Modeling the distribution of trees species by diameter class in southwestern Amazonia

Abstract

The aim of the study was to analyze the prediction of the distribution of forest tree species, on a local scale, using occurrence data grouped by diameter class. To estimate the distribution of species the maximum entropy method (Maxent) and the occurrence data are forest management plan was used. The predictor variables were: elevation, slope, height above the nearest drainage (HAND), and normalized difference vegetation index (NDVI) and dot density of forest species. Six predictor variables were selected by species by method of all possible regressions. The models, by species and diameter class, had an average good performance (AUC = 0.7; omission rate = 8.8%), demonstrating the viability of predicting the distribution of species by diameter class. The most important predictor variables were altitude, NDVI, densities Amburana acreana and Clarisia racemosa. However, further studies are needed to clarify whether there is an interaction between forest species or share the same habitat. According to the models, trees of the species Astonium lecointei, Clarisia racemosa and Ceiba pentandra with diameter at breast height (DBH) ≥ 100 cm are more likely to occur in localized higher elevations. This modeling procedure is efficient to increase knowledge about habitat preferences and geographical distribution of species in the landscape by DBH.