The ecological niche as a tool for predicting potential areas of two pine species

Abstract

Modelling the optimum ecological niche after the species potential distribution, a tool to predict potential production sites is an option for delimitation of best seed stands as Forest Germplasm Production Areas (UPGF) The aim of this study was to model the potential distribution of Pinus pseudostrobus and P. oocarpa in Chiapas, Mexico by mapping topographic, climatic, edaphic, ecological and ecological niche models (MaxEnt). Data for 220 site presence points for P. oocarpa and 52 for P. pseudostrobus were obtained from the World Information Network on Biodiversity, Global Biodiversity Information Facility, the Missouri Botanical Garden and Herbario Nacional de México [National Herbarium of Mexico (MEXU)]. The potential distribution for each species was modelled after 500 and 1 000 iterations through Logistic, Cumulative and Cloglog regressions. Statistical validation was performed with 28 % of the data for each species through the technique Crossvalidate and Bootstrap. The best fit model was the Logistic type with crossvalidate type validation. Values of area down the curve (AUC) for estimated and validated data were 0.882 (P. oocarpa) and 0.947 (P. pseudostrobus). The most influent variables for species presence or absence were altitude with 84.5 y 97.3 % for P. oocarpa and P. pseudostrobus, respectively. Results from the best model allowed the delimitation of optimum sites to establish UPGF.