Riverine nitrate source apportionment using dual stable isotopes in a drinking water source watershed of southeast China.

It is crucial to quantitatively track riverine nitrate (NO3-) sources and transformations in drinking water source watersheds for preventing current and future NO3- pollution, and ensuring a safe drinking water supply. This study identified the significant contributors to riverine NO3- in Zhaoshandu reservoir watershed of Zhejiang province, southeast China. To achieve this goal, we used hydrochemistry parameters and stable isotopes of NO3- (δ15N-NO3- and δ18O-NO3-) accompanied with a Markov Chain Monte Carlo mixing model to estimate the proportional contributions of riverine NO3- inputs from atmospheric deposition (AD), chemical nitrogen fertilizer (NF), soil nitrogen (SN), and manure and sewage (M&S). Results indicated that the main form of riverine nitrogen in this region was NO3-, constituting ~60% of the total nitrogen mass on average (total organic nitrogen ~37% & ammonium ~3%). Variations in the isotopic signatures of NO3- demonstrated that microbial nitrification of NF, SN and M&S was the primary nitrogen transformation process within the Zhaoshandu reservoir watershed, whereas denitrification was minimal. A classical dual isotope bi-plot incorporating chloride concentrations suggested NF, SN and M&S were the major contributors of NO3- to the river. Riverine NO3- source apportionment results were further refined using the Markov Chain Monte Carlo mixing model, which revealed that AD, NF, SN and M&S contributed 7.6 ± 4.1%, 22.5 ± 12.8%, 27.4 ± 14.5% and 42.5 ± 11.3% of riverine NO3- at the watershed outlet, respectively. Finally, uncertainties associated with NO3- source apportionment were quantitatively characterized as: SN > NF > M&S > AD. This work provides a comprehensive approach to distinguish riverine NO3- sources in drinking water source watersheds, which helps guide implementation of management strategies to effectively control NO3- contamination and protect drinking water quality. SUMMARY OF THE MAIN FINDING FROM THIS WORKS (CAPSULE): We utilized NO3- stable isotope analysis and a Markov Chain Monte Carlo mixing model to quantify riverine nitrate pollution sources in a drinking water source watershed in Zhejiang province, southeast China. Markov Chain Monte Carlo mixing model output showed that NF, SN and M&S were the dominant sources of riverine NO3- during the sampling period in Zhaoshandu watershed. Uncertainty analysis characterized the variation strength associated with contributions of individual nitrate sources and indicated the greatest uncertainty for SN, followed by NF, M&S and AD