Inclusion of soil arsenic bioaccessibility in ecological risk assessment and comparison with biological effects.

The purpose of this study was to conduct an ecological risk assessment (ERA) for meadow voles (Microtus pennslvanicus) found at three arsenic contaminated sites in Nova Scotia, Canada (as well as two background locations) and to compare the numeric results to measured biomarkers of exposure and effect. The daily intake of arsenic by meadow voles was determined by three separate calculations: estimated daily intake (EDI), bioaccessible estimated daily intake (BEDI, with bioaccessibility of soil included), and actual daily intake (ADI, which is calculated with arsenic concentrations in the stomach contents). The median bioaccessibility of arsenic in soils from the contaminated locations was significantly greater than at background locations. The bioaccessible arsenic concentration in soil from all samples (both contaminated and background) was significantly less than the total concentration. Use of site-specific bioaccessibility (hazard quotients=38 at Upper Seal Harbour (USH); 60 at Lower Seal Harbour (LSH); and 120 at Montague tailings (MONT)) and stomach arsenic contents (hazard quotients=2.1 at USH; 7.9 at LSH; and 6.7 at MONT) in the ERA resulted in lower numeric risk than compared to risk calculated with 100% bioavailability (hazard quotient=180 at USH; 75 at LSH; and 680 at MONT). Further, the use of bioaccessibility on the calculation of risk was aligned with biomarker results (changes in glutathione and micronucleated erythrocytes) in voles captured at the sites. This study provides evidence that using site-specific bioaccessibility in ERAs may provide a more realistic level of conservatism, thereby enhancing the accuracy of predicting risk to wildlife receptors. Furthermore, when numeric risk assessments are combined with site-specific biological data (i.e., biomarkers of exposure and effect), both lines of evidence can be used to make informed decisions about ecological risk and site management.