MODELING OF THE DIAMETER DISTRIBUTION FOR FOREST SPECIES IN A FRAGMENT OF MIXED OMBROPHYLLOUS FOREST

This study aimed to evaluate the performance of the probability density functions (pdf) Weibull with 3 parameters (3P) and Beta adjusted by Maximum Likelihood method to describe the diameter distribution of some representative species from a fragment of Araucaria Forest. The study area is part of the National Forest Irati (FLONA), where 25 permanent sample plots (100 m x 100 m) were established with a total sampled area of 25 ha, and for the fittings, we used data from all trees with diameter at breast height (DBH) greater than or equal to 10 cm. Adjustments were made for the 20 species with the highest importance value (VI) with data grouped into classes of 5 cm DBH. The nonlinear programming (NLP) was used in order to minimize the deviations between observed and estimated values. To evaluate the adjustments, we used the Goodness-of-fit tests Kolmogorov-Smirnov and Hollander-Proschan, in addition to the Standard Error of the Estimate (%) and Reynolds Index. The results indicated that Weibull 3P and Beta functions presented satisfactory results, and both could be recommended to evaluate the diameter distribution of the selected species and that the use of NLP has not improved adjustments in all cases.