Association to drivers
Within-population genetic diversity (nucleotide diversity π and heterozygosity He) was not influenced by lake area (π: Spearman’s rho = 0.03, p = 0.95, He: rho = -0.06,p = 0.88), connection (π: rho = 0.43, p = 0.25, He: rho = 0.53, p = 0.15), salinity (π: rho = 0.37, p = 0.33, He: rho = 0.24, p = 0.53) (Supplemental Table 6, Supplemental Figure 7). However, there appeared to be a trend towards higher nucleotide diversity with lower water temperatures (rho = -0.61, p = 0.08), but not for heterozygosity (rho = -0.20, p = 0.60).
Mantel tests indicated no correlation between the geographic and genetic distance matrices over all filter options (Figure 4A, Supplemental Table 7, r = 0.007, p = 0.504). Finding no correlation refutes the isolation-by-distance hypothesis and indicates other factors might explain the distribution of S. diversicolor genetic diversity. However, the genetic distance matrix also did not correlate with matrices of environmental distance (r = 0.002, p = 0.503, Figure 4B) or connection distance (r = 0.041, p = 0.441, Figure 4C).