Uso de modelos estatísticos para classificação de sítios em povoamentos de Pinus taeda na região de caçador - Sc

Abstract

This study aimed to test different statistical models for fitting site index curves and check if the anamorphic curves generated were satisfactory to classify the loblolly pine (Pinus taeda L.) plantations in the region of Caçador, Santa Catarina State, Brazil. The data of dominant height (hdom) used in this study were obtained from permanent plots of fixed area and rectangular shape, with areas of 500 and 625 square meters. The plots were distributed randomly in the stands to cover the largest possible variability of productivity. The selection of the best equation was made by the following fitting and precision statistics: adjusted coefficient of determination (R2aj), standard error of estimation (Syx), standard error of estimate in percentage (Syx%) and graphical distribution of the residuals in percentage. The monomolecular model showed the best performance, so it was chosen for the construction of the site index curves by guide curve method. After construction of the site index curves, the stability of the curves was also tested, which indicated the anamorphic pattern of the curves generated by the equation for site classification of loblolly pine plantations in the region of Caçador, Santa Catarina State, Brazil.