Hierarchical cluster analysis of arsenic and fluoride enrichments in groundwater from the Datong basin, Northern China

Hierarchical cluster analysis of arsenic and fluoride enrichments in groundwater from the Datong basin, Northern China

Journal of Geochemical Exploration 118 (2012) 77–89 Contents lists available at SciVerse ScienceDirect Journal of Geochemical Exploration journal ho...

2MB Sizes 19 Downloads 62 Views

Journal of Geochemical Exploration 118 (2012) 77–89

Contents lists available at SciVerse ScienceDirect

Journal of Geochemical Exploration journal homepage: www.elsevier.com/locate/jgeoexp

Hierarchical cluster analysis of arsenic and fluoride enrichments in groundwater from the Datong basin, Northern China Junxia Li, Yanxin Wang ⁎, Xianjun Xie, Chunli Su School of Environmental Studies & State Key Laboratory of Biogeology and Environmental Geology, China University of Geosciences, 430074 Wuhan, China

a r t i c l e

i n f o

Article history: Received 21 August 2011 Accepted 3 May 2012 Available online 12 May 2012 Keywords: Hierarchical cluster analysis Fluoride Arsenic Groundwater Datong basin

a b s t r a c t To better understand the occurrence of high F and As in groundwater of the Datong basin, a total of 486 groundwater samples were collected for hierarchical cluster analysis (HCA) of eighteen hydrochemical parameters. Groundwater samples can be divided into thirty-six and nineteen groups for shallow and deep groundwater, respectively. Results show that high F samples in shallow groundwater contain F as high as 22 mg/L and mainly occur in the discharge area in the basin center, and the highest F concentration of deep groundwater samples is 8.3 mg/L which mainly occur in the western mountain front area. The groundwater with elevated HCO3 concentration favors F enrichment in the Datong basin. Nearly all of the high F samples are oversaturated with respect to calcite and undersaturated with respect to fluorite, indicating that fluorite solubility is a limit for F enrichment. Besides, evapotranspiration has a stronger effect on fluoride enrichment, especially for the shallow groundwater. For the both F and As enrichment samples in deep groundwater, the desorption of Fe-(hydr)oxides is suggested to be the major mechanism. High As samples of shallow and deep groundwater mainly occur between Senggan River and Huangshui River. The highest arsenic concentration reaches up to 469 μg/L, and all samples of high arsenic groundwater have low concentrations of NO3 and SO4, indicating the prevailing reducing conditions in the aquifer system at Datong. The reductive dissolution of Fe-(hydr)oxides driven by sulfate reduction and biodegradation of organic matters is postulated to be the major process controlling arsenic enrichment in groundwater. © 2012 Elsevier B.V. All rights reserved.

1. Introduction High arsenic (As) and high fluoride (F) groundwater have been widely documented all over the world in recent decades, including As-affected areas in Inner Mongolia of China (Deng et al., 2009), Bangladesh (Nickson, et al., 1998) India (Farooq, et al., 2010), Mexico (Roy, et al., 2004), Hungary (Rowland, et al., 2011), Vietnam (Berg et al., 2007), Argentina (Bhattacharya, et al., 2006) and Chile (Romero, et al., 2003) and F-affected areas in Yuncheng and Taiyuan of China (Currell et al., 2011; Guo et al., 2007), Ghana (Apambire et al., 1997), India (Jacks et al., 2005), Jordan (Rukah and Alsokhny, 2004), Mexico (Grimaldo et al., 1995), East Africa (Gaciri and Davies, 1993), Gaza Strip (Shomar et al., 2004), and Spain, Iran, Algeria, Australia, Egypt, and South Korea (Chae et al., 2007). In the arid/semi-arid regions of northern China, groundwater is the major source for domestic and agricultural water supply. With the fast social and economic development, demand for groundwater has been increasing in recent years. Unfortunately, naturally occurring high As and high F groundwater have threatened the health of local residents in Datong (Guo and Wang, 2004; Li, 2001; Su, 2006).

⁎ Corresponding author. Tel.: + 86 27 67883998; fax: + 86 27 87481030. E-mail address: [email protected] (Y. Wang). 0375-6742/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.gexplo.2012.05.002

In the Datong basin, As and F contents in groundwater frequently exceed the maximum permissible limit concentration of 10 μg/L and 1.5 mg/L, respectively as recommended by WHO and the Chinese government (Li, 2001; Su, 2006). The recorded highest As and F concentrations were up to 1820 μg/L and 39 mg/L, respectively (Su, 2006). Due to long term intakes of high As and F contaminated groundwater, a large population has suffered from arsenocosis and fluorosis (Wang et al., 2009). The As-affected areas cover 63 villages with more than 5000 confirmed arsenocosis patients, which make up 12.1% of exposed population. In the most serious As-affected Daying and Shuangzhai villages, the results of preliminary survey carried out by the local disease control agency indicate that the mortalities were up to 48.6% and 29.5%, respectively in all surveyed samples from these two villages (Su, 2006). The major symptoms of As poisoning are keratosis, skinpigmentation problems, and in some serious cases cancers of the lung, skin and bladder (Su, 2006; Wang et al., 1998). Fluoride poisoning causes dental fluorosis, and even loss of mobility (Su, 2006). In the Datong basin, it has been reported that more than 170,000 people were drinking high F groundwater with 82,000 dental fluorosis patients and 8500 skeletal fluorosis patients, respectively (Su, 2006). Since the As poisoning cases were reported at Datong in 1990, many studies have been done to study the geological, hydrological, geochemical, hydrogeochemical, biogeochemical and mineralogical factors of high As groundwater in the Datong basin (Guo and Wang, 2004,

78

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

Fig. 1. a: Location of the study area and sampling sites; b: hydrogeological section along the I–II lines in A (Han, 2008).

2005; Xie et al., 2008, 2009a, 2009b, 2011). High As groundwater mainly occurred in lacustrine aquifers containing high organic matter (1.0% organic carbon) and was characterized by high pH and low SO4

concentration (Guo and Wang, 2005; Wang et al., 2009). Iron oxides and hydroxides were thought to be the natural sources, and reducing conditions were also believed to be associated with As enrichment in

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

groundwater of the Datong basin (Xie et al., 2009a,b). However, the precise mechanisms and processes of As release and mobilization in As-affected aquifers are still uncertain. In comparison to As, less work has been done on hydrogeochemistry of high F groundwater in the Datong basin. Although the sources of F in groundwater have been well understood in many parts of the world, more detailed work is still needed to characterize the complex processes controlling its migration from F-bearing aquifer matrix into groundwater (Jayawardana, et al., 2012; Msonda, et al., 2007; Zhao, et al., 2007). As a useful tool, the multivariate statistical techniques have been widely used to analyze hydrogeochemical data (Cloutier et al., 2008; Güler et al., 2002; Singh et al., 2004, 2005; Yidana, 2010; Yidana et al., 2010). Multivariate statistical methods cannot only help to understand the cause-and-effect relationships, but also provide information about hydrochemical processes (Yidana et al., 2010). Therefore, in this study, hierarchical cluster analysis (HCA) has been used in combination with conventional hydrochemical graphs to study the main geochemical processes resulting in As and F enrichments in groundwater of the Datong basin. The main objectives of this study are to 1) delineate the spatial distribution of As and F enrichments in the groundwater from the Datong basin; 2) assess biogeochemical indicators and potential mechanisms of high As in the groundwater; 3) evaluate hydrogeochemical factors and processes controlling on high F in the groundwater; and 4)analyze the links between the high As groundwater and the high F groundwater. 2. The study area Datong basin is located in the northern part of Shanxi Province, China. This semi-arid area has an annual average precipitation of about 225–400 mm and evaporation above 2000 mm, with 75%–85% of rainfall occurring in July and August. The annual air temperature ranges between −14.9 and 22.8 °C with an average temperature of 6.5 °C. Ephemeral Senggan River and its tributaries are the main surface water bodies. In this area, groundwater generally flows from northwest to southeast in the central basin and from the mountain front to the central parts of the basin. Bedrock around the Datong basin mainly includes Archean gneiss and Quaternary basalt in the north, Cambrian–Ordovician limestone and Carboniferous–Permian–Jurassic sandstone and shale in the west (Guo et al., 2003). Archean gneiss and granite sporadically outcrop in the northwest. The thickness of the Quaternary sediments increases

79

from 200 m at the margin to 2700 m in the central part of the basin, and the grain sizes of the sediments generally decrease. The sediments from the central parts are mainly sandy loam and silt, lacustrine and alluvial–lacustrine sandy loam, silty clay and clay with high contents of organic matters, while those at the basin margin are mostly alluvial– pluvial gravel and sand (Fig. 1). The Quaternary groundwater system can be basically divided into 3 groups, upper, middle and lower aquifers, with the corresponding depths of 5–50 m, 50–150 m and more than 150 m (Guo and Wang, 2004). Groundwater is recharged mainly by vertically infiltrating meteoric water in the basin and laterally penetrating groundwater in bedrock fractures along the basin margin. In addition, the seepage of non-perennial river water and irrigation return flow is recharge sources of groundwater in the basin. Evapotranspiration and artificial abstraction are the major ways of shallow groundwater discharge (Wang et al., 2009).

3. Methodology 3.1. Sampling and analytical methods Four hundred and eighty-six water samples including 392 shallow groundwater samples (b50 m) and 94 deep groundwater samples (≥50 m) were taken from the Datong basin during August 2009 (Fig. 1). The samples were collected in three 500 mL pre-cleaned polythene bottles. Temperature, pH and electrical conductivity (EC) were measured in the field, and the total alkalinity was determined using titration method within 24 h after sampling. Water samples were filtered through 0.45 μm PVDF membranes (Millipore). Samples for cation and trace metallic element analyses were acidified using ultra-purified HNO3 to pH b 2. Samples for anion analysis were stored in bottles without acidification after filtering. Major cations and some trace elements such as Fe, Pb, Ba, Sr, and Mn were determined using ICP-AES (IRIS Intrepid II XSP). Anions including F, Cl, NO3 and SO4 were analyzed using IC (Metrohm 761 Compact). As concentration was analyzed using AFS-2202 with a dual-channel atomic fluorescence spectrophotometer. The analytical reproducibility of major element concentrations varies between 95 and 105%. The average errors of ICP-AES for trace elements determined are less than 5%. The average error for As analysis is less than 10% by AFS-2202. All analyses were accomplished at the State Key Laboratory of Biogeology and Environmental Geology of China University of Geoscience in Wuhan.

Fig. 2. Dendrogram for all the shallow groundwater samples.

80

Table 1 Descriptive statistics of the different clusters with samples of more than four of shallow groundwater in the study area (in mg/L except pH, As in μg/L, and Ec in μs/cm). Chemical type

195

Various

S2

13

Na–Mg–HCO3–Cl, Na–Mg–HCO3, Na–HCO3, Mg–Ca–Na–HCO3

S3

23

Na–HCO3–Cl, Na–Mg–HCO3–Cl, Na–HCO3–Cl–SO4

S4

5

Na–Cl–HCO3, Na–Cl–HCO3–SO4

S5

15

S7 S9

4 7

S11

16

Na–HCO3, Na–HCO3–Cl, Na–Mg–HCO3

S12

17

Na–HCO3–Cl, Na–HCO3, Na–HCO3–Cl–SO4

S14

12

Na–Cl–HCO3, Na–Cl, Na–Mg–Cl

Na–Mg–HCO3, Na–HCO3, Na–HCO3–Cl, Mg–Ca–HCO3 Na–Mg–Ca–HCO3, Ca–HCO3 Na–Mg–HCO3–Cl, Ca–Mg–HCO3, Na–HCO3

S15

5

Ca–Mg–SO4–HCO3–Cl, Ca–Na–Mg–SO4–Cl, Mg–Na–Ca–HCO3–SO4⁎

S16

6

Na–Mg–HCO3, Na–HCO3–Cl, Ca–HCO3

S19

20

Na–Cl, Na–Mg–Cl, Na–Cl–HCO3

⁎ The feature value of the cluster.

Mean Max Min Mean Max Min Mean Max Min Mean Max Min Mean Max Min Mean Mean Max Min Mean Max Min Mean Max Min Mean Max Min Mean Max Min Mean Max Min Mean Max Min

pH

Ec

As

F

7.87 8.70 7.07 7.82 8.03 7.60 7.66 7.98 7.29 8.27 8.41 8.07 7.81 8.16 7.44 7.71 7.79 8.46 7.41 8.11 8.39 7.86 7.93 8.41 7.31 7.69 7.90 7.38 7.33 7.50 7.12 8.03 8.39 7.67 7.47 7.84 7.09

1160 4740 340 1320 2210 592 2820 3910 1910 3290 4100 2070 1060 2410 504 806 1160 2250 497 1140 2950 345 4500 8240 1230 7220 9030 5590 2590 2960 2310 1040 1800 375 11,300 22,600 6660

5.4 102 b 0.1 1.8 4.2 0.2 2.8 12.7 0.2 34.1 114 0.2 9.5 95 0.6 0.9 1.8 3.9 b 0.1 286⁎ 470⁎ 165⁎

1.36 4.21 b0.01 1.69 4.51 0.67 2.19⁎ 6.04⁎

640 2350 205 757 1550 297 1630⁎ 2300⁎

0.78 2.59⁎ 3.42⁎ 1.67⁎ 1.65 4.40 b0.01 1.23 1.57 3.96 b0.01 1.03 2.97 b0.01 5.80⁎ 10.4⁎ 2.10⁎ 1.36⁎

1060 1930 2640 1150 594 1120 381 419 703 1340 343 567 1570 238 2730⁎ 4730⁎

9.2 59.3 b 0.1 25.6⁎ 122 0.4 0.8 0.8 0.8 1.7 2.7 0.9 6.5 90.7 0.2

3.47 b0.01 0.61 0.89 0.41 2.19 5.37 0.16 1.51 6.06 b0.01

TDS

741 3710⁎ 5030⁎ 2880⁎ 1840 2090 1480 571 911 215 6800⁎ 10,700⁎ 3390⁎

Cl

NO3

SO4

HCO3

K

Na

Ca

Mg

Fe

Ba

Mn

Pb

78.0 580 5.31 103 325 10.8 260 464 142 459 638 142 74.8 336 18.4 32.5 94.1 234 10.8 90.3 351 10.8 469 1030 42.5 1050⁎ 1590 492 171 218 135 65.0 124 6.12 2120⁎

29.8 204 b0.01 57.3 148 10.5 213 566 b0.01 46.8 213 b0.01 27.6 144 b0.01 51.2 53.2 182 b0.01 0.80 3.58 b0.01 140 498 b0.01 45.8 207 b0.01 148. 328 21.4 30.9 98.5 3.90 372 1120⁎

123 884 b 0.01 146 385 23.5 342 716 78.1 478 711 b 0.01 79.6 237 21.7 48.0 107 222 41.2 33.0 352 b 0.01 597 1510 65.8 1110 1260 922 832 1140 5490 118 267 14.1 1800 2770 598

396 1130 159 420 644 274 628 959 379 608 973 462 425 731 199 316 461 1120 243 483 886 283 1150 1790 590 741 1080 568 397 806 218 348 446 216 779 1540 497

3.38 39.7 b0.01 3.04 16.5 0.44 6.61 17.7 2.75 3.90 7.97 1.64 2.19 5.10 b0.01 1.68 3.47 8.69 0.40 1.56 3.70 b0.01 12.7 99.4 0.34 5.41 9.33 3.32 4.04 6.22 1.17 1.76 3.17 b0.01 34.0 212 1.68

115 657 6.36 115 377 24.1 312 529 105 541 661 274 110 381 17.1 35.4 104 352 7.61 168 440 41.1 771 1480 213 834 1310 633 100 129 57.1 101 190 6.88 1480 2900 581

46.8 137 8.75 50.5 98.5 29.6 80.9 146 39.0 24.8 46.1 4.79 48.5 158 13.1 57.0 49.8 71.0 34.1 16.0 27.9 4.57 31.3 84.9 5.75 70.7 109 47.8 252 339 181 35.8 57.2 14.0 202 380 32.1

42.5 172 8.29 54.5 106 23.3 106 182 51.2 67.8 128 41.7 41.0 95 20.8 35.5 60.9 124 25.0 29.1 70.4 9.62 90.1 258 17.8 222 302 123 134 184 115 44.4 77.6 8.98 405 773 128

0.021 0.275 b 0.001 0.042 0.110 b 0.001 0.021 0.125 b 0.001 0.010 0.039 b 0.001 0.042 0.148 b 0.001 0.007 0.029 0.083 b 0.001 0.007 0.059 b 0.001 0.034 0.152 b 0.001 0.021 0.053 b 0.001 0.026 0.071 0.003 0.234 0.714 b 0.001 0.032 0.145 b 0.001

0.637 1.84 0.068 1.79⁎ 2.57⁎ 1.21⁎

0.028 0.339 b0.001 0.010 0.088 b0.001 0.009 0.095 b0.001 0.0259 0.119 0.002 0.386 0.587 0.188 0.003 0.036 0.121 0.004 0.063 0.297 b0.001 0.086 0.546 b0.001 0.039 0.150 b0.001 0.032 0.151 0.001 0.760 1.018 0.591 0.064 0.400 b0.001

0.432 0.814 0.046 0.581 0.803 0.385 0.653 0.806 0.337 0.417 0.502 0.341 0.389 0.644 0.241 1.017⁎ 0.389 0.562 0.266 0.389 0.544 0.303 0.411 0.602 0.236 0.408 0.509 0.325 0.647 0.812 0.332 0.499 0.645 0.313 0.361 0.604 0.024

3270 1070

b0.01

0.921 1.84 0.331 1.40 2.35 0.984 0.452 0.907 0.147 0.543 0.437 0.755 0.226 0.641 1.30 0.064 0.580 1.17 0.256 0.642 0.967 0.304 0.585 0.948 0.183 0.944 1.97 0.235 0.614 1.55 0.063

Sr 1.04 3.73 0.130 2.15 4.04 0.608 0.973 1.60 0.431 1.47 1.98 0.822 3.41⁎ 6.84⁎ 0.555 1.02 6.73⁎ 10.0⁎ 3.65⁎ 0.709 1.50 0.388 1.05 3.96 0.210 1.01 1.68 0.499 1.94 4.67 0.668 3.18⁎ 6.50⁎ 0.787 2.36 14.2 0.168

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

No. S1

1.18 1.10 1.41

1.36 1.61 0.417 1.56 0.261 0.536 0.800 1.40 1.07 13.7⁎ 0.396 7.73⁎

0.252 0.388 0.273 0.409 0.762 0.668 0.503 0.275 0.288 0.319 0.488 0.357 0.305 0.310 0.568

In this study, squared Euclidean distances are used to divide parameters into initial clusters, while the between-groups linkage method is used to link the resulting initial clusters (Cloutier et al., 2008; Yidana, 2010; Yidana et al., 2010). Meanwhile, hierarchical cluster analysis (HCA) has been accomplished using the SPSS 17.0 code. For components with the concentration below the detection limit, the concentration was denoted with the detection limit value before the multivariate statistical analysis. In order to achieve the objectives of normal distribution and homogeneity, the data were standardized to their corresponding Z scores, which are obtained by subtracting the mean value of the normal distribution from each data and dividing by the standard deviation of the distribution, Zi = (Xi − mean) / s (Davis, 1986). For all multivariate statistical analyses, data standardization is the necessary step, because the calculation of Euclidean distance will be affected severely by the parameters with the highest variances (Cloutier et al., 2008; Güler et al., 2002).

0.144 0.481 0.668 1.16

0.262 0.395 0.579 0.793 0.700 0.738 1.15 0.260 0.205 1.49 4.16⁎

b 0.001 0.022 b 0.001 0.013 0.047 0.064 b 0.001 b 0.001 b 0.001 0.018 0.030 0.370 0.003 b 0.001 3.82⁎ 2.48

69.1 120 74.2 56.0 34.6 29.0 50.0 130 101 31.0 365 691 152 111 21.1 70.1 54.8 32.1 78.9 16.3 20.1 22.5 56.4 30.9 39.6 88.9 218 79.7 44.8 38.7 190 332 428 40.6 362 64.1 390 764 754 127 651 1850 376 758 20.2

4.1. Hierarchical cluster analysis (HCA)

5.41 1.15

1270 353 1480 2570⁎ 2530⁎ 583 3990 9590 1950 3410 250

0.01 1.41 3.05 0.23 5.58 0.57 1.84 0.86 2.00 0.53 3.28 1.88 22.0⁎

1200 1690 1810 651⁎

276 304 509 120 131 47.0 289 342 320 72.4 1320 1860 355 585 11.8

141 12.6 0.01 77.4 b 0.01 b 0.01 4.66 0.73 0.92 1.48 168 129 241 572 20.6

201 383 355 95.8 413 86.9 550 1010 692 136 1180 4460 418 494 17.4

319 795 706 358 614 191 327 530 1240 347 440 738 644 1040 235

92.9 88.1 61.4 2.32 1.42 10.4 5.81 3.68 3.91 2.25 3.14 6.21 3.40 327⁎

4. Results and discussion

2420 786 2400 412⁎ 356⁎ 895 6940 14,400 6630 4640 500

33.7 5.2 3.4 0.8 5.1 3.4 30.8 1.2 42.4 4.1 5.0 1.0 0.1 15.8 1.0 7.77 7.63 8.06 7.64 7.47 7.41 7.64 8.00 S17 S18 S20 S21 S22 S23

S13

S8 S10

⁎ The feature value of the cluster.

2110 3050 3310 12,000⁎

7.86 7.91 8.06 7.43 8.47⁎ 9.07⁎ 8.92⁎

Na–HCO3–Cl Na–HCO3–Cl Na–Cl–HCO3 Mg–Ca–Na–HCO3–Cl Na–HCO3 Na–Mg–HCO3 Na–Cl–SO4–HCO3 Na–SO4–Cl–HCO3 Na–HCO3 Na–HCO3 Na–Mg–Cl Na–Cl–SO4 Na–Mg–HCO3–Cl Na–HCO3–Cl Ca–Na–Mg–HCO3 DT-133 DT-423 DT-085 DT-339 DT-383 DT-284 DT-251 DT-147 DT-140 DT-265 DT-484 DT-357 DT-215 DT-195 DT-244 S6

81

3.2. Data preparation for the multivariate statistical analysis

0.010 0.004 b 0.001 0.001 0.003 0.003 0.003 0.001 0.140 0.355 0.003 0.327 0.011 0.005 0.310

Ba Fe Mg Ca Na K HCO3 SO4 NO3 Cl TDS F As Ec pH Chemical type ID

Table 2 Descriptive statistics of the different clusters with samples of less than four of shallow groundwater in the study area (in mg/L except pH, As in μg/L, and Ec in μs/cm).

Mn

Pb

Sr

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

4.1.1. Shallow groundwater A total of 353 shallow groundwater samples are analyzed for 18 hydrochemical parameters. The groundwater samples are classified following the procedure: (1) samples with the larger similarity are grouped at first; (2) groups of samples are joined using a linkage rule of between-groups linkage method as mentioned above; (3) repeat the first and second steps until all observations are classified. The results of the HCA performed on the 353 shallow groundwater samples are shown in Fig. 2. Fewer or greater number of clusters could be defined by moving the position of the phenon line up or down in the dendrogram (Güler et al., 2002). In this study, the cluster number is not confirmed directly by phenon line. The water samples can be basically divided into two groups: one group with TDS greater than 3000 mg/L (high TDS group) and the other with TDS less than 3000 mg/L (low TDS group) (Table 1). In the high TDS group, the phenon line is drawn across a scaled linkage distance of six, implying that all of the samples under this line are recognized as the same cluster. According to this method, samples in the high TDS group can be divided into five clusters (S19 to S23). But notably, the four clusters from S20 to S23 just have four samples (DT-357, DT-215, DT-195 and DT-244, respectively). These samples have abnormally high concentrations of Sr, F, K and Fe, respectively (Table 2), which could be related to the addition of domestic sewage and industrial pollution (Guo and Wang, 2004). Most of the samples in S19 occur in the center of the basin which belongs to the discharge area (Fig. 4). The samples in this cluster usually have high TDS values indicating the effects of strong evapotranspiration and water–sediment interactions. In addition, several samples in this cluster contain high NO3 concentration as high as 1120 mg/L, suggesting that agricultural activities have a great impact on groundwater quality in this area. Our recent study indicated that irrigation practice can cause surface water vertical recharge groundwater in this area (Xie et al., 2012). Combined with irrigation practice, the utilization of NO3-bearing fertilizers can contribute to the high NO3 concentration in groundwater, especially in the central part of the basin. In the low TDS group, the phenon line is defined as a distance of three, and produces a total of 18 clusters, naming S1 to S18, respectively. S1 is the biggest cluster containing 55% of samples which will be discussed in the next paragraph. S2 and S18 members have the same characteristics with high Ba concentrations ranging from 1.21 to 4.16 mg/L, which are classified according to the different levels of the Ba concentration. The same reason can also be applied to the S5, S9, S16 and S17 with high Sr concentrations ranging from 0.55 to 13.67 mg/L. S7 presents the samples

82

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

Fig. 3. Dendrogram for the S1, the biggest cluster of shallow groundwater samples.

with high Pb concentrations (mean: 1.02 mg/L). S8 and S13 aggregate three samples with inconsistent value between TDS and EC, which would be eliminated in the following analysis. S10 is discriminated due to abnormal pH value approaching or exceeding the WHO recommended standard values for taste purposes. These abnormal samples may have been impacted by anthropogenic activities. The complex hydrochemical type makes S15 special, and the samples mainly occur in the area between Datong County and Huairen County where there are large scale coal mining activities. The major ions of the samples from this area are dominated by SO4, HCO3, Ca and Mg, and the mean pH value is 7.33. Coal mining drainage could contribute sulfate to groundwater and decrease pH (Guo and Wang, 2005). On the other hand, these samples mainly occurring in the front of the mountain contain Ca and Mg as the major cations. Exceptionally high As and F concentrations are found in the samples of S14. And S3, S4, S12 are composed of high F groundwater samples and S11 consists of high As groundwater samples. In order to eliminate the effect of the abnormal data and plot the spatial variation effectively, the dendrogam for the cluster S1 is drawn (Fig. 3). The phenon line of it is defined across a distance of two to create eight clusters naming S1-1 to S1-8 to label high F samples with lower TDS and few high As samples (Table 3). It can be seen that high As groundwater clusters mainly include S11, S1-6 and S1-4 (Tables 1 and 3). All of high As groundwater samples are distributed at the southeastern bank of the Senggan River. The clusters reflecting high F samples are S3, S4, S12, S21, S1-2, S1-7, S1-8 with 92 samples, which mainly occur along the Senggan River and Huangshui River (Fig. 4). It is interesting that 84% of S1-2 samples have TDS lower than the

recommended value (1000 mg/L) of WHO for drinking water, and the concentrations of all detected components in groundwater samples from S1-1 are lower than the WHO recommended values for these components. All the samples in S1-1 are distributed in the margin areas of the basin (Fig. 4). 4.1.2. Deep groundwater The results of hierarchical cluster analysis and the features of every cluster for deep groundwater are presented in Figs. 5 and 6, where the phenon lines are defined as the distances of two and seven, respectively. Within these reasonable distances, 85% of the deep groundwater samples can be mainly divided into four clusters by combining two dendrograms (all medium-deep groundwater samples and the biggest cluster in Fig. 5): high F samples, high As samples, high F and As samples, and fitness samples. High F groundwater samples are included in the clusters D2, D3, D4, D5, D9, D10 and D1-2, which mainly occur in the mountain front area. High As groundwater samples are located in between Senggan River and Huangshui River (Fig. 7), which is similar to the case of the shallow groundwater. High-As deep groundwater samples are loaded on the D3 and D1-4. Compared with the shallow groundwater, deep groundwater contains lower As concentration and has a limited distribution area. Notably, the samples from D3 contained both high F and As concentrations (Fig. 7). Similar to shallow groundwater, the deep fitness groundwater group named D1-1 mainly occurs in the margin areas of the basin (Fig. 7). As discussed above, several different types of groundwater are classified by HCA. Therefore, HCA proves to be a useful technique to offer scientifically reasonable classification of groundwater samples based on logical calculation. The results are actually useful for optimization of future groundwater monitoring strategy in the basin by selecting representative locations for different groups of samples. 4.2. Fluoride in groundwater

Fig. 4. Regional distribution of different clusters in shallow groundwater samples.

In shallow groundwater, seven out of thirty-six clusters (S3, S4, S12, S21, S1-2, S1-7, S1-8) are characterized by high F concentration according to the results of HCA. Seven high F groundwater clusters (D2, D3, D4, D5, D9, D1-2) are detected out of all nineteen clusters in deep groundwater. The F concentration in high F shallow groundwater varies from 0.77 to 22 mg/L with an average value of 3.2 mg/L, and that of deep groundwater ranges from 1.09 to 8.3 mg/L with an average value of 2.6 mg/L. F is abundant in some hydroxyl minerals such as muscovite, biotite and apatite, and high F concentration in groundwater is mainly derived from mineral weathering (Jacks et al., 2005). Granite contains 0.05–0.14% of F, which is much higher than in other types of rock (Pertti and Backman, 2000). When granitic powders are reacted with purified water over a long period, F concentration in solution can reach up to 7 mg/L (Abdelgawad et al., 2009). Chae et al. (2007) also showed that F concentration in groundwater in Korea is very closely related to granite. However, the geochemical survey conducted by Zhao et al. (2007) in the Datong basin shows that the F content in Quaternary

Table 3 More detailed statistics of the first cluster of shallow groundwater in the study area (in mg/L except pH, As in μg/L, and Ec in μs/cm). Chemical type

118

Main cations: Na, Ca, Mg; main anion: HCO3

S1-2

26

S1-3

8

Na–HCO3, Ca–Mg–HCO3

S1-4

4

Na–HCO3–Cl, Ca–HCO3

S1-5

10

S1-6

4

Na–Cl–HCO3, Na–Ca–HCO3

S1-7

5

Na–HCO3–Cl, Na–Mg–Cl–HCO3

S1-8

20

Na–Mg–HCO3–Cl, Na–HCO3, Na–Cl–HCO3, Mg–Na–HCO3–Cl

Na–Mg–HCO3, Na–HCO3, Na–Ca–HCO3–Cl

Na–HCO3, Na–HCO3–Cl

Mean Max Min Mean Max Min Mean Max Min Mean Max Min Mean Max Min Mean Max Min Mean Max Min Mean Max Min

pH

Ec

As

F

TDS

Cl

NO3

SO4

HCO3

K

Na

Ca

Mg

Fe

Ba

Mn

Pb

Sr

7.85 8.57 7.07 7.98 8.70 7.57 7.89 8.18 7.64 7.97 8.34 7.59 7.85 8.31 7.48 7.87 7.98 7.73 7.66 7.86 7.53 7.89 8.33 7.57

719 1820 350 1520 2520 841 726 1080 340 1190 1810 675 993 1880 555 1270 2640 524 3030 4000 2030 3020 4740 1870

2.6 30.0 b 0.1 4.3 19.1 0.3 4.1 7.7 b 0.1 25.7⁎ 40.8⁎ 6.3 4.0 14.5 0.5 79.4⁎ 102⁎ 57.2⁎ 11.2⁎ 36.1⁎

1.09 3.27 b0.01 2.20⁎ 4.21⁎ 0.44 0.83 2.04 b0.01 0.52 1.12 b0.01 1.26 3.25 b0.01 0.41 0.83 0.05 1.58⁎ 2.48⁎ 0.76 2.37⁎ 3.91⁎ b0.01

28.8 144 5.75 133 290 49.6 42.2 110 5.31 135 261 67.9 77.9 175 15.1 131 341 12.3 294 373 152 235 580 52.4

18.5 73.2 b0.01 59.6 144 b0.01 12.1 31.0 b0.01 2.07 4.51 b0.01 53.6 172 b0.01 2.51 6.72 b0.01 46.0 108 6.96 59.7 204 b0.01

59.3 337 b 0.01 195 358.9 45.0 80.4 183 19.4 122 188 83.0 122 380 28.6 68.8 120 14.1 629 884 498 308 523 117

309 557 159 421 632 229 364 551 239 354 520 247 370 715 256 325 396 266 667 849 515 861 1130 592

2.45 8.10 b0.01 3.46 15.6 b0.01 4.69 14.7 b0.01 1.18 2.41 0.03 12.6 39.7 0.51 1.16 3.13 b0.01 4.21 7.39 1.81 4.73 16.3 1.04

46.8 187 6.36 134 249 37.4 81.7 188 11.6 111 212 16.9 103 231 19.5 112 286 25.0 410 577 251 451 657 177

46.9 137 12.1 47.3 81.3 14.6 45.7 77.1 14.6 66.2 122 18.0 51.1 108 21.9 46.8 51.5 40.3 86.0 132 37.6 31.8 70.0 8.75

29.3 69.9 8.29 74.5 121 27.9 30.6 58.0 15.6 35.7 49.9 26.6 49.7 100 22.5 37.8 84.5 11.1 106 172 33.1 72.2 128 16.9

0.016 0.102 b0.001 0.020 0.106 b0.001 0.009 0.025 b0.001 0.014 0.055 b0.001 0.128 0.275 0.013 0.021 0.069 b0.001 0.022 0.094 b0.001 0.011 0.072 b0.001

0.617 1.42 0.085 0.484 1.03 0.068 1.17⁎ 1.51⁎

b 0.1 4.0 16.1 b 0.1

388 972 205 858 1380 523 479 868 239 649 877 424 653 1060 275 562 1080 250 1920⁎ 2350⁎ 1580⁎ 1590⁎ 2310⁎

0.021 0.210 b 0.001 0.017 0.092 b 0.001 0.182 0.339 0.053 0.045 0.149 0.001 0.023 0.098 0.002 0.063 0.131 0.005 0.031 0.057 b 0.001 0.012 0.060 b 0.001

0.450 0.814 0.164 0.409 0.708 0.046 0.409 0.495 0.270 0.542 0.769 0.414 0.390 0.645 0.246 0.431 0.515 0.324 0.426 0.707 0.240 0.364 0.576 0.229

1.06 3.73 0.130 0.975 1.76 0.317 1.45 2.61 0.616 0.674 1.14 0.310 0.769 1.57 0.247 1.18 2.74 0.307 0.767 1.24 0.299 1.08 2.38 0.308

1050

0.570 1.29 1.84 0.900 0.577 1.01 0.160 0.528 0.809 0.344 0.830 1.31 0.292 0.615 1.16 0.170

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

No. S1-1

⁎ The feature value of the cluster.

83

84

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

Fig. 5. Dendrogram for all the deep groundwater samples.

basalt rocks from the Datong basin reaches up to 740 mg/Kg, while the Archeozoic complex only contains 255 mg/kg. In addition, the loess in the western mountain front areas and lacustrine sediment in the center of the basin also can be a source of F in the groundwater since the F contents are up to 895 mg/kg in loess and 609 mg/kg in sediment (Zhao et al., 2007). Thus, the Tertiary basalt rocks and Quaternary sediment may be the primary sources of F in the groundwater and may also be responsible for the pattern of distribution of high F samples in shallow and deep groundwater at Datong. Combining Figs. 4 and 7, it therefore suggests that F-enriched basalt and loess in the western mountain are the primary

sources of F in deep groundwater, and the Quaternary lacustrine sediment in the central part of the basin may be the major source of F in shallow groundwater. Apart from high F sources, geochemical processes are also important in the genesis of high F groundwater. The saturation indices calculated by PHREEQC indicates that all the groundwater samples are undersaturated with respect to fluorite apart from three exceptions (Fig. 10a and b). Moreover, F concentration is negatively correlated with the Ca in groundwater samples, especially for deep groundwater, suggesting that fluorite solubility draws a limit for F concentration in

Fig. 6. Dendrogram for the D1, the biggest cluster of deep groundwater samples.

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

85

Fig. 7. Regional distribution of different clusters in deep groundwater samples.

groundwater (Fig. 10b). From Piper's trilinear diagrams (Fig. 9), it can be seen that the predominant cation for all high F samples is Na, and the major anions are Cl and HCO3, and HCO3 for the shallow groundwater and deep groundwater, respectively. The Na–HCO3 type groundwater should have provided favorable conditions for F enrichment (Wang et al., 2009). On the one hand, higher HCO3 concentration is prone to precipitation of calcite, promoting the dissolution of F-enriched minerals under weaker fluorite solubility limit. The results of PHREEQC calculation show that the groundwater in the Datong basin is supersaturated with respect to calcite. On the other hand, groundwater with Na–HCO3 type is usually alkaline. F can be exchanged by OH from clay minerals or micas, such as muscovite and biotite under an elevated pH condition (Guo et al., 2007). Compared with deep groundwater, high F samples in shallow groundwater have higher Ec values, Na and Cl concentrations, which may be related with evapotranspiration and halite dissolution (Table 4; Fig. 8). For the shallow groundwater, evapotranspiration has a stronger effect, increasing the Ca and F concentrations in small degree under the permission of mineral saturation indices. When reaching the saturation point of calcite, the activity of Ca would decrease by precipitation as CaCO3. It's favorable for the dissolution of F-loaded minerals, which leads to gradually increase in fluorite SI and F concentrations in the shallow groundwater until reaching saturation point with fluorite (Fig. 10c) (Handa, 1975). Thus, Fig. 9. Piper diagram of high-F groundwater at Datong basin: (a) shallow groundwater samples; (b) deep groundwater samples.

Fig. 8. Variation of F with Cl in high fluoride groundwater samples from the Datong basin.

evapotranspiration may partly account for the high F concentration in the shallow groundwater. It is interesting that several deep groundwater samples from the center of the basin contain high As and F concentrations, ranging from 18.7 to 300 μg/L with a mean value of 192 μg/L and from 1.09 to 2.87 mg/L with a mean value of 1.99 mg/L, respectively (Fig. 7). All the samples are also typical Na–HCO3 type groundwater (Fig. 9). As discussed above, high pH values and HCO3 content are favorable for F enrichment in groundwater. The activity of Ca may decrease by cation exchange on the surface of the clay minerals and precipitation of calcite, and favors the F enrichment (Guo and Wang, 2004). In our previous studies, Fe-(hydr)oxides are regarded as the important sources of As in the groundwater at Datong (Xie, et al., 2009b). A sorption experiment

86

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

Fig. 10. Variation of F with Ca and fluorite SI in shallow groundwater and deep groundwater from the Datong basin.

designed by Kim et al. (2012) showed that Fe-(hydr)oxides in the sediment can also be an important host for F. F can be exchanged with structural oxygens on the surfaces of Fe-(hydr)oxides and be sorbed on the positive surfaces (Davis and Kent, 1990). With the pH increase, the adsorption capacity of Fe-(hydr)oxides on F and As decreases, releasing both of them into groundwater (Streat et al., 2008). 4.3. Arsenic in groundwater All the high As groundwater samples fall into groups of hierarchical cluster analysis. The high As groundwater samples are mainly included in S11, S1-4, S1-6, and D3, D1-4. The maximum concentration of As reaches up to 469 μg/L and 300 μg/L in shallow and deep groundwater, respectively (Table 1 and Fig. 5). Compared with fitness groundwater samples, all the high As groundwater samples occur under moderately

Table 4 Correlation coefficients between fluoride and other hydrochemical data in shallow and deep groundwater.

As TDS Cl NO3 SO4 HCO3 K Na Ca Mg pH Ec Fe Ba Mn Pb Sr

Shallow groundwater

Deep groundwater

0.060 0.454 0.409 0.303 0.383 0.481 0.045 0.454 − 0.204 0.350 0.093 0.553 0.063 − 0.033 0.188 − 0.055 0.079

0.130 0.226 0.104 − 0.127 0.229 0.295 − 0.001 0.265 − 0.329 0.136 0.358 0.228 − 0.040 − 0.068 − 0.044 0.270 0.033

reducing conditions, with negative Eh value that can be as low as −289 mv (Xie et al., 2009b) and low nitrate concentrations under detection limit (0.01 mg/L) (Figs. 11a and 12a). The detected low nitrate concentration in high As groundwater could be due to abiotic reduction processes or dissimilatory nitrate reduction (Jang and Liu, 2005). If so, positive correlation should exist between As and nitrate in groundwater. However, from Table 5 it can be seen that the correlation is weak and moderate for shallow and deep groundwater, respectively. It may imply that besides NO3, other electron acceptors are available in the aquifer system to promote oxidation of organic matter, which plays an important role to the generation of anoxic condition in high As aquifer system (Agusa et al., 2006; Farooq et al., 2011, 2012; Stüben et al., 2003;). For instance, SO4 in groundwater may function as the electron acceptor in high As aquifer (Guo et al., 2008), since SO4 concentrations in high As groundwater are mostly below detection limit (b0.01 mg/L) with a median value of b0.01 mg/L (Figs. 11b and 12b). In contrast, the SO4 concentrations in fitness shallow groundwater ranges from b0.01 to 883.5 mg/L with a median value of 66.48 mg/L. Except for the lower SO4 concentration, our previous studies showed that there is a positive correlation between δ 34S values and As concentration, indicating the importance of sulfate reduction in the aquifers (Xie et al., 2009b). The deep groundwater samples contain HCO3 as high as 1080 mg/L, and a strong correlation has been observed between As and HCO3 in deep groundwater (Table 5). Saturation indices (SI) of selected minerals calculated by PHREEQC show that all the groundwater are oversaturated with respect to dolomite and calcite, suggesting that the elevated HCO3 concentrations in groundwater not only derive from the dissolution of carbonate minerals, and may partly originate from the oxidation of organic matter driven by the microbes in sediments and in aquifers (Oremland and Stolz, 2005). As a matter of fact, the DOC contents of the aquifer sediments from Datong can be up to 1.0% in lacustrine black silty clay and dark gray clay (Wang et al., 2009). The Fe2O3 content in the lacustrine sediments in Datong basin ranges between 3.3 and 6.2 wt.% and the samples with highest Fe2O3 concentration is black clay which has the highest As concentration (Guo et al., 2008; Xie et al., 2009a,b;). Under reductive environment,

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

87

Fig. 11. Variation of As with NO3, SO4 and Fe in shallow groundwater from the Datong basin.

As can be released into groundwater through redox reactions such as the dissolution of As-enriched Fe-oxides (Anawar et al., 2003; Kim et al., 2002; McArthur et al., 2001, 2004). Therefore, As and Fe concentrations should have a positive correlation. The mean Fe concentration of high As samples in shallow groundwater is 7.2 μg/L, while that of fitness samples is 15.6 μg/L. However, the expected close correlation between Fe and As does not exist, which was also observed by Nickson et al. (2000) (Table 5). Microbial reduction of Fe(III) does not necessarily release all Fe(II) into groundwater due to retention in the solid phase or formation of Fe(II)-bearing minerals (Fredrickson et al., 1998), keeping some amount of Fe in the solid phase but releasing As into the aqueous phase (Figs. 11c and 12c). Organic matters and As-bearing Fe-(hydr)oxides are also likely to be enriched in the fine lacustrine sediments. Moreover, biogeochemical processes play an important role in As mobilization in the aquifers (Guo, et al., 2011). Thus, the reductive dissolution of Fe-(hydr)oxides driven by sulfate reduction and biodegradation of organic matters could be the major process of As enrichment in groundwater at Datong.

Fig. 12. Variation of As with NO3, SO4 and Fe in deep groundwater from the Datong basin.

5. Conclusions In this study, hierarchical cluster analysis is applied to groundwater samples from the Datong basin to delineate the spatial distribution of some special elements. The results demonstrate that all high As samples occur in the southeastern side of the Senggan River in Shanyin county, while high-F deep groundwater samples distribute in the western mountain front areas and that of shallow groundwater mainly occur in the discharge zone of the basin. Both high F and As concentrations were detected in several deep groundwater. The hydrogeochemical characteristics of high As groundwater and high F groundwater are also evaluated in the Datong basin. It suggests that the F concentration in all high F groundwater is controlled by fluorite solubility, and the evapotranspiration is also an essential factor for F enrichment in the shallow groundwater. Both high As and F groundwater have elevated pH values and belonged to NaHCO3 type. Resorption from Fe-(hydr)

88

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89

Table 5 Correlation coefficients between arsenic and other hydrochemical data in shallow and deep groundwater.

F TDS Cl NO3 SO4 HCO3 K Na Ca Mg pH Ec Fe Ba Mn Pb Sr

Shallow groundwater

Deep groundwater

− 0.023 0.287 0.449 − 0.368 − 0.125 0.432 − 0.151 0.490 − 0.431 − 0.011 0.298 0.345 − 0.116 − 0.026 0.242 − 0.144 − 0.159

0.608 0.396 0.331 − 0.507 − 0.430 0.732 − 0.381 0.674 − 0.721 − 0.275 0.615 0.594 − 0.178 0.161 0.007 − 0.047 0.023

oxides under elevated pH conditions is one of the major processes releasing As and F into groundwater. High As groundwater are characterized by low concentrations of NO3 and SO4, and high concentrations of HCO3, suggesting the strong reducing condition in the study area which can also be proved by negative values of Eh. The high HCO3 concentrations maybe related to the oxidative degradation of organic matter creating reducing environment favorable for As mobilization in the aquifers. Our previous studies showed that the As loaded iron hydroxide is the main source of As in aquifers. Therefore, the relationship between As and other parameters further supports microbially mediated reductive dissolution of Fe-(hydr)oxides as the dominant mechanism of As release to groundwater. Acknowledgment The research work is supported by the National Natural Science Foundation of China (40830748 and 40902071), and the Ministry of Education of China (111 project and SRFDP). We would like to thank Dr. Chandrasekharam Dornadula and the editor for their constructive comments and suggestions on this manuscript. References Abdelgawad, A.M., Watanabe, K., Takeuchi, S., Mizuno, T., 2009. The origin of fluoriderich groundwater in Mizunami area, Japan — Mineralogy and geochemistry implications. Engineering Geology 108, 76–85. Agusa, T., Kunito, T., Fujihara, J., Kubota, R., Minh, T.B., Trang, P.T.K., Iwata, H., Subramanian, A., Viet, P.H., Tanabe, S., 2006. Contamination by arsenic and other trace elements in tube well water and its risk assessment to humans in Hanoi, Vietnam. Environmental Pollution 139, 95–106. Anawar, H.M., Akai, J., Komaki, K., Terao, H., Takahashi, T., Ishzuak, T., Safiulah, S., Kato, K., 2003. Geochemical occurrence of arsenic in groundwater of Bangladesh, source and mobilization process. Journal of Geochemical Exploration 77, 109–131. Apambire, W.B., Boyle, D.R., Michel, F.A., 1997. Geochemistry, genesis and health implication of fluoriferous groundwater in the upper regions, Ghana. Environmental Geology 33, 13–24. Berg, M., Stengel, C., Trang, P.T.K., Viet, P.H., Sampson, M.L., Leng, M., Samreth, S., Fredericks, D., 2007. Magnitude of arsenic pollution in the Mekong and Red River Deltas—Cambodia and Vietnam. Science of the Total Environment 372, 413–425. Bhattacharya, P., Claesson, M., Bundschuh, J., Sracek, O., Fagerberg, J., Jacks, G., Martin, R.A., Storniolo, A.R., Thir, J.M., 2006. Distribution and mobility of arsenic in the Río Dulce alluvial aquifers in Santiago del Estero Province, Argentina. Science of the Total Environment 358 (1–3), 97–120. Chae, G.T., Yun, S.T., Mayer, B., Kim, K.H., Kim, S.Y., Kwon, J.S., Kim, K., Koh, Y.K., 2007. Fluorine geochemistry in bedrock groundwater of South Korea. Science of the Total Environment 385, 272–283. Cloutier, V., Lefebvre, R., Therrien, R., Savard, M.M., 2008. Multivariate statistical analysis of geochemical data as indicative of the hydrogeochemical evolution of groundwater in a sedimentary rock aquifer system. Hydrogeology Journal 353, 294–313. Currell, M., Cartwright, I., Raveggi, M., Han, D., 2011. Controls on elevated fluoride and arsenic concentrations in groundwater from the Yuncheng basin, China. Applied Geochemistry 26, 540–552.

Davis, J.C., 1986. Statistics and Data Analysis in Geology. John Wiley & Sons Inc., New York. Davis, J.A., Kent, D.B., 1990. Surface complexation modeling in aqueous geochemistry. Reviews in Mineralogy and Geochemistry 23, 177–260. Deng, Y., Wang, Y., Ma, T., 2009. Isotope and minor element geochemistry of high arsenic groundwater from Hangjinhouqi, the Hetao Plain, Inner Mongolia. Applied Geochemistry 24 (4), 587–599. Farooq, S.H., Chandrasekharam, D., Berner, Z., Stüben, D., 2010. Influence of traditional agricultural practices on mobilization of arsenic from sediments to groundwater in Bengal delta. Water Research 44 (19), 5575–5588. Farooq, S.H., Chandrasekharam, D., Norra, S., Berner, Z., Eiche, E., Thambidurai, P., Stüben, D., 2011. Temporal variations in arsenic concentration in the groundwater of Murshidabad District, West Bengal, India. Environmental Earth Science 62, 223–232. Farooq, S.H., Chandrasekharam, D., Abbt-Braun, G., Berner, Z., Norra, S., Stüben, D., 2012. Dissolved organic carbon from the traditional jute processing technique and its potential influence on arsenic enrichment in the Bengal Delta. Applied Geochemistry 27 (1), 292–303. Fredrickson, J.K., Zachara, J.M., Kennedy, D.W., Dong, H.L., Onstott, T.C., Hinman, N.W., Li, S.M., 1998. Biogenic iron mineralization accompanying the dissimilatory reduction of hydrous ferric oxide by a groundwater bacterium. Geochimica et Cosmochimica Acta 62, 3239–3257. Gaciri, S.J., Davies, T.C., 1993. The occurrence and geochemistry of fluoride in some natural waters of Kenya. Journal of Hydrology 143, 395–412. Grimaldo, M., Borja-Aburto, V.H., Ramírez, A.L., Ponce, M., Rosas, M., Diaz-Barriga, F., 1995. Endemic fluorosis in San Luis Potosí, Mexico. Environmental Research 68, 25–30. Güler, C., Thyne, G., McCray, J.E., Turner, K.A., 2002. Evaluation and graphical and multivariate statistical methods for classification of water chemistry data. Hydrogeology Journal 10, 455–474. Guo, H., Wang, Y., 2004. Hydrogeochemical processes in shallow Quaternary aquifers from the northern part of Datong basin, China. Applied Geochemistry 19 (1), 19–27. Guo, H., Wang, Y., 2005. Geochemical characteristics of shallow groundwater in Datong basin, northwestern China. Journal of Geochemical Exploration 87, 109–120. Guo, H., Wang, Y., Shpeizer, G.M., Yan, S., 2003. Natural occurrence of arsenic in shallow groundwater, Shanyin, Datong basin, China. Journal of Environmental Science and Health, Part A: Toxic/Hazardous Substances & Environmental Engineering 38, 2565–2580. Guo, Q., Wang, Y., Ma, T., Ma, R., 2007. Geochemical processes controlling the elevated fluoride concentrations in groundwaters of the Taiyuan basin, Northern China. Journal of Geochemical Exploration 93 (1), 1–12. Guo, H., Yang, S.Z., Tang, X.H., Li, Y., Shen, Z.L., 2008. Groundwater geochemistry and its implications for arsenic mobilization in shallow aquifers of the Hetao basin, Inner Mongolia. Science of the Total Environment 393, 131–144. Guo, H., Zhang, B., Li, Y., Berner, Z., Tang, X., Norra, S., Stüben, D., 2011. Hydrogeological and biogeochemical constrains of arsenic mobilization in shallow aquifers from the Hetao basin, Inner Mongolia. Environmental Pollution 159 (4), 876–883. Handa, B.K., 1975. Geochemistry and genesis of fluoride containing groundwaters in India. Ground Water 13, 275–281. Han, Y., 2008. Characters of the groundwater flow field and hydrochemistry field in Datong Basin. Geological Survey Resource 32, 138–146 (in Chinese with English abstract). Jacks, G., Bhattacharya, P., Chaudhary, V., Singh, K.P., 2005. Controls on the genesis of some high-fluoride groundwater in India. Applied Geochemistry 20, 221–228. Jang, C.S., Liu, C.W., 2005. Contamination potential of nitrogen compounds in the heterogeneous aquifers of the Choushui River alluvial fan, Taiwan. Journal of Contaminant Hydrology 79, 135–155. Jayawardana, D.T., Pitawala, H.M.T.G.A., Ishiga, H., 2012. Geochemical assessment of soils in districts of fluoride-rich and fluoride-poor groundwater, north-central Sri Lanka. Journal of Geochemical Exploration 114, 118–125. Kim, M.J., Nriagu, J., Haack, S., 2002. Arsenic species and chemistry in groundwater of southeast Michigan. Environmental Pollution 120, 379–390. Kim, S.H., Kim, K., Ko, K.S., Kim, Y., Lee, K.S., 2012. Co-contamination of arsenic and fluoride in the groundwater of unconsolidated aquifers under reducing environments. Chemosphere 87, 851–856. Li, Y., 2001. Environmental chemistry study of waterborne poisoning in Shanyin, Shanxi province. M.Sc. Thesis. China University of Geosciences, Wuhan, P.R. China (in Chinese, with English Abstract). McArthur, J.M., Ravenscroft, P., Safiullah, S., Thirlwall, M.F., 2001. Arsenic in groundwater, testing pollution mechanisms for sedimentary aquifers in Bangladesh. Water Resources Research 37, 109–117. McArthur, J.M., Banerjee, D.M., Hudson-Edwards, K.A., Mishra, R., Purohit, R., Ravenscroft, P., Cronin, A., Howarth, R.J., Chatterjee, A., Talukder, T., Lowry, D., Houghton, S., Chadha, D.K., 2004. Natural organic matter in sedimentary basins and its relation to arsenic in anoxic groundwater: the example of West Bengal and its worldwide implications. Applied Geochemistry 19, 1255–1293. Msonda, K.W.M., Masamba, W.R.L., Fabiano, E., 2007. A study of fluoride groundwater occurrence in Nathenje, Lilongwe, Malawi. Physics and Chemistry of the Earth 32, 1178–1184. Nickson, R.T., McArthur, J.M., Burgess, W.G., Ahmed, K.H., Ravenscroft, P., Rahman, M., 1998. Arsenic poisoning of Bangladesh groundwater. Nature 395, 338. Nickson, R.T., McArthur, J.M., Ravenscroft, P., Burgess, W.G., Ahmed, K.M., 2000. Mechanism of arsenic release to groundwater, Bangladesh and West Bengal. Applied Geochemistry 15, 403–413. Oremland, R.S., Stolz, J.F., 2005. Arsenic, microbes and contaminated aquifers. Trends in Microbiology 13, 45–49. Pertti, L., Backman, B., 2000. The occurrence and geochemistry of fluorides with special reference to natural waters in Finland. Geological Survey of Finland, Report of Investigation 149.

J. Li et al. / Journal of Geochemical Exploration 118 (2012) 77–89 Romero, L., Alonso, H., Campano, P., Fanfani, L., Cidu, R., Dadea, C., Keegan, T., Farago, T.M., 2003. Arsenic enrichment in waters and sediments of the Rio Loa (Second Region, Chile). Applied Geochemical 18 (9), 1399–1416. Rowland, H.A.L., Omoregie, E.O., Millot, R., Jimenez, C., Mertens, J., Baciu, C., Hug, S.J., Berg, M., 2011. Geochemistry and arsenic behaviour in groundwater resources of the Pannonian basin (Hungary and Romania). Applied Geochemical 26 (1), 1–17. Roy, A., Kordas, K., Lopez, P., Rosado, J.L., Cebrian, M.E., Vargas, G.G., Ronquillo, D., Stoltzfus, R.J., 2004. Association between arsenic exposure and behavior among first-graders from Torreón, Mexico. Environmental Research 111 (5), 670–676. Rukah, Y.A., Alsokhny, K., 2004. Geochemical assessment of groundwater contamination with special emphasis on fluoride concentration, North Jordan. Chemie der ErdeGeochemistry 64 (2), 171–181. Shomar, B., Müller, G., Yahya, A., Askar, S., Sansur, R., 2004. Fluorides in groundwater, soil and infused-black tea and the occurrence of dental fluorosis among school children of the Gaza Strip. Journal of Water and Health 2, 23–35. Singh, K.P., Malik, A., Mohan, D., Sinha, S., 2004. Multivariate statistical techniques for the evaluation of spatial and temporal variation in water quality of Gomti River (India)—a study. Water Research 38, 3989–3992. Singh, K.P., Malik, A., Sinha, S., 2005. Water quality assessment and apportionment of pollution sources of Gomti River (India) using multivariate statistical technique, a case study. Analytica Chimica Acta 538, 355–374. Streat, M., Hellgardt, K., Newton, N.L.R., 2008. Hydrous ferric oxide as an adsorbent in water treatment part 3: batch and mini-column adsorption of arsenic, phosphorus, fluorine and cadmium ions. Process Safety and Environmental Protection 86, 21–30. Stüben, D., Berner, Z., Chandrasekharam, D., Karmakar, J., 2003. Arsenic enrichment in ground water of West Bengal, India, geochemical evidence for mobilization of As under reducing conditions. Applied Geochemistry 18, 1417–1434. Su, C., 2006. Regional hydrogeochemistry and genesis of high arsenic groundwater at Datong basin, Shanxi Province, China. Ph.D. Thesis. China University of Geosciences, Wuhan, P.R. China (in Chinese, with English Abstract).

89

Wang, J., Zhao, L., Wu, Y., 1998. Environmental geochemical study on arsenic in arseniasis areas in Shanyin and Yingxian, Shanxi province. Geoscience 12 (2), 243–248 (In Chinese with English abstract). Wang, Y., Shvartsev, S., Su, C., 2009. Genesis of arsenic/fluoride-enriched soda water, a case study at Datong, northern China. Applied Geochemistry 24 (4), 641–649. Xie, X., Wang, Y., Su, C., Liu, H., Duan, M., Xie, Z., 2008. Arsenic mobilization in shallow aquifers of Datong basin: hydrochemical and mineralogical evidences. Journal of Geochemical Exploration 98 (3), 107–115. Xie, X., Wang, Y., Duan, M., Liu, H., 2009a. Sediment geochemistry and arsenic mobilization in shallow aquifers of the Datong basin, northern China. Environmental Geochemistry and Health 31, 493–502. Xie, X., Ellis, A., Wang, Y., 2009b. Geochemistry of redox-sensitive elements and sulfur isotopes in the high arsenic groundwater system of Datong basin, China. Science of the Total Environment 407, 3823–3835. Xie, X., Wang, Y., Su, C., 2011. Hydrochemical and sediment biomarker evidence of the impact of organic matter biodegradation on arsenic mobilization in shallow aquifers of Datong basin, China. Water, Air, and Soil Pollution 223 (2), 483–498 2012. Xie, X., Wang, Y., Su, C., Li, J., Li, M., 2012. Influence of irrigation practices on arsenic mobilization, evidence from isotope composition and Cl/Br ratios in groundwater from Datong basin, northern China. Journal of Hydrology 424–425, 37–47. Yidana, S.M., 2010. Groundwater classification using multivariate statistical methods, Southern Ghana. Journal of African Earth Sciences 57 (5), 455–469. Yidana, S.M., Yakubo, B.B., Akabzaa, T.M., 2010. Analysis of groundwater quality using multivariate and spatial analyses in the Keta basin, Ghana. Journal of African Earth Sciences 58 (2), 220–234. Zhao, L., Wu, S., Zhou, J., Wang, J., Wang, J., 2007. Geochemical investigation on the endemic As and F poisoning in Datong basin. Earth Science Frontiers 14 (2), 225–235 (in Chinese with English abstract).