Arsenic retention and release from ferrihydrite and tropical soils during iron and sulfate reduction

Soils and surface sediments are enriched with arsenic bearing iron (hydr)oxides throughout sedimentary basins of southeast Asia; arsenic concentrations typically range from 10 to 40 mg/kg (DW). Seasonal monsoons result in drastically fluctuating water levels, which drive wide swings in redox processes within soils and near-surface sediments. Extensive flooding results in anaerobic soil/sediment conditions, with ensuing reduction of iron (hydr)oxides and sulfate. Aqueous concentrations of arsenic vary with redox conditions and processes and when elevated, result in migration through the soil profile to the deeper subsurface. Using ceramic-cup lysimeters and passive samplers (peepers), we measured a suite of reduced aqueous constituents, including arsenic, within soil profiles in the lower Mekong delta, Cambodia. Within a zone of seasonal wetting/drying, we observed increased aqueous concentrations of arsenic in surface soils during periods of flooding. Arsenic release is coincident with the production of Fe(II) and, to a lesser degree, S(-II) in the porewater. These results indicate that arsenic mobilization occurs in near-surface soil environments upon soil wetting, and that multiple chemical processes (arsenic, iron, and sulfate reduction) may contribute to arsenic release. These measurements were therefore complimented with a suite of model column experiments utilizing arsenic-loaded ferrihydrite-coated sand and bacteria capable of iron, sulfur, and/or arsenate reduction. Column studies indicate that biologically-induced iron reduction by S. putrefaciens (CN32) leads to arsenic sequestration within biotransformed solids, relative to abiotic control columns. At high surface loadings of arsenic, biotransformation of ferrihydrite during sulfidogenesis induced by D. vulgaris (Hildenborough) also leads to arsenic sequestration within column solids; however, low surface loadings of arsenic resulted in arsenic mobilization during rapid dissolution of ferrihydrite and subsequent formation of iron sulfide (relative to high surface arsenic loadings). Finally, the greatest amount of arsenic was mobilized during As(V) reduction by B. benzoevorans, which reduces As(V), but is incapable of Fe(III) or sulfate reduction. Our results indicate that arsenic release in subsurface environments is controlled by complex pathways of mineral biotransformation and redox processes, and that As(V) reduction may be the largest factor governing arsenic release under the imposed column reaction conditions.