Table 3. The effect of random terms on species presence.
| Models | ||||||
|---|---|---|---|---|---|---|
| Summary statistics | Random terms | Full | -Year slope | -Treatment slope | -Species intercept | -Plot/area intercept |
| Standard deviation | Plot/site intercept | 0.29 | 0.29 | 0.29 | 0.22 | |
| Site intercept | 0.18 | 0.18 | 0.19 | 0.13 | ||
| Species intercept | 2.15 | 2.15 | 2.19 | 2.11 | ||
| Year slope | 0.00 | 0.00 | 1.78 | 0.00 | ||
| Treatment slope | 0.52 | 0.52 | 1.98 | 0.51 | ||
| AIC | 3,693 | 3,691 | 3,704 | 4,311 | 3,727 | |
| χ2 | 0.00 | 12.23 | 619.94 | 37.93 | ||
| p-value | 1.000 | 0.000 | 0.000 | 0.000 | ||
Note:
The response variable is the number of species within a sample. The full model contains fixed terms for treatment, year and treatment x year, random intercepts for plot within each site and species, and random slopes for treatment and year within each species. The table shows the effects of removing each random term (intercept or slope) from the full model. It lists the standard deviations attributed to each remaining random term, as well as the summary statistics of that model. The AIC values are the Akaike Information Criteria values for each model, the χ2 values are the test statistics obtained by comparing the given model to the full model with likelihood ratio tests, while the p-values indicate the level of significance of the test statistics. The degree of freedom is one for all χ2 except for the ‘-plot/site’ model where it is two. Significant results are highlighted in bold.