Commentary: Hormesis and Ecological Risk Assessment

Gary M. Rand, Ph.D.

Florida International University

Ecotoxicology and Risk Assessment

Southeast Environmental Research Center

& Department of Environmental Studies

3000 NE 151st Street

N. Miami, Fl 33181

Tel: 305-919-5869

Fax: 305-919-5887


The manuscript by P.M. Chapman presents an overview of the potential "future roles that hormesis could legitimately and usefully play in ecotoxicology and ecological risk assessment". The Arndt- Schulz Law and the term "hormesis" are both used to describe a U-shape at the low end of dose-response curves. In essence, at low stressor levels (or concentrations) organisms not only repair stressor-induced damage but also overcompensate and reduce background stress or damage more effectively than before the stress occurred. Therefore, at the low end (i.e., low concentrations or levels) of the dose (or concentration)- response curve there would be less of a biological effect than in the untreated controls but as dose increases, the capacity to overcompensate would be saturated (or overwhelmed) and adverse toxicological effects would occur. The reciprocal of this U-shaped curve shows an area of stimulation above the controls at low stress levels and then a region of toxicity as levels increase. The hormesis hypothesis has thus far not achieved much recognition in ecotoxicology or in the ecological risk assessment (ERA) process.

Terms and principles are discussed along with their implications. Chapman points out that problems associated with defining hormesis in ecotoxicology relate to inadequate study design (e.g., orientation towards higher dose testing to define NOECs). Ecotoxicity tests must incorporate multiple doses below the NOEC to assess hormesis. If more low doses are not added in toxicity tests hormesis maybe impossible to differentiate from background changes. Obviously adding more doses to regulatory toxicity test designs will take a monumental task. However, if the biological significance and relevance of hormesis in biological systems are delineated along with its potential link to the ERA process and risk management decisions, multiple low dose testing would be required and accepted in regulatory testing.

Hormesis has not been studied in detail because of the obvious short-comings of ecotoxicology. Ecotoxicology is a relatively new discipline that has been largely dominated over the last 25 years by laboratory- and field-testing in response to regulatory requirements. Until recently there has been little influence of and interaction of ecotoxicology with basic ecology. Furthermore, since development of numerical values (e.g., NOEC, LOEC, etc.) have been the major focal endpoints of ecotoxicity testing the mode of action of many chemical stressors in ecological receptors is largely unknown. Cause-effect relationships are studied in ecotoxicity tests using standardized protocols, often with little relevance to actual systems. Although it may be difficult even with microcosm and mesocosm studies to simulate and replicate natural systems it is imperative that we understand the modes action of chemical stressors in different ecological receptors before chemicals are registered. The mode of action of most pesticides in registration programs is only defined in target and mammalian species. Understanding the presence and significance of hormesis in biological systems will assist in defining and understanding mode of action. Chemical stressors that elicit hormesis in organisms may very well be operating at more than one mode of action; one mode of action at a low level of stressor exposure and another at a higher level of the stressor. Alternatively, if a stressor does not elicit hormetic responses (i.e., at low levels) this may be indicative of a single mode of action at only higher levels of stressor exposure.

Hormetic responses have been predominantly studied and observed in the laboratory. They have not been adequately studied in the field. Examples of ecosystem hormesis responses are mentioned (e.g., increased phytoplankton production as a function of copper exposure). However, are these hormetic responses or indirect (secondary) effects resulting from a direct effect? Hormesis will obviously be easier to study in the laboratory than in the field. It will have great application in understanding the stress syndrome in cause-effect studies. Whether it will be useful in the ERA process as Gentile and van der Schalie (2000) point out is not clear. In the ERA process, field and/or simulated field studies are important. However, the emphasis in the field should not be on cause-effect relationships but rather on understanding typical patterns and distributions to detect regularities in ecosystem dynamics and the nature of xenobiotic versus natural changes. If the latter is indeed accepted then hormesis may play only a minor role in the ERA process.

Chapman points out that hormesis is an adaptation (genetic) versus an acclimation response and that because of this it appears "to occur without appreciable metabolic cost". In the literature the energetic endpoints that have typically been used to detect toxicant-induced stress include adenylate energy charge, energy reserves (e.g., glycogen, lipids, protein, etc.), enzymes linked to intermediary metabolism, scope for growth, etc. Is there sufficient data available to conclude that hormesis is an adaptative response based on using these endpoints? Have hormesis studies been done with multiple generations in a species?

The concluding section on paradigm shifts raises some very significant issues that the scientific community is generally aware of and in agreement with- less importance should be placed on point estimates, greater understanding of the needs and ecology of organisms (e.g., toxicity occurs typically at higher doses but can occur when essential elements are left out) in ecotoxicity tests, and use of ecologically relevant toxicity studies. The ecotoxicology community should encourage greater participation of ecologists in research and at SETAC meetings. Ecotoxicity studies should be designed to not only incorporate more (and realistic) exposure concentrations but also to investigate: the interaction of chemical and non-chemical stressors including the significance of background low level concentrations of chlorinated hydrocarbons in the environment and their interaction with other stressors, biological effects assessment in oligotrophic versus mesotrophic and eutrophic systems.