Investigation of current and future potential distribution of Astragalus gossypinus in Central Iran using species distribution modelling

Abstract

The potential impact of climate change on Astragalus gossypinus in Central Iran based on grid map 2.5 arc min was analyzed. A stratified sampling was applied through a geographic information system to pick up 587 sample sites (prevalence 0.39). For each sampling site, the presence or absence of given species together with environmental variables was recorded. Two novel statistical techniques, logistic regression tree (LRT) and nonparametric multiplicative regression (NPMR), were used to examine environmental variables related to the current species distribution. Using these models, maps of current potential distribution and potential distribution for a climatic change scenario (2CO2) were generated. Both statistical techniques produced strong and useful models, but NPMR identified a much smaller subset of relevant predictor variables. The model demonstrated that the occurrence of A. gossypinus is highly probable when the precipitation of the wettest month is between 30 and 50 mm and the mean temperature of the wettest quarter is between −2 and +4 °C, but much lower outside this range. Under double-CO2 climatic scenario, predicting a moister and slightly warmer climate in Central Iran, A. gossypinus is expected to move north-eastwards with a decreasing area of distribution.