Outlier Detection
For S. umbilicalis , 143 loci (1.2%) were identified by at least three methods, with 52 and 14 of these loci identified by four and five methods, respectively (Table 3). Observed heterozygosity was lower than expected and FIS was positive for all populations (Table S6). FST values ranged from 0-0.368 at outlier loci compared to 0-0.026 at neutral loci, and 80.3% and 81.8% of these pairwise comparisons were significant (p=0.001) in the neutral and outlier datasets, respectively (Fig. 2A; Table S7). For N. lapillus , 24 loci (0.6%) were identified by at least three methods, with only one locus identified by four methods (Table 3). Observed heterozygosity was higher in N. lapillus and values were similar to expected heterozygosity across all sites (Table S8). FIS was lower in N. lapillus in comparison toS. umbilicalis (Table S8). FST values ranged from 0.015-0.637 in outlier loci compared to 0.033-0.362 in neutral loci, and all pairwise comparisons were significant (p=0.001), with the exception of the north Wales sites (Llanddulas & Great Orme West, p=0.211) in the neutral marker dataset (Fig. 3A, Table S9). All downstream analyses of adaptive population structure were carried out on these outlier datasets. Our resulting neutral marker datasets contained 11,457 and 3,820 loci for S. umbilicalis and N. lapillus , respectively.