TABLE A2.
Results of averaged generalised additive models from Model Groups 3 and 4, each model modelling the prevalence of a single prey type
| Model group | Variable | Prey type | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Amphibian | Avian | Bullhead | Crustacean | Cyprinid | Eel | Insecta | Mammal | Marine | Salmonid | Stickleback | ||
| 3 | Age class | 1 | 0.46 | 0 | 1 | 0.37 | 0.44 | 0.58 | 1 | 0.1 | 0.75 | 0.09 |
| 3 | Sex | 0.57 | 0.19 | 1 | 0 | 1 | 0.13 | 0.42 | 1 | 0.11 | 0 | 0.38 |
| 3 | Body condition | 0.85 (*) | 0.71 | 0 | 1 | 0.65 | 1 (*) | 0.68 | 1 (*) | 0.29 | 1 (*) | 0 |
| 3 | Body length | 1 (*) | 0.14 | 1 (*) | 1 | 0.39 | 0.5 | 0.46 | 0.43 (.) | 0.11 | 1 (*) | 0.89 (.) |
| 3 | Sex: Age class | 0.26 | 0.06 | 0 | 0 | 0 | 0.13 | 0.35 | 1 | 0 | 0 | 0 |
| 3 | Sex: Body condition | 0 | 0.06 | 0 | 0 | 0.05 | 0 | 0.15 | 0 | 0 | 0 | 0 |
| 3 | Sex: Body length | 0.45 | 0 | 0.45 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 3 | Age class: Body length | 1 | 0 | 0 | 0 | 0 | 0.29 | 0.35 | 0.34 | 0 | 0.53 | 0 |
| 3 | Age class: Body condition | 0.18 | 0.06 | 0 | 1 | 0.09 | 0 | 0 | 0.34 | 0 | 0.21 | 0 |
| 3 | Body condition: Body length | 0 | 0 | 0 | 0 | 0.08 | 0 | 0 | 0 | 0 | 0 | 0 |
| 3 | Month | 1 | 0 | 1 | 1 | 1 | 1 | 0.58 | 0 | 0 | 0.8 | 0 |
| 3 | Year | 0.5 (.) | 0.11 | 0.5 (*) | 0.5 | 0.5 | 0.5 (**) | 0.32 | 0 | 0.25 | 0.6 (.) | 1 |
| 3 | Region | 1 | 0 | 0 | 1 | 1 | 0 | 0 | 1 | 0 | 1 | 0.15 |
| 3 | Distance from coast | 1 (+,**) | 1 (−) | 1 (***) | 1 | 0.89 | 1 (***) | 0.33 | 0 | 1 (*) | 0 | 0.23 |
| 4 | Reproductive status | 0 | NA | 0 | 0 | 0 | 0 | 0 | 0 | NA | 0 | 0 |
Relative variable importance (RI) is shown for each variable, for each modelled prey type, significance (p) is indicated as: *<.05, **<.01, ***<.001. Model Group 4 (females only) revealed that reproductive status was not an important predictor of the prevalence of prey types (although note that for some prey types [marked NA] models were not run due to small sample sizes when males were removed). The remainder of the variables for Model Group 4 are therefore not shown (e.g., age class, sex, etc) as these were all incorporated in model group 3.