Nitrate source apportionment using a combined dual isotope, chemical and bacterial property, and Bayesian model approach in river systems

Abstract

AbstractNitrate (NO3−) pollution is a serious problem worldwide, particularly in countries with intensive agricultural and population activities. Previous studies have used δ15N‐NO3− and δ18O‐NO3− to determine the NO3− sources in rivers. However, this approach is subject to substantial uncertainties and limitations because of the numerous NO3− sources, the wide isotopic ranges, and the existing isotopic fractionations. In this study, we outline a combined procedure for improving the determination of NO3− sources in a paddy agriculture‐urban gradient watershed in eastern China. First, the main sources of NO3− in the Qinhuai River were examined by the dual‐isotope biplot approach, in which we narrowed the isotope ranges using site‐specific isotopic results. Next, the bacterial groups and chemical properties of the river water were analyzed to verify these sources. Finally, we introduced a Bayesian model to apportion the spatiotemporal variations of the NO3− sources. Denitrification was first incorporated into the Bayesian model because denitrification plays an important role in the nitrogen pathway. The results showed that fertilizer contributed large amounts of NO3− to the surface water in traditional agricultural regions, whereas manure effluents were the dominant NO3− source in intensified agricultural regions, especially during the wet seasons. Sewage effluents were important in all three land uses and exhibited great differences between the dry season and the wet season. This combined analysis quantitatively delineates the proportion of NO3− sources from paddy agriculture to urban river water for both dry and wet seasons and incorporates isotopic fractionation and uncertainties in the source compositions.