Table 1 Multiple regression models between characteristics of the climate within grid cells and mean phylogenetic and specific components of butterfly and dragonfly colour lightness in Europe.

	Model	Weight	Surface	Phylogenetic component					AIC	BIC	r ²
				Thermal 1	Thermal 2	Prec 1	Prec 2	Topo 1
Butterflies	Env	None	Dorsal	0.44*	0.02	−0.13*	−0.22*	−0.07	4,186	4,225	0.34
			Ventral	0.67*	0.16*	−0.08*	−0.01	0.11*	3,615	3,653	0.53
	Env	Richness	Dorsal	0.36*	−0.05	−0.16*	−0.20*	−0.04	3,234	3,272	0.41
			Ventral	0.58*	0.09*	−0.12*	−0.03	0.10*	2,843	2,881	0.56
	Env (spatial)	None	Dorsal	0.51*	0.07	−0.03	−0.05	−0.08*	4,083	4,127	0.38
			Ventral	0.71*	0.07	−0.10*	0.02	0.10*	3,580	3,624	0.54
Dragonflies	Env	None	Dorsal	0.73*	0.46*	0.06	0.07*	0.14*	2,986	3,025	0.71
	Env	Richness	Dorsal	0.64*	0.39*	0.06*	0.05*	0.12*	2,627	2,666	0.64
	Env (spatial)	None	Dorsal	0.57*	0.24*	−0.11*	−0.06*	0.11*	2,382	2,426	0.79

	Model	Weight	Surface	Specific component					AIC	BIC	r ²
				Thermal 1	Thermal 2	Prec 1	Prec 2	Topo 1
Butterflies	Env	None	Dorsal	0.20*	0.15*	−0.21*	−0.15*	0.25*	4,423	4,462	0.24
			Ventral	0.60*	0.12*	−0.01	−0.02	0.15*	3,983	4,021	0.41
	Env	Richness	Dorsal	0.31*	0.09*	−0.18*	−0.10*	0.21*	3,473	3,511	0.35
			Ventral	0.57*	0.06*	−0.05	−0.02	0.17*	3,017	3,056	0.52
	Env (spatial)	None	Dorsal	0.22*	0.23*	−0.10	−0.05	0.19*	4,316	4,360	0.29
			Ventral	0.58*	0.00	−0.07	−0.02	0.12*	3,950	3,994	0.43
Dragonflies	Env	None	Dorsal	0.55*	0.30*	−0.14*	0.01	0.21*	3,917	3,956	0.51
	Env	Richness	Dorsal	0.58*	0.25*	−0.09*	0.02	0.16*	3,526	3,565	0.49
	Env (spatial)	None	Dorsal	0.38*	0.16*	−0.25*	−0.10*	0.16*	3,778	3,822	0.55

AIC, Akaike information criterion; BIC, the Bayesian information criterion; Topo 1, first principal component of variables related to the terrain topography.
For each model, the standardized coefficients of the predictors (slope), the Akaike information criterion 573 (AIC), the Bayesian information criterion (BIC), and the coefficient of determination (adjusted R574 squared) are shown. Results from weighted least-squares regressions with species richness in each grid 575 cell as the weighting factor and models to account for spatial autocorrelation are also provided (spatial). 576 Thermal 1 and Thermal 2: first two principal components of temperature- and insolation-related variables 577 (Supplementary Table 5). Prec 1 and Prec 2: first two principal components of precipitation-related 578 variables. Topo 1: first principal component of variables related to the terrain topography. Positive 579 coefficients indicate an increase in the components (warmer, higher temperature seasonality, wetter, 580 higher precipitation seasonality, higher elevation), negative values a decrease (butterflies: n=1,825; 581 dragonflies: n=1,845). *=significant at p < 0.001 (t-test).

Search