Ecological Risk Assessment - Part 6 pps

Part VI
Risk Management
Risk characterization serves to inform risk management. At minimum, ecological risk asses-
sors must report the results of their risk characterization to the risk manager and any
stakeholders who are involved in the decision-making process (Chapter 35). In some contexts,
ecologi cal risk assessors are also involv ed in the decision-m aking process (Chapter 36). To the
extent that a formal analytical process is involved, risk assessors must, at least, present results
that support that decision analysis. Because decisions are based on human health, legal,
economic, ethical, and political considerations as well as ecological considerations, ecological
risk assessors should be prepared to help integrate all of those considerations to support the
risk manager (Chap ter 37 and Chapt er 38). Once the managem ent decisio n has be en made,
ecologi cal risk assessors may be invo lved in monito ring the results (Chapter 39).
This part of the process of environmental assessment and management is the least com-
fortable for most ecological risk assessors. However, if they are to be successful in influencing
environmental management, environmental scientists must be prepared to engage with the
social sciences and deal with the politics of decision making.
ß 2006 by Taylor & Francis Group, LLC.
ß 2006 by Taylor & Francis Group, LLC.
35
Reporting and
Communicating
Ecological Risks
It may not seem very important, I know, but it is, and that ’s why I’m bothering telling
you so.
Dr. Seuss
It is impor tant to distinguish the reporting of risk assessment results from communicating
risks. Reporting ecological risks involves creating a document that will support the decision-
making process by providing information for cost–benefit analysts or decision analysts, by
informing the decision makers, by informing stakeholders such as manufacturers and respon-
sible parties, and by serving as a basis for defense of the decision should it be challenged in
court or sent for review to the National Research Council or similar body. Communicating

ecological risk is a process in which risk assessors personally convey their findings to decision
makers and stakeholders. Risk communication may also include written products, but they
should be short statements (e.g., a one-page fact sheet) to prepare the audience or for the
audience members to take away as a reminder of the message. Risk reporting and commu-
nication are similar in the need to convey results clearly to the intended audiences. However, a
report must meet the needs of all likely audiences, while oral communication should be
tailored to specific audiences.
35.1 REPORTING ECOLOGICAL RISKS
The form in which ecological risks are reported is an oft-neglected aspect of the practice of
ecological risk assessment. The EPA’s guidance for risk characterization states that a report
of risk assessment results must be clear, transparent, reasonable, and consistent (Science
Policy Council 2000). Consid erations for achievi ng these goals are listed in Box 35.1. Ho w-
ever, the goals of being brief (for clarity) and transparent are conflicting. If sufficient detail is
presented for the reader to fully understand how the results were derived and to replicate
them, the resulting multivolume report will be thicker than anyone will care to read. As
discus sed in Chapt er 5, simply justify ing the assignmen t of dist ribut ions to parame ters may
result in a sizable report. However, some critics have advocated more complete risk charac-
terizations including multiple alternative risk estimates (Gray 1994). For ecological risk
assessments, this means reporting not only risk esti mates for all lines of evidence in all
types of evidence for each endpoint, but also results for alternative assumptions within a
line of evidence.
ß 2006 by Taylor & Francis Group, LLC.
The usual solution to this conflict between brevity and transparency is the executive
summary. Unfortunately, executive summ aries attempt to summarize the entire assessment
and are seldom sufficient to stand alone if the ‘‘ex ecutive’’ is the risk manager. A report of
results that ne glected methods but presented risks in adequate detail for decision making
would probably be more useful in most cases. In addition to summarizing the results, a report
to the risk manager should explain the major issues, any co ntroversies, and relevant prece-
dents. Ideally, the contents and level of detail would be worked out between the risk assessors
BOX 35.1

Clear, Transparent, Reasonable, and Consistent Risk Characterizations
For clarity:
.
Be brief.
.
Avoid jargon.
.
Make language and organization understandable to risk managers and informed lay people.
.
Explain quantitative results.
.
Fully discuss and explain unusual issues specific to a particular risk assessment.
For transparency:
.
Identify the scientific conclusions separately from policy judgments.
.
Clearly articulate major differing viewpoints or scientific judgments.
.
Define and explain the risk assessment purpose (e.g., regulatory purpose, policy analysis,
priority setting).
.
Describe the approaches and methods used.
.
Explain assumptions and biases (scientific and policy) and their influence on results.
For reasonableness:
.
Integrate all components into an overall conclusion of risk that is complete, informative,
and useful in decision making.
.
Acknowledge uncertainties and assumptions in a forthright manner.

.
Describe key data as experimental, state-of-the-art, or generally accepted scientific know-
ledge.
.
Identify reasonable alternatives and conclusions that can be derived from the data.
.
Define the level of effort (e.g., quick screen, extensive characterization) along with the
reason(s) for selecting this level of effort.
.
Explain the status of peer review.
For consistency:
.
Follow statutory requirements, guidelines, and precedents.
.
Describe how the risks posed by one set of stressors compare with the risks posed by a
similar stressor(s) or similar environmental conditions.
.
Indicate how the strengths and limitations of the assessment compare with past assessments.
Source: Adapted from EPA (U.S.Environmental Protection Agency), Guidelines for Ecological Risk Assess-
ment, EPA=630=R-95=002F, Risk Assessment Forum, Washington, DC, 1998; and Science Policy Council,
Risk Characterization Handbook, EPA 100-B-00-002, US Environmental Protection Agency, Washington,
DC, 2000.
ß 2006 by Taylor & Francis Group, LLC.
and risk manager. Routine assessments, such as those for new chemicals, may have a standard
form or format for reporting to the risk manager.
The needs of users other than the decision maker constitute a more serious conflict with the
call for brevity. Cost–benefit analysts or decision analysts need detai led results to support
their analyses. Risk assessments prepared by a responsible party must present data and
methods in detail so that regulators can review their acceptability. Risk assessments prepared
by regulators must present data and methods in sufficient detail that the responsible party can

review their acceptability. In either case, the report must be sufficiently detailed to withstand
legal scrutiny. As a result, the report of a complex ecological risk assessment may fill a library
shelf. Simply providing data, models, and analytical results on a CD or DV D can help, but
creative solutions to the problem are need ed. The use of hypertext is promising in that it
would allow a person reading a brief summary of the risk assessment to bore into a topic to
the depth that is appropriate to interest s and needs. However, creating a large hypertext
document is not quick or easy, and many people do not like to read from a computer screen.
35.2 COMMUNICATING ECOLOGICAL RISKS
Risk communication is the process of conveying the results of risk assessments to decision
makers, stakeholders, or the public, and of receiving and responding to their comments.
(Some documents define risk communication to include consultation during planning and
problem formulation.) It goes beyond the issue of reporting the results of the risk assessment
in a clear and useful manner to actually conveying the results to a skeptical audience. It is
difficult for two reasons. First , like any quantitative and scientific subject, it is difficult to
convey to those who do not have the requisite training or experience. Risk assessments may
be particularly difficult to explain, because they combine biological and physical sciences with
mathematics and statistics. Second, situations that require risk assessments are often emo-
tionally charged. Peopl e’s health, livelihood, and property values are typically at stake, and
distrust is typically high. Most of the literature of risk communication is directed at issues of
managing emotions and gaining trust with respect to health risks, independent of the nature
or quality of the technical message (NRC 1989; Fisher et al. 1995; Lundgren and McMakin
1998). Those issues will not be treated here, because emotional investment in ecological issue s
is usually lower and is likely to be qualitatively different. In many if not most ecological risk
assessment cases with high levels of emotional investment , such as restrictions on harvesting,
water withdrawals, or land use to protect resource species or endangered species, the issues
are largely economic and the strong emotions are largely held by those who generate the risk.
Fishermen, logger s, ranchers, and farmers are reluctant to believe that their activities damage
the environment in ways that would justify restriction of their acti vities. Research is needed to
guide risk communication in such situations.
The technical communication problems are more severe for ecological risk assessors than

for health risk assessors. Ecological risk assessors not only deal with unfamiliar scientific and
mathematical concepts, but also often estimate risks to unfamiliar entities and attributes.
Decision makers and stakeholders know full well what a human is and have both knowledge
of, and empathy for, the various fates that befall humans. However, many will not know what
a spectacled eider or an Atlantic white cedar bog is, much less the implications of changes in
their nesting success or in hydro-period. Humans have a sense of their own inherent value and
their value to their family and community, but have little knowledge or appreciation of
equivalent values of ecological entities. Hence, much of ecological risk communication is a
matter of education. This education may require a little honest salesmanship as well as the
basic description of an entity and attribute. This may involve the use of attr active photo-
graphs and explanations of values associated with the endpoint. Failure to protect the
ß 2006 by Taylor & Francis Group, LLC.
environment should not occur because the decision makers lack a vivid unde rstanding of
what may be lost or gained.
A related communication problem is the greater familiarity of most decision makers with
the relatively simple methods and results of human health risk assessments. The numerous
endpoints and multiple types of evidence employ ed in ecological risk assessments make them
seem complex and ambiguous. This problem may be alleviated as health risk assessments
begin to use multiple lines of evidence and to estimate the range of public health outcomes.
However, in the meantime, decision makers tend to focus on aspects of ecological
risk assessment that seem familiar. As a result of that tendency and the natural affinit y
of people for mammals and birds, risks expressed in the familiar terms of effects on
survival or reproduction of such species tend to be inordinately influential. For other end-
points, it is important to explain not only what they are and how they respond but also why
they are assessed using unfamiliar methods and models. As far as possible, use analogies to
health risk assessment. For example, analyses of biological survey data in risk assessments
can be described as ecological epidemiology, and species sensitivity distributions can be
described as dose–response models for ecological communities. Finally, the unfamiliarity of
ecological methods and models often leads decision makers to ask whether they are official
methods or have been used previously in decision making. Therefore, it is important to be

prepared to cite guidance and precedents. If a genuinely novel method or model is used, be
prepared to compare its results to those of more familiar methods or models and to explain
the advantages of the innovation.
A more pervasive problem is the inherent difficulty of conveying scientific and mathemat-
ical concepts to people who are not trained in those fields. As Cromer (1993) explains, science
is difficult to do and to convey, because it constitutes uncommon sense. It has been suggested
that this is because the human mind evolved to extrapolate directly from experience, which
works routinely but results in flawed logic in complex or unfamiliar inferences (Pinker 1997;
Dawes 2001). Further, even when reasoning carefully, the mind deals more easily with some
sorts of information and problems than othe rs (Pinker 1997; Anderson 1998, 2001). From
these generalizations, some advice can be derived.
Avoid probabilities: People, including scientifically trained experts, have difficulty with
probabilities, but understand an d manipulate frequencies relatively easily (Gigerenzer and
Hoffrage 1995; Gigerenzer 2002). Whenever possible without distorting the results, translate
probabilities to frequencies when communicating risks. This has the ancillary advantage of
forci ng you to de termine exactly what you mean by a probab ility (Ch apter 5).
Use discrete units: Although most properties of nature are continuous, the mind divides the
continua of time and space into events and objects. For example, Bunnell and Huggard (1999)
reduced the spatial continuum of forests to a hierarchy of units (patches, stands, landscapes,
and regions), which could be more easily described to forest managers than sets or areas of
forest.
Use categories: We not only discretize continuous variables, but also lump the units into
like categories to which we assign names. Hence, ‘‘folk biology is essentialistic’’ (Pinker 1997).
Long after ecology has revealed the variance in species composition over space, we continue
to name vegetation and ecosystem types (e.g., mixed mesophytic forest). Similarly, it is often
easier to communicate the frequencies of categories (e.g., high, moderate, or low) associated
with ranges of a continuous variable (e.g., flow) than to co mmunicate the meaning of the
variable’s probability density.
Use few categories : People tend to divide entities or events into only two or three categories
such as drought, normal, and flood for flow regimes.

Tell stories: Information imbedded in a narrative is more convincing and remembered
longer. Every conceptual model is a potential story.
ß 2006 by Taylor & Francis Group, LLC.
Use multiple modes: Individual audience members will respond differently to verbal, dia-
grammatic, pictorial, or other modes of representation. By using multiple modes of presen-
tation you are more likely to achieve comprehension with one. In particular, photographs of
the effects being discussed such as photos comparing plants grown in site and reference soils
or depictions of the diversity of fish species in a disturbed and reference stream can make
unfamiliar effects vivid. In addition, repetition increases comprehension and retention and
repetition using multiple modes avoids boredom. Similarly, in presentations it is advanta-
geous to include speakers with different styles to engage the diverse members of the audience.
Use case studies: Even if your assessment is generic (e.g., national risks from mercury in
coal combustion emissions), illustrate it with a particular case (e.g., loons in the boundary
waters). People are more willing to extrapolate from one real case to many than from an
abstraction to any real case.
Simplify carefully: Scientific concepts can be complex and difficult. It is often advantageous
to use a simple analogy to convey a complex system. However, it is important to prepare
qualifiers (e.g., of course it is not really that simple), because knowledgeable members of the
audience, particularly those who are opponents, will pounce on an ‘‘oversimplification’’
(Schneider 2002).
Remember the human analogy: While you are talking about risks to otters from eating fish,
many in the audience will be thinking about what it implies for people eating fish. Be prepared
for the questions that such analogies imply and avoid statements that would alarm those who
are focused on the human analogy or that contradict the human health risk assessment.
Avoid personalizing: Although personalizing a situation makes the message more vivid (e.g.,
‘‘I would not let my daughter eat fish from that river’’), it is a bad strategy for risk assessors.
Present the results of your analysis and let the risk managers and stakeholders infer the
personal implications.
Of course, this advice is superceded by the advice to know your audience. For example,
former US EPA administrator Ruckleshaus (1984) preferred that risk assessment results be

presented as cumulative distribution functions, contradicting some of the previous advice.
This example also serves to remind assessors not to talk down to risk managers, stakeholders,
or the public. An individual who is not an environmental scientist and does not know what
Bayesian means or what a Pimephales promelas is should not be treated as unintelligent. A
condescending or contemptuous attitude will be detected and will result in dismissal of your
message. On the contrary, people can unde rstand and appreciate descriptions of complex
natural systems if they are well presented. The renowned conservation biologist Daniel
Janzen has said of ecological complexity: ‘‘Audiences soak it up. I was told that it’s too
complicated and a lot of people won’t understand it. Bullshit. They understand it perfectly
well’’ (Allen 2001).
Risk communication is an opportunity to ensure that your efforts do some good. The
audience is less likely to be hostile than in human health risk communication. Because
ecological risks are not personal threats, audiences are open to learning a bit about nature
and how plants and animals interact with contamination or disturbance. You have an
opportunity not only to explain your results but also to educate and even entertain. In that
way, you can help to create a constituency for good ecological management.
ß 2006 by Taylor & Francis Group, LLC.

36
Decision Making
and Ecological Risks
It is not industry, development or the nation’s growing population that poses the greatest threat
to the environment, it is shortcomings in the political process that perpetuate environmental
degradation.
Howard R. Ernst (2003)
In the conventio nal environmental and ecological risk assessment frameworks, risk assessors
communicate their results to risk managers and then leave the room. Decision making is
viewed as a policy-based and science-informed political process that is best left to those who
have political authority. This description is accurate in many cases. However, to clarify
technical issues and avoid misunderstanding, risk assessors may be, and should be, involved

in the decision-making process. In some cases, distinct analyses are performed after the
risk assessment and before the decision by eco nomists or experts on decision analysis. In
such cases, risk assessors are likely to be involved in supporting or helping to perform
those analyses. In addition, even when risk assessors are not involved, they should have
some understanding of the decision-making process so that their results may be as useful as
possible and to avoid unrealistic expectations about their ability to determine the outcome.
The bases for decision making discussed in this chapter are limited to those that are risk-
related. As discus sed in Chapt er 2, so me environm ental de cisions are made based on best
available technology or some other criterion that does not include risk. It must be borne in
mind that politics is the ultimate criterion. Any of the decision criteria discussed below may be
overridden by political ideology or political expediency.
36.1 PREVENTING EXCEEDENCE OF STANDARDS
The simplest and least ambiguous decision criterion is that if an environmental standard is
violated, action must be taken. To the extent that standards are based on ecological risks
(Chap ter 29), these are risk- based decision s.
36.2 PREVENTING ADVERSE EFFECTS
The national policy of the Netherlands states that exposure to substances should not result in
adverse effects on humans or ecosystems (VROM 1994). Similarly, the US Clean Water Act
prohibits ‘‘impairment of the physical, chemical, or biological integrity of the Nation’s
waters.’’ Such policies provide the most clearly risk-based criteria for decision making.
Given a definition of adverse effects, one can determine whether they are expected to occur
or whether the probability of occurrence is excessive, and that is sufficient grounds for action.
ß 2006 by Taylor & Francis Group, LLC.
36.3 MINIMIZING RISKS
Com parative risk assessment s (Ch apter 33) provide a basis for decision makers to ch oose the
acti on that present s the least risk to human health or the e nvironm ent. Com parative risk
estimat es can be based on a techni cal risk charact erization as discus sed in Chapter 33.
Alternat ivel y, it can be viewed more broad ly as a type of policy an alysis that includes relative
risks a nd psychosoci al preferen ces co ncerning different types of risks in a process of stake-
holder consensus buildin g. That approach can be thought of as a microscal e a pplication to a

specif ic decision of the macros cale pr ocess of using compara tive risk asses sment for priority
setting (Sect ion 1.3.1) (Andr ews et al. 2004).
36.4 ASSURING ENVIRONMENTAL BENEFITS
Often, eithe r expli citly or impl icitly, actions to protect the environm ent are judged to be
defensi ble if they are exp ected to resul t in adeq uate benefi ts to the environm en t. Bene fits are
the complem ent of ad verse effe cts an d their risks. In the context of risk assessment , benefits
are avoidanc e of a risk of adverse effe cts. In the context of remediati on and restorati on, risks
are the probabil ities that ben efits will not be realized or that damage will occur due to poor
planning or executio n, or chance events. Bene fits may be judged in term s of absolute benefi ts
of an action, relative be nefits of alternati ve actions (Sect ion 33.1. 5), or be nefits relat ive to
costs (Section 36.6).
Concer ns for the benefi ts of actio ns may be express ed when a program has been ong oing
for some time, costs have be gun to mount, an d questio ns arise abou t the cost-eff ectiven ess of
the expend itures. An examp le is the exp enditur e of $580 milli on on 38 pro jects to remediate
con taminate d sedim ents in the Laur entia n Great Lake s witho ut a clear linkag e to increa ses in
ben eficial uses (Za rull et al. 1999). At some point, faith in su ch large exp enditur es fails
withou t evidence of benefits. As a resul t, the Great Lakes Water Qual ity Boar d reco mmended
the de velopm ent of be tter method s to qua ntify the relationshi p be tween sed iment co ntamin-
ation and use impai rment s, and to mon itor the ecologi cal be nefits and bene ficial uses of
remedi ated sites.
36.5 MAXIMIZING COST-EFFECTIVENESS
Cost e ffectiven ess an alysis identifi es the relative moneta ry costs of diff erent mean s of ach iev-
ing a standar d, an a cceptable risk level, or other goal. The decisio n maker co uld then choose
the least-cos t method. A v ariant of this idea that ha s had impor tant environm ental impl ica-
tions in the Un ited States is the principles and guidelines framew ork of the Army Corps of
Engin eers which maximize s the net nationa l econ omic benefi ts of project alternati ves, as long
as they do not cau se signi ficant environm ental deg radation (USA CE 1983).
36.6 BALANCING COSTS AND BENEFITS
Cost–b enefit an alysis is increa singly ap plied to environm ental regula tory and remedial ac-
tions . The decisi on mod el is that the public benefits of regula tions should exceed the c osts of

compliance to the regulated parties. In some cases, it is sufficient to show that monetary and
nonmonetary benefits qualitatively balance the costs. However, strict interpretations of cost–
ben efit requir ements allow only mon etary be nefits. As discus sed in Sectio n 38.3, this require-
ment to monetize the benefits of ecological entities and processes can be a serious impediment
to environmental protection.
ß 2006 by Taylor & Francis Group, LLC.
36.7 DECISION ANALYSIS
Decision analysis comprises a diverse set of concepts and methods for informing a decision.
Classic formal methods define the decisions to be made, the alternatives, goals, possible
outcomes, their values to the decision maker (utility metr ics), and their probabilities so as
to calculate the expected value or utility of each alternative (Clemen 1996). An early example
of an environmental application of formal decision analysis is an analysis of further research
vs. remedial alternatives for polychlorinated biphenyl (PCB)-contaminated Salt Creek, Indi-
ana (Parkhurst 1984). Other methods are less quantitative, less focused on quantif ying the
expected consequences, and more focused on clarifying the decision for the decision maker or
stakeholders. In general, they include more considerations than the other decision-making
approaches discussed in this chapter, and, because they do not require monetizing the
decision criteria, decision analyses can include noneconomic criteria (Stahl et al. 2002).
Utilities may be scaled in standard categories (e.g., low, medium, or high) or in terms of an
environmental goal (e.g., hectares of wetland or abundance of game fish).
Although decision analysis has a large literature and commercial software for its imple-
mentation, it is rarely used in environmental regulation or management. In part, this is
because the explicit inclusion and quantification of considerations other than risk is likely
to anger some stakeholders (Hattis and Goble 2003). Also, decision makers have little
incentive to go through the trouble of formal decision analysis when they have been succ essful
using precedents and their own judgment.
The US Army Corps of Engineers has begun to use multicriteria decision analysis to
support decisions concerning the management of contaminated sediments (Linkov et al.
2006). In an example from the Cocheco River, New Hamshire, the alternatives for sediment
disposal were cement manufacture, flowable fill, wetland restoration, and upland disposal

cell. The multiple criteria were cost, environmental quality, ecological habitat, and human
habitat. A mail survey of stakeholders provided weights for the criteria that could be
combined to produce a multicriteria score for each alternative. This sort of decision analysis
serves primarily to inform the decision maker about preferences of stakeholder groups based
on a structured elicitation process rather than the usual stakeholder meetings.
36.8 MISCELLANEOUS AND AD HOC CONSIDERATIONS
Many, if not most, environmental management decisions are made without any explicit
criteria or any formal analysis of how the alternati ves relate to criteria. Decision makers
consider the information co ncerning risks, benefits, and costs, they may consult with stake-
holders, they pay particular attention to legal and regulatory constraints and precedents, then
they test the political winds, consult their gut, and make a decision.
ß 2006 by Taylor & Francis Group, LLC.

37
Integration of Human Health
Risk Assessment
People are the probl em, but any solution which does not serve peop le will fail .
M arty Matl ock (unpubli shed present ation)
Inevitab ly, when both human health and the environm ent are threatened by a common
hazard, human health concern s dominat e asses sment and decisi on-maki ng pro cesses. Eco-
logic al risk asses sors can use the concern for hum an health to their advantag e by using
wildlife as senti nels for health effec ts, by integ rating the ecological assessment with the healt h
asses sment so as to combine resourc es, and by sh owing how ecologi cal effe cts influence
human healt h and welfar e.
37.1 WILDLIFE AS SENTINELS
Wildlif e may serve as sentinels, thereby stre ngtheni ng the case for current or futur e risks to
humans (NRC 1991; Burk hart and Gardne r 1997; Peter 1998; Sheffield et al. 1998; van der
Schali e et al. 1999; Col born an d Thay er 2000; Fox 2001). An exa mple of wildlife a s sentinels is
the obs ervation of thyroi d patho logy in wildli fe due to halogenat ed organic chemi cals that led
to studies in hum ans (Fox 2001; Karmau s 2 001). How ever, most report s of the use of an imal

sentin els of he alth effects have not provided the types and levels of e vidence ne eded before
healt h decisio ns can be based on sen tinel responses (Ra binowitz et al. 2005). Althou gh there
are diff iculties in extra polating from wildlife to humans (Stahl 1 997), they are conc eptually no
more severe than those associated with extra polating from labo ratory rats to humans.
Wildlife species are likely to be effective sentinels if they have a common source and route
of exposure with humans but are more exposed, more sensitive, or more readily monitored.
They are, in general, likely to be more exposed and therefore likely to respond more quickly
and severe ly than huma ns (Bo x 2.1) . In general , the use of wildli fe as sentinels may be just ified
by the following factors:
Common routes: Wildlife may feed on the same organisms as humans, particularly in the
case of piscivorous wildlife and subsistence or recreational fishermen. Similarly, wildlife may
consume contaminated soil, as do children.
Stenophagy: Wildlife species are generally less omnivorous than humans and therefore
species that consume a contaminated food item are likely to be more exposed.
Local exposure: Wildlife species do not obtain food or water from outside their home range,
so they are more exposed to contaminated locations.
Same mixture: Wildlife species are exposed to the same mixture of contaminants as humans
living in, or consuming foods from, the same contaminated system.
ß 2006 by Taylor & Francis Group, LLC.
Nonen viron mental source s : W ildlife spe cies do not have the occu pationa l or lifestyle expo s-
ures that con found epidemi ological studie s of humans.
Var iance : Most human populati ons are geneti cally divers e due to immigra tion and vary in
their diets , religious practices, oc cupatio ns, use of recrea tional and medic inal drugs, etc.
Thes e sources of varia nce, whi ch do not occur in wildli fe, tend to hide effe cts of environm en-
tal exp osures.
Avai lability : W ildlife may be readily sampl ed, an alyzed, and necrop sied.
Wh ile wi ldlife are more commonl y sentin els for humans, in some cases humans may be
senti nels for wildli fe. Birth, death, and diseas e records , whi ch may be used to detect environ-
menta l effects in hum ans, a re not availab le for wi ldlife. In additio n, in some cases, humans
may be exposed and affected wher e equ ivalent wi ldlife are rare or ab sent (Fox 2001). As a

resul t, a field termed conserva tion medic ine or conserva tion healt h has developed to jointly
study effe cts on human an d nonhum an organis ms of environm ental pollutan ts and pathogens
(Wei nhold 2 003). This approach may result in the de velopm ent of integrate d sen tinels. For
exampl e, harbor seals have been pro posed as sen tinels for other mari ne mamm als a nd
human s (Ros s 2000).
A databas e of studi es from the biomedi cal literatu re of the use of an imal surroga tes can be
found at http: == www.canar ydatabas e.org.
37.2 INTEGRATED ANALYSIS OF HUMAN
AND ECOLOGICAL RISKS
For practi cal reasons , the methodo logies for human health and ecologi cal risk assessment
were developed independently. However, for several reasons , the need for a more integrated
practice of risk assessment has become clear. These issues led the World Health Organization
to develop an integ rated framew ork for health and eco logical risk assessment (Sect ion 3.2.1)
(WHO 2001; Suter et al. 2003).
37.2.1 COHERENT EXPRESSION OF ASSESSMENT RESULTS
Decision makers must make a single decision with respect to an environmental hazard that is
beneficial to both human health an d the environment, but this goal is impeded by the
presentation of incoherent results of health and ecological risk assessments. The results of
independent health and ecological risk assessments may be inconsistent and the bases for the
inconsistency may be unclear because the results of the health and ecological risk assessments
are based on different spatial and temporal scales, different degrees of conservatism, or
different assum ptions, such as assumed parameter values or assumed land use scenarios. As
a result, decision makers may find it difficult to decide whether, for example, the reported
risks to humans are sufficient to justify taking a remedial action that will destroy an
ecosystem. As another example, consider a decision to license a new pesticide that poses an
increased risk to humans and a decreased risk to aquatic communities relative to a current
pesticide. If the ecological risk estimates are based on expected effects on a spatially dist rib-
uted community while the health risks are based on provision of a margin of safety on an
effect level for a hypothetical maximally exposed individual, the two estimates of risk cannot
be compared. Finally, if variance and uncertainty are not estimated and expressed equiva-

lently for health and ecological risks, a decision maker cannot determine the relative need for
additional research to support future assessments. For example, variance in aqueous dilution
should be either included or excluded in both assessments, and, if it is included, the same
estimates should be used. Integration of health and ecological assessments can avoid these
impediments to defensible decisions.
ß 2006 by Taylor & Francis Group, LLC.
An integ rated compara tive risk asses sment of co ntaminated sedim ent managem ent illus-
trates this issue (Dr iscoll et a l. 2002). The asses sment used a common set of alternati ves an d
assum ptions in a co mmon he alth and ecologi cal risk asses sment that highli ghted co mmon
resul ts (the no-act ion alte rnative had the high est risk for both endp oints) and differences
(island disposa l ha d the highest ecologi cal risk of the remedi al alternati ves but had relative ly
low healt h risk).
37.2.2 INTERDEPENDENCE
Ecologica l an d hum an he alth risks are interd ependen t (Lubchen co 199 8; Wilson 1998b).
Humans dep end on nature for foo d, water purificat ion, hydrologi c regula tion, an d other
produ cts and servi ces, which are dimin ished by the effects of toxic chemica ls or other
distu rbances. In ad dition, ecologi cal injur ies may resul t in increa sed hum an exposures to
contam inants or oth er stre ssors. For exampl e, addition of nutri ents to aq uatic ecosyst ems an d
the resul ting ch anges in algal communi ty structure may infl uence the oc currence of water-
borne diseas es such as cholera as well as toxic algae such as red tides . The need to asses s
ecologi cal risks in or der to estimat e indir ect effe cts on hum an health is particular ly apparent
in the asses sment of climate change (Bernar d and Ebi 2001).
37.2.3 QUALITY
The scientif ic quality of asses sments is impro ved through shari ng of informat ion an d tech-
niques between asses sment scient ists in diffe rent fields. For exampl e, in assessment s of
contam inate d sites, human he alth asses sors may use defaul t uptake fact ors to estimat e
plant upt ake, una ware that ecologi cal asses sors are measur ing co ntaminan t concen trations
in plants from the sit e. The da ta sets av ailable for the safet y evaluat ion of ch emicals in hum an
food an d drinki ng water are relat ively large and are used to suppo rt intens ive asses sments. In
contras t, ecologi cal risk assessment s for chemi cals have relative ly small data sets an d few

resourc es to perfor m assessment s, even though the recepto rs include thousands of specie s
such as plants, invert ebrates, and verte brates. Integr ation of efforts may help to alle viate these
imbalan ces in quality.
37.2.4 EFFICIENCY
Integr ation of human health and eco logical risk asses sments offer s signi ficant increa ses in
efficie ncy. In fact, isolated asses sments are inherent ly inco mplete when both human s an d
ecologi cal systems are pot entially at risk. For exampl e, the proce sses of contam inant release,
trans port, an d transform ation are common to a ll recept ors. Although only humans showe r in
water and only aqu atic organisms respire water, the process es that intr oduce the con tamin-
ants to water, degrade or transform them, and parti tion them among pha ses are co mmon to
both. There fore, there are clear advantag es in an integ rated exposure mod el. The develop-
ment of risk assessment methods, which takes into account insights from both human and
ecological risk assessments, will lead to improvements that can benefit both disciplines.
Integrated analysis of toxic risks will be facilitated by the increasingly mechanistic charac-
ter of toxicology. Because the structure and function of vertebrate cells are highly conserved,
the mechanism of action of a chemical is likely to be the same in all vertebrate species and
even the effe ctive co ncentra tion at the site of actio n is likely to be effe ctively con stant (Sect ion
23.1.6) (Escher an d Herm ens 2002). Hence, when toxico logy is suff iciently mechan istic, the
need for toxicity testing should decline, and a common effects analysis approach should serve
for both humans and other vertebrates. However, this approach would increase the demand
for toxicokinetic modeling to estimate site of action concentrations.
ß 2006 by Taylor & Francis Group, LLC.
37.3 ENVIRONMENTAL CONDITION AND HUMAN WELFARE
One view of risk asses sment an d man agement is that we protect human healt h as well as the
nonhum an environm ent, an d the only con nection betw een them is throu gh environm entally
media ted or trans mitted threat s to hum an healt h (Sect ion 37.2.2). An alte rnative view is that
we mu st consider not just human health but also human welfar e, and that human welfar e is
influ enced by environm ental co nditions.
Envi ronment al quality infl uences hum an welfar e throu gh the pro vision of ecosystem ser-
vices . Humans obtain a wide range of be nefits from the environm ent, often termed service s of

nature, which make human life possible and enh ance its qua lity (Daily et al. 1997, 2002). At the
most basic level , plants capture the en ergy of the sun, produce food for human s an d their
live stock, and con vert our CO
2
into O
2
. Othe r life su pport functi ons of ecosyst ems include
purif ication of wat er, air, an d soil , and cyclin g of water an d nutri ents. In addition, ecosystems
prod uce a variety of goods, su ch as tim ber, fisheries , biomas s fuels , and med icinal he rbs, roots,
and spices . Thes e servi ces are estimat ed to be worth more than the entire mon etary ec onomy
(Co stanza et al. 1997; Naeem et al. 1999). The concep t of rest oring services is alrea dy contai ned
in the Natur al Resour ce Dama ge Ass essment pro visions of some US laws (Sect ion 1.3.9).
A more subtle aspect of this issue is impr ovement in the qualit y of huma n life provided by
nature (Kea ch 19 98). An obv ious examp le is the regener ative effe ct of a vacati on spent
fish ing, hunt ing, bird-watch ing, hik ing, phot ographing na ture, or simply visiti ng natural
places . Ho wever, day-to-day con tact with nature such as wal king in park s, feedi ng birds,
and observin g trees, flowers , an d butterfl ies may be more impor tant to the qua lity of life. The
impor tance of this relationshi p to people’s qua lity of life is reflected in various beh aviora l and
econ omic measur es includin g the billions spent on feedi ng birds and the value of houses on
large lots in the outer sub urbs or on bodi es of wat er. The invers e of this relation ship is the
con sternati on people feel when they observe trees dying along highw ays, dead fish along a
river , or even images of oiled birds on the television .
Fina lly, cultural values are often related to aspect s of the environm ent. Thi s is parti cularly
true of indige nous peoples whose cultures requir e the use of certain natural resourc es and the
presence of c ertain environm ental featu res (Harris an d Harper 2000). The same is true of
nonin digenous cu ltures, althoug h in less obv ious ways. Exa mples from the United States
include the ba ld eagle as a nationa l emble m and uns poiled wide-open spaces as a back drop
for the na tional mythos of cowboys an d hearty pion eers. For ests have played a sim ilar role in
Germ an culture (Scham a 1995).
It is impor tant to distinguis h risks to nonhum an organ isms, populati ons, and communi ties

from risks to human welfar e through loss of ec osystem services. Most environm ental legi sla-
tion protects ecological entities irrespective of any requirement to demonstrate benefits to
human welfare (EPA 2003c). However, when they can be demonstrated, risks to hum an
welfare could significantly supplement ecological and health risks as justifications for envir-
onmental protection. At least, they provide a basis for estimating economic benefits of
protect ive acti ons (Section 38.2) .
37.4 SUMMARY
To become more influential in environmental decision making, ecological risk assessors must
collaborate with, and supplement, health risk assessors. The performance of risk assessments
for reductions in human welfare and indirect effects on human health is a large task that
cannot be readily performed by ecological risk assessors. Health risk assessors must be
encouraged to look beyond direct toxic effects to a broader view of risks to humans.
Integration will require movemen t from both directions.
ß 2006 by Taylor & Francis Group, LLC.
38
Integration of Risk, Law,
Ethics, Economics, and
Preferences
To derive conclusions about action, we need some statements about how the world works and
some statements about what we believe are good and right.
Randall (2006)
Risks are ne ver the sole basis for decision making (Chapter 36). Risks are more accepta ble
when there are countervailing benefits, and actions to reduce risks are more acceptable if there
is a clear legal mandate, legal precedent, or public support. Increasingly, formal analyses of
costs and benefits are required. Therefore, integration of the risk assessment with the eco-
nomic analysis should be planned during the problem formulation. However, it is also
important to be aware of the limitations of economic decision criteria and of the existence
of other decision criteri a. Environmental law, environmental economics, and environmental
ethics are large and complex fields in their own rights, which are barely touch ed on here. This
chapter is intended to simply create an awareness among ecological risk assessors of these

other considerations so that they have an idea of how their risk estimates must be combined
with these considerations during the decision-making process.
38.1 ECOLOGICAL RISK AND LAW
Risk assessment may contribute to both criminal and civil law. It may be used to demonstrate
compliance with environmental laws or failure to comply. It may also be a tool to establish
injuries in civil legal actions. The language of the law lends itself to concepts of risk: ‘‘more
likely than not,’’ ‘‘the preponderance of evidence,’’ or ‘‘beyond a reasonable doubt.’’ When laws
provide clear and specific legal mandates, legal compliance is a sufficient justification for action.
For example, the US Endangered Species Act provides clear legal protection for species that are
listed as threatened or endangered and for their critical habitat. Efforts to restore the bald eagle
and peregrine falcon were not subject to cost–benefit analyses or public surveys. Other laws
such as the US Clean Water Act do not include economic considerations in its mandate to
restore the ‘‘physical, chemical, and biological integrity of the Nation’s waters,’’ but the
interpretation of these vague terms leaves room for balancing of interests in defining criteria
and standards. However, once legal standards are established, risk assessment may simply
estimate the probability that the standard will be exceeded. Other phrases in laws such as
‘‘reasonably achievable’’ create a mandate for balancing the environment against economic
costs. Hence, the legal context of an assessment determines the degree of confidence required
for an action and the extent to which risks must be balanced against other considerations.
ß 2006 by Taylor & Francis Group, LLC.
38.2 ECOLOGICAL RISK AND ECONOMICS
In the United States and many other nations, efforts to improve environmental quality and
protect nonhuman populations and ecosystems are increasingly subject to cost–benefit tests.
This practice is based on welfare economics, which is based on the premise that when
resources are limited, the general welfare is best achieved if individuals are free to maximize
their individual welfare (also termed utility) through free markets for goods and services. In
an ideal market, the decisions of rational individuals (including corporations as legal indi-
viduals) will lead to an efficient outcome. This invisible hand of the market often fails,
particularly with respect to the environment. The main reason for these market failures is
the absence of a market for environmental goods such as clean air or wild songbirds.

Similarly, the services of nature such as water purification and soil formation are performed
without anyone receiving a utility bill from nature. Polluters may destroy these goods and
services without buying the right in an efficient market, replacing the lost goods and services,
or compensating the users. Finally, for commonly held resources such as fisheries, there is an
incentive for overexploitation. The economic gain goes to the irresponsible individual, but the
resource loss is shared by all, leading to ‘‘the tragedy of the commons’’ (Hardin 1968).
Regulation is needed to compensate for these market failures. To ensure that regulation is
not excessive, welfare economists devised the cost–benefit analysis, whi ch creates a pseudo
market. Costs of regulation are assumed to be justified if they buy an equivalent benefit of
environmental goods and services.
An obvious difficulty in this concept, even for welfare economists, is that environmental
benefits are difficult to define and enumerate, much less quantify in monetary terms. Alter-
native approaches to estimating monetary benefits of environmental protection are listed in
Table 38.1. All have severe limitations . The reveal ed prefer ence methods requir e that the
value of a resource be quantified in terms of some moneta ry expenditure to use the resource,
such as using the cost to recreationists of visiting an ecosystem as an estimate of the value of
the ecosystem. Clearly, such methods address a small fraction of the value of nature. Stated
preference methods, particularly contingent valuation, are more commonly used to value the
environment, because they use surveys to create an entirely hypothetical market. In particu-
lar, contingent valuation is used to estimate the monetary damages that polluters must pay to
replace lost ecosystem services under Natural Resource Damage Assessment (Kopp and
Smith 1993). These survey-based methods may be applied to any use or nonuse value, but
they suffer from a number of problems including:
.
The public has little understanding of most environmental resourc es and services.
.
Even if they understand the resource or service, they may not have relevant well-defined
values to be elicited by a survey.
.
Even if respondents have well-defined values, the values may not encompass the full

utility of the resource or service.
.
Any attempt to educate the survey participants is likely to bias their responses.
.
The respondents have no experience in pricing or paying for valued ecological resources
or services.
.
It is not known whether respondents would actually pay the stated amount if a market
could be created.
.
The number of resources or services that can be addressed is limited by the patience of the
survey participants.
.
People may not be willing to pay anything, because they belie ve the responsible party
should pay.
.
People may opt out because they object to putting a price on nature.
ß 2006 by Taylor & Francis Group, LLC.
These problems can introduce significant biases as well as uncertainty. For example, the last
two problems described in the list cause people who are strongly pro-environment to opt out,
thereby biasing the sample.
If one of these cost–benefit techniques is applied to an environmental decision, it is
incumbent on ecological risk assessors to support that analysis. If a revealed preference
TABLE 38.1
Methods for Estimating Monetary Values of Environmental Goods and Services
Method Description Examples
Revealed preference methods (can estimate use values only)
Market When environmental goods are traded in
markets, their value can be estimated
from transactions

The benefits of an oil spill cleanup that
would result in restoration of a
commercial fishery can be projected
from changes in markets for fish, before
and after the spill, and their effects on
fishermen and consumers
Production function The value of an environmental good or
service can be estimated when it is
needed to produce a market good
If an improvement in air quality would
lead to healthier crops, the value of the
improvement includes, e.g., the
reduction in fertilizer costs to produce
the same amount of agricultural crops
Hedonic price method The value of environmental characteristics
can be indirectly estimated from the
market, when market goods are affected
by the characteristics
If an improvement in air quality improves
a regional housing market, its value
includes increases in housing value,
which can be measured by statistically
estimating the relationship between
house prices and air quality
Travel cost method The value of recreational sites can be
estimated by examining travel costs and
time
The value of a recreational fishing site to
those who use it can be estimated by
surveying visitors, to determine the

relationship between the number of
visits and the costs of time and travel
Stated preference methods (can estimate both use and nonuse values)
Contingent valuation
method
Individuals are surveyed regarding their
willingness to pay for a specifically
described nonmarket good
In a telephone survey, respondents are
directly asked their willingness to pay,
via a hypothetical tax increase, for a
project that would reduce runoff,
improving the health of a particular
stream
Conjoint analysis Survey respondents evaluate alternative
descriptions of goods as a function of
their characteristics, so the
characteristics can be valued
In a mail survey, hypothetical alternative
recreational fishing sites are described
by type of fish, expected catch rate,
expected crowding and round-trip
distance; respondents’ preferences are
used to calculate value for changes in
each of the characteristics
Source: Bruins, R.J.F., Heberling, M.T., eds., Integrating Ecological Risk Assessment and Economic Analysis in
Watersheds: A Conceptual Approach and Three Case Studies, EPA=600=R-03=140R, Environmental Protection
Agency, Cincinnati, OH, 2004.
ß 2006 by Taylor & Francis Group, LLC.
method is used, the assessment endpoints must be identified and quantified in terms of the

good or service that is provided. For example, if the market value of fish is used by
the economists, reductions in harvestable mass of fish should be estimated. If a stat ed
preference method is used, endpoints must be defined that are potentially understood and
valued by the public. This may require converting a primary effect such as death of ugly forest
lepidopteran larvae into valued secondary effects such as reduced abundance of beautiful
birds, moths, and butterflies.
To ecologists, assigning monetary values to the natural environment is likely to be some-
what repugnant, and the techniques employed often seem scientifically suspect. However, in a
decision context that requires cost–benefit analysis, nonparticipation of ecological risk asses-
sors is likely to result in mini mizing or even ignoring environmental benefits other than
improved human health. This requires a different approach to risk assessment and decision
making than the pur ely legal approach. It is not enough to establish that no unacceptable
effects will occur. Rather, one must be able to estimate the risks of specific effects and the
benefits of avoiding or remediating those risks. When working in an economic decision
context, it is important to engage the economists in the pro blem formulation process and to
understand their needs and methods (Bruins and Heberling 2004). An extensive summary of
environmental economics can be found in van den Bergh (1999), a review of economics for
watershed ecological risk assessment is provided by Bruins and Heberl ing (2005), and relevant
guidance from a US regulatory perspective can be found in National Center for Environ-
mental Economics (2000) and Science Policy Council (2002).
Analysis of economic and other benefits of ecological risk reduction can lead to a broadening
of the scope of assessments. For example, the restoration of riparian vegetation in agricultural
areas is justified under the Clean Water Act in terms of improved water quality. However,
riparian communities have other benefits such as provision of habitat for birds (Deschenes et al.
2003). While the problem formulation for a risk assessment focuses on the endpoints that relate
to the legal mandate, the accounting of benefits to justify the cost of riparian restoration is not
constrained in that way. Just as all of the costs of requiring waste treatment, remediation, or
restoration are identified and summed, so too should all of the benefits be identified and
summed, not just those that were endpoints for the assessment. In fact, it is hard to know
where to stop. The protection of an ecosystem results in a nearly infinite list of benefits. All of the

species have value, all of the functions have value, and every visible or audible detail of an
ecosystem has at least some esthetic value. The fact that the costs of regulation are relatively
easily and completely tabulated (e.g., the costs of building and operating a waste treatment
plant), while the benefits to human health and the environment are incompletely identified and
estimated, has led to the idea that what is really practiced is ‘‘complete cost–incomplete benefits
analysis’’ (Ackerman 2003). In fact, the costs to industry are routinely overestimated, largely
because regulation leads to development of lower cost treatment technologies or waste mini-
mization and reuse (Ruth Ruttenberg and Associates 2004).
In the face of highly uncertain benefits, it may be desirable to abandon cost–benefit analysis
in favor of an insurance-based approach. For example, if the benefits of a policy such as
greenhouse gas control are highly uncertain, a risk management approach can justify the cost
of some control to avoid the risk of catastrophic losses. This approach is equivalent to the
precautionary principle, but is based on economic rather than ethical principles.
A failure of con tingent valuation and the other techniq ues in Table 38.1 is that they treat
people strictly as consumers. People are also citizens who contribute to the form ation of laws
as discussed in Sectio n 38.1 (Sagof f 1988). A pe rson as a pur ely self-i nterested econo mic entity
may be willing to pay very little to protect streams in other states, but that same person as a
citizen of a democracy may agree that the political process should protect streams nationally,
leading to legal criteria and standards. In addition, welfare economics ignores the fact that
ß 2006 by Taylor & Francis Group, LLC.
people are ethical and act in ways that are rational but do not maximize their economic
welfare.
38.3 ECOLOGICAL RISK AND ETHICS
Welfare economics is not the only intellectual system for supporting environmental protec-
tion. Formal ethics may be applied. The fundamental difference between considering the
environment in an ethical manner rather than an economic manner, no matter how broad, is
illustrated by Routley’s last man argument (Schmidtz and Willott 2002). Consid er that you
are the last person left in the world and you have little time to live. If you decided that it
would be fun to cut down the last surviving redwoods, would there be anything wrong with
that? Resource value, services of nature, and even esthetics would be irrelevant. Yet, I hope

that most readers would agree that the act would be wrong. What would make it wrong is
some ethical principle. While the public has been responsive to appeals for environmental
protection that are effectively ethical, ethics has had little influence as a formal decision
support tool relative to law or economics. This is in part because there is no generally
accepted system of ethics analogous to classic welfare economics.
Ethicists have no standard classification system, but ethical principles and systems gener-
ally fall in the following categories:
Motivist: The acceptability of an act depends on the intent of the individual. This is the
basis of the distinction between murder and manslaughter and for the requirement of ‘‘intent
to deceive’’ in financial fraud. It has little relevance to modern environmental ethics, but may
have been important in traditional cultures that allowed use and even waste of resources if it
was done with the proper intent as displayed by rituals and incantations (Krech 1999).
Consequentialist: The acceptability of an act depends on its effects. This is the basis for
utilitarian ethics and its offspring, welfare economics. However, consequentialist and even
utilitarian interests may extend well beyond economic welfare. People may be willing to give
up economic resources out of sympathy for nonhuman organisms or in order to avoid
repugnant experiences.
Deontological: The acceptability of an act depends on its nature. This concept of ethics is
related to concepts of duty, obligation, and rights. It is associated with Kant, but its best-
known expression in the environmental literature is Aldo Leopold’s land ethic, which ex-
presses a human obligation to protect the land, by which he meant ecosystems. Those who
ascribe moral standing to nonhuman organisms, populations or ecosystems are taking a
deontological stance.
Environmental ethics a re well summarized in Blamey and Comon (1999), and the range of
positions in environmental ethics is represented in Schmidtz and Willott (2002).
The implications for risk assessment of ethical standards are less clear than those for legal
or economic standards. Deontological ethics would tend to promote protective standards
rather than the balancing of interests suggested by consequenti alist ethics. However, some
consideration of consequences is nearly inevitable. For example, if we have a duty to both
aquatic and terrestrial ecosystems, the decision to sacrifice a terrestrial ecosystem for sewage

sludge disposal must be based on a determination that the consequences of not generating
the sludge or of disposing of it in the ocean would be more severe or at least less acceptable.
Also, since different versions of the concept of duty toward the environment are conflicting,
consequentialist ethics may be required to decide between them. For example, animal rights
advocates condemn Leopold’s land ethic because it implies management of anima ls for the
sake of ecosystems. Therefore, a decision to protect an ecosystem by harvesting herbivores
ß 2006 by Taylor & Francis Group, LLC.
that are not natural ly control led woul d be based on a judgme nt ab out the consequ ences of not
harvest ing.
38.4 ECOLOGICAL RISK, STAKEHOLDER PREFERENCES,
AND PUBLIC OPINION
Trad itions an d public opin ions infor m politics, which in turn drives the legal proc ess and
influ ences how individual managem ent decisi ons are made. In add ition, stakehol ders may be
involv ed in de cision making as well as parti cipating in plann ing and pro blem form ulation.
This trend ha s been encou raged in recent publica tions of advisor y pan els on risk assessment
(Nation al Rese arch Counci l 1994; The Pr esidential =Congr essiona l Com mission on Ris k
Ass essment and Risk M anageme nt 199 7). This advice is based on concerns that de cisions
that are not accepte d by the affected parties will be polit ically unacceptabl e and may be
blocke d or delayed. While these process es are usually infor mal, de cision a nalytic methods
that utilize multiple utilit ies can make this process as rigor ous as cost–be nefit an alysis without
requir ing the co nversion of prefer ences to moneta ry units (Brauer s 2003).
Fro m an ecologic al perspect ive, increa sing stakeho lder influen ce is unfort unate in that it
tends to accentua te hum an health con cerns and dimi nish ecologica l issue s. The respon sible
parti es do not pus h for ecologi cal protection, the member s of the publ ic who are sufficien tly
moti vated to participat e are primarily those with health co ncerns, and en vironme ntal advo-
cacy groups tend to adopt the public’ s healt h con cerns or are sim ply absent. The stakehol ders
who do ca re passio nately about ecologi cal issues tend to be those who use or harvest na tural
resourc es (e.g ., logger s, ranchers , fishermen, farm ers) and theref ore are opposed to pro tec-
tion. Hence, in most stakeholder processes, nobody speaks for the trees. Ecological risk
assessors who participate in stakeholder-informed decision processes should be prepared to

make the case for the environment by clearly presenting the results of their assessments in a
way that the stakeho lders can relate to (Sect ion 34.2). If a man agement decision is of
sufficient importance, national surveys of public opinion may serve to balance the interests
of stakeholders. An example is the current conflict in the United States over oil development
in the Arctic National Wildlife Refuges. Such ecological issues of national importance, for
which environmental advocacy organizations can mobilize public opinion, are rare. In routine
assessments, risk managers who are public officials must represent the public’s interests,
based on law and policy.
38.5 CONCLUSIONS
Most of this book has been concerned with how to estimate risks in an accurate and unbiased
manner, addressing the endpoints of concern at the right spatial and temporal scales, and
presenting results clearly and appropriately. The critical final step is ensuring that the results
are influential in the full context of the decision. This chapter has described other consider-
ations that may influence the interpretation of ecological risks. Ecological risk assessors who
simply drop their results on the risk manager’s desk are almost guaranteed to be frustrated by
how little their work is reflected in the decision. Ecological risk assessors must learn to work
with lawyers, economists, policy analysts, and others who have the ear of risk managers. For
most of us, introverted ecologists who prefer dealing with the complexities of nature than with
the complexities of human emotions or institutions, this is more of a challenge than calcu-
lating integrals of exposure and effects distributions.
ß 2006 by Taylor & Francis Group, LLC.
39
Monitoring the Results
of Risk Management
Many imp acts cannot be foreseen and planni ng must therefore provide for monitor ing an d
adapta tion .
Holl ing (1978)
Results of ecologi cal risk asses sments are unc ertain and regula tory and remedial action s may
have unexp ected effe cts. Ther efore, risk-based decisi ons are not guarant eed to have the
desir ed outco me. Environme ntal monit oring can reveal the actual outco me and guide furt her

asses sments, decision s, and acti ons. Unfor tunately, the e ffects of environm en tal managem ent
actio ns on ecosyst ems are seldom monit ored. Rather, it is typic ally assum ed that the problem
is resol ved, an d asses sors an d man agers move on to the next pro blem. This is ostens ibly a
reasonabl e appro ach. M anagers choose actio ns that they belie ve will be effica cious, and it
seems reasonabl e to assum e, at least, that una ssessed and unremed iated sit es are worse.
How ever, the efficacy of remedi ation an d restorati on techni ques is often unc ertain, they
may be poorly c arried out, an d, in any case, the law of uni ntended consequ ences ap plies.
A fun dament al questi on to be an swered by monito ring is whet her exposure and effects have
been reduced by a remedi al action? W hile it might seem self-evid ent that treatment of
effluents or remova l of co ntaminated media woul d reduce exposu res, studies do not always
bear that out. In particu lar, dredging of polychl orinated bip henyl (PCB)-cont amina ted
sedim ents has not, in some cases, reduced exposure as measur ed by accumul ation in bival ve
mollu scs and fis h (Rice and White 1987; Voi e et al. 2002). In effe ct, the mass of mate rial
decreas ed but availab ility to the biota did not. Further, remedi ation may red uce exposure but
not eliminat e toxicity . Follow ing dre dging an d capping of the Laur itzen Channel of Sa n
Francis co Bay, Calif ornia, concentra tions of chlori nated pe sticides decli ned in sediment an d
decreas ed in trans plante d mu ssels at one site but increa sed at another (Anderson et al. 2000 ).
One year afte r remedi ation, the be nthic invertebrat e co mmunity co ntained few sp ecies an d
individ uals, and the sedimen t was more toxic to amphipods than prior to remedi ation.
The regis tration of new pestic ides and new indust rial c hemicals lend s itself to pos t-man-
agement mon itoring an d asses sment. This practi ce is partic ularly co mmon for pe sticides. In
the United State s surveilla nce studi es are somet imes requ ired of the manufa cturer . Bec ause
applications are replicate treatments, such studi es are relatively easy to design. In fact, the
availability of this option was a reason that requirements for preregistration field testing of
pesticides have been largely eliminated in the United States (Tuart and Maciorowski 1997).
Because effluent permitting based on concentrations of individual chemicals does not
accoun t for c ombined toxic effects, effluent toxic ity testing (Sect ion 24.4) was develop ed as
a means of monitoring the success of effluent pe rmits. However, because effluent test s are
periodic, they are likely to miss episodes of high toxicity due to treatment failures, temporary
ß 2006 by Taylor & Francis Group, LLC.

changes in a process, or other incidents. Continuous biological monitoring of treated effluents
has been proposed, but has not yet been adopted (Figure 39.1).
Effluent permitting may also fail to protect the environment because of the combined
effects of multiple effluents and of nonpoint sources. In the United States, biological mon-
itoring and bioass essment programs by the states and tribes are intended to detect those
combined effects and provide a basis for regulating total pollutant loading, i.e., total max-
imum daily loads (TMDL findings) (Houck 2002). The primary limitation on this ecoepide-
miol ogical ap proach (C hapter 4) is the high cost of perfor ming adequate mon itoring at
enough points at sufficient frequencies for the many water bodies of a state or nation.
The results of removing an exotic species should be monitored, because removal does not
necessarily resolve ecological problems. For example, removal of livestock from Sa nta Cruz
Island, California, and from one of the Mariana Islands resulted, in each case, in explosive
growth of a formerly uncommon exotic weed and suppression of native plants (Simberloff
2003). Hence, monitoring is necessary to assure that ecosystem management goals are
achieved as well as to confirm that the species has been removed.
Similarly, ecological restoration projects are often physically but not ecologically success-
ful. In particular, wetland restoration projects often succeed in creating a wetland but fail to
establish diverse wetland communities that support target species such as clapper rails or that
adequately perform wetland functions such as nutrient retention.
When designing post-management monitoring programs, it is important to carefully con-
sider the goals. The first goal will be to determine whether the endpoint attributes that
prompted the management action were restored. For example, the assessment of the Laur-
itzen Channel remediation monitored changes in contamination of sediment and mussels, in
Surface water biomonitor or upstream biomonitor for a water treatment plant
Water treatment plant biomonitor
Influent wastewater biomonitor
Effluent wastewater biomonitor
1.
2.
3.

4.
1
2
3
4
Industrial
site
Water
treatment
plant
Untreated
wastewater
Wastewater
treatment
plant
(influent)
Treated
wastewater
(effluent)
Upstream
dilution
water
for
biomonitor
River
FIGURE 39.1 Diagram of a system to biologically monitor the toxicity of waste water. (From van der
Schalie, W.H., Gardner, H.S., Bantle, J.A., De Rosa, C.T., Finch, R.A., Reif, J.S., Reuter, R.H., et al.,
Environ. Health Persp., 107, 309–315, 1989. With permission.)
ß 2006 by Taylor & Francis Group, LLC.
toxic ity to inverteb rates, an d in inverteb rate communi ty struc ture, but co uld not answe r the

questi on of efficacy be cause the endpo int assem blage, benthic fish, was not mo nitored
(Ande rson et al. 2000). The secon d common goal is to determ ine the cause of an y manage-
ment failure, whic h requir es that the contam inant levels a nd other inter media te causal
parame ters be mo nitored along wi th the end points. Thi s can be difficul t, because failure
often occu rs due to some process that asses sors did not con sider. For exampl e, the Laurit zen
Channel asses sment did not include pos sible sources of sedimen t contam ination other than
the Unite d Heckat horn Supe rfund sit e (Ande rson et al. 2000). Therefor e, it is impor tant to
develop conceptual models of alternative candidate causes and identify variables that distin-
guish them and could be monitored.
Media toxicity tests (Sect ion 2 4.5) can play an impor tant role in monit oring to determ ine
efficacy. By testing treated or remediated media, it is possible to distinguish ongoing effects
that are due to residual toxicity from effects of other causes. Such tests can even precede
remediation, if treated media are available from pilot-scale studies. For example, solvent
extraction of soils contaminated primarily with PCBs removed 99% of the PCBs, but the
toxicity of the soil to earthworms and plants was unchanged or even increased, depending on
the species and test endpoint (Meier et al. 1997).
If monitoring reveals that the ecological system is not recovering or is still impaired after
sufficient time has been allowed for recovery, an ecoepidemiological study should be con-
ducted to determine the cause (Chap ter 4). Resid ual impairme nt may be due to fail ure of the
remedial action to sufficiently reduce exposure to the agent of concern or to the effects of
other hazardous agents. This analysis requires that the residual impairment be clearly defined,
that plausible candidate causes be listed, and that analyses be performed based on spatial and
temporal associations of candidate causes with the impairment, experimental results,
and mechanistic understanding. Successful causal analysis requires that the candidate
causes and the residual impairment be monitored concurrently at the impaired site and at
reference sites so that spatial and temporal associations can be determined. Once the
most likely cause has been determined, an ecological risk assessment of the new remedial
alternatives can inform a decision concerning appropriate management actions.
Another goal of monitoring management results is to improve understanding of what
assessment practices are successful and to determine how and why other assessment practices

fail, so that the practice of ecological risk assessment can advance. It is also valuable to
determine whether failures tend to be overprotective or underprotective. There is evidence for
both. For example, studies of toxic effects on striped bass embryos contaminated from
maternal PCB burdens led the EPA Region II (2000) to conclude that effects on striped
bass populations were likely in the Hudson River. However, a careful analysis of long-term
striped bass population data showed no effect of PCBs on year-class strength, apparently due
to density-dependent compensatory processes (Barnthouse et al. 2003). In contrast, continu-
ous monitoring of Prince William Sound has shown long-term ecological effects that were not
predicted by assessments performed shortly after the Exxon Valdez spill (Peterson et al. 2003).
In particular, short-term toxicity studies of the water-soluble fraction of the oil—thought to
be the toxicologically active fraction— resulted in predictions of no significant risk to fish.
However, exposures of salmon eggs to persistent 3- to 5-ring hydrocarbons in field sediments
and in the laboratory caused increased mortality for years after the spill. Hence, limitations in
the state of the science and the need to employ simplifying assumptions in ecological risk
assessments have caused errors in both directions.
In sum, monitoring of the results of management actions is essential to ensuring that the
goals of those actions are achieved. In addition, real progress in ecological risk assessment
depends on the feedback provided by well-designed and conducted monitoring programs.
Neglected areas of environmental chemistry and toxicology such as the toxicology of reptiles
ß 2006 by Taylor & Francis Group, LLC.