Figure 3 Composition of hydrogen and oxygen isotopes in surface water and groundwater in the Huashan watershed

3.3 Qualitative Analysis of Nitrate Sources

A high NO3-/Cl- molar ratio with low Cl- in groundwater is supposed to be affected by intensive agricultural activities (Liu et al. 2006), while low NO3-/Cl- molar ratio with high Cl- in groundwater is typical considered to be influenced by manure and municipal sewage (Yue et al. 2017). The NO3-/Cl- molar ratio versus Cl- concentration (μmol/L) dual-logarithmic plot is widely utilized in numerous studies to identify the origin of groundwater NO3- concentration (Torres Martínez et al., 2021; Li et al., 2022). As shown in Figure 4, soil organic nitrogen is an important source of nitrate in the study area. Groundwater in April 2022 was influenced by sewage. There was some overlap between agricultural inputs and soil organic nitrogen, indicating that NO3- in the study area was also influenced by agricultural inputs. This conclusion is consistent with the results shown in Figure 5.
Figure 4 Relationships between NO3-/Cl-molar ratio and Cl- molar concentration in Huashan watershed
Dual isotope approach (δ15N-NO3- and δ18O-NO3-) was generally applied to identify nitrate source by comparing the isotopic signals between the water samples and the possible nitrogen source (Nestler et al. 2011). As shown in fig.5, the observed isotopic signal of surface water samples in November 2021 were mainly located in the expected ranges of M&S and SN, and the isotopic values of groundwater mainly fall in the NF and SN intervals, indicating M&S and SN sources of nitrate for surface water and the NF and SN sources of nitrate for groundwater in winter. The δ15N-NO3-and δ18O-NO3- values of surface water in April 2022 (spring) mainly fell in the NF and SN intervals, while groundwater fell mainly in M&S, SN and NF. Therefore, nitrate in spring surface water comes mainly from NF and SN, while groundwater comes from M&S, SN and NF, which is consistent with the analysis results of water chemistry methods. The seasonal difference in nitrate sources between November 2021 (winter) and April 2022 (spring) mainly due to the application of chemical fertilizers (ammonia-nitrogen fertilizers) during planting period in spring. Half a month before sampling, the nitrogen fertilizers were washed into the river by a heavy rainfall (107 mm) which led to the NF source of nitrate in the river in spring. Furthermore, some samples fall outside the endmember mixing frame, as shown in the bottom-left corner of the figure 5, which may be due to isotope fractionation. It is known that fractionation increases with decreasing temperature (Mariotti et al., 1981; Yun et al., 2011). During the sampling period, the temperature was relatively low with an average of 3℃, which could have led to the preferential utilization of14N and 16O during the conversion of Ammonium fertilizer to nitrate, resulting in lower δ15N and δ18O values in the nitrate infiltrating into groundwater, consistent with the findings of Yu et al. (2020).
Figure 5 δ15N-NO3- versus δ18O-NO3- diagram. Typical ranges of δ15N-NO3- and δ18O-NO3- values for different nitrate sources were taken from Xue et al. (2009).
The isotopic signals and the contents of nitrate were general applied to identify denitrification processes in water as the ratios of fractionation factors (Δδ15N/Δδ18O) usually vary from 1.3 to 2.1 (Xue et al., 2012). As shown in Figure 6b, δ15N-NO3- is positively correlated with δ18O-NO3- with a slope of 0.67 for groundwater samples in winter, confirmed the existence of denitrification. No significant correlation was observed between δ15N-NO3- and Ln[N-NO3-] for groundwater samples in winter may be because the mixed source of groundwater and its dilution effect on NO3- (Fig 6a). No relation was observed between δ15N-NO3- and δ18O-NO3- for groundwater samples in spring and surface water samples in winter and in spring, suggesting no denitrification existed. The DO content of surface water samples and part of groundwater samples in spring generally higher than 6 mg/L also suggested the same viewpoint, as denitrification generally occurred in an anaerobic environment with DO content lower than 4 mg/L (Shang et al., 2020). Water samples with lower dissolved oxygen concentrations were predominantly observed during winter and located in proximity to the outlet of the watershed for groundwater. High nitrogen isotopes and low DO concentrations are typically observed at the watershed outlet, thus denitrification in groundwater is mainly observed in the outlet area during winter (Li et al. 2019). The relationship between DO and pH, as shown in Figure 6c, reflects the same conclusion. In addition, related studies have shown that 2/3 of the oxygen atoms bound during nitrification come from H2O and 1/3 from O2, and the nitrate δ18O-NO3- produced by nitrification ranges from -10‰ to 10‰ (Kendall et al, 2007). The δ15N-NO3- of surface waters in the Huashan watershed lie in this interval range, indicating that denitrification is non-existent in surface waters (Fig 6b). The water samples of groundwater in both phases, 30% of which had δ15N-NO3- beyond this range, and these points were mainly distributed downstream of the watershed, indicating the presence of denitrification in groundwater downstream of the watershed. Overall, denitrification was not detected in the winter and spring surface water or in the spring groundwater of the Huashan watershed. Weak denitrification was found in the winter groundwater, mainly located near the outlet of the watershed.