AbstractNitrate (NO3−) pollution is a serious global problem, and the quantitative analysis of its sources contributions is essential for devising effective water-related environmental-protection policies. The Shengjin Lake basin, located in the middle to lower reaches of the Yangtze River in China was selected as the research area in our study. We first grouped 29 surface water samples and 33 groundwater samples using cluster analysis, and then analyzed potential nitrate sources for each dataset of δ15N–NO3− and δ18O–NO3− isotope values by applying a Bayesian isotope-mixing model. Our results show that the nitrogen pollution in the surface-ground water in the study area seriously exceeded to class V of the Environmental Quality Standard of Surface Water of China. The NO3− in surface water from the mid-upper reaches of the drainage basin mainly originates from soil nitrogen (SN) and chemical fertilizer (CF), with contribution rates of 48% and 32%, respectively, and the NO3− in downstream areas mainly originates from CF and manure and sewage (MS), with contribution rates of 48% and 33%, respectively. For the groundwater samples, NO3− mainly originates from MS, CF, and SN in the mid-upper reaches of the drainage basin and the northside of Dadukou near the Yangtze River, with contribution rates of 34%, 31%, and 29%, respectively, whereas NO3− in the lower reaches and the middle part of Dadukou mainly originates from MS, with a contribution rate of 83%. The nitrogen conversion of surface water in lakes and in the mid-upper reaches is mainly affected by water mixing, while the groundwater and surface water in the lower plains are mainly affected by denitrification. The method proposed in this study can expand the ideas for tracking nitrate pollution in areas with complex terrain, and the relevant conclusions can provide a theoretical basis for surface and groundwater pollution control in the hilly basin of Yangtze River.
Spatial variability of nitrate pollution and its sources in a hilly basin of the Yangtze River based on clustering