Population genetics
We estimated the pairwise of nucleotide diversity, π and the proportion of segregant sites, θw using the R package PopGenome (Pfeifer et al., 2014). Values of F ST were calculated with StAMPP 1.5.1 Weir and Cockerham’s unbiased estimator to obtain 95% confidence intervals by performing 5,000 bootstraps (Pembleton et al., 2013; Weir and Cockerham, 1984). The IBD (Isolation by Distance) was performed using the twelve locations from Chile. The significance of correlations between distance matrices in the Mantel test was evaluated using 10,000 permutations. The geographic distances were calculated using latitude/longitude of all relevant sampling points using the function of GenAlEx v 6.5 (Peakall and Smouse, 2012). Population pairwiseF ST values were used to build the genetic distance matrices for this test. We predict the number of generations since the most recent common ancestor of any pair of lineages using the mutation parameter (θw = 2N θµ) as previously described (Ruderfer et al., 2006). We used the mutation rate (µ) previously reported from laboratory estimates on S. cerevisiae (Lynch et al., 2008).