Table 2.
Summary of model selection results for each bear management area for analysis of Alberta grizzly bear inventory data (2004–8).
| BMA | Detection (g0)/scale of movement (σ) | Density surface predictor (AICc weight) | ||||
|---|---|---|---|---|---|---|
| Females | Males | Females | Males | |||
| g0 | σ | g0 | σ | |||
| 2 | CC (−) | TRI (−) | TRI (+) | RSF*Risk (0.57) RSF (0.32) |
RSF*Risk (0.48) RSF (0.35) RSF + Risk(0.17) |
|
| 3 | TRI (+) | TRI (+) Trend (−) |
RSF (0.75) | RSF (0.98) | ||
| 4 | TRI (~ +)A | Trend (+) | TRI (+)A | RSF (0.86) | RSF (0.70) | |
| 5 | TrendA (~ +) | heterogeneity | Risk (0.65) | Risk (0.44) RSF + Risk (0.41) |
||
| 6 | TRI (+)A | TRI (+)A | Risk (0.81) | Risk (0.48) Constant (0.37) |
||
AThese covariates were tied for support with AICc scores that were greater than the constant model by less than 2 AICc units.
SECR site covariates include CC (Canopy closure), TRI (terrain ruggedness) and Trend (linear trend over sampling sessions). Heterogeneity refers to a mixture model with 2 classes. Supported density surface models (∆AICc < 2) along with AICc weights are also shown for each BMA and sex of bear. Detailed model selection results, as well as SECR parameters for each model, are given in Appendix S1.