Relative bioavailability and bioaccessibility and speciation of arsenic in contaminated soils.

Environ Health Perspect. 2011 Nov;119(11):1629-34.
Bradham KD, Scheckel KG, Nelson CM, Seales PE, Lee GE, Hughes MF, Miller BW, Yeow A, Gilmore T, Serda SM, Harper S, Thomas DJ. 
National Exposure Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina, USA.
BACKGROUND: Assessment of soil arsenic (As) bioavailability may profoundly affect the extent of remediation required at contaminated sites by improving human exposure estimates. Because small adjustments in soil As bioavailability estimates can significantly alter risk assessments and remediation goals, convenient, rapid, reliable, and inexpensive tools are needed to determine soil As bioavailability. OBJECTIVES: We evaluated inexpensive methods for assessing As bioavailability in soil as a means to improve human exposure estimates and potentially reduce remediation costs. METHODS: Nine soils from residential sites affected by mining or smelting activity and two National Institute of Standards and Technology standard reference materials were evaluated for As bioavailability, bioaccessibility, and speciation. Arsenic bioavailability was determined using an in vivo mouse model, and As bioaccessibility was determined using the Solubility/Bioavailability Research Consortium in vitro assay. Arsenic speciation in soil and selected soil physicochemical properties were also evaluated to determine whether these parameters could be used as predictors of As bioavailability and bioaccessibility. RESULTS: In the mouse assay, we compared bioavailabilities of As in soils with that for sodium arsenate. Relative bioavailabilities (RBAs) of soil As ranged from 11% to 53% (mean, 33%). In vitro soil As bioaccessibility values were strongly correlated with soil As RBAs (R² = 0.92). Among physicochemical properties, combined concentrations of iron and aluminum accounted for 80% and 62% of the variability in estimates of RBA and bioaccessibility, respectively. CONCLUSION: The multifaceted approach described here yielded congruent estimates of As bioavailability and evidence of interrelations among physicochemical properties and bioavailability estimates.