Fig. 2: Odds ratios of non-native species occurrence in local assemblages in response to LU-type and LU-intensity.

a a logistic generalised linear mixed effects model with LU-type, LU-intensity and their interaction as predictors (n = 11,693); b–g the effect of LU-type modelled across all taxa (b) (n = 11,713) and for each taxon separately (c–g) (n = 407, 3978, 1114, 762 and 4453; respectively). All p-values are based on two-sided tests. Odds ratios from contrasts to the reference levels (‘Primary under Minimal use’ (a) or ‘Primary vegetation’ (b–g)) are back-transformed from the logarithmic scale of the models, respectively. Shown are mean odds ratios (dots) and standard errors (bars). Asterisks indicate significant differences from the reference levels (p values: *<0.05; **<0.01 and ***<0.001). The NA in a and f indicates intense use in urban areas and spiders in plantation, which were removed from the analysis since all of the assemblages had non-native species (leading to total separation in the models, see “Methods” section for details). The NA in e, f and g indicates mammals, spiders and vascular plants in urban areas, and mammals in cropland, which could not be included in the analysis due to low sample size. Silhouette illustrations for the taxa are from PhyloPic (http://phylopic.org), contributed by various authors under public domain license.