Microhabitat selection by the Beale’s eyed turtle (Sacalia bealei) and conservation implications

Understanding the microhabitat preferences of the critically endangered Beale’s eyed turtle (Sacalia bealei) is essential for its conservation. We investigated habitat selection patterns across breeding (April–June, November) and non-breeding (July–October) seasons in the Huboliao National Nature Reserve, China, using radio telemetry and quadrat sampling. By comparing used and control quadrats, we identified key ecological drivers of habitat choice. Our findings revealed that S. bealei prefers habitats with deep stream pools, abundant stone caves, greater fruit abundance, and high benthic species diversity and abundance. Seasonal variations in habitat use were evident, with breeding habitats characterized by lower woody plant density, thinner leaf litter, higher elevations, and increased stone cave availability. To validate these patterns, we applied the Random Forest and LASSO regression models, which confirmed that concealment and resource-related factors were the strongest predictors of habitat selection. The high classification accuracy of both models (Random Forest: 86.11 %; LASSO: 88.89 %) reinforced the robustness of our findings. These results underscore the vulnerability of S. bealei to habitat fragmentation and human disturbance, given its reliance on specific microhabitat conditions. We propose targeted conservation strategies, including riparian vegetation restoration, artificial shelter creation, and community-based anti-poaching initiatives, to mitigate threats. This study has significant implications for the conservation of this critically endangered species, providing vital insights not only for its protection but also contributing to global efforts in turtle and tortoise conservation. By filling a key gap in research, our work offers valuable guidance for the conservation of similar species worldwide.