The Implications of Hormesis to Ecotoxicology and Ecological Risk Assessment

John H. Gentile, Ph.D.

Senior Scientist

University of Miami

Center for Marine and Environmental Analysis

Rosentstiel School of Marine and Atmospheric Science

4600 Rickenbacker Causeway

Miami, Fl 33149-1098

Tel: 305-361-4152

Fax: 305-361-4077


The contribution by Chapman provides an important perspective on the potential role of hormesis in ecotoxicology and ecological risk assessment. Though there are many issues presented in this paper, I will limit my comments to the detection, measurement and interpretation of hormetic responses in ecotoxicology and ecological risk assessments and issues of scale, uncertainty, and predictability.

The strength of the paper lies in it's rationale and recommendations for a more comprehensive study of the role of hormesis within the field of ecotoxicology, particularly in light of proposed changes to the analysis and interpretation of ecotoxicity data and their use in regulations (Chapman et al. 1996, Chapman 1998, 1999). However, there are several issues that need to be addressed explicitly if this "paradigm shift," as the author characterizes it, is to be successful.

Given the evidence of hormetic-like responses for a wide range of organisms, endpoints, and stressors (Calabrese and Baldwin 1997) it is tempting to assume that hormesis is a generalized ecological phenomenon. In fact, we have neither a theoretical or mechanistic basis for such an assumption. What we do have is a similarity in stressor-response curves that may have totally different underlying mechanisms. As Chapman points out, Stebbing (1998, 2000) has suggested that control systems regulate growth in organisms and their behavior and tolerance to differential load (e.g. stressor intensity and duration) may provide a basis for the interpretation of hormesis, acquired tolerance to stress, and the effects of combinations of stressors. Stebbing's work begins to provide the type of mechanistic framework for a generalized theory for hormesis that is amenable to a variety of testable hypotheses. If successful, this would provide the theoretical foundation and credibility that proponents of hormesis are seeking and which Chapman rightly suggests is a necessary step in gaining more widespread acceptance.

The ecotoxicity component of Chapman's paper explores the potential role of hormesis in the current debate on the utility of commonly used regulatory endpoints (e.g., NOECs, ECs, etc.) (Chapman 1998, 1999). The author makes several important points regarding the detection and measurement of hormetic responses including: 1) the need for a change in both the design of ecotoxicity tests to account for potential non-linearity at low concentrations; 2) the more widespread application of existing analysis methods (Van Ewijk and Hoeskatra, (1993), specifically generalized linear models (Kerr and Meador, 1996, Bailer and Oris, 1997); and 3) the development of additional biostatistical tools. I would like to expand this discussion to include an overarching issue that the author briefly mentioned: the role of response variability and statistical power in interpreting the significance of hormetic responses.

While the magnitude of stimulatory response attributable to hormesis has been known to range as much as several fold, the prevailing wisdom suggests an average of 30-60% stimulation above the control (Calabrese and Baldwin 1997). The immediate question that comes to mind is "What is the significance of this change?" which in turn suggests that we need to put these values into a context that compares the typical sources of variability in ecotoxicity tests and in natural systems.

An examination the variability of "unperturbed systems" under controlled laboratory conditions suggests that on average up to a 20% range in control survival is acceptable (Chapman, this issue). If this is true, do we have the "power" in our current experimental designs to detect a delta of 10-40% (hormetic range) change in survival with any degree of confidence? This is not a trivial issue when one considers that there often is in excess of a two-fold range in survival from intra-, and inter- laboratory calibration studies. The same question of other endpoints commonly used in ecology, such as growth and reproduction.

From experimental data we know that the variability in reproduction is likely to be greater than survival for aquatic organisms (Suter 1993).

In addition there is natural variability which can have diurnal, annual, and inter-annual components that are likely to exceed what has been observed under controlled laboratory conditions and potentially mask any hormetic responses. Given this variability and the diversity of species in an ecological community, it will confound the interpretation and importance of hormesis in multi- species ecological assessments. However, if protection of a single threatened or endangered species is the goal, then hormesis may well be important to consider as long as one uses an appropriate demographic model to interpret the population-level significance of any non-linear/non-monotonic toxicological responses (Caswell 2000).

As Chapman points out, what differentiates ecological assessments from health assessments is scale and complexity. Unlike health assessments that deal with individuals of a single population or sub-population, ecological assessments deal with populations of single species but most often multispecies communities, ecosystems, and landscapes. While it appears plausible that hormetic effects can be incorporated into assessments of populations of individual species (e.g., threatened or endangered) it is seems unlikely that a similar successful application can be made at higher orders of ecological scale and complexity (Gentile and van der Schalie 2000). Consequently, this limits the role of hormesis in the field of ecology and ecological risk assessments unless the theoretical basis for such paradigms as the "intermediate disturbance hypothesis" (Tansley and Adamson 1925, Collins et al., 1995) can be developed and applied more fully (Bartell 2000).

Chapman presents an interesting hypothesis for the use of hormesis in risk assessment. He suggests that it can be treated similarly to that for essential elements. Using this analogy, deficiencies and high concentrations can cause adverse effects while intermediate low concentrations can be stimulatory. While I agree in principal with this concept, I do not share the author's conviction that identifying hormetic thresholds "effectively replaces uncertainty factors." While it might when considering NOEC or ECs, there are clearly other issues such as extrapolation across endpoints, from tested to untested species, laboratory to field, etc., that will require continued use of uncertainty/safety factors.

Finally, being able to a priori forecast which stressor-response relationships are the most likely to result in a hormetic response is the nexus of the problem. The ability to make this forecast in turn is dependent either upon a large empirical data base on hormetic responses or an underlying "mechanistic theory" (e.g., Stebbing's growth control mechanism) that allows one to judge from a chemical's structure whether it will have hormetic activity and, if so, what is the appropriate biological response to employ. This "structure-activity" like data base would permit the a priori determination of whether the potential for hormesis existed and the desirability of incorporating it into ecotoxicty based criteria and standards and ecological risk assessments (Gentile and van der Schalie 2000). This, in my opinion is where future research on hormesis should be directed. Without this basic information it will continue to be difficult to focus effort and resources on those assessments where hormesis might make a significant difference.


Bailer AJ, Oris JT. Estimating inhibition concentrations for different response scales using generalized linear models. Environmental Toxicology and Chemistry 1997; 16: 1554-1559.

Bartell SM. Are ecosystems hormetic? Human and Ecological Risk Assessment 2000; 6:237-243.

Calabrese EJ, Baldwin LA. The dose determines the stimulation (and poison): Development of a chemical hormesis database. International Journal of Toxicology 1997; 16:545-559.

Calabrese EJ, Baldwin LA. History of chemical hormesis. Human and Experimental Toxicology 2000; 19: 2-31.

Caswell H. Matrix Population Models, 2nd Edition. Sinauer Associates, Sunderland, MA, 2000.

Chapman PM. New and emerging issues in ecotoxicology ­ The shape of testing to come? Australasian Journal of Ecotoxicology 1998; 4:1-7.

Chapman PM. Whole effluent toxicity testing ­ usefulness, level of protection, and risk assessment. Environmental Toxicology and Chemistry 1999; 19:3-13.

Chapman PM. Integrating toxicology and benthic ecology: Putting the eco back into ecotoxicology. Marine Pollution Bulletin 2001b (in press).

Collins SL, Glenn SM, Gibson DJ. Experimental analysis of intermediate disturbance and initial floristic composition: decoupling cause and effect. Ecology 1995; 76: 486-492.

Gentile JH, van der Schalie WH. Hormesis and ecological risk assessment: fact or fantasy? Human and Ecological Risk Assessment 2000; 6:227-236.

Kerr DR, Meador JP. Modelling dose response using generalized linear models. Environmental Toxicology and Chemistry 1996; 15:395-401.

Stebbing, ARD. Maia Hypothesis-Growth Control and Toxicology. Human and Ecological Risk Assessment, 2000; 6:301-311

Stebbing ARD. A theory for growth hormesis. Mutation Research 1998; 403:249-258.

Suter GW II. Ecological Risk Assessment. Lewis Publishers, Chelsea, MI, 1993.

Tansley AG, Adamson RS. Studies on the vegetation of the English chalk. III. The chalk grasslands of the Hampshire-Sussex border. Journal of Ecology 1925; 13:177-223.

Van Ewijk PH, Hoekstra JA. Calculation of the EC50 and its confidence interval when subtoxic stimulus is present. Ecotoxicology and Environmental Safety 1993; 25:25-32.