Abstract
Landscape genomics is a rapidly growing field with recent advances in both genotyping efficiency and statistical analyses that provide insight towards local adaptation of populations under varying environmental and selective pressure. Chinook salmon ( Oncorhynchus tshawytscha ) are a broadly distributed Pacific salmon species, occupying a diversity of habitats throughout the northeastern Pacific with pronounced variation in environmental and climate features but little is understood regarding local adaptation in this species. We used a multivariate method, redundancy analysis ( RDA ), to identify polygenic correlations between 19 703 SNP loci and a suite of environmental variables in 46 collections of Chinook salmon (1956 total individuals) distributed throughout much of its North American range. Models in RDA were conducted on both rangewide and regional scales by hierarchical partitioning of the populations into three distinct genetic lineages. Our results indicate that between 5.8 and 21.8% of genomic variation can be accounted for by environmental features, and 566 putatively adaptive loci were identified as targets of environmental adaptation. The most influential drivers of adaptive divergence included precipitation in the driest quarter of the year (Rangewide and North Coastal Lineage, anova P = 0.002 and 0.01, respectively), precipitation in the wettest quarter of the year (Interior Columbia River Stream‐Type Lineage, anova P = 0.03), variation in mean diurnal range in temperature (South Coastal Lineage, anova P = 0.005), and migration distance (Rangewide, anova P = 0.001). Our results indicate that environmental features are strong drivers of adaptive genomic divergence in this species, and provide a foundation to investigate how Chinook salmon might respond to global environmental change.