Using δ 15 N and δ 18 O Signatures to Evaluate Nitrate Sources and Transformations in Four Inflowing Rivers , North of Taihu Lake

Taihu Lake is the third largest freshwater lake in China. Due to rapid economic development and excessive nutrient discharges, there is serious eutrophication in the northern part of the lake. Nitrogen (N) is one of the key factors for eutrophication in Taihu Lake, which mainly comes from the rivers around the lake. Samples from four inflowing rivers were analysed for δ15N and δ18O isotopes in December 2013 to identify the different sources of nitrogen in the northern part of Taihu Lake. The results indicated that the water quality in Taihu Lake was clearly influenced by the water quality of the inflowing rivers and nitrate (NO3-N) was the main component of the soluble inorganic nitrogen in water. The soil organic N represented more than 70% of the total NO3-N loads in the Zhihugang. Domestic sewage was the major NO3-N source in the Liangxi river, with a contribution of greater than 50%. Soil organic N and domestic sewage, with contributions of more than 30% and 35% respectively, were the major NO3-N sources in the Lihe river and Daxigang river. Denitrification might be responsible for the shifting δN-NO3 and δO-NO3 values in the Daxigang river, and a mixing process may play a major role in N transformations in the Lihe river in winter. The results of this study will be useful as reference values for reducing NO3 pollution in the inflowing rivers in the north of Taihu Lake.


Introduction
Taihu Lake is the third largest freshwater lake in China, with an area of 2338 km 2 and an average depth of 2 m.The average growth rate of Gross Domestic Product (GDP) was 13.46% and the average growth rate of the population was 5.24‰ from 1991 to 2008 in Taihu Lake basin.This growth has resulted in large quantities of nutrients being discharged into Taihu Lake, which have caused eutrophication of the lake [1].After years of environmental regulation, eutrophication problems are still serious.It has been reported that the main source of pollution into Taihu Lake was from the rivers around the lake, and nitrogen (N) was the main pollutant of these rivers [2][3][4].Paired-sample T test analysis indicated that the input of total nitrogen (TN) to Taihu Lake was different (p = 0.034 < 0.05) between the 1990s and the early 21st century, and the inputs of TN to Taihu Lake were still rising (Figure 1) [5,6].Some research illustrated that the inflowing rivers in the west or northwest of Taihu Lake were seriously polluted, and more than 70% of the TN entering Taihu Lake was from the northern rivers [7].Therefore, it is important and necessary to study the sources of N in the northern rivers.With the rapid development of the economy and the population increases in recent decades, excess nitrogenous substances have drained into the rivers [8] from synthetic N fertilizers [9], animal wastes and manure, sewage, and atmospheric deposition [10], which has caused adverse effects on the water's ecological environment, such as eutrophication and the degradation of ecosystems (National Research Council, 2000).According to the date of pollutants of Taihu Lake in 1998, the pollutants from domestic sewage and agriculture accounted for 25% and 28% of TN, respectively [11].It is necessary to identify the sources and transformations of nitrate in water to control nitrate pollution.
Water 2017, 9, 345 2 of 15 synthetic N fertilizers [9], animal wastes and manure, sewage, and atmospheric deposition [10], which has caused adverse effects on the water's ecological environment, such as eutrophication and the degradation of ecosystems (National Research Council, 2000).According to the date of pollutants of Taihu Lake in 1998, the pollutants from domestic sewage and agriculture accounted for 25% and 28% of TN, respectively [11].It is necessary to identify the sources and transformations of nitrate in water to control nitrate pollution.
Figure 1.Total Nitrogen (TN) input and output throughout the years in Taihu Lake [2,4].
Surface water systems are very vulnerable to nitrate pollution.The sources of nitrate are mainly rain, chemical fertilizers, sewage and animal wastes, and nitrate derived from nitrification.Due to the distinct isotopic characteristics of nitrate, stable isotope techniques have been used to identify the sources of nitrate-nitrogen [12,13].The range of δ 15 N values of artificial fertilizers generally extends from −6‰ to +6‰.There are some natural factors and biological activities that cause small variations in the N isotopic ratios of soil organic N, such as soil depth, vegetation types, mineralization and nitrification.Therefore, δ 15 N values for soil organic N generally range from 0‰ to +8‰ [14,15].The N isotopic ratios in sewage and manure are more highly enriched than those of the other NO3 − -N sources.During the storage, treatment, and application of sewage and animal wastes, the ammonia in the sewage evaporates, causing 15 N enrichment in the residual NH4 + -N.The remaining NH4 + -N is subsequently oxidized into δ 15 N-enriched nitrate.As a result of this process, the δ 15 N values of sewage generally range from +8‰ to +20‰ [16,17].
However, as a result of fractionation, the N isotope composition of NO3 − could change during transformation [18].For example, ammonification and denitrification results in N isotope fractionation between organic matter and NH4 + [19,20].Therefore, the NO3 − -N source cannot be identified reliably by using the δ 15 N-NO3 − technique on its own.In order to solve this problem, a dual isotope method with analysis of δ 15 N and δ 18 O values of nitrate (i.e., δ 15 N-NO3 − and δ 18 O-NO3 − ) has been used to identify NO3 − -N sources [21,22].Researchers have found that δ 18 O-NO3 − is not only helpful for identifying sources of NO3 − -N, but it is also useful to help distinguish between NO3 − -N from atmosphere deposition, biological processes in soil and chemical fertilizers, and to study biological denitrification processes in water [23][24][25][26][27][28] .It has been suggested that δ 18 O was more useful for separating atmospheric NO3 − -N deposition from biologically-produced soil NO3 − -N than δ 15 N [28].The δ 18 O values in NO3 − -N fertilizers range from +17‰ to +25‰ and the δ 18 O values of atmospheric NO3 − -N deposition are higher than 60‰ [29].
Many studies have used N and O stable isotope ratios to identify the sources of nitrate in Taihu Lake.In those studies of Taihu Lake, the researchers found that the value of 15 N-NO3 − in winter was lowest of the four seasons, and human activities were the main source of nitrate of Taihu Lake in winter, but agriculture was the main source of nitrate of Taihu Lake in spring and summer [30,31].In some studies, some researchers found that in winter, the biodegradation was weak and the N and O stable isotopes were similar from year to year in Taihu Lake basin [32].Some research found that the main source of N pollution of water in Taihu basin was not the N fertilizer from farmland but the domestic sewage and human and animal excreta discharged into the water, and the atmospheric Surface water systems are very vulnerable to nitrate pollution.The sources of nitrate are mainly rain, chemical fertilizers, sewage and animal wastes, and nitrate derived from nitrification.Due to the distinct isotopic characteristics of nitrate, stable isotope techniques have been used to identify the sources of nitrate-nitrogen [12,13].The range of δ 15 N values of artificial fertilizers generally extends from −6‰ to +6‰.There are some natural factors and biological activities that cause small variations in the N isotopic ratios of soil organic N, such as soil depth, vegetation types, mineralization and nitrification.Therefore, δ 15 N values for soil organic N generally range from 0‰ to +8‰ [14,15].The N isotopic ratios in sewage and manure are more highly enriched than those of the other NO 3 − -N sources.During the storage, treatment, and application of sewage and animal wastes, the ammonia in the sewage evaporates, causing 15 N enrichment in the residual NH 4 + -N.The remaining NH 4 + -N is subsequently oxidized into δ 15 N-enriched nitrate.As a result of this process, the δ 15 N values of sewage generally range from +8‰ to +20‰ [16,17].However, as a result of fractionation, the N isotope composition of NO 3 − could change during transformation [18].For example, ammonification and denitrification results in N isotope fractionation between organic matter and NH 4 + [19,20].Therefore, the NO 3 − -N source cannot be identified reliably by using the δ 15 N-NO 3 − technique on its own.In order to solve this problem, a dual isotope method with analysis of δ 15 N and δ 18 O values of nitrate (i.e., δ 15 N-NO 3 − and δ 18 O-NO 3 − ) has been used to identify NO 3 − -N sources [21,22].Researchers have found that δ 18 O-NO 3 − is not only helpful for identifying sources of NO 3 − -N, but it is also useful to help distinguish between NO 3 − -N from atmosphere deposition, biological processes in soil and chemical fertilizers, and to study biological denitrification processes in water [23][24][25][26][27][28] .It has been suggested that δ 18 O was more useful for separating atmospheric NO 3 − -N deposition from biologically-produced soil NO 3 − -N than δ 15 N [28].
Many studies have used N and O stable isotope ratios to identify the sources of nitrate in Taihu Lake.In those studies of Taihu Lake, the researchers found that the value of 15 N-NO 3 − in winter was lowest of the four seasons, and human activities were the main source of nitrate of Taihu Lake in winter, but agriculture was the main source of nitrate of Taihu Lake in spring and summer [30,31].In some studies, some researchers found that in winter, the biodegradation was weak and the N and O stable isotopes were similar from year to year in Taihu Lake basin [32].Some research found that the main source of N pollution of water in Taihu basin was not the N fertilizer from farmland but the domestic sewage and human and animal excreta discharged into the water, and the atmospheric deposition was also another source of the surface waters [33].But, in some studies, the results indicated that the non-domestic sewage (not the domestic sewage) was the primary source of nitrate pollution in Taihu Lake [34].The nitrate of groundwater was also a risk of lake eutrophication and water safety [35], but not the main nitrate source in Taihu Lake [36].These studies have covered many fields, such as the nitrate source of Taihu Lake in different seasons, the reliability of the method to trace the nitrate source, and so on.However, the results of these studies have shown some inconsistencies with the main source of nitrate in Taihu Lake in the same season or the same region, and the contribution of the main sources has not been quantified either.
In order to obtain more reliable results, samples were collected in winter and N and O stable isotopes were used to identify and quantify the nitrate sources in four rivers (Zhihugang river, Liangxi river, Lihe river, and Daxigang river) flowing into the lake.We hope the results of this study will be useful as reference values for reducing NO 3 − -N pollution in the inflowing rivers in the north of Taihu Lake, and assist the management of land-use and pollution-source control in the watershed.

Study Area
The Zhihugang river and the Liangxi river connect with Meiliang Bay, which is located in the northwest of Taihu Lake.Land in the watershed is mainly used for growing crops, industry and human settlements.According to data for the period from 2005 to 2007, the mean TN concentrations in the Zhihugang and Liangxi rivers were 5.38 mg/L and 6.90 mg/L respectively [37].The Lihe and the Daxigang rivers are located to the northeast of Taihu Lake; they flow into Gonghu Bay.These catchments are dominated by forests and cropland, and land bordering the rivers which is used for construction [38].Basic information about the Zhihugang, Liangxi, Lihe, and Daxigang rivers is shown in Table 1.

Sample Collection and Measurement
Water samples were collected from the Zhihugang (ZH), Liangxi (LX), Lihe (LH), and Daxigang (DX) rivers from 5 to 15 December 2013.Sampling sites ZH1, LX1, LH1, and DX1 were located in the upper reaches of the rivers and were surrounded by many industries and houses.Sampling sites ZH4, LX4, LH4, and DX4 were located in the estuaries to Taihu Lake.Sampling sites ZH2, LX2, LX3, LH2, LH3, DX2, and DX3 were adjacent to settlements and cropland.Sampling sites ZH3 and LH3 were located in industrial areas (Figure 2).pH and dissolved oxygen (DO) were determined directly in situ using a multi-parameter water quality monitoring instrument (METTLER TOLEDO, SevenGo Duo pro, SG68, America).Calibration of sensors was performed before the measurements.Ammonium (NH 4 + -N) was measured with Nessler' reagent; nitrate nitrogen (NO 3 − -N) and total nitrogen (TN) were determined through ultraviolet spectrophotometry and the alkaline potassium persulfate oxidation-UV spectrophotometric method (GB3838-2002) respectively.Chloride (Cl − ) were measured using silver nitrate titration.Detection limits were 0.025, 0.02, 0.003 and 0.5 mg/L for NH 4 + -N, NO 3 − -N, TN, and chloride (Cl − ) respectively.Total organic carbon (TOC) concentrations were determined by a TOC-V Analyser (SHIMADZU -TNM-1, Japan), with an estimated detection limit of 0.05 µg/L.
Water 2017, 9, 345 4 of 15 0.5‰) international reference materials for δ 15 N and δ 18 O were used to calibrate the sample data after correction for blanks.The sample detection error was 0.2‰.The isotope results are expressed in δ units defined as: where R represents the 15 N/ 14 N or 18 O/ 16 O ratio, expressed as δ 15 N and δ 18 O.δ 15 N and δ 18 O were expressed relative to air and the Vienna standard mean ocean water (V-SMOW), respectively.

Physicochemical Characteristic of Water Samples
Information on the water chemistry and concentrations of dissolution inorganic nitrogen (DIN) of the samples are presented in

Physicochemical Characteristic of Water Samples
Information on the water chemistry and concentrations of dissolution inorganic nitrogen (DIN) of the samples are presented in Table 2.The pH of the Zhihugang, Liangxi, Lihe, and Daxigang rivers ranged from 7.72 to 7.94 (mean = 7.79), 8.46 to 8.58 (mean = 8.51), 8.08 to 8.53 (mean = 8.30), and from 8.09 to 8.84 (mean = 8.42), respectively.The four rivers were slightly alkaline.DO in the Zhihugang, Liangxi, Lihe, and Daxigang Rivers ranged from 5.90 to 9.24 mg/L (mean = 7.23 mg/L), 10.49 to 10.73 mg/L (mean = 10.60 mg/L), 8.37 to 10.07 mg/L (mean = 9.47 mg/L), and from 8.66 to 11.38 mg/L (mean = 9.84 mg/L), respectively.The method of Paired-sample T test indicated that DO in Zhihugang river was significantly different from Liangxi river and Li river (p = 0.022 and p = 0.042, respectively), and was not significantly different with Daxigang river (p = 0.082); and there was a significant difference between Liangxi river and Lihe river (p = 0.046), but not with Daxigang river (p = 0.314).were lower in the section that fed the Liangxi river than in the section that fed the Zhihugang and Daxigang rivers [44] and there is lower density of population and factories.NH 4 + -N and NO 3 − -N concentrations were higher at the source of the Lihe river (LH1) than at the other sampling sites in this river; this is because LH1 is located at the junction of two rivers, one of which (the Liangtang river) received inputs from the Grand Canal.Out of all the rivers, there was more variation in the NH 4 + -N and NO 3 − -N concentrations in the Zhihugang river, mainly because it is the longest of the four rivers, and there are more pollution sources with a wide range of NH 4 + -N and NO 3 − -N concentrations along the river [45][46][47].The NH 4 + -N and NO 3 − -N concentrations were higher in Meiliang Bay than in Gonghu Bay, and the NH 4 + -N and NO 3 − -N concentrations were higher at ZH4 and LX4 than at LH4 and DX4, because of mixing processes.NO 3 − -N concentrations were higher than NH 4 + -N concentrations in the four rivers, and NO 3 − -N in the four rivers was not from the nitrification of NH 4 + -N in water, but from the external pollution sources.Pearson correlation analysis (shown in Appendix A, Table A1) showed that there were no strong positive relationships between DO and NO 3 − -N or negative relationships between DO and NH 4 + -N, but there was a strong positive relationship between NO 3 − -N and NH 4 + -N (p < 0.01).The low temperatures (≈5 ± 3 • C) were not beneficial for nitrification (Brookshire, et al., 2011), indicating that nitrification in surface water might be not intense in our study.As reported, water quality and level of eutrophication in Meiliang Bay were more severe than in Gonghu Bay [48,49].It indicated that the water quality of rivers had significant effect to the lake, and it is necessary to study the nitrate source and N transformation of the rivers.

Nitrate Sources and Nitrogen Transformation
The values of δ 15 N-NO  2), which indicates that the NO 3 − -N source of these four rivers is very complicated.Because of the wide variability of NO 3 − -N from different sources, in this study we used the δ 18 O-NO 3 − ratio to help identify NO 3 − -N sources [15,52].The δ 18 O-NO 3 − values in our samples were lower than 15‰, further suggesting that neither atmospheric precipitation [29] (δ 18 O-NO 3 − > 60‰,) nor nitrate-fertilizer is major sources of riverine nitrate in our study.After one-way ANOVA analysis of δ 15 N-NO 3 − and δ 18 O-NO 3 − in the four rivers, it was found there were no significant differences of δ 15 N-NO 3 − (p = 0.071 > 0.05), but significant differences of δ 18 O-NO 3 − (p = 0.032 < 0.05) in the four rivers.Thus, we inferred that the sources of nitrate in the four rivers might be different.Figure 3 shows that there may be differences between the NO 3 − -N sources in the four rivers.
The NO 3 − -N that primarily originates from sewage has δ 15 N-NO 3 − values of nearly 10‰, such as were reported for sites ZH1, ZH3, ZH4, and DX1.These sites could be assigned to one source.Water samples with a low NO 3 − -N content, such as those at sites LH3, LH4, and DX3, have δ 15 N-NO 3 − values between +3‰ and +8‰, suggesting that the NO 3 − -N might be originally derived from nitrification of ammonium in precipitation, fertilizer, sewage and manure, as well as soil organic N.But the value of δ 18 O-NO 3 − at these sites was from −12.26‰ to +2.56‰, that was much less than the value of δ 18 O-NO 3 − coming from the nitrification of ammonium (+25‰-+75‰) in precipitation and fertilizer (+17‰-+25‰) [53][54][55].Thus we inferred that the source of nitrate of these sites was from soil organic N (−10‰-+15‰) [56].The low water temperature (≈5 ± 3 • C) and the high dissolved oxygen (DO > 6 mg/L) was not beneficial for denitrification, and there was no significant negative correlation between NO 3 − and δ 15 N-NO 3 − by Pearson correlation analysis, so we inferred that the denitrification at these points was not the main reason for the low concentration of nitrate nitrogen and high δ 15 N-NO 3 − values in the water [57,58].So the samples with low NO 3 − -N concentrations have δ 15 N-NO 3 − values greater than 10‰, which indicates that the NO 3 − -N was the result of a mixture of sewage and soil organic N; this was the case at sampling sites ZX2, LX1, LX2, LX3, LX4, LH1, LH2, DX2, and DX4.
Water 2017, 9, 345 7 of 15 δ 15 N-NO3 − values greater than 10‰, which indicates that the NO3 − -N was the result of a mixture of sewage and soil organic N; this was the case at sampling sites ZX2, LX1, LX2, LX3, LX4, LH1, LH2, DX2, and DX4.Several influence factors were found, which influence isotopic compositions of NO3 − -N sources, such as mixing, ammonification, nitrification, denitrification and so on.Although each NO3 − -N source has its own distinctive isotopic composition, the mixture may lead the intermediate values.
The ammonification and nitrification results in some fractionation of δ 15 N-NO3 − , the enrichment factors are between −1‰ to +1‰ and 12‰ to 29‰ [59].Microbial denitrification may result in some fractionation of δ 15 N-NO3 − and δ 18 O-NO3 − , however the enrichment factors are within −40‰ to −5‰ and −18‰ to −8‰ respectively [60,61].Therefore, Cl − may be a useful indicator of sewage and manure, because of its stability in water and its resistance to physical, chemical and biological processes.Therefore, the NO3 − -N/Cl − ratio can be used to obtain further information about the effects of N transformation processes, such as dilution or denitrification [21,53].The Cl − concentrations in the Zhihugang river gradually declined along the river (Table 2), and there was generally a positive correlation (Figure 4, R 2 = 0.7223) between Cl − and NO3 − -N, which indicates that mixing and dilution processes had a major effect on NO3 − -N transport.A positive relationship (Figure 5) was found between δ 15 N-NO3 − and δ 18 O-NO3 − , with a linear regression slope of 0.8628 in the Zhihugang river, which indicated that denitrification occurred in the Zhihugang river but less significantly.There was little variation in Cl − concentrations in the Liangxi river (Table 2), and Cl − and NO3 − -N in the Liangxi river were not correlated (Figure 4, R 2 = 0.4555), which indicates that mixing and dilution processes had limited effects on NO3 − -N transport.There was a negative correlation between δ 15 N-NO3 − and δ 18 O-NO3 − in the Liangxi river (Figure 5, k = −0.4337),which suggests that denitrification may not be the reason for the shift in the δ 15 N-NO3 − and δ 18 O-NO3 − values, and δ 18 O-NO3 − values indicates a mixing of different nitrate sources.Algae would easily grow in the Liangxi river because of its slow flow, and assimilation by algae would lead to enrichment of residual NO3 − -N with heavy isotopes [62,63]; this helps explain the low NH4 + -N and NO3 − -N concentrations, and the high δ 15 N-NO3 − values, in the Liangxi river.
The ranges of Cl − concentrations in Lihe river exhibited no large variation, but a positive correlation (Figure 4, R 2 = 0.6497) between Cl − and NO3 − -N had been observed, indicating that the mixing process had an effect on nitrate transportation in Lihe river.No negative relationship between NO3 − -N and δ 15 N/ δ 18 O-NO3 − was observed (not shown) in the Lihe river.But a positive relationship was found between δ 15 N-NO3 − and δ 18 O-NO3 − with the slope of the linear regression of to −8‰ respectively [60,61].Therefore, Cl − may be a useful indicator of sewage and manure, because of its stability in water and its resistance to physical, chemical and biological processes.Therefore, the NO 3 − -N/Cl − ratio can be used to obtain further information about the effects of N transformation processes, such as dilution or denitrification [21,53].The Cl − concentrations in the Zhihugang river gradually declined along the river (Table 2), and there was generally a positive correlation (Figure 4, R 2 = 0.7223) between Cl − and NO 3 − -N, which indicates that mixing and dilution processes had a major effect on NO 3 − -N transport.A positive relationship (Figure 5) was found between δ 15 N-NO 3 − and δ 18 O-NO 3 − , with a linear regression slope of 0.8628 in the Zhihugang river, which indicated that denitrification occurred in the Zhihugang river but less significantly.There was little variation in Cl − concentrations in the Liangxi river (Table 2), and Cl      was 3.42 (mg/L), 0.79 (mg/L), 0.41 (mg/L) and 1.20 (mg/L) respectively.f i are the contributions of the respective sources of NO 3 − -N.Q i was the annual runoff of four rivers (Table 1).Results of the calculations are shown in Table 4. Table 4 shows that the annual amounts of NO 3 − -N in the Zhihugang river and Daxigang river were higher than the annual amounts of TN in the two rivers (in Table 1).This could be because the annual amount of TN in the four rivers in Table 1 was the net amount of TN, but the result of the Table 4 was the gross amount.But this result indicates that nitrate pollution was the main nitrogen pollution in the Zhihugang and Daxigang rivers, and the N transformations in the two rivers were complicated, such as the denitrification.So, pollution control and harnessing in the Zhihugang and Daxigang river basins should be strengthened.Annual amounts of NO 3 − -N from sewage in the four rivers discharging into Meiliang Bay and Gonghu Bay were 34.8 to 207.6 ton and 70.7 to 98.9 ton respectively.Some reports have shown that the annual average proportions of the domestic sewage source in Meiliang Bay and Gonghu Bay were 17.2% and 15.3% respectively [34].Based on the results of Zhen et al. [34] and Wang et al. [40], annual total amounts of NO 3 − -N from sewage discharging into Meiliang Bay and Gonghu Bay were about 357 and 253 ton respectively.The sewage was not the main NO 3 − -N source in Zhihugang river, but the maximum annual amount of NO 3 − -N from sewage accounted for about 51.5% of annual total amounts NO 3 − -N imported into Meiliang Bay.
Therefore, it is very necessary to pay attention to the control of sewage pollution in the watershed of the Zhihugang river.

Conclusions
Results from this study of four rivers to the north of Taihu Lake show that the water quality of the rivers to the northwest of Taihu Lake was worse than that of rivers in the northeast (according to the national quality standards for surface waters, China (GB3838-2002)).Also, NO 3 − -N was the dominant inorganic N species in the four rivers during the sampling season.It means that we need more strict protection measures, emission standards and wastewater treatment with the abilities of denitrification to protect the rivers to the northwest of Taihu Lake.Because it was winter, we ignored the contribution of atmospheric NO 3 − -N to rivers.Soil organic N, with a contribution from 75.06% to 79.47%, was the major NO 3 − -N source in the Zhihugang.Domestic sewage, with a contribution from 56.70% to 71.13%, was the major NO 3 − -N source in the Liangxi river.Dilution and mixing processes may play a major role in N transformations in the Zhihugang river, but might contribute to assimilation by algae in the Liangxi river.Soil organic N and domestic sewage were the main NO 3 − -N sources in the Lihe and Daxigang rivers, with contributions greater than 30% and 35% respectively.Denitrification might result in the enrichment of heavy isotopes of NO 3 − -N in the Daxigang river, but mixing process may play a major role in N transformations in the Lihe river.The maximum annual amount of NO 3 − -N from sewage in the Zhhugang river was highest of the four rivers.Therefore, while sewage was not the main NO 3 − -N source in the Zhihugang river, it is very necessary that we pay attention to the control of sewage pollution in the watershed of the Zhihugang river.

from 1 .
09 to 4.75 mg/L (mean = 2.52 mg/L), 0.13 to 0.46 mg/L (mean = 0.35 mg/L), 0.10 to 0.51 mg/L (mean = 0.27 mg/L), and from 0.24 to 1.61 mg/L (mean = 0.72 mg/L), respectively.The highest concentration of NH 4 + -N in four rivers appeared at ZH3, LX1, LX4, LH1, DX1, and lowest concentration appeared at ZH2, LX2, LH3, DX4, respectively.NO 3 − -N concentrations in the Zhihugang, Liangxi, Lihe, and Daxigang rivers ranged from 2.40 to 4.45 mg/L (mean = 3.43 mg/L), 0.76 to 0.83 mg/L (mean = 0.79 mg/L), 0.29 to 0.53 mg/L (mean = 0.42 mg/L), and from 0.43 to 3.31 mg/L (mean = 1.20 mg/L), respectively.The highest concentrations of NO 3 − -N in four rivers appeared at ZH1, LX2, LH1 and DX1, and the lowest concentrations appeared at ZH3, LX4, LH2 and DX3.TN concentrations in the Zhihugang, Liangxi, Lihe, and Daxigang rivers ranged from 4.45 to 7.50 mg/L (mean = 6.12 mg/L), 1.18 to 1.35 mg/L (mean = 1.26 mg/L), 0.71 to 1.35 mg/L (mean = 1.08 mg/L), and from 1.14 to 5.12 mg/L (mean = 2.31 mg/L) respectively.By one-way analysis of Variance (ANOVA) analysis of NH 4 + , NO 3 − and TN of four rivers(p = 0.015 < 0.05, 0.001 < 0.05 and 0.00 < 0.05, respective), the concentration of NH 4 + , NO 3 − and TN in four rivers had significant differences, and it might indicate that the sources in the four rivers were different.Variations of the NH 4 + -N, NO 3 − -N, and TN concentrations in the Zhihugang river, Liangxi river, Lihe river, and Daxigang rivers are shown in Figure A1 (shown in Appendix A).Because of the extensive area the Grand Canal basin covers and the large number of industries and residential areas within it, NH 4 + -N and NO 3 − -N concentrations are high in the Grand Canal.The NH 4 + -N and NO 3 − -N concentrations at site ZH1 in the Zhihugang river and at site DX1 in the Daxigang river were higher than at the other sampling sites because of inputs from the Grand Canal and high density population and factories around the sample sites.The NH 4 + -N and NO 3 − -N concentrations at LX1 were lower than those at ZH1 and DX1, because the NH 4 + -N and NO 3 − -N concentrations in the Grand Canal

− and NO 3 − 3 −
-N in the Liangxi river were not correlated (Figure 4, R 2 = 0.4555), which indicates that mixing and dilution processes had limited effects on NO 3 − -N transport.There was a negative correlation between δ 15 N-NO 3 − and δ 18 O-NO in the Liangxi river (Figure 5, k = −0.4337),which suggests that denitrification may not be the reason for the shift in the δ 15 N-NO 3 − and δ 18 O-NO 3 − values, and δ 18 O-NO 3 − values indicates a mixing of different nitrate sources.Algae would easily grow in the Liangxi river because of its slow flow, and assimilation by algae would lead to enrichment of residual NO 3 − -N with heavy isotopes [62,63]; this helps explain the low NH 4 + -N and NO 3 − -N concentrations, and the high δ 15 N-NO 3 − values, in the Liangxi river.The ranges of Cl − concentrations in Lihe river exhibited no large variation, but a positive correlation (Figure 4, R 2 = 0.6497) between Cl − and NO 3 − -N had been observed, indicating that the mixing process had an effect on nitrate transportation in Lihe river.No negative relationship between NO 3 − -N and δ 15 N/ δ 18 O-NO 3 − was observed (not shown) in the Lihe river.But a positive relationship was found between δ 15 N-NO 3 − and δ 18 O-NO 3 − with the slope of the linear regression of 3.95 (Figure 5, R 2 = 0.8526) in Lihe river, which indicated that denitrification might be responsible for the shifting δ 15 N-NO 3 − and δ 18 O-NO 3 − values, although the slope was bigger than that reported for denitrification (1.3-2.1)[64-66].No clear positive correlation (Figure 4, R 2 = 0.2651) between Cl − and NO 3 − -N was observed, but a gradual decline of Cl − concentrations was observed in the Daxigang river, which may indicate that dilution plays a major role in the variation of Cl − and NO 3 − .A strong positive (Figure 5) between δ 15 N-NO 3 − and δ 18 O-NO 3 − , with a linear regression slope of 1.0797 of the Daxigang river (Figure 5, R 2 = 0.378), which indicated that denitrification might be mainly responsible for the shifting δ 15 N-NO 3 − and δ 18 O-NO 3 − values.Water 2017, 9, 345 8 of 15 3.95 (Figure 5, R 2 = 0.8526) in Lihe river, which indicated that denitrification might be responsible for the shifting δ 15 N-NO3 − and δ 18 O-NO3 − values, although the slope was bigger than that reported for denitrification (1.3-2.1)[64-66].No clear positive correlation (Figure 4, R 2 = 0.2651) between Cl − and NO3 − -N was observed, but a gradual decline of Cl − concentrations was observed in the Daxigang river, which may indicate that dilution plays a major role in the variation of Cl − and NO3 − .A strong positive (Figure 5) between δ 15 N-NO3 − and δ 18 O-NO3 − , with a linear regression slope of 1.0797 of the Daxigang river (Figure 5, R 2 = 0.378), which indicated that denitrification might be mainly responsible for the shifting δ 15 N-NO3 − and δ 18 O-NO3 − values.

Table 1 .
Hydrologic data of the four studied rivers and annual amounts of contaminants imported.

Table 2 .
Physicochemical characteristics and isotopic compositions of water in four studied rivers (mean values ± SD).

Table 4 .
Import of nitrate in the four rivers.