POTENCIAL DISTRIBUTION OF Pinus herrerae Martínez AT WESTERN JALISCO

Abstract

Based upon its endemism and natural restricted areas at west Jalisco, potential zones were delimited for the distribution of Pinus herrerae. 391 trees were geo-referred which were translated into the same number of accessions of the species to make an environmental description; eight climatic variables were interpolated variables, as well as nine of the environmental kind. With this information, the MaxEnt model was use to classify, in a probabilistic way, its potential distribution. The study area was located at the west of Sierra de Quila. The validation of the model was determined with the Receiver Operating Characteristic (ROC) technique, which is applied in distribution models based on algorithms of presence only. To determine the most important climate and environmental variables that influence the species distribution, the Jackknife test was used. According to the area under the operational curve (AUC) of ROC, both interpolations in the training and testing curves were placed the closest possible, which meant to reach a fit in the model by using 0.959 and 0.958 climate variables. On the other side, with the interpolation of the climatic data with the environmental variables, the highest values for the training (0.973) and testing data (0.971) values were obtained. It can be concluded that the potential distribution of P. herrerae is favored when it grows in a humic cambisoil soil, at an altitudinal range from 1 985 to 2 227m, an average annual precipitation of 882 mm and an average annual temperature of 18 °C.