Major River Science Drivers and Challenges
The nation faces many complex challenges in its stewardship of natural resources. Human landscape alterations over the last few hundred years have significantly changed the form and functioning of river systems. Population and economic growth have dramatically increased the competing demands for limited river-based services that are central to the growth and health of the nation’s economy and quality of life (including water supply, navigation, recreation, flood control, and hydroelectric power). Simultaneously, expectations to maintain and restore the natural functioning of the nation’s river ecosystems are increasing. Indeed, trade-offs between the impacts on ecosystems, sustainable allocation of water resources, and economic interests are often at the center of regional and interstate management problems. Furthermore, innumerable rivers flow through federal lands (such as the national parks, national forests, wildlife refuges, and military bases) and thus the government also has a legal interest in river policy and management. Current projects such as restoring the Florida Everglades and Pacific Northwest fisheries, decreasing Mississippi-Gulf hypoxia, and minimizing the impacts of urbanization on streams and rivers, reflect a pervasive national need for science-based information to support policy and management decisions that affect the nation’s rivers.
The impact of human landscape alterations on the valuable goods and services rivers provide have given rise to public policy debates about how rivers should be optimally managed, debates that would benefit from science-based information. Humans have utilized rivers for centuries, and in the process have altered their form and function. These alterations have, however, become so pervasive they are having cumulative impacts that society can no longer ignore. Historically, the incremental effects of human activities on rivers have been
managed as de minimis local perturbations. Yet from the ubiquitous presence of pharmaceuticals in natural waters to the wholesale diversion and consumptive use of water resource systems, the cumulative impacts of local management decisions have surpassed anticipated consequences. Furthermore, they confound the traditional management frameworks that guide policy and management decisions. Therefore, the national need for a new integrated multidisciplinary river science is more compelling now than ever before.
Within this chapter, we explore these challenges that are driving our current policies and management practices and thus our need for river science: ecological restoration (including dam removal), relicensing of hydropower facilities, invasive species, water allocation, climatic variability, urbanization and other land-use changes, and water quality. We also mention briefly the economic value of river ecosystem services, a matter that is particularly relevant when the case for river science is rationalized based on the values rivers provide. We conclude with an outline of the characteristics that river science needs to have to confront the individual unmet challenges and overall cumulative effects that human activities have on river ecosystems.
ECOLOGICAL RESTORATION AND DAM REMOVAL
Human and natural actions have caused the loss or degradation of riverine habitat. Throughout the country, thousands of ecological restoration efforts are being undertaken to improve water quality, manage riparian zones, improve habitat, and stabilize streambanks (Bernhardt et al., 2005). Billions of dollars are being spent on small projects including an Ecosystem Initiative on the Platte River (Box 2-1) and billions more on major restoration projects in the Everglades, coastal Louisiana, the California Bay and Delta, and the upper Mississippi River. Yet the science of river restoration is still in its infancy.
Dam removal is a high-profile form of river restoration. Because dams with significant storage capacity dramatically alter riverine flows—creating lakes where rivers once flowed and fundamentally altering the downstream flow regime—they have had enormous impacts on ecological patterns and processes in rivers. Currently in the United States, the National Inventory of Dams lists 76,000 dams that exceed 2 meters in height and estimates 2 million more of less than 2 meters in height (Graf, 1999; http://crunch.tec.army.mil/nid/webpages/nid.cfm). For many of these dams, their original uses have long disappeared and they stand only to hold back river water and are thus a repository for collecting river sediments, nutrients, and contaminants. Almost 500 dams were removed during the 20th century, most of which were less than 5 meters in height (Poff and Hart, 2002). Removal of these small dams has been accelerating in recent years as their economic viability declines, dam owners realize their increasing legal and financial liabilities, and the government recognizes the environmental benefits of their removal.
Restoration of Biological Habitat: The Platte River, Central Nebraska
Braided reaches of the Platte River in central Nebraska provide important habitat for migratory and nesting birds, including three endangered or threatened species: the whooping crane (Grus americana), the northern Great Plains population of the piping plover (Charadrius melodus), and the interior least tern (Sterna antillarum athalassos) (NRC, 2005). Further to the east, the Platte River provides habitat for the pallid sturgeon (Scaphirhynchus albus), which is also endangered. A considerable amount of the riverine and wetland habitat available to these species prior to human settlement has now been lost or altered. Much of this loss can be attributed to changes in streamflow hydrology caused by reservoirs and water-diversion systems, some of which lie far upstream of the areas of interest.
The USGS has been involved in hydrologic studies of the Platte River system for many years. The current effort—known as the USGS Platte River Ecosystem Initiative—involves all four disciplines within the USGS (Biological Resources, Geologic, Geographic, and Water Resources). Its objectives are to (1) provide a better understanding of migratory and resident birds and the ecology of their habitats and of the physical processes that influence these habitats, and (2) use this knowledge to evaluate the effects of different management strategies on individual species and habitats (http://mcmcweb.er.usgs.gov/platte/index.html). The initiative includes eight project elements, each targeting specific concerns or questions regarding the hydrology and/or ecology of the Platte River system. Collectively this work will improve the understanding of linkages between hydrology, geomorphology, biological communities, and ecosystem processes, which can then be used to develop strategies to sustain or rehabilitate the riparian ecosystem of the central Platte River.
However, the loss of the benefits associated with a dam together with the costs of removal must be balanced against the benefits created by the removal. This requires an understanding of the river system’s likely response to dam removal. The sediment accumulated behind the dams, which may contain toxic materials, must be considered and plans made either for its disposal or to address the effects of its movement into the downstream channel. If restoration of aquatic habitat is a goal, how will this be accomplished and what other actions will this restoration require? How does one measure or assess the benefits of the removal? Even at dams that are no longer in use, the overall costs and benefits and impacts of removal must be developed. The science to support these analyses is in its infancy and limited by the small number of case studies (see Boxes 2-2 and 2-3) where dam removal has been followed by adequate monitoring.
Among scientific and many practitioner communities, there is a strong consensus that this restoration should be by adaptive iteration—try an approach, monitor to see how it works, and then adjust if necessary. However, legal and
Dam Removal: The Elwha and Glines Canyon Dams, Washington
The proposed plan to dismantle the Elwha and Glines Canyon dams on the Elwha River in the Olympic Mountains, Washington, represents the most ambitious dam removal project in the United States. Authorization to remove these two dams comes from the Elwha River Ecosystem and Fisheries Restoration Act, passed by Congress in 1992 to restore the natural system. As the name of the act implies, the primary purpose of dam removal is to restore the ecosystem and native salmon fisheries of the Elwha River system. This “natural” system restoration goal implies a state that is often difficult to define because of natural variability and a lack of ecosystem information prior to dam construction. USGS scientists have played important roles in evaluating the potential impacts of dam removal in two key areas: (1) assessing the fate of the sediment stored behind the dams and (2) assessing the suitability of in-channel habitats and water-quality characteristics for restoring ecosystem processes.
During a lake drawdown experiment in 1994, USGS hydrologists made bed load and suspended load measurements in the vicinity of a delta formed at the head of the upper reservoir (Lake Mills) to evaluate the potential mobility of sediment stored behind the dams (Childers et al., 2000). The results of that work were used to develop a sediment transport model for managing the movement of sediment into and through the channel reach downstream of the dams.
In a related study, USGS hydrologists collected baseline data to assess nutrient concentrations and habitat characteristics throughout the basin (Munn et al., 1999). They found that nutrient concentrations were quite low (often below detection limits) suggesting that nutrients from salmon carcasses brought into the system by the salmon’s natural life cycle would greatly benefit the ecosystem if additional instream habitat were restored. They also found that, while most of the tributary channel network is very steep (> 16% slope) and unlikely to provide much spawning and rearing habitat, the main stem of the Elwha River upstream of the dam contains large areas of potential habitat.
logistical issues often discourage such an approach. Often funds are available only for a short period of time or for a single restoration action. Further, because the majority of these projects have had little or no monitoring after restoration activities, we know little about which restoration approaches are most effective or even if projects are completed as designed (Hassett et al., 2005).
Research on how best to measure restoration effectiveness and how to quantify multiple river ecosystem services in meaningful ways is critical. Palmer and Allan (2006) recommended that a coordinated national study evaluating the effectiveness of different restoration approaches, particularly those that are expensive and highly interventionist (e.g., channel reconfiguration projects) be com-
Dam Removal: The O’Shaughnessey Dam, Hetch Hetchy Valley, California
O’Shaughnessey Dam that floods the Hetch Hetchy Valley, California, has been controversial for the past 100 years. It is the only major dam built in a national park and was strongly opposed by John Muir, who compared the Hetch Hetchy Valley in beauty and uniqueness to Yosemite Valley. Despite the opposition, San Francisco has managed to maintain its hold on Hetch Hetchy because of scarcity of water supplies in northern California. The removal of O’Shaughnessey Dam and the restoration of the river and the valley have received serious consideration because of two studies that show that with optimal use of other dams in the region, the 250,000-acre-foot-capacity dam could be removed with only a minor loss of 20,000 acre feet of water (Null, 2003; Rosekrans et al., 2004). In addition, the recreational pressure in Yosemite Park is very high, especially in the valley. Demand for recreation in a parallel valley in future years would clearly add to the economic base of the region. There will be additional costs from the removal of O’Shaughnessey Dam, primarily due to the loss of hydropower and the need for substantial additional treatment of San Francisco’s water if stored in downstream dams.
The sediment disposal problem that hinders many dam removal proposals is minor in this case, as the largely granitic High Sierra catchment has yielded very little sediment over the past 90 years. There are even plans to replace the original trees by identifying their stumps if they become exposed. Surveys during the last major drought showed the feasibility of this restoration.
However, the debate on the removal of the dam has moved from one about water and scarcity to one that pits the interests of the City of San Francisco against those of the local regions around Hetch Hetchy that would benefit from continued growth in the tourist trade. Since very little water is lost, these impacts can now be assessed on a purely financial basis, taking into account the environmental benefits from a Hetch Hetchy restoration. In the long run, given the increase in the scarcity value of prime recreational sites, and the ability to effectively store water elsewhere, the removal of the dam is likely. Additional information on the impacts to the river environment and other effects from the removal of the dam by the USGS would provide a reference against which the optimal interests of the nation in this case could be judged.
pleted and that the USGS was poised to undertake such research. Other research frontiers include determining how best to restore ecosystem processes under highly constrained conditions that surround dams and levees or in urban settings, how to move some aquatic ecosystems beyond “restoration” to boost their ability to perform functions of value to society, and how to identify feedbacks associated with critical thresholds beyond which river restoration is not possible (Palmer and Bernhardt, 2006).
RELICENSING OF HYDROPOWER FACILITIES
Hydropower dams are a source of considerable discussion because they are subject to periodic relicensing by the Federal Energy Regulatory Commission (FERC). In the relicensing process, dam operators must discuss, among other things, the impact of their operations on the riverine environment and steps, if any, they will take to mitigate negative impacts. Citizen groups and environmental nongovernmental organizations frequently raise questions about these impacts and ask why the dams should not be removed or have their operations modified. Many of the questions center on the timing and volume of releases from the structures and the impact these have on water quality and instream habitat and aquatic species below the dams. As with dam removal, these questions rest on knowledge of the present state of the river and the ability to predict future conditions. Over the years, approaches for modeling habitat changes downstream of dams have evolved from relatively simple measurements of minimum wetted habitat for select aquatic species to more complex analyses. These more involved analyses consider how the altered flow regime modifies the normal distribution of high and low flows and investigates how the flows influence a much broader suite of species and affect the entire aquatic ecosystem. Still more work is needed to gain a predictive understanding of the linkages between flow variations and river habitat in highly regulated rivers.
Deliberate and unwitting human actions have brought thousands of alien invasive species to this country and spread them across the aquatic and terrestrial landscape. Saltcedar has infested watercourses throughout the western United States, causing declines of native riparian trees, and is now the object of a multimillion-dollar eradication campaign (Morisette, 2006) (Box 2-4). The Asian carp, originally brought to this country with government approval to clean the beds of southern catfish farms, escaped and now threatens fisheries in the Mississippi and Illinois Rivers and possibly the Great Lakes (Kolar and Lodge, 2002). The zebra mussel, which infests the Great Lakes and major rivers such as the Mississippi and Ohio, has cost municipalities, utilities, and the fishing industry billions of dollars.
These and thousands of other plants and animals species have entered the country attached to packing crates, interspersed with agricultural imports, or in the ballast discharged into our ports and harbors. Insufficient knowledge exists about how these species spread and, therefore, how they can be eradicated or, more plausibly, how their spread can be limited (International Joint Commission, 2004). These exotic species often flourish in highly altered river ecosystems. Therefore, much thought needs to be given to how river and riparian restoration can be used to minimize the success of these exotics, and how much restoration is needed to accomplish a given level of reduction.
Invasive Species: Tamarisk Invasion in the Western United States
Riparian areas bordering streams and rivers in arid regions of the western United States have been profoundly changed by the invasion of non-native plants, such as Eurasian saltcedar or tamarisk (Tamarix ramosissima). Since its introduction in the 1800s, Tamarisk is now the third most common woody riparian plant species in western river ecosystems in North America (Friedman et al., 2005). Comparisons of early and modern-day photographs of floodplains and riparian segments of semiarid streams illustrate vividly how native plants such as sandbar willow (Salix exigua) and Fremont cottonwood (Populus fremontii) have been completely replaced by tamarisk.
In the 1960s, the USGS initiated field studies on rivers throughout the western United States to evaluate the role of tamarisk and other phreatophytes in cycling water through riparian areas. The results indicated that floodplains covered with mature tamarisk can transpire 1-3 meters of water per year (Weeks et al., 1987), which far exceeds annual precipitation rates. In contrast to native riparian plants, tamarisks have a high leaf area index (leaf area per unit of ground area); thus, they are highly efficient in cycling water between the land surface and the atmosphere (Shafroth et al., 2005).
More recently, USGS scientists at the BRD Science Center in Fort Collins, Colorado, have been examining interactions among streamflow, fluvial geomorphology, and riparian vegetation, including environmental factors that favor non-native plants such as tamarisk (http://www.usgs.gov/invasive_species/plw/). One such project uses data and information from 500 long-term gaging stations in 17 western states. Data from these sites will be used to relate the abundance of native and non-native woody riparian plants to the timing and magnitude of streamflow, channel geometry, salinity, and climate. The long-term objective of this work is to identify the factors influencing the current distribution of native and non-native plants and provide resource managers with information and tools for predicting the spread of non-native species.
Research on the life history characteristics of the invaders and how to predict the rate of their spread as a function of river hydrology and network configuration is critical. This should be coupled with research on the consequences invasive species have on river biodiversity and ecosystem processes.
WATER ALLOCATION AND REALLOCATION
Recently, five years of severe drought in the western United States brought attention to the allocation of water among water users. Forecasts that the drought might have continued and would have required a reallocation of the waters of the Colorado River heightened political attention to allocation issues. Coupled to large-scale issues, such as those of the Colorado River, are numerous other conflicts over the allocation of flows in smaller streams. Efforts to protect the sil-
very minnow in the Rio Grande River through directed low flows led to years of lawsuits with the City of Albuquerque and the Albuquerque-Bernalillo County Water Utility Authority, who were concerned about maintaining a reliable water supply. Similar efforts are also underway in the Klamath Basin in Oregon where agricultural interests are in conflict with the government over the need to provide flows for the threatened and endangered species in the lakes and rivers (suckers and coho salmon). In this case, an NRC study concluded that the scientific information available to make these decisions was scarce, and so in its absence, resource managers were forced to rely more heavily on their judgment and follow the precautionary principle in prescribing flow requirements (NRC, 2004b).
While many water allocation conflicts are concentrated in western states (Box 2-5 provides another example), they are becoming more common in the east as well. Periodic drought combined with rapidly growing urban areas has led to conflicts in major river basins, such as the Apalachicola-Chattahoochee-Flint (Box 2-6) and the Potomac.
Some of the research needed to address such issues is not, strictly speaking, river science, but rather involves water-use practices (e.g., conservation, conjunctive use of groundwater and surface water, water reuse). However, a better understanding of the habitat needs of species throughout their life cycle and the
Water Allocation and River Restoration: The San Joaquin River, California
The San Joaquin River, the second longest river in California, has been the focus of a recent water allocation controversy. When the Friant and Millerton Dam came online in the 1950s, the flow in the river was reduced substantially, and a once thriving salmon run became extinct. Recently, a coalition of 13 environmental and fishing groups (http://www.nrdc.org/media/pressreleases/040827.asp) acquired a court order to restore water in the San Joaquin River to quantities sufficient to support a salmon run. The suit has been in the courts since 1988. Currently, there is a wide divergence of views on how much water would be needed, ranging from about 200 to 500 KAF (1000 acre feet, 1 KAF = 1.2335 million cubic meters). In addition, the methods with which flows can be restored in the river differ widely. Proponents of restoration claim the majority of flows can be obtained by more efficient operation and conservation in the river basin. Opponents contend that increase in flows can come only by the reduction of irrigated croplands, which consequently would have a high cost in foregone production and associated job loss.
An impartial source of technical information on flow requirements needed to restore the fishery would help resolve this difficult water allocation problem. A multidisciplinary assessment of the amount and timing of releases, the amount and quality of the habitat that could be created, and the response of salmon to these conditions is needed to assess management alternatives.
Water Allocation and Urban Growth: Atlanta, Georgia, and the Apalachicola-Chattahoochee-Flint Basin
As populations grow and consumptive use of water increases, concerns of water shortages, which have long been common in the western United States, are becoming increasingly common in the east. In Atlanta, Georgia, the population has more than tripled in the last 50 years and demand for water is expected to increase by 50 percent by 2020. Most growth is occurring north of the city center, where suburban developments with large lawns, water-thirsty plants, and swimming pools are increasing. Atlanta’s residents share water from the Apalachicola-Chattahoochee-Flint (ACF) river basin with residents of Alabama and Florida. Multiple stakeholders throughout the region all lay claim on river waters, including navigation interests, hydroelectric power generators, and industrial water users. Additionally, for ecological purposes Florida residents have been proponents for maintaining natural flow regimes.
As of early 2006, many streams and rivers in the basin are flowing at less than half their usual rate (http://water.usgs.gov/waterwatch/). Therefore, reservoir levels are dropping, and “water wars” have arisen between hydroelectric power interests, lakeside homeowners, water recreation enthusiasts, and residents who rely on the reservoirs for drinking water. Drought conditions have further exacerbated the problems and have led to emergency restrictions on water use. New water- and land-use policies are expected in the coming decade.
The environmental implications of these water problems and thus the most effective policies are, however, still unclear. These issues are well suited for an interdisciplinary science approach, as is the development of scientific solutions for increasing the efficiency of water distribution systems that allow for protecting ecological water needs while meeting human needs.
relationships among habitat, hydrology and hydraulics, geomorphology, and other characteristics are key to providing the scientific foundation for evaluating the ecological consequences of flow management decisions.
The ecological and sociological impacts of climatic variability have been and will continue to be significant on seasonal, annual, and decadal timescales. While there may not be complete agreement on the anthropogenic contribution to climate change, the evidence of significant climate variation both from the modern instrumental record and from historical reconstructions from tree rings and other paleoclimate data is indisputable. Streamflow variability in the western United States is strongly tied to large-scale interannual and interdecadal climate oscillations (Redmond and Koch, 1991; Kahya and Dracup, 1993; Dracup and Kahya, 1994); low-frequency streamflow variations over the last 500 years have also been shown to be related to large-scale climatic variations in the
Pacific and Atlantic Oceans (Hidalgo, 2004). Reduced rain or snowfall (e.g., Mote et al., 2005) or changes in snowmelt timing (Box 2-7) can severely impact agricultural production and threaten to increase conflicts over scarce water resources. Conversely, in other parts of the United States, observed increases in heavy precipitation events in recent decades (Karl and Knight, 1998; Kunkel et al., 1999; Groisman et al., 2001) might increase flooding in those areas. Climatic variability changes the flow regime of river systems, which is a “master variable” controlling river geomorphology and ecology (Poff et al., 1997). For instance, the temporal sequence of floods can affect the configuration of the river channel and floodplain (Schumm and Lichty, 1963). In the arid Southwest, episodes of channel down-cutting are associated in part with changes in the fre-
Climatic Variability: Earlier Snowmelt in the Western United States
The hydrologic response to climate change in the 21st century may be most apparent in western mountains. Historical streamflow records for the western United States indicate that the timing of the snowmelt spring pulse is earlier by about 10 days than in previous decades (Regonda et al., 2005; Stewart et al., 2005). The trend is widespread throughout the Sierra Nevada. Timing changes are related to long-term warming trends during the winter and spring across the western United States (Dettinger and Cayan, 1995). Warming in mountainous areas affects not only the timing of the spring melt but also the partitioning of precipitation into snow or rain. In catchments sensitive to changes in this partitioning, such as the American River, changes in the frequency of floods have also been observed (NRC, 1999a).
Although large-scale interannual and interdecadal climate variations related to the El Niño/Southern Oscillation (ENSO) and the Pacific Decadal Oscillation (PDO) account for part of the streamflow trends, climate change during the late 20th century is also significant (Stewart et al., 2005). Projected changes in climate over the 21st century predict a climate that would result in greater shifting in seasonality of streamflows, with more precipitation as rain, more frequent flooding, earlier loss of mountain snowpack, and lower summer and autumn baseflows (Dettinger et al., 2004; Knowles and Cayan, 2004). Given the sensitivity of snow-fed western mountain rivers to air temperature and precipitation variability, the impact of large-scale climate change would be significant in this region.
Understanding and predicting the impacts of climatic variability and change on river flow regimes and flood frequencies is only one of the challenges. A changing hydrologic regime would also put a greater strain on water resources management. In the western states the competition for water is already intense and is represented by numerous users with interests in both consumptive and nonconsumptive water use. Recorded variations in flows representing climatic conditions of the recent past will be insufficient to guide decisions on future operations. A science-based understanding of climate changes and their impacts on watershed hydrology will be needed to manage water resources during the 21st century.
Climatic Variability: Trends in High and Low Flows
Concerns over the hydroclimatic effects of climate change have led to widespread speculation that increases in temperature may accelerate the hydrologic cycle, producing more extreme floods and severe droughts. USGS hydrologists Harry Lins and Jim Slack (1999) examined this by studying trends in streamflow in the conterminous United States during the 20th century. They used historical records from 395 climate-sensitive gaging stations to evaluate differences in high, medium, and low flows; each of the stations had continuous daily records from 1944 to 1993 (50 years), and 34 had records of 80 years or more.
Results showed that since the 1940s, values of annual minimum to median flows have generally been rising throughout most of the United States; the number of stations with upward trends in low flows is much higher than the number exhibiting downward trends. However, there has been little change in high flows. For annual maximum daily discharge, only 35 stations showed a significant trend, and these were roughly balanced between positive and negative trends. Increases tended to be in the East and upper Midwest; decreases were scattered throughout the country. McCabe and Wolock (2002), using 400 sites in the conterminous United States measured during 1941-1999, reached similar conclusions, but found a “step” increase in annual minimum and median daily streamflow around 1970, followed by stabilization of the new regime.
Long-term trends for the western United States may, however, be toward lower overall discharge. Using an ensemble of 12 climate models to predict relative changes in runoff in the 21st century, Milly et al. (2005) projected an estimated 10-30 percent decrease in runoff in midlatitude western North America by the year 2050. Such changes in sustainable water availability would have regional-scale consequences for both our economies and ecosystems (Milly et al., 2005).
quency of large floods during the Holocene (Hereford, 2002). Variations in the magnitude and frequency of high and low flows (Box 2-8) affect the creation, availability, and quality of river habitat. Climate extremes and low flow conditions can also combine to create acute conditions for aquatic species by raising water temperatures above threshold limits. Therefore, an improved understanding of the role of large-scale climate oscillations and trends on streamflow timing and seasonality is needed.
With an improved ability to forecast short-term seasonal and interseasonal climate variations, there are more opportunities to better manage water resources for aquatic species, particularly in the context of adaptive management experiments (Pulwarty and Redmond, 1997; Pulwarty and Melis, 2001). Furthermore, a better understanding of the impacts of climate variations on riparian vegetation and river aquatic habitat is needed to implement effective restoration measures or control invasive plant species.
URBANIZATION AND OTHER LAND-USE CHANGES
Just as climatic variability inhibits our ability to rely on past conditions to develop future climate predictions, ongoing changes in urban occupancy and rural land use hinder our ability to predict future land-use impacts. The quantity and quality of river flows are tied directly to land-use activities (Allan, 2004). Seemingly minor changes in land use can create significant changes in the pattern of runoff that reaches streams and rivers. Therefore, our ability to accurately predict and manage floods or to estimate water availability is limited by our understanding of these minor changes. Current federal regulations governing flood insurance generally do not require the consideration of future conditions, although there is a movement toward including predictions of future riverine hydrology in flood stage predictions.
Changes in the quantity and quality of water because of rural land-use modifications and urbanization can severely affect downstream environmental conditions on both regional and local scales. These changes lead to habitat modifications that affect both aquatic and terrestrial species dependent on access to clean water (Moglen et al., 2004). Regional modifications in rural landuse activities have changed how the landscape regulates the flow of water to streams (DeFries et al., 2004). These regional shifts in land management have occurred throughout the 20th century. Examples include reforestation in the Southeast, changes in grazing management in the interior West, and alterations in logging practices in the western coastal states. On a more local scale, a study of lowland streams in western Washington, Booth and Jackson (1997) found the onset of aquatic-system degradation occurred at relatively minor levels of urban development (i.e., an effective impervious area in a watershed of 10 percent). Furthermore, the interactive effects of urbanization and climate change may be critical in some areas; urbanization can cause a dramatic rise in summer stream temperatures with large temperature spikes during rainstorms, which would likely be exacerbated in regions that experience more severe storms in the future.
Thus, the science community needs to develop a systematic understanding of the relationships between landscape changes, sediment fluxes, and ecosystem functions and services (NRC, 2001). To do this, one has to be able to distinguish between human-induced changes and natural variations in the water cycle; such work would include both field studies and model development (Hornberger et al., 2001). Field studies would logically include observations from watersheds for which good hydrological information is available and where land-use changes are documented (NRC, 2004d).
Since the passage of the Clean Water Act in 1972, the water quality of the nation’s waterbodies has generally improved. The Cuyahoga River no longer catches fire, the Potomac River is no longer labeled a national disgrace, and
Lake Erie now supports a sport fishery and is no longer considered moribund. The Clean Water Act is estimated to have prevented discharge of almost 700 billion pounds of pollutants per year, including over 1 billion pounds of toxic pollutants such as heavy metals, over 470 billion pounds of nonconventional pollutants such as nutrients and salts, and 220 billion pounds of conventional pollutants such as suspended solids (USEPA, 2002a).
However, the quality of our nation’s water remains at risk. The Environmental Protection Agency’s 2000 National Water-Quality Inventory indicated only 61 percent of assessed stream miles fully support all of their designated uses, and of these, 8 percent are considered threatened for one or more uses. In the remaining 39 percent of assessed reaches, one or more designated uses are impaired by pollution or habitat degradation. The situation is worse for lakes and estuaries; some form of pollution or habitat degradation impairs 45 percent of the assessed lakes and 51 percent of assessed estuarine waters by area (http://www.epa.gov/305b/). The primary water-quality problems include bacteria, nutrients, metals (especially mercury), and siltation. Their sources include runoff from agricultural lands, sewage treatment plants, and hydrologic modifications, such as channelization, flow regulation, and dredging (http://www.epa.gov/305b/).
This situation is made more complex by the increasing number of pollutants and the poorly understood chemical mixes that ensue, especially if they combine with other molecules in the environment to create endocrine disruptors. Increases in nutrient enrichment, hormones, and pharmaceutical products threaten our ability to identify new pollutants and to clean the waters for human, fish, and wildlife use (NRC, 2000, 2004c).
Gaps in our understanding of river water quality are numerous. Research needs include ecotoxicological studies for species of interest and contaminants of concern, studies of the fate and transport of lesser understood emerging contaminants, and investigations into the role of the hyporheic zone in transforming, adsorbing, and biodegrading these contaminants and nutrients such as nitrogen.
VALUING RIVER ECOSYSTEM SERVICES
The economic value of the different components of river science can act both as a constraint on the implementation of changes in river science policy, but more importantly, as a consistent method of comparing the social value of different river science actions. Economic approaches to valuing river ecosystem services are particularly relevant when the case for river science is rationalized based on the value rivers provide, and thus ultimately underlies all the drivers and challenges noted above. Given, however, that the USGS does not have an economic analysis capacity, and the development of one is not envisaged in the near future, we do not expand on this here but briefly address valuation methods and their challenges in Appendix A.
CHARACTERISTICS OF RIVER SCIENCE
A science that can begin to address some of the above drivers and challenges must have two important qualities. First, it must be multidisciplinary and integrative and second, it must be process-based and predictive. These qualities are discussed in the next sections.
Multidisciplinary and Integrative
In the past, much of the science needed for making policy and management decisions affecting river systems was supplied from traditional disciplinary science. For example, hydrologists provided information on flood hazards that guided management of flood-prone areas, and river forecasting models that allowed reservoir operators to make decisions on releasing flows when a flood is occurring. The work of geomorphologists on channel-forming flows guided efforts to control stormwater from urbanizing watersheds, where changes in the frequency of high flows leads to channel widening or incision. And biologists documented declines in aquatic species and birds that feed on them. The legacy of this disciplinary work continues to benefit the public in decision making today. Furthermore, advances in traditional disciplinary sciences focusing on rivers and river processes will continue to provide new insights for better management.
Yet there are needs for science-based information on rivers that cannot be met solely from traditional disciplinary approaches. All aspects of a river’s physical, biochemical, and ecological systems depend upon each other. For example, floods mobilize channel-bed sediment, while depositing silt and clay locally in floodplains. Flooding provides critical, if transient, habitat for wetland and riparian species, while also potentially disrupting the life cycles of some upland ecosystem species. During a flood, the nutrient processing efficiency of the hyporheic zone (the subsurface interface between groundwater and surface water) changes according to modifications of channel geometry and bed topography.
Additionally, cumulative impacts of human alterations pose challenges traditional disciplinary science is incapable of solving. Historically, river and watershed management have focused on local-scale problems with direct cause and effect relationships. For instance, science has contributed to solving soil erosion problems at local scales through the plot-scale study of erosion, leading to the development and widespread use of conservation tillage and best management practices. Local problems of elevated nitrogen loads in streams have also been addressed through plot-scale research on manure application rates. While this research has been extremely valuable, it is insufficient to answer questions about the processes affecting sediment movement in major rivers or hypoxia in the Gulf of Mexico. Rivers integrate the multitude of spatially distributed, small-scale alterations to the landscape and waterways, and awareness of such cumulative impacts must guide the vision for river science.
Therefore, despite inherent difficulties, guiding the emergence of a distinct river science needs to be a multidisciplinary and integrative endeavor. To understand the functions of rivers and the impacts of human alterations, river science should synthesize information from biology, geology, chemistry, and the fluid mechanics and physics governing water and sediment transport at multiple scales. Thus, the greatest challenge is to determine which components of the very large number of physical, chemical, and biological parameters in a river system should be studied to evaluate most comprehensively their interrelationships.
While interdisciplinary research is not a new phenomenon, and river ecologists have long recognized the important role of hydrology and geomorphology in governing ecological processes (Vannote et al., 1980; Stanford et al., 1996; Benda et al., 2004), truly interdisciplinary work on rivers is still in its infancy. To clarify, throughout this report the term “interdisciplinary” is defined as the intentional effort to integrate across disciplines (combining both multidisciplinary and integrative); this is distinct from “multidisciplinary,” which refers to the involvement of many disciplines but does not inherently imply integration.
In the last decade, we have seen increased emphasis on interdisciplinary research, with funding for this work coming from new programs established within existing outlets and agency sources. Some examples of recently established interdisciplinary programs include the National Science Foundation’s Biocomplexity in the Environment Program, Environmental Protection Agency’s Water and Watersheds Program, and National Oceanic and Atmospheric Administration’s Climate and Global Change Program. Other proposals for developing major interdisciplinary research programs at the National Science Foundation are on the table and may soon be underway; these include National Ecological Observatory Network (NEON) and a merged Collaborative, Large-scale, Engineering Analysis Network for Environmental Research (CLEANER)–Consortium of Universities for the Advancement of Hydrologic Science, Incorporated (CUAHSI) observatory initiative. Many of the federal agencies involved in resource management or scientific research have likewise branched out to form interdisciplinary teams, programs, or initiatives emphasizing integrated research in hydrology, ecology, and engineering. Still, an even greater degree of interaction is needed today in the study of rivers.
Process-Based and Predictive
Changes to river systems during the last 200 years have been drastic and pervasive. The nation has evolved from an agrarian society of a few million inhabitants to a country of about 300 million of whom 75 percent live in urban areas. As noted earlier, there are some 76,000 dams greater than 2 m in height and perhaps 2 million smaller structures. Rivers in urban and suburban areas are connected to complex and sometimes aging stormwater and sewage infrastructure, making it difficult to define how water is routed to channels or even what
constitutes the watershed. Overall, the channels and watershed landscapes of most of the nation’s rivers have been so modified that the concept of a “natural river system” sensu stricto reference state may no longer exist.
Despite this difficult context, many of today’s environmental policy decisions call for science-based information about the likely responses of river systems to changes in both natural forcing and human drivers. Empirical equations have helped improve the hydrological science knowledge of these responses. Generally speaking, regression-based approaches have been extremely useful in determining the likely responses within a defined context. The predictive abilities of these descriptive equations are, however, limited when the environment exceeds the normal bounds. The most challenging problems demand information about river systems’ responses to future conditions that are outside the range of historical observations and experience, such as extreme events like Hurricane Katrina.
River science must, therefore, be structured and conducted to provide a process-based and predictive understanding of river systems. This understanding must go beyond methods that have commonly been used in the past to guide policy and management decisions, that is, operational or pragmatic predictions based on empirical associations such as regression relationships. Rather, these associations should be used to help uncover the processes beneath the trends. Sound policy decisions require a sufficient understanding of river systems so as to offer sound, credible, testable predictions of river systems’ responses to new and previously unobserved forcing that could accompany climate change, excessive groundwater extractions, large-scale land-use conversion, hydrologic alterations from urbanization and stormwater management and, ironically, restoration actions designed to mitigate or reverse some of the above forcings.
CONCLUSION AND RECOMMENDATION
The understanding necessary to assess the complex changes of river systems to management alternatives will not come easily, especially in the midst of uncertainty. Therefore, there is a compelling national need for a new approach to studying rivers. River science—an emerging discipline distinct from traditional disciplinary sciences but still supported by their activities—can provide a vision for organizing scientific endeavors to address these unique challenges.
Recommendation: USGS river science activities should be driven by the compelling national need for an integrative multidisciplinary science structured and conducted to develop a process-based predictive understanding of the functions of the nation’s river systems and their responses to natural variability and the growing, pervasive, and cumulative effects of human activities.