Identification of Low Inorganic and Total Grain Arsenic Rice Cultivars

Jun 23, 2009 - Williams , P. N.; Islam , M. R.; Adomako , E. E.; Raab , A.; Hossain , S. A.; Zhu , Y. G.; Feldmann , J.; Meharg , A. A. Increase in ri...
0 downloads 0 Views 236KB Size
Download PDF
Environ. Sci. Technol. 2009, 43, 6070–6075

Identification of Low Inorganic and Total Grain Arsenic Rice Cultivars from Bangladesh G A R E T H J . N O R T O N , * ,† M. RAFIQUL ISLAM,‡ CLAIRE M. DEACON,† FANG-JIE ZHAO,§ JACQUELINE L. STROUD,§ STEVE P. MCGRATH,§ SHOFIQUL ISLAM,‡ M. JAHIRUDDIN,‡ JOERG FELDMANN,| ADAM H. PRICE,† AND ANDREW A. MEHARG† 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, Soil Science Department, Rothamsted Research, Harpenden, Hertfordshire AL5 2JQ, U.K., and College of Physical Sciences, Chemistry, University of Aberdeen, Meston Walk, Aberdeen AB24 3UE, U.K.

Received April 14, 2009. Revised manuscript received June 9, 2009. Accepted June 16, 2009.

For the world’s population, rice consumption is a major source of inorganic arsenic (As), a nonthreshold class 1 carcinogen. Reducing the amount of total and inorganic As within the rice grain would reduce the exposure risk. In this study, grain As was measured in 76 cultivars consisting of Bangladeshi landraces, improved Bangladesh Rice Research Institute (BRRI) cultivars, and parents of permanent mapping populations grown in two field sites in Bangladesh, Faridpur and Sonargaon, irrigated with As-contaminated tubewell water. Grain As ranged from 0.16 to 0.74 mg kg-1 at Faridpur and from 0.07 to 0.28 mg kg-1 at Sonargaon. Highly significant cultivar differences were detected and a significant correlation (r ) 0.802) in the grain As between the two field sites was observed, indicating stable genetic differences in As accumulation. The cultivars with the highest concentration of grain As were the Bangladeshi landraces. Landraces with red bran had significantly more grain As than the cultivars with brown bran. The percent of inorganic As decreased linearly with increasing total As, but genetic variation within this trend was identified. A number of local cultivars with low grain As were identified. Some tropical japonica cultivars with low grain As have the potential to be used in breeding programs and genetic studies aiming to identify genes which decrease grain As.

Introduction Inorganic arsenic (As) is a nonthreshold, class 1 carcinogen (1). It is found in rice grain, the dietary staple for ∼50% of * Corresponding author e-mail: [email protected]; tel: +44(0)1224272700. † Institute of Biological and Environmental Sciences, University of Aberdeen. ‡ Bangladesh Agricultural University. § Rothamsted Research. | College of Physical Sciences, University of Aberdeen. 6070

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009

the world, at concentrations that constitute a health concern in countries which rely on rice as a staple food (2-5). Household studies have shown that for Bangladesh and West Bengal, India, countries that both suffer from elevated As in drinking water, ∼50% of inorganic As (or total where the As was not speciated) came from rice ingestion (2, 6-8). Recently, rice has been identified as an important inorganic As exposure source in the Red River Delta, Vietnam, where groundwater has high As (9). Rice, specifically, is problematic with respect to As accumulation, as its anaerobic growth conditions mean that As is more available from the soil (10), leading to rice being much more efficient at assimilating As into grain than other cereal crops (11). As speciation in rice grain is dominated by inorganic As and dimethylarsinic acid (DMA) (12). Rice cultivars with high inorganic As are more of a risk to human health than rice with higher levels of DMA (13). Evidence suggests that inorganic As, as a percentage of total grain As, decreases as total grain As content increases (14), with some studies indicating that there may be a genetic component to As speciation (3, 13, 15). It has been demonstrated that there is a genetic component to total As in pot trial studies (12, 16), also the same phenomenon has been observed in field trials using a small number of genotypes (17). Genetic variation has also been observed in As tolerance within rice (18, 19). One way to reduce As exposure through rice would be to identify the extent of genetic variation in the concentrations of inorganic and total grain As in cultivars developed for specific agronomic conditions, such as the Bengal Delta. This would immediately identify which widely grown cultivars have low grain As, thus indicating cultivars which should be grown to produce grain with low As. These cultivars would also be suitable genetic stock for further breeding of lower As cultivars. Using two field sites in Bangladesh with different As contamination, we investigated if there is genetic variation in total grain As concentration across a wide spectrum of 76 rice cultivars, including local landraces and cultivars resulting from Bangladeshi breeding programs. A subset of these cultivars was then analyzed for As speciation within the grain to establish if speciation is controlled by environmental or genetic factors.

Methods Field Sites. A field trial was conducted, with rice germinated in December 2007 and harvested in May 2008, at two field locations in Bangladesh, in Faridpur (latitude 23°35.105′ and longitude 89°47.122′) and Sonargaon (latitude 23°40.086′ and longitude 90°36.37′). Both sites had a history of dry season (boro) irrigation with As contaminated water extracted from tubewells. For the Faridpur field, the tubewell was installed in 1990 and had an average soil As content, measured by taking 24 samples systematically across the trial area, of 29.6 ( 7.2 mg kg-1 and a tubewell water As concentration of 198 ( 31 µg L-1. For the Sonargaon field, the tubewell was installed in 2001 and had an average soil As content, measured by taking 24 samples systematically across the trial area, of 10.3 ( 2.2 mg kg-1 and a tubewell water As content of 331 ( 13 µg L-1 (Figure S1 in the Supporting Information). The field site in Faridpur was under continually flooded conditions, with irrigation every 3 days. The field site in Sonargaon was irrigated every 2 days, which resulted in alternative wet-dry cycles. Both field sites were fertilized with 70 kg of N/ha (split over three equal applications), 20 kg of P/ha, 50 kg of K/ha, 15 kg of S/ha, and 2 kg of Zn/ha. The field sites were also were sprayed with a single application of Furadan. 10.1021/es901121j CCC: $40.75

 2009 American Chemical Society

Published on Web 06/23/2009

Element concentrations and a description of the soils at both field sites are presented in Table S1 in the Supporting Information. Plant Material. A set of 76 cultivars (for a full list of genotypes, see Table S2 in the Supporting Information) consisting of Bangladesh Rice Research Institute (BRRI) improved cultivars (n ) 19), diverse cultivars previously used to generate permanent rice mapping populations (n ) 24), and local landraces (n ) 33) were used in this study. The local landraces were then subdivided into two groups: those which had grain with a brown bran (n ) 12) and those which had grain with a red bran (n ) 21). All of the nonlandrace cultivars had brown bran, except for one of the mapping population parents (Kasalath) which had red bran. Plants were sown in a randomized complete block design with three replicates. In each replicate, each genotype was planted in a single row of 2 m with 10 hills, each hill (one seedling) 20 cm apart and each row 20 cm apart. The central six plants were harvested and pooled together for As analysis, the yield is also expressed as the combined yield from these six plants. To separate the test genotypes, two hills of the check cultivar Minikit were planted at each end of the 10-hill test rows. Between every row of test genotypes, one row of Minikit was planted. Polishing Rice Grain. To remove the outer bran of the whole grain rice, 10 landrace cultivars (5 with red bran and 5 with brown bran) from the Faridpur field site were dehusked and then polished until 20% of the biomass had been removed. Local custom is that red rice is milled to white. This percentage loss of biomass was required to remove the red coloring of the red bran rice. To determine the percentage of As lost due to polishing for these cultivars, As in the polished grain was calculated and expressed as a percentage loss of As in the polished grain compared to the whole grain. Grain Sample Preparation and Analysis for Total As. For all digests, trace element grade reagents were used, and for quality control, replicates of certified reference material (rice flour [NIST 1568a]) were used; spikes and blanks were included. Rice grain samples were dehusked, and 0.2 g was weighed (either polished or unpolished grains) into 50 mL polyethylene centrifuge tubes. Samples were digested with concentrated HNO3 and H2O2 as described in ref 20. Briefly, total As analysis was performed by inductively coupled plasma mass spectrometry (ICP-MS, Agilent Technologies 7500). As standards with the appropriate ranges were made from 1000 mg L-1 ICP-MS grade As stock solution. All standards and blanks contained 10 µg L-1 indium as the internal standard and were made in 2% nitric acid. Analysis was performed as described in ref 20. As Speciation. A subset (n ) 25) of the test cultivars were selected for speciation, consisting of 8 local landrace cultivars, 6 BRRI cultivars, and 11 cultivars used as parents of mapping populations. The local landraces were divided between red (n ) 4) and brown (n ) 4) bran rice. The samples were dehusked and powdered, and 0.2 g was weighed into a 50 mL polyethylene centrifuge tube. Samples were extracted with 1% HNO3 as described in ref 20. As speciation of DMA and inorganic As was performed by anion exchange highperformance liquid chromatography (HPLC)-ICP-MS. Details of the protocol for As speciation analysis are given in ref 20.

Results For the determination of total As concentration in the grain, the recovery for the certified reference material (rice flour [NIST 1568a]) was 96%. For the speciation, the average inorganic As was 190 and 107 µg kg-1 for the total organic species (DMA and monomethyl arsonic acid), with an average extraction efficiency of 103%, which is comparable to previous studies using the same certified reference material (21).

Grain Total As Concentration. At the Faridpur field site, 72 of the 76 cultivars were successfully grown to harvest grain. The As in the grains ranged between 0.16 and 0.74 mg kg-1 with a mean concentration of 0.41 mg kg-1 (full list of As concentrations for the genotypes is presented in Table S2 in the Supporting Information). There was a significant genotypic effect for grain As concentration at this site (P < 0.001, F ) 5.57). The local landraces with red bran (n ) 21, mean As ) 0.57 mg kg-1, SD ) 0.08) had significantly more As (P < 0.001, F ) 25.6) than the local landraces with brown bran (n ) 12, mean As ) 0.42 mg kg-1, SD ) 0.08). The five genotypes with the highest concentration of grain As were Kazla boro, Chhola boro, Jagli, Rata boro, and Chandaina (all landraces with red bran), and the five with the lowest genotypes for grain As were Lemont, Azucena, Moroberekan, CT 9993-5-10-1-M, and BR 3 (all parents of mapping populations except BR3, which is an improved BRRI cultivar). At the Sonargaon field site, all 76 cultivars were successfully grown to harvest. The As concentration in the grains ranged between 0.07 and 0.28 mg kg-1 with a mean As concentration of 0.17 mg kg-1 (full list of As concentrations for the genotypes is presented in Table S2 in the Supporting Information). There was a significant genotypic effect for grain As concentration (P < 0.001, F ) 7.95) at this site. The local landraces which had red bran (n ) 21, mean As ) 0.25 mg kg-1, SD ) 0.03) had significantly more As (P < 0.001, F ) 56.57) than the local landraces with brown bran (n ) 12, mean As ) 0.16 mg kg-1, SD ) 0.034). The five genotypes with the highest concentration of grain As were Goa Bish, Kazla boro, Tulsi boro, Jagli boro, and Lafai (all landraces with red bran). The five with the lowest genotypes for grain As were Dumsia 81, Azucena, BRRI dhan 47, Lemont, and Moroberekan. BRRI dhan 47 is an improved BRRI cultivar; Dumsia 81 is a landrace with brown bran, and the other three genotypes are parents of mapping populations. While there was significantly more As in the grain of the rice plants grown at the Faridpur compared to Sonargaon, there was a highly significant correlation between the two data sets (Figure 1) (r ) 0.802, n ) 72, P < 0.001). When the cultivars were considered as four subgroups (BRRI, parents of mapping populations, landraces with brown bran, and landraces with red bran), there was a significant difference between the total grain As concentrations in these cultivars (Figure 2). Also, there was a field site by cultivar subgroup interaction. In both sites, the landraces with red bran had the highest grain As, and at the Faridpur site, the landraces with brown bran had grain As higher than the parents of the mapping populations and the BRRI cultivars. However, at the Sonargaon site, the landraces with brown bran were not significantly different in grain As than in the parents and the BRRI cultivars. The yield of the genotypes at the Faridpur site ranged between 8.0 and 117.6 g with a mean yield of 55.9 g. There was a significant correlation between the yield of the rice genotypes and the As concentration in the grain (r ) 0.753, n ) 72, P < 0.001) (Figure 3). The yield of the genotypes at the Sonargaon site ranged between 17.0 and 221.8 g with a mean of 119.1 g. Again, there was a significant correlation between the yield of the rice genotypes and the As concentration in the grain (r ) 0.741, n ) 76, P < 0.001) (Figure 3). There was no significant difference between the percentage of grain As lost when the bran was removed between red bran rice and brown bran rice (P ) 0.074). The As concentration of both types of rice was reduced by approximately 45% when the outer 20% of the biomass was removed. The removal of 20% of the biomass was required to remove the bran of the red bran rice, which is done before consumption of red bran rice. The removal of this amount of biomass reduced the As concentration by a much larger degree than VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6071

FIGURE 1. Correlation of total unpolished grain arsenic for each cultivar at both field sites in Bangladesh (r ) 0.802, n ) 72, P < 0.001).

FIGURE 2. Total unpolished grain arsenic levels in the four subgroups of rice cultivars. Data are presented as the mean ( standard error of the mean (SEM). in previous studies, where 7% of the biomass was removed and this only resulted in a 10% reduction in grain As (22). As Speciation. There was a significantly higher percentage of inorganic grain As at the Sonargaon site compared to the Faridpur site (P < 0.001, F ) 93.1; Figure 4A). A full list of the percentages of inorganic As and DMA of the selected genotypes is presented in Table S3 in the Supporting Information. At the Faridpur site, the inorganic As percentage ranged from 63 to 82% and there was a significant genotype effect for both percentage of inorganic As (P ) 0.014, F ) 2.20) and percentage of DMA (P ) 0.006, F ) 2.45). At the Sonargaon site, the inorganic As percentage ranged from 80 to 96% and there was a significant genotypic effect for percentage of inorganic As (P < 0.001, F ) 4.14) and percentage of DMA (P ) 0.009, F ) 2.28). For the parents of the mapping population cultivars, monomethyl arsenic acid (MMA) was detected in the samples; however, for the BRRI and landrace cultivars, the MMA peak was below the detection limit. At the Faridpur field site, MMA ranged from 2.1 to 3.4% and there was a significant genotype effect (P < 0.001, F ) 6.04). At the Sonargaon site, MMA ranged from 3.4 to 5.4%, and there was a genotype effect (P ) 0.006, F ) 3.28). When the data of both sites were analyzed together, there was a significant difference in percentage of inorganic grain 6072

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009

As between the subgroups of rice cultivars (P ) 0.019, F ) 3.70) and a significant subgroup by field site interaction (P < 0.001, F ) 11.9). There was a significantly higher percentage of DMA in the grain of the cultivars grown on the Faridpur site compared to the Sonargaon site (P < 0.001, F ) 104.12; Figure 4B). There was no significant difference between percentage of DMA in the grain within the subgroups of rice cultivars; however, there was a subgroup by field site interaction, which indicated that the subgroups showed different trends between both sites. When the data from both sites were combined, a significant negative correlation between the percentage of inorganic As and total As (r ) -0.483, n ) 25, P < 0.001) with a corresponding positive correlation between the percentage of DMA and total grain As was observed (r ) 0.572, n ) 25, P < 0.001) (Figure 5). However, the landraces with red bran deviate from this trend; analysis of variance on the reciprocal of the total grain As from the regression of best fit indicates that the landraces with red bran have significantly more percentage of inorganic As than predicted from the general regression model (P < 0.001, F ) 5.35). This is also visible in Figure 5 where all but one of the landraces with red bran are above the line of best fit for percentage of inorganic As.

FIGURE 3. Correlation of total unpolished grain arsenic and yield at the Faridpur and Sonargaon field sites in Bangladesh: Faridpur (r ) 0.753, n ) 72, P < 0.001); Sonargaon (r ) 0.741, n ) 76, P < 0.001).

FIGURE 4. Percentage of inorganic arsenic (A) and percentage of DMA (B) versus the study sites. Cultivars were broken down into subgroups of BRRI improved cultivars (n ) 6), landraces with red (n ) 4) and brown (n ) 4) grain, and cultivars used to generate mapping populations (n ) 11). Data are presented as the mean ( SEM.

Discussion Mean grain total As was approximately 2.3 times greater for Faridpur than for Sonargaon. This difference in mean grain total As concentration between the two sites is reflected in

the As concentration in the soil at the two sites; Faridpur has 2.9 times more total As in the soil compared to the Sonargaon site, whereas the tubewell water used for irrigation in the latter contained 67% more As than the former. There was a wide range in the concentrations of total grain As between the cultivars at each site, with a 4.6-fold range at the Faridpur site and a 4-fold range at the Sonargaon site. Importantly, there was a high degree of correlation between the two sites across all cultivars (Figure 1). This information, combined with the fact that there is a genotype effect for grain As concentration at both sites, shows that there is strong genetic control of grain As. This is in agreement with a previous field study looking into the variation of multiple elements within the rice grain, in which a 3-fold variation in As concentration between nine different japonica cultivars was observed (17). Genotypic variation in As concentration has also been detected in the roots and shoots of rice plants grown in hydroponics (23), as well as quantitative trait loci being detected for root, shoot, and grain As concentrations (16). It has previously been found that there was a genetic component to total As present in grain of different cultivars (12). The findings of the current study, comparing BRRI bred rice versus local landraces, also found a difference based on genetic background, with the improved BRRI cultivars assimilating less As than landraces. The BRRI cultivars used in this study were developed from a wide range of stock from abroad (Table S4 in the Supporting Information). Only the local landrace Latishail (which is not included in this study) was used as parent material for developing the BR 3 variety. Therefore, the BRRI cultivars had total As and percentage of DMA similar to those of the parent population. Considering the concentrations of total As in the whole grain, the local landraces could be split into two distinct groups, those grains that have red bran having high As and those with brown bran having low As. The reason for this difference is presently unclear. When the grain was polished (to a degree which removed the color of the bran), the same percentage of grain As was lost between the cultivars with red and brown bran. This indicates that the concentration difference of the red and brown bran cultivars is not limited to just the bran. In this study, a number of interesting genotypes have been identified that could potentially be used in breeding programs. The most notable of these are the cultivars VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6073

FIGURE 5. Correlation of percentage of DMA, inorganic arsenic grain arsenic, and total grain arsenic at both field sites: DMA (r ) 0.572, n ) 25, P < 0.001); inorganic As (r ) -0.483, n ) 25, P < 0.001). Azucena, Lemont, and Moroberekan. These cultivars (all tropical japonicas (24)) were all ranked among the lowest five genotypes in both sites. While these cultivars are unadapted to the growing conditions at these sites (all had low yields), they could be exploited as donors of genes to produce plants with low grain As concentrations. Within genotypes adapted to growing conditions in Bangladesh, two of the BRRI cultivars were identified as having low grain As at both sites. The BRRI cultivar BR 3 had the 5th lowest grain As at Faridpur and the 8th lowest at Sonargaon although in both sites its yield was not good. BRRI dhan 47 was the 4th lowest grain As at the Sonargaon site and the 16th lowest at the Faridpur site and had good yield in both sites. These improved cultivars, especially BRRI dhan 47, appear to be ideal genotypes to be used currently in As contaminated regions in Bangladesh. When the total As concentrations in the grains and the yield of the plants were compared, a strong positive correlation was observed. In general, cultivars which had a larger yield also had higher grain As. The consequences of this strong positive correlation between yield and grain As could be very detrimental to human health, as cultivars which produce higher grain yields are more likely to be grown and, therefore, grain total As in the rice for human consumption will be high. There is an urgent need to study this relationship further since some of the reasons, which are currently unknown but theoretically numerous (e.g., local adaptation to the soil environment leading to high As tolerance and high As uptake or differences in flowering or grain filling time as it relates to fluctuating soil As), may have major consequences for efforts to breed low grain arsenic. In whole grain rice, the inorganic As and DMA content are correlated with grain total As concentrations in opposite patterns (Figure 5), with a positive linear relationship between percentage DMA and total grain As, and a negative correlation with percentage inorganic As content vs total grain As. Thus, the higher the As content in the grain, the greater the organic proportion. It is possible that methylation of As, as a mechanism of detoxification, is enhanced by increased As accumulation in rice. A recent study identified a similar trend in polished (white) market and whole grain (mixture of market, field and pot experiments) rice, though noting that polished rice had ∼10% lower As content than whole grain rice (14). The lower total As content in white rice compared to whole grain rice and the percentage contribution of inorganic As to that total was found as arsenite and, in 6074

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009

particular, was elevated in bran (14). All of these factors need to be taken into account when considering As speciation in rice, particularly as it is white rice that is consumed by most of the world’s population. Despite the general observation that the proportion of inorganic As is regulated largely by the total As in the grain, our results clearly showed a genetic component regulating As speciation since deviation from the linear relationship was detected in red bran rice. Red rice had less DMA than would have been predicted by their total As concentration; therefore, this observation does not look promising for breeding low inorganic As rice, as an increase in the proportion of DMA to total As is desirable, as DMA is less toxic to humans than inorganic As. Interestingly, one cultivar, CT 9993-5-10-1-M, at the Faridpur site did have a higher percentage of DMA than expected compared to the grain total As concentration. However, further investigation of this genotype across more field environments with high concentrations of As is needed to confirm this observation. In summary, there was a wide range of genetic variation for total whole grain As, and the detected variation was consistent between the two field sites. There were also higher concentrations of grain As in the landraces compared to the improved BRRI cultivars; the landraces could be split into two categories, plants that had red bran having grain As higher than those with brown bran. In addition, there was a strong relationship between total grain concentration and percentage of DMA, although a degree of genetic variation was also observed. Ultimately, understanding why these differences occur in total grain and inorganic As content at a genetic level should enable enhanced selection for low total and inorganic As grain content cultivars, through marker assisted breeding or genetic modification.

Acknowledgments This work was funded by BBSRC-DFID Grant BBF0041841. The authors would like to thank The International Rice Research Institute for providing the seeds of the parents of the mapping population and the Bangladesh Rice Research Institute for providing the seeds of BRRI varieties and landraces of boro rice.

Supporting Information Available Information on the soil properties at the two field sites and a list of genotypes used in the study is available. The parents of the BRRI cultivars are also presented in this information.

This material is available free of charge via the Internet at http://pubs.acs.org.

(12)

Literature Cited (1) National Research Council. Arsenic in drinking water-2001 Update; National Academy Press: Washington, D.C., 2001. (2) Mondal, D.; Polya, D. A. Rice is a major exposure route for arsenic in Chakdaha block, Nadia district, West Bengal, India: A probabilistic risk assessment. Appl. Geochem. 2008, 23, 2986– 2997. (3) Meharg, A. A.; Williams, P. N.; Adomako, E.; Lawgali, Y. Y.; Deacon, C.; Villada, A.; Cambell, R. C. J.; Sun, G.; Zhu, Y. G.; Feldmann, J.; Raab, A.; Zhao, F. J.; Islam, M. R.; Hossain, S.; Yanai, J. Geographical Variation in Total and Inorganic Arsenic Content of Polished (White) Rice. Environ. Sci. Technol. 2009, 43, 1612–1617. (4) Rahman, M. M.; Owens, G.; Naidu, R. Arsenic levels in rice grain and assessment of daily dietary intake of As from rice in Ascontaminated regions of Bangladesh. Environ. Geochem. Health 2009, 1–8. (5) Signes-Pastor, A. J.; Mitra, K.; Sarkhel, S.; Hobbes, M.; Burlo´, F.; De Groot, W. T.; Carbonell-Barrachina, A. A. Arsenic speciation in food and estimation of the dietary intake of inorganic arsenic in a rural village of West Bengal, India. J. Agric. Food Chem. 2009, 56, 9469–9474. (6) Williams, P. N.; Islam, M. R.; Adomako, E. E.; Raab, A.; Hossain, S. A.; Zhu, Y. G.; Feldmann, J.; Meharg, A. A. Increase in rice grain arsenic for regions of Bangladesh irrigating paddies with elevated arsenic in groundwaters. Environ. Sci. Technol. 2006, 40, 4903–4908. (7) Ohno, K.; Yanase, T.; Matsuo, Y.; Kimura, T.; Hamidur Rahman, M.; Magara, Y.; Matsui, Y. Arsenic intake via water and food by a population living in an arsenic-affected area of Bangladesh. Sci. Total Environ. 2007, 381, 68–76. (8) Kile, M. L.; Houseman, E. A.; Breton, C. V.; Smith, T.; Quamruzzaman, Q.; Rahman, M.; Mahiuddin, G.; Christiani, D. C. Dietary arsenic exposure in Bangladesh. Environ. Health Perspect. 2007, 115, 889–893. (9) Agusa, T.; Kunito, T.; Minh, T. B.; Kim Trang, P. T.; Iwata, H.; Viet, P. H.; Tanabe, S. Relationship of urinary arsenic metabolites to intake estimates in residents of the Red River Delta, Vietnam. Environ. Pollut. 2009, 157, 396–403. (10) Xu, X. Y.; McGrath, S. P.; Meharg, A. A.; Zhao, F. J. Growing rice aerobically markedly decreases arsenic accumulation. Environ. Sci. Technol. 2008, 42, 5574–5579. (11) Williams, P. N.; Villada, A.; Deacon, C.; Raab, A.; Figuerola, J.; Green, A. J.; Feldmann, J.; Meharg, A. A. Greatly enhanced arsenic shoot assimilation in rice leads to elevated grain levels compared

(13) (14)

(15)

(16) (17)

(18) (19)

(20)

(21) (22)

(23) (24)

to wheat and barley. Environ. Sci. Technol. 2007, 41, 68546859. Williams, P. N.; Price, A. H.; Raab, A.; Hossain, S. A.; Feldmann, J.; Meharg, A. A. Variation in arsenic speciation and concentration in paddy rice related to dietary exposure. Environ. Sci. Technol. 2005, 39, 5531–5540. Zavala, Y. J.; Gerads, S.; Gu ¨ rleyu ¨ k, H.; Duxbury, J. M. Arsenic in rice: II. Arsenic speciation in USA grain and implications for human health. Environ. Sci. Technol. 2008, 42, 3861–3866. Meharg, A. A.; Lombi, E.; Williams, P. N.; Scheckel, K. G.; Feldmann, J.; Raab, A.; Zhu, Y.; Islam, R. Speciation and localization of arsenic in white and brown rice grains. Environ. Sci. Technol. 2008, 42, 1051–1057. Li, R. Y.; Stroud, J. L.; Ma, J. F.; McGrath, S. P.; Zhao, F. J. Mitigation of arsenic accumulation in rice with water management and silicon fertilization. Environ. Sci. Technol. 2009, 43, 3778–3783. Zhang, J.; Duan, G. Genotypic difference in arsenic and cadmium accumulation by rice seedlings grown in hydroponics. J. Plant Nutr. 2008, 3, 2168–2182. Cheng, W. D.; Zhang, G. P.; Yao, H. G.; Wu, W.; Xu, M. Genotypic and environmental variation in cadmium, chromium, arsenic, nickel, and lead concentrations in rice grains. J. Zhejiang Univ. Sci. B 2006, 7, 565–571. Dasgupta, T.; Hossain, S. A.; Meharg, A. A.; Price, A. H. An arsenate tolerance gene on chromosome 6 of rice. New Phytol. 2004, 163, 45–49. Norton, G. J.; Nigar, M.; Williams, P. N.; Dasgupta, T.; Meharg, A. A.; Price, A. H. Rice-arsenate interactions in hydroponics: A three-gene model for tolerance. J. Exp. Bot. 2008, 59, 2277– 2284. Sun, G.; Williams, P. N.; Zhu, Y.; Deacon, C.; Carey, A.; Raab, A.; Feldmann, J.; Meharg, A. A. Survey of arsenic and its speciation in rice products such as breakfast cereals, rice crackers and Japanese rice condiments. Environ. Int. 2009, 35, 473–475. Raab, A.; Baskaran, C.; Feldmann, J.; Meharg, A. A. Cooking rice in a high water to rice ratio reduces inorganic arsenic content. J. Environ. Monit. 2009, 11, 41–44. Sun, G. X.; Williams, P. N.; Carey, A. M.; Zhu, Y. G.; Deacon, C.; Raab, A.; Feldmann, J.; Islam, R. M.; Meharg, A. A. Inorganic arsenic in rice bran and its products are an order of magnitude higher than in bulk grain. Environ. Sci. Technol. 2008, 42, 7542– 7546. Zhang, J.; Zhu, Y.; Zeng, D.; Cheng, W.; Qian, Q.; Duan, G. Mapping quantitative trait loci associated with arsenic accumulation in rice (Oryza sativa). New Phytol. 2008, 177, 350–355. Garris, A. J.; Tai, T. H.; Coburn, J.; Kresovich, S.; McCouch, S. Genetic structure and diversity in Oryza sativa L. Genetics 2005, 169, 1631–1638.

ES901121J

VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6075