib. Sci. 138. NicOL, J. A. 1969. Bioluminescence. In W. S. Hoar and D. J. Randall (editors), Fish physiology, Vol. 3, p. 355- 400. Academic Press, New York. Parr, A. E. 1928. Deepsea fishes of the order Iniomi from the waters around the Bahama and Bermuda Islands. Bull. Bingham Oceanogr. Collect. Yale Univ. Vol. 3, Artie. 3, 193 p. Taning, a. V. 1928. Sjnopsis of the scopelids in the North At- lantic. Vidensk. Medd. Dan. Naturhist. Foren. Kjobenhaven 86: 49-69. APPENDIX Material examined Taaningichthys minimus University of Southern California, RV Velero IV- Stn. 11168, 31 July 1966, 32° N/120° W, 350 m, 10-ft IKMT, 1 (36 mm), LACM 9705. Stn. 11185, 2 Aug. 1966, 29° N/118° W, 375 m, 10-ft IKMT, 1 (47 mm), LACM 9650. International Indian Ocean Expedition, RV Anion Briiuu, Cruise III: Stn. 156, 6 Sept. 1963, 29° S/60° E, 122 m, 10-ft IKMT, 1 (45 mm), LACM 31320. University of Hawaii, Institute of Marine Biology: LACM 31574, 11 Sept. 1969, Hawaiian waters, 380 m, 6-ft IKMT, 2 (55-57 mm). LACM 31575, 30 Oct. 1969, Hawaiian waters 780 m, 6-ft IKMT, 2 (63-65 mm). LACM 31576, 13 Nov. 1969, Hawaiian waters, 575 m, 10-ft IKMT, 2 (50-60 mm). Scripps Institution of Oceanography: SIO 57-86, 12 May 1955, 29° N/125° W, 700 m, 10-ft IKMT, 1 (42 mm). SIO 62-430, 24 Aug. 1962, 29° N/130° W, 600 m, 10-ft IKMT, 1 (52 mm). SIO 68-490, 22 Sept. 1968, 29° N/178° W, no depth, 10-ft IKMT, 1 (46 mm). SIO 69-341, 27 Mar. 1969, 13° N/110° W, 1,100 m, 10-ft IKMT, 1 (42 mm). Woods Hole Oceanographic Institution: RHB stn. 1112, 17 June 1965, 22° N/70° W, 200 m, 10-ft IKMT, 1 (28 mm). RHB stn. 1735, 8 July 1968, 28° N/67° W, 870 m, 10-ft IKMT, 1 (40 mm). Zoological Institute, Academy of Sciences of the USSR, Leningrad: RV Vitiaz stn. 4885, 20 Dec. 1960, 17° S/71° E, 2,700 m, RT, 1 (27 mm). RV Vitiaz stn. 5127, 28 Oct. 1961, 13° N/154 ° W, 1,000 m, CN, 1 (48 mm). National Marine Fisheries Service, RV Horizon: Cruise H6204, stn. 100-140, 15 Apr. 1962, 28° N/ 124° W, 1,676 m, 10-ft IKMT, 1 (53.5 mm). Cruise H6204, stn. 110-160, 17 Apr. 1962, 35° N/ 124° W, 1,676 m, 10-ft IKMT. 1 (50 mm). U.S. National Museum, Ocean Acre material: Stn. 2-2N, 6 Mar. 1967, 32°26' N/63°44' W, 140 m, 6-ft IKMT, 1 (21 mm). Stn. 3-3N, 4 July 1967, 33°4' N/64°37' W, 1,060 m, 10-ft IKMT, 1 (52.5 mm). Stn. 3-4N, 4 July 1967, 33°10' N/64°45' W, 480 m, 10-ft IKMT, 1 (37 mm). Stn. 3-6N, 5 July 1967, 33°9' N/64°33'W, 250 m, 10-ft IKMT, 3 (36-45.2 mm). Stn. 3-13N, 6 July 1967, 32°54' N/64°45' W, 161 m, 10-ft IKMT, 1 (20.5 mm). Stn. 4-9A, 4 Sept. 1967, 31°52' N/63°58' W, 479 m, 10-ft IKMT, 1 (49 mm). Stn. 4-9B, 4 Sept. 1967, 31°52' N/63°58' W, 479 m, 10-ft IKMT, 1 (42 mm). Stn. 4-16C, 6 Sept. 1967, 32° N/64° 17' W, 500 m, 10-ft IKMT, 1 (41 mm). Stn. 6-7B, 26 Apr. 1967, 31°47' N/63°53' W, 155 m, 10-ft IKMT, 1 (33.5 mm). 75 FISHERY BULLETIN: VOL. 70, NO. 1 Stn. 6-15B, 28 Apr. 1967, 32°13' 10-ft IKMT, 1 (33 mm). Stn. 6-1.5P, 28 Apr. 1967, 32°13' 10-ft IKMT, 1 (34.5 mm). Stn. 6-18P. 29 Apr. 1967, 32° 14' 10-ft IKMT, 1 (30 mm). Stn 6-26.\, 30 Apr. 1967, 32°18' 10-ft IKMT, 1 (29 mm). Stn. 7-14X, 8 Sept. 1967, 32°12' 10-ft IKMT, 1 (45 mm). Stn. 7-15N, 8 Sept. 1967, 32°21' 10-ft IKMT, 1 (42 mm). N/63°51' W, 160 m, N/63°51' W, 160 m, N/63°46' W, 750 m, N/63°55' W, 200 m, N/63°25' W. 250 m, N/63°29' W, 450 m, Taatihigichthys hathyphilm University of Southern California, RV Eltanin: Stn. 947, 27 Jan. 1964, 67°31' S/90°26' W, 2,690 m, 3-m IKMT, 1 (67 mm), LACM 10424. Stn 1724, 18 July 1966, 40°06' S/149°55' W, 1,180 m, 3-m IKMT, 1 (60 mm), LACM 11247. Universitv of Southern California, RV Velero IV: Stn. 8959, 17 Oct. 1963, 33° N/119° W, 900 m, 10-ft IKMT, 1 (69 mm), LACM 6435. Stn. 10607, 10 June 1965, 33° N/119° W, 900 m, 10-ft IKMT, 1 (61 mm), LACM 6723. Stn. 966, 15 May 1964, 33° N/118° W, 750 m, 10-ft IKMT, 1 (72 mm), LACM 8525. Stn. 8238 25 Oct. 1962, 33° N/118° W, 10-ft IKMT, 1 (64 mm), LACM 9036. Stn. 9860, 25 June 1964, 33° N/118° W, 750 m, 10-ft IKMT, 1 (60 mm), LACM 9089. Stn. 11538. 21 June 1967, 32° N/118° W, 1,300 m, 10-ft IKMT, 1 (66 mm), LACM 9676. Stn. 11539, 21 June 1967, 33° N/118° W, 950 m, 10-ft IKMT, 1 (63 mm), LACM 9677. Stn. 11617, 16 Aug. 1967, 31° N/118° W, 1,130 m, 10-ft IKMT, 1 (68 mm), LACM 9682. Stn. 11312, 25 Jan. 1967, 28° N/116° W, 1,325 m, 10-ft IK.MT. 1 (58 mm), LACM 9708. Stn. 10373, 23 Feb. 1965, 33° N/118° W, 10-ft IKMT, 2 (61-68 mm), LACM 9764. Stn. 11696, 12 Oct. 1967, 32° N/118° W, 860 m, 10-ft IKMT, 1 (67 mm), LACM 9796. Stn. 11733, 8 Nov. 1967, 20° N/106° W. 1,400 m, 10-ft IKMT, 1 (62 mm), LACM 300.34. Stn. 11767, 16 Nov. 1967, 24° N/109° W, 1,500 m, 10-ft IKMT, 1 (57 mm), LACM 30045. Stn. 12066, 12 Apr. 1968, 26° N/114° W, 1,300 m, 10-ft IKMT, 1 (61 mm), LACM 30075. Stn. 12072, 14 Apr. 1968, 29° N/118° W, 750 m, 10-ft IKMT, 1 (65 mm), LACM 30079. Stn. 12184, 24 July 1968, 31° N/119° W, 820 m 10-ft IK.MT, 1 (50 mm), LACM 30271. Stn. 12597, 17 Jan. 1969, 32° N/120° W, 770 m, 10-ft IKMT, 1 (67 mm), LACM 30348. Stn. 12392, 11 Oct. 1968, 32° N/118° W, 1,110 m, 10-ft IKMT, 1 (61 mm), LACM 30403. Stn. 12349, 12 Sept. 1968, 32° N/118° W, 1,400 m, 10-ft IK.MT, 1 (67 mm), LACM 20598. Stn. 12491, 20 Nov. 1968, 29° N/118° W 910 m, 10-ft IKMT, 2 (61-65 mm), LACM 30609. Stn. 13385, 28 Oct. 1969, 28° N/118° W, 780 m, 10-ft IKMT, 1 (68 mm), LACM 30886. Smithsonian Oceanographic Sorting Center, RV Anton Brunn, Cruise III and VI: Label no. 7033, 18 Aug. 1963, 4° N/60° E, 2,120 m, 10-ft IKMT, 1 (34 mm), LACM 31292. Label no. 7057, 23 Aug. 1963, 5° S/60° E, 2,030 m, 10-ft IKMT, 1 (26 mm), LACM 31303. Label no. 7083, 6 Sept. 1963, 29° S/60° E, 1,150 m, 10-ft IKMT, 7 (44-55 mm), LACM 31320. Label no. 7173, 23 May 1964, 8° N/65° E, 2,850 m, 10-ft IKMT, 1 (44 mm), LACM 31344. Label no. 7177, 23 May 1964, 7° N/65° E, 940 m, 10-ft IKMT, 1 (36 mm), LACM 31345. Label no. 7204, 27 May 1964, 2° N/65° E, 1,250 m, 10-ft IKMT, 1 (46 mm), LACM 31358. Label no. 7217, 28 May 1964, 14° S/65° E, 2,250 m, 10-ft IKMT, 1 (53 mm), LACM 31361. Label no. 7265, 4 June 1964, 12° S/64° E, 1,930 m, 10-ft IKMT, 1 (38 mm), LACM 31375. Label no. 7273, 6 May 1964, 14° S/65° E, 880 m, 10-ft IKMT, 1 (47 mm), LACM 31376. Label no. 7305, 24 June 1964, 24° S/65° E, 3,500 m, 10-ft IKMT, 1 (57 mm), LACM 31401. Label no. 7312, 25 June 1964, 24° S/65° E, 1,100 m, 10-ft IKMT, 1 (56 mm), LACM 31404. Office de la Recherche Scientifique et Technique Outre- Mer, Noumea, Ne\v Caledonia : RV Cnridc, Cruise I, stn. 36A, 23 Sept. 1968, 0°2' N/ 137°51' W, 950 m, 10-ft IKMT, 3 (51-60 mm) , LACM 31439. RV Cnride, Cruise I, stn. 39A, 24 Sept. 1968, 0°14' N/ 138°17' W, 1,130 m, 10-ft IKMT, 1 (46 mm), LACM 31440. RV Cnride, Cruise I, stn. 69A, 29 Sept. 1968, 0°5' N/ 144°41' W, 580 m, 10-ft IKMT, 1 (32 mm), LACM 31446. RV Caride, Cruise I, stn. 74 A, 29 Sept. 1968, 0°/ 145°41' W, 820 m, 10-ft IKMT, 1 (58 mm), LACM 31448. RV Cnride, Cruise I, stn. 77A, 30 Sept. 1968, 0°25' N/ 146° 17' W, 1,110 m, 10-ft IKMT, 1 (48 mm), LACM 31450. RV Cnride, Cruise I, stn. 78A, 30 Sept. 1968, 0°2' S/ 146°29' W, 1,280 m, 10-ft IKMT, 1 (42 mm), LACM 31451. RV Caride, Cruise III, stn. 17, 7 Feb. 1969, 11°17' S/ 142°47' W, 1,050 m, 10-ft IKMT, 1 (60 mm), LACM 31459. RV Cnride, Cruise III, stn. 18, 8 Feb. 1969, 11'='7' S/ 142°35' W, 1,050 m, 10-ft IKMT, 1 (53 mm), LACM 31460. RV Cnride, Cruise III, stn. 60, 18 Feb. 1968, 0°12' S/ 139°19' W, 850 m, 10-ft IKMT, 1 (55 mm), LACM 31464. RV Caride, Cruise III, stn. 64, 19 Feb. 1969, 0°/ 140°9' W, 900 m, 10-ft IKMT, 1 (48 mm), LACM 31467. RC Caride, Cruise III, stn. 68, 10 Feb. 1969, 0°/ 140°42' W, 1,080 m, 10-ft IKMT, 1 (29 mm), LACM 31469. RV Caride, Cruise III, stn. 122, 24 Feb. 1970, 0°3'N/ 147°2' W, 1,100 m, 10-ft IKMT, 2 (45-51 mm), LACM 31478. RV Caride, Crui.se III, stn. 200, 2 March 1970, 0°1' N/ 154°14' W, 930 m, 10-ft IKMT, 1 (46 mm), LACM 31491. RV Caride, Crui.se III, stn. 204, 2 March 1970, 0°/ 154°25' W, 1,160 m, 10-ft IKMT, 1 (34 mm), LACM 31492. RV Cyclone, Cruise III, stn. 8, 4 May 1967, 2°13' S/ 169°47' E, 1,125 m, 10-ft IKMT, 1 (37 mm), LACM 31501. RV Cyclone, Cruise III, stn. 17, 5 May 1967, 4°23' S/ 169°52' E, 1,090 m, 10-ft IKMT, 1 (51 mm), LACM 31505. RV Santo, Cruise 68, stn. 6, 20 July 1968, 16°17' S/ 166°40' E, 1,395 m, 10-ft IKMT, 1 (52 mm), LACM 31528. 76 DAVY: REVIEW OF LANTERNFISH GENUS Taaningichthys U.S. National Mu.seum: Ocean Acre material: Stn. 3-2N, 4 July 1967, 33° N/64°45' W, 1,425 m, 10-ft IKMT, 1 (44 mm). Stn. 3-llN, 5 July 1967, 33° N/64°40' W, 1,920 m, 10-ft IKMT, 2 (53-58 mm). Stn. 6-lOB, 27 Apr. 1967, 31°59' N/63°43' W, 900 m, 10-ft IKMT, 1 (55 mm). Stn. 6-24N, 30 Apr. 1967, 32°13' N, 63°40' W, 750 m, 10-ft IKMT, 1 (67 mm). Stn. 7-13N, 8 Sept. 1967, 32°18' N/63°30' W, 1,500 m, 10-ft IKMT, 1 (36 mm). N/21° W, N/24° W, N/24° W, Carlsberg Foundation, Dana collections: Dana stn. 1156 VII, 25 Oct. 1921, 25° 2,000 m, S 150, 1 (37 mm). Dana stn. 1159 II, 29 Oct. 1921, 18° 2,000 m, S 150, 1 (43 mm). Dana stn. 1159 III, 29 Oct. 1921, 18° 1,500 m, S 150, 1 (35 mm). Dana stn. 1181 III, 21 Nov. 1921, 13° N/57° W, 1,500 m, S 150, 1 (28 mm). Dana stn. 1217 III, 29 Jan. 1922, 19° N/79° W, 1,500 m, S 150, 1 (45 mm). Dana stn. 1342 I, 15 May 1922, 34° N/70° W, 2,250 m, E 300, 1 (43 mm). Dayia stn. 1365 IX, 8 June 1922, 32° N/42° W, 2,500 m, E 300, 1 (55 mm). Zoological Institute, Academy of Sciences of the USSR, Leningrad: RV Lyra stn. 50, 25 Mar. 1966, 3° N/120° W, 1,000 m, CN, 1 (43 mm). RV Lyra stn. 3717, 9 Jan. 1957, 3° N/128° E, 1,250 m, CN, 1 (35 mm). RV Vitiaz stn. 4183, 6 Dec. 1958, 40° N/127° W, 675 m, CN, 1 (65 mm). RV Vitiaz stn. 4189, 7 Dec. 1958, 40° N/133° W, 1,000 m, CN, 1 (49 mm). RV Viiiaz stn. 4939, 4 Feb. 1961. 9° N/87° E, 1,000 m, CN, 1 (39 mm). National Institute of Oceanography, Surrey, England: NIO stn. 4687, 30 Aug. 1961, 29°57' N/32°3' W, 800 m, IKMT, 1 (50 mm). NIO stn. 4746, 30 Sept. 1961, 29°59' N/22°56' W, 1,100 m, IKMT, 2 (38-44 mm). NIO stn. 5799, 19 Oct. 1965, 28°9' N/14°9' W, 675 m, IKMT, 1 (48 mm). NIO stn. 5810, 7 Nov. 1965, 28°4' N/13°51' W, 800 m, IKMT, 1 (28 mm). NIO stn. 5813, 10 Nov. 1965, 28°5' N/14°ll' W, 950 m, IKMT, 1 (56 mm). NIO stn. 6687, 7 Mar. 1968, 20°37' N/22°56' W, 1,000 m, RMT8, 1 (56 mm). NIO stn. 7072, 30 Oct. 1969, 20°27' N/25°32' W, 1,000 m, RMT8, 2 (32-55 mm). NIO stn. 7079, 3 Nov. 1969, 17°40' N/27°6' W, 1,000 m, RMT8, 1 (24 mm). NIO stn. 7089 #54, 22 Nov. 1969, 17°47' N/25°22' W, 1,000 m, RMT8, 2 (54-77 mm). NIO stn. 7089 #55, 22 Nov. 1969, 17°47'N/25°22' W, 2,000 m, RMT8, 1 (46 mm). British Museum of Natural History: Rosaura collection, 26 June 1969, 17° N/86° W, 1,100 m, S 200, 1 (57 mm). Rosaura collection, 26 June 1969, 11° N/76° W, 1,200 m, S 200, 1 (46 mm). National Marine Fisheries Service, RV Horizon: Cruise H6204 stn. 120 • 70, 23 Apr. 1962, 26° N/117° W, 1,676 m, 10-ft IKMT, 1 (65 mm). Woods Hole Oceanographic Institution : RHB stn. 977, 26 Feb. 1963, 1° S/27° W, 10-ft IKMT, 1,100 m, 1 (65 mm). RHB stn. 979, 28 Feb. 1963, 3° S/29° W, 10-ft IKMT, 1,100 m, 1 (57 mm). RHB stn. 1603, 6 Oct. 1967, 39°46' N/70°30' W, 10-ft IKMT, 1,000 m, 1 (65 mm). Taauingichthys paurolychniis University of Southern California, RV Velero IV: Stn. 10675, 28 Aug. 1965, 29° N/118° W, 1,625 m, 10-ft IKMT, 1 (44 mm), LACM 9350. Stn. 11187, 2 Aug. 1966, 29° N/118° W, 1,720 m, 10-ft IKMT, 1 (87 mm), LACM 9567. Stn. 11257, 21 Oct. 1966, 29° N/118° W, 940 m, 10-ft IKMT, 1 (77 mm), LACM 9408. Stn. 11628, 18 Aug. 1967, 32° N/119° W, 1,300 m, 10-ft IKMT, 1 (33 mm), LACM 9693. Stn. 12331, 24 Aug. 1968, 29° N/118° W, 1,100 m, 10-ft IKMT, 1 (55 mm), LACM 30284. Stn. 12340, 26 Aug. 1968, 32° N/118° W, 1,130 m, 10-ft IKMT, 1 (17 mm), LACM 30591. Stn. 12475, 18 Nov. 1968, 28° N/119° W, 900 m, 10-ft IKMT, 1 (42 mm), LACM 30606. Stn. 12483, 19 Nov. 1968, 28° N/119° W, 2,080 m, 10-ft IKMT, 3 (52-81 mm), LACM 30382. Stn. 12592, 15 Jan. 1969, 32° N/120° W, 1,950 m, 10-ft IKMT, 3 (23-82 mm), LACM 30429. Stn. 12593, 16 Jan. 1969, 32° N/120° W, 1,920 m, 10-ft IKMT, 1 (91 mm), LACM 30430. Stn. 12594, 16 Jan. 1969, 32° N/120° W, 1,250 m, 10-ft IKMT, 1 (79 mm), LACM 30431. Stn. 12786, 16 Mar. 1969, 32° N/118° W, 1,350 m, 10-ft IKMT, 1 (66 mm), LACM 30423. Stn. 12791, 17 Mar. 1969, 32° N/118° W, 1,200 m, 10-ft IKMT, 1 (64 mm), LACM 30428. Smithsonian Oceanographic Sorting Center, RV Anto7i Briuni, Cruise III and VI: Label no. 7057, 23 Aug. 1963. 4° S/60° E, 2,030 m, 10-ft IKMT, 1 (51 mm), LACM 31303. National Marine Fisheries Service, RV Horizon: Cruise H6204 stn. 60.60, 26 Mar. 1962, 37° N/123° W, 1,863 m, IKMT, 1 (86 mm). Cruise H6204 .stn. 60.140, 29 Mar. 1962, 35° N/129° W, 1,863 m, IKMT, 1 (77 mm). Cruise H6204 stn. 80.90, 18 Mar. 1962, 33° N/123° W, 2,234 m, IKMT, 1 (29 mm). Cruise H6204 .stn. 100.60, 13 Apr. 1962, 31° N/119° W, 1,676 m, IKMT, 1 (68 mm). Cruise H6204 stn. 100.80, 13 Apr. 1962. 30° N/120° W, 1,676 m, IKMT, 1 (82 mm). Scripps Institution of Oceanography: SIO 54-95, 23 June 1954, 23° N/119° W, 2,500 m, 10-ft IKMT, 1 (49 mm). SIO 60-283, 12 Aug. 1960. 28° N/135° W, 3,000 m, 10-ft IKMT, 1 (44 mm). SIO 60-284, 13 Aug. 1960, 29° N/132° W, 3,000 m, 10-ft IKMT, 1 (71 mm). SIO 64-11, 3;) Jan. 1964, 24° N/113° W. 5,300 m, 10-ft IKMT, 1 (78 mm). SIO 66-31, 5 Apr. 1966, 29° N/117° W, 4,000 m, 10-ft IKMT, 1 (84 mm). SIO 67-52, 22 Apr. 1967, 30° N/117° W, 4,000 m, 10-ft IKMT, 1 (80 mm). 77 FISHERY BULLETIN: \0L. 70, NO. I SIO r>7-102, 10 June 1967, 35° N/123° W, 2,200 m, 10-ft Zoological Institute, Academy of Sciences of the USSR, IK.MT, 2 (65-71 mm). Leningrad: SIO 70-19, 17 Dec. 1969. 31° N/119° W, 4,000 m, 10-ft RV Akademik Kurchntov stn. 233, 13 Sept. 1968, IKMT, 1 (67 mm). 17° S/86° W, 2,000 m, CN, 1 (87 mm), 39908. SIO 70-20, 17 Dec. 1969:31° N/119° W, 4,000 m, 10-ft ^ , . , ,, IKMT '^ (79-95 mm) Zoological Museum, University of Copenhagen: ' " Dana stn. 1217 I, 29 Jan. 1922, 19° N/79° W, 2,000 m, E 300, 1 (57 mm), P2330669. National Institute of Oceanography, Surrey, England: Museu Municipal do Funchal, Madeira: NIO .sta. 7089 #55, 22 Nov. 1969. 17°47' N/25°22' W, RV Discovery 4742, 20 Sept. 1961, 32°42' N/16°32' W, RMT8, 3 (45-73). 1,700 m, IKMT, 1 (49 mm), MMF 22115. 78 SOME LIFE HISTORY CHARACTERISTICS OF COHO SALMON OF THE KARLUK RIVER SYSTEM, KODIAK ISLAND, ALASKA Benson Drucker^ ABSTRACT This paper contains data on some life history characteristics of the coho salmon of the Karluk River system, Kodiak Island, Alaska: age, fecundity, length, and egg size of adults; and migration charac- teristics, age, and size of smolts. The greater age at maturity of Karluk coho salmon (4 and 5 years) because of the longer freshwater residence of the juveniles is unique among reported North American stocks and may result in greater freshwater mortality but less marine mortality because the smolts are larger when they enter the ocean. Fecundity of Karluk coho salmon also differs from that reported for other North American stocks in that they are extremely fecund — more similar to Asiatic stocks of the Kamchatka Peninsula. Coho salmon, Oncoi^hynchus kisutch, are widely distributed along the Pacific coast of North America and occur in commercially harvestable quantities from northern California to north- western Alaska. About one-third of the total North American commercial catch comes from Alaska waters, where from 1960 to 1968 the average annual catch of 16 million pounds was valued at almost $3.5 million to the fishermen." The amount of biological research on coho salmon in Alaska is small, and published scientific re- ports on Alaska coho salmon stocks are very few. In this paper I present data on some life his- tory characteristics of the coho salmon of the Karluk River system. This system is located on the southwest side of Kodiak Island, Alaska, at approximately lat 57° N and long 154° W and includes Karluk Lake, tributaries to the lake. Thumb and O'Malley Lakes, and Karluk River (Figure 1). Information is presented on age, fecundity, length, and egg size of coho salmon adults; and migration characteristics, age, and ^ National Marine Fisheries Service, Auke Bay Fish- eries Laboratory, Auke Bay, Alaska 99821 ; present ad- dress: National Marine Fisheries Service, Technical Advisory Division, Interior Building, Washington, D.C. 20235. " Nelson, Richard C. 1968. Alaska catch and pro- duction, commercial fisheries statistics. Alaska Dep. Fish Game, Stat. Leafl. 17. 29 p. (Unpublished.) Figure 1. — The Karluk River system, Kodiak Island, Alaska. Manuscript accepted July 1971. FISHERY BULLETIN: VOL. 70, NO. 1, 1972. 79 FISHERY BULLETIN: VOL. 70, NO. I size of coho salmon smolts. These life history features of Karluk coho salmon are compared with those reported for various coho salmon stocks from other areas on both the Asiatic and North American sides of the Pacific Ocean. In addition, the effect of a prolonged juvenile fresh- water residence, a feature unique to the Karluk system, on freshwater and marine survival is discussed. All of the data on Karluk coho salmon were collected incidentally during studies of sockeye salmon. 0. nerka, the dominant salmon species in the Karluk system; much of the in- formation on other Alaska stocks is from un- published administrative and progress reports. Because I intend to discuss differences between coho salmon in the Karluk system and those in other areas, a description of general features of the life cycle of coho salmon stocks is appropri- ate. Typically, the adults enter streams and rivers from late summer to November and spawn in late fall and early winter. Some Asiatic stocks, however, spawn as late as mid-March (Smirnov, 1960). The progeny emerge as fry in the spring following spawning and reside in rivers or lakes for 1 or 2 years before going to sea as smolts. In some areas the seaward smolt migration begins in late winter (Chapman, 1961; Smoker, 1953), but in most areas it takes place from April to August (Godfrey, 1965). The salmon grow rapidly in the ocean, and the adults return to the streams and rivers to spawn 12 to 18 months later. However, a significant per- centage of male coho salmon, particularly in their southern range of distribution in North America (California), mature precociously (6 to 9 months after they enter salt water) and return to spawn the same year that they mi- grated to sea (Shapovalov and Taft, 1954). These fish are known as jack salmon. METHODS The data for adult coho salmon of the Karluk system were obtained from fish from the 1966 escapement that were captured at the adult counting weir or caught by sport fishermen at the outlet of Karluk Lake about 300 yards up- stream from the weir site. All fish were mea- sured for length (mideye to fork of tail) with a caliper to the nearest millimeter. Mideye-fork length was used because of morphological changes that occur as the fish matures, partic- ularly the elongation of the snout. Ovaries for fecundity samples were removed from all females and were preserved in 10% Formalin solution for at least 48 hr. The eggs were then hand-counted to get total egg counts. The diameters of some eggs from the fecundity samples were measured. These eggs were removed directly from the ovary, water hardened, and placed in Stockard's solution. The diameters were then measured with a vernier measuring microscope calibrated to 0.01 mm. The ages of adult fish were deter- mined by reading scales that had been taken halfway between the lateral line and the poster- ior insertion of the dorsal fin. The data for smolts were obtained from fish captured in 1956, 1965, and 1968 in fyke nets fished on the downstream side of the adult counting weir. Fork lengths were taken to the nearest millimeter with a steel millimeter ruler, and weights were taken to the nearest tenth of a gram on a triple-beam balance. As with the adults, the ages of smolts were determined from scales taken halfway between the lateral line and the posterior insertion of the dorsal fin. AGE OF COHO SALMON The average age composition of coho salmon for several systems in northern and southern latitudes of North America and Asia is shown in Table 1. The differences from the northern to southern latitudes in age composition is sim- ilar to that noted by Marr (1943) and possibly represent a geographic cline. The Karluk system had three freshwater age classes," two of which were decidedly predom- inant (Table 1). The three age classes, 43, 54, and 65, designate fish that went to sea in their third, fourth, and fifth years of life and returned to spawn after being at sea for about 1 year. "Age classes are designated according to the system developed by Gilbert and Rich (1927). A 43 coho salmon is in its fourth year of life. It went to sea as a smolt at the beginning of its third year, having spent two growing seasons in fresh water. 80 DRUCKER: COHO SALMON OF KARLUK RIVER SYSTEM Table 1. — Age class composition of stocks of coho salmon from North America and Asia, arranged geographically from north to south. Area Percent coho salmon in age class— 2i 2a 33 33 4i 4a 43 53 65 Reference North America Nome River, Alaska Unalakleet River, Alaska Yukon River, Alaska Yukon River, Alaska Cook Inlet River, Alaska Resurrection Boy, Alaska Bear Creek, Alaska Dairy Creek, Alaska Mendenhall River, Alaska Hood Bay Creek, Alaska Karluk River, Alaska Sashin Creek, Alaska Port Herbert, Alaska Stikine River, Alaska Chignik River, Alaska Ketchikan River, Alaska Quatsino Bay, British Columbia Fraser River, British Columbia Georgia Strait, British Columbia West coast, Vancouver Island, British Columbia Langara Island, Georgia Strait, British Columbia Columbia River, Wash. Waddell Creek, Calif. Asia East coast of Kamchatka, USSR: Kamchatka River Lake Ushki Kyrganik River Paratunko River Avachin Gulf, Solevarko Bay Kalyger River West coast of Kamchatka, USSR: Kikhchik River Bolshaya River Ozernaia River Kukhtui River (Okhotsk) 2.6 1.5 1.3 0.1 2.0 0.6 O.I 6.1 18.4 1.2 Tr. Tr. Tr. Tr. Tr. Tr. 29.4 37.9 55.6 38.7 40.0 30.3 27.1 383.3 12.0 46.5 8.0 0.7 0.4 0.6 18.0 20.0 45.2 23.2 70.8 95.0 96.5 97.1 97.5 0.2 97.9 83.9 81.6 55.7 4.3 93.6 57.4 80.2 27.5 Tr. 0.3 Tr. Tr. 100.0 69.3 100.0 32.8 Tr. Tr. Tr. Tr. Tr. Tr. Tr. Tr. 1.0 70.6 62.1 44.4 58.1 60.0 68.8 71.1 16.7 80.0 47.5 56.9 77.0 76.0 51.9 72.4 29.2 0.1 3.0 0.2 0.4 0.4 0.7 0.2 0.9 9.7 43.1 92.9 6.4 42.6 19.8 72.5 30.1 67.2 3.2 0.9 1.8 1.9 6.0 41.7 5.0 4.0 4.4 1.4 2.8 Godfrey (1965) Godfrey (1965) Godfrey (1965) Gilbert (1922) Godfrey (1965) Logan (1963,i 1964=) Logan (1964) = Logan (1964) = (') Armstrong (1970) Present study Crone (1968) Crone (1968) Godfrey (1965) Israel (1933) Godfrey (1965) Godfrey (1965) Godfrey (1965) Godfrey (1965) Godfrey (1965) Pritchard (1940) Marr (1943) Shapovalov and Toft (1954) Godfrey (1965); Gribanov (1948) Gribanov (1948) Gribanov (1948) Gribanov (1948) Gribanov (1948) Gribanov (1948) Gribanov (1948) Godfrey (1965); Gribanov (1948); Semko (1954) Gribanov (1948) Godfrey (1965) In Dingell-Johnson project report, 1962-63, Vol. 4: In Dingell-Johnson project report, 1963-64, Vol. 5: 175-194, Alaska 133-151, Alaska 1 Logan, Sidney M. 1963. Silver salmon studies in the Resurrection Bay area Dep. Fish Gome, Sport Fish Div., Juneau, Alaska. (Unpublished.) = Logan, Sidney M. 1964, Silver salmon studies in the Resurrection Boy area Dep. Fish Game, Sport Fish Div., Juneau, Alaska. (Unpublished.) 3 The high percentage of age 3^ fish is atypical and not representative of Resurrection Bay streams. Dairy Creek juveniles ore reared in a brackish water lagoon rather than in the stream itself, resulting in 1-year smolts. (Personal communication, Sidney Logan, Area Management Biologist, Alaska Department of Fish and Game, Soldotna, Alaska, March 15, 1971.) * Collected by author in 1966. Fifty-seven percent of the Karluk fish had mi- grated in their third year and 42 9f in their fourth year; only l^c had migrated in their fifth year (Table 1). Although the freshwater residence of fish in the Karluk escapement var- ied from 2 to 4 years, Karluk coho salmon, like those in all other systems, returned to spawn after being at sea for 12 to 18 months. The presence of large numbers of fish( 42%) that had spent 3 years in freshwater residence (age class 64) is unique to the Karluk system. Fish of age class 64 have been found in other Alaska river systems, i.e., the Yukon River (Gil- bert, 1922), Resurrection Bay and Bear Creek (see footnotes 1 and 2, Table 1) , Hood Bay Creek (Armstrong, 1970), Sashin Creek and Port Herbert (Crone, 1968), and Chignik River (Israel, 1933) ; but the proportion of 64 fish in the total runs to these systems is small — usually less than 5% (Table 1). The age composition of stocks of coho salmon from systems on the Kamchatka Peninsula, 81 FISHERY BULLETIN: VOL. 70. NO. 1 USSR, is similar to that of coho salmon in the northern areas of the west coast of North Amer- ica. The main age classes are 82 and 43 (Table 1 ) . The ratio of one age class to the other varies, however, from year to year and from area to area (Gribanov, 1948; Semko, 1954). An additional comparison of the age compo- sition of coho salmon from northern to southern latitudes is shown in Figure 2, which gives the percent age composition of the major age classes from five geographical areas along the west coast of North America. In California, the southern limit of the range of coho salmon, the major age class is 82, but jack salmon (age 22) contribute significantly to the runs. The 82 age class is still dominant in Washington, but the number of jack salmon is less and 43 fish are starting to appear. North of Washington to central British Colum- bia, more than 95 ^r of the fish are age 82, and there are only traces of other age classes, mainly the 43 class. From central British Columbia and northward through Alaska, the primary age class is 43; 82 fish are the secondary class and 54 fish are found in small numbers. In Alaska, the increase in total age is the result of juvenile coho salmon residing an additional year in fresh water before migrating to sea. Possible exceptions to the dominance of the 43 age class in Alaska are the Ketchikan River, Dairy Creek, Yukon River, and Karluk River systems (Table 1). In the first three river systems, 82 fish are the dominant age class and 43 fish the secondary class. The sizes of the samples from these systems were small, however (less than 25 fish). In the Kar- luk system, although 43 fish were dominant, 64 fish rather than 82 fish were the secondary age class (Figure 2). 100 "50 CALIFORNIA WASHINGTON "n 11^ I ^11 BRITISH COLUMBIA ALASKA KARLUK RIVER i h h \ \ S h S h S ^ S ^ AGE CLASS Figure 2. — Average age composition of coho salmon runs along the west coast of North America by geographical area (minor age classes omitted). The presence of older fish (bi) in northern latitudes may be a result of the juveniles being reared in lakes rather than rivers. Typically, coho salmon spawn in rivers or tributaries to rivers and the emerging fry reside in these areas until they migrate to sea. In contrast, in some Alaska river systems where 54 fish are part of the run (Table 1 and Figure 2), the juveniles migrate from spawning grounds to lakes before migrating to sea. It appears that some of the juveniles that reside in lakes (lake type) go to sea at an older age than those that reside in rivers (river type)/ NUMBER AND SIZE OF EGGS In this section, information is presented on fecundity (number of eggs contained in a fe- male) as a function of latitude and length, the relative numbers of eggs in right and left ovaries, and egg size in relation to length and fecundity. Fecundity and factors related to it form the basis for determining the reproductive potential of a spawning stock and subsequent survival from egg to young. Knowledge of variations in fe- cundity and egg size is of increasing importance in fish stocking and fish rehabilitation programs. Size of egg may be useful in predicting the con- dition, or hardiness, of developing fry. Because the fecundity of fish differs among geographic areas, the reproductive potential must be deter- mined for each stock. FECUNDITY AS A FUNCTION OF LATITUDE The average fecundity for both North Amer- ican and Asiatic stocks of coho salmon is con- siderably higher in fish from northern latitudes than in those from southern latitudes (Table 2 and Figure 8) , Coho salmon from Alaska river systems (with the exception of the two small samples from Port Herbert and Sashin Creek) * Personal communication, 1969, Charles J. DiCos- tanzo, Chief, Salmon Investigations, National Marine Fisheries Service, Auke Bay Fisheries Laboratory, Auke Bay, Alaska 99821. 82 DRUCKER: COHO SALMON OF KARLUK RIVER SYSTEM Table 2. — Average fecundity of coho salmon stocks from North American and Asiatic river systems, arranged geographically from north to south. Area Lati- tuda Average number of eggs Reference North America Swonson River, Alaska Bear Creek, Alaska Dairy Creek, Alaska Karluk River, Alaska Pasogshak River, Alaska Scshin Creek, Alaska Port Herbert, Alaska Namu River, British Columbia Fraser River, British Columbia Nile Creek, British Columbia Cultus Lake Hatchery, British Columbia Port John, British Columbia Cowichan River, British Columbia Oliver Creek, British Columbia Beadnell Creek, British Columbia Seattle, Wash. Winter Creek, Wash. Fall Creek, Alsea River, Ore. Scott Creek, Calif. Asia East coast of Kamchatka, USSR: Ushki Hatchery Kamchatka River Paratunka River West coast of Kamchatka, USSR: Bolshoya River Sakhalin Island, USSR: Tymi River 61° N 3,378 Engel (1966)i 60° N 4,115 Lawler (1963,= 1964^) 60° N 4,177 Engel (1965),*; Lawler (1963) 57° N 4,706 Present study 57° N 4,510 Marriott (1968) = 56° N 2,868 Crone (1968) 56° N 2,565 Crone (1968) 54° N 3,002 Foerster and Pritchard (1936) 53° N 3,152 Foerster and Pritchard (1936) 49° N 2,310 Wickett (1951) 49° N 2,300 Foerster and Ricker (1953) 49° N 2,313 Hunter (1948) 48° N 2,329 Neave (1948) 48° N 2,267 Foerster (1944) 48° N 2,789 Foerster (1944) 47° N 3,141 Alien (1958) 47° N 2,447 Salo and Bayliff (1958) 44° N 1,983 Koski (1966) 37° N 2,336 Shapovalov and Toft (1954) 56° N 5,282 Gribanov (1948) 56° N 4,883 Gribanov (1948) 53° N 4,350 Gribanov (1948) 53° N 4,638 Semko (1954) 52° N 4,570 Smirnov (I960) 1 Engel, Larry J. 1966. Egg-take investigations in Cook Inlet drainage and Prince William Sound. In Federal aid in fish restoration, 1965-66 progress report. Vol. 7: 109-116, Alaska Dep. Fish Game, Sport Fish Div., Juneau, Alaska. (Unpublished.) - Lav^ler, Robert E. 1963. Silver salmon egg taking investigations in Cook Inlet drainage. In Ding- ell-Johnson project report, 1962-63, Vol. 4: 161-173, Alaska Dep. Fish Game, Sport Fish DIv., Juneau, Alaska. (Unpublished.) ^ Lawler, Robert E. 1964. Egg take investigations in Cook Inlet and Prince William Sound. In Dingell-Johnson project report, 1963-64, Vol. 5: 123-132, Alaska Dep, Fish Game, Sport Fish Div., Juneau, Alaska. (Unpublished.) * Engel, Larry J. 1965. Egg take investigations in Cook Inlet drainage and Prince William Sound. In Dingell-Johnson project report, 1964-65, Vol. 6: 155-163, Alaska Dep. Fish Game, Sport Fish Div., Ju- neau, Alaska. (Unpublished.) = Marriott, Richard A. 1968. Inventory and cataloging of the sport fish waters In southwest Alaska. In Federal aid in fish restoration, 1967-68 progress reporf. Vol. 9: 81-93. Alaska Dep. Fish Game, Sport Fish Div., Juneau, Alaska. (Unpublished.) are more fecund than coho salmon from more southerly areas in North America (Figure 3). Stocks of coho salmon from Asiatic river systems are extremely fecund, even more so than North American stocks in more northerly latitudes. The high fecundity of Karluk River coho salmon more closely resembles the fecundity of Asiatic stocks than North American ones. Contrary to these findings for coho salmon, Rounsefell (1957) suggests that for the genus Oncorhynchus, salmon in southern latitudes may be more fecund than those in northern latitudes because of ". . . the higher age at maturity, and therefore slower growth rates, from south to north." Rounsefell found that the amount of time juvenile sockeye salmon spent in fresh water had no effect on fecundity, but the amount of time the adults spent at sea did have an effect: adult sockeye salmon that spent 2 years at sea had higher fecundity counts than fish of the same size that spent 3 years at sea. With coho salmon, however, the greater age at maturity is not due to increased time in the ocean but to increased time in fresh water. 83 FISHERY BULLETIN: VOL. 70, NO. 1 63' 59' 57' 55' 53' ui5|. o 3 <49' -I I a 47' o z 45' 43"- 39' 37' 35' 0' © \F © © ®® ® © ® @ ® ® ® ®^A ® ® ® 3 4 5 6 7 8 9 10 I I 12 13 14 15 16 17 IB 19 20 21 22 23 24 _1 SWftNSON RIVER, ftLASKA BEAR CREEK. ALASKA DAIRY CREEK .ALASKA KARLUK RIVER, ALASKA PASA6SMAK RIVER, ALASKA PORT HERBERT, ALASKA SASHIN CREEK, ALASKA KAMCHATKA RIVER, USSR USHKI HATCHERY, USSR NAMU RIVER. BRITISH COLUMBIA ERASER RIVER, BRITISH COLUMBIA PARATUNKA RIVER. USSR BOLSHAYA RIVER. USSR TYMI RIVER, USSR CULTUS LAKE, BRITISH COLUMBIA NILE CREEK. BRITISH COLUMBIA PORT JOHN. BRITISH COLUMBIA OLIVER CREEK, BRITISH COLUMBIA COWICHAN CREEK, BRITISH COLUMBIA BEADNELL CREEK, BRITISH COLUMBIA MINTER CREEK, WASHINGTON SEATTLE, WASHINGTON ALSEA RIVER, WASHINGTON SCOTT CREEK, CALIFORNIA _1_ _1_ 1,000 2,000 3,000 4,000 NUMBER OF EGGS 5,000 6,000 Figure 3. — Average fecundity of various stocks of echo salmon from North America and Asia. fork length by the method of least squares. The result may be expressed by the equation y = —7,503.55 + 195.51X, where Y is the esti- mate of number of eggs and X is the mideye- fork length of female salmon (Figure 4). The mean number of eggs for the sample was 4,706 (range 1,724 to 6,906); the mean length was 62.1 cm (range 46.6 to 69.8 cm). 2 3,000 Figure 4. — Relation of fecundity to length of coho salmon sampled at Karluk v^reir, 1966. FECUNDITY AS A FUNCTION OF LENGTH The presence of a positive relation between fecundity and length in the genus Oncorhynchus is well known (Gilbert and Rich, 1927; Foerster and Pritchard, 1941; Allen, 1958; Hartman and Conkle, 1960). For fish in general, the relation of fecundity to length is logarithmic (Y = aX^) over a wide range of lengths. For salmon, how- ever, the narrow range in length at maturity permits this relation to be described adequately by a straight line of the foi'm Y = a + bX (Foer- ster and Pritchard, 1941; Rounsefell, 1957). I counted the total number of eggs in 49 coho salmon from the Karluk River and calculated the relation between number of eggs and mideye- It is difficult to determine if the high fecun- dity of coho salmon of the Karluk system (Fig- ure 3) is due to greater fecundity per unit length or simply to the fact that coho salmon from Kar- luk are very large. The average lengths of fe- male coho salmon from various spawning streams along the Pacific coast of North Amer- ica are quite variable and do not seem to follow any set geographic pattern (Table 3). More- over, Karluk fish were measured from mideye to fork of tail, and direct comparisons of lengths with coho salmon from other areas are difficult to make because of variability in the types of measurements used. For instance lengths re- ported from areas other than Karluk include tip of snout to fork of tail (fork length), tip of snout to tip of tail (total length), and tip of snout to base of tail (standard length). 84 DRUCKER: COHO SALMON OF KARLUK RIVER SYSTEM Table 3.— Average lengths of female coho salmon from river systems along the Pacific coast of North America and Asia, arranged geographically from north to south. • Area North America Yukon River, Alaska Swanson River, Alaska Resurrection Boy, Alaska Dairy Creek, Alaska Brooks River, Alaska Karluk River, Alaska Sashin Creek, Alaska Port Herbert, Alaska Nomu River, Britisli Columbia Fraser River, British Columbia Seattle, Wash. Minter Creek, Wash. Columbia River, Wash. Deer Creek, Oreg. Flynn Creek, Oreg. Needle Branch, Oreg. Scott Creek, Calif. WadcJell Creek, Calif. Asia East coast of Kamchatka, USSR: Kamchatka River Kyrganik River Kalyger River Avachin Gulf Paratunka River West coast of Kamchatka, USSR: Kikhchik River Bolshaya River Ozernaia River Average length Fork Total Standard Mideye- fork Refer 63.4 62.0 67.2 __ 72.8 — 58.8 70.5 " 67.8 __ _. 69.0 64.0 63.4 ~ 74.6 70.7 __ 69.4 67.6 66.3 — _ 63.9 60.9 68.1 61.0 __ 55.4 __ 59.4 — 58.6 57.4 __ 62.6 __ 62.1 62.1 Gilbert (1922) Engel (I966)i Logon (1965) = Engel (1965)^ (') Present study Crone (1968) Crone (1968) Foerster and Pritchard (1936) Foerster and Pritchard (1936) Allen (1958) Salo (1955) Marr (1943) Koski (1966) Koski (1966) Koski (1966) Shapovalov and Toft (1954) Shapovalov and Taft (1954) Gribanov (1948) Gribanov (1948) Gribanov (1948) Gribanov (1948) Gribanov (1948) Gribanov (1948) Gribanov (1948) Gribanov (1948) ' See footnote 1, Table 2. 2 Logan, Sidney M. 1965. Silver salmon studies in the Resurrection Boy area. In Dingell-Johnson project report, 1964-65, Vol. 6: 129-145, Alaska Dep. Fish Game, Sport Fish Div., Juneau, Alaska. (Unpublished.) ■' Lowler, Robert E. 1964. Egg take investigations in Cook Inlet and Prince William Sound. In Dingell-Johnson project report, 1963-64, Vol. 5: 123-132, Alaska Dep. Fish Gome, Sport Fish Div., Juneau, Alaska. (Unpublished.) * Eicher, George J., Jr. The effects of laddering a falls in a salmon stream. National Marine Fisheries Service, Auke Bay Fisheries Laboratory, Auke Bay, Alaska, 5 p. (Unpublished.) NUMBER OF EGGS IN RIGHT AND LEFT OVARIES The numbers of eggs in the right and left ovaries of the genus Oncorhynchtis are usually quite variable. Rounsefell (1957) noted that although the rate of maturation of eggs from Karluk Lake sockeye salmon was the same in both ovaries of the same fish, the number of eggs in each ovary varied. Eguchi and his co-workers (Rounsefell, 1957) found no significant diflFer- ences in the numbers of eggs in the two ovaries in chum salmon, 0. keta, in Japanese waters. Helle (1970) found the same lack of a significant difference in a sample of pink salmon, 0. gor- buscha, from Olsen Bay, Alaska, in 1963. Sock- eye salmon from Brooks Lake, Alaska, in 1957 and 1958 and from Karluk Lake in 1958 had more eggs in the left ovary than in the right (Hartman and Conkle, 1960). At Bare Lake, Alaska, sockeye salmon had more eggs in the right ovary than in the left (Nelson, 1959). I compared the numbers of eggs from the right and left ovaries of Karluk River coho salmon (Table 4) by means of a ^ test for paired ob- servations. The differences between the num- bers of eggs in the right and left ovaries were significant (t = 2.60; df = 31; P = 0.05). In 31 fecundity samples, 71% had more eggs in the right ovary than the left. I could not find comparable information on comparisons between the numbers of eggs in the ovaries of coho salmon from other areas. 85 FISHERY BULLETIN: VOL. 70, NO. 1 Table 4. — Numbers of eggs in right and left ovaries from coho salmon collected at the outlet to Karluk Lake, 1966. Sample number 1 2 3 4 5 6 7 8 9 10 11 12 13 U 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Average Mideye-fork length (cm) Number of eggs Righl ovary 50.9 61.4 65.0 56.9 62.7 65.6 69.8 64.7 62.6 60.2 59.3 60.2 62.3 61.2 60.1 63.5 64.5 67.2 60.2 65.2 64.0 65.1 66.0 62.9 65.8 64.6 63.6 66.7 63.1 61.9 64.8 1,640 2,265 3.005 2,213 2,322 3,001 3,559 2,258 2,546 2,433 2,243 2,331 2,481 2,620 2,067 2,581 2,473 2,604 2,044 2,824 2,608 2,491 2,926 2,266 3,047 2,726 2,721 3,104 2,981 2,176 2,340 Left ovary 1,403 1,918 2,876 2,083 2,337 3,147 3,347 1,884 2,501 2,220 2,225 2,161 2,283 2,539 2,000 2,221 2,546 3,233 1,813 2,697 2,659 2,501 3,174 2,280 2,878 2,579 2,563 2,997 2,843 2,125 2,521 Total 3,043 4,183 5,881 4,296 4,659 6,148 6,906 4,142 5,047 4,653 4,468 4,492 4,764 5,159 4,067 4,802 5,019 5,837 3,857 5,521 5,267 4,992 6,100 4,546 5,925 5,305 5,284 6,101 5,824 4,301 4,861 r.oo • 680 ' 660 - . • • • 6.40 is.zo ;;;6.oo - • • • • • • • • • • • UJ Z 5 5.80 o - • / • 13 £5.60 - • 25.40 • S 5.20 - 500 - • 4.60 - 0.0 U — 1— 1 1 1 1 _l fill ■ III 63.0 2,545 2,469 5,015 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 MIOEyE-FORK length Icm) Figure 5. — Relation of mean egg diameter to mideye- fork length of female coho salmon, Karluk River, 1966. length of the fish. Allen (1958) in his studies of coho salmon in Green River, Wash., also found no relation. The average diameters of eggs were plotted against number of eggs in individual fish to de- termine if a relation existed between the fecun- dity of a female and the size of her eggs (Figure 6). For 24 females, the egg diameter ranged RELATION OF EGG SIZE TO LENGTH AND TO FECUNDITY The average diameter of eggs obtained from the fecundity samples from Karluk River was plotted against the length of the female coho salmon from which the samples were taken ( Fig- ure 5) to determine if there was a relation be- tween the size of a female and the size of her eggs. The size of eggs increases as they mature, and so the eggs used had to be in the same stage of maturation. I therefore selected only females beginning to show secondary sexual character- istics and containing eggs that could not be readily expressed from the body cavity. For 25 females the eggs varied in size from 4.92 to 6.88 mm. (mean 6.11 mm) ; lengths varied from 50.4 to 69.8 cm (mean 62.0 cm). No relation was found between the size of the egg and the • 6.80 - 6.60 - • . 640 - • • • • §6.20 a: - • • • • • ;^6.00 - • • • UJ Z < 580 _ • o • • o • S5 60 - z • uj5.40 - Z 5.20 - 5.00 - • 4.80 - 0.0 U — 1 — 1,- 1 _l L. 1111 1 1 1 1 1,600 2;600 3,600 4;600 5^00 6,600 NUMBER OF ECGS 7,600 Figure 6. — Relation of mean egg diameter to number of eggs in female coho salmon, Karluk River, 1966. 86 DRUCKER: COHO SALMON OF KARLUK RIVER SYSTEM from 4.91 to 6.87 mm (mean 6.11 mm); fecun- dity ranged from 2,855 to 6,906 (mean 4,766) . No relation was found: eggs from a female with low fecundity were not necessarily large, nor were those from a female with high fecundity necessarily small. Allen (1958) reported similar findings. Unlike the general relation that exists between fecundity and length (i.e., larger fish are more fecund), the size or fecundity of the females I sampled apparently had no relation to the size of the eggs. Large, fecund females had a wide range of egg sizes (Figure 5) . Thus, the larger number of eggs in large females may be due to a larger body cavity that allows more eggs to develop rather than to the fish having smaller eggs. COHO SALMON SMOLTS Smolts of coho salmon, like those of other salmon that live for a while in fresh water be- fore migrating to sea, migrate seaward at a par- ticular season and under particular light inten- sities. This migration and the associated envi- ronmental factors and information on age and size of migrating smolts are discussed in this section. SEASONAL MIGRATION Coho salmon juveniles reside in Karluk Lake for 1 to 4 years before they migrate to sea as smolts. From 1961 to 1967 the migration began in mid-May and was usually over by early July (Figure 7). Although most coho salmon migrate in the spring (Hamilton and Andrew, 1954; Taft, 1934; Gharrett and Hodges, 1950; Semko, 1954) , several exceptions do exist. In the Paratunka River, Kamchatka Peninsula, the migration ex- tends from the end of May to the end of August (Gribanov, 1948) ; in several streams in Oregon it extends from late winter to May (Chapman, 1961); in some streams in western Washington it runs from early winter to late spring ( Smoker, 1953) ; and at Waddell Creek, Calif., small num- bers of atypical migrants migrate in the fall and early winter (Shapovalov and Taft, 1954) . The 100 -| 1 1 T — I 1 1 1 1 rr 15 20 25 30 4 9 14 19 24 29! 4 9 14 19 MAY ' JUNE ' JULY Figure 7. — Cumulative seasonal migration of coho salm- on smolts from Karluk Lake, 1961-67. number of coho salmon smolts involved in the early or late parts of these migrations, however, represents only a small percentage of the total number of smolts in each migration. The warming of the water after the ice breaks up is of major importance in initiating the sea- ward migration of smolts. Hartman, Heard, and Drucker (1967) found this to be a major factor in the migration of sockeye salmon in lakes of southwestern Alaska; and Logan (see footnote 2, Table 1) found that the coho salmon smolt migration in Bear Lake, Alaska, did not start until the ice cover on the lake was gone and the water temperature had risen to 4.2° C. Ninety percent of the Bear Lake coho salmon smolts had migrated to sea when water temperatures ranged between 5° and 13.3° C. Coho salmon smolts apparently migrate over a greater 87 FISHERY BULLETIN: VOL. 70, NO. I temperature range than sockeye salmon, whose migration generally ends when water tempera- tures reach about 10° C (see footnote 2, Table 1) . At Karluk Lake, for each year from 1961 to 1968 (excluding 1964) the date by which 50' > of the coho salmon smolts had migrated was later than the comparable date for sockeye salm- on smolts (Table 5). The difference in time of the two migrations ran from 6 to 19 days (aver- age 11 days). Not only did more of the coho salmon smolts migrate later than the sockeye salmon smolts, but the coho salmon smolts usu- ally migrated during a period of relatively warm- er water, when the abundance of migrating sockeye salmon smolts had greatly diminished. Similarly, Foerster and Ricker (1953) found that the coho salmon smolt migration in Cultus Lake and Sweltzer Creek, British Columbia, al- ways followed the sockeye salmon smolt migra- tion by about 10 days. Although the seasonal timing of the outmigra- tion of coho salmon smolts may vary from system to system, it is relatively consistent within a particular system. When time of migration is plotted against latitude, a definite south to north cline in time of migration becomes evident (Fig- ure 8). Coho salmon smolts migrate later in the season in northerly systems than in more southerly ones. More than a month separates the midpoint of smolt migration from the central coast of California (lat 37° N) to the Gulf of Alaska (lat 60° N). This relation also applies for the Asiatic side of the Pacific Ocean. DIEL PATTERN OF MIGRATION The transformation of juvenile coho salmon from either lake- or stream-type residents to Table 5. — Dates by which 50% of the coho and sockeye salmon smolts migrated from Karluk Lake, 1961-68. Year Sampling period 50% migration date Coho salmon Sockeye salmon 1961 May 25 to June 29 June 10 June 2 1962 May 17 to June 21 June 10 May 29 1963 May 18 to July 6 June 12 June 6 1964 May 17 to July 6 June 2 June 3 1965 May 16 to July 16 June 18 June 6 1966 May 18 to July 2 June 15 June 4 1967 May 18 to June 29 Juno 5 May 27 1968 May 17 to June 26 June 12 May 24 60' 55' HOOD BAY CREEK, ALASKA* SASHIN CREEK, ALASKA* LAKELSE LAKE, BRITISH COLUMBIA* 50* o z UJ ^< (E (J 40" 35° 0' BEAR CREEK, ALASKA* *LAKE EVA, ALASKA •karluk LAKE, ALASKA *BOLSHAYA RIVER, KAMCHATKA, USSR *CULTUS LAKE, BRITISH COLUMBIA *MINTER CREEK, WASHINGTON *WADOELL CREEK. CALIFORNIA 10 15 20 MAY 25 31 /I 10 15 20 25 JUNE Figure 8. — Average date when 50% of the coho salmon smolts had migrated from river and lake systems along the Pacific coast of North America and Asia. smolts is associated with avoidance of light and increasing nocturnal activity (Hoar, Keenley- side, and Goodall, 1957; Hoar, 1958; Smirnov, 1960). Although most of the migration of smolts to salt water occurs during the darkest hours of the night, some occurs during the daytime. At Karluk Lake, for instance, during some years almost 40% of the coho salmon smolts migrated in the daytime — between 0600 and 1800 hr (Fig- ure 9). In other coho salmon rivers, the per- centage of smolts that migrate seaward during daylight is quite variable. In the Bolshaya River in Kamchatka, during the years 1944-47, 6.3 to 50.0% of the age 1 smolts"^ and 8.8 to 73.2% Fish that go to sea in their second year. 88 DRUCKER: COHO SALMON OF KARLUK RIVER SYSTEM 67 MEAN JUNE 22) 22 ) TIME PERIOD Figure 9. — Migration of echo salmon smolts by time period from Karluk Lake for the years 1961-67, of the age 2 smolts migrated in daylight (Semko, 1954.) At Bear Creek, Alaska, 49.8^r of the smolts migrated between 0400 and 2000 hr in 1962 (see footnote 1, Table 1), and W/c mi- grated between 0900 and 1700 hr in 1964 (see footnote 2, Table 3). AGE In 1956, 1965, and 1968, scale samples taken from seaward-migrating coho salmon smolts at the Karluk Lake weir revealed that the dom- inant ages were 2 and 3 and that the age com- position was similar between years (Table 6). The freshwater age composition determined from the scales of adults collected in 1966 was strikingly similar to the age composition of sea- Table 6. — Freshwater age composition of Karluk Lake coho salmon as determined from smolt and adult scale samples. ter Perce nf compos tlon Freshwa age From smolt sea es From adult scales 1956 1965 1968 1966 1 1.4 3.0 3.0 2 44.5 51.5 48.5 56.9 3 49.1 43.9 42.5 41.7 4 4.9 1.5 6.0 1.4 ward-migrating smolts in 1965 — smolts that pro- duced the adults in 1966 (Table 6). The only group missing from the adult scale sample but present in small numbers in the smolt scale sam- ples was age 1 (fish that went to sea in their second year of life). Fish from this age class could have been missing in the returning adults 89 FISHERY BULLETIN: VOL. 70. NO. 1 because (1) they had poor ocean survival be- cause of their small size; (2) the young over- wintered in the river and migrated as age 2 smolts the following year; or (3) the young never migrated at all. SIZE Only lengths were measured in 1956, and lengths and weights were taken in 1965 and 1968. A summary of average size data by age class is presented in Table 7. Average lengths for comparable age classes were greater in 1968 than in 1956 and 1965. Average weights, with the exception of age 3 fish, were less in 1968 than 1965. Differences in lengths and weights bet\N'een smolts for the two comparable years (1965 and 1968) are reflected in the condition factor (K), or coefficient of condition, which indicates the relative well-being of the fish. In 1965 all age groups had K values greater than 1.0000; the range was 1.0544 to 1.3695. In 1968 all K values were under 1.0000; the range was 0.9187 to 0.9600. Information on the size of coho salmon smolts from other spawning systems is presented in Table 8. This table gives information for natural or "wild" populations and not for artificially hatched or reared stocks. Karluk Lake coho salmon smolts were generally as large as smolts of the same age from other areas or larger. POSSIBLE EFFECTS OF INCREASED FRESHWATER RESIDENCE ON SURVIVAL OF COHO SALMON The extended period of freshwater residence resulting in coho salmon smolts of age 3 occurs in many systems but seems to be significant only at Karluk. It is interesting to hypothesize what effect a prolonged freshwater residence has on the an- nual return of adult coho salmon at Karluk Lake. Is an increased freshwater residence advanta- geous or disadvantageous to survival of each year class? What effect is there on marine sur- vival of coho salmon if they take up ocean resi- dence at an older age and consequently a larger size? One means of answering these questions is to examine freshwater and marine survival rates for coho salmon from other areas. Survival from egg to smolt (fresh water) and smolt to returning adult (marine) are shown in Table 9 for some areas in California, Oregon, Washing- ton, and British Columbia. Both freshwater and marine survival for age 1 smolts from these areas are quite variable: 0.13 to 12.00% and 3.77 to 11.79% respectively. The survival data in Table 9 pertain to stocks in which the smolts were primarily age 1 when they migrated to sea, and the application of these data to more northern stocks in which the smolts are mostly older and larger when they migrate must be done with caution. The small population of age 2 smolts from Sweltzer Creek in British Columbia (Table 9) is of interest because these fish are more compar- able to Karluk smolts, in that they may possibly have had a period of lake residence. Marine survival of these older, larger fish was high. Of 72 fin-clipped migrating age 2 smolts, 19 (26%) returned 5 or 6 months later as 83 fish (Foerster and Ricker, 1953). Although marine survival for these age 2 smolts might have been lower if they had spent another year in the ocean, it Table 7. — Average length, weight, and condition factor of coho salmon smolts by age from Karluk Lake, 1956, 1965, and 1968. Age Length 1956 1965 Weight Condition factor Length Weight Condition factor Length 1968 Weight Condition factor mm 1 106.8 2 139.7 3 151.1 4 165.4 112.5 136.3 141.7 177.0 g 19.5 28.2 30.7 63.9 1.3695 1 .0544 1 .0790 1.1523 114.8 140.1 160.4 181.8 e 13.9 26.4 38.5 56.2 0.9187 0.9600 0.9329 0.9353 90 DRUCKER: COHO SALMON OF KARLUK RIVER SYSTEM Table 8. — Average fork lengths of coho salmon smolts of ages 1 to 4 from river and lake systems along the Pacific coast of North America and Asia, arranged geographically from north to south. Area and year Age Reference North America Hood Bay Creek, Alaska 1968 83.0 1969 79.0 Karluk River (Karluk Lake), Alaska 1956 106.8 ^ 1965 112.5 1968 114.8 Bear Creek, Alaska 1962 106.3 Sweltzer Creek (Cultus Lake), British Columbia _. 110-120 1939 - Minter Creek, Wash. 1940 296.3 1953 !'99.7 Deer Creek, Oreg. 1960 88,7 Flynn Creek, Oreg. 1960 88.1 Waddell Creek, Calif. 1933 113.5 1934 113.3 1935 113.1 1936 116.6 1937 114.8 1938 112.4 1939 112.4 1940 109.5 1941 103.1 Asia Bolshoya River, Kamchatka, USSR 85.0 96.0 91.0 139.7 136.3 140.1 118.7 291.6 130.0 151.1 141.7 160.4 150.8 165.4 177.0 181.8 Armstrong (1970) Armstrong (1970) Present study Present study Present study Logan^ Foerster and Ricker (1953) Foerster and Ricker (1953) Sab and Bayliff (1958) Salo and Bayliff (1958) Chapman (1961) Chapman (1961) Shapovalov Shapovalov Shapovalov Shapovalov Shapovalov Shapovalov Shapovalov Shapovalov Shapovalov and Toft and Taft and Taft and Toft and Taft and Taft and Taft and Taft and Taft (1954) (1954) (1954) (1954) (1954) (1954) (1954) (1954) (1954) Semko (1954) 1 Personal communication, Sidney M. Logan, Fishery Biologist, Alaska Department of Fish and Game, May 3, 1967. ^ Primarily age 1 fish. Table 9. — Average freshwater and marine survival for coho salmon from various streams along the Pacific coast of North America, arranged geographically from north to south. Streams Nile Creek, British Columbia Hooknose Creek, British Columbia Sweltzer Creek, British Columbia Sweltzer Creek, British Columbia Sweltzer Creek, British Columbia Minter Creek, Wash. Deer Creek, Oreg. Waddell Creek, Calif. Percent survival Fresh water Marine Reference Egg to Aga 1 smolt aga 1 smolt to adult 1.40 6.00 Wicketf (1951) 1 1.30 11.79 Godfrey (1965) I 10.13 8.07 Foerster and Ricker (1953) I 2,^0.33 _. Foerster and Ricker (1953) 1 *26.39 Foerster and Ricker (1953) 3.22 3.77 Salo and Bayliff (1958) 12.00 _^ Chapman (1961) 1.35 4.95 Shapovalov and Taft (1954) 1 Before piscivorous fishes were controlled. " After piscivorous fishes were controlled. ^ Geometric mean. * Age 2 fish only. 91 FISHERY BULLETIN: VOL. 70, NO. 1 nevertheless was considerably higher than for any of the age 1 smolts. In the absence of knowledge of survival rates for the more northern populations of coho salm- on, an examination of the effect of increased freshwater residence on sockeye salmon, the dominant species of salmon in the Karluk system, is of value. Sockeye salmon juveniles at Karluk Lake have long been known to reside in the lake a year or more longer than do sockeye salmon in other areas (Gilbert and Rich, 1927) . In most Alaska systems, sockeye salmon smolts migrate at the beginning of their second or third year of life, but at Karluk Lake most sockeye salmon smolts migrate at the beginning of their third or fourth years. Possibly the factor (s) respon- sible for the 1-year holdover of juvenile sockeye salmon in the lake may also be responsible for the holdover of juvenile coho salmon. Freshwater sui'vival of sockeye salmon at Kar- luk Lake is extremely poor, but marine survival is good. During the late 1920's and early 1930's, freshwater survival was less than 1 % and ocean survival was about 21% (Barnaby, 1944). In recent years, freshwater survival has dropped to less than 0.5% and ocean survival has in- creased to about 40%.' Ricker (1962) modi- fied Barnaby's data by applying a marking mor- tality factor derived from his Cultus Lake studies and determined that the older, larger smolts have greater marine survival and that Barnaby's ori- ginal estimate of 21% survival was too low. Average marine survival by freshwater age for the years 1926 and 1929-33 were as follows: age 1 smolts, 18.3%; age 2, 27.4%; age 3, 34.2%,, and age 4, 33.3%. Ricker attributed the high ocean survival to the large size of the smolts when they entered salt water. The larger size of the sockeye salmon smolts at the time of sea- ward migration, however, is offset by a greater total freshwater mortality due to their prolonged stay in the lake. I have shown that in the more northern lat- itudes coho salmon usually reside a minimum of one extra year in fresh water before they migrate ' Unpublished data on file at National Marine Fish- eries Service Auke Bay Fisheries Laboratory, Auke Bav. Ala.ska 99821. to sea. Generally, a longer period of freshwater residence will result in greater freshwater mor- tality but lower marine mortality because the fish are larger when they enter the ocean. Most likely, as with Karluk Lake juvenile sockeye salmon, an extra year in the lake for juvenile coho salmon probably results in a greater total freshwater mortality. Total marine mortality, however, may be less for coho salmon than for sockeye salmon because the coho salmon gener- ally spend less time at sea before returning to spawn (12 to 18 months rather than 24 to 30 months). SUMMARY AND CONCLUSIONS Both the freshwater and total ages of adult coho salmon increase from southern to northern latitudes. In California, the southern portion of the coho salmon's range, fish of ages 32 and 22 are in the majority, but in the northern areas, ages 43 and 32 predominate. Karluk coho salmon, however, are unique, in that although age 43 fish are still the primary age class, the age 32 fish are replaced by age 54, so that age 54 fish account for 42% of the run. In no other North Amer- ican or Asiatic coho salmon stock for which in- formation is available is such a large percentage of the run composed of 54 fish. The increase in total age of coho salmon from south to north is associated with the increased time the juveniles spend in fresh water. The small numbers of age 54 fish in several Alaska stocks may represent juveniles that live in lakes rather than rivers. Fecundity generally increases from south to north, and Karluk coho salmon are the most fecund of any North American stock and closely parallel the highly fecund Asiatic stocks from the Kamchatka Peninsula. In Karluk coho salmon, there is a relation between number of eggs and length but no relation between egg size and length or egg size and fecundity. Egg counts are significantly higher in the right ovary than in the left. Coho salmon smolts generally migrate after the ice breaks up and the water warms. In North America, the migration is earlier in southern latitudes than northern ones. The coho salmon 92 DRUCKER: COHO SALMON OF KARLUK RIVER SYSTEM migration at Karluk Lake is primarily nocturnal, although some daytime migration does occur. A prolonged freshwater residence by juvenile coho salmon in Karluk Lake should result in a greater total freshwater mortality, but the re- sulting larger smolts should have a lower total marine mortality. Coho salmon at Karluk may have an even lower marine mortality than sock- eye salmon, in part because the coho salmon spend less time at sea. LITERATURE CITED Allen, G. H. 1958. Notes on the fecundity of silver salmon {Oncorhynchus kisutch). Prog. Fish-Cult. 20: 163-169. Armstrong, R. H. 1970. Age, food, and migration of Dolly Varden smolts in southeastern Alaska. J. Fish. Res. Board Can. 27: 991-1004. Barnaby, J. T. 1944. Fluctuations in abundance of red salmon, Oncorhynchus nerka (Walbaum), of the Karluk River, Alaska. U.S. Fish Wildl. Serv., Fish. Bull. 50: 237-295. Chapman, D. W. 1961. Factors determining production of coho salm- on, Oncorhynchus kisutch, in three Oregon streams. Ph.D. Thesis, Oreg. State Univ., Cor- vallis, 214 p. Crone, R. A. 1968. Behavior and survival of coho salmon, On- corhynchus kisutch (Walbaum), in Sashin Creek, southeastern Alaska. M.S. Thesis, Oreg. State Univ., Corvallis, 79 p. FOERSTER, R. E. 1944. Appendix IV. Report for 1943 of the Pa- cific Biological Station, Nanaimo, B.C. Annu. Rep. Fish. Res. Board Can., 1943: 22-26. FoERSTER, R. E., and A. L. Pritchard. 1936. The egg content of Pacific salmon. Biol. Board Can., Prog. Rep. Pac. Biol. Stn. Pac. Fish. Exper. Stn. 28: 3-5. 1941. Observations on the relation of egg content to total length and weight in the sockeye salmon {Oncorhynchus nerka) and the pink salmon (O. gorbuscha) . Trans. R. Soc. Can., Ser. 3, Sec. 5, 35: 51-60. FOERSTER, R. E., AND W. E. RiCKER. 1953. The coho salmon of Cultus Lake and Sweltzer Creek. J. Fish. Res. Board Can. 10: 293-319. Gharrett, J. T., AND J. I. Hodges. 1950. Salmon fisheries of the coastal rivers of Ore- gon, south of the Columbia. Oreg. Fish Comm., Contrib. 13, 31 p. Gilbert, C. H. 1922. The salmon of the Yukon River. Bull. U.S. Bur. Fish. 38: 317-332. Gilbert, C. H., and W. H. Rich. 1927. Investigations concerning the red-salmon runs to the Karluk River, Alaska. Bull. U.S. Bur. Fish. 43, Part II: 1-69. Godfrey, H. 1965. Salmon of the North Pacific Ocean— Part IX, Coho, chinook and masu salmon in offshore waters. 1. Coho salmon in offshore waters. Int. North Pac. Fish. Comm., Bull. 16: 1-39. Gribanov, V. I. 1948. Kizuch [Oncorhynchus kisutch (Walb.)]: (biologichestii ocherk). (The coho salmon (On- corhynchus kisutch (Walbaum) — A biological sketch.) Izv. Tikhookean. Nauchn.-issled. Inst. Rybn. Khoz. Okeanogr. 28: 43-101. (Fish. Res. Board Can., Transl. Ser. 370.) Hamilton, J. A. R., and F. J. Andrew. 1954. An investigation of the effect of Baker Dam on downstream-migrant salmon. Int. Pac. Salmon Fish. Comm., Bull. 6, 73 p. Hartman, W. L., and C. Y. Conkle. 1960. Fecundity of red salmon at Brooks and Kar- luk Lakes, Alaska. U.S. Fish Wildl. Serv., Fish. Bull. 61: 53-60. Hartman, W. L., W. R. Heard, and B. Drucker. 1967. Migratory behavior of sockeye salmon fry and smolts. J. Fish. Res. Board Can. 24: 2069- 2099. Helle, J. H. 1970. Biological characteristics of intertidal and fresh-water spawning pink salmon at Olsen Creek, Prince William Sound, Alaska, 1962-63. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 602, iv + 19 p. Hoar, W. S. 1958. The evolution of migratory behaviour among juvenile salmon of the genus Oncorhynchus. J. Fish. Res. Board Can. 15: 391-428. Hoar, W. S., M. H. A. Keenleyside, and R. G. Goodall. 1957. Reactions of juvenile Pacific salmon to light. J. Fish. Res. Board Can. 14: 815-830. Hunter, J. G. 1948. Natural propagation of salmon in the central coastal area of British Columbia. Fish. Res. Board Can., Prog. Rep. Pac. Coast Stn. 77: 105- 106. Israel, H. R. 1933. On the life history of the silver salmon On- corhynchus kisutch (Walbaum) of Chignik River, Alaska. M.S. Thesis, Stanford Univ., Palo Alto, Calif., 58 p. KosKi, K. V. 1966. The survival of coho salmon (Oncorhynchus kisutch) from egg deposition to emergence in three Oregon coastal streams. M.S. Thesis, Oreg. State Univ., Corvallis, 84 p. 93 FISHERY BULLETIN: VOL. 70, NO. I Mark, J. C. 1943. Age, length, and weight studies of three species of Columbia River salmon (Oncorhynchus keta, O. gorbuscha, and O. kisutch). Stanford Ichthyol. Bull. 2: 157-197. Nea\'E, F. 1948. Fecundity and mortality in Pacific salmon. Trans. R. Soc. Can., Ser. 3, Sect. 5, 42: 97-105. Nelson, P. R. 1959. Effects of fertilizing Bare Lake, Alaska, on growth and production of red salmon (0. nerka). U.S. Fish Wildl. Serv., Fish. Bull. 60: 59-86. PRITCHARD, A. L. 1940. Studies on the age of the coho salmon (O71- corhynchus kisutch) and the spring salmon (Oncorhynchus tshuwytscha) in British Columbia. Trans. R. Soc. Can., Ser. 3, Sect. 5, 34: 99-120. RiCKER, W. E. 1962. Comparison of ocean growth and mortality of sockeye salmon during their last two years. J. Fish. Res. Board Can. 19: 531-560. ROUNSEFELL, G. A. 1957. Fecundity of North American Salmonidae. U.S. Fish Wildl. Ser\'., Fish. Bull. 57: 451-468. Salo, E. 0. 1955. Silver salmon, Oncorhynchus kisutch, sur- vival studies at M inter Creek, Washington. Ph.D. Thesis, Univ. Wash., Seattle, 183 p. Salo, E. 0., and W. H. Bayliff. 1958. Artificial and natural production of silver salmon {Oncorhynchus kisutch) at Minter Creek, Washington. Wash. Dep. Fish., Res. Bull. 4, 76 p. Semko, R. S. 1954. Zapasy zapadnokamchatskikh lososei i ikh promyslovoe ispolzovanie. (The stocks of west Kamchatka salmon and their commercial utili- zation.) Izv. Tikhookean. Nauchn.-issled. Inst. Rybn. Khoz. Okeanogr. 41: 3-109. (Fish Res. Board Can., Transl. Ser. 30.) Shapovalov, L., and A. C. Taft. 1954. The life histories of the steelhead rainbow trout (Salmo gairdneri gairdneri) and silver salmon (Oncorhynchus kisiitch) with special ref- erence to Waddell Creek, California, and rec- ommendations regarding their management. Cal- if. Fish Game, Fish. Bull. 98, 375 p. Smirnov, a. I. 1960. K. kharakteristike biologii razmnozheniia i razuitiia kizhucha — Oncorhynchus kisutch (Wal- baum). (The characteristics of the biology of reproduction and development of the coho On- corhynchus kisutch (Walbaum).) Vestn. Mosk. Univ., Ser. VI Biol. Pochvoved., p. 9-19. (Fish. Res. Board Can., Transl. Ser. 287) Smoker, W. A. 1953. Stream flow and silver salmon production in western Washington. Wash. Dep. Fish., Fish. Res. Pap. 1: 5-12. Taft, A. C. 1934. California steelhead experiments. Trans. Am. Fish. Soc. 64: 248-251. WiCKETT, W. P. 1951. The coho salmon population of Nile Creek. Fish. Res. Board Can., Prog. Rep. Pac. Coast Stn. 89: 88-89. 94 DEVELOPMENTAL RATES AT VARIOUS TEMPERATURES OF EMBRYOS OF THE NORTHERN LOBSTER (Homarus americanus MILNE-EDWARDS) Herbert C. Perkins' ABSTRACT The rates of development, time from extrusion to hatching at various temperatures, and differential de- velopmental rates at the same temperature of lobster embryos are presented. The eyes of the embryos were measured to monitor the rates and degree of embryo development. Herrick (1890, 1896) discussed developmental rates for lobster embryos in the early stages at 20° to 22° C. Templeman (1940) determined the times required at various temperatures for lobster eggs to reach the 16-cell stage, and up to the formation of eye pigment. The information from these studies is valuable for determining the rates of early development in lobster egg- embryos but is not adequate for the accurate assessment of developing embryos once eye pig- ment has been formed. By monitoring the rate of development of lobster embryos throughout the embryonic period at various temperatures one can predict hatching times of larvae and con- trol hatching times by manipulating the water temperature in tanks holding egg-bearing fe- males, so that larvae can be available over a wide period of time for use in experiments. This paper presents the rates of development and time required to complete the embryonic period by lobster embryos at various temperatures and a method of continually monitoring that develop- ment. The work was conducted at the National Marine Fisheries Service, Biological Laboratory, Boothbay Harbor, Maine, as part of the Labora- tory's investigation of the early life history of the lobster. METHODS AND MATERIALS Most of the egg-bearing lobsters used in this study came from the offshore canyons of the ^ National Marine Fisheries Service, Northeast ' Fish- eries Center, Boothbay Harbor Laboratory, W. Boothbay Harbor, ME 04575. continental shelf, south and east of New England. A few came from the Boothbay Harbor area and are so noted. All egg-bearing females were kept in tanks at seasonal water temperatures or in water warmed to various constant temperatures. Water from the laboratory's seawater system was piped to the heated tanks at rates consistent with the capacity of the heaters. Salinity aver- aged 31'/( and ranged from 29 to Z2%c throughout the study period. Five egg-bearing females were kept in a tank through which natural seawater at seasonal tem- perature was circulated during the development- al period of their eggs. The purpose of holding these females at seasonal temperatures was to determine the rates of development of their em- bryos in a natural temperature regime. Four- teen female lobsters from the offshore canyons, with recently extruded eggs (eggs in prenaupliar condition), were kept at constant temperatures from 6.9° to 24.6° C. The primary objective at the higher temperatures (20°-24.6° C) was to force the eggs to hatch before the time they would do so at seasonal temperatures. Of fur- ther interest was the rate of development of the embryos at constant, rather than fluctuating, temperatures, and the time required for the eggs to hatch at these temperatures from a given point in their development. The rates and extent of development of the embryos were determined by measuring the size of their eyes. Measurements were made to the nearest micron with an ocular micrometer in a dissecting microscope at a magnification of 50 x . When measuring an eye, I took its greatest width and greatest length, combined these figures and Manuscript accepted July, 1971. FISHERY BULLETIN: VOL. 70, NO. 1, 1972. 95 FISHERY BULLETIN: VOL. 70, NO. 1 divided by two. The resulting figure was used as an index of development only and was not meant to represent the actual increase or growth of the eye. Samples of 15 to 20 eggs were re- moved from the periphery of an egg mass when- ever a measurement was desired (usually once a week) . From these samples five eggs were se- lected randomly, and one eye of each embryo measured. A mean of these five measurements was used as a working figure or index. Eggs were taken from the peripheral layer of the mass as these are the furthest advanced in develop- ment and are the first to hatch. Variation in the eye measurements of eggs from this layer ranges from zero to o^C When the eyes first appear and are large enough to measure, they are but thin crescents, darkly pigmented and surrounded by a halo of lighter material. The dark crescents only were measured. The eyes are very distinct for most of the developmental period and are easily measured, the crescents gradually becoming tear-drop in shape. The in- dex of the eye is about 70/x when it is first mea- surable; the index is about 560/i at hatching. All eyes were measured after the eggs had been preserved in a 5/f solution of Formalin in sea- water. Preservation in Formalin caused sig- nificant swelling in the eggs themselves but had no determinable effect on the size of the eyes. year, water temperatures were no higher than 6° C. Embryonic development during this pe- riod ceased in some of the egg masses and was barely discernible in others, at least by the method of eye index measurement. Squires (1970) , using the amount of yolk material in the eggs as a criterion, reported a standstill in devel- opment during the winter in embryos of New- foundland lobsters. tfi 600 I 500 y 400 Z 500 S 200 o 100 Eye Pigment Stofis 600 500 400 xa 200 100 0 JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG Figure 1. — Trends of development of the embryos of five female lobsters held in water of seasonal temperature at the Boothbay Harbor Laboratory and the temperature cycle during the period. The lines showing develop- mental trends are derived from plots of periodic measure- ments of the embryos' eye indices. The lines are num- bered in accordance with the age of the egg mass; the lower the number, the older the egg mass. RESULTS AND DISCUSSION TRENDS OF EMBRYONIC DEVELOPMENT IN WATER OF SEASONAL TEMPERATURES The developmental patterns, from onset of eye pigment to hatching, of the embryos of the five females held at seasonal temperatures are shown in Figure 1, as is the cycle of water temperature for the same period. The trends of embryonic development were plotted by using periodic (usu- ally weekly) measurements of the eye index of each egg mass. These five lobsters had extruded their eggs in the laboratory tanks so the age of each egg mass was known. From the latter part of November to the first of May of the following DIFFERENTIAL RATES OF DEVELOPMENT The developmental rates of lobster embryos appear to be governed not only by their thermal environment, but by the age or extent of de- velopment at which they are subjected to that environment. During the experiments I con- ducted the older or more advanced embryos de- veloped at slower rates than those less advanced, though all were maintained in the same tank. The oldest egg mass of the five females held under seasonal conditions was extruded 7 weeks before the youngest, but the total time for development of the younger egg mass was 414 weeks less than the older; the younger embryos had developed considerably faster. Measurements of the eye index of embryos in all five egg masses were made for the first time on November 7. In Fig- 96 PERKINS: DEVELOPMENTAL RATES OF NORTHERN LOBSTER EMBRYOS LlI $ a: UJ 0. z o a: <_) X UJ Q LlI O UJ to < UJ a: o 60 50 40 30 20 10 - (r=.952,P<.OI) _L. J_ _1_ (r= .995, P <.0I) _L J 50 100 150 200 250 300 350 400 450 500 EYE INDEX AT START (MICRONS) Figure 2. — Line A represents the different rates of in- crease of the eye index, in microns per week, of the em- bryos of the five lobsters held in the same tank, under seasonal conditions, from November 7, until hatching. Line B represents the different rates of eye increase of the embryos of seven females held at a constant tem- perature of 22.6° C. Size of eye index at the starting time is plotted against the corresponding rate of increase of the eye index up to the time of hatching. Table 1. — Lobster number, carapace length of female, age of eggs 10 January, increase of eye index of embryos from 10 January to 26 March, and the total develop- mental time for the embryos of the five female lobsters held under seasonal water conditions at the Boothbay Harbor Laboratory. Lobster Carapace number 'f"9\'i (mm) Area of capture Age of eggs Increase of Total 10 January , eye index weeks (weeks) 7J";''°"'/.T/^' u /?• ' 10 Jan. -26 Mar. hotchmg 1 97 Boothbay Harbor 29 0.00 51.4 2 94 Boothbay Harbor 26 0.46 50.6 3 147 Veatch Canyon 24 0.65 50.0 4 124 Hudson Canyon 23 1.49 49.4 5 94 Boothbay Harbor 22 2.52 47.0 surable increase in development during this time, whereas some development was noted in the em- bryos of the other egg masses. In fact, the em- bryos in the oldest egg mass showed no measur- able increase from the second week of December to the middle of the following April. The num- ber of weeks during the winter in which no de- velopment could be measured, for each egg mass, was as follows (as in Table 1, lobsters are num- bered according to the age of their egg mass) : weeks 'Number of "dormant' Lobster number during winter 1 18 2 14 3 8 4 6 5 0 ure 2(A) the rate of eye increase (microns per week) of the embryos in each egg mass for the remainder of the developmental period is plotted against the corresponding eye index taken on November 7. The same result is obtained if the age of an egg mass is substituted for its eye index on the abscissa. Rates of increase were calculated by dividing the total increase of the eye index by the total number of weeks the em- bryos took to complete development after No- vember 7. The increase in eye index (microns per week) of embryos in each egg mass, for the period January 10 to March 26, is presented in Table 1, During this time the water temper- ature ranged between 0.1° and 1.5° C; the mean was 1.0°. This was the coldest 10-week interval of the developmental period. The embryos in the oldest egg mass exhibited no noticeable or mea- The rates of increase (microns per week) in eye index for the embryos of seven lobsters held at a constant temperature of 22.6° C are indicated in Figure 2(B). These females came from offshore canyons of the continental shelf, off New England, The ages of these egg masses were not known, but the eye index of the embryos in each was measured before the females were placed in the warm water. The increase in eye index of the embryos in each egg mass was mon- itored weekly until hatching. Times and rates pertain only to the time spent at 22.6° C. Al- though one might expect that in a given time in- terval, at the same temperatures, younger em- bryos would develop faster than older ones, it might also be expected that the younger embryos would assume the slower growth rate of the older when they eventually reached the same age. Of 97 FISHERY BULLETIN: VOL. 70, NO. 1 the five females held in water of seasonal tem- perature, the youngest eggs were at lower tem- peratures than the older at the same age, making assessment of differential developmental rate difficult. However, of the seven egg masses held at 22.6° C the embryos of least development at the start developed faster at comparable levels of development than the more advanced embryos. The differential rate of development of lobster embryos, at the same temperature, seems to im- ply that in a given population where extrusion of eggs may be somewhat staggered in time among the females, hatching of the eggs would occur during a more limited period, providing the egg-bearing females occupied the same ther- mal environment. SOr 70- X u t- X 60 50 40 » 30 20 - 10 • = OBSERVED VALUES O « CALCULATED VALUES (•) 10 12 14 16 le 20 22 TEMPERATURE OF WATER (°C ) 24 26 Figure 3. — The number of weeks for lobster eggs to complete the embryonic period at various temperatures. Line A represents the time required from onset of eye pigment in the embryos; line B represents the time re- quired from extrusion to hatching. Points in paren- theses indicate times required at the mean temperature of a fluctuating thermal environment. RATES OF DEVELOPMENT AT VARIOUS TEMPERATURES The times required for the embryos to hatch at various temperatures are shown in Figure 3. Line A represents the time required for the em- bryos to hatch after the formation of eye pig- ment; line B represents the time required from extrusion to hatching. Most of the points in each line indicate the time required to complete de- velopment at constant temperatures. A few (points in parentheses), representing the time required for total development at the mean tem- perature of a fluctuating thermal environment, have been included as well. For example, the average time required for total development of the eggs of the five females held under seasonal conditions was 49.7 weeks. The mean water temperature during the period was 8.1° C. These values are virtually the same as would be expect- ed if the water was held constantly at 8° C. All values showing time from onset of eye pig- ment to hatching were observed. The times re- quired from extrusion to hatching at five tem- peratures were also observed. To find the unknown time required from extrusion to hatch- ing, at other temperatures, I used the following equation: At = Az Xi X2 where Ai was the observed time from onset of eye pigment to hatching at 20° C; A2 was the observed time from extrusion to hatching at 20° C; Xi was the observed time required from Table 2. — Number of weeks required from extrusion to onset of eye pigment, onset of eye pigment to hatching, and to hatching at certain temperatures, at salinities near 31%c. Weeks required from Water temperature C C) Extrusion to onset of eye pigment Onset of eye pigment to hatching Extrusion to hatching 5 10 15 20 25 40 9 S A 3 120 30 18 12 9 160 39 23 16 12 98 PERKINS: DEVELOPMENTAL RATES OF NORTHERN LOBSTER EMBRYOS onset of eye pigment to hatching at a given tem- perature; and A'2 is the unknown time required from extrusion to hatching at the same temper- ature as Xi. Templeman (1940, p. 74) used a similar method to find unknown developmental rates. The requisite times for development of lobster embryos at certain temperatures are summarized in Table 2. The relationship be- tween water temperature and the average in- crease in eye index of lobster embryos, in mi- crons per week, is linear at temperatures be- tween 5° and 25° C. The index of the embryonic eye must increase to approximately 560^t at hatching. If eggs are encountered with eyed embryos, their eye index may be subtracted from 560 and the difference divided by the value calculated from the following equation: Y = —8.3151 + 2.6019 (X) where Y is the increase of the eye index in mi- crons per week, and X is the developmental tem- perature. The resulting quotient is the average number of remaining weeks required for the em- bryos to hatch, depending on genetic variation and the differential rate of development noted earlier. SUMMARY 1. Once eye pigment has been formed, the course and rate of development of lobster em- bryos may be monitored by the periodic mea- suring of the eye of the embryos. 2. Lobster embryos develop differentially, under the same thermal conditions, depending on their age or extent of development when they are subjected to a given thermal environment. 3. As water temperature has a direct effect on the developmental rate of lobster embryos, that rate may be manipulated by adjusting the water temperature of holding tanks to insure periodic hatches of larvae throughout the year. LITERATURE CITED Herrick, F. H. 1890. The development of the American lobster, Homarus americarms. Johns Hopkins Univ. Circ. 9: 67-68. 1896. The American lobster : A study of its habits and development. Bull. U.S. Fish Comm. 15: 1-252. Squires, H. J. 1970. Lobster (Homarus americanus) fishery and ecology in Port au Port Bay, Newfoundland, 1960- 65. Proc. Natl. Shellfish. Assoc. 60: 22-39. Templeman, W. 1940. Embryonic developmental rates and egg-lay- ing of Canadian lobsters. J. Fish. Res. Board Can. 5: 71-83. 99 PRELIMINARY STUDIES OF SELECTED ENVIRONMENTAL AND NUTRITIONAL REQUIREMENTS FOR THE CULTURE OF PENAEID SHRIMP' Lowell V. Sick, James W. Andrews, and David B. White" ABSTRACT Types of substrate, type of aeration, and stocking density were compared as prerequisities for high-den- sity culture studies with penaeid shrimps. Neither sand-shell substrate nor brick subdivisions of cul- ture tank bottoms produced significantly higher survival rates than bare fiber glass tanks. Forced air supplied via airstones proved to be a more suitable form of aeration than did physical agitation of the water column in culture tanks by high-pressure nozzles. Survival rates of 80 to 90% were achieved when biomass densities did not exceed 40 g/m^. Semipurified pelleted diets (i.e., containing defined chemical ingredients plus one or more natural products) having a complement of nutrients including minerals and vitamins, various ratios of shrimp to fish meal, protein hydrolysates, and such diets fed at three percentages of total biomass daily were compared for their ability to produce increases in growth. Diets without fish or shrimp meal sustained biomass while those diets having the highest proportion of shrimp to fish meal in addition to added vitamins produced over 60% increase in total biomass over a 3-month period. Animals fed a combina- tion of yeast, soy, and casein hydrolysates increased 39% in biomass over the same period of time while those fed each of the above hydrolysates during the 3-month period separately showed only an average of 18% increase in weight. Feeding shrimp with a fish-shrimp base with added vitamins at a rate of 15% daily of the total biomass produced a 164% increase in weight with 95 to 100% survival during the 3-month period. Using semipurified pelleted diets, a food conversion ratio of 5.5 was obtained. Establishing selected preliminary environmental and nutritional requirements for penaeid shrimp re- sulted in the successful and reproducible production of major biomass increases with relatively high sur- vival rates and low food conversion ratios. The harvest of commercial shrimp suffers great seasonal variability and has failed to keep pace with ever-increasing domestic and export de- mands (Surdi and Whitaker, 1971). In order to supplement the natural harvest and provide a year-round supply of shrimp, several attempts have been made to culture shrimp in natural ponds, restricted portions of bays and estuaries, and laboratory tanks. In general, these efforts have had limited success and have explicitly il- lustrated the need for more accurately defining the nutritional and environmental requirements ^ This work is a result of research sponsored by NOAA Office of Sea Grant, U.S. Department of Commerce, under Grant #GH-73. ^ Skidaway Institute of Oceanography, 55 West Bluff Road, Savannah, GA 31406. necessary for culturing these species. Although pond culture has produced annual crops of shrimps (Villadolid and Villaluz, 1951; Lunz, 1967; Wheeler, 1967; Broom, 1969; Moore and Elan, 1970') , production has been minimal and highly variable. Attempts to obtain commercial quantities of shrimp by stocking enclosed por- tions of estuaries have to date not yielded pro- duction results (American Fish Farmer & World Aquaculture News, 1970) . During recent labora- tory studies, Subrahmanyam and Oppenheimer (1969) were able to maintain shrimp in labora- tory tanks using a pelleted diet consisting of fish meal, stickwater, and vitamins. However, the Manuscript accepted July 1971. FISHERY BULLETIN: VOL. 70, NO. 1, 1972. " Moore, W. R., and L. L. Elan. 1970. Salt water pond research. Tex. Parks Wildl. Dep., Austin. (Processed.) 101 FISHERY BULLETIN: VOL. 70. NO. 1 total biomass increase of shrimp fed their best diet for 6 weeks was only approximately 50% higher than initial biomass. Better results (on an individual weight basis) using Penaeus duo- rarum were obtained from animals grown on a sand substrate than those grown in bare tanks. The greatest promise for economical shrimp production lies in determining the exact nutri- tional requirements and developing an inexpen- sive artificial diet from feedstuffs for these spe- cies. Current commercial practices in Japan employ chopped clam (predominantly Tapes semidecussata, Reeve) as a diet for rearing shrimp. Despite the high market price for cul- tured shrimp in Japan (the retail price of cul- tured shrimp ranged from $4 to $10 per pound in 1970), shrimp farming there tends to be a marginal enterprise because of the high cost of a clam diet. However, in other parts of the world where shrimp does not command such a luxury price, the use of a high-value product such as clam for shrimp feed is prohibited. Pelleted diets (i.e., pellets containing all the chemical ingredients thought to be important to animal growth) have been designed consisting of purified soybean meal, glucose, sucrose, starch, glucosamine, chitin, cellulose, soybean oil, citric acid, succinic acid, amino acid, minerals, vita- mins, and cholesterol (Kanazawa et al., 1970). After growing penaeids on such diets, the ani- mals were in excellent physiological condition, but in the best group, total biomass increase was only 72% of the control group fed chopped clam. Thus, little progress has been made toward estab- lishing nutritional and environmental require- ments that will yield optimum growth (total biomass increase) and survival of penaeid shrimp. In the present study, an attempt was made to develop a suitable experimental culture system which could serve as a model for future nutri- tional and environmental studies. Several en- vironmental factors were examined, and, as a result, environmental conditions were created which would allow acceptable growth and sur- vival. Having first established suitable culture conditions, several diets were evaluated in pre- liminary studies of the nutritional requirements of shrimp. MATERIALS AND METHODS Both environmental and nutritional studies were conducted in round fiber glass culture tanks measuring approximately 1 m deep by 1 m in diameter and equipped with a venturi type cen- ter drain which maintained a water depth of 0.75 m. Three replicates were maintained for all treatments. Water (ranging in salinity from 26.8 to 29.3^r) from the Skidaway River was filtered through an oyster shell and sand filter to re- move major food particles. Filtered water was heated to 30° C in a stainless steel heat exchang- er and jetted into each tank at a rate of 1.9 liters/min through flow-control nozzles which were aimed so that the agitation of the water column in each tank was minimal. Temper- ature ranged from 25° to 28°C in each tank throughout the experimental period. White shrimp (P. setiferus) obtained from the Savannah, Ga., river and tributary systems were used in all environmental experiments, and brown shrimp (P. aztecus) obtained from the Tampa Bay, Fla., area were used in the nutri- tional studies. Shrimp weighing 4 db 0.8 g (mean and standard deviation based on 480 weighed shrimp) were selected from the above stock and used in all environmental studies (10 animals per tank) and fed pelleted diets (Table 1, Diet 1) at a rate of 5% of their biomass daily, on a dry weight basis. Shrimp were weighed each week and the percent increase or loss re- corded on a wet weight basis. SUBSTRATE STUDY Sand-shell substrates suitable for burrowing, subdivisions of tank bottoms, and bare fiber glass tank bottoms were provided for replicate groups of shrimp, and relative survival rates among the treatment groups were compared over a 5-week period. Sand-shell substrates were placed di- rectly onto the tank bottom in one group, and in another group, the same substrate was placed on a perforated platform 10 cm above the tank bot- tom, allowing a flow of water through the drain below the sand surface. Such an arrangement was designed to test the eflfect of decreasing the 102 SICK, ANDREWS, and WHITE: REQUIREMENTS FOR SHRIMP CULTURE S U p4 0) a 3 '•B O 3 6 -t-> CO »-H OS C _o '+3 •rH U -*-> ;3 C >H s o «M s 10 o "S b tn -£ -M "s a 03 O (D ^ M O ■o c cS CO ,2 3 OS •E cS > ,__, c3 -t^ C o 1 p lO I O lO I I I I p CO r O ' CO I lo to ' CO 00 CM o p lO I CN CO CN to 1 o >o ■"I- CNl 1 CO CN 1 1 p lO * 1 CO CN 1 1 1 r 1 1 1 1 >0 1 I 1 1 1 1 1 1 1 1 1 1 1 1 p 1 1 1 1 1 lO 1 1 1 1 1 1 o t 1 1 1 1 1 1 1 1 1 1 1 1 O lO 1 oS ' ' ' CM ' ' ' CO ' CO CM CM >o ' ' ' CJ CM ' to 1 1 1 rill 1 >o ■o 1 1 O >0 1 CO ' ' ' cs 1 1 1 1 CO ^ ' ' -^^ ci ' O I 1 1 1 1 1 o 1 >o o 1 1 O vo 1 CO ' ' ' ' ' ' 00 MD CO ' ' Tf CM ' to 1 1 1 1 1 T P 1 lO >o 1 1 O lO 1 00 ' ' ' CM 1 1 ' CC) ' oo CN od CM ' ' ^ CM ' 1 1 1 1 CN O 1 1 I 1 1 1 1 1 1 p 1 1 1 1 1 1 1 1 lO I I 1 (III 1 1 1 1 1 1 1 p 1 1 1 1 UO 00 o ^ 1 1 1 1 CO d > ' ' 1 1 1 Tf CN CO to CM CN O ^_ CN CO CO p lO 1 1 1 1 t 1 o >o O- C3 — d 6 O o CO 6 ' ' ' '''■*■ CM CN o CO 6 ' ' ' 111^ CN IX 2 S i 3 5 "^ E ^ E 2 — o p o p ^ i) o o -D >- *(U ^ if "^ *= c o^ ■- o^ O .± — p — a -^ c a> O (U c "^ D> p ^ .5 a. "' — n — USO«Uco5lEu^<^VUooU>U o E o O E 5 >^ 9 o- o co o- 00 o 6 CO u CO o s? CO CO cs o c o- o o X o o "O o> lO — o o D E o „ 2 s? *- lA O^ ( 5° ^.ro? 05 c £ cmB^ >2 ai — ^ O E'^ o o E o^ O u ^ CO 0) ■^ O w 0>>CN 5. . o 103 FISHERY BULLETIN: VOL. 70, NO. 1 buildup of anaerobic conditions. Division of tank bottoms into sundry tunnels and levels was created by specific placement of bricks and clay drain tiles. AERATION STUDY Aeration provided by jetting streams of fil- tered seawater into respective tanks was com- pared to aeration supplied by bubbling air through airstones into tanks in which water was continuously added with no agitation of the water column for an 8-week period. Two airstones were placed in each tank and valve-regulated air lines controlled the pressure at approximately 4 psi. Oxygen levels were monitored periodically and used along with survival rates as a basis for evaluation of replicate groups aerated by each method. STOCKING DENSITY STUDY Survival data were compared among triplicate tanks stocked at 10, 20, and 40 shrimp per m^ for an 8-week period. These densities of ap- proximately 40, 80, and 160 g/m- were chosen on the basis of data provided in pond and lab- oratory culture of penaeid shrimp (Broom, 1969; Subrahmanyam and Oppenheimer, 1969). PRELIMINARY NUTRITIONAL STUDY Triplicate groups of ten 4 g brown shrimp (P. aztecus) were fed a series of pelleted diets. Growth data (biomass increase) was used as a means of evaluation. Diets examined consisted of those patterned after Japanese purified diets (i.e., diets containing only chemical ingredients) (Table 1, Group I) (Diet 1 was conducted for 5 weeks and Diets 2, 3, and 4 for 11 weeks each) ; a second group of semipurified diets (i.e., con- taining defined chemical ingredients but contain- ing one or more natural products) providing four combinations of levels of protein, fat, shrimp, and fish meal (Group II) (conducted for 11 weeks) ; and a third group designed to compare the nutritional value of casein, yeast, and soy hydroly.sates (Group III) (conducted for 6 weeks). All of these groups were fed at 5% Table 2. — Percent of pellet dissolved over time and at at three concentrations of binder. (Values are means and standard deviation on two replicates with Diet 1.) Percent b inder added Hours (collagen) 6 12 24 1 3 5 10 13 ± 1.2 11 ±0.8 10 ±0.6 14 ±0.9 10 ± 0.6 10± 1.1 18 ± 1.7 10 ±0.6 10 ± 1.0 of their respective biomass daily. In addition, Diet 6 was fed at 5, 10, and 15% of biomass (Group IV) (conducted for 6 weeks). Combined environmental factors which pro- duced best survival in each of the environmental experiments (i.e., culture conditions consisting of bare fiber glass tank bottoms, supplied aera- tion, and a stocking density of approximately 40 g/m-) were used in all nutritional studies. This combination oflfered a maximum potential for an increase in biomass and therefore allowed accurate evaluation of diflferences among diets tested. Although survival in bare fiber glass tanks was not significantly different from sand substrates, the fact that bare tanks were simpler to maintain dictated that they be used for the nutritional studies. Prior to starting nutritional studies, the phys- ical properties of pelleted diets were evaluated for acceptability as shrimp food. Pellet consis- tency was determined according to its ability to resist dissolution over a given period of time, and texture and size were chosen according to animal performance when presented several choices. Collagen' proved to be a suitable bind- ing agent. Using an experimental design with time and collagen levels as variables, a pellet with 5% collagen added as a binder was found to offer optimum consistency over a 24-hr im- mersion in salt water (Table 2). Percent dis- solution was measured by taking dry weights after 6, 12, and 24 hr of immersion (no shatter- ing of pellets was observed, and all loss of weight was therefore assumed to be from dissolution). Animals were observed to feed most readily on * Supplied on an experimental basis by the Hides and Leather Division of the U.S. Department of Agriculture Eastern Utilization Laboratory in Philadelphia, Pa. 104 SICK, ANDREWS, and WHITE: REQUIREMENTS FOR SHRIMP CULTURE pellets 0.3 cm in diameter by approximately 1.5 cm in length and which sink in water, and hence, pellets having these characteristics were used in both environmental and nutritional experiments. RESULTS AND DISCUSSION SUBSTRATE STUDY A survival rate of 80% was obtained after 5 weeks in tanks without substrate, 80 to 90% survival was maintained over much of the dur- ation of the experiment among both treatments having sand-shell substrates, and less than 60% survival occurred among tanks having brick sub- divisions (Figure 1). Although P. setiferus is reported to burrow less than either P. duorarum or P. aztecus (Anderson, 1966; Perez Farfante, 1969) , it apparently was able to avoid predation, especially during the highly vulnerable moulting period, quite successfully with or without a sand substrate, since 5-week survival data among the two sand-shell treatments and the bare tank bot- tom treatment were not significantly different (P < 0.05) (Duncan, 1955). If the type of shelter is a factor in increased survival for penaeids maintained under culture conditions, the brick subdivisions should have enhanced sur- vival. However, the markedly high mortality 100 80 ■ c 60 40 20 SiJND-SHELL SUBSTRATE ON PLATFORM SAND-SHELL ON TANK BOTTOM NO SUBSTRATE BRICK SUBSTRATE Figure 1. — Mean and standard error for percentage of animals surviving after 5 weeks of growth on four dif- ferent substrates. rate among this group, significantly different from the other three treatments (P < 0.05), may have resulted from either failure of the shrimp to behaviorally segregate and thus fully utilize this protection or from physical abrasion against the sharp and coarse brick surface. Al- though there may have been toxic substances in the brick materials, the bricks were carefully washed and assumed to be otherwise inert in any chemical effect they may have had on the animals. Although differences in volume of water caused by placing various substrates in their respective treatments was not controlled for, it was felt that these differences in a running water system were not critical to the survival of shrimp. Dif- ferences in bottom area among the treatments caused by placement of different types of sub- strate were neither controlled for nor measured but were also thought to be negligible compared to differences found among treatment groups. The high degree of cannibalism noted by Subrah- manyam and Oppenheimer (1969) in tanks with- out substrate was not observed in any groups. AERATION STUDY The group having oxygen supplied by injecting air through airstones had significantly higher survival rates (P < 0.05) when compared with a treatment aerated by agitation of the water column (Figure 2) . Although the average oxy- gen levels were similar between the two treat- ments (3.4-6.8 ppm), such levels in tanks aer- ated by high-pressure nozzles often dropped for short intervals due to clogging of the nozzles with silt and biological debris. Electrical power failures which affected water flow but not the compressed air supply (equipped with stand-by DC power) also caused intermittent drops in oxygen levels. Such short-term irregularities may have been more critical to shrimp toler- ances than is indicated from reference to average oxygen level values, per se. Also, at the rela- tively high temperatures maintained throughout the study, short drops in oxygen levels could have been very critical. Decreased survival in tanks with agitation of the 0.75-m water column may also have resulted from physical agitation of the animals. lor FISHERY BULLETIN: VOL. 70, NO. 1 100 Figure 2. — Mean and standard error for percentage of animals surviving after 8 weeks of growth with two types of aeration. STOCKING DENSITY STUDY Stocking densities higher than 40 g/m^ pro- duced proportionally higher mortalities indi- cating an appi'oximate carrying capacity for this particular culture system (Figure 3). If shrimp were stocked at 40 g/m-, a population of 32 g/m^ remained after 8 weeks. Similarly, when shrimp 100 80 a 60 3 40 20 '^^■a — -A ^ >" L ^^7 T K^^^^ ^^^ '' 'H 1 o— 0 80 •— • 160 1111 4 WEEKS Figure 3. — Mean and standard error for percentage of animals surviving after 8 weeks of growth at three stocking densities. were stocked at 80 g/m^, a relatively stable pop- ulation of approximately 52 g/m^ existed during the final 2 weeks of study. Similarly, mortality during the first 8 weeks among a population orig- inally stocked at 160 g/m^ created a population of approximately 80 g/m^, but in this case the survival rate was still declining after 8 weeks of growth. Therefore, a carrying capacity (max- imum biomass obtainable) for this culture sys- tem may have been somewhere between 32 and 80 g/m2. Although such a carrying capacity would de- pend on the particular culture system, applicable calculations, utilizing data from a laboratory culture study (Subrahmanyam and Oppenhei- mer, 1969) and a pond culture experiment (Broom, 1969), indicate a similar carrying capacity for populations of other systems (ponds, embayments, and laboratory tanks). In the case of the laboratory study, best survival was obtained when shrimp were stocked initially at 34 g/m^, yielding a biomass of 27 g/m^ at the termination of the experiment. Likewise, best survival and an increase in biomass occurred in the pond culture study when initial stocking den- sities were below 20 to 30 g/m^. Recent data from a commercial operator in Central America indicates that, regardless of stocked biomass, the carrying capacity ranged from 5.5 to 7.3 g/m^ (Smitherman and Moss, 1970). Such evidence suggested that final production expectations should be considered in choosing initial stocking densities. PRELIMINARY NUTRITIONAL EXPERIMENTS A comparison of several groups of diets (Table 1) revealed that semipurified diets with casein as the major source of protein (Group I), only produced an average of 18% increase in biomass above stocked biomass levels. Group II, having fish and shrimp meal as additional sources of protein, produced approximately 63% growth on the best diet. Group III diets comparing hydro- lyzed proteins yielded only 39% growth on the best diet, and animals fed at a rate of 15% of their total biomass (Group IV) increased their initial biomass 164%. In addition to increased 106 SICK, ANDREWS, and WHITE: REQUIREMENTS FOR SHRIMP CULTURE growth above that obtained in the environmental studies, survival was increased from 80 to 90% to 90 to 100% through the information acquired from the above nutritional comparisons. Shrimp obtained little if any sustenance from organic or settled detritus since the continuous flow of filtered seawater through the tanks kept the system relatively free of siltation and ex- traneous growth. Furthermore, starved animals were not able to sustain their initial biomass level beyond 2 weeks, and the populations in all three replicate tanks had died after 7 weeks. Cannibalism appeared to be prevalent among starved organisms, and the decline in weight was undoubtedly moderated due to growth of animals preying upon dead shrimp. Diets in Group I with casein as the major pro- tein source produced little growth above susten- ance. Diet 1 with an added mineral mix yielded a significantly higher biomass increase above ini- tial weights at the 5% feeding level than either Diets 2 or 3 which lacked the mix. In addition to the mineral mix. Diet 3 lacked sodium gluta- mate, glycine, citric acid, and succinic acid and correspondingly caused a loss in biomass over the 3-month growth period. Although the above results showed Diet 1 to be significantly different (P < 0.05) from the other diets after the first month of study, results are somewhat confounded with initial differences in stocked biomasses and poor response in general. These sustenance biomass levels represented far less growth increase than the 72% of control obtained by Kanazawa et al. (1970) and may be due to a lack of cholesterol in our study. Since many crustaceans are not able to synthesize cholesterol (Van Den Oord, 1964; Dall, 1965; Zandee, 1967), including recent evidence for shrimp (Kanazawa et al., 1971) , the lack of this entity undoubtedly was attributed to the poor performance of these diets. Shrimp fed on Group II diets averaged a 37 to 63% increase in growth. Although results from Diets 5, 6, and 7 were not significantly dif- ferent (P < 0.05), Diet 6, which consisted of a high ratio of shrimp to fish meal and a low level of casein, yielded greatest biomass increases. Total biomass decreased in diets having a de- crease in percentage of shrimp meal. Growth from Diet 8, which contained blended shrimp muscle and lower levels of shrimp and fish meal, was statistically less (P < 0.05) than the other three diets. Again, the control group of starved shrimp was not able to sustain its initial weight and declined in biomass after the first 2 weeks. Group III, consisting of yeast, casein, and soy protein hydrolysate diets, produced an average biomass increase of 18 to 39% (Table 1). The combination of diets containing casein, soy, and yeast hydrolysates produced significantly better (P < 0.05) growth than individual hydrolysates. Since results from this group were not better than results after 6 weeks from Diet 5 which was similar except it contained only intact pro- tein, these data indicate that hydrolyzed proteins are not utilized more efficiently than intact pro- teins. Comparing food supplied at 0, 5, 10, and 15% of total biomass using Diet 6 illustrated that growth was directly proportional to an increase in feeding rate (Group IV), and may reflect the natural feeding habit of the species. While the population of starved animals disappeared after 8 weeks, the treatments fed at 5% of their biomass increased 58% over their initial weight; those fed at 10% of their biomass increased 109%; and those fed at 15% biomass gained 164%. The above results indicate that penaeids are capable of consuming large amounts of food. This may be a reflection of their natural tendency to continuously graze upon large quantities of benthic material rather than feed periodically as would a strict carnivore. Although pellets used in all experiments were textured to main- tain consistency in solution for 24 hr, some shat- tering may occur as shrimp gnaw at them and thus some food may be lost through flushing, thus decreasing the efficiency of ingestion as feed levels are increased. Although growth was directly proportional to an increase in feeding rate, feeding at low levels was still justified in attempting to determine nu- tritional requirements of shrimp. The benthic material normally grazed upon is low in energy content and is often of relatively poor nutritional content. Feeding at lower fed levels but with food of proper nutritional value could conceiv- ably produce growth comparable to higher fed 107 FISHERY BULLETIN: VOL. 70, NO. 1 levels of natural or formula diets presently- known. Food conversion ratios (FCR) (weight of food fed for 6 weeks/ weight increase) were cal- culated from results in Group IV (calculated on a dry weight basis). Feeding at 109r biomass yielded an FCR of 6.7 and growth increase of 109 ^r. On the other hand, feeding at the 15 ^f level produced a 164 ''r growth increase and an FCR of 5.5. Such FCR, although not compar- able to those obtained for vertebrates such as the 1.6 or less for catfish (Andrews, in press), nonetheless represent a significant decrease over the FCR of 10 or greater reported for shrimp fed on natural foods (Fujinaga, 1963) . Further refinement of the FCR for penaeids can un- doubtedly be obtained through procurement of a more suitable pellet, better understanding of exact nutritional requirements of specific nu- trients, and more information on ingestion and assimilation phenomena. SUMMARY 1. Environmental conditions yielding 80 to 90% survival in the intensive tank culture of penaeid shrimp encompassed a combination of either no substrate or sand substrate on elevated platforms, air supplied externally by an aeration system, and population density of 40 g/m^. 2. Diets having balanced complements of pro- teins, lipids, carbohydrates, amino acids, fatty acids, minerals, and vitamins produced only sus- tained biomass levels. 3. Diets having 69.5^; of the total diet as shrimp meal produced growth increases of 63%. 4. Examination of soy, casein, and yeast hy- drolysates revealed that a combination of each produced 39 Sr growth increase while an average of 18% resulted from feeding each hydrolysate separately. Hydrolyzed proteins did not yield better growth than intact proteins. 5. Feeding at 5, 10, and 15 s; of the animals' biomass daily yielded directly proportional growth. A growth increase of 164% was achieved with a fish meal and shrimp meal diet fed at 15% of biomass daily. 6. Using semipurified pelleted diets, food con- version ratios were reduced by nearly half of that reported for penaeids feeding on clam and other natural foods. 7. Establishing selected preliminary environ- mental and nutritional requirements for penaeid shrimp resulted in reproducible production of major biomass increase with relatively high sur- vival and low food conversion ratios. 8. Results from these studies have allowed us to design facilities and experiments for future work with environmental and nutritional factors. Development of an inexpensive diet which will yield rapid and maximum growth will be an es- sential requirement for economical production of penaeid shrimp, ACKNOWLEDGMENTS The authors wish to sincerely thank Lee H. Knight and his engineering crew for their night and day effort to establish and maintain the fa- cilities and auxiliary power units that were es- sential for this study. In addition, we are grate- ful to Harry Carpenter and his crew for their efforts in general construction and maintenance of our mariculture facilities. LITERATURE CITED American Fish Farmer & World Aquaculture News. 1970. First cultured shrimp harvested at Florida farm. Am. Fish Farmer World Aquacult. News 2(1): 7. Anderson, W. W. 1966. The shrimp fishery of the southern United States. U.S. Fish Wildl. Serv., Fish Leafl. 589, 8 p. Andrews, J. W. In press. The stocking density and water require- ments for the culture of channel catfish in intens- ively stocked tanks. Foodstuffs. Broom, J. G. 1969. Pond culture of shrimp on Grand Terre Island, Louisiana, 1962-1968. Gulf Caribb. Fish. Inst. Proc. 21st Annu. Sess., p. 137-151. Dall, W. 1965. Studies on the physiology of a shrimp, Meta- penaetis sp. (Crustacea: Decapoda: Penaeidae). IV. Carbohydrate metabolism. Aust. J. Mar. Freshwater Res. 16: 163-180. 108 SICK, ANDREWS, and WHITE: REQUIREMENTS FOR SHRIMP CULTURE Duncan, D. B. 1955. Multiple range and multiple F tests. Bio- metrics 11: 1-42. FUJINAGA, M. 1963. Culture of Kuruma-shrimp (Penaeus japon- icus). Curr. Aff. Bull. Indo-Pac. Fish. Counc. 36: 10-11. Kanazawa, a., M. Shimaya, M. Kawasaki, and K. Kashiwada. 1970. Nutritional requirements of prawn — I. Feed- ing on artificial diet. Bull. Jap. Soc. Sci. Fish. 36: 949-954. Kanazawa, A., N. Tanaka, S. Teshima, and K. Kashiwada. 1971. Nutritional requirements of prawn — II. Re- quirements for sterols. Bull. Jap. Soc. Sci. Fish. 37: 211-215. Lunz, G. R. 1967. Farming the salt water marshes. Proceed- ings of the marsh and estuary management sym- posium, p. 172-177. Louisiana State Univ., Baton Rouge. Perez Farfante, I. 1969. Western Atlantic shrimp of the genus Peyi- aeiis. U.S. Fish Wildl. Serv., Fish. Bull. 67: 461-591. Smitherman, R. 0., AND D. D. Moss. 1970. Fish culture survey report for Panama. Ala. Agric. Exp. Stn., Auburn, 71 p. PB 195 912. SUBRAHMANYAN, C. B., AND C. H. OPPENHEIMER. 1969. Food preferences and growth of grooved penaeid shrimp. In H. W. Youngken, Jr. (editor) Food-drugs from the sea, proceedings 1969, p. 65- 75. Mar. Technol. Soc, Wash., D.C. SuRDi, R. W., AND D. R. Whitaker 1971. Shellfish situation and outlook 1970 annual review. Natl. Oceanic Atmos. Adm. Natl. Mar. Fish. Serv., Curr. Econ. Anal. S-20, 39 p. Van Den Oord, A. 1964. The absence of cholesterol synthesis in crab, Cancer pagurus L. Comp. Biochem. Physiol. 13 : 461-467. ViLLADOLID, D. v., AND D. K. ViLLALUZ, 1951. The cultivation of Sugpo {Penaeus monodin Fabricus) in the Philippines. Philipp. J. Fish. 1(1): 16-28. Wheeler, R. S. 1967. Experimental rearing of postlarval brown shrimp to marketable size in ponds. Commer. Fish. Rev. 29(3) : 49-52. Zandee, D. I. 1967. Absence of cholesterol synthesis as contrasted with the presence of fatty acid synthesis in some arthropods. Comp. Biochem. Physiol. 20: 811-822. 109 METHOD OF DETERMINING CAROTENOID CONTENTS OF ALASKA PINK SHRIMP AND REPRESENTATIVE VALUES FOR SEVERAL SHRIMP PRODUCTS Carolyn E. Kelley' and Anthony W. Harmon^ ABSTRACT An extraction method is described for estimating the amount of carotenoids in pink shrimp. The carot- enoid index is useful as a measure of quality and as an indicator of changes during storage. Values for several shrimp products are reported. The carotenoid content of Alaska pink shrimp is affected by many conditions and can be used as an index of the general quality of canned shrimp and of the changes in quality of frozen shrimp during storage. It has been also used as a factor in determining optimum peeling charac- teristics of shrimp (Collins and Kelley, 1969) and in selecting desirable retorting conditions (Kelley, 1971'). Color differences in shrimp at different seasons and in different areas may be important in harvesting and marketingpractices. The carotenoid in Alaska pink shrimp is pri- marily astaxanthin. Both total astaxanthin and astacin, its oxidation product, are measured by the method to be described, which wa'fe developed for use with frozen Alaska king crab (Ravesi, 1965') and adapted to Alaska pink shrimp which contain more interfering protein and moisture than crab. ' National Marine Fisheries Service, Fishery Products Technology Laboratory, Kodiak, AK; present address: 609 Schoenbar, Ketchikan, AK 99916. ^ Formerly, National Marine Fisheries Service, PMsh- ery Products Technology Laboratory, Kodiak, AK; pre- sent address: Department of Chemistry, Oklahoma State University, Stillwater, OK 74074. \ Kelley, C. E. 1971. Carotenoid content of pink shrimp: Effect of retorting conditions. National Ma- rine Fisheries Service, Fishery Products Technology Lab- oratory, Kodiak, Alaska. (Unpublished manuscript.) * Elinor Ravesi. 1965. Effect of processing and fro- zen storage on the carotenoid pigments of Alaska king crab. Unpublished manuscript filed at NMFS, Kodiak, Alaska. Manuscript accepted August 1971. FISHERY BULLETIN: VOL. 70, NO. I, 1972. METHOD OF DETERMINING CAROTENOID CONTENTS To 50 g of blended meat add approximately 10 g of silica gel and 100 ml of the proper ace- tone solution: 1. 759r acetone for canned shrimp with liquor. 2. 65% acetone for frozen cooked or raw meats. 3. 50 ^r acetone for raw shrimp with shells on. The silica gel, which serves as a filter aid, is not essential but makes subsequent extraction and filtration easier. Blend just enough to ensure complete mixing and filter through a medium porosity fritted glass funnel, maintaining suc- tion until dripping ceases. Rinse container and filter as needed with 50 "^r acetone. Discard colorless filtrate and blend residue about 2 min with 15 to 20 g anhydrous sodium sulfate and 100 ml of 1:1 2-propanol: chloroform. Filter and re-extract with 50 ml solvent one or two times as needed to get a colorless meat. Use 2- propanol: chloroform as rinse solution during these extractions. Transfer filtrate to 500 ml round bottom flask and strip the solvent, using a rotating vacuum evaporator. Add 5 to 10 ml chloroform and evaporate to dryness. Dissolve residue in enough pure cyclohexane to wash sides of flask and add 10 to 15 g aniiydrous sodium sulfate. Let set for a few minutes and filter through sodium sulfate on a fine porosity fritted 111 FISHERY BULLETIN; VOL. 70, NO. I glass funnel, washing: sodium sulfate with cyclo- hexane to remove all traces of color. If filtrate is clear, dilute to 100 ml. If it appears hazy, repeat the filtration, allowing solution to remain in fresh sodium sulfate for a brief time. Read at 474 nm on spectrophotometer, using cyclo- hexane as a blank. The precision of the method was determined by analyzing 11 identical samples in quadrupli- cate on 11 different days. Twenty-two cans of the same code of canned shrimp were blended in a Waring blendor, and the homogeneous mix- ture was sealed in cans and frozen at — 60° F. For each day's sampling, two cans of the frozen mixture were thawed and blended together. Different lots of solvents were used at inter- vals to determine the sensitivity to slight vari- ations in solvents. The solvent lot was not criti- cal but the cyclohexane used in the spectropho- tometer should be carefully distilled within a few days of use. We used a Gilford modification of a Beckman DU spectrophotometer' which gives readings with three place accuracy. The range of absorbance readings was 0.420 to 0.452, the average was 0.436, and the standard devi- ation was 0.008. The carotenoid content, expressed as the carot- enoid index, is a calculated value based on dry weight. The solids content of the shrimp was determined by the Association of Official Agri- cultural Chemists method (Horwitz, 1965: 346), using 5 to 10 g of the blended meat sample and heating at 105° C for 18 to 24 hr. The carotenoid index represents the absorbance (A) in 100 ml of solvent of the carotenoids from 1 g of dry sample, measured in a 1-cm cell. It is calculated as follows: C, = (A^ 474 nm in 100 ml cyclohexane) (100) (50-g wet sample) {% dry weight) The absorbance reading of a shrimp sami)le with average moisture content is roughly 10 times larger than the carotenoid index; there- fore the carotenoid index equivalent of the stan- dard deviation is slightly less than 0.001. The amount of carotenoid can also be ex- pressed as grams of pigTnent/gram tissue by using the extinction coefficient of 2150, as re- ported by Kanemitsu and Aoe (1958). The amount of astacin present in fresh shrimp is small and since the extinction coefficients of astacin and astaxanthin for calculation purposes do not introduce significant error for routine analytical work, the calculation would be: grams pigment/1 g tissue = (A aj^47^nm) (100 ml) 100 (50 g) (d) (2150) wiiere d is the cell width in centimeters. This could be converted to dry tissue weight by multi- plying by the percent of solids in the sample. CAROTENOID CONTENTS OF VARIOUS TYPES OF SHRIMP SAMPLES Table 1 gives carotenoid values of various types of shrimp samples described. Most of the data were collected as part of some other project so these samples are from several lots of shrimp caught at different times of the year. Only those grouped together in the table can be accurately compared with each other. All data, however, represent an average figure for the given sample and may be used to compare types of sample products or processing methods. Tablk 1. — Carotenoid values for 11 shrimp samples. Sampling conditions Carotenoid index Cause of color differences indicated by data ° The use of trade names is merely to facilitate de- scription and does not imply endorsement of a product. 1. Raw tails, shells on 0.237 Raw meats 0.086 2. Whole cooked, hand-peeled meats, frozen 0.112 3. Precooked, machine-peeled meats, frozen 0.086 After 6 months' storage , After 12 months' storage 0.070 0.062 Storage time 4. Precooked, machine-peeled canned 2-day iced, machine-peeled, canned 0.073 0.039 Precook versus Ice held conditioning 5. 1-day not iced, precooked, machine-peeled, canned 0.080 2-day iced, precooked, machine-peeled, canned 3-day iced, precooked, machine-peeled, canned 0.066 0.064 Time of ice holding 112 KELLEY and HARMON: CAROTENOID CONTENTS EXPERIMENTAL PROCEDURE 1. The shrimp were frozen whole as soon as possible after being caught, then were shipped to the laboratory. They were partially thawed, weighed, and separated according to weights. The shrimp used were about 80 count. All were headed and some were hand peeled to obtain meats. The tails with shells on and the peeled meats were refrozen as needed until analyses could be made. 2. Whole cooked, hand-peeled frozen shrimp meats were obtained from a commercial proces- sor. This is the conventional, cocktail style product. 3. Precooked, machine-peeled shrimp were produced under experimental conditions in a commercial plant. Shrimp were landed within 24 hr of catching, held overnight without ice, and precooked at 165° F for 10 sec, 110° F for 2 min, and machine peeled. The meats were col- lected at the end of the inspection belt and frozen in cans without vacuum. Analyses were made within a few days, after 6 months, and after 12 months of 0° F storage. 4. Precooked, machine-peeled canned shrimp were produced as described above except they were routinely retorted. The 2-day iced, ma- chine-peeled shrimp are a standard commercial pack from the same lot of shrimp. 5. The 1-day not iced; 2-day iced ; and 3-day iced, precooked machine-peeled, canned shrimp were also experimentally produced in a commer- cial plant. The 1-day not iced shrimp were held in the wooden boxes in which they were landed. The 2- and 3-day iced shrimp were held in large tanks and ice added as needed to keep them cool. All of these shrimp were precooked at 165° F for 10 sec, 110° F for 2 min, and then routinely peeled and canned. All samples were analyzed using the previ- ously described method of determining carote- noid contents. The averages reported in Table 1 represent 3 to 12 determinations under the given s,ampling conditions. ' Some of the factors which cause differences in the carotenoid content of shrimp are shown in Table 1. These include several processing variations which can be controlled by processors and fishermen. CONCLUSIONS The method of determining carotenoid content described is simple and precise and may be used on a variety of shrimp product forms. The carotenoid index for Alaska pink shrimp varies from 0.267 in raw tails to 0.059 in ice held, machine-peeled canned shrimp. Correla- tion with subjective color rating is quite good (Collins and Kelley, 1969) . At the higher color levels found in raw, hand-picked, or precooked shrimp, small differences are difficult to detect visually and the determination of the carotenoid index becomes even more useful in evaluating samples. Since the carotenoid content is usually closely correlated with other quality characteristics, the carotenoid determination may be useful in making decisions about the best ways to process or handle shrimp. LITERATURE CITED HORWITZ, W. (chairman and editor). 1965. Official methods of analysis of the Associ- ation of Official Agricultural Chemists. 10th ed. Association of Official Agricultural Chemists, Washington, D.C., xx + 957 p. Collins, J., and C. Kelley. 1969. Alaska pink shrimp, Pandalus borealis: Ef- fects of heat treatment on color and machine peel- ability. U.S. Fish Wildl. Serv., Fish. Ind. Res. 5: 181-189. Kanemitsu, T., and H. Aoe. 1958. Studies on the carotenoids of salmon. I. Identification of the muscle pigments. Bull. Jap. Soc. Sci. Fish. 24: 209-215. 113 A DESCRIPTION OF YOUNG ATLANTIC MENHADEN, Brevoortia tyrannus, IN THE WHITE OAK RIVER ESTUARY, NORTH CAROLINA Robert M. Lewis/ E. Peter H. Wilkens/ and Herbert R. Gordy^ ABSTRACT Atlantic menhaden exhibit three different stages — larva, prejuvenile, and juvenile — during their stay in the estuary. For specimens collected from the White Oak River estuary, N.C., the length-weight rela- tion was logg Y = -8^1104 + 3.6050 (log^ X) for larvae, log^ Y = -16.9638 + 6.3083 (log^ X) for prejuveniles, and logg Y = —5.2298 + 3.1452 (logp X) for juveniles, where Y =: weight in mg and A' = length in mm. Larvae and prejuveniles concentrated in the low salinity-freshwater zone upstream. Juveniles tended to move downstream toward the higher salinity water. Condition factors of larvae and prejuveniles increased toward the higher salinity zone. Atlantic menhaden spawn and hatch in coastal oceanic waters from Maine to Florida. The larvae enter estuaries where they transform in- to juveniles near the freshwater zone. The re- lation of young menhaden length and weight to time, salinity, and location within an estuary provides insight on the environmental require- ments of menhaden during a critical phase in their life cycle. We collected young menhaden from a small estuary in North Carolina, from March to September 1969 with a tidal net (Lewis et al., 1970) to study changes in the length- weight relation. The lower portion of the White Oak River estuary (28 sq km) is shallow with depths from 0 to 3.0 m and distances from opposing shores from 1 to 3 km. The intracoastal waterway crosses the lower estuary and is maintained at a depth of approximately 4 m. The upstream portion narrows into a river up to 4.6 m deep. During the study period we generally found that the change from brackish to fresh water oc- curred between 18 and 24 km upstream from Bogue Inlet. The exact location of this low sa- linity zone was influenced by tide, rainfall, and direction and speed of the wind. The mean tidal ^ National Marine Fisheries Service, Atlantic Coastal Fisheries Center, Beaufort, NC 28516. " National Marine Fisheries Service, Southeast Fish- eries Research Center, 75 Virginia Beach Drive. Miami, FL 33149. Manuscript accepted August 1971. FISHERY BULLETIN: VOL. 70, NO. 1. 1972. range at Bogue Inlet is 2.2 ft (0.67 m) (U.S. Coast and Geodetic Survey, 1970). At 21 km upstream the average river flow is 14.7 cfs (0.42 mVsec) (North Carolina State Board of Health, 1954) . A map showing the location of the White Oak River estuary is shown in Lewis and Mann (1971). We selected 14 stations from Bogue Inlet to 34 km upstream (Wilkens and Lewis, 1971). Stations ranged from 2 to 5 km apart and were selected to be representative of the various sa- linities encountered. We also sampled in fresh water to determine how far menhaden move up- stream. Salinity measurements were taken within 1 m of the surface. During collections, spot checks of salinity between the surface and bottom indicated that in this shallow estuary thorough mixing generally occurred. Diff"er- ences between measurements at one location were due to flooding and ebbing tides. Our menhaden collections ranged from 15 to several thousand individuals. We measured and weighed all fish to the nearest 0.5 mm total length and 0.1 mg in collections containing less than 26 and subsampled the larger collections. Since both length and weight variances in the sub- samples were small, we considered our estimates of length and weight to be reliable. Our mea- surements of total length were based on the greatest dimension between the most anteriorly projecting part of the head and the farthest tip 115 FISHERY BULLETIN: VOL. 70. NO. I of the caudal fin when the caudal rays are squeezed together (Hubbs and Lagler, 1949). We separated young menhaden into three stages on the basis of body form and the length- weight relation of individuals within each stage. Length and weight ranges of all the fish used in the study are given in Table 1, An illustra- tion of each stage (larva, prejuvenile, and ju- venile) that occurs during the first year in the estuary is shown in Figure 1. Allometric growth, with stanzas for larvae, prejuveniles, and juveniles is shown in Figure 2. The inflection points, indicating change in slope, are 30 and 38 mm for length, and 70 and 469 mg for weight. We considered specimens less than 30 mm and 70 mg as larvae; they are long and slender, and even at 30 mm total length the body depth is only 4 mm or less. In the next group, Table 1. — Lengths and weights of Atlantic menhaden from the White Oak River estuary, N.C., arranged in order of increasing weight classes. Weight Length range Number of menhaden Weight Length range Number of menhaden mg mm mg mm 0.0- 4.9 8-16 23 100.0- 199.9 29-34 38 5.0- 9.9 14-20 46 200.0- 299.9 33-37 26 10.0-14.9 17-21 43 300.0- 399.9 35-38 6 15.0-19.9 20-23 48 400.0- 499.9 37-43 5 20.0-24.9 20-24 45 500.0- 599.9 39-41 10 25.0-29.9 22-26 32 600.0- 699.9 40-44 14 30.0-34.9 23-26 30 700.0- 799.9 41-44 11 35.0-39.9 24-27 31 800.0- 899.9 44-45 6 40.0-44.9 25-28 29 900.0- 999.9 45-49 4 45.0-49.9 25-29 10 1,000.0-1,499.9 47-54 41 50.0-54.9 26-29 20 1,500.0-1,999.9 53-60 27 55.0-59.9 27-29 26 2,000.0-2,499.9 58-62 16 60.0-64.9 27-31 26 2,500.0-2,999.9 61-66 11 65.0-69.9 27-31 19 3,000.0-3,499.9 68-71 5 70.0-74.9 28-31 22 3,500.0-3,999.9 71-74 8 75.0-79.9 29-32 15 4,000.0-4,499.9 75-77 4 80.0-84.9 29-32 12 4,500.0-4,999.9 76-82 5 85.0-89.9 29-31 7 5,000.0-5,499.9 81 3 90.0-94.9 28-32 U 5,500.0-5,999.9 81-83 2 95.0-99.9 29-32 6 B Figure 1.— Atlantic menhaden (a) larva 27.0 mm total length (TL) ; (b) prejuvenile 32.0 mm TL; and (c) ju- venile, 64.0 mm TL. The alimentary tracts are shown as they were visible in the preserved specimens used in drawings. 116 LEWIS, WILKENS, and GORDY: YOUNG ATLANTIC MENHADEN o o o X o 7.0 10 0 9.0 B.5 8.0 7.5 7.0 6.5 6.0 6.5 6.0 4,5' 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 LENGTH (MM) 20.0 30 0 40.0 60.0 80.0 100.0 10,000.0 I I I I Y = -5. 2298 + 3.1452 X Juveniles Y = -16 9638 + 6. 3083X Prejuveniles Y=-8.1104+3.6050X Larvae I I 5.000.0 260 0 O X 100.0 O . 10.0 I no 0 2.2 2,4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 LENGTH (LOGe MM) Larval menhaden were most abundant in March, prejuveniles in late March and April, and juve- niles by the beginning of May (Wilkens and Lewis, 1971). Large catches of larval and pre- juvenile menhaden within the freshwater-low salinity zone (Table 2) suggest that favorable conditions for growth are present. Condition factors {W/V, where W = weight in mg, L = length in mm, and 5 = value for the slope of weight on length for each growth stanza) of larvae and prejuveniles increased with time as the result of growth and develop- ment. The apparent lack of growth of larvae and prejuveniles in the low salinity-freshwater zone during April is probably due to large num- bers entering this zone, putting on fast growth, moving out of the zone, and being replaced by new groups (Tables 2 and 3). Juveniles, which have the same body form as adults and which are scattered in schools throughout the estuary, showed no change in condition factor with time or salinity. Figure 2. — Regression of weight on length for larval, prejuvenile, and juvenile Atlantic menhaden collected in White Oak River estuary, N.C., in 1969. (We separated the lengths and weights into three groups after visual observation of the data and fish. Lines were then fitted by least squares regression based on data in each group.) prejuveniles, we included specimens from 30 to 38 mm and 70 to 469 mg. In this stage there is a rapid increase in body depth, but little increase in length. Fish above 38 mm and 469 mg we classed as juveniles. Huntsman' found that the relation between length and weight is similar for juveniles and adults. Both stages have a similar body form, only their color and size being dif- ferent. We did not find any adults in our estu- arine study. Larvae enter the lower estuary and move up- stream to the freshwater-low salinity zone where they go through a prejuvenile stage before com- pleting their transformation into juveniles. ' Huntsman, Gene R. 1971. Growth by year class of Atlantic menhaden. (Unpublished manuscript.) NMFS Center for Estuarine and Menhaden Research, Beaufort, NC 28516. LITERATURE CITED HUBBS, C. L., AND K. F. Lagler. 1949. Fishes of the Great Lakes Region. Cran- brook Inst. Sci., Bull. 26, 186 p. Lewis, R. M., W. F. Hettler, Jr., E. P. Wilkens, and G. N. Johnson. 1970. A channel net for catching larval fishes. Chesapeake Sci. 11: 196-197. Lewis, R. M., and W. C. Mann. 1971. Occurrence and abundance of larval Atlantic menhaden, Brevoortia tyrannus, at two North Carolina inlets with notes on associated species. Trans. Am. Fish. Soc. 100: 296-301. North Carolina State Board of Health. 1954. The White Oak River Basin. N.C. State Board Health, PoUut. Surv. Rep. 2, 122 p. U.S. Coast and Geodetic Survey. 1970. Tide tables. East Coast of North and South America including Greenland, 1971. U.S. Coast Geod. Surv., 290 p. Wilkens, E. P. H., and R. M. Lewis. 1971. Abundance and distribution of young At- lantic menhaden, Brevoortia tyrannus, in the White Oak River estuary. North Carolina. Fish. Bull., U.S. 69: 783-789. 117 FISHERY BULLETIN: VOL. 70, NO. 1 Table 2. — The distribution and mean total length (mm) of menhaden by date collected, kilometers upstream from Bogue Inlet, and salinity i%o) in the White Oak River estuary, March-August 1969. Date 1969 Mar. 17 Mar. 27 Apr. 1 Apr. 9 Apr. 16 May 1 May 14 May 27 June 26 July 16 Aug. 27 16 kilometers: Salinity __ 0 1.5 4.1 8.0 5.2 13.7 15.4 1.8 10.2 15.1 Abundance index* 4.3 298.4 48.8 111.0 279.1 O.I 5.2 4.2 4.6 0.3 Mean total length — 28.5 28.9 30.0 29.2 32.1 — 42.4 53.4 58.8 — 18 kilometers: Salinity 1.5 0 0.2 0.1 3.8 1.8 7.4 11.7 0.2 3.0 8.9 Abundance index 9.6 0.4 1,053.4 497.7 526.5 155.2 0.4 1.1 16.3 6.6 0 Mean total length 29.9 — 28.5 27.5 30.8 30.5 — — 43.5 51.0 — 21 kilometers: Salinity 0.2 0 0 0 1.2 4.2 — __ Abundance index 48.7 0 54.2 65.3 0 __ 0.5 — , _- . Mean total length 30.8 0 28.3 27.2 — — — — — — — 24 kilometers: Salinity 0 0 0 0 0 0 1.5 4.4 0 0 0.5 Abundance index 13.3 1.8 6.0 1.5 1,533.3 392.6 0.7 13.9 0.4 1.4 I.O Mean total length 30.1 29.6 29.2 26.6 28.5 29.4 — 42.6 — — ~ 28 kilometers: Salinity 0 0 0 0 0 0 0 0.6 0 0 0 Abundance index 0 0.5 1.4 0.2 0.6 0 O.I 12.1 0.3 0.2 0 Mean total length — — 29.5 — — — — 41.8 — — — 31 kilometers: Salinity 0 __ 0 0 0 0 0 0 0 Abundance index 0 0.3 10.3 0 0 0 1.0 0 Mean total length — — — 29.1 — — — — — — — ^ Abundance index is the number of young menhaden for 100^ of water. Table 3.— ■ Mean condition factors of young Atlantic menhaden collected in the White Oak River estuary, N.C., in 1969. Salinity {%„) 0 0.1-0.9 1. 0-1.9 2.0-2.9 3.0-3.9 4.0-4.9 5.0-5.9 >6.0 Larvae Prejuveniles Juveniles 1969 Mar. 17 Mar. 27 Apr. I Apr. 9 Apr. 16 May 1 Mar. 17 Mar. 27 Apr. 1 Apr. 9 Apr. 16 May 1 May 27 May 27 June 26 July 16^ 0.292 0.293 0.302 0.291 0.332 0.310 0.331 0.317 0.355 0.321 0.358 0.299 0.394 0.376 0.342 0,364 0.381 0.396 0.402 0.497 0.303 0.309 0.326 0.253 0.331 0.326 0.357 0.444 0.305 0.387 0.435 .555 .551 .526 0.314 0.390 0.599 0.540 0.476 0.420 0.517 0.547 0.542 0.516 0.498 0.507 0.535 ' Sample sizes after July 16 were too small to show trends 118 GROWTH OF PREMIGRATORY CHINOOK SALMON IN SEAWATER Bernard M. Kepshire, Jr., and William J. McNeil^ ABSTRACT A potential demand exists in sea farming for premigratory juvenile Pacific salmon that have been acclimated to seawater. This paper reports experiments on growth of premigratory chinook salmon (Oncorhynchus tshawytscha) acclimated to water of 33^r salinity and lower and describes a simple mathematical model to evaluate rate of growth. Although chinook salmon raised in these experiments experienced low mortality in water of high salinity, their growth slowed. Reasons for slow growth at high salinity are discussed. Pacific salmon reproduce in fresh water, but only two species — pink (Onchorhynchus gorbuscha) and chum (0. keta) salmon — survive direct transfer as fry from fresh water to full-strength seawater (Weisbart, 1968). The ocean serves as the early nursery ground for these two spe- cies. The other species — including sockeye (0. nerka) , coho (O. klsutch) , and chinook (0. tshawytscha) salmon — require freshwater nur- sery areas. Juvenile salmon undergo a period of adjust- ment when they first enter the sea in order to regulate water and salts in body fluids and tis- sues. This adjustive phase for chum salmon fry lasts about 30 hr and is characterized by an immediate depression of activity, increased con- centration of salts in body fluids, and dehydra- tion of body tissues (Houston, 1959) . A slightly longer adjustive phase of 36 to 40 hr has been reported for yearling coho salmon (Conte et al., 1966; Miles and Smith, 1968). Early exposure to water of low salinity can "trigger" the physiological adaptation to sea- water of salmon species which typically remain in fresh water for several months as juveniles. Acclimation of premigratory young chinook salmon to water of 30^( salinity by exposing them to gradual increments in salinity has been described by Wagner et al. (1969). Black (1951), Coche (1967), and Otto (1971) found ^ Department of Fisheries and Wildlife, Oregon State University, Marine Science Center, Newport, OR 97365. also that coho salmon fry were better able to tolerate water of high salinity after having first been exposed to water of low salinity. Other evidence suggests that the growth of juvenile coho and chinook salmon is influenced by salinity. Coho salmon fry were observed by Canagaratnam (1959) to grow faster in water of 12 to 18^;, than in fresh water. Otto (1971) reported faster growth of juvenile coho salmon at 5 and 10'/,r salinity than at higher salinities or in fresh water. Bullivant (1961) found no significant diflPerence in growth of juvenile chi- nook salmon in water of 0 and W/U salinity. However, Bullivant's fish grew more slowly at 35/^f salinity than at the two lower salinities. This paper reports comparisons of the growth of juvenile chinook salmon raised in water rang- ing in salinity from 0 to 33/i:f . The experiments were conducted at the Oregon State University Port Orford Marine Research Laboratory, Curry County, Oreg. GENERAL PROCEDURES Two groups of chinook salmon used in these experiments were obtained as eyed eggs from the Fish Commission of Oregon, Elk River Hatchery, in winter 1969. Group I fish were divided into five subgroups of 200 each on Feb- ruary 24 (47 days after hatching). Group II fish were divided into six subgroups of 300 each on March 5 (18 days after hatching). Indi- vidual subgroups were introduced to water of Manuscript accepted August 1971. FISHERY BULLETIN: VOL. 70, NO. 1, 1972. 119 FISHERY BULLETIN: VOL. 70, xNO. 1 increasing salinity according to the schedules outlined in Tables 1 and 2. Both groups of fish received the Oregon Moist Pellet diet. The young salmon were fed five times daily beginning 30 days after hatching. After the fish had attained an average weight of 1 g, the frequency of feeding was reduced to three times daily. They were provided more food than they would consume at each feeding. Fish were raised in 100-gal pl>^vood tanks which were lined with fiber glass. Water was introduced to each tank at the rate of one-half gallon per min. Incoming fresh and salt water were premixed in head tanks to obtain desired salinities. Salinities were calculated from the proportions of premixed seawater and fresh water, and density of water in fish tanks was measured periodically with hydrometers to in- sure that salinities remained at their calculated levels. The first experiment (Group I fish) began on January 8 with newly hatched alevins. Group I fish first received food on February 7, and se- lected subgroups were exposed to water of 9 or lT/(( salinity beginning on February 24. The five subgroups were first weighed on February 27. The experiment ended on May 7. The second experiment (Group II fish) began on February 16 with newly hatched alevins. All six subgroups of fish were first exposed to water of 5 or 9'/, salinity on March 6 while still in the alevin stage, and they remained at these salinities for 18 days. The fish were first fed on March 18 and first weighed on April 7. The experiment ended on May 6. Mortality of the 11 subgroups of fish during the test periods ranged from 0 to 6 /r of the orig- inal number of fish placed in the tanks. Even the maximum mortality {6^r ) was considered to have no appreciable effect on the comparisons of growth. The average wet weight of fish in each sub- group was determined at 14-day intervals from random samples of 30 fish. Excess water was blotted from anesthetized fish before weighing. Fish were weighed separately in a flask contain- ing a known weight of water and were returned to their respective tanks after each weighing. Table 1. — Exposure of Group I chinook salmon to water of increasing salinity. Date of hatch — January 7, 1969. Subgroup Age (days after hatchi water' !ng) at which of given sal fish inlty were placed 47 66 80 la lb 0 9 %. 0 17 0 17 Ic 17 17 33 Id 17 24 33 le 9 24 33 ' Temperature of incoming water averaged 10.7° C for fresh water and 10.8° C for seawater. Table 2. — Exposure of Group II chinook salmon to water of increasing salinity. Date of hatch — February 15, 1969. Aga (days after hatchi ng) at which fi sh were placed Subgroup in wafer' ot given sal ini ty 18 36 54 66 %, %, ^o 0/ AW ila 5 18 18 18 Mb 9 18 18 18 lie 5 18 25 33 lid 5 18 25 33 lie 9 18 25 33 llf 9 18 25 33 ' Temperature of incoming water averaged 11.9° C for fresh water and 12.0° C for seawater. OBSERVATIONS ON GROWTH Growth rate was calculated for each subgroup from the periodic measurements of wet weight. Growth was assumed to be exponential over each period considered, and a value for the daily in- crement in body weight, which can be expressed as a iiercent of body weight per day, was ob- tained from the expression Wt (1 + h) (1) where Wt Wo h t weight at the end of the period, weight at the beginning of the period, the compounded daily increment of body weight, and days. It is convenient to solve equation (1) for (1 + h) by converting the terms to common logarithms and taking the antilog, i.e. log (1 + h) = log Wt log Wo (2) 120 KEPSHIRE and McNEIL: GROWTH OF PREMIGRATORY CHINOOK To clarify the concept of daily increment of body weight, fish that can maintain an increase in body weight of 2.0 and 3.0 "^r per day, for example, will double their weight in approxi- mately 35 and 23 days, respectively. Fish held in water of 0, 17, and IS'/cc salinity grew at a faster rate and were heavier at the end of the experiments than fish of the same age transferred to water of 24, 25, and SS'/cc salinity. The observed mean weight of fish in individual subgroups is plotted against age of fish in Fig- ure 1. Equation (2) can be rewritten in linear form to calculate statistics which are useful for making comparisons of rate of growth among test groups of fish. The linear model is: (log Wt — log Wo) = log (l + h) t (3) Slope of the regression line is given as log (1 + h). This model requires the regression line to pass through the origin since (log Wt — log T^o) = 0 ait = 0. Group I fish were weighed on six occasions over a period of 70 days. We have estimated values of log (1 + h) and h for each of the five subgroups of Group I by calculating the five regressions of (log Wt — log Wo) on t. Because the weight of Group II fish was measured on only three occasions, we have not applied a sim- ilar analysis to the second experiment. Application of regression methods to obser- vations on Group I fish indicates that fish in fresh water and water of 17^f salinity (Sub- groups la and lb) grew at a significantly faster rate than fish exposed to water of 33%r salinity (subgroups Ic, Id, and le). Equations for the five subgroups are given in Table 3 along with the 95 '^r confidence interval estimates of log {1 + h) and the approximate confidence interval estimates of h. Figure 2 shows growth curves for the fastest (Subgroup la) and slowest (Sub- group le) growing fish. The periodic measure- ments of weight are plotted in Figure 2 to show their correspondence with the growth curves calculated by use of equation (1). 30 2.5 - 2.0 1.5 5 1,0 0.5 Salinity O 0% O 17-18%. A 33 7c _1_ _1_ _1_ 50 64 78 92 106 Age (days after hotching) 120 Figure 1. — Growth in weight of experimental subgroups of juvenile chinook salmon. DISCUSSION Chinook salmon used in these experiments were exposed to salt water much earlier in life than they normally would experience in nature. Group I fish were acclimated to high salinity (24%r) 66 days after hatching and 36 days after commencement of feeding. Group II fish were acclimated to high salinity (25'/, o) 54 days after hatching and 24 days after commencement of feeding. There were only 66 deaths (3.7%) among the 1,800 fish which had been exposed to salinities of 24, 25, and 33/{o for periods of 25 and 54 days. The average rate of growth in water of high salinity (24'/cc and above) varied between 2.1 and 2.3% increment in body weight per day. These fish doubled their weight in 30 to 33 days. The average rate of growth in water of low sa- linity (17%f. and 0%o) was 2.6 and 2.7% per day. These fish doubled their weight in 26 to 27 days. Although these experiments demonstrate that juvenile chinook salmon can be acclimated to full- strength sea water in an early age, it is apparent that water of high salinity causes a reduced rate 121 FISHERY BULLETIN: VOL. 70, NO. 1 Table 3. — Regression of (log W^ — log Wq) on time for Group I fish. The approximate 95 "Jf confidence interval estimates of h are taken from the confidence limits of log (1 + h). Sub- group Regression equation 95% confidence limits of log (I + A) Approximate 95% confidence limits of A la (log If^ - log W^) = 0.01 168< lb (log IV ^- log (Cg) = 0.011 18« Ic (log »'j — log ffp) = 0.01006< Id (log ;fj - log W^) = 0.00945< le (log Jfj - log W^) = 0.00880i 0.01168 ± 0.00099 0.01118 ± 0.00084 0.01006 ± 0.00087 0.00945 It 0.00077 0.00880 ± 0.00042 2.7 ± 0.2%/day 2.6 ± 0.2%/day 2.3 ± 0.2%/day 2.2 ± 0.2%/day 2.1 ± 0.1%/day of growth. Reduced growth may come about in part because the young salmon expend more en- ergy to maintain an osmotic homeostasis in water of high salinity than in water of low sa- linity. Chinook salmon blood is isotonic with water of salinity between 10 and 13'/r (Coche, 1967). Houston (1959) thought that the increased ener- gy demands for osmoregulation combined with possible inhibition of electrolyte-sensitive com- ponents of the neuromuscular system might con- tribute to reduced growth of young salmon in water of high salinity. There is the further possibility that endocrine systems which are as- sociated with osmoregulation and growth in water of high salinity are not fully functional in premigratory juvenile salmon (Saunders and Henderson, 1970). O Subgroup la (h = 2.7% /day) 0 Subgroup le (h=2 l7o/day) ~ 3 E 2 2 28 42 56 I 70 Time (days after first weighing) Figure 2. — Calculated growth curves for chinook salmon from subgroups la and le as calculated from equation (1). The observed growth is plotted to illustrate cor- respondence with calculated curves, "h" is the com- pounded daily increment of body weight. The acclimation of premigratory chinook, coho, and sockeye salmon to seawater may find future applications in aquaculture. Possibilities include the early release of young salmon from hatcheries into open ocean pastures to reduce costs of feeding and handling and to increase hatchery production. Other possibilities are to pen young salmon in saltwater bays or estuaries (Garrison, 1965; Mahnken et al., 1970) or to place them in raceways receiving waste salt water from coastal thermal-electric stations (McNeil, 1970). Large-scale aquaculture systems, similar to one under development in the Canadian Maritime Provinces (Gunstrom, 1970), would most likely benefit from early acclimation of juvenile salmon to seawater. The release of premigratory ju- venile chinook salmon acclimated to seawater should also be tested at hatcheries equipped with seawater pumping systems. The eflFects of early acclimation on ocean survival is unknown, but the greater availability of food and space in the ocean than in freshwater conceivably would pro- vide potential advantages to juvenile salmon which had been acclimated to seawater. ACKNOWLEDGMENTS Research on acclimation of juvenile salmon to seawater is administered by the Oregon State University Agricultural Experiment Station. Funds are provided by the National Oceanic and Atmospheric Administration's Sea Grant Pro- gram (Contract No. GH97) and National Ma- rine Fisheries Service (Project No. AFC-55). We wish to express our appreciation to Robert Courtright, Director of the Oregon State Uni- versity Port Orford Marine Research Labora- 122 KEPSHIRE and McNEIL: GROWTH OF PREMIGRATORY CHINOOK tory, for assistance and guidance with this project. We also wish to acknowledge helpful comments on the content and organization of this paper by Dr. Lauren R. Donaldson, Uni- versity of Washington, Anthony J. Novotny, Na- tional Marine Fisheries Service, and Harry H. Wagner, Oregon State Game Commission. LITERATURE CITED Black, V. S. 1951. Changes in body chloride, density, and water content of chum {Oncorhynchus keta) and coho (0. kisutch) salmon fry when transferred from fresh water to sea water. J. Fish. Res. Board Can. 8: 164-177. BULLIVANT, J. S. 1961. The influence of salinity on the rate of oxy- gen consumption of young Quinnat salmon (On- corhynchus ishawytscha) . N.Z. J. Sci. 4: 381-391. Canagaratnam, p. 1959. Growth of fishes in different salinities. J. Fish. Res. Board Can., 16: 121-130. COCHE, A. G. 1967. Osmotic regulation in juvenile Oncorhynchus kisutch (Walbaum). I. The salinity tolerance of 50-day-old fry. Hydrobiologia, 29: 426-440. CoNTE, F. P., H. H. Wagner, J. Fessler, and C. Gnose. 1966. Development of osmotic and ionic regulation in juvenile coho salmon {Oncorhynchus kisutch). Comp. Biochem. Physiol. 18: 1-5. Garrison, R. L. 1965. Coho salmon smolts in 90 days. Prog. Fish- Cult. 27: 219-230. Gjjnstrom, G. K. 1970. Canadian mariculture facility begins oper- ation. Am. Fish Farmer 2(1) : 8-11. Houston, A. H. 1959. Locomotor performance of chum salmon fry {Oncorhynchus keta) during osmoregulatory adaption to sea water. Can. J. Zool. 37: 591-605. AHNKEN, C. V. W., A. J. Novotny, and T. Joyner. 1970. Salmon mariculture potential assessed. Am. Fish Farmer 2(1): 12-15, 27. McNeil, W. J. 1970. Heated water from generators presents fish- culture possibilities. Am. Fish Farmer 1(11): 18-20. Miles, H. M., and L. S. Smith. 1968. Ionic regulation in migrating juvenile salmon, Oncorhynchus kisutch. Comp. Biochem. Physiol. 26: 381-398. Otto, R. G. 1971. Effects of salinity on the survival and growth of pre-smolt coho salmon {Oncorhynchus kisutch). J. Fish. Res. Board Can., 28: 343-349. Saunders, R. L., and E. B. Henderson. 1970. Influence of photoperiod on smolt develop- ment and growth of Atlantic salmon {Salmo salar). J. Fish. Res. Board Can., 27: 1295-1311. Wagner, H. H., F. P. Conte, and J. L. Fessler. 1969. Development of osmotic and ionic regula- tion in two races of Chinook salmon Oncorhynchus tshawytscha. Comp. Biochem. and Physiol. 29: 325-341. Weisbart, M. 1968. Osmotic and ionic regulation in embryos, alevins, and fry of the five species of Pacific salmon. Can. J. Zool. 46: 385-397. 123 EFFECT OF ENCROACHMENT OF WANAPUM DAM RESERVOIR ON FISH PASSAGE OVER ROCK ISLAND DAM, COLUMBIA RIVER Richard L. Majors and Gerald J. Paulik° ABSTRACT The filling of Wanapum Reservoir in 1964 flooded the lower sections of the three fish ladders at Rock Island Dam, 61 km upstream from Wanapum Dam on the Columbia River. To maintain fish passage under the new hydraulic conditions, the lower portions of the center and left-bank fish ladders of Rock Island Dam were rebuilt and a new sequence of spill patterns inaugurated. The effectiveness of these modifications was evaluated by comparing results from a series of tagging experiments conducted in 1964-65 on spring chinook salmon (Oncorhynchus tshawytscha) and sockeye salmon (0. nerka) with the results of similar experiments in 1954-55 before Wanapum Dam was built. These comparisons indicated fish passage over Rock Island Dam had improved substantially between 1954-55 and 1964-65; tagged fish traveled over the dam in a shorter time, and higher percentages of the tagged groups were sighted passing over the dam under postencroachment conditions. Successful reproduction of Pacific salmon {On- corhynchus spp.) and steelhead trout (Salmo gairdneri) requires that sufficient numbers of adults in suitable physical condition reach the spawning grounds. Serious consequences can result from delays en route. Thompson (1945) , for example, showed that very few of the sockeye salmon (0. nerka) that were delayed more than 12 days by the Hell's Gate rock slide (Fraser River, British Columbia) reached their spawn- ing grounds. Thompson also suggested that shorter delays reduced the reproductive capa- bility of the survivors. Similarly, man-made fa- cilities such as hydroelectric dams, even though equipped with fish-passage facilities, can act as barriers and thus delay or otherwise interfere with the migratory behavior of salmonids on their way to the spawning grounds. One of the primary goals of the agencies re- sponsible for conserving the fish resources of the Columbia River is to seek ways of minimizing the eflfects of dams on the migration and spawn- ing success of the river's populations of salmo- nids. Although a variety of solutions to the ^ National Marine Fisheries Service Fisheries North- west Center, 2725 Montlake Boulevard East, Seattle, WA 98102. " Center for Quantitative Science, University of Wash- ington, Seattle, WA 98195. problem of dams impeding the passage of mi- grating spawners have been proposed and a number of these have been tried in the field, the pool type of ladder has proven to be the only practical means of passing large numbers of adult salmonids over the dams on the Columbia River. Many research studies aimed at improv- ing fish passage have been conducted over the past several decades. One result of this research has been the introduction of a number of im- provements in design and operation of the pool- and-weir ladder. In some cases fish passage over ladders can be substantially improved by modi- fication of spill patterns (Leman and Paulik, 1966). The present study was designed to evaluate the eflfectiveness of modifications in the fish lad- ders at Rock Island Dam and changes in the spill pattern which were made after the lower portions of the ladders were flooded by the res- ervoir of Wanapum Dam. This type of problem is apt to become more common as all existing sites for hydroelectric dams are utilized and the reservoir of one dam begins to encroach on the tailrace of the dam immediately above. Rock Island Dam, completed in 1934, was the first dam built on the Columbia River. It is in central Washington and about 725 km above the river's mouth (Figure 1). Originally, the dam Manuscript accepted July 1971. FISHERY BULLETIN: VOL. 70, NO. 1, 1972. 125 FISHERY BULLETIN: VOL. 70. NO. I Figure 1. — The Columbia River and locations important to the present study. was provided with two pool- and weir-type fish ladders, one adjacent to each bank. A third was added near the middle of the dam in 1936. Rock Island Dam was modified during 1951- 53. Six new generating units were added to the powerhouse (on the left side of the dam looking downstream), the reservoir was raised about 3.7 m, and regulating lift gates were installed in spillway bays 16-37 on the right side of the dam (Figure 2). Turbine discharge was in- creased from about 793 to 2,265 m^ per sec, the fish ladders were altered to enable them to function at the new reservoir level, and the at- traction flow at the entrance to the left ladder was increased to counteract the increased dis- charge from the turbines. Although the fishery agencies requested changes at the lower end of the right ladder to provide better entrance con- ditions and additional attraction flow, no imme- diate action was taken to implement these re- quests. The Federal Power Commission, in granting permission for the modification of the dam, re- FiGURE 2. — Rock Island Dam showing locations of the fish ladders, powerhouse, and spillway bays. served the right to require alteration of the low- er end of the right-bank ladder if substantial evidence were presented to show that alteration was necessary to protect runs of anadromous fish. Any such alteration was to begin before December 1, 1960. To determine whether the dam caused loss or delay to the runs and whether loss or delay was associated with the right-bank fishway, tagging studies were conducted in 1954-56. French and Wahle (1966) summarized the results as fol- lows: Point estimates of sockeye salmon losses ranged from 0 to 42 percent. Tagging results (one season only) on spring chinook salmon indicated a loss of fish re- leased below the right bank ladder, but no loss when total tag returns from below and above dam releases were compared; data failed to show the dam caused losses of simimer chinook. Tagged salmon released below the dam were delayed 2 to 4 days. Altering the right bank fishway may cause more fish to use it, but there is no clear evidence that such alteration will reduce overall loss or delay.^ Although the evidence did not indicate that a major overhaul of the right-bank ladder was justified, one relatively minor change was made. In 1956, a concrete wall was built at the entrance to the ladder. This wall replaced a cyclone fence " Seasonal races of chinook salmon (0. tahawytcha) in the Columbia River system are classified as spring, summer, or fall chinook depending on the time of year that the adults enter the river to spawn. 126 MAJOR and PAULIK: ENCROACHMENT OF WANAPUM DAM RESERVOIR which, at high tailwater elevations, appeared to impede the entry of fish into the ladder. The wall also eliminated the surging action across the ladder that occurred under certain combi- nations of spill pattern and tailwater elevation. Even after these changes, fishery biologists continued to voice concern over the effectiveness of the right-bank ladder. Some held that hy- draulic conditions at the entrance to the right- bank ladder impeded fish passage over the ladder. In 1958, in response to the continuing concern about the fish-passage conditions at Rock Island Dam, the owners of the dam — Puget Sound Pow- er and Light Co. and Public Utility District No. 1 of Chelan County — financed a study to determine if fish passage could be improved by manipulat- ing the spill pattern. The study showed that fish could be guided to either the right or center ladders by spilling adjacent to the respective ladder. The experi- ment also indicated, but not conclusively, that when low tailwater elevations prevailed, spilling from gates 35, 36, and 37 (immediately adjacent to the right-bank ladder) confused the fish and interfered with their entry into the ladder (Leman and Paulik, 1966). The construction in 1960-64 of Wanapum Dam, 61 km downstream, brought about a fur- ther change in fish-passage conditions at Rock Island Dam. The lower portions of the fish lad- ders at Rock Island Dam were flooded by Wan- apum Reservoir. This condition, in the judg- ment of the Federal Power Commission, required certain modifications of the left and center fish ladders. The modifications were completed by 1963 — 1 year before Wanapum Reservoir was filled. The Commission also directed the owners of Rock Island Dam and of Wanapum Dam (Pub- lic Utility District of Grant County) to develop, in cooperation with representatives of the U.S. Department of the Interior and the Washington State Departments of Fisheries and Game, a program for the study and evaluation of the further effects of encroachment by Wanapum Reservoir on fish passage at Rock Island Dam. It is important to note that the question at hand was the effect of the encroachment of Wan- apum Reservoir on fish passage at Rock Island Dam and not the effects of Wanapum Dam on fish passage in the broader sense, i.e., passage over Wanapum Dam itself and passage through the newly formed forebay. Representatives of the participating agencies formed the "Rock Island Study Group" and en- gaged the junior author as a consultant to serve as chairman of the group. A major segment of the research program, initiated and supervised by the study group, consisted of a series of tag- ging experiments conducted under postencroach- ment conditions. These experiments were so designed that the results would be comparable to results available from preencroachment tag- ging in 1954-55 (French and Wahle, 1966).' The field work was conducted by experienced personnel of the National Marine Fisheries Ser- vice (formerly the Bureau of Commercial Fish- eries) under the supervision of the senior author. In this paper we describe tagging experiments at Rock Island Dam in 1964-65 and compare the results to those obtained in the earlier (1954-55) study. The primary purpose is to estimate the differences between the times required for sock- eye and spring chinook salmon (0. tshawytscha) to move from tagging sites below Rock Island Dam to the counting stations in the three (left, center, and right) fish ladders before encroach- ment and after encroachment. EXPERIMENT RATIONALE The basic experimental measures obtained from this type of tagging are (1) elapsed time from the release of tagged fish below Rock Island Dam to the sighting of tagged fish as they passed through a counting station near the top exit of the fish ladders and (2) the percentage of each release group passing over the dam. The elapsed times include (a) the time, if any, required for tagged fish to recover from possible effects of tagging, (b) the time required to locate and enter the fish ladders, and (c) the time required to ascend the ladders. A statistical analysis of the preencroachment tagging was employed to de- termine adequate sample sizes and release * French and Wahle also tagged in 1956, but because the tagged fish were released in a different manner and at different locations than in any other year, the 1956 experiments were excluded from our comparisons. 127 FISHERY BULLETIN: VOL. 70. NO, 1 frequencies needed in postencroachment tagging to be do^f certain of detecting a change of one- half day and 99 -^r certain of detecting a change of a full day in elapsed times, if such changes occurred between 1954-55 and 1964-65. If we assume that the basic condition of the tagged fish and the time required for tagged fish to re- cover from possible effects of tagging did not diflfer significantly between the 1954-55 and 1964-65 experiments, it follows that changes in elapsed times could be attributed to the ability of tagged fish to find and ascend the fishways. The efficiency of the fish-passage system at Rock Is- land Dam could thus be compared under pre- and postencroachment conditions. Although travel times were expected to be the most sensitive measure of encroachment effects, it is obvious that any significant drop in the percentage of tagged fish passing over the dam would indicate severe stress under postencroachment conditions. It might seem unrealistic at first to assume that tagged fish recovered from the possible effects of tagging equally well in the pre- and postencroachment phases of the study. One might expect, for example, that tagged fish were released into faster moving water in 1954-55 and into slower moving water in 1964-65 and that, accordingly, the tagged fish required longer to recover from the effects of tagging in the earlier phase of the study than in the latter. If this were true, we might have ended up measuring diflferences in recovery time of tagged fish rather than differences in the efficiency of the Rock Island Dam fish ladders. Although water velocities were not measured at the release sites, velocities measured in a model of Rock Island Dam (Ward, 1965)' were not uniformly diflferent under postencroachment conditions than under preencroachment condi- tions. In fact, velocities at the measuring point nearest the right-bank release site on the simu- lated model were generally higher after en- croachment than before. On the other hand, at the station closest to the left-bank release site, Ward, David A. 1965. Hydraulic model studies of the Rock Island fish attraction facilities. Wash. State Univ., Pullman, Div. Ind. Res., Inst. Technol Res. Rep. 65/9-4.3. Vol. 1—20 p., 29 fig., Append. I-II; Vol. 11—23 fig. (Processed.) velocities were higher at lower river flows and about the same at higher flows after encroach- ment. Observations made during the 1964-65 exper- iments revealed that large numbers of tagged fish tended to remain close to shore in protective eddies. According to French, tagged fish be- haved similarly during the 1954-55 experiments." These observations tend to support the assump- tion that tagged fish recovered from tagging equally well in 1954-55 and 1964-65. The flooding of the lower portions of the fish ladders at Rock Island Dam by Wanapum Reser- voir was not the only factor aflfecting fish pas- sage that changed between 1954-55 and 1964-65. Riprap was added to the left bank of the river below the dam, and the left and center fish lad- ders were modified extensively. Figures 3 and 4 show Rock Island Dam before and after Wana- pum Reservoir had been filled. New spill pat- terns designed to enhance fish passage were in eflfect throughout most of the 1964 and all of the 1965 tagging. The basic pattern was developed from findings of the 1958 study (Leman and Paulik, 1966) and modified slightly after ex- periments with the model of Rock Island Dam (Ward, 1965, see footnote 5). METHODS AND MATERIALS The basic experimental procedure was as fol- lows: salmon were trapped as they passed over the left ladder, transported to the release sites approximately 300 m below Rock Island Dam on either side of the river, then tagged and re- leased. Fish counters at the dam recorded the tags as the tagged fish passed the counting boards after reascending the ladders. TAGGING Two diflferent traps were used to capture the salmon. Sockeye were captured as they entered a trap placed at the upstream edge of the count- ing board in the left-bank fish ladder. Chinook salmon, which would not enter this trap, had to " Personal communication, Robert R. French, Fi-shery Biologist, Natl. Mar. Fish. Serv., Northwest Fish. Cent., Seattle, Wash. 128 MAJOR and PAULIK: ENCROACHMENT OF WANAPLM DAM RESERVOIR Figure 3. — Rock Island Dam before Wanapum Dam had been built. Figure 4. — Rock Island Dam after Wanapum Dam had been l)ui!t. Note how the rocks and the lower portion of the right-bank fish ladder, visible below the dam in Figure 3, have been inundated in Figure 4. 129 FISHERY BULLETIN: VOL. 70, NO. 1 be taken in a larger floating trap positioned at the upstream end of the ladder. A conventional 1,000-gal (3.79-m3) tank truck transported sockeye salmon in 1964 and 1965 and Chinook salmon in 1965 but was not available for the 1964 chinook salmon experiments. In- stead, we used 1.2 m by 1.2 m by 1.1 m plywood boxes, equipped with aeration systems and mounted on -Vfton (680-kg) trucks. These units were suitable for transporting fish to the adja- cent left-bank release site but inadequate for moving more than seven or eight fish per unit to the opposite bank via Wenatchee, Wash., — a 48-km trip that took about 1/2 hr. Because of this limitation, only one-half as many chinook salmon were released on the right bank as on the left bank in 1964. Each batch of fish liberated was distinctively marked. Several types and colors of tags were used. Petersen plastic disks were used either alone or in combination with plastic bars and vinyl streamers. Nickel pins, inserted through the body just below the dorsal fin, provided the attachment. Tags were applied in pairs, so that the same color and type of tag showed on both sides of the fish. Tagging time seldom exceeded 30 sec per fish. ARTIFICIAL MANIPULATION OF SPILL PATTERN, AUGUST 3-5, 1964 The spill pattern throughout most of the 1964 and 1965 tagging was developed from results of experiments at Rock Island Dam in 1958 (Leman and Paulik, 1966) with subsequent refinements from a model study in 1964 and 1965 (Ward, 1965, see footnote 5). On August 3-5, 1964, however, gates 16 to 18 (adjacent to the center ladder) were closed and gates 34, 36, and 37 (adjacent to the right-bank ladder) were opened. This departure from the recommended spill pat- tern was undertaken to measure its eflfect on the passage of tagged fish over the dam. TAG OBSERVATION AND DATA RECORDING Four steps were taken to insure the accuracy of the tag observations: First, hydroscopes (floating "windows") were installed over the counting boards on the right- and left-bank lad- ders to suppress glare and surface disturbance. Second, all fish counters were tested for color blindness. Third, samples of tags were mounted on the tally boards to facilitate instant recogni- tion and recording of the tags. Fourth, fish counters were systematically rotated between ladders to distribute any bias by the counters be- tween the ladders. The gates in the fish ladders were open and the passing fish were counted 16 hr a day — 5:00 AM to 9:00 PM — during these experiments. The half-day units used to measure travel times were adapted to the counting schedule. Fish observed during the same 8-hr period in which they were released were assigned a travel time of one-quarter day or 0.5 half-day. Fish re- leased just before noon (as most were) were given a travel time of 0.5 half-day, if observed the same day, or a time of 1.0 half-day if ob- served the next morning. Tag observations were grouped by tag combination, ladder, and travel time in half days. Data were punched on IBM cards — one card containing the release and recovery data for each fish. The numbers of tagged salmon released below Rock Island Dam and later observed passing the dam in 1954, 1955, 1964, and 1965 are summarized in Table 1. Table 1. — Numbers of salmon that were tagged and re- leased below Rock Island Dam in 1954, 1955, 1964, and 1965, and the numbers and percentages of tagged fish that were later observed passing over the dam's fish ladders.^ Species of salmon Year and ^released l^'elow Tagged fish observed dam passing dam no. no. % Spring chinook 1954 155 60 38.7 1955 157 94 59.9 1964 103 93 90.3 1965 311 285 91.6 Sockeye 1954 1,485 1,176 79.2 1955 1,176 793 67.4 1964 951 895 94.1 1965 679 623 91.8 ' The numbers of spring chinook salmon released in 1954 and 1955 differ from those reported by French and Wahle (1966). They used the July 13 date suggested by Fish and Hanavan (1948) as the termination of the spring run and the beginning of the summer run. We used the scale method described by Koo and Isarankura (1967) to determine that the dotes of least overlap between the two races were July 30, 1954, and July 8, 1955. We believe that our separations, based on the more recent study, are the more accurate. 130 MAJOR and PAULIK: ENCROACHMENT OF WANAPUM DAM RESERVOIR TAG OBSERVATIONS AT ROCK ISLAND DAM, 1964 AND 1965 The 1964 and 1965 data are treated by spe- cies— spring Chinook salmon first, followed by sockeye salmon. In a later section these data will be compared to the 1954 and 1955 results to determine the effects of encroachment of Wanapum Reservoir on fish passage at Rock Island Dam. SPRING CHINOOK SALMON The tag release and tag observation data for 1964 and 1965 are presented in Table 2. In- cluded are the date and location of release, the number of fish tagged, the release area, and the number and mean travel time of tagged fish sub- sequently observed passing Rock Island Dam. Logarithmically transformed data are used throughout this paper for analysis of travel time; means are geometric means. Three times as many chinook (311) were tagged in 1965 as in 1964 (103). The small number released and the short duration of the tagging period in 1964 (May 19-27) can be at- tributed to difficulties in completing the access roads, difficulties with the trapping and trans- portation systems, and a shortage of fish in the left-bank ladder — the ladder where the trap was located. Tagging was from May 16 to June 9, 1965. The 1965 data are grouped, somewhat arbitrarily, into five time periods. Percentage Observed The overall percentages of tagged chinook salmon subsequently observed passing over Rock Island Dam were 90.3 in 1964 and 91.6 in 1965. Variability among the release groups was high, ranging from 67 to 129^; in 1964 and 60 to 115% in 1965. It is noteworthy that the number ob- served exceeded the number released for 6 of the 24 releases over the two tagging seasons. Table 2. — Numbers of chinook salmon that were tagged, released below Rock Island Dam and the numbers and mean travel times of tagged fish that were later observed passing over the dam's fish ladders, 1064 and 1965. Nurr iber and mean travel t ime in half-days of tagged fish observed possi ng dam Number of f ish tagged Period and 1 released be. low dam Left ladder Center ladder Right ladder Total' Year Dote Lefl ■ bank Ri ght bank No. Travel tii me No. Travel time No. Travel time No. Travel time 1964 May 19 17 10 14.8 5 16.8 7 8.5 22 12.8 May 19 ._ 15 6 6.0 6 8.7 1 4.0 13 6.9 May 20 13 __ 7 13.2 6 13.4 2 14.8 15 13.5 May 20 __ 17 6 18.1 3 5.3 4 6.6 13 10.0 May 21 27 __ 9 10.5 7 4.2 4 8.1 20 7.2 May 21 5 1 7.0 3 12.6 0 __ 4 10.9 May 27 9 — 4 9.9 0 -- 2 8.7 6 9.5 1965 May 16 5 1 2.5 3 3.4 1 2.5 5 3.0 May 18 5 __ 2 11.5 2 4.0 1 2.0 5 5.3 1 May 21 19 12 6.8 4 2.7 3 4.0 19 5.1 May 22 41 — 33 5.4 9 15.8 5 5.7 47 6.6 May 23 11 8 7.0 3 5.2 0 - — 11 6.4 May 24 9 __ 4 4.7 0 __ 2 2.5 6 3.8 II May 24 __ 25 17 7.7 6 4.7 3 2.5 26 6.0 May 25 8 _. 5 3.1 2 22.5 2 12.5 9 6.6 May 26 15 — 8 4.8 0 — 1 0.5 9 3.7 May 27 30 23 8.2 4 5.9 5 3.9 32 7.0 III May 28 28 — 16 9.1 4 15.5 5 3.7 25 8.3 IV May 31 21 10 6.7 2 5.2 2 6.5 14 6.5 June 1 __ 27 14 6.4 4 4.1 4 2.4 22 4.9 June 4 10 5 2.7 3 8.0 0 ._ 8 4.1 V June 4 22 11 6.9 2 6.6 5 2.3 18 5.0 June 8 22 12 3.9 3 5.1 5 3.2 20 3.9 VI June 9 13 — 8 6.5 1 20.5 0 — 9 7.3 1 The total number observed ma) ,' exceed th le J lumber tagc ,ed. See text (p. 132) for expl anation. 131 FISHERY BULLETIN: VOL. 70, NO. 1 Possible explanations include the misidentifica- tion of tagfs by the counters and multiple obser- vations of the same tagged fish that passed over the dam. fell back, and survived to pass over the dam ajrain. The "falling: back" of salmon over dams is a frequent occurrence on the Columbia River (Johnson, 1965),' although recent studies^ have shown the magnitude of such fallback is not large (usually less than 5'^r ). The i^erceiitages of tagged fish recovered by release location (releases pooled by location within years) were not consistent for the 2 years. In 1964 the percentage of tagged fish released on the left bank and subsequently observed ex- ceeded that of the right bank— 95.5 to 81.1. In 1965 the comparable percentages were 88.4 and 94.9, respectively. Distribution by Ladder Of the tagged fish sighted in 1964, nearly one- half (46.2 "^r ) chose the left ladder; the center and right ladders lured 32.3 and 21.5^/r , respec- tively. In 1965, 66.3 ^r chose the left ladder, 18.2 vr the center ladder, and 15.4% the right ladder. Distribution between ladders was basic- ally the same for each release site within, but not between, years. The percentages observed in the left ladder were 47.6 and 43.3 for the left- and right-bank releases in 1964 but con- siderably higher (67.2 and 65.5) for fish re- leased from the left and right banks, respective- ly, in 1965. However, in both years similar per- centages of the fish not using the left ladder chose the center ladder (60.0 in 1964 and 54.1 in 1965) . Between-period comparisons are possible for 1965 only when the percentage of tagged fish using the left-bank ladder remained very con- sistent from i)eriod to period, varying only from 61.5 to 69.0. ' Johnson, James H. 1965. Fal'.back of adult chinook salmon at Ice Harbor Dam spillway, May 1964. Final Report to U.S. Army Corps of Engineers for Research Contract No. DA-45-164-CIVENG-63-286. Bur. Commer. Fish., Fi.sh-Passage Research Program, Seattle, Wash., 16 p. (Processed.) * Personal communication with Charles Junge of the Oregon Fish Commission with regard to experiments with tagged chinook salmon at Bonneville Dam during 1970. Travel Time from Release to Observation in Fish Ladders Travel times — by date of release, release lo- cation, and ladder in which the tagged spring chinook salmon were sighted — are presented in Table 2. Because of the small numbers of fish involved in the 1964 tests, their value is limited. The 1965 experiments provided the most sen- sitive analysis of the time required for tagged fish to pass over Rock Island Dam under en- croachment conditions. Results of analysis of variance tests of the hypothesis of no difference in mean travel time between fish released on the right and left banks in 1965 are summarized in Table 3. Regardless of how the data were grouped — whether travel times of the right-bank and the left-bank re- leases were compared period by period, whether adjacent periods were combined, or whether all periods were pooled — no statistically significant diff"erences were found. It is noteworthy, how- ever, that the mean passage time for fish released from the right-bank site was less than for fish released at the left-bank site. Thus, fish released from the right bank were finding and passing over the ladders at least as fast as, if not faster than, those released from the left bank. Table 3. — Analysis of variance tests of the hypothesis that spring chinook salmon, tagged and released on the left bank below Rock Island Dam in 1965, traveled over the Rock Island Dam fish ladders equally as fast as fish released on the right bank below the dam. Periods Mean travel tinne In half-days, all ladders combined (1 f-statistics Left-bank Right-bank and 268 df) releases releases 1 6.081 5.130 0.32 N.S.i II 4.630 6.159 0.93 N.S. III 8.249 7.009 0.29 N.S. IV 6.454 4.940 0.48 N.S. V 4.054 5.048 0.21 N.S. VI II 7.342 3.855 2.03 N.S. {1 /"-statistics and 279 df) 1 and 5.609 5.789 0.03 N.S. III and IV 7.553 6.078 0.84 N.S. V and VI 5.553 4.380 0.50 N.S. (1 /•'-statistics and 283 df) l-VI 6.097 5.486 0.63 N.S. ' N.S. = Not significant. 132 MAJOR and PAULIK: ENCROACHMENT OF WANAPUM DAM RESERVOIR Table 4. — Analysis of variance tests of the hypothesis that spring chinook salmon, tagged and released below Rock Island Dam in 1965, traveled over the right fish ladder at Rock Island Dam equally as fast as those traveling over the left and center ladders. Ladders Mean trave release 1 time in ho areas combi ilf days, ined /■-statistics' compared Left ladder Center ladder Right ladder ■ (1 and 268 df) Right vs. left Right vs. center 6.210 6.929 3.396 3.396 10.60** 9.86** 1 ♦♦ = Highly significant at the 0.01 level, reject hypothesis of equal travel times, and conclude that travel time was significantly less through the right ladder. We also analyzed the 1965 data on a ladder by ladder basis. Although the data, even when compared on a period by period basis, are too limited to provide a sensitive comparison of the mean passage times between fish using the dif- ferent ladders, they did reveal that in every peri- od fish moved over the right ladder faster than over either the center or left ladder. Mean pas- sage times for right ladder versus the left ladder and right ladder versus the center ladder (all periods and both release sites pooled) are com- pared in Table 4." The data have been adjusted for simultaneous tests according to the method described by Dunn (1961). The diff"erences shown in Table 4 are highly significant. Thus the hypothesis that spring chinook salmon trav- eled from the tagging sites over the right ladder equally as fast as over the left ladder is strongly rejected as is the same hypothesis for the right versus the center ladder. In both cases the mean passage times are significantly less for fish using the right ladder. This means that ladder choice was the only observed factor clearly affecting passage time. Similar trends were noted in 1964. Mean passage time over the right ladder was less than for either the left or center ladders. The overall travel time in 1965 (5.8 half-days) was shown by an analysis of variance test to be significantly less than that of 1964 (9.8 half- days) . The F-value for this test was 17.72 with 1 and 165 degrees of freedom. * Because the passage of fish over the right-bank fish ladder had been a source of controversy among fishery biologists, we directed special attention to the right-bank ladder in the present study. SOCKEYE SALMON Fewer sockeye salmon were tagged in 1965 than in 1964. Analysis of the 1964 data revealed that the precision desired could still be achieved if sample sizes were reduced from 100 to 75 fish per release in each year. The tagging season was divided into five periods. With one excep- tion, each of these periods contained releases from the left and right banks. During period IV in 1965, both releases were from the left bank. Tagging was from July 15 to August 5 in 1964 and from July 14 to August 4 in 1965. Tag release and tag observation data are presented in Table 5. Percentage Observed The percentages of tagged sockeye from indi- vidual releases observed passing Rock Island Dam were similar for 1964 and 1965. The per- centage ranged from 84.3 to 120.8 in 1964 and from 81.2 to 98.8 in 1965. Overall, 94.1% of the tagged fish were observed in 1964 and 91.8% in 1965. Percentages observed from left-bank releases were not significantly diflferent from those released on the right bank in either year — 94.8 (left bank) versus 93.3 (right bank) in 1964 and 91.5 (left bank) versus 92.0 (right bank) in 1965. Distribution by Ladder Distribution by ladder was similar in both years. Of the tagged fish sighted, 55.6% used the left ladder in 1964 and 53.1% in 1965. The center ladder took 20.7 and 23.6% and the right ladder 23.7 and 23.3% in the 2 years, respec- tively. For the left- and right-bank releases, respectively, the percentages using the left lad- der were 54.1 and 57.4 in 1964. Comparable percentages were 55.4 and 50.7 in 1965. Fish not choosing the left ladder were fairly evenly distributed between the center and right ladders in both years. The between-period consistency of the per- centage of tagged sockeye salmon using the pre- ferred left ladder was less evident than for chinook salmon. For sockeye salmon, the 133 FISHERY BULLETIN: VOL. 70, NO. I Table 5. — Numbers of mean travel times of sockeye salmon that were tagged, released below Rock Island Dam and the numbers, and tagged fish that were later observed passing over the dam's fish ladders, 1964 and 1965. Period Dote Number of f and released ■ish bel tagged ow dam N umber and mean travel 1 fish observed time in passing half-days of dam tagged Left ladder Cent er ladder Right ladder Totali Year Left bank Right bank No. Travel time No. Travel time No. Travel time No. Travel time 1964 1 July 15 July 16 July 17 76 24 113 21 16 48 2.7 2.2 5.9 16 9 29 2.2 1.9 4.8 40 7 29 2.4 2.0 6.8 77 29 106 2.4 2.1 5.8 II July 21 July 22 July 23 104 70 89 35 59 48 4.9 3.4 5.7 8 16 20 4.9 2.8 3.3 16 15 16 3.0 4.1 4.8 59 90 84 4.3 3.4 4.8 III July 28 July 30 92 80 45 56 1.6 5.0 25 20 1.3 3.3 19 11 3.2 5.5 89 87 1.7 4.6 IV July 31 August 3 85 70 — 53 38 2.3 4.4 16 10 3.4 2.8 13 15 3.3 5.6 82 63 2.7 4.4 V August 4 August 5 52 96 53 26 4.8 4.1 13 6 3.3 2.8 16 15 5.6 3.1 82 47 4.6 3.6 1965 1 July 14 July 15 71 68 38 38 3.1 4.2 8 9 3.8 2.3 23 20 2.4 3.5 69 67 2.9 3.7 II July 19 July 20 81 76 51 43 4.6 4.1 7 6 4.2 2.9 22 23 4.7 2.9 80 72 4.6 3.5 III July 27 July 28 75 75 36 28 3.7 2.5 19 28 2.8 2.1 13 16 3.4 1.9 68 72 3.4 22 IV July 29 July 30 69 66 25 25 4.3 2.0 18 19 4.5 2.7 13 13 4.1 2.6 56 57 4.3 2.4 V August 3 August 4 60 38 29 18 2.0 3.1 19 14 3.0 3.5 1 1 2.5 2.0 49 33 2.4 3.3 1 The total number observed may exceed the number tagged. See text (p. 132) for explanation. percentage varied from 40.1 to 62.8 in 1964 and from 44.2 to 61.8 in 1965. For chinook salmon, the range was 61.5 to 69.0 ''r in 1965, the only year in which adequate data were obtained. Travel Time from Release to Observation in Fish Ladders Travel time by date of release, release location, and ladder for the 1964 and 1965 experiments are presented in Table 5. Following analysis of the 1964 data, we will examine the 1965 exper- iments (p. 135). Analysis of variance tests of the hypothesis of no difference in travel time between left- and right-bank releases (all ladders combined) are given in Table 6. Because there were no re- leases from the right bank in period IV, the test compares the two left-bank releases. Within- release group variances were pooled to form an overall pooled estimate of the variance with 883 degrees of freedom. Because the mean travel times did not differ significantly between the Table 6. — Analysis of variance tests of the hypothesis that sockeye salmon, tagged and released on the left bank below Rock Island Dam in 1964, traveled over the Rock Island Dam fish ladders equally as fast as fish released on the right bank; the testing period IV involves two left-bank releases. Period Release locations compared Mean travel time in half-days, all ladders combined f-stotisticsi (1 and 883 df) Left bank Right bonk 1 Left vs. right 2.331 5.787 72.95 •• II Left vs. right 3.394 4.617 8.71 ** III Left vs. right 1.746 4.619 69.38 ** IV Left (July 31) vs. 2.651 14.55" Left (Aug. 3) 4.350 -- V Left vs. right 4.628 3.604 3.11 N.S. 1 «♦ = Highly significant at the 0.01 level, reject hypothesis of equal travel times, and conclude that travel time was significantly less for fish released from one bank than for those released on the other bank. N.S. = Not significant at the 0.05 level, accept hypothesis of equal travel times. July 15 and 16 left-bank releases, these releases were combined. The July 21 and 23 right-bank releases were similarly tested and combined. Note that in periods I to III, travel times for the left-bank releases were significantly less than 134 MAJOR and PAULIK: ENCROACHMENT OF WANAPUM DAM RESERVOIR for the right-bank releases. In period IV when both releases were from the left bank, the travel time of fish released on August 3 was signifi- cantly higher than that of fish released on July 31. Differences were not significant in period V when fish were released from both left and right banks. There is little doubt then that in periods I to III (when the spill pattern recommended by the study group was in operation) , fish released from the left bank were finding the ladders and trav- eling over the dam faster than those released on the right bank. During the last two periods (V and VI), the flows were intentionally switched from the center to the right and back again. The effects of this change on fish passage are discussed in greater detail in a later section. Further examination of Table 5 reveals two additional characteristics about the movement of tagged sockeye salmon over Rock Island Dam in 1964. First is the consistency of the relative passage times between ladders for fish from a given tag release. This means that regardless of how rapidly or slowly fish from a particular release moved, they did so more or less uniformly at all three ladders. Spring chinook salmon (Table 2) varied much more than did the sock- eye in this respect. Second, in five of six com- parisons fish released on the left bank negotiated the right-bank fish ladder faster than did fish released on the right bank. Considered jointly, these two features suggest that sockeye were capable of rapid lateral movement in the area downstream from the dam and that the passage of fish released on the right bank was somehow delayed whether they chose the left, center, or even the adjacent right-bank ladder. Table 7 depicts the travel times by period and ladder, ignoring release sites. Corresponding statistical tests of the hypothesis of no difference in travel times between the right and center and between right and left fish ladders are included. For these tests, a pooled estimate of the error variance with 880 degrees of freedom was com- puted from within-ladder variances for the 15 groups. Because we tested left versus right and center versus right ladders simultaneously, using the same within-period data, we modified the ^-test to control the type I error according to the method suggested by Dunn (1961). Only one difference is significant. In period III, the mean passage time through the center ladder was less than through the right ladder. In general, however, passage time does not ap- pear to be influenced by the ladder chosen. Analysis of variance tests of the hypothesis of no difference in travel time between fish released on the right and left banks (all ladders com- bined) in 1965 are presented in Table 8. The conclusions from these tests are mixed. Fish traveled over the dam faster from the left-bank release site than from the right-bank release Table 7. — Analysis of variance tests of the hypothesis that sockeye salmon, tagged and relased below Rock Island Dam in 1964, traveled over the right fish ladder at Rock Island Dam equally as fast as those traveling over the left and center ladders. Mean trove 1 time in half-d( 3ys, Period Ladders _ compared release areas combined — <-statistlcsi Degrees of freedom Left ladder Center ladder Right ladder 1 Left vs. right 4.018 3.530 0.981 N.S. (1, 159) Center vs. right — 3.358 3.530 -0.329 N.S. (1, 125) II Left vs. right 4.453 3.887 0.967 N.S. (1, 187) Center vs. right — 3.324 3.887 -0.893 N.S. (1, 89) III Left vs. right 2.989 __ 3.874 -1.494 N.S. (1, 129) Center vs. right — 2.014 3.874 -3.321 * (1, 73) IV Left vs. right 3.042 4.399 -2.043 N.S. (1, 117) Center vs. right — 3.155 4.399 -1.461 N.S. (1, 52) V Left vs. right 4.547 4.211 0.434 N.S. (1, 108) Center vs. right — 3.132 4.211 -1.215 N.S. (I, 48) ' N.S. = Not significant at the 0.05 level, accept hypothesis of equal travel times. * = Significant at the 0.05 level, reject hypothesis of equal travel times, and conclude that travel time through the center ladder was significantly less than through the right ladder. 135 FISHERY BULLETIN: VOL. 70, NO. 1 Table 8. — Analysis of variance tests of the hypothesis that sockeye salmon, tagged and released on the left bank below Rock Island Dam in 1965, traveled over the Rock Island Dam fish ladders equally as fast as fish released on the right bank. Mean travel time in half-days. Period all ladc lers combined /■-statistics' (1 and 613 df) Left-bank Right-bank releases releases 1 2.943 3.674 3.239 N.S. II 4.589 3.543 4.915 • III 3.374 2.182 12.867 *♦ IV 2.363 4.319 25.511 *• V 2.368 3.251 3.841 • 1 N.S. = Not significant at the 0.05 level, accept hypothesis of equal travel times. , . , i i ^■ * = Significant at the 0.05 level, reject hypothesis of equal travel times, and conclude that travel time for fish released on one bank was signifi- cantly less than for fish released on the other bank. •* = Significant at the 0.01 level, reject hypothesis, and conclude as above. site in period I but not significantly so. In peri- ods II and III, fish released on the right bank moved over the dam significantly faster than their left-bank counterparts. In periods IV and V, statistically significant differences were found only in the other direction, e.g., left-bank re- leases were faster than right-bank releases. Thus, there is no clear superiority of one re- lease location over the other. The effect of re- lease location on relative and absolute travel times changed from period to period. In con- trast, tagged sockeye salmon released from the left-bank site in 1964, before the spill was in- tentionally modified, moved past the dam faster than their right-bank counterparts. The main difference between the 2 years seems to be the decreased jiassage time for the right-bank re- leases of 1965 which, for every comparable peri- od, moved over the dam faster than their 1964 counterparts. Overall travel times (3.6 half-days in 1964 and 3.2 half-days in 1965) did not differ significantly despite the better performance by the right-bank releases. Next, it is appropriate to examine the effect of ladder choice on mean travel time in 1965 by period and with release areas pooled. The basic data and the corresponding tests of sig- nificance are given in Table 9. No significant differences were found. Ladder choice did not appear to influence travel time. The same re- sult was noted in 1964. SPILL PATTERN MANIPULATION On August 3, 4, and 5, 1964, spill was shifted from gates adjacent to the center ladder to the gate on the far right side of the dam. During this 3-day period, two groups of tagged fish (Au- gust 3 and 5) were released from the left bank and one (August 4) from the right bank. We will consider the left-bank releases first. The re- lease of August 3 was subjected to 3 days of the modified spill condition, whereas the release of August 5 was subjected to 1 day of the same condition. The left-bank release of July 31 pro- vided a crude "control" (no eflfect of modified Table 9.— Analysis of variance tests of the hypothesis that sockeye salmon, tagged and released below Rock Island Dam in 1965, traveled over the right fish ladder at Rock Island Dam equally as fast as those using the left and center ladders. Period Ladders compared Mean travel time in half-days, release areas combined Lefl ladder Center ladder Right ladder /"-statistics^ Degrees of freedom 1 Right vs. left 3.642 Right vs. center — II Right vs. left 4.354 Right vs. center -- III Right vs. left 3.096 Right vs. center -- IV Right vs. left 2.946 Right vs. center -- V Right vs. left 2.400 Right vs. center __ 2.906 3.577 2.340 3.464 3.201 2.868 3.051 N.S. 117) 2.868 0.004 N.S. 58) 3.639 1.903 N.S. 137) 3.639 0.006 N.S. 56) 2.502 1.766 N.S. 91) 2.502 0.155 N.S. 74) 3.288 0.402 N.S. 74) 3.288 0.080 N.S. 71) 2.236 0.019 N.S. 47) 2,236 0.472 N.S. 33) N.S. = Not significant at the 0.05 level, accept hypothesis of equal travel times. 136 MAJOR and PAULIK: ENCROACHMENT OF WANAPUM DAM RESERVOIR spill) for the two "experimental" releases on August 3 and 5. The percentag-es of the tagged fish observed in the right ladder were 15.9, 23.8, and 31.9 for the releases of July 31 (0-day modified spill), August 3 (3-day modified spill), and August 5 (1-day modified spill), respectively. Compara- ble percentages were 19.5, 15.9, and 12.8 for the center ladder and 64.6, 60.3, and 55.3 for the left ladder. Travel times for the three releases averaged 2.7, 4.4, and 3.6 half-days respectively. Thus, it appears that spilling from the right side tended to attract fish released on the left bank to the right ladder but at the expense of in- creasing overall travel time. The change in the spill pattern did not have a significant effect on tagged fish released on the right bank. For the July 30 "control" re- lease, mean travel time was 4.6 half-days; for the August 4 "experimental" release it was 4.6 half-days. Slightly more fish from the August 4 release (19.5^0 were attracted to the right lad- der than from the July 30 release (12.69r). Most striking is the similarity between the left-bank release of August 3 and the right-bank release of August 4. The percentages of tagged fish using various ladders for the releases of August 3 and 4 were: (see Table 5) left, 60.3 and 64.6; center, 15.9 and 15.9; and right, 23.8 and 19.5. Overall travel times were 4.4 and 4.6 half-days. Travel times by ladder were similar —4.4 and 4.8 half-days for the left, 2.8 and 3.3 for the center, and 5.6 and 5.6 for the right ladder. In summary, the departure from the basic spill pattern tended to attract fish to the right-bank ladder, especially those released on the left bank and in so doing, increased the overall travel time. These experiments support the eflScacy of the basic spill pattern as compared to the other pat- tern tested. COMPARISON OF PREENCROACH- MENT AND POSTENCROACHMENT TAGGING STUDIES The effect of the encroachment of Wanapum Reservoir on fish passage at Rock Island Dam is best measured by comparing the results of the pre- and postencroachment tagging studies. We shall consider spring chinook salmon first, followed by sockeye salmon. Three measure- ments— percentage observed, distribution by ladder, and travel time — provide the basis of our analysis. SPRING CHINOOK SALMON The results of the 1954 and 1955 tagging studies with spring chinook salmon are presented in Tables 10 and 11. Comparable data for 1964 and 1965 are in Table 2. Percentage Observed The overall percentages of tagged spring chi- nook salmon observed passing Rock Island Dam were 38.7 and 59.9 in 1954 and 1955; they were 90.3 and 91.6 in 1964 and 1965. Although some of the significant increase may represent better tag retention or increased survival brought about by improved conditions for fish passage during the postencroachment study, it is likely that the precautions we took to improve the tag observations also were important. It is interesting to note that the greatest in- crease occurred for fish released from the right- bank site. In 1954 and 1955, sightings from right-bank releases were only 37.5 and 51.9%, whereas in the postencroachment years — 1964 and 1965— they were 81.1 and 94.9%. Few fish were released on the left bank in 1954, but the increase in the percentage of tagged fish ob- served for the other years (from 77.6 in 1955 to 95.5 and 88.4 in 1964 and 1965) while sig- nificant, is not as dramatic as for the right-bank releases. Distribution by Ladder The percentages of tagged spring chinook salmon in the left-bank fish ladder were 61.7, 74.5, 46.2, and 66.3 in 1954, 1955, 1964, and 1965, respectively. For the right ladder the percent- ages were 25.0, 16.0, 21.5, and 15.4 for the 4 years, respectively. This means that 13.3% used the center ladder in 1954, 9.6% in 1955, 32.3% in 1964, and 18.2% in 1965. Thus, there was no 187 FISHERY BULLETIN: VOL. 70, NO. 1 Table 10. — Tag release and observation data for spring chinook salmon seen passing over fish ladders at Rock Island Dam, 1954 and 1955.' Dote Tagging location Number of fisli fagged and released Number of fagged fish observed Left ladder Center ladder Right ladder Total 1954 June 23-25 Juno 29-July 2 July 7-9 July 13 July 16-22 July 29-30 1955 June 7-9 June 14-15 June 17-21 June 28-29 June 30-July 1 July 5-6 July 7-8 Right bonk Right bonk Right bank Right bank Left bank Right bank Left bank Right bank Right bank Left bank Right bank Right bank Right bank Left bank Right bank 26 36 57 26 3 7 8 14 13 9 19 15 19 17 43 6 5 7 15 3 1 3 7 7 10 6 10 3 8 16 3 2 1 2 0 0 0 1 1 0 1 0 2 3 1 7 2 5 0 0 1 1 1 2 2 2 0 2 1 4 ily 7-8 Right bank 43 16 1 4 ' The total number observed may exceed the number tagged. See text (p. 132) for explanation. 16 9 13 17 3 2 4 9 10 12 9 10 7 12 21 Table 11. — Travel time of tagged spring chinook salmon from tagging areas below Rock Island Dam to the Rock Island fish ladders. 1954 and 1955. Date 1954 Tagging location Mean travel time in half-days, all ladders combined June 23-25 Right bank 12.3 June 29-July 2 Right bank 7.4 July 7-9 Right bank 14.9 July 13 Right bank 13.8 July 16-22 Left bank 15.8 July 29-30 Right bank 10.6 1955 June 7-9 Left bank 6.5 Right bonk 19.4 Juno 14-15 Right bank 12.4 June 17-21 Left bank 24.2 Right bank 21.8 June 28-29 Left bank 20.3 June 30-July I Right bank 9.2 July 5-6 Left bank 19.6 July 7-8 Right bank 12.8 marked and repeatable difference in distribution by ladder between the postencroachment years, 1964 and 1965, and the preencroachment years, 1954 and 1955. The center ladder took a dis- proportionate share of the fish in 1964 (at the expense of the left ladder), but this was less pronounced in 1965. Travel Time from Release to Observation in Fish Ladders Apparently no large-scale mortalities and no great losses of tags were caused by encroachment (these assumptions are supported by the high percentages of tagged fish subsequently seen passing Rock Island Dam). The overall eflfect of encroachment is then best measured by com- paring the travel times between the pre- and postencroachment tagging studies. Although we have already shown that the 1965 travel time (5.8 half-days) was significantly less than the 1964 travel time (9.0 half-days), this difference does not overshadow the fact that both values are well below the comparable figures (12.4 and 16.0 half-days) for 1954 and 1955, respectively. We can only conclude that travel time of spring chinook has decreased markedly since encroach- ment. SOCKEYE SALMON Results of the 1954 and 1955 tagging studies with sockeye salmon are given in Table 12. Com- parable data for 1964 and 1965 are presented in Table 5. 138 MAJOR and PAULIK: ENCROACHMENT OF WANAPUM DAM RESERVOIR Table 12. — Tag releases, tag observations, and travel times for sockeye salmon seen passing over fish ladders at Rock Island Dam, 1954 and 1955. Period Date Tagging location July 7-8 Right bank July 16, 20 Left bank July 21-22 Right bank July 22-23 Left bonk July 27-29 Right bank July 28-30 Left bank August 3 Right bank August 6-12 Left bank August 5-12 Right bank July 19-20 Right bank July 21 Left bank July 22 Right bank July 26-29 Left bonk July 26-28 Right bank August 2-3 Left bank August 2 Right bank August 5-10 Left bank August 4-1 1 Right bonk Number of fish togged and released Nunnber of tagged fish observed' Mean travel time in half-days 22 23.3 168 12.5 151 10.1 196 8.4 75 7.7 111 7.3 80 6.6 128 4.8 245 6.4 118 7.3 46 7.0 6 14.9 224 7.0 151 5.2 119 7.1 12 3.8 38 7.1 79 7.5 1954 IV 1955 III IV 22 119 155 272 146 246 89 174 262 123 59 46 298 227 129 24 93 177 The number observed may exceed the number tagged. See text (p. 132) for explanation. Percentage Observed The overall percentages of tagged sockeye salmon observed passing Rock Island Dam in 1964 and 1965 (94.1 and 91.8, respectively) were significantly higher than those recorded in 1954 and 1955 (79.2 and 67.4). A similar change was noted for chinook salmon. As we mentioned earlier in discussing the results with chinook salmon, two factors — increased tag retention and improved facilities for observing and reporting tagged fish — probably contributed to the in- creased percentages of tagged fish that were observed. Unlike chinook salmon, for which a greater part of the increased percentage of fish sighted could be attributed to fish released on the right bank, the improvement in sockeye salmon was of the same magnitude for releases made on both banks. Distribution by Ladder Percentages of tagged sockeye salmon in the left ladder were 55.3, 64.3, 55.6, and 53.1 in 1954, 1955, 1964, and 1965, respectively. For the right ladder, the percentages were 13.2, 12.4, 23.7, and 23.3; for the center ladder they were 31.5, 23.3, 20.7, and 23.6. As with chinook salm- on then, the distribution of tagged sockeye salmon by ladder in postencroachment years, 1964 and 1965, did not differ significantly or at least consistently so from that observed in pre- encroachment years, 1954 and 1955. Travel Time from Release to Observation in Fish Ladders Comparisons of the mean travel times by period and area of release for 1954 versus 1964 and 1965 and for 1955 versus 1964 and 1965 are presented in Figure 5. On only 1 of 34 oc- casions was the travel time in the preencroach- ment tagging year less than for the correspond- ing postencroachment tagging year. The dif- ference was not significant. As with chinook salmon then, we found the time required by tagged sockeye salmon to pass the fish ladders at Rock Island Dam was con- siderably less after the onset of encroachment than before. 139 MAJOR and PAILIK: ENCROACHMENT OK WANAPl M D.VM RESERVOIR 25r LEFT-BANK RELEASE AREA — • 1954 — O 1955 * » 1964 A A 1965 RIGHT- BANK RELEASE AREA I II III IV V RELEASE PERIOD M III IV RELEASE PERIOD Figure 5. — Time required for tagged sockeye salmon to move over the Rock Island Dam fish ladders from the tagging areas below the dam, 1954, 1955, 1964, and 1965. SUMMARY AND CONCLUSIONS The lower portions of the fish ladders at Rock Island Dam were flooded by the reservoir of Wanapum Dam. At the direction of the Federal Power Commission, the fish ladders were mod- ified to maintain or enhance fish passage and a study was developed to evaluate the adequacy of the modifications. In 1964 and 1965 over 2,000 spring- chinook and sockeye salmon were tagged and released below Rock Island Dam; their subsequent move- ment over the fish ladders w^as noted. Three features — travel time, the percentage of tagged fish observed, and the distribution of tagged fish by ladder — were compared with data obtained in a similar tagging study in 1954 and 1955. Results clearly indicate that the fish passage over Rock Island Dam was better in 1964 and 1965 than in 1954 and 1955. Travel times were significantly shorter and higher percentages of tagged fish were sighted passing over the ladders under postencroachment conditions. LITERATURE CITED Dunn, 0. J. 1961. Multiple comparison among means. J. Am. Stat. Assoc. 56(293) : 52-64. Fisn, F. F., AND M. G. Hanavan. 1948. A report upon the Grand Coulee fish-mainte- nance project 1939-1947. U.S. Fish Wildl. Serv., Spec. Sci. Rep. 55, 63 p. French, R. R., and R. J. Wahle. 1966. Study of loss and delay of salmon passing Rock Island Dam, Columbia River, 1954-56. U.S. Fish Wildl. Serv., Fish. Bull. 65: 339-368. Koo, T. S. Y., AND A. Isarankura. 1967. Objective studies of scales of Columbia River chinook sa.\rtwn, Oncorhynchus tshawytsclia (Wal- baum). U.S. Fish Wildl. Serv., Fish. Bull. 66: 165-180. LEMAN, B., AND G. J. Paulik. 1966. Spill pattern manipulation to guide migrant salmon upstream. Trans. Am. P^ish. Soc. 95 : 397-407. Thompson, W. F. 1945. Effect of the obstruction at Hell's Gate on the sockeye salmon of the Fraser River. Int. Pac. Salmon Fish. Comm., Bull. 1, 175 p. 140 SCALE FEATURES OF SOCKEYE SALMON FROM ASIAN AND NORTH AMERICAN COASTAL REGIONS Kenneth H. Mosher' ABSTRACT Photographic plates of sections of sockeye salmon scales, with descriptions, and frequency tables of the number of circuli in the freshwater and first ocean zones illustrate the variations in scale features of fish over the range of the species in coastal regions of Asia and North America. Suggestions are also given for using these data to determine the geographical origin of sockeye salmon taken in offshore areas of the North Pacific Ocean and adjacent waters. Sockeye salmon {Oyicorhynchus nerka) are val- uable food fish of the Bering Sea and the north- ern part of the North Pacific Ocean. They spawn in coastal streams of Asia and North America but spend a portion of their lives feed- ing in oceanic areas. Upon the onset of sexual maturity, they migrate from the ocean, enter their natal streams, spawn, and then die. Be- cause sockeye return to natal streams to spawn, the species is divided into hundreds of individual populations (each from its own geographical area), which are self-reproducing units or "stocks." A major goal in fisheries research and man- agement of the sockeye salmon resource is to obtain enough spawning fish within each stream to provide the maximum catch to the fishery and to insure the perpetuation of each stock. This goal is difficult to attain in fishing areas where management agencies are uncertain of the geo- graphic area of origin of the stocks of fish that are being caught. Consequently, methods for de- termining the area of origin of sockeye salmon taken beyond their natal streams are needed. A number of methods have been used to de- termine the area of origin of sockeye salmon taken in offshore and coastal areas. These in- clude morphological studies (Fukuhara et al, 1962; Landrum and Dark, 1968), parasitologi- cal studies (Margolis, 1963), serological studies ' National Marine Fisheries Service, Northwest Fish- eries Center, 2725 Montlake Boulevard East. Seattle, WA 98102. (Ridgway, Klontz, and Matsumoto, 1962), tag- ging studies (Hartt, 1962, 1966; Kondo et al., 1965), and scale studies (Krogius, 1958; Kubo, 1958'; Kubo and Kosaka, 1959'; Henry, 1961; Mosher, Anas, and Liscom, 1961; and Mosher, 1963, 1968). Scale studies have become one of the most popular and successful methods; scale features, for example, are routinely used by in- vestigators of the International Pacific Salmon Fisheries Commission as one element in a tech- nique to determine the natal streams of sockeye taken near the mouth of the Fraser River and are also routinely used by investigators of the National Marine Fisheries Service (NMFS, formerly the Bureau of Commercial Fisheries) to determine continent of origin of sockeye salm- on taken in the Bering Sea and the central North Pacific Ocean. No detailed information, however, has been published on the variations in scale features among fish from diff'erent spawning regions along the Asian and North American coasts. Krogius (1958) specifically mentioned the need for an atlas illustrating scales from difl["erent Manuscript accepted September 1971. FISHERY BULLETIN: VOL. 70, NO. I, 1972. " Kubo, T. 1958. Study of sockeye salmon stocks by means of the growth pattern of scales (preliminary re- port). Fac. Fish., Hokkaido Univ. (Hakodate). Part I - 15 p. of Japanese text; Part II - 2 pi., 16 fig. in English. (Transl. of Part I, Int. North Pac. Fish. Comm. Doc. 206), 9 p. (Processed.) ^ Kubo, T., and J. Kosaka. 1959. A study of 5,^ age group red salmon stocks by scale growth formula [in Japanese with English abst., headings, tables, and fig. legends.] Suisan cho (Fisheries Agency of Japan), (Int. North Pac. Fish. Comm. Doc. 826), 27 p. (Processed.) 141 FISHERY BULLETIN: VOL. 70, NO. 1 areas and included in her paper were many pictures of scales of sockeye salmon of Asian stocks. A photographic atlas of sockeye salmon scales (Mosher, 1968) was the first step in de- termining racial origins, and it should be avail- able for reference when the present paper is studied. The purpose of this paper is to show varia- tions in age and scale characteristics among adult fish from various coastal areas over the range of the species so that workers planning to col- lect and analyze scale data to determine origin of sockeye taken at sea and in coastal waters are informed about scale features that are linked to various geographic localities. This report consists of two principal parts. The first comprises (1) photographs of sections of scales of adult sockeye salmon, as plates, for each freshwater age group from various areas over the range of the species; (2) frequency tables of the number of circuli in the freshwater and first ocean zones for fish taken from the var- ious areas; and (3) descriptions of the scales of sockeye salmon from the various areas. The second part is concerned with the selection of scale features for the determination of the origin of fish taken in offshore waters. METHODS AND MATERIALS In the preparation of this paper I was con- cerned with (1) the selection of scale samples of fish from various geographical areas, (2) the selection of scale features that are linked to var- ious stocks or geographical areas, and (3) the method of preparation of plates from photo- graphs of selected sockeye scales. I have dis- cussed each of these items separately. SELECTION OF SCALE SAMPLES OF FISH FROM VARIOUS GEOGRAPHICAL AREAS An important consideration in deciding which stocks of sockeye salmon to include in this paper was the relative number of fish produced in the various localities over the range of the species shown in Figure 1. Study of catch data seems to be the best way to determine the most abun- I50°E 60°N — 60°N Figure 1. — Approximate range of sockeye salmon in and around the North Pacific Ocean and adjacent seas. The distribution in the northern Bering and Chukchi Seas was estimated to include the northernmost known spawning streams on both continents. Sockeye salmon may be found in many streams within the range shown, but in only a few streams in some areas. Atkinson et al. (1967) shows detailed maps of streams where sockeye salmon have been known in the United States. The distribution at sea varies within and between years, depending on many factors, Manzer et al. (1965), Hartt (1962, 1966), and Kondo et al. (1965). In addition to the above refer- ences, Hanamura (1966, 1967) and Aro and Shepard (1967) were also sources of data for this figure. dant stocks because the catch is roughly propor- tional to the production of fish in an area. Table 1 shows the average catch for the 3 years, 1966- 68, the statistics for which are complete for Asia Table 1.— Sockeye salmon catch, average of 1966-68. Area Thousands of fish Metric tons Total all areas 27,297.2 70,387.0 Asia 9,512.6 20,972.6 Japan' 8,527.0 17,988.6 USSR 985.6 2,984.0 North America 17,784.6 49,414.4 Canada (British Columbia) 5,676.4 United States 12,108.2 North of Bristol Bay 2.8 Bristol Bay- 5,715.0 Alaska Peninsula^ 1,625.5 Cook Inlet 1,465.7 Copper River area* 832.2 Southeastern Alaska^ 952.1 Washington and Oregon" 1,514.9 ' Japan has no stocks of sockeye salmon. - Includes north side of Alaska Peninsula. ■' Includes Aleutian Islands, south side of Alaska Peninsula, Chignik, and Kodiok Island, ' Includes Resurrection Bay, Prince William Sound, Copper and Bering Rivers, ^ Includes Yakutat. " Includes the Columbia River (32.2 thousand fish). Source of data: International North Pacific Fisheries Commission, 1966, 1967, and 1968; and supplemental catch statistics supplied to the INPFC by the USSR. 142 MOSHER: SCALE FEATURES OF SOCKEYE SALMON and North America from International North Pacific Fisheries Commission (INPFC) sources. The approximate ranking on each continent of the importance of the coastal areas is as follows: Asia (from Hanamiira, 1906 and Krogins, 1958) Ozernaya River 1 Kamchatka River 2 Bolshaya River 3 Paratunka River 4 Apuka River 5 Okhota and Kukhtuy Rivers 6 North America {from catch data of Table 1) Bristol Bay 1 British Columbia 2 Alaska Peninsula 3 Washington and Oregon 4 Cook Inlet 5 Southeastern Alaska 6 Copper River area 7 Columbia River 8 North of Bristol Bay 9 The abundance and catch of sockeye salmon in most areas can fluctuate widely between years, however. Some of these variations in catch are revealed in Appendix Table 1 of the catch of the 5 years, 1964-68, which series includes one high production year in Bristol Bay — 1965. The distribution of spawning streams in Asia extends from approximately lat 66° N near the Anadyr River southward to the tip of the Kam- chatka Peninsula and the Kurile Islands, and westward to the Okhota and Kukhtuy Rivers on the northern coast of the Okhotsk Sea (Hana- mura, 1966, 1967). Berg (1948) indicated that the species was very rare in northern Hokkaido Island. Krogius and Krokhin (1956) concluded that approximately 90 ';r of the total sockeye catch along the Far Eastern Coast of the USSR was produced in the Ozernaya and Kamchatka Rivers of the Kamchatka Peninsula. The distribution of North American spawning streams extends from the Noatak River of Kot- zebue Sound in Northern Alaska, southward to the Columbia River of Oregon and Washington (Aro and Shepard, 1967; Atkinson et al., 1967). The streams can be conveniently grouped into three major geographical areas for study: (1) the Columbia River to and including British Columbia: (2) Bristol Bay, Alaska, and areas north of Bristol Bay; and (3) the area between Bristol Bay and British Columbia. In many years each area contributes about one-third of the North American catch. Normally, the catch of sockeye salmon north of Bristol Bay is insig- nificant in relation to the number of fish taken in the Bay, but consumption and barter of salm- on is substantial, especially by residents along the Kuskokwim River. Thus, plates of representative scales of fish from southern Kamchatka, Bristol Bay, and the areas north of Bristol Bay, central and south- eastern Alaska, British Columbia, and the Co- lumbia River (the coastal areas listed in Table 1) are included in the first part of this paper. The scale samples used in a previous study (Mosher, 1968) with a few samples, which have recently become available from additional areas, were used for this study.' The areas from which these samples were collected are listed in Table 2. Figure 2 shows the approximate location of the areas mentioned in the text and on the plates, SELECTION OF SCALE FEATURES LINKED TO VARIOUS STOCKS OR AREAS My previous paper (Mosher, 1968) shows in detail the features of sockeye salmon scales and the range of variations in many characters. This paper continues the study of sockeye salmon scales to show the relation of many of the var- iations to locality and how these variations in scale characters can be used to identify the main- land origin of sockeye salmon taken in offshore waters. A number of age groups have been found in all populations of sockeye salmon that have been studied. These age groups are based on the number of years the fish lived in fresh water and in the ocean. Over the geographical range of the species, individuals with scales showing freshwater ages from 0. to 4., ocean ages from .1 to .4, and total ages of 0.1 to * Contributions of the following agencies to the salmon scale sampling program are gratefully acknowledged: The Alaska Department of Fish and Game, Juneau, Alaska; the Fisheries Research Board of Canada, Na- naimo, B.C.; the Fisheries Agency of Japan, Tokyo, Japan; and the Fish Commission of Oregon, Portland, Oreg. In addition, special thanks are given to Dr. I. Lagunov of the Pacific Institute of Fisheries Research and Oceanography (TINRO), Petropavlovsk. Kamchat- ka, USSR, who kindly supplied a series of samples from USSR streams. 143 FISHERY BULLETIN: VOL. 70, NO. 1 Table 2. — Geographical areas wliere scales were taken from sockeye salmon. (Scales are available at the Na- tional Marine Fisheries Service, Northwest Fisheries Center, Seattle, Wash.) Asia 11. Alaska Peninsula area!' 16. Ketchikan orea- 1. Kamchatka River A. King Cove A. Portland Canal 2. Paratunka River A. Dalnee Lake B. Chignik C. Kodiak Island B. Moira Sound (1) North Arm 3. B. Blizhnee Lake Bolshaya River (1) Karluk (2) Red River (2) Kegan Creek 4. Ozernaya River (3) Frazer River C. Karta Bay 5. Okhtosk Sea 12. Cook Inlet orea- D. DolomI Lake and Stream 6. North Pacific Ocean and Bering Sea from A. Cook Inlet Fishery E. Hugh Smith Stream areas near the Kamcha tkc 1 Peninsula B. Kenoi River F. Clarence Strait C. Susitno River G. Eek Boy North America (1) Fish Creek H. Hetta Bay Nicholas Bay 7. Nome area A. Salmon Lake (2) Judd Lake (3) Alexander Creek 1. B. Unalakleet River D. Kasilof River J. Nichols Bay 8. Kuskokwim River E. Fish Creek, Knik Arm, Kenai Peninsula K. Klowock Creek 9. Bristol Bay A. Togiok River F. Skilok Lake G. Tustamina Lake (Bear Creek) L. M. Klokas Lake Deweyville B. Nushagak-Wood River System' H. Upper Russian Lake 17. British Columbia C. Kvichak Riveri 13. Copper River" A. Nass River D. Naknek River' A. Haley Creek B. Skeeno River (1) Brooks Lake 14. Yakutat C. Rivers Inlet (2) Branch River A. Situk River D. Smith Inlet E. Egegik River' 15. Petersburg area- E. Alert Bay F. Ugashik River' A. Kasheets F. Nimpkish River G. Bear River B. Stikin© River G. Eraser River 10. Aleutian Islands C. Salmon Boy 18. Columbia River A. Attu D. Tohlton Lake A. Main stem B. Adak E. Red Bay B. Wenatchee River C. Unalaska F. Port Houghton C. Okanogan River ' Samples from the five major Bristol Bay rivers were taken each year since 1954. but there is only one sample for some of the less important areas. - Additional samples were taken, but numbers of scales were small. There ore samples for most other areas for a number of years. ""3.4 occur, but in most localities most fish are in affe groups 1.2, 1.3, 2.2, and 2.3. The number of years spent in fresh water varies within and between many spawning areas and influences many of the scale characters (Mosher, 1963, 1968). Only a few adult fish of age 0. or 4. are found, some fish of age 3. are found in a few areas, but fish of age 1. and 2. are present in substantial numbers in most areas. Table 3 shows the percentage freshwater age composition of sockeye salmon in stream samples (except as noted) from areas around the North Pacific Ocean available at the NMFS Northwest Fisheries Center. Many features of sockeye salmon scales can be used in racial studies. Scientists at the NMFS Northwest Fisheries Center in Se- attle have examined about 50 diff"erent features ■^ Age designation follows the European system, Koo (1962a) : the number of winters the fish spent in fresh water since hatching, a decimal point, and the number of winters the fish spent in the ocean (see Mosher, 19G8, p. 259 and 262). such as counts of circuli, measurement of zones and portions of zones, and various ratios based on these counts and measurements; but only scale characteristics with the greatest differ- ence between Kamchatkan and Bristol Bay stocks have been described in our publications (Mosh- er et al., 1961; Mosher, 1963, 1968; Anas, 1964; Anas and Murai, 1969). In all of our studies — published and unpublished — the best features for racial studies have been in the fresh- water and first ocean zones of the scale. Because of the large number of Asian and North American spawning areas and the large number of age groups in some areas, it is evi- dent that space is not available for examples of scales representative of each area and age group over the geographic range of the species. Scales from numerous areas are similar in many char- acters; consequently I will group together scales from fish of more than one stock or spawning area that have a relatively similar ap])earance. The number of circuli in the freshwater and first ocean zones (counted as indicated on p. 36 144 MOSHER: SCALE FEATURES OF SOCKEYE SALMON 60°N 60°N 50° Figure 2. — Approximate 1. Okhota River 2. Kukhtuy River 3. Okhotsk Sea 4. Bolshaya River 5. Ozernaya River 6. Paratunka River (Dalnee and Blizhnee Lakes) 7. Kamchatka River 8. Apuka River 9. Anadyr River 10. Attu Island 11. Adak Island 12. Unalaska Island location of areas mentioned in the text and on the plates 13. Bear and Sandy Rivers 25. Kodiak Island 14. Ugashik River 26. Cook Inlet 15. Egegik River 27. Copper River area 16. Naknek River 28. Yakutat Bay 17. Kvichak River 29. Petersburg area 18. Nushagak-Wood River 30. Ketchikan area system 31. Nass River 19. Togiak Bay 32. Skeena River 20. Kuskokwim River 33. Rivers Inlet 21. Norton Sound- 34. Smith Inlet Yukon River 35. Nimpkish River 22. Kotzebue Sound 36. Fraser River 23. Noatak River 37. Columbia River 24. Chignik Bay and 37 of Mosher, 1963) are shown for the var- ious areas as frequency tabulations in Tables 4 and 5 for the age 1. fish; in Tables 6 and 7 for the age 2. fish; and in Table 8 for the age 3. fish. Inspection of Tables 4 and 6 shows that the mean number of circuli in the freshwater zone varies among some geographical areas, but that in the first ocean zone (Tables 5 and 7) there is a cline in number of circuli from least in the Adak Island fish to most in the central British Colum- bia areas of Rivers and Smith Inlets and the Nimpkish River. A decrease in the mean num- ber of circuli among stocks from central British Columbia southward to the Columbia River is also found. The Asian fish and those north of Bristol Bay have slightly more circuli, on the av- erage, than those of Adak Island and Bristol Bay. The scales from sockeye taken from certain geographic areas have similar frequency distri- butions of circuli in the freshwater or first ocean zone. These similarities are the basis for di- viding the coast of North America into certain broad areas. When I discuss the various geo- graphic areas, reference will be made to the appropriate frequency table. PREPARATION OF PLATES The scale plates for the report were made as follows: (1) The scale images produced by a scale projector like the one I described (Mosher, 1950) , at 82 x magnification, were photographed with a 35-mm single lens reflex camera on me- dium speed, fine grained film and processed to accentuate the contrast by minimum exposure 145 FISHERY BULLETIN: VOL. 70. NO. 1 Table 3. — Percentage freshwater age composition of samples of sockeye salmon from geographical areas of the North Pacific Ocean from those listed in Table 2. Locality Yeari Freshwater age Number in 0. 1. 2. 3. 4. sompla ~ — _ Percent - Asia* Ozernaya R. 1959 0.8 53.3 45.1 0.8 124 Bolshaya R. 1964 1.2 94.1 3.5 1.2 — 84 Kamchatka R. 1964 2.2 90.2 6.5 1.1 — 92 Parotunka R.: Dalnee Lake 1964 __ 2.9 44.3 52.8 — 70 Blizhnee Lake 1958 0.6 78.5 20.3 0.6 172 Okhotsk Sea 1957 51.5 42.1 6.4 — 501 Off S.E. Kamchatka 1965 — 21.9 59.2 16.9 2.0 201 North America: Kuskokwim 1959 70.5 28.7 0.8 — 122 Bristol Boy: Ugashik R. 1966 39.6 60.4 — 318 Egegik R. 1966 5.9 86.9 7.2 __ 305 Naknek R. 1966 32.0 67.2 0.8 — 356 Kvichak R. 1966 4.0 95.7 0.3 351 Nushagck R. 1966 94.7 5.3 __ — 322 Togiak R. 1955 3.3 80.1 16.6 __ __ 307 Bear R. 1959 __ 1.6 95.0 3.4 119 Aleutian Islands: Attu Isl. 1956 1.3 38.8 52.9 6.6 0.4 227 Adak Isl. 1956 _. 30.6 60.5 8.0 0.9 213 Unalaska Isl. 1956 38.9 37.5 22.9 0.7 144 Alaska Peninsula: King Cove 1957 45.7 52.7 1.6 __ 182 Chignik R. 1957 __ 40.0 59.3 1.7 175 Karluk R. 1959 48.1 47.2 4.7 106 Red River 1959 _. __ 85.1 14.9 107 Cook Inlet 1959 _. 83.9 16.1 118 Cook Inlet Fish Cr. 1959 __ 98.3 1.7 120 Southeastern Alaska: Copper R. 1959 __ 87.7 12.3 __ 122 Yakufat R. 1958 __ 31.5 65.6 2.9 105 Petersburg 1964 4.6 81.8 13.0 0.6 323 Ketchikan 1964 0.3 84.1 15.6 __ 352 British Coiunnbio: Mass R. 1964 35.0 65.0 __ __ 320 Skeena R. 1964 ._ 98.7 1.3 297 Rivers Inlet 1961 __ 98.0 2.0 248 Smith Inlet 1961 _. 100.0 .. __ 255 Nimpkish R. 1967 __ 64.5 30.4 5.1 79 Fraser R. 1964 0.7 98.0 1.3 __ 153 Columbia River 1964 0.2 88.1 11.7 — — 416 ' The year selected is the most complete for that area, is in a series with adjacent areas, or is the only year a sample is available in our series (Table 2). ^ Historical data on age composition of samples from Ozernaya, Bolshaya, and Kamchatka Rivers, and Lake Dalnee for a number of years from 1931 to 1960 are available in Hanamuro (1966). and maximum development (Adams, 1952; Mortensen, 1947). (2) Positive prints of sec- tions that showed the important features of the anterior field of the scales were made from these negatives on high contrast enlarging pa- per. (3) The positive prints of the scale sec- tions were assemV)led by area groups on mount- ing sheets and photographed to provide the plates. The scales were all photographed at the same magnification. Consequently, the rel- ative size of the scale features on the plates reflects the relative size of the scale features themselves. As indicated in my previous paper (Mosher, 1968), the texture, contrast, and distinctness of circuli vary greatly both on individual scales and between the scales of the same and different fish. Some scales can be photographed to show the features clearly; other scales, especially from some localities with many closely spaced and broken circuli, do not provide clear photographs of all features. 146 MOSHER: SCALE FEATURES OF SOCKEYE SALMON o m u >> w 'u > o 'o o C3 CO -^ o o w ID bo c3 o u 0) c o ;-l O) .—I CO CO 3 «H O c _o 3 jQ 'C -1^ to •i-H >. u c Ol 3 cr a> »-i «H (D bo 03 -1^ C O) o ;-! O) PL. ■J 0-" 5^ E 1^ n) "1 iL V ,*_ u >- , 0) a 1) n U u O 0) 0> IE U to D C "D D u- O 2 DO CQ_c — t Oei O D > »0 CN <> O O 00 p "O "O lO p ^ ■^' CO CN O ^o o* CO to CO CO o ^ "vr o 00 CO ' o I lO o ' O CO »0 »0 iq »0 lO O --- Ci CN oi C^^ lO CN Qs O; CN ^ •— »0 >0 to ^ -o cs d CO v6 CN ^ 00 N; o* rx — ' CO "^ -^r <> UO lO o p d — ^ CS CN T t I I rs CO o p d CN r< CO CO^CO^cq—lCO*— CO 00'-;fOrxCMC^CO I dddd — cs'^'o'^ 'Oi^-^dwo-^d ' O"— oi'^o-^^Jcod dcoTir — roco'^Tro^'^o^corvco ■^■^uoo^o^ioiocs'o o.'o^'Oio'^co'^dco<> ^•^ojdddddd uo IT) O "O ^ "0 I I I [ CO lO S3 CO 00 — ' ' ' ' O *— ^ CN O lO'OOiOO'OOpio I * * lo-^wSco-^-^coKco^ d ' — (N CS — o^oo^ooooooio UOOOsQWOCMpCO^ csdco'O'oco — d I I I C'jcoco"^oq»-;pv> I I I I — ^C'JTroov^'O^d ' ' ' ' >OCNlO^C00'0;UOiO S^CM — CSP'— ;0^'OCM0v ' ' ' 'ocof\rx^o^6 "OTtioo^co — 'OUOTt-^ O-Tooo — — ^CMO'Ococoro * • sJM-'csrs.'rx'ococorocN S cS d o C t- 00 a c * ^o vo rv 00 cs o ^CNCO'^iOSSrN.OOO-O — CNCO^'OSSN.OOOO "— E o.: D-^oorxsco *- S) wo "O "O "O yo^ 0*00^0 H M « ^ to (0 147 FISHERY BULLETIN: VOL. 70, NO. I CO U3 0> s o *-l SQ u a >, S o •c > C o CI] u 0) >. u o 03 be C] o c o (u as -- 2 (fl « 3 3 -t-> CO -3 >. C 0) 3 cr O) »^ «t-i 0) bo rt +^ C (U u n < ■5-° 5 0,> ^1 0) (1) :3o >- a: — 0) u_ iZ. U O 0) 31 'c U o c T) O < — 2"^ O -a .•^ o z D O o o , On o o > E o i^ t I I o^pcocooocqcNoicN Tt o^ C^ I I I I I 1 I I I I I CO >q ■'T t I I I I I I I I I * * I I I I I t i*otr>p»oop ^pp'Oioiqp'Oioio * ♦ * + I lOtoO'OiOiO^P^'^ PPPP ' * * * III IIO*— CvlOOcOCMCS *Oltl|t — CM CNJ ^ — I ico oo-— "^csiococqo^p ^(Nrv^'-;co o-K loo^ChCM-^oocoCNsq co i pu") Ou^opv>iOp'<)'^ I ^ ^ ^ c^ — 0-— ON^cocN o OsN.(>p*orocM00'^, "^ cqGOco^N.cs(N^. ] I I ] ocNvooCNcocoOcdrx wScocN — 0000 ' ^Tj-— _<>ocorN.p '^"Ipch'^. I ' ' ' oo^f^cocvisD— cdoio — o ' ' ' ' — -- CM — ■" '^. CNCN^CN iO'-;CNfOpr^ I I I I lTfCN*00> CO— ;0"^'0of^'^ '.-^coc*^io cno-o^^OiocsO 'cor^f^o locoioCN — ocqr^p 'ooO'^ '^fCNo^ — coc^iooi — — r— cs •— — iCM— ;rN. ooco^o^>;C^^o'>'ocq i t i '-■coco ^6cNrxor^o"^^o ' ' ' — .— CN »— . — o "— cMco-^'O'^ihsCOO^o •-csco'^iONor^coCho — csco'^io^Ji^cocso E^.t2 00 O CN C t- X ffl o * _Q E o o ■ • • ■^ 010 "o ^c^c^c^ o M c9 ^ la o 148 MOSHER: SCALE FEATURES OF SOCKEYE SALMON Table 6. — Percentage frequency distribution' of circuli in the total freshwater zone of age 2. sockeye salmon col- lected in various years from 1956 to 1967. Number Asia Alaska British Columbia of circuli Ozernaya River- Blizhnee Lake^ Dal nee Lake^ Far North* Bristol Bay^ Attu Island* Adak Island" Un- alaska Island" Chignik" Cook Inlet** Karluk River" Yakutat= Ketch- ikan" Nass River"^ Nimp- kish River^i . - PtTCftlt - 9 — 0.5 — ~ — 0.4 — — — — — — — — — . 10 2.9 2.4 __ ._ 1.2 ._ __ __ _. .. 1 __ 10.3 __ 5.5 8.5 2.4 __ — •• 2 0.3 18.6 __ 9.3 16.1* __ __ 6.0 __ _^ 3 0.7 20. 1» __ 1.2 __ 12.9 13.3 0.3 ^_ 11.9 __ __ _^ 4 0.7 16.2 __ 4.5 0.2 14.0* 7.7 3.0 20.2* 0.5 __ __ 5 2.4 11.3 8.2 1.9 12.1 __ 0.9 7.3 6.9 19.0 4.0 — -• __ 6 5.6 6.9 ._ 12.7 4.4 8.8 3.7 9.7 10.2 0.5 13.1 8.5 1.1 -« 7 9.4 4.4 16.4* 6.8 7.9 0.2 6.9 12.5 11.8* 4.0 11.9 11.5 5.7 8 14.6 3.9 16.4* 10.0 7.4 0.4 7.9 12.1 11.5 10.0 8.3 15.5 10.2 1.0 9 16.3* 2.9 13.9 11.9 4.8 0.9 7.9 7.3 10.9 12.5 3.6 16.5* 13.6 3.1 20 14.6 1.5 — 9.4 13.1 3.3 2.5 10.2* 2.4 11.5 12.0 2.4 13.5 19.3* 5.2 1 13.2 0.5 7.0 14.8* 2.6 5.8 10.2* 0.4 11.8* 15.0 1.2 n.5 18.2 5.2 2 9.7 6.6 14.1 2.4 9.7 6.9 _. 10.5 18.5* ._ 9.5 10.2 3.1 3 5.9 3.3 11.3 2.6 13.1 6.9 0.4 5.9 14.5 5.5 8.0 2.1 4 4.2 0.4 7.4 1.4 15.5* 9.2 0.8 2.6 7.0 __ 1.5 8.0 2.1 5 2.1 __ 3.0 0.7 15.3 9.7 0.4 1.3 3.5 4.5 2.1 6 0.3 __ __ 0.8 0.9 13,3 7.9 __ 0.3 2.0 0.5 1.1 3.1 7 0.7 0.2 0.4 10.2 5.1 0.3 0.5 1.0 ^_ 5.2 8 5.3 __ 6.5 2.8 __ 0.7 __ __ 0.5 ^_ 4.2 9 10.5 __ __ 3.2 1.9 __ 0.3 __ 2.1 30 — 11.2 — — — 1.3 1.4 — — — — — — 3.1 1 12.5 0.7 0.5 __ __ .. 4.2 2 16.4* ._ __ 0.5 _. __ 7.3 3 16.4* __ __ 0.4 12.5* 4 __ 10.5 _. 0.4 __ 11.5 5 5.9 __ __ 0.2 __ 6.2 6 __ __ 5.3 __ __ __ __ 4.2 7 __ __ 3.9 __ __ _^ __ __ — 5.2 8 __ __ 1.3 __ _.. __ 5.2 9 — — — — — — -- — — — — — — — 2.1 Number of 72 51 38 61 216 105 139 54 62 76 50 21 50 22 24 fish 1 Actual 2 1959. 3 1958. * 1957. s 1966. « 1956. frequencies ; smoothed 1 occordi ng to Her iry (1961). ^ 1961. 8 1963. 0 1965. 10 ,964. >i 1967. * Indicates modes. DESCRIPTION OF SCALES FROM FISH OF VARIOUS AREAS The scale photographs are shown in three major series: ages 1., 2., and 3. Representative scales from each broad coastal area with rela- tively similar scales (including specific scale types from some areas of small production, when necessary) are shown on one plate. Two ex- ceptions to this grouping are made: (1) dis- tinctive scales from age 1. fish from North Amer- ican areas are shown on two plates and those for age 2. fish on one plate; and (2) scales from fish from North American areas north of Bristol Bay of ages 1., 2., and 3. are shown on the same plate. Scales of age 0. fish from all areas are shown on one plate, and those of age 4. fish from all areas are shown on another plate. A reference to the appropriate frequency tables of the number of circuli in the freshwater and first ocean zones (Tables 4 to 8) is made for each area group. These tables should be re- ferred to as scales from each group are dis- cussed. 149 FISHERY BULLETIN: VOL. 70, NO. I Table 7. — Percentage frequency distribution' of circuli in the first ocean zone of age 2. sockeye salmon collected in various years from 1956 to 1967. Number Asia Alaska British Columbia of Ozernaya Blizhnee Dal nee For Bristol Attu Adak Un- alaska Chignik'' Cook Karluk Yakutat'' Ketch- Noss Nimp- kish circuli River* Lake* Loke^' North' Bay=* Island" Island" Island" Inlet" River" ikan'" River- Riverii Pgrcint - II ._ — — 0.9 — — — — — — — 2 __ __ — -- 2.7 -- — — — — — — _. 3 __ __ __ 0.1 5.9 0.2 — — — — __ 4 __ 0.3 0.4 10.0 1.2 — __ __ _. __ 5 _. 1.2 0.8 12.6 3.4 __ __ __ _. _. 6 __ 3.4 1.2 13.3* 5.4 __ __ __ __ __ _. 7 2.0 2.0 8.0 2.6 13.2 7.0 .,_ __ __ 8 5.3 7.0 13.1 5.1 10.8 9.6 __ __ ._ ._ .. 9 0.3 7.9 9.8 17.9 7.3 8.4 13.7 __ 0.7 0.5 __ 20 1.7 1.0 13.8 12.7 20.9* 12.6 7.4 14.8* 0.4 3.0 1.5 — — — — 1 3.1 5.9 21.1* 16.4* 17.8 16.9* 5.8 11.2 4.0 6.9 4.5 0.5 2 6.6 11.8 21.1* 15.6 10.5 14.9 3.8 9.5 8.1 9.5 7.5 2.0 _. _, 3 16.0 13.2 13.8 13.5 4.5 13.2 2.3 8.9 8.9 10.9 8.5 2.4 3.5 4 24.0* 14.7 7.9 10.7 1.5 11.5 1.5 6.4 11.3 15.5 12.5 14.3 5.0 1.1 1.0 5 22.2 18.6* 5.3 6.6 0.5 6.7 0.7 3.9 14.5 18.4* 15.0 25.0* 10.0 3.4 3.1 6 14.2 16.7 2.0 3.7 0.2 2.7 0.2 2.0 16.1 13.8 14.5 21.4 15.0* 8.0 5.2 7 7.3 9.3 __ 1.6 0.1 1.4 0.2 1.2 16.9* 9.2 16.5* 16.7 14.5 20.5 6.2 8 3.5 4.9 __ 0.4 _. 1.3 0.1 1.1 13.3 6.9 13.5 13.1 13.0 29.5* 8.3 9 1.0 2.5 __ __ 0.9 0.4 5.6 3.3 5.0 6.0 12.5 21.6 16.7 30 1.0 — — — 0.3 — — 0.8 1.3 0.5 1.2 10.0 9.1 22.9* 1 ... 0.5 — — — — — 0.7 — — 6.5 2.3 16.7 2 3 4 5 ~ — ~ — — — — — — — ~ — 4.5 2.5 0.5 1 1 8.3 5.2 2.1 1.0 — — ~ ~ — — — — — — — — 1 . 1 2.3 1.1 6 7 ~~ ~~ ~~ ~ — — :: — :: — — *""■ ~~ — 2.1 1.0 Number of 72 51 38 61 216 191 203 140 62 76 50 21 50 22 24 fish ' Actual frequencies smoothec 1 accord i ing to Henry (1961). " 1961. = 1959. » 1963. ■■> 1958. « 1956. * 1957. i» 1964. s 1966. 11 1967. « 1956, ages 1 and 2 combined. * Indicates modes. KEY TO THE PLATES To compensate for the reduction of the original scale photographs to fit the printed page, use a 3 to 5x reading or magnifying glass to study them. The area enclosed by the first or central circu- lus is the focus or central platelet of the scale. The long black pointers near the focus indi- cate winter marks in the freshwater growth zone. A black stub pointer, if present, indicates the end of plus or transitional growth. If plus growth is i)resent, the circuli between the outer- most long black pointer and the black stub point- er are plus growth circuli. If no plus growth is present, the outermost black pointer indicates the end of freshwater growth as well as the last winter in fresh water. The white pointers bordered by black indicate the first winter mark in the ocean growth. The circuli between the end of the freshwater growth (or plus growth, if present) and this pointer are ocean-growth circuli and record the first year's growth in the ocean (the first ocean growth zone). The more widely spread circuli of this zone were deposited from May or June to Sep- tember or October (the summer growth) , where- as the more closely spaced circuli near the pointer were deposited during the autumn, winter, and early spring months (the winter growth). If a small white pointer'is present, it indicates an adventitious check in the first ocean growth 150 MOSHER: SCALE FEATURES OF SOCKEYE SALMON Table 8. — Percentage frequency distribution' of circuli (A) in the total freshwater zone and (B) in the first ocean zone of age 3. sockeye salmon collected in various years from 1955 to 1964. (A) (B) Number of circuli Asia North America Asia North America Asia" Blizhnea Loke^ Dalnee Lake^ Bristol Bay=^ Korluk River" Asia2 Blizhnea Lake^ Dalnee Lake* Bristol BayS Karluk River" Percent . . _ )2 _. 0.6 __ __ ._ __ .. 3 _^ 4.1 — — — — __ __ 4 __ 9.9 __ .„_ 0.4 5 0.5 15.1 __ 1.2 .. 6 2.9 19.2* __ .__ _, 3.3 _. 7 8.8 17.4 — ■_.. __ __ „_ 7.1 __ 8 15.9 11.6 __ 2.7 12.1 _. 9 18.1* 9,9 1.0 1.0 8.1 17.9 0.3 20 16.9 8.1 "- 2.0 — 2.9 — lO.I 19.6* 1.9 1 14.5 3.5 4.0 1.0 5.9 11.5 15.4 4.8 2 7.8 0.6 __ 10.0 3.2 10.8 1.2 16.9 10.8 7.6 3 2.2 — . — 12.0 6.6 16.4 5.8 18.9* 6.7 13.0 4 1.2 .— __ 13.0 9.5 20.1* 12.2 15.5 2.9 20.2 5 1.5 „ 19.0* 11.4 19.1 15.1 10.8 1.7 20.9* 6 1.2 _. __ 17.0 13.3 13.5 16.3 4.7 0.8 14.8 7 1.7 __ 10.0 12.7 6.4 18.0* 0.7 -- 8.6 8 2.5 __ 0.7 6.0 12.1 2.2 16.3 __ __ 3.8 9 2.0 __ 4.1 3.0 14.2* 1.2 10.5 __ 1.9 30 1.2 8.1 2.0 10.2 0.5 4.1 — 1.6 1 0.7 8.1 I.O 4.1 0.6 _. 0.6 2 0.2 __ 6.1 .. 1.2 _^ __ __ __ __ 3 __ __ 8.1 0.3 _^ _ — __ __ 4 5 ~ — 11. 5» 8.1 ~ ~ — — ~ ~ ~ 6 7 ~ — 6.8 11. 5* ~ ~ ~ — — — — 8 9 40 1 2 3 4 5 6 7 8 ^ — — 11. 5» 8.1 4.1 0.7 ~ ~ — — ~ ~ — — — 0.7 1.4 0.7 — ~ ~ — ~ Number of fish 102 43 37 25 79 102 43 37 25 79 ^ Actual frequencies smoothed according to 2 1962, from an area off southeast Kamch 3 1958. ' 1964. « 1955 and 1957 combined. » 1961. • Indicates modes. Henry (1961), latka. zone. Adventitious checks in other zones are not noted. AGE 0., ALL AREAS (Plate I) As noted previously, age 0. sockeye salmon are not common anywhere since fish of this spe- cies normally live for one or more years in a lake before migrating to the sea. Consequently, it was not possible to assemble frequency dis- tributions of the number of circuli in the first ocean zone for age 0. fish. It appears, however, that usually there are a few more circuli in the first ocean zone of scales of age 0. than on scales from fish of the same geographical area that have lived one or more years in fresh water. A few individuals of this age have been found at some time in almost every locality. Gilbert 151 FISHERY BULLETIN: VOL. 70, NO. 1 KAMCHATKA R. COPPER R. NUSHAGAK R. PETERSBURG NASS R SMITH INLET Plate 1. — Age 0., all areas. 152 MOSHER: SCALE FEATURES OF SOCKEYE SALMON KUSKOKWIM R. :^'^*fe&, AGE ^^^ Plate 2. — Ages 1., 2., and 3., Alaskan areas north of Bristol Bay. 153 FISHERY BULLETIN: VOL. 70, NO. 1 KAMCHATKA R. OZERNAYA R. BOLSHAYA R. OKHOTSK SEA LAKE DALNEE Plate 3. — Age 1., Asia. 154 MOSHER: SCALE FEATURES OF SOCKEYE SALMON UGASHIK R. NAKNEK R. EGEGIK R. KVICHAK R. NUSHAGAK R. TOGIAK BAY Plate 4. — Age 1., Bristol Bay. 155 FISHERY BULLETIN: VOL. 70, NO. 1 ><" -T^r^-.-rv- ATTU ISLAND :~ --«.:*5'S CHIGNIK R. UNALASKA RED R. KARLUK R. COOK INLET Plate 5. — Age 1., Aleutian Islands to Cook Inlet. 156 MOSHER: SCALE FEATURES OF SOCKEYE SALMON COPPER R. YAKUTAT A PETERSBURG KETCHIKAN Plate 6. — Age 1., Copper River to southeastern Alaska. 157 FISHERY BULLETIN: VOL. 70, NO. 1 MASS R. SKEENA R. FRASER R. COLUMBIA R Plate 7. — Age 1., British Columbia and the Columbia River. 158 MOSHER: SCALE FEATURES OF SOCKEYE SALMON FISH CREEK UNALASKA '' NIMPKISH R B Plate 8. — Age 1., North American areas with distinctive scales, Fish Creek type. 159 FISHERY BULLETIN: VOL. 70. NO. 1 A RIVERS INLET A NIMPKISH R. B Plate 9. — Age 1., North American areas with distinctive scales, Rivers Inlet type. 160 MOSHER: SCALE FEATURES OF SOCKEYE SALMON KAMCHATKA R. BOLSHAYA R. LAKE BLIZHNEE LAKE DALNEE' Plate 10. — Age 2., Asia. 161 FISHERY BULLETIN: VOL. 70, NO. 1 :,^w. UGASHIK R. NUSHAGAK R BEAR R Plate 11. — Age 2,, Bristol Bay. 162 MOSHER: SCALE FEATURES OF SOCKEYE SALMON ADAK ISLAND V.Udl KARLUK R. COOK INLET Plate 12. — Age '2., Aleutian Islands to Cook Inlet. 163 FISHERY BULLETIN: VOL. 70, NO. 1 COPPER R. NIMPKISH R. Plate 13. — Age 2., Copper River to the Columbia River. 164 MOSHER: SCALE FEATURES OF SOCKEYE SALMON FRAZER LAKE FISH CREEK COLUMBIA R. RIVERS INLET Plate 14. — Age 2., North American areas with some distinctive scales. 165 FISHERY BULLETIN: VOL. 70, NO. 1 OZERNAYA R. BOLSHAYA R OKHOTSK LAKE BLIZHNEE LAKE DALNEE Plate lo. — Age /^^V;it; Jf' *Vt ■^ Figure 3. — Developmental stages of Sphyraena borealis. Specimens A, B, and C were laboratory reared; specimen D was collected in a plankton net. A. 9.4 mm SL; B. 12.3 mm SL; C. 14.5 mm SL; D. 21.0 mm SL. in a 10.9-mm SL larva. All specimens had 24 vertebrae. Six branchiostegals were present on the 7.4- mm SL larva and a seventh had developed on the 9.1-mm SL specimen. Cleithra were present but poorly ossified at 7.4 SL; they were well de- veloped at 9.1 mm SL. Ossification of the caudal region began at 7.4-mm SL. At this size, the hypural bones were developing and some rays could be distinguished in the ventral half of the caudal finfold. All caudal elements were beginning to develop by 10.9 mm SL and were easily recognized on a 12.1-mm SL specimen. The last three vertebrae (including the urostyle) contributed to support of the caudal fin; both neural and haemal spines of the antepenultimate vertebra (22nd) sup- ported developing accessory rays of the caudal fin, as did the haemal spine of the penultimate (23rd) vertebra. A total of 6 hypural bones 190 HOUDE: NORTHERN SENNET Table 1. — Summary of morphometric data from laboratory-reared (L) larvae and from wild-caught (W) larvae and juveniles of Sphyraena horealis. (Specimens between broken lines are undergoing notochord flexion.) Except for total and standard lengths, measurements are proportional values, the ratio of the character relative to standard length. Specimen No. Total length Standard length Preonal length Isf predorsal 2nd predorsal length length Head length Snout length Tip lower jaw Eye diameter mm mm 3 (U 2.6 2.6 0.68 4 (L) 4.0 3.8 .66 5 (L) 4.2 4.1 .63 7 (L) 4.3 4.2 .56 6 (L) 4.5 4.3 .66 9 (L) 5.2 5.0 .65 0 (L) 5.5 5.3 .67 8 (L) 5.8 5.6 .68 13 (L) M (L) 17 (L) 15 (L) 18 (L) 16 (L) 19 (L) 24 (W) 23 (W) 20 (L) 28 (W) 26 (W) 30 (W) 29 (W) 21 (W) 25 (W) 22 (W) 32 (W) 31 (W) 10.2 12.2 12.9 13.4 13.7 13.7 14.7 15.9 16.5 16.7 19.4 19.1 19.7 21.2 24.2 24.6 33.9 66.8 70.0 9.4 10.9 11.3 11.9 12.1 12.3 12.7 13.7 14.4 14.5 16.9 17.0 17.1 18.4 21.0 21.5 30.6 59.6 62.9 0.13 --. __ .14 0.03 __ 0.08 .19 .04 __ .08 .19 .04 __ .08 .21 .05 __ .08 .26 .07 <0.01 .08 .26 .07 <.01 .08 .27 .08 <.01 .07 11 (L) 7.7 7.4 .66 .. ._ .28 .10 <.01 .07 27 (W) 7.9 7.4 .71 .30 .12 .02 .08 12 (L) 9.5 9.0 .69 — 0.67 .31 .12 .01 .08 .70 .70 .71 .69 .69 .66 .68 .72 .72 .63 .70 .69 .70 .7C .71 .70 .70 .71 .72 0.50 .44 .44 .47 .49 .48 .47 .47 .48 .46 .47 .49 .46 .46 .46 .45 .68 .68 .70 .68 .68 .66 .70 .71 .70 .71 .70 .70 .70 .70 .72 .69 .70 .70 .71 .32 .33 .36 .31 .33 .32 .34 .32 .36 .35 .33 .32 .32 .33 .34 .32 .32 .32 .32 .13 .14 .12 .13 .12 .13 .13 .12 .15 .14 .14 .14 .13 .14 .15 .14 .14 .14 .14 <.01 .02 .02 .01 .03 .02 .05 .04 .03 .04 .04 .04 .05 .06 .03 .06 .03 .03 .09 .08 .08 .08 .08 .08 .08 .07 .08 .08 .07 .06 .06 .07 .08 .07 .07 .06 .06 16.0- 14.0 s S 12.0- X O 10.0 z UJ "* ».0 ^ '-OH z < I- 4 0- 0 "1 I I I I I I 6 a 10 1? 14 16 It DAYS AFTER HATCHING 20 — 1 — 22 Figure 4. — Growth of Sphyraena borealis larvae reared in the laboratory at an average temperature of 24.0° C. were formed near the posteroventral surface of the urostylar vertebra. There were 3 epurals and 2 pairs of uroneurals ossifying dorsal to the urostyle. The 17 principal caudal rays were sup- ported by the hypural bones. Principal caudal ray support was as follows: hypural 1 support- ed rays 1 to 3; hypural 2 supported rays 4 to 7; hypural 3 supported ray 8; hypural 4 supported rays 9 to 11 ; hypural 5 supported rays 12 to 15; hypural 6 supported rays 16 and 17. A 17.0- mm SL specimen differed from smaller speci- mens only in having the first principal caudal ray partially supported by the haemal spine of the penultimate vertebra as well as the first hypural bone. All elements were at least partially ossi- fied at 17.0 mm SL; both the epurals and uro- neurals were the most poorly ossified bones of the caudal skeleton at this stage. Hollister (1937) examined caudal skeletons of adult and 191 FISHERY BULLETIN: VOL. 70, NO. 1 Table 2. — Summary of meristic data (spines and rays) from laboratory-reared (L) and wild-caught (W) larvae of Sphyraena borealis. (Dashes indicate elements were present but could not be accurately counted.) Specimen Standard length Principal Caudal rays Anal fin 2nd dorsal fin Isf dorsal fin Pel vie fin Pectoral fin No. Left Right Left Right mm 3 (L) 2.6 0 0 0 0 0 0 0 0 4 CD 3.8 0 0 0 0 0 0 0 0 5 (L) 4.1 0 0 0 0 0 0 0 0 7 (L) 4.2 0 0 0 0 0 0 0 0 6 (U 4.3 0 0 0 0 0 0 0 0 9 (L) 5.0 0 0 0 0 0 0 0 0 10 (L) 5.3 0 0 0 0 0 0 0 0 8 (L) 5.6 0 0 0 0 0 0 0 0 27 (W) 7.4 0 0 0 0 0 0 0 0 11 (L) 7.4 7 0 0 0 0 0 0 0 12 (L) 9.0 16 8 8 0 0 0 0 0 13 (L) 9.4 16 9 9 0 0 0 0 0 14 (L) 10.9 17 10 10 3 0 0 0 0 17 (L) 11.3 17 11 10 5 5 __ 8 6 15 (L) 11.9 17 11 10 3 _^ 18 (L) 12.1 17 11 10 5 __ __ 8 8 16 (L) 12.3 17 11 10 4 5 __ 8 6 19 (L) 12.7 17 10 10 5 — 10 24 (W) 13.7 16 11 10 N 5 12 12 23 (W) 14.4 17 __ 10 5 5 6 12 20 (L) 14.5 17 11 10 5 6 12 10 28 (W) 16.9 17 11 10 ._ _^ _^ ' _« 26 (W) 17.0 17 11 10 5 6 6 11 12 30 (W) 17.1 17 10 10 5 _^ , 29 (W) 18.4 17 11 10 21 (W) 21.0 17 11 10 __ . 25 (W) 21.5 16 11 10 5 6 6 12 12 22 (W) 30.6 17 11 10 5 6 6 12 12 juvenile S. borealis, S. piaidilla, and S. barra- cuda. She found that all were similar and showed that a progressive fusion of caudal ele- ments occurred in barracudas as they grew. Fusion of hypurals 2 and 3, 4 and 5, and of the urostyle with hypural bones was apparent in her large specimens. Fin Development Newly hatched larvae had a prominent larval finfold (Figures 1 and 2) that appeared granular because of many bubbles or small inclusions dis- tributed throughout. These inclusions were not illustrated, except in the newly hatched larva (Figure IB) , but were present until larvae grew to about 9.5 mm SL. Fin ray development essentially was completed at 13.5 mm SL (Table 2). Fan-shaped pectoral fins without rays developed at 3.8 mm SL 1 day after hatching (Figure 2A). Rayed fins devel- oped in the following sequence: caudal, anal and second dorsal, first dorsal, pelvics, and pectorals. The caudal fin rays began to develop at 7.4 mm SL when the notochord started to flex (Figure 2D). All 17 principal caudal rays were present on a 10.9-mm SL larva. Accessory caudal rays (ray- lets) also were developing at 10.9 mm SL. Their number varied from 6 to 8 dorsally and 6 to 9 ventrally on specimens up to 30.6 mm SL. The anal and second dorsal fins were represented only by opaque areas in the finfold at 7.4 mm SL, but rays of these fins were developing at 9.0 mm SL; posteriormost rays developed before the more anterior rays and 9 rays were present in each fin at 9.4 mm SL (Figure 3A). A full comple- ment of 2nd dorsal (I, 9) and anal (I, 9 or 10) elements was present at 10.9 mm SL. Spines of the first dorsal fin appeared at 10.9 mm SL and a full complement of 5, located over verte- brae 5 to 7, was present on a 12.1-mm SL larva. The 5 spines were more heavily ossified and lo- cated over vertebrae 6 to 8 on a 17.0-mm SL specimen. Pelvic fin buds formed on larvae as small as 9.4 mm SL, but rays did not begin to develop until 11.3 mm SL. Pectoral fin rays also began to develop at this length. All pelvic (I, 5) and pectoral (12) elements were not present on 192 HOUDE: NORTHERN SENNET all larvae until about 13.5 mm SL. Remnants of the larval finfold persisted until 12.5 mm SL (Figure 3B). Pigmentation Except where specifically mentioned, referen- ces to pigment are to melanophores. Xantho- phores were common on larvae, and both silver iridophores and blue chromatophores were pre- sent. (See Fujii, 1969, for chromatophore ter- minology.) Some variations in melanophore patterns were present among S. borealis larvae of similar size but the following description gives the typical sequence of development. Melanophores and xanthophores were present on newly hatched larvae but the latter faded after preservation. Small melanophores were distributed in a dorso-lateral and ventro-lateral row on each side of the larva (Figure IB) . The two rows converged just above the yolk sac and ran anterior as a single row to the posterior cephalic region, joining a series of melanophores located over the hindbrain. Other melanophores were present on the anterior half of the oil glob- ule and near the posterior of the yolk sac. One to two days after hatching (about 3.8 mm SL) , melanophores became larger and more nu- merous (Figure 2A) . Those located in the later- al rows and on the cephalic region became stel- late. They were more numerous on the yolk sac and two small melanophores appeared near the developing mouth. A series of two to six small, contracted melanophores were noted near the tip of the notochord. The eye became pig- mented at 2 days after hatching. As larvae developed, both melanophores and xanthophores became more numerous. Xantho- phores were distributed over much of the body and consisted of elongate yellow cells forming a loose network on the body. In life, larvae ap- peared green because of the presence of both yellow and black pigments. Blue iridophores on the hindgut and some iridescent pigment in the eyes also were present. By 7 days after hatching (about 5.3 mm SL, Figure 2C) stellate melanophores had appeared over the brain, on the tip of the upper jaw, lower jaw, angle of the jaw, ventral margin of the opercular region and over the foregut. Each dorso-lateral row of stellate melanophores was well developed while the ventro-lateral rows were condensing into a single ventro-medial row, posterior to the anus. A mid-lateral series of melanophores was developing posterior to the anus. At 9 days after hatching (about 7.4 mm SL) pigmentation became more intense (Figure 2D) . About 5 melanophores were present in the de- veloping caudal fin. The fleshy tip of the lower jaw began to become darkly pigmented on some specimens at this stage. The extent and inten- sity of pigmentation continued to increase on older larvae (Figures 3A and 3B). No changes in pattern were observed, except for develop- ment of a line of pigment that bisected the eye, and a migration of melanophores from the tip of the developing urostyle to the ventral margin of the hypural plate. When specimens were 20 to 22 days old (about 12.5 to 14.5 mm SL), the juvenile pattern of dorsal and lateral blocks of pigment began to ap- pear (Figure 3C). Stellate melanophores also developed in the second dorsal and anal fins of individuals of this size. Longer specimens from plankton collections had the typical juvenile pig- ment pattern (Figure 3C) (cf. de Sylva, 1963; Figure 4). GROWTH AND MORTALITY Larvae of S. borealis grew most rapidly dur- ing the first 14 days after hatching, but more slowly during the next 7 days (Figure 4). Larvae were 2.6 mm SL at hatching, averaged 5.5 mm SL at 7 days after hatching, 11 mm at 14 days, and about 13.5 mm at 21 days. The decrease in growth rate of the sennets during the third week may have been caused by the scarcity of fish larvae in their diet. Average growth rate during the rearing experiment was about 0.5 mm per day. A total of 78 sennet eggs were incubated but only about 50 ';r hatched. From this small num- ber of hatched larvae, 9 survived until 14 days after hatching, when they had developed most of the characters of juvenile sennets. Mortality in the first 14 days included 11 larvae preserved 193 FISHERY BULLETIN: VOL. 70, NO. 1 for describing larval development. No larvae survived beyond 22 days after hatching, prob- ably because of the inadequate diet. TRANSFORMATION Transformation from the larval to the early juvenile stage occurred at about 13.5 mm SL for S. boreal'is. At that size, fin ray development was complete and there were no remnants of the larval finfold. Additional evidence that this size marked the end of the larval period was the occurrence of keeled scales on specimens at 14.5 mm SL and the appearance of juvenile pig- ment patterns between 12.5 and 14.5 mm SL. COMPARISONS Among the four early life history series of barracudas that have been described in the lit- erature, newly hatched larvae of S. borealis most closely resemble those of S. p'mguis from Jap- anese waters (Shojima et al., 1957; Uchida et al, 1958) , except for size at hatching. In both spe- cies, the oil globule, at hatching, is located at the anterior end of the yolk mass; neither develop melanophores in the finfold. In contrast, the oil globule is located at the posterior end of the yolk mass in newly hatched larvae of S. sphyraena (Vialli, 1956) and S. argentea (Orton, 1955) and both have melanophores in the finfold. Eggs of S. pinguis (0.69 to 0.82 mm diameter) were considerably smaller than those of S. sphy- raena (1.11 to 1.15 mm), S. borealis (1.22 to 1.24 mm), and S. argentea (about 1.5 mm). It is not surprising that at hatching, larvae of S. pinguis were smaller than those of S. borealis, measuring 1.75 mm as compared to 2.6 mm. Both species were similarly pigmented at hatch- ing, but differed significantly by 45 and 66 hr. Melanophores were larger and more concen- trated in S. pinguis larvae than in S. borealis at the same stage of development. Older larvae of S. borealis and S. sphyraena had similar pigment patterns, but those of S. sphyraena had a more developed lower jaw tip. Three-day-old larvae of S. argentea from Cali- fornia waters had a distinctive band of melano- phores on the tail portion of the body just pos- terior to the anus that was lacking in S. borealis. Postlarvae of S. barracuda were described and illustrated by de Sylva (1963) . They diflfer sub- stantially from S. borealis in being deeper bodied and having a relatively longer snout. Pigmenta- tion of S. barracuda and S. borealis larvae be- tween 5.5 and 11.9 mm SL also differs somewhat in its detail. BEHAVIOR Newly hatched larvae of S. borealis drifted about the rearing tank making only occasional feeble swimming attempts when disturbed. At 2 days after hatching, larvae maintained a horizontal position and began swimming actively with short darting motions. Feeding activity was first observed 3 days after hatching. Sennet larvae usually fed using the S-flex behavior pre- viously described for many species of clupeiform larvae (e.g., Rosenthal, 1969; Schumann, 1965). Occasionally, however, a sennet larva would strike at a food organism without first flexing its body and examining the item. Zooplankton organisms less than 100 yu, in body width were the initial food of larvae. No stomach analyses were carried out on sennet larvae, but most of the food which was placed in the tank were cope- pod nauplii and copepodites. Sennet larvae continued to feed on small zoo- plankton organisms until 10 days ^fter hatching, even though larger plankton, including fish larvae, and nauplii of brine shrimp were present in the tank beginning 7 days after larvae hatched. At 10 days, large plankton (about 300 to 400 [x body width) was accepted as food, as were some unidentified, newly hatched fish larvae about 2 mm in length. A sennet larva would approach a tiny fish larva, assume an '^ y-d FiGURR 7. — .S. purpiiratim with e.xtendod podia ju.st prior to initial contact witli tho arm of a leather star. Figure 8. — The defensive behavior of S. purpuratus im- mediately after being touched by D. imbricata. 212 ROSENTHAL and CHESS: PREDATOR-PREY RELATIONSHIP side Mission Bay, San Dieg-o County (lat 32° 45'30" N; long 117°14'30" W), and the other was from the study site off Pt. Loma. Asteroid tube feet were selected as the biotic stimuli be- cause of the known effectiveness of this tissue in eliciting avoidance reactions in other inverte- brate species (Bullock, 1953). Coarse, washed sand grains were used as the abiotic control. Purple urchins were placed individually into glass bowls which contained seawater, and then an asteroid tube foot, or sand grain, was dropped onto the urchin's test. Each urchin was tested for a 2-min period using a different stimulus on each run. Fresh seawater was placed into the bowls prior to each test. The tube feet of D. imbricata elicited gaping and erection of pur- ple urchin globiferous pedicellariae in 28 out of 30 test animals. These 28 individuals had an average reaction time of 8 sec, with a range be- tween 2 and 20 sec. None of the S. purpuratus displayed gaped globiferous pedicellariae when either Pisaster giganteus tube feet or sand grains were presented to them. Additional asteroids that were found in the sublittoral zone off Pt. Loma were also tested. The tube feet from Astrometis sertulifera, Pa- tiria miniata, and Pycnopodiu helianthoides evoked the pedicellariae response in S. purpiir- atus; however, no response was elicited when eight other asteroid species were tested (Table 1). It is interesting to note that the former three species were the only other sea stars, be- sides D. imbricata, that have been observed feed- ing on live S. purpuratus off Pt. Loma. Table 1. — A list of sea stars found off Pt. Loma, Cal- ifornia. All 12 species were individually used to test the globiferous pedicellariae response in S. purpuratus. Species Reaction Astrometis sertulifera Astropecten armafus Dermasterias imbricata Henricia leviuscuta lAnckia columbiae Mediaster aequalis Patiria miniata Pisaster brevispinus Pisaster giganteus Pisaster cchraceus Orlhasterias koehleri Pycnopodia helianthoides + + + + pedicellariae response no pecJicellariae response The erection and gaping of the globiferous pedicellariae initially occurred only on the area of the urchins' test which was directly stimu- lated by the sea star's tube foot. Jensen (1966) found that a single tube foot from the sea star, Marthasterias glacialis, activated the globiferous pedicellariae only in a restricted area on the test of the sea urchin, Psanimechinus miliaris, whereas the sea star arm tip caused a res]3onse from all globiferous pedicellariae. In contrast, we found that the arm tip of D. imbricata acti- vated the pedicellariae of S. purpuratus only in the region of the stimulus. The defensive response and recognition of predatory stimuli was so acute in S. purpuratus that an arm of a P. giganteus placed on one side of a purple urchin's test, and a D. imbricata arm positioned on the opposite side elicited a response from the globiferous pedicellariae only in the area of leather star contact. Defensive use of globiferous pedicellariae by sea urchins when disturbed by predatory aster- oids has been described by Prouho (1890), Jen- nings (1907), Jensen (1966), Mauzey et al. (1968), and Rosenthal and Chess (1970). Jen- sen (1966) reported that the poison contained in globiferous pedicellariae of Psammechinus miliaris was not strong enough to paralyze a M. glacialis; however, it did have an irritating ef- fect on the sea star which caused it to retreat from the urchin. We found that in some lab- oratory situations, globiferous pedicellariae bites on the arms of D. imbricata caused localized withdrawal of gills or papulae, and a shortening or retraction of the affected arm. Despite this irritation, 90 Sr of all the leather stars we found feeding on purple urchins had from one to over 300 sea urchin pedicellariae attached to their epidermis. Certain groups of aquatic organisms have been observed to respond to chemical signals or alarm substances emitted by injured conspecifics (von Frisch, 1941; Pfeiffer, 1963; Snyder and Sny- der, 1970). Recently, Snyder and Snyder (1970) found that the tropical sea urchin, Diadema aji- tillarum, exhibited an alarm or escape response when stimulated with the juices of injured mem- bers of its own species. Laboratory and field 213 FISHERY BULLETIN: VOL. 70, NO. I tests were conducted to determine if a similar alarm response existed in S. jmrpuratiis. A purple urchin within a group of urchins was crushed underwater, and the reactions of neigh- boring conspecifics noted for 5 min. In both situations we observed no change in movement or alteration in behavior which could be con- sidered alarm oriented following injury to a con- specific. In place of an alarm response we occa- sionally noted an entirely different reaction from S. purpuratus in the laboratory. If a leather star was disturbed while feeding on a purple urchin or moved away from an urchin test following predation, occasionally other S. purpiirahis in the aquaria approached the conspecific and scav- enged the remains. DISCUSSION The behavioral responses exhibited by S. pur- puratus when it is disturbed by D. imbricata suggest a well-developed predator-prey rela- tionship. In most instances purple urchins erected globiferous pedicellariae when touched by the four sea stars {D. hnbricata, P. helian- thoides, A. sertulifera, and P. miniata) which are known to prey upon them. In contrast, no evasive or defensive responses were noted in the same purple urchins when they were touched by eight additional asteroid species. It appears as though S. 2)urpuratus either responds to sea stars that are biochemically similar, or through selection the urchin has acquired the ability to recognize particular asteroid species as potential predators. The predator-prey association which exists be- tween these two species off Pt. Loma, California may be a regional phenomenon, since the rela- tionshi]) has not been reported from other local- ities along the Pacific Coast. However, from the responses in both laboratory and field situations we believe that the occurrence is probably more widespread than indicated in the literature. The large number (41 '/r ) of .S. purpuratus we found included in the overall diets of leather stars off Pt. Loma, as opposed to the total exclusion of this species in the diets of leather stars off Wash- ington state as reported by Mauzey et al. (1968) is extremely puzzling to us. We can only spec- ulate at this time on what could account for this variation in feeding behavior. Selection of po- tential prey by D. imbricata may be determined by the following conditions: (1) Prey density and availability, (2) search time or the time re- quired by the sea star to find and capture suit- able prey, (3) taste or gustatory preferences of the sea star, and (4) some form of associative learning by D. imbricata. Strongyloceyitrotus purpuratus appeared to be available to D. imbricata on almost a continuous basis within the study area, since the population of purple urchins was estimated to have a mean density value of 30.2/m2. Predator search and capture time also seems to be related to the den- sity and distribution patterns of the prey, as well as to the avoidance tactics employed by these potential prey. Encounters between the two species on uniform substratum usually resulted in the escape of S. purpuratus; however, when the purple urchins occupied depressions, holes, or crevices along the sea floor, they became more vulnerable to asteroid predation. In response to asteroid predation S. purpuratus has appar- ently evolved countermeasures such as evasive movement, and defensive utilization of spines and poisonous globiferous pedicellariae. The large number (90'^r) of feeding leather stars with purple urchin pedicellariae attached to their epidermis might lead one to suspect that these appendages are ineffective as a defensive mech- anism. The pedicellariae, however, appear to act as an irritant that in certain situations halts the pursuit of a leather star and thus allows the urchin to escape. Marler and Hamilton (1966, p. 142) stated that "there is evidently a subtle and dynamic balance between these different evasive characteristics of the prey species on the one hand and the abilities of the predators to overcome them on the other". The taste or gustatory preferences of individual D. imbricata as compared to a leather star population has not been explored. From our observations off Pt. Loma we would expect that at least a few purple urchins would show up in the diets of D. imbricata off Wash- ington, even if other species were "preferred" above 5. purpuratus. Possibly before leather stars prey on live urchins there is a learning 214 ROSENTHAL and CHESS: PREDATOR-PREY RELATIONSHIP process involved before the sea star recognizes or associates specific stimuli with food. Tin- bergen (1960) proposed that learning was in- volved in the feeding behavior of insect-eating birds, and that initial non-acceptance of specific insects by these birds was due to an unfamiliarity with these forms as prey. He further related initiation of feeding on a new food item with chance experience and prey density. Tinbergen (1960) suggested that the predator acquires a "specific search image" for the prey after being sufficiently impressed with it from frequent chance encounters. Holling (1958 and 1965) studied predation on the cocooned pupae of saw- flies by shrews and mice, and suggested that as- sociative learning was an important component in the feeding behavior of these small mammals. Unfortunately, associative learning has been studied in only a relatively small number of low- er animal (invertebrate) groups. Evans (1968) discussed this form of learning in cephalopods, insects, annelids, and flatworms. There is some evidence to suggest that associative learning exists in echinoderms. Landenberger (1966) found that the sea star P. giganteus learned to associate a light stimulus with food. The asso- ciation apparently disappeared when the re- sponse to the light stimulus was no longer re- warded with food. If associative learning, with food as a reinforcement or reward, is a compo- nent in the feeding behavior of D. rmbricata, then it might account for the presence of purple urchins in the diets of leather stars oflF Pt. Loma. This area contained a large number of highly accessible S. purpuratus, and yet at the same time appeared to be practically devoid of many of the sessile or sedentary invertebrates that these sea stars are reported to feed on. Derma- sterias imbricata probably responds to a small class of chemical and/or tactile stimuli; how- ever, only through associative learning and ex- perience can it exploit an evasive prey species such as S. purpuratus. The leather star may not acquire the experience necessary to capture live iS. purpuratus in other subtidal areas that contain alternate prey in greater abundance, since these forms are more accessible and pos- sibly can account for the total nutrient require- ments of D. imbricata. ACKNOWLEDGMENTS We especially wish to thank W. D. Clarke, P. K. Dayton, T. A. Ebert, H. M. Feder, and H. R. Melchior for stimulating discussions and critical evaluation of this manuscript. We also wish to thank Virginia Moore who prepared Figures 1, 5, and 6. Westinghouse Ocean Research Labora- tory assisted in financial support of this study and the National Marine Fisheries Service gen- erously provided laboratory facilities at the Southwest Fisheries Center, La Jolla, California. LITERATURE CITED Bullock, T. H. 1953. Predator recognition and escape responses of some intertidal gastropods in presence of star- fish. Behav. 5: 130-140. Evans, S. M. 1968. Studies in invertebrate behavior. Heinemann Educational Books Ltd. London. 110 p. Feder, H. M. 1956. Natural history studies on the starfish, Pi- snstei- ochracens (Brandt, 1835) in the Monterey Bay area. Doctoral dissertation, Stanford Uni- versity. 294 p. 1959. The food of the starfish, Pisaster ochraceus, along the California Coast. Ecol. 40: 721-724. 1970. Growth and predation by the ochre sea star, Pif;aster orhracevs (Brandt), in Monterey Bay, California. Ophelia. 8: 161-185. Feder, H. M., and A. M. Christensen. 1966. Aspects of asteroid biology. In R. A. Boo- lootian (editor), Physiology of Echinodermata. p. 87-127. Interscience Publishers, New York. Fisher, W. K. 1928. Asteroidea of the North Pacific and adjacent waters. Part II, Bull. U.S. Nat. Mus., 76: 245 p. 1930. Asteroidea of the North Pacific and adjacent waters. Part III, Bull. U.S. Nat. Mus., 76 : 356 p. Frisch, K. von. 1941. Uber einen Schreckstoff der Fischhaut und seine biologische Bedeutung. Z. Vergl. Physiol. 29: 46-145. Holling, C. S. 1958. Sensory stimuli involved in the location and selection of sawfly cocoons by small mammals. Can. J. Zool. 36: 633-635. 1965. The functional response of predators to prey density and its role in mimicry and population regulation. Mem. Entomol. Soc. Can. 45:1-60. Jennings, H. S. 1907. Behavior of the starfish Asterias forreri De Loriol. Univ. Calif. Publ. Zool. 4(2):53-185. 215 FISHERY BULLETIN: VOL. 70, NO. 1 Jensen, M. 1966. The response of two sea urchins to the sea star Marthasterias glncialis (L.) and other stim- uli. Ophelia. 3:209-219. Kjerskog-Agersborg, H. p. 1918. Bilaterial tendencies and habits in the twenty-rayed starfish Pycnopodia helianthoides (Stimpson). Biol. Bull. Mar. Biol. Lab., Woods Hole. 35: 232-254. Landenberger, D. E. 1966. Learning in the pacific starfish Pisaster gi- ganteus. Anim. Behav. 14:414-418. 1968. Studies on selective feeding in the Pacific starfish Pisaster in Southern California. Ecol. 49:1062-1075. Leighton, D. L., L. G. Jones, and W. J. North. 1966. Ecological relationships between giant kelp and sea urchins in Southern California. Proc. 5th Intl. Seaweed Symp., p. 141-158. Pergamon Press, Oxford. Leighton, D. L. 1971. Grazing activities of benthic invertebrates in Southern California kelp beds. In W. J. North (editor), The Biology of Giant Kelp Beds (Ma- crocystis) in California, p. 421-453. Verlag von J. Cramer, Germany. Marler, p., and W. J. Ha:\iilton, IIL 1966. Mechanisms of animal behavior. J. Wiley and Sons, New York. 771 p. Mauzey, K. p., C. Birkeland, and P. K. Dayton. 1968. Feeding behavior of asteroids and escape re- sponses of their prey in the Puget Sound region. Ecol. 49:603-619. North, W. J., and J. S. Pearse. 1970. Sea urchin population explosion in Southern California coastal waters. Science 167(3915) :209. Paine, Robert T. 1969. The Pisaster-Tcgida interaction: prey patches, predator food preference and intertidal community structure. Ecol. 50:950-961. Pfeiffer, W. 1963. The fright reaction in North American fish. Can. J. ZooL, 41:69-77. Prouho, H. 1890. Du role des pedicellaires gemmiformes des oursins. C. r. hebd. Seanc. Acad. Sci., Paris III: 62-64. RiCKETTS, Edward F., and Jack Calvin. 1962. Between Pacific Tides. Stanford Univ. Press, Stanford, 516 p. Rosenthal, R. J., and J. R. Chess. 1970. Predation on the purple urchin by the leather star. Calif. Fish and Game, 56:203-204. Rosenthal, Richard J. 1971. Trophic interaction between the sea star Pisaster giganteus and the gastropod Kelletia kelletii. Fish. Bull. U.S. 69:669-679. Snyder, N., and H. Snyder. 1970. Alarm response of Diadema antillarum. Scence. 168(3928) :276-278. Tinbergen, L. 1960. The natural control of insects in pine woods. Arch. Neerl. Zool. 13:265-343. Wood, L. 1968. Physiological and ecological aspects of prey selection by the marine gastropod Urosalpinx ci- nerea (Prosobranchia: Muricidae). Malacol. 6: 267-320. 216 COMPARISON OF FOREGUT CONTENTS OF Sergestes similis OBTAINED FROM NET COLLECTIONS AND ALBACORE STOMACHS David C. Judkins and Abraham Fleminger^ ABSTRACT Sergestes similis, an oceanic shrimp, was taken at a number of locations in the eastern North Pacific, principally in the California Current region. The contents of foreguts from shrimp caught by net during the day and night and those of shrimp eaten by free-swimming fish were compared. In all three categories of foreguts the predominant prey were adult specimens of the larger, common calanoid cope- pods which typically inhabit the upper 200 to 300 m in the California Current region. However, be- cause the diversity of calanoid species and the numbers of fish scales, calanoids, and euphausiids were appreciably greater in the foreguts of net-caught S. similis than in fish-caught samples, it appears likely that S. similis feeds in collecting nets under tow. Sergestes similis Hansen is an abundant pelagic shrimp endemic to North Pacific waters of bo- real-temperate influence (Pearcy and Forss, 1969; Judkins, unpublished data). Examina- tion of its stomach contents indicates it is pre- daceous and feeds primarily on copepods and euphausiids (Renfro and Pearcy, 1966). There is an expanding body of evidence that, in the freshwater environment, predation by planktivores is size selective and determines, in part, the composition of zooplankton communi- ties (Brooks and Dodson, 1965; Brooks, 1968; Dodson, 1970 ; Hall, Cooper, and Werner, 1970) . Size-selective feeding by an abundant oceanic carnivore such as S. similis may play an im- portant role in limiting the abundance of an array of prey species within a size range and, hence, in determining the composition of the zooplankton community within its habitat. The first step in determining the impact of S. siynilis as a predator in the community is to identify and to enumerate the prey species it utilizes. In this report we identify and enumerate the foregut contents of S. similis from net tows and albacore stomachs taken at a number of local- ities in the eastern North Pacific. To determine if the results were affected by feeding in the ^ Scripps Institution of Oceanography, University of California, San Diego, La JoUa, CA 92037. net or by diurnal changes in feeding intensity and diet, comparisons were made between three categories of specimens: day-net samples, night- net samples, and fish-stomach samples. MATERIALS AND METHODS About two thirds of the 270 foreguts with contents examined in this study were obtained from 5. similis collected by nets, principally Isaacs-Kidd midwater trawls, over several sea- sons (Table 1). The remaining one third were obtained from troll-caught albacore taken in July-August 1968 (Laurs and Nishimoto, per- sonal communication). With the exception of one net tow taken in the Gulf of California all of the collections were made in the northeast Pa- cific Ocean between lat 31° and 53° N (Table 1). Carapace lengths (measured from the tip of the rostrum to the dorsal mid-point of the pos- terior margin) of the net-caught shrimp ranged from 6.5 to 17.0 mm, with a median of 11.3 mm. Carapace lengths of fish-caught shrimp ranged from 6.3 to 12.5 mm, with a median of 8.3 mm. Because many fish-caught shrimps were partially digested, it was necessary to estimate their cara- pace lengths from the lengths of their foreguts (Judkins, unpublished data). The foreguts were removed intact from the specimens with fine forceps and placed on glass Manuscript accepted September 1971. FISHERY BULLETIN: VOL. 70, NO. I, 1972. 217 FISHERY BT-XLETIN: VOL. TO. SO. 1 Taxlx 1. — Material examined. ^^irion Uflituda '»: — -- -.- r^r. k' :: A-c -i" •t;' 3S- 15:° OJ- - .0 13 - : . - '-■.■" ::6-27 Mar. i?cu: 37° 37' 123° 37- 2ri--jjri 1 ri, 9 V,'.' 26 Mo- 1962 36' 34* 125° 4c- j."';-." -" 293 K) ijj' Ir r- ^rl 2A' 18' 120' 4S' , - . - . 2=6 )0 k'.'" ;- :;: 3<* 07' 121' oe* c :: :;; 3J0 >a i CMT -- Apr. Irol 31* 21' IIT" 27« 0229-0414 298 3 Z .rsa Major 10 2-t-Snalkm ■= Aj9- 23 Ai^- 1?6< /J' OT 3S' 155" 155° 00- CO- 0327-O416 2216-7738 245 10075 19 4 f«BV II BrIi 4 -jAs- T» U rsb- :=66 ■AS" 12" 148* 37" 0622-0656 450 n io.fr 20 Jebi. 1965 29= 113* 2245O507 i.roo ■: :o V-r>— *- -r ozrr 13 Cv~ 13 Ajg. 1951 53° 35' li6° Co- 1345-1655 450 6 Zc:: - 60J^ ilv* 27 Mar. 1962 37- 07' ■ 24° ^5. :.D- 3-0750 298 4 i2.1» C'/" 28 Mar. 1962 36' 17' 126° 3 " ■ ; £.39-1776 l,fl63 3 KTO iKxia -.7 M3-. 1962 33* 54- 121° sy 09AS-1424 1 IVi? 5 • J-sc Mo'fsr 250Bep BMOC 2? Aug. 1964 /.£= 15' 155' 02' 1640-1809 300400 n >- 5 roicr 11 Si-rh 3 T,ir+=r T U Feb. 1966 ■c?" 54' 148° 23' 1519-lSv? — M S»0 67-101 icwcr 9 Jj-» 1967 35* ^& 122° 16' r«77-1534 2,200 m S»0 67-102 aoa 13 JjTB 1967 35- 19- 123° 06- 0705-1403 7 7(10 » S>0 67-' 13 ma U Jure 1967 37* 20- 123° 42" 0558-1150 7;>(W B 65 '» jILOf. i2:»<-n 90.6: iKwrr 6 Apr. 1962 72' 2i' 11 ' JD' ^4304)629 '-^5 3 ■^ S*0 66^1 IKMT 21-22 May 1966 jiC° 35' 125' 52- 2D03-O543 1.125 11 _o SO 67-C lowrr 2a Apr. 1967 31* 2& ■■7° 54' 1535-1 923 4i)U) 10 J 31 - ' .: ';-::- I"; TrcHi 16 JlfI/ 196E — 36° ~ 125' 1815 il H2 17 July 196B -34- — 124° , ^ 2 ;*■ IS Jjty 1968 ~35' — 126° 1213 24 ; li; 18 Juiy 196B -36" - 126° IHU 10 ^ ?2 19 JL-Iy "95: -39° ~ 127' _ ^ 17 § 179 20 Juty I960 ~42' ~ 127° 2 - 140 21 July 1963 — .: , : ~ 127' 4 -= 58 2 Aug. 196S — .! -i° ~ 135' 5 — ? 12 Aug. 1968 -45" ~ 124' n55.S 15 2i4 18 Aiig 1965 ~3 6' ~ 125° 0615 10 100 s joart OMJ Cdd. *il tttCiomim and Btomv. .vtiA. ImckBr, 1951. microscope slid^ where their contents were emptied into 100 *'r g^lycerol. All observations were made with aid of a stereomicroscope or a compound microscope at various magnifications as needed from 16 x to 1000 x . The contents were examined to identify and enumerate all particles. Se%eral drops of lactic acid tinted with chlorozol Black E were added to the glycerol to facilitate identification of the comj)onents. Lactic acid aids in clearing the preparation. The tint stains the components and is especially useful for arthropod cuticle. The identitj- of crustacean prej* was often established from the morphologv* of mandibles, genital seg- ments, fifth legs, or other diagnostic features. In Table 1, collection numbers grouped bj' the designation "night-net" represent a set of 10 samples varjing in number from 3 to 10 shrimp each and collected by nets that were towed ex- clusively between sunset and sunrise. Collection 218 JUDKJSS aad nOOSGEBiz FOtLECCT CONTEXTS OF Sergfita timiEt numbers designated "day-net" comprise a set of nine samples of from 3 to 10 shrimp each and collected by nets that were towed exclusively be- tween sunrise and sunset. "Twilight-net" col- lection numbers represent three samples collec- ted by nets which were sr*:^ >"v towed at sunset or sunrise. Collection n.: designated '*fish stomachs" represent 10 samples of 2 to 24 shrimp taken from stomachs of albacore. The albacore were taken during daylight hours by trolling lines; exact times of capture were not available for five of the si)ecimens. Included in statistical analyses were all classes of food occurring in 10% or more of any of the sets of night-net, day-net, and fish samples. Cal- anoid genera occurring in 10% or more of any of the sets of samples were considered separately. The small number of twilight-net samples was not included in statistical comparisons of the sampling categories, but the details of this set are given in Table 2. Overall mean? (no. prey foregut) and frequencies (*■>) were calculated for the three sets of samples that were compared. In addition, means were calculated for each night-net, day-net, and fish sample. The Mann- Whitney f '-Test, a test designed to estimate the significance of differences in median values of two samples was used to compare median sam- ple means between day- and night-net samples and between combined day- and night-net sam- ples and fish samples. RESLTTS Ingested items inc t euphausiids, ostra- cods, amphipods, ch£.e: _ . ths, and fish scales, but the principal identif t ::>mponents in aU sampling categories were aduit calanoid cope- pods (Table 2). The composition of foregut contents from different geographical regions (North Pacific Subarctic, Xorth Pacific Drift, and the Califomia Current) were very similar and hence were not considered separately. The diversity of calanoids was much greater in aU categories of net samples than in fish samples: 42 species were identified in net samples, but only 7 were found in fish samples. Foregut contents varied widely in the extent of maceration by digestion and mastication. curzeiiee in the f or^;ats of Sergestes ghmUs. ^..zr: Dtoy- T«Stg«*. Rdi '•6* -;er «tef C-ceccc- JifMMj.-ir -^rrrci 4 Jf--z.ti.-7j:i: 1 JfrU-^s 1 ■3 C^iMmmi cristmems 1 C^Umms tmcHtms 20 1 3 Commas tlmwuhrui 2 CtUaas IS 9 3 CmmJscif UfmrnsU 2 m CtmJ^Kim I 7 Ck^m£mM "l CUmxacJaaai fmnstmi 1 O^miisadmams ^mmftrjfmi 5 "3 CSmx^svuImbbs 3 2 Caiytmna 11 CmtytOm T Eaialxm»i fos{B 5 1 "2 EaealMm 4 1 Emtimdf 8 7 "4 "T EmtMrdUtakkn 1 2 EaeUrellM mtratm ~ 1 Gmet^mas ~I Gonial fmmeemi 2 • Cmi&u 4 " CmmiM 2 HebrrmriMUmi mkynaBs 2 "2 HtUwmUkimi 2 2 1 Lmeicmtm "T MetriSm tmitmrndm 1 ilitriJim pacif 17 W "5 B itftri£m 20 II 2 Omcmem 3 1 ~2 1 OkUmM I 1 Parme»lMmaLt fmrrms 1 P^mtJmKU 4 Pmmduutm 2 PkrHtru 1 ~7 "2 Wi'ii ■■«■■■ immCt E ~1 1 "T 2 WrwawM sjptan I ~T ~l PiemwwmMmatm II « 4 ~i 6 !• 2 ^^^^rt^^^TT w^swSbs 13 - KtimtJau 7 J. BmnaikMams ^ 1 ScafUaltmmt mm^tas ~I SeaLr^hriaBm akfst^t ~I SewlmthrirtOm dbatMa "T SeaUckkriuOm n^ "i 3 SidtcitkntdU mmta 3 1 SpimmcJammi 1 3 ~I — ~x — — "1 r«/nci«cte Htfiamf 1 rmJfaduKtm "2 "T umdenfofied adamani 14 !4 1 II — — I Eiu^SrnoiuiESifcxBO Tiysmm^essm 12 unouEinffirnGsdl 19 16 "6 "2 *L^tuuKjUUU C»«*«nM 2 1 unndenlined ~6 1 5 Amp^iqpada^ unudeniCvffiea S 9 6 Chod&u|j|iiuWio MAfartiBal 12 1 2 Rsn soDies 17 16 3 __ Rdieggs I — . ITiuiunmi'r ■ f-z 2 , __ Bo&ifarc- , t II , 1 ***«°»- ----- 1 1 — 1 — liWUQSCZI ^— — ^ Gnved r 1 UWdotfr- 30 22 Ml 49 rahri nur- ■^r- .- . :c -~ 5- ag 219 FISHERY BULLETIN: VOL. 70. NO. 1 Specimens in the same sample frequently varied from one another, and the degree of decomj^o- sition in each specimen usually was not uniform. Lightly digested prey were largely intact, and their internal organs still visible. Moderately digested prey contained obviously macerated tis- sue and were often incompletely fragmented. Prey in an advanced state were highly frag- mented, and soft tissues were absent. Digestion of foregut contents tended to be less advanced in night-net than in day-net samples. However, most foreguts in both categories contained ma- terial in an advanced state of digestion. Five groups of ingested material, calanoids, euphausiids, chaetognaths, fish scales, and un- identifiable matter, occurred in lO^r or more of at least one of the three sampling categories (Figure 1). Unidentifiable material consisted primarily of crustacean fragments and matted aggregations of fibrous and granular debris. Of identifiable groups, calanoids were usually most abundant in all three categories (Table 3) . The four identifiable groups of prey organisms appeared in higher numbers in net samples than in fish samples, diff"erences between the median values, with the exception of chaetognaths, being significant at P ^ 0.05. Frequencies, overall means, and median sample means of calanoids and euphausiids in day-net samples were less than were those in night-net samples. However, differences be- tween the day and night median values of the four identifiable groups were not significant (P > 0.50). Twilight-net samples do not ap- pear to diflfer appreciably from day- and night- net samples. Calanoid genera occurring in more than 10 '^r of the three categories (Figure 2) were Metridia CALANOIDS EUPHAUSIIDS FISH SCALES CHAETOGNATHS UNIDENTIFIED Figure 1. — Frequencies of ingested items occurring in lO'/c or more of night-net (black), day-net (open) and fish (strippled) samples. (principally pacifica), Calaviis (principally pa- cificus), Pleuromamma (principally borealis), Rhincalanus (all nasutus) , Euchaeta, Eucalanus (principally himgii) , Scolecithricella, and Can- dacia (principally hipinnata) . These genera have been found typically in the uppermost 200 to 300 m of the California Current region (Flem- inger, unpublished data). The overall mean number of Calanus in day- and night-set samjjles are notably diflferent (Table 4) . In night-net samples Calanus ranked with Metridia, the two dominating the list of Tablf, 3. — Median of sample means and overall mean of number of prey per foregut. Prey Night-net Day-net Fish Median Range Mean Median Range Mean Median Range Mean Calanoids 2.40 0-12.50 4.23 2.25 0.67-5.70 • 2.68 0.63 0-3.20 0.96 Euphausiids .33 0- 2.90 .66 .10 0-1.50 .32 0 0- .12 .02 Fish scales .10 0- 1.00 .31 .23 0-3.00 .37 0 0 Chaetognaths .20 0- .30 .15 .16 0- .33 .14 0 0- .50 .02 Number samples 10 9 10 Total number foreguts 86 65 88 220 JUDKINS and FLEMINGER: FOREGUT CONTENTS OF Sergestes limilis METRIDIA CALANUS PLEUROMAMMA RHINCALANUS EUCHAETA EUCALANUS SCOLECITHRICELLA CANDACIA ZJ Figure 2. — Frequencies of calanoid genera occurring in 10% or more of night-net (black), day-net (open), and fish (strippled) samples. calanoid prey. In day-net samples Calanus num- bers are considerably below those of Metridia which remained high. However, the difference between the median sample means of Calanus in day- and night-net samples is not significant (P > 0.20) . This is also true of the other seven genera. Metridia was numerically dominant in fish samples, although it occurred in lower numbers than in net samples. Differences between the median sample means of Metridia in net and fish samples, however, are not significant (P > 0.40). Differences between net-and-fish- caught S. similis in the median values of Euca- lanus, Euchaeta, and Rhincalanus are significant at P values falling between 0.10 and 0.20. Dif- ferences among the remaining calanoid genera are significant at P values between 0.02 and 0,10. DISCUSSION Metridia jyacifica, Calanus pacificus, Pleuro- mamma abdominalis , Rhincalanus nasutu^, Eu- cakinus bungii californicus, and Candacia bipin- nata are among the 24 most abundant and frequently occurring of 176 calanoid species found in zooplankton samples collected on Cal- COFI Cruises 5804, 5807, 5810, and 5901 (Flem- inger, 1967). Significantly, these are also the principal species of six of the eight genera most frequently occurring in the foreguts examined in the present study. Adults of these species exceed 3 mm in length. It appears, then, that the principal prey of S. similis are the more abundant, relatively large, adult copepods in- habiting the uppermost 200 to 300 m in the Cal- ifornia Current region. Additional items found in both net and fish samples are euphausiids, chaetognaths, ostra- cods, amphipods, and radiolarians. Of these, Table 4. — Median of sample means and overall mean in calanoid genera occurring in more than 10% of foreguts. [Means expressed as number of prey per foregut.] Prey Night-net Day-net Median Range Mean Median Range Mean Median Fish Range Mean Metridia 0.30 0-4.10 0.95 0.56 0- Calanus .40 0-5.90 1.16 0 0 Pleuromamma .30 0-1.10 .36 .10 0 Rhincalanus .20 0-2.00 .35 0 0 Euchaeta .10 0- .20 .10 0 0 Eucalanus 0 0- .30 .10 0 0 Scolecithricella .10 0- .40 .13 .10 0 Candacia 0 0 0 Number samples 10 10 Total number foreguts 86 65 1.60 .80 .80 .30 .40 .50 .30 0.78 .17 .15 .09 .14 .11 .12 0 0.36 0 0 0 0 0 0 .10 10 88 0-2.90 0- .09 0- .06 0- .50 0.69 0 .02 0 .01 0 0 .11 221 FISHERY BULLETIN: VOL. 70. NO. I only euphausiids were reported in a previous study on the diet of S. similis (Renfro and Pearcy, 1966). A large percentage of shrimp of all sampling categories contained mixtures of granular and fibrous debris. The unidentifiable state of this material may not be entirely the result of di- gestive processes, but of the decomposed nature of the material at the time of its ingestion. Ser- gestes similis, like S. lucens (Omori, 1969), may scavenge decomposing dead material in addition to taking living prey. Another possibility is that this material represents the remains of prey stomach contents, e.g., those of euphausiids. There is a tendency for day-net overall means to be less than those of night-net samples (Table 3). This and the trend toward less advanced digestion in night-net samples suggest more in- tense feeding activity at night, as was reported for S. lucens (Omori, 1969) . The lesser average number of calanoids per foregut in day-net samples may be attributed primarily to the no- tably fewer Calanus in that category (Table 4). Although the median values are not statisti- cally different, day and night differences in over- all mean numbers of the two calanoid species most frequently occurring in net samples, Me- tridia pacifica and Calamis pacificus, are notable in that they agree with differences in the vertical distributions of these two species. In the Cal- ifornia Current region south of lat 33° N, Cal- anus pacificus usually concentrates in and near the thermocline during the day and disperses throughout the mixed layer at night. M. pa- cifica, on the other hand, occurs in the vicinity of the thermocline at night and disperses down- ward during the day (Fleminger, unpublished data). Studies have shown S. similis to be con- centrated between the surface and 200 m at night and between 250 and 500 m during the day (Barham, 1957; Pearcy and Forss, 1966). The diurnal vertical range of M. pacifica, then, seems to correspond more closely with that of S. similis than does the diurnal range of C. pacificus. Ser- gestes similis probably has access to quantities of M. pacifica during both day and night. Most likely, S. similis encounters and feeds upon con- centrations of C. pacificus primarily at night after the sergestid has ascended to shallower depths. However, without knowledge of di- gestive rates, these considerations are specu- lative. The large numbers of fish scales in net samples and their complete absence in fish-stomach sam- ples strongly suggest that feeding in the net has occurred. Although the source of these scales cannot be ascertained, it seems probable that they are the highly deciduous scales of lantern- fish captured in the net with the shrimp. The significantly greater numbers of euphausiids and chaetognaths and the higher diversity of cala- noids in net samples may also be indicative of feeding after capture. Many preserved net- caught S. similis have been observed by one of us (Judkins) to have fish scales, chaetognaths, and small crustaceans packed into their mouth- parts and sometimes gripped in their mandibles. Special conditions under which fish-caught shrimp might have been feeding before capture (related perhaps to time or depth) may also have contributed to the observed disparities between net and fish samples. Albacore are thought to feed primarily during daylight hours and prob- ably most intensively in the early morning and early evening (Iversen, 1962). Time of capture by trolling of the albacore that we examined is available for only about half of the specimens. Of these about half were taken in morning day- light hours and the remainder were from the late afternoon and early evening. If all of the fish-caught shrimp were captured during the day, the relatively small quantities of identifiable components might reflect less intensive feeding by the shrimp at the time they were ingested by the albacore. The generally smaller size of fish-caught spec- imens is probably not a factor; small net-caught shrimp (carapace length less than 10 mm) con- tained numbers of fish scales, euphausiids, and chaetognaths proportionally as high as larger net-caught individuals. Differences amongst sergestid specimens in the length of time spent in an albacore stomach also appear to be a neg- ligible factor. Foregut contents of nearly in- tact fish-caught shrimp do not differ appreciably from those of extensively decomposed shrimp. In general, the various samples of foreguts from adult S. similis that were analyzed provide 222 JUDKINS and FLEMINGER: FOREGUT CONTENTS OF Sergestes similis vide a coherent pattern of prey organisms con- sisting principally of the larger, commoner spe- cies of calanoid copepods inhabiting the upper- most 200 to 300 m of the California Current region. However, it is likely that feeding by S. similis in the net while it is being towed may explain the greater number and diversity of prey items in net-caught shrimp. ACKNOWLEDGMENTS We thank Dr. R. M. Laurs and R. N. Nishi- moto, NMFS Fishery Oceanography Center, La Jolla, for providing us with the albacore-caught S. similis. This research was supported by the Marine Life Research Group, SIO, and by the National Science Foundation, including the Sea Grant College Program. LITERATURE CITED Barham, E. G. 1957. The ecology of sonic scattering layers in the Monterey Bay area. Stanford Univ., Hopkins Mar. Stn. Tech. Rep., No. 1, 182 p. Brooks, J. L. 1968. The effects of prey size selection by lake planktivores. Syst. Zool. 17: 272-291. Brooks, J. L., and S. I. Dodson. 1965. Predation, body size and composition of plankton. Science (Washington) 150: 28-35. Dodson, S. I. 1970. Complimentary feeding niches sustained by size-selective predation. Limnol. Oceanogr. 15(1) : 131-137. Fleminger, a. 1967. Distributional atlas of calanoid copepods in the California Current region, part II. California Cooperative Oceanic Fisheries Investigations Atlas No. 7, 213 p. Hall, D. J., W. E. Cooper, and E. E. Werner. 1970. An experimental approach to the production dynamics and structures of freshwater animal communities. Limnol. Oceanogr. 15(6): 839-928. Isaacs, J. D., and L. W. Kidd. 1953. Isaacs-Kidd midwater trawl. Univ. Calif., Scripps Inst. Oceanogr., Equip. Rep. 1 (SIO Ref. 53-3), 21 p. Iversen, R. T. 1962. Food of albacore tuna , Thunntis germo (Lacepede), in the central and northeastern Pa- cific. U.S. Fish Wildl. Serv., Fish. Bull. 62: 459-481. McGowAN, J. A., and D. M. Brown. 1966. A new opening-closing paired zooplankton net. SIO Ref. 66-23, 56 p. Omori, M. 1969. The biology of a sergestid shrimp, Sergestes lucens Hansen. Bull. Ocean Res. Inst., Univ. Tokyo 4, 83 p. Pearcy, W. G., and C. A. Forss. 1966. Depth distribution of oceanic shrimps (De- capoda; Natantia) off Oregon. J. Fish. Res. Board Can. 23: 1135-1143. 1969. The oceanic shrimp Sergestes similis off the Oregon coast. Limnol. Oceanogr. 14(5) : 755-764. Renfro, W. C, and W. G. Pearcy. 1966. Food and feeding apparatus of two pelagic shrimps. J. Fish. Res. Board Can. 23: 1971-1975. Tucker, G. H. 1951. Relation of fishes and other organisms to the scattering of underwater .sound. Mar. Res. 10(2): 215-238. 223 NOTE LONGEVITY AND GROWTH OF TAGGED KING CRABS IN THE EASTERN BERING SEA During the period 1957 through 1959, the Bureau of Commercial Fisheries (now the National Ma- rine Fisheries Service) released 32,328 tagged king crabs, Paralithodes camtschatica, in the eastern Bering Sea. Since then several thou- sands have been recovered, 23 of which exceed previously reported maximum ages for this spe- cies (Table 1) . These crabs were originally cap- tured by Bureau research vessels, marked with serially numbered plastic spaghetti tags inserted through the isthmal muscle (method described in Alaska Fisheries Board and Alaska Department of Fisheries, 1955: 34-43) , and released immedi- ately. Crabs were recaptured by the crab tangle net fisheries of Japan and the Soviet Union, and recapture data were provided by the two nations as part of annual exchanges of scientific infor- mation called for by bilateral fishing agreements with the United States. We estimated the ages of the crabs at release from size and age data for young king crabs published by Weber (1967). Weber found that both male and female king crabs in the eastern Bering Sea mature at about 95 mm carapace length; that males reach this size in 5 years and females in 514 years (assuming a hatching date of late April to early May) ; and that im- mature crabs longer than 60 mm increased about 16 mm per molt. Growth curves for both sexes were similar up to the fourth year of life (length of 80 mm) , but after the females become ma- ture they grow slower. The males, however, continue to grow about 16 mm per molt through- out the rest of their lives (Weber and Miyahara, 1962; Hoopes and Greenough, 1970). Total age was estimated for all four males, and length data are available for three of the four (Table 1). One male released in 1957 was recaptured in 1968, having been at liberty for 11 years. This crab was estimated to have been 6 years old at time of release, and if this esti- mate is correct, it was 17 years old at time of recapture. If the average growth per molt was 16 mm, this crab molted three or four times dur- ing the 11 years between release and recovery, and the other two crabs for which growth data are available molted only once or twice in 9 years. This molting frequency is much lower than that reported by Weber and Miyahara (1962), but we have no explanation for the reduced rate of molting in the two crabs. Before this study, the oldest known-age king crab reported in the literature was a male that was tagged near Ko- diak Island and recaptured 20 miles from the release location 6 years and 4 months later (Powell, 1965) . Powell estimated that this male was 7 years old when tagged and 13 years old when recaptured. Table 1. — Carapace lengths and estimated ages of 4 male and 19 female king crabs tagged in the eastern Bering Sea by the National Marine Fisheries Service in 1957, 1958, and 1959 and recovered 1966, 1968 and 1969. Sex and tag no. Release date Recovery date Carapace length at- Estimated age at- Release Recovery Release years 6 Recovery Males B5178 9/9/57 9/16/68 mm 108 mm 170 years 17 B7013 4/29/58 5/18/66 124 1 7 15 B7364 5/2/58 3/30/68 109 132 6 15 C8909 5/14/59 9/25/68 108 133 6 15 Females B7535 5/3/58 5/21/68 97 138 5 15 B7560 5/3/58 6/26/68 94 152 5 15 B7567 5/3/58 6/8/69 99 149 5 16 B7906 5/4/58 5/10/66 109 1 6 14 B8008 5/4/58 4/30/68 110 155 6 15 B80I5 5/4/58 5/18/66 108 1 6 14 B8552 5/11/58 3/28/69 108 149 6 16 B9I50 5/26/58 5/8/69 90 138 5 15 C2IH 6/17/58 5/8/69 130 162 8 18 C2508 6/17/58 4/30/68 99 140 5 14 C256I 6/17/58 6/21/68 114 157 7 17 C2785 6/17/58 5/9/68 77 136 4 13 C2804 6/17/58 5/20/68 108 150 6 15 C3871 6/27/58 5/5/68 97 162 5 14 C3969 6/27/58 5/19/68 94 147 5 14 C4I06 6/27/58 6/14/68 107 137 6 15 C4423 6/28/58 3/29/68 104 141 6 15 C4837 6/29/58 9/ 1/69 102 136 5 16 C5961 7/5/58 5/17/68 91 120 5 14 1 Not avo liable. 225 Length and age data are available for 17 of the 19 females (Table 1). One attained an es- timated age of 18 years, having been at liberty for 10 years after being tagged. If these 17 females molted once each year, their growth per molt ranged from 3.1 to 7.2 mm (length) ; the average increment was 4.7 mm. This aver- age value is similar to other reported annual growth increments for female king crabs in Alaska waters— 4.4 mm (Powell, 1967), 3.9 mm (Gray, 1963), 5.0 mm (Bright, Durham, and Knudsen, 1960), and 4.0 mm (Sakuda, 1959). Literature Cited Alaska Fisheries Board and Alaska Department of Fisheries. 1955. Alaska Fisheries Board and Alaska Depart- ment of Fisheries Annual Report, 1954, 92 p. Bright, D. B., F. E. Durham, and J. W. Knudsen. 1960. King crab investigations of Cook Inlet, Alaska. Dep. Biol., Allan Hancock Found., Univ. South. Calif., Los Angeles, 180 p. Gray, G. W., Jr. 1963. Growth of mature female king crab, Parali- thodes camtschatica (Tilesius) . Alaska Dep. Fish Game, Inf. Leafl. 26, 4 p. HOOPES, D. T., AND J. W. Greenough. 1970. King crab research. Int. North Pac. Fish. Comm Annu. Rep. 1968: 116-124. Powell, G. C. 1965. Tagged king crab recaptured six years after release in the North Pacific. Trans. Am. Fish. Soc. 94: 95. 1967. Growth of king crabs in the vicinity of Ko- diak Island, Alaska. Alaska Dep. Fish Game, Inf. Leafl. 92, 106 p. Sakuda, H. M. 1958. Observations of molting female king crabs {Paralithodes camtschatica). U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 274, 5 p. Weber, D. D. 1967. Growth of the immature king crab, Parali- thodes camtschatica (Tilesius). Int. North Pac. Fish. Comm. Bull. 21: 21-53. Weber, D. D., and T. Miyahara. 1962. Growth of the adult male king crab, Parali- thodes camtschatica (Tilesius). U.S. Fish Wildl. Serv., Fish. Bull. 62: 53-75. David T. Hoopes and John F. Karinen %lational Marine Fisheries Service Alike Bay Fisher'ies Laboratory Auke Bay, AK 99821 226 JAPANESE DISTANT- WATER FISHERIES: A REVIEW HiROSHi Kasahara^ ABSTRACT Most of the industrialized fisheries of Japan have developed under a licensing system controlled by the central government. Limitations on entry and the allocation of re- sources based on a variety of social and economic considerations have resulted in the development of an extremely diversified industry. The postwar expansion of distant-water fisheries greatly accelerated the exploitation of resources in the North Pacific, as well as in many other areas of the world, and has caused numerous international conflicts. In addition to regional conventions for high seas fisheries, various bilateral agreements have been negotiated to cope with problems arising from jurisdictional claims by coastal states. While supporting narrow limits of national jurisdiction and the concept of free access to high seas fishing, Japan has accepted different forms of allocation as a means to accommodate the conflicting interests of the nations involved. Her domestic institutions and organization of the industry have helped the government make pragmatic arrangements with other nations. Whether or not a general agreement on fishery issues can be reached at the new Law of the Sea Conference, Japan will face more and harder international negotiations in view of the general trend of coastal states claiming broader zones of national jurisdiction. Each of the main sectors of the Japanese fishing industry, including inshore fisheries, offshore fisheries, distant-water fisheries, and aquacultui'e, now operates under severe constraints. Although the total catch of distant-water fisheries is still increasing due largely to intensified pollack fishing in the Pacific, long-term prospects for further ex- pansion do not appear bright. Little progress has been made in the utilization of abundant resources of unconventional species. Thus, the rapid growth of domestic fish- ery production is unlikely to continue. Increased joint ventures and other business ar- rangements in foreign countries may provide a partial solution. Import decontrol for fishery products would contribute substantially to meeting immediate problems of supply shortage. This paper was originally drafted to provide, as part of the NORFISH study under the Wash- ington Sea Grant Program which is supported by the National Oceanic and Atmospheric Ad- ministration, some background information on the development and the present status of the Japanese high seas fisheries, particularly those which have bearing on various international arrangements in the North Pacific. Since, how- ever, discussions on the future regimes of the sea have been carried out with increasing in- tensity, the emphasis of the paper has shifted somewhat from descriptive information to a more analytical study of the international fishery ^ Contribution No. 360, College of Fisheries, Univer- sity of Washington. '' College of Fisheries, University of Washington, Se- attle, WA 98195. problems faced by the Japanese government and industry, as well as the courses of action they are likely to take in response to future changes in international regulatory regimes. The im- portance of the topic in considering future international arrangements for fisheries is ob- vious, for the Japanese and Soviet distant-water fisheries have been among the major sources of international conflicts over fishery matters in various parts of the world. Although emphasis is on the North Pacific, developments in the rest of the world are also covered to the extent that they have bearing upon the situation in the North Pacific. The present paper is not a comprehensive study of the Japanese fishing industry to ex- amine closely all sectors of the industry, in- cluding inshore, coastal, and distant-water fish- eries, as well as processing and marketing Manuscript accepted January 1972. FISHERY BULLETIN: VOL. 70, NO. 2, 1972. 227 FISHERY BULLETIN: VOL. 70, NO. 2 aspects. It emphasizes those fisheries the devel- opment of which have had substantial effects on international regulation of fisheries, except for such passing references to other elements of the industry as considered relevant.' One of the serious problems about writing an English paper concerning Japanese fisheries is that fishery institutions in Japan are extremely complex and difficult to understand. It is al- most impossible for foreigners to fully compre- hend institutional aspects of the Japanese fishing industry without having been in the country for some time. English literature in this area is meager. Yet, domestic institutions for fisheries have had such tremendous effects on develop- ment of all Japanese fisheries, including those in distant waters, that it is often irrelevant to discuss their problems without having some understanding of the institutional framework in which they operate. To obtain some general in- formation on this aspect, the readers of this pa- per are referred to four English papers: Oka, Watanabe, and Hasegawa (1962), Kasahara (1964), Comitini (1967), and Herrington (1971). All high seas fisheries discussed in this paper are rigidly regulated by what is called "the li- censing system." The system controls the acti- vities of each fishery through restrictions on the total number of licenses to be issued, size of ves- sels to be used, area of fishing, method of fishing, and often species to be taken. Although the ac- tual regulations under this system differ from fishery to fishery, a common, and most important, feature is direct control on the number and types of vessels to be used for a particular fishery. The justifications used by the government for imposing the limited entry system on offshore fisheries has varied. Among apparent objectives ' The author excluded whaling from the present paper due to lack of time. The history of the whaling industry is a story of its own, and may better be dealt with as a separate topic. Two nations, Japan and the Soviet Union, are more responsible than others for the present state of baleen whale stocks. Their recovery in the Ant- arctic would take many years even with restrictions more severe than tho.se currently enforced. The relative im- portance of whaling in the Japane.se fishing industry has decreased rapidly in recent years. While the omission of the topic affects the comprehensiveness of first two sections, its inclusion would not change greatly the sub- stance of the last section. are: protection of inshore fisheries against off- shore fishing, reduction of competition and pre- vention of disputes between different groups of offshore fishermen, stabilization of fishing con- ditions, maintenance of profitability, conserva- tion of resources, prevention of international disputes, and others. Degree of success in achieving these objectives has also differed from case to case, but there is no question that the system has served as a powerful and convenient means to control each fishery and introduce such changes as considered desirable by the Japanese fishery administration.* Practically all offshore and distant-water fish- eries discussed in this paper are regulated by the central government. While legal authority is vested in the Minister of Agriculture and For- estry, the Fishery Agency (Suisancho) , which is subordinate to the Ministry of Agriculture and Forestry, has in fact full power to control all major fisheries. There still exist a large num- ber of small fisheries regulated by the provincial authorities, but they have practically no inter- national implications, except those operating off the southernmost part of the Kurile chain and in Korean Straits. Chapter 3 of the Fishery Law, as amended in 1963, provides that anyone wishing to be engaged in any of so-called "des- ignated fisheries" must be licensed by the Min- ister of Agriculture and Forestry, and spells out principles under which such licenses are issued. Administrative ordinances specify the designat- ed fisheries and the types of regulation under which they operate. This category includes practically all important fisheries carried out in waters far from the home islands, as well as the coastal trawl fisheries conducted by medium vessels and purse-seine fisheries by medium and * Scholars in North America approach the question of limited entry in fisheries mainly from the point of view of economic efficiency. Application of limited en- try in the Japanese fishery administration is based on rnuch more diversified considerations. The transfera- bility of licenses, which is an essential condition for maintaining economic efficiency under this system, has been subject to increasing constraints in Japan. While the old fishery law established fishing rights as freely transferable private properties, the new fishery law (1949) specifically prohibited transfer. The new law also prohibited in principle the transfer of licenses for offshore fisheries; in reality licenses were still trans- ferable in most cases; but the 1963 revision of the law further restricted the transferability of licenses. 228 KASAHARA: JAPANESE DISTANT-WATER FISHERIES large vessels. The only major fisheries which do not fall in this category are the saury and squid fisheries, which are regulated by separate ordinances." Measures to further control the operation of the designated fisheries are provided in policies for licensing which are issued by the Ministry from time to time. Licensees of each of the designated fisheries are obliged to submit reports of their operation according to the for- mat specified by the Minister. In short, practically all important Japanese fisheries are strictly controlled by the central government under the licensing system. The administration of fisheries under this system is, of course, subject to pressures from diflferent groups in the industry, including large fishing companies, vessel owner associations, and fish- ermen's associations, but changes in fishery pol- icies are brought about only through this cen- trally controlled system. The system is also eflfective in accommodating such changes as the government and industry consider necessary for meeting new international developments. It is also responsible for the coexistence of many dif- ferent types of fishing operation for catching the same species. For example, the present pat- tern of Japanese trawl fishing in the North Pacific, which employs all types of trawl gear and vessels of enormous size range, could not have developed without continuous manipulation of the system by the government under pres- sures from various sources. REVIEW OF SELECTED HIGH SEAS FISHERIES Most of the important Japanese fisheries are conducted on the high seas as defined by Japan. This review includes only those which have had or are likely to have international problems. Table 1 compares the catches (round weight) of such fisheries. Some of them, particularly trawl fishing by medium vessels and purse sein- ing, are basically coastal, but they too have in- ternational implications in relation to the fish- eries of Korea, China, and the Soviet Union. Table 1.— Catches of marine fisheries, 1969. Categories Catch (Metric tons X 10^) High seas salmon fisheries 123 Tunc and skipiack fisheries^ 586 Trawl fisheries Distant water^ 2,290 China Seas 304 Coastal 608 Mothership crab fisheries 44 Distant-water longline/gill-net fisheries^ 27 Large and medium vessel purse-seine fisheries 1,041 Mackerel angling fishery 134 Saury fishery 51 Squid angling fishery 467 All other fisheries* 2,301 Aquaculture^ 473 Total 8,449 Whaling 2,510 6,668 627 blue whale units sperm whales (head) small whales (head) " These fisheries are still not subject to strict limited entry, but saury vessels and larger squid boats must be approved by the Minister and thus are subject to var- ious regulations established by the central government. 1 Longline and pole-and-line (excluding catches by vessels smaller than 20 gross tons). 2 Not including the China Seas, ^ Other than tuna longline fisheries. * Including inshore and coastal fisheries, as well as collection of clams (weight with shell) and seaweeds. ^ Including oysters (with shell) and seaweeds. Source: Ministry of Agriculture and Forestry (Japan) (1971). TRAWL FISHERIES Developments in the Prewar Period A wide variety of fishing operations can be found even within one technical category, trawl- ing. The coastal waters of Japan are crowded with a vast number of small draggers employing a great many diflferent types of gear; over 800 Danish seiners and nearly 200 pair trawlers operate on the continental shelf and slope around and near the Japanese islands; a few otter trawlers and about 670 pair trawlers fish in the East China Sea (including the Yellow Sea) ; a fleet of motherships and factoryships, with trawl catchers of various types, is sent to the Bering Sea and the northeast Pacific, and a large num- ber of independent trawlers to waters oflf Kam- chatka and the northern Kuriles; many large stern trawlers operate in the Bering Sea and the northeast Pacific, as well as in West African waters; an increasing number of Japanese trawlers are found in the Atlantic waters oflf North America; and a few vessels trawl in waters oflF New Zealand and South Arabia. A brief review of historical sequences of develop- ment of trawl fisheries in Japan will help one understand how such a complicated pattern has emerged in this particular sector of the industry. 229 FISHERY BULLETIN: VOL. 70. NO. 2 Different types of draggers for bottom fish have existed for several hundred years, and some of the primitive kinds can still be seen in in- shore waters. The enactment of a law providing various incentives for development of offshore fisheries in 1898 and the introduction of two European-type trawlers (one imported from England and one constructed in Japan) in 1908 marked the beginning of modern trawl fishing in Japan. The fishery expanded very rapidly, the number of otter trawlers reaching 136 by 1912. This resulted in serious conflicts with in- shore fishermen, prompting the government to issue trawl fishery regulations establishing large closed areas in coastal waters and to stop the ap- plication of subsidies to trawlers under the pro- motion law. These measures forced otter trawlers to move into the East China Sea, leading to the discovery of new abundant resources of groundfishes, par- ticularly highly valued porgies (sparids). The colonial administration of Korea immediately established large closed areas to shut out these trawlers from its coastal waters, thus pushing the fishery farther offshore. High costs of op- eration and overproduction, combined with a great demand for large vessels during World War I, resulted in a drastic reduction of trawl vessels, with only six remaining in 1918. The government, in 1917, established a new set of regulations and limited the total number of otter trawlers to 70, with a minimum size of 200 gross tons. This maximum number of 70 remained unchanged for many years for otter trawlers in the East China Sea (including the Yellow Sea). Among the primitive methods of catching groundfishes, winddriven dragging and hand- hauling bottom fishing were considered most efficient in early years. With the introduction of powered vessels, the latter method developed into one somewhat similar to Danish seining. This fishery expanded very rapidly beginning in the 1910's and has since been a major source of conflicts between fishermen in inshore and coastal waters. The number of powered drag- gers exceeded 2,000 in the 1920's and became subject to new regulations in an attempt to con- trol expansion and reduce conflicts with inshore fishei^men. Meanwhile, the method of trawling by two vessels was introduced in 1920 and this fishery, too, began to expand at a rapid rate. Pair trawlers immediately started fishing in the East China Sea; the government then took action to control pair trawling in waters west of long 130 °E under a separate set of regulations, the practice still in effect today. Although both pair trawling and Danish sein- ing became subject to ministerial regulations, the authority to issue licenses for these fisheries still lay in the prefectural governors. As a re- sult, the expansion of the pair-trawl fishery in the China Sea continued, the number of its ves- sels reaching nearly 700 plus some 400 operating from the Japanese fishing bases in mainland China and Taiwan. In 1933, the authority to issue licenses for both pair trawling and Danish seining was transferred to the central govern- ment. The government then instituted a long- term plan to reduce these vessels, particularly Danish seiners in waters east of long 130°E, which were causing serious overfishing and con- flicts with inshore fishermen. The plan was im- plemented for several years with some success, but with numerous problems arising from the increasing number of illegal vessels and viola- tions of closed areas. The power to issue li- censes (east of long 130°E) was transferred back to the provincial governments during World War II and remained there until 1947 when it was again given to the central government. A new cycle of various efforts to control the ex- pansion of Danish seining and pair trawling and reduce the numbers of these vessels began in 1950-51, when the nation was still under oc- cupation. Regulatory measures taken during the pre- war years to control the trawl fisheries of Japan established a pattern for allocating groundfish resources to different types of trawling: the stocks in coastal and nearby waters largely to the Danish-seine fishery (in the richest grounds off northern Honshu and Hokkaido) and partly to the pair-trawl fishery (in the western part of Japan) ; the stocks in the China Sea mainly to the pair-trawl and partly to the otter-trawl fishery; and the stocks in distant- water grounds to the otter-trawl fishery using large vessels. 230 KASAHARA: JAPANESE DISTANT-WATER FISHERIES Fishing in inshore waters was left largely to miscellaneous fisheries, including primitive drag- gers of various types. To a considerable degree, this pattern has persisted to the present, ex- cepting some major changes in the allocation of fishing grounds in the northern North Pacific in- cluding west Kamchatka, the Bering Sea, and the Gulf of Alaska, as will be mentioned later. The coastal trawl fishery, mainly by Danish seines and partly by pair trawls, still remains the most difficult one from the point of view of fishery administration. Due to the long-established vested interests of different groups of vessels operating from different bases, the allocation of fishing grounds is extremely complex, as illus- trated in Figure 1. In addition, there are closed areas around the home islands, some of which are rather extensive, different closed seasons applied in different areas, minimum depth limits in some places, as well as restrictions on the fishing bases each vessel can use for landing her catches. Post- World War II Developments East China Sea traivl fishery. — Most of the otter trawlers and many of the pair trawlers were sunk by American submarines during the war, in most cases while serving as military transport vessels, and only eight otter trawlers were left when the war was over. To meet the serious shortage of food after the war, the gov- ernment provided many incentives for recon- structing and expanding the fishing industry. The China Sea being the best trawl area in the nearby waters, the fisheries there recovered very quickly in spite of the so-called MacArthur Line limiting their fishing grounds to a narrow area of the continental shelf of the East China Sea. Numerous violations occurred and the area was later expanded slightly, but it was with the com- ing into force of the peace treaty in 1952 that the main fishing area became legally available to the Japanese trawl fishery. By that time, however, 58 otter trawlers and 783 pair trawlers had been licensed, with the total fishing power far exceeding that of the prewar years. The catch per unit of effort, which had shown a remarkable recovery during the war time, started to decline sharply. Fur- FiGURE 1. — Allocation of coastal trawl fishing grounds (from Norin Keizai Kenkyusho, 1965). Closed areas are not shown in the figure. thermore, due to international disputes with South Korea and mainland China, various re- strictions were imposed on fishing operations. The expansion of the fisheries of mainland China (estimated to take roughly 70% of the total groundfish catch from the East China Sea, in- cluding the Yellow Sea) also affected the Jap- anese catch. Increased fishing for China Sea shrimp {Penaeus orientalis) improved the sit- uation for a while, but the relative importance of the East China Sea grounds decreased rapidly as trawl fishing expanded into more distant areas, particularly the Bering Sea. Many of the otter-trawl licenses were used for obtaining new licenses for distant-water fishing by larger ves- sels, and there were only seven otter trawlers operating in the East China Sea by 1969. The number of pair trawlers also decreased, but to a much lesser extent. The use of pair trawlers as catchers of the Bering Sea mothership fish- ery also contributed toward reducing fishing pressure. Fishing in the South China Sea was also resumed in 1952 but ceased almost com- pletely as the main fishing area, the Gulf of Tonkin, became unaccessible due to the Vietnam War. 231 FISHERY BULLETIN: VOL. 70, NO. 2 It is important to know how the Japanese gov- ernment encouraged the license holders of the trawl fisheries to move into distant waters. A policy for the otter-trawl fishery was established as early as 1950; those wishing to use otter trawlers currently licensed for fishing in the China Seas, or those proposing to abolish licen- ses for China Sea fishing in return for trawling in distant waters, were given priorities. A new policy on the replacement of distant-water trawl licenses (issued in 1967) is summarized in Table 2 as an interesting example of license conversion. Coastal traivl fishery. — The trawl fisheries in coastal and inshore waters fall in two categories in the fishery administration of Japan: (1) in- shore fisheries conducted by various primitive types of draggers of less than 15 gross tons each and (2) those by Danish seiners of over 15 gross tons and pair trawlers, pair trawlers being much less important except in the western part of Japan. What was referred to as the coastal trawl fishery in this section is the latter.' By the time the authority to license the coastal trawl fishery in waters east of long 130 °E was again transferred to the central government, Japan was left with some 2,500 vessels plus a sub- stantial number of illegal boats, and the number further increased to a maximum of over 2,800 in 1951, when a new plan for reducing them was instituted. ' The Japanese word for this category literally means "offshore powered trawl fishery." Table 2. — Requirements for replacing a distant-water trawl vessel (in the North Pacific waters, including the Bering Sea, east of long 170°E, the Atlantic waters south of lat 40°N, and other distant areas) with a larger vessel. Gross tonnage of existing license less than 550 550-1,000 over 1,000 less than 550 less than 550 550-1,000 Gross tonnage of new license Licenses to be abolished up to 550 up to 1,000 over 1,000 less than 1,000 over 1,000 over 1,000 Source: Fishery Aoency of Japan (1970). None None None (a) One or more distant-water trawlers, or (b) One or more Danish seiners (or pair trawlers) east of long 130°E with minimum aggregate tonnage of 50 RGT; or (c) One or more pair trawlers west of long 130°E; or (d) One or more large or medium purse seiners with minimum ag- gregate tonnage of 100 RGT; or (e) One or more distant-water tuna longllners with nninlmum aggre- gate tonnage of 300 RGT; or (f) One tuna mothership with three or more deck-loaded catchers. (a) Same as above; or (b) Two or more Danish seiners (or pair trawlers) east of long 130°E with minimum oggregate tonnage of 100 RGT; or (c) Same as above; or (d) Two or more large or medium purse seiners with minimum aggre- gate tonnage of 150 RGT; or (e) One or more distant-water tuna longliners with minimum aggre- gate tonnage of 6(X) RGT; or (f) One tuna mothership with three or more deck-loaded catchers. (a) Soma as above; or (b) One or more Danish seiners (or pair trawlers) east of long 130° E with minimum aggregate tonnage of 50 RGT; or (c) Same as above; or (d) One or more large or medium purse seiners with minimum aggre- gate tonnage of 50 RGT; or (e) One or more distant-water tuna longliners with minimum aggre- gate tonnage of 300 RGT; or (f) One tuna mothership with one or more deck-loaded catchers. 232 KASAHARA: JAPANESE DISTANT-WATER FISHERIES Various measures were taken, including the combining of gross tonnages of smaller vessels to license a larger vessel, tighter control on il- legal trawlers, compensations for giving up trawl fishing, and preferential licensing for transfer to other fisheries which were still in the process of expansion. During 1953-54, 285 licenses were transferred to other fisheries with compensa- tions, a substantial number entering the tuna longline fishery. During 1954-56, when the salmon mothership fishery was still expanding rapidly, a large number of trawlers were con- verted into salmon catchers. Thus, a total of 910 licenses were taken out of the coast trawl fishery during 1953-56, with a total gross ton- nage of 225,500 tons' (Norin Keizai Kenkyusho, 1965). The most effective measure taken to reduce vessels operating in coastal waters, however, has been the expansion of trawl fishing grounds, which began in 1954. Danish trawling was ex- panded successfully into waters around the southern Kuriles, oflfshore banks in the Japan Sea, waters along Sakhalin and the Japan Sea coast of the Soviet Union, and, finally, waters around the northernmost part of the Kurile chain and both coasts of Kamchatka. Expansion into the northern Kuriles and Kamchatka waters marked a new era for Japanese land-based trawl fishing. By then, the Bering Sea trawl fishery, both by mothership fleets and large independent otter trawlers, was in full blast, and the mother- ship trawl fishery in waters off west Kamchatka had also started. A separate set of regulations, therefore, was established for fishing by trawl vessels licensed under the category of the coastal trawl fishery (see footnote 6) . Great operation- al difl^culties were encountered by the vessels en- gaged in fishing in these areas during the initial period, for they were largely from the existing fleet of coastal Danish seiners. Priorities for licensing were given to those having vested in- terests in waters around Hokkaido. Fishing area was originally defined as north of lat 48°N, ^ The following numbers of trawl licenses were trans- ferred to other fisheries, either converting vessels or giving up licenses in return for constructing new boats : 888 to the mothership salmon fishery as catchers, 102 to the tuna longline fishery, and 14 to the purse-seine fishery. east of long 148°E, and west of long 170°E, but was later expanded eastward to long 170 °W with the western boundary moved to long 153°E. The fishery has grown very rapidly since 1963, and the present fleet consists of nearly 200 ves- sels (now called "Hokutensen," meaning vessels transferred to the north), most of them newly built stern trawlers (the upper limit of their size is set at 350 gross tons). The total catch of the fleet is nearly comparable to that of the entire mothership trawl fishery in the Bering Sea. The main fishing grounds are still in west Kamchatka and the northern Kuriles, but the amount of fish taken from east Kamchatka and the Bering Sea is also considerable. Out of the total catch of 768,000 metric tons in 1969, 670,000 tons were Pacific pollack {Theragra chalco gramma) . A second government plan to further reduce trawl fishing in coastal waters (the third in the history of Japanese fishery administration) started in 1962, again through the transfer of licenses to other fisheries. By that time, how- ever, most of the other fisheries had reached or were reaching a point of saturation, and the ef- fects of this plan were not too great. Some 30 licenses were transferred to the tuna and skip- jack fisheries; a few licenses were issued for trawling in West Africa at the expense of those for coastal trawling. Some remarks may be appropriate for the handling of the inshore trawl fishery. Emphasis of the fishery administration was on reducing the number of vessels through compensations and subsidies. Over 30,000 vessels existed in 1950, of which only 7,000 carried licenses issued by prefectural governments, the remainder be- ing illegal vessels. The central government established policies and guidelines for the hand- ling of this fishery, which included the definition of inshore draggers (called small bottom drag- gers) as vessels of less than 15 gross tons each; the establishment of nationwide limits on the total number, the combined gross tonnage, and the aggregate horsepower; the establishment of a target for reduction, etc. During the period 1956-61, a total of 2,342 vessels were scrapped to be used for "tsukiiso" (objects sunk in shallow waters to attract fish), 2,379 diverted to other 233 FISHERY BULLETIN: VOL. 70, NO. 2 fisheries, and 75 converted to transport boats. But as of 1969, there still existed about 29,000 vessels, indicating that the reduction plan was not very successful. The fishery, however, is of relatively minor importance in the Japanese in- dustry, its total production in 1969 being only 262,000 metric tons, roughly half of which con- sisted of shellfishes. Mothership trmvl fisheries in the Bering Sea and adjacent areas. — In the Japanese regulatory system, a mothership means a vessel with pro- cessing facilities aboard which is accompanied by one or more fishing vessels. Most of the motherships do not fish themselves, but large fishing vessels, such as factory stern trawlers, or large tuna longliners, are also defined as moth- erships if they are used for processing catches delivered by smaller fishing vessels. The trawl fisheries in the Bering Sea, Kamchatka, the Aleu- tians, and the Gulf of Alaska consist of three licensing categories: "the Northern Seas Moth- ership Trawl Fishery"; "the Northern Seas Trawl Fishery" conducted by independent trawl- ers; and "Hokutensen," mentioned above. Catches by these three categories in 1969 were 862,000 metric tons, 373,000 tons, and 768,000 tons respectively, the combined total being 2,0 million metric tons. Trawl fishing in the Bering Sea was carried out even before and during World War II. As early as 1933, two fish meal factoryships with catchers were sent to Bristol Bay, The oper- ation stopped after 1937 due largely to unprof- itable fish meal trade. A freezer mothership operated in the Bering Sea in 1940 and 1941; a mothership operation for frozen and salted fish was conducted in waters off west Kamchatka during the war. The postwar mothership trawl fishery began in 1954 with two freezer mother- ships, accompanied by catcher boats, mostly otter trawlers, to produce frozen flounders, particu- larly yellowfin sole (Limanda aspera) in the Bristol Bay area. The number of freezer moth- erships increased to four in 1956, and a fish meal factoryship entered the fishery in 1958, as well as a mothership bottom-longline fleet. By 1961, the fishery expanded to include five fish meal factoryships (including one for processing Atheresthes for oil and meal) with 138 catchers, and 18 freezer motherships with 154 catchers. Three of the 18 motherships were specialized for shrimp fishing in an area near the Pribilof Islands, and eight (some of them were rather small) combined trawling, longlining (for hal- ibut and sablefish), and gillnetting (for her- ring). The trawl catchers were from those li- censed for fishing in the China Sea and coastal areas and included all three types: otter trawl- ers, Danish seiners, and pair trawlers. The Ber- ing Sea trawl fisheries started as summer oper- ations, but the season has since been extended, and some ships have been seen in the Bering Sea throughout the winter in the most recent years. For regulatory purposes, the Bering Sea was divided into several areas to allocate fishing grounds among different fleets taking into ac- count their historical interests. The next few years witnessed marked changes in the Bering Sea mothership trawl fishery. The yellowfin sole stock went down very quickly, as might have been expected for a hitherto unex- ploited flounder stock being subject to extremely intensive fishing, and also from past experience in waters along the Soviet coast (Kasahara, 1961) . Thus, the operation of fish meal factory- ships became unprofitable; this coincided with a slump in fish meal markets, both international and domestic. The number of factoryships man- ufacturing fish meal decreased from five (in- cluding one producing fish meal from Atheres- thes) in 1961-62 to only two in 1963 (including one making a substantial amount of fish meal from Atheresthes), with the catch of flounders decreasing from 467,000 metric tons to less than 100,000 tons, A sharp decrease in the halibut catch, combined with a decline in the sablefish catch, made longlining less profitable. A sud- den increase in herring production resulted in a market crisis. The shrimp fishery near the Pribilof Islands reached its peak in 1963 and declined rapidly thereafter. Meanwhile, empha- sis has shifted from yellowfin sole to pollack, which is perhaps the most abundant species of commercial fish in the region. The introduction of a mechanized process to make minced pollack meat further boosted fishing for this species, and the catch has shown a phenomenal increase 234 KASAHARA: JAPANESE DISTANT-WATER FISHERIES 7n t for the abstention pro- visions of the North Pacific fisheries treaty, the Japan-United States crab agreement, and reg- ulatory measures recommended by INPFC for Bering Sea halibut, Japanese fisheries in the eastern half of the Pacific Ocean have been rel- atively free of restrictions. The Japanese gov- ernment initially controlled the expansion of groundfish fishing into waters south of the Alas- kan Peninsula on the basis of unofficial discus- sions with the United States, but the situation changed in 1965 when the government issued regular licenses to a substantial number of ves- sels. Negotiations for resolving problems arising from the establishment of an exclusive fishery jurisdiction zone by the United States in 1966 (3-12 miles) began in January 1967, and an agreement came into effect in May 1967. While the United States took the position that an ex- clusive fishery zone could be established by a domestic law, Japan held the view that such a zone had no legal basis without an international agreement. As in the case of the crab dispute, the governments shelved their legal positions and worked out practical arrangements. The agreement covered a wide variety of fish- ing activities, both within and outside the fishery zone. For example, Japan was permitted to con- tinue the established fisheries within the fishery zone for crabs off the Pribilof Islands, groundfish along the Aleutians except during certain periods in certain areas, whales along the Aleutians and the Gulf of Alaska except between long 150°W and 163°W, salmon ofl^ the Aleutian Islands west of long 175°W, and tunas except in waters around the Hawaiian Islands and off the main- land coast. Certain areas within the zone were also designated for loading and support activi- ties. In turn, Japan undertook to refrain from fishing in certain areas of international waters during the first part of the halibut season and during the main crab pot fishing season. The agreement has since been revised twice, the most recent revision (effective January 1971) being summarized in Figures 12 and 13. (In compar- ison, the United States-Soviet agreement is sum- marized in Figures 14 and 15.) It is obvious that the United States has tried, with some success, to reduce the eflFects of for- eign fishing in international waters on impor- tant domestic fisheries in turn for allowing for- eign fishermen to continue their fishing in areas within the exclusive fishery zone where such fishing does not seriously aflfect domestic fish- eries. The United States has also avoided con- 256 KASAHARA: JAPANESE DISTANT-WATER FISHERIES 1 — r c « •H M u in = 3 13 u c (0 o 4) +J i : ■J N : O Ui 3 f1 D O ■; O M :! Z B ;; M Z ^ o :: S u ii S z □ H Eh S o 8 o a; Q a: z E-i a o ^ o: o N >. M 3 O 3 '■+J C o bo a "s c j« O T-l 01 ^ ■> dJ 0) •2 W S a> 0) Q CO to O) to J- *^ o to t« tn 0) »4 3 w S bo s CIS <3 a P 257 FISHERY BULLETIN; VOL. 70, NO. 2 44° JAPAN PERMITTED LOADING OPERATIONS WITHIN CONTIG- UOUS ZONE OFF DESTRUCTION ISLAND iCrays Harbor JAPAN WILL REFRAIN FROM TRAWLING AND LONGLINING IN THIS ZONE IN ,^:::::::::::;\s>wiiiapa Ha INTERNATIONAL WATERS LANDWARD OF THE ISOBATH OF 110 METERS (60 FATHOMS) j\ Asloria Tillamook Head / Warrenton Garibaldi Portland! Cape Lookout ^ Tillamook °^y f Willamette River Figure 13. — Arrangements under the Unit- ed States- Japan fisheries agreement (De- cember 1970), off the Pacific Northwest (taken from Commercial Fisheries Review, 1971a). 126° eluding a long-term agreement which might af- fect her position with respect to global negoti- ations now being held. in the form of an international treaty effective over a period of 5 years. This was a relatively minor dispute. Mexico. — When Mexico declared, in 1967, a 3-mile exclusive fishery zone beyond her 9-mile territorial sea, Japan entered into negotiations with Mexico to protect her vested interest in tuna longline fishing in waters between 9 and 12 miles. There was no agreement on the legality of the Mexican claim, but practical arrangements were worked out so that, except in areas particularly important to sport fishing, Japan was able to continue longline fishing in waters between 9 and 12 miles without exceeding the amount of effort in the previous years. The agreement was New Zealand. — New Zealand declared a 9-mile exclusive fishery zone beyond her 3-mile terri- torial sea in 1966. Japan entered into negotia- tions with New Zealand to protect her longline porgy ("tai") fishery within the newly claimed area. The agreement reached in 1967 allowed Japanese fishing to continue in waters between 6 and 12 miles without increasing the number of vessels, nor their size, until the end of 1970 (for 5 years after the establishment of the exclusive 258 KASAHARA: JAPANESE DISTANT-WATER FISHERIES r -■ 1 1 1 1 I 1 1 — T^^viiii^ — 1 1 ' ! 1 I 1 — r •* '. '-KJ^T •-.<-■ Q 2 ^^^■^ 'fc>-^>^ 2 O < U .-•.>i^rf^ T^^^rS^ ~ O 2 •••' -v^i^w^ £>i''^ _ 2 M •• ,iij2(rj\e>i/7<^^ rf MX ' ^i^^Lirx^L " IH y en M * ; jr- ' jil^ v*^^ H _ < MS .--.^ t'\Jr7vffir^'. '■ 2 b. K ^ ^^"^ / - t- o \ V •/ / £-■ K \ .-: y' / r-^ ^ Ml- \ K*-."V / r S Oi [J \_***"V^ 2 < 2 ...\ 7 w u o • • V/ / / f z < ' M M 2 ^ - CU > N ", ^-OC / / 1 X O en > tn y in I 1 ^ / II — (j^ >-A LT) q; 2 ME-. ^ MO X — < S g 8 "••• > M Q D ' f / / I ft. < O ft. to 2 >. rn a; s 2 « b: ft] 1 - ?; s '-' 'I Oft] a: ►J O w n .4/ IN FR N INT NES; JANU APRI 21, I BUT D X : _ •jf« -^ -5 M O P D >. ►J N « K rt < — a yy V ft. « X X a H Ln l/xis/ ft] < ft] E-. H S ft] tn D ft. -I _ H /<^:^i. A / / / ej ci] en 1/1 ct: o X / / ►J X -H -H X J ►J ft] E-. E-. en M ^J H Eh >. 3 M 2 to tn r-- < E 4 < /I " A. / 1 m M D D Q en o tj o x 1< - 'T^;"/!.'- ■' "^ j6 / / Eh S tn D O < in ^ 1 ■ i'^^-;."". .; /\Sr^ ** Wry > — ft] Oi < < S M M X ft] J^ Ci Q PriSH-^' **1-'^ fr'v» y?\ > Eh Eh • • • « < z ^ 1. **'^^5^'rt«l\ o M < .H (N (*i »r 3 M .. W < qH. t-^^CH Hrt^ ^y^ ^-— ^ en 3 3 X fc □ J 2 -v. ■ H t^ Q 2 a M < ■ "•.vWl Q M o 2 9 2 M J • .••■vV\ \ U s < J a tn • :• ■.<. /r fc ^^>. * .*" " vjC^jrn \ t> W J LO J oi M ■ • • ''TVA; ^Tx^i^«%r^ — ^ H to 2 en M tn w *. ■ -Tt*'.- / M 2 O M M fH iwi ■ ■ • ON M tn < / '•-'**v ^^< (K O .•i/.-*. 2 O < a < O tj •. .:J^ wpcntni2t,< ; .A • ."^^ •■flir*"' y C/l H a. o .~J w O M •■..- / - M H •. U 2 'H rs) m T in ^>f^ "P k-^ 6 o z o ^^^^■■•j ^ V* . .V ■ ." ^ M u ^-^'^ — ^ '•'•'* CSi^ . .* • ^■^^^^ ■ '. :.:'^N s. W / . A • -^-^ij"- ^ \^ / ■^ ■■^eC^^ » / . .-.••^ i/ /A ^i^ .^^ n \ 1 - il i \r\ - "^ ^id -^ wi;:W:':'x-:-:-:-x':':':': :•:•:•:•: ^^*M ^^JHS:?:^:?:?:?:::?:: " - ^^ / S-g-^IS^ - /^ ^^^^"^ * 1/ ■«• • - f / 6- « / ^ ^ Ul ^^ / ■;;i:i!::.:. " ' -* ..::.;: ^s. « "^oi r-^ / :::::::::::: ■ ::::::: \. Hw^v- ■' - - • ■■•wl'ilv.U qff\ 2 U Q / ^^^^?:-;, fi ■^-■■'^-' ■.■■.■■.■:■:■■.■:■:■ ^s,,^ M M Z / Bi' * 'J §g is ..--(^ / ^ Q O Q 2 W 4jABcV^' / -/ a U 2 3 1 " «>'**> / T"" 1 } Q Z en 3 \ / ft >l^^ U M to M [J H X M a \ / -A \\^ - MM !-; 3 Eh I \ / ^\^ _ MM M J X ftj \ o o 1 - 0- 2 •■ 2 «: • S ^ O 2 ;;::;;.:::::;:::;;;;, .,::.f::;;;;: O ft. D 2 H en M ft. 2 b en H H O 3 M _CJ 2 (N M 3 f- 2 O D \ M 2 u • • • 2 M H ft] J a W J M 3 O ►J iiiiiiiiipiiii ■N .-<■ ^> ftj 2 M (N rn X en :■■/ <<^ • O 0. O en ftj CQ ft] 0^ ;4^-' en O ESI -.1 M Z ft. o U CQ - i;i;;!i----^ll:l"" • -V N X ft. Q S t-\ •■.•\ o u a, en en H O X H ei) en 2 O m IJJ ■■A :::::::;::::: :■: •::::;:::::: < tS] a: H a - •••\ ft. ft] X a. z >. cN r- p.] i_i -^ ^1 .^1 O PL4 X E-. Q vD o; a: < XX in H H H en Pi en :;■;;;;;::;:;::■;;:;«::;:::;;;!: M M 2 8 n X - :::;:::::;::;;;7t:«:;::::::::: ::::::: i:;::::;::::::;*:;::::::::::;!:;-:"" a VvV WATER METE R FIR CH 17 CH 22 Q O Qi vD w cj Q X a -H t^ 2 H 2 Eh 2 3 n >^ M tH o ■-( o o; - t. M M M M en • • r > CO H Q a; < ^ c-j H 2 o o < 2 ft] ft] S M en S z < 0. en > * > 3 ■-< f^ f^ ■^ • 'C"' k d^ ." A V o o , , ,:•^- . - .1. -1 L .i m a J eeJ 01 3 c n u -H n a U U) 11 nj «i ft. S ^H s s o o o u 0* c3 p3 o u '0 Company, Ltd. In addition to fishing, fish processing, and mar- keting, engaged in a wide variety of other ac- tivities. Gross sales $472 million. Number of employees 10,890 including 6,880 aboard ships. Number of vessels 486, with a combined gross tonnage of 225,000. Major fishing acti- Whaling, the mothership vities salmon fishery, the mother- ship trawl fishery, trawl fisheries in distant waters (including the Bering Sea and northeast Pacific) and the East China Sea, over- seas shrimp ventures, and others. Nippon Suisan Company Ltd. In addition to fishing and fish processing, a sub- stantial interest in transport business, with four large vessels carrying oil and ore. Gross sales $195 million. Number of employees 7,959 including 3,960 aboard ships. Number of vessels 131, with a combined gross tonnage of 396,000. Major fishing acti- Whaling, trawl fisheries in vities distant waters (including the Bering Sea and north- east Pacific) and the East China Sea, the mothership trawl fishery, the mother- ship salmon fishery, the mothership crab fishery, overseas shrimp ventures, and others. Nichiro Gyogyo Company, Ltd. Before the war, the company monopolized Japanese salmon fishing from the Russian territory. Now engaged in diversified activities. Gross sales $155 million. Number of employees 5,800 including 3,190 aboard ships. Number of vessels 113, with a combined gross tonnage of 98,000. Major fishing acti- The mothership salmon fish- v^^i^s ery, the mothership crab fishery, the mothership trawl fishery, trawl fisher- ies in distant waters (in- cluding the Bering Sea and northeast Pacific), overseas shrimp ventures, tuna fish- eries, and others. Kyokuyo Hogei Company, Ltd. Started as a whaling company but has since diver- sified its activities. Gross sales $85 million. Number of employees 3,640 including 1,810 aboard ships. Number of vessels 48, with a combined gross ton- nage of 100,000. Major fishing acti- Whaling, the mothership vities salmon fishery, trawl fish- eries in distant waters (in- cluding the Bering Sea and northeast Pacific) , overseas shrimp ventures, the tuna longline fishery, crab fish- ing, and others. Among these four companies, they own prac- tically the entire whaling business excepting minor operations by two small companies, 8 out of 11 salmon motherships,-"" roughly two-thirds of the crab fisheries in the eastern Bering Sea, three-quarters of the mothership crab fishery in west Kamchatka, some 80 Sr of the Bering Sea mothership trawl fishery, most of the large stern trawlers operating in the North Pacific, West Africa, and other distant waters, as well as much of the shrimp ventures abroad. It means that the government can handle the industry aspects of most of the international problems concerning these fisheries by communicating with these and a few other companies (some of them subsidiar- ies of the big ones) . The government sometimes has forced them to conduct joint operations. Thus, one of the two mothership crab fleets fish- ing in the eastern Bering Sea is managed jointly by four companies, the other by five companies. Representatives of these companies, particularly the first three, often participate in international negotiations. The structure of fishery trade associations in Japan is rather complicated. The following is a brief description of some of the associations concerned with international aspects of Japa- nese fisheries: "" Salmon catcher boats are largely owned by small companies and individual vessel owners. 264 KASAHARA: JAPANESE DISTANT-WATER FISHERIES Japan Fishery Association (Daisui) generally represents the interest of larger fishing com- panies engaged in offshore and distant-water fisheries. It is often represented in important international fishery negotiations. It develops in- dustry policies on international fishery problems and also deals with specific disputes. It also ar- ranges for long-term, low-interest loans for the development of distant-water fisheries. The present chairman of the association is one of the most experienced Japanese in international fish- ery negotiations. National Federation of Fishery Co-operatives (Zengyoren) provides nationwide representa- tion for Japanese fishery co-operatives. Two of the main areas of activity are the procurement and distribution of duty-free diesel oil and fish marketing, but the federation is involved in in- ternational fishery negotiations from time to time. Federation of Japan Tuna Fishery Co-opera- tives (Nikkatsuren) is the most powerful asso- ciation for Japanese tuna fisheries, participated in by tuna and skipjack vessel owners through their local cooperatives,"" and is involved in most of the international negotiations concerning tuna fisheries. Together with Zengyoren, the feder- ation signed the Banda Sea agreement with In- donesia. The federation is making an effort to restrengthen the Japanese tuna longline fish- ery with substantial success. It buys in when the market is weak. It has been campaigning for increased domestic consumption of the tuna species that have been mainly exported, result- ing in an appreciable increase in the consumption of canned albacore tuna. It has promoted con- sultations with the tuna industries of South Korea and Taiwan. It plans to institute vol- untary restrictions, mainly closed seasons, on fishing for southern bluefin tuna (in effect as of October 1971). It compiles the most complete statistics of the longline fishery available in Japan. ^^ Tuna operators not eligible to cooperative member- ship under the Japanese fishery cooperative law, mainly companies operating large tuna boats, are organized under the Japan Tuna Fishery Association (Nikkatsu- kyokai). Nikkatsuren and Nikkatuskyokai always work together. Federation of Japan Salmon Fishery Co-op- eratives (Nikkeiren) represents salmon catcher boat owners and is mainly concerned with catch quotas for the mothership salmon fishery and ne- gotiations with mothership owners for profit sharing (formerly for selling prices). There are also associations representing such other salmon fisheries as the land-based drift-net fish- ery and the land-based longline fishery. National Federation of Medium Trawlers rep- resents bottom trawl fisheries in waters east of long 130°E, including the category called "Ho- kutensen" (see page 233). As the importance of "Hokutensen" increases, the association is now concerned about the condition of the pollack stocks in the northern areas (Kamchatka, North Kuriles, and Bering Sea) on which the entire fishery is based. Japan Trawler Fishery Association repre- sents trawl fisheries (largely by pair trawlers) in the China Sea. They have been concerned with problems with South Korea and the People's Republic of China. Japan Deep-sea Trawlers Association repre- sents companies operating large distant-water trawlers, and has been active in negotiations with Mauritania (Chairman of the association served as the Japanese chief delegate). It has made arrangements for exploratory trawl fishing for new grounds, and has engaged in planning the production of pollack minced meat ("sur- imi"). Except the first two mentioned in the above list, these associations represent the interests of specific fisheries and provide a convenient means of communication between the government and industry in connection with international nego- tiations involving such fisheries. IMPACT OF JAPANESE FISHING The expansion of the Japanese and Soviet fisheries has caused more international fishery problems than any other single factor. To be fair, some credit should be given to these two nations for their contribution towards develop- ment of new fishery resources all over the world. Japan and the Soviet Union, for example, have developed new resources in the Bering Sea and 265 FISHERY BULLETIN: VOL. 70. NO. 2 adjacent areas which now support an annual combined yield of perhaps 3 million metric tons and which would have remained unexploited or grossly underexploited without their effort. Japan has developed the tuna resources of the world ocean exploitable by longline. Again along with the Soviet Union, Japan initiated large-scale exploitation of groundfish resources along the west coast of Africa. The Soviet Union has been most active in developing new resources in the northwest Atlantic area. Japan has developed other resources, though not as great as those mentioned above, in various parts of the world. On the international scene, how^ever, Japan has seldom been given credit for her contribu- tion towards resource development, for the im- pact of Japanese distant-waters fisheries on the resources in international waters, some of which are also utilized by coastal states, was such that many nations look upon Japanese fishing, along with Soviet fishing, as one of the major factors responsible for the depletion of fishery resources on a global basis. There are many obvious cases in which Japan should be blamed for overexploi- tation of the resources that were either utilized by other states at the same time or were con- sidered important potential resources for them. Japan and the Soviet Union are largely respon- sible for the present state of the Antarctic whale stocks; Japan obviously overexploited many of the important stocks in the East China Sea; she overfished the yellowfin sole stock in the eastern Bering Sea, which was also an important re- source for the Soviet Union; the impact of off"- shore salmon fishing on the Soviet salmon stocks is apparent, although no critical assessment has been carried out; many of the crab stocks in the Bering Sea and Kamchatka have been overex- ploited to varying degrees; some of the stocks of porgies (sparids) in West Africa have been overfished by the trawl fisheries of Japan and some other nations. In other instances, Japanese fishing has not had any substantial eff"ect on the fisheries of the coastal states concerned, as is the case with the pollack fishery in the Bering Sea, much of the tuna and skipjack fishing, deepwater trawling, fishing for cuttlefish and octopus in northwest Africa, herring fishing in the eastern Bering Sea, squid fishing oflf New York, etc. But even in those cases, the w^ay new resources have been developed by the Japanese looks frightening to many other nations. A new resource may be exploited to a maximum level within several years, sometimes in 2 or 3 years. Emphasis shifts from one resource to another, or from area to area. The way Japanese trawl fisheries in the Bering Sea and adjacent areas are being expanded mainly based on one species, pollack, makes biologists wonder how long the resource can support the fisheries and what would hap- pen if the pollack stock collapsed suddenly. This new pattern of fishing, characterized by concentration of effort through large fleet op- erations and shift of emphasis from one resource to another, may not necessarily be a bad strategy from the point of view of maintaining the total production and the profitability of the industry. But it is not acceptable to many other nations be- cause it is contradictory to the established prin- ciples of management based on the concept of maximum sustainable yield and, more important, because such a pattern of fishing can be adopted only by nations having well-organized distant- water fisheries. If a nation is unable to partici- pate in the utilization of a resource for techno- logical or economic reasons, she would rather keep it undeveloped than see some other nation exploit it. There is little doubt that the development of Japanese and Soviet distant-w^ater fisheries has had very appreciable effects on the international fishery regimes. The impact of these fisheries, whether real or imaginary, has been one of the major factors motivating unilateral jurisdiction- al claims by coastal states. This applies, for ex- ample, to actions taken by the United States, Canada, South Korea, some of the Latin Amer- ican nations, many of the West African states, and even some of the Southeast Asian nations. Even the Soviet Union has taken unilateral ac- tions to protect its fisheries against Japanese high seas activities. In addition to these events, the expansion of Japanese and Soviet fisheries has been at least partially responsible for a worldwide trend for coastal states to justify var- ious forms of jurisdictional control as effective means to deal with international fishery prob- 266 KASAHARA: JAPANESE DISTANT-WATER FISHERIES lems. Whether or not the Law of the Sea Con- ference can produce a general agreement on this matter, some principle to the above effect is likely to emerge as a consensus of the majority. Exactly to what extent the development of dis- tant-water fisheries has contributed to this gen- eral trend is difficult to assess. It should also be pointed out that some of the European nations, particularly the east European, have followed the example of Soviet fishery development, though on a smaller scale, and have accelerated the trend for extension of coastal jurisdiction. FUTURE PROBLEMS CHANGES IN REGIMES FOR FISHERIES The purpose of this section is to make pre- dictions, based on past performance, on how the Japanese government and industry might re- spond to possible changes in international re- gimes for marine living resources. First, a brief analysis of the changes in inter- national regimes that are most likely to take place appears appropriate (Kasahara, in press). The first preparatory meeting of the Law of the Sea Conference (scheduled for 1973), held in March 1971, made it clear that fishery problems were among the most controversial issues con- cerning uses of the ocean. One of the reasons for this is the fact that fisheries are important to many of the developing countries, which com- prise the overwhelming majority of United Na- tions membership. Another factor, which may be more important, is the very nature of fishery problems. It is perhaps useful to note how well some of the major uses of the sea have been served by the existing regimes based largely on the traditional concept of free access. These in- clude transportation, which is the most impor- tant use of the sea, communication, scientific re- search, and recreation. Even the exploitation of mineral resources has not caused insolvable international conflicts. Although developing na- tions might look upon such freedoms as inequit- able because of their limited participation, little real damage has been done in those aspects of use of the sea. The major exceptions to this general notion are fishing and pollution. Except for pollution from sea accidents, most of marine pollution originates in areas within the limits of national jurisdiction rather than beyond. This leaves fishing as the most controversial issue. Free access to fishing on the high seas may have served for increasing food production from the sea, but it has resulted in numerous inter- national conflicts and necessitated almost con- tinuous negotiations between nations all over the world. Most of the actions taken to extend na- tional jurisdiction in one form or another have been motivated by a desire to control use of living resources. Fishery interests have also created such new concepts as an exclusive fishery zone, preferential rights of coastal states, as well as the allocation of resources in international waters. Judging from the nature of recent fishery con- flicts and discussions in the United Nations sys- tem, one of the predominant trends will obviously be further extension of coastal jurisdiction over the exploitation of living resources. Such a trend will continue regardless of the outcome of the Law of the Sea Conference. Extension of coastal jurisdiction might take the form of broader territorial zones, or preferential rights of coastal states. National claims might also be expanded through a new definition of living re- sources subject to the existing continental shelf convention and/or a new sea-bed treaty now under consideration. It is also possible that some nations might translate the new regime for sea- bed resources into a regime for the control of living resources in superjacent waters. There is no question that most of the devel- oped nations would prefer a relatively narrow territorial sea as a general rule from the point of view of minimizing potential hazards to im- portant nonextractive uses of the sea, particu- larly shipping and navigation. The probability of coastal states taking unilateral actions to re- strict the right of passage for nonmilitary pur- poses is rather remote, since practically all na- tions are beneficiaries of this right, and such actions would result in retaliatory measures of various kinds. Nevertheless, under certain cir- cumstances, some nations might possibly take such actions for economic gains. However small 267 FISHERY BULLETIN: VOL. 70, NO. Z the probability might be, the stake is big enough for a substantial number of nations to try to block a proposal for a territorial sea wider than 12 miles, or, failing this, to refuse to sign any treaty containing such a provision. Thus, chances are slim for an effective global treaty specifying a territorial sea broader than 12 miles to come out of the pi'oposed 1973 conference. This will not, of course, prevent some nations from extending their territorial seas through unilateral claims. If any effective global agree- ment on fishery matters should come out of this conference, however, it would be based on the principle of separating out the question of juris- diction over fisheries from the total package of national jurisdictions comprising sovereignty. The conference may not result in an overall agreement on fishery issues, but it is quite likely that there will be a general recognition of special rights of coastal states in terms of exclusive fish- ery jurisdiction or other forms of preferential allocation of resources. Such a principle will be supported not only by developing nations but also many of the developed nations. One way of protecting fishery interests of coastal states beyond the territorial sea would be the recognition of exclusive fishery jurisdic- tion within a certain zone, perhaps defined in terms of a fixed distance and/or a depth. It would be up to the particular coastal state wheth- er it chooses to allow foreign fishermen to fish within the zone under conditions set by the coast- al state. Some coastal states might prefer to allow foreign fishing for the resources that are not utilized or grossly underexploited by their own fishermen, probably charging foreign ves- sels a substantial fee. Arrangements might also be made for such resources to be developed from coastal bases as a condition for allowing for- eign fishing. Another way of protecting the interests of coastal states would be for coastal fisheries to be given preferential rights (including a right to adopt and implement conservation measures which would be binding on foreign vessels) to all resources within a certain zone beyond the territorial limit. This would involve problems of determining what portions of such resources or catches therefrom should be allocated to the coastal fisheries concerned, including the ques- tion of whether the coastal state should have a right to control the exploitation of the resources that are not used by them to any substantial degree. Under this principle, the formula to be adopted would perhaps vary from case to case. Preferential fishing rights might also be applied to specific resources important to the coastal states without establishing a fixed zone. This would involve such additional questions as the determination of major areas of distribution of the resources concerned, and the eff"ect of for- eign fishing for other resources on the particular resources in the same area. Among the three alternatives mentioned above, more nations might favor the first to en- sure a greater degree of control and simplicity of implementation. The main question in this case would be how the zone should be defined. Some of the nations supporting this idea may still be thinking in terms of a distance of 12 nau- tical miles from the shore for their exclusive fishery zone, with a narrower territorial sea. Some others are apparently considering varying distances to meet the specific situations. A sub- stantial number of nations seem to favor much greater distance, up to 200 miles, and/or to the outer edge of the continental shelf. A small number of nations might prefer pre- ferential fishing rights for specific resources that are important to their coastal fisheries. This would be a rather complex concept and a variety of problems would arise from its implementation. Many different formulae could be considered. The existing arrangement for yellowfin tuna in the eastern tropical Pacific may fall in this gen- eral category in that allowance is made, within the total catch limit, for vessels of smaller car- rying capacities. Various bilateral fishery agree- ments between the United States and nations op- erating distant-water fisheries off her coast also include provisions for reducing the adverse ef- fects, on coastal fisheries, of foreign fishing on the high seas. The treatment of anadromous fishes, particularly salmon, and marine mammals returning to land for breeding might also be con- sidered a special case in this general category. Different formulae are in practice to handle such a case. For North American salmon, the absten- 268 KASAHARA: JAPANESE DISTANT-WATER FISHERIES tion principle prohibits fishing by Japan in the eastern half of the North Pacific and Bering Sea. Catches of Asian salmon, on the other hand, have been shared by the Soviet Union and Japan. A system of product distribution has been ap- plied to the harvesting of North Pacific fur seals. Along with the general trend of extension of coastal jurisdiction, there will also be a contin- uing trend for more bilateral and multilateral fishery agreements between the nations directly concerned. International agreements solely for conservation, that is, for the purpose of maxi- mizing the total catch, have become less and less attractive to most nations, and emphasis has shifted to arrangements combining systems of allocation with conservation measures. The question of national quotas, particularly for the heavily exploited stocks, will undoubtedly become one of the most critical issues of fishery negoti- ations in the future. National quota systems are now being discussed even by some of the inter- national commissions which originally did not envisage them, as is the case with the Inter- American Tropical Tuna Commission or the In- ternational Commission for the Northwest At- lantic Fisheries. There is no established set of principles as to how the allowable total catch from a stock or stocks should be divided among the nations exploiting such a stock or stocks in waters beyond the limits of national jurisdiction, nor as to what allowance should be made for new entry. It is not very likely that the Law of the Sea Conference would come up with any specific formula to divide the limited catch. It is pos- sible, however, that discussions at the confer- ence might result in the general acceptance of the establishment of national quotas as a prin- ciple of international regulation of fisheries with- out spelling out details to implement it (such details would be left to bilateral or multilateral agreements between the countries concerned) . In any case, changes likely to take place in the regimes for regulation of fisheries, with a predominant trend for extension of national jur- isdiction by coastal states, may result in more in- ternational negotiations rather than less. In many parts of the world, such as Southeast Asia, the Gulf of Mexico and the Caribbean, the South Pacific Islands, West Africa, or even in much of Europe, the question of determining the boundaries between areas of national jurisdic- tion of neighboring states would become enor- mously complicated and, in some cases, might never be solved. Negotiations for the handling of historical rights of noncoastal states, as well as of neighboring coastal states, in the extended area of national jurisdiction of each state, would also take time. In many regions, regional ar- rangements of various kinds would have to be negotiated among neighboring coastal states to accommodate each other's fishing activity. With- out such arrangements, the development of the fisheries of coastal states would be hampered greatly, and the proper management of stocks of fish crossing several national boundaries would become impossible. In the present polit- ical environment, I doubt that the countries con- cerned could agree to a single regional conven- tion for each region. In most areas, a complex network of bilateral and semiregional agree- ments would develop. The enforcement of these arrangements would also be difficult and costly. POSSIBLE RESPONSE The question of how Japan might respond to likely changes in international regimes for fish- eries is, to a substantial degree, answered by what she has done in the past in response to various claims by other nations (see section on International Arrangements) . If the Law of the Sea Conference results in a global conven- tion providing for extensive coastal jurisdiction or broad preferential rights of coastal states, it is unlikely that Japan will be a party to such a convention. She would then regard actions taken by member states of the convention as uni- lateral. In the past, Japan has responded to unilateral actions in a variety of ways. When she did not have much vested interest in the zone claimed and the nation claiming the zone was not prepared for negotiating the issue, Japan voluntarily refrained from fishing in the zone while refusing to recognize the claim. When her vested interest was very substantial, Japan en- tered into negotiations with the country con- cerned. In some cases, such as the Japan-South 269 FISHERY BULLETIN: VOL. 70, NO. 2 Korea controversy, the Japanese government did not stop fishing vessels of its nationals from entering the claimed zone, resulting in the sei- zure of many vessels. In most cases, however, practical arrangements of various kinds were agreed upon, sometimes after long negotiations, as described before, Japan has not challenged any fishery claim by force, and, except for the Japan-South Korea and Japan-Soviet contro- versies, no real diplomatic crisis has developed from fishery issues. The future trend in this respect will be about the same. Japan would do her best to protect her fishery interest against unilateral claims with whatever trade offs available to her, both within and outside the purview of fisheries, but would still seek a pragmatic solution to settle the issue. If Japan has no vested interest in the area claimed, she might voluntarily refrain from entering the zone for fishing while officially refusing to recognize the claim. The same would perhaps apply to Japan's reaction to claims based on the concept of preferential rights of coastal states. Such concepts as the allocation of resources, the division of catches therefrom, or the distri- bution of benefits, have already been applied ex- tensively to fishery arrangements involving Ja- pan. Although Japan would not recognize these as internationally accepted legal concepts, she would not object to practical arrangements which would have the same eflFects. The appli- cation of a limited entry system has never been a problem to Japan because of the very nature of her domestic regulations, as outlined in the introduction of this paper. In most of the bi- lateral agreements she has made so far, the num- ber (and in many cases the size as well) of the vessels to operate in a designated area is limited. Any substantial change in the definition of shelf resources to include more living resources currently exploited would not be recognized by Japan officially. The main reason for Japan not to sign the 1958 continental shelf convention was the inclusion of living resources. The pattern of bilateral negotiations for problems that might arise from this source would be about the same as that for problems from extended fishery jur- isdiction. She would do her best to protect the vested interest of her fishing industry. The possibility of general recognition of a special right to anadromous species, particularly salmon, would be a matter of great concern to Japan, as high seas salmon fishing is still one of the most important sectors of the Japanese fishing industry. During the Law of the Sea Conference, the establishment of a special right to anadromous species may be proposed by some nations as one of the principles of international regulation of fisheries. This might receive rath- er broad support, not only because of problems of Pacific salmon but also in view of recent de- velopments in offshore salmon fishing in the At- lantic. Again, Japan would not join a conven- tion including such a provision. But if the United States, Canada, and the Soviet Union should claim, on the basis of such a convention, a special right to anadromous species for the main purpose of eventually eliminating high seas salmon fishing, Japan would be in a difficult po- sition to protect her interest in salmon fish- ing. The idea of establishing a world agency for regulating all high seas fisheries has been talked about by idealistic people, but by now it is widely recognized that this is not feasible, nor even de- sirable. We can pretty well eliminate this pos- sibility from our consideration of fishery prob- lems in the foreseeable future. In short, it is unlikely that Japan could take any definite course of action to cope with an in- creasing number of international problems she is going to face. She must be prepared for more and harder negotiations to find a practical so- lution to each of the problems. In the North Pacific, Japan will have to keep negotiating with the Soviet Union for salmon, crabs, and herring, and probably for some of the groundfishes in the future. Negotiations have become increas- ingly difficult as additional regulatory measures have been proposed by the Soviet Union every year. As the U.S. king crab fishery in the Ber- ing Sea is expanding with the Japanese quota being reduced, the future of the Bering Sea king crab fishery also looks bad. Continuous pres- sure will come from the United States and Can- ada to provide their coastal fisheries with a 270 KASAHARA: JAPANESE DISTANT- WATER FISHERIES greater amount of protection against Japanese fishing for groundfishes and shrimp. Pollack, the main species for the trawl fisheries in the Bering Sea and Kamchatka waters, might be- come a serious international problem in the near future. Japanese fishing pressure is still mount- ing ; the Soviet catch is increasing; South Korea is building a number of stern trawlers in Japan with a view to rapidly increasing her participa- tion in pollack fishing. The Japanese trawl fish- ery along the west coast of Africa will face fur- ther international problems as more African nations take measures to extend fishery jurisdic- tion. Most of the bilateral agreements Japan has concluded in recent years are of short dur- ation, and it may be difficult to continue these on the same terms. The Japanese tuna industry might still be able to compete with the Taiwanese and Korean fisheries by taking advantage of rapidly expand- ing domestic markets, but a substantial increase in the catch of the longline fishery is not likely. Major eflforts are being made to develop a purse- seine fishery similar to that of the United States and to increase the production of skipjack, which is at present an underexploited resource; but international regulations will gradually be ap- plied to many of the tuna fisheries. In the east- ern tropical Pacific, the present pattern of tuna fishing is hkely to lead to a system of national quotas. Tuna fisheries in the Atlantic will also be subject to some international regulatory sys- tem in the future. Eventually there might be a regime of worldwide regulation covering all ma- jor tuna fisheries. Trawl fishing in the North- west Atlantic will also be subject to further re- strictions through bilateral and multilateral arrangements. Whaling both in the Antarctic and the North Pacific will have to be further curtailed. International fishery problems faced by the Soviet Union are not too different from those confronting Japan, except in the Northwest Pa- cific where the Soviet Union finds herself in the position of a coastal nation seeking protection against Japanese fisheries. It is interesting to note that their responses to unilateral claims have not been too difl^erent from those of Jap- an. FUTURE OF THE INDUSTRY The phenomenal growth of the Japanese econ- omy has greatly increased demand for high- quality foods, particularly animal protein pro- ducts. The per capita consumption of animal protein increased by 19 ^r in the 5-fiscal year pe- riod of 1963-68. About 58% of the animal pro- tein intake is still from seafoods, including whale meat. During the same period, the per capita expenditure for fishery products increased by 10% per annum in cities and 13.2% per annum in rural areas (Anonymous, 1969). Markets have also developed for a greater variety of fish- ery products. Imbalance between demand and supply has been increasing constantly, pushing up prices sharply. Pressure for increased fish supply is still quite strong in Japan. Domestic Production What alternatives are available for Japan to meet this problem? First, let us examine the possibilities of increasing the domestic supply of fish. Figure 16 indicates the trend for pro- duction by four sectors of the Japanese marine fishing industry. Divisions between the sectors, except aquaculture, are somewhat arbitrary. 4n 48 1954 1955 1956 1957 1958 1959 1960 4 145 413 145 41 202 111 1961 262 2 I 1 1962 387 4 2 2 1963 660 4 2 1 1 1964 526 12(13) 6(7) 2 3 1 1965 385 6(8) 3(4) 2(3) 1 1966 1967 1968 1969 Unknown 508 521 883 1,395 2 Total 6,590 Total, 1961-^5 only 9(11) 3(4) 9 5(6) 2 58(65) 4(6) 3 1 (1) 6 3 5(6) 28(33) 13(15) 11(13) 7(3) ' Number recaptured ore for groups A-C (see Appendix); numbers for groups A-D, If different, are shown in parentheses. 0-12 12-24 24-36 36-48 MONTHS AT LIBERTY Figure 5. — Number of returns of white marlin tagged from 1961 to 1965 in waters north of lat 32 °N, plotted by time at liberty. BLUE MARLIN Since 1954, 486 blue marlin have been tagged in the western North Atlantic; 3 have been re- covered, all near their respective release points. One blue marlin released oflF Chub Cay, Bahamas, in August 1968, was recaptured oflF nearby An- dros Island the following December; one re- leased off La Guaira, Venezuela, in August 1966, was recaptured in the same area in October 1968; and one released off Biloxi, Miss., in June 1969, was recaptured 5 months later off Sabine Pass, La., 350 miles to the west. The returns indicate that meaningful information about blue marlin can be obtained if sufficient numbers are tagged. LONGLINE CATCHES METHODS Data on the catch of marlins by the Japanese longline fishery in the Atlantic Ocean have been published by Shiohama, Myojin, and Sakamoto (1965) and by the Fisheries Agency of Japan (1966, 1967a, b, 1968, 1969). Catches of white and blue marlins, and those of other billfishes and tunas, are listed in those publications by 5°- quadrangles for each month, beginning in June 1956. Although catches in the categories "black marlin" and "striped marlin" also are listed, systematists do not presently recognize that black marlin, Makaira indica,^ and striped mar- lin, Tetrapturus audax, occur in the Atlantic Ocean. We do not know whether those catches were misidentified or whether these species do, in fact, occur in the Atlantic. The catches listed in the two categories were too few to affect the conclusions of our study even if they were really white or blue marlin. For each species, the catch per unit of effort (CUE), in fish per 100 hooks, was calculated for each 5°-quadrangle-month stratum in the period 1956-67 for which data were available. These CUE's were computed by dividing the number of fish caught in each such stratum by 1 % of the number of hooks fished in it. To show seasonal distribution, average (unweighted) ^ Ueyanagi et al. (1970) report occasional catches of M. indica in the equatorial and southeastern Atlantic. 290 F MATHER, JONES and BEARDSLEY: MARLINS IN ATLANTIC CUE'S for each 5°-quadrang-le-month stratum were obtained by summing its yearly CUE's and dividing by the number of years. The Japanese longline fishery in the Atlantic Ocean is directed primarily toward catching yel- lowfin tuna, Thunnus albacares, and albacore, T. alalunga, (Wise and Le Guen, 1969). Since marlins form only a small part (<3%) of the total catch of scombroid fishes, fishermen prob- ably do not select specific fishing areas for mar- lins or adapt their fishing gear to catch marlins more eflfectively than other species. Possibly the catch rates for marlins are influenced by variations in the availability or the catchability of the fish, but the effects of such variations can- not be- distinguished on the basis of the available data. We believe, therefore, that the catch rates by the longline fishery represent reasonably well the relative apparent abundance of marlins in the areas and seasons of intensive fishing. On this basis, we discuss distribution of white mar- lin and blue marlin in the Atlantic Ocean in the next sections. WHITE MARLIN The catches of the wide-ranging Japanese longline fishery show that white marlin are dis- tributed over a much broader area than that in- dicated by returns from fish tagged in the sport fishery in the western North Atlantic Ocean. White marlin have been caught in all consistently fished areas of the Atlantic Ocean from lat 40 °N to lat 40 °S (Figure 6). The available data in- dicate that white marlin are scarce in both the north and south temperate zones in their respec- tive winters (December-February and June- August) . Catch rates above 0.5 fish per 100 hooks are reported more often in the western than in the eastern Atlantic Ocean. We therefore conclude that although the distribution of white marlin is oceanwide, the species is more abundant in the western Atlantic. Sport fishermen report that white marlin are often concentrated at the edge of the continental shelf. Data from the longline fishery support this conclusion, inas- much as the catch rates are generally higher in 5 "-quadrangles adjacent to, or including, the edge of the shelf than in quadrangles in mid- ocean. No evidence has been found to date to suggest that the relative apparent abundance of white marlin in the Atlantic Ocean has been markedly affected by the longline fishery (Wise and Le Guen, 1969). North Atlantic In winter (December-February) white marlin are concentrated in the eastern Caribbean Sea and off the north coast of South America as far south as the equator. In spring some of these fish move northward into the Antilles Current beginning in March, and others move westward into the western Caribbean beginning in May and June. The northward movement of the first group conforms to the migrational pattern de- duced previously from tag returns, but the sec- ond group appears to move into the northern and western Gulf of Mexico in summer, a pattern not supported by tagging data. High catch rates in October, 300 to 1,200 miles oflF the east coast of the United States, support the hypothesis derived from tag returns — that white marlin summering off the U.S. mid- Atlantic coast move offshore in fall. In the Gulf of Mexico, however, white marlin are relatively abundant through October. South Atlantic In summer (December-February) white mar- lin are concentrated in the central South Atlantic Ocean and off the coast of Brazil. In the latter area, catch rates are among the highest known, occasionally reaching 3.8 white marlin per 100 hooks. In autumn (March-May) catch rates are generally lower than in summer and the marlin are not caught as far south. Because large areas of the South Atlantic are not fished in the au- tumn, catch rate data may not reflect a true pic- ture of distribution. From June through Sep- tember white marlin in the South Atlantic are concentrated in the South Equatorial Current and off the southwest coast of Africa. This sea- son is the only one in which white marlin ap- parently are concentrated along the coast of Africa. From September through November, 291 FISHERY BULLETIN: VOL. 70, NO. 2 "/ ^/JilJI \ WE \"'" \ I ^"^^^r*^ / ^*/*l' *l'l III* 1 1 PaliA^' »v^ " EW / 1 ' \ / ^^.S'^— \ 1/ /y^l^S^/. /./././. . . . ~7 /il^^^mJWpkfLj It f j ^T\\*i'jjJjJ*jj 'I It i n"^^ "^^^t o . o o o 5" \j *v AUG/ y WWW J 1 0 . . . . TH \\W\\\ fW o o • o •/o/o^ .\\\\\a A\° • — rhk \\\\\^ -iAW ° o • •/•/o/o/o/L-r^ \\\\\\ yA\\\ 1 1 m \\\\\\ A\\\\ //////// xwwSWWW , ////////.., / ^/7 III \ \ \ ^ \ / ^7./....'ooo\ r 1 yri^Ss -^X 1 r«ii«*-^' wV*\ / W^'r/'r'rhi \ / ^v-j'^-A "//^^i^^^*/'r ' 'I'r ' 1/ ' /n^r^^'l'rrrr • " • ° Li I ^'^Mf^*r\r r r * ° • • "^ / / 7m°MW-^*/ ' / dL .^^^^^ • is...pp^ J *^Qc • • o . . 0 . ^ ° • • •ft \\\\>v JUN/l- °o o mii WWW 1 o o o ■ olc WW\\\ rd °° /o/c ,^\\\\ \\ PA / /■/■A \\\\\\ J\°\°\\ / / /°A--> \\\\\\ /\\\ I ' ff^ \\\\\\ {\\\\\\\ //////// \\\\\\\\\\\\\\ \ L. Figure 6. — Distribution and apparent relative abundance of white marlin in the Atlantic Ocean. Data are from records of the Japanese longline fishery, 1956-67. The catch per unit of effort (CUE) for each month in the rec- tangle is the arithmetic mean of the CUE of each month that the area in the rectangle was fished in the 12-year period. CUE is the number of fish caught per 100 hooks. 292 MATHER, JONES and BEARDSLEY: MARLINS IN ATLANTIC "/ *7-n UK \ / Jf 1 1 I 1° 1 1 I ^../'rnH-H o\ \i Ms.^ ^ V/tTW ■• hoo °\f /I/lhJI-n- ff^UTfj^l / "^r"^^"^^ ^ , . . . . * ^^Mf-I^Tl w ^« \ \ 1 ^"^Jc • ' \ \ \ \\. - - . _ JLL • _ VVrrK MARfF* • y \ \ \ \ \ v\ /• !• * < >i la \ \ \ \ \ W rTi I * /■i \ \\ \\ \k Mil ; -K '" ' \ \ \\\ » «A*\ \ I /A^ \\\\\\ ^\\V- , ////tT; \\\\\\ A\\\\\ 1 //////// \\\N\M\\\\\ L / ^/Jlljj \ \\JS ^ \ / yj rrrrrl'V ° / /t/ / 1*1 ! 1 rl • "IfT'T'Ti / / / 1 1 1 1 ' • W /m.J*0^IpI' •I'l /• o • 0 . L ^^Wm^rhl A A ? TSIj ^4^-^*/ * / i=T Si^:--^ i5 m^oc^M - IB • . . . o o/ol \ \\\\ \y ru\.u • • • o o/o/o/ol \ \ \\\ ^ ^\ \o\"l* o o o \\°\ \ l\-r* \ \ \\ \ \ /\\ \ \ \ /o///o77; \\\\\^ ay\°\\\ , //////// WWIII/'j. / ^/Jiijj 1 M kT ^ 1 1 't /^^^ _• ^*^^^ • ^,/ \i 80 ^X. /T 70 - • 1 1 \ f 60 c • \ S 50 \ 1 40 ■y \j ^ 5 ! 30 •-V ^ 20 3: 10 -L 1 1 1 1 r OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY Figure 3. — The change in percentage in the average catch rate in the Boothbay area as compared with the average catch rate in the coastal area. larvae were obtained in the Boothbay area that were larger than those recently hatched in the western coastal sector. The movement of larvae in the spring was marked by their accumulations in locations receiving shoreward intrusions of coastal water and by the relation of their dis- tribution with that of surface salinity. The in- trusions resulted from eddies formed by the spa- cing of river discharge along the western sector of the coast (feathered arrows, Figure 4) . The configurations of isohalines and lines of equal larval density throughout the coast were often similar. Such similarity indicated that the mag- nitude of station-to-station differences in catch paralleled the station-to-station differences in salinity. These differences were accumulated for coastal stations from west to east and their fre- quency distributions were plotted on probability paper for two cruises having widely different salinity distributions (Figure 5). The cumula- tive frequencies of salinity and catch were alike for a given cruise, but differed between cruises. We infer from these results that the larvae are carried by currents, indicated by the station-to- station differences in salinity. The precise di- rection of drift cannot be ascertained because it is not possible to determine whether the salinity distribution is causing the current or is the effect of the current, perhaps generated by winds. 3U FISHERY BULLETIN: VOL. 70, NO. 2 '^^. \,^- Cape Ann 0.5 Figure 4. — (a) Isohalines (%c) during a spring cruise, March 28-April 13, 1967. Feath- ered arrows indicate shoreward intrusions of coastal water (from Graham, 1970b), and unfeathered arrows indicate current directions along the coast (from Bumpus and Lauzier, 1965). (b) Isolines of larval catch rates (no/100 m^) for the above cruise. catches made at the three outer estuarine, three lower estuarine, and two upper estuarine sta- tions, and plotting them in a time series. Peaks in larval abundance progressed from outer to upper estuarine stations with time, suggesting an inshore movement of the larvae (A, B, and C in Figure 6). Two such progressions were obtained in 1966 and one in both 1964 and 1965. In 1964, larvae passed the outer and lower sta- tions between our scheduled cruises and ap- peared first at the upper stations as a peak in abundance in early October. In spring, larvae were always more abundant at the upper estu- arine stations, and progressions in peaks of abundance up the estuary were not apparent. LARVAL LENGTHS Boy FEB.2I - MAR. 4 y^-"-'""" Pleotont Boy 1966 ( f^^^^'^ P«noDlCOl Boy ^ ,.-•• ' :' /C MAY 7-15 .'■-■'^V^ 1968 Cope Smoll Cqp« Porpoix j^/^ ^ SALINITY LARVAE / ? Cope Ann 5 10 20 30 40 50 60 70 80 90 95 98 99 STATION DIFFERENCES (Percent Accumulated ) 99.8 99.9 Figure 5. — Accumulated frequency distributions of sta- tion-to-station differences from west to east along the coast. The cruise in 1968 began to the east of that in 1966. Larval movements up the estuaries during autumn were easier to detect in the Boothbay area because larvae were transported inshore primarily by tidal flows (Graham and Davis, 1971 ) . Tidal flows followed the inshore-offshore axes of the estuaries and embayments; thus the landward movement was detected by grouping The rate of increase in average length varied seasonally and geographically. Data from the coastal and Boothbay areas showed a marked 312 GRAHAM, CHENOWETH, and DAVIS: LARVAL HERRING 12 10- I eh o z 4 O 2 - SEP. OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY 173 12 10 lO o z < 1965 -66 N=2462 SEP. OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY 12 r 10 I- 2 • Ot A> 218 B • 1966 -67 N- 1539 SEP. OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY FiGxmE 6. — Seasonal changes in larval abundance in the Boothbay area during 1964-67. Values are plotted for outer (A), lower (B), and upper estuarine (C) stations. increase in length in the autumn followed by a more gradual increase in the winter and spring (Figure 7) . During the autumn and spring the increase in length accompanied a comparable increase in weight, but not in the winter when the larvae were very thin (Chenoweth, 1970). In the coastal area the increase in length 35 p 30- 251- 20 15 10 2 X z < > < < liJ 5- 0 COASTAL AREA 1962 - 67 _L J I I L J I SER OCT. NOV. DEC. JAN. FEB. MAR. A PR. MAY 40r 35 1- 30 25 20 15 10 5 1964- 65 v"'-. l965-66-7<:-''V ^1966-67 BOOTHBAY AREA I ■ I I L I I I SER OCT NOV. DEC. JAN. FEB.MAR.APR.MAY Figure 7. — Seasonal change of mean larval length in the Boothbay and coastal areas, 1962-67. 313 FISHERY BULLETIN: VOL. 70. NO. 2 appeared asymptotic while the average length fluctuated above and below 32 mm in the Booth- bay area from March through May in 1965 and 1966. Differences in the average length also oc- curred between the eastern and western sectors of the coast. In late September to mid-October, the mean lengths did not differ greatly along the coast (Figure 8) . By early November the mean length was larger in the eastern sector than in the western sector, but this difference diminished in winter; by spring mean lengths were similar along the coast. z z z UJ < > < Z < UJ r Sep. 27- Oct. 14,1962 N = 393 20 - 10 • t • J 1 • 1 1 rocf. 18-27, 1966 N = 2 ,232 ?0 - • .••• • • •• 99 10 1 •.V. 1 1 1 rOcl. 30-Nov 6 1962 N=759 20 - • .. • ••. 10 • 1 1 1 1 •. 20 10 .•' •Nov. 1-17, 1966, N=I,0I2 40 r Mor. 16-23, 1966 N=l,844 30 40 r Mar. 28-Apr. 13, 1967 . N = 87l ll I I L_ L I I 1_ WEST c.«>V c.<>V EAST WEST EAST *» V /'«<>■' *«■%' C.«>V'^ r&V ^ ,<-> ^%-^ ^^cf .o-oA ^oV STATIONS Figure 8. — Variations in the average length of larval herring from west to east along western coast of the Gulf of Maine. The seasonal change in mean length (Figure 7) was influenced by the movements of larvae, their avoidance of our sampling gear, and their departure from the vicinity of our sampling sta- tions after metamorphosis. Broods of larval herring moved shoreward in the autumn from the coastal water, but the size at which they in- itially entered the Boothbay area varied. This is illustrated by data from 1964 and 1965 ; data from 1966 were similar to those of 1965. In autumn 1964, two broods of larvae (labeled a and b in Figure 9) initially entered the Boothbay area; the more abundant had a size mode of 13 mm, the other was recently hatched and had a size mode of 9 mm. By mid-October these two broods were of equal abundance. In November a third brood (c) was detected, during a coastal cruise, that had a modal length of 13 mm, equal to the modal length of larvae initially entering the Boothbay area. Presumably, the addition of this group in part slowed the shift in modal length to only 3 mm from mid-November to late December. These variations in size resulted from the location of the Boothbay area within an east-west coastal zone of transition for hatch- ing time. Also, some of the larvae hatched in the eastern sector of the coast at an earlier time and were carried westward along the prevailing currents (Graham, 1970b) and then into the Boothbay area. In autumn 1965, a single brood (d) of re- cently hatched larvae with a modal length of 7 mm initially entered the area. In October, this brood was still the most important contrib- utor to the inshore area, since the modal size from a coastal cruise in October coincided with the seasonal progression of larval size within the Boothbay area. A second group (e) of re- cently hatched larvae entered the area in early November. In the spring it was obvious that we were failing to catch the larger larvae and the spring peak in catch rates should have been consider- ably larger than that recorded for daytime tows. We assumed that this failure was due to their avoidance of our gear because we captured more of these larger larvae at night when they could not see the gear. Day and night cruises were alternated in the Boothbay area from January through April during 1965 and 1966. The size ranges of larvae captured during these cruises were the same, but the modal length was much larger for larvae captured during the night cruises. The length-frequency curves for all larvae from the day and night cruises showed two distinct modes, one at 33 mm and the other at 40 mm (Figure 10). At 36 mm the percent- ages of larvae in the day and night catches were identical. For larvae larger than 36 mm the percentages increased for night catches and decreased for day catches, indicating that avoid- ance increased progressively until the larvae grew to 40 mm. Initial nighttime catches of 314 GRAHAM, CHENOWETH, and DAVIS: LARVAL HERRING D MODE AUTUMN DEC. 7-11 N = 266 I BOOTH BAY □ COASTAL 1964-1965 WINTER SPRING FEB. 2- MAR. 5 N = I46 ~r~i 1 1 — r -T r 20rOCT. 27- N = II3 u. ,0^0^- N=476 . DEC. 23-30 N:I84 -t 1 1 r MAR. 23-24 N:|,24l NIGHT MAR. 30-APR. 6 N = 792 JAN. 5-27 N = I30 APR. 13-20 L N=370 JAN. 31- FEB. 7 N = 720 APR. 27-28 . N = 2,694 NIGHT 5 20 25 30 35 40 45 50 LENGTH(MM) 5 10 15 20 25 3b LENGTH (MM) J AUTUMN Q60 30r il SEP. 27-29 N = 240 1965 - 1966 WINTER UA. DEC. 6-8 N=II9 MAR. 17-23 N=l,844 DEC. 22-29 N = 99 -I 1 r NIGHT 'f'^ — T r MAR. 21-23 N=l,260 JAN. 4-20 p N = 85 T 1 r MAR. 31- APR. 4 N = 55l FEB. 15-MAR. 4 N=l,795 APR. 18-20 N=I74 NOV. 8-10 ■ 1 = 267 . FEB. 23- MAR. 17 N= 149 -| 1 r 10 15 20 25 30 35 40 45 LENGTH (MM) APR. 27-28 N^BI LENGTH (MM) 20 25 30 35 40 45 50 LENGTH (MM) Figure 9. — Size distribution of larval herring during 1964-66. Only the most obvious modal lengths are indicated for the data from the Boothbay and coastal areas and those modes discussed in the text are labeled a to e. 315 FISHERY BULLETIN: VOL. 70, NO. 2 10 8 UJ o < 6 - NIGHT N = 6,650 \ DAY J N = 2.874 UJ 1 • 1 • 1 *** u ~ 1 • f • f • \ : oc \ : UJ 4 / \ : a. f 1 / / \i f • 1 • f • \\ 2 f • f • / • \\ r" — ' Vv 0 r-^ " 4/....-- 1 1 \ ■■-.i 20 30 40 LENGTH (MM) 50 Figure 10. — Length-frequency curves of larvae captured in the Boothbay area at night and during the day. those larvae larger than 36 mm occurred earlier in the year and their peak in abundance occurred later than that of larvae captured during the day ( Figure 11). In 1964, losses by avoidance began in December. Failure to catch larvae began later the following year (in January) because the larvae that initially entered the Boothbay area were smaller than those that entered in 1964 (Figure 7) . Catches of larvae larger than 36 mm in length in the Boothbay area were si- milar to those obtained with buoyed and an- chored nets set in the Sheepscot estuary at night during 1966. This similarity indicated the ease with which larvae were captured at night. The buoyed nets strain water at a much lower velo- city than the trawl. In late April larvae were occasionally observed schooled in shallow coves. By late May large numbers of these fish were frequently observed in the shallow coves whereas the numbers cap- tured at our sampling stations had declined. This change in distribution was coincident with the period of metamorphosis (about 41-50 mm SL) from larval to juvenile form. Metamor- phosed herring were found in the Boothbay and other inshore areas but not in the outer coastal waters during the summer (Davis and Graham, 1970). Variations in the lengths of larvae in our samples related to their shoreward movements, avoidance of gear, and departure from our sam- pling stations, caused discrepancies when esti- mating the real growth of larvae in their envi- ronment. Some estimate of growth rate may be obtained, however, from changes in modal length at certain times of the year. During 1965, the modal lengths of a single abundant group of larvae may be traced from the autumn into De- cember. This group was not greatly influenced by mixing with other groups of different sizes and avoidance of our gear was not yet important. From late September to early December, the E O O oc UJ a. JO.OO / CATCHES • A 10.00 TRAWL NET o« -\ BUOYED NET A / * / • / o \ 1.00 NIGHT / / A DAY / V \ 0.10 • / / 1 A / O / 0 0.01 1 1 1 1 1 JAN. FEB. MAR. APR. MAY Figure 11. — Monthly progression of larval catches (lar- vae 36 mm and larger) in the Boothbay area from trawl tows at night and during the day and in the upper Sheepscot estuary from buoyed and anchored nets at night. 316 GRAHAM, CHENOWETH, and DAVIS: LARVAL HERRING ^ larvae appeared to grow about 1 mm every 5 days. Similarly, in 1964 modal lengths may be traced from late September to early November with an apparent larval growth of 1.3 mm every 5 days, but then the mode became difficult to identify. DISCUSSION SEASONAL CHANGES IN ABUNDANCE AND LARVAL LENGTH CAUSED BY MORTALITY After larvae hatched and accumulated in the coastal bays and estuaries, their abundance was determined by the rate at which they died. The autumn mortality was especially severe, for the catch declined in the Boothbay area (Figure 6) despite the addition of successive broods to the area throughout the autumn. This high mor- tality was also indicated by the failure of length modes to persist into the winter. The number of larvae in a given mode apparently was reduced with time and coincidently with differential growth until the larvae were too sparse in the catches to form a distinguishable mode. In 1966 the disappearance of a mode of relatively large fish from the catch caused a sharp drop in the mean larval length during November (Figure 7). Although the mean length in- creased either with subsequent growth of the re- maining larvae or with the addition of larvae to the area, it remained below the mean lengths of the other years throughout the winter. Win- ter mortalities were not as high as those in the autumn. Graham and Davis (1971) determined mortalities from December to January for larvae captured in the Sheepscot estuary during 1964- 67 and recently for mortalities for 1968-69. Es- timates for the 6 years varied from 22% to 52% for 15-day intervals and appeared statistically reliable with the largest spread in the 0.95 fidu- cial interval in 1968 (27.7-36.9%) , and the smal- lest in 1966 (22.0-22.8%). Extensive reductions in our catch rates in the late spring occurred from avoidance of the gear by the larvae and their departure from our sam- pling stations. Catches at the upper estuarine stations in the Boothbay area were larger than those at the lower and outer stations. Progres- sive peaks in abundance, that were present in the autumn, were absent from outer to upper estuarine stations in the spring. One explana- tion for the lack of progression is that the larger larvae moved landward so rapidly as to be un- detected in the spring. Another and more likely explanation is that their mortality was sufficient- ly low to permit an accumulation of larvae at the landward extremity of their movement where numbers always greatly exceeded the number of larvae moving into the area. Estimates of annual mortality were based on winter mortality (Graham and Davis, 1971) be- cause measurements of total mortality during a given year were impracticable as they were in- fluenced by larval movement in the autumn and in the spring. For most year classes, a higher winter mortality recorded in the upper end of the Sheepscot estuary coincided with a smaller maximum catch there in the subsequent spring. Also, higher winter mortalities usually coincided with a poorer condition or well-being of the lar- vae (Chenoweth, 1970) for the Boothbay area (Figure 12). The causes of larval mortality along the west- ern coast of the Gulf of Maine were not deter- mined, but inferences were made by Chenoweth (1970) and by Sherman and Honey (1971). Essentially, they suggested that winter mortal- ity might be related to lower lethal temperatures, inhibition of feeding by low temperatures, and a scarcity of food. Sherman (personal communi- cation) found in recent studies that the larval guts were frequently occluded by parasites, which may cause death. SEASONAL CHANGES IN DISTRIBUTION RELATED TO LARVAL SOURCES After hatching, the larvae shifted their dis- tribution from spawning areas to the coastal bays and estuaries. These destinations were apparent from our catches, but not all the sources or spawning areas were determined. The sources of larval herring in the Gulf of Maine, including the vvestern coast, their movements from these 317 FISHERY BULLETIN: VOL. 70, NO. 2 SHEEPSCOT ^ 20 64 65 66 67 68 69 YEAR CLASS Figure 12. — A comparison among year classes of winter mortality of larvae in the Sheepscot estuary (upper panel), the decline in condition of larvae in the Booth- bay area (center panel), and the subsequent spring catch rates in the Sheepscot (lower panel). sources, and the resulting changes in their dis- tributions have received attention from other investigators. Spawning grounds of herring were located in the past by surveying the distribution of recently hatched larvae (Tibbo et al., 1958), and more recently by collecting eggs off the spawning grounds (Noskov and Zinkevich, 1967). Migra- tions of larvae from these grounds were traced by assuming that they were transported simi- larly to a particle of water. Thus, their paths of migration from the grounds were located with- in the residual currents at the surface. Boyar et al. (1971) reviewed such evidence for the Georges Bank-Gulf of Maine area and decided that larvae may enter the eastern sector of our sampling area from the spawning grounds off southwestern Nova Scotia, and that some larvae from Nova Scotia may be carried along the coast of Maine as far as Cape Cod, Mass. In a con- current paper, lies (1971) reported on the dis- persion of larvae from southwest Nova Scotia and concluded from his data that the larvae were transported into the Bay of Fundy where they were retained during the winter. Boyar et al. (1971) also suggested that larvae entered the western sector of our coast from Jeffreys Ledge, Cashes Ledge, Stellwagen Bank, and other areas collectively just offshore of the coastal sec- tor. The larvae from these two sources, Nova Sco- tia and the ledges and banks in the southwestern Gulf of Maine, possibly contributed to the coastal larval population in the spring. Such a contri- bution would partially explain the high catch rates obtained in the Boothbay area where larvae accumulated in the autumn and spring. The autumnal movements of larvae into the area sub- sided by early December. The catch in the area at that time was less than the peak catch ob- tained in the subsequent spring, indicating that larvae present in autumn could not account for all of the larvae found in the same area in the spring. Much of the spring catch, therefore, may be formed by emigrants. Das (1968, see footnote 2) also discovered a similar emigration by examining length-frequency distributions of larvae in the coastal area of southwest Nova Scotia. And Sameoto (1971)^ reached a similar conclusion from catches of larval herring enter- ing St. Margaret Bay on the southeastern coast of Nova Scotia. Another explanation for the differences in the early winter and the spring catches in our sam- * Sameoto, D. D. 1971. The distribution of herring (Clupea harengus L.) larvae along the southern coast of Nova Scotia with some observations on the ecology of herring larvae and the biomass of macrozooplankton on the Scotian Shelf. Fish. Res. Board Can., Tech. Rep. 252, 72 p. 318 GRAHAM, CHENOWETH. and DAVIS: LARVAL HERRING pling areas may be deduced from the residual surface currents. In winter, the direction of these currents is usually offshore; in spring, the currents are often directed inshore. Thus, lar- vae swept offshore in the winter might be re- turned to the coast in the spring. However, we did not detect any concerted movement by the larvae offshore in the winter and suspect that the assumption that larvae are transported si- milarly to a particle of water is often an oversim- plification. The factors controlling the move- ment and retention of the larvae in shoal water must be investigated to understand the possibil- ities of their transport. ANNUAL CHANGES IN ABUNDANCE AND THE FISHERY 90r r 80 UJ o K liJ a. 70 £ 60 t- (/) in V) < -> 50 o IT < 40- • 66 • 68 1 67 1 65 .64 20 30 40 50 60 LARVAL MORTALITY (PERCENT) 70 The goal of this research was to choose an estimate of larval abundance or its correlate that could be used to predict the annual recruitment of immature 2-year-old herring to the sardine fishery of the western coast of the Gulf of Maine. Three different types of estimates were chosen: (1) winter mortality, (2) condition in the win- ter, and (3) maximum abundance in the spring. Sampling should continue over a number of years to determine whether the relations between the three measures have substance and whether any one or all of them are pertinent to predicting the abundance of 2-year-old herring. A tentative comparison between the percent- age of 2-year-old fish taken in the fishery with the percentage of winter mortality of the cor- responding year class during 1964-68 is shown in Figure 13. Years of low mortality were usu- ally related to subsequent greater percentage of 2-year-old fish in the fishery. Estimates of winter condition are important because they provide an insight as to the cause of larval mortality. Because winter condition correlates with winter mortality, larval deaths are probably caused by debilitating factors such as disease, starvation, or parasitism. But, lack of agreement during 1969 (Figure 12) would involve other factors as well, such as predation or sudden and transient effects of man's activities within the coastal environment. Figure 13. — Comparison between the percentage of 2- year-old herring captured in the Maine sardine fishery and the winter larval mortality for a given year class. Age composition of the fishery was determined by John E. Watson (personal communication). The catch also includes fish from the adjacent Canadian coast. Estimates of abundance in the spring are nec- essary as well as winter mortality estimates be- cause the number of larvae surviving until spring theoretically depends upon the initial number of larvae present by the end of autumn. During each of the years in which winter mortality was estimated in this study, the autumnal abundance was reduced to approximately the same level by early winter. During years of very successful hatching, the autumn mortality might not be suf- ficient to reduce the number of larvae to a level common to that of previous years. The sub- sequent spring abundance would then be deter- mined by the initial number of larvae present in early winter as well as the winter mortality. The estimates of spring abundance in the Boothbay area and in the coastal area between Cape Small and Penobscot Bay (Figure 14) did not agree with the estimates obtained with buoyed and anchored nets in the Sheepscot for year classes 1964-65 (Figure 12, bottom panel). We do not understand the reason for this difference. To date, monitoring of the winter mortality 319 FISHERY BULLETIN: VOL. 70, NO. 2 and condition and spring abundance of larvae has been largely confined to the Boothbay area; the possibility of extending monitoring to other areas of the coast is being investigated. Never- theless, events in the ecology of the larvae in the Boothbay area could represent those of the entire coast of the Gulf. Seasonal changes in larval abundance are similar along the coast and are comparable to those in the coastal area off south- west Nova Scotia (Das, 1968, see footnote 2) and on Georges Bank (Boyar et al., 1971). Yearly changes in oceanic conditions along the coast and in the offshore Gulf of Maine are also related (Colton, 1968). Further correlation between ecological events in the Boothbay area and those in other areas of the Gulf, at least adjacent areas, is evidenced by the agreement of the winter mor- tality estimates obtained in the Sheepscot and the subsequent spring abundance there. Larvae captured in the spring include emigrants from areas other than Boothbay. Correct forecasts of a poor fishery, to date, coincided with poor re- cruitment to the fishery; unfortunately, fore- casts have not been made during years of good recruitment in the western sector. ^■^"1 BOOTHBAY llllL a. V) l.5n 10 I 0,5 COASTAL 64 65 66 67 68 69 YEAR CLASS Figure 14. — Peak catches for larval herring in the spring among year classes 1964-69 ; in the coastal area between Cape Small and Penobscot Bay and in the Boothbay area. SUMMARY OF ABUNDANCE AND DISTRIBUTION The abundance and distribution of larval her- ring along the western coast of the Gulf of Maine is determined by their movements and mortality. Two inshore movements, one in the autumn and the second in the spring, are separated by a pe- riod of larval dispersal. Larvae hatch along the coast in autumn and penetrate the bays and estu- aries. Their inshore movement decreases by early winter and in midwinter they disperse; at this time concentrations of larvae are infrequent inshore and along the coast. A second shore- ward movement begins with the advent of spring. Larvae which hatched the previous au- tumn along the coast, and probably some which hatched beyond our sampling area, aggregate when making this shoreward movement. The inshore movement is completed by the end of spring when the majority of the larvae have assumed their adult form. Mortality during the inshore movement in the autumn is very high. Although it is lower in the winter, mortality during this season may determine the abundance of larvae in the spring because the numbers are reduced by early winter to a relatively common level each year. The low- er winter mortality may be related to the dis- persal of the often weakened larvae; dispersal would reduce intraspecific competition for food and space. The lowest mortality occurs in the spring when the larvae are in good condition as shown by their ability to avoid high-speed trawls with large mouth openings. The movement of the larvae shoreward with- in the complex currents of the coast is a striking feature of their coastal ecology. LITERATURE CITED Anthony, V. C, and H. C. Boyar. 1968. Comparison of meristic characters of adult Atlantic herring from the Gulf of Maine and ad- jacent waters. Int. Comm. Northwest Atl. Fish., Res. Bull. 5:91-98. BiGELOW, H. B., AND W. W. WELSH. 1925. Fishes of the Gulf of Maine. Bull. U.S. Bur. Fish. 40(Part I),567p. 320 GRAHAM. CHENOWETH. and DAVIS: LARVAL HERRING BOYAR, H. C. 1968. Age, length, and gonadal stages of herring from Georges Bank and the Gulf of Maine. Int. Gomm. Northwest Atl. Fish., Res. Bull. 5:49-61. BoYAR, H. C, R. R. Marak, F. E. Perkins, and R. A. Clifford. 1971. Seasonal distribution of larval herring, Clu- pea harengus harengus Linnaeus, in Georges Bank-Gulf of Maine area, 1962-70. Int. Comm. Northwest Atl. Fish., Res. Doc. 71/100, 11 p. BUMPUS, D. F., and L. M. Lauzier. 1965. Surface circulation on the continental shelf off eastern North America between Newfoundland and Florida. Ser. Atlas Mar. Environ. Am. Geogr. Soc. Folio 7, 8 plates, [4] p. Chenoweth, S. B. 1970. Seasonal variations in condition of larval herring in Boothbay area of the Maine coast. J. Fish. Res. Board Can. 27:1875-1879. CoLTON, J. B., Jr. 1968. Recent trends in subsurface temperatures in the Gulf of Maine and contiguous waters. J. Fish. Res. Board Can. 25:2427-2437. CoLTON, J. B., Jr., and R. F. Temple. 1961. The enigma of Georges Bank spawning. Limnol. Oceanogr. 6:280-291. Davis, C. W., and J. J. Graham. 1970. Brit herring along Maine's coast. Commer. Fish. Rev. 32(5) :32-35. GeHRINGER, J. W., AND W. ARON. 1968. Field techniques. In D. J. Tranter (editor) , Reviews on zooplankton sampling methods, p. 87- 104. UNESCO Monogr. Oceanogr. Methodol. 2 (Part 1). Goode, G. B. 1884. The herring tribe. The herring — Clupea harengus. In G. B. Goode and associates. The fisheries and fishery industries of the United States. Section I. Natural history of useful aquatic animals, p. 549-568. Gov. Print.' Off., Wash. [D.C.] Graham, J. J. 1970a. Temperature, salinity, and transparency observations, coastal Gulf of Maine, 1962-65. U.S. Fish Wildl. Serv., Data Rep. 42, 43 p. 1970b. Coastal currents of the western Gulf of Maine. Int. Comm. Northwest Atl. Fish., Res. Bull. 7:19-31. Graham, J. J., and H. C. Boyar. 1965. Ecology of herring larvae in the coastal waters of Maine. Int. Comm. Northwest Atl. Fish., Spec. Publ. 6:625-634. Graham, J. J., and C. W. Davis. 1971. Estimates of mortality and year-class strength of larval herring in western Maine, 1964- 67. Cons. Perm. Int. Explor. Mer, Rapp. P.-V. Reun. 160:147-152. Graham, J. J., and G. B. Vaughan. 1966. A new depressor design. Limnol. Oceanogr. 11:130-135. Graham, J. J., and P. M. W. Venno. 1968. Sampling larval herring from tidewaters with buoyed and anchored nets. J. Fish. Res. Board Can. 25:1169-1179. Iles, T. D. 1971. The retention inside the Bay of Fundy of herring larvae spawned off the southwest coast of Nova Scotia. Int. Comm. Northwest Atl. Fish., Res. Doc. 71/98, 7 p. NOSKOV, A. S., AND V. N. ZiNKEVICH. 1967. Abundance and mortality of herring, Clupea harengus Linnaeus, on Georges Bank according to the results of egg calculation in spawning areas in 1964-66. Int. Comm. Northwest Atl. Fish., Res. Doc. 67/98, 16 p. RiDGWAY, G. J., R. D. Lewis, and S. W, Sherburne. 1969. Serological and biochemical studies of her- ring populations in the Gulf of Maine. Int. CounC. Explor. Sea C. M. 1969/24, 13 p. Sherman, K. 1970. Seasonal and areal distribution of zooplank- ton in coastal waters of the Gulf of Maine, 1967 and 1968. U.S. Fish Wildl. Serv., Spec. Sci, Rep. Fish. 594, 8 p. Sherman, K., and K. A. Honey. 1971. Seasonal variations in the food of larval herring in coastal waters of central Maine. Cons. Perm. Int. Explor. Mer, Rapp. P.-V. Reun. 160: 121-124. TiBBO, S. N., J. E. H. Legare' L. W. Scattergood, and R. F. Temple, 1958. On the occurrence and distribution of larval herring {Clupea harengus L.) in the Bay of Fundy and the Gulf of Maine. J. Fish. Res. Board Can. 15:1451-1469. 321 AN ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES Michael D. Dahlberg^ ABSTRACT The distribution of fishes in relation to ecological factors was studied in a Georgia estuary and adjoining beach and coastal plain creek waters to establish fundamental base lines for the ecology of a relatively undisturbed section of the coast. The distributions of 168 fish species were related to nine recognizable habitats, temperature, and salinity. Length frequencies and spawning seasons were determined for most of the trawled species that contributed significantly to fish production in the estuary. Collecting by a variety of techniques permitted evaluations of types of life cycles, of diversity in the various aquatic habitats, and of distribution patterns within the estuarine complex and adjoining waters. Numbers of species collected in nine aquatic habitats were as follows: beach - 114, lower reach of the estuary - 100, high marsh - 56, upper and middle reaches - 61, oligohaline creek - 40, fresh water - 39, tidal canal - 22, low- salinity tidal pool - 22, high-salinity tidal pool - 37. Estuaries are highly productive and support im- portant sport and commercial fisheries. A ma- jority of the nation's commercial finfish and shellfish species and many coastal sport fishes utilize the estuarine environment during at least part of their life cycle. Estuaries are important recreational areas, especially for fishing, partly because of their proximity to civilization. Un- fortunately, this proximity and lack of pollution controls have resulted in mass degradation of the nation's estuaries through pollution, filling, and dredging. Pre-pollution studies are essen- tial for the precise evaluation of the ecological impact of stresses on estuaries. The normal functioning of estuaries must be understood be- fore scientists can evaluate the effects of var- ious stresses on estuaries. There is an urgent need to determine the significance of coastal ha- bitats to the various life history stages of coastal fishes. The central Georgia coast presents an opportunity to study the ecology of fishes in a relatively undisturbed estuary and establish fundamental base lines for the detection and evaluation of pollution. Aspects of fish ecology that have been selected for examination include ^ University of Georgia, Marine Institute, Sapelo Is- land, GA 31327; present address: Virginia Institute for Scientific Research, 6300 River Road, Richmond, VA 23229. (1) distribution of fishes in relation to recogniz- able habitats, salinity, and temperature, (2) size frequencies and spawning seasons of many abundant species, (3) diversity of fishes in each of nine habitats, and (4) types of life cycles. Except for the Brunswick and Savannah re- -gions, Georgia estuaries are relatively free of pollution. However, certain Golden Isles along the Georgia coast are being considered for strip mining for titanium and phosphate. A nuclear power plant is under construction on the lower Altamaha River. Recent developments in the estuary studied herein include a kraft paper mill, a shrimp culture farm that may not be completed, and an interstate highway that may alter tidal flushing. A complete picture of species distributions and size in coastal waters can be obtained only if all major habitat types are examined (Springer and Woodburn, 1960) . Extensive trawling and seining were undertaken to achieve this goal. Comparisons with other studies are complicated by the difference in types and numbers of hab- itats sampled and differences in collecting meth- ods. There has been little work on the ecology of fishes of the U.S. Atlantic coast between Cape Hatteras, N.C., and northern Florida. Tagatz (1968) surveyed the fishes of St. Johns River, Manuscript accepted January 1972. FISHERY BULLETIN: VOL. 70, NO. 2, 1972. 323 FISHERY BULLETIN: VOL. 70, NO. 2 Fla. Tagatz and Dudley (1961) studied the sea- sonality of fishes in four coastal habitats near Beaufort, N.C. In South Carolina, Bearden (1961) published a field guide to the common marine fishes, compiled an unpublished list (1961) of marine fishes, and surveyed the elas- mobranchs (Bearden, 1965). Lunz and Schwartz (1970) published an 18-year study of South Carolina coastal fishes. Anderson (1968) surveyed the fishes caught by shrimp trawling from South Carolina to northeastern Florida from 1931 to 1935. His data are not tabulated here because many spe- cies complexes were identified only to genus and his collections were from regions other than the estuary under study. Miller and Jorgenson (1969) studied the seasonal abundance and length frequencies of fishes collected by seining at a beach habitat on St. Simons Island and at two high marsh stations, one near Jekyll Island and one near Meridian, Ga. They also listed the fishes collected at a freshwater station in the Altamaha River. Dahlberg and Heard (1969) surveyed the common elasmobranchs of the Georgia coast. Dahlberg and Odum (1970) demonstrated the trawl diversity of Georgia es- tuarine fishes collected over 14 months. Dahl- berg (1971a)' presented an annotated list of the Georgia estuarine and coastal fishes. A section of an unpublished ecological survey (Dahlberg, 1971b) ' was the basis for this report. LOCATION AND DESCRIPTIONS OF HABITATS Composition and diversity of fish species in nine aquatic habitats along the Georgia coast are compared in this study. Salinity, tempera- ture, and some aquatic plants and animals that are characteristic of these habitats are given in Table 1. = Dahlberg, M. D. 1971a. An annotated list of Geor- gia coastal fishes. In An ecological survey of the coastal region of Georgia, p. 255-300. Unpublished report to National Park Service from University of Georgia In- stitute of Natural Resources, Athens. ' Dahlberg, M. D. 1971b. Habitat and diversity of the fishes in North and South Newport Rivers and adja- cent waters. In An ecological survey of the North and South Newport Rivers and adjacent waters with respect to possible effects of treated kraft mill effluent, p. 36-121. Unpublished report to Georgia Water Quality Control Board from University of Georgia Marine Institute, Sapelo Island. Table 1. — Salinity, temperature, characteristic plants and animals of habitats studied, exclusive of freshwater creek habitat. Sapelo Island Btach Salinity— Range was 25.0 to 31.3%o except when flood waters reduced salinity. Measured at 6.8 to 7.7%,, along south end of beach and l5.2%o in surf near Big Hole at low tide in April 1970. Plants— Sea oais (Uniola pankulata) are the nnost conspicuous plant on dunes along the beach. Invertebrates seined or observed— Ghost crab (Ocypode quadrata), polychaete worm tubes (Onuphii microcephala, and Diopatra cupria), horseshoe crab (Limulus Polyphemus), hermit crabs (Clibanarius vittatus, and Pagurus longicarpus) , gastropods (Busycon carica, and Nassarius), sea cucumber (Thy- one hriareus), sand dollar (Mellita quinquiesperjorata) , isopod (Aegothoa oculata), white shrimp (Penaeus setijerus), shrimp (PalatmonetfS pugio) , blue crab (Callinectes sapidus), and squid (Lolliguncula brevis). Lower Reach of the Estuary - Trawl Stations 1-9, 12-14 Salinity-Averages for these stations ranged from 21.4 to 28.9 %, from 1967 to 1970 (Dahlberg et al., 1971, see text footnote 3). Temperature— Averages fc" the 12 stations ranged from 8.1° (January) to 31.2°C (August). Plants— Cord gross (Spartina alternifiora) is the dominant plant. Invertebrates— Commercial species are white shrimp, brown shrimp (Penaeus aztecus), and pink shrimp (P. duorarum), American oyster (Crassostrea virtinica), and blue crab. Other common species are too numerous to list here. High Marsh Salinlty-Generally 15 to 30^. Plants— Sfarfina alternifiora and /uncus roemerianus ore characteristic. Middle and Upper Reaches Salinity-Range was 11.7 to 29.0^ and average was 21.2%<, at lowermost station, trawl station 10. Range was 0.3 to 18.7^ and average was 5.3^ at uppermost station (F). Temperafure-Ronge was 8,3° to 31.7°C and average was 20.9°C of the lowermost station, trawl station 10. Range was 9,0° to 30,0°C and average was 21.5°C at uppermost station, station F. Aquatic plants— Sfar/ina alternifiora, Juncus roemerianus, toll cord gross (Spartina cynosuroides), marsh elder (Iva jrutescens), bulrush (Scripus robustus), primrose-willow (Jussiaea). 324 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES Table 1. — Continued. Sapelo Island Beach — Cont. Semiaquatic invertebrates— CAca pugnax, Sisarma, U. pugilator. Aquatic invertebrates that v^ere seined, in decreasing order of abundance— Pi2/a *j • Ik a; en s u 0) X 6r «J o m V >> »H X >4 a s. c a 3 »4 •o IKS 05 ^ »3 o rt — u 01 •sz o cS 3 CO C o o .« <^ CO- =? .. bo •^ IS c! - o .. « "o N O .. '^ ^ '^ 03 'rt ^- £ U -^ Vh -^3 O <4H , O tn C bo o -l-> 3 ns p -M >. ^ 1 1 ^ 00 3 rt rJ ^ 73 So CO '^ OS "^ ri<: ^ 5 w as - 'E w (D O _ Ch CO < C o CO bo Ol "-5 CO 73 O ^ 03 'V bo c -t-> o s t> CIS *•, A! 09 Ill a> V * M CO y «C M •s 03 n ^ o X 3 CO XI «*-( »H ■. .M «is C« P -i*: 1 OJ 1 ii> (N a V >-3 c ^ "nl H X 03 u C tn CO OJ - ^ 1-5 3-73 c o . 0! M C « bos c -a PL| >H ID s O 5"^ c4 O <5 ? «^ J? 2 -:^ -2 5 "S 4) " bo 45 _e ^ ^^ IS X X XX X X XX XXXXXX XX XX XX XX X X X X X X X X X X X XX X X X X XX X X X X X X X X X X XX X X X X X X XXX X XXX (U -.J <: 3 .5 rS <3 O C3 •D o :^ -« -« -c -e i> ^ «-. Vj Vj ".J O n B e e e T3 ^ 0) » * o e 6 T) UJ OJ ^ 0 11 - ^ .t: ^ ° t; o CO Q^ <; <; o 3 I" i: < < 03 ♦ E < o -C a < ^ "5 CO 15 0 O J^ 5 e ~ o -o O V ^^ u ^ ■ D II; 'a k. -2 ,a £ O e •o -«: 3 E e »-v ^ r 3 3 3 3 0 "8 3 3 3 a k. '1:1 >i -^1 ■V( •^ -n ft. c K e e C |£ ^. a 3 3 3 3 3 i- O 1*, ^ ^ ^ t*. t*. > * « * ♦ * (J X X X X X XX X S SS^SSS Sx X X X XXXX X X X XXXXXX X XXX X X X X X X XX X X X X X X SS8 S X X X X X XX XXX XX X X X Xxxxxx X XXx XX XX X X xxgxxxx SCO o y E O tfl o .S S. " <3 Vj I EE 5 5 2 "S -e c « o -e 1 ^ ■: O ■fi •? o U S S s ^rf «.> v. O y 13 i^:§ «U3 -2 § tt, 05 00 * -» * ♦ -D -« -D - >. O. ^ LU -n ^ " ^ a a '5) ^ ^ « a ^ ^ ^ n ^ (C t^ o ^^ u -n 'f " o. «i, O c r >■ CO W c o 3 §-s a to ^ -o rS -S S S; a; 3 * * 5 ":= S a ^ > 41 < ^ -^ oi a; t/o c * 0 > D. < 00 4> 3 O c o S 1^ . E o £ £ ■= a a c ^ ,y 2 O O M -a -< a a E E E E .^ a g E E E E E o 5 U -S -:2 a a «. '^ « o U I (u J to © t3 *- D ^ ^ ^ -£ -t: -*■ fi "3 vj T >* ^ --I *4 a <\ E o O S c a:< , Q, a. 0, 0, 0, includes the North Newport River and lower parts of its tributaries from the lower end of Carrs Neck Creek (trawl station 10) upstream to station C. The upper reach extends from Payne Creek to the mouth of Riceboro Creek (station F) . I generally treat these two sections as a unit for convenience. Because salinity varies so greatly, this section of the estuary only roughly corresponds to the zones recognized by Carriker (1967). He rec- ognized salinity ranges of 5 to \%%c for the up- per reach and 18 to 25Ar for the middle reach. Habitat 5: Oligohaline creek (Figure 6). — The North Newport River originates at the con- fluence of Peacock and Riceboro Creeks. Oligo- haline sampling stations were located in the oligohaline section of Riceboro Creek (Figure 2) . Most fishes were collected in Riceboro Creek at station G and the lower part of Crossroads Creek where it joins Riceboro Creek near the town of Crossroads. Crossroads Creek was recently di- verted into Riceboro Creek above the station by the State Highway Department, thus eliminating the station. Salinity and temperature ranges are given in Table 1. A temperature-salinity diagram (Figure 7) indicates that salinity gen- erally increased with rising tide except when fresh water prevailed throughout the tidal cycle. In winter water temperatures were lowest at low tides. Habitat 6: Freshwater creek. — Seven fresh- water or limnetic stations in Coastal Plain creeks, free from the influence of salt water, are called "freshwater stations" in the text. The first six of these stations are in the North New- port River drainage in Liberty County and were sampled on 13 May 1968. The seventh is in the Ogeechee drainage in Chatham County and was occupied on 9 March 1969. Location of these collections are as follows: (1) Upper South Newport River at U.S. Highway 17; (2) Head- waters of Payne Creek on U.S. Highway 17, 2 miles west of Riceboro; (3) Tributary of Rice- boro Creek about 2.5 miles north of Crossroads on a county road ; (4) Peacock Creek, 2 miles southeast of Mcintosh, on dirt road off U.S. Highway 82; (5) Goshen Swamp Creek on U.S. Highway 82, 2 miles south of Flemington; (6) Peacock Creek at Solomon Temple, between 330 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES (^*-»-*S Figure 5. — A muddy sand bar exposed at low tide at the mouth of Payne Creek where fishes were seined. Spartina altertiiflora is visible on the right and the "back marsh" contains S. alterniflora and Juncus roemerianus. Hammocks and larger islands in this region are bordered by salt marshes and tidal creeks. Figure 6. — Riceboro Creek near high tide at the Seaboard Railroad bridge. A variety of invertebrates are associated with the diverse intertidal flora, roots, and debris found in this oligohaline creek. 331 FISHERY BULLETIN: VOL. 70, NO. 2 31 27 - 23 - o e UJ o: Z) UJ 0. UJ I- 19 15 - II - 1 1 1 1 A8 •S 1 1 1 1 A9 "a6 •^ +6,7 _ A;>4 • 9 A--IO- -^•~ A 3 — • 1 - • II Al - ^)2#I2 12 1 ▲ 1 1 1 1 1 1 1 8 10 12 14 Vo SALINITY voo Figure 7. — Temperature-salinity diagram for Riceboro Creek at U.S. Highway 17; Numbers 1-4 are months of January through April 1969, and 6-12 are June through December 1968. Circles are high tide readings and tri- angles are low tide readings. Cross is high tide value for June and low tide value for July. Midway and Mcintosh on dirt road off U.S. High- way 82; (7) Tributary of Ogeechee River at U.S. Highway 17 near the Ogeechee River bridge. Habitat 7: Tidal canal — This station is a runoff ditch or canal located in the marsh on the west side of Sapelo Island, 3.4 miles north of the Sapelo airport. This ditch is open to a poly- haline river (Mud River) at all tide levels and its mud banks are under several feet of water at high tide. Water depth was 3 to 4 ft at low tide in the deepest section, which was located at the end of a road culvert on the salt marsh side of a road. Habitat 8: Low-salinity tidal pools — Two small pools located next to the road on the west side of Sapelo Island are treated together in the text. These pools are located 4.1 miles north of the Sapelo airport on the west side of the road. Runoff water flooding through a culvert opening into the lower and larger pools has eroded the bottom to a depth of approximately 5 ft. A higher culvert opening to the upper and smaller pool rarely has runoff water. This pool is 3 to 4 ft deep. At low tide only trickles of water connect the two pools and drain the lower pool through the marsh. Located behind the high marsh, these pools are flooded by Mud River only on high tide. Habitat 9: High-salinity tidal pools (Figure 8) . — These are a series of artificial pools located on the south end of Sapelo Island along a road leading to the Sapelo lighthouse, hence the local name lighthouse ponds. The pools are located behind a high marsh characterized by Spartina alterniflora and Juncus roemerianus. The pool sampled on the west side of the road is flooded by waters from South End Creek and the Ma- rine Institute's boat basin. Two pools sampled on the east side of the road are connected at high tide and are flooded by waters from Deans Creek. Palaemonetes piigio was abundant in seine collection, except during the coldest weeks when the water temperature approached 8°C. At this time the fishes were also scarce or absent, and the relatively sterile and clear water allowed the author to see the bottom of the pools where- as the water was very turbid during other months. METHODS I seined shallow estuarine habitats (habitats 1, 5, 7, 8, and 9 and 4 in part) with a 35-ft (10.7-m) seine having 14-inch (6.35-mm) bar mesh; a 10-ft (3.05-m) version was used in freshwater creeks. For habitat 3 I used the rec- ords of Miller and Jorgenson (1969) who col- lected with 40- and 70-ft (12.2- and 21.4-m) bag seines. Their records for St. Simons Island Beach are included in the list for habitat 1. Habitat 2 (trawl stations 1-9, 12-14 in Fig- ure 1) and the lower part of habitat 4 (trawl 332 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES Figure 8. — Tidal "borrow pool" on west side of road to lighthouse on Sapelo Island. Zonation of dominant plants along road bank generally from left to right in picture : wax myrtle (Myrica cerifera) , groundsel-tree (Baccharus halimifolia) , marsh elder {Iva frutescens), and sea ox-eye (Borrichia frutescens) . Juncus roemerianus and Spartina alterniflora are dominant along the pool bank. There is an abundance of crabs (Uca spp. and Sesarma sp.) in both the harder and softer substrates between the pool and the road. stations 10 and 11) were sampled on RV Kit Jones cruises, each station for 15 min with a 20-ft (6.1-m) otter trawl which had 114-inch (32-mm) stretched mesh in the bag. Each of the 14 trawl stations was sampled about 37 times from January 1967 through December 1969 ex- cept for stations 6, 8, 12, and 13, which were sampled 33 or 34 times and then discontinued. P'ive trawl samples were lost for various reasons including snags at station 9 on two occasions. All Sapelo Island habitats (1, 7, 8, and 9) were sampled during all four seasons and about 20 times from April 1967 through February 1970. Habitat 4 was sampled primarily with a 20-ft (6.1-m) trawl and a 35-ft (10.7-m) seine. The two trawl stations (10 and 11) are the upper- most of the 14 stations sampled with the RV Kit Jones from January 1967 through December 1969. Seining stations were sand bars that were exposed only near low tide. Regular stations were at the lower end of Carrs Neck Creek (mile point 10 = 16.1 km) and the mouth of Payne Creek (mile point 15.4 = 24.8 km). These sta- tions were seined approximately monthly from June 1969 to May 1970. In addition 22 collec- tions were made between the upper end of Carrs Neck Creek and mouth of Riceboro Creek with the seine, a cast net, and a 10-ft (3.05-m) trawl towed by an outboard motor boat. In habitat 5 most fish collections were seined at low tide at a muddy sand bar (station G) be- tween the Seaboard Railroad tracks and the effluent outfall of the Interstate Paper Corp. Trawling (10-ft trawl) was most fruitful in the winter. Seine and dip-net collections were made at the Crossroads station. A few additional collections were made with a 100-ft (30.5-m) seine in habitat 1, by angling in habitats 1 and 2 (Dahlberg and Heard, 1969), 333 FISHERY BULLETIN: VOL. 70, NO. 2 and by dip net in habitat 8 and in Riceboro Creek at the Seaboard Railroad. All specimens were preserved in lO^^f formal- dehyde solution and analyzed in the laboratory except when numbers in trawls were too large to retain more than a sample. Representatives of all species were retained at the Marine Institute. The phylogenetic order and names recom- mended by Bailey (1970) are followed herein. HABITATS OF COASTAL FISHES ORDER SQUALIFORMES - SHARKS Ten species of sharks (Table 2) were col- lected in this study and by Dahlberg and Heard (1969). Nearly all specimens, bonnethead ex- cepted, were caught with fishing poles at the beach and in the sounds. Only three sharks were trawled, one each of the spiny dogfish {Squalus acanthias) , bonnethead (Sphyrna ti- huro), and the blacktip shark {Carcharhinus limbatus). Most shark species have distinct seasonal mi- gratory patterns. Carcharhinid sharks appar- ently migrate into the estuary during the warm months as all six were collected from June to September. The spiny dogfish is a cold-water species that migrates into Georgia estuaries during the coldest months (Dahlberg and Heard, 1969); it was collected only in January and February. ORDER RAJIFORMES AND RAYS SKATES The Atlantic guitarfish (Rhinobatos lentigi- nosus) and clearnose skate (Raja eglanteria) are most common in the ocean and were not col- lected farther up the estuary than the sounds. Although the guitarfish was collected only in May and June, the skate was taken throughout the year. Three stingrays of the genus Dasyatis (Table 2) are common in the sounds and shallow waters along the beaches. The Atlantic stingray (D. sabina) was the most abundant of the three spe- cies taken by trawling and angling. It was com- monly trawled in the middle reach of the estu- ary at salinities as low as 9.9^f. This species is euryhaline (Gunter, 1956). The smooth butterfly ray (Gymnura micrura) was collected by trawl in the estuary as far up as the middle reaches but not at salinities lower than 24.4%c. ORDER SEMIONOTIFORMES Lepisosteidae - gars Two species of Lepisosteus were taken in the study area. The Florida gar (L. platyrhincus) occurred only at freshwater station 2. The eury- haline longnose gar (L. osseus) was abundant and was often seen or trawled in fresh and brack- ish waters throughout the year. The longnose gar has been collected in the ocean off Georgia. ORDER ELOPIFORMES Elopidae - tarpons Small ladyfish {Flops saurus) were common in enclosed waters of tidal pools and the tidal ditch. Only two specimens were collected in open waters, one at Sapelo Beach and one in the upper reach of North Newport River. The lady- fish was common in both the high- and low-sa- linity pools at a salinity range of 0.1 to 28.7%f. It entered the pools in May and remained until November when the temperature was 19.9°C. Small tarpon (Megalops atlantica) were com- mon in enclosed waters of tidal pools and the tidal ditch, and large tarpon were often hooked by anglers along Sapelo Beach. Young were col- lected only from July to October at a temperature range of 20.0° to 31.9°C and were more common in the low-salinity pools than the high-salinity pools; the recorded salinity range was 0.1 to 24.8^f. Rickards (1968) found tarpon at Sa- pelo Island from July to November at salinities of 0.0 to 22.3;^r and temperatures of 16° to 36°C. Albulidae - bonefishes Inclusion of the bonefish (Albula vulpes) in Table 2 is based on a record from St. Simons Beach (Miller and Jorgenson, 1969). 334 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES ORDER ANGUILLIFORMES Anguillidae - freshwater eels The catadromous American eel {Anguilla ro- strata) was widely distributed and abundant in the freshwater habitat and the high-salinity pools, but absent from trawl stations. All small specimens (54-170 mm) occurred in low salini- ties (0-1. 9^f) in the freshwater and oligohaline creeks, and the low-salinity pools. Large speci- mens (230-470 mm) occurred primarily in saline waters (13.1^30.3%c) but one was collected in fresh water. Ophichthidae - snake eels One speckled worm eel (Myrophis punctatus) was collected with a dip net in Riceboro Creek under the Seaboard Railroad bridge. The 243-mm specimen was taken at a salinity of 0%c and temperature of 24°C. Miller and Jorgenson (1969) also reported it from the beach and high marsh habitats. Three shrimp eels {Ophichthus gomesi) were trawled in the lower reach at stations 4, 7, and 9. They probably burrow and easily escape the trawls. I have also seen specimens of the pale- spotted eel (0. ocellatus) from the Georgia coast. ORDER CLUPEIFORMES Clupeidae - herrings Three anadromous species of shad (Alosa) are listed (Table 2) for the habitats along their mi- gratory route. I collected the American shad (A. sapidissima) and hickory shad (A. medioc- ris) only in saline waters. Both of these species and the blueback herring (A. aestivalis) spawn in the nearby Altamaha River (Godwin and Adams, 1969). Young American shad were occasionally col- lected from the lower reach to the upper reach from December to April. Since the spawning season of the American shad is from March through May in the Altamaha River (Godwin and Adams, 1969) , 60- to 114-mm specimens col- lected from December to March were approxi- mately 1 year old, and a 29-mm shad collected in April was recently spawned. Both age groups were represented in the upper reach but only individuals of age group I were taken in the middle and lower reaches. Two species and a hybrid of menhaden {Bre- voortia) (Dahlberg, 1970) occur within the es- tuary. The Atlantic menhaden {B. tyrannus) occurred in eight habitats. Compact schools of adults occurred in the lower reach of the estuary and along the beaches from spring through fall. Smaller numbers were present in the sounds in the winter. Juveniles (29-42 mm) were often collected from June to September in the upper reach where salinity ranged from 0.5 to 16.8%o and temperature, from 28.4° to 30.8°C. Juve- niles were also collected in the oligohaline Rice- boro Creek and this species is known to occur in freshwater (Gunter, 1956). Young menhaden (30-99 mm) also occurred in the tidal pools and tidal ditch in May and June. The yellowfin menhaden {B. smithi) was col- lected in the high-salinity pools, tidal canal, and along the beach. Its absence from trawl col- lections supports my theory (Dahlberg, 1970) that B. smithi is a bay or shallowwater species. The occurrence of this species apparently is re- stricted by low temperature as it first appeared in May and was present until water temperature decreased to approximately 20°C in November. Its absence after November may result from a southward migration since it is common along Florida in the winter (Dahlberg, 1970). The hybrid (B. smithi x B. tyrannus) was the least common of the menhadens. It was col- lected with B. smithi on 5 May 1969 in the tidal canal which had a salinity of 18.8%r and a tem- perature of 29 °C. Hybrids were collected also along the beach and in the Marine Institute's boat basin at Sapelo Island. The gizzard shad (Dorosoma cepedianiim) and threadfin shad (D. petenense) are euryha- line species that are important forage fishes in Georgia reservoirs. The gizzard shad is known from the upper reach and the oligohaline creek from two trawl collections made on 19 November 1969. These large gizzard shad (205-220 mm) were collected at water temperatures of 15.5° and 14.1°C, and others were caught in the Ma- rine Institute's boat basin (lower reach habitat) in the summer of 1969. 335 FISHERY BULLETIN: VOL. 70, NO. 2 The occurrence of the threadfin shad at Sapelo Island may have resulted from recent introduc- tion of this species in the Altamaha and Savan- nah River drainages by the Georgia Game and Fish Commission. Movement to the north through estuarine waters is also possible. This species was taken in water as saline as 29.8%o and over a temperature range of 26.0° to 30.6°C. The scaled sardine (Harengula pensacolae) and Atlantic thread herring (Opisthonema oglin- um) were occasionally collected in the higher salinity waters of the beach, lower reach, and high marsh habitats. The Spanish sardine (Sardinellu anchovia) was reported from St. Simons Beach by Miller and Jorgenson (1969). Engraulidae - anchovies Of four species of anchovies known from Georgia, only the striped anchovy (Anchoa hep- setus) and bay anchovy (A. mitchilli) were common and widely distributed in the estuary. The bay anchovy was one of the most abundant fish species in trawl and seine collections. It was often collected in fresh waters of the oligo- haline creek and upper and middle reaches, but was absent from protected waters with low sa- linity. The bay anchovy was present through- out the year at a temperature range of 7.9° to 32.0°C. The striped anchovy was found in few- er habitats than the bay anchovy, and it occurred only during warmer seasons, May to November, at a temperature range of 15.7° to 30.8°C. Two tropical species {A. cubana and A. lyole- pis) were collected at St. Simons Beach (Miller and Jorgenson, 1969) but were not collected at Sapelo Beach. ORDER SALMONIFORMES Esocidae - pikes The pickerel (Esox americanus) was collected at five freshwater stations. On 16 December 1969, one specimen was trawled in lower Rice- boro Creek where the salinity was O.S%c and the temperature, 7.4 °C. Umbridae - mudminnows The little-known eastern mudminnow {Umbra pygmaea) was collected at freshwater station 6, the Ogeechee River tributary. This Coastal Plain species is rare in collections from Georgia waters. ORDER MYCTOPHIFORMES Synodontidae - lizardfishes Only one species (Synodus foetens) of the lizardfishes ranges inshore in Georgia estuaries. Six specimens were collected by trawls in 1967 and one by a seine in 1969 in the lower and mid- dle reaches from June to November when the salinity was 16.7 to 32.0;^c and temperature was 16.3° to 31.5°C. ORDER CYPRINIFORMES Cyprinidae - minnows and carps Three species of the freshwater cyprinids were collected. Some large carp {Cypyinus carpio) were caught in the lower Atlamaha Riv- er during the spring and some apparently moved to Doboy Sound and off Sapelo Beach with the flood waters that reduced salinities to 6.8%f along Sapelo Beach. Three large carp were found dead on Sapelo Beach in April 1970. Some sur- vived in a low-salinity tidal creek that was tem- porarily dammed on Sapelo Island until June when they were found dead. The alien carp is not known from the North Newport River head- waters. The golden shiner {Notemigonus crysoleucas) was common in the freshwater habitat, and three juveniles (20-24 mm) were collected in the oligohaline creek on 4 September 1969 when the salinity was 0%o and temperature was 26.4°C. The golden shiner is a common bait and forage fish in Georgia. The little-known taillight shiner (Notropis maculatus) was taken only at freshwater sta- tion 4. Catostcmidae - suckers The lake chubsucker (Erimyzon sucetta) was collected at freshwater stations 1, 4, and 6. 336 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES ORDER SILURIFORMES Ictaluridae - freshwater catfishes Three species of freshwater catfishes were collected. The yellow bullhead (Ictalurns na- talis) was common in freshwater stations. A 16-mm tadpole madtom {Noturus gyrinus) wandered downstream from its typical fresh- water habitat to the Crossroads station where the salinity was 1.9%c. The white catfish (/. catus) ranged down the estuary to the middle reach (trawl stations 10 and 11) when the sa- linity was reduced to 11.7 and 9.2%o and temper- atures were 11.1° and 10.8°C, in January and February. White catfish were seined and trawled (10-ft trawl) on nine occasions in the oligohaline creek at salinities less than l%o and at a temperature range of 7.4° to 28.4°C. Large samples including specimens over a length range of 40 to 300 mm, trawled in January and Feb- ruary 1970, suggest that large numbers of white catfish moved into the oligohaline creek when temperatures were 8.6° and 10.0°C. The low temperatures could also reduce their mobility and ability to avoid the trawl. Ariidae - sea catfishes The sea catfish {Arius felis) and gafftopsail (Bagre marinus) were most abundant in the lower reach of the estuary but also occurred in the middle reach and beach habitats. Most of the sea catfish (Arius) migrated to the ocean in the late autumn and winter and returned to the estuary and beaches in the spring (Table 3) . Males orally incubated the marble-sized eggs in June and July. Young with and without yolk sacs were incubated in August until they were 42 mm long or longer. Young lost the yolk sac when they were 37 to 40 mm long. The gafftop- sail also is an oral incubator, but none of the few adults collected contained eggs or young. This species is common in the warm months but scarce in the winter. ORDER PERCOPSIFORMES Amblyopsidae - cavefishes The rare swampfish (Chologaster cornuta) was collected at freshwater station 7. Table 3. — Length-frequency distribution of the sea cat- fish, Arius felis, collected with a trawl, 1967. There was no June collection. Length Apr. May July Aug. Sept. Oct. Nov. mm 25-30 4 31-35 31 36-40 104 4 41-45 123 19 46-50 90 65 51-55 45 93 8 56-60 10 72 26 61-65 2 55 92 1 66-70 1 11 92 2 71-75 1 3 1 3 51 76-80 3 3 3 22 81-85 3 5 86-90 2 24 91-100 27 2 1 101-110 7 20 111-120 1 53 2 121-130 1 37 16 131-140 7 7 16 I 141-150 7 1 3 151-160 10 8 2 161-170 2 13 3 171-180 10 7 8 181-190 11 3 6 191-200 11 9 6 1 200-210 9 5 3 211-220 5 6 2 221-230 3 1 2 231-240 3 2 241-250 3 251-260 2 261-280 2 Aphredoderidae - pirate perches The pirate perch (Aphredoderus sayanus) was collected at freshwater stations 3, 4, 5, and 7, and one specimen was taken in fresh water in the oligohaline creek. ORDER BATRACHOIDIFORMES Batrachoididae - toadfishes The oyster toadfish (Opsanus tau) was often collected in small numbers in trawls in the lower and middle reaches of the estuary throughout the year. This toadfish is sometimes seen in oyster reefs, and one specimen was found in a fouling community on the underside of a floating dock. The collection of only one specimen in the beach habitat can be attributed to the lack of cover. Salinity ranged from 12.4 to 32.0%o and temperature, from 28.4° to 30.8 °C. 337 FISHERY BULLETIN: VOL. 70. NO. 2 ORDER GOBIESOCIFORMES Gobiesocidae - dingfishes The habitats of the small skilletfish (Gobiesox strumosiis) are similar to those of toadfish. It is usually associated with oyster reefs or bottoms that provide cover, especially shell bottoms. The skilletfish occurred in small numbers in the lower and middle reaches throughout the year. ORDER LOPHIFORMES Antennariidae - frogfishes Two species of frogfishes (Table 2) collected at St. Simons Beach (Miller and Jorgenson, 1969) are stragglers from offshore. The sargas- sumfish {Histrio histrio) is of ten associated with sargassum weed, which drifts onto Sapelo Beach. ORDER GADIFORMES Gadidae - codfishes The southern hake (Urophycis floridanus) and spotted hake {U. regius) were common in the trawl collections in the lower and middle reaches from January to May. These southern representatives of a group that inhabit cold water first entered the estuary when the tem- perature was 8.2° to 10.1°C in January, and they remained until it rose to 24.3° to 24.6°C. Sa- linity and temperature ranges were 12.4 to 30.6/^c and 9.0° to 24.6°C for U. floridanus, and 14.5 to Sl.V/cc and 8.2° to 24.3°C for U. regius. Miller and Jorgenson (1969) also reported four U. regius from St. Simons Beach. Ophidiidae - cusk-eels This family is represented by one species, the striped cusk-eel (Rissola marginata) , in Georgia estuaries. Small numbers of this burrowing spe- cies were collected throughout the year by trawl- ing in the lower and middle reaches at salinities of 11.6 to 32.6/^f and temperatures of 8.0° to 30.0°C. ORDER ATHERINIFORMES Belonidae - needlefishes Several needlefishes range to the Georgia coast, but only the Atlantic needlefish {Strongy- lura marina) is common in Georgia estuaries. It is occasionally seen around docks in the lower and middle reaches and specimens were seined at the beach, high marsh, and middle reach ha- bitats. Specimens (52-315 mm) were collected from May to October at 22.9° to 30.9 °C. Cyprinodonidae - killifishes These small fishes are found almost entirely in shallowwater habitats. The euryhaline sheepshead minnow {Cyprinodon variegatus) was present throughout the year at salinity and temperature ranges of 0.8 to 34.0^f and 7.9° to 31.9°C in the tidal pools, beach, high marsh, and tidal canal habitats. The golden topminnow (Fundulu^ chrysotu^) was collected at freshwater stations 3 and 5. The marsh killifish (F. confluentus) occurred at freshwater station 1 and at Crossroads but was common only at an artesian well and in the low-salinity tidal pool where it occurred through- out the year. Recorded salinity and temperature ranges were 0 to 24.4^f and 7.8° to 29.3°C. The mummichog (F. heterocUtus) occurred at all the habitats except the freshwater habitat. Although it is a shallowwater species, two col- lections were made at trawl stations (10 and 12) that were close to the marsh. The mummichog was common at most seine stations but was rare along the beach. Recorded salinity and temper- ature ranges were 0 to 34%. and 7.8° to 32.2 °C. The southernmost record of the spotfin killi- fish {F. luciae) was reported by Miller and Jor- genson (1969). The striped killifish (F. majalis) was pre- sent at the Sapelo Beach throughout the year, but was scarce during the coldest months. It was also common in the high marsh and high- salinity tidal pools, and uncommon in the tidal ditch. The striped killifish was not taken at sand bars in the middle reach although it was col- 338 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES lected at low salinities. Salinity and tempera- ture ranges were 6.8 to 34.0%. and 7.0° to 32.2°C. The last three cyprinodonts listed on Table 2 are freshwater species characteristic of the Coastal Plain. The bluefin killifish {Lucania goodei) is also common in fresh water on Sapelo Island. Poecilidae - livebearers The euryhaline mosquitofish (Gambusia af fin- is) occurred at seven habitats and was abundant at the freshwater, oligohaline creek, and low-sa- linity tidal pool stations. It was collected on the beach only when the salinity was 6.8%o, but it was not uncommon at higher salinities. Salinity and temperature ranges were 0 to 34%p and 7.3° to 30.7°C. The least killifish {Heterandria formosa) is a freshwater species that occasionally wanders into estuarine waters. The highest salinity rec- ord was 4.0%o. The sailfin molly (Poecilia latipinna) was common in the shallow protected waters of the tidal pools, tidal ditch, and an artesian well on Sapelo Island. It occasionally occurred in the high marsh, beach, and oligohaline creek habi- tats. Temperature and salinity ranges were 7.3° to 32.2°C and 0 to 34%t>. Atherinidae - silversides The rough silverside {Membras martinica) was common at the beach and high marsh habi- tats. It was attracted to artificial lights at night and collected in large numbers with a dip net in the lower reach. Because of their pelagic nature and small size, only one was collected in a trawl. The Atlantic silverside (Menidia menidia) was found in the four high salinity habitats and in the middle and upper reaches of the estuary. This silverside was common only at high salin- ities, but it occurred at a sand bar in the middle reach when the salinity was 0.9%. on 4 September 1969. Although collected at a temperature range of 7.0° to 31.5°C, this silverside left the sand bars of the middle and upper reaches at temper- atures below 12°C, The tidewater silverside {Menidia beryllina) was common in the freshwater and oligohaline creek habitats at salinities of 0 to 7.9%c. It ranged to the middle reach when the salinity was 0.1%f. One was collected at Sapelo Beach when the sahnity was 6.8 to 7.7%.; this was the only occasion the two species of Menidia were col- lected together although their habitats and sa- linity tolerances overlap. ORDER GASTEROSTEIFORMES Syngnathidae - pipefishes and seahorses One lined seahorse {Hippocampus erectus) was collected by trawl in Johnson Creek (station 4) in May 1969 at a salinity of 27.1%. and a tem- perature of 22.3°C. The northern pipefish {Syngnathus fuscus) was taken sporadically by trawl and seine throughout the year. It occurred in four habitats at a salinity range of 0 to 31.3%c. The chain pipefish {Syngnathus louisianae) was uncommon in all six habitats where it was found. It was taken by trawl and seine through- out the year, but was less common than the northern pipefish. The chain pipefish was col- lected only 12 times and at a salinity range of 0.7 to 31.6%o. ORDER PERCIFORMES Centropomidae - snooks The snook {Centropomu^ undecimalis) was represented by young (23-81 mm) collected in protected waters of the tidal pools and tidal ditch. These were collected from June to No- vember at salinity and temperature ranges of 0 to 22.1%. and 23.0° to 28.6°C. Linton and Ric- kards (1965) collected 64 juveniles (24.1- 74.9 mm long) at Sapelo Island in 1963 and 1964. Their low temperature record of 18°C occurred in November. Serranidae - sea basses Of the diverse serranids found on the Georgia coast, only two species range inshore to the estu- ary. Young of the black sea bass {Centropristis striata) and rock sea bass (C. philadelphica) 339 FISHERY BULLETIN: VOL. 70, NO. 2 were collected throughout the year in the lower reach, mostly over shell bottoms at trawl sta- tions 3 to 6. Centrarchidae - sunfishes Twelve centrarchids (Table 2) were common in the creeks of the Coastal Plain. Six species (Centrarchus macropterus, Lepomis auritus, L. gulosus, L. macrochirus, L. piinctatus, and Mi- cropterus sahnoides) were also collected in the oligohaline creek. Although sunfishes have marked tolerance for salinity (Bailey, Winn, and Smith, 1954), none of these were found at a salinity above 0.5^c. Percidae - perches The swamp darter (Etheostoma fusi forme barratti) was collected only at freshwater sta- tions 3, 4, and 5. Pomatomidae - bluefishes The bluefish (Pomatomus saltatrix) is often caught by anglers in the beach and lower reach habitats but was rarely taken in this study. Only four young (115-196 mm) were collected, one in each season of the year. Echeneidae - remoras A sharksucker (Echeneis naucrates) is re- corded for the lower reach and beach habitats since one was attached to a lemon shark that was caught from the beach in Doboy Sound (Dahl- berg and Heard, 1969). Carangidae - jacks and pompanos Nine species of carangids occurred primarily in the beach and lower reach habitats. The horse- eye jack {Caranx latus) is represented by four specimens from St. Simons Beach (Miller and Jorgenson, 1969). The crevalle jack (C. hippos) was occasionally caught by anglers in the lower reach of the estuary and a few juveniles were seined as far up the estuary as the upper reach when the salinity was 10.8/^f . Five small spec- imens (24-77 mm) were collected in the summer and autumn. The Atlantic bumper (Chloroscombrus chry- surus) and leather jacket (Oligoplites saurtis) were mostly caught at the high-salinity stations. The lowest salinity recorded for the leather jack- et was 16.S%c. Both occupied shallow waters and the bumper was also caught in trawls. The lookdown (Selene vomer) and Atlantic moonfish (Vomer setapinnis) were found in the beach habitat generally from May to November. Lookdowns were occasionally trawled in the sounds. Young of the commercially important Florida pompano (Trachinotus carolinus) and young permit (T. falcatus) were common in the beach habitat and occasionally wandered to the high marsh. Young pompano were present from spring to autumn. The palometa (T. goodei) was present in summer and autumn at St. Si- mons Beach (Miller and Jorgenson, 1969), and the permit followed the same pattern at Sapelo Beach. Lutjanidae - snappers The gray snapper (Lutjanus griseus) spawns offshore, and the young have occasionally been collected in the beach, high marsh, and tidal pools habitats. This primarily tropical species was collected from August to November. The low- est salinity recorded for this species was 13.1%o. Gerreidae - moj arras Three species of moj arras, Irish pompano (Diapterus olisthostomus) , spotfin mojarra (Eucinostomus argenteus) , and flagfin mojarra (E. melanopterus) , were collected. The flagfin mojarra was represented in one collection that contained all three species. This unusual collec- tion was at a sand bar in the oligohaline creek on 22 October 1969 when the salinity was 0.7%o and the temperature was 24°C. The Irish pom- pano was represented by juveniles collected at six shallowwater habitats. They occurred from July to November at salinities and temperatures of 0.7 to 31.3%f and 19.4° to 31.8°C. The spotfin mojarra was collected in seven habitats by seining and was the only mojarra collected (twice) by trawling. This species was 340 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES taken from July to November at salinities and temperatures of 0.7 to 31.3%r and 12.7° to 31.1°C. Dr. C. L. Hubbs identified this species in two collections from Sapelo Beach and Riceboro Creek, and I believe that neither of the closely related species, E. gula or E. jonesii, occurred in my collections. Pomadasyidae - grunts The primarily tropical grunts are represented by the pigfish (Orthopristis chrysoptera) in Georgia estuaries. The pigfish was trawled in the lower and middle reaches from June to De- cember and was also collected at the beach. A minimum salinity of 15.4%^ was recorded. Sparidae - porgies The sheepshead {Archosargus probatoceph- alus) provides an important sport fishery around docks in the lower reaches and also occurred in the high marsh and beach habitats. The small pinfin (Lagodon rhomboides) was rare but wide- spread (six habitats). Small pinfish (20-87 mm) were present from May to September, and most were collected in the intertidal pools. Sciaenidae - drums This is the most important family of fishes in Georgia estuaries to sport fishermen. Sciae- nids are the most numerous fishes in terms of numbers available to trawls (Anderson, 1968), and probably most abundant in terms of biomass in trawl collections. Silver perch (Bairdiella chrysura) occupied a variety (8) of habitats. Adults were common only in trawl collections in the lower and middle reaches. The silver perch spawned primarily in April and May (Table 4) . The smallest spec- imens were collected in the lower reach and high-salinity pools in May and June. Two age groups (Table 4) were distinct from May to July. Young grew rapidly from May to October in a wide variety of habitats. They apparently use the whole estuary and also the beach waters as a "nursery ground." Salinity and tempera- ture ranges for this species were 1.3 to 34. 1%^ and 7.5° to 32°C. The spotted seatrout (Cynoscion nebulosus) are among the most important estuarine fishes to Georgia anglers. They were occasionally trawled in small numbers in the lower and middle Table 4. — Length-frequency distribution of the silver perch, Bairdiella chrysura, collected by trawling and seining, 1967-68. Length Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Jan. Feb. mm 16-20 3 21-25 13 6 26-30 92 6 31-35 20 14 10 36-40 8 46 4M5 1 85 2 46-50 1 85 8 51-55 3 66 9 2 56-60 43 10 1 61-65 34 15 4 5 66-70 8 9 7 15 1 71-75 4 6 7 35 1 2 76-80 4 13 48 3 2 2 81-85 1 2 4 14 64 5 4 3 5 86-90 1 2 1 14 65 6 5 3 4 91-95 3 7 2 1 6 60 6 4 10 4 96-100 2 1 9 2 1 3 52 4 13 9 1 101-105 1 8 20 5 36 2 8 5 106-1 10 4 6 3 I 14 1 1 111-115 2 7 1 4 2 116-120 1 1 2 1 1 121-125 1 126-130 1 131-135 1 136-140 1 341 FISHERY BULLETIN: VOL. 70, NO. 2 Table 5. — Length-frequency distribution of weakfish, Cynoscion regalis, collected by trawling, 1967-68. None were collected in March. Length Jan. Feb. Apr. May July Aug. Sept. Oct. Nov. Dee. Jan. mm 11-20 4 2 2 21-30 7 42 25 18 31-40 27 93 66 18 6 41^0 12 150 55 7 24 51-60 1 122 24 5 30 61-70 1 151 17 5 50 1 1 71-80 1 157 16 9 61 3 81-90 1 1 141 21 4 66 18 1 91-100 1 122 28 1 36 24 4 2 101-110 5 1 2 1 95 29 22 18 7 111-120 1 4 50 41 5 14 5 1 121-130 1 1 2 11 18 35 9 2 3 1 131-140 1 7 13 25 8 1 141-150 2 16 6 10 3 1 151-160 9 4 1 161-170 3 1 171-180 1 4 1 181-190 1 191-200 1 1 201-210 I 211-220 1 221-230 231-240 1 reaches throughout the year. They were rarely seined in shallowwater habitats although many are caught with fishing poles along the beaches and salt marshes. Spotted seatrout are known to spawn in and spend their whole life in the estuary (Tabb, 1966). Juveniles were found as far up the estuary as the upper reach at a salinity of 0.5%c. During 1967-69, silver seatrout (Cynoscion nothus) entered the lower reach of the estuary in May and stayed until July or August. Weakfish (Cynoscion regalis) apparently spawned from April to August (Table 5) . With the exception of May samples, age groups are difficult to recognize because of the protracted spawning season. Young weakfish were collect- ed in six habitats. Adults and young were abun- dant only in trawl collections in the lower and middle reaches. Although most abundant in high-salinity waters, young weakfish occurred at salinities as low as &.&%[ in the upper reach. Weakfish were conspicuously scarce in the cold months, December to April. The banded drum (Larimus fasciatus) was occasionally collected in the lower reach through- out the year and was collected three times along the beach. It was restricted to high salinities, 22.0 to 34.1%o. Length frequencies of the spot (Leiostomus xanthurus) were based on trawl and seine col- lections (Table 6). Adults were common in deeper waters and juveniles dispersed to eight shallowwater and deepwater habitats. Young spot (11-85 mm) were among the most numer- ous fishes of the oligohaline creek where they were collected seven times from April to July at a salinity range of 0.2 to S.l%c. Young were also common in the tidal pools and tidal ditch. Seine collections indicated that the spot spawned primarily from January to April. Two age groups were distinct from February to May. The three species of kingfishes (Menticirrhus) have marked similarities and differences in their ecologies. All three occurred in the beach and lower reach habitats and were rare in the high marsh. Young southern kingfish (M. america- nus) were also taken in the middle and upper reaches. It is the only kingfish that was found in low salinities. Four juveniles (19-36 mm) were collected in the upper reach in July and August at salinities of 1.5 to 7.9%c, much lower than the low-salinity limits observed by Gunter (1961). Seining along Sapelo Beach took young south- ern kingfish that had been spawned primarily from April to August (Table 7). Young and 342 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES Table 6. — Length-frequency distribution of the spot, Leiostomus xanthurus, collected by seine, 1967-69 (in paren- theses) and by trawl in 1967. None were collected in September. Length Jan. Feb. Mar. Apr. May Juna July Aug. Oct. Nov. Dec. mm , 11-15 (2) (3) (3) 16-20 (10) C14) (2) 21-25 (6) (30) (4) (1) 26-30 (12) (3) (1) 31-35 (2) (6) (4) 36-40 (1) (4) 1 (10) 41-45 (2) 3 (11) (2) (7) 46-50 (1) 2 (10) (5) 51-55 (1) 3 (9) (3) 2 (9) (2) 56-60 (5) 2 (9) (3) 5 (13) 61-65 (4) (5) 6 (6) (1) 66-70 4 2 (3) (3) 7 (3) 1 71-75 11 10 (1) (1) 4 (3) 76-80 12 14 2 (1) 2 (3) 5 3 81-85 17 12 4 (2) 2 (1) 86-90 18 14 3 2 I 2 (1) 91-95 14 18 1 2 2 2 96-100 19 16 3 4 2 2 1 101-105 23 18 9 2 4 1 2 I 106-110 14 7 7 4 7 1 2 111-115 21 8 7 2 8 2 116-120 13 8 11 12 1 121-125 9 8 7 3 9 1 126-130 10 5 9 2 10 2 1 131-135 5 5 10 5 9 2 2 1 136-140 1 2 5 2 9 2 1 141-145 2 4 3 9 5 1 1 1 146-150 2 1 1 4 6 2 3 151-155 7 1 156-160 1 1 2 161-165 1 2 1 1 166-170 1 adults also occupied the lower and middle reach- es. Although generally common, adults were rare in March and April apparently because they migrate to the ocean where they spawn in off- shore waters (Bearden, 1963). Gulf kingfish (M. littoralis) spawned primar- ily from April to September, judging from 18- to 22-m specimens that were taken along the beach from May to October. The Gulf kingfish was common only along the beach from May to November in my collections. The northern kingfish (M. saxatilis) was the least common of the kingfishes. Young (17- 50 mm) occurred along the beach in April and May, and five larger specimens (58-140 mm) were trawled in the lower reach from May to August. The Atlantic croaker (Micropogon undalatus) somewhat resembles the spot (Leiostomus xan- thurus) in population size and distribution in the estuary. Table 8 indicates that croakers spawned primarily from September to April. Bearden (1964) found that in South Carolina croakers spawned almost entirely in the ocean, and larvae were found from October through May. Young croakers (16-80 mm) were col- lected in seven habitats, including six shallow- water habitats, at salinities down to 2.7%c. Un- like spots, young croakers were not abundant in shallow water and they did not occur in fresh water. Adult croakers were common in trawl catches in the lower and middle reaches, espe- cially May through August when they were not spawning. The population decline in autumn 343 FISHERY BULLETIN: VOL. 70, NO. 2 Table 7. — Length-frequency distribution of the southern kingfish, Menticirrhus americanus, collected at Sapelo Beach by seine, 1967-68 (parentheses) and by trawls, 1967. Length Jan. Feb. Mar. Apr. May Juna July Aug. Sept. Oct. Nov. Dec. mm 11-15 1 (4) (2) 16-20 1 (33) (1) 2 (2) (1) 1 (15) 21-25 1 (48) (13) 2 (1) 1 (1) 4 (23) 26-30 (14) (24) 19 (14) 21 (2) 11 (7) 31-35 (2) (45) 38 (28) 49 (2) 12 (3) (1) 36-40 (19) 48 (33) 41 11 6 (2) 1 41-45 1 (2) 30 (11) 30 5 18 1 46-50 39 (3) 26 9 (1) 27 51-55 (1) 15 (1) 15 (2) 28 6 21 2 1 56^0 (2) 29 9 17 3 1 61-65 1 (1) 14 20 8 14 5 1 66-70 2 (1) 15 (3) 16 (2) 19 21 7 11 3 ■ 4 71-75 1 2 12 4 10 8 7 76-80 4 3 12 4 7 7 13 81-85 2 6 1 18 16 10 9 6 5 86-90 3 12 1 13 4 4 14 7 4 91-95 4 7 11 9 2 10 6 3 96-100 4 13 7 10 2 5 6 5 101-105 4 6 11 6 2 11 5 2 106-110 I 8 1 1 6 1 6 4 111-115 1 5 2 2 2 10 4 4 116-120 1 6 1 2 1 1 4 5 2 121-125 1 2 1 1 6 I 126-130 1 5 2 1 2 4 2 2 131-135 2 3 2 4 136-140 2 1 1 3 1 I 141-150 3 1 1 4 1 1 3 2 2 151-160 2 1 1 2 2 1 5 2 161-170 1 2 1 I 171-180 2 2 1 181-190 1 1 1 1 2 191-200 1 2 3 201-210 1 1 1 211-220 I I 221-230 2 1 241-250 1 1 was the result of seaward migration for spawn- ing. Bearden (1964) found ripe croakers 3 to 30 miles offshore. Correlation of movements with spawning is complicated by the protracted (8 months) spawning period. Large black drums (Pogonias cromis) are occasionally caught by fishermen in the lower reach. I collected only a few black drums, all small (19-130 mm), mostly in the high-salinity pools. The red drum {Sciaenops ocellata) is one of the most popular estuarine sport fishes because of its large size and abundance in the beach and lower reach habitats. The red drum was taken in only six seine collections and no trawl collec- tions. The smallest (36 and 37 mm) specimens occurred in November. The star drum (Stellifer lanceolatus) was the most abundant species in the lower reach habi- tat in 1967 (Dahlberg and Odum, 1970) an^ also in estuaries near Brunswick in 1933-35 (Ander- son, 1968). It is a small species and may ac- count for less biomass than the spot or croaker. Young of the year, apparently spawned from May or June to September (Dahlberg and Odum, 1970), accounted for most of the numbers. 344 I DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES Table 8. — Length-frequency distribution of the Atlantic croaker, Micropogon undulatus, 1967 and January and Feb- ruary 1968. Most were collected with a trawl. Seine and cast net collections are indicated by parentheses. Length 1967 1968 Jan. Feb. Mar. Apr. May July Aug. Sept. Oct. Nov. Dec. Jon. Feb. mm 16-20 (5) (1) 21-25 1 (I) (9) 26-30 3 1 (6) 31-35 1 2 (6) 36-40 2 2 2 (1) 41-45 8 1 2 2 4(W0 1 10 1 4 1 51-55 2 17 1 1 1 1 1 1 56-60 1 1 7 1 1 1 2 61-65 2 2 8 5 5 1 1 66-70 1 20 7 11 1 71-75 1 4 17 6 2 76-80 2 2 \ 1 23 13 5 1 81-85 1 2 25 12 2 8 1 1 1 86-90 1 1 8 18 3 2 2 1 3 91-95 3 7 43 3 5 1 3 3 1 96-100 I 7 47 7 1 2 1 101-105 4 70 6 1 1 106-110 81 10 4 111-115 1 74 13 1 1 S 116-120 2 1 50 9 2 1 A 121-125 1 1 1 41 10 1 1 1 126-130 1 2 16 5 2 1 131-135 2 1 4 4 1 1 2 136-140 3 3 1 141-145 3 1 146-150 1 151-170 1 1 Most star drum left the estuary during the cold months. Kyphosidae and Pomacentridae These two families were represented only by stragglers from offshore (Table 2). They were collected at St. Simons Beach (Miller and Jor- genson, 1969). Ephippidae - spadefishes Small Atlantic spadefishes (Chaetodipterus faher) were common in trawls in the lower and middle reaches from June to October at a tem- perature range of 20.1° to 32.0°C. They were uncommon in shallowwater habitats and were not collected at salinities below 9.9%o. Mugilidae - mullets Striped mullet (Mugil cephaliis) and white mullet (M. curema) were widespread and abun- dant in the estuary. Both species generally oc- cupied shallow water near the surface. The striped mullet was collected in eight shallow- water habitats, and it often enters inland rivers. The white mullet occupied the same estuarine habitats with the exception of the oligohaline creek. This difference may be attributed to sa- linity preferences since the white mullet was taken at salinities down to 5.0%o and the striped mullet was often collected at salinities below 0.5%c. The striped mullet was common or abundant in all the habitats in which it occupied. The white mullet was common in the beach and high- salinity pool habitats. The striped mullet ap- parently has a greater temperature tolerance since it was collected throughout the year at a temperature range of 7.0° to 31.7°C. The white mullet was absent from collections from Jan- uary through March and occurred at a temper- ature range of 15.0° to 32.2°C. Length frequencies indicated that the striped mullet spawned from September through April and the white mullet from March through Sep- tember, with some overlap in March, April, and September. 345 FISHERY BULLETIN: VOL. 70. NO. 2 Sphyraenidae - barracudas Young of two species of barracudas occurred in the beach habitat. One guaguanche (Sphy- raena guachancho) was collected at Sapelo Beach in October 1967. Three southern sennets (S. piciidilla) were collected at St. Simons Island Beach in May (Miller and Jorgenson, 1969). Uranoscopidae - stargazers The southern stargazer (Astroscopus y-grae- cum) was occasionally trawled in the lower and middle reaches. Young were seined along the beach and in the upper and middle reaches at salinities as low as 12.5%o. Blennidae - combtooth blennies Three species of blennies (Table 2) are com- monly associated with oyster reefs or patches of oyster shells in the lower reaches of the es- tuary. The feather blenny (Hypsoblennitis hentzi) was occasionally trawled in the lower and middle reaches and one crested blenny (Hy- pleurochilus geminatus) was trawled in the low- er reach. All three species occasionally occurred along the beach (Miller and Jorgenson, 1969) generally in association with shell or debris. My data suggest that the feather blenny oc- curred primarily in deep water in the cold months and that they migrated to the oyster reefs in the warm months where they rear their eggs inside of gaping oyster shells. Three male striped blennies (Chasmodes bosquianus) , 58 to 64 mm, were rearing embryos inside of gaping oyster shells on 24 April 1970 and 19 May 1970 when water temperature was 27° to 27.5°C. Eleotridae - sleepers The fat sleeper {Doi^mitator maculatus) was collected only in the low-salinity tidal pools from May through November in 1967. Gobiidae - gobies The lyre goby {Evorthodus lyriciis) was col- lected once in the high-salinity pools previous to my collections. All four species of Gohioyielliis were rarely encountered. The darter goby (G. boleosoma) was collected three times in the high- salinity tidal pools and was rarely found in the beach or high marsh habitats. Only five spec- imens of the sharptail goby (G. hastatus) were collected — three in trawls, one in the Marine In- stitute's boat basin, and one in the high-salinity pools. This goby was found only in September 1967 and November 1969. The freshwater goby (G. shufeldti) occurred in the low-salinity tidal pools from May 1967 to February 1968. This goby was also reported in the high-salinity waters of St. Simons Beach and in fresh water of Altamaha River (Miller and Jorgenson, 1969). The emerald goby (G. smaragdus) and green goby (Micro gobius thalassinus) are two little- known species that were collected only in the high-salinity tidal pools. The former occurred there only in September and November and the latter only in May and November. Two species of scaleless gobies {Gobiosoma) were most abundant in oyster reefs and patches of oyster shells where they laid their eggs inside gaping oyster shells during the warm months. The naked goby (G. bosci) occurred in six hab- itats in addition to oyster reefs and was the most abundant goby in the estuary. It was col- lected by hand in gaping oyster shells and by seining, and three were trawled. The salinity range was from fresh water to 30.8%c.. Only juveniles (18-22 mm) were found at salinities of less than 2%c. The seaboard goby {Gobiosoma ginsburgi) was not collected at regular stations except for specimens from the lower reach that were taken from the stomachs of hakes ( Urophycis) . Males were found rearing embryos inside of gaping oyster shells. On 4 March 1969, a large number of seaboard gobies and a few naked gobies were found in burrows in an eroding clay bank at Sapelo Island Beach. Trichiuridae - cutlassfishes The Atlantic cutlassfish (Trichiurus lepturus) was occasionally trawled throughout the year in the lower reach but was rare along the beach. 346 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES Scombridae - mackerels and tunas King and Spanish mackerels are often caught offshore by sport fishermen during the warm months. Young Spanish mackerel (Scomberom- orus maculatus) occasionally ranged into the high marsh, beach, and lower reach habitats, but they were rarely collected by seine or trawl be- cause of their speed. One king mackerel {S. cavalla) was also collected along the beach (Mil- ler and Jorgenson, 1969). Stromateidae - butterfishes The butterfish (Peprilus triacanthus) and harvestfish (Peprilus alepidotus) are primarily marine fishes that occasionally occur in trawl catches in the middle and lower reaches and in beach seine hauls. They have a wide temper- ature tolerance as they were collected through- out the year. Butterfish did not occur at salin- ities below 19.5%o and harvestfish below 22.S%c. Triglidae - searobins Four species of searobins {Prionotus carolin- us, P. evolans, P. scitulus, and P. tribulus) were present in trawl collections in the lower reach throughout the year, and all but P. carolinus were trawled in the middle reach. Three of these, excluding P. evolans, were occasionally taken in the beach habitat (Miller and Jorgen- son, 1969). A salinity of 15.3%c for P. evolans was the lowest I found occupied by searobins. One P. salmonicolor was collected in Sapelo Sound. ORDER PLEURONECTIFORMES Bothidae - lefteye flounders The lefteye flounders were represented by eight species that were found only in habitats 1 to 5 (Table 2) . Species of Paralichthys reach a large size and two of them are important sport and commercial fishes whereas the four other bothids are small. The ocellated flounder (An- cylopsetta quadrocellata) and windowpane (Scophthalmus aquosus) migrated into the low- er and middle reaches during the winter and spring, December or January to May. Water temperature ranges were 8.0° to 26.0 °C for the ocellated flounder and 8.8° to 25.7° for the win- dowpane. These species were seasonally re- placed by the bay whiff (Citharichthys spilop- terus) , which was occasionally trawled in the lower and middle reaches from May to October at a temperature range of 26.0° to 31.5°C. Strag- glers also occurred in the beach and oligohaline creek habitats. The fringed flounder (Etropus crossotus) was common in trawls in the lower reach throughout the year. There are records from the beach hab- itat, and a 16-mm specimen was seined in the upper reach when the salinity was 0.5%o. Two species of Paralichthys occur primarily offshore and rarely move inshore. The gulf flounder (P. alblgutta) was trawled only three times. The broad flounder (P. squamilentus) was represented by only one juvenile from the beach habitat. The summer flounder (P. dentatus) and south- ern flounder (P. lethostigma) enter the com- mercial and sport fisheries of the coast. The summer flounder was most abundant in the lower reach and was rarely trawled in the middle reach. The southern flounder was much more abundant at the middle reach stations than in the lower reach. Its tolerance or preference for lower sa- linities was also demonstrated by its distribution up the estuary to Riceboro Creek where it was often collected in fresh water. Soleidae - soles Thehogchoker (Trinectes maculatus) is a ma- rine or brackish-water fish that often spends considerable time and travels considerable dis- tances in fresh water. It was common from the lower reach to the freshwater habitat. The hog- choker was present throughout the year in Rice- boro Creek where the water was usually fresh. Cynoglossidae - tonguefishes The blackcheek tonguefish (Symphurus pla- giusa) was rarely collected outside of the lower and middle reaches where it was one of the most 347 FISHERY BULLETIN: VOL. 70, NO. 2 abundant species. A low-salinity record of 0.1%c was obtained from Riceboro Creek. ORDER TETRAODONTIFORMES Three marine species (in three families) in this order were represented mostly by small numbers of juveniles. Nearly all specimens were collected in the lower reach and beach hab- itats. One planehead filefish (Monacanthus hispidus) was recorded for a low salinity of 11.87ff. The planehead filefish was in the estuary from April to September at 21.8° to 31.3°C. The northern puffer {Sphoeroides maculatus) and striped burrfish {Chilomycterus schoepfi) were in the estuary from April to November or De- cember when the temperature was reduced to approximately 11° to 16°C. LIFE CYCLES OF ESTUARINE SPECIES Many fishes found in estuaries follow the ma- rine-estuarine life cycle pattern described by Gunter (1967). They spawn in the ocean and the young enter the estuarine and beach waters where the salinity is reduced. The estuary ap- parently provides them with a nursery ground that is rich in food and a refuge from certain predators, diseases, and parasites that do not thrive in the rigors of highly variable salinities and temperatures. Young of coastal species often have greater tolerance to reduced salinities than adults (Gunter, 1961). Young Atlantic menhaden even require low salinities for devel- opment (June and Chamberlin, 1959). Gunter (1967) noted: "The preponderant macroorga- nisms, both in numbers of species and individu- als, are mostly motile species which undergo the general type of life history described above. In southern waters these are the mullet (Mugil) , menhaden, croakers (sciaenids), shrimp and crabs. Vast numbers of these animals may be found in estuaries at one time or another and in general the very smallest sizes are found in the lower salinities." Some species that spawn in high-salinity waters but were represented pre- dominantly by young in two low-salinity estu- arine habitats (oligohaline creek and low-salini- ty tidal pools) included the spot, striped mullet, hogchoker, southern flounder, ladyfish, tarpon, and snook. The Atlantic menhaden, silver perch, and Atlantic croaker were euryhaline and rep- resented in the upper reach primarily by young. Since estuaries are being destroyed by pollu- tion, dredging, and filling, it is important to rec- ognize which species are found in estuaries as adults or young. Dependence of the young of marine species on the estuaries is the basis of the nursery ground concept (Gunter, 1967). I expand the nursery ground concept to include all species that are commonly represented by young (defined herein as sexually immature) in estuaries, whether they were spawned in the ocean, estuary, or fresh water. A total of 168 species is listed for coastal waters (Table 2). The number of species re- corded for the estuary is 136 when the numbers of species found only in the freshwater habitat (17) and only along the beach (15) are excluded. The beach and estuary are treated as a single complex here because of their similar fish species compositions. The estuary and beach complex functions as a nursery ground to various degrees for 78 species. Some young sharks and rays that were commonly caught by angling (Dahlberg and Heard, 1969) are included in the compilation. Two species that spawn in fresh water, the anadromous American shad and the white catfish, are also included. The family Sciaenidae includes 13 species that have young dependent on the rich estuarine waters. These sciaenids are the most important group of sport fishes on the Georgia coast and they are poten- tially important commercial fishes. Atlantic menhaden that are reared in the estuaries are caught off'shore in large numbers with purse seines. Vitally important in the food chain are forage species such as the bay anchovy, Atlantic silverside, and rough silverside. Some organisms that are indigenous to bay or estuarine waters, or at least normally complete their life cycle in these waters, include certain "copepods and planktonic species," several spe- cies of molluscs including the American oyster, certain gobioid and cyprinodontid fishes, and a palaemonid shrimp (Gunter, 1967). I have not found fish species that are restricted by salinity tolerances to estuarine waters throughout their 348 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES life cycle, but many normally complete their life cycle in the estuary. Some species that typically pass their complete life cycle in the estuary, at least in Georgia estuaries, include those species I later consider to be characteristic of the oyster reefs and certain cyprinodontid fishes such as the marsh killifish, spotfin killifish, and mum- michog. The spotted seatrout possibly belongs to this category but adults are common along the beach at times and northern populations mi- grate offshore in the winter (Tabb, 1966). There are species that typically complete their life cycle in the ocean or fresh water that are either regular or accidental visitors to the estuary. Another life cycle pattern is exhibited by the anadromous shads (Alosa) which "spawn in Georgia rivers, including the Altamaha River (Godwin and Adams, 1969). The American eel is the only catadromous species in the estu- ary. DIVERSITY OF COASTAL HABITATS Diversity of habitats is considered here in order to determine the importance of the var- ious habitats to the fish community and also to explore the possible relationship of diversity and stability. A simple index, number of species, is used here in a comparison of habitats (Table 9) . To be objective, I define stragglers as those species that are represented by only one collec- tion in a habitat. Further studies would be Table 9. — Number of species recorded for nine Georgia coastal habitats. To eliminate the influence of strag- glers, species recorded (collected or seen) once in the estuary are subtracted from total. Habitat Number of species Species recorded once Species recorded more than once 1. Beach 114 19 95 2. Lower reach 100 4 96 3. High marsh 56 17 39 4. Middle and upper reaches 61 12 49 5. Oligohaline creek 40 21 19 6. Freshwater creek 39 0 39 7. Tidal canal 22 7 15 8. Low-salinity tidal pool 22 4 18 9. High-salinity tidal pool 37 7 30 needed to confidently ascertain which species are naturally rare in their preferred habitats. Dis- counting stragglers removes a large percentage of the accidental species and makes diversity comparison more meaningful. These diversity values are only roughly comparable because of the differences in sampling effort and gear. The beach habitat produced the highest diver- sity— 114 or 95 species if stragglers are dis- counted. A high diversity of clupeids, carangids, sciaenids, and bothids accounted for 36 species. Tagatz and Dudley (1961) recorded only 40 fish species from a Beaufort, N.C., beach. Gunter (1958) recorded 44 species from Texas beach station, and Springer and Woodburn (1960) re- corded 48 from a beach near Tampa Bay. The latter consider the beach notable for harboring few species compared to other coastal habitats. A low diversity would be expected because the beach offers little niche variety or cover. The higher diversity I report may be attributed to several factors. A large number of species that are typical of other marine and estuarine habi- tats are occasionally found along the beach. An- other factor is the inclusion of eight shark spe- cies that were caught while fishing from the beach. A third factor is that Miller and Jorgen- son (1969) sampled more extensively than in other studies noted herein. They found 98 spe- cies at St. Simons Beach; this total includes 38 species which did not occur at Sapelo Beach. Species that I consider to be eurj^thermal in the estuary were collected in both winter and summer and usually in all four seasons, but not necessarily every month. Species that are eury- thermal in the beach habitat include the rough silverside (Miller and Jorgenson, 1969), Atlantic silverside, striped killifish, bay anchovy, and striped mullet. Some species that were abun- dant only in the warm months include the bump- er, pompano, white mullet, southern kingfish, and gulf kingfish. Numbers of species and indi- viduals were considerably reduced during the cold months. The lower reach ranked high in diversity. Most of its 100 species were caught in trawls but seven shark records are based on Dahlberg and Heard (1969). Since only four species are ranked as stragglers, diversity of the charac- teristic species is similar to the beach habitat. Trawl collections in the region of Cedar Key, Fla., yielded only 63 species (Reid, 1954). This 349 FISHERY BULLETIN: VOL. 70, NO. 2 lower diversity is partially the result of less sam- pling effort. Dahlberg and Odum (1970) noted the abun- dance of species at the 14 trawl stations over the first 14 months of this study. We found that the most numerous species for the first 12 months of the study were star drum (15,209 individu- als), weakfish (2,454), blackcheek tonguefish (2,193), sea catfish (1,681), southern kingfish (1,345), silver perch (1,133), bay anchovy (1,090), spot (1,004), Atlantic croaker (896), and spotted hake (467) . In shrimp trawl catch- es in an estuary near Brunswick, 1931-35, An- derson (1968) found the following order of decreasing abundance: star drum, Atlantic croaker, spot, fringed flounder, weakfish, sea catfish, anchovy species, gaflftopsail catfish, and kingfish species. In both studies the star drum was the most abundant species. Certain differ- ences (e.g., tonguefish, silver perch, fringed flounder, gaflFtopsail catfish) may be related to spatial or temporal changes in populations, size of trawl, and mesh size. Miller and Jorgenson (1969) recorded 56 spe- cies, including 39 collected more than once, for the high marsh habitat. Species that were eury- thermal in the high marsh also were character- istic of the beach habitat. However, they found the mummichog to be more abundant than the striped killifish in the marsh. Collections with trawls and seines both con- tributed heavily to the high diversity (61) of the upper and middle reaches. Some of the spe- cies trawled in the middle reach may not occur in the lower salinity upper reach. Species that were eurythermal in the middle reach and com- mon in trawl catches in the middle reach include the Atlantic stingray, bay anchovy, silver perch, spot, southern kingfish, Atlantic croaker, hog- choker, blackcheek tonguefish, and oyster toad- fish. Species that were common in trawl catch- es only during the warm months include the sea catfish, weakfish, star drum, and Atlantic spade- fish. Species that were common in trawl catches only during the cold months include the two hake species, spotted seatrout, ocellated flounder, and southern flounder. Gunter (1967) and others have pointed out that young of marine species are predominant in brackish water. Young fishes that I consider to be eurythermal in the shallow waters (col- lected by seine, 10-ft trawl, and cast net) of the upper and middle reaches include the Atlantic menhaden, silver perch, spot, southern kingfish, croaker, striped mullet, white mullet, southern flounder, hogchoker, and blackcheek tonguefish. The longnose gar, striped anchovy, bay anchovy, mummichog, tidewater silverside, Atlantic sil- verside, northern pipefish, and chain pipefish were also characteristic of this region. The oligohaline section of Riceboro Creek has a low diversity (40 species) especially when the stragglers (21 species) are considered. Tagatz and Dudley (1961) recorded 38 species including 12 freshwater species at an oligohaline station in the Neuse River, N.C. Fishes of the oligo- haline creek include 15 freshwater species, 20 euryhaline marine species, 4 anadromous species (Alosa and Dorosoma) , and 1 catadromous spe- cies (Anguilla) . Populations of Dorosoma spe- cies that occur in saline waters can be considered anadromous as Bailey et al. (1954) have done, but these are primarily freshwater species, at least in Georgia. Characteristic species of the oligohaline creek include the longnose gar^ bay anchovy, white catfish, mummichog, mosquito- fish, tidewater silverside, bluegill, striped mullet, hogchoker, southern flounder, and spot. Numbers of fish species decrease up the estu- ary until the stable freshwater habitat is reached. Most of the 39 species recorded for the freshwater habitat are freshwater species that are characteristic of Coastal Plain waters. Others are migratory or euryhaline marine spe- cies, including the anadromous shads (Alosa), catadromous American eel, tidewater silverside, northern pipefish, and hogchoker. Fishermen report catching the anadromous striped bass (Morone saxatiUs) in Riceboro Creek but we have no records. Miller and Jorgenson (1969) reported 48 spe- cies from a freshwater station in the lower Altamaha River. In addition to species that I re- port, they recorded some marine species, fresh- water fishes that usually occur in large rivers and reservoirs, and others that are probably absent from the North Newport River drainage. The tidal canal and low-salinity tidal pools are 350 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES quite similar in diversity and location but some interesting differences exist. The canal is more open to high-salinity water and to the marsh. Larger individuals of some species, such as the striped mullet, silver perch, and spot, entered the canal but not the more isolated pools. The pools contained small species and small individu- als of large species. Species characteristic of both habitats include the ladyfish, tarpon, sheeps- head minnow, mummichog, mosquitofish, sailfin molly, snook, striped mullet, white mullet, and spot. Species that were restricted to the pools include the marsh killifish, fat sleeper, and fresh- water goby. The greater diversity (37) of the high-salinity tidal pools, compared to the low-salinity pools, is probably related to their higher salinity and greater accessibility from habitats of high di- versity (beach and lower reach) . As noted pre- viously, salinities and temperatures were similar in the two series of high-salinity pools on the east and west sides of the road to the Sapelo Island lighthouse. Conspicuous faunal differ- ences in the two series of pools may be related to water depth. In the shallow pools of the east side there were large numbers of cyprinodonti- form fishes (including the sailfin molly, mum- michog, sheepshead minnow, mosquitofish, and striped killifish) and spotfin mojarra. These were all found in much smaller numbers in the deep pool on the west side of the road. The deep pool produced larger numbers of young silver perch, young spot, and gobies. The sharp- tail goby, emerald goby, and green goby were not found in the shallow pool. Large numbers of striped mullet, white mullet, ladyfish, and bay anchovy were found in the shallow and deep pools. These semi-isolated tidal pools of estuaries have 'received little attention in ichthyological studies although they are nursery grounds for many species. Kilby (1955) compared "inner pools" and "outer pools" for two regions on the Florida Gulf coast. He found a higher diversity (36 and 26 species) in the outer pools than in the lower salinity inner pools (19 and 28 spe- cies) . He did not find significantly more species in open water as I have, because of differences in collecting methods. The ecological stability of a habitat is gener- ally related to its species diversity. Therefore, it would appear that the low-salinity habitats, tidal pools, and tidal canals would be more vul- nerable to pollution than the other habitats. This suggests that locating factories or developments on lower reach or beach habitats would be less likely to damage the fish populations. However, other factors must be considered. For example, the beaches are especially sensitive to develop- ment because removal of the stabilizing beach plants results in rapid erosion. ADDITIONAL COASTAL HABITATS I consider the oyster habitat to include the oyster reefs of the lower reach and smaller patches of oysters in tidal creeks. The oyster habitat was located near the low tide level and was sampled by hand. Fishes that remain with- in the interstices between oysters at low tide I consider characteristic. In general order of de- creasing abundance, these are: naked goby, feather blenny, skilletfish, seaboard goby, striped blenny, oyster toadfish, and crested blenny. The mummichog was often observed swimming in the vicinity of oysters. At high tide a number of sciaenids migrate to submerged oyster reefs where they provide good fishing. The naked goby, feather blenny, mummichog, and oyster toadfish (one specimen) were also associated with the fouling community on the underside of floating docks in the lower reach. Few species occur in strictly freshwater hab- itat on Sapelo Island. Collections were made in a pond at the Marine Institute and ditches at artesian wells. The largemouth bass and blue- gill were introduced and are well established in ponds. There are specimens in the University of Georgia Fish Collection of the yellow bullhead and warmouth that were collected on Sapelo Island. These species probably were introduced. Five species that were common in a ditch at an artesian well, locally called Flora Bottom, are the mosquitofish, sailfin molly, marsh killifish, least killifish, and bluefin killifish (Lucania good- ei). Of these, the first three are euryhaline. The least killifish and bluefin killifish may be native to Sapelo Island, or they could have dis- 351 FISHERY BULLETIN: VOL. 70, NO. 2 persed across the estuary from the mainland since salinities are sometimes reduced to brack- ish in the estuary. The stabilization pond of the Interstate Paper Corp. at Riceboro provided a unique study hab- itat. Fishes were collected on 12 August 1969, 14 April 1970, 2 June 1970, and 6 July 1970. Fishes that were intentionally introduced in 1968 and 1969 and subsequently collected by seining include the largemouth bass, bluegill, and warmouth. Personnel of the paper company also collected adults of the following introduced species: brown bullhead (Ictalurus nebulosus) and redear sunfish (Lepomis microlophus) . I collected four specimens that appeared to be hy- brids of the bluegill and redear sunfish. Golden shiners are native to the area but since they were not collected until June and July 1970, they may have been introduced by fishermen. Native species include the mosquitofish, which was abundant, and the least killifish, which was represented by one specimen. I collected juve- niles of warmouth, sunfish, mosquitofish, golden shiner, and bass that were spawned in the pond. DISTRIBUTION PATTERNS OF COASTAL FISHES The 168 coastal fish species exhibit a variety of distribution patterns in relation to the de- scribed habitats. The 86 species that were re- stricted to one or two habitats included large numbers of (1) freshwater species, (2) elasmo- branchs and teleosts that were primarily ma- rine and stenohaline, and (3) rare species. The 30 species that occupied three habitats included only one elasmobranch {D. sabina) , three fresh- water species, and some uncommon and steno- haline species. The large number of species in these categories indicates that a majority of the coastal species occupy only a relatively small number of the coastal habitats. However, the low diversity estuarine habitats (habitats 5, 7, 8, and 9) may be widespread and important nur- sery grounds to estuarine species that are impor- tant as sport, commercial, or forage fishes, e.g., tarpon, anadromous shads, Atlantic menhaden. white catfish, tidewater silverside, silver perch, spotted sea trout, weakfish, spot, Atlantic croak- er, striped and white mullets, and southern flounder. Euryhaline species were defined as species that occur in both fresh water and pure seawater (Gunter, 1956). Although many of the species studied are euryhaline, none were found in all nine habitats. Six species occurred in all hab- itats except the freshwater habitat, and all of these sometimes occur in fresh water. These widely adapted species are the Atlantic men- haden, mummichog, spotfin mojarra, silver perch, spot, and striped mullet. The mosquito- fish, croaker, and white mullet occurred at seven habitats. The mosquitofish was absent from the deepwater habitats. The white mullet and croaker were absent only from low-salinity hab- itats. ACKNOWLEDGMENTS Much support for this study was derived from a contract with the Georgia Water Quality Con- trol Board with funds provided by the Interstate Paper Corp. of Riceboro, Ga. The National Sci- ence Foundation provided funds for the ship sup- port. Mrs. Joyce Swanberg identified representative plants using Radford, Ahles, and Bell (1968) but she used Muenscher (1944) to identify Elatine. G. C. Miller identified Prionotus pectoralis, now a synonym of P. salmonicolor. Dr. C. L. Hubbs identified some Diapterus olisthostomus, Eucin- ostomus argenteus, and E. melanopterus; Dr. E. Herald, some Syngnathns fuscus and S. louisia- nae; Dr. C. R. Gilbert, Evorthodus lyricus; and R. W. Heard, Jr., many invertebrates. Valuable field assistants included P. M. Glenn, J. C. Conyers, W. B. Sikora, J. Switten, and C. Durant. Capt. J. Rouse competently navigated the RV Kit Jones through the precarious estu- arine rivers. Mrs. Lorene Gassert drew the il- lustrations. Drs. F. J. Schwartz, G. Gunter, and C. E. Dawson made many valuable comments on the manuscript. 352 DAHLBERG: ECOLOGICAL STUDY OF GEORGIA COASTAL FISHES LITERATURE CITED Anderson, W. W. 1968. Fishes taken during shrimp trawling along the South Atlantic Coast of the United States, 1931-35. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 570, 60 p. Bailey, R. M. (chairman). 1970. A list of common and scientific names of fishes from the United States and Canada. Am. Fish. Soc, Spec. Publ. 6, 149 p. Bailey, R. M., H. E. Winn, and C. L. Smith. 1954. Fishes from the Escambia River, Alabama and Florida, with ecologic and taxonomic notes. Proc. Acad. Nat. Sci. Phila. 106:109-164. Bearden, C. M. 1961. Common marine fishes of South Carolina. Contrib. Bears Bluff Lab. 34, 47 p. 1963. A contribution to the biology of the king whitings, genus Menticirrhns, of South Carolina. Contrib. Bears Bluff Lab. 38, 27 p. 1964. Distribution and abundance of Atlantic croaker, Micropogon undulatus, in South Carolina. Contrib. Bears Bluff Lab. 40, 23 p. 1965. Elasmobranch fishes of South Carolina. Con- trib. Bears Bluff Lab. 42, 22 p. Carriker, M. R. 1967. Ecology of estuarine benthic invertebrates: a perspective. In G. H. Lauff (editor), Estuaries, p. 442-487. Am. Assoc. Adv. Sci. Publ. 83. Dahlberg, M. D. 1970. Atlantic and Gulf of Mexico menhadens, ge- nus Brevoortia (Pisces: Clupeidae). Bull Fla. State Mus. Biol. Sci. 15:91-162. Dahlberg, M. C, and R. W. Heard, IIL 1969. Observations on elasmobranchs from Georgia. Q. J. Fla. Acad. Sci. 32:21-25. Dahlberg, M. D., and E. P. Odum. 1970. Annual cycles of species occurrence, abun- dance, and diversity in Georgia estuarine fish pop- ulations Am. Midi. Nat. 83:382-392. Godwin, W. F., and J. G. Adams. 1969. Young clupeids of the Altamaha River, Geor- gia. Ga. Game Fish. Comm., Contrib. Ser. 15 : 1-30. Gunter, G. 1956. A revised list of euryhalin fishes of North and Middle America. Am. Midi. Nat. 56:345-354. 1958. Population studies of the shallow water fishes of an outer beach in south Texas. Publ. Inst. Mar. Sci. Univ. Tex. 5:186-193. 1961. Some relations of estuarine organisms to sa- linity. Limnol. Oceanogr. 6:182-190. 1967. Some relationships of estuaries to the fish- eries of the Gulf of Mexico. In G. H. Lauff (ed- itor), Estuaries, p. 621-638. Am. Assoc. Adv. Sci. Publ. 83, Heard, R. W., and W. B. Sikora. In press. A new species of Corophium Latreille, 1806 (Crustacea: Amphipoda) from Georgia brackish waters. June, F. C, and J. L. Chamberlin. 1959. The role of the estuary in the life history and biology of Atlantic menhaden. Proc. Gulf Caribb. Fish. Inst., 11th Annu. Sess., p. 41-45. Kilby, J. D. 1955. The fishes of two Gulf coastal marsh areas of Florida. Tulane Stud. Zool. 2:175-247. Linton, T. L., and W. L. Rickards. 1965. Young common snook of the coast of Georgia. Q. J. Fla. Acad. Sci. 28:185-189. Lunz, G. R., and F. J. Schwartz. 1970. Analysis of eighteen year trawl captures of seatrout (Cynoscion sp.: Sciaenidae) from South Carolina. Contrib. Bears Bluff Lab. 53, 29 p. Miller, G. L., and S. C. Jorgenson. 1969. Seasonal abundance and length frequency distribution of some marine fishes in coastal Geor- gia. U.S. Fish Wildl. Serv., Data Rep. 35, 102 p. Muenscher, W. C. 1944. Aquatic plants of the United States. Corn- stock Publ., Ithaca, N.Y., 374 p. Radford, A. E., H. E. Ahles, and C. R. Bell. 1968. Manual of the vascular flora of the Caro- linas. Univ. N.C. Press, Chapel Hill, 1183 p. Ragotzkie, R. a., and R. A. Bryson. 1955. Hydrography of the Duplin River, Sapelo Island, Georgia. Bull. Mar. Sci. Gulf Caribb. 5: 297-314. Reid, G. K., Jr. 1954. An ecological study of the Gulf of Mexico fishes, in the vicinity of Cedar Key, Florida. Bull. Mar. Sci. Gulf Caribb. 4:1-94. Rickards, W. L. 1968. Ecology and growth of juvenile tarpon, Me- galops atlanticus, in a Georgia salt marsh. Bull. Mar. Sci. Gulf Caribb. 18:220-239, Springer, V. G., and K. D. Woodburn. 1960. An ecological study of the fishes of the Tam- pa Bay area. Fla. State Board Conserv. Mar. Lab., Prof. Pap. Ser. 1, 104 p. Tabb, D. C. 1966. The estuary as a habitat for spotted seatrout, Cynoscion nebulosiis. In A symposium on estu- arine fisheries, p. 59-67. Am. Fish. Soc, Spec. Publ. 3. Tagatz, M. E. 1968. Fishes of the St. Johns River, Florida. Q. J. Fla. Acad. Sci. 30:25-50. Tagatz, M. E., and D. L. Dudley. 1961. Seasonal occurrence of marine fishes in four shore habitats near Beaufort, N.C, 1957-60. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 390, 19 p. 853 THE RELATIONSHIP BETWEEN THE SUMMER FOOD OF JUVENILE SOCKEYE SALMON, ONCORHYNCHUS NERKA, AND THE STANDING STOCK OF ZOOPLANKTON IN ILIAMNA LAKE, ALASKA' ' Stephen H. Hoag^ ABSTRACT The foregut contents of juvenile sockeye salmon in samples taken at night by tow net in the limnetic area of Iliamna Lake consisted primarily of zooplankton and rarely of in- sects. The number of organisms per foregut was correlated with the estimated zoo- plankton density between 0 and 100 m. Cyclops and Bosmina were the dominant zoo- plankters in both foregut and zooplankton samples. The zooplankton hauls contained a greater percentage of calanoid copepods than the fish foreguts. Food selectivity was indicated but appeared to be minimal. Fry (age 0) foreguts contained a lesser percentage of Cyclops and a greater percentage of Bosmina than did yearling (age I) foreguts. Juvenile sockeye salmon {Oncorhynchus nerka) spend 1 or 2 years in Iliamna Lake before mi- grating to sea. They occupy the littoral from the time they emerge from the gravel, in late winter or early spring, until mid-July, when they move to the limnetic area where they re- main until migrating to sea in the spring of the following or second year (at age I or age II). A similar change in distribution was found in Lake Aleknagik in the Wood River system (Pel- la, 1968) and is probably common for all juvenile sockeye salmon in the lakes of Bristol Bay. Pre- ious food studies, summarized by Rogers (1968) , indicated that juvenile sockeye salmon feed pri- marily on insects in the littoral and on zooplank- ton in the limnetic area. The standing crop of zooplankton is usually used as an indicator of food availability for zoo- plankton feeders. However, differences in size, agility, and visibility of the zooplankters may in- validate this assumption. The objectives of this study were: (1) to de- termine the food of juvenile sockeye salmon and ' Contribution No. 353, College of Fisheries Univer- sity of Washington, Seattle, WA 98195. " Work on this study was supported by the U.S. Fish and Wildlife Service, Contract Nos. 14-17-0005-82 (B) and 14-17-0005-129 (B). * Formerly, Fisheries Research Institute, University of Washington ; presently with the International Pacific Halibut Commission, University of Washington, Seattle, WA 98195. Manuscript accepted November 1971. FISHERY BULLETIN: VOL. 70, NO. 2, 1972. (2) to compare the composition of the diet with that of the estimated standing stock of zooplank- ton during the summers of 1966 and 1967. The diets of fry (age 0) and yearlings (age I) were also compared. The population density of ju- venile sockeye salmon in Iliamna Lake was high during these years as the escapement into the Kvichak River system was exceptionally large (24.3 million fish) in 1965. THE ENVIRONMENT Iliamna Lake is the largest lake in Alaska, with an area of 2,622 km' and an average depth of 44 m. It empties into the Kvichak River, which flows into Bristol Bay. The Lake was divided geographically into four sampling areas (Figure 1). Areas I and II have a mean depth of 34 m and an even bottom of glacial till. Area III is much deeper (mean depth 74 m) and has a highly variable, glacially scoured bottom. Area IV is made up of islands and isolated bays and also has a highly variable bottom contour. Twenty-nine fishes have been identified in the Kvichak River system (Bond and Becker, 1963) , but only the sockeye salmon is of commercial im- portance. The threespine stickleback, Gaster- osteus aculeatus, the most abundant of the res- ident species, may compete with the juvenile sockeye salmon for food. 355 FISHERY BULLETIN: VOL. 70, NO. 2 Scale m kilometers Figure 1. — Sampling areas of Iliamna Lake. The following species of zooplankton occur commonly in the limnetic area of the Lake: Di- aptomus gracilis, Erytemora yukonensis, Cy- clops scutifer, Bosmina coregoni, Daphnia long- iremis, and Holopedium gibherum (Lenarz, 1966). Juvenile sockeye salmon and zooplankton ex- hibit a similar dial vertical migration. Echo- grams have shown a movement toward the sur- face at dusk, a concentration usually at less than 10 m at night and a downward movement of fish at dawn. Pella (1968) described a similar diel vertical migration for juvenile sockeye salmon in Lake Aleknagik, Alaska. About half of the zooplankton population occurs above 15 m at midnight and above 50 m at noon in Iliamna Lake (Fowler and Lenarz, 1965). MATERIALS AND METHODS Fry (age 0) and yearlings (age I) were col- lected with a tow net, having an opening of 2.7 m' and a cod end mesh size of 0.2 cm, and suspended between two boats spaced 15 m apart with 30.5 m of tow line. Each tow was 20 min in duration at a speed of about 1.5 m/sec. Al- ternate tows were made at 1.5 and 6 m, and occa- sionally deeper tows were made when fish were observed at greater depths on the echo sounder. Most samples were taken during darkness (nor- mally between 2100 and 0300 hr). During the day catches were very small at depths of 30 m, and major fish concentrations usually were not seen on the echo sounder. Fishing near the bot- tom with monofilament, small-meshed gill nets was tried without success. Fry and yearlings were preserved separately in 10% Formalin within 15 min after capture. No regurgitation of the stomach contents was observed. The fish were measured and the stomachs removed several months later. The stomach was divided at the major bend, and only the contents of the foregut were exam- ined so that bias from unequal digestive rates among different food items would be minimized. The foregut contents from all the fish in a sample were combined and mixed with water of a known volume. Two subsamples were taken with a 1-ml bulb pipette, and the organisms identified and counted under a low-powered dissecting microscope. Each organism was identified as either Cyclops, Daphnia, Bosmina, Holopedium, calanoid copepods, or insects. Nauplii and roti- fers were seldom observed and were not counted. The total number of each food item in the sam- ple was estimated by multiplying the number of each food item in the two subsamples by the appropriate factor. The variance between sub- samples was less than 1% of the total variance among samples within date and area and was ignored. The samples ranged from 1 to 45 fish but usually contained between 15 and 25 fish. The results from each sample were weighted by the number of fish per sample and grouped by area and sampling period. The sampling periods were: /' 1. Late summer 1966 (August 15-September 10). 2. Early summer 1967 (June 20-July 20) . 3. Late summer 1967 (August 10-September 5). The zooplankton sampling (described by Lenarz, 1966) was undertaken for a separate study and diff"ered spatially and temporally from the young fish sampling. Samples were taken with a conical, nylon net of No. 6 mesh attached to a 0.5-m ring by vertical hauls, either from 100 m to the surface or from the bottom to the surface if the depth was less than 100 m. The zooplankters were identified and counted simi- 356 HOAG: SUMMER FOOD OF JUVENILE SOCKEYE SALMON larly to the organisms in the foregut. Samples were grouped by area and sampling period. The sampling periods were: 1. Late summer 1966 (August 10-26). 2. Early summer 1967 (June 18-30). 3. Late summer 1967 (August 22-September 15). RESULTS The numbers of fry and yearlings and the numbers of zooplankton samples by area and sampling period are listed in Table 1. Catches of yearlings in 1966 and of fry in the early sum- mer of 1967 were small and were excluded from the analysis. Both fry and yearlings occurred in varying numbers in nearly all samples during the late summer of 1967. The mean fork length by area and sampling period ranged from 50 to 58 mm for fry and from 71 to 96 mm for year- lings. Table 1. — Summary of samples of fish and zooplankton by area and sampling period. f 10- o Yearling Y = - 420 33 + 0 I 65 x • Fry Y-199 16+ 0 04 x 3 6 9 Number of organisms per cubic meter (tnousands) Figure 2. — Relationship between the number of organ- isms per foregut and the number of zooplankton per cubic meter. Sampling period Number of Number of Number of Area Age group fish fish zooplankton samples examined samples FEEDING ACTIVITY Most juvenile sockeye salmon (94%) con- tained some food. Temporal and spatial dif- ferences in the mean number of organisms in the foregut of juvenile sockeye salmon were apparent although the number of organisms per foregut varied considerably between samples within areas and sampling periods (Table 2). The number of organisms consumed per foregut increased from early to late summer and from area IV to area L The mean number of organisms per foregut by age group and the mean number of organisms per cubic meter in each area and sampling pe- riod are plotted in Figure 2. The positive slopes, significant at P = 0.05, indicate that feeding was in proportion to the abundance of zooplank- ton. The slope for fry is less than for yearlings probably due to the smaller foregut capacity of fry. Fry probably require a lower food concen- tration to become satiated, and the number of organisms per foregut may be approaching an Table 2. — Mean number of organisms per foregut and variance between samples by age group, area, and sampling period. Late summer 1966 1 Fry 7 22 9 II 7 62 9 iir 7 116 6 IV 17 385 4 Early summer 1967 Yearling 7 14 9 II 8 113 9 III 7 39 6 IV 8 81 4 Fry 7 43 9 II a 107 9 III 7 150 6 IV 7 160 6 Late summer 1967 Yearling 8 43 9 II a 105 9 III 7 120 6 IV 7 90 6 Total 127 1,650 116 Aga group Areo 1 Area II Area III Area IV Sampling period « * S 2 X X S 2 I X \' Late summer 1966 Fry 715 72,895 685 172,641 525 113,384 502 77,863 Early summer 1967 Yearling 222 46,625 497 157,397 338 8,910 242 50,485 Late summer 1967 Fry 649 66,511 429 31,179 414 20,784 360 35,023 Yearling 1,323 444,252 680 39,222 660 114,339 496 72,992 357 FISHERY BULLETIN: VOL. 70, NO. 2 :3 I ft ft ro- se 5.0- 4.0 200 400 600 800 1000 Mean number of organlsnns per fry foregut Figure 3. — Relationship between the mean number of organisms per foregut in fry and yearlings. asymptotic upper limit with respect to food abundance in the Lake. The relationship might have been nonlinear for both fry and yearlings if lower and higher zooplankton densities had occurred. The relationship between the number of or- ganisms per foregut and fish size within each age group was not examined because size range of fry and yearlings in any sampling area was small. However, the yearlings contained an average of 40 Yr more organisms than the fry in the late summer of 1967, probably because of their larger size and feeding capacity. The feed- ing intensities of both age groups are compared in Figure 3. The correlation coefficient, r = 0.67, between the mean number of organisms per fry foregut and the natural logarithm of the mean number of organisms per yearling foregut, significant at P = 0.05, indicates an exponential relationship between the feeding intensity of the two age groups. This was expected as fry tend to approach an upper limit of food intake (Fig- ure 2). COMPOSITION OF THE DIET AND THE ZOOPLANKTON SAMPLES The food of juvenile sockeye salmon in the limnetic area of Iliamna Lake consisted primar- ily of zooplankton. Insects averaged less than 1% of the total number of organisms in the fore- gut except for yearlings from area I during the early summer of 1967, when they constituted 42.6 % . Cyclops and Bosmina were usually dom- inant in the zooplankton hauls and in the fish foreguts, and averaged 75% in the zooplankton hauls and 90% of zooplankton in the fry and yearling foreguts (Table 3). Yearlings con- tained a higher percentage of Cyclops and a low- er percentage of Bosmina than fry in late sum- mer 1967 when both age groups were sampled. The percentage of calanoid copepods averaged higher in the zooplankton hauls (13%) than in the fish foreguts (3%) and was consistently higher for all areas and sampling period. Daph- nia and Holopedmm occurred in most samples but only constituted 11% of the organisms in the zooplankton hauls and 7% in the fish foreguts. To compare the percentage composition of or- ganisms in the zooplankton hauls with that of the foregut contents, a "two-way crossed" anal- ysis of variance (fish-zooplankton by area) was performed for each sampling period and for each organism {Cyclops, Bosmhia, and calanoid cope- pods). Daphnia and Holopedium were not tested. Mean squares and degrees of freedom are shown in Table 4. The percentage compo- sition of the foregut contents and the zooplank- ton hauls (Table 3) were significantly different except for Cyclops in late summer 1966 and early summer 1967. However, differences in the zoo- plankton hauls and the foregut contents were not consistent in all areas for Cyclops and Bos- mina as indicated by the significant interaction between the fish-zooplankton samples and areas. Interaction was not significant for the calanoid copepods. A modification of Tukey's test (Snedecor, 1956, p. 251) showed a significant difference 358 HOAG: SUMMER FOOD OF JUVENILE SOCKEYE SALMON (P = 0.05) in the percentage of Cyclops and Bosmina in fry foreg-uts, yearling foreguts, and zooplankton hauls in late summer 1967. The percentage of calanoid copepods in fry and year- ling foreguts did not differ significantly, but both differed significantly from that in the zooplank- ton hauls. Spatial and temporal changes in the composi- tion of the zooplankton catches were compared with changes in the fish foreguts. In the zoo- plankton hauls, the percentage of Cyclops gen- erally decreased and the percentage of Bosmina generally increased from early to late summer and from area IV to area I. The composition Table 3. — Mean percentages of Cyclops, Bosmina, and calanoid copepods in the foregut of fry and yearlings (insects excluded) and in the zooplankton samples by area and sampling period. Sampling period/ area Cyclops Bosmina Calanoids Fry Year ling Zoo- plankton Fry Year ling Zoo- plankton Fry Year- ling Zoo- plankton % % % % % % % % % Late summer 1966 Area 1 45.1 __ 42.0 45.6 __ 38.1 4.0 __ 7.1 Area II 29.4 __ 41.0 55.5 _^ 40.9 4.9 __ 6.5 Area 111 38.1 _. 47.5 49.6 _^ 38.6 1.2 __ 6.8 Area IV 26.8 — 62.0 64.5 — 24.9 1.0 — 7.0 Early summer 1967 Area 1 _^ 58.9 72.4 28.9 5.7 __ 7.5 13.1 Area 11 __ 79.0 78.5 __ 4.6 3.2 _^ 1.1 11.0 Area 111 _. 87.9 84.6 __ 4.2 1.2 __ 1.7 9.4 Area IV — 92.3 77.8 — 1.8 1.0 — 4.5 17.9 Late summer 1967 Area 1 21.0 47.2 27.5 68.3 44.0 34.6 2.5 2.3 17.0 Area 11 34.6 54.7 14.0 54.3 34.4 45.7 2.3 2.4 16.4 Area 111 75.2 84.8 38.4 18.2 6.8 27.0 2.8 2.6 21.1 Area IV 77.6 84.7 36.1 18.0 9.9 23.5 2.3 3.7 24.0 All sampling periods/area 43.5 73.6 51.8 46.8 16.8 23.7 2.6 3.2 13.1 Table 4. — Mean squares and degrees of freedom for a "two-way crossed" analysis of variance by sampling pe- riod and zooplankton. Cyclops Bosmina Calanoids Mean Degrees of freedom Mean Degrees of freedom Mean Degrees square square square freedom Late Summer 1966 Treatments (zooplankton, fry) 65.9 1 644.3** I 33.0** 1 Areas 117.4** 3 14.0 3 1.5 3 Interaction (treatment X area) 127.7** 3 106.8* 3 1.6 3 Residual 20.6 58 28.1 58 .6 58 Early summer 1967 Treatments (zooplankton, yearlings) Areas 3.4 178.6** 1 3 100.7** 109.2** 1 3 166.6** 16.8* I 3 Interaction (treatment X area) Residual 67.8 30.8 3 50 57.2** 10.6 3 50 5.8 4.8 3 50 Late summer 1967 Treatments (zooplankton, fry, yearlings) Areas 1,514.5** 1,086.8** 2 3 615.0** 887.5** 2 3 386.1** 6.0 2 3 Interaction (treatment X area) Residual 121.0** 26.9 6 77 97.5** 25.0 6 77 3.5 3.5 6 77 F test significant a\ P = 0.05. F test significant ai P = 0.01. 359 FISHERY BULLETIN: VOL. 70, NO. 2 Table 5.- -Rank correlation coefficient and zooplankton h auls for percentage of five by area and sampling organisms period. in fish foreguts Sampling per iod Samples - Area 1 II III IV Late summer 1966 Early summer 1967 Late summer 1967 Fry-zooplankton Yearllng-zooplankton Fry-zooplankton Yearllng-zooplankton 0.975*** .675 .700 .600 0.975* .425 .225 -.025 Significant at P = 0.20. Significant at P = 0.10. Significant at Z' = 0.05. 0.800* .300 .900* .825* 0.800* .700 1.00**' 1.00**' of organisms in the yearling foreguts showed a similar change; for Cyclops and Bosmina r rr 0.66 (significant at P = 0.10) and 0.71 (significant at P = 0.05), respectively. Cor- relation was not significant for fry (r = 0.01 and 0.22, respectively) . The percentage of cal- anoid copepods in the fish foreguts remained constant in spite of an increase from early to late summer in the zooplankton hauls (Table 3) . Although diflferences were significant in the percentage composition of organisms in the zoo- plankton hauls and in the fish foreguts, the fish fed predominantly on those organisms which were most abundant in the zooplankton hauls. Rank correlation coefficients (?d) were used in comparing the percentage of Cyclops, Bosmina, calanoid copepods, Daphnia, and Holopedlum in the zooplankton hauls and in the foreguts, and showed very good correlation in several areas and sampling periods and, with one exception, were always positive (Table 5). COMPARISON OF FRY AND YEARLING DIETS Cyclops and Bosmina were the major food items in the diet of both fry and yearlings al- though the percentage of the former was greater and the percentage of the latter smaller in the yearling foreguts than in the fry foreguts (Table 3) . Differences in the foregut contents may re- sult from slight diflferences in habitat and hence from diflferences in available food. To minimize this possibility only those sam- ples that included at least five fry and five year- lings were examined. The relationship between the two age groups in the percentage of Cyclops and Bosmina in the foreguts appeared linear (Figures 4 and 5) and a regression analysis was performed. The slope (b) was significantly greater than zero (P = 0.05) for both zoo- plankters; thus, the two age groups have similar food habits. However, the hypothesis 6 = 1 was rejected (P = 0.05) ; therefore, I con- cluded that fry consume more Bosmina and less Cyclops than yearlings. The positive intercept in Figure 4 and b less than 1 in Figures 4 and 5 indicate that the percentage of Cyclops and Bos- mina vary less in yearling than in fry foreguts. DISCUSSION AND CONCLUSIONS Cyclops and Bosmina were the most abundant, and calanoid copepods, Daphnia, and Holope- dlum, the least abundant zooplankters in the fish foreguts and in the zooplankton hauls. How- ever, the percentage composition of organisms in fish foreguts diflfered significantly from those in the zooplankton hauls in that: fry contained more Bosmina; yearlings contained more Cy- clops; fry and yearlings contained less calanoid copepods than the zooplankton hauls. Discrepancies in estimates of available food and diet are due to sampling error and selective feeding. Spatial and temporal diflferences in the sampling of zooplankton and fish, and diel and depth variations in available food and feeding activity probably accounted for some discrepan- cy. Fowler and Lenarz (1965) established diel and depth variations in the percentage compo- sition of the standing zooplankton stock in the lake. Northcote and Lorz (1966) showed diel changes in the food of resident sockeye salmon (kokanee) in Nicola Lake, British Columbia, 360 I HOAG: SUMMER FOOD OF JUVENILE SOCKEYE SALMON IOOt lOOn «> ■S 41 25 50 75 Percentage of Cyclops in fry diet 100 25 50 75 Percentage of Bosmina in fry diet 100 Figure 4. — Relationship between the percentage of Cyclops in the foregut of fry and yearlings. and Narver (1970) showed diel and depth changes in the food of juvenile sockeye salmon in Babine Lake, British Columbia. Food selection depends on characteristics of the feeder and food items (Ivlev, 1961) and prob- ably occurs in some degree for all species. In this study, the cause of the discrepancy between the composition of the foregut and zooplankton samples cannot be attributed specifically either to sampling error or to selective feeding. What- ever the cause, the degree of discrepancy was small, and I concluded that the zooplankton sam- ples generally reflect available food in Iliamna Lake. This is in contrast to Narver ( 1970) , who found that juvenile sockeye salmon strongly selected numerically less abundant zooplankters in Babine Lake. ACKNOWLEDGMENTS Dr. Ole A. Mathisen gave advice and encour- agement during the study and critically reviewed the manuscript; Drs. Robert L. Burgner, David W. Narver, and Donald E. Rogers critically re- viewed the manuscript; Dr. Tor B. Gunnerod conducted the zooplankton sampling and re- FiGURE 5. — Relationship between the percentage of Bosmina in the foregut of fry and yearlings. viewed the manuscript; and Orra E. Kerns, Jr. and John W. Anderson assisted in the collection of fish samples. LITERATURE CITED Bond, C. E., and C. D. Becker. 1963. Key to the fishes of the Kvichak River sys- tem. Fish. Res. Inst. Coll. Fish. Univ. Wash. Circ. 189, 9 p. Fowler, C. W., and W. H. Lenarz. 1965. Iliamna Lake limnetic zooplankton studies. In J. D. McPhail (editor), Research in fisheries .... 1964, p. 16-18. Coll. Fish. Fish. Res. Inst. Univ. Wash. Contrib. 184. Ivlev, V. S. 1961. Experimental ecology of the feeding of fishes. (Translated from the Russian by Douglas Scott.) Yale Univ. Press, New Haven, 302 p. Lenarz, W. H.^ 1966. Population dynamics of Cyclops scidifer and an evaluation of a sampling program for estimat- ing the standing crop of zooplankton in Iliamna Lake. M.S. Thesis, Univ. Washington, Seattle, 120 p. Narver, D. W. 1970. Diel vertical movements and feeding of underyearling sockeye salmon and the limnetic zooplankton in Babine Lake, British Columbia. J. Fish. Res. Board Can. 27:281-316. 361 FISHERY BULLETIN: VOL. 70, NO. 2 NORTHCOTE, T. G., AND H. W. LORZ. 1966. Seasonal and diel changes in food of adult kokanee (Oncorhynchus nerka) in Nicola Lake, British Columbia. J. Fish. Res. Board Can. 23:1259-1263. Pella, J. J. 1968. Distribution and growth of sockeye salmon fry in Lake Aleknagik, Alaska, during the sum- mer of 1962. In R. L. Burgner (editor), Further studies of Alaska sockeye salmon, p. 45-111. Univ. Wash. Publ. Fish., New Ser. 3. Rogers, D. E. 1968. A comparison of the food of sockeye salmon fry and threespine sticklebacks in the Wood River lakes. In R. L. Burgner (editor), Further stu- dies of Alaska sockeye salmon, p. 1-43. Univ. Wash. Publ. Fish., New Ser. 3. Snedecor, G. W. 1956. Statistical methods. 5th ed. Coll. Press, Ames, 534 p. Iowa State 362 SMALL-SCALE DISTRIBUTIONS OF OCEANIC DIATOMS E. L. Venrick^ ABSTRACT A sampling study was designed to investigate small-scale abundance fluctuations of diatoms over a distance of 10 miles. It was carried out at three depths in each of two oceanic environments of the North Pacific. Significant nonrandom distributions were observed. The intensity of aggregation varied with species and with depth. An expression for the approximate confidence intervals for single observations was derived from the 5th and 95th percentiles of the observed frequency distributions. Statistical analysis of the fluctuations of Nitzschia turgiduloides indicated a pattern of distribution with a scale of 1 mile. This may be associated with internal waves in the region of the thermocline. Knowledge of small-scale distributions of or- ganisms in the ocean is important for evaluation of data based on widely spaced samples, and, hence, is essential for design of efficient sam- pling programs. Moreover, abundance fluctua- tions on even the smallest scale relate directly to the ecology of the species, and an understand- ing of the magnitude and scale of such fluctua- tions is an important step toward the under- standing of a species' relationship to its environment and to other species within its community. Evidence indicates that the distribution of phytoplankton in the ocean may be highly aggre- gated (Bainbridge, 1957) . A few attempts have been made to sample small-scale aggregations and to investigate quantitatively their density and spacing and the environmental factors which influence them (e.g., Hasle, 1954; Holmes and Widrig, 1956; Barnes and Hasle, 1957; Cassie, 1959a, b, 1960; Bernhard and Rampi, 1965). Although these studies applied a wide variety of statistical procedures to a range of spatial and temporal scales, all of the phytoplankton species studied were reported to have aggre- gated distributions. However, such studies have all been conducted in the nearshore environ- ments. If, as has been suggested (Cassie, 1957) , the contagious distributions of plankton reflect ^ Marine Life Research Group, Scripps Institution of Oceanography, University of California, San Diego, La Jolla, CA 92037. heterogeneities in the environment, then the re- sults of such studies may not be applicable to the more homogeneous environments of the open ocean. The study described in this paper was carried out in two oceanic environments of the North Pacific. While it was primarily designed to give a quantitative estimate of the precision of sam- ples collected for an extensive study of oceanic diatoms, the results have general interest. LOCATION OF STUDY Closely spaced samples were taken twice dur- ing Scripps Institution of Oceanography Expe- dition URSA MAJOR, August-September, 1964. Station 23 (lat 49°07'N, long 155°31'W) was located in the Central Subarctic Pacific and Sta- tion 5 (lat 37°00'N, long 155°02'W) in the Cent- ral Pacific (Dodimead, Favorite, and Hirano, 1963) ; both regions were removed from the ef- fects of either neritic environments or the North Pacific Transition Domain. The phytoplankton of the Central Subarctic consisted primarily of diatoms which reached densities in excess of 5,000 cells/100 ml. A total of 27 diatom species was recorded, of which Nitzschia turgiduloides comprised 68-92 9f of the population, and Dentic- ula seminae an additional 9-20%. The maximum density of diatoms at the Central Pacific station was only 30 cells/100 ml. The dominant spe- cies, Hemiauliis hauckii, contributed 20% of the Manuscript accepted September 1971. FISHERY BULLETIN: VOL. 70, NO. 2, 1972. 368 FISHERY BULLETIN: VOL. 70, NO. 2 total diatom population, and 16 additional spe- cies were recorded. Subsequent samples from this region indicated that the major diatom pop- ulation was below the depth range of the sam- pling study. METHODS At each location a line of stations was posi- tioned with respect to a 10-m drogue. In the Subarctic Pacific, 10 stations were sampled; in the Central Pacific, 11. The stations were spaced at randomly ordered intervals of 0.5, 1.0, and 2.0 nautical miles, covering a total distance of 10.5 miles. At each station samples were col- lected from depths of 10, 35, and 50 m (wire out) . Samples were collected with 3-liter Van Dorn water samplers. The organisms in samples of 400 ml were preserved with 10 ml of 10% basic Formalin. Aliquots of 50 or 100 ml were settled for 24 hr and the diatoms identified and enu- merated under the inverted microscope accord- ing to the sedimentation procedure of Utermohl (1931). The species used for the analysis of distributions were those for which experience has shown the problems of identification and enumeration are negligible. In the case of chain-forming species, the statistical analyses were based upon the numbers of chains per aliquot. STATISTICAL PROCEDURES The variance of the cell counts includes the variability introduced into the data during the preparation and enumeration of the subsample, as well as any spatial heterogeneity. If these are independent, the sums of squares will be additive: oototal =^ OOsuhsaniple + OOspatial and the total observed variance is given by: s^otai = SStotai /degrees of freedom. In a brief preliminary study it was demon- strated that, for all species but one, the varia- bility introduced into the data at either the initial or final subsampling stage was no greater than random (Poisson) expectation and may be ap- proximated by the mean count. This agrees with the results of other workers (Holmes and Widrig, 1956; Lund, Kipling, and Le Cren, 1958). The single exception was Nitzschia turgiduloides, foj which the total introduced var- iability was SbX. It has been shown (Venrick, 1971) that the expected total variance of a series of counts from a randomly distributed population is given by: cr^ = \ (Wal— 1) (Wss) (Upop) (1— 77) ^al + [(Wss— l)(npop)(l— -^) X (fal+ ^^al)] + (Wpop-l)[Za, + ^Xa, + n. f '^jf- ^al] f /[Wal) (Wss) (Wpop) — 1] where na\, Wss, and Wpop are the numbers of aliquots per subsample, the numbers of sub- samples per sample, and the number of samples collected from each depth; /i is the ratio of sample volume to subsample volume, fi is the ratio_of subsample volume to aliquot volume; and Xa\ is the mean number of cells (or chains) per aliquot. Substituting t?ai = 1, Wss — 1, and , 3000 __ ,, . . ,.. ^ /i = "400" ~ ' expression simplifies to: r2 = X^i + J^ Zal + -73-yr- Xal where /2 = 8.0 for 50 ml aliquots and 4.0 for 100 ml aliquots. For N. turgiduloides, the ap- propriate expected variance is: a^ = 35Xai + n f- J Xni. At each depth, the observed variance, s^, may be compared with the expected variance, ar^, and the probability of departure from random ex- pectation determined by means of the ratio 364 VENRICK: SMALL-SCALE DISTRIBUTIONS OF OCEANIC DIATOMS §2/0-2 r= xVdf. For 10 and 11 samples xM0.05)/df values are 1.88 and 1.83 respec- tively. Species for which the s-/a^ ratio exceeds the xVdf value are considered to have aggre- gated distributions. Ratios greater than 1.63 and 1.60 respectively were significant at the 0.10 level. With the small number of degrees of freedom involved, the maximum variance attainable by species with mean counts less than 0.2 is too small to give an ^/a^ ratio significant at better than the 0.10 level. For rarer species, a runs test on presence and absence (Tate and Clelland, 1959) was used to give additional information about distribution patterns. RESULTS The detection of aggregation in a population is influenced by interaction between volume and spacing of field samples and the scale of aggre- gation of the population (Grieg-Smith, 1964), and by the proportion of the initial sample which is ultimately enumerated (Venrick, 1971). Thus, the specific results of this study are strictly pertinent only to this sampling design, and they must be interpreted accordingly. The results of these studies are presented in Tables 1 and 2. Within the Subarctic region, 8 of the 24 distributions were significantly ag- gregated at the 0.05 level, and two additional species at the 0.10 level. At every depth the species with contagious distributions were the most abundant ones, with the exception of N. turgiduloides at 10 m. It is likely that spatial variability of this species was obscured by the large sampling error. Aggregations of the dom- inant species would result if they had outgrown, in situ, the other species. The fewest aggre- gated distributions occurred at 10 m. This was the only sampled depth within the mixed layer, and presumably, wind-driven turbulence was sufficient to keep all but the most rapidly di- viding species distributed randomly. Within the Central Pacific, only 3 of the 20 distributions were significantly nonrandom, at the 0.05 level. The runs test, significant at the 0.10 level, indicated that five additional species were aggregated. In this region, aggregation did not appear to be related to the abundance of the species. Concordance tests were used to investigate the agreement of species with respect to fluc- tuations of abundances between samples. At Subarctic Station 23 there was significant con- cordance (P < 0.05) between all species at each of the three depths, indicating that species tended to respond to, or be influenced by, their environment in the same manner. In contrast, there was no concordance between species at any depth at Central Pacific Station 5. PRECISION OF SINGLE SAMPLES ESTIMATES OF ABUNDANCE If the frequency distribution of organisms in the field can be fitted to a theoretical distribu- tion, confidence limits on single observations can be derived from the variance of that distribu- tion. Some workers (e.g., Winsor and Clarke, 1940; Barnes and Hasle, 1957) have success- fully used logarithmic transformations to nor- malize abundance data. This procedure was successful for some of the diatom species under consideration in this study. (Normality was tested with normal-probability paper.) The transformation, however, was not successful for all species at all depths and thus a general use of parametric statistics on log-transformed data was not justifiable. The observed frequency distributions of the aggregated species were satisfactorily predicted by the negative binomial distribution (Ans- combe, 1949). Values of k_ (estimated from the expression k = X^/ (s^ — X) for the aggregated species ranged from 0.15 to 13.30. The com- parisons between the predicted and the observed cumulative frequency distributions were made with Kolmogorov-Smirnov tests (Tate and Clel- land, 1959) ; none were significantly different at the 0.10 level. There are available transfor- mations which normalize negative binomial dis- tributions (Anscombe, 1948). However, these transformations depend upon knowledge of the value of k and thus are applicable only to this particular set of data and not to observations of other species or observations from other en- vironments. 365 FISHERY BULLETIN: VOL. 70, NO. 2 Table 1. — Results of sampling study at Subarctic Pacific Station 23. Species \. 10-m depth Nitzickia turgiduloides^ Denticula seminae^ Chaetoceros atlanticus^ DactyliosoUn mediterraneus^ Coscinodiscus marginatus Rhizosolenia Thalassiothrix longissima Rhizosolenia hebetata hiemalis Corethron criophilum II. 35-m depth Nitzschia turgiduloides'^ Denticula seminae^ DactyliosoUn meditcrraneus'^ Chaetoceros attanticus^ Thalassiothrix longissima Rhizosolenia hebetata hiemalis Rhizosolenia alatd^ Corethron criophilum Coscinodiscus marginatus III. 50-m depth Nitzschia turgiduloides^ Denticula seminae^ DactyliosoUn mediterraneus^ Corethron criophilum Coscinodiscus marginatus Chaetoceros atlanticus^ Rhizosolenia hebetata hiemalis al ^Vo-2 513.5 53.8 8.A 4.4 2.1 1.6 1.5 1.3 1.1 801.0 119.3 23.1 3.0 1.5 1.4 0.9 0.9 0.8 330.6 39.0 7.7 0.7 0.6 0.2 0.1 17,981.23 61.44 9.59 5.02 2.40 1.83 1.71 1.48 1.26 28,048.62 136.24 26.38 3.43 1.71 1.60 1.03 1.03 0.91 11,581.91 50.04 9.88 0.90 0.77 0.26 0.13 20,764.30 271.28 8.26 9.30 0.76 2.26 1.16 1.56 0.98 122,374.22 1,966.23 367. 1 1 17.55 3.16 2.04 0.76 0.98 1.28 291,104.04 279.33 65.30 1.13 0.93 0.40 0.10 1.15 4.42 0.86 1.85 0.32 1.23 0.68 1.05 0.78 4.36 14.43 13.92 5.12 1.85 1.28 0.74 0.95 1.41 25.13 5.58 6.61 1.26 1.21 1.54 0.77 <0.00I <0.10 <0.001 <0.001 <0.001 <0.001 <0.10 <0.00I <0.001 <0.001 • Statistics based on numbers of chains per aliquot. The expression which was ultimately chosen to estimate the precision of single observations was derived empirically from the 5th and 95th percentiles of the frequency distributions. For any single count, x, of a nonrandomly distrib- uted species at a single depth, it was found that the expression 0.3x ^ X < 3.2a: included the observed population mean X 90% of the time. The expression was conservative for species with nonaggregated distributions. When all species were considered, the expression included the population mean 95% of the time. The expression gave satisfactory results for es- timates of mean numbers of cells of chain-form- ing species (F -^ 0.13) and for mean total diatom abundances (P — 0.14). The use of this expression is demonstrated in Table 3, where it has been applied to two samples from 35-m depth at Subarctic Station 23, 366 VENRICK: SMALL-SCALE DISTRIBUTIONS OF OCEANIC DIATOMS Table 2. — Results of sampling study at Central Pacific Station 5. Species al (T" ^Vo-= 1. 10-m deptS H emiaulus 2.4 Asterolampra marylandica 1.0 Rhizosdenia hebetata semispinal 0.7 Asteromphalus hep tact is 0.4 Chaetoceros dadayi^ 0.4 Nitischia sicula 0.3 Mastogloia rostrata 0.3 11. 35-m depth Hemiaulus hauckii^ 1.2 Nitischia sicula 0.7 Asteromphalus heptactis 0.6 Mastogloia rostrata 0.4 Bacteriastrum sp.i 0.4 Chaetoceros bacteriastroides 0.4 war. Chaetoceros bacteriastroides 0.4 Asterolampra marylandica 0.1 Chaetoceros dadayi^ 0.1 III. 50-m depth Nitzschia sicula 0.9 Hemiaulus hauckii^ 0.8 Chaetoceros bacteriastroides''- 0.4 Asteromphalus heptactis 0.2 Bacteriastrum sp.i 0.2 Chaetoceros bacteriastroides 0.2 var. Asterolampra maryla ndica 0.1 Mastogloia rostrata 0.1 3.08 1.28 0.90 0.51 0.51 0.38 0.38 1.54 0.90 0.77 0.51 0.51 0.51 0.51 0.13 0.13 1.15 1.03 0.51 0.26 0.26 0.26 0.13 0.13 3.45 l.OC 3.42 0.45 0.45 0.21 0.41 4.56 0.42 0.45 0.27 1.45 0.45 0.45 0.09 0.09 1.70 1.17 0.27 0.16 0.16 0.16 0.09 0.09 1.12 — 0.78 <0.10 (r) 3.80 <0.001 0.88 — 0.88 <0.10 (r) 0.55 — 1.08 — 2.96 = 0.001 0.47 <0.10 (r) 0.58 - 0.53 <0.10 (r) 2.84 <0.01 0.88 — 0.88 — 0.69 — 0.69 — 1.48 .. 1.14 <0.10 (r) 0.53 — 0.62 — 0.62 — 0.62 — 0.69 -. 0.69 1 Statistics based on numbers of chains per aliquot. r Nonrandomness indicated only by runs test. For the more abundant species, the 95% con- fidence limits which can be placed around a single sample are extremely broad. However, without replicate samples, this interval cannot be significantly reduced. For species represent- ed in a sample by fewer than five cells, the con- fidence interval given by the empirically derived expression is smaller than that obtained from the assumption of a Poisson distribution. For these rarer species, it is recommended that the confidence interval around a single sample be derived from the assumption of a Poisson dis- tribution (Fisher and Yates, 1957). ESTIMATES OF DIVERSITY The variability of individual species in the field determines the precision with which a 367 FISHERY BULLETIN: VOL. 70, NO. 2 single sample estimates the structure of the assemblage. The phytoplankton association within the Subarctic Pacific had a low diversity and significant concordance between species. As a result, the species showed a high degree of consistency of relative abundances within sam- ples. In every sample Nitzschia turgiduloides was the numerically dominant species, Denticula seminae was the second dominant, and one of two species, Chaetoceros atlanticus or DactyU iosolen mediterraneus was third in abundance. Thus, a single sample appeared to give a pre- cise estimate of the structure of a less diverse assemblage, even though the large between-sam- ple variability decreased the precision of the estimate of absolute abundances of single species. In contrast, the phytoplankton association in the upper 50 m of the Central Pacific had a high diversity and lacked concordance between spe- cies. At 10, 35, and 50 m, respectively, three, seven, and five species were dominant in at least one sample. Thus, a single sample from a di- verse assemblage gave an imprecise estimate of the relative abundances of the component spe- cies. Table 3. — Confidence interval about single samples. (95% confidence intervals about single samples, x, calculated from the expression 0.3x — 3.2a; and compared with the population mean density as estimated by the mean of 10 samples, X.) X Substaf 'on e' Substation h^ Species X Cc nfidence interval * Confidence interval Station 23, 35 m,- 50-ml aliquots Nitzschia cells 3,018.0 1,481 444.3-4,739.2 4,654 1,396.2- 4,892.8 turgiduloides chains 801.0 484 145.2- ,548.8 1,252 375.6- 4,006.4 Denticula cells 395.8 175 52.5- 560.0 624 187.2- 1,996.8 seminae chains 119.3 87 26.1- 278.4 169 50.7- 540.8 Dactyliosolen chains 23.1 8 2.4- 25.6 64 19.2- 204.8 mediterraneus Chaetoceros cells 7.7 I 0.9- 9.6 25 7.5- 80.0 atlanticus chains 3.0 I 0.3- 3.2 11 3.3- 35.2 Rhizosolenia cells 1.6 0 _^ __ 3 0.9- 9.6 hebetata chains 1.4 0 __ _^ 2 0.6- 6.4 hiemalis Thalassiothrix cells 1.5 2 0.6- 6.4 1 0.3- 3.2 longissima Coretkron cells 0.9 I 0.3- 3.2 1 0.3- 3.2 eriophilum Total cells 3,491.1 1,704 511.2-5,452.3 5,445 1,633.5-17,424.0 > Substations e and h separated by 3.5 nautical miles. ANALYSIS OF PATTERNS In the analysis of patchiness and its causal factors, the size and shape of a patch often re- ceives primary consideration. This approach is hampered by the difficulty of accurately de- fining a patch, particularly where, as in the ocean, one can rarely see the patch as a physical entity. An alternate approach is to examine the scale on which a population shows consistent spatial distribution, regardless of the degree of contagion. Since the detection of nonrandom- ness depends upon the interaction of the size and distribution of the samples with the pop- ulation distribution, if the scale of sampling is systematically altered, the observed population variance may change, and those sampling scales which produce maximum variances may indi- cate scales of heterogeneity in the population distribution. The six sets of 10 and 11 samples were con- sidered as sets of 45 and 55 pairs of samples separated by intervals of 0.5, 1.0, 1.5, . . . 10.5 miles. For all nonrandomly distributed species, the variance was calculated between each pos- sible pair of samples and averaged for each in- terval. Thus, for the set of 10 Subarctic sam- ples, in which three pairs were separated by 0.5 mile, four pairs by 1.0 mile, two pairs by 1.5 miles, etc., s^o.s is an average of three variances, s^i.o an average of four variances, s-1.5 an aver- 368 VENRICK: SMALL-SCALE DISTRIBUTIONS OF OCEANIC DIATOMS age of two variances, etc. The average varian- ces were plotted against the sampling interval, i. In the case of one species, A^. turgiduloides, this technique revealed a periodicity of si^ with peaks separated by 1-mile increments of the sampling interval (Figure 1 b-d). This indi- cates a pattern of heterogeneity on a scale of 1 mile which was not apparent from a direct plot of abundances (Figure la). The periodi- city was best developed at 35 m (runs test sig- nificant at P < 0.001) where it centered about the population variance, as measured by the total variance of the 10 samples. The periodicity was also highly significant (runs test, P = 0.01) at 10-m depth. At 50 m the variance showed sig- nificant periodicity (P = 0.10) only when the sample from substation h was omitted from the calculation. The high population densities of N. turgiduloides, and other species, encountered at substation h were comparable to densities ob- served at shallower depths, and may represent another scale of patchiness imposed upon the deeper populations by vertical mixing. The horizontal pattern observed in N. turgid- uloides was most highly developed along the top of the seasonal thermocline, which, at Station 23, extended between 30 and 50 m. Internal waves travelling along the thermocline produce a reg- ular series of vertical displacements, which may occur on a scale of 1 mile. In species with strong vertical gradients of density at the top of the thermocline, such circulation patterns would produce regular horizontal fluctuations of abundance, such as were observed in the present study. The effect of vertical displacement on the less strongly stratified species may have been obscured by their horizontal variations. This technique has been successfully used to investigate patterns of terrestrial vegetation (Grieg-Smith, 1964) . Once scales of heterogen- eity have been defined, those environmental pa- rameters that vary on scales of similar magni- tude may be sought as possible determinants of the species patterns. Because this approach is not limited to factors which can be measured simultaneously, it is very flexible. It is appli- cable not only to parameters in eflfect at the time of sampling, but also to those whose effect on phytoplankton was exerted some time in the past, and which cannot therefore be directly cor- related with abundance. It may for instance prove to be a useful tool for examining the ef- fect of vertically migrating herbivores on the standing stock of phytoplankton. SUMMARY AND CONCLUSIONS Of the distributions examined in the present study, less than half showed significant aggre- gation. For these species the 90% confidence interval about a single sample, x, could be esti- mated from the interval 0.3x — 3.2a:. This ex- pression was conservative for the nonaggregated species. The inability to establish contagion for the majority of the species investigated in the pre- sent study does not prove randomness on this or other scales. However, the prevalence of nonaggregated distributions lends support to the hypothesis that the oceanic environment is less complex than that of the nearshore region. In the oceanic environment, the numerous process- es which bring about local variations in abund- ance of phytoplankton appear to proceed more slowly relative to the randomizing turbulent processes. In such an environment, only the most important local processes produce a mea- sureable effect, and, thus, these may be rela- tively easily isolated for further study. ACKNOWLEDGMENTS I am grateful to Professor E. W. Fager for his help with the statistical analysis, and for his criticism of the first draft of this paper. The work was based on part of a dissertation submitted in partial fulfillment of the require- ments for the Ph. D. degree at the University of California at San Diego (Scripps Institution of Oceanography) . The work was supported in part by Scripps Institution of Oceanography and the Institute of Marine Life Research Program, the Scripps Institution of Oceanography's part of the California Cooperative Oceanographic Fisheries Investigations, which are sponsored by the Marine Research Committee of the State of California, and by the National Science Founda- tion Grant GB 2861. 369 FISHERY BULLETIN: VOL. 70. NO. 2 E o in c a. 35 m / 1200 c *% ; 't -• \ : V 1000 |\ 1 • '•--.. 800 600 400 i 200 40 30- c\j — 20- I 23456789 10 drogue distance (miles) i 35 m i ■Tt- 10- • ^ n 2 TT 3 5 6 7 1 I I I 9 10 'O 90- 7S- 60- (VJ - |tP 45 30- 15- b. lOm 1 i ; I Sn 5 .'«. I i I I I I I I I I I I I I I I I I 23456789 10 i (miles) I (miles) ( miles) Figure 1. — Small-scale distribution patterns of Nitzschia turgiduloides. (a) Fluctuations in the abundance of chains, 35 m. (b-d) Fluctuations of the mean, two-sample variance, s^^, with increasing sample interval, i, com- pared with total variance, s^^, of 10 samples: (b) 10 m; (c) 35 m; (d) 50 m; analysis run with (dotted line) and without (dashed line) substation h. 370 VENRICK: SMALL-SCALE DISTRIBUTIONS OF OCEANIC DIATOMS FLORAL REFERENCES Asterolampra marylandica Ehrenberg, 1845 (Hustedt, 1930, p. 485, Figures 270-271). Asteromphalus heptactis (Brebisson) Ralfs, 1861 (Hus- tedt, 1930, p. 494, Figure 277). Chaetoceros atlanticus Cleve, 1873 (Hustedt, 1930, p. 641, Figures 633-636). Chaetoceros bacteriastroides Karsten, 1907 (Karsten, 1907, p. 390, Table 44, Figure 2). Chaetoceros dadayi Pavillard, 1913 (Cupp, 1943, p. 109, Figure 64). Corethron criophilum Castracane, 1886 (Hendey, 1937, p. 325, Plates 7-8). Coscinodiscus marginatus Ehrenberg, 1841 (Cupp, 1943, p. 55, Figure 19, Plate 1, Figure 3). Dactyliosolen mediterraneus H. Peragallo, 1892 (Hus- tedt, 1930, p. 556, Figure 317). Denticula seminae Simonsen et Kanaya, 1961 (Simon- sen and Kanaya, 1961, p. 503, Figures 26-30). Hemiaulus hauckii Grunow, 1880-1885 (Cupp, 1943, p. 168, Figure 118). Mastogloia rostrata {'^a.Wich) Hustedt, 1933 (Hustedt, 1933, p. 572, Figure 1007). Nitzschia sicula (Castracane) Hustedt, 1958 (Hasle, 1964, p. 38— Figures 11-13; Plate 5, Figure 8; Plate 13, Figure 14 ; Plate 14, Figure 22 ; Plate 16, Figures 1-5). Nitzschia turgiduloides Hasle, 1965, p. 28 — Plate 12, Figures 9-14; Plate 13, Figures 3-6). Rhizosolenia alata Brightwell, 1958 (Cupp, 1943, p. 90, Figure 52). Rhizosolenia hebetata (Bailey) Gran f. hiemalis Gran, 1904 (Cupp, 1943, p. 88, Figure 50-A). Rhizosolenia hebetata (Bailey) Gran f. seinispina Gran, 1904 (Cupp, 1943, p. 88, Figure 50-B). Thalassiothrix longissima Cleve et Grunow, 1880 (Cupp, 1943, p. 184, Figure 134). LITERATURE CITED Anscombe, F. J. 1948. The transformation of Poisson, binomial and negative-binomial data. Biometrika 35:246-254. 1949. The statistical analysis of insect counts based on the negative binomial distribution. Bio- metrics 5:165-173. Bainbridge, R. 1957. The size, shape and density of marine phy- toplankton concentrations. Biol. Rev. (Cam- bridge) 32:91-115. Barnes, H., and G. R. Hasle. 1957. A statistical examination of the distribution of some species of dinoflagellates in the polluted inner Oslo Fjord. Nytt Mag. Bot. 5:113-124. Bernhard, M., and L. Rampi. 1965. Horizontal microdistribution of marine phy- toplankton in the Ligurian Sea. Bot. Gothob. 3:13-24. Cassie, M. C. 1957. The sampling problems with particular ref- erence to marine organisms. Proc. N. Z. Ecol. Soc. 4:37-39. 1959a. Micro-distribution of plankton. N. Z. J. Sci. 2:398-409. 1959b. Some correlations in replicate plankton samples. N. Z. J. Sci. 2:473-484. 1960. Factors influencing the distribution pattern of plankton in the mixing zone between oceanic and harbour waters. N. Z. J. Sci. 3:26-50. Cupp, E. E. 1943. Marine plankton diatoms of the West Coast of North America. Bull. Scripps Inst. Oceanogr. Univ. Calif. 5:1-237. Dodimead, a. J., F. Favorite, and T. Hirano. 1963. Salmon of the North Pacific Ocean. Part II. Review of oceanography of the Subarctic Pacific Region. Int. North Pac. Fish. Comm. Bull. 13, 195 p. Fisher, R. A., and F. Yates, 1957. Statistical tables for biological, agricultural and medical research. 5th ed. Oliver and Boyd, Edinburgh, 138 p. Grieg- Smith, P. 1964. Quantitative plant ecology. 2nd ed. Aca- demic Press, N.Y., 256 p. Hasle, G. R. 1954. The reliability of single observations in phy- toplankton surveys. Nytt Mag. Bot 2:121-137. 1964. Nitzschia and Fragilariopsis species studied in the light and electron microscopes. I. Some marine species of the groups Nitzschiella and Lanceolatae. Skr. utgitt Nor. Vidensk. -Akad. Oslo I. Mat. -Naturvidensk. Kl., Ny Ser. 16, 48 p. 1965. Nitzschia and Fragilariopsis species studied in the light and electron microscopes. II. The group Pseudonitzschia. Skr. utgitt Nor. Vidensk.- Akad. Oslo I. Mat. -Naturvidensk. Kl. Ny Ser 18, 45 p. Hendey, N. I. 1937. The plankton diatoms of the southern seas. Discovery Rep. 16:151-364. Holmes, R. W., and T. M. Widrig. 1956. The enumeration and collection of marine phytoplankton. J. Cons. 22:21-32. Hustedt, F. 1930. Die Kieselalgen Deutschlands, Osterreichs und der Schweiz mit Berucksichtigung der iibrigen Lander Europas sowie der angrenzenden Meeresgebiete. In Dr. L. Rabenhorsts Krypto- gamen-Flora von Deutschland, Osterreich und der Schweiz, Vol. 7 (Part 1), 920 p. 1933. Die Kieselalgen Deutschlands, Osterreichs und der Schweiz mit Beriicksichtigung der iibrigen Lander Europas sowie der angrenzenden Meeresgebiete. In Dr. L. Rabenhorsts Krypto- gamen-Flora von Deutschland, Osterreich und der Schweiz, Vol. 7 (Part 2, No. 4), p. 433-576. 371 FISHERY BULLETIN: VOL. 70. NO. 2 Karsten, C. 1907. Das Indische Phytoplankton nach dem Ma- terial der deutschen Tiefsee-Expedition 1898-1899. Wiss. Ergeb. Dtsch. Tiefsee-Exped. Dampfer "Valdivia" 1898-1899 2 (Part 2) :221-548. Lund, J. W. G., C. Kipling, and E. D. LeCren. 1958. The inverted microscope method of estimat- ing algal numbers and the statistical basis of estimations by counting. Hydrobiologia 11:143- 170. SiMONSEN, R., AND T. KaNAYA. 1961. Notes on the marine species of the diatom genus Denticula Kiitz. Int. Rev. ges. Hydrobiol. 46:498-513. Tate, M. W., and R. C. Clelland. 1959. Nonparametric and shortcut statistics in the social, biological, and medical sciences. Inter- state Printers and Publishers, Inc., Danville, 111., 171 p. Utermohl, H. 1931. Neue Wege in der quantitativen Erfassung des Planktons. Verb. Int. Ver. theor. angew. Limnol. 5:567-596. Venrick, E. L. 1971. The statistics of subsampling. Limnol. Oceanogr. 16:811-818. WiNSOR, C. P., and G. L. Clarke. 1940. A statistical study of variation in the catch of plankton nets. J. Mar. Res. 3:1-34. / 372 LABORATORY OBSERVATIONS ON THE EARLY GROWTH OF THE ABALONE, HALIOTIS SORENSENI, AND THE EFFECT OF TEMPERATURE ON LARVAL DEVELOPMENT AND SETTLING SUCCESS David L. Leighton^ ABSTRACT The influence of temperature on larval development rate and growth of juveniles of the white or Sorensen's abalone, Haliotis sorenseni, was investigated using a thermal gradient apparatus. While larvae developed most rapidly at 20°C, most settled juveniles at that temperature did not survive. At 15-16 °C, however, the operculate veliger stage was attained in 72 hr and settlement of advanced individuals occurred in 9 days. No settling was observed at 10°C. Juveniles maintained at 15-19°C and provided mixed diatoms as food showed marked variability in growth rate; at 130 days shell length ranged from 4.0 to 8.0 mm (average 5.5 mm). Two distinctly different patterns of shell pigment distribution emerged with continued growth. Approximately 60% of the juveniles were bicolored, red and yellow-green, while the remainder had an even tone of red-violet. The description by Carlisle (1962) of the troch- ophore and early veliger stages of the red aba- lone, Haliotis rufescens Swainson, has hereto- fore been the only published information on lar- val development of an eastern Pacific species of Haliotis. No account of larval development through settlement and juvenile growth of any of the seven American species of abalones exists in the literature. Details of larval morphogen- esis and an estimate of growth during the first year of life for the northeastern Atlantic H. tu- berculata Linnaeus were given by Crofts (1929, 1937) . Japanese workers have reported obser- vations on early development and growth in several of their native species; H. gigantea Chemnitz (Murayama, 1935), H. discus Reeve and H. sieboldii Reeve (Ino, 1952) , and H. di- versicolor supertexta Lischke (Oba, 1964). Recent interest in mass culture of commer- cially important species in Japan and the United States has prompted more critical studies of growth and nutrition of abalones. Advances by Japanese workers in the field of abalone cul- ture were reported by Imai (1967) and Ryther (1968). However, no comparable research in- ^ Scripps Institution of Oceanography, University of California, San Diego, La Jolla, CA 92037. Manuscript accepted December 1971. FISHERY BULLETIN: VOL. 70, NO. 2, 1972. formation has been published heretofore from any abalone culture facility in California. This work, done in conjunction with abalone culture efforts of California Marine Associates, Cayu- cos, Calif., describes early growth of the white or Sorensen's abalone, Haliotis sorenseni Bartsch. H. sorenseni exhibits an unusual pattern of distribution. Common to southern California and northern Baja California, it appears in abundance along the mainland only at the north- ern and southern extremes of its range (Santa Barbara to Pt, Conception and Pta. Eugenia to Pta, Abreojos, respectively). Elsewhere it is islandic, occurring at Santa Cruz, Santa Bar- bara, San Clemente, Santa Catalina, Los Cor- onados, and Cedros Islands (Cox, 1962) , Adults attain a shell length of about 25 cm and are characterized by an orange-tan epipodium with foliose epipodial papillae, brown cephalic tenta- cles, and a deeply cupped, light-weight, and scar- free shell. Respiratory apertures are strongly fluted. The species is prized by commercial and sport divers for its white and tender edible por- tion, the right shell muscle. Sexually mature //. sorenseni have been dif- ficult to obtain in good condition since this is a relatively fragile species and losses in transit 373 FISHERY BULLETIN: VOL. 70. NO. 2 have been common. On several occasions gravid adults have been retained in our tanks and in- duced to spawn, but such attempts were never successful in obtaining their larvae. It is con- sidered, therefore, of importance to document a first success with the species. METHODS Ten adult H. sorenseni were collected at Santa Catalina Island and transported by ship to San Diego, February 18, 1971." Four possessed gonads sufficiently mature to respond to spawn- ing stimuli. Spawnings were induced on Feb- ruary 21 and 26 and March 1, using the thermal shock method (Ino, 1952), but in no case did fertilization occur in greater than about 5% of the eggs liberated. Observations of larval de- velopment and growth were carried out with progeny (approximately 1,000 trochophore lar- vae) obtained March 1 from a single male and female pair. Fertilized eggs were 190 to 200 /a in diameter. Washing of eggs to remove excess sperm, mucus, and other possibly contaminating substances re- leased during or prior to spawning was achieved by repeated suspension in filtered seawater (Mil- lipore' filters, pore size 0.45 jx) following settling and decantation. After incubation of the eggs at 12 ± 2°C for 24-36 hr, larvae hatched at the trochophore stage. Because the swimming trochophores exhibited a strong negative geo- taxis, they concentrated at the surface where they were easily drawn into a Pasteur pipette for transfer to experimental containers. The effect of temperature on rate of develop- ment, settlement, and growth of recently settled juveniles was investigated using an aluminum thermal gradient block (Thomas, Scotten, and Bradshaw, 1963) with compartments bored to accommodate a series of 100-ml beakers. Each compartment could be illuminated from beneath to permit growth of algae within beakers. Con- " Collections were made at a depth of about 70 ft by T. Tutschulte and were timed to sample the peak of the reproductive season at the Isthmus station. ° Reference to trade names in this publication does not imply endorsement of commercial products by the Na- tional Marine Fisheries Service. tainers, arranged in duplicate, were maintained at 10, 12, 14, 16, 18, and 20°C (each ± 0.5°C). The Pyrex beakers received 80 ml of Millipore- filtered and ultra-violet light treated seawater (Loosanoff and Davis, 1963). Twenty 72-hr veliger larvae were introduced into each togeth- er with 1 ml of a suspension of the food orga- nism, Nitzschia sp,, containing approximately 10,000 cells. To guard against salinity changes due to evaporation or condensation, each beaker was covered with a paraffin sheet, and the entire block insulated with foam styrene. Inspection was made on alternate days to determine the number of larvae surviving, settlement, and growth rate. The water was changed once weekly and new food supplied. Approximately 500 larvae were distributed among five 1-liter Pyrex beakers at the time the thermal gradient erperiment was begun. These containers received a combination of about 10 species of pennate diatoms collected from illu- minated aquarium surfaces in the laboratory (chiefly Nitzschia, Grammatophora, and Navic- ula). A thermal environment of 15-17°C was maintained. DESCRIPTION OF LARVAE The reader seeking details of morphogenesis in haliotids should refer to illustrations and text as provided for Japanese species by Ino (1952) and Oba (1964), early development of which closely parallels American members of the genus. In this paper an efl!"ort is made to indicate dis- tinctive features of H. sorenseni larvae, Trochophores were subcylindrical in outline, bore a distinct prototroch and were yellow-tan in color. At 15-16°C, the roughly symmetrical shape of the trochophore was soon lost and by 48 hr after fertilization the early veliger form was attained with the incipient cap-shaped shell and lobular velum becoming defined. At 72 hr, the larval shell had taken the typical gastropo- dan snail-form complete with operculum while shell musculature, viscera, eye-spots, and velum were well developed. At the same time the foot and cephalic tentacles began to diflE"erentiate (Figure 1). Tissue pigmentation was predom- inantly beige, velar fringes were yellow, and the 374 I LEIGHTON: EARLY GROWTH OF THE ABALONE Figure 1. — Three-day veliger larva of Haliotis sorenseni. (Shell diameter, 270 jj.) a. shell, b. digestive gland, c. velum, d. eye spot, e. incipient cephalic tentacle, f. foot, g. operculum. digestive gland was maroon. Swimming of ve- ligers at this stage carried them throughout the water column. Settlement occurred at 9-10 days at 15°C. Velar dystrophy, with loss of ciliated fringes, took place gradually, and larvae retained the ability to swim if dislodged for several days after initial settlement. INFLUENCE OF TEMPERATURE ON DEVELOPMENT RATE Larvae confined to 100-ml beakers in the thermal gradient block appeared, except for the temperature efl^ects to be described, to be normal in appearance and behavior and were similar to larvae maintained in liter beakers at 15-17 °C. Development was most rapid at higher temper- atures; some individuals settled as early as the 7th day at 20°C and the 8th day at 18°C. By the 15th day almost all larvae at the two highest temperatures had settled, yet none had settled at 10°C (Table 1, Figures 2 and 3). By the 25th day most larvae at 10°C had succumbed, having progressed only to the late veliger stage. Settling rate varied within groups, and even at 18 °C approximately 5% did not settle. Observations at the end of the 36-day exper- iment revealed that survival was best at 16° and 18°C, somewhat less at 20° and least at 14° and below. Larvae maintained at 10°C did not survive beyond 25 days and most of those at 12° failed at 25 to 30 days. Size attained by settled juveniles corresponded with survival; at 36 days shell lengths as great as 1.2 mm were attained at 18°C, about 1.0 mm at 20° and 16°C, and only 0.6 mm at 12° and 14°C. GROWTH OF JUVENILE H. SORENSENI Approximately 100 settled juveniles were ob- tained from the larvae reared in liter Pyrex beakers at 15-17°C. These individuals were carefully dislodged by a jetting stream of water and teasing needle and transferred to larger con- tainers to observe growth and behavior under a variety of food and water flow situations. Food quality and quantity was not limited to in- sure maximum growth, hence a wide variety of food organisms was provided (pennate diatoms and filamentous red, green, and brown algae) . Table 1. — Number of Haliotis sorenseni larvae surviving and settled at 15 days in thermal gradient experiment. Series 1 Series II Temperature Surviving Settled Survivors settled Surviving Settled Survivors settled °C ± 0.5° No. No. % No. No. % 10 s 0 0 18 0 0 12 7 4 57.2 15 10 66.6 14 15 11 73.4 IS 14 77.7 16 12 11 91.5 14 11 78.5 18 9 9 100.0 17 15 88.4 20 7 7 100.0 13 11 84.6 375 FISHERY BULLETIN: VOL. 70, NO. 2 lOOi- > < O in o 111 ill 111 < > < Z 20 - o cr UJ a 10 14 16 TEMPERATURE CO 18 Figure 2. — Percentage of larvae firmly settled at 15 days. Points are averages for duplicate observations. Settlement was considered complete when larvae could no longer be dislodged by jetting a stream of water from a pipette. A marked variation in growth rate existed, even among those individuals which were pro- vided surfaces (such as plastic and glass beak- ers) on which diatom growth appeared to be quite uniform. In one case, 19 individuals were retained in a liter beaker and provided mixed diatoms. At an age of 85 days some had at- tained a shell length of 4.5 mm while others were as small as 1.4 mm. The largest had formed four respiratory pores, yet the smallest was just beginning to form the first. Size attained at 100 days (15-19°C) ranged from 3.0 to 5.6 mm (mean, 4.25, SD ± 0.63 mm, n = 19). By 130 days the range was 4.0 to 8.0 (mean, 5.53 ram, SD lb 1.01 mm) in the same group (Figure 4). DESCRIPTION OF JUVENILE SHELL FEATURES The most conspicuous feature of recently set- tled and minute haliotids (less than 500 fx long) is the asymmetrical outline resulting from dex- tral growth of the persistomial shell. Through 1 o 15 Ijj u) 10 <0 >- g 5 ± X ± 12 14 16 18 TEMPERATURE CO 20 Figure 3. — Days to settlement of at least 75% of Haliotis sorenseni larvae from groups held at five different tem- peratures. Not more than 75% of the larvae at 12°C successfully settled, and none of those at 10 °C reached this stage. Si- Figure 4. — Growth of larval and juvenile Haliotis sor- enseni in 1-liter beakers provided mixed diatoms as food. Ranges of shell length are shown by vertical lines. 376 LEIGHTON: EARLY GROWTH OF THE ABALONE continued growth the asjTiimetry is reduced and the shell form, as viewed from above, is almost circular until the "notch stage" is reached. For- mation of the first respiratory aperture is ini- tiated by separation of two portions of the mantle at the right anterior margin interrupting other- wise uniform marginal shell deposition. The "notch" so formed is a convenient growth mark, reached by the most rapidly growing H. soren- seni at an age of 55 days and a shell length of 2.0 or 2.1 mm (Figure 5). In the juvenile H. sorenseni observed, shell pigmentation was a pale violet-pink becoming more vivid as the notch stage was approached. In most cases a cyan-blue flare spread from the apex to the right shell margin. After comple- tion of the first respiratory pore, a dichotomy in pigment pattern emerged. Although all ju- veniles developed from eggs of a single female, fertilized by a single male, approximately 40% developed an even red-violet shade throughout the greater extent of the shell while the majority became increasingly bicolored with growth. The bicolored pattern consisted of a rich red zone along the left edge of the shell (extending to the aperture row) and pigmentation of blue, green, and yellow over the broader surface right of the aperture row. In individuals of both color- ation patterns, a conspicuous ivory-white patch remained at the position of attachment of the right shell muscle until obscured by nacre depo- sition at about 8 mm. The apex, in turn, became increasingly white. Elevated apertures, typical of the species, ap- pear even in first-pore individuals. More con- spicuous is the ridge bordering the left shell margin. It is this ridge that forms the sharp corner and relatively straight left anterior mar- gin (Figure 6). DISCUSSION Establishment of specific characteristics by which larvae and juveniles of the seven cooccur- ring California species of Haliotis may be dis- tinguished must await results of studies of fine structure of larval shells. Examination by scan- ning electron microscope has revealed sufficient detail of fine crystalline and basement structure Figure 5. — Juvenile Haliotis sorenseni at "notch" stage (2.0 mm, 55 days). Shell is still slightly transparent and prominent anatomical features are readily seen from the dorsal aspect, a. cephalic tentacle, b. mouth, c. in- dentation at point of formation of first respiratory pore ("notch"), d. right shell muscle, e. intestine, f. digestive gland, g. epipodial tentacle. of larval shells in certain other prosobranchs (Fretter and Pilkington, 1971). Gross shell morphology alone is inadequate to differentiate larvae of even distantly related species. Pig- mentation of velar and visceral portions may provide distinctive features for recognition of some species. Pigments derived from parental yolk appear to be retained by trochophore and veliger larvae of Haliotis. Among California species, ovarian tissue is dark green in H. rufescens, H. crach- erodii, H. walallensis, and H. kamtschatkana assimilis. Correspondingly, larvae of these spe- cies are conspicuously green. The remaining three species found locally, H. fulgens, H. corru- gata, and H. sorenseni, produce eggs of brown, olive, and beige color, respectively. Their larvae 377 FISHERY BULLETIN: VOL. 70, NO. 2 may be expected to reflect these pigments ac- cordingly. I have examined larvae of H. fulgens which were generally brown but with green velar margins and those of H. sorenseni, which, as described above, were beige with yellowish velar margins. Tissues of larval H. corrugata are light yellow-green while velar fringes are a darker shade of green. Color of the digestive gland in planktotrophic prosobranch larvae has been shown to reflect diet (Fretter and Montgomery, 1968). Since HaUotis veliger larvae are lecithotrophic, color of the digestive gland may have diagnostic value. The maroon-colored digestive gland of H. soren- seni veligers appears to be distinctive; I have not observed other than green or brown in the digestive glands of other haliotid veligers. Although development of larvae and growth of settled juvenile H. sorenseni were more rapid at higher temperatures, survival was reduced at 20°C. More advanced juveniles, reared initially at 15-20°C, did not appear adversely aff"ected by temperatures as high as 25 °C. Possibly thermal tolerance limits are more restricted in larvae and recently settled individuals. Certainly other factors could have influenced success at higher temperatures. Despite weekly changes of water, bacterial, algal, and protozoan growth together with a build up of metabolites and reduced oxy- gen tensions in the relatively small volume of water used in the thermal gradient study could have influenced the results. Larval H. sorenseni were not successful at 10- 12°C. Water temperatures within the bathy- metric range of the species may fall to 12° and occasionally to 11°C (e.g., at depths of 130 ft oflf Santa Catalina Island, T. Tutschulte, per- sonal communication). Therefore the 10° or 12° C bathyisotherms may limit the depth to which larvae of H. sorenseni may successfully settle and grow. The results of the temperature block study indicate the range 14-18°C may be optimal for H. sorenseni larvae — an outcome not unexpected in view of the prevailing conditions in the natural environment from Pt. Conception to central Baja California. Information on growth throughout the first days of life is available for several species of HaUotis. Ino (1952) reported that H. discus attained only 1.25-1.40 mm at 100 days and that the first respiratory pore was not formed until 130 days. In contrast, Oba (1964) observed rapid early development of H. diversicolor super- texta. In that study, trochophores hatched at only 6 hr, veligers developed within 11 hr and settled by 2 days. The first respiratory pore was formed as soon as 23 days after fertilization. Interpolating from his growth curve, 100-day juveniles ranged from 8 to 13 mm (Oba et al, 1968). In Ino's study, water temperatures de- clined from 18° to 10°C through the course of observations while Oba's work was carried out during the summer and fall when temperatures ranged between 20° and 28 °C. Whether the dif- ferent development rates in these two species reflect specific contrasts or thermal influences is not clear. In another study (Kan-no and Ki- kuchi, 1962), H. discus hannai was reared at a relatively constant intermediate temperature (19-20°C). This species also exhibited rapid early development, settling in 3 days. Juveniles reached 11 mm in length at 100 days and 26 mm at 180 days. The rather extreme variability in growth rate observed in H. sorenseni has also been found in H. rufescens (Leighton, unpublished data). Comparable variability is reflected in Oba's growth curve for H. diversicolor supertexta. DiflFerences in growth rate may reflect variation in food availability. In the present study, care was taken to provide uniform food distribution and feeding conditions. Yet within a single container, even within sampled subareas, a full spectrum of size variation could be found. The hypothesis may be advanced that gametogenic inequalities (e.g., yolk allotment) may be in- volved giving greater advantage to some indi- viduals over others. Indeed, mortalities were more common among the smaller and presuma- bly less active members. Inherent variability in growth rate may be expected in nature to be complicated by differ- ential quality and availability of food. Multi- modal size-frequency distributions obtained for juvenile HaUotis populations in the field have been concluded to reflect multiple spawnings and recruitment waves (Leighton and Boolootian, 1963; Boolootian, Farmanfarmaian, and Giese, 378 LEIGHTON: EARLY GROWTH OF THE ABALONE Figure 6. — Photograph of shells of juvenile Haliotis sorenseni representing the bicolored (above) and red pattern (below) . Shells range in length from 6.6 to 8.4 mm. 379 LEIGHTON: EARLY GROWTH OF THE ABALONE 1962). The evidence from this study suggests the assumption of uniform growth within a pop- ulation of juvenile Haliotis is untenable and that conclusions regarding settlement date estimated from sizes of members of a sample must be drawn cautiously. ACKNOWLEDGMENTS This study was carried out under Sea Grant No. GH-52 to the University of California, San Diego, using aquarium facilities at the National Marine Fisheries Service, Southwest Fisheries Center, La Jolla. I wish to thank Dr. Reuben Lasker for his advice and criticism of the man- uscript. LITERATURE CITED BooLOOTiAN, R. A., A. Farmanfarmaian, and A. C. GlESE. 1962. On the reproductive cycle and breeding hab- its of two western species of Haliotis. Biol. Bull. (Woods Hole) 122:183-193. Carlisle, J. G. 1962. Spawning and early life history of Haliotis rufescens Swainson. Nautilus 76:44-49. Cox, K. W. 1962. California abalones, family Haliotidae. Calif. Dep. Fish Game, Fish Bull. 118, 133 p. Crofts, D. R. 1929. Haliotis. Liverpool Mar. Biol. Comm. Mem. 29, 174 p. 1937. The development of Haliotis tuberculata, with special reference to organogenesis during torsion. Philos. Trans. R. Soc. Lond., Ser. B Biol. Sci. "208:219-268. Fretter, v., and M. C. Montgomery. 1968. The treatment of food by prosobranch vel- igers. J. Mar. Biol. Assoc. U.K. 48:499-520. Fretter, V., and M. C. Pilkington. 1971. The larval shell of some prosobranch gas- tropods. J. Mar. Biol. Assoc. U.K. 51:49-62. IMAI, T. 1967. Mass production of molluscs by means of rearing the larvae in tanks. Venus 25(3-4) :159- 167. Ino, T. 1952. Biological studies on the propagation of Jap- anese abalone (genus Haliotis). [In Japanese, English summary.] Bull. Tokai Reg. Fish. Res. Lab. 5, 102 p. Kan-no, H., and S. Kikuchl 1962. On the rearing of Anadara broughtonii (Schrenk) and Haliotis discus hannai Ino. Bull. Mar. Stn. Asamushi 11(2) :71-76. Leighton, D., and R. a. Boolootian. 1963. Diet and growth in the black abalone. Hal- iotis cracherodii. Ecology 44:227-238. Loosanoff, V. L., and H. C. Davis. 1963. Rearing of bivalve moUusks. Adv. Mar. Biol. 1:2-130. MURAYAMA, S. 1935. On the development of the Japanese abalone, Haliotis gigantea. Coll. Agric. Tokyo Imp. Univ. J. 13:227-233. Oba, T. 1964. Studies on the propagation of an abalone, Haliotis diversicolor supertexta Lischke — II. On the development. [In Japanese, English synopsis.] Bull. Jap. Soc. Sci. Fish. 30:809-819. Oba, T., H. Sato, K. Tanaka, and T. Toyama. 1968. Studies on the propagation of an abalone, Haliotis diversicolor supertexta — III. On the size of the one-year-old specimen. [In Japanese, English synopsis.] Bull. Jap. Soc. Fish. 34:457- 461. Ryther, J. H. 1968. The status and potential of aquaculture. Volume 1 : Particularly invertebrate and algae culture. Am. Inst. Biol. Sci., Wash., D.C., 261 p. Distributed by Clearinghouse Fed. Sci. Tech. In- form., Springfield, Va., PB 177 767. Thomas, W. H., H. L. Scotten, and J. S. Bradshaw. 1963. Thermal gradient incubators for small aquatic organisms. Limnol. Oceanogr. 8:357-360. 381 EXPLOITATION EFFECTS UPON INTERSPECIFIC RELATIONSHIPS IN MARINE ECOSYSTEMS' Saul B. Saila and James D. Parrish" ABSTRACT Due to man's continuing efforts to extract greater harvests of marine organisms from the world ocean, it is becoming increasingly important to be able to predict the conse- quences of exploitation on complex assemblages of organisms. These assemblages, or ecosystems, consist of predator and prey organisms in various interacting combinations. Preliminary evidence available from studies of marine invertebrate communities in coastal areas has indicated that removal of grazing herbivores or predators at various levels results in lower species diversity and greater instability of the ecosystem. In order to permit a quantitative evaluation of the effects of various rates and types of exploitation on interspecific relationships, model ecosystems were constructed utilizing a subset of graph theory as applied to network analysis. A basic ecological trophic unit was for- mulated, and these units were combined to form more complex model ecosystems. In par- ticular, a hypothetical four species system of interacting predator and prey organisms was analyzed to demonstrate the consequences of varying certain model coefficients, espe- cially rates of exploitation. It was shown that nonselective exploitation tended to main- tain stability of the system better than highly selective exploitation. A hypothetical example of an empirical approach for examining changes in community structure was also demonstrated. Much of the present theory of fisheries science as well as many practical fisheries management techniques are based on the concept of a single species or unit stock (Beverton and Holt, 1957; Ricker, 1958). This approach continues to be useful in describing and predicting the behavior of fisheries consisting primarily of a single spe- cies. Recently, Walters (1969) developed a de- terministic computer simulation model for deter- mining optimum harvest strategies based on a unit stock. However, modern fishing seems to be progressing toward exploitation on many spe- cies of the larger animals in aquatic ecosystems. To some extent this is due to the tendency to- ward reduction to meal of many species of fishes. Some obvious areas for the future development ' Part of this work is a result of research sponsored by U.S. Department of Commerce, National Oceanic and Atmospheric Administration, Office of Sea Grant, under Grant #2-35190 and a National Institutes of Health Fel- lowship, 5 FOl GM48175-02, General Medical Sciences to JDP. ^ Graduate School of Oceanography, University of Rhode Island, Kingston, RI 02881. Manuscript accepted January 1972. FISHERY BULLETIN: VOL. 70, NO. 2, 1972 of marine fisheries will be in tropical or sub- tropical waters, such as parts of the Indian Ocean as well as the southeast and southwest Atlantic. These areas are characterized by a relatively greater diversity of species than the traditional fishing grounds of north-temperate regions. It can reasonably be expected that some species will be exploited intensively from them. It therefore seems desirable to attempt to better understand trophic relations of ecosys- tems consisting of several interacting species in order to develop management techniques for these systems. This increased understanding includes the eff"ects of various kinds of exploi- tation strategies as well as environmental per- turbations on these ecosystems. It is suggested that stresses applied to ecosystems may produce somewhat similar eff'ects, regardless of their origin. Some of the large volume of literature con- cerning diversity and stability in terrestrial as well as aquatic ecosystems has been reviewed at a recent symposium (Woodwell and Smith, 383 FISHERY BULLETIN: VOL. 70, NO. 2 1969). One of the generalizations which appeared to emerge from the s^-mposium was that ecosystems characterized by high species diversity tended to be relatively stable. Elton (1958) has shown that if a predator has several alternate prey species to utilize, it will persist even if one of the prey becomes very scarce. Thus, it seems as if there is some correlation be- tween diversity and stability. In the case of marine invertebrates, there is some observational evidence (Paine, 1963) to indicate that marine predators at high levels in the food chain feed on more species of prey than do those on lower levels. Observations, such as the above mentioned, have led some ecologists to suggest that high level predators might contrib- ute more to community stability than the lower level predators. Removal of predators from rocky shore inver- tebrate communities (Paine, 1966) resulted in a reduction of the species diversity of the an- imal community. In addition, removal of graz- ing herbivores from rocky shores has resulted in the rapid growth of some of the formerly eaten plant species and a change in community com- position toward lower species diversity (Jones, 1948; Southward, 1964; Paine and Vadas, 1969). The observations and experiments of Paine ( 1966) also indicated that diversity among competing species of marine invertebrates could be decreased by removal of predators in some in- stances. A theoretical dynamic analysis (Par- rish and Saila, 1970) of a trophic subweb using Lotka-Volterra type interactions offered some support to Paine's conclusions. With the exception of some pioneering con- ceptual work by Larkin (1963, 1966) in describ- ing models for interspecific competition and ex- ploitation applied to natural fisheries, very little seems to have been done in an effort to predict the effects of man's activities on aquatic com- munities consisting of several interacting spe- cies. One of the reasons for this appears to lie in the degree of complexity required to establish and express all the basic interrelationships in such a system (Mann, 1969). Recently, Men- shutkin (1969) suggested graph theory as a use- ful tool for minimizing some of the difficulties of constructing models of interacting systems if certain simplifying assumptions, such as lin- earity and steady-state conditions, could be tolerated. Recognizing that any set of mathe- matical equations represents at best a crude ap- proximation of the actual behavior of complex ecosystems and that empirical values of coeffi- cients for complex models are largely unavail- able, we have proceeded under the assumption that the simplest models should first be explored and carefully evaluated before proceeding to more elaborate formulations. In addition, it is believed that the simplicity of the methods de- scribed herein may enhance their utility, espe- cially when considering the initial effects of ex- ploitation or environmental modification on in- teracting ecosystems. The objectives of this work were to: (a) in- troduce a subset of graph theory as used in net- work analysis; (b) describe a graph theoretic formulation of a basic ecological trophic unit, and to demonstrate some effects of predation and ex- ploitation on model ecosystems consisting of these units; and (c) demonstrate some other uses of graph theory with a view toward stim- ulating further interest in its applications. BACKGROUND AND DEVELOPMENT By definition, a graph is a set of vertices (nodes) connected by a set of edges (branches) . If the graph has polarity or direction, the edges have arrows, and the graph is said to be directed. In this report we are concerned only with direct- ed graphs. Two very simple directed graphs are illustrated in Figure 1. The ecological graphs utilized herein are based largely on graph theoretical techniques of net- work analysis, for which the theory has been clearly and concisely presented by Mason and Zimmermann (1960). To analyze a network, each edge connecting two vertices is given a co- efficient, a "transfer function" or "branch trans- mission." The "transmission" from one vertex A to a distant one C can then be expressed as a combination of these individual coefficients. The important principle is that the value of any ver- tex is the sum of the directed inputs, regardless of the outputs. In the very simple case of Fig- 384 SAILA and PARRISH: EXPLOITATION EFFECTS (a) ab D (b) Figure 1. — An illustration of two simple directed graphs. A "self-loop" is shown in part (b) of the figure. ure 1(a), the value of B is equal to the input from A plus the input from C: B = aA -^ cC. (1) Similarly, the value of C is equal to the only- input: C = bB. (2) Substituting B from Equation (1) into Equation (2) and solving for C gives: C = abA 1—bc (3) Thus the ratio of the value of C to the value of A is: C ■■ ab (4) 1 — be ' which is the transmission from A to- C. This type of term is used later as a biomass ratio or "trophic efficiency." It can easily be seen that the graph in Figure 1 simply represents a set of linear algebraic equa- tions. Such sets of equations can, of course, be solved classically. However, solution by inspec- tion of some graphs or parts of graphs is pos- sible. For example, in the graph of Figure 1 (a) , observe that: C = {a X b) X A + {c X b) X C. (5) The graph can be simplified to that shown in Figure 1 (b) . A "self-loop" has been created that has the effect of making the value of C to be _ , , L — ., __ — _ times what it 1 — loop transmission coefficient would have been without the loop. The situ- ation becomes only slightly more complex when the transmission from A to D is considered. The value of D can be obtained from the value of C in Equation (3) as: D eC = e abA be (6) Or, making use of the known effect of a self- loop, it is possible to simply see by inspection of Figure 1(b) that: D = {abe)A abA 1 — bc 1 — bc (7) These simple principles and techniques are considered adequate for formulating some use- ful trophic graphs. GRAPH OF A TROPHIC UNIT Graph theory has been applied to the analysis of a variety of problems in engineering, oper- ations research, and the social sciences (Berge, 1958; Busacker and Saaty, 1965; Kaufmann, 1967; Harary, 1969). Its use in biological sci- ences has been much more limited. However, Benzer (1959) and Maruyama and Yasuda (1970) have applied these concepts to genetics, and Landau (1955) and Trucco (1957) have used graph theory in describing animal behavior- al problems. Menshutkin (1969) appears to have been the first to apply graph theory to the study of communities of aquatic organisms. He used graph theory to derive expressions to illus- trate the relationship of the biomass of harvested organisms (fish) to primary production under 385 FISHERY BULLETIN: VOL. 70. NO. 2 specific conditions. He graphed the relationship between a prey and a predator as shown in Fig- ure 2(a). For simplicity, all the vertices (cap- ital letters) can be standardized in energy units (or energy per unit time). The lower case co- efficients are dimensionless constants or have a dimension of reciprocal time with values between zero and 1.0. Vertices and coefficients are listed in Table 1. Subscript 1 refers to the prey and subscript 2 refers to the predator. When an- other trophic level is added later, use of these subscripts can be easily extrapolated. For ex- ample, Di2 in Figure 2(a) is the amount of Prey 1 accessible to Predator 2; D24 in a later graph is the amount of Prey 2 accessible to Predator 4. The symbols used in Figure 2 are further de- fined in the following manner. Table 1. — Description of the vertices and coefficients utilized in model development. p = (net) production Q ^^ loss by "respiration" "respiration" coefficient Q = qB B = biomass M m = loss by natural mortality natural mortality coefficient M = mB F 1 ;_ loss by exploitation (fishing) exploitation coefficient F = IB U R = loss in undigested or unassimilcted food actual food ration y = (1 - k)R iR = food assimilated (gross production) k = digestion or assimilation coefficient P = kR - Q D-i2 dvi = accessible food accessibility coefficient Dvi = disiBi H12 = maximum ration (most a predator hv, = would ever consume) maximum ration coefficient H\3 = hisB'ji /Jia = a\2H\2 -\- ^i^Sia 6ii = D12 — H12 laia feeding coefficient ^-© (a) q,(-1)(1) 1-m2-f2 di2(1)bi2 hi2(-1)bi2 di2bi2 hi2(ai2-bi2) (b) (c) Figure 2. — Trophic graphs of Species 2 preying on Species 1. Part (a) illustrates Menshutkin's (1969) original formulation, and parts (b) and (c) represent the successive application of network analysis to obtain the basic trophic unit. 386 SAILA and PARRISH: EXPLOITATION EFFECTS P in the graph is what is usually called net pro- duction (an energy rate). Net production P, is equal to gross production, i.e., assimilated food, kR, minus respiration, Q. Summing inputs at the graph vertex P2, the value of P2 is: P2 = k2Ri2 — Q2 . (8) Respiration, Q, is expressed as the product of biomass and a "respiration coefficient," q: Q = qB, where B is the biomass (standing energy crop) of a species. M is natural mortality, considered propor- tional to biomass; Ml = niiBi, M2 = m2B2. (9) Constancy of these coefficients is assumed. F is fishing mortality which is used if an ex- ploited population is considered. Death due to any other specific cause can be separately con- sidered in a similar manner. U is energy in the undigested (unassimilated) portion of food eaten, and k is the "digestion coefficient." The predator assimilates a frac- tion k of the ration R eaten, and the remaining energy, U = (1 — k)R, is lost. Upon first inspection of the graph, it may be disconcerting to see vertices representing quan- tities such as biomass (energy) in the graph with vertices representing quantities such as production (energy per unit time). The con- fusion is resolved by realizing that the graph is not a pure flow network. It merely shows some assumed relationships, and at each vertex the same rules apply. For example, at vertex Bi in Figure 2(a) : Bi = ( + l)Pi + (—l)Ri2 + (1 — m,)5,, (10) or. Pi — Bi + Bi — miBi = R12. (11) Net production — Natural mortality = Re- mainder eaten by predators. Thus interpreted, the graph represents the re- lationships correctly. An important feature of this formulation is the attempt to approximate the density depen- dence of feeding rate. Formulations for species interactions such as the classic equations of Lotka and Volterra express rates of change of the number or biomass of a species as products of coefficients and numbers or biomass of the interacting species. This approach has involved the assumption that feeding rate is independent of the abundance of prey and it is an oversim- plification which results in an inherently un- stable system. The experimental work of Ivlev (1961) provided a density-dependent feeding re- lation: R = H {1 -VP ), (12) R where: H = the "maximum ration" of the predator, the most it would ever eat (or the maximum rate at which it would feed) no matter how much food were available; the "actual ration" of the pred- ator, the amount actually eaten (or the rate at which it feeds) under an actual condition of food availability; the density or biomass of the prey population; a coefficient. and P V A linear approximation of this relationship, fol- lowing Menshutkin (1969), can be used in the graph model. A parameter 812 is defined as: 812 = D12 — H12 , (13) where: D12 — the amount of the prey biomass accessible to the predator D12 = ^12^1. A constant of propor- tionality to prey biomass is as- sumed. H12 = as defined above. Since H12 is obtained as a fraction of pred- ator biomass {H12 = ^12^2), the assumption is introduced that all predator individuals feed at the same rate. The "actual ration" of the predator is then de- fined as: R12 =ai2//i2 + 612812 , (14) where ai2 and 612 are fractional coefficients. 387 FISHERY BULLETIN: VOL. 70, NO. 2 12 (16) (17) This linear expression is used to approximate the following modification of Ivlev's exponential relation: -ID R = H{1 — e ~^)' ^^^) where D and H are defined as before. The implementation of Equation (14) in the graph of Figure 2 (a) is seen by summing inputs at vertex 812: S12 = ( + l)Z>i2 + { — l)Hx2 = D12 — H and at vertex R12: R12 = ttviHio + bi2 812 . Although lacking in mathematical rigor, this linear approximation can be made to give rea- sonably accurate results over a limited range of prey density, and it is considered to be an improvement over the simple density-indepen- dent assumption. Figure 3 provides a sample comparison of an Ivlev exponential relationship according to Equation (15) with the linear ap- proximation of Equation (14). The coefficients a and b should, of course, be chosen in any real case to approximate either a desired analytic function of known utility or a set of data on feeding observations. The "network analysis" techniques described previously were applied to the graph of Figure 2(a) to produce the simplified graphs shown as Figure 2(b) and Figure 2(c). The derivations § 04 Z I s OS 10 ACCESSIBLE PHEY MAXIMUM RATION Figure 3. — Linear and exponential approximations of feeding behavior. used do not require Pi or mi. Furthermore, Mi, M2, and U2 cannot give "inputs" at any vertex since they are all directed outward. Therefore, the above parameters were eliminated with no effect on the solutions. Vertices P2 and Q2 were absorbed using graph theory network techniques to produce Figure 2(b). The same figure shows the similar absorption of vertices Dri, 812, and Hv2. Parallel inputs to a vertex can be combined. In this case, the two self-loops at vertex B2 were combined, and the two edges from B2 to R12 were combined. The resulting simplified graph, Fig- ure 2(c) , is the most basic graph that expresses the assumed relationships. The above formulation describes a two-species predation model where Species 2 preys on Spe- cies 1. At this trophic level, and for the form- ulations to be used, the term "predation" is applied in its broadest sense. Since the formula- tion does not make use of production, mortality or any other vital property of Species 1, Species 1 is really just a resource. It could be vegetation, or with some reinterpretation of coefficients, even living space. Clearly, the above graph is a building block from which a variety of more complex food webs can be constructed. Only limited applications of this concept are made in the following material, and its validity awaits the test of further applications. SOME MODELS AND THEIR INTERPRETATION Since relatively little observational informa- tion is available concerning the important prob- lem of community interactions, it was believed that a model study such as this might assist in a further understanding when additional ob- servational data are taken. Competitive and predatory interactions, with and without exploitation, were examined using trophic graphs made from the building block developed previously. Figure 4 shows Species 2 and Species 3 competing in their utilization of resource Bu A relation was derived for the ra- tio of the biomass of each of the competitors to that of the resource: B2/B1 and B3/B1. In either case this was done by writing the very simple linear equations for each of two vertices 388 SAILA and PARRISH: EXPLOITATION EFFECTS 1-1X12-012-^2 hi2(ai2-bi2) hi3(ai3-bi3) 1-m3-q3-f3 Figure 4.- -Trophic graph of Species 2 and Species 3 preying on Species 1. and solving them simultaneously in the classical manner. For competitor Species 2, Ri2 was writ- ten as the sum of its inputs in the graph, and B2 was written as the sum of its own inputs. In this case there are two equations in the three variables B2, Bi, and Rn. R12 was eliminated B2 k2d\2b\2V Bs ~ ksdisb.iV (18: V v = ma + qz + fs — kshisian - = ni2 + q2 + f2— k2hn{an - -&12) to give B2/B1 in terms of coefficients. B^/Bi was derived in a similar manner, and division gave the ratio of the biomass of the two com- peting species B2/B3 as follows: where: Inspection of Equation (18) reveals that if the two species compete exactly equally, or are ex- ploited equally, the ratio is unity. This is en- tirely the expected result. By giving one species or the other a competitive edge in one or an- other of the coefficients, it is apparent that the B2/B3 ratio can be changed. The simplest subweb involving predation on two competing species is shown in Figure 5. In this subweb Species 4 preys on Species 2 and Species 3, and Species 2 and Species 3 prey on Species 1. The procedure for deriving the ra- tios B2/B1 and B3/B1 was exactly as described above. That is, an equation was written for each l-mj-qj-fj l-m.-q.-f. l-rtij-qj-fs Figure 5. — Trophic graph of a 4-species subweb. In this case Species 4 preys on Species 2 and Species 3, and Species 2 and Species 3 prey on Species 1. 389 FISHERY BULLETIN: VOL. 70, NO. 2 vertex except Bi, and the equations were solved simultaneously. In this case there are seven equations, and the work of classical solution was not excessive. The result was found to be: B2 \V _k.,d,obr2iXY + Z) W' — k^dubniXY' + Z') (19) where: W = W = X = Y = Y' = Z = Z' = k3k4dl3hl3d34b3ih24 (Ci2i — 624) k2k4di2bv2d2ib2ih34{(iu — ^34) kih24ia2i — 624) + k4h34{a34 — ^34) W4 — <74 — A kshn (ai3 — bn) — (^34634 — W3 — Q3 k2hi2 (ai2 — 612) — 6^24624 — W2 — 92 k4d34b34h34iCl34 &34) k4d24b24h24 iCl24 &24) h h Questions of interest here were the effects on biomass ratios of the competitors as a function of various competitive coefficients and exploita- tion, and the difference in these effects with and without predation on the competitors. "Coef- ficient" values from Menshutkin (1969) were introduced for the coefficients for predation by Species 4 on Species 2 and Species 3 (the same coefficients for both — equal predation). Basical- ly the same coefficients were used for the com- petition of Species 2 and Species 3 as well. Coefficients were held constant except for the one whose effect was being considered. Using such values, the equations were simplified, and in most cases Species 3 was then given the nom- inal value of the competitive variable of interest while the value of that variable for Species 2 was made to vary above and below the nominal. This range of variation of Species 2 was expressed as the ratio coefficient 2/coefficient 3. The same process was performed for the earlier formula- tion without predation (Equation 18) . Thus ra- tios B2/B3 were obtained from both cases — with and without predation. A brief examination was made of the effect of various exploitation strategies on the relative stability of two model ecosystems, one with pre- dation and one without predation. These sys- tems are described by Equations (18) and (19), and stability was measured by the change in bio- mass. Figure 6 illustrates the results of various types of exploitation on the two systems. It is Figure 6. — Effects of predation and exploitation on mod- el ecosystem stability as measured by biomass ratios. Curve A illustrates a 4-species subweb in which there is no exploitation of the top predator. Curve B illus- trates a 3-species subw^eb with no top predator. Curve C illustrates a 4-species subweb with exploitation of the top predator as well as prey species 2 and 3. All nu- merical values of coefficients are from Menshutkin (1969). The nominal value of f^ was taken as 0.3. apparent from an examination of this figure that the most stable conditions examined involved predation as well as exploitation of the predator and the prey species. However, the system in- volving no top predator seemed to be more stable under exploitation of both prey species than the system involving predation, but with no exploi- tation of the top predator. For different types of competitive advantage of one species over the other, the effect of pre- dation on biomass ratios may be very different. Figure 7 demonstrates the effect of unequal com- petition in the coefficient d, which relates to the availability of the resource to Species 2 and Spe- cies 3. Without predation, the ratio B2/B3 of biomasses of the competitors is always the same as their d ratio. With predation, the ratio takes the much different form indicated. The values used for the dn/di3 ratio ranged from 2.7 to 0.37. This range of values produces a full range of B2/B3 ratios, from the point where Species 3 becomes extinct, to the point where Species 2 be- comes extinct. For the coefficient d, the results are not dependent upon the absolute value of d. 390 SAILA and PARRISH: EXPLOITATION EFFECTS Figure 7. — An illustration of the effect of predation (as measured by biomass ratios) on competition as mea- sured by changes in the ratio of the resource accessibility coefficient d for the two species. The results shown in Figure 7 clearly indicate that some of the competitive coefficients have a very large influence on the relative stability of interacting systems. They suggest that if the stress of exploitation or other environmental stresses interact with other model coefficients as, for example, in a simple predator-prey interac- tion, the system may respond very violently, with the rapid extinction of one or the other of the competing species. In some instances it may be desirable to have some rough empirical measure of the stability of exploited ecosystems consisting of interacting species. As Margalef (1969) has indicated, an adequate measure of community stability must include a measure of diversity as well as a mea- sure of persistence. Furthermore, Margalef at- tempted to formulate a generalized mathematical model for their interdependence. It is suggested that an additional application of graph theory may also be utilized to provide some en\pirical indication of stability and persistence of com- munities subjected to either environmental or exploitive stresses, assuming certain types of background information are available. Consider the following hypothetical example. Three communities of fishes (A, B, C) are sub- jected to various levels of exploitation. Assume that some crude index of diversity or community structure has been established which permits identification of the three communities as mu- tually exclusive groups. Assume that the three communities are sampled again during the course of a year, and that the frequency of sam- ples which resemble the previously defined com- munity as well as the frequency of samples which resemble the other two communities are listed. These frequencies can be displayed in the form of a network as shown in Figure 8. The data 3 2 130 60 10 85 — 1 50 200 100 M: 90 06 04 05 .65 .30 05 .10 .85 M' 8150 0970 0880 0925 .4555 .4520 0925 .1530 .7545 Po= [■ 50 200 100 G •] 8150 .0970 .0880 0925 .4555 4520 0925 .1530 .7545 68.50 111.25 170.25 G Figure 8. — Example of a hypothetical network showing the frequencies of samples resembling their initial struc- ture as well as those of the other two community struc- tures. In this example Fg is the matrix of frequencies at the end of the first sampling period, M is the cor- responding probability matrix, M^ is the square of the probability matrix and Pg is the vector of frequencies by community type. P^M^ is the matrix-vector product expressing the expected new frequencies by community type at the end of the second sampling period under the assumption that the probability matrix remains constant during the time interval. 391 FISHERY BULLETIN: VOL. 70, NO. 2 from the network is presented as a matrix (Fo of Figure 8) of frequencies wliich was normal- ized to form a probability matrix (M of Fig- ure 8). This probability matrix M is one in which the i, j entry gives the proportion of the samples from community Vi which resembled community Vj during the sampling period. An important theorem concerning probability mat- rices states that if B and C are probability mat- rices, so is their product BC. A corollary to this theorem states that if M is a probability matrix, then so is every power M", for any positive in- teger n. If the assumption is made that the probability matrix M remains constant over time, then if one knows the initial frequency matrix Fo and the probability matrix M, it is possible to find the sample frequency distribution at a subsequent time tn by finding the nth power of M and then forming the product PoM" where Po is the vector of row sums equal to the initial vector of frequencies by community type. In our case the frequency in year 2 is A = PoM". This matrix-vector multiplication is illustrated in the lower part of Figure 8. It is suggested that this derived frequency might be useful as the expected value basis for comparison with the sample observations made during subsequent years, if it is assumed the probability matrix remains constant over time, CONCLUSIONS The examples provided in this study are given primarily to illustrate the wide range of possi- bilities for the use of graph theory in studying the stability of interacting competitive and pred- atory relationships. The tentative results of this model study suggest that a nonselective exploi- tation strategy, which includes both predator and prey organisms, may be "best" from the point of view of maintaining community stability in complex ecosystems. The high desirability of obtaining experimental values for certain coef- ficients was also pointed out. The limitations of a linear, steady-state model are many and obvious. However, if such a model can sometimes be utilized to provide approximate results suitable for use in practical management at the early stages of marine ecosystem manage- ment, then the model is a worthwhile tool, and the method utilized has some merit. If the meth- od (graph theory) can be used not only to obtain some basic insight into system behavior but can also be used as an empirical tool, then it seems particularly worthwhile. Both these possibilities seem to await the results of future imaginative development. LITERATURE CITED Benzer, S. 1959. On the topology of genetic fine structure. Proc. Natl. Acad. Sci. 45:1607-1620. Berge, C. 1958. Theorie des graphes et ses applications. Dunod, Paris, 275 p. Beverton, R. J. H., AND S. J. Holt. 1957. On the dynamics of exploited fish popula- tions. Fish. Invest. Minist. Agric. Fish. Food (G.B.) Ser. II, 19, 533 p. BUSACKER, R. G., AND T. L. SAATY. 1965. Finite graphs and networks; an introduc- tion with applications. McGraw-Hill, N.Y., 294 p. Elton, C. S. 1958. The ecology of invasions by animals and plants. Methuen Monograph. J. Wiley, N.Y., 181 p. Harary, F. 1969. Graph theory. Addison-Wesley Publ. Co., Reading, Mass., 274 p. Ivlev, V. S. 1961. Experimental ecology of the feeding of fishes. (Translated from the Russian by D. Scott.) Yale Univ. Press, New Haven, 302 p. Jones, N. S. 1948. Observations and experiments on the biology of Datella vulgata at Port St. Mary, Isle of Man. Proc. Liverp. Biol. Soc. 56:60-77. Kaufmann, a. 1967. Graphs, dynamic programming, and finite games. (Translated by H. G. Sneyd.) Academic Press, N.Y., 484 p. Landau, H. G. 1955. On dominance relations and the structure of optimal societies, III: a condition for a score structure. Bull. Math. Biophys. 15:143-148. Larkin, p. a. 1963. Interspecific competition and exploitation. J. Fish. Res. Board Can. 20:647-678. 1966. Exploitation in a type of predator-prey re- lationship. J. Fish. Res. Board Can. 23:349-356. Mann, K. H. 1969. The dynamics of aquatic ecosystems. Ad. Ecol. Res. 6:1-81. 392 SAILA and PARRISH: EXPLOITATION EFFECTS Margalef, R. 1969. Diversity and stability : a practical proposal and a model of interdependence. In G. M. Wood- well and H. H. Smith (editors), Diversity and stability in ecological systems, p. 25-27. Brook- haven Symp. Biol, No. 22. Maruyama, T., and N. Yasuda. 1970. Use of graph theory in computation of in- breeding and kinship coefficients. Biometrics 26: 209-219. Mason, S. J., and H. J. Zimmermann. 1960. Electronic circuits, signals, and systems. J. Wiley, N.Y., 616 p. Menshutkin, v. V. 1969. [Graph theory applied to aquatic organism communities.] Zh. Obshch. Biol. 1969(1) :42-49. (Transl. available Natl. Tech. Inf. Serv., Spring- field, Va., as JPRS 47,689.) Paine, R. T. 1963. Trophic relationships of 8 sympatric preda- tory gastropods. Ecology 44:63-73. 1966. Food web complexity and species diversity. Am. Nat. 100:65-77. Paine, R. T., and R. L. Vadas. 1969. The effects of grazing by sea urchins Stron- gyloceutrohis spp., on benthic algae populations. Limnol. Oceanogr. 14:710-719. Parrish, J. D., AND S. B. Saila. 1970. Interspecific competition, predation and spe- cies diversity. J. Theor. Biol. 27:207-220. Richer, W. E. 1958. Handbook of computations for biological sta- tistics of fish populations. Fish. Res. Board Can., Bull. 119, 300 p. Southward, A. J. 1964. Limpet grazing and the control of vegetation on rocky shores. In N. J. Crisp (editor) , Grazing in terrestrial and marine environments, p. 265-274. Blackwells, Oxford. Trucco, F. 1957. Topological biology: a note on Rashevsky's transformation T. Bull. Math. Biophys. 19:19-21. Walters, C. J. 1969. A generalized computer simulation model for fish population studies. Trans. Am. Fish. Soc. 98:505-512. Woodwell, G. M., and H. H. Smith. 1969. Diversity and stability in ecological systems. Brookhaven Symp. Biol., No. 22. 393 NITROGEN EXCRETION BY ANCHOVY (ENGRAULIS MORDAX AND E. RINGENS) AND JACK MACKEREL (TRACHURUS SYMMETRICUS)' James J. McCarthy^ and Terry E. Whitledge^ ABSTRACT Teleost fish have been shown to excrete a variety of nitrogenous substances among which are ammonia, urea, and creatine. Previous reports show values for excretion of near- shore or bottom fish but not of pelagic species. Two species of anchovy and jack mackerel were placed in chambers and their nitro- genous excretion products were measured. Ammonia, urea, and creatine accounted for 82% of the total nitrogen excreted by Engraulis mordax and the identified fraction was 83% ammonia, 16%> urea, and 1% creatine. The significance of pelagic fish as a source of ammonia and urea in California coastal waters is discussed. On the basis of the major end product of their protein catabolism, animals are classified as am- monotelic, ureotelic, and uricotelic. Although these categories can be useful in evolutionary considerations (Baldwin, 1964), they are some- what arbitrary in that the excreta of most ani- mals contain a mixture of ammonia, urea, and uric acid. Mammals, elasmobranch fish, some amphibians, and some reptiles are considered to be ureotelic; teleost fish, some amphibians, and most invertebrates are ammonotelic; and birds and some reptiles are uricotelic (Baldwin, 1964). In ureotelic animals urea is produced via the ornithine cycle. It is, however, unlikely that this metabolic pathway is operative in non- ureotelic organisms such as some teleosts, which excrete substantial quantities of urea. Brown and Cohen (1960) have shown that no marine teleost has the enzymes necessary for the first two steps in the ornithine cycle. The complete complement of ornithine enzymes has been found in the coelacanth however (Brown and Brown, 1967). Arginase is present in the livers of teleost fish and hence dietary arginine has been sug- '■ Contribution #623 of the Department of Oceanog- raphy, University of Washington, Seattle, WA 98105. ^ Formerly at Scripps Institution of Oceanography, La Jolla, California 92037; present address: 213 Ma- caulay, Department of Earth and Planetary Sciences, The Johns Hopkins University, Baltimore, MD 21218. ^ Department of Oceanography, University of Wash- ington, Seattle, WA 98105. gested as a source of urea (Hunter and Dauph- inee, 1924-1925; Hunter, 1929), but this sug- gestion is considered unlikely since arginine is an essential amino acid for this group of or- ganisms (Forster and Goldstein, 1969). Purine catabolism also has been suggested as a means of urea formation via uricolysis (Brunei, 1937), and in support of this Goldstein and Forster (1965) found uricolytic activity in the livers of five species of marine and freshwater teleosts. Delaunay (1929), Grafflin and Gould (1936), Grollman (1929), and Smith (1929) determined the composition of the urine of eight species of marine teleosts and their results were summa- rized by Scheer and Ramimurthi (1968). The proportions of various components of the total nonprotein urinary nitrogen varied greatly within, as well as between, species (ammonia varied from 0.5 to 9.6?f , urea varied from 0.1 to 30.8%, creatine varied from 6.5 to 61.7%, and amino-N varied from 4.0 to 21.4%). In 1929 Smith used a divided chamber to permit separate determinations of the nitrogen released by the gills and by the kidneys, and his results showed that essentially all of the ammonia and urea released originated from the gills. In ad- dition to these compounds, the branchial excreta consisted of amine or amine oxide derivatives while the less diffusible nitrogenous end products such as creatine, creatinine, and uric acid were excreted solely by the kidneys. It is therefore Manuscript accepted December 1971. FISHERY BULLETIN: VOL. 70, NO. 2, 1972. 395 FISHERY BULLETIN: VOL. 70. NO. 2 apparent that a release rate with ecological sig- nificance would have to consider both branchial and renal release. Since the only data available from which one can calculate nitrogenous release rates for ma- rine teleosts are for nearshore or bottom fish such as the sculpin, the starry flounder, and the blue sea perch (Wood, 1958) , the following work was undertaken using significant pelagic species in order to assess the importance of marine teleosts as a source of ammonia, urea, and cre- atine in the euphotic zone. METHODS Experiments with the northern anchovy, Engraulis mordax, were conducted in August 1970, in the laboratories of the Fj'shery-Ocean- ography Center, La Jolla, California. Between experiments the fish were kept in a tank of flow- ing seawater and, unless specified otherwise, were fed frozen brine shrimp daily. For each experiment ten fish were placed in a 32-liter circular Plexiglas chamber similar to that de- scribed by Lasker (1970). During the first ex- periment the chamber was darkened and 24 °C seawater flowed through it at a constant and de- termined rate. Beginning at the time the fish were introduced, effluent was sampled at 10-min intervals for 40 min. The fish had not been fed for 24 hr. For experiments 2 and 3 the flowing system was not used. The chamber was filled initially with seawater and during the ex- periments was exposed to room light. The tem- perature during experiment 2 was 24°C and water samples were taken at 10-min intervals for 40 min. The fish had been fed 30 min prior to their placement in the chamber. The temper- ature during experiment 3 was 21.5°C and water samples were taken at 10-min intervals for 70 min. The fish had not been fed for 48 hr and ju.st after the 30-min sampling, 2.49 g of frozen brine shrimp were thawed and added to the chamber (the first and second portions of ex- periment 3 will be referred to as 3a and 3b). An appropriate control was run to determine the eff'ect of the brine shrimp on the ammonia, urea, creatine, and total nitrogen concentrations. Ten fish from the group used for experiments 1 and 2 and the ten used for experiment 3 were sacrificed for weight determinations. Wet weight determinations were made after blotting the specimens on filter paper, and dry weight was determined after they had been dried at 60 °C for 96 hr. Experiments with the Peruvian anchovy, Engraulis ringens, were conducted off" the coast of Peru in April 1969, during the RV Thompson cruise 36. The methods of collection were de- scribed by Whitledge and Packard (1971). Re- lease rates for urea and ammonia were deter- mined for each of three fish in separate one-liter volumes of 0.45/a m Millipore^ filtered seawater at 15°C. Each experiment was run until the animal died (approximately 90 min). Experiments with the jack mackerel, Tra- churus symmetricus, were conducted off the coast of California in July 1970, during a cruise of the RV Alpha Helix for the Institute of Marine Resources Food Chain Research Group. Of three specimens caught with lure and line, two were placed in a 42-liter Plexiglas deck tank which was continually flushed with surface sea- water, and the third, which had been injured when caught, was sacrificed for a weight deter- mination. After approximately 5 hr the fish were transferred to a similar tank recently cleaned and filled with 19°C surface seawater which had been filtered through a 173 /um nylon mesh. Samples for ammonia and urea analyses were taken at the beginning and at the conclusion of the experiment 2 hr later. The fish were kept another 20 hr and released. For samples from the E. ringens experiments ammonia was determined by a modified form of Johnston's (1966) rubazoic acid method and urea by a urease hydrolysis coupled with the rubazoic acid method. In all other experiments ammonia was determined by the phenolhypo- chlorite method (Solorzano, 1969) and urea by the urease method (McCarthy, 1970). All de- terminations were made in duplicate immedi- ately after each experiment. Creatine concen- trations were measured using a fluorescent com- plex with alkaline ninhydrin (Whitledge and Dugdale, in press). Total nitrogen was deter- mined with the ultraviolet oxidation technique 396 McCarthy and whitledge: nitrogen excretion of Armstrong, Williams, and Strickland (1966) as described by Strickland and Parsons (1968), but modified by the addition of sodium hypo- chlorite (chlorine bleach, 5,25% active ingre- dients) . One ml sodium hypochlorite per 100 ml sample was found to increase the percentage of conversion of creatine to nitrate. If added in excess, the sodium hypochlorite will react with the hydrogen peroxide to release oxygen. The conversion of a standard solution of creatine to nitrate in the digester was 43.4%, but the addi- tion of sodium hypochlorite increased this val- ue to 95.0%. Urea standards showed 96-100% conversion without addition of sodium hypo- chlorite. All chambers were washed with either fresh or distilled water immediately prior to each ex- periment and experimental periods were kept to a minimum in an effort to reduce possible effects of bacteria. Antibiotics interfere with both of the ammonia methods mentioned above, and could not be used to reduce bacterial ac- tivity. Wood (1958) showed that in fish excre- tion experiments running as long as 24 hr, bac- teria on the surface of the fish do not affect the results and the concentrations of released nitrogen compounds were unchanged for an ad- ditional 24 hr. The data were analyzed statistically using the Mann Whitney U and the Tukey-Siegel tests to compare the medians and variabilities respec- tively. Significant differences between exper- iments were at the 0.05 level. RESULTS The release rates for ammonia, urea, creatine, total nitrogen and the ammonia-urea ratios for all experiments are given in Table 1. The re- sults of the E. mordax experiments represent the mean of the 10-min interval samples for each experiment. An example of the data ob- tained in one of these experiments is shown in Figure 1. At the conclusion of the first E. mor- dax experiment all of the fish had acquired dark- ened dorsal coloration and one was locked into the panic response described below for E. ring ens. Immediately after release into the large holding tank, food was added and all ex- cept the one panicked fish, which died a few hours later, fed normally and regained normal coloration. During the other experiments with E. mordax, the fish behaved normally and re- tained their normal coloration. The fish in the E. mordax experiment 3b were restricted from feeding in their normal frenzied manner because of the size of the chamber and at the con- clusion a small portion of the food remained uneaten. Creatine and total nitrogen values for the second E. mordax experiment indicate that Table 1. — Excretion of nitrogenous compounds by Engraulis viordax, E. ringens, and Trachurus syimnetricus. /tg at N/mg dry wt/day Species Ammonia Urea Creatine Total nitrogen Ammonia-N Urea-N Engraulis mordax starvecJ one day (I) 0.074 0.020 3.70 fed be- fore ex- periment (2) 0.185 0.036 0.003 0.273 5.14 starved 2 days (3a) 0.055 0.021 0.002 2.62 feeding (3b) 0.076 0.023 0.003 3.45 Engraulis ringens (1) 0.240 0.083 0.118 0.507 2.90 (2) 0.281 0.089 0.111 0.700 3.16 (3) 0.171 0.057 0.104 0.346 3.00 Trachurus symmetricus 0.090 0.017 5.29 397 FISHERY BULLETIN: VOL. 70, NO. 2 1 2.00 r 10 20. TIME 30. 40. ( MIN.) Figure 1. — Nitrogen excretion by 10 specimens of En- graulis mordax in experiment 2. Specimens were fed 30 min before experiment. ammonia, urea, and creatine accounted for ap- proximately 82.7, 16.0, and 1.3Sr, respectively, of the sum of the ammonia, urea, and creatine values. This sum was approximately 82% of the total nitrogen released. The fish used in the E. mordax experiments 1 and 2 had a mean wet weight of 13.9 g and a mean dry weight of 3.9 g per animal; those used in the third exper- iment had mean wet and dry weights, respectiv- ly, of 15.2 g and 4.3 g per animal. In the E. ringens experiments, the sum of the individually measured nitrogen compounds ac- counted for 69 to 96 9*^ of total nitrogen excreted. Relative amounts of ammonia, urea, and creatine averaged 44.6, 14.5, and 21.4%, respectively, while 19.5% was unidentified. Although the specimens of E. ringens appeared in good health for the first few hours after capture, their condi- tion deteriorated rapidly after the experiment began. One lived for 83 min, another for 85 min, and the third for 105 min. The behavior pat- tern was a panic response in which the animals tried repeatedly to swim downward into the bot- tom of the container and eventually did consider- able damage to their heads, resulting in broken blood vessels in their eyes and nasal regions. The dorsal coloration changed from a greenish- grey to jet black during the deterioration, and once an animal locked into this behavior pattern there was apparently no way to reverse it and death was inevitable. The mean wet weight and dry weight per fish for these experiments was 7.0 and 1.9 g, respectively, and nitrogen was found to be 10.9% of dry weight. The results for the T. symmetriciis represent a single value for both fish in the same chamber. Prior to, during, and in the 20 hr after the ex- periment the fish swam about the chamber in a calm manner. The wet weight of the sacri- ficed fish was 190.0 g. DISCUSSION The fish in the E. mordax experiment 2 were the most recently fed, and Figure 1 shows that urea, ammonia, and creatine were released at approximately constant rates over the experi- mental period. Table 1 shows that these release rates for ammonia, urea, and the amm.onia-urea ratio were the highest of all the E. mordax ex- periments. The urea release rates for the other experiments (1, 3a, and 3b) were all similar, but the ammonia release rates and the ammonia- urea ratios appear to be related to the length of the starvation period. For the E. mordax experiments, the statistical tests indicated that the ammonia release rates in experiments 1 and 3a and in experiments 1 and 3b were not significantly different within each pair with respect to central tendency. On the other hand there were significant differences between the pairs of ammonia release rates in experiments 1 and 2, 2 and 3a, and 2 and 3b. The variabilities for ammonia release rates for all of the experiments were similar. With re- gard to the urea release rates, only those for experiments 1 and 2 were significantly different with respect to location of central tendency and none of the experiments differed from each other with respect to variability. These statistics imply that the effect of feeding is rapidly apparent in the ammonia release rates while it appears more slowly and to a lesser de- 398 McCarthy and whitledge: nitrogen excretion gree in the urea release rates; this is clearly reflected in the ammonia-urea ratios. When the period of starvation was the greatest (experi- ment 3a) the ratio was the lowest, and when the starvation was least the ratio was the great- est. The ratio for experiment 3b presumably would have increased to a value comparable to that for experiment 2 when the time since feed- ing had become equal for both. Upon thawing, the brine shrimp which were used to feed E. mordax liberated ammonia, urea, and creatine. If these substances were retained in large quantities by the shrimp, the results of experiments 2 and 3b could have been af- fected. Such interference is, however, unlikely since only the ammonia release rate increased with feeding in experiment 3. If this increased rate of release had resulted from ammonia lib- erated from the shrimp after ingestion by the fish, a comparable increase in the rate of urea release would have been expected since the quantities of urea liberated by the shrimp were approximately equal to those of ammonia. It is important to note that since experiment 3 was conducted at a temperature 2.5 °C lower than the other experiments, a correction of the measured rates should be made in order to com- pare them properly. Since an appropriate Qio value was not available, the correction was not made. Presumably, however, a Qio for an ammo- nia release rate would be similar if not identical to that for urea, and the ammonia-urea ratio would not be changed with a temperature cor- rection. The lack of both a temperature cor- rection and an appropriate relationship between body size and nitrogen excretion also makes the comparison of data between different species dif- ficult. The E. ringens experiments produced the high- est release rates (particularly for creatine) and the lowest ammonia-urea ratios. This can pro- bably be attributed to the high level of activity and/or the poor health of the specimens. Wood (1958) ran experiments for 24 hr at 12°C, and from his Table 2 and an approxima- tion of 28% wet weight = dry weight, one can calculate ammonia and urea release rates. These calculated mean release rates are 0.0141, 0.0179, and 0.0065 fjcg at ammonia-N + urea-N/mg dry weight/day, for the sculpin {Leptocottus armcu- tus) , the starry flounder (Platichthys stellatus) , and the blue sea perch {Taeniotoca lateralis) re- spectively. These rates are nearly an order of magnitude lower than those reported here, but a temperature correction (assuming a Qio of 2) would bring them within approximately a factor of five. The calculated ammonia-urea ratios for the sculpin, the starry flounder, and the blue sea perch are 3.09, 7.13, and 1.26 respectively. Since the fish used in Wood's studies were nearly ten times larger than the anchovies, were maintained for some time prior to the experiments on a diet of lingcod muscle, and for the experiments were enclosed in chambers barely larger than the fish, it is difl^cult to compare the results of the dif- ferent sets of data. Ammonia and urea are important plant nu- trients and it is of interest to examine the sig- nificance of fish excretion as a source of these substances in the sea. Whitledge and Packard (1971) estimated that nitrogen excretion by the herbivorous E. ringens in the near surface waters of the Peru Current is an order of mag- nitude greater per unit volume of water than zooplankton excretion and they suggested, on the basis of these rates and measured rates of nitrogen uptake by phytoplankton, that fish ex- cretion may be a major source of the ammonia utilized by phytoplankton in this area. Oflf the coast of southern California the con- tribution of the fish community in the regenera- tion of ammonia and urea can also be estimated. Integrated values of phytoplankton nitrogen utilization at three stations in the euphotic zone off the coast of San Diego (Stations 1, 4, and 6, McCarthy* averaged 0.073 /xg at ammonia-N/li- ter/day and 0.066 /xg at urea-N/liter/day. For the area included in the California Cooperative Oceanic Fisheries Investigations (CalCOFI) survey, the total biomass of the most common species of near-surface fish [northern anchovy (Engraidis mordax), Pacific hake (Merluccius productus) , jack mackerel (Trachurus symmet- 7'icus) , Pacific saury {Cololabis saira), and al- * McCarthy, J. J. The uptake of urea by natural pop- ulations of marine phytoplankton. Manuscript in prep- aration. 399 FISHERY BULLETIN: VOL. 70, NO. 2 bacore {Thunniis alalunga)] is estimated at 20-25 X 10" metric tons wet weight (Dr. P. E. Smith, personal communication). Using an area of 70 x 10^°m-, a depth of 140 m, an average excretion rate from Table 1 (E. mordax, exper- iments 1 and 2, and T. symmetricus) , and a conversion factor (dry wt = 28 9r wet wt) ap- proximate production rates of 0.0075 /x.g at ammonia-N liter/day and 0.0015 /xg at urea-N/ liter day can be calculated. These rates would account for 10 '"f of the ammonia and 2% of the urea utilized by the phytoplankton. Other investigators— Harris, 1959; Dugdale and Goering, 1967 and 1970; and Martin, 1968 — have attempted to balance ammonia utiliza- tion by phytoplankton and excretion by zoo- plankton in Long Island Sound, the Bermuda re- gion, the Peru Current and Narragansett Bay respectively. Further calculations can be made for the area off southern California to compare the significance of fish ammonia and urea excre- tion to that of zooplankton. A zooplankton stand- ing crop estimate of 0.125 mg dry wt/liter was calculated from 10 years of data collected in the California Current as part of the CalCOFI pro- gram by multiplying the mean catch by a factor of 3 to compensate for the biomass of the smaller zooplankton lost through the 0.505 mm mesh (Dr. P. E. Smith, personal communication). Using average excretion rates for recently fed zoo- plankton {Calaniis helgolandims, Calanus chil- enis, and Clausocalanus sp.) of 0.73 ixg at ammo- nia-N/mg dry wt/day and 0.36 fj-g at urea-N/mg dry wt/day (McCarthy, 1971) average regener- ation rates of 0.090 (xg at ammonia/liter/day and 0.045 /jLg at urea-N/liter/day can be cal- culated. These rates would account for 123% of the ammonia and 68 Sf of the urea uti- lized per day. If on the other hand, micro- zooplankton and zooplankton biomass data col- lected from April through September 1967 in the same approximate area as two of the three sta- tions used for the phytoplankton utilization cal- culations are applied (Beers and Stewart, 1970; Mullin and Brooks, 1970) the calculated regen- eration would be 0.020 ixg at ammonia-N/liter/ day and 0.010 /xg at urea-N/liter/day. These rates account for 27*"/ of the ammonia and 15% of the urea utilized by the phytoplankton. Hence, the fish contribution would amount to 8-27% of the ammonia and 3-13 ^r of the urea released by both groups of organisms. There undoubtedly are, however, situations in which large fish or schools of fish (the mean density of an E. mordax school is estimated as the equivalent of 1,300 10-cm fish/m^ Dr. P. E. Smith, personal com- munication) are more important than zooplank- ton in supplying ammonia and urea to a partic- ular parcel of water. Obviously these calculations are based on many simplified assumptions. Other than the fact that the zooplankton and fish biomass esti- mates are averages for larger areas and longer periods of time than can be represented by the phytoplankton utilization rates, perhaps the most poorly based assumption is the application of mean excretion rates for three zooplankton spe- cies and two fish species to the entire zooplankton and fish populations. More reliable estimates of excretion rates are needed for smaller species of zooplankton and larger species of fish. ACKNOWLEDGMENTS We are grateful to Dr. 0. Holm-Hansen for the Trachurus symmetriais specimens, to Dr. Reuben Lasker for the Engraulis mordax spec- imens, laboratory facilities, and helpful advice and to Drs. Paul E. Smith and John A. McGowan for advice and encouragement. This work was supported by Federal Water Quality Administra- tion Grant 16010 EHC to Dr. R. W. Eppley, and the National Science Foundation under grants GB-8648 and GB-18568 to the University of Washington and GB-24816 to Scripps Institution of Oceanography for operation of the Alpha Helix Research Program. LITERATURE CITED Armstrong, F. A. J., P. M. Williams, and J. D. H. Strickland. 1966. Photo-oxidation of organic matter in sea water by ultra-violet radiation, analytical and other applications. Nature (Lond.) 211:481-483. Baldwin, E. 1964. An introduction to comparative biochemistry. 4th ed. Cambridge Univ. Press, Lond., 179 p. 400 McCarthy and whitledge: nitrogen excretion Beers, J. R., and G. L. Stewart. 1970. Numerical abundance and estimated biomass of microzooplankton. In J. D. H. Strickland (edi- tor) , The ecology of plankton off La Jolla, Cal- ifornia, in the period April through September, 1967, p. 67-87. Bull. Scripps Inst. Oceanogr. Univ. Calif. 17. Brown, G. W., Jr., and S. G. Brown. 1967. Urea and its formation in coelacanth liver. Science (Wash., D.C.) 155:570-573. Brown, G. W., Jr., and P. P. Cohen. 1960. Comparative biochemistry of urea synthesis. 3. Activities of urea-cycle enzymes in various higher and lower vertebrates. Biochem. J. 75:82- 91. Brunel, a. 1937. Metabolisme de I'azote origine purique chez les Poissons et les Batraciens. 1. — Catab- olisme de I'azote d'origine purique chez les Selaciens. Bull. Soc. Chim. Biol. 19:805-826. Delaunay, H. 1929. Sur I'excretion azotee des poissons. C. R. Soc. Biol. 101:371-372. Dugdale, R. C, and J. J. Goering. 1967. Uptake of new and regenerated forms of nitrogen in primary productivity. Limnol. Ocean- ogr. 12:196-206. 1970. Nutrient limitation and the path of nitrogen in Peru Current production. Anton Bruun Rep. 4:53-58. Texas A and M Press. FoRSTER, R. P., and L. Goldstein. 1969. Formation of excretory products. In W. S. Hoar and D. J. Randall (editors) , Fish physiology, Vol. I, p. 313-350. Academic Press, N.Y. Goldstein, L., and R. P. Forster. 1965. The role of uricolysis in the production of urea by fishes and other aquatic vertebrates. Comp. Biochem. Physiol. 14:567-576. Grafflin, a. L., and R. G. Gould. 1936. Renal function in marine teleosts. II The nitrogenous constituents of the urine of sculpin and flounder, with particular reference to trimeth- ylamine oxide. Biol. Bull. (Woods Hole) 70:16-27. Grollman, a. 1929. The urine of the goosefish (Lophius pisca- torius) : Its nitrogenous constituents with special reference to the presence in it of trimethylamine oxide. J. Biol. Chem. 81:267-278. Harris, E. 1959. The nitrogen cycle in Long Island Sound. Bull. Bingham Oceanogr. Collect. Yale Univ. 17:31-65. Hunter, A. 1929. Further observations on the distribution of arginase in fishes. J. Biol. Chem. 81:505-511. Hunter, A., and J. A. Dauphinee. 1924-25. Quantitative studies concerning the dis- tribution of arginase in fishes and other verte- brates. Proc. R. Soc, Ser. B 97:227-242. Johnston, R. 1966. Determination of ammonia in seawater as rubazoic acid. I.C.E.S.C.M. Hydrog. Comm. N:10. Lasker, R. 1970. Utilization of zooplankton energy by a Pa- cific sardine population in the California Current. In J. H. Steele (editor), Marine food chains, p. 265-284. Oliver and Boyd, Edinburgh. Martin, J. H. 1968. Phytoplankton-zooplankton relationships in Narragansett Bay. III. Seasonal changes in zoo- plankton excretion rates in relation to phyto- plankton abundance. Limnol. Oceanogr. 13:63-71. McCarthy, J. J. 1970. A urease method for urea in seawater. Limnol. Oceanogr. 15:309-313. 1971. The role of urea in marine phytoplankton ecology. Ph.D. Thesis. Scripps Institution of Oceanography, La Jolla, Calif., 165 p. MuLLiN, M. M., and E. R. Brooks. 1970. Production of the planktonic copepod, Cal- anus helgolandicus. In J. D. H. Strickland (ed- itor). The ecology of plankton off' La Jolla, Cal- ifornia, in the period April through September, 1967, p. 89-103. Bull. Scripps Inst. Oceanogr. Univ. Calif. 17. SCHEER, B. T., AND R. RAMIMURTHL 1968. Nonprotein nitrogen in urine. In P. L. Alt- man and D. S. Dittmer (editors), Metabolism, p. 540-547. Fed. Am. Soc. Exp. Biol., Bethesda, Md. Smith, H. W. 1929. The excretion of ammonia and urea by the gills of fish. J. Biol. Chem. 81:727-742. SOLORZANO, L. 1969. Determination of ammonia in natural waters by the phenolhypochlorite method. Limnol. Ocean- ogr. 14:799-801. Strickland, J. D. H., and T. R. Parsons. 1968. A practical handbook of seawater analysis. Fish. Res. Board Can., Bull. 167, 311 p. Whitledge, T. E., and R. C. Dugdale. In press. Creatine in sea water. Limnol. Oceanogr. Whitledge, T. E., and T. T. Packard. 1971. Nutrient excretion by anchovies and zoo- plankton in Pacific upwelling regions. Invest. Pesq. 35:243-250. Wood, J. D. 1958. Nitrogen excretion in some marine teleosts. Can. J. Biochem. Physiol. 36:1237-1242. Wood, J. D. 1958. Nitrogen excretion in some marine teleosts. Can. J. Biochem. Physiol. 36:1237-1242. 401 ELECTROPHORETIC INVESTIGATION OF THE FAMILY SCORPAENIDAE Allyn G. Johnson, Fred M. Utter, and Harold O. Hodgins' ABSTRACT Thirty-one species of three genera of the family Scorpaenidae were separated into 17 groups based on starch gel electrophoretic comparison of muscle proteins and six enzy- matic systems. This study concluded that relatively greater similarity existed between the Pacific Sebastes and the Atlantic Sebastes than between either and the other genera. Ten of the 27 species of Pacific Sebastes tested had unique biochemical profiles which may be useful for identification of specimens. The family Scorpaenidae contains several ge- nera in the Pacific and Atlantic Oceans. On the Pacific coast of North America there are four genera — Sebastes," Sebastolobus, Scorpaena, and Scorpaenodes. The genus of Pacific Sebastes contains over 50 species (Tsuyuki et al, 1968). In this genus new species and ex- tensions of known distribution ranges have been described in recent years (Westrheim, 1965; Westrheim and Tsuyuki, 1967; Nishimoto, 1970; Tsuyuki and Westrheim, 1970). At present there are difficulties in showing taxonomic re- lations and, in some instances, in making positive identification of specimens using morphometric and meristic methods, although taxonomic re- lations can be obtained by biochemical methods. Starch gel electrophoresis — developed by Smith- ies (1955) — coupled with histochemical proce- dures (Hunter and Markert, 1957) is one of the best biochemical techniques for taxonomic studies. Scorpaenid muscle proteins and hemoglobin were investigated by starch gel electrophoresis by Tsuyuki et al. (1968). They suggested that the electrophoretic evidence did not support the separation of the two genera Sebastodes and Sebastes. Chu (1968), using disk electrophore- sis of muscle proteins, found different patterns ^ National Marine Fisheries Service, Northwest Fish- eries Center, Seattle Laboratory, 2725 Montlake Boule- vard East, Seattle, WA 98102. ^ In this paper we follow the designation of Bailey (1970) and Chen (1971), considering Sebastodes as Se- bastes. Members of the genus Sebastes that were col- lected along the Pacific Coast of North America are sig- nified in this paper as Pacific Sebastes. in two out of eight species of Sebastodes. Altuk- hov and Nefyodov (1968) demonstrated serum protein diff'erences between Sebastes marinus and S. mentella using agar gel electrophoresis. This paper reports the findings of our inves- tigation of proteins and six enzyme systems found in the skeletal muscle or liver of members of the family Scorpaenidae. Our study involved 27 species of Pacific Sebastes, 2 of the Atlantic Sebastes, and 1 each of Sebastolobus and Helico- lenus. We present information on the relative biochemical similarity between genera and a key which separates 10 of the 27 Pacific Sebastes species studied. This was not a genetic study per se but a research which demonstrated repeat- able biochemical diff"erences between species. The observed constancy of biochemical charac- ters examined within a species in samples taken at diflferent ages, depths, and geographic loca- tions is evidence that the reported diff'erences between species are, indeed, genetic. Alternate explanations for such repeatable expression of proteins under the above conditions seem much less likely. MATERIAL AND METHODS Sampling data including location, species, and number of individuals collected are given in Table 1. Most samples were frozen quickly after cap- ture, but in some instances were kept on ice for short periods; all samples were kept frozen at — 20 °C after receipt at the laboratory until tested. Extracts were prepared by mixing equal Manuscript accepted November 1971. FISHERY BULLETIN: VOL. 70, NO. 2, 1972. 403 FISHERY BULLETIN: VOL. 70, NO. 2 Table 1. — Location and number of specimens of Scor- paenidae collected, 1968-70. Location^ Total Species ■ — number ABGDEFof fish Pacific Sibastes S. aleutianus 10 — 6 — -- — 16 S. alutus 217 - 843 .. „ — 1,060 S. auriculatus — 76 — — — — "76 S. aurora 3 — — — — — 3 S. br^vispinis 5 — 40 „ — — 45 S. catnaematicus — — — 3 3 S. caurinus .. 283 — — — — 283 S. chlorostictu! — — — 1 — ' S. crameri 2 — 16 — — — 18 S. diploproa 14 — — — — — 14 S. tlongatui .. 297 96 .. .. - 393 S. entomilas 2 — 2 ._ — — 4. S. flavidus 8 — — — — — 8 S. helvomaculatus 5 .. 19 .. ._ ._ 24 S. levii — — — — 1 — 1 S. malisfr .. 25 .. ._ — — 25 S. milanops __ 28 — — — — 28 S. paucispinis 2 15 — 1 — 18 S. pinniger „ — 24 .. ._ __ 24 S. proriger 9 — 100 .. - - 109 S. reedi 1 — 110 .. - - HI S. rvbtrrimus 5 27 5 .. — — 37 S. rubivinctus 5 -- 34 „ ._ -- 39 S. saxicola 5 — -- ._ 1 _- 6 S. vjihoni — 1 — — — ' S. variegatus — — — — 1 ' S. zacentrus 1 — 37 — — — 38 Atlantic Sfbastes S. marinus — 9 — — 9 S. viviparous — 10 — — 10 Sebaitolobus alascanus „ — 100 ._ — — 100 Heticottnus dactylopterus -. — — 10 — — 10 1 A = Pacific Coast of Washington and Oregon, 1968-70; B = Puget Sound, Wash., 1968-70; C = Queen Charlotte Sound, B.C., Canada, June 1970; D = West Coast of Britain and Ireland, August 1970; E = Avila Beach, Calif., October 1970; F = Cape Ommaney, Alaska, April 1970. volumes of tissue and phosphate-buffered physi- ological saline solution (pH 7.4) into uniform pastes with glass rods. The extracts were tested by electrophoresis without further treatment by (1) drawing them into 1/4,-inch X 3/16-inch filter paper inserts (Schleicher and Schuell grade S and S No. 470)', placed on the surface of the tis- sue-saline mixture, and (2) placing the inserts into starch gels. Electrophoresis in starch gel followed the methods of Kristjansson (1963). All but two of the biochemical systems were resolved well using a buffer system described by Markert and Faulhaber (1965). Lactate dehydrogenase and ' Reference to trade names in this publication does not imply endorsement of commercial products by the Na- tional Marine Fisheries Service. phosphoglucomutase were best resolved by using the buffer system described by Ridgway, Sher- burne, and Lewis (1970) . Gels consisted of 35 g starch plus 250 ml of buflfer. A voltage of 300 was applied for 10 min; sample inserts were removed and 400 v applied until indicator dye markers reached a point 6 to 9 cm anodal to the origin. The gels were cooled during electro- phoresis by placing ice packs on glass plates on top of the gels. After electrophoresis, bands re- flecting enzyme activity were detected by the fol- lowing methods: Tetrazolium oxidase (TO) (after Brewer, 1967, and Johnson, Utter, and Hodgins, 1970b) : 5 mg phenazine methosulfate (PMS) 3 mg p-nitro blue tetrazolium (NET) 40 ml tris-citrate buffer (0.03 M tris, 0.005 M citric acid, pH7.0) L-alpha-glycerophosphate dehydrogenase (aGPDH) (after Nyman, 1967, and Johnson, Utter, and Hodgins, 1970a) : 5 mg PMS 3 mg NET 5 mg NAD + 100 mg L-alpha-glycerophosphate 40 ml tris-citrate buffer Lactate dehydrogenase (LDH): 10 mg PMS 5 mg NET 5 mg NAD + 20 ml of 0.5 M sodium lactate solution 40 ml tris-citrate buffer Peptidase A (after Lewis and Harris, 1967, and Lewis and Truslove, 1969) : 10 mg DL valyl-leucine 1 mg horseradish peroxidase 5 mg 0-dianisidine in 10 ml acetone 0.5 ml M MgCla 1 mg Bothrops atrox venom 40 ml tris-citrate buffer Phosphoglucomutase (PGM) (after Spencer, Hopkinson, and Harris, 1964) : 100 mg glucose-1-phosphate (dipotassium salt) 5 mg NADP 5 mg PMS 3 mg NET 20 units glucose-6-phosphate dehydrogenase 404 JOHNSON, UTTER and HODGINS: ELECTROPHORETIC INVESTIGATION Table 2. — Classification of species of Sec irpaenidae into various groups by means of biochemical characteristics. Biochemical characteristics Species Muscle Peptidase A 'Biochemical TO GPDH 1 DH - group pattern^ LLTl 1 (Fast zone) II (SI ow zone) Pacific Sebastes S. elongatus 2 F E c c c \» S. tntomelas 2 S E c c c l|2 S. aurora 3 F E c b d ll|2 S. chloroitictus 4 F E B a c IVa S. aleutianus A F E C c e V S. zacentrus 4 F E C c c V S. caurinus 4 F F, S C d c VI S. diptoproa 4 F E C b b Vila S. helvomacutatus 4 F E B c c VHP S. maliger 4 F F C d c VI S. Tubirrimus 4 F E C c c V S. rubrivinctus 4 F E C c c V S. saxicola 4 F F C c c 1X2 S. auriculatus 4 F E, F C d c VI S. brevispinis 4 F E C c c V S. fiavidus 4 S E C c c X S. melanops 4 s E C c c X S. pinniger 4 s E c c c X S. prorigcr 4 s E c c c X S. wilsoni 4 s «- -_ M _-. _« S. variegatus 4 s E c a c X|a S. caenaematicus 4 s E c c c X S. alutus 4 s F, S c c c Xll» S. crameri 4 vs E c c c XIII S. paucispinis 4 vs E c c c XIII S. recdi 4 vs E c c c XIII S. livis _« F E c a c XlVa Sebastolobus alascanus A vs D A b a, e XV3 Atlantic Sebastes S. marinus B s E, F B c c XVI S. viviparous B s E B c c XVI Heticotenus dactylopterus C s,vs C B e b XVI l» 1 Modified after Tsuyuki et ol., 1968. a Species with unique biological characteristics. ^ Pattern of the single specimen tested has not been described. 0.5 ml 1 M MgCl2 40 ml tris-citrate buffer Isocitrate dehydrogenase (ICDH): (a) NADP dependent (after Opher, Leonard, and Miller, 1969): 30 mg DL sodium isocitrate 5 mg NADP 0.5 ml 1 M MgCl2 5 mg PMS 5 mg NBT 40 ml tris-citrate buffer (b) NAD + dependent: Same formulation as NADP dependent, but substituting 10 mg NAD + for 5 mg NADP Muscle protein detected by nonspecific protein staining using 1% nigrosin-buffalo black in solution of 1:4:5 acetic acid: methanol: water and destained with a 1:4:5 solution of acetic acid, methanol, and water. ENZYME AND PROTEIN PHENOTYPES TETRAZOLIUM OXIDASE (TO) Interspecific variation of TO was previously reported in the genus Sebastes (Pacific) by Johnson et al. (1970b), where three anodal mo- bilities were observed in 15 species studied: Fast (F),Slow (S), and Very Slow (VS). These findings are expanded in the present study (Table 2, Figure 1). The F band occurred in 15 of the 27 Pacific Sebastes species; the S band was present in 9 species and the VS band in 3 species. Only the S band occurred in both At- lantic Sebastes species. The VS band was found 405 FISHERY BULLETIN: VOL. 70, NO. 2 •■scss'«i«s:^"sr i Origin— I > c — I I 1 :3 c Figure 1. — Band in starch-gel illustrating the four tetrazolium oxidase phenotypes, F, S, S-VS, and VS, detected in the family Scorpaenidae. The following samples are shown: 1, 5 Helicolenus dactylopteriis (S-VS), 2 Sebastes reedi (VS), 3 Sebastes caurinus (F), 4 Sebastes alutus (S), and 6 Sebastes reedi (VS). in Sebastolobus alascanus.' Helicolenus dactyl- opteriis was polymorphic for the S and VS bands; of the 10 samples tested, two exhibited a three- banded phenotype having the S and VS bands in addition to another band of intermediate mo- bility, whereas the rest had only the single S band. The three-banded phenotype suggests that two TO alleles are segregating in Helicole- nus and that tetrazolium oxidase functions as a dimer in scorpaenids. This interpretation is consistent with TO polymorphisms observed in salmonids (Utter, 1971) where three-banded phenotypes were observed in heterozygous rain- bow trout (Salmo gairdneri) and chinook salmon (Oncorhynchus tshaivytscha) . L-ALPHA -GLYCEROPHOSPHATE DEHYDROGENASE (aGPDH) Evidence for a polymorphic dimer having two alleles — Fast (F) and Slow (S) — were described * We used liver extracts of this species for detection of TO activity because muscle extracts failed to develop TO bands. We assume that this is a valid comparison because of parallel TO activity between liver and muscle observed in other scorpaenid species. All other scorpae- nid enzymes tested were extracted from skeletal muscle. in S. alutus (Johnson et al,, 1970a) . In addition to the F and S bands, three faster aGPDH bands have been observed among the scorpaenids that we have tested: E, D, and C,° listed according to increasing mobility (Figure 2 and Table 2). Additional aGPDH bands invariably occurred, regardless of phenotype, when electrophoresis proceeded beyond a 6-cm anodal migration of the dye marker. These bands are presumably arti- facts of electrophoresis and did not alter our in- terpretation of enzyme variations. This phe- nomenon was also noticed by McCabe, Dean, and Olson (1970) in aGPDH variants of skipjack tuna (Katsmvomis pelamis) . In Pacific Sebastes, 19 species were monomor- phic for the E band. S. auricukitus was poly- morphic for the E and F bands. S. caurinus as well as S. alutus were polymorphic for F and S bands. S". maliger and S. saxicola were mono- morphic for the F band. In the Atlantic Se- bastes, S. viviparous was monomorphic for the E band and S. mar inns was polymorphic for the E and F bands. The D and C bands were mon- omorphic Sebastolobus alascanus and Helicole- nus dactylopterus, respectively. ^ The separation of aGPDH bands C and D depends on optimal electrophoretic conditions. 406 JOHNSON, UTTER and HODGINS: ELECTROPHORETIC INVESTIGATION -L therefore called these regions peptidase A-I and i peptidase A-II. iggjg^ Origin I 8 Figure 2. — Bands in starch-gel illustrating four L-alpha glycerophosphate dehydrogenase phenotypes (C, D. E, F) detected in the family Scorpaenidae. The following samples are shown: 1, 5 Helicolenus dactylopterus (C), 2, 6 Sebastolobus alascanus (D), 3, 7 Sebastes rubri- vinctus (E), and 4, 8 Sebastes alutus (F). LACTIC DEHYDROGENASE (LDH) Muscle LDH was resolved as a single anodal band in each scorpaenid species we tested. This agrees with studies of Wilson, Kitto, and Kaplan (1967), who found single anodal bands of mus- cle LDH in two scorpaenid species, Sebastes ma- rinus and Scorpaenopsis gibbosa. The electro- phoretic mobilities were distinct in each species. LDH bands of three different mobilities (A, B, and C) were found in our sampling (Figure 3 and Table 2) . No poljonorphisms were detected. All but two Pacific Sebastes species expressed the C band. The B band was found in S. helvo- maculatiis and S. chlorostictus. The B band was found in two Atlantic Sebastes species and Heli- colenus dactylopterus. Only Sebastolobus alas- canus expresses the LDH A band. PEPTIDASE Peptidase staining occurred in two anodal re- gions for all species tested (Figure 4, Table 2). Both regions are developed with the dipeptide valyl-leucine, which is the specific substrate for peptidase A in mammals (Lewis and Harris, 1967; Lewis and Truslove, 1969). We have + I— Origin Figure 3. — Bands in starch-gel illustrating the three phenotypes of lactate dehydrogenase detected in the fa- mily Scorpaenidae. The following species are shown: 1, 4 Sebastolobus alascanus (A), 2, 5 Sebastes helvomac- tilatus (B), and 3, 6 Sebastes alutus. Five diflferent bands (a, b, c, d, e) were ob- served in the peptidase A-I (fast) zone. In Pa- cific Sebastes, peptidase A-I bands were ex- pressed as follows: P - S. chlorostictus, S. levis, and S. variegatus; P - S. caurinus, S. auricula- tus, and S. maliger. S. marinus had band P as did 9 of the 10 S. viviparous tested; Sebastolobus alascanus had band P ; and H. dactylopterus had band P. The aberrant Sebastes viviparous sam- ple had a single P band but corresponded to S. viviparous in all other systems tested. The sig- nificance of the variant is unclear. It may re- flect an intraspecies genetic variant (although multiple bands would be expected if this were the case) or perhaps a sibling species. Because only muscle samples were available for Atlantic Sebastes, identification of subtle morphological differences between individuals was not possible. Bands of five different mobilities (a, b, c, d, e) were also observed in the peptidase A-II (slow) zone. Band IP was expressed in all but two Pa- cific Sebastes tested; band IP was found in S. 407 FISHERY BULLETIN: VOL. 70. NO. 2 Figure 4. — Bands in starch-gel illustrating the various phenotypes of Peptidase A de- tected in the family Scorpaenidae. The following species are shown: 1 Helicolenus dactylopterus (IP, I^), 2 Sebastes caurinus (11^, Ic), 4 Sebastes variegatus (IP, Ja), 5 Sebastologus alascaniis (11^' ^ jb)^ 6 Sebastes diploproa (IP, P), and 7 Sebastes aurora (lid, lb). aurora and band IP in S. diploproa. Band IP was found in both Atlantic Sebastes species, and Helicolenus dactylopterus possessed band IP. Two bands representing the extremes of pepti- dase A-II mobilities — IP and IP — ^were ex- pressed in all Sebastolobus alascanus individuals tested. These bands are presumed to be fixed rather than polymorphic because of their invar- iant expression and may reflect gene duplication. PHOSPHOGLUCOMUTASE (PGM) PGM polymorphism was reported in Sebastes alutus, where two allelic bands — A and B — were described (Johnson, Utter, and Hodgins, 1971). In extending these observations here to addition- al scorpaenid species a third band — A' — has also been found which migrates somewhat faster than the A band (Figure 5). PGM is the most polymorphic of the scorpae- nid enzymes that we have investigated (Table 3). In Pacific Sebastes polymorphism was found in 10 species for the A and B bands and in 1 species for the A and A' bands. Twelve species of Pacific Sebastes were monomorphic for the A band, one for the B band, and one for the A' band. In other scorpaenid species, Se- bastes marinus was polymorphic for the A and B bands, and S. viviparous was monomorphic for «JW.-. A' A B Origin Figure 5. — Bands in starch-gel illustrating three mobil- ities of phosphoglucomutase detected in the family Scor- paenidae. The following species are shown: 1, 4 Sebastolobus alascanus (A') , 2, 5 Sebastes caurinus (A) , and 4, 6 Sebastes reedi (B). the A band. H. dactylopterus was polymorphic for the A and B bands, and Sebastolobiis alas- canus was monomorphic for the A' band. We assume that these variants reflect allelic diff"er- ences although further study is needed for some species. Also, the limited number of samples 408 JOHNSON, UTTER and HODGINS: ELECTROPHORETIC INVESTIGATION tested for some species that were listed as mon- omorphic are too few to preclude the possibility of polymorphism. ISOCITRATE DEHYDROGENASE, NADP DEPENDENT (ICDH NAD?) We tested for both NAD- and NADP-depen- dent ICDH in the 31 species studied and found activity only for the latter form. It is assumed that this represents cytoplasmic ICDH activity (Opher et al, 1969). Two anodal mobilities of ICDH were detected: the band of H. dactylop- terus migrated slightly faster than the band of the other species (Figure 6). No activity was detectable in extracts of Sebastolobus alascanus. Activity was highly labile in all species, requir- ing testing on the same day that the extraction was made. It may be that S. alascanus has an even more labile form of ICDH than the other species tested. Table 3. — Phosphoglucomutase phenotypes in muscle samples from species of Scorpaenidae.^ + i Phenotypes Species B AB A AA' A' Pacific Sebastes S. aUutianus ^_ + + __ _^ S. alutus + + + __ -_ S. auriculatus » .« + _^ _- S. aurora __ + + __ _- S. brevispinis + -f- + S. caurinus _• _• + S. chlorostictus _^ -.-. + S. crameri _^ + + __ __ S. elongatus + + + __ __ S. entomela! _^ _^ + __ __ S. flavidus „_ „^ + __ __ S. helvomaculatus _^ _. + + + S. levis .« _^ + — M _^ S. maliger -• _^ + _.. __ S. melanops -« .. + S. paucispinis w> + + ^_ S. pinniger _« + + S. proriger + + + -- S. reeii + _^ __ __ S. ruberrimus _« — * + -- S. rubivinctus _.« __ + S. saxicola — w _^ + __ _. S. lacentrus „^ _^ + __ S. caenacmaticus „.. + + __ _- S. variegatus _^ _^ + __ Atlantio Sebastes S. niarinus _^ + + _^ __ S. viviparous —a __ + _^ __ Sebastolobus alascanus __ __ __ __ + Helicolenus dactylopterus + ~ + — — Origin- 2 3 4 5 ^ PGM in our samples of S. diploproa and S. viitsoni did not develop. Figure 6. — Isocitric dehydrogenase (NADP dependent) bands found in the family Scorpaenidae. Samples 1, 3, 5 are Sebastes alutus and samples 2, 4 are Helicolenus dactylopterus. MUSCLE PROTEIN A satisfactory separation of muscle protein bands was obtained by permitting the dye mark- er to migrate 9.0 cm anodally from the origin. These bands were separated into two regions — A and B (Figure 7). Distinct protein patterns occurred in region A, which differ between genera as well as within the genus Sebastes (Pacific) (Table 4). S. aurora has a unique pattern (bands 1, 4) which differed from the other Pacific Sebastes species (bands 1,3). The intergeneric differences in re- gion A were: Sebastes (Pacific) — bands 1, 4 and 1, 3; Sebastes (Atlantic) — bands 2, 6; Heli- colenus— bands 3, 7; and Sebastologus — 5, 7. A band (X) which migrated more anodally than band 7 was found in some Sebastes alutus. We assume this band (X) to be an artifact as it did not appear in repeated tests. The most anodal band (8) was found in all samples tested. Cor- responding region A patterns were not described by Tsuyuki et al. (1968) in instances where the same species were tested and may arise from differences in methodology such as buffer sys- tems (Rasmussen, 1969). 409 Genus FISHERY BULLETIN: VOL. 70, NO. 2 Table 4. — Intergeneric comparison of muscle protein bands of Scorpaenids. Protein bands Subgroup^ Region B^ a bed e i g h i i k m n o p + + + - - + + + - .. + + - - + - - + - - + - + Region A 1 2 3 7 8 Pacific Stbastes Pacific Stbastes Pacific Stbastes Atlantic Sebastes Sebastolobus Helicolenus + + + + + + + + - + + + __ _ - + - + + - - + - + + - - + - + + - + - + - - + - + + + + + -I- + - + + + -. + + - + - + .. + + + + + 1 Pacific Sebastes subgroups after Tsuyuki et al., 1968. " Alphabetical classification after Tsuyuki et al., 1968. i 8 Region -t A -I "^^/gffi 3iaM Origin — 2 3 ■^k+l h f Region B lenus. Bands b and c stained weakly in our gels and failed to show in some individuals (Figure 7). The slowest anodal bands were f and h which occurred only in S. aurora. Band i oc- curred in all species tested except S. aurora, S. elongatus, and S. entomelas. On the other hand, S. entomelas and S. elongatus were the only species having the j band, bands j and k being polymorphic in S. elongatus (first reported by Tsuyuki et al., 1968). Band k was present in all genera but Sebastolobus, which — in turn — was the only genus expressing band 1. Simi- larly, bands m and p — present in other genera — were absent in Sebastolobus, which uniquely ex- pressed band o. Our methods were unable to detect band q, reported by Tsuyuki et al. in At- lantic Sebastes and Sebastolobiis. 12 3 4 5 I Region A (enlarged ) Figure 7. — Muscle protein bands in starch-gel: 1 Heli- colenus dactylopterus (A-3, 7), 2 Sebastolobus alascanus (A-5, 7), Sebastes marinu^ (A-2, 6), 4 Sebastes aurora (A-1, 4), and 5 Sebastes alutus (A-1, 3). The protein patterns in region B were similar to those described by Tsuyuki et al. (1968) , who described 16 bands (a-p) that varied between genera and species. Three cathodally migrating bands (a, b, c) occurred in Pacific Sebastes (ex- cept S. aurora), Atlantic Sebastes, and Helico- COMPARISON OF VARIATION BETWEEN GENERA A comparison of the total variation between genera suggests some possible relations. The greatest similarity was between the Pacific Se- bastes and Atlantic Sebastes where all the elec- trophoretic patterns of the Atlantic Sebastes were found in one or more species of the Pacific Sebastes, except for the protein bands of region A. Pacific Sebastes and Sebastolobus exhibited common bands for PGM, TO, peptidase A-I, and protein B-i. Pacific Sebastes and Helicolenus shared common bands for LDH, PGM, and pro- tein bands of region B. Helicolenus and one spe- cies of Pacific Sebastes possessed a common pep- tidase A-II band. Helicolenus and Sebastes had common bands in LDH, PGM, and protein region B. Helicolenus and Sebastolobus shared only pro- 410 JOHNSON, UTTER and HODGINS: ELECTROPHORETIC INVESTIGATION tein bands B-i and A-7. Only protein band B-i was common to Sebastolobiis and the Atlantic Sebastes (Tables 2, 4, and 5). When the total amount of common patterns between genera is considered, we agree with Tsuyuki et al. (1968) that there is relatively greater similarity between the Pacific Sebastes and the Atlantic Sebastes than between either and the other genera studied. S. aurora was found to have relatively the same degree of dif- ference between itself and the other Pacific Se- bastes species as there was between the Atlantic Sebastes and the Pacific Sebastes. This agrees with the findings of Tsuyuki et al. (1968) who suggested that S. aurora should possibly be ele- vated to the generic level because of its degrees of difference. The interpretation of similarity based on electropherograms must be done with caution as only amino acid substitutions which change the net charge of the polypeptide chain can be detected. Table 5. — Summary of intergeneric enzymatic similarity in Scorpaenidae.^ X indicates the occurrence of com- mon bands between one or more species of the genera compared. Gei nera Genus and enzyme Pacific Sebastes Atlantic Sebastes Sebi istolobus Mel Uolenus Pacific Sebastes TO X X __ aGDPH X _^ _^ LDH X _^ X Peptidase A-l X X _^ Peptidase A-ll X _^ X ICDH X _^ —v PGM X X X Atlantic Sebastes TO X ^^ ^^ aGDPH X __ __ LDH X _^ X Peptidase A-l X ._ -« Peptidase A-ll X — « _^ ICDH X __ — M PGM X — X '^ No common bands were found between Sebastolobus and Helicolenus. VARIATION WITHIN PACIFIC SEBASTES Combining the enzyme and protein variations in the Pacific Sebastes resulted in 10 of the 27 Pacific Sebastes species having unique biochem- ical profiles (Table 2). These species were S. elongatus, S. entomelas, S. aurora, S. chloros- tictus, S. diploproa, S. helvomaculatus, S. saxi- cola, S. variegatus, S. alutus, and S. levis. Some species were represented by only a few samples — therefore further sampling may reveal vari- ation in these profiles. PGM was not included in these profiles because of its high degree of polymorphism in the genus. A new species, S. reedi, was reported by Westrheim and Tsuyuki (1967) that resembles S. cramerl, S. alutus, and S. proriger but was readily separable from these when morphology and biochemical methods were employed. Our study found that S. reedi and S. crameri were identical with respect to muscle protein and five enzyme systems but differed in PGM. This suggests that S. reedi may be more closely related to S. crameri than to the other species. Three species, S. caurinus, S. maliger, and S. auriculatus, had profiles that difl["ered only in the enzyme aGPDH, which was monomorphic in S. maliger (F band) but polymorphic for the F and S bands in S. caurinus. All three species have the peptidase A-P band which was found in no other Sebastes species. These three species are very similar in morphology and habitat pref- erences. In certain areas of Puget Sound, Wash., hybridization between the three may occur, whereas in other areas they remain separate be- cause of behavioral differences.' Investigation of biochemical and morphological characteristics of these species may provide valuable informa- tion on the processes of speciation. The amount of polymorphism of aGPDH and PGM in the family Scorpaenidae could prove to be useful for the identification of breeding pop- ulations and verification of species and subspe- cies. On the basis of morphometric data, Bar- sukov (1964) suggested that two subspecies exist in Sebastes alutus (S. a. alutus — off the Pacific coast of North America and S. a. paucispinus — from Honshu Island, Japan, to perhaps Bristol Bay, Alaska). Westrheim (1970) suggested that S. alutus had a southern and a northern type of fish off the coast of North America — the south- * C. R. Hitz, National Marine Fisheries Service, Fish- ery Biologist, Exploratory Fishing and Gear Research Base, Seattle, Wash., personal commun., April 1971. 411 FISHERY BULLETIN: VOL. 70, NO. 2 ern type south of Dixon Entrance and the north- ern type North of Dixon Entrance and in the Gulf of Alaska, Differences in gene frequencies would add to the support of their separations. This approach may also prove useful in studying com- plexes such as found in S. aleutianus, S. reedi, and S. diploproa (Tsuyuki et al., 1968). SUMMARY An investigation of muscle protein and six en- zymatic systems by starch-gel electrophoresis was presented. Samples of 31 species of three genera of the family Scorpaenidae were com- pared which resulted in the conclusion that a relatively greater similarity existed between the Pacific Sebastes and the Atlantic Sebastes than between the other genera. Ten of the 27 species of Pacific Sebastes had unique profiles when the systems were compared. ACKNOWLEDGMENTS We are especially grateful to the following persons who provided samples and valuable in- formation: Dr. I. Barrett and S. Kato (National Marine Fisheries Service, Southwest Fisheries Center, La Jolla, Calif.) , C. R. Hitz, B. G. Patten, H. H. Shippen, K. E. Thorson, and K. D. Waldron (National Marine Fisheries Service, Northwest Fisheries Center, Seattle, Wash.), Dr. A. C. DeLacy (University of Washington, Seattle, Wash,), Dr. A. Jamieson (Ministry of Agricul- ture, Fisheries and Food, Lowestoft, Suffolk; England), and S. J. Westrheim (Fisheries Re- search Board of Canada, Nanaimo, B.C., Can- ada). LITERATURE CITED Altukhov, JU. p., and G. N. Nefyodov. 1968. A study of blood serum protein composition by agar-gel electrophoresis in types of redfish (ge- nus Sebastes). Int. Comm. Northwest Atl. Fish., Res. Bull. 5:86-90. Bailey, R. M. (chairman). 1970. A list of common and scientific names of fishes from the United States and Canada. Am. Fish. Soc, Spec. Publ. 6, 150 p. Barsukov, V. V. 1964. Vnutrividovaya izmenchivost' u Tikhookean- skogo morskogo okunya (Sebastodes alutus Gil- bert) [Intraspecific variability in the Pacific rock- fish (Sebastodes alutus Gilbert)]. Tr. Vses. Nauchn.-issled. Inst. Morsk. Rybn. Khoz. Okean- ogr. 49 (Izv. Tikhookean. Nauchn.-issled. Inst. Morsk. Rybn. Khoz. Okeanogr. 51) : 231-252. (Transl., 1968, in Soviet fisheries investigations in the northeast Pacific, Part II, p. 241-267. Avail. Natl. Tech. Inf. Serv., Springfield, Va., TT 67-51204.) Brewer, C. J. 1967. Achromatic regions of tetrazolium stained starch gels: inherited electrophoretic variation. Am. J. Hum. Genet. 19:674-680. Chen, L. 1971. Systematics, variation, distribution, and bi- ology of rockfishes of the subgenus Sebastomus (Pisces, Scorpaenidae, Sebastes). Bull. Scripps Inst. Oceanogr. Univ. Calif. 18, 115 p. Chu, R. 1968. Identification of commercially used fish found in the Pacific by disk electrophoresis. J. Assoc. Off. Anal. Chem. 51:743-746. Hunter, R. L., and C. L. Markert. 1957. Histochemical demonstration of enzymes sep- arated by zone electrophoresis in starch gels. Sci- ence (Wash., D.C.) 125:1294-1295. Johnson, A. G., F. M. Utter, and H. 0. Hodgins. 1970a. Electrophoretic variants of L-alpha-glycero- phosphate dehydrogenase in Pacific ocean perch (Sebastodes alutus). J. Fish. Res. Board Can. 27:943-945. 1970b. Interspecific variation of tetrazolium oxi- dase in Sebastodes (rockfish). Comp. Biochem. Physiol. 37:281-285. 1971. Phosphoglucomutase polymorphism in Pa- cific ocean perch, Sebastodes alutus. Comp. Bio- chem. Physiol. 39 B: 285-290. Kristjansson, F. K, 1963. Genetic control of two per-albumins in pigs. Genetics 48:1059-1063. Lewis, W. H. P., and H. Harris. 1967. Human red cell peptidase. Nature (Lond.) 215:351-355. Lewis, W. H. P., and G. M. Truslove. 1969. Electrophoretic heterogeneity of mouse er- ythrocyte peptidases. Biochem. Genet. 3:493-498. Markert, C. L., and I. Faulhaber. 1965. Lactate dehydrogenase isozyme patterns of fish. J. Exp. Zool. 159:319-332. McCabe, M. M., D. M. Dean, and C. S. Olson. 1970. Multiple forms of 6-phosphogluconate dehy- drogenase and alpha-glycerophosphate dehydro- genase in the skipjack tuna, Katsuwonus pelamis. Comp. Biochem. Physiol. 34:755-757. NiSHIMOTO, J. 1970. Western range extension of the rosethorn 412 JOHNSON, UTTER and HODGINS: ELECTROPHORETIC INVESTIGATION rockhsh, Sebastes helvomaculatus (Ayres). Calif. Fish Game 56:204-205. Nyman, C. 1967. Protein variations in Salmonidae. Rep. Inst. Freshwater Res. Drottingholm 47:5-38. Opher, a. W., J. W. Leonard, Jr., and J. M. Miller. 1969. Isoenzymes of isocitrate dehydrogenase in cytoplasm of human cells. Exp. Med. Surg. 27: 256-266. Rasmussen, D. I. 1969. Molecular taxonomy and typology. BioSci- ence 19:418-420. RiDGWAY, G. J., S. W. Sherburne, and R. D. Lewis, 1970. Polymorphism in the esterases of Atlantic herring. Trans. Am. Fish. Soc. 99:147-151. Smithies, 0. 1955. Zone electrophoresis in starch gels: Group variations in the serum proteins of normal human adults. Biochem. J. 61:629-641. Spencer, N., D. A. Hopkinson, and H. Harris. 1964. Phosphoglucomutase polymorphism in man. Nature (Lond.) 204:742-745. TSUYUKI, H., AND S. J. WESTRHEIM, 1970. Analyses of the Sebastes aleutianus — S. mel- anostomiis complex, and description of a new scor- paenid species, Sebastes caenaematicus, in the northeast Pacific Ocean. J. Fish. Res. Board Can. 27:2233-2254. TsuYUKi, H., E. Roberts, R. H. Lowes, W. Hadaway, AND S. J. WESTRHEIM. 1968. Contribution of protein electrophoresis to rockfish (Scorpaenidae) systematics. J. Fish. Res. Board Can. 25:2477-2501. Utter, F. M. 1971. Tetrazolium oxidase phenotypes of rainbow trout (Salmo gairdneri) and Pacific salmon (On- corhynchus spp.). Comp, Biochem. Physiol. 39 B:891-895. -^ WESTRHEIM, S. J. 1965. Northern range extensions for four species of rockfish (Sebastodes goodei, S. helvomaculatus, S. rubrivinctus, and S. zacentrus) in the North Pacific Ocean. J. Fish. Res. Board Can. 22:231- 235. 1970. Survey of rockfishes, especially Pacific ocean perch, in the northeast Pacific Ocean, 1963-66. J. Fish. Res. Board Can. 27:1781-1809. WESTRHEIM, S. J., AND H. TSUYUKI. 1967. Sebastodes reedi, a new scorpaenid fish in the northeast Pacific Ocean. J. Fish. Res. Board Can. 24:1945-1954. Wilson, A. C, G. B. Kitto, and N. O. Kaplan. 1967. Enzymatic identification of fish products. Science (Wash., D.C.) 157:82-83. 413 THE SYNERGISTIC EFFECTS OF TEMPERATURE, SALINITY, AND MERCURY ON SURVIVAL AND METABOLISM OF THE ADULT FIDDLER CRAB, UCA PUGILATOR' Winona B. Vernberg and John Vernberg- ABSTRACT Gill tissues of fiddler crab, Uca pugilator, were the major site of mercury concentration; lesser amounts accumulated in the hepatopancreas and green gland. Very small amounts were found in the carapace and muscle tissue. No significant differences in the amount of mercury in tissues of males and females were found. A concentration of mercury sublethal to fiddler crabs under optimum conditions of temperature and salinity greatly reduced survival times when crabs were placed under conditions of temperature and salinity stress. Males were more susceptible to the synergistic effects of mercury in combination with environmental stress than were females. Metabolic rates of male and female fiddler crabs were affected by prolonged exposure to mercury both under optimum environmental conditions and under temperature and salinity stress. Metabolic rates of males were more adversely affected than those of females. Estuaries are an extremely important part of the marine environment. Yet often an estuary becomes so grossly polluted that much of the biota is destroyed before it is recognized that the quality of water affects the biology of such an area. Part of the problem is the subtleness of the effects of sublethal concentrations of man- introduced pollutants. In low concentration the pollutant may have no observable effect on a given population of animals if environmental conditions remain at an optimum. However, when another environmental parameter becomes stressful, it may synergistically interact with the sublethal concentration of pollutant and the or- ganism dies. Many estuaries are polluted, and since one of the chief characteristics of estuaries is the rather extreme environmental fluctuations that occur throughout the year, knowledge of synergistic interaction on estuarine animals is important in the preservation of estuarine eco- systems. ' This study was supported by Grant No. 18080 FYI from the U.S. Environmental Protection Agency. ^ Belle W. Baruch Coastal Research Institute and Department of Biology, University of South Carolina, Columbia, SC 29208. Manuscript accepted December 1971. FISHERY BULLETIN: VOL. 70. NO. 2, 1972. This study was undertaken to determine the effect of a sublethal concentration of mercury on the metabolism of adult male and female fid- dler crabs, Uca pugilator (Bosc), maintained under optimum and stressful conditions of tem- perature and salinity, and the synergistic effects on survival of this species with sublethal con- centration of mercury in combination with sa- linity and thermal stress. This species was se- lected because it is one of the more abundant and ecologically important species in an estu- arine ecosystem. MATERIALS AND METHODS Crabs used in this study were collected in the Georgetown, S.C., area during the fall and winter months. After collection the animals were brought into the laboratory where they were maintained in plastic boxes containing a thin layer of seawater having a salinity of ap- proximately 30/'ff. All crabs were kept in con- stant temperature boxes at 25°C and on a 12-hr light-dark photoperiod for at least 2 weeks. Crabs were fed on Clark's fish pellets every third 415 FISHERY BULLETIN: VOL. 70, NO. 2 day; the water was changed after each feeding. Preliminary studies were undertaken to deter- mine the amount of HgCl2, an inorganic mercury compound, that could be added to the water with- out killing the crabs. A concentration of 9 X 10-^ M HgCl2 was found to be sublethal for crabs which were kept under optimal conditions of temperature and salinity. Under these condi- tions crabs survived for a 2-month period with only slight mortality. The experiment was terminated at this point. The initial concen- tration of mercury, 9 X 10 "'^ M HgCl2 was 0.18 ppm Hg (or 0.18 mg/liter seawater). Tissues of crabs were analyzed for mercury following exposure to 0.18 ppm mercury in 30^c seawater at a temperature of 25°C for 1, 3, 7, 14, and 28 days. Tissues of crabs maintained under the same conditions but without added mercury were also analyzed. Tissues were removed from 10 crabs for each assay and frozen immediately. The concentration of mercury in each tissue was then determined on a Perkin-Elmer Mercury Analyzer System-50.' The techniques were based on the Environmental Protection Agency method developed by the Analytical Quality Control Laboratory, using dilute nitric acid to digest the samples. Determinations were made by South Carolina State Board of Health person- nel. Five tissues were assayed: gill, hepatopan- creas, green gland, abdominal muscle, and car- apace. None of the tissues were kept frozen for more than 1 week. Tissues from 20 males and females (two determinations each) were assayed for each of the five experimental time exposures. Since the amount of mercury proved to be essentially the same in tissues of both males and females, all data were pooled. To determine the synergistic effects of the normally sublethal concentrations of mercury and stressful environmental factors, crabs accli- mated to 25°C, 30^^( seawater were placed in seawater with a salinity of 5'/!, containing 0.18 ppm mercury or in 5'/ic seawater without mercury and maintained at either 5°C or 35°C. At each experimental temperature, four groups of ani- ' Reference to trade names in this publication does not imply endorsement of commercial products by the National Marine Fisheries Service. mals were used. Thus at 5°C, 5^f , one group of 30 males and a second group of 30 females were used as controls; in the experimental group 30 males and 30 females were maintained under the same conditions except the water contained 0.18 ppm mercury. The same procedure was followed at 35°C and in a salinity of 5%c. Survival of both experimental and control crabs was followed for 28 days or until 50 9f of any one group had died. The temperatures of 5°C and 35 °C were selected since they represent low and high temperature extremes which fiddler crabs experience season- ally in South Carolina marshes. A salinity of 5%f is also encountered by them in the field. Oxygen consumption of control and experi- mental animals was determined by means of a Gilson respirometer using respiration flasks with a volume of approximately 140 cc. Base-line oxygen consumption measurements were made on 10 males and 10 females in untreated sea- water (30;^c) at 25°C. These same crabs were then maintained under the same conditions but with 0.18 ppm Hg added to the water, and me- tabolic determinations made on days 1, 3, 7, 14, 21, and 28. Only medium-sized crabs in the in- termolt stage were used to avoid any variation due to molting or metabolic size relationships. Oxygen consumption rates were also deter- mined on crabs exposed to mercury in combina- tion with temperature and salinity stress. Me- tabolic measurements were made on crabs main- tained in 5/^r'f seawater at 5°C (control crabs) and crabs kept in 5%c seawater at 5°C with 0.18 ppm Hg added to the water (experimental crabs) . Oxygen consumption rates were then determined after 1 and 3 days exposure for ex- perimental crabs and 1, 3, and 7 days for control crabs. These conditions proved too stressful for most of the crabs to survive longer periods of time. The same experimental procedures were followed for crabs kept in 5^f at 35°C with and without added mercury. Since these conditions were less stressful than the combination of low salinity and low temperature, it was possible to measure the metabolic rate of these crabs on days 1, 3, 7, 14, and 21 for experimental animals and to day 28 for control crabs. All results are expressed as />tl iters of oxygen consumed per hour per gram live weight. 416 VERNBERG and VERNBERG: SURVIVAL AND METABOLISM OF FIDDLER CRAB RESULTS MERCURY UPTAKE BY TISSUE Mercury was not detected in the untreated seawater, although there were detectable traces of mercury found in the Clark's fish pellets fed to the crabs. The hepatopancreas of the control animals (animals collected in the same region as experimental animals and maintained in sea- water without mercury addition) had approxi- mately 0.03 ppm mercury, but no mercury was found in any of the other tissues. Within the first 24 hr after exposure to 30%c seawater at 25 °C containing an initial concentration of 0.18 ppm mercury, however, gill tissue contained 1.73 ppm mercury; the amount of mercury in this tissue increased steadily with continued ex- posure (Figure 1). Of the five tissues assayed for mercury content, gill tissue was found to have the highest concentration. Mercury also accumulated in the hepatopancreas and green gland although much less rapidly and at a lower concentration level (Figure 1). Lower amounts of mercury were found in abdominal muscle tis- sues and in the carapace; after 28 days exposure to water containing mercury, levels were approx- imately 1 ppm. 14 " i/ 12 - / / / 10 - / Is a - / V ^ ca ^'■^ 3 _ ^-' o 6 - IT ^■'"^ UJ ^^ s 4 - ■^ 2 """^ l^s 1 1 1 28 DAYS Figure 1. — Mercury in tissues of Vca pugilator after exposure of the crabs to 9 X lO-^ m HgCla (0.18 ppm Hg) in 30%o seawater at 25 °C for varying lengths of time. • • Control females -~^ o o Experimentol females • Control moles '~ o Experimental moles 18 20 22 24 26 Figure 2. — Mortality of Uca pugilator in b%c seawater at 5° C with and without the addition of 9 X 10"'' M HgClg (0.18 ppm Hg). LETHAL LEVELS Preliminary studies established that under optimum conditions of temperature (25 °C) and salinity (30^r) the crabs could survive for pro- longed periods of time (at least 2 months) in sea- water having an initial concentration of 9 X 10"'^ M HgCl2 (0.18 ppm mercury). Under tempera- ture and salinity stress, however, this concentra- tion of mercury significantly shortened survival time. For example, under conditions of low tem- perature (5°C) and low saHnity (5%f), such as could occur following heavy winter rains, the crabs could not survive as long as under condi- tions of high temperature and low salinity. In winter animals without the added stress of a pollutant, 50 yi of the females survived 21 days but 50% of the males were dead within 7 days. Under the same temperature and salinity con- ditions with the addition of 0.18 ppm mercury, males survived 6 days, but 50% of the females died by day 8 (Figure 2). Under conditions of low salinity (5%o) and high temperature (35°C), conditions very apt to occur following the heavy rains associated with a summer hurricane, both male and female U. pugilator can survive with very little mortality for at least 28 days (Fig- ure 3) . With the addition of 0.18 ppm mercury, however, survival times of both males and fe- males are reduced. Under conditions where crabs were maintained at this high temperature and low salinity in water containing mercury, 50% of the males had died by day 17, while 50% of the females survived to day 26 (Fig- ure 3). 417 FISHERY BULLETIN: VOL. 70, NO. 2 0' 10 20 g 30 ^ 40 i 50 t 60 - O 2 70 ■ 80 - 90 - • •Control females o o Experimental females • •Control moles o o Experimental moles 14 DAYS 22 Figure 3. — Mortality of Uca pugilator in h'/ir seawater at 35 °C with and without the addition of 9 X 10-7 m HgCla (0.18 ppm Hg). METABOLIC EFFECTS Although a low level concentration of mercury- was not lethal to the crabs under optimum envi- ronmental conditions, metabolic rates of these crabs were affected, especially those of males. Initially, metabolic rates were established for both males and females at 25°C in 30%. seawater, and the rates for males and females were essen- tially the same (Figure 4). After the base-line rate was determined, the same animals were then maintained at 25 °C in ^O'/ic seawater with the 200 X » 2 < O UJ a. 100 80 60 40 (ft UJ 20 tFemoles I " 1 J L I 3 7 II 14 17 DAYS 21 28 Figure 4. — Oxygen uptake rates of male and female Uca pugilator maintained in 30%r seawater containing 9 X 10-7 M HgCla (0.18 ppm) at 25°C. The base-line rate is represented by the first set of data points on the left. The vertical bar through each mean value is the standard error. addition of mercury, and metabolism of the crabs was determined periodically for 28 days. The metabolic rate of the males remained essentially unchanged through day 3. By day 7, however, the metabolic rate had dropped to 32% of that of untreated crabs ; by day 21 the rate had de- creased by 48%; and by day 28 the rate was 20% lower than the base-line value. In the fe- male, oxygen uptake values also decreased by day 7, but by day 14 the metabolic rate returned to the base-line rate and remained essentially at this level through day 28. Although initially the same, the rate of oxygen uptake of males was significantly lower than that of the females after 21 days in this sublethal concentration of mercury, and the metabolic rate of the males had not returned to the same level as it was be- fore the crabs were placed in mercury by the end of the 28 day experimental period (Figure 4) . Both males and females, however, continued to survive for another month under the same mercury regime as before without any signifi- cant increase in mortality. Under conditions of low temperature (5°C) and salinity (5%p) stress, not only did females survive much longer than males, but also the females were better able to maintain a steadier rate of oxygen uptake (Figure 5). The metab- olic rate and pattern of the experimental female crabs were similar to those of the control female crabs. The metabolic rate of male experimental crabs was not significantly different from that of the female experimental or male and female control crabs, after a 1-day exposure to mercury, but by day 3 the rate dropped markedly (Fig- ure 5). Oxygen uptake rates of female control crabs maintained in low-salinity water {^'/u) and at high temperature (35°C) were relatively con- stant over a 28-day period and tended to be higher than that of control male crabs (Figure 6) . The metabolic rates of experimental female crabs remained fairly constant for the first 7 days and then declined rapidly. The uptake rates of experimental male crabs declined steadily from day 1 and tended to be lower than those of the females throughout the remainder of the time period (Figure 6). 418 VERNBERG and VERNBERG: SURVIVAL AND METABOLISM OF FIDDLER CRAB 30 - I o ^ 20 I- UJ < (r o cr UJ Q. a: I O (0 (T \^ 20 10 9 EXPERIMENTAL 10 CONTROL Females 3 DAYS Figure 5. — Oxygen uptake rates of male and female Uca pugilator maintained at 5° in 5%o seawater with and without the addition of 9 X 10-^ m HgClg or 0.18 ppm. The vertical bar through each mean value is the standard error. DISCUSSION Fiddler crabs are capable of rapidly removing mercury from their surrounding aqueous media and retaining it in their tissues. However, not all tissues concentrate mercury to the same de- gree. The rapid accumulation and large con- centration of mercury in gill tissue of fiddler crabs and the lesser but significant amounts of mercury found in the hepatopancreas and green gland are similar to results obtained in experi- ments involving other heavy metals. Bryan (1966), for example, found the highest concen- tration of zinc in the gills and hepatopancreas. He related these concentrations to the fact that excess zinc can be stored in the hepatopancreas in the crab Carcinus maenus and concentrated and excreted across the gills. 300 « 200 UJ 5 < 100 EXPERIMENTAL ;j-i--4 -. -«■« 1*o/es >.. IT ui 300 a. q: X \ o 200 q: UJ _l 3. 100 CONTROL Females 14 DAYS 21 26 Figure 6. — Oxygen uptake rates of male and female Uca pugilator maintained at 35°C in 5%c seawater with and without the addition of 9 X 10-7 m HgClg or 0.18 ppm. The vertical bar through each mean value is the standard error. Uca pugilator can adapt quickly to a wide range of adverse environmental fluctuations (Vernberg and Vernberg, 1970). The sudden changes in temperature and salinity that do oc- cur usually do not persist for prolonged periods of time, and conditions usually ameliorate with- in a week or two. Results presented in this pa- per indicate that the crabs can withstand low salinity and high temperature better than low salinity coupled with low temperature, findings consistent with the earlier generality proposed byPanikkar (1940). Further, Lockwood (1962) stated that since ionic regulation is thermally influenced, organisms survive dilute medium more successfully when the rate of ion uptake as compared with ion loss increases faster with increasing temperature. Under both sets of conditions, however, the added stress of con- centrations of mercury that are sublethal under optimum conditions adversely aflfected survival rates under stressful conditions and more mark- edly in males than in females. Bryan (1971) has speculated that the increased lethality of a heavy metal under stressful conditions is in some way related to changing rates of absorption. Our data are another example of the principle 419 FISHERY BULLETIN: VOL. 70, NO. 2 that multiple environmental factors, each at a sublethal level, interact synergistically to cause death. Earlier papers, especially the classic paper of McLeese (1956), emphasized the lethal role of "normal" environmental factors, whereas we have demonstrated the importance of pollut- ants as part of the "normal" environment of many species. Under optimal conditions of temperature and salinity mercury generally decreased metabolic rates of the males; effect on metabolic rates of females was much less pronounced. This dif- ferential effect of mercury on the metabolism of males and females is difficult to understand. On an interspecific basis, differences between re- sistance of larvae oi Artemia salina and Elminius viodestus to mercury have been related to dif- ferences between rates of uptake rather than of tissue resistance (Corner and Rigler, 1958). However, since the amount of mercury in tissues of both male and female fiddler crabs was essen- tially the same, these differences would not ap- pear to be related to differences in uptake of the mercury. Under conditions of thermal and sa- linity stress without the addition of mercury the metabolic rate of the female crabs tended to be more stable and less depressed than the rate of male crabs. The addition of mercury to the already stressful conditions accentuated these differences. Our results indicate, then, that a concentration of mercury that is sublethal under optimum con- ditions of temperature and salinity, may greatly reduce the ability of the population to survive under normally stressful conditions of temper- ature and salinity flux. ACKNOWLEDGMENTS We are grateful to Ms. Gary Clark and Ms. Barbara Caldwell for technical assistance and to Dr. Lamar Priester of the State Board of Health for mercury analyses, LITERATURE CITED Bryan, G. W. 1966. The metabolism of zinc and ^szn in crabs, lobsters and freshwater crayfish. Symp. radio- ecological concentration processes, Stockholm, Sweden, p. 1005-1016. Pergamon Press, Oxford, 1971. The effects of heavy metals (other than mercury) on marine and estuarine organisms, Proc. R. Soc. Lond., Ser. B Biol. Sci. 177:389-410. Corner, E. D. S., and F. H. Rigler. 1958. The modes of action of toxic agents. III. Mercuric chloride and N-amylmercuric chloride on crustaceans. J. Mar. Biol. Assoc. U.K. 37: 85-96. LOCKWOOD, A. P. M. 1962. The osmoregulation of Crustacea. Biol. Rev. (Camb.) 37:257-305. McLeese, D. W. 1956. Effects of temperature, salinity and oxygen on the survival of the American lobster. J. Fish. Res. Board Can. 13:247-272. Panikkar, N. K. 1940. Osmotic properties of the common prawn. Nature (Lond.) 145:108, Vernberg, F. J., and Vernberg, W. B. 1970. The animal and the environment. Holt, Rinehart and Winston, 398 p. 420 LENGTH-WEIGHT RELATIONSHIP, FOOD HABITS, PARASITES, AND SEX AND AGE DETERMINATION OF THE RATFISH, HYDROLAGUS COLLIEI (LAY AND BENNETT)' Allyn G. Johnson^ and Howard F. Horton^ ABSTRACT In the fall and winter of 1965-1967, 292 ratfish (Hydrolagus colliei) were collected from four locations off the Pacific coast of Oregon. Specimens were examined for length-weight relationships, food habits, parasites, and a method of sex and age determination. Equa- tions describing the body weight-body length (snout to vent) relationships were log weight = log —4.3217 -f 3.0546 log length for males, and log weight = log —4.1692 + 2.9720 log length for females. The food organisms most important to ratfish were shrimp (Pandalus and Crago), mollusks (Musculus and Amphissa) , and echinoderms (Bri- saster). Two occurrences of cannibalism were found in ratfish collected off Cape Arago, Oreg. Infestations by Gyrocotyle ranged from 29 to 66% among samples from the four locations. The copepod, Acanthochondria sp., was attached to the claspers of seven males from Cape Arago. Eye-lens weights (wet and dry), vertebral radii, basal sections of the dorsal spine and left pectoral fin, and body-length frequencies were studied, but no accurate method of age determination was found. Tritors on the posterior side of the vomerine dental plate may be indicative of age, but the precise relationship was not determined. The ratfish, Hydrolagus colliei (Lay and Ben- nett) , is a member of the class Chondrichthyes, order Chamaeriformes, and family Chimaeridae (Bailey, 1970) . Distributed from western Alas- ka to northern Baja California (Koratha, 1960) , this cartilaginous fish is the only chimaeroid found on the Pacific coast of Canada and the United States, Ratfish are of little economic value, but their liver oil is an excellent lubricant and could be used commercially (Clemens and Wilby, 1961), Ratfish are an important source of food for such commercial fishes as soupfin sharks, Galeorhinus zyopterus (Nakatsu, 1957), spiny dogfish, Squalus acanthias (Alverson and Stansby, 1963), and Pacific halibut, Hippoglos- ^ Technical Paper No. 3144, Oregon Agricultural Ex- periment Station, Corvallis, Oreg. ^ Department of Fisheries and Wildlife, Oregon State University, Corvallis, OR 97331; present address: Na- tional Marine Fisheries Service, Northwest Fisheries Center, 2725 Montlake Boulevard East, Seattle, WA 98102. ' Department of Fisheries and Wildlife, Oregon State University, Corvallis, OR 97331. sus stenolepis (Thompson, 1915). Conversely, ratfish are commonly caught in the trawls of commercial fishermen who consider them a nui- sance. Our study was conducted to help fill the need for more information on the general biology of this primitive fish (Bigelow and Schroeder, 1953; Crescitelli, 1969). Little information has been published on the food habits of H. colliei. Dean (1906) reported that Chimaera (Hydrolagus) colliei fed on small isospondylous fishes, opisthobranchs, annelids, crustaceans, mollusks, squids, and nudibranchs. Olsson (1896) and Legendre (1944) reported that Chimaera monstrosa fed on mollusks, dec- opods, annelids, amphipods, echinoderms, and coelenterates. Several studies have been conducted on the p&rasites of H. colliei. Wardle (1932) reported that most ratfish contained a pair of Gyrocotyle urna in the anterior region (spiral valve) of the intestine. Lynch (1945) concluded that G. wma should be divided into G. urna and G. fimbriata. Koratha (1960) examined two H. colliei from Manuscript accepted January 1972. FISHERY BULLETIN: VOL. 70, NO. 2, 1972. 421 FISHERY BULLETIN: VOL. 70, NO. 2 Baja California and found two digenea {Otodist- omum sp.) and one cestode from the intestine, one hirudinean { Branc hellion &p.) from the skin surface, and three copepods {Chondracanthus epacthes) and one monogenea from the gills. A monogenea reported from the gills of ratfish from Washington waters was Octobathrium leptogaster (Bonham, 1950). Wilson (1935) reported the copepods Acanthochondria clavata and A. epachthes from the claspers and gills, re- spectively, of H. colliei from California waters. Kabata (1968) could not accept Wilson's (1935) identification of A. clavata, because the species had never been found outside the North Sea, and described A. holocephularum from the clasp- ers of ratfish captured off British Columbia. Most morphometric studies of ratfish have been descriptive or histological in nature. San- ford, Clegg, and Bonham (1945) studied the liv- er oil and vitamin A content of 35 ratfish cap- tured off Tatoosh Island, Wash. These factors were related later to size and sex of ratfish by Pidlaoan (1950). Halstead and Bunker (1952) described the venom apparatus and anatomy of the dorsal spine of ratfish. They concluded that the venom of H. colliei was not capable of in- flicting fatal injuries to man. A histological study of the digestive tract and of the pituitary of H. colliei was conducted by Clothier (1957) and Sathyanesan (1965), respectively. Stanley (1961) performed a morphometric study of the genital systems of H. colliei and found that summer was the peak reproductive period although one-third of the females and all of the males evidenced reproductive activity throughout the year. Sexual maturity was at- tained at 24-25 cm (S-V)' for females and 18.5- 20 cm (S-V) for males. METHODS Four collections totaling 292 ratfish were made by otter trawl off the coast of Oregon at depths ranging from 50 to 120 fm from 1965 to 1967. The collection from off Newport (iV = 189) was frozen while specimens collected off Cape Blanco {N = 44), Cape Arago {N = 35), and Astoria {N = 24) were preserved in 10% Formalin. All ratfish were examined as follows: Sex was determined by inspection of the gonads; snout to vent length was measured in milli- meters, and total weight was measured in grams; all specimens were examined for internal and ex- ternal parasites; alimentary canals were exam- ined along their entire length for food items; and dental plates, dorsal spines, left pectoral fins, and a piece of the vertebral column were decal- cified, sectioned frozen, treated with Delafield's hematoxylin stain, and examined for growth structures indicative of age. Both eye lenses were removed from specimens in the Newport and Astoria collections. The wet and dry weight of each lens was determined to the nearest ten thousandth of a gram. For wet-weight deter- minations, lenses were stored in 10% Formalin for 1 month, removed and blotted, and immedi- ately weighed. For dry weight determinations, the lenses were then desiccated at 80°C for 82 hr and reweighed. The 82-hr drying period was determined from a curve of weights of 10 lenses dried at 80 °C and weighed at progressive time intervals. The 82-hr period assured evapora- tion to a stable weight. Most statistical analyses of body length-body weight and body length-eye-lens weight relation- ships were performed on a CDC 3300 computer" utilizing program FISH 6669 in the Department at Oregon State University. RESULTS AND DISCUSSION There was a highly significant correlation (P = 0.01) of body length to body weight for male and female ratfish in the large Newport collection and for the aggregate of each sex col- lected (Table 1). Based on the coefficient of determination (r^) (Croxton, 1953), more than 87% of the variation in weight in males and more than 96% in females was attributable to the variation in length of the ratfish. The length- * Body length measured from the tip of the snout to the anterior edge of the vent. ° Reference to trade names in this publication does not imply endorsement of commercial products by the National Marine Fisheries Service. 422 JOHNSON and HORTON: RATFISH, ASPECTS OF BIOLOGY Table 1. — Data to describe length -weight relationship (log weight = log a + 6 log length) for male and female ratfish collected off Oregon during 1965-67. Location Sex Sample size Constant log a Constant b Sign. Level of r {P = 0.01)* 2»» Newport Newport Totalt Totalt Male Female Male Female 128 56 175 112 —2.0168 —3.1384 -4.3217 —9.1692 2.0447 2.5336 3.0546 2.9720 * From Table X in Quenouille (1952). ** Coefficient of determination (Croxton, 1953). t Composite of collections from Newport, Astoria, Cope Arogo, and Cape Blanco. 0.9352 .9824 .9917 .9943 0.234 .361 .210 .257 0.8746 .9651 .9835 .9886 weight relationship for male and female ratfish collected off Newport is defined and illustrated in Figure 1. A taxonomic list of all food organisms identi- fied from the alimentary canals of 283 ratfish is presented in Table 2. The table also contains lists of the relative importance of food items by the frequency of occurrence and numerical methods (Lagler, 1956) and gives the locations of the collections in which the food items were found. Based on these data, ratfish appear to be op- portunistic feeders. The most important food items (>10% occurrence) were shrimp {Pan- dalus and Crago) , mollusks (Mtcsculus and Am- phissa) , and echinoderms (Brisaster) . In gen- eral, young and adult ratfish ate the same foods. Dean (1906) found seaweed in the alimentary canals of ratfish, but we did not find any plant materials in the specimens we examined. In the Cape Arago collection, ratfish were eaten by ratfish. One egg capsule and a caudal fin were eaten by two large females (280 mm). We are not aware of any previous record of can- nibalism in ratfish. Of the teleostomi, flatfish appeared to be taken most frequently by ratfish. The two flatfish found were Hippoglossoides elassodon and Esopsetta jordani. In the food habits study, we did not use a volumetric method of examination because in many alimentary canals only shells and frag- ments remained. Also, materials such as car- apaces of shrimp have a large surface to volume relationship which causes them to displace little water or to float. Ratfish often void ingested matter between capture and landing, making volumetric measurements inaccurate. Dean (1906) commented on this habit, and we noticed it in ratfish captured by hook and line. In ad- 3.2 - 3.0 I 2 8 o Uj *2.6 >- Q O (D2 4 o O _1 2.2 0 = FEMALES LOG WEIGHT= -3.1384+2.5336 LOG LENGTH r= 9824, r2".965l A = MALES LOG WEIGHT = -2,0168 +2 0447 LOG LENGTH r=9352, r2=.B746 . J I 2.1 2.2 2.3 2.4 LOG BODY LENGTH (S-V) (mm) 2.5 Figure 1. — Length -weight relationship of male and fe- male ratfish collected off Newport, Oreg., 1965. dition, we found Gyrocotyle in the mouths of some trawl-caught ratfish, indicating that the contents of the alimentary canals recently had been voided. According to Lynch (1945), Gyrocotyle would normally be restricted to the anterior section of the intestine (spiral valve) . The parasites found on or in the ratfish were Gyrocotyle urna, G. fimhriata, and the copepod, Acanthochondria sp. Table 3 lists the frequency of Gyrocotyle occurring in the four collections of ratfish. Both G. fimhriata and G. urna ap- peared in about equal numbers in the Newport collection, but only G. fimhriata occurred in the Astoria and Cape Arago collections. The fre- quency of infestation by Gyrocotyle in young fish from the Cape Blanco collection was 30%. We did not find evidence of mass infestation by G?/rocoi^^e, as suggested by Wardle (1932). The young fish (<50 mm) contained from zero to two Gyrocotyle each. The voiding of canal con- tents by the ratfish interfered with obtaining an accurate estimation of the degree of infestation. 423 FISHERY BULLETIN: VOL. 70, NO. 2 Table 2. — Taxonomic list' and relative importance of food organisms identified from the alimentary canals of 283 ratfish (Hydrolagus colliei)- collected off Oregon during 1965-1967. Numerical method* Location* Frequency ot occurrence (%) Organisms Organisms %- A N CA CB Phylum Annelida Class Polychaeta Jphrodita 4.4 10 1.0 X X X Unidentified 2.2 10 1.0 X X Phylum Mollusca Class Gastropoda ^mphiiia 33.4 1,016 74.5 X X X X Amygdalum 2.2 9 1.0 X Yoldia 2.6 6 0.0+ X Atuiculus 24.6 66 4.9 X X X X Leptofecttn 0.9 2 0.0+ X Picten 1.8 7 0.0+ X Cardiomya 6.1 27 2.0 X Calliostoma 0.4 1 0.0+ X Scarlesia 0.4 1 0.0+ X Class Scaphopoda Dentalium 3.1 7 0.0+ X X Phylum Arthropoda Class Crustacea Livonica 4.8 26 2.0 X X X Crago il2.3 32 2.3 X Pandalus 20.2 81 5.9 X X X Canctr 0.9 2 0.0+ X X Chioncctes 0.4 1 0.0+ X Unidentified 3.3 13 1.0 X Phylum Echinodermata Class Echinoidea Brisaster 11.0 25 1.8 X X X Strongylocentrotus 0.9 2 0.0+ X X Phylum Chordato Class Chondrichthes Hydrolagus 0.9 2 0.0+ X Class Osteichthyes Pleuronectidao 0.7 2 0.0+ X X Unidentified 5.5 13 1.0 X X X Unknown 0.4 1 0.0+ X Total 1,362 98.4+ 1 After Smith et al. (1954) , Barnes (1963), and Bailey (1970). 2 Of the 283 ratfish 224 contained food organisms. 3 Fragments (less than one half an animal) were recorded as one ndividual. * Locations were: A = Astoria, N = Newport, CA = Cape Arac o, and CB = Cape Blanco. Gyrocotyle were found lodged in the folds of the spiral valve and were not embedded in the in- testinal wall, thus making expulsion by violent intestinal movements possible. In the Cape Arago collection, 7 of the 21 adult male ratfish had from two to eight Acanthochon- dria sp. attached to the free ends of their clasp- ers. The immature males and the females did not carry this copepod. The species is similar to, but not the same as, A. compacta." * Personal communication, Dr. Satyu Yamaguti, Belts- ville, Md., June 13, 1967. An unidentified fungus, which occurred on the intestine of 29% of the Newport collection, was not necessarily a parasite as it may have devel- oped in the interval between capture and preser- vation. No visible lesions or other damage were noticed on the body or alimentary canal surfaces of the ratfish in which the fungus occurred. The fungus appeared to be of nonseptate, white, fil- amentous type. Sex and relative age of ratfish can be deter- mined by examination of the secondary sex char- acteristics. Males possess a frontal tenaculum, prepelvic tenacula, and claspers, whereas fe- 424 JOHNSON and HORTON: RATFISH, ASPECTS OF BIOLOGY Table 3. — Gyrocotyle found in 283 ratfish (Hydrolagus colliei) collected off the coast of Oregon, 1965-1967. Location Total number of Gyrocotyle'^ G. fimbriata G. Unidentified Newport Astoria Cape Arago Cape Blanco 50(34) 8( 7) 40(23) 18(11) 62(39) 4( 2) 38(25) Number of Percent ratfish infested" examined 184 53.2(60.0) 24 29.2(31.8) 35 65.8(79.4) 40 30.0(35.4) ^ Number in parenthesis is the number of ratfish Infested. ^ Number in parenthesis is the percent infestation when alimentary canals that contained neither food nor para- sites were excluded. males do not possess these structures but de- velop paired oviducal openings not possessed by males. Development of these structures can be used to separate ratfish into young, immature, and adult age groups. Young males have a frontal tenaculum streak and diminutive claspers (Figure 2) ; immature males have a small fron- tal tenaculum and claspers which are not perfo- rated at their free ends (Figure 3) ; and mature males have a well-developed frontal tenaculum and well-developed claspers which are perforated at their free ends (Figure 4). Young females have no oviducal openings (Figure 2) ; the ovi- ducts of immature females have small openings (Figure 3) ; and mature females have oviductal openings which are large, elongated, and swollen M METRIC Figure 2. — Regions of young ratfish showing (A) the frontal tenaculum streak, and (B) the small claspers of the male; and (C) the absence of a frontal tenaculum streak, and (D) the absence of oviductal openings in the female. 425 FISHERY BULLETIN: VOL. 70, NO. 2 D B. "f yt , A- 'W\ '.ifuVc ■ ■ ' ! ■ i M ■ 1 "■ * '*■"'■■ i' lii!i|lilil!i|i;iiil. Figure 3. — Regions of immature ratfish showing (A) the frontal tenaculum, (B) the prepelvic tenacula, and (C) the claspers of the male; and (D) the absence of a frontal tenaculum, and (E) the presence of small openings to the oviducts of the female. (Figure 4) . Attempts to age ratfish on the basis of length-frequency distributions were incon- clusive. When dry eye-lens weight was compared to body length for specimens in the Newport col- lection, lens weight was positively correlated with increasing body length for males (A^ = 128), but was not correlated with body length for females (N' = 56) (Figure 5), There was no difference in the dry weights of the right and left eye lens at the 95^^ level of confidence. Wet eye-lens weights were similarly related to body lengths with the coefficient of determination (r-) for males being 0.8788 and 0.9292 for the right and left eye lens respectively, and for females being 0.0017 and 0.0133 for the right and left eye lens respectively. We can offer no logical explanation for the lack of positive correlation between eye-lens weight and body length for female ratfish. Data for males and females were processed simulta- neously and were consistent, by sex, for left and right and for wet and dry eye-lens weights. Because all but eight females exceeded 230 mm in length, most eye-lens growth may take place between birth (30-40 mm [Stanley, 1961]) and maturity (240-250 mm [Stanley, 1961]). The possibility of decreasing density of the lens with size (and maturity) should be investi- gated. 426 JOHNSON and HORTON: RATFISH, ASPECTS OF BIOLOGY D Figure 4. — Regions of mature ratfish showing (A) the frontal tenaculum, (B) the prepelvic te- nacula, and (C) the claspers, of the male; and (D) the absence of a frontal tenaculum, and (E) the presence of well-developed openings to the oviducts of the female. In general, there was an increase in the size of other body parts (teeth, vertebrae, base of left pectoral fin, and base of dorsal spine) with in- creasing body length. We did not find any lay- ering or structures in these body parts which were sufficiently correlated with body length to provide a possible means of age determination. The number of tritors (horizontal ridges) on the posterior side of the left vomerine dental plate was compared to the respective body length of male and female ratfish in the Cape Blanco and Newport collections. In general, the num- ber of tritors increased with increasing body length (Figure 6) . Two problems arose in using this structure as a basis for age determination: (1) No comparison to known-aged fish was pos- sible. (2) The amount of wear on these ridges per unit of time was not known. ACKNOWLEDGMENTS Drs. C. E. Bond, R. E. Millemann, R. C. Simon, and J. D. Hall, Department of Fisheries and Wildlife, Oregon State University, provided counsel and reviewed the manuscript. Dr. W. S. Overton, Department of Statistics, Oregon State University, provided direction in the statistical analyses. Special thanks are due personnel of the Fish Commission of Oregon for assistance in obtaining ratfish for this study. James Mee- han and Gary Milburn were particularly helpful. 427 FISHERY BULLETIN: VOL. 70, NO. 2 o. .000 X o UJ z bJ _i -.333 • -RIGHT LENS OF FEMALES. r2=.000l D-RIGHT LENS OF MALES. r2=.8944 O-LEFT LENS OF FEMALES. r2=.0242 A-LEFT LENS OF MALES. r2=.9l69 A > cr o o o _ figyi ^^ I 2.00 2.25 2.50 LOG BODY LENGTH (S-V) (mm) Figure 5. — Log dry eye-lens weight compared to log body length for ratfish collected off Newport, Oreg., 1965. 10 en O ^ cr q: 10 UJ CD =) z 5 MALES FEMALES W ^:^ ^h — RANGE .^MEAN ■^PLUS OR MINUS ONE STANDARD DEVIATION OF THE MEAN I I _1_ 50 100 150 200 250 BODY LENGTH (S-V) (mm) 300 Figure 6. — Comparison of the number of tritors on the posterior side of the vomerine dental plate to body length of ratfish collected off Oregon, 1965-1967. LITERATURE CITED Alverson, D. L., and M. E. Stansby. 1963. The spiny dogfish (Squalus acanthias) in the northeastern Pacific. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 447, 25 p. Bailey, R. M. (chairman). 1970. A list of common and scientific names of fish- es from the United States and Canada. Am. Fish. Soc, Spec. Publ. 6, 149 p. Barnes, R. D. 1963. Invertebrate zoology. W. B. Saunders, Phila., 632 p. BiGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the western North Atlantic. Part Two. Sawfishes, guitarfishes, skates and rays [and] chimaeroids. Mem. Sears Found. Mar. Res., Yale Univ. 1, 588 p. BONHAM, K. 1950. Some monogenetic trematodes of Puget Sound fishes. In M. H. Hatch (editor), Studies honoring Trevor Kincaid, p. 85-103. Univ. Wash. Press, Seattle. Clemens, W. A., and G. V. Wiley. 1961. Fishes of the Pacific coast of Canada. 2d ed. Fish. Res. Board Can., Bull. 68, 443 p. Clothier, G. E. 1957. The histology of the digestive tract of the chimaeroid fish Hydrolagua colliei (Lay and Ben- nett). M.S. Thesis, Oregon State Coll., Corvallis, 43 p. Crescitelli, F. 1969. The visual pigment of a chimaeroid fish. Vision Res. 9:1407-1414. Croxton, F. E. 1953. Elementary statistics with applications in medicine and the biological sciences. Dover Publ., N.Y., 376 p. Dean, B. 1906. Chimaeroid fishes and their development. Carnegie Inst. Wash. Publ. 32, 194 p. Halstead, B. W., and N. C. Bunker. 1952. The venom apparatus of the ratfish, Hydro- lagus colliei. Copeia 1952:128-138. Kabata, Z. 1968. Some Chondracanthidae (Copepoda) from fishes of British Columbia. J. Fish. Res. Board Can. 25:321-345. KORATHA, K. J. 1960. Preliminary report on the parasites of a chimaeroid fish, Hydrolagus colliei (Lay and Ben- nett) from the eastern Indo-Pacific with comments on the phylogenetic significance of parasitic dis- tribution in Holocephali. (Abstr.) J. Parasitol. 46(5) Sect. 2:14. Lagler, K. F. 1956. Freshwater fishery biology. 2d ed. Wm. C. Brown, Dubuque, 421 p. Legendre, R. 1944. Notes biologiques sur Chimaera monstrosa L. Bull. Soc. Zool. Fr. 69(1) : 10-17. Lynch, J. E. 1945. Redescription of the species Gyrocotyle from 428 JOHNSON and HORTON: RATFISH, ASPECTS OF BIOLOGY Lynch — Cont. the ratfish Hydrolagus colliei (Lay and Bennett), with notes on the morphology and taxonomy of the genus. J. Parasitol. 31:418-446. Nakatsu, L. M. 1957. A review of the soupfin shark fishery of the Pacific coast. Commer. Fish. Rev. 19(12) :5-8. Olsson, p. 1896. Sur Chimaera inonstrosa et ses parasites. Mem. Soc. Zool. Fr. 9:499-512. PiDLAOAN, N. A. 1950. The vitamin A potency of liver oil and the oil yield in the ratfish (H. colliei) of Puget Sound, and the relationship of these factors to size and sex of the fish. M.S. Thesis, Univ. Washington, Seattle, 45 p. QUENOUILLE, M. H. 1952. Associated measurements. Butterworths Sci. Publ., Lond., 242 p. Sanfobd, F. B., W. Clegg, and K. Bonham. 1945. The ratfish. Fish. Mark. News 7(11) :6-7. Sathyanesan, a. G. 1965. The hypophysis and hypothalamo-hypophys- ial system in the chimaeroid fish Hydrolagics colliei (Lay and Bennett) with a note on their vascularization. J. Morphol. 116:413-449. Smith, R. I., F. A. Pitelka, D. A. Abbott, and F. M. Weesner. 1954. Intertidal invertebrates of the central Cali- fornia coast: S. F. Light's "Laboratory and field text in invertebrate zoology." Revised ed. Univ. Calif. Press, Berkeley, 446 p. Stanley, H. P. 1961. Studies on the genital systems and reproduc- tion in the chimaeroid fish Hydrolagus colliei (Lay and Bennett). Ph.D. Thesis, Oregon State Univ., Corvallis, 94 p. Thompson, W. F. 1915. A preliminary report on the life-history of the halibut. B.C. Rep. Comm. Fish. 1914, p. 76- 99. Wardle, R. a. 1932. The cestoda of Canadian fishes. I. The Pa- cific coast region. Contrib. Can. Biol. Fish., New Ser. 7:221-243. Wilson, C. B. 1935. Parasitic copepods from the Pacific coast. Am. Midi. Nat. 16:776-797. 429 OCEAN DISTRIBUTION, GROWTH, AND EFFECTS OF THE TROLL FISHERY ON YIELD OF FALL CHINOOK SALMON FROM COLUMBIA RIVER HATCHERIES Kenneth A. Henry^ ABSTRACT Data are presented depicting the distribution of some stocks of Columbia River hatchery fall Chinook salmon in the northeast Pacific Ocean. These are based on recoveries of marked fish. Also presented are the apparent growth histories for fish from the Kalama and Spring Creek Hatcheries as well as a graphic population model for Columbia River fall chinook salmon. Finally, the effect on total yield for varying troll fishing mortalities for the 3-, 4-, and 5-year-old fish is analyzed and the results depicted in three-dimensional yield diagrams. In most instances, at least at the lower levels of estimated natural mor- tality, troll fishing of the younger fish reduced total yield. A cooperative program was undertaken between the various fishery agencies in the United States and Canada to estimate the contributions to var- ious fisheries of fall chinook salmon (Oncorhyn- chiis tshawytscha) produced by a number of Co- lumbia River hatcheries. Approximately 10% of the output from 12 hatcheries for the brood years 1961-64 was marked by the removal of certain fins. The design of this experiment, including a detailed account of the procedures used in cal- culating the number of marked fish recovered, and details of mark recoveries for the 1961 brood have been reported by Worlund, Wahle, and Zim- mer (1969). Details of mark recoveries for the 1962 and 1963 broods were given by Rose and Arp (1970)' and by Arp et al. (1970', respec- tively. Cleaver (1969) made a detailed analysis ^ National Marine Fisheries Service Northwest Fish- eries Center, 2725 Montlake Boulevard East, Seattle, WA 98102. ' Rose, J. H., and A. H. Arp. 1970. Contribution of Columbia River hatcheries to harvest of 1962 brood fall chinook salmon {Oncorhynchus tshawytscha) . Bur. Commer. Fish., Columbia Fish. Program Off., Appraisal Section, Portland, Oreg. 27 p. (processed). ' Arp, A. H., J. H. Rose, and S. K. Olhausen. 1970. Contribution of Columbia River hatcheries to harvest of 1962 brood fall chinook salmon (Oncorhynchus tsha- wytscha). U.S. Dep. Commer., Natl. Oceanic Atmos. Ad- min., Natl. Mar. Fish. Serv., Columbia Fish. Program Off., Portland, Oreg., Econ. Feasibility Rep. 1. 33 p. (processed). of the 1961 brood based on mark recoveries. His analysis included estimating ocean mortality rates, maturity schedules, and, for the Spring Creek and Kalama fish, the effect of no ocean fishing on total yield. Henry (1971) made a similar analysis for the 1962 brood and, wher- ever possible, compared the results with those obtained for the 1961 brood. The general re- lease and recovery areas covered by this marking program are shown in Figure 1. Although a total of 12 hatcheries was involved in this study, fish from only two hatcheries, Spring Creek and Kalama, received specific identifying marks for all 4 years. Each year fish from two of the other ten participating hatcheries (but two different ones each year) also received special identifying marks. In ad- dition to specific marks for four hatcheries each year, a certain proportion of the output from all participating hatcheries had a common mark. This mark — that is a composite of all the hatch- eries — is referred to as the general mark. Thus, it is possible for only Spring Creek, Kalama, and general mark fish to analyze mark recoveries from all four brood years. In the report, I have analyzed the mark re- coveries for Spring Creek, Kalama River, and general marked Columbia River hatchery fall chinook salmon from the four consecutive brood Manuscript accepted January 1972. FISHERY BULLETIN: VOL. 70. NO. 2, 1972. 431 FISHERY BULLETIN: VOL. 70, NO. 2 FiGLTRE 1. — Sampling areas in marine fisheries and release locations (inset) for the Columbia River hatchery eval- uation program (from Lander, 1970). 432 HENRY: FALL CHINOOK SALMON years (1961, 1962, 1963, and 1964) . I have dem- onstrated some of the differences in ocean dis- tribution, contributions to various fisheries, and growth that occur between hatcheries — even within hatcheries — for different brood years. I also have presented a graphic population model for Columbia River fall chinook salmon. This model depicts all the factors affecting these stocks of salmon throughout their life history. Finally, within the general framework of the mortality estimates developed for the 1961 brood (Cleaver, 1969) and 1962 brood (Henry, 1971), I have analyzed the portion of the population model pertaining to the commercial fisheries to determine the overall effect on total yield for varying levels of ocean troll fishing mortalities on the 3-, 4-, and 5-year old fish. RECOVERIES BY AREA The ocean distribution of mark recoveries and contributions to the various fisheries was quite different for Kalama, Spring Creek, and general marked fish. Pulford (1970) listed the contri- bution of Columbia river hatchery fall chinook (all hatcheries combined) to the various fisheries along the Pacific Coast for 1966 only. Lander (1970) analyzed the distribution along the coast and contribution to the various fisheries in con- siderable detail for each hatchery; his data in- cluded sampling through 1966 only. In Table 1 are listed the calculated total recoveries by geo- graphical area for the 1961 through 1964 brood years of special marked Kalama and Spring Creek fall chinook salmon and general marked Table 1. — Calculated total recoveries, by age of fish and type of mark, for fall chinook salmon of the 1961-64 broods that were marked and released at Columbia River hatcheries. Spri ing Creek mark Kalama t ■nark General i mark Brood and age (year; i) age (years) age (years) area 2 3 4. 5 Total 2 3 4 5 Total 2 3 4 5 Total 1961 Alaska » 0 0 0 0 « 5 35 4 44 « 0 7 0 7 Brif. Col. « 841 164 5 1,010 ♦ 441 480 91 1,012 » 4,106 1,871 218 6,195 Wash. Com. 4 1,084 82 0 1,170 0 149 142 7 298 0 3,241 455 41 3,737 Wash. Sporf 152 431 97 0 680 21 78 100 9 208 375 1,681 416 67 2,539 Oreg. Ocean 4 130 24 0 158 0 21 4 0 25 0 396 36 0 432 Col. River 22 518 685 17 1,242 0 38 399 111 548 72 2,158 3,544 176 5,950 Calif. * 25 0 0 25 » 2 0 0 2 * 23 0 6 29 Total 182 3,029 1,052 22 4,285 21 734 1,160 222 2,137 447 11,605 6,329 508 18,889 1962 Alaska 0 0 0 0 0 0 0 4 0 4 0 0 5 2 7 Brif. Col. 0 75 90 9 174 0 162 155 23 340 51 1,183 802 48 2,084 Wash. Com. 0 150 33 5 188 0 47 15 0 62 8 973 130 4 1,115 Wash. Sport 34 140 24 0 198 0 76 7 8 91 163 540 108 27 838 Oreg. Ocean 0 11 3 0 14 0 8 4 0 12 2 37 3 0 42 Col. River -10 272 85 0 367 6 21 60 10 97 50 1,216 606 24 1,896 Calif. 0 0 0 0 0 0 0 0 0 0 0 6 0 0 6 Total 44 648 235 14 941 6 314 245 41 606 274 3,955 1,654 105 5,988 1963 Alaska 0 0 0 0 0 0 5 14 0 19 0 0 9 • 9 Brit. Col. 23 557 224 6 810 0 233 195 45 473 55 4,464 2,246 201 6,966 Wash. Com. 0 381 38 0 419 0 107 71 5 183 5 3,227 464 10 3,706 Wash. Sport 120 329 103 0 552 138 139 53 12 342 1,189 2,569 , 451 46 4,255 Oreg. Ocean 0 63 10 0 73 2 27 13 0 42 4 652 165 3 824 Col. River 29 164 275 15 483 7 32 44 47 130 121 1,182 2,418 317 4,038 Calif. 0 4 0 0 4 0 0 0 0 0 0 12 12 15 39 Total 172 1,498 650 21 2,341 147 543 390 109 1,189 1,374 12,106 5,765 592 19,837 1964 Alaska 0 0 # « 0 0 0 « • 0 0 0 « * 0 Brit. Col. 7 589 432 16 1,044 0 45 478 64 587 10 1,339 1,446 92 2,887 Wash. Com. 0 906 127 0 1,033 0 65 37 5 107 4 2,268 354 3 2,181 Wash. Sport 244 581 70 10 905 38 44 26 0 108 483 1,506 249 0 2,212 Oreg. Ocean 127 171 27 0 325 0 10 13 0 23 151 614 170 0 935 Col. River 15 447 504 24 990 0 3 58 56 117 19 1,001 1,204 169 2,393 Calif. 0 8 2 0 10 0 0 0 0 9 0 0 1 0 1 Total 393 2,702 1,162 50 4,307 38 176 612 125 951 667 6,728 3,424 266 11,085 • No sampling. 433 FISHERY BULLETIN: VOL. 70, NO. 2 Columbia River fall chinook salmon. No single fin marks are included. These same data are depicted in Figures 2-4 for calculated numbers of marked fish recovered and in Figures 5-7 as percentages of the total number of marks re- covered by each age group within each brood year. More Spring Creek mark recoveries (Figure 2) were recovered as age 3 fish than at any other age. Both the age 2 and age 3 recoveries came principally from Washington State fisheries, whereas most of the age 4 and age 5 recoveries were from the British Columbia troll fishery and the Columbia River gillnet fishery. For all ages combined, Washington State fisheries had the most recoveries of Spring Creek fish with about equal numbers recovered in the British Columbia and Columbia River fisheries. From 18 to 35% of the total recoveries were made in British Co- lumbia fisheries (Figure 5). No Spring Creek fish were recovered from the Alaska fisheries. The recoveries of marked Kalama fish were distributed somewhat differently than marked Spring Creek fish (Figure 3). Although re- coveries of marked 2-year-old fish occurred pri- marily in the Washington sport fishery, 3- and 4-year-old recoveries and the recoveries for all ages combined came principally from the British Columbia troll fishery. Most of the recoveries occurred with the 3- and 4-year-oId fish. British Columbia fisheries had from 40% to 60% of the total recoveries (Figure 6), a considerably higher percentage than for Spring Creek fish. There were very few recoveries from the Alaska fisheries. General mark recoveries seemed to be more similar to Spring Creek recoveries than to Ka- lama recoveries, indicating that most of the fish from the participating hatcheries had an ocean distribution and maturity schedule more like Spring Creek fish than Kalama fish. For the general mark recoveries, the 2-year-old recov- eries were mainly from the Washington sport fishery (Figure 4). The Washington fisheries accounted for 38-53% of the 3-year-old mark recoveries, whereas both the 4- and 5-year-old recoveries came mainly from the British Co- lumbia troll fishery and the Columbia River gill- Fishing Area Age 2 Age 3 Age 4 Age 5 All Ages Alaska BritishColumbia Washington Oregon California NS NS P NS ] a o o n- Alaska BritishColumbia Washington Oregon California 1 1 ? 1 m CM Wf a> Alaska British Colunnbia Washington Oregon California : Alaska BritishColumbia Washington Oregon California 0 200 0 400 800 1200 0 NUMBERS 400 0 OF FISH 20 1000 Figure 2. — Calculated total recoveries (in numbers of fish) of special marked Spring Creek hatchery fall chinook salmon, by age, in different fishing areas, 1961-64 brood years (Columbia River recoveries are shown as the shaded por- tion of the Oregon recoveries). NS = no sample; * = less than 10 recov- eries. 434 HENRY: FALL CHINOOK SALMON Fishing Area Age 2 Age 3 Age 4 Age 5 All Ages Q O Alaska British Columbia Washington Oregon Colif ornio NS NS ] NS * ■j(- 1 O 1 1 m 1 ID 1 ■X- ^ ^^^^^^^ ^^^^1 * Alaska British Columbio Washington Oregon California * ■X- * * 1 1 1 k J i Alaska British Columbia Washington Oregon California * 1 i 1 1 1 1 , 1 1 1 J 1 1 Q O Alaska British Columbia Washington Oregon California ■-I — 1 NS NS 1 * 1 O 1 1 (U ~j 1 1 * I 1 * — 1 1 - 100 0 200 400 0 200 NUMBERS OF FISH 400 0 100 600 Figure 3. — Calculated total recoveries (in numbers of fish) of special marked Kalama hatchery fall chinook salmon, by age, in different fishing areas, 1961-64 brood years (Columbia River recoveries are shown as the shaded por- tion of the Oregon recoveries). NS = no sample; * = less than 10 recov- eries. Fishing Area Age 2 Age 3 Age 5 All Ages Alaska British Columbia Washington Oregon CaHfornia NS NS ? NS Alaska BritishColumbio Washington Oregon California o Alaska BritishColumbio Woshington Oregon Colifornio 1 , * * 8 1 1 , 1 1 m 1 1 1 1 u> ^^^ ^^^^^^ ^^^ ^^^^ * * ♦ ♦ Q O O Alaska BritishColumbio Washington Oregon Colifornio 1 ■_ 1 I NS NS * 1 1 1 1 m 1 m^ wm ^^ 0) 1 1 1 1 1 0 900 0 1,800 3,600 5,400 0 1,800 NUMBERS OF FISH 180 0 4,000 Figure 4. — Calculated total recoveries (in numbers of fish) of general marked Columbia River hatchery fall chinook salmon, by age, in different fishing areas, 1961-64 brood years (Columbia River recoveries are shown as the shaded por- tion of the Oregon recoveries). NS = no sample; * = less than 50 recov- eries. 435 FISHERY BULLETIN: VOL. 70, NO. 2 Fishing Area Age 2 Age 3 Age 4 Age 5 All Ages Alaska BritishColumbia Washington Oregon Californio NS NS NS u o Alaska BntishColumbio Washington Oregon California n n o 1 CD 1 1 1 ID m ^^^^ mm ^^^ ^:;:i: " ft::::::;::™; III 50 0 963 BROOD 50 - iiii nil 1 ^M U 964 BROOD 50 ~ ill ill iiii 0 234523452345 AGE (YEARS) Figure 8. — Calculated total ocean recoveries (in per- cent) , by age, of special marked Spring Creek, Kalama, and general marked Columbia River hatchery fall chi- nook salmon, 1961-64 brood years. GRAPHIC POPULATION MODEL FOR COLUMBIA RIVER FALL CHINOOK To more clearly understand all the forces af- fecting a stock of fish, it often is convenient to depict the population in a simulated flow chart or block diagram. Shapiro and Andreev (1969) show such a diagram for chum salmon (0. keta) . With some modification, their diagram also would be applicable to Columbia River fall Chi- nook salmon as shown in Figure 11. This dia- gram is based on the following stock character- istics. There are five age groups, four of which are capable of entering the spawning groups, and each brood year diminishes in numbers as ^ 10 MALES 1961 BROOD FEMALES 5 2 3 AGE (YEARS) Figure 9. — Average weight (kilograms) by age and sex, of 1961-64 brood Spring Creek and Kalama special marked hatchery fall chinook salmon caught in the Co- lumbia River gillnet fishery. Shaded columns represent data for Kalama salmon — clear columns, Spring Creek. Numbers indicate sample size. a result of natural mortality (M,) and ocean fishing mortality {Fi). Ocean fishing mortality occurs only on the 2 + , 3 + , and 4 + -year-old- fish. Growth and fecundity diflFer for each age group, and entry into the spawning part of the population is aflfected by a probability (Pi) . The spawning group is diminished by a river fishing mortality (RFi) . After river fishing mortality, some portion of the spawning fish (k,) are re- moved for artificial propagation. Survivals to recruitment resulting from artificial and natural reproduction are taken to be difl"erent. The number of recruits (R) is dependent on a sur- vival relationship for the eflfective fecundity, i.e., R/E. EFFECT OF TROLL FISHING ON YIELD A complete analysis of the Columbia River fall Chinook population, as depicted in Figure 11, is beyond the scope of this paper. However, I 438 HENRY: FALL CHINOOK SALMON - 3 •C + + 1.0 SPRING CREEK B 16 o. 12 J L KAL AMA B H.O + 0.5 I - 0 - -0.5 - 20 16 ' 12 4- J 1 L 0 to w/3 (kg) 2 4 6 AGE (YEARS) 2 4 6 AGE (YEARS) Figure 10. — Successive stages of fitting Spring Creek and Kalama Hatchery fall Chinook salmon weight data from Figure 8 to the von Bertalanffy growth equation. Part A shows the relation between average weight at age t and age i + 1 — point on the dotted line where Wf/i = w^hf + i is an estimate of W ^Vi. Part B shows the relation between age and a logarithmic function of the weight used to estimate the value of k (i.e., the slope). Part C shows the calculated growth curve based on the values computed from Parts A and B. want to analyze a part of the model, the part dealing with the ocean and river fisheries, to show the calculated effect the troll fishery has on total yield for a particular brood year. In these analyses, I have used data developed for the 1961 and 1962 brood years for the Spring Creek and Kalama fall chinooks, based on studies by Cleaver (1969) and Henry (1971), respec- tively. Where there were some data missing for the 1962 brood Spring Creek fish, I used 1961 data. The starting point for the analyses is at age 1+ in Figure 11. It is apparent from Figure 11 that given Fu Mi, Pi, and RF,, as well as data on average weights of fish at each age, total yields from the ocean and river fisheries for a given number of recruits (R) can be calculated. Henry (1971) gives estimates of Fi and Pi for various values of Mi as well as values for RFi for 1961 brood Spring Creek and Kalama fish and 1962 brood Kalama fish. Lack of river recoveries of 5-year- old fish prevented estimates of these values for 1962 brood Spring Creek chinook salmon. Henry also 'ists average weight data, by age, for both 439 FISHERY BULLETIN: VOL. 70, NO. 2 Figure 11. — Block diagram of a population model for Columbia River fall chinook salmon (adapted from Sha- piro and Andreev, 1969). Mi =■ natural mortality rate. Fj = ocean fishing mortality rate. RFi = river fishing mortality rate. Pj = probability of being a spawTier. ki = percentage of spawners taken for artificial reproduction. El = effective fecundity/1,000 eggs. R = number of recruits. 1 + , 2+, 3 + , 4+ = age of fish. the ocean and river recoveries of these two stocks of fish. The procedures followed in these analyses were: given a certain number of recruits (1,000) and a certain Mi (natural mortality) value, and assuming M2 = M3 = M4, 1 calculated the total yield for all ages (numbers caught X average weight) from both the ocean and river fisheries using the Pi values (where Pi is the proportion of the ocean population entering the river to spawn) for the M, as given by Henry (1971). I let the Pi and RFi values (where RFi is the river fishing mortality) remain con- stant but varied the Fi's (where Fi is the ocean fishing mortality) from 0 to 1.8. This procedure resulted in a 3-dimensional yield diagram (i.e., Fs, Fi, and Fo) . The computations were done on an IBM 1130* computer and I wrote the pro- gram so the computed yields were plotted di- rectly in even yield planes by a CalComp plotter. Regardless of the natural mortality rate as- sumed for these computations, it appears that total yields under actual conditions were below the potential yields for both Spring Creek and Kalama fall chinook. In Figure 12 are depicted the calculated maximum potential yields for var- ious values of natural mortality (Mi) and the ocean fishing values {Fi) needed to achieve these maximum yields as well as the calculated yields based on estimated fishing mortality rates for various values of M as given by Henry (1971). It is only with the higher levels of natural mor- tality (>0.60) or with the 1962 brood Spring Creek fish that a troll fishery on the 3-year-olds (F3) would have increased yield. At all levels of natural mortality shown, maximum yield re- quired maximum troll fishing effort on the 5- year-old fish {F-,) . The calculated total yield varied considerably depending on the F and M values used. The 3- dimensional outputs for M = 0.24 and Ft — 0 to 1.8 for the two hatcheries for the 1961 brood year are shown in Figures 13 and 14. I have included a yield diagram for each stock to show the differences in yield that can be generated in this type of analysis as well as to emphasize the differences between these two groups of salmon. The calculated yield planes shown in these figures are actually planes of equal yield passing through the block diagrams. Each yield plane shown consists of points that repre- sent all possible combinations of ocean fishing mortality (Fs, F4, and F:,) — such that total yield will equal the value shown. In Figure 13, for 1961 brood Spring Creek fish (M = 0.24), the maximum yield is at point A (Fa = 0, F4 = 0, and Fs = 1.8) and is slightly over 6,300, the same value as shown in Figure 12. Yield dia- grams calculated for M = 0.96 were similar to * Use of trade names does not imply endorsement by the National Marine Fisheries Service. 440 HENRY: FALL CHINOOK SALMON z> (T O tiJ IT O O O 6,000 3,000 0 6,000 3>000 ft 0 6,000 < f- o 3,000 - 1961 BROOD SPRING CREEK 0 0 i.e 0 0 1.8 1.8 0 1.8 -1.8- 0 18 F,- 0 F4. 0 1961 BROOO KALAMA 0 0 1.8 0 0 1.8 0 0 1.8 -1.8' *l.8" 1.8 1962 BROOD KALAMA ^5- 0 0 1.8 0 0 1.8 yield = 0 = 1.8 1.8 stimatedj 'potentio actual II 1 1 yield .MOKimum 1.8 '1.8- 1.8 .24 .45 60 .72 NATURAL MORTALITY .96 Figure 12. — Calculated maximum potential yields (kil- ograms) per 1,000 recruits for the 1961 brood Spring Creek and Kalama and the 1962 brood Kalama hatchery fall Chinook salmon compared with yields calculated for estimated rates of actual ocean fishing mortality (F values shown are those producing maximum potential yields) . those shown except that they depicted lower yields, as would be expected with a higher na- tural mortality. Similar analyses also were made from the 1962 brood data, but those yield diagrams are not depicted. In a similar manner to that shown for obtain- ing the maximum yield, the calculated total yield for any combination of Fs, Fa, and F-, could be estimated from these graphs. However, the purpose of these graphs is not to estimate single yield values but rather to show the overall effect on yield of varying values of Fs, Fa, and Fs. Thus, Figure 13 shows that the total yields di- minish as Fz and Fa increase but that Fs has little effect on total yield. Under these condi- tions, any troll fishing on 3- and 4-year-old fish would reduce the total yield. It might be well to re-emphasize that these calculations are based on the assumption of a constant M for the period analyzed. Variations in M during this period would, of course, affect the calculated total yield depending on when the variations occurred and their magnitude. How- ever, a differential M before entry into the troll fishery, i.e., at age 2 and younger, would not affect this analysis. There is considerable variation in the two yield diagrams, and the reader will have to ex- amine each figure to see in detail the effect of varying levels of troll fishing for that particular brood and level of natural mortality. In gen- eral, at the lower levels of natural mortality (Af = 0.24) , increased troll fishing on the young- er age group (2 + and 3 + ) results in reduced yields, whereas the reverse of this generally occurs at the higher level of natural mortality (Af = 0.96). In other words, at the younger ages and the lower natural mortalities it appears that the rate of growth exceeds the loss due to natural mortality and, consequently, yield is in- creased by letting the salmon reach an older age before harvesting them. It should be pointed out that even if no fish had been caught in the ocean, the number returning to the river would have been less than the sum of river entry and ocean catch since some of the fish caught in the ocean fishery would have died from natural causes. The values of M and F that I used, although arbitrary, are believed to cover the range of realistic possible values. Obviously, any values of M could be inserted into the program, but it was felt that M = 0.24 and M = 0.96 would depict extremes. Also, Ft values greater than 1.8 could be used, but I arbitrarily limited them at this upper boundary because F values of 1.8 considerably exceed any of the F values calcu- lated for either the 1961 or 1962 brood data (Henry, 1971). The calculated total yields for 1961 and 1962 brood Spring Creek and Kalama fall chinook salmon for various values of natural mortality — if troll fishing could be restricted to only one age group; i.e., troll fishing on only the 3-, 4-, or 5-year-old fish — are shown in Figures 15 and 16. The values for river fishing intensity and proportion spawning are the same as used in the previous analyses. The yields shown are the 441 FISHERY BULLETIN: VOL. 70, NO. 2 Figure 13. — Calculated total yield diagrams (in kilograms per 1,000 recruits) for the 1961 brood Spring Creek hatchery fall chinook salmon with constant natural mortality (M = 0.24), varying ocean troll fishing mortality rates (F3, F4, and F5), and the proportion spawning and river fishing intensity remaining constant. Yields are shown by 100-kilogram plane intervals. 442 HENRY: FALL CHINOOK SALMON of^ Figure 14. — Calculated total yield diagrams (in kilograms per 1,000 recruits) for the 1961 brood Kalama hatchery fall chinook salmon with constant natural mortality (M = 0.24), varying ocean troll fishing mortality rates (F3, F^, and F5) , and the proportion spawning and river fishing intensity remaining constant. Yields are shown by 100- kilogram plane intervals. 443 FISHERY BULLETIN: VOL. 70, NO. 2 1961 BROOD M = 24 45 60 .72 .96 ^5 ^4 ----F3 6,000 4,500 V W) V \-~., \."~--.._ i 3,000 V---. o UJ a: g 1,500 t» - 6,000 o u 4,500 "^ . 1962 BROOD \ y 3,000

i t u Pi o rrt -u >. rt ^ X 5 -tj rt Q^ SM O -kJ -a s O S2 o s > q= (1; ""' U [ [ U 1 1 o ■"—a t—l 03 S i-) CO (D3M puD inojjoas pguodj qs!}3n|g jspunO|j iods J3>|D0J3 3HUD|4V 3p]SJ3A|!S D!(UDHV HS]J |DAJD|(SOd/|DAJD-| S|!DU9 sqojs en|g sjapaa} 4!Sodap jaqjo sjapaaj Ja4|!J Jamo Aaol|dud Aog ja||nui paduts uapDHuauj d]4udhv (!}}>U0M>VIVJ) dLUjjqs (sn}vu?j) dLuuqs do||DDS Aog oiop pJDH jaisAo uDDuaoiv DunD^OjaVM saqoj3!UJ siqdojiOjaiaH uoj>(UD|doo2 uo;>(UD|dojAqd saiAqdida puo ssDj6|ag saiAL|d|da puD aD6|D D|L)(uag SSDJ6 4|DS DjoijojDjUJ puo snjuiap a4D|nD!(jDd papuadsn^ DJO|pj3!UJ puD S4isod3p 4uauJ!pas e3D^jn9 4uaui!pas pDO|-pag sjuaoiipas eao^jnsqng a|qn|os |D]4!4Sj34U| 3IUD6JO a|qn|os 3IU0] a|qn|OS ■52 .t:_o oi o tn J3 E C --E u C o D) g 0) Hi u (D-rj (U o n n (rt (U rt -n • — (T) ■— ^ n IJ o > J3 1- V -D _0 0) D n I/) to CO to z a £ -;v (1) >. S -c S "o a i i o 'g- L E u 51-2 - °>:£ 2 §• i: c ^ >- O dJ 0) _c (/> CQ tu Q- tj -4- ^ ■ -ij 5 s o ^ QJ M- (D E i 2 P ^ p E^o,cS E i_o^ o a .2 o -O o ■- M (1) T) E a a E E ^ (U O t- 1- o -n ^ ■*- x5^cQOO 964 WOLFE and RICE; CYCLING OF ELEMENTS IN ESTUARIES 5 S t. "^ a> 5 « i! -- p. ^ = — 09 E O "^ 0) D — -D IT o u _J < < a S h! D a> E r E o o a Q tj y> 2 I I o 3 2 :£ o ^-c □ □ r 6 O T) (Wolfe, 1971) ; whereas *^^Zn may be irreversibly adsorbed onto estuarine sediments (Johnson, Cutshall, and Osterberg-, 1967) . The actual flux of any given element throug-h the complex inter- woven network of pathways tentatively identi- fied in Table 2 is in a state of dynamic equilibri- um afl^ected by system imports and exports and by a number of environmental variables, to be discussed in the following sections. INPUTS AND OUTPUTS OF ELEMENTS TO AND FROM ESTUARIES System inputs and outputs consist of the phys- ical translocation of elemental reservoirs (sedi- ment, water, and biotic components) across the boundaries of the estuarine system (Table 2), and fluxes of other materials (nonreservoirs) which influence the distribution or movement of elements among the system reservoirs. Rainfall is a dominating system input, which introduces small amounts of some elements directly as aero- sols (Gorham, 1961), but exerts its greatest in- fluence in the form of runoff". The volume and flow rate of runoff in turn determine the amount and composition of elements leached from the land and the size of the bed load of eroded sedi- ment transported into the estuary. The volume and flow of runoff" also interact with the tidal volume to affect the flushing rate for the estuary. Tidal flushing continuously removes from the estuary a fraction of the dissolved and suspended materials, including the planktonic reservoirs. The tidal volume also represents an input of ocean water, but since seawater contains less of most metallic elements than estuarine waters, net loss from the estuary probably results. If, on the other hand, the coastal or estuarine waters contained an excess of uncomplexed organic ma- terial, seawater might represent an input of sol- uble ionic metallic elements which could accumu- late in the estuary through sedimentation, or be removed from the estuary as organic complexes, either by flushing or bioaccumulation. Variable wind speed and direction affect water circulation patterns particularly in shallow estu- aries and thereby vary the "normal" flushing characteristics of an estuary. Circulation within 965 FISHERY BULLETIN: VOL, 70, NO. 3 the estuary obviously determines also the physi- cal distribution of certain reservoirs within the system, and this aspect of cycling compounds the difficulties of developing a realistic model. Application of the term "ecosystem" to an estu- ary, however, implies that an estuary behaves as a discrete, identifiable unit and, despite the clumped distributional patterns exhibited by or- ganisms in the environment, spatial distribution of reservoirs within the system is assumed insig- nificant; i.e., the system model is homogeneous. In addition to tidal flushing of planktonic or- ganisms, biotic inputs and outputs include the contribution of terrestrial primary productivity, brought into estuaries as dissolved organic ma- terial and particulate organic detritus in runoff, and the seasonal migrations of larger inverte- brates and fish into and from estuarine waters. A final input which must be considered con- sists of human waste materials. Organic waste materials affect productivity and biological spe- cies structure of the ecosystem and probably also aflfect the availability of metallic elements through organic complexing. In addition, large amounts of certain metallic elements are intro- duced into many estuaries directly in industrial and municipal effluents. FACTORS INFLUENCING DISTRIBUTIONS AND TRANSFERS OF ELEMENTS AMONG THE VARIOUS RESERVOIRS IN ESTUARIES The various groups of biota, the dissolved and particulate components in the water column, and sediment compartments, represent the major reservoirs per se for metallic elements, but the interaction of these compartments and regula- tion of their sizes are influenced greatly by other variable characteristics or components of estu- arine ecosystems including inputs of energy and auxiliary materials or factors which aflfect res- ervoir sizes or transfer rates within the system. These factors represent "nondynamic state vari- ables" in the modeling terminology of Kowal (1971). These characteristics may be categor- ized as: (1) those that affect productivity in the trophic web and (2) those that afl'ect the physical state or rates of exchange between com- ponents for the specific elements of interest. The influences of major nondynamic state variables are summarized briefly in Table 3. Solar energy is perhaps the best example of the first type of input. Solar energy drives the primary productivity supporting all the biotic reservoirs and establishes the basic temperature regime for the entire system, which is a dominant factor influencing the rate of material transfer processes, biological and nonbiological. The sea- sonal variability of temperature can readily be incorporated into systems models according to a sinusoidal function. This general approach has been described and applied by Williams (1969). In the first category also are dissolved gases, e.g., CO2 and O2, and inorganic nutrients, espe- cially the various forms of P and N. Certain dissolved organics may also be included here, although absolute requirements for these have not been demonstrated in the nutrition of estu- arine organisms, and dissolved organic-metal complexes constitute a major reservoir already considered. Acidity, or pH, is an important fact- or in that it acts with temperature to control CO2 solubility and thereby aflTects primary produc- tivity, but pH also falls into the second category because it determines the equilibrium distribu- tion of metals in adsorption reactions, i.e., ex- change between water and sediments, or between water and biological surfaces; and in chelation reactions, including metal-organic complexes in the dissolved phase. Wind is another significant energy input in shallow estuaries. Wind increases turbulence and suspension of particulate matter which pro- motes exchange of elements between water and sediments and affects the size of the suspended particulate reservoir subject to flushing. Salinity is also a determinant in the distribu- tion and cycling of elements. Salinity, as a mea- sure of ionic strength, affects the adsorption equilibria established between sediments and water and biological surfaces and water, and probably also influences the configuration of pro- teins and other biological polymers which form complexes with metals. This mechanism may be the basis of salinity eflfects on active transport processes in ionic and osmotic regulation in es- 966 WOLFE and RICE: CYCLING OF ELEMENTS IN ESTUARIES Table 3. — Role of nondynamic state variables. Factor Compartment and sign (+ or — ) ( Df correlation Fe, Al, Mn, etc. Suspended particulate + Nutrients (nitrate. , nitrite. All primary producers and microbial ammonia, phos phate, and heterotrophs + silicate) Solar insolation All primary producers + Temperature Exchange rotes Adsorption Respiration + + Season of year Inputs and outputs via mi grot ions ± Wind speed and Suspended particulate + wind direction Flushing export rote ^ Dissolved carbon dioxide All primary producers + Dissolved oxygen Respirotion + pH Surface adsorption + Salinity Surface adsorption Ion-exchange rates + Precipitation Inputs of sediment bed-load + Suspended particulate + Dissolved ionic + Dissolved organic + tuarine organisms. In addition, salinity pro- vides a measure of bulk components of seawater, e.g., Ca and K, which may compete directly with trace components for adsorption sites of biologi- cal accumulation, e.g., Sr and Cs. For example, increasing salinity decreases the accumulation rate and concentration factor for '"Cs in estu- arine clams (Wolfe and Coburn, 1970) . Salinity is correlated also with the concentration of '"Cs in lower-salinity estuarine water, possibly as a result of mass action on the sediment-water ex- change equilibrium (Wolfe, 1971). Important dissolved trace components include Fe, Al, and Mn, whose concentrations depend less on salinity than on watershed characteristics and runoff. At the pH, Eh, and ionic strength characteristic of seawater, these elements form insoluble hy- droxides which flocculate and provide adsorptive surfaces for other elements. Such a coprecipi- tation process may be an important determinant in the distribution of fallout '""Ru in estuaries (Wolfe and Jennings, in press). For any given estuary the chemical environ- ment and flushing characteristics are affected greatly by the geological and chemical charac- teristics of the watershed area, the type of es- tuary (stratified vs. well-mixed, etc.), and the physical size, shape, and orientation of the es- tuarine basin. Such characteristics must be con- sidered when different estuaries are compared. A set of biological processes accompanies each trophic interaction and determines the efficien- cies for transfers of biomass between compo- nents. These are feeding or ingestion rate, di- gestion rate, assimilation efficiency, respiration, and growth efficiency. Each of these character- istics may be influenced in turn by various en- vironmental factors, such as temperature and sa- linity (Peters and Boyd, in press; Peters and Angelovic, in press). Assimilation and respiration maintain the bio- mass of the biological reservoirs for elements, but the accumulation and retention of most me- tallic elements are probably not directly correl- ated with the assimilation and respiration of carbon. For dynamic modeling of the cycling of metallic elements in the estuarine biota, how- ever, it is essential to consider the changes in biomass as well as changes in concentration of the element of interest. The interrelation of •^^Zn-excretion and respiration has been examined directly for pinfish, Lagodon rhomboides, (Hoss, 1971). Lowman et al. (1971) computed assimi- lation efficiencies ("conversion factors") for the transfer of several metallic elements from ocean- ic phytoplankton to zooplankton, based upon 967 FISHERY BULLETIN: VOL. 70, NO. 3 relative concentration factors in zoo- and pliyto- plankton and a mean carbon assimilation of 50^/r . On this basis, assimilations for 24 elements ranged from 1 to 85 ''r — indicating nonparallel- ism with carbon assimilation — but Lowman went on to say: "The major weakness in this method of calculating conversion factors for a variety of elements is the uncertainty of the accuracy of the concentration factors for phytoplankton and zooplankton." Similar uncertainty exists for published concentration factors for various elements in estuarine organisms, especially since environmental conditions (and therefore ele- mental concentrations in estuarine water) are subject to such wide variation. A further com- plication is the uncertainty concerning the pro- portion of total elemental intake represented by food sources. According to Polikarpov (1966), marine animals satisfy their requirements for most elements by direct absorption from the sur- rounding water. Considerable experimental evi- dence, however, supports the importance of food as a source of elements for many organisms (Rice, 1963; Hoss, 1964; Baptist and Lewis, 1969) . Atoms adsorbed directly from the water onto body surfaces, whether internal or external, do contribute to the concentration factor but have no relevance to assimilation of food. Sur- face adsorption frequently results in assimila- tion, however, as in phytoplankton (Goldberg, 1952) and the mantle epithelium of Pelecypoda (Nakahara and Bevelander, 1967). For organ- isms with well-defined and easily analyzed in- ternal tissues, e.g., crustacean and fish muscle, internal concentrations of elements probably rep- resent the assimilated fraction, but for many smaller or less differentiated organisms, internal tissues cannot readily be separated from adsorp- tive surfaces. Assimilation efficiency may de- pend also on biochemical composition of the food — at least at certain trophic levels. For example, the assimilation of ^^Zn by human subjects was 35 "^r from a diet of whitefish (Honstead and Brady, 1967) and 13. 5''/^ from oysters (Honstead and Hildebrandt, 1967) , showing a high (though perhaps coincidental) positive correlation with the protein content of the foodstuff (Wolfe and Rice, 1968). The concentration of an element in represent- atives of a population of organisms is a function of the turnover time for the element and the av- erage life span of the organism. Long-lived or- ganisms probably achieve a steady state for the turnover of most elements after the cessation of growth — and if the availability of the element from the organism's environment is stable. In organisms with rapid growth and high popula- tion turnover, net accumulation probably pro- ceeds for most metallic elements throughout the entire short life span of the organism, and steady state is not reached before the organism is con- sumed by the next trophic level. The environ- mental and physiological factors determining the steady state conditions are not known, however. Many organisms may accumulate metallic ele- ments far in excess of their biological require- ments (Wolfe, 1970b; Pequegnat, Fowler, and Small, 1969), and accumulation of metals may continue independent of the biological necessity in some cases until available reaction sites (e.g., between metal and proteins or tissue surfaces) are saturated. This process is suggested also by the increasing concentration of mercury with age (or size) in various fish (Westo, 1969; Bache, Gutenmann, and Lisk, 1971). In a dynamic estuarine system, where envi- ronmental levels of metallic elements are subject to rapid fluctuations, the organismic response to environmental change must be identified. Infor- mation of this nature is sorely lacking in the lit- erature. Pringle et al. (1968) tested the re- sponse of oysters to various increased experi- mental levels of lead, and after 49 days exposure, accumulation had proceeded in direct relation to availability of lead in the environment. Other data (Chipman, Rice, and Price, 1958; Preston, 1967; Wolfe, 1970a) suggest that concentration factors for Zn in oysters are inversely related to zinc content of water, implying that net ac- cumulation would diminish or cease at some high environmental concentration (low concentration factor in oyster) and steady state would be esta- blished. In these cases, however, the variability of instantaneous uptake of the element can only be inferred from the amounts contained after a long period of accumulation. In natural ecosys- tems, fallout radioisotopes appear in estuarine organisms very quickly after initial entry of the 968 WOLFE and RICE: CYCLING OF ELEMENTS IN ESTUARIES isotopes into the ecosystem (Wolfe and Schelske, 1969; Wolfe and Jennings, in press). A high rate of instantaneous uptake has also been dem- onstrated for many organisms and many ele- ments in experiments on radioisotope accumula- tion, but, in most cases, concentrations of stable element counterparts for the radioisotopes were undetermined so that rates of flux for the stable element could not be computed from the observed flux of radioactivity. Although the accumulation and turnover of a radioisotope in a single com- ponent can be modeled mathematically inde- pendent of the stable element chemistry (Reichle, Dunaway, and Nelson, 1970), it is the flux of stable elements which determines the movement of radionuclides among the various components of an ecosystem and investigators should rou- tinely collect data on the stable element compo- sition of the compartments involved in their ac- cumulation studies. In this way, experimentally observed radioisotopic accumulation rates can be used in conjunction with the specific activity, i.e., the concentration ratio of radioisotope to total element, to determine rates of elemental turn- over. It seems probable that instantaneous up- take of an element is a direct function of avail- able environmental levels whereas instantaneous loss is a direct function of accumulated amounts. The instantaneous uptake rate is also a function of other environmental variables. For example, in the estuarine clam Rangia cuneata, the instan- taneous uptake rate and the equilibrium concen- tration of ^"Cs increase with temperature and decrease with salinity (Wolfe and Coburn, 1970). Salinity, temperature, pH, and total Zn also influenced the accumulation of ^^Zn by var- ious estuarine organisms under experimental conditions (Duke et al., 1969). Net accumulation (or net loss) results when instantaneous uptake exceeds (or is less than) instantaneous loss, and the physiology and me- tabolism of the organism determine the residence times required for passage through its many al- ternate internal compartments and pathways. Retention times are usually discussed in terms of biological half-life. ( See for example Baptist, Hoss, and Lewis, 1970.) Since organisms have several compartments simultaneously interact- ing with the environment, retention typically consists of several components with diff'erent rates. One might expect the faster rates to be associated with surface adsorption reactions, in- termediate rates with excretion of unassimilated material as feces, and slow rates with the turn- over of the assimilated and metabolized fraction of the elemental content. Although the relative amounts of an accumulated radioisotope involved with different retention components can be de- termined for the particular conditions and time period of accumulation and loss used in the ex- periment, these amounts will not be represent- ative of stable element pools unless all of the in- ternal compartments are equally labeled, i.e., to a uniform specific activity. Thus, long-term ac- cumulation experiments under conditions of constant specific activity are required (Cross, Willis, and Baptist, 1971). The individual re- tention components for an element probably will have to be considered for each important reser- voir in modeling the overall flux of that element in the ecosystem. We have discussed several aspects of elemental cycling in estuaries and have demonstrated the incompleteness of man's knowledge of how an estuary operates as a system with many integral, smoothly functioning components. We believe, however, that many of the unsolved problems which have presented themselves will be realist- ically resolved only by a holistic approach to eco- logical research. The foregoing discussion rep- resents an eff"ort to conceptualize the elemental cycling system that operates in our southeastern coastal zone estuaries. Prior recognition of the complexity and integrity of the system as a whole provides an improved basis for planning mean- ingful research on the transfer processes be- tween individual components of an estuary. Con- siderable research is required before we can act- ually quantify the reservoirs, routes, and rates of elemental flux involved in this preliminary model of these ecosystems. The complexity of the ecosystem defies precise quantification of all the reservoirs and all the transfer rates under any set of environmental conditions. In such a system, however, the predictability of the mag- nitude and variability of any elemental reservoir depends upon recognition of all the interactions impinging upon that reservoir. Research now 969 FISHERY BULLETIN: VOL. 70, NO. 3 in progress at the Atlantic Estuarine Fisheries Center will enable us to estimate the size ranges for most reservoirs of manganese, iron, and zinc in our local estuarine system. The major gaps in our present knowledge lie in: 1. Determining the relative amounts of dif- ferent physico-chemical forms of an element or radioisotope in natural waters, their relative sta- bilities, and the ease of interconversion between the various forms. 2. Determining the relative biological availa- bilities of these different physico-chemical forms to various types of biota. 3. Determining trophic structure of the entire ecosystem. The , xo\e of microorganisms — as sources of metallic elements to consumers in detritus-based food chains, as producers of or- ganic-metal complexes, and as remineralizers of metals previously incorporated into plant or animal tissues — is particularly poorly under- stood. 4. Determining feeding rates and assimilation efficiencies for carbon and metallic elements at each major trophic interaction. 5. Determining biological retention of metallic elements in the major organisms consumed by man. 6. Determining the interactions of variable environmental parameters on reservoir size and transfer rates at each step in the overall system. As further information becomes available, this preliminary systems model will be refined and tested as to its adequacy for describing the flux of manganese, iron, and zinc in our local estua- arine system. Maintenance of this sort of ho- listic viewpoint toward ecological function and continuous updating of existing conceptual mo- dels will provide the most reliable basis for ra- tional management of man's releases of toxic heavy metals and radionuclides — or indeed, of any contaminant additions to the environment. LITERATURE CITED Bache, C. a., W. H. Gutenmann, and D. J. Lisk. 1971. Residues of total mercury and methylmerc- uric salts in lake trout as a function of age. Science (Wash., D.C.) 172:951-952. Baptist, J. P., and C. W. Lewis. 1969. Transfer of 65Zn and siCr through an estu- arine food chain. In D. J. Nelson and F. C. Evans (editors). Symposium on Radioecology, p. 420-430. USAEC CONF-670503. Oak Ridge, Tenn. Baptist, J. P., D. E. Hoss, and C. W. Lewis. 1970. Retention of siCr, s^Fe, soCo, ssZn, s^Sr, 95Nb, 147 mjn and ^''I by the Atlantic croaker (Microp- ogon undulatus). Health Phys. 18:141-148. Barber, R. T., and J. H. Ryther. 1969. Organic chelators: Factors affecting pri- mary production in the Cromwell Current upwel- ling. J. Exp. Mar. Biol. Ecol. 3:191-199. Chipman, W. a., T. R. Rice, and T. J. Price. 1958. Uptake and accumulation of radioactive zinc by marine plankton, fish, and shellfish. U.S. Fish Wildl. Serv., Fish. Bull. 58:279-292. Cross, F. A., and J. H. Brooks. In press. Concentrations of Mn, Fe, and Zn in ju- veniles of five estuarine-dependent fishes. In D. J. Nelson (editor). Third National Symposium on Radioecology. USAEC Oak Ridge, Tenn. Cross, F. A., T. W. Duke, and J. N. Willis. 1970. Biogeochemistry of trace elements in a coast- al plain estuary: Distribution of manganese, iron, and zinc in sediments, water, and polychaetous worms. Chesapeake Sci. 11:221-234. Cross, F. A., J. N. Willis, and J. P. Baptist. 1971. Distribution of radioactive and stable zinc in an experimental marine ecosystem. J. Fish. Res. Board Can. 28:1783-1788. Duke, T. W., J. N. Willis, and T. J. Price. 1966. Cycling of trace elements in the estuarine environment. I. Movement and distribution of zinc 65 and stable zinc in experimental ponds. Chesapeake Sci. 7:1-10. Duke, T. W., J. N. Willis, T. J. Price, and K. Fischler. 1969. Influence of environmental factors on the concentrations of •'^Zn by an experimental com- munity. In D. J. Nelson and F. C. Evans (edi- tors). Symposium on Radioecology, p. 355-362. USAEC CONF-670503. Oak Ridge, Tenn. Duke, T. W., J. N. Willis, and D. A. Wolfe. 1968. A technique for studying the exchange of trace elements between estuarine sediments and water. Limnol. Oceanogr. 13:541-545. Epstein, E. 1965. Mineral metabolism. In J. Bonner and J. E. Varner (editors), Plant biochemistry, p. 438-466. Academic Press, N.Y, Goldberg, E. D. 1952. Iron assimilation by marine diatoms. Biol. Bull. (Woods Hole) 102:243-248. 1963. The oceans as a chemical system. In M. N. Hill (editor). The sea: Ideas and observations on progress in study of the seas, Vol. 2, p. 3-25. Interscience Publ., N.Y. 970 WOLFE and RICE: CYCLING OF ELEMENTS IN ESTUARIES Gordon, D. C, Jr. 1966. The effects of the deposit feeding polychaete Pectinaria gouldii on the intertidal sediments of Barnstable Harbor. Limnol. Oceanogr. 11:327- 332. GORHAM, E. 1961. Factors influencing supply of major ions to inland waters, with special reference to the at- mosphere. Geol. Soc. Am. Bull. 72:795-840. HONSTEAD, J. F., AND D. N. BRADY. 1967. The uptake and retention of ^^p and ^^Zn from the consumption of Columbia River fish. Health Phys. 13:455-463. HONSTEAD, J. F., AND P. W. HiLDEBRANDT. 1967. Uptake and retention of zinc-65 from certain foods. Health Phys. 13:649-652. Hoss, D. E. 1964. Accumulation of zinc-65 by flounder of the genus Paralichthys. Trans. Am. Fish. Soc. 93: 364-368. 1971. Routine energy requirements of a population of pinfish (Lagodon rhomboides Linnaeus) in the Newport River Estuary, North Carolina. Ph.D. Thesis, North Carolina State Univ., 78 p. Johnson, V., N. H. Cutshall, and C. L. Osterberg. 1967. Retention of ^^Zn by Columbia River sedi- ment. Water Resour. Res. 3:99-102. Joseph, A. B., P. F. Gustafson, I. R. Russell, E. A. SCHUERT, H. L. VOLCHOK, AND A. TAMPLIN. 1971. Sources of radioactivity and their character- istic. In Radioactivity in the marine environment, p. 6-41. Nat. Acad. Sci., Wash., D.C. Kowal, N. E. 1971. A rationale for modeling dynamic ecological systems. In B. G. Patten (editor), Systems anal- ysis and simulation in ecology. Vol. 1, p. 123-194. Academic Press, N.Y. LowMAN, F. G., D. K. Phelps, R. McClin, V. Roman de Vega, I. Oliver de Padovani, and R. J. Garcia. 1966. Interactions of the environmental and bio- logical factors on the distribution of trace elements in the marine environment. In Disposal of radio- active wastes into seas, oceans and surface waters, p. 249-266. Int. At. Energy Agency, Vienna. Lowman, F. G., T. R. Rice, and F. A. Richards. 1971. Accumulation and redistribution of radio- nuclides by marine organisms. In Radioactivity in the marine environment, p. 161-199. Nat. Acad. Sci., Wash., D.C. Lowman, F. G., and R. Y. Ting. In press. Turnover of cobalamin-^'Co and ionic 58Co in the clam Donax denticulatus Linne. In D. J. Nelson (editor) , Third National Symposium on Radioecology. USAEC Oak Ridge, Tenn. Mason, B. 1958. Principles of geochemistry. 2d ed. Wiley, N.Y., 310 p. Nakahara, H., and G. Bevelander. 1967. Ingestion of particulate matter by the outer surface cells of the mollusc mantle. J. Morphol. 122:139-146. Pequegnat, J. E., S. W. Fowler, and L. F. Small. 1969. Estimates of the zinc requirements of ma- rine organisms. J. Fish. Res. Board Can. 26:145- 150. Peters, D. S., and J. W. Angelovic. In press. The effect of temperature, salinity and food availability on growth and energy utiliza- tion of juvenile summer flounder, Paralichthys dentatns. In D. J. Nelson (editor). Third Na- tional Symposium on Radioecology. USAEC Oak Ridge, Tenn. Peters, D. S., and M. T. Bo1(T). In press. The effect of temperature, salinity, and food availability on the feeding and growth of the hogchoker, Trinectes maculatiis. J. Exp. Mar. Biol. Ecol. POLIKARPOV, G. G. 1966. Radioecology of aquatic organisms: The ac- cumulation and biological effect of radioactive sub- stances. (Translated from the Russian by Scripta Technica, Ltd.) Reinhold, N.Y., 314 p. PoMEROY, L. R., R. E. Johannes, E. P. Odum, and B. Roffman. 1969. The phosphorus and zinc cycles and produc- tivity of a salt marsh. In D. J. Nelson and F. C. Evans (editors). Symposium on Radioecology, p. 412-419. USAEC CONF-670503. Oak Ridge, Tenn. Preston, A. 1967. The concentration of ^sZn in the flesh of oys- ters related to the discharge of cooling pond ef- fluent from the C.E.G.B. Nuclear Power Station at Bradwell-on-Sea, Essex. In B. Aberg and F. P. Hungate (editors), Radioecological concentration processes, p. 995-1004. Pergamon Press, Oxford. Pringle, B. H., D. E. Hissong, E. L. Katz, and s. t. mulaw^ka. 1968. Trace metal accumulation by estuarine mol- lusks. J. Sanit. Eng. Div., Proc. Am. Soc. Civil Eng. 94:455-475. Reichle, D. E., p. B. Dunaway, and D. J. Nelson. 1970. Turnover and concentration of radionuclides in food chains. Nucl. Saf. 11:43-55. Rhoads, D. C. 1967. Biogenic reworking of intertidal and subtidal sediments in Barnstable Harbor and Buzzards Bay, Massachusetts. J. Geol. 75:461-476. Rhoads, D. C, and D. K. Young. 1971. Animal-sediment relations in Cape Cod Bay, Massachusetts II. Reworking by Molpadia oolitica (Holothuroidea). Mar. Biol. (Berl.) 11:255-261. Rice, T. R. 1963. Review of zinc in ecology. In V. Schultz and A. W. Klement, Jr. (editors), Radioecology, p. 619-631. Reinhold, N.Y. 1965. Annual report of the Bureau of Commercial Fisheries Radiobiological Laboratory, for the fiscal 971 FISHERY BLLLETIN: VOL. 70, NO. 3 year ending June 30, 1963. U.S. Fish Wildl. Serv., Circ. 204, 44 p. Rice, T. R., and D. A. Wolfe. 1971. Radioactivity — chemical and biological as- pects. In D. W. Hood (editor), Impingement of man on the oceans, p. 325-379. Wiley-Interscience, N.Y. Riley, J. P., and R. Chester. 1971. Introduction to marine chemistry. Academic Press, N.Y., 465 p. Underwood, E. J. 1962. Trace elements in human and animal nutri- tion. 2d ed. Academic Press, N.Y., 430 p. Westo, G. 1969. Mercury in fish. Fish. Res. Board Can., Transl. Ser. No. 1350, 8 p. Williams, R. B. 1969. A table of mean effective temperatures for the metabolism of biological systems subjected to sinusoidal cycles in temperature. J. Theoret. Biol. 24:240-245. Wolfe, D. A. 1970a. Levels of stable Zn and ""Zn in Crassostrea virginica from North Carolina. J. Fish. Res. Board Can. 27:47-57. 1970b. Zinc enzymes in Crasi^ostrca virginica. J. Fish. Res. Board Can. 27:59-69. 1971. Fallout cesium-137 in clams (Rangia cuneata) from the Neuse River estuary. North Carolina. Limnol. Oceanogr. 16:797-805. Wolfe, D. A., and C. B. Coburn, Jr. 1970. Influence of salinity and temperature on the accumulation of cesium-137 by an estuarine clam under laboratory conditions. Health Phys. 18:499- 505. Wolfe, D. A., and C. D. Jennings. In press. Iron-55 and ruthenium-103 and -106 in the brackish-water clam, Rangia cuneata. In D. J. Nelson (editor). Third National Symposium on Radioecology. USAEC Oak Ridge, Tenn. Wolfe, D. A., and T. R. Rice. 1968. Safe levels of radioactivity in aquatic envi- ronments. Scientia 103:469-487. Wolfe, D. A., and C. L. Schelske. 1969. Accumulation of fallout radioisotopes by bi- valve molluscs from the lower Trent and Neuse Rivers. In D. J. Nelson and F. C. Evans (editors), Symposium on Radioecology, p. 493-504. USAEC CONF-670503. Oak Ridge, Tenn. 972 A LABORATORY STUDY OF PARTICULATE AND FILTER FEEDING OF THE PACIFIC MACKEREL, SCOMBER JAPONICUS Charles P. O'Connell and James R. Zweifel^ ABSTRACT In laboratory feeding trials Pacific mackerel, Scomber japonicns, averaging 147 g in weight did not respond to Artemia nauplii, but did capture Artemia adults by biting (particulate feeding) when density was 1 or 2/liter and by filtering when density was 22 to 112/liter. Particulate feeding is described by Np = 60.3f/?, where N is the number of Artemia ingested in t minutes at D numbers per liter. Filter feeding is described by Nf- = 23,788 (1 — e-o.ooastvD)^ where 23,788 is an asymptotic estimate of the number of Artemia in the digestive tract at full capacity. The results suggest that the mackerel utilizes only the larger of the planktonic crusta- ceans, such as euphausiids, in the sea. For the relatively low average densities of such organisms the derived equations indicate that the mackerel could not obtain its daily nutritional requirement, estimated to be 8% of body weight, in less than 24 hr of feeding. Though the daily requirement could be obtained in much shorter periods, perhaps by filter feeding, if such crustaceans are encountered in aggregations of considerably higher density than reflected by area averages, it is probable that the mackerel must often de- pend in part on such larger organisms as fish to fulfill its needs. Comparison of the mackerel ingestion rates to those for the smaller northern anchovy indicates that while the individual mackerel may generally capture a greater proportion of the large crustaceans encountered than the anchovy, the proportion captured would have a relatively lower nutritional value for the mackerel. The Pacific mackerel, Scomber japonicus, is one of several pelagic schooling fishes of the eastern temperate Pacific which feed on zooplankton, but it does not depend entirely on zooplankton. Fitch (1956) reported that stomach contents contained about 30*;/ larval and juvenile fish by volume, with the remainder composed largely of such crustaceans as mysids, copepods, and euphau- siids. Frey (1971) commented that larval and juvenile fish appear to be the most important food, but that the mackerel relies heavily on euphausiids at times. Hatanaka et al. (1957) showed that S. japonicus in coastal regions of Japan consume mainly small anchovies in the late summer and autumn and euphausiids in other seasons. The biomass of euphausiids con- ' National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Southwest Fisheries Center, La Jolla, CA 92037. sumed annually was estimated to be three or four times that of anchovies. In addition to being the dominant element in the diet of S. japonicus, the larger crustaceans are an important class of food for a number of other pelagic schooling fishes. The jack mack- erel {Trachurus syynynetricus) is known to feed heavily on small fish and squid at times (Fitch, 1965), but about 10 ''/c of stomach contents by volume is euphausiids (Carlisle, 1971) . The Pa- cific sardine (Sardinops caerulea) feeds largely on copepods smaller than those consumed by the jack mackerel but is occasionally gorged on euphausiids, and these average about 5% of stomach contents by volume (Carlisle, 1971). The northern anchovy {Engraulis mordax) con- sumes phytoplankters and small zooplankters, but large copepods and euphausiids appear to be the most important food items (Loukashkin, 1970). Thus while all of these species prey on Manuscript accepted April 1972. FISHERY BULLETIN: VOL. 70, NO. 3, 1972. 973 FISHERY BULLETIN: VOL. 70. NO. 3 large crustaceans, the latter do not necessarily make an equivalent contribution to their diets. It is probable that the segment of the biota rep- resented by such crustaceans in the sea is a food source of different potential for different fishes. Assessing the food potential of large crusta- ceans for the Pacific mackerel, or any of the above teleosts, requires a knowledge of the den- sity levels of such crustaceans in the sea and of the rates at which the organisms can be cap- tured in relation to density. Leong and O'Con- nell (1969) determined by a laboratory study the rates at which the northern anchovy feeds on Artemia nauplii by filtering and on Artemia adults by biting (particulate feeding), and O'Connell (1972) showed that the two kinds of feeding activity in a small school varied with the relative abundance of the two sizes of Ar- temia. The present study was undertaken to obtain comparable information for the Pacific mackerel, using Artemia as the food. The re- sults are discussed in respect to the densities of large crustaceans in the sea and in respect to the feeding rates of the anchovy. METHODS Feeding trials were carried out in an arrange- ment of two plastic pools with a connecting trough and gates (Leong and O'Connell, 1969; O'Connell, 1972). The pools were supplied with a continuous flow of filtered sea water and each contained 4.5 m^ of water at a depth of 0.61 m (2 ft). They were under a 12-hr day, 12-hr night cycle of illumination, but all trials were carried out during the day. Temperature varied from 16° to 19°C. A school of about 160 mackerel was acclima- tized in the pool arrangement with the gates open for about 2 weeks and then confined to one pool before the study started. The fish averaged 222 mm (202 to 247 mm) in standard length and 152 g (100 to 225 g) in weight. The latter average is based on weights taken at the end of each feeding trial and included digestive tract contents, which were later found to range from 0.1 to 21.8 g. These values were subtracted from the measured fish weights to obtain estimates of the weights before feeding. The subsequent analysis involves the adjusted weights, which averaged 147.3 g. All feeding trials were preceded by at least 24 hr without feeding, which was enough to pro- duce virtually empty digestive tracts in prelimi- nary investigations. For each trial the water flow was turned off, and four fish were diverted from the holding pool to the trough and then admitted to the prepared food situation in the other pool where they were allowed to feed for a given number of minutes. Closing of the gate after admission initiated the time period and in- troduction of a hinged crowder, which was rap- idly closed to trap and remove the fish, terminat- ed the time period. The fish were killed in less than 2 min after introduction of the crowder, immediately measured and weighed, then placed in jars of lO^/r Formalin with the body cavity opened. Though the study is concerned primarily with feeding on Artemia adults, a few feeding trials were carried out with newly hatched Arteinia nauplii, which averaged 0.65 mm in length, to document the response of the mackerel to very small crustaceans. The densities in the water were estimated from subsamples, and quantities in the digestive tracts were estimated by total counts or volumetric aliquoting, as described by Leong and O'Connell (1969). Feeding on Artemia adults is based on digest- ive tract contents of four fish in each of 28 trials, with each trial representing a different combina- tion of food density and feeding time (Table 1). The food situation was established for each trial by placing a given wet weight of Artemia adults in the experimental pool and dispersing it gently with wide-mesh dip nets just before admitting the fish. The weights given under food quantity are the wet weights introduced at the start of each trial. These were selected after prelimi- nary observation to include the highest and the lowest levels that could be managed conveniently. The multiplication factors given in certain cells of the table indicate the number of times the specified weight was introduced into the pool dur- ing the trial, e.g., 20 g was introduced six times, or at 15-min intervals, during the 90-min trial. This was done to sustain the nominal food den- 974 O'CONNELL and ZWEIFEL: FEEDING OF PACIFIC MACKEREL Table 1. — The array of trials with respect to food density and feeding duration. Number/liter is the nominal den- sity for the weight introduced. Quantity of Artemia Minutes of feeding Total weight Number liter 2.5 5 10 30 60 90 g 10 20 200 590 1,010 1 2 22 66 112 1 I 1(3X)' 1(6X) 1 1 I(2X) 1(4X) 1111 1111 111-. 1(9X) 1(6X) 1 1 ' The multiplication factors show the number of times, at equal inter- vals, that the specified weight was introduced during the trial to maintain tha nominal density. See text. sity over long periods at the lower food density- levels. Introduction of food during the trial, and even brief dispersion with the wide-mesh net, did not appear to distress the fish and scarce- ly interrupted their feeding activity. The intervals of food introduction for the low- er density levels were determined from a prelimi- nary study. During a 30-min trial with 20 g of Artemia, e.g., feeding activity declined sharply after 15 min, and average digestive tract con- tents did not differ from that for a trial of 15 min. A trial of 30 min with two introductions of 20 g showed approximately double the digest- ive tract contents of the above two. The 10-min interval for 10 g introductions was established from similar considerations. Food availability remained high throughout the trials at the three higher density levels. The nominal densities of Artemia shown in Table 1 are based on the volume of the pool and a conversion factor of 500 Artemia adults/g wet weight. The conversion factor is an average derived from samples from the several batches of Artemia delivered to the aquarium. Portions of these samples also indicated that individuals averaged 4.5 mm in length and 0.48 mg dry weight. The quantities of A^'temia in the digestive tracts of the fish were estimated separately for the mouth, the esophagus and stomach combined, and the anterior and posterior halves of the in- testine. The contents of the esophagus and stomach were weighed. Numbers in the various sections were estimated by total count where quantity was low, but by counts from aliquots where the quantity was large. These data indi- cated that the weight of ingested Artemia, de- termined 2 to 6 months after preservation, was 25 to 50 9^ lower than the weight of the live organisms. Digestion and leaching by Formalin may have been responsible for such loss. Since the weights of food in the digestive tracts could not be considered representative of the weights of food consumed, the analysis was carried out on the numbers of organisms in digestive tracts. Values for all sections of the digestive tract were pooled to obtain estimates of total digestive tract contents, the primary entity in analysis. The numbers found in the mouth varied from less than 1 to 53% and averaged 18% of total diges- tive tract contents. This percentage, further- more, varied with food density, increasing from 4% at the 10 g level to 41% at the 1,010 g level. It is presumed that the contents of the mouth were accumulations to be swallowed and/or ma- terial regurgitated during capture. Proportions in the intestine were considerably lower and will be described later. RESULTS RESPONSE TO ARTEMIA NAUPLII Three trials were carried out with newly hatched Artemia nauplii as the only available food. In two cases the average densities of nauplii were about 190/liter and 230/liter, and groups of four fish showed no recognizable feed- ing activity during 30-min periods of exposure. In the third case density was a little more than 200/liter and a group of approximately 50 fish showed no feeding activity. The stomachs of five fish from these trials contained no nauplii, al- though a few nauplii were found on the gill rakers of two of them. One trial was carried out with Artemixi adults and nauplii both present in the water. The den- sity of adults was 22/liter and the density of nauplii was 185/liter. After feeding for 5 min, four fish averaged 2,334 Artemia adults and 823 Artemia nauplii (635 to 1,083) in the mouth, stomach, and esophagus combined. The bulk of both sizes of organism was in the stomachs. From these results it is evident that while the Pacific mackerel did not respond to nauplii alone, 975 FISHERY BULLETIN: VOL. 70, NO. 3 it did ingest them along with the larger Artemia when both were present in the water. However, the dry weight of the nauplii ingested would be about 0.1% of the dry weight of the adults in- gested. It appears, also, that the uptake of nauplii was only about 4% as efficient as the uptake of adult Artemia, i.e., the number of nauplii per fish represents 4.4 liters of water in the pool, while the number of adults per fish rep- resents 106 liters. Yasuda (1963) obtained results comparable to the above for S. japonicus 120 to 130 mm in length. He found that the mackerel did not eat brine shrimp (0.38 mm in length and presumably nauplii) as did anchovies {Engraulis japonica) and even horse mackerel {Trachurus japonicus) of approximately the same length. The spacing between gill rakers and gill raker processes was shown to be smaller in the latter two species than in the Pacific mackerel. RESPONSE TO ARTEMIA ADULTS The 28 trials in which Artemia adults were the only available food showed that feeding was particulate at the two lower density levels and filtering at the three higher density levels. Par- ticulate feeding is the capture of individual or- ganisms by directed biting. Filter feeding is the process of straining organisms from the water as it passes through the gill rakers while the mouth remains open. The duration of mouth opening was 1 to 3 sec, and such mouth openings occurred almost rhythmically 15 to 20 times/min. Filtering, however, was sustained for only a lim- ited time, and this time varied inversely with density level of the food. It lasted 30 min at 22 Artemia /Mier, 20 min at 66 Artemia /Mier, and about 15 min at 112 Artemia /Wier. At these times the fish noticeably reduced swimming speed and shifted to particulate feeding. Though par- ticulate feeding appeared to be less vigorous under these circumstances than at the two lowest food densities, complete cessation of feeding did not occur in any of the trials. The trial groups of four fish evidenced some discomfort upon encountering the highest food density, 112 Artemia /Mier. Swimming speed and coloration showed less increase than at other densities, and filtering intervals were shorter and less rhythmic. A school of about 70 fish in- troduced to this highest density, on the other hand, exhibited strong rhythmic filtering ac- companied by marked intensification of color and increased swimming speed. Though digestive tract contents from the fish of this group were not significantly greater than for the four-fish group after 5 min of feeding, the larger group virtually eliminated the available food in this time while the smaller group did not noticeably afl^'ect its density. Preliminary analysis indicated that there was no confounding of size of fish with density levels and suggested that the larger fish tended to con- sume slightly more food at all densities. Hence, estimates of total digestive tract quantities were standardized to the average adjusted weight of all fish (147.3 g) by simple proportion. The means and standard errors of the standardized numbers are shown for each trial in Table 2. The greatest change in a trial average resulting from standardization was 20% of the original estimate, and the change was 5% or less for half of the trials. Standardization also aflJ'ected the estimates of variability, but only to the extent that the coefficients of variation averaged 1% higher. A separate analysis for each of the food den- sity levels indicated that digestive tract contents increased proportionately with time at the two lower densities, but increased exponentially to- ward an asymptotic value at the three higher densities. It was also evident that the density- specific rates of increase varied directly with density for the two lower levels and with the square root of density for the three higher levels. The modes of food accumulation for the two den- sity ranges can therefore be expressed as and Np = atD fl ^e-^'^A (1) (2) where Np = number of organisms in the di- gestive tract after t minutes of particulate feeding, 976 OCONNELL and ZWEIFEL: FEEDING OF PACIFIC MACKEREL Table 2. — The average fish weight and the mean (Z) and standard error (SE) of the standardized number of Artemia in the digestive tract for each trial. Density Number/ lifer Feeding duration Number fish Average fish weight Number in digestive tract SE 22 66 112 min 2.5 5 10 30 60 90 2.5 5 10 30 60 90 2.5 5 10 30 60 90 2.5 5 10 30 60 90 2.5 5 10 30 S 125 130 136 134 135 150 149 148 149 123 157 149 136 159 141 135 165 152 159 139 158 174 158 168 145 155 139 147 149 299 572 2,280 3,430 5,620 254 695 1,530 4,636 5,790 11,325 745 2,013 4,399 10,893 15,664 18,079 1,502 4,041 5,232 9,863 12,676 15,723 3,079 3,617 7,629 15,845 60 168 86 373 342 216 37 81 163 583 684 1,204 263 101 270 790 648 1,143 197 409 858 990 484 605 936 447 982 2,600 Nf = number of organisms in the di- gestive tract after t minutes of filter feeding, D = the nominal density of food in numbers per liter, Nx = asymptotic number of organisms in the digestive tract at full ca- pacity, a = specific feeding rate, and f3 = instantaneous feeding rate. Fitting all of the data in the two low density series with equation (1) and those in the three high density series with equation (2) resulted in a satisfactory fit for the low density group but not for the high density group. Calculated values tended to be lower than trial values for the 22 Artemia/liter series, and higher than trial values for the 66 Artemia/liter series, particu- larly for the longer time periods. The difl^culty arises from the fact that average quantities in the digestive tracts were lower for the 66 Ar- temia/\iter level than for the 22 Artemia/liter level in the 30-, 60-, and 90-min trials. The rea- son for this is not known, but examination of the quantities in the anterior and posterior halves of the intestine for all trials (Table 3) offers a plausible explanation. Table 3. — Average number of Artemia in anterior (A) and posterior (P) halves of intestine for each trial. Time Density 1 22 66 112 2.5 A P 5.0 A P 10 30 60 90 A P A P A P A P 0 0 0.3 0 1.8 0 50 0 0.7 0 141 4 15 0 15 0 50 0 51 0 165 45 209 339 9 0 108 0 57 0 320 36 830 720 890 385 33 0 101 0 215 10 1,077 327 569 714 937 722 300 0 175 0 467 59 1,265 245 At 22 Artemia/Uter the maximum quantity in the intestine is reached at 60 min, and at 66 A/"^ew/a/liter the maximum is reached at 30 min. Fluctuations in the two halves of the intestine thereafter suggest posterior movement of ma- terial and intermittent elimination. This is cor- roborated by visual observations made during the trials. The earliest detected defecations were at 50 min for the 22 Artemia/liter level and at about 30 min for the 66 Artemia/liter level. No defecation was detected in the trials at 112 Artemia/liter, which did not go beyond 30 min. The quantities in the intestine suggest that rate of movement of material into the posterior part of the intestine approached the maximum at 66 Artemia/liter and that defecation might not start any sooner at 112 Artemia/liter than at 66 Artemia/liter. From these data it is reasonable to suppose that beyond 30 min the time-specific losses by defecation would be greater for the 66 Artemia/ liter series than for the 22 Artemia/liter series, and negligible for the two lower density levels. On the supposition that the greatest under- estimates of total amounts consumed occurred in the three longest time periods of the 66 Ar- temia/liter series, equation (3) was refitted to 977 FISHERY BULLETIN: VOL. 70, NO. 3 Table 4. — Estimated parameter values and 95% con- fidence limits for density series separately and pooled under particulate and filter feeding. Particulate feeding ArtemiaA't^er Intercept Slope (a) 1 2 Pooled 46.8 (-316; 410) 181.3 (-1,004; 1,367) 0 assumed 61.5 (53.5; 69.5) 58.5 (45.9; 71.1) 60.3 (56.3; 64.3) Filter feeding ArtemiaAi^er Asymptote (A'oo' Instantaneous feeding rate (|3) 22 20,322 (16,736; 23,908) 66 15,384 (13,991; 16,778) 112 22,050 (16,999; 27,102) Pooledi 23,788 (21,802; 25,775) 0.0052 (0.0031; 0.0074) 0.OO45 (0.0035; 0.0060) 0.004O (0.00(29; 0 O059) 0.O036 (0.O032; 0.0042) 1 The pooled array under filter feeding does not include the 30-, 60-, and 90-min trials for 66 Artemia/\\\er. the data for the higher density levels with these three trials removed. The estimated parameters are shown, along with those for the lower density levels, and also for the density levels individually, in Table 4. Since the individual series under particulate feeding did not have intercepts that differed from zero, the parameters for the pooled array were estimated with the intercept assumed to be zero. Goodness of fit for the combined data in both density groups was judged satisfactory; in Figure 1 the calculated curves are compared to the standardized trial means and standard er- rors for each density series. The equations for the two feeding modes can be stated as Np = 60.ZtD (3) and N, 23,788 I -[^ g-o.oosetVD j M) The asjrmptotic level, 23,788 Artemia, is in- dicative of the maximum capacity of mackerels at an average weight of 147 g. On the basis of the wet weight of Artemia, 500 individuals/g, maximum capacity would be 48 g, or 32% of fish weight. This is about double the greatest weight of fish food removed from the esophagus and stomach, but the two kinds of estimate are not necessarily inconsistent. As suggested earlier, the weights of the digestive tract contents may have underestimated the weights of Artemia in the live state by 25 to 50% because of digestion and leaching in Formalin. The estimate given here is based on weight of live organisms. Since digestion is going on during protracted feeding, 112/1. (o) 66/1. (A) 22/1 (•) 2 /I. (A) 1/I.(D) 0 5 10 60 MINUTES Figure 1. — The number of Artemia in digestive tracts for different density levels in the water and for different feeding durations. The lines for 1 and 2 Artemia/Mier were calculated from equation (3) and the lines for 22, 66, and 112 Artemia/Mier were calculated from equation (4). The symbols associated with each line show aver- ages and standard errors for the trials in that density series. Standard errors smaller than 450 are not shown. 978 OCONNELL and ZWEIFEL: FEEDING OF PACIFIC MACKEREL averag-e capacity might well be less than 32% of weight for the asymptotic number specified. The asymptotic number, which was most strongly influenced by the trials at 22 Artemia/ liter, may slightly underestimate the total num- ber that can be ingested because of losses by defecation. DISCUSSION It is evident that Artemia adults approximate the smaller crustacean sizes utilized by the mack- erel and that the fish resorts to filter feeding to increase the rate of consumption as density exceeds some level where particulate feeding becomes relatively inefficient. The ecological meaning of this feeding pattern is indicated by considering the resulting relation between food density and rate of food accumulation in the digestive tracts in respect to 1 ) the daily nutri- tional requirement of the species, 2) the density levels of crustaceans in the sea, and 3) the feed- ing rates of the northern anchovy (Engraulis mordax) . Hatanaka et al. (1957) showed that S. japon- icus in coastal regions of Japan tend to utilize small fish in late summer and autumn but to rely largely on euphausiids and other crustaceans of similar size in other seasons. They concluded that mackerel a little over 1-year-old and aver- aging 149 g in weight required 8% of their body weight per day in crustaceans to sustain the growth rate observed in nature, which was esti- mated as 0.42% of body weight per day. The feeding functions derived in the present study indicate that the times required for the mack- erel to obtain this daily requirement at the A?'- temia densities tested, or their equivalent for euphausiids, would be Artemia/liter 1 2 22 66 112 Euphausiids/liter 0.7 1.3 15 44 75 Minutes 97 49 17 10 7 The equivalent densities for euphausiids are given on the assumption that the mackerel feed- ing rate is keyed to concentration in terms of biomass rather than to numbers per unit volume as such. Euphausiids near the surface at night in the eastern Pacific, largely Euphausia pacifica, average 6 mm in body length (O'Connell, 1971), and individuals of this length are 3 mg wet weight (Lasker, 1966), or 50% more than the A7'temia. The feeding times given above are relatively short, but the Artemia densities are much higher than those reported for comparable organisms in the sea. Brinton (1962) showed average den- sities of E. pacifica, largely juveniles, to be about 0.02/liter near the surface at night oflf southern California. O'Connell (1971) showed an aver- age of 0.03/liter over much the same region, with perhaps 5 to 10% of the area having den- sities approaching 0.1/liter at any one time. These estimates would be elevated, perhaps doubled, if other large crustaceans were added on a biomass equivalent basis. The highest of these area densities would enable the mackerel to ob- tain its daily nutritional requirement in about half a day of particulate feeding, but the more commonly prevailing level would not permit the mackerel to obtain its requirement within the space of a day. In all probability the mackerel obtains much of its needs from euphausiids and other crusta- ceans of similar size, but there is good evidence that it depends to some extent on larger organ- isms, such as fish up to one-third of its own body length (Hatanaka and Takahashi, 1960), to se- cure the daily requirement over a reasonable length of time. Though stomachs tended to be fuller and growth better during the season when the mackerel feeds primarily on fish (Hatanaka et al., 1957) , it must be remembered that the 8% daily feeding requirement used here pertains to maintenance and growth when euphausiids were the primary food. If the feeding rates in- dicated for higher densities of Artemia apply for biomass equivalents of much larger organisms, the large capacity of the mackerel suggests that relatively infrequent encounters with such or- ganisms would sustain the daily requirement in an average sense. The present study and that of Kariya and Takahashi (1969) indicate that feeding can be expected to continue towards full capacity regardless of state of fullness when food becomes available. Though mackerel may have to depend in part on larger organisms, it is possible that the food 979 FISHERY BULLETIN: VOL. 70, NO. 3 potential of the larger crustaceans, and even very young fish, is greater than indicated by average area densities. The definite filtering response of the mackerel, which appears to be an adap- tation for capturing the smaller organisms it utilizes at a greater rate than would be possible by biting activity when density is high, implies that high densities are a factor of some conse- quence in the feeding ecology of the species. One possible explanation is that the various kinds of food organisms tend to be distributed conta- giously, with aggregations of considerably high- er density than reflected by area averages. As Ivlev (1961) demonstrated with carp fry, feed- ing rate can be expected to increase with degree of aggregation for a fixed quantity of food or- ganisms. Euphausiids have been observed in schools and breeding swarms (Brinton, 1962). More than likely the filtering response is evoked in mackerel with empty or near-empty stomachs by densities not much above the equivalent of 1 or 2 Artemia/Mter, where rate of eff"ective biting must already be on the order of 60 or more per minute. At 4 to 5 Artemia /liter the daily nu- tritional requirement could be obtained in less than 45 min of filter feeding. Whereas Artemia adults represent the smal- lest organisms consumed by the mackerel, they represent the largest organisms commonly con- sumed by the anchovy (Loukashkin, 1970); the latter also consumes phytoplankton and crusta- ceans less than 1 mm in length by filter feeding. The wet weight quantities of Artemia adults consumed at different densities (Table 5), based on the present study for the mackerel and on Leong and O'Connell (1969) and O'Connell (1972) for the anchovy, suggest the diff"erences in utilization and nutritional value of large crustaceans for equivalent age groups of the two species. The anchovy requires far less food than the mackerel to meet its daily nutritional require- ment, and can obtain the necessary quantity in about 20 min when Artemia adults are at den- sities of 1 or more per liter. In this same period the mackerel consumes more than the anchovy at all densities, but not enough to satisfy the daily requirement. When considered in terms of the weight of the two fish, the greater quan- Table 5. — Comparison of Artemia adults consumed (mg wet weight) by the 1 -year-old mackerel and anchovy for different density levels of Artemia in the water.^ Item Mackerel Anchovy Age in years 1-1- 1- Weight ± g 147 4 Nutritional requirement: Percent body weight 8 6.8 mg 11,784 270 Minutes of feeding for nutritional requirement at: 1 Artemia/Wler 97 20 5 Artemia/\\\eT 35 20 Mg Artemia consumed in 20 min with Artemia at: 0.1 Artemia/Wter 240 -152 1 Artemia/Uter 600 270 5 Artemia/\\\eT 9,600 270 Mg Artemia consumed/g \ fish weight in 20 min with Artemia 1 at: 0.1 Artemia/\her 1.6 -38 1 Artemia/\Her 17 68 5 Artemia/Uier 65 63 1 Calculations for the anchovy at age 1-f- and weight 7 g show the nutritional requirement and amounts consumed in 20 min at 1 and 5 Artemia/Uter to be 475 mg, but all other values ore the same as for the 4-g anchovy. titles consumed by the mackerel have a rela- tively lower nutritional value than those con- sumed by the anchovy, except, perhaps, at very high densities. The extreme diflference in quantities consumed by the two species at high densities is attribu- table to the filtering capability of the mackerel, and the failure of the anchovy to filter feed on the larger crustaceans, regardless of density. No meaning can be attached to the small diflference at the lowest density level because calculation of the values involved considerable extrapolation. It is nevertheless probable that the anchovy tends to remove large crustaceans at a lower rate than the mackerel at such levels. O'Connell (1972) showed that the feeding activity of the anchovy is likely to be divided between these and the smaller crustaceans captured by filtering if the larger organisms are less than about 5 to 8% of the dry weight of total zooplankton concen- tration. From the above considerations it is tentatively hypothesized that the individual mackerel gen- erally utilizes a greater proportion of the large crustaceans in the sea than does the individual anchovy, but that the proportion utilized gen- erally has a relatively low nutritional value for the mackerel. 980 O'CONNELL and ZWEIFEL: FEEDING OF PACIFIC MACKEREL LITERATURE CITED Brinton, E. 1962. The distribution of Pacific euphausiids. Bull. Scripps Inst. Oceanogr. Univ. Calif. 8:51-269. Carlisle, J. G., Jr. 1971. Food of the jack mackerel, Trachurus sym- metricus. Calif. Fish Game 57:205-208. Fitch, J. E. 1956. Pacific mackerel. Calif. Coop. Oceanic Fish. Invest., P-og. Rep. 1955-1956:29-32. 1965. Offshore fishes of California. Calif. Dep. Fish Game Booklet, 79 p. Frey, H. W. (editor). 1971. California's living marine resources and their utilization. Calif. Dep. Fish Game, [Sacramento], 148 p. Hatanaka, M. a., and M. Takahashl 1960. Studies on the amounts of the anchovy con- sumed by the mackerel. Tohoku J. Agric. Res. 11:83-100. Hatanaka, M., K. Sekino, M. Takahashi, and t. ichimura. 1957. Growth and food consumption in young mack- erel, Pne2imatophorus japonicus (Houttuyn). Tohoku J. Agric. Res. 7:351-368. IVLEV, V. S. 1961. Experimental ecology of the feeding of fishes. (Translated from the Russian by D. Scott.) Yale Univ. Press, New Haven, 302 p. Kariya, T., and M. Takahashi. 1969. The relationship of food intake to the stom- ach contents in the mackerel, Scomber japonicus. [In Japanese, English synopsis.] Bull. Jap. Soc. Sci. Fish. 35:386-390. Lasker, R. 1966. Feeding, growth, respiration, and carbon utilization of a euphausiid crustacean. J. Fish. Res. Board Can. 23:1291-1317. Leong, R. J. H., AND C. P. O'Connell. 1969. A laboratory study of particulate and filter feeding of the northern anchovy (Engraulis mor- dax). J. Fish. Res. Board Can. 26:557-582. LOUKASHKIN, A. S. 1970. On the diet and feeding behavior of the northern anchovy, Engraulis mordax (Girard). Proc. Calif. Acad. Sci., Ser. 4, 37:419-458. O'Connell, C. P. 1971. Variability of near-surface zooplankton off southern California, as shown by towed-pump sampling. Fish. Bull., U.S. 69:68*1-697. 1972. The interrelation of biting and filtering in the feeding activity of the northern anchovy {Engraulis mordax). J. Fish. Res. Board Can. 29:285-293. Yasuda, F. 1963. The food selectivity of some plankton feed- ers, with regard to the amount and size of bait. Rec. Oceanogr. Works Jap., New Ser. 7:57-64. 981 COLOR PATTERNS OF SPINNER PORPOISES (STENELLA CF. S. LONGIROSTRIS) OF THE EASTERN PACIFIC AND HAWAII, WITH COMMENTS ON DELPHINID PIGMENTATION William F. Perrin' ABSTRACT In this paper, the second of a series on the morphology of small pelagic delphinids by the author, the color patterns of forms of a tropical spinner porpoise Stenella cf. S. long- irostris (Gray) 1828 from the far eastern Pacific, from the waters between North America and Hawaii, and from Hawaiian waters are described and illustrated. The patterns can be analyzed in terms of discrete component systems, and most geographical variation appears to be in a "dorsal field system" overlying a basic general pattern. The overlay is darkest and most extensive in the easternmost form considered and lightest and least extensive in the Hawaiian form. The color patterns of three other delphinids, Stenella graffmani, Delphinus sp., and Tursiops truncatus are analyzed with the component ap- proach, and possible pattern element homologies are defined. During the course of collecting materials and data for population studies of porpoises involved in the seine fishery for tunas in the eastern trop- ical Pacific (Perrin, 1970a), I have during the last several years had the opportunity to observe and, less often, to photograph the color patterns of a large number of small pelagic delphinids of several species. Emerging patterns of variation and their possible implications for the systemat- ics of these very poorly known mammals have prompted me to undertake detailed analyses of this body of observations. I have previously (Perrin, 1970b) described the color pattern of the eastern Pacific spotted porpoise Stenella graffmani (Lonnberg) 1934. The primary pur- pose of this paper is to describe the ontogeny and geographical variation of the color patterns of spinner porpoises of the eastern tropical Pa- cific and Hawaii. Some description is included ' National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Southwest Fisheries Center, La Jolla. CA 92037. also of variation in the shape of the dorsal fin and correlated features. The study is based on observations on 157 specimens from the eastern Pacific (Figure 1), of which 41 were photographed, and on photo- graphs and observations of approximately 35 wild and captive animals from Hawaiian waters. Many of the specimens, including most of those from the far eastern Pacific east of Clipperton Island, were examined minutes after capture; others were examined after being held in frozen storage for up to 8 months. Degree of sexual development was determined by histological ex- amination of gonads, with the exception of the large male (Figure 8) from Acapulco, of which only a photograph is available and which was undoubtedly a mature male. Skeletons of spe- cimens of many of the animals illustrated were placed in museums, and the museum numbers are given in the figure captions. The spinner porpoise of the far eastern Pa- cific has been referred to Stenella microps (Gray) 1846 (Miller and Kellogg, 1955; Hester, Manuscript accepted March 1972. FISHERY BULLETIN: VOL. 70. NO. 3, 1972. 983 FISHERY BULLETIN: VOL. 70, NO. 3 160° I 140° 30' 20' 10' Hawaiian Is AA 160° 140° Figure 1. — Localities and numbers of specimens examined. Circles in eastern area indicate "eastern spinners"; triangles indicate "whitebelly spinners." Hunter, and Whitney, 1963; Nishiwaki, 1967; Pilson and Waller, 1970), but others regard S. microps as a synonym of S. longirostris (Gray) 1828 (Hershkovitz, 1966; Rice and SchefFer, 1968; Harrison, Boice, and Brownell, 1969) . The spinner porpoise of Hawaii has been referred to jS. longirostris (Hershkovitz, 1966; Nishiwaki, 1967; Tomich, 1969) and to S. roseiventris (Wagner) 1853 (Brown, Caldwell, and Caldwell, 1966; Fraser, in Morris and Mowbray, 1966; Rice and Scheffer, 1968) . The type localities for S. microps and S. longirostris are unknown; S. roseiventris was described from the Banda Sea, Indonesia. The spinner porpoise of the far off- shore areas of the eastern Pacific — called "white- belly spinner" by fishermen — has not to my knowledge been previously described or referred to any named species. No critical review of the Figure 2. — Calf of eastern spinner. Male, 105 cm total length, from lat 10°30'N, long 92°56'W, April 4, 1968. Perrin field no. CV83; speci- men not saved. Shape of dorsal fin is distorted by angle of photograph, should be higher. Photographed minutes after death. Figure 3. — Lateral (a) and ventral (b) views of subadult eastern spin- ner. Female, 166 cm, from 2l°43'N, 106°47'W, February 17, 1967. Ma- rine Mammal Biological Laboratory (Seattle) field no. 1967-102. Skel- etal specimen in MMBL collection. Photographed minutes after death. Figure 4. — Adult eastern spinner. Female, 166 cm, from 12°51'N, 93°18'W, April 9, 1968. Perrin field no. CV7; specimen not saved. Pho- tographed minutes after death. 984 PERRIN: COLOR PATTERNS OF SPINNER PORPOISES A 985 FISHERY BULLETIN: VOL. 70, NO. 3 genus has been accomplished since True's work on the Delphinidae published in 1889, when the existence of spinner porpoises in eastern Pacific and Hawaiian waters was not yet known. Evi- dence presented below concerning geographical variation in pigmentation supports the concept of a single species, highly variable geographi- cally, but taxonomic and nomenclatorial decisions must be based on adequate review of osteological characters, as well as on a broader range of ex- ternal characters than is considered here. The use of Stenella cf. S. longirostris therefore is provisional pending the outcome of more exten- sive studies underway by me and others. A si- milar situation obtains for the spotted porpoises, and use here of Stenella graffmani for the spotted porpoise (s) of the eastern Pacific is also pro- visional. DEVELOPMENT AND INDIVIDUAL VARIATION This account of ontogeny is divided into two parts, one describing development of color pat- tern as inferred from specimens from the more easterly, relatively nearshore population (s) of spinner porpoise referred to below as the "east- ern spinner," and another for specimens of the "whitebelly spinner" of the more westerly, far- ther offshore areas of the eastern Pacific (Fig- ure 1). EASTERN SPINNER Length at birth is 75 to 85 cm (Harrison et al., 1969). The smallest specimen examined was 80 cm long. The smallest specimen photo- graphed was 105 cm long (Figure 2). At this size, the animal is predominantly dark gray. The gray of the dorsum grades imperceptibly into white around the genital region and in a smaller area in the axillary region. A dark gray flipper band from flipper base to the eye-gape region is demarcated above by a narrow, very light line running from behind the posterior insertion of the flipper to the eye and below by a sharp boundary with a light gray gular region. The boundary between the two shades of gray runs forward between the eye and the end of the gape, becoming obscure in the furrow at the base of the melon. The very light gray below grades anteriorly into darker gray, so that the upper and lower sides of the snout are of about the same shade as the dorsal field. The gape is edged with very dark gray sharply delineated above for about the last third of the gape and below for about the last half of the gape but grading distally above and below into the generally dark gray of the snout. A small eye patch of similar shade is present, and a narrow eye stripe, also of dark gray, runs from the eye patch forward to join the gape mark near the apex of the melon. A similar mark extends from the blowhole to the apex. The margin of a very faintly defined dor- sal cape (not visible in the photographs) runs from near the apex of the melon to behind the dorsal fin, passing high over the eye and dipping slightly below the fin to yield a saddle effect dis- cernible only upon very careful scrutiny of a freshly caught animal. Low on the side and adjacent to the genital slit, an elongate, smudge- like mark extends obliquely for several centi- meters along a line that if extended in both di- rections would run from the eye to behind the anus. All appendages are on both surfaces the same gray as the dorsum. The major point of individual variation in animals of this and larg- er size is in the extent of the ventral and ax- illary white areas; in some individuals they are larger and may even be confluent, with the dorsal margin extending back from the axillary area to run into the higher genital white area. The pattern persists as described into sub- adulthood (subadult being defined as an animal of adult or near-adult size but sexually imma- ture) without change aside from overall dark- ening (Figure 3) . In subadulthood the margins of the ventral and axillary white areas begin to become speckled, and in adults (Figures 4, 5, 6, 7, 8, and 9) the dorsal field, by now very dark gray, appears to encroach on the white areas with spots and blotches to yield a very speckled ap- pearance below (Figure 10). As in younger animals, the major feature subject to individual variation appears to be the extent of the whrte areas, although in no adult specimens examined were the genital and axillary areas confluent. 986 PERRIN: COLOR PATTERNS OF SPINNER PORPOISES i ^ 00 to > ft 0} 0 O o '3 -4-> o CO «c ^ o o .S . M 1-H 'C 13 , C3 u J' X (U s o 4) > +3 -M Ph c4 rt tj CO OJ 9 «H -o TJ 00 -M u << £ to o Q) a> s -tJ o tH •♦H a ^ -u P 13 3 C £ S <; S 0! ^ 1-1 00 (U v> > ft ^ o O 2 -i-> o (ll s £8 0) "3 SH "O T3 ^ 0(5 -)-> -4-> < 2 to 05 o iH tH us o to c 991 FISHERY BULLETIN: VOL. 70, NO. 3 13) more closely resembles the Hawaiian spinner than does the adult, in that the margin of the ventral white field is relatively smooth. The pat- tern of the fetus of the eastern spinner (Figure 25) is also very close to the Hawaiian pattern, with the tripartite effect of cape, lateral field, and high ventral field being very pronounced. The question of which sort of developmental cline is involved, paedogenesis to the west or gerontogenesis to the east, is however yet an open one and must be settled by consideration of ad- ditional lines of morphological and zoogeograph- ic evidence. Pattern component analysis. — Comparison of the color pattern of a partially albinistic subadult whitebelly spinner (Figure 14) with that of a normally pigmented individual of the same sex and nearly the same length (Fig- ure 13) affords insight into the mechanisms of pattern formation. In the lighter-than-normal animal, the cape, eye patch and stripe, gape mark, dorsal fin, and flukes are as in the normal specimen. The lateral field, flipper band, and oblique genital mark are obscure, and the flippers are white on both surfaces. White brushings sweep up from the edge of the cape at about midlength toward the dorsal fin. From these facts, the inference can be made that the full normal pattern is the result of the combined ef- fect of two independent pigmentation systems, one involving the cape and accessory stripes, eye and gape marks, dorsal fin, and flukes, and the other involving lateral field, flipper band, and flippers. While the genital mark is not apparent in the albinistic animal, the developmental pat- tern of obliteration through development of the lateral field suggests that it belongs to the cape system. The difference among the Hawaiian, whitebelly, and eastern spinners can be inter- preted schematically in terms of these hypothet- ical systems (Figure 26) , allowing a clearer pic- ture to emerge of the possible patterns of geo- graphical variation involved. Since the lateral field appears in this sense to be the lateral por- tion of a more extensive dorsal field overlaying the cape, it is referred to below as an aspect of a "dorsal field overlay." Among these three forms, the "dorsal field Figure 13. — Lateral (a), dorsal (b) , and ventral (c) views of subadult whitebelly spinner. Male, 164 cm, from 9°47'N, 133°25'W, August 11, 1970. Perrin field no. WFP65 ; U.S. National Museum 396017 (complete skeleton). Rostrum damaged. Photographed aft- er frozen for several months and thawed in water. Figure 14. — Lateral (a) and dorsal (b) views of albinistic subadult white- belly spinner. Male, 154 cm, from 10°19'N, 135°38'W, August 5, 1970. Perrin field no. WFP67; U.S. National Museum 396020 (complete skeleton). Compare with normally pigmented ani- mal in Figure 13. Note dark flukes and light flippers. Photographed after frozen for several months and thawed in water. overlay" is darkest and most extensive in the easternmost form, and least so in the Hawaiian form. It is not yet known whether the three forms represent modes in a geographically con- tinuously varying single species or two or more reproductively isolated allopatric or sympatric populations. OTHER DELPHINIDS The color patterns of other closely related del- phinids can be similarly dissected into hypothet- ical component systems. The patterns of three forms of which I have directly examined live or freshly captured dead specimens are analyzed in Figure 27, with an attempt to define possible ho- mologies for the component systems among the species. As the first step in breaking down the patterns into supposedly homologous compo- nents, I attempted to define for each a cape with associated markings, similar to that in the al- binistic spinner described above. In the eastern Pacific spotted porpoise, Stenel- la graffmani, (Figure 27, top) the general cape system is very close to that of the spinners, with the cape extending farther ventrad. The dorsal overlay is less extensive than in any of the spin- ners, and the flipper band runs forward to the gape rather than to the eye. The area of con- tact between the flipper band and the gape mark in this and in similar spotted porpoise from other 992 PERRIN: COLOR PATTERNS OF SPINNER PORPOISES 993 FISHERY BULLETIN: VOL. 70, NO. 3 parts of the world is emphasized in many indi- viduals by a zone of lighter pigmentation (see Figures 4 and 6 in Perrin, 1970b; Figure 1 in Best, 1969; and Figure 3 in Nishiwaki, Naka- jima, and Kamiya, 1965), which supports the idea of separate origin of the two marks. In the spotted porpoises, the combined pattern systems are overlaid with discrete dorsal and ventral spot systems that to greater or lesser extent obliterate them (Perrin, 1970b). In Delphinus spp. (Figure 27, middle) the dorsal field overlay is less extensive anteriorly than the cape, resulting in invasion of the cape by the ventral field and yielding a four-part criss- cross pattern with zones of black, buflF, gray, and white. The deduction to be made is that the whitebelly region represents total lack of pig- ment, the buff "thoracic patch" (terminology of Mitchell, 1970) represents the color yielded by the pigment of the cape alone, the gray "flank patch" (of Mitchell) that of the pigment of the dorsal field overlay alone, and the black dorsal- most area that of the combined eflfect of the pig- ments of the cape and the dorsal field overlay. Chromatographic analysis of pigments present in the various regions and comparison of the re- sults with those of similar tests for other del- phinids would contribute to verification or re- jection of the suggested homologies. The pattern of Tursiops truncatus (Figure 27, bottom) is very close to that of the spinners in every respect, with the flipper band running to the eye and demarcated dorsally by a narrow light line. The dorsal field overlay of varying darkness extends ventrad to about the same de- gree as in the eastern spinner. Observation of wild bottlenose porpoise in the eastern temper- ate and tropical Pacific and examination of some of the thousands of published photographs of animals from around the world lead me to believe that the main component of geographical var- iation, as for the spinner porpoises, is the extent and darkness of the dorsal overlay. In Ha- waiian Tursiops that I have observed, the vent- ral margin was high and sharply defined exactly as in the Hawaiian spinner. The general cape system common to the four forms discussed above corresponds roughly to the "saddled pattern" described by Mitchell Figure 15. — Lateral (a) and ventral (b) views of subadult whitebelly spin- ner. Male, 166 cm, from 10°N, 129° or 136°W, August 10-12, 1970. Perrin field no. WFP71; U.S. National Mu- seum 396023 (complete skeleton). Pho- tographed after frozen for several months and thawed in water. Figure 16. — Lateral (a) and ventral (b) views of adult whitebelly spinner. Male, 169 cm, from 9°47'N, 133°25'W, August 11, 1970. Perrin field no. WFP53; U.S. National Museum 896031 (complete skeleton). Photo- graphed after frozen for several months and thawed in water. Figure 17. — Adult whitebelly spinner. Male, 174 cm, from 9°47'N, 133°25'W, August 11, 1970. Perrin field no. WFP76; U.S. National Museum 396170 (complete skeleton). Photo- graphed after frozen for several months and thawed in water. (1970) as a generalized and probably primitive pattern within the Delphinidae. The "saddled pattern" depicted in his Figure 7, however, is definitely not that of the spinner porpoises, which he included in a group of "saddled" species in his Figure 17. I concur with Mitchell in his se- lection of the "saddled" condition as a good can- didate for a primitive pattern, insofar as his definition of that pattern category pertains to the "general cape system" described here. Lines of evidence presented here leading to this conclu- sion are the possession in common of a relatively invariant cape system by several delphinid spe- cies and the tendency of the cape to be partially marked, distorted, or obliterated in a varying fashion within a species or species group not so much by alteration of its intrinsic form but more by interaction with more plastic overlying systems or (as Mitchell pointed out in the case of the "spinal blaze") by subtraction through invasion by areas of nonpigmentation. 994 PERRIN: COLOR PATTERNS OF SPINNER PORPOISES 995 ACKNOWLEDGMENTS Specimens were collected and examined through the cooperation of the owners, captains, and crews of the tunaboats Carol Virginia, Conte Bianco, Connie Jean, Conquest, Anne M, Pacific Queen, and Queen Mary. C. J. Orange, J. S. Leatherwood, and G. D. Sharp collected speci- mens. D. W. Rice provided the photographs used in Figure 3. J. H. Prescott provided photo- graphs used in Figure 8. K. S. Norris provided photographs used in Figure 24, and arranged for a visit to Hawaii to observe wild and captive spinner porpoise. I am indepted to Bonnie Dalzell for the ideas concerning the possible sig- nificance of the canted fin and ventral keel in the eastern spinner. K. S. Norris, E. D. Mitchell, D. W. Rice, C. 0. Handley, Jr., and D. K. Cald- well read the manuscript. E. D. Mitchell oflFered especially valuable criticism. LITERATURE CITED FISHERY BULLETIN: VOL. 70, NO. 1 Figure 18. — Adult whitebelly spinner. Male, 177 cm, from 9°47'N, 133°25'W, August 11, 1970. Perrin field no. WFP54; U.S. National Mu.5eum 396032 (complete skeleton). Photo- graphed after frozen for several months and thawed in water. Figure 19. — Adult whitebelly spinner. Male, 178 cm, from 10°N, 128° or 136°W, August 10-12, 1970. Perrin field no. WFP52; U.S. National Museum 396030 (complete skeleton). Photographed after frozen for several months and thawed in water. Figure 20. — Dorsal surface of flipper of speci- men depicted in Figure 17. Figure 21.— Ventral surface of flipper of speci- men depicted in Figure 19. Best, P. B. 1969. A dolphin (Stenella attenuata) from Durban, South Africa. Ann. South Afr. Mus. 52:121-135, Plates 11-17. Brown, D. H., D. K. Caldwell, and M. C. Caldwell. 1966. Observations on the behavior of wild and captive false killer whales, with notes on associ- ated behavior of other genera of captive delphin- ids. Los Ang. Cty. Mus. Contrib. Sci. 95, 32 p. Harrison, R. J., R. C. Boice, and R. L. Brownell, Jr. 1969. Reproduction in wild and captive dolphins. Nature (Lond.) 222:1143-1147. Hershkovitz, p. 1966. Catalog of living whales. U.S. Natl. Mus. Bull. 246, 259 p. Hester, F. J., J. R. Hunter, and R. R. Whitney. 1963. Jumping and spinning behavior in the spin- ner porpoise. J. Mammal. 44:586-588. Miller, G. S., Jr., and R. Kellogg. 1955. List of North American recent mammals. U.S. Natl. Mus. Bull. 205, 954 p. Mitchell, E. 1970. Pigmentation pattern evolution in delphinid cetaceans: an essay in adaptive coloration. Can. J. Zool. 48:717-740. Morris, R. A., and L. S. Mowbray. 1966. An unusual barnacle attachment on the teeth of the Hawaiian spinning dolphin. Nor. Hvalfan- gst-tid. (Norw. Whaling Gaz.) 55:15-16. Nishiwaki, M. 1967. Distribution and migration of marine mam- mals in the North Pacific area. Bull. Ocean Res. Inst., Univ. Tokyo 1:1-64. Nishiwaki, M., M. Nakajima, and T. Kamiya. 1965. A rare species of dolphin (Stenella attenu- ata) from Arari, Japan. Sci. Rep. Whales Res. Inst., Tokyo 19:53-64, Plates I-VI. Perrin, W. F. 1968. The porpoise and the tuna. Sea Front. 14: 166-174. 1970a. The problem of porpoise mortality in the U.S. tropical tuna fishery. Proc. Sixth Annu. Conf. Biol. Sonar Diving Mammals, Stanford Res. Inst., Menlo Park, Calif., p. 45-48. 1970b. Color pattern of the eastern Pacific spotted porpoise Stenella graffmani Lonnberg (Cetacea, Delphinidae). Zoologica (N.Y.) 54:135-142, Plates I-VII. Perrin, W. F., and J. R. Hunter. 1972. Escape behavior of the Hawaiian spinner porpoise (Stenella cf. S. longirostris) . Fish. Bull., U.S. 70:49-60. PiLSON, M. E. Q., AND D. W. Waller, 1970. Composition of milk from spotted and spin- ner porpoises. J. Mammal. 51:74-79. 996 PERRIN: COLOR PATTERNS OF SPINNER PORPOISES 19 ■^•aiafti: 997 FISHERY BULLETIN: VOL. 70, NO. 3 Rice, D. W., and V, B. Scheffer. Spec. Publ. 57, 238 p. 1968. A list of the marine mammals of the world. True, F. W. U.S. Fish. Wildl. Serv., Spec. Sci. Rep. Fish. 579, 1889. Contributions to the natural history of the 16 p. cetaceans, A review of the family Delphinidae. TOMICH, P. Q. Bull. U.S. Natl. Mus. 36, 191 p., Plates I-XLVII. 1969. Mammals in Hawaii. Bernice P. Bishop Mus. Figure 22. — Lateral (a) and ventral (b) views of adult whitebelly spinner. Female, 179 cm, from 3°20'N, 110° 44'W, June 24, 1970. Perrin field no. WFP79; U.S. National Museum 396173 (complete skeleton). Rostrum damaged. Photographed after frozen for several months and thawed in water. Figure 23. — Adult whitebelly spinner. Female, 169 cm, from 8°19'N, 119° 15'W or 3°20'N, 110°44'W, June 11 or 24, 1970. Perrin field no. WFP80; U.S. National Museum 396174 (complete skeleton). Photographed after frozen for several months and thawed in water. 998 PERRIN: COLOR PATTERNS OF SPINNER PORPOISES 999 FISHERY BULLETIN: VOL. 70, NO. 3 1000 PERRIN: COLOR PATTERNS OF SPINNER PORPOISES Figure 24. — Lateral (a), dorsal (b), ventral (c), and close-up lateral (d) views of Hawaiian spinner porpoise. Subadult male, 179 cm, from Waikiki Beach, Oahu, May 25, 1970. Oceanic Institute field no. OI70- 35 ; specimen stolen during prep- aration. Photographed 3 hours after death. H '^ 0) > M S-l o .)i a :^ &^ 03 «t-4 S >.^ ■3 JS rC ^ ^ T3 03 o .0 OS ^ o o3 u o 'a c 03 O) "o «f-l u o p ^ 03 1001 FISHERY BULLETIN: VOL. 70, NO. 3 Z ^ < U < Of lU Z o 09 to >> 09 V c o o ^ C ■Ft 09 a> w •i-« o a > ^ < o P^ c^ a LU u z < a. > o C X lU Q 'S. •0 tn z lU t4H lU U. o ^ oe .1 m .•? g < g Q .M pt - ^y' O C- -' .t'.'-v"' »t 9% /'' o ^*^ e^! ^ "o _ 43 ? u C s u W P 1002 PERRIN: COLOR PATTERNS OF SPINNER PORPOISES r-- %,-,0 s-^ > u 0) o -o p. p o "« a. n.? (M ^ t/ ■t-> W 3 t: p M -<- fe O I- 1003 APPARENT ABUNDANCE OF SOME PELAGIC MARINE FISHES OFF THE SOUTHERN AND CENTRAL CALIFORNIA COAST AS SURVEYED BY AN AIRBORNE MONITORING PROGRAM James L. Squire, Jr.^ ABSTRACT From September 1962 through December 1969, commercial aerial fish spotter pilots esti- mated tonnage of species observed during flights off the southern and central California coast. Observations of fish and the aircraft's flight route were recorded on special charts. These data were analyzed using 10-minute-longitude by 10-minute-latitude "block areas." A total of over 17,593 flight hours was involved, surveying 57,628 block areas — 37,186 during the day and 20,442 during the night. Data from each block area were used to compute diurnal and nocturnal variation in apparent abundance and an annual index of apparent abundance. Pacific bonito, Sarda chiliensis, and yellowtail, Seriola dorsalis, were observed in greater frequency and quantity during the day, and the northern anchovy, Engraulis mordax; jack mackerel, Trachiirus symmetriciis \ and Pacific mackerel. Scomber japoniais, were observed in greater frequency and quantity during the night. Pacific barracuda, Sphy- raena argentea, was observed in greater quantity at night but more frequently during the day. Between 1963 and 1969 indexes of apparent abundance declined for jack mackerel. Pacific mackerel, Pacific sardine. Pacific bonito, Pacific barracuda, and yellowtail and increased slightly for the northern anchovy. The index closely follows estimates of total abundance for the Pacific mackerel, a species for which reliable estimates of total abundance are available. From observations of the catch trends in the bonito fishery, the index appears to be little aff'ected by changes in economic demand. Its trends in apparent abundance are evident before they are reflected in catches and are useful in the evaluation of catch variations in underutilized resources. In a search for more efficient fishing methods, many of the fisheries throughout the world that catch pelagic surface schooling species are using aircraft to locate and guide the fleet to the schools and in some cases to direct the catching oper- ation (Gushing, Devoid, Marr, and Kristjonsson, 1952). In some areas of the United States the services of the fish spotter are vital to the suc- cess of the commercial fleet which depends in part on the aircraft scouting the fishing grounds to obtain current information on the location of near-surface schooling fish (Squire, 1961), At times commercial aerial fish spotters assist the * National Marine Fisheries Service, Southwest Fish- eries Center, La Jolla. CA 92057, Manuscript accepted March 1972. FISHERY BULLETIN: VOL. 70, NO. 3, 1972. sport fishing fleet by advising them of the lo- cation of desirable marine game species. Data obtainable by techniques of aerial obser- vation have been used by fishery biologists to gain information on distribution and abundance of pelagic near-surface schooling fish, Sette (1949) investigated the possibilities of aerial scouting for sardines off southern Cali- fornia in search of a method that would provide information useful in estimating abundance yet be free of the availability influence. Aerial scouting was conducted during the day, and com- mercial fishing was conducted at night. As a result the spotting data were deemed less reli- able than those obtained from the commercial fishery. Jones and Sund (1967), using commercial fish 1005 FISHERY BULLETIN: VOL. 70, NO. 3 spotter aircraft in a search for tuna schools in the same area surveyed by a research vessel, found that the aircraft was about two and one half times more efficient than the vessel at lo- cating fish schools. An evaluation of aircraft by the U.S. Navy, for making biological obser- vations, indicated that for whales the frequency of sighting averaged about 20 times greater than that from ships (Levenson, 1968). From 1956 to 1964 the California Department of Fish and Game conducted monthly survey flights along the California coast from San Fran- cisco to Mexico. Data were published as flight reports in chart form showing the aircraft's flight track, areas surveyed, notes on species ob- served, and number of schools and their geo- graphical location. Large variations in the num- ber of schools visible over a short-time period appeared to limit the usefulness of the data, and the surveys were discontinued in 1964. The av- erage number of schools sighted per flight was determined by Wood (1964),° and from these data a comparison was made of the relative abundance of northern anchovy, Engraulis mor- dax, schools for the period 1956 through 1963. Limitations on flights to nearshore areas during daylight and low search time in any one area restricted the potential of these surveys for de- termining the apparent abundance of the many pelagic species found off the California coast. Fish spotter aircraft range over a large geo- graphical area, and during these flights they may observe concentrations of several species of pe- lagic fish. Many times these fish are not caught for one or more reasons, such as fishing boats not equipped with proper nets, concentrations are small, species is not economically desirable, and fishing boats are not capable of reaching fish within a reasonable length of time. However, the fish spotters are able to identify these con- centrations of fish. Species commonly observed by the aerial fish spotters within the survey area were northern anchovy, jack mackerel, Trachurus symmetricus; Pacific bonito, Sarda chiliensis; Pacific mack- • Wood, R. 1964. Aerial surveys along the California coastline 1956 to 1963. Document V prepared for the Marine Research Committee meeting, March 6 1964, San Pedro, Calif., 2 p. [Processed.] erel, Scomber japonicus; Pacific sardine, Sardi- nops sagax; bluefin tuna, Thunnus thynniis; Pacific barracuda, Sphyraena argentea; white seabass, Cynoscion nobilis; and yellowtail, Seri- ola dor sails . The majority of fish spotting effort is directed toward the location and catching of jack mack- erel. Pacific mackerel. Pacific bonito. Pacific sar- dine, and in recent years the northern anchovy. Of these five species the Pacific sardine and Pa- cific mackerel are most economically desirable with jack mackerel, Pacific bonito, and northern anchovy of descending importance. To increase knowledge on the apparent abun- dance of pelagic near-surface marine life, the Tiburon Marine Laboratory initiated a pelagic fish monitoring program in cooperation with aerial fish spotter pilots who are active in spot- ting for the southern and central California coastal commercial fishery. These cooperators were individuals with specialized training and experience. When assisting the commercial fleet, fishing success is dependent upon accurate identification of schooling species by the spotters. They have considerable experience in estimating the weight of fish schools, and they are consid- ered to be quite accurate in the estimation of weight. There are a number of variables that affect the statistical accuracy of fish spotter data which are difficult to evaluate, such as individual dif- ference in ability of pilots to locate fish, deter- mine species, and estimate school size, and esti- mate total tonnage available in a fishing area. Variation in estimating school size probably has more effect on the data than the other variables. However, since at least five experienced observ- ers were used in the program during each year, it was assumed that reasonable annual averages were obtained. This report consists of an analysis of aerial fish spotter data for the period September 1962 through December 1969 to determine if, for the species commonly observed, it can be used to: (1) compute an accurate index of apparent abundance and (2) obtain a trend in the appar- ent abundance of pelagic near-surface species and in particular those of underutilized re- sources. 1006 SQUIRE: PELAGIC MARINE FISHES OFF SOUTHERN AND CENTRAL CALIFORNIA METHODS AND PROCEDURES Five fish spotter pilots were contracted to re- cord observations of pelagic species, giving loca- tion, number of schools, estimated tonnage of each school or groups of schools, counts of large marine animals, and flight track for each survey flight. Two pilots usually covered the Santa Barbara • Channel and Santa Barbara Islands north to Estero Bay and occasionally into Mon- terey Bay. The remaining three spotter pilots normally sui-veyed the area from west of Los Angeles to southwest of San Diego and occasion- ally offshore to San Clemente Island, Cortez Bank, and San Nicholas Island. Flight oper- ations were conducted during daylight hours or on nights during the dark phase of the moon at elevations of 500 to 1,200 ft (152 to 365 m) above the sea surface. TECHNIQUES OF OPERATION Specific observation of a fish school has three phases: (1) distinguishing a school, (2) identi- fying the species, and (3) estimating weight of the school. The detection of near-surface schools during the day is dependent upon the pilot's ability to distinguish subtle color and light in- tensity differences in the water. Detection of schools at night is possible only during the dark period of the moon and depends on the pilot's ability to discern gradation of light intensity. Bioluminescence of planktonic organisms agi- tated by schooling fish indicates by a dull glow the location and size of the school. Species are identified during the day on the basis of a com- bination of two or more of the following char- acteristics: color of school or individual fish, shape of school, and behavior and size of indi- viduals within the school. At night, species identification is based on shape of the luminous area and behavior of the schooling fish under un- disturbed conditions, or by the behavior of the school after being subjected to a stimulus from an external source such as a flash from the air- craft's landing light. At first observations were recorded by the pilots on small portable tape recorders. This method was unsatisfactory, and recorders were replaced with three charts covering the coastal waters from the Coronado Islands, Mexico, north to Half Moon Bay, Calif. The charts were com- pleted by the pilots after each flight and were submitted quarterly to the National Marine Fisheries Service. Figure 1 illustrates the type of information recorded by the fish spotter pilot. PROCESSING OF OBSERVATION DATA Each chart was overlayed with a 10-minute- longitude by 10-minute-latitude grid, numbered according to the California Department of Fish and Game "Block area" statistical system (Clark, 1935). With the gridded chart, the ob- sei-vation and flight track data could be con- veniently tabulated and coded for subsequent computer analysis. California Department of Fish and Game statistical code numbers were assigned to each of the 27 species of marine animals observed. The computer output grouped data by species, year, week, block area, day or night observation, number of schools, tons per school, and tons per block area. The data for block areas were later combined into 11 larger grouped block areas or "zones" lettered A through K (Figures 2 and 3) . These zones were selected to outline important geographical areas where fish were commonly observed. The following criteria were used in tabulating the data from the flight charts: 1. Groups of schools which were indicated on the flight chart as covering more than one block area were listed for each block holding part of the group. For example, if one group of schools (10 schools, 15 tons per school, total 150 tons) overlapped two block areas equally, each area was credited with having 5 schools at 15 tons per school, equalling 75 tons per block area. 2. If only one school was shown overlapping two block areas, the school was assigned to the block area having the greatest portion of the school. 3. If a large area of fish was indicated involv- ing more than two block areas and only a total tonnage estimate made, the tonnage was credited to the block areas in proportion to area outlined. 1007 FISHERY BULLETIN: VOL. 70, NO. 3 MARINE RESOURCE MONITORING SYSTEM Aircraft M /Sf^J> Date 7'^-^tfTI m e lllA'-f(i£-o Tach. ♦im>j-y^^.2yTn^yj//7 U.S. DEPARTMENT OF COMMERCE PilOt 0^^*1n>i. 'j'ruI./M^ National Oceanic and Atmoapharic Adminiatration Notional Marine Fisheries Service Soothwest Fisheries Center LA JOLLA Laboratory Santa Monica t >U#. /3' LOS ANGELES [^untington Beacti Newport Btact> ;?o >»«.'* .aguna Beach kOana Pt. >an Clemente Oceantid* tanner bonk eortez -; j>anii' tanner bank eortez' '-.bank 10 20 30 jeOMB IziSFS 50 Figure 1. — Flight chart for southern California area showing typical flight track and fish and mam- mal observations. Block area grid is overlayed on chart for coding observations. 1008 SQUIRE: PELAGIC MARINE FISHES OFF SOUTHERN AND CENTRAL CALIFORNIA 480 479 V47e r X563 H04 5CB B02 N 501 SI2 511 Rin BOB 508 V507 EY 820 519 518 817 isie ^ fie 9 sea se7 Vmonter I 834 J i 833 632% T 840 639 K 538 5fiO 849 848 V 566 566 R«4 563 562 861 V56O 604 603 60C \ L60I 610 609 60S 607 \ turtDPA a Av — 618 617 616 I 616 ^ 626 9i9 E 2^21 623 J 622 ^35 634 3 633 632 ^631 640 639 636 __J637 6^S_ Jg^^ Q^^ c^ ,643 Figure 2. — Block areas grouped into zones (A-B), selected to outline the more important coastal fishing areas. 1009 FISHERY BULLETIN: VOL. 70. NO. 3 SAh 655 M\ 880 879 ^^7 S7 N BS9 SM SS7 658 t 654 jTABA 852 RBAR 91 ijj 801 \g2 1 firn i/ 667 886 88S k e»i 690 689 rii 710 896 io» r 6aff 884 nr> ^ ^^^" ^ Ik OS ANGELES 714 rn TU r \7M TOT ro« 706 704 703 702 7nA 733 732 7 31 7 30 L J 121^ ^720 719 r\ 18 752 751 750 749 746 747 74 6 745 74 4 742 0 741 740 7S9 736 771 770 769 766 767 r "766 765 764 763 76? Hi S07 806 1 752 758 757 ■6 ■IS ■4 t 613 ■2 811 am 608 1 805 804 803 802 83(1 829 827 82^ B2!> 824 S.Z3 A J 6£J. 843 pa 7 BS8 MM BK4 an 3 asi aan i 849 867 848 847 846 845 844 JI842 874 B7S S7'2--. •>r-. 870 aaa aaa e^s ns 964 863 862 861 MsAN DIEGO AAJ C T UQ -889 aaa 884 883 881 aeo 67 8 K r 897 »I8 K • 916 r Figure 3. — Block areas grouped in zones (C-K), selected to outline the more important coastal fishing areas. 4. Whales, porpoises, and sharks were re- corded as numbers of individuals observed. 5. When the flight track entered any portion of a block area, the block area was credited for the purpose of determining observation efl!"ort as having a "block area flight." RESULTS During survey flights, 20 species of fish were observed and identified. A number of other marine species (mammals, invertebrates) were observed and all are listed in Table 1. DISTRIBUTION OF FLIGHT OBSERVATION EFFORT A total of 17,593 flight hours were logged by spotter pilots during the survey period. The number of block area flights from September 1962 through the end of 1969 totaled 57,628 with 37,186 block areas surveyed during the day and 20,442 at night. Distribution of day and night block area flights by year and by zone is shown in Table 2. NOCTURNAL AND DIURNAL VARIATION IN NUMBER OF SIGHTINGS AND TONNAGE To determine criteria concerning the frequen- cy of observation during the day and night for each of the species more commonly observed, the ratios in numbers of sightings and tonnages ob- served were calculated for the period September 1962 through December 1966. Information on the diurnal and nocturnal frequencies and mag- nitude of occurrence for each species is of im- portance in evaluating which observation (day 1010 SQUIRE: PELAGIC MARINE FISHES OFF SOUTHERN AND CENTRAL CALIFORNIA Table 1. — Species of fish and other marine animals observed during surveys. Fish: Basking shark (Cetorhinus maximus) White shark (Carcharodon carcharias) Northern anchovy (Engraulis mordax) Pacific sardine (Sardinops sagax) Pacific bonito (Sarda chiliensis) Jack mackerel (Trachurus symmttricus) Pacific mackerel (Scomber japonicus) Pacific barracuda (Sphyrama argentta) Yellowtail (Siriola dorsalis) Whita seobass (Cynoscion nobilis) Bluefin tuna (Thunnus thynnus) Albacora tuna (Thunnus atalunga) Yeliowfin tuna (Thunnus albacares) Skipjack tuna (Katsuwonus pelamis) Jacksmelt (Atherinopsis calijornimsis) Ocean sunfish (Mola mola) Striped bass (Morone saxatilis) Pacific saury (Cololabis saira) Swordfish (Xiphias gladius) Striped marlin (Tttrapturus audax) Mammals: Gray whale Pilot whole Blackfish (killer whale) Porpoise and dolphin Seals and sea lions Invertebrates: Squid Jellyfish or night) might be the more significant in eval- uating the trend of apparent abundance. These data were calculated during an earlier part of the study (1962-1966) to evaluate the method of using aerial fish spotter data. The total amount of fish estimated to have been seen by the aerial fish spotters during the period 1962-1966 was 5,289,521 tons of the following species: northern anchovy, 4,550,218 tons; jack mackerel, 335,794 tons; Pacific bonito, 238,247 tons; Pacific mackerel, 103,464 tons; and yel- lowtail, 1,955 tons. Annual sightings of each species per block area flight were expressed as a percentage of all block area flights day and night (% day/ % night) and are shown in Table 3. The ratio of diurnal and nocturnal sightings was obtained by dividing the percentage of day sightings by the percentage of night sightings. Ratio values greater than 1.00 indicate a greater number of sightings during day, less than 1.00 indicate greater number of sightings during the night. To determine the day and night diff'erences in the tonnage observed for each species, the Table 2. — Observation effort (day/night) in block area flights by zone for the period September 1962 through 1969. [Data are presented in number of block area flights (day/night).] Zone 1962 1963 1964 1965 1966 1967 1968 196? Total A 41/62 175/78 125/86 102/84 239/9 361/0 585/0 86/29 1,714/348 B 174/1117 220/75 283/156 471/252 770/90 610/13 519/0 126/10 3,173/713 C 104/126 470/658 632/680 892/743 1,860/495 1,016/559 1,014/281 2,130/718 8,118/4,260 D 12/71 137/167 409/518 485/434 1,268/385 813/533 1,000/291 1,942/712 6,066/3,111 E 0/10 63/96 40/263 48/58 108/30 40/51 101/109 87/97 493/714 F 0/6 15/32 23/37 2/106 37/21 15/53 35/194 79/48 206/497 G 0/2 454/496 394/610 481/404 1,358/434 874/596 616/676 735/1,500 4,939/4,718 H 0/0 291/282 363/336 387/349 723/206 435/305 368/469 303/576 2,870/2,523 1 0/0 366/155 477/157 395/258 814/208 614/281 544/378 613/348 3,823/1,785 J 0/0 463/198 586/300 550/128 847/113 1,087/154 672/294 932/188 5,137/1,375 K 0/0 18/0 13/0 13/0 69/9 65/10 75/32 106/1 350/52 Total 331/394 2,672/2,237 3,336/3,143 3,826/2,816 8,120/2,000 5,936/2,555 5,826/3,070 7,139/4,227 37,186/20,442 Grand total 725 4,909 6,479 6,642 10,120 8,491 8,896 11,366 57,628 Table 3. — Annual sightings per block area flight in percentage (day/night) and day/night averages and ratios. Species 1962 1963 1964 1965 1966 Day/night Average Ratio Northern anchovy 3.3/21.6 Pacific bonito 6.3/ 6.3 Jack mackerel 0.3/ 1.0 Pacific mackerel 0.0/ 6.9 Pacific sardine 1.2/ 1.5 Ydllowtail 0.0/ 0.0 Pacific barracuda 1.8/ 0.0 8.7/19.7 7.3/ 5.1 3,5/ 9.9 4.2/ 9.6 1.5/ 2.3 1.0/ 0.4 0.9/ 0.6 7.8/21.5 9.7/ 3.4 5.0/ 6.7 2.0/ 3.5 0.8/ 2.1 0.2/ 0.0 0.5/ 0.6 4.9/11.8 8.3/ 1.7 3.3/ 6.2 0.3/ 1.2 0.1/ 0.4 0.2/ 0.0 0.7/ 0.0 5.4/25.0 6.3/ 2.5 1.8/ 5.7 0.1/ 2.0 0.2/ 0.4 0.2/ 0.0 0.2/ 0.1 6.0/19.9 7.5/ 3.8 2.7/ 6.1 1.3/ 4.6 0.7/ 1.3 0.3/ 0.0 0.8/ 0.2 0.30 1.99 0.47 0.28 0.57 4.00 3.15 1011 FISHERY BULLETIN: VOL. 70, NO. 3 amount observed (day or night) in each zone was divided by the number of block area flights (day or night) within the zone. The average number of tons observed per block area flight for each zone, the average number of tons observed for all zones combined, and ratios of day and night tonnages observed are shown in Table 4. Ratios were obtained by dividing the tons per block area flight (day) by tons per block area flight (night). Therefore, ratios greater than 1.00 indicate greater tonnage during the day, less than 1.00 indicate greater tonnage during the night. AVERAGE WEIGHT OF FISH SCHOOLS Average weight of schools was computed for the period September 1962 through December 1966 from all data having estimates of individual schools by weight. As previously indicated, some tonnages were given by areas, not by num- bers of schools and tonnages of each school. The average tonnage per school is listed for each spe- cies in Table 5. INDEX OF ANNUAL APPARENT ABUNDANCE An index of annual apparent abundance was calculated for observations during day and night for each zone and for all zones combined from September 1962 through December 1969 for the northern anchovy. Pacific bonito, jack mackerel. Pacific mackerel. Pacific sardine, yellowtail, and Pacific barracuda. Marr (1951) defined the term apparent abundance as "abundance as af- fected by availability, or the absolute number of fish accessible to a fishery." This definition of apparent abundance most nearly describes the type of index calculated in this paper. For convenience in calculating this index, four arbitrary tonnage ranges were selected for each species. Tonnage ranges for each species were selected to cover the entire range of observed tonnages that may occur in any one block area. The midpoint tonnage of each range was divided by 100 for the northern anchovy and by 10 for Pacific bonito, jack mackerel. Pacific mackerel, and Pacific sardine to provide a tonnage range value (X) of convenient size to be used in the Table 4. — Day/night differences in tonnage and ratios observed based on average tons observed per block area flight in each zone for the period September 1962-1966. Species one 1 zone Tons/block area flight Day/nighf day/n, ight ratio Northern ai -ichovy Zone A 478.8/636.1 0.75 B 148.5/832.2 0.17 C 45.4/386.4 0.11 D 32.6/337.5 0.09 E 9.8/214.0 0.04 F 3.9/ 39.2 0.01 G 105.0/197.3 0.53 H 64.4/ 75.8 0.84 1 10.8/ 92.3 0.11 J 32.6/237.1 0.13 K 13.5/ 11.1 1.21 Average- ■all zones 79.8/299.7 0.26 Pacific boni to Zone A 0.6/ 0.0 B 0.5/ 0.2 2'50 C 15.8/ 7.6 2.00 D 12.8/ 2.1 6.09 E 44.5/ 0.1 445.00 F 0.6/ 0.0 G 5.8/ 0.9 6..44 H 5.5/ 2.2 2.50 1 7.0/ 0.8 8.75 J 27.6/ 11.3 2.44 K 0.2/ 0.0 Average- •all zones 10.9/ 2.3 473 Jack mackerel Zone A 41.7/ 12.2 3.48 B 25.3/ 44.8 0.56 c 2.5/ 6.0 0.41 D 9.2/ 34.6 0.27 E 2.5/ 19.2 0.13 F 8.0/169.4 0;04 G 2.3/ 6.1 0.37 H 1.7/ 8.2 0.20 1 5.8/ 17.0 0.34 J 4.0/ M.5 0.34 K 0.0/ 0.0 Average- ■all zones 7.7/ 18.4 0^41 Pacific mackere il Zone A 0.0/ 0.0 B 0.0/ 1.1 oToo C 0.8/ 5.7 0.14 D 2.5/ 3.8 0.65 E 0.5/ 6.4 0.07 F 50.1/ 50.3 0.99 G 2.0/ 2.0 1.00 H 11.1/ 5.9 1.88 1 1.7/ 2.4 0.70 J 2.7/ 8.4 0.32 K 3.2/ 0.0 Average- all zones 2.6/ 5.2 o'.so Pacific sard ine Zona A 9.5/ 4.5 2.11 B 0.6/ 3.9 0.15 C 0.1/ 0.1 1.00 D 0.2/ 0.2 1.00 E 3.0/ 57.4 0.05 F 16.6/ 6.1 2.72 G 0.2/ 1.2 0.16 H 0.2/ 7.0 0.02 1 0.0/i 0.0 J 0.1/ 1.1 0^90 K 0.0/ 0.0 Average— all zones 0.7/ 4.1 0^7 Yellowtail (Note, small number of observaf ions, zone data omitted.) Average- all zones 0.09/0.02 4.50 Pacific barracu( da (Note, small number of observat ions, zone data omitted.) Average— all zones 0.06/0.29 0.20 1012 SQUIRE: PELAGIC MARINE FISHES OFF SOUTHERN AND CENTRAL CALIFORNIA Table 5. — Average weight per school (data from Sep- tember 1962 through December 1966). Species Northern anchovy Pacific sardine Jack mackerel Skipjack tuna Albacore Bluefin tuna Pacific bonito Pacific mackerel Yellowtail White seabass Pacific barracuda Total tons = Avg. No. schools obs. 192,047.5 5,261 - 36.5 5,140.5 194 = 26.5 44,545 1,846 = 24,1 260 14 = 18.6 73 18.2 17.9 4 7X)92 396 = 38,435 2,244 = 17.1 10,948 649 = 16.9 754.5 53 =^ 14.2 234 = 4.9 47 834.5 4.5 = Avg. tons/school 184 index formula. Midpoints were not reduced for Pacific barracuda and yellowtail. Range of ob- served tonnage and X values are shown in Table 6. The following formula was used to calculate annual indexes of apparent abundance, day and night, for each species by zone and the day/night index of annual average apparent abundance for each species. Index of apparent abundance = ^ NiXi + N2X2 + N3X3 + N4X4 Nt where: N1.2.3A — number of block area flights in which the species oc- curred at value Zi,2.3,4. Xi,2,3,4 — tonnage range values. Nt — total number of block area flights in the zone during the year. Day and night indexes of apparent abundance for each zone and the annual average day/night indexes of apparent abundance for all zones are listed in Table 7. DISCUSSION AND SUMMARY A direct, precise measure of total abundance is most desirable for the management of pelagic marine species. However, at the present time and into the foreseeable future, this degree of accuracy in the measurement of total abundance cannot be attained. Therefore, pelagic resource mangement will be required to rely on an indirect measure of total abundance. Some observations on the relation between the index of apparent abundance and changes in estimates of total abundance can be made. For the years 1963 through 1969, either separately or combined, some data are available giving estimates of total abundance, spawning biomass, or indexes of abundance for such species as the northern an- chovy. Pacific mackerel. Pacific sardine, jack mackerel, and Pacific bonito. However, all such estimates were calculated from data obtained from such measures as catch, eff"ort, catch com- position, fecundity, and egg and larval counts. No direct measurements of total abundance were Table 6. — Range of tonnage and tonnage range values (X). Species Observed tonnage X Anchovy 0-400 2 400-1,000 7 1,000-10,000 55 10,000-20,000 150 Pacific bonito 0-50 2.5 50-150 10 150-1,000 57.5 I,OO0-5JOO0 300 Jack mackerel 0-50 2.5 50J3O0 17.5 300-1,000 65.5 1,000-2,000 150 Pacific mackerel 0-20 1 20J100 6 100-250 17.6 250-500 37.5 Pacific sardine Oh 100 5 100-500 30 500-2,000 125 2,000-4,000 300 Pacific barracuda 0-10 5 10^0 20 30-80 55 80-160 120 Yellowtail 0-5 2.5 5-10 7.5 10-30 20 30-60 45 1013 FISHERY BULLETIN: VOL. 70, NO. 3 Table 7. — Annual average indexes of apparent abundance for both day and night observations. cates no flight observations in zone. Indexes given as day/night. Dash (__) indi- Zone 1962 1963 1964 1965 1966 1967 1968 1969 Northern anchovy A 0.51/1.22 10.98/13.28 0.80/7.32 2.51/2.46 ■2.02/ 6.33 1.15/ - 0.28/ - 0.00/ 0.00 B 3.25/3.18 6.17/13.02 4.08/3.08 1.85/5.75 0.29/15.31 0.28/ 0.30 0.73/ _- 0.03/ 0.00 C 0.05/0.99 0.40/ 2.70 0.90/2.85 1.82/8.63 0.27/ 2.79 2.91/ 3.03 0.05/ 1.76 0.87/ 1.93 D 0.00/2.97 1.24/10.02 0.23/4.75 0.48/3.67 0.14/ 6.73 0.63/ 5.12 0.05/ 3.24 0.41/ 1.96 E _-/0.00 0.00/ 1.05 0.00/3.14 0.00/0.12 1.09/ 0.23 3.15/ 0.11 0.18/ 0.12 0.32/ 0.67 F .-/O.OO 3.66/ 1.71 5.75/0.37 0.00/0.58 0.00/ 2.61 0.00/ 0.00 0.00/ 0.60 0.00/ 0.00 G ../1. 00 0.51/ 1.81 0.54/6.21 1.42/3.23 (2.24/ 1.78 2.40/ 8.32 0.55/ 2.11 5.01/ 8.57 H — /— 0.14/ 0.12 0.38/1.00 0.05/1.90 2.28/ 1.43 5.45/ 1.60 0.65/ 0.58 6.35/ 3.26 1 — /-_ 0.14/ 0.00 0.05/0.11 0.02/0.34 0.28/ 2.53 2.87/ 0.71 0.18/ 0.21 0.78/ 0.49 J ._/.. 0.78/ 0.67 0.85/6.08 0.00/0.00 0.45/ 1.55 0.16/20.16 0.52/ 1.11 0.55/ 3.39 < -./.. 0.00/ - 0.00/- 0.00/- 0.00/ 0.22 0.00/ 0.70 2.44/ 0.06 0.00/ 0.00 Average all zones 1.79/1.99 1.64/ 2.99 1-.03/3.90 0.96/4.18 Pacific bonito 0.84/ 3.62 1.78/ 4.30 0.33/ 1.46 1.30/ 4.35 A 0.00/0.00 0.00/ 0.00 0.50/0.00 0.00/0.00 0.04/ 0.00 0.07/ - 0.00/ - 0.00/ 0.00 B O.Ol/O.OO 0.00/ 0.01 0.06/0.01 0.02/0.00 0.43/ 0.00 0.01/ 0.00 0.00/ — 0.00/ 0.00 C 0.62/0.25 3.63/ 1.51 2.40/0.82 2.18/0.40 1.38/ 0.83 0.51/ 1.02 0.36/ 1.36 0.29/ 0.27 D 0.83/0.63 0.65/ 1.88 1.40/0.12 0.85/0.10 1.27/ 0.54 0.38/ 0.80 0.53/ 0.15 0.24/ 0.23 E ../O.OO 0.02/ 1.15 0.50/0.00 6.81/0.00 4.30/ 0.00 0.00/ 0.00 0.00/ 0.02 0.00/ 0.00 F _./0.00 0.00/ 0.00 0.43/0.00 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 G --/O.OO 0.31/ 0.37 0.30/0.00 0.50/0.01 1.19/ 0.16 0.67/ 0.39 0.12/ 0.13 0.12/ 0.16 H „/— 0.85/ 0.22 2.12/0.53 2.06/0.32 0.03/ 0.00 0.00/ 0.02 0.68/ 0.24 0.09/ 0.06 1 ../— 0.00/ 0.45 3.46/0.54 0.40/0.13 0.06/ 0.02 0.12/ 0.00 0.02/ 0.03 0.02/ 0.05 J — /— 3.67/ 0.23 1.10/0.06 1.71/0.46 4.96/ 0.11 3.27/ 0.12 1.70/ 0.69 0.64/ 0.51 K _-/-. 0.13/ - 0.00/- 0.00/— 0.00/ OjOO 45.73/ 0.00 0.26/ 0.00 0.75/ 0.00 Average all zones 0.23/0.19 1.62/ 0.58 1.62/0.28 1.26/0.19 Jack mackerel 1.34/ 0.35 1.35/ 0.34 0.43/ 0.35 0.26/ 0.18 A 0.00/0.08 1.81/ 0.00 8.32/2.89 0.78/ 1.67 4.29/25.72 0.81/ - 0.91/ - 0.03/ 0.00 B 0.02/0.00 3.48/ 1.37 6.65/6.67 1.03/ 1.22 0.74/ 8.78 0.78/ 1.34 0.70/ - 0.00/ 0.00 C 1.59/0.45 0.96/ 0.49 0.83/0.56 0.14/ 0.33 0.06/ 0.03 0.04/ 0.33 0.13/ 0.02 0.11/ 0.17 D Q.9Q/145 1.79/13.77 2.72/S.36 1.44/ 1.40 0.24/ 1.62 0.41/ 0.24 0.21/ 0.20 0.15/ 0.33 E -./O.OO 0.27/ 4.74 0.93/2,31 0.72/ 1.55 0.16/ 7.38 0.00/ 2.15 0.24/ 4.09 0.00/ 4.41 F ..72.91 11.16/39.70 1.63/2.31 0.00/17.41 0.00/27.91 0.00/38.84 10.87/27.77 3.50/13.08 G -/O.OO 1.23/ 1.76 0.64/0.99 0.02/ 0.29 0.00/ 1.05 0.00/ O.IO 0.02/ 0.00 0.13/ 0.26 H .-/.. 0.56/ 1.34 0.01/1.11 0.71/ 0.43 0.04/ 2.18 0.01/ 0.33 0.00/ 0.26 0.00/ 0.59 1 -/- 2.11/ 5.53 0.17/1.86 1.35/ 0.75 0.19/ 2.37 0.04/ 3.27 0.00/ 1.27 0.04/ 1.58 J -/-_ 0.67/ 0.54 0.72/1.51 0.86/ 0.96 0.00/ 0.06 0.00/ 0.03 0.08/ 1.07 0.06/ 0.30 K .-/-. 0.00/ — 0.00/- 0.00/ - 0.00/ o;oo 0.00/ 1.75 0.00/ 1.09 0.00/ 0.00 Average all zones 0.51/0.46 1.41/ 2.98 1.62/2.18 0.71/ 1.36 Pacific mackerel 0.28/ 1.94 0.20/ 1.41 0.30/ 2.25 0.11/ 0.65 A 000/ 0.00 0.00/ 0.00 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ - 0.00/ -. 0.00/ 0.00 B 0.30/ 0.03 0.00/ 1.56 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ - 0.00/ 0.00 C 0.00/ 3.15 0.23/ 0.37 0.03/0.30 0.04/ 0.28 0.01/ 0.35 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 D 0.50/ 0.30 1.41/ 2.20 0.49/0.23 0.05/ 0.08 0.00/ 0.01 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 E — / 3.75 0.43/ 0.67 0.02/0.74 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.03 0.00/ 0.00 F — / 0.00 5.40/ 6.59 3.15/3.62 0.00/ 0.23 0.00/ 2.95 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 G .-/ 18.75 0.79/ 0.46 0.78/0.13 0.00/ 0.44 0.00/ 0.04 0,00/ 0.00 0.00/ 0.00 0.00/ 0.00 H ../ - 2.19/ 1.62 0.14/0.36 0.00/ 0.02 0.00/ 0.02 0.00/ 0.00 0.00/ 0.00 0.00/ 0.03 1 -/ - 0.76/ 0.40 0.01/0.53 0.00/ 0. 1 1 0.00/ 0.09 0.00/ 0.02 0.00/ 0.00 0.00/ 0.04 J ../ - 0.83/ 1.44 0.75/0.29 0.00/ 0.36 0.00/ 0.06 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 K .-/ .. 1.94/ .. 0.00/- 0.00/ .- 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 Average ell zones 0.17/ 1.26 0.79/ 0.91 0.33/0.33 0.01/ 0.19 0.00/ 0.14 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 1014 SQUIRE: PELAGIC MARINE FISHES OFF SOUTHERN AND CENTRAL CALIFORNIA Table 7.- —Continued. Zone 1962 1963 1964 1965 1966 1967 1968 1969 Pacific sardine A 0.00/ 0.00 1.40/ 3.78 1.22/0.05 0.04/ O.OO 0.00/ 0.00 0.00/ -_ 0.00/ ._ 0.00/ 0.00 B 0.05/ 2.13 0.31/ 0.00 0.26/1.63 0.00/ 0.00 0.00/ 0.38 0.00/ 0.00 0.00/ _- 0.00/ 0.00 C 0.04/ 0.00 0.02/ 0.00 0.00/0.05 0.07/ 0.02 0.00/ 0.00 0.00/ 0.03 0.00/ 0.00 0.00/ 0.00 D 0.00/ 0.28 0.07/ 2.36 0.23/0.00 0.00/ O.OO 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 E ../ 12.50 2.22/ 0.83 1.75/7.87 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ O.OO 0.00/ 0.00 F ../ 0.00 0.00/ 0.00 13.47/1.62 0.00/ O.OO 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 G — / 0.00 0.08/ 0.30 0.17/0.13 0.03/ 0.00 0.00/ 0.09 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 H _-/- 0.06/ 0.67 0.00/2.27 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.02 0.00/ O.OO 1 — /— 0.00/ 0.00 O.OO/O.OO 0.00/ 0.00 0.00/ 0.02 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 J ../__ 0.12/ 0.05 0.22/0.01 0.00/ 1.09 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 K __/- 0.00/ __ 0.00/_- 0.00/ __ 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 Average all zones 0.04/ 1.00 0.22/ 0.50 0.27/1.03 0.02/ 0.05 Pacific barracuda 0.00/ 0.04 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 A 0.00/ 0.00 0.00/ 0.00 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ .- 0.00/ -. 0.00/ 0.00 B 0.00/ 0.00 0.00/ 0.00 0.00/0,00 0.00/ 0.00 0.24/ 0.00 0.00/ 0.00 0.00/ - 0.00/ 0.00 C 0.00/ 0.00 0.34/ 0.43 0.45/0.48 0.25/ 0.00 0.02/ 0.00 0.02/ 0.00 0.03/ 0.01 0.04/ 0.00 D 0.00/ 0.00 0.00/ 0.00 0.00/0.01 ■0.00/ 0.00 0.01/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 E _-/ 0.00 0.00/ 0.00 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 F ../ 0.00 0.00/ 0.00 0.00/0.00 10.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 G _./ 0.00 0.00/ 0.1 1 0.00/0.00 0.02/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 H — / ._ 0.00/ 0.00 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.11/ 0.00 0.00/ 0.00 0.00/ 0.00 1 ../ .- 0.00/ 0.00 0.00/0.31 0.40/ 0.00 0.02/ 0.00 0.01/ O.OO 0.00/ 0.00 0.00/ 0.00 J ../ .. 0.15/ 0.00 0.00/0.00 0.00/ O.OO 0.00/ 0.00 0.00/ 0.00 0.00/ O.OO 0.04/ 0.00 K --/ .. 0.04/ .. 0.00/— 0.00/ -_ 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 Average all zones 0.00/ 0.00 0.08/ 0.15 0.08/0.12 0.10/ 0.00 Yellov^rtail 0.03/ 0.00 0.00/ 0.00 0.01/ 0.00 0.02/ 0.00 A 0.00/ 0.00 0.00/ 0.00 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ - 0.00/ .. 0.00/ 0.00 B 0.00/ 0.00 0.00/ 0.00 0.00/0.00 0.00/ 0.00 0.07/ 0.00 0.00/ 0.00 0.00/ _- 0.00/ 0.00 C 0.00/ 0.00 0.03/ 0.04 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 D 0.00/ 0.00 0.36/ 0.00 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ O.OO E -/ 0.00 0.00/ 0.00 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ O.OO F -./ 0.00 0.00/ o.oo 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ O.OO 0.00/ 0.00 G .-/ 0.00 0.00/ 0.00 0.00/0.00 0.01/ 0.00 0.00/ 0.00 0.00/ O.OO 0.00/ 0.00 0.00/ 0.00 H -_/ __ 0.00/ 0.00 0.00/0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 1 ../ - 0.55/ 0.64 0.13/0.04 0.34/ 0.07 0.39/ 0.00 0.09/ 0.00 0.00/ 0.00 0.00/ 0.00 J .-/ .. 0.63/ 0.10 0.00/0.00 0.16/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 K .-/ __ 0.00/ .- 0.00/- 0.00/ _- 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 Average all zones 0.00/ 0.00 0.19/ 0.06 0.01/0.00 0.06/ 0.00 .0.04/ 0.00 0.00/ 0.00 0.00/ 0.00 0.00/ 0.00 made which would provide a precise count of the~ population size. From data for 1962 through 1966 the diurnal/ nocturnal ratios in tonnage observed and in sightings indicate that the northern anchovy, jack mackerel, Pacific mackerel, and Pacific sar- dine were observed more frequently and in great- er quantities at night. Pacific bonito and yellow- tail were observed more frequently and in great- er quantities during the day. However, Pacific barracuda were observed in greater quantity at night but more frequently during the day. Indexes of apparent abundance for day and night observations and variations in total com- mercial catch (Lyles, 1963, 1964, 1965, 1966; Keilman and Allan, 1969) during the years 1963 through 1969 are shown in Figures 4 through 7. The records for 1962 were incomplete (program initiated in September 1962) and were not con- sidered in the discussion of the index. In consideration of the day/night ratios, the indexes reflect the following: Pacific sardine (Figure 4) — The Pacific sar- dine is observed in greater quantity and more frequently during the night; therefore,* the night index should provide a better measure of the sardine's apparent abundance. The night index declined from 0.50 in 1963 to an index of less 1015 FISHERY BULLETIN: VOL. 70, NO. 3 than 0.00 in 1969. Some positive index values of less than 0.005 will be recorded as 0.00. QOO 1963 1964 1965 1966 1967 1968 1969 Figure 4. — Total catch and average index value for the Pacific sardine. night; therefore, the night index should be the best measure of apparent abundance. The night index declined from 2.98 in 1963 to 1.36 in 1965; however, the index level showed an increase in 1968 to 2.25 and a sharp decline in 1969 to 0.65. 300 2O0 1.00 0.00 1963 1964 1965 1966 1967 1968 1969 Pacific mackerel (Figure 5) — Data show that Pacific mackerel are observed in greater fre- quency and abundance during the night; there- fore, the night index should be a better indi- cator of apparent abundance. The night index declined sharply from 0.91 in 1963 to 0.14 in 1966 and continued the decline to less than 0.00 in 1969. 20 z o I 10 u I- < o _J < fe 5 PACIFIC MACKEREL INDEX DAY NIGHT TOTAL CATCH v_... 00 050; 1963 1964 1965 1966 1967 1968 ■—'000 1969 Figure 5. — Total catch, average index values for the Pacific mackerel. Jack mackerel (Figure 6) — Observation of data show that jack mackerel is sighted more frequently and in greater abundance during the Figure 6. — Total catch and average index values for the jack mackerel. Pacific honito (Figure 7) — Bonito are ob- served in greater frequency and abundance dur- ing the day; therefore, the day index better represents any changes in apparent abundance. The day index showed only a slight decline dur- ing the 1963-1967 period, declining from 1.62 to 1.34. However, since 1967 the index has de- clined rapidly to a low in 1969 of 0.18. 300 2 50 10 PACIFIC BONITO ^^^^\^ 9 j \^ - INDEX gfJ,T '-'~ / ^-—-"^''^ O 8 / K TOTAL CATCH / tn 7 1 z 1 o 1 •- 6 1 \ / 1 X 5 o s ^ ^'7 \ k 3 / \ o / \ •- 2 -•^.^ ^ ^ r\ 1 ' 1 1 1 2.00. > ui 1.50 z 00 0.50 000 1963 1964 1965 1966 1967 1968 1969 Figure 7. — Total catch and average index values for the Pacific bonito. 1016 SQUIRE: PELAGIC MARINE FISHES OFF SOUTHERN AND CENTRAL CALIFORNIA Northern anchovy (Figure 8) — Data indicate that the northern anchovy is observed more fre- quently and in abundance during the night; therefore, the night index should better reflect the apparent abundance of this species. The night index increased from 2.99 in 1963 to 4.30 in 1967, declined in 1968 to 1.46, and increased to a high level of 4.35 in 1969. 1963 1964 1965 1966 1967 1966 1969 Figure 8. — Total catch and average index values for the northern anchovy. Yellowtail and barracuda — Indexes have de- clined for both species (see Table 7) ; however, the frequency of observation was low, and no comparisons can be made with trends in abun- dance. The relation between the trend of the apparent abundance index and of the trend of abundance estimates, where available, are discussed for the following species: Pacific sardine — During the period 1963-1969, the Pacific sardine population continued to de- cline to a very low level, and the population is now at a fraction of that calculated for the 1930's and 1940's. Since no recent annual estimates are available (latest is 190,000 tons for 1959), a direct comparison of the estimates of total abundance with the index of apparent abun- dance cannot be made; however, the trend of the index follows closely the downward trend of the commercial catch. Pacific mackerel — The Pacific mackerel fishery in southern California has been the subject of comprehensive research by the California De- partment of Fish and Game for many years. For purposes of comparison between trends of the index and population estimates, the more recent data for Pacific mackerel provides the best source of comparative statistics. The Pacific mackerel catch has declined to a low level in recent years, and the trend of the index follows closely the catch decline (see Fig- ure 5). Population estimates have been calcu- lated by a number of workers. Blunt and Par- rish (1969) summarized the knowledge of this fishery and reported estimates of total spawning biomass of 160 million pounds in 1963. Blunt' (personal communication) computed revised es- timates for California waters using a modifi- cation of the Murphy method (Murphy, 1966). Revised figures indicate a spawning biomass of 64.5 thousand tons in 1962 and 78.5 thousand tons in 1963 reducing to less than 5,000 tons in 1968, an 84% or more, decline from 1963. The night index follows this 847^ decline in estimated spawning biomass with an 89% index decline from 1.26 in 1963 to 0.14 in 1966 and to less than 0.00 in 1968. Jack mackerel — Jack mackerel total abun- dance estimates are derived from egg and larval surveys. Ahlstrom (1968) estimated the adult spawning population in 1951-1954 for the Cal- ifornia area to be between 1.4 and 2.4 million tons and that the resource was "much to moder- ately underutilized." In 1968 he estimated the population level to be approximately the same as was found in the earlier years. The commercial fishery has experienced a sub- stantial decline in catch over the past years and has extended its fishing grounds further offshore. Ahlstrom (1968) indicated the spawning popu- lation is centered in the oceanic waters. Blunt (1969) reported that in this offshore area the population is comprised of mature adults, some reaching the age of 30 years. The young fish remain inshore until 3 to 6 years old and then inhabit the offshore waters where they are out- side the range of the normal fishery. The night aerial index shows a decline in apparent abun- * C. E. Blunt, Jr., California Department of Fish ana Game, Marine Resources Brancsh, 1416 Ninth St., Sac- ramento, CA 95814. 1017 FISHERY BULLETIN: VOL. 70, NO. 3 dance of these inshore younger fish, and this de- cline follows the downward catch trend. Pacific bonito — The Pacific bonito has been abundant in southern California waters since the advent of the "warm years" of 1957-1958. Its northward latitudinal range into southern California waters is influenced by environmental conditions (Radovich, 1963). Figures on total abundance are not available. However, an in- dex of abundance off southern California was de- veloped from party boat catches by Radovich (1963) . It is believed that party boat catches re- flect the bonito's general abundance oflf southern California within certain limits. These limits were not defined and are due to the anglers pref- erence for fishing more desirable species such as barracuda, yellowtail, albacore, etc., when they are available rather than fishing for bonito. Blunt (personal communication; see footnote 3 ) continued Radovich's study and calculated an in- dex based on catch per angler day for the years 1962 through 1968. Values are as follows: 1962, (1.7), 1963 (1.5), 1964 (2.4), 1965 (1.4), 1966 (0.9), 1967 (0.6), 1968 (1.4). The highest party boat index level was in 1964, The index shows an increase for 1968 and diff"ers from the index of apparent abundance; however, this index has shown an overall decrease since 1964. The day index of apparent abundance shows an overall decrease from 1963 through 1969. The Pacific bonito is classed as "much underutilized" by Ahlstrom (1968). Due to economic factors and a decline in catches of jack and Pacific mack- erel, the catch of bonito increased sharply in 1966 and 1967 with only a slight reduction in catch during 1968 and 1969. In contrast to this increase in catch level, the index of apparent abundance has shown a substantial decrease from 1.26 in 1965 to 0.26 in 1969. A consider- able reduction in catch was experienced in 1970 (Lester A. Keilman, personal communication) as the total catch declined to 4,600 tons, about one-half the 1969 catch. Northern anchovy — Studies by Ahlstrom (1968) estimated a total population of 1.8 to 2.3 million tons in 1958, increasing to 4.5 to 5.6 mil- lion tons in 1968. This estimate was based on data from larval counts and shows a population increase of approximately 8.6 times during the period 1951 through 1968. No total population estimates for the successive years 1963 through 1969 are available; however, Ahlstrom does state that the larval counts show that the popu- lation is somewhat variable from year to year and that the population reached a plateau in about 1962. Since annual abundance estimates are not available for the years 1963 through 1969, a direct comparison with the aerial spotter index cannot be made. The northern anchovy has increased substantially in abundance during the past decade and is classed as an underutilized species (Ahlstrom, 1968) . Since it has been sub- jected only to minor fishery, except in 1969, it is generally agreed that the population level con- tinued to be at high level throughout the years 1963-1969. The trend of the night aerial index of apparent abundance shows an overall increase during the years 1963 through 1969. The only significant change in apparent abundance was in 1968 when the index declined sharply; how- ever, in 1969 it again increased to a high level. Wide fluctuations in anchovy relative abun- dance were noted by Wood (1964, see footnote 2) during aerial surveys from 1956 to 1963. Future observations will determine if this paralleling of catch and apparent abundance will continue. Since the survey area covers an area common to the anchovy, trends in the annual index should be of use in evaluating catch variations and re- flect the trend of total abundance in this under- utilized resource. In summary, for the geographical area nor- mally surveyed by the aerial fish spotter, the author believes these data represent a reasonable index of apparent abundance. Like all other measures presently available, the true relation of the index of apparent abundance to total abundance for each species cannot be deter- mined. However, from all data available con- cerning the Pacific mackerel, a species for which considerable and more reliable data on the adult population are available, the trend of the index follows the downward trend of the total abun- dance estimate. The index shows little effect from fluctuations in economic demand, as shown in data for the Pacific bonito. Trends in the abundance level of the Pacific bonito within the survey area are evident before they are reflected 1018 SQUIRE: PELAGIC MARINE FISHES OFF SOUTHERN AND CENTRAL CALIFORNIA in catches, and it appears to be a useful index in the evaluation of catch variations and long- term trends in total abundance in underutilized pelagic surface schooling resources. ACKNOWLEDGMENTS The author wishes to acknowledge the coop- eration and interest of aerial fish spotter pilots who participated in the program from 1962 through 1969; Edward Burden, Leon Burden, Jack Mardesich, Paul Mardesich, Joseph Miles, John Bourgois, Jack Whalen, John O'Conner, Rodger Hillhouse, and Tony Marinkovich. The studies of Br. Oscar E. Sette in the late 1940's must be recognized as one of the earliest applications of scientific aerial fish observations in the continuing search for new methods and techniques for obtaining abundance estimates of pelagic fishery resources. Reynold A. Fredin of the National Marine Fisheries Service, Northwest Fisheries Center, Seattle, Wash., gave valuable assistance through suggestions on statistical processing methods for the aerial spotter data, as did Norman Abramson' of the California Bepartment of Fish and Game, Marine Resource Laboratory, Long Beach, Calif. The processing of the data by the Statistical Section of the California Be- partment of Fish and Game, Marine Resources Laboratory under the direction of Edward Greenhood was appreciated, as were suggestions of the staff of the Statistical Section on original coding of the data. LITERATURE CITED Ahlstrom, E. H. 1968. An evaluation of the fishery resources avail- able to California fishermen. In: D. Gilbert (editor) , The future of the fishing industry of the United States, p. 65-80. Univ. Wash. Publ. Fish., New Ser. 4. Blunt, C. E., Jr. 1969. The jack mackerel {Trachiirus symmetricus) resource of the eastern North Pacific. Calif. Coop. Oceanic Fish. Invest. Rep. 13:45-52. * Now with National Marine Fisheries Service, South- west Region, Terminal Island, Calif. Blunt, C. E., Jr., and R. H. Parrish. 1969. The Pacific mackerel fishery: A summary of biological knowledge and the current status of the resource. Calif. Dep. Fish Game, MRO Ref. 69-7, 25 p. Clark, G. H. 1935. Logs on California trawlers. In The com- mercial fish catch of California for the years 1930-1943, p. 37-43. Calif. Div. Fish Game, Fish Bull. 44. Gushing, D. H., F. Devold, J. Marr, and F. Kristjonsson. 1952. Some modern methods of fish detection. FAO Fish. Bull. 5:95-119. Jones, A. C, and P. N. Sund. 1967. An aircraft and vessel survey of surface tuna schools in the Lesser Antilles. Commer. Fish. Rev. 29(3) :41-45. Keilman, L. a., and T. S. Allen. 1969. California fisheries 1969. U.S. Fish Wildl. Serv., Bur. Commer. Fish., Terminal Island, Calif., 40 p. Levenson, C. 1968. Factors effecting biological observations from the ASWEPS aircraft. U.S. Nav. Oceanogr. Offi., Informal Rep., 68-102, 6 p. Lyles, C. H. 1965. Fishery statistics of the United States, 1963. U.S. Fish Wildl. Serv., Stat. Dig. 57, 522 p. 1966. Fishery statistics of the United States, 1964. U.S. Fish Wildl Serv., Stat. Dig. 58, 541 p. 1967. Fishery statistics of the United States, 1965. U.S. Fish Wildl. Serv., Stat. Dig. 59, 756 p. 1968. Fishery statistics of the United States, 1966. U.S. Fish Wildl. Serv., Stat. Dig. 60, 679 p. Marr, J. C. 1951. On the use of the terms abundance, availa- bility and apparent abundance in fishery biology. Copeia 1951:163-169. Murphy, G. I. 1966. Population biology of the Pacific sardine (Sardinops caerulea). Proc. Calif. Acad. Sci., Ser. 4, 34:1-84. Radovich, J. 1963. Effects of water temperature on the distri- bution of some scombrid fishes along the Pacific coast of North America. In H. Rosa, Jr. (editor), Proceedings of the World Scientific Meeting on the Biology of Tunas and Related Species, La Jolla, Calif., U.S.A., 2 - 14 July 1962. FAO Fish. Rep. 6:1459-1475. Sette, 0. E. 1949. Methods of biological research on pelagic fisheries resources. Indo-Pac. Fish. Counc, Proc. 1st meet, p. 132-138. Squire, J. L., Jr. 1961. Aerial fish spotting in the United States commercial fisheries. Commer. Fish. Rev. 23 (12) :l-7. 1019 AN APPLICATION OF YIELD MODELS TO A CALIFORNIA OCEAN SHRIMP POPULATION Norman J. Abramson' and Patrick K. Tomlinson^ ABSTRACT Two types of yield models were utilized to analyze fishery data from California's northern- most bed of ocean shrimp, Pandalus jordani. The Schaefer form of stock production model was applied to catch and effort data for the years 1954 through 1969. Age-struc- tured catch data for 1955 through 1968 were analyzed by the Murphy method to obtain mortality rates and biomass estimates. Catchability coefficients and a growth curve were also estimated. Attempts to fit spawner-recruit models to estimates obtained from the age-structured catch data were inconclusive; so, age specific mortality and growth estimates were only used to fit a yield-per-recruit model. After comparing the results from the two models, the Schaefer model was deemed most suitable for managing this fishery. The model estimated the maximum sustain- able yield at 2.46 million pounds. A strategy for managing the fishery under a quota system was proposed. The fishery for ocean shrimp, Pandalus jordani, in California has a unique importance despite the fact that it does not rank high among the State's fisheries in terms of pounds landed or value of the landings. This unique importance exists since the fishery developed after discovery of the shrimp beds by the California Department of Fish and Game's exploratory fishing and be- cause it has been under continuous quota control by the California Fish and Game Commission since the fishery's inception in 1952 (Dahlstrom 1961, 1970). It is also the only California com- mercial fishery whose catch is fully regulated under a quota system. This paper is limited to a discussion of the population and fishery which range along the coast from the mouth of the Mad River in Cal- ifornia to the Rogue River in Oregon. This fishery consists primarily of regulated Califor- nia vessels, but there is a small Oregon fleet not covered by California's regulations while fishing ^ California Department of Fish and Game, Opera- tions Research Branch, Long Beach, Calif.; present address: National Marine Fisheries Service, Tiburon Fisheries Laboratory, P.O. Box 98, Tiburon, CA 94920. - California Department of Fish and Game, Opera- tions Research Branch, La Jolla, Calif.; present ad- dress: Inter- American Tropical Tuna Commission, P.O. Box 271, La Jolla, CA 92037. north of the California border. Three smaller populations which occur farther south in Cali- fornia are not considered here. Initially, quotas were set arbitrarily at one- fourth the estimated biomass on the bed. Bio- mass was originally estimated from an exam- ination of commercial catch data and later from research vessel cruise data. In later years, quota recommendations were at least partially directed toward allowing what was deemed an appropri- ate spawning stock to remain at the end of the season. Spawning stock values were based on estimated preseason year class abundance and estimated survival over the fishing season. Estimating procedures which assume commer- cial or research fishing gear catches all shrimp in the water column above the swept path must inherently be negatively biased since escape- ment over, around, and through the gear occurs. The methods just discussed are of this type. A more complete account of the basis for quota recommendations prior to 1969 is found in Dahl- strom (1961, 1970), Dahlstrom and Gotshall (1969), and Gotshall (in press). Over the history of this fishery substantial amounts of data have been collected. Of rel- evance to this paper are catch and effort data, estimated age and sex composition of landings, Manuscript accepted March 1972. FISHERY BULLETIN: VOL. 70, NO. 3, 1972. 1021 FISHERY BULLETINt VOL. 70, NO. 3 and research vessel biomass estimates for 1965 through 1968. These data were used in applying a stock production model and a dynamic pool model. The general characteristics of these two models were discussed by Schaefer and Beverton (1963) under the designations of "Schaefer Ap- proach" and "Beverton-Holt Approach," re- spectively. STOCK PRODUCTION MODEL From an operational viewpoint stock produc- tion models possess the advantage of requiring only catch and effort data, which are usually available at relatively little expense, for their fitting. Another desirable characteristic is the inclusion of density dependent effects, even though they are treated grossly and population response to density is assumed to be instanta- neous. Pella and Tomlinson (1969) discuss the assumptions implicit in the model. The most notable fisheries application of this type model was to yellowfin tuna of the eastern Pacific by Schaefer (1954, 1957), who developed a method for fitting the model to a population in a non- equilibrium state, Pella (1967) examined a number of methods for estimating parameters of the Schaefer model and concluded that a surface searching tech- nique for minimizing the summed, squared de- viations between observed catches and catches predicted by an integrated form of the Schaefer model was generally most satisfactory. Pella and Tomlinson (1969) generalized the Schaefer model to allow asymmetry in the inte- grated form and gave the population growth rate as dt = HP^it)-KP(t)-qfit)Pit), (1) where H, K, m, and q are constants. P{t) rep- resents the population size at time t, f(t) is the fishing intensity at t, q is the catchability co- efficient, and m determines the amount of asym- metry in the equilibrium yield curve. In the Schaefer model, m = 2 and the equilibrium curve is a parabola. The integral of (1) from time 0 to t with / constant is •(K + q/)(l-m)t 1-m J (2) X e- and Pella and Tomlinson (1969) used a numer- ical approximation of C{t) = /: qf{t)P{t)dt (3) for computer calculation of expected catch over the interval. Pella (1967) gives the integrated form of (3) for the Schaefer model A computer program, GENPROD, (Pella and Tomlinson, 1969) for fitting the generalized model to catch and effort data uses the criterion of least squares between observed and predicted catches. Fox (1971) discusses least squares for estimating parameters in (2) and suggests al- ternatives which may be preferable to that used by GENPROD. CATCH AND EFFORT DATA Catch and effort data have been collected since the beginning of the fishery in 1952, but data from the first 2 years of the fishery are not used in this study because there was little effort and low catch-per-effort values indicated that fish- ermen had not fully acquired the skills needed for successfully catching shrimp. California landings were obtained from market receipts, and effort by California vessels was obtained from compulsory logbooks carried by all Cali- fornia trawlers. Oregon landings and effort were supplied by the Oregon Fish Commission (Jack Robinson, Oregon Fish Commission, per- sonal communication). California vessels were restricted to use of beam trawls until otter trawls became legal in 1963. Oregon vessels have used otter trawls since their entry into the fishery in 1960. A correction factor was used to convert California beam trawl effort to otter trawl effort for 1954 to 1962. Fishing power of beam trawls relative to otter trawls was estimated from 40 pairs of catch- per-hour statistics. These paired statistics con- sisted of the average weekly catch-per-hour for 1022 ABRAMSON and TOMLINSON: APPLICATION OF YIELD MODELS each gear within a 10-fm depth interval bounded by a 10-min by 10-min block area. The data were collected during 1960 through 1962 when Oregon vessels were using otter trawls and Cal- ifornia vessels were still restricted to beam trawls. California Department of Fish and Game trawler logbooks and information supplied by the Oregon Fish Commission were the sources of the records (Tom Jow, California Department of Fish and Game, personal communication). With otter trawl taken as the standard gear, the relative log fishing power of beam trawls was computed by Robson's (1966) method ex- cept the two gear types were used in a manner analogous to his treatment of individual ves- sels. If the logarithm of catch-per-hour is nor- mally distributed and the other assumptions of Robson's model hold, then his method produces Bi, an unbiased estimate of relative log fishing power, (3i, for the tth gear. However, exp (Bi) is a biased estimate of exp (/?,). An unbiased estimator for exp {{30 is given by Laurent (1963) as 00 exp (A) = rexp(5i)iri + v^ _t^ i = i (n—k—1)' [v(Bi)V (n—k—1) (n—k+l) ... (n—k + 2j—S)]' (4) X where v (Bi) is an unbiased estimate of the var- iance of Bi with n — k — 1 degrees of freedom. Robson's method provides v{Bi) and our com- puter program for calculating fishing power carries the series expansion in (4) to 15 terms. This computer program is described by Berude and Abramson (1972) and a FORTRAN listing is contained in Abramson (1971). The estimated fishing power of beam trawls relative to otter trawls in the shrimp fishery was 0.71; all beam trawl effort used in this study was adjusted by that factor. FITTING THE PRODUCTION MODEL Usable catch and effort data covered a period of 16 years, each divided into open and closed seasons. Each season was treated as a sep- arate interval in the fitting procedure and thus population estimates were obtained at 32 points in time. Table 1 shows catch, adjusted effort, and time for the series of seasons used to fit the generalized production model. When initially fitting GENPROD to the data the parameters representing optimum effort (Fopt), catchability coefficient (q), maximum catch-per-effort (C/max), and the ratio of initial population to maximum population (r) were un- restricted. Pella and Tomlinson (1969) give these parameters as transformations of those in ( 2 ) . The equation was fitted with the parameter m taking values from 1.4 to 2.6 by increments of 0.2. Results showed that number or distri- bution of data points was not sufficient to de- termine the value of m with any degree of pre- cision and that very small population estimates accompanied by excessively large q values were being obtained. The first problem was handled by setting m = 2, since the symmetric or Schaefer model seemed best in face of the uncertainty. The catchability coefficient was fixed at a value which minimized the sum of the squared deviations be- tween GENPROD'S estimates of Pit) and re- search vessel cruise estimates of population bio- mass at seven time points when both were avail- able. The research vessel biomass estimates were obtained from surveys conducted in the spring and fall of 1965, 1966, and 1967 and the fall of 1968 (Gotshall, in press). Gotshall's catch in weight per standard haul was expanded on an areal basis to provide estimates for the entire survey area; as mentioned previously, these are negatively biased. Based on this pro- cedure, q = 8.5 X 10 ~^ was the best value. The final fit of the Schaefer model was made with GENPROD's computing parameters KK and AT set equal to 5 and 10, respectively. KK is re- lated to the fineness of the surface searching procedure, and N involves the accuracy of the numerical integration used to estimate expected catch. These computing parameters are ex- plained fully in Pella and Tomlinson (1969). GENPROD estimated a maximum equilibrium catch (Cmax) of 2.46 million pounds, an effort level required to obtain this catch under equi- librium conditions (Fopt) of 6,049 otter trawl 1023 FISHERY BULLETIN: VOL. 70. NO. 3 Table 1. — Estimates of Schaefer model parameters, observed catch and effort, predicted population size and catch, population and catch in millions of pounds, effort in thousands of hours. Tima interval Population size end of interval Applied effort Observed catch Predicted Catch/effort Begin End catch Observed Predicted May 54 Sept. 54 Aug. 54 Apr. 55 8.63 9.07 0.206 0.0 0.169 0.0 0.150 0.0 0.819 0.727 May 55 Nov. 55 Oct. 55 Apr. 56 8.82 9.11 0.733 0.0 0.505 0.0 0.557 0.0 0.689 0.760 May 56 Oct. 56 Sept. 56 Apr. 57 8.57 9.00 1.11 0.0 0.896 0.0 0.836 0.0 0.803 0.750 May 57 Nov. 57 Oct. 57 Apr. 58 8.59 8.96 1.05 0.0 0.748 0.0 0.783 0.0 0.713 0.746 May 58 Oct. 58 Sept. 58 Apr. 59 8.18 8.76 1.61 0.0 1.14 0.0 1.17 0.0 0.706 0.726 May 59 Oct. 59 Sept. 59 Mar. 60 7.83 8.45 2.01 0.0 1.69 0.0 1.41 0.0 0.841 0.702 Apr. 60 Nov. 60 Oct. 60 May 61 7.36 8.22 2.90 0.0 1.80 0.0 1.93 0.0 0.623 0.667 Jun. 61 Dec. 61 Nov. 61 Mar. 62 7.75 8.21 1.70 0.0 1.46 0.0 1.15 0.0 0.859 0.677 Apr. 62 Nov. 62 Oct. 62 Mar. 63 6.39 7.23 4.70 0.0 2.98 0.0 2.87 0.0 0.635 0.611 Apr. 63 Nov. 63 Oct. 63 Apr. 64 5.82 6.91 4.85 0.0 2.30 0.0 2.66 0.0 0.475 0.549 Moy 64 Nov. 64 Oct. 64 Apr. 65 6.63 7.56 2.28 0.0 1.20 0.0 1.31 0.0 0.525 0.575 Moy 65 Nov. 65 Oct. 65 Apr. 66 6.17 7.20 4.14 0.0 1.62 0.0 2.39 0.0 0.392 0.578 May 66 Nov. 66 Oct. 66 Feb. 67 6.13 6.85 3.76 0.0 1.61 0.0 2.12 0.0 0.427 0.563 Mar. 67 Nov. 67 Oct. 67 Apr. 68 6.22 7.24 3.71 0.0 2.26 0.0 2.05 0.0 0.608 0.553 May 68 Nov. 68 Oct. 68 Feb. 69 6.72 7.36 2.54 0.0 2.67 0.0 1.50 0.0 1.052 0.592 Mar. 69 Nov. 69 Oct. 69 Apr. 70 6.03 7.09 4.82 0.0 3.11 0.0 2.71 0.0 0.644 0.563 Parameter estimates max 2.46 opt 6.05 p opt 4.79 1 8.5 X 10- opt -E 0.407 U max 0.814 ^max 0.884 9.58 -1.07 X 10-T K -1.03 hours, and an optimum population size (Popt) of 4.79 million pounds. Other parameter estimates, as defined by Pella and Tomlinson, and the com- plete output from the program are shown in Table 1. Figure 1 shows both the expected catch as predicted by the model and the observed catch plotted against time. The fit appears to be gen- erally quite good, although it has worsened dur- ing the most recent 5 years. The statistic R, derived by Pella and Tomlinson to measure the improvement in estimating catch from this mod- el rather than from the mean catch, was 0.91. However, a somewhat spurious R is obtained when intervals with no catch are included in the data. This occurs because the model always predicts a zero catch from zero effort and the arithmetic mean cannot make such a prediction. Recalculating R from only periods when effort was applied yielded 0.75. Figure 2 shows the fitted line (w = 2) for catch per unit effort versus effort in the equi- librium state and the observed catches per hour by year. However, the population should not 1024 ABRAMSON and TOMLINSON: APPLICATION OF YIELD MODELS have been in equilibrium during the period studied since the level of effort fluctuated from year to year. The actual catch exceeded the estimated max- imum equilibrium yield (2.46 million pounds) 30 / / V^ / 25 1 \ *'• / . i V\ #. a. 1 \ ''f-% \l \ Z IS o 13 -1 z 10 ^\l: U : / , / i /.-■ • ' ^ s7 A ^>^ 05 1 • Predicted Coich / 0 , . , , , 1 1962 YEAR 1968 69 1970 during the period 1954 through 1969 only three times (Table 1): 1962, 1968, and 1969. Effort has always been substantially below the esti- mated level which would produce the maximum sustainable yield, A literal interpretation of these results would indicate the population has been underexploited until recently. It is a problem in actual management situa- tions to deduce how well a model such as this represents a population. In years when the ob- served catch-per-unit effort deviates substan- tially from the corresponding expected value, it cannot be determined whether deviations are due to an actual departure from the expected popu- lation size or due to a temporary change in the catchability. In the management strategy which we will discuss later, we are assuming the population size is being predicted correctly by the model and we are essentially ignoring devi- ations between the observed and expected catch insofar as they may represent actual population deviations. Figure 1. — Ocean shrimp catches predicted by GEN- PROD and observed catches for the years 1954 through 1969. (/> a z o Q. I 3 o I tu Q. X o 1 68 X 1 1 1— 1 1,000 - - 800 EJ X X59 -'" X56 X " »\ 55 56 \ 69 X X 62 - 600 - 6^0-- 67 - 400 X 64 66 X 63 \^ X55 \^ - 200 - \.,; 2,000 4,000 6,000 8,000 EFFORT— HOURS 10,000 12,000 Figure 2. — Fitted model (m = 2) for catch-per-hour as a function of hours under equilibrium conditions. DYNAMIC POOL MODEL Catch data by age categories, both in weight and numbers, were utilized to estimate mortal- ity, growth, and recruitment parameters neces- sary in a dynamic pool model. AGE-STRUCTURED CATCH DATA Catches from the population were landed at Eureka and Crescent City, Calif., and Brookings, Oreg. Landings data were obtained mainly from the Pacific Marine Fisheries Commission Data Series (1965-1969). Catches south of the California-Oregon border were recorded in that publication in tables for PMFC Area 96, but those from north of the border were included in, but did not comprise all of, the catch reported from PMFC Area 88. Catches within PMFC Area 88 south of the Rogue River were obtained from the Oregon Fish Commission (Jack Rob- inson, Oregon Fish Commission, personal com- munication). Catches made in the more recent years were obtained from the California Depart- ment of Fish and Game Shellfish Program (Dan- iel Gotshall and Walter Dahlstrom, California 1025 FISHERY BULLETIN: VOL. 70, NO. 3 Department of Fish and Game, personal com- munication). Virtually all catches were made during single day trips. Landings were stratified into port-months, with Eureka-Crescent City as "California" and Brookings as "Oregon." Relative age frequency and weight at age were determined from samples of most port-month catches. Values used for California strata not sampled were either the average of preceding and following strata or the nearest sampled strata of the same season. The Oregon Fish Commission provided values for all Oregon strata. Several methods of drawing samples from within strata were used by California. For all but very recent years, the methods were equiva- lent to assuming a simple random sample of shrimp from within strata. These sampled shrimp were aged by carapace length measure- ments, and the fraction falling into a specific age group determined its relative frequency. In re- cent years a simple random sample of boatloads was assumed drawn, and the length composition of a subsample from each boatload was weighted by the estimated number of shrimp in the load. Estimates by strata, done separately for Oregon and California, were combined to obtain the val- ues in Table 2. The average weight at age was determined by two methods: (1) the aged shrimp were placed into length frequency groups, a length- weight key was used to convert length to weight, and average weight for each age group was calculated; (2) the aged shrimp were weighed and an average weight computed directly for each age group. The study of aged catch data was performed for the 1955 through 1968 sea- sons. All aged shrimp fell into age groups 0, I, II, or III, but the 0 group was rare and omitted from the study. Catch by age category for 2,598 million shrimp (22.88 million pounds) harvested during 1955 through 1968 are listed by month in Table 2. During the first 7 of these years, the fishery was active during 39 months and captured an esti- mated 954 million shrimp, excluding age 0, yield- ing a monthly average of 24.5 million. These shrimp weighed about 8.25 million pounds, av- eraging 212,000 lb. per month of fishing and 0.0086 lb. per shrimp. The fishery was active during 46 months of the second 7 years and caught an estimated 1,644 million shrimp, ex- cluding age 0, for a monthly average of 35.7 mil- lion. These weighed about 14.63 million pounds, averaging 318,000 lb. per month and 0.0089 lb. per shrimp. The relative frequencies in num- bers during the first 7 years were: 0.559 for age I, 0.422 for age II, and 0.019 for age III. During the second 7 years the frequencies were 0.495 for age I, 0.463 for age II, and 0.042 for age III, The reliability of the age frequency val- ues is uncertain due to the aging method. GROWTH CURVE A growth in weight curve was obtained em- pirically by plotting average weights of shrimp by month and age for all seasons 1955 through 1968 (Tables 2 and 3, Figure 3). Dahlstrom (1970) and Gotshall (California Department of 025 CO Q z o 0. X o 010 I I I I I I I I I I I I I I I I I I I I I I I I I 00025 - -kIk! / .;.rf^ Seasons 1955— 1959 Seasons 1960-1968 -Empirical Growth Curve I I I I I I I I I I I I I I I I I I I I I I I I I X < 2 o z _L_L AGE I AGE H AGE M Figure 3. — Ocean shrimp growth in weight by month from sampling commercial landings. Seasons included are 1955 through 1968. 1026 ABRAMSON and TOMLINSON: APPLICATION OF YIELD MODELS Table 2. — Aged catch^ and catch-per effort (C.P.E.) statistics, [Pounds and numbers in thousands.] c.« A^= Relative S^°- Month „^9f„ fre- son group ^^^^^^ Average weight (lb.) C.P.E. numbers Pounds Numbers Spr.. Ar,» Relative 'r; Month ^3- fre- y """^ quency Average weight (lb.) C.P.E. numbers Pounds Numbers 1955 May 1 0.226 0.0055 9.9 4.3 782 1958 July 1 .492 0.0053 29.7 30.9 5842 1 .747 .0118 32.7 30.4 2586 I .496 .0108 29.9 63.3 5890 1 1 .027 .0169 1.2 1.6 93 1 1 .012 .0152 .7 2.2 142 June 1 .408 .0046 26.2 21.0 4531 Aug. 1 .596 .0064 64.8 121.7 18992 1 .576 .0118 37.0 75.3 6397 1 .394 .0115 42.9 144.3 12555 >ll 1 .016 .0172 1.0 3.1 178 1 1 .010 .0161 1.1 5.1 319 July 1 .425 .0055 38.4 30.8 5603 Sept. 1 .786 .0068 70.8 209.9 30649 1 .565 .0122 51.0 90.8 7449 1 .209 .0118 18.8 95.9 8150 II 1 .010 .0172 .9 2.3 132 1 1 .005 .0159 .5 3.1 195 Aug. 1 .378 .0057 28.8 28.0 4897 1959 May 1 .599 .0055 87.2 104.0 18919 1 .603 .0122 46.0 95.3 7812 1 .355 .0112 51.7 126.0 11213 11 1 .019 .0172 1.4 4.2 246 1 1 .046 .0159 6.7 23.1 1453 Sept. 1 .588 .0064 49.9 47.7 7441 June 1 .748 .0062 87.8 268.8 43542 1 .397 .0123 33.7 62.0 5024 1 .240 .0115 28.2 160.6 13971 II 1 .015 .0172 1.3 3.3 190 1 1 .012 .0167 1.4 11.6 699 Oct. 1 .588 .0067 19.3 2.1 309 July 1 .595 .0066 53.1 211.4 31921 I .397 .0125 13.0 2.6 208 1 .382 .0118 34.1 241.1 20494 II 1 .015 .0175 .5 .1 8 1 1 .023 .0161 2.1 19.9 1234 1956 May 1 .342 .0044 45.2 19.7 4479 Aug. 1 .648 .0069 47.9 91.1 13120 1 .608 .0099 80.4 78.8 7963 1 1 .329 .0120 24.3 80.3 6661 II 1 .050 .0156 6.6 10.2 655 11 1 .023 .0164 1.7 7.6 466 Juno 1 .140 .0051 13.1 16.1 3166 Sept. 1 .720 .0073 66.8 207.3 28403 1 .833 .0093 78.2 174.4 18839 ■1 1 .258 .0123 23.9 125.7 10178 II 1 .027 .0159 2.5 9.7 611 II 1 .022 .0164 2.0 14.2 868 July 1 .140 .0051 13.0 13.7 2697 1960 May 1 .601 .0042 45.1 48.2 11464 1 .843 .0097 78.5 157.7 16240 1 1 .382 .0100 28.7 73.2 7287 11 1 .017 .0154 1.6 5.0 327 II 1 .017 .0153 1.3 5.0 324 Aug. 1 .161 .0057 13.0 27.8 4857 Juno 1 .689 .0050 53.2 94.0 18882 I .818 .0110 66.0 271.2 24680 1 1 .289 .0115 22.3 91.5 7920 II 1 .021 .0169 1.7 10.7 634 II 1 .022 .0162 1.7 9.8 603 Sept. 1 .230 .0064 11.4 14.2 2221 July 1 .798 0055 78.5 321.7 58428 I .753 .0115 37.3 83.6 7271 1 1 .192 .0114 18.9 160.5 14058 II 1 .017 .0164 .8 2.7 164 II 1 .010 .0172 1.0 12.6 732 1957 May 1 .366 .0053 29.5 22.0 4159 Aug. 1 .700 .0060 63.4 281.4 47073 1 .629 .0104 50.7 74.4 7147 1 1 .276 .0110 25.0 204.0 18560 II 1 .005 .0167 .4 .9 57 II 1 .024 .0162 2.2 26.1 1614 June 1 .592 .0057 33.4 21.9 3807 Sept. 1 .699 .0066 53.5 210.6 31760 1 .403 .0110 22.7 28.5 2592 1 1 .264 .0122 20.2 146.6 11995 II 1 .005 .0167 .3 .5 G2 II 1 .037 .0168 2.8 28.3 1681 July 1 .592 .0063 33.0 11.2 1785 Oct. 1 .698 .0076 33.3 51.3 6790 I .403 .0116 22.5 14.1 1215 J 1 .280 .0132 13.4 35.9 2724 11 1 .005 .0167 .3 .3 15 II 1 .022 .0179 1.0 3.8 214 Aug. 1 .652 .0070 60.9 229.0 G2515 1961 June 1 .454 .0052 48.8 52.3 10053 I .343 .0122 32.0 208.6 17105 1 1 .531 .0112 57.1 131.5 11759 11 1 .005 .0161 .5 4.0 249 II 1 .015 .0162 1.6 5.4 332 Sept. 1 .597 .0072 40.0 52.4 7235 July 1 .441 .0063 36.1 79.0 12559 1 .386 .0122 25.8 57.0 4678 1 1 .549 .0126 44.9 196.7 15635 II 1 .017 .0161 1.1 3.3 206 II 1 .010 .0170 .8 4.8 285 Oct. 1 .597 .0074 50.9 9.0 1221 Aug. 1 .337 .0069 25.8 118.9 17195 1 .386 .0125 32.9 9.9 790 1 1 .643 .0131 49.2 428.3 32809 11 I .017 .0167 1.4 .6 35 II 1 .020 .0172 1.5 17.6 1021 1958 May I .429 .0048 30.1 41.6 8648 Sept. 1 .400 .0076 31.6 62.0 8152 1 .541 .0101 38.0 110.0 10906 1 1 .574 .0135 45.3 158.1 11698 II 1 .030 .0150 2.1 9.1 60S II 1 .026 .0190 2.1 10.1 530 June 1 .429 .0055 37.2 82.3 14987 Oct. 1 .222 .0073 15.7 23.4 3181 ! .541 .0108 46.9 203.2 18899 1 1 .758 .0146 53.5 158.5 10860 II 1 .030 .0159 2.6 16.6 1048 II 1 .020 .0227 1.4 6.5 287 1027 FISHERY BULLETIN: VOL. 70. NO. 3 Table 2. — Continued. son Aga group Relative fre- quency Averoga weight (lb.) C.P.E. numbers Pounds Numbers ^,nn" Month son Aga group Relative fre- quency Averaga weight (lb.) C.P.E. numbers Pounds Numbers 1961 Nov. 1 .633 0.0066 33.1 4.8 728 1964 Sept. 1 .548 0.0071 29.5 9.7 1359 11 .345 .0121 18.0 4.8 397 II .431 .0128 23.2 13.7 1069 III .022 .0192 1.2 .5 25 III .021 .0179 1.1 .9 52 1962 Apr. 1 .460 .0043 38.6 44.8 10512 Oct. 1 .548 .0076 26.7 17.8 2351 II .465 .0093 39.1 99.0 10626 11 .431 .0131 21.0 24.3 1849 III .075 .0141 6.3 24.1 1714 III .021 .0186 1.0 1.7 90 May 1 .460 .0047 41.8 121.8 25684 1965 May 1 .612 .0049 29.4 98.2 20029 II .465 .0100 42.3 259.1 25964 II .275 .0098 13.2 88.4 9000 III .075 .0148 6.8 62.0 4188 III .113 .0141 5.4 52.3 3698 June 1 .460 .0050 30.9 80.2 15998 June 1 .704 .0048 47.7 168.8 35439 II .465 .0113 31.2 182.9 16172 II .266 .0109 18.0 146.3 13390 III .075 .0180 5.0 46.8 2608 III .030 .0154 2.0 23.3 1510 July 1 .414 .0053 22.4 86.4 16208 July 1 .868 .0056 61.5 456.1 81360 II .537 .0120 29.0 253.3 21024 II .119 .0115 8.4 128.8 11154 III .049 .0179 2.6 34.3 1918 III .013 .0164 .9 19.9 1219 Aug. 1 .560 .0059 32.6 249.2 42011 Aug. 1 .858 .0069 35.6 177.7 25734 II .395 .0129 23.0 382.9 29633 II .126 .0130 5.2 49.1 3779 111 .045 .0231 2.6 78.0 3376 III .016 .0172 .7 8.2 480 Sept. 1 .560 .0068 41.2 281.6 41305 Sept. 1 .865 .0075 30.9 105.2 14112 U .410 .0128 30.2 387.7 30241 II .124 .0132 4.4 26.7 2023 III .030 .0180 2.2 39.7 2213 III .011 .0185 .4 3.3 179 Oct. 1 .550 .0064 51.3 95.1 14976 Oct. 1 .807 .0074 30.6 48.7 6545 II .380 .0136 35.4 140.8 10347 II .176 .0137 6.7 19.5 1427 III .070 .0177 6.5 33.7 1906 III .017 .0211 .6 2.9 138 1963 Apr. 1 .160 .0043 11.2 7.5 1751 1966 May 1 .147 .0050 7.2 11.4 2273 II .770 .0094 54.0 79.3 8427 11 .795 .0093 38.9 114.4 12295 III .070 .0142 4.9 10.9 766 III .058 .0151 2.8 13.5 897 May 1 .162 .0034 10.6 20.7 6028 June 1 .230 .0053 11.1 70.7 13397 II .750 .0092 49.0 257.7 27909 11 .735 .0100 35.6 426.8 42812 III .088 .0177 5.7 57.9 3273 III .035 .0158 1.7 32.2 2039 June 1 .171 .0038 8.9 26.1 6811 July 1 .252 .0062 13.4 94.5 15364 II .730 .0096 37.9 280.4 29077 II .717 .0100 38.1 437.6 43714 III .099 .0155 5.1 61.1 3943 III .031 .0116 1.6 22.0 1890 1963 July 1 .165 .0055 6.8 39.5 7236 Aug. 1 .292 .0064 9.9 33.4 5223 11 .725 .0114 29.7 361.7 31796 II .687 .0119 23.3 146.1 12289 III .110 .0154 4.5 74.4 4824 III .021 .0161 .7 6.1 376 Aug. 1 .274 .0061 11.1 99.6 16226 Sept. 1 .410 .0072 12.6 36.1 5022 II .674 .0125 27.3 497.1 39915 II .586 .0127 18.0 91.4 7178 III .052 .0176 2.1 54.1 3079 III .004 .0181 .1 .9 49 Sept. 1 .289 .0061 12.5 46.8 7687 Oct. 1 .424 .0072 15.8 19.4 2682 II .661 .0119 28.7 209.3 17582 II .567 .0134 21.1 47.9 3587 III .050 .0163 2.2 21.7 1330 III .009 .0206 .3 1.2 57 Oct. 1 .330 .0068 14.4 20.1 2943 1967 Mar. 1 .674 .0042 51.7 10.4 2482 II .660 .0127 28.9 74.8 5886 II .159 .0094 12.2 5.5 586 III .010 .0108 .4 1.0 89 ■III .167 .0142 12.8 8.7 615 1964 May 1 .460 .0059 26.4 65.6 11102 Apr. 1 .730 .0044 60.7 45.6 10444 II .523 .0110 30.1 139.3 12622 II .176 .0101 14.6 25.5 2518 III .017 .0165 1.0 6.8 410 III .094 .0158 7.8 21.2 1345 June 1 .430 .0071 23.5 95.2 13352 May 1 .723 .0048 47.3 11.9 2461 II .550 .0107 30.0 182.3 17078 II .176 .0110 11.5 6.6 599 III .020 .0206 1.1 12.8 621 III .101 .0162 6.6 5.6 344 July 1 .505 .0069 24.1 68.8 9924 June 1 .739 .0055 61.8 294.7 53862 II .474 .0124 22.6 115.9 9315 >ll .225 .0115 18.8 188.9 16399 III .021 .0196 1.0 8.1 413 III .036 .0152 3.0 39.8 2624 Aug. 1 .548 .0081 30.4 183.7 22554 July 1 .804 .0063 79.0 672.5 106859 II .431 .0132 23.9 233.4 17739 II .172 .0120 16.9 273.8 22860 III .021 .0186 1.2 16.0 864 HI .024 .0157 2.4 50.2 3190 1028 ABRAMSON and TOMLINSON: APPLICATION OF YIELD MODELS Table 2. — Continued. Sea- son Month „^9f„ group Relative fre- quency Average weight (lb.) C.P.E. numbers Pounds Numbers Sea- son Month Age group Relative fre- quency Average weight (lb.) C.P.E. numbers Pounds Numbers 1967 Aug. 1 .868 0.0071 61.9 359.7 50317 1968 July .380 0.0065 40.9 154.7 23947 .107 .0131 7.6 81.5 6203 .611 .0120 65.7 460.6 38504 III .025 .0171 1.8 24.7 1449 III .009 .0186 1.0 10.5 567 Sept. 1 .840 .0079 30.0 76.8 9734 Aug. .277 .0066 15.0 46.1 6982 .130 .0138 4.6 20.8 1506 .685 .0120 37.1 206.8 17267 III .030 .0195 1.1 6.8 348 III .038 .0180 2.1 17.2 958 Oct. 1 .784 .0084 24.4 15.1 1807 Sept. .161 .0056 9.4 4.2 761 .171 .0154 5.3 6.1 394 .814 .0108 47.5 41.6 3848 III .045 .0239 1.4 2.5 104 III .025 .0141 1.5 1.7 118 1968 May I .193 .0054 17.1 77.9 14303 Ocl. .164 .0062 10.1 .5 81 .726 .0109 64.4 584.2 53801 .812 .0105 50.2 4.2 402 III .081 .0166 7.2 99.4 6003 III .024 .0142 1.5 2 12 June 1 III .268 .715 .017 .0059 .0112 .0175 46.6 124.3 3.0 153.0 776.9 28.9 26091 69609 1655 1 Catches of 0 age groups are not included. Table 3. — Average weight in pounds by age. From aged catch landed in northern California and southern Ore- gon, 1955-1968. Month Age I Aga II Age III March 0.0038 0.0092 0.0146 April .0043 .0098 .0152 May .0049 .0104 .0158 June .0055 .0110 .0164 July .0060 .0116 .0168 August .0060 .0121 .0174 September .0065 .0127 .0180 October .0070 .0132 .0185 November .0074 .0135 .0190 January .0086 .0140 Fish and Game, personal communication) indi- cated that shrimp grow faster in the open season than during the closed season. Hence, the em- pirical curve was drawn to show seasonal dif- ferences in the growth rate. A more objective fit of the data could be obtained, but it would not alter the results enough to change the con- clusions contained herein. The curve shows relatively constant (linear) growth in weight during the open season, but slower growth during the closed period. The shrimp apparently do not approach an asymp- totic weight prior to reaching maximum age in the fishery, and growth in weight could be de- scribed as linear during the exploited phase. Obviously, there was considerable variation, in- creasing with age. Annual average count per pound for ages I, II, and III combined varied from 94 in 1961 to 142 in 1965 (Figure 4). Monthly values varied from 76 to 155 with an average for all years of 114. B WO«lHL1 VALUES ■ SEASOMIL *ven«GE$ ■ AVCRASE Figure 4. — Average size of ocean shrimp in the landings by month, season, and overall. Because of the variation exhibited by the size at age data, it is possible that significant random or systematic errors are contained in the age composition data and that the subsequent anal- yses of these data will be correspondingly af- fected. 1029 FISHERY BULLETIN: VOL. 70. NO. 3 PARAMETER ESTIMATION WITH THE MURPHY METHOD Reference Values We used the generalized Murphy catch equa- tion (Tomlinson, 1970) to analyze aged Catch data. Gotshall (in press) provides estimates of natural mortality and biomass based upon a fishery independent randomized trawling scheme ( Abramson, 1968) . Since the biomass estimates are inherently negatively biased regardless of the catchability of shrimp and the mortality es- timates may deviate from the population pa- rameters in either direction, we decided to choose a natural mortality which would provide Murphy Method biomass estimates of a magnitude simi- lar to those obtained from the randomized trawling scheme. An annual natural mortality coefficient of M = 1.44 applied to all age groups yielded the ap- propriate biomass estimates. This is within the range of the M values given in Gotshall's (in press) Table 6 and cannot be considered signifi- cantly different from those estimates in view of the sizes of the standard errors shown in his Table 9. Constructing Catch Ratios Ratios of number caught in month t + 1 to number caught in month i were calculated for all age III catches, giving values useful for with- in-season estimation of fishing mortality. To estimate across the closed seasons, the ratio of catch at age III in the first catch-month of season t + 1 to catch at age 11 in the last catch- month of season i was calculated. For example, with 2 seasons and 3 months in each season, the ratios computed by this scheme would be: E(l)=C„,(2)/C„i(l);/2(2)=C„r(3)/C,„(2); R(3)=C„r(4)/C„(3); R(4)-C„i(5)/C,„(4); i2(5) =Ciii(6)/Ciii(5), where the catches used represent monthly catches by age (Table 4) and a closed season exists between months 3 and 4. An additional assumption is that the exploi- tation rate (E) during the last month of each season is equal for ages II and III. Thus in the example, £"11(3) =£7111(3). This assumption is necessary to allow estimation across the closed season. Using these ratios for age III within season and age III to age II between seasons and assum- ing various exploitation rates for the last month of fishing in 1968, it was possible to make nu- merous estimates of the exploitation rates at age III during each month of fishing from 1955 to 1968. The Murphy method with backward calculation (Tomlinson, 1970) was used. The technique is similar to one used by Murphy (1965, 1966), except that Murphy used years instead of months, combined some age groups within years, had no years without catches, and did not treat year classes separately. The data were separated into catches from year classes 1952 through 1967. Using the same hypothetical example as before (Table 4), the catch data can be put in the order Ci(l), Ci(2), Ci(3), 0, C„(4), C„(5), C„(6), 0, C„i(7), Cm (8), Cm (9). The catch ratios are computed as Ci(2)/C,(l), C,(3)/Ci(2), 0, C„(4)/Ci(3), C„(5)/C„(4), C„(6)/C„(5), 0, Cm(7)/C„(6), Cm (8) /Cm (7), Cm (9) /Cm (8). Since these catches all came from the same cohort, the Murphy method can be used to estimate £"1(1), £"1(2),..., £111 (8) , given that Em (9) is known. The previous analysis of age III data gave esti- mates of E at age III during the last month of fishing in each season, and these were used as starting values for backward calculation on each year class from 1952 through 1965. It was nec- essary in estimating E for the 1966 and 1967 year classes to choose values which gave an av- erage population size in 1968 similar to the re- sults obtained from fitting the Schaefer model. Additional Modifications and Assumptions Two additional assumptions fundamental to the results are: (1) ages II and III were ex- ploited at the same rate, on the average, over the entire time period; (2) the catchability co- efficient (q), computed as monthly catch-per- eff'ort in weight divided by estimated average population weight for the combined age groups during the month, was reasonably constant over the entire time period. In order to satisfy these two assumptions, it was necessary to alter some 1030 ABRAMSON and TOMLINSON: APPLICATION OF YIELD MODELS Table 4. — Hypothetical structure of age-structured shrimp catches and exploitation rates as arranged for analysis by the Murphy method. Season Catch- month Catches by ages Exploitation rates by ages Ago 1 Aga II Age III Age 1 Aga II Aga III 1 1 Cjd) Cjj(I) Sll'l' £j(I) V £jjj(I) I 2 Cj(2) Cjj(2) ^ni'2) £■^(2) £jj(2) £jjj(2) 1 3 Closed season Cjj(3) Sll<3) £j(3) £jj(3) £jjj(3) 2 4 Cj(4) Cjj(4) ^II/^) £j(4) £jj(4) £'jjj(4) 2 5 Cj(5) Cjj(5) ^in<5) £j(5) £jj(5) £„j(5) 2 6 Closed Cj(6) season Cjj(6) ^iii'^) £j(6) £jj(6) fijjjW 3 7 Cj(7) Sl'^> ^ni'7) £j(7) £jj(7) 5„j(7) 3 8 Cj(8) Cjj(8) ^ni'S) £j{8) £jl(8) fjjj(8) 3 9 Cj(9) Cjj(9) ^iii'') £j{9) £jj{9) 5jll(9) of the E values from the age III analysis used as starting values for the year class solutions. An additional problem occurred which result- ed in some final changes that were arbitrary and difficult to explain. For some years, especially 1955 through 1959, estimates of population size were quite low and q very high. It was demon- strated that a good transfer from age III to age II across the closed season did not occur for the year classes involved. Therefore, with year classes 1953 through 1958, 1962, 1963, and 1966, the estimation from the last catch-month at age II to the first catch-month at age I disregarded estimates during age III. It is hoped that the fi- nal result justifies these arbitrary decisions. It was also noted from the dots on Figure 3 that a growth curve from the sample data (Table 2) for seasons 1955 through 1959 indicates faster growth during the closed season than during the open season. This seems extremely doubtful in light of other contrary evidence and indicates that the problem was caused by inaccurate aging. Since age III shrimp make up such a small frac- tion of the catch and population biomass, it was not considered to seriously discredit final results. Fishing Mortality Estimates Estimation of monthly instantaneous fishing mortality coefficients, {F) , was accomplished for each age group in each month by applying the Murphy method, as described above, to catches in numbers (Table 2) . Since M = 0.12 was used as monthly instantaneous natural mortality for all months and ages, monthly exploitation rates, E, and monthly survival rates, s, may be obtained from E f\i — e-^f'+o-i2)1/(F + 0.12), and s = e -(F + 0.12) The estimates of F (Table 5) varied consider- ably, but age I was always exploited at a rate lower than ages II and III. During the 7 years, 1955-1961, average estimated F was 0.015 for age I, 0.056 for age II, and 0.057 for age III. In the 7 years, 1962-1968, F{1) = 0.023, F{ll) = 0.116, and F(III) = 0.159. Averages for all 14 years were F (I) = 0.019, F(II) = 0.088, and F(III) = 0.089. Thus, as previously stated for a condition of estimation, ages II and III were exploited at about the same rates. Converting fishing mortality to exploitation (Table 6), it was estimated that the fishery was removing about 5% of ages II and III and 1% of age I each month of fishing. Fishing intensity increased over the years and during 1962-1968 exploitation was nearly double that of 1955-1961 for each age. During the period 1961-1967, July and August were the most important months, followed by May, June, and September, while April and October were of little importance. Average F (Table 10) during these years, for 1031 FISHERY BULLETIN: VOL. 70, NO. 3 Table 5. — Monthly instantaneous fishing mortality coefficients. Year Month Age group Year Month Age group 1 II III 1 II III 1955 May 0.001 0.011 0.006 1963 Apr. 0.002 0.026 0.030 June .003 .032 .013 May .007 .105 .161 July .004 .044 .011 June .009 .140 .270 Aug. .004 .055 .023 July .011 .204 .555 Sept. .007 .042 .021 Aug. .029 .386 .762 Oct. .001 .002 .001 Sept. .016 .267 .812 Mean .0033 .0310 .0125 Oct. Mean .007 .0116 .123 .1787 .100 .3843 1956 May .003 .021 .042 June ' .002 .058 .046 1964 May .017 .071 .021 July .002 .060 .029 June .024 .119 .038 Aug. .005 .112 .066 July .021 .080 .029 Sept. .003 .040 .020 Aug. .055 .198 .072 Mean .0030 .0582 .0406 Sept. Oct. .004 .008 .015 .030 .005 .010 1957 May June .003 .003 .022 .009 .014 .009 Mean .0215 .0855 .0292 July .002 .005 .005 1965 May .017 .069 .298 Aug. .038 .080 .096 Juno .035 .128 .174 Sept. .010 .026 .098 July .098 .137 .190 Oct. .002 .005 .020 Aug. .037 .058 .098 Mean .0097 .0245 .0433 Sept. Oct. .024 .013 .036 .030 .044 .040 1958 May June .006 .012 .039 .081 .071 .154 Mean .0373 .0763 .1407 July .005 .030 .026 1966 May .002 .057 .097 Aug. .020 .077 .068 June .017 .262 .301 Sept. .038 .060 .050 July .023 .423 .456 Mean .0162 .0574 .0738 Aug. Sept. .009 .010 .183 .142 .140 .022 1959 May .016 .037 .090 Oct. .006 .090 .030 June .042 .055 .053 Mean .0112 .1928 .1743 July .036 .098 .113 Aug. .017 .038 .052 1967 Mar. .001 .002 .030 Sept. .043 .070 .120 Apr. .007 .011 .078 Mean .0308 .0596 .0856 May Juna .002 .045 .003 .100 .024 .228 1960 May .003 .030 .019 July .108 .180 .432 June .015 .038 .042 Aug. .063 .062 .325 July .055 .080 .060 Sept. .014 .018 .110 Aug. .052 .133 .168 Oct. .003 .005 .040 Sept. .042 .109 .241 Mean .0304 .0476 .1584 Oct. .010 .030 .040 Mean .0303 .0700 .0783 1968 May June .025 .054 .098 .162 .207 .074 1961 June .009 .048 .021 July .059 .116 .030 July .013 .077 .031 Aug. .020 .064 .060 Aug. .020 .209 .137 Sept. .002 .017 .009 Sept. .011 .098 .090 Oct. .001 .002 .001 Oct. .005 .114 .059 Mean .0268 .0765 .0635 Nov. .001 .005 .006 Mean .0098 .0918 .0590 1962 Apr. May June July Aug. Sept. Oct. .006 .018 .013 .015 .045 .053 .022 .034 .099 .016 .122 .231 .354 .179 .040 .120 .094 .085 .193 .171 .200 Mean .0246 .1564 .1290 1032 ABRAMSON and TOMLINSON; APPLICATION OF YIELD MODELS Table 6. — Monthly exploitation rates. Year Month Age group Year Month Age group 1 II III 1 II III 1955 May 0.0004 0.0107 0.0056 1963 Apr. 0.0017 0.0246 0.0282 June .0029 .0301 .0121 May .0067 .0943 .1402 July .0041 .0408 .0103 June .0086 .1230 .2235 Aug. .0040 .0505 .0218 July .0104 .1744 .4037 Sept. .0069 .0387 .0194 Aug. .0267 .3028 .5062 Oct. .0003 .0019 .0009 Sept. .0147 .2211 .5282 1956 May .0030 .0195 .0385 Oct. .0064 .1090 .0898 June .0024 .0531 .0422 1964 May .0162 .0644 .0199 July .0023 .0547 .0266 June .0223 .1054 .0347 Aug. .0046 .0995 .0599 July .0192 .0730 .0270 Sept. .0024 .0370 .0187 Aug. .0501 .1698 .0656 1957 May June .0031 .0032 .0203 .0085 .0134 .0086 Sept. Oct. .0036 .0070 .0141 .0279 .0048 .0094 July .0017 .0045 .0046 1965 May .0162 .0629 .2433 Aug. .0349 .0720 .0860 June .0328 .1130 .1509 Sept. .0091 .0241 .0883 July .0880 .1206 .1635 Oct. .0017 .0047 .0187 Aug. .0346 .0528 .0878 1958 May June .0058 .0114 .0362 .0736 .0642 .1346 Sept. Oct. > .0222 .0119 .0338 .0279 .0407 .0370 July .0051 .0281 .0240 1966 May .0024 .0524 .0870 Aug. .0188 .0695 .0623 June .0160 .2179 .2458 Sept. .0348 .0549 .0460 July .0210 .3262 .3471 1959 May June July .0146 .0385 .0332 .0346 .0504 .0881 .0812 .0482 .1011 Aug. Sept. Oct. .0082 .0090 .0055 .1577 .1247 .0810 .1230 .0208 .0279 Aug. .0160 .0356 .0482 1967 Mar. .0013 .0022 .0277 Sept. .0396 .0638 .1067 Apr. .0063 .0107 .0703 1960 May June July Aug. Sept. Oct. .0076 .0141 .0501 .0481 .0385 .0097 .0277 .0350 .0727 .1172 .0976 .0279 .0181 .0387 .0552 .1457 .2024 .0370 May Juno July Aug. Sept. Oct. .0017 .0414 .0968 .0573 .0133 .0028 .0029 .0895 .1554 .0569 .0166 .0050 .0219 .1929 .3323 .2622 .0983 .0370 1961 June July Aug. Sept. Oct. Nov. .0083 .0119 .0185 .0101 .0045 .0012 .0442 .0696 .1778 .0880 .1016 .0047 .0290 .0290 .1207 .0810 .0541 .0056 1968 May June July Aug. Sept. Oct. .0233 .0492 .0538 .0187 .0023 .0003 .0876 .1409 .1033 .0587 .0157 .0019 .1767 .0676 .0281 .0552 .0081 .0009 1962 Apr. May June July Aug. Sept. Oct. .0061 .0171 .0121 .0140 .0416 .0483 .0208 .0311 .0885 .0686 .1084 .1946 .2820 -1550 .0371 .1064 .0842 .0767 .1657 .1485 .1712 June to September, was 0.16 for age II, 0.21 for age III, and only 0.03 for age I. Biomass Estimation The Murphy method produces estimates of population size in numbers at the beginning of each catch interval. The present study also re- quired estimates of biomass. Murphy (1966) stated he computed biomass by dividing the catch in weight by the appropriate estimate of E from his analysis of numbers in the catch. This would result in a positively biased biomass estimate, since it is equivalent to multiplying the number alive at the beginning of a catch interval by the average weight during the in- terval. Two ways of computing the correct estimates of biomass utilizing Murphy's method to esti- mate numbers are possible; 1033 FISHERY BULLETIN: VOL. 70, NO. 3 (1) multiply the estimated average weight at the beginning of each interval by the number estimated for the population by the Murphy method. That is, P*ij = Cyu'*ii/.&ij = estimated biomass to be- gin interval j, age i, w*ij = estimated average weight for age i at beginning of interval j, Cij = estimated number of age i caught dur- ing interval j. (2) Multiply the average number of age i alive during interval j by the average weight of age i individuals during this interval. That is, Pij = NijiVij = estimated average biomass of age i, during j. Wii = average weight in the catch of age i, ^ during interval j. Nij = average number alive during interval j of age i. For this study, the second method was used with average population numbers, Na, being given by Nii = Nii{l-e-Yu)/itiZ<}) Nii = Cii/Eij, ti = fraction of year elapsed during inter- ^ val j; tj — 1/12 for all intervals. Zij = total annual instantaneous mortality coefficient during j. Total population biomass for ages I through III was computed as Pj = V NijWij — average biomass available i = l during interval j, and the catchability coefficient from qj = ^CijWij/iPjf }); fi is effort expended during interval j. Estimates of within-season monthly popula- tion biomass varied from a high of 12.0 million pounds in May 1955 to a low of 3.4 million pounds in October 1964 (Table 7) . Population changes estimated by the Murphy method follow trends estimated by the Schaefer model (Figure 5), except Schaefer model estimates exhibit con- siderably less within season change. This dif- ference in range of within season change was caused by the different ways in which the two models treat growth and recruitment. The Schaefer model assumes a continuous process for combined growth and recruitment, whereas the Murphy method treats growth as continuous (Figure 3) and recruitment as instantaneous (Table 7). Estimates of monthly catchability (q) (Table 7) had extreme variation and showed an average within season increase (Figure 6). However, the within season changes were inconsistent and obscured by the variation. Monthly estimates of q varied from 21.3 x 10"^ in June 1968 to 3.8 X 10-5 in May 1955. Yearly averages had less variation and appeared to show no long-term trend (Figure 7). Average q over all months was about 9.0 X 10"^ which agreed closely with the value 8.5 x 10-^ used for the Schaefer model. Spawning Biomass and Recruitment Female spawning biomass consisted of all ages II and III shrimp plus some fraction of age I shrimp. Some data from commercial catch samples on the fraction of age I shrimp func- ~1— MURPHY METHOD ^ — SCHAEFER METHOD 55 56 57 58 59 60 61 62 63 6.1 65 66 67 SB 69 70 YEAR Figure 5. — Comparison of annual maximum and min- imum population sizes as estimated by the Schaefer model and the Murphy method. 1084 ABRAMSON and TOMLINSON: APPLICATION OF YIELD MODELS Table 7.- — Ocean shrimp ] population bioma; ss in tl tiousands 3 of poun( is by age j ind mo nth. Year Month Ages Total Est. 9X108 Year Month Ages Total Est. 1 II III 1 II III 9X106 1955 May 9,100 2,670 265 12,035 38 1963 Apr. 4,101 3,000 358 7,459 84 June 6,788 2,318 237 9,343 61 May 2,892 2,448 360 5,700 104 July 7,119 2,051 208 9,378 91 June 2,830 2,008 227 5,065 95 Aug. 6,538 1,731 181 8,450 89 July 3,546 1,772 134 5,452 81 Sept. 6,467 1,481 157 8,105 51 Aug. 3,466 1,289 71 4,826 92 Oct. 5,942 1,300 140 7,382 41 Sept. 2,979 785 27 3,791 120 1956 May 6,265 3,770 246 10,281 107 Oct. 2,939 609 10 3,558 132 June 6,385 3,008 212 9,605 86 1964 May 3,790 1,970 317 6,077 83 July 5,649 2,637 175 8,461 101 June 3,972 1,538 341 5,851 87 Aug. 5,619 2,432 164 8,215 101 July 3,349 1,438 278 5,065 92 Sept. 5,570 2,089 134 7,793 66 Aug. 3,363 1,177 222 4,762 123 1957 May June 6,701 6,434 3,425 3,155 67 58 10,193 9,647 68 46 Sept. Oct. 2,542 2,378 908 810 183 167 3,633 3,355 145 148 July 6,229 2,940 51 9,220 51 1965 May 5,676 1,281 176 7,133 49 Aug. 6,067 2,625 42 8,734 95 June 4,764 1,147 134 6,045 75 Sept. 5,405 2,207 34 7,646 82 July 4,657 942 105 5,704 80 Oct. 4,872 1,975 29 6,876 118 Aug. 4,751 852 84 5,687 57 1958 May June 6,710 6,739 2,811 2,501 129 108 9,650 9,348 58 80 Sept. Oct. 4,409 3,831 732 651 75 73 5,216 4,555 57 73 July 5,704 2,096 84 7,884 62 1966 May 4,481 1,999 140 6,620 67 Aug. 6,053 1,885 75 8,013 116 June 4,131 1,628 107 5,866 75 Sept. 5,574 1,598 62 7,234 99 July 4,191 1,035 48 5,274 92 1959 May June July 6,650 6,441 5,895 3,372 2,922 2,459 256 222 176 10,278 9,585 8,530 113 93 92 Aug. Sept. Oct. 3,804 3,754 3,332 799 645 533 43 40 39 4,646 4,439 3,904 76 73 103 Aug. 5,337 2,084 146 7,567 86 1967 Mar. 7,401 2,371 293 10,065 51 Sept. 4,830 2,795 119 6,744 121 Apr. 6,853 2,246 274 9,373 57 1960 May June July Aug. Sept. Oct. 5,987 6,217 5,894 5,378 5,048 4,967 2,453 2,419 2,001 2,537 1,343 1,196 255 276 209 155 117 96 8,695 8,912 8,104 7,070 6,508 6,259 57 62 82 97 100 71 May June July Aug. Sept. Oct. 6,708 6,566 6,212 5,741 5,407 5,049 2,154 1,895 1,522 1,309 1,172 1,148 236 174 116 76 62 62 9,098 8,635 7,850 7,126 6,641 6,259 51 70 94 80 48 51 1961 June July Aug. Sept. Oct. Nov. 5,875 6,234 5,983 5,749 4,885 3,893 2,737 2,566 2,054 1,614 1,391 960 172 155 128 112 110 80 8,784 8,955 8,165 7,475 6,386 4,933 105 90 104 119 145 93 1968 May June July Aug. Sept. Oct. 3,105 2,853 2,636 2,295 1,694 1,667 5,991 4,804 3,971 3,219 2,474 2,116 480 389 347 286 192 170 9,476 8,046 6,954 5,800 4,360 3,953 95 213 154 100 134 155 1962 Apr. May June July Aug. Sept. Oct. 6,888 6,714 6,197 5,764 5,519 5,358 4,265 2,956 2,631 2,424 2,076 1,660 1,095 785 601 518 500 404 404 232 168 10,445 9,863 9,121 8,244 7,583 6,685 5,218 59 73 66 63 73 106 177 tioning as females were made available from unpublished sources, but a good method for pre- dicting the fraction of age I shrimp that would function as females was not found. Thus^ a simple mean was computed from the data avail- able for years 1957 through 1967 (Table 8) . It is assumed that this mean proportion (0.33) pre- dicts the fraction of the biomass of age I shrimp alive in September that will be females and that the sum of the September biomass of ages II and III, plus the fraction of age I functioning as fe- males in September, is directly proportional to spawning biomass during the spawning season. Recruitment was defined as the number of age I shrimp alive on May 1 of each year. Thus, the female biomass in September of season i is proportional to the biomass which will spawn sometime prior to May of season i+1 and the progeny of this spawning will be recruited to the fishery at the beginning of season i + 2. Two 1035 FISHERY BULLETIN: VOL. 70, NO. 3 10 ^ ' 1 Q-Avefoge 0-volues not used ■ e " - - , . ^ ' 1/^^^ '"'^ 0 e 1 ! ! - J J : L . . . I 1 1 L Figure 6. — Within-season changes in catchability coef- ficient (g). Line connects seasonal mean values. Circled points were not used to compute means. I I 1 1 ' — o o Monthly Volues Igl- • ■• Seosonol Averoges Average I (XIO'l n \ T" MAY MAY MAY MAY MAY MAY MAY MAY MAY MAY MAY MAY MAY MAY 55 5€ 57 58 59 60 61 62 63 64 65 66 67 68 MONTH AND YEAR Figure 7. — Monthly catchability coefficients (q). Dashed line shows 1955-1968 mean. Table 8. — Estimated percentage of numbers of shrimp functioning as spawners at age I. Yeor Percent Year Percent 1957 43 1963 21 1958 17 1964 36 1959 36 1965 51 I960 42 1966 24 1961 11 1967 54 1962 30 Mean 33 models for predicting recruitment from popu- lation biomass were tried. Model I : Ri+2 = aSi e-^^i ; Model II: Ri+2 = cSi e-^Pi+i where Ri + 2 = Si = Pi+i = number of age I shrimp on May 1, season i + 2. average biomass of functioning fe- males during September of sea- son i. average total biomass (ages I, II, and III) during September of season a, b, c, and d are constants. Model I assumes the number of eggs produced is proportional to spawning biomass and that survival from egg to recruitment is influenced by this same spawning biomass. Model II as- sumes the number of eggs is proportional to spawning biomass and that survival from egg to recruit is a function of average biomass com- peting for the population space. September of season i + 1 was selected for Model II because this seemed likely to be proportional to the av- erage biomass encountered by age 0 shrimp, and data were available for all Septembers. May 1 was selected for recruitment since most seasons opened on this date. Both models of recruitment were fitted by using transformations and a linear model (Paul- ik and Gales, 1965) . The transformed equations are: Model I : \oge{Ri+2/Si) = loge(a) — bSr, Model II: log. (i?, + 2/5;) = loge(c) — dPi+i. Estimates of recruitment by the Murphy method varied from a high of 1.5 billion shrimp on May 1, 1962, to a low of 0.6 billion on May 1, 1968. Spawning stocks producing recruitment varied from 4.5 million pounds in September 1959 to 1.8 million pounds in September 1963 (Table 9). The range in recruitment observed at any giv- en spawning stock size was very large relative to the range in size of spawning stock, and the fitting of Model I did not result in a meaningful 1036 ABRAMSON and TOMLINSON: APPLICATION OF YIELD MODELS Table 9. — Recruits vs. spawners and population biomass. Year (i) ^^i + 2 '^ 3P l+l 1955 1,341.6 3,772 7,793 1956 1,491.0 4,061 7,646 1957 1,295.8 4,025 7,234 1958 1,508.4 3,499 6,744 1959 ^1,365.6 4,508 6,508 I960 1,502.0 3,126 7,475 1961 899.7 3,623 6,685 1962 685.3 3,095 3,791 1963 1,236.4 1,795 3,633 1964 947.1 1,930 5,216 1965 1,447.6 2,262 4,439 1966 613.9 1,924 6,641 ^ R.,„ = Recruits in millions on May 1 of year i+2. 2 S. = Spawners in thousands of pounds during Seotember of year i. * P. = Population in thousands of pounds during September of year i-f-1- * Estimated from June. relationship (Figure 8). Model II, which con- siders the effect of the population competing with the prerecruits, did not account for the vari- ation in recruitment either. Consequently, a realistic spawner-recruit relationship could not be determined from the available data. Yield per Recruit Because a well-defined spawner-recruit rela- tionship could not be determined the use of a self-generating model of the dynamic pool type, such as Walters (1969), is not feasible. We can, however, utilize the age-structured catch data to examine this type of model under the assump- tion that recruitment is constant. We feel that the greatest confidence can be placed in the estimates of instantaneous fishing mortality (Fa) for 1961 through 1967 (Table 5) . For this reason, these values were combined to yield average monthly values (Fu) . The aver- ages were computed as simple arithmetic means to give vectors of average fishing mortality by month and age for April through October (Table 10), and allow for computations of yield per re- cruit. Yield per million recruits was computed by step-wise integration (Ricker, 1958; Paulik and Baylifl[", 1967). For a season of / months, a year class would be exposed to fishing for n = SI months and protected for 3(12 — I) months. This would give a total lifetime after 5 06 SPAWNING FEMALES (Millions of Pounds) Figure 8. — Number of recruits on May 1 of year i+2 produced by spawning biomass of September, year i. Smooth curve represents Model I of the text. recruitment of L = 36 months. The yield can be expressed as L 3 12 Y = X y^^ = XX ^^^^'' ^ ^ 12(t-l) +3 k = l i = l 3 = 1 where Wij — average weight taken from the empir- ical growth curve, Cii — L^ Eij = number caught in month j of year i, Eij = Fij(l — e~^k)/Zk = monthly exploita- tion rate in month j of year i Z^ = (Fij + M) = total monthly instan- taneous mortality (note that Z was previously used Table 10. — Mean monthly instantaneous fishing mortal- ity coefficients, F^j, by age group. Month Age group 1 II III April 0.005 0.024 0.049 May .011 .067 .120 June .022 .123 .162 July .041 .175 .254 August .037 .190 .247 September .019 .133 .179 October .009 .082 .068 1037 FISHERY BULLETIN: VOL. 70, NO. 3 to represent the annual mortality coefficient) , fc-i L^ = R exp \ —^ Zh 1 = number sur- h=l Zh R M vivors to begin interval k, = M during months closed to fishing. = number of recruits = 1,000,000. = 0.12. The yields at various levels of fishing intensity were predicted by multiplying all Fa by a con- stant equal to the intensity change desired and recalculating catches for all months. Estimates of expected yield in numbers and expected av- erage weight per shrimp were also provided by this procedure. By setting appropriate values of F = 0, yields for various seasons and entry ages were computed. It M — 0.12 is the monthly instantaneous mortality coefficient and if growth in weight at age is taken from the empirical growth curve (Table 3), a year class of shrimp that is not fished will reach its maximum biomass during the period July to August as age I. The bio- mass will then decline rapidly and by July to August as age II it will be about one-half the maximum. The estimated yield per recruit for the period 1961 to 1967 was 0.00165 lb. per shrimp. Since the average annual catch during that 7-year pe- riod was 1.918 million pounds, it would have required an average of 1.162 billion recruits on April 1 to support the catch. The Murphy meth- od estimates an average recruitment on May 1 of about 1.155 billion. Thus, the analysis of yield per recruit is in good agreement with the Murphy method results. Given 1.155 billion recruits on April 1, it would require a yield per recruit of 0.00216 lb. per shrimp to obtain a total harvest approxi- mately equal to the maximum sustainable yield estimated from the Schaefer model. To have obtained a yield-per-recruit of 0.00216 during those 7 years would have required an increase in fishing mortality of about 75% (Figure 9). This additional yield could not have been ob- tained by shortening the season or changing age at recruitment unless a substantial increase in fishing mortality accompanied the changes (Fig- ures 9 and 10) . With the distribution of fishing effort observed during 1961-1967, the average total monthly instantaneous fishing mortality (^ 2 Fij) operating against a year class during 3 seasons was estimated to be 2.0176. While maintaining total fishing mortality at 2.0176, AUG. JULY AGE I JUNE MAY ^APR. >- IT Z ^MAR. FEB. CLOSED JAN. 5 DEC. UJ < Lnov. pOCT. SEPT. AUG. AGE I JULY JUNE MAY — APR. — 1 — r 500 " / TOO / T 1 1 1 900 100 T — r r __I300 T "1 1 1500 ' - // / ■^"^^^^ ^ 1700 - / / / / / y^ "^ / / / ^.^ 1800 - - - / , / ) J ^^^_^ 1900 - / / / / y\^ ^^^^^S-^^ ^_^^,_— - 2000 / / / / /^^^c^^^^^^ ^ 2200 // / // y'y^ ^^__.,-— 2400 - 7 1 /// / /^ ^,^ 2600 ^i. / / 1 1 1 (-1 1 1 1,5 2.0 2.5 3.0 MULTIPLIER OF F Figure 9. — Yield in pounds per million recruits as a function of age at entry into the fishery and fishing mortality. Fixed population parameters used were 1961- 1967 means. < IS z — MAY 15 o o to < APRIL 15 ~i 1 1 1 1 1—1 1 1 1 1 1 1 1 1 r 500 1200 1500 1700 1900 2100 2300 2400 2500 2600 2700 J I I L J I I I I I I I I 05 10 15 20 25 30 35 40 MULTIPLIER OF F Figure 10. — Yield in pounds per million recruits as a function of season opening date and monthly fishing mortality coefficient. October 31 season closing date assumed and fixed population parameters used were 1961-1967 means. 1038 ABRAMSON and TOMLINSON: APPLICATION OF YIELD MODFXS the annual yield could theoretically be increased to about 2.8 million pounds by shifting fishing mortality so that I's and II's suffered equal rates. This would involve a 75% reduction in fishing mortality at ages II and III and assumes that the population with the new age structure would continue to produce 1.155 billion recruits. Such a change would also produce a reduction of 26% in the average size of shrimp in the catch and pose the problem of how the mortality pattern could be so altered. INTER-MODEL COMPARISONS Because we were unable to determine a spawn- er-recruit relationship and produce a self-gen- erating form of the dynamic pool model, a realistic comparison of the results from the two types of models is not possible. In addition, the yield-per-recruit model treats natural mortality and growth parameters as constant while in the Schaefer model they are components of density dependent terms. It is of interest to note that the biomass esti- mates obtained from the age-structured catch data by the Murphy method are in general agree- ment with the corresponding estimates of the Schaefer model. Although this does not compare the yield-per-recruit and Schaefer models, we feel it indicates some support of the Schaefer model from a semi-independent source. Another point of agreement between the yield-per-recruit and Schaefer models was that, given the average recruitment over the 1961-67 period, the former required a 75% increase in fishing mortality to produce the Schaefer model's maximum sustain- able yield while the average annual effort ex- pended during that period would require a 68% increase to reach the optimum effort level of the Schaefer model. However, as can be seen from Figure 9, maximum yield-per-recruit is predict- ed to occur at a much higher effort level under the previously mentioned assumption of constant parameters. It seems clear from the foregoing discussion of results relative to the two models that man- agement procedures should be based on the Schaefer model at the present time. PROPOSED MANAGEMENT STRATEGY Fitting an equation such as the Schaefer model to a set of actual catch and effort data may be viewed as merely an interesting exercise unless one has to make actual management recommen- dations based upon the results. Then the situ- ation becomes somewhat sticky. It is obvious that a simple deterministic model such as Schaef- er's will not precisely describe the dynamics of a fish population. At best, thei'e will be fluctu- ations in recruitment, growth, and catchability which will cause some consternation to the man- ager attempting to use such a model. In the case of the shrimp fishery, the manage- ment strategy we propose treats the Schaefer model estimates as exactly correct, responds to indicated deviations from the optimum popula- tion size in a relatively arbitrary but conserva- tive manner, and integrates the Oregon and California fishing. This conservative strategy- attempts a gradual reduction in the biomass when the model estimates it to be above Popt and a rapid increase in the stock size when it is es- timated to be below Po„t. To formulate this pro- cedure, let Q be the catch quota (California + Oregon) and CeiP) = HP- — KP be the equi- librium yield obtainable from a population of size P. With P{t) the population when the next fishing season commences. Q Pit) - P opt + Ce ( Pit) + Popt )■■ Pit) > Popt, 0 = Pit) — Popt + CelPit)]] Pit) < Popt. When the model predicts the stock is in the sur- plus condition we are, then, proposing to harvest one-half of this surplus plus the predicted sus- tainable yield at the point midway between Pit) and Popt. A predicted stock deficit evokes a procedure which harvests the sustainable yield at Pit) minus the amount by which the stock falls short of Popt. For example, the 1970 Cal- ifornia shrimp quota of 3.4 million pounds was set by the above method with Pit) — Popt = 1039 FISHERY BULLETIN: VOL. 70. NO. 3 7.1 _ 4.8 = 2.3 and Ce (5.9) = 2.3 for a rec- ommended yield of 3.4 million pounds (Table 1). It was assumed the Oregon fleet's catch from Oregon waters would be negligible. A more radical strategy such as harvesting all of the surplus stock could be employed, but the attendant risks would be higher. These risks would include a possible disturbance of whatever stability exists in the population, par- ticularly with reference to age structure. It might also be argued that the observed catch- per-effort should be used to adjust P(t) before making the quota calculation described above. Here again, a substantial risk would be involved if the observed catch-per-effort were much high- er than the expected since with our methods it could not be determined whether such an anom- aly was due to abnormal catchability or to a real increase in the stock size. The quota-setting procedure we recommended above does res])ond in a limited way to a higher than expected catch. Since the fishery operates under a quota, a catch- per-eflFort which is higher than expected will result in the quota being filled with a lower than expected amount of eflfort and usually in a short- er time period. An examination of (2) shows that this will increase P(t) and thus result in a larger quota for the season beginning at time t. This fishery must be carefully followed in the future to observe how well the model based up- on current parameters describes the observed catch and effort pattern. An equation such as this which is fitted to data from only 16 years cannot be considered definitive from a statistical estimation viewpoint and, in addition, there is a chance the population parameters will actually be changing. For example, one cannot avoid speculating about the effect of the large Pacific coast hake fishery on the shrimp natural mor- tality rate. Since hake may be a substantial predator upon shrimp (Gotshall, 1969), a reduc- tion of the hake population due to a large fishery might increase the abundance of ocean shrimp. Beyond the technical management problems which we have discussed at length, there is the institutional problem of a single state attempting to manage an interstate fishery. While the catch from Oregon waters by Oregon-based vessels has usually been so small that it affects the popula- tion negligibly, at times it has been substantial. A sustained change in conditions could nullify the effect of California's quota mechanism. ACKNOWLEDGMENTS Many biologists of the Department of Fish and Game collected the data used in this paper. Daniel W. Gotshall and Walter A. Dahlstrom were of special help to us in obtaining the data and in freely passing on to us their knowledge of this shrimp fishery and of shrimp life history. We also wish to thank Catherine L, Berude for programming and computing assistance. LITERATURE CITED Abramson, N. J. 1968. A probability sea survey plan for estimating relative abundance of ocean shrimp. Calif. Fish Game 54:257-269. Abramson, N. J. (compiler). 1971. Computer programs for fish stock assess- ment. FAO Fish. Tech. Pap. 101, [149 p.] Berude, C. L., and N. J. Abramson. 1972. Relative fishing power, CDC 6600, FOR- TRAN IV. Trans. Am. Fish. Soc. 101-133. Dahlstrom, W. A. 1961. The California ocean shrimp fishery. Pac. Mar. Fish. Comm., Bull. 5:17-23. 1970. Synopsis of biological data on the ocean shrimp Pandalus jordani Rathbun, 1902. In M. N. Mistakidis (editor). Proceedings of the world scientific conference on the biology and culture of shrimps and prawns, Mexico City, Mexico, 12- 14 June 1967, p. 1377-1416. FAO Fish. Rep. 57. Dahlstrom, W. A., and D. W. Gotshall. 1969. Will the shrimp boats keep a comin'? Out- door Calif. 30(3): 1-4. Fox, W. W., Jr. 1971. Random variability and parameter estima- tion for the generalized production model. Fish. Bull., U.S. 69:569-580. Gotshall, D. W. 1969. The use of predator food habits in estimating relative abundance of the ocean shrimp, Pandalus jordani Rathbun. In M. N. Mistakidis (editor). Proceedings of the world scientific conference on the biology and culture of shrimps and prawns, Mexico City, Mexico, 12-14 June 1967, p. 667-685. FAO Fish. Rep. 57. In press. Estimates of population size, mortality rates and growth rates of northern California ocean shrimp, Pandalus jordani 1965 - 1968. Calif. Dep. Fish Game, Fish Bull. 155. 1040 ABRAMSON and TOMLINSON : APPLICATION OF YIELD MODELS Laurent, A. G. 1963. Lognormal distribution and the translation method: Description and estimation problems. J. Am. Stat. Assoc. 58:231-235. Murphy, G. I. 1965. A solution of the catch equation. J. Fish. Res. Board Can. 22:191-202. 1966. Population biology of the Pacific sardine {Sardinops caerulea) . Proc. Calif. Acad. Sci., Ser. 4, 34:1-84. Pacific Marine Fisheries Commission. 1965-1969. Data series: Crab and shrimp section. Portland, Oreg. Paulik, G. J., "and W. H. Bayliff. 1967. A generalized computer program for the Ricker model of equilibrium yield per recruit- ment. J. Fish. Res. Board Can. 24:249-259. Paulik, G. J., and L. E. Gales. 1965. Weighted linear regression for two variables, IBM 709, FORTRAN II. Trans. Am. Fish. Soc. 94:196. Pella, J. J. 1967. A study of methods to estimate the Schaefer model parameters with special reference to the yellowfin tuna fishery in the eastern tropical Pacific Ocean. Ph.D. Thesis, Univ. Washington, Seattle, 155 p. Pella, J. J., and P. K. Tomlinson. 1969. A generalized stock production model. [In English and Spanish.] Inter-Am. Trop. Tuna Comm., Bull. 13:419-496. Richer, W. E. 1958. Handbook of computations for biological sta- tistics of fish populations. Fish. Res. Board Can., Bull. 119, 300 p. ROBSON, D. S. 1966. Estimation of the relative fishing power of individual ships. Int. Comm. Northwest Atl. Fish., Res. Bull. 3:5-14. Schaefer, M. B. 1954. Some aspects of the dynamics of populations important to the management of the commercial marine fisheries. Inter-Am. Trop. Tuna Comm., Bull. 1:25-56. 1957. A study of the dynamics of the fishery for yellowfin tuna in the Eastern Tropical Pacific Ocean. [In English and Spanish.] Inter-Am. Trop. Tuna Comm., Bull. 2:245-285. Schaefer, M. B., and R. J. H. Beverton. 1963. Fishery dynamics — their analysis and inter- pretation. In M. N. Hill (editor). The sea: Ideas and observations on progress in the study of the seas. Vol. 2, p. 464-483. Interscience Publ., N.Y. Tomlinson, P. K. 1970. A generalization of the Murphy catch equa- tion. J. Fish. Res. Board Can. 27:821-825. Walters, C. J. 1969. A generalized computer simulation model for fish population studies. Trans. Am. Fish. Soc. 98:505-512. 1041 OBSERVATIONS ON SCALE PATTERNS AND GROWTH OF THE PACIFIC SARDINE REARED IN THE LABORATORY Makoto Kimura and Gary T, Sakagawa^ ABSTRACT Scale patterns and growth of Pacific sardine (Sardinops caerulea) were studied with lab- oratory-reared fish held for 24 months. All scales examined after the fourth month had accessory marks. The first accessory mark was formed in August to October and the second accessory mark, in May to August. The accessory marks were indistinguishable from annuli that formed in November to March. Five possible causal factors of mark formation were investigated — temperature, sa- linity, gonad index, condition factor, and relative growth rate. Growth rate showed the best correlation. The accessory mark was formed during a period of change in growth rate and the annulus during a period of relatively constant growth rate. An estimate of the body length-scale radius relation indicated that scales first increase in size at a body length of 33 mm. Growth in length was rapid from the start of the experiment to the fourth month, after which the increase was gradual. The average instantaneous rate of growth was about 0.47/month during the first 4 months and about 0.03/month thereafter. It was concluded that the abrupt increase in Lj that was recorded in the 1940's for the sardine population was probably caused by errors in aging, owing to a change in scale readers and scale-reading criteria. The ages of fish age II and older were probably underestimated by the recent scale readers. In the course of aging Pacific sardine (Sardinops caerulea) from scales, the senior author ob- served, from comparison of back-calculated lengths with growth curves, that accessory marks or false annuli occurred frequently and were easily mistaken for true annuli. This was contrary to the findings of Walford and Mosher (1943:8-9), who reported that the annulus was "present on all normal scales of an individual" whereas an accessory mark was only rarely pre- sent on "all the scales of an individual." Since misidentifying an accessory mark as an annulus can aflfect the estimated age of a fish, a labora- tory experiment was conducted to determine (1) the frequency of occurrence of accessory marks, (2) the time of accessory mark formation, (3) possible factors that may cause accessory mark formation, (4) the time of annulus for- mation, and (5) the seasonal pattern of sar- ' National Marine Fisheries Service, Southwest Fish- eries Center, La JoUa, CA 92037. dine growth. The experiment was initiated in May and was terminated approximately 24 months later. Results from the first 12 months of the experiment were reported by Kimura (1970) , who showed that an accessory mark was present on scales of 4.5- to 5.0-month-old fish. This report is a more comprehensive presenta- tion of results from the entire experiment. METHODS REARING The collection of sardine eggs and the hatch- ing and rearing of the young in the laboratory for the first 12 months were described by Kimura (1970) . The general procedures, including those used after the 12th month, are briefly reviewed as follows. In May 1968, sardine larvae were hatched from eggs collected in plankton tows ofl? San Diego, Calif., and the larvae were held in a Manuscript accepted April 1972. FISHERY BULLETIN: VOL. 70, NO. 3, 1972. 1043 FISHERY BULLETIN: VOL. 70, NO. 3 polyethylene bag, which was suspended in a 4.6-m diameter pool (13.2 kliter) . The pool also contained about 15 northern anchovy (EngrauUs mordax) larvae. A 1,500-w mercury lamp that was suspended about 1 m above the bag provided illumination and a source of heat. The larvae were fed daily with plankton collected in Mis- sion Bay, Calif., and live Artemia. On the 29th day the larvae were released into the pool. Feed- ing continued with plankton and live Artemia up to the 45th day, after which frozen Artemia and dry trout food were substituted. The daily ration was about 10% of the estimated total weight of fish alive on a given day, based on the average weight of fish sampled in the previous month. On the 31st day the density of sardine larvae was noticeably reduced. Since few dead larvae were found during the previous days, the ancho- vies (average length 66 mm) were suspected of feeding on the sardines (average length 24 mm) . Analysis of stomach contents of three anchovies revealed that they were indeed feeding on the sardines.^ During the first few months of the experiment, water was added to the pool only periodically to remove surface scum. But beginning on the 61st day, seawater was circulated continuously through the pool at an initial rate of about 19 liter/min. This rate was increased on the 180th day to about 34 liter/min. the third month and continuing through the 12th month, about 20-24 fish/month were sampled. After the 12th month, the sampling rate was re- duced to 11-14 fish/month until an unusually large mortality from unknown causes reduced the population to 12 fish in the 16th month. Two more samples were taken thereafter; in one sam- ple four fish were measured for length and re- turned to the pool. After the 24th month, the seven surviving fish were sampled and the ex- periment was terminated. Standard length, total weight, gonad weight, and six scales removed from the body area at the tip of the pectoral fin (see Walford and Mosher, 1943:4) were obtained from most sam- ples. Water temperature was recorded daily on a thermograph. The monthly mean temperature ranged from 15.3° to 25.0°C and was generally 2.9 °C higher than the monthly mean surface temperature recorded off the Scripps pier. La Jolla (Kimura, 1970), site of the water intake for the experimental pooP (see Lasker and Vly- men, 1969), Salinity measurements were made by daily ti- trations from surface water sampled off Scripps pier. The mean values ranged from 33.09 to 34.38^r. It was assumed that water salinity was the same off the Scripps pier and in the experi- mental pool. SAMPLING Fish were sampled and sacrificed at various intervals during most of the duration of the ex- periment. About 7-17 larvae were sampled daily from the polyethylene bag during the first month. During the second month 6-26 larvae were sam- pled primarily on a weekly basis. Starting with SCALE MEASUREMENTS Scales were placed between two glass slides and viewed on a scale projector that magnified them 30 times. All marks that appeared to be annuli were recorded and the widths from focus to mark and focus to margin in the anterior field were measured. ^ To our knowledge there have not been previous re- ports on predation of sardine larvae by anchovies. The significance of our observation is that predation by an- chovies may have contributed to the decline of the Pa- cific sardine population. Our anchovies were hatched from eggs and were 44 days older than the sardines. This difference in age (size) between the species is also found in the wild; the northern anchovy has a peak spawning period during January-March and the sardine, during April-June (Ahlstrom, 1966). The young of both species coexist in the California Current; thus it is con- ceivable that predation by anchovies is an important source of sardine natural mortality. ANALYSIS Gonad weight was expressed as a percentage of total weight. This was designated a gonad index, or relative measure of sexual development. The relative "fatness" of a fish was estimated ' The experimental pool was located at the Southwest Fisheries Center, about 1 km from the Scripps pier. 1044 KIMURA and SAKAGAWA: SCALE PATTERNS OF PACIFIC SARDINE Table 1. — Results of aging of 205 known age, laboratory-reared sardines from scales. Total number of fish Total number of marks Actual Estimated Age (months) Number wi th: Number with Percent 0 mark 1 mark 0 mark 1 mark 2 marks 3 marks ment 5 20 1 20 0 20 0 0 0 0 6 20 1 20 0 20 0 0 0 0 10 15 2 0 15 0 12 3 0 20.0 11 21 2 0 21 0 10 11 0 52.4 12 24 2 0 24 0 18 6 0 25.0 13 11 3 0 11 0 3 8 0 72.7 14 12 3 0 12 0 2 6 4 83.3 15 14 3 0 14 0 0 1.2 2 lOO.O 16 64 3 0 64 0 4 50 10 93.7 18 4 3 0 4 0 lO 3 1 lOO.O Table 2. — Frequency and percent (in parentheses) of Pacific sardines with various number of scale marks, as determined from scale reading. (Sardines were reared in the laboratory for 24 months.) Age Number of fish No Accessory Annulus Accessory Annulus (months) mark mark 1 1 mark II II I 3 3 (lOO) 2 8 8 ClOO) 3 17 17 (100) 4 9 4 (44) 5 (56) 5 20 20 tlOO) 6 20 20 (100) 7 20 20 (100) 10 (50) a 20 20 (lOO) 13 (65) 9 20 20 (lOO) 18 (90) 10 20 20 (100) 20 (100) 11 2,1 21 (100) 21 (lOO) 12 24 24 (lOO) 24 (lOO) 13 1.1 11 (100) 11 (100) 4 (40) 14 12 la (loo) 12 (100) 6 (60) 15 14 14 (100) 14 (100) 10 (100) 16 64 64 (100) 64 (100) 64 (100) 18 4 4 (100) 4 (VOO) 4 (100) 24 7 7 (100) 7 (100) 7 (100) 7 (100) by a condition factor (MacGregor, 1959). The factor (K) is calculated as a ratio of the weight in grams (W) to the cubic power of length in millimeters (L): K = (W/L^) x 10^. Back-calculated lengths from scale measure- ments were derived by two methods. Method one was by direct proportion, Lm = (Sm/Sr)L, where Lm = back-calculated length at time of mark (m) formation, Sm = scale width from focus to mark m, Sr —' scale width from focus to margin, and L = length at sampling. Meth- od two was also by direct proportion but with a correction factor (c) for body length when scales first increase in size in the Pacific sar- dine. The method uses the equation Lm = c + (Sm/Sr) (L — C). ACCESSORY MARKS AND ANNULI FREQUENCY OF OCCURRENCE An annulus is generally believed to form at annual intervals, usually as the result of a slow- ing down in growth such as occurs in winter for temperate species. An accessory mark, on the other hand, is believed to occur, if at all, at ir- regular intervals (Walford and Moser, 1943), the causes of which are unknown. We examined the scales of the laboratory- reared fish and were unable to distinguish be- tween the two types of marks. To test if other scale readers similarly could not distinguish ac- cessory marks from annuli as we did, a series of scales from the laboratory-reared fish was mixed 1045 FISHERY BULLETIN: VOL. 70. NO. 3 with scales collected from wild fish and three ex- perienced scale readers read the scales (Table 1) . The percent error was as large as 100%, which indicates that accessory marks can indeed be easily misidentified as annuli. TIME OF FORMATION As many as f oiir marks were observed on some scales. The percentage of fish with various num- bers of marks was tabulated for each sample (Table 2) . The results show that the first mark was formed from August to October (4-5 months old), the secbnd from November to March (6-10 ANNULUS |< 1 months old) , the third from May to August (IS- IS months old), and the fourth sometime after November (18 months old) but probably before January (20 months old). This January date for the fourth mark was deduced from two scales collected in late December from the bottom of the pool, but the data are not given in Table 2, The two scales had four distinct marks, but it is not known whether the scales were from one or two fish. The bottom of the pool was cleaned daily. Based on the above criteria of an annulus and accessory mark, the second and fourth marks are annuli, and the first and third marks are ac- cessory marks. The interval between annuli was about 12 months, and that between acces- sory marks was only 9 months. ACCESSORY < n C0NDITI0r4 FACTOR RELATIVE GROWTH RATE GONAD INDEX WATER SALINITY WATER TEMPERATURE Figure 1. — Four possible factors that may be associated with mark formation on scales of Pacific sardines. Periods of mark formation are delineated by vertical lines. 1046 KIMURA and SAKAGAWA: SCALE PATTERNS OF PACIFIC SARDINE POSSIBLE FACTORS AFFECTING MARK FORMATION Water temperature and salinity, gonad index, condition factor, and growth rate were analyzed to determine whether they affected mark for- mation (Figure 1) . Growth rate appeared to be best correlated with mark formation. Kimura (1970) used monthly growth incre- ments* for the first year to show that formation of the accessory mark was associated with max- imum growth, and formation of the annulus was associated with the onset of rapid growth. His choice of growth increments for his analysis was not ideal because the magnitude of an increment is dependent on the size of the fish. We there- fore chose to use the instantaneous rate of growth, g = (In Lt — In Lo)/(* — U) , Ricker ( 1958) on a 30-day basis to analyze the data from the entire experiment. The results (Figure 1) show that the accessory marks formed during periods of change in growth rate, whereas the annulus formed during a period of relatively con- stant growth rate. GROWTH BODY LENGTH-SCALE RADIUS RELATION A body length-scale radius relation was fitted by least squares to data from 283 fish. A straight line of the form Y = 32.856 + 9.030Z, where Y = standard length and X — scale radius, was calculated (Figure 2) . The intercept of the line, or 32.856, is an estimated body length when scales first increase in size in the Pacific sardine. This estimate is probably too high, because scales with several circuli were observed on 26- to 30- mm long fish. Landa (1953) reported 12 positive intercept values (68-191 mm) and 1 negative value ( — 102 mm) for body length-scale radius rela- tions of fish caught by the commercial fishery in the 1940's. Compared to our estimate, his es- 180 160- 140 120 100 o IT 80 60 40 20 6 8 10 SCALE RADIUS (mm) * In Kimura's Figure 3 the notations for weight and length increments are mislabeled. The increments are not percentages but absolute values. Figure 2. — Body length-scale radius relation for 283 laboratory-reared Pacific sardines. timates are considerably larger. One reason for the difference is Landa used data from large fish (mean lengths of 186-228 mm) , whereas we used data from small fish (mean length of 115 mm). This suggests that a relation over the entire size range of sardines may be nonlinear, although it is linear over a short segment of the curve. The parameters of a linear relation could hence vary, depending on the segment of the curve being ex- amined. WEIGHT-LENGTH RELATION Data from 326 fish were used to estimate the weight-length relation. The relation (Figure 3) appears to underestimate the average weight of fish greater than 135 mm long. This is probably because the weight-length relation was based on data from individual fish whereas the data points in Figure 3 represent average weights for 5-mm groupings of lengths. Clark (1928) estimated a weight-length rela- tion for sardines landed at San Pedro, Calif., in the 1920's. Her estimate was compared to ours 1047 FISHERY BULLETIN: VOL. 70, NO. 3 90r 80 70 60- 50 40 30 20 10 0^M_ / — CLARK (1928) 30 50 70 90 110 130 STANDARD LENGTH (mm) 150 170 190 Figure 3.— Weight-length relation for the Pacific sardine, (Figure 3) and found to be significantly differ- ent based on analysis of covariance (F = 18.02, df = 1, 237). The laboratory-reared fish were appreciably heavier for a given length than sar- dines caught in the 1920's. This may be attrib- uted to several causes, among them diff'erence in diet, in amount of exercise, and in the range of sizes sampled. SEASONAL GROWTH PATTERN Growth of the Pacific sardine has been well documented by several investigators (e.g., Wal- ford and Mosher, 1943; Phillips, 1948; Felin, 1954; Clark and Marr, 1955) . Most of the stud- ies have concentrated on estimating growth based on scale readings of fish caught by the commercial fishery. Another method of estimating growth is by rearing experiments. Although we recognize the limitation of laboratory vs. natural condi- tions, we believe that estimates of growth of lab- oratory-reared sardines can indicate the general trend in growth in the wild. We therefore esti- mated growth of our laboratory-reared sardines. As shown in Figure 4, growth in length was rapid from the start of the experiment to the fourth month, after which the increase was more gradual. The average instantaneous rate of growth was about 0.47/month during the first 4 months and about 0.03/month during the fifth to 24th month. In contrast, growth in weight increased somewhat exponentially during two phases: during the first 4 months and again dur- ing the fifth to 14th month (Figure 5). Walford and Mosher (1943) reported the standard lengths of juvenile sardines caught in monthly samples in the late 1930's. Although the date of birth, and hence the exact age, of ANNULUS -I * ACCESSORY MAY J 0 I M A M J J 10 II 12 13 14 AGE (MONTH) Figure 4. — Growth in length of Pacific sardines reared in the laboratory for 24 months. Mean length is represented by a circle, and one standard deviation is shown on each side of the mean. The sample size is also indicated. The first accessory mark occurred in August-Oc- tober, and the second in May-August. The first annulus formed in November-March, and the second apparently in December. 1048 KIMURA and SAKAGAWA: SCALE PATTERNS OF PACIFIC SARDINE FiGUKE 5. — Growth in weight of Pacific sar- dines reared in the laboratory for 24 months. Mean weight is represented by a circle, and one standard deviation is shown on each side of the mean. The sample size is also indicated. 'MAY J J A S 0 N 0 JAN F M A M J J A S 0 N D JAN F M A M 0 1 i i 4 b 6 1 8 9 10 II 12 13 AGE (MONTH) 14 15 16 17 le 19 20 21 22 23 24 their fish was not known, we compared their data for the 1937 and 1938 year classes with ours (Figure 6). The results indicate that although growth of the 1937 year class was fast, growth was similar in fish caught in the 1930's and in the laboratory-reared fish, Marr (1960) presented data on the average length at time of first annulus formation (Li) and showed that there was a sharp change in Li to a higher level with the 1944 year class landed at San Pedro. Using his data, we calcu- lated separate estimates of average L\\ one for the 1934-43 year classes, and another for the 1944-57 year classes. The estimates are 101.3 and 131.5 mm, respectively. Compared with our estimate of 103.0 mm, growth of the 1934-43 year classes was almost identical to that for the lab- oratory fish, and growth for the 1944-57 year classes appears to have been faster than that for the laboratory fish. This faster growth may be an artifact and is discussed in a later section. (a) 1937 YEAR CLASS ,^°s. J MAYJJflSONOJANFMAMJJASONDJANFMAM 0 I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 AGE (MONTH) Figure 6. — Comparison of growth of juvenile sardines caught in the 1930's with growth of laboratory-reared sardines. Data for the 1937 and 1938 year classes are from Walford and Mosher (1943). BACK-CALCULATED LENGTHS Back-calculated lengths were computed for two samples, taken in September (16th month) and May (24th month). We reasoned that the back-calculated lengths would give an independ- ent estimate of the time of mark formation, if growth is assumed to follow that shown in Fig- ure 4. The average back-calculated lengths based on two methods are shown in Table 3. The first method gave lower estimates than the second. Both methods, however, gave back-calculated lengths that were reasonably similar to average lengths of samples collected during the period when the marks formed. 1049 FISHERY BULLETIN: VOL. 70, NO. 3 Table 3. — Back-calculated length of sardines sampled on two dates. The two methods of back-calculating length are discussed in detail in the text; in general, method 1 is by direct proportion, and method 2 is by direct proportion with a correction factor for the body length-scale radius relation. Sampling date Method Averag< 3 length (mm) at: Accessory mark 1 Annulus 1 Accessory Mark II Annulus II Margin size September May 1 2 1 2 71.5 88.9 75.1 93.6 99.9 1 10.6 104.7 116.8 135.6 137.8 137.4 142.3 153.1 155.0 143.3 143.3 160.3 160.3 27 27 7 7 DISCUSSION MARK FORMATION ON SCALES Walford and Mosher (1943) indicated that on sardine scales an accessory mark was disting- uishable from an annulus by its finer sculpturing and its rare occurrence on all scales of an indi- vidual. The results of our study on sardines reared in the laboratory showed that accessory marks occurred on all scales examined after the fourth month, and they were generally indisting- uishable from annuli. But for some scales in which the accessory mark was distinguishable from an annulus, the identifiable characteristic was the fine sculpturing mentioned by Walford and Mosher. Interruptions in the growth pattern are gener- ally assumed to form marks on scales (e.g., see Van Oosten, 1957). It is also widely assumed that the driving mechanism behind mark formation is temperature, through the influ- ence of a fish's metabolism (Brown, 1957). Our results indicate that mark formation is un- related to temperature, but appears to be related to growth rate. Hogman (1968) obtained some- what similar results in his experiments with coregonids. He found that formation of marks was closely related to growth and somewhat re- lated to temperature. However, Hogman indi- cated that light period may be the primary driv- ing mechanism. Bilton and Robins (1971) found that mark formation was correlated with in- crease in food supply, but not with resumption of feeding after starvation in sockeye salmon. Their experiments with light period proved in- conclusive. It thus appears that mark forma- tion in fishes is probably related to growth, al- though the actual driving mechanism (s) have so far not been clearly identified. ERROR IN AGING Scales are routinely used to age and study growth of the Pacific sardine. Our results indi- cate that extreme caution must be exercised in aging because of the presence of accessory marks on scales. Furthermore, since the annulus is formed during the winter, the actual age of a fish at time of first annulus formation may vary depending on its date of hatching. In our study the first annulus was laid down after the sixth month for fish hatched in May. May is the mid- dle of the heavy spawning season for the Pacific sardine, but the season extends from March to October (Kramer and Smith, 1971). Kimura (1970)° conducted an experiment to test the consistency of early and recent scale readers in aging sardines from scales. He found that the abrupt increase in Li, as reportedly ob- served in the 1940's by Marr (1960) for sardines landed at San Pedro, may have actually been caused by a change in scale readers and in criter- ia used in reading scales. We compared our av- erage length at annulus formation with the av- erage Li of fish aged by the early (Li = 101.3 mm) and recent (Li = 131.5 mm) scale readers, and discovered that the early readers probably aged sardines correctly, whereas the recent read- ers probably underestimated the age of age II ^ Kimura, M. 1970. Possible errors in locating the first scale annulus and in estimating the length of Pacific sardines. Manuscript filed at National Marine Fisheries Service, Southwest Fisheries Center, La Jolla, CA 92037. 1050 KIMURA and SAKAGAWA: SCALE PATTERNS OF PACIFIC SARDINE and older fish. If this was the case, and assuming that growth has not appreciably changed, then the number of fish that was incorrectly aged as age I by recent scale readers was at most equal to the number of fish correctly aged as age I. This is deduced from the fact that the presumably overestimated Li of 131.5 mm is about midway between our Li of 103 mm and L2 ( — length at time of second annulus formation) of about 160 mm. Back-calculated lengths for Pacific sardines (Marr, 1960) have been based on Method 1. A statistical test of the intercept of our body length-scale radius relation (Figure 2) indicated that the intercept is not significantly different from zero (t = 1.831, df = 282). Although this indicates that Method 1 is acceptable, the back-calculated lengths at first annulus are ap- preciably and significantly different between Method 1 and Method 2 (Table 3). It is there- fore advisable that Method 2 be used since it is the better procedure (Ricker, 1969). We conclude from these results that a bias may be present in published accounts of the age comp- osition of the Pacific sardine catch, which in turn may have affected studies on the population dy- namics of the Pacific sardine. We recommend that the method of aging Pacific sardines be re- evaluated, perhaps with the consideration of modifying the scale method so that the scale reader is made aware of the length of fish being aged or even utilizing other hard parts for aging, and that appropriate steps be taken to eliminate aging errors in the historic records on age comp- osition of the catch. We realize that this task will not be easy, but it may be worthwhile be- cause of the frequent use of the records to an- alyze fisheries hypotheses. ACKNOWLEDGMENTS We thank C. E. Blunt, Jr., J. E. Hardwick, and J. S. MacGregor for participating in our test to determine whether accessory marks can be misidentified as annuli. P. Paloma and A. Saraspe assisted in hatching and rearing the sardines. D. Kramer, W. H. Lenarz, J. S. Mac- Gregor, and A. M. Vrooman kindly reviewed the manuscript and gave helpful suggestions. J. R. Thrailkill drew the figures. We are indebted to these individuals. LITERATURE CITED Ahlstrom, E. H. 1966. Distribution and abundance of sardine and anchovy larvae in the California Current region off California and Baja California, 1951-64 : A summary. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 534, 71 p. BiLTON, H. T., AND G. L. Robins. 1971. Effects of starvation, feeding, and light pe- riod on circulus formation on scales of young sock- eye salmon (Oncorhynchus nerka). J. Fish. Res. Board Can. 28:1749-1755. Brown, M. E. 1957. Experimental studies on growth. In M. E. Brown (editor). The physiology of fishes, Vol. I, p. 361-400. Academic Press, N.Y. Clark, F. N. 1928. The weight-length relationship of the Cali- fornia sardine (Sardina caerulea) at San Pedro. Calif. Div. Fish Game, Fish Bull. 12, 58 p. Clark, F. N., and J. C. Marr. 1955. Population dynamics of the Pacific sardine. Calif. Coop. Oceanic Fish. Invest., Prog. Rep. 1953- 1955:11-48. Felin, F. E. 1954. Population heterogeneity in the Pacific pil- chard. U.S. Fish Wildl. Serv., Fish. Bull. 54:201- 225. HOGMAN, W. J. 1968. Annulus formation on scales of four species of coregonids reared under artificial conditions. J. Fish. Res. Board Can. 25:2111-2122. KiMURA, M. 1970. Formation of a false annulus on scales of Pacific sardines of known age. Calif. Coop. Oceanic Fish. Invest., Rep. 14:73-75. Kramer, D., and P. E. Smith. 1971. Seasonal and geographic characteristics of fishery resources, California Current region — VII. Pacific sardine. Commer. Fish. Rev. 33(10) :7-ll. Landa, a. 1953. The relationship between body length and scale length in five year-classes of the Pacific pil- chard or sardine, Sardinops caerulea (Girard, 1854). Pac. Sci. 7:169-174. Lasker, R., and L. L. Vlymen. 1969. Experimental sea-water aquarium. Bureau of Commercial Fisheries Fishery-Oceanography Center, La Jolla, California. U.S. Fish Wildl. Serv., Circ. 334, 14 p. 1051 FISHERY BULLETIN: VOL. 70, NO. 3 MacGregor, J. S. 1959. Relation between fish condition and popula- tion size in the sardine {Sardinops caerulea) . U.S. Fish Wildl. Serv., Fish. Bull. 60:215-230. Mark, J. C. 1960. The causes of major variations in the catch of the Pacific sardine Sardinops caerulea (Gir- ard). Proceedings of the World Scientific Meet- ing on the Biology of Sardines and Related Spe- cies, p. 667-791. FAO, Rome. Phillips, J. B. 1948. Growth of the sardine, Sardinops caerulea, 1941-42 through 1946-47. Calif. Div. Fish Game, Fish Bull. 71, 33 p. RiCKER, W. E. 1958. Handbook of computations for biological sta- tistics of fish populations. Fish. Res. Board Can., Bull. 119, 300 p. 1969. Effects of size-selective mortality and sam- pling bias on estimates of growth, mortality, pro- duction, and yield. J. Fish. Res. Board Can. 26: 479-541. Van Oosten, J. 1957. The skin and scales. In M. E. Brown (edi- tor) , The physiology of fishes, Vol. I, p. 207-244. Academic Press, N.Y. Walford, L. a., and K. H. Mosher. 1943. Studies on the Pacific pilchard or sardine {Sardinops caerulea). 2. — Determination of the age of juveniles by scales and otoliths. U.S. Fish. Wildl. Serv., Spec. Sci. Rep. 20, 19 p. 1052 UNSTRUCTURED MARINE FOOD WEBS AND "POLLUTANT ANALOGUES" John D. Isaacs' ABSTRACT The several species of fish living in the Gulf of California have been shown to possess quite different concentrations of cesium (and cesium in respect to potassium) than the same species of fish living in the Salton Sea. The Salton Sea fish display simple trophic steps of concentration, whereas those in the Gulf all show about the same levels. These differences are reasonably well explained by simplified trophic models of the two envi- ronments. The concentration factor found in the known and describable food chain of the Salton Sea, applied to a model of an assumed unstructured food web in the Gulf, leads to reasonable results. This suggests that study of the concentrations in marine organisms of such natural trace substances as cesium may lead to an understanding of the trophic position of the organisms, and hence constitute "pollutant analogues" that may yield a better understanding of the existing or potential distribution of pollutants in marine organisms. Young (1970) found that the cesium content of the muscle of the several species of fish in the Salton Sea (California) increased by about a factor of 2 (average 2.15) in each of the suc- cessive steps in the simple linear food chain ex- isting in that isolated marine environment and that the cesium/potassium (Cs/K) ratio simi- larly increased by about a factor of 3 (average 3.1). Results both for normal and radioactive cesium were in agreement. Young further dem- onstrated that the entry of the cesium into the food chain was via the phytoplankton, only. However, Young found that the same several species of fish in the Gulf of California displayed no such successive steps of increase, but rather, that the muscle tissue of all contained about the same concentration of cesium (with the suppos- edly more primary feeding mullets surprisingly having a somewhat greater than average con- centration of cesium) . Young's results are sum- marized in Table 1, where it can be seen that the Cs/K ratio is elevated by about a factor of ^ University of California, San Diego, Scripps Institu- tion of Oceanography, La JoUa, CA 92037. 16 between the primary food and the fish in the Gulf of California. Young suggested that these differences might be the result of a complex food web in the Gulf of California, and reported some evidence for this in the stomach contents of these Gulf fish. DEVELOPMENT OF AN UNSTRUCTURED FOOD WEB MATRIX AND EQUATIONS The following is an attempt to develop a highly simplified model of a generalized food web, to discuss some of its implications, and to test its validity by using real data. Let us first assume that in a food web each transfer of organic material (or energy), or of individual elements such as cesium, from one component of the web to another can be char- acterized by the following quantities: Ki = a. coefficient of conversion of matter (or energy) in food into living tissue 7^2 = a coefficient of conversion of matter (or energy) in food into irretrievable form (e.g., by respiratory combus- tion or mineralization) Manuscript accepted May 1972. FISHERY BULLETIN: VOL. 70, NO. 3, 1972. 1053 FISHERY BULLETIN: VOL. 70. NO. 3 Table 1. — Summary of Young's results. Salfoi ■) Sea Gulf of California Item Cs/K ratio X 10« Concentration factor relative to food Cs/K rotio X 106 Concentration factor relative to food Water 3.5 ± 0.1 0.8Q3 ± 0.047 Algae 4.0 ± 1.0 10.823 ± 0.047 Invertebrate' 3.8 ± 8.5 IX) - 2.1 Mullet 9.0 ± 1.1 2.3 ± 0.7 (algae) IS.^ ± 1.7 18.5 ± 2.0 Shad 14.8 ± 3.1 2.7 ± 0.7 (Invertebr.)* Sorgcxroaker as.o ±2.7 5.5 (invertebr.)' 13.2 ± 1 0 16.0 ± 1.2 Corvina 57.7 ± 7.7 2X) ± 0.3 (croaker) 10.4 ± 0.8 12.7 ± 1.0 Average 3.1 ± 0.8 12.8 ± 1.1 15.6 ± 1.3 ^ Cs/K for digae token as Cs/K for water. ' Concentration factor involves values for specific invertebrates within the range shown. Ks = a, coefficient of conversion of matter (or energy) in food into nonliving but re- trievable form (e.g., organic detritus or dissolved organic matter) For each component Ki + K2 + K3 = l (1) 0 < Ki < 1, where i = 1, 2, or 3. These coefficients are assumed to apply to the nutrient processes of all heterotrophs in the system, microorganisms, invertebrates, and ver- tebrates alike. Further they are assumed to have constant mean values for each ingested com- ponent of a food web in which plant production is the only primary source. Under these assumptions, some of the princi- pal characteristics of a generalized food web can be represented by a matrix such as Figure 1. Each division along an abscissa represents the conversion of a fraction, Kt, of any arbitrary unit of food into living tissue. Therefore, the numbers on the horizontal axis are powers of Ki. Each division along an ordinate similarly rep- resents the conversion of a fraction, Kz, of any arbitrary unit of food into retrievable but non- living organic matter, and the numbers on the vertical axis are, thus, powers of Kz. These numbers can be used to name points in the mat- rix as in any set of cartesian coordinates. One use of the matrix can be illustrated by consid- ering a unit of food. Mm, (e.g., in a copepod or in an edible nonliving particle) at 2, 3. If it is con- sumed by a heterotroph, the fraction Ki is con- verted into tissue and the fraction K^ into non- living retrievable matter. The sum of these two fractions Mm {Ki + Ks) is now found at 3, 4. Its trajectory in the matrix is always to the right and down.^ All points on a line at right angles to this trajectory (e.g., a diagonal line) can be considered to represent organic matter that has undergone the same number of steps of con- version from its origin as tissue of autotrophic organisms. The expression at the end of each diagonal line in Figure 1 is the sum of all frac- tions of an original unit of organic matter along that diagonal. The number of possible paths from the origin to each point in the matrix is easily counted and is the sum of the two numbers of units shown at the point, the lower right-hand number being the number of paths resulting in living matter (or energy) at that point, and the upper left- hand number being the number of paths result- ing in nonliving but recoverable matter (or en- ergy) at the point. It will be seen that this doubly infinite matrix is composed of two superimposed binomial (Pascal) triangles, and can be summed along diagonals resulting in two infinite series of bi- nomials of the form: a + a{a + b) + a{a + b)~ + a{a + 6)^ . . a(a+ b)". The sum of such a series is a 1—ia + b) (2) and for the sum of living matter (or energy), ^ Cases may be considered in which either Ki or K^ ^ 0 and motion is either only to the right or only down. Ki = 0 is, of course, not a permanently viable condition. 1054 ISAACS: FOOD WEBS AND "POLLUTANT ANALOGUES" LIVING STEPS K| n- A^ J A. +-^ >^ .+-^ V^ to a CO 3r ..^ .5- < UJ v^ tr ^-^^ ^r'^ ■k^ ^ 0 / 21 +-\ 0 ■7^ / / / I recoverable materlal- RULE FOR CONSTRUCTION OF MATRIX * Taken as the number of paths leading to the point. ■t-^ 0+ b living material c + d Figure 1. — Matrix representing principal characteristics of a gen- eralized food web. a = K\,'b — Ks, and for the sum of recoverable matter (or energy), a = K3, b = K\. These series can be further modified and summed in several ways. We will define a term, Mo, where Mo is an increment of initial input periodically introduced into the system at inter- vals equal to the time taken by one average step in the food web. Under the assumption that all steps are equally probable, the diagonals along which the matrix has been summed mark equal intervals of time. Thus, allowing the quantity Mo to be successively introduced at the origin of the matrix, the total quantity of material in the entire web, above the phytoplankton level, (Mt) for steady state becomes: Mt = Mt' + Mt" (3) where Mt = total in living material, and Mt" = total in nonliving recoverable Mt' Mt'' Mt Mo Mo material; l—(Ki + Ks) MoK, M0K3 K2 '^2 (4) (5) (6) FURTHER VARIATIONS OF THE FOOD WEB MATRIX ANALYSIS Later I will delete earlier steps in the matrix series. In this case, a more general solution of equation (4) involves a summation of the matrix above any step. Here the partial sum, 1055 FISHERY BULLETIN: VOL. 70, NO. 3 (7) where p is the number of steps of the series eliminated. Other types of formulations and summations are, of course, possible. For example, differen- ces in the time for living-to-living and recov- erable-to-living steps may be incorporated by multiplying the constant appropriate to the slow- er step by another constant representing the in- verse ratio of the times of the two respective processes. As another example, the average content of each point in time in the matrix can be derived by dividing each term of the series by the num- ber of points along the particular diagonal. Other manipulations of this matrix will be suggested later. CALCULATION OF THE COEFFICIENTS OF THE PREDACEOUS FOOD WEB For the purpose of examining Young's results on cesium, we will proceed with equations (4), (5), and (7) and test the hypothesis that the carnivorous fishes of the Gulf of California are members of an unstructured food web. That is to say that they randomly draw from the mean composition of the food web above the plant level. Since, from the Salton Sea results, the Cs/K ratio increases by a factor of about 3 for each living step, we can say that Kic ^ SKif, for organic food material (8) and since the invertebrates in the Salton Sea which feed partially on detritus show no sub- stantially diflferent increase in Cs/K ratio, I will assume that a double step from living to recov- erable to living material also results in an in- crease in the Cs/K ratio of only about 3, and, hence, that K, 3c Ksf (9) where subscript c denotes the coefficients Ki and Ki applicable to the Cs/K ratio, and subscript / those applicable to food. For the conditions in the Gulf of California, it is clear that, with the exception of the mullet, the fish do not consume plant material directly, and we can eliminate the first step in the series, and thus, for non-herbivorous omnivores, p — I, and equation (7) becomes M. Mo K,iKr + Ks) (10) '" 1—iKr+K,) Let us also define the Cs/K ratio in the plant material as: then Mpc Mpf Mac ^' l-iKrc+Ksc) = 16^0 (11) l—(Ku + Ksf) where the coefficient 16 is approximately the mean increase in the Cs/K ratio from algae to predaceous fish in the Gulf (Table 1), and sub- scripts are as in equations (8) and (9). Substituting (8) and (9), equation (11) for foo(f becomes S9Ki' + Kxi52Ks—7) —lBKz(l—Ks) = 0.(12) Assumption of a range of reasonable values of Ka results in the following: ^3 Ki Mt'/M, Mt'VMo Mf'/Mt" 0.0 0.18 0.22 0.0 ^■s 0.1 0.20 0.29 0.14 2.10 0.2 0.19 0.31 0.33 0.94 0.3 0.18 0.35 0.58 0.60 0.4 0.16 0.36 0.91 0.40 Values of ratios MZ/Mo (living material to in- put) , Mt'VMo (recoverable to input) and Mt/Mt" (living to recoverable) are derived, but it is not known what values of these ratios might exist in nature. The values of Ki, however, are within the range of values of conversion commonly found in laboratory experiments and culture operations involving growing animal populations. We can thus conclude that the assumption of K In the remainder of this paper, K^ ^ K^f, and K^^ = 3/- 1056 ISAACS: FOOD WEBS AND "POLLUTANT ANALOGUES" an unstructured food web in the Gulf of Cali- fornia in explanation of Young's findings on pre- daceous fish leads to reasonable estimates of the coefficient of conversion of food. PELAGIC VERSUS SHALLOW WATER POPULATIONS Other variations or testable conclusions or manipulations of these equations can be under- taken. For example, we can derive the relative biomass of two similar environments where, in one, the recoverable matter approaches zero. These might be a near-surface population ovej deep water where Ks-^ 0 because of rapid sink- ing of inert particles compared with a similar population over shallow water where K3 could have any of the values shown in the table above. From equation (4) Mt' pelagic Mt shallow X (1 - Mo pelagic Mo shallow K. fc)' (13) For the range of values found in the Gulf K3 = 0.0 Ks = 0.1 Ks = 0.2 Ks = 0.3 K3 = 0.4 Ki = 0.18 Ki = 0.20 Ki = 0.19 Ki = 0.18 Ki = 0.16 (1- K: k' = 1.00 = 0.88 = 0.75 = 0.62 = 0.52 THE MULLET ANOMALY The high Cs/K ratio of the mullet is difficult to explain. It is generally thought that the mul- let is largely herbivorous. Three possible factors may give rise to the anomalous levels: First, the mullet in the Gulf of California may ingest far more organisms with the detrital material than is generally supposed, and these organisms may be part of a microscopic multistep food web that generates a high Cs/K ratio. No such effect was apparent in the Salton Sea, however, pos- sibly because of the paucity and simplicity of invertebrate infauna in the detrital mat. Sec- ondly, some members of the detrital community in the Gulf may take up inorganic cesium from the sediment, unlike those of the Salton Sea. Third, the high fat level of the mullet results in their possessing a caloric level, as related to organic material, that is 1.6 that of the other fish. This argues that it might be more signifi- cant to normalize cesium to calories than to dry organic weight, since more food intake (and hence more cesium) possibly is required for a given increment of growth in a fish with high fat content, with the potassium better normalized to fat free dry weight. This latter assumption would bring the Cs/K ratio in the mullet into the lower ranges of that in predaceous fish. Pursuing this latter possibility, the adjusted Cs/K ratio for Gulf mullet can be set at about 10, and considering that the composition of the det- rital feeding creatures (Md) should be [from food represented by equation (5)] then the Cs/K ratio should be (14) Mdc KicK. 3c l-{Krc-{^Kzo)_ ^ ^0^^ K\jKzf (15) Substituting equations (8) and (9), as before, this becomes K, = 1 -^ K^ (16) which is within the range of values for these con- stants previously derived (e.g., Ks = 0.30, Ki = 0.18). In simultaneous solutions of equations (12) and (16) values of Ks are extremely sen- sitive to the values of concentration. The values of Ki, however, are essentially stable. CONCENTRATION FACTORS It should be noted in this discussion that the high concentration of cesium in the mullet may be a specific case of a more general principle, which is that the concentration factor (/c) of a trace substance in a single step may be inversely 1057 FISHERY BULLETIN: VOL. 70, NO. 3 related to the coefficient of conversion of food (in terms of material). Certainly the upper value of fc is limited to the total trace substance intake "diluted" by the new grov^rth and hence, ^''^ l^f (17)* This argues that a creature with a low coefficient of conversion for food can show high concentra- tions of trace elements or pollutants for this reason alone. Such lower coefficients of conversion and high concentration of trace elements may result from definitive growth, sickness or abnormality or, as probably in the case of the mullet, when food is converted into a more energetic form. Conclu- sions as to cause and effect of pollutant trace ele- ment concentrations in creatures with abnormal- ities must thus be considered with caution. eral, and hence into the existing or potential distribution of pollutants. ADDENDUM Several reviewers of this paper have suggest- ed that I expand the treatment of trophic types in an unstructured food web. Dominance of any highly complex trophic types is incompatible with an unstructured food web hypothesis. However, the following may be considered com- patible (from equations 4, 6, or 7 as appropri- ate): Strict herbivores (19) CONCLUSIONS The assumption and analysis of an unstruc- tured food web is reasonably consistent with findings on the concentration of the element cesium (in respect to potassium) in the Gulf of California. It suggests that unstructured ma- rine food webs may be common and that the dis- tribution of natural trace elements, such as cesium, may give insight into food webs in gen- * This can be derived more formally as follows : let Cf be the concentration in the food; then 9/ where q^ is quantity of the substance and q^ is total quantity of food material. Then the concentration in the organism. since /c = C„ = Co QcK \c K, WTi-^'^^i , and Kic ]^ 1 , then /c}> ^ Omnivores Kt M. = Mo i_(^;^^3) (20) Particle feeders (detritus + phytoplankton) * = ^»T^Hfc^^ <21) Detrital feeders Ma - Mo 1 — (^1 + Ks) Strict predators /Ti^ Non-herbivorous omnivores M„ = Mo 1— {Ki + Ks) (14) Feeders on the detrital milieu (detritus and detrital feeders) M... = M. ff -,]/. X ^f^ (22) ^- = ^» T^TTTrrrKT ^^s) (10) '' Equation 20 could well have been employed instead of eauation (10) in the previous treatment. Differences in results would be small, however. 1058 ISAACS: FOOD WEBS AND "POLLUTANT ANALOGUES" These expressions can be employed to gen- erate expressions for the coefficient R, or the increase in the concentration of any component passing through the food web and introduced only via plants. The derivation follows that of equation (11) above, where j is the coefficient of increase over the conversion of food, as in equation (8) above, where j ^ 3. Also, for in- terest, is shown the coefficient N, which is the Jimiting or maximum increase in the biomass of each trophic type resulting from a unit input into the food web (i.e. Mo = 1), evaluated from the above equations. Strict herbivores Omnivores Particle feeders Detrital feeders Feeders on the detrital milieu Strict predators Non-herbivorous omnivores * for Ki = 0.18, Ks = 0.30, j = 3 It should be noted that the coefficient N indi- cates only the potential increase of the biomass of each trophic type, as related to productivity input. The potential population (standing crop) depends upon how closely this maximum is approached, and the fecundity, mortality, etc. of the appropriate organisms. It should also be noted that the usefulness of the total assumptions of this model can be tested by the determination of R for an appropriate R Equation No. R* N* j (24) 3.0 0.18 jK2 (25) 9.7 0 35 1 jKt K^ jK^a-jK,) (26) 5.5 0 28 (1 /^i)(l i^i K^) 3K2 (27) 9.7 0 10 1 jKi Kz JK2 (1 + yxi) (28) 12.7 0 12 (1 +Ki)(l— jKi- K,) j'K2 (29) 10.4 0 06 1 jKi Ks JK2 (jKr + Ks) (30) 15.0 0 17 K, + Ks)il-3k^- -K,) trace element, such as Cs, in samples of well defined trophic types. LITERATURE CITED Young, D. R. 1970. The distribution of cesium, rubidium, and potassium in the quasi-marine ecosystem of the Salton Sea. Ph.D. Thesis. Scripps Inst, of Oceanogr., Univ. California San Diego, La JoUa. 1059 IN MEMORIAM: OSCAR ELTON SETTE Oscar Elton Sette died on July 25, 1972, at Palo Alto, Calif. He was 72. On September 7th his ashes were scattered at sea from the research vessel David Starr Jordan. The last issue of this journal {Fishery Bulletin, Vol. 70, No. 3) was dedi- cated to Elton Sette. His colleagues who contributed papers to that thick number make up a roster of some of the most able scientists in the field of marine research today, yet it was woefully incomplete, because many who had wished to contribute w^ere unable to do so. The number had been planned by his friends as a surprise gift to Elton (but how it would have benefited from his quick editorial eye !) : it must now stand as a memorial. Towards mid-July, Phil Roedel, Director of the National Marine Fish- eries Service, wisely broke the 2 years of secrecy surrounding the "Sette number" and sent Elton the table of contents and a copy of Pat Powell's warm portrait. Elton was pleased. Since July 1970, Elton had been a member of the Editorial Committee of the Fishery Bulletin. This is the first number not to carry his name. As- suredly, it will not be the last to show his influence on the course of marine research. We had planned a gift to him: he left one to us — a standard of high and unflagging professional competence. Not many men leave as much. Reuben Lasker Scientific Editor Thomas A. Manar Chief, Scientific Publications Staff 1061 TEMPERATURE AND PHYTOPLANKTON GROWTH IN THE SEA Richard W. Eppley^ ABSTRACT The variation in growth rate with temperature of unicellular algae suggests that an equation can be written to describe the maximum expected growth rate for temperatures less than 40 °C. Measured rates of phytoplankton growth in the sea and in lakes are reviewed and compared with maximum expected rates. The assimilation number (i.e., rate of photosynthetic carbon assimilation per weight of chlorophyll a) for phytoplankton photosjTithesis is related to the growth rate and the carbon/chlorophyll a ratio in the phytoplankton. Since maximum expected growth rate can be estimated from tempera- ture, the maximum expected assimilation number can also be estimated if the carbon/ chlorophyll a ratio in the phytoplankton crop is known. Many investigations of phytoplankton photosynthesis in the ocean have included measures of the assimilation number, while fewer data are available on growth rate. Assimilation numbers for Antarctic seas are low as would be expected from the low ambient temperatures. Tropical seas and temperate waters in summer often show low assimilation numbers as a result of low ambient nutrient concentrations. However, coastal estuaries with rapid nutrient regeneration processes show seasonal variations in the assimilation number with temperature which agree well with expectation. The variation in maximum expected growth rate with temperature seems a logical starting point for modeling phytoplankton growth and photosynthesis in the sea. Temperature does not seem to be very important in the production of phytoplankton in the sea. For example, Steemann Nielsen (1960) has written, "Recent investigations have shown, however, that the direct influence of tempera- ture on organic production in the sea is fairly insignificant." Other reviewers of photosyn- thesis in the sea likewise give little or no con- sideration to a role of temperature and Steemann Nielsen's statements find widespread endorse- ment in the published data on geographic and seasonal variation in marine phytoplankton photosynthesis. In response to this, the reader may ask at least two questions: (1) Why is temperature of so little importance and (2) why would any- body write a review on temperature and phy- toplankton growth in the oceans? Several an- swers to the first question have appeared in the literature and some of these will be discussed ^ Institute of Marine Resources, University of Cal- ifornia, San Diego, P.O. Box 109, La Jolla, CA 92037. in this account. I have two answers for the second question. The purpose of this review is (1) to suggest maximum growth and photo- synthetic rates which might be reasonably ex- pected for natural marine phytoplankton and (2) to point out the interrelationship among growth rate, photosynthetic assimilation num- ber (i.e., rate/chlorophyll), and carbon/chlor- ophyll a ratios in the phytoplankton. What follows is an attempt to show that tem- perature sets an upper limit on phytoplankton growth rate and on the rate of photosynthesis per weight of chlorophyll, and that this upper limit can be predicted from a knowledge of tem- perature and the carbon and chlorophyll content of the plants. It can perhaps be inferred, from above, that phytoplankton growth in the oceans seldom ap- proaches the upper limits imposed by the tem- perature of the water and that temperature does not figure prominently in simulation models for primary production in the marine environment. Other factors effect reduced rates of growth and Manuscript accepted May 1972. FISHERY BULLETIN: VOL. 70, NO. 4, 1972. 1063 FISHERY BULLETIN: VOL. 70, NO. 4 photosjTithesis and diminish the potential pro- duction of phytoplankton. Nevertheless, earlier reviewers have been able to generalize on sev- eral aspects of the relation between phytoplank- ton growth and temperature (see especially Tailing, 1957; Steemann Nielsen, 1960; Ichi- mura and Aruga, 1964; Yentsch and Lee, 1966; Ichimura, 1968). Culture experiments have re- v^ealed that clones of a species isolated from cold or warm seas may differ in their optimum tem- perature for growth (Braarud, 1961; Hulburt and Guillard, 1968). VARIATION IN SPECIFIC GROWTH RATE ifji) WITH TEMPERATURE IN LABORATORY CULTURES OF UNICELLULAR ALGAE Much of the available data on the specific growth rates of algae in culture have been as- sembled by Hoogenhout and Amesz (1965). Growth rates for marine phytoplankton fall in the same range of values as those for freshwater algae, and there are no obvious distinctions be- tween marine and freshwater unicellular algae rt'ith respect to the variation of specific growth rate (/x) with temperature. Hence data for algae from the two media will not be segregated. Specific gro\\i;h rate is defined as the rate of increase of cell substance per unit cell substance i/N dN/dt — [x. Since dN/dt depends upon the rate of metabolic processes, one expects some temperature variation of ix if conditions are otherwise favorable for growth (i.e., if light and nutrient supply are not growth-rate limiting). This variation can be seen in Figure 1. Data 3f Figure 1 were selected from Hoogenhout and Amesz (1965) as representing, as nearly as pos- sible, growth rates measured under conditions such that temperature would be rate limiting. Figure 1 shows much variation in /.i among spe- cies at a given temperature. Most of this re- sults from differences in cell size (Williams, 1964; Eppley and Sloan, 1966; Werner, 1970) and in the concentration of photosynthetic pig- ments within the cells of the different species (Eppley and Sloan, 1966). It has been mechanically impossible to iden- tify each of the points on the Figure by species (approximately 130 species or clones were in- cluded, some for several temperatures). No doubt, by further literature search, the entire area beneath the line of maximum expected growth rate could be filled in. It is perhaps surprising and a tribute to the quality of the measurements from many laboratories that only three of nearly 200 values were rejected as being unrealistically high. Inclusion of these three spurious values would only be an embarrassment to the authors rather than a critique of the va- lidity of the line of maximum expected growth rate presented. Not plotted in Figure 1 are values of yu, for Chlamydomonas mundana photoassimilating ac- etate (Eppley and Macias R., 1962), Chlorella pyrenoidosa 7-11-05 for which fx was computed for increase in cell substance uncoupled from cell division (Sorokin and Krauss, 1962), or for the photosynthetic bacteria listed by Hoogenhout and Amesz (1965). Values for these slightly exceed the line of maximum expected /x. Figure 1 is limited to algae growing photoautotrophic- ally with carbon dioxide and water. Two general trends are noted in Figure 1: (1) There is a gradual and exponential increase in /x with temperature up to about 40°C. Tem- perature data above 40°C, obtained with thermo- philic, blue-green algae show no further increase in IX (Castenholz, 1969) . Such temperatures are outside the range encountered in the ocean and will not be further discussed. (2) Values of fx below 40°C seem to fall within an envelope and it is possible to draw a smooth curve, i.e., a line of maximum expected value, to describe the up- per limit of IX to be expected at a given temper- ature. An approximate equation for this line is: logio fx = 0.0275T — 0.070 (1) where T is temperature in degrees Celsius. Equation (1) gives a Qio for growth rate of 1.88, slightly lower than expected from the Qio for photosynthesis measured in natural waters (Tailing, 1955, gives Qio = 2.3; Williams and Murdoch, 1966, give Qio = 2.25 ; Ichimura, 1968, gives Qio = 2.1) or the Qio for growth rate of laboratory cultures suggested earlier (Eppley and Sloan, 1966, give Q,o = 2.3). 1064 EPPLEY: PHYTOPLANKTON AND TEMPERATURE 12 - 10 o ■o cr> 8 o LjJ 6 < cr X I- ^ 4 O oc •ib- -- ,3- "I . • I 2 2 • -.-i- 10 20 30 TEMPERATURE °C 40 Figure 1. — Variation in the specific growth rate (m) of photoautotrophic unicellular algae with temperature. Data are all for laboratory cultures. Growth rate is expressed in dou- blings/day. Approximately 80 of the points are from the compilation of Hoogenhout and Amesz (1965). That listing is restricted to maximum growth rates observed, largely in con- tinuous light. The figure also includes additional data, mostly for cultures of marine phy- toplankton, from the following sources : Lanskaya (1961), Eppley (1963) , Castenholz (1964, 1969), Eppley and Sloan (1966), Swift and Taylor (1966), Thomas (1966), Paasche (1967, 1968), Hulburt and Guillard (1968), Jergensen (1968), Smayda (1969), Bunt and Lee (1970), Guillard and Myklestad (1970), Ignatiades and Smayda (1970), Polikarpov and Tokareva (1970). The latter papers include about 50 strains of marine phytoplankton. The line is the growth rate predicted by Equation (1), i.e., the line of maximum expected /i. Small numbers by points indicate the number of values which fell on the point. I will avoid speculation on possible reasons why such a curve would include algae from a wide variety of taxonomic groups, including both eucaryotic and procaryotic cell types, cells with different complements of photosynthetic pig- ments, and diverse morphologies. Nevertheless, the curve and Equation (1) appear to be useful as a generalization of maximum /x to be expected for photosynthetic unicellular algae. Equation (1) is essentially a van't Hoff for- mula and can be expressed in the more typical form 1065 FISHERY BULLETIN: VOL. 70, NO. 4 fi = 0.851 (1.066)' (la) McLaren (1963) discussed the choice of a tem- perature function and preferred the formula of Belehradek DetonulQ confervacea H- a (T — a)' (lb) where a, b, and a are constants. A virtue of this equation, among the three monotonic func- tions discussed by McLaren (1963) is that a, the scale positioning factor, represents a "bio- logical zero" for the process. A graph of log (/x) vs. log (T — a) assumes linearity for appro- priate values of a. Fitting values from Equa- tion (1) at T = 0, 10, 20, and 30 degrees gave linear graphs if a were ^ — 40 degrees. For a — — 40, a and b were approximately 2.46 X 10~^ and 3.45, respectively. Figure 1 can be made more understandable by comparing fx vs. temperature curves for a few selected species for which fairly complete data are available (Figure 2) . Each of these species has a different optimum temperature and the maximum growth rate for each approaches the line of maximum expectation. Such "fx vs. tem- perature" curves typically show a gradual di- minution of fx as temperature decreases from the optimum, but an abrupt decline at supraoptimal temperatures. Temperature optima and the upper critical temperature can be shifted somewhat by alter- ing environmental conditions. For example, the salinity of the culture medium influences these parameters in euryhaline Dunaliella tertiolecta (Figure 3). Note, however, that only one salt concentration gives the unique maximum growth rate of about 5.0 doublings/day. The figures can be criticized as being limited with respect to the number of species included. Furthermore many of them represent "labora- tory weed" species and relatively few are ecolog- ically significant ocean phytoplankton. Happily this shortcoming is temporary and information on important planktonic species is growing (see Figure 1 legend). Use of Figures 1 and 2 or Equation (1) for insight as to maximum expected values of /a in the sea presumes that natural marine phyto- plankton are autotrophic. But it is conceivable, o T3 < 5 o 20 30" TEMPERATURE 40 50 Figure 2. — Growth rate vs. temperature curves for five unicellular algae with different temperature optima: Detonula confervacea (Guillard and Ryther, 1962; Smayda, 1969), Skeletonema costahnn (Jorgensen, 1968), Dityhnn brightwellii (Paasche, 1968), Dunaliella tertiolecta (McLachlan, 1960; Ukeles, 1961; Eppley, 1963; Eppley and Sloan, 1966), Chlorella pyrenoidosa. (Sorokin and Krauss, 1958, 1962). although perhaps unlikely in the sea, that hetero- trophic nutrition might lead to values of /-t higher than predicted above, as appears to be the case when one compares doubling times of heterotrophic and photosynthetic bacteria or autotrophic vs. photoheterotrophic growth rates of the sewage alga Chlamydomonas mundana. Equation (1) has been useful in this labora- tory for predicting the maximum dilution rates ("washout rates") for continuous cultures. In the few organisms examined here the value of fjL at washout was slightly higher than the max- imum rate observed in batch cultures of the or- ganism, but within the envelope of values pre- dicted by Equation (1). 1066 EPPLEY: PHYTOPLANKTON AND TEMPERATURE Rates of growth given by Equation (1) are much higher than those which permit the op- eration of mass cultures at maximum efficiency of light utilization or nutrient removal. Max- imum production will be achieved when the product of /x and standing stock is a maximum, and light is likely to be limiting growth at some depth in the culture under these conditions (see, for example, Ketchum, Lillick, and Redfield, 1949; Myers and Graham, 1959). The data of Figures 1 and 2 apply to cultures grown with continuous illumination (or with optimum daylength for those in which /^ passes through a maximum at intermediate daylength [Castenholz, 1964; Paasche, 1968]). This les- sens the utility of the data for predictive pur- poses with natural phytoplankton exposed to seasonally varying daylength since the daylength for maximum fx varies among species (Table 1). Efforts to generalize on the influence of day- 5 - //A ^ 4 1( / \* \ ^ "'°\A — V-^ o ^^ ■D G \ ^ (A B^ \. ° 2 3 - A =) \ o \ 3 \ Ul \ 1- \ < \ cc \ I 2 - \ 1- \ 5 \ o \ oc \ 1 - 1 I M Na CI 25 30 35 TEMPERATURE 40 45 Figure 3. — Growth rate vs. temperature curves for Dunaliella tertiolecta measured in culture media con- taining different salt concentrations (R. W. Eppley and F. M. Macias, unpublished data). Table 1. — Daylength resulting in maximum growth rate for some algae which show depressed growth rate in continuous light. Some species which showed maximum fi in 24 hr light/day are shown for comparison. Day- Growth Temper- Organism length rate' ature Reference (hr) ''max (°C) Ditylum brighttvellii 16 2.1 20 Paasche (1968) Nilzschia turgidida 16-24 2.5 20 Paasche (1968) Fragilaria sp. 24 1.7 11 Castenholz (1964) Biddulpkia sp. 15 1.5 11 Castenholz (1964) Synedra sp. 15-24 1.2 11 Castenholz (1964) Melosira sp. 15-24 0.7 11 Castenholz (1964) Units are doublings/day. length on /x have not been very successful since the daylength allowing maximum /x at a given temperature seems to vary with the intensity of illuinination (Tamiya et al. 1955; Terborgh and Thimann, 1964). A proportion between fi and the number of hours of light/24 hr is often as- sumed but this can be only a first approximation. Use of Figure 1 and Equation (1) for insight on the behavior of natural phytoplankton re- quires the further assumption that the organ- isms present are reasonably adapted to ambient temperatures and are, preferably, at a temper- ature somewhat less than optimum. Aruga (1965a) has shown this to be so for the phjrto- plankton of a pond on the University of Tokyo campus. Smayda (1969) has discussed his own and earlier observations on the distribution of phytoplankton in nature where temperature optima for growth in laboratory cultures were invariably higher by several degrees than the water temperature in which the species flourish. Figure 2 suggests that /jl for suboptimal tem- peratures will be only slightly lower than would be predicted from the maximum n for the species given a temperature coefficient (Qio) for growth of about 2. However, some organisms show a critical lower temperature, above the freezing point of water, below which no growth occurs. Ukeles (1961) has listed such lower critical temperatures for several species, and see Smayda (1969) for another example. Temperatures in excess of the optimum for growth result in a much steeper decline in /jl with increasing tem- perature than do suboptimal temperatures; growth in this thermal region would be risky 1067 FISHERY BULLETIN: VOL. 70, NO. 4 if the ambient temperature were subject to fluctuations of a few degrees. ESTIMATES OF THE SPECIFIC GROWTH RATES OF PHYTOPLANKTON IN THE SEA REVIEW OF METHODS Measurement of the phytoplankton specific growth rate in nature is not a routine procedure both because of the lack of widely accepted meth- odology and because the utility of such data is not well appreciated. J.W.G. Lund, J. F. Tailing, L. A, Lanskaya, T. J. Smayda, J. D. H. Strick- land, and S. Ichimura and his colleagues have been the pioneers in such measurement in nat- ural waters while R. W. Krauss and J. Myers have promoted the measurement of ix for lab- oratory cultures. Minimal values of /u, can be calculated from rates of increase of cell concentration or of chlorophyll during the spring bloom in temperate waters, although advection, diffusion, and graz- ing complicate their interpretation. Recent ex- amples of this technique are provided by Bunt and Lee (1970), Pechlaner (1970), and Happey ( 1970) . Samples of water can also be incubated in bottles for cell counting at intervals (see, for example. Tailing, 1955; Smayda, 1957). In oligotrophic waters the period of growth neces- sary to allow a precise estimation is likely to result in the depletion of nutrients and the grad- ual diminution of yu, with time. In rich water if growth were extensive, changes in jx would be expected as a result of the decrease in effective illumination in the bottles due to self-shading. Short-term incubations of less than 24 hr may be complicated by diel periodicity in the property measured, by synchronous cell division, or in- sufficient change for meaningful calculations. Such problems are eased in shipboard cultures provided with adequate nutrients for growth, but here rates may be unreasonably high if am- bient nutrient or light levels in the natural water are not duplicated. Estimates of /u. are obtained routinely in terms of ^"^N-nitrogen assimilation rate per unit par- ticulate nitrogen in the sample, but such rates will underestimate fjL to the extent that the par- ticulate nitrogen analyzed includes detrital and other nonphytoplankton nitrogen (Dugdale and Goering, 1967). Carbon assimilation rates per unit phyto- plankton carbon have also been calculated but suffer from the errors inherent in measuring the latter as well as from the uncertain reality of incubation conditions (Riley, Stommel, and Bumpus, 1949; McAllister, Parsons, and Strick- land, 1960; McAllister et al., 1961; Antia et al., 1963; McAllister, Shah, and Strickland, 1964; Strickland, Holm-Hansen, Eppley, and Linn, 1969). What is needed is an instantaneous method not confounded by the complexities of long incubation either in situ, in enclosed ves- sels, or in shipboard cultures. Unfortunately, no such method is in view. In this laboratory two methods have been em- ployed for estimating the carbon content of the crop. In the first of these, all the cells in the sample are counted and their dimensions mea- sured so that the cell volume of each species can be calculated (see Kovala and Larrance, 1966, for dealing with cell shape problems) . The carbon content of a cell is then computed from its volume, or "plasma volume," using empirical equations developed from laboratory culture (Mullin, Sloan, and Eppley, 1966; Strathmann, 1967). The carbon in each species is then ob- tained from the concentration of cells of that species, and the total carbon of all species is summed. Several applications of this method have been published (Strickland, Eppley, and Rojas de Mendiola, 1969; Holm-Hansen, 1969; Eppley, Reid, and Strickland, 1970; Reid, Fug- lister, and Jordan, 1970; Zeitzschel, 1970; Beers et al., 1971; Hobson, 1971; Eppley et al., in press) . In the second method, only recently put into practice, the adenosine triphosphate (ATP) content of particulate matter retained on a fine porosity filter is determined (Holm-Hansen and Booth, 1966) . The ATP is apparently restricted to living cells but may include contributions from bacteria, protozoans, and other colorless micro- organisms, as well as phytoplankton, even if larger animals are removed by passing the sam- ple through netting. However, phytoplankton appear to be predominant in water samples from 1068 EPPLEY: PHYTOPLANKTON AND TEMPERATURE the euphotic zone judged from the rough pro- portionality of ATP to chlorophyll. Estimates derived from ATP appear to agree well with those given by the first method (Holm-Hansen, 1969) and the ratio C/ATP approximates 250. In determining an average fi for the phyto- plankton the carbon content, as measured above, is taken at the beginning of the photosynthesis measurement to give phytoplankton carbon at time zero (Co). The measured daily rate of photosynthetic carbon assimilation, assumed to represent net carbon assimilation (AC), is then added to the carbon content after a day's growth. The specific growth rate is then calculated as: for Pyrocystis species in situ (E. Swift, Uni- versity of Rhode Island, personal communica- tion). Changes in cell morphology related to cell division probably give the least ambiguous estimates of /u where advection and sinking are not serious problems and when a parcel of water can be followed over time. The time course of change in valve diameter in diatoms seems to be out of favor for estimating /a since valve di- ameter in cultures may not decrease in a regular way or always be proportional to the number of cell divisions. Methods of measuring microbial growth rates were recently reviewed by Brock (1971). 1 , ,Co + AC .n\ ^ = T ^°^^ ^ C^ ) ^2) to give fx in doublings of cell carbon per day. It should be straightforward to compute /u, using ATP determined initially and after 24-hr incubation, and this has been done at least once (Sutcliffe, Sheldon, and Prakash, 1970). We have used chlorophyll a values, before and after 24- or 48-hr incubation, to compute /a but the results were poor due to the plasticity of cell chlorophyll a content and the difficulty of pro- viding incubation conditions sufficiently close to those in nature to maintain constant cell chlor- ophyll a per cell or per weight of carbon (Eppley, 1968). Increase in the total volume of particulate matter, determined with an electronic particle counting and sizing machine, can also be used to compute fx (Parsons, 1965; Cushing and Nich- olson, 1966; Sheldon and Parsons, 1967). This method holds much promise when changes are large enough to be significant/)ver background levels of particulates. The cost of the machines is a serious drawback to wider use, and the prob- lems in proper incubation of the sample to mimic conditions in the sea are as serious here as in the other incubation methods. Sweeney and Hastings (1958) used the per- centage of paired dinoflagellate cells in cultures as a measure of the time of day of cell division and this has been used at sea (R. Doyle, Duke University, personal communication). A vari- ation on this theme has allowed estimates of /u. RESULTS OF GROWTH RATE MEASUREMENTS IN THE NATURAL PHYTOPLANKTON AT DIFFERENT TEMPERATURES In their classic paper of 1949, Riley, Stommel, and Bumpus expressed photosynthetic rate as the daily carbon assimilation per unit plant car- bon, a measure readily calculated as fx in dou- blings/day. They used Baly's equation as a model. This equation includes temperature as a variable influencing photosynthetic rate. The constants in the equation were computed from Baly's compilation of data on Chlorella cultures and detached leaves, and from Jenkin's 1937 data for a culture of Coscinodiscus incubated at var- ious depths in the sea. I have calculated ex- pected values of ^ using their Equation 6 for different levels of total incident radiation (Fig- ure 4). It is seen that the Baly equation is rel- atively insensitive to temperature, in comparison to Figure 1, and gives values inconsistent with the results from laboratory cultures. Bunt and Lee (1970) provide a unique set of data on the photosynthetic rates of Antarctic phytoplankton which grow under the ice layer, an environment with low ambient light and with temperature approximately — 2°C. They also provide seasonal values of the particulate carbon and chlorophyll a concentration. A maximum, midsummer, value of /a was less than 0.5 dou- blings of cell carbon/day. Most of the data which allow estimates of ^ are from nutrient-poor waters, such as are found 1069 FISHERY BULLETIN: VOL. 70, NO. 4 O 1.0 ly/min A 0.53 ly/min n 0 05 ly/min 10 20 TEMPERATURE »C 30 Figure 4. — Growth rate vs. temperature relationship predicted by the Baly equation as used by Riley, Stom- mel, and Bumpus (1949). Three different levels of total radiant energy are included for the Baly equation: 1.0 (circles), 0.53 (triangles), and 0.05 ly/min (squares). The line of maximum expectation predicted by Equation (1) is drawn for comparison (no symbols). in the Sargasso Sea, the Eastern Tropical Pa- cific, and southern California coastal waters. Exceptions are jx estimates obtained from up- welling regions off Peru (Strickland, Eppley, and Rojas de Mendiola, 1969; Beers et al., 1971) and southwest Africa (Hobson, 1971) where nutrient limitation is not a factor reducing yu. A summary of estimated values of [x, as av- erage doublings of cell carbon/day in the euphotic zone, is provided in Table 2. The re- cent data are based upon simulated in situ tech- niques usually involving 24-hr incubation in order to obtain photosynthetic rates free of er- rors resulting from diel periodicity in metab- olism. The list of values given is not inclusive but is, hopefully, representative. Mean values of fjL in the Peru Current showed little variation and averaged about 0.7 doubling/day. Values of this magnitude are consistent also with esti- mates from ^^N-labeled nitrate assimilation rates measured by R. C. Dugdale, J. J. Goering, and Table 2. — Some estimates of the average specific growth rate of phytoplankton in the euphotic zone for various regions. Temperatures indicated are for the surface or the average in the mixed layer. Location Temper- ature Growtti rate as doublings/day Reference (°C) Measured Max. expected Oligotrophic waters Sargasso Sea _. 0.26 _. Riley, Stommel, and Bumpus (1949) Florida Strait -_ 0.45 _^ Riley, Stommel, and Bumpus (1949) Off the Corolinas _. 0.37 «a Riley, Stommel, and Bumpus (1949) Off Montauk Pt. __ 0.35 _^ Riley, Stommel, and Bumpus (1949) Off southern California July 1970 20 0.25-0.4 1.5 Eppley et al. (in press) Apr. -Sept. 1967 12-21 0.7 avg 0.9-1.6 Eppley et al. (1970) Peru Current Nutrient-rich waters Apr. \966 17-20 0.67 avg 1.5 Strickland, Eppley, and Rojas de Mendiola (1969) June 1969 18-19 0.73 avg 1.4 Beers et al. (1971) Off S.W. Africa ._ 1.0 avg .. Calculated from Hobson (1971) Western Arabian Sea 27-28 >1.0 avg 2.4 Calculated from Ryther and Menzel (1965b) ' From Equation (1) assuming ^ will be one-half the value calculated as expected if daylength Is 12 hr and /t is directly proportional to the number of hours of light per day. 1070 EPPLEY: PHVTOPLANKTON AND TEMPERATURE co-workers (University of Washington, 1970) in tiie Peru upwelling region. The maximum values of /jl observed in depth profiles off Peru approached those expected from Figure 1 if the effect of daylength is considered (Figure 5) but were lower as a result of low insolation brought about by continuous cloud cover. Depth profiles of (jl roughly parallel those for photosynthetic rate per weight of chlorophyll a and both show diminished rates with depth as a result of decreasing light. Figure 5 also shows a depth profile of /a for the North Central Pacific where /u. was depressed because of low ambient nutrient concentrations. Enrichment experiments suggested that growth rate was limited at two stations by low concen- trations of both nitrogen and phosphorus and at a third station by nitrogen alone (Perry, in press) . Thomas (1970b) and Thomas and Owen (1971) reported values of fju for 10 m depths in the eastern tropical Pacific Ocean. In situ /u, was estimated to be about 0.2 doubling/day re- sulting from low ambient nitrogen concentra- tion. Shipboard cultures were enriched with various concentrations of nitrogen (nitrate and ammonium), and the variation of fx with nitro- gen concentration was determined (Thomas, 1970b), Maximum values of /x were 1.1-1.5 doublings/day. In many cases nutrient limitation (in the up- per surface waters) or light limitation (in deep- er waters and in nonstratified water where ver- tical mixing may reduce the effective light level to which the phytoplankton are exposed) ap- pears to decrease /x. The values expected from Figure 1 are not realized under such conditions and fi shows little or no temperature-dependence. Table 3 presents growth rates measured by three different methods (i.e., from the velocity of nitrogen assimilation per unit particulate ni- trogen, from the photosynthetic carbon assim- ilation rate per unit phytoplankton carbon, where the carbon content of the phytoplankton crop was determined from ATP, and from cell concentration and cell volume). Growth rates from the three methods usually agree within a factor of two, but more precise methods are de- sirable. The value from ^''N assimilation rate 100 GROWTH RATE (doublings/day) 0.2 0.4 0.6 0.8 1.0 50 1.2 1.4 ~i 1 T 1 r- I 10 NO Me Figure 5. — Variation in growth rate of natural marine phytoplankton with depth in the Peru Current, June 1969, and in the subtropical North Pacific central gyre, November 1971 (this laboratory, unpublished). The "light depth" of the ordinate was calculated as the ra- diant energy at depth as a percentage of that at the surface so that data from the two regions, with euphotic zone depths of about 30 and about 150 m, could be com- pared. The calculated line is based on Equation (1) for 19 °C with the assumption that light limits growth rate below the surface. The iJ.„^a\ from Equation (1) was multiplied by (7/2.5 + H where / is the radiant energy at depth as percent of surface. The half-satura- tion constant of 2.59c is low (see Rodhe, 1965) and sug- gests that the Peru Current phytoplankton were "shade adapted." Hence, measured /x would be less than ex- pected from Equation (1), in spite of abundant nu- trients. In the North Pacific study enrichment exper- iments and other data suggested limitation of phyto- plankton growth rate by both nitrogen and phosphorus concentration (Perry, Renger, Eppley, and Venrick, un- published data). There the temperature in the mixed layer was 22 °C and the maximum expected value would be slightly greater than shown for the Peru Current. 1071 FISHERY BULLETIN: VOL. 70, NO. 4 Table 3. — Some comparison of the average growth rate of phytoplankton in the euphotic zone in southern California coastal waters using different methods of estimation. Station (g Photo- synthetic rata C/mVday) Standing stock Growth rat. by e (doubli method ing/dc jy) Month From From ATP cell vol. (g C/m2) (a) (b) (0 June 1970 4 0.53 2.4 2.0 0.28 0.33 0.13 7 1,05 3.1 3.25 0.42 0.40 July 1970 1 1.37 8.4 _^ 0.26 0.19 6 1.10 4.4 __ 0.32 0.22 10 0.36 3.6 __ 0.13 0.21 19 1.76 5.9 5.38 0.37 0.40 0.15 1 Methods: (a) /i from photosynthetic rate and ATP X 250 = standing stock as carbon. (b) A from photosynthetic rate and standing stock carbon computed from cell numbers and cell volumes. (c) II computed from assimilation rate of nitrate -\- ammonium + urea per unit particulate nitrogen. Data for method (c) from McCarthy (1971) and Institute of Marine Resources (1972, see text footnote 2). Other data are unpublished values from this laboratory. Surface water temper- atures were 18°-20°C. Maximum expected growth rates would be about 1.5 doublings/day. tends to be lower than those from "C assimila- tion rate because no correction was made for the detrital nitrogen in the particulate matter, while detrital carbon is not a complication in the other methods. Low growth rates in these samples resulted from nitrogen limitation. Rates of nitrogen assimilation per weight of particulate N were measured in the Sargasso Sea and Peru upwelling regions (Dugdale and Goering, 1967; Dugdale and Maclsaac, 1971), and in the eastern tropical Pacific Ocean (Goer- ing, Wallen, and Nauman, 1970) which allow estimates of yu,. As is readily seen from the above discussion and the values of Tables 2 and 3 we have very little data at hand to properly evaluate the role of temperature in determining maximum rates of phytoplankton growth in the sea, and whether Figures 1 and 2 and Equation (1) are useful guides for field work. It is hoped that this lack will stimulate more effort to make growth rate measurements. Most needed are /x values for cold waters and warm, nutrient-rich waters. Meantime let us turn to lakes and ponds. Ad- ditional growth rate data are available and the influence of temperature on growth rate is often apparent. Since growth rates seem comparable in laboratory cultures for freshwater and ma- rine unicellular algae, as noted earlier, /x vs. temperature in lakes should be of equal interest to limnology and oceanography. Some data are given in Table 4 which confirm low jx values in cold water and a variation in ^ with temperature in outdoor ponds. The phytoplankton growth rates in lakes which show a variation in ^ with temperature were usually measured in the spring as the waters were gradually warming but before nu- trients were depleted to levels limiting to the rate of phytoplankton growth (cf. Cannon, Lund, and Sieminska, 1961). Presumably simi- lar data could be gathered for nutrient-rich estuaries or for temperate, coastal sea areas where sufficient warming occurs to obtain a reasonable range of temperatures before strati- fication and nutrient depletion become severe. The seasonal succession of phytoplankton in coastal ocean waters has been much studied, and the change in the phytoplankton crop from pre- dominantly diatoms to flagellates, with the on- set of nutrient depletion, would be accompanied by a marked decrease in growth rate. One may judge the magnitude of change from the com- parison of /i in the Peru Current with /x in the North Pacific central gyre (Figure 5). INTERRELATION BETWEEN SPECIFIC GROWTH RATE OF PHYTOPLANKTON AND ASSIMILATION NUMBER The specific growth rate of phytoplankton in laboratory cultures is often measured from the rate of increase in the concentration of cells in the culture when cell counts are determined over a time interval, i.e., H- = 1 M logs (3) 1072 EPPLEY: PHYTOPLANKTON AND TEMPERATURE This can also be expressed as fX = 1 Ai log2 ( ]^^^^ ) (4) where Ni is the initial cell concentration, No the cell concentration after an interval of time, M, and AN" is N2 — Nu To determine jx from analogous carbon units one needs the initial carbon content of the phytoplankton (Ci) and either the carbon content after a time interval A^, i.e., C2, or a measure of carbon assimilation by the phytoplankton during the time interval, i.e., AC. It will be assumed that the "C method of measuring phytoplankton photosynthesis (Steemann Nielsen, 1952) in fact measures AC, the net increase in particulate carbon in the phy- toplankton. This is indicated by several studies with laboratory cultures which include two or more independent measures of the rate of carbon assimilation by the phytoplankton cells (Antia et al., 1963; McAllister et al., 1964; Eppley and Sloan, 1965; Ryther and Menzel, 1965a; Strick- land, Holm-Hansen, Eppley, and Linn, 1969). Then fi can be calculated from carbon data from Equation (2). The evaluation of fi requires a measurement of photosynthetic rate as carbon and an estimate of the carbon content of the phytoplankton at the initiation of the measure- ment. Direct methods for the latter are not usu- ally suitable because of detrital carbon in na- tural waters and indirect methods must often serve (see earlier discussion of methods of measuring fi) . A convenient way of expressing photosynthetic rate per unit phytoplankton standing stock is the "assimilation number," i.e., the rate of photosynthetic carbon assimilation per weight of chlorophyll ft. The terms "assim- ilation ratio" and "photosynthetic index" are common synonyms for assimilation number. If the carbon/chlorophyll a ratio in the phytoplank- ton is known, its carbon content can, of course, be calculated from chlorophyll measurements. Usu- ally this is not the case and considerable effort has been expended to derive such estimates (see, for example, Harris and Riley, 1956; Cushing, 1958; Wright, 1959; Steele and Baird, 1961, 1962; Lorenzen, 1968; Eppley, 1968; Zeitzschel, 1970; Hobson, 1971). An equation expressing fx (as doublings of cell carbon /day) in terms of the assimilation number per day and the carbon/ chlorophyll ration of the phytoplankton can be derived from Equation (2) as H- 1 M , C/Chl. a + AC/Chl. a . ,_, ^°^^ ^ OTChTi )• ^^) This equation is useful in that it directly relates the assimilation number, i.e., the photosynthetic rate per weight of chlorophyll (AC/Chl. a), the carbon/chlorophyll a ratio of the phytoplankton Table 4. — Phytoplankton growth rates in lakes and ponds. Organism Temper- ature (°C) Growtli rate as doublings/day Reference Measured Max. expected^ 1-m depth only Astenonella jormosa 5 0.8 1.2 Average In the lake Tailing -(1955) Stephanodiscus hantichii Asterionella formosa Stephanodiscus rotula 2-4 5 8 15 0.3 1.1 0.3 1.2 0.25 1.4 0.7 2.2 \n outdoor ponds Pechlaner (1970) Happey (1970) Happey (1970) Happey (1970) Chlorella ellipsoidea 7 15 0.15 1.3 0.65 2.2 Tamlya et al. (1955) 25 1.4 4.1 1 From Equation (1). 1073 FISHERY BULLETIN: VOL. 70, NO, 4 (C/Chl. a), and /jl. Figures 6 and 7 show this relationship graphically where the calculated as- CO CJ t- O & C/CHL = 30 Q C/CHL = 60 X C/CHL = 90 X C/CHL = 120 0 1 2 SPECIFIC GROWTH HATE (DOUBLINGS/DRY) Figure 6. — Photosynthetic rate (assimilation number/ day) vs. the specific growth rate of the phytoplankton computed from Equation (5). Photosynthetic rate is expressed as milligrams carbon assimilated per day per milligram chlorophyll a and is shown for several values of the ratio carbon/chlorophyll a in the phytoplankton crop (30, 60, 90, and 120 g/g). t— o A C/CHL = 30 CD C/CHL = 60 X C/CHL = 90 X C/CHL = 120 1 2 SPECIFIC GROWTH RRTE (D0U6LINGS/DRT1 similation number per day (Figure 6) or per hour (Figure 7) is graphed as a function of fi for different carbon/chlorophyll a ratios in the crop. Carbon/chlorophyll ratios of Figures 6 and 7 are typical of the Peru upwelling region (C/Chl. a 30-40) (Lorenzen, 1968; Strickland, Eppley, and Rojas de Mendiola, 1969; Beers et al., 1971) and the Western Arabian Sea (Rjrther and Menzel, 1965b), or low^nutrient surface waters off southern California (90-100) (Eppley, 1968; Strickland, 1970); and of sur- face waters in the North Pacific central gyre (120-150) (Institute of Marine Resources, un- published data) . The marked dependence of the assimilation number upon the carbon/chloro- phyll a ratio of the phytoplankton is noteworthy, although little discussed in the literature. It is interesting that assimilation numbers greater than about 15 per hour (see Figure 7) are rarely reported in the literature and one wonders whether this is because of disbelief in the va- lidity of the data or because high fx and high C/Chl. a are somehow mutually exclusive in nature. The latter is most likely since such high assimilation rates and high fx would place ex- treme demands for nutrients, such as N and P, on the environment and could not long be sus- tained without massive nutrient input. Even at southern California sewage outfalls where high rates of nutrient input prevail we found low values for /x. These low values apparently result from the buildup of high phytoplankton crops which maintain low-nutrient levels in the surface waters such that growth rate is nitrogen- limited (Institute of Marine Resources).' Fur- thermore, high C/Chl. a ratios seem to be typical of nutrient depleted cells which grow slowly. For example, carbon/chlorophyll a ratios in- creased from 30 to over 150 with increasing nitrogen limitation of growth in N-limited chemostat cultures of marine phytoplankton (Thomas and Dodson, in press; Institute of Marine Resources'). Figure 7. — Same as Figure 6, but photosynthetic rates (assimilation numbers) were calculated per hour, rather than per day, assuming 12 hr light per day (i.e., values of Figure 6 were divided by 12). ° Institute of Marine Resources. 1972. Eutrophica- tion in coastal waters: nitrogen as a controlling factor. Final Rep. U.S. Environ. Prot. Agency, Proj. #16010 EHC. Inst. Mar. Resour., Univ. Calif., San Diego. 67 p. 1074 EPPLEY: PHYTOPLANKTON AND TEMPERATURE THE VARIATION OF ASSIMILATION NUMBER WITH TEMPERATURE IN THE SEA The maximum expected values of /u, at dif- ferent temperatures, from Equation (1), can be used to predict maximum assimilation numbers to be expected in the sea (as grams carbon/gram chlorophyll a per time). Combining Equations (1) and (5) gives rise to Figures 8 and 9 to show assimilation numbers per day and per hour for different C/Chl. a ratios in the phytoplank- ton. Actual rates would be lower than those shown for the reasons already discussed and would require the growth of small-celled phy- toplankters with light essentially saturating for photosynthesis and with adequate nutrient con- centrations. Aruga (1965b) presents graphs of assimilation numbers vs. temperature, with var- ious light levels, for Sce7iedesmus sp. grown at 20°C. His curves resemble these in form. The question of the influence of daylength upon /ji is ignored in Figure 8 and needs further investigation before generalities may be drawn. In Figure 9 it was assumed that /jl in natural phytoplankton assemblages will be one-half the value calculated from Equation (1) since that function assumes continuous light rather than natural illumination of, on the average, 12 hr light and 12 hr dark. There are several reasons why the dramatic potential effects of temperature on assimilation number are not often observed in oceanic studies and why so little variation in assimilation num- bers has been observed (cf. Ryther and Yentsch, 1958; Curl and Small, 1965). One of these is that growth at different temperatures results in shifts in the chemical composition of phyto- plankton. Increased C/Chl. a ratios at low tem- perature would tend to increase assimilation numbers in cold waters over those predicted by Figures 8 and 9 and a constant C/Chl. ft ratio cannot be assumed. Steemann Nielsen and Jorgensen (1968a, b) point out that while the lowering of the tem- perature of a culture of Skeletonema costatum reduced the growth rate (by an amount to be expected from Figure 1 and Equation (1)), the A C/CHL = 30 Q C/CHL = 60 X C/CHL = 90 X C/CHL = 120 10 20 TEMPERflTURE IN DEGREES C. Figure 8. — The variation in maximum expected rate of photosynthesis (assimilation number) with temperature. Rates were computeci by combining Equations (1) and (5) and are expressed as milligrams carbon/milligram chlorophyll a/day. Continuous light was assumed. i C/CHL = 30 a C/CHL = 60 X C/CHL = 90 E C/CHL = 120 10 20 TEMPERRTUBE IN DEGREES C. Figure 9. — Maximum expected photosynthetic rate (as- similation number) from Equations (1) and (5) with the assumption that the growth rate will be one-half the value predicted by Equation (1) to adjust for natural daylength averaging 12 hr light/day. Photosynthetic rates are expressed as milligrams carbon assimilated/ milligram chlorophylll a/hour. This figure gives values more in line with ocean measurements than does Fig- ure 8. 1075 FISHERY BULLETIN: VOL. 70, NO. 4 photosynthetic rate at light saturation was de- creased by a lesser amount. Assimilation num- bers for S. costatum at 2° or 8°C were higher than would be expected from Figure 9, if it were assumed that a constant C/Chl. a ratio was main- tained at all temperatures. They observed that cells at low temperature contained greater amounts of photosynthetic enzymes and of or- ganic matter than at higher temperatures. For example, S. costatum assimilated 10.2 picogram (pg) carbon/cell in one generation at 20°C, but 19.5 pg at 7°C (Jorgensen, 1968). The carbon content of a cell nearly doubled between 20° and 7°C. Dunaliella tertiolecta cells were likewise larger at low temperature than at high temper- atures as were cells of Ditybmi brightivellii (Table 5). This phenomenon seems to be gen- eral for mesothermal marine phytoplankton, but data for cold water species are not available. Fluctuations in C/cell and in the C/Chl. a are about twofold over the 10°-15°C range studied (Table 5). Steele and Baird (1962) reported high C/Chl. ft ratios in winter in Loch Nevis and suggested that they resulted from low light "etiolation." One wonders if low winter tem- peratures might also play a role in this. We have seen that low temperature reduces the assimilation number and promotes increased carbon/chlorophyll a ratios. Similar effects re- sult from nutrient deficiency and were well doc- umented by McAllister, et al. (1964). An in- fluence of nutrient deficiency on fx, was shown also in Figure 5 for the North Pacific and was noted in the eastern tropical Pacific (Thomas, 1970b). Low assimilation numbers for phyto- plankton photosynthesis in nutrient-impover- ished waters are well known (Curl and Small, 1965) and are clearly shown by Ichimura (1967; see his graph of assimilation number vs. phos- phate concentration in the waters of Tokyo Bay). Caperon, Cattell, and Krasnick (1971) reported 10 year increases in assimilation num- bers in Kaneohe Bay, Oahu, Hawaii (from ap- proximately 6-8 to 11-13 between 1960 and 1970) which attended increased rates of waste dis- charge into the bay. Hepher (1962) found as- similation numbers of about 4 in unfertilized fish ponds while values in fertilized ponds av- eraged about 7.6. Furthermore, there are many examples of enhanced "C-assimilation rates in shipboard enrichment experiments in response to nutrient additions. A recent report is that Table 5. — Carbon content of a cell and carbon/chlorophyll a ratios in phytoplankton cultured at different temperatures. Organism Carbon/cell aCM./a Temper- afura Reference SkeUtonema costatum 19.5 _^ 7 Jorgensen (1968) 16.5 10 Jorgensen (1968) 12.7 __ 15 Jorgensen (1968) 10.2 — 20 Jorgensen (1968) Ditylum brightwellii''- 1600 41 5 Checkley (1972)= 1500 43 7.5 Checkley (1972)2 1330 50 10 Checkley (1972)2 720 25 15 Checkley (1972)2 — 20 14.5 Strickland, Holm-Hansen, Eppley, and Linn (1969) 680 14 20 Eppley, Holmes, and Paasche (1967) DunalieHa ttrtiotecta^ 41.8 33 12 Eppley and Sloan C1966) 35.6 29 16 Eppley and Sloan (1966) 25.9 25 19.5 Eppley and Sloan (1966) 28.2 24 20 Eppley and Sloan (1966) 25.3 26 21 Eppley and Sloan (1966) 22.5 16 25 Eppley and Sloan (1966) ' D. brightwellii was cultured with irradiance 0.05 cal/cm^/min with periodic illumination 12L : 12D by Checkley (1972, see footnote 2 below). Values are for samples at the begin- ning of the light period 2 Checkley, D. 1972. carbon to chlorophyll a Resour., La Jolla, Calif ^ D. terlioUcta was The effect of the variation of growth temperature on the ratio of 1 a laboratory culture of Ditylum brightwellii. Univ. Calif., Inst. Mar. (Unpubl. manuscr.) cultured under continuous light with irradiance 0.07 cal/cm2/min. 1076 / EPPLEY: PHYTOPLANKTON AND TEMPERATURE of Glooschenko and Curl (1971). These authors, and Thomas (1969, 1970a), found no enhance- ment in waters in upwelhng regions, but assim- ilation numbers were increased in response to nutrient additions in oligotrophic subtropical water. Malone (1971a, b, c) found assimilation numbers in eutrophic waters to be nearly an order of magnitude greater than those in oligo- trophic surface waters of the subtropical and tropical Pacific. It has so far proved difficult to sort out the effects on assimilation number of low light and low temperature in seasonal studies of natural waters. Phytoplankton cultures grown with either low light or low temperature show low maximum photosynthetic rates per chlorophyll a at light saturation (Pmax) and low saturating intensity (h) for photosynthesis (Tailing, 1957; Steemann Nielsen and Hansen, 1959, 1961; Ichimura, 1960; Yentsch and Lee, 1966). Thus some of the effects on assimilation number usu- ally attributed to low light levels may, in cold waters, result also from low temperature. Bunt and Lee (1970) were able to sort out the two variables in their study of diatom growth under the ice in Antarctica by comparing a station with clear ice to one with snow cover. Photosyn- thetic rate and growth rate were considered to be light -limited at the station with snow cover but temperature-limited at the clear ice station (see also Saijo and Sakamoto, 1964, for photo- synthesis vs. depth curves in ice-free and ice- covered lakes). Assimilation numbers in Antarctic waters are low. Many values are less than 1.0 per hour (Mandelli and Burkholder, 1966; Home, Fogg, and Eagle, 1969; Bunt and Lee, 1970). Saijo and Kawashima (1964) found an average value of 1.2 mg C/mg Chi. a/hr which they attributed to low temperatures and to a deep mixed layer (resulting in a low average irradiance seen by a cell) . Water temperature in these studies was usually in the range — 2° to 1°C. El-Sayed and Mandelli (1965) gave a range of 1.0 to 3.6 for the assimilation number in surface samples over a temperature range — 1.75° to 6.0°C. Assim- ilation numbers of 4-5 were found in Drake Passage and Bransfield Strait where water tem- perature was usually about 1°C (El-Sayed, Mandelli, and Sugimura, 1964). All these val- ues are compatible with assimilation numbers predicted by Figure 9. Besides shifts in carbon/chlorophyll a ratios with temperature and the effects of nutrient lim- itation and light on assimilation number there is yet another factor which tends to obscure the expected seasonal changes in assimilation num- ber with temperature. This comes about as a result of the variation in growth rate and assim- ilation number with cell size. By passing a water sample through netting one can conveniently separate the phytoplankton into two size cate- gories: the larger cells and diatom chains re- tained by the net (the netplankton) and the smaller cells and chains which pass through the net (the nanoplankton). Malone (1971a, b, c) has recently compared assimilation numbers of the two size fractions and cites earlier studies. Invariably, the nanoplankton showed higher assimilation numbers than the netplankton, as would be expected from laboratory studies (cited earlier) which show a regular diminution in growth rate with increasing cell size. He fur- ther showed that netplankton are relatively more abundant during upwelling in coastal waters off California (Malone, 1971b). Chain-forming di- atoms seem to be characteristic of the rich waters of temperate regions during the spring bloom. Yentsch and Ryther (1959) have shown a pro- gressive increase in the ratio nanoplankton/net- plankton with seasonally increasing temperature off New England. Tropical, warm, oligotrophic waters have been shown repeatedly to contain a high proportion of nanoplankton (see refer- ences cited by Malone and by Sutcliffe et al., 1970). The causes of such seasonal succession of phy- toplankton species is one of the significant problems in the study of marine phytoplankton. One can only speculate on possible contributing factors. For example, the high (relative) sink- ing rates of large-celled species and long diatom chains suggest that suspension and buoyancy are more significant problems for large cells than small (Munk and Riley, 1952; Smayda, 1970). Hence stratification, reduced mixing, and the im- position of a seasonal thermocline would tend to discourage large forms. Perhaps the most ele- 1077 FISHERY BLXLETIN: VOL. 70, NO. 4 gant work in such problems is that of Lund and colleagues on diatom succession in the English lakes. Artificially mixing a lake in summer, when it would normally be stratified, permitted a bloom of Melosira italica, a diatom which usu- ally sinks out of the water column upon the for- mation of a thermocline in late spring (Lund, 1971). Another factor which tends to select against large-celled species in low-nutrient waters re- sults from a low surface 'volume ratio and a con- sequent inability to absorb nutrients from low concentration (Munk and Riley, 1952). This generalization has been confirmed in laboratory experiments on the kinetics of nutrient absorp- tion where large-celled species showed higher half-saturation constants (Ks) for nitrate and ammonium uptake than small-celled species (Eppley, Rogers, and McCarthy, 1969). Simi- larly, the Ks for assimilation of vitamin B12 by phytoplankton depends on cell size (Carlucci, 1972)." The argument with respect to netplankton vs. nanoplankton and the expected seasonal changes in assimilation number with temperature can be summarized as follows: (1) Nanoplankton show higher assimilation numbers (and growth rates) than do netplankton. This generalization results both from observations of natural phy- toplankton and from studies of variations with cell size in laboratory cultures. (2) Increasing insolation in the spring results in increased water temperatures, and often in stratification and seasonal thermoclines. Nutrients in the mixed layer then tend to be depleted and often rather quickly, except in very shallow water where regenerative activities of microorganisms in sediments can maintain adequate nutrient levels for rapid phytoplankton growth. (3) Stratification of the water column tends to dis- courage the growth of large-celled species and long chain diatoms, because (a) reduced vertical mixing may result in their sinking out of the water column and (b) they are less effective in ^ Carlucci, A. F. 1972. Saturation constants for vitamin assimilation by phytoplankton. (Unpubl. manuscr.) absorbing nutrients from low ambient concen- trations than are nanoplankton. (4) Both sea- sonal increase in temperature and in the ratio of nanoplankton /'netplankton should increase assimilation numbers for photosynthesis except where growth and i^hotosynthetic rates are re- duced by nutrient limitation. Nanoplankton would be expected to be more abundant, relative to netplankton, in oligotro- phic waters because of their lower sinking rates and lower Ks values for nutrient absorption. Hence, phytoplankton of warm, oligotrophic tropical waters would be expected to show high assimilation numbers (and growth rates) except for effects of nutrient limitation. One can begin to understand from all this why a graph of as- similation number vs. temperature for observa- tion of natural phytoplankton usually fails to show the relationship expected from Figure 9, and why so much current work emphasizes the role of nutrient concentrations in phytoplankton growth in the sea. Some exceptional marine waters which do show the expected relationship betw^een assim- ilation number and temperature are shallow coastal estuaries where nutrient regeneration on the bottom maintains a high nutrient input to the overlying water. Examples reported for the east coast of the United States are Barlow, Lorenzen, and Myren (1963), Williams and Murdoch (1966), and Mandelli et al. (1970). Both of the latter papers show graphs of assim- ilation number vs. temperature which match beautifully the relation expected in Figure 9. Williams and Murdoch's data fall between the C/Chl. a 30 and 60 lines, with an indication of higher C/Chl. a ratio in winter, as expected. Mandelli et al. present two graphs, one for di- atoms and the other for dinoflagellates. Assim- ilation numbers of the latter are higher than those for diatoms and fall on the line in Figure 9 for C/Chl. a = 30. They also show the seasonal change in relative numbers of diatoms and dino- flagellates; the latter are more abundant at high- er temperatures. Williams and Murdoch (1966) cite several other studies which show parallels between phy- toplankton production in shallow marine waters and temperature over the seasons. The Danish 1078 EPPLEY: PHYTOPLANKTON AND TEMPERATURE results are reviewed also by Steemann Nielsen (1960). Few of these earlier works included chlorophyll a measurements, however, and as- similation numbers are not reported. Ichimura (1967) found a close relation be- tween temperature and assimilation number for a station well within Tokyo Bay, but not at a station in deeper water. Nutrient limitation was postulated for the outer station. Some of the values for assimilation number and its variation with temperature which can be conveniently summarized are provided in Table 6. One might have hoped, by comparison of the data with values expected from Figure 9, to check up on the quality of one's colleagues' work and to find some reported values outside the bounds of reasonable expectation. Happily, only one of the papers reviewed gave unrealistically high assimilation numbers and these were not repeated in subsequent studies by that author. IMPLICATIONS FOR SIMULATION MODELS OF PHYTOPLANKTON PRODUCTION As pointed out by Patten (1968) and others, mathematical models are usually designed to be accurate or alternatively, realistic, but seldom are both. It can be seen from the preceding discussion that attempts to compute photosyn- thetic rates from temperature would generally be inaccurate, and unrealistic as well, unless radiant energy and concentrations of essential nutrients were also considered. In the past, models of photosynthesis have often included a term for the maximum rate of photosynthesis at light saturation which is widely acknowledged to be temperature-dependent. In Steele's (1962) model Pmax is a constant and is expressed in units "carbon assimilation rate per unit plant carbon" Table 6. — Assimilation numbers measured in different ocean regions in comparison with maximum expected values taken from Figure 9. A similar table is given by Saijo and Ichimura (1962) for pelagic and coastal seawaters and lakes. Assimilation number (mg C/mg Chl./hr) Temper- ature CO Region — Measured Max. expected if C/Chl. = Source 30 60 90 Cold Seas Antarctic avg <2.5 1.0 2J0 3.0 -2-2 (1) Subarctic North Pacific 0.4-2 1.4 2.7 4.0 2-6 (2) North Atlar\tic 3.5 1.4 2.7 4.0 A-6 (3) 3.5 1.7 3.4 5.1 9 (3) 4 3.1 6.3 9.4 16 (3) Upwelling Regions Peru Current <7.5 4.6 9.2 17-20 (4) Peru Current 5 4.6 9.2 M20) (5) S.W. Africa <6.5 4.6 9.2 1(20) (6) Cromwell Current 5.3 5.1 10.3 21 (7) 10 8.0 16.0 25 (7) Tropical Seas Madagascar avg 3.8 8.0 16j0 M25) (8) Caribbean avg 6.3a 8.0 16.0 1(25) (9) Tropical Pacific avg 3.7b avg 2.3a 8.0 16.0 M25) (9) Off Puerto Rico avg 1.5b <13 8.0 16.0 H25) (10) Western Arabian Sea avg 4.4 11.7 23.4 <28 (11) ■> Assumed temperature. Sources: (1) Saijo and Kawashima, 1964; El-Sayed and Mandelli, 1965; Mandelli and Burkholder, 1966; Home et al. 1969; Bunt and Lee, 1970; (2) Biological station, Nanoimo (1970. Biological, chemicol and physical data First Canadian Trans-Pacific Oceonographic Cruise March to May 1969. Fish. Res. Board Can., Manuscr. Rep. 1080, 92 p). (3) Steemann Nielsen and Hansen, 1959, for light-saturated rate; (5) Lorenzen, 1968, average over the euphotic zone; (7) Barber and Ryther, 1969, average over the euphotic zone; (8) Sournia, 1968; (9) Malone, '971a. Values designated by "a" are for nanoplankton, "b" values for netplankton; (10) Burkholder, Burkholder, and Almodovar, 1967; (11) Ryther and Menzel, 1965b, average for euphotic zone. 1079 FISHERY BULLETIN: VOL. 70, NO. 4 and is equivalent to a specific growth rate of about 1.1 doubling-s/day. Such a value would be appropriate for temperate waters, but prob- ably not for polar or eutrophic tropical waters. But to make Pmax a function of temperature would probably add unnecessary complexity for modeling purposes, although it would add real- ism. However, the use of constant values makes the model restrictive geographically (see, for example. Parsons and Anderson's, 1970, use of the model of Steele and Menzel, 1962, for the subarctic North Pacific). A plant physiologist would perhaps prefer to approach modeling phytoplankton growth in the sea in as physiologically realistic way as possible and to let the computer handle the complexity. But it is questionable how realistically this can now be accomplished or what insight would thereby result. Equation (1) of this paper can be considered a model of sorts and its apparent universality is appealing. Comparing its predictions as to growth rate and assimilation number with data from natural phytoplankton shows, moreover, the magnitude of diflference between potential plant growth and reality, as it is now best esti- mated. The gulf between real and maximum ex- pected values shows how significant are the other environmental factors which affect phytoplank- ton: radiant energy, nutrient concentrations, grazing, and mixing processes. All of these parameters have been successfully treated in models since the 1940's (see Patten's summary review, 1968 ; Parsons, Giovando, and LeBras- seur, 1966; Dugdale and Goering, 1967). A physiologically realistic model might begin with a relation between temperature and max- imum expected growth rate, as in Eppley and Sloan (1966). In that paper the variations in growth rate among species were rationalized by including the chlorophyll concentration per unit cell volume (a parameter not readily measur- able in assemblages of mixed species, but sus- ceptible to averaging) . This parameter seemed also to compensate for the sun-shade alterations of phytoplankton photosynthesis when used to calculate radiant energy absorbed by a cell's pig- ments. However, the problem of daylength could not be adequately handled for species which grow faster with a few hours darkness each day than in continuous light. None of the models proposed for primary productivity simulation has attempted to treat diel periodicity in the metabolic processes of phytoplankton. Nor is the alteration of chem- ical composition attendant to growth with lim- iting concentrations of nutrients or to variations with irradiance or temperature treated. One suspects that the simple models now available can be satisfactory for describing the major features of regional phytoplankton pro- duction. Realistic physiological models will probably remain in the "special purpose" cate- gory for the insight of those familiar enough with the subject to use them as guide to their own research. Nevertheless, it is admitted, giv- en the current popularity of modeling, that neither the reader nor the author may be able to resist for long the temptation to combine Equation (1) with a realistic function for nu- trient assimilation rate vs. ambient concentra- tion, a function for the dependence of ix and assimilation number upon light, and a suitable function for describing eflTects of mixing, in line with critical depth theory, and to try it with his favorite set of oceanic data. ACKNOWLEDGMENTS I am grateful to Mrs. Elizabeth Stewart for computer calculations and graphs, to Mrs. Vir- ginia Moore for drawing the inked figures, and to Ms. Janice Walker for typing the manuscript. I thank my colleagues Dr. 0. Holm-Hansen, David Checkley, and Dr. James T. McCarthy for use of unpublished data, and E. H. Renger and Mrs. Gail Hirota for expert analytical services. This study was supported by the U.S. Atomic Energy Commission Contract No. AT (11-1) GEN 10, P.A. 20. LITERATURE CITED Antia, N. J., C. D. McAllister, T. R. Parsons, K. Stephens, and J. D. H. Strickland. 1963. Further measurements of primary produc- tion using a large-volume plastic sphere. Limnol. Oceanogr. 8:166-183. 1080 EPPLEY: PHYTOPLANKTON AND TEMPERATURE ARUGA, Y. 1965a. Ecological studies of photosynthesis and matter production of phytoplankton. I. Seasonal changes in photosynthesis of natural phytoplank- ton. Bot. Mag. (Tokyo) 78:280-288. 1965b. Ecological studies of photosynthesis and matter production of phytoplankton. II. Photo- synthesis of algae in relation to light intensity and temperature. Bot. Mag. (Tokyo) 78:360-365. Barber, R. T., and J. H. Ryther. 1969. Organic chelators: Factors affecting pri- mary production in the Cromwell Current up- welling. J. Exp. Mar. Biol. Ecol. 3:191-199. Barlow, J. P., C. J. Lorenzen, and R. T. Myren. 1963. Eutrophication of a tidal estuary. Limnol. Oceanogr. 8:251-262. Beers, J. R., M. R. Stevenson, R. W. Eppley, and E. R. Brooks. 1971. Plankton populations and upwelling off the coast of Peru, June 1969. Fish. Bull., U.S. 69: 859-876. Braarud, T. 1961. Cultivation of marine organisms as a means of understanding environmental influences on pop- ulations. In M. Sears (editor), Oceanography, p. 271-298. AAAS (Am. Assoc. Adv. Sci.), Publ. 67. Brock, T. D. 1971. Microbial growth rates in nature. Bact. Rev. 35:39-58. Bunt, J. S., and C. C. Lee. 1970. Seasonal primary production in Antarctic sea ice at McMurdo Sound in 1967. J. Mar. Res. 28:304-320. Burkholder, p. R., L. M. Burkholder, and L. R. Almodovar. 1967. Carbon assimilation of marine flagellate blooms in neritic waters of southern Puerto Rico. Bull. Mar. Sci. 17:1-15. Cannon, D., J. W. G. Lund, and J. Sieminska. 1961. The growth of TabeUaria flocculosa (Roth) Kutz. var. flocntlosa (Roth) Knuds. under nat- ural conditions of light and temperature. J. Ecol. 49:277-287. Caperon, J., S. A. Cattell, and G. Krasnick. 1971. Phytoplankton kinetics in a subtropical estuary: Eutrophication. Limnol. Oceanogr. 16: 599-60*7. Castenholz, R. W. 1964. The effect of daylength and light intensity on the growth of littoral marine diatoms in cul- ture. Physiol. Plant. 17:951-963. 1969. Thermophilic blue-green algae and the ther- mal environment. Bact. Rev. 33:476-504. Curl, H., Jr., and L. F. Small. 1965. Variations in photosynthetic assimilation ratios in natural, marine phytoplankton com- munities. Limnol. Oceanogr. 10(Suppl.) :R67- R73. Gushing, D. H. 1959. The estimation of carbon in phytoplankton. Cons. Perm. Int. Explor. Mer, Rapp. P.-V. Reun. 144:32-33. Gushing, D. H., and H. F. Nicholson. 1966. Method of estimating algal production rates at sea. Nature (Lond.) 212:310-311. DUGDALE, R. C, AND J. J. GOERING. 1967. Uptake of new and regenerated forms of nitrogen in primary production. Limnol. Ocean- ogr. 12:196-206. DUGDALE, R. G., AND J. J. MacISAAC. 1971. A computation model for the uptake of ni- trate in the Peru upwelling region. Invest. Pesq. 35:299-308. El-Sayed, S. Z., and E. F. Mandelli. 1965. Primary production and standing crop of phytoplankton in the Weddell Sea and Drake Passage. Antarct. Res. Ser. 5:87-106. El-Sayed, S. Z., E. F. Mandelli, and Y. Sugimura. 1964. Primary organic production in the Drake Passage and Bransfield Strait. Antarct. Res. Ser. 1:1-11. Eppley, R. W. 1963. Evaluation of certain marine algal flagellates for mass culture. Tech. Doc. Rep. No. SAM- TDR-63-91. USAF School of Aerospace Medi- cine. Brooks Air Force Base, Tex. 1968. An incubation method for estimating the carbon content of phytoplankton in natural sam- ples. Limnol. Oceanogr. 13:574-582. Eppley, R. W., A. F. Carlucci, 0. Holm-Hansen, D. KiEFER, J. J. McCarthy, and P. M. Williams. In press. Evidence for eutrophication in the sea near southern California coastal sewage outfalls, July, 1970. Calif. Coop. Oceanic Fish. Invest., Rep. Eppley, R. W., R. W. Holmes, and E. Paasche. 1967. Periodicity in cell division and physiological behavior of Ditylum brightwellii, a marine plank- tonic diatom, during growth in light-dark cycles. Archiv. Microbiol. 56:305-323. Eppley, R. W., and F. M. Macias R. 1962. Rapid growth of sewage lagoon Chlamydom- onas with acetate. Physiol. Plant. 15:72-79. Eppley, R. W., F. M. H. Reid, and J. D. H. Strickland. 1970. Estimates of phytoplankton crop size, growth rate, and primary production. In J. D. H. Strick- land (editor), The ecology of the plankton off La JoUa, California, in the period April through September, 1967, p. 33-42. Bull. Scripps Inst. Oceanogr. Univ. Calif. 17. Eppley, R. W., J. N. Rogers, and J. J. McCarthy. 1969. Half-saturation constants for uptake of ni- trate and ammonium by marine phytoplankton. Limnol. Oceanogr. 14:912-920. 1081 FISHERY BULLETIN: VOL. 70, NO. 4 Eppley, R. W., and p. R. Sloan. 1965. Carbon balance experiments with marine phytoplankton. J. Fish. Res. Board Can. 22:1083- 1097. 1966. Growth rates of marine phytoplankton: Correlation with light absorption by cell chloro- phyll a. Physiol. Plant. 19:47-59. Glooschenko, W. a., and H. Curl, Jr. 1971. Influence of nutrient enrichment on photo- synthesis and assimilation ratios in natural North Pacific phytoplankton communities. J. Fish. Res. Board Can. 28:790-793. GoERiNG, J. J., D. D. Wallen, and R. a. Nauman. 1970. Nitrogen uptake by phytoplankton in the discontinuity layer of the eastern subtropical Pa- cific Ocean. Limnol. Oceanogr. 15:789-796. GuiLLARD, R. R. L., and S. Myklestad. 1970. Osmotic and ionic requirements of the ma- rine centric diatom Cyclotella nana. Helgolander wiss. Meeresunters. 20:104-110. GuiLLARD, R. R. L., and J. H. Ryther. 1962. Studies of marine planktonic diatoms. I. Cyclotella nana Hustedt and Detonula confervacea (Cleve) Gran. Can. J. Microbiol, 8:229-239. Happey, C. M. 1970. The effects of stratification on phytoplank- tonic diatoms in a small body of water. J. Ecol. 58:635-651. Harris, E., and G. A. Riley. 1956. Oceanography of Long Island Sound, 1952- 1954. VIII. Chemical composition of the plankton. Bull. Bingham Oceanogr. Collect. Yale Univ. 15:315-323. Hepher, B. 1962. Primary production in fishponds and its ap- plication to fertilization experiments. Limnol. Oceanogr. 7:131-136. HoBSON, L. A. 1971. Relationships between particulate organic carbon and micro-organisms in upwelling areas off southwest Africa. Invest. Pesq. 35:195-208. Holm-Hansen, 0. 1969. Determination of microbial biomass in ocean profiles. Limnol. Oceanogr. 14:740-747. Holm-Hansen, O., and C. R. Booth. 1966. The measurement of adenosine triphosphate in the ocean and its ecological significance. Lim- nol. Oceanogr. 11:510-519. HOOGENHOUT, H., AND J. AmESZ. 1965. Growth rates of photosynthetic microorgan- isms in laboratory cultures. Arch. Mikrobiol. 50:10-24. HoRNE, A. J., G. E. Fogg, and D. J. Eagle. 1969. Studies in situ of the primary production of an area of inshore Antarctic sea. J. Mar. Biol. Assoc. U.K. 49:393-405. HULBURT, E. M., and R. R. L. GuiLLARD. 1968. The relationship of the distribution of the diatom Skeletonema tropicum to temperature. Ecology 49:337-339. ICHIMURA, S. 1960. Photosj-nthesis pattern of natural phyto- plankton relating to light intensity. Bot. Mag. (Tokyo) 73:458-467. 1967. Environmental gradient and its relation to primary productivity in Tokyo Bay. Rec. Ocean- ogr. Works Jap., New Ser. 9:115-128. 1968. Phytoplankton photosynthesis. In D. F. Jackson (editor). Algae, man, and the environ- ment, p. 103-120. Syracuse Univ. Press. Syra- cuse, N.Y. ICHIMURA, S., AND Y. ARUGA. 1964. Photosynthetic natures of natural algal com- munities in Japanese waters. In Y. Miyake and T. Koyama (editors), Ken Sugawara festival vol- ume. Recent researches in the fields of hydro- sphere, atmosphere and nuclear geochemistry, p. 13-37. Maruzen Co., Tokyo. IGNATIADES, L., AND T. J. SmAYDA. 1970. Autecological studies on the marine diatom Rhizosolenia fragilissima Bergon. I. The influ- ence of light, temperature, and salinity. J. Phy- col. 6:332-339. Jenkin, P. M. 1937. Oxygen production by the diatom Coscino- disciis excentricKS Ehr. in relation to submarine illumination in the English Channel. J. Mar. Biol. Assoc. U.K. 22:301-343. J0RGENSEN, E. G. 1968. The adaptation of plankton algae. II. As- pects of the temperature adaptation of Skeleto- nema costatum. Physiol. Plant. 21:423-427. Ketchum, B. H., L. Lillick, and A. C. Redfield. 1949. The growth and optimum yields of unicel- lular algae in mass culture. J. Cell. Comp. Phy- siol. 33:267-279. Kovala, p. E., and J. D. Larrance. 1966. Computation of phytoplankton cell numbers, cell volume, cell surface and plasma volume per liter, from microscopical counts. Univ. Wash. Dep. Oceanogr., Spec. Rep. 38, 21 p. Lanskaya, L. a. 1961. Fission rate of plankton algae of the Black Sea in cultures. In C. H. Oppenheimer (editor). Marine microbiology, p. 127-132. C. C. Thomas, Springfield, 111. LORENZEN, C. J. 1968. Carbon/chlorophyll relationships in an up- welling area. Limnol. Oceanogr. 13:202-204. Lund, J. W. G. 1971. An artificial alteration of the seasonal cycle of the plankton diatom Melosira italica subsp. subarctica in an English lake. J. Ecol. 59:521- 533. 1082 EPPLEY: PHYTOPLANKTON AND TEMPERATURE McAllister, C. D., T. R. Parsons, K. Stephens, and J. D. H. Strickland. 1961. Measurements of primary production in coastal sea water using a large-volume plastic sphere. Limnol. Oceanogr. 6:237-258. McAllister, C. D., T. R. Parsons, and J. D. H. Strickland. 1960. Primary productivity and fertility at Station "P" in the north-east Pacific Ocean. J. Cons. 25:240-259. McAllister, C. D., N. Shah, and J. D. H. Strickland. 1964. Marine phytoplankton photosynthesis as a function of light intensity: a comparison of methods. J. Fish. Res. Board Can. 21:159-181. McCarthy, J. J. 1971. The role of urea in marine phytoplankton ecology. Ph.D. Thesis, Univ. California, San Diego. McLachlan, J. 1960. The culture of Dunaliella tertiolecta Butcher — a euryhaline organism. Can. J. Microbiol. 6: 367-379. McLaren, I. A. 1963. Effects of temperature on growth of zoo- plankton, and the adaptive value of vertical mi- gration. J. Fish. Res. Board Can. 20:685-727. Malone, T. C. 1971a. The relative importance of nannoplankton and netplankton as primary producers in tropical oceanic and neritic phytoplankton communities. Limnol. Oceanogr. 16:633-639. 1971b. The relative importance of nannoplankton and netplankton as primary producers in the Cal- ifornia Current system. Fish. Bull., U.S. 69:799- 820. 1971c. Diurnal rhythms in netplankton and nan- noplankton assimilation ratios. Mar. Biol. (Berl.) 10:285-289. Mandelli, E. F., and p. R. Burkholder. 1966. Primary productivity in the Gerlache and Bransfield straits of Antarctica. J. Mar. Res. 24:15-27. Mandelli, E. F., P. R. Burkholder, T. E. Doheny, and R. Brody. 1970. Studies of primary productivity in coastal waters of southern Long Island, New York. Mar. Biol. (Berl.) 7:153-160. Mullin, M. M., p. R. Sloan, and R. W. Eppley. 1966. Relationship between carbon content, cell volume, and area in phytoplankton. Limnol. Oceanogr. 11:307-311. Munk, W. H., and G. a. Riley. 1952. Absorption of nutrients by aquatic plants. J. Mar. Res. 11:215-240. Myers, J., and J. R. Graham. 1959. On the mass culture of algae. IL Yield as a function of cell concentration under continuous sunlight irradiance. Plant Physiol. 34:345-352. Paasche, E. 1967. Marine plankton algae grown with light- dark cycles. I. Coccolithns huxleyi. Physiol. Plant. 20:946-956. 1968. Marine plankton algae grown with light- dark cycles, n. Ditylum hrightwellii and Nitzsch- ia turgidtda. Physiol. Plant. 21:66-77. Parsons, T. R. 1965. An automated technique for determining the growth rate of chain-forming phytoplankton. Limnol. Oceanogr. 10:598-602. Parsons, T. R., and G. C. Anderson. 1970. Large scale studies of primary production in the North Pacific Ocean. Deep-Sea Res. 17: 765-776. Parsons, T. R., L. F. Giovando, and R. L. LeBrasseur. 1966. The advent of the spring bloom in the eastern subarctic Pacific Ocean. J. Fish. Res. Board Can. 23:539-546. Patten, B. C. 1968. Mathematical models of plankton production. Int. Rev. gesamten Hydrobiol. 53:357-408. Pechlaner, R. 1970. The phytoplankton spring outburst and its conditions in Lake Erken (Sweden). Limnol. Oceanogr. 15:113-130. Perry, M. J. In press. Alkaline phosphatase activity in subtrop- ical central North Pacific waters using a sensitive fluorometric method. Mar. Biol. (Berl.) Polikarpov, G. G., and A. V. Tokareva. 1970. The cell cycle of the dinoflagellates Peridin- iutn trochoideum (Stein) and Gonyaulax polyedra. Hydrobiol. J. 6(5):54-57. Reid, F. M. H., E. Fuglister, and J. B. Jordan. 1970. Phytoplankton taxonomy and standing crop. In J. D. H. Strickland (editor), The ecology of the plankton off La Jolla, California, in the period April through September, 1967, p. 51-66. Bull. Scripps Inst. Oceanogr. Univ. Calif. 17. Riley, G. A., H. Stommel, and D. F. Bumpus. 1949. Quantitative ecology of the plankton of the western North Atlantic. Bull. Bingham Ocean- ogr. Collect. Yale Univ. 12(3):1-169. RODHE, W. 1965. Standard correlations between pelagic photo- synthesis and light. In C. R. Goldman (editor), Primary productivity in aquatic environments, p. 365-381. Mem. 1st. Ital. Idrobiol. 18(Suppl.). Ryther, J. H., and D. W. Menzel. 1965a. Comparison of the i^C-technique with di- rect measurement of photosynthetic carbon fixa- tion. Limnol. Oceanogr. 10:490-491. 1965b. On the production, composition, and distri- bution of organic matter in the Western Arabian Sea. Deep-Sea Res. 12:199-209. 1083 FISHERY BULLETIN: VOL. 70, NO. 4 Ryther, J. H., AND C. S. Yentsch. 1958. Primary production of continental shelf waters off New York. Limnol. Oceanogr. 3:327- 335. Saijo, Y., and S. Ichimura. 1962. Some considerations on photosynthesis of phytoplankton from the point of view of produc- tivity measurement. J. Oceanogr. Soc. Jap., 20th Anniv. Vol.: 687-693. Saijo, Y., and T. Kawashima. 1964. Primary production in the Antarctic Ocean. J. Oceanogr. Soc. Jap. 19:22-48. Saijo, Y., and M. Sakamoto. 1964. Photosynthetic production of phytoplankton in some ice covered lakes in the Central Japan. In Y. Miyake and T. Koyama (editors), Ken Sugawara festival volume. Recent researches in the fields of hydrosphere, atmosphere and nuclear geochemistry, p. 289-303. Maruzen Co., Tokyo. Sheldon, R. W., and T. R. Parsons. 1967. A practical manual on the use of the Coulter Counter in marine science. Coulter Electronic Sales Co. of Canada, Toronto, 66 p. Smayda, T. J. 1957. Phytoplankton studies in lower Narragan- sett Bay. Limnol. Oceanogr. 2:342-359. 1969. Experimental observations on the influence of temperature, light, and salinity on cell division of the marine diatom, Detonula confervacea (Cleve) Gran. J. Phycol. 5:150-157. 1970. The suspension and sinking of phytoplankton in the sea. Oceanogr. Mar. Biol. Annu. Rev. 8: 353-414. SOROKIN, C, and R. W. KRAUSS. 1958. The effects of light intensity on the growth rates of green algae. Plant Physiol. 33:109-113. 1962. Effects of temperature & illuminance on Chlorella growth uncoupled from cell division. Plant Physiol. 37:37-42. Sournia, a. 1968. Variations saissonieres et nycthemerales du phytoplancton marin et de la production primaire dans une bale tropicale, a Nosey-Be (Madagas- car). Int. Rev. gesamten Hydrobiol. 53:1-76. Steele, J. H. 1962. Environmental control of photosynthesis in the sea. Limnol. Oceanogr. 7:137-150. Steele, J. H., and I. E. Baird. 1961. Relations between primary production, chloro- phyll and particulate carbon. Limnol. Oceanogr. 6:68-78. 1962. Further relations between primary produc- tion, chlorophyll, and particulate carbon. Limnol. Oceanogr. 7:42-47. Steele, J. H., and D. W. Menzel. 1962. Conditions for maximum primary production in the mixed layer. Deep-Sea Res. 9:39-49. Steemann Nielsen, E. 1952. The use of radio-active carbon (C^*) for measuring organic production in the sea. J. Cons. 18:117-140. 1960. Productivity of the oceans. Annu. Rev. Plant Physiol. 11:341-362. Steemann Nielsen, E., and V. K. Hansen. 1959. Light adaptation in marine phytoplankton populations and its interrelation with tempera- ture. Physiol. Plant. 12:353-370. 1961. Influence of surface illumination on plankton photosynthesis in Danish waters (56° N) through- out the year. Physiol. Plant. 14:595-613. Steemann Nielsen, E., and E. G. J0rgensen. 1968a. The adaptation of plankton algae. I. Gen- eral part. Physiol. Plant. 21:401-413. 1968b. The adaptation of plankton algae. III. With special consideration of the importance in nature. Physiol. Plant. 21:647-654. Strathmann, R. R. 1967. Estimating the organic carbon content of phytoplankton from cell volume or plasma vol- ume. Limnol. Oceanogr. 12:411-418. Strickland, J. D. H. (editor). 1970. The ecology of the plankton off La Jolla, Cal- ifornia in the period April through September, 1967. Bull. Scripps Inst. Oceanogr. Univ. Calif. 17:1-103. Strickland, J. D. H., R. W. Eppley, and B. Rojas de Mendiola. 1969. Phytoplankton populations, nutrients and photosynthesis in Peruvian coastal waters. [In Spanish and English.] Bol. Inst. Mar Peru (Callao) 2:1-45. Strickland, J. D. H., 0. Holm-Hansen, R. W. Eppley, AND R. J. Linn. 1969. The use of a deep tank in plankton ecology. I. Studies of the growth and composition of phy- toplankton crops at low nutrient levels. Limnol. Oceanogr. 14:23-34. Sutcliffe, W. H., Jr., R. W. Sheldon, and A. Prakash. 1970. Certain aspects of production and standing stock of particulate matter in the surface waters of the northwest Atlantic Ocean. J. Fish. Res. Board Can. 27:1917-1926. Sweeney, B. M., and J. W. Hastings. 1958. Rhythmic cell division in populations of Gonyaulax polyedra. J. Protozool. 5:217-224. Swift, E., 5th, and W. R. Taylor. 1966. The effect of pH on the division rate of the coccolithophorid Cricosphaera elongata. J. Phy- col. 2:121-125. Talling, J. F. 1955. The relative growth rate of three plankton diatoms in relation to underwater radiation and temperature. Ann. Bot., New Ser. 19:329- 341. 1084 EPPLEY: PHYTOPLANKTON AND TEMPERATURE 1957. Photosynthetic characteristics of some fresh- water plankton diatoms in relation to underwater radiation. New Phytol. 56:29-50. Tamiya, H., T. Sasa, T. Nihei, and S. Ishibashi. 1955. Effect of variation of daylength, day and night-temperatures, and intensity of daylight up- on the growth of Chlorella. J. Gen. Appl. Mi- crobiol. 4:298-307. Terborgh, J., AND K. V. Thimann. 1964. Interactions between day-length and light intensity in the growth and chlorophyll content of Acetabularia crenulata. Planta (Berl.) 63: 83-98. Thomas, W. H. 1966. Effects of temperature and illuminance on cell division rates of three species of tropical oceanic phytoplankton. J. Phycol. 2:17-22. 1969. Phytoplankton nutrient enrichment exper- iments off Baja California and in the eastern equatorial Pacific Ocean. J. Fish. Res. Board Can. 26:1133-1145. 1970a. On nitrogen deficiency in tropical Pacific Ocean phytoplankton : photosynthetic parameters in poor and rich water. Limnol. Oceanogr. 15: 380-385. 1970b. Effect of ammonium and nitrate concentra- tion on chlorophyll increases in natural tropical Pacific phytoplankton populations. Limnol. Oceanogr. 15:386-394. Thomas, W. H., and A. N. Dodson. In press. On nitrogen deficiency in tropical Pacific oceanic phytoplankton. II. Photosynthetic and cellular characteristics of a chemostat-grown di- atom. Limnol. Oceanogr. Thomas, W. H., and R. W. Owen, Jr. 1971. Estimating phytoplankton production from ammonium and chlorophyll concentrations in nu- trient-poor water of the eastern tropical Pacific Ocean. Fish. Bull., U.S. 69:87-92. Ukeles, R. 1961. The effect of temperature on the growth and survival of several marine algal species. Biol. Bull. (Woods Hole) 120:255-264. University of Washington. 1970. Biological production in upwelling ecosys- tems. Data Report. R/V T. G. Thompson Cruise 36 (PISCO). Univ. Wash. Dep. Oceanogr., Spec. Rep. 42, 97 p. Werner, D. 1970. Productivity studies on diatom cultures. Helgolander wiss. Meeresunters. 20:97-103. Williams, R. B. 1964. Division rates of salt marsh diatoms in re- lation to salinity and cell size. Ecology 45:877- 880. Williams, R. B., and M. B. Murdoch. 1966. Phytoplankton production and chlorophyll concentrations in the Beaufort Channel, North Carolina. Limnol. Oceanogr. 11:73-82. Wright, J. C. 1959. Limnology of Canyon Ferry Reservoir. II. Phytoplankton standing crop and primary pro- duction. Limnol. Oceanogr. 4:235-245. Yentsch, C. S., and R. W. Lee. 1966. A study of photosynthetic light reactions and a new interpretation of sun and shade phyto- plankton. J. Mar. Res. 24:319-337, Yentsch, C. S., and J. H. Ryther. 1959. Relative significance of the net phytoplank- ton and nanoplankton in the waters of Vineyard Sound. J. Cons. 24:231-238. Zeitzschel, B. 1970. The quantity, composition and distribution of suspended particulate matter in the Gulf of California. Mar. Biol. (Berl.) 7:305-318. 1085 COMPARATIVE STUDY OF FOOD OF SKIPJACK AND YELLOWFIN TUNAS OFF THE COAST OF WEST AFRICA' Alexander Dragovich and Thomas Potthofp^ ABSTRACT Stomach contents of 711 skipjack tuna (Katsmvonus pelamis) and 132 yellowfin tuna {Thunnus albacares) captured in 1968 by live bait and trolling off the coast of West Africa were examined. A marked taxonomic similarity was noted between the organisms in the diets of the two tunas. Frequency of occurrence, displacement volume, and num- bers of each food item identified are presented for each species of tuna. Fishes, mollusks, and crustaceans were the principal foods with fishes generally dominant. The most prominent fish families were Acanthuridae, Carangidae, Dactylopteridae, Gempylidae, Gonostomatidae, Lutjanidae, MuUidae, Priacanthidae, Scombridae, Serranidae, and Trichiuridae; mollusks were chiefly cephalopods (squids), and crustaceans consisted mostly of macrozooplankton. Juvenile tunas were present in the diet of both species of tunas. Estimates of the size of forage organisms were primarily based on displacement vol- umes. In the majority of observations, food organisms displaced less than 1.0 ml and the displacement volumes of stomach contents varied for skipjack tuna from 0.1 to 20.0 ml and for yellowfin tuna from 0.1 to 60.0 ml. Spearman's rank correlation analysis was used to test for a relation between the food type (in volume and frequency of occurrence) and the lengths of skipjack and yellowfin tunas. Significant correlations were noted between the size of skipjack tuna and both the volume and the frequency of occurrence of forage fish. A comparison between the findings of our study and that of other food studies off the coast of West Africa showed greater taxonomic similarity in tuna forage when the studies were made in the same general area and that only several types of food were of primary importance in each given area. Seasonal changes in taxonomic composition of forage organisms were also discussed. The method used to evaluate food organisms consisted of ranking the organisms ac- cording to their dispersal indices, abundance indices, and biomass contribution. Stomato- pods, the amphipod Phrosina semiluyiata, Teuthoidea, Carangidae, Serranidae, and megalopal stages were most important constituents of food throughout the investigation area. The principal surface tuna fishery in the trop- ical Atlantic Ocean is located off the coast of West Africa (Jones, 1969). One of the major tasks of the Southeast Fisheries Center, Miami Laboratory, has been the study of the biology and ecology of tunas and tunalike fishes in the ' Contribution No. 218, National Marine Fisheries Service, Southeast Fisheries Center, Miami Laboratory, Miami, Fla. ^ National Marine Fisheries Service, Southeast Fish- eries Center, Miami Laboratory, 75 Virginia Beach Drive, Miami, FL 33149. Manuscript accepted February 1972. FISHERY BULLETIN: VOL. 70, NO. 4, 1972. tropical Atlantic Ocean. In view of the recog- nized importance of food as an ecological factor in the life history of tunas, one project of this investigation consisted of a study of the food and feeding habits of skipjack (Katsmvonus pelamis) and yellowfin {Thunmis albacares) tunas — the two predominant species in com- mercial catches in those waters. We describe and compare the food of skipjack and yellowfin tunas and discuss the relative im- portance of different forage organisms. We compare our findings with those of other investi- gators working in the same general area. This 1087 FISHERY BULLETIN: VOL. 70, NO. 4 information may be used to study the relation- ship betv\^een the distribution of food organisms and occurrence of tuna schools. Most of the information up to 1969 on food of various tunas off the west coast of Africa may be found in the review of studies of tuna food in the Atlantic Ocean by Dragovich (1969) . Dragovich (1970) also reported on the food of skipjack and yellowfin tunas off the west coast of Africa. MATERIALS AND METHODS Samples on which the present report is based were collected during February, March, April and September, October, November of 1968 on two cruises (UN6801 and UN6802) of the re- search vessel Undaunted of the Bureau of Com- mercial Fisheries (now National Marine Fisher- ies Service) (Figure 1). All tunas sampled for 10- 5- 0- 5- 10° 15- 20' Figure 1. — Shaded area shows lo- calities where stomachs of skipjack and yellowfin tunas were collected. 20 1088 DRAGOVICH and POTTHOFF: FOOD OF SKIPJACK AND YELLOWFIN TUNAS this study were caught by pole and line or by trolling (Table 1) . A total of 711 stomachs from skipjack tuna and 132 from yellowfin tuna were examined. The skipjack tuna studied varied in fork length from 36 to 63 cm and the yellowfin tuna from 52 to 94 cm (Figure 2), Sampling of catches for stomach samples was carried out as other requirements of the program and circumstances permitted. Immediately after completion of the morphometric work aboard the ship the stomachs were removed by opening the abdominal cavity and by severing them from the intestine and the esophagus. Each stomach was pierced in several places to allow Table 1. — Distribution of skipjack and yellowfin tuna stomachs collected during 1968 from the eastern tropical Atlantic Ocean, identified by month, cruise, and method of capture. February UN680U March UN6801 April UN6801 September October UN6a022 UN6802 November UN6802 Tota With food Empty With food Empty With food Empty With food With Empty food Empty With food Empty With food Empty Method of capture Skipjack tuna 20 78 292 36 70 4 104 69 25 5 511 142 Live bait 41 8 3 3 3 Yeillowfin tuna 47 11 Trolling 67 24 4 18 1 109 5 Live bait 4 9 1 3 1 17 1 Trolling 1 UN68CI) = RV Undaunted 6831 cruise. 2 UN68a2 = RV Undaunted 6832 cruise. SKIPJACK z HI UJ Q. (/) u. O 70 cc UJ '0- OQ S so Z 40 30 JO- 10- YELLOWFIN UN-6801 N=431 UN-6801 N=84 39 41 43 45 47 49 SI S3 5S 57 59 52 54 56 51 to 62 64 66 61 70 72 74 ^ IS UN-6802 N=280 ''^T■ 3t 40 42 44 46 41 50 52 S4 56 St 60 62 63 65 67 69 71 73 75 77 79 «1 «3 IS 87 19 91 93 FORK LENGTH (cm) Figure 2. — Length-frequency distribution of skipjack tuna and yellowfin tunas from which stomachs were collected. 1089 FISHERY BULLETIN: VOL. 70, NO. 4 penetration of Formalin' and placed in a labeled polyethelene bag containing lO-^r Formalin. In the laboratory the stomachs were first clas- sified into those containing food and those that were empty. The stomach contents were then identified to the lowest possible taxonomic units which were subsequently sorted, counted, and their displacement volumes measured. Length measurements were taken of many forage or- ganisms, particularly fishes. Bait fishes were found in some of the stomachs, but they were not considered as part of the regular diet of skipjack and yelloA\iin tunas; therefore, stom- achs which contained only bait were considered empty. Stomachs that contained parasitic trem- atodes were also considered empty. This study was no exception in regard to dif- ficulties encountered in the identification of for- age organisms (Dragovich, 1969) . In numerous instances the identification of ingested fishes, particularly juvenile tunas, was made from ver- tebrae using methods employed by PotthoflP and Richards (1970) . Cephalopod identification was particularly difficult since many diagnostic ex- ternal characters usually are the first destroyed during digestion. The following methods of analysis were used: 1) the volumetric method — the individual vol- ume of each taxon and the total aggregate vol- ume of broad taxonomic groups, 2) frequency of occurrence method — the frequency of occur- rence of a food item and of broad taxonomic groups, and 3) numerical method — number of individuals in the same taxonomic group. Spearman's rank correlation test, x" test of homogeneity, and paired ^-test of diff"erence be- tween the means were used. A method consist- ing of ranking of food organisms according to their geographic distribution, relative abun- dance, and biomass was also employed. COMPOSITION OF FOOD Fishes, crustaceans, and cephalopods were the three principal food categories found in stom- ' Use of trade names does not imply endorsement by the National Marine Fisheries Service. achs of both skipjack and yellowfin tunas (Fig- ure 3). Food items that do not fall into these three categories consisted of mollusks other than cephalopods, salps, polychaetes, and siphono- phores. Other mollusks and salps were found in both species of tunas; polychaetes and siphono- phores were present only in stomachs of skip- jack tunas. A checklist of all food items, num- ber of organisms, frequency of occurrence, dis- placement volumes, and length measurements of some organisms are presented according to the cruises in Appendix Tables 1 to 4. Fishes were represented in the diet of skipjack and yellow- fin tunas by 90 different taxa, crustaceans by 45, and mollusks by 24. The percentage composition of five food cat- egories in terms of number, volume, and fre- quency of occurrence is shown in Figure 3. Fish was the dominant food item by volume for both species of tunas, except for yellowfin tuna cap- tured during UN6802, when cephalopods were dominant. Fish occurred most frequently in the diet of both species of tunas sampled during UN6801; however, crustaceans occurred most often in the collections from UN6802. In the diet of yellowfin tuna, fishes were numerically the most important food items during both cruis- es; in the diet of skipjack tuna, fishes were the most important by numbers during UN6801, but crustaceans were most numerous during UN6802. The group of forage organisms classed as other mollusks consisted primarily of pteropods and heteropods. Salps, polychaetes, and siphono- phores were the principal components of the group of forage organisms classed as miscella- neous— this group was not prominent by vol- ume, frequency of occurrence, or by numbers. FISHES Fishes utilized as food consisted mainly of postlarval and juvenile forms of pelagic and reef fishes. Some adult fishes, primarily Vinciguerria nimharui, were also present in the diet of both species of tuna. Although fishes were repre- sented by a larger number of families, only a few families were important in terms of volume, frequency of occurrence, and numbers. 1090 DRAGOVICH and POTTHOFF: FOOD OF SKIPJACK AND YELLOWFIN TUNAS NUMBER SKIPJACK TUNA UN -6801 VOLUME FREQUENCY OF OCCURRENCE Fishes Crustaceans I Cephalopods | Other mollusks | Miscellaneous | Fishes Crustaceans Cephalopods | Other mollusks I Miscellaneous I UN -6802 I I YELLOWFIN TUNA NUMBER UN- 6801 VOLUME FREQUENCY OF OCCURRENCE Fishes Crustaceans I Cephalopods I Other mollusks | Miscellaneous | Fishes Crustaceans Cephalopods Other mollusks Miscellaneous PERCENT UN-6802 50 PERCENT 100 0 50 PERCENT Figure 3. — Percentage of total food (by five categories) in stomachs of skipjack and yellowfin tunas captured during cruises UN6801 and UN6802 off the west coast of Africa. Food items are represented in terms of numbers, volumes, and frequency of occurrence. For UN6801, fish families Acanthuridae, Ca- rangidae, Dactylopteridae, Gempylidae, Gonos- tomatidae, Lutjanidae, Mullidae, Priacanthidae, Scombridae, and Serranidae ranked high in terms of volume and frequency of occurrence for both species of tunas. Owing to the large numbers of V. nimbaria in the diet of both spe- cies of tunas, the family Gonostomatidae was the most important forage item for both species in terms of volume. In the diet of skipjack tuna, 1091 FISHERY BULLETIN: VOL. 70, NO. 4 important contributors by volume were Gonosto- matidae, 44.7%; Engraulidae, 8.99f ; Mullidae, 7.9%; Gempylidae, 2.7%; Serranidae, 2.5%; Lutjanidae, 2.0%; Scombridae, 1.6% ; Carangi- dae, 1.6%; and Priacanthidae, 1.5%. Important contributors by volume to the diet of yellowfin tuna were Gonostomatidae, 22.1%; Mullidae, 14.8%; Tetragonuridae, 6.3%; Carangidae, 3.8%; Paralepididae, 1.5%; Priacanthidae, 1.2%; and Scombridae, 1.0%. The remaining fish families contributed less than 1% per fam- ily for both species of tunas. The high volumet- ric contribution by the family Tetragonuridae was due to the large size of only four Tetrago- nurus cuvieri, which were found in a single stomach of a yellowfin tuna. During UN6802 most important fish families by volume and by frequency of occurrence were Carangidae, Gempylidae, Paralepididae, Scom- bridae, and Trichiuridae. Serranidae and Scor- paenidae were prominent in the diet of skip- jack tuna, but entirely absent in the diet of yellowfin tuna. Important contributors by vol- ume in the diet of skipjack tuna were Paralep- ididae, 28.4%; Percoidei, 8.6%; Carangidae, 3.1%; Serranidae, 1.5%; Trichiuridae, 1.4%; Gempylidae, 1.3%; and Scombridae, 1.2%. In the diet of yellowfin tuna important contributors by volume were Exocoetidae, 9.6% ; Alepisauri- dae, 5.6%; Carangidae, 2.7%; TrachyiDteridae, 2.6%; Scombridae, 2.5%; and Percoidei, 1.4%. The remaining fish families and suborders in the diet of both species of tunas contributed less than 1 % per taxon in terms of volume. The rel- atively high contribution by the families Exo- coetidae and Alepisauridae was due to the large volumes of only three forage fish (Appendix Table 4). From our data we see that some of the prominent forage fish families for both spe- cies of tunas were common to both cruises and that others were important during only one cruise (Appendix Tables 1-4). CRUSTACEANS As shown in previous publications (Drago- vich, 1969, 1970), crustaceans, because of their high numbers and high frequency of occurrence, were important components of tuna food. Crus- taceans found in tuna stomachs during both cruises were similar. The majority were larval stomatopods, hyperiid amphipods, and different types of megalopae or their equivalents. The highest number (32) of taxa was noted in the diet of skipjack tuna during UN6801, while in the diet of yellowfin tuna for the same cruise, 20 different taxa were identified — 16 of these were common in the diet af both species of tunas. During UN6802, 22 different taxa were identi- fied in the diet of skipjack tuna and 10 in the diet of yellowfin tuna — 7 were common to both species of tuna. Stomatopods were not iden- tified further than order. Phronima sedentaria, Phrosina semUimata, and Brachyscelhis spp. were the most common amphipods in both tunas for both cruises. Megalopal stages probably con- sisted of many species, but due to the lack of taxonomic literature, they were not identified further than class or family. A variety of anomurans and caridean shrimp were consumed by both species of tunas. Dar- danus pectinatus (Glaucothoe) was the most im- portant anomuran for both tunas during both cruises. Carideans were more prominent during UN6801 than during UN6802. Euphausia hanseni was eaten by both tunas during UN6801. During UN6802, E. hanseni occurred in high numbers in the diet of skipjack tuna but was entirely absent in the food of yel- lowfin tuna. Since E. hanseni are of minute size, they were probably accidentally ingested or the skipjack tuna were filter feeding. The same explanation may be applied to other or- ganisms of similar size found in the stomachs of both species of tunas, for example, copepods and isopods. Another explanation is that some of the euphausiids, copepods, or isopods could be the remains of stomach contents of other fishes ingested by tunas. Phyllosoma occurred in low numbers in the diet of both species during both cruises. The identified forms were Panulirus rissoni, Scyl- larus arctus, Scyllariis sp., and Scyllaridea sp. MOLLUSKS Cephalopods formed the bulk of the molluscan food of both species of tunas during both cruises. 1092 DRAGOVICH and POTTHOFF: FOOD OF SKIPJACK AND YELLOWFIN TUNAS Teuthoidea (squid) were the most important by volume and by frequency of occurrence in the diet of both species. Most of the squid belong to the family Ommastrephidae. Among iden- tified omasterphids, Ornithoteuthis antillarum was most frequently encountered. This species was especially numerous in the food of skipjack tuna during UN6802. Octopoda were less nu- merous and occurred with less frequency than Teuthoidea. The displacement volume of some of the Octopoda (Argonauta argo and A. sp.) was very large. Five specimens of A. argo con- sumed by yellowfin tuna during UN6802 dis- placed 165.5 ml — more than all other mollusks combined for that cruise or all the fishes for that cruise (Appendix Table 4). Among other mollusks, pteropods and hetero- pods were found in the stomachs of skipjack tuna during both cruises. They were absent in the food of yellowfin tuna during UN6802 and occurred only in two stomachs during UN6801. A heteropod, CavoUnia longirostris, occurred in high numbers in the diet of skipjack tuna during UN6801. In terms of volume, both of these mollusks were of minor importance. JUVENILE TUNAS AS FOOD OF SKIPJACK AND YELLOWFIN TUNAS Knowledge on the distribution and abundance of juvenile tunas and tunalike fishes is very lim- ited because existing collection methods for juveniles are inadequate. This information is very important, however, as an aid in identifying spawning seasons and areas of tunas. One of the major sources of juvenile tunas is from stomachs of adult tunas. Juvenile tunas and tunalike fishes were present in the diet of both species of tunas sampled on both cruises. As many as 20 juvenile tunas were found in a single tuna stomach. The most frequently en- countered and the most numerous juvenile tunas were Auxis spp. and little tunny (Euthynnus alletterahis) (Table 2). Specimens of Auxis spp. were found in both species of tunas during both cruises. Specimens of E. alletteratus were present in the diet of both species of tunas, but only during UN6801. All the remaining species of juvenile tunas occurred infrequently in small numbers. Katsuivonus pelamis and Thunnus Table 2. — Occurrence of juvenile scombrids in the stomachs of skipjack and yellowfin tunas during cruises UN6801 and UN6802. Totail number Standard length (mm) Number of juveniles in 1 single stomach Frequency of occurrence Di: splacement volumes Range Mean a Number Percent ml Percent Skipjack tuna UN6801 Unidentified Scombridae 4 _^ __ 1,2 3 0.8 0.5 <0.1 Auxis spp. 53 12-37 29 1,2,3,4, 9 29 8.1 9.5 0.3 Euthynnus alletteratus ISO 10-68 29 1,2,3,4, 5, 10 66 18.5 26.1 1.0 Katsuwonus pelamis 2 20-32 26 1 2 0.6 0.3 0.3 Thunnus spp. 2 34-47 41 Skipjack tuna 1 UN6802 2 0.6 1.4 <0.1 Auxis spp. 33 15-63 31 1,6, 16 10 SjO 8.6 0.9 Sarda sarda 4 25-43 34 1 4 2.0 2.3 0.2 Scomber japonicus 1 — - 40 Yellowfin tuna 1 UN6801 1 0.5 ■1.3 0.1 Auxis spp. 7 12-50 22 1,2 5 6.0 1.4 <0.1 Euthynnus alletteratus 58 11-70 33 1,2,3,4, 5,6,8 21 25.3 le.o 0.7 Katsuwonus pAamis 1 38 38 1 1 1.2 0.2 "> ''"•'"-^y^^- ' f i' _c C 13 ^ CO w "3 dJ >> ^3 >< C m c3 CQ '•—i o _a S '2 to V o TT ^ -> (^ a. tf 111 ^ I >• <9 0. a. ui u 0> o z o a. CO I- < > o X -I a. < I- < 1 >. o ^ ? i U (L < t^ CO « I- is 0. < Ol CO S o o « 5 e I < V CO X t < iij •*■ ^ u CO X o o K z m X < >- CO 2 Id -I 2. 5 < -I CO I o bJO 0.0. O U t- < O. 4 Z uj O. CO Zoo * 5 S X < i CO S >L< O 2 z Ui CO o a: o. > u CO ■ o )0 u CO * K <0 K K Z Z UJ < < > s -i Z UJ Q. — o: 111 CO CO ■ o kJ O o. a. o o r o I o 1 o ! o 1 o 1 o 1 o 1 o I o 1 o o at « N <0 lO ♦ K) CM (*»^) SiN31N00 HOVI^OIS M ^ A a u bo -t-> -S u o C4 CO 0) c xi (V M > 03 o XI « § 4) «H N -U o ii 8 a o in o «> > lO b ^ CD E C ♦ lO s "5 ♦ ^' s _ a> 0) «■ a- ^^^ cS — . E O O o E CD Qi TO lO M C CO 23 CO f^ S < .y ^ -J ^ m o 2 " .1 .22 UJ 1^ M §1 o CO « CM o :zfi^^ ':^<^:m ^#^t^^^>?^ v Figure 5. — Bottom feeding by a swimming fish. A swimming fish (A) visually fixes on a shrimp on the bottom, partially brakes its forward motion (B), and tilts toward the sand. As the caudal fin beats downward, the mouth opens, the opercula spread, and the fish moves ahead to ingest the shrimp (C, D). its mouth several times. Then, without lunging or striking at the prey, the fish would swim away after a few seconds. We found feeding intention movements to oc- cur after the fish had been feeding. For in- stance, in one case, after we had introduced 1,100 g of shrimp into the tank (29 days after the last feeding of 4,400 g of shrimp), the fish began to feed within 3 min. One fish, after eat- ing 13 shrimp within a 22-min period, exhibited 1133 FISHERY BULLETIN: VOL. 70, NO. 4 an intention movement 3 min later. Then four more shrimp were ingested during the next 20 min, followed 2 min later by two additional in- tention movements. A second fish ingested 15 shrimp within 35 min and 2 min later made an intention movement. A third fish, after ingest- ing 16 shrimp in 26 min, made an intention move- ment 1 min later, then fed immediately on one shrimp, and 2 min later exhibited an intention movement. These movements may have been related to a reduction of feeding motivation as a result of satiation. FRIGHT RESPONSE We observed what was apparently a "fright" response to a sudden stimulus under two differ- ent circumstances. In one case, there was a mal- function of the dimmer lights which caused the sudden onset of the daylight lights. A swim- ming fish immediately dropped to the bottom where it remained resting on the sand surface. In another case, an observer above the tank waved his arms as a fish moved about near the water surface. The fish immediately dropped to the sand, darkened, and assumed a rigid posture. The head and caudal fin lay flat, but the dorsal and anal fins were arched in two places along their length. The fish remained in this posture for about 45 sec during which it slowly lowered first the anterior, then the posterior sections of the dorsal and anal fins until they were flat. The flounder then swam about 2 m away and buried. In both instances, the initial response of the fish to a fright stimulus was to drop to the bottom and remain motionless. Burying occurred as a secondary response. DISCUSSION Previous descriptions of the habits of sum- mer flounder have characterized them as primar- ily bottom-oriented, except for occasional sorties to the surface in pursuit of prey (Bigelow and Schroeder, 1953:267-270). Furthermore, this species has been described as being relatively immobile, except while feeding or during the normal migratory period (Ginsberg, 1952). In our laboratory observations, while the fish fre- quently searched for and captured prey on the sand and also remained quiescent on the bottom for long periods, they would also frequently use the water column for swimming, prey search, and feeding. In fact, during one part of our study, the fish swam and glided for extended periods throughout all levels of the tank, seldom resting in any one position. The gliding behavior we observed could play an important role for the animal in the sea. After reaching the surface, the fish could travel considerable distances with little or no active swimming movements, using natural negative buoyancy and body shape to full advantage. Po- sitioning of the fins and body would control for- ward speed and distance traveled. Although there might have been a sacrifice in speed, the gliding would represent a saving in energy as compared with that required to swim the same distance. Gliding would also enable the animal to search for and capture prey in the water col- umn more efficiently, since it could approach a prey with less gross movement than would occur during active swimming. This might lessen the chance of eliciting escape responses from the prey due to visual or mechanical stimuli. An- other adaptive advantage of gliding in food search might be related to the fact that the head was steady, thus making it easier to keep the prey in the visual field. Although the summer flounder were primarily day-active, we observed burying, feeding, shambling, and swimming both day and night. Similar to the summer flounder, turbot (Scoph- thalmus maximus) swim and are active on the bottom primarily during the day, although both activities may occur at night to a lesser degree (de Groot, 1971). Verheijen and de Groot (1967) and de Groot (1971) established that plaice (Pleuronectes platessa) and flounder (P. flesus) showed a nocturnal pattern of swimming in the upper water column, while during the day they would shamble or swim over the bottom searching for food. Kruuk (1963) and de Groot (1971) found that the sole (Solea vulgaris) also had a nocturnal period of high activity under both natural and artificial light. The method of burying in the summer flounder is similar to Kruuk's (1963) description of "dig- ging-in" in the sole. In the sole as well as the 1134 OLLA, SAMET, and STUDHOLME: SUMMER FLOUNDER winter flounder, Pseudopleuronectes american- us, (McCracken, 1963) and starry flounder, Platichthys stellaUis, (Orcutt, 1950), burying could be induced as a direct response to a sudden disturbance, such as a change in light intensity or moving object. In the summer flounder, ap- parently the primary response to a fright stim- ulus is to assume a stationary and sometimes rigid posture on the bottom. This is followed by burying as essentially a secondary response. There are numerous descriptions of the sen- sory mechanisms utilized by diff"erent groups of flatfishes during feeding (see de Groot, 1971, for review). Since summer flounder are primarily day-active, it was not surprising that vision played a primary role in prey selection and cap- ture. According to de Groot (1971), this is ap- parently characteristic of Bothidae including brill (Scophthalmiis rhombiis) and turbot, which he designated as visual day-feeders, largely de- pendent on visual stimuli for locating prey. De- spite the fact that the summer flounder would also be categorized in this manner, we did ob- serve feeding at night. Although the light level of 2.5 X 10~^ mc (as measured at 1 m below the surface) fell slightly below the lO^-lO"- mc level cited by Blaxter (1970) as the range in which most visual feeders cease active feeding, it was possible that vision was still being utilized. The summer flounder, winter flounder, (011a, Wicklund, and Wilk, 1969), and lemon sole, Microstomus kitt, (Steven, 1930) may rest on the bottom with head up while actively moving their eyes. In the latter two species, the fish may be searching for food and will lunge for- ward from this position to strike at and capture prey. While we considered summer flounder in this position to be alert and responsive, it was also apparent that this was not necessarily in- dicative of a prefeeding strike. Although Gins- berg (1952) stated that summer flounder lie in wait for passing prey, we found that the fish, even in this alert "head-up" posture, never lunged from a resting position at a prey, even though it was only a few centimeters away, but always preceded prey capture by active search- ing. While we do not understand the role yawning plays in the behavior of the summer flounder, we did find evidence that it was associated with changes in activity. Rasa (1971) found that yawning in the yellowtail demoiselle was asso-* ciated with an increased excitement level. She postulated that the strong muscle contraction that occurs during yawning could serve to in- crease the blood flow and oxygen to the body musculature and thereby facilitate the onset of the animal's activity. It may also be conceiv- able that yawning movements may act to flush sand or debris from the gill areas, one function suggested for yawning in Pacific bonito, Sarda chllieiisis, (Magnuson and Prescott, 1966). ACKNOWLEDGMENTS We wish to express our grateful appreciation to A. J. Mansueti for her illustrations and A. D. Martin for his help throughout all phases of the study. LITERATURE CITED BiGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53:1-577. Blaxter, J. H. S. 1970. Light: Fishes, hi O. Kinne (editor), Ma- rine ecology: A comprehensive, integrated trea- tise on life in oceans and coastal waters. Vol. I, Part 1, p. 213-285. Wiley-Interscience, Lond. Ginsberg, I. 1952. Flounders of the genus Paralichthys and re- lated genera in American waters. U.S. Fish Wildl. Serv., Fish. Bull. 52:267-351. Groot, S. J. de 1964. Diurnal activity and feeding habits of plaice. Cons. Perm. Int. Explor. Mer, Rapp. P.-V. Reun. 155:48-51. 1969. Digestive system and sensorial factors in relation to the feeding behaviour of flatfish (Pleuronectiformes). J. Cons. 32:385-394. 1971. On the interrelationships between morphol- ogy of the alimentary tract, food and feeding be- haviour in flatfishes ( Pisces : Pleuronectiformes) . Neth. J. Sea Res. 5:121-196. Kruuk, H. 1963. Diurnal periodicity in the activity of the com- mon sole, Solea vulgaris Quensel. Neth. J. Sea Res. 2:1-28. Magnuson, J. J., and J. H. Prescott. 1966. Courtship, locomotion, feeding, and miscel- laneous behaviour of Pacific bonito (Sarda ckil- iejisis). Anim. Behav. 14:54-67. 1135 FISHERY BULLETIN: VOL. 70. NO. 4 McCracken, F. D. 1963. Seasonal movements of the winter flounder, Pseudopleuronectes americanus (Walbaum), on the Atlantic coast. J. Fish. Res. Board Can. 20: 551-586. Olla, B. L., W. W. Marchioni, and H. M. Katz. 1967. A large experimental aquarium system for marine pelagic fishes. Trans. Am. Fish. Soc. 96: 143-150. Olla, B. L., R. Wicklund, and S. Wilk. 1969. Behavior of winter flounder in a natural habitat. Trans. Am. Fish. Soc. 98:717-720. Orcutt, H. G. 1950. The life history of the starry flounder, Plat- ichthys stellatus (Pallas) . Calif. Div. Fish Game, Fish Bull. 78, 64 p. Poole, J. C. 1964. Feeding habits of the summer flounder in Great South Bay. N.Y. Fish Game J. 11:28-34. Rasa, O. A. E. 1971. The causal factors and function of 'yawning' in Microspathodon chrysurus (Pisces: Pomacen- tridae). Behaviour 39:39-57. Smith, R. W. 1969. An analysis of the summer flounder, Paral- ichthys dentatus (Linnaeus) population in the Delaware Bay. Masters Thesis, Univ. Delaware, Newark, 72 p. Steven, G. A. 1930. Bottom fauna and the food of fishes. J. Mar. Biol. Assoc. U.K., New Ser. 16:677-[707]. Verheijen, F. J., and S. J. de Groot. 1967. Diurnal activity pattern of plaice and floun- der (Pleuronectidae) in aquaria. Neth. J. Sea Res. 3:383-390. 1136 THE EFFECTS OF TEMPERATURE AND PHOTOPERIOD ON REPRODUCTIVE CYCLING IN THE ESTUARINE GOBIID FISH, GILLICHTHYS MIRABILIS Victor L. De Vlaming^ ABSTRACT Investigations were undertaken at several different times during the year to examine the effects of various photoperiods and constant-temperature regimes on reproductive function in the longjaw goby, Gillichthys mirabilis, with the intent of evaluating the influence of these factors in regulation of the annual sexual cycle. Testicular regression occurs at any time during the year when fish are exposed to constant temperature of 24°C and above, independent of photoperiod. Similar results were obtained with female fish, but 22 °C is the thermal threshold. It is concluded that the gonadal regression ob- served in the Alviso population of this species during the summer months is a conse- quence of increasing temperature. At high temperatures, the transformation of sper- matogonia to spermatocytes is blocked, and in females vitellogenesis is inhibited. The degree of gonadal regression is temperature-dependent. Gonadal recrudescence is de- pendent on low temperatures (10°-20°C) and will not occur if fish are exposed to high temperatures (24°C or above) regardless of photoperiod. At low temperatures, short photoperiods accelerate recrudescence. Between January and June spermatogenesis and oogenesis are maintained at temperatures between 10° and 21°C; long photoperiods are more effective in this respect, but not essential. Termination of the reproductive season in this species is not endogenously timed. Regression is not "obligatory" since gonadal involution does not occur at the "normal" time if fish are exposed to temper- atures of 20 °C or below. The survival of any species in a seasonally chang- ing environment is dependent on the develop- ment of mechanisms that permit it to adjust physiological functions to changes in the envi- ronment. Studies of reproductive timing and how the environment influences this timing are of importance in understanding the ecology of any species. Compared with the wealth of information available on the systematics, ethology, and physiology of fishes, there is little knowledge concerning how external factors regulate their reproduction. Some investigations have been undertaken to elucidate the role of environmental factors in regulating the reproductive cycles of various teleosts. The relationship of environ- * Department of Zoology, University of California, Berkeley, Calif.; present address: Department of Bi- ology, Wehr Life Sciences Building, Marquette Univer- sity. Milwaukee, WI 53233. Manuscript accepted May 1972. FISHERY BULLETIN: VOL. 70, NO. 4, 1972. mental factors to the reproductive cycles of gobies has not received experimental consider- ation. Moreover, experimental work with the environmental control of teleost reproductive cycles has been confined to fewer than 20 species representing only 8 families. Photoperiod and temperature are presumed to be the most important factors (i.e., the most studied) influencing the neuroendocrine centers that control gonadotropin secretion in teleosts (de Vlaming, 1972a). The experimental condi- tions employed in a majority of the previous studies, however, are diverse and the results contradictory. In fact, most of the experimental work was too poorly controlled and too brief in duration to allow proper assessment of the role of the environment in synchronizing fish reproduction (de Vlaming, 1972a). The subject of the present study is the long- jaw goby, Gillichthys mirabilis. It is distributed 1137 FISHERY BULLETIN: VOL. 70, NO. 4 from central California south to Magdalena Bay, Baja California, and the Gulf of California south to Mulege on the west coast, and south to Agia- bampo Bay on the east coast (Barlow, 1963). The typical habitat of this species is the inter- tidal region of coastal sloughs. Barlow (1961, 1963) discussed the systematics and some as- pects of the ecology of G. mirahilis. The popu- lation of Gillichthys used in these studies occurs in the Alviso salt ponds located at the southern end of San Francisco Bay, Calif. Carpelan (1957) described seasonal changes in the hy- drobiology of these ponds. De Vlaming (1972b) described the reproduc- tive cycle of G. mirahilis and suggested that sea- sonal temperature changes may be involved in regulating sexual cycling in this species. The spawning period is protracted, extending from December to June. Gonadal regression occurs in July; the gonads remain regressed during August and September. Gonadal recrudescence begins in late September, reaching completion by early December. The aim of the present study was to determine the effects of various light and constant temper- ature regimes on gonadal function in G. miror bills, with the hope of evaluating the influence of these factors in regulation of the annual sex- ual cycle. The phenological data on reproductive cycling in this species presented by de Vlaming (1972b) was used as a basis for these studies. Some of the previous studies with teleosts have shown that the effect of the environmental synchronizer (s) varies with the stage of gonadal maturity. Consequently, the effects of photo- period and constant temperature treatments were examined during different phases of the gonadal cycle. MATERIALS AND METHODS Samples of G. mirabilis were captured in the Alviso habitat at several different times during the year and thus in different phases of gameto- genesis. Since males were more abundant in these samples, a greater number of experiments were conducted with this sex. Several fish from each sample were sacrificed, and the gonads ex- amined at the time of capture; these fish served as a reference for the experiments that followed. In the following discussion the fish sacrificed from the samples from nature will be referred to as initial controls. In many of the experiments, samples of fish from the natural habitat were collected and sacrificed upon termination of the experiment; these fish will be referred to as terminal controls. To facilitate quantification of gonadal response, animals of approximately equal size were utilized in these experiments. Experimental fish were maintained in 56- (no more than 10 fish per tank) or 132-liter (no more than 18 fish per tank) tanks. Recirculating filtered seawater was used in all of these exper- iments. The bottom of each tank was covered with fine gravel. The experimental tanks were housed in constant temperature rooms (± 1.5°C). Temperatures selected for these experiments are within the range normally ex- perienced by this species during the year. Various photoperiods were also employed in these experiments. Light was provided by 20-w warm-white fluorescent bulbs suspended above the tanks. Salinity was maintained at a con- stant 35 %c, and pH between 8.0 and 9.5 (these pH's are consistent with those experienced by the fish in the Alviso ponds). The fish in these experiments were provided with a varied diet consisting of brine shrimp, chopped fish, boiled egg white, and beef kidney and liver; all fish ate voraciously. Upon termination of each experiment, the weight and standard length of each fish were re- corded. Gonads were weighed and prepared for histological examination in the same manner as previously reported (de Vlaming, 1972b) . Grav- imetric data are expressed in absolute weights since it was shown (de Vlaming, 1972b) that gonadal weight is independent of body weight (and length) in the size range used. Spermato- genesis and oogenesis were divided into six and five recognizable phases (Tables 1 and 2) to fa- cilitate quantitative evaluation of gametogenetic activity. Statistical comparisons of gonadal weights be- tween experimental groups were made by using the Mann- Whitney U test (Siegel, 1956, p. 184- 193). This nonparametric test is suitable for small sample sizes and can be used to determine 1138 de VLAMING: CONTROL OF REPRODUCTION IN CILLICHTHYS Table 1. — Criteria used for histologically staging testes of Gillichthys mirabilis. Stage Histological characteristics of testes "Regressing testis." Seminiferous lobules characterized by large numbers of pyknotic nests of degenerating cells (sper- matozoa, spermatids, and spermatocytes); phagocytes observed free within the lobu'les. "Quiescent testis." Seminiferous lobules smaH in diameter. Germinal epithelium corfslsts of ortly spermatogonia. Lumen of the lobules contain only fev^ residual spermatozoa, and the sperm duct Is coHopsed. "Mitotic phase." Same as Stage I, wirh the exception that mitotic figures are observed in the spermatogonia. "Meiotic phase or active spermatogenesis." Testicular lobules larger than In Stages 1 and 2; germinal epithelium consists of spermatogonia, spermatocytes, and spermatids. "Prespawning testis." Seminiferous lobules large and dis- tended vi/ith sperm. Germlnol epithelium consists of relatively few spermatogonia. "Postspawning testis." Seminiferous 'lobules small and con- tain relatively few sperm; sperm duct expanded and containing re'sidual sperm. Table 2. — Criteria used for histologically staging ovaries of Gillichthys mirabilis. Stage Histologica'l characteristics of ovaries I "Regressing ovary." Atretic follicles predominate in fhe ovary. Only nonytolky oocytes and oogonia present. II "Quiescent phase or phase of oogonial proliferation." Ovary characterized by nonyolky oocytes with a basophilic cytoplasm, ond a diameter of less tfian 75 /i. Granulosa not fully or- ganized around the developing oocytes. III "Phase of active vltellogenesis." Ovary characterized by developing yolky oocytes whose diameter is between 75 /i and 640 m- Granulosa fully organized around the oocytes. IV "Prespawning condition." Ovary characterized by oocytes whose diameter Is in excess of 640 fi. Yalk vesicles abundant. V "Postspawning condition." The ovary is wine-red In color; the tunica albuginea thick, highly vascularized, and folded. Postovulatory follicles predominate in the ovary. The stroma of fhe ovary opipears disorganized, yet highlly vascularized. whether two independent groups have been drawn from the same population. RESULTS EFFECTS OF CONSTANT TEMPERATURE AND PHOTOPERIOD ON FISH WITH REGRESSING GONADS To examine the influence of low and high temperature treatments at different photoperi- ods on gonadal recrudescence, fish with regres- sing testes (Stage 0) and ovaries (Stage I) were collected in July 1967, Fish were exposed to 13°C, at a short (8L/T6D) or long (15L/9D) photoperiod, or to 27 °C at a short photoperiod (8L/16D). The effects of these treatments on testicular and ovarian weights are summarized in Figure 1. The testes (Stage 0) and ovaries (Stage I) of all of the initial controls were regressing. After 57 days, testicular and ovarian weights increased significantly (P < 0.01) at both photoperiods at 13°C, and were also significantly greater (P < 0.01) than those of fish from nature sacri- ficed at the same time. Ovarian weights of fe- males exposed to 8L/16D at 13°C were signifi- cantly greater (P < 0.01) than those of fish exposed to 15L/9D at the same temperature. Testes of all fish at 13°C were in the meiotic phase of spermatogenesis (Stage 3) whereas those of all fish in the September sample from nature were only in Stage 2 (mitotic prolifer- ation phase). The ovaries of all fish at 13°C were in phases of vltellogenesis (Stage III); the oocytes of fish at BL/IBD were, however, in a ?60 1> 50 r-,0 f 30 2 20 3 y ^ lO- 0 r^ i^ <¥^ \h Initial Control (IS July) 8L 8L I5L 27° 13° 9 S«plember 8L 27° 27 Now 2000- 1800- 1600 i 1200 * 1000 c 1 800 > o 600 400 200 0 m ji= Initial Control Ui July) rfl 8L BL I5L 27° 13° 9 Seplennber Jl 8L 27° 27 Nov Figure 1.— Effect of 13°C at a long (15L/9D) photoperiod and 13° and 27°C at a short (8L/ 16D) photoperiod on testicular and ovarian weight in Gillichthys mirabilis. Mean gonadal weight is illustrated by histograms; the mean is bracketed by one standard error. Shaded histograms represent gonadal weight of samples collected in nature; open histograms, experi- mental groups. Light (hours per 24 hr), tem- perature, and dates on which fish were sacrificed are recorded below the histograms. Sample sizes are indicated atop each histogram. 1139 FISHERY BULLETIN: VOL. 70, NO. 4 later stage of vitellogenesis. The ovaries of all fish in the September nature sample were in the quiescent phase (Stage II), After 57 days at 27°C, testicular weights were significantly lower (P < 0.05) than those of the initial July controls. The testes of all fish were in the quiescent phase (Stage 1), even after 134 days. In contrast, the testes of all fish in the November sample from nature were in Stage 3. Ovarian weights of fish exposed to 27°C (both the September and November samples) did not differ significantly from those of the initial July controls, but they were significantly lower (P < 0.01) than ovarian weights of both 13°C groups. The ovaries of the 27°C treated fish revealed only resting oocytes (Stage II). How- ever, the ovaries of fish in the November sample from nature were in Stage III, IV, or V. Thus, low temperatures promote gonadal re- crudescence in Gillichthys, independent of photo- period. Short photoperiods may accelerate the rate of gonadal recrudescence at low tempera- tures. A short photoperiod, however, failed to initiate gonadal recrudescence at high temper- ature. High temperatures act by blocking vitel- logenesis and the transformation of spermatogo- nia to spermatocytes. The influence of constant temperature and photoperiod on testicular recrudence was exam- ined again in July 1968, using 13° or 20°C, with a short (8L/16D) and a long (15L/9D) photo- period. A fifth group was exposed to 24°C at a long photoperiod (15L/9D). Each group was sampled after 45, 70, and 120 days (Figures 2 and 3). In the initial July controls testes were regres- sing (Stage 0). At 24°C testicular weights re- mained low throughout the experiment; in Sep- tember and November testicular weights at this temperature were significantly lower (P < 0.05) than those of the initial July controls. More- over, at 24 °C testes remained in the regression or quiescent phase (Stage 0 or 1) throughout the experiment (Figure 3). At 20°C testicular weights remained essen- tially the same as in the initial July controls 90 80 70 '60 .50 40- •5 30- 20 10- Iniliol Controls (15 July) i I 4a i i: 1 i 1 ■ ■ A (^ ffi I5L 8L I5L 8L I5L 24° 20° 13° 28 August 15L 8L I5L SL I5L 24° 20° 13° V 22 September th * I5L 8L I5L 8L I5L 24° 20° 13° , I November Figure 2.— EflFect of 24°, 20°, and 13°C treatments at short (8L/16D) and long (15L/9D) photoperiods on testicular weight in Gillichthys mirabilis. Mean testicular weight is represented by histograms; the mean is bracketed by one standard error. Shaded histograms represent testicular weight of samples collected in nature; open histograms, experimental groups. Light (hours given per 24 hr), temperature, and dates on which fish were sacrificed are recorded below the histograms. For sample sizes, see Figure 3. 1140 de VLAMING: CONTROL OF REPRODUCTION IN GILLICHTHYS Figure 3.— Effect of 24°, 20°, and 13°C treatments at short (8L/16D) and long (15L/9D) photoperiods on testicular his- tology in Gillichthys mirabilis. I.C. refers to initial controls (15 July) and N, to samples from natural population. Each box represents the testicular condition of one fish. 1=1 _H_ _H_ P P .M. H R I5L BL 15L 9L l5L 24° 20° 13° 28 Auqust I5L 8L I5L BL I5L 24° 20° 13° 22 SepiemDef ISL 6L I5L BL I5L 24° 20° 13° with one exception. Testicular weights of the 20 °C group at 8L/16D sacrificed in August were significantly greater (P < 0.05) than those of the July controls. With the exception of this same group, active spermatogenesis was not in- itiated in fish exposed to 20°C. Mitotic prolifer- ation of spermatogonia was, however, stimulated by this treatment. In contrast, active spermato- genesis was initiated by August at 13°C, regard- less of photoperiod. Active spermatogenesis was not initiated in the natural population until after 22 September. Testicular weights of both groups at 13°C sacrificed in September were significantly greater (P < 0.01) than those of the initial July controls and those of the September sample from nature. Some photoperiod effect was evident at 13°C since testicular weights at 8L/16D were significantly higher (P < 0.05) than those of the 15L/9D group by September. By November, the testes of a majority of the fish in the 13°C-8L/16D group were in the prespawning condition (Stage 4) whereas the testes of all fish in the 13°C- 15L/9D group were in Stage 3; testicular weights of these two groups were also signifi- cantly diflferent (P < 0.01). These data indicate that 24°C inhibits testic- ular recrudescence by blocking the transforma- tion of spermatogonia to spermatocytes and also retards mitoses in the spermatogonia. Low tem- peratures promote testicular recrudescence; the rate of recrudescence at a low laboratory tem- perature was faster than in the natural popula- tion. At low temperatures, short photoperiods accelerate testicular recrudescence. With the exception of the one sample at 20°C-8L/16D sacrificed in August, 20°C stimulates little or no testicular recrudescence, only mitotic prolifer- ation of spermatogonia. EFFECTS OF CONSTANT TEMPERATURES AND PHOTOPERIOD ON FISH IN STAGES OF ACTIVE GAMETOGENESIS (MAY) Responses of fish in phases of active gameto- genesis (in May) were examined by exposing two groups of fish for 67 days to 13°C, at a short (8L/16D) and a long (15L/9D) photoperiod, and 27°C at a long photoperiod (15L/9D) . The efl["ects of these treatments on gonadal weights are summarized in Figure 4, At the beginning of treatment, testes and ov- aries were in Stages 3 and III, respectively. In the July sample from nature, sacrificed with the experimentals, ovarian (Stage I) and testicular (Stage 0) regression was occurring. Testes and ovaries of fish at 27°C regressed as in nature; in both sexes gonad weights were significantly lower (P < 0.01) than initial levels. In both groups at 13°C spermatogenetic ac- tivity remained at the initial levels and the testes did not show the regression seen in nature or at 27°C. Ovarian weights in the 13°C group at a long photoperiod were significantly lower (P < 0.05) than those of the initial May sample 1141 FISHERY BULLETIN: VOL. 70, NO. 4 Figure 4.— Effect of 27° and 13°C treat- ments at short (8L/16D) and long (15L/9D) photoperiods on testicular and ovarian weight in Gillichthys mirabilis. Mean gonadal weight is represented by histograms; the mean is bracketed by one standard error. Shaded histograms represent gonadal weight of samples collected in nature; open histo- grams, experimental groups. Light (hours given per 24 hr), temperature, and dates on which fish were sacrificed are recorded below the histograms. Sample sizes are in- dicated atop each histogram. 80 70 e 60- 2 50 a> S 40 I 30 £ 20 10 0 * r--i'" * Initiol Control (15 May) I5L I5L 8L 27° 13° 20 July 2000 1800 1600 a* ^ 1400 2 1200 a> * lOOO c ° 800 o > ° 600- 400- 200- 0 rfl Iniliol Control (15 Mayl M 15L I5L 8L 27° 13° 20 July and those of the 13°C group at a short photo- period. Ovaries in both 13°C groups were, how- ever, in active vitellogenesis (Stage III). These results could indicate a photoperiod influence in vitellogenesis. But one must observe caution in interpreting these data since this difference in ovarian condition may simply be a problem of beginning experiments with fish in various stages of oogenesis. Data presented by de Vlam- ing (1972b) revealed the nonsynchrony of ga- metogenesis in this species (i.e., fish in different stages of gonadal development were common in monthly samples between November and June) . High temperatures apparently cause testicular and ovarian regression in spring, at least when the photoperiod is long, whereas low tempera- tures prevent gonadal regression and are re- quired for spermatogenesis and vitellogenesis. EFFECTS OF CONSTANT TEMPERATURES AND PHOTOPERIOD ON FISH IN STAGES OF ACTIVE SPERMATOGENESIS (JANUARY) In the previous experiment the influence of temperature and photoperiod was examined during the spawning season in spring. Whether fish respond similarly in winter (near the onset of the spawning season) is also of interest, so in January 1968 fish were exposed to 27°, 20°, and 13°C, at a short (8L/16D) or a long (15L/9D) photoperiod for 30-39 days. The in- fluence of these treatments on testicular weight and histology is presented in Figures 5 and 6, respectively. Testes of the initial January controls were in Stages 3, 4, or 5. As observed in May, testes regressed rapidly at 27°C; testes weights were significantly less (P < 0.01) than those of the January controls and February sample from nature. Histological examination confirmed that regression had occurred (testes in Stage 1). No photoperiod effect was seen at 27 °C. 70 E 50 o- 40 i U V. 20 I- 10 A k A A A A Intttol Contfots (15 January) v I5L 8L I5L 8L 27° 20° I5L 8L 13° 15 Februory I5L 8L Figure 5.— Effect of 27°, 20°, and 13°C treatments at short (8L/16D) and long (15L/9D) photoperiods on testicular weight in Gillichthys mirabilis. Mean testic- ular weight is represented by histograms; the mean is bracketed by one standard error. Shaded histograms represent testicular weight of samples collected in na- ture; open histograms, experimental groups. Light (hours given per 24 hr), temperature, and dates on which fish were sacrificed are recorded below the histo- grams. For example sizes, see Figure 6, 1142 de VLAMING: CONTROL OF REPRODUCTION IN GILLICHTHYS S 4 a_ _B_ 3_ _s_ _B- H R N' I5L 8L ISL 8L I5L BL 27° 20° 13° 15 Feb'uory 2'' feb Figure 6.— EflFect of 27°, 20°, and 13°C treatments at short (8L/16D) and long (15L/9D) photoperiods on testicular histology in Gillichthys mirabilis. I.C. refers to initial controls (15 January) and N, to sample from natural population. Each box represents the testicular condition of one fish. Testicular weights in the 20°C groups re- mained at approximately the initial level. Both groups at 20°C had significantly heavier (P < 0.01) testes than those of the 27°C groups. The testes of all fish at 20°C advanced to Stages 3 or 4 by 24 February, and there was no clear photoperiod eflfect. Testicular weights of the 13°C groups did not diflfer significantly from the initial January controls but were significantly greater than those of the 20 °C group at a short photoperiod (P < 0.05) and those of the 27°C groups (P < 0.01). Spermatogenic condition of the testes in the 13°C groups was essentially the same as in the January controls and Febru- ary sample from nature (Figure 6). The results of this experiment indicate that spermatogenesis is maintained at 13°C and 20°C (independent of photoperiod) but that this pro- cess may occur at a slower rate at 20°C than at 13°C. This difference in effect of these two tem- peratures must be accepted with some reserva- tion because the variability of gonadal develop- ment in the initial controls could introduce a degree of bias into the results. Regardless of photoperiod, testicular regression occurred at 27°C. EFFECTS OF HIGH TEMPERATURES ON GONADAL FUNCTION IN DIFFERENT SEASONS Experiments using high temperatures were initiated at different times during the year to determine whether there is a seasonal variation in gonadal susceptibility to such treatment. The conditions employed and results obtained in these experiments are summarized in Table 3. Experiments I and II indicate that 24°C is a sufl^ciently high temperature to initiate testic- ular regression within a relatively short period. With either a long (15L/9D) or a short (8L/16D) photoperiod (Experiment III), 25°C stimulates the completion of testicular regres- sion within 21 days; testes of these fish were in the quiescent phase ( Stage 1 ) . A temperature Table 3. — EflFect of various high temperature and photoperiod treatments on testicular weight in Gillichthys mirabilis. Experi- Length of Temper- Photo- Testicu lar weight (mg) (5 ± SE) ment Date treatment (Days) ature (°C) period (L/D) numbers Initial controls After treatment (») 1 May 69 21 24 12/12 79.1 ^ 5.2 58.0 ± 3.1** (7) II Feb. 70 15 24 14/10 79.1 ^ 9.5 34.4 ± 2.5** (9) III Feb. 68 21 25 15/9 43.7 ± 3.4 25.0 ± 1.6* (7) Feb. 68 21 25 8/16 43.7 ^ 3.4 23.2 ± 4.8* (7) IV July 70 30 27 10/14 22.3 ^ 7.8 18.1 ± 1.4 (8) V Nov. 69 14 27 12/12 64.8 -t- 8.8 30.0 ± 2.7** (8) VI Apr. 70 8 27 12/12 97.8 ± 7.4 A6.7 ± 3.9** (10) VII July 68 41 24 12/12 33.1 ± 3.2 28.9 ± 2.2 (7) July 68 41 28 12/12 33.1 ^ 3.2 18.1 ±1 2.2* (8) VIII June 68 21 27 12/12 52.2 -*- 3.9 26.2 ± 2.7** (7) Juna 68 21 30 12/12 52.2 -f- 3.9 14.8 ± 3.1** (8) IX Apr. 68 8 26 12/12 62.2 i 5.4 40.1 ± 4.2** (7) Apr. 6a 14 28 12/12 62.2 ■+■ 5.4 24.3 ± 1.8** (9) Apr. 68 8 32 12/12 62.2 ± 5.4 15.6 ± 2.8** (8) •Significantly less {P < 0.05) than initial controls. "•Significantly less {P < 0.01) than initial controls. 1143 FISHERY BULLETIN: VOL. 70, NO. 4 of 25 °C does not, however, block mitotic prolif- eration of spermatogonia. Experiment IV (Table 3) indicates that 27°C blocks the initiation of testicular recrudescence. Testicular regression is initiated within 14 days at 27°C (Experiment V); the testes of all fish were in the regression phase (Stage 0). More- over, Experiment VI suggests that testicular re- gression is initiated within 8 days at 27°C; the testes of all fish were regressing (Stage 0) . When initial controls are undergoing testicu- lar regression, 28°C stimulates a more rapid completion of regression than does 24°C (Ex- periment VII). This is evident since the testes of a majority of the fish at 28°C were in the quiescent phase (Stage 1) whereas those of the 24 °C group were in Stage 0. A 30°C tempera- ture (Experiment VIII) causes a more complete testicular regression than does a 24°C tempera- ture (i.e., testicular weights of the 30°C group were significantly lower (P < 0.01) than those of the 27 °C group). Both of these treatments caused the completion of testicular regression within 21 days; the testes of a majority of fish in both groups were in the quiescent phase (Stage 1). Similarly, 32°C stimulates a more rapid and more complete testicular regression than 28°C (Experiment IX) ; testicular weights after 8 days were significantly lower (P < 0.01) in the 32°C group than those of the 28°C group. Testes of all fish at both temperatures were re- gressing (Stage 0). After 14 days at 28°C, testicular weights were still significantly higher (P < 0.05) than testicular weights after 8 days at 32°C (Experiment IX). With the exception of the experiments in which the initial controls were fish with regres- sing gonads, all high temperature treatments summarized in Table 3 caused a significant de- crease in testicular weights. These data indicate that the rate of testicular regression and the degree of testicular regression are temperature dependent. These data also indicate that under laboratory conditions testicular involution is in- itiated relatively soon after exposure to high temperatures and that regression can be com- pleted rapidly in this species. Experiments similar to those discussed above were conducted with female Gillichthys (Table 4). Beginning with fish in active vitellogenesis (Stage III), 25°C initiates ovarian regression (Stage I) within 21 days, independent of photo- period (Experiment 1). Ovarian weights in the two photoperiod groups were not significantly different. Beginning with fish having regres- sing ovaries (Experiment 2), 27°C caused the completion of ovarian risgression. Experiment 3 suggests that ovarian regression is complete within 21 days at 30°C; the ovaries of a majority of the initial controls were in phases of active vitellogenesis (Stage III), whereas following treatment, the ovaries of all fish were in the quiescent phase (Stage II). Experiment 4 (Table 4) indicates that 32°C stimulates a more rapid rate of ovarian regres- sion than does 28°C; ovarian weights of the 32° and 28°C groups were significantly difl^erent (P < 0.01) after 8 days of treatment. Ovarian regression was occurring in all fish. After 14 days at 28°C, further ovarian regression oc- curred, but involution had still not been com- pleted. Therefore, temperatures between 25° and 32°C initiate ovarian regression within a rel- atively short time; all high temperature treat- ments caused a significant decrease in ovarian weights. Apparently the completion of ovarian regression is temperature dependent, EFFECTS OF 21°-22°C AND PHOTOPERIOD TREATMENT ON GONADAL FUNCTION The previous experiments suggest that gonad- al regression occurs at 24°C but not at 20°C. To define more precisely the thermal threshold for gonadal involution, the efl^ects of 21°-22°C treat- ments were examined in April, May, September, and November. The conditions employed and results obtained in these experiments are sum- marized in Table 5. Beginning in April with fish in active sperm- atogenesis or in the prespawning condition (Stage 3 or 4), there was no change in testes after 17-day treatment at 22°C (15L/9D) . How- ever, after 30-day treatment at 22°C and a short photoperiod (10L/14D), there was a significant decrease (P < 0.01) in both ovarian and testic- ular weights (Experiment 11 — Table 5) . Testes 1144 de VLAMING: CONTROL OF REPRODUCTION IN GILLICHTHYS Table 4. — Effect of various high temperature and photoperiod treatments on ovarian weight in Gillichthys mirabilis. Experi- menf number Length of treatment (Days) Temper- ature (°C) Photo- period Weight of ovaries (mg) (X ± SE) Initial controls ♦Significantly less [P < 0.05) than initiail controls. "♦Significantly less {P < 0.01) than initial controls. After treatment (n) 1 21 25 15L/9D 1,744 ± 422 381 ± 81^* (7) 21 25 8L/16P 1,744 ± 42G 480 ± 108** (8) 2 30 27 lOL/MD 484 ± 98 188 ± 4* (8) 3 21 30 12L/1'2D 1,638 ± 219 190 ± 76** (8) 4 14 28 12L/KD 3,04i2 ± 305 654 ± 181** (6) 8 28 121/120 3,042 ± 305 1,790 ± 423** (6) 8 32 12L/V2D 3,042 ± 305 871 ± 3'12** (8) of the May controls were in active spermatogen- esis (Stage 3) or in the prespawning conditions (Stage 4) ; ovaries of the May controls were in active vitellogenesis (Stage III). The testes of six of the eight fish at the short photoperiod, however, were regressing (Stage 0); ovaries in this group were regressing (Stage I) or in the quiescent phase (Stage II). In contrast to the effects of short photoperiod at 22°C, a long photoperiod (in May) did not initiate testicular regression. Although testic- ular weights in this group were significantly less (P < 0.05) than those of the initial May con- trols, the testes of all fish were in active sper- matogenesis (Stage 3). Ovarian weights of fish at a long photoperiod were also significantly low- er (P < 0.01) than those of the initial May con- trols; ovarian regression (Stage I) was occur- ring in all fish. Beginning in September, 26-day exposure to a short photoperiod (10L/14D) at 21°C stim- ulated a significant increase (P < 0.05) in ovar- ian and testicular weights when compared to gonadal weights in the initial controls (Exper- iment 12 — Table 5). In contrast, neither ovar- ian nor testicular weights in the long photope- riod group were significantly altered. The testes of the September controls were in Stage 2, and the ovaries of this group were in early stages of vitellogenesis ( Stage III ) . After short photo- period treatment, the gonads of all fish were in the meiotic phase of spermatogenesis (Stage 3) or vitellogenesis (Stage III) ; the long photope- riod, however, did not stimulate spermatogen- esis (testes in this group were in Stage 1) and caused ovarian regression (Stage I). Beginning in November (Experiment 13), testicular weights were maintained at the initial level for 21 days at 22°C and a short photoperiod (10L/14D) ; testes of the initial controls and the experimental fish were in Stage 3 or 4. A long photoperiod at 22°C, however, caused the initi- ation of testicular regression (Stage 0); testic- ular weights in this group were significantly Table 5. — Effect of 21° and 22°C treatments on testicular and ovarian weight in Gillichthys mirabilis. Experi- ment number 10 11 11 12 12 13 13 Beginning data April May May September September November November Length of treatment (Days) 17 30 30 26 26 21 21 Tempera- ture (°C) Photo- period Gonadal v>/eight (mg) (x ± SE) Initial controls 22 15L/9D moles: 97.8 ± 7.4 22 10L/14D males: 88.4 ± 7.1 females: 1,1263 ± 186 22 15L/9D males: 86.4 ± 7.1 females: 1,283 ± 186 21 lOL/'MD males: 42.8 ± 7-4 females: 233 ± 39 21 15L/9D moles: 42. 8 ± 7.4 females: 233 ± 39 22 15L/9D males: 68.5 ± 9.7 22 10L/14D males: 68.5 ± 9.7 After treatment (n) 99.2 It 8.8 (10) 4Q.3 ± 2.3** (8) 681 ± 242** (7) 70.6 ± 6.n* (8) 631 ± 2:16** (6) 61.5 ± 7.8* (8) 364 ± 44* (6) 36.0 ± 4.2 (7) 166 ± 21 (6) 46.9 ± 6.3** (10) 79.5 ± 5.1 (8) •Significantly different {P < 0.05) from initial controls. •'Significantly different (P < 0.01) from initial controls. 1145 FISHERY BULLETIN: VOL. 70, NO. 4 - lower (P < 0.01) than those of the initial No- vember controls. In spring spermatogenesis occurs at 22°C only if the photoperiod is long and in autumn only if the photoperiod is short. This temperature causes ovarian regression regardless of photo- period, suggesting that females may be more sensitive to temperature than males. Similarly, 21°C will promote the initiation of testicular and ovarian recrudescence only if photoperiod is short. Apparently then, the effects of photoperi- od at these temperatures are seasonally variable. EFFECTS OF PHOTOPERIOD AT 20oC ON FISH WITH REGRESSING OR QUIESCENT GONADS Experiments reported above showed that go- nadal recrudescence will not occur at 20°C if treatment is initiated in July but that gameto- genesis is maintained at this temperature at other times during the year. Thus, the influence of 36-day 20°C treatment at various photope- riods on gonadal recrudescence was examined in August 1968 (Figure 7). The gonads of the initial August controls were in Stages 0 and 1. Neither testicular nor ovarian weights in any of the experimental groups varied significantly from gonadal weights in the initial August con- trols, and there were no significant differences in gonadal weights among the experimental groups. Differences in gonadal histology were, nonetheless, evident. Testes of all fish collected from nature in August and September were in the regression phase (Stage 0). In all photo- period groups testicular regression was com- plete, and spermatogonial proliferation (Stage 2) or spermatogenesis (Stage 3) was occurring. The testes of 4 of 10 fish at a 15L/9D photo- period, however, were in the quiescent phase (Stage 1). A majority of ovaries from fish col- lected from nature in August and September were in the quiescent phase (Stage II) . In each experimental group, the ovaries of some fish were in the early phases of vitellogenesis (Stage III); with the exception of one fish, vitellogen- esis was initiated in all females at 15L/9D. These results indicate that gonadal recrudes- cence is initiated at 20°C if treatment is begun 30 20 it) .J« J<8) (8) ^Jrt (Id) rti w l&l 200 « Initial Controls (15 August) Nature 8L/I6D I2L/I2D I5L/9D Sorriple 21 September Figure 7.— Effect of 36-day 20 °C treatment at various photoperiods on ovarian and testicular weight in Gil- lichthys mirabilis. Shaded histograms illustrate mean ovarian weights ; open histograms illustrate mean testic- ular weights; the means are bracketed by one standard error. Photoperiod treatments are given below the his- tograms. Sample sizes are indicated atop each histo- gram. in August. Recrudescence was initiated after 36 days in all photoperiod groups, but a long photoperiod (15L/9D) was most eflfective in females and least effective in males. EFFECTS OF 16° C TREATMENT ON GONADAL FUNCTION Gonadal recrudescence does not normally oc- cur if Gillichthys is exposed to 20°C in July. To examine whether recrudescence could be initi- ated in July at a slightly lower temperature, the eflfect of 16°C treatment was determined. Before examining the effects of this temperature treat- ment on recrudescence, an experiment was con- ducted in January to be certain that gametogen- esis could be maintained at this temperature. The ovaries of the initial January controls (Table 6) were in Stage III, IV, or V; the testes of fish in this sample were in Stage 4 or 5. Neither testicular nor ovarian weights were sig- nificantly altered by this 90-day treatment, nor was there an evident change in gonadal activity as judged by histological examination. The testes and ovaries of the initial controls of the July experiment (Table 6) were regres- 1146 de VLAMING: CONTROL OF REPRODUCTION IN GILLICHTHYS sing (Stages 0 and I) . Following 80-day treat- ment at 16°C, ovarian and testicular weights were significantly greater (F < 0.01) than those of the initial controls. Testes of the experimen- tal fish were in Stage 3 or 5. Ovaries of fish in the 16°C group were in Stage II or III. The results of these experiments indicate that spermatogenesis and vitellogenesis are main- tained at 16° C. In fish with regressing gonads (collected in July), 16°C treatment initiates spermatogenesis and oogenesis. EFFECTS OF LONG PHOTOPERIOD TREATMENT (15°C) ON FISH WITH REGRESSING TESTES In July 1969 an experiment was initiated to determine whether a long photoperiod could block recrudescence at an "intermediate" tem- perature. All of the initial controls were under- going testicular regression (Stage 0); the mean (±: SE) testicular weight (mg) in this sample was 32.3 it 2.2. Fish were exposed to 15°C for 30 days at a 15L/9D photoperiod. At the ter- mination of the experiment a sample of fish was taken from nature; the mean testicular weight (± SE) in this group was 20.7 ± 3.2, and the testes of these fish were regressing (Stage 0). Testicular weights (x ± SE = 58.3 ± 7.3) of the 15°C group were significantly greater (P < 0.01) than those of the initial July controls and the September sample from nature. The testes of nine fish in the 15°C group were in Stage 3, and those of three were in Stage 4. These data suggest that 15°C treatment initiates testicular recrudescence, even when the photo- period is long. Table 6.— Effect of 16°C treatment (12L/12D) on testicular and ovarian weight in Gillichthys mirahilis. Beginning Length treatment (Days) Sex Gonadal weight (mg) (X ± SE) data Initial controls After treatment («) January July 90 ao Males Females Males Females 71.6 ± 4.9 1,321 + 295 32.3 ± 2.2 256 ± 23 66.7 ir 4.0 1,136 ± 238 54.5 ± 4.1** 438 ± 34** CIO) (1!) (10) (10) EFFECTS OF 12° AND 20°C TREATMENT ON FISH IN STAGES OF ACTIVE GAMETOGENESIS Data presented above suggest that gameto- genesis may be maintained more efl^ectively at 13°C than at 20°C. Beginning in March 1968, fish were exposied to 20° and 12°C for 21 days to examine this possibility (Table 7). The ovaries of the initial March controls were in phases of active vitellogenesis (Stage III) or the pre- spawning or postspawning condition (Stage IV or V) ; the testes of this group were in Stage 3. Table 7.— Effect of 21-day 12° and 20°C treatments on ovarian and testicular weights in Gillichthys mirabilis. Experimental group Gonadal (X weight (mg) ± SE) Males (n) Females (n) Initial controls (March) 59.3 ± 6.1 (11) 1,038 ± 198 Oil) 12°C treatment 74.7 ± 6.3* (10) 1,428 ± 106 (8) 20°C treatment 58.8 ± 5.7 (7) 1,375 ± 123 (8) Terminal controls (April) 62.2 ± 5.4 (8) 3,043 ± 297 (18) **Significantly greater [P < O.Ol) than initial controls. *Significantly greater (P < 0.05) fhan initial controls. Ovarian weights in the two experimental groups did not differ significantly from those of the initial March controls, but they were sig- nificantly less (P < 0.01) than ovarian weights of the sample taken from nature at the time of sacrifice (April). Ovaries of the 20° and 12°C groups were in Stage III or IV; the ovaries of a majority of the April sample from nature were in Stage IV. Testicular weights in the 20°C group were not significantly altered by treatment, but those of the 12°C group were significantly greater (P < 0.05) than the testicular weights of the initial March controls. The testes of fish in the 12°C group were in Stage 3 or 4; however, the testes of four of seven fish at 20°C were in Stage 2 and the remainder in Stage 3. These data indicate that at a 12L/12D photo- period, vitellogenesis and spermatogenesis will occur at both 12° and 20°C but that the lower temperature is more eff'ective. Neither of these treatments, however, was as effective in promot- ing vitellogenesis as the factors acting on the natural population. 1147 FISHERY BULLETIN: VOL. 70, NO. 4 INHIBITION OF TESTICULAR REGRESSION BY LOW TEMPERATURE An experiment was initiated in June 1968 to resolve whether low temperature treatment in combination with a short photoperiod could pre- vent gonadal regression at the "normal" time (early July). The testes of the beginning con- trols were in the prespawning or postspawning conditions (Stages 4 or 5). Fish were exposed to 10° or 20°C at a 10L/14D photoperiod for 21 days. The testes of fish collected from nature at the time of sacrifice (July) of the experimen- tals were regressing (Stage 0). Testicular weights in the 10°C group (x ± SE = 60.7 ± 4.1) and the 20°C group (61.0 ± 6.2) did not vary significantly from those (52.2 ± 3.9) in the initial June controls, but those of both groups were significantly greater (P < 0.01) than tes- ticular weights of the July sample from nature (32.1 ± 3.1). The testes of 10 fish exposed to 10°C and 6 fish to 20°C were in active spermato- genesis (Stage 3), and 4 in each group were in the prespawning condition (Stage 4). Thus, testicular regression does not occur at the nor- mal time when fish are exposed to temperatures between 10° and 20°C. COMPARATIVE EFFECTS OF 10° AND 18°C TREATMENT ON GONADAL RECRUDESCENCE To determine whether there is a differential eflect of 10° and 18°C on gonadal recrudescence, a 21-day experiment was initiated in August 1970 with fish having regressing or quiescent gonads (Stages 0 and 1). The effects of these treatments on gonadal weight are presented in Table 8. Table 8.— Effect of 21-day 10° and 18°C treatments (13L/11D) on ovarian and testicular weights in GUlich- thys mirabilis. Experimental Gonadal (X weight (mg) ± SE) Males (») Females in) Initial controls (August) 10°C treatment 18°C treatment 32.5 ± 2.1 42.6 ± 4.4* 41.7 ± 7.3* (10) (8) (8) 280 ± 19 339 ± 48 386 ± 33* (8) (8) (8) •Significantly greater [P < 0.05) than initial controls. A significant increase (P < 0.05) in testicu- lar weights occurred at both 10° and 18°C; tes- ticular weights in the two groups were not sig- nificantly different. Spermatogenesis was ini- tiated in all fish in both experimental groups. Ovarian weights in fish exposed to 18 °C were significantly greater (P < 0.05) than those of the initial August controls; however, ovarian weights in the 10°C group were not significantly different than those of the initial controls or those of the 18°C group. Nonetheless, the ova- ries of all fish in both experimental groups were in early phases of vitellogenesis (Stage III). These data indicate that there is little or no difference in the rate of initiation of gonadal recrudescence at 10° and 18°C. The possibility exists, however, that following the initiation of recrudescence, the rate of testicular and ovarian growth could be different at the two tempera- tures. DISCUSSION In previous work (Barlow and de Vlaming, 1972; de Vlaming, 1972b), the gonadal cycle of Gillichthys was observed to be closely correlated with seasonal changes in several environmental variables. Gonadal regression occurs as day- length begins to shorten and temperature reach- es a seasonal maximum. The initiation of go- nadal recrudescence coincides with the decline in temperature and the continued decrease in daylength. A majority of the spawning in this species occurs when daylength and temperature are increasing. The data presented here sug- gest that temperature may be important with regard to reproductive cycling, photoperiod act- ing only to modify the responses to temperature. GONADAL REGRESSION In these laboratory experiments, constant tem- peratures of 24°C and above cause testicular and ovarian regression at any time of the year re- gardless of photoperiod (i.e., photoperiod does not seem to have an influence on gonadal regres- sion) . As temperature is increased above 24°C, shorter treatment periods are required to attain the ovarian and testicular quiescent phases. 1148 de VLAMING: CONTROL OF REPRODUCTION IN GILLICHTHYS High temperatures bring about ovarian regres- sion by inhibiting vitellogenesis and causing atresia of all yolky oocytes. In males, high tem- peratures apparently increase the rate of mei- otic divisions and the process of spermiogenesis; this is apparent because the testes of fish sac- rificed soon after the initiation of high temper- ature treatment are characterized by large num- bers of secondary spermatocytes and spermatids. Fish sacrificed after longer treatments at high temperatures, however, are characterized by testes with only primary spermatogonia, sug- gesting that high temperatures inhibit the trans- formation of spermatogonia into primary sper- matocytes. High temperatures also cause the "flaking oflF" of cysts of spermatocytes from the germinal epithelium into the lumen of the tes- ticular lobules. Moreover, pyknotic degeneration of spermatocytes, spermatids, and spermatozoa occurs at high temperatures, followed by phago- cytosis of cellular debris. Mitotic proliferation of spermatogonia is inhibited above 25°C, but treatment at 27°C for 15 days does not inacti- vate the sperm remaining in the testes. Weisel (1948) showed that the spermatozoa of Gillich- thys remain alive in vitro for 2 weeks at 2°-4°C, but at 24°-26°C they are immobile in 33 hr. In Gillichthys the termination of the reproduc- tive season is apparently not endogenously timed. Regression is not "obligatory" since low temper- ature treatments (regardless of photoperiod) prevented gonadal involution at the "normal" time (July). These studies imply that the re- productive cycle of this species is primarily under exogenous regulation. A similar situa- tion has been reported in the cyprinodontid, Epiplatys bifasciatus which occurs in the Zio River and Lagoon of Lome of the Republic of Togo, Africa (Loiselle, 1969). Gillichthys is thus apparently a potentially continuous breeder but has a reproductive cycle imposed upon it by the increased temperatures of summer. Al- though little information is available on the causes of termination of reproductive cycles, differences are evident. For example. Bagger- man (1957) suggested that since none of the experimental conditions she tested could main- tain continuous breeding in Gasterostetis acule- atus, termination of the cycle is endogenously controlled. The rate of postspermatogonial re- gression is also accelerated by warm tempera- tures, and low temperatures retard the rate of sexual regression in F^mdulus heteroclitus (Lofts, Pickford, and Atz, 1968). GONADAL RECRUDESCENCE In Gillichthys gonadal recrudescence does not occur at constant temperatures of 24°C or above, regardless of photoperiod. Long-term experi- ments indicate that gonadal recrudescence is not initiated in males or females after nearly 4 months at high temperatures (comparable to summer temperatures). High temperatures also retard the early phases or intermediate phases of gametogenesis in Fundulus heterocli- tus (Burger, 1939), Phoxinus laevis (Bullough, 1939) , female Apeltes quadracus (Merriman and Schedl, 1941), Enneacanthus obesus (Harring- ton, 1956), female Fundulus confluentus (Har- rington, 1959), Couesius plumbeus (Ahsan, 1966), and Cymatogaster aggregata (Wiebe, 1968). Experiments, begun in winter and spring with the longjaw goby in phases of active gameto- genesis, indicate that gonadal activity is main- tained at 20°C over a wide range of photoperi- ods; long photoperiods may be more eff'ective in this regard, but more experiments are needed to prove conclusively the influence of photoperi- od. Beginning in July with fish having regres- sing testes, mitotic proliferation of spermato- gonia was stimulated, especially with a short photoperiod, but complete recrudescence did not occur at 20°C. However, beginning in August with fish having regressing testes, recrudescence did occur at 20°C; a short photoperiod was more eflfective in this respect. Beginning in Septem- ber, testicular and ovarian recrudescence is in- itiated at 21°C, but only with a short photoperi- od. The rate of spermatogenesis was, however, extremely low at these temperatures. Thus, a shift in gonadal responsiveness to 20°C appar- ently occurs between July and August. The ex- periment beginning in July was continued for 3 months without the initiation of spermatogenesis whereas the experiment beginning in August was terminated after a much shorter time. The 1149 FISHERY BULLETIN: VOL. 70, NO. 4 adaptive significance of the "refractoriness" to 20°C in July may be to prevent "premature" initiation of spermatogenesis should tempera- tures decrease somewhat during the regression phase. Although the data presented here are by no means conclusive, they suggest that photoperiod has a variable effect at 20°C. Long photoperiods may be more effective in maintaining spermato- genesis whereas short photoperiods seem to pro- mote a faster rate of recrudescence at 20°C. Moreover, active spermatogenesis is maintained at 22°C only with a long photoperiod. Differ- ences are evident between the sexes since vitel- logenesis does not occur at this temperature re- gardless of photoperiod. Since a majority of spawning in GilUchthys occurs when daylength is increasing and recrudescence occurs when daylength is decreasing, the variation in the ef- fects of photoperiod seen here seem reasonable. That 20°C is not as effective as lower temper- atures in maintaining gametogenesis or promot- ing gonadal recrudescence is consistent with the phenological and climatic data presented by de Vlaming (1972b). Average daily temperatures are below 20°C from early October to mid-June; it is during this period that most of the gonadal activity occurs in this species. Complete gonadal recrudescence in both male and female GilUchthys occurred at 12° and 13°C, suggesting that the decreasing temperatures in autumn are primarily responsible for regulating this process. Constant low temperature treat- ment promoted a faster rate of recrudescence than occurred in the natural population. Diurnal increases in temperature in the natural habitat during autumn may account for the slower rate of gonadal recrudescence. Short photoperiods augmented the rate of recrudescence at low tem- peratures in both males and females. Thus, the decreasing photoperiod in autumn probably fa- cilitates, but is not essential for, the effects of decreasing environmental temperatures in pro- moting gonadal recrudescence. With a 12L/12D photoperiod, gonadal recru- descence was initiated within 21 days at 10° and 18°C. This suggests that gonadal response in this species is relatively fast; rapid mobilization of energy for gonadal recrudescence may be characteristic of species that spawn more than once in a season. Whether recrudescence was more rapid at 10° or 18°C was not investigated, nor are there sufficient data to indicate whether an optimal temperature exists for the comple- tion of gonadal recrudescence. However, the data presented here show that gonadal recrudes- cence will occur over a wide temperature range (10°-20°C). This gonadal responsiveness to a wide range of temperatures may reflect an adap- tation to the labile thermal environment of this species. Active gametogenesis was maintained at low temperatures regardless of photoperiod. After treatment at a low temperature in May (Figure 4) , more male fish were in the prespawning con- dition at the longer photoperiod. In the same experiment, however, a short photoperiod was more effective in maintaining vitellogenesis (Figure 4) . Perhaps these results reflect a true photoperiod influence, but because of the vari- able gonadal conditions of the beginning controls no conclusive statements can be advanced. In fact, experiments begun in January with fish revealing less variable gonadal conditions indi- cated that long and short photoperiod treatments maintained active gametogenesis equally well at low temperature. Photoperiod influences, how- ever, could vary between January and May. The question arises as to why all fish treated at low temperatures did not progress to the pre- spawning condition. One possible explanation is that many of the experiments discussed here were relatively short term. In many of the long-term experiments some of the fish could be stripped of ripe eggs and sperm. It is likely, however, that physical factors other than tem- perature and photoperiod, and also social factors, influence final gonadal maturation and spawn- ing in this species. Indeed, Weisel (1947) has indicated that GilUchthys is territorial and has an elaborate courtship behavior. In addi- tion, de Vlaming (1971b) has shown that sa- linity changes and changes in the availability of food can alter the rate of gametogenesis in GilUchthys. Thus, the failure of many of the experimental fish to completely attain the pre- spawning condition was probably due to the ab- sence of certain conditions in the laboratory sit- 1150 de VLAMING: CONTROL OF REPRODUCTION IN GILLICHTHYS uation. Unfortunately, no data are available on the effects of exogenous factors on the fre- quency of spawning, fecundity, egg size, or sur- vivorship of the fry in this species, nor are data available on the influence of environmental fac- tors on the spawning act. For a complete under- standing of the role of the environment in the physiology and ecology of reproduction in Gil- lichthys, examination of these parameters is needed. The data presented here indicate that temper- ature may be the proximate factor regulating the Gillichthys reproductive cycle. Zambrano (1971) found that the secretory activity of the pituitary gonadotrophic cells is altered by tem- peratures, and this provides further support for this suggestion. However, since fish in their natural habitat experience large diurnal temper- ature fluctuations, the experiments reported here are not conclusive. In addition, Gillichthys is capable of regulating its body temperature by behavioral means (de Vlaming, 1971a). These experiments do set the physiological limits with regard to the influence of temperature on reproduction. Understanding the ecological meaning of temperature in reproductive cycling requires experimentation with diurnally fluctu- ating temperature; experiments of this nature are reported elsewhere (de Vlaming, 1972c), ACKNOWLEDGMENTS I am particularly indebted to Dr. Paul Licht for his continued interest and encouragement in this research. I am grateful to Dr. Licht and Dr. George Barlow for reading an initial draft of this manuscript and making many insightful suggestions, JoNell Biancalana also deserves special thanks for her assistance and encourage- ment in this research, I appreciate the assistance of Geraldine Ard and Abbey Reeder in typing this manuscript and Emily Reid in preparing the figures pre- sented herein. This work was supported in part by a research grant from the Graduate Division of the University of California and a National Institutes of Health predoctoral fellowship. LITERATURE CITED Ahsan, S. N. 1966. Effects of temperature and light on the cyclical changes in the spermatogenetic activity of the lake chub, Couesius plumbeus (Agassiz), Can. J. Zool. 44:161-171. Baggerman, B. 1957. An experimental study on the timing of breeding and migration in the three-spined stickle- back (Gasterosteus aculeatus L.) Arch. Neerl. Zool. 12:105-317. Barlow, G. W. 1961. Intra- and interspecific differences in rate of oxygen consumption in gobiid fishes of the genus Gillichthys. Biol. Bull. (Woods Hole) 121: 209-229. 1963. Species structure of the gobiid fish Gillich- thys mirabilis from coastal sloughs of the eastern Pacific. Pacific Sci. 17:47-72. Barlow, G. W., and V. L. de Vlaming. 1972. Ovarian cycling in longjaw gobies, Gillich- thys mirabilis, from the Salton Sea. Calif, Fish Game 58:50-57, Bullough, W, S, 1939. A study of the reproductive cycle of the minnow in relation to the environment. Proc. Zool. Soc. (Lond.), Ser. A, 109:79-102, Burger, J. W. 1939. Some experiments on the relation of the ex- ternal environment to the spermatogenetic cycle of Fundulus heteroclitus (L.). Biol. Bull. (Woods Hole) 77:96-103. Carpelan, L. H. 1957, Hydrobiology of the Alviso salt ponds. Ecol- ogy 38:375-390, DE Vlaming, V. L. 1971a. Thermal selection behaviour in the estu- arine goby, Gillichthys mirabilis Cooper. J. Fish Biol. 3:277-286. 1971b. The effects of food deprivation and salinity changes on reproductive function in the estuarine gobiid fish, Gillichthys tnirabilis. Biol. Bull. (Woods Hole) 141:458-471. 1972a. Environmental control of teleost reproduc- tive cycles: A brief review. J. Fish Biol. 4:131- 140, 1972b. Reproductive cycling in the estuarine gobiid fish, Gillichthys mirabilis. Copeia 1972:278-291, 1972c. The effects of diurnal thermoperiod treat- ments on reproductive function in the estuarine gobiid fish, Gillichthys mirabilis Cooper. J. Exp, Mar. Biol. Ecol. 9:155-163, Harrington, R, W., Jr. 1956. An experiment on the effects of contrasting daily photoperiods on gametogenesis and repro- duction in the centrarchid fish, Enneacanthus obesus (Girard). J. Exp. Zool. 131(3) :203-223, 1151 FISHERY BULLETIN: VOL. 70, NO. 4 1959. Effects of four combinations of temperature and daylength on the ovogenetic cycle of a low lat- itude fish, Fuiiduhis confluentus Goode & Bean. Zoologica (N.Y.) 44:149-168. Lofts, B., G. E. Pickford, and J. W. Atz. 1968. The effects of low temperature, and Cortisol, on testicular regression in the hypophysectomized cyprinodont fish, Fundulus heteroclitus. Biol. Bull. (Woods Hole) 134:74-86. LOISELLE, P. V. 1969. The biology of Epiplatys bifasciaUis (Stein- dachner, 1881) (Teleostomi: Cyprinodontidae: Rivulinae) in southern Togo. J. Am. Killifish Assoc. 6:40-63. Merriman, D., and H. p. Schedl. 1941. The effects of light and temperature on gametogenesis in the four-spined stickleback, Apeltes quadracus (Mitchell). J. Exp. Zool. 88: 413-449. SlEGEL, S. 1956. The Kruskal-Wallis one-way analysis of var- iance by ranks. In S. Siegel, Nonparametric sta- tistics for the behavioral sciences, p. 184-193. McGraw-Hill, N.Y. Weisel, G. F., Jr. 1947. Breeding behavior and early development of the mudsucker, a gobiid fish of California. Copeia 1947:77-85. 1948. Relation of salinity to the activity of the spermatozoa of Gillichthys, a marine teleost. Physiol. Zool. 21:40-48. Wiebe, J. P. 1968. The effects of temperature and daylength on the reproductive physiology of the viviparous seaperch, Cymatogaster aggregata Gibbons. Can. J. Zool. 46:1207-1219. Zambrano, D. 1971. The nucleus lateralis tuberis system of the gobiid fish Gillichthys mirabilis. III. Functional modifications of the neurons and gonadotropic cells. Gen. Comp. Endocrinol. 17:164-182. 1152 KINDS AND ABUNDANCE OF FISH LARVAE IN THE EASTERN TROPICAL PACIFIC ON THE SECOND MULTIVESSEL EASTROPAC SURVEY, AND OBSERVATIONS ON THE ANNUAL CYCLE OF LARVAL ABUNDANCE Elbert H. Ahlstrom^ ABSTRACT This is the second and concluding paper dealing with kinds and abundance of fish larvae in the eastern tropical Pacific based on collections made on EASTROPAC survey cruises. Main emphasis is placed on the composition and abundance of fish larvae on the second multivessel EASTROPAC cruise, occupied by three research vessels during August-Sep- tember 1967. This cruise, spaced 6 months after EASTROPAC I, affords interesting comparisons of composition and relative abundance of fish larvae during two contrasting periods of the year. Counts of fish larvae per haul on EASTROPAC II were about 50% greater than on EASTROPAC I; species composition, however, was strikingly similar in the two surveys. A portion of the EASTROPAC pattern, lying between long 119° to 98° W and lat 20°N to 3°S, was covered on four additional monitoring cruises — providing coverage of this more restricted area on six cruises, spaced at 2-month intervals, between February 1967 and January 1968. Essentially the same kinds of fish larvae were taken on each of the six coverages of the monitoring pattern, and for most species the range in relative abun- dance during the annual cycle was 3X or less. This report deals with the composition and relative abundance of fish larvae in the eastern tropical Pacific Ocean collected on the second multivessel survey cruise made as part of EAS- TROPAC, during August-September 1967. For brevity, the cruise is referred to in this report as EASTROPAC II (ETP II) . This cruise, con- ducted 6 months after EASTROPAC I (ETP I), deals with the composition of fish larvae at a contrasting period of the annual spawning cycle in tropical waters (Ahlstrom, 1971). Three research vessels participated in ETP II: Washington operated by the Scripps Institution of Oceanography occupied the outer pattern, Undaunted of the National Marine Fisheries Service occupied the middle pattern, and Rock- away operated by the Coast Guard took the in- ner pattern (Figure 1). ' National Marine Fisheries Service, Southwest Fish- eries Center, P.O. Box 271, La Jolla, CA 92037. Manuscript accepted February 1972. FISHERY BULLETIN: VOL. 70, NO. 4, 1972. The coverage during ETP II was less exten- sive than that of the four vessels of ETP I. One major line of stations of ETP I was omitted from ETP II, that along long 126°W, and the coverage below the equator was abbreviated in the two outer patterns, with lines ending at lat 10°S or 5°S. Comparison of the composition, relative abundance, and distributions of fish larvae at diflferent periods of the year in tropical waters is a primary objective of this report. The major comparison is with fish larvae obtained on ETP I (Ahlstrom, 1971); all of the 355 ETP II col- lections and an equivalent number of ETP I collections can be used to show similarities and diflferences in the composition of the larval fish fauna during two contrasting periods of the year. A portion of the EASTROPAC pattern was occupied by the National Marine Fisheries Ser- vice research vessel, David Starr Jordan, on 1153 FISHERY BULLETIN: VOL. 70, NO. 4 90" 80' Figure 1. — Location of plankton stations occupied by three research vessels participating in the second multivessel EASTROPAC survey (ETP II). Symbols for vessels indicated in legend above. Samples collected by Was/iingrfon are numbered as 45.000 series (for example 45.016, 45.140, 45.387), samples from J/jidawnied as 46.000 series, from Rockaway as 47.000 series. monitoring cruises, spaced at bimonthly inter- vals between the multivessel cruises; coverages equivalent to the monitoring pattern were sum- marized for ETP I and ETP II, in order to com- pare the results of six bimonthly coverages of the same area (Figure 2). The monitoring pat- tern lacked coverage in the nearshore portion of the EASTROPAC pattern that was occupied by the inner vessel on multivessel surveys. Ad- ditional seasonal information about composition and relative abundance of fish larvae in this area was supplied by a "zig-transect" of 50 sta- tions occupied by the RV Oceanographer of the Environmental Science Services Administration during November 1967 — 2 to 3 months after ETP II coverage of this area (Figure 2), The methods of making zooplankton hauls on ETP II were identical to those previously de- scribed for ETP I (Ahlstrom, 1971). This pa- per deals solely with collections obtained from oblique hauls made with a net, 1-m mouth di- ameter, constructed of 505 fi nylon (Nitex) cloth, with approximately a 5: 1 ratio of effective straining surface, i.e., pore area to mouth area. 1154 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Figure 2. — Location of monitoring pattern (large solid circles), occupied between multivessel EASTROPAC cruises at 2-month intervals by David Stai-r Jordan, and of zig-transect pattern (triangles) occupied by Oceanographer during November 1967, superimposed on ETP II pattern. As on ETP I, this net was paired in an assembly frame with a finer-meshed 0.5-m net. Each tow attempted to obtain a uniform sampling of zoo- plankton in the water column between the sur- face and approximately 200-m depth. The net assembly was lowered to depth by paying out 300 m of towing cable and then retrieved at a uniform rate. The depth reached by the net was estimated from the angle of stray (departure from the vertical) of the towing cable. The av- erage depths of haul taken by the three research vessels are summarized in Table 1. Only slightly over two-thirds of the hauls were lowered to depths of 200 m or more and about one-eighth were taken at depths shallower than 180 m. Based on variation in depths sampled, speed of hauling was controlled less consistently on ETP II as compared with ETP I. Four plankton collections were made each day with the paired net-assembly, at about 6-hr in- tervals. Timing of hauls was not coordinated between research vessels (Table 2), Usually Rockaway spaced the four hauls at approximate- ly 0500, 1030, 1700, and 2300; hauls from Wash- 1155 FISHERY BULLETIN: VOL. 70, NO. 4 Table 1. — Depths of paired oblique pankton hauls taken by the three research vessels in EASTROPAC II (net lowered by paying out 300 m of towing cable). Average depth of haul Number of hauls token at each depth interval from Washington Undaunted Rockaway All (m) 45.000 Series 46.000 series 47.000 Series vessels 80.1- 90.0 ._ 1 __ 1 90.1-100.0 _. _• — 100.1-110.0 _. __ — — 110.1-120.0 __ 1 — 1 120.1-130.0 __ 2 2 130. 1-140 jO 1 __ 1 140.1-150.0 I _^ 3 4 150.1-160.0 2 4 1 7 160.1-170.0 __ 5 4 9 170.1-180.0 7 4 6 17 180.1-190.0 6 4 3 13 190 1-200.0 24 16 14 54 200.1-210.0 36 22 52 110 210.1-220.0 28 28 56 112 220.1-230.0 5 6 7 13 230.1-240.0 1 _^ 1 2 240.1-250.0 ^^ _» 1 1 250.1-260.0 __ _. 1 1 260.1-270.0 — 2 ~ 2 Total Ill 95 149 355 Table 2. — Hour of day that paired oblique plankton hauls were made from the three research vessels par- ticipating in EASTROPAC II (midtime of haul used). Hour of day Number of hau each hour Is take of the n during day Washington Undaunted R ockaway All vessels OOOl^lOO 10 10 8 28 0101-0200 4 0 1 5 0201-0300 1 0 3 4 0301-0400 0 1 3 4 0401-0500 0 10 14 24 0501-0600 5 8 15 28 0601-0700 12 3 4 19 0701-0800 5 1 0 6 0801-0900 2 0 1 3 0901-1000 1 0 6 7 1001-1100 1 0 21 22 1101-1200 10 14 6 30 1201-1300 11 11 1 23 1301-1400 3 0 1 4 1401-1500 3 0 2 5 1501-1600 2 0 12 14 1601-1700 0 0 16 16 1701-1800 10 22 4 36 1801-1900 9 3 1 13 1901-2000 5 0 0 5 2001-2100 2 0 3 5 2101-2200 1 0 5 6 2201-2300 0 0 11 11 2301-2400 14 12 11 37 Total 111 95 149 355 ington usually were taken at approximately 0630, 1200, 1800, and 2400; and hauls from Undaunted at 0500, 1200, 1730, and 2400. At least some hauls were taken during every hour of the day. EFFECTIVENESS OF SAMPLING FISH LARVAE IN DAYLIGHT HAULS AS COMPARED WITH NIGHT HAULS Catches of fish larvae for selected families in day hauls compared to night hauls or to hauls taken within ± 1 hr of sunrise or sunset are summarized in Table 3. For all categories of larvae combined, the catch was 212.0 larvae per daytime haul and 480.4 larvae per night haul, a difference in catch of 2.27 X. Hauls taken within ± 1 hr of sunrise or sunset had an av- erage catch of 347.0 larvae, intermediate be- tween day and night catches. Difference between day and night collections was somewhat less than for ETP I; on that sur- vey the average count of larvae in night hauls was 2.76 X as many as in day hauls (Ahlstrom 1971, Table 4). On both surveys gonostomatid larvae had the most marked differences in catches between night hauls and day hauls: 4.3 X as many, on the average, in night collec- tions compared with day on ETP I, 2.9 X as many in night collections on ETP II. Night-day diflferences in catch per haul of myctophid lar- vae were less marked between the two surveys: 3.0 X on ETP I as compared with 2.6 X on ETP II. Night-day differences in average catches of bathylagid larvae were similar on the two multi- vessel surveys: 1.5 x as many per haul on the average, in night collections compared with day collections. Sternoptychid larvae, which were sampled almost as well in day hauls as in night hauls on ETP I, showed a somewhat greater night-day diflference on ETP II: 1.7 X for ETP II as compared with 1.2 x for ETP I. Scombrid larvae were taken in lesser num- bers per haul in both day and night hauls on ETP II compared with ETP I; in contrast to ETP I, however, (where a night-day difference of 3.7 X was observed) no difference was ob- tained in night and day collections on ETP II. 1156 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE S3 © I. u 1. c 3; a > 5 U — © n 0 S cs CO CN rx --^ o lo CM rj (^ — u-J ■* f^" — -* w" — CO +1' Z D> CD C t- c a > 3 < c c 0 _ o .t; 1/) f> 5 - i-_2 c « o ~z ^ D D*^ X o )r © « -0--rr t ;^ 0 -3 o-c Z a IZ © Q, U O) c n V ° I la- © Q. U O >- O ^ II 3 - _ © ,2S zR-" a — o D ,^ *0 CN CO OOOOO'^O'OCNCN cs fo o r\ lo lO CO O^ CO CO CN CN — 00 ^0 CM -^ ■ CO 3 o CO r\ -^ — CN — r^ CM K ^^C>lOC9"*COOO'^T\CN'^'-OCN^O^^OCOCN iC^co-O'^'O'ooo-oto-o-* — — — — CNCM^ — rx'Trs ^/)C^^-.■>»•>qcN^q'tcqc^■OCDKcOCOcO oCN-^rcicNC-itvcb^-^oocDciocJoo'^oodc^-o ~ O CN lO CN ^ — CN h^ "^ ^o vo rv o -o n CO o § CN - CO CN 2 " r\^xoc^^cocN^^■^ -^t^O--COC0^0. a o c ■S o .J O ^ -H o © D Oc^lj30Sc2c)5Smi2H;i6 sj^ChuioCOC^<^^OCOCNOo'oOOOCOOOCNc6 CO ^ -O CN O — — so ^ rv CN sO D © > 0 > _D i_ _o -D © ^ tA C ^ 1157 FISHERY BULLETIN: VOL. 70, NO. 4 WATER TEMPERATURES ON EASTROPAC II Water temperatures were available at 1-m intervals from the surface to about 750-m depth for each station at which an STD was used for determination of salinity and temperature. I selected three depths for tabulation and study: surface, 10 m, and 50 m. STD readings were available for 347 of the 355 plankton stations taken on ETP II. A chart of surface temper- ature for ETP II will be included in the EAS- TROPAC Atlas. To facilitate discussion of distributions of fish larvae, I have found it convenient to divide the EASTROPAC area into quadrants with the north-south division at the equator and the east- west division at long 100 °W. I will use these divisions when discussing distribution of tem- peratures on ETP II, except for separating out a narrow band of water at the equator (lat 2°N to 2°S). Within a quadrant the temperatures are summarized by 5° latitude, except near the equator (Table 4). In some parts of the ETP II pattern, the ther- mocline was considerably deeper than 50 m, so that the temperature at 50 m was similar to that at the surface. At a few stations in the north- east quadrant, where the thermocline was almost at the surface, the temperature at 10 m was 5° to 10° C lower than at the surface. At most sta- tions in this quadrant the thermocline was con- siderably shallower than 50-m depth, as attested by marked differences in temperature between the surface and 50 m (Table 5). Mixed layer depths on ETP II were illustrated in Blackburn, Laurs, Owen, and Zeitzschel (1970) — Figure 7 on page 27. A brief summary of the tempera- ture structure is given by quadrant. NORTHEAST QUADRANT, EXCEPT WITHIN 2o LATITUDE OF EQUATOR Surface temperatures in this quadrant were high, ranging between 25.4° and 29.8°C (av- erage 27.2°C) , Temperatures at 10 m were usu- ally the same or within ±0.5°C of the surface, although 10 stations showed differences of more than 1°C and 7 of those were 4.6° to 10.1°C lower. These marked differences are indications of very shallow thermoclines; five contiguous Stations along long 88°W offshore from Pun- tarenas, Costa Rica, had such near-surface thermoclines. At most stations the thermocline was shal- lower than 50 m; at 79 of 91 stations, the tem- perature at 50 m was 5° to 15°C lower than at the surface, and at half of these the temperature was between 10° to 15°C lower at 50 m. NORTHWEST QUADRANT, EXCEPT WITHIN 2° LATITUDE OF EQUATOR Surface temperatures in this quadrant, 24.8° to 29.7°C (average 27.6°C), were similar to those of the inshore quadrant. Highest surface temperatures, averaging 28.3°C, were encoun- Table 4. — Range of temperatures at surface, 10 m, and 50 m summarized by 5° latitude or smaller intervals for both offshore (long 100°-119°W) and inshore (coast to long 98°W) for EASTROPAC II. Offshore: long 100°-119°W Inshore: coast to long 98°W Lotitude No. stn. Ranga in temperature {°C) at: No. stn. Ron ge in temperature) (°C) at: Surface 10 m 50 m Surface 10 m 50 m 15°-20°N 32 25 7-29.7 25.5-29.6 17.7-28.6 1 29.8 29.6 23.1 10°-15°N 30 26.3-29.8 26.0-29.4 16.7-27.9 23 26.4-29.7 19.7-29.6 14.0-26.7 5°-I0°N 20 25.8-27.7 25 8-27.8 16.5-27.4 42 25.8-28.9 16.3-28.9 13.5-23.6 2°- 5°N 15 24.8-26.6 24.8-26.4 29.3-26.2 25 25.2-26.8 25.3-27.0 15.2-25.9 Total 2''-20°N 97 24.8-29.7 24.8-29.6 16.5-28.6 91 25.2-29.8 16.3-29.6 13.5-26.7 Equator 2°N-2''S 20 19.5-25.0 19.0-25.0 15.7-22.4 30 16.4-25.9 15.6-25.9 13.7-17.4 2°- 5°$ 14 20.6-23.0 20.5-23.0 14.0-22.6 19 18.4-22.2 17.4-21.9 14.2-20.2 5°-10°S 16 22.9-24.9 22.7-24.9 22.4-24.9 30 16.3-21.5 16 3-21.5 14.6-21.4 10°'I5°S 0 ._ 30 15.4-21.4 15.1-21.4 13.9-21.2 Total 2°-15°S 30 20.6-24.9 20.5-24.9 14.0-24.9 79 15.4-22.2 15.1-21.9 13.9-21.4 1158 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Table 5. — Summary of temperature differences within upper 50 m depth (differences in temperature at the surface and at two selected depths — 10 m and 50 m) summarized by quadrants. Area No. stn. Diffe the su rence rface in 1 and temperature at at 10-m depth Difference in 1 the surface and temperature at at 50-m depth 0°C 0.1°-0.5°C 0.6° ,1.0' 'C 1 .r-5.o°c 5.r '-lo.rc o°c 0.1 °-I.0°C i.r °-5.0' 'C 5.]° '-IO.O°C 10.1 °-15.0 = C NE Quadrant (lot 2°-16°N, coast to long 98 °W) 90 38 39 3 4 6 0 2 9 40 39 NW Quadrant (lot 2°-2.0''N, long 100^119°W) 97 65 26 5 I • 0 6 14 46 27 4 Equator (lot 2°N-2°S, coast to long 98°W) 30 19 8 1 2 0 0 0 10 19 1 Equator (lot 2°Ni2''S, long 100-1 19°W) 20 9 9 2 0 0 0 1 13 6 0 SE Quadrant (lot 2°-15''S, coast to long 98°W) 79 61 IS 2 1 0 20 31 27 1 0 SW Quadrant (lot 2"'-10°S, long 100-1 19°W) 30 17 12 I 0 0 9 15 5 1 0 tered between lat 15° and 20°N. Temperatures at 10 m were usually the same as at the surface, and in only one instance was the difference as great as 1.3 °C. Temperatures at 50 m were identical to, or within 1°C of, the surface tem- peratures at about 20 9r of the stations, all lo- cated between lat 2° and 10°N — these were sta- tions with deep thermoclines. Temperature differences between the surface and 50 m ex- ceeded 5°C at about 35% of the stations. EQUATOR, LAT 2oN TO 2oS Surface temperatures were variable, with 9.5°C range (16.4° to 25.9°C). Lowest surface temperatures, undoubtedly resulting from up- welling, were encountered seaward of the Ga- lapagos Islands, between long 92° and 98°W, lat 0.5°N to 2.0°S, but cold water was also en- countered farther offshore. Thermoclines were shallow at most stations inshore from the Galap- agos Islands, the difference between surface and 50 m exceeded 5°C at about 63% of the stations, but the surface water was warmer than offshore. SOUTHEAST QUADRANT, EXCEPT WITHIN 2° LATITUDE OF EQUATOR Water temperatures were much lower in this quadrant than in the northeast quadrant. Few surface temperatures were as high as 20 °C, and the average surface temperature was 18.7°C. The thermocline was usually deep. At 65% of the stations the temperature at 50 m was the same as that at the surface or was not more than 1°C colder. SOUTHWEST QUADRANT, EXCEPT WITHIN 2o LATITUDE OF EQUATOR Surface temperatures ranged from 20.6° to 24.9°C (average 21.0°C), Temperatures at 10 m were usually the same as that at surface or within 0.5°C. Temperatures at 50 m were identical to the surface at 30% of stations and within 1°C of the surface temperature at 80% of stations, indicative of a region of deep ther- mocline. In summary, water temperatures were much higher north of the equator, than south of the equator. Surface temperatures averaged 8.5 °C higher in the northeast quadrant than in the southeast quadrant. Offshore the differences were almost as great; surface temperatures averaged 6.6°C higher in the northwest quad- rant than in the southwest quadrant. As noted in the paper dealing with ETP I col- lections, information is mostly lacking on depth distribution of fish larvae in the eastern trop- ical Pacific. Because of the marked variation in depth of thermocline encountered in different 1159 FISHERY BULLETIN: VOL. 70, NO. 4 parts of the EASTROPAC pattern, ranging from near-surface to deep, it is anticipated that depth distribution of larvae will be markedly affected by the temperature structure. A care- fully planned study of depth distribution of larvae in the eastern tropical Pacific in relation to temperature and thermocline depth is badly needed. Lacking this, it is difficult to meaning- fully relate larval distributions to temperature. A REVIEW OF SIGNIFICANT PAPERS DEALING WITH ADULT FISHES OF THE EASTROPAC AREA A working knowledge of the adult fishes of an oceanic region is a necessary prerequisite to meaningful study of the fish larvae of that re- gion. Most larval series, initially, are estab- lished by working backwards from larger iden- tified specimens (late-larvae or early juveniles) to early-stage larvae. Until recently shore fish- es of the eastern tropical Pacific were much better known than deep-sea fishes, e.g., studies by Meek and Hildebrand (1923, 1925, 1928) for Panama and Hildebrand (1946) for Peru. Shore fishes, however, were not an important element of the EASTROPAC ichthyoplankton. A major contribution to our knowledge of eastern Pacific fishes was made by Carman (1899), who worked up the fishes collected on the Albatross Expedition of 1891 to the west coasts of Mexico, Central and South America, and off the Galapagos Islands, Carman dealt with 180 species of fish, most new to science; about a third of these were pelagic, oceanic fishes. Included among the latter are the two most common pelagic fishes, Diogenichthys la- ternatus and Vinciguerria lucetia, in the eastern tropical Pacific, based on their abundance as larvae. The second oceanographic expedition of the Pawnee to the eastern Pacific in 1926 added ma- terially to our knowledge of the deep-sea fishes. Several of the species described by Parr (1931) from these collections are common as larvae in EASTROPAC plankton hauls, including Bathyl- agus nigrigenys, Diaphus pacificus, Larnpanyc- tus idostigma, L. parvicauda, and Scopelarch- oides yiicholsi. The New York Zoological Society sponsored several expeditions to the eastern Pacific which stimulated papers on Pacific Myctophidae by Beebe and Vander Pyle (1944) and on ceratioid fishes by Beebe and Crane (1947). The paper on myctophids contains information on taxon- omy, biology, and zoogeography of 24 species of myctophids of which none were new. The paper by Beebe and Crane on deep-sea ceratioid fishes dealt with 24 species belonging to six families, of which 10 were new. The ceratioid fishes of the Gulf of Panama had received attention previously: the Danish research vessel Dana had occupied several very productive stations in the Gulf of Panama in 1922, from which Regan (1926) described 18 species of ceratioids, mostly new. Bertelsen (1951) reported taking early life history stages of 23 kinds of ceratioids from the Gulf of Pan- ama in Dana collections from its round-the- world expedition of 1928-30. Information on fishes off Peru was obtained on the Yale South American Expedition of 1953, Morrow (1957a) gave an annotated list of 104 shore fishes, 21 new to the Peruvian fauna, and Morrow (1957b) gave an annotated list of 18 mid-depth fishes. Bussing (1965) reports on 15 pelagic trawl hauls made on Eltanin cruises taken off the South American coast in 1962 and 1963 between lat 3° and 35°S, The collections contained 100 species, representing 33 families. Four trawl hauls were made within the EASTROPAC area; only one yielded substantial numbers of speci- mens. This was Eltanm Station 34 at lat 7°45' to 7°48'S, long 81°23'W, from which 45 species were obtained, Haedrich and Nielsen (1966) provided an- notated identifications of 32 species (21 fami- lies) of fishes from stomachs of Alepisaurus collected at 19 stations by exploratory longline fishing from the Japanese RV Shoyo Maru. Four collections were obtained within the EAS- TROPAC area, and the other 15 between lat 20° and 40°S, Craddock and Mead (1970) reported on col- lections made along two transects through the 1160 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE southern portion of the Peru Current off Chile at lat 34°S. They provide annotated identifi- cations of 133 species. Although these transects were south of the EASTROPAC area, many of the species also occur in the EASTROPAC area. Parin (1971) reports on collections of mid- water fishes of the Peru Current zone collected on the fourth cruise of RV Akademik Kurchatov. He lists about 150 species representing 33 fami- lies. Collections were obtained between lat 5°N and 30°S, in a broad coastal band, extending offshore to long 90 °W, Distributions are illus- trated for 24 species. In addition to the above, a number of ref- erences dealing with particular species of genera or families of eastern Pacific fishes are cited in the body of the text, or were referred to in Ahl- strom (1971). NUMBER OF FISH LARVAE OBTAINED ON EASTROPAC II Fish larvae were obtained in all collections (355) made with the 1-m plankton net on ETP II; counts of larvae per haul ranged from 1 to 2,864, and averaged 347. Four collections con- tained 10 or fewer larvae, and 22 collections contained 1,000 or more specimens in each (Table 6). Abundance of fish larvae according to latitude and proximity to shore within the EASTROPAC pattern is summarized in Table 7. The same grouping of stations by latitude (5° except near the equator) and longitude (inshore-oflFshore) is used as in Table 4 (temperature summary table). Subtotals provide a rough separation into quadrants. Larvae were taken in greatest abundance in the northeast quadrant, particularly between lat 5° and 10°N; in this latter area, with an average surface temperature of 27.1°C, larvae averaged 639 per haul. Larvae were less abun- dant in the southeast quadrant, with numbers decreasing southward and averaging only 118 larvae per haul between lat 10° and 15°S (av- erage surface temperature, 18.1°C), Larvae were much less abundant offshore, in the northwest quadrant, averaging slightly over 40% as many per haul as in the inshore (north- east) quadrant. Surface temperatures, how- ever, were quite similar. Near the equator (lat 2°N to 2°S), larvae were moderately abundant inshore (434 per haul), and the decrease in the abundance off- shore was not as marked as in other areas (362 per haul). This is not surprising, since this was an area of upwelling. In the southwest quadrant (lat 2° to 10°S, long 110° to 119°W), there was a decrease in abundance toward the south. However, this quadrant was poorly sampled on ETP II. When compared to inshore coverage of the same lat- itude (lat 2° to 10°S), abundance of larvae per haul averaged about 62% as many. Fish larvae averaged more per haul on ETP II as compared with comparable coverage on ETP I, 347.0 versus 231.9 larvae per haul; the in- crease in abundance was reflected in all parts of the EASTROPAC pattern. The majority of large collections of fish larvae were made at stations with shallow thermo- clines and relatively high mixed layer water Table 6. — Relative numbers of fish larvae obtained over the three vessel patterns oc- cupied on EASTROPAC II; last column gives comparable counts for EASTROPAC I. No. of fish larvae per haul Washington 45.000 Series Undaunted 46.000 Series Rockaway 47.000 Series All patterns — EASTROPAC II Comparable Coverage — EASTROPAC 1 0 1-10 n-100 101-1,000 1,001 and over 0 1 29 78 3 0 0 16 72 7 0 3 27 107 12 0 4 72 257 22 4 6 122 214 9 Total 111 95 149 355 355 Average no. larvae per haul 224.0 400.3 404.7 347.0 231.9 1161 FISHERY BULLETIN: VOL. 70, NO. 4 Table 7. — Total catches of fish larvae (actual counts) taken on EASTROPAC 11 sum- marized by latitude (5° except near equator) and longitude (offshore or inshore). Offshore: long 100M19°W Inshore: coast to long 98°W Latitude Number stations occupied Number per of larvae haul Number stations occupied Number per of 1 hou arvae 1 Range Mean Range Mean 15°-20°N 32 31-1,048 219.0 1 130 130.0 lO'-Ii'N 30 58- 481 159.1 23 37-2,242 435.4 5°-10°N 21 2M,128 237.4 42 61-2,864 639.1 2° S'-U 16 93-1,217 359.6 28 141-1,975 555.1 Total 2''-20"'N 99 21-1,128 227.5 94 37-2,864 558.8 Equator 2°N-2''S 21 30-1,506 361.5 30 1-1,513 434.4 2°- 5°S 14 79' 817 268.5 20 6-1,061 431.4 5°-10''S 16 8- 472 178 2 30 27-1,002 287.3 I0°-I5°S _^ __ 31 4- 579 118.0 Totol 2°-l'5°S 30 8. 817 220.4 81 4-1,061 258.1 Grand total 150 8-1,506 244.8 205 1-2,864 421.8 temperatures. Over 60% of the larger collec- tions of fish larvae (750 or more larvae) were taken at stations with mixed layer temperatures in excess of 26°C and mixed layer depths of 35 m or less. Unfortunately, information is lacking on the depth distribution of fish larvae in the eastern tropical Pacific in relation to thermocline depth, hence it is not known whether most kinds were limited in depth distribution to the upper mixed layer, as reported for California Current waters (Ahlstrom, 1959). KINDS OF FISH LARVAE OBTAINED ON EASTROPAC II With some interesting exceptions, the same kinds of larvae were obtained on ETP II as on ETP I, and Table 8, the principal summary table covering ETP II larvae, contains essentially the same families as its counterpart for ETP I. The table lists 53 families and 6 composite categories including 3 orders or suborders and a catchall category — other identified. For the latter, com- position by families is given in subsequent tables or in text discussions. Altogether, fish larvae of 82 families were represented in ETP II col- lections. As on ETP I, larvae of 10 families contributed over 907o (91.0 on ETP II) of the total; 9 of these families were among the first 10 on both EASTROPAC surveys and had simi- lar rankings. The first 10 families on ETP II were as follows: Myctophidae, 52.0% ; Gonos- tomatidae, 19.7%; Sternoptychidae, 6.0%; Bathylagidae, 4.8%; Bregmacerotidae, 2.5%; Paralepididae, 2.0%; Nomeidae, 1.2%; Melam- phaidae, 1.1%; Engraulidae, 1.1%; and Idia- canthidae, 0.6% . Engraulidae is the only family on this list that did not rank among the first 10 on ETP I. The sole displacement from the pre- vious list is the family Scombridae, which slipped in ranking from fifth in ETP I to twen- tieth in ETP II. Of the remaining 9%, 2.3 %o were too damaged (disintegrated) to identify, 0.7% could not be identified (these were mostly very small larvae), and the remainder, about 6%, belonged to the other 72 families. The displacement of Scombridae from among the 10 most abundant families on ETP II left only one perciform family, Nomeidae, among those contributing 1 % or more of the total. Only a moderate number of perciform families have, become widely distributed in off"shore oceanic waters. Among these, larvae of Gempylidae con- tributed 0.3 %r of the total on ETP II; Apogoni- dae, 0.2%; Chiasmodontidae, 0.2% ; Coryphae- nidae, 0.1%; Trichiuridae, 0.1%; and Brami- dae, 0.1%. The basic data on the kinds and number of fish larvae obtained in the 355 ETP II collections are contained in Appendix Tables 1 to 6. These are keyed to Table 8 and to other tables in this report. The data presented in this paper represent but the first step in utilizing eggs and larvae col- lections for resource evaluation. 1162 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Table 8. — Occurrences and counts of fish larvae taken in oblique 1.0-m plankton hauls on the second multivessel EASTROPAC survey (EASTROPAC II), summarized by family or larger grouping and by research vessel. Family or Basic station data contained in Appendix Table no. Distribution Washington 45.000 Series Undai 46.000 inted Series Rockaway 47.000 Series Total EASTROPAC II larger grouping^ By family By genus ~" snown in Figure no. No. No. No. No. No. No. No. No. larvae or larger grouping or species positive hauls larvae positive hauls larvae positive hauls larvae positive hauls t Albulidae 4 0 0 0 0 2 9 2 9 *2 Clupeidae 4 3 0 0 7 185 2 85 9 270 *3 Engraulidae 4 3 4 15 2 3 29 1,342 35 1,360 *4 Argentinidae 3 14 21 14 32 4 5 32 58 *5 Bathylagidae 1 3 4 74 352 90 1,277 134 4,262 293 5,891 *6 Gonostomatidae 1 3 5-7 1.08 9,079 94 7,095 140 8X>81 342 24,255 *7 Sternoptychidae 1 75 1,882 74 1,941 128 3,562 277 7,385 *8 Astronesthidae 3 5 10 16 IS 30 17 28 42 74 *9 Chauliodontidae 1 8 10 25 10 15 36 167 S6 207 *10 Idiacanthidae 1 9 49 275 62 219 7Q 301 131 795 *1 1 Other Stomiatoidel 1 3 60 245 64 219 86 570 210 1,034 12 Chlorophthalmidae 0 0 3 5 1 3 4 8 *I3 Evermannellidae 3 10 9 47 7 19 1 1 17 67 *14 Myctophidae 1 2 3,8, 11-14 111 9,546 95 21,082 146 33,381 352 64,009 15 Neoscopelidae 5 5 2 3 5 7 12 15 *I6 Paralepididae 1 3 91 497 80 1,002 76 1,036 247 2,535 *17 Scopelarchidae 1 39 103 45 92 50 103 134. 298 *I8 Scopelosauridae 3 10 0 0 11 46 29 344 40 390 *I9 Synodontidae 4 5 1 1 2 6 11 53 14 60 20 Alepisauridae 1 1 1 2 a 2 4 5 *21 Anguilliformes 5 14, 15 16 33 30 42 35 76 81 151 *22 Melamphaidae 1 16 79 262 83 408 122 695 284 1,365 *23 Trachichthyidae 3 0 0 0 0 11 70 11 70 24 Holocentrldae 3 10 0 0 0 0 3 10 *25 Bregmacero+idae 1 63 379 47 1,624 50 1,059 160 3,062 26 Macrouridae 1 1 3 3 5 5 9 9 *27 Scomberesocidae 3 14 0 0 0 0 27 153 27 153 28 Exocoetidae 1 15 26 22 33 23 87 59 146 29 Trachypteridae 1 10 10 16 20 20 29 46 59 *30 Apogonldoe 1 28 178 24 73 14 32 66 283 31 Balistidoe 3 1 1 3 6 1 I 5 8 32 Blenniidae 0 0 0 0 5 6 5 6 33 Bramidae 1 21 31 17 24 29 41 67 96 *34 Carangidae 4 17 8 28 a 59 20 137 36 224 35 Carapidae 4 0 0 0 0 7 7 7 7 36 Chiasmodontidae 1 25 46 25 45 50 146 100 237 *37 Coryphaenidae 1 37 56 34 62 38 67 109 185 *38 Gempylidae 3 4 36 59 30 64 46 247 112 370 *39 Gobiidae 4 S 65 11 33 37 286 53 384 40 Labridae 4 10 9 21 9 10 14 26 32 57 41 Mugilidae 4 0 0 0 0 5 16 3 16 *42 Nomeidae 1 68 357 64 391 97 712 229 1,4<&0 *43 Ophidiidae 0 0 7 9 31 72 38 81 44 Polynemidae 4 2 21 2 5 3 5 7 31 45 Pomacentridae 4 1 6 0 0 4 5 3 11 4AA 4fsv' v> FISHERY BULLETIN: VOL. 70, NO. 4 Table 12. — Relative abundance and percentage contribution of fish larvae of the 10 most common families within that portion of EASTROPAC area covered on six successive bimonthly cruises between February 1967 and Jan- uary 1968. Family ETP multi (Feb.- vessel |i Mar.) ETP monitoring cruise #20 (Apr. -May) ETP mor cruise (June- litoring #30 July) ETP multi (Aug.- vessel II 2 Sept.) ETP monitoring cruise #50 (Oct.-Nov.) ETP monitoring cruise #60 (Dec-Jan.) Six cruises monitoring - ETP area Averaga Average Average Averoge Average Average Average no. per % no. per % no. per % no. per % no. per % no. per % no. per % haul haul haul haul haul haul haul Myctophidae 104.9 49.0 100.4 50.7 67.6 37.9 116.4 45.6 103,7 57.1 58.7 44.1 91.2 47.5 Gonostomatidae 48.6 22.7 50.6 25.6 67.3 37.7 77.9 30.5 32.6 18.5 33 7 25.3 51.6 26.9 Sternoptychidae 14.2 6.6 17.1 8.6 9.5 5.3 12.8 5.0 V3 5 7.6 12.9 9.7 13.3 7.0 Bathylagidae 6.4 3.0 3.9 2.0 2.9 1.6 49 1.9 3.2 1.8 2.6 2.0 4.0 2.1 Paralepididae 4.8 2.3 4.1 2.1 5.3 3.0 8.2 3.2 3.2 1.8 4.8 3.6 5.1 2.7 Nomeidae 3.6 1.7 35 1.8 5.1 2.9 4.2 1.7 1.5 0.9 2.4 1.8 3.4 1.8 Bregmacerotidoe 6.6 3.1 1.3 0.7 2.6 1.4 3.4 1.3 3.1 1.7 2.4 1.8 3.2 1.7 Idiacanthidae 2.8 1.3 2.9 1.5 2.0 1.1 3.0 1.2 1.1 0.6 1.5 1.1 2.2 1.1 Melamphaidae 1.5 0.7 1.8 0.9 1.6 0.9 2.9 1.1 1.7 1.0 1.3 1.0 1.8 0.9 Scombridae 2.4 1.1 1.4 0.7 1.0 0.6 0.6 0.2 1.3 0.8 1.4 1.1 1.4 0.7 Other 18.1 8.5 10.8 5.4 13.5 7.6 20.8 8.2 14.5 8.2 11.4 8.5 14.8 7.7 Total 213.9 100.0 197.8 100.0 178.4 100.0 255.1 99.9 176.4 100.0 133.1 lOO.O 192.0 100.1 1 ETP I -stations 11.022-11.118 (35), a ETP II - stations 45.016-45.114 (41), 12.002-12.109 (50), 12.209-12.264 (24), and 13.187-13.265 (28), 45.191-45.365 (37), 46.002-46.069 (36), and 46.079-46.132 (27). in all cruises. The average abundance per haul ranged from 9.5 larvae (June-July) to 17.1 (April-May), a range of less than two times. Bathylagid smelts were represented in the monitoring pattern by a single species, Bathyl- agus nigrigenys Parr. Average abundance of larvae per haul ranged from 2.6 (December- January) to 6.4 (February-March) and aver- aged 4.0 larvae. Larvae of Bathylagidae usually ranked fifth in abundance. Paralepididae usually ranked fourth in rela- tive abundance; the lowest average abundance per haul was 3.2 larvae in October-November, and the highest was 8.2 larvae in August-Sep- tember. Nomeidae ranked variously fifth to eighth in relative abundance, with an overall ranking of sixth. The range in average abundance per haul was from 1.5 (October-November) to 5.1 larvae (June-July) and averaged 3.4 larvae. Bregmacerotidae showed the widest variation in abundance, 1.3 larvae (April-May) to 6.6 lar- vae (February-March); consequently they ranked variously between fourth and tenth in relative abundance. Larvae of the most com- mon species of Bregmaceros within the monitor- ing pattern, B. hathy master, tend to cluster with occasional samples having rather large numbers of larvae. Variability in sampling due to chance encounters of clusters of larvae could be of greater magnitude than that resulting from ac- tual changes in reproductive activity during the year. Idiacanthidae, usually ranked eighth in abundance per haul from 1.1 (October-Novem- ber) to 2.9 larvae (April-May), with an overall average of 2.2 larvae per haul. Melamphaidae ranked variously between sev- enth and tenth, with an overall rank of ninth. The lowest abundance, 1.3 larvae per haul in December-January, was less than half the high- est, 2.9 larvae in August-September. Scombridae in the monitoring area ranked either ninth or tenth in relative abundance of larvae on all cruises; the lowest average abun- dance, 0.6 larvae per haul in August-September, is only a fourth of the highest average value, 2.4 larvae in February-March. Larvae of these 10 families made up over 92% of the fish larvae in the monitoring pattern. In all instances, larvae of all principal families were taken throughout the year. The spread between the highest and lowest abundance val- ues for larvae of these principal families of fish- es was usually less than three times, and for Myctophidae and Sternoptychidae, was less than double. A similar seasonal pattern of abundance was observed for individual genera or species (Table 13). I found it helpful to arrange the 18 ge- 1168 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE 0) « ; o9~ PI ■- (U 2q E ® '. •- ui *- I- v.U Els'? to E * UJ o ■ — I- >^j; o IS. IS. O; CO o ^ CM d ■» q CO d CO -— '^iOu^"^pc0CN»O Csd^OJCNO'— CD »0 CO CO p CO ^ o a> CD cs CO CN ^ ^ d — d o' -■ CM -^ o -^ -^ -^ o S d ~ cc> (^^ p --^ w^ cq -^ ^ d — — — d d lO c^ d d 4 ^ IV. ■«<; p CN d d CM *0 Ov CO CM d '-^ d :; d "I O Q CO ^' — -: d CN -"t '— d d CO CNJ |Ss CO ^ d d CM — d '-^ CMlO^-^pNO^OCMCO cs o' — CO — ■ d — ' d — CO -q CO ^ d d '^ (U B e o p -q O; CO d — ^ pco •qi^iqrs.p — o.aj CO ^' o6 o' ^ eg CM ^' ^ d M3 iq Is. .q — d d ^ CC3 CO q p iq d —• CM CM d — d Is. rs. CO — d d ^ CO p ■<* CO '^ o. EE5"c-«£ ^ ^ --- a; U cj O '^ CQ Q ■-J O O _ , '"J -c.> *j H «. o ^ a: ^ E Q. f «• S:fe;c«bs -^o!!: < nera and species of this table according to the magnitude of the seasonal change in abundance that each displayed. Seasonal range in relative abundance between highest and lowest average number of larvae per haul: Less than 2X : Notolychnus valdiviae 2.1 to 3X : Bathylagus nigrigenys, Diplophos tae- nia, Vinciguerria lucetia, Idiacan- tlms sp., Bathophilus filifer, Dioge- nichthys laternatus, Hygophum atratiim, Hygophum proximum, Notoscopeliis resplendens, Tripho- turus sp. 3.1 to 4X : Cyclothone spp., Chauliodus sp., Go- nichthys tenuiculiis, Coryphaena spp., Howella pammelas 4.1 to 5X : Symbolophorus evermanni, Auxis spp. Larvae of the above 18 genera and species were taken on all cruises throughout the year. Obviously, reproduction is a continuous process for all of these, varying in amount at different seasons of the year and at different latitudes, but never stopping entirely. Larvae of two species dominated the collec- tions from the monitoring pattern: those of the myctophid Diogeuickthys laternatus and of the gonostomatid, Vinciguerria lucetia. Togeth- er these two species made up 44 to 56% of the total larvae in the monitoring pattern (Table 14). The highest average abundance of larvae of the lanternfish, Diogenichthys lateryiatus, 68.6 larvae in August-September, was 2.5 x as much as the lowest average abundance, 31.8 larvae in December-January. Larvae of this species ranked first in abundance between October and May, but were less abundant than larvae of Vinciguerria lucetia during June-September. Larvae of Vincigueriia lucetia had a range of 2.4 X between their highest average abundance per haul, 71.9 larvae in August-September, and lowest, 30.0 larvae in December-January. A small but consistent change in abundance with season is evident for this species, with the peak period in June-September, the minimal period in October-January, and intermediate abundance of larvae in February-May. The monitoring cruises were valuable in per- mitting us to establish the seasonal patterns of 1169 FISHERY BULLETIN: VOL. 70, NO. 4 Table 14. — Percentage contributions of larvae of the two most abundant species to the total catch of larvae in the monitoring pattern. Cruise Percentage contribution of larvae of designation ^'"'^ °^ ^"'"^^ Vincigu^rna Diogenichthy: Combined lucetia laternatus EASTROPAC 1 Feb. -Mar. 20.2 23.9 44.1 20.000 series Apr.-May 23.2 Z\Z 55.0 33.000 series June-July 35.0 21.0 56.0 EASTROPAC 11 Aug. -Sept. 28.6 26.9 55.1 50.000 series Oct. -Nov. 17.2 35.3 52.5 60.000 series Dec. -Jan. 22.5 23.9 46.4 Annual 24.6 27.2 51.8 reproduction in oceanic, tropical fishes. Except for this, little more was gained from the repeated coverages that was not evident from any one of the six coverages. The same species composition was observed in all six cruises, and even the rel- ative abundance of the various constituents did not change much. The similarity between cruis- es also extended to the geographic distributions of the various constituents which changed but little over time. COMPARISON WITH RV OCEANOGRAPHER ZIG-TRANSECT Although the average number of larvae per haul was almost identical for the Oceanographer collections and equivalent ETP II collections, 488.5 versus 487.8 larvae, and the kinds of larvae obtained were strikingly similar, the relative abundance of several categories was somewhat more variable than in the monitoring pattern (Table 15). Similarities and differences in relative abun- dance of larvae during the two coverages can be shown from a consideration of the 10 families with highest abundance in the Oceanographer collections (Table 16). Myctophfdae. — The difference in relative abundance of Myctophidae larvae between Oceanographer and ETP II collections, 194.1 versus 273.9 larvae per haul, is almost entirely due to difference in relative abundance of larvae of Diogenichthys laternatus. Over 509^ of D. laternatiis larvae on ETP II were taken in four contiguous stations between lat 5°40' and 7°44'N, with three collections exceeding 1,000 larvae and the largest with 2,505 larvae. In- terestingly, the two Oceanographer collections containing more than 1,000 D. laternatus larvae were taken between lat 6° and 7°N; these were the only two stations occupied between lat 5°40' and 7°44'N by Oceanographer in contrast to four collections from this rich area on ETP II. Gonostomatidae. — The difference in relative abundance of Gonostomatidae larvae in the two occupancies of the zig-transect was again due principally to a single species, Vinciguerria lucetia. Although twice as many larvae of this species were taken in Oceanographer collections, an examination of the station record revealed that one collection, OP.036 with 2,046 larvae, accounted completely for the difference. Clupeidae. — It is necessary to examine the species composition to evaluate differences be- tween the two coverages (Table 17). Larvae of the sardine, Sardinops sagax, were taken in six contiguous stations near the Galapagos on ETP II, and averaged 29 larvae per positive haul, whereas only 1 sardine larva was obtained from the same area by Oceanographer. This species appears to have a period of peak spawn- ing with reduced reproduction at other periods of the year. The contrast between the two col- lections of thread herring, Opisthonema sp., made at the station nearest the Mexican coast along long 92 °W is the largest observed in EASTROPAC collections— 2,730 larvae in the Oceanographer sample versus one larva in the ETP II collection. The larvae in the Ocean- ographer collection were intermediate-sized, 6 to 12.5 mm. Even allowing for the circumstance that clupeid larvae often occur in patches, this exceptionally large collection of larvae must have been obtained at a peak period of spawning of thread herring. Bathylagidae. — Larvae of the two species of bathylagid smelts that occur in the area covered by the zig-transect, were similar in distribution and relative abundance in the two coverages. Larvae of Bathylagus nigrigenys were taken in all but three collections on each coverage, and average abundance per haul was almost identi- cal, 26.3 versus 26.6 larvae (Table 17). Larvae of Leiiroglossus stilbius urotranus had a more restricted distribution on both coverages, occur- ring between about 7°S and the equator, at sta- 1170 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Table 15. — Comparison of composition of catches of fish larvae in Oceanographer zig- transect, occupied in November-December 1967, with equivalent coverage by EASTRO- PAC II vessels during August-September 1967. Oceanographer (50 stations) Equivalent ETP II coverage (48 stations) Family or larger grouping No. positive hauls No. larvae Average no. per haul No. positive hauls No. larvae Average no. per haul Clupeidae 2 2,737 54.7 7 185 3.9 Engraulidae 6 760 15.2 15 381 7.9 Argentinidae 0 0 0 2 2 <0.1 Bothylagidae 47 2,308 46.2 45 2,005 41.8 Gonostomatidae 47 4,386 87.7 45 2,386 49.7 Sternoptychidae 39 976 19.5 40 1,098 22.8 Chauliodontidae 16 47 0.9 9 19 0.4 Idiacanthidae 15 27 0.5 28 58 1.2 Other stomlatoidei 28 209 4.2 32 274 5.7 Myctophidae 50 9,706 194.1 47 13,149 273.9 Parolepididae 28 556 11.1 25 320 6.7 Scopelarchidae 15 138 2.8 9 36 0.8 Scopelorsauridae 11 14 0.3 19 32 0.7 Synodontidoe 1 3 0.1 3 7 0.1 Anguilliformes 10 18 0.4 16 23 0.5 Melamphaidae 43 243 4.9 40 274 5.7 Bregmacerotidao 1 470 9.4 9 1,455 30.3 Macrouridae 3 3 0.1 4 4 0.1 Exocoetidae 5 \6 0.3 6 \\ 0.2 Scomberesocidae 1 I <0.1 6 7 0.1 Trachypterldoe 7 11 0.2 8 M 0.2 Apogonidae 1 1 <0.1 5 14 0.3 Bramidae 13 22 0.4 7 10 0.2 Carangidae 2 354 7.1 5 56 1.2 Chiasmodontidae 27 65 1.3 13 38 0.8 Coryphaenidae 8 8 0.2 8 13 0.3 Gempylidae 7 21 0.4 9 30 0.6 Gobiidae 6 AO. 0.8 12 35 0.7 Labridae 3 3 0.1 3 3 0.1 Nomeidae 37 185 3.7 27 155 3.2 Ophidiidae 6 10 0.2 5 6 0.1 Sciaenidae 4 34 0.7 4 96 2.0 Scombridae 7 82 1.6 10 41 0.9 Scorpaenidae 9 14 0.3 11 86 1.8 Serranidae 2 28 0.6 6 9 0.2 Sphyraenidae 1 1 95 140 8,081 342 24,255 338 18,380 1177 FISHERY BULLETIN: VOL. 70, NO. 4 90° 80" Figure 5. — Distribution of larvae of the gonostomatid, Diplophos taenia (open circle with dot), of the stomiatoid family, Astronesthidae (open triangle with dot), and of the synodontid genus, Synodus spp. (open square with dot). Small solid circles represent other stations occupied on ETP 11. not replicated on ETP II. In these collections Cyclothone lan^ae occurred in 111 of 127 col- lections, with an average abundance per col- lection of 8.7 larvae. Danaphos oculatus (Garman) (1 occurrence, 1 larva) A single large larva was taken at the northern end of the Washington pattern at lat 19°16'N, long 118°56'W. Information obtained from Cal- ifornia Current and NORPAC collections indi- cates that Danaphos is a temperate water spe- cies, occurring most commonly in collections obtained from the central water mass of the North Pacific in hauls which sampled to depths greater than 140 m. Diplophos taenia (GUnther) (57 occurrences, 136 larvae) Larvae of Diplophos taenia afford a striking example of similarities in distribution, frequency of occurrences, and relative abundance in the 1178 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Figure 6. — Distribution of larvae of three kinds of gonostomatids. Records of occurrence of larvae of Gonostoma spp. shown as open square with dot, of Ichthyococcns irregularis as open circle with dot, and of Yarrella argenteola as open triangle with dot. Small solid circles represent other stations occupied on ETP II, two EASTROPAC multivessel cruises. Larvae were obtained in 57 collections from both ETP II and equivalent ETP I; on both surveys the majority of larvae were taken to the north of lat 10°N, particularly on the coastward-oriented portion of the station line terminating off Aca- pulco, Mexico, and that terminating off Man- zanillo, Mexico (Figure 5, and Ahlstrom, 1971, Figure 4). Larvae of this species were taken in moderate numbers, seldom more than 5 per haul; the average number per haul on ETP II was 0.38 larva versus 0.44 larva on equivalent ETP I. Gonostoma sp. (11 occurrences, 22 larvae) At least two kinds of gonostomatid larvae have been referred to Gonostoma, the more common being larvae of G. elongatum Gunther. The dis- tribution of Gonosto7na larvae on ETP II is shown in Figure 6; 8 of 11 occurrences were in a compact group in the southern, inshore por- tion of the ETP pattern (between lat 13° and 15°S, offshore to long 88°W) . 1179 FISHERY BULLETIN: VOL. 70, NO. 4 Figure 7. — Distribution of larvae of the gonostomatid, Vinciguerria spp. on ETP II. Collections of 1-100 larvae are shown as open circles with dot in center, collections of 101 or more larvae as large solid circles; negative hauls are shown as small solid circles. Ichthyococcus sp. (46 occurrences, 76 larvae) Maurolicus muelleri (Gmelin) (47 occurrences, 773 larvae) All Ichthyococcus larvae taken on ETP II were similar in appearance and have been re- ferred to /. irregularis Rechnitzer and Bohlke. Although widely distributed (Figure 6), all lar- vae were obtained between lat 12°N and 4°S; only three collections of Ichthyococcus larvae were taken in the outer pattern occupied by Washington. Larvae of M. muelleri ranked third in abun- dance among gonostomatid larvae. As on ETP I, (Ahlstrom, 1971, Figure 4) larvae of this spe- cies were sampled in a rather narrow equatorial belt, and none were taken seaward of long 112°W. This again is a striking instance of the similarity in distribution of larvae on the two multivessel cruises. Although the incidence of occurrences of Maurolicus larvae was almost as 1180 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE high in ETP I as in ETP II, 43 positive hauls as compared with 47, the average number of larvae per positive haul was much higher on ETP II— 16.4 larvae versus 6.1 larvae. Vinciguerria spp. (341 occurrences, 22,198 larvae) As in ETP I, larvae of Vinciguerria spp. ranked second in overall abundance, exceeded only by larvae of the myctophid, Diogenichthys laternatus (Garman). They were obtained throughout the EASTROPAC pattern, occur- ring in 96 Cf of the collections (Figure 7). Av- erage abundance of larvae per haul was about one-third greater than in ETP I: 62.5 versus 47.2 larvae. Larvae of two species of Vinciguerria occur within the ETP II pattern, although most were those of V. lucetia Garman. As commented upon for ETP I, larvae of V. nimbaria (Jordan and Williamson) were taken principally in the South Pacific central water mass, to the south of about lat 5°S. On ETP II this distribution involves about 20 collections only. Yarrella argenteola (Garman) (18 occurrences, 33 larvae) Larvae of Y. argenteola were taken in a lim- ited area shoreward or immediately south of the Galapagos Islands between lat 2°N and 5°S (Figure 6). No metamorphosing specimens were observed, although larvae as large as 16 mm were represented in the collections. As noted in the introductory section, only one specimen of Yarrella was obtained on ETP I, in contrast to the 18 occurrences on ETP II. Adults of this species were recorded from within the area cov- ered on ETP II by Morrow (1957b), Grey (1960), Bussing (1965), and Parin (1971). 7. STERNOPTYCHIDAE (277 occurrences, 7,385 larvae) As in ETP I, hatchetfish larvae ranked third in abundance. Although hatchetfish larvae con- tributed almost identical percentages of the total larvae in ETP II as in comparable ETP I (5.99% versus 5.989r), the average number of larvae per haul, 20.8 versus 13.9, reflected the greater relative abundance of larvae on ETP II. As noted for ETP I, hatchetfish larvae are more fragile than most kinds, and a portion of the larvae are too damaged to identify, except to family. Even so, identification to genus was made for most ETP II collections, and in these, larvae of Sternoptyx sp. contributed about 85% of the total and larvae of Argyropelecus (mostly A. lychnus Garman), the remainder. Baird (1971) in his revision of the family Sternop- tychidae recognized three species of Sternoptyx, with S. obscura Garman the common species in the eastern tropical Pacific; however, he includ- ed one record of S. diaphana Hermann from within the area surveyed on ETP II. 8. ASTRONESTHIDAE (42 occurrences, 74 larvae) Astronesthid larvae were taken in about four times as many collections as on equivalent ETP I. Most larvae had an equatorial distribution be- tween lat 8°N and 5°S; only two larvae occurred elsewhere (Figure 5). Three distinctive kinds of astronesthid larvae were taken. 9. CHAULIODONTIDAE (56 occurrences, 207 larvae) Although larvae of ChauUodus sp. were taken in a comparable number of hauls on ETP II and ETP I (56 versus 59 occurrences), more larvae were obtained on ETP II (207 versus 134 lar- vae). The majority of ChauUodus larvae on ETP II were taken in the inner half of the ETP pattern, below the equator — 34 collections con- taining 165 specimens were obtained from this quadrant (Figure 8) . In other parts of the ETP pattern somewhat fewer larvae were taken than on ETP I. As on ETP I, the majority of pos- itive hauls contained 1 to 3 larvae (41 of 56 hauls) ; even so, a higher proportion of the hauls on ETP II contained somewhat larger numbers of ChauUodus larvae, i.e., 6 to 26 lar- vae per haul. 1181 FISHERY BULLETIN: VOL. 70, NO. 4 Figure 8. — Distribution of larvae of the stomiatoid genus Chauliodus sp. (open square with dot, 1-10 larvae, closed square, 11 or more larvae), of the myctophid, Hygophum proximum (open circle with dot), and of the bothid flat- fish, Bothus leopardinus (open triangle with dot). Small solid circles represent other stations occupied on ETP II. 10. IDIACANTHIDAE (181 occurrences, 795 larvae) Larvae of Idiacanthus sp. were taken in over half of the plankton hauls made on ETP 11; there was an increase in frequency of occurrence of Idiacanthus larvae as compared to equivalent ETP I, but not in actual abundance of larvae. Larvae of Idiacanthus were most abundant in the inshore quadrant to the north of the equator and least abundant in the offshore quadrant south of the equator (Figure 9). All larger collections of larvae (11 to 43 larvae per haul) were taken to the north of the equator, usually within 600 miles of the coast. 11. OTHER STOMIATOIDEI (210 occurrences, 1,034 larvae) Included under other Stomiatoidei in Table 9 are larvae of two stomiatoid families: Stomia- tidae and Melanostomiatidae. In Appendix Table 1, the category "other Stomiatoidei" also includes the family Astronesthidae. In Appen- dix Table 3, counts are given for three principal 1182 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Figure 9. — Distribution of larvae of the stomiatoid genus Idiacavthufi sp. on ETP II. Collections of 1-10 larvae are show^n as open circles with dot in center, collections of 11 or more larvae as large solid circles; negative hauls are shown as small solid circles. constituents: Astronesthidae, Bathophiliis filifer (Garman), and Stomias sp. Stomias larvae (43 occurrences, 177 larvae) were most abundant in the inner pattern. Lar- vae of three categories of Melanostomiatidae were identified to the genus or species level. The most common of these were larvae of Bathoph- ihis filifer (Garman) (104 occurrences, 310 lar- vae). Larvae of Etistomias spp. (10 occurren- ces, 19 larvae) represented several species, whereas larvae of Leptostomias sp. (8 occur- rences, 17 larvae) were those of a single species. Approximately half of the stomiatoid larvae (140 occurrences, 511 larvae) were not identi- fied below the subordinal level. These were mostly small or damaged specimens; some of the unidentified stomiatoid larvae possibly are those of Malacosteidae. 13. EVERMANNELLIDAE (17 occurrences, 67 larvae) The majority of evermannellid larvae were 1183 130" 120° 110" I I I I • 1—1 — I — I — \ — I — I — I — r 100° FISHERY BULLETIN: VOL. 70, NO. 4 90° 80° T I I IS' I r~\ I I — I — \ — I — I — rn — i — i — r-i — i — i — r~i — i — i — i — i — i — i — i — i — i i i 20' 10" I0° — 1111 ©0©0_XMANZANILLO • Socorro I ^^N. 6? Clippcrton I -0- -4- Q A I ^ 20° r-irGolopogos 20' J L_l I I I I I L I I I L I I I J I I I I I i I L J I \ L I I I 130° 120° 110° 100° 90° 80° Figure 10. — Distribution of larvae of the myctophiform families Evermannellidae (open triangle with dot), and Scopelosauridae (open squares with dot, 1-25 larvae, closed squares, 26 or more larvae) and of the perciform family, Labridae (open circle with dot) ; negative hauls are shown as small solid circles. taken on the outer line of stations along long 119° W; the remainder were taken in an equa- torial band between lat 2°S and 4°N (Figure 10). This distribution is less widespread than that encountered on ETP I; however, 17 of the records of occurrence on ETP I were in the southern portion of the pattern not covered on ETP II. 14. MYCTOPHIDAE (352 occurrences, 64,009 larvae) Larvae of Myctophidae were more abundant on ETP II than on ETP I; the increase in abundance of myctophid larvae per haul in ETP II over ETP I was 1.63 x . Much of the increase was due to the greater abundance of larvae of the dominant species, Diogenichthys laternatus (Garman), although a number of kinds of myc- tophid larvae were taken in somewhat greater abundance, and only a few kinds were taken in lesser numbers per haul (Table 19) . To show changes in relative abundance of myctophid lar- vae between the two multivessel cruises, I have arranged the more common kinds in order of their relative abundance on ETP II as compared 1184 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Table 19. — Frequency of occurrence and relative abundance of the kinds of myctophid larvae on EASTROPAC II, and for equivalent coverage on EASTROPAC I. Washington Undaunted Rockaway EASTROPAC 11 Equivalent Myctophid genera or species 45.000 series 46,000 series 47.000 series tota. 1 EASTROPAC 1 total No. positive hauls No. larvae No. positive hauls No. larvae No. positive hauls No. larvae No. positive hauls No. larvae No. ^. * Benthosema fanamense 3 72 2 88 8 971 13 1,131 7 l/)27 Benthosema suborbitaU 1 1 1 1 0 0 2 2 7 7 Centrobranchus sp. 1 2 0 0 0 0 1 2 0 0 • Ceratoscopelus townsendi complex 12 365 0 0 1 24 13 389 37 349 * Diaphus spp. 73 938 53 1,113 51 392 177 2,433 168 1,931 * Diogenichthys atlanticus 1 1 0 0 3 9 4 10 6 7 * Diogenichthys laternatus 92 4,661 90 16,440 138 25,865 320 46,966 302 24,315 Diogenichthys sp. 0 0 0 0 2 4 2 4 0 0 * Gonichthys tenuiculus 15 25 27 99 64 169 106 293 88 226 * Hygophum atratum 38 335 10 A6 18 140 66 521 85 629 * Hygophum proximum 54 499 15 75 8 50 77 624 55 448 * Lampaiena spp. 7 8 3 6 0 0 10 14 15 27 Lampanyctus spp. 84 I,01<3 72 1,629 135 2,692 291 5,334 271 5^2612 * Lepidophane! pyrsobolus complex 16 53 12 73 8 12 36 138 13 41 * Lobianchia sp. 5 8 2 2 3 5 10 15 10 26 * Loweina laurae 10 15 10 14 5 8 25 37 31 41 * Myctophum aurolaternatum 37 85 41 144 70 445 148 674 145 S29 * Myctophum asperum 16 118 10 62 0 0 26 180 (M (1) * Myctophum nitidulum 25 300 43 274 66 717 134 1,291 V-) (1) * Myctophum other 11 27 6 13 0 0 17 40 117 l/MC * Notolychnus valdiviat 36 147 31 247 33 140 100 534 106 605 * Notoscopetus resplendens 14 28 29 198 35 156 78 382 54 231 * Protomyctophum sp. 5 7 12 22 8 15 25 44 33 74 * Symbolophorus evermanni 43 248 38 140 74 434 155 822 132 906 * Triphoturus spp. 23 40 27 1312 94 652 144 824 111 356 Unidentified myctophid larvae 33 86 33 94 50 217 116 397 1 15 295 Disirttegroted myctophid larvae 79 464 42 170 84 274 205 908 155 876 Total myctophid larvae \\\ 9,546 95 21,08(2 146 33,381 352 64,009 346 39,249 '^ Not 'Separately tabulated. to ETP I (comparable coverage, identical num- ber of samples). Genus or species of myctophid ^'^^ / Hygophum atratum,-reinhardti 0.83 Notolynchus valdiviae 0.88 Symbolophoms evermanni 0.91 Lampanyctus spp. 1.01 Benthosema panatnense 1.10 Ceratoscopelus iownsendt-complex 1.11 Diaphus spp. 1.26 Myctophum aurolaternatum 1.27 Gonichthys tenuiculus 1.30 Hygophum proximum 1.39 Myctophiim spp. (other than M. aurolaternatum) 1.45 Notoscopetus resplendens 1.66 Diogenichthys laternatus 1.93 Triphoturus spp. 2.31 No. in ETP I Benthosema panamense (Taning) (13 occurrences, 1,131 larvae) Although larvae of this species ranked fifth in abundance among myctophid larvae, they were collected in a relatively narrow coastal band, no wider than 200 miles (Figure 11). A similar pattern of inshore, clumped distribu- tion was encountered on ETP I (Moser and Ahl- strom, 1970, Figure 45). Benthosema suhorhitale (Gilbert) (2 occurrences, 2 larvae) Only two specimens of the larvae of Bentho- sema suhorhitale were taken on ETP II. Larvae of this species only recently have been positively identified. The larval series was initially estab- lished by Dr. H. G. Moser from Dana material. Larvae are strikingly similar to Electrona 1185 FISHERY BULLETIN: VOL. 70, NO. 3 90" 80° Figure 11. — Distribution of larvae of two species of myctophid lanternfishes. Records of occurrence of larvae of Bentliosema pananiense are shown as open triangles with dot for collections of 1-100 larvae, and as closed tri- angles for collections containing 101 or more larvae; records of occurrence of larvae of Myctophnm nitidulum are shown as open circles with dot for collections of 1-25 larvae, and as large solid circles for collections containing 26 or more larvae; negative hauls are shown as small solid circles. larvae, and earlier were confused with larvae of this genus. Most larvae included in Electrona sp. in the ETP I compilation were those of this species. The majority of occurrences of the larvae of this species on ETP I was in the southern, offshore portion of the ETP pattern, not covered on ETP II. Ceratoscopelus townsendi-complex (13 occurrences, 389 larvae) Abbreviated coverage of the southern portion of the EASTROPAC pattern, with coverage lim- ited to hit 10°S or 5°S on offshore lines, cut down markedly on the occurrences of larvae of Cera- toscopelus, as comi^ared with ETP I: 13 occur- rences as com]:)ared with 110. All occurrences but one of Ceratoscopelus larvae on ETP II were obtained in the outer pattern, occupied by Washingtov : 2 at the two northernmost stations along long 119°W, and 10 in the southern por- tion of the pattern between lat 6° and 10°S along long 119° and 112°W (Figure 12). Both clus- 1186 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE 130° 120' I lo- go Figure 12. — Distribution of larvae of two species of myctophid lanternfishes. Records of occurrence of larvae of Ceratoscopehis townsendi-complex are shown as open triangles with dot for collections of 1-25 larvae and as closed triangles for collections of 26 or more larvae; records of occurrence of larvae of Gonichthys tenuiculus are shown as open circles with dot; negative hauls are shown as small solid circles. ters of larvae occurred in the central water mas- ses of the North and South Pacific. Diaphus spp. (177 occurrences, 2,433 larvae) Larvae of Diaphus rank third in abundance among myctophid genera, exceeded only by Diogetikhthys and Lampanyctns. Although Diaphus larvae were taken in half the collections made on ETP II, occurrences and nonoccur- rences tended to be clustered. Almost two-thirds of Diaphus larvae were obtained to the north of lat 10°N on the four outer station lines; these were predominantly larvae of D. pacificus Parr. The largest area of nonoccurrence was off Peru, between lat 5° and 15°S; here Diaphus larvae were absent from 42 consecutive stations, 47.081 to 47.197. Larvae of the subgenus Diaphus, which are quite distinctive, made up about 10% of the total. 1187 FISHERY BULLETIN: VOL. 70, NO. 4 Juveniles and adult Diaphiis, separated from micronekton hauls made on ETP I, have been identified, with the cooperation of Robert Wis- ner of Scripps Institution of Oceanography: 15 species were represented in the collection made by Argo, David Starr Jordan, and Alaminos on ETP I. D. pacificus was, by far, the most abundant species, occurring in more collections and in larger numbers than other species of Diaphus. This species occurs in a broad coastal belt, 600 to 800 miles wide, from lat 20°N to the vicinity of the equator. Six species were taken offshore, between lat 5° and 20°S, in the South Pacific central water mass, including D. rolfbo- lini Wisner, D. brachycephalus Taning, D. fra- gilis Taning, D. jenseni Taning, D. schmidt Taning, and D. spleMdidiis (Brauer). Five spe- cies were taken in an oflfshore equatorial belt, between lat 10°N and 5°S, including D. garmani Gilbert, D. malayanus Weber, D. termophilus Taning, D. lucidus Goode and Bean, and D. lut- keni Brauer, the latter showing some admixture with central water mass species. Species be- longing to subgenus Diaphus, tentatively iden- tified by Wisner as D. longleyi Fowler and D. mollis-nanus complex had quite widespread dis- tributions. Now that the species composition of adult Diaphus has been clarified, life history series can be determined for the more common kinds. Diogenichthys latematus (Garman) (320 occurrences, 46,966 larvae) Larvae of D. latematus were outstandingly abundant, making up 38.1 9f of the total fish larvae obtained on ETP II. Almost twice as many D. latematus larvae were taken in equiv- alent coverage of the EASTROPAC region on ETP II as on ETP I; 46,966 versus 24,315 lar- vae. The number of collections that contained D. laternatus larvae, however, was not much different: 302 of 355 in ETP I as compared with 320 of 355 in ETP II. Almost one collection in three from ETP II contained over 100 D. la- tematus larvae, and 19 collections contained over 500 larvae. Of these larger collections, 13 of 19 were taken between lat 5° and 10 °N. As on ETP I, larvae of D. latematus were not taken in collections made within the central water mass of the South Pacific (Figure 13). Diogenichthys atlanticus (Taning) (4 occurrences, 10 larvae) Larvae of this species were taken more fre- quently on ETP I (29 occurrences, 92 larvae); however, all but six of these occurrences were in the portion of the ETP I pattern that was not covered on ETP II. The four records on ETP II were taken between lat 9° and 15°S, with two occurrences in the transitional waters of the Humboldt Current and only one occurrence off- shore in the central water mass. Larvae of this species were commonly taken on MARCHILE VI off Chile (12 occurrences, 100 + larvae). Gonichthys tenuiculus (Garman) (106 occurrences, 293 larvae) Larvae of Gonichthys had rather similar dis- tributions and frequency of occurrences in the two multivessel EASTROPAC surveys. The ma- jority of larvae were obtained in the inner pat- tern occupied by Rockaway, with highest fre- quency of occurrences in an equatorial belt between lat 5°N and 5°S (Figure 12). Hygophum atratum (Garman) {66 occurrences, 521 larvae) The less extensive coverage on ETP II elim- inated the area in which H. reinhardti (Liitken) larvae were taken on ETP I, and only larvae of H. atratum were observed in ETP II collections. Larvae of H. atratum were spottily distributed, occurring mostly in three clusters of stations: 1) between lat 15° and 20 °N in the Washington pattern, 2) between lat 10° to 15°S in the Rock- away pattern, and 3) an equatorial band between lat 5°N and 5°S along long 119°, 112°, and 105°W. 1188 AHLSTROM: KIND AND ABLNDANCE OF FISH LARVAE Figure 13. — Distribution of larvae of the myctophid Diogenichthys Intcrnntus on ETP II. Three orders of abun- dance are shown. Open circles with dot represent counts of 1-100 larvae, large solid circles represent counts of 101-500 larvae, and large solid circles with bisecting line represent counts of 501 or more larvae; negative hauls are shown as small solid circles. Hygophum proximum (Becker) (77 occurrences, 624 larvae) The distribution of larvae of H. proximum again is illustrated (Figure 8) to show the marked similarity in distribution to ETP I ( Ahl- strom, 1971, Figure 10). Larvae of this species were decidedly more abundant in the offshore pattern occupied by Washington (55 occurren- ces, 499 larvae). As noted earlier, larvae of H. jiroximimi were taken in somewhat greater abundance in ETP II as compared to equivalent ETP I (1.39 X ) . Fully half of the occurrences and specimens of H. proximum larvae on ETP I was in the unreplicated portion of ETP I cov- erage, i.e., on the offshore line of stations along long 126°W and in the offshore southern por- tion of the pattern. There were three occur- rences of larvae on ETP II in the southern part of the Rockaivny pattern in transitional waters of the Humboldt Current; larvae were not ob- tained from this area on ETP I. 1189 FISHERY BULLETIN: VOL. 70. NO. 4 Lampadena sp. (10 occurrences, 14 larvae) Larvae of Lamapadena sp. were taken on the three offshore lines in two groups — one occur- ring between lat 3° and 8°N and the other in the central water mass of the South Pacific be- tween lat 7° and 10°S. A similar distributional pattern was obtained on ETP I ; however, the more extensive coverage of the South Pacific central water mass on the earlier survey pro- vided better distributional information for the southern component. Lampanyctus spp. (291 occurrences, 5,334 larvae) Lampanyctus larvae rank second in abundance and in frequency of occurrence among the myc- tophid genera represented in the eastern trop- ical Pacific. Lampanyctus larvae were most abundant between lat 5°N and 5°S and least common between lat 10° and 20°N. The six col- lections of Lampanyctus larvae that contained over 100 specimens per collection were taken be- tween the equator and lat 5°N. Three kinds of Lampanyctus larvae dominated over most of the EASTROPAC pattern. Although identification to the species level are tentative as yet, these three kinds of larvae are almost certainly those of L. idostigma Parr, L. omostigma Gilbert, and L. parvicauda (Parr) — three widespread trop- ical species of Lampanyctus. A quite different assemblage of Lampanyctus larvae was taken in the moderate number of stations occupied in the South Pacific central water mass. with common characteristics attributed to L. pyrsoholus. These workers considered Alcock's poorly described L. pyrsoholus as unidentifiable. Instead they identified their material with L. photothorax (Parr), L. longipes (Brauer), and L. indicus Nafpaktitis and Nafpaktitis. L. pho- tothorax was taken in four ETP I collections between lat 15° and 20°S in the offshore pattern occupied by Argo. The specimens from the eastern Pacific agree closely with the description and illustration of this species in Nafpaktitis and Nafpaktitis (1969). These workers gave 7 + 4 as the usual combination of AO photo- phores on specimens from Indian Ocean ma- terial. In the EASTROPAC area all specimens examined had 6 + 4 AO photophores. The widely distributed species in the EAS- TROPAC area is either L. longipes (Brauer) or a species closely related to L. longipes. The eastern Pacific form has similar luminous patch- es to those described for L. longipes from the Indian Ocean except for the luminous tissue on the head of males and the size of the infracaudal gland on some larger specimens. Luminous patches developed on the head were restricted to a single wide pair of luminous patches. On some larger specimens the infracaudal gland began under the last AO photophore and was conspic- uously larger than those observed by Nafpaktitis and Nafpaktitis (1969) on Indian Ocean mate- rial. AO photophores were usually 5 + 4; gill raker counts were 5 + 1 + 11 to 13. Two kinds of Lepidophanes have been ob- served in the EASTROPAC area, although only one kind was taken commonly. Larvae of the latter have been assigned to L. longipes (?). Lepidophanes pyrsoholus complex (36 occurrences, 138 larvae) Lobianchia spp" (10 occurrences, 15 larvae) An examination of the juvenile and adult spe- cimens of Lepidophanes collected on ETP I has shown that two closely related species are pre- sent— one with a very restricted distribution and the other with a widespread distribution. Naf- paktitis and Nafpaktitis (1969) found three species of Lepidophanes from the Indian Ocean Larvae of Lobiauchia, although uncommon in the eastern tropical Pacific, have a fairly wide- spread distribution in two separated areas: 1) in an equatorial belt between lat 3°S and 6°N (8 occurrences) and 2) in the transitional wa- ters of the Humboldt Current. In the latter area, two occurrences were recorded at about lat 12° 1190 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE to 13°S along long 88°W, and three additional records were obtained at MARCHILE VI sta- tions (not included in above totals). At least two species, L. gemellari (Cocco) and L. dumer- ili (Bleeker), and perhaps a third, are involved. Loweina laurae (Wisner) (25 occurrences, 37 larvae) Wisner (1971) has separated the eastern Pa- cific species of Loweina from L. vara (Liitken). Although the two species are basically quite sim- ilar, Wisner points out that L. laurae has a somewhat longer head, 27.3 to 30.7% of SL versus about 25.7%, and a somewhat larger eye, averaging about 8% of SL versus about 6%. Wisner gave the distribution of L. laurae in the eastern Pacific as between lat 30 °N and 30 °S and westerly to long 150°W. Of the 25 occurrences of larvae of L. laurae on ETP II, all but one occurred in a broad equa- torial band between lat 7°N and 6°S (Figure 14). The isolated record was on the southern- most line of stations oriented normal to the coast occupied by Rockaway. This distribution is sim- ilar to that illustrated for ETP I (Moser and Ahlstrom, 1970, Figure 51). In equivalent cov- erage on ETP I, 31 stations yielded 41 larvae. It should be noted that larvae of Loweina from EASTROPAC appear to be identical with those identified as L. rara from other oceans; hence larval evidence does not support the separation of the eastern Pacific form as a separate spe- cies. Myctophum spp. (217 occurrences, 2,185 larvae) Larvae of the genus Myctophtim ranked fourth in abundance and third in frequency of occurrence. Larvae of M. aurolaternatum Gar- man (148 occurrences, 674 larvae) were taken more frequently but in lesser amounts than lar- vae of M. nitidulum-complex (134 occurrences, 1,291 larvae). Larvae of M. aurolaternatum were taken in all parts of the EASTROPAC pat- tern, but in largest numbers between the equator and lat 5°N. Most larvae of M. nitidulum-com- plex were taken in a broad equatorial band be- tween lat 8°N and 5°S (Figure 11). The dis- tribution, however, had a southerly extension to the bottom of the pattern in the area of the Humboldt Current. Larvae of M. asperum Richardson (26 occurrences, 180 larvae) were taken in an offshore equatorial tongue, extending seaward from long 98 °W to its widest extent (lat 2°S to 7°N) along long 119°W (Figure 3). The remainder of Myctophum larvae (17 occur- rences, 40 larvae) belong to two and possibly three species. One group of these occurred in the offshore equatorial tongue, along with larvae of M. asperum; the other group occurred be- tween lat 7° and 10 °S in the offshore Washington pattern. The latter group includes larvae of both M. lychyiohium Bolin and M. brachygnathos (Bleeker). Only larvae of M. aurolaternatum were sep- arately tabulated for equivalent ETP I coverage (145 occurrences, 529 larvae). Both the dis- tribution of M. aurolaternatum larvae and their frequency of occurrence were similar for the two multivessel surveys, although abundance was moderately greater on ETP II, 1.9 versus 1.5 larvae per haul. This pattern of greater abun- dance on ETP II also held for the remainder of the larvae of Myctophum, 4.3 versus 2.9 larvae per haul. Notolychnus valdivtae (Brauer) (100 occurrences, 534 larvae) Larvae of the wide-ranging oceanic species are seldom taken closer to shore than 200 miles. On ETP II, the majority of records were from an equatorial tongue that extended between lat 10°S and 10°N in the oflFshore Washington pat- tern, but shoreward of this (long 105° to 85°W) the distribution narrowed to between lat 2°S and 8°N, with the majority of occurrences between lat 2° and 6°N. A second group of larvae were sampled in the southern portion of the Rockaway pattern between lat 9° and 15°S. Only two oc- currences of Notolychnus larvae were noted in 85 stations occupied by all vessels between lat 20° and 10°N. Distribution of Notolychnus 1191 FISHERY BULLETIN: VOL. 70, NO. 4 Figure 14. — Distribution of larvae of the myctophid, Loweuut laurae (open circle with dot), of the scomberesocids, Scomberesox saurus (open diamond with dot), and Cololabis adocctus (open hexagon with dot) and of the anguil- liform families Congridae (open triangle with dot) and Nemichthyidae (open square with dot) ; negative hauls are shown as small solid circles. larvae was illustrated for ETP I coverag'e (Ahlstrom, 1971, Figure 11). In the portion of ETP I pattern also covered on ETP II, frequency of occurrence and distribution of Notolychnus larvae were quite similar: 1.7 versus 1.5 larvae. Notoscopelus resplendens (Richardson) (78 occurrences, 382 larvae) As on ETP I, most larvae of N . resplendefis were taken in an equatorial belt, between lat 5°N and 5°S (65 occurrences, 364 larvae). A second center of occurrence was at the southern i)ortion of the Rockaicay pattern between lat 9° and 15°S. Except that the distribution of the main group of Notoscopelus larvae is more definitely centered on the equator, the distribution of lar- vae of Notoscopelus and Notolych)ii(s are quite similar. No larvae of Notoscopelus were taken north of lat 6°N. Moderately more larvae of Notoscopelus were taken on ETP II, 1.1 versus 0.7 larvae per haul. 1192 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Protomyctophum sp. (25 occurrences, 44 larvae) For most kinds of myctophids, the distribu- tional patterns of larvae are so similar in the two multivessel EASTROPAC survey cruises that distributional information from ETP II merely reinforced that obtained on ETP I. Dis- tribution of Protomyctophrim larvae affords an- other example of this. All but two of the occur- rences lie between lat 10°N and 5°S, the zone in which all Protomyctophum larvae were ob- tained on ETP I. As noted in Ahlstrom (1971) , the larvae were all of a kind, belonging to a per- haps undescribed species of Protomyctophum, subgenus Hierops. Wisner (1971) described two new species of Protomyctophum, subgenus Hierops from the eastern Pacific: P. chilensis from off Chile about lat 33 °S and P. beckeri from the vicinity of the Hawaiian Islands. It is not known as yet whether the form from EASTRO- PAC is referable to either of these. Symbolophorus evermanni (Gilbert) (155 occurrences, 822 larvae) Larvae of Symbolophorus were absent from a wide coastal strip off Mexico and a narrower coastal strip off Peru, but were taken at most stations in the remainder of the ETP II pattern. The distribution was rather similar to that il- lustrated for ETP I (Ahlstrom, 1971, Figure 12); the frequency of occurrence was slightly lower on equivalent ETP I (37% positive hauls versus 44%), but the average abundance per haul was slightly higher (2.6 versus 2.3 larvae). However, in the ETP I stations without counter- parts in ETP II, frequency of occurrence was higher than in the remainder of the ETP I pat- tern (63% versus 37%) and average abundance per haul was higher (4.5 versus 2.6 larvae). Triphoturus spp. (144 occurrences, 824 larvae) Larvae of Triphoturus oculeus (Carman) were taken in most hauls made between lat 5°N and 15 °S off Ecuador and Peru and offshore to the vicinity of the Galapagos Islands. Larvae of this species, which appear to be more exclu- sively restricted to the transition waters of the Humboldt Current than are those of other myc- tophids sampled in the EASTROPAC pattern, also may exhibit the most marked seasonal change in relative abundance. Other Tripho- turus larvae, sampled mostly offshore, were tak- en in slightly lesser abundance than on ETP I. 16. PARALEPIDIDAE (247 occurrences, 2,535 larvae) Larvae of Paralepididae ranked sixth in abundance and contributed over 2% of the total. Larvae were taken throughout the ETP II pat- tern, but most commonly in an equatorial band between lat 5°N and 5°S; all collections of lar- vae exceeding 25 larvae per haul were obtained from this band. Fewest larvae were taken in the southern portion of the inner pattern, below about lat 7°S. Because of limited coverage of the South Pacific central water mass on ETP II, no material was obtained of Sudix atrox Rofen (see Ahlstrom, 1971, Figure 7 for distribution of larvae of this species on ETP I) . A detailed study of the species composition of the paralep- idid material from EASTROPAC surveys has not been made. 17. SCOPELARCHIDAE (134 occurrences, 298 larvae) Larvae of Scopelarchidae were taken through- out the area surveyed on ETP II. As noted for ETP I (Ahlstrom, 1971, p. 32-33), larvae of five or six kinds of scopelarchids were obtained, usually in small numbers per haul. On ETP II, only 6 of 133 positive hauls contained over 5 larvae (6 to 12 larvae), and over 80% of the hauls contained 1 to 3 larvae per haul. 18. SCOPELOSAURIDAE (40 occurrences, 390 larvae) Larvae of Scopelosauridae were taken in more 1193 FISHERY BULLETIN: VOL. 70, NO. 4 hauls and in much larger numbers than on equiv- alent ETP I (6 occurrences, 13 larvae). As shown in Figure 10, most occurrences were in an equatorial band between lat 5°N and 5°S and offshore to long 105°W; the five hauls containing 25 or more larvae were obtained within 2° of the equator. Only one kind of Scojjelosaurus larva was obtained on ETP II. Larvae of Sco- pelosauriis superficially resemble paralepidid larvae — both have elongate larvae with a short gut that increases in relative length in older lar- vae. However, Scopelosaiirus larvae differ in several significant ways from paralepidid larvae, Scopelosaurus larvae never develop patches of pigment above the intestinal tract, whereas these patches are a striking feature of paralepidid lar- vae; the eyes of Scopelosaurus larvae are nar- rowed, whereas they are round in most paral- epidid larvae; also the intestinal tract does not increase in relative length nearly as much in old- er stage Scopelosaurus larvae as in paralepidid larvae. 19. SYNODONTIDAE (14 occurrences, 60 larvae) Larvae of Synodus spp. occurred in a coastal band along the extent of the ETP II pattern (Figure 5). Six species of Synodus are known to occur in the eastern Pacific. Several kinds of Synodus larvae were taken in the EASTRO- PAC collections, mostly small specimens. Until more older-stage larvae are obtained, it will not be possible to work out life history series. 21. ANGUILLIFORMES (EEL LEPTOCEPHALI) (81 occurrences, 151 larvae) Eel leptocephali, although conspicuous mem- bers of the larval fish fauna, are not common in the EASTROPAC pattern: they contributed only 0.12% of the total ETP II larvae. Lepto- cephali of seven families of true eels of the order Anguilliformes, suborder Anguilloidei, were identified from the micronekton net collections of ETP II. The micronekton net collections from ETP I contributed three times as many lepto- cephali as the regular net hauls; a total of 10 families was represented in the combined ETP I collections, including the 7 discussed below and in addition Derichthyidae, Muraenesocidae, and Nettastomidae. The record of occurrence and counts by family of eel leptocephali on all pos itive stations is contained in Appendix Table 5, and summarized in Table 20. The distributions of larvae of the seven families taken in ETP II collections are shown in Figures 14 and 15. Congridae (28 occurrences, 42 larvae) This family ranked first in frequency of oc- currence among eel leptocephali and second in relative abundance. Most congrid larvae were identifiable to genus. The breakdown was as follows: ylWosoma sp. (5 occurrences, 8 larvae), Bathyconger sp. (3 occurrences, 4 larvae), Gnathopis sp. (1 occurrence, 1 larva), Hilde- brandia (10 occurrences, 18 larvae) , Paraconger (4 occurrences, 5 larvae), and genus uncertain (6 occurrences, 6 larvae). All but two occur- rences were from north of the equator, and most specimens were taken in a broad coastal band. However, offshore oceanic occurrences of cong- rid leptocephali were more frequent on ETP I than on ETP II. Moringuidae (3 occurrences, 3 larvae) One occurrence of leptocephali of the morin- guid genus Neoconger was off Manzanillo, Mex- ico, the other two near Panama Bay. Muraenidae (5 occurrences, 6 larvae) Although adults of Muraenidae are known to have a wide distribution in the eastern Pacific, the few leptocephali taken on ETP II were con- fined to a narrow tongue extending offshore be- tween lat 7° and 10°N in the northeast quad- rant. !l 1194 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE 130" 120" 110" 20 00° 90° 80° Figure 15. — Distribution of eel leptocephali of the anguilliform families: Moringuidae (open diamond with dot), Muraenidae (open triangle with dot), Opichthidae (open circle with dot), Serrivomeridae (open square with dot), and Xenocongridae (open hexagon with dot) ; negative hauls are shown as small solid circles. Table 20. — Familial composition of eel leptocephali taken on the second multivessel EASTROPAC survey, summarized by vessel pattern. Washington Undaunted Rock away Total 45.000 series 46.000 series 47.000 series EASTROPAC 11 Family No. positive hauls No. larvae No. positive hauls No. larvae No. positive hauls No. larvae No. positive hauls No. arvae Congridae 3 6 9 11 16 25 28 42 Moringuidae 1 1 0 0 2 2 3 3 Muraenidae 0 0 3 3 2 3 5 6 Nemichthyidae 4 5 9 9 6 7 19 21 Ophichthidae 2 8 7 12 17 29 26 49 Serrivomeridae 6 8 0 0 0 0 6 8 Xenocongridae 0 0 2 2 2 2 4 4 Family unknown 4 5 4 5 5 8 13 18 Total 16 33 30 42 35 76 81 151 1195 FISHERY BULLETIN: VOL. 70, NO. 4 Nemichthyidae (19 occurrences, 21 larvae) Although eels of this family are widely distrib- uted in offshore oceanic waters, most occurren- ces of leptocephali (14 of 19) were in the north- east quadrant, between lat 0° and 10°N. Ophichthidae (26 occurrences, 49 larvae) Ophichthid leptocephali were taken in a broad coastal band between Manzanillo, Mexico, and Central Peru (lat 10°S). They ranked first in relative abundance among eel leptocephali and second in frequency of occurrence. Serrivomeridae (6 occurrences, 8 larvae) Most occurrences of serrivomerid leptocephali (5 of 6) were on the outer line of the ETP II pattern, along long 119°W, and the remaining occurrence was along long 112°W. In contrast to nemichthyid leptocephali which may grow to 300 or 400 mm long, leptocephali of Serrivomer- idae rarely exceed about 60 mm. Xenocongridae (4 occurrences, 4 larvae) The few occurrences of leptocephali of Chlop- sis, the sole representative of this family, were within 4° of the equator. 22. MELAMPHAIDAE (284 occurrences, 1,365 larvae) Larvae of Melamphaidae ranked fourth in frequency of occurrence, eighth in relative abundance. Larvae were distributed through- out the ETP II pattern (Figure 16) and oc- curred in 80% of the collections. Most collec- tions contained only moderate numbers of lar- vae— the average number of larvae per positive haul was only 4 to 8. The majority of hauls containing larger numbers of larvae (11 or more per haul) were taken within 5° of the equator (Figure 16). Melamphaid larvae were repre- sented by four genera: Melamphaes, Scopelog- adus, Scopeloberyx, and Poromitra. 23. TRACHICHTHYIDAE (11 occurrences, 70 larvae) The big-headed larvae of a representative of this family were taken at 11 stations on the two inner lines of the Rockaway pattern, between about lat 2° to 8°S (Appendix Table 3). They appear to be larvae of Trachichthys mento Gar- man, initially described from the Gulf of Pan- ama. Bussing (1965) supplemented Garman's description, utilizing 53 specimens (55 to 104 mm) collected at Eltaniyi Station 34 at lat 07°45' to 07°48'S, long 81°23'W. Parin (1971) also obtained material of this species in the eastern tropical Pacific from oflf South America. 25. BREGMACEROTIDAE (160 occurrences, 3,062 larvae) Larvae of Bregmacerotidae ranked fifth in abundance and contributed 2.5% of fish larvae on ETP II. The majority of larvae was taken to the north of the equator, with three inshore collections contributing over 70% of the total. These collections of 927, 753, and 511 larvae were exclusively Bregmaceros bathymaster Jordan. Larvae of this species were distributed in a broad coastal band in the northern half of the EAS- TROPAC pattern. As noted in the ETP I re- port, larvae of five species of Bregmaceros are distributed in the eastern tropical Pacific. 27. SCOMBERESOCIDAE (27 occurrences, 153 specimens) Two species of Scomberescocidae were taken on ETP II — Scomberesox saurus L. (18 occur- rences, 52 specimens) and Cololabis adocetus Bohlke (9 occurrences, 101 specimens). The 1196 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE 130' 120° 110° 20 Figure 16. — Distribution of larvae of the beryciform family Melamphaidae on ETP II. Collections of 1-10 larvae are shown as open circles with dot, collections of 11 or more larvae as large solid circles; negative hauls are shown as small solid circles. word "specimen" is used intentionally because some juveniles as well as larvae are included in the above counts. A number of the specimens were x-rayed in order to obtain vertebral counts to verify identification. All occurrences of the small tropical saury, Cololabis adocetus, were along long 95°W at nine contiguous stations (Figure 14) ; surface temperatures ranged be- tween 19.5° and 21.5°C at these stations. Scom- beresox larvae occurred in a broad coastal belt, shoreward of C. adocetus, extending from near the equator to the southernmost line occupied on ETP II (Figure 14); surface temperatures ranged between 15.8° and 19.5°C at these sta- tions. Actually Scomberesox eggs and larvae were commonly taken in the pattern occupied by YelcJw off Chile as part of ETP II— MAR- CHILE VI. Collections obtained from surface tows as well as from oblique net hauls were available from MARCHILE VI. Five short lines of stations normal to the trend of the Chilean coastline were occupied on MARCHILE VI, be- tween lat 18°30' and 33°S. Scomberesox eggs and larvae were sampled best in surface hauls. Scomberesox eggs were taken in 17 of 20 surface hauls and Scomberesox larvae in 10 surface 1197 FISHERY BULLETIN: VOL. 70, NO. 4 Table 21. — Measurements of eggs of Scomberesox saurus collected on EASTROPAC II, including collections made off Chile by Yelcho (MARCHILE VI). Collection Type of haul Locality of collection Number Range in egg diameter (mm) Average diameter (mm) Surf wot temper ace er Lot S Long W eggs measured atura MAR. 5.4 Surface 33°05.3' 73''20.5 25 2.41-2.67 2.52 12.50 MAR. 4.4 Oblique 28''30.6' 72''43.2 23 2.39-2.65 2.52 12.09 ETP 47.177 Oblique 06°35.0' 85°08.5 31 2.31-2.60 2.44 18.14 MAR. 4.1 Surface 28°30.2' 7r40.1 30 2.26-2.62 2.43 11.93 MAR. 3.2 Surface 23">42.5' 7I°35.0 25 2.24-2.45 2.36 14.39 ETP 47.145 Oblique 14° 17.8' 83 ""03.7 14 2.26-2.43 2.35 18.28 MAR. 1.8 Surface 18°27.6' 73 "06.1 25 2.24-2.51 2.34 15.81 (10 m) ETP 47.134 Oblique 12°56.5' 79°27.8 16 2.21-2.48 2.34 16.62 ETP 47.103 Oblique 10°09.0' 82°D8.5 25 2.26-2.45 2.34 18.32 ETP 47.107 Oblique 09°50.0' 80°53.0 16 2.26-2.46 2.34 17.74 MAR. 2.1 Surface 20°09.0' 70''31.8 25 2.15-2.45 2.33 15.77 (10 m) MAR. 1.4 Surface 18°32.0' 7r42.0 25 2.17-2.45 2.32 15.92 (10 m) MAR. 2.4 Surface 20° 10.8' 71°33.2 25 2.19-2.45 2.32 15.74 (10 m) hauls on MARCHILE VI. Hence young of Scomberesox have a north-south extent off South America of at least 1,860 miles. Scomberesox eggs are approximately round and occur singly — lacking the attachment fila- ments characteristic of most eggs of fishes in the suborder Exocoetoidei (see in this regard Orton, 1964). The egg .shell, however, is ornamented with minute closely spaced swellings. Eggs from 13 collections were measured (eggs measured in widest dimensions as they were not truly spheri- cal); the data are summarized in Table 21. The range in egg size was from 2.15 to 2.67 mm; the range in egg diameter means for the 13 col- lections was from 2.32 to 2.52 mm. Eggs in the majority of collections (9 of 13) were quite sim- ilar in average diameters, ranging between 2.32 and 2.36 mm. Three of the four collections of eggs with larger average diameters were taken on the southernmost two lines of the Yelcho pat- tern. However, the collection of eggs made nearest to the equator (lat 6°35') also was in this group of larger eggs. Howella pammelas (Heller and Snodgrass). Larvae of this species were most common to the north of the equator in a broad band ex- tending oflFshore between 0° and lat 9°N. Only three occurrences were found to the north of this band and 11 to the south. This species was not limited in its distribution to particular water masses. 34. CARANGIDAE (36 occurrences, 224 larvae) Larvae of the pilotfish, Nauc rates diictor (L.), with 18 occurrences, 27 larvae (Figure 17), was the most widely distributed carangid on ETP II. Over half of the carangid larvae were obtained at two coastal stations — 45 larvae at 46.135 and 69 larvae at 47.527. As on ETP I, a number of kinds of carangid larvae were taken, including Chloroscombrus orqueta Jordan and Gilbert, Selene brevoorti (Gill), and Caranx spp. 30. APOGONIDAE (66 occurrences, 283 larvae) This family contains both oceanic and coastal species. Larvae of coastal apogonids were taken in four hauls off Central America and northern South America. The remainder of the larvae (62 occurrences, 278 larvae) were those of 37. CORYPHAENIDAE (109 occurrences, 185 larvae) Larvae of the dolphin, Coryphaeva spp., were taken almost exclusively to the north of the equator (105 occurrences, 180 larvae) on ETP II; three of the four occurrences to the south of the equator were at stations immediately ad- jacent to the equator. Coryphaena larvae were 1198 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Figure 17. — Distribution of larvae of the carangid, Naucrates ductor (open circle with dot), of the tetragonurid, Tetragonuriis spp. (open square with dot), and of the trichiurids, Lepidopus sp. (open diamond with dot) and Trichiurus lepturus (open triangle with dot) ; negative hauls are shovNTi as small solid circles. taken throughout the coverage on ETP I but at a lesser proportion of the stations; in equivalent coverage, 97 larvae were taken in 67 collections on ETP I. Dolphin larvae provide one of the more striking examples of a marked difference in the distributional pattern of larvae as be- tween ETP I and ETP II. Eggs and newly hatched larvae have been de- scribed for Coryphaena hipjnirus L. by Mito (1960). The eggs are 1.2 to 1.6 mm, with a single oil globule 0.3 to 0.4 mm. The larvae are heavily pigmented, even at hatching. We have not found distinguishing characters to separate the larvae of C. hippurus from those of C. equi- selis, hence have labelled our material as Cory- phaena spp. After the vertebral column is devel- oped, definitive identification can be made: C. hippurus has 31 vertebrae, C. equiselis has 33 (Collette, Gibbs, and Clipper, 1969). Parin (1968) reports that C. hippurus reproduces only in the littoral zone and that C. equiselis is the offshore spawner. If this pattern of spawning holds for the eastern Pacific, then the majority of larvae taken on EASTROPAC cruises were those of C. equiselis. As noted for this family in the first EASTROPAC paper (ETP I, 38) the 1199 FISHERY BULLETIN: VOL. 70, NO. 4 majority of the specimens obtained were early- stage larvae, hence spawned in the area of col- lection. 38. GEMPYLIDAE (112 occurrences, 370 larvae) Two kinds of gempylid larvae were widely dis- tributed in the EASTROPAC area on ETP II: larvae of Gempylus serpens Cuvier and Valen- ciennes (71 occurrences, 152 larvae) and Nea- lotus tripes Johnson (66 occurrences, 218 lar- vae). Larvae of both species had a higher frequency of occurrence and greater abundance on ETP II. Distribution within the EASTROPAC area also was different in the two multivessel surveys. The more widespread distributional pattern for Gempylus serpens was observed from the wider- ranging ETP I survey. Over a third of the oc- currences of Gempylus larvae were in the por- tion of the ETP I pattern not replicated on ETP II (Ahlstrom, 1971, Figure 13). Larvae were taken throughout the ETP I pattern with as many records from south of the equator as to the north. In contrast, only three collections were made to the south of the equator on ETP II (Figure 4), but many more collections of Gem- pylus larvae were obtained to the north of the equator, particularly in the inner pattern occu- pied by Rockaway. Changes in distribution of larvae of Nealohis tripes in the two surveys were not as marked as for G. serpens. On both surveys the majority of the occurrences of Nealotus larvae were in the inner half of the ETP pattern; heaviest concentration of larvae on ETP II was in an equatorial band between circa lat 5°N and 3°S. Fewer Nealotus larvae were taken in the inner pattern off Peru, between circa lat 3° and 15°S, as compared with ETP I (ETP I, Figure 7). 39. GOBIIDAE (53 occurrences, 384 larvae) Larvae of several families of shore or bottom fishes have a much more widespread oceanic dis- tribution than would be anticipated from the distribution of adults. In the EASTROPAC area this applies particularly to larvae of Go- biidae, Scorpaenidae, Labridae, Bothidae, and Cynoglossidae. Based on pigmentation and meristics a minimum of eight kinds of goby larvae were taken. 42. NOMEIDAE (229 occurrences, 1,460 larvae) And other Stromateiodei (14 occurrences, 16 larvae) Four families of stromateoid, fishes were taken on EASTROPAC cruises: Amarsipidae, Nomeidae, Stromateidae, and Tetragonuridae. Three of these families contain oceanic species that are widely distributed in offshore waters; only fishes of the family Stromateidae are con- fined to coastal waters. Important papers deal- ing with stromateoid fishes include Grey (1955), Haedrich (1967, 1969), Haedrich and Horn (1969),-" and Horn (1970). In the EASTROPAC area, only the nomeids were common, occurring in about two-thirds of the collections made on ETP II. Larvae were obtained of two genera, Cubiceps and Psenes; larvae of the former were the more abundant, larvae of the latter were more diversified as to species represented. Larvae of a species in the family Stromateidae, Peprilus medius (Peters) , were taken at a single station on ETP II, 46.135 (2 larvae), but a larger collection was obtained at Oceanographer Station OP 168 (16 larvae). Larvae of Tetragonuridae (11 occurrences, 12 larvae) occurred in an equatorial band be- tween lat 2°N and 7°S, seaward of the Galapagos Islands (Figure 17) . As noted in the first EAS- TROPAC report, larvae of two species were taken: T. cuvier i Risso and T. atlanticus Lowe. Two specimens of Amarsipus carlshergi, de- scribed by Haedrich (1969) as a monotypic rep- resentative of a new family Amarsipidae, were ' Haedrich, R. L., and M. H. Horn. 1969. A key to the stromateoid fishes. Woods Hole Oceanogr. Inst. Ref. #69-70, 46 p. (Unpublished manuscr.) 1200 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE obtained on ETP II, and five specimens previ- ously had been taken on ETP I. These had been identified as Centrolophus-Mke with the notation that they probably represented an undescribed form. Identification of the material as Amarsi- pus carlsbergi was made by Dr. Michael H. Horn. Since little is known about this species in the eastern Pacific, I am listing all catch locali- ties. ETP II = Station 45.346 at lat 14°38.2'N, long 109°37.1'W, Sept. 8, 1967, 1 specimen, 26.2 mm; Station 47.272 at lat 11°20.8'N, long 88°00.5'W, Aug. 31, 1967, 1 specimen, 15.0 mm. ETP I = Station 11.066 at lat 06°49.8'N, long 118°55.5'W, Feb. 3, 1967, 1 specimen, 10.3 mm; Station 11.114 at lat 02°37.8'S, long 119°02.3'W, Feb. 7, 1967, 1 specimen, 30.0 mm; Station 11.306 at lat 12°03.5'N, long 126°00'W, Feb. 27, 1967, 1 specimen, 16.0 mm; Station 12.059 at lat 09°31.5'N, long 105°02.0'W, Feb. 22, 1967, 1 spe- cimen, 7.2 mm; Station 12.246 at lat 06°12.0'N, long 112°00.5'W, Mar. 16, 1967, 1 specimen, 7.3 mm. 43. OPHIDIIDAE (38 occurrences, 81 larvae) A number of kinds of larvae of this complex family were taken on ETP II, mostly in a coastal band between Acapulco, Mexico, and central Peru, but six occurrences were in a loose cluster about the Galapagos Islands. Only one kind has been identified to genus as yet; this is a form with conspicuously large pectorals (11 occur- rences, 15 larvae) whose larvae were clustered in the Gulf of Panama or immediately seaward. Dr. Daniel Cohen of the National Marine Fish- eries Service has identified larger specimens (small juveniles) as Brotula sp. A character- istic of this genus observed on several specimens was the presence of two ural centra in the "urostyle." Garman (1899) described 22 spe- cies of ophidiid-brotulids from the eastern trop- ical Pacific, few of which have been retaken sub- sequently. However, the variety of kinds of ophidiid larvae in our material attests to a speciose fauna. 47. SCOMBRIDAE (55 occurrences, 248 larvae) Scombrid larvae were markedly less abundant in ETP II as compared with similar coverage on ETP I (163 occurrences, 1,840 larvae). The majority of scombrid larvae from ETP II were those of Aiixis sp. (34 occurrences, 151 larvae) or were too small to identify with cer- tainty (30 occurrences, 84 larvae) . The remain- ing scombrid larvae included the wahoo, Acan- thocybium solanderi (Cuvier) (2 occurrences, 3 larvae) from Stations 45.065 and 46.004; the mackerel. Scomber japoyiicus Hottuyn, (2 oc- currences, 4 larvae) from near the Galapagos Islands; bigeye tuna, Thimnus obesus Lowe, (1 occurrence, 1 larva) ; skipjack, Katsuwonus pe- lamis (Linnaeus), (2 occurrences, 2 larvae); yellowfin tuna, Thimnus albacares (Bonnater- re), (2 occurrences, 2 larvae). Scombrid larvae were given to W. Klawe of the Inter-American Tropical Tuna Commission for identification. 52. TRICHIURIDAE (49 occurrences, 186 larvae) In the ETP I contribution, I pointed out the similarity in appearance of larvae of Diplospinus multistriatus Maul and those of Gempylus ser- pens, and the problems this raised about the distribution of genera between Gempylidae and Trichiuridae and perhaps about the need for two families. Treating larvae of the two families separately in this paper was done only for con- venience. The problems raised in the first ETP contribution still need to be solved. Three kinds of trichiurid larvae were obtained on ETP II: larvae of D. multistriatus Maul, Trichiurus lepturus (L.), and Lepidopus sp. The distribution of larvae of D. multistriatus (25 occurrences, 69 larvae) was strikingly sim- ilar on the two multivessel cruises (Figure 4 and Ahlstrom, 1971, Figure 14) . On ETP II, all but two occurrences were in a compact group at the southern inner half of the ETP pattern between circa lat 8° and 15°S and offshore to long 95 °W. Most ETP I collections of larvae of this species were obtained from this same 1201 FISHERY BULLETIN: VOL. 70, NO. 4 general area. The remaining two occurrences on ETP II were obtained at the northern, outer end of the pattern, again similar to the distribution of Diplospinus larvae on ETP I. On ETP II, there were no occurrences of Diplospinus larvae between these two widely separated groups; on ETP I two specimens were taken at intermediate localities. Larvae of this species have been ob- tained in a number of collections made in the North Pacific central water mass, with best dis- tributional information from the NORPAC Ex- pedition of August 1955. It is not taken in Cal- ifornia Current waters, hence the distribution in the Humboldt Current waters off Peru does not have a mirror-image replication in the Cal- ifornia Current, as has been found for a number of species. Larvae of Trichiunis lepUirus (20 occurren- ces, 106 larvae) were taken in a coastal band on ETP II (Figure 17). Eggs of this species are readily identified and occurred in many of the hauls containing Trichiunis larvae and in some additional hauls. Interestingly enough, larvae of this species were not obtained in ETP I col- lections, hence this is another exception to the general pattern of year-long reproduction by tropical pelagic fishes. Unlike larvae of Gempy- lus or Nealotus, which were widely distributed in the EASTROPAC area, larvae of this species appear to have a restricted, coastal distribution. Larvae of Lepidopiis sp. (3 occurrences, 9 larvae) were taken in contiguous stations at about lat 5°S off Peru (Figure 17). Larvae of Lepidopiis were taken in more hauls on ETP I (7 occurrences, 25 larvae, Ahlstrom, 1971, Fig- ure 14), all located between the equator and lat 5°N and offshore to long 92°W. This change in area of spawning of Lepidopiis from north of the equator on ETP I to the south of the equator on ETP II may not be significant, because of the paucity of positive hauls. If real, one can only surmise as to whether the two pop- ulations were discrete, with separate spawning seasons on the two sides of the equator. 53. BOTHIDAE (70 occurrences, 690 larvae) Bothid larvae occurred in more hauls than on ETP I (70 versus 56 occurrences) and in larger numbers (690 versus 199 larvae). The species composition, however, was similar (Table 22). A short section will be devoted to each of the forms listed in this table. Bothus leopardinus (Gunther) (27 occurrences, 97 larvae) Only larvae of B. leopardinus have been ob- Table 22. — Frequency of occurrence and relative abundance of larvae of flatfishes, Pleuronectiformes, on the second multivessel EASTROPAC survey, summarized by vessel pattern. Washiiis,ton 45.000 series Undai 46.000 tnted series Rochazvay 47.000 series Total EASTROPAC II Flatfish larvae No. positive hauls No. larvae No. positive hauls No. larvae No. positive hauls No. larvae No. positive hauls No. larvae BOTHIDAE Bothus leopardinus 2 2 15 45 10 50 27 97 Citharichthys-Etropus I 1 5 35 11 34 17 70 Cyclopselta sp. 0 0 4 26 9 12 13 38 Engyophrys sancli-taurentii 0 0 0 0 3 3 3 3 Monolene sp. 0 0 0 0 1 1 1 1 Sy acium ovale 6 15 17 201 32 264 55 480 Other Bothidae 1 1 0 0 0 0 1 1 Total Bothidae 7 19 28 307 35 364 70 690 CYNOGLOSSIDAE Symphurus spp. 2 5 16 109 38 243 56 362 Total Pleuronectiformes 7 24 30 416 46 612 83 1,052 1202 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE tained in EASTROPAC collections. Although B. mancus (Broussonet) has been recorded from Clarion Island, off the west coast of Mexico (Norman, 1934), larvae of this species have not been obtained. Larval material of B. man- cus has been examined from the vicinity of the Hawaiian Islands, and it differs from B. leopar- dinus in developing pigment on late stage larvae. Larvae of B. leopardinus were distributed in a broad coastal band between Manzanillo, Mex- ico, and lat 4°N (Figure 8). This distribution is more restricted than that found on ETP I (Ahlstrom, 1971, Fig. 10). On ETP I, there were nine occurrences between lat 5°N and 6°S, as compared with a single occurrence on ETP II. of at least 32 mm before transformation. The opercular spination is more strikingly developed on the Pacific form, and the pelvic fins become markedly more elongate, extending almost to the base of the caudal fin, whereas the fins attain only about 40% of this length proportionately in C. flmbriata. Engyrophrys sancti-laurentii (Jordan and BoUman) (3 occurrences, 3 larvae) Only three larvae of this species were obtained on ETP II, two from the vicinity of the Gulf of Panama and one from near Puntarenas, Costa Rica. Citharichthys-Etropus (17 occurrences, 70 larvae) Although labeled Citharichthys-Etropus as for ETP I, the larvae taken on ETP II probably rep- resent two species of Citharichthys, one with three elongated dorsal rays, the other with two elongated rays. Larvae of the latter were taken below the equator, either off Ecuador or near the Galapagos Islands (9 occurrences, 48 larvae). The form with three elongated dorsal rays was distributed in a coastal band between Manzanil- lo, Mexico, and Ecuador (8 occurrences, 22 lar- vae). Cyclopsetta sp. (13 occurrences, 38 larvae) Larvae of Cyclopsetta sp. occurred in a broad coastal band between lat 15°N and circa lat 5°S. The larvae have been identified tentatively as C. querna (Jordan and Bollman). A development- al series was recently described by Gutherz (1970) for an Atlantic species of this genus, C. fimbriata (Goode and Bean). The Pacific and Atlantic species are similar in having opercular spination, a pair of sphenotic spines on the head, and nine or so elongated dorsal rays. They dif- fer in several interesting respects. C. fimbriata transforms at a much smaller size, 14.0 mm, whereas the Pacific species can attain a length Monolene sp. (1 occurrence, 1 larva) A 16-mm specimen was obtained at Station 47.520. Larvae of Monolene develop a single, prominent elongated dorsal ray (2nd fin ray) — this ray was 6 mm long. Its meristics — D.82, A. 63, Vert. 39 — would fit Monolene asaedai Clark (Perkins, 1963) and possibly M. dubiosa Garman. The other two eastern Pacific species, M. maculipinna Garman and M. danae Bruun, have higher fin ray counts. Morrow (1957b) reported taking a 65-mm larva of M. maculipinna off Peru in a pelagic trawl fishing to 152-fm depth over rather deep water (1,300 fm) , Mor- row's specimen had the following meristics: D.98, A. 79, Vert. 43. Monolene danae Bruun (1937) was described from a juvenile taken in a pelagic trawl off Panama by the Dana in 1922. Sy actum ovale (Giinther) (55 occurrences, 480 larvae) Although larvae of 5. ovale were the most common bothid flatfish collected on both ETP I and ETP II, it was decidedly more abundant in ETP II as compared with ETP I (24 occur- rences, 84 larvae) . Larvae of Syacium occurred in a broad coastal band between Manzanillo, Mexico, and Ecuador (Figure 3); only three collections were obtained to the south of the 1203 FISHERY BULLETIN: VOL. 70, NO. 4 Table 23. — Familial composition of Lophiiform larvae taken on the second multivessel EASTROPAC survey, summarized by vessel pattern. Washington Undaunted Rockaway Totol 45.000 series 46.000 series 47.000 series EASTROPAC II Family No. positive hauls No. larvae No. positive hauls No. larvae No. positive hauls No. larvae No. positive hauls No. larvae Caulophrynidoe 0 0 2 2 0 0 2 2 Centrophrynidae 1 1 0 0 0 0 1 1 Cerotiidae 0 0 2 3 3 5 5 8 Gigantactinidoe 8 9 5 5 5 13 18 27 Himantolophidae 4 5 1 1 3 3 8 9 Linophrynidae 2 2 0 0 19 32 21 34 Melanocetidae 3 3 13 18 17 23 33 44 Oneirodidae n 11 13 18 23 53 47 82 Unidentified ceratiods 9 11 7 7 10 16 26 34 Antennariidoe 0 0 1 1 0 0 \ 1 Lophiidae 0 0 1 1 0 0 1 1 Total Lophiiform 25 42 33 56 56 145 114 243 equator. Most larvae of Syacium were under 5 mm in standard length, and few were as large as 9 mm. At the latter size, the adult comple- ment of fin rays were present in all fins except the pectoral, and the vertebral column was com- pletely ossified. The vertebral count in speci- mens examined was 10 + 25. 54. CYNOGLOSSIDAE (56 occurrences, 362 larvae) Larvae of Symphurus spp. were taken in a broad coastal band between Manzanillo, Mexico, and northern Peru. Symphurus larvae were taken in slightly less hauls than on ETP I (56 versus 63 occurrences), but in slightly larger numbers (362 versus 304 larvae). Two kinds of Symphurus larvae were widely distributed, and three or four additional kinds occurred spar- ingly. Of the two common forms, one developed two elongated dorsal rays and the other six elon- gated dorsal rays. 56. LOPHIIFORMES (114 occurrences, 243 larvae) Lophiiform larvae were accumulated during the identification and enumeration of ETP II larvae, and then studied as a unit. Ten families were represented (Table 23) . All but two of the specimens belonged to the subfamily Ceratioidei, a group of fishes whose ontogeny and taxonomy were dealt with in the impressive contribution of Bertelsen (1951). Ceratioid fishes have the most striking sexual dimorphism found in fishes. The males are parasitic in some ceratioids, free- living in others, but always quite small. Bertel- sen showed that sex can be determined in the late larval stage; a papilliform illicium develops on the head of the female, but not on the male. A major achievement of Bertelsen was defining the distinguishing characteristics of larvae of all 10 ceratioid families. His work makes it pos- sible to identify larger ceratioid larvae to the family level with assurance; however, small cer- atioid larvae are much more difficult to identify because they have few distinguishing characters. Although Bertelsen worked out life history se- ries to the generic or species level within all cer- atioid families, ontogeny of the less common gen- era and species still remains unknown. The ceratioids are a particularly difficult group in which to work out new developmental series. These cannot be based on larvae alone but must include transforming and adolescent specimens, preferably of both sexes, as well as adults. The EASTROPAC material, almost ex- clusively larvae, is inadequate for this purpose. Distributions of larvae are shown for five cer- atioid families (Figure 18), as noted in the dis- cussion of families. Most kinds of ceratioid lar- vae are quite rotund, hence aptly described as butterballs. 1204 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Figure 18. — Distribution of larvae of the lophiiform families Caulophrynidae (Caidophryne jordani) (open hex- agon with dot), Gigantactinidae {Gigantactis sp.) (open triangle with dot), Himantolophidae (Hiinantoloplms sp.) (open diamond with dot), Linophrynidae (2 or more genera represented) (open square with dot), and Melanoceti- dae {Melanocetus spp.) (open circle with dot) ; negative hauls are shown as small solid circles. Caulophrynidae (2 occurrences, 2 larvae) Bertelsen referred all material of Caulophryn- idae to a single species, Caidophryne jordani Goode and Bean. This is the only ceratioid fish known to develop pelvic fins. The two occur- rences (Figure 18) were north of the equator in the pattern occupied by the middle vessel. Centrophrynidae (1 occurrence, 1 larva) A single specimen was obtained of Centro- phyrne spinulosa Regan and Trewavas in the oflF- shore pattern (Station 45.325). Larvae of this species develop a digitiform barbel on the throat, a character unique to the species. 1205 FISHERY BULLETIN: VOL. 70, NO. 4 Ceratiidae (5 occurrences, 8 larvae) Larvae were obtained of two species of Cer- atiidae, Cryptopsaras couesi Gill and Ceratias holboelli Kroyer. Bertelsen had previously re- corded larvae of C. couesi from the eastern trop- ical Pacific, but not of C. holboelli. Ceratiid larvae are peculiarly "humpbacked," and the larger larvae of females develop "caruncles" on their backs. The caudal ray count in ceratioid fishes is constant at nine, except for two species that develop only eight caudal rays — C. couesi is one of these. Gigantactinidae (18 occurrences, 27 larvae) Larvae of Gigantactis sp. were taken in a tri- angular-shaped wedge, broadest offshore (Fig- ure 18) . Even small larvae of this family can be identified with certainty, because of the large size of the pectoral fins. Himantolophidae (8 occurrences, 9 larvae) Larvae of Himantolophidae were taken to the north of the equator, between lat 2° and 10 °N in all vessel patterns (Figure 18), Larvae are similar to Bertelsen's series for H imantolophus groenlandicus Reinhardt, and he recorded spec- imens from Panama. Two additional species of H imantolophus have been described from Panama or vicinity: H. azuerlucens Beebe and Crane and H. rostratu^ Regan. I have recorded the EASTROPAC larvae simply as Himantol- ophus sp. Linophrynidae (21 occurrences, 34 larvae) Several kinds of linophrynid larvae were taken, of which three were common — larvae of Borophryne apogon Regan, of the Linopht^ne macrorhinu^ group, and of the type designated by Bertelsen as "Hyaloceratis." All but two oc- currences of linophrynid larvae were in the in- ner pattern shoreward of the Galapagos Islands (Figure 18) . Most linophrynid larvae are more elongate than other ceratioid larvae and also have the lowest D and A counts, usually D3 and A3. Melanocetidae (33 occurrences, 44 larvae) At least two kinds of Melanocetus larvae were obtained on ETP II, with most specimens refer- able to M. polyactis Regan and the remainder to M. johnsoni Giinther. Most records of Mela- nocetus were from the northeast quadrant of the EASTROPAC pattern (Figure 18). Oneirodidae (47 occurrences, 82 larvae) At least one-third of the ceratioid larvae taken on ETP II were referable to the family Oneiro- didae. Bertelsen (1951) recorded seven kinds of oneirodid larvae belonging to six genera from collections made off Panama. All but one of these were taken in ETP II, together with a new record for the eastern Pacific. Oneirodid larvae sam- pled on ETP II included Chaenophryne draco- group, Chaenophryne longiceps-group, Dolopich- thys sp., Micropolichthys microlophus (Regan), Oneirodes eschrichti-group, Oneirodes melano- cauda Bertelsen, and Pentherichthys sp. Ber- telsen could identify some oneirodid larvae only to species groups, including the three listed above. Bertelsen included 24 nominal species in the Oneirodes eschrichti-group, most of which were possibly synonyms. Perhaps the most interesting record of an oneirodid larva from ETP II was of Oneirodes melanocauda from Station 47.008, off Panama. A male, 9.5 mm TL (6.5 mm SL), agreed in all essential characters with Bertelsen's description. This is one of the more heavily pigmented cer- atioid larvae. The fin counts were D6, A4, P19, C9. Bertelsen based his description of 0. we/a- nocauda on four specimens, 8 to 21 mm TL, the 1206 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE largest a metamorphosing female. These were obtained in the South China Sea, Indian Ocean, and Caribbean Sea. The EASTROPAC record is the first from the Pacific. The caudal fin is usually unpigmented in cer- atioid larvae, but caudal pigment is developed on several kinds of oneirodid larvae. Larvae of O. melanocauda have stippled pigment near the outer margin of the caudal rays. Larvae of Penthrichthys sp. have pigment sprinkled over much of the caudal fin rays. A third kind of oneirodid larva with streaks of caudal pig- ment between the rays was taken at Station 47.250 (ontogenetic series yet to be worked out) . The larvae of Pentherichthys from the east- ern Pacific are probably referable to P. atratus (Regan and Trewavas) . Collections were made at six stations in the inner pattern between lat 2° and 8°N. The 10 specimens ranged in total length from 3.2 to 7.0 mm. Bertelsen remarked on the paucity of small specimens of Penther- ichthys in the Dana material; only 2 of 19 spec- imens were under 7.5 mm in total length. Antennariidae (1 occurrence, 1 larva) The specimen, taken at an inshore Station, 46.132, on the middle pattern, was 7.5 mm SL and had fin counts of D-II + 1-13, A7, PIO, V5, C9. These counts could apply equally to species in the genera Histrio or Antennarius. Lophiidae (1 occurrence, 1 larva) A specimen of a lophiid larvae was obtained in the middle pattern at Station 46.145. This specimen, 15.5 mm SL (25.0 mm TL), had the following counts: D-II + I + III — 8, A6, P16/17, V6, C8. This specimen is referable to the genus Lophiomus. Carman described two species of Lophiomus from the eastern Pacific with identical counts to the above. Norman in his unpublished synopsis considered the genus monotypic with Carman's species as junior syn- onyms of L, setigerus Vahl. The third dorsal spine is rather widely separated from an anteri- or group of two spines and a posterior group of three. The last ray in both the dorsal and anal fins was bifurcate to the base, differing in this respect from the last ray in ceratioid fishes, which is single. The larvae had two spines above the eye on either side of the head, differing in this character from the published larval series for Lophius piscatorius and L. americanus (Taning, 1923) ; the pectoral fins were consider- ably smaller and compact. 57. OTHER IDENTIFIED (23 occurrences, 51 larvae) Two of the families, Amarsipidae and Stro- mateidae, have been discussed in the section deal- ing with Nomeidae and other Stromateiodei (No. 42). Other families included under "other iden- tified" include Eutaeniophoridae (2 occurrences, 2 larvae), Cadidae (7 occurrences, 10 larvae), Callionymidae (2 occurrences, 2 larvae), Fistu- lariidae (1 occurrence, 1 larva), Gerridae-E^- cinostomus sp. (4 occurrences, 15 larvae) , Micro- desmidae (4 occurrences, 6 larvae), Pomadasyi- d'cXQ-Anisotremus sp. (2 occurrences, 7 larvae), and Tetradontidae (1 occurrence, 1 larva). ACKNOWLEDGMENTS I wish to thank the many scientists who par- ticipated on EASTROPAC cruises for their care in collection and preservation of the plankton collections, and the technicians who laboriously sorted out fish eggs and larvae from the 1.0-m oblique plankton hauls for their thoroughness and patience. I especially wish to thank Eliza- beth Stevens for her careful identification of the fish larvae obtained on the four EASTROPAC monitoring cruises made by the David Starr Jordan, Kenneth Raymond for preparing the distribution charts, Elaine Sandknop and Amelia Gomes for their aid in many aspects of the work, such as preparation of cleared and stained spec- imens and x-raying of juvenile and adult spec- imens. H. GeoflFrey Moser worked closely in 1207 FISHERY BULLETIN: VOL. 70, NO. 4 studies of larvae of Myctophidae; W. L. Klawe identified the tuna larvae. I also wish to thank Daniel Cohen for his help in identifying juvenile specimens of Brotula, Michael Horn for his help in identifying specimens of Amarsipidae, Robert Wisner for his help in identification of the Diaphus fauna of the EASTROPAC area, and Solomon Raju for his help with eel lep- tocephali. I wish particularly to thank David Kramer, H. GeoflFrey Moser, and Walter Matsu- moto for reviewing the manuscript. LITERATURE CITED Ahlstrom, E. H. 1959. Vertical distribution of pelagic fish eggs and larvae off California and Baja California. U.S. Fish Wildl. Serv., Fish. Bull. 60:107-146. 1971. Kinds and abundance of fish larvae in the eastern tropical Pacific, based on collections made on EASTROPAC I. Fish. Bull., U.S. 69:3-77. Baird, R. C. 1971. The systematics, distribution, and zooge- ography of the marine hatchetfishes (family Sternoptychidae). Bull. Mus. Comp. Zool. Har- vard Univ. 142:1-128. Beebe, W., and J. Crane. 1947. Eastern Pacific Expeditions of the New York Zoological Society. XXXVII. Deep-sea ceratioid fishes. Zoologica (N.Y.) 31:151-182. Beebe, W., and M. Vander Pyl. 1944. Eastern Pacific expeditions of the New York Zoological Society. XXXIII. Pacific Myctoph- idae. (Fishes). Zoologica (N.Y.) 29:59-95. Bertelsen, E. 1951. The Ceratioid fishes - Ontogeny, taxonomy, distribution and biology. Dana Rep. Carlsberg Found. 39, 276 p. Blackburn, M., R. M. Laurs, R. W. Owen, and B. Zeitzschel. 1970. Seasonal and areal changes in standing stocks of phytoplankton, zooplankton and micro- nekton in the eastern tropical Pacific. Mar. Biol. (Berl.) 7:14-31. Bruun, a. F. 1937. Monolene danae, a new flatfish from Panama, caught bathypelagically. Ann. Mag. Nat. Hist., Ser. 10, 19:311-312. Bussing, W. A. 1965. Studies of the midwater fishes of the Peru- Chile Trench. In G. A. Llano (editor), Biology of the Antarctica Seas II, p. 185-227. Antarctic Res. Ser. 5, Natl. Acad. Sci.-Natl. Res. Counc. Publ. 1297. Collette, B. B., R. H. Gibbs, Jr., and G. E. Clipper. 1969. Vertebral numbers and identification of the two species of dolphin (Coryphaena) . Copeia 1969:630-631. Craddock, J. E., and G. W. Mead. 1970. Midwater Fishes from the eastern South Pa- cific Ocean. Scientific Results of the SE Pacific Exped. Anton Bruun Rep. no. 3, 46 p. Einarsson, H., and B. Rojas de Mendiola. 1967. An attempt to estimate annual spawning in- tensity of the anchovy (Engraulis ringens Jen- yns) by means of regional egg and larval surveys during 1961-1964. Calif. Coop. Oceanic Fish. In- vest. Rep. 11:96-104. Garman, S. 1899. Reports on an exploration off the west coasts of Mexico, Central and South America, and off the Galapagos Islands, in charge of Alexander Agassiz, by the U. S. Fish Commission steamer "Albatross," during 1891, Lieut. Commander Z. L. Tanner, U. S. N., commanding. XXVL The fishes. Mem. Mus. Comp. Zool. Harvard Coll. 24, 431 p. Grey, M. 1955. The fishes of the genus Tetragonurus Risso. Dana Rep. Carlsberg Found. 41, 75 p. 1960. A preliminary review of the family Gonos- tomatidae, with a key to the genera and the de- scription of a new species from the tropical Pa- cific. Bull. Mus. Comp. Zool. Harvard Coll. 122: 55-125. GUTHERZ, E. J. 1970. Characteristics of some larval bothid flatfish, and development and distribution of larval spotfin flounder, Cyclopsetta fimbriata (Bothidae). Fish. Bull., U.S. 68:261-283. Haedrich, R. L. 1967. The stromateoid fishes: Systematics and a classification. Bull. Mus. Comp. Zool. Harvard Univ. 135:31-139. 1969. A new family of aberrant stromateoid fishes from the equatorial Indo-Pacific. Dana Rep. Carlsberg Found. 76:1-14. Haedrich, R. L., and J. G. Nielsen. 1966. Fishes eaten by A/episfinniS (Pisces, Iniomi) in the southeastern Pacific Ocean. Deep-Sea Res. 13:909-919. Hildebrand, S. F. 1946. A descriptive catalog of the shore fishes of Peru. U.S. Natl. Mus. Bull. 189, 530 p. Horn, M. H. 1970. Systematics and biology of the stromateid fishes of the genus Peprilus. Bull. Mus. Comp. Zool. Harvard Univ. 140:165-261. Meek, S. E., and S. F. Hildebrand. 1923. The marine fishes of Panama. Part I. Field Mus. Nat. Hist., Publ. 215, Zool. Ser. 15:1-330. 1208 AHLSTROM; KIND AND ABUNDANCE OF FISH LARVAE 1925. The marine fishes of Panama. Part II. Field Mus. Nat. Hist., Publ. 226, Zool. Ser. 15:331-707. 1928. The marine fishes of Panama. Part III. Field Mus. Nat. Hist, Publ. 249, Zool. Ser. 15: 709-1045. MiTO, S. 1960. Egg development and hatched larvae of the common dolphin-fish Coryphaena hippiirus Linne. Bull. Jap. Soc. Sci. Fish. 26:223-226. Morrow, J. E. 1957a. Shore and pelagic fishes from Peru, with new records and the description of a new species of Sphoeroides. In Studies in ichthyology and oceanography off coastal Peru, p. 5-54. Bull. Bingham Oceanogr. Collect. Yale Univ. 16(2). 1957b. Mid-depth fishes of the Yale South Amer- ican Expedition. In Studies in ichthyology and oceanography off coastal Peru, p. 56-70. Bull. Bingham Oceanogr. Collect. Yale Univ. 16(2). MOSER, H. G., AND E. H. AHLSTROM. 1970. Development of lanternfishes (family Myc- tophidae) in the California Current. Part I. Species with narrow-eyed larvae. Bull. Los Ang. Mus. Nat. Hist., Sci. 7, 145 p. NaFPAKTITIS, B. G., AND M. Nafpaktitis. 1969. Lanternfishes (family Myctophidae) collect- ed during cruises 3 and 6 of the R/V Anton Bruun in the Indian Ocean. Bull. Los Ang. Cy. Mus. Nat. Hist., Sci. 5, 79 p. Norman, J. R. 1934. A systematic monograph of the flatfishes (Heterosomata). Vol. I, Psettodidae, Bothidae, Pleuronectidae. Br. Mus. (Nat. Hist.), Lond., 459 p. Orton, Grace L. 1964. The eggs of scomberesocid fishes. Copeia 1964:144-150. Parin, N. V. 1968. Ichthyofauna of the epipelagic zone. [In Russian.] Izdatel'stvo "Nauka," Moscow, 186 p. (Transl. Israel Program Sci. Transl. 1970, 206 p. available Natl. Tech. Inf. Serv., Springfield, VA 22151 as TT 69-59020.) 1971. On the distributional pattern of midwater fishes of the Peru Current Zone. [In Russian.] Akad. Nauk SSSR, Tr. Inst. Okeanol. 89:81-95. Parr, A. E. 1931. Scientific results of the Second Oceanograph- ic Expedition of the "Pawnee" 1926. Deepsea fishes from off the western coast of North and Central America. Bull. Bingham Oceanogr. Col- lect. Yale Univ. 2(4), 53 p. Perkins, H. C. 1963. Redescription and second known record of the bothid fish, Monolene asaedai Clark. Copeia 1963:292-295. Regan, C. T. 1926. The pediculate fishes of the suborder Cera- tioidae. Dan. "Dana"-Exped. 1920-22, Oceanogr. Rep. 2, 45 p, 13 plates. Taning, a. V. 1923. Lophius. Rep. Dan. Oceanogr. Exped. 1908- 10 Mediterr. Adjacent Seas 2, Biol. A. 10, 30 p. Wisner, R. L. 1971. Descriptions of eight new species of mycto- phid fishes from the eastern Pacific Ocean. Copeia 1971:39-54. 1209 FISHERY BULLETIN: VOL. 70, NO. 4 Appendix Table 1. — Counts of fish larvae, tabulated by family, for all stations occupied on the second multivessel EASTROPAC survey (EASTROPAC II). 1 i o 2 t o (U i a) g B 1 o i u § u CO i -8 o "3 ■a o . 1 1 E 1 ! 2 c Bathylagidae Paralepididae "3 a o O m T3 3 a "3 Bregmacerotidae Exocoetidae Trachypteridae Apogonidae Bramidae Chiasmodontidae Coryphaenidae Nomeidae Scombrtdae Other identified larvae ^ J ^ i . 3 1 4 3 1 2 1 54 '►b.023 61 106 i . 2 2 8 I 2 2 • 189 ^5.02'f 86 9b 1 ] * . 3 2 2 1 3 5 204 45.026 40 30 9 4 4 10 3 2 3 1 L 2 2 112 45.U28 24 22 L c ) 9 4 3 1 3 • 8 1 1 84 •VS-OSO 34 60 b 2 1 2 * • • 9 2 3 121 45.032 158 lU 14 1 7 3 1 12 5 2 2 17 10 10 353 45.034 146 37 8 ; ? 1 1 4 2 9 I 1 9 224 45.035 40 29 7 2 3 4 4 1 • 9 3 1 107 45.(j37 41 25 1 10 >■ t 10 1 • 2 11 L 2 1 3 124 45.039 77 82 7 : J 1 4 12 4 11 2 5 5 1 216 45.041 29 49 2 ^ 8 1 4 4 9 • 1 » • 10 1 122 45.043 17 16 2 ■3 3 1 5 4 6 3 3 1 67 45.044 50 41 1 3 4 4 5 2 4 • 4 1 5 127 45.046 43 57 2. 43 2 1 6 6 6 1 5 2 5 2 1 205 45.048 25 21 Vc 9 2 1 1 6 1 2 1 L 1 2 4 92 45.050 19 3 ( 1 1 2 2 • • » • • • 2 1 40 45.(j51 87 15 I'- 4 1 . 1 » • • 1 2 • 125 45.053 75 19 31 2 3 11 2 I • 4 3 151 45.054 44 3 It 1 , 4 24 • • I 1 97 45.056 22 6 2i 1 2 23 L • 2 5 7 90 45.058 146 lb 1. 2 I 1 6 2 11 I 1 I 1 * • 6 5 228 45.060 02 4 i 5 f 5 1 3 5 29 . 1 3 1 » • 3 7 loS 45.063 29 46 t 2 1' » 1 3 q U 3 2 2 2 6 9 138 45.065 116 374 c 6 2' . 25 1 2< ? 1 2 3 5 3 11 23 634 45.067 105 584 3" ? Q 2 4> i 3 56 • • I 1 ? 3 1 5 I 31 11 992 45.071 42 '*'J » • -> 2 . 1 L 1 1 3 5 1 121 45.073 65 6a 1' I 2 5 1 . 2< !^ • • 3 16 203 45.078 n 107 K 1 6 3 } 1 • 4 2 212 45.033 37 26 1. I . 5 • • . 1 • 2 4 4 105 45.036 65 77 e ' 1 ? 3 9 3 • • 2 21 191 45.090 47 67 J 1 i 8 5 , , I I 2 12 156 45.094 150 107 u 5 6 27 1 > I ? . 2 1 2 33 346 45.098 43 10 2 a • • • 1 • 10 85 45.102 136 32 i< I 5 1 24 • • I L 1 16 288 45.106 265 21 2 ? 5 4 1 • 1 • « L 2 22 354 45.110 545 32 18f I 5 12 5 1 , , 7 5 9 817 45.114 40 2C 5 ? 2 • • • 5 16 138 45.117 92 49 8< 1 4 2 • • » • 6 2 20 263 45.121 55 98 20< 4 3 2 • • • • 2 25 399 45.125 41 325 3^ 1 1 • • 1 6 2 • 416 45.127 21 39 3 • • 6 • • 7 2 3 10 122 45.129 22 i 2' • • • • i • < k • 6 63 45.131 73 337 2: • • , , > t ? 1 2 3 25 472 45.133 323 33 2. 1 / ♦ . 5 5 7 » • 4 . L 3 3 416 45.135 165 17 1' 2 1 3 2 t « . L 1 10 227 45.137 11 • 1, ? 1 1 2 • • > • « , • • 27 45.139 24 2 1 1 3 • , , , . • 3 36 45.146 106 8 7 4' J a 11 10 • • » • 4 . 2 20 44 338 1210 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 1. -Counts of fish larvae, tabulated by lamily, for all stations occupied on the second multivessel EASTROPAC survey (EASTROPAC II) .—Continued. K a 2 a S o 1 i o O 1 s > o. § 2 O §■ u w Melamphaidae Bre^macerotidae Exocoetidae Trachypteridae Apogonidae Bramidae Chlasmodontidae eS ■a § o Nomeidae Scombridae Other identified larvae Clj i is g •a 1 cd 1 B •1 s ■a .3 ■a O H 'fS.iba 8 • » • ♦ • • • • • • • 8 ^5.165 152 65 212 1 6 : . 1 ^ 2 8 • • • 3 2 12 471 45.167 86 6 1( • 2 2 2 1 • 10 119 45.169 56 15 1 • . . • • 8 82 45.171 26 1 1. • • t > • • \ 1 9 50 45.173 10 • . • • * • • • « 13 45.175 62 27 Z' 1 • L 1 1 2 3 127 45.177 70 14b 4( 5 • 1 7 3 1 281 45.179 59 33 ( 1 1 1 • 9 113 45.183 26 U 7( . A • 2 1 2 123 45.187 255 14 8' i< !> 3 • 2 3 2 382 45.191 144 lb 8C 3 17 • 1 4 272 45.194 49 4 41 } 11 • » • 4 113 45.198 251 21 3 ? 2? • 2 11 350 45.202 53 19 t £ ' 9 • 2 • 92 45.206 22 3 L • » • 3 30 45.283 102 13 £ I • 2 3 129 45.287 25 82 I 3 « » • • 114 45.289 157 125 I 16 • 1 1 2 306 45.293 157 58 i: t- i . 11 1 I ? 1 1 2 2 265 45.297 231 107 I' I 1? ? 17 2 6 1 I . 1< ) L K • 8 440 45.301 24 26 i; I ^ i i 5 • J L ; ? . 1 L 1 4 • 93 45.305 119 1040 1" i I ' 1 1 ( ) . I 6 10 1 9 1217 45.309 44 150 2, t i 4 1 1 t 1 5 . . 3 2 2 7 251 45.313 49 36 c i. •■ t I 6 • * t ■> , , 1 9 119 45.316 41 34 t L 2 • \ I ♦ . • 1 2 99 45.319 32 46 r i 3 • 3 t t • • 2 I 111 45.321 45 13 2( i \ • 7 I t . • 1 1 94 45.323 51 4 L ? • 7 L • 1 1 74 45.325 1002 28 1' hi I 3 1 7 L 6 8 I 4 3 1128 45.329 69 4 1 } 1 • 3 1 4 1 3 110 45.331 235 5 1] i 1 5 • J 1 10 1 1 290 45.333 108 28 ic i. '■> 1 • 5 1 25 . 13 • 201 45.335 65 32 H 1. > ( J 3 3 1 L 9 8 2 155 45.337 49 12 2 u 3 1 12 ( 3 • 2 ; 14 1 16 158 45.339 26 8 It ) < \ t I • 7 » 3 1 3 ] 5 9 9 100 45.341 47 70 r 1 J 4 4 3 < ? 1 1 144 45.343 17 32 i 1 I ? 2 • Z i ? 3 4 2 77 45.344 21 37 > t 1 3 1 J 1 11 4 99 45.346 19 19 I ] 4 , 3 £ ? 4 2 58 45.348 30 120 4 4 . » • 2 3 166 45.350 70 297 1 • . , 6 1 5 383 45.352 29 66 J 2 • 1 1 3 • 10 116 45.356 69 153 1 L 6 1 3 19 ' ) 1 2 • 282 45.358 98 107 I • I 4 6 5 > 2 3 7 242 45.360 18 3r ? 2 , 9 ? > 4 3 4 91 45.362 36 31 t 3 8 • > 8 < > 17 8 1 128 45.365 99 55 1 • . 2- r <; ► 103 15 6 311 45.367 23 10 r 1 • 5 • 1 19 • 73 138 45.369 101 106 5 10 1 2 2 = i . 7 i; ! 61 11 15 357 45.371 120 239 \ 1 5 4 3 10 i: 5 . 1 . 7 t > 10 6 11 446 45.373 117 149 1' t 1 . 2 2 13 <: ) 10 3 • 321 1211 FISHERY BULLETIN; \0L. 70, NO. 4 Appendix Table 1. — Counts of fish larvae, tabulated by family, for all stations occupied on the second multivessel EASTROPAC survey (EASTROPAC U) .—Continued. K m o I— < H < H w o oi -a a 3 o •§ % m o a o O .a o a § 0) o •1-4 ■3 o o 2 s o 2 o I Id o a) 00 •a a 73 fi cd ft 0) cd <1> •« TJ •s 0 (-] d ID J3 0 U a 0) (U H ■3 P9 CD T3 cd •c (U "O v cd a> i 1: •s 0 J3 0 g 0 0 bo 0 cd 0 ot i3 ti 0. M H < n o s) T3 a o 73 O e l-l Id a ^^ > . 3 1 5 13 2 2 1 2 7 1 2 2 • 125 46.026 206 140 I . 2 . 5 6 4 11 3 1 4 13 3 2 • 403 46.028 412 157 IC ) . 9 2 12 b ] 10 4 . L . 3 15 2 2 1 8 655 46.030 05 7 1. I . 3 2 8 • < 2 2 1 t ? • 7 131 46.032 9 • 3 • « • 1 * « • , • 1 1 L . 1 • • 21 46.034 152 411 r 1 • 3 5 3 L 1 9 . I 1 1 ] L . 7 2 3 248 46.036 436 127 2f 1 7 8 14 4 1 1 1 3 1 3 • 6 634 46.038 20 7 i( • 3 1 ] 2 1 • » • • • • • 3 49 46.040 25 20 t 1 • 4 1 • I - • • 3 • 4 69 46.042 62 133 < • 4 • 2 2 1 i • • 1 1 1 4 218 46.044 89 96 2 1 1 1 5 1 1 • • L . 1 2 U 232 46.046 78 57 2, 2 1 2 3 3. , 2 L 1 8 • 11 191 46.048 63 5^ 2' • 10 3 5 4 1 4 L . 1 3 3 4 181 46.050 250 184 1 1 12 24 I 7 11 . . 11 L 1 ^ t . 3 4 15 547 46.052 319 217 2. • 7 18 13 4. . 11 L . 1 < ? . 13 3 5 631 46.054 211 101 2 3 20 2 3 L 6 9 . » • 2 2 3 14 497 46.055 100 73 IC 1 5 18 4 4 4 . 3 5 4 * • • • 16 249 46.057 751 649 2 10 70 2 3 2 3 5 . 6 2 • 1506 46.059 494 504 5 2 59 2 1 L 1 t I . 22 • 5 1098 46.061 140 98 3 1 1 3 23 1 » • • . 14 • 72 366 46.063 48 19 f i i , 2 10 • * » • • • • 5 20 116 46.065 404 97 1' 5 4 50 2 6 2 • 8 592 46.067 269 121 2( 2 5 18 1 » • • 2 3 20 461 46.069 116 55 2 • 22 53 4 • • 5 33 20 331 46.071 72 27 3' • 14 22 • < » • • 4 3 16 192 46.075 35 21 3 • 5 17 4 i • • • • • 6 120 46.077 36 19 1 1 1 3 1 1 • 3 • • 79 46.079 408 30 I' 21 23 31 2 » • • I . 4 • 1 536 46.082 61 46 1 ^ 17 15 2 • • 1 L . 8 • 8 176 1212 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 1. -Counts of fish larvae, tabulated by family, for all stations occupied on the second multivessel EASTROPAC survey (EASTROPAC 11).— Continued. 2 < H 1 t 1 0) 1 s a s Stemoptychidae Chauliodontidae Idiacanthidae ■s 2 CH 1 o 1 -3 a -A 0) 1 2 at "3 §" o to si 2 '5 1 1 0 a u PQ ■a 0 1 Trachypteridae Apogonidae Bramidae 01 0 £ 1 u d •a 1 cd 0 1 i z (t) 1 u w 0) cd ^^ •s s g 2 43 0 1 § "S p a t, T3 TO (S 0) cd t cd 3 H 46.004 45 39 13 2 21 21 i • • 6 1 8 159 46.086 10? 2 3i 25 -> 33 26 3 a 4 8 • 1 443 46.088 124 6 'J 16 6 8 27 • • 3 4 1 9 258 46.09'J 99 24 47 b 4 2H 6 • 7 • • 143 369 46.U92 100 29 36 4 3 10 4 . 1 4 • 4 2 200 46.094 300 156 87 7 3 21 8 3 » • 3 1 1 595 46.096 240 249 54 ? 13 24 5 5 3 3 2 7 8 4 628 46.098 73 42 3 5 8 y 4 3 . • 1 2 3 15 197 46.100 125 8 41 . ? 1 2 9 • • 1 1 6 202 46.102 334 56 69 1 4 3 10 9 7 . • • 4 3 4 508 46.104 109 48 36 4 « 10 1 2 1 2 2 1 12 232 46.106 115 41 19 1 9 1 2 2 7 L 2 4 4 210 46.108 250 38 45 7 27 • • • 20 4 5 1 18 415 46.110 162 10 22 • 5 • 3 • 2 3 15 i • 230 46.112 562 75 14 I lb 3 • • 3 14 49 • • 738 46.114 31 6 2 1 3 • . • 3 7 • • 61 46.116 14 t> 3 • • 1 3 • 7 1 • 2 37 46.118 176 85 6 1 3 7 5 4 13 1 4 3 321 46.120 80 100 13 -3 1 9 1 , 2 12 . 10 • • 237 46.122 17 29 2 4 4 ? 1 10 8 • • 85 46.124 19 22 • 2 3 4 1 12 6 3 3 79 46.126 63 90 13 1 10 7 7 • 4 . 20 10 • 237 46.128 113 131 11 9 H U 9 1< 1 12 7 7 3 343 46.130 46 5-i 5 11 6 6 1( 1 9 5 20 9 183 46.132 26 2b • 5 9 • / • 17 11 12 19 1 130 46.134 9 12 • 8 • • 51 2 • 20 93 18 8 684 46.135 81 12 • 8 . • 92 1 5 6 149 24 4 1218 46.137 201 2M 1 1 24 . 7 c • 4 . 278 • 7 561 46.139 175 r>t 1 • 11 1 6 • 15 3 15 * 1 283 46.141 430 20 4 2 18 1 4 't 1 6 8 3 3 518 46.143 113 l3 31 • 10 • 2 • 7 I 3 2 5 201 46.145 282 119 24 2 3 8 3 2 2 4 2 4 2 1 489 46. 147 484 41 49 1 41 1 4 • 5 L 4 • 6 642 46.149 1333 14 30 10 22 . 7 • 6 4 2 2 1432 46.151 1 i58 9 24 4 13 • 8 • • 1 • • 1420 46.153 2561 165 81 3 4 • 7 • 4 , 18 • 16 2864 46.155 973 140 67 w 6 1 8 2 3 } 11 • 3 1234 46.157 490 27 70 7 6 5 8 7 ; 2 • 7 2 12 660 46.159 71 U 32 t 2 1 2 } 2 • 6 • 1 134 46.161 350 t^l 73 2 3 12 15 • L 12 1 3 531 46.163 290 Ma 64 4 10 12 13 1 L • 6 2 41 540 46.165 90 36 60 • 18 15 7 3 ] I • 9 5 9 261 46.167 195 40 49 I 3/ 7 11 Z i 5 7 2 25 386 46.169 163 2 36 30 4 52 46 10 I 4 4 10 563 46.171 255 55 15 C 44 2 8 7 • 72 • 60 545 46.173 33 15 6 • 10 6 1 • 5 • 2 78 46.175 52 7b 4 • ?1 3 2 • 2 • I 162 46.177 124 33 17 2 45 16 10 • 26 • 6 281 46.179 213 110 54 4 135 1 9 2? • i 110 3 14 688 46.181 52 58 13 } 2 8 132 13 14 1 21 18 4 339 46.183 85 21 29 ■» ^ 11 47 I! 9 t 2 2 5 • 4 235 46.185 262 9 7 5' + 10 45 12 9 3 2 6 • 3 505 1213 FISHERY BULLETIN-: VOL. 70, NO. 4 Appendix Table 1. — Counts of fish larvae, tabulated by family, for all stations occupied on the second multivessel E ASTRO PAC survey (EASTROPAC U) .—Contijiued. 2; < H w <0 at ■a •S •S -S e s m o a o O J3 O i ■^ s 3 o (D •a 2 a S o 35 s o 01 y 'E •a 01 • ^ t . 29 I 10 463 J . 127 18 22 567 1 I L . 12 2 35 347 . 1 r . 82 7 21 580 2 3 ) .114 6 50 731 3 . i; » . 33 4 9 231 • • J 3 . 22 5 19 365 1 3 L . 38 6 9 480 1 ) . 34 2 • 370 . . ? . 65 11 23 587 9 6 102 . 29 2 257 . 127 20 1002 3 5 162 I 3 132 . 258 1 283 . 62 2 84 . 227 1 282 2 1 15 4 • 21 9 • 134 . 293 4 305 . 35 2 38 . 99 1 110 . 96 • 137 . 46 • 158 . 20 • 52 1 2 48 9 • 47 . 18 9 90 4 9 89 3 • • 2 52 2 4 5 4 181 6 3 8 119 1 6 • 11 50 ( ) 6 8 5 100 1214 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 1. — Counts of fish larvae, tabulated by family, for all stations occupied on the second multivessel EASTROPAC survey (EASTROPAC U).— Continued. 1 s v CD 2 a a) 1 1 1 0) i -§ o 3 23 2 • 11 518 47.175 327 95 't 2 < f 16 1 • 5 15 1 • 470 47.177 119 IOC 10 • 6' i 157 4 1 1 19 . 12 • 40 532 47.179 228 14 1 2 1 I 139 i 2 3 3 • 42 499 47.181 150 32 7 1 • 1< I 59 1 • • . 11 • 3 276 47.185 350 31 45 1 3 1< > 140 23 ^ » 5 1 . 27 • 14 658 47.189 28 5 8 • i 3 97 3 1 4 • 5 • 26 185 47.193 2 2 • • I I . • • • « 6 47.197 97 9 13 2 2 ^ 123 5 I 2 17 2 8 306 47.^01 392 152 24 1 l< i 140 20 10 9 2 8 • 54 824 47.205 109 4U 27 • 2 5 152 20 17 2 4 8 3 125 541 47.213 769 440 43 14 ? 58 26 25 2 • 43 . 43 5 40 1513 47.217 44 0 90 4't 0 ♦ 3D 20 4 5 . 2b • 11 677 47.221 452 36 5^ 13 ? 30 43 9 i 3 . 17 • 14 676 47.225 512 151 53 5 1 34 31 ) 6 2 15 . 16 2 13 860 47.229 496 2 72 83 14 ( 3 17 11 1 9 ; 1 5 . 21 • 3 942 47.233 1441 249 95 ' 42 ( 3 14 31 '. > 18 1 3 L 16 5 37 1975 47.237 100 17 18 t 8 3 • 1 7 • 3 165 47.240 201 46 51 5 3 1 . 16 . 13 • 1 348 47.242 953 94 18 2 J 20 6 ' . t 8 I 14 • 6 1138 47.244 90 5 129 16 1 ? 22 ? 5 1 12 I 26 2 • 1135 47.246 111 8 6 I 2 1 • 12 4 • 2 148 47.250 297 18 • ( :. 3 • • 5 . 15 3 1 349 47.254 245 27 11 I 8 3 , 8 . 15 2 • 326 47.258 516 3t) 6 3 ■ • 6 I 109 37 26 752 47.268 55 15 1 1 6 7< 1 • b 29 3 • 191 47.^72 21 6 5 • • 2 • 11 I 18 12 9 86 47.276 193 5 208 6 ? 45 3 , 5 I 29 • 2242 47.278 54 3b • 6 a 1 1 :? 8 • 111 47.280 175 8 3 ? 14 5 • 1 1 • 209 47.283 184 6 6 4 2 • 3 5 2 212 47.286 1272 15 70 3 18 I 3 , 1 . 12 5 1403 47.288 1311 44 17 . 13 ? 11 2 • 2 5 I 5 1414 47.290 186 11 93 2 3 4 10 3 , 2 • 2 • 3 319 47.292 276 30 46 6 7 9 , I • 2 5 9 3 3 402 47.295 702 101 103 1 4 J 7 15 S L 1 1 ? 1 4 ? 17 4 4 985 47.297 552 136 64 4 1 b 17 11 . 16 3 1 2 3 • 13 . 26 1 11 882 47.301 438 54 71 5 S H 10 I 15 ? 1 • 1 6 8 2 • 634 47.304 92 21 27 1 1 4 9 •^ • 2 > • 2 6 2 3 180 47.306 309 68 !0 5 1 S 15 10 ' 12 • 1 I 1 I 3 13 . 18 4 18 570 47.310 780 154 54 9 3 3 18 51 ? 8 1( J • • t ' 5 i 2 36 . 42 22 18 1249 47.314 241 83 18 6 1 1 24 14 2 1 1 ? 1 I } 4 13 . 24 5 1 456 47.318 73 58 17 4 1 D 2 3 22 2 1 • • • i ' 1 17 4 • 15 248 47.322 152 33 14 > • 5 61 i 3 24 » • • . 1 13 14 11 4 338 1215 FISHERY BULLETIN: VOL. 70, NO. 4 Appendix Table 1. -Counts of fish larvae, tabulated by family, for all stations occupied on the second multivessel EASTROPAC survey (EASTROPAC U) .—Continued. K 2 < d t o >> i § c o O 1 w 1 Xi o 1 t— i other Stomiatoldet Bathylagidae Paralepididae Scopelarchidae Melamphaidae 01 3 .4-* 2 0) o di Exocoetidae Trachypteridae Apogonidae Bramidae Chiasmodontidae Coryphaenidae Nomeidae d a •c 'C 1 in cd t .2 •s r 1 u 3) J3 o 10 s 2 c p i 1 'S 5 0) en « Si .2 •a ^.7.326 217 4J U . I 152 15 1 58 ? 1 4 . 26 . 50 582 'tl.iiO 436 148 80 • 11 137 56 1 14 12 . 16 . 18 15 947 '*7.3i'* 165 29 32 • 9 53 21 3 2 4 321 ^7.338 78 19 51 • ^ 24 8 4 • • • 7 199 ^7.3^2 302 2^ 62 3 ' 19 32 1 S 4 4 . 10 472 47.3^5 856 45 7H 2 U 26 b . 11 • • • . 15 1061 -^7.349 28m 61 4 9 . i 16 * 1 1 1 2 426 47.351 684 66 85 5 10 60 3 3 11 6 . 13 966 47.354 71 37 • • 1 • » • • • • 7 168 47.357 239 67 27 1 4 14 i. J 5 1 2 . 15 390 47.359 68 5ti 2i • ' lu 1 1 2 • • • 5 176 47.362 81 4 34 11 3 • 3 1 > • • 139 47.364 96 7 4 • 3 • 2 K • • 116 47.367 97 24 22 6 . U • 1 3 1 3 180 47.369 13 4 1 • 3 • 1 » * • 2 26 47.371 49 12 1 • I 2 • 2 1 2 70 47.373 260 239 49 1 2 3 ? 1 > • • . 16 2 579 47.376 52 76 16 3 9 • 5 k • • 1 5 173 47.379 7 56 9 3 b • 1 2 4 3 • 95 47.382 35 114 20 7 } H • 4 5 1 4 205 47.415 30 10 4 « 2 • 1 a L » • • 2 • 49 47.430 82 109 7 1 t 1 ► . L > • • 7 • 210 47.432 14? 101 21 • • 2 1 ? K • * 2 2 276 47.436 10 2 11 . 3 2 • t > 1 2 25 2 64 47.438 100 35 14 1 } 1 7 6 8 1 I 11 28 2 1 215 47.4'»0 34 109 30 3 10 1 1 L . 3 . 12 8 • 212 47.443 56 17 57 • 2 1 1 3 » • • 6 1 146 47.446 152 4 2 • L 2 • 4 , . 8 ■ 173 47.450 570 85 126 10 12 ?C 41 10 2 41 12 936 47.454 200 24 75 1 0 41 40 6 3 4 5 407 47.458 32 5 6 • i 42 37 3 1 3 8 141 47.462 • • • • • « > • • • • • I 47.466 64 8 3 , 3 7 14 2 • • « 30 1 152 47.470 81 110 7 • 6 4 1 • • < 5 • 215 47.478 38 36 34 • 3 6 4 1 3 1 • 130 47.486 194 38 29 • L 11 12 2 112^ 11 4 324 47.490 238 24 25 • I 13 20 7 ? 2 2 1 7 17 364 47.494 77 10 40 • 2 I 7 • • * • 2 141 47.498 366 75 111 • 7 1 ' 6 1 3 6 592 47.501 454 165 bO 2 « 13 3 1 : 7 28 772 47.504 628 59 91 1 7 20 4 1 • • ^ ! 44 4 28 895 47.507 305 77 33 • 6 6 3 9 3 1 3: > 7 26 518 47.509 9a6 65 186 2 2 26 5 • 2 8 ' 10 35 1331 47.511 346 2 4 ■ 11 1 1 • • i 1 3 372 47.513 80 4 . , « 3 2 1 • • i 2 2 94 47.515 35 11 14 • 7 3 • • « 2 1 80 47.517 34 36 1 • 1 3 • • « 3 4 85 47.520 55 HO • • 3 L 6 • • • 4 44 201 47.523 4 3D • • 7 . L 3 • ■ « 3 8 63 47.525 7 10 . • 6 L 2 ) 1 • ► 13 52 47.t>2/ 550 liL • • ■J > 3 75: i 1 5 2^ 580 52 12 2118 1216 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 2. — Myctophid larvae, tabulated by genus or species, for all stations occupied on the second multi- vessel EASTROPAC survey (EASTROPAC II). w m z z o t— I < H o c bJ) o a a S a a Q. o s: -4-1 c a a 3 3 H 3 O s 3 3 a 3 C IS X o 0! D. s 3 a o ho >. X 3 x; a o b£l >. s a a I a o -a "a, 0) c o s a 2 o s a o o >> u Qi J3 a a a 3 2 O >> m c 0) T3 rt g > a •n > (H 3 O §. Q, o o to o a m a 3 a 3 o >> a 2 o 1 a 3 CO a o >, S a a o 01 o CO 3 01 a t+H o CO '^ ■a jC c a u a T3 to s CO •o 3 a 3 o a ■a o H 45.016 i S 3 x: a o tan >> EC C4l 3 CS .3 > CO B CU ■a rt a > 9- ■n Ol ^ > 3 o _>> "o o Z a CO s 3 x: a o o >, s o 3 o x; a 5^ o e t/3 a a CO XI a 2 u >. a T3 ■T3 XJ o a B XI s X3 a o 4-» O •a & CO 5 00 •o I- u a o H 45.163 ^►5.165 45.169 45, 45, 45, 45, 45, 171 173 175 177 179 45.183 45.187 45.191 45. 194 45.198 45.202 45.206 45.283 45.287 45.289 45.293 45.297 45.301 45 45 45 45 ,305 ,309 ,313 .316 45.319 45.321 45. 45. 45. 45. 45. 45. 323 325 329 331 333 335 45.337 45.339 45.341 45.343 45.344 45, 45, 45, 45, 45, 45, 45, 45, 45. 45. 45. 45. 45. 346 348 350 352 356 358 360 362 365 367 369 371 373 32 23 39 14 244 115 27 116 10 4 26 68 127 163 6 15 10 33 10 14 19 32 96 7 49 218 84 47 37 9 17 6 3 6 12 39 17 40 46 9 1 5 53 32 59 1 23 6 7 5 I 12 43 12 5 8 13 3 35 24 10 56 22 74 13 18 5 29 6 29 16 13 12 5 48 43 9 5 5 58 20 2 7 5 4 5 30 9 16 4 15 26 9 15 10 9 25 4 24 49 5 26 39 12 34 50 36 36 15 17 9 11 2H 20 1 4 44 16 3 14 1 5 7 4 1 7 1 1 2 1 12 2 1 6 1 12 4 4 7 2 13 2 1 11 3 1 1 4 12 7 3 5 7 • 5 10 39 1 4 5 1 5 • 5 I 4 1 6 4 2 8 3 8 152 86 56 26 10 62 70 59 26 255 144 49 251 53 22 102 25 157 157 231 24 119 44 49 41 32 45 51 1002 69 235 108 65 49 26 47 17 21 19 30 70 29 69 98 18 36 99 23 101 120 117 1218 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 2. — Myctophid larvae, tabulated by genus or species, for all stations occupied on the second multi- vessel EASTROPAC survey (EASTROPAC II) .—Continued. PQ o I— < < M a. a. 3 o B >, (U s O n. XI B a at v K a a o ■J" !3 h u >, n, (U 4) XI a =1 o •a n! B a S 2 o >> ^_^ a 5 "1 T3 g o > a T3 g: rt 1) t4 m CO CO 3 M a 3 O 8- o CO o o o o 2 Z 3 x: a. 2 o >. a 2 o ^^ o x; & o B tn a a CO T3 t ■T3 CD CD S a o U >, 8 73 d (U ■a §• B 73 (U ■4-> CI! .s m "3 t o 20 42 42 59 37 7 6 12 8 2 22 2 63 85 13 19 • 1 1 7 2 2 34 67 199 98 40 19 26 46 7 2 1 • 26 31 1 1 114 75 38 60 150 53 343 32 80 5 117 8 357 20 12 1 19 2 31 1 69 3 51 1 50 1 163 13 207 21 126 23 65 15 366 241 133 2 50 71 36 14 27 189 84 164 73 46 64 21 34 19 14 30 6 285 28 35 2 12 > • • 2 14 94 I 103 27 1 30 9 10 16 17 33 60 55 89 17 18 23 25 163 1 285 76 1 » • 160 38 64 206 30 L 2 . 3 3 L 1 412 85 9 6 > . 1] 3< 5 5 5 . 14 L 1 3 L 1 152 436 20 L 2 25 2 3 . 4 L : > I 62 6 J 4 89 16 I 3 78 5 3 1 » • 63 8 1' I . 4 « i: > i 3 1 9 2 8 250 5 2 ) . i( 1 1 3 I z: J i: 3 : > 1 9 • . 11 319 • > L 3 5 1 I i< i L . 15 • 10 211 • I 1 2 1 L ( > • 3 100 4 1 I 3 r r . r ? 2 8 3 L ) r 3 . 8 22 3 751 • ) 11 3 . 5 K 3 3 ( 3 1< 3 • 3 17 3 20 4 94 • • > • • 1 I L ; > , , 1 25 148 • • • • • • t ♦ • A 3 48 7 > 18 1' * 1 • 5< b 2' t • 1 3 2 404 3 7 • • ? 3 I 1 3 3 269 2 1 ? • * 4 • 116 2 » £ • I . L • 1 10 72 • 1 1 L • a « • 35 • » • • • 4 • 36 4 2 2 1 3 . 3: 1 . 1 13 ] 2 408 1 < 1 4 ♦ ■ } . . , 5 2 61 1219 FISHERY BULLETIN: VOL. 70, NO. 4 Appendix Table 2. — Myctophid larvae, tabulated by genus or species, for all stations occupied on the second multi- vessel EASTROPAC survey (EASTROPAC II) .—Co7itinued, w a D § CO O "S P bo O a a en CO a I a E M u a a M CO 3 x: CO _3 "3 u 3 C B CO >. XI s o c 8 s 1 C3 >< o c« a a S 1 3 C J2 o. a o o bn tuD >, >> as S a a CO c 0) 13 a a S ►J o SI o 'O a. CO c XI a, o a CSI (U I a .3 a 1-1 > ^^ a. CO s a o *.» o >> s o o 3 o j3 a o "o S >, CO a a a S o CO T3 en "3 >i JS S- s % H o >1 T3 >. F (U a ■o "S ^ e OS 0) S 0 •a jc: CO 5 :^ Q J3 a u >> a o 46.0 84 46.086 46.088 46.090 46.092 46.094 46.096 46 46 46 46 46 46 .098 ,100 ,102 ,104 ,106 .108 46.110 46.112 46.114 46.116 46.118 46.120 46.122 46.124 46.126 46.128 46.130 46.132 46.134 46 46 46 46 46 46 .135 .137 .139 .141 .143 .145 46.147 46. 149 46. 46. 46. 46. 46. 46. 46. 46. 46. 151 153 155 157 159 161 163 165 167 46.169 46.171 46. 46. 46. 46. 46. 46. 46. 173 175 177 179 181 183 185 23 68 69 72 88 196 144 50 105 273 70 110 241 151 53 0 12 9 50 19 11 1 9 3 2 201 160 435 112 269 480 1328 1358 2505 939 474 59 268 214 48 131 89 182 27 44 105 182 20 50 178 9 20 23 10 4 25 37 27 15 5 8 10 13 14 1 10 4 5 49 20 3 1 17 28 7 25 34 8 1 4 8 14 5 9 36 7 81 16 16 5 111 56 11 19 46 97 37 22 3 11 3 3 3 3 5 23 3 2 3 2 1 13 4 3 1 2 1 1 2 2 4 5 3 30 2 1 1 5 9 8 25 6 3 4 10 15 1 1 I 4 1 39 12 6 3 3 5 4 16 1 2 3 1 1 6 1 3 13 14 4 6 4 11 5 1 3 1 2 1 6 4 11 5 1 2 4 11 45 102 124 99 100 300 240 73 125 334 109 115 250 162 562 31 14 176 80 17 19 63 113 46 26 9 81 201 175 438 113 282 484 1333 1358 2561 973 490 71 350 290 90 195 163 255 33 52 124 213 52 85 262 1220 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 2. — Myctophid larvae, tabulated by genus or species, for all stations occupied on the second multi- vessel EASTROFAC survey (EASTROPAC II) .—Continued. m •A s t CO o c O >1 c CO 01 O n. J3 e C nl > K 3 a o bo >, •s. a a CO c 0) T3 ol a E o! a CO a) c c3 JS a o a > s 3 a 2 o >> % a a S 3 a S u >1 m E! 0) T3 C > a ■n ol > Sh a CO 3 O -C D. O "o Si s >> w a a •a j3 (0 1 "2 % t^ >> CO T3 >. B T3 ^ S T3 J2 a C -a t3 S 2 «4H s S 0) T3 • 97 47.005 161 33 9 33 1 19 6 • 17 1 • 287 47.008 205 31 5 4 2 4 3 1 » • 257 47.011 117 22 13 2 3 3 1 • 161 47.019 287 37 2 39 I 2 2 » • 370 47.022 148 11 • 2 2 5 169 47.025 73 5 1 2 • 83 47.028 175 66 5 I 7 9 1 264 47.032 148 38 12 3 6 1 • 208 47.034 95 23 1 7 4 6 • 139 47.035 171 21 3 9 12 ) 1 1 223 47.040 129 38 11 3 10 6 • 6 205 47.049 42 9 1 • < 4 3 2 1 64 47.053 105 89 1 4 19 18 11 ! 3 9 265 47.057 192 50 I 23 1 6 12 . 11 300 47.061 22 12 3 2 4 23 1 9 » • 79 47.065 31 22 1 3 2 12 • 22 » • 99 47.069 29 6 2 3 35 . 2! 3 > 5 . 30 140 47.070 50 11 . I 17 , 10 t • 90 47.074 98 4 L 1 . 26 . L 5 2 138 47.078 5 • • • . 3 2 » • 11 47.082 70 10 11 6 . 15 119 47.086 212 10 6 18 1 248 47.090 33 17 2 2 • 57 47.094 96 * • 3 • K • 100 47.097 9 • • < 1 1 • 16 47.099 6 7 • » • 16 47.101 39 6 L • > • 48 47.103 • 1 • K • 2 47.105 • 4 • » • 4 47.107 30 70 6 • 109 47.109 I 4 • » • 6 47.113 • • • • • 0 47.124 • 7 3 » • 10 47.128 ■ 31 4 • 35 47.132 15 53 ! 12 » • 85 47.134 5 10 L 1 1 21 47.137 4 4 1 2 : J 1 17 47.139 3 7 1 ) 1 > • 16 47.141 12 5 2 1 1 24 47.143 19 2 » • • » 1 • 23 47.145 • 2 . 6 2 1 2 15 47.147 2 2 . 16 ^ 1 > 8 ; ) 1 44 47.149 12 7 . 10 . 6 3 40 47.151 6 1 5 3 L . 4 • 22 47.153 6 3 I 1 L L , 3 3 30 1221 FISHERY BULLETIN: VOL. 70, NO. 4 Appendix Table 2. -Myctophid larvae, tabulated by genus or species, for all stations occupied on the second multi- vessel EASTROPAC survey (EASTROPAC U) .—Co7iti7iued. PQ Q D. a. B o >> a 6 Hi J S- Q. to to 3 to 3 & 3 O B 3 3 a 0! 4.1 u ' ca a. s 3 J3 a a >> a 3 j3 a o bi] >> X o. a 0] rt c (U T3 rt a S ►J 10 "o J2 O to ;.! >> c ctj x; a o a > x; to a (0 ■a d a 01 > a -n c3 > u 3 o _>> "o *.» o Z '^ to a 3 3 u o a _o "o a >> en a a to to 3 ;4 2 o a to a. a to •a >, JS 9* a -t Q. 2 o >> ID TJ >> a 2 fc: T3 J3 4-> c ■s ^ 2 ■a .a 1" s o A7. ^7. U7, ^7, ^7. ^7. ^^7. ',7. 47. ^^7. A7. 47. 47. 47. 47. 155 157 159 162 164 166 168 171 173 175 177 179 181 185 189 193 47.197 47.201 47.205 47.213 47.217 47. 47. 47. 47. 47. 47, 47. 47. 47. 47. 47, 47, 47. 47, 47, 47, 47. 221 225 229 233 237 240 242 244 246 250 2 54 258 268 2 72 2 76 278 280 47.283 47.286 47.288 47 47 47 47 47 47 47 47 47 47 47 .290 .292 .295 .297 .301 .304 .306 .310 .314 .318 ,322 72 9 • 9 11 4 50 179 265 60 190 126 281 20 2 56 148 51 595 360 425 416 265 1206 86 182 931 837 103 295 224 11 55 21 1904 45 170 177 1260 1297 172 266 62 7 432 382 56 241 552 145 49 111 8 4 2 7 3 12 35 15 13 4 6 16 75 28 121 36 6 54 145 146 5 6 16 58 6 1 3 19 402 17 8 3 7 10 10 6 4 28 51 25 16 16 116 62 8 11 11 1 5 3 2 3 12 2 4 4 2 1 1 17 67 13 1 16 14 19 5 15 30 23 5 4 1 4 3 8 12 7 4 19 44 18 4 2 4 9 13 4 2 13 2 28 89 19 6 4 6 8 10 1 4 2 5 3 2 6 19 1 13 17 6 4 11 11 1 3 20 33 16 27 10 7 13 19 6 I 4 5 7 1 3 30 37 27 21 12 34 4 10 38 3 9 6 3 6 4 2 1 16 6 12 3 5 3 7 16 5 0 109 17 11 25 28 6 66 251 327 119 228 150 350 28 2 97 392 109 769 440 452 512 496 1441 100 201 953 905 111 297 245 516 55 21 1935 54 175 184 1272 1311 186 276 702 552 438 92 309 780 241 73 152 1222 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 2. — Myctophid larvae, tabulated by genus or species, for all stations occupied on the second multi- vessel EASTROPAC survey (EASTROPAC U) .—Continued. a u m D Z z o < H DD 1 g to o C O ■q a a 10 to B o CIS a S . S Myctophum spp. (other) Notolychnus valdiviae Notoscooelus resolendens Protomyctophum sp. Symbolophorus evermannl Triphoturus spp. Other Identifiecf myctophids Unidentified myctophids Disintegrated myctophids Total myctophids 47.326 60 42 u 3 : i . . . 1 2 52 5 . 20 , 11 2 217 'tl.^^0 193 53 13 20 72 3C , 2 51 2 436 47.334 92 10 6 2 19 13 . 4 16 1 165 47.338 42 15 • 1 2 • 16 78 47.342 227 8 • 2 57 1 5 302 47.345 813 15 • 1 12 2 8 1 856 47.349 271 2 1 2 5 3 288 47.351 bbZ 7 1 5 5 H • 684 47.354 1 66 • • , • 1 ' 71 47.357 179 21 • 8 12 1 14 3 239 47.359 60 3 1 2 • 68 47.362 69 2 4 • 1 81 47.364 76 7 • 2 ' 96 47.367 71 12 • 1 4 97 47.369 11 • • 1 • 13 47.371 31 4 • 1 3 49 47.373 155 15 4' . 23 • 1 £. L > 260 47.376 32 3 1 4 • 52 47.379 3 I • . • . 1 7 47.382 5 2 4 5 6 1 7 I 35 47.415 6 4 1 • 2 1 7 4 30 47.430 3 23 I 6 • 4 2 35 8 82 47.432 • 34 2^ 12 ? 14 • 1 1 29 > 10 '. 142 47.436 5 1 , . • 1 1 L 10 47.438 31 49 • 14 100 47.440 10 9 • 10 I 34 47.443 54 1 • • • 56 47.446 139 1 4 « 2 2 152 47.450 500 19 6 3 . 17 11 6 570 47.454 153 25 2 1 4 S [ 200 47.458 10 7 1 « , C ? 32 47.462 • • • • » • 0 47.466 60 5 1 4 . L < 84 47.470 39 16 • 14 . t 1 81 47.478 19 10 . 1 • 1 , 1 1 38 47.486 150 12 1 1 1 1 ) 2 t 9 194 47.490 199 • 2 , . 2 2 ) ' , ( 12 238 47.494 63 4 . • • ) L 1 4 77 47.498 296 16 2 2 4 . i: ) 1 3 1. 7 366 47.501 401 33 3 2 L 6 454 47.504 568 15 1 • ) L 21 628 47.507 272 23 4 • 3 305 47.509 955 24 3 1 L 2 986 47.511 344 2 • . 346 47.513 80 • • . 80 47.515 34 1 • . 35 47.517 32 2 • , 34 47.520 26 9 U • . 55 47.523 • 4 • 4 47.525 1 1 • . 7 47.527 . • 52' 2' • • 1 550 1223 FISHERY BULLETIN: VOL. 70, NO. 4 Appendix Table 3. — Counts of selected categories of fish larvae, tabulated by station, EASTROPAC II. pa ^ S CO D Clj 2 t4 Z ^H O 3 E- < .s Kl > a. s o ■s "o >> u 3 CO d O cU b£ c O 0) !^ CO 3 0) s -J Z 01 l4 cU ?n .^ (U « 3 CO -I B 01 73 ffl -1 a -C CD a ca o 5 y m ffi T3 T3 T3 Li lU XI B O .ij CO a m CO 3 •.;^ l:^ 9- a u CO CO 0) a 3 B 4^ a CO CO 3 CO CO 3 .S CO 3 3 >1 3 Q 3 » a o "a J3 ca S •3 O h (U (U t< CD O Z Q H H CO 0) ^ o bc ffl ta o u ■a o (D >> ■o 0) C« o CO .i3 o ■2 13 tH Cl! CO o H m o H ^5.016 45.020 45.021 45.023 45.024 45.026 45.028 45.030 45.032 45.034 45.035 45.017 45.039 45.041 45.043 45.044 45.?46 45.048 45.0 50 45.051 45.0 53 45.054 45..056 45.058 45.060 45.063 45.065 45.067 45.071 45.073 45.078 45.0d3 45.086 45.090 45.094 45.09R 45.102 45.106 45.110 45.114 45.117 45.121 45.125 45.127 45.129 45.131 45.133 45.135 45.137 45.139 45.140 12 77 5 15 95 27 21 54 10 8 35 27 24 77 49 16 3ft 56 21 2 B 7 2 4 1 1 1 43 366 560 39 65 105 26 77 65 100 15 78 21 27 15 43 91 301 3 8 2 32 5 3 6 1 12 5 7 24 12 3. J 10 1 3,-. 4 9 7 1 2 10 1 3 5 4 2 1 4 3 K t 2 1 5 12 i 1 4 2 7 12 39 1 1 10 1 3 2 4 24 4 1 1 3 28 1 1 2 2 1 14 87 7 25 110 100 38 36 61 129 39 34 35 90 58 23 45 59 23 4 19 22 4 8 24 12 58 397 685 47 73 110 30 81 84 116 18 86 28 47 22 52 104 327 41 3 339 36 18 0 2 52 1224 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 3. — Counts of selected categories of fish larvae, tabulated by station, EASTROPAC II. — Continued. Pi n S a. Q, rt ^^ 01 u c .a >> > u fe a d B o to a J3 to n, c< o J3 S o m s ^ g o e« S 0) V rt OJ •n n •o ni h o rt 3 m h al lU a! CO ki n O lU o o t< ^ a o o s u CO ^ to Q 3 s t^ 3 to .s 3 & o o rt a IS S to 3 ti D. 3 f^ a (l; to 3 !.< -^ 3 3 s J3 ca o h u H H b •2 -0 a _2 1 t> 1 2 o to o H 45.163 4b. 165 45.167 45.169 45.171 45. 173 45.175 45. 177 45.179 45.183 45.187 45.191 45. 194 45.198 4 5 45 45 45 45 45 ,202 I? 06 ,283 ,287 ,289 ,293 45.297 4 5.301 45.3 05 45.309 45.313 45.316 45.319 45.321 45.323 45.325 45.329 45.331 4 5.333 45.335 45.337 45.339 45.341 45.343 45.344 45.346 45.34B 45.350 45.352 45.356 45.358 45.360 45.362 45.365 4 5.3 57 45.369 45.371 45.373 33 3 5 20 141 32 13 13 10 4 16 15 1 R 79 114 d5 103 2 4 1035 148 34 3 1 45 9 2 1 7 2 27 2 7 12 5 69 32 3 4 1'^ 12C 296 65 147 55 y r" 1('3 231' 146 11 4 16 1 4 2 3 1 i 5 6 1. 7 3 13 1-. 4 15 3 1 4 1 3 1 13 8 1 5 1 1 2 0 66 6 15 1 0 28 151 33 20 33 20 7 32 24 2 17 84 127 67 129 41 1046 158 38 38 48 15 9 72 8 20 47 51 14 18 74 36 45 21 123 298 69 165 109 45 37 56 17 108 253 164 1225 FISHERY BULLETIN: VOL. 70, NO. 4 Appendix Table 3. — Counts of selected categories of fish larvae, tabulated by station, EASTROPAC II. — Continued. (4 H n 2 s Eh CO tn s- bl d m B u u CO c o s 3, .c rt o C) x > 03 > U ffl be O 3 (1) ►J .3 a a 2 o j3 a a 13 1 (-1 CO o o o u o u cu Hi T3 o i3 •o -1 a XI m D nl o 1 03 § o o. o ■a S o O -J nl eft 3 3 S a 0) o. a -a a o "o OS is at O o H ^5.375 45.377 A5.379 45.3S1 ^►5.383 45.335 45.387 39 7^ 797 Ih 6C 4?8 11 7 43 95 812 253 26 61 433 Ci^UlSE ^6 46.002 46.004 46.006 46.007 46.009 46.011 46.013 46.015 46.017 46.019 46.020 46.022 46.024 46.02 6 46.02 8 46.030 46.032 46.034 46.036 46.038 46.040 46.042 46.044 46.046 46.048 46.050 46.052 46.054 46.055 46.C57 46.059 46.061 46.063 46.065 46.067 46.069 40.071 46.075 46.077 46.079 46.082 18 116 Id iH 22 69 47 16 4 147 7 8 14 21 ni 154 7 35 9 1 4 19 129 9 5 56 53 175 ?06 177 66 610 480 8 5 16 76 104 52 26 20 17 2 3 10 36 3 1 4 1 1 36 2 3 / In 13 1 1 1 1 5 1 10 1 5 5 12 e 7 3 4 1 1 10 12 7 10 2 3 2 4 5 22 14 5 1 23 1 7 36 14 10 21 130 21 28 24 75 57 19 11 154 92 16 33 145 171 16 1 49 136 10 23 137 98 61 67 201 235 202 78 665 526 116 21 103 129 82 45 26 23 65 74 1226 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 3. — Counts of selected categories of fish larvae, tabulated by station, EASTROPAC II. — Continued. i i CO o 0) .s > d §■ o a o j3 O > to 1 i a to 1 > a 'tb. 084 17 1 21 46. u86 66 7 33 46. 0H8 27 8 46. 090 17 4 46. 092 13 3 46. 094 145 i 46. 096 226 21 13 46. 09« 4C 2 46. 100 6 2 46. 102 46 3 46. 104 46 . 46. 106 34 9 46. 108 36 27 46. 110 10 5 46. 112 75 16 46. 114 6 3 46. 116 5 • 46. 118 65 3 46. 120 100 1 46. 122 29 4 46. 124 22 2 46. 126 98 10 46. 128 130 8 46. 130 53 11 46. 132 26 5 46. 134 12 8 46. 135 12 8 46. 137 2a 24 46. 139 50 11 46. 141 20 18 46. 143 14 10 46. 145 ua 38 46. 147 41 41 46. 149 11 22 46. 151 4 13 46. 153 163 4 46. 155 132 6 46. 157 24 6 46. 159 IC ? 46. 161 3 t 12 h 46. 163 ec 10 46. 165 2i lb 46. 167 25 37 46. 169 14'= 52 46. 171 13 4'. 46. 173 1 9 46. 175 1 21 46, 177 2C 4'5 46. 179 o2 135 46. 181 51 132 46. 183 1 47 46. 185 9C 45 o 3 cfii 3 o, 2 o Q a H o o o > -«^ x: o 1) ■O 0) ta T3 CO "1 S' J3 to Q CD 1 m K 01 Li CD §^ U O as §■ u en m o o m u S o 3 oi 1^ fO 01 3 c ;2 u & en m 3 g 3 S CO 1 D V (0 3 CO CO 3 U 3 g ;> a Q _g ^n n o o a JS c« S •a o i3 V o Z Q H H cd T3 i cH 0) CO ■a cd P3 c« ■a i o '3 cd is « E o •o B u I— I cu 5 o 2 19 1 169 2 24 3 . 11 2 ? A 7 10 85 41 14 7a 5 24 67 274 72 29 36 163 270 52 10 67 51 52 69 15 93 10 5 89 103 33 24 109 142 66 35 20 20 54 64 41 27 157 83 37 22 169 156 40 15 64 103 56 82 290 103 25 96 80 247 191 78 148 1227 FISHERY BULLETIN: VOL. 70, NO. 4 Appendix Table 3. — Counts of selected categories of fish larvae, tabulated by station, EASTROPAC II. — Continued. a m 8 s CO Q. a. a. V a o a a. S o XI Q o > J3 a o b ■a to 3 g- n CO D ol n fi a PQ 55 m a cU Tl •a Sh o 3 (0 0! lU «: h O (1) a 1 o o g. Q a a CO CO 3 Sh 3 C §4 u o bo 3 1 ^ c o 1) CO o nl is X3 CO ■a h 01 CO o H m o H • • • 60 • • • 15 A6.187 ^6.189 39 5 17 3 CRUISE A7 '(7.001 47.005 47.008 47.011 47.019 47.022 47.025 47.028 47.032 47.034 47.035 47.040 47.049 47.053 47.057 47.061 47.065 47.069 47.070 47.074 47.078 47.082 47.086 47.090 47.094 47.097 47.099 47.101 47.103 47.105 47.107 47.109 47.113 47.124 47.128 47.132 47.134 47.137 47.139 47.141 47.143 47.145 47.147 47.149 47.151 47.153 61 203 49 26 221 27 58 117 137 36 70 76 92 94 106 18 91 49 17 14 4 6 11 5 2 8 8 9 11 16 2 14 76 27 3 9 13 2 1 4 4 6 7 3 5 2 4 14 15 7 20 8 26 27 50 U 26 21 4 122 12 176 7 294 61 79 527 77 7 2 1 3 5 13 16 5 lf> 15 4 17 7 14 5 3 1 10 3 2 5 59 55 6 1 19 4 1 7 3 8 47 16 1 1 1 11 5 6 18 13 14 21 20 2 79 219 62 30 258 45 63 133 159 57 106 101 145 205 271 55 122 207 233 337 71 86 541 89 11 1 3 3 3 11 11 1 0 0 0 15 28 14 21 34 22 19 97 48 13 35 1228 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 3.— Counts of selected categories of fish larvae tabulated by station, EASTROPAC U.— Continued. n 15 a a. m c (U CO o s 5 o! o > j3 > cj o m hi O U 3 0 & 0 CO T3 o o CO a a tn O > c i c CO a a m u o 3 .a c d S to *^ CO o J3 a 'q o o ca T3 <0 a o h n 3 g- j= CD a CJ o a cU H W m CS ■o a u > <0 ■a a o o CO ■3 a CO I a a a Cli ■a j3 O !c o cc! H ClJ •a ccl CI] o cd i! CO o u ■d o ^.7.326 47.330 '•7.33'^ 47.338 47.342 47.345 47.349 47.351 47.354 47.357 47.359 47.362 47.364 47.367 47.369 47.371 47.373 47.376 47.379 47.382 47.415 47.430 47.432 47.436 47.438 47.440 47.443 47.446 47.450 47.454 47.458 47.462 47.466 47.470 47.478 47.486 47.490 47.494 47.493 47.501 47.504 47.507 47.509 47.511 47.513 47.515 47.517 47.520 47.523 47.525 47.527 33 129 28 19 24 44 61 66 e7 66 56 4 7 22 3 7 237 72 55 103 7 108 69 2 34 107 17 lb 21 5 6 3 26 30 14 10 65 148 49 75 64 2 3 9 36 PO 3 0 10 13? 143 123 45 19 17 26 16 60 14 I'J 3 3 11 3 ii 2 9 2 1 1 2 23 41 4/ 7 6 6 11 13 2 1 13 2 'J 6 26 11 9 14 3 5 2 6 11 2 107 9 S 7 1 IC 8 6 1 5 1 2 10 2 1 5 43 16 4 2 22 15 7 6 7 37 1 26 3 2 2 16 236 311 89 51 54 77 79 127 87 85 69 7 10 38 8 17 259 85 66 111 11 114 103 29 64 129 26 14 149 68 52 0 42 119 45 53 37 12 78 184 79 85 93 13 4 18 37 83 37 16 135 1230 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 4. — Summary of occurrences and numbers of larvae of 23 categories, limited in distribution to a broad coastal band or around offshore islands or banks. a u a w nl T3 5 % CO 3 CD .9 a 3 O hi a) a S Cj rt 0) e M a rt 0) CO ■a c 1 1 ■§ M ■a ^ o Q ta c Ol M U m to 3 D. H a , •o !3 CO §• (1) 73 a u 1 to J3 o a § CO Li Si d a Q * j3 o ••-t >: >: W O O M VI Q) 3 TI 2 V . s o O ft ft CU o % a o u en lU cH T3 t4 U eel T3 ■§• cd O I o 45.339 45.343 45.350 45.358 45.360 45.362 45.365 45.367 45.369 45.371 45.373 45.375 45.377 45. 379 45.3iil c3 ?1 5 23 1 1 II 2 1 1 2 10 2 19 1 1 1 1 16 90 13 53 7 8 2 II I 2 ;ruise 46 46.002 46.004 46.006 46.007 46.009 46.011 46.042 46.052 46.U59 46.0 82 46.108 46.110 46.112 46.114 46.116 46.118 46.120 46. 122 46.124 46.126 46.128 46.130 46.132 46. 1 34 46.135 46.137 12 46.139 46.141 46.147 46.149 46.151 46.153 46.155 46.157 46.163 46.165 4 22 2 3 13 1 1 1 10 11 1 13 132 2 14 1 6 6 76 3 2 1 2 45 2 45 48 1 I 18 23 32 1 3 2 I 2 I 1 2 2 1 3 1 1 3 12 45 5 1 1 8 5 4 15 3 2 4 91 143 276 7 4 4 I 16 1 1 1 3 1231 FISHERY BULLETIN: VOL. 70, NO. 4 Appendix Table 4. -Summary of occurrences and numbers of larvae of 23 categories, limited in distribution to a broad coastal band or around offshore islands or hanks.— Co7itinued. a. w m S D 2 Z O < CO J3 o O 3 03 fl s D. Si 3 hi 0> o rt cJ Q 0) P CO 3 > M CD CO 3 C a 0) o o cU E J= .3 3 3 U3 T3 O ^ ^ cfl < u Ol M 0) 3 .3 a CO ■a CO a <1) ci 0 cS > m ca a rt > 3 8 > j3 1i) c ' ' CO F CO ^ Q 3 3 O o c JS j: O o cS >1 CQ O O M a a Li 3 Q. 3 a P f o CO CO o ■o Cli .3 Li Li o ^ 0) 3 ci! U w -1 3 cd •o E 0) m v to 3 c Q V o o m B J3 .3 -1 3 ffi ■a JO t, ki ■*-» Gj < w O 03 ;h (U 3 o CD cj •a •a 3 a nl •a ^ o e >> w CO CO 3 d h 9- , T3 (0 a CO o u CO 9< cc! > O a :3 CO 3 3 ^ XI "o Si D a o > >i > W U U do w CO 3 5J a ?1 o oj o ■a fa .3 !m h o J2 cd 3 td U u ►J 01 •a bO 3 zi •a a _>. a. •ff CU CS CD T3 •a •a 73 U oj c Oi a n! y* a o Si ,v o CD u< CO CO CO O 5 o H ^7. 179 ^7.181 ^7.185 ^►7.189 ^.7.197 ^7.2C1 47.229 47.^33 47.237 47.240 47.242 47.244 47.246 47.250 47.254 47.2b8 47.268 47.272 47.276 47.278 47.283 47.286 47.288 47.290 47.292 47.295 47.301 47.304 47.306 47.310 47.314 47.322 47.330 47.334 47.349 47.357 47.436 47.446 47.450 47.466 47.470 47.501 47.504 47.507 4 7.5 09 47.511 47.513 47.515 47.517 47.520 47.523 47.525 47.527 1 13 1 83 10 1 1 23 1 36 2 13 1 3 1 180 142 69 36 5 2 15 13 2 1 1 1 10 19 2 11 2 7 6 5 9 7 5 1 3 8 84 21 13 17 3 3 9 3 1 1 1 1 2 5 5 2 1 1 1 1 1 1 1 1 3 2 1 2 5 5 2 1 1 3 43 7 9 568 1233 FISHERY BLLLETIN: VOL. 70, NO. 4 Appendix Table 5. — Numbers and kinds of eel leptocephali (Anguilliformes) obtained on the second multivessel EASTROPAC survey (EASTROPAC II), tabulated by family for all positive hauls. S H Eel leptocephali Congridae Moringuidae Muraenidae Nemichthyidae Ophichthidae Serrivomeridae Xenocongridae 3 Eel leptocephali Congridae Moringuidae Muraenidae Nemichthyidae Ophichthidae Serrivomeridae Xenocongridae o 45 .018 2 2 • • 46 .157 I 1 • 45 .023 1 L • • 46 .161 1 » • • I 45 .030 1 • I 46 .165 1 1 45 .058 1 • I 46 .169 1 1 45 .063 1 • I 46 .177 1 45 .065 2 • I ? 47 .001 3 45 .067 2 • i I 47 .008 1 45 .071 1 • L 47 .011 3 2 45 .073 1 • I 47 .019 11 4 . , 45 .140 2 • I 47 .028 2 1 45 .313 1 • 47 .032 1 1 45 .362 1 1 • 47 .035 1 45 .365 12 ^ 7 47 .040 6 3 45 .367 3 • I 47 .049 2 2 45 .371 1 1 47 .053 2 45 .379 1 « 47 .069 46 .032 1 • 47 .070 46 .034 1 • L 47 .107 46 .046 2 • * I 47 .179 46 .065 1 ] • 47 .197 46 .077 1 ] • 47 .217 46 .086 2 L • 4 47 .229 46 .098 1 • t L 47 .233 1 46 .106 1 L • 47 .242 1 46 .108 1 L • « 47 .244 46 .110 2 1 47 .254 46 .112 1 L . t 47 .258 46 .114 1 . , 47 .268 46 .120 1 I 47 .276 46 .122 1 1 . ^ 47 .283 46 .124 2 2 47 .286 I . \ 46 .128 1 1 . < 47 .297 • • 4 46 .130 2 -, . , 47 .304 2 1 46 .134 2 2 47 .314 3 2 1 46 .135 3 3 47 .438 2 2 46 .137 1 1 • * 47 .478 46 .139 4 ■, 2 47 .504 46 .143 1 1 47 .509 46 .145 2 1 • 47 .525 46 .153 1 L • 1 47 .527 5 46 .155 I L • * 1234 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 6. — Numbers and kinds of lophiiform larvae obtained on the second multivessel EASTROPAC survey (EASTROPAC II), tabulated by family for all positive hauls. •i iz; I ^►5 (^5 a, o ■B ►J 0) a o o o 0) ■3 o ^ a> O ■a ig a 0 03 01 .4-3 s o 1 o O a) i5 3 ^ g- till pa u w M S 5 '3 1 1 1 1 1 1 1 1 Q & ^ ^ •g 0) o ni ^ ^ ^ B W PQ n! ■g a ^ :s ^ 1 'S c 3 ■■ a o Z 31 £ o a ►J o H t4 6 t l-l ■§ 3 .4-1 c 3 d D > U be a Q CO 3 o H OP .001 8 • " . 1 • • • 1 • • 17 OP .002 2 • « « 200 • ■ • • * • • • • 202 OP .003 4 * • • 74 • « • • • • 1 • 79 OP .005 32 9 3 . I 2 • 3 • • 2 52 OP .007 50 19 11 1 • 2 3 • • 2 88 OP .009 41 9 12 1 18 8 1 • • • 91 OP .011 50 19 6 3 4 2 7 5 1 9 106 OP .013 70 22 20 11 3 2 3 1 • • 132 OP .015 46 15 24 5 1 1 • • 2 2 98 OP .017 13 1 1 5 2 2 1 2 • 2 29 OP .019 20 10 8 16 1 3 • 3 • 64 OP .021 10 7 4 6 2 4 • 1 2 36 OP .023 251 53 15 61 5 • t • • 2 393 OP .025 206 31 8 19 3 3 1 • 1 275 OP .027 248 110 111 41 ) ! 9 13 1 « 30 589 OP .029 168 146 153 26 If i 3' 5 12 2 1 56 623 OP .032 321 154 213 1 54 i 1" J 7< ! 16 7 1 1 12 886 OP .036 543 2047 487 1 36 If 3 5" 12 10 2 • 25 323a OP .040 113 126 176 6 1' 3 . 11 1 • 8 468 OP .044 100 64 38 5 I. 5 7 , 17 • 1 265 OP .048 2 30 103 119 47 2' . 5' ♦ 2C ) 13 4 . 18 • 97 732 OP .052 101 52 51 12 b 3' t 3 3 L 12 . 14 5 20 315 OP .056 62 20 108 4 1 > 4( 3 2( ) 10 L 3 5 5 30 321 OP .060 43 26 65 8 ] > 7 2 I 1 2 2 3 185 OP .064 387 51 96 17 • « r 1 2< 3 2 17 2 L 3 7 • 10 655 OP .068 119 40 54 34 i b 1 L 7 6 . . 12 • 3 306 OP .072 109 7 30 13 . 3 I 4 1 4 » . 4 • 4 187 OP .076 86 22 18 a 3 L 6 3 1 3 1 5 180 OP .080 127 67 49 20 t I 8 1 8 1 • 4 307 OP .084 105 12 46 7 . 8 1 3 • 14 222 OP .088 70 28 34 . 29 3 2 3 1 7 221 OP .092 17 6 28 2 1 1 9 • 1 75 OP .096 64 7 6 . 15 ? 1 _ • ft • 2 5 2 124 OP .100 92 4 6 , « ^ • ft • 1 • • 114 OP .104 60 16 2 3 1 ft • 2 3 3 109 OP .108 95 44 31 7 ' b 2 3 2 1 ft • I 4 6 235 OP .112 320 100 75 6 I 1 4 15 1 4 8 • 6 552 OP .116 383 123 43 . 18 . 2 ? 11 3 2 7 1 5 621 OP .120 110 6 6 . 27 5 ft • 3 2 3 165 OP .124 296 3 4 . 67 2 3 L 4 4 • • 386 UP .128 1325 181 20 . 133 3 4 2 5 1 11 1697 OP .132 1454 229 15 . 115 3 3 L 8 7 1 10 1856 OP .136 343 14 4 . 17 • • • 5 3 3 • 12 408 OP .138 411 9 15 . 54 3 5 4 ? 1 • 3 510 OP .155 37? 37 52 . 20 1 5 L 1 ? 3 • 25 522 OP .160 220 152 27 . 11 7 2 5 3 I 5 3 33 • 35 510 OP .162 255 155 6 • • • 4 2 2 . 15 4 3 31 1 5 525 OP .164 19 10 1 3 • 1 • • • 9 1 I 2 • 4 61 OP .166 4 2 1 3 • 2 • • • 2 3 • 1 18 OP .168 131 18 6273 3 47 7 • 2 47 D • . 1 3 704 11 5 4567 1236 AHLSTROM: KIND AND ABUNDANCE OF FISH LARVAE Appendix Table 7B. — Myctophid larvae, tabulated by genus or species, stations occupied by Oceanographer on zig-transect. 2 £ Ri a 1 Q) -4^ 9 -!-> § E 3 1« S ID (1) m 0) 4-> 1 § o o 0) 3 O '3 E 13 3 s g t4 Q g 3 o, a ft m 3 •3 2 > 1 CQ CO s o '3 CO 15 S 5 c t c a! CO a) c a o ft CO o S-i c e 3 ft B 3 s 3 ft > § O "o § S J3 1 o c ^ s i D s 3 J2 a) -4-* O ■♦-> O CO o H a) n o >, > oj rt o ^ o § o o M pq O Q Q O K K J -) ►J l-l kJ fe! fei z. iz; ft (0 s 3 1^ o o 0- o ft ft .94 (0 ^ 3 -a ■s 4-> 1 3 c .a D Q H OP .001 • • • ••o«««^ • • • 1 1 8 OP .002 • • • 2 • • • • 2 OP .003 • • • 2 • • • 2 4 OP .005 2 • 21 1 7 1 32 OP .007 1 8 • 12 L . 16 12 , 50 OP .009 1 8 t • . 15 4 3 4 6 41 UP .011 . 10 • 11 7 13 2 7 50 OP .013 . i ! 3 15 • 18 C 5 . 8 13 5 1 70 OP .015 5 I 2 5 » t > . 19 5 5 2 46 OP .017 3 a 4 • • 1 • 5 13 OP .019 5 1 3 7 3 1 • 20 OP .021 1 • 8 • • 1 • • 10 OP .023 . 188 , 47 1 10 4 • 251 OP .025 . 180 • . 21 , , 5 • • 206 OP .027 . 213 I 6 . 16 2 2 • 8 248 OP .029 . 113 I 4 12 . 10 , L . 2 4 20 168 OP .032 2 229 ' ^ • . 12 J 25 . L . 5 27 • 13 321 OP .036 . 487 . 19 7 24 • 5 543 OP .040 . 65 . 25 . 17 • 2 113 OP .044 1 43 4 f 4? • • 100 OP .048 1 51 . 91 > 47 9 2 16 230 OP .052 2 51 . 27 . 11 2 1 6 101 OP .056 1 18 5 . 15 2 5 14 62 OP .060 3 11 7 . 18 2 2 43 OP .064 5 253 : 65 L 38 t 11 5 387 OP .068 4 24 . 24 > 27 . 1' 3 8 119 OP .072 . 16 . 23 L 47 1 13 109 Op .076 5 24 : 3 . 31 5 8 86 OP .080 5 46 . 32 I 30 5 127 OP .084 . 55 . 15 L 15 . 1 12 105 OP .088 5 26 . 20 L 5 , i t . 4 2 70 OP .092 6 6 2 • 17 OP .096 3 41 5 3 7 64 OP .100 2 77 7 » • , 5 92 OP .104 1 20 . 15 . 14 . 1 7 60 OP .108 4 32 < . 11 I 26 2 ' y • • 4 95 OP .112 4 256 1 . 39 3 1 2 ? . 3 4 320 OP .116 . 347 1 . 16 ( 5 5 3 L . 1 • 383 OP .120 . 97 t , 1 I 3 2 L . 1 1 110 OP .124 . 265 . 14 i 2 6 : I 1 1 2 296 OP .128 11188 < . 88 , < J 1 5 1 9 2 11 132 5 OP .132 11300 . 122 . 1' > 1 2 1 I 4 3 » 1454 OP .136 . 338 4 • • 343 OP .138 . 403 7 • 411 OP .155 . 346 . 18 • 372 OP .160 2 135 . 65 • 15 220 OP .162 . 206 . 45 t • 255 OP .164 . 19 ► • • 19 OP .166 2 1 • 4 OP .168 43 . 86 2 • 131 1237 FISHERY BULLETIN: VOL. 70, NO. 4 Appendix Table 7C. — Counts of selected categories of fish larvae, tabulated by station, for Oceanographer on zig-transect. 2 O H < H CO bf ■a I a O s .2 .2 ^ Q, o a 05 s J3 D O a o J3 £ c ts o > ffl 55 0) "2 (1) I o H s d Q 2? Li a 9 "4 ■a O Q 13 V 0) « ►J Z (U "3 'he c u a U 'S o O Vi I) d o 73 <1> 'c V Cll o o 0) c (4 in ^ Q. ,^ D. , •o Q. ^ ca a o 0) > o 9 ;h 3 X ^ S 1 m G > •2 bi 8, .037 13 47. ON 119 04.0 VI I-IO 1755-OT 198 3.53 45 .206 00 17. 2S 111 56.5 VIH- 28 1203-D 212 3.65 45 .039 13 00. ON 119 03.6 VI I-IO 2400-N 193 3.12 45 .283 00 04. 2N 111 59.3 IX- 2 1321-D 174 2.49 45 .041 12 31. ON 119 03.5 VI I-U 0534-NT 203 3.83 45 .287 00 32. 5N 112 04.5 IX- 2 1926-N 237 4.99 45 .043 11 54. 5N 119 02.7 VI I-ll 1338-D 177 2.74 45 .289 00 53. 8N 112 03.3 IX- 2 OOOl-N 194 4.02 45 .044 U 29. ON 119 00.5 VI I-U 1820-NT 207 3.56 45 .293 01 26. 2N 112 01.4 IX- 3 0521-OT 202 3.81 45 .046 10 51. 7N 119 02.2 VI I-ll 0005-N 216 3.78 45 .297 02 07.9>J 111 58.3 IX- 3 1209-0 207 3.30 45 .048 10 15. 7N 119 03.3 VI 1-12 0601-DT 177 2.67 45 .301 02 47. 3N 112 02.0 IX- 3 18H-NT 217 3.95 45 .050 09 34. 3N 119 03.0 VI 1-12 1446-D 201 3.54 45 .305 03 25. 8N 112 00.3 IX- 3 2335-N 209 3.50 45 .051 09 10. 2N 118 59.5 VI 1-12 2051-N 201 3.31 45 .309 04 08. 6N 111 56.3 IX- 4 0501-OT 222 3.82 45 .053 08 46. 6N 118 58.3 VI 1-13 0105-N 203 3.19 45 .313 04 55. ON 111 53.5 IX- 4 1154-0 192 2.93 45 .054 08 28. 2N 118 57.0 VI 1-13 0716-0 209 3.38 45 .316 05 5 8. 7N 111 57.8 IX- 4 1746-NT 215 3.53 45 .056 07 53. 2N 118 51.3 VI 1-13 1505-D 203 3.28 45 .319 07 01. 8N 112 05.4 IX- 5 0520-0 217 3.35 45 .058 07 24. 8N 118 47.2 VI 1-13 2030-N 191 2.97 45 .321 07 40. 2N 112 06.8 IX- 5 1149-0 194 2.77 45 .050 06 50. ON 118 55.3 VI 11-14 0151-N 211 3.40 45 .323 08 18. ON 112 01.0 IX- 5 1756-NT 194 2.58 45 .063 05 55. 5n 118 56.5 VI 1-14 1053-D 209 3.46 45 .325 09 01. 8N 111 54.8 IX- 5 0007-N 156 2.02 45 .065 05 23. 2N 118 51.2 VI 1-14 1901-NT 224 3.90 45 .329 10 18. 7N HI 44.0 IX- 5 1150-D 202 2.85 45 .067 04 53. 5N 119 06.8 VI 1-14 2345-N 222 3.88 45 .331 10 49. 5N 111 39.5 IX- 6 1805-NT 175 2.45 45 .071 04 06. ON 119 15.3 VI 1-15 0730-0 219 3.61 45 .333 11 27. ON HI 57.0 IX- 6 2341-N 156 2.52 45 .073 03 34. 2N U9 14.7 VI 1-15 1536-0 210 3.79 45 .335 12 05. 7N 112 09.0 IX- 7 0605-DT 215 3.95 45 .078 02 48. 5N 119 11.0 VI 1-15 0001-N 201 3.71 45 .337 12 27. BN 111 36.1 IX- 7 1145-0 207 3.17 45 .083 02 02. ON 119 03.2 VI 1-16 0640-DT 213 3.64 45 .339 12 58. 4N HI 14.3 IX- 7 1757-V(T 216 3.11 45 .086 01 33. 3N 119 02.2 VI 1-16 1229-D 212 4.18 45 .341 13 33. 4N 110 50.0 IX- 7 0005-N 187 3.01 45 .090 00 57. 2N 119 04.5 VI 1-16 1901-NT 209 4.15 45 .343 13 56. 3N 110 34.0 IX- 8 0805-3 193 2.79 45 .094 00 23. 5N 119 01.5 VI 1-17 0051-N 201 3.81 45 .344 14 12. 5N 110 23.0 IX- 8 H46-D 144 1.80 45 .098 00 13. 5S 118 58.0 VI 1-17 0740-0 205 3.50 45 .346 14 38. 2N 109 37.1 IX- B 1841-N 135 1.71 45 .102 00 56. OS 118 51.0 VI 1-17 1253-D 171 1.96 45 .348 14 57. 4N 109 04.3 IX- 8 2345-N 199 3.15 45 .106 01 42. 2S 118 43.8 VI 1-17 1901-NT 202 3.02 45 .350 15 16. ON 108 37.5 IX- 9 0535-DT 206 3.22 45 .110 02 22. OS 118 43.5 VI 1-17 0036-N 196 4.02 45 .352 15 45. ON 108 08.5 IX- 9 1141-0 217 3". 55 45 .114 03 12. OS 118 43.2 VI 1-18 0800-0 212 3.26 45 .356 16 53. ON 107 03.0 IX- 9 2334-N 186 2.87 45 .117 03 38. OS 118 48.5 VI 1-18 1237-0 222 3.52 45 .358 17 20.5NI 106 29.8 IX- 10 0535-DT 208 3.40 45 .121 04 14. 3S 118 50.0 VI 1-18 1841-NT 191 2.86 45 .360 18 01. ON 105 41.0 IX- 13 1248-0 216 3.68 45 .125 04 45. 8S 118 53.0 VI 1-19 0243-N 216 4.20 45 .362 18 16. 5N 105 20.5 IX- 10 1828-N 199 3.27 45 .127 05 10. OS 118 54.5 VI 1-19 0745-0 192 3.31 45 .365 18 48. 5N 104 44.0 IX- 10 2351-N 207 3.48 45 .129 05 41. BS 118 54.9 VI 1-19 1231-0 215 3.91 45 .367 19 00. ON 104 50.0 IX- 13 1455-0 212 3.57 45 .131 06 19. 5S 118 58.0 VI 1-19 1831-NT 205 3.83 45 .369 19 05. 3N 105 09.2 IX- 13 1925-N 209 3.53 45 .133 06 57. BS 119 01.5 VI 1-19 2344-N 219 3.47 45 .371 19 12. 5N 105 40.5 IX- 14 0024-M 203 3.33 45 .135 07 26. BS 119 04.0 VI 1-20 0535-NT 199 3.56 45 .373 19 19. 5N 106 13.8 IX- 14 0601-D 221 3.63 45 .137 08 12. OS 119 06.3 VI 1-20 U25-D 214 3.79 45 .375 19 34. 3N 106 57.2 IX- 14 1148-0 184 2.74 45 .139 08 47. BS 119 00.0 VI 1-20 1718-DT 200 4.19 45 .377 19 36. 2N 107 37.5 IX- 14 1810-NT 195 3.31 45 .140 09 44. 7S 118 59.0 VI 1-21 0016-N 217 3.95 45 .379 19 38. 5N 108 24.0 IX- 14 2351-N 204 3.27 45 .163 09 56. 9S 111 59.5 vi 1-23 1153-D 217 3.82 45 .381 19 41. 7N 109 08.0 IX- 15 0520-OT 192 2.82 1239 FISHERY BULLETIN': VOL. 70, NO. 4 Appendix Table 8. — Station data: latitude and longitude, date of collection, time of day, depth of haul, and standardized haul factor. — Continued. 2 < CO •a 1 i ■a a co OS rH i s a a X 1 O t Q 1 1 X ■a 1 m z z o < H 1 It ►J 5" 1 o i» o> rH Q 1 1 1 s a a * u o X 1 X Q 1 a: •a g ■■B U ■o 45 .383 19 48. ON 109 56.7 IX-15 1126-D 212 3.37 46 .102 05 36. ON 097 54.0 IX- 2 2341-N 210 3.21 ttS .385 19 53. ON 110 46.2 IX-15 1755-NT 200 3.07 46 .104 06 09. ON 097 45.0 IX- 3 0611-0 199 2.81 ttS .387 19 58. 7N HI 25.2 IX-15 2328-N 206 3.22 46 46 .106 .108 06 07 56. 4N 35. ON 097 098 48.0 00.0 IX- 3 IX- 3 121I-D 1742-N 213 215 3.45 3.29 CRUISE CO en y-t CM o 1 09 ^ g ! 1 "3 1 S ■^ s X § _^ § as B % j3 •3 ■a s h ■B ■3 ^ ^ ct m z a g 1 g 4(. O ■■B I 1 3 X 1 1 < B ■■3 ■| 1 S t 1 < 1 ■ a> 1 s % t H oS o ci o 0) H a 3 a o o CO ►J ~i Q X Q M CO -) J a X Q M 47 .032 04 01. ON 079 54.0 VI I 1- 4 0136-N 195 2.77 47 .201 01 26. OS 085 03,5 VIII-24 0249-NT 209 3.64 47 .034 03 11. DN 079 41.0 VII I- 4 0730-D 196 2.77 47 .205 00 39. 5S 085 04.0 VIII-24 0835-0 207 3.61 47 .035 02 57. ON 079 39.0 VII I- 4 1351-D 205 3.13 47 .213 00 47. ON 084 55.0 VII 1-24 2141-N 192 3.03 47 .040 01 41. DN 079 22.0 VI I I- 4 2010-N 202 2.69 47 .217 01 31. ON 084 55.0 VIII-25 0326-NT 211 3.51 47 .049 01 00. ON 082 00.0 VII I- 5 1239-0 215 3.67 47 .221 02 08. ON 084 57.2 VIII-25 0836-D 211 3.44 47 .053 00 18. ON 081 57.0 VII I- 5 2104-N 193 2.91 47 .225 02 47. ON 084 58.5 VII 1-25 1500-DT 200 3.22 47 .057 00 22. OS 082 00.0 VII I- 6 0211-N 210 3.38 47 .229 03 33. ON 084 54.0 VII 1-25 2053-N 209 3.63 47 .061 01 11. OS 082 02.0 VII I- 6 0748-D 193 3.09 47 .233 04 21. ON 084 49.0 VIll-26 0248-NT 207 3.26 47 .065 01 46. OS 081 58.0 VII I- 6 1333-D 202 2.48 47 .237 05 06. 5N 084 45.0 VIII-26 0835-0 213 3.76 47 .069 02 27. OS 081 50.0 VII I- 6 2006-N 148 1.94 47 .240 05 41. ON 084 56.5 VIII-26 1336-0 205 3.51 47 .070 03 06. OS 082 01.0 VII I- 9 2103-N 193 2.53 47 .2 42 06 21. ON 084 54.5 vni-26 1957-N 146 ■ 2.10 47 .074 03 58. OS 082 02.0 VI I I -10 0305-NT 203 3.35 47 .2 44 06 59. ON 084 54.0 VIII-27 0156-N 207 3.58 47 .078 04 39. OS 082 03.0 VII I-IO NOT QUANTITATIVE 47 .Zttb 07 42. 5N 085 04.0 VIII-27 0825-0 204 3.52 47 .082 05 21. OS 082 02.5 VII I-IO 1358-D 204 3.09 47 .250 08 26. 5N 085 05.0 VIII-27 1430-0 210 3.61 47 .086 06 04. OS 082 00.0 VII I-IO 1941-N 208 3.06 47 .2 54 08 56. 5N 085 01.0 VIII-27 1946-N 224 4.26 47 .090 06 47. OS 081 58.0 VII I-ll 0211-N 217 3.79 47 .258 09 26. 5N 084 52.0 VII 1-28 0055-N 213 3.27 47 .094 07 28. OS 081 56.5 VII I-ll 0756-D 218 3.40 47 .268 11 58. ON 088 02.0 VIII-31 0919-D 191 3.30 47 .097 08 U.OS 082 01.0 VII I-ll 1314-D 205 3.20 47 .272 11 20. 8N 088 00.0 VIII-31 1441-0 195 3.10 47 .099 08 48. OS 082 04.0 VI I I-ll 2111-N 214 3.39 47 .276 10 54. ON 088 10.0 VIlI-31 2201-N 183 3.20 47 .101 09 29. OS 082 05.0 VII 1-12 0211-N 220 3.63 47 .278 10 21. ON 088 18.0 IX- 1 0345-DT 219 3.69 47 .103 10 09. OS 082 09.0 VII 1-12 0751-0 213 3.51 47 .2 80 09 44. 5N 088 14.0 IX- 1 0941-0 201 3.46 47 .105 10 02. OS 081 34.0 VII I-I2 1246-0 204 3.20 47 .283 09 04. ON 088 04.0 IX- 1 1512-DT 209 3.80 47 . 107 09 50. OS 080 53.0 VII 1-12 2042-N 196 2.81 47 .2 86 08 15. ON 087 52.0 IX- 1 2117-N 215 3.81 47 .109 09 35. OS 080 15.0 VI I 1-13 0215-N 205 3.23 47 .288 07 29, ON 087 44.0 IX- 2 0330-NT 215 3.91 47 .113 09 22. OS 079 39.0 VI I 1-13 0733-0 216 3.61 47 .290 06 47. ON 087 57.0 IX- 2 0949-0 207 3.32 47 .124 12 13. OS 077 39,0 VI I 1-16 1340-D 200 3.05 47 .292 06 20. ON 087 57.0 IX- 2 1514-OT 207 3.48 47 .128 12 29. OS 078 04.0 VI I 1-16 2216-N 200 3.00 47 .295 05 16. ON 087 57.0 IX- 2 2201-N 208 3.53 47 .132 12 44. OS 078 52.0 VI I 1-17 0304-NT 213 2.50 47 .297 04 38. ON 087 57.0 IX- 3 2246-N 213 3.39 47 .134 12 56. OS 079 28.0 VI I 1-17 0746-0 211 3.27 47 .301 04 00. ON 088 02.0 IX- 4 0353-OT 212 3.50 47 .137 13 10. OS 080 13.0 VI I 1-17 1341-0 210 3.17 47 .304 03 22. ON 088 04.0 IX- 4 0847-D 211 3.46 47 .139 13 27. OS 081 01.0 VII 1-17 1922-N 209 3.24 47 .306 02 44. 8N 087 59.0 IX- 4 1407-0 211 3.52 47 .141 13 38. OS 081 45.0 VII 1-18 0107-N 211 3.38 47 .310 02 02. ON 088 03.0 IX- 4 2027-N 204 3.32 47 .143 13 58. OS 082 25.0 VII 1-18 0719-D 213 3.53 47 .314 01 18. ON 088 06.0 IX- 5 0201-N 220 3.95 47 .145 14 18. OS 083 05.0 VII 1-18 1232-0 213 3.47 47 .318 00 33. 5N 088 02.0 IX- 5 0829-0 216 3.90 47 .147 14 33. OS 083 41.0 VI I 1-18 2121-N 217 3.70 47 .322 00 13. 5S 088 07.0 IX- 5 1446-0 177 2.37 47 .149 14 43. OS 084 21.0 VII 1-19 0235-N 214 3.61 47 .326 00 59. OS 088 06.5 IX- 5 2053-N 150 2.01 47 .151 14 56. OS 085 00.0 VI I 1-19 0806-0 216 3.75 47 .3 30 01 44. 5S 088 08.2 IX- 6 0255-NT 185 3.10 47 .153 14 17. OS 085 03.0 VII 1-19 1259-D 210 3.53 47 .334 02 23. 5S 088 02.3 IX- 6 0837-0 222 4.51 47 . 155 13 35. OS 085 00.0 VII 1-19 1857-N 214 3.60 47 .338 03 04. OS 088 03.5 IX- 6 1421-D 212 4.21 47 .157 12 44. OS 085 07.0 VI I 1-20 0335-NT 209 3.38 47 .342 03 47. 5S 088 03.5 IX- 6 2139-N 215 4.06 47 .159 12 06. 5S 084 59.0 VI I 1-20 0756-0 214 3.82 47 .345 04 29. OS 087 57.0 IX- 7 0334-NT 209 3.77 47 .162 11 30. OS 085 01.3 VII 1-20 1349-0 207 3.60 47 .349 05 09. 5S 088 02.0 IX- 7 0923-0 217 3.95 47 . 164 10 43. 2S 085 04.0 VI I 1-20 2146-N 215 3.65 47 .351 05 53. OS 087 59.0 IX- 7 1513-OT 174 2.60 47 . 166 10 01. OS 085 05.3 VII 1-21 0258-NT 211 3.57 47 .354 06 36. 5S 087 59.0 IX- 7 2057-N 207 3.35 47 . 168 09 16. 8S 085 07.3 VI I 1-21 0817-D 209 4.06 47 .357 07 18. OS 088 01.0 IX- 8 0207-N 205 3.32 47 .171 08 35. OS 085 02.3 VII 1-21 1325-D 212 3.61 47 .3 59 08 07. OS 088 03.0 IX- 8 0846-0 211 3.65 47 . 173 07 54. 8S 085 06.0 VI I 1-21 1945-N 216 3.61 47 .362 08 55. OS 088 04.0 IX- 8 1421-0 206 3.34 47 .175 07 16. OS 085 07.3 Vl I 1-22 0123-N 211 3.80 47 .3 64 09 40. OS 088 02.0 IX- 8 2011-N 163 2.93 47 .177 06 35. OS 085 08.5 VII 1-22 0754-0 217 2.79 47 .367 10 22. OS 088 02.0 IX- 9 0207-N 209 3.23 47 .179 05 49. 7S 085 00.0 VI I 1-22 1301-D 221 4.39 47 .369 11 02. OS 087 58.0 IX- 9 0756-D 215 3.61 47 . 181 05 17. OS 085 01.0 VI I 1-22 2101-N 213 3.85 47 .371 11 48. OS 088 00.0 IX- 9 1346-D 205 3.18 47 . 185 04 28. 5S 085 00.0 VI I 1-23 0309-NT 212 3.82 47 .373 12 31. OS 088 03.0 IX- 9 2136-N 225 4.02 47 . 189 03 51. OS 085 01.0 VII 1-23 0837-0 211 3.57 47 .376 13 13. 8S 088 01.0 IX-10 0311-NT 209 3.40 47 .193 02 58. 8S 085 01.0 VII 1-23 1456-DT 220 4.15 47 .379 13 57, OS 087 57,0 IX-10 0902-D 211 3.69 47 . 197 02 10. OS 085 03.0 VII I-?3 2055-N 222 3.97 47 . 382 14 47. OS 087 59,0 IX-10 1451-DT 205 3.26 1241 FISHERY BULLETIN': \0L. 70, NO. 4 Appendix Table 8. — Station data; latitude and longitude, date of collection, time of day, depth of haul, and standardized haul factor. — Continued. t- ■-I ■s o » OS ■s ^ g 1 1 g 1 1 fe •z, •mi* ^ 1 •O 2 i" "i ■s 1 1 1 1 3k ■3 2 s 1 1 1 * I o« o i 1 < 1 ° 1 1 J3 ■s. 73 J hJ Q SI Q m CO J Q K Q m 47 .415 11 06. 5S 095 01.5 IX-13 0745-0 211 3.14 OP .017 10 31. OS 084 54.0 XI-16 NOT QUANTiraTIVE 47 .430 09 27. OS 094 59.0 IX-14 2100-N 209 3.04 OP .019 09 40. 5S 084 52.8 XI-16 1225-D 214 3.98 47 .432 10 14. 7S 095 02.0 IX-15 0253-N 229 3.41 OP .021 08 57. IS 084 52.7 XI-16 1827-N 240 4.09 47 .436 08 32. OS 095 01.5 IX-15 1225-D 237 4.51 OP .023 07 57. 8S 084 53.6 XI-16 0008-N 193 3.38 47 .438 07 39. 5S 095 08.5 IX-15 2040-N 214 3.80 OP .025 07 13. IS 084 52.7 XI-17 0611-D 205 3.41 47 .440 06 41. OS 095 09.0 IX-16 0258-N 256 4.84 OP .027 06 07. 8S 084 55.1 XI-17 1153-0 206 3.65 47 .443 05 55. OS 095 08.3 IX-16 0908-0 249 4.26 OP .029 05 15. 9S 084 53.3 XI-17 1801-N 208 3.32 47 .446 05 01. 5S 095 05.5 IX-16 1510-0 173 2.99 OP .032 04 12.9S 084 55.8 XI-17 0002-N 216 4.01 47 .450 04 14.05 095 01.0 IX-16 2036-N 216 4.54 OP • 036 03 34^35 084 56.1 XI-18 0601-D 230 4.03 47 .454 03 27. 3S 095 00.5 IX-17 0242-N 215 4.08 OP .040 02 30^7S 084 57.9 XI-18 1141-D 201 3.44 47 .458 02 40. 5S 095 00.5 IX-17 0817-D 161 2.31 OP .044 01 57. OS 084 58.0 XI-18 1722-N 218 4.19 47 .46 2 01 54. OS 095 04.0 IX-17 1356-D 180 2.68 OP .048 00 41. IS 084 57.9 XI-18 2331-N 174 2.77 47 .466 01 01. OS 095 08.5 IX-17 2101-N 210 5.19 OP .052 00 04. IN 084 57.9 Xl-19 0619-D 204 3.60 47 .470 00 24. OS 095 09.2 IX-18 0424-OT 166 1.98 OP .056 00 17. 8S 085 37.8 XI-19 1141-D 223 4.16 47 .478 01 05. 2N 094 57.7 IX-19 1601-OT 205 3.43 OP .060 00 30. 6S 086 12.8 XI-19 1753-N 209 3.75 47 .486 02 32. 5N 094 42.0 IX-19 0313-NT 205 3.00 OP .064 00 54. 2S 087 09.3 XI-19 2329-N 227 4.29 47 .490 03 16. 3N 094 40.8 IX-19 0835-D 222 3.61 OP • 068 01 10. 4S 087 52.2 XI-20 0532-D 216 3.78 47 .494 03 58. ON 094 59.0 IX-19 1349-D 216 3.65 OP .072 01 37. 8S 088 47.9 XI-20 1123-D 216 3.66 47 .498 04 44. ON 094 55.0 IX-19 2207-N 213 3.51 OP .076 01 56. 8S 089 26.0 XI-20 1743-N 213 3.87 47 .501 05 36. ON 094 55.5 IX-20 0319-NT 206 3.90 OP .080 02 20. 7S 090 24.7 XI-20 2331-N 226 3.99 47 .504 06 26. 5N 094 58.5 IX-20 0853-D 219 3.39 OP .0 84 02 36. 6S 091 08.8 XI-21 0518-D 211 4.17 47 .507 07 19. ON 094 57.5 IX-20 1427-D 209 3.19 OP .088 02 59. 5S 092 02.8 XI -21 1056-D 216 3.00 47 .509 08 05. ON 095 02.0 IX-20 1955-N 207 2.87 OP .092 02 11. 6S 092 06.1 XI-21 1747-N 211 3.73 47 .511 08 56. 5N 095 04.0 IX-21 0142-N 200 2.83 OP .096 01 07. 5S 092 03.9 XI-21 2329-N 219 3.99 47 .513 09 49. ON 095 05.0 IX-21 0722-D 193 2.44 OP .100 00 24. OS 092 05.0 XI-22 0527-D 199 2.89 47 .515 10 45. 5N 095 04.0 IX-21 1321-D 217 3.58 OP .104 00 46. 5N 092 05.8 XI-22 1140-0 214 3.63 47 .517 11 36. ON 095 00.5 IX-21 2101-N 211 3.48 OP .108 01 24. 8N 092 08.2 XI-22 1730-N 215 4.07 47 .520 12 33. 2N 094 57.0 IX-22 0249-N 205 3.13 OP .112 02 35. 5N 092 03.7 XI-22 2330-N 216 3.96 47 .523 13 16. ON 095 00.0 IX-22 0748-D 210 3.72 OP • 116 03 15. ON 092 00.3 XI-23 0517-0 205 3.63 47 .525 14 11. ON 095 01.0 IX-22 1330-D 204 2.92 OP .120 04 2 3. 6N 091 58.8 XI-23 1139-D 217 3.60 47 .527 15 00. 3N 094 59.0 IX-22 1858-N 204 3.01 OP OP .124 .128 05 06 08. 7N 08. 4N 091 091 56.7 58.9 XI-23 XI-23 1733-N 2333-N 203 203 3.50 3.40 CRUISE OP OP .132 06 48. 3N 092 00.4 XI-24 0515-D 184 2.92 OP .001 09 17. 4S 079 41.9 XI-14 0228-NT 198 3.00 OP .136 07 5 7. 4N 092 02.8 XI-24 1157-D m 4.03 OP .002 09 41. 9S 080 28.4 XI-14 0836-0 255 3.88 OP • 138 08 12. 3N 092 03.2 XI-24 1716-N 193 3.10 OP .003 09 53. 6S 080 50.6 XI-14 1231-0 206 3.06 OP • 155 09 03. IN 092 00.4 XI-25 1153-D 216 3.78 OP .005 10 14. OS 081 26.0 XI-14 1818-N 192 3.41 OP .160 10 14. 9N 091 59.5 XI-25 2335-N 203 3.51 OP .007 10 40. OS 082 20.5 XI-14 0008-N 191 3.28 OP .162 11 09. 5N 092 00.3 XI-26 0517-D 181 3.C9 OP .009 10 58. 4S 082 59.4 XI-15 0626-0 191 2.63 OP .164 11 43. 5N 091 59.6 XI-26 1242-0 216 4.24 OP .011 11 26. 2S 083 51.6 XI-15 1216-0 179 2.87 OP .166 12 24. ON 092 01.1 XI-26 1713-N 204 3.40 OP .013 11 58. IS 084 54.6 XI-15 1948-N 205 2.68 OP .168 13 21. 2N 091 59.4 XI-26 2157-N 212 3.83 OP .015 11 18. 9S 084 54.4 XI-15 0001-N 194 3.37 1242 DESCRIPTION OF BLACK SEA BASS, CENTROPRISTIS STRIATA (LINNAEUS), LARVAE AND THEIR OCCURRENCES NORTH OF CAPE LOOKOUT, NORTH CAROLINA, IN 1966 Arthur W. Kendall, Jr.^ ABSTRACT Larvae of black sea bass collected during RV Dolphin ichthyoplankton surveys of the mid- Atlantic continental shelf are described. Development of most meristic characters occurs between 6 and 10 mm standard length. The larvae are identified by characteristic ventral pigment patterns, body shape, meristic counts, and lack of extensive armature. The 147 larvae were taken during cruises from June to November 1966, from 4 to 82 km from shore. They were found in tows from the surface to 33 m in water varying in surface temperature from 14.3° to 28.0°C and surface salinity from 30.3 to 34.6^.. Black sea bass are of considerable economic im- portance and occur along most of the Atlantic coast of the United States, Although they were first studied in the late 1800's little is reported on their early life history. Spawning is reported to take place in May off North Carolina and in mid-May and June off New Jersey and southern New England (Bigelow and Schroeder, 1953; Miller, 1959). Wilson (1891) described their embryology as part of an incomplete monograph on the species but did not describe the larvae or provide diagnostic characteristics to identify eggs. Hoff (1970) figured a black sea bass egg and prolarva from artificially reared specimens but gave no written description (Figure 1). The eggs are pelagic, clear, round, and 0.9 to 1.0 mm in diameter. They have a smooth shell, narrow perivitelline space, and a single oil globule. They hatch in 75 hr at 16°C and in 38 hr at 23°C (Wilson, 1891; Hoff, 1970). The larvae re- main inadequately described although Pearson (1941) identified specimens collected at the mouth of Chesapeake Bay as black sea bass by comparing them with a known series from south- ern New England using the ventral pigment pattern and fin ray counts. Apparently Merri- ' National Marine Fisheries Service, Middle Atlantic Coastal Fisheries Center, Sandy Hook Laboratory, High- land, NJ 07732. man and Sclar (1952) had access to the same or similar specimens as Pearson because they pointed out differences between black sea bass and silver hake, Merluccius bilinearis, larvae. 0. E. Sette's notes, made in connection with his work on Atlantic mackerel, Scomber scomhrus, early life history contained a mention of black sea bass in a description of bluefish, Pomatomus saltatrix, larvae. Larvae of black sea bass have been identified from other collections of ichthyo- plankton along the east coast of North America (Perlmutter, 1939; Herman, 1963) with no ref- erence to means of identification. Figures of juveniles, ranging from 39 to 58 mm total length (TL), are shown in Bean (1888), Hildebrand and Schroeder (1928), and Fowler (1945) and reproduced here (Figure 2). Three species of Centropristis occur along the Atlantic coast; C. striata is the most widespread and occurs from the Gulf of Maine to the Florida Keys (Miller, 1959). Rock sea bass, C. phila- delphica, occurs along the Atlantic coast south of Chesapeake Bay, and bank sea bass, C. ocyur- us, is found generally offshore south of Cape Hatteras. All three species also occur in the Gulf of Mexico, C. striata as the subspecies C. s. melana (southern sea bass). Black sea bass generally occur over hard bottoms and migrate along the middle Atlantic coast shoreward and northward in summer and oflfshore and south in Manuscript accepted May 1972. FISHERY BULLETIN: VOL, 70, NO. 4, 1972. 1243 FISHERY BULLETIN: VOL. 70, NO. 4 Figure 1. — Previously illustrated black sea bass eggs and prolarva. A) egg, 23 hr after fertilization at 23°C, from Hoff (1970, Figure 8) ; B) egg, 65 hr after fertilization at 16°C, from Wilson (1891, Figure 151) ; C) prolarva, 54 hr after hatching at 23°C, 2.01 mm TL, from Hoff (1970, Figure 9). winter. C. striata is the only serranid expected to spawn on the continental shelf between Ches- apeake Bay and Cape Cod, Mass. (Miller, 1959). Among east coast serranids the unique fin ele- ment counts for Centropristis for dorsal (X, 11) and anal (III, 7) fins allowed me to determine that I had larvae of this genus. The modal pec- toral fin ray count of 18 for black sea bass is distinctive among Centropristis and seen on larg- er larvae. The pigment patterns on larger lar- vae, whose fin complements were complete, were seen in smaller larvae which appeared to be de- veloping the meristic characters of black sea bass. This was the rationale for identifying the larvae described here as Centropristis striata. On five ichthyoplankton surveys by the RV Dolphin between June and November 1966, we collected larval black sea bass at stations between Barnegat Bay, N.J., and Cape Lookout, N.C. This paper describes these larvae and their oc- currences. PROCEDURES Collecting methods and hydrographic data from the 1965-66 RV Dolphin ichthyoplankton survey are reported in detail by Clark et al. (1969). Gulf V plankton tows were taken at 92 stations on eight cruises between Cape Cod, Mass., and Cape Lookout, N.C. The oblique tows covered 4.6 km with one net fishing from the surface to 15 m and, simultaneously, a second, from 18 to 33 m. Samples fixed in 5% buffered 1244 KENDALL: BLACK SEA BASS LARVAE Figure 2. — Previously illustrated juvenile black sea bass. A) 39 mm SL, from Bean (1888, Figure 12) ; B) 42.4 mm SL, from Fowler (1945, Figure 263) ; C) 58 mm TL (?) from Hildebrand and Schroeder (1928, Figure 144), 1245 FISHERY BULLETIN: VOL. 70, NO. 4 Formalin" were returned to the laboratory where fish eggs and larvae were removed. Black sea bass larvae were identified using criteria described, separated from the rest of the larvae in the Dolphin collections, counted and measured in standard length (SL). Black sea bass eggs are not well described, and several of their characteristics apply to many other species, so they were not identified in our samples. Other young black sea bass examined included one larva (13.0 mm SL)^ and several juveniles (37- 73 mm SL) .' Body proportions were measured to the nearest 0.1 mm on selected larvae in Formalin on a depression slide with an ocular micrometer. The base points for larval measure- ments approximate those used by Ahlstrom and Ball (1954) except body depth, which was mea- sured at the junction of the cleithra, and stan- dard length, measured to the distal ends of the hjq^urals when formed. Base points for mea- surements of juveniles follow Hubbs and Lagler (1958), We determined meristic counts on se- lected specimens lightly stained with alizarin red. Osteological examination was made from specimens cleared and stained following Cloth- ier's (1950) method. Michael P. Fahay illus- trated the larvae (Figures 3 and 9). DESCRIPTION OF LARVAE In the following description, features useful in identifying black sea bass larvae are empha- sized rather than those demonstrating general teleostean development. The approach follows Ahlstrom and Ball (1954) in that each feature is at once traced through its development within the size range (2-13 mm) of the available larvae. Four areas of development are described: ar- mature, body shape, meristic characters, and pigment patterns. Stages of development of black sea bass larvae are illustrated in Figure 3. '' Reference to trade names in the publication does not imply endorsement of commercial products by the Na- tional Marine Fisheries Service. ^ Collected on October 5, 1967, at Corson Inlet, N.J., by Walter S. Murawski, Jr., New Jersey Department of Conservation and Economic Development. * Inshore seining collections of fish from New Jersey taken by Dr. Albert E. Parr. ARMATURE Among larvae of serranids which have been described there is diversity of development of armature. Species of Epmephelus develop an- terior dorsal and pelvic spines nearly as long as the larva. These spines are barbed and serrated. Preopercular spines are also well developed (Sparta, 1935; Mito, Ukawa, and Higuchi, 1967; Presley, 1970) . Larvae of other genera are less ornate and the relative length of fin spines is near that of the adults in some (Bertolini, 1933) . Black sea bass larvae are among the serranids with little development of armature. No fin spines are either serrated or pronounced. Pelvic fin spines do not reach the vent; the dorsal and anal fin spines are shorter than the rays. Four to seven short, widely spaced spines are present on the posterior margins of the preopercle and opercle on larvae longer than 5 mm. The three small spines on the opercular flap of the adult form at 8 mm (Figure 3C). Preopercular ser- rations characteristic of the adult develop early in the juvenile stage. BODY SHAPE Changes from larval to adult body form take place over a narrow size range, and the extent of development among fish with similar standard lengths varies. Some body proportions of ju- venile and adult black sea bass given by Miller (1959) are compared to those of the developing larvae in this section and in Figures 4 to 6. Be- tween 2 and 5 mm the body proportions remain fairly constant with a slight increase in snout and eye length, and body and caudal peduncle depth, relative to standard length. Most head and body proportions increase significantly be- tween 5 and 6 mm then level oflf as they approach those of the adult. Caudal peduncle depth and total length, relative to standard length, continue to increase through the larval stage as the caudal fin develops. Body depth, head length, and pre- anal length proportions increase through the ju- venile stage. The adult black sea bass is robust with a large terminal mouth and large head. The back is slightly elevated anteriorly. The dorsal fins are contiguous and the pectorals and pelvics 1246 KENDALL: BLACK SEA BASS LARVAE Figure 3. — Black sea bass larvae. A) 5.1 mm SL; B) 6.2 mm SL; C) 7.9 mm SL. 1247 FISHERY BULLETIN: VOL. 70, NO. 4 are large. The caudal fin outline varies from rounded to trilobed, with one upper ray produced in larger specimens. Head Length Between 2 and 4 mm head length averaged about 33 9f of SL. At 5 mm it reached 37-38% where it remained through 12 mm. Almost all values lie between 30 and 40% , except in a few larvae smaller than 5 mm where precise measure- ment is difficult. The juveniles demonstrate a continuing trend toward a longer head ranging from 34 to 45% of SL. Miller (1959) gives 40- 41% as the proportion in his specimens (Fig- ure 4A). Eye Length Eye length remained constant throughout larval development at about 9-10 Cr of SL. Most juveniles as well as Miller's (1959) adult spe- cimens also ranged from 9 to 10% (Figure 4B) . Snout Length In larvae between 2 and 6 mm snout length in- creased from 6 to 11% of SL where it holds through adulthood (Miller, 1959) (Figure 4C). 35r B < Q Z 10 K 40 35 - C - SNOUT LENGTH • • • • • • • • . B - EYE LENGTH • • • • • » • • 1 I I 1 1 1 1 ■ A- HEAD LENGTH - • • • • •• • • • NUMBERS 1-5 6- 10 OF FISH • .*. > 10 • 20 30 40 50 60 70 MEAN STANDARD LENGTH (mm) 80 "ADULT Figure 4. — Body proportions of black sea bass associ- ated with the head plotted as percentages of standard length. A) head length; B) eye length; C) snout length. Each point represents the mean of several ob- servations. "Adult" points from Miller (1959). 65 - A - PRE AN AL LENGTH 50- BODY DEPTH 10 20 30 40 50 MEAN STANDARD LENGTH (mm) NUMBERS OF FISH 1 - 5 • 5- 10 • > 10 • 60 70 30 ADULT Figure 5. — Body proportions of black sea bass associated with the trunk plotted as percentages of standard length. A) preanal length; B) body depth. Each point repre- sents the mean of several observations. "Adult" points from Miller (1959). Preanal Length The preanal length increases from about 50 % of SL at 5 mm to 58% at 10 mm. During ju- venile development it increases to nearly 65% (Figure 5A). Body Depth Relative to standard length, body depth in- creases from about 25 to 27% during larval de- velopment. During the juvenile stage it con- tinues to increase to about 30%- (Figure 5B). The adult proportion is about 34% (Miller, 1959). Total Length Total length is 102% of SL from 2 to 5 mm. As caudal fin development proceeds total length becomes a larger portion of standard length, reaching 120% in our largest larva. It remains constant through juvenile development at about 125% of SL (Figure 6A). Caudal Peduncle Depth From 4 mm where caudal peduncle depth is 6% of SL it increases steadily through larval 1248 KENDALL: BLACK SEA BASS LARVAE 15 r B- CAUDAL PEDUNCLE DEPTH • • • • • • • 10 r 1— 2 -J • • Q a: < 130 Q z < "120 U- O 2 110 A- TOTAL LENGTH • • • • • • U • ^ NUMBERS OF FISH °- 100 1-5 • 6- 10 • > 10 1 — ■ -J • 10 20 30 40 50 00 70 MEAN STANDARD LENGTH (mm) 80 -ADULT Figure 6. — Body proportions of black sea bass associ- ated with caudal fin development plotted as percentages of standard length. A) total length; B) caudal pe- duncle depth. Each point represents the mean of sev- eral observations. "Adult" points from Miller (1959). development to about 14% where it remains in adults (Miller, 1959) (Figure 6B). DEVELOPMENT OF MERISTIC CHARACTERS Most of the meristic features of black sea bass larvae develop within a length range of 4 mm (Table 1). Larvae 5 mm long have undifferen- tiated finfolds with a few caudal rays, buds of paired fins, some gill rakers and branchiostegal rays, and an incompletely ossified vertebral col- umn (Figure 3A). By 7 mm most median and paired fin elements and branchiostegal rays have formed, some gill rakers are visible, and the ver- tebral column is ossified (Figure 3B). Other than gill rakers and scales, adult complements of meristic characters are reached by 9 mm (Fig- ure 3C). The following descriptions roughly follow the sequence of attainment of adult char- acters. Table 1. — Development of meristic characters of larval black sea bass. Caudal fin rays Dorsal fin Anal Spines fin Rays Pectoral fin rays Pectoral fin elements Vertebrae Branchio- stegal rays SL Sec. Primary Sec. Gill rakers (mm; - Upper Upper Lower Lower Spines Rays 3^ 3.4 4.0 4.1 4.5 2 4.5 4 4.6 1 1 3 7 5 4.8 4.9 9 4 5.0 10 4 8 5.0 6 S 1 12 5 6 5.2 6 6 2 14 5 5 5.3 8 6 15 5 5 5.5 6 7 15 5 6 5.5 5 4 5 6 5 3 11 5 5 5.9 6 4 5 5 7 18 3 6 5.9 7 5 7 6 7 5 20 6 8 6.0 8 7 8 8 8 9 3 22 6 8 6.1 5 4 6 9 7 7 22 5 8 6.3 8 7 2 10 8 9 3 16 6 9 6.5 9 7 2 9 8 2 6 12 5 23 7 a 6.6 9 8 1 8 ^ 2 8 14 4 23 6 3 6.7 9 8 1 14 11 2 8 16 4 24 7 11 6.7 9 8 8 11 2 8 12 4 23 6 7 7.0 9 8 2 9 10 3 7 10 5 24 7 10 7-1 9 8 2 10 ilO 3 7 14 5 24 7 10 7.7 9 8 3 10 8 3 7 13 6 24 7 11 8.7 4 9 8 4 10 U 3 7 18 6 24 ■ 7 10 9.9 4 9 8 3 10 11 3 7 19 6 24 7 11 10.6 9 9 8 8 10 11 3 7 18 6 24 7 14 11.0 9 9 8 8 10 111 3 7 18 6 24 7 16 11.8 5 9 8 5 10 M 3 7 18 6 24 7 15 '■ Damaged. 1249 FISHERY BULLETIN: VOL. 70, NO. 4 Teeth At 5 mm, widely spaced small conical teeth are visible on the premaxillaries (Figure 3A). By 6.5 mm teeth are fairly closely spaced all along the premaxillaries and medially on the dentaries (Figure 3B). By 10 mm the teeth on the pre- maxillaries are very closely spaced and slightly recurved; those on the dentaries are enlarged and more widely spaced posteriorly but resemble those on the premaxillaries anteriorly. Axial Skeleton Ossification of the vertebral column proceeds posteriorly beginning between 4.5 and 5.0 mm. Neural and hemal spines form concurrently with their associated vertebrae. By 5 mm the anteri- or 12 vertebrae have ossified. By 6.5 mm all of the vertebrae are ossified except two or three an- terior to the urostyle. The urostyle ossifies at 6 mm. The penultimate and antepenultimate vertebrae ossify last at 7 mm (Table 1). The first caudal supports to ossify are the medial four hypurals at 6.5 mm, Hypural 1 is ossified at 8.0 mm and hypural 6 is ossified at 8.5 mm. The uroneural and epurals form at 10 mm. The structure of the caudal region at 11.8 mm (Figure 7) varies only slightly from the typical perciform type described by Gosline (1961). As in other serranids, the two separate uroneurals have fused, but no fusion of the hypurals has oc- curred. Otherwise the caudal skeleton is typical, having three epurals, a urostyle with a single os- sification, and the hemal arch on the antepenulti- mate and penultimate vertebrae being autog- enous (not fused to the vertebrae). There are 15 branched segmented rays supported by six hypurals, 7 ventrally on hypurals 1, 2, and 3 and 8 dorsally on hypurals 4, 5, and 6. One ray dorsal and one ray ventral to these are also segmented but not branched. About eight raylets form dorsally and an equal number form ventrally an- terior to the segmented rays. If the first hypural is considered a parhypural (Nybelin, 1963), black sea bass have only two ventral hypurals. Branchiostegal Rays The first branchiostegal rays form at about 4.5 mm. By 5 mm four to six rays have formed and at 6.5 mm the adult complement of seven rays is reached, with the medial ones being the last to form (Table 1). Caudal Fin From 2 to 4 mm the finfold is symmetrical around the tip of the notochord. Actinotrichs are formed adjacent to the posterior 10 9f of the UNBRANCHEO RAY Figure 7. — Caudal skeleton of a black sea bass larva 11.8 mm SL. APU = antepenultimate vertebra; EP = epural; HS = hemal spine; HY ^ hypural ; NS = neural spine ; PU = penultimate vertebra; UN = uroneural; UR = urostyle. UNBRANCHEO RAY 1250 KENDALL: BLACK SEA BASS LARVAE notochord. Between 4 and 5 mm the anlage of the base of the caudal fin starts to form ventral to the notochord, just anterior to its tip (Figure 3A). By 5.5 mm the notochord is slightly up- turned, the developing hypural region appears bilobed and about nine primary caudal rays are formed. The caudal fin at 6 mm has a rounded homocercal outline and eight segmented prin- cipal rays in the superior lobe and seven in the inferior lobe (Figure 3B) ; the rays are branched and some secondary procurrent rays are present. At 8 mm the dorsal hypurals are slightly longer than the ventral ones and the rays and raylets are more clearly defined and approaching the adult complement (Table 1). Anal Fin Between 2 and 5 mm there is an undifferen- tiated finfold in the area of the anal fin. Between 5 and 6 mm, fin rays start to form in the anal finfold between the vent and the most prominent pigment spot on the ventral surface of the trunk (Figure 3A). By 6 mm about six rays and one anal spine are seen (Figure 3B). The finfold posterior to the fin is reduced. By 7 mm the three anal spines are formed. The second spine is first to form and is most prominent through- out development (Figure 3C), The third spine forms as a ray; by 7 mm the spinous form is apparent. The first spine is smaller than the others and forms last at 6.5 mm. By 7 mm the adult complement of seven anal rays is reached and some are branched (Table 1). Dorsal Fins The undifferentiated dorsal finfold extends from the nape to the caudal region at 4 mm. By 5 mm the finfold becomes elevated about half- way back on the body where fin ray development begins (Figure 3A). Rays and spines develop along the dorsal fin base and the dorsal finfold posterior to the dorsal fin disappears between 5.5 and 6.5 mm. The anterior spines and poste- rior rays develop at a smaller size than inter- mediate fin elements. The first four dorsal spines are visible by 6 mm. The second through fourth are longer and remain so. The first dorsal spine is about half as long as the second. The second and third spines are the same length at 6.5 but, by 10 mm, the third has become 1.5 times longer than the second. The fourth spine^ the longest, is slightly longer than the third. The final complement of 10 spines and 11 rays is attained by 8.7 mm (Table 1). The rays are branched and seg- mented by 8 mm (Figure 3C). Median Fin Supports Anterior interneurals and the interhemal sup- porting the second anal spine ossify concurrently at about 8.5 mm. By 10 mm most of the inter- neurals and interhemals are formed. The an- terior two interneurals fuse to support the first two dorsal spines (Figure 8) . The first two anal spines are supported by one interhemal, appar- ently formed by fusion of two elements (Figure 8). Pelvic Fins Buds of the pelvic fins are seen on 4-mm lar- vae. Fin rays form between 4 and 6 mm. Rays and spines are first seen between 5 and 6 mm (Figure 3B). At 8 mm the adult complement of one spine and five rays has formed, with the smooth spine two-thirds as long as the longest ray (Table 1). Pectoral Fins Pectoral fin buds are present on the smallest larvae (2.1 mm) examined. The early pectorals change little until fin rays appear between 5 and 6 mm. By 6.5 mm the rays are mostly formed and the adult complement of 18 or 19 rays is reached by 9 mm (Table 1), Gill Rakers At 5 mm gill rakers appear as a few tubercles on the gill arches. By 6 mm there are nine rakers on the lower limb and none on the upper. At 10.6 mm there are 4 rakers on the upper and 10 on the lower limb. The adult complement of about 10 upper and 18 lower is reached in ju- veniles. 1251 FISHERY BULLETIN: VOL. 70, NO. 4 Scales Scale formation occurs at a size between the largest larva (13.0 mm), which is scaleless, and the smallest juvenile (37 mm) in our collections. PIGMENT PATTERNS Pigmentation on Formalin-preserved black sea bass consists of a few melanophores in char- acteristic positions, mainly along the ventral part of the larva (Figure 9). Other kinds of larvae in the collections with spots in similar po- sitions had meristic counts and body shapes ap- proximating black sea bass. However, the rel- ative size of the various spots in combination with examination of the sequence of development of meristic characters, the distinctive fin element counts on larger larvae, and body shape assured separation of black sea bass from other larvae. Head Region At 5 mm one spot usually forms ventrally on the median cartilage between the dentaries and urohyal. A second forms posterior to this on some specimens (Figure 9). A spot forms at 4 mm on each angular (Figure 3A). A charac- teristic transverse dentritic spot forms immedi- ately anterior to the symphysis of the cleithra. Usually there is a spot between the bases of the pelvic fins. Dorsally there is a variable number of spots irregularly spaced on the posterodorsal covering of the cerebellum, and generally a pair of spots internally on the posterior surface of the midbrain. Between 12 and 13 mm a band of minute melanophores develops from the angular, past the eye, to the anterior part of the cerebel- lum. There are also several larger spots on the anterior halves of the cerebral hemispheres and a group of spots which originate at the eye and extend posteriorly to the opercular flap. Gut Region Considerable internal pigmentation develops in the dorsal area of the gut cavity, mostly on the surface of the viscera. These are large spots but superficially not readily definable. In lar- vae up to about 6 mm, this pigment reaches the exterior as a large intense spot on the posterior region of the renal tract (Figures 3 A and B). In some specimens there is a smaller spot just anterior to the vent and another one about mid- way between the origin of the pelvic fins and the vent, along the midventral line (Figure 9). Trunk and Caudal Region Occasionally a few irregular spots occur dor- sally on the trunk about midway on the body. Be- tween 12 and 13 mm, a series of about six groups i(/?'^=r^=r: ^«^4^^^i^- ■ Figure 8. — Cleared and stained black sea bass larva 11.8 mm SL. 1252 KENDALL: BLACK SEA BASS LARVAE Figure 9. — Ventral view of pigment on a 5.4-mm black sea bass larva. 1253 FISHERY BULLETIN: VOL. 70, NO. 4 of small melanophores develops along the lateral line. At this length a few spots also develop at the origin of the dorsal fin. As the anal fin rays form, spots develop at some of their bases. A prominent spot forms early, near the posterior end of the anal fin; it is followed by four to six smaller spots along the midventral line that ex- tends to the base of the caudal fin where there is another large spot. Before notochord flexion there is a prominent spot on the developing fin rays (Figure 9). This persists as the fin devel- ops and reaches a position at the bases of some of the ventral rays of the caudal fin. As the caudal rays develop, some less prominent spots appear on rays dorsally and ventrally from the first spot. The pigment in the caudal region and along the midventral line persists through larval development. LARVAL OCCURRENCES Black sea bass larvae were taken on five cruises from June to November 1966, between Sandy Hook, N.J., and Cape Lookout, N.C. (Ap- pendix Table) . Of 39 tows containing black sea bass larvae, 20 had only one larva, and only two had more than 20 larvae; thus only limited in- ference about their distribution and relative abundance can be derived from these data. Lar- vae were taken in both shallow (0-15 m) and deep ( 18-33 m) tows from 4 to 82 km from shore. Water depths in the areas of capture ranged from 15 to 51 m. Surface temperature varied from 14.3° to 28.0°C; that on the bottom from 8.3° to 26.6°C. Surface salinity varied from 30.3 to 34.6%c. Seasonally, there seems to be some northerly progression of spawning. During June one larva was taken off Oregon Inlet, N.C, and one off Paramore Island, Va. (Figure 10). During August, larvae occurred from Cape Henlopen, Del, to Ocracoke Inlet, N.C, being most abun- dant off Maryland and Virginia (Figure 11). Figure 10. — Occurrence of black sea bass larvae in shallow (0-15 m) (open symbols) and deep (18-33 m) (solid symbols) tows. Three RV Dolphin cruises, 1966. 1254 KENDALL: BLACK SEA BASS LARVAE BLACK SEA BASS LARVAE CRUISE D-66-10 AUG. 5-26, 1966 BLACK SEA BASS LARVAE CRUISE D-66-12 SEPT. 28 - OCT. 20, 1966 LARVAE/TOW • NONE D 1-5 SHALLOW ■ 1-5 DEEP \... Figure 11. — Distribution and abundance of black sea bass larvae in August (left) and October (right) from RV Dolphin cruises. 1255 FISHERY BULLETIN: VOL. 70. NO. 4 During September only the four northern tran- sects were sampled and the only larva taken was off Barnegat Inlet, N.J. (Figure 10). During October larvae ranged from Great Egg Inlet, N.J., to Currituck Light, N.C. In November four larvae occurred between Cape Henlopen, Del., and Ocracoke Inlet, N.C. (Figure 10). The larvae ranged from 2,1 to 13.0 mm SL (Figure 12). Among the cruises the mean size of larvae was nearly constant. The length-fre- quency curve indicates that we undersampled small larvae (< 4 mm), probably because they were extruded through the meshes of the net. Comparisons of mean lengths of larvae taken during day and night and in shallow and deep tows for the individual cruises were made. Lar- vae in the deep net were slightly larger than those in the shallow net. Day-night differences in size were inconclusive (Table 2). Geograph- ic variations in size are not apparent. A comparison of fish caught per successful tow in shallow and deep tows taken during day and night shows that more fish were taken in shallow than in deep tows and more were taken at night than during day (Table 2). Our data on size and diurnal and depth distribution of the larvae indicate several things. More and slightly larger larvae were taken at night than during the day. Also, larvae in the deep net were slightly larger than those in the shallow net. More larvae were taken in the shallow than STANDARD LENGTH Figure 12. — Length distribution of black sea bass larvae from RV Dolphin cruises, 1966. in the deep net. Thus it appears that larvae are more abundant near the surface (0-15 m) than deeper (18-33 m) waters and that visual warning allows larger larvae to escape in near surface waters, particularly during the day. DISCUSSION The general distribution of larvae and juve- niles of black sea bass can be deduced from their occurrence in several disjunct studies of estu- aries and coastal waters of the Atlantic coast. Early larvae (< 6 mm) have been found at the mouth of Chesapeake Bay (Pearson, 1941), in Long Island Sound (Perlmutter, 1939), and in Narragansett Bay (Herman, 1963). The lack of black sea bass larvae in some intensive sur- Table 2. June Aug. Sept. Oct. IMov. -Black sea bass larvae from 1966 RV Dolphin cruises. Total numbers, numbers per positive tow, and lengths in shallow and deep tows by cruise and time of day. Day Night Day Night Day Night Day Night Day Night Time of day Shallow Deep Cruise Na. Length (mm) No. Length (mm) No. /tow Range Mean No. /tow Range Mean 32 72 1.0 6.4 9J0 1.0 1.8 I.O 1.0 4.6 2.1- 5.5 2.8-10.7 5.4 4.3- 6.3 3.5- 6.1 7.5 5.2 4.6 3.7 5.4 5.4 5.5 5.1 7.5 5.2 2 16 4 2 2.0 1.0 3.2 IX) 2.0 5.5- 5.9 5.4- 7.9 2.5-11.8 3.6- 7.3 7.6-, 13.0 5.7 6.7 6.3 5.5 10.3 Day total or mean 42 3.5 2.1- 7.5 4.2 4 1.3 5.4-13.0 6.8 Night total or mean 79 6.1 2.8-10.7 5.4 22 2J2 2.5-11.8 6.1 Grand total or mean 121 4.8 2.1-10.7 5.0 26 2.0 2.5-13.0 6.5 1256 KENDALL: BLACK SEA BASS LARVAE veys along the Atlantic coast is remarkable (e.g., Merriman and Sclar, 1952; Wheatland, 1956; Richards, 1959; Massmann, Joseph, and Norcross, 1962; Marak et al., 1962). Ichthyo- plankton sampling in more enclosed areas, such as Indian River, Del. (Pacheco and Grant, 1965) and Sandy Hook Bay, N.J. (Croker, 1965), has failed to reveal larvae. Juveniles (25- 75 mm) have been taken from saline areas of estuaries from Florida (Tagatz, 1968), Mary- land (Schwartz, 1961, 1964), Delaware (de Sylva, Kalber, and Shuster, 1962), New York (Perlmutter, 1939; Greeley, 1939; Richards, 1963), Rhode Island (Herman, 1963), and Mas- sachusetts (Lux and Nichy, 1971). Bean (1888) reported that young about 1 inch (25 mm) long were common in Great Egg Harbor Bay, N.J., and Nichols and Breder (1927) reported 20-mm fish over oyster beds off Staten Island, N.Y., in August. Massmann et al. (1962) found one 43-mm juvenile in the ocean off Virginia. Black sea bass spawn offshore along the coast from Florida to New England. Spawning takes place earlier in the southern part of the range than in the northern part; in May off North Carolina (Smith, 1907); in late May off Chesa- peake Bay (Hildebrand and Schroeder, 1928); and into early summer off southern New Eng- land (Bigelow and Schroeder, 1953). At least some of the young, less than 30 mm, enter open estuaries near inlets where they spend their first summer associated with hard bottoms such as oyster shells (Nichols and Breder, 1927; Arve, 1960; Richards, 1963). Young leave the estu- aries during fall and return during spring. Enough return to estuaries in subsequent years to support fisheries there. From references made to black sea bass abundance around the turn of the century (Bean, 1888; Smith, 1898; Sherwood and Edwards, 1901), it seems that present stocks in the northern part of the range are diminished. This decrease in abundance may be associated with decrease in oyster beds (Arve, 1960). Commercial catch records show recent catches near the historical mean, but indicate a shift in abundance from the New York-Dela- ware to the Chesapeake region (Lyles, 1967). Our data on offshore occurrences of larvae complement work in estuaries where early stages have been found. However, a definitive picture of the early life history of this species is still lacking. The small numbers of larvae taken in this survey do not seem consistent with the pop- ulation size and extent of adult black sea bass along the coast. Possibly we sampled in a year when spawning was unsuccessful or our sam- pling was not effective for capturing black sea bass larvae in proportion to their abundance. The pelagic existence of this fish is short. Larvae longer than 13 mm were not taken, pre- sumably because near that size they assume de- mersal or estuarine habits. Few of these late larvae and early juveniles have been collected, and it is still not known what part of the pop- ulation may enter estuarine waters and what part remains at sea. The routes and mechanisms of larval transport from spawning grounds to nursery areas are also unknown. The known seasonal distribution of larvae shows that black sea bass spawn over a long period. The range of juvenile sizes taken in individual samples also indicates a long spawn- ing season. Details of the suggested northward progression of spawning need clarification. In- tensive sampling of the water column and bot- tom offshore, at inlets, and in open estuaries could resolve these deficiencies in our knowledge of black sea bass life history. LITERATURE CITED Ahlstrom, E. H., and 0. P. Ball. 1954. Description of eggs and larvae of jack mack- erel {TracJucrus symmetriciis) and distribution and abundance of larvae in 1950 and 1951. U.S. Fish Wildl. Serv., Fish. Bull. 56:209-245. Arve, J. 1960. Preliminary report on attracting fish by oyster-shell plantings in Chincoteague Bay, Mary- land. Chesapeake Sci. 1:58-65. Bean, T. H. 1888. Report on the fishes observed in Great Egg Harbor Bay, New Jersey, during the summer of 1887. Bull. U.S. Fish Comm. 7:129-154 Plates MIL Bertolini, F. 1933. Uova, larvae e stadi giovanili di teleostei: familia 2: Serranidae. Fauna e flora del Golfo di Napoli, Monogr. 38:310-331, Plates 20-21. 1257 FISHERY BULLETIN: VOL. 70, NO. 4 BiGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53:1-577. Clark, J., W. G. Smith, A. W. Kendall, Jr., and M. P. Fahay. 1969. Studies of estuarine dependence of Atlantic coastal fishes. U.S. Bur. Sport Fish Wildl., Tech. Pap. 28, 132 p. Clothier, C. R. 1950. A key to some southern California fishes based on vertebral characters. Calif. Div. Fish Game, Fish Bull. 79, 83 p. Croker, R. a. 1965. Planktonic fish eggs and larvae of Sandy Hook estuary. Chesapeake Sci. 6:92-95. de Sylva, D. p., F. a. Kalber, Jr., and C. N. Shuster, Jr. 1962. Fishes and ecological conditions in the shore zone of the Delaware River estuary, with notes on other species collected in deeper water. Univ. Del. Mar. Lab., Inf. Ser., Publ. 5, 164 p. Fowler, H. W. 1945. A study of the fishes of the southern pied- mont and coastal plain. Acad. Nat. Sci., Phil., Monogr. 7, 408 p., 313 fig. GOSLINE, W. A. 1961. The perciform caudal skeleton. Copeia 1961 : 265-270. Greeley, J. R. 1939. Section II. Fishes and habitat conditions of the shore zone based upon July and August sein- ing investigations. In A biological survey of the salt waters of Long Island, 1938. Part II, p. 72- 91. N.Y. State Conserv. Dep. Suppl. 28th Annu. Rep. Herman, S. S. 1963. Planktonic fish eggs and larvae of Narragan- sett Bay. Limnol. Oceanogr. 8:103-109. Hildebrand, S. F., and W. C. Schroeder. 1928. Fishes of Chesapeake Bay. Bull. U.S. Bur. Fish 43(1): 1-366. Hoff, F. H., Jr. 1970. Artificial spawning of black sea bass, Cen- tropristes striatus melanus Ginsburg, aided by chorionic gonadotrophic hormones. Fla. Dep. Nat. Resour. Mar. Res. Lab., Spec. Sci. Rep. 25, 17 p. HuBBS, C. L., AND Karl F. Lagler. 1958. Fishes of the Great Lakes region. Revised ed. Cranbrook Inst. Sci. Bull. 26, 213 p. Lux, F. E., and F. E. Nichy. 1971. Number and lengths, by season, of fishes caught with an otter trawl near Woods Hole, Massachusetts, September 1961 to December 1962. U.S. Natl. Mar. Fish. Serv., Spec. Sci. Rep. Fish, 622, 15 p. Lyles, C. H. 1967. Fishery statistics of the United States, 1965. U.S. Fish Wildl. Serv., Stat. Dig. 59, 756 p. Marak, R. R., J. B. Colton, Jr., D. B. Foster, and D. Miller. 1962. Distribution of fish eggs and larvae, temper- ature, and salinity in the Georges Bank-Gulf of Maine area, 1956. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 412, 95 p. Massmann, W. H., E. B. Joseph, and J. J. Norcross. 1962. Fishes and fish larvae collected from Atlantic plankton cruises of R/V Pathfinder March 1961 - March 1962. Va. Inst. Mar. Sci., Spec. Sci. Rep. 33, 3 p., 5 tables. Merrimann, D., and R. C. Sclar. 1952. The pelagic fish eggs and larvae of Block Island Sound. In Hydrographic and biological studies of Block Island Sound, p. 165-219. Bull. Bingham Oceanogr. Collect. Yale Univ. 13(3). Miller, R. J. 1959. A review of the seabasses of the genus Centropristes (Serranidae). Tulane Stud. Zool. 7:33-68. MiTO, S., M. Ukawa, and M. Higuchi. 1967. On the larval and young stages of a ser- ranid fish, Epinephelus akaara (Temminck et Schlegel). [In Japanese, English synopsis.] Bull. Naikai Reg. Fish. Res. Lab. 25:337-347. Nichols, J. T., and C. M. Breder, Jr. 1927. The marine fishes of New York and southern New England. Zoologica (N.Y.) 9:1-192. Nybelin, 0. 1963. Zur Morphologie und Terminologie des Schwanzskelettes der Actinopterygier. Ark. Zool., Ser. 2, 15:485-516. Pacheco, a. L., and G. C. Grant. 1965. Studies of the early life history of Atlantic menhaden in estuarine nurseries. Part I — Sea- sonal occurrence of juvenile menhaden and other small fishes in a tributary creek of Indian River, Delaware, 1957-58. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 504, 32 p. Pearson, J. C. 1941. The young of some marine fishes taken in lower Chesapeake Bay, Virginia, with special ref- erence to the gray sea trout Cynoscion regalis (Bloch). U.S. Fish Wildl. Serv., Fish. Bull. 50: 79-102. Perlmutter, a. 1939. Section I. An ecological survey of young fishes and eggs identified from tow-net collections. In A biological survey of the salt waters of Long Island, 1938. Part II, p. 11-71. N.Y. State Cons. Dep. Suppl. 28th Annu. Rep. Presley, R. F. 1970. Larval snowy grouper, Epinephelus niveatus (Valenciennes, 1828), from the Florida Straits. Fla. Dep. Nat. Resour. Mar. Res. Lab., Leafl. Ser. 4 (Part 1, 18), 6 p. 1258 KENDALL: BLACK SEA BASS LARVAE Richards, S. W. 1959. Pelagic fish eggs and larvae of Long Island Sound. In Oceanography of Long Island Sound, p. 95-124. Bull. Bingham Oceanogr. Collect. Yale Univ. 17(1). 1963. The demersal fish population of Long Island Sound. Bull. Bingham Oceanogr. Collect. Yale Univ. 18(2) :1-101. Schwartz, F. J. 1961. Fishes of Chincoteague and Sinepuxent Bays. Am. Midi. Nat. 65:384-408. 1964. Fishes of Isle of Wight and Assawoman Bays near Ocean City, Maryland. Chesapeake Sci. 5 : 172-193. Sherwood, G. H., and V. N. Edwards. 1901. Notes on the migration, spawning, abun- dance, etc., of certain fishes in 1900. Bull. U.S. Fish Comm. 21:27-33. Smith, H. M. 1898. The fishes found in the vicinity of Woods Hole. Bull. U.S. Fish Comm. 17:85-111. 1907. The fishes of North Carolina. N.C. Geol. Econ. Surv. 2, 453 p. Sparta, A. 1935. Contributo alia conoscenza dello svillupo nei percidi. Uova ovariche mature di Epinephelns gauza L. e stadi post-embrionali e larvali di Epinepliehis alexandrinus Cuv. e Val. Mem. R. Com. Tallassogr. Ital. 224:1-12. Tagatz, M. E. 1968. Fishes of the St. Johns River, Florida. Q. J. Fla. Acad. Sci. 30:25-50. Wheatland, S. B. 1956. Pelagic fish eggs and larvae. In Ocean- ography of Long Island Sound, 1952-1954 p. 234- 314. Bull. Bingham Oceanogr. Collect. Yale Univ. 15. Wilson, H. V. 1891. The embryology of the sea bass (Serranus atrarius). Bull. U.S. Fish Comm. 9:209-277, Plates LXXXVIII-CVII. 1259 FISHERY BULLETIN: VOL. 70, NO. 4 CIS u O <1> u cS C C3 O >. CO CO cri o > 1^ 09 < Q U Ph s c Q u o F \- o (U c o 5 t: op o D E^o 1^ *T .- LU <§^ Si JP u O CO — CJ rj C7 0 CO ^ CO CO uS CN c>c-'0-'*"Otx — r^ir»coo-roc-5'0'o'0'orvo ocopcc>^_pcsc>|S)^, co*OcoN.C>;K*oio^cqco O^ — o — '-^^'-O — — ■'r----'-csco-o ^oj)rj)roo;>c^forpcj)cj)coo^c^coc^rococ^c^ooco O^K^'Oiop'^COrvcOOOCN'^CO(NOvCN'^*000 cococococoncococococococOcocDcocoe^C^coco d o ^7 CS CN 00 CO K CN CJ -^ ^ ^ CN -^ CN C^ O* CN CO CO Q -T ■— ■UOCOCpCOCO"**''*t^OUO I — .— --CvjC^'^'yC^CN ll*i.0OOOTv.l^N.N.CNC>. D> U) D> O) O) O D> c c c c c c c o o o o o o o O) D) O) D) c c c i: O ( 0> 0> O) 0> _ c c C c o o o o o oocyooocooococo^c/>cot>ocotoc/^c>ot>otoc/^ ceo O O 0) z z ^ CO CO oooo*^'^rocooqu^iO<)Nq wS £>»' 00 CO O CO "^ o O O" CN TJ- '^J- p p "^ ^O ^ •o *o ^ ^ ^O ^ ' — •—•— — ■— CM "Q^dQ CO cococococococococococo 00 '^Ch'^O^CSO'— osoo^o cj) c^co^pC7coc7coco^^cJ^op CM CMcqcMcqrxospCN^"^oo CO cocococorocoWcocococo CO c^ CO CO CM CM O 0) (U o P