Chemical Stressors and Ecological Risk
Abstract
Risk assessment is a critical part of the regulatory aspect of ecotoxicology. Risk assessors use scientific findings and scientific wild-ass guess to estimate the risk to nontarget organisms and the environment due to contaminant exposures. Risk managers then consider the advice of the assessors and determine, along with other factors such as economics, human populations, etc., what steps need to be taken to remediate the risks or not.
Keywords
Risk analysis; risk management; hazard; stressor; receptor; toxicity reference value; hazard quotient; competing
Introduction
What do we mean by ecological risk assessment? Whenever contaminants are present in the environment, they pose some degree of potential harm to that environment, its plants and animals, and the people that live there. This potential harm is called risk or ecological risk (also see the following discussion). The risk from any occurrence of contaminants in the environment may be negligible, indeed below the level that humans can measure, or it may be substantial. The instances of environmental catastrophes described in this book are examples of the latter.
A branch of ecotoxicology, called risk assessment, evaluates the harm that may come from using or storing contaminants in the environment. Risk analysts use mathematical, biological, and chemical models to predict what may happen under various scenarios with the goal of choosing an optimal program of use or storage. Sometimes, they have to make a scientific wild-ass guess (SWAG) when data are not available. For instance, risk from a landfill may vary depending on what type of underlayment or capping is used to seal the contents of that landfill from discharging into the environment. The risk analyst may be involved in making choices of these materials and in predicting what may happen to nearby rivers, streams, or air should the landfill fail in preventing a discharge. This is just one of many different functions of risk assessment.
Many journal reviews and reference books that focus on practices and methods for evaluating risks associated with chemicals in the environment have been published recently (see Suter, 2006; Suter et al., 2000). Not surprisingly, risk analysts have not achieved complete agreement on how technical findings of ecotoxicology should be applied to questions related to risks associated with environmental chemicals. Indeed, beyond this introduction, your future study will discover that ecotoxicology affords a range of opportunities to apply its technical findings and methods to the risk evaluation process. How those technical findings and methods are applied to questions focused on chemicals in the environment vary widely, and each implementation of a risk-based analysis will have its champions and critics. However, here we focus on conceptual common ground and do not enter the fray of expert opinion regarding the technical analysis of risks associated with chemicals in the environment. At least that is our goal.
Our introductory overview of the evaluation of environmental risks begins with a brief review of terms and some historic context regarding evaluation of risks within environmental settings, and conceptual frameworks for such evaluations focused on environmental chemicals. We then consider a snapshot of sorts focused on the technical processes deployed for an evaluation of risks of naturally occurring chemicals or those released to the environment directly or indirectly through human.
Terms of the Tradecraft
For this introduction we will opt for terms that follow consensus-based definitions, often having regulatory or legal precedence. Risk is characterized as the probability of occurrence of an adverse effect on human health or the environment as a result of exposure to a hazard. In general, a hazard is defined as a condition, process, or state that adversely affects the environment. We focus here on environmental hazards that manifest themselves as physical or chemical pollution in air, water, and soils and can cause widespread harm to a range of life forms, including humans. Environmental hazards may take many forms, ranging from occupational and work-related hazards to those typically considered as environmental hazards involving fish and wildlife, as a category of events focused on ecological outcomes. Hazards for our present purposes are regarded as physical events or processes (eg, landslides), or chemical releases to air, water, and soils. In general then, environmental hazards are frequently linked as potential causes of harm to biotic components of the environment, including adverse effects on humans. As such, environmental hazards range from those associated with events that are linked to intentional or accidental release materials to the environment (eg, routine applications of agrichemicals or mishaps resulting in chemicals released to soil, water, or air), natural hazards (eg, earthquakes, floods, or extreme weather events such as hurricanes or tornados), and events linked to use of a hazardous technology (eg, nuclear power for generation of electricity). Throughout this book, we have focused on environmental chemicals and the role that ecotoxicology plays in describing the relationships among those chemicals and the abiotic and biotic components of the environment. Our primary considerations for the balance of this chapter are characterizing how hazards and risks may be linked to chemicals released to soil, water, and air.
Historical Perspectives on Chemical Stressors and Ecological Risks
Risk assessment can deal with any type of stressor. A stressor is any physical, chemical, or biological “agent” potentially capable of inducing a negative response from exposure. Stressors may adversely affect specific natural resources or entire ecosystems, including plants and animals, as well as the environment with which they interact. The history related to the evaluation of hazards and risks of environmental stressors—chemical, physical, and biological—is relatively short, yet the concepts of hazard evaluation and risk assessment have long been in practice, including evaluations of multiple stressors (eg, Menzie et al., 2007; Piggott et al., 2015). Although hazards and risks of chemicals in the environment initially brought the practice of risk assessment to a wide range of users, much of the practice advanced today will be considered in a relatively general context, providing guidance rather than prescriptions for implementation (Suter, 2006; US EPA, 2015a). These generic approaches are process driven, and their implementation necessarily involves specific application of a wide range of analytical tools determined, in part, by the nature of the risks being considered. Many of those tools and underlying technical findings that frame risk-based questions are directly linked to outcomes of ecotoxicology. Most practitioners focused on questions related to chemicals in the environment consider the risk-assessment paradigm initially advanced by the National Academy of Science (1983). This was a starting point that motivated numerous investigations, including those focused initially on ecological risks associated with chemical stressors (Warren-Hicks et al., 1991; Suter et al., 2000) and those applications concerned with multiple stressors occurring across a range of spatiotemporal contexts (eg, Graham et al., 1991).
With these terms in place for hazards and risks, we will characterize risk analysis as a process that strives to qualitatively or quantitatively estimate environmental risk, whereas risk assessment more often than not incorporates elements of administrative and management policy that reflect, in part, an individual’s or organization’s level of risk tolerance or risk acceptability when interpreting outcomes of an analysis of risks. Risk analysis incorporates qualitative or quantitative estimates of selected attributes of environmental hazards—for example, adverse effects and events or conditions that lead to or modify such adverse effects—in order to characterize how environments or exposed populations of humans or fish and wildlife might respond by manifesting adverse effects linked to exposure. Risk analysis becomes entangled with risk assessment, particularly within the settings of risk management, wherein: (1) decisions are developed regarding hazards of environmental chemicals, biota exposed to those hazards, and adverse effects associated or potentially associated with those exposures; (2) mitigation options are crafted as part of the implementation of those decisions; and (3) outcomes of those decisions are evaluated through monitoring and surveillance activities (Fig. 13.1). We will return to the process represented in Fig. 13.1 later in this chapter and illustrate how topics in previous chapters may be viewed within the context of this process.
For environmental chemicals, outcomes of a comprehensive evaluation of risks potentially serve to better inform resource managers and others regarding the nature and magnitude of adverse effects to humans (eg, residents, workers, and recreational visitors), and fish or wildlife exposed to chemicals occurring in soil, sediment, water, or air. Indeed, increased awareness and interest in managing environmental risks associated with chemicals and other environmental stressors has gained visibility across a range of governmental and nongovernmental organizations, industrial, and other interests, as well as the general public. Information regarding risks gained by risk managers benefits their decision-making processes with respect to protecting humans and the environment from chemicals and a wide range of other environmental stressors. Indeed, we are all “risk managers” regardless of our roles as regulators (eg, federal or state officials serving to protect the environment), industrial or business authorities whose practices potentially impact the environment, or private citizens making decisions regarding chemical use, directly or indirectly contributing to the release of chemicals to the environment, and through their participation in the environmental regulatory process, contributing to development of risk and environmental policies governing their use.
Generalized Process for Evaluating Environmental Risks
An evaluation of environmental risks relies heavily on scientific format. However, an increasing awareness of the sociopolitical elements of the environmental risk assessment and risk management processes has become well-established as part of the overall process (eg, Gowdy and Erickson, 2005; Brown and Timmerman, 2015).
The focus at this point will be on ecotoxicological ties that shape a generalized process for evaluating ecological risks. Chapter 12 will demonstrate the linkages between ecological and human-health risks, particularly in environmental settings that illustrate potential for sharing common frames of exposure to chemicals in soil, sediment, water, or air. Regardless of the context, whether it is conditioned on ecotoxicological frames of reference or on toxicological frames of reference specific to human-health applications, risks linked to environmental chemicals are strongly conditioned by: (1) chemical quantities occurring in exposure matrices—that is, what concentration of chemical(s) occur in a given matrix; (2) the toxicity of the chemical(s); (3) pathways linking a receptor and chemical(s) in an exposure matrix or, more commonly, multiple environmental matrices with multiple chemical constituents; and (4) duration of exposure between receptor and matrix contaminated with environmental chemical(s). A receptor in this context is any biological agent, usually plants or animals, considered representative of organisms that are likely to be encountered at the location of interest and likely exposed to environmental chemicals of concern. As we have seen from previous chapters, each of these conditioning factors will shape chemical risks that are linked to concentration and to duration of exposure. For example, risk analysis is dependent on ecotoxicological data that characterize a chemical’s acute and chronic toxicity through the application of methods previously considered in Chapter 2.
Steps in Risk Assessment Processes
Based on the four conditioning qualities identified previously, we will revisit our generalized framework for implementing an evaluation of risks summarized in Fig. 13.1. Regardless of the language and terms applied to characterizing the assessment, evaluations of ecological and human health risks may be considered as a relatively simple four-component process that consists of establishing the context, identification, analysis, and evaluation. Establishing the context involves characterizing the site of concern—is it aquatic, terrestrial, atmospheric, etc.; are we concerned with soil, sediment, water, or air? Risk identification categorizes the species of concern and the chemicals that may be involved. Risk analysis attempts to determine to what degree the environment has or is likely to be impaired and whether biota are being affected. Risk evaluation brings in other factors such as economics, cost versus benefit scenarios, and determines whether and to what extent remediation should occur. The last category on the diagram, risk treatment may or may not occur depending on the results of the evaluation. As such, the analytical process assures technical outcomes that inform risk-based, resource management decisions with a level of detail required by the resource manager. Depending on the resource management questions and technical issues driving the evaluation of risks, assessment and monitoring needs may be satisfied through analysis of existing data and information, or designed field and laboratory studies may be undertaken as part of the assessment and monitoring activities (Suter, 1991; Karr and Chu, 1997).
In its simplest summary, the analysis, assessment, and management of risks are captured by a stepwise, iterative process wherein: (1) questions are formulated; (2) observations or experiments are conducted to provide answers that address those questions; and (3) decisions are made given the answers to the questions that initiated the process. Either decisions that result from the initial assessment may yield sufficient information so that management can evaluate potential outcomes of a particular management action, or the analysis process might be augmented to address critical data gaps identified as outcomes of the initial process. In risk-assessment language, data developed during the first round of investigation were not sufficient to support management decisions within the level of uncertainty reflected by the risk-tolerance of the decision-makers; thus, a second round of investigative studies would be indicated. The assessment process might also uncover a need to conduct a follow-up study to determine the effectiveness of remediation or perhaps a monitoring strategy for a defined period of time would be seen as being necessary.
In general, and regardless of the specific risk evaluation process being initiated, be that regulatory (Suter, 2006; US EPA, 2015b) or nongovernmental in kind (Suter, 2006), a planning and scoping stage serves to define the purpose and scope of any evaluation of risk. For risk evaluations focused on environmental chemicals, the risk analysis and assessment process often begins by collecting existing data or initiating reconnaissance investigations intended to characterize the nature and extent of chemical release to the environment, as well as to gather existing information and data regarding the fate and effects of the chemicals of concern. In pursuing existing data and information, we gain a better understanding of how these chemicals might behave in the future. Based on these preliminary assessments, a risk assessor then considers the frequency and magnitude of exposures, and identifies pathways potentially linking sources of environmental chemicals to humans and ecological receptors. This phase of the generalized process is most frequently termed “analysis of exposure,” which is then considered relative to an “analysis of effects.”
Analysis of effects is given different names by regulatory agencies or in peer-reviewed literature (Suter, 2006; National Academy of Science, 2009; Kapustka and Landis, 2010). Regardless of the name, however, each implementation of an analysis of effects is linked to the chemicals present and their toxicities, particularly because they serve as estimates of the nature and magnitude of possible adverse effects. Ideally, evaluations of risks would be based on reliable and complete data and existing information regarding the nature and extent of contamination, fate and transport processes, the magnitude and frequency of exposures to biological receptors in the field, and a chemical’s toxicity. Unfortunately, having all of the desired information is the exception. In general, existing data and information on actual exposures and effects are limited. Thus, all risk estimates are uncertain to varying extents. When risk estimates are so uncertain that they exceed the tolerance of managers, those conducting the assessment may have to delve deeper into the literature or obtain additional relevant data in an iterative process, as indicated in the generalized process summarized in Fig. 13.1.
Once an analysis of exposures and effects is completed and risk has been assessed, risk management (risk management can combine risk evaluation and treatment) takes center stage in any project’s lifecycle. Risk management is a relatively straightforward process that focuses on how to protect human health and ecological receptors. For example, identifying risk-based wastewater discharge limits that are acceptable for mining activities or identifying clean-up objectives for hazardous materials previously released to the environment can be elements of risk management.
Our overview of an evaluation of risk has been generalized. Yet, much of the practice implemented today follows regulatory guidance initially developed in the late 1970s and early 1980s with a focus on human health (Suter, 2006), which was followed shortly after by a similar process for evaluating risks associated with environmental chemicals when ecological receptors, principally fish and other aquatic biota, wild birds, mammals, and species of concern were the principal receptors of interest to federal and state regulatory agencies in United States, Canada, or Europe (Suter, 2006). For example, the US EPA provides guidance for conducting risk assessments with some processes focusing on human health, others on ecological concerns. Although the guidance for each of these regulatory-driven practices differs due to the receptors of primary concern, and there are relatively subtle distinctions between stages in the processes called for evaluating risks to human health and ecological concerns, our generalized process summarized in Fig. 13.1 easily accommodates such nuanced regulatory differences (Table 13.1).
Table 13.1
Comparison of Generalized Framework With Regulatory Frameworks Guiding Evaluations of Risks for Human-health or Ecological Concerns
| Generalized Process for Evaluating Risks (Fig. 13.1) | Regulatory Process Evaluating Risks Related to Human Healtha (National Academy of Science, 1983) | Regulatory Process for Evaluating Risks Related to Ecological Concernsb (US EPA 1992, 1998) |
| Problem identification and specification: Information collected to help determine what biological systems are at risk, including special status species (eg, Endangered Species Act) | Hazard identification: Examines whether a stressor has the potential to cause harm to humans and/or ecological systems, and if so, under what circumstances | Problem formulation: Information is gathered to help determine what, in terms of plants and animals, is at risk and what needs to be protected |
| Analysis: Determines chemical concentrations in exposure matrices (abiotic and biotic), target receptors, and biological effects in receptors; determines exposure, particularly frequency, timing, and levels of contact with a stressor | Dose-response assessment: Examines the numerical relationship between exposure and effects | Analysis: Determines what plants and animals are exposed and to what degree they are exposed, and if that level of exposure is likely or not likely to cause harmful ecological effects |
| Exposure assessment: Examines what is known about the frequency, timing, and levels of contact with a stressor | ||
| Characterization, synthesis, and interpretation: Determines numerical relationships between exposure and effects; development weight and strength of evidence about nature and extent of the risk from exposure to environmental stressors; complete sensitivity and uncertainty analysis | Risk characterization: Examines how well the data support conclusions about the nature and extent of the risk from exposure to environmental stressors | Risk characterization: Involves risk estimation, risk description, wherein risk estimation combines exposure profiles and exposure-effects and risk description provides important information for interpreting the risk results and identifies a level for harmful effects on the plants and animals of concern |
aHuman health risk assessment includes four basic steps and is generally conducted following various EPA guidance documents, with early phases of the process involving preliminary assessments and a range of planning and scoping activities.
bEcological risk assessments focus on evaluating how likely it is that the environment may be impacted as a result of exposure to one or more environmental stressors such as chemicals, land change, disease, invasive species, and climate change. Ecological risk assessment includes three phases, and is generally conducted following the Guidelines for Ecological Risk Assessment (US EPA, 1998).
Within regulatory contexts, ecological risk assessments are more widely implemented and support a variety of regulatory and resource management programs, including the regulation of hazardous waste sites, industrial chemicals, and pesticides, as well as the management of watersheds or other ecosystems affected by multiple chemical, physical, or biological stressors. As such, ecological risk assessment has been applied to projects that forecast likelihoods of future effects (prospective) or evaluate the likelihood that effects are caused by past exposure to stressors (retrospective). Information from ecological risk assessments may then be used by risk managers for follow-up, such as communicating to interested parties and the general public, limiting activities related to the ecological stressor, limiting use of a given chemical, or developing a monitoring plan to determine if risks have been reduced or whether an ecosystem is recovering. Thus, the current practice of ecological risk assessment can be applied to many “what-if” analyses related to forecasts of changes associated with various mitigation options considered for development under the auspices of risk management.
Assigning Risk
However, trying to simplify risk management to its core requires our brief return to the underlying characteristics of risk. Mathematically, risk is expressed in terms of the probability of hazard (as a numeric value bound by 0 and 1, where an event with “probability 0” never happens and an event with “probability 1.0” always happens, or as categorical estimates wherein ordinal assignments of risks are characterized based on experience or survey outcomes). For example, we might judge a given action as having risk rated as high, moderate, or low, and we subsequently gauge our risk management decision accordingly.
All risks are accompanied by consequences, most often characterized by frequency of occurrence (such as a numeric or categorical estimate of time) and severity (such as a numeric or categorical estimate of adverse impact, US EPA, 2009). A risk-based decision hinges on the definition of acceptable risk associated with a management action intended to offset risks linked to particular hazards. For example, management alternatives may range from no action to control chemicals released to the environment to costly management actions undertaken to attain a level of acceptable risk associated with release of chemicals, say, to a surface-water body such as a stream or river. Risk management, then, is the process of identifying and controlling hazards, where controls are intended to eliminate hazards or, more often than not, reduce their risk and attendant consequences to an acceptable level.
A clear example of the interactive role between assessment and management of risk was presented by Diamond et al. (2015). Their focus was on chemicals of emerging concern (CECs), such as pharmaceuticals and personal care products that can be discharged from wastewater treatment plants (WWTP) because there is no currently available technology to remove these CECs from discharge water. However, their concepts can apply to other contaminants as well. Consider collecting data on CECs from many different WWTPs, some of which discharge high concentrations of CECs and others that discharge lesser amounts (Fig. 13.2A). Risk managers may decide to take steps to remediate the most polluted sites, as indicated by the yellow bar.
Alternatively, data may be gathered on the probability that an area is subject to risk (Fig. 13.2B). Signs to look for might include physiological responses, or signs at the population or community level. For example, one common community risk would be reduced species richness with the species known to be most sensitive absent from collections. The risk-management team might choose to remediate only those sites where risk is greatest. In the diagram, the most polluted sites (Fig. 13.2A) coincide with sites that show effects (Fig. 13.2B), but this need not be the case in all examples. Perhaps the most polluted sites still have concentrations that are below effects levels. Finding sites with no ecological impairment and relatively high concentrations of certain CECs, or those that are ecologically impaired with low CEC concentrations, could provide evidence that specific CECs are unlikely to pose a risk to aquatic life in similar situations. When concentrations of CECs are low but ecological damage is occurring, the managers may send out an assessment team to explore the possibility that other, non-CEC stressors are having an impact. When the concentration and effects data coincide (ie, there is evidence of harm at the sites with the highest concentrations), risk managers are asked if that harm is acceptable or not.
A third approach that has been used in some assessments is to develop information on potential thresholds based on measured CEC levels (or class) associated with documented ecological effects (Fig. 13.2C). In this case, water concentrations of selected CECs are measured and then compared with data from laboratory or previous field studies to determine if there is an overlap between known effect concentrations and measured concentrations. This process could lead to a threshold-setting exercise in which a regulatory agency, such as the US EPA, established acceptable standards of contamination. We caution that it may be too early to establish acceptable standards for most CECs due to a lack of information at this time.
Finally, a fourth approach that might be used to screen sites is to compare the range of a specific physiological effect, such as the production of vitellogenin in male fish or the average fecundity of fish for screening for endocrine-disrupting effects in relation to CEC concentration (Fig. 13.2D). This fourth approach could evaluate under what conditions certain physiological indicators of exposure, such as endocrine responses, could be translated into higher-level impacts or could be used to identify characteristics of water bodies (in addition to the CECs of interest) that correlate with relevant biological measures. In this approach, managers attempt to identify sites where CEC exposures would present a risk or identify stressors, in addition to the CECs of interest, that predict sites with potential biological impacts.
Hazards and Risks
Depending on the contingencies that drive the environmental risk-assessment process—here, our focus has been environmental chemicals—the evaluation of risks may be categorized as hazard-based or as risk-based to inform resource management decisions. The distinction between these ends of the technical-analysis spectrum are largely dependent on empirical data and information, with a hazard-based process generally reflecting limited empirical data, while a risk-based analysis will be more data rich. Risk-based evaluations tend to be more probabilistic in their derivation, whereas hazard-based evaluations tend to be more experience-dependent expressions of risk, based largely on expert judgements and a variety of analytical tools. In practice, a mix of hazard-based and risk-based tools are frequently used for evaluating environmental chemicals, particularly when the analysis focuses on chemicals that are not well known with respect to their toxicity or exposure potentials, or when multiple stressors are considered and chemicals are only a subset of the possible hazards. Given this brief overview of hazard-based and risk-based approaches likely underlying risk management decisions, we close our brief consideration of hazards and risks with a short consideration of a frequently deployed method for applying technical findings derived from ecotoxicological research to assessment and management of risks linked to exposures of environmental chemicals; namely, the analysis of exposure through food-chain analysis and the subsequent estimation of risk by comparing this estimate of exposure to a toxicity reference value.
Exposure Models and Food-Chain Analysis for Birds and Mammals
Exposure models for terrestrial and semiaquatic vertebrates, principally birds and mammals, usually focus on materials transfer and remain a favorite regulatory application for evaluating exposure comparable to the process implemented in human-health risk evaluations (Linder and Joermann, 2001; Linder et al., 2004). As indicated in Chapter 12, models are part and parcel of the skill-set commonly used in risk evaluation and have become more commonly used as the disciplines of ecotoxicology and risk analysis have developed in recent years (Pastorok et al., 2001). Here, we simply ask you, the reader, to entertain some mathematical reasoning without fear and loathing because conceptually, Eq. (13.1) simply provides a mathematical explanation of how a wild bird or mammal can be exposed to sources of environmental chemicals:
(13.1)
where
E = Exposure concentration or exposed dose
T = Total time and space over which the concentrations in various microenvironments or habitats are to be averaged
Cijk = Concentration in microenvironment k that is linked to environmental matrix i by pathway j
tk = Time and space that accounts for a receptor’s contact with specific microenvironment or habitat k
This equation simply translates to “you are what you eat and drink”; that is—and please recall, risk is an iterative process—as a first estimate of exposure for some chemicals, drinking water and ingesting food likely serve as dominant sources in pathways of exposure. Most often, Eq. (13.1) is simplified, and becomes a unitless narrative (Eq. (13.2)):
(13.2)
where
ED = Exposed dose
Dered = Dermal- or cutaneous-exposed dose
Inhed = Inhalation-exposed dose
f(Inhed) = Exposed dose coincidental to inhalation
Inged = Ingestion-exposed dose
f(Inged) = Exposed dose coincidental to ingestion
DWed = Drinking water consumption exposed dose
f(DWed) = Exposed dose coincidental to drinking water consumption
Ultimately, given practical matters and all too frequently, the relatively sparse to nonexistent data available, risk analysts further simplify, which yields a food-chain dominated exposure model (Eq. (13.3)):
(13.3)
wherein chemicals contributed by dermal and inhalation routes of exposure are considered negligible to the receptor. In the first iteration, the fraction of exposed dose that is absorbed is frequently assumed to be 100%, or otherwise estimated using available literature values to account for efficiencies in the uptake of material across surfaces such as the walls of the gut or other barriers between external and internal environments or across cell membranes. Although physiologically oversimplified, food-chain models lump together “whole-animal” functions of water consumption, foraging rate, food intake, and food processing through parameters that have been predefined by the regulatory agency. These functions provide an estimate of exposure referred to as exposed dose or site-exposure level.
For our first iteration estimate of risk, we need an estimate of toxicity—here, referred to as a toxicity reference value (TRV)—derived in a similar manner to exposed dose or site-exposure level, which regulatory agencies such as US EPA have developed and published in support of their regulatory programs (eg, US EPA, 2003). These TRVs can provide a better understanding of how exposure to selected chemicals in the environment translates into risks.
Estimating Risks Using Simple Ratio Estimators
A commonly applied approach for evaluating ecological risks from environmental chemicals relies on a simple ratio estimator (US EPA, 2015c) that estimates potential for adverse effects to an ecological receptor, such as wild bird or mammal, by comparing the receptor’s estimated levels of exposure—recall exposed dose or site-exposure level from above—with appropriate TRVs. Most TRVs for chemicals commonly encountered in environmental settings of regulatory interest have been derived from published toxicity studies, wherein toxicity values such as no-observed-adverse-effect level (NOAEL) or lowest-observed-adverse-effect-level (LOAEL) (see Chapter 1) were identified and compiled. Then a representative value was selected as a benchmark TRV for regulatory applications. TRVs are both chemical-specific and species-specific, and the comparison to estimate risk takes the form of a ratio, referred to as the hazard quotient (HQ) (13.4):
(13.4)
If the value of HQ is less than or equal to 1.0, risk is considered negligible to low, and no unacceptable effects will occur in the exposed population of receptors. In contrast, if the value of the HQ exceeds 1.0, then an unacceptable impact may occur, and risk will be moderate to high. Within regulatory settings, food-chain modeling has become a primary tool for evaluating exposure in birds and mammals. However, depending on the animal-at-risk and its life history (eg, early-life stages tend to be more sensitive to exposure than adults), the tool may be relatively ineffective for evaluating potential adverse effects and risks under field conditions. Animals with life history attributes that markedly differ from wild birds and mammals may not fit well with the HQ concept. For example, if dermal exposures are assumed to be unimportant, this may be very misleading for amphibians whose semipermeable skin plays a critical role in respiration and may be an important route of exposure for some environmental contaminants (Linder et al., 2010).
Cause-and-Effect Relationships, Multiple Stressors, Uncertainty, and Risks
As became apparent in our brief overviews of effects of chemicals on population (see Chapter 10) and community and ecosystem (see Chapter 11) levels of organization, the more pieces of the puzzle, the more likely we are to encounter questions that defy simple answers. Indeed, complexities in biological systems across levels of organization or spatial scales through time will make the use of models that rely on technical findings (see Chapter 12) more difficult. Regardless of the innovation that continues to develop in the ever-expanding discipline of ecotoxicology, and the publication of new methods and high-quality data that fill data gaps critical to evaluating risks, environmental and resource managers focused on evaluating effects of environmental chemicals on biological systems, especially for large spatial scales or for systems considered of “special status,” uncertainty must be considered in lockstep with characterization of risks. Thus, to close our chapter focused on risk and chemical stressors we briefly identify issues related to risks and uncertainty, particularly because those are entangled with cause-and-effect analysis.
With multiple stressor exposures, there are increased potentials for confounding analyses of cause-and-effect relationships between environmental chemicals and biological responses—especially at population and community levels of organization (see Chapter 11). Confusion increases as the number of chemicals at low concentrations increases. This can happen, for example, with the discharge of permitted wastewater releases following treatment at a municipal water treatment facility. Consequently, the simplicity of laboratory-derived characterizations of cause-and-effect relationship will likely be inadequate to explain field situations.
The number of examples of such entanglements involving multiple stressors, exposures in the field, and confounded cause-and-effect characterizations for environmental effects are many. However, we have conceptually identified categories of risks that, in part, help address these circumstances and enable our efforts to characterize uncertainties to follow suit. For example, risks may be viewed as being of various types: competing risks, aggregate risks, cumulative adverse effects, and cumulative risks, residual risks, and landscape or regional risks. We can also speak of vulnerability analysis, most often within the context of proactive risk management.
Types of Multiple Risks
Competing Risks
Competing risks—regardless of the level of biological organization considered—are very common in natural resources. These exist in land or water resources that are generally challenged by a range of chemical, physical, and biological stressors in the field. Each of these stressors may potentially be characterized as a hazard capable of resulting in a loss of system integrity and may trump chemical hazards such as, for example, when physical aspects of wildfire overwhelm the hazards from chemicals (eg, dioxins, furans) released by the fire. Simply said, regardless of a system’s complexity or its level of biological organization, if system sustainability is the desired outcome of a resource-management plan, managers must consider multiple, interacting strategies such as chemical-management practices, land-use and water-use practices, or management of biological agents, such as invasive species or disease organisms. This is because a system’s health may be at risk from numerous hazards that act independently or in combination to adversely impact that system.
Cumulative Adverse Effects and Cumulative Risks
When evaluating chemical and physical stressors as part of an environmental risk assessment, cumulative risk is narrowly defined as risk associated with exposure to all chemical and physical stressors with a common mechanism of toxicity. For chemical hazards, cumulative risk assessments combine risks from multiple exposure pathways and routes for all substances that have a common mechanism of toxicity. For example, within a multiple-stressors context, such risks would be identified by cumulative adverse effects associated with a loss of system integrity yielding similar, if not identical, end states.
Aggregate Risks
Aggregate risks share attributes of competing and cumulative risks and may be regarded as the sum total of all risks from multiple sources. Simply defined, aggregate risks are those that are due to single stressors that come from many different sources and collectively add to the total stress on a system. For example, chemicals may weaken the immune systems of organisms and make them more susceptible to viral or bacterial infections. Weakened animals may be easier for predators to catch and kill. The bacteria, viruses, and predators may have always been in the environment, but it was not until after the additive stress from chemicals that they could harm the biota.
For evaluating hazard-specific competing risks or aggregate risks, all possibilities may be seen as racing to see which occurs first. For biological and ecological systems, as well as complex systems in general, networks of stressors may occur in series (one after another), parallel (occurring together), or in a variety of other patterns with each other (Westbury et al., 2007; Olff et al., 2009). In addition, many systems may be characterized by marked interdependence and interconnectedness, which means biological or ecological systems will necessarily reflect operational structures that make management difficult or even impossible to analyze (Luo and Magee, 2011).
Residual Risks
Residual risk is the risk remaining after control measures have been identified and selected for managing hazards. To manage residual risks, a system’s multiple stressors and their cumulative risks must be well-characterized. Similarly, the uncertainties associated with residual risks must be understood because management of these risks generally responds to uncertainties and those risks considered acceptable and manageable. Assessment of residual risks is an important step in the iterative process of risk assessment.
To address these residual risks, resource management often focuses on the system at risk through a top-down, holistic approach and seeks to resolve problems by identifying proximal causes related to maintaining system integrity. In contrast, much of the analysis focused on risks associated with multiple stressors and much of the regulatory focus-guiding risk management activities related to natural resources relies on bottom-up approaches. A bottom-up approach is generally characterized by efforts to identify whether suspected causes are key to resolving resource management problems, then moves to eliminate those risks without determining the extent to which that solution will resolve the management problem. For multiple risks potentially requiring conflicting solutions, a bottom-up approach contributes to management conflicts when priorities are set among risks or to evaluate whether a risk management action is likely to succeed.
Landscape and Regional Risk
Landscape and regional risk assessments have increasingly been incorporated into the environmental decision-making process (eg, Westbury et al., 2007; Olff et al., 2009; Kapustka and Landis, 2010). These risks occur at spatial scales larger than the area of interest as, for example, an entire watershed at the regional level and surrounding agricultural fields at the landscape level (Fig. 13.3). Each higher level has inputs into the lower levels and understanding the risks at higher levels will facilitate understanding of all the risks at the site level. While the methods and tools applied in such analyses vary depending on the scope and spatial scale of the problem, decision making may benefit from the melding of risk disciplines focused on evaluating multiple stressors and their effects on biological and ecological structure and function (eg, Chen et al., 2013). Much of the impetus for developing these larger spatial-scale processes resulted from increasing applications of the environmental risk-assessment process, which provides a systematic approach to evaluating undesired effects linked to exposures and environmental stressors, including chemicals, land-use changes, altered hydrology, and climate change. Larger-scale approaches to evaluating multiple stressors can benefit from an ecological risk assessment framework, since the analysis can reflect (1) explicit consideration of scale and spatial organization during problem formation; (2) spatial heterogeneity in exposure characterization; and (3) extrapolation from small-scale studies to broad-scale effects, representing adverse effects from widely dispersed chemicals, such as those released to the atmosphere. From a risk-management perspective, opting into a nested, spatial hierarchy for evaluating risks could also provide maps or other visualization techniques for risk communication, while informing environmental decision makers.
Given this integrated spatiotemporal setting, a vulnerability analysis is more easily developed, again at various spatiotemporal scales.
Vulnerability Analysis
Within the context of environmental risk assessment, vulnerability is defined as a measure of a system’s likelihood to experience adverse effects due to exposures from environmental perturbations or stressors, such as chemicals (Ippolito et al., 2010). Not surprisingly, these environmental perturbations are referred to as threats. Landscape or regional vulnerability analysis is comparable to chemical stressor evaluations focused on individual species or their populations. For instance, benthic invertebrates may be more vulnerable than open-water biota to poorly water-soluble materials released in aquatic environments. Sensitivities to environmental stressors, such as chemical contaminants may predispose some organisms as receptors or target species that are regarded as representative species for guilds of similarly exposed biota. Parallel to risk analysis, vulnerability analysis focuses on the quality of a resource at risk, including the frequency of occurrence and intensity of the hazards.
Regardless of the spatial scale, for many environmental stressors a fundamental limitation to analysis of risk is lack of data. Yet, in the likely event that data are not sufficient for a fully implemented quantitative analysis, reliance on categorical tools to evaluate hazards can be used in risk-management practices to develop a better understanding of the uncertainties associated with sparsely available empirical data; for example, as part of a hazard-based evaluation or risks.
Uncertainty
As Brewer and Gross (2003) observed, the desire to forecast risks and consequences linked to systemwide change encounters many challenges, prominent among them is accounting for uncertainly in models of ecological and physical processes. Such challenges confront risk analysts and natural resource managers who work within a network of multiple stressors and multiple hazards. Two general types of uncertainty—aleatory uncertainty and epistemic uncertainty—affect the characterization of risks, especially within their roles in influencing risk management. Epistemic uncertainty is also referred to as subjective uncertainty, parameter uncertainty, or model-specific uncertainty, and more simply stated, reflects uncertainty linked to our state of knowledge. On the other hand, aleatory uncertainty, frequently referred to as stochastic uncertainty, relates to the inability to fully characterize a model of a system that represents higher levels of development than those detailed by basic events in a process. Often, such sources of uncertainty are confused with statistical uncertainty, which is largely a complex of measurement error, observational error, and model-dependent assumptions underlying any particular method applied to data analysis. In any process, basic events more often than not consist of lower-level events, each potentially characterized by processes subject to failure or loss of integrity.
When applied to science-based discussions regarding risks, uncertainty often reflects a level of complexity that is inadequately appreciated. For example, in analysis of complex adaptive systems, such as ecosystems, the evaluation of model, parameter, or aleatory uncertainty is often based on expert opinion out of necessity (Krueger et al., 2012). Some types of uncertainty are more easily quantified than others, although a complete quantitative treatment of all types of uncertainty is often not achievable. For example, uncertainty arising through error, bias, and imprecise measurement, and uncertainty arising through inherent variation in natural parameters can be addressed through sampling in the field or in other forms of acquiring data. Uncertainty can be estimated by the quality and quantity of data obtained and can be regulated to some degree by creating study designs wherein the amount and type of data that will be accepted are defined prior to the study. For example, do we need 10 samples or 100? Do we measure to 1 g/L or 1 μg/L? In contrast, epistemic uncertainty arises through a lack of knowledge or scientific ignorance, which reflects uncertainty related to the state of knowledge (or rather lack of knowledge); thus, epistemic uncertainty is frequently termed irreducible uncertainty. Using the previous examples, we might not have any idea of how many samples will be needed or what level of accuracy in measurements are required. Each of these types of uncertainty undoubtedly exists in every analysis or forecast that forms part of risk characterization. Regardless of resource management needs, and in many respects the driver behind much of the research focused on uncertainty analysis as part of risk characterization, we close with a brief overview of cause-and-effect analysis. Although a longtime research interest in many disciplines, cause-and-effect analysis remains a critical element linked to the evaluation of environmental risks.
Cause-and-Effect Analysis
Robert Koch (1843–1910) was a prominent physician and pioneering microbiologist/epidemiologist who developed four postulates, all of which must be true for an infectious organism to be responsible for a disease: (1) the organism must always be present, in every case of the disease; (2) it must be able to be isolated from a host containing the disease and grown in pure culture; (3) samples of the organism taken from pure culture must cause the same disease when inoculated into a healthy, susceptible animal in the laboratory; and (4) the organism must be isolated from the inoculated animal and be identified as the same original organism first isolated from the originally diseased animal. To some degree, these postulates pertain to ecotoxicology with a few modifications, mostly dealing with the terminology of replacing organism with chemical or contaminant, and replacing pure culture with something like exposed population.
With exposures to multiple stressors, cause-and-effect analysis frequently calls for the application of Koch’s postulates (Rothman, 2012) as part of the analytical process, but the evaluation of cause and effect for complex systems goes well beyond that process which was critical to identifying numerous causative agents of infectious disease (eg, cholera, tuberculosis, typhoid, diphtheria, and others) in the late 19th century. Indeed, exposures to multiple stressors complicate the process outlined by Koch’s postulates. However, early foundations of cause-and-effect analysis that preceded those efforts focused on communicable diseases that demonstrated the value of a systematic, problem-solving approach. When applied to exposures involving multiple stressors, evaluations of cause-and-effect relationships may still be accomplished but they require a degree of complexity similar to that of the system being addressed.
Integrated field and laboratory investigations and manipulative studies have contributed to the process of cause-and-effect analysis for evaluating multiple-stressor exposures. With varying levels of effect, these studies moved from exploratory studies to those more focused on establishing relationships among specific stressors in environmental exposures, as, for example between cancer and the presence of polycyclic aromatic hydrocarbons (PAHs). Through experience and many studies, scientists have developed postulates similar to those of Koch and formalized these into a discipline called “ecoepidemiology” (Susser and Susser, 1996a,b; Susser, 2004).
Cause-and-effect analyses lead, in part, to an emphasis on risk management practices that ultimately stem from both estimates of exposures to chemicals in the field and from toxicity benchmarks derived in the laboratory as inputs critical to risk characterization. Cause-and-effect analyses assumes various guises, depending upon their specific applications and implementations For example, they may lead to identification of a disease agent or identification of a root cause of failure when your computer operating system malfunctions (Pearl, 2009). However, a common foundation binds these various applications because all are characterized by response and explanatory variables, and risk factors that influence the expression of response (or nonresponse, as it may be). Explanatory and response variables may be direct or indirect in their association (Fig. 13.4), with multifactorial systems being complicated by interactions among component factors that influence the expression of response. The degree of complication in multifactorial systems yields various categories of cause. For example, a frequent categorization of cause used in epidemiological studies identifies factors as necessary or sufficient. These terms are most easily illustrated in simple systems, wherein a cause is sufficient if it inevitably yields an effect. In multifactorial processes such as disease or an otherwise compromised state of a complex system, sufficient cause nearly always occurs as a set of interacting component causes, where one component is commonly described as the cause. A necessary cause must always be present to produce a specific effect, outcomes classically observed in epidemiological studies focused on human health or investigations of fish or wildlife disease outbreaks (eg, Myxobolus cerebralis must always be present for a diagnosis of whirling disease in salmonid fishes). However, M. cerebralis may be present in very low titers and no disease is expressed, so the mere presence of M. cerebralis is not sufficient to cause the disease; it must be at some higher level to induce the signs of the disease. Yet many infectious and noninfectious diseases—the latter frequently of concern to chemical exposures—may be produced by different sufficient causes that may or may not have component causes in common. Uncomplicated infectious disease is frequently characterized by a disease agent that serves as a necessary cause, and in some instances, a factor may be “necessary and sufficient,” depending on the specific process being considered. In ecotoxicology. a particular contaminant, say p,p′-DDE, may be necessary in all species to induce a specific kind of eggshell thinning, but the concentration of DDE sufficient to produce thinning is species-specific. Infectious diseases are often host-specific.
In multifactorial processes such as those characteristic of environmental exposures to chemicals, factors may be necessary, sufficient, neither, or both. Component causes in a multifactorial system are generally characterized as predisposing factors, precipitating factors, reinforcing factors, and enabling factors. In characterizing failures in biological systems, predisposing factors are those that increase susceptibility (eg, that of a host to a disease agent or process). For example, the manifestation of disease in a host is frequently influenced by its immune status. Similarly, within an ecotoxicological context, a predisposing factor might be age, as that factor may be critical to the development of metabolic detoxification systems. Precipitating factors are those that are associated with the definitive onset of response, but are not sufficient in the absence of a necessary cause (eg, whirling disease may be precipitated by infection with M. cerebralis in susceptible individuals). Reinforcing factors are those that aggravate the expression of response, which in the case of disease agents might be repeated exposures in a genetically predisposed, susceptible individual or in an ecotoxicological context, to a behavioral factor that exacerbates exposure to chemically contaminated prey (eg, failure of avoidance behaviors). Enabling factors tend to be less clearly characterized than other categories of component factors because they are those components of exposure that facilitate the expression of response (eg, dry years may enable disease outbreaks to occur in grassland habitats that usually have a low incidence of disease, such as those reliant on arthropod vectors, or reduced prey-base may enable disease outbreaks predicated on malnutrition of the host). An example in an ecotoxicological context would be high uptake of heavy metals in waters with low calcium concentrations.
Epidemiological cause-and-effect models approach ecological complexity when disease processes are considered within a field setting where multiple stressors are common features of exposure. In such settings, simple linear models of cause and effect may be of limited usefulness, since multifactorial systems are characterized by having factors with varying intensity that interact at various levels in the system. These multifactorial systems are adaptive and highly dynamic, yielding a “web of causation” (Krieger, 1994), in which both epidemiology and ecotoxicology wrestle with exposures to multiple stressors of various types—biological, chemical, or physical. Indeed, beyond simple cause-and-effect analysis focused on identification of a single disease-causing agent or characterization of effects linked to a single toxicant, exposures in the field must necessarily acknowledge the ever-present role of confounding factors that inevitably produce spurious associations among variables and potentially mask real cause-and-effect relationships.
Study Questions
1. What is meant by risk in ecotoxicology? What poses the risk, and what is at risk?
2. Identify at least five types of stressors on populations of organisms. Can we say that any one stressor is more important than others for organisms in the same group (eg, terrestrial mammals)?
3. What is the difference between risk analysis, risk assessment, and risk management?
4. Outline the steps of a generalized risk analysis process.
5. Under what conditions might a manager accept some degree of risk? (Ie, if the manager decides that she can tolerate some risk.)
6. What are the four ways of making management decisions based on risk assessment as described by Diamond et al. (2015)?
7. What is a hazard-based assessment? How does it differ from a risk-based assessment?
8. Discuss with your class the concept of multiple stressors. What are some of these stressors? Which model, single or multiple stressors, better reflects real-world situations? How do multiple stressors interact to make an assessment more complex?
9. Describe how competitive, cumulative, aggregate, and residual risks are similar and different from each other.
10. What contributes to uncertainty in a risk assessment? How does this uncertainty affect the development of management decisions?