GEOCHEMICAL FATE OF ARSENIC IN CATTLE-DIP VAT SITE SOILS FROM AUSTRALIA

Arsenic (As) is a naturally occurring element with soil background concentrations ranging between 0.1 to 40 mg/kg. Anthropogenic activities such as mining, smelting of ore, and widespread use of pesticides have resulted in higher than background concentrations of As in the environment. In the past, dipping vats were used to control the tick population in cattle by filling the concrete structure with a slurry of arsenic trioxide, a common inorganic arsenical pesticide. Not all vat structures were concrete lined, which resulted in contamination of soils under the vat. Also, as the slurry lost its potency, it had to be changed several times a year. The used slurry was then applied directly to the neighboring soils, resulting in further spreading of As contamination. In the absence of a universally accepted “soil model” on As bioavailability, many baseline risk assessments of As-enriched sites assume that all (100%) As present in the soil is bioavailable. Because several geochemical forms of As are insoluble in human gastric/intestinal juices, these forms are not likely to be bioavailable. Hence, the assumption of 100% bioavailability generally overestimates the actual human health risk, thereby increasing potential site clean-up expenses. For the present study, five soil samples were obtained from former cattle dip vat sites in Australia. These soils were characterized for soil properties reported to influence soil retention of As, namely, pH, salinity, clay content, total P, total Ca+Mg, soil organic matter, and oxalate extractable Fe+Al concentrations. After initial characterization, contaminated soils were subjected to an appropriate in-vitro gastrointestinal (IVG) method in order to determine human bioavailability of As. Soil samples were also subjected to a sequential extraction procedure in order to determine the operationally-defined geochemical forms of As in soils. The extracted As fractions were evaluated against the various soil properties. Linear regression techniques were employed to obtain the correlations between IVG-extractable As and operationally-defined forms of As. Our results indicate that bioavailability of As is determined by its geochemical fate, and that total As is not a good predictor of As bioavailability.