Fig. 4

Summary of the results of discriminant analysis of principal components (DAPC). Panel a represents the Bayesian Information Criterion (BIC) for each value of K. The minimum value of BIC before the first increase or stabilization indicates the best-supported number of genetic clusters (K = 4 in this case). Panel b represents an ordination plot for the first two discriminant axes. Each dot represents one individual and colors and inertia ellipses indicate their assignment to one of the four genetic clusters inferred by DAPC. The up-right graph inset displays the variance explained by the principal component axes used for DAPC (in dark grey). The bottom-right inset displays in relative magnitude the variance explained by the two discriminant axes plotted (in dark grey). Panel c represents whether the individuals (rows) were correctly assigned (based on discriminant functions) to the genetic cluster where they were included a priori (columns) by K-means analyses used to infer the best-supported clustering solution. Colors represent membership probabilities to each genetic cluster (red = 1, orange = 0.75, yellow = 0.25, white = 0) and blue crosses indicate the cluster where the individuals were originally assigned by K-means analyses. Generally, the DAPC classification of individuals is consistent with their assignment to the clusters originally identified by K-means analyses (i.e. blue crosses are on red rectangles). Panel d shows the number of individuals from each population (rows) assigned to each of the four inferred genetic clusters (columns). The size of black squares is proportional to the number of individuals assigned to the different clusters (up-right legend). Population codes are described in Table 1