• Identification and application of molecular markers to chum salmon population genetics

      Garvin, Michael R.; Gharrett, A. J.; Tallmon, David; Smoker, William (2008-08)
      I developed a new technique, DEco-TILLING (an adaptation of Eco-TILLING), to discover useful single nucleotide polymorphisms (SNPs) rapidly and inexpensively. Some chum salmon (Oncorhynchus keta) populations have declined in Western Alaska and in other areas of their geographic range. Possible reasons include climate shifts, by- catch in fisheries, and other perturbations. Genetics offers powerful tools that can be used to monitor this species on the high seas in stock mixtures aiding management of by- catch identification and other contributors to declines. Single nucleotide polymorphisms are a genetic marker that can be easily and rapidly surveyed on many individuals. Tools like SNPs offer advantages in discriminating stocks in mixtures. However, tens or hundreds of informative SNPs must be discovered from among the millions in the chum salmon genome. Available discovery methods introduce ascertainment bias into this process, which can result in uninformative SNPs or the failure to identify useful ones. I incorporated and improved a genotyping assay to screen SNPs in thousands of individuals for a tenth of the cost of the standard available assay, and improved an assay to resolve the phase of linked SNPs. I show that the SNPs that I discovered are informative and can be use for mixed stock analysis.
    • Population Genetics And Mixed Stock Analysis Of Chum Salmon (Oncorhynchus Keta) With Molecular Genetics

      Garvin, Michael R.; Gharrett, A. J.; Kruse, Gordon; Pella, Jerome; Tallman, David (2012)
      Chum salmon (Oncorhynchus keta) are important for subsistence and commercial harvest in Alaska. Variability of returns to western Alaskan drainages that caused economic hardship for stakeholders has led to speculation about reasons, which may include both anthropogenic and environmental causes in the marine environment. Mixed stock analysis (MSA) compares genetic information from an individual caught at sea to a reference baseline of genotypes to assign it to its population of origin. Application of genetic baselines requires several complex steps that can introduce bias. The bias may reduce the accuracy of MSA and result in overly-optimistic evaluations of baselines. Moreover, some applications that minimize bias cannot use informative haploid mitochondrial variation. Costs of baseline development are species-specific and difficult to predict. Finally, because populations of western Alaskan chum salmon demonstrate weak genetic divergence, samples from mixtures cannot be accurately assigned to a population of origin. The chapters of this thesis address three challenges. The first chapter describes technical aspects of genetic marker development. The second chapter describes a method to evaluate the precision and accuracy of a genetic baseline that accepts any type of data and reduces bias that may have been introduced during baseline development. This chapter also includes a method that places a cost on baseline development by predicting the number of markers needed to achieve a given accuracy. The final chapter explores the reasons for the weak genetic structure of western Alaskan chum salmon populations. The results of those analyses and both geological and archaeological data suggest that recent environmental and geological processes may be involved. The methods and analyses in this thesis can be applied to any species and may be particularly useful for other western Alaskan species.