levels of iron may also be involved in the formation of toxic algal blooms (Graneli et al. 1986; Cosper et al. 1990). These issues are discussed in detail in Appendix A.
The effects of toxic substances on ecological systems have proven more difficult to study than the effects of nutrients or than human health effects. In general, the science of ecological risk assessment for toxic substances is not as well developed as that for human health risk assessment. The two techniques, however, involve the same fundamental principles. In the case of ecological risk assessment, the causative agent is generally referred to as a stressor, and adverse effects are identified as stresses on an ecosystem. Sometimes, ecological risk assessment can be easier than human health risk assessment. For example, controlled experiments can be performed directly on the systems of concern eliminating the need for extrapolation from high doses to low doses or sensitive subpopulations. These advantages are, for the most part, however, outweighed by the greater complexity in organization and response of ecosystems to stresses.
Several indicators of the health of an ecosystem can be used to assess the hazard of a particular stress. These indicators include rates of primary production or other processes; trophic structure; survival of sensitive species; species diversity; and population of fisheries and shellfish stocks, as well as endangered species of birds and mammals. Specific measures include population counts, growth, survival, reproduction, and recruitment.
Responses to toxic chemical stresses can take place at four levels of biological organization: 1) biochemical and cellular, 2) organismal, 3) population, and 4) community and ecosystem. Within each of these levels, there are multiple potential endpoints that could be considered. Not all responses are disruptive in nature, and they do not necessarily result in degeneration at the next level of organization. Only when the compensatory or adaptive mechanisms at one level begin to fail do deleterious effects become apparent at the next level (Capuzzo 1981). However, failures at various levels are often exceedingly difficult to discern, and so changes in populations or in ecosystems may occur without any change at the organismal level ever being detected. In general, ecological risk assessments should be performed using the most sensitive measure of stress. For aquatic ecosystems, change in community structure is an important ecological concern and appears to be sensitive to toxic chemical response (Schindler 1987, Howarth 1991). A variety of toxic agents predictably cause changes, with loss of sensitive species (e.g., amphipods) which will result in domination by weedy or opportunistic species (e.g., capitellid worms). Species diversity usually decreases although this is generally a very insensitive measure of change compared to loss of sensitive species (Howarth 1991). Table 4.1 shows the response levels of marine organisms to chemical contaminants at the four levels of ecological organization.