A mechanistic comparison of two constructed wetlands responding to atrazine in agricultural runoff: Divergent removal efficiencies and eco-physiological strategies.

Two constructed wetlands (CWs) included horizontal subsurface flow (HSFCW) and surface flow constructed wetland (SFCW), were established to evaluate performance and underlying mechanisms in removing atrazine and conventional pollutants from non-point source pollution. The introduction of atrazine resulted in a notable decrease in the removal efficiencies for COD, TP and NH4+-N across CWs. Notably, when hydraulic loading rate (HLR) was reduced, a significant recovery was observed in HSFCW, with the removal rates of COD (76.85 ± 12.82%), TP (81.97 ± 2.87%), and NH4+-N (54.14 ± 4.11%) reaching or surpassing the baseline levels of the atrazine-free. In contrast, the removal efficiencies only partially recovered, remaining below their pre-atrazine levels in SFCW. Although the atrazine removal efficiencies reached 50.83 ± 4.86% in SFCW and 59.33 ± 7.50% in HSFCW, removal mechanisms differed. In SFCW, photosynthetic oxygen production and microbial degradation of atrazine were critical steps. Photosynthetic bacteria like Rhodoplanes etc. improved the elimination of organic compound under oligotrophic conditions, and NH4+-N oxidation by creating aerobic condition. Due to the simplicity of their metabolic network, SFCW exhibited compromised resilience when exposed to shock loads or contaminants. In HSFCW, plant uptake, substrate adsorption and microbial decomposition of atrazine, and stress resistance of system played an important role. The anti-adaptive bacterium (Luteimonas, Crenothrix, Sediminibacterium) enabled the HSFCW to maintain robust removal efficiencies for both atrazine and conventional pollutants even under pollutants stress. Four atrazine degradation intermediates were identified and possible metabolic routes were postulated. These findings offer critical insights for optimizing CW designs to mitigate non-point source agricultural pollution.