| Summary: | Investigating the role of natural selection in driving adaptive diversification has become a central theme in evolutionary ecology as advancements in genome typing technologies provide new approaches for identifying the genetic basis of phenotypic diversity in non-model organisms. I used population-based genome scans to investigate the recent divergence of shore- and stream-spawning ecotypes in Kokanee salmon (Oncorhynchus nerka). These ecotypes co-exist in many post-glacial lakes throughout their range, and exhibit distinct reproductive behaviors, spawning habitat preference, and life history traits. Several algorithms were used to test for statistical outliers across five replicate ecotype pairs of Kokanee salmon. Among 50 expressed sequence tag (EST)-linked and anonymous microsatellite loci, signatures of directional selection were observed at 15 loci, including four loci that exhibited outlier behaviour across two or more lakes. The inconsistency of parallel patterns suggests that either several different genes or genetic pathways underlie ecotype divergence or outlier-detection methods are prone to Type II error when selection is weak. Nonetheless, population structure and differentiation at outlier loci is distinct from that of neutral loci, which infers that outliers may be under selection. Annotations of EST-linked outliers suggest that energy metabolism and pathogen resistance may be involved in initiating and maintaining barriers to gene flow between these two reproductive ecotypes. Within a Kokanee fisheries management context, markers associated with adaptive genetic variation would be very useful since neutral microsatellite markers cannot distinguish these recently diverged ecotypes (<10,000 years). Currently, the absolute abundance of shore-spawning Kokanee cannot be determined using conventional methods and ecotypes cannot be determined for angled fish to estimate ecotype-specific harvest rates. Here, I assess the accuracy and power of outlier loci in distinguishing shore- and stream-spawning Kokanee using mixed stock analyses and individual assignment tests. In general, outlier loci outperform neutral loci and simulations suggest that management-relevant levels of accuracy (>90%) may be achieved with sufficient baseline sampling and ecotype differentiation. Thus, genome scans can be useful in identifying informative markers for recently diverged stocks. === Irving K. Barber School of Arts and Sciences (Okanagan) === Biology, Department of (Okanagan) === Graduate
|
|---|
