Table 4.

Summary of the traits-impervious surfaces model.

Predictor	Coefficient (± SE)	z-value	p value (adj)
Proportion of impervious surfaces	− 0.20 (0.05)	− 3.89	< 0.0001
Traits, no interaction
Nesting: above versus below	0.02 (0.18)	0.09	0.92
Lectism: specialist versus generalist	− 0.54 (0.17)	− 3.16	0.002
Sociality: solitary versus social	− 0.96 (0.23)	− 4.08	< 0.0001
Size: large versus small	− 0.17 (0.16)	− 1.02	0.31
Traits, interacting with proportion of impervious surfaces
Nesting: above versus below	0.31 (0.03)	11.4	< 0.0001
Lectism: specialist versus generalist	− 0.09 (0.03)	− 3.03	0.002
Sociality: solitary versus social	− 0.08 (0.03)	− 2.68	0.007
Size: large versus small	− 0.02 (0.02)	− 0.69	0.49

Predictor		Chisq	p value
Biogeographical zone		56.877	< 0.0001

p values in bold are significant (α = 5%). The significance of the categorical variable “Biogeographical zone” was evaluated using a Likelihood Ratio Test. This model is a mixed-effect model, the random variables included are the sites ID, nested in the sampling categories, as well as the species names. Coefficients are given with their standard error. Traits in bold are the reference modalities. p value(adj) means that p-values were adjusted with the False Discovery Rate adjustment method.