Modelagem de biomassa e carbono do componente florestal em sistema integração lavoura-pecuária-floresta a partir de características dendrométricas e físico-químicas obtidas por intervenção silvicultural

Abstract

The Integrated Crop-Livestock-Forestry (ICLF) system has been showing indicators of sustainability, such as the reduction of greenhouse gas emissions by improving the quality of pastures and, consequently, by providing organic matter that improves the quality of the soil, reflecting an increase in productivity and animal welfare, guaranteeing the provision of important environmental services on rural properties. The aim of this study was to model the biomass and carbon of the forest component in an ICLF system based on dendrometric and physical-chemical characteristics obtained through silvicultural intervention. In a Technological Reference Unit, located in the municipality of Mojuí dos Campos, Pará, three forest species were implemented in 2010, two of which are native to the Amazon biome, Bertholletia excelsa (Amazon Castanha) and Dipteryx odorata (Cumaru), and an exotic one, Khaya grandifoliola (African mahogany), in plots of 0.92 ha spaced 7 m apart x 5 m. After the first thinning of the trees with undesirable phenotypic characteristics, carried out in 2021, dendrometric and physicochemical characteristics were assessed, as well as biomass and carbon modeling of each forest component in the ICLF system. The analyses included diameter at breast height (DBH), total height (Ht), crown diameter (CD), crown length (CL), crown projection area (CA), crown percentage (CP) and other morphometric indices. Carbon content was determined in leaves, branches and stem, while wood density (ρ) was obtained at different humidity levels (apparent, anhydrous and basic) with samples from 85 thinned trees, totaling 425 wood discs. Biomass was calculated from the moisture content of the leaves and thin branches and by cubing the thick branches and stem. Carbon was obtained from the biomass and carbon content of each compartment of the tree. Six mathematical models were adjusted using the biomass and carbon data, with DBH and Ht as predictor variables. Criteria such as adjusted coefficient of determination (R²adj.), standard error of estimate (Syx%), Akaike (AIC) and Bayesian (BIC) information criteria, Willmott's concordance index (d) and graphical analysis of residuals were used to select the best model. The species with the highest dendrometric characteristics was K. grandifoliola, with (DBH = 32.98 cm and Ht = 27.17 m). B. excelsa stood out for having the highest CD, with an average of 24.04 mm and CA of 462.13 m². The highest carbon content was found in the leaves of D. odorata (52.14%) and B. excelsa (50.68%). It can be seen that K. grandifoliola stores the highest carbon content in the stem (49.01%). Wood densities varied between species, with D. odorata showing the highest values (ρ apparent = 0.99 g.cm-³; ρ anhydrous = 0.91 g.cm-³; ρ basic = 0.83 g.cm-³). The Spurr model was the most efficient, with R²adj.= 0.92, Syx% = 26.27 for biomass, and R²adj.= 0.92, Syx% = 32.05 for carbon. At 11 years, the forest component stored an average of 254.25 kg (92.12 kg.ha-¹) of biomass and 120.00 kg (43.48 kg.ha-¹) of carbon. The K. grandifoliola species had the highest biomass (454.21 kg) and carbon (227.07 kg) stocks. The results highlight the efficiency of the ILPF system in capturing and storing carbon, especially in young plantations. Silvicultural practices consolidate the livestock production system, which is subject to certification for providing important environmental services through the adoption of sustainable practices in the Amazon.