Chromosomal rearrangements (e.g., inversions, fusions, and translocations) have long been associated with environmental variation in wild populations. New genomic tools provide the opportunity to examine the role of these structural variants in shaping adaptive differences within and among wild populations of non-model organisms. In Atlantic Salmon (Salmo salar), variations in chromosomal rearrangements exist across the species natural range, yet the role and importance of these structural variants in maintaining adaptive differences among wild populations remains poorly understood. We genotyped Atlantic Salmon (n = 1429) from 26 populations within a highly genetically structured region of southern Newfoundland, Canada with a 220K SNP array. Multivariate analysis, across two independent years, consistently identified variation in a structural variant (translocation between chromosomes Ssa01 and Ssa23), previously associated with evidence of trans-Atlantic secondary contact, as the dominant factor influencing population structure in the region. Redundancy analysis suggested that variation in the Ssa01/Ssa23 chromosomal translocation is strongly correlated with temperature. Our analyses suggest environmentally mediated selection acting on standing genetic variation in genomic architecture introduced through secondary contact may underpin fine-scale local adaptation in Placentia Bay, Newfoundland, Canada, a large and deep embayment, highlighting the importance of chromosomal structural variation as a driver of contemporary adaptive divergence.

Juvenile Atlantic Salmon, young-of-the-year (YOY) and parr (ages 1 to 2+), were collected by electrofishing during the period July to September of 2017 and 2018 from 26 rivers around Placentia Bay, Newfoundland, Canada. Individuals were genotyped by Centre for Integrative Genetics (CIGENE, Ås, Norway) using a 220K bi-allelic single nucleotide polymorphism (SNP) Affymetrix Axiom array developed for Atlantic Salmon as described in Barson et al. (2015). SNPs were filtered for a minor allele frequency (MAF) cut-off of 0.01, a missingness threshold of 0.05 across samples, and non-biallelic loci within each year sampled. SNPs were retained only if they passed filtering in both years resulting in a total of 139,038 SNPs.