*3.2. Characteristics of DIC and δ13CDIC in the Pingzhai Reservoir*

Influenced by the dissolution of carbonate rocks, the water in the study area is weakly alkaline, under which condition HCO3 − is the main form of DIC [31]. In this study, HCO3 − is used to represent the concentration of DIC in the water. The variation in HCO3 − concentration in the Pingzhai Reservoir and inflow rivers ranged from 1.30 to 3.45 mmol/L, with an average of 2.32 mmol/L. The average concentrations of HCO3 − in river, reservoir, and outlet water were 2.52 mmol/L, 2.11 mmol/L, and 2.97 mmol/L, respectively. The seasonal variation was as follows: dry season (2.69 mmol/L) > normal season (2.44 mmol/L) > wet season (1.82 mmol/L). The average annual flows of the NYR, SGR, and ZWR were 1.35 × <sup>10</sup><sup>9</sup> <sup>m</sup>3·a<sup>−</sup>1, 5.89 × 108 <sup>m</sup>3·a<sup>−</sup>1, and 5.22 × 108 <sup>m</sup>3·a<sup>−</sup>1, respectively. Flow data from BSR and HJR are lacking. According to Equation (1), combined with the HCO3 − concentration at the monitoring section, it can be calculated that the HCO3 − fluxes of NYR, SGR, and ZWR were 2580.25 t·a−1, 700.62 t·a−1, and 722.52 t·a−1, respectively. The <sup>δ</sup>13CDIC value ranged from −3.1 to −18.4%, with an average value of −9.9%, and the average <sup>δ</sup>13CDIC values in river, reservoir, and outlet water were −12.0%, −8.3%, and −12.4%, respectively. In terms of time, the <sup>δ</sup>13CDIC value was negative in the wet season and positive in the dry season. There was no correlation between HCO3 <sup>−</sup> and δ13CDIC in the normal and dry periods except for a negative correlation between HCO3 <sup>−</sup> and δ13CDIC in the wet season (r = −0.692, *p* < 0.01). This is due to the strong photosynthesis during the wet season. Phytoplankton absorb the DIC in water and fractionate δ13CDIC, resulting in a decrease in HCO3 <sup>−</sup> concentration and an increase in the δ13CDIC value. Due to the influence of the dam, stable stratification is formed in the water body due to temperature differences in summer, which blocks the material exchange between the surface water body and the deep-water body. However, with the disappearance of stable stratification of water temperature in the normal and dry periods, the high concentration of HCO3 − at the bottom diffuses to the surface water with water turnover and internal circulation, resulting in a high concentration of HCO3 − in the surface water.

#### *3.3. Characteristics of NO3* <sup>−</sup>*, δ15N-NO3* <sup>−</sup>*, and δ18O-NO3* − *in the Pingzhai Reservoir*

The variation in the NO3 − concentration in the Pingzhai Reservoir ranged from 1.92 to 17.47 mg/L, with an average of 11.46 mg/L, and the time variation was wet season (12.68 mg/L) > dry season (12.09 mg/L) > normal season (11.02 mg/L). The average NO3 − concentrations in the river, reservoir area, and outlet were 11.33 mg/L, 10.68 mg/L, and 15.4 mg/L, respectively. According to the average annual flow and concentration of NO3 −, the NO3 <sup>−</sup> fluxes of the NYR, SGR, and ZWR points were calculated as 210.74 t·a<sup>−</sup>1, 25.33 t·a<sup>−</sup>1, and 54.76 t·a<sup>−</sup>1, respectively. The <sup>δ</sup>15N-NO3 − values in the dry season ranged from 13.0 to 17.1%, and the δ15N-NO3 − values in the wet and normal seasons ranged from 3.4 to 8.3% and 0.4 to 2.9%, respectively. This seasonal variation indicates different nitrate sources. The δ18O-NO3 − value was the highest in the wet season, intermediate in the normal season, and the lowest in the dry season.

#### **4. Discussion**

#### *4.1. Spatial Variation and Influencing Factors of Dissolved Inorganic Carbon and Nitrate*

Subject to different environmental factors, carbon and nitrogen elements and isotopes will show different spatial and temporal changes (Figure 3). Compared with the reservoir area, the δ13CDIC value of the river is more negative, and the DIC concentration is higher, which is due to the large proportion of farmland and woodland in the river flow area. The proportion of cultivated land area in the Baishui and Hujia River Basins is more than 50%, and the proportion of woodland area in the Shuigong River Basin is nearly half [24]. The δ13CDIC value of the water body inherits the δ13CDIC value of vegetation and soil CO2. The amount of soil CO2 produced by plant root respiration into the water body increased, which led to a negative δ13CDIC value and an increase in DIC concentration. In addition, the Narong and Zhangwei Rivers flow through cities and towns, and human activities (industrial production and agricultural cultivation) discharge sewage with negative δ13CDIC values into the river, which will also cause a negative δ13CDIC value in river water [32].

Compared with the reservoir area, the δ13CDIC value at the outlet of the reservoir was negative, and the DIC concentration was higher. The spatial variation in δ13CDIC values was negative in the river, positive in the reservoir area, and negative in the outlet. The DIC concentration first increased, then decreased, and then increased again. Yuan et al. (2021) studied DIC concentration and δ13CDIC in cascade reservoirs of the Yunnan section of the Lancang River and found that the DIC concentration was high in the river, low in the reservoir, and high at the outlet during the wet and dry seasons, and the δ13CDIC value first showed an increasing and then decreasing trend [4]. In the karst area of Southwest China, the DIC concentration of the Hongjiadu Reservoir in the Wujiang River Basin is higher at the outlet than in the river and reservoir area, and the δ13CDIC value is negative in the river, positive in the reservoir area, and negative at the outlet. The DIC concentration in the rainy season is lower than that in the dry season, and the δ13CDIC value in the rainy season is higher than that in the dry season [33]. All of these parameters showed the same change trend in the Pingzhai Reservoir basin, which was due to the following: Under the influence of the subtropical monsoon climate, the rainfall in the study area is mostly concentrated in spring and summer, the increase in rainfall leads to an increase in runoff into the reservoir, and the DIC concentration decreases due to the influence of water dilution. Generally, photosynthesis on the water surface is relatively active, and aquatic phytoplankton will produce approximately 18–20% isotope fractionation while absorbing CO2 through photosynthesis [34]. Therefore, the DIC concentration of surface water in the reservoir area is low, and the δ13CDIC value is positive, while the photosynthesis of the bottom water is weak, and the degradation of organic matter at the water–sediment interface produces CO2 with poor 13C, which increases the DIC concentration and causes the δ13CDIC value to be negative. As a result, the DIC value of the lower water is higher than that of the upper water, and the δ13CDIC value of the lower water is lower than that of the upper water. The discharge mode of the Pingzhai Reservoir is bottom discharge, so bottom water with high DIC concentrations and low δ13CDIC values is injected downstream. There are also relevant studies that show that the DIC concentration in the water body is lower in the rainy season than in the dry season, and the δ13CDIC value in the rainy season is also lower than that in the dry season, which may be related to the water nutrient level, reservoir operation mode, vegetation cover conditions, and tributary inflow [4,35,36].

**Figure 3.** Variations of δ13CDIC and DIC, δ15N-NO3 <sup>−</sup>, δ18O-NO3 −, and NO3 − along the course of water bodies in different seasons ((**a**) is the variation characteristic of δ13CDIC; (**b**) is the variation characteristic of DIC; (**c**) is the variation characteristic of δ15N, δ18O; (**d**) is the variation characteristic of NO3 −).

The concentration of NO3 − has a more direct impact on the deterioration of water quality. In the river section, the concentration of nitrate in the Nayong and Hujia Rivers was higher, which is a consequence of the discharge of domestic sewage and livestock breeding wastewater increasing the concentration of nitrate in the water body. The Shuigong River sampling site, which was not surrounded by residential and cultivated land, was minimally affected by anthropogenic activities and therefore had the lowest nitrate concentration. In the reservoir area, the water was fully mixed, and the nitrate concentration differed little among the sampling points in different seasons. Similar to the distribution of the NO3 <sup>−</sup> concentration, the δ15N-NO3 <sup>−</sup> and δ18O-NO3 − values also showed large differences in the river section and small differences in the reservoir section. It is mainly affected by crop cultivation and fertilization, domestic sewage discharge, and other factors. Fadhullah et al. (2020) used nitrate nitrogen and oxygen isotopes to study the source of nitrate in the Bukit Merah Reservoir in Southeast Asia and found that industrial and mining production and agricultural expansion in the upstream river disturbed the value of nitrate nitrogen and oxygen isotopes in the river water. The nitrogen and oxygen isotopes in the reservoir area indicated that the nitrate concentration is affected by atmospheric deposition [37]. Studies on nitrate isotopes in Chaohu Lake and its rivers showed that nitrification of soil organic nitrogen and soil erosion caused changes in nitrate isotopes [38].

#### *4.2. Sources of Dissolved Inorganic Carbon and Nitrate*

The DIC concentration and δ13CDIC value in water record and reflect the geochemical behavior and cycling characteristics of carbon. Because carbon is affected by many factors in the cycling process and different sources of carbon have different isotopic values, the δ13CDIC value can be used to trace the source of DIC. Related studies have shown that DIC in natural water mainly comes from atmospheric CO2 (including CO2 in atmospheric precipitation), dissolution of carbonate rock, and biogenic CO2 (plant root respiration and organic matter decomposition) dissolution in water [38,39]. The study area is located in the acid rain region of Southwest China, the pH value of rainwater is weakly acidic, the DIC concentration is low [40], and the partial pressure of water CO2 (*p*CO2) in the region is higher than that of atmospheric CO2 [23,35]. The influence of atmospheric precipitation and atmospheric CO2 on DIC and δ13CDIC values is not considerable. Therefore, the DIC in water is mainly derived from the weathering of carbonate rocks and the oxidative decomposition of organic matter.

The sources of nitrate in aquatic ecosystems mainly include atmospheric deposition, soil organic nitrogen, chemical fertilizer, domestic sewage, and livestock manure. Atmospheric deposition of δ15N-NO3 <sup>−</sup> ranges from −8% to +15%, and of <sup>δ</sup>18O-NO3 − ranges from +60% to +95%. The nitrogen and oxygen isotope values of nitrate fertilizer range from −5% to +5% and +17% to +25%, respectively. The nitrogen and oxygen isotopes of nitrate produced by fertilizer and deposition of NH4 +, soil organic nitrogen, sewage, and livestock manure range from −10% to +25% and −10% to +10%, respectively [41–44]. Combined with the δ15N-NO3 <sup>−</sup> and δ18O-NO3 − values of the water body in the study area, various nitrate sources could be determined (Figure 4). During the normal season, δ15N-NO3 − and δ18O-NO3 − were distributed in soil organic nitrogen and ammonium fertilizer endmembers, indicating that nitrification of soil organic nitrogen and ammonium fertilizer was the main source of nitrate in water. In the dry season, δ15N-NO3 <sup>−</sup> and δ18O-NO3 − are distributed in domestic sewage endmembers and livestock manure, and δ15N-NO3 − is higher. Ren et al. (2021) found that the δ15N value of some points was higher in a study of groundwater in the Zhaotong Basin, Yunnan Province [8]. Lin et al. (2019) also found a similar phenomenon in a study of the Illinois River in Chicago, USA [45]. This is because the volatilization of NH4 <sup>+</sup> in sewage releases a large amount of lighter isotopes, making the remaining NO3 <sup>−</sup> enriched in heavy δ15N. The δ15N-NO3 <sup>−</sup> and δ18O-NO3 − were distributed in the nitrate fertilizer endmembers during the wet season. It may be that inorganic nitrogen from the oxidation and decomposition of nitrate fertilizer enters the water with surface runoff. In general, agricultural activities (ammonium fertilizer, nitrate fertilizer, soil organic nitrogen, livestock manure, and sewage) were important sources of nitrate in the Pingzhai Reservoir. Chemical fertilizers (ammonium fertilizer, nitrate fertilizer) can improve soil fertility and increase crop yield. From 1997 to 2005, the amount of chemical fertilizer applied in China increased from 7.07 million tons to 26.21 million tons [46]. However, the irrational use of chemical fertilizers and the lack of farmland management often lead to nitrogen loss and nitrate pollution. Due to the periodic storage and discharge of the reservoir, a water-level fluctuation zone will form around the reservoir, and the water-level fluctuation zone is also the area where agricultural nonpoint source pollution and soil erosion often occur [47]. According to Han et al. (2016), the amount of soil erosion in the Yangtze River basin and Wujiang River basin reached 1.4 × <sup>10</sup><sup>10</sup> t/a and 1.2 × <sup>10</sup><sup>9</sup> t/a in 2014, respectively [48]. Soil erosion in the water-level fluctuation zone will transport a large amount of nutrients and soil organic nitrogen to rivers and reservoirs. This part of the lost N will be converted into nitric acid through nitrification and will participate in the geochemical cycle, which will affect the chemical composition of rivers and reservoirs, especially the concentrations of Ca2+ and HCO3 − [16]. Nitric acid participates in the dissolution of carbonate rocks and directly or indirectly releases CO2 into the atmosphere, and these anthropogenic carbon source emissions offset part of the natural carbon sink in the natural process [20]. Therefore, it is necessary to consider the impact of external acid input on the weathering of carbonate rocks in the basin.

**Figure 4.** Distribution characteristics of δ15N-NO3 <sup>−</sup> and δ18O-NO3 − in water. The isotopic values of various nitrate sources are based on Kendall et al. (2007) [44].

## *4.3. Weathering of Carbonate Rocks and C–N Coupling Relationship in the Pingzhai Reservoir Basin*

The average <sup>δ</sup>13CDIC of CO2 produced by root respiration of C3 vegetation is −27% because the migration and diffusion of CO2 in the soil layer will produce approximately 4% isotope fractionation, so the δ13CDIC value of soil CO2 produced by plant respiration and organic matter oxidation decomposition is approximately −23%, and the <sup>δ</sup>13CDIC value of marine carbonate rocks in karst areas is 0% [49–51]. According to the stoichiometric relationships of carbonate rock dissolution, with the participation of carbonic acid (H2CO3), 2 mol HCO3 − will be produced in the process of carbonate rock dissolution, of which 1 mol is derived from atmospheric/soil CO2 in the watershed and 1 mol from carbonate rocks (Equation (2)) [52].

$$\text{Ca}\_{\text{(1-x)}}\text{Mg}\_{\text{x}}\text{CO}\_{3} + \text{CO}\_{2} + \text{H}\_{2}\text{O} = (1-\text{x})\text{Ca}^{2+} + \text{xMg}^{2+} + 2\text{HCO}\_{3}^{-}\tag{2}$$

In the process of carbonate rock dissolution by carbonic acid, carbon isotopes will produce +9% fractionation, so the water body δ13CDIC theoretical value is approximately −14% [53]. The variation in <sup>δ</sup>13CDIC in the Pingzhai Reservoir water ranges from −3.13% to −18.42%, with an average of −9.92%. Compared with the theoretical value, the <sup>δ</sup>13CDIC actual value of the water body is more positive, which indicates that in addition to carbonic acid, there are other sources of acid, such as nitric acid from human activities, which contribute to the dissolution of carbonate rocks in the basin (Equation (3)). In addition, the [Ca2+ + Mg2+]/[HCO3 −] equivalent ratios in the water samples in the study area were all greater than 1 (1.37 on average), and the results indicate that Ca2+ and Mg2+ in the water sample are surplus relative to HCO3 <sup>−</sup>, that is, Ca2+ and Mg2+ have additional sources [54]. In addition to the carbonate rocks dissolved by H2CO3 and HNO3, the weathering and dissolution of silicate rocks in the watershed can also produce Ca2+, Mg2+, and HCO3 − ions (Equations (4) and (5)).

$$\rm Ca\_{(1-x)}Mg\_xCO\_3 + HNO\_3 = (1-x)Ca^{2+} + xMg^{2+} + NO\_3^- + HCO\_3^- \tag{3}$$

$$\mathrm{CaMg\_{(1-x)}Al\_2Si\_2O\_8} + 2\mathrm{H\_2CO\_3} + 2\mathrm{H\_2O} = \mathrm{xCa^{2+}} + (1-\mathrm{x})\mathrm{Mg^{2+}} + 2\mathrm{HCO\_3^-} + 2\mathrm{SiO\_2} + 2\mathrm{Al(OH)\_3} \tag{4}$$

$$\text{Na}\_{\text{k}}\text{K}\_{(1-\text{x})}\text{Al}\_{2}\text{Si}\_{2}\text{O}\_{8} + \text{H}\_{2}\text{CO}\_{3} + \text{H}\_{2}\text{O} = \text{xNa}^{+} + (1-\text{x})\text{K}^{+} + \text{HCO}\_{3}^{-} + 3\text{SiO}\_{2} + \text{Al}(\text{OH})\_{3} \tag{5}$$

According to the dissolution equation of silicate rock, the Ca2+, Mg2+, and HCO3 − produced by the dissolution of silicate rock by carbonic acid are calculated according to onefifth and one-half of the molar concentration of SiO4, respectively [11]. The contributions of silicate weathering to Ca2+, Mg2+, and HCO3 − in the Pingzhai Reservoir basin ranged from 0.97% to 2.79% and 2.43% to 6.98%, with average values of 1.23% and 3.07%, respectively, accounting for a small proportion. This indicates that the weathering of carbonate rocks is the main process controlling the material composition and geochemical cycle of water bodies in the basin.

To show that nitric acid is indeed involved in the weathering of carbonate rocks in the basin, the ratio relationship between [Ca2+ + Mg2+]/[HCO3 <sup>−</sup>] and δ13CDIC was established based on the ion and carbon isotope data in water samples. According to Figure 5, most of the sampled data fall between the endmembers of H2CO3 and HNO3 dissolved carbonate rocks. This shows that H2CO3 and HNO3 are jointly involved in the dissolution of carbonate rocks in the basin and have an impact on the value of carbon and nitrogen elements and ion concentration in the water body [51].

**Figure 5.** The relationship between [Ca2+ + Mg2+]/[HCO3 <sup>−</sup>] and δ13CDIC in different seasons.

In addition to carbonate rock dissolution by carbonic acid, nitric acid produced by nitrification in agricultural activities also participates in carbonate rock dissolution. Assuming that carbonic acid and nitric acid participate in the dissolution process with equal molar ratios, the C–N coupling formula of carbonate rock dissolution can be obtained from Equation (6):

$$\text{Ca}(\text{a} + \text{b})\text{Ca}\_{\text{x}}\text{Mg}\_{\text{(1}-\text{x)}}\text{CO}\_{3} + \text{aH}\_{2}\text{CO}\_{3} + \text{bHNO}\_{3} = (\text{a} + \text{b})\text{aCa}^{2+} + (\text{a} + \text{b})(\text{1} - \text{x})\text{Mg}^{2+} + \text{bNO}\_{3}^{-} + (\text{2a} + \text{b})\text{HCO}\_{3}^{-} \tag{6}$$

where a and b represent the coefficients of H2CO3 and HNO3 involved in the dissolution of carbonate rocks, respectively.

According to Equation (6), the molar ratio of (Ca2+ + Mg2+)/HCO3 − in water should be 2/3 (0.67), and the molar ratio of (Ca2+ + Mg2+)/HCO3 − in the water sample of the Pingzhai Reservoir basin was 0.17–0.98, with an average of 0.60. This indicates that the dissolution of carbonate rocks in the study area is controlled by C–N coupling, but H2CO3 and HNO3 dissolve carbonate rocks according to the molar ratio of 1:1. Therefore, the amount of (Ca2+ + Mg2+) and HCO3 − released by HNO3 dissolved carbonate rocks can be calculated according to the C–N coupling equation. The calculation results show that the proportion of HCO3 − produced by HNO3 dissolving carbonate rocks ranges from 1.50% to 13.35%, with an average of 8.38% (Table 2). The proportion of (Ca2+ + Mg2+) ranges from 4.04% to 12.06%, with an average of 7.84%. Among the different periods, the proportion of HCO3 <sup>−</sup> and (Ca2+ + Mg2+) produced by HNO3 dissolution during the wet season is the highest, with average values of 10.32% and 9.90%, respectively. The proportion in the dry season is the lowest, with averages of 7.29% and 5.21%, respectively. The average values in the normal season are 7.54% and 8.42%, respectively. This is because dam construction slows down or even stops the water flow, and the abundant water volume and appropriate temperature in the wet season make the water-rock interaction more complete. In addition, the amount of NO3 − fertilizer from agricultural activities increased during the wet season, which combined with H<sup>+</sup> to generate HNO3 to participate in the dissolution of carbonate rocks.

**Table 2.** The ratio of HCO3 <sup>−</sup> and (Ca2+ + Mg2+) produced by HNO3 dissolution of carbonate rocks and the measured and theoretical values of δ13CDIC.


Note: mean ± standard deviation (SD).

The δ13CDIC produced by carbonate rocks dissolved by carbonic acid is approximately −14%, and the HCO3 − produced by carbonic acid-dissolved silicate rocks is all from soil CO2, so the <sup>δ</sup>13CDIC is approximately −23%. Carbonate rocks dissolved by nitric acid do not consume soil or atmospheric CO2, and all the HCO3 − produced comes from carbonate rocks, so the <sup>δ</sup>13CDIC is approximately −0% [23,40,55]. The theoretical value of <sup>δ</sup>13CDIC in the Pingzhai Reservoir basin can be estimated by Equation (7):

$$
\delta^{13}\mathcal{C}\_{DIC-T} = f\_{\rm cc}\delta^{13}\mathcal{C}\_{\rm cc} + f\_{\rm cs}\delta^{13}\mathcal{C}\_{\rm cs} + f\_{\rm nc}\delta^{13}\mathcal{C}\_{\rm nc} \tag{7}
$$

where *δ13CDIC-T* represents the theoretical value of δ13CDIC in water; *fcc*, *fcs*, and *fnc* represent the contribution proportions of carbonate rock dissolved by carbonic acid, silicate rock dissolved by carbonic acid, and carbonate rock dissolved by nitric acid to HCO3 − in water, respectively; *δ13Ccc*, *δ13Ccs*, and *δ13Cnc* represent the values of *δ13CDIC* generated by carbonate rock dissolved by carbonic acid, silicate rock dissolved by carbonic acid, and carbonate rock dissolved by nitric acid, respectively. The calculated results show that *<sup>δ</sup>13CDIC-T* ranges from −14.06% to −12.72%, with an average value of −13.10%. The measured *<sup>δ</sup>13CDIC* ranges from −3.13% to −18.42%, with an average value of −9.32%. The measured *δ13CDIC* values in river and reservoir areas are more positive than the theoretical values in different periods, and the reservoir area is the most positive during the wet season. This is because the water flow in the reservoir area is slow and receives more light, and the photosynthesis of aquatic organisms absorbs DIC and causes isotope fractionation, making the water body δ13CDIC value positive [56]. Compared with the river, the proportion of HCO3 <sup>−</sup> and (Ca2+ + Mg2+) produced by the dissolution of carbonate rocks by nitric acid in the reservoir water is larger, which reflects that dam construction promotes the water-rock interaction and the retention of ionic substances.

As mentioned above, carbonate rock dissolved by carbonic acid in karst systems consumes atmospheric/soil CO2 to form HCO3 −, one part of which is used by aquatic organisms [26,57], and the other part enters the ocean with rivers for sedimentation. In addition to carbonic acid, nitric acid is also involved in the dissolution of carbonate rocks in the Pingzhai Reservoir. Nitrogen from chemical fertilizers (ammonium fertilizer and nitrate fertilizer), soil organic nitrogen, sewage, and livestock manure is lost and converted to HNO3 (Figure 6). As a result, the concentration of HNO3 in the water increases, and the carbonate rocks are dissolved by both carbonic and nitric acids. In this study, the average proportions of HCO3 <sup>−</sup> and (Ca2+ + Mg2+) produced by HNO3-dissolved carbonate rocks were 8.38% and 7.84%, respectively. However, this part of HCO3 − does not come from atmospheric/soil CO2 but rather from carbonate rocks and does not constitute a carbon sink. Therefore, the environmental effect of C–N coupling not only causes water nitrate pollution but also reduces carbon sinks.

**Figure 6.** The C-N coupling cycle in the karst zone (adapted from Hu et al., 2017 [58]).

#### **5. Conclusions**

By monitoring the hydrochemistry and δ13CDIC, δ15N-NO3 <sup>−</sup>, and δ18O-NO3 − in the Pingzhai Reservoir and its inflow rivers, we analyzed the temporal and spatial variation in water chemistry and carbon and nitrogen isotopes, explored the source of carbon and nitrogen elements in water, and analyzed the C–N coupling in water. The results show that the dominant cation in the water of the Pingzhai Reservoir is Ca2+, which accounts for 68% of the total cations. The dominant anion is HCO3 −, accounting for 65% of the total anions. The hydrochemical type was HCO3-Ca. Dissolved inorganic carbon, nitrate, and their isotopes have different spatial and temporal variations. From the river to the reservoir area and then to the outlet, the concentration of HCO3 − increased first, then decreased, and then increased again, while the δ13CDIC value was negative first, then positive, and then negative again. The values of NO3 <sup>−</sup>, δ15N-NO3 <sup>−</sup>, and δ18O-NO3 − were different in each stage of the river, which were mainly affected by dam construction and water storage, surrounding land-use mode, crop cultivation and fertilization, domestic sewage discharge, and other factors. According to the characteristics of carbon and nitrogen isotopes, the HCO3 − in the water of the study area is mainly derived from the weathering of carbonate rocks and the oxidative decomposition of organic matter. Nitrate mainly comes from agricultural activities, including chemical fertilizer (ammonium fertilizer and nitrate fertilizer), soil organic nitrogen, sewage, and livestock manure in the normal season, dry season, and wet season. The input of nitrate caused the C–N coupling cycle of hydrogeochemistry in the Pingzhai Reservoir basin and disturbed the water-rock interaction. The average proportions of HCO3 <sup>−</sup> and (Ca2+ + Mg2+) produced by HNO3 dissolved carbonate rocks were 8.38% and 7.84%, respectively, but this part does not constitute a carbon sink. The proportion of HCO3 <sup>−</sup> and (Ca2+ + Mg2+) produced by the dissolution of carbonate rocks by HNO3 in

reservoir water was relatively large. This reflects the full water-rock interaction and retention effect due to the construction of the dam. Therefore, successful fertilizer application experience and farmland management practice should be learned, controlling nitrogen input from agricultural activities. Using new technologies can increase the capacity of domestic wastewater treatment and limit the discharge of sewage into rivers and reservoirs and prevents soil organic nitrogen loss. In addition, a limitation of the study is that it is lacking an analysis of the impact of geological conditions on the geochemical characteristics of carbon and nitrogen elements. This will be addressed in subsequent studies.

**Author Contributions:** Z.Z.: Resources, Data curation, Formal analysis, Project administration, Funding acquisition. J.K.: Conceptualization, Methodology, Software, Writing—original draft. F.Z.: Visualization. Y.Z.: Supervision. J.X.: Investigation. C.W.: Validation. All authors have read and agreed to the published version of the manuscript.

**Funding:** National Natural Science Foundation of China (42161048); National Natural Science Foundation of China (41661088); Science and Technology Plan Project of Guizhou Province (Qiankehe Jichu [2020]1Y154) financially supported this study.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** This study was supported financially by the State's Key Project of Research.

**Conflicts of Interest:** The authors declare no conflict of interest.
