Identifying optimized on-the-ground priority areas for species conservation in a global biodiversity hotspot

Threatened species are inadequately represented within protected areas (PAs) across the globe. Species conservation planning may be improved by using public species-occurrence databases, but empirical evidence is limited of how that may be accomplished at local scales. We used the Three Parallel Rivers Region of China as a case to investigate the utility of public species data in improvement in conservation planning. We mapped the distribution of each species as suitable habitat ranges using species distribution models (for 261 plants and 29 animals with >= 5 occurrences) or as point locations (for 591 plants and 328 animals with <5 occurrences). Systematic conservation planning was then applied to identify three optimized portfolios of priority conservation areas (PCAs) for achieving increasing targets of 17, 31, and 50% of the total study area. We then compared the distributions of PCAs in this study with those in two existing PCA datasets. PCAs in this study covered greater areas in the southeastern highly-disturbed regions and along valleys of great rivers than two existing datasets that had a focus on intact ecosystems in remote mountain areas. The three portfolios of PCAs had some overlap with two existing PCA datasets, with the overlapping area accounting for 26.4-39.0% of the total areas of our PCAs. Our PCAs could complement existing PCAs by identifying more priority areas in developed landscapes; this is critical for protecting biodiversity in such areas as they face greater pressures. PCAs in this study received a much lower PA coverage (32.9-43.1%) than existing PCAs (60.2-60.8%) because of biased PA distribution toward mountain areas. Our results suggest that conservation planning based on limited public species data could improve local-scale priority-setting practices. The analysis supports effective integration of species targets in China's new national park system by identifying optimized networks of PCAs.