Main Article Content
The accuracy of diameters and volumes estimates, along the stem of Pinus taeda provided by segmented and non-segmented polynomial taper models, were compared and the effect of trees diameter classes on the quality of the fittings was analysed. 58 trees of Pinus taeda cubed in the farms of PISA - Papel de Imprensa S/A, located in Jaguariaíva county (PR) were separated in 8 diameter classes, and two segmented and two non-segmented polynomial models were fitted for each diameter class and for the total group of data. The segmented models tested were: Clark et al. (1991); and Max & Burkhart (1976). The non-segmented models were: Hradetzky (1976); and Goulding and Murray (1976). Diameters and volumes provided by the four models, fitted with and without the diameter classes control, were analysed by a two-way ANOVA in a randomized block experimental design. Models presented different behaviour for estimating diameters and volumes along the stem of Pinus taeda, and accurate estimates of diameter and volume along the stem profile require diameter classes control. The model proposed by Clark et al. (1991) presented more flexibility than the others, because it can be fitted with or without diameter classes control, except for the diameter class with 32.5 cm and for trees with diameter larger than 45cm. Hradetzky (1976) model, fitted by diameter classes is a simple but accurate model for estimating stem diameter and volume.