Incorporating Crown Dimensions into Stem Height and Basal Area Growth Models for African White Wood (Triplochiton scleroxylon)
Four crown dimensions (crown diameter, crown projection area, crown length and crown ratio) were each, incorporated into nonlinear individual tree total height and basal area increment models for African white wood (Triplochiton scleroxylon K. Schum). The basic height/basal area growth model was formulated as a function of the tree size (i.e. total height or basal area), age and tree form (i.e. taper which is the ratio of diameter at breast height to total height). These increment models, with and without crown dimension were fitted to a modelling data set and the statistical significance of each of the crown dimensions was examined. All the models were then compared for predictive ability using an independent validation data set. The results obtained were similar for both the total height and basal area increment models. All the crown dimensions except crown projection area were found to be significant in the basic models (i.e. models without crown dimensions) in terms of model fit. However, for predictive ability, increment models (for both total height and basal area) with crown ratio term were found to have the smallest bias and the greatest prediction precision of all the models examined both with and without crown dimension. Therefore, for tree total height and basal area prediction of African white wood, models with added crown ratio term were found to be most appropriate.