Table 3.

Tests of the slope, intercept and shift along a common slope for bivariate relationships among the three growth forms (L for liana, SS for scandent shrub and T for trees; Table 1) using standardized major axis regression analyses as implemented within SMATR software.

y ~ x	Slope				Intercept				Shifts along the common slope (P)
	L	SS	T	P	L	SS	T	P
log Kt ~ log WD	−4.17			0.191	0.34a	0.23a	−0.44b	<0.001	<0.001
log Kt ~ log VD	−0.79			0.222	3.26a	2.77ab	2.67b	0.004	<0.001
log Kt ~ log Dh	2.09			0.337	−2.17	−2.37	−2.32	0.358	<0.001
log SLA ~ log LT	−0.88ab	−0.67a	−1.55b	0.048
log C ~ log SLA	−0.13			0.472	2.95	2.95	2.95	0.937	0.019
log δ13C ~ log SLA	0.09			0.84	1.27b	1.27ab	1.28a	0.039	0.235
log K ~ log SLA	1.02			0.573	−1.07	−1.08	−1.04	0.824	0.035
log SLA ~ log WD	−2.26			0.253	1.36	1.52	1.41	0.145	0.015
log C ~ log Kt	−0.08			0.339	2.80a	2.78a	2.74b	0.002	<0.001
log SLA ~ log Kt	0.62			0.427	1.06b	1.29b	1.59a	0.003	<0.001
log K ~ log Kt	0.63			0.191	−0.07b	0.24b	0.58a	<0.001	<0.001

Significant differences (P ≤ 0.05) are indicated in bold. Different lowercase letters identify growth forms that significantly differed in post hoc tests. See Table 2 for trait abbreviations. The unit of δ¹³C was converted to positive when calculating log δ¹³C ~ log SLA relationship.