Table 1.

Averaged models describing effects of covariates on worker size variation, queen–worker dimorphism (where (a) and (b) represent models excluding and including worker size variation, respectively) and colony size in the Attini. Regression coefficients and CIs are reported from best (ΔAICc < 7) PGLS models from full candidate sets (see the electronic supplementary material, table S4a, b and c). Bold type indicates significant covariates. β = model averaged regression slope (95% CIs), pCI = 95% CI for the regression slope from 1000 models including parameters with W > 0.4 from a sample of 1000 equally likely trees; W = cumulative AICc weight over all models from the full candidate model set.

covariates	worker size variation	queen–worker dimorphism (a)	queen–worker dimorphism (b)	colony size
(intercept)	β = 1.806 (−0.554, 4.165), pCI ± 0.000	β = 1.659 (1.644, 1.675), pCI ± 0.031	β = 0.923 (−0.957, 2.802), pCI ± 0.010	β = 4.230 (23.954, 12.414), pCI ± 0.146
colony size	β = 0.392 (0.227, 0.559), pCI ± 0.000, W = 1.00	β = 0.159 (0.042, 0.276), pCI ± 0.001, W = 0.85	β = 0.135 (−0.030, 0.301), pCI ± 0.005, W = 0.60	—
worker size variation	—	—	β = 0.235 (−0.127, 0.598), pCI ± 0.011, W = 0.52	—
queen–worker dimorphism	β = 0.108 (−0.497, 0.713), pCI± 0.000, W = 0.48	—	—	—
mean diurnal temperature range	β = −0.015 (−0.035, 0.0058), pCI ± 0.000, W = 0.69	—	—	β = −0.021 (−0.071, 0.029), pCI ± 0.001, W = 0.40
isothermality	—	β = −0.004 (−0.044, 0.036), W = 0.28	β = −0.003 (−0.046, 0.038), W = 0.33	β = 0.022 (−0.126, 0.170), pCI ± 0.001, W = 0.50
temperature seasonality	—	—	—	—
precipitation seasonality	β = 0.013 (−0.017, 0.043), pCI ± 0.000, W = 0.041	β = 0.0009 (−0.014, 0.016), W = 0.28	β = 0.004 (−0.009, 0.016), W = 0.36	β = −0.015 (−0.210, 0.179), pCI ± 0.001, W = 0.62
isothermality × precipitation seasonality	—	—	—	β = 0.003 (−0.0002, 0.006), W = 0.13
latitude	—	—	—	β = 0.009 (−0.077, 0.096), W = 0.19