Technology Description

Compounds at installations, particularly ranges, are often found in heterogeneous distributions in the environment. Additionally, valued wildlife species use the environment in very biased ways that are linked to habitat requirements. Current means to estimate exposure lose the spatial heterogeneity through the calculation of a single upper-range mean value for each contaminant (upper 95% confidence level of the mean value or highest concentration detected). That value is used to represent the concentration in soil and sediment throughout the site of concern. From there, exposure is modeled for a generalized single representative of a species, and any influences of habitat preferences are lost. Although likely protective, this process confounds and exacerbates exposure estimates and leads away from predictive estimates of risk. Recently, two spatially explicit exposure models have been developed within the Army Environmental Quality Technology program that include spatial aspects of chemical distributions in the environment. Additionally, they incorporate virtual movements for a population of individuals in the environment that include habitat preferences and produce a probabilistic estimate of risk. The Spatially Explicit Exposure Model (SEEM) calculates exposure and risk for terrestrial wildlife species, and FishRand calculates tissue concentrations for fish. SEEM has been tested at two small arms range sites for songbirds and was shown to be predictive within a factor of three, whereby the conventional deterministic model was between 10- to 500-fold off from field observations. This project will further test and refine these models—SEEM for its utility in predicting risk for small mammals and FishRand for fish. Field verification of results will be based on blood lead toxicity reference values for mammals and fish concentrations of persistent organic substances for fish. Additionally, phenotypic biomarkers for effect will be investigated.