Estimating Exposure of Terrestrial Wildlife to Contaminants

This report presents a general model for exposure of terrestrial wildlife to contaminants (Sect. 2), methods for estimating parameters of the model (Sect. 3), species specific parameters for endpoint species on the Oak Ridge Reservation (ORR) (Sect. 4), and a sample application (Sect. 5). Exposure can be defined as the coincidence in both space and time of a receptor and a stressor, such that the receptor and stressor come into contact and interact (Risk Assessment Forum 1992). In the context of ecological risk assessment, receptors include all endpoint species or communities identified for a site [see Suter (1989) and Suter et al. (1994) for discussions of ecological endpoints for waste sites]. In the context of waste site assessments, stressors are chemical contaminations, and the contact and interaction are uptake of the contaminant by the receptor. Without sufficient exposure of the receptor to the contaminants, there is no ecological risk. Unlike some other endpoint assemblages, terrestrial wildlife are significantly exposed to contaminants in multiple media. They may drink or swim in contaminated water, ingest contaminated food and soil, and breath contaminated air. In addition, because most wildlife are mobile, moving among and within habitats, exposure is not restricted to a single location. They may integrate contamination from several spatially discrete sources. Therefore, exposure models for terrestrial wildlife must include multiple media. This document provides models and parameters for estimating exposure of birds and mammals. Reptiles and amphibians are not considered because few data exist with which to assess exposure to these organisms. In addition, because toxicological data are scarce for both classes, evaluation of the significance of exposure estimates is problematic. However, the general exposure estimation procedure developed herein for birds and mammals is applicable to reptiles and amphibians. Exposure models must be appropriate to the assessment endpoints. The models presented herein are models of the exposure of individual organisms, but except for threatened and endangered species, all the wildlife endpoints for the ORR are for populations (Suter et al. 1994). The use of organism exposures is appropriate because of the need to integrate exposure estimates with exposure-response information which is expressed as organism-level responses. The conversion of individual exposure to population effects occurs in the risk characterization. Conceptually, the conversion of organism-level exposures to the population level can be made in two ways. First, it may be assumed that there is a distinct population on the site so that the exposure of the population is the exposure of all the individuals. This assumption is appropriate for small organisms on large sites, particularly if the site constitutes a distinct habitat that is surrounded by inappropriate habitat. For example, a grassy site surrounded by forest or industrial development might support a distinct population of voles. The risks to that population can be estimated directly from the exposures of the individual organisms. Second, it may be assumed that a certain number of individuals are exposed to contaminants out of a larger population. For example, a certain proportion of a deer herd may forage on a site or a pair of hawks may hunt on a site. The estimated exposure of these individuals will result in estimation of certain effects on those individuals, and the resulting population risks will need to be characterized. In either case, the organism level exposure models are appropriate