Introduction

Groundwater is an essential part of water resources for human survival and economic development in many regions over the world (Li et al. 2015). It is especially valuable in arid regions where there is only limited availability of precipitation and surface water resources (Li et al. 2016). Groundwater quality which relates closely to human health and economic development has become as important as its quantity due to the demand for safe water (Li et al. 2014). Nowadays, groundwater pollution has been reported in many areas around the world. The major contaminants include heavy metals (Hejabi et al. 2011), nitrate (Mahvi et al. 2005), etc.

In twentieth century, the generally high concentration of NO3 − in groundwater was found in many arid areas all over the world (Marret et al. 1990; Faillat and Rambaud 1991; Barnes et al. 1992). Excess levels of nitrate in groundwater can cause methemoglobin disease and cancer, posing an extreme threat to human health especially to the development of children’s health (Kendall and Aravena 2000; Knobeloch et al. 2000; Weyer et al. 2001). As an important material in the nitrogen cycle and widely participating in physical, chemical, and biological reactions in groundwater, nitrate has attracted extensive attention (Canter 1997; Galloway et al. 2003).

Arid and semi-arid area, which is prone to high concentration of nitrate under the condition of extreme drought, can serve as the main storage area for nitrogen (Walvoord et al. 2003; Smith et al. 2000; Schaeffer et al. 2003). Turpan-Hami area, or Tu-Ha area, located in the northwest arid areas of China, which is an important part of the Silk Road economic belt (Li et al. 2015), is the largest nitrate accumulation area in China (Qin et al. 2012a, b). Such high concentrations of nitrate in the groundwater can cause serious health risk to local residents. Therefore, it is of scientific and realistic significance to investigate and find out the distributions and sources of nitrate in groundwater for the water resources utilization and management.

In recent years, the nitrogen and oxygen isotopic techniques have been well developed and widely applied to trace and distinguish the source and formation conditions of nitrate (Deng et al. 2007; Wang 1997; Wigand et al. 2007; Lee et al. 2008; Mattern et al. 2011), which can largely avoid the uncertainty of a single isotope tracer. Many studies have shown that the determination of δ18O and δ15N in NO3 − is not only an effective tool to identify the different sources of NO3 −, but also a useful way for the study of migration and transformation of NO3 − in the groundwater (Seiler 2005; Widory et al. 2005; Umezawa et al. 2008). Recently, some studies focus on the in-depth research about the source of NO3 − in groundwater and zone of aeration in Xinjiang and Badain Jaran Desert by using isotopic techniques (Gates and BÖHLKE 2008; Ge et al. 2014; Qin et al. 2012a, b). These studies indicate that the sedimentation of atmospheric aerosol is the main source of NO3 − in the nitrate deposits in the Tu-Ha basin, Xinjiang (Qin et al. 2008, 2012a, b). However, the scant attention has been paid to the concentration and source of nitrate in groundwater in the Tu-Ha area. In present study, the hydrogeological, isotopic, and hydrogeochemical information were integrated in an attempt to explore the source, migration, and transformation of nitrogen in the groundwater in the Tu-Ha area. The data evaluation methods and results of this study could be useful to the protection and management of groundwater resources in Tu-Ha area and other arid and semi-arid areas.

Study area Description

Natural Geography

The study area, which was divided into two workspaces, Turpan area and Hami area, is located in the Turpan-Hami area, Xinjiang, China. The geographic coordinates are from 41°40′ to 42°40′ N and 90°00′ to 94°00′ E (Fig. 2). In the study area, the topography goes up from northwest to southeast. The peneplain hilly terrain is widely distributed in the south of the study area and the plain terrain covers large areas in the north. South Lake Gobi Desert, Geshun Gobi Desert, Big Depression, and Kumutage Desert are located in the study region. Hami City is 100 km away on the northeast of the study area; Aydingkol Lake and Turpan City are 200 km away to the northwest of the region.

The study area has the typical continental climate, and is one of the regions with most intense continental climate in the world. The earth’s surface, where only a small amount of drought-resistant plant and animal could grow, e.g., ephedra, hay, and antelope, is covered with broad gobi gravel. The region lacks rain and has a large difference in temperature between day and night. The annual average temperature is 9.8 °C and the annual rainfall is from 30 to 57.5 mm. The annual average evaporation is from 2245 to 2879 mm, and is much larger than the precipitation in the study region. Due to scarce rainfall, the surface runoff is not developed and the arid desert landscape exists everywhere in the region. Although there is no perennial river, the dry valleys are widely distributed and mostly formed by flood erosion. The river water which distributes in the mountainous area gradually disappears in the gobi after flowing into basin.

Hydrogeology

According to the geological structure, lithology, aquifer and hydrodynamic characteristics, etc., the groundwater in Hami area could be divided into four types: loose sediment pore water, clastic rock cranny pore water, carbonate fissure water, and bedrock fissure water. The area of loose sediment pore water is relatively large and mainly distributed in the intermountain depression and valleys area. The thickness of the loose debris can be up to 7 m. The lithology of clastic rock cranny pore water is mainly tertiary glutenite and mostly exist in the intermountain depression. Carbonate fissure water mainly distributes around Yamansu Iron ore and has a relatively high water content. Bedrock fissure water is the main type of groundwater in the Hami area, and it mainly includes weathering fissure water and structural fissure water. Within the depth of several meters below the surface, weathering fissures develop best and the development degree gradually weakens with the increase of depth. The development depth of rock weathering fissure water in the study area is from 10 to 30 m and only few areas can reach 100 m. Previous studies showed that the groundwater generally flows from southeast to northwest, based on the measurements of water depth in 30 wells in the Hami area (Guo et al. 2014, 2016). Combining the data of the water depth from the previous and present studies, the groundwater flow directions were roughly drawn (Fig. 1).

Fig. 1
figure 1

The hydrogeological map of Hami area

Full size image

Groundwater Sampling and Analysis

Sampling Point Layout

In the study area, totally 19 groups of samples were collected. The workspace in the southeast of Turpan basin is located in south of the Kumutage desert and north of the Kuruketag mountain. 8 groups of groundwater samples were collected in this workspace and the sample numbers were labeled from SS12 to SS19. There were 2 groups of spring water, 4 groups of shallow groundwater, and 2 groups of deep groundwater. The other workspace is located in the southeast of the Hami basin, as a typical study area. This area is located in the southeast of South Lake Gobi Desert and the south of Hami city. 11 groups of groundwater samples in this workspace were collected and the sample numbers were labeled from YMS01 to YMS11. Among them there were 3 groups of spring water, 7 groups of shallow groundwater, and 1 group of deep groundwater. We analyzed the concentrations of major ions, including Cl−, SO4 2−, CO3 2−, HCO3 −, Na+, K+, Mg2+, and Ca2+, for all the samples collected in the Tu-Ha area. The hydrogen and oxygen stable isotopes of 15 groundwater samples were measured, and nitrogen and oxygen stable isotopes of nitrate for 19 groups of samples were tested. The concentrations of NO3 −–N, NO2 −–N and NH4 +–N for 15 groups of samples were measured in the study area (Fig. 2).

Fig. 2
figure 2

Locations of the groundwater samples

Full size image

Isotopes and Hydrochemistry Analysis

The major ions of groundwater in this study were analyzed by ion chromatography (Anionic chromatography DX-120 and Cationic chromatography ICS900) and the data were assessed according to national standard of the People’s Republic of China: Quality standard for groundwater (GB/T 14848-93). A charge balance check showed that the error is <10 % for all samples except for one sample YMS08. The error rate of sample YMS08 is as high as 86 %. Due to high concentrations of Na+ that exceeded the detection limitation of instrument, Na+ was not measured in the SS16 sample. Therefore, the ions data of groundwater were not convincible for YMS08 and SS16 samples in this study. Liquid water isotope analyzer was used for the analysis of hydrogen and oxygen isotopes. The NO3 −–N, NO2 −–N, and NH4 +–N concentrations were determined by spectrophotometer according to “Standard methods for the examination of water and waste water”(State Environmental Protection Administration of China, 2002). The nitrogen and oxygen stable isotopes in NO3 − of groundwater were analyzed by isotope mass spectrometer (MAT 253) in the Institute of geographical science and resources, Chinese academy of sciences (the precision of δ15N–NO3 and δ18O–NO3 is ±0.3 ‰).

Data Analysis

By using the method of the relational graph of hydrogen and oxygen stable isotopes in groundwater, obtained by Origin 9.0 software (Seifert 2014), the recharge source and discharge of groundwater in Tu-Ha area were analyzed. The chemical composition of groundwater was studied by using the Piper diagram that was drawn by AqQA software (Geoffrey et al. 2004). The correlation of groundwater chemical characteristics was calculated by SPSS20 software and the ion contour maps of NO3 −–N, NO2 −–N, NH4 +–N, and TDS in the study area were drown by Surfer11 (Hasanah et al. 2013). Combining with the correlation and the ion contour map, the characteristics of distribution and the circulation of nitrogen in groundwater were discussed. The source characteristics of nitrate in groundwater in the study area were identified by the diagram of δ15N and δ18O in NO3 − (Origin 9.0).

Results And Discussion

A statistically correlation analysis was conducted to assess the correlation between the major ions. From the formation of correlation matrix, the correlation coefficient of each groundwater component can be clearly seen, and the NO3 −–N, NO2 −–N, and NH4 +–N sources can be inferred from the value of the correlation coefficient among the components. According to the correlation coefficient matrix of groundwater chemical composition (Table 1), NO3 −–N, NO2 −–N, NH4 +–N, and total nitrogen have higher correlation coefficient with each other and also have a higher correlation coefficient with the major ions compositions (e.g., Cl−, SO4 2−, Na+). Especially, the correlation coefficient is over 0.91 among NO3 −–N, NO2 −–N, NH4 +-N, and Cl−, TDS. The concentrations of NO3 −–N, NO2 −–N, NH4 +–N, and total nitrogen increase with the rising of TDS and the change of NO3 −–N, NO2 −–N, NH4 +–N is almost the same with the major ions compositions of groundwater. Meanwhile, the correlation coefficient among NO3 −–N, NO2 −–N, NH4 +–N and pH is very low. It indicated that the formation of high level nitrogen in groundwater, which is mainly affected by strong evaporation, is almost the same as that of high concentration of Cl−, Na+, and SO4 2−.

Table 1 Correlation coefficient (R) matrix of groundwater chemical composition in Tu-Ha area
Full size table

Distribution Characteristics of Nitrogen

The main hydrochemical components of groundwater in Tu-Ha area are Cl−, Na+, and SO4 2−, but it can be illustrated that the contents of NO3 −–N, NO2 −–N, and NH4 +–N are very high. In the groundwater samples of Tu-Ha area, the concentrations of NO3 −-N vary from 23.29 to 1819.49 mg/L, NO2 −–N vary from 0.006 to 102.25 mg/L and NH4 +–N vary from 0.02 to 2.53 mg/L. However, except for sample SS13, the hydrochemical types of the rest samples are invariant with the changes of NO3 −–N, NO2 −–N, and NH4 +-N concentrations in study area. According to the national standard of “the People’s Republic of China: Quality standard for groundwater” (GB/T 14848-93), the NO3 −–N concentration in all samples exceeds the standard of class III, equivalent to the drinking water standard (class III:5 mg/L<NO3 −–N ≤ 20 mg/L, class IV:20 mg/L<NO3 −–N ≤ 30 mg/L, class V:NO3 −–N>30 mg/L). Meanwhile, YMS02, YMS03, and YMS04 belong to class V, and the rest 12 groups of samples belong to class V. The NO3 −–N concentration of SS13 sample is the highest, being as much as 1819.49 mg/L. The site of SS13 is located in the local large granite mining area, and human activities may have potential influences on the groundwater flow and water quality, leading to the unusually high concentration of NO3 −–N in groundwater.

In the typical study area (Hami area), covering approximately 3700 km2, the flow direction of groundwater is roughly from southeast to northwest. TDS constantly increases from the southeast to the northwest along the groundwater flow direction. Meanwhile, the concentrations of NO3 −–N, NO2 −–N, and NH4 +–N also increase from southeast to northwest (Fig. 3). The enrichment of nitrate, nitrite, and ammonia has similar trend with TDS and all have the similar trends of rising along the groundwater flow direction.

Fig. 3
figure 3

Contour graph of Nitrogen and TDS in the Hami area: a Contour graph of TDS(g/L); b Contour graph of NO3 −–N(mg/L); c Contour graph of NO2 −–N(mg/L); d Contour graph of NH4 +-N(mg/L)

Full size image

The Isotopic and Chemical Characteristics of Groundwater

According to the diagram showing the relationship between hydrogen and oxygen stable isotopes of groundwater (δD and δ18O), we discussed the hydrogeochemical formation of groundwater. Compared with the Global Meteoric Water Line (GMWL) (Fig. 4), we found that all the points of groundwater samples lie in the lower right of the GMWL and are far away from the line. A linear regression line was fitted using the data of δD and δ18O of all the samples, which is approximately δD = 3.6δ18O–38.6 with R 2 = 0.924, representing an evaporation line. The concept of evaporation line comes from the fact that the evaporation will cause waters deviating from the GMWL and distributing along the lines with the slopes less than 8 (usually between 3.5 and 6 or from 4 to 6) (Friedman et al. 1962). The relatively enrichment of heavy isotopes in the groundwater could be caused by non-equilibrium evaporation process (Clark and Fritz 1997). The result indicated that the main recharge source of groundwater is from atmospheric precipitation and the groundwater in study area has experienced strong evaporation and concentration effects.

Fig. 4
figure 4

Plot of δD versus δ18O in Tu-Ha area

Full size image

It can be seen from the Piper diagram of groundwater chemistry (Fig. 5) that the main chemical types of groundwater are Cl·SO4–Na and Cl–Na (except SS13), confirming that groundwater chemistry is mainly influenced by strong evaporation in the study area. In arid and semi-arid area, the groundwater levels are relatively shallow and evaporation becomes the main discharge of groundwater. Under the effect of evaporation, the moisture of groundwater is constantly evaporated and salts with lower solubility precipitates with the continuation of time. In contrast, the salts with higher solubility would stay in the water (e.g., NaCl), and become the main components of groundwater with high TDS. As shown in Table 2, among the shallow groundwater, the chemical type of YMS01, YMS08, YMS09, and SS12 is Cl–Na, YMS02, YMS03, YMS04, and YMS07 is Cl·SO4–Na water. For the spring water, the type of YMS06, YMS08, and SS18 is also Cl·SO4–Na and YMS10 is Cl–Na water. In the deep groundwater, the type of YMS05 is SO4·Cl–Na, SS13 is NO3–Na and SS14 is Cl·SO4–Na water. Meanwhile, the type of groundwater chemistry is gradually from Cl·SO4–Na to Cl–Na along the direction of groundwater flow in the typical research area and the TDS ranges from 4 to 100 g/L (Table 2).

Fig. 5
figure 5

Piper graph of water chemistry in Tu-Ha area

Full size image
Table 2 Statistics of major ions and water depths, Ion concentration unit(mg/L)
Full size table

Source Characteristic of Nitrate

Cl− is widely distributed in various kinds of water body, and it is relatively stable. Due to the stability of Cl−, it has often been used as atmospheric precipitation inert tracer. The ratio of NO3 −–N/Cl− can standardize the influence extent of evapotranspiration and also can be used as a basic index for distinguishing the nitrate sources of atmospheric precipitation from other sources (Gates et al. 2008). According to the previous study in the northwestern region of China, the reference value of NO3 −–N/Cl− which could represent the value without human activities is 0.05–0.22 (Ma and Edmunds 2006, Ma et al. 2009; Li 2014; Pan 2014; Huang 2005; Dang et al. 2010). The figure shows that the concentration of NO3 −–N increases with the rise of concentration of Cl− (except SS13) (Fig. 6), while the NO3 −–N/Cl− remained at a low value ranging from 0.02 to 0.05 (Table 3). And besides SS13, there is high correlation between the source of NO3−– N and that of Cl−. The reason for the high NO3 −–N/Cl− ratio of SS13 (up to 21.6) is that SS13 is located in the quarry mining area, and large-scale human exploitation activities possibly lead to the increasing content of nitrate. Also it shows that except SS13 the main nitrate sources of groundwater samples in study area is the natural source (atmospheric sedimentation), not the non-natural sources produced by human activities.

Fig. 6
figure 6

Plot of NO3 −–N versus Cl− in Tu-Ha area

Full size image
Table 3 NO3 −-N, NO2 −-N, and NH4 +-N data of groundwater in Tu-Ha area,(mg/L)
Full size table

On the basis of the eigenvalue of nitrogen stable isotope (δ15N–NO3 −) and oxygen stable isotope (δ18O–NO3 −), the source of nitrate can be determined because the isotopes of different nitrate sources have different eigenvalues. In the atmospheric precipitation (dry precipitation and wet precipitation), the eigenvalue of δ15N–NO3 − ranges from −15 to +15 ‰ and the eigenvalue of δ18O–NO3 − ranges from +18 to +70 ‰. In the values of inorganic fertilizers, the δ15N–NO3 − ranges from −4 to +4 ‰ and the δ18O–NO3 − ranges from 17 to +25 ‰. In mineralization of soil organic matter (microbiological nitrification nitrogen), the δ15N–NO3 − ranges from +4 to +8 ‰ and the δ18O–NO3 − ranges from −10 to +10 ‰. In manure or sewage, the δ15N–NO3 − ranges from +8 to +22 ‰ and the δ18O–NO3 − ranges from +3 to +8 ‰ (Heaton 1990; Kendall 1998b, 2007; Choi et al. 2003). In 19 groups of the samples in Tu-Ha area, the δ15N–NO3 − ranges from −0.6 to +31 ‰ and mainly between +4.1 and +19.3 ‰, and the δ18O–NO3 − ranges from +16.3 to +37.4 ‰. The δ15N–NO3 − of most groundwater samples is within the range of atmospheric precipitation as previously shown. Except for SS15, the δ18O–NO3 − of the rest samples is within the range of atmospheric precipitation. Compared with the previous research which showed that the long sedimentation of atmospheric aerosol is the main source of NO3 − in the nitrate deposits in Tu-Ha basin (Qin et al. 2008, 2012a, b), we found that the values of δ15N–NO3 − and δ18O–NO3 − of groundwater samples are similar to that of the nitrate deposits in Qin’s research. According to the previous and the present research, the nitrate in Tu-Ha area is mainly derived from atmospheric precipitation, and local or regional isotope differences in nitrate deposits may be attributed to post-depositional processes, and the difference in aridity and associated microbial activities.

The previous studies have suggested a diagram of δ15N versus δ18O to trace the sources of nitrate in water by taking the nitrogen stable isotope as abscissa and the oxygen stable isotope as ordinate (Kendall et al. 1998; Kendall and McDonnell 1998; Kendall and Aravena 2000). Accordingly, the source of nitrate could be more accurately inferred via the diagram of δ15N and δ18O in the Tu-Ha area. Most samples but YMS01, YMS04, YMS05, YMS10, and SS19 in the Tu-Ha area are located in the nitrate distribution area caused by atmospheric precipitation (Fig. 7), indicating that the source of nitrate is mainly from atmospheric precipitation. According to the site investigation, there are no human activities and livestock, and wild animals are very scarce around YMS01, YMS04, YMS05, YMS10, and SS19 samples. Meanwhile it can also be proved from the mass-independent fractionation effect of oxygen isotope in the nitrate deposits of Tu-Ha area that the nitrate is produced by atmospheric photochemical reaction (Qin et al. 2008; Li et al. 2010). In the sample of SS15, which is located near the line of denitrification, it suggests that the phenomenon of local denitrification exists.

Fig. 7
figure 7

Plot of δ15N versus δ18O in Tu-Ha area (Kendall et al. 1998; Kendall and McDonnell 1998; Kendall and Aravena 2000)

Full size image

In the process of denitrification, the concentration of nitrate is low and δ15N–NO3 − is high because the 15N–NO3 − becomes abundant in the rest of nitrate after 14N–NO3 − participates in the reaction. Meanwhile, it also causes the fractionation of oxygen stable isotope in NO3 − and the enrichment of 18O–NO3 − in the rest of the nitrate. According to the analysis of the concentration relationship between δ15N, δ18O, and NO3 −–N, the denitrification can be clearly judged. If δ15N and δ18O increase with the decrease of nitrate concentration and a linear relationship presents between δ15N, δ18O, and the rest of ln [NO3 −–N], denitrification will occur in the aquifer of study area (McKeon et al. 2005; Chen et al. 2006). Therefore, the changes of δ15N–NO3 − and 18O–NO3 − could be used as an indicator of denitrification. It shows that the changes of δ15N–NO3 −, 18O–NO3 −, and the rest of ln [NO3 −–N] do not present a linear relationship (Figs. 8, 9), and the δ15N–NO3 − and 18O–NO3 − do not increase according to the proportion of 1.3:1 to 2.1:1 (Aravena and Robertson 1998; Böttcher et al. 1990) (Fig. 7). The DO concentrations of groundwater change between 1.82 and 3.73 mg/L, and most of them are higher than 2 mg/L which is not suitable for denitrification. Thus, we can speculate that denitrification does not take place in Tu-Ha area.

Fig. 8
figure 8

Plot of δ15N versus ln [NO3 −–N]

Full size image
Fig. 9
figure 9

Plot of δ18O versus ln [NO3 −–N]

Full size image

When NH4 +–N transfers to NO3 −–N through the nitrification in groundwater, 1/3 of oxygen in demand is from atmosphere and the rest 2/3 of oxygen is from the oxygen of nitrification environment (generally refers to the water)(Andersson and Hooper 1983; Hollocher 1984; Mariotti et al. 1981). Therefore, the enrichment degree of 18O in the nitrate can be used to determine whether the nitrification happens, because the δ18O in the nitrate formed by nitrification is closed to the δ18O in local groundwater. Due to the relative stability of oxygen isotopic composition in the atmosphere, the δ18O in nitrate formed by nitrification mainly depends on the oxygen isotopic composition of local groundwater (Andersson and Hooper 1983; Mariotti et al. 1981). Therefore, the oxygen isotopic composition in nitrate formed by microbiological nitrification can approximately be calculated using the following equation: δ18O–NO3 − = 2/3(δ18O–H2O) + 1/3(δ18O–O2). The value of δ18O-O2 is +24 ‰ (Hollocher 1984; Durka et al. 1994) and the δ18O-H2O of groundwater in study area ranges from −9.2 to +0.6 ‰. The δ18O–NO3 −of groundwater in Tu-Ha area calculated by the formula ranges from +1.87 to +8.4 ‰. However, the actual value of δ18O–NO3 − in study area ranges from +16.3 to +37.4 ‰, inconsistent to the calculated values of δ18O–NO3 −. Many studies have shown that the optimum pH range of nitration ranges from 6.4 to 7.9 and either higher or lower value of pH will have different influences on the activity of nitrifying bacteria which can destroy the balance of nitration. In this case, the alkaline water or the acid water will retard the nitrification (Qiu et al. 1997). The pH of groundwater in study area ranges from 8.34 to 9.62 which is far beyond the optimal pH for microbiological nitrification. Meanwhile, the most suitable ORP of nitrification is higher than 250 mV. However, the ORP of groundwater in study area is 20~150 mV. To sum up, it can be determined that the nitrate of groundwater is not formed by microbiological nitrification in the study area.

Conclusions

In this study, we explored the hydrochemical characteristics and enrichment of nitrogen in groundwater in the Tu-Ha area by using the hydrochemical, hydrogen, oxygen, and nitrogen stable isotopes methods. The following conclusions could be drawn:

In the Hami area, the main recharge source of groundwater is from atmospheric precipitation and the groundwater has experienced strong evaporation in study area. The hydrochemistry type is gradually from Cl·SO4–Na to Cl–Na along the groundwater flow direction, and the concentration change of nitrate, nitrite, ammonia is consistent with that of TDS.

The formation of high level nitrogen in groundwater is consistent to that of Cl−, Na+, and SO 2−4 , which is mainly affected by strong evaporation. There is no obvious transformation between nitrate, nitrite, and ammonia of groundwater in the study area. The NO3 −–N/Cl− ratio of most of samples ranges from 0.02 to 0.05, indicating that the nitrate of groundwater samples in the study area is mainly from natural sources (atmospheric sedimentation), not from the non–Natural sources produced by human activities.

The δ15N–NO3 − ranges from −0.6 to +31 ‰ and mainly in a range of +4.1 to +19.3 ‰, the δ18O–NO3 − ranges from +16.3 to +37.4 ‰. The δ15N–NO3 − and δ18O–NO3 − of most groundwater samples are within the range of atmospheric precipitation. The source of nitrate could be inferred mainly from atmospheric precipitation, and the nitrate in groundwater is not formed by denitrification or microbiological nitrification in the study area.