| Summary: | Due to a dramatic increase in livestock, the impact of grazing on animal communities is a global conservation concern, especially in temperate forests. Recognizing the interaction between livestock and wildlife is essential for developing more effective and holistic strategies for wildlife conservation and sustainable livestock management. In this study, we conducted systematic camera trapping at 129 sites within the central portion of the Giant Panda National Park in China to assess the interspecific relationships between livestock and 21 sympatric wild species. Using the joint species distribution model, we fitted a null model and a constrained model to investigate the distinction between the spatial distribution relationship of species (as raw associations) and the interactions between livestock and wildlife (as residual associations). The results of raw associations showed that livestock exhibited a significant positive spatial co-occurrence with Temminck's tragopan, Himalayan porcupine, Chinese serow, Reeves's muntjac, and forest musk-deer (p < 0.05). Meanwhile, livestock exhibited negative spatial associations with greater hog badger, giant panda, Chinese goral, and Sichuan takin (p < 0.05). The variance partitioning of the constrained model showed that environmental covariates explained 77.1 % of the species co-occurrence pattern, with 35.5 % variation attributed to geography (including elevation, slope and aspect), 22.2 % to the distance from settlement. After accounting for environmental filtering, the residual associations revealed that there were no significant posterior probabilities between livestock and any of the species (p > 0.05), indicating no strong evidence of negative interactions between livestock and wildlife. Since livestock grazing is a source of income for residents, grazing management policies within the Giant Panda National Park should not prohibit all grazing practices.
|
|---|