Differences in genetic architecture between continents at a major locus previously associated with sea age at sexual maturity in European Atlantic salmon.

Abstract Wild Atlantic salmon populations show heritable variation in the proportion of fish that mature early, as grilse, after one winter at sea. Grilse are undesirable in aquaculture, and in hatchery-supported recreational fisheries, but their frequency could be reduced using marker-assisted selection. SNP markers near vgll3 on Ssa 25 are associated with 30–40% of the variation in early maturity in European (EU) Atlantic salmon. Our goal was to determine whether this association occurred in a major North America (NA) aquaculture strain from the Saint John River (SJR). Grilse reared entirely in freshwater tanks were heavier and longer two years post-hatching than normally-maturing post-smolts of the same sex. We first genotyped a large sample from a single year-class (YC-80) on a custom 80 SNP assay that included the four SNPs near vgll3. When no associations were found, our study was extended to include a genome-wide analysis (GWA) of another year-class using a custom 50 K SNP chip (YC-50 K). The GWA found no associations on Ssa 25 but did detect a SNP with experiment-wide significance on its homeolog, Ssa 21. The new SNP explained 6% of the variance in grilsing. Its association was not a by-product of the greater weights and lengths of grilse. GWA showed that the top SNP for weight was on Ssa 01/23. The early-maturing allele at vgll3 had a higher frequency of homozygotes in European populations than in the North American populations. We conclude that differences in the genetic architecture of traits between highly diverged populations, such as these from different continents, can require re-estimation of marker-trait relationships. Highlights • Early post-smolt maturity is undesirable in salmon aquaculture but could be reduced through marker-assisted selection. • SNP markers on Ssa25 that are highly associated with maturity in Europe were not associated with maturity in North America. • Instead a SNP on Ssa21 was highly associated with maturity but was not associated with the covariate weight. • Genetic architecture differences between highly-diverged populations may require re-estimation of marker-trait associations. [ABSTRACT FROM AUTHOR]