## Introduction

Water plays an important role in ecological systems, and is an indispensable component of human lives. As open environmental spaces, rivers are more likely to be polluted1. This issue has received a great deal of attention, because concentrations of heavy metals in rivers continued to rise above the regional background value under constant anthropogenic disturbances, and has caused potential impacts on human health and the ecological balance2,3,4,5,6. Heavy metals in watershed ecosystems are difficult to be degrade and characterized by strong concealment, resulting a lasting harm and difficult to removal7. They exhibited different degrees of risk, impacts by point and non-point surface pollution, and characteristics of bioaccumulation and anthropogenic dominant sources4. Most studies have focused on lakes, river basins, and bays3,4,7,8, and few have been conducted from the perspective of water bodies in mining areas.

Antimony (Sb) is the ninth most mined industrial metal in the world, and Sb and its compounds are considered to be emerging pollutants9. As the largest antimony mining area in the world, the Xikuangshan area (XKS) has hosted hundreds of antimony mining, processing, and smelting enterprises throughout history, and also produced a great deal of exposed waste residue disposal sites. The main mineral in the deposit is stibnite (Sb2S3), with trace amounts of pyrite (FeS2), pyrrhotite (Fel-xS), and sphalerite (ZnS)10. Many studies have been conducted on the spatial distribution, migration characteristics, bioavailability and risk assessment of soil heavy metals from XKS in recent years11,12,13,14,15,16,17, but researches on heavy metal pollution in water bodies were few, and only focused on the Sb and/or As elements4,11,14,18. Pollution of Sb and its associated metals in water bodies has seldom been studied.

Therefore, the goals of this study were (1) to systematically investigate the distributions of Sb and its associated elements in the surface water in XKS; (2) to assess the levels of the pollution using both the Nemerow multi-factor index (NI) and the modified heavy metal pollution index (m-HPI); (3) to use the ecological risk index (ERI) to assess the ecological hazards; and (4) to assess the health risks of heavy metals to adults and children through direct ingestion and skin contact. The results of this study provide a basis for the precise remediation of regional watersheds pollution and the protection of positive cycle of the ecological environmental in mining areas.

## Materials and methods

### Study area

The XKS is located in Lengshuijiang city, Hunan Province, China, with a total area of 70 km2 and an altitude of 200–400 m15. The terrain is high in the north and low in the south. This area has a subtropical continental monsoon climate, with an average annual temperature of 16.7 °C and an annual rainfall of 1354 mm11. The river system in this area is well developed, including the Qingfeng (QF) River and the Lianxi (LX) River (Fig. 1). The Qingfeng River runs through the north part of the XKS, where the Tanjiachong (TJC) stream and the Batangshan (BT) stream converge. The Lianxi River is located in the south part of the XKS, and it can be divided into upstream (Chuanshan creek (CS) and Feishuiyan (FSY)) and downstream (Lianxi River) components.

### Sampling and analysis

In order to evaluate the changes in the heavy metal concentrations in the surface water of the XKS area, a total of 35 samples with a depth of 0–5 cm were collected from the rivers in May 2020 according to the Technical Specifications Requirements for Monitoring of Surface Water and Waste Water (HJ/T91-2002) (Fig. 1). And 15 water samples were collected in QF river, including 9 samples in TJC tream (TJC-1–9), 3 samples in BT Stream (BT-10–12), and 3 samples in the confluence and downstream (QF-13–15). A total of 20 samples were collected in the LX river, including 11 samples from FSY creek (FSY-16–26), 7 samples from CS creek (CS-27–33) and 2 samples from the intersection and downstream (LX34–35). To ensure the location of sampling point are accurate, portable global positioning systems (GPS) were used. All samples were filtered using a pre-weighted fiberglass filter (0.45 um) to remove suspended particles, then stored in sealed containers at temperatures below 5 °C and shipped back to the laboratory.

Based on the monitoring data from the local environmental monitoring stations, as well as the local residents’ habits regarding domestic water and agricultural irrigation, the pH value, potassium permanganate index (CODMII), chemical oxygen demand (COD), five-day biochemical oxygen demand (BOD5), ammonia nitrogen (NH3–N), total phosphorus (TP), total nitrogen (TN), and heavy metal contents (Sb, As, Mn, Pb, Zn, Hg, and Cd) were determined. A Phs-3c pH meter was used to measure the pH value of the surface water. The CODMII of the water was determined using the acidic potassium permanganate method (GB 11892-1989). The COD was determined using dichromate method (HJ 828-2017), and the BOD5 was determined using dilution and inoculation tests (HJ 505-2009). The NH3-N, TP, and TN concentration were determined using Nessler's reagent colorimetric method, the ammonium molybdate spectrophotometric method, and alkaline potassium persulfate digestion UV spectrophotometry method, respectively. The Sb, As and Hg contents were determined using the atomic fluorescence method (AFS 8220), Mn and Zn contents determined using flame atomic absorption spectrophotometry (TAS-990AFG), and the Pb and Cd contents were determined using graphite atomic absorption spectrophotometry (AA240Z).

Throughout the sample collection, transportation, storage and testing processes, all of the operations were carried out in accordance with the Technical Specifications Requirements for Monitoring of Surface Water and Waste Water (HJ/T91-2002). The entire quality control process was implemented using 20% quality control samples and 10% parallel samples to ensure the precision, accuracy, and reproducibility of sample determination of the data.

### Heavy metal pollution in surface water

#### Modified heavy metal pollution index

The m-HPI method19 is proposed based on the heavy metal pollution index (HPI)20 and the heavy metal evaluation index (HEI)21. The HPI method takes the highest desirable concentration (MDL) and the maximum permissible concentration (MPL) of the different metals into consideration, while the HEI method uses the maximum allowable concentrations (MAC) of the metals to characterize the water quality. However, due to the update of the heavy metal supervised guideline, the categories of some metals have been redefined. Thus, the methods in the reference the previous guidelines have some limitations when applied in some fields, but the m-HPI does not5.

This m-HPI method uses the positive index (PI) and negative index (NI) to evaluate the water quality5. The PI reflects the contribution of metals to the water quality when the metal concentrations are lower than the threshold values. And when the PI values of two water samples are similar, the NI can further classify the quality grades. The m-HPI is calculated as follows:

$${\text{NI}} = \left\{ {\begin{array}{*{20}c} {0,} & {{\text{when}}\; n_{1} = 0} \\ {\mathop \sum \limits_{i = 1}^{{n_{1} }} {\text{mHPI}}_{i} } & {n_{1} > 0} \\ { - 1,} & {{\text{when}} \;M_{i} = {\text{DBL}}} \\ \end{array} } \right.$$
$$PI = \left\{ {\begin{array}{*{20}c} {0,} & {{\text{when}} \;n_{2} = 0} \\ {\mathop \sum \limits_{i = 1}^{{n_{1} }} {\text{mHPI}}_{i} } & {n_{2} > 0} \\ {0,} & {{\text{when}} M_{i} = DBL } \\ \end{array} } \right.$$

where, $${\text{mHPI}}^{i}$$ is the modified heavy metal pollution index, n is the number of total metals (n = 7), and n1 and n2 are to be treated as independent and mutually exclusive subsets such that n = n1 + n2. DBL denotes ‘below the detection limit’.

$${\text{m - HPI}}^{i} = \omega_{i} Q_{i}$$
$$\omega_{i} = \frac{{W_{i} }}{{\mathop \sum \nolimits_{i = 1}^{n} W_{i} }}$$
$$W_{i} = \frac{1}{{I_{i} }}$$
$$Q_{i} = \frac{{M_{i} - I_{i} }}{{I_{i} }}$$

$$\omega_{i}$$ is the relative weight of a factor, and is inversely proportional to the highest desirable concentration ($$I_{i}$$), $$M_{i}$$ is the measured concentration of metal $$i$$, and $$I_{i}$$ refers to the Chinese drinking water standard22 and the maximum allowable concentration of the World Health Organization (WHO) standard23. $$Q_{i}$$ is not a modulus, and it will be zero when $$M_{i}$$ < $$I_{i}$$, which means the samples pass the quality standard based on a specific parameter. When $$M_{i}$$ > $$I_{i}$$, $$Q_{i}$$ will be positive. For all of the metals $$M_{i}$$ = $$S_{i}$$, the upper limit of PI (UL) was calculated to be 54.51. The water quality was classified into four categories5: excellent (− 1 ≤ NI ≤ 0, PI = 0), very good (-1 ≤ NI ≤ 0, 0 ≤ PI ≤ UL/2), good (− 1 ≤ NI ≤ 0, UL/2 ≤ PI ≤ UL), and unacceptable (NI ≤ 0, PI > UL).

#### Nemerow index

The Nemerow multi-factor index is a comprehensive evaluation method 4,24, and it is calculated as follows:

$$NI = \left[ {\left\{ {\left( {M_{i} /I_{i} } \right)_{{{\text{mean}}}}^{2} + \left( {M_{i} /I_{i} } \right)_{\max }^{2} } \right\}/n} \right]^{\frac{1}{2}}$$

where, $$\left( {M_{i} /I_{i} } \right)_{{{\text{mean}}}}$$ and $$\left( {M_{i} /I_{i} } \right)_{{{\text{max}}}}$$ are the respective mean and maximum values of $$\left( {M_{i} /I_{i} } \right)$$ for metal $$i$$ in water. NI can be divided into four grades4,24: no pollution (NI < 1), slight pollution (1 ≤ NI < 2.5), moderate pollution (2.5 ≤ NI < 7), and severe pollution (NI ≥ 7).

### Ecological risk

The ERI was used to evaluate the potential ecological risk posed by the heavy metals in the surface water system 4,25. It is calculated as follows:

$${\text{ERI}} = \mathop \sum \limits_{i = 1}^{n} \left[ {T_{i} \times \left( {M_{i} /I_{i} } \right)} \right]$$

where, $$T_{i}$$ is the biological toxicity factor of metal $$i$$ (Sb = 5, As = 10, Mn = 1, Pb = 5, Zn = 1, Hg = 40, and Cd = 30)25,29. The ERI is divided into four categories: low risk (ERI < 110), moderate risk (110 ≤ ERI < 200), considerable risk (200 ≤ ERI < 400), and very high risk (ERI ≥ 400).

### Human health risk assessment

Health risk assessment of heavy metals in aquatic environments usually considers two pathways, direct ingestion and dermal contact4,26,27. The daily exposure doses of direct ingestion ($${\text{ADD}}_{{{\text{ing}}}}$$) and dermal absorption ($${\text{ADD}}_{{{\text{derm}}}}$$) for adults and children are:

$${\text{ADD}}_{{{\text{ing}}}} = \frac{{M_{i} \times {\text{IR}} \times {\text{ABS}}_{i} \times {\text{EF}} \times {\text{ED}}}}{{{\text{BW}} \times {\text{AT}}}}$$
$${\text{ADD}}_{{{\text{derm}}}} = \frac{{M_{i} \times {\text{SA}} \times {\text{KP}}_{i} \times F \times {\text{EF}} \times {\text{ET}} \times {\text{CF}} \times {\text{ED}}}}{{{\text{BW}} \times {\text{AT}}}}$$

where, $$M_{i}$$ is the measured concentration of metal $$i$$ (mg/L); $$IR$$ is the intake rate (L/day, 2.0 for adults and 0.64 for children). $$SA$$ is the exposed skin area (in cm2) exposed to surface water (18,000 for adults and 6,600 for children); $$ET$$ is the exposure time (hours/day, 0.58 for adults, 1.00 for children); $$ED$$ is the duration of exposure (70 years for adults and six years for children). $$EF$$ is the exposure frequency (days/year), and its value is 350; $$BW$$ is the bodyweight of the residents (65 kg for adults and 20 kg for children). $$AT$$ is the average exposure time in days (25,550 days for adults and 2190 days for children). $$KP_{i}$$ is the dermal permeability coefficient of metal $$i$$ (Table 1). $$ABS_{i}$$ is the absorption factor of the stomach and intestines (Table1). $$F$$ is the ratio of the skin contact surface with water, and its value is 0.9. $$CF$$ is the volume conversion coefficient for water (1L/1000 cm3,10–3). All of the parameter values except $$M_{i}$$ are reference with the related articles 4,26,27,28.

The total potential non-carcinogenic risk (HI) is evaluated using the hazard quotient (HQ) for non-carcinogenic metals. When HQ or HI > 1, the adverse effects on human health should be considered 28.

$${\text{HQ}}_{i} = \left( {{\text{ADD}}_{{{\text{ing}}}}^{i} /RfD_{{{\text{ing}}}}^{i} } \right) + \left( {{\text{ADD}}_{{{\text{derm}}}}^{i} /RfD_{{{\text{derm}}}}^{i} } \right)$$
$$HI = \mathop \sum \limits_{i = 1}^{n} HQ_{i}$$

$${\text{HQ}}_{i}$$ is the hazard quotient of metal $$i$$$$RfD_{ing}^{i}$$ and $$RfD_{{{\text{derm}}}}^{i}$$ are the reference doses of metal $$i$$ through direct ingestion and dermal contact, respectively, and $$RfD_{{{\text{derm}}}}^{i} = RfD_{ing}^{i} \times {\text{ABS}}_{i}$$.

The total carcinogenic risk (TCR) was evaluated using the single health risk index (CR) for carcinogenic metals. When CR > 0.1, the cancer risk is very high; 10–3 < CR ≤ 0.1 indicates a high risk; 10–4 < CR ≤ 10–3 indicates a moderate risk; 10–6 < CR ≤ 10–4 indicates a low risk; and CR ≤ 10–6 indicates an ignorable risk 26,27,28..

$${\text{CR}}_{i} = \left( {{\text{ADD}}_{{{\text{iderm}}}}^{i} \times {\text{SF}}_{{{\text{derm}}}}^{i} } \right) + \left( {{\text{ADD}}_{{{\text{derm}}}}^{i} \times {\text{SF}}_{{{\text{derm}}}}^{i} } \right)$$
$${\text{TCR}} = \mathop \sum \limits_{i = 1}^{n} {\text{CR}}_{i}$$

$${\text{CR}}_{i}$$ is the carcinogenic risk of metal $$i$$$$SF_{ing}^{i}$$ and $$SF_{derm}^{i}$$ are the slope factors of metal $$i$$ through direct ingestion and dermal contact, respectively; and $$SF_{derm}^{i} = SF_{ing}^{i} /ABS_{i}$$.

## Results

### Changes in water quality indices and heavy metal concentrations in XKS

The results of the water quality indices and heavy metal concentrations of the surface water in XKS are presented in Table 2. The pH of the water was weakly alkaline (7.17–9.08). The average CODMII, COD, BOD5, NH3–N, TP, and TN values were 1.79 mg/L (0.90–3.92 mg/L), 13.51 mg/L (9.00–16.00), 2.91 mg/L (2.00–3.60 mg/L), 0.13 mg/L (0–0.40 mg/L), 0.48 mg/L (0–4.34 mg/L), and 2.58 mg/L (0–4.34 mg/L), respectively. Compared to the class II quality standard for surface water in China, the CODMII, COD, BOD5 and NH3-N values met the quality requirements (Table 2), while the TN and TP values were far beyond the standard of class II and they even exceeded class V (Table 2).

The concentrations of Sb, As, Mn, Pb, Zn, Hg, and Cd in the surface water ranged from 0.0009 to 5.33 mg/L, BDL to 19.60 mg/L, ~ DBL, ~ DBL, DBL to 0.44 mg/L, DBL to 0.0053 mg/L, and DBL to 0.0036 mg/L, respectively. The mean concentrations of Sb (1.05 mg/L), As (1.06 mg/L), and Hg (0.00084 mg/L) were higher than the class III limits of the surface water quality standard, and some were even beyond the class V limits, while Zn (0.01 mg/L) and Cd (0.00013 mg/L) met the standards (Table 2). Compared with previous studies on surface water in XKS, the concentration of Sb and As is higher than that reported by Zhu Jing et al. (2009) (Sb: 5.93–7.56 mg/L, As: 0.0011–0.0073 mg/L) and Fu et al. (2016) (Sb: DBL ~ 0.16 mg/L, As: DBL ~ 0.01 mg/L), but lower than Guo et al.12 (Sb: 0.016 ~ 38.29 mg/L, As: 0 ~ 0.50 mg/L), Guo et al.11 Sb: 0.005 ~ 309 mg/L, As: 0 ~ 18.39 mg/L). Obviously, the surface water pollution in XKS has been aggravated with years of mining activity. some comprehensive treatment measures in recent years can effectively curb the aggravation of pollution, but they cannot completely eliminate the pollution.

Pearson correlation analysis (Table 3) of the metals and related water quality indexes revealed that Sb was significantly positively correlated with As (r = 0.707, P < 0.01) and Hg (r = 0.536, P < 0.01), and TN was positively correlated with Sb in the surface water (r = 0.391, P < 0.01). The pH-As (r = 0.643, P < 0.01) and TP–Sb (r = 0.785, P < 0.05) were significantly positively correlated. From the perspective of the spatial distribution, the variations in the surface water quality indexes (Fig. 2a) were also consistent with the metal concentrations (Fig. 2b). However, the heavy metal concentrations and types were different in the rivers due to the distribution of the smelters and waste dumps. The As content of the watershed of Qingfeng River was significantly higher than the Sb content, while the opposite was observed in the Lianxi River watershed. The metal concentrations of the following samples were high: samples 5 to 9 from the lower reaches of Tanjiachong stream close to the waste rock pile; samples 18 to 26 from the middle to lower reaches of Feishuiyan stream where it flows past a smelting plant; and samples 29 to 30 from the left tributaries of the upper reaches of Chuanshan Stream adjacent to the smelter and the waste rock pile.

### Heavy metal pollution assessment

The NI (the Nemerow index) and m-HPI methods were used to evaluate the water pollution by heavy metals in XKS (Fig. 3). Based on the m-HPI water quality scale, the water quality of 2.85–88.57% of the sampling sites in the study area was good to very good, and only 8.57% of the sites had the worst water quality. The NI values ranged from 0.90 to 407.49, with an average of 80.80. According to the NI water quality classification standard, about 5.70% of the sampling sites were insignificantly to lightly polluted, and 7.14% and 87.16% were moderately and severely polluted, respectively. The pollution level assessed using the NI method was more severe than that assessed using the m-HPI method (Fig. 3), which is contrary to the results for groundwater quality in Indian using the same methods4. In the evaluation using the NI, the influence of the maximum value will be overemphasized and the other factors will be artificially ignored30. Sb and As are the most important pollutants in the area. Several studies have shown that Sb contents of food, water and soil are several times greater than their As contents11. Therefore, the impact of the high Sb content may be exaggerated, while those of other trace metals may be weaker. However, both results indicate that the quality of samples 23, 29, and 30 was the worst, and the pollution level was far higher than that of the other samples. It is necessary to be alert to the release of metals in these areas.

### Ecological risk assessment

The ERI values in XKS ranged from 22.69 to 7351.20, with an average of 1200.40. In the study area, 17.14% of the samples had a low ecological risk, 11.43–65.71% of the samples had a moderate to high risk, and more than 50% had a very high risk (Fig. 4). Sb and As were the main pollutants, followed was Hg. The biotoxicity factors of Hg and Cd were the highest, with contributions of 43.48% and 32.61%, respectively, so they are the main risk elements. The detection rates of Hg and Cd in the water were 91.43% and 8.57%, respectively. The Cd contents in the water were low in most places, but at sampling sites 23, 29, and 33, Cd posed a threat to the ecology, with ERI values of ≥ 110. The contributions of As, Sb, and Pb were 10.87%, 5.43%, and 5.43%, respectively, and their detection rates in the surface water were 94.29%, 100%, and 0.00%, respectively. In addition, 71.43% and 17.14% of the samples had a moderate or greater ecological risk related to Sb and As (ERI ≥ 110), respectively. Studies have shown that Sb and As in surface water are toxic to organisms and humans at a certain level11,31. In particularly, near the antimony mining area, the acute and chronic potential risk levels of Sb in the surface water reached 35.00% and 40%, respectively, posing a serious threat to the water ecosystem31.