Individual Tree Basal Area Increment Models for Brazilian Pine (Araucaria angustifolia) Using Artificial Neural Networks

This research aimed to develop statistical models predict basal area increment (BAI) for Araucaria angustifolia using Artificial Neural Networks (ANNs). Tree species were measured their biometric variables and identified at the level. The data subdivided into three groups: (1) intraspecific competition with A. angustifolia; (2) first group of that causes interspecific (3) second angustifolia. We calculated both dependent independent distance described indices, considering impact stratification. Multi-layer Perceptron (MLP) ANN was structured modeling. main results that: (i) input size most significant, allowing us explain up 77% BAI variations; (ii) spatialization competing trees contributed significantly representation competitive status; (iii) separate each improved performance models; (iv) besides competition, also proved be important consider. developed showed precision generalization, suggesting it could describe a common in native forests Southern Brazil potential upcoming forest management initiatives.