The objective of this study was to evaluate taper models for Corymbia citriodora at 55 months of age in the southern state of Tocantins. Thirty-five trees were used to obtain the diameters with and without bark at the positions 0.2 m, 0.4 m, 0.7 m, 1.3 m, 2.7 m and from this point forward every 2 meters to a diameter of about 1.0 cm in bark. For the evaluation of the models, the following statistics were used: adjusted coefficient of determination, standard error of the estimate, graphical distribution of the residuals, mean absolute deviations, standard deviation of the differences, sum of the squares of relative residues, mean of the percentage deviations, square root of mean error, bias and multiple correlation coefficient. In addition, we evaluated the feasibility of using the binary variable "Tx" in the simultaneous adjustment for data with and without shell. It was concluded, therefore, that the Muhairwe I model obtained the best results for the statistics used, as well as the confirmation of the feasibility in obtaining the diametric adjustment with a single equation through the inclusion of the variable Tx.
|
