Table 2. Explanatory models for the residual connectance according to path analyses in Fig. 3 with direct and indirect coefficients and their relative contributions to the explained variation (R2), following Lewinsohn [49].
| Residual connectance | Explanatory variables | Correlation (r) | Effect | ||
|---|---|---|---|---|---|
| Direct (d) | Indirect (i) | Total (e = d + i) | |||
| All herbivores | Land use intensity | -0.19 | -0.36 | 0.17 | -0.19 |
| Proportion of exotic plants | 0.01 | 0.35 | -0.06 | 0.29 | |
| AvTD of herbivores | 0.22 | 0.66 | - | 0.66 | |
| Plant taxonomic span | -0.14 | -0.18 | 0.04 | -0.14 | |
| Herbivore taxonomic span | 0.09 | -0.61 | 0.62 | 0.01 | |
| Endophages | Land use intensity | -0.21 | -0.50 | 0.30 | -0.20 |
| Proportion of exotic plants | 0.13 | 0.78 | -0.10 | 0.68 | |
| AvTD of herbivores | 0.29 | 0.79 | - | 0.79 | |
| Plant taxonomic span | -0.30 | -0.48 | 0.24 | -0.24 | |
| Herbivore taxonomic span | 0.26 | -0.72 | 0.80 | 0.08 | |
| Exophages | AvTD of herbivores | 0.22 | 0.57 | - | 0.57 |
| Herbivore taxonomic span | 0.06 | -0.45 | 0.48 | 0.03 | |
Only variables with significant total effects are presented. The “direct effect” (d) expresses how much a given variable changes in response to changes in another variable while controlling for the effect of all other variables in the model. The ‘‘indirect effect’’ (i) expresses the influence of a given variable on another variable that is mediated by one or more variables through causal relationships presented in the model.