SEARCH

SEARCH BY CITATION

Keywords:

  • Arsenic;
  • Soil;
  • Bioavailability;
  • In vitro method;
  • Risk assessment

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

Encroachment of residential development on agricultural lands in the United States where arsenical pesticides were extensively used prior to the 1990s has increased the potential for human exposure to arsenic (As), a group A carcinogen. Soil ingestion by children is a critical issue in assessing health risks from exposure to As-enriched soils. In the absence of a universal “soil model” on As bioavailability, many baseline risk assessment studies use the assumption that all (100%) As present in soil is bioavailable. However, As exists in many geochemical forms as dictated by soil chemical properties. Because As bioavailability is a function of soil speciation, using total soil arsenic values potentially overestimates human health risk, thereby increasing site cleanup expenses. A laboratory incubation study was conducted to estimate in vitro As bioavailability as a function of soil properties in four chemically variant soil types contaminated with sodium arsenite pesticide. Results demonstrate that As speciation in certain soils translates to significant lowering of As bioavailability and hence potential cancer risk.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

The United States Environmental Protection Agency (U.S. EPA) has classified inorganic arsenic (As) as a group A human carcinogen (Southworth 1995). In the late 1980s and early 1990s, the U.S. EPA banned the usage of many inorganic As-based pesticides, but a vast volume of agricultural lands likely has already been contaminated. Murphy and Aucott (1998) estimated that approximately 15 million pounds of As was applied to New Jersey soils alone between 1900 and 1980 and reported the discovery of As-contaminated soils in residential developments located on former apple orchards. Similar situations exist in Texas, where cotton fields contain above-normal levels of As due to years of widespread application of As-based pesticides, mostly used as defoliants during cotton harvesting (Hudak 2000). Organic forms of As are considered noncarcinogenic and are still being used on agricultural lands. However, transformation of As from organic to inorganic forms is possible in soil environments (Rodriguez 1998). Rapid encroachment of suburban development on lands previously used for agricultural purposes in fast-expanding metropolitan areas has tremendously increased the potential for human contact with arsenic-enriched soils in the past two decades. The importance of considering soil ingestion from incidental hand-to-mouth activity by children playing in the backyards has been repeatedly emphasized in recent studies assessing public health risks associated with long-term exposure to low-level As-contaminated systems (Calabrese et al. 1989).

A critical parameter for realistic health risk assessment in As-containing soils is an estimate of “bioavailable” As, which is the extent of absorption of a chemical into the bloodstream from the gastrointestinal tract, lungs, or skin (Halmes and Roberts 1997). A majority of the U.S. EPA-mandated baseline risk assessments studies of superfund sites have used the conservative estimate that all (100%) As present in soil is bioavailable by equating As bioavailability in water with that in soils. However, limited available in vivo data show that bioavailability of As in soils is significantly less than that in water (Ng et al. 1993; Groen et al. 1994; Rodriguez et al. 1999). Using an input value of 100% bioavailability generally overestimates the actual health risk, thereby elevating the potential site cleanup expenses. Moreover, As exists in many geochemical/mineralogical forms, many of which are stable and/or insoluble in human gastric/intestinal juices and hence are not likely to be bioavailable. Therefore, evaluation of As bioavailability requires accurate, case-specific information on the geochemical fate of As. The major reason why As bioavailability is generally assumed to be 100% is the tremendous cost associated with performing in vivo bioavailability studies for every reported As contamination case. In vivo studies typically involve animal models (e.g., rats, rabbits, monkeys, and pigs); disadvantages in conducting such studies include expense, specialized facilities/personnel requirement, and time. Several in vitro methods to estimate bioavailable As have been developed in recent years that simulate in vivo data with varying degree of success (Sheppard et al. 1995; Rodriguez et al. 1999). However, these studies generally do not focus on the issue of variability in soil properties and its effect on bioavailable As concentration. A universally accepted “soil model” to predict As bioavailability is still lacking.

The current study aims at addressing the issue of soil variability on human bioavailability of As. The primary objective of this study was to reevaluate the validity of the “one size fits all” model currently practiced in the majority of baseline risk assessments where the issue of soil variability and its impact on As bioavailability is not given any consideration. In this study, we present data obtained from a 4-month incubation study using four chemically variant soils contaminated with a sodium arsenite pesticide using an apposite in vitro gastrointestinal model to estimate human bioavailability of As. The soil-specific in vitro bioavailability data were used to calculate the potential cancer risk as opposed to total soil As concentrations. The effect of soil properties on As bioavailability is discussed.

Table Table 1.. Selective properties of the soils. Data represent mean of three replicates
Properties ImmokaleeMillhopperPahokee MuckOrelia
  1. a EC = electrical conductivity.

  2. b CEC = cation exchange capacity.

  3. c SOM = soil organic matter.

pH 6.06.45.98.2
ECa (μs/cm) 59145558203
CECb (Cmol/kg) 7772,35618,9083,810
SOMc (%) 0.844.3885.42.39
As (mg/kg) 15.016.514.017.0
P (mg/kg)Mehlich 34.013436.062.0
 Total2084,8756,8121,688
Ca+Mg (mg/kg)Mehlich 326688610,0053,099
 Total1,1783,15540,80013,125
Fe+Al (mg/kg)Oxalate667041,957384
 Total2124,7456,0106,060

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

Soil sampling, preparation, and characterization

Four soil types were used: the Immokalee series, with more than 93% sand and minimum As retention capacity; the Millhopper series, a sandy loam with relatively high concentration of Fe/Al-oxides; the Pahokee Muck series, high in organic matter in addition to high Fe/Al and Ca/Mg content; and the Orelia series, a sandy clay loam with high pH and high Ca/Mg content. The soils were incubated with the pesticide for a period of 4 months. Arsenic bioavailability results at 0 time (immediately after spiking the soils with pesticides) and after a 4-month incubation period are reported.

Surface soil (0–15 cm) from the Immokalee series was collected from Southwest Florida Research and Education Center, Immokalee, Florida; the Millhopper series soil was collected from the University of Florida campus at Gainesville, Florida; the Pahokee Muck was collected from Everglades Research and Education Center at Belle Glade, Florida; and the Orelia series soil was collected from Corpus Christi, Texas. Soil samples were air-dried and passed through a 2-mm sieve. The soils were characterized for pH, electrical conductivity, and particle size using standard protocols (Klute 1996). Water content was estimated by the gravimetric method, and organic matter content was determined using the loss-on-ignition method (Klute 1996). Exchangeable cations were extracted in 1 M ammonium acetate (pH 7.0), and cation exchange capacity was determined by removal of ammonium ions (Rhoades 1982). Plant-available Ca, Mg, and P were extracted by Mehlich III solution (Mehlich 1984). Oxalate-extractable Fe and Al, was obtained using Tamm's reagent (Klute 1996). Total P was extracted using the ignition method (Saunders and Williams 1955). Total recoverable Ca, Mg, Fe, Al, P, and As was obtained by soil digestion according to U.S. EPA method 3050B (USEPA 1996).

Phosphorus was measured colorimetrically by ultraviolet/visible light spectrophotometer using the molybdate-ascorbic acid method (Watanabe and Olsen 1965). Iron was determined colorimetrically according to Olson et al. (1982) as a complex with 1,10-phenanthroline reagent. The Ca, Mg, and Al were analyzed using flame atomic absorption spectrometry (FAAS). Arsenic was analyzed using graphite furnace atomic absorption spectrometry (GFAAS).

Soil amendments

Two hundred grams of soil each from the Immokalee, Millhopper, Pahokee Muck, and Orelia series were spiked with sodium arsenite at the rate of 45 mg As/kg soil. Water content was maintained at 70% of the water holding capacity of the soils and was measured in triplicate once every month in randomly picked samples. Pesticide was thoroughly mixed with soils, which were then stored in tightly sealed bags at room temperature. The soils were aerated regularly, and constant water content was maintained by adding water (if necessary) followed by thorough hand mixing. Arsenic was extracted from the soils immediately after pesticide amendment and after a 4-month incubation period using in vitro procedure as described next.

In vitro procedure

Bioavailable As was estimated following the method of Rodriguez et al. (1999) with certain modifications (Sarkar and Datta 2003). The reactions were carried out in 250-ml beakers in a 37°C water bath to simulate body temperature. Anaerobic conditions were maintained by passing argon gas through the solutions. The pH of the solutions was continuously monitored. Constant mixing of the solution was maintained using a stirrer to simulate gastric mixing. The gastric solution consisted of 0.15 M NaCl and 1% porcine pepsin (Sigma Chemical, St. Louis, MO, USA). One gram of soil sample was added to 150 ml of gastric solution, and the pH of the solution was adjusted to 1.8 using 1 N HCl. The solution was incubated for 1 h at the end of which 10 ml of solution were collected, centrifuged at 5,000 rpm for 30 min, and analyzed by GFAAS.

Risk calculation using bioavailability data

The amount of soil ingested by a child is generally estimated as 100 mg/d. Based on this value, chronic daily intake (CDI) was calculated using Equation 1:

  • equation image((1))

where CDI is chronic daily intake of As (mg) per day, AsCS is As concentration in soil (mg/kg), and 0.0001 is the amount of soil (kg) ingested by a child per d.

The potential cancer risk was calculated as the product of CDI and the appropriate slope factor. This value is defined as the excess cancer risk (ECR):

  • equation image((2))

where ECR is the estimate of the excess lifetime probability of developing cancer resulting from exposure to As (dimensionless), CDI is chronic daily intake (mg/kg/d), and SF is slope factor of As for carcinogenic effects (mg/kg/d).

Slope factor is defined as a plausible upper-bound estimate of the probability of a response per unit intake of the chemical over a lifetime. For As, the slope factor was assumed to be 1.5 mg/kg/d, obtained from U.S. EPA's Integrated Risk Information System (IRIS) (USEPA 2001) and Health Effects Assessment Summary Tables (HEAST) (USEPA 1997) databases.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

Soil properties

Human health risk from exposure to As-enriched soils is associated only with those forms of soil As that are potentially extractable by the human gastrointestinal juices. Arsenic may exist in several geochemical forms in soils depending on soil chemical properties; some of these forms may not be bioavailable. The soil properties that most likely to influence As retention are listed in Table 1. Soils from the Immokalee, Millhopper, Pahokee Muck, and Orelia series were characterized to determine their pH, salinity, soil organic matter (SOM) content, cation exchange capacity (CEC), and total and available P, Mg, Ca, Fe, and Al. Immokalee soil is a sandy spodosol with low Fe/Al, Ca/Mg, and P contents. Being sandy and lacking positively charged surfaces (e.g., amorphous Fe/Al oxides), the Immokalee soil is likely to have minimal As retention capacity (Pierce and Moore 1980; Oscarson et al. 1981), thereby potentially increasing the bioavailable fraction of As. This soil was used as a control to study the effects of the variant soil chemical properties of the other soils. The Millhopper soil has a pH similar to Immokalee but much higher concentrations of Fe, Al, and P; the Pahokee Muck soil has 85% SOM and much higher concentrations of Fe, Al, Ca, and Mg; and the Orelia soil, in addition to having higher concentrations of Fe, Al, Ca, and Mg, has a much higher pH compared to Immokalee soil. The soils also varied widely in their salinity (measured as electrical conductivity [EC]) and the CEC, both likely to affect As geochemistry. The higher the CEC, the greater the amount of positive charge on the surface and the higher the potential of the As oxyanions to form electrostatic bonds with the positively charged surface sites. According to Chen et al. (1999), the major factors controlling trace metal concentrations in soils are clay content, organic carbon content, pH, CEC, and Fe, Al, Ca, Mg, and P concentrations. Because both As and P occur as oxyanions in environmental systems and have similar chemical properties, high P content of the Millhopper soil could result in desorption of retained As. Reportedly, As is strongly adsorbed onto Fe and Al oxides (Jacobs et al. 1970; Barringer et al. 1998); hence, Millhopper, Pahokee Muck, and Orelia soils are likely to have strong As retention capabilities. Generally, sorption of As decreases with increasing pH (Adriano 2001). This can be attributed to the negative surface charge on the adsorptive surface at higher pH as well as the negative charge of As oxyanions (Wasay et al. 2000). Certain components of SOM (such as fulvic acid) tend to complex As, thereby making it more soluble and hence bioavailable (Gough et al. 1996). On the other hand, adsorption of As by certain other components of SOM, such as humic acids, is high in the pH range of 5 to 7 and when the humic acids have high ash and calcium content (Mok and Wei 1994). Humic acids can contribute more to the retention of As in acidic environments than do clays and some metal oxides, thereby lowering its ultimate bioavailability. The major retention sites on the humic acids at low-pH systems are the amine groups (Thanabalasingam and Pickering 1986). Moreover, As bound to the Ca/Mg fraction in Millhopper, Pahokee Muck, and Orelia soils has the potential solubilize in the highly acidic environment of human stomach, thus becoming bioavailable. All the soils studied had similar native As concentration, ranging between 14 mg/kg in Pahokee Muck and 17 mg/kg in Orelia.

Estimation of bioavailable arsenic

Bioavailability of arsenic is typically assessed using animal (in vivo) models, which are rather complicated, lengthy, and expensive procedures that require highly trained personnel. Consequently, in vitro methods (“beaker models”) have been designed to chemically simulate metal availability in the human gastrointestinal system. Ruby et al. (1993) developed the physiologically based extraction test, or PBET, to predict metal bioavailability taking into consideration metal solubility in gastric and intestinal juices. Rodriguez et al. (1999) modified the method by introducing iron hydroxide gel in nylon bags to simulate intestinal absorption of metals. Sarkar and Datta (2003) used Fe-oxide-coated filter papers instead of iron hydroxide gels to better mimic the physical properties of the intestinal membrane to estimate As absorption.

Results of the in vitro gastrointestinal bioavailability studies obtained from soils contaminated with 45 and 450 mg/kg As are shown in Figures 1 and 2, respectively. For the 45-mg/kg As-spiked soils, approximately 100% of the total As was bioavailable at 0 time (immediately after pesticide spiking) in the Immokalee and Millhopper soils, whereas about 90 and 65% of As was bioavailable in the Pahokee Muck and Orelia soils (Figure 1). After 4 months equilibration, the percent bioavailability dropped to about 72% in the Immokalee soil. However, the number was significantly lower in the case of Millhopper (39%) and Pahokee Muck and Orelia (44% each) after the pesticide was allowed to equilibrate with the soil. The potentially irreversibly adsorbed As fraction rendered a significant portion of total soil As unavailable to the human gastrointestinal system in Millhopper, Pahokee Muck, and Orelia soils. Interestingly, although Pahokee Muck soil has higher concentrations of Fe and Al compared to Millhopper (Table 1), the latter was more efficient in As retention as indicated by the lower percent bioavailability (Figure 1). This is likely to be caused by the high amount of organic matter in Pahokee Muck that potentially solubilizes As.

Table Table 2.. Cancer risk calculation based on bioavailability of arsenic in pesticide-applied soils. Data represent mean of three replicates
   CDIb (mg/kg/d)ECRc
Soil seriesAsCSa (mg/kg)Bioavailability (%)TotalBioavailableTotalBioavailable
  1. a AsCS = As concentration in soil.

  2. b CDI = chronic daily intake.

  3. c ECR = excess cancer risk.

Immokalee4572.120.450.320.680.49
 45074.514.503.356.805.03
Millhopper4539.030.500.180.680.26
 45076.194.503.436.805.14
Pahokee Muck4543.990.450.200.680.30
 45082.294.503.706.805.50
 4544.10.450.200.680.30
Orelia45060.674.502.736.804.10

Similar results were obtained for the 450-mg/kg As-spiked soils, although the reduction in As bioavailability was not as marked as in the lower contamination system (Figure 2), except for the control soil from Immokalee series. After 4-month equilibrations, As bioavailability decreased by 24, 18, and 29%, respectively, in Millhopper, Pahokee Muck, and Orelia soils. One possible explanation for this behavior is potential lack of soil-pesticide equilibration at such high levels of contamination. Possibly, 4 months was not enough time to promote static equilibration; we are currently aging the soils for a year to verify this assumption. The impact was more prominent in the chemically complex soils with higher reactivity toward As. It is also possible that the soil sorption sites were saturated at such high concentrations and started desorbing As, thereby making it more bioavailable. Rodriguez (1998) also found that As bioavailability was not a function As loading rate in the contaminated media studied.

No attempt was made in the present study to quantify the various aqueous arsenic species in the in vitro solutions. However, it would be worthwhile to quantify these species, particularly since arsenite (reduced form) is more mobile and toxic compared to arsenate (oxidized form).

thumbnail image

Figure Figure 1.. Arsenic bioavailability in pesticide-applied soils (45 mg As/kg) using in vitro method at time 0 and after 4-month equilibration. Data represent mean of three replicates.

Download figure to PowerPoint

Risk calculations

Remedial decisions are usually based on risk calculations and the identification of unsafe levels of human exposure. In accordance with U.S. EPA regulations, risk to human health is calculated in terms of excess cancer risk, which is the additional risk above background risk of developing cancer as a result of a specific exposure. Using the data obtained from in vitro bioavailability studies, the reduction in potential cancer risk was calculated using soil-specific bioavailability data as opposed to total As data (Table 2). Even in a sandy soil like Immokalee, which has minimal As retention capacity, the bioavailability of As (76%) was significantly reduced as a result of As speciation in soil, resulting in lowering of cancer risk to 0.51, as compared to the expected value of 0.68 if total soil As concentrations were used. In a soil having high concentration of Fe/Al, such as Millhopper, bioavailability was only 39 % of total As, lowering the cancer risk to a much greater extent (0.26). In Pahokee Muck and Orelia soils, the cancer risk was reduced to 0.30, reflecting the reduction in the As bioavailability value to 44%. Such data provide encouraging evidence in favor of our hypothesis that soil properties influence human bioavailability of As and hence need to be considered while performing human health risk assessment from exposure to As-enriched soils.

thumbnail image

Figure Figure 2.. Arsenic bioavailability in pesticide-applied soils (450 mg As/kg) using in vitro method at time 0 and after 4-month equilibration. Data represent mean of three replicates.

Download figure to PowerPoint

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

The results obtained in this study demonstrate the importance of considering site-specific bioavailability data rather than total As data for risk calculations. Soil properties had a marked impact on the bioavailability of As. After adequate equilibration of pesticide-applied soils, soil chemistry dictated bioavailability of As in all the four soils studied. Arsenic was more strongly retained in Millhopper, Pahokee Muck, and Orelia compared to Immokalee, and the potentially irreversibly adsorbed As fraction rendered a significant portion of total soil As unavailable to the human gastrointestinal system. A reduction of bioavailability due to soil speciation of As resulted in significant reduction of cancer risk.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

We acknowledge the Texas Higher Education Coordinating Board Advanced Research Program for funding this study. The Center for Water Research, UTSA, is acknowledged for providing the support for a technical assistant and for the administrative assistance of Hermina Simpson. Thanks are also due to E. Hanlon, T. Obreza, J. Matocha, G. Snyder, J. Walker, and Larry Schwandes for providing the soils used in this study.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES
  • Adriano DC. 2001. Trace elements in terrestrial environments: Biogeochemistry, bioavailability, and risks of metals. New York (NY), USA: Springer.
  • Barringer JL, Szabo Z, Barringer TH. 1998. Arsenic and metals in soils in the vicinity of the Imperial Oil Company superfund site, Marlboro Township, Monmouth County, New Jersey. West Trenton (NJ), USA: U.S. Geological Survey. USGS WRI report 98–4016. 251 p.
  • Calabrese E, Barnes R, Stanek EJ. 1989. How much soil do young children ingest: An epidemiologic study. Reg Toxicol Pharmacol 10: 123137.
  • Chen M, Ma LQ, Harris WG. 1999. Baseline concentrations of 15 trace elements in Florida surface soils. J Environ Qual 28: 117381.
  • Gough LP, Kotra RK, Holmes CW, Briggs PH, Crock J, Fey DL, Hegeman PL, Meter AL. 1996. Chemical analysis results for mercury and trace elements in vegetation, water, and organic-rich sediments. Washington DC: U.S. Geological Survey. Open file report 96–091. 29 p.
  • Groen K, Vaessen HA, Kliest J, de Boer JL, van Ooik T, Timmerman A, Vlug, RF. 1994. Bioavailability of inorganic arsenic from bog ore-containing soil in the dog. Environ Health Perspect 102: 182184.
  • Halmes NCH, Roberts SM. 1997. Technical report 97–02. Gainesville (FL), USA: University of Florida Center for Environmental and Human Toxicology.
  • Hudak PF. 2000. Distribution and sources of arsenic in the southern high plains aquifer. J Environ Sci Health 35: 899913.
  • Jacobs LW, Syers JK, Keeney DR. 1970. Arsenic-phosphorus interactions on corn. Soil Sci Soc Am Proc 33: 279282.
  • Klute A. 1996. Methods of soil analysis: Part 1: Physical and mineralogical methods. Madison (WI), USA: Soil Science Society of America.
  • Mehlich A. 1984. Mehlich No 3 soil test extractant: A modification of Mehlich No 2 extractant. Commun Soil Sci Plant Anal 15: 14091416.
  • Mok WM, Wei CM. 1994. Mobilization of arsenic in contaminated river waters. In: NriaguJO, editor. Arsenic in the environment part I: Cycling and characterization. New York (NY), USA: Wiley. p 99117.
  • Murphy EA, Aucott M. 1998. An assessment of the amounts of arsenical pesticide used historically in a geographic area. Sci Total Environ 218: 89101.
  • Ng JC, McDougall KW, Imray P, Hertle A, Seawright AA. 1993. Arsenic contaminated soil: A study of the biological availability of the element in comparison with sodium arsenite, sodium arsenate, and calcium arsenate. In: The 12th Australian Symposium on Analytical Chemistry/3rd Environmental Chemistry Conference; 1993 Sep 26-Oct 1; Perth, Australia.
  • Olson RV, Ellis R Jr, Kunze GW, Dixon J. 1982. Chemical and microbiological properties. In: PageAL, editor. Methods of soil analysis: Part 2. Madison (WI), USA: American Society of Agronomy. p 301312.
  • Oscarson DW, Huang PM, Defosse C, Herbillon A. 1981. Oxidative power of Mn (IV) and Fe (III) oxides with respect to As (III) in terrestrial and aquatic environments. Nature 291: 5051.
  • Pierce ML, Moore CB. 1980. Adsorption of arsenite on amorphous iron hydroxide from dilute aqueous solution. Environ Sci Technol 14: 214216.
  • Rhoades JD. 1982. Cation exchange capacity. In: PageAL, editor. Methods of soil analysis: Part 2. Madison (WI), USA: American Society of Agronomy. p 149158.
  • Rodriguez RR. 1998. Bioavailability and biomethylation of arsenic in contaminated soil and solid wastes [dissertation]. Stillwater (OK), USA: Oklahoma State University.
  • Rodriguez RR, Basta NT, Casteel S, Pace L. 1999. An in-vitro gastro-intestinal method to estimate bioavailable arsenic in contaminated soil and solid media. Environ Sci Technol 33: 642649.
  • Ruby MV, Davis A, Link TE, Schoof R, Chaney RL, Freeman GB, Bergstrom P. 1993. Estimation of lead and arsenic bioavailability using a physiologically based extraction test. Environ Sci Technol 27: 28702877.
  • Sarkar D, Datta R. 2003. A modified in-vitro method to assess bioavailable arsenic in pesticide-applied soils. Environ Pollut 126: 363366.
  • Saunders WMH, Williams EG. 1955. Observations on the determination of total organic phosphorus in soils. J Soil Sci 6: 254267.
  • Sheppard SC, Evenden WG, Schwartz WJ. 1995. Ingested soil: Bioavailability of sorbed lead, cadmium, cesium, iodine, and mercury. J Environ Qual 24: 498505.
  • Southworth RM. 1995. Land application pollutant limit for arsenic. Part 503. Washington DC: U.S. Environmental Protection Agency.
  • Thanabalasingam P, Pickering WF. 1986. Arsenic sorption by humic acids. Environ Pollut Ser B Chem Phys 12: 233246.
  • [USERA] U.S. Environmental Protection Agency. 1996. Test methods for evaluating solid waste, SW 846 (3rd ed.). Washington DC: USEPA, Office of Solid Waste and Emergency Response.
  • [USEPA] U.S. Environmental Protection Agency. 1997. The health effects assessment summary tables—Annual update (HEAST). Washington DC: USEPA. EPA/540-R-97–036. PB97–921199.
  • [USEPA] U.S. Environmental Protection Agency. 2001. The integrated risk information system (IRIS). Cincinnati (OH): USEPA, Environmental Criteria and Assessment Office.
  • Wasay SA, Parker W, Van Geel PJ, Barrington S, Tokunaga S. 2000. Arsenic pollution of a loam soil: Retention form and decontamination. J Soil Contam 9: 5164.
  • Watanabe FS, Olsen SR. 1965. Test of an ascorbic acid method for determining phosphorus in water and NaHCO3 extracts from soil. Soil Sci Soc Am Proc 29: 677678.