Countrywide Distribution Modelling of the Persian Leopard Potential Habitats on a Regional Basis in Iran

Abstract

This chapter is dedicated to assessing the Persian leopard potential distribution in Iran on a regional basis that aims to address four objectives and a null hypothesis. Objectives are concerning (1) estimation of the leopard potential distribution, (2) possibility of a major fragmentation in the Persian leopard range in Iran as first mentioned by Sanei et al. (2016), (3) prediction of landscape corridors which can improve the distribution pattern connectivity and (4) the main environmental variables that contribute to assessing the predictive maps. The null hypothesis addresses the variability of permutation importance of the environmental factors in accordance with the regional variability of environmental characteristics. Due to the variability of the environmental characteristics across the country and the leopard putative range which includes almost 30 provinces out of 31, the area has been innovatively divided into five significantly dissimilar regions as discussed in the previous chapter. Subsequently, MaxEnt modelling is conducted in a regional context using a total of 17 variables including 12 natural and 5 human factors together with more than 550 well distributed leopard occurrence data in all regions. Environmental variables have been tested for possible correlation prior to the modelling procedures. Area under the curve (AUC) was used to test the model fit to the data set. Jackknife test was performed to assess the contribution of environmental variables to the MaxEnt models. Fifteen replications with test percentage of 20% were used for validation. Additional evaluation of the predictive models was conducted by assessing the potential habitat distribution maps via the expert/local knowledge of 150 individuals from all five regions. Findings support that the Persian leopard range in Iran is in the process of a major fragmentation to the northern and the southern parts. Accordingly, two landscape corridors providing vital linkages to connect leopard potential habitats in a metapopulation scale are identified. Developed predictive maps in this chapter are a basis for the researches presented in Chaps. 5, 6 and 7. Authors believe that MaxEnt modeling on a regional basis has considerably improved the accuracy of the predictive maps that eventually formed the countrywide potential distribution of the Persian leopard potential habitats in Iran.