Skip to main content
. 2017 Aug 11;7(18):7462–7474. doi: 10.1002/ece3.3292

Table 1.

Minimum‐adequate mixed‐effects model for the effects of pathogen damage, tree species richness, plant traits, and plot characteristics on herbivory (R 2m = 14.3%; R 2c = 48.5%)

Predictor Std. Est. SE df t p
(Intercept) 2.05 0.07 38 30.3 <.001
Site B 0.56 0.04 275 13.6 <.001
Elevation (log) 0.06 0.02 220 2.9 .004
Slope (log) 0.04 0.01 231 2.6 .009
Pathogen damage (log) 0.07 0.01 10,100 10.8 <.001
Tree species richness (log) 0.04 0.01 110 2.8 .006
Initial wood volume (log) 0.05 0.01 9,993 6.9 <.001
PC1morph 0.11 0.06 30 1.9 .070
PC2morph −0.11 0.05 30 −2.1 .044
PC2chem 0.04 0.06 29 0.6 .550
Niche breadth (log) 0.05 0.07 32 0.7 .472
Niche marginality (log) 0.10 0.05 30 2.0 .053
Pathogen damage:PC2chem −0.03 0.01 10,120 −3.8 <.001
Pathogen damage:niche breadth −0.02 0.01 10,090 −3.5 <.001
Tree richness:wood volume 0.02 0.01 6,311 2.7 .008
Pathogen damage:tree richness −0.01 0.01 6,725 −2.1 .039

Nonsignificant terms retained in the minimal models are italicized; log‐transformed predictors are indicated by (log); colons indicate interactions between two predictors. PC1 and PC2 = scores of the first and second principal component of a PCA on morphological or chemical leaf traits (see Tables S1, S2).