Variability of heavy metal content in soils of typical Tibetan grasslands

Abstract

Relatively high contents of heavy metals were recently reported in the high-altitude Tibetan Plateau (TP) environment, but the source and distribution characteristics of heavy metals in grassland environments of the TP remain unclear. Here, we report the contents of Hg, Cd, As, Pb, Ni, Cr, Cu, and Zn in soils of typical grasslands from the western to eastern parts of the TP, and the factors regulating accumulation of these heavy metals in the soil. Results show a large degree of variability of the eight heavy metals in the topsoil (0–20 cm) of different grasslands of the TP. Distribution characteristics of the heavy metals in the subsoil (20–40 cm) from the different grasslands were similar to those in the topsoil. Concentrations of As, Cd, Cu, and Zn in grassland soil tended to decrease as longitude increased, whereas Hg content displayed an increasing trend with increasing longitude after 88°E. These observations may be related to different sources of the heavy metals in the soils. Our results suggest that Cu, Zn, Cr, Cd, Ni, and As were mainly derived from natural sources, whereas anthropogenic activities could be responsible for the accumulation of Pb and Hg in the soil. Positive correlation between average annual rainfall amounts and soil Hg content may suggest the contribution of precipitation to soil Hg content. However, accumulation and distribution of heavy metals in the grasslands also depend on soil characteristics which could influence the mobility of heavy metals. These findings have important implications for an understanding of the occurrence and accumulation of soil heavy metals in grasslands of high-altitude environments.

---

1. Introduction

Heavy metal pollution is of great concern due to its toxic and irreversible characteristics.^1–3 Soil pollution by heavy metals can pose serious health risks in food chains, through the translocation of toxic metals from soils to the edible parts of crops,^7,8 as well as threatening soil biodiversity and ecological functions.^9–11 Heavy metals in soils mainly originate from the weathering of parent materials and through anthropogenic emissions such as industrial waste, mining and smelting activities, and traffic.^4–6 Heavy metals in terrestrial ecosystems are also important sources of pollution of aquatic environments downstream, through surface runoff.^12–14 Additionally, some heavy metals or heavy-metal-containing compounds may be emitted into the atmospheric environment through volatilization (e.g., elemental Hg and monomethylarsine).¹⁵ Recently, much effort has been devoted to the study of heavy metals in areas of intensive human activity, which are important sources of pollution of the surrounding environments.^16–18 However, the source and distribution characteristics of heavy metals in remote high-altitude regions remain unclear.

The Tibetan Plateau (TP), occupying 2.5 million km², is one of the most remote regions on earth and is considered to receive little impact from human activities. The TP has a mean altitude of greater than 4000 m above sea level (a.s.l.), and much of the TP lies above the atmospheric boundary layer.¹⁹ Soils in the TP have originated from various parent materials, mainly involving residue deposits, glacial deposits and other sediments of the quaternary,²⁰ and are associated with various vegetation types. The diverse origins of the resulting soils and the complicated geology in this region have resulted in a range of different contents of soil elements.^21,22 The high-altitude TP plays an important role in the distribution and transport of pollutants, and heavy metals can also be transported over long distances in the TP.^23–25 The TP is adjacent to very densely populated and rapidly industrializing regions, including eastern China, India, and Nepal. Thus, anthropogenic activities from neighboring regions may lead to increasing inputs of heavy metals into the TP by long distance transport.^26–28 In addition, increasing urbanization and tourism-related activities could increase inputs of heavy metals into the soil in this region.²² Earlier studies reported concentrations of heavy metals in TP soils based on data of scientific expeditions conducted several decades ago.²⁰ Recently, large-scale investigations of heavy metals in topsoils of the TP have also been undertaken,^22,28 suggesting that natural sources have been the main contributors of heavy metals in the region. However, soils from the earlier studies were under various vegetation types, including coniferous forest, alpine meadow shrub, and other mixed types, which may obscure the identification of heavy metal sources in the TP,²⁰ since there were differences in the translocation of metals from soils to plants among the different ecosystems, as well as differences in the filtering of pollutants from atmospheric deposition. The occurrence of heavy metals in soils also depends on the edaphic and climatic conditions.^29–31 Therefore, it is essential to understand the accumulation characteristics and driving factors for soil heavy metals in a consistent type of ecosystem. Grassland ecosystems, covering more than 60% of the Qinghai-Tibetan Plateau, are highly sensitive to environmental disturbance.^32,33 Thus, it is particularly significant to investigate the content and distribution characteristics of soil heavy metals in typical grasslands of the TP.

The objective of this study was to investigate the contents and variability of soil heavy metals in typical grasslands from eastern to western areas of the TP, as well as examining how environmental variables influence the accumulation of heavy metals in the soil. All the soil samples were collected from typical grasslands from the TP, and the concentrations of Hg, Cr, Cd, As, Cu, Zn, Ni, and Pb in them were measured and linked to environmental variables. The data from this study are compared with previously reported results both from this region and from other regions. In particular, characterization of the accumulation of heavy metals in the topsoil and subsoil is discussed. Multivariate analysis was used to identify potential sources of heavy metals in typical grasslands from the TP. Additionally, by establishing the relationships between heavy metals and the environmental factors examined, we propose the major drivers of heavy metal occurrence in the grassland soils of the TP.

2. Materials and methods

2.1 Study area and sampling

The climate in the TP is dominated alternately by the Indian monsoon in summer and westerly winds in winter, and the mean annual precipitation changes from ∼50 mm to ∼4000 mm from the northwest to the southeast of the TP. In this study, 23 typical grassland sites with similar landscape characteristics were selected along a longitude gradient of the TP ranging from about 84°E to 95°E (Fig. S1†). Sampling was carried out during August 2013, and the sites were identified using a global positioning system (GPS). Considering the large variability in natural samples, we collected three replicate samples at each site to enhance the representativeness of the soil samples. Each sample was produced by compositing three subsamples from different plots. At each site, topsoil (0–20 cm) and subsoil (20–40 cm) were collected to investigate the migration of heavy metals in soil profiles. Consequently, a total of 69 (23 × 3) topsoil and 69 subsoil samples were obtained. After sampling, these soil samples were sealed in plastic bags and shipped to the laboratory where they were freeze-dried, manually ground, and subsequently passed through a 0.149 mm sieve. Detailed site information and soil properties are given in Table S1.†

2.2 Soil heavy metal analysis

For Hg analysis, 0.50 g of each soil sample was digested with 10 mL of mixed acid solution (2 mol L⁻¹ HNO₃ and 4 mol L⁻¹ HCl) in a Teflon tube at 100 °C for 2 hours. Digestates were diluted with deionized water to 50 mL volume, after which Hg in the solution was determined by cold vapor atomic fluorescence spectrometry (CVAFS) following reduction by SnCl₂. The method detection limit was 0.2 ng L⁻¹. For other heavy metals, 0.3 g of soil was digested by trace metal grade acids (9.0 mL of concentrated HNO₃ and 3.0 mL of concentrated HF) using a MARS microwave digestion system (CEM, USA) according to U.S.EPA method 3052.³⁴ The HF in the digested solution was removed by evaporating the liquids to near dryness, and then the remaining residues were re-dissolved using 3% HNO₃. The concentrations of Cr, Pb, Cd, Cu, Zn, As, and Ni in the final solution were determined by a 7700X Inductively Coupled Plasma-Mass Spectrometer (Agilent, USA). Two standard reference materials GBW-07401 (GSS-1) and GBW-07405 (GSS-5) were included in the analytical procedure for heavy metals for quality assurance/quality control, and procedural blanks were performed every 10 samples. Recovery of target heavy metals in the standard reference materials was reasonably good (81–118%).

2.3 Data analysis

Descriptive statistics of heavy metals in the topsoil, including maximum, minimum, mean and median concentrations, were performed to reflect the degree of discrete distribution of the target metals. The top enrichment factor (TEF) was evaluated by the ratio of the metal concentrations in the topsoil to subsoil. Correlation analysis (Pearson) was performed to explore the relationships between standardized heavy metal concentrations and longitude. Spearman's correlation analyses were carried out to assess relationships between geochemical characteristics, heavy metals, longitude, and climatic factors. Principal component analysis (PCA) was performed to identify sources of heavy metals in soils of the TP. PCA is widely used for distinguishing between different sources of heavy metals in complex matrices by reducing the number of observed variables.^5,6,35 The tests for normality on raw and log-transformed data were performed before PCA. To make the results more easily interpretable, PCA with VARIMAX normalized rotation was applied, which can maximize the variances of factor loadings across variables for each factor. Cluster analysis (CA) was then performed to further identify heavy metals with similar distribution patterns. All of the statistical analysis was conducted using SPSS 19.0. Relationships were considered statistically significant at p < 0.05.

3. Results and discussion

3.1 Heavy metal contents in soils of typical grasslands from the TP

Soil heavy metal contents and basic characteristics are shown in Table 1. There was large variability of soil water content in different grasslands, ranging from 3.1% to 69.3%. The average soil pH changed from 4.9 to 8.2, and relatively high organic carbon contents were observed in grassland soils from the eastern TP. The mean concentrations of Hg, Cr, Ni, Cu, Zn, Cd, Pb, and As were 0.078 mg kg⁻¹, 51.89 mg kg⁻¹, 26.11 mg kg⁻¹, 24.69 mg kg⁻¹, 75.06 mg kg⁻¹, 0.144 mg kg⁻¹, 19.43 mg kg⁻¹, and 18.40 mg kg⁻¹, respectively (Table 1). There was significant variability in the contents of Hg, As, and Cd (Fig. S2†). Owing to low background values in the TP, none of the heavy metals exceeded the limit values for pollution according to the Chinese Environmental Quality Standard for soil (GB 15618-1995). Among the eight heavy metals, the concentrations of Hg, Cu, and Zn are comparable to the background values for world soils,²¹ while the mean concentrations of the other elements are a little lower than the background values for world soils. These differences may be mainly attributable to differences in the parent materials from which the soils have been developed, as previously discussed.^20,22 Interestingly, compared to background values for Tibetan soils, we found higher mean concentrations of Hg and Cd. The increased element concentrations may be attributable to long-distance transport of metals from surrounding regions, given that atmospheric circulation is one of the most common pathways for the movement of heavy metals into terrestrial ecosystems.^36–38 In addition, data from earlier studies are from soils under various vegetation types, which could have influences upon the identification of heavy metal sources.

Table 1 Concentrations of heavy metals soil properties in topsoil of typical grasslands from the Tibetan Plateau (mg kg⁻¹). SWC, soil water (%); SOC, soil organic carbon (g kg⁻¹); clay, percentage of clay component of soil texture (%); CEC, cation exchange capacity (cmol kg⁻¹)

	Minimum	Maximum	Mean	Median	Background value of Tibetan soils^a	Recently reported values^b	Background values^c
a Data from the study in 1993 by Chen and Tian.b Data from the study in 2012 by Shen et al.c Data from the study in 1979 by Bowen.
Hg	0.001	0.194	0.078	0.080	0.026	0.037	0.060
Cr	4.06	188.41	51.89	47.91	77.40	155.54	70.00
Ni	15.81	80.75	26.11	23.16	32.10	55.86	40.00
Cu	17.62	42.43	24.69	23.97	21.90	24.27	30.00
Zn	33.61	107.97	75.06	74.80	73.70	75.59	90.00
Cd	0.091	0.312	0.14	0.134	0.080	0.141	0.350
Pb	8.72	59.48	19.43	17.79	28.90	32.15	35.00
As	2.56	33.97	18.40	17.65	18.70	19.27	7.20
pH	4.69	8.98	7.00	7.32	—	—	—
SWC	2.33	82.49	18.67	9.68	—	—	—
SOC	0.66	23.98	6.27	3.35	—	—	—
Clay	1.42	70.11	48.39	56.80	—	—	—
CEC	3.29	46.34	14.42	11.42	—	—	—

---

In order to understand spatial effects on soil heavy metal concentrations for these grasslands, we investigated the relationship between the soil heavy metal concentrations and the longitude (Fig. 1). There are differential distribution patterns for the heavy metals in the grassland soils along longitudinal gradients from about 84°E to 95°E. All the heavy metals in the subsoil exhibited similar distribution characteristics to those in the topsoil. The concentrations of As, Cd, Zn, and Cu in topsoil negatively correlated with the longitude, but there was no significant relationship between Cr, Ni, and Pb concentrations and the longitude. Differentially, we found significant positive relationships between the Hg content in both topsoil and subsoil and the longitude from 88°E to 95°E. The effect of longitude on the distribution characteristics of heavy metals in the grasslands may be related to different sources and accumulation mechanisms in soils from different sites. For example, the higher Hg content in the eastern TP than in the western TP may be partially explained by greater rainfall amounts (Table S3†), which could result in more Hg wet deposition into soils. However, we did not note any significant correlation between Hg and longitude west of 88°E, which could be associated with the relatively low mean annual rainfall, ranging from 100 to 280 mm (Table S1†). These observations are in line with a recent study highlighting that precipitation is an important source of Hg in the southeastern TP,^19,39 though dry deposition may play a role in influencing the soil Hg concentration. Here, we used average annual precipitation to estimate the overall contribution of rainfall to the soil Hg pool, given that Hg wet deposition through precipitation is seasonal.³⁹ On the other hand, As, Cd, Zn, and Cu tended to decrease with increasing longitude. This contrasting observation may be ascribed to increased loss by leaching of heavy metals from the topsoil of the grasslands, resulting from an increase in average annual precipitation with increasing longitude.^20,40,41 At the same time, surface runoff may also be a pathway for movement of heavy metals into downstream ecosystems,^42,43 and this is also supported by previous findings of increasing metal contents downstream of the TP.¹⁴

Fig. 1 Relationships, for the Tibetan Plateau, between concentrations of heavy metals in the soil and longitude. Solid lines represent significant relationships (p < 0.05), while dashed lines represent no significant relationships (p > 0.05). Pearson's r and p values for the correlations are given in the figure panels. For Hg, a correlation was indicative of soil samples from sites at longitudes greater than 88°E (longitude), while no significant correlation was found for longitudes less than 88°E.

3.2 Possible sources of heavy metals in the TP

The TP, named as the “the third polar region”, is the most remote region in the word, and it is also a model ecosystem for environmental research, due to the pristine environment and low level of influence from anthropogenic activities. In the present study, we applied PCA with VARIMAX normalized rotation to identify possible sources of the heavy metals in soils of typical grasslands from the TP. We extracted four major factors, accounting for over 87.9% of the total variation in total heavy metal concentrations (Table 2). The commonality varied from 71% for Zn to 99% for Hg, indicating that the four extracted factors well represent the eight heavy metals examined in this study. The rotated component matrix shows that Cu, Zn, Cd, and As were strongly associated in factor 1, with high loading ranging from 0.65 to 0.88, which explains 52.3% of the total variation. These heavy metals are also highly associated according to Table 3, suggesting that they probably have a similar source in the soils. A previous study suggested that heavy metals such as Cu and Zn may exit synchronously from soil parent materials,²² and they could be derived from weathering of exposed bedrocks. Cr and Ni display high loading in factor 2, accounting for 15.5% of the total variation. As shown in Table 3, Cr is highly correlated to Ni (r² = 0.55, p < 0.01), and they may come from ultramafic rocks since soils developed from these rocks usually contain high levels of Ni and Cr.^18,22 Our results show that factor 3 is dominated by Pb, with 12.5% of the total explained variation, and Pb in the surface soils may come from atmospheric deposition that originated from traffic emissions and industrial point sources.^18,26,44 The lack of significant correlation between soil Pb concentrations and soil water or average annual precipitation may suggest that Pb accumulation in the soil could be a result of atmospheric dry deposition. Factor 4 is mainly responsible for Hg and accounted for 7.3% of the total variation. Studies have suggested comparable Hg inputs through atmospheric wet deposition from distant sources.^18,39,45,46 Long-distance transport of Hg through atmospheric circulation is one of the most important pathways for Hg accumulation in the soils of the TP.³⁹ Thus, we consider factor 4 to be an anthropogenic emission source dominated by wet deposition, which may be partially supported by a positive relationship between average annual precipitation and the soil Hg content of the TP (Table 3). However, it is difficult to estimate the contribution of dry deposition to soil heavy metals based on the present study.

Table 2 Rotated component matrix of heavy metals in the soils of the Tibetan Plateau^a

	1	2	3	4	Commonality
a PCA factor loadings greater than 0.46 are shown in bold.
Hg	−0.20	−0.10	0.01	0.97	0.99
Cr	0.13	0.96	0.13	−0.03	0.95
Ni	0.35	0.89	−0.02	−0.11	0.93
Cu	0.88	0.26	−0.01	−0.16	0.87
Zn	0.65	0.46	0.16	−0.22	0.71
Cd	0.77	0.14	0.37	−0.16	0.76
Pb	0.26	0.09	0.95	0.02	0.97
As	0.79	0.34	0.32	−0.02	0.84
Initial eigenvalue	4.20	1.24	0.99	0.759
Percent of variance	52.56	15.54	12.45	7.32
Cumulative percent	52.56	68.09	80.54	87.86

---

Table 3 Spearman's correlations matrix of heavy metals and environmental variables in topsoil of the Tibetan Plateau. p values below 0.05 are in bold^a

	Hg	Cr	Ni	Cu	Zn	Cd	Pb	As
a *p < 0.05 and **p < 0.01.
Cr	−0.12
Ni	−0.33**	0.65**
Cu	−0.32*	0.55**	0.84**
Zn	−0.35**	0.60**	0.75**	0.75**
Cd	−0.32**	0.35**	0.61**	0.68**	0.64**
Pb	−0.21	0.49**	0.57**	0.67**	0.65**	0.73**
As	−0.24*	0.46**	0.71**	0.75**	0.73**	0.83**	0.73**
pH	−0.15	−0.19	0.22	0.26*	0.10	0.20	−0.11	0.35**
SWC	−0.05	0.24*	−0.01	−0.03	−0.04	0.19	0.02	−0.25**
SOC	0.06	0.20	−0.11	−0.16	−0.13	−0.22	0.02	−0.30**
Clay	−0.03	0.28*	−0.00	0.15	0.12	−0.07	0.11	−0.02
CEC	−0.09	0.35*	−0.07	−0.05	0.08	−0.05	0.11	−0.13
Precipitation	0.30*	0.23	−0.05	−0.09	−0.10	−0.28*	−0.11	−0.36**
Humidity	0.08	0.25*	0.07	0.06	0.05	−0.14	0.06	−0.22
Longitude	0.19	0.22	−0.12	−0.23	−0.19	−0.31*	−0.12	−0.45**
Altitude	−0.08	−0.22	−0.35**	−0.14	−0.20	0.12	0.10	0.02

---

The first three rotated components are visualized in Fig. 2a, showing the associations among soil heavy metals in typical grasslands from the TP. In addition, hierarchical cluster analysis performed in this study reveals possible homogeneous groupings of the heavy metals (Fig. 2b). This analysis provides further support for our hypothesis that Pb and Hg clearly came from sources that were different from those of the other heavy metals in this region. In order to further evaluate the effect of anthropogenic activities on soil heavy metal accumulation, we compared the concentrations of the eight heavy metals in the topsoil with those in the subsoil. The results show relatively high average contents of Hg, Zn, and Cd in the topsoil (Fig. 3), suggesting possible enrichment of heavy metals in the topsoil. As we know, Pb and Hg are the most toxic pollutants that can enter into soils through atmospheric deposition; thus anthropogenic activity is probably a major factor influencing the contents of Pb and Hg in the grassland soils.^22,44 The concentrations of other heavy metals in the two soil layers are similar, and associated with natural sources of these metals in the soils of the TP.²² The top enrichment factor (TEF) was used to evaluate the impact of human activities on heavy metal concentrations in the topsoil;^40,47,48 this can eliminate the effect of parent materials by using the concentrations of heavy metals in deeper soil, rather than mean background values for world soils.⁴⁸ Generally, the deeper the subsoil, the more representative the TEF is. However, due to the great hardness of deep soils in the grasslands of the TP, we only collected a 20–40 cm soil layer as the subsoil in this study. The results show no strong top enrichment of any of the heavy metals along the longitudinal gradient, and most of the TEF values are below 1.5, except for Hg (Fig. 4). A plausible reason for these observations could be the leaching of the metals along the soil profiles, which lessens the differences in heavy metal concentrations between the topsoil and the subsoil of a typical TP grassland. Thus, not only topsoils but also subsoils are affected by the heavy metals, and how these pollutants impact upon the health of the ecosystem of the TP should also be a concern.

Fig. 2 Multivariate analyses of heavy metals in topsoil of the Tibetan Plateau. (a) 3-D plot for PCA loading of heavy metals. (b) Hierarchical dendrogram of the heavy metals obtained by the single-linkage clustering method.

Fig. 3 Boxplot of heavy metals in the topsoil and subsoil of the Tibetan Plateau.

Fig. 4 Top enrichment factors (TEF) of heavy metals in the soils of the Tibetan Plateau.

3.3 Factors influencing the accumulation of heavy metals in the soils

The accumulation of heavy metals in soils depends upon their inputs and outputs. According to the PCA results, Pb and Hg could originate mainly from human activities, while the other heavy metals could be derived from natural sources. In this study, climatic and edaphic factors were examined in relation to concentrations of soil heavy metals in order to explore their impacts on accumulation in the topsoils of TP grasslands. Correlations between the eight heavy metals and the variables examined are also included in Table 3. We found a positive relationship between average annual precipitation and Hg concentration, which further supports the previous argument that rainfall could be one of the most important pathways for entry of Hg into grassland soils of the TP.³⁹ However, there are negative correlations between precipitation and Cd and As concentrations, indicating that relatively high rainfall amounts may lead to loss of these metals from the subsoil. Considering the marginal difference in metal concentrations between topsoil and subsoil, these heavy metals may enter into deeper soils by leaching,²⁰ since most soils in the TP are sandy soils with substantial pores for water migration and metal mobility.⁴⁹ In addition, relatively high precipitation may lead to migration of metals from the grasslands to downstream environments through surface runoff.^42,43 Thus, precipitation is an important factor influencing the occurrence of heavy metals in TP grassland soils.

A large body of literature has demonstrated the important role of soil characteristics in regulating speciation and mobility of heavy metals,^50,51 although the extent and mechanism of this effect may vary among different soils on account of complicated reactions (e.g., dissolution and precipitation) in the soil system. Relationships between soil properties and heavy metals could, at least in part, explain the distribution characteristics of heavy metals in typical TP grassland soils. In this study, we found positive correlations of pH with As and Cu (p < 0.05). However, the effect of pH on the accumulation of heavy metals depends also on the pH range. For example, at a low pH of 3–5 or at a high pH above 8, an increased pH could increase the mobility of dissolved Hg²⁺. For As, our results agree with previous studies, suggesting that As precipitates easily and accumulates readily in high-pH soils.^52,53 It seems that the content of soil organic carbon (SOC) had a negative effect on soil As concentration (Table 3), though several studies have suggested that SOC has a positive effect on metal accumulation through the adsorption of metals to soils.^51,54 In fact, organic matter can supply organic chemicals to the soil solution, which may then serve as chelates and increase metal availability to plants.⁵⁵ Consequently, an enhanced SOC content reduced the metal content in topsoil. However, though these soil various properties show associations with heavy metals, our study suggests that atmospheric deposition and natural sources are the two most important factors controlling the occurrence and accumulation of heavy metals in the grassland soils of the TP.

4. Conclusions

This study provides a new insight into the occurrence and accumulation of heavy metals in soils of typical grasslands from the TP. The content and variability of soil heavy metals in grasslands from the TP is predominantly governed by natural sources, though human activities through atmospheric deposition may increase Pb and Hg accumulation in the soil. The occurrence of heavy metals in the grasslands can be influenced to some extent by soil physical and chemical characteristics, as well as by climatic conditions. With the rapid economic and industrial development of surrounding countries, ecosystems in the TP may face more serious threats of pollutant emission through atmospheric circulation. These findings significantly advance our understanding of the source and accumulation of heavy metals in grasslands of high-altitude environments.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (41571453 and 41230857). We are also grateful to Dr Jun-Tao Wang for his assistance in soil sampling. We would like to thank Dr A. M. Jubb for the English correcting of this manuscript.

References

1. K. E. Giller, Nature, 1988, 335, 676.
2. P. Abrahams, Sci. Total Environ., 2002, 291, 1–32.
3. C. Li, S. Zhang, J. Yang, Y. Shi, W. Yu, S. Liang, J. Song, Q. Xu, Y. Chen and J. Hu, RSC Adv., 2015, 5, 102332–102339.
4. B. J. Alloway, in Heavy metals in soils, Springer, 2013, pp. 11–50.
5. C. Zhang, Environ. Pollut., 2006, 142, 501–511.
6. T. Chen, Q. Chang, J. Liu, J. G. P. W. Clevers and L. Kooistra, Sci. Total Environ., 2016, 565, 155–164.
7. F.-J. Zhao, Y. Ma, Y.-G. Zhu, Z. Tang and S. P. McGrath, Environ. Sci. Technol., 2015, 49, 750–759.
8. H. Zhang, X. Feng, T. Larssen, L. Shang and P. Li, Environ. Sci. Technol., 2010, 44, 4499–4504.
9. Y.-R. Liu, J.-J. Wang, Y.-M. Zheng, L.-M. Zhang and J.-Z. He, Microb. Ecol., 2014, 68, 575–583.
10. J. Li, Y.-B. Ma, H.-W. Hu, J.-T. Wang, Y.-R. Liu and J.-Z. He, Front. Microbiol., 2015, 6, 31.
11. W. Maliszewska, S. Dec, H. Wierzbicka and A. Woźniakowska, Environ. Pollut., Ser. A, 1985, 37, 195–215.
12. S. Audry, J. Schäfer, G. Blanc and J.-M. Jouanneau, Environ. Pollut., 2004, 132, 413–426.
13. E. F. da Silva, S. F. Almeida, M. L. Nunes, A. T. Luís, F. Borg, M. Hedlund, C. M. de Sá, C. Patinha and P. Teixeira, Sci. Total Environ., 2009, 407, 5620–5636.
14. X. Huang, M. Sillanpää, B. Duo and E. T. Gjessing, Environ. Pollut., 2008, 156, 270–277.
15. C. L. Miller, D. B. Watson, B. Lester, J. Y. Howe, D. H. Phillips, F. He, L. Liang and E. M. Pierce, Environ. Sci. Technol., 2015, 49, 12105–12111.
16. B. Wei and L. Yang, Microchem. J., 2010, 94, 99–107.
17. G. Yaylalı-Abanuz, Microchem. J., 2011, 99, 82–92.
18. Y. Hu, X. Liu, J. Bai, K. Shih, E. Y. Zeng and H. Cheng, Environ. Sci. Pollut. Res., 2013, 20, 6150–6159.
19. H. Yang, R. W. Battarbee, S. D. Turner, N. L. Rose, R. G. Derwent, G. Wu and R. Yang, Environ. Sci. Technol., 2010, 44, 2918–2924.
20. X. P. Zhang, W. Deng and X. M. Yang, J. Asian Earth Sci., 2002, 21, 167–174.
21. H. J. M. Bowen, Environmental Chemistry of the Elements, Academic Press, 1979.
22. J. Sheng, X. Wang, P. Gong, L. Tian and T. Yao, Environ. Sci. Pollut. Res., 2012, 19, 3362–3370.
23. J.-J. Shao, J.-B. Shi, B. Duo, C.-B. Liu, Y. Gao, J.-J. Fu, R.-Q. Yang, Y. Cai and G.-B. Jiang, RSC Adv., 2016, 6, 541–546.
24. M. Loewen, S. Kang, D. Armstrong, Q. Zhang, G. Tomy and F. Wang, Environ. Sci. Technol., 2007, 41, 7632–7638.
25. Y.-R. Liu, J.-X. Dong, L.-L. Han, Y.-M. Zheng and J.-Z. He, Environ. Pollut., 2016, 209, 53–59.
26. J. Luo, J. She, P. Yang, S. Sun, W. Li, Y. Gong and R. Tang, Ecotoxicology, 2014, 23, 1086–1098.
27. X. Wang, H. Yang, P. Gong, X. Zhao, G. Wu, S. Turner and T. Yao, Environ. Pollut., 2010, 158, 3065–3070.
28. C. Li, S. Kang and Q. Zhang, Environ. Pollut., 2009, 157, 2261–2265.
29. I. M. Sokolova and G. Lannig, Clim. Res., 2008, 37, 181–201.
30. D. P. Krabbenhoft and E. M. Sunderland, Science, 2013, 341, 1457–1458.
31. C. Gast, E. Jansen, J. Bierling and L. Haanstra, Chemosphere, 1988, 17, 789–799.
32. S. Piao, J. Fang and J. He, Clim. Change, 2006, 74, 253–267.
33. H. Chen, Q. Zhu, N. Wu, Y. Wang and C.-H. Peng, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, E93.
34. E. USEPA, Test Methods for Evaluating Solid Waste, 1995.
35. A. Facchinelli, E. Sacchi and L. Mallen, Environ. Pollut., 2001, 114, 313–324.
36. T. Berg and E. Steinnes, Environ. Pollut., 1997, 98, 61–71.
37. J. Gunawardena, P. Egodawatta, G. A. Ayoko and A. Goonetilleke, Atmos. Environ., 2013, 68, 235–242.
38. A. J. Dore, S. Hallsworth, A. G. McDonald, M. Werner, M. Kryza, J. Abbot, E. Nemitz, C. J. Dore, H. Malcolm and M. Vieno, Sci. Total Environ., 2014, 479, 171–180.
39. J. Huang, S. Kang, Q. Zhang, J. Guo, M. Sillanpää, Y. Wang, S. Sun, X. Sun and L. Tripathee, Environ. Pollut., 2015, 206, 518–526.
40. K. Page, M. J. Harbottle, P. J. Cleall and T. Hutchings, Sci. Total Environ., 2014, 487, 260–271.
41. W. Hartley, R. Edwards and N. W. Lepp, Environ. Pollut., 2004, 131, 495–504.
42. Z. He, M. Zhang, D. Calvert, P. Stoffella, X. Yang and S. Yu, Soil Sci. Soc. Am. J., 2004, 68, 1662–1669.
43. M. Huber, A. Welker and B. Helmreich, Sci. Total Environ., 2016, 541, 895–919.
44. H. Yongming, D. Peixuan, C. Junji and E. S. Posmentier, Sci. Total Environ., 2006, 355, 176–186.
45. L. Burger Chakraborty, A. Qureshi, C. Vadenbo and S. Hellweg, Environ. Sci. Technol., 2013, 47, 8105–8113.
46. J. Huang, S. Kang, S. Wang, L. Wang, Q. Zhang, J. Guo, K. Wang, G. Zhang and L. Tripathee, Sci. Total Environ., 2013, 447, 123–132.
47. T. Sterckeman, F. Douay, N. Proix and H. Fourrier, Environ. Pollut., 2000, 107, 377–389.
48. P. Blaser, S. Zimmermann, J. Luster and W. Shotyk, Sci. Total Environ., 2000, 249, 257–280.
49. A. Kessler and H. Rubin, J. Hydrol., 1987, 91, 187–204.
50. S. Sauve, W. Hendershot and H. E. Allen, Environ. Sci. Technol., 2000, 34, 1125–1131.
51. F. Zeng, S. Ali, H. Zhang, Y. Ouyang, B. Qiu, F. Wu and G. Zhang, Environ. Pollut., 2011, 159, 84–91.
52. P. H. Masscheleyn, R. D. Delaune and W. H. Patrick Jr, Environ. Sci. Technol., 1991, 25, 1414–1419.
53. M. Komárek, A. Vaněk and V. Ettler, Environ. Pollut., 2013, 172, 9–22.
54. R. Shrestha, R. Fischer and M. Sillanpää, Int. J. Environ. Sci. Technol., 2007, 4, 413–420.
55. F. Vega, E. Covelo, M. Andrade and P. Marcet, Anal. Chim. Acta, 2004, 524, 141–150.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra23868h

---