Spatial dependence of dendrometric variables in different ages and sample intensities on a eucalyptus stand

Abstract

Understanding the spatial variations of dendrometric variables in a continuous forest inventory is essential to support management actions, in addition to allowing sample intensities that reflect accuracy with lower inventory cost. The objective of this study was to evaluate the spatial dependence structure of dendrometric variables over time and in different sample intensities in eucalyptus stands. The hypotheses considered were that the spatial dependence structure of dendrometric variables changes along the growth in eucalyptus stands and the sample intensity influences this structure. The variables diameter at 1.30 m of soil, basal area, total height, mean height of dominant trees and volume of wood were obtained in continuous forest inventory, at 3.5, 4.5 and 5.5 years, from 80 plots (400 m²) distributed randomly in the stand (394 ha) located in Abaeté, Minas Gerais state. Were evaluated the sample intensities of one plot every 4.9 (n = 80), 7.3 (n = 54), and 16.4 (n = 24) hectares. The highest sample intensity was reference for the others. The spherical, exponential and gaussian semivariance models were adjusted to the experimental semivariogram, where the best fit model was used by ordinary kriging in the spatialization of the analyzed variables. The results showed a predominance of strong spatial dependence of dendrometric variables, regardless of age and sample intensity, especially for mean height of dominant trees and volume of wood. The lower sample intensity influenced negatively the spatial dependence of the basal area at all ages. The structure of spatial dependence is not influenced by the increase in the age of the stand, and the geostatistical analysis of these variables is recommended in a continuous forest inventory, considering the sample intensity of one plot of each 16.4 hectares.

Keywords:
Continuous forest inventory; Semivariance; Kriging; Geoestatistic