Defining the sources and the fate of nitrate by using dual isotopes and a Bayesian isotope mixing model: Water-nitrate management in cascade dams of Lancang river.

A multi-isotopes approach involving the use of stable nitrate isotopes (δN-NO and δO-NO) combined with stable water isotopes (δD-HO and δO-HO) and SIAR model as tracers can help identify the nitrogen source and understand the transformation process in a river-cascade reservoirs system. In this study, we identify the potential impact of the N source in the Lancang River basin, clarified the seasonal variations in the isotope values and estimated the probability distribution and proportional contribution of multi-terminal NO-N sources using Bayesian isotope mixing model. In addition, we investigate the factors that led to the seasonal variations of the stable isotopes and evaluated the relationship between the uncertainty of the contribution ratio of the N sources and isotopic variations in the river-cascade reservoirs system. The NO-N is the main component of DIN (dissolved inorganic nitrogen), accounting for 68.1 % of DIN. The ratios of δN-NO and δO-NO ranged from +4.2 ‰ to +10.3 ‰, and from +5.9 ‰ to +9.3 ‰ in the Lancang River. The δD-HO and δO-HO of the surface waters ranged from -109.47 ‰ to -76.44 ‰, and from -15.13 ‰ to -11.61 ‰, respectively. The SIAR model analysis results show that nitrification of livestock and poultry manure is the main source of NO-N in the upstream natural reach, accounting for 40.2 %. There is little difference between the wet season and the dry season. Nitrification of soil organic nitrogen is the main source of NO-N in the cascade development reach, accounting for 42.3 %. The contribution rate of atmospheric precipitation to nitrate concentration in both sampling periods is low (<5 %). This study provides a useful insight for reservoir water environmental managers to verify cascade development river pollution contributors and to better apply remedial solutions.