Results of model selection examining the effect of roadside vegetation cutting and environmental variables on moose browsing.

Four generalized linear mixed-effects models included in model selection to determine which environmental explanatory variables influenced the proportion of moose browsed plants along roadsides. The variables plot number nested within site id were included as random effects in all models.
a Models are ranked with Akaike Information Criterion, corrected for small sample size (AICc)
b Key: k, number of parameters; LL, log-likelihood; Marginal R2, Nakagawa and Schielzeth’s Marginal R2 where the fixed factors alone explain the proportion of variance; Conditional R2, Nakagawa and Schielzeth’s Conditional R2 where both the fixed and random factors explain the proportion of variance; ΔAICc, the difference in the AICc; ωAICc, model weights.
Results of model selection examining the effect of roadside vegetation cutting and environmental variables on moose browsing.