Table 3.

Phylogenetically-controlled regression analyses between t_m and other variables

	Long lived speciesa		Herbaceousb		Woodyc
Variable	R ‡	ΔR2 †	R	ΔR2 †	R	ΔR2 †
Stature	0.162*	--	0.152NS		0.029NS
Growth form d	0.239***	0.057	NC		NC
Perenniality e	NC		0.423***		NC
Pollination mode f	0.082NS		0.027NS		0.020NS

^a N = 230 ^b N = 76 ^c N = 185

0.01<P < 0.05 and ^NS for P > 0.05.

^d Herbaceous species were coded 0 and woody species 1; ^e Short-lived species (annuals and biannuals) were coded 0, and long-lived species (perennials species) were coded 1; ^f The mode of pollination was coded 0 for abiotic (wind- or water-dispersed) and 1 for biotic (animal-dispersed)

^‡ correlation coefficient of simple regression analyses

^† multiple regression analyses (stepwise option using all significant variables from the simple regression analyses): ΔR² represents the gain of R² (proportion of variation explained by the variables) when the variable is incorporated into the model, "--" indicates that the variable does not enter in the model, after the integration of the most explicative variables.