Implications of regulating environmental contaminants on the basis of wildlife populations and communities

To the Editor:

Ecotoxicological research examines toxic effects at different scales, from geographic regions to individual organisms [1], cutting across fields from landscape ecology to biochemistry. Most regulatory decisions involving wildlife and environmental contaminants have been based on overt injury to individuals observed in either field or laboratory studies. With the exception of the banning of DDT, regulatory decisions in general have not been based on effects on wildlife populations or communities. This may change soon. We have participated in recent conferences and workshops (Environmental Contaminants and Terrestrial Vertebrates: Effects on Populations, Communities, and Ecosystems, University of Maryland, College Park, MD, October 19–21, 1998) that discussed whether wildlife mortality should be considered unimportant unless it can be related to community structure or large-scale population trends. Some biologists also have asserted that mortality from environmental contaminants should be put on the same basis as other forms of mortality, such as disease or predation.

For several years, ecotoxicologists have been deliberating the appropriate end points for risk assessments. The most recent guidance on superfund sites from the U.S. Environmental Protection Agency (Issuance of final guidance: Ecological risk assessment and risk management principles for superfund sites, October 7, 1999) states as Principle 1 that “Superfund's goal is to reduce ecological risks to levels that will result in the recovery and maintenance of healthy local populations and communities of biota.” This principle may be difficult to apply. We have written this letter to encourage discussion on the implications of applying population and community concepts to the regulation of chemicals and sites.

Ecotoxicologists will continue to disagree over whether the individual, the population, or the community is the appropriate ecological unit to consider in regulatory assessments. The biologist who studies local populations, for example, may be quick to dismiss as unimportant the findings of a pathologist who has found lesions in individual animals. That same population biologist, in turn, may find his work on changes in local populations dismissed by a population biologist who models regional populations. A community ecologist is likely to go one step further in determining what is ecotoxicologically meaningful, and be willing to accept the loss of some populations if community function remains unchanged. More often than not, these differences of opinion are related to the training and professional circumstances of the biologist. Because we cannot single out any of these views as the most legitimate on scientific grounds, we suggest that regulatory decisions ought to be based, at least in part, on the public's views. Unless ecotoxicologists are willing to let their field become irrelevant to society, they should be aware of what the public thinks is an appropriate basis for regulation.

The debate over the use of DDT illustrates what the public values. When Americans became upset over the deaths of robins on a midwestern university campus after the robins ate earthworms contaminated with DDT [2], the public did not demand a demonstration that a measure of ecosystem function showed a statistically significant perturbation. No economist had to estimate the cost of a few dozen dead robins. The public simply thought that finding poisoned robins meant that we were doing something to the environment that we should not be doing. Without the need for a formal risk assessment, the public was upset and reacted to finding robins strewn belly-up across lawns.

The public's reaction to finding tumors in fish from polluted waters provides a similar illustration. An economist could estimate the loss associated with fish tumors, based on the cost of fish from the market, the cost of fuel for a boat, and so forth. However, to the unfortunate fisherman repulsed by the sight of a tumor bulging from the mouth of a fish, this kind of assessment would be irrelevant. He will be left with a lingering feeling of repulsion and may never return to that fishing spot. Although modeling fish population dynamics may help make sound environmental decisions, it would not change that fisherman's outlook. Economists and regulators would prefer to evaluate environmental injury objectively and quantitatively, rather than with the imprecision of emotion. However, the powerful feelings evoked by fish tumors and dead robins drive the public's willingness to clean up polluted soil and water.

Biologists use the word “population” to cover dissimilar concepts [3], complicating attempts to apply population biology to risk assessment. Although population has a simple definition—organisms in a species living in the same area—the term “area” has been used by biologists to mean anything from a small local site to the entire range of the species. Fundamental differences in thinking lie behind these meanings that are obscured when different concepts of population are lumped together under the same banner. The concept of a population at a small site is not very different from the concept of individual animals at that site. The concept of a population over a large area leads risk assessment in a new direction. When a risk assessment goes beyond the limits of a contaminated area to model regional populations, the toxic effect of the contaminant may be an inconsequential part of the model compared to other ecological variables, variables that often are poorly understood. The results of such a risk assessment will depend more on the viability of the species chosen rather than on the presumed toxic effects. We disagree with this way of thinking and suggest that if toxicologic findings of mortality and serious sublethal effects are widespread enough, they should be sufficient to move the regulatory process, regardless of regional population trends. Would our concerns about the toxic damage at Love Canal have been reduced if risk assessors had put the effects at this site in the context of all of New York State? Would we be less concerned about a local fish kill if we tallied the fish remaining in nearby rivers? Why then, would a regional approach make sense for wildlife? In our opinion, dieldrin and DDT are probably the only pesticides for which we might attempt to argue for significant regional effects on wildlife populations [4]. If regulators apply a broad regional population approach then we know in advance that it is senseless to even evaluate wildlife mortality at local sites. In our opinion, the public would not agree with that approach, particularly that segment of the public living near polluted sites.

The public makes value judgments about wildlife mortality. Most Americans seem willing to accept waterfowl mortality from hunting, for example, but not waterfowl mortality from ingesting spent lead shot. The public does not consider all kinds of wildlife deaths equal, and it cares more about some species than about others. This brings us to community effects, the level above population effects. Community-level indices and functional analyses may hide effects on individual species. For example, eliminating peregrine falcons (Falco peregrinus) from the East Coast through pesticide poisoning would have virtually no effect on avian species diversity indices. Replacing California condors (Gymnogyps californianus) poisoned by environmental contaminants with local vultures would scarcely change energy dynamics and nutrient flow through the ecosystem. Protecting community functions will not necessarily protect particular species. In proposing the use of community concepts in ecotoxicology, biologists seem to be accepting the proposition that the elimination of wildlife species is acceptable if those species would be replaced by other species performing similar ecological functions. Have the implications of this proposition been considered adequately?

We do not advocate the extreme position that every change in a bioindicator and every death of an animal should be considered significant by regulators. Inevitably, some animals will be poisoned as the price of widespread use of chemicals by our society. We also feel that the tools of population and community biology are often helpful when studying wildlife toxicology, especially for those few pollutants, such as acid rain, that are truly regional. Ecotoxicologists would agree that the most serious contaminant effect is one so severe that widespread populations are reduced for a long time. Regulators should consider population and community data if available. However, our main point is that demonstrating harmful effects on wildlife at the regional population or community level should never become a prerequisite for regulatory action.

The scientific research community, by nature, tends to downplay the more obvious effects of contaminants and to embrace new methods and concepts. The public, by nature, tends to care mainly about environmental damage it can see and animals it considers important. The public has a greater reaction to a picture in a newspaper of a bird deformed by pollution than to technical reports on population models and community function. This difference in outlook between scientific researchers and the public is at the heart of this letter. We are reminded of a down-to-earth pathologist at our Center, who patiently listened to the complex research plans put forth by biologists, and then humbly inquired why we did not just “look for sick and dead animals.”