Salnionid Ecosystems of the North Pacific

Edited by

William J. McNeil Daniel C. Himsworth

Published hy

OREGON STATE UNIVERSITY PRESS AND OREGON STATE UNIVERSITY SEA GRANT COLLEGE PROGRAM Corvallis, Oregon Library of Congress Cataloging in Publication Data Main entry under title:

Salmonid ecosystems of the North Pacific. 1. Pacific salmon-Ecology. 2. Fishes-Ecology. 3. Fish-colture-Northwest, Pacific. I. McNeil, \\'illiam J. 11. Himsworth, Daniel C., 1953- QL638.S%24 597'.55 80-14800 ISBN 0-87071-335-3

@ 1980 by the Oregon State University Press Printed in the United States of America All rights reserved

The Oregon State University Sea Grant College l'rogram is supported cooperatively by the National Oceanic and Atmo~pl~eric Administration, U.S. Department of Commerce, ljy the State of Oregon, and hy participating local governments and private industry.

This pi~blicationwas supported in part by the National Marine Fisheries Service.

Popl~latioll Structures of Indigenous S:~lmonidSprcio of rlic Pacific Norrhwcsr Frccl XI. Utter, Dollald Campton, Stcwart Grant, George hlilner, James Seeb, and Lisa Wishard

Electrophoretic techniques provide an investigator with the capability to collect Studies at the National blarine Fisheries large amounts of data with relative ease. Service Northwest and Alaska Fisheries No other known method is as convenient for Center (NWAFC) (Hodgins 1972) to define the making genetic comparisons among congeneric genetic structure of salmonid populations or conspecific populations. Utter et al. had a biochemical basis by the mid-1960s. (1974) and Allendorf and Utter (1979) These early studies (Utter et al. 1974) be- describe methods for collecting and inter- came the foundation of more recent applica- preting genetic data by electrophoresis. tions of electrophoretic techniques to Useful references on histochemical staining evaluate genetic variations of salmonid include Siciliano and Sha~(1975) and Harris proteins. Genetic variations defined by and Hopkinson (1976). electrophoretic studies have been reviewed by Utter et al. (1973), Utter et al. (1974), Statistical mcthods for analyzing data Utter et al. (1976), and Allcndorf and on gene variations among loci, individual Utter (1979). fish, populations and species come mostly from procedures developed in studying The more recent genetic researcl~at NWAFC DrosophiZa and man. The tla rdy-IVeinberg has centered on populations of SaZmo and principle (Stern 1943) is the foundation Oncorhgnchils sp. native to the Pacific North- of many of the analytical processes. It west. Although our understanding of the states that the expected genotypic propor- population structures of salmon and trout tions at a polymorphic locus are the square remains incomplete, our capability to of the allelic frequencies in large, random examine specific loci for genetic differences mating populations where no selective is increasing steadily. This paper summa- difference exists among genotypes. Signifi- rizes current knowledge on genetic varia- cant deviations from expected Ilnrdy-Keinhcrg tions in natural populations of sockcye proportions may be the result of factors salmon (Oncorizgnchus ner::a), pink salmon such as drawing the sample from a rniuturc (0. go~huschc*),chum sa lmon (0. kcta), of pop1~1:ltions h;lvirlg tliffcrc~lt;11 lc.lic chinook salmon (0. tsha?~?,t~cFa), coho sa 1111orl l'rcclucnci cs ; a ~I~I:I1 1 n111nl)cr of parcnt:; (0. kisutch), rainbow (stcelhcad) trout giving risc to all i 11l)rctl 11op~rJiitioll; 01- 4Salrno gtrirdneri) and coas ta 1 cut thro:it selection. trout (S. clarki ctarki). Statistical tcsts for significant diffcr- ences between two groups of inclividua 1s :I rc made either from gellotypic or nl lcl ic frcquencics. Tests incli~dcmeasuring Geographic patterns of genetic variations deviations from expected 1 lardy-Wei nl)el-g in protein systems among salmon and trout proportions, chi -squsrc tcsts for intlcprr~~l- populations can be used to identify different ence and measuring differences of :~llrlic populations. Protein systems used to evalu- frequencies from normal approsirnations OF ate differences among populations are chosen l~inomialdata. Significant cli ffcrenccs on the basis of their simple inheritance, occurring at a single locus or at scvcrnl and genetic differences among populations loci can bc used to diffcrenti:~te popul:~- are evaluated statistically by comparing tions. Summing thc diffcrcnccs of allclic the frequencies of genotypes or alleles. . frequencies at several loci ~)roviJes3 286 SALMONID ECOSYSTEMS OF THE NORTH PACIFIC basis to calculate the degree of genetic nllelic frequencies at polymorphic loci, simi1;lrity or differencc between two popula- ever1 though the saotples include fish from tions. Roger-.; (1972) and Nei (1972) different year classes. Allelic frequency describe two methods for doing this. Calcu- comparisons supporting this assunlption are lated values range from 0 (complete genetic presented in this report and describe dif- differcnce of loci examined) to 1 (complete ferent ages, generations and year classes genetic identity of loci examined). Dendro- of salmonid populations for which relatively grams are sometimes constructed from extensive data have been obtained. Enzymes matrices of such pairwise comparisons anlong used in these studies are listed in Table 1. populations (Sneath and Sokal 1973).

A1 lelic frequencies slust remain stable RESUI.TS AND DISCUSSION within a population over time to usc electrophoretic data to define genetic -Comparisons of Allelic Frequencies Among differences among populations and to' Year Classes and Generations estimate contributions of separate popula- tions to mixed fisheries. Estimates takcn Changes in ailelic frequencies of pink at different times may be combined to in- salmon have been observed at two locations crease the precision of statistical esti- in . Table 2 presents data for mators where frequencies rc~nainstable. loci (ACP-Z and MDH-3) from five consecu- Unstable frequencies require annual tive generations of an early run of pink sampling of spawning populations and render salmon at the Dungeness River (Aspinwall the electrophoretic data virtually useless 1974, Seeb and Grant 1976). These data, for managing fisheries on mixed stocks. along with data for two other loci (PGM-I and AAT-3) obtained front the Dungeness River Allelic frequencies are assumed to remain and floodsport Hatchery populations, provide stablc in large populations of salmonids an opportunity to study changes in allelic over succeeding generations and a~nongyear frequencies. classes. This assumption is based largely on data where adult and juvenile fist1 from There were no observed changes in allelic the same stream tend to express similar frequencies for the common alleles of AGP-2

---- Table 2. Protein Enzyme Sgstems Used in EZectrophoretic Cor~arisonsof Genotypes

Enzyme Abbreviation

Albumin Alb Alcohol dehydrogenase ADH Alpha glycerophosphate dei~ydrogenase AGP Aspartate amino transferase. AAT Creatine kinase Ck N-acetyl-B-D-Galactosamj nidasc GA 1. Glutamir-yyruvic transaminase G I3'r Isocitrate dchydrogenase Inti Lactate dehydrogenase I.Dli blnlntc dehydrogcn:~sr t.11)11 Flalic enzyme MI I'cpt idase PIiP 6 phosphogluconate dehydrogen:!se 6PCD Phosphoglucose i son~erase I'G I Phosphogl~rcomutase I'CM ~~hos~homannoscisomerase I'tl l Sorbitol dehydrogenasc SDll Tctrazolium oxj dese TO Transferri n Tfn Glycyl lcucine peptidnse G L Leucyl glycylglycine pcptidnse LGG Phenylalanylproline peptidasc I'hAP FRED M. IITrER, DONALD CAMPTON, STEWART GRANT, GEORGE MILNER, JAKES SEEB AND LISA WISHARD 287

Table 2. AZZetic Frequencies uith (in Parentheses) 95 Percent Confidence Intervals Over Sequential Generations for Four Loci i77 Pink ScrZzr~n. Collected frm Two Areas of Washington.

Location Year AGP- 2 PGM- 2 - Hoodsport 1973~. .959 (. 027) Hoodsport 1975' .956 (. 025) Dungeness 196g1 .987 (. 067) Dhngeness 1.971I .957 (. 030) Dungeness 1973~ .943(. 055) Dungeness 1975' .932(. 054) Dungeness 1977 .923(. 025)

Data from Aspinwall 1974. 2 Data from Seeb and Grant 1976. and MDII-3. Significant changes occurred for each of these biochemical systems. both within and between populations at the These four exceptional frequencies (out of PGM-2 and AAT-3 loci where frequencies in 37 data points) are somewhat higher than the the Dungeness population approached frequen- one out of 20 such observations that would cies in the tloodsport population at both be expected from chance alone at the 95 loci. These results were puzzling at first. percent confidence level in a stable, random However, an examination of Washington Depart- mating population. But the fish returning ment of Fisheries records indicated that fry to the Skamania tlazchery do not represent from the tloodsport Hatchery had been released such a population. The hatchery run was from the Dungeness Hatchery in 1976 (1975 derived from native fish of the Washougal brood year). Thus, the run returning to River and also included fish from the the Dungeness Hatchery in 1977 included Klickitat River approximately 50 miles up- survivors from the release of fry originating stream (personal communjcat ion, James from Hoodsport plus progeny from natural Morrow, Washington State Department of spawners (Foster et al. 1977). Results are Game). Each of the founding year classes readily explained by: 1) similar frequencies comprised a different mixture of progenitor of AGP-Z and MD2-3 alleles in the Dungeness stocks. Some differences among year and the Hoodsport pink populations, which classes of such a heterogeneous stock would should remain stable over generations; 2) therefore be expected des~itefactors tend- different frequencies of PGA-2 and AAT-3 ing to reduce these differences such as the alleles between the two populations; and 3) overlap among year classes and the capability an alteration of the PGKZ and AAT-3 of steelhead for mu1tiple- year spawning. frequencies in the 1977 brood year for pink Under these conditions significantly dif- salmon returning to the Dungeness Hatchery ferent frequencies between some col lections as a direct result of large plantings of are therefore anticipated. Evidence of Hoodsport pink salmon into the Dungeness allelic frequency stability among cohorts River. over time is srm in the MDH-3 frequencies of the juveniles sampled in 1973 aritl 1974 A second set of data on changes in allelic and in the adults of 1970 and 1977, consitl- frequcnci cs pertains to stcelhead. Suaoner- eririg the prctlo~ain;irlt thrre-year spar1 from run steclheatl from the Skamania Hatchery juvcni le to adult. 'I'hese data suggest that (Washington State Department of Game) on the the initial inequalities of al.~lic Washougal River near Vancouver, Washington frequencies expected among year classes may have been sampled for five years to study be approaching a point of equilibrium. the genetic effect of transplanting stocks. Figure 1 shows that most frcquencies of the Finally, a third set of data exists common allele of a given locus were not concerning a1lel ic frequencies froni the significantly different (P -05) between four loci in juveni lc nlld ad~iltwinter- rtln collections. However, the frequencies of steelhead trout from the Ii'asliington Ikpnrt- the common allele of MDH-3 from three ment of Game Cll;~ml)ersCreek tlatchery collections and of the common allele of TO (Tacoma, I'lashington) (Fig. 2). Ch:~ml,crs from one collection lie outside of the Creek tlatchery stock w:~s prinl:~rily dcr ivcd confidence intervals of two or more samples from a natural run of Chambers (:reek during 288 SALMONID ECOSYSTEMS OF THE NORTH PACIFIC

~igurc1. Allelic frequencies and 95 percent confidcnce intervals at five loci for juvenile and adult Washougal steelhcad trout from the Skamania Hatchery over a five year interval. the latc 1940s (personal communication, --Gcnetic Structures of Populations within Art West rope, Iyashington State Department - of Game). Only a single data point (MDH-3) was found to lie outside of the 95 percent This section describes the better known confidencc intervals. (A single aberrant aspects of the genetic variation among observation out of 30 lies well within the snlmonicls native to western North America. cxpected range for this level of confidencc). Five salmon spccics and two trout species The data indicate that the Oharnhers Creek arc considered. stec1hc:id stock presently appcilrs to be close to :in cquilibrium state with rcgard Tllc numlwr of currently known polymorphic to allclic frequencies. systems in tl~esespecies (Appendix Tablc 1) is approximntcly doublc the number of l'hc three sets of data support the assump- variant systcl~~sknown in rainbow trout five tion of consistencies in nllclic frcquencics years ago (Uttcr et al. 1974). This suh- betwecn gcncrations and anlong (overlapping) stnntial increasc in our knowledge OF year classes. It is important that allelic gcrlctic viiriation, and thc likcljhood of frequency records over time be maintained, add it ional variants bcing rcvealccl, suggests whcrevcr possiblc, for salmonicl populations. that n largc reservoir of genetic variation Such data are necessary to document the exists, which is potcntinlly availal~lefor level of gcnetic consistency within popula- exan~irlirlg genct ic structures oE s;ilmon ids. tions, and can provide valuablc and unique insights into the causes of changes in popu- lations as they occur. FRED M. UTTER, DONALD CAMPTON, STEWART GRANT, GEDRGE IIILHEF!, JAPES SEEB AND LISA \.IISHArlD 289

-a- -a- 75t -* -Juvenile

Figure 2. Allelic frequencies and 95 percent confidence intervals at four loci for' juvenile and adult steelhead trout from the Chambers Crcek Hatchery over s seven yesr interval.

alleles than in populations from western : Because more prccise Alaska and Asia (Utter et al. 1974). These knowledge was needed about the origins of differences were also found in non-.tnadro- sockeye salmon harvested in the Japanesa I mous populations of sockeyc (kokanee) : It high seas fishery, this species was the is now assumed that high frequencies of initial primary target of immunologi.ca1 and these two allelcs are typical of sockeye electrophoretic studies. populations in the southern part of this spccies North American range. Subseqilen t Early electrophoretic data on sockcye samplings, however, have proven that this were not vcry useful becausc alleles of two ;issun~ptionis an ovcrsimpl if icat iori (I'ie. hil~hlypolymorphic systems (LDH-4, E'GM) 3). Data from popillat ions s:~n~plcdbout 11u:i rd occurrcd at similar frequencies in both through the 1:rnser R ivcr clr-a irlagc. into American and the ~siclticstocks (Ilodgins !\lashingtor1 ilrc cons istrnt wi tll thc corlccpt et al. 1969, Utter and tlodgins 1970, of a singlc major popul;~tiongl-OLI~ (Utter Hodgins and Utter 1971, Altukhov 1375). and tlodgins 1970, tlodgirls nrlcl Ilttcr 1971). However, genetic differences among sockeye flowever, two col lcct ions of nn:~tlromo~~s populations have subsequently been observed sockeyc from Colun~hiaKivrr tl*ibut;~ries to increase southeastwardly from the Copper (blay and Utter 1974) and collections of River drainage in Alaska into rcgions wherc kokanee from Issaquah Creek near Seattle stocks do not intermingle significantly (personal communication, James Secb, Wash- with Asiatic stocks. Sampling of sockeye ington Department of Fisheries) have allellic salmon populations from the Skeena frequencies that differ from any other known River (northern British Columbia) southward sockeye population. Fish from these latter to and the Quinault Kivcr locations possess high frequencies of PCt: (Washington coast) indicated higher fre- variants typically found in populations from quencies of the common LDH-4 and PGM western Alaska and Asia. The Issaquah Creek 290 SALMONID ECOSYSTEMS OF THE NORTH PACIFIC

Figure 3. Major population units of sockeye salmon identified through different frequencies of allelic proteins. kokanee population is also more polynlorphic the Cook Inlet region of Alaska (Grant et al. for the same LDE-4 allele that occurs in the in press), the Port Alberni region of British more western populations. Columbia River and Columbia (Allendorf and Mitchell 1977)) and Issaquah sockeye also possess variant the drainage near Seattle (per- alleles that have not been observed else- sonal communication, James Seeb, Washington where in this species. MDH-3 and TO Department of Fisheries) have shown considerably variants were found in Columbia River sock- greater complexity of allelic frequencies eye, and a very high frequency of an LDH-2 within drainages than expected. Hetero- variant was found in Issaquah Creek kokanee. geneity among three drainages supporting sockeye in Cook Inlet was lowest in the river The occurrence of previously undetected system containing a single lake (Kasilof alleles in the southeastern extreme of River) and highest in two river systems con- this species' natural range is puzzling. taining a complex of lakes (Kenai and Susitna One possible explanation is a diphyletic Rivers). The Port Alberni investigations origin of Washington's sockeye populMjons were limited to three lakes in a single following the last glacial era (about drainage where differcnces in the frequcncies 10,000 years ago). One group may have of EJGM, MDII-3 :~nd SDIl allelcs were sufficient origir~atedfrom northern populations with to distinguish fish collected from each of the second group descending from southern the lakes. The most detailed data were stocks that may hnvc existed during the obtained from the Lakc Washington drainage last glacial period. (personal communication, James Seeb , Wash- ington Department of Fisheries), which sup- The siniilarity.of allelic frequencies ports both anadromous and non-anadromous observed within a particular region coupled populations including Issaquah Creek ko- with low average heterozygosity observed kanee. There are. large genetic differences in sockeye when compared to other spccies among populations of Lake Washington sock- (Utter et al. 1973, Allcndorf and Utter eye. Samples from two consecutive years 1979) caused us to expect little variation suggest that simulataneous spawning of in allelic frequencies among different anadromous and kokanee populations in Is- spawning populations within a given drainage. saquah Creek 'occurs without significant Recent surveys of sockcye populations in gene flow between groups. FRED M. UTTER, DONALD CMPTON, STEWART GRANT, GEORGE MILNER, JAMES SEEB P.ND LISA WISHFRD 291

grcat surprise. h%at surprising are thc ---Pink Salmon: Pink salmon populations is are characterized by a substanti;tl degree . high levels of hetc~rozygosityin pink of randomness of allelic frequencies over salmon populations because periodic "bottlr- large geographic areas. Some genetic dif- necks" in the absence of gene influx from ferences have been observed among odd-year other populations are expected to reduce populations in Alaska and the Pacific North- levels of heterozygosity (Crow and Kimura west, including the Fraser River. An /OH-B 1970). It is possible that sufficient variant is virtually al~sentfrom even-year straying occurs among pink salmon popula- populations in Alaska, but occurs at varying tions to maintain reasonably high levels frequencies in odd-year populations (Aspinwall of heterozygosity, but that this straying 1974, Johnson et al. 1976). Frequencies of is insufficient to overcome the periodic AGP variants appear to be significantly alteration of a1 lel ic frequencies brought lower in Fraser River arld Puget Sound popu- about hy the "bottlenecks .I1 lations than in either even- or odd-year populations returning to Alaska. ' Chum Sale: Genetic studies of chum salmon populations were initiated at the For a particular year class, differences same time as sockeye studies because the observed between even and odd years within Japanese fish both species extensively on a particular stream are generally greater the high seas in the same areas. Chum than differences that occur between streams. salmon have played a subordinate role to This peculiar distinction has been recorded sockeye in studies because they are by three independent investigations not as important to American fishermen. (Aspinwall 1974, Seeb and Wishard 1977a, tlowever, some genetic data on Asiatic chum Johnson et al. 1978) and appears to be a salmori populations have been published direct reflection of the rigid two-year life (Numachi et al. 1972, Altukhov 1975). cycle of pink salmon. Differences contrast- ing ~dd-~karwith even-year fish include Three major groups of chum salmon popu- general ly higher frequencies in odd-year lations are evident from the allelic distri- fish for variant alleles of the MT-3, bution of two polymorphic loci that both MDII-3 and PGM loci and lower frequencies Asiatic and American workers examined. of MDH- 2 variants. Variants for LDH-Z are found in virtually all populations examined from southeastern Attributes of pink salmon populations Alaska through Japan (Utter et al. 1973, that distinguish them from other salmonid Altukllov 1975, Seeb and ifishard 19775). ?!US-3 species include: 1) relatively high amounts variants are found at polymorphic frequencies of genetic variation within populations (frequencies greater than .01) in Asiatic (Allendorf and Utter 1979) ; 2) reasonably collections (Numachi et al. 1972, Altukhov large fluctuations of allelic frequencies 1975, Seeb and lvishard 1977b'). Yeither among breeding groups within limited geo- LDH-Z nor MDH-3 variants have been observed graphic areas; and 3) minimal patterns of from the Fraser River southward despite similarity that are useful for defining intensive sampling of Puget Sound populations broad geographic population units for even recently. Thus, three major groups -- or odd-year populations. It is possible tentatively defined as Asiatic, Alaskan and to interpret these characteristics within American -- are identified on the basis of the context of the species' life history. variants at two loci (Fig. 4). Discontinuity Pink salmon spawn largely in the lower of sampling precludes any precise definition reaches and intertidal areas of small of bo~~ndariesat this time. coastal streams that are periodically subjccteii to drastic fluctuations in Othcr variants have bccn ~fsefulfor tenrperature, and quality and quantity of examining genetic structures of cll~lnl sa 1111on water. Sorne streams are also occasionally populations in greater detail. Studics by modified through shifts of spawning strata Uttcr et ill. (1973) indicrrtcrl :a lack of resulting from earthquakes such as that genetic variation in cl~umsrrlri~on from centered in the Prince William Sou~ldarea Washingtori based on observations of -70 loci. of Alaska in 1964 (Thorsteinson et al. 1971). However, morc rcccnt st~~dieshnvc rcvca I cd Population sizes vary drastically under at least six polymorphic loci in W:~sl~ir,gton such diverse conditions (Royce 1362). populations (AAT-2, 2, and 3; :UH; I::.'T; C:r...':rl) . Allelic frequencies of populations subjected There appears to be tli st i r~t~>op~~litt ions OF to periodic "bottlenecks" are expected to Puget Sound clium salmon (Sccb and Ni sh;r I-d fluctuate randomly as observed in pink 1377h). Different frcquenci cs of I.!lF/ a l l clcs salmon (Crow and Kimura 1970). Thus, the tend to separate stocks from north xnd so~ith relatively large degree of randoniness Puget Sound, and an important late run OF observed for allelic frequency distributions fish entering the Nisqual ly Rivcr is among pink salmon populations comes as no characteri zed by except ionally high t'rcqucn- 292 SALMONID ECOSYSTEMS OF THE NORTH PACIFIC

Chum

Figure 4. Major populatioli units of churn salrnori identified through different frequencies of allelic proteins. cies of 6PDG variants. A recent study of characterized by uniformly high frequencies chwn salmon populations of the Yukon and of PGI-2 variation and usually by the Kuskokwim drainages (Utter 1978) has indicated presence of PGM-2 variation (Utter et al. an apparent randomness of allelic frequencies 1976). Only low frequencies of PGI-2 within these two drainages except for a variants and no PGf4 variants have been slight and consistently different frequency observed in the rather extensive sampling of PMI variants between the drainages. of Columbia River and Puget Sound chinook populations. Preliminary data from the Chinook Salmon: Chinook salmo~ipopula- east and west coasts of Vancouver Island tions were examined intermittently through suggests that these characteristics do not collection efforts that focused on sockeye extend into British Columbia populations. and chum salmon. This work, along with reports from other workers, resulted in a Chinook salmon runs from the Columbia relatively early awareness of two polymorphic River have been examined for several years. systems in chinook salmon, TO and MDH-3 Unlike coastal runs, no clearly identifiable (Bailey et al. 1970, Utter 1971). A stronger groups of populations have been determined. cmphasis has subsequently been directed Fa1 1 chinook salnlon fro111 hatcheries on the toward studying chinook salmon populations, lower Colurr~l)i;tRiver share theC:conlinorl a1 lclcs particularly those of the Pacific Northwest. of thc TO ant1 I'MI loci. This rnny be more a These offorts considerably expanded the rcflectiori of the incliscrinlinnnt exchange nunibcr of identified polymorphic systelns oi cggs among hatchcrics than a close and a more comprehensive understanding of affinity of ancestral stocks (Simon 1972). the genetic structure of chinook salmon Sprjng chinook fronl the lower Columbia populations (Flay 1975, Kristiansson and River and its major tributaries tend to MacIntyre 1976, Utter et al. 1976, blilner bc quite di stinct from fall chinook ret~lrnilig 1978, Utter 1978). to this region, principally through con- siderably higher f rcquencies of the common Chinook salmon populations that extend 7'0 allele. at lcast from the !.lad River drainage in California through the Quillayute drainage Chinook populations of the upper Columbia in Washington share similar genetic and Snake r-ivcrs are tentatively identifiable characteristics. These populations are either through distinctive frequencies of FRED N. UTTER, DONALD CARPTON, STEWART GRANT, GEORGE YILNER, JAF?ES SEE0 AND LISA WISHARD 293

PPVI and TO alleles or the prcscnce of low allclcs may be more a reflection of conlplex frequency variants at other loci that have bacteriostatic properties of different not been seen in other populations. l'hese trarlsfcrrirl alleles than of ancestral rela- differences must be regarded as provisional tionships (Suzumoto et al. 1977, I'ratschner because they are generally based on single 1978). observations separated by large distances. Most other loci screened from collcctions A few chinook salmon popula t i ons have of taken throughout the Pacific also been examined in the Yukon ant1 Northwest are mollornorphic. A variant a1 lcle Kuskokwim rivers in Alaska. I'herc appears of LUH-4 occurs at moderate frequencies (up to be genetic similarity characterized by to .lo) in streans in south Puget Sound and high frequencies of the common allele of tlood Canal, but is virtually absent in polyntorphic loci among the populations collections fron othcr Pacific Northwest from tributaries of both rivers. These streams (Play 1975). Two unusual variants populations were dissimilar from all, other at the PGM and LDH-2 loci occurred at chinook populations examined except for polymorphic frequencies (greater than -05) populations in the upper Snake River drainage. at the Feather River Hatchery, California (Utter unpublished data, Flay 1975). Both It is interesting to compare genetic of these variants may occur widely in the variation within and among chinook salmon southern range of coho. At least some populations relative to the overall distri- coho populations from the Fraser River and bution of the species. The Columbia River Puget Sound have polymorpllic frequencies is in the center of distribution with of variant alleles for AGPD-3, ASPD-4 and major populations both northward and south- two peptidase loci (Seeb u~publisheddata, ward. The Yukon and Kuskokwim rivers, on Utter unpublished data). Present data are the other hand, are near the northern insufficient to define any geographic extremity of the range of chinook salmon patterns for these variants. Coho salmon in North America (Atkinson et al. 1967). persist as the species wit,^ t.>e lowest It h3s long been postulated that populations incidence of detected polymorphism in spite near the center of the range should have of these additional polymorphic loci. fluch higher levels of genetic variation than of the developmental work on methods to populations near the periphery because of detect polymorphism has been directed at the greatcr potential gene flow (Mayr 1970). this species because of its apparent high Genetic data from chinook salmon support incidence of monomorphism. such a hypothesis. The Kuskokwim River and Yukon River populations are generally The low levels of genetic variation distinguished from those of the Columbia observed among coho populations can be River by lower amounts of genetic variation readily explained if this species had among individuals and populations. Only discontinuous distribution and restricted those populations at the inland extremities habitats. These factors would contribute of the Columbia River drainage (which them- to reduced gene flow among populations ;is selves are peripheral populations within a in sockeye salmon, \%.hichhas a similaf lor< large river system) resemble the Alaskan average heterozygosity (.A1 lendorf and Utter populations with regard to reduced levels 1978). However, coho salmon, more than any of genetic variation. of its congeneric species. occupies n contin~runiof populations in diverse h:ll>i tats over a broild gcogl-;tl)hic range (:\ro and --Coho - Salmon:-- The coho salmon lies the lowest average heterozygosity vn luc oF ttl, Shepard 19h7). 'Thc spec i cs has 1.c;itli ly five Pacific salmon species in North An1eric.a adapted to t rarlsplnrit;~ti or1 ant1 tin tcllel-y (Allendorf and Utter 1979) with only :I contlitiorts (prr-sorlal comrrrcrnirat ion, It. single highly polymorphic locus - transfcr- I'rcsscy, Nntj on:^ 1 .\Inrille I:istierics Scl-vice) rin - among 24 loci examincd in a broad and is present 1y mol-c abutid:int thnt~ survey of populations from California through historically over m~rcli of its r;lrigc clue to Alaska. Populations from the Columbia and hatchery propngat ion. It a~pe:it-s ;~rlon;~lo:is Fraser rivers can be distinguished from that such an ubiquitous and ad;ipt;~blcspccics other populations on the basis of a single as the coho sllould cxprcss litt lc gcneric allele occurring in high frequencies in variati.on in relat ion to othcr sri lmon spc~:i.cs. the former group contrasted with varying, Two opposing possi bi 1 it.ics arc suggcstcd: but somewhat equal,. frequencies of three 1) the level of gcnct ic variation ct~rre~ltly alleles in the latter group (Utter et al. indicated: by protein loci is not ;I valid 1970, Utter et ak. 1973, May and IJtter 1974, reflection of gcnetic variation ovrr the Hay 1975, Seidel 1976, Utter et al. 1976, remainder of the coho salmon gcnome; and Suzumoto et al. 1977). Experimental data 2) the coho salmon h:~s cvolve~ln gcnomc suggest that the distribution of transferrin possessing little gcnctic v;iriatioll hut a h?gh ly adaptable nhenotypc. 294' SALMON1 D ECOSYSTEMS OF THE NORTH PAC1FIC

Rainbow Trout: The genetic constitutjon northern California. Moderate to low of rainbow trout has been st~rdiedmore frcqucncies of LDH-4 varialits and moderate intensively than any otlier snlmonid species. i'rccluencics of 7'0 variants typify this Considcr;lbly greater genetic variation of group. An inland group, found exclusively proteins has been reported for rainbow in the Fraser arid Columbia river drainages trout than for Pacific salmon (Utter and east of the Cascade range is identified through Hodgins 1972, Utter et al. 1973). !lore very high frequencies of LDH-4 variants recent studies of anadromous populations and low frequencies of 70 variants. This (steelhend) in Washington involving more major division presumably reflects two than 30 loci (Allendorf 1975) have supported distinct lines dating into the last glacial this coriclusion. Continuirig studies of era (Allendorf 1975). The inland group is rainbow trout place emphasis on more precise postulated to havc descended from fish definition of Columbia River populatior~s migrating into a large freshwater impound- (hlilner 1977, 1979). A more extended geo- ment resulting from the glacial diversion graphic survey of the gross population of the upper drainages of the Columbia structures of steelhead populations hosed and Fraser rivers. The coastal group on data from two polymorphic loci (LDH-4 presumably descended from Asiatic or and TO) has recently been published (Utter American stocks that existed outside tlte and A1 lelidorf 1977). glacial mass.

The current data for frequencies of LDH-4 Neither the tendency toward anadrolny nor and TO variants define at least two major the timing of upstream migration appear to geographic units of rainbow trout popula- be characteristics of distinct evolutionary tions (Fig. 5). A coastal group tentatively lines. Allelic frequencies of both resident extends at least from Kodiak Island south- and migratory populations of a particular ward through the Plad River drainage of region were invariably similar (e-g. steel-

Major Sf ee/heud Popu/af/bn Groups

Figure 5. Major geographic units of rainbow trout popt~lationsdcfincd I)y frequencies of LDff-4 and TO variants. (From Utter and Allcndorf 1977)- FRED M. UTTER, DONALD CAPPTON, STEWART GRANT, GEORGE MILNER, JAWS SEEB AND LISA WISHT,ED 295

head of the upper Fraser River and "Kamloops" its confluence with the Snake River and rainbow trout), while summer and winter-run some poptrlations of the Snake River steelhead of a particular drainage tended to drainage. resemble one another more than they resembled populations of adjacent drainages (Allendorf ---Cutthroat Trout: Data have been collcctcd 1974; Thorgaard 1977a, 19773. Thus, the sporadically over a number of years from terms "steelhead, "summer-ruri," and coastal cutthroat trout, primarily in con- "winter-run" are useful for the description junction 14i th ongoing studies of rainbow of life history patterns of a particular trout. Most of these observations were population, but are not an indication of directed toward determining biochemical close evolutionary relationships among differences between cutthroat trout and populations of different areas. rainbow trout (Utter et al. 1973, Utter et al. 1976, Campton 1980). Cutthroat The high frequency of LDH-4 variants in and rainbow are very closely related, occur the inland rainbow trout group separates sympatrically in many waters and are it from the. golden trout (SaZmo aqunbonita) morphologically indistinguishable in early and red-band trout. Present evidence stages of life history. The ability to indicates that these inland trout groups electrophoretically identify individuals :of are closely related to rainbow trout (Gold the two species and their hybrids has 19771, but the absence of LDH-4 variation proven to be a valuable asset in. stream (Utter and Allendorf 1977) indicates a surveys where both species occur syrilpatric- divergence from the inland rainbow trout ally, and is routinely used by the populations that predates at least the last Washington State Department of Game (WSDG) period of glaciation.' for this purpose (Burns et al. 1977). Although we have studied coastal cuttnroat The boundaries of the coastal group are trout at the species level for a number of uncertain because of the absence of sampling years, virtually nothing was known about at the northern and southern extremes of the genetic structure of populations of distribution. The aberrant chromosome this species prior to 1976 when a fornal counts Tliorgaard (19773) observed in steel- investigation was initiated' to examine head from the Elad River in California indi- this question in conjunction with the cates a division at this point that was establishment of different hatchery lines. not reflected in the protein data. Non- These studies have been described by anadromous hatchery strains Allendorf (1975) Campton (1980), and have now reached the examined showed differences between inland point where some interesting facts have and coastal groups. These populations emerged co~~cerningt ~e population structure presumably descended from fish taken from of coastal cutthroat'trout. the PlcCloud River drainage of California during the last century (LlacCrimmon 1971) Although cutthroat trout occur from and appear to reflect a third major group northern California through Prince Wi1 liam of rainbow trout populations that may Sound, Alaska, the present data are restrict- converge with southern populations of the ed to two proximal regions of xestern coastal group. Washington -1lood Canal and northern Yuget Sound. These two regions have been examined The major population groups of rainbow in considerable detail. Some collections trout described to this point are virtually included tributary streams within minor certain to have a considerable degree of drainages and adjacent areas in a given internal structuring that will become stream. Coastal cutthroat trout have apparent as more populatiolls and loci. are prove11 to hc tlic most polylnorphir s;i ln~onid cxamined, and as complementary data arc spccics stutlictl to tlatc. O\+r 50 percent collccted by other methods. Such str-ucturi~ig of more tli:~ri -30 loci cx:~lninctl WCI-c poly- has nlrcady been indicntrd in the coastal nrorpl~ic in one or more of tlie popular i on5 group through the cytogenetic studies of exanlincd. Averiige lirtcroxygos i ty \,;I lucs Tliorgaard (1977a, 197731, and the presence . ill this group of pupu1:ltions are we1 1 in of apparently unique protein variants in excess of .lo. certain drainages (Allendorf 1975). Direct evidence for a substructure of the inland This detailed sampling and testing, group was recently found in a comparison couplerl with known uniquc aspects of tho lifc of three collections from widely separated history rclativc to othcr inctigcnous ss1n1o11- areas of the upper Columbia River drainage ids, has led to new knowledge of this through differing frequencies of peptidase spccics' biology that has dire<-t imp1ic:t- variants (kfilner 1977). The pattern of tions for its nlan;igolacnt. ,I major- finding variation suggests a clear separation of from the protein data \<:IS ;I gelieral llnity populations from the Columbia River above of populations within cnrh I-cgion cor~ti-;~stc~I 296 SALMONID ECOSYSTEMS OF THE NORTH PACIFIC with a clifference between regions of a chemical genetic examinations of these fishes magnit~rdesimilar to that separating the (Cold 1977, Cnrnpton 1980). SaZrno cZarki coast:~land inland groups of rainbow trout. presently inclucles a rather broad assortment The linear distance separating the two of population units that are bound together regions is small, but the coastal distance through morl)hological similarities but appear is large because of the hundreds of miles to be rather widely diverged on the basis of of coastline separating the Stillaguanish cytological, biochemical and geographical River drainage at the southern extreme of evidence. Althougl) it is beyond the scope the north Puget Sound region and tlood Canal of this paper to review this situation in Tagging studies show that coastal cutthroat detail, it is irnportarrt to emphasize that the trout avoid open, deep water areas such as coastal cutthroat trout is a distinct taxo- those separating lfood Canal and north Puget nomic unit from all other groups of Satrno Sound (Johnston and Mercer 1976). This indigenous to western North America, includ- characteristic is consistent with the ing other groups of SaZmo ctarki. The genetic difference separating pop~~lations diploid chrolnoson~en~unber (2N = 65) (Gold of the two regions. 1977) is four higher than any other cutthroat trout group. Coastal cutthroat are restricted The apparent absence of gene flow between to coastal drainages from Alaska through Hood Canal and north Puget Sound populations California and are largely sympatric with is sufficient evidence to preclude the use coastal steelhead trout populations. A com- of a single hatchery stock derived from parison of the genetic similarities (measured either region to enhance fishing in both by about 30 protein loci) among coastal regions. Interbreeding of native and cutthroat trout, "west-slope" cutthroat trout strongly diverged exogenous fish carries (from Pacific drainages of the Columbia the risk of disrupting highly adapated g?ne River east of the Cascade Crest), "Yellow- pools, giving rise to suboptimally adapted stone" cutthroat trout and coastal steelhead progeny, and an artificial dependence on trout is particularly interesting. Each of continued hatchery supplen~entationto these four groups are about equally diverged support the fishery (see Reisenbichler and from one another, indicating a similar time blclntyre 1976). . interval since comnlon ancestry (Campton, 1980). It appears, then, that at least One collection of yearling fish from Howe some cutthrost trout lineages have diverged Creek of the Hood Canal region did not con- to the species level, and that a thorough form to the above picture of much greater multidiscipli~laryreexamination of the genetic similarity within a region than systematics of Pacific species of SaZmo is between regions. In fact, the collection appropriate at this time. formed a separate cluster from the remainder of the Hood Canal populations and the cluster of north Puget Sound populations (Campton GENERAL CONSIDERATIONS REGARDING DATA 1980). This population was also characterized AND BREEDING STRUCTURE by a low average heterozygosity value relative to all other anadromous coastal cutthroat The present data amplify earlier conclu- trout populations examined. These peculiari- sions from a less extensive survey of three ties were consistent with a hypothesis of species (coho and chinook salmon and rain- one or more population "bottlenecks" bow trout) where a uniqueness of population affecting the allclic frequencies and levels structuring within a species was noted of genetic variation of this population (Allrndorf and Uttcr 1979). It has become couplcd with an absence of gene flow from i11crc;is ingly cvi dent that gcnetic data adjacent popl~lations. Such a sitwition derived for one species are i!~suFficient would be expected in non-:lnsdro~nous (rrsident) for drawing infercnccs regarding structures populations in a small stream possessing a of closely rclatcd s;~lmonidspccics. Dif- small pop~ilationand experiencing little or fcrcnces in genetic structllre becon~emore no outside gene flow into the populations. evident frorn the data presented for seven Campton (1980) postulated that cutthroat species in this rcport,which also discusses residing in Howe Creek represcntrd such an possiblc causes and pr;lcticnl implications isolated population. This hypothesis was of gcnetic diffcrences. In this section we verified in 3 group of older fish from Howe will discuss four questions related to Creek, that had the same genetic charactcris- gcneral aspects of gcnetic differences. tics as young of the ycar. What are the principal cadses of these A question concerning the systematic --difGrent patterns of variation in genetic status of cutthroat trout native to different ---structure? This qucstion lies at the center areas of the western United States has arisen of one of the fundamental controversies of as a result of cytogenetic and recent bio- modern biology, i.c., is such eenetic varia- FRED M. UTTER, DONALD CAMPTON, STEWART GRANT , GEORGE MILNER, JAMES SEEB AND LISA WISHARD 297

tion predominantly a reflectior~of processes invalidate this broad generalization for of natural selection that directly affect salmonids and suggest that it may be. the protein loci examined, or is most of generally inappropriate. this genetic variation more an echo of random events such as genetic drift or What arc the ultimate limits of popul?- migration? The controversy is presently lation definition based on genetic varia- unresolved, in spite of the intense effort tions of proteins? This relate'; and thought that has been directed toward directly to the preceding question. The asolution for more than a decade, mainly major factors inf lucncing the frequencies because of the difficulties in obtaining of single gene variants in a population reliable estimates of sizes of natural inclr~de1) the number of breccling indiviclu- populations (Lewontin 1974). 111 large, als in a particular generation; 2) the dcyrcc random breeding populations, the effects of genetic isolation from other populations; of natural selection are much more 3) the amount of time elapsed since gene apparent for loci where different coeffici- exchange occurred from other populations; ents of selection exist for a set of geno- and 4) the differential inf lucncc of types, while random factors tend to pre- natural selection on variants of a particu- dominate in the distribution of genotypes lar protein. in small populations (Nei 1975). The limited fecundity and reasonably small We have already explained that the first population sizes of salmonids places this three factors probably account for most of group of organisms among those where geno- the variation observed among salmonid typic distributions of natural populations populations. Each factor is interrelated would be principally a reflection of random and varies from species to species and, to factors. The above argument does not some degree; among populations within resolve the issue of selection vs. neutral- species. ity of polymorphic protein loci, but does provide a functional framework for inter- Pink salmon populations are a good preting most of the allelic frequencies example of the apparent complexity of those observed in natural populations of salmonids. genetic interactions. Pinks may be influ- Thus, the distribution of protein variants enced sufficiently by periodic restrictions described in this report for seven salmonid of population size within local breeding species (with the possible exception of units to override the potentially stabilir- transferrin variants of coho salmon) has ing effects of gene flow between units from been interpreted to result predominantly possible straying. This combination of from random processes. This interpretation genetic'dynamics could explain the relatively is supported by the following observations: high amount of genetic variation observed in 1) the major patterns of distribution pink salmon populations and the minimal degree apparently reflect geographic rather than of geographic structuring. The effects of environmental variables; 2) distinct allelic time since gene exchange are apparent through frequencies are observed in even and odd- the larger differences seen betl~eenycar year pink salmon from the same stream; 3) classes than are generally observed within consistent allelic frequencies are observed them. The consistency over generations of among age groups, year classes and genera- gene frequencies of Wasliingtorl pink salmon tions in "closed" populations; and 4) populations (excepting changcs brought about successfully transplanted fish obviously by transplantation) occurrecl within :> pcrio~l influence gene frequencies of the next when none of thcsc population% were affected generat ion in "open" populations' (see Table by drastic rcd~rctior~sof hrcccli~~gi n~livitlua 1s. 2) - 'rhe potential il~stabi1 ity of ;\lnskan pinh salmor~popul;~tioas mi nimi :cs- tlre v~~lurof It is apparent that nllclic frequency gene freq~rcrlcyc[ati~ for cst inlat i II~;rnrc.ht r;~l data are highly useful indicators of the rc1:ltionships among populations or dcf i ni ng genetic structures of salmonid species and broad geographic population units. Ilo~~cvcr, that the structure of each specics is stable local differences anlong ~>olx~lations. unique. This distinct structuring precludes regardless of c:lirsc , arc st11 1 potclltial ly generalizing from one specics to another. useful for measuring interact ions ;inlong Based on protein data collected from dif- these populations (sce Table 2). ferent populations of American eels (Anguilla rostmta), Williams ct al. (1973) statcd Anaciromous populations of coastal cut- ". ..the recognition of separate Mendelian throat trout, in contrast to pink salmon, populations on the basis of significantly tend to form distinct population units on a different allelic frequencies in a number regional basis that appear to rcflc.ct of other commercially important species is ancestral rclat ionships. 'I'h is contrast is highly questionable. .'I Our findings surprising at first glance bcc;l~rseof lift history similarities of the two species. Will protein data replace traditional Both species spawn in smnllcr coastal methods-- for studying l~opulationstruct=? stl-ealns and, in a given generation, are '1'11e nnique propert i & of genetic variants potentially affected by drastic retiirctions of proteins (i.e. simple inheritance coupled in the number of spawning individuals. with case of detection) permit genetic However, a ni~niber of differences exist be- definitions at the individual, population twecrl the life histories of the two species and species level that were previoltsly that appear to influence their genetic ir~~possibleto obtain. These capabilities structures. Coastal cutthroat trout arc have opened up a new dimension in the study capable of breeding niore than once and fish and understanding of natural populations. of different ages participate in the breeding 'rllese exprrndecl captbi 1 it ies are general ly activities of a given year. Thus, an cornplementary to existing nlethodologies. apparent "hottleneck" for coastal cuttllroat Extcrnal marks and tags remain indispensable trout is not nearly as drastic an event fl-om tools because of their visibility. Coded the point of view of gcne frequency ,stability wire tags combine relative ease of applica- as it is for pink salmon. A secon,l major tion and detection with a high potential differe~lccis that coastal cutthroat trout information content. Intrinsic differences never venture into deeper saltwater and such as those observed in scale characters the potential for gene flow over large (Anas and Flurai 1969) and trace element distances is therefore much more restricted. composition (Calaprice 1971) are primarily Both of these life history diffcrenccs are indicators of environmental differences consistent with the gene frequency patterns and, as such, are completely complementary observed in coastal cutthroat trout. to genetic variants of proteins. Genetic variants of proteins are therefore, just These two examples demonstrate that one of the valuable tools available for knowledge of the ultimate resolving powers obtaining information about structures of gcne frequency data in natural popula- and movements of fish populations. It is tions is dependent upon knowledge of the important that both the capabilities and species' life history in question. This limitations of a particular tool be properly dependency does not reduce the value of understood so that the appropriate set of gene frequency data for defining population rncthods can be implemented in a particular structures. Rather, the dependency works study. Nevertheless, genetic data based in both directions. As more is known about on protein variation remain singularly the genetic structure of a species, more superior to any other existing method for can be dcduced about its life history defining the genetic structures of populations. patterns and vice versa.

How much data are required before a pattern of variation among populations of a species becomes evident? Again, there is General consistencies of allelic and no single, simple answer to this question phenotypic ,frequencies at many electro- any more than to the more general question phoretically detected loci have been of the number of points required to resolve observed within closed salmonid populations trends from graphic projections of any set 1) throughout life cycles; 2) over succes- of data. A logical approach is to initially sive generations; and 3) among year classes collect samples over as broad a geographic where overlap occurs. These data support the range as possible and to subsequently focus assi~lnptionchat pat terns of genetic varin- with more intense sampling on rnore restricted tion .observed among populations of these areas. Single data points are insufficient species tend to reflect life history rather to define major population groups because than environmental variables : This assump- of the possibility that differences observed tion is the basis for interpreting known from a particular group of fish may be patterns of variation in sever] saImonid localized aberrations from the effects of species irldigenous to the I'ac ific Northwest genctic drift. Two collections from (sockeye, churn, pink, chinook and coho proximal populations of a species having salmon,nnd rainbow ;ind coastal ctrtthroiit similar frequencies at one or more loci trout). Each specics h:is a characteristic that differ significantly from frequencies and irnique pattern of variation arid it is of these loci in other populations are thero- thercforc inappropriate to general izc with fore regarded as the absolutc minimum for rcgard to popirlat ion structure between even positive identification of a distinct group such closely related forms as the two trout of populations. species. A major factor affecting the patterns and amounts of genetic variation is the effective populatioll size. This parameter is, in turn, affected by a number FRED M. UTTER, DONALD CAMPTON, STEWART GRANT, GEORGE MILNER, JAP4ES SEEB AND LISA WISHARC 299 of variables such as discrete or variable and determination of the genetic control year classes, single or multiple spawnings, of duplicate loci through inheritar~ce preferred spawning habitat, dcgree of studies and the examination of popula- anadromy, degree of immigration and emigra- tions. Pagcs 415-432 in Plarkert, C. L. tion, population stability, location of (ed.), Isozymes, Volume 4: Genetics population relative to conspecific popula- and evolution. Academic Press, N.Y. tions and time since divergence of two populations. The tcndency to form distinct Altukhov, Y. P. 1975. Population genetics populations in a particular drainage appears of fishes. Transl. Fish. blar. Serv. to be strongest in thc coastal cutthroat Trnnsl. Ser. Lo. 3548. and weakest in the pink salmon. Anas, K. E., and S. Murai. 1969. Usc of scale characters and a discriminant function for classifying sockeye salmon (Oncorhynchus nerka) by continent of The outstanding technical assistance of origin. Int. North Pac. Fish Conlm. Paul Aebersold and David Tee1 is gratefully Bull. 26: 157- 192. acknowledged. The value of their efforts extended well beyond the collection and Aro, K. V., and Ef. P. Shepard. 1967. analysis of data. They have been jointly Pacific salmon in Canada. Int. North responsible for procedures that produce Pac. Fish Comm. Bull. 23:225-327. reliable identification of previously un- detected polymorphic loci and an overall Aspinwall, N. 1974. Genetic analysis of increase in the efficiency of data collec- North American populat-ions of the pink tion. salmon, Oncorhynchus gorbusciza, possible evidence for the neutral mutation- random drift hypothesis. Evol. 28:295- LITERATURE CITED 305.

Allendorf, F. W. 1975. Genetic variability Atkinson, C. E., H. J. Rose, and T. 0. in a species possessing extensive gene Duncan. 1967. Pacific salmon in the duplication: Genetic interpretation of United States. Int. North Pac. Fish duplicate loci and examination of genetic Comm. Bull. 23:43-224. variation in populations of rainbow trout. Ph.D. Thesis. Univ. of Wash., Bailey, C. S., A. C. lfilson, J. E. H~lver, Seattle. and C. L. Johnson. 1970. Multiple forms of supernatant malate dehydrogenasc Allendorf, F. W. and N. Mitchell. 1977. in salmonid fishes. J. Biol. Chem. 245: Genetics of Alberni Inlet sockeye salmon 5927- 5940. populations. Report to Pacific Biologi- cal Station, Nanaimo, British Columbia. Burns, D., W. Young, and D. Hanson. '1977. Progress report for the Fish Flanagement Allendorf, F. W., and F. kt. Utter. 1973. Division, Itash. State Dept. Came, Gene duplication within the family Sept. 30. Salmonidae: Disomic inheritance of two loci reported to be tetrasomic in rain- Calaprice, J. R. 1971. X-ray spectrometric bow trout. Genetics 74: 647-654. and multivariate analysis of sockcye salmon (&~.~14~pnc::zusilt?l'ic~) from Allendorf, F. W., and F. M. Utter. 1976. di fferent gcogl-:~phic rcgitms. .I. 1:i >I). Gene duplication in the family Salmonidae. Kes. Bd. Can. 26:369-377. 111. Linkage between two duplicated loci coding for aspartate aminotrnnsfernsc Cnn~pton, D. E. 1950. Gcnet ic str~~cturc in the cutthroat trout (Salmo ctarkil. of sen-run cut throat trout (.?23n;o Hereditas 82:19-24. ctarki clar*;?:) populations in the I'ugrt Sound region. P1.S. I'hrsis. Univ. of Allendorf, F. W., and F. M. Utter. 1979. Wash., Senttlc. Population genetics. Pages 407-454 in tloar, W. S. and D. J. Randall (eds.), Crow, J. F., and !I. kimura. 1970. .-In Fish physiology. Uol. VI1. Academic introdrlct ion to populnt i on gcnct ics Press, N.Y. theory. Il:~rper, N.Y.

Allcndorf, F. W., F. PI. Utter, and B. P. Engel, W., J. Op't Ilof. and U. \Self. l!I;O. May. 1975. Gene duplication within Geduplikation durch polyploide

the family Salmonidae. 11. Detection Evolution: ' die Isocnzy-c dcr Sorbit-

. -.-- ~ ~ . . ~. . . ~ ~ . .------.. . . - .. .- .. . . ~-" ---. 300 SALMONID ECOSYSTEMS OF THE NORTH PACIFIC

dchydrogenase bei herings-und lachsarti- coastal rivers. Trans. Am. Fish. Soc. gen Fi schen (Isospondyli) . tluo~angenetik 105: 620-623. 9:157-163. 1,ewontin. R. C. 1974. The gcnetic basis Foster, B., V. Fletcher, and B. Kiser. 1977. of evolutionary change. Columbia Univ. 1376 hatcheries statistical report of Press, N.Y. production and plantings. Wash. Dep. Fish. Prog. Rep. No. 30.. MacCrimmon, H. R. 1971. World distribution of rainbow trout (SaZmo gairdneril. J. Gold, J. R. 1977. Systematics of western Fish. Res. Bd. Can. 28:663-704. North American trout (SaZmol, with notes on the redband trout of Sheepheaven May, B. 1975. Electrophoretic variation in Creek, California. Can. J. Zoo. the genus Oncorhynchus: The methodology, genetic basis, and practical applications Grant, W. S., G. B. Milner, P. ~rasriowski, to fisheries research and management. and F. M. Utter. In press. Biochemical M.S. Thesis. Univ. of Wash., Seattle. genetic variation in sockeye salmon, (Oncorhynchus nerka): tho use of bio- May, B., and F. M. Utter. 1974.. Biochemi- chemical markers for stock identifica- cal genetic variation of the genus tion in Cook Inlet, Alaska. J. Fish. Oncorhynchus in Pacific Northwest popula- Res. Bd. Canada. tions: A progress report and program summary. Interim report of contractual Harris, li., and D. A. Hopkinson. 1976. work to Wash. Dep. Fish. Handbook of enzyme electrophoresis in human genetics. American Elsevier, N.Y. May, B., F. M. Utter, and F. W. Allendorf. 1975. Biochemical genetic variation in Hodgins, H. 0. 1972. Serological and bio- pink and chum salmon: Inheritance of chemical studies in racial identification intraspecies variation and apparent of fishes. Pages 199-208 in Simon, R. C., absence of interspecies introgression and P. A. Larkin (eds.), The stock following massive hybridization of concept in Pacific salmon, H. R. Machiillan hatchery stocks. J. Hered. 66:227-232. lectures in fisheries. Univ. British Columbia. Elayr, E. 1970. Populations, species and evolution. Belknap Press, Cambridge, Hodgins, H. O., W. E. Ames, and F. M. Utter. Mass. 1969. Variants of lactate dehydrogenase isozymes in sera of sockeye salmon Mi lner, G. B. 1977. Biochemical genetic (Oncorhynch7~snerkal. J. Fish. Res. Bd. variation: Its use in mixed fishery Can. 26: 15-19. analysis. Pages 10-15 in Hassler, T. J., and R. R. VanKirk (eds.), Genetic iiodgins, H. O., anu F. hi. Utter. 1971. implications of steelhead management. Lactate dehydrogenase polymorphism of Spec. Rep. 77-1. sockeye salmon (Oncorhynchus nerka). Cons. Int. Explor. Mer., Rapports et Milner, G. B. 1978. Final report, Columbia proces verbaux; 161:lOO-101. River stock identification study. Sub- mitted to U.S. Fish and Widl. Serv. Johnson, K. R., R. F. Donelly, V. K. tiershberyer, and D. E. Bevan. 1978. Milner, G. B. 1979. Biochemical genetic lde~itificationof Kodiak Island pink variation in Columbia Hivcr steelhead salmon populations based on biochemical trout (Salmo gairdneri richardson) gcnetic variation. Univ. !Vash. Coll . populations and its use in mixed fishery Fish. Rep. FRI-UW-7801. analysis. Ph.D. Thesis. Univ. of Wash., Seattle. Johnston, J. M., and S. P. Mercer. 1976. Sea-run cutthroat in saltwater pens: Nei, PI. 1972. Genetic distance between Broodst. -k development and extended populations. Anicr. Natur. 106: 283-29-7. juvenile rearing (with a life history compendium). Wash. State Game Dep. Nei, M. 1975. Molecular population genet- Fish. Rep. Projcct AFS-57-1. ics and evolutioli. Nort-tlolland, Anist crdam. Kristiansson, A. C., and J. D. McIntyre. 1976. Genetic variation in chinook Numachi, K., Y. Pfatsumiya. and R. Sato. salmon (Oncorhynchus tshmy tschal from 1972. Duplicate genetic loci and the Columbia River and thrcc Oregon variant forrns of malate dchydrogen:~se FRED M. UTTER, DONALD CARPTON, STEWART GRANT, GEORGE FILNER, J?.!LS SEEB AND LISA WISHPPD 301

in chum salmon and rainbow trout. Bi111. Siciliano, PI. J., and C. R. Shaw. 1976. .rap. Soc. Sci. Fish. 3,9:699-706. Separation and visunlizatjon of enzymes on gels. I'dges 185-209 1:rl Smith, I. Perriard, J., A. Scholl, and H. Eppenherger. (cd.), Chromatographic and clectrophoretic 1972. Comparative studies on creatine techniques, Vol . 2, Zone electrophoresis . kinase isozymes fro111 skeletal-m~~scle Ileinemann, London. and stomach of. trout. J. Exp. Zoo. 182: 119-126. Simon, R. C. 1972. Gene frequency and the stock problcm. Pages 161-196 in Simon, Pratschner. G A. 1977. Rclat ive R. C., and P. A. Larkin (cds.), l'he resistance of six transferrin phenotypes stock concept in Pacific salmon. Irrst. of coho salmon to cytophagosis, Fish. Urriv. British Columbia, lrancouver. furunculosis and vibriosis. M.S. Thesis. Univ. of Wash., Seattle. Sneath, P. tl., and K. R. Sokal. 1973. Numerical taxonomy. Freeman, San Royce, W. F. 1962. Pink salmon fluttua- Francisco. tior~sin Alaska. Pages 15-34 in Wilimovsky, N. J. led.), H. R. MacMillan Stern, C. 1943. The Hardy-1Seinberg law lectures in fisheries, Symposium on Sci. 97:137-138. pink salmon: Suzumoto, R. K., C. B. Schreck, and J. D. Rogers, J. S. 1972. Fleasures of genetic F4cIntyre. 1977. Relative resistances similarity and genetic distance. Univ. of three transcerrin genotypes of coho Texas Publ. 7213: 1.45-153. salmon IQncorhynchus kisutch) and their gematological responses to bacterial Reinsenbichler, R. R., and J. D. McIntyre. kidney disease. J. Fish. Res. Bd. Can. 1977. Genetic differences in growth and 34: 1-8. survival of juvenile hatchery and wild s teelhead trout, Sa Zmo gairdneri. J . Thorgaard, G . H. 1977a. Chromosome Fish. Res. Bd. Can. 34:123-128. studies of steelhead. Pages 20-25 in Massler, T. J. , and R. R. VanKirk (eds .) , Reinitz, G. L. 1977. Inheritance of muscle Genetic implications of steelhead and liver types of supernatant NADP- mar~agement. C3 lif. Coop. Fish. Res. dependent isocitrate dehydrogenase in Unit. Spec. Rep. 77-1. rainbow trout (Saho guirdrceri). Bio- chem. Genetics 15:445-454. Thorgaard, G. H. 1977b. Chromosome re- arrangements and sex chromosomes in the Seeb, J., and IV. S. Grant. 1976. Biochemi- rainbow trout and sockeye salmon. Ph.D. cal genetic variation in coho, chinook, Thesis. Univ. of itash., Seattle. chum and pink salmon: The use of elcctro- phoretic markers in stock identification. 'rhorsteinson, F. V., J. H. Helle, and D. G. Wash. Dep. Fish. Final rep. service Birkhol z. 1971. Salmon survival in contract 711. intertidal zones of the Prince Viilliam Sound streams in uplifted and subsided Seeb, J. and L. N. Wishard. 1977a. Genetic area::. Pages 194-219 ir, The great characterization of prince William Sound Alaska earthquake of 1964: Biology. pink salmon populations. Report to Nat. Acad. Sci. STRN 0-309-01604-5. Alaska Dep. Fish Game. Utter, F. M. 1971. Tc-trnzol iunr oxi~I:~se Seeb, J., and L. N. Wishard. 1977b. 771e pllcnotypes of r;l i nhow tl.otrt f:itr (~!c-J use of' biocbemicc~l genetics in thc grc.irdnt.r~i)and I':I~ i f ic s:~1 mon (ilni:,rll- management of Pacif i c sa lmorr stocks : hynchu~sp. ) Camp. Uioctlem. I'lrysiol . Genetic marking and mixed fishery 398: 891-895. analysis. Wash. Dep. Fish. Final rep. service contract 792. Utter, F. hi. 1978. Gc~reticvariants of proteins in churn and cllirrook salmon Seidel, P. R. 1977. Early life history from the Bering Sea ;ind thc Yukon and study of three Washington Department of Kuskokwinl rivers: Rcport of an;lIyscs Fisheries coho salmon (Oncorhgnchus of 1976 collections. Rrp. to R.E.F.N. kisutch) populations by interracial Div., Nort'h~ccst and .\l:lska I'isll; Cc~rtcr hybridization. M.S. Thesis. Univ. of 2nd the Alaska Pep. Fish .11rd (;:lnrc. Wash., Seattle. 302 SALMONID ECOSYSTEMS OF THE NORTH PACIFIC

Utter, F. PI., and F. W. Allendorf. 1977. !~lilli:tms, C. C., R. K. Kochn, and J. B. Detol-nlination of the brced ing structure blitton. 1973. Genetic differentiation -of steelhead populations tllrough gene without isolation in the American eel, frequency analysis. Pages 10-15 in ArzguiLla rostrata. Evol . 27: 192-204. Hassler, 7'. J. and R. K. VanKirk (cd~.) . Genetic ilnplications of steelhend Wright, J. E., J. R. Heckman, and L. M. management. Calif . Coop. Fish. Res. Athertor). 1975. Genetic and develop- Unit. Spec. Rep. 77-1. mental :~nalysesof LDIl isozymes in trout. Pages -375-401 in Isozymes, Vol. 111. Utter, P. H., F. #. Allendorf, and M. 0. Academic Press, N.Y. Hodgins. 1973. Genetic variability and relationships in P;~cificsalmon and related trout based on protein varia- APPENDIX tions. Syst . Zoo. 22: 257-270. Symbols Used in Appendix Table 1 Utter, F. PI., F. W. Allcndorf, and B. May. 1970. The use of protein v:~riationin A -no variants seen the rnnnagenlent of salmonid populations. B -infrequent variants seen Pages 373-384 in Taber, R. C., and F. C -polymorphic above one percent allelic Runnel1 (eds.), Proceedings of the 41st frequency North American Wildlife Conference. .. . . -no data available E -eye Utter, F. t.f., F. W. Allendorf, and B. May. G -gut In preparation. The genetic basis of H -heart creatine kinase isozymes .in skeletal L -liver muscle of salmonid fishes. M -skeletal muscle (WH-M is presumed to be mitochondria1 locus in skeletal Utter, F. M., W. E. Ames, and 11. 0. Hodgins. muscle) 1970. Transferrin polymorphism in coho S -serum salmori IOr,corhynchus kisutch). J. Fish. Res. Bd. Can. 27:2371-2373. Abbreviated protein systems include:

Utter, F. M., and 11. 0. Hodgins. 1970. ALL, -albumin Phosphoglucomutase polporphisrn in ADH -alcohol dehydrogenase sockeye salmon. Comp. Biochem. Physiol A@ -alpha glycerophosphate dehydrogenase 36: 195-199. AAT -aspartate amino trans ferase CK -creatine kinase Utter, F. bl., and H. 0. Ilodgins. 1972. GAL -M-acetyl-B-D-Galactosaminidase Biochemical genetic variation at six GPT -glutamic-pyruvic transaminase loci in four stocks of rainbow trout. IDH -isocitrate dehydrogenase Trans. Am. Fish Soc. 101:494-502. LDH -lactate dehydrogenase MDH -malate dehydrogenase Utter, F. bl., H. 0. Hodgins, and F. W. ME -malic enzyme Allendorf. 1974. Biochemical genetic PEP -peptidase* studies of fishes: potentialities and GPGD -6 phosphogluconate dehydrogenase limitations. Pages 213-237 in Biochemi- PGI -phosphoglucose isomerase cal and biophysical perspectives in PGM -phosphoglucomutase marine biology, Vol. 1. Academic Press, PMI -phosphomannose isomerase N.Y. SDH -sorbit01 dehydrogenase * TO -tetrasolium oxidase (also commonly Utter, F. bl., 11. 0. Hodgins, F, W. Allendorf, called superoxide dismutase-SOD) A. G. Johnson, and J. L. Mighell. 1973. Tfn -transferrin Biochemical variants in Pacific salmon and rainbow trout: Their inhcritancc *different peptidase for PEP include and application in population studies. glycyl leucine (GL), lcucyl glycyiglycine Pages 329-339 in Schroder, J. ti. (ed.), (LGG) , and phenylalanylproline [PhAP) . This Genetics and mutagenesis of fish. summary cxcludcs esterase and hemoglobin Sprinter-Verlag, Berlin. variants, which we do not regularly examine because of problems involving repeatability Utter. F. M., H. 0. Hodgins, and A. G. and instability. Johnson. 1972. Biochemical of genetic differences among species and stocks of fish. Pages 98-101 in Annual Report for 1970. Int. North Pac. Fish. Comm. Appendix Table 1. Smczry of Variant Protein Systenis in Seven Salmonid Species.

Tissue Sock- -?&- Rain- Cut- Locus Distribution eye Pink Chum Coho nook bow throat Reference Sources

Alb S,E ...... C A ... C . . . Altukhov 1975, unpublished deta of G. A. E. Gall and B. Bentley.

ADH L A A A B C C C Allendorf et al. 1976, Milne 1978, Campton 1980. Seeb and Wishard 1977b, Utter and Hodgins 1972, Aspinwall 1974, May et al. 1975, Utter 1978, unpublished data of Seeb, Aebersold and Wishard.

May 1975, Allendorf and Utter 1976, May et al. 1975, Allendorf 1975, Seeb and Wishard 1977a.

CK-A1,Z Utter et al. in preparation, Perriard 1972.

-C1,2 GAL ...... C...... Aebersold and Wishard unpublished data. GPT- 2 Grant et al. in preparation, Canpton 1980. Allendorf and Utter 1973, Seeb and Wishard 1977a. Campton 1980, Flay 1975, Milner 1978, Reinitz 1977.

Utter and Hodgins 1972, Utter et al. 1973, Wright et al. 1975, Play 1975, Campton 1980, Jol~nson ct nl. 1978, Hilncr 1978, Sech and \Vish;n.d 197%.

C I:

C C Bailey et al. 1970, Uttcr and Hodgins 1973, Numachi ct 31. 1972, Aspinwall 1974, Altuhl~ov 1975, May C C 1975, Seeb and Wishard 1977~1,Compton 19SO, blilncr 1976, 'reel and Mi lncr ur~pub2isi:ctf to...... Appendi~Table 2. Smary of Variant Protein Systems in Seven Salmonid Species-continued.

Tissue Sock- Chi- Rain- Cut- Locus Distribution eye Pink Chum Coho nook bow throat Rcferencc Sources.

FIE- I hl A C ... A A A A Seeb and Wishard 1977a, Johnson et al. 1978, Campton 1980, Milner 1978, Gall and Bentley - 2 M A A ... A C C B unpublished data.

PEP- 1 (GL) M, L,E A .A A C C C C Milner 1977, Campton 1980, unpublished data of all authors of this paper. - 2 (GL) E A A A C A A A - 3 (LGG) M,L,E A A C B C ...... -4(PhAP) M,L.E ...... B...... ,

6PGD M.L,E A C C A A A C May 1975, Seeb and Wishard 1977a, 1977b; Campton 1980.

PGT- 1,2 M A C A B C B C May 1975, Allendorf 1975, Utter et al. 1976, Seeb and Wishard 1977a, Campton 1980. - 5 !.I, E ,L A A A B A C B

POI- 1 M C C A B C C C Utter and Hodgins 1970, 1972; May 1975, Seeb and Grant 1976, Kristiansson and McTntyre 1976, Campton -2 L,E C A ....B ... C C 1980, unpublished data of all authors of this paper.

P;lI FI, L,E A C C B C B C May 1975, Seeb and Grant 1976, Milner 1978, Campton 1980.

SDH- 1, 2 L A A A B C B C Engel et al. 1970.

TO- 1 L ,!1 B A A A C C C Utter 1971, May and Utter 1974, Campton 1980.

- 2 H ...b ...... C ...... Utter 1978.

Tfn S, E A A A C A C C Utter et al. 1970, Utter and Hodgins 1972, Campton 1980.