Mercury Biogeochemical Cycle in Yanwuping Hg Mine and Source Apportionment by Hg Isotopes

Although mercury (Hg) mining activities in the Wanshan area have ceased, mine wastes remain the primary source of Hg pollution in the local environment. To prevent and control Hg pollution, it is crucial to estimate the contribution of Hg contamination from mine wastes. This study aimed to investigate Hg pollution in the mine wastes, river water, air, and paddy fields around the Yanwuping Mine and to quantify the pollution sources using the Hg isotopes approach. The Hg contamination at the study site was still severe, and the total Hg concentrations in the mine wastes ranged from 1.60 to 358 mg/kg. The binary mixing model showed that, concerning the relative contributions of the mine wastes to the river water, dissolved Hg and particulate Hg were 48.6% and 90.5%, respectively. The mine wastes directly contributed 89.3% to the river water Hg contamination, which was the main Hg pollution source in the surface water. The ternary mixing model showed that the contribution was highest from the river water to paddy soil and that the mean contribution was 46.3%. In addition to mine wastes, paddy soil is also impacted by domestic sources, with a boundary of 5.5 km to the river source. This study demonstrated that Hg isotopes can be used as an effective tool for tracing environmental Hg contamination in typical Hg-polluted areas.


Introduction
Mercury (Hg) is a highly toxic heavy metal that can travel a long distance in the atmosphere and is therefore considered a global pollutant [1]. The toxicity of Hg depends on its chemical form. The elevated levels of Hg in the air are mostly attributed to industrial emissions, such as coal burning, Hg mining, gold mining, wastes incinerators, and cement production [2]. Methylmercury (MeHg) is neurotoxic, and it has the ability to bioaccumulate and become ultimately biomagnified in the food web. Humans are exposed to MeHg mainly through the consumption of food [3][4][5]. The Minamata Convention went into effect in August 2017 to reduce the effects of Hg exposure on human health [6,7].
The Wanshan Hg Mine is considered the "capital of Hg" in China. Since 2002, mining activities have been banned at the site due to the depletion of Hg resources and the environmental implications [8,9]. However, long-term Hg mining activities have produced a large amount of mine wastes, which are an important source of Hg pollution in the surrounding atmosphere and surface water. Most Hg calcine piles are distributed at the source of the river. Under external forces, such as rainwater leaching, surface runoff, and wind erosion, the Hg from the mine wastes is released and enters the downstream water system [10][11][12]. Therefore, evaluating the ecological risks caused by Hg mines is crucial for local ecological restoration.
The mine wastes from Hg mines can diffuse into the surrounding environment through water and atmospheric transportation. The paddy soils are more heavily contaminated by Hg in Hg mining areas compared with other areas [13]. Among the crops grown in the

Study Area
The Wanshan Hg Mine is located in Guizhou Province, southwest China ( Figure 1a). Mineralization at the Wanshan Hg Mine is primarily associated with thin-layered, laminated, fine-grained, dolomite or limestone beds of the mid-Cambrian age. The wall rocks are intensively altered by silicification, dolomitization, calcification, subordinate bituminization, and pyritization [32]. The primary ore mineral in the Hg deposits is cinnabar, with less metacinnabar [33].The Yanwuping Hg Mine (YMM) is one of the largest Hg mines in the Wanshan area. The Yanwuping Hg Mine is hilly and karstic, and it is located at an altitude of 340-1010 m. The climate is subtropical humid, with an annual rainfall of 1200-1400 mm and an annual temperature of 15 • C [34].
The Yanwuping Hg Mine's historic Hg extraction facility and about 3.1 × 10 5 m 3 of mine wastes are located at the upper Wengman River [35]. In 2011, the government renovated the YMM and tailing dams, but 1.3 × 10 4 m 2 of the calcine deposits remained. The Wengman River (Figure 1b) originates in the YMM zone and belongs to the Yangtze River basin, which has an average summer depth of 1 m and is directly affected by upstream mine wastes [36].
Mine wastes, surface-layer soils, and deep-layer soils were collected at Yanwuping Mercury Mine. W1-16 and S1-6 were the water and paddy soils sampling sites, respectively.
The total gaseous Hg sampling sites were the same as those for mine wastes, soils, and paddy soils. Mine wastes, surface-layer soils, and deep-layer soils were collected at Yanwuping Mercury Mine. W1-16 and S1-6 were the water and paddy soils sampling sites, respectively.
The total gaseous Hg sampling sites were the same as those for mine wastes, soils, and paddy soils.

Sample Collection
Water and atmospheric samples were collected and monitored twice, in December 2021 and August 2022, due to the high seasonal variability of the various indicators in the river water and atmosphere. The interannual variability in the soil Hg is not significant. Thus, soil and mine wastes samples were collected only once, in December 2021. There are two main types of mine wastes: calcines, the residues of Hg ore after high-temperature calcination, and waste rock, which is lower-grade surrounding rock [37]. Because most of the site has been restored, a total of 75 samples were collected from the surface layer and below 30 cm, and the difference between the restored area and bare area was compared and evaluated. Among them, 42 samples comprised surface soil, calcines, and waste rock, and 33 samples comprised deep soil, calcines, and waste rock. During the same period, the soil samples from the paddy fields downstream of the YMM were collected. For each site, a final sample composed of 3-5 subsamples was collected using the diagonal sampling method (15 paddy soil samples; Figure 1b). The collected soil, calcine, and waste rock samples were kept in clean polyethylene bags, air-dried, ground, and passed through a 200-mesh sieve, followed by total Hg (THg) and THg isotopic analysis.

Sample Collection
Water and atmospheric samples were collected and monitored twice, in December 2021 and August 2022, due to the high seasonal variability of the various indicators in the river water and atmosphere. The interannual variability in the soil Hg is not significant. Thus, soil and mine wastes samples were collected only once, in December 2021. There are two main types of mine wastes: calcines, the residues of Hg ore after high-temperature calcination, and waste rock, which is lower-grade surrounding rock [37]. Because most of the site has been restored, a total of 75 samples were collected from the surface layer and below 30 cm, and the difference between the restored area and bare area was compared and evaluated. Among them, 42 samples comprised surface soil, calcines, and waste rock, and 33 samples comprised deep soil, calcines, and waste rock. During the same period, the soil samples from the paddy fields downstream of the YMM were collected. For each site, a final sample composed of 3-5 subsamples was collected using the diagonal sampling method (15 paddy soil samples; Figure 1b). The collected soil, calcine, and waste rock samples were kept in clean polyethylene bags, air-dried, ground, and passed through a 200-mesh sieve, followed by total Hg (THg) and THg isotopic analysis.
The YMM downstream rainwater and surface water of the Wengman River were sampled for unfiltered THg, filtered dissolved Hg (DHg), DHg isotopes, particulate Hg (PHg) isotopes, anions, and cations. The THg and DHg isotopes samples were acidified with ultrapure hydrochloric acid, the cation samples with distilled nitric acid, and the anion samples without acid. The water samples were sealed in double-layer polyethylene bags, sent back to the laboratory, protected from light, and stored in a refrigerator at 4 • C. The analytical tests were completed within 28 days.
The total gaseous Hg (TGM) concentrations at the YMM and downstream paddy field sites were monitored 48 times using a portable RA-915+ Zeeman Hg Analyzer (Lumex, Saint Petersburg, Russia). The Lumex instrument's detection limit was 0.5 ng/m 3 . The instrument instantaneously displays the TGM concentrations per second, and each sampling point dataset represents an average monitoring time of at least 5 min in the field [37].

Analytical Methods
Approximately 0.1 g of the mine wastes and soil samples (dry weight) were digested with a mixture of HNO 3 and HCl (v:v = 1:3) for 2 h in a water bath at 95 • C. BrCl was added to the samples, and they were stored for 24 h for the conversion of all forms of Hg to Hg 2+ , followed by the addition of acidic SnCl 2 to the solution to reduce the Hg ions to Hg 0 . They were analyzed using cold-vapor atomic absorption spectrometry (CVAAS, F732-S, Shanghai Huaguang Instrument Factory, Shanghai, China). The detection limit of this method was 0.1 µg/L.
To determine the concentrations of THg and DHg in a water sample, BrCl was added to the sample and allowed to oxidize for 24 h. The Hg ions in the solution were then reduced to Hg 0 using acidic SnCl 2 . The samples were preconcentrated into gold tubes and were later tested using a cold-vapor atomic fluorescence spectrophotometer (CVAFS, Tekran 2500, Tekran, Toronto, Ontario, Canada). The detection limit of this method was 0.1 µg/L. The THg in the water passing through a 0.45 µm filter is defined as DHg; subtracting the DHg from the THg yields the concentration of PHg in the water [34]. The anions and cations were analyzed by automated Dionex ICS-90 ion chromatography (Dionex, Sunnyvale, CA, USA) and an inductively coupled plasma optical emission spectrometer (ICP-OES, Varian, Palo Alto, CA, USA), respectively [38].
The Hg isotopic composition was analyzed using Neptune Plus MC-ICP-MS (Thermo Fisher Scientific, Waltham, MA, USA)at the State Key Laboratory of Environmental Geochemistry, the Institute of Geochemistry, the Chinese Academy of Sciences, following the method described by Yin et al. [39]. The total soluble Hg (TSHg) of the Hg mine wastes was extracted using a leaching experiment, as the Hg isotopes were to be tested along with the digested soil sample [37]. To ensure the minimum Hg concentration required for the DHg isotopes analysis of aqueous samples, each filtered water sample was pre-enriched into 5 mL of 40.0% aqua regia absorbent solution (v:v, HNO 3 : HCl = 2:1), as shown in the method established by Li et al. [40]. For the Hg isotopes of PHg in the water samples, 2-5 L of water was filtered through a high-temperature purified Teflon membrane and freeze-dried. The Hg in the membrane was extracted into 5 mL of 40.0% anti-aqua regia absorbent solution using a tubular muffle furnace [41].

Hg Isotopes Analysis
The Hg isotopic composition was calculated using the formula presented by Blum and Bergquist (2007). Mass-dependent fractionation is expressed as delta (δ), and the results were calculated as follows: where xxx is 199, 200, 201, 202, or 204. Mass-independent fractionation is expressed as "∆", and it was calculated using the following equations: In this study, the binary mixed model was used to calculate the two sources of DHg of River Water No. 1. The calculations were performed using Equations (5) and (6) [1,42,43]: δ 202 Hg 3 = δ 202 Hg 1 × F 1 + δ 202 Hg 2 × F 2 (5) 1= F 1 + F 2 (6) where F represents the percentage of the pollution source, subscript 1 represents the TSHg, subscript 2 represents the mountain spring water DHg, and subscript 3 represents the River Water No. 1 DHg. When calculating the two sources of River Water No. 1 PHg using a binary mixing model, subscript 1 represents the Hg mine wastes, subscript 2 represents the mountain spring water PHg, and subscript 3 represents the River Water No. 1 PHg. The fractions of Hg in the paddy soil were derived from rainwater sources, river water sources, and geological background sources, and they were calculated using a triplemember mixing model as follows: δ 202 Hg soil = δ 202 Hg rain × F rain + δ 202 Hg river × F river + δ 202 Hg nat × F nat (7) ∆ 199 Hg soil = ∆ 199 Hg rain × F rain + ∆ 199 Hg river × F river + ∆ 199 Hg nat × F nat (8) 1 = F rain + F river + F nat (9) where the subscripts rain, river, and nat represent rainwater sources, river water sources, and geological background sources, respectively, and F rain , F river , and F nat represent the percentages of rainwater sources, river water sources, and geological background sources, respectively.

Data Analysis
Statistical analysis of the data, including means, standard deviations, and t-tests, was performed using IBM SPSS Statistics 26.0 (IBM, Armonk, NY, USA) and Microsoft Excel 2019 software (Microsoft, Redmond, WA, USA) (statistical significance = p < 0.05). Origin 2021 (OriginLab, Northampton, MA, USA) was used for the graphical demonstration of the data, and Arcmap 10.7 (ESRI, RedLands, CA, USA) was used to plot the spatial distributions by inverse distance weighting.

Hg Pollution in Mine Wastes
Considerable variation was observed in the THg concentrations of the YMM mine wastes. Except for the highest Hg concentration of 1.98 × 10 4 mg/kg found in surface mine wastes, the THg concentrations in the remaining samples showed a geometric mean of 38.4 mg/kg, with a range of 1.60-358 mg/kg. For the deep mine wastes, the THg concentrations showed a geometric mean of 46.8 mg/kg, with a range of 14.5-1.07 × 10 3 mg/kg. The considerable variations in the calcine THg concentrations may be attributed to the different retort furnaces used at the YMM. As the early smelting methods were not so advanced, inadequate ore burning resulted in low Hg recovery and a high concentration of calcines. The advancement in smelting technology led to adequate ore roasting, increased Hg recovery (≥95.0%), and a lower Hg concentration in the calcines [47]. The THg concentrations in the deep mine wastes analyzed in this study were higher than that of the local bedrock (0.35 mg/kg) [32], which was similar to the previous study also conducted at the Wanshan Hg Mine (geometric mean of THg concentrations: 49.0 mg/kg; THg concentration range: 4.15-825 mg/kg) [33]. The THg concentrations in 38.5% (15/39) of the YMM mine wastes samples exceeded the second-type soil pollution risk screening value (38.0 mg/kg) [48], and for 17.9% (7/39) of the samples, the THg concentrations exceeded the soil pollution risk control value of the second category of construction land (82.0 mg/kg) [48]. This demonstrates that a significant proportion of Hg persists even after the high-temperature melting of Hg ore [29].
The THg concentrations in the soils covered by restoration showed a geometric mean of 7.70 mg/kg, with a range of 1.68-139 mg/kg. The THg concentrations in the surface samples and deep samples correlated significantly (p < 0.05), indicating that the surface soil has been polluted by calcines in the lower layer. The soil THg concentrations were much higher than the agricultural land soil pollution risk control value (4.00 mg/kg, 6.5 < pH ≤ 7.5) [49], which is 70 times higher than the Guizhou Province soil background value of 0.110 mg/kg [50]. A comparison of the calcine area before and after the YMM restoration is shown in Figure 2a,b. The distribution of the Hg pollution in the surface and deep layers reveals that the most serious Hg pollution occurs in the exposed calcines areas. The mine wastes in the YMM are still the primary source of Hg pollution in the surrounding ecosystem. The exposed calcines seriously impact the local ecological environment by continually releasing Hg into the atmosphere, entering surface water bodies, and leaching into downstream farming soil [8,33]. 2023, 11, x FOR PEER REVIEW 6 of 16 the data, and Arcmap 10.7 (ESRI, RedLands, Cal, USA) was used to plot the spatial distributions by inverse distance weighting.

Hg Pollution in Mine Wastes
Considerable variation was observed in the THg concentrations of the YMM mine wastes. Except for the highest Hg concentration of 1.98 × 10 4 mg/kg found in surface mine wastes, the THg concentrations in the remaining samples showed a geometric mean of 38.4 mg/kg, with a range of 1.60-358 mg/kg. For the deep mine wastes, the THg concentrations showed a geometric mean of 46.8 mg/kg, with a range of 14.5-1.07 × 10 3 mg/kg. The considerable variations in the calcine THg concentrations may be attributed to the different retort furnaces used at the YMM. As the early smelting methods were not so advanced, inadequate ore burning resulted in low Hg recovery and a high concentration of calcines. The advancement in smelting technology led to adequate ore roasting, increased Hg recovery (≥95.0%), and a lower Hg concentration in the calcines [47]. The THg concentrations in the deep mine wastes analyzed in this study were higher than that of the local bedrock (0.35 mg/kg) [32], which was similar to the previous study also conducted at the Wanshan Hg Mine (geometric mean of THg concentrations: 49.0 mg/kg; THg concentration range: 4.15-825 mg/kg) [33]. The THg concentrations in 38.5% (15/39) of the YMM mine wastes samples exceeded the second-type soil pollution risk screening value (38.0 mg/kg) [48], and for 17.9% (7/39) of the samples, the THg concentrations exceeded the soil pollution risk control value of the second category of construction land (82.0 mg/kg) [48]. This demonstrates that a significant proportion of Hg persists even after the high-temperature melting of Hg ore [29].
The THg concentrations in the soils covered by restoration showed a geometric mean of 7.70 mg/kg, with a range of 1.68-139 mg/kg. The THg concentrations in the surface samples and deep samples correlated significantly (p < 0.05), indicating that the surface soil has been polluted by calcines in the lower layer. The soil THg concentrations were much higher than the agricultural land soil pollution risk control value (4.00 mg/kg, 6.5 < pH ≤ 7.5) [49], which is 70 times higher than the Guizhou Province soil background value of 0.110 mg/kg [50]. A comparison of the calcine area before and after the YMM restoration is shown in Figure 2a,b. The distribution of the Hg pollution in the surface and deep layers reveals that the most serious Hg pollution occurs in the exposed calcines areas. The mine wastes in the YMM are still the primary source of Hg pollution in the surrounding ecosystem. The exposed calcines seriously impact the local ecological environment by continually releasing Hg into the atmosphere, entering surface water bodies, and leaching into downstream farming soil [8,33].

Atmospheric Hg
The spatial distribution of the TGM at the YMM showed significant variations (Figure 3a,b). In wintertime, the TGM concentrations averaged 24.1 ± 6.90 ng/m 3 , with a range The TGM concentrations at the exposed calcines areas were found to be the highest both in winter and summer, and the TGM concentrations in summer were much higher than those in winter (winter: 45.0 ng/m 3 ; summer: 700 ng/m 3 ), while the TGM concentrations at the restored sampling point were much lower (winter: 23.6 ng/m 3 ; summer: 153 ng/m 3 ). The results show that mine wastes are still an important emission source of atmospheric Hg pollution, and remediated measures could effectively reduce the Hg emission flux at the interface between the mine wastes and air [51]. Compared with other Hg mining areas in China, the TGM concentrations in the YMM (43.3-700 ng/m 3 ) were much higher than those in the Xunyang (7.40-410 ng/m 3 ) and Wanshan (13.5-309 ng/m 3 ) Hg mining areas [33,52]. However, the concentrations were much lower than those in the Xiushan Hg mining area (29.0-4.21 × 10 4 ng/m 3 ) [53]. The mean concentration of TGM was three times higher when compared with the air quality reference standard of 50.0 ng/m 3 set by the Ministry of Environmental Protection of China, and it may pose a potential risk to local residents [54]. Therefore, the Hg emissions from mine wastes should be strictly controlled to reduce the environmental risks.

Atmospheric Hg
The spatial distribution of the TGM at the YMM showed significant variations ( Figure  3a,b). In wintertime, the TGM concentrations averaged 24.1 ± 6.90 ng/m 3 , with a range of 10.1-45.0 ng/m 3 . In summertime, the TGM concentrations averaged 153 ± 129 ng/m 3 , with a range of 43.3-700 ng/m 3 . The TGM concentrations at the exposed calcines areas were found to be the highest both in winter and summer, and the TGM concentrations in summer were much higher than those in winter (winter: 45.0 ng/m 3 ; summer: 700 ng/m 3 ), while the TGM concentrations at the restored sampling point were much lower (winter: 23.6 ng/m 3 ; summer: 153 ng/m 3 ). The results show that mine wastes are still an important emission source of atmospheric Hg pollution, and remediated measures could effectively reduce the Hg emission flux at the interface between the mine wastes and air [51]. Compared with other Hg mining areas in China, the TGM concentrations in the YMM (43.3-700 ng/m 3 ) were much higher than those in the Xunyang (7.40-410 ng/m 3 ) and Wanshan (13.5-309 ng/m 3 ) Hg mining areas [33,52]. However, the concentrations were much lower than those in the Xiushan Hg mining area (29.0-4.21 × 10 4 ng/m 3 ) [53]. The mean concentration of TGM was three times higher when compared with the air quality reference standard of 50.0 ng/m 3 set by the Ministry of Environmental Protection of China, and it may pose a potential risk to local residents [54]. Therefore, the Hg emissions from mine wastes should be strictly controlled to reduce the environmental risks. Previous studies have reported that the TGM released from pollution sources can settle into paddy soil after migration [55,56]. After the TGM monitoring in the downstream paddy fields, the following trends were noticed (Figure 3c). The results indicate that the TGM concentrations gradually decreased within 5.5 km of the YMM both in winter and summer (winter: 5.79 ng/m 3 ; summer: 21.4 ng/m 3 ). However, the TGM Previous studies have reported that the TGM released from pollution sources can settle into paddy soil after migration [55,56]. After the TGM monitoring in the downstream paddy fields, the following trends were noticed (Figure 3c). The results indicate that the TGM concentrations gradually decreased within 5.5 km of the YMM both in winter and summer (winter: 5.79 ng/m 3 ; summer: 21.4 ng/m 3 ). However, the TGM concentrations were still higher than the global background (1.50-1.60 ng/m 3 ) [33,57]. The TGM concentrations considerably increased at around 5.5 km, with the highest value of 68.7 ng/m 3 in summer, and then gradually decreased with the increasing distance. The increase in the TGM in  [37,60]. The highest Hg concentration was found at the W1 site near the YMM (Figure 4a,b). The THg concentration in the winter surface water at this site exceeded the standard limit of 1000 ng/L stipulated by China's Class V surface water environmental quality standard [61], indicating the direct influence of the Hg mine wastes. In order to reduce the impact of the upstream calcine leachate on the downstream water system, Xu et al. [36] designed and built a weir 1.5 km away from the YMM, which can intercept 40.4% of the THg per year and significantly reduce the THg concentrations in the river. The upstream calcine leachate is mainly composed of PHg. In this study, the proportions of the water PHg to THg in the winter and summer flowing through the weir decreased by 76.5% and 52.9%, respectively, indicating that the removal ratios were higher than those presented by Xu et al. [36]. The difference between the two studies might be attributed to the flow of the river water, as previous studies have reported that water flow is the main factor that influences the transport and migration of Hg [53,62]. In this study, the average THg concentrations in the summer samples (34.4 ng/L) were lower than those of the winter samples (101 ng/L). This might be because the summer samples were collected after a heavy rain event and the erosive, and leaching effect of the rainwater was not significant. The river flows in summer were higher than those in winter, which mainly showed that the dilution effect resulted in lower Hg concentrations in the summer samples. However, the highest concentration still exceeded the threshold limit of 100 ng/L set by China's Class III surface water environmental quality standard [61].
The above studies have proven that the weir can indeed cause the particulate matter to settle, which is because the water flow slows down and the suspension time increases, thereby reducing the Hg pollution downstream. However, Xu et al. [36] only monitored before and after the weir, and they did not set sampling points downstream of the Wengman River. The results of this study show that at 4-5 km downstream of the Wengman River, the THg concentrations decreased to 10.3-11.3 ng/L, as a large amount of PHg had settled. The proportion of DHg to THg increased to 55.4%-98.9%, indicating the PHg sedimentation effect. The Hg concentrations at 4-5 km were close to the mean concentration of 7.09 ng/L in the tributaries, which indicated the baseline concentration of the surface water in this area. Except for W14, the water Hg concentrations in the tributaries in this study were similar to those of Qiu et al. [35] (tributary: 3.00-17.0 ng/L). After 5.5 km, the Hg concentration in the river water gradually increased, but it did not exceed the limit of 100 ng/L stipulated by China's Class III surface water environmental quality standard [61]. The Hg concentration at 6.5 km of the tributary (W14 sample) in wintertime reached 85.0 ng/L, exceeding the 50.0 ng/L limited stipulated by China's Class II surface water environmental quality standard [61]. This indicated that the surface water after 5.5 km may be impacted by other external sources of Hg.  This study analyzed the anions and cations of the Wengman River, and it found that the Na + and Cl − concentrations increased significantly after 5.5 km in both winter and summer. The increase in the Na + concentration was much higher than that of Cl − (Figure 4c). Figure 4d shows that the ratio of Cl − :Na + in the sample at 1 km was 1:1, indicating that it was mainly derived from the dissolution of evaporite rock. However, the ratio of the water Cl − :Na + downstream gradually fell below the 1:1 ratio line. The ratio of Cl − and Na + ranged from 0.232 to 0.876, which indicates the contribution of sources other than the dissolution of evaporite rocks [63,64]. Because Liulongshan Township is 5.5 km away, with a dense population, domestic activities have significant impacts on the water chemistry [65]. Cl − is not affected by physical, chemical, or biological processes, and it is a good indicator of anthropogenic activities, such as the use of agricultural fertilizers, animal manure, and domestic sewage [66]. The Wanshan area belongs to the karst landform, and the main rock type is carbonate rock. The Na + concentration in the surface water is relatively low. The increase in Na + relative to Cl − may be due to silicate weathering (e.g., plagioclase) and the effect of the input from domestic pollution sources. In order to further elucidate the contribution of both to the excess Na + , this study used the molar ratio bivariate plots of the Na + -normalized Ca 2+ and Mg 2+ and Na + -normalized Ca 2+ and HCO3 − distributions ( Figure  S1) to identify the contribution of rock weathering to the ion source in the river [67,68]. As shown in Figure S1, the ionic composition of the river water in the study area was mainly located near the weathered end element of the carbonate rock, with a trend towards the weathered end element of the silicate rock. This indicates that the contribution of silicate rocks to river water ions is small. Therefore, this study concluded that the increase in Na + concentration relative to the Cl − concentration after 5.5 km from the Wengman River may be caused by domestic pollution [69,70]. This domestic pollution may come from domestic This study analyzed the anions and cations of the Wengman River, and it found that the Na + and Cl − concentrations increased significantly after 5.5 km in both winter and summer. The increase in the Na + concentration was much higher than that of Cl − (Figure 4c). Figure 4d shows that the ratio of Cl − :Na + in the sample at 1 km was 1:1, indicating that it was mainly derived from the dissolution of evaporite rock. However, the ratio of the water Cl − :Na + downstream gradually fell below the 1:1 ratio line. The ratio of Cl − and Na + ranged from 0.232 to 0.876, which indicates the contribution of sources other than the dissolution of evaporite rocks [63,64]. Because Liulongshan Township is 5.5 km away, with a dense population, domestic activities have significant impacts on the water chemistry [65]. Cl − is not affected by physical, chemical, or biological processes, and it is a good indicator of anthropogenic activities, such as the use of agricultural fertilizers, animal manure, and domestic sewage [66]. The Wanshan area belongs to the karst landform, and the main rock type is carbonate rock. The Na + concentration in the surface water is relatively low. The increase in Na + relative to Cl − may be due to silicate weathering (e.g., plagioclase) and the effect of the input from domestic pollution sources. In order to further elucidate the contribution of both to the excess Na + , this study used the molar ratio bivariate plots of the Na + -normalized Ca 2+ and Mg 2+ and Na + -normalized Ca 2+ and HCO 3 − distributions ( Figure S1) to identify the contribution of rock weathering to the ion source in the river [67,68]. As shown in Figure S1, the ionic composition of the river water in the study area was mainly located near the weathered end element of the carbonate rock, with a trend towards the weathered end element of the silicate rock. This indicates that the contribution of silicate rocks to river water ions is small. Therefore, this study concluded that the increase in Na + concentration relative to the Cl − concentration after 5.5 km from the Wengman River may be caused by domestic pollution [69,70]. This domestic pollution may come from domestic sewage, domestic waste (such as batteries, thermometers, pigment and paint residues, fluorescent lamps, and so on), etc. [59,71].

Source Apportionment by Hg Isotopes
The river water THg increased significantly when it flowed through the Hg mining area (Figure 4a,b), which is consistent with previous literature [9,72]. In winter, the δ 202 Hg and ∆ 199 Hg values in the river water samples downstream averaged −0.29‰ ± 0.30‰ (−0.71~0.11‰, n = 7) and −0.02‰ ± 0.07‰ (−0.12~0.06‰, n = 7), respectively (Table S1). Mercury isotopes were used to trace the source of Hg in the upstream water of the Wengman River. The results showed that the main contributing sources included Hg mine wastes and mountain spring water. This study assumes that the Hg from mine wastes and mountain spring water enters the River Water No. 1 sampling site in a rapid mixing process and that no significant MDF and MIF will occur during this process. The pollution sources of the DHg in the River Water No. 1 sample mainly include the TSHg from the Hg mine wastes and the DHg from the mountain spring water. For the TSHg of the Hg mine wastes, the DHg of the mountain spring water, and the DHg of River Water No. 1, the δ 202 Hg values were −0.90‰, −1.57‰, and −1.25‰, respectively, and the ∆ 199 Hg values were all close to zero. Many previous studies have demonstrated the usefulness of end-member mixing models for Hg-source tracking in water environments [73,74]. In this study, the relative contributions of the different sources for the River Water No. 1 DHg were calculated using a binary mixing model. The relative contribution ratio of the mine waste TSHg to DHg was 48.6%, and that of the mountain spring water was 51.4%.
The pollution sources of the PHg in River Water No. 1 mainly included the Hg mine wastes and mountain spring water. The δ 202 Hg values in the Hg mine wastes, the PHg of mountain spring water, and the River Water No. 1 PHg were −0.35‰, −1.78‰, and −0.48‰, respectively, and the ∆ 199 Hg values were all close to zero. The observed ∆ 199 Hg values in Hg mine wastes are consistent with previous studies [29,75,76]. In this study, the relative contributions of the two pollution sources to the River Water No. 1 PHg were calculated using a binary mixing model. The relative contribution ratio of the Hg mine wastes was 90.5%, and that of the mountain spring water was 9.50%.
The THg in the River Water No. 1 was 1.25 × 10 3 ng/L, while DHg accounted for 3.00%, and PHg for 97.0%. It was calculated that the Hg mine wastes directly contributed 89.3% to the river Hg pollution at the No. 1 site, indicating that the erosion of Hg mine wastes by runoff is the main process of the Hg pollution in the rivers near the Hg mine. This shows that the upstream water of the Wengman River is seriously polluted with Hg. The government needs to remediate the mine wastes left at the site and reinforce the tailings dam, which could reduce the Hg pollution in the downstream river and the health risk to local residents.

Paddy Soil Hg Pollution
The THg concentrations in the paddy soil downstream of the YMM averaged 3.58 ± 1.82 mg/kg with a range of 1.49-8.51 mg/kg. Compared with other Hg mining areas, the paddy soil THg concentrations at the YMM were lower than those in China's Xunyang (1.30-750 mg/kg), Xiushan (0.45-68.0 mg/kg), and Wanshan (0.50-188 mg/kg) Hg mining areas [52,53,77]. However, they were still much higher than the agricultural soil pollution risk screening value (0.60 mg/kg, 6.5 < pH ≤ 7.5) and Guizhou soil Hg background value (0.110 mg/kg) [49,50]. The results showed that the downstream paddy soils are still seriously polluted by THg and indicated serious ecological risks. It is critical to identify the sources and contributions of Hg pollution in paddy soil. Therefore, preventive measures should be taken to control the Hg release. The obtained results provide a reliable theoretical and scientific basis for the treatment and safe utilization of Hg-contaminated soil.
The trends of the soil THg in the downstream paddy fields in this study were not consistent with those presented by Xu et al. [53], who reported that, with the increase in the distance from the Hg mining area, the soil THg concentrations tended to decrease. However, in this study, a different trend was noticed. A decreasing trend was observed prior to the 5.5 km distance; however, after 5.5 km, the THg concentrations at sites S4 and S5 significantly increased. The variation in the THg concentration in the downstream paddy soil is consistent with that of the TGM ( Figure S2). Apart from this, a significant correlation was found between the TGM and paddy soil THg (p < 0.05). This indicates that atmospheric dry and wet depositions play a vital role in the Hg pollution of paddy soil [35]. The paddy fields are located along the banks of the Wengman River, and the local people have been using the Hg-contaminated river water for irrigation for a long time, which could also be a key source of the Hg pollution in the paddy soil [18,78].

Source Apportionment by Hg Isotopes
Previous studies have shown that soils can preserve the isotopic fingerprints of Hg pollution sources [42,79,80]. The paddy field downstream of the YMM are located on both sides of the Wengman River. There was a significant correlation between the TGM and THg in the paddy soil (p < 0.05), indicating that dry and wet atmospheric depositions are an important source of Hg pollution in the paddy soil. The contribution of wet deposition is of key importance as shown in Figure 5a,b. Pribil et al. [47] stated that the Hg in the soils to the north and east of the mining area may be the result of atmospheric deposition, geological background influence, and gaseous Hg emissions from calcines piles during Hg processing. The irrigation of paddy fields with Hg-contaminated river water is also one of the key sources [18]. water to the paddy soil increased to 62.0% at 6 km, which is consistent with the increase in the Hg concentrations in the river water, indicating the influence of domestic pollution sources. The contribution ranges were 30.0~62.0% by river water, 4.00~17.0% by rainwater, and 34.0~53.0% by geological background. With the increase in distance, the contribution ratio of the river water gradually decreased, and the contribution ratio of the rainwater and geological background sources increased again, indicating that the paddy soil pollution in this range was mainly attributed to domestic pollution sources.

Conclusions
The YMM is still the primary source of Hg pollution in the surrounding ecosystem, and especially of the significant atmospheric Hg emissions in the exposed calcines area. Peak concentrations of THg were observed at the river upstream, and the source apportionment for the DHg and PHg in the river water using a binary mixing model demonstrated that mine wastes were the main source of Hg in the surface water. The TGM concentrations downstream showed specific spatial distributional characteristics, indicating a large amount of Hg emissions from Hg mine wastes and unidentified domestic pollution sources. This study calculated the contributions of the river water, atmospheric wet dep- The Rain Water No. 1 sample was collected 2 km away from the mining area. Due to its proximity to the Hg mining area, the Hg mine wastes have a greater impact. As shown in Figure 5a, the Hg isotopes compositions were δ 202 Hg = −0.51‰ ± 0.05‰ and ∆ 199 Hg = −0.10‰ ± 0.04‰. The Rain Water No. 2 sample was collected at a distance of 6 km from the mining area, and we noticed that the Hg mine wastes had less of an impact. The ∆ 199 Hg in the rainwater was positive, which was similar to the Hg isotopic value of the rainwater in Guiyang (δ 202 Hg: −0.44~−4.27‰, ∆ 199 Hg: 0.19~1.16‰) [81]. The δ 202 Hg (δ 202 Hg = −0.34‰ ± 0.05‰) and ∆ 199 Hg (∆ 199 Hg = 0.30‰ ± 0.04‰) values in the rainwaters are presented in Figure 5b. The mean values of the δ 202 Hg and ∆ 199 Hg in the paddy soil were −0.73‰ ± 0.11‰ (−0.91~−0.56‰, n = 12) and 0.03‰ ± 0.05‰ (−0.05~0.10‰, n = 12), respectively. According to Song et al. [6], the mean values of the δ 202 Hg and ∆ 199 Hg in the paddy soils in the Wanshan area were −1.26‰ ± 0.06‰ (−1.30~−1.21‰, n = 2) and −0.07‰ ± 0.10‰ (−0.14~0.00‰, n = 2), which were used as the background values in this study (Table S2).
As shown in Figure 5a,b, the combined characteristics of the ∆ 199 Hg and δ 202 Hg indicate that the paddy soil is a ternary mixture of different sources, such as rainwater, river water, and geological background sources. Therefore, the δ 202 Hg and ∆ 199 Hg of the corresponding point samples were used to trace the source of the Hg pollution in the paddy soil. This study calculated the relative contributions of the three sources using a ternary mixed model, as shown in Figure 5c. The results showed that the exogenous input of the Hg pollution in the paddy soil can be divided into two parts. The first part is the area contaminated by Hg mining activities. Within 5.5 km of the YMM, the river water is mainly polluted by Hg mine wastes, while the paddy soil pollution at 2 km is mostly attributed to river water, reaching 86.0%. The contribution range of river water was 28.0~86.0%, while in the case of rainwater, the range was 7.00~25.0%. The contribution range of the geological background was 7.00~60.0%. With the increase in distance, the Hg contribution of the river water gradually decreased, while the contribution ratio of the rainwater and geological background sources gradually increased. The paddy fields in this range were mainly contaminated by the Hg mine. The second part is the domestic Hg-polluted area. After 5.5 km from the YMM, the contribution of the river water to the paddy soil increased to 62.0% at 6 km, which is consistent with the increase in the Hg concentrations in the river water, indicating the influence of domestic pollution sources. The contribution ranges were 30.0~62.0% by river water, 4.00~17.0% by rainwater, and 34.0~53.0% by geological background. With the increase in distance, the contribution ratio of the river water gradually decreased, and the contribution ratio of the rainwater and geological background sources increased again, indicating that the paddy soil pollution in this range was mainly attributed to domestic pollution sources.

Conclusions
The YMM is still the primary source of Hg pollution in the surrounding ecosystem, and especially of the significant atmospheric Hg emissions in the exposed calcines area. Peak concentrations of THg were observed at the river upstream, and the source apportionment for the DHg and PHg in the river water using a binary mixing model demonstrated that mine wastes were the main source of Hg in the surface water. The TGM concentrations downstream showed specific spatial distributional characteristics, indicating a large amount of Hg emissions from Hg mine wastes and unidentified domestic pollution sources. This study calculated the contributions of the river water, atmospheric wet deposition, and geological background to the paddy soil Hg pollution by Hg isotopes, which were also verified by the spatial distributions of the river water Hg, river water anions and cations, TGM, and paddy soil Hg. The study shows that the paddy soil upstream at 5.5 km is mainly polluted by the Hg mine, while domestic sources are the main contributors after a distance of 5.5 km. This study provides an important scientific basis for the source control of Hg in the surface water and paddy fields in Hg mining areas. This is needed to control the Hg emissions from mine wastes to the river water and atmosphere, which would finally reduce the Hg bioaccumulation in agricultural crops and the associated human health risks in Hg-polluted areas.

Conflicts of Interest:
The authors declare no conflict of interest.