Tree species diversity impacts average radial growth of beech and oak trees in Belgium, not their long-term growth trend

Forest Ecosystems

Table 3 Parameter estimates and model evaluation of the ln (BAI) date models for beech and oak

Beech (n = 3095)					oak (n = 3978)		Df	p > \|t\|
Fixed effects	Estimate	SE	Df	p > \|t\|	Estimate	SE	Df	p > \|t\|
Intercept	3.4389	0.1030	2941	< 0.0001	8.0459	0.0620	3904	< 0.0001
poly (Dp)¹	11.0569	2.7688	2941	0.0001	12.2897	2.9592	3904	< 0.0001
poly (Dp)²	−16.1505	1.5582	2941	< 0.0001	−10.0370	1.5731	3904	< 0.0001
BAL	−2.1*10⁻⁵	3.2*10⁻⁶	70	< 0.0001	−1.80*10^− 5	3.1*10^− 6	68	< 0.0001
SW_BA	0.2054	0.0528	70	0.0002
poly (year)¹	4.3896	1.2802	2941	0.0006	4.6529	1.2740	3904	0.0003
poly (year)²	3.0445	0.8343	2941	0.0003	2.7191	0.8727	3904	0.0018
poly (year)³	−1.6641	0.7491	2941	0.0264
SE random effect	Intercept	Dp	error		Intercept	Dp	error
	0.6272	0.0015	0.3993		0.2220	0.0007	0.3799
Model evaluation	R²f	R²m	rRMSE	AIC	R²f	R²m	rRMSE	AIC
	0.644	0.397	11.159	2208	0.580	0.351	3.014	1573

Dp (previous year diameter, cm), BAL (basal area of trees larger than the cored tree, m²), SW_BA (scaled Shannon diversity index), poly (Dp) and poly (year) estimates of orthogonal polynomial of Dp and year respectively with superscript indicating the degree, R²f (pseudo- R² of the full model), R²m (pseudo- R² of the marginal model), rRMSE (relative root mean squared error) and AIC (Akaike Information Criteria)