Table 1.
Model 1: The effects of habitat and treatment on the proportion of artificial nests dug up, showing the fixed effects that were included in each version of the model
| Model | (Intercept) | Habitat | Nest depth | Habitat:Nest depth | df | logLik | AICc | Delta | Weight | Marginal R2 | Conditional R2 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Model 1.2 | −3.37599 | + | + | NA | 6 | −93.290 | 200.331 | 0.000 | 0.734 | 0.164 | 0.439 |
| Model 1.1 | −4.27503 | + | + | + | 8 | −91.613 | 202.356 | 2.026 | 0.266 | 0.222 | 0.484 |
| Model 1.3 | −2.12057 | + | NA | NA | 4 | −105.964 | 220.727 | 20.397 | 0.000 | 0.089 | 0.363 |
| Model 1 | −1.45160 | NA | NA | NA | 3 | −108.776 | 224.023 | 23.692 | 0.000 | 0.000 | 0.364 |
| Model 1.4 | −2.20247 | NA | + | NA | 4 | −119.715 | 248.230 | 47.899 | 0.000 | 0.068 | 0.276 |
Model formula: glmer(cbind(success, failure) ~ fixed effects + (1|Sett/Transect ID), family = Binomial) [NB: Model 1.4 random term is (1|Sett)].
Models were selected using AICc model selection, and the models with a delta AICc < 3 were kept (Table 2). The model formula is shown, with “success” representing the number of artificial nests that were dug up, and “failure” representing the number left in the ground. All models had the same random terms, except for Model 1.4, which only included “sett” to allow model convergence.