Bayesian methods for inference in populations of hybridizing fish

Eric C. Anderson

To be presented in session "Issues in Fisheries I"

A pervasive concern in the conservation of salmonids is hybridization between different populations and species. For example, small, wild populations might encounter substantial gene flow from fish straying from hatcheries. Also, the use of genetic data in helping to delineate "evolutionarily significant units" may be complicated by naturally-occurring hybridization between closely related species. However, ongoing hybridization between genetically dissimilar populations leads to predictable changes in the frequency of multilocus genotypes, and so may be investigated using genetic marker data. I will review recent advances in Bayesian methods for analyzing genetic data from populations containing hybrids. I illustrate these versatile methods by using them to estimate allele frequencies and to identify coastal cutthroat trout, steelhead trout, and their hybrids in samples of juvenile fish taken from Oregon and Washington.