Spatiotemporal changes in dissolved organic matter chemodiversity and its interaction with microbial composition in heavy-metal polluted river

Riverine dissolved organic matter (DOM) constitutes a key reactive carbon reservoir and critically regulates biogeochemical cycling in terrestrial-aquatic ecosystems. However, the chemodiversity of riverine DOM and its interactions with microorganisms in ecologically fragile watersheds remain poorly understood. Here, we investigated the spatiotemporal variations in the chemical composition of surface water DOM and its interactions with microbial communities during a hydrologic year in a heavy-metal polluted river of southern China by integrating optical spectroscopy, ultrahigh-resolution mass spectrometry (FT-ICR MS), and high-throughput sequencing. The results showed that riverine DOM was predominantly composed of protein-like components (tryptophan-like C2 (peak T) and tyrosine-like C3 (peak B), accounting for 69.6% ± 8.87%, based on the EEM-PARAFAC analysis), demonstrating a primarily microbial origin. Moreover, FT-ICR MS indicative of P-containing compounds exhibited higher concentrations in spring and summer, whereas CHONS compounds were more abundant in autumn and winter, indicating the significant influence of human activities on riverine DOM chemodiversity. The molecular composition of DOM (e.g., saturated compounds and lignin) and the structure of the microbial community (including dominant taxa and their relative proportions) exhibited significant spatiotemporal variations. These fluctuations were particularly pronounced at sites S2 and S3, located near the mining area. Seasonal variation in both DOM composition and microbial community structure exceeded spatial heterogeneity. Furthermore, redundancy analysis (RDA) and mantel test analysis indicated that DOM composition and microbial community was influenced by various environmental factors (e.g., pH, temperature, metals) and that terrestrial humic-like C1 and protein-like C2/C3 were significantly correlated with the microbial community composition. DOM components exhibited stronger correlations with fungal communities than bacterial communities, suggesting that fungal communities displayed greater structural and functional responsiveness to changes in DOM composition. This study highlights the underlying mechanism of interactions between riverine DOM chemodiversity and microbes in heavy metal-polluted rivers, and provides important implications for watershed carbon cycling management under such anthropogenic stress.