Ecosystem vulnerability: A complex interface with technical components

This special edition of Environmental Toxicology and Chemistry contains technical papers from a symposium, Modeling and Measuring the Vulnerability of Ecosystems at Regional Scales for Use in Ecological Risk Assessment and Risk Management, which was held August 17–20, 1998, in Seattle, Washington, USA. This symposium was organized and co-sponsored by the U.S. Environmental Protection Agency, SE-TAC, and the ASTM's Committee E47 on Biological Effects and Chemical Fate to discuss problems and explore potential solutions to integration of data and models across scales of time and space, levels of organization, and multiple stressors. It provided an important venue for fostering communication across interdisciplinary fields, especially between the social sciences and the natural sciences. The ultimate aim was to improve ecological risk assessment and management practices. The hope was that, from the technical presentations and the dialogues generated by hearing the perspectives of researchers, regulators, and various public stakeholders, we could determine the proper tools and approaches for identifying ecosystems at risk to man-made and natural stressors and guide the development of research strategies that will help us preserve or restore those systems. Problems encountered in addressing vulnerability and management of ecological systems are related to defining appropriate questions and experimental designs for different assessment scales because data are often at multiple levels of biological organization or at single subecosystem levels at discrete regions within ecosystems. Data at the ecosystem level, by nature of the scale, are often aggregated and tend to have less detail, which is problematic when evaluating impacts or exposures. Frequently, models are used, but they tend to address specific factors such as exposure versus effects, water versus land, and local versus regional geography. The organizing committee sought interested experts to share their findings and experiences and to help identify both the state of the science and the information gaps that hinder progress in making effective tools and guidance available to the various communities of potential users. The committee hoped through the process to address monitoring research focused on development of indicators at various levels of biological organization for multiscale ecosystem evaluation and the development of common modeling frameworks and calibration of models to improve the integration of those models. Research reported in the symposium focused on novel approaches to integrating landscape, exposure, and effects and risk management and risk communication tools applicable to watershed and larger scale pollution issues. The traditional technical focus of Environmental Toxicology and Chemistry precluded inclusion of several of the social science- and decision science-oriented presentations. In many ways, this is unfortunate because it is precisely through enlightened dialogue across these disciplines that major advances in environmental management are likely to occur. Nevertheless, the papers published in this issue report important scientific and technological developments in the assessment of environmental systems and present several of the themes of the session. Three papers in this issue represent the theme of model integration. Eloy addresses long-range transboundary effects of air pollutants on forest ecosystems in Europe. The Loux paper models diel temperature effects on the exchange of mercury between the atmosphere and underlying waters. Morton et al. discuss methods for modeling cumulative episodic exposure to chemical stressors in the conduct of aquatic risk assessments. Three more papers discuss aspects of scale integration. Specifically, Dyer et al. explain approaches to assessing multiple stressors over large geographic areas, Hutchinson and McIntosh present a case study of integrated risk assessments in New Zealand, and Lassiter et al. address the vulnerability of Mid-Atlantic ecosystems to climate change. The ecological risk assessment theme is represented by five papers. Gordon and Majumber discuss empirical stressor–re-sponse relationships for prospective risk analysis based on their work in the Eastern Cornbelt. The paper by Cormier et al. (pp. 1120–1126) presents approaches for estimating exposure criteria values for PAH metabolite concentrations in fish. Jones and Gordon discuss how watershed-based strategies can be implemented into local, state, and national policy. Wen-te explains a proximity-based measure of land-use impacts. Birge et al. test the hypothesis that metal body burden in selected sentinel organisms can be used as an indicator of ecological integrity and allows extrapolation from site-specific to regional systems over time. The indicators and linkages among indicators themes are represented by three papers in this issue. Norton et al. discuss the use of biological assessments to discriminate among types of stress to ecosystems in the Eastern Cornbelt ecoregion. Landis and McLaughlin highlight those indicators that predict the dynamics of ecological systems and demonstrate their utility in ecological risk assessment. Culp and Lowell describe their approach to using indicators from several trophic levels, in concert with other measurements, in the conduct of retrospective risk assessments on large Canadian rivers. Five papers in this issue represent the case studies session and theme. Three of these, Cormier et al. (pp. 1082–1096), Schubauer-Berigan et al. (pp. 1097–1105), and Cormier et al. (pp. 1127–1135), present complementary pieces of the planning and conduct of an integrated risk assessment of the Big Darby Creek (Ohio, USA) area and the potential application of the study findings to other Midwestern watersheds. Carpenter and Lunetta present stressor profiles developed from interpolation of monitoring network point data, modeling projections, and the processing of continuous data sets. The paper by Detenbeck et al. tests both geographically dependent and independent watershed classification schemes for ecological risk assessment. The U.S. Environmental Protection Agency Office of Research and Development's new ecological research strategy aligning Agency research more closely with the risk assessment paradigm of the National Academy of Sciences appears in a paper by Linthurst et al. Public lands. In his keynote address, Dr. William Brown, Science Advisor to Bruce Babbitt, Secretary of the Department of the Interior, focused on the inventory and monitoring of biological resources on public lands, commenting that such efforts are much discussed but little coordinated and seldom put to effective use. The Interior Department has launched an effort to move forward with an initiative involving lands of the Bureau of Land Management, the Fish and Wildlife Service, and the National Park Service, which collectively manage lands covering about 20% of the United States. The purposes of the effort are to inform the public on the condition and status of these lands and to secure the information needed to implement the laws for which these agencies are responsible. The U.S. Geological Survey is coordinating the effort, which is expected to begin with a pilot project in fiscal year 1999. The initial categories identified for measurement are landscape/community, riparian systems, threatened and endangered species, invasive species, and amphibians. Historical perspective. Mr. Henry Lickers, Director of the Department of the Environment of the Mohawk Council of Akwesasne, told the story of how, when just a young biologist, he was called upon by the Council to investigate all the impacts caused by construction of the St. Lawrence Seaway on the people of Akwesasne. He proudly turned in his report, only to find, on further research, that all of his findings had been documented more than a century before through the knowledge of the fishermen, trappers, and farmers of his community. The lesson? Do not ignore the knowledge that is readily available from people who do not consider themselves scientists but who observe nature and its changes in their day-to-day lives and who represent a valuable, but largely untapped, resource. Mr. Lickers also described the changes that occurred in the area with the displacement and loss of the people who had utilized the land according to a spiritual and political system that strove for harmony among the peoples and the land. The “new people” didn't view the Earth as their mother or the land as part of their family, so the acclaimed engineering marvel of the St. Lawrence, coupled with other exploitation of the lands, seriously hurt the Mohawk people who depended on the River. The Department of the Environment of the Mohawk Council of Akwesasne has developed a way of looking at community, which they call Naturalized Knowledge Systems, to acquire knowledge about the place where they live. This approach can use new tools and adapt while still maintaining the traditional principles of a people. It is represented by six basic principles: The Earth is Our Mother, cooperation is the way to survive, knowledge is powerful only if it is shared, the spiritual world is not distant from Earth, responsibility is the best practice, and everything is connected to everything. The successful implementation of the Naturalized Knowledge System relies on a balance of respect, equity, and empowerment. Scale. Dr. James MacMahon of Utah State University, past president of the Ecological Society of America, described how various elements of scale become important when one attempts to aggregate data on the condition of ecosystems at the watershed or regional level or to evaluate the potential impacts of stressors to such systems. Indeed, the appropriate scale for many studies depends as much on the question being asked as on the system being studied. He reminded us that, to understand ecological systems, one must look to the level of biological organization below the unit of study for the important processes and at the level above for context. Dr. MacMahon also cautioned that average values of biological variables in studies across spatial scales often are misleading and that scaling up from plot studies to make generalizations is often difficult (and dangerous) because of the spatial and temporal variation in systems. He pointed out that studies of a system over long time scales may, in fact, be studies of fundamentally different systems because so much changes with time in nature. Finally, he warned that, even if we are successful in creating a proper structure for an ecological system, it does not guarantee that normal functions will exist. Focus and definition. Dr. Kenneth Dickson, Director of the Institute of Applied Sciences at the University of North Texas, had been charged by the organizing committee to challenge the meeting participants to focus on the state-of-the-science/art of measuring the vulnerability of ecosystems to stressors at various scales. He observed that few of the scheduled presentations were at the regional scale—the proposed focus of the symposium—probably because the historical focus of assessments has been at the site or watershed levels. He challenged the group to focus their thinking on stressors (chemical, physical, and biological) that operate at the regional scale and on the identification of ecological endpoints that are appropriate at that scale. Dr. Dickson also urged participants to develop a consensus definition of vulnerability, as it was evident that there were many different definitions among the presenters. This lack of agreement could serve as a major impediment to identifying gaps in understanding and data. In subsequent discussions at the symposium, the following definitions emerged. (1) System vulnerability is a characteristic of an entity or system to respond to stressors over time and space. It is estimated by an analysis of characteristics affecting exposure and sensitivity of receptors (endpoints) to stressors. It is a function of receptor sensitivity and environmental factors (which include biological, physical, and chemical). (2) Ecosystem vulnerability is the potential of an ecosystem to modulate its response to stressors over time and space, where that potential is determined by characteristics of an ecosystem that include many levels of organization, such as a soil, a bioregion, a tissue, a species, an organism, a stream reach. It is an estimate of the inability of an ecosystem to tolerate stressors over time and space. Balance. Dr. Elaine Dorward-King of Rio Tinto Technology in the United Kingdom described the factors that must be considered in balancing natural resource development and ecosystem protection. Natural resource managers, in both industry and government, face land use and environmental management decisions that require not only scientific, but economic, social, cultural, and political contributions. Sustainable environmental management decisions require incorporation of all these types of information while involving all people with legitimate interests early in project definition and data collection. While ecosystems are vulnerable to numerous threats, including pollution from current or past anthropogenic activities, the overwhelming threat is from human population density and associated uncontrolled development, leading to depletion of natural resources (water, soil, biota) and degradation or loss of wildlife-supporting habitats. The predominant considerations in many parts of the world for environmental management decisions and ecosystem protection include understanding and establishing land and resource use priorities, establishing time frames for management, using comparative valuation of ecosystems and comparative risk assessment, and clarifying where decision authority resides (local, regional, national). It is a challenge for scientists to include all these considerations and communicate effectively with the nontechnical public or decision makers. It is, however, our responsibility to interpret and communicate data/results/conclusions in a manner that is understandable and useful. Also, scientists must communicate and involve the public early and not presume they understand the social and cultural parameters of a new environment. The opportunity to intervene to protect vulnerable ecosystems and to influence the management of others is an immediate one and must not be squandered. Four different ecosystems vulnerable to degradation or change, ranging from an arid, high-altitude saline lake system to a tropical rainforest, were discussed in the context of the management decisions that must be made and the strategies that are being developed to provide the necessary technical and socioeconomic information. Communication and Management. Walt Galloway of the U.S. Environmental Protection Agency's National Health and Environmental Effects Laboratory summed up research issues involved in the vision of an ideal future for risk communication and management, which are (1) assuring stakeholders an early and frequent voice in value-driven, place-based decision making, (2) driving responsibility down to the lowest level, (3) providing a transparent and jargon-free decision-making process, (4) defining what or who actually constitutes the community and the decision maker, and (5) identifying linkages among communities. One suggestion was the application of large-group, collaborative, decision-making processes to community-based efforts to better achieve the goals of regional assessments. The participants recognized the strong correlation between the consensus recommendations of their work groups and the six basic principles of Naturalized Knowledge Systems cited earlier. A workshop subsequent to the symposium identified a number of major, cross-cutting issues and need areas and many specific issues within each. The major issues/needs included community management (local, community based), improved databases and analysis, stressor/response relationships, modeling, issues of scale (spatial and temporal), uncertainty in ecological risk assessment, and the concept of vulnerability. Community. Community management must recognize that community extends as far as the most distant stressor source. The local human community is good at identifying local scale and some broader impacts and should be involved in discussions of the issues of concern and their scale. They must educate (and be educated) as part of community involvement using jargon-free dialogue and must learn how to articulate their concerns. While humans define the values by which we determine ecosystem health goals, we must recognize that ecosystem health values go beyond merely human values and that humans are both stressors and receptors in a community and that, with their biases, define the values by which we determine ecosystem use goals. Finally, the community must realize that action may need to occur at a different scale than the problem recognition. Databases and analysis. Issues related to databases and analysis focused on lack of complete data on geographic areas and the need for better stressor information (beyond first-order approximations) that allow linkages of physical, chemical, and biological (exposure and effects) data over landscapes. Data from only a few long-term monitoring programs are available; what is there is rarely balanced in time and space. Analysis should be focused on better integrating data from different levels, scales, and processes and standardizing, on a single scale, data for soil mapping to promote interregional comparisons. The nation as a whole must commit to collecting time-series, remotely sensed data on a national scale. Stressor/re-sponse relationships, as well as stressor/vulnerability relationships, must be considered when assessing ecosystem vulnerability, and the roles of indicator measures as they relate to describing exposure, response, and vulnerability must be defined. Stressor/response curves are needed on nonchemical stressors, with more biological focus, e.g., invasive species. Better understanding of interactions and modulating factors across scales and levels of organization is necessary. To prioritize management action, we must develop the ability to disaggregate multiple stressors, interactions, and modulating factors to prioritize for management action. Models. The inability to experiment at landscape levels may force a shift to rely only on models. Models are needed that relate stress and variation over space (and levels of organization) and incorporate spatial information and address multiple stressors, media, modulating factors, and receptors. Special care must be taken with complex models, however, as they require better data and design parameters (focused with sensitivity analysis). Scale. This term should be used with care. For example, to cartologists, larger scale means closer in or greater resolution. The scale of the assessment must match (+ or — one level) the scale of the system analyzed. Finally, and this leads to the next topic, scale affects uncertainty, which in turn determines our ability to make is a need to estimate uncertainty associated with and e.g., of on data greater than The nature of the processes to be using indicators need to report the and uncertainty of indicators should be The of extrapolation need to be estimated and should be used Finally, the of from multiple must be and defined. the among groups to manage environmental for perspectives or enlightened human the dialogue at this symposium was It is our hope that the papers in this technical issue will greater among scientists to their interests to the broader in vulnerability and systems this special involved many who their time. The and to to the than provided of papers. The Committee Dickson, and were especially in the process The of the were long of ecological planning and a of the To all a