Water quality-carbon coupling in a human-impacted river network: Contrasting CO2 supersaturation and CH4 sink strength along oxidizable loading and seasonal gradients.

Routine river monitoring often evaluates water quality separately from greenhouse-gas disequilibrium, limiting its use for integrated carbon assessment. We investigated dissolved CO2 and CH4 at 16 sites across the Liao River during July and December 2024, a wetter-than-normal sampling year, and related their patterns to hydrochemistry, EEM-PARAFAC-derived dissolved organic matter (DOM) descriptors, and 16S rRNA-based microbial attributes. CO2 was supersaturated at all sites, indicating persistent source potential across the river network, whereas CH4 was predominantly undersaturated, with 27 of 32 observations below atmospheric equilibrium and only localized positive patches. CO2 supersaturation was most closely associated with an oxidizable loading-DOM optical quality gradient: CODMn provided the clearest water-quality signal, while HIX, BIX, and ureolysis further reflected DOM processing conditions. The CO2 response increased most strongly within the observed CODMn range of approximately 2.4-4.5 mg L-1 and remained elevated above ∼5 mg L-1. In contrast, CH4 sink strength was most strongly associated with seasonal temperature structure, with BIX and dissolved oxygen representing related DOM freshness and oxygen-background dimensions. Most microbial diversity and FAPROTAX-inferred functional signals covaried with the same thermal-DOM-oxygen background, except for the selective CO2-related signal of ureolysis. These results indicate that routine CODMn monitoring and DOM optical diagnostics can help identify reaches associated with elevated CO2 supersaturation, whereas CH4 assessment requires seasonally resolved monitoring and targeted evaluation of localized source-like conditions.