Quantifying multi-interface runoff-driven nitrogen loss and its control effects in karst hillslopes

Abstract

Runoff-driven rapid nitrogen loss poses a risk for groundwater degradation and non-point source pollution in aquatic ecosystems. However, the pathways and controlling factors of runoff nitrogen loss remain poorly understood, particularly in complex hillslope dominated by lateral subsurface runoff. In this study, nitrogen concentrations (total dissolved nitrogen-TDN, NO₃^--N, NH₄⁺-N) and total loads driven by multi-interface runoff (surface, subsurface, and epikarst) were quantified during 49 natural rainfall-runoff events across the karst hillslopes. Furthermore, the effects of land use types (cropland, forage grassland, and shrubland) and soil thickness on nitrogen loss were distinguished. Consistent with expectations, surface runoff exhibited the highest nitrogen concentration; however, due to its low runoff coefficient (∼15 %), the TDN loss flux in surface runoff (∼2.47 kg ha⁻¹ yr⁻¹) was significantly lower than those of lateral subsurface (∼6.56 kg ha⁻¹ yr⁻¹) and epikarst (∼6.51 kg ha⁻¹ yr⁻¹) runoff. Nitrogen concentration in subsurface and epikarst runoff increased with rainfall intensity (9.9 ∼ 100 mm d⁻¹); however, TDN and NO₃^--N concentrations (0.63 ± 1.48 and 0.39 ± 0.98 mg L⁻¹) were observed in rainfall events exceeding 100 mm d⁻¹, indicating that a diluted effect occurs during rare extreme rainfall events. Relationships between event runoff yield and nitrogen loss fluxes showed positive trends for TDN, NO₃^--N and NH₄⁺-N, suggesting an overall transport limitation. Event runoff yield emerged as the primary driver of nitrogen export variability across land uses. Additionally, soil thickness regulated nitrogen loss pathways, with shallow-soil cover reduce TDN and NO₃^--N loss in surface runoff by 2.8–3.4 times compared to deep-soil cover, attributed to its higher permeability. Accordingly, our results highlight that lateral subsurface flow drives nitrogen loss and is influenced by multiple environmental factors in karst hillslopes.