Logging in the Amazon using reduced impact methods has high potential. However, one of the problems of forest management is an accurate estimation of the wood stock of a forest. Thinning equations allow for the analysis of the volume and fluctuations in diameter along the trunk of a tree, making it possible to estimate the volume, diameter, or height based on known dimensions of the trunk. Therefore, this study aimed to adjust and evaluate two thinning equations in a forest management area in the Western Amazon, in the municipality of Xapuri, state of Acre, Brazil, comparing them and selecting the one that presented the best performance for the study area. The study involved sampling in the Chico Mendes Extractive Reserve, utilizing 53 work units and employing the Kozak and linearized Demaerschalk models. These models were assessed based on their determination coefficient, standard error percentage, and graphical analysis of residues. The Demaerschalk model exhibited superior performance, with R² adj. of 92.88% for commercial height and 95.68% for total height. This research contributes to efficiently estimating tree dimensions, essential for sustainable forest management in the Western Amazon. The application of the Demaerschalk model provides precise measurements, enhancing the accuracy of volume calculations and facilitating effective forest management practices.
Keywords: allometry, Amazon, Demaerschalk model, dendrometry, Kozak model