Fig 1.
Sampling locations of corkwing wrasse analysed in this study (for details of sample abbreviations see Table 1).
The location and geographical extension of the two major sandy areas in Norway, Jæren and Lista, are indicated in the inset.
Table 1.
Sample information and summary statistics for genetic variability in corkwing wrasse samples.
n = sample size, A = average number of alleles per locus ± standard deviation, Ar = average allelic richness per locus ± standard deviation. HS = gene diversity. FIS = deviation from Hardy-Weinberg genotype proportions. Numbers in bold indicate locus deviating significantly from Hardy-Weinberg expectation at 5% level after the False Discovery Rate approach [42].
Table 2.
Pairwise comparisons among corkwing wrasse sample localities within and across the three genetic breaks.
Values below the diagonal are FST estimates for all sample pairs and above the diagonal are the corresponding P-values. Numbers in bold indicate statistical significant tests at 5% level after the False Discovery Rate approach [47].
Table 3.
Pairwise FST estimates within and among geographical regions.
Fig 2.
Maps of posterior probabilities of corkwing wrasse individual membership defined by GENELAND under a) uncorrelated (K = 2) and b) correlated (K = 4) alleles model. Plots representing assignment of pixels to each cluster (highest probability in light yellow and zero probability in red) and map of estimated posterior probability of population membership are presented for each model.
Fig 3.
Unrooted neighbour joining tree based on Nei et al.’s [54] Da distances estimated for sixteen corkwing wrasse populations at nine microsatellite loci.
Numbers of the nodes indicate the bootstrap support, based on 100 000 replicates.
Fig 4.
Results of multiple matrix regression with randomization (MMRR) analysis.
Plots show a) the relationship of environmental and geographical distance, b) isolation-by-distance (IBD), c) isolation-by-environment (IBE), and d) multiple regression for the effects of geographical (ßDistance = 0.05) and environmental (ßEnvironment = 0.95) distances on genetic distance.
Table 4.
Results of the partial Mantel tests and MMRR analysis comparing the contribution of geographical (IBD) and environmental distance (IBE) to genetic differentiation for the whole Scandinavia, and separately for southern (Skagerrak) and West coast samples.
r = coefficient of correlation, P = P-values (significant values in bold).