Table 1 Results of the multimodel inference across two spatial scales for cytb.

Scale	Independent variables	AICc	R²	R² adj.	ΔAICc	wAICc	Moran’s I	P value (Moran’s I)
Grid cells	Phylogenetic diversity	−853.15	0.385	0.382	0	0.403	−0.008	0.438
	Phylogenetic diversity + temperature variability	−852.63	0.39	0.384	0.516	0.311	−0.006	0.425
	Phylogenetic diversity + precipitation variability	−851.5	0.387	0.38	1.652	0.176	−0.01	0.458
	Phylogenetic diversity + temperature variability + precipitation variability	−850.52	0.39	0.38	2.628	0.108	−0.013	0.48
Zoogeographic regions	Phylogenetic diversity + precipitation trend + precipitation variability	−270.707	0.443	0.387	0	0.535	0.0561	0.24
	Phylogenetic diversity +precipitation trend	−269.916	0.381	0.341	0.789	0.361	0.06	0.228
	Precipitation trend + precipitation variability	−267.428	0.334	0.291	3.278	0.103	0.125	0.126

All possible combinations of variables identified as significant from the hierarchical partitioning approach (see Supplementary Table 5) were evaluated using linear models. Models that contained highly collinear variables (Pearson’s |r| > 0.7) were excluded. The ranking of the models is based on the corrected Akaike’s information criterion (AICc). We retained only models with ΔAICc ≤ 5 compared to the best model (ΔAIC = 0) for each spatial scale. Residual spatial autocorrelation was computed for each retained explanatory model across spatial scales using Moran’s I and 10,000 simulations (P value).

Quick links

Search