Otolith microchemistry and chemoscape matching accurately reconstructed the spawning ground utilization of river fish

Freshwater habitats are globally degraded, threatening ecosystem functioning and services. Fish spawning grounds are vital for population maintenance and biodiversity. However, detailed empirical data on fish habitat use in large rivers remain scarce due to tagging limitations. This study reconstructed spawning ground utilization of Clupisoma yunnanensis in the upper Nu-Salween River by matching otolith microchemistry with high-resolution water chemistry maps (chemoscapes). We analyzed otolith and water element ratios (Mg:Ca, Mn:Ca, Sr:Ca, Ba:Ca) using electron probe microanalysis (EPMA) and inductively coupled plasma mass spectrometry (ICP-MS), respectively, and established their quantitative relationships. These data were then applied in a random forest (RF) model to reconstruct fish spawning ground distribution, with validation performed through environmental niche analysis. Results revealed significant core-to-edge otolith variations and spatiotemporal water chemistry heterogeneity, indicating diverse habitat use. The model identified downstream areas as primary spawning grounds with 98% accuracy, and niche analysis showed 83% concordance with traditional observations. This work establishes a novel biogeochemical framework integrating otolith microchemistry, chemoscapes, and machine learning to accurately reconstruct critical fish habitats. The approach resolves the spawning ecology of C. yunnanensis and offers a transferable method for conservation in data-scarce basins worldwide.