Table 1.

Summary of statistical tests using mixed-effects models to determine H:D allometric variation for log-transformed tree height (H, m) and mean centered diameter (D, cm) with random intercepts and slopes, plot as a random factor, and shade tolerance (not included in any of the best-fit models), angiosperm/gymnosperm (Clade), evergreen/deciduous (Phenology), and bioclimatic variables: altitude (m), mean temperature (°C × 10), mean precipitation (mm), precipitation seasonality, mean diurnal temperature range (Bio2), and isothermality (Bio3). For brevity, the best-fit model is reported for each functional or floristic group. AIC = Akaike's information criterion; n = number of tree individuals in each subsetted dataset

Model	Marginal R2	Conditional R2	AIC	Model structure	Fixed effects	Coefficient
All	0.55	0.75	−260,190	H ∽ D + Phenology + Altitude + SeasonPrecip + MeanTemp	Phenology	−0.06
n = 2,770,803					Altitude	−0.0001
					Season Precip	−0.0032
					Mean Temp	−0.0010
Gymnosperms	0.49	0.79	−348,174	H ∽ D + Altitude + Mean Precip	Altitude	−0.0001
n = 1,270,306					Mean Precip	0.0001
Angiosperm	0.60	0.77	−114,826	H ∽ D + Phenology + Bio3 + Bio2	Phenology	−0.22
n = 1,500,497					Bio3	−0.01
					Bio2	−0.0013
Evergreen	0.51	0.79	−320,509	H ∽ D + Clade + Altitude + Mean Precip	Clade	0.29
n = 1,292,397					Altitude	−0.0001
					Mean Precip	0.0001
Deciduous n = 1,478,406	0.60	0.76	−138,801	H ∽ D + Bio3	Bio3	−0.01