Analysis of microsatellite DNA resolves genetic structure and diversity of chinook salmon (Oncorhynchus tshawytscha) in California’s Central Valley Michael A. Banks, Vanessa K. Rashbrook, Marco J. Calavetta, Cheryl A. Dean, and Dennis Hedgecock Abstract: We use 10 microsatellite DNA markers to assess genetic diversity within and among the four runs (winter, spring, fall, and late fall) of chinook salmon (Oncorhynchus tshawytscha) in California’s Central Valley. Forty-one pop- ulation samples are studied, comprising naturally spawning and hatchery stocks collected from 1991 through 1997. Maximum likelihood methods are used to correct for kinship in juvenile samples and run admixture in adult samples. Through simulation, we determine the relationship between sample size and number of alleles observed at polymorphic microsatellite markers. Most samples have random-mating equilibrium proportions of single and multilocus genotypes. Temporal and spatial genetic heterogeneity is minimal among samples within subpopulations. An F ST of 0.082 among subpopulations, however, indicates substantial divergence among runs. Thus, with the exception of our discovery of two distinct lineages of spring run, genetic structure accords with the diverse chinook life histories seen in the Central Val- ley and provides a means for discrimination of protected populations. Résumé : Nous nous sommes servis de dix marqueurs microsatellites de l’ADN pour estimer la diversité génétique dans et entre les quatre remontes (hiver, printemps, automne et fin de l’automne) de quinnat (Oncorhynchus tshawyts- cha) dans la vallée centrale de la Californie. Nous avons étudié 41 échantillons de populations, représentant des pois- sons à reproduction naturelle et des stocks d’écloserie, prélevés entre 1991 et 1997. Les méthodes du maximum de vraisemblance ont servi à faire la correction pour la parenté génétique dans les échantillons de juvéniles et pour le mé- lange des remontes dans les échantillons d’adultes. Par la simulation, nous déterminons la relation entre la taille de l’échantillon et le nombre d’allèles observés aux marqueurs polymorphes microsatellites. La plupart des échantillons présentent des proportions équilibrées d’appariement aléatoire de génotypes à un et à plusieurs locus. L’hétérogénéité génétique temporelle et spatiale est minime entre les échantillons au sein des sous-populations. Un F ST de 0,082 entre les sous-populations indique toutefois une divergence substantielle entre les remontes. Ainsi, à l’exception de notre dé- couverte de deux lignées distinctes dans la remonte du printemps, la structure génétique concorde avec les divers cy- cles biologiques du quinnat observés dans la vallée centrale et fournit un moyen de distinguer les populations protégées. [Traduit par la Rédaction] Banks et al. 927 Introduction Many salmon populations throughout the Pacific North- west have been extirpated and, of those remaining, the ma- jority is at risk of extinction (Nehlsen 1994). At the southern limit of its range in North America, the Pacific salmon Oncorhynchus is particularly vulnerable, owing primarily to dry climate and human competition for water but also to in- tensive exploitation and habitat disturbance. In California, three salmonid stocks are already protected, and listing of all remaining stocks has been proposed (National Marine Fish- eries Service (NMFS) 1998, 1999). Salmon conservation concerns are particularly acute in California’s Central Val- ley, habitat for four spawning runs of the chinook salmon (Oncorhynchus tshawytscha) and the source of water for two thirds of the state’s inhabitants and its enormous agricultural industry. The stock at greatest risk, the Sacramento River winter-run chinook salmon, was listed as threatened by the state of California in 1989, when run size fell below 200, and as endangered by the federal government in 1994 (NMFS 1994). Population genetic data are increasingly important in managing salmonid populations on the brink of extinction (Waples 1995). There is a need to identify protected stocks in mixed ocean harvests and in rivers or estuaries where dams or water diversions imperil out-migrating juveniles. Identification of broodstock in propagation programs is Can. J. Fish. Aquat. Sci. 57: 915–927 (2000) © 2000 NRC Canada 915 Received March 8, 1999. Accepted February 17, 2000. J15051 M.A. Banks, 1 V.K. Rashbrook, M.J. Calavetta, 2 C.A. Dean, and D. Hedgecock. Bodega Marine Laboratory, University of California at Davis, P.O. Box 247, Bodega Bay, CA 94923, U.S.A. 1 Author to whom all correspondence should be addressed. e-mail: mabanks@ucdavis.edu 2 Current address: Hitachi Genetic Systems, 1201 Harbor Bay Parkway, Suite 150, Alameda, CA 94502, U.S.A.