Measures to improve the accuracy of species distribution models for effective conservation of endangered species

Accurate identification of ecologically sensitive areas, particularly those harboring endangered species, is essential for ensuring that development projects proceed without causing irreversible biodiversity loss. Species distribution models (SDMs) are pivotal tools in addressing this issue, yet they are constrained by challenges such as the overestimation or underestimation of potential habitats and inadequate representation of species-specific ecological traits, especially for endangered species. To address these limitations, this study refines the SDMs for the endangered species Mankyua chejuense by employing a novel approach that incorporates weighted occurrence data and optimized pseudo-absence point selection. These advancements not only mitigate the intrinsic biases of SDMs in habitat estimation but also provide a more nuanced reflection of the species’ ecological characteristics. This improved modeling framework holds substantial implications for developing more effective national and regional biodiversity conservation strategies, and supports the broader application of this refined approach to other species in the future.