DOI: 10.1002/clen.70237 ISSN: 1863-0650

Socioeconomic Drivers Reshape Dissolved Organic Matter (DOM) Sources in Highly Regulated Rivers: Evidence From the Yellow River Basin

Ji Qi, Haojing Zhang, Dongping Liu, Shixiang Zhang, Yan Hao, Huibin Yu

ABSTRACT

Accurate pollution source identification is critical for effective watershed management in highly regulated river basins, where traditional water quality indicators often lack source specificity or fail to capture the complex influence of socioeconomic drivers. This study developed an integrated methodological framework combining three‐dimensional excitation‐emission matrix (3D‐EEM) fluorescence spectroscopy, a Bayesian mixing model (MixSIAR), and partial least squares structural equation modeling (PLS‐SEM) to quantify dissolved organic matter (DOM) sources and elucidate their anthropogenic driving mechanisms in the Ningxia reach of the Yellow River. PARAFAC analysis resolved five fluorescence components, while MixSIAR and PLS‐SEM were utilized to apportion sources and examine multidimensional socioeconomic impacts. MixSIAR results revealed that aquaculture non‐point sources (ANS) dominated the basin (53.6%–66.1%), with a clear longitudinal transition from upstream autochthonous inputs (27.3% in Zhongwei) to intensified urban point sources peaking in the regional capital, Yinchuan (38.6%). PLS‐SEM demonstrated that unified socioeconomic factors—urbanization, population density, and GDP—significantly drive both water quality (0.575) and DOM composition (0.524), while a nonsignificant pathway (0.187) between them underscores DOM's superior sensitivity to anthropogenic disturbances. This integrated framework substantially enhances the resolution of source apportionment, providing a robust and transferable diagnostic tool for tracking environmental impacts. Overall, the findings highlight the necessity of source‐oriented management prioritizing aquaculture control and urban discharge regulation. Future research should incorporate finer‐resolution geospatial data and localized land‐use indicators to further refine these complex driving mechanisms and support adaptive, data‐driven watershed governance and restoration efforts.

More from our Archive