SNP‐array reveals genome‐wide patterns of geographical and potential adaptive divergence across the natural range of Atlantic salmon (Salmo salar)

Abstract

Atlantic salmon (Salmo salar) is one of the most extensively studied fish species in the world due to its significance in aquaculture, fisheries and ongoing conservation efforts to protect declining populations. Yet, limited genomic resources have hampered our understanding of genetic architecture in the species and the genetic basis of adaptation to the wide range of natural and artificial environments it occupies. In this study, we describe the development of a medium-density Atlantic salmon single nucleotide polymorphism (SNP) array based on expressed sequence tags (ESTs) and genomic sequencing. The array was used in the most extensive assessment of population genetic structure performed to date in this species. A total of 6176 informative SNPs were successfully genotyped in 38 anadromous and freshwater wild populations distributed across the species natural range. Principal component analysis clearly differentiated European and North American populations, and within Europe, three major regional genetic groups were identified for the first time in a single analysis. We assessed the potential for the array to disentangle neutral and putative adaptive divergence of SNP allele frequencies across populations and among regional groups. In Europe, secondary contact zones were identified between major clusters where endogenous and exogenous barriers could be associated, rendering the interpretation of environmental influence on potentially adaptive divergence equivocal. A small number of markers highly divergent in allele frequencies (outliers) were observed between (multiple) freshwater and anadromous populations, between northern and southern latitudes, and when comparing Baltic populations to all others. We also discuss the potential future applications of the SNP array for conservation, management and aquaculture.

A genome scan for selection signatures comparing farmed Atlantic salmon with two wild populations: Testing colocalization among outlier markers, candidate genes, and quantitative trait loci for production traits

“…Search for outlier loci in North American populations without first removing fish with putative European ancestry could have led to the detection of a large number of false‐positive outlier loci in the analysis because of large differences in average allele frequencies at the loci on the same 6K SNP chip (Bourret, Kent, et al., 2013) that was used here. We believe that the well‐documented use of fish with European ancestry in salmon breeding populations in Maine, just south of the Bay of Fundy region where SGRP was developed (Glebe, 1998) justifies the use of our K  = 2 (Fig.…”

“…Recent outlier locus studies of salmonid fishes have primarily compared wild populations at different spatial scales (Freamo, O'Reilly, Berg, Lien, & Boulding, 2011; Vasemägi & Primmer, 2005) or have correlated outlier loci of wild populations with specific environmental variables (Bourret, Kent, et al., 2013; Bourret, Dionne, Kent, Lien, & Bernatchez, 2013; Narum, Campbell, Kozfkay, & Meyer, 2010; Perrier, Bourret, Kent, & Bernatchez et al., 2013). Of more relevance to Atlantic salmon domestication are four outlier studies comparing aquaculture strains with wild populations from the same region as their putative founder population(s) (Gutierrez, Yáñez, & Davidson, 2016; Karlsson, Moen, Lien, Glover, & Hindar, 2011; Mäkinen, Vasemägi, McGinnity, Cross, & Primmer, 2015; Vasemägi, Nilsson, & Primmer, 2005).…”

Estimating contemporary effective population size in non-model species using linkage disequilibrium across thousands of loci

“…To investigate how the above methods can be applied to real populations, genetic data from 28 populations of Atlantic salmon (Bourret et al, 2013) were used to estimate N e . From the published data set, we excluded populations with mean sample sizes o30 and samples that combined individuals from multiple years, as these issues can also bias estimates of N e .…”

“…We restricted our analysis to loci present on the linkage map of Lien et al (2011) and the linkage groups of this map were treated as chromosomes. Any loci identified by Bourret et al (2013) as putatively under selection, or with minor allele frequency o0.05, were excluded. We placed the r 2 values for each pair of loci into sets of 'unlinked' (loci on different chromosomes), and 'linked' (loci colocated on a chromosome) following Larson et al (2014).…”

“…Using these gross measures of genome architecture and our simulated data, we develop approximate bias corrections to account for the effects of linkage. We illustrate our results by calculating biased and putatively unbiased estimates of N e from a published data set that includes 28 populations of Atlantic salmon (Salmo Salar) sampled across their range in Europe and North America (Bourret et al, 2013) and an Atlantic salmon linkage map. We also show that, although high-throughput sequencing techniques will increase precision of N e estimates, the lack of independence of the vast numbers of pairwise comparisons means that precision inN e does not increase nearly as fast as naive expectation.…”