Summary of model selection for each dependent variable, showing the models with ΔAICc <2 and the subsequent model.

ΔAICc—differences in AICc relative to the lowest value of AICc of all models; AICcWi—Akaike weight of model i; Wi / Wk—ratio between the weight of model i and the weight of the null model K; G1—local vegetation; G2—proximal landscape; G3—Broad landscape; G4—Combined effect; β0—intercept; β1, β2 and β3—parameters associated with the respective variables; LV—local vegetation; PPA—Proximal landscape proportion of agricultural cover; PLC—Proximal landscape configuration; PLD—Proximal landscape diversity; BPA—Broad landscape proportion of agricultural cover; BLC—Broad landscape configuration; BLD—Broad landscape diversity.Summary of model selection for each dependent variable, showing the models with ΔAICc