SCOPE 53 - Methods to Assess the Effects of Chemicals On Ecosystems

A critical first step in studying the effects of chemicals in ecosystems is making a clear statement of the ecosystem values to be protected. Historically human societies have placed value on ecosystems based on the ability of the ecosystem to provide products and services deemed necessary by the society. These values have varied between societies and over time, but all have had an economic basis. An awareness of the historical emphasis on economic benefits and a desire to communicate the wide range of benefits provided by ecosystems in meaningful ways have resulted in the development of the Ecological Benefits Paradigm. This paradigm, which classifies ecosystem benefits or values as use versus non-use benefits, is introduced in Section 2.1 and discussed in greater detail in Chapter 19. 

This paradigm facilitates environmental management decisions by providing a means of comparing and prioritizing essentially dissimilar environmental situations, and the efforts and resources available to manage and protect these ecosystems.

To maintain the availability of the desired ecological benefits, the conditions of an ecosystem that is capable of supplying the benefits must be described. Environmental managers must be able to characterize the desired condition (health) of the ecosystem in a way that will allow assessment of nominal (desirable or acceptable ecological condition) versus subnominal (undesirable or unacceptable ecological condition) states, so that they will know when a problem develops that deserves attention. Section 2.2 describes methods to characterize ecosystems. Environmental managers must also be able to clearly state the ecosystem quality objectives (Section 2.3) to be achieved so that the desired benefits may be maintained.

To protect the ecosystem and maintain the desired benefits, full analysis of the historical, current, or potential environmental contamination threatening the ecosystem is necessary. The steps in conducting comprehensive environmental assessment studies are presented in Section 2.4. Ecological risk assessment is offered as a tool that environmental managers may use to compare environmental risks and, thereby, more effectively prioritize actions and the allocation of resources to protect or restore ecosystems.

If risks are not translated into familiar or relevant terms for use by policy makers, little may be done to manage or reduce them. The importance of effectively communicating the results of environmental monitoring and assessment and the decisions based on environmental studies is discussed below.

2.1 THE ECOLOGICAL BENEFITS PARADIGM

Degradation of ecosystems is generally agreed to be undesirable. Many reasons exist to preserve an ecosystem, and they vary according to the needs of the interested parties. Some of the numerous motivations to preserve the integrity of ecosystems, discussed in Chapter 1, are social and cultural, aesthetic and ethical, and the traditionally important economic motivations. Terrestrial and aquatic ecosystems provide a wide array of goods and services for human use; historically, the desire to maintain the flow of these goods and services has been the impetus for action to preserve and protect ecosystems. Because most environmental policies and management actions are initiated on the basis of perceived degradation or loss of these goods or services, ascribing value to ecosystems based on their ability to provide such benefits affords a useful framework to evaluate the analytical methods available to study the effects of chemicals on ecosystems. Historically, most evaluations of ecological effects of chemical contamination have related impacts to effects on the supply of products and services of importance to human cultures. Recently, a framework was developed that attempts to classify ecosystem values in terms of the benefits they provide to humans as use versus non-use benefits. This framework, known as the Ecological Benefits Paradigm, is discussed in greater detail in Chapter 19 of this document. Some benefits can be quantified easily, because they relate to the production of commodities and provision of services and amenities. Examples of these uses that are directly related to market benefits are food crops, water resources, and construction materials. Use benefits that do not provide marketable commodities or services, such as recreational opportunities, are referred to as non-market use. The non-use benefits, such as those that relate to ecological quality (e.g., biodiversity) are more difficult to quantify; in the Ecological Benefits Paradigm, they are known as "neglected benefits." Increasingly, the protection of ecosystems for "neglected benefits" is being recognized by scientists and environmental managers. Table 2.1 presents the types of benefits assigned to ecosystems by the Ecological Benefits Paradigm.

The Ecological Benefits Paradigm is useful to environmental managers by providing a method to prioritize ecosystem benefits, and, thereby, help to allocate the limited resources available for monitoring, managing, and protecting ecosystems. The relative value of protecting an ecosystem depends, in part, on the needs of the local human population; often the benefits that one group considers important are not valued by another group. These varying value systems may result in conflicting priorities for environmental management. For example, in some countries, the expanding industrialized populations seeking recreational opportunities may place demands on the same undeveloped areas that other, less industrialized indigenous groups have relied on for subsistence. Another example is seen in the situation that occurred in Minimata Bay, Japan. Protecting Minimata Bay from mercury contamination was important to the local population who relied heavily on fish from the bay for food. If the people living near Minimata Bay had obtained their food from other sources, they would not have been affected by mercury contamination in the bay, and the relative value of monitoring and protecting the ecosystem may not have been considered worth the cost based solely on the market benefits of the ecosystem.

Table 2.1. Benefits of the Ecological Benefits Paradigm

Ecological, as well as market benefits, provided by an ecosystem should be considered in developing environmental policy. Although these so-called neglected benefits have not been neglected or misunderstood by ecologists, they have been virtually ignored in the development of national policies, partially due to the difficulty in characterizing or quantifying them. Furthermore, these non-use benefits can be quantified only by indirect economic methods (e.g., economic benefits of a recreational area), and cannot be related easily to small changes in ecological condition. If environmental policy makers wish to emphasize benefits that can be affected by small or incremental changes in ecological condition, they must focus on benefits for which the most direct measures are available. Unfortunately, this situation could lead analysts to continue to ignore ecological benefits that are difficult to quantify. Considering quantifiable benefits exclusively, however, is acceptable only for preliminary analysis because other benefits could easily be so important as to significantly alter the outcome of the analysis.

Although applying the Ecological Benefits Paradigm pushes ecological science closer to what some consider an undesirable tendency to value ecosystems in a purely economic manner, its use could improve environmental policy making and management strategies by considering all ecological goods and services. This framework would allow ecologists to describe ecological benefits to policy and decision makers in an easy to understand format, and would help clarify policy makers' preferences by stating them in terms of scientifically defensible data and not as subjective values.

An example of how the use of the Ecological Benefits Paradigm might facilitate decisionmaking by policy makers is when considering the management decision to harvest trees from a hypothetical old-growth forest. The authors point out that the greatest benefits from a virgin old-growth forest are realized from non-market benefits. Once the decision has been made to harvest the old-growth trees, most of the non-market benefits have been lost, but the market benefits from timber sales have been realized. Continuing the analysis provides an important insight: after the trees are harvested, not only are the non-market benefits lost, so are the market benefits. The resulting benefits profile of the harvested forest is much smaller than the benefits profile of the intact forest. 

2.2 CHARACTERIZING THE CONDITION OF ECOSYSTEMS

2.2.1 DESCRIBING THE CONDITION OF ECOSYSTEMS

To evaluate an ecosystem's ability to provide the desired benefits, ecologists, managers, and policy makers  must have a reliable method to characterize ecosystem condition. Defining a "healthy" ecosystem provides a baseline against which to evaluate the current condition of an ecosystem and to predict the potential effects of changes in that condition over time (i.e., trends in resources). 

Some ecologists have attempted to describe ecosystem "health" in terms of human health. However, this is difficult or even impossible, because the kinds of analytical tools used in medical science do not exist in ecological science. However, two distinct features of ecosystems, stability and resilience, may be potentially useful in describing a measure of "health" in ecosystems (Pimm, 1984; Holling, 1986). Costanza (1992) identified six major concepts derived from the studies of ecosystem stability and resilience that could provide a framework for characterizing ecosystem health. These concepts define ecosystem health as: homeostasis, absence of disease; diversity or complexity; stability or resilience; vigour or "scope for growth;" and balance between system components.

As Costanza and Principe state, these concepts, while useful, are limited in their ability to describe complex ecosystems. They do, however, provide a basis to develop a more practical definition of ecosystem "health" that can be described as: a system free from "distress syndrome" (defined as the irreversible processes of system breakdown), stable, and self-sustaining. A "healthy" ecosystem is one that is active, maintains its autonomy over time, and is resilient to stresses (Haskell et al., 1992). 

2.2.2 STRUCTURAL AND FUNCTIONAL CHARACTERISTICS TO EVALUATE THE CONDITION OF ECOSYSTEMS 

To characterize ecosystems, the structural and functional properties of ecosystems must be differentiated, particularly in the context of evaluating the effects of chemical or other stresses on an ecosystem. To properly characterize an ecosystem, both structural and functional parameters must be examined. Structural parameters (e.g., productivity, species composition, and demographic descriptors) and functional parameters (e.g., nutrient cycling and trophic structure) are interdependent, but have distinct characteristics. Structural parameters describe the parts that make up an ecosystem, that is, they simply tell what is there. Functional parameters describe the actions or inner-workings of a system and how the components work. Structural characteristics, descriptive of ecosystem components, generally are easier to assess (with the exception of underground microbiology), because fewer parameters can be measured over a shorter time period. Structural properties correlate with functional properties, that are expressed as rates, and must be measured in a more complicated manner. Measuring a functional property such as decomposition of plant matter on a forest floor requires taking multiple, complex measures over time; yet describing microbial composition is all the more difficult. 

Functional redundancies can buffer ecosystems from the effects of perturbations within specific populations. Thus if groups of organisms are killed, their functions may be assumed by other groups of organisms. The overall effect is that the particular ecosystem functions remain unchanged, while the structure of the ecosystem changes substantially. Structured populations are able to resist stress-induced changes through compensatory alterations in growth and reproduction, which also may affect the structural characteristics of the ecosystem. These mechanisms for mediating stress have limitations, and, if the magnitude or duration of the stress exceeds these limitations, the condition of the ecosystem may decline. Environmental management strategies often give higher priority to sustaining ecosystem function than to preserving ecosystem structure; however, an ecosystem in which essential functions are performed by a severely reduced community may be more sensitive to stress in the future.

Several functional and structural properties of ecosystems can be monitored to evaluate ecosystem condition. Examples of functional parameters proven to be valuable indicators of stress are primary and secondary productivity, decomposition, and mineralization. Primary productivity, the rate at which green plants convert light energy into organic matter, provides the energy essential for an ecosystem to operate, and is dependent on abiotic factors (e.g., light, temperature, humidity, soil structure), biotic factors, and trophic interactions. Secondary productivity, the rate at which herbivores utilize the energy from green plants and, in turn, supply energy for carnivores, is dependent on primary productivity, but is affected by other factors as well. Primary and secondary productivity may be influenced both directly and indirectly by chemical stresses. Decomposition of organic matter depends on climate, chemical composition of the substance, and type and quantity of decomposer organisms. The rate of decomposition appears to be a valuable indicator of any kind of ecosystem stress as any chemical stressor affecting these factors will alter decomposition rate. Mineralization rates of nitrogen, sulphur, and phosphorus compounds are also critical functional indicators of ecosystem health. These functional parameters are discussed in greater detail elsewhere in this document. 

Table 2.2. Structural and functional characteristics of aquatic ecosystems

Structural properties of ecosystems may also be useful indicators of ecosystem conditions. These may include: changes in species composition and abundance (i.e., increase in stress-tolerant species or decrease in native species with a corresponding increase in introduced or exotic species); reduced biodiversity; shorter, less complex food webs; reduced population density; and reduced genetic diversity.

Tables 2.2 and 2.3 summarize some structural and functional characteristics that are useful indicators of the condition of some major aquatic and terrestrial systems. These parameters may be monitored to evaluate changes in ecosystems or to assess trends that may suggest the need for concern. 

Table 2.3. Structural and functional characteristics of terrestrial ecosystems

  2.3 SETTING AND MAINTAINING QUALITY OBJECTIVES FOR ECOSYSTEMS 

Having defined the values of an ecosystem and identified its characteristics, environmental managers can set ecosystem quality objectives or quantitative goals for acceptable conditions. Quality objectives can be described as the specific ecosystem conditions that must be achieved or maintained to ensure that the desired state of an ecosystem is maintained, and that the desired ecosystem benefits are available. Examples are habitat continuity and maintenance of species parameters. Such objectives provide a standard with which to compare the effects of chemical stress and to evaluate the effectiveness of corrective actions.

A wide variety of natural and anthropogenic stressors can influence ecosystem quality objectives. While this book focuses on chemical and physical stressors, the impact of naturally occurring and non-chemical stressors should not be underestimated. Habitat destruction is perhaps the most important and devastating ecosystem stressor; but from the point of view of this book, non-chemical stressors are important only to the extent that they interact with chemical stressors and to the extent that the effects of chemical and non-chemical stressors must be separated from one another to select appropriate regulatory and remedial strategies. Table 2.4 summarizes some common stressors that influence ecosystem quality objectives.

Table 2.4. Stressors that can influence ecosystem quality objectives

Synergistic interactions between anthropogenic and natural stressors or among pollutants may increase the difficulty of identifying the principal cause of injurious effects. Many examples of interactions between non-chemical and chemical stressors exist. Generally, an ecosystem already weakened by stress is much more susceptible to the effects of additional stress. For example, a forest stressed by drought is more likely to be affected by either acid rain or sulphur dioxide (SO₂) than one that has been receiving an adequate amount of water. Experiments have proven that Norway spruce seedlings are quite sound after several weeks of exposure to high levels of SO₂ (200 ppb), whereas they are severely stressed if they experience drought at the end of the exposure period. Seedlings exposed to SO₂ recover from drought slower than those that endure drought without being exposed to SO₂. In the Vosges region of northeastern France, a seven-year field assessment of mineral nutrition in spruce and fir growing in very desaturated soils showed that magnesium content is generally low, but is much lower in dry years. Furthermore, terrestrial mammals and birds stressed with high body burdens of organochlorines have been shown to be more susceptible to disease and to have a large proportion of deformities. 

2.3.1 SETTING QUALITY OBJECTIVES 

Identifying and establishing quality objectives for an ecosystem are more easily accomplished by evaluating the baseline conditions of the pristine ecosystem. If historical data are unavailable to provide precontamination information, relying on data collected at carefully selected reference sites may be necessary to compare against the impacted study areas. These areas should be as similar to the study area as possible to eliminate variables in the extrapolation of data from the reference site to the impacted sites. However, some uncertainty will always exist with this approach.

Local conditions must be considered when setting quality objectives. Local conditions at one location may result in that ecosystem responding at a different rate than at another, or may result in reduced or increased resilience. For instance, the type of underlying bedrock present may provide some buffering of the effects of acid rain surrounding one lake and not at another.

Departures from ecosystem quality objectives can be defined as effects, which may vary in magnitude and occur over different time scales. Furthermore, some effects may be reversible, while others may persist for long periods. Environmental managers must decide how great an effect needs to be over a specific area, and how long it must persist before regulatory and remedial actions become necessary. Accidental releases of chemicals can be useful study exercises to observe effects over a gradient of exposures in a way that would normally be impossible. The opportunities for research on the effects of petroleum products on coastal ecosystems, birds, and mammals provided by accidental oil spills are discussed elsewhere in this volume. 

2.3.2 MAINTAINING QUALITY OBJECTIVES

Identifying the potential effects of chemical contaminants before a release occurs has become vital in maintaining ecosystems. A thorough evaluation of a chemical prior to its commercial use has become a common regulatory policy in many countries, and is designed to prevent damage to ecosystems. In addition, efforts are being extended to implement waste-limiting technologies to minimize industrial discharges. The introduction of between 1000 and 2000 new chemicals yearly necessitates evaluation of potential environmental toxicity of these chemicals (Xu and Pang, 1992). Prior to industrial-scale production, new chemicals are tested to varying degrees to define potential environmental effects. Such information can help to prevent environmental contamination. Often, a tiered structure of decision-making is designed to define criteria for higher level testing. 

Because of the enormous volume of chemicals requiring evaluation, test procedures must be cost-effective, rapid, reproducible, and ecologically relevant. Current methods to evaluate ecotoxicity are largely laboratory-based, and include quantitative structure-activity relationship (QSAR) approaches, acute toxicity testing with selected species, multiple species testing, and life-cycle analyses (Xu and Pang, 1992).

Identifying potential effects of chemicals prior to release to the environment is founded on two principal methodologies: (1) controlled studies to evaluate the potential effects on specific ecosystems and (2) methods to provide an early warning of adverse effects in an ecosystem (i.e., monitoring programs). Controlled studies range from those of simple, acute toxicity to the more complex ones with microcosm and mesocosm. Attempts to predict more ecologically relevant changes frequently involve the use of microcosms and mesocosms, in which effects on food chains and ecosystem structure and function can be evaluated. The use of microcosm and mesocosm studies to evaluate contaminant effects on aquatic and terrestrial ecosystem is discussed in Chapter 3. Biomonitoring strategies (i.e., the bioavailability of chemicals is a measure of the extent to which a chemical accumulates in an organism) may serve as early warning signals of significant chemical contamination of an ecosystem. The use of biomarkers can also provide an indication of the need for regulatory action, and can assist in evaluating the effectiveness of remedial action. Biomonitoring strategies and biomarkers are discussed in Chapter 3. 

A policy approach to maintaining environmental quality objectives is to define "exclusion zones" around a chemical release or discharge. Such areas permit limited deviations from quality objectives that are deemed to be acceptable (Peakall, 1992). Their definition provides a balance between the sometimes conflicting needs of industrial production and protecting the ecosystem. The economic benefits of allowing an exclusion zone are weighed against protecting the ecosystem, and sometimes the economic benefits can be determined as outweighing the damage to the ecosystem. 

The time scales of chemical contamination are also important, particularly for persistent pollutants, because several years may be needed for contaminants to be reduced to nominal levels (Burger and Gochfeld, 1992). In some instances, ecosystem recovery to within quality objectives can be accelerated by restoration procedures. For example, the River Thames (England) has historically been negatively impacted by stormwater carrying untreated sewage,  which led to reduced  dissolved oxygen levels and to frequent dieoffs of aquatic organisms, leading subsequently to drastically reduced biodiversity in the river. Eventually, only a few species remained that are known to be tolerant of low oxygen. By controlling the stormwater surges to the river, the biodiversity in the river has been gradually increasing (i.e., pollution-tolerant organisms are being replaced by organisms less tolerant).  

2.4 STUDYING THE EFFECTS OF CHEMICALS

2.4.1 ECOLOGICAL RISK ASSESSMENT AS A TOOL

Ecological risk assessment is a structured approach with four components that provide a systematic method to assess the effects of chemical contaminants on ecosystems. The four components of ecological risk assessment are:

1. Problem formulation: to identify and describe the ecological damage and resources potentially at risk.

2. Exposure assessment: to describe the magnitude, duration, frequency, and routes of contaminant exposure to potential receptors.

3. Ecological effects assessment: to identify the nature of the hazards associated with the contaminant(s) and to quantify the relationship between exposure stress and receptor characteristics.

4. Risk characterization: to integrate the exposure and hazard information by estimating the likely incidence of an effect under the conditions described.

The ecological risk assessment process is most useful to obtain the information necessary to identify the potential hazards from contamination, determine the relative risks associated with contamination, and facilitate management of risks in a manner that best protects social, cultural, political, and economic benefits associated with the area. The process allows the comparison and prioritization of various essentially dissimilar local situations by focusing on the potential risk to biota.

Ecological risk assessment is an approach to guide the process of evaluating chemically contaminated areas. This section provides a discussion of essential steps of any ecological study to evaluate chemical contamination of an area. These generalized steps are: problem formulation, study design, and data analysis using statistical and modelling applications.

2.4.2 PROBLEM FORMULATION

2.4.2.1 Establishment of goals

The first step in diagnosing the cause of an environmental problem is to state clearly the questions to be answered. A clearly focused statement of goals assists in the selection of test biota, sampling sites, sampling and analytical methods, etc. Generally, the questions centre on identification of contaminants in the area, whether their levels are elevated above background levels, their bioavailability, and their potential to injure biota.

Table 2.5. Characteristics of an area to plan an ecological risk assessment

2.4.2.2 Identification and characterization of chemical stressors

Chemical stressors must be identified and characterized early in the study design process. An understanding of the most likely pathways of movement of contaminants through the ecosystem helps guide the sampling process. Chemical characteristics will sometimes help predict a contaminant's fate and transport and its temporal and spatial movement through site media. Knowing where the chemical is most likely to be compartmentalized and whether it is likely to bioaccumulate in biota helps to determine the most appropriate sampling and  analysis methods.  

2.4.2.3 Characterization of the ecosystem potentially at risk 

Thoroughly characterizing an area before conducting an ecological assessment can result in substantial savings in time and resources. Relevant characteristics of an area to be evaluated before deciding which assessment methods to address the desired endpoints and yield relevant data are listed in Table 2.5. Narratives discussing the importance of each characteristic are available in other references; this section provides only a listing of the specific characteristics.

2.4.2.4 Identification of relevant endpoints 

The concept of ecological endpoints, described as expressions of the values or benefits to be protected, provides a useful method to guide the study design (Suter, 1989). A clear statement of the endpoints helps focus the study, and ultimately facilitates communication of the study results. Assessment endpoints are described as formal expressions of the actual ecological values to be protected, referred to as environmental characteristics at risk from contamination. Characteristics of good assessment endpoints are: social relevance, biological relevance, unambiguous descriptor, measurable or predictable; capable of being damaged by contaminant, and logically related to the final decision (adapted from Suter, 1989, 1993). Examples are species diversity and reproductive integrity of a valued species. 

Measurement endpoints are quantitative expressions of the assessment endpoints, that is, the actual measurements of the characteristics that are to be protected. Criteria of good measurement endpoints are: corresponding to or predictive of the assessment endpoint; readily measured; appropriate to the scale, exposure pathway, and temporal dynamics of the study area; low in natural variability; diagnostic; broadly applicable; and standardized. They also have existing data sets (Suter, 1989, 1993). Examples of measurement endpoints to quantify the specific assessment endpoint of species diversity are species abundance and composition. An example of a measurement endpoint for reproductive integrity of a valued species is age/size class structure of the population.  

2.4.3 STUDY DESIGN 

2.4.3.1 Types of sampling designs

Sampling locations may be determined quantitatively by random or non-random methods. Conclusions from random methods are the more reliable. Common types of random sampling designs are simple random and stratified random. The simple random method is most useful in areas that are highly uniform in habitat patterning, such as grasslands; however, sites not uniform in contaminant concentration, or other characteristics, make purely random sample site selection less efficient. Cluster sampling is generally used at a chemically contaminated study site. 

Stratified random sampling is a modification of random sampling in which similar habitats within the larger study area are selected, and sites in those similar habitats are selected randomly. Stratified random sampling provides the opportunity to conduct some statistical analysis while increasing sampling efficiency. Examples of stratified random sampling are studies of the different levels in a forest ecosystem and studies of aquatic systems where the area is first stratified by salinity then sampled randomly.

If quantitative data are not required for statistical analysis, non-random sampling may be conducted, in which specific study sites are located along a transect or on a predetermined grid within the area. While efficient, non-random sampling may not provide data suitable for statistical analyses, thereby introducing immeasurable uncertainty in the conclusions. 

2.4.3.2 Test method selection 

No single type of test can provide all of the information necessary to determine the extent and magnitude of environmental contamination. Therefore, several types of tests should be employed. Test methods typically include field surveys, chemical analysis of media samples, toxicity tests (laboratory and field applications), and bioaccumulation studies. Additional studies such as biomarkers and microcosm and mesocosm studies may also be included. The advantages and requirements of numerous test methods for aquatic and terrestrial ecosystems  are discussed in Chapter 3. Only together as part of a carefully designed field or laboratory study can these tools provide the information necessary to establish linkages between contamination and adverse effects in biota.  

2.4.3.3 Quality assurance procedures 

Any environmental data collection program is only as good as the data it collects. Thus, any program must be designed with an adequate level of quality assurance (QA) to ensure that the data collected are of the quality and quantity needed to answer the questions for which the study was designed. The first step in ensuring quality is to develop data quality objectives (DQOs) as part of the design process. DQOs are qualitative or quantitative statements of the  level of uncertainty that a decision maker is willing to accept in decisions made  with environmental data. 

Development of clearly focused DQOs very early in the study design process can save time and money, while  producing the kind of data that address the major issues of a study. The logical process for DQO development is described by Neptune and Blacker (1990). Their process, presented in Figure 2.1, was modified for use in chemically contaminated areas, and in the US contains six major steps (USEPA, 1993a).

Figure 2.1. The modified logic process to develop DQOs (USEPA, 1993a)

After DQOs have been developed and a study has been designed to meet these data objectives, the study details need to be developed. Typically, the DQOs are produced via a  quality assurance project plan (QAPP). QAPPs provide a link between the DQOs and the specific steps to be taken to achieve the data objectives. QAPPs are required by the USEPA in all of their data collection activities (USEPA, 1984). Elements generally found in USEPA's QAPP are:

1. description of the goals and objectives of the study;

2. organization of the project and quality assurance responsibilities; 

3. DQOs and sampling strategy and design;

4. sampling procedures;

5. sample custody, storage and shipping procedures;

6. calibration procedures;

7.  analytical procedures;

8.  procedures for data management (i.e., data reduction, validation, and reporting); 

9. internal quality control checks;

10. quality assurance audits;

11. quality assurance reports;

12. preventative maintenance procedures; 

13. data assessment procedures; and

14. corrective actions.

Other tools to ensure the quality of a data collection project include the use of performance evaluation samples (i.e., analysis of a sample with a known concentration of a substance), and development of standard operating procedures for the specific procedures to be used routinely in the project (i.e., sample collection and analysis procedures). Very large studies may employ other quality tools such as Quality Management Plans, data quality audits, and management and performance audits by outside parties. These approaches are generally applicable to most data collection projects, and help to ensure, along with proper quality  control measures, that  he data generated are of the quality and quantity needed to answer the questions posed by the study initiator and decision maker. 

2.4.4 STATISTICAL ANALYSIS AND MODELLING 

Statistical analysis is a necessary step to determine the importance of the data collected and to evaluate whether the data are sufficient to answer the questions posed by the study. Integration and synthesis of the data  collected into assessments of environmental condition are critical. The use of standard parametric procedures for analysis of normally distributed populations may be used in many applications. However, environmental data sets are frequently not normally distributed as is usually found with the life sciences. As such, special non-parametric (distribution-free) statistical procedures are commonly required. Environmental data also are frequently characterized by large measurement errors (random and systematic), missing or questionable data points, data near or below the detection and quantitation limits, and complex spatial and temporal trends and patterns (Gilbert, 1987). Thus, to minimize these factors as much as possible, statistical analyses should be planned carefully before the study is initiated.

Due to the complexity of environmental data, the presentation of results is critical. Graphical methods to display data should be employed wherever possible. Cumulative distribution functions (CDFs) are a recommended method to display data. CDFs display information on central tendency (mean, median) and data range in a graphical format. An example of a CDF is provided in Figure 2.2. 

The use of geographic information systems (GISs) is also recommended, because of its ability to display spatial and temporal data using a mapping format. Geographic information systems are especially useful to monitor a regional scale, and are discussed in Chapter 4.

Figure 2.2 Example of a CDF: a confidence interval of 90 percent for an estimated benthic index for the Virginian Province (adapted from USEPA, 1993b)

As an integral component of the assessment process, mathematical models are used to integrate data or to generate missing data. They may be either statistical, such as an analysis of variance, or mechanistic (a quantitative description of the relationship between certain parameters). Mechanistic models are most often thought of when the "model" is used. These models are applied to describe for a substance an organism's exposure, its toxicity, or its transport and fate. Examples of their use in ecological assessment include establishing links between the measured responses in receptors of concern to population-level responses and to information on the spatial extent of contamination. Models may also be used to incorporate the effects of chemicals on ecosystem structure and processes and to predict the likely fate and transport of chemical contaminants. Models can be categorized according to their structure. Some of the more complex models attempt to depict the cycling of materials such as nutrients or metals among biotic and abiotic components of the ecosystem, and are often associated with chemical loadings and mass balance studies or studies of the ability of ecosystems to process chemicals. Another category of models emphasizes individual processes, such as the transfer of chemicals between the different media or the assimilation of toxicants or nutrients by the biota. Another advantage of models is that they allow scientists to verify their assumptions about the way an ecosystem works. The use of mathematical modelling to characterize risk at the regional scale is discussed in Chapter 4. 

2.5 COMMUNICATING RISK 

The effective communication of environmental data informs interested parties in relevant and understandable terms about the scientific data, analysis, interpretation, and recommendations that resulted in the environmental decision. Communicating an environmental decision is just as important to the success of an initiative as the ecological study on which the decision is based. Attempts to maintain ecosystem quality objectives have a much greater chance of success if the decision and the reasons behind it are effectively conveyed to interested parties. Communicating environmental risks brings the management strategy full circle by linking the assessment results to the efforts to protect the ecosystem benefits of concern. 

Each society has a characteristic decision-making process: from a highly focused and hierarchical to broadly-based requiring active public participation. Since target audiences are different, communications pathways must be tailored to fit each. Although the following discussion focuses on a broadly-based process, the principle of matching the communications medium and the form of the message to the audience and the specific level of decision making is applicable to all situations.  

2.5.1 TARGET AUDIENCES

Communicating scientific information effectively requires identifying target audiences and using language and media that the audiences can most easily understand. Audiences for environmental monitoring and assessment information are varied, and can include:  

1. environmental managers, administrators, and regulators who require information to assess both risk and the effectiveness of regulations and management practices;

2.  political leaders and government agency officials who set public policies and allocate financial resources;

3. industrial leaders and private sector professionals who advise and determine the choices of technology, raw materials, and plant location for economically productive activities;

4. environmental organizations, community groups, and private citizens who influence decisions in the public and private sectors through advocacy, voting procedures, and lobbying activities;

5. university faculty and students, public school administrators, and teachers who determine the curricula that can be a means of educating the public in environmental issues; and 

6. researchers in the scientific community who determine the needs for additional scientific endeavour.

The diversity of audiences and their differing needs for scientific, technical, or interpreted information mandates a varied communications initiative.

2.5.2 CONTENT 

Information must be provided in a context relevant for the level of the decision making expected of the target audience. The need to collect data, perform analyses, or derive conclusions should be explained clearly within the framework of the decision-making process. The specific data collected, their organization and analysis, and the conclusions and interpretation can be presented at a level of technical detail appropriate for the audience and its decision-making responsibilities. The sources and types of data, their strengths and weaknesses, significance of the findings, and alternative interpretations provide a basis to evaluate the reliability and conclusiveness of the findings. 

Although most communication is presented through textual material, it is enhanced in clarity, persuasiveness, and understanding by the use of well-designed maps, charts, graphs, and other visual methods. For example, the presentation of the results of mathematical models, although highly technical in content, is an effective method to convey the dynamic nature of ecological functions and structures. 

2.5.3 METHODS 

The method of communicating data should vary with the audience being addressed. Scientific and technical data are initially entered, analysed, and stored in computerized data bases. Computer programs can also be developed to organize, structure, and present the data in a multilayered format for use at variable levels of technical complexity by audiences with different needs and skills. In addition, the data and interpretations may be presented in regular scientific and technical journals for peer review, critique, and validation, and through agency technical reports for program assessment, regulatory needs, and archiving. For wider use, data summaries and interpretations may be published in professional journals read by the relevant users. Scientific background articles in the general press and subscription periodicals and monographs, edited volumes, books, and textbooks inform the public, and serve as educational material for schools and universities. News reports appearing in the popular press, television, and radio are effective to achieve public awareness, but not to convey the full depth of technical issues of environmental problems. The accuracy of these initial reports, however, is critical in forming public attitudes; unfortunately, misinformation or misinterpretation of information is often very difficult to correct. Finally, oral presentations of reports to scientific societies, professional meetings, community groups, legislative hearings, and court proceedings are important additions to more formal communications. The various methods for communicating ecosystem data and the audiences for which the methods are appropriate are presented in Table 2.6. 

Table 2.6 Methods to communicate ecosystem data to various audiences

2.5.4 TIMING OF COMMUNICATIONS 

Proper timing to communicate monitoring and assessment information is vital to the overall success of management decisions. Input to databases for regulatory and assessment purposes is ongoing and not time sensitive. Status reports to agencies responsible for management are periodic as required by the program, and are time sensitive, but not urgent (i.e., years). Publication in scientific journals is appropriate to research schedules that have a long timeframe (e.g., months). Background articles in the public media are ongoing, but more effective when placed in the timeframe of pending public decision-making. 

The timing of communication is essential when viewed in the context of critical decision-making, such as research planning and financing, pending legislation, annual budget authorizations, and appropriation cycles (when lobbying efforts are maximal and effective), in relation to current or planned tort proceedings, in local or regional economic planning and development hearings and decisions, and in the course of formulating environmental impact assessments. Ensuring that monitoring information is communicated in a proper and timely manner is an essential component of any research and monitoring program. 

2.6 REFERENCES 

Burger, J., and Gochfeld, M. (1992) Trace element distribution in growing feathers: additional excretion in feather sheaths. Arch. Environ. Contam. Toxicol. 23, 105-108.

Costanza, R. (1992) Toward an operational definition of ecosystem health. In: Costanza, E., Norton, G.G., and Haskell, B.D. (Eds.) Ecosystem Health: New Goals for Environmental Management, pp. 236-253. Island Press, Washington, D.C.

Gilbert, R.O. (1987) Statistical Methods for Environmental Pollution Monitoring. Van Nostrand Reinhold, New York, 320 pp.

Haskell, B.D., Norton, B.G., and Costanza, R. (1992) What is ecosystem health and why should we worry about it? In: Costanza, E., Norton, G.G., and Haskell, B.D. (Eds.) Ecosystem Health: New Goals for Environmental Management, pp. 1-18. Island Press, Washington, D.C.

Holling, C.S. (1986) The resilience of terrestrial ecosystems: local surprise and global change. In Clark, W.C., and Munn, R.E. (Eds.) Sustainable Development of the Biosphere. Cambridge University Press, Cambridge, Melbourne, New York. 

Neptune, D. and Blacker, S. (1990) Applying the data quality objective (DQO) process to research operations. In: Hart, D. (Ed.) Proceedings of the Third Annual Ecological Quality Assurance Workshop, pp. 5-22. Environment Canada, Burlington, Ontario, and US Environmental Protection Agency, Environmental Monitoring Systems Laboratory, Las Vegas, Nevada.

Peakall, D. (1992) Animal Biomarkers as Pollution Indicators. Chapman and Hall, London, England, 291pp.

Pimm, S.L. (1984) The complexity and stability of ecosystems. Nature 307, 321-326.

Suter, G.W., II (1989) Ecological endpoints. Chapter 2 in: Warren-Hicks, W., Parkhurst, B.R., and Baker, S.S., Jr. (Eds.) Ecological Assessment of Hazardous Waste Sites. Report No. EPA 600/3-89/013. US Environmental Protection Agency, Environmental Research Laboratory, Corvallis, Oregon.

Suter, G.W., II (1993) Ecological Risk Assessment. Lewis Publishers, Chelsea, England, 538 pp.

USEPA (US Environmental Protection Agency) (1984) Policy and Program Requirements to Implement the Mandatory Quality Assurance Program. Order No. 5360.1. U.S. Environmental Protection Agency, Office of Administration and Resources Management, Washington, D.C., 3 April.

USEPA (1993a) Biological Assessment at Hazardous Waste Sites: An Overview of Issues and Methods. Proceedings of Seminar Series, 10 February 1993. Prepared for US Environmental Protection Agency, Office of Solid Waste and Emergency Response, Toxics Integration Branch, Washington, D.C., by Tetra Tech, Inc., Baltimore, Maryland.

USEPA (1993b) Virginian Province Demonstration Report: EMAP-Estuaries: 1990. Report No. EPA/620/R-93/006. US Environmental Protection Agency, Office of Research and Development, Washington, D.C.

Xu, X., and Pang, S. (1992) Briefing of activities relating to the studies on the environmental behavior and eco-toxicity of toxic organics. J. Environ. Sci. (China) 4(4), 3-9.