Modelagem e estimativa de biomassa acima do solo em floresta no estado do Pará

Abstract

Understanding and quantifying forest biomass are essential steps for efficient natural resource planning. Furthermore, obtaining detailed information on biomass allows us to better understand the role of ecosystems as carbon reservoirs, contributing to more accurate estimates of carbon emissions and storage, fundamental factors in the context of global warming. This study aimed to evaluate the biomass stock in the Iriri State Forest (FLOTA), located in the municipality of Altamira, Pará. The research was carried out using indirect methods, using allometric equations to improve monitoring techniques, covering a total area of 434,889 hectares. The inventory was conducted in 2013, using a cluster sampling approach, with 27 clusters installed, each containing 8 subunits of 4,000 m², totaling 216 sampling units. Forest structure analyses followed conventional methodologies, including descriptive statistical calculations and assessments of horizontal structure, considering absolute and relative density, absolute and relative dominance, absolute and relative frequency and the Importance Value Index (IVI). Furthermore, the Schumacher-Hall volumetric model, developed specifically for the region, was used, which presented satisfactory statistical criteria. The 10 species with the highest IVIs were selected, whose basic densities were consulted in the Brazilian Timber Database. To determine the real aboveground biomass, 12 allometric models were tested, 6 single-input and 6 double-input. The best model was chosen based on the adjusted coefficient of determination, residual standard error of the estimate and mean percentage deviation (MPD). The results indicated that, among the 10 species analyzed, Tetragastris panamensis presented the highest IVI. The distribution of biomass by diameter classes followed an inverted "J" pattern, indicating a greater concentration of individuals in the smaller diameter classes and a reduction in frequency as the diameter increases. Among the families, Fabaceae had greater representation due to its greater number of individuals, while Bertholletia excelsa stood out due to its high capacity to accumulate biomass. The statistical analysis indicated an average of 114.91 Mg ha⁻¹ of above-ground biomass, with a confidence interval between 104.08 Mg ha⁻¹ and 125.74 Mg ha⁻¹, and a relative sampling error of 9.42% at a 95% confidence level. In allometric modeling, diameter at breast height (DBH) showed a strong correlation with commercial height (0.62) and actual biomass (0.86), indicating that these variables significantly influence biomass. Among the 12 models tested, model 11, with variables β₀ (-4.2495), β₁ (2.1086) and β₂ (0.7293), obtained the best performance, presenting a high coefficient of determination (98.08%), a recalculated standard error of the estimate of 35.87% and a DMP% of -1.94%. These results prove the effectiveness of using tree volume and basic wood density in allometric modeling to quantify forest biomass.