Aquatic Ecosystems: Defined by Hydrology. Holistic Approaches Required for Understanding, Utilizing, and Protecting Freshwater Resources
Diane M. McKnight Institute of Arctic and Alpine Research University of Colorado, Boulder
The scientific disciplines of hydrology and limnology are distinct though closely connected. Limnology is an integrative discipline, applying physics, chemistry, and biology to the study of inland aquatic ecosystems. Inland aquatic ecosystems include streams, rivers, lakes, reservoirs, and wetlands and possess many diverse characteristics. The focus of this paper is how hydrology defines aquatic ecosystems, especially the ecosystem boundaries and the fluxes of water, solutes, organisms, and detrital organic matter across boundaries. For millennia, human civilizations have used knowledge of hydraulics and hydrology to distribute water resources for agricultural, municipal, recreational, and power generation purposes. In contrast, the basic concepts of ecology have been established only during this century, and their application to the management of water resources, through watershed management or ecosystem management approaches, is just now coming into practice. Undoubtedly, freshwater is a strategic resource and advances in hydrology, limnology, and water resource research and management will be required in the future. Broader education of all scientists and engineers involved in the study and management of freshwater is one way to promote greater knowledge and expertise and to achieve more complete, holistic resolutions of water resource issues. Further, it is critical to maintain and expand networks of monitoring programs that provide long-term records of physical and chemical characteristics of inland aquatic ecosystems. These records allow us to extend current understanding of aquatic ecosystems to the appropriate hydrologic time scales and will allow for resolution of water resource issues through a process of "knowledge-based" consensus at local and regional scales.
Freshwater is a strategic resource upon which human society depends. Water is consumed for drinking water, industrial uses, and irrigation, and flow regulation is instrumental to power generation and flood protection. Aquatic ecosystems provide fish and other food resources as well as recreational resources, which are highly valued by communities. All of these uses depend not only upon having a sufficient quantity of water but also on the quality of water, that is, the physical and chemical characteristics of water. As we gain a greater scientific understanding of aquatic ecosystems, it is apparent that the quality of aquatic ecosystems influences not only the resident aquatic biota but also the quality of a water body as a water resource. These interactions can be represented as healthy aquatic ecosystems providing "goods and services" to the other users of water resources, as shown in Figure 1.
In many areas of the world and in some regions of the United States, water is now a limited strategic resource. Over the next 25 to 50 years, constraints on the
expansion of human societies and on economic development associated with limitations in water quantity and quality are expected to become more severe. There will not be magic bullets, quick fixes, or easy answers to the water resource challenges that lie ahead. Restructuring, optimization, and otherwise fine-tuning of existing water resource systems to achieve multiple goals will be required, as well as increasing the efficiency of irrigation and other consumptive uses (Postel, 1997). In the United States, federal, state, and local governments provide an infrastructure to put into place changes at a scale commensurate with that of large river basins. Some steps have been made in this direction, such as the recent landmark agreement that regulates flow in the Colorado River to achieve power generation and ecological goals. The governmental infrastructure and the commitment to multiple goals are important assets but are, in and of themselves, insufficient to meet the water resource challenges of the future. Clearly, these assets must be matched with greater knowledge in all aspects of aquatic sciences and engineering.
We need not only more in-depth understanding of hydrologic, chemical, biological, and ecological processes but also more detailed characterization of local and regional hydrology and aquatic ecology. Both predictive understanding of processes and knowledge of detailed characteristics are needed for effective fine-tuning at the scale of an individual reservoir, river, lake, wetland, or ground water aquifer. Management of water resources has been and will continue to be a governmental function. In some regions, private nongovernmental bodies, such as the Hudson River Foundation, also are playing an active role in research and management. Given that greater scientific knowledge will continue to promote progress in water resources and other civic issues, we can make the case that now is the time for federal, state, and local governments to invest in aquatic science. In response to the argument that a greater investment in research and training in aquatic science is not affordable within the current constrained governmental budgets, it can only be pointed out that the consequences of lack of knowledge and scientifically ill-founded or unsustainable practices are potentially disastrous as demands for water resources continue to grow. Simply put, greater knowledge of aquatic science will provide for the future, while lack of knowledge may have dire consequences.
Together, the disciplines of hydrology and limnology comprise a healthy portion of aquatic science. In this paper, hydrology is intended to include the study of the hydrologic cycle and physical processes controlling the movement of water in surface and ground waters. Limnology is broadly defined as an integrative discipline applying physics, chemistry, and biology to study inland aquatic ecosystems, which include streams, rivers, lakes, reservoirs, and wetlands. In order to meet the water resource challenges of the future, investments should be made now to advance these two disciplines and to provide comprehensive training of scientists and engineers in these disciplines.
Hydrology and limnology are intimately related. Despite their different histories and etiologies, there are some common approaches that can facilitate
interaction and interdisciplinary research. In fact, the most exciting areas of research currently include those based upon integration of hydrologic and ecological approaches. This paper will conclude with recommendations for revitalizing education in limnology to promote greater knowledge of hydrologic processes among future limnologists. Similar recommendations are presented for incorporating limnology and ecology into the training of future hydrologists and water resource engineers.
Different Histories of Hydrology and Limnology
For millennia the progress of human societies has been tied to the development and use of water resources. Developing cities and agrarian communities put to good use the basic idea that water flows downhill, rainfall to river to ocean. Engineering know-how preceded the quantitative sciences of hydraulics and hydrology by many centuries. The skills to route water through channels, aqueducts, and pipes were well developed in Roman times. The plumbing systems of the Roman baths are masterpieces of civil engineering, one of which can be observed in an operational state in the museum in Bath, England (Cunliffe, 1993). The large reservoir, sluices, and lead pipe drain were constructed on the frontier of the Roman Empire beginning in the first century A.D., 1,500 years before the equations governing flow in pipes were developed by Bernoulli. In the new world, the Anasazi, the ancient Pueblo people of the southwest, moved water by sluices and man-made canals around 1000 A.D. Thus, hydraulics and hydrology were derived from physics and geology but were developed with an extensive foundation of empirical knowledge from observation and engineering experience. Further, new conceptual advances were rapidly employed as water resource systems were developed and refined.
The most remarkable changes to aquatic ecosystems, as a result of human activity, came with the development of agriculture. These alterations have accelerated in recent centuries and were generally not intended to improve water quality or the health of an aquatic ecosystem. In this century the motivation for extensively studying aquatic ecosystems stems from problems associated with this inadvertent degradation. Issues that have received considerable attention over the past 50 years include the discharge of nutrients from wastewater treatment plants, leading to eutrophication, and the atmospheric release of sulfur and nitrogen oxides during fossil fuel combustion, leading to acid rain and acidification of water bodies far removed from the sources of pollution.
In contrast to hydrology, limnology is based upon a concept that is less than a century old. In 1991 Science (Pool, 1991) listed the top 20 most significant scientific concepts of our century. The central theme of ecology, that "all life is connected," was listed with three other biological concepts (all organisms are made of cells, life is based on a genetic code, life evolves through natural selection). Continental drift, during which the Earth's surface is in continual change,
was also identified as a new significant concept. Together these concepts have radically changed how we view and understand the natural world. The term ecosystem was introduced by Alfred George Tansley in 1935 and was presented as a basic unit of nature in the continuum from the atom to the universe (Golley, 1993). Although the ecosystem concept is to some extent based upon the descriptive information provided by naturalists and on the work of community ecologists, it represents an integration of biology with the more quantitative sciences of physics and chemistry (Mayr, 1982).
Although limnology is often viewed as a subset of ecology, the two fields developed in parallel (Golley, 1993). While the ecosystem concept is credited to Tansley, limnologists had previously voiced similar ideas that coupled the biota to their physical environment. Forbes' description of "the lake as a microcosm" in 1892 was further developed by the work of Birge and Juday. Lindeman was the first to explicitly apply Tansley's ecosystem concept in his classic work on Cedar Bog Lake. From his study of a small lake, he concluded that the ecosystem
was the fundamental unit of trophic-dynamics, during which energy is transferred from primary producers (photosynthesizing algae and plants) to heterotrophs (grazers and predators). Figure 2 presents a diagram from his paper in American Midland Naturalist illustrating this concept.
The next major advance in limnology was the application of the small watershed approach to quantitatively understand the relationships between biological, chemical, and hydrologic processes in a watershed. This work began in 1963 with the Hubbard Brook Ecosystem Study, a collaborative effort between ecologists, hydrologists, and geochemists. The study quantified the input and output of chemicals to and from the watershed in constructing a nutrient budget for a forest ecosystem in New England (Likens et al., 1977). It has served as a basis for understanding management practices for New England forests and the effects of acid rain in the region. The streamflow and water quality records for Hubbard Brook are the primary data sets for this ongoing study.
In 1980 Vannote et al. presented a theoretical concept describing how ecological characteristics would vary along a ''continuum'' from a headwater stream to an intermediate-order stream to a large river. This concept, called the River Continuum Concept (RCC), linked the hydrologic and geomorphic processes controlling networks of streams and rivers to processes controlling basic ecosystem parameters such as the ratio of primary production (photosynthetic production) to respiration. A schematic diagram presenting the RCC is presented in Figure 3. This concept has provided a unifying theme for understanding streams and rivers, and new hypotheses have enriched this basic concept. The applicability of the RCC has now been evaluated at a global scale.
The watershed approach has been implemented in many areas of the United States and around the world. In addition to organizations conducting research, the Western Water Policy Review Advisory Commission estimates that there are currently 550 watershed stakeholders' groups in the western United States alone. The RCC was used in the design of the National Water Quality Assessment (NAWQA) program of the U.S. Geological Survey, and the scale of the river basins studied in NAWQA matches that of the RCC. Currently, "ecology" and "watershed" are household words, and the ecosystem concept is taught to school-children, indicative of the impact and relevance of the discipline. If, in the future, we continue to apply other areas of limnology to freshwater issues, the term limnology may also become widely recognized.
For a given inland aquatic ecosystem, our understanding of hydrologic and ecological processes varies greatly. Ground water provides a striking example. The science of ground water hydrology is well established, with a broad knowledge base, including detailed physical models that allow quantitative study of ground water flow systems (Freeze and Cherry, 1979). On the other hand, the science of ground water ecology has made great advances in the past 20 years but is still in a "fledgling" state. Much remains to be learned in order to apply ground water ecology to the resolution of ground water contamination problems. The
current interest in introducing designer microbes to carry out in situ degradation of organic contaminants in ground water illustrates the pressing need for understanding ground water ecosystems.
In the introductory chapter to their book River and Stream Ecosystems , Cummins et al. (1995) point out that:
A hallmark of flowing-water studies during the 1980s and 1990s has been their interdisciplinary nature. The interactions involve biological stream ecologists, hydrologists, geomorphologists, microbiologists, and terrestrial plant ecologists, all interacting to develop generalizations concerning riverine ecosystems.
As a result of the integrative nature of limnology, it may be true that aquatic ecologists have a greater familiarity with hydrology than vice versa. Another reason for this disparity may be that the time scales over which hydrology controls aquatic ecosystems are so long that these controls are not observed by hydrologists during short-term field studies. However, now that dams have been built and hydrologists are involved in studies of geomorphic and ecological changes associated with the modified flow regimes and lowered water tables, this situation is likely to change. The recent experimental flood in the Grand Canyon is an exciting example of collaboration between hydrologists and ecologists that has led to better understanding of these long-term interactions (Collier et al., 1997).
Definition of an Aquatic Ecosystem
A new term has emerged to describe a more holistic approach toward management of natural resources—ecosystem management. This term implies a more comprehensive approach than those of the past, which appear piecemeal in retrospect. One of the fundamental concepts of ecology is the ecosystem, broadly defined as an assemblage of species and the dynamic physical environment that they inhabit (Golley, 1993). The ecosystem concept is most useful because the boundaries of the ecosystem (which encompass the control volume) are flexible and can be drawn to address the specific question being asked (Odum, 1953). Thus, an aquatic ecosystem can be defined as a drop of water on a leaf, a small puddle in a forest, a shallow aquifer, or the entire drainage basin of the Mississippi River. There is no right or wrong dimension to an aquatic ecosystem; the specification of the boundaries is only judged relative to the scientific question or the resource issue at hand. If the question is the loss or gain of solutes as rainwater moves through a forest canopy, then defining the ecosystem as the water pooled on a leaf may be useful (Sollins et al., 1980). If the question at hand is the management of flooding and water quality issues in the Mississippi River, consideration of the entire basin may be necessary (Bayley, 1995).
The boundaries serve to distinguish between the flux of organisms and constituents into and out of the ecosystem and interactions and feedback processes
occurring within the ecosystem. Materials produced within the ecosystem are "autochthonous," and materials entering from outside the ecosystem are "allochthonous." In deciding how to set the boundaries, a rule of thumb is that once an organism or constituent has exited across a boundary it no longer influences processes occurring within the ecosystem. This is a ''gone is gone" criterion, corresponding to organisms or constituents having dripped off the leaf or having been discharged into the Gulf of Mexico. The discrete nature of a lake or an upland stream is conducive to applying ecosystem boundaries, and the directionality of flow is another aspect of aquatic ecosystems that can make them straightforward to define compared to terrestrial ecosystems.
However, the flow-through nature of aquatic ecosystems can present conceptual challenges. To understand these challenges, the temporal dimension of an aquatic ecosystem must be considered as well as the physical dimensions. The hydrologic time scale is determined by the movement of water through the control volume, represented by the residence time. There are also time scales associated with other changes in the physical environment, such as the solar cycle (which drives photosynthesis, evapotranspiration, and other biogeochemical processes). For chemical and biological processes, time scales may vary by orders of magnitude. The time scales of chemical processes depend upon the properties of the major solutes, mineral, and gas phases, as well as the kinetics of important chemical reactions. Within the diverse biota of an aquatic ecosystem, biological time scales range from doubling times of less than days for bacteria and algae to the much longer, more structured life histories of the aquatic plants, invertebrates, and fish. In general, the hydrologic residence time of water, which controls the transport of solutes, organisms, and suspended particulates, may be short relative to the time scales of dominant ecosystem interactions between organisms or between organisms and their environment. As a result, in many aquatic ecosystems the flux of water, material, and organisms moving through the control volume in a unit of time characteristic of important processes may be large relative to the quantity contained within the control volume.
Another challenge is that in inland aquatic ecosystems the important interactions often involve stationary organisms and moving chemical constituents. In streams, for example, algae growing on rocks (periphyton) take up dissolved nutrients from the overlying water and some benthic invertebrates living on the streambed feed on suspended organic material in the overlying water. Similarly, in ground waters biogeochemcial conditions are controlled by bacteria attached to aquifer materials as biofilms. Because of the interactions between moving and stationary components of the ecosystem, the problem of high water flux through the control volume of an aquatic ecosystem cannot be solved by using a frame of reference that moves with the flow. The interactions between flow regime, stream habitat, and habitat requirements for important aquatic species are central to implementation of ecosystem management approaches for streams and rivers (Richter, 1995).
Examples of Hydrology Controlling Aquatic Ecosystems
There have been exciting conceptual advances in aquatic ecology in the past two decades. Many of these advances have been made by quantitatively addressing the hydrologic interactions occurring at ecosystem boundaries (ecotones). These advances would not have occurred without the active participation of hydrologists. In the following sections, examples from current limnological research are presented that illustrate the role of hydrology in defining aquatic ecosystems. The examples begin at the largest scale, that of large river systems, and continue to finer scales, such as the transitional zones in ground waters.
The Comprehensive River Continuum Concept
It is beyond the scope of this paper to discuss in detail the important ideas that make up the more comprehensive River Continuum Concept. In general, these additions have refined our understanding of the hydrologic linkages occurring (1) in the water column and streambed sediments (such as nutrient spiraling and the hydraulic food chain model) (Power et al., 1995), (2) within the flood-plain (flood pulse hypothesis), and (3) within the hyporheic zone (the underlying substrate near the stream where water is also flowing in the downstream direction). Only the nutrient spiraling concept and the flood pulse concept will be described to illustrate critical hydrologic interactions.
Nutrient spiraling is a concept that integrates the flow-through aspect of streams with the important biogeochemical role of the substrate and organisms attached to the streambed (Newbold et al., 1981). A given molecule present as a solute in flowing water is converted to a particulate form as it is taken up by streambed biota. Molecules thus taken up are then released from the same biota via excretion or death and returned as solutes to the flowing water, resulting in a nutrient cycle or spiral. For specific molecules of interest, the rate of biological processes in the streambed is compared to the hydrologic transport processes to determine the overall rate of nutrient cycling. The nutrient spiraling concept allows one to assess how stream ecosystems will respond to different flow and substrate conditions (see Figure 4).
In the flood pulse hypothesis the seasonal advance and retreat of water on the floodplain is seen as an important environmental feature to which the biota have adapted (see Figure 5) This hypothesis emphasizes the connection of the stream ecosystem to its floodplain, and implicates flooding as a major control on aquatic biota such as fish. Aquatic ecosystems in the United States have experienced considerable change as a result of flow regulation and other hydrological alterations in rivers. Controls are so pervasive in many rivers that natural, wild rivers are few and far between. In the "Freshwater Imperative" research agenda (Naiman et al., 1995), the topic of understanding "modified hydrologic regimes" was assigned a high priority. The recent successful experimental flood in the
goals. Such approaches can be used across the country when modifying flow regimes on regulated rivers to gain healthier stream communities.
Lake Ecosystems and Climate Change
Temperate lakes respond to climatic forcing. Thus, research on lake ecosystems should be included in research on human-accelerated environmental change. Current studies have considered both how lakes are changing in response to the changing climatic conditions of the last century and how lake ecosystems may change if there is a directional shift in the climate driven by the buildup of carbon dioxide or its radiative equivalent in the atmosphere. Both of these considerations have generally led to new insights into the functioning of lake ecosystems, especially in the context of wintertime processes for which fewer field measurements of dynamics have been carried out.
Circulation and stratification of physical properties in lakes are controlled by temperature, wind speed, and the clarity of the water. These characteristics in turn can influence the distribution of organisms and ecological interactions. Fish populations provide one example of how temperature regimes in the lake can define the habitat for species of fish with different thermal tolerances. DeStasio et al. (1996) used climate predictions from a global circulation model to provide climate parameters mimicking a doubled carbon dioxide atmosphere for a set of four lakes in Wisconsin. They predicted the new summer "niche volumes" for fish species tolerating cold, cool, and warm water temperatures. Because of increased onset of stratification, increased epitimnetic summer temperatures, and a longer duration of stratification, suitable thermal habitats were more abundant for all fish types and all climate change scenarios considered, as shown in Figure 6. The modeling exercise also showed that surface water temperatures may exceed upper lethal limits for some warm and cool water fish. The timing of ice cover is another climate-driven parameter that strongly influences the annual ecological cycle in temperate lakes. Primarily from newspaper accounts, long-term records of ice-on and ice-out dates are available for many lakes in the central United States. Analysis of these records shows a progressive trend of decreasing ice-out dates; an example of a typical record is shown in Figure 7. The regional trend of the duration has also been studied using remote sensing data, one of the concrete ways to quantify the changes in freshwater ecosystems that are driven by changing climatic conditions.
Beyond the observed trends in the duration of ice cover, a fundamental physical understanding of the processes linking climate and ice cover is critical to predicting the changes in ice cover that might occur in response to the buildup of carbon dioxide in the atmosphere. In turn, these changes could be used as input to ecosystem response models. Physical limnologists have been making progress in this area of research, and the schematic diagram in Figure 8 illustrates the
temperature than the ice-on date and that increased snow cover can cause a delay in the ice-out date. Thus, temperatures and precipitation measurements during both winter and spring are important parameters for climate simulations.
Lake sediments are integrative records of processes occurring in a given year. Sediments contain sensitive records of conditions in the surrounding watershed and water column. Various biological, physical, and chemical indicators can be interpreted together to construct the past climate and hydrologic conditions in a watershed. These efforts are aided by the tools of paleolimnology. In particular, knowledge of the habitat range of diatom and chrysophyte fossils has been useful. This is an example in which greater biological knowledge (e.g., improvements in the taxonomy of diatoms) can be instrumental to improvements in paleohydrology and applied areas of hydrology such as hazard assessment.
Ground Water Ecosystems
Ground water is the largest reservoir of liquid freshwater in the world. It is an important water resource that is affected by a range of human activities (Freeze and Cherry, 1979). Although the influence of microorganisms on the geochemistry of ground water has been recognized for some time, it has only been in the
past several decades that ground water has been studied as an aquatic ecosystem (Ghiorse, 1997). Ground water ecosystems differ from surface water ecosystems in that the surface area of the solid phases is very high relative to the volume of the water. Microbial populations responsible for influencing the ground water chemistry reside on these surfaces, except for a small portion of free-living bacteria and viruses that are transported by ground water flow. The distribution of microorganisms in sediments is represented schematically in Figure 9. Ground water ecosystems are inherently different from surface water ecosystems because photosynthesis is not possible. Chemotrophic processes are certainly important in some ground waters, and are generally limited by the availability of substrates such as carbon sources and terminal electron acceptors.
As emphasized in the Introduction, measuring the rates of important processes is critical when studying an aquatic ecosystem. This is particularly challenging in ground water for two reasons. One is the inaccessible nature of the ground water ecosystem itself. Often in ground water studies the greatest expense in terms of field resources is the drilling of the wells. This is a common experience among ecologists, chemists, and hydrologists who are studying ground water systems. In fact, the cost saving in using an existing well may be a factor that promotes interdisciplinary collaboration. Even when core material has been retrieved under the most careful conditions, the disturbances in the fine-scale structure of the material (see Figure 9) are bound to be large. These disturbances
are large enough to influence the rates of microbial processes during incubation experiments, in which the active microbial populations within the core material are exposed to different substrates or contaminants.
Rate measurements are also problematic in ground water ecosystems because ground waters are typically oligotrophic. The concentrations of substrates are very low, causing the microorganisms to grow at very slow rates. Even though the water flow rates are much slower than in surface waters, these flow rates can still be substantial compared to the growth rate of a bacterial cell in the interstices of a sand grain (Smith and Garabedian, 1998). In addition to bacteria and viruses, protozoan grazers occur in ground water ecosystems, and the rates of grazing relative to the growth rates of the bacterial populations are very hard to measure or estimate.
With the advance of ground water ecology, the initial view of the subsurface habitat as uniform was replaced by a three-dimensional view of the subsurface, which contains biogeochemical gradients of contaminants. It was recognized that these gradients were controlled by microbial oxidation/reduction processes, in which the sequence of electron acceptors used by degrading microorganisms corresponds to a spatial sequence in biogeochemical conditions in the subsurface (Smith, 1996).
Through collaborations with ground water hydrologists, the methods developed to study ground water flow have been extended to determine the in situ rates of microbial processes. The basic approach is to quantify the hydrology using conservative tracers and solute transport model, and then evaluate the rate of a microbially mediated reaction by comparing the conservative tracer simulations with simulations that include microbial processes. Figure 10 shows the arrival of bromide and methane in an experiment conducted to study microbial oxidation of methane (Smith and Garabedian, 1998). The experiment took place at the Otis Air Force base research site, which is a sand and gravel aquifer contaminated with dilute, treated sewage. In this experiment, methane was coinjected with bromide into the aquifer, and the attenuation of the methane concentration relative to the bromide allowed for quantification of the oxidation rate of methane in this aquifer. Although such field experiments are quite resource intensive, they provide for the coupling of the hydrologic and microbial processes, yielding more definitive information about the process rates in ground water ecosystems. These field-scale experiments are a critical tool for studying ground water contamination and designing in situ ground water remediation.
Toward ''Knowledge-Based" Consensus: Future Science and Information Needs
In the introductory discussion the link between water resources and social progress was emphasized. This link has been operational in the United States during this century and has also been maintained by advances in aquatic science
and technology. In many regions of this country, city planners projected expanding populations for their cities and developed large systems of reservoirs to provide water supply for the future. Their foresight has benefited the current metropolitan residents, who have derived both clean water and recreational use from man-made reservoirs. Also in this century, there have been significant improvements in drinking water quality through the chlorination of drinking water, resulting in an overall increase in public health. Regulations coming into effect in the next few years are more sophisticated, addressing the production of harmful chlorinated organic compounds during chlorination of drinking water.
While advances were made in some areas of water resources in the mid-century, the quality of many surface waters deteriorated because of municipal and industrial pollution. Implementation of federal legislation has resulted in significant improvements in water quality in many rivers. For numerous Midwestern cities the revitalization of the downtown areas has been centered on developing riverside parks where residents and weekend visitors from the suburbs can stroll, jog, or rollerblade alongside a river that was once seriously
contaminated. Although restoration of wetlands and remediation of ground water have been more limited, similar benefits to the public good are anticipated.
What will be required to continue this progress as pressures on water resources build? Many competing interests will be involved in resolving these pressures, and the involvement of stakeholders at the local and regional levels will result in decisions that are specific to a particular aquatic ecosystem. Generic water quality standards will be replaced by site-specific or regional aquatic life criteria. Recent first steps in regional assessments include the report on aquatic ecosystems to the Western Water Policy Review Advisory Commission (Minckley, 1997) and the regional assessment of the impacts of climate change in North America (Cushing, 1997). Federal legislation and regulations alone will not be adequate. Greater scientific knowledge, a more broadly trained cadre of aquatic scientists and engineers, and sufficiently detailed and accurate monitoring networks and databases for the nation's freshwater resources are necessary.
Educational and Professional Issues
Recent assessments of the status of education in limnology have identified major improvements needed for the future (Wetzel, 1995; National Research Council, 1996). Over the past quarter century, limnology has grown into a truly integrative science, fulfilling the vision of the founders of limnology. With this growth, limnology has become dispersed among the academic departments in a manner that varies greatly among institutions. This fragmentation of limnology courses within a university decreases the probability that a student will have easy access to course work that would form a solid foundation for a career in linmology. Further, the lack of emphasis in limnology by one department may translate into inadequate support for laboratory and field training for undergraduate students studying limnology. The problem of availability of courses is especially acute at the undergraduate level. At many universities, undergraduates were not able to take limnology courses because the enrollment demand could not be accommodated by resources assigned for instruction in limnology.
The lack of breadth in graduate training in water-resource-related fields is also a concern. For example, it is not uncommon in civil and environmental engineering departments for students to be required to master hydrology and water chemistry but to have no requirement to learn aquatic ecology. Students may not encounter ecosystem management approaches and aquatic life criteria as targets in remediation until after graduation. The Committee on Opportunities in the Hydrologic Sciences made recommendations to improve education in limnology by establishing regional centers of limnology through the creation of strong aquatic science departments offering comprehensive undergraduate majors in limnology and by establishing strong interdepartmental programs with an opportunity to specialize in limnology. Other recommendations centered on strength-
ening limnology as a profession through certification programs for limnologists offered by professional societies and establishing a limnologist job classification within the federal government's hiring system. This would help to define for water resource managers the types of training that would be useful in implementing an ecosystem management approach for a water resource.
Aquatic Ecosystem Monitoring and Characterization
As has been emphasized in this paper, the important way in which hydrology defines aquatic ecosystems is in defining the flow of water, solutes, organisms, and detritus through the ecosystem over a range of temporal and spatial scales. Hydrochemical data are a key ''raw material" in the coupling of aquatic ecology and hydrology. Just as long-term hydrologic records have been important in designing reservoir systems for flood control and water supply, long-term hydrochemical records such as for the Hubbard Brook watershed are important in understanding water quality trends. Thus, in order to "take an ecosystem approach," some relevant hydrologic and chemical information is needed about a given aquatic ecosystem or a similar nearby ecosystem. The specific data needs are dependent upon the specific question or issue. The willingness of the stake-holders to participate in a management plan may be influenced by the quality and the credibility of the database upon which the ecosystem analysis is based.
For these reasons now is not the time to be consolidating and merging monitoring networks and degrading the geographical resolution of the current networks. The seriousness of this issue is reflected in the decreasing trend in the number of stream-gaging stations operated by the U.S. Geological Survey; in 1990, 7,400 stations were operated, and in 1997 the number of stations decreased to 6,800 (Carlowicz, 1997). It is time to protect and preserve existing monitoring networks that provide basic data on flow and chemistry with a degree of reliability that has established the network as a source of credible data. For a "knowledge-based" consensus to take hold among different competing interests involved in a water resource issue, some hydrologic records are required to form a basis for interpretation of other information.
Aquatic ecosystems are defined by hydrologic processes. The great advances that have been made in limnology in the past several decades have been fueled by interactions with hydrologists. Research programs in limnology should be greatly expanded such that the scientific benefits of these interactions can be realized in a timely manner. Through expanded research, we will develop a more holistic approach towards addressing the freshwater issues of the future. Thus, the interaction of limnologists and hydrologists should be encouraged, and con-
tinued efforts should be made to influence members of the engineering community who discourage the participation of limnologists (such as Lyon, 1997).
To foster this ongoing collaboration, we must strengthen undergraduate and graduate education in limnology. In addition to changes within the universities, establishment of a "limnologist" professional track in the federal and state governments would help to advance limnology as a profession. It is also important to continue the long-term monitoring programs that provide the hydrologic and chemical data needed for ecosystem management. We should not let the budget ax fall now when these data can be used effectively to achieve "knowledge-based consensus" in the management of specific freshwater resources.
I acknowledge helpful discussions with D. Niyogi, A. Brown, R. Smith, J. Cole, J. Baron, and G. Hornberger, which provided ideas and insight for the manuscript.
Bayley, P. B. 1995. Understanding large river-flood plain ecosystems. Bioscience 45: 153–158.
Carlowicz, M. 1997. USGS retrenches, regroups after weathering budget hazards. EOS 78(16):165–168.
Collier, M. P., R. H. Webb, and E. D. Andrews. 1997. Experimental flooding in Grand Canyon. Sci. Am. 276(1):82–89.
Cummins, K. E., C. E. Cushing, and G. W. Minshall. 1995. Introduction: An overview of stream ecosystems. In River and Stream Ecosystems. Ecosystems of the World Volume 22. Amsterdam, New York: Elsevier.
Cunliffe, B. 1993. The Roman Baths: A View over 2000 Years. Bath, England: Bath Archeological Trust.
Cushing, C. E. 1997. Fresh Ecosystems and Climate Change in North America: A Regional Assessment. New York: Wiley.
DeStasio, B. T., Jr., D. K. Hill, J. M. Kleinhans, N. P. Nibbelink, and J. J. Magnuson. 1996. Potential effects of global climate change on small north-temperate lakes: Physics, fish, and plankton. Limnol. Oceanogr. 41:1136–1149.
Freeze, R. A., and J. A. Cherry. 1979. Groundwater. Englewood Cliffs, N.J.: Prentice-Hall.
Ghiorse, W. C. 1997. Subterranean life. Science 275:789–790.
Golley, F. B. 1993. A History of the Ecosystem Concept in Ecology. Binghampton, N.Y.: VailBallou Press.
Johnson, B. L., W. B. Richardson, and T. J. Naimo. 1995. Past, present, and future concepts in large river ecology. Bioscience 45(3):134–141.
Likens, G. E., F. H. Bormann, R. S. Pierce, J. S. Eaton, and N. M. Johnson. 1977. Biogeochemistry of a Forested Ecosystem. New York: Springer-Verlag.
Lindeman, R. L. 1941. Seasonal food-cycle dynamics in a senescent lake. Am. Midl. Nat. 26:636–669.
Lyon, W. A. 1997. A research agenda or an ideological treatise? Review of the Freshwater Imperative. Rivers 5(4):304–307.
Mayr, E. 1982. The Growth of Biological Thought. Cambridge, Mass.: Harvard University Press.
Minckley, W. L. (ed). 1997. Aquatic Ecosystems Symposium. A report to the Western Water Policy Review Advisory Commission. Arizona State University, Tempe.
Minshall, G. W., R. C. Peterson, K. W. Cummins, T. L. Bott, J. R. Sedell, C. E. Cushing, and R. L. Vannote. 1983. Interbiome comparison of stream ecosystems dynamics. Ecol. Monographs 53:1–25.
Naiman, R. J., J. J. Magnuson, D. M. McKnight, and J. A. Stanford. 1995. The Freshwater Imperative: A Research Agenda. Washington, D.C.: Island Press.
National Research Council. 1996. Freshwater Ecosystems: Revitalizing Educational Programs in Limnology. Washington, D.C.: National Academy Press.
Newbold, J. D., J. W. Elwood, R. V. O'Neill, and W. VanWinkle. 1981. Measuring nutrient spiraling in streams. Can, J. Fish. Aquat. Sci. 38:860–863.
Odum, E. P. 1953. Fundamentals of Ecology. Philadelphia: W. B. Saunders.
Pool, R. 1991. Science literacy: The enemy is us. Science 251:266–267.
Postel, S. 1997. Dividing the waters. Technol. Rev. 100(3):54–62.
Power, M. E., A. Sun, G. Parker, W. E. Dietrich, and J. T. Wooton. 1995. Hydraulic food-chain models. Bioscience 45:159–167.
Richter, B. D. 1995. Integrating science in applied aquatic ecosystem management. Bull. NABS 12(1):82.
Smith, R. L. 1996. Determining the terminal electron-accepting reaction in the saturated subsurface. In Manual of Environmental Microbiology. C. J. Hurst, G. R. Knudsen, M. J. Mclnerney, L. D, Stetzenbach, and M. V. Walter, eds. Washington, D.C.: ASM Press.
Smith, R. L., and S. P. Garabedian. 1998. Using transport model interpretations of tracer tests to study microbial processes in ground water. Pp. 94–123 in Mathematical Modeling in Microbial Ecology, A. L. Loch, J. A. Robins and G. A. Milkier, eds. New York: Chapman and Hall.
Sollins, D., C. C. Grier, F. M. McCorison, K. Cromack, Jr., R. Fogel, and R. L. Fredriksen. 1980. The internal element cycles of an old growth Douglas-fir ecosystem in western Oregon. Ecol. Monogr. 50:261–285.
Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. The river continuum concept. Can. J. Fish. Aquat. Sci. 37:130–137.
Vavrus, S. J., R. H. Wynne, and J. A. Foley. 1996. Measuring the sensitivity of southern Wisconsin lake ice to climate variations and lake depth using a numerical model. Limnol. Oceanogr. 41:822–831.
Wetzel, R. G. 1995. Training of aquatic ecosystem scientists: Continue to languish or accept our responsibilities? Water Resour. Updates 98:15–20.