Modelos alométricos diâmetro-altura para florestas oligotróficas do Norte da Amazônia brasileira

Abstract

The study of the D-H (diameter-height) relationship in tropical forests improves our understanding of biomass stocks and the role they play in mitigating the effects of climate change in the regional and global context. In this sense, the objective of this study was to adjust D-H allometric models for different types of oligotrophic forests in the center-south of Roraima, northern Brazilian Amazonia. The study was based on measuring the diameter and height of 350 trees on a 2.25 km with 5m wide transect. The transect intersects a hydro-edaphic gradient where three forest types (campinarana, ecotone and ombrófila) are defined by environmental factors (soil class and duration of flooding). The five mathematical models tested were non-linear: Weibull, Michaelis-Menten, Log1, Log2 and Log3. The selection criteria of the models were the standard error of residues (RSE), the adjusted determination coefficient (R²adj) and the Akaike Information Criterion (AIC). The specific allometric models that define the total height of each of the three forest types investigated are asymptotically different (p = <5%). Analysis of the residues of the more parsimonious non-linear models showed a tendency to overestimate the total heights for the three forest types. The best fit model (Michaelis-Menten) indicated that general models previously published for tropical regions using diameter as an independent variable may overestimate heights based on stem diameter. The general Weibull models overestimated the heights in the study area in (17.7-105.8%) in the three forest types (Treed shade-loving, ecotone and ombrófila), except in the diameters > 60 (cm) in the forest type of the ombrophilous where the South American model underestimated the heights (13.2-67.6%), overestimating the highest heights in the campinarana, however, in the ecotone area, it overestimated the heights in the diameters <= 20 (cm) and underestimated them in diameters> 38 (cm), the heights were also overestimated in diameters <= 20 (cm), and underestimated in diameters > 25 (cm). It is concluded that the diameter-height relationship differs between the phytophysiognomy of the campinarana with (3-4 months of flood) in relation to the ecotone area (1-2 months of flooding) and the ombrophilous forest (without flooding) , indicating that more restricted environments (campinaranas established in areas with longer flooding periods and poorer soils) have lower biometric standards than forest types with lower restrictions (ombrophilous and ecotone forest located in free zones or with a shorter time period of flooding).