Concentrations and characteristics of selenium in soil samples from Dashan Region, a selenium-enriched area in China

Abstract

In order to clarify the total selenium (Se) content and species of Se in the soils of Dashan Region, a typical Se-enriched area in China, 69 soil samples with different land-use purposes were collected, and the concentrations of total Se, water-soluble Se, exchangeable Se, acid-soluble Se, organic-bound Se and residual Se and the physical and chemical properties of these soils were investigated. The total Se content ranged between 0.31 and 7.65 mg kg⁻¹ (dry weight). The concentrations of water-soluble, exchangeable, acid-soluble, organic-bound and residual forms of Se were 7.71–26.8 μg kg⁻¹, 24.7–93.3 μg kg⁻¹, 112–1430 μg kg⁻¹, 26–765 μg kg⁻¹ and 1300–5330 μg kg⁻¹, respectively. The concentrations of total Se were positively correlated with the contents of clay and soil organic matter, but were negatively correlated with pH values. The contents of bioavailable Se, which consists of water-soluble Se and exchangeable Se, were negatively correlated with the contents of soil organic matter, but positively correlated with pH values. The results suggest that pH value and the content of organic matter are key factors that regulate the bioavailability of Se in soils of Dashan Region.

Key words:

• Dashan Region
• selenium speciation
• soil
• soil texture
• total selenium

1. INTRODUCTION

Selenium (Se) is an essential trace element for humans and other animals. Se deficiency can lead to human diseases, such as Keshan disease (an endemic cardiomyopathy) and Kashin–Beck disease (a type of osteoarthritis; Coppinger and Diamond 2001). Excess intake of Se can also be toxic to organisms (Zhu and Zheng 2001; Moreno Rodriguez et al. 2005). For example, acute Se toxicity can cause respiratory, gastrointestinal and cardiovascular problems, while chronic exposure to high levels of Se can result in hair loss, fragile nails, mental problems, garlic-smelling breath, and excessive tooth decay and discoloration (Pedrero and Madrid 2009). A number of studies have proven the pivotal role of Se in anti-aging, repairing cells, improving human immunity and preventing carcinogenesis. Furthermore, Se is also an integral component of various enzymes such as glutathione peroxidase (GSH-Px) and thioredoxin reductase (TR), which participate in the antioxidant protection of cells (Rotruck et al. 1973; Birringer et al. 2002).

Se is ubiquitous in various environmental medias including rocks, soils, plants, the aquatic system and the atmosphere. Se enters the environment through volcanic activity, burning of fossil fuels, weathering of rocks and soils, groundwater transportation, precipitation of minerals, adsorption and desorption of elements, chemical or bacterial redox reactions, and metabolic uptake and release by plants and animals (McNeal and Balistrieri 1989). The global mean concentration of total Se in soil is 0.4 mg kg⁻¹, ranging from 0.01 to 2 mg kg⁻¹ (Dungan and Frankenberger 1999). In England, total Se in soils ranges from 0.1 to 4 mg kg⁻¹ (Hawkesford and Zhao 2007). In northwest India, the average concentration of Se in soils is 3.1 mg kg⁻¹ (Bajaj et al. 2011). In New Zealand, the average concentration of Se in soils is 0.5 mg kg⁻¹ (Oldfield 1999). In the semi-arid Central Spain, the concentration range of Se in soils is 0.17 to 0.39 mg kg⁻¹ (Moreno Rodriguez et al. 2005). In China, the mean concentration of Se in soils is 0.29 mg kg⁻¹. According to the classification of Tan (1996), the levels of Se in China could be classified into six grades: deficient (< 0.125 mg kg⁻¹), marginal (0.125–0.175 mg kg⁻¹), moderate (0.175–0.450 mg kg⁻¹), enriched (0.45–2.00 mg kg⁻¹), high (2–3 mg kg⁻¹) and excessive (> 3 mg kg⁻¹). About 72% of the counties in China present a lack of Se (< 0.1 mg kg⁻¹) in soil, and about one third of these counties exhibit serious shortages of Se (< 0.02 mg kg⁻¹) in the soil (Zhao et al. 2005). The amount of bioavailable selenium (B-Se) seems to be the key to determining the Se content in plants (Ellis and Salt 2003). Hence, the content of Se in plant tissues is mainly determined by the concentration of B-Se in soil (Combs 2001; Hawkesford and Zhao 2007). The proportion of B-Se varies in different areas. The proportion of B-Se in soils of Enshi is less than 5% (Yuan et al. 2012). In England, the proportions of B-Se in soils are 1.1–3.4% (Stroud et al. 2010). In Punjab, India, the proportions of B-Se in soils are close to 50% (Bajaj et al. 2011).

The concentrations of Se in soils vary with different land use; Zhang et al. (2005) found that the concentration of Se in a forest area (1.36 mg kg⁻¹) was more than 4 times higher than that in farmland (0.36 mg kg⁻¹), and almost 2 times higher than that in grassland (0.67 mg kg⁻¹) of the Hong Kong Special Administration Region. Yanai et al. (2012) found that the concentration of Se in upland (0.56 mg kg⁻¹) was about 1.24 times higher than that of paddy (0.45 mg kg⁻¹) in Japan. These phenomena might be caused by the different soil physicochemical properties and anthropogenic activities of different land-use types (Zhang et al. 2005; Yanai et al. 2012).

Dashan Region is located in the south of Anhui province, China. It is known as a Se-enriched area, together with the other two Se-enriched areas Enshi, Hubei and Ziyang, Shanxi, China. Similarly to the other two areas, Se in soils of Dashan Region mainly comes from carbon-siliceous sediment (also called “stone coal”) which consists of Se-rich carbonaceous chert and carbonaceous shale (Yuan et al. 2013). This carbon-siliceous sediment contains a high content of Se (the average concentrations of Se in stone coal of Enshi, Ziyang and Dashan Region are 329 mg kg⁻¹, 20 mg kg⁻¹ and 37.4 mg kg⁻¹, respectively), and the weathering products might contain clay minerals, carbonate minerals, silicate minerals, etc. (Yan 1993; Wu et al. 2007b). Furthermore, the distribution of Se in soils of Dashan may also be influenced by human activities, and weathering and eluviation of the parent soil material (Zhu et al. 2008). The parent soil material of Dashan is limestone (Wu et al. 2007b), and soils of this region are classified as yellow brown soils (Cambisols, Alisols; FAO–UNESCO 1990). In 1960–1970, many selenosis incidents occurred in humans and animals in Enshi and Ziyang (Yang et al. 1983; Zhao et al. 1993); unlike these two areas, no selenosis incidents have occurred in Dashan Region to date. However, the health risk caused by chronic exposure of high levels of Se cannot be ignored in Dashan.

As yet, no study has clarified the total Se content and its species in the soils of Dashan Region. The objectives of this study are (1) to determine the levels of total Se and Se species of the soils in Dashan Region, and (2) to reveal the relationships between total Se and Se species with soil general properties, respectively.

2. MATERIALS AND METHODS

2.1. Study site and sample collection

The soil samples (0–15 cm) were collected in a mountainous area of Anhui Province, China. The sampling region is located at the intersection of Shitai and Qimen counties (Fig. 1). It contains Dashan village, Shili village, Shuangkeng village and Xianyu village in Shitai County, and liuyuan village in Qimen County (117°19′54″~117°24′53″E, 30°00′00″~30°03′06″N). The sampling area is about 46.22 km². The annual average temperature in the sampling area is 16°C, the annual average rainfall is 1626.4 mm and the total average annual evaporation is 1256.2 mm (Wu et al. 2007b).

Figure 1 Study site (located in Chizhou, Anhui Province).

Sixty-nine soil samples (0–15 cm) were collected (Table 1), with 12 samples collected in farmland, 28 samples collected in forest area and 29 samples collected in tea gardens. All of the samples were zipped in polyethylene bags and transported to the laboratory. A portable global positioning system (GPS) device was used to locate the sampling sites. In the laboratory, soil samples were air dried, homogeneously ground and sieved through a 0.15-mm mesh before analysis.

Table 1 Number of soil samples for each soil type and land use

Soil type^†	Relevant soil type^‡	Land use	Number of samples
Yellow soil	Alisols, Acrisols, Cambisols	Woodland	16
		Tea garden	11
Yellow brown soil	Cambisols, Alisols	Woodland	10
		Tea garden	18
Red soil	Cambisols	Woodland	2
Paddy soil	Anthrosols	Farmland	12

†By classification of soil in China (Gong 1994).

‡By classification of soil from FAO–UNESCO (FAO-UNESCO 1990).

2.2. Measurement of soil properties

The following soil properties were detected for all soil samples: pH value (H₂O), content of organic matter (OM) and soil texture. The pH (H₂O) values were measured using a PHS-3C digital pH-meter (Shanghai Yueping Instrument Co., China). OM contents were determined titrimetrically by the oxidation method using potassium dichromate–sulfuric acid (Ahmed et al. 2008). Soil texture was measured with a Beckman Coulter Ls 230 laser granularity apparatus after OM was decomposed with 30% Hydrogen peroxide (H₂O₂) (Beckman Coulter, Inc., USA).

2.3. Digestion of soil sample

The measurement of total Se was accomplished by the use of previously established methods (Gao et al. 2011). Briefly, 0.5–3 g of sample was digested overnight in a 50-mL conical flask by 10 mL of concentrated acid mixture nitric acid (HNO₃) / perchloric acid (HClO₄)=4:1 volume/volume). The mixture was then heated at 100°C for 1 h, 120°C for 2 h and then 180°C for 1 h, using an electrical plate. The samples were then heated at 210°C until no white fume appeared. The remaining solution was cooled down to room temperature, and 5 mL of hydrochloric acid (HCl, 12 M) was added to convert Se⁶⁺ to Se⁴⁺, for about 4 h. Finally, the solution was adjusted to 25 mL with Milli-Q water for Se analysis.

2.4. Sequential extraction and determination of Se species

Subject to the method detection limits, the contents of bioavailable-Se cannot be detected in soils with low levels of total Se. Thus, to illuminate the species distribution of Se in the soils of Dashan Region, only 24 soil samples (containing all soil types) which had relatively high levels of Se (> 0.3 mg kg⁻¹) were picked up for the measurement of Se speciation using a sequential chemical-extraction method which has often been used to evaluate the geochemical behaviors of Se (Sharmasarkar and Vance 1995; Mao and Xing 1999; Zhang et al. 2002; Zhao et al. 2005; Zhu et al. 2008a; Wakim et al. 2010). The details of extraction were as follows. Ten milliliters of Milli-Q water was added to 1 g of soil. The suspension was shaken (200 cpm) for 1 h and then centrifuged (4000 rpm) for 30 min. The supernatant was separated from the residue for the determination of water-soluble selenium (W-Se). Next, 10 mL of potassium phosphate monobasic (KH₂PO₄)-Potassium hydrogen phosphate anhydrous (K₂HPO₄) (0.1 M, pH = 5) solution was added to the residue, and the suspension was shaken for 1 h to extract the exchangeable selenium (E-Se). Then, the residue was treated with 10 mL of HCl (3 M) at 90°C for 50 min with intermittent shaking to extract acid-soluble selenium (A-Se). Finally, the residue was treated with 10 mL of potassium persulfate (K₂S₂O₈) (0.1 M) at 90°C for 2 h with intermittent shaking to extract organic-bound selenium (O-Se). For all extractions, the extractant was separated by centrifugation in the same way as for the determination of W-Se. The material remaining was digested in a mixture of 15 mL concentrated nitric acid (HNO₃), 8 mL hydrofluoric acid (HF) and 2 mL HClO₄, heating to below 170°C for the determination of residue selenium (R-Se). All of the prepared samples were analyzed with hydride generation atomic fluorescence spectrometry (HG-AFS 9230, Beijing Titan Instrument Co., China). HG-AFS, which uses Sodium borohydride (NaBH₄) (or Potassium borohydride, KBH₄) as a reducing agent, will reduce Se⁴⁺ to Hydrogen selenide (H₂Se) in HCl solution. The irradiation of a hollow cathode lamp excites the electrons in atoms of Se and causes them to emit fluorescence. The strength of the fluorescence increases with the increasing content of Se. The instrumental parameters are listed in Table 2. The detection limit of the HG-AFS method for samples was 0.08 μg L⁻¹ solution (Gao et al. 2011; MoH-PRC 2012).

Table 2 Operating parameters of the hydride generation atomic fluorescence spectrometry (HG-AFS) instrument

Parameters	Settings
Negative high voltage of the photomultiplier (V)	300
Lamp current (mA)	100
Atomization temperature (°C)	800
Atomizer height (mm)	8
Flow rate of carrier gas (mL min^−1)	500
Flow rate of shield gas (mL min^−1)	1000
Read mode	Peak area
Measurement method	Standard curve
Delay time (S)	1
Read time (S)	15
Injection volume (mL)	2

2.5. Quality control

Confidence in the measurements was qualified by using blank controls, certified reference materials and duplicates (Bettinelli et al. 2000). National standard reference material GSS-1 (soil) was used to check the recoveries of Se. The recoveries of the standard reference material ranged from 85.5 to 118%, with the relative standard deviation (RSD) being 10.76%. The detection limit (DL) of the instrument was 0.08 μg L⁻¹ solution.

3. RESULTS AND DISCUSSION

3.1. Soil properties and concentration of total Se

The properties of the soil samples are presented in Table 3. The soil samples collected from Dashan Region were all clay loam. pH values of the soil samples ranged from 4.09 to 7.18. The OM contents ranged from 0.73 to 24.1 g kg⁻¹.

Table 3 Soil properties and concentration of total selenium (Se) in Dashan region.

	Total Se (mg kg^−1)		Clay (%)		pH (H2O)		OM (mg kg^−1)
	Mean	Range	Mean	Range	Mean	Range	Mean	Range
Dashan Region (n = 69)	1.50	0.31–7.65	19	12.9–29.8	5.02	4.09–7.18	7.24	0.73–24.1
Farmland (n = 12)	0.97	0.31–1.87	17.50	14.2–20.4	5.61	4.62–7.18	4.79	2.27–6.49
Woodland (n = 28)	1.63	0.32–7.65	18.96	12.9–29.8	4.92	4.09–5.71	8.30	0.73–24.13
Tea garden (n = 29)	1.58	0.36–5.54	19.58	15.4–29.2	4.88	4.13–6.48	7.26	2.62–14.73

OM, organic matter.

Se is a non-metallic element naturally present in the earth’s crust; the global average concentration of Se in soil is 0.4 mg kg⁻¹, ranging from 0.01 to 2 mg kg⁻¹ (Dungan and Frankenberger 1999). The concentrations of total Se in soils of Dashan Region ranged from 0.31 to 7.65 mg kg⁻¹. According to the classification of Tan (1996), 71% of soils collected from Dashan Region were enriched with Se, and 10, 12 and 7% of soil samples had moderate, high and excessive levels of Se, respectively.

Compared with previous measurements of total Se in soils of other areas of China, the average concentration of total Se in soils from Dashan Region was almost 2 times higher than that in Hong Kong (0.76 mg kg⁻¹; Zhang et al. 2005); 4 times higher than that in Guizhou Province (0.37 mg kg⁻¹; He 1996) and Ziyang of Shanxi Province (0.32 mg kg⁻¹; Zhao et al. 1993), which was named the second most Se-enriched region in China; 5 times higher than that in the mainland of China (0.29 mg kg⁻¹; Liu 1996); and more than 4 times lower than that in Enshi (6.33 mg kg⁻¹; Zhang and Ma 2008), China, which is the “World Capital of Selenium” (Table 4; Yuan et al. 2012). Compared with previous studies in other countries, the levels of total Se in soils of Dashan were also at the high end for the world; details can be found in Table 4 (Dhillon and Dhillon 1999; Maksimovic et al. 1992; Fordyce et al. 2000; Zhang et al. 2005; Zhang and Ma 2008; Yamada et al. 2009; Bajaj et al. 2011; Yuan et al. 2012). The high levels of Se found in the soils of Dashan Region might be caused by the weathering of carbon-siliceous sediment (Yuan et al. 2013). According to a study by Zhu and Zheng (2001), Se is present mainly in the carbon-siliceous sediment which is exposed in the environment of Dashan Region. The soils weathered from local rocks by the heavy dissection by streams, rains and air in Dashan Region inevitably retain their inherent properties (Li et al. 2008). Therefore, the soils in Dashan Region are rich in Se.

Table 4 Comparison of concentrations of selenium (Se) in soils of Dashan Region with previous data

Country	Regions	Mean (mg kg^−1)	Range (mg kg^−1)	Reference
China	Dashan Region (n = 69)	1.50	0.31–7.65	This study
	Farmland (n = 12)	0.97	0.31–1.87	This study
	Woodland (n = 28)	1.63	0.32–7.65	This study
	Tea garden (n = 29)	1.58	0.36–5.54	This study
	Enshi	6.33	–	(Zhang and Ma 2008)
	Hong Kong Special Administration Region	0.76	0.07–2.26	(Zhang et al. 2005)
	Guizhou Province	0.37	0.06–1.33	(He 1996)
	Ziyang	0.32	–	(Zhao et al. 1993)
	Mainland China	0.29	0.05–0.99	(Liu 1996)
New Zealand	Whole country	0.5	–	(Oldfield 1999)
Yugoslavia	Whole country	0.2	–	(Maksimovic et al. 1992)
Japan	Paddy	0.43	–	(Yamada et al. 2009)
	Upland	0.51	–
Sri Lanka	Low goiter incidence(<10%) village group	0.23	–	(Fordyce et al. 2000)
	Moderate goiter incidence(10–25%) village group	0.88
	High goiter incidence(>25%) village group	1.12
India	Punjab	1.19	–	(Dhillon and Dhillon 1999)
	Himachal Pradesh	0.4
	Andhra Pradesh	0.13
	Karnataka	0.16
Worldwide		0.40	0.10–2.00	(Zhang et al. 2005)

In the present study, soil samples were collected from three different types of land use, and the highest concentration of Se was found in a soil sample collected in woodland, while the lowest concentration was found in a sample from farmland (Table 4). According to our study, woodland had the highest content of Se (0.32–7.65 mg kg⁻¹), followed by tea garden (0.36–5.54 mg kg⁻¹) and farmland (0.31–1.87 mg kg⁻¹; Table 4). The concentrations of Se in soils in woodland were similar to those of tea gardens, but were about 1.7 times higher than those of farmland. Our finding was similar to Zhang et al. (2005), who found that the average Se concentration in woodland (1.36 mg kg⁻¹) was 4 times higher than that in farmland (0.36 mg kg⁻¹) of Hong Kong. The difference of Se content among the three different land use types might be caused by different soil properties and anthropogenic activities. The OM and clay contents of woodland and tea gardens were higher than those of farmland, while the pH values of woodland and tea gardens were lower than those of farmland. According to the research of Zhang et al. (2005), the contents of Se were positively correlated with contents of OM and clay, but negatively correlated with pH values. This may be one of the reasons for the different concentrations of the three different land-use types. On the other hand, compared with the farmland, which was mostly paddy field, the intensity of anthropogenic activity in the tea garden and woodland was relatively weak. Water-flooding management of farmland in Dashan Region could induce a decrease of total Se content in the surface soils because Se in surface soils can migrate downward in the soil profile of paddy fields, either as inorganic Se or as Se compounds in fulvic acids (Kang et al. 1991).

3.2. Different speciation of Se in Dashan region

The bioavailability of different species of Se provides important information for understanding the geochemical behavior of Se. Generally, the W-Se and E-Se are considered, as the B-Se, A-Se and O-Se are regarded as the transferable Se, and the R-Se is regarded as the un-bioavailable Se. Overall, the B-Se, consisting mainly of selenate (SeO₄^2–) and selenite (SeO₃^2–) in soil, is a key factor for Se accumulation in plant (Ellis and Salt 2003). The transferable Se in soil provided a potential Se source for plant utilization (Zhao et al. 2005; Zhu et al. 2008a). The residual Se, which means the Se present in silicate, primary and secondary minerals, cannot be released into the environment easily or extracted by normal methods with extractants such as KH₂PO₄–K₂HPO₄, HCl, etc.; generally, it was extracted by a mixture of HNO₃, HF and HClO₄ (Wang et al. 2013). Therefore, the residual Se cannot be utilized by plants directly (Wu et al. 2004).

The contents of B-Se in soils of Dashan Region ranged from 32.41 to 120.1 μg kg⁻¹, with an average value of 58.45 μg kg⁻¹ (Table 5). The average content of B-Se in soils of Dashan Region was about 5 times higher than that found in Ziyang (ranging from 2.21 to 56.38 μg kg⁻¹, with an average value of 12.35 μg kg⁻¹; Zhao et al. 1993), but was more than 70 times lower than that found in Enshi (which ranged from 960 to 22,260 μg kg⁻¹, with an average value of 4234 μg kg⁻¹; Yuan et al. 2012). Compared with other regions of the world, the contents of B-Se in soils of Dashan Region were lower than those found in soils of England (ranging from 245 to 590 μg kg⁻¹; Stroud et al. 2010), but high than those found in soils of Kyoto prefecture, Japan (ranging from 5.5 to 40.2 μg kg⁻¹, with an average value of 12.35 μg kg⁻¹; Yamada et al. 2012). The proportions of B-Se in soils of Dashan Region were 0.82~5.95%, which were higher than those in Ziyang (0.1–0.9%; Zhao et al. 1993), but comparable to those in Enshi (< 5%; Yuan et al. 2012) and in other farmland of China (1.1–6.7%; Tan et al. 2002). The proportions of B-Se were also similar to those in England (1.1–3.4%; Stroud et al. 2010), in Benton Lake (2.6–7.3%; Zhang and Moore 1996) and Fresno County (3.3%) of the USA (Martens and Suarez 1997), and in Kyoto prefecture of Japan (1.7–11.6%; Yamada et al. 2012), but were much lower than those in Punjab, India, where the proportions of B-Se in soils were close to 50% (Bajaj et al. 2011). The difference of the proportions of B-Se in soils may be due to the different soil properties and different sources of Se in soil (Wang et al. 2012).

Table 5 Content and proportion of selenium (Se) species in soils in Dashan region compared with previous data

Country	Regions	Se speciation	Content (μg kg^−1)	Proportion (%)	Reference
China	Dashan Region (n = 24)	W-Se	7.71–26.8	0.28–1.45	This study
		E-Se	24.7–93.3	0.54–4.50	This study
		A-Se	112–1430	5.45–24.4	This study
		O-Se	26–765	1.01–14.6	This study
		R-Se	1300–5330	62.6–88.8	This study
		B-Se	32.41–120.1	0.82–5.95	This study
	Ziyang	B-Se	2.21–56.38	0.1–0.9	(Zhao et al. 1993)
	Enshi	B-Se	960–22260	< 5	(Yuan et al. 2012)
	Farmland in China	B-Se	–	1.1–6.7	(Tan et al. 2002)
England	Rothamsted	B-Se	–	1.4	(Tolu et al. 2011)
	Whole country	B-Se	245–590	1.1–3.4	(Stroud et al. 2010)
USA	Benton lake	B-Se	–	2.6–7.3	(Zhang and Moore 1996)
	Fresno County	B-Se	–	3.3	(Martens and Suarez 1997)
Japan	Kyoto prefecture	B-Se	5.5–40.2	5.12	(Yamada et al. 2012)
India	Punjab	B-Se	–	50	(Bajaj et al. 2011)

Se, Selenium; W-Se, water-soluble Se; E-Se, exchangeable Se; A-Se, acid-soluble Se; O-Se, organic-bound Se; R-Se, residue Se, B-Se, bioavailable S.

3.3. Correlations between Se contents and soil properties

There was a significant positive correlation between levels of total Se and soil OM contents (r = 0.615, p < 0.01; Table 6). However, significant negative correlations were found between OM contents and levels of W-Se, E-Se and B-Se in the soils (r = −0.621, r = −0.552, r = −0.583 respectively; p < 0.05). A previous study indicated that OM in soil played an important role in the accumulation of Se; it would act as a pool to increase the absorption of soil to selenate and selenite (Gustafsson et al. 1993). Therefore, the soils with high OM contents presented high levels of total Se, while the levels of B-Se were relatively low in Dashan Region.

Table 6 Correlation among total selenium (Se), different species of Se, pH value, organic matter (OM) and clay content in soils

	Total Se	W-Se	E-Se	A-Se	O-Se	R-Se	B-Se	pH	OM	Clay
Total Se	1
W-Se	0.198	1
E-Se	0.292	0.845**	1.
A-Se	0.852**	0.339	0.477*	1
O-Se	0.783**	0.481*	0.633**	0.901**	1
R-Se	0.978**	0.092	0.158	0.729**	0.650**	1
B-Se	0.280	0.901**	0.993**	0.460*	0.618**	0.148	1
pH	−0.512**	0.566**	0.707**	0.147	0.280	−0.196	0.695**	1
OM	0.615**	−0.621**	−0.552*	−0.162	−0.253	0.260	−0.583**	−0.623**	1
Clay	0.539**	0.220	0.135	0.611**	0.573**	0.587**	0.160	−0.222	0.068	1

*Significant at the 0.05 level. **significant at the 0.01 level.

W-Se, water-soluble Se; E-Se, exchangeable Se; A-Se, acid-soluble Se; O-Se, organic-bound Se; R-Se, residue Se, B-Se, bioavailable S.

There was a negative correlation between concentrations of total Se and pH values of soil (r = −0.512, p < 0.01), while significant positive correlations were found between pH values and levels of W-Se, E-Se and B-Se (r = 0.566, r = 0.707, r = 0.695 respectively; p < 0.01). In the soil with high air permeability, the conversion of Se between SeO₃^2– and SeO₄^2– was mostly affected by the pH value of soil. In the acidic soil, Se was mainly in the form of SeO₃^2–, which had lower solubility than SeO₄^2– and was difficult for plants to absorb or utilize. With an increase of the soil pH value, SeO₃^2– in the soil would be easily oxidized to SeO₄^2–, which could be easily dissolved in water and utilized by plants (Qu et al. 1998; Li and Wang 2002).

Furthermore, a significant positive correlation was also found between levels of total Se and clay contents of soil (r = 0.539, p < 0.01), but no significant correlations were found between clay contents of soil and levels of W-Se, E-Se and B-Se. In the study of Se in soils of Hong Kong, a significant positive correlation was also found between levels of total Se and clay contents (r = 0.570, p < 0.01), which suggested that soils with high contents of clay had a high ability for Se accumulation (Zhang et al. 2005).

In conclusion, the levels and environmental fates of Se in Dashan Region were influenced by the physicochemical properties of soil, which were related to the climate and intensity of anthropogenic activities of this region. Firstly, the parent soil material seemed to be the key to determine the contents of total Se in Dashan. The high level of total Se in Dashan Region was mainly attributed to the carbon-siliceous sediment, which contains high levels of Se. Secondly, during the process of soil formation in Dashan Region, the effect of the parent soil material on the content of Se degraded gradually, while the other factors, including climate, anthropogenic activities and different land-use types, would play a more important role in the content of total Se and different species of Se. Dashan region has a subtropical rainy and humid monsoon climate, which accelerates the weathering of the parent soil material including carbon-siliceous sediment. The minerals existing in the carbon-siliceous sediment which contain a high concentration of Se may break down; then Se would enter into the environment and enhance the content of B-Se in the soil (Wu et al. 2007b). Furthermore, the different intensity of human activities in Dashan Region is another factor which might impact the content of Se in soil of this region. The contents of OM, which acted as a pool of SeO₄^2– and SeO₃^2–, were high in the land where the anthropogenic activities were weak (Wu et al. 2007a); thus, the concentrations of total Se in farmland were lower than those of different land-use types. On the other hand, the relatively high temperature (annual average temperature is 16°C) and intense rainfall (annual average rainfall is 1626.4 mm) made the base cations leach more easily, thus increasing the soil acidity of Dashan Region (Chen 2010). As mentioned above, high soil acidity would inhibit the oxidization of SeO₃^2–. Hence, the B-Se was low in the soil of Dashan Region, while the total concentration of Se was high.

ACKNOWLEDGMENTS

This study was supported financially by the National Natural Science Foundation of China (Project Nos. 41203075 and 41201186), the Natural Science Foundation of Anhui Province (Project No. 1308085QB31) and the Science and Technology Research Foundation of Anhui Province (Project No. 12010402133).