AbstractThe strict homing of sockeye salmon Oncorhynchus nerka results in reproductively isolated populations that often spawn in close proximity and share rearing habitat. High spawning fidelity enables these populations to adapt to local conditions, resulting in a wide range of life history characteristics and genetic variation within individual watersheds. The Wood River system in southwestern Alaska provides a pristine, well‐studied system in which to examine fine‐scale population structure and its influences on juvenile life histories. Adult sockeye salmon spawn in lake beaches, rivers, and small tributaries throughout this watershed, and juveniles rear in five nursery lakes. We genotyped 30 spawning populations and 6,066 migrating smolts at 45 single nucleotide polymorphism loci, two of which are candidates for positive selection in the study system. We show that there is significant genetic structure (FST = 0.032) in the Wood River lakes and that divergence is generally related to spawning rather than nursery habitat (hierarchical analysis of molecular variance; P < 0.05). Four groups of populations were identified based on genetic structure and used to determine the composition of unknown mixtures of migrating smolts using a Bayesian modeling framework. We demonstrate that smolt migration timing is related to genetic structure; stream and river populations dominate catches in early June, while beach spawners and the populations in Lake Kulik are more prevalent from mid‐June to early September. Age‐2 smolts are primarily produced by the Lake Kulik and beach spawning populations, showing that genetic differences may reflect divergent freshwater and migration life history strategies. These results indicate that local adaptation to spawning habitat influences genetic divergence in the Wood River lakes, affecting both adult and juvenile life stages of sockeye salmon.
Read full abstract