Efeito da classificação das árvores em grupos arquitetônicos na estimativa de biomassa florestal acima do solo

Abstract

Many equations for estimating biomass have already been developed, however , they are not yet sufficiently accurate, constituting challenges around the efficiency of indirect methods of obtaining this variable. Some researchers suggest the inclusion of canopy variables in the regression models, due to the relevance of this component in the composition of the total biomass. However, it is known and the complexity associated with field collection and inclusion of these variables in statistical models. This context prompted the following work question: "Is the allocation of biomass above ground influenced by the architectural pattern of the canopy of trees?”. In this perspective, this study aims to verify whether the classification of trees in architectural groups improves the accuracy of the above-ground forest biomass estimates. With the following specific objectives : (i) Check if there is a difference between the groups in terms of canopy architecture; (ii) Check if there is a pattern of behavior of the dendometric variables and the biomass variables in the groups; (iii) Fit regression models for above-ground forest biomass by architectural group. For this, an architectural classification of trees from native forests located in the south of Brazil was carried out, using the following criteria: a) sympodial or monopodial; b) crown axes predominantly orthotropic or plagiotropic. Subsequently, a descriptive statistic was made, followed by MANOVA, discriminant analysis and cluster analysis. Finally, fit and statistical evaluations of the performance of two regression models were performed. The results showed that: (i) the sympodial trees differ of monopodials in terms of canopy architecture; (ii) in the sympodial trees, it was not possible to find a pattern of the biomass distribution nor of the relation between height of the morphological inversion point and crown length; (iii) monopodial trees, show a pattern of behavior d the distribution of biomass between stem and crown, and between the height of the morphological inversion point and crown length in the largest individuals; (iv) biomass estimates are more accurate when carried out separately from and monopodials sympodial.