Stock identification of Fraser River sockeye salmon using microsatellites and major histocompatibility complex variation

The utility of DNA-based variation for stock identification was evaluated for Fraser River sockeye salmon Oncorhynchus nerka. For this evaluation, the variation at 14 microsatellite loci and one major histocompatibility complex (MHC) locus was determined from approximately 13,000 fish from 47 populations in the drainage. Genetic differentiation among the populations was observed, the overall F-ST value for the 14 microsatellite loci surveyed being 0.054 and that for the MHC locus being 0.215. The variation among regions and populations within regions was approximately 20 times as great as that of the annual variation within populations for the microsatellite loci and 28 times as great for the MHC locus. The power of a microsatellite locus for population-specific identification in simulated Mixture samples was positively correlated with the number of observed alleles at the locus. Analysis of simulated Mixtures indicated that the mean percentage error of estimated stock compositions wits less than 1% per population, with a standard deviation of approximately 3% for a mixture sample size of 150 fish. Estimated stock compositions of a known sample of radio-tagged sockeye salmon indicated that the mean percentage error was 1% per population and 1% per run-time group. The use of DNA variation to estimate stock compositions in the management of the 2002 Fraser River sockeye salmon fisheries indicated the early arrival of the late run in the Fraser River. Stock identification based on DNA-level variation will probably become the preferred method in Pacific salmon applications in the near future.