Eagle view on the habitat distribution of snow leopards (PANTHERA UNCIA) SPECIES

the total global population of this big cat species is estimated to be around 4080-6590. They spread across an area of 2 million square km with most of them being found in China, followed by Mongolia and India. The animal favors a rugged terrain interspersed with steep slopes, ridges, gullies, rocky outcrops, mostly an elevation of of 3000-4500m m. The habitat of this cat is in the vulnerable ecosystem of the mountainouse regions which face the threats of climate change and global warming, making it all the precarious. Apart from these, the devlopmental activities and pastoralism puts pressure on this already fragile ecosystem, making the habitat and the cats all more susceptible to endangerment (S. Hazra et al., 2017) The current study area cover Afghanistan, Discussion and Result: The estimated areas of predicted habitas of snow leopard based on three models General Additive Model(GAM),Randome Froest(RF) and Maximum Entropy(MaxEnt) were modeled (Fig 1). Then the mean of three models calculated and depicted in figur 2. The models were evaluated using AUC and comparasion with other studies result, despite the low availablity of presence data the model results were quite good. The model over predict in few instances like northern face of Hindukush mountain in Afghanistan. The relatively small number of ground observation could be a majore reason. The standard deviation between three models depicted in Fig 3., and input variables correlation depicted in Fig 4. Method and Data: For predictors bioclimatic data obtained from WorldClim and other data which listed as follow: land cover and altitude acquired from other sources (web source). Slope and aspect were generated using DEM data in Arcmap. All raster layer were resampled to 30 arc-sec(~1 Km) resolution to correspond to the original resolution of the WorldClim data. Altogether I used 9 bioclimatic data that are as follow: 1-Annual Mean Temperature, 2-Mean Diuranl Range, 3-Mean Temperature of Warmest Quarter, 4-Mean Temperature of Coldest Quarter, 5-Precipitation Seasonality, 6-Land Cover,7-DEM, 8-Aspect, 9-Slope. For presence data, 68 snow leopard occurrence records from GBIF(the Global Biodiversity Information Facility) that covers snow leopard range countries in Asia were used. Overall, from above mentioned variables elevation, landcover, slope, Annual precipitation contributed the most to the model. For modeling, three model were used to predict the species habitat distirbution that they are as follow: maximum entropy (MaxEnt), Random Forest (RF) and General Additive Model (GAM).