Table 3.
Best‐fit wolverine‐habitat models across each model set. Comparing across the best‐fit, or minimum adequate, models (1, 13, 17, 22) suggests that a combination of the natural and anthropogenic variables included in the cumulative effects model (24) best‐explain patterns of wolverine frequency
| Model no., set | Variables | Res. deviance | Res. df | AIC | ΔAIC | AIC weight | −2LL |
|---|---|---|---|---|---|---|---|
| 1, Landcover | DENSECON + MIXED + OPENCON + SHRUB + HERB + REGEN + SNOW.ICE | 65.01 | 83 | 213.78 | 6.04 | 0.05 | 201.78 |
| 13, Human Disturbance | LININD | 85.18 | 89 | 221.96 | 14.22 | 0.00 | 219.96 |
| 17, Abiotic | SP.SNOW | 94.392 | 89 | 231.17 | 23.43 | 0.00 | 229.17 |
| 22, Biotic | FOX + COYOTE | 107.53 | 88 | 246.31 | 38.57 | 0.00 | 242.31 |
| 24, Cumulative Effects | DENSECON + MIXED + SHRUB + HERB + REGEN + SNOW.ICE + BLOCKURB + LININD + LINRECL + SP.SNOW + COYOTE + FOX | 48.631 | 78 | 207.74 | 0 | 0.95 | 183.74 |
Models highlighted in bold represent the best‐fit out of each model set.