J. Cent. South Univ. (2020) 27: 1211-1223

DOI: https://doi.org/10.1007/s11771-020-4361-7

Assessment of water quality and safety based on multi-statistical analyses of nutrients, biochemical indexes and heavy metals

JIANG Dong-yi(蒋东益)¹, WANG Yun-yan(王云燕)^{1, 2, 3}, LIAO Qi(廖骐)^{1, 2, 3},LONG Zhe(龙哲)⁴, ZHOU San-yang(周三羊)⁵

1. School of Metallurgy and Environment, Central South University, Changsha 410083, China;

2. Chinese National Engineering Research Center for Control & Treatment of Heavy Metal Pollution, Changsha 410083, China;

3. Water Pollution Control Technology Key Lab of Hunan Province, Changsha 410004, China;

4. School of Information Science and Engineering, Central South University, Changsha 410083, China;

5. Hunan Province Environmental Monitoring Centre, Changsha 410014, China

Central South University Press and Springer-Verlag GmbH Germany, part of Springer Nature 2020

Abstract: The purpose of this research was to better understand the water quality status of the Xiangjiang River, China, and to evaluate the risks posed by the river water. Precisely, ten water quality parameters including pH, dissolved oxygen (DO), Escherichia coli (E. coli), potassium permanganate index (COD_Mn), dichromate oxidizability (COD_Cr), five-day biochemical oxygen demand (BOD₅), ammonia nitrogen (NH₄⁺-N), total phosphorus (TP) and fluoride (F^-) as well as metal(loid)s (Pb, Hg, Cd, As, Zn, Cu and Se) were monitored monthly in 2016 at 12 sampling sites throughout the Hengyang section of the Xiangjiang River. Concentrations of all parameters were presented according to rainy and dry seasons. They were compared with Chinese surface water standards and WHO drinking water limits to assess the sustainability of the river water status. Principal component analysis (PCA) revealed different pollution sources in different seasons. Dual hierarchical cluster analysis (DHCA) was applied to further classify the water quality variables and sampling sites. Besides, a risk assessment was introduced to evaluate the carcinogenic and non-carcinogenic concerns of heavy metal(loid)s to human health. This research will help to optimize water monitoring locations and establish pollution reduction strategies on the preservation of public safety.

Key words: surface water; water quality; human health risk; the Xiangjiang River

Cite this article as: JIANG Dong-yi, WANG Yun-yan, LIAO Qi, LONG Zhe, ZHOU San-yang. Assessment of water quality and safety based on multi-statistical analyses of nutrients, biochemical indexes and heavy metals [J]. Journal of Central South University, 2020, 27(4): 1211-1223. DOI: https://doi.org/10.1007/s11771-020-4361-7.

1 Introduction

Water is the most fundamental resource for sustaining life, and it is a renewable but not unlimited resource. China is one of the most water resource shortage countries in the world [1-3]. Owning to the water resources scarcity and water pollution, China is confronted with supply and demand problems under the fast development of the society now [4]. Besides, water resource is further influenced by climate change [5, 6]. In the context of global climate change, rising temperature, sea level, and the intensity and frequency of extreme weather events such as drought and floods, global water flows will impact [7] and challenge water availability and increase exposures to unsafe water. For example, floods and heavy storms will show an increasing trend worldwide and enhance the risk of waterborne diseases [8]. Besides, rising temperature and salinity change the aquatic environment status [9]. Some studies have found that the movement of pesticides and the distribution of pharmaceuticals and nutrient pollution correlated with climate change significantly [9-11]. Although the biodegradation of chemicals becomes more rapid in the water at higher temperatures, the toxicity of many chemicals will also increase because of the increasing accumulation rates in aquatic organisms via food and respiratory or other body surfaces [9]. Therefore, climate change, along with growing anthropogenic activities, has caused a lot of stress and has escalated concern for pollution control and management of world water resources [10, 12, 13].

In China, water scarcity, uneven distribution, and widespread pollution have posed significant challenges to the sustainable development of the society and economy [14]. It is worth noting that discharges of chemicals and pollutants from agricultures, industries, and domestic activities into rivers threaten water security, water quality, and water-dependent biodiversity [15-18]. A report on the national freshwater environment in 2016 showed that only 67.8% of surface water sections (sites), 71.2% of the river basins, and 66% of lakes (reservoirs) can meet the standard of drinking water [19]. The year 2016 marked the beginning of “The 13th Five-Year Plan” for China and was expected to be a pivotal year to fully implement the Water Pollution Prevention and Control Plan for Major River Basins (2016-2020) [20]. The plan specified the priorities for water environment protection of the Beijing-Tianjin-Hebei region and the Yangtze River Economic Belt. As one of the largest tributaries of the Yangtze River [21], the Xiangjiang River delivers great wealth (more than 80% of provincial GDP) to people of Hunan Province. It bears a heavy burden (more than 60% of pollution). Since 2013, the first “Three-Year Action Plan” has launched a series on the preservation and protection of the Xiangjiang River, which aims to reduce the pollution sources and improve water quality. Until 2016, the quality of the water environment in the Xiangjiang River Basin has been improved and the metal(loid)s cadmium (Cd), lead (Pb), chromium (Cr(VI)), mercury (Hg), and Arsenic (As) decreased by 54.6%, 52.8%, 36.8%, 15.1%, and 4.4%, respectively, compared to 2015 [22]. The second “Three-Year Action Plan” which aimed at further improving the water quality in the main streams and tributaries of the Xiangjiang River began in 2016. In this study, water quality in 2016 was assessed based on the concentration of nutrients, biochemical indexes, and heavy metal(loid)s monitored in an important industrial section of the Xiangjiang River.

Among researches for the Xiangjiang rivers, ZENG et al [23] assessed the water quality of the Xiangjiang River in terms of 7 trace elements (As, Cd, Cu, Pb, Se, Zn, and Hg), and found that the upstream of the Xiangjiang River was a relatively high polluted area. CHAI et al [3] discovered that mining and smelting anthropogenic activities caused pollution of Cd, Pb, Zn, and Cu during their assessment on surface sediments of the Xiangjiang River. TANG et al [15] found the unique behavior of the chromium in the sediments. However, these studies were lack of the combination of season comparisons and risk assessments. Human health risks were evaluated in our research to demonstrate the ecosystem health of the river waters. We believe that our results will be helpful and meaningful for water resource administration and management.

2 Materials and methods

2.1 Study area

The Xiangjiang River is regarded as the mother river of Hunan Province. It flows through the most developed and urbanized regions including Yongzhou, Hengyang, Zhuzhou, Xiangtan, Changsha and Yueyang cities. Its length is 948 km, and the basin covers 94721 km², supporting approximately 30 million residents [23]. In 1995, the Chinese government secured a World Bank loan to conduct a waterways project entitled “Inland Waterways Multipurpose Project” in the Hengyang section of the Xiangjiang River [24]. The basin projects, including reservoir construction, embankment reinforcement, and vegetation and water environment protection, enhanced flood control capability, boosted the socioeconomic development of the region, and improved the ecological environment of the basin [25]. From 2004 to 2008, a decreasing trend was found for heavy metal(loid)s and an increasing trend occurred for organic factors on an annual scale, and the seasonal pattern was only observed for natural factors [24]. A comprehensive study of water quality and water health needs to be conducted in the Hengyang section because it is one of the leading industrial cities and covers the largest basin of the Xiangjiang River.

2.2 Sample collection and analytical methods

Samples were collected monthly at 12 sites in the Hengyang section of the Xiangjiang River from January to December 2016. Sampling sites were located at hydrological monitoring stations to ensure that representative and comprehensive water quality information was collected. Quality assurance and quality control in the laboratory were implemented and covered the detailed procedures of operation and calibration of standards. All samples were filtered by 0.45-μm micropore membranes, and then transferred into pre-conditioned acid- washed polyethylene containers. Until analysis, samples were acidified with concentrated nitric acid and then refrigerated at 0-5 °C with special labels and seals. For the sample collection and preservation of samples, national standard [26] issued by the National Health Commission of the People’s Republic of China and the Standardization Administration of the People’s Republic of China was carried out to ensure the standardization and the integrity of all the samples. For the determination and analysis of the sample, strict Chinese detection standards and methods were followed, and detailed methods are given in Table 1. The determination of E. coli followed the instruction book issued by the Ministry of Environmental Protection of the People’s Republic of China [39]. In addition, triplicate samples were collected, and their standard deviations were within 5%.

Figure 1 Sampling sites on Hengyang section of Xiangjiang River, China

Table 1 Variables and methods of determination

2.3 Statistical analysis methods

2.3.1 Principal component analysis

Principal component analysis (PCA) is a useful tool that broadly utilized in water quality data analysis to find the inner connections between variables from the original dataset [40-47]. PCA data reduction techniques to divide the original dataset into principal components (PCs) [48, 49]. PCs containing eigenvalues of >1 remained; the result of the Kaiser-Meyer-Olkin was >0.5; the result of the Bartlett’s test was <0.001 [50, 51].

2.3.2 Dual hierarchical cluster analysis

Dual hierarchical cluster analysis (DHCA) is a combination of two-hierarchical cluster analysis (HCA) regarding metal(loid)s and sites, which presented to be a thermodynamic graph with cluster lines. HCA was executed based on Euclidean and Ward’s methods with a cut-off line to distinguish the cluster [3, 23, 52]. The thermodynamic graph that we created can show the relative concentration for one element in different sites, and the cluster lines can demonstrate the inner relationship of metal(loid)s and sites.

2.3.3 Human health risk assessment

The human health risks assessment is an effective way of explaining the public health situation [53]. Exposure to metal(loid)s in the Xiangjiang river water would happen from the incidental ingestion of river water, the inhalation of particulates emitted from the river water, and dermal absorption through skin exposure [54]. Risk characterization is quantified by carcinogenic risk and non-carcinogenic risk [50]. The calculation formulae are presented in Table 2. ADD_ingestion is the average daily dose by ingestion, μg·kg^-1·d^-1; DAD_dermal is the dermally absorbed dose, μg·kg^-1·d^-1; C_w is the average concentration of the metal(loid)s in water, μg·L^-1; HQ is the hazard quotient; RfD is the reference dose, μg·kg^-1·d^-1; HI is the hazard index; CR is the carcinogenic risk. HQ>1 indicates that concerns should be made on human health. The reasonable range of CR pointed out by the USEPA is 10^-6 to 10^-4 [56]. The values of specific parameters are shown in Table 3. BW is the body weight; ED is the exposure duration; AT is the average time; IR is the ingestion rate; SA is the exposed skin surface area; EF is the exposure frequency; ET is the exposure time; ABS_GI is the gastrointestinal absorption factor; K_p is the dermal permeability coefficient in water; CSF is the cancer slope factor. The equations in our study and the specific parameters values are referred from USEPA and the official Chinese guideline (HJ 25.3-2014) [56, 58].

Table 2 Summarized equations of human health risk assessment recommended by USEPA

Table 3 Specific value of parameters in risk assessment and references

3 Results and discussion

3.1 Seasonal variation of water quality parameters

The water quality parameters including nutrients, biochemical indexes, and heavy metal(loid)s were seasonally monitored in the Hengyang section of the Xiangjiang River, and the mean value and range (min-max) are shown (Table 4) in both the dry and rainy seasons. It is clear that heavy metals Pb, Hg, Cd, Zn, Cu, and the metalloids As, Se were all at levels of μg/L, much lower than DO, COD_Mn, COD_Cr, NH₄⁺-N, TP and F^-, which were all at levels of mg/L. Mean value concentrations of metals followed the order: Zn>Cu>As>Pb>Cd>Se>Hg, and the mean value of the other elements except pH and E. coli followed the order: COD_Cr>DO>COD_Mn>BOD₅>NH₄⁺-N≈ F^->TP. The mean value of E. coli was 4838.6 and 7147.1 CFU/L in dry and rainy seasons, respectively. Although E. coli generally meets the Chinese limit (10000 CFU/L), the possibility of exceeding this limit occurred more in the dry season (8.9%) than in the rainy season (8.3%). The highest E. coli levels occurred at S9 (70692 mg/L) in the rainy season and at S7 (5138.49 mg/L) in the dry season, which is respectively approximately 7 and 5 times higher than the Chinese limit. Except for E. coli, there was a minor difference in the mean value of other parameters between the dry and rainy seasons, which meets the Chinese limit and the WHO limit. However, the minimum and maximum values of metal(loid)s showed significant seasonal differences. In particular, the highest As occurred as 33.37 μg/L at S4 in the rainy season and as 20.77 μg/L at S12 in the dry season; the highest Zn occurred as 44 μg/L at S6 in the rainy season and as 136.67 μg/L at S10 in the dry season; and the highest Cu occurred as 25 μg/L at S9 and S10 in the rainy season and as 70 μg/L at S5 in the dry season.

Table 4 Water quality variables concentrations in Hengyang section of Xiangjiang River, China

It was observed that the maximum concentration of As meets the Chinese limit but is approximately 3.3 and 2.1 times higher than the WHO limit. The maximum concentration of other metal(loid)s meets both the Chinese and the WHO limits. The difference of As, Zn, and Cu between the rainy and dry seasons may result from less anthropogenic activity and more rainfall in the rainy season, and more agricultural and mineral activity in the dry season [61, 62]. Moreover, the very low concentration of Hg (0.014 μg/L) indicated that there were few Hg discharges into the river and low atmospheric deposition of Hg [63]. Excluding metal(loid)s, there are no WHO guidelines for other parameters, meaning no significant health risks. Therefore, metal(loid)s are of great concern to ensure water safety.

3.2 Multivariate statistical analysis

PCA was implemented based on the mean value of the 16 parameters revealing that there are six principal components in both rainy and dry seasons, explaining 73.03% and 75.3% of the total variance, respectively. As shown in Table 5, in the rainy season, PC1 contributes 16.76% of the total variance with loadings of NH₄⁺-N (0.80) and Cu (0.55), which may be attributed to a combination factor of nutrient input and chemical behavior. PC2 contributes 15.5% of the total variance and is characterized by loadings of COD_Mn (0.80), Hg (0.82), and Cd (0.61), which might be originated from smelting industries. PC3 explains 14.11% of the total variance of Se (0.75), which might be attributed to natural erosion. PC4 explains 10.8% of the total variance of DO (0.91) and BOD₅ (0.76), which may be attributed to nutrient input. PC5 explains 7.43% of the total variance of Zn (0.93), and PC6 explains 7.43% of the total variance of E. coli (0.77), which may be originated from the domestic sewage. In the dry season, PC1, accounting for 18.54% of the total variance, is comprised of COD_Mn (0.74), Zn (0.79), and Cu (0.67), which might be originated from smelting industries. PC2 explains 14.15% of the total variance of Hg (0.82) and Se (0.76), which might be originated from the anthropogenic activities and natural erosion. PC3 explains 11.56% of the total variance of E. coli (0.50) and TP (0.90), which may be attributed to nutrient input. PC4 explains 11.28% of the total variance of F^- (0.50) and Cd (0.80),which might be originated from smelting industries. PC5 explains 10.48% of the total variance of BOD₅ (0.77), Pb (0.62), and NH₄⁺-N (0.54), which may be attributed to a hybrid element of anthropogenic activities including metal and organic effects. PC6 explains 9.28% of the total variance of COD_Cr (0.86), which may be originated from chemical impacts. Therefore, there were differences in principal components between rainy season and dry season. The six principal components mainly consist of DO, E. coli, COD_Mn, BOD₅, NH₄⁺-N, Hg, Cd, Zn, Cu and Se in the rainy season, while consist of E. coli, COD_Mn, BOD₅ Hg, Cd, Zn, Cu, Se, COD_Cr, TP, F^- and Pb in the dry season. The differences between the rainy and dry seasons may result from the lower anthropogenic activity and more rainfall in the rainy season. In contrast, more agricultural and mineral activity may occur during the dry season [61]. Moreover, more variables found in the dry season indicate that water quantity would significantly influence water quality by changing the physical, chemical, and biological parameters.

Table 5 Principal components analysis for physical– chemicals of the Xiangjiang River, China

To further reveal and discuss the inner relationship of parameters and sites, dual hierarchical cluster analysis (DHCA) was then followed, and sampling sites and parameters were each divided into three categories. According to DHCA results in the rainy season (Figure 2), in the vertical dendrogram, Cluster 1 contains PC1, PC2,PC3, PC5, TP, and COD_Cr. Cluster 2 consists of PC4. Cluster 3 includes Pb, pH, F^-, As, and PC6. In the horizontal dendrogram, Cluster 1 contains S1, S2, S4, S6, S7, S8, and S12, and Cluster 3 contains S3, S5 and S11. Clusters 1 and 3 concentrated on the upper part of the Hengyang section of the Xiangjiang River. Cluster 2 consists of S9 and S10, which concentrated on the middle part of the Hengyang section of the Xiangjiang River. Based on results in the dry season (Figure 3), in the vertical dendrogram, Cluster 1 contains PC3, F^-, As, Pb, pH, and Hg. Cluster 2 consists of PC1, Se, COD_Cr, and Cd, and Cluster 3 includes BOD₅, NH₄⁺-N, and DO. In the horizonal dendrogram, Cluster 1 contains S1, S2, S4, S6, S7, S8, S11 and S12, and Cluster 3 contains S3 and S5. Clusters 1 and 3 concentrated on the upper and lower part of the Hengyang section of the Xiangjiang River. Cluster 2 consists of S9 and S10, which concentrated on the middle part of the Hengyang section of the Xiangjiang River.

Figure 2 Dual hierarchical cluster analysis results of water parameters and sampling sites for rainy seasons (a) and dry seasons (b)

It was found that the DHCA results of sampling sites were almost the same in the rainy and dry seasons. In particular, S3 and S5 are in the same cluster, and S9 and S10 are in the same cluster, and the other sites are in another cluster. We found that S6-S8 are located in an urban area of Hengyang City, which would more likely arise the pollution from domestic sewage. S1-S2 are near the Shuikoushan industrial area, which would likely arise metal-related pollution. The clusters for water parameters were generated by the similarities and the interplay of elements [52], which may provide the classification of the pollution source. The main pollution source that we concluded from our research were smelting industries, domestic sewage, and natural erosion. Comparing the clusters of water parameters in the rainy and dry seasons, BOD₅ and DO are in the same cluster, which may be due to the fact that their inverse relationships were widely found in river water [64]. In addition, metal(loid)s Cu, Cd, Zn, and Se are in the same cluster, together with COD_Mn and COD_Cr. The cluster of pH and E. coli also includes typical heavy metal(loid)s As, Pb, and Hg, implying that E. coli is related to these heavy metal(loid)s. Therefore, the pollution of Cu, Cd, Zn, and Se may originate from the same source, while As, Pb, and Hg have another pollution source. The clusters for sampling sites were extracted to show geographical features, which can provide a reliable classification of sampling sites and may support to establish an optimal sampling strategy in the future by saving costs of building sampling sites [48]. Metals, including heavy metal(loid)s, can be hazardous not only to aquatic life but also to humans exposed to the water, so it’s necessary to consider the human health risks caused by metal(loid)s with different seasons being considered.

3.3 Human health risk assessment

According to the above results, metals including heavy metals (namely, Pb, As, Hg, Cd, Zn, Cu, and Se) are the main parameters of potential pollution. Nutrient factors such as NH₄⁺-N, and TP as well as biochemical factors such as E. coli and BOD₅, are very detrimental to water quality. Still, they could be degraded under natural erosion. However, metals are more likely to be gathered in the water or sediment of the river water, which poses a threat to residents living around. Table 6 presents the results of human health risk assessment, including HQ from oral ingestion and dermal absorption, HI and CR.

For both the dry and rainy seasons, HQ_ingestion of As is close to 1 for adults and higher than 1 for children. Followed by Cd, its HQ_ingestion reaches 10^-2 order of magnitude for both adults and children, and HQ_ingestion of Cu also reaches the same magnitude only for children. This indicates that As and Cd should be seriously concerned about their risks to human health, and Cu is also of concern for children. In addition, HQ_dermal of Pb, Hg, Cd, As, Zn, Cu, and Se are far below 1 in both the rainy and dry seasons, indicating that there are low potential risks through dermal absorption. It was noticed that the hazards via dermal absorption follow the order of As, Cd>Hg>Cu, Se>Pb, Zn. Comparing HQ_ingestion and HQ_dermal, since only HQ_ingestion of As is greater than 1, we conclude that As posed severe health concerns to the local residents through oral ingestion. For the HI results, As had a value higher than 1 for children in both the rainy and dry seasons, indicating that As is the most critical hazard in the Xiangjiang River. The HI values for Cd for both adults and children and HI of Cu for children reached 10^-2 order of magnitude, and HI values of other metals are of 10^-3 order of magnitude. Similar to the HQ_ingestion result, As and Cd are of concern for adults, and along with Cu are of concern for children. As shown in Table 6, CR of As for children is slightly higher than for adults in both the rainy and dry seasons but is in the reasonable range (10^-6-10^-4) recommended by the USEPA. Other studies find similar results in the sediments or soils of this river basin [65-68]. They also found the aged people and left-behind children are all faced with the carcinogenic risk from the surface water in the Middle Chinese Loess Plateau [69]. Among the water risk research, LI et al [70] and WANG et al [71] discovered that As and Cr contribute the most human health risks in the Shahu Lake Tourist Area or Ordos basin of northwest China, which is very similar to our findings in the Xiangjiang River. However, uncertainties existing in the risk characterisation methodology were mentioned by the USEPA and other documents. Uncertainties resulting from the quantification of water and dermal contact factors, varied exposure conditions due to different ages and receptors, and temporal and spatial variations in contaminant concentrations. Despite the uncertainties, the comprehensive research of water quality assessment should be noticed, and WU et al [72] also found that the health risk assessment should be carried out together with general groundwater quality assessment in order to improve the reliability.

Table 6 Human health risk assessment results

4 Conclusions

A comprehensive investigation of water quality in the Hengyang section of the Xiangjiang River was provided by analyzing different types of parameters through multiple analysis approaches.

1) The results show that almost all parameter values were higher in the dry season, and these higher values may be caused by decreased rainfall and increased anthropogenic activity. Mean values of metals followed the order: Zn>Cu>As>Pb>Cd> Se>Hg, and the mean values of the other parameters, except pH and E. coli, followed the order: COD_Cr > DO>COD_Mn>BOD₅>NH₄⁺-N≈F^->TP. E. coli raised concerns because its concentration value is approximately 7 and 5 times higher than the China limit in dry seasons and rainy seasons, respectively. It also needs special concern because its highest value is very close to the Chinese limit.

2) In addition, PCA in multivariate analyses extracted six main components, consisting of DO, E. coli, COD_Mn, BOD₅, NH₄⁺-N, Hg, Cd, Zn, Cu and Se in the rainy season and consisting of E. coli, COD_Mn, BOD₅, Hg, Cd, Zn, Cu, Se, COD_Cr, TP, F^-, and Pb in the dry season. These findings indicate the water quantity would significantly influence water quality by changing the physical, chemical, and biological parameters.

3) DHCA further divided sampling sites and water quality parameters into three categories allowing the classification of the pollution source, which explains the correlation of parameters and may inform the building of future monitoring sites.

4) Moreover, the human health risk assessment indicates that As was the most important pollutant, and Cd was of concern for adults and, along with Cu, was of concern for children. The hazards via dermal absorption follow the order of As, Cd>Hg> Cu, Se>Pb, Zn. However, uncertainties in human risk assessment should be of concern in the future. Therefore, the government should provide comprehensive pollution alleviation strategies to control industrial and urban pollution with particular attention paid to As and E. coli.

References

[1] CHAI Li-yuan, WANG Zheng-xing, WANG Yun-yan, YANG Zhi-hui, WANG Hai-ying, WU Xie. Ingestion risks of metals in groundwater based on TIN model and dose-response assessment-A case study in the Xiangjiang watershed, central-south China [J]. Science of the Total Environment, 2010, 408(16): 3118-3124. DOI: 10.1016/ j.scitotenv.2010.04.030.

[2] WANG Zheng-xing, CHAI Li-yuan, WANG Yun-yan, YANG Zhi-hui, WANG Hai-ying, WU Xie. Potential health risk of arsenic and cadmium in groundwater near Xiangjiang River, China: A case study for risk assessment and management of toxic substances [J]. Environmental Monitoring and Assessment, 2011, 175(1-4): 167-173. DOI: 10.1007/s10661-010-1503-7.

[3] CHAI Li-yuan, LI Huan, YANG Zhi-hui, MIN Xiao-bo, LIAO Qi, LIU Yi, MEN Shui-hui, YAN Ya-nan, XU Ji-xin. Heavy metals and metalloids in the surface sediments of the Xiangjiang River, Hunan, China: Distribution, contamination, and ecological risk assessment [J]. Environmental Science and Pollution Research, 2017, 24(1): 874-885. DOI: 10.1007/s11356-016-7872-x.

[4] MAS-PLA J, MENCIO A. Groundwater nitrate pollution and climate change: learnings from a water balance-based analysis of several aquifers in a western Mediterranean region (Catalonia) [J]. Environmental Science and Pollution Research, 2018, 26: 2184-2202. DOI: 10.1007/s11356- 018-1859-8.

[5] ADHIKARI A, HANSEN A J. Climate and water balance change among public, private, and tribal lands within greater wild land ecosystems across north central USA [J]. Climatic Change, 2019, 152: 551-567. DOI: 10.1007/s10584-018- 2351-7.

[6] NELLEMANN C, HAIN S, ALDER J. In dead water: merging of climate change with pollution, over-harvest, and infestations in the world’s fishing grounds [M]. Arendal: United Nations Environment Programme, 2008.

[7] KIBRIA G. World rivers in crisis: Water quality and water dependent biodiversity are at risk-threats of pollution, climate change & dams development [J]. Research Gate, 2015, 11: 1-11. DOI: 10.13140/RG.2.1.1791. 5365/2.

[8] FUNARI E, MANGANELLI M, SINISI L. Impact of climate change on waterborne diseases [J]. Annali dell’Istituto Superiore di Sanita, 2012, 48(4): 473-487. DOI: 10.4415/ANN_12_04_13.

[9] SHEAHAN D. Impacts of climate change on pollution [EB/OL]. [2006] http://www.mccip.org.uk/media/1399/ pollution-report-from-cefas.pdf.

[10] ALAM M J, DUTTA D. Predicting climate change impact on nutrient pollution in waterways: A case study in the upper catchment of the Latrobe River, Australia [J]. Ecohydrology, 2013, 6(1): 73-82. DOI: 10.1002/eco.282.

[11] LIN Hui-ju, CHEN Lei-lei, LI Hai-pu, LUO Zhou-fei, LU Jing, YANG Zhao-guang. Pharmaceutically active compounds in the Xiangjiang River, China: Distribution pattern, source apportionment, and risk assessment [J]. Science of the Total Environment, 2018, 636: 975-984. DOI: 10.1016/j.scitotenv.2018.04.267.

[12] WU Yi-ping, LIU Shu-guang, YAN Wen-de, XIA Jiang-zhou, XIANG Wen-hua, WANG Ke-lin, LUO Qiao, FU Wei, YUAN Wen-ping. Climate change and consequences on the water cycle in the humid Xiangjiang River Basin, China [J]. Stochastic Environmental Research and Risk Assessment, 2016, 30(1): 225-235. DOI: 10.1007/s00477-015-1073-x.

[13] SIVAKUMAR B. Global climate change and its impacts on water resources planning and management: assessment and challenges [J]. Stochastic Environmental Research and Risk Assessment, 2011, 25(4): 583-600. DOI: 10.1007/ s00477-010-0423-y.

[14] CAI Bei-ming, ZHANG Bing, BI Jun, ZHANG Wen-jing. Energy’s thirst for water in china [J]. Environmental Science and Technology, 2014, 48(20): 11760-11768. DOI: 10.1021/es502655m.

[15] TANG Jing-wen, CHAI Li-yuan, LI Huan, YANG Zhi-hui, YANG Wei-chun. A 10-year statistical analysis of heavy metals in river and sediment in Hengyang segment, Xiangjiang river basin, China [J]. Sustainability, 2018, 10(4): 1057. DOI: 10.3390/su10041057.

[16] FEI Jiang-chi, MIN Xiao-bo, WANG Zhen-xing, PANG Zhi-hua, LIANG Yan-jie, KE Yong. Health and ecological risk assessment of heavy metals pollution in an antimony mining region: A case study from South China [J]. Environmental Science and Pollution Research, 2017, 24(35): 27573-27586. DOI:10.1007/s11356-017-0310-x.

[17] TANG Jing-wen, LIAO Ying-ping, YANG Zhui-hui, CHAI Li-yuan, YANG Wei-chun. Characterization of arsenic serious-contaminated soils from Shimen realgar mine area, the Asian largest realgar deposit in China [J]. Journal of Soils and Sediments, 2016, 16(5): 1519-1528. DOI: 10.1007/s11368-015-1345-6.

[18] WANG Zhen-xing, CHEN Jian-qun, CHAI Li-yuan, YANG Zhi-hui, HUANG Shun-hong, ZHENG Yu. Environmental impact and site-specific human health risks of chromium in the vicinity of a ferro-alloy manufactory, China [J]. Journal of Hazardous Materials, 2011, 190(1-3): 980-985. DOI: 10.1016/j.jhazmat.2011.04.039.

[19] Ministry of Environmental Protection of the People’s Republic of China. 2016 Report on the State of the Environment in China [R]. Beijing: MEP, 2017. (in Chinese)

[20] Ministry of Environmental Protection of the People’s Republic of China. The water pollution prevention and control plan for major River Basins (2016-2020) [R]. Beijing: MEP, 2017. (in Chinese)

[21] ZHANG Lei, QIN Yan-wen, ZHENG Bing-hui, LIN Tian, LI Yuan-yuan. Polycyclic aromatic hydrocarbons in the sediments of Xiangjiang River in south-central China: occurrence and sources [J]. Environmental Earth Sciences, 2013, 69(1): 119-125. DOI: 10.1007/s12665-012-1939-x.

[22] LI Z. “No.1 Key Project” Controlling Xiangjiang pollution [EB/OL]. [2016]. http://politics.gmw.cn/2016-04/26/content_ 19860306.htm.

[23] ZENG Xiao-xia, LIU Yun-guo, YOU Shao-hong, ZENG Guang-ming, TAN Xiao-fei, HU Xin-jiang, XI Hu, HUANG Lei, LI Fei. Spatial distribution, health risk assessment and statistical source identification of the trace elements in surface water from the Xiangjiang River, China [J]. Environmental Science and Pollution Research, 2015, 22(12): 9400-9412. DOI: 10.1007/s11356-014-4064-4.

[24] ZENG G M, YUAN X Z, YIN Y Y, YANG C P. A two- dimensional water-quality model for a winding and topographically complicated river [J]. Journal of Environmental Management, 2001, 61(1): 113-121. DOI: 10.1006/jema.2000.0401.

[25] WU Yun-qing, SHAO Dong-guo, XIAO Yi. A comprehensive benefit evaluation for the Xiangjiang River basin rehabilitation project [C]// Proceedings of the International Association of Hydrological Sciences and the International Water Resources Association Conference. Guangzhou, China: International Association of Hydrological Science, 2008: 8-10.

[26] National Health Commission of the People’s Republic of China. Standard examination method for drinking water- collection and preservation of water samples (GB/T 5750.2-2006) [R]. Beijing: NHC, 2006. (in Chinese)

[27] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of total arsenic-silver diethyldithiocarbamate spectrophotometric method (GB/T7485-1987) [R]. Beijing: MEP, 1987. (in Chinese)

[28] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of copper, zinc, lead and cadmium-atomic absorption spectrometry (GB/T 7475-1987) [R]. Beijing: MEP, 1987. (in Chinese)

[29] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of total mercury-cold atomic absorption spectrophotometry (HJ597-2011) [R]. Beijing: MEP, 2011. (in Chinese)

[30] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of 65 elements-inductively coupled plasma-mass spectrometry (HJ700-2014) [R]. Beijing: MEP, 2014. (in Chinese)

[31] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of Ph value- glass electrode method (GB/T 6920-1986) [R]. Beijing: MEP, 1986. (in Chinese)

[32] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of dissolved oxygen-electrochemical probe method (HJ 506-2009) [R]. Beijing: MEP, 2009. (in Chinese)

[33] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of permanganate index (GB/T 11892-1989) [R]. Beijing: MEP, 1989. (in Chinese)

[34] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of chemical oxygen demand-dichromate method (GB/T 11914-1989) [R]. Beijing: MEP, 1989. (in Chinese)

[35] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of biochemical oxygen demand after 5 days-dilution and seeding method (HJ 505-2009) [R]. Beijing: MEP, 2009. (in Chinese)

[36] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of ammonia nitrogen-salicylic acid spectrophotometry (HJ 536-2009) [R]. Beijing: MEP, 2009. (in Chinese)

[37] Ministry of Environmental Protection of the People’s Republic of China. Water quality-determination of total phosphorus-ammonium molybdate spectrophotometric method (GB/T 11893-1989) [R]. Beijing: MEP, 1989. (in Chinese)

[38] Ministry of Environmental Protection of the People’s Republic of China. Water quality-Determination of inorganic anions-Ion chromatography method (HJ/T84-2001) [R]. Beijing: MEP, 2001. (in Chinese)

[39] Ministry of Environmental Protection of the People’s Republic of China. Water and wastewater monitoring and analysis method [R]. Fourth edition. Beijing: China Environmental Science Press, 2002. (in Chinese)

[40] BENGRAINE K, MARHABA T F. Using principal component analysis to monitor spatial and temporal changes in water quality [J]. Journal of Hazardous Materials, 2003, 100(1-3): 179-195. DOI: 10.1016/S0304-3894(03)00104-3.

[41] WU Jian-hua, LI Pei-yue, QIAN Hui, DUAN Zhao, ZHANG Xue-di. Using correlation and multivariate statistical analysis to identify hydrogeochemical processes affecting the major ion chemistry of waters: Case study in Laoheba phosphorite mine in Sichuan, China [J]. Arabian Journal of Geosciences, 2014, 7(10): 3973-3982. DOI: 10.1007/s12517-013-1057-4.

[42] WU Jian-hua, LI Pei-yue, WANG Dan, REN Xiao-fei, WEI Miao-jun. Statistical and multivariate statistical techniques to trace the sources and affecting factors of groundwater pollution in a rapidly growing city on the Chinese Loess Plateau [J]. Human and Ecological Risk Assessment, 2019: 1-19. DOI: 10.1080/10807039.2019.1594156.

[43] LI Pei-yue, TIAN Rui, LIU Rong. Solute geochemistry and multivariate analysis of water quality in the Guohua phosphorite mine, Guizhou province, China [J]. Exposure and Health, 2019, 11(2): 81-94. DOI: 10.1007/s12403- 018-0277-y.

[44] TZIRITIS E P, DATTA P S, BARZEGAR Rahim. Characterization and assessment of groundwater resources in a complex hydrological basin of central Greece (Kopaida basin) with the joint use of hydrogeochemical analysis, multivariate statistics and stable isotopes [J]. Aquatic Geochemistry, 2017, 23(4): 271-298. DOI: 10.1007/ s10498-017-9322-x.

[45] BARZEGAR R, ASGHARI M A, ADAMOWSKI J, HOSSEIN A. Assessing the potential origins and human health risks of trace elements in groundwater: A case study in the Khoy plain, Iran [J]. Environmental Geochemistry and Health, 2019, 41: 981-2001. DOI: 10.1007/s10653-018- 0194-9.

[46] BARZEGAR R, ASGHARI M A, SOLTANI S, FIJANI E, TZIRITIS E, KAZEMIAN N. Heavy metal(loid)s in the groundwater of Shabestar Area (NW Iran): Source identification and health risk assessment [J]. Exposure and Health, 2019, 11: 251-265. DOI: 10.1007/s12403-017- 0267-5.

[47] SIMEONOV V, STRATIS J A, SAMARA C, ZACHARIADIS G, VOUTSA D, ANTHEMIDIS A,SOFONIOU M, KOUIMTZZIS T. Assessment of the surface water quality in Northern Greece [J]. Water Research, 2003, 37(17): 4119-4124. DOI: 10.1016/S0043-1354(03) 00398-1.

[48] ZHANG Qi, LI Zhong-wu, ZENG Guang-ming, LI Jiang-bing, YUAN Qing-shui, WANG Ya-mei, YE Fang-yi. Assessment of surface water quality using multivariate statistical techniques in red soil hilly region: a case study of Xiangjiang watershed, China [J]. Environmental Monitoring and Assessment, 2009, 152(1-4): 123-131. DOI: 10.1007/ s10661-008-0301-y.

[49] SINGH K P, MALIK A, MOHAN D, SINHA S. Multivariate statistical techniques for the evaluation of spatial and temporal variations in water quality of Gomti River (India)—A case study [J]. Water Research, 2004, 38(18): 3980-3992. DOI: 10.1016/j.watres.2004.06. 011.

[50] LI Si-yue, ZHANG Quan-fa. Risk assessment and seasonal variations of dissolved trace elements and heavy metals in the Upper Han River, China [J]. Journal of Hazardous Materials, 2010, 181(1-3): 1051-1058. DOI: 10.1016/ j.jhazmat.2010.05.120.

[51] LIU Chen-wuing, LIN Kao-hung, KUO Yi-ming. Application of factor analysis in the assessment of groundwater quality in a blackfoot disease area in Taiwan [J]. Science of the Total Environment, 2003, 313(1-3): 77-89. DOI: 10.1016/S0048-9697(02)00683-6.

[52] LAURSEN J, MILMAN N, PIND N, PEDERSEN H, MULVAD G. The association between content of the elements S, Cl, K, Fe, Cu, Zn and Br in normal and cirrhotic liver tissue from Danes and Greenlandic Inuit examined by dual hierarchical clustering analysis [J]. Journal of Trace Elements in Medicine and Biology, 2014, 28(1): 50-55. DOI: 10.1016/j.jtemb.2013.08.003.

[53] RANJBAR J A, RIYAHI B A, SHADMEHRI T A, JADOT C. Spatial distribution, ecological and health risk assessment of heavy metals in marine surface sediments and coastal seawaters of fringing coral reefs of the Persian Gulf, Iran [J]. Chemosphere, 2017, 185: 1090-1111. DOI: 10.1016/ j.chemosphere.2017.07.110.

[54] WU B, ZHAO D Y, JIA H Y, ZHANG Y, ZHANG X X, CHENG S P. Preliminary risk assessment of trace metal pollution in surface water from Yangtze River in Nanjing section, China [J]. Bulletin of Environmental Contamination and Toxicology, 2009, 82(4): 405-409. DOI: 10.1007/ s00128-008-9497-3.

[55] Ministry of Environmental Protection of the People’s Republic of China. Technical guidelines for risk assessment of contaminated sites (HJ25.3-2014) [R]. Beijing: MEP, 2014. http://kjs.mee.gov.cn/hjbhbz/bzwb/jcffbz/201402/t20140226_268358.shtml. (in Chinese)

[56] United States Environmental Protection Agency (USEPA). Risk assessment guidance for superfund (RAGS). Volume I. Human health evaluation manual (HHEM). Part E. Supplemental guidance for dermal risk assessment [EB/OL]. [2004-07]. https://www.epa.gov/risk/risk-assessment- guidance-superfund-rags-part.

[57] RODRIGUEZ-PROTEAU R, GRANT R. Toxicity evaluation and human health risk assessment of surface and ground water contaminated by recycled hazardous waste materials [M]// Kassim T.A. Water pollution. Berlin: Springer, 2005. DOI: 10.1007/b11434.

[58] OSBORNE P. Water pollution [J]. Utilities Law Review, 1999, 10(1): 17-19. DOI: 10.1002/(SICI)1099-1808(199901/ 02)10:1<17::AID-ULR120>3.0.CO;2-W.

[59] State Environment Protection Administration of China, Environment quality standard for surface water (GB3838-2002) [R]. Beijing: SEPAC, 2002. (in Chinese)

[60] World Health Organization. Guidelines for drinking-water quality, 4th edition, incorporating the 1st addendum [R]. Geneva, WHO, 2017. https://www.who.int/water_sanitation_ health/publications/drinking-water-quality-guidelines-4-including-1st-addendum/en/.

[61] GIRI S, SINGH A K. Risk assessment, statistical source identification and seasonal fluctuation of dissolved metals in the Subarnarekha river, India [J]. Journal of Hazardous Materials, 2014, 265: 305-314. DOI: 10.1016/j.jhazmat.2013.09.067.

[62] CHEN Yan, WANG Ling-qing, LIANG Tao, XIAO Jun, LI Jing, WEI Hai-cheng, LIN Lin. Major ion and dissolved heavy metal geochemistry, distribution, and relationship in the overlying water of Dongting Lake, China [J]. Environmental Geochemistry and Health, 2019, 41: 1091-1104. DOI: 10.1007/s10653-018-0204-y.

[63] QU Li-yin, HUANG Hong, XIA Fang, LIU Yuan-yuan, DAHLGREN Randy A, ZHANG Ming-hua, MEI Kun. Risk analysis of heavy metal concentration in surface waters across the rural-urban interface of the Wen-Rui Tang River, China [J]. Environmental Pollution, 2018, 237: 639-649. DOI: 10.1016/j.envpol.2018.02.020.

[64] NKWOJI J A, IGBO J K, ADELEYE A O, OBIENU J A, TONY-OBIAGWU M J. Implications of bioindicators in ecological health: Study of a coastal lagoon, Lagos, Nigeria [J]. Agriculture & Biology Journal of North America, 2010, 1(4): 683-689.

[65] LI Huan, CHAI Li-yuan, YANG Zhi-hui, LIAO Qi, LIU Yi, OUYANG Bin. Seasonal and spatial contamination statuses and ecological risk of sediment cores highly contaminated by heavy metals and metalloids in the Xiangjiang River [J]. Environmental Geochemistry and Health, 2019, 41: 1617-1633. DOI: 10.1007/s10653-019-00245-2.

[66] LI Huan, YANG Jin-qin, YE Bin, JIANG dong-yi. Pollution characteristics and ecological risk assessment of 11 unheeded metals in sediments of the Chinese Xiangjiang River [J]. Environmental Geochemistry and Health, 2019, 41: 1459-1472. DOI: 10.1007/s10653-018-0230-9.

[67] LI Fei, ZHANG Jing-dong, JIANG Wei, LIU Chao-yang, ZHANG Zhong-min, ZHANG Cheng-de, ZENG Guang- ming. Spatial health risk assessment and hierarchical risk management for mercury in soils from a typical contaminated site, China [J]. Environmental Geochemistry and Health, 2017, 39(4): 923-934. DOI: 10.1007/ s10653-016-9864-7.

[68] SUN Guang-yi, LI Zhong-gen, LIU Ting, CHEN Ji, WU Ting-ting, FENG Xin-bin. Rare earth elements in street dust and associated health risk in a municipal industrial base of central China [J]. Environmental Geochemistry and Health, 2017, 39(6): 1469-1486. DOI: 10.1007/s10653-017-9982-x.

[69] HE Xiao-dong, LI Pei-yue. Surface water pollution in the middle Chinese loess plateau with special focus on hexavalent chromium (Cr⁶⁺): Occurrence, sources and health risk [J]. Exposure and Health, 2020: 1-17. DOI: 10.1007/ s12403-020-00344-x.

[70] LI Pei-yue, FENG Wei, XUE Chen-yang, TIAN Rui, WANG Si-ting. Spatiotemporal variability of contaminants in lake water and their risks to human health: A case study of the Shahu lake tourist area, northwest China [J]. Exposure and Health, 2016, 9: 213-225. DOI: 10.1007/s12403- 016-0237-3.

[71] WU Jian-hua, ZHANG Yu-xin, ZHOU Hui. Groundwater chemistry and groundwater quality index incorporating health risk weighting in Dingbian County, Ordos basin of northwest China [J]. Geochemistry, 2020. DOI: 10.1016/ j.chemer.2020.125607.

[72] WU Jian-hua, ZHOU Hui, HE Song, ZHANG Yu-xin. Comprehensive understanding of groundwater quality for domestic and agricultural purposes in terms of health risks in a coal mine area of the Ordos basin, north of the Chinese Loess Plateau [J]. Environmental Earth Sciences, 2019, 78: 446. DOI: 10.1007/s12665-019-8471-1.

(Edited by ZHENG Yu-tong)

中文导读

水质及安全测评基于对营养元素、生化指标和重金属元素的多元统计分析

摘要：为更好地了解湘江的水质状况，并评估河水带来的风险，于2016年按月采样了湘江衡阳段的12个采样点，并分析了各采样点中7种重金属元素(Pb, Hg, Cd, As, Zn, Cu, Se)含量以及9种相关水质参数(pH, DO, E. coli, COD_Mn, COD_Cr, BOD₅, NH₄⁺-N, TP, F^-)。按照雨季和干旱季节分析了参数，并将其与中国标准和世界卫生组织饮用水限值进行了比较，以评估河水状况的可持续性。主成分分析(PCA)显示了不同季节的不同污染源。双重聚类分析(DHCA)对水质变量和采样地点进行了进一步分类。此外，引入了人类健康风险评估来评估重金属的健康风险。这项研究将有助于优化水质监测点的建设，并建立与维护公共卫生有关的污染修复策略。

关键词：表层水；水质；人类健康风险；湘江

Foundation item: Projects(2018YFC1801805, 2018YFC1903301) supported by National Key R&D Program of China; Project(51825403) supported by National Science Fund for Distinguished Young Scholars, China; Project(2019SK2281) supported by Key R&D Program of Hunan Province, China

Received date: 2020-01-14; Accepted date: 2020-04-07

Corresponding author: WANG Yun-yan, PhD, Professor; Tel: +86-13975808192; E-mail: wang.yunyan@outlook.com; ORCID: 0000- 0001-5842-2781