Table 1.

Proportion of genetic variance explained by geographic or climatic variation

	DA			Fst/(1-Fst)
Model	Geography	Climate	Shared	Geography	Climate	Shared
geog + [clim1,clim2]	0.4001	−0.0278	0.0660	0.1557	0.0078	0.0969
geog + clim1	0.4838*	0.0102	−0.0176	0.2332	0.0578	0.0194
geog + clim2	0.3392*	−0.0841	0.1270	0.1010	−0.1316	0.1516
geog + [temp,seas,prec]	0.8526***	0.4035*	−0.3864	0.2404	0.1287	0.0122
geog + prec	0.4832*	0.0095	−0.0170	0.2321	0.0569	0.0205
geog + seas	0.4865*	0.054	−0.0204	0.3068	0.0545	−0.0545
geog + temp	0.4917*	−0.0295	−0.0256	0.2047	−0.0928	0.0479
geog + [prec,seas]	0.3636	−0.047	0.1026	0.1007	−0.0349	0.1518
geog + [prec,temp]	0.5093	−0.0238	−0.0432	0.1870	−0.0294	0.0655
geog + [seas,temp]	0.6909***	0.2009	−0.2248	0.3160	0.0240	−0.0635

Geographic distance (geog) is the Euclidean distance among sampling sites and climate is the proportion of variance explained by the different climatic variables included in each model. Clim1 and clim2 are the first two principal components of the entire 19 variable BioClim dataset, explaining 97 % of the variance in total. Temp, seas, and prec are the first principal components for temperature, seasonality, and precipitation BioClim variables, respectively. The proportion of variance shared by both geography and climate is reported in the columns labeled ‘shared’. High ‘shared’ values would indicate collinearity between ecological and geographic parameters; negative values are an artifact of subtracting adjusted R² values to derive the shared proportion of variance [83]

*p < 0.5; **p < 0.01; ***p < 0.001