3.1 Distribution of grey wolf and golden jackal
Among all the model’s RF and GLM represented the highest and lowest performance in predicting habitat suitability for both species, respectively (Table 1). All five employed models produced good discriminating power; however, the models’ accuracy was better for golden jackal compared with grey wolf. For grey wolf, elevation, topographic roughness and distance from agriculture lands were the strongest predictors; however, for golden jackal, distance from agriculture lands, human settlements and topographic roughness were the most important variables predicting occurrence in the study area (Table S1).
Golden jackal showed a positive association with increasing distance from agriculture lands, roads, human settlements and elevation (Figure S1). Besides, it showed a decrease in occurrence rate with increasing NDVI, roughness and distance from dumpsites (Figure S1). Grey wolf had a positive association with increasing distance from agriculture lands, roads, human settlements, NDVI (Figure S2). Also, grey wolf showed a decrease in occurrence rate with increasing roughness (Figure S2).
Prediction of the ensemble models for grey wolf and golden jackal revealed that large parts of the landscape had the potential to support the occurrence of both species (Figure 2). However, the predicted suitable areas for gray wolf were more concentrated and spatially demarcated. 75 percent of the area was suitable for grey wolf and slightly less (74%) for golden jackal (Figure S3).
Fig 2. Predicted suitability of the study area for grey wolf and golden jackal based on the combined result of five SDMs.
Table1: Accuracy evaluation of the different models’ models (TSS, AUC) used to predict distribution of grey wolf and golden jackal in central Iran.
3.2 Core habitat:
Our connectivity simulation modeling for grey wolf revealed that core habitats are extensive, concentrating in the study area’s southern parts. Among the identified core habitat, eight are more extensive than 2000 km2. The largest and most important core area (C1 in Figure 3) is of 49800 km2 and is located in the south part of the landscape (Figure 3). The second largest and most important core area, based on size (6200 km2) and strength (sum of kernel value), occurred in the southwestern part of the landscape (Haftadgholleh and Alvand protected areas and Rasvand Wildlife Refuge). Also, an average of 37.84 % of the identified core habitats for the grey wolf is covered by Protected Areas Networks (Table 2). The highest overlap between core habitats and Protected Areas networks was observed in the southern part of the landscape which three CAs (Haftadgholleh and Alvand protected areas and Rasvand Wildlife Refuge) were covered by the most important identified core areas. The largest and most important core area (C1 in Fig 4) for golden jackal, according to size (38800.20 km2) and strength, is in the south parts of the study area (Figure 4 and Table 2). Among the predicted core habitats of this species, 18.6 % are covered by IUCN Protected Areas Networks.
Fig 3: Grey wolf core areas at dispersal ability 50, 100, 150 and 200 km respectively and network of protected areas and roads.
Fig 4: Golden Jackal core areas at dispersal ability 50, 100, 150 and 200 km respectively and network of protected areas and roads.
Table 2. The extent and percent of core habitats covered by current conservation networks for grey wolf and golden jackal in Central Iran. The median value of habitat suitability for presence points was used as the threshold to define the highly suitable habitats.