Spatial Distribution and Conservation Prioritization of Medicinal Gymnosperms in China Using an Optimal Set‐Cover Approach

To elucidate the geographical distribution patterns and hotspots of medicinal gymnosperms in China, providing a scientific basis for formulating conservation strategies for this group, we compiled 17,999 occurrence records for 148 medicinal gymnosperm species native to China. Species were categorized into all, endemic, threatened, and nationally key protected medicinal gymnosperms groups. Distributions were analyzed across 943,100 × 100 km grid cells. Priority conservation areas were identified using an optimal algorithm based on an integer linear programming set-cover formulation, which minimizes grids required to represent all species at least once, and compared with the Dobson algorithm. Conservation gaps were assessed by overlaying priority grids with national nature reserves and national parks. Medicinal gymnosperms showed a "more in the south, less in the north" distribution pattern, concentrated in mountainous areas and provincial borders. The distribution hotspots for all and endemic medicinal gymnosperms were in the Hengduan Mountains, while that for threatened and nationally key protected medicinal gymnosperms was in areas such as northern Guangxi. The optimal algorithm identified 41 priority conservation grids, mainly in areas like the border between Guizhou and Guangxi, outperforming the Dobson algorithm (which required 14% more grids on average). Overlay analysis revealed eight conservation gap grids, including high-priority areas in the Hengduan Mountains. Current national reserves inadequately protect medicinal gymnosperms. Targeted conservation measures are needed based on identified gaps in this study. The optimal algorithm provides a resource-efficient conservation tool. However, several limitations should be acknowledged. The use of 100 km grid cells, derived largely from county-level records, may overestimate species' distribution areas and mask fine-scale patterns. Uneven collection may also introduce spatial sampling biases. In addition, our analysis only considered national reserves, potentially overlooking local initiatives. Future work should incorporate higher-resolution distribution data and sub-national conservation designations to refine priority assessments.