Phase-specific behavior of PAHs and OPAHs across the sediment-water continuum in a nutrient-enriched agricultural river.

Agricultural rivers receive nutrients together with polycyclic aromatic hydrocarbons (PAHs) and oxygenated PAHs (OPAHs), yet whether their nitrogen-associated behavior differs across the sediment-water continuum remains unclear. Using paired surface-water and sediment samples from the Zhangshi River, China, we compared the occurrence, partitioning, and environmental associations of PAHs and OPAHs, and evaluated transformation-related interpretations experimentally. ΣPAHs and ΣOPAHs ranged from 12.4-84.1 and not detected (ND)-189 ng/L in water, and from 14.5-934 and 5.11-156 ng/g in sediments, respectively. Parent PAHs provided a transport-retention reference: aqueous PAHs were associated with hydrometeorological conditions, whereas sedimentary PAHs were linked to organic-carbon-rich retention settings. Random forest coupled with Tree SHapley Additive exPlanations (RF-TreeSHAP) revealed opposing nitrate (NO₃⁻)-associated patterns for OPAHs across phases. Positive SHAP contributions to aqueous OPAHs occurred mainly above a dataset-specific NO₃⁻ transition near 0.79 mg/L under oxygenated, warmer, light-exposed conditions, whereas sedimentary OPAHs showed more positive contributions below approximately 0.43 mg/L NO₃⁻ under organic-carbon-rich interface conditions. Covariance and partial-correlation analyses indicated that the aqueous NO₃⁻-OPAH association was embedded within a broader temperature-dissolved oxygen (DO) and seasonal context, while sedimentary OPAHs reflected hydrological and depositional/nitrogen-cycling backgrounds rather than a direct NO₃⁻ response. Controlled fluorene (FLO) experiments showed enhanced reactive oxygen species generation with NO₃⁻, a 15.4% increase in the apparent phototransformation rate constant, and formation of 9-fluorenone, supporting the aqueous interpretation. Quantitative polymerase chain reaction and grain-size analyses constrained the sedimentary interpretation to a broader nitrogen-cycling and depositional context. These results provide a testable phase-comparative framework for evaluating nutrient-contaminant interactions in nutrient-impacted rivers.