Tamanho de parcela amostral para inventários de espécies não madeireiras da Amazônia Central

Abstract

This study aimed to verify if there is a plot size that provides precise stock estimates from multiple non-timber species, and thereby increase viability of forest management plans. Data from forest inventory of a permanent plot of 30 ha at Tropical Forestry Experiment Station of the National Institute for Amazon Research were used, which included all individuals with DBH ≥ 5 cm. The non wood species selected were: Açai (Euterpe precatoria), Bacaba (Oenocarpus bacaba), Bacabinha (O. minor), Acariquara Branca (Geissospermum sericeum), Acariquara Roxa (Minquartia guianensis), Breu Branco, Breu Vemelho and Cipó-titica (Heteropsis spp.). The roots of Heteropsis spp. were counted and classified as mature and immature. From 0.04 to 1.5 ha plots were tested on simulations of forest inventory using simple random sampling at 20% of intensity. The simulations were performed for each species in three diameter class for all tree species, palms, and finally, trees plus palm trees. Were proposed plot sizes that result precise estimates (smaller sampling errors) of basal area and abundance, lower coefficient of variation and whose form was appropriate to the species characteristics. None of the 22 plot sizes tested resulted estimates with sampling error of less than 20% for all species selected in this study area. Such precision was possible only for some species (Breu branco, Breu vermelho and Heteropsis spp.), for groups of life forms (tree species) and when included both life forms (trees and palm trees).