Effect of soil properties on arsenic fractionation and bioaccessibility in cattle and sheep dipping vat sites

Historical use of high arsenic (As) concentrations in cattle/sheep dipping vat sites to treat ticks has resulted in severe contamination of soil and groundwater with this Group-A human carcinogen. In the absence of a universally applicable soil As bioaccessibility model, baseline risk assessment studies have traditionally used the extremely conservative estimate of 100% soil As bioaccessibility. Several in-vitro, as well as, in-vivo animal studies suggest that As bioaccessibility in soil can be lower than that in water. Arsenic in soils exists in several geochemical forms with varying degree of dissolution in the human digestive system, and thus, with highly varying As bioaccessibility. Earlier batch incubation studies with As-spiked soils have shown that As bioaccessibility is a function of soil physicochemical properties. We selected 12 dipping vat soils collected from USA and Australia to test the hypothesis that soil properties exert a significant effect on As bioaccessibility in As-contaminated sites. The 12 soils varied widely in terms of soil physico-chemical properties. They were subject to an As sequential fractionation scheme and two in-vitro tests (IVGS and IVGIA) to simulate soil As bioavailability in the human gastrointestinal system. Sequential As fractionation results showed that the majority of the As measured in the dipping vat soils resided either in the Fe/Al hydroxide fraction, or the Ca/Mg fractions, or in the residual fraction. Water-extractable As fraction of the 12 soils was typically < 10% of the total, reaching values up to 23%, indicating minimal leaching potential, and hence, lower risk of As-contamination from exposure to groundwater, typically used as drinking water in many parts of the world. Partial individual correlations and subsequent multiple regression analyses suggested that the most significant soil factors influencing As bioaccessibility were total Ca + Mg, total P, clay content and EC. Collectively, these soil properties were able to explain 85 and 86% of the variability associated with the prediction of bioaccessible As, using IVGS and IVGIA in-vitro tests, respectively. This study showed that specific soil properties influenced the magnitude of soil As bioaccessibility, which was typically much lower than total soil-As concentrations, challenging the traditional risk assessment guideline, which assumes that soil As is 100% bioaccessible. Our study showed that total soil As concentration is unlikely to provide an accurate estimate of human health risk from exposure to dipping vat site soils