AJUSTE DE MODELOS PARA ESTIMATIVA DO DIÂMETRO DE COPA EM UMA ÁREA DE REFLORESTAMENTO MISTO NA MATA ATLÂNTICA

Abstract

The main objective of this study was to fit regression models to estimate canopy diameter from circumference at Breast height in an Atlantic Forest area located in Laranjeiras county, SE, Brazil. The study was carried out using 30 permanent plots of 600 m² (20m × 30m), systematically distributed, in a recovery area, belonging to the company Votorantim Cimentos SA. In each plot, the measurements were taken of the circumference at the breast height (CBH), for individuals with CBH > 15 cm and cup diameter. From the CAP values, the diameter at the breast height of each individual was estimated. Subsequently, regression models were fitted between the values of diameter at breast height and canopy diameter considering all the species found and for the three most abundant species. The models were evaluated by: adjusted coefficient of determination, standard error of the estimate, the F-value and the graphical analysis of residues. The general model with the best performance presented a coefficient of determination of 0.5 and a standard error of the estimate of 34.8%. When considering specific models by species Cassia grandis Lf., Enterolobium contortisliquum Vell. Morang. and Schinus terebinthifolius it was obtained coefficients of determination of 0.70, 0.55 and 0.41, respectively; and standard error of estimation ranging from 10% to 22%. The relationship between canopy diameter as a function of CBH follows a linear trend and can be estimated by diameter at breast height, which will imply as a thinning indicator and thus predict the maximum number of trees per unit area as they go growing. There is a significant improvement in the models when adjusted by species.