Environ. Sci. Technol. 2003, 37, 229-234
Arsenic Contamination of Bangladesh Paddy Field Soils: Implications for Rice Contribution to Arsenic Consumption ANDREW A. MEHARG* AND MD. MAZIBUR RAHMAN School of Biological Sciences, University of Aberdeen, Cruickshank Building, St. Machar Drive, Aberdeen, AB24 3UU, UK
Arsenic contaminated groundwater is used extensively in Bangladesh to irrigate the staple food of the region, paddy rice (Oryza sativa L.). To determine if this irrigation has led to a buildup of arsenic levels in paddy fields, and the consequences for arsenic exposure through rice ingestion, a survey of arsenic levels in paddy soils and rice grain was undertaken. Survey of paddy soils throughout Bangladesh showed that arsenic levels were elevated in zones where arsenic in groundwater used for irrigation was high, and where these tube-wells have been in operation for the longest period of time. Regression of soil arsenic levels with tube-well age was significant. Arsenic levels reached 46 µg g-1 dry weight in the most affected zone, compared to levels below 10 µg g-1 in areas with low levels of arsenic in the groundwater. Arsenic levels in rice grain from an area of Bangladesh with low levels of arsenic in groundwaters and in paddy soils showed that levels were typical of other regions of the world. Modeling determined, even these typical grain arsenic levels contributed considerably to arsenic ingestion when drinking water contained the elevated quantity of 0.1 mg L-1. Arsenic levels in rice can be further elevated in rice growing on arsenic contaminated soils, potentially greatly increasing arsenic exposure of the Bangladesh population. Rice grain grown in the regions where arsenic is building up in the soil had high arsenic concentrations, with three rice grain samples having levels above 1.7 µg g-1.
Introduction The digging of tube-wells for drinking water supply into aquifers elevated in arsenic in Bangladesh and West Bengal has been described as the greatest mass poisoning in human history (1), with 36 million people exposed to elevated arsenic in their drinking water (2). Arsenic groundwater levels in some areas of Bangladesh groundwater arsenic concentrations approaches 2 mg L-1 (3, 4). It is predicted that 200 000270 000 people will die of cancer from drinking arsenic contaminated drinking water in Bangladesh alone (5). Arsenic contaminated groundwater is not just used for drinking water but is also widely used for irrigation of crops, and particularly for the staple food paddy rice (Oryza sativa L.), which provides 73% of calorific intake for Bangladeshi’s (6). Groundwater is used extensively to irrigate rice crops in * Corresponding author phone: ++44 (0)1224 272264; fax: ++44 (0)1224 272703; e-mail:
[email protected]. 10.1021/es0259842 CCC: $25.00 Published on Web 11/20/2002
2003 American Chemical Society
Bangladesh, particularly during the dry season with 75% of the total cropped area given over to rice cultivation and 83% of the total irrigated area used for rice cultivation (7). Background levels of arsenic in soils from limited surveys conducted in Bangladesh rice paddy fields range from 4 to 8 mg As kg-1 (8, 9). In areas with elevated arsenic in the groundwater irrigated with contaminated water, soil level can reach up to 83 µg g-1 arsenic according to one report (9) and up to 57 µg g-1 arsenic in another survey (8). If arsenic levels building up in paddy soils leads to elevated arsenic in rice grain, then the amount of arsenic ingested by the inhabitants of this region could be considerably more than previously thought. There has been considerable investigation into drinking water contamination in Bangladesh/West Bengal, with increasing numbers of epidemiological studies. However, to date, no studies have been published which consider other potential arsenic exposure routes to these populations (10). Food surveys on the daily arsenic intake in the United States and Europe (11-13) showed, second to fish products, rice is a major dietary source of arsenic. In countries that have a rice subsistence diet, the importance of dietary exposure to arsenic through rice could be considerable, as shown by the study of Schoof et al. (14). Levels of arsenic in rice grain are typically 0.05-0.4 µg g-1 for North America, Europe, and Taiwan (11-14). Arsenic levels in rice grain reached 0.7 µg g-1 in rice grown on paddy soils containing 68 µg g-1 arsenic in China (15), showing the potential for arsenic contamination of rice grain from contaminated paddy soils. The problem of arsenic contaminated groundwaters is not just restricted to Bangladesh/West Bengal. Other regions in SE Asia such as China, Vietnam, Thailand, and Taiwan have been found to have high levels of arsenic in groundwaters and soil (2). Paddy rice is also the staple food for these regions. This study presents a survey of arsenic levels in paddy soils collected over an extensive area of Bangladesh. Contamination of soil was related to tube-well arsenic levels, depth, and age to understand the mechanism of soil contamination by arsenic. Arsenic levels in Bangladesh produced rice, were determined to calculate baseline exposure of the population to rice derived arsenic. Dietary arsenic exposure was then modeled using these data.
Materials and Methods All chemicals used in the sample preparation and analysis were obtained from BDH (Poole, Dorset, UK) and were of analytical (Analar) grade or better. Collection of Soil and Rice Grain Samples. Soil and rice samples were collected during the period of JanuaryFebruary, 2001. Soil samples were collected from 27 administrative districts of Bangladesh, and a total of 71 samples were obtained. The location of these districts is shown in Figure 2d. Surface soils (0-15 cm) were obtained from paddy field plots of 1-5 ha, with three samples collected from each field combined to give a composite sample. Samples were air-dried and 2 mm sieved. Ripe rice grains were collected from nine Bangladesh rice varieties from field trials conducted by the Bangladesh Rice Research Institute (BBRI), Gazipur District, Bangladesh. Further samples of known rice varieties were collected from field trials in western Bangladesh. Analysis. Soils were oven dried (70 °C) before analysis. Husks were removed from the rice grains and grain oven dried (70 °C). Soil and grain samples were then finely ground VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
229
The tubes were then placed on a heating block, and the temperature was raised to 60 °C. After adding 3 mL of 30% hydrogen peroxide, the temperature was gradually raised to 120 °C, and the samples were allowed to digest for 3 h. The digests were then cooled, diluted in deionized water and made up to 50 mL and then filtered through a Whatman no. 42 filter paper. Arsenic in soil digests was quantified by graphite-furnace atomic absorption spectrometry using a Perkin-Elmer 3300 instrument. Arsenic content of grain digests was determined by flow-injection hydride generation atomic absorption spectrometer using a Perkin-Elmer AAnalyst 300 interfaced with the FIAS 100 hydride generator. Hydride generation was used for grain samples due to its lower limits of detection (0.4 µg L-1). Prior to analysis by hydride generation, digests were evaporated to dryness and then dissolved in 10% HCl containing 1% KI and 0.5% ascorbic acid to reduce As(V) to As(III).
FIGURE 1. Geographical distribution of arsenic levels in individual Bangladesh paddy soils. Grey dotted lines on the map represent district borders. in a ball mill. Subsamples were weighed (0.1-0.2 g) into quartz glass digestion tubes and then digested using a nitric acid-hydrogen peroxide procedure. Five milliliters of nitric acid was added to each tube and allowed to stand overnight.
All instruments were calibrated using matrix-matched standards. In each analytical batch at least, two reagent blanks, one internationally certified reference material, one spike, and five duplicate samples were included in the acid digests to assess precision and accuracy of the chemical analysis. Certified reference material GBWO7405 (soil) and GBWO7604 (poplar leaf) were obtained through the U.K. Laboratory of Government Chemists. Tube-Well Data. Tube-well information (year constructed, depth, arsenic levels) were obtained from the British Geological Survey (BGS) web-site (3) where they have placed the information to be used by other researchers. We analyzed our soils data by calculating average soil concentrations in each district. The BGS tube-well data was similarly averaged per district for comparison with the soils data.
FIGURE 2. Geographical distribution of arsenic levels in groundwaters (a), tube-well depth (b), tube-well age (c), and arsenic levels in paddy soils (d) averaged per administrative district. The data for tube-well arsenic, well depth, and well age were obtained from the BGS(3). The position of each point is the average longitude and latitude grid references from the BGS data. 230
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003
Statistics. All statistics were conducted using Minitab v.13 (State College, PA).
Results and Discussion All samples were above instrumental limits of detection. Recovery of arsenic from the poplar leaf reference material was 78% (n ) 5) and for the soil was 88% (n ) 8). The presented data have not been corrected for these recoveries. Arsenic in Paddy Soils. Arsenic levels in the 0-15 cm surface paddy soils varied between 3.1 and 42.5 µg g-1 in our survey (Figure 1, Table 1). These are in agreement with more geographically limited surveys of Bangladesh soils. Alam and Sattar (8) found that arsenic levels in surface soils (0-15 cm, the same as the soils reported here) ranged from not detectable to 31.8 µg g-1 and that arsenic levels in soil were higher in the 15-30 cm soil, with levels reaching 56 µg g-1. They found that soil arsenic levels were correlated with local well water concentrations, suggesting that the soils had become contaminated through irrigation with arsenic contaminated water. Ullah (9) reports, also in 0-15 cm surface soil, levels up to 83 µg g-1 arsenic in another survey of Bangladesh soils and suggests that these soils were contaminated through irrigation with arsenic contaminated groundwater. The highest levels of arsenic in the survey reported here were in districts to the west of the country, particularly the districts of Meherpur, Chuadanga, Kushtia, and Rajbari (Table 1), all having one or more soil samples with >30 µg g-1 (Figure 1). Soils in the range of 20-30 µg g-1 arsenic are scattered throughout the central belt of Bangladesh, mostly associated with the high zones of groundwater arsenic contamination (Figure 2a). The groundwater contamination by arsenic in Bangladesh has been extensively investigated, with thousands of wells tested (3). Data from this survey has been presented as district averages in Figure 2. The number of wells tested per district ranged from 15 to 110. The results of our soil survey have been presented alongside these data for comparative purposes. The soils collected from the Bangladesh border with West Bengal are the most elevated and are associated with an area which has >50 µg L-1 of arsenic in the groundwater. The zone of high arsenic in groundwaters basically covers the bottom half of central and western Bangladesh (Figure 2a). The results presented here show that low levels of arsenic in soils can be found in zones with high levels in the groundwaters, such as central Bangladesh (Figure 2). It could be that agronomic practices differ with respect to the volume of irrigation water used in different regions, leading to differential contamination with arsenic of paddy soils. Also, the age and depth of the tube-wells need to be considered as arsenic will accumulate in the soil with increased time period of irrigation with contaminated waters, and the depth of the tube-wells is related to arsenic concentrations in the well water (3). The map of tube-well age (Figure 2b) showed that in the areas with high soil arsenic (Figure 2d), average year of tubewell construction was early. Thus, this zone would have had soils irrigated with contaminated arsenic for a relatively long period of time. Linear regression of tube-well age against paddy soil arsenic levels (Figure 2) was significant (P ) 0.048). Similar regressions with tube-well depth (P ) 0.505) and tube-well arsenic levels (P ) 0.684) were not significant (regressions not shown). The regression with age shows scatter, but this is not surprising given the range of factors that may contribute to paddy soil contamination and that the data were for district averages, with some districts being represented by only a few soil samples (see Table 1). The fact that tube-well depth and arsenic concentrations in the well water were not significantly correlated with soil arsenic levels may be due to a number of reasons. Highest
arsenic levels are found in the shallow wells (3), yet the districts with the shallowest wells on average are in the region of the country (the northeast corner) with the lowest well arsenic levels. This may be because the wells are not deep enough to reach the reductive zone where mobile arsenic is generated (16), or it may simply be due to the fact that the aquifer chemistry differs compared to other regions of the country. Deep wells (>250 m) also have low arsenic levels, but these are generally restricted to the coastal districts (Figure 2b). This complex geographical relation between well depth and well water arsenic levels is possibly a reason soil arsenic did not correlate well with well depth. Can the amount of water added to paddy field during rice irrigation lead to substantial increases in soil arsenic levels? For lowland rice production in Asia, typically 150-200 mm of water are required to prepare land for rice planting, but this value can rise to 900 mm (17). Water added to the crop during growth may vary between 500 and 3000 mm. Rice in Bangladesh is grown during the dry season (Boro) and wet season (Aman). During dry season rice production this water comes from rivers or tube-wells. In the wet season, there is much less demand on groundwater or surface water resources, with up to 225 mm of water from these sources required to supplement rainwater, though this figure is probably higher as it does not take account of seepage loses (18). Using a conservative estimate of 1000 mm of irrigation water added to a paddy field per annum, this equates to 1000 l added to 1 m2 of paddy field. If the field is solely irrigated from tube-wells, and that water contains 0.1 mg L-1 arsenic, 100 mg will be added per m2 per annum. Assuming infiltration of that 100 mg to 10 cm and a bulk density of 1, soil arsenic levels will rise by 1 µg g-1 per annum. The oldest average tube-well age recorded was 1986 in a region with average tube-well water between 0.05 and 0.1 mg L-1 (Figure 2). Our soil sampling was conducted in 2001, with the conservative estimate of 1000 mm irrigation water per annum, arsenic levels could have risen by at least 7.5-15 µg g-1 on average. Older tube-wells, or higher levels in the irrigation water (levels greater than 1 mg L-1 have been recorded (3)), or higher levels of groundwater application would considerably increase this figure. Potential losses of arsenic from paddy field system must be considered. Arsenic can be volatilized as arsines from soils, though at low rates (19). Irrigation water can percolate to depth, by-passing surface soils although the soil will effectively act as an exchange column for arsenic, and water would have to be channeled through large cracks for the arsenic not to be removed. Irrigation water could also be lost through cracks in the paddy field dams, escaping to surface waters. However, as water is a valuable resource, water loss is closely managed. An alternative hypothesis to explain the pattern of arsenic in the soils is that soil arsenic may be derived, at least in part, from a geogenic origin, leading to differential levels of arsenic in soils derived from different parent materials. As the soils are sediment derived, there is a possibility that the arsenic enriched sediments which give rise to the groundwater contamination also outcrop on the surface, providing the substrate from which soils are formed. Nickson et al. (16) showed that arsenic in aquifer sediment ranged from 5 to 30 µg g-1, not dissimilar, but lower, than the range that we observed for surface soil samples (Table 1). Acharyya et al. (20) quote a range of 9.5-12 µg g-1 for clay rich aquifer sediments from West Bengal, and 3.8-4.8 µg g-1 for sands. Chowdhury et al. (21) conducted an extensive sediment survey of 2235 sediment samples from 112 boreholes in West Bengal, sampled at 3-6 m depth intervals. Only 85 of their samples were above the detection limit of 10 µg g-1, 3.8% of the total measured. Given that 8.5% of our samples were above 30 µg g-1 (Table 1), it can be argued that the elevated soil arsenic reported here and by others (8, 9) was not VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
231
TABLE 1. Arsenic Levels in Individual Paddy Fields in Bangladesh, Listed by Administrative Divisions and Districts division
soil series
district
As content (µg g-1)
Barisal Barisal Barisal Barisal Barisal Chittagong Chittagong Chittagong Chittagong Chittagong Chittagong Chittagong Chittagong Chittagong Chittagong Chittagong Chittagong Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Dhaka Khulna Khulna Khulna Khulna Khulna Khulna Khulna Khulna Khulna Khulna Khulna Khulna Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi Rajshahi
Bagalkati Barisal Barisal Nilkamal Ramgati Burichang Chandina Debidwar Mirsharai Pahartali Rawjan Burichang Chandina Chandina Debidwar Debidwar Tippera Belabo Tejgaon Ghatail Shilmondi Sonatala Kendua Jadurpara Lokdeo Melandaha Shilmondi Shilmondi Sonatala Sonatala Sonatala Sotiakhali Tarakanda Chandra Khilgaon Melandaha Shilmondi Sonatala Ghior Ishurdi Dumuria Ganges silt Ghior Ishurdi Golapur Golapur Sara Sara Golapur Sara Gangachhara Jamun Lauta Lauta Nijhuri Nijhuri Baliadangi Tista silt Bhabanipur Gopalpur Gopalpur Sara Atwari Ruhia Gopalpur Amnura Chandra Gangachhara Polashbari Baliadangi Pirgachha
Bagalkati Barisal Barisal Bhola Bhola Chandpur Chandpur Chandpur Chittagong Chittagong Chittagong Comilla Comilla Comilla Comilla Comilla Laksham Gazipur Gazipur Jamalpur Jamalpur Jamalpur Kishoregonj Manikgonj Mymensingh Mymensingh Mymensingh Mymensingh Mymensingh Mymensingh Mymensingh Mymensingh Mymensingh Tangail Tangail Tangail Tangail Tangail Chuadanga Chuadanga Jessore Jessore Kushtia Kushtia Meherpur Meherpur Meherpur Meherpur Rajbari Rajbari Bogra Bogra Bogra Bogra Bogra Bogra Dinajpur Kurigram Naogaon Naogaon Nawabgonj Nawabgonj Panchagarh Panchagarh Pbna Rajshahi Rangpur Rangpur Rangpur Thakurgaon Thakurgaon
24.5 19.3 26.1 16.3 16.8 18.4 6.8 7.3 6.5 7 8.6 21.6 9.2 9.9 3.1 11 7.4 14.6 10.9 16.5 16.4 13.4 9 13.5 8.2 6 17.1 17.3 9.8 12.8 25.4 15.8 9.3 5.7 24 9.6 19.2 12.8 29 33.3 21 23.3 42.5 34.2 30.2 36.6 10.8 15.7 30.7 23.9 15.5 9.3 4.9 5.1 5.3 6.3 11.7 9.6 24.3 26.7 20.9 15.7 8.1 9.6 14.4 7.8 6.5 11.5 7.6 9.1 12.4
232
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003
TABLE 2. Arsenic Levels in Different Bangladesh Rice Varieties Grown by the Bangladesh, Rice Research Institute (BBRI), Gazipur, Bangladesh. rice variety
grain arsenic (µg g-1)
BR14 BR11 BR23 BR24 BR28 BR29 BR31 BR32 BR33
0.073 0.043 0.056 0.089 0.067 0.18 0.092 0.044 0.206
TABLE 3. Arsenic Levels in Rice Varieties Collected from Different Districts of Bangladesh district
variety
grain arsenic (µg g-1)
Bogra Bogra Bogra Bogra Dinajapur Naogaon Nawabgonj Nawabgonj Mymensingh Rangpur Rajshahi Rajshahi Rajshahi
khatobada pajam swarna BR11 BR11 BR11 BR11 BR11 BR8 BR11 gocha kalia swarna
0.058 0.082 0.096 0.104 0.203 1.835 1.747 1.775 0.078 0.185 0.075 0.117 0.096
primarily of due to the geogenic origin of the soils. The frequency of elevated arsenic levels was much greater in soils than in sediment, and generally the soil levels were higher than these sediment concentrations. Other anthropogenic sources need to be considered, such as fertilization of the fields with human wastes. In areas of high arsenic in the drinking water, human urine, and feces will also be elevated in arsenic. Considering all these factors, there is evidence from the current study, and the study of Alam and Sattar (8) and Ullah (9) that arsenic levels have been raised through irrigation of paddy soils with arsenic contaminated groundwaters, at least in some regions of Bangladesh. Further detailed studies on mass balances of arsenic in paddy field systems will reveal the extent and mechanisms of arsenic buildup in Bangladesh paddy fields. Arsenic Levels in Rice. The consequences for elevated arsenic in paddy soils must be considered. Samples collected from Gazipur District at the Bangladesh Rice Research Institute had an average level of 0.092 µg g-1 dry wt (average of 11 different rice varieties), with the highest concentration being 0.21 µg g-1 (Table 2). The soils of Gazipur were not highly contaminated with arsenic and this district well represented in the soil survey (Table 1). Arsenic levels in rice ranged from 0.03 to 0.11 µg g-1 in North American cooked rice (N ) 18) (13) and in another study on raw rice this range was 0.2-0.46 µg g-1 with a mean of 0.303 µg g-1 wet weight (N ) 4) (12). For Taiwan values ranged between 0.063 and 0.2 µg g-1 dry weight (N ) 4) (14). The levels found for Gazipur were in keeping with these previous studies (11-14). Rice grain collected in western Bangladesh Districts had arsenic levels ranging from 0.058 to 1.83 µg g-1. The arsenic levels found in Bogra, Dinajapur, Mymensingh, Rangpur, and Rajshai had arsenic levels within the same range as the field trials at Gazipur (Table 3). All these districts had low levels of soil arsenic in the survey reported here (Table 1) The three grain samples collected from Nawabgonj and
FIGURE 3. Regression of average paddy soil arsenic level verses average age of tube-well per district for the data presented graphically in Figure 2. Naogoan had levels of arsenic ranging between 1.75 and 1.83 µg g-1, an order of magnitude higher than the samples from other districts. The districts of Nawabgonj and Naogoan had high levels of arsenic in paddy soils (Table 1). Reports on the arsenic grain content of rice growing on contaminated soils are limited. Two glasshouse studies by Abedin et al. (22, 23) show that arsenic levels in grain can reach 0.7 µg g-1 dry wt. The highest concentrations were reached when irrigating rice with 8 mg L-1 arsenic (as arsenate), giving a final soil arsenic concentration on harvest of 210 µg g-1. However, when plants were irrigated with 1 mg L-1 arsenic, grain levels rose to 0.47 µg g-1 compared to 0.33 µg g-1 in the non-arsenic treated controls (23). This grain arsenic rose to 0.51 µg g-1 when irrigated with 2 mg L-1 arsenic. The final soil concentrations on harvest were 34.5, 56.6, and 70.8 µg g-1 for 0, 1, and 2 mg L-1 arsenic treatments. Groundwaters can reach and exceed 1 µg L-1 arsenic in Bangladesh (3), and arsenic levels in Bangladesh soils from this study and those of Alam and Sattar (8) and Ullah (9) are within this range used in the experiments of Abedin et al. (23). These glasshouse trials show the potential for rice grain becoming elevated in arsenic under field conditions. Field trials by Xie and Huang (15) on Chinese arsenic polluted paddy soils showed that rice could accumulate up to 0.725 µg g-1 dry wt arsenic when grown on soils containing 68 mg kg-1 arsenic. Again, the level of soil contamination Xie and Huang’s (15) is within the range reported for Bangladesh soils, with a maximum reported value 83 µg g-1 (Ullah).9 A Taiwanese field survey showed that rice grain grown on paddy soils containing 6.9-7.5 µg g-1 of arsenic had an arsenic concentration of 0.2 µg g-1 dry weight (14). The arsenic levels reported here for Nawabgonj and Naogaon, to the authors knowledge, are the highest recorded to date in scientific literature for rice grain. The sample size was limited, and there is a clear imperative for a more detailed study of arsenic levels in Bangladesh rice grain, focusing on the contaminated districts in the west of the country, to establish the extent of this problem. Second to fish, rice is the major dietary source of arsenic to the general public in North America and Europe (11-13), accounting for 34% of arsenic exposure in babies, the group that eat the most rice (13). In Bangladesh, where rice is the subsistence food, the contribution of arsenic to dietary exposure, including contaminated drinking water, is considerable. The calculated daily human intake of arsenic from rice has been modeled for Bangladesh (Figure 4). With a drinking water intake of 0.1 mg L-1 (which is considered highly elevated and a major health threat), arsenic intake from rice will account for 17.3 and 29.6% of arsenic consumption if rice contained 0.1 and 0.2 µg g-1 of arsenic,
FIGURE 4. Percentage contribution of rice to daily arsenic intake. The percentage daily As intake was calculated using a rice consumption rate of 0.42 kg d-1 (25) and that 2 L of groundwater were consumed per day (26). Other potential dietary sources of arsenic were ignored for this calculation. The solid line is for 0.01 mg L-1, dotted line for 0.1 mg L-1, and dashed line for 1 mg L-1 arsenic in drinking water. The equation used to calculate these curves was as follows: 100*(level of arsenic in grain * 0.42 kg-1/ (level of arsenic in grain * 0.42 kg-1 + level of arsenic in drinking water * 2 L)). respectively. These grain values are typical of what has been observed in a range of studies in Asia, Europe, and N. America (11-13). We report similar values for a non-arsenic contaminated region of Bangladesh (Table 2). The model of arsenic consumption raises concern as they show that “background” levels of arsenic in rice contribute considerably to dietary exposure in elevated groundwater zones. Where arsenic levels in rice increased from cultivation on arsenic contaminated soils, such as Naogaon and Nawabgonj districts, the consequences for arsenic consumption are considerable. The figure of 0.5 µg g-1 grain arsenic for rice grown on soils with 70 µg g-1 arsenic (23) and 0.7 µg g-1 grain arsenic from rice grown on soil containing 68 µg g-1 arsenic (15) would give figures of 51 and 60% contribution by rice to dietary exposure if the population were drinking 2 L of 0.1 mg L-1 arsenic contaminated drinking water. If a drinking water arsenic concentration of 1 mg L-1 is considered, 0.5 µg g-1 arsenic in grain would contribute to 10% of arsenic ingestion. At 2 µg g-1 grain arsenic, the levels recorded in Naogaon and Nawabgonj disticts, 98%, 80%, and 30% of arsenic intake would be from rice at drinking water concentrations of 0.01, 0.1, and 1 mg L-1, respectively. At lower levels of arsenic in groundwater (0.01 mg L-1), grain derived arsenic is the dominant source of arsenic exposure. High levels in tube-well water would result in high levels in the paddy soils. Thus exposure from rice will increase in the most affected areas. These findings suggest that ingestion of rice is a major source of arsenic exposure in Bangladesh and elsewhere in regions with subsistence rice diets. The bioavailability of arsenic in rice must be addressed to understand the importance of arsenic exposure from this food source. Arsenic is present in rice grain predominantly as inorganic arsenic (arsenate and arsenite) and methylated species (monomethyl arsonic acid, dimethyl arsinic acid) (12, 13, 24). All these species are toxic and readily assimilated into the blood stream. The quantity of these species mobilized from rice products into the blood stream is not known. Extraction under harsh acidic heated digestion conditions effectively mobilizes these arsenic species from rice (12, 13, 24). Bioavailability of rice derived arsenic in human stomachs needs to be assessed. Survey has shown that paddy soils of Bangladesh can have high levels of arsenic, and this elevation is due to VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
233
irrigation with contaminated water. Even if a safe source of drinking/irrigation water was found for the affected areas, the paddy soils irrigated with arsenic contaminated water will be contaminated indefinitely. If a safe drinking water source is successfully introduced to Bangladesh, such as rainwater harvesting or treatment of tube-well water, it is likely that such technologies will not provide the volume, and/or will be too expensive, to be used to produce clean irrigation water. Thus tube-well water may be still be used to irrigate rice crops. It is imperative that the contribution of rice to dietary arsenic exposure is investigated in Bangladesh, as evidence is presented here that arsenic is elevated in rice grain grown on contaminated soil. This should be through grain surveys and through studies on rice from paddy fields where the arsenic concentration in the irrigation waters are monitored. Once the scale of the problem is understood, the question of how to decrease dietary exposure from arsenic still exists.
Literature Cited (1) Smith, A. H., Lingas, E. C.; Rahman, M. Bull. World Health Org. 2000, 78, 1093. (2) Nordstrum, D. K. Science 2002, 296, 2143. (3) British Geological Survey http://www.dainichi-consul.co.jp/ english/ article/DFID sum.html). 2000, (Nov. 11, 2000). (4) Tondel, M.; Rahman, M.; Magnuson, A.; Chowdhury, I. A.; Faruquee, M. H.; Ahmad, S. A. Environ. Health. Persp. 1999, 107, 727. (5) WHO. http://www.who.imt/inf-fs/en/fact210.ht, l 2001, (may 30/2001). (6) Ninno, C.; Dorosh, P. A. Agric. Econ. 2001, 25, 337. (7) Dey, M. M.; Miah, M. N. I.; Mustafi, B. A. A.; Hossain, M. In Rice research in Asia: Progress and priorities; Evenson, R. E, et al., Eds.; CAB International, Wallingford, UK and International Rice Research Institute: Manila, Philippines, 1996; p 179. (8) Alam, M. D.; Sattar, M. A. Water Sci. Technol. 2000, 42, 185. (9) Ullah, S. M. In Abstracts: International conference on arsenic pollution of groundwater in Bangladesh: causes, effects and
234
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003
(10) (11) (12)
(13) (14)
(15) (16) (17) (18) (19) (20) (21)
(22) (23) (24) (25) (26)
remedies, 8-12 February 1998, Dhaka Community Hospital, Dhaka, Bangladesh, 1998; p 133. Adeel, Z. Glob. Environ. Change 2002, 12, 69. Robberecht, H.; Cauwenbergh, R. V.; Bosscher, D.; Cornelis, R.; Deelstra, H. Eur. Food Res. Technol. 2002, 214, 27. Schoof, R. A.; Yost, L. J.; Eickhoff, J.; Crecelius, E. A.; Cragin, D. W.; Meacher, D. M.; Menzel, D. B. Food Chem. Toxicol. 1999, 37, 839. Tao, S. S.-H.; Bolger, P. M. Food Add. Contam. 1998, 16, 465. Schoof, R. A.; Yost, L. J.; Eickhoff, J.; Crecelius, E. A.; Irgolic, K.; Goessler, W.; Guo, H. R.; Greene, H. Hum. Ecol. Risk Asses. 1998, 4, 117. Xie, Z. M.; Huang, C. Y. Commun. Soil Sci. Plant Anal. 1998, 29, 2471. Nickson, R.; McArthur, J.; Burgess, W.; Ahmed, K. M.; Ravenscroft, P.; Rahman, M. Nature 1998, 395, 338. Bouman, B. A. M.; Tuong, T. P. Agric. Water Manag. 2001, 49, 11. Jensen, J. R.; Mannan, S. M. A.; Uddin, S. M. N. Agric. Water Manag. 1993, 23, 199. Woolson, E. A. Weed Sci. 1997, 25, 412. Acharyya, S. K.; Chakraborty, P.; Lahiri, S.; Raymahashay, B. C.; Guha, S.; Bhowmik, A. Nature 1999, 401, 545. Chowdhury, T. R.; Basu, G. K.; Mandal, B. K.; Samanta, G.; Chowdhury, U. K.; Chanda, C. R.; Lodh, D.; Lal Roy, S.; Saha, K. C.; Roy, S.; Quamruzzaman, Q.; Charaborti, D. Nature 2000, 401, 545. Abedin, Md. J.; Cresser, M.; Meharg, A. A.; Feldmann, J.; CotterHowells, J. Environ. Sci. Technol. 2002, 36, 962. Abedin, Md. J.; Cotter-Howells, J.; Meharg, A. A. Plant Soil 2002, 240, 311. Heitkemper, D. T.; Vela, N. P.; Stewart, K. R.; Westphal, C. S. J. Anal. Atomic Spectrosc. 2001, 16, 299. Baffes, J.; Gautam, M. Food Policy 2001, 26, 515. Das, D. et al. Analyst 1995, 120, 917.
Received for review July 19, 2002. Revised manuscript received October 15, 2002. Accepted October 24, 2002. ES0259842
