Fig. 9: Environmental variables driving species community changes (zeta diversity changes).

a I-splines mean values of predictors from 30-rounds calculation with Sørensen-equivalent metric for seven environmental variables with their respective contributions to zeta diversity. Environmental contributions are shown for six different zeta orders (i.e. ζ_i) here referring to plots compared together (from ζ2 to ζ50), with the main contributively environmental variables highlighted for each zeta order. We observed a shift in the key forest features explaining community changes according to geographic scale, with canopy openness and altitude as main environmental drivers at local scale (ζ2), replaced by tree-related microhabitats diversity and deadwood amount at large scales (from ζ28 to ζ50). b Proportion (in r²) of the total zeta-diversity variance explained by environmental variables, distance, undistinguishable variables or unexplained, according to zeta order. The eight tested environmental variables accounted for 15–20% or the total variance in community changes regardless of zeta order, with distance explaining up to 20% of the total variance as well depending on the different zeta orders considered.