Nonlinear models in the study of the cedar diametric growth in a seasonally dry tropical forest

Abstract

The continuous decrease of forest cover in Brazil and the growing environmental pressure on forest-originate products increase the concern with the protection of the native forests remnants and endangered species such as the cedar (Cedrela fissilis Vell.), which has great economic prominence in the country due to the its excellent wood quality. Therefore, studying in detail its growth as a basis for its sustainable management in the forests is necessary. Based on the knowledge that nonlinear models are traditionally well fit for biological processes, this study aimed to compare the performance of the Brody, Gompertz, Logistic and von Bertalanffy models in reconstructing the cedar diametric growth. To that end, we used data from the mean diameter at breast height accumulated over time, obtained from non-destructive samples of 85 trees in an environmental preservation area. The models were fit using the R software considering the autoregressive error structure as of first order. All models demonstrated a good fit, although the Brody model was the best to describe the diametric growth of cedar over time in the studied region.