3
An Overview of Risk Assessment

JOHN D. STARK

Ecotoxicology Program, Department of Entomology, Washington State University

The world we live in is a risky place. Every day the act of getting out of bed and facing the world exposes us to potential harm or even death. From the hole in the ozone layer, global warming, consumption of coffee and alcohol, the drive in the car to work, danger is around every corner—the message is loud and clear: We are at risk! We are deluged with reports from the media about the risks of many of the things we do, consume, or are exposed to every day. Not only are humans at risk, the very world we live in is at risk, and to top it off, we are to blame. Some risks are easy to quantify. The connection of obesity to diabetes, smoking to lung cancer, excessive consumption of alcohol to liver damage, and the time spent driving a car to getting in an accident are quite straightforward and easily quantifiable. However, what about the low levels of pesticides in our diets, exposure to radon in our homes, or exposure to electromagnetic fields? These risks are much harder to quantify. What are the risks associated with the introduction of an exotic species to a country or new geographical region that does not already have this species? How does the introduction of exotic species impact humans, crops, domesticated animals, and the habitats that we wish to protect?

The risks of some behaviors or chemicals to human health are certainly real, but risk has become for some a new-age religion. For example, survivalist groups had formed because they were convinced that the world as we know it was coming to an end due to the Y2K computer bug.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference 3 An Overview of Risk Assessment JOHN D. STARK Ecotoxicology Program, Department of Entomology, Washington State University The world we live in is a risky place. Every day the act of getting out of bed and facing the world exposes us to potential harm or even death. From the hole in the ozone layer, global warming, consumption of coffee and alcohol, the drive in the car to work, danger is around every corner—the message is loud and clear: We are at risk! We are deluged with reports from the media about the risks of many of the things we do, consume, or are exposed to every day. Not only are humans at risk, the very world we live in is at risk, and to top it off, we are to blame. Some risks are easy to quantify. The connection of obesity to diabetes, smoking to lung cancer, excessive consumption of alcohol to liver damage, and the time spent driving a car to getting in an accident are quite straightforward and easily quantifiable. However, what about the low levels of pesticides in our diets, exposure to radon in our homes, or exposure to electromagnetic fields? These risks are much harder to quantify. What are the risks associated with the introduction of an exotic species to a country or new geographical region that does not already have this species? How does the introduction of exotic species impact humans, crops, domesticated animals, and the habitats that we wish to protect? The risks of some behaviors or chemicals to human health are certainly real, but risk has become for some a new-age religion. For example, survivalist groups had formed because they were convinced that the world as we know it was coming to an end due to the Y2K computer bug.

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference Before we can discuss risk assessment, we must have a definition of risk. At its simplest, risk is the probability that harm will occur from a specific act. This definition covers a lot of ground. One can imagine just about anything having a risk associated with it even if it is slight. RISK AND TRADE BARRIERS The process of living exposes all organisms to various risks. For humans some of these risks are self-imposed such as smoking tobacco, which increases the risk of developing lung cancer. However, the risks that are important to trade differ somewhat from the examples mentioned above. Sanitary and phytosanitary (SPS) procedures have been established by many countries to protect their agricultural economy and natural environment (Gray et al., 1998). The goal of these procedures is to limit the entry of foreign pests and diseases in their respective countries. Risks associated with SPS measures are broken down into three categories: Direct food risks—additives, contaminants, toxins, or disease-causing organisms in food. Some examples are hormones and antibiotics in beef, pesticide residues in crops, aflatoxin in grains, Escherichia coli, Salmonella, Listeria, and botulism toxin in various foods, and food additives (colorings, flavor agents, etc.). Introduction of exotic organisms—plant-or animal-carried diseases, pests, diseases, or disease-causing organisms. Some examples are various insect pests that are introduced in produce such as the Mediterranean fruit fly, species that are introduced in something other than a commodity such as the Asian longhorn beetle, infectious agents such as prions that cause mad cow disease, and weed pests like purple loose strife. Damage caused by exotic organisms—by entry, establishment, or spread of pests. Here the concern is the actual damage that may be inflicted on agricultural industries. WHAT IS RISK ASSESSMENT? The National Academy of Sciences defines risk assessment as "the determination of the probability that an adverse effect will result from a defined exposure" (NRC, 1983). Contrary to popular belief, risk assessment is not a science but rather a combination of science and expert judgment. Scientific data are used to develop an assessment of risk, but the risk assessor does not often have extensive data and has to make a judgment call. Furthermore, the type of data available are not uniform for each risk assessment. For some chemicals, for example, a complete toxicological profile will be available whereas for others the data may be much more limited.

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference The process of risk assessment varies from agency to agency within the United States depending on the type of risk being evaluated: pesticide residues in food, hazardous waste sites, introduction of exotic species, etc. However, the same basic principles are usually followed and consist of four steps: (1) hazard identification, (2) dose-response assessment, (3) exposure assessment, and (4) risk characterization. Hazard Identification The first step in risk assessment is to determine whether the agent in question is hazardous. If the agent being evaluated is a chemical, then basic information about its toxicity is required. If the agent is an organism, then basic information about its biology and life history are required. To put this into perspective, at least for chemicals, the toxicity of several common chemicals to rats or mice is listed in Table 3-1. Dose-Response Assessment Here a characterization of the relationship between the dose or concentration and the incidence of adverse effects in exposed populations is developed. This is based solely on scientific data. Dose can be thought of as chemical concentration or the number of individuals of an exotic species that are introduced to a geographic area over time. The most commonly used measure of toxic effect is the LD50. The LD 50 is a statistically derived measure of the dose-response relationship and is an estimate of the lethal dose that causes 50 percent mortality of a group of organisms being studied. Other dose-response measures are estimated in the same manner; for example, a dose that causes 50 percent reduction in offspring. If a large enough group of organisms is exposed to increasing concentrations of a poison, a sigmoid curve is obtained when the cumulative percent affected (dead) is plotted against a dose or concentration (Figure 3-1). At low concentrations no effect is observed, but as the concentration increases, some of the organisms begin to respond. The highest concentration where no effect is observed is the no observable effect concentration or level. The threshold is the lowest dose that elicits a response or the lowest observable effect concentration. Eventually a dose is reached that kills all of the organisms being evaluated (maximum effect). It is at this point that an increase in dose can have no further effect. It is difficult to derive the LD50 or the slope of the dose-response line from a sigmoid curve. Also, data points along the dose-response line below the threshold and above the maximum effect do not provide data that can be used in the estimation of the LD50. Methods have been developed to straighten the dose-response curve and estimate the LD50. The first statistical approach for dose-response data was proposed by Trevan (1927), but many

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference Table 3-1. Acute LD50a Values of Selected Common Chemicals   Oral LD50 (rat or mouse) Chemical (mg chemical/kg body weight) Botulism toxin <0.001 mg/kg Aflatoxin B1 9 mg/kg Sodium fluoride 180 mg/kg Tylenol 338 mg/kg Diazinon 350 mg/kg Aspirin 1,500 mg/kg Malathion 2,800 mg/kg Table salt (sodium chloride) 3,750 mg/kg Ethanol (alcohol) 10,600 mg/kg a LD50 is the lethal does that kills 50 percent of a population. FIGURE 3-1. Dose-Response Relationship

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference modifications have been made over time (Finney, 1971). Acute toxicity data are usually analyzed by probit or logit analysis (Finney, 1971). SELECTING TOXICOLOGICAL ENDPOINTS: WHAT DO WE EVALUATE? Toxicological data used in risk assessments fall under two categories—acute and chronic exposure data. Acute data are generated after a single exposure to a chemical for a short time period. Chronic data are generated after repetitive exposure to concentrations over many days to a lifetime. Mortality is the endpoint of interest for many acute studies whereas life span, reproduction, weight gain, cancer, and birth defects are of interest in chronic studies. To establish the amount of pesticide that can be ingested over a lifetime without causing illness, the lowest no observable effect level (lowest value for the endpoints studied—cancer, offspring, life span, etc.) is divided by a safety factor of 10–1,000 which results in the reference dose (RfD). The RfD is the dose or concentration below which daily aggregate exposure over a lifetime will not pose an appreciable risk to human health. What is an appreciable or acceptable risk? One new case of cancer in 1,000,000 people is considered acceptable. The reason that the safety factor varies has to do with the type of data available. Very little human toxicological data are available because we do not conduct toxicity studies with humans. What human data we do have usually come from accidents, suicides, or worker exposure. Therefore, extrapolation from animal data is often necessary. If human epidemiological data are available, then a safety factor of 10 might be used. If animal data are the only data available then a factor of 10 for the lack of human data is multiplied by 10 for animal data, resulting in a safety factor of 100 (10 × 10). The type of animal data available also reduces the safety factor. Chronic data result in a lower risk factor than acute data. Exposure Assessment A measure or estimate of the intensity, frequency, and duration of exposure agents is estimated in the exposure assessment part of risk assessment. All potential routes of exposure are considered in exposure assessment. For example, the likelihood of contact with the chemical through exposure to contaminated soil, water, air and/or food is evaluated. For chemicals this involves characterization of the exposure setting. The following questions are then asked: Where is the chemical likely to be found: water, soil, air, food? Are the organisms at risk aquatic or terrestrial or both? How will they be exposed? Through drinking, eating, breathing, dermal contact?

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference How long will the exposure last? A measure of the amount of chemical likely to be encountered in an environmental medium such as river water is estimated using data and mathematical models. For organisms such as disease organisms, the likelihood of exposure of a susceptible population over time is estimated. Risk Characterization— What Are the Consequences? Is There a Problem? Risk characterization integrates the toxicity data and exposure assessment to arrive at probabilities of effects occurring. There are several approaches to risk characterization. Perhaps the simplest is the quotient or ratio method (Barnthouse et al., 1986; Urban and Cook, 1986; Nabholz, 1991): The estimated concentration likely to be encountered is divided by the concentration estimated to cause a toxicological effect to arrive at the quotient. Quotients of 1 or greater imply a risk whereas quotients lower than 1 indicate less risk. For example, the LD50 for a pesticide to a fish species is 0.075 mg/l. The estimated environmental concentration is 0.1 mg/l. Using the quotient method, we find that 0.1/0.075 = 1.33, which means that this pesticide poses a risk to the fish species. DETERMINISTIC RISK ASSESSMENT The most commonly used methods of risk assessment today are deterministic and probabilistic risk assessment. A risk assessment based on a point estimate is called a deterministic risk assessment. Deterministic risk assessments are based on a single estimate of exposure (usually the worst-case scenario) and therefore do not provide information about variability and uncertainty that may be associated with a risk. The quotient method mentioned above is a type of deterministic risk assessment. However, deterministic risk assessment is often based on a tiered decision-making progress whereby a series of decisions are made based on the outcome of a previous result. As an example, imagine that a pesticide is registered for use on a hypothetical crop (crop A). The maximum allowable residue for the pesticide on crop A is 5µg pesticide/g of crop. An assumption is made that all of crop A is always sprayed with the maximum amount of pesticide allowed by the pesticide label, and thus the residue of pesticide present on the crop is also at a maximum. If the highest consumption of crop A is 10 g/kg body weight/day, then to arrive at the risk characterization, 5 µg pesticide/g of crop A × 10 g crop A/kg body weight/day = 50 µg pesticide/kg body weight/day. Exposure is then compared to the RfD. If

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference the RfD is 75 µg pesticide/kg body weight/day, then there is no appreciable risk associated with this pesticide. Note that the assumptions made all err on the high side (are conservative). Refinements are made using the tiered approach that may result in less conservative (more realistic) estimates of risk. PROBABILISTIC RISK ASSESSMENT An assessment based on the probability of occurrence is called a probabilistic risk assessment. This method gives a measure of risk and the associated probabilities of their occurrence. Using the pesticide and crop example above, data collected by the Food and Drug Administration (FDA) indicate that only 50 percent of the population consume crop A on any given day, and the amount consumed varies from 2, 4, 6, 8, and 10 g/kg body weight/day. Pesticide residues on the crop that are treated with the pesticide vary from 1, 2, 3, 4, and 5 µg pesticide/g of crop. Data from the Environmental Protection Agency indicate that only 25 percent of the crop is sprayed with the pesticide in question. The above data are run through a computer program (Monte Carlo simulation is an example of such probabilistic approaches) and the following exposures are generated: 78 percent of the population is not exposed to the pesticide, 1 percent is exposed to the highest exposure level (50 µg/kg body weight), 5 percent is exposed to 40 µg/kg body weight, 7 percent is exposed to 30 µg/kg body weight, and 9 percent is exposed to 10 µg/kg body weight. The output is a distribution of risk values with a probability assigned to each estimated risk. Variability and uncertainty associated with the risk are part of the assessment (Hattis and Burmaster, 1994; Rai and Krewski, 1998). The general consensus among risk assessors is that probabilistic methods result in a risk assessment that is more realistic than a deterministic risk assessment. PROTECTING HUMANS, PLANTS, AND WILDLIFE Human Health— Pesticide Residues in Food Obviously, protecting human health is the major concern for many risk assessors. Protection of human health, however, must be looked at in several different ways. The first and most obvious is direct protection, which is protection from disease-causing organisms and poisonings. Examples might be protection from diseases caused by organisms such as E. coli, Salmonella, Listeria, and parasites; and protection from toxins such as aflatoxin, botulin, pesticides, and hormones. To protect human health, consideration must also be

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference given to clean water and air. Also, indirect protection involves protection of the human food supply and thus protection of crops and domestic animals. As mentioned above, the reference dose is the amount of pesticide residue that can be ingested daily (daily allowable intake) by an average adult without an appreciable risk to human health. How Safe Is Our Food Supply in Terms of Pesticide Residues? The percentage of foods that in 1997 contained pesticide residues is presented in Table 3-2 (FDA, 1998). Fruits had the highest percentage of residues and dairy products had the lowest. Interestingly, imported produce tended to have lower residues than commodities that originated in the United States. This is clearly the opposite of public perception. As can be seen from the data presented in Table 3-2, very little of the agricultural commodities sold in the United States contain pesticide residues that are above the residue tolerance. However, pesticide RfD values are generated separately for each pesticide registration. The problem is that no one knows whether exposure to low levels of many pesticides (all at or below the RfD) can cause health problems. The reason for this is that the cost to do multiple exposure studies is prohibitive. To illustrate this problem, imagine that the cost to conduct a toxicological study for one chemical is $1,000. If we were to evaluate the toxicity of 10 chemicals including all possible combinations of these chemicals, the cost is 10 factorial (10!) × $1,000 or $3,628,800,000. The Food Quality Protection Act (FQPA) was enacted in 1996. This law changed the way that the United States deals with pesticide residues in food. Prior to this law, residues for each pesticide were considered separately. The FQPA mandates that pesticides with like modes of action be lumped together. For example, residues of all organophosphate insecticides are added together to Table 3-2. Pesticide Residues in Agricultural Commodities, 1997   Percent Commodity with Residues   Within Residue Tolerance Above Residue Tolerance Commodity Domestic Imported Domestic Imported Grains and grain products 40 13 0 1 Milk, dairy products, eggs 3 11 0 0 Fruits 55 38 1 1 Vegetables 28 35 2 1 Fish, shellfish, other aquatic products 32 6 0 0   Source: Adapted from FDA (1998).

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference come up with the total amount of exposure. This will result in the exceeding of tolerances for many classes of pesticide. The law then dictates that the total residue must be reduced. Pesticide uses on certain crops will almost certainly be reduced or eliminated all together in order to reduce total crop residues. Therefore, U.S. farmers stand to lose certain pesticides. However, pesticides that may be banned or their uses restricted in the United States will still be used in foreign countries. Some of these countries have developed accurate knowledge of the timing of sprays so that no residue is detected. Plants, Domestic Animals, and Wildlife When we think of plant protection, crops obviously come to mind, but we must also think of the risk posed to plant species that are not crops. Arthropod species and disease organisms that attack plants and weed species that compete with native species and other plants that we wish to protect are the major risk concerns. Domestic animals are also susceptible to arthropod pests such as biting flies and disease organisms. Even weeds can be a big problem for our domestic animals. For example, some weeds are toxic to cattle. No one would argue that protection of humans, domestic animals, crops, and wildlife from harm is important. However, when it comes to protecting wildlife, what should be protected? People in general like birds of prey, songbirds, sea mammals, salmon, and other fishes. But what about spiders, algae, and worms; are they not important as well? The most important species for ecosystem function may not be at the top of the food chain (the large predators). Our biases influence science and the funding of scientific research. Nowhere is this more evident than in environmental research. If two grants are submitted to a granting agency, the first dealing with determining the risks of pesticides to eagles and the second determining the risks of pesticides to soil-dwelling nematodes, guess which grant will get higher priority? Does this mean that eagles are more important to ecosystem function than nematodes? Not necessarily! In fact the nematode species in question may be more important than the eagle, but we assign a value to living things whether we realize it or not. Spiders, mites, algae, insects, and worms are just not high on our list of important species, yet the loss of these very organisms may be devastating to ecosystem function. Loss of eagles, a top predator, may not have much of an impact. Most people would agree, however, that the loss of eagles is unacceptable. We place a very high value on their presence in our world. Values are not the same for everyone. Wolves are a prime example. Ranchers hate them but conservation biologists and environmentalists love them.

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference Introduction of Exotic Species Risk assessment was initially developed by the insurance industry in an attempt to determine life expectancies. A high degree of sophistication has been obtained in risk assessment of chemicals in food and the environment. However, risk assessment for biological hazards is much less developed than for chemical hazards (Powell, 1997). This is due in part to the fact that the risk that introduced species pose is less subject to quantification. For many species we just do not know how well they will adapt to a new environment or whether they will change food sources or evolve. They may also interact with existing species in unpredictable ways. One of the problems with conducting risk assessment of exotic species is determining how many organisms are necessary for establishment. Is one pregnant female of a potential insect pest enough to establish a population? Some pest species are parthenogenetic, that is, they are all females and produce clones of themselves. Thus, only one surviving individual may be enough to establish a population. For others species, many individuals may have to be introduced over time for establishment to occur. A great deal of knowledge about pest biology is therefore essential in developing pest risk assessment (Gray et al., 1998). When it comes to trade of agricultural commodities, quarantines may be put in place that limit export from a particular geographical area (Gray et al., 1998). Postharvest disinfestation procedures, such as fumigation, may also be required (Stark, 1994) as well as inspections at points of export and import (Armstrong and Paull, 1994). Products may be banned if the risk is perceived to be very high or if there is no way to guarantee pest-free produce. The following are some important questions that are asked by risk assessors about potential exotic pest species: Is a pest species present in an exporting country? Can the pest develop on hosts in the importing country? Can the pest species survive transport to the importing country? Are there quarantine treatments in place in the exporting country? How effective are the quarantine treatments? Can the pest exist in the climate of the importing country? We should all be very concerned about the movement of species from one country to another because great economic and environmental damage can happen when exotic species arrive in a new geographic area. One of the greatest threats to wildlife is the introduction of exotic species because some of these species can outcompete native species and change the structure of communities of organisms. Exotic species do not always enter a country directly on or in an agricultural commodity. An example of two species that have invaded the United States not through produce but related to commerce are the Asian long-horn beetle that entered the United States through shipping pallets originating in China, and the zebra mussel that entered the Great Lakes through the discharge

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference of water ballast from ships originating in Europe. The zebra mussel, first discovered in 1988 in Lake St. Clair, is thought to have originated in the Caspian Sea. By 1990, zebra mussels were found in all the Great Lakes. Zebra mussels are pests because they close off water supply pipes of various industries and power plants. They are also destroying native mussel populations through competition and directly by attaching to native mussels (Hebert et al., 1991; Hunter and Bailey, 1992; Nalepa, 1994; Schloesser and Nalepa, 1994; Ricciardi et al., 1995). The Asian long-horn beetle was first discovered in the United States in Brooklyn, New York, in 1996 and has since been reported in Long Island, Chicago, and Bellingham, Washington. This species attacks hardwood trees with a preference for maples. Adult beetles chew holes in the bark and lay eggs. The larvae hatch and eat the bark, making tunnels as they grow. Mature larvae pupate and then the adult emerges from the tree by chewing through the bark. The Asian long-horn beetle kills the trees that it infests and thus is a very serious pest that could devastate many of our hardwood trees. Transport of disease organisms is also a major issue in SPS measures. We only have to look at the recent outbreaks of E. coli-related foodborne illnesses in the United States to realize that food safety is a major concern worldwide. The recent outbreak of mad cow disease in the United Kingdom resulted in trade barriers being erected in other European Union nations. The presence of aflatoxins in grain and peanuts has also been a risk issue. RISK ASSESSMENT OF GENETICALLY ENGINEERED ORGANISMS An area that is already becoming a major trade issue is the importation of genetically modified organisms. One of the worries associated with the use of genetically modified organisms is the spread of genes from one species to another (Kareiva and Stark, 1994). Thus, a crop that is engineered to tolerate herbicides might transfer this gene through pollen to weed species resulting in weeds that are also resistant to a herbicide. The safety of food that has been genetically modified has also recently been called into question. In fact, trade of genetically modified agricultural commodities is presently being debated at the international level. HOW CAN WE BE FOOLED? UNPROVABLE RISKS As mentioned above, because of physical and financial limitations we do not know if exposure to multiple pesticide residues causes health problems. An argument could be made that food that contains several pesticide residues, even if they are at or below the RfD, could cause a health risk. It would be difficult to disprove this argument. A product that may appear harmless might actually cause a problem. And toxicologists know that exposure to low levels of poisons

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference can actually result in increased vigor. How do we figure this into a risk assessment? FUTURE PROBLEMS-SCIENTIFIC ARGUMENTS ABOUT RISK ASSESSMENT Individual versus Population-Level Effects One of the current debates in risk assessment has to do with the endpoints used to evaluate toxic effects. Toxicologists usually study the effects of chemicals in individuals. However, what happens at one level of organization (individuals) does not necessarily translate to another level (populations). The National Research Council (1981) has recommended that chemicals should be studied at the population-, community-, and ecosystem-level, yet few researchers have adopted approaches for the evaluation of chemical effects at levels of organization higher than the individual (Kareiva et al., 1996; Stark et al., 1997; EPA, 1998; Suter, G.W. II. 1999). One thing that may occur at the population level that cannot be accounted for by examining individual mortality and reproduction is ''population compensation." For example, when individuals are removed from the population after exposure to a chemical, survivors have more resources available and may reproduce at a greater rate; offspring may also be larger and more vigorous. Thus, populations may be less susceptible than we predict based on studies with individuals. On the other hand, effects can be masked at the population level and loss of genetic diversity may occur. Population Structure and Susceptibility Susceptibility of a population may be greatly influenced by the structure of the population at the time of exposure. Toxicological studies are almost always conducted for one life stage or age. However, populations in nature often consist of a mixture of stages and ages. A recent study has indicated that the effect that pollutants have on populations is greatly influenced by the initial structure of the population at the time of exposure (Stark and Banken, 1999). These findings have implications for ecological risk and protection of wildlife. Susceptibility of populations in the wild may be greater or less than predicted depending on population structure. CONCLUSIONS Risk assessment is a valuable tool that combines science and expert judgment. Increasingly more sophisticated means of risk assessment have been developed, particularly in the areas of human health and the environment. However, risk assessment of exotic species is much less developed, and more work is needed in this area.

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference Trade disputes over food contaminants such as hormone and pesticide residues and exotic pest introductions have occurred in the past and may continue to be trade issues in the future, particularly in light of the FQPA. However, the issue that will probably dominate future trade disputes is genetically modified organisms. REFERENCES Armstrong, J., and R.E. Paull. 1994. Introduction. In Insect Pests and Fresh Horticultural Products: Treatments and Responses. R. E. Paull and J. Armstrong eds. Wallingford, U.K: C.A.B. International. Barnthouse, L.W., G.W. Suter II, S.M. Bartell, J.J. Beauchampp, R.H. Gardner, E. Linder, R.V. O'Neill, and A.E. Rosen. 1986. User's Manual for Ecological Risk Assessment. ORNL Publication No. 2679. Oak Ridge, Tenn.: Oak Ridge National Laboratory. EPA (U.S. Environmental Protection Agency). 1998. Guidelines for Ecological Risk Assessment . EPA/630/R-95/002F. Risk Assessment Forum. Washington, DC.: U.S. Environmental Protection Agency. Finney, D.J. 1971. Probit Analysis, 2nd ed. Cambridge, U.K.: Cambridge University Press. FDA (Food and Drug Administration). 1998. FDA Pesticide Program: Residue Monitoring. Available online at http://vm.cfsan.fda.gov/~dms/pes97rep.html Gray, G.M., J.C. Allen, D.E. Burmaster, S.H. Gage, J.K. Hammitt, S. Kaplan, R.L. Keeney, J.G. Morse, D.W. North, J.P. Nyrop, M. Small, A. Stahevitch, and R. Williams. 1998. Principles for conduct of pest risk analyses: report of an expert workshop. Risk Analysis 18:773–780. Hattis, D. and D.E. Burmaster. 1994. Assessment of variability and uncertainty distributions for practical risk analysis. Risk Analysis 14:713–730. Hebert, P.D.N., C.C. Wilson, M.H. Murdoch, and R. Lazar. 1991. Demography and ecological impacts of the invading mollusc Dreissena polymorpha . Canadian Journal of Zoology 69:405–409. Hunter, R.D., and J.F. Bailey. 1992. Dreissena polymorpha (zebra mussel): colonization of soft substrata and some effects on unionid bivalves. The Nautilus 106(2):60–67. Kareiva, P.K., and J.D. Stark. 1994. Environmental risks in agricultural biotechnology. Chemistry and Industry 2:52–55. Kareiva, P., J.D. Stark, and U. Wennergren. 1996. Using demographic theory, community ecology, and spatial models to illuminate ecotoxicology. Pp.13–23 in Ecotoxicology: Ecological dimensions. L. Maltby and P. Grieg-Smith, eds. London: Chapman & Hall. Nabholz, J.V. 1991. Environmental hazard and risk assessment under the United States Toxic Substances Control Act. Science of the Total Environment 109/110:649-665. Nalepa, T.F. 1994. Decline of native unionid bivalves in Lake St. Clair after infestation by the zebra mussel, Dreissena polymorpha. Can. J. Fish. Aquat. Sci. 51:2227–2233. NRC (National Research Council). 1981. Testing for Effects of Chemicals on Ecosystems. Committee to Review Methods for Ecotoxicology, Commission on Natural Resources. Washington, D.C.: National Academy Press. NRC (National Research Council). 1983. Risk Assessment in the Federal Government: Managing the Process . Washington, D.C.: National Academy Press.

OCR for page 51
Incorporating Science, Economics, and Sociology in Developing Sanitary and Phytosanitary Standards in International Trade: Proceedings of a Conference Powell, M. 1997. Science in Sanitary and Phytosanitary Dispute Resolution. Discussion Paper 97–50. Washington, D.C.: Resources for the Future. Rai, S.N. and D. Krewski. 1998. Uncertainty and variability analysis in multiplicative risk models. Risk Analysis 18(1):37–45. Ricciardi, A., F.G. Whoriskey, and J.B. Rasmussen. 1995. Predicting the intensity and impact of Dreissena infestation native unionid bivalves from Dreissena density. Can. J. Fish. Aquat. Sci. 52:1449–1461. Schloesser, D.W., and T.F. Nalepa. 1994. Dramatic decline of native unionid bivalves in offshore waters of western Lake Erie after infestation by the zebra mussel, Dreissena polymorpha. Can. J. Fish. Aquat. Sci. 51:2234–2242. Stark, J.D. 1994. Chemical fumigants. Pp. 69–84 in Insect Pests and Fresh Horticultural Products: Treatments and Responses, R.E. Paull, and J. Armstrong, eds, Wallingford, U.K.: C.A.B. International. Stark, J.D., and J.A. O. Banken. 1999. Importance of population structure at the time of toxicant exposure. Ecotoxicology and Environmental Safety 42:282–287. Stark, J.D., L. Tanigoshi, M. Bounfour, and A. Antonelli. 1997. Reproductive potential: It's influence on the susceptibility of a species to pesticides. Ecotoxicology and Environmental Safety 37:273–279. Suter, G.W. II. 1999. A framework for assessment of ecological risks from multiple activities. Human and Ecological Risk Assessment 5(2):398–413. Trevan, J.W. 1927. The error of determination of toxicity. Proc. Royal Soc. B 101:483–514. Urban, D.J., and N. Cook. 1986. Ecological Risk Assessment. EPA 540/9–85-001. Office of Pesticide Programs. Washington, D.C.: U.S. Environmental Protection Agency.