The spatial distribution of species plays a very important role in the management planning, protection, restoration, development, and eva‎luation of ecosystems. The lack of data and the large size or inaccessibility of the study area limit studying the habitat distribution of plant species. Species distribution modeling (SDM) can be used to overcome these limitations. This study aims to generate potential distribution maps of Glycyrrhiza glabra L. using Maximum Entropy Methods (MaxEnt), support vector machine (SVM), and Bayesian belief network (BBNs) in Chaharmahal and Bakhtiari province (with an approximate area of 16419 km2). Physiographic data such as elevation, direction, and slope) of the study area was extracted from the Digital Elevation Model (DEM), and the climatic layers were obtained from the Chelsa database (1981-2010). The occurrence of G. glabra was recorded randomly in 130 points (65 presence and 65 absence points) and the mathematical relationship between the occurrence of the species and environmental factors was explored using MaxEnt, BBNs, and SVM. The most effective variables in the habitat suitability of the species were identified by running the MaxEnt model with all the environmental variables and conducting a jackknife test. A new MaxEnt model was then built using the selected variables that had the highest contribution to the species occurrences. Temperature annual range (bio7), precipitation of coldest quarter (bio19), precipitation of driest quarter (bio17), mean diurnal range (bio2), and precipitation of driest month (bio14) were identified as the most important variables in determining the species habitat suitability based on the MaxEnt model. The produced MaxEnt model had high accuracy, as indicated by ROC area under curve plots (0.83). The BBNs model was developed using the same variables as the MaxEnt model along with elevation as the experts believe that it is an important variable that affects the species occurrence. The sensitivity analysis of the BBNs model showed that precipitation of coldest quarter (bio19), temperature annual range (bio7), precipitation of driest quarter (bio17), precipitation of driest month (bio14), elevation and mean diurnal range (bio2) were the most important variables in determining G. glabra habitat suitability. The results of the BBNs model showed that more than half of the province meets the climatic requirements of the species. In the next step, the SVM model was implemented with the same environmental variables as the previous models. The Kappa coefficient and ROC area under curve plots of SVM model were 0.87 and 0.98, respectively, and indicated that this model had an excellent agreement with the field data. According to the SVM model, 28% of the study area (4480 km2) had good and excellent climatic potential for growing the species. The response curves of the species obtained from the SVM model with respect to physiographic factors showed that the highest occurrence probability of this species occur in areas with an altitude of 1500 to 2000 meters above sea level. Also, G. glabra is more likely to occurs in areas with a slope between 0 and 10 degrees, and its occurrence decreases with the increase of the slope. This species is mainly found in the margins of agricultural and rainfed lands.

حسين بشري , مصطفي تركش اصفهاني

رضا جعفري , پوراندخت گلكار تفت