Climate-driven habitat loss of Chinese sturgeon in the Northwestern Pacific and implications for marine protected area design

The conservation of endangered anadromous fishes depends on understanding their full-lifecycle habitat requirements, yet the marine phase of Chinese sturgeon ( Acipenser sinensis ) remains a critical knowledge gap despite over 90% of its lifespan occurring at sea. We integrated multi-source occurrence data (satellite tagging, bycatch records, environmental DNA, and online databases) and applied an ensemble modeling framework (Biomod2 with eleven algorithms) to predict habitat suitability across the Northwestern Pacific Ocean (100°E–150°E, 10°N–55°N). After collinearity filtering, seven key environmental drivers were retained, among which mean temperature contributed most, followed by maximum pH, total suspended matter and mean silicate. The ensemble model achieved high predictive performance (median TSS = 0.944) and identified the Changjiang Estuary and coastal waters of Zhejiang and Jiangsu provinces as current distribution hotspots. Under the high-emission scenario (SSP5–8.5), suitable habitats are projected to disappear completely by 2100, accompanied by a significant northward contraction. These projections should be regarded as a potential worst-case scenario under high-emission conditions rather than a deterministic forecast. These findings demonstrate that business-as-usual emissions would eliminate viable marine habitat for this critically endangered species. Conservation priorities should therefore focus on (1) protecting benthic prey resources in the Changjiang Estuary and Zhoushan Archipelago as core refugia, and (2) transitioning from static marine protected areas to dynamic, climate-adaptive zoning informed by real-time habitat predictions. The ensemble modeling approach developed here is transferable to other data-limited, endangered marine species facing similar climate threats.