Understanding primate habitat selection at the micro-scale is crucial for effective conservation, especially for species with specialized habitat requirements and narrow distributions. However, traditional survey methods often struggle to obtain habitat data for target species in complex terrains due to accessibility limitations and insufficient precision. We employed UAV-borne LiDAR technology to investigate the habitat selection patterns of François’ langur ( Trachypithecus francoisi ), an endangered primate that primarily inhabits cliff areas of karst valleys in southwestern China. Based on the location information of François’ langurs recorded by ground transect surveys and UAV observations, we established presence and control quadrats and collected LiDAR point cloud data for these quadrats. We extracted 13 habitat variables covering horizontal structure, vertical structure, and topographic features, and applied Random Forest and Generalized Linear Model combined with SHAP analysis to conduct precise habitat selection analysis. The results indicate that topographic features (elevation, slope) are the primary variables influencing the distribution of François’ langurs, followed by horizontal and vertical structural features (canopy cover, vegetation vertical distribution). Compared with control quadrats, the quadrats used by François’ langurs are characterized by lower elevation, steeper slope, greater canopy cover, and simpler vegetation vertical distribution. Our study emphasizes that the restoration of François’ langur habitat should not focus exclusively on dense, mature forests with complex vegetation structures on both sides of the valley. Instead, efforts should be made to create and maintain secondary forest composite ecological landscapes with specific structures on both sides of steep cliffs. These findings provide an important scientific basis for habitat management and conservation planning of this endangered species.
周源 et al. (Fri,) studied this question.