Fig. 4: Estimates of continuous long-distance gene flow and isolation by distance (IBD) across the breeding range of the Atlantic puffin estimated from 71 individuals across 12 colonies.
From: Complex population structure of the Atlantic puffin revealed by whole genome analyses
a Effective migration surfaces inferred by the program EEMS using the average distance between pairs of individuals calculated in ANGSD by sampling the consensus base for each individual at 1,093,765 polymorphic nuclear sites. Darker reds indicate reduced migration across those areas, while darker blues highlight higher migration rates than the global mean. Different colonies are indicated using colors consistent with those in Fig. 1. b Correlation between genetic (Slatkin’s linearized FST) and geographic (Least Cost Path—only over water) distance presented after removing the Spitsbergen, Bjørnøya, Isle of May, and Canadian individuals. The diagonal line visualizes the result of the multiple regression on distance matrices (MRM) analysis (slope and y-intercept). The Mantel test between genetic and geographic distance (R = 0.775, P = 0.012, nColonies = 7) was significant and 60.08% of the variance in Slatkin’s linearized FST was explained by geographic distance (regression coefficient of linear IBD model = 0.76 × 10−6, P = 0.006, nColonies = 7). A two-dimensional kernel density estimation (kde2d) highlights dense groups of data points, thus substructure in the genomic landscape pattern. Analyses were conducted and results visualized in R using the ecodist, marmap and MASS packages. The dataset(s) needed to create this figure can be found at https://doi.org/10.6084/m9.figshare.14743323.
