MODELAGEM DA OCORRÊNCIA DE OCOS E DE EQUAÇÕES DE VOLUME GENÉRICAS E ESPECÍFICAS PARA ESPÉCIES MANEJADAS NA AMAZÔNIA

Abstract

The success of the management of native forests is conditioned to the knowledge of forest characteristics and the improvement of management techniques. Two issues deserve importance, because they directly interfere with forest management: (i) the occurrence of hollows in the trees and (ii) the precision of the estimates of volumetric production. In this sense, the objective was, respectively: (i) to evaluate and model the occurrence of hollows in commercial trees managed in the Tapajos National Forest (TNF), in the state of Pará, Brazil and (ii) to evaluate the performance of specific commercial volume equations by species, comparing them with generic equations for the Forest Management Area (FMA) and per Annual Production Unit (APU). For the study of both questions, a database composed of 30.026 trees of 43 species, inventoried, harvested and cubed in 10 APUs in the TNF, was used. For the study of the occurrence of hollows, the incidence of hollows at the species and tree level was evaluated, as well as the structural characteristics of the hollows. Logistic regression models were adjusted by species according to the diameter measured at 1.30 m from the ground (DBH) to estimate the probability of the hollows occurrence. Equations were also tested to estimate the hollow diameter as a function of DBH. In order to improve the methods of estimating commercial volumes, simple entry equations were selected by species (specific equations) and for FMA and APUs (generic equations). The equations selected were compared using precision measurements. The monetary revenue forecast error was also calculated considering the volumetric production estimates. It was found that 24% of the harvested trees contained hollows, and the species showed different degrees of propensity to the occurrence of hollows. It was identified that the larger the tree size, the greater the possibility of hollows occurring and proportionally larger are the hollows. Most of the trees hollows harvested had small hollows, reflecting a percentage of hollow volume that is not expressive in relation to the total volume of harvested wood. Patterns of occurrence of hollows along the commercial stem were observed, which depend on the species and the size of the trees. Most of the trees contained hollows distributed throughout the stem or in the region of the base of the stem. The logistical models tested by species adequately represented the relationship between the DBH and the probability of hollows occurrence, being valid for use in forest inventory data. It was identified that the species have limit diameters from which the probability of hollows occurrence is significant, which can be considered in the selection of trees for harvest. Predictive models of the size of the hollow were also developed and are valid for use in forest inventory phase. Although hollows are common in commercial trees, knowledge of their characteristics and patterns of occurrence, as well as prediction tools, can assist managers in control of trees hollows. Regarding the commercial volume equations, it was found that the use of specific equations results in more accurate estimates of production, which, consequently, minimize errors in the forecast of monetary revenue, reducing the risks of economic losses. Therefore, in addition to being statistically valid, equations by species represent the best alternative for obtaining more accurate estimates of the commercial volume in TNF, recommending its use.