Amazon Forest provides ecological environmental services, which guide and regulate the ecosystemic activities for its maintenance. It concentrates the largest biodiversity, and is known as the largest rainforest in the world, where it influences in many ways, because it is an excellent store of biomass and atmospheric carbon, while reducing the incidence of the greenhouse effect. The allometric relationships between height-diameter interfere in the estimation of above-ground biomass, thus influencing the carbon stock stored by the forest. For this purpose, the objective of the work was to generate hypsometric equations for two types of forest typologies and to investigate how these estimates influence the inference of biomass. The work was carried out in nine 0.5 ha plots, five of which were found in the terra firme forest area and four in the lowland forest of the Amazonian estuary. Through forest inventory, 2525 individuals were identified and measured in the two forest types. With total height and diameter data, the local hypsometric models were validated and compared the total estimated heights with the heights estimated by the adjusted pantropical model of Chave et al 2014. As it was not possible to obtain any destructive sample of biomass at these sites, the biomass reference values were based on allometrics with height measured in the field, and with the coefficients of variation (CV%) it was inferred what would be the efficiency of each model biomass estimation. On a local scale, the local rectangular-hyperbule (CV = 0.38%) and weibull (CV = 0.65%) local models were the best fit among the local hypsometric models for terra firme and floodplain forests, specifically. In the analysis for two forests together, on a more regional scale, the local weibull model (CV = -0.86%) was the one that obtained the best fit. In the estimative of biomass at the two local and regional scales, the local hypsometric models were the most efficient (local model for terra firma: bias = -0.153%, CV = 35.30%, local model of the várzea: bias = 0.01%, CV = 53.91%, local model of both forests: bias = 0.03%, CV = 62.05%). In addition, local models are best suited for height estimates at local and regional scales, taking into account that pantropical models are efficient for a broader pattern, ie global level
|
