Fig. 4

Bayesian cluster analysis. a Determination of the optimal value of K by the Evanno method from Structure Harvester. b Mean likelihood [L(K) + −SD] over 20 runs assuming K clusters (K = 1–10). c Bar plot representations of Bayesian STRUCTURE analysis of S.scrofa samples with K = 4. d Output from CLUMPAK, visualizing major modes for K = 4 from the individual-based clustering performed in STRUCTURE. Each vertical line represents one individual and the colour shows the proportion of each individual assigned to each of the four genetic clusters