Non-Linear Models Applied to Volume Estimation of Peking Wood (Caryocar villosum)

Objective: Our purpose here is to compare nonlinear models applied to estimate the wood volume of the species Caryocar villosum (popularly known as Pequiá) to indicate which one has the best goodness of fit for the location studied. Theoretical Framework: For any planning of sustainable forest management in the Amazon, Brazilian legislation requires inspection and control procedures such as the calculation of the wood volume. It is acceptable to use both the Form Factor and specifically developed equations. In the current literature, there are several volumetric models used to estimate the commercial wood volume. However, commonly used models are not particular to species and localities. Method: All data were obtained through forest inventory with all individuals located in the Caxiuanã National Forest, PA. We evaluated the models through the metrics of goodness of fit: least error of estimation, Akaike’s Information Criterion, coefficient of determination (R2 and R2adjusted), root-mean squared error, profile likelihood, intrinsic curvature and parameter effects curvature, and k-folds cross-validation. Results and Discussion: The model chosen to estimate the volume of the species Caryocar villosum was Husch model which resulted in the equation where V is the commercial wood volume and d is the diameter at breast height (DBH). Research Implications: Equation to estimate the commercial wood volume of Pequiá (Caryocar villosum) in the Caxiuanã National Forest. Originality/Value: Environmental compliance in Brazilian Amazon rainforest with a model which has the main advantage of only needing the diameter variable to estimate the volume.