Creation and utilization of novel genetic methods for studying and improving management of Chinook salmon population

As a major component of fisheries in the northern Pacific Ocean, Chinook salmon (Oncorhynchus tshawytscha) are of significant management concern. Their anadromous life history, in which adult fish migrate to their natal streams, leads to populations (stocks) that are genetically distinguishable and, ideally, would be managed independently. Many of these stocks, particularly at the southern end of the species' range, have experienced serious declines, which has motivated widespread hatchery production and supplementation. The physical coded-wire tagging (CWT) program currently used to track hatchery fish, and ultimately to supply information for cohort-based fishery harvest models, is increasingly ineffective and can no longer sustain the data demands of fishery managers and scientists. Also, current genetic tools utilizing microsatellite markers do not scale well to the enormous number of fish that need to be analysed, have error rates that are too high for individual- and pedigree-based methods, and genotype inconsistently across laboratories, creating an impediment to interjurisdictional collaboration. However, the next generation of genetic markers, single nucleotide polymorphisms (SNPs), do have low enough error rates and are amenable to the high-throughput genotyping required for ocean fishery stock identification and large-scale tagging of hatchery fish via pedigree reconstruction. Here we describe the successful identification of 117 novel SNP loci using genomic data from a sister salmonid taxon and demonstrate their substantial power for discriminating five major stocks of salmon from the three largest basins on the Pacific coast of North America. We then assemble a panel of 96 SNP loci and genotype over 8000 individuals from 69 distinct populations for construction of a baseline for genetic stock identification (GSI) and show that it has, effectively, near-maximum power for discriminating most Chinook salmon stocks captured in mixed-stock fisheries off the coasts of California and Oregon. This baseline is used to confidently assign over 2000 ocean-caught Chinook to their source population and demonstrate over 99% concordance between the GSI assignments and identifications from CWTs recovered from these fish. The same panel of SNPs is also used to implement a large parentage-based tagging (PBT) experiment at one of the most productive hatcheries in the Central Valley of California. PBT involves genotyping reproducing adults and using their genotypes as intergenerational genetic tags that are recovered through parentage inference with their progeny. By genotyping over 12,000 individuals from six complete brood years, we show that the large number of resulting pedigrees effectively provide the same age and stock information as traditional CWTs, but also can be used to inform hatchery breeding practices, estimate the heritability of physical traits and eventually can serve as the basis for detailed linkage maps and associated mapping of quantitative trait loci. The genetic resources developed here are a substantial improvement over current methods and are fundamentally changing the way salmon populations are studied, monitored and managed.