Surface Water Sample

Surface water samples were collected at a depth of about 10cm and filtered through pre-washed 0.45 μm Millipore nitrocellulose filters. The initial portion of the filtration was discarded to clean the membrane, and the following ones were acidified to pH<2 with ultra-purified 6 M HNO₃ and then stored in plastic bottles for trace metal analyses. Cleaning of plastic bottles was carried out by soaking in 20% (v/v) HNO₃ for 24 h and then rinsing with milli-Q deionized water (~18 MΩ/cm resistivity; TOC<5 mg/L) from Milli-Q (Millipore, Direct 8).

Field indicators were determined by a Portable Multi-parameter Water Quality Monitoring Analyzer (EXO1, USA). Metallic elements were detected by an Inductively Coupled Plasma Mass Spectrometer (NexION 300X, USA) with a detection limit of 0.06~0.41μg/L and a recovery rate of 86.8~99.8%. Hg was detected by the Atomic Fluorescence Spectrophotometer (AFS-3100, China) with a detection limit of 0.01 μg/L and a recovery rate of 93.0~104.0%. The contents of cations (K⁺, Na⁺, Ca²⁺, and Mg⁺) were analyzed by a Microwave Plasma Atomic Emission Spectrometer (Agilent 4200, USA) with a detection limit of 1.94~6.61μg/L and a recovery rate of 95.6~108.7%.

The contents of anions (F^-, Cl^-, SO₄^2-, and NO₃^-) were analyzed by Ion Chromatography (ICS900, USA) with a detection limit of 0.006~0.018 mg/L and a recovery rate of 99.1~116.3%. The CO₃^2- and HCO₃^– were analyzed by acid-base titration (methyl orange as an indicator and diluted hydrochloric acid titration) with an error of less than 5%.

Fig. 1.

Sampling sites in the study area

Sediment Sample

After being naturally dried, sand and gravel were removed from the sediment. It was ground before passing through a mesh sieve with a pore size of 0.075 mm. According to the literature method [45-47], sediment samples were digested by a Microwave Digestion Instrument (MARS6, USA). Metallic elements in the sediment were detected by an Inductively Coupled Plasma Mass Spectrometer with a detection limit of 0.04~2.0 mg/kg and a recovery rate of 75~120%. Hg in the sediment was detected by an Atomic Fluorescence Spectrophotometer with a detection limit of 0.002 mg/kg and a recovery rate of 80%.

Physical and Chemical Properties Analysis

According to the study results (Table 2), the main rivers in the SRYR were alkalescence, and the temperature difference was large. The pH range was 8.2~8.8. The water temperature range was 7.5~18.6℃. The dissolved oxygen (DO) range was 7.0~9.3 mg/L. The redox potential (ORP) range was 16.0~109 mV. The conductivity range was 333~5326 μS/cm. The total dissolved solids (TDS) range was 145~2875 mg/L.

Similar to the SRYR, the main rivers in the SRLR were alkalescence, and the temperature difference was large. The pH range was between 8.3 and 8.6. The water temperature range was 9.7~17.3℃. The DO range was 7.6~8.8 mg/L. The ORP range was 22.3~56.9 mV. The conductivity range was 709~2160 μS/cm. The TDS range was 423~1281 mg/L.

The average value of TDS in the source region was 647.3 mg/L and 702.2 mg/L, both of which were much larger than the world average of 150 mg/L [48]. It might be caused by geotectonic movement [34]. The uplift of the Qinghai-Tibet Plateau is the result of tectonic movement, which caused landform ups and downs, and led to the exposure of fresh rocks to accelerate the weathering process [49, 50]. It had caused an increase in the content of ions, such as K⁺, Na⁺, Ca²⁺, and Mg²⁺, entering the river basin through weathering, and TDS was closely related to these ions. Therefore, the tectonic movement had affected the TDS content, making it higher than the world average. The total hardness (TH) range in the SRYR was 279~693 mg/L, and the average value was 647.3 mg/L. The TH range in the SRLR was 338~812 mg/L, and the average value was 620.2 mg/L. It was much larger than the average value of the Yangtze River and the Lancang River basins (291.6 mg/L and 164.4 mg/L), which indicated that the water hardness in the source area was relatively high [51-53].

Most of the river water in the source region was fresh water, with medium or high hardness, and only a small fraction was brackish water with high hardness (Figure 2). This might be caused by complex topography and a changeable local climate [54]. It was reported that the Qinghai Provincial Government implemented an emergency drainage project in 2019. The high salinity lake water located in the upper reaches was guided into the Chumaer River, which led to an increase in salinity in the Chumaer River [55, 56].

Fig. 2.

TDS and TH distribution of river water in the source region

Fig. 3.

Proportion of anions and cations content (a. Cation Percentage; b. Anion Percentage)

Characteristics of Ionic Content and Spatial Distribution in Water

Analysis of Hydrochemical Characteristics and Main Control Factors

The Piper plot (Figure 6) provides a way to visualize the compositions and water types of rivers and therefore, can show the relative importance of different weathering regimes [29]. On the cation diagram, samples from Zhaqu and Dangqu lie in the lower-left corner, where the Ca values are high (more than 50%), revealing that the dominant cation in these samples is calcium. On the contrary, samples from the Chumaer River and Tuotuo River fall on the lower-right corner, where the dominant ion is sodium, showing the characteristics of evaporites. The values of samples from the Tongtian River are relatively scattered, but still show that the dominant ions are Ca and Na. On the anion diagram, the samples from Dangqu and a part of the Tongtian River fall on the left, showing carbonate rock weathering features. However, samples from the Chumaer River, the Tuotuo River, and a part of the Tongtian River fall on the lower-right corner, revealing chloride-dominated features. The values of samples from Zhaqu are relatively scattered. According to the Piper plot (Figure 6), the main hydrochemical types in the source region are Na-Cl and Ca-SO₄•HCO₃, but they are diverse in different rivers in the SRYR. The hydrochemical type of the river water in Dangqu is Ca-SO₄•HCO₃, which indicates that it may be affected by the evaporites, carbonate, and silicate rocks [72]. That of river water in the Tuotuo River and Chumaer River is Na-Cl, affected by evaporite dissolution [34, 73, 74]. In the Tongtian River, the hydrochemical types are Na-Cl and Ca-HCO₃. That of the trunk stream is Na-Cl, which has the same hydrochemical type as the Chumaer River and Tuotuo River, indicating that their ion has a great impact on that of the Tongtian River.

Fig. 4.

Metallic elements’ average content of river water in different regions

Fig. 5.

Distribution of metallic elements in different rivers in the SRYR

Gibbs defined the mechanisms controlling world surface water chemistry as three end-members: atmospheric precipitation, rock dominance, and the evaporation-crystallization process. To distinguish the contribution of the three end-members, TDS versus Na⁺/(Na⁺+Ca²⁺) or Cl^-/(Cl^-+HCO₃^-) are used [75]. The Gibbs schematic diagram is shown in Figure 7. The river water of the source region is characterized by moderate or high concentrations of TDS, indicating that the dominant factor is rock weathering and evaporation-crystallization processes, and none of the rivers are affected by precipitation [76]. The diagram reveals that the river waters of Dangqu are characterized by low ratios of Na⁺/(Na⁺+Ca²⁺) and Cl^-/(Cl^-+HCO₃^-) (less than 0.5) and moderate concentrations of TDS (about 100 to 400 mg/L), typical of rock-dominant rivers. Chumaer River is characterized by high ratios of Na⁺/(Na⁺+Ca²⁺) and Cl^-/(Cl^-+HCO₃^-) (approaching 1) and high concentrations of TDS, showing that the controlling factor is evaporation-crystallization processes. However, Tuotuo River and Tongtian River are in an intermediate state, having moderate values of TDS and ratios of Na⁺/(Na⁺+Ca²⁺) and Cl^-/(Cl^-+HCO₃^-), as is the river in the SRLR. They are controlled by rock weathering and evaporation-crystallization processes [32]. These results and conclusions are consistent with the Piper plot (Figure 6).

Fig. 6.

Piper plot of major ion composition

Fig. 7.

Variation of the ratio of Na⁺/(Na⁺+Ca²⁺) and Cl^-/(Cl^-+HCO₃ ^-) as a function of TDS in the Gibbs diagram (a. TDS-Na⁺/(Na⁺+Ca²⁺); b.TDS-Cl^-/(Cl^-+HCO₃ ^-))

Analysis of Ion Sources in Water

The way of ions’ sources in river water can be divided into natural or anthropogenic. Natural sources are atmospheric transmission and rock weathering. Anthropogenic sources are influenced by human activities such as mineral development, agriculture, and animal husbandry, and so on [70, 77, 78]. The source of river ions can be determined by calculating the ratio of different ions.

(2) Rock Weathering

Rock weathering with different lithologies can produce different ions, which can reflect how that affects ions [83]. Na⁺ and K⁺ in natural water mainly come from evaporites or silicate rock weathering. Ca²⁺ and Mg²⁺ may come from evaporites, carbonate, or silicate rock weathering. Cl^- and SO₄^2- mainly come from evaporite weathering. And HCO₃^- mainly comes from carbonate rock weathering [84, 85].

In the SRYR, the molar equivalent ratios of Cl^-/Na⁺ in the river water were less than or equal to one (Figure 8), which indicated that Na⁺ and Cl^- were greatly influenced by the evaporites. It should be paid special attention to that the Cl^-/Na⁺ ratio of Dangqu and some areas of the Tongtian River was located at the lower right side of the isogram, which indicated that those areas might be affected not only by evaporites, but also by silicate rock weathering. The ratio of all samples was located above the (Ca²⁺+Mg²⁺)/ HCO₃^- and (Ca²⁺+Mg²⁺)/(SO₄^2-+HCO₃^-) isograms, and closer to the (Ca²⁺+Mg²⁺)/(SO₄^2-+HCO₃^-) isograms, indicating that Ca²⁺ and Mg²⁺ were affected not only by carbonate rock weathering, but also by evaporites [86]. The ratios of the Dangqu and part of the Tongtian River were located around the HCO₃^-/(Cl^-+SO₄^2-) isogram, which indicated that those areas were affected by carbonate rocks and evaporites [87]. While the ratios of Chumaer River, the Tuotuo River, and part of the Tongtian River were located at the lower right side of the HCO₃^-/(Cl^-+SO₄^2-) isogram, showing that those areas were effected by evaporites more than carbonate rocks.

Fig. 8.

The molar equivalent ratios of different ions (a. Cl^-/Na⁺; b. (Ca²⁺+Mg²⁺)/HCO₃ ^-; c. (Ca²⁺+Mg²⁺)/(SO₄ ^2-+HCO₃ ^-); d. HCO₃ ^-/(Cl^-+SO₄ ^2-))

In the SRLR, the molar equivalent ratio of Cl^-/Na⁺ was about 1 and the ratios of (Ca²⁺+Mg²⁺)/HCO₃^- and (Ca²⁺+Mg²⁺)/(SO₄^2-+HCO₃^-) were located at the upper left of the isogram, which indicated that Na⁺ and Cl^- were effected by evaporites and Ca²⁺ and Mg²⁺ were influenced by carbonate rocks and evaporites. In addition, according to the ratios of HCO₃^-/(Cl^-+SO₄^2-), the main ions were more affected by the evaporites than carbonate rocks.

(3) Human Activities Affect

In recent years, with the development of tourism and the increase of grazing activities in Tibetan areas, the hydrochemical characteristics in the source region may change accordingly [88]. The NO₃^- can serve as an indicator of the degree of impact of human activities on the water environment [48, 89]. The NO₃^- content in the river water of the source region was low as a whole. In the SRLR, the NO₃^- content ranged from 1.77 to 2.88 mg/L, with an average of 2.19 mg/L. In the SRYR, that was 0.40~2.71 mg/L, with an average of 1.46 mg/L. The NO₃^- concentration in the natural background value is generally 0.08~0.16 mg/L [90]. In the SRLR and the SRYR, the NO₃^- concentration is higher than the natural background value. Therefore, it may be affected by human input.

To further investigate the impact of human activities on the river water in the source region, the ratios of NO₃^-, SO₄^2-, and Cl^- normalized by Na⁺ were used to distinguish between human sources and nature sources. However, there was no obvious correlation between NO₃^-/Na⁺, SO₄^2-/Na⁺, and Cl^-/Na⁺ in river water (Figure 9), which showed that the sources of NO₃^- were different from SO₄^2- and Cl^-. The Cl^- and SO₄^2- mainly come from evaporite weathering, which belongs to a kind of natural source. Therefore, the NO₃^- may come from human inputs. Anthropogenic sources are usually divided into industrial sources and agricultural sources. Generally, industrial sources are associated with higher NO₃^-/Cl^- and SO₄^2-/ Ca²⁺, while agricultural sources are characterized by low NO₃^- and high Cl^-. However, the geographical location of the study area is special, with an average elevation of more than 4000 m, and no industrial sources were found in the study area during the field investigation, so the input of agricultural sources was considered. In Figure 9, the NO₃^-/Na⁺ was obviously much lower than the Cl^-/ Na⁺ which meant high Cl^-/NO₃^-. In other words, it meant low NO₃^- and high Cl^-. Therefore, the river water of the source region was affected by agricultural sources to a certain extent [54]. Due to the low content of NO₃^-, it was speculated that human activities had a relatively small impact on the ions.

Fig. 9.

The ratios of NO₃ ^-, SO₄ ^2- and Cl^- normalized by Na⁺ (a. SO₄ ^2-/Na⁺-NO₃ ^-/Na⁺; b. Cl^-/Na⁺-NO₃ ^-/Na⁺)

Content and Distribution Characteristics of Metallic Elements in Sediments

Metallic elements are a special group of ions because heavy metals have persistence and potential toxicity to the environment, which can change the activity processes of microorganisms and can be accumulated through sediments and aquatic food chains [91]. The content of metals in sediments is an important factor affecting river water, because metals can be transformed into each other in sediment water. Analyzing the content, species, and distribution characteristics of metal elements in sediments is helpful to determine the pollution degree of metals in rivers and their potential pollution risks. Therefore, the analysis of metallic elements in sediment is very important for assessing the state of the river.

Compared with other metallic elements, the average contents of Ti, Mn, Fe, Sr, and Ba in the sediment of the source region were relatively higher (Figure 10). In the SRYR, the average content of Fe in the sediment was 19758 mg/kg and Ti was 2581 mg/kg. In the SRLR, that of Fe was 18736 mg/kg and Ti was 2533mg/kg. It was mainly due to the high background value of some metallic elements in river sediments in the source region [46, 92].

Fig. 10.

The box-plot of the metallic elements’ average contents in sediment

In the SRYR, the average contents of Cd, Mo, Sr, Se, Cu, Co, Fe, Mn, Cr, V, Ti, and Li in the sediments of different rivers were basically flat, while those of the others were different in spatial distribution (Figure 11). In the sediments of the Dangqu River, the average contents of Hg, As, Ba, Sb, Ni, and Be were higher than those in the Tuotuo River, Chumaer River, and Tongtian River. In the sediments of the Chumaer River, the average contents of Hg (below the detection limit), As, Pb, Sb, Zn, Ni, and Be were lower than those in other rivers. In the sediments of the Tuotuo River, Hg, and Ba were below the detection limit, and the average contents of other metallic elements were flat with those of other rivers. In the sediments of the Tongtian River, only the average content of Hg was obviously higher than that of other rivers, and others were flat or lower.

Compared with the average background value of Chinese river sediment [93, 94], in the sediments of the Dangqu River, the average contents of Pb, Ba, Zn, Cu, Co, Fe, Mn, V, and Ti were obviously lower, Cd, Mo, and Ni were basically flat, and Hg, As, Sb, Sr, Se, Cr, Be, and Li were higher. Except for Se and Cr, the average contents of other metallic elements in the Chumaer River were lower than the average background value of Chinese river sediment. In the Tuotuo River, the average contents of As, Sr, Se, and Cr in sediments were obviously higher, while Sb, Cd, Mo, and Cr were flat and other metallic elements were lower. In the Tongtian River, the average contents of Hg, As, Sr, Se, and Cr in sediments were relatively higher, while Sb, Cd, Mo, and Li were flat. The other metallic elements were lower than the average background value of Chinese river sediment.

In the SRLR, the average contents of Cr and Cd in sediments were slightly higher than the background values of sediments in the China River system (Figure 12), while V, Ni, Cu, Zn, Pb, Mn, and Fe were lower. Some studies indicate that it was the rocks containing Cd and Cr that affected the content of metallic elements in sediments [95, 96].

As shown in Figure 12, the average contents of metallic elements in the sediments in the SRYR were basically flat compared to those in the SRLR. Compared with the source regions of the Yellow River and other rivers in alpine and cold regions, the average contents of V, Ni, and Zn in the sediments of the source regions were slightly lower, while Cr was slightly higher. Cd was lower than that in the source region of the Yellow River, but higher than that in rivers in alpine and cold regions. Pb, Mn, and Fe were higher than those in the source region of the Yellow River, but lower than those in rivers in alpine and cold regions.

Analysis of the Source of Metallic Elements in Sediments

Correlation analysis can be used to infer the homology between various metallic elements and reveal the source relationship of ions based on the significant level of correlation [60, 92]. Considering that the amount of samples affects the analysis accuracy, the metallic elements in the sediments of rivers in the source region measured by our study team in the past five years (2019~2022) were selected to complete the relevant analysis.

Fig. 11.

Distribution of metallic elements in river sediments in the SRYR

Fig. 12.

Metallic elements’ average contents in sediments of different areas

The Pearson Correlation Coefficient is shown in Table 5. In the SRLR, the results show that there was a significant positive correlation (p<0.05) between Fe and Ti, V; Pb, V, Co, Sr, Sb, Co, Cr, and Ni were pairwise correlated; Cu, Sr, Cb, and Sb were pairwise correlated; Be, Mn, and Ni were pairwise correlated. It indicated that these metallic elements had a certain degree of correlation with each other, and their sources might have similarities and homology. In the SRYR, Li and Be were significantly positively correlated with most ions; T, V, Mn, Fe, Co, and Cu were pairwise correlated; Mn and Co were significantly positively correlated with Ni, Se, and Pb; there was a significant positive correlation in Se-Sr and Sr-Cd-Pb. It indicated that these metallic elements were highly correlated (p<0.01) and might have similar sources. However, not all metallic elements had a significant correlation, which indicated that their sources were not completely consistent.

In order to further analyze the main sources of metallic elements, the principal component analysis method was used to analyze the sediments in the source region. Under the premise of a 95% confidence level, three principal components (eigenvalues>1) of sediments were extracted, respectively. In the source region of the Lancang River, the contribution rates of the first and second principal components were 67.15% and 20.65%, and the cumulative contribution rate was 87.8%. In the SRYR, the contribution rates of the first and second principal components of sediments were 45.97% and 35.05%, and the cumulative contribution rate was 81.02%.

The results of the principal component analysis are shown in Table 6. In the SRLR, the first principal component contribution rate of the metallic elements in river sediments (67.15%) was much higher than that of other principal components. The metallic elements with a higher load were Be, Ti, V, Cr, Fe, Co, Ni, Cu, Zn, Sr, Cd, Sb, and Pb, and the metallic elements of the second principal component were Li and Mn. In the SRYR, the first principal components of metallic elements with a higher load in river sediments were Li, Be, V, Mn, Fe, Co, and Ni, and the metallic elements of the second principal component were Zn, Sr, Cd, Sb, and Pb. According to the correlation analysis, these metallic elements had a significant correlation, which indicated that their sources were similar or the same. According to the literature research, Fe and Mn were abundant macro-elements in the crust [97, 98]; Ni and Co were controlled by the geological background of the Qinghai-Tibet Plateau [99]; Pb, Zn, and Ba were often used as identification elements of traffic pollution sources [100, 101]. Although the values of Pb, Zn, and Ba in the study area were lower than the background values of sediments in China, there was a significant correlation between Sr and Sb with Pb and Zn, while the values of Sr and Sb were higher than the background values of sediments in China. Therefore, Sr and Sb might be affected by traffic pollution sources. In conclusion, it was considered that the metallic elements in the sediments of the source region were controlled by both natural and traffic pollution.