Genome wide association analyses
To identify the potential associations of genomic regions with dispersal probability in the Helgeland house sparrow metapopulation, a GWA analysis on dispersal was carried out using the R package RepeatABEL (Rönnegård et al., 2016). In the GWA model, sex was included as fixed factor to control its effect on dispersal whereas hatch year and natal island were used as random factors to account for these environmental effects, and finally, the genomic relatedness matrix (GRM) was also added as a random effect to account for the genetic population structure and relatedness between individuals. The GWA analysis was carried out with a logistic model where dispersal was a binary response variable (disperser = 1; resident = 0). To reveal the SNPs that potentially may be in linkage disequilibrium (LD) with genetic variants affecting dispersal propensity, Bonferroni correction was applied with a family-wise error rate (FWER) of 0.05. The genes flanking any SNP markers associated with dispersal were determined using the annotated house sparrow genome (Elgvin et al., 2017), and any annotation of these genes were investigated.