Hormesis in Ecological Risk Assessment: a Useful Concept, a Confusing Term, and/or a Distraction?

Charles A. Menzie, Ph.D.

Menzie-Cura & Associates

1 Courthouse Lane

Chelmsford, MA 01824

Tel: 978-322-2856

Fax: 978-970-2791

Email: charliemen@aol.com

This paper is written in response to Chapman (2001) who provides useful insights into the application of hormesis concepts to ecotoxicology and ecological risk assessment. I found myself agreeing with some points, disagreeing with others, and wondering whether the term or the concept of hormesis was useful for scientific inquiry and environmental policy.

The term ­ which refers to stimulatory responses at low doses and inhibitory or otherwise negative effects at high doses has had an interesting history (Calabrese and Baldwin, 1999). In formulating my comments I considered the reasons why hormesis is of interest and I surmise it is largely because of the substantial uncertainty associated with low-dose exposures. This uncertainty stems, in part, from extrapolations currently made in human health risk assessments wherein data are available at high or intermediate doses but the effects of interest are those associated with much lower doses. Scientific policy decisions have typically involved the application of models to extrapolate from the data range to these low doses. In the case of cancer risk, these models have typically assumed no-threshold effects and linear extrapolations. Much debate has centered on such extrapolations. The uncertainties associated with application of these models are among the reasons why Deisler (2000) encouraged the exploration of hormesis.

Assuming that policy and technical issues surrounding low-dose extrapolations in human health risk assessment are the genesis of recent discussion concerning hormesis, how do such discussions bear on ecotoxicology and ecological risk assessment? I explore this by first separating the term "hormesis" from the concept that low, intermediate, and even high levels of exposure/stress in ecological systems can have apparent stimulatory or beneficial effects. There are numerous cases where such stimulatory or potentially beneficial effects occur but they are not explicitly gathered under the term hormesis and I am not certain that applying that term to describe these various phenomena serves a useful purpose. It is useful to recognize these phenomena when assessing exposure/stress at the individual, population, or ecosystem level. The fields of ecotoxicology and ecological risk assessment are moving slowly in this direction. This does not require a paradigm shift as Chapman suggests but rather a more complete understanding of the relationships between exposure and effects at various organizational scales. The need for developing a better understanding of such relationships is receiving considerable attention and is not ignored. What lags the current scientific inquiry is a shared understanding of these phenomena by many in the technical arena as well as many policy makers. Obviously this can also affect the degree to which funding is provided for research in the area.

I've organized my comments around three categories with varied environmental policy implications:

1. hormetic phenomena that are currently recognized and used in environmental policy;

2. hormetic phenomena associated with natural defense mechanisms that could be important but are currently debated or rarely considered when setting environmental policy;

3. hormetic phenomena associated with ecotoxicological investigation.

Hormetic phenomena are currently incorporated into ecological risk assessment where such responses are known as in the case of essential nutrients. Zinc is a good example. This metal is an essential nutrient at low doses but can be toxic at high doses in soil, sediment, and aquatic environments. This hormetic phenomena is explicitly considered in current environmental policy. The recent U.S. Environmental Protection Agency's effort to develop ecological soil screening levels (Eco SSLs) recognizes that certain metals such as zinc have stimulatory effects within a certain dose range and toxic effects at high levels. This information is being used in a policy sense to guide the development of appropriate Eco SSLs for zinc and other metals known to have stimulatory/beneficial effects at low doses. The importance of numerous natural phenomena ­ e.g., periodic flooding and fire ­ are also recognized for their long-term stimulatory and beneficial effects at the community and ecosystem level. As a result, allowance of some level of stress (e.g., fire and flood) has gained acceptance in various environmental policies related to land and water management. The mechanisms underlying simulation of plants by zinc and re-growth of forests following fire are very different and for this reason the application of the term "hormesis" may confuse rather than clarify. Nevertheless, what they share in common is the important idea that exposure and response are not always directly related (i.e., the idea that less exposure is better.) In fact, many cases can be found in the ecological literature that point to the importance of chemical, radiation, and physical stresses on the manner in which populations, communities, and ecosystems develop and are sustained. This is the way the world works and such phenomena are generally well recognized by ecologists and evolutionary biologists. With respect to the occurrence of these well-recognized hormetic phenomena, perhaps it is beneficial to underscore that "less is not always better and may in some cases be worse." This is a matter of education and discussion that may need to occur on a case by case basis. However, nesting all these phenomena under the term "hormesis" might confuse the idea with a common mechanistic process and could detract from addressing the varied and important underlying processes.

My second category concerns situations where hormetic phenomena may be present but are rarely considered in environmental policy or are the subject of ongoing debate. This includes situations where: a) exposure to a stressor has occurred during a species evolutionary history such that the organism has developed defense mechanisms, b) the stressor is not associated with a clear benefit (e.g., such as an essential nutrient) and c) the stressor would usually be considered harmful in the current regulatory environment. Examples of such stressors include exposure to low doses of radiation, metals, petroleum hydrocarbons and pyrogenic hydrocarbons (e.g., polycyclic aromatic hydrocarbons.) Most species have experienced exposure to one or more of these stressors during their evolutionary history and to various degrees have evolved physiological or other mechanisms to deal with the stress. Within communities of organisms (e.g., bacteria and fungi) some species have developed these mechanisms sufficiently to use the stress as a source of nutrient as in the case of biodegradation of petroleum hydrocarbons. (Note such biodegradation mechanisms ­ stimulation at lower doses where toxicity may exist at high levels ­ are currently incorporated into a number of environmental policies related to risk assessment and remediation and fit into the first category.)

For this second category, possible environmental policy considerations include:

1. species can benefit from stimulation of their defense mechanisms following low-level exposures to a chemical, biological, or physical stressor; and,

2. species have a capacity to deal with exposures to low levels of exposure because they have evolved defense strategies.

It is reasonable to presume that organisms that have developed physiological defense mechanisms might benefit by having those systems stimulated by low dose exposures. Stimulation of such metabolic, sequestration, and immune defense mechanisms can be shown in the laboratory as well as in organisms collected from the field. Certainly, stimulation of immune systems often follows low-level exposure to a chemical or biological stress. Such stimulation readies organisms for dealing with variations in exposure during their life history. While the importance of stimulating an organism's defense mechanisms is clear, the value and acceptability of having this stimulation occur as a result of exposure to anthropogenic sttresses is less so. Chapman (2001) states that ­ given the possibility of hormesis ­ the current emphasis on zero discharge of any chemical may well be less than beneficial to at least some if not many organisms. In other words, low level exposures associated with anthropogenic stresses such as wastewater discharges could have a potential beneficial rather than deliterious effect on natural systems. My initial reaction is that it would be difficult to incorporate these potential positive effects into environmental policy for two reasons. First, it would be difficult to demonstrate that natural populations were in need of this anthropogenically-induced stimulation. Second, it would be difficult to demonstrate a benefit beyond allowing the organisms to deal with the anthropogenic exposure.

The idea that species have evolved defense mechanisms that buffer them against stressors has already been considered in some environmental policies, most notably those that account for "assimilative capacity". This refers to the ability of an organism, population, or system to withstand certain levels of stress. While the concept of assimilative capacity embodies a variety of mechanisms, it includes consideration of metabolic and sequestration processes such as those involved in the breakdown of hydrocarbons and the binding and storage of metals. Proponents argue that organisms, populations, and systems are able to assimilate a certain level of environmental stress because they have systems designed to handle these stresses. If the assimilative capacity can be known, then it is possible to use this information to manage waste disposal and other activities in a manner that does not impair individual populations or ecosystem function. Opponents argue that there are large uncertainties associated with the behavior of natural systems and that anthropogenic environmental stress might exact energy and fitness costs that are not readily apparent. The degree to which the environment can assimilate or buffer anthropogenic stress will be an ongoing science policy debate. A better understanding of how organisms and systems respond to low levels of anthropogenic stress including hormesis can help inform that debate.

Chapman (2001) argues that hormesis in ecotoxicology is a particularly important issue that is not well understood. The major advantage appears to be that explicit consideration of hormesis provides a boundary for judging low-level exposures. The presence of hormesis could support the perspective that some level of contaminants can exist in the environment without having harmful effects. This is consistent with the concept of assimilative capacity. Demonstrating hormesis in ecotoxicology testing would presumably provide a technical basis for arguing the technical validity of incorporating assimilative capacity concepts in environmental policy. I agree with Chapman that this is an area where additional research and technical policy development are needed. The major hurdle for gaining acceptance is likely to be the uncertainty associated with extrapolating laboratory studies of hormesis to natural populations in the field.

Based on the forgoing discussion and reviews of Chapman (2001), Gentile and van der Schalie (2000) and Bartel (2000) I offer the following conclusions concerning hormesis in ecotoxicology and ecological risk assessment:

1. Recent discussions of hormesis have highlighted the fact that "less exposure/stress is not always better and in some cases can be worse." This is an important lesson with associated environmental policy implications. Clear examples can be found where such knowledge has been incorporated into environmental decisions. However, there could be other cases where this lesson should apply that are currently unrecognized. Recognizing such cases would benefit from a greater understanding by policy makers and scientists of the diversity of exposure response relationships.

2. Exposure response relationships can vary considerably depending on the stressor, number of stressors, receptors, and spatial and temporal scales. As is clear in Chapman (2001) and Bartell (2000) a variety of these relationships exhibit "hormesis". However, application of that term to relationships with very different underlying mechanisms may confuse rather than clarify. It may be more productive to develop and communicate the idea that exposure and response do not always involve the familiar relationships presented in toxicology texts and policy documents. Discussions of hormesis can serve to promote this idea. However, promoting "hormesis" as an organizing construct could detract from the key idea and the diversity of underlying mechanisms.

3. For chemicals commonly considered harmful with no apparent benefit (e.g., they are not essential nutrients), the presence of hormetic phenomena is most likely to be useful for gaining acceptance of environmental policies that consider the assimilative capacity of organisms, populations, or systems. It is less likely that laboratory observations of stimulation relative to controls can be argued as a benefit that should be taken into account in environmental policy.


Bartel, S.M. 2000. Are ecosystems hormetic? Human and Ecological Risk Assessment. 6: 237-243.

Calabrese, E.J. and L.A. Baldwin. 1999. Tales of two similar hypotheses: the rise and fall of chemical and radiation hormesis. BELLE Newsletter 8(2).

Chapman, P.M. 2001. Implications of hormesis to ecotoxicology and ecological risk assessment.

Deisler, P.F. 2000. A personal perspective on hormesis. Human and Ecological Risk Assessment. 6: 245-248.