Identifying patterns of dispersal, connectivity and selection in the sea scallop, Placopecten magellanicus, using RADseq-derived SNPs

Abstract

Understanding patterns of dispersal and connectivity among marine populations can directly inform fisheries conservation and management. Advances in high-throughput sequencing offer new opportunities for estimating marine connectivity. We used restriction-site-associated DNA sequencing to examine dispersal and realized connectivity in the sea scallop Placopecten magellanicus, an economically important marine bivalve. Based on 245 individuals sampled rangewide at 12 locations from Newfoundland to the Mid-Atlantic Bight, we identified and genotyped 7163 single nucleotide polymorphisms; 112 (1.6%) were identified as outliers potentially under directional selection. Bayesian clustering revealed a discontinuity between northern and southern samples, and latitudinal clines in allele frequencies were observed in 42.9% of the outlier loci and in 24.6% of neutral loci. Dispersal estimates derived using these clines and estimates of linkage disequilibrium imply limited dispersal; 373.1 ± 407.0 km (mean ± SD) for outlier loci and 641.0 ± 544.6 km (mean ± SD) for neutral loci. Our analysis suggests restricted dispersal compared to the species range (>2000 km) and that dispersal and effective connectivity differ. These observations support the hypothesis that limited effective dispersal structures scallop populations along eastern North America. These findings can help refine the appropriate scale of management and conservation in this commercially valuable species.

Keywords: RADseq; connectivity; dispersal; outlier loci; population genomics; population structure; sea scallop; single nucleotide polymorphism.

Figure 1

Map of 12 sea scallop (P. magellanicus) collection locations from the Northwest Atlantic. Site MDA (Mid‐Atlantic Bight) represents the middle of several nearby collection locations grouped as one population

Figure 2

Results from (a) the Bayesian test for selection completed using the program BayeScan and (b) the hierarchical island model test for selection completed using the program Arlequin for 7163 loci sequenced in 12 populations of P. magellanicus. BayeScan outliers are defined as all loci with a q‐value higher than .05 (highlighted in red). Arlequin outliers are defined as the loci that fall above the simulated 5% quantile of F _ST versus Heterozygosity (q ≤ .05, highlighted in blue)

Figure 3

Map of the proportion of each of the 12 P. magellanicus populations assigned to two population groups (blue and red) identified in the program STRUCTURE using outlier loci and the ΔK method to select the optimal number of genetic clusters in the data

Figure 4

Plots of individual admixture for 12 populations of P. magellanicus at K = 2 determined using the program STRUCTURE and the ΔK method to select the optimal number of genetic clusters in the data. Results are presented for all populations at (a) all loci, (b) neutral loci and (c) outlier loci as well as four north populations at (d) all loci, (e) neutral loci and (f) outlier loci

Figure 5

Principal components analysis plots for (a) all loci, (b) neutral loci and (c) outlier loci in 12 populations of P. magellanicus

Figure 6

Neighbour‐joining trees for Cavalli–Sforza and Edwards chord distance (D _c) between 12 populations of P. magellanicus for (a) neutral loci and (b) outlier loci. North populations are highlighted in grey, south populations in white and bootstrap values greater than 50% are shown

Figure 7

Isolation by distance plot of F _ST/1−F _ST versus population pairwise distance for 12 populations of P. magellanicus using (a) approximate current‐based distance (all loci: p < .05, outlier loci: p < .01) and (b) shortest ocean‐based distance (all loci: p < .05, outlier loci: p < .001) for all loci (red squares) and outlier loci (blue triangles)

Figure 8

(a) Heat map of population‐specific standardized allele frequencies for 48 clinal outlier loci in 12 populations of P. magellanicus. (b) Plot of clines in allele frequency in 12 populations of P. magellanicus as a function of the distance in kilometres from the furthest north population (SUN) for clinal neutral loci (n = 123, 24.6% of tested loci, grey) and clinal outlier loci (n = 48, 42.9% of tested loci, black)