Distribution Characteristics and Influencing Factors of Uranium Isotopes in Saline Lake Waters in the Northeast of Qaidam Basin

Four saline lakes in the northeast of Qaidam Basin were selected to explore the distribution characteristics and influencing factors of uranium isotopes in lake waters with high evaporation background. The 238U concentration and the activity ratios of 234U/238U ([U/U]AR) showed that there was no significant change in the same lake, but there was a certain degree of difference in the distribution between different lakes. We found that aqueous 238U concentration within a certain range increased with an increase in TDS (total dissolved solid) and salinity, as was also the case with pH. As in natural waters, the pH affects the speciation of 238U, but TDS and salinity affect the adsorption process of aqueous 238U. Further, the replenishment of water will also affect the uranium isotope concentration for lakes, but it is not the main influencing factor for saline lakes. Therefore, we suggest that pH is the dominant factor affecting changes in aqueous 238U concentration of the sampled saline lakes. The [U/U]AR in these saline lakes are closely related to the input water and the associated water–rock interactions involving sediments, atmosphere dust, and organic material, etc. during the evolution stage, metamorphous degree, and hydrochemistry of the saline lakes. Lake water samples collected in the maximum and minimum discharge water period, were used to evaluate the seasonal distribution characteristics of aqueous 238U, and we found that 238U concentration did not show an evident change with the seasons in these saline lakes. If the 238U concentration and [U/U]AR can remain consistent during a period of time, then the sediment ages and/or sedimentation rates could be determined by lake sediment and/or biogenic carbonate in future, thus allowing for the accurate reconstruction of the paleoclimate and paleoenvironment.


Introduction
In natural conditions, uranium (U) mainly exists as three isotopes: 234 U, 235 U, and 238 U. Among them, 235 U and 238 U have long half-lives with high abundance, and the content ratio of about five orders of magnitude lower than that of 238 U [2]. Uranium isotopes are widely used as a typical tracer in the dating of various oceanographic processes and in the study of sources of dissolved solutes in river basins [3][4][5][6][7][8].
Studies of the distribution characteristic of uranium isotopes have mainly focused on glacial sedimentation at the two poles and oceanic circulation [9,10] and recent research has still mainly focused on the oceans. Existing research results show that the uranium concentration of seawater has a relatively consistent, linear relationship with salinity [11,12]. Some scholars have tried to convert oceanographic methods to inland areas, but inland research has focused on rivers and estuaries [13][14][15], river bends [16], and lakes [17], while very few studies have explored the uranium content and uranium isotopes in saline lakes.
Limited studies on hydrological systems in saline lakes have been conducted and research methods have been systematically developed. The research dates back to Osmond, et al. [18] who established a two-end mixing model based on the relationship between 234 U/ 238 U, 1/U, and U/TDS (total dissolved solids) in the water of Sambhar lake. They found that one end of the lake had a low uranium content and low salinity, while the other end had the opposite. Yadav et al. [19]. explored the U/TDS relationship in different aqueous media (including river water, lake water, and inter-crystal brine) of the Sambhar saline lake during different periods of the year, finding that there was a strong linear relationship between aqueous U concentration and TDS, which was consistent with the two-end mixing model established by Osmond et al. [18]. Moreover, Yadav, et al. [19] mentioned that the seasonal distribution difference of uranium isotopes was mainly controlled by the process of water dilution and evaporation. Borole et al. [20] studied the two major water systems of Narbada and Tapti in the west coast of India and showed that the abundance ratio of the aqueous uranium content to the sum of the main cations was similar to that found in typical crustal rocks. Due to the process of evaporation and solute concentration in saline lakes, the aqueous uranium content is generally higher than that in freshwater lakes.
Reports of saline lakes in high uranium environments are limited. Since the 1950s, Zheng [21] has carried out extensive and in-depth research on saline lakes in the Qinghai-Tibet Plateau. Zhang [22] found that, in the mining of lithium, boron, potassium, and magnesium from the Qaidam Basin, both uranium and thorium could be enriched, meaning that uranium enrichment is one of the basic characteristics of the saline lakes of northwestern China. The concentration of uranium of soda lakes in Mongolia is 57~15,000 µg/L [23], 320 µg/L in Gasikule salt lake in Qaidam Basin [24], and 13.4 µg/L in Qinghai Lake [25]. The potential mechanism of high uranium content in a saline lake has also been explored. For example, Benjamin et al. [23] and Isupova et al. [26] found that there exist high concentrations of uranium(U) in soda lake basins with groundwater-fed hyperalkaline. Hence, uranium-rich groundwaters are concentrated by evaporation and the U(VI) is chelated by CO 2− 3 to form the highly soluble UO 2 (CO 3 ) 4− 3 in water, which is the main cause of the enrichment of uranium content in lakes in northwestern Mongolia. Strakhovenko et al. [27] showed that biogeochemical processes have an impact on the mobility of uranium species. Andersson et al., [28] showed that the behavior of colloids suspended phase and colloids are beneficial for the removal of U at very low salinities. Koch-Steindl and Pröhl, [29] indicate that the mobility of U is influenced by sorption and complexation processes on inorganic soil constituents such as clay minerals, oxides, organic matter, and biological fixation and transformation. Fredrickson et al. [30] indicated that the Mn (III/IV) oxides in particular may significantly inhibit the microbial reduction of U(VI) and its subsequent precipitation as UO 2 (s). Belli et al. [31] thought that a certain number of OM and/or competing cations of iron (such as Ca 2+ and Mg 2+ ) may inhibit the adsorption effect of uranium complex on iron hydroxide.
In high-salinity environments, such high uranium features make it possible to extract uranium resources from saline lake brine. In addition, no matter when it is processed, boron lithium potassium salt products or waste liquid discharge will affect the product quality or pollute the environment.
Thus, it is necessary to determine the distribution characteristics of aqueous uranium isotopes in lakes with different salinity gradients in order to understand the influencing factors of uranium isotope distribution in lake waters. Previous research has shown that ion concentrations, TDS, salinity, pH, and biogeochemical process are factors that influence the aqueous uranium content distribution and the [ 234 U/ 238 U] AR . For this study, we selected several lakes of different salinities with low human disturbance in the northeast of Qaidam Basin of the Qinghai-Tibet Plateau for water chemistry (anions, cation, TDS, salinity, and pH) and uranium isotope analysis to gain insight on the overall distribution pattern and possible seasonal variations in the distribution characteristics of uranium in the saline lakes of China. In particular, we wanted to gain insights into the controlling factors and underlying mechanism of aqueous uranium concentrations in the lakes and to explore the potential environmental applications of uranium isotopes with the goal of promoting the widespread use of radioactive isotopes as environmental indicators.

Sampling Locations
There are many lakes in the Qaidam Basin and most of them are saline lakes. The water samples in this study were collected from four lakes of different salinities in the northeast of Qaidam Basin: Gahai lake (GH), Keluke lake (KLK), Tuosu lake (TS), and Xiligou lake (XLG). In order to further explore the influence of water sources on the uranium isotope of KLK lake, we obtained some water samples from the Bayin river (BY). These sampling locations are shown in Figure 1. Two sampling sites were established in XLG: one (XLG 1) in the center of the lake, and the other (XLG 2) in a shallow pool next to the lake, with the pool water mainly originating from natural precipitation with water quality indicators similar to those of KLK. The XLG 2 pool samples can therefore be regarded as freshwater lake samples. All five sampling sites could be roughly divided into two categories: freshwater lake and saline lakes, with the former consisting of KLK and XLG 2, and the latter GH, TS, and XLG 1. The study area belongs to the typical alpine continental climate, characterized by a large temperature difference between day and night, low precipitation, and high drought throughout the year. The main vegetation types are alpine meadows and alpine steppes. The overall elevation of the area is about 3000 m.
Minerals 2020, 10, x FOR PEER REVIEW 5 of 21 covered with ice coverage. Thus, these sampled lake waters could show typical saline lake water uranium distribution characteristics. The surface water dozens of meters off the shore was collected into HDPE (High Density Polyethylene) plastic bottles (Nalgene 2002-0032) from the depth of 10 cm below surface water by hand with Derma Free ® Vinyl Gloves. We collected other samples from various depths using PMMA (Polymethyl Methacrylate) water collectors and placed into HDPE plastic bottles (Nalgene, 2002-0032). After collection, about 1000 mL of each water sample was filtered through a mixed cellulose filter membrane (0.8-μm pore size, Millipore AAWP04700) with a manual vacuum filter system According to data from the National Meteorological Station of Delhi city, Han [32] determined that the mean annual precipitation over the period of 1961-2017 was 185.1 mm, and the mean annual evaporation ranged between 2242.8-2439.4 mm. Precipitation is uneven during the year, showing a unimodal distribution. Maximum precipitation occurs in June-July, and the minimum precipitation occurs in December. The total area of GH is about 37 km 2 with a mean water depth of 8 m and a maximum water depth of 15 m. The lake basin has no perennial surface rivers, relying on atmospheric precipitation and subsurface runoff to supply water. TS has a total area of about 180 km 2 and a maximum water depth of 25.7 m, with the water mainly originating from the drainage of the TS waters, atmospheric precipitation, and groundwater [21,24].
The Bayin river originates from the northern foot of the ZongWu Long Mountain which has a total length of 188 km and a drainage area of 7281 km 2 . The main water supply source in the upper reaches of the Bayin river is natural precipitation, and in lower reaches is groundwater. After the Bayin river flows out of the ZongWu Long Mountain, a great deal of river water infiltrates into the ground water because the sediments on the riverbed in the Gobi Desert are mostly gravel. Then, the river passes through Delhi city. The infiltration process of river water was prevented by the hardening of the riverbed in Delhi city. After the river flows out of the city, the infiltration process of the river recommences. When the groundwater flows near a tree, a large amount of groundwater overflows into a pool of springs and becomes the main supply water source of the Bayin river, because of the lithological thinning and the blocking effect of the Denan hills. Then, the river water and the groundwater flow into the Keluke lake. Eventually, the Bayin river leaves Keluke lake and flows into Tuosu lake [33,34].

Sampling Design and Methods
The samples were collected in three batches. The analysis of the first batch of samples showed no significant differences in the horizontal or vertical distribution of aqueous 238 U within the same lake, indicating a homogenous distribution of 238 U in the water column, such kind of phenomenon has also been showed in the published Qinghai Lake data [25]. Based on this feature, we suggest that one single sample was deemed representative of the whole water column. Therefore, we collected only limited samples in the second and third of batch samplings. The sampling information of each sample was presented in Table 1.
The sampling area is shown in Figure 1. We selected these sampling periods because water inputs northeast of the Qaidam Basin exists in greater variation, and during other seasons are almost covered with ice coverage. Thus, these sampled lake waters could show typical saline lake water uranium distribution characteristics.
The surface water dozens of meters off the shore was collected into HDPE (High Density Polyethylene) plastic bottles (Nalgene 2002-0032) from the depth of 10 cm below surface water by hand with Derma Free ® Vinyl Gloves. We collected other samples from various depths using PMMA (Polymethyl Methacrylate) water collectors and placed into HDPE plastic bottles (Nalgene, 2002-0032). After collection, about 1000 mL of each water sample was filtered through a mixed cellulose filter membrane (0.8-µm pore size, Millipore AAWP04700) with a manual vacuum filter system (Nalgene, 300-4100; Nalgene, 6133-0010), and the filtrate was stored for later analysis.

Analytical Methods
Samples analysis for water chemistry parameters (anions, cations, salinity, TDS, and pH) were carried out at the Key Laboratory of Surface System and Environmental Carrying Capacity of Shaanxi Province, Northwest University. Anion and cation concentrations of the samples were determined using two Dionex AQUION ICs [35,36]. Conductivity (ms/cm), TDS (g/L), salinity (% ), and pH were determined using a Mettler-Toledo Five Easy Plus pH meter. Uranium isotopes in the samples were measured at the Xi'an Jiaotong University Isotope Laboratory. Methods of investigation of uranium isotopes mainly include chemical separation and instrument analysis. The chemical separation process mainly includes the digestion process, centrifugal coprecipitation, ion exchange resin separation, uranium isotope collection, and purification. The steps involved are as follows. To analyze the U isotopic compositions of the water, the acidified aliquots of each sample were spiked with 233 U-236 U-229 Th, preconcentrated and digested with HNO 3 and HClO 4 , and then co-precipitated with Fe oxyhydroxide. The Fe precipitate was moved to a centrifuge tube for centrifugation and rinsed with deionized H 2 O (>18 MΩ) to eliminate the large seawater ions. The precipitate was then dissolved in 14N HNO 3 (1N = 1M) and moved to a Teflon beaker, before being dried and dissolved in 7N HNO 3 for anion-exchange chromatography with AG1-X8, 100-200 mesh size resin and a polyethylene frit. Separation was perfromed on Teflon columns with a 0.5 mL column volume (CV). The first separation was performed on Teflon columns with 4CV 7N HNO 3 , 3 CV of 8N HCl (to get rid of Fe and Th fraction), and 3 CV of deionized H 2 O (to gather the U fraction). The U fractions were then dried with 2 drops of HClO 4 and brought to volume with 7N HNO 3 . The final U fractions were then dried with 2 drops of HClO 4 and dissolved in weak nitric acid for analysis on the mass spectrometer. The concentrations of 234 U, 235 U, and 238 U were calculated by isotope dilution using the nuclide ratios determined on a Thermo-Finnigan Neptune mass spectrometer. All measurements were performed using a peak-jumping routine in ion counting mode on the discreet dynode multiplier behind the retarding potential quadrupole. Each sample measurement was bracketed by the measurement of an aliquot of the run solution, which was utilized to adjust for the instrument background count rates on the measured masses. For U, the measurement uncertainties involved propagated errors from ICP-MS isotope ratio measurements, spike concentrations, and blank corrections. The procedural blanks for chemical and mass spectrometric analyses at the Laboratory of Isotope Geochemistry in Xi'an Jiao Tong University are approximately 830 ag (2.2 × 10 6 atoms) for 234 U, 29 fg (7.7 × 10 7 atoms) for 235 U, 2 pg (5.2 × 10 9 atoms) for 238 U. The methods used here have been fully described by Cheng et al. [1,37] and Shen et al. [38,39].

Results
The water chemistry results of each sampled lake are presented in Table 2, including the aqueous uranium isotope concentration and the [ 234 U/ 238 U] AR . Given that 238 U is the dominant uranium isotope, accounting for 99.275% of the total U content, the uranium concentration in this paper refers to the 238 U concentration, expressed as 238 U hereafter. In general, freshwater lakes (KLK, XLH 2) are lower than saline lakes (GH, TS, XLG 1) for all measured parameters.
(a) Given that the salinity in the GH waters was above the detection limit of the instrument, we estimated these GH samples salinity by converting the measured conductivity as described below, resulting in an estimated salinity of 72.36% for the GH samples.
Conductivity can be converted to salinity between 0-40 • C using the following conversion formula: where y NaCl is the salinity calculated as NaCl, x is the conductivity in µs/cm, and t is the water temperature in • C. Salinity calculated using the above formula is given in units of ppm, and must be subjected to a basic unit conversion for expression in % .
(b) These contents of HCO − 3 and CO 2− 3 have been calculated by these two formulas [40].

Water Chemistry Characteristics in the Northeast of Qaidam Basin
As shown in a Piper diagram (Figure 2), the cations in these lake waters of the northeast of Qaidam Basin were mainly Na + , jointly accounting for 78.5%, 54.5%, 73.6%, 72.4%, and 71.0% of the total cations (TZ + ) in GH, KLK, TS, XLG1, and XLG2 lakes, respectively. Na + and K + are followed by Mg 2+ with the second highest concentration, while Ca 2+ had the lowest cation concentrations measured. Of note, the XLG 2 waters had a slightly higher concentration of Ca 2+ than Mg 2+ . The anions were mainly Cl − , accounting for 76.5%, 52.2%, 63.4%, and 67.9% of all anions (TZ − ) in GH, KLK, TS, XLG1, and XLG2 lakes, respectively. Cl − was followed by SO 2− 4 with the second highest concentration, while HCO − 3 , CO 2− 3 , NO − 2 , and NO − 3 jointly had the lowest anion concentrations measured. The concentration of an ion (or ion pair) remained essentially the same throughout the same water body.

Water Chemistry Characteristics in the Northeast of Qaidam Basin
As shown in a Piper diagram (Figure 2), the cations in these lake waters of the northeast of Qaidam Basin were mainly Na + , jointly accounting for 78.5%, 54.5%, 73.6%, 72.4%, and 71.0% of the total cations (TZ + ) in GH, KLK, TS, XLG1, and XLG2 lakes, respectively. Na + and K + are followed by Mg 2+ with the second highest concentration, while Ca 2+ had the lowest cation concentrations measured. Of note, the XLG 2 waters had a slightly higher concentration of Ca 2+ than Mg 2+ . The anions were mainly Cl − , accounting for 76.5%, 52.2%, 63.4%, and 67.9% of all anions (TZ − ) in GH, KLK, TS, XLG1, and XLG2 lakes, respectively. Cl − was followed by SO

Distribution Characteristics of Uranium Isotopes in These Lake Waters in the Northeast of Qaidam Basin
Multiple sampling sites were set at different depths in KLK and TS so that the horizontal and vertical distribution characteristics of 238 U could be analyzed for these lakes. The horizontal

Distribution Characteristics of Uranium Isotopes in These Lake Waters in the Northeast of Qaidam Basin
Multiple sampling sites were set at different depths in KLK and TS so that the horizontal and vertical distribution characteristics of 238 U could be analyzed for these lakes. The horizontal distribution characteristics of the 238 U concentration and the [ 234 U/ 238 U] AR in the KLK and TS waters are shown in Table 3.
Aqueous 238 U concentration of the freshwater lakes in the northeast of Qaidam Basin were generally low. In contrast, the concentration of 238 U in these saline lake waters were generally higher. There is no significant difference in the [ 234 U/ 238 U] AR between freshwater lakes and saline lakes, with the values generally ranging from 1.460 to 1.798. A comparison of all five sampled lakes data revealed that 238 U (TS) was the highest of all, while [ 234 U/ 238 U] AR (TS) was the lowest. There was no significant horizontal distribution difference among the three sampling sites for the 238 U concentration and the [ 234 U/ 238 U] AR .  Table 4. Three sampling sites were set in KLK. At sampling sites 1 and 2, water samples were collected at three depths referred to as T (surface), M (middle), and B (bottom) from the surface water to the bottom water, while only two sampling depths (T and M) were collected at sampling site 3.
In summary, in natural lakes, in the absence of interference from other aqueous factors, the 238 U concentration increased with an increased water depth, while the [ 234 U/ 238 U] AR showed an initial increase, then decreased gradually with depth. However, the ranges of these changes were very small, indicating that aqueous 238 U concentration and the [ 234 U/ 238 U] AR were essentially evenly distributed vertically and horizontally within the same lake.

Influence of Water Sources Supply on the Uranium Isotopes of Lake
In order to further explore the influence of water sources supply on the uranium isotope of lake, we get some water samples from different sections of the Bayin river (BY) and KLK at the same single day, and the Bayin river is the only incharge river of KLK lake. Results of 238 U concentration and [ 234 U/ 238 U] AR in the BY river and KLK waters are shown in Table 5. We found that the 238 U concentration and [ 234 U/ 238 U] AR in different sections of BY river are not the same. The 238 U concentration is highest in the upper reaches (6.0 µg/L) and the lowest in the lower reaches (2.8 µg/L) of natural waters. However, [ 234 U/ 238 U] AR shows the opposite of the distribution of 238 U concentration. The Bayin river passes through Delhi city. We collected the highest 238 U concentration samples (6.2 µg/L) from the watercourse of Delhi city.
The Bayin river flows into the KLK from the northeast and mixes with the original lake water. Two sampling results show that the average 238 U concentration of KLK lake water is about 5.5 µg/L, which is slightly lower than the upper reaches of the Bayin River and is significantly higher than the lower reaches. The Bayin river flows out the KLK from the southwest. The 238 U concentration of watercourse connecting KLK and TS is 5.6 µg/L. This is basically consistent with that of KLK lake water.

Seasonal Distribution Characteristics of Uranium Isotopes
We explored the seasonal distribution characteristics of U in these lake waters in the northeast of Qaidam Basin. The results from the first batch of samples (December 2017), with the minimum recharge water volume in winter, indicated that the aqueous 238 U concentration of the same lake remained at a relatively stable level. However, it is unclear whether the contribution of exogenous inputs (e.g., atmospheric precipitation and river inputs) could affect the lake water 238 U concentration and the [ 234 U/ 238 U] AR and the effects of exogenous inputs should be clarified before interpreting the environmental implications of the above indicators. Therefore, GH and TS were selected for a second round of sampling in summer with the maximum recharge water volume and measurement in order to conduct a seasonal comparison of the aqueous 238 U concentration and the [ 234 U/ 238 U] AR and the results are shown in Table 6.
We found the 238 U concentration of GH and TS changed slightly with the season. In winter (December), 238 U concentration of GH is 18.4 µg/L, in summer (June), the result is 16.2 µg/L. We also see the same pattern in TS, where 238 U concentration of TS decreased from 35.5 µg/L to 34.7 µg/L in from winter to summer. Whether it is GH or TS, the [ 234 U/ 238 U] AR in different seasons is approximative. From winter to summer, the maximum discharge water volume variation, the [ 234 U/ 238 U] AR of GH vary from 1.792 to 1.798, and the result of TS vary from 1.463 to 1.462.

Factors Influencing the Concentration Distribution Characteristics of 238 U in These Lake Waters in the Northeast of Qaidam Basin
Based on the existing literature, we hypothesized that TDS, salinity, and pH may account for the differences in the spatial distribution of aqueous 238 U concentration. The factors that are likely to affect the distribution of 238 U concentration in these lake waters in the northeast of Qaidam Basin are addressed in more detail as follows. For the convenience of discussion, we will use the parameters measured in KLK and TS as representative of the freshwater and saline lakes of the region, respectively.

Response Relationship of 238 U Concentration to TDS and Salinity in These Lake Waters in the Northeast of Qaidam Basin
Water Chemistry parameters of these lake waters show that the major cation is mainly Na + and the major anion is mainly Cl − . This means that TDS and salinity may be similar in the saline lake from the northeast of Qaidam Basin. As shown in Figure 3, the relationship between TDS and salinity of these lake waters clarifies this perspective. Minerals 2020, 10, x FOR PEER REVIEW 12 of 21 Therefore, we selected TDS as the research object to investigate the response relationship of 238 U concentration in these lake waters to TDS. The relationship of 238 U concentration to TDS in these lake waters in the northeast of Qaidam Basin is shown in Figure 4. When it comes to the relationship of 238 U concentration to TDS in these lake waters in the northeast of Qaidam Basin, in the absence of interference from exogenous factors, we found a higher lake water TDS in a certain range would lead to a higher concentration of 238 U. This indicates that the enrichment process of uranium involves a series of evaporation and concentration steps. We could speculate that the migration and the ability of uranium to transform in water increases with an increase of the TDS. Shi [41] suggested the main reason for this phenomenon is that 238 U is difficult to be adsorbed in an environment with high TDS. The mobility of U is largely controlled by their ability to form complexes both with inorganic (F − , Cl − , PO 3− 4 , CO 2− 3 ) and organic ligands. This has been described in detail for U [42,43]. Studies in the ocean [44] and the Jiu long River Estuary [45] also showed that the aqueous 238 U concentration increased linearly with an increase in TDS. Therefore, we selected TDS as the research object to investigate the response relationship of 238 U concentration in these lake waters to TDS. The relationship of 238 U concentration to TDS in these lake waters in the northeast of Qaidam Basin is shown in Figure 4. Therefore, we selected TDS as the research object to investigate the response relationship of 238 U concentration in these lake waters to TDS. The relationship of 238 U concentration to TDS in these lake waters in the northeast of Qaidam Basin is shown in Figure 4. to TDS in these lake waters in the northeast of Qaidam Basin, in the absence of interference from exogenous factors, we found a higher lake water TDS in a certain range would lead to a higher concentration of 238 U. This indicates that the enrichment process of uranium involves a series of evaporation and concentration steps. We could speculate that the migration and the ability of uranium to transform in water increases with an increase of the TDS. Shi [41] suggested the main reason for this phenomenon is that 238 U is difficult to be adsorbed in an environment with high TDS. The mobility of U is largely controlled by their ability to form complexes both with inorganic (F − , Cl − , PO 3− 4 , CO 2− 3 ) and organic ligands. This has been described in detail for U [42,43]. Studies in the ocean [44] and the Jiu long River Estuary [45] also showed that the aqueous 238 U concentration increased linearly with an increase in TDS. interference from exogenous factors, we found a higher lake water TDS in a certain range would lead to a higher concentration of 238 U. This indicates that the enrichment process of uranium involves a series of evaporation and concentration steps. We could speculate that the migration and the ability of uranium to transform in water increases with an increase of the TDS. Shi [41] suggested the main reason for this phenomenon is that 238 U is difficult to be adsorbed in an environment with high TDS. The mobility of U is largely controlled by their ability to form complexes both with inorganic (F − , Cl − , PO 3− 4 , CO 2− 3 ) and organic ligands. This has been described in detail for U [42,43]. Studies in the ocean [44] and the Jiu long River Estuary [45] also showed that the aqueous 238 U concentration increased linearly with an increase in TDS.
The results of the KLK and TS samples are compared in Figure 5. The results of the KLK and TS samples are compared in Figure 5. These results indicate that TDS are not the only factors affecting 238 U concentration in the waters of terrestrial lakes. The other possibility is there is a certain threshold. When above a TDS range, the 238 U concentration will decrease with the increase of TDS [22]. This should be the reason for GH with high TDS but low 238 U concentration. In summary, the positive relationship of aqueous 238 U concentrations to the factors of TDS may exist only when TDS are lower than certain thresholds, and if the impacts of other factors are deemed negligible. In other words, we speculate that, within a certain range of TDS, the aqueous 238 U concentration increases with the increase of TDS, but this will decrease rather than increase once beyond the range. More research is required to determine the veracity of this speculation, and to understand the underlying mechanism of thresholds for TDS that affect aqueous 238 U concentrations.

Response Relationship of 238 U Concentration to pH in These Lake Waters in the Northeast of Qaidam Basin
Most of these lakes in the northeast of Qaidam Basin are saline lakes with a pH generally above 7.0 (i.e., basic). As shown in Figure 6, the 238 U concentration in these lake waters in the northeast of Qaidam Basin increased with pH. These results indicate that TDS are not the only factors affecting 238 U concentration in the waters of terrestrial lakes. The other possibility is there is a certain threshold. When above a TDS range, the 238 U concentration will decrease with the increase of TDS [22]. This should be the reason for GH with high TDS but low 238 U concentration. In summary, the positive relationship of aqueous 238 U concentrations to the factors of TDS may exist only when TDS are lower than certain thresholds, and if the impacts of other factors are deemed negligible. In other words, we speculate that, within a certain range of TDS, the aqueous 238 U concentration increases with the increase of TDS, but this will decrease rather than increase once beyond the range. More research is required to determine the veracity of this speculation, and to understand the underlying mechanism of thresholds for TDS that affect aqueous 238 U concentrations.

Response Relationship of 238 U Concentration to pH in These Lake Waters in the Northeast of Qaidam Basin
Most of these lakes in the northeast of Qaidam Basin are saline lakes with a pH generally above 7.0 (i.e., basic). As shown in Figure 6, the 238 U concentration in these lake waters in the northeast of Qaidam Basin increased with pH.
Since KLK is an open lake, it was excluded from the following discussion of close lakes. For the close saline lakes, the pH results of these sampled lake waters show that pH (XLG 2) : 7.53 < pH (GH) : 7.84 < pH (XLG 1) : 8.51< pH (TS) : 8.87. The 238 U concentration of different waters samples were 238 U (XLG 2) : 12.6 µg/L < 238 U (GH) : 16.2 µg/L< 238 U (XLG 1) : 19.8 µg/L < 238 U (TS) : 34.7 µg/L. We can see the aqueous concentration of 238 U of these lake waters increased with pH, which is attributed to the fact that weakly basic conditions would enhance the leaching of uranium in rocks and thereby increase the migration of uranium into water [41]). The pH value of lake waters depends mainly on the contents of HCO − 3 and CO 2− 3 [40]. The results show that the pH value of these lakes increases with an increase in the contents of HCO − 3 and CO 2− 3 . Under basic conditions, uranium is mainly present in the form of UO 2 (CO 3 ) 4− 3 . As the decrease of pH, UO 2 (CO 3 ) 4− 3 decompose into UO 2 (CO 3 ) 2− 2 slowly in water. Two highly mobile forms will eventually become one of the stable forms (UO 2 CO 3 , UO 2 (OH) − 3 , and UO 2 (OH) + ) when the pH is low enough. A study by Mochizuki [46] on Lake Biwa in Japan showed that the change of lake waters pH affected the leaching process of uranium at the sediment/water interface, which would directly affect the aqueous 238 U concentration. Since KLK is an open lake, it was excluded from the following discussion of close lakes. For the close saline lakes, the pH results of these sampled lake waters show that pH (XLG 2): 7.53 < pH (GH): 7.84 < pH (XLG 1): 8.51< pH (TS): 8.87. The 238 U concentration of different waters samples were 238 U (XLG 2): 12.6 μg/L < 238 U (GH): 16.2 μg/L< 238 U (XLG 1): 19.8 μg/L < 238 U (TS): 34.7 μg/L. We can see the aqueous concentration of 238 U of these lake waters increased with pH, which is attributed to the fact that weakly basic conditions would enhance the leaching of uranium in rocks and thereby increase the migration of uranium into water [41]). The pH value of lake waters depends mainly on the contents of HCO − 3 and CO 2− 3 [40]. The results show that the pH value of these lakes increases with an increase in the contents of HCO slowly in water. Two highly mobile forms will eventually become one of the stable forms (UO2CO3, UO2(OH) − 3 , and UO2(OH) + ) when the pH is low enough. A study by Mochizuki [46] on Lake Biwa in Japan showed that the change of lake waters pH affected the leaching process of uranium at the sediment/water interface, which would directly affect the aqueous 238 U concentration.
Under natural conditions, aqueous 238 U concentrations are affected by TDS, and pH, which jointly make different lakes in the same region have different aqueous 238 U concentrations. In order to identify the major controlling factors of aqueous 238 U concentration of saline lake, we performed a correlation analysis of aqueous 238 U concentration with TDS and pH. The results showed that the 238 U concentration was affected by pH to a far greater extent than by TDS. The 238 U concentration in these sampled lakes vary with the range from 12.6 μg/L to 34.7 μg/L. The pH only changes from 7.53 to 7.88, but TDS changes from 0.63 g/L to 39.31 g/L. Namely, the slightly pH variation can result in the bigger variation of 238 U concentration in the sampled saline lake waters than that of TDS.

Response Relationship of 238 U Concentration to Organic Matter in These Lake Waters in the Northeast of Qaidam Basin
Studies have also confirmed the 238 U concentration could be related with biogeochemical process (i.e., the presence of organic matter (OM) and its interaction with iron and manganese oxides/hydroxides) [28][29][30]. The sorption of uranium complexes on iron hydroxides is significantly reduced in the presence of organic matter and/or competing cations, such as Ca 2+ and Mg 2+ [31]. Other studies have shown that the concentration of U(VI) in the aqueous phase drastically decreases during bacterial processes because of the formation of insoluble U(IV) compounds [47,48]. In addition, humic substances (HS), including fulvic acid and humic acid, have a strong complexing ability for 238 U of all types of organic matter. If the bottom sediment of a lake is covered with a large amount of organic Under natural conditions, aqueous 238 U concentrations are affected by TDS, and pH, which jointly make different lakes in the same region have different aqueous 238 U concentrations. In order to identify the major controlling factors of aqueous 238 U concentration of saline lake, we performed a correlation analysis of aqueous 238 U concentration with TDS and pH. The results showed that the 238 U concentration was affected by pH to a far greater extent than by TDS. The 238 U concentration in these sampled lakes vary with the range from 12.6 µg/L to 34.7 µg/L. The pH only changes from 7.53 to 7.88, but TDS changes from 0.63 g/L to 39.31 g/L. Namely, the slightly pH variation can result in the bigger variation of 238 U concentration in the sampled saline lake waters than that of TDS.

Response Relationship of 238 U Concentration to Organic Matter in These Lake Waters in the Northeast of Qaidam Basin
Studies have also confirmed the 238 U concentration could be related with biogeochemical process (i.e., the presence of organic matter (OM) and its interaction with iron and manganese oxides/hydroxides) [28][29][30]. The sorption of uranium complexes on iron hydroxides is significantly reduced in the presence of organic matter and/or competing cations, such as Ca 2+ and Mg 2+ [31]. Other studies have shown that the concentration of U(VI) in the aqueous phase drastically decreases during bacterial processes because of the formation of insoluble U(IV) compounds [47,48]. In addition, humic substances (HS), including fulvic acid and humic acid, have a strong complexing ability for 238 U of all types of organic matter. If the bottom sediment of a lake is covered with a large amount of organic matter, 238 U could precipitate after forming organic complexes with HS, thereby reducing the concentration of 238 U in the lake waters [49,50]. Some researches mentioned that Qinghai is one of the few provinces in China where natural Artemia is distributed and GH is the largest saline lake producing Artemia in Qaidam Basin of Qinghai Province [51][52][53]. There could exist a layer of organic matter attached to the bottom of GH. If we speculate that there exists an evident biogeochemical process in GH, then the decomposition of organic matter and release of CO 2 is possible, which will result in the carbon isotope variation of DIC (Dissolved Inorganic Carbon) and/or biological shell in lake. The carbon and oxygen isotope research of DIC and shell in GH, showed that there is a limited effect on lake water chemistry from biogeochemical process with shallow lake water environment [54,55]. Because the limited data about organic matter effect and the Fe/Mn element result of the sampled lakes, here, we will not present much more discussion.
Based on the above discussion about the relationship between 238 U concentration, TDS, pH, and organic matter, we infer that the dominant factor leading to the low 238 U concentration in GH waters was also pH, although other factors such as TDS and organic matter could have non-negligible effects. There exists a negative correlation between pH value and TDS when TDS is between 50 and 310 g/L for salt lakes [22]. Under natural conditions, when TDS reaches a certain level, pH value will gradually decrease with increasing TDS, and the concentration of aqueous 238 U will be lowered too. Therefore, we conclude that pH is the dominant factor affecting the changes in aqueous 238 U concentration in the sampled saline lakes. The pH values of these saline lakes are not only controlled by the input water, but associated with water-rock interaction during the lake evolution stage, metamorphous degree, and hydrochemistry type [22,56].

Response Relationship of 238 U Concentration to Water Sources Supply in Open Lake Waters
Taking "KLK-Bayin river" as an example to discuss the response relationship of 238 U concentration to water supply sources in an open lake, one of the most important sources of uranium isotopes in water bodies, we note that a large amount of debris is produced by physical erosion and chemical weathering of mountain masses on both sides of rivers.
The upper reaches of the Bayin river are characterized by rolling mountains and abundant rainfall. The leaching effect of the river makes the uranium isotope content of water in the upper reaches generally higher than others. The physical erosion and chemical weathering of mountain masses on both sides of rivers from upper reaches to lower reaches gradually weakened due to the smooth terrain. Because of the large amount of gravel in the riverbed of the Gobi Desert, a large amount of river water supplies groundwater. This makes the uranium content of water in the middle reaches slightly lower than that in the upper reaches The watercourse hardening of the Bayin river in Delhi city prevents the infiltration process of the river. There are many materials used for watercourse hardening, such as concrete, marble, and so on. This may be one of the main reasons why the uranium isotope content in of Bayin river in Delhi city is higher than others. A large amount of spilt groundwater has become the main source of water for the Bayin river, due to the change of lithology of the watercourse and the obstruction of Denan hills. The dilution of groundwater may result in a significant reduction in uranium isotope content of lower reaches for Bayin river [33,34].
Research [21,24] shows that the evaporation of KLK is greater than precipitation every year under such a high evaporation environment. This means that there is a large amount of replenishing water into KLK every year under such high evaporation environment. The results of samples show that the 238 U concentration of Bayin river near the estuary (2.8 µg/L) is lower than that for KLK (5.5 µg/L). However, the result of the watercourse near the exit of KLK is 5.6 µg/L, which indicates certain effect on aqueous 238 U concentration for KLK.

Distribution Characteristics and Influencing Factors of the [ 234 U/ 238 U] AR
There is a certain degree of difference between waters and sediment in the radioactivity of 234 U and 238 U under natural conditions. At present, the activity of the number of atoms of a nuclide is the number of decay events per unit of time, referred to as the [ 234 U/ 238 U] AR In theory, the rate at which 238 U decays into 234 U is the same as the decay rate of 234 U. When the so-called radioactive equilibrium is reached, the [ 234 U/ 238 U] AR is equal to 1 [57]. However, the actual [ 234 U/ 238 U] AR in natural waters is usually greater than 1. Scott [58] made a statistical compilation of the data published before 1982 about 238 U concentrations and the [ 234 U/ 238 U] AR in the water of main rivers on five continents and found that the radioactivity ratio of 234 U to 238 U was basically 1.2-1.3, which was indicative of radioactive disequilibrium between uranium isotopes attributed to the properties of the rock formations and soils in the rivers. Chabaux et al. [59] calculated the mean [ 234 U/ 238 U] AR of the world's major rivers to be 1.17, confirming the above conclusion.
Under natural conditions, the enrichment of 234 U in water is related to the α-decay of 238 U nuclides and there are two main ways. When 238 U in a mineral undergoes α-decay at the interface between a uranium-bearing mineral and waters, the daughter 234 Th will be ejected out of the crystal lattice due to the recoil effect. 234 Th has a short half-life (24 day) and decays into 234 U, which readily forms complex ions and thereby enters the waters, resulting in an increase in aqueous 234 U concentration. This process is called the "α-recoil process". When the 238 U radioactive decay occurs, the daughter 234 U will be placed into the crystal lattice destroyed by the 238 U radioactive decay.
The measured [ 234 U/ 238 U] AR values of these sampled lakes are shown in Figure 7.
that the radioactivity ratio of 234 U to 238 U was basically 1.2-1.3, which was indicative of radioactive disequilibrium between uranium isotopes attributed to the properties of the rock formations and soils in the rivers. Chabaux et al. [59] calculated the mean [ 234 U/ 238 U]AR of the world's major rivers to be 1.17, confirming the above conclusion. Under natural conditions, the enrichment of 234 U in water is related to the α-decay of 238 U nuclides and there are two main ways. When 238 U in a mineral undergoes α-decay at the interface between a uranium-bearing mineral and waters, the daughter 234 Th will be ejected out of the crystal lattice due to the recoil effect. 234 Th has a short half-life (24 day) and decays into 234 U, which readily forms complex ions and thereby enters the waters, resulting in an increase in aqueous 234 U concentration. This process is called the "α-recoil process". When the 238 U radioactive decay occurs, the daughter 234 U will be placed into the crystal lattice destroyed by the 238 U radioactive decay.
The measured [ 234 U/ 238 U]AR values of these sampled lakes are shown in Figure 7. U]AR ≠ 1), which is due to frequent waterrock interaction (i.e., sediments, atmosphere dust and organic material etc.) that promotes the production and release of 234 U into the waters due to the α-recoil process. This increases the aqueous 234 U concentration and thereby causes the [ 234 U/ 238 U]AR to significantly deviate from the equilibrium.

GH
Based on the current data, we only could show that there are big differences of [ 234 U/ 238 U]AR in the sampled different lakes located in the same regional background, but show consistent [ 234 U/ 238 U]AR in vary sites and depths in any one lake. Further, the [ 234 U/ 238 U]AR in all sampled saline lakes is above 1.5 which is higher than the reported average value (1.17) of river from the world [58], lower than the reported value (~2.5) of lakes in Namshi-Nur and Tsagan-Tyrm [64]and that of rivers in same region [25]. These results mean that the [ 234 U/ 238 U]AR of these saline lakes may be related to the input water and the water-rock interaction during the evolution stage, metamorphous degree, and   AR 1), which is due to frequent water-rock interaction (i.e., sediments, atmosphere dust and organic material etc.) that promotes the production and release of 234 U into the waters due to the α-recoil process. This increases the aqueous 234 U concentration and thereby causes the [ 234 U/ 238 U] AR to significantly deviate from the equilibrium.
Based on the current data, we only could show that there are big differences of [ 234 U/ 238 U] AR in the sampled different lakes located in the same regional background, but show consistent [ 234 U/ 238 U] AR in vary sites and depths in any one lake. Further, the [ 234 U/ 238 U] AR in all sampled saline lakes is above 1.5 which is higher than the reported average value (1.17) of river from the world [58], lower than the reported value (~2.5) of lakes in Namshi-Nur and Tsagan-Tyrm [64] and that of rivers in same region [25]. These results mean that the [ 234 U/ 238 U] AR of these saline lakes may be related to the input water and the water-rock interaction during the evolution stage, metamorphous degree, and hydrochemistry type. The input water of saline lakes in the northeast of Qaidam Basin is mainly based on natural precipitation, river water, and groundwater [56,65]. The factors affecting the [ 234 U/ 238 U] AR in water-rock interaction during the evolution stage, metamorphous degree, and hydrochemistry type could be related with the redox conditions, pH of water, and biogeochemical process (i.e., the presence of organic matter (OM) and its interaction with iron and manganese oxides/hydroxides and/or competing cations, such as Ca 2+ and Mg 2+ ) during evolution stage, metamorphous degree, and hydrochemistry of the saline lakes. This will be resolved. By analyzing the major and trace element geochemistry distribution and the [ 234 U/ 238 U] AR of the suspended particles filtered from the water of these sampled lakes in future.

Seasonal Effects on 238 U Concentration and the [ 234 U/ 238 U] AR in Lakes and Their Potential Environmental Impact
The concentration of 238 U in natural lake waters and the [ 234 U/ 238 U] AR are the best tracers for determining changes in the water environment. The main source of 238 U in lake waters are water-rock interaction and rock weathering, in addition to other sources such as exogenous inputs from rivers, atmospheric precipitation, and groundwater recharge. The radioactive disequilibrium of 234 U/ 238 U is the best evidence of the water-rock interaction of 238 U. The 238 U concentration and the [ 234 U/ 238 U] AR of various sites in a lake remain almost consistent. However, this will be affected by seasonal variation due to an input of precipitation mainly occurring in the rainy season. The annual precipitation in GH and TS is subject to significant variation, with precipitation mainly concentrated in May-September every year [21,32], which is likely to induce the seasonal variation of aqueous 238 U concentration. However, compared with the total water volume in GH and TS, the effect of seasonal precipitation variation on aqueous 238 U concentration was insignificant. Based on the seasonal effect on the result of 238 U concentration in GH and TS (Table 5), the seasonal change range of 238 U concentration in GH is 11.96%, and the result is 2.25% for TS. Therefore, we conclude that the subtle seasonal difference of 238 U concentration in the lake waters are due to the dilution effect of precipitation and/or groundwater.
In summary, the concentration of 238 U in natural lake waters and the [ 234 U/ 238 U] AR remain almost consistent for a certain period of time. Studies of U-series disequilibria in lakes mainly include the determination of sediment ages and/or of sedimentation rates [66]. Stable U-and Th-series nuclides in lakes is the basis of the determination of sediment ages. The ability of obtaining precise sediment ages is critical to the accuracy of the reconstruction of paleoclimate and paleoenvironment.
U-and Th-series nuclides in lakes are closely related. U-and Th-series nuclides of lake waters may give us some information about geochemical behavior and mixing characteristics of lakes [6]. There are many environmental changes recorded by lake sediments. We can chronologically decipher such records by measuring the U-Th nuclides of these sediments.

Conclusions
Four lakes in the northeast of Qaidam Basin were selected to represent a sequence of waters salinity in order to test the effect of lake salinity on aqueous uranium isotopes. A total of 43 water samples were taken during three different sampling campaigns and analyzed for water chemistry and uranium isotope concentrations. The aqueous 238 U concentration of the freshwater lakes was generally low, while it was generally high in these saline lakes, which is a response to the combined effects of TDS and pH.
Furthermore, pH is the dominant factor affecting changes in aqueous 238 U concentration of the sampled saline close lakes, as the lake waters pH could directly affected the speciation of 238 U, with lower water pH leading to more fractions of 238 U into stable forms, and thereby smaller aqueous 238 U concentration. With the exception of pH value, the replenishment of water likely exerts a certain effect on the uranium isotope concentration for open KLK fresh lake located in the northeast of Qaidam Basin in the same high evaporation environment background as close saline lakes.
The [ 234 U/ 238 U] AR of each lake was greater than one, ranging roughly from 1.460 to 1.798, which was likely due to the α-recoil of 238 U at the water-rock interface and the leaching-precipitation process. The degree of consistency in the 238 U and [ 234 U/ 238 U] AR values across various sites in a lake indicated that aqueous 238 U and [ 234 U/ 238 U] AR were evenly distributed in the same lakes.
The [ 234 U/ 238 U] AR in these saline lakes are closely related to the input water and the associated water-rock interactions involving sediments, atmosphere dust, and organic material etc. during the evolution stage, metamorphous degree, and hydrochemistry of these saline lakes. If the concentration of 238 U in natural lake waters and the [ 234 U/ 238 U] AR remain almost consistent for a certain period of time. The potential environmental impact is that which is the base of the determination of sediments ages in lake and/or of sedimentation rates could be determined by lake sediment and/or biogenic carbonate in future. Further, we can chronologically decipher the environmental change records in them by measuring the U-Th nuclides of these sediments.