Modelagem em nível de povoamento para Eucalyptus sp. com estratificação em classes de precipitação pluviométrica

Abstract

The objective of this study was to perform the modeling of growth and yield for Eucalyptus sp. stand, considering areas stratified by pluviometric precipitation classes. The data used were obtained from Eucalyptus grandis and Eucalyptus urophylla clonal plantations from permanent plots of a continuous forest inventory, with ages from 15 to 84 months, located in the northeastern of Bahia State, Brazil. Initially, it was fitted hypsometric and volumetric models, where the latter one fitted by the traditional way and also stratified by diameter classes, and subject to the F test (Graybill, 1976) to detect possible statistical differences between the estimates. Then, models were tested to estimate dominant height as a function of age, using the data without stratification (traditional) and stratified by precipitation classes. To examine the possibility of statistical differences between the estimates of dominant height in each class, it was used the Dette and Neumeyer (2001) test. For the growth and yield modeling process, it was used the Clutter (1963) model. Then, the initial basal area was based on its arithmetic mean at each site specified, both for the adjustment without stratification (traditional method) and for the stratified area by precipitation. In order to investigate possible statistical differences of the estimates of yield between stratified and non stratified data with their respective observed values, 120 plots were selected and subject to the F test (Graybill, 1976). The results showed that there were no statistical differences between the individual volume estimates by the traditional method and stratified by diameter classes with their respective observed values. The Dette and Neumeyer (2001) test showed statistical difference between the estimates of dominant height in the different classes of precipitation. However, in terms of projection of yield, there were no statistical differences between the two studied alternatives, where both provided estimates close to the observed values, with slight superiority of the non stratified data by precipitation classes.