Health and ecological risk of heavy metals in agricultural soils related to Tungsten mining in Southern Jiangxi Province, China

Study area

Dayu County is located in the southwest of Jiangxi Province, China (E114°–144.44°, N25.15°–25.37°) (Fig. 1). This region contains the highest concentration of tungsten deposits globally, with considerable tungsten-tin mineral occurrences (Zhang et al., 2021). The climate is a typical subtropical humid monsoon with distinct four seasons and abundant rainfall. The annual average temperature is between 19.1 °C and 20.8 °C, and the annual average precipitation is 1,318.9 mm. The topography rises in the northern, western, and southern parts of the county, whereas the central and eastern sections form a hilly basin surrounded by mountains on three sides that opens eastward. The Zhangjiang River crosses the county from west to east. The tungsten mines are distributed across the northern mountainous terrain, and farmland occupies the eastern plain. Long-term mineral development has become an important source of heavy metal contamination in this region.

Soil sampling and chemical analysis

Soil samples were taken at a depth of 0–20 cm at six selected areas in four towns (Fujiang, Huanglong, Qinglong, and Chijiang) along the Zhangjiang River (Fig. 1) on sloping land, sloping plain transition land, and plain land. These sits were chosen considering their agree influenced by the tungsten mines: Near Qinglong has a large-scale tungsten mine called Piaotang Tungsten Mine, Fujiang has an important historical tungsten mining area in Xihua Mountain, Huanglong locates no tungsten mine and Chijiang is an important rice planting area far from tungsten mining areas. The soil types in the study area are classified as red soil, yellow soil, and alluvial soil. The six sampling sites were termed FJ-N, FJ-S, HL, QL, CJ-W, and CJ-E; FJ-N and FJ-S are sloping land with artificial terraced fields located along the Fujiang River, a main tributary of the Zhangjiang River, located 4 and 2 km, respectively, from historical tungsten mining areas (Xihuashan, Fig. 1), respectively, HL is a control site far away from mining areas, QL is close to active northern mining areas, and CJ-W and CJ-E are plain sites with traditional paddy rice cultivation. To obtain representative samples, at each site, four to six subsamples were obtained within a 10 × 10-m grid at the center and at the four corners of an ideal square and combined to obtain one composite sample (weight: 1 kg) per site, resulting in a total of 72 soil samples and 26 corresponding rice grain samples. We removed stones, plant debris, and other residues from the soil samples. Due to the influence of terrain, the distance between sampling plots was 50–200 m. Soil samples were obtained using wooden shovels and stored in polyethylene bags. In the laboratory, all samples were air-dried at room temperature, crushed, homogenized, and sieved through a 100-mesh nylon mesh prior to further analysis. Rice grains were placed in a clean, sunny, and ventilated room on a clean wooden board and frequently turned to prevent mold growth. After drying, grain samples were weighed, sieved through a 100-mesh nylon mesh, and stored in polyethylene bags before digestion.

The concentrations of metals (As, Cd, Cu, Cr, Pb, Mo, W, and Zn) in soils were analyzed using an inductively coupled plasma-mass spectrometer (ICP-MS, iCAP Q; Thermo Fisher Scientific, Waltham, MA, USA). Prior to analysis, the soil and rice samples were digested by a H₂SO₄-HNO₃-HF mixture using a Microwave Digestion System (Multiwave PRO, Anton Paar, Graz, Austria) (Li et al., 2006). A blank control, duplicate samples, and standard reference materials (standard sample of lateritic, GSS-5) were employed to ensure data quality. The concentrations of heavy metals were computed as the average values of duplicate samples. The precision and accuracy of the method is expressed in standard deviation and relative error from GSS-5 (n = 12), respectively. The standard deviation (SD%) of As, Cd, Cr, Cu, Mo, Pb, W, Zn is 2.47, 0.02, 1.77, 2.07, 0.16, 2.37, 0.25, 2.60, resepectively. The relative error (RE%) of As, Cd, Cr, Cu, Mo, Pb, W, Zn is 1.12, 2.56, 0.99, −0.26, 2.79, 0.11, −0.55, 1.15, respectively.

In the soil, pH, total organic carbon (TOC), cation exchange capacity (CEC), and metal oxides (Fe₂O₃, CaO, Na₂O, and K₂O) were also measured. The soil pH was measured using a pH meter in a 1:2.5 soil: water mixture, and the soil TOC was determined via a Torch TOC combustion analyzer (multi N/C 2100; Analytik Jena, Jena, Germany). The CEC was determined as described elsewhere (Chen & Huang, 2021), and the levels of Fe₂O₃, CaO, Na₂O, and K₂O were measured via X-ray fluorescence spectrometry (XRF, EDX-8000; Shimadzu, Kyoto, Japan).

Assessment of soil heavy metal contamination

(1) Geo-accumulation (I_geo) index

The geo-accumulation (Igeo) index was employed to assess the pollution degree of heavy metals in soil (Muller, 1969), according to Eq. (1):

(1) $${\mathrm{I}}_{\mathrm{g}\mathrm{e}\mathrm{o}}=\mathrm{l}\mathrm{o}\mathrm{g}2\left[{\displaystyle \frac{{\mathrm{C}}_{\mathrm{i}}}{1.5{\mathrm{B}}_{\mathrm{i}}}}\right],$$where Ci is the mean concentration of the element in the examined soil, and Bi is the geochemical background value in the soil of Jiangxi Province (As: 10.4, Cd: 0.10, Cu: 20.8, Cr: 48.0, Pb: 32.1, Mo: 0.30, W: 4.93, Zn: 69.0, mg/kg) (China National Environmental Monitoring Centre (CNEMC), 1990). The soil pollution levels can be partitioned into seven classes based on the values of I_geo: Class 0 (unpolluted; I_geo ≤ 0), Class 1 (unpolluted to moderately polluted; 0 < I_geo ≤ 1), Class 2 (moderately polluted; 1 < I_geo ≤ 2), Class 3 (moderately to heavily polluted; 2 < I_geo ≤ 3), Class 4 (heavily polluted; 3 < I_geo ≤ 4), Class 5 (heavily to extremely polluted; 4 < I_geo ≤ 5), and Class 6 (extremely polluted; I_geo > 5) (Loska et al., 1997).

(2) Contamination factor (CF)

The level of soil contamination is expressed in terms of a contamination factor (CF), calculated following Eq. (2):

(2) $\mathrm{C}\mathrm{F}={\displaystyle \frac{{\mathrm{C}}_{\mathrm{i}}}{{\mathrm{B}}_{\mathrm{i}}}}$where Ci is the mean concentration of the element in the examined soil, and Bi is the background value of metals in the soil of Jiangxi Province (As: 10.4, Cd: 0.10, Cu: 20.8, Cr: 48.0, Pb:32.1, Mo: 0.30, W: 4.93, Zn: 69.0, mg/kg) (China National Environmental Monitoring Centre (CNEMC), 1990). The contamination factor CF < 1 refers to low contamination, 1 ≤ CF < 3 indicates moderate contamination, 3 ≤ CF ≤ 6 indicates considerable contamination, and CF > 6 indicates very high contamination.

(3) Pollution load index (PLIs)

The overall metal load in soils from each site was computed using the pollution load index (PLIs) (Tomlinson et al., 1980) according to Eq. (3):

(3) $\text{PLIs}={\left({\text{CF}}_{\text{1}}\times {\text{CF}}_{\text{2}}\times {\text{CF}}_{\text{3}}\times \dots {\text{CF}}_{\text{n}}\right)}^{\frac{1}{\text{n}}}$

According to Tomlinson, CF_n is the contamination factor of the metal n in the sample. The PLIs provide simple but comparative means for assessing site quality, where a value of PLIs < 1 denotes low contamination, 1 < PLIs < 2 denotes moderate contamination, 2 < PLIs <3 denotes heavy pollution, and PLIs > 3 denotes extremely heavy pollution.

Ecological risk assessment regarding metals in soils

The ecological risk factor ( ${\mathrm{E}}_{\mathrm{r}}^{\mathrm{i}}$) is a measurement of the potential ecological risk (RI) of a given metal and defined as Eq. (4):

(4) ${\text{E}}_{\text{r}}^{\text{i}}={\text{T}}_{\text{r}}^{\text{i}}\left(\frac{{\text{C}}_{\text{r}}^{\text{i}}}{{\text{B}}_{\text{i}}}\right),$where B_i is the background value of metals in soil in Jiangxi Province (As:10.4, Cd:0.10, Cu:20.8, Cr:48.0, Pb:32.1, Mo:0.30, W:4.93, Zn:69.0, mg/kg) (China National Environmental Monitoring Centre (CNEMC), 1990), ${\mathrm{C}}_{\mathrm{r}}^{\mathrm{i}}$ is the heavy metal content in the soil, ${\mathrm{T}}_{\mathrm{r}}^{\mathrm{i}}$ is the toxic-response factor, and the ${\mathrm{T}}_{\mathrm{r}}^{\mathrm{i}}$ values of Cr, Cu, Zn, As, Cd, Pb, and W are 2, 5, 1, 10, 30, 5, and 2, respectively (Zheng et al., 2020). Hakanson (1980) introduced the potential ecological risk index (RI) as a method to quantitatively evaluate the risk of soil heavy metal pollution. The RI represents the integrated potential ecological risk, calculated as the sum of individual risk factors ( ${\mathrm{E}}_{\mathrm{r}}^{\mathrm{i}}$) for each metal contaminant (Hakanson, 1980), and is defined as follows Eq. (5):

(5) $\mathrm{R}\mathrm{I}=\sum {\mathrm{E}}_{\mathrm{r}}^{\mathrm{i}}$

According to Hakanson (1980), ${\mathrm{E}}_{\mathrm{r}}^{\mathrm{i}}$ and RI represent potential ecological risks through different levels (Table 1).

Health risk assessment of heavy metals in paddy rice

As Dayu County is a major rice production area in China, with rice being a staple food for a large part of the population, heavy metal pollution must be considered. The health risk assessment model developed by the United States Environmental Protection Agency (USEPA) was used to evaluate the carcinogenic and non-carcinogenic risks of heavy metals (United States Environmental Protection Agency, 2011), using Eqs. (6)–(10) of the average daily dose (ADD_i), hazard quotient (HQ) and total hazard quotient index (HI), cancer risk (CR) and total cancer risk (TCR):

(6) $$\mathrm{A}\mathrm{D}{\mathrm{D}}_{\mathrm{i}}={\displaystyle \frac{{\mathrm{C}}_{\mathrm{i}}\times \mathrm{I}\mathrm{R}\times \mathrm{E}\mathrm{F}\times \mathrm{E}\mathrm{D}}{\mathrm{B}\mathrm{W}\times \mathrm{A}\mathrm{T}}}$$

(7) $$\mathrm{H}{\mathrm{Q}}_{\mathrm{i}}=\sum {\displaystyle \frac{\mathrm{A}\mathrm{D}{\mathrm{D}}_{\mathrm{i}}}{{\mathrm{R}}_{\mathrm{f}}\mathrm{D}}}$$

(8) $$\mathrm{H}\mathrm{I}=\sum \mathrm{H}{\mathrm{Q}}_{\mathrm{i}}$$

(9) $$\mathrm{C}{\mathrm{R}}_{\mathrm{i}}=\sum \mathrm{A}\mathrm{D}{\mathrm{D}}_{\mathrm{i}}\times \mathrm{S}{\mathrm{F}}_{\mathrm{i}}$$

(10) $$\mathrm{T}\mathrm{C}\mathrm{R}=\sum \mathrm{C}{\mathrm{R}}_{\mathrm{i}},$$where HQ_i and HI are the non-carcinogenic health risks of single and combined metals, respectively, CR and TCR are the carcinogenic health risks of single and combined metals, respectively (Hui et al., 2021), ADD_i refers to the daily intake of heavy metal i in paddy rice (mg/kg/day), C_i is the mean value of metal i in rice (mg/kg), EF is the exposure frequency (365 days/year), ED is the exposure duration (30 years), IR indicates the daily intake amount of rice (kg/day), which was 410.13 g/day according to Ministry of Environmental Protection of China (2013), BW is the body weight (60.6 kg), AT refers to the average exposure time (365 × ED days for non-carcinogenic risk; 365 × 70 days for carcinogenic risk), R_fD is the corresponding reference dose of the metals (As: 0.0003, Cd: 0.001, Cr: 0.003, Cu: 0.04, Mo: 0.005, Pb: 0.004, and Zn: 0.3) (mg/kg/day), SF is the slope factor of the metals (As: 1.5, Cd: 6.1, Cr: 0.5, and Pb: 0.0085) (mg/kg/day).

The translocation factor (TF) was used to assess the bioaccumulation abilities of the heavy metals in rice, as described in Eq. (11):

(11) $$\mathrm{T}\mathrm{F}={\mathrm{C}}_{\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e}}/{\mathrm{C}}_{\mathrm{s}\mathrm{o}\mathrm{i}\mathrm{l}}$$

Statistical analysis

The differences in the heavy metal contents in soils from different sampling sites were determined via one-way analysis of variance (ANOVA). A Shapiro–Wilk test and Leven’s test were performed to ensure the normality and homogeneity of the data prior to ANOVA (p < 0.05 or p < 0.01). Pearson’s correlation analysis was applied to evaluate the influences of soil factors on the heavy metals in the soils. Statistical analyses were conducted in SPSS 26.0 (IBM, Armonk, NY, USA) and Origin 2021b (OriginLab Corporation, Northampton, MA, USA).

Concentrations and spatial distribution of heavy metals in soil

Figure 2 and Table S1 show the concentrations and basic statistics of the heavy metals in the tested topsoil samples. The mean concentrations of As, Cd, Cu, Cr, Pb, Mo, W, and Zn exceeded their respective average background values (ABV) of Jiangxi Province soil (China National Environmental Monitoring Centre (CNEMC), 1990), accounting for 87.5%, 98.6%, 90.7%, 86.1%, 76.4%, 100%, 97.2%, 70.8%, and 20% in the samples, respectively. The levels of Mo and W surpassed the background values by 34.1 and 19.3 times, respectively, and the Cd concentration exceeded the background level by 6.23 times. The mean values of As, Cu, Cr, Pb, and Zn were 1.33–3.21 times higher than the background values. The coefficient of variation (CV) represents the spatial dispersion and degree of variation of heavy metal concentrations, and the larger its value, the greater the impact of external factors on the pollution level (Pan et al., 2016). The CV values for Cd and Mo indicated strong spatial variation (CV > 100%), likely due to anthropogenic activities (Zhang, 2020). The levels of Cu, Cr and Zn showed a moderate spatial variation (21% < CV < 50%), and for As, Pb, and W, we observed a high variability(50% < CV < 100%). The risk screening values for agricultural soil contamination are 30, 0.3, 50, 150, 250, and 200 mg kg⁻¹ for As, Cd, Cu, Cr, Pb, and Zn, respectively (Ministry of Ecology and Environment of China (MEEC), 2018). For Mo and W, the screening values of 3 mg kg⁻¹ refer to the Dutch standard (Dutch Ministry of Housing, Spatial Planning and the Environment (VROM), 2000). In the present study, the mean As and Cd concentrations were 1.11 and 2.07 times, respectively, their corresponding MEEC limits. However, compared to the Dutch standard, the As and Cu levels indicated mild contamination, whereas the levels for Mo and W suggested strong pollution. In this study, As, Cd, Mo, and W were the major pollutants, similar to the findings for residential aeras near a Russian Russia tungsten mine (Timofeev, Kosheleva & Kasimov, 2018). Compared with the findings for a northern tungsten mine in China (Wu, Wang & Chen, 2020), in this study, the contents of heavy metal were much lower, except for Mo and Cr. The heavy metal concentrations reported for a southern tungsten mining area were not significantly different to those found in the present study, which could be explained by the soil properties of both sites (Guo et al., 2017; Zheng et al., 2020). The significant lower concentration of metals in the study area compared with the northern tungten mining area may be resulted in the distincition of mining methods, climate types of the two mining areas. Heavy metal pollution in mining areas in southern China is mostly concentrated in river valleys. Due to large rainfall and long duration, heavy metal migration is fast, which is not conducive to accumulation in the soil (Chen, Fan & Long, 2024). In the north, it is semi-arid with strong winds, and the dust coverage from open-pit mining is wide and not easily lost, resulting in the accumulation of heavy metals in the soil for a long time (Mao, Ma & Zou, 2016).

The mean concentrations of the eight tested heavy metals followed the order FJ-S > QL > FJ-N > HL > CJ-E > CJ-W, with a significant distribution throughout the Zhangjiang River basin (Zheng et al., 2020). The spatial distribution of heavy metal concentrations is shown in Figs. S1 to S8. This indicates that the degree of pollution in the sampling areas was significantly impacted by the distance from the tungsten mining area. Sites FJ-S and FJ-N were only 2 and 4 km, respectively, away from the abandoned mining area (Xihuashan) (Wang et al., 2009), with a history of 100 years of mining, resulting in long-term metal accumulation. Site QL was closer to the current mining area Piaotang (Han, Zhu & Wang, 2019), with more active metal accumulation. As a control point in the valley on the southern bank of the Zhangjiang River, HL was not only far from historical tungsten mining areas but also was not impacted by the operating mines. Therefore, the concentrations of heavy metals were close to those detected in traditional agricultural areas.

Correlation between heavy metals and soil properties

Table 2 shows the soil properties for the different sampling sites. Of all soil samples, 73.7% were weakly acidic. The pH values ranged from 3.74 to 6.67, with an average of 5.24 ± 0.46. The average pH value was similar to the background value determined for Ganzhou area (Chen et al., 2019; Wang et al., 2015, 2019). The highest and lowest mean pH values were observed for FJ-N (5.37) and HL (4.93), respectively. This variation can be attributed to differences in resident population and human activities, which can notably influence the soil properties (Shen et al., 2017; Vega et al., 2004).

The concentration of total organic carbon (TOC) in soil samples from site FJ-S was highest, with a mean value of 2.50%. In comparison, soil samples from sites QL, CJ-W, and CJ-E sites showed lower TOC levels, with mean values of 1.83%, 1.85%, and 2.05% respectively. The fractionation and composition of TOC is of great importance when determining the adsorption capacity of soils. The higher TOC content at site FJ-S suggests a greater capacity to adsorb pollutants, likely due to higher inputs of organic matter, such as vegetation litter, expanding the carbon pool. The lower TOC levels in the other sampling sites indicate a more limited organic matter accumulation, constraining the adsorptive potential of the contaminants (Xiao et al., 2017). However, at site QL, with high heavy metal concentrations, an increased level of human activities could result in a low TOC level, which was the reason for the bioavailable metal supplementation via sewage effluent or mineral fertilizers (Nacke et al., 2013).

The concentrations of alkali metal oxides (K₂O and Na₂O) were below the levels of the continental crustal abundances in China (Li et al., 2014). The content of K₂O was significantly higher at FJ-S, QL, and CJ-W (mean value: 2.54%, 2.43%, and 2.54%, respectively), whereas that of Na₂O was significantly higher at FJ-S (mean value: 0.33%). The CaO concentration was higher at sites FJ-S, QL and CJ-E (mean values: 0.25%, 0.26%, and 0.24%, respectively), but this heavy metal was strongly positively correlated with pH (Table S2). At site FJ-N, levels of the above four metallic oxides were lowest. Sites CJ-W and CJ-E had the lowest Fe₂O₃ concentration (mean values: 2.55% and 2.88%, respectively).

Based on the Pearson’s correlation coefficients (Table 3), Mo was significantly positively correlated with pH (p < 0.05), whereas Cr and Zn were highly significantly positively correlated with TOC (p < 0.01). The Cr was highly significantly positively correlated with CEC (p < 0.01) and significantly positively correlated to Fe₂O₃ (p < 0.05). These results indicate that higher pH values can facilitate the accumulation of Mo in soil, whereas Cr and Zn are mainly accumulated in the TOC. Consequently, rapid organic degradation leads to rapid Zn cycling (Liu & Han, 2021). The Cr levels slightly fluctuated, and Cr was positively correlated with Fe₂O₃ and TOC, indicating that it was derived from natural sources. The heavy metals As, Cd, Cu, Mo, Pb, W, and Zn were highly significantly positively correlated with K₂O and Na₂O (p < 0.01), whereas CaO was significantly positively correlated with Mo and Pb (p < 0.05) and highly significantly positively correlated with Cu and Zn (p < 0.01). As seen in Table S5, CaO was strongly positively correlated with pH and CEC, thereby affecting the bioavailability of heavy metals such as Mo, as a result of mine wastewater treatment (Chen, Zhu & Huang, 2007; Meng et al., 2021). This leads us to infer that K₂O and Na₂O may be principal factors in the accumulation and migration of heavy metals, such as As, Cd, Cu, Mo, Pb, W, and Zn, in soil, whereas CEC and TOC play an important role in Cr accumulation. According to previous studies, K₂O, Na₂O, and CaO are mainly derived from human activities. Whilst K₂O accumulation is the result of the use of inorganic and organic fertilizers (Lambers & Barrow, 2020), Na₂O and CaO are the main agents in the treatment of tungsten ore mining wastewater (Zhang et al., 2012). This suggests that long-term tungsten ore smelting and agricultural fertilizing have a significant impact on the soil heavy metal content, which has also been reported in previous studies (Hui et al., 2021; Zhang et al., 2017).

Pollution and risk assessment of heavy metals in soil

The geo-accumulation index (I_geo) has been widely used to assess the pollution status of metals in sediments and soils (Sawe, Shilla & Machiwa, 2019; Timofeev et al., 2020; Zahra et al., 2014). In the present study, the mean values of I_geo in soil followed the order Mo > W > Cd > As > Pb > Cr > Zn (Table S3). Regarding the pollution classes, Mo and W (with mean values of 2.59 and 2.51, respectively) pollution was moderate to heavy, whereas Cd pollution was moderate (class 3), and for As and Cu, the levels indicated no to moderate pollution. The mean contamination factors (CFs) (Rastegari Mehr et al., 2017) for As, Cd, Cu, Cr, Pb, Mo, W, and Zn were 2.98, 5.56, 1.83, 1.31, 1.67, 15.16, 16.55, and 1.41, respectively (Table 4). The contamination factors for Cu, Cr, Pb, and Zn indicated moderate contamination (Inengite, Abasi & Walter, 2015) and were significantly lower compared to those for the other tested metals; this leads us to infer that they were derived from natural sources rather than human activities. Although As pollution was moderate, it was close to being considerable. The highest contamination factors were observed for Mo and W, indicating very high contamination (mean values:15.16 and 16.55, respectively), which was due to the long-term tungsten mining in this area (Wu, 1993). In contrast, Cd pollution was moderate to high. The main heavy metals found across all sampling sites were Mo, W, and Cd. As their levels greatly exceeded the ABV, they need to be taken into consideration regarding environmental pollution (Shi et al., 2022).

Figure 3 shows the I_geo spatial distribution of the tested heavy metals. Soil I_geo values for As, Cu, Cd, and Zn below 0 were found in sites CJ-W and CJ-E, indicating that agricultural soils were largely unaffected by these metals. For Mo and W, the Igeo values were higher than 2 and lower than 4 in sites FJ-N, FJ-S, HL, and QL, classifying these sites as moderately to heavily polluted and heavily polluted, respectively. For As, moderate pollution was observed, whereas for Cu, Pb, and Zn, the sites were categorized as unpolluted to moderately polluted. Additionally, Cd pollution was moderate to heavy at sites FJ-S and QL. Based on the I_geo findings, sites FJ-S and QL received varying amounts of heavy metals, except Cr. The PLI findings indicated extremely heavy pollution (PLIs > 3) at sites FJ-S and QL, mostly with Cd, Mo, and W (Table 4). Sites CJ-W and CJ-E exhibited low pollution levels compared to the other sites. The heavy metal accumulation period of site FJ-S was prolonged, and the tailing pond located in the valley was prone to leakage, resulting in the pollution of both surface water and groundwater (Reutova et al., 2022). Site QL served as a busy transportation hub for tungsten ore, and the accumulation of heavy metals was worsened by irrigation with sewage (Lu, Zheng & Lai, 1997).

Table 5 shows the ecological risk posed by heavy metals. The ecological risk factor ( ${\mathrm{E}}_{\mathrm{r}}^{\mathrm{i}}$) values for Cr, Pb and Zn were below 40 in all sampling aeras. However, for As, the ${\mathrm{E}}_{\mathrm{r}}^{\mathrm{i}}$ values were close to 40 in sites FJ-S and QL (mean: 41.32 and 39.25, respectively), indicating a moderate risk. The ${\mathrm{E}}_{\mathrm{r}}^{\mathrm{i}}$ values of W (mean: 35.04, 63.98, 34.99 and 36.72) had reached or cloesd to a moderate risk in sites FJ-N, FJ-S, HL and QL near the Tungsten mining area. For Cd, the risk degree ranged from considerable risk to very high risk, indicating serious contamination and human health risks. The RI values indicated a considerable risk at FJ-S (value: 456) and a low risk at CJ-E (value: 124), whereas for the other sites, a moderate risk was detected. Risk analysis revealed Cd as the primary contaminant of concern, with sites FJ-S and QL exhibiting the highest risk quotients and contamination levels. Regarding other heavy metals such as As and W, some sites were moderately polluted. The overall ecological risk pattern highlights Cd contamination, especially at sites FJ-S and QL, requiring targeted risk management.

Concentrations and transfer of metals in paddy rice

Compared to the concentrations of heavy metals for paddy rice (Table 6) (As, p < 0.01; Cr, Pb, Mo, Zn, p < 0.05; Cd, Cu, p > 0.05), the mean values of Cd, Cu, Mo, Pb, and Zn were the highest in FJ-S, whereas the As concentration was highest in FJ-N. In CJ-W, the heavy metal concentrations for paddy rice were the lowest. The total concentration of heavy metals in paddy rice followed the order FJ-S > FJ-N > QL > CJ-E > HL > CJ-W. For each metal of accumulation in paddy rice followed the order Zn > Cu > Cd > As > Cr > Pb > Mo; the Cr concentration was largely stable. Based on these results, large amounts of heavy metals had accumulated in the farmland next to the historical mining area, which further exacerbated metal accumulation in rice. Active mining areas also release large amounts of heavy metals into the soil. As heavy metals are affected by rainfall and enter irrigation water, they accumulate in rice plants (Wu, Wang & Chen, 2020). Based on the TF values, there were differences in the concentration accumulation values (Table S4). Site QL showed the highest TF values for As, Cd, Cu, Mo, and Zn, whereas for Pb and Cr, the highest values were found for CJ-E. For all tested metals, the values were below 1, except for Cd, indicating that paddy rice had a poor ability to accumulate metals in soil. The TF values for Cd were above 2 in FJ-S, HL, and CJ-E and above 5 in QL, indicating serious accumulation. This results is coincident with the study of Liu et al. (2023) which indicates that the paddy rice has high TF and BAF values (mean:0.99 and 0.91, respectively) of Cd. Yang et al. (2021) reached the opposite conclusion when studying the enrichment of Cd in rice in southern karst geological areas. The high calcium carbonate content in the soil of the karst geological area, which may impede the accumulation of Cd by rice, may explain their different result.

Soil properties and composition play crucial roles in metal accumulation (Chen et al., 2016; Tunc & Sahin, 2017). In this study, Pearson correlation analysis revealed a significant association of Na₂O and K₂O with the concentration and transfer of specific metals such as Cd. However, this contradicts the expected positive relationship between Na₂O, K₂O, and Cd (Table S5), which suggests that Na₂O and K₂O might hinder metal accumulation in paddy rice by precipitating metals in soils. Previous research has indicated that sodium and potassium are essential nutrients for plant growth, suggesting that they may induce metal accumulation in plants by promoting plant growth (Shrivastav et al., 2020; Xu et al., 2020). This could potentially explain the positive influence of K₂O and Na₂O on Cd accumulation observed in this study. Furthermore, Na₂O and K₂O might also enhance Cd accumulation in paddy rice by altering the structure of microorganisms (Yang et al., 2020; Wang et al., 2021; Zheng et al., 2021). Nevertheless, this study did not assess indicators related to rice growth and soil microorganisms. Our future research will continue to investigate the impact of K₂O and Na₂O on Cd accumulation, considering rice growth and soil microorganisms. Additionally, we aim to further explore the primary influencing factors of heavy metal enrichment in rice in mining areas.

Assessment of the health risks posed by heavy metals in paddy rice

Figures 4, 5 and Table S6 show the results of the health risk assessment. The average daily dose (ADD) values for non-carcinogenic risk were highest for Cd, Cu, and Zn at site FJ-S, exceeding 0.001 mg/kg/day. However, As was the only metal with an ADD above the 0.001 mg/kg/day reference limit, specifically at site FJ-N. The hazard quotient (HQ) values for As were above 1 across all study areas, whereas that for Cd was below 1 only at CJ-W. At HQ values above 1, there are potential non-carcinogenic effects on humans (United States Environmental Protection Agency, 2011). The order of As HQ was FJ-N > FJ-S > QL > CJ-E > HL > CJ-W (values: 4.03, 3.32, 2.48, 2.47, 2.14, and 1.63, respectively), accounting for 50.88%, 25.40%, 26.05%, 34.69%, 30.01%, and 44.17% of the hazard index (HI), respectively. The levels of Cd HQ followed the order FJ-S > QL > HL > CJ-E > FJ-N > CJ-W (values: 7.99, 5.61, 3.61, 3.18, 2.41, and 0.77), accounting for 61.13%, 58.93%, 50.63%, 44.66%, 30.43%, and 20.87% of the HI, respectively. The HI value exceeded 10 at site FJ-S, indicating health risks and risk of chronic poisoning. The total cancer risk (TCR) was significant across all sites, with As, Cd and Cr cancer risk values(CR) above 1 × 10⁻⁴ a⁻¹ (Dong et al., 2022), indicating high carcinogenic risk. Further, As and Cd may represent a considerable potential ecological risk at some sites (FJ-S and QL), based on the pollution characteristics of agricultural soil in mining zones (Du et al., 2015). According to the conclusion above, it is crucial to focus on the heavy metal pollution situation at sites FJ-S and QL, reduce the planting of edible crops, and perform soil remediation.