Using dual isotopes and a Bayesian isotope mixing model to evaluate sources of nitrate of Tai Lake, China

Identification and quantification of sources of nitrate (NO3–) in freshwater lakes provide useful information for management of eutrophication and improving water quality in lakes. Dual δ15N- and δ18O-NO3– isotopes and a Bayesian isotope mixing model were applied to identify sources of NO3– and estimate their proportional contributions to concentrations of NO3– in Tai Lake, China. In waters of Tai Lake, values for δ15N-NO3– ranged from 3.8 to 10.1‰, while values of δ18O ranged from 2.2 to 12.0‰. These results indicated that NO3– was derived primarily from agricultural and industrial sources. Stable isotope analysis in R called SIAR model was used to estimate proportional contributions from four potential NO3– sources (agricultural, industrial effluents, domestic sewage, and rainwater). SIAR output revealed that agricultural runoff provided the greatest proportion (50.8%) of NO3– to the lake, followed by industrial effluents (33.9%), rainwater (8.4%), and domestic sewage (6.8%). Contributions of those primary sources of NO3– to sub-regions of Tai Lake varied significantly (p < 0.05). For the northern region of the lake, industrial source (35.4%) contributed the greatest proportion of NO3–, followed by agricultural runoff (27.4%), domestic sewage (21.3%), and rainwater (15.9%). Whereas for the southern region, the proportion of NO3– contributed from agriculture (38.6%) was slightly greater than that contributed by industry (30.8%), which was similar to results for nearby inflow tributaries. Thus, to improve water quality by addressing eutrophication and reduce primary production of phytoplankton, NO3– from both nonpoint agricultural sources and industrial point sources should be mitigated.