Table 1.
Summary of linear mixed models for height- and biomass-based relative growth rates, root, stem, and leaf mass ratios, root-shoot ratio, %survival, and %herbivory over the entire study period (19 months). The design included two fixed effects: forest type (two levels: seasonal or aseasonal) and species distribution (two levels: widespread or aseasonal); and two random factors: species (16 levels) and plot (20 levels, 10 plots per site). Herbivory and root–shoot ratio data were log10-transformed and seedling survival data were square root-arcsin-transformed to meet model assumptions. Because of significant forest × distribution interaction terms in all analyses, data were re-coded creating a single, combined variable (ForDist) with four levels. Significant P-values are bolded
| Trait | DFnum | DFden | F-value | P-value |
|---|---|---|---|---|
| Height growth | 3 | 1048 | 4.01 | 0.0075 |
| Biomass growth | 3 | 1071 | 5.93 | <0.0001 |
| Root mass ratio | 3 | 1071 | 5.18 | 0.0015 |
| Leaf mass ratio | 3 | 1071 | 8.82 | <0.0001 |
| Stem mass ratio | 3 | 1071 | 1.69 | 0.167 |
| Root–shoot ratio | 3 | 1071 | 3.77 | 0.0104 |
| Survivorship | 3 | 1060 | 6.32 | <0.0001 |
| Herbivory | 3 | 1060 | 5.84 | <0.0001 |