Identification of sources and transformations of nitrate in the Xijiang River using nitrate isotopes and Bayesian...

Coupled nitrogen and oxygen isotopes of nitrate have proven useful in identifying nitrate sources and transformation in rivers. However, isotopic fractionation and low-resolution monitoring limit the accurate estimation of nitrate dynamics. In the present study, the spatio-temporal variations of nitrate isotopes (¹⁵N and ¹⁸O) and hydrochemical compositions (NO₃^- and Cl^-) of river water were examined to understand nitrate sources in the Xijiang River, China. High-frequency sampling campaigns and isotopic analysis were performed at the mouth of the Xijiang River to capture temporal nitrate variabilities. The overall values of δ¹⁵N-NO₃^- and δ¹⁸O-NO₃^- ranged from +4.4‰ to +14.1‰ and from -0.3‰ to +6.8‰, respectively. The results of nitrate isotopes indicated that NO₃^- mainly originated from soil organic nitrogen (SON), chemical fertilizer (CF), and manure and sewage wastes (M&S). The negative correlation of nitrate isotopic values with NO₃^-/Cl^- ratios suggested the importance of denitrification in NO₃^- loss. The results of Bayesian model with incorporation of isotopic fractionation during the denitrification showed that SON and CF contributed to the most (72-73%) nitrate in the wet season; whereas approximately 58% of nitrate was derived from anthropogenic inputs (M&S and CF) in the dry season. The nitrate flux was 2.08 × 10⁵ tons N yr^-1 during one hydrologic year between 2013 and 2014, with 86% occurring in the wet season. Long-term fluctuations in nitrate flux indicated that nitrate export increased significantly over the past 35 years, and was significantly correlated with nitrate concentrations. The seasonal pattern of nitrate dynamics indicated the mixing of nitrified NO₃^- and denitrified NO₃^- between surface flow and groundwater flow under different hydrological conditions. Overall, the present study quantitatively evaluates the spatio-temporal variations in nitrate sources in a subtropical watershed, and the high-frequency monitoring gives a better estimate of nitrate exports and proportional contributions of nitrate sources.