Influence of plot size in Weibull’s diameter distribution function precision in a primary forest in Central Amazon

Abstract

This research aimed to evaluate how the observed data of Amazonian forest can be fitted to Weibull’s probabilistic function and the influence of plot size to the fitness as well. The Weibull probabilistic function is an important tool for forest management. In this study it was used 15 different plot sizes, installed in the INPA’s Experimental Site ZF2 in Manaus, replicated 5 times for the larger sizes and 10 times for the smaller ones. It was collected the diameter of adult trees (DBH > 10 cm), and distributed into 11 diameter classes, with 10 cm interval. Weibull parameters were estimated by Maximum Likelihood method, and the probability of the expected frequency was estimated for each diameter class. The difference between observed and expected frequency generated a set of Chi-squared (χ²) for a 5% critical level (α = 0.05). The main results were: (i) the observed data were fitted by Weibull function appropriately; (ii) the precision obtained by the Weibull function was not influenced by plot size; and (iii) only one function is good enough for Manaus forests, independently of plot size.