Metodologia Complementar para Avaliação de Modelos de Afilamento do Fuste de Eucalipto (Evaluating Nonlinear Taper Functions for Different Eucalyptus Genotypes)

Functions which predict taper tree can be used to provide accurate information to forest management. Wrong models providing erroneous data can lead to substantial economic losses. This study compares non-linear taper models for Eucalyptus sp. genotypes using an additional assessment, in order to determine which best performed data. This study also aimed to analyze if there was significant effect of genetype on profile of tree bole. 70 sample-trees were collected to each treatment. The treatments were five different Eucalypstus sp. genotypes under the same stand density (3x3m). Six non-linear taper models were fitted: Biging (1984), Demaerschalk (1972), Garay (1979), Kozak (1988), Max and Burkhart (1976) and Ormerod (1973). The variables measured was: diameter at breast height (DBH), total height (Ht) and diameter (di) at different heights above ground. In order to evaluate the accuracy of model fitting, the study used the correlation coefficient, standard error of estimate, and other supporting statistics, such as: mean Bias, root mean square error (RMSE), mean absolute deviation (MAD), and graphical analysis of error. The additional evaluation, Islam et al. (2009) modified, embraced all model performances into a "categorizing graphic", in which the best performances could be identified . All models showed good estimates, but Kozak (1988) described the diameters along the bole with the best accuracy. The study also showed that there is significant effect of genotype on eucalyptus stem taper.