The quantitative identification of nitrate sources is of great significance for the control of non-point source pollution and the comprehensive management of water resources in watersheds. δ ¹⁵N-NO ₃ ^- and δ ¹⁸O-NO ₃ ^- isotopes combined with the Bayesian isotope mixing model were widely used as effective methods to identify nitrogen sources. In this study, a total of 60 surface water samples and 82 groundwater samples were collected in study area from November 2021 to October 2022, and atmospheric deposition (AD), chemical nitrogen fertilizer (NF), soil nitrogen (SN), and manure and sewage (M&S) were determined as the potential nitrate sources. Source identification by SIAR indicated that in November 2021 the M&S was the main contributor of nitrate to surface water (mean 38.1%), while NF was the main contributor to groundwater (mean 39.8%). In April 2022, NF contributed the most to surface water (38.3%), while groundwater mainly originated from SN (29.4%) and MS (29.8%). The uncertainty analysis showed that the greatest uncertainties were in SN and NF, followed by M&S and AD. Sensitivity analysis showed that the changes in the nitrate isotopic composition of M&S had the greatest effect on the results for δ ¹⁵N, whereas only the mean values of oxygen isotope values of AD had a greater effect on the results for δ ¹⁸O. The sensitivity analysis results can optimize the sampling scheme and improve the accuracy of the model predictions. Additionally, the contributions of soil nitrogen and nitrogen fertilizer to nitrate in surface water and groundwater reached 58% and 64%, respectively. Therefore, optimizing fertilizer and irrigation management is necessary to improve nitrogen use efficiency in watershed management.