Context: Habitat loss and fragmentation are main drivers of biodiversity loss and are occurring at an accelerating rate globally. Carnivores are often subject to persecution and conflict with humans, typically reside in low densities and require large areas for core habitat and dispersal, which makes them especially vulnerable to habitat loss and fragmentation.
Objectives: This paper identifies, maps and analyses habitat core areas and connectivity linkages for three mountain-residing carnivore species of high conservation importance in Iran: Persian leopard (Panthera pardus saxicolor), Eurasian lynx (Lynx lynx) and Pallas's cat (Otocolobus manul). Methods: We used ensemble habitat suitability modeling, and compared the accuracy of ensemble models with seven separate models based on AUC and TSS. We applied resistant kernel and factorial least-cost path analyses to identify population core areas and corridors across the full distributions of the three species in Iran. Results: Mean annual temperature, vegetation greenness (NDVI), and slope were among the most important predictor variables for all three species. We found ensemble modeling outcompeted all single-method models in terms of AUC. We found low overlap between predicted corridor locations of our modeled species with Protected Areas. Conclusions: Given the fragmented populations of our studied species in Iran, conserving them will require integrated landscape-level management to protect corridors and enhance connectivity, especially outside of Protected Areas. Optimized landscape management to conserve these species will likely promote conservation of montane landscapes and their inhabitants in Iran.