<b>A Pleistocene biogeography in miniature: the small-scale evolutionary history of</b><b> </b><b><i>Podarcis</i></b><b> </b><b><i>lusitanicus</i></b><b> </b><b>(Squamata, Lacertidae)</b>

A heatmap with the F_ST values was built using the ggplot2 and reshape R packages (Wickham, 2007, 2011). To test for isolation-by-distance (IBD), we conducted a spatial Analysis of Principal Components (sPCA; Jombart, Devillard, Dufour, &amp; Pontier, 2008) under the adegenet R package (Jombart, 2008). The sPCA yields scores summarizing both genetic variability and the spatial structure among individuals or sites using Moran’s eigenvector maps (MEMs). First, the sPCA spatial variation is modelled through a connection network using a K-nearest neighbors approach. A test is provided for global structure, defined as entities being more genetically similar to their immediate neighbors than expected in a random spatial distribution, with the null hypothesis that allele frequencies are distributed at random on the connection network. The test relied on a comparison of the observed test statistic, which summarizes correlations of alleles to vectors in the matrix of global MEMs, to the distribution of test statistics obtained through a Monte Carlo randomization procedure using 10⁴ permutations (Jombart et al., 2008).