Predicting crown width and length using nonlinear mixed-effects models: a test of competition measures using Chinese fir (Cunninghamia lanceolata (Lamb.) Hook.)

Table 4 Lack-of-fit statistics for CW—DBH basic models

Function no	Fitting statistics of the candidate crown width models						Cross-validation
Function no	− 2LL (df)	AIC	BIC	Bias (m)	RMSE (m)	R²	NMSE	PRESS
[CR1]^b	2576.04 (3)	2582	2598	0.00	0.57	0.4566	0.5535	48.7260
[CR2]^c	2585.43 (4)	2593	2615	0.00	0.57	0.4584	0.5525	48.6443
[CR3]^c	2595.93 (5)	2606	2632	0.00	0.57	0.4613	0.5508	48.5197
[CR4]^b	2552.41 (3)	2558	2574	0.00	0.57	0.4623	0.5505	48.4941
[CR5]^b	2584.56 (3)	2591	2606	0.01	0.58	0.5021	0.5623	49.5767
[CR6]^b	2597.36 (3)	2603	2619	0.00	0.58	0.4823	0.5667	49.9972
[CR7]^b	2601.31 (3)	2607	2623	− 0.00	0.58	0.4176	0.5699	50.1528
[CR8]^b	2601.31 (3)	2607	2623	− 0.00	0.58	0.4176	0.5617	50.0737
[CR9]^b	2601.31 (3)	2607	2623	− 0.00	0.58	0.4176	0.5680	50.1595
[CR10]^b	2563.32 (4)	2571	2593	− 0.00	0.57	0.4490	0.5560	48.9412

CR, crown width; − 2LL (df), − 2 log-likelihood (degree of freedom); AIC, Akaike information criterion; BIC, Bayesian information criterion; Bias, absolute bias; RMSE, root mean square error; R², coefficient of determination; NMSE, normalized mean square error; PRESS, predicted the error sum of squares
^bP value for each coefficient is lower than 0.05 in a T-test
^cP values for one or more than one coefficient are higher than 0.05 in a T-test

ISSN: 1297-966X