ARTICLE pubs.acs.org/est
Assessing the Labile Arsenic Pool in Contaminated Paddy Soils by Isotopic Dilution Techniques and Simple Extractions Jacqueline L. Stroud,† M. Asaduzzman Khan,† Gareth J. Norton,‡ M. Rafiqul Islam,§ Tapash Dasgupta,|| Yong-Guan Zhu,^ Adam H. Price,‡ Andrew A. Meharg,‡ Steve P. McGrath,† and Fang-Jie Zhao*,† †
Rothamsted Research, Harpenden, Hertfordshire AL5 2JQ, U.K. Institute of Biological and Environmental Sciences, University of Aberdeen, Cruickshank Building, St. Machar Drive, Aberdeen, AB24 3UU, U.K. § Department of Soil Science, Bangladesh Agricultural University, Mymensingh-2202, Bangladesh Institute of Agricultural Science, Calcutta University, 35 B.C. Road, Kolkata 700 019 West Bengal, India ^ Research Centre for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China
)
‡
bS Supporting Information ABSTRACT: Arsenic (As) contamination of paddy soils threatens rice cultivation and the health of populations relying on rice as a staple crop. In the present study, isotopic dilution techniques were used to determine the chemically labile (E value) and phytoavailable (L value) pools of As in a range of paddy soils from Bangladesh, India, and China and two arable soils from the UK varying in the degree and sources of As contamination. The E value accounted for 6.221.4% of the total As, suggesting that a large proportion of soil As is chemically nonlabile. L values measured with rice grown under anaerobic conditions were generally larger than those under aerobic conditions, indicating increased potentially phytoavailable pool of As in flooded soils. In an incubation study, As was mobilized into soil pore water mainly as arsenite under flooded conditions, with Bangladeshi soils contaminated by irrigation of groundwater showing a greater potential of As mobilization than other soils. Arsenic mobilization was best predicted by phosphate-extractable As in the soils.
’ INTRODUCTION Elevated levels of arsenic (As) in groundwater is a serious problem in south Asia, particularly the Bengal delta.1 This groundwater is extracted to irrigate rice crops during the dry (boro) season, leading to accumulation of As in the paddy soils.24 Field studies have shown that As accumulation in paddy soils can cause substantial decreases in rice yield.5 Elsewhere, some paddy soils can be contaminated with As from mining activities or geogenic sources.6 Furthermore, rice is an efficient accumulator of As, resulting in an elevated transfer of As from soil to the food chain.7,8 This may pose significant health risk for populations relying on rice as the staple food.9 A number of recent studies have investigated As accumulation and attenuation in paddy fields irrigated with As-contaminated groundwater.2,1012 However, factors that influence As mobility and phytoavailability in paddy soils are not well understood. It is known that As is mobilized under anaerobic conditions primarily due to a reductive dissolution of Fe oxides/hydroxides, which provide the main sorption sites for As in soil.1315 However, the extent of As mobilization varies among different studies.5,14,15 It is possible that this mobilization may be related more closely to r 2011 American Chemical Society
the size of the labile As pool than to the total concentration in the soil. The labile pool of a metal or metalloid can be determined using isotopic dilution techniques.16,17 In these techniques, a small quantity of a tracer isotope (radioactive or stable) is introduced to soil, which redistributes itself among the solution and the exchangeable phases in the same way as the other nontracer isotopes of the same element. By measuring of the activity of the tracer isotope and the concentration of the element in the soil solution, the isotopically exchangeable pool (E value) of the element can be quantified; this pool is often referred to as the labile pool.16,17 Alternatively, a test organism (e.g., plant) may be introduced to take up the tracer and nontracer isotopes, which then allows estimation of the L value, representing the bioavailable pool.16,18 Arsenic E values have been determined in mine- contaminated and laboratory-spiked soils, with the labile As pool ranging between 0.4 and 73% of the total soil As;1922 Received: December 6, 2010 Accepted: April 8, 2011 Revised: March 31, 2011 Published: April 19, 2011 4262
dx.doi.org/10.1021/es104080s | Environ. Sci. Technol. 2011, 45, 4262–4269
Environmental Science & Technology however, no values have been reported for paddy soils. There have been also no reports of arsenic L values in the literature. Because As speciation is sensitive to changes in the redox potential, isotope dilution measurements may be subject to the influence of soil water conditions in paddy soils.16 Understanding this effect may provide valuable information that can be used for paddy field water management. The objectives of the present study were to (1) determine the labile pool of As (E value) in a range of paddy soils contaminated with As from different sources and to different degrees; (2) determine the phytoavailable pool of As (L value) using rice as the test plant grown under both aerobic and anaerobic conditions; and (3) investigate if flooding-induced As mobilization can be predicted using the E value or other simple extraction methods. For comparison, two UK arable soils were included in the study.
’ MATERIALS AND METHODS Soils. Sixteen paddy soils from Bangladesh, India, and China, and two UK arable soils were collected from the plow layer (020 cm). The source of As varied from As-contaminated irrigation water (14 sites), mining/smelter runoff (1 site), to geogenic source (3 sites). Selected soil properties, such as pH, organic C, total N, available P (by 0.5 M NaHCO3 extraction, the Olsen method), and total As concentrations (by aqua-regia digestion), were determined following standard methods of soil analysis. The soils had a range of physicochemical properties (Table 1). All soils were air-dried and sieved to 99%, 37 MBq/mL, Oak Ridge National Laboratory) was added to each sample and shaken for a further 48 h (isotopic exchange period). All tubes were centrifuged for 15 min at 2100g. The supernatant was filtered through 0.2-μm before γ-spectrometry counting of 73As activity (1480 Wizard, Wallac) and As speciation analysis by high performance liquid chromatography (HPLC, Agilent 1100, Agilent Technologies) coupled to ICP-MS using anion-exchange separation as described
ARTICLE
previously.15 All analyses included blanks and solutions spiked with the 73As so that the total amount of radioisotope added to the soil slurry could be determined. E values were determined in all soils. In a number of selected samples, As(V) and As(III) were separated by HPLC and collected for γ-spectrometry counting to determine the 73As radioactivities of As(V) and As(III). An attempt was also made to measure As E values under anaerobic conditions by filling the tubes with N2 gas and incubating the soil suspensions for 2 weeks under gentle shaking. However, arsenate remained the predominant As species at the end of the incubation, suggesting that anaerobic conditions did not develop. L Value Determination. The five soils selected for L value determination included two each from Bangladesh and China and one from India. Soil (90 g dry weight) was weighed into 100-mL black plastic pots in quadruplicate. The soil was hydrated with deionized water to 40% water holding capacity before spiking of 73As to avoid the disproportionate distribution of the spike on external adsorption sites of microaggregates causing underestimation of L values.25 A spike (50 μL) of 73As (300 kBq mL1) as As(V) was added to the soil and mixed thoroughly with a plastic spatula. Two conditions were applied to the soils: (1) aerobic—where soil was wetted to 70% water holding capacity (WHC) and maintained for the duration of the experiment or (2) flooded—where soil was flooded to have 2 cm water above the surface of the soil for the duration of the experiment. The experiment was carried out at two time points, with pots planted at 0 d or 8 weeks after 73As and water additions. Each pot was planted with two emerging (5 day old) rice seedlings (Oryza sativa L. cv. Shatabdi). The planted pots were placed in a controlled environment room with day/night temperatures 28/23 C, day/night relative humidity 80/60%, and light period 12 h per day with fluorescence lamps to maintain light intensity of 550 μmol m2 s1. Soilwater conditions were maintained with daily additions of water for 35 days, and the soils were fertilized with a basal nutrient solution of 120 mg N kg1 as NH4NO3, 30 mg P kg1 as K2HPO4, 75.5 mg K kg1 as KCl, and 25 mg of S kg1 as MgSO4. After 35 days, the above-ground tissue of rice was removed, rinsed, cut into small pieces, and analyzed using γ-spectrometry. The plant samples were then oven-dried (60 C) for 48 h and digested with HNO3/HClO4 and analyzed for As concentration by ICP-MS. A certified reference material (NIST 1568a rice flour) was used for quality control, which gave a mean value ((SD) of 274 ( 13 μg As kg1 compared with the certified values of 290 ( 30 μg As kg1. To check the initial γ-spectrometry analysis on the whole plant (due to potential measurement complications from heterogeneous plant tissue), a 2-mL portion of the digest was γ-counted, and the total activity in the digest compared to the original plant tissue and was within 100 ( 8%. Arsenic Mobilization in Soil Pore Water. Fourteen soils were used in a laboratory incubation experiment to determine As mobilization and speciation in soil pore water. Triplicates of 500 g of soil were placed in plastic pots and flooded with deionized water to maintain a 2-cm layer of standing water above the soil surface. Soils were placed in a constant temperature room (25 C). A Rhizon soil solution sampler was inserted to each pot for the collection of soil pore water at 3, 7, 14, 21, and 30 days after flooding. Soil pore water was analyzed for pH, Eh, Fe concentration, and As species. Eh was measured immediately after pore water collection using a combined platinum and silver/silver chloride electrode system. Soluble Fe was determined by a colorimetric method26 and As species were quantified by HPLC-ICP-MS.15 4263
dx.doi.org/10.1021/es104080s |Environ. Sci. Technol. 2011, 45, 4262–4269
a
4264
Bangladesh
Bangladesh
India
India
China China
UK
UK
Faridpur (B11)
Sonargoan (B12)
De Ganga (I1)
Nonaghata (I2)
Chenzhou (C1) Qiyang (C2)
Rothamsted (U1)
Woburn (U2)
irrigation
geogenic
geogenic
mining geogenic
irrigation
irrigation
irrigation
irrigation
irrigation
irrigation
irrigation irrigation
irrigation
irrigation
irrigation
irrigation
irrigation
7.2
6.8
6.0
7.6 7.3
7.2
6.4
7.1
7.9
7.8
7.7
7.3 7.6
7.4
7.9
7.1
6.9
6.5
73.1
9.6
43.6 23.1
11.5
3.9
13.1
15.7
26.4
20.8
22.2 21.5
21.5
17.5
15.4
15.5
17.5
24
200
12
15 13
13
27
12
5.9
33
50
158 119
78
10
9.2
100
87
10209
1703
2483 3835
789
462
914
610
1657
2138
4369 4267
3120
1455
959
4538
3938
362
(mg kg1)
amorphous Fe
total As
46
6.8
60 79
6.2
17
12
34
13
13
73 77
138
62
10
72
114
4.0
(mg kg1)
0.7
0.2
2.8 1.9
0.9
0.2
2.0
3.3
1.2
1.2
4.4 4.8
12.7
5.0
0.8
5.5
8.4
0.5
(mg kg1)
phosphate-extractable As
A subset of the data (basic soil properties, total and phosphate-extractable As in the soils B112) was published in ref 31.
Bangladesh
Bangladesh Bangladesh
Paranpur (B7) Paranpur (B8)
Bangladesh
Bangladesh
Dhumrakandi (B6)
Badarpur (B10)
Bangladesh
Dhumrakandi (B5)
Badarpur (B9)
Bangladesh
Bangladesh
Dhumrakandi (B4)
Fatehpur (B2)
Fatehpur (B3)
Bangladesh
Bangladesh
Shenbag (B1)
7.5
available P (mg kg1)
(g kg1)
pH
organic C
As source
(label)
country
soil sampling site
Table 1. Selected Soil Properties and Arsenic Characterizationa
8.5
0.9
13.5 13.0
0.8
2.6
2.4
4.4
2.6
3.1
20.9 25.2
43.7
15.9
1.6
22.9
36.4
0.5
(mg kg1)
oxalate-extractable As
As E value
2.9
0.6
9.0 9.3
0.7
2.8
2.6
5.9
2.9
2.5
11.8 16.4
17.2
7.9
1.3
8.3
11.8
0.6
(mg kg1)
As E value
6.2
9.2
15.2 11.9
10.7
16.3
17.8
17.2
21.4
19.2
16.2 21.4
12.5
12.7
13.2
11.6
10.4
15.2
As % of total As
Environmental Science & Technology ARTICLE
dx.doi.org/10.1021/es104080s |Environ. Sci. Technol. 2011, 45, 4262–4269
Environmental Science & Technology
ARTICLE
Figure 1. Effect of soil water conditions on As L values (reported as % of total As) for the 5 soils determined with no prior incubation (a) or with 8 weeks prior incubation (b) (mean ( S.E.). See Table 1 for soil labels.
Assumptions, Calculations and Statistical Analysis. All isotope dilution analysis relies on several assumptions: that the isotope addition does not perturb the existing As equilibrium in the soil, that the isotope mimics the distribution of the target chemical, the isotope is mixed evenly though the soil (for L value determination), and that the “equilibrium” period, while a snapshot in time, is taken at the plateau of rapid exchange reactions.16 All values relating to 73As were corrected for the decay by counting the stock solution and the samples at the same time. The E value was calculated according to Hamon et al. as follows:16 Csol V E¼ R W Csol
where Csol is the concentration of As in solution (μg/mL), C*sol is the concentration of the radioisotope remaining in solution after equilibrium (Bq/mL), R is the total amount of radioisotope that was added to each sample (Bq/mL), and V/W is the ratio of solution to sample (mg/g). The L value of As was calculated according to Larsen.18 activity applied per pot L¼ ðAs in plant material activity of plant As in rice seedÞ The As in the rice seed used in the experiment was determined according to Oliver et al.27 where rice seedlings were germinated and shoots, roots, and seed shells were separated, dried, and digested. The percentage of seed As transferred to shoots was determined and used to correct the L values. Although the 73As introduced to the soils was in the form of arsenate, it was likely to undergo the same species transformations as the nontracer isotope of As in the soils depending on the soil water conditions, thus simplifying the calculations of L values.16 To calculate the proportion of labile As, the labile soil As concentration (E or L value) was divided by the total soil As concentration and expressed as a percentage. Linear regression analysis was performed to investigate the relationships between E value or soil pore-water As and other soil properties, using Genstat (version 12; VSN International Ltd., Hemel Hempstead, UK).
’ RESULTS Soil As Concentrations. Total As concentrations of the soils used in this study varied from 4 to 138 mg As kg1, and the sources of As included geogenic, mining, and irrigation water (Table 1). Ammonium phosphate extracted 1.117.8% of the total As. Generally, the percentage of phosphate-extractable As was larger in the paddy soils from the Bengal delta (6.117.8%, except the soil from De Ganga (I1) having a low value of 1.1%) than soils from China and the UK (1.54.8%). Ammonium oxalate extracted 11.832.9% of the total As from the soils with no clear difference between different sources of As contamination. E Values. During the 48-h equilibration following 73As addition, only arsenate was detected in the soil slurry, meaning that the E value determined was a measurement of isotopically exchangeable arsenate. Among the 18 soils, the E values accounted for between 10.4 and 21.4% (n = 14) of total As in paddy soils from the Bengal delta impacted by As from irrigation; 11.8 and 15.2% (n = 2) in paddy sites with mining/geogenic As sources, and 6.2 and 9.2% (n = 2) in UK arable soils with geogenic As sources (Table 1). In all but one soil, the E values were larger than the concentrations of phosphate-extractable As, but were more comparable with those of the oxalate-extractable As (the E values being 32134% of the oxalate-extractable As). Soil properties that influenced As E values were investigated using multiple regression analysis. The following regression equations show that the significant parameters affecting the As E values were total or oxalate-extractable As and pH:
E value ¼ 13:0 þ 0:12Total As þ 1:9pH ðR 2 adj ¼ 0:875, n ¼ 18, P < 0:001Þ E value ¼ 15:9 þ 0:37Oxalate As þ 2:5pH ðR 2 adj ¼ 0:885, n ¼ 18, P < 0:001Þ L Values. The L value was measured for the 5 paddy soils using rice, which was grown either immediately or 8 weeks after 73As addition. Across all soil and treatment combinations, the As L values accounted for 630% of the total soil As (Figure 1). Analysis of variance showed highly significant effects (P < 0.001) of soil, water regime (aerobic vs flooded, and incubation time (0 and 8 weeks), as well as significant (P = 0.005) interactions between soil and water regime. At both incubation time points, 4265
dx.doi.org/10.1021/es104080s |Environ. Sci. Technol. 2011, 45, 4262–4269
Environmental Science & Technology the same trend in L values was found across the 5 paddy soils. The L value was similar in the aerobic and flooded treatments of Nonaghata soil (I2), but in other four soils the flooded conditions produced 63110% higher L values than the aerobic conditions. Eight-week incubation led to greater decreases in the L value under aerobic conditions (mean decrease 33%) than under anaerobic conditions (mean decrease 9%), suggesting greater fixation of 73As under aerobic conditions. Comparing the two isotope dilution methods, the E values were similar to the aerobic L values measured without incubation after 73As addition, with the exception of Nonaghata soil in which both aerobic and flooded L values were more than double the E value. Arsenic Mobilization into Soil Pore Water. Fourteen soils were incubated under flooded conditions for 30 days to determine the extent and time-course of As mobilization into the soil pore water. Except soils U1 and B1, pore-water Eh decreased to 60 mg kg1) due to irrigation. In contrast, two Bangladeshi soils with low concentrations of total As (
