ABSTRACT
The objective of this study was to select a model of taper considering the influence of stratification of data in four different diameter classes. Six models of taper were evaluated for a commercial plantation of Eucalyptus camaldulensis Dehnh having 53 months, located in the south central region of Tocantins. In this selection, in addition to the distribution of residuals analysis, statistical criteria were used: average absolute deviation, bias, sum of the squares of the relative residue and square root of the mean error. In general, the best models of taper were the Hradetzky, Schoepfer and Biging. Already using data stratified by diameter class, the best model to estimate the stem profile in Eucalyptus camaldulensis Dehnh was to Hradetzky followed by Schoepfer model.
Keywords:
Forest Biometrics; commercial forest and taper