Abstract:

Aims: The rapid expansion of linear infrastructure poses a significant threat to biodiversity. Developing large-scale optimization pathways for wildlife passages is crucial for regional biodiversity conservation. 

Method: We employ the Maximum Entropy (MaxEnt) model, with emphasis on representative wildlife species in Xinjiang, to predict potential habitats and assess the importance of each. The minimal cumulative resistance (MCR) model computes the least-cost paths influenced by linear infrastructure. We then accurately pinpoint the necessary locations and numbers of passages by integrating current and planned roads and fences to improve habitat connectivity, ultimately proposing an optimized spatial layout for wildlife passages. 

Results: Our findings reveal notable ecological differences in the habitat distributions of various representative species. Carnivores are predominantly found in alpine and subalpine regions, while large herbivores show a preference for arid desert areas. High-resistance zones are mainly located where railways and highways intersect protected areas. Our analysis identified 2,494.98 km of passage regions and 4,314 individual passages. Higher technical grades of road are associated with longer passage regions and a greater number of passages. The total length of passages in important regions surpasses that of key regions, but passage density is significantly higher in key regions. 

Conclusion: Given these findings we suggest implementing differentiated passage construction based on the technical grade of linear infrastructure. Further, enhancing passage layout optimization outside protected areas to effectively preserve regional ecological connectivity is necessary to maintain biodiversity.

Key words: wildlife passage, road, fence, MaxEnt model, MCR model