Isolation By Distance
As an initial test of IBD, we implemented various classical statistics based on Fst values (Slatkin 1993; Rousset 1997; Hutchison and Templeton 1999; Diniz-Filho et al. 2013). We calculated pairwise Fst using the approach of Weir and Cockerham (1984) for individuals, populations, and lineages. For individuals, we used a Mantel test with 999 permutations to evaluate significance of the overall relationship between geographic and genetic distance. For populations, we examined between- and within-lineage slopes to estimate the relative prominence of local versus regional structure. Finally, we calculated overall fixation between lineages as a basic estimator for potentially species-level divergence (Roux et al. 2016). For individual and population-based analyses, we used linearizedFst =Fst /1-Fst.
As an additional test of both IBD and isolation by environment (IBE), we used generalized dissimilarity modeling (GDM) in the R package ‘gdm’ (Ferrier et al. 2007; Mokany et al. 2022). We used the pairwiseFst matrix by site as our measure of genetic dissimilarity, and the 9 continuous ecological variables, elevation, and geographic distance (lat/long; see SI for list) as the environmental predictors. We fit a GDM to these data and assessed variable inclusion using estimated coefficients greater than zero, and variable importance and significance using 1,000 permutations. We also evaluated model significance and percent deviance explained.