that all the contaminant mass is readily available. In practical terms this means that no special adjustments are made to account for bioavailability processes when exposures are estimated. If explicit consideration of bioavailability processes is to become more frequent, the uncertainties inherent in their measurements must be addressed and reduced, if possible.

Some general sources of uncertainty associated with bioavailability processes include:

  • a lack of knowledge about how physical, chemical, and biological processes acting at the level of soil and sediment particles influence the binding and release of chemicals;

  • variations in soil and sediment characteristics at various spatial scales;

  • a lack of knowledge about how biota modify bioavailability of chemicals in soils and sediments that come into contact with external membranes (e.g., skin) or that are taken into the body (e.g., digestive systems), and whether information obtained for one species is representative of another;

  • variations in chemical form or properties (e.g., redox state of metals or diffusive rates for organics);

  • physical, chemical, or biological changes that might, at some point in the future, change the bioavailability of a chemical.

Given these multiple sources of uncertainty, regulatory agencies have been cautious about moving away from default assumptions concerning bioavailability processes in risk estimates. It is not clear whether there is too much uncertainty associated with bioavailability tools for regulatory agencies to feel comfortable about more explicitly incorporating their results into exposure estimates. Input received by the committee indicates that there is disagreement over this issue. An individual who has a strong precautionary stance might argue against replacing certain default assumptions (e.g., of 100 percent availability) to account for site-specific bioavailability processes. On the other hand, someone who sees large trade-offs among alternatives that hinge on bioavailability considerations would likely support their inclusion in specific situations. Risk assessment practitioners well versed in uncertainty and probabilistic analyses might argue that the uncertainties could be identified and taken into account, thereby providing more complete information to the risk manager.

Explicit incorporation of information on bioavailability processes has occurred in ecological and human health risk assessments for particular types of problems and chemicals where the uncertainty has been relatively low due to extensive testing of certain contaminants and processes. Examples include exposure of humans to lead in soils (oral), and to polychlorinated biphenyls (PCBs) in soils (dermal); leaching of soil contaminants to groundwater; exposure of benthic invertebrates to non-polar organic chemicals (e.g., polyaromatic hydrocarbons or PAHs) in sediments; and site-specific determinations of bioavailability via up-

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement