ABSTRACT
Ecological risk assessors analyze the effects of human actions on the natural environment. Like other risk assessors, they have been concerned primarily with predictive assessments, those that estimate the nature, probability, and magnitude of effects of proposed actions such as the marketing of a new chemical product or the release of a new effluent. Retrospective risk assessments differ from predictive assessments principally in that the environment is already affected and can be observed and measured. Whether the problem is to establish a cause for an observed effect (effects driven) or to identify an effect associated with a source (source driven), four questions arise in ecological epidemiology: Is there a real effect as opposed to natural variation in ecological properties? What is its cause? What is its magnitude? What are its ultimate consequences? If biomarkers are to become useful in ecological risk assessment, they must be linked to the other measures and indicators that contribute to retrospective assessments.
