OVERVIEW: MEASURES OF ENVIRONMENTAL PERFORMANCE AND ECOSYSTEM CONDITION
Peter C. Schulze and Robert A. Frosch
". . . everything is an indicator of something but nothing is an indicator of everything. "
(Cairns et al., 1993, p. 6)
No metrics were necessary to recognize a problem when the Cuyahoga River caught fire in Cleveland. Likewise, any casual observer would realize that something is wrong with the Aral Sea, where commercial fishing vessels lie stranded in their ports dozens of kilometers inland. In both cases, shortsighted human behavior led to dire, readily detectable environmental consequences.
The problem with relying on such blatant evidence is that environmental deterioration becomes critical before a response is even contemplated, let alone implemented. Rather than wait for disasters to happen, one would prefer to avoid problems in the first place. This, however, requires the ability to predict the environmental consequences of human activities, avoid activities with unacceptable impacts, and document the environmental consequences of other activities. Any sophisticated attempt to predict and detect the environmental impact of human activities requires appropriate measurement methods.
Since the modern environmental movement picked up steam in the 1960s, we as a society have effectively addressed some of the most obvious or straightforward environmental effects of human activities. Gasoline sold in the United States no longer contains lead, hence, there is less lead in the air. One rarely sees a belching smokestack. The Cuyahoga River, much cleaner now, is not likely to catch fire again.
The human impacts on the environment that receive the most attention today
are often either obscure or complex. Human impacts on stratospheric ozone and greenhouse gases provide examples. Unlike a belching smokestack, the effect of chlorofluorocarbons on stratospheric ozone is apparent only to those with the expertise to understand the critical atmospheric chemistry. Meanwhile, a plethora of human activities contribute greenhouse gases to the atmosphere and thereby have the potential to alter climates. As a result of such obscure and complex relationships, it is difficult to predict the future environmental consequences of human activities.
As more obscure and complex relationships between humans and the environment are recognized, those charged with reducing and monitoring human impacts face greater challenges. They must shift their attention from obvious and relatively simple impacts (e.g., lead in gasoline or oil floating on a river) to more complicated processes (e.g., the effects of manufactured chemical compounds on animal development and human health [Sharpe, 1995]). Thus, they need to improve their ability to predict the ecosystem consequences of changes in human activities. As usual, better predictive abilities will require better measurement abilities.
One can envision a continuum of progress, from a past when there was little concern for the obvious environmental impacts of human activities, to the present when some obvious effects have been reduced but new understanding has led to new concerns, and, finally, to a future when comprehensive understanding of the environmental implications of human activities makes it possible to eliminate any activities that have unacceptable environmental consequences. The latter is probably too much to expect, given the diversity of ecological impacts, the potential for interaction between these impacts, and the complexity of the ecosystems involved, but it is a worthy target.
Over the past few decades, two new groups of professionals have emerged. One is responsible for managing and reducing the environmental impacts of human activities. The other is responsible for assessing and monitoring the conditions of ecosystems that are affected by human activities.
Those charged with managing and reducing environmental impacts have developed a suite of metrics for gauging environmental performance. Those who assess the status of affected environments have developed metrics for determining ecosystem conditions. However, surprisingly little interaction appears to take place between those working to improve environmental performance and those trying to monitor the condition of ecosystems. With so little interaction, those who work to improve environmental performance are rarely able to assess the marginal environmental effects of their improvements. Meanwhile, those who measure the condition of impacted ecosystems often lack information on the particular human activities that are responsible for changes in ecosystem conditions. As attention turns to more obscure and complex environmental impacts, the lack of communication between these two groups could substantially impede progress. How can a plant manager know which of two alternatives in a production process
would be least harmful to the environment? How can a designer know whether a compound made of material x would be better or worse for the environment than a compound made of material y? In some cases, the answers to these questions are straightforward, but in many cases they are not.
Ideally, a change in environmental performance could be measured in units of ecosystem condition. For example, computer design might be assessed a priori in terms of its expected per-unit impact on, for example, the water quality in streams that receive effluent from the computer factory. Recent efforts have attempted to extend life-cycle assessment procedures toward this goal (Steen and Ryding, 1992), but those efforts have been criticized because of the difficulty of ranking the environmental significance of different types of impacts. In other words, there is inadequate information on how different impacts affect ecosystems. As a result, critics are unconvinced that improvements in environmental performance measures will correlate with improvements in ecosystem conditions (Field and Ehrenfeld, this volume). Absent any correlation, efforts to improve environmental performance could be ineffective or even possibly counterproductive.
In other cases, there may be insufficient information to confirm that adherence to environmental performance standards does safeguard impacted ecosystems. For example, federal regulations require the use of bioassays to measure the toxicity of effluents released directly into waterways (Goulden, this volume). The presumption is that if the effluent bioassay results meet the environmental performance standard, then the ecosystem will not be harmed. Goulden argues that this is not a safe assumption. This appears to be another case of insufficient collaboration and cooperation between managers of environmental performance and managers of ecosystems. Hart (1994, p. 111) notes that the Intergovernmental Task Force on Monitoring Water Quality determined that ". . . for every dollar invested in programs and infrastructure designed to reduce water pollution, less than two-tenths of one cent (or 0.2 percent) was spent to monitor the effectiveness of such abatement programs!" Field and Ehrenfeld and Goulden essentially argue that one can not assume that an improvement in environmental performance, as measured by life-cycle assessment or effluent bioassays, leads to even an incremental improvement in ecosystem condition. Clearly, there is ample room for better coordination and collaboration between students of environmental performance and students of ecosystem condition.
Although it is probably too much to expect a comprehensive ability to predict how each particular human activity affects ecosystems, existing evidence suggests that collaboration between ecologists and engineers can be a powerful means of simultaneously achieving engineering objectives and environmental goals. Shen (1996) and Lindstedt-Siva et al. (1996) describe relevant examples from the fields of oil exploration and water management. In both cases, explicit environmental objectives served as engineering design constraints. Environmental scientists identified precise design criteria that engineers then used. The
projects described by these authors have not yet been completed, so convincing evidence of success must await studies of the actual impacts of those designs. Nevertheless, the designs appear to satisfy the particular ecological constraints that served as design criteria.
Strang and Sage (this volume) describe a similar collaboration, but one with a 30-year track record. Since 1965, Eastman Chemical Company (and its predecessors) has worked closely with aquatic scientists from the Academy of Natural Sciences in Philadelphia to improve the environmental performance of the Tennessee Eastman Division, a facility that releases effluents to the South Fork Holston River. Staff members of the Academy of Natural Sciences periodically assess the condition of the river. Eastman personnel then use the assessment results to help determine what specific objectives are appropriate for efforts to improve environmental performance. As Eastman modifies facilities and practices to improve performance, subsequent studies by the Academy of Natural Sciences document resulting changes in the river's condition. This procedure, which depends on a strong collaboration between corporate managers and aquatic ecologists, has led to a long record of continual improvement in both environmental performance and ecosystem condition. Eastman used what it learned from the Tennessee Eastman Division to design facilities on the White River in Arkansas. Studies of that river by the Academy of Natural Sciences have shown no differences between the ecosystem conditions upstream and downstream from the Eastman plant.
The close relationship between the properties of liquid effluents and the condition of receiving waters undoubtedly facilitated the success of the collaboration between Eastman Chemical Company and the Academy of Natural Sciences. Many other situations involve a less direct connection between environmental performance and ecosystem condition. Nevertheless, Strang and Sage's account confirms the potential for progress in environmental performance and ecosystem condition when corporate managers and environmental scientists collaborate closely.
This volume is intended to facilitate that progress by reporting on a variety of important metrics that have been developed to assess environmental performance and the condition of ecosystems. Our hope is that the discussion of these various indicators in one volume will foster not only further refinement of the particular measurement techniques but also more communication between users of the two sets of measures so that examples such as the one described by Strang and Sage will accumulate rapidly. The remainder of this chapter provides a brief summary of some of the important metrics that are used to assess environmental performance and ecosystem condition.
Examples of Environmental Performance Metrics
Measurement methods are being developed for a variety of purposes, from tracking the impact of a soda can to gauging the performance of national economies.
Life-cycle assessments attempt to summarize the environmental impacts of a product through its entire life-cycle, from the extraction of the raw materials through manufacturing, use, and disposal. Such information can serve as the basis of a search for design alternatives to reduce the environmental impact of a particular product or to compare several different technologies designed to serve the same function.
Although simple in principle, life-cycle analyses can be difficult to complete in practice because they require large amounts of information and frequently involve assessments of the relative importance of qualitatively different types of environmental impacts (Field and Ehrenfeld; Hocking, this volume). Comparing two alternatives that have qualitatively different environmental impacts is one situation in which collaboration between engineers and environmental scientists could be most useful. Engineers have the expertise to develop design options. Environmental scientists will have more information (although often not enough) about the environmental consequences of releasing different wastes. Even if it will never be possible to predict precisely how different design options will affect the environment, it seems self-evident that life-cycle analysts would benefit from the insights of ecosystem experts and that ecosystem experts would profit from an understanding of design options and production processes. If nothing else, such information could help set priorities for ecosystem research.
Field and Ehrenfeld elaborate on the difficulty of putting life-cycle assessment into practice. They explain the serious limitations that arise due to the inability to rank the importance of qualitatively different environmental impacts associated with different technologies or design alternatives. These limitations notwithstanding, the authors emphasize that life-cycle assessments help to illuminate the differences in the environmental properties of various technologies.
Hocking argues that the problem of ranking qualitatively different types of environmental impacts can be solved by using energy requirements as a basis for comparing different technologies or alternative designs. He notes that differences in the emissions characteristics of two technologies can often be overcome by the expenditure of energy to reduce the emissions of the poorer performing technology. He illustrates the insights that can result from an energy-based life-cycle analysis by comparing the energy requirements of ceramic, plastic, and paper cups.
Allenby and Graedel (this volume) focus on the extensive data requirements of conventional life-cycle analyses. They argue that if data requirements are too great, life-cycle analyses simply will not be performed. They choose instead a qualitative checklist approach and show how it can be used to guide the site selection and design of corporate facilities.
Todd (this volume) notes that Allenby and Graedel's decision-support matrix is the type of tool that many managers lack. Managers may wish to identify
means of pollution prevention, for example, but not have the information they need to identify opportunities and make good decisions. Todd explains that the information needed to make good environmental decisions may be either lacking entirely or unreliable. She proposes a set of guidelines for developing an effective information and measurement system to support environmental decision making.
Plant-Based or Organization-Based Measurements
Raw Flux Measurements
Perhaps the most important raw flux measurement system in operation is the Toxic Release Inventory (TRI). TRI was mandated by the 1986 Emergency Planning and Community Right-to-Know Act. The law requires companies to publicly report releases of each of more than 300 chemicals. Many people believe that the TRI has been remarkably effective in reducing environmental impacts. When companies and people living near them have learned of the quantities of emissions from plants, those companies have often chosen to voluntarily reduce their emissions. The beauty of TRI is that no one is required to act in response to the TRI data, but many have done so anyway. Critics have emphasized that the TRI does not distinguish chemicals on the basis of their relative toxicity and have questioned the accuracy of emissions reports.
3M's Waste Ratio
3M calculates a simple ''waste ratio" to assess the environmental performance of their operations (Zosel, this volume):
(waste)/(waste + products + by products).
The waste ratio uses mass as a common currency that can be followed over time. The advantages of the waste ratio include its simplicity and its limited data requirements. Using mass balances, the waste ratio can be calculated from information on the mass of products, by-products, and wastes. Most of these data are already collected for other purposes, or can be calculated from existing data and process engineering relationships.
The major disadvantage of the waste ratio is that it characterizes waste purely on the basis of mass. A kilogram of nontoxic waste has the same effect on the waste ratio as a kilogram of highly toxic waste. Nevertheless, initial results suggest that the waste ratio can be a valuable tool for improving environmental performance. From the time 3M Corp. introduced the ratio in 1990 through 1995, it reduced wastes 32.5 percent worldwide.
Effluent bioassays measure the survival, growth, or reproduction of aquatic organisms exposed directly to effluents (Goulden). Effluent bioassays are frequently required by government regulation. Their advantage is that they directly measure the response of organisms to whatever mixture of materials is being released to the aquatic environment. Strictly speaking, these bioassays measure environmental performance, but they have properties more characteristic of measures of environmental condition. For example, like most biologically based measures of ecosystem condition, the test organisms respond to the complete suite of effluent components. Thus, with the use of effluent bioassays, humans need not make assumptions about the relative toxicity of different effluent components. However, it can be difficult to determine why the organisms respond as they do or even to what they are responding. Finally, as Goulden argues, the widespread use of effluent bioassays risks a sort of complacency if bioassay results are extrapolated carelessly. As Goulden explains, an effluent could damage an ecosystem even if it does not appear to be toxic in conventional effluent bioassays.
National or Regional Measurements
Natural Resource Accounting
From an environmental perspective, the widely used measure of national economic health, gross national product (GNP), has two important shortcomings: It does not include depreciation charges for depletion of the natural resources that form the basis of production, and it does not incorporate the costs of environmental externalities (environmental consequences of market transactions that are not reflected in market prices). Repetto et al. (this volume) explain that "a country could exhaust its mineral resources, cut down its forests, erode its soils, pollute its aquifers, and hunt its wildlife and fisheries to extinction, but measured income would not be affected as the assets disappeared."
Natural resource accounting methods attempt to overcome the environmental shortcomings of traditional accounting calculations (Daly and Cobb, 1989; Solorzano et al., 1991). These methods ascribe prices to various forms of environmental degradation and then add those values into conventional accounts, thereby providing more comprehensive accounts of the consequences of economic activities. Repetto et al. use a case study of Indonesia to illustrate natural resource accounting. They show that although Indonesia's GNP increased by an average of 7.1 percent per year from 1971 to 1984, the increase drops to an average of 4.0 percent per year when a few major forms of environmental depletion are factored in.
Repetto et al. further describe the insights from natural resource accounting that accrue when the method is applied to particular components of a nation's economy. They show that conventional accounting calculates the net economic
value of a Pennsylvania corn and soybean farm as $75 per acre per year, but that the value drops to only $2 per acre per year when calculated by natural resource accounting methods. Finally, Repetto et al. demonstrate that estimates of industrial productivity growth can be spuriously low if they do not account for emissions reductions. Using natural resource accounting methods, they estimate productivity growth in the electric power industry to be two to three times as high as conventional estimates of productivity growth during the same period.
National Materials Use and Waste Production
Ayres and Ayres (this volume) and Wernick and Ausubel (this volume) focus on aggregate waste generation by the U.S. economy. Ayres and Ayres make detailed calculations to estimate this aggregate waste production. They use a mass balance approach that combines information on the import and extraction of materials in the United States with information on production of goods. They calculate waste generation by subtracting the quantity of materials in goods produced from the materials imported and extracted domestically. This is similar to 3M's system of calculating the mass of products, by-products, and wastes produced in its plants (Zosel), but it is applied to the nation as a whole. Such information can be instructive for a variety of reasons. First, measurements of this type provide a comprehensive baseline by which to measure future progress in waste reduction. Second, they tally the various types of waste production in the various industries. Third, these measurements may help to identify previously overlooked opportunities to reduce waste production because although data are regularly collected on materials extraction and production of products, rates of waste production often become apparent only through deliberate, detailed calculation.
Wernick and Ausubel suggest a variety of metrics that can be used to track changes in national environmental performance. Their metrics measure a variety of features of national materials use and waste production, such as aggregate consumption of materials per capita, ratios of uses of various fuels, consumption of materials per unit of economic production, growth-versus-harvesting ratios for natural resources, and inputs of agricultural chemicals per unit of agricultural production. Cox and Offutt (this volume) describe recent efforts to establish similar metrics for farming and ranching. Such measures should prove useful for tracking national and sectoral progress in waste reduction and identifying opportunities for improvements within various sectors of the economy.
Global Estimates of Environmental Performance: Appropriation of Primary Production
An interesting aggregate measure of global human impacts is Vitousek et al.'s (1986) estimate of the proportion of net primary production1 "appropriated"
by humans. This is one of the few performance measures that not only attempts to quantify total human impacts, but does so in terms that are related directly to the environment's potential for support. Vitousek et al. (1986) estimate that humans appropriate 25 percent of global net primary production and almost 40 percent of terrestrial net primary production. By appropriation, they mean the sum of direct use of plants for food, fuel, fiber, and timber and the reduction in primary production that would otherwise occur through alteration of ecosystems by deforestation, decertification, paving, or other types of conversion to a less productive condition. They conclude that ". . . with current patterns of exploitation, distribution, and consumption, a substantially larger human population—half again its present size or more—could not be supported without co-opting well over half of terrestrial NPP [net primary production]" (p. 373). Regardless of whether this estimate is accurate, human use of resources and the impact of such usage on the life of the planet is clearly massive and growing.
Examples of Ecosystem Condition Metrics
Like performance measures, a diverse collection of methods is now used to assess the condition of environments or their components. These range from simple physical and chemical measurements to measures of the condition of individual organisms and methods for assessing the condition of entire ecological communities. Some conventional ecological measures, such as the biomass or size structure of a population, are useful as measures of ecosystem condition. Others, such as the Index of Biotic Integrity (IBI) (Karr, 1981), were developed for the express purpose of measuring ecosystem condition.
An extensive literature describes the relative merits of various measures for various applications. We do not attempt to review that literature here, but merely describe briefly some important methods. Introductions to the literature are provided by Schindler (1987, 1990) and Cairns et al. (1993).
Physical and Chemical Measures
Environmental conditions have traditionally been assessed with physical or chemical measurements, such as the pH of water or the temperature of air. These measures can be very informative and will surely remain important. However, it is often difficult to predict the ecosystem consequences of a change in physical or chemical conditions. In addition, conventional physical or chemical measures do not always detect important changes (Karr, 1991; Yoder and Rankin, this volume). As a result of these limitations, there is a trend toward reliance on a more balanced combination of physical and chemical measures plus direct biological criteria for assessing ecosystem condition.
Measurements of the Condition of Individual Organisms and Populations
Physiological, Histological, and Demographic Measures
There are a variety of methods for assessing the condition of individual organisms that can provide evidence of environmental degradation. These include measurements of body burdens of various compounds (e.g., polychlorinated biphenyls, mercury), the prevalence of cancers or deformities, and the concentrations of enzymes that are synthesized in response to environmental contaminants. Other types of environmental impacts can be detected from studies of the size or structure of populations. These latter measures can be particularly useful for monitoring the status of harvested populations such as fish or trees.
Like the closely related effluent bioassays described above, ambient bioassays expose test organisms to a stimulus. In effluent bioassays, the stimulus is an effluent. In ambient bioassays, the stimulus is usually water from a polluted or potentially polluted source, such as a river. Like effluent bioassays, ambient bioassays measure the survival, growth, or reproduction of test organisms. Whereas effluent bioassays are used to assess the toxicity of particular effluents, ambient bioassays assess the cumulative toxicity of point and nonpoint sources of pollutants after their dilution by, for example, a body of water. Stewart (this volume) describes the insights yielded by ambient bioassays and the considerations that are necessary when evaluating ambient bioassay data.
Measurement of the Condition of Entire Ecological Communities
Particular species have long been used as "indicators" of ecosystem condition. Indicator species are organisms whose sensitivity to pollution makes them useful as a tool for detecting polluted sites. The concept is useful, but reliance on a particular species makes for a crude measure; the indicator species is either present or absent. Absence is not necessarily a result of local conditions; the species may never have had the opportunity to colonize the site. In addition, particular indicator species may help detect particular environmental impacts, but they are not likely to be suitable for detecting a wide range of impacts.
Information on the presence or absence of indicator species has frequently been supplemented with various basic ecological measurements such as species richness (e.g., the number of fish species in a particular section of a river), the abundance of various organisms, or more formal ecological measures of species diversity. Most of these measures assume that analysts have good information on the characteristics of relatively undisturbed reference sites in the region of inter-
est. Strang and Sage demonstrate how such measurements have helped to document improvements in the conditions of the rivers they studied.
Index of Biotic Integrity
Recently, the concepts of indicator species and species diversity have been elaborated with the development of so-called multimetric biotic indices. These measures assess the overall condition of an ecosystem through simultaneous use of a variety of metrics. One such index is Karr's (1981) IBI. Karr (1992) argues that the multivariate nature of natural systems dictates that effective measures of ecosystem condition be based upon a variety of relevant biological attributes but that, to be usable, a comprehensive measure cannot require data for an endless number of system properties. The IBI "represents a synthesis of a dozen distinct hypotheses about the relationship between attributes of biological systems under varying influence from human society" (Karr, 1992, p. 233). The IBI is now widely used in North America and Europe to assess the condition of stream fish communities. Karr's original IBI has been modified to apply the same approach to other organisms, such as stream invertebrates. Carriker (this volume) and Yoder and Rankin apply Karr's approach to assess the conditions of streams and reservoirs managed by the Tennessee Valley Authority and the Ohio Environmental Protection Agency.
Any efforts to limit human environmental impacts should include two goals. The first should be to minimize the undesirable environmental impact per unit of human activity. The second should be to ensure that the cumulative impact of all human activities is compatible with the persistence of all critical ecosystem conditions and processes. Profound uncertainties will complicate efforts to achieve these goals, but they are appropriate targets.
Environmental performance measures are key tools in identifying opportunities to move toward the first of these goals. Measures of ecosystem condition are vital in charting progress toward the second goal. However, these two sets of tools will be most useful if they can be used to accurately predict the consequences of human activity on affected ecosystems. To achieve this, they must be refined and coordinated such that predictions can be based on the product of the environmental impact per unit of an activity multiplied by the scale of that activity. Such information would be useful for distinguishing acceptable and unacceptable impacts. Users of performance and condition measures should examine the potential for finding or developing relationships between the two types of indicators. Ideally, it will eventually become possible to measure environmental performance in terms that can be related directly to the consequences for affected ecosystems. In essence then, the environmental impact of a given activity could
be expressed in units of ecosystem condition. We hope the papers in this volume will foster progress toward that ideal.
Cairns, J., Jr., P. V. McCormick, and B. R. Niederlehner. 1993. A proposed framework for developing indicators of ecosystem health. Hydrobiologia 263:1-44.
Daly, H. E., and J. B. Cobb, Jr. 1989. For the Common Good. Boston: Beacon Press.
Hart, D. D. 1994. Building a stronger partnership between ecological research and biological monitoring. Journal of the North American Benthological Society 13:110-116.
Karr, J. R. 1981. Assessment of biotic integrity using fish communities. Fisheries 6:21-27.
Karr, J. R. 1991. Biological integrity: A long-neglected aspect of water resource management. Ecological Applications 1:66-84.
Karr, J. R. 1992. Ecological integrity: Protecting Earth's life support systems. Pp. 223-238 in Ecosystem Health: New Goals for Environmental Management, R. Costanza et al., eds. Washington, D.C.: Island Press.
Lindstedt-Siva, J., L. C. Soileau IV, D. W. Chamberlain, and M. L. Wouch. 1996. Engineering for development in environmentally sensitive areas: Oil operations in a rain forest. Pp. 141-162 in Engineering within Ecological Constraints, P. C. Schulze, ed. Washington, D.C.: National Academy Press.
Schindler, D. W. 1987. Detecting ecosystem responses to anthropogenic stress. Canadian Journal of Fisheries and Aquatic Sciences 44(suppl. 1):6-25.
Schindler, D. W. 1990. Experimental perturbations of whole lakes as tests of hypotheses concerning ecosystem structure and function. Oikos 57:25-41.
Sharpe, R. M. 1995. Another DDT connection. Nature 375:538-539.
Shen, H. W. 1996. Engineering studies based on ecological criteria. Pp. 177-186 in Engineering within Ecological Constraints, P. C. Schulze, ed. Washington, D.C.: National Academy Press.
Solorzano, R., R. deCamino, R. Woodward, J. Tosi, V. Watson, A. Vasquez, C. Villalobos, J. Jimenez, R. Repetto, and W. Cruz. 1991. Accounts Overdue: Natural Resource Depreciation in Costa Rica. Washington, D.C.: World Resources Institute.
Steen, B., and S. O. Ryding. 1992. The EPS Enviro-Accounting Method: An Application of Environmental Accounting Principles for Evaluation and Valuation of Environmental Impact in Production Design. IVL Report B 1080. Göteborg, Sweden: Swedish Environmental Research Institute.
Vitousek, P. M., P. R. Ehrlich, A. H. Ehrlich, and P. A. Matson. 1986. Human appropriation of the products of photosynthesis. BioScience 34:368-373.