Coefficient estimates from final models for each species.

Plots are ordered by mean effect size (calculated as the coefficient divided by the standard error of that coefficient) across all models, with highest effect size on the left and lowest on the right. Species are shown in increasing order of prevalence (top to bottom). Colours are to distinguish between plots only, and do not have any inherent meaning. Only variables included in the final model are shown (hence missing values in panel f), while interactions are not shown. Axes vary in scale between subplots.