Evaluating the Sources and Fate of Nitrate in the Alluvial Aquifers in the Shijiazhuang Rural and Suburban Area, China: Hydrochemical and Multi-isotopic Approaches

The identification of the sources and fate of NO contaminants is important to protect the water quality of aquifer systems. In this study, NO contaminated groundwater from the drinking water wells in the Shijiazhuang area, China, was chemically (NO /Cl − ratio) and isotopically (δ 15 NNO3, δ 18 ONO3 and δ 13 CDOC; δ 2 HH2O, δ 18 OH2O) characterized to identify the sources of NO and address subsequent biogeochemical processes. The positive correlations between dominant anions and cations suggested that the dissolution of calcium carbonate and gypsum minerals was the most effective process in the groundwater. Elevated concentrations of NO , Cl − and Mg 2+ could be related to the wastewater irrigation and usage of fertilizers. The natural water in the study area originated primarily from precipitation and experienced a limited extent of evaporation, as demonstrated by measurements of δ


Introduction
Shijiazhuang (SJZ) City is a representative area, which has been developed from a rural area to a metropolitan city in China.In most suburban and rural regions of SJZ, groundwater is the major source of drinking water due to the inadequate centralized water supply system.Nitrate contamination in groundwater caused by intensive land-use activities has become a major environmental concern in this area [1,2].To prevent the deterioration of groundwater quality due to nitrate contamination, it is important to discern the major sources and fate of nitrate [3][4][5].
Some studies have been conducted to identify the nitrate contamination of groundwater using stable isotopes near the city of SJZ [6][7][8].In the wastewater-irrigated area (the southern part of the study area), a stable N isotope of nitrate indicated that the nitrate in the groundwater originated from wastewater or a mixture of pig manure and wastewater [6,8].However, the single isotope approach applied in these studies could not fully interpret the transformation of nitrogen in the groundwater.In the non-wastewater-irrigated area, a combination of δ 15 NNO3 and δ 18 ONO3 was used to trace NO sources in the shallow groundwater, suggesting that NH fertilizer and manure were the main sources of NO in the groundwater; nevertheless, denitrification seriously interfered with determining the source of nitrate [7].
In addition, previous studies drew different conclusions on whether microbial denitrification occurred.Chen et al., found a good linear relationship of δ 15 N and the logarithm of the residual NO that was consistent with denitrification of NO in groundwater in the southern part of the study area [6].Similarly, Liu and Chen observed a linear relationship with a slope of 0.644 (ɛN/ɛO = 1.5) between δ 15 NNO3 and δ 18 ONO3 values (4.6‰-9.7‰and 2.3‰-7.8‰,respectively) confirming the occurrence of denitrification [7].Others, however, suggested the absence of denitrification based on constant NO /Cl − values, high DO values (1.20-5.97mg/L) and a limited range of δ 15 NNO3 values [8].In general, it is difficult to specify the varied sources and processes that affect the isotope composition of NO due to the isotope fractionation combined with similar isotopic values of various NO sources in SJZ area.
Similarly, isotopic composition of nitrate used to constrain the sources and fate of N also appears to have certain limitation in other study cases [4,5,9,10].δ 15 NNO3 was not sufficient to estimate potential sources of nitrate contamination in rural regions in Korea since the concentration and δ 15 NNO3 of the unconfined groundwater nitrate exhibited seasonal variations due to seasonal changes in agricultural activities [11].Later, the same author proposed that the impacts of various land-use activities on the nitrate contamination of groundwater could be more precisely inferred from long-term data on the concentration and δ 15 NNO3 of nitrate [12].Since denitrification and mixing processes limited the identification of NO sources by δ 15 NNO3 in the groundwater, δ 15 NNO3 was combined with co-migrating isotopic tracers (δ 11 B and 87 Sr/ 86 Sr) to distinguish various NO sources under intensive agricultural activities in Brittany (France) [13].In addition, based on the level of natural NO in the groundwater, anthropogenic contamination (inorganic fertilizer) can be discriminated from the soil organic N using δ 15 NNO3 and δ 18 ONO3 isotopic values in the Mt.Vulture area (Italy) [14].These studies indicated that, for better understanding of the various sources and transformations of NO in groundwater, it is important to couple isotopic composition of NO with multiple lines of evidence.
Dissolved organic carbon (DOC) is ubiquitous in groundwater [15].Changes in concentration and composition of DOC reflect biogeochemical sources and reaction processes that determine the chemical composition of groundwater [16][17][18].Numerous studies have described the importance of organic carbon in controlling the occurrence of denitrification.Several studies relate denitrification activity to DOC concentration [19,20].The availability of organic carbon limits the denitrification rate [17,[21][22][23][24]. On the other hand, biogeochemical activities can be inferred from elevated δ 13 CDOC values [25,26].Breukelen et al., found that dissolved organic carbon (DOC) concentration decreased in association with increasing δ 13 CDOC values in a landfill leachate plume, supporting the occurrence of degradation and denitrification potentially being the dominated redox process at the top fringe [27].It is hypothesized that the combination of δ 13 CDOC and δ 15 NNO3 isotopic values would provide a sensitive method for tracing denitrification process in groundwater.
The NO /Cl − method has been proven to be a valid and effective tool to distinguish the effect of N removal processes by dilution from denitrification [28,29].In addition, the NO /Cl − in groundwater can be used as a source indicator because different sources of nitrate and chloride have different ratios in most cases [18,30].Nishikawa et al., reported that the NO /Cl − ratios for the native groundwater, imported water, and septage collected in Warren ground-water basin were completely different.Based on the NO /Cl − ratios and NO concentrations, a three-part mixture of native ground water, imported water, and septage was identified [18].Liu et al., showed the variation of the NO /Cl − molar ratios with Cl − concentrations in the surface and groundwater samples as well as sewage samples in both the summer and winter seasons [30].Two NO source end members were distinguished as agricultural inputs with low-Cl/high-NO /Cl − ratios and municipal inputs with high-Cl/low-NO /Cl − ratios close to sewage.
In this study, the characteristics of δ 15 NNO3, δ 18 ONO3, and δ 13 CDOC were investigated in the unconfined Quaternary aquifers of SJZ, China.Traditional hydrochemical and isotopic parameters such as chemical composition (anion, cation, DOC concentrations and NO /Cl − ratio), δ 2 HH2O, and δ 18 OH2O of groundwater were also measured.The aim of this study is to combine δ 15 NNO3, δ 18 ONO3, and δ 13 CDOC data with the NO /Cl − ratio to understand biogeochemical processes in an alluvial aquifer where there are multiple potential NO sources, and to identify the mixing process of different NO sources in this region.

Description of the Study Area
The Shijiazhuang (SJZ) area is located in the piedmont pluvial plain of Hutuo River (Figure 1).The region is middle-latitude continental semiarid monsoon climate with mean annual temperature of 12-13 °C, mean annual precipitation of 500-600 mm, and potential evaporation of 1100-1800 mm [31,32].Historically, the Hutuo River played an important role in the formation of the alluvial fan at the front edge of the piedmont plain zone in the study area [33].Today, the river is disconnected and severed due to drought and interception of water by the Huangbizhuang reservoir.The Shijin and Dongming canals are mainly used for agricultural irrigation.Xiao River is a seasonal stream originating from a mountainous area.Over the past few decades, a quantity of industrial and domestic wastewater flowed into Xiao River through Dongming Canal.The aquifers in this area are mainly composed of the multi-layer loose deposits of Quaternary with a depth of 350-450 m in the piedmont (Figure 2).The regional Quaternary aquifers can be divided into four groups: The Holocene formation (Q4), the late Pleistocene formation (Q3), the middle Pleistocene formation (Q2), and the early Pleistocene formation (Q1) [34].Aquifer 1 (A1) is an unconfined shallow aquifer dominated by coarse-grained sand with the bottom boundary at 10-20 m deep, in which groundwater has been drained.Aquifer 2 (A2), a 50 m thick shallow confined aquifer (30-80 m depth), contains sandy gravel and medium to fine sand, and is the main source of groundwater for drinking water supplies.Aquifers 3 (A3) and 4 (A4) are both deep confined aquifers, consisting of 90 m thick sandy gravel (80-200 m depth) and 50-60 m thick cemented sandy gravel (300-370 m depth), respectively [1,34].The shallow aquifers (A1 and A2) can be regarded as one system as a result of the natural hydraulic connection between them [31].Aquifers 3 and 4 can be recognized as the deep unit.The shallow aquifers are over developed and contaminated, increasing the need to pump clean groundwater from the deep aquifers.Several components make up groundwater recharge in the shallow groundwater aquifer unit: Precipitation, irrigation return flow, infiltration of surface water in rivers and canals, and lateral inflow from the Taihang Mountains [35].The deep aquifer is mainly recharged by infiltration of paleo-precipitation under cold climatic conditions with modern vertical inflow from the shallow aquifer as another source.
Currently, groundwater discharge in the study area is mainly through intensive localized pumping in SJZ City and widespread pumping via small capacity wells for irrigation in the surrounding country areas [31,32,36].Groundwater exploitation has resulted in serious water-level decline, forming a 400 km 2 groundwater cone of depression in the shallow aquifer [37,38].The water table is currently more than 10 m below ground surface; thus, evaporation of groundwater is generally considered negligible [39].

Sampling and Analytical Methods
Groundwater samples from domestic wells and surface water samples from the Huangbizhuang reservoir, Hutuo River and Shijin Canal were collected in September 2011 (Figure 1).Samples were collected unacidified for analysis of concentrations and values for anion, deuterium, and oxygen isotopes.Samples were acidified to pH < 2 with concentrated HNO3 for cation analysis.To obtain sufficient dissolved NO for N and O isotope analyses, five liters of water were taken from each sampling site.For dissolved organic carbon (DOC) concentration and C isotope analysis, 500 mL water samples were collected in amber glass bottles.All these samples were stored below 4 °C and transferred to the lab for chemical and isotopic analyses.The temperature, pH values and electrical conductivity (EC) were measured on site.Alkalinity was determined by titration within 24 h after sampling.Samples were filtered through 0.45 μm membrane filters prior to measurement of ion and DOC concentration, and C, N, O and H isotopic analyses.All these measurements were performed at China University of Geosciences (Wuhan), China.
All stable isotopic compositions are reported in per mil relative to their corresponding international standards (VPDB for 13 C, V-SMOW for 2 H and 18 O, AIR for 15 N), using the delta notation: δ(‰) = [(R − Rstd)/Rstd] × 1000, where R and Rstd are the isotope ratio of the sample and the standard, respectively.
For analysis of N and O isotopes in nitrate, a modification of the procedure described by Silva (2000) was used for the conversion of dissolved NO to AgNO3 salt [41].The water samples were filtered through 0.45 μm membrane filters in the laboratory.The Barium chloride was added to precipitate sulfate and phosphate in 5 L samples.Then the solution was filtered, agitated with activated carbon to remove dissolved organic matters, passed through a cation exchange column to remove all cations, neutralized with Ag2O, filtered to remove the AgCl precipitate, and freeze-dried to obtain solid AgNO3 [41].About 0.5 mg AgNO3 was transferred to silver capsules and loaded into the auto-sampler coupled to the EA (Flash 2000, Thermo Finnigan).N2 and CO generated after AgNO3 pyrosis at 1325 °C were separated in a chromatographic column at 60 °C, and then introduced into MAT 253 (Thermo Finnigan) where δ 15 NNO3 and δ 18 ONO3 were determined with a total uncertainty of ±0.3‰ for δ 15 NNO3 and of ±0.5‰ for δ 18 ONO3, respectively, for solutions of KNO3 standard processed through the entire column procedure.
To measure the C isotope of DOC, 200 mL water from each sample was concentrated to a volume of about 1 mL at 40 °C in a round-bottomed flask by rotary evaporation, and then acidified to pH 2 by adding 2-3 drops of 85% phosphoric acid to remove the dissolved inorganic carbon (DIC) by agitation.Flask was rinsed with two 1.5 mL aliquots of ultra-pure deionized water, resulting in a final volume of 3.5-4.5 mL.One hundred μL of the concentrated sample was transferred to silver capsules (8 mm × 5 mm) and dried at 70 °C.More details about the procedure for preparation were described in a previous study [42].
The δ 13 CDOC analyses was carried out using a MAT 253 (Thermo Finnigan) after combustion in a Flash EA via open split at a temperature of 1020 °C, with a precision of ±0.25‰.

Chemical Composition
The geochemical parameters and major ion composition of surface water (SW), shallow groundwater (SGW), and deep groundwater (DGW) samples are summarized in Table 1.Chemical data obtained are also plotted on a Piper diagram (Figure 3).The SW samples were alkaline with a pH of 8.1 to 9.2.Water temperatures ranged from 18 to 26 °C.Electrical conductivities (EC) varied from 577 to 717 μS/cm.Concentrations of dissolved organic carbon (DOC) in the SW samples ranged from 4.1 to 22.1 mg C/L.The dissolved oxygen (DO) in SW01 and SW03 were 4.7 and 4.8 mg/L, respectively, while the DO concentration in SW02 was lower (2.3 mg/L).The concentrations of cations and anions in most SW samples followed the order: Ca 2+ > Mg 2+ > Na + + K + and SO > HCO > Cl − + NO , except anions in the sample SW4 ( HCO > SO > Cl − + NO ).The pH values of SGW samples ranged from 7.0 to 7.6, and temperatures ranged from 16 to 26 °C.Conductivities varied from 795 to 1199 μS/cm.The DOC concentrations of most SGW samples varied in the range of 3.9 to 10.2 mg C/L, except for samples of SGW5 and SGW12 with values of 20.2 and 15.8 mg C/L, respectively.The DO concentration in SGW samples ranged from 1.1 to 3.1 mg/L.The orders of dominant cations and anions in SGW samples were Ca 2+ > Mg 2+ > Na + + K + and HCO > SO > Cl − + NO , except the order of anion in sample SGW5 was SO > HCO > Cl − + NO and in SGW10-13 was HCO > Cl − + NO > SO .In particular, the HCO3.Cl-Ca.Mg type of groundwater was generally found in the southeastern part of the study area where the wastewater was used for irrigation during the1960s to 2000s [6].The DGW samples had pH values of 7.5 to 7.9 and temperatures between 17 and 23 °C.Most samples had low conductivities ranging from 464 to 589 μS/cm, whereas sample DGW5 had a high value of 994 μS/cm.The DGW samples had relatively low DOC concentrations in a narrow range of 3.1 to 5.5 mg C/L.The DO concentration in DGW samples ranged from 1.5 to 2.6 mg/L.The orders of dominance of cations and anions in DGW samples were Ca 2+ > Mg 2+ > Na + + K + and HCO > SO > Cl − + NO .
NO concentrations of local surface water and groundwater varied significantly.Concentrations of NO in SW samples varied in a narrow range from 2.1 to 2.6 mg/L.The contents of NO in the SGW and DGW samples were determined in the range of 5.7 to 26.2 mg/L (average 15.5 mg/L) and 0.5 to 5.9 mg/L (average 3.2 mg/L), respectively.All samples had lower NO content than the statutory limit (50 mg/L for NO ) for drinking water (GB5749).However, the NO concentrations of most groundwater samples were significantly higher than the regional background value (4.4 mg/L in 1950s) in the groundwater.

Hydrochemical Characteristics of Water
Dissolution is the most effective process in groundwater chemistry and could be considered the first step in the hydrochemical evolution.As shown in Figure 4a,b, the molar ratios of Ca 2+ /HCO are close to 0.5, while the molar ratios of (Ca 2+ + Mg 2+ )/HCO are greater than 0.5.This indicates that the dissolution of calcium carbonate minerals is one of the most important hydrochemical processes [43].The linear correlations between Ca 2+ and SO , Na + and SO suggest that SO in groundwater may be mainly derived from the dissolution of gypsum and mirabilite (Figure 4c,d).High SO concentration in the surface water might be the result of coal mining drainage from mountainous areas [36].The shallow groundwater (HCO3.Cl-Ca.Mg type) with low SO concentration in the wastewater-irrigated area may be mainly affected by wastewater and precipitation (Figure 4c).The concentration of Mg 2+ and Cl − showed a linear relationship indicating that Mg 2+ and Cl − have a common source which could be attributed to the anthropogenic activities such as wastewater irrigation or usage of fertilizers (Figure 4e) [6,28,44].The plot of Na + and Cl − shows that the increased Cl − concentration is not closely related with the dissolution of halite or input of sea salt (Figure 4f).
In Figure 5a,b, NO concentrations show strong correlations with Mg 2+ and Cl − concentrations indicating that they originate from the same sources.The Mg 2+ and Cl − concentrations of the groundwater in the wastewater irrigation area were higher than those in the non-wastewater irrigated area, but the NO concentrations of groundwater in both areas fall within the similar range.This demonstrates that wastewater irrigation is not the only source of NO in groundwater.The concentrations of HCO and SO are not correlated well with those of NO (Figure 5c,d), since the levels of HCO and SO are controlled primarily by natural processes, and the concentration of NO is influenced mainly by anthropogenic inputs.The water samples all plot on or close to the local meteoric water line (LMWL) in Figure 6, suggesting that the surface water and groundwater originate from local precipitation.Most of the water samples have isotope compositions similar to the weighted mean isotope compositions (WMIC) of rain water in SJZ that has the composition of δ 2 HH2O = −55‰ and δ 18 OH2O = −7.8‰(n = 142) calculated with data from 1985 to 2003 obtained from the Global Network for Isotopes in Precipitation (GNIP) dataset (available at http://isohis.iaea.org)(Figure 6).This isotopic similarity implies that the shallow groundwater was derived by direct inflow or infiltration of modern rainwater.The deep groundwater with relatively negative δ 2 HH2O and δ 18 OH2O reflects a paleorecharge effect under cold climatic conditions.The surface water and groundwater samples on the dashed line below LMWL, with d-excess values of −2.2‰ to +14.8‰, suggest these samples experienced minor evaporation.The two surface water samples from Huangbizhuang (SW1) and Hutuohe reservoirs (SW2), with d-excess values of −15.2‰ and −11.6‰, suggest long-term strong evaporation of Huangbizhuang and Hutuohe reservoirs under the semi-arid climate.The groundwater samples (SGW1 and SGW9) were collected near the Huangbizhuang and Hutuohe reservoirs, indicating the seepage from Huangbizhuang and Hutuohe reservoirs are the major source of this groundwater.The DGW1 could be recharged by Hutuo River in an old and cold period.As shown in Figure 7, some of the shallow groundwater samples have similar δ 18 OH2O values and Cl − concentrations with the surface waters, suggesting that these samples can be recharged by the surface water.In addition, wastewater contributes a considerable amount of Cl − to the shallow groundwater in the mixing process.Moreover, deep groundwater quality could be influenced by seepage from shallow groundwater.The surface water sample (SW2) might be subjected to intensive evaporation resulting in high Cl − concentration and δ 18 OH2O value.

Evidence of Denitrification
Characterizing biogeochemical processes in groundwater is a key issue to understand the sources and behavior of NO .Microbial denitrification can decrease NO concentration and cause significant alterations of the isotopic composition of NO .Various studies have demonstrated that the heavy isotopes 15 NNO3 and 18 ONO3 become enriched with an ɛN/ɛO ratio that ranges from 1.3 to 2.1 in the residual NO pool during microbial denitrification [20,21,46,47].Although the individual isotopic enrichment factors for δ 15 NNO3 and δ 18 ONO3 during denitrification may vary according to specific field conditions, the fractionation ratio (ɛN/ɛO) appears to remain constant.In this study, most of the samples show a positive trend in a δ 15 NNO3 and δ 18 ONO3 diagram (Figure 8) with an ɛN/ɛO ratio of proximately 2, suggesting microbial denitrification might occur.There were more positive δ 15 NNO3 and δ 18 ONO3 values and lower NO concentrations in deep groundwater than those in shallow groundwater (Figure 9a,b).This indicates that the higher extent of denitrification in the deep groundwater than that in the shallow groundwater.Denitrification processes take place under reducing environments, usually in the saturated zone.This process is linked either to organic matter oxidation or sulfide oxidation [10,20,49].It is believed that denitrification followed by sulfide oxidation cannot occur in carbonated aquifers where pH is over 7 [50].Thus, DOC is the main energy source for microbial metabolism in the groundwater.The denitrification reaction involving carbon as electron donors can be described as follows: (1)  (1), the residual DOC and NO will become increasingly enriched in the heavy isotopes ( 13 C and 15 N) rather than in the initial DOC and NO In the subsurface, a variety of biogeochemical reactions occur with DOC attenuation in the water infiltration and flow processes [22], while denitrification is the dominant attenuation process of NO in groundwater [10].As shown in Figure 10, synergistic changes in δ 13 CDOC and δ 15 NNO3 values of the groundwater samples imply that variation of δ 15 NNO3 values are related to biogeochemical processes.It's reasonable to propose denitrification as the most possible process.SGW7 and SGW8, deviating from the positive relationship between δ 13 CDOC and δ 15 NNO3 values, may be due to the import of manure (δ 13 CDOC = 27.3‰,[51]) in the Zhengding and Beigaoying areas where agricultural inputs could occur.Several samples (SGW9, DGW1, DGW5 and SW4) stray from the denitrification trend and approach SGW7 and SGW8, indicating these samples are concurrently influenced by the manure pollution and denitrification.The denitrification process occurs in the surface water (SW2) which has several favorable conditions such as low concentration of dissolved oxygen (2.3 mg/L) [23,28,52] and high concentration of DOC (22.1 mg C/L).The surface water with relatively high δ 15 NNO3 values and low δ 13 CDOC values may be the result of preferential uptake of light nitrogen isotopes by algae, leading to enrichment of heavy nitrogen isotope in the residual NO [53][54][55].In general, the high δ 13 CDOC and δ 15 NNO3 values highlight that a limited extent of microbial denitrification occurred mainly in the deep groundwater and part of the shallow groundwater and surface water.The low concentration (typically < 5 mg C/L) and bioavailability of residual DOC may be one of the most important reasons for the low extent of denitrification in the study area [56].The possible sources of nitrate in the groundwater in this area include agricultural input including NO fertilizer, NH fertilizer and manure, domestic and industrial wastewater, soil organic N, and atmospheric deposition.The characteristic N and O isotopic compositions of these possible nitrate sources are displayed in Figure 8.
In this study, denitrification poses the most difficulties for simple applications of nitrate isotopes.However, according to the mixing model reported in the previous study [9,57], combining the denitrification trend (ɛN/ɛO ratio) with the theoretically-expected δ 18 ONO3 value could preliminarily deduce the initial δ 15 NNO3 values of the NO in the water samples.In the case of the NO of groundwater originating from soil organic N, manure, NH fertilizer and wastewater, the dominant process related to NO throughout the hydrochemical profile is nitrification.Generally, NO generated via microbial nitrification incorporates two oxygen atoms from water and one from atmospheric O2 [58,59].Therefore, the expected δ 18 O value of NO may be estimated as follows: (2) The δ 18 O signature of soil O2 is equivalent to that of atmospheric O2 which is about +23.5‰ [9].The δ 18 OH2O of shallow groundwater in the investigated aquifer system ranges from −8.5‰ to −6.0‰ (Table 2).Consequently, the expected δ 18 ONO3 value for groundwater NO − 3 derived from nitrification of alternative sources in soils should be in the range of +2.2‰ to +3.8‰.The theoretically-expected value is less than the measured δ 18 ONO3 values (+5.4‰ to +9.1‰).Theoretically, both denitrification and contribution of NO in precipitation to natural water can lead to the increase in δ 18 ONO3 values.In this study, if denitrification is the only process for the elevated δ 18 ONO3 values, the initial δ 15 NNO3 values of DGW samples should be extremely depleted (less than −20‰) which are deviated from the common values −10‰ to +20‰) [4] (Figure 8).Thus, the relatively high δ 18 ONO3 values of DGW samples cannot be simply explained by the denitrification process, suggesting the existence of other process.As shown in Figure 8, denitrification, coupled with mixing processes between precipitation and surficial sources (soil organic N, wastewater, manure and NH + 4 fertilizer) can be responsible for the higher δ 18 ONO3 values of the water samples.In addition, the δ 15 NNO3 value(s) of possible surficial source(s) can be roughly estimated between 2‰ and 5‰ according to the mixing model.The typical δ 15 NNO3 values of soil N ranged from 0 to 8‰ and NH fertilizers from −6‰ and +6‰ [4].In the SJZ area, the δ 15 NNO3 value of effluents N in the wastewater irrigated area was 2.5‰ [6], while the δ 15 NNO3 value of NO in the unsaturated zone was 5.7‰ [7].The reported values of δ 15 NNO3 ranged between +8.7‰ and 14.4‰ for the compost-applied area and between +4.5‰ and +8.5‰ for the area where urea was applied with compost [12].In general, soil N, NH fertilizer, wastewater, and the mix of NH fertilizer and manure could be possible NO sources.
For the shallow groundwater, soil organic N and precipitation cannot be the main origins of NO since NO concentrations of most samples are much higher than the natural NO concentration (4.4 mg/L [60]) (Figure 9a).Thus, NO in the shallow groundwater mainly derives from anthropogenic sources.In the non-wastewater-irrigated area, NH + 4 fertilizer could be the dominant NO source due to the initial δ 15 NNO3 values entirely falling in the range of NH + 4 fertilizer.In the wastewater-irrigated area, wastewater is the primary NO source.The relatively low NO /Cl − ratios of these samples collected from the wastewater-irrigated area can be another important piece of evidence for wastewater sources (Figure 11).For the deep groundwater, the samples with relatively low δ 18 ONO3 values might be contaminated by wastewater, while the samples with relatively high δ 18 ONO3 values could be attributed to a mixture of precipitation and soil organic N contribution (Figure 8).The mixing of different NO sources can be characterized by the pattern of NO /Cl − ratios and Cl concentrations (Figure 11).Agricultural inputs (high NO /Cl − ratios and low Cl concentration) are the dominant NO sources in the Zhengding and Beigaoying areas.The samples collected from the southern part of the study area are characterized by lower NO /Cl − ratios but high Cl − concentrations, confirming that long-term wastewater irrigation has a significant impact on the groundwater quality [6].The surface water and deep groundwater are both primarily originated from natural sources (precipitation and soil organic nitrogen) with little influence of anthropogenic activities, and show low NO /Cl − ratios and Cl − concentrations.However, the samples SW2 and DGW5 are polluted by wastewater and show increased Cl − concentration.All data points are included in a mixing triangle, with end members represented by wastewater, natural sources, and agricultural inputs.Several points located in the middle of the mixing triangle, suggesting varied sources, can jointly aggravate the nitrate pollution.

Conclusions
This study focuses on the sources and fate of NO in the groundwater used as drinking water in the rural and suburban areas of SJZ, China.
The dissolution of the calcium carbonate and gypsum minerals was the most effective process in the groundwater chemistry.Hydrochemical data suggested that NO in the groundwater could be related Denitrification affects the transformation of NO in the groundwater supported by the fractionation ratio (ɛN/ɛO) and the pattern of δ 13 CDOC and δ 15 NNO3.Our study clarifies that a low extent of denitrification could be ubiquitous in the groundwater, and the effect of denitrification in the deep groundwater is more significant than that in the shallow groundwater.
For the shallow groundwater, NO contamination is predominantly derived from NH + 4 fertilizers in non-wastewater irrigated areas; in contrast, the NO source can be mainly attributed to wastewater in wastewater irrigated area.For the deep groundwater samples, NO could primarily originate from precipitation and soil organic N. In addition, some of them were influenced by wastewater.In this study, the mixing processes of various NO sources are confirmed by dual isotopes (δ 15 NNO3, and δ 18 ONO3) and hydrochemistry (NO /Cl − ratio and Cl concentration) in the drinking water wells of SJZ.

Figure 1 .
Figure 1.Illustration of the study area and sampling points.(a,b) are location maps of the study area; and (c) sampling map.Line I-I' indicates location of hydrogeologic cross section in Figure 2. (SW: Surface water; SGW: Shallow groundwater; DGW: Deep groundwater).

Figure 2 .
Figure 2. The hydrogeologic cross section in Shijiazhuang.The range of study area is marked by rectangle with dashed lines.The Quaternary aquifers are divided into four groups named as Aquifers 1-4.

Figure 6 .
Figure 6.Plot of δ 2 H against δ 18 OH2O for surface and ground water samples from the Shijiazhuang area.Local meteoric water line (LMWL, solid line) and evaporation lines (dashed line and dotted line) are shown for comparison.The LMWL is defined by Jia et al. [45].(SW: Surface water; SGW: Shallow groundwater; DGW: Deep groundwater; WMIC: Weighted mean isotope composition).

Figure 8 .
Figure 8. Isotopic values of surface water and groundwater dissolved NO plotted with the ranges of the potential NO sources in the study area [4,9,48].(SW: Surface water; SGW: Shallow groundwater; DGW: Deep groundwater).

Figure 10 .
Figure 10.Plot of δ 13 CDOC and δ 15 NNO3 in surface water and groundwater.The rectangle with gray shadow indicates the synergic trend between δ 13 CDOC and δ 15 NNO3.The two ellipses with dashed lines highlight the values deviating from the trend due to mixing with different sources and assimilation by microbial activities.(SW: Surface water; SGW: Shallow groundwater; DGW: Deep groundwater).

2 .
Sources of NO in Groundwater

Figure 11 .
Figure 11.Variations of NO /Cl − molar ratios with Cl − molar concentrations of surface water and groundwater.The solid lines represent the theoretical mixing lines constructed by using the manure, municipal wastewater, and precipitation as end members.(SW: Surface water; SGW: Shallow groundwater; DGW: Deep groundwater).
or usage of fertilizers.The isotopic data (δ18 O and δ 2 H) indicated that the groundwater originated primarily from local precipitation and experiences different in extent of evaporation.

Table 1 .
Chemical compositions and field data of the surface water and groundwater in Shijiazhuang.

Table 2 .
Environmental isotopes data of the field sampling.