TABLE 1.
Bayesian linear regression models and coefficients
| Model | β 1 (SD) | β 2 (SD) | μ (SD) | DIC |
|---|---|---|---|---|
| GenDist ~ Environment + Geography | 0.173 (0.013) | 0.297 (0.0126) | −0.0028 (0.0457) | 601.242 |
| GenDist ~ Barrier + Geography | 0.364 (0.0367) | 0.301 (0.0132) | −9.63e−05 (0.0419) | 679.119 |
| GenDist ~ Barrier + Environment | 0.465 (0.0408) | 0.207 (0.0151) | 0.000258 (0.0538) | 921.633 |
| GenDist ~ Geography | 0.327 (0.0135) | −0.00132 (0.0494) | 759.724 | |
| GenDist ~ Environment | 0.229 (0.0156) | 0.00428 (0.061) | 1,026.883 | |
| GenDist ~ Barrier | 0.533 (0.0435) | 0.000125 (0.0491) | 1,105.838 |
Predictor variables were standardized using a z‐score prior to modeling. Genetic distance (GenDist) was calculated as Nei's D A. Environmental distance is a multivariate distance matrix of degree days less than zero and growing season precipitation. Geography is a pairwise geographic distance matrix. The smallest DIC indicates the best model.