At the dawn of a new millennium, the United States is taking stock of where it has come from and where it is headed in years to come. The quality of the nation's water resources is of great interest because it is so integrally linked to the long-term availability of water that is clean and safe for drinking and recreation and that is suitable for industry, irrigation, and habitat for fish and wildlife. Historically, "availability" of water could be viewed simply as an issue of quantity. Water management has focused largely on controlling or alleviating impacts of droughts and floods. However, with escalating population growth and increasing demands for multiple water uses, availability is now measured in terms of quantity and quality-both of which are critical to the long-term sustainability of the nation's communities and ecosystems.

Before the Clean Water Act was implemented in 1972, many surface waters in the United States were recipients of uncontrolled "point" discharges of sewage and industrial waste. (Point source contamination can be traced to specific end-of-pipe points of discharge or outfalls, such as from wastewater treatment plants, factories, or combined sewers.) Since the 1970s, these discharges have been regulated by the U.S. Environmental Protection Agency (EPA), states, and tribes. Although violations still occur, this legislation has had a positive effect on water quality. The overwhelming majority of water-quality problems are now caused by diffuse "nonpoint" sources of pollution from agricultural land, urban development, forest harvesting, and the atmosphere.(n1) Reauthorization of the Clean Water Act in 1987 added provisions to begin addressing these nonpoint sources. However, nonpoint pollution sources are more difficult to control and evaluate than point sources because they have many diffuse and widespread origins. The amount of pollution they deliver varies hour to hour and season to season, making it difficult to quantify the sources. The demanding steps of pinpointing and quantifying the sources are the first steps to effective control. Evaluation of the success of these provisions and of the general state of U.S. water resources is limited quantitatively because information is pieced together from a variety of sources, many of which are not designed to characterize nationwide water quality. State water-quality data have been inconsistent in types of information, analytical methods, and temporal and spatial scales.(n2) Therefore, still uncertain are the precise extent of water-quality problems, the nature and location of the severe problems, and the location of high-quality waters that need to be protected.(n3)

Although a complete national picture of water-quality problems is not yet a reality, the last decade of studies by the National Water-Quality Assessment (NAWQA) Program of the U.S. Geological Survey (USGS) has documented significant contaminant patterns in some of the United States's most important river basins and aquifers.(n4) These patterns have important implications for water-quality management and can be used to help guide and inform state and national water policies. Policies that take advantage of these science-based insights while continuing to fill information gaps wiPl help advance regulations and drinking-water standards and guidelines to reflect actual environmental patterns, identify key sources of nonpoint pollution, and improve investments in monitoring and management across the nation's diverse settings (see the box on this page).

An important general pattern emerging from the last decade of NAWQA studies is the prevalence of contaminants in streams and ground water. For example, at least one pesticide was found in more than 90 percent of water and fish samples collected from streams and in more than half of shallow wells sampled in agricultural and urban areas. Similarly, at least one volatile organic compound (VOC), widely used and produced in the manufacture of many products-including refrigerants, plastics, adhesives, paints, and petroleum-was detected in almost half the wells sampled in urban areas.

Concentrations of contaminants are generally at low levels-almost always below current EPA drinking-water standards. However, the risk to humans and the environment from present-day levels of contaminant exposure remains unclear. Exposure is complicated because individual compounds seldom occur alone. Streams and ground water in areas with significant agricultural or urban development almost always contain complex mixtures of VOCs, nutrients, pesticides, and their chemical breakdown products. For example, more than half of all stream samples contained five or more pesticides, and nearly one-quarter of all ground-water samples contained two or more. Chemical breakdown products, which can have similar or even greater toxicities than parent compounds, are often as common in the environment as parent compounds. Atrazine, a common farmland herbicide, and its breakdown product, desethylatrazine (DEA), were found together in about 35 percent of stream samples and about 25 percent of ground-water samples from agricultural areas (see Figure 1 on page 12).

Exposure to contaminants is also complicated by the lengthy periods of low concentrations that are often punctuated by brief seasonal pulses of much higher concentrations. In streams that drain much of the farmland throughout most of the United States, concentrations of nutrients and herbicides are highest during spring runoff following chemical applications. Exposure during extreme storm events can have overriding effects on the quality of streams and respective receiving bodies. For example, high flows in the Susquehanna, Potomac, and James Rivers during January 1996 carried almost half of the phosphorus and one-quarter of the nitrogen that is typically transported to the Chesapeake Bay in an average year.(n5)

Possible risks implied by these patterns of exposure have not been fully evaluated for several reasons. First, many contaminants and their breakdown products do not have drinking-water standards or guidelines. Water-quality standards and guidelines are generally maximum acceptable concentrations of contaminants for protecting humans, aquatic life, or wildlife. They are established by the United States and other nations, international organizations, and some states and tribes. For the U.S. effort, precedence was largely given to standards and guidelines established by EPA, although some states may have different standards and guidelines that take priority for the particular water bodies. Specifically, only about half the pesticides (46 of the 83 measured compounds) and VOCs (27 of 60 measured compounds) have current EPA standards. In addition, current standards do not yet address exposure to contaminant mixtures and the possibility that the presence of multiple compounds, even at low concentrations, may have adverse cumulative health effects. Furthermore, standards and guidelines are usually based on expected long-term exposure to constant concentrations rather than lengthy periods of low concentrations punctuated by brief pulses of high concentrations. And finally, various possible impacts on aquatic organisms, such as on reproductive, nervous, and immune systems, have not been tested. Many of the 20 most frequently detected pesticides in studies by NAWQA, although detected at relatively low concentrations, are suspected endocrine disrupters that have the potential to affect reproduction or development of aquatic organisms or wildlife by interfering with natural hormones. For example, recent studies show that concentrations of pesticides below current aquatic toxicity levels may affect the ratio of estrogen to testosterone in both male and female carp. The hormone ratio, which is sometimes used as an indicator of potential impacts on the endocrine system, was significantly lower at sites with the highest pesticide concentrations. At this point, the lower hormone ratios are not definitively associated with measurable declines in fish populations; however, they are a signal that further investigation is needed.(n6)

Effective water-resource management requires improved information on the actual patterns of contaminant occurrence and exposure. Water-quality standards and guidelines should reflect the most current knowledge as this information is developed and should address the many complexities of contaminant occurrence, such as mixtures, breakdown products, and seasonal pulses. Monitoring programs should also reflect existing environmental conditions, including seasonal variations and contaminant mixtures. Specifically, monitoring programs cannot rely on one or two samples per year and should be designed to simultaneously measure multiple compounds in one sample. In addition, the programs should also measure streamflow to make it possible to determine contaminant transport and to place the information in its hydrologic context.

Key Sources of Nonpoint Pollution

Patterns of occurrence provide new perspectives about key sources of nonpoint source contaminants. For example, urban areas, which cover less than 5 percent of land in the continental United States, traditionally have not been recognized as important contributors to nonpoint-source contamination, especially when compared with agricultural land, which covers more than 50 percent of the United States. However, various trace elements and some nutrients, pesticides, and VOCs commonly used around homes and gardens and in commercial and public areas occur widely throughout urban streams and ground water (see the box on page 13).(n7)

The types and concentrations of compounds found in certain bodies of water are closely linked to land use in surrounding areas. In fact, water in urban settings seems to have a characteristic chemical makeup or "signature" that is different than that found in agricultural or other settings. The chemical makeup correlates with the chemicals used in this setting or watershed. For example, some of the highest concentrations of phosphorus and insecticides, including those currently used (such as diazinon, carbaryl, and malathion) and those historically used (such as DDT (dichlorodiphenyltrichloroethane), dieldrin, and chlordane) were detected in urban streams, whereas some of the highest concentrations of nitrogen and herbicides, particularly those most heavily used (such as atrazine, metolachlor, alachlor, and cyanazine), were detected in streams and shallow ground water in agricultural areas. The direct connection between chemical use and contamination was demonstrated in some upper Midwest streams in 1994. These streams quickly showed increased acetochlor concentrations and decreased alachlor concentrations after acetochlor partially replaced alachlor as an herbicide used on corn (see Table 1 on page 14).

This linkage suggests that increased care in chemical use can go a long way toward improving water-quality conditions. Reducing the amount of chemicals used and applying these chemicals more efficiently are two of the most effective ways to reduce contaminant levels in both urban and agricultural settings. Unfortunately, current chemical-use information is generally insufficient-and in urban areas essentially unavailable-for local and regional water-resource management and decisionmaking. Improved tracking of chemical use is needed to definitively attribute specific pollutants to different sources in nonpoint runoff and support management actions.

Effective solutions also depend on the commitment of a multitude of individuals because nonpoint pollution results largely from everyday activities. This presents a challenge because the consequences of individual actions (such as application of chemicals on lawns and disposal of household chemicals) on the degradation of water resources are not well understood. In addition, in the agricultural community, there is some economic incentive to reduce chemical use and improve application efficiency. But, for individual homeowners who apply chemicals in much smaller quantities and less frequently, there is little economic incentive to optimize the use of chemicals. Excessive application more often than not has minimal economic impact to the individual homeowner, whereas the collective environmental impact can be large. Creative solutions are needed to help minimize negative impacts of chemical use, including increased communication to citizens and private industry on how certain choices can affect water resources and the environment.

Chemical use, land use, and population density are not the sole predictors of water quality. Natural features, such as geology, hydrology, and soils, and land practices, such as tile drainage and irrigation, govern vulnerability to contamination because they affect the movement of chemicals over land and into aquifers. Concentrations of contaminants can thereby vary significantly in different regions of the United States, or even locally within a basin, despite similar land-use settings and chemical use.

For example, NAWQA studies show some of the highest concentrations of nutrients and pesticides in ground water of sand and gravel or in karst aquifers (areas that underlie well-drained farmland). These natural geologic features readily transmit water and are common in the Central Valley of California and parts of the Northwest, Great Plains, and mid-Atlantic regions. In contrast, ground-water contaminants underlying farmland in parts of the upper Midwest are barely detectable, despite similar high rates of chemical use. This is partly because the ground water is "protected" by relatively impermeable and poorly drained soils and glacial till that cover much of the region. In addition, tile drains and ditches commonly provide quick pathways for chemical transport to streams, which minimize the downward movement of contaminants to ground water.

Local and regional vulnerability is also governed by naturally occurring toxic minerals. For example, concentrations of arsenic (a toxic mineral in certain rocks and soils) are usually highest in ground water in the West. Parts of the Midwest and Northeast also have elevated arsenic in ground water. Timely information on arsenic patterns is particularly important because EPA is considering reducing the maximum contaminant level allowed in drinking water (currently 50 parts per billion (ppb), which is equal to one drop in an Olympic-size swimming pool), citing risks for developing bladder and other cancers. EPA has proposed a new standard of 5 ppb. Slightly more than 13 percent of the 18,850 samples that were evaluated by the NAWQA program exceed the proposed drinking water standard, compared with less than 1 percent exceeding the current standard.(n8)

Hydrology and basin characteristics, including stream channel size, can also affect contaminant vulnerability by controlling the magnitude and timing of contaminant transport. For example, in the Mississippi Basin, closer proximity of nitrogen sources to large streams and rivers increases the ultimate transport of nutrients to the Gulf of Mexico. This is because nitrogen is not removed as readily in the large streams and rivers by natural processes as in the smaller tributaries and is, therefore, much more likely to reach a coastal area if it originates close to a large river.(n9) As a result, some watersheds in the Mississippi Basin are much more significant contributors of nitrogen to the Gulf of Mexico than others, despite similar nitrogen sources or similar distances from the gulf.

The critical role of natural features and their local and regional control of water quality raise issues about how best to manage chemical transport to water. Controlling nonpoint source pollution may be best achieved through targeted and thoughtful actions based on local and regional vulnerability rather than uniform treatment of contaminant sources. This approach requires combined knowledge of chemical use, contaminant occurrence, transport, and local and regional cause-and-effect controls. By linking the four, one can better anticipate sensitive areas of concern, set priorities in streams, aquifers, or watersheds that are most vulnerable to contamination, and increase the cost-effectiveness of policies designed to protect water resources in diverse settings.

The success of such an approach has been demonstrated in Washington State where an empirically based, quantitative vulnerability model was used to guide the best level of protection to critical aquifers, resulting in increased cost-effectiveness of monitoring. Specifically, the Washington State Department of Health, in concert with USGS, assessed the vulnerability of public water-supply wells to pesticide contamination based on geology, well depth, and land-use activities. USGS information on pesticide contamination at low levels of detection enabled the state health department to identify aquifers with low susceptibility to contamination and obtain waivers for quarterly monitoring required under the Safe Drinking Water Act. By using the information to meet EPA requirements for safe drinking water, Washington State was able to save at least $6 million in costly additional monitoring. This is an annual savings of as much as $70 per household tapping public supply wells that were granted full monitoring waivers.(n10)

The interconnections between living systems and the air, land, and water that support them can no longer be ignored. Unintended impacts on the environment can result because these natural compartments interact with one another; thus, improvements in one resource may adversely affect another. A policy must consider consequences far beyond the immediate impacts on the resources that it was designed to protect. For example, in the 1970s, methyl tert-butyl ether (MTBE) was not regarded as an environmental contaminant, but as a compound that could help improve air quality in many urban areas by oxygenating gasoline and allowing it to burn cleaner. Characteristics of MTBE, such as its high solubility in water and persistence in the subsurface, were not considered in determining such use. As a result, humans are now confronted with the unintended consequence of widespread, albeit low, MTBE contamination in much of the urban shallow ground water in the United States, often in close proximity to a large number of community water supplies. A recent preliminary analysis of about 26,000 community water supplies (CWS) indicates that approximately 9,000 CWS wells in 31 states have a leaking underground storage tank within 1 kilometer. Not all of these sites will be a significant source of MTBE to ground water and to the CWS wells. However, 9,000 is enough to indicate that the actual number of CWSs that may be ultimately affected should be identified.(n11)

Environmental water policies focus primarily on surface water and overlook the critical nonpoint contributions of contaminants to streams and coastal waters from ground water and the atmosphere. However, ground-water issues for water managers continue to grow in importance in many parts of the United States. For example, in the last decade, people have learned that more than half of the water and nutrients that enter the Chesapeake Bay first travel through the ground-water system.(n12) Consideration of ground-water contributions is needed in water-resource programs, such as in the increasingly significant state programs designed to establish total maximum daily loads (TMDLs) for impaired streams. Exclusion of ground water may prevent a full accounting of all available sources and could limit the effectiveness that TMDLs could have in future stream restoration and protection.

The atmosphere can also be a major source of contaminants. For example, as much as 25 percent of the nitrogen entering the Chesapeake Bay comes from the atmosphere.(n13) Almost every pesticide that has been investigated has been detected in air, rain, snow, or fog throughout the country at different times of the year.(n14) Consideration of atmospheric contributions is critical for effective management of water resources, and because contributions can cross state boundaries, full implementation of strategies often requires regional or national involvement.

Long-range interstate transport of atmospheric contaminants is not the only reason that local and state environmental controls and management by themselves are not sufficient to address water issues. Rarely do decisions on the effects of land use or human actions in individual watersheds consider the cumulative or overall impact on the quality of the downstream resource and receiving coastal water. For example, the Gulf of Mexico experiences low oxygen (hypoxia) largely as a result of increased nutrient concentrations that contribute to excessive growth of algae and other nuisance plants.(n15) Studies conducted in the last decade show that a considerable amount of total nitrogen originates from watersheds in the Mississippi River Basin very distant from the Gulf of Mexico (see Figure 2 on page 16).

Policies must therefore take into account the local watersheds and the larger water resource network that connects them. This requires decisionmaking based on a rational geographic or hydrologic structure that recognizes the multiple orders of nested, progressively larger watersheds, as opposed to one exclusively based on political boundaries. Such an approach would accommodate local watershed needs and management perspectives, in combination with a broader systems perspective that targets priority watersheds within the larger network for action and incorporates out-of-boundary issues, such as atmospheric and regional ground-water contributions. This approach requires a new way of doing business-one that can circumvent institutional barriers and governmental divisions that currently prevail at all levels and foster collaboration among individuals, private and nongovernment organizations, and local, state, interstate, and federal institutions.

As we enter the new millennium, a complete picture of national water quality is still out of grasp. During the next 5 to 10 years, as national water-quality programs mature and water-quality information is better integrated among government and nongovernment organizations, the United States will be better positioned to assess whether its waters and aquatic life are improving or further degrading. In the meantime, the nation can move forward with confidence, based on comparable and consistent quantitative assessments across the United States, on some critical next steps to manage its waters and improve today's water-resource strategies and policies. For these strategies to be effective, they should

• reflect actual environmental conditions, including effects of contaminant mixtures, breakdown products, and seasonal high concentration pulses;
• account for local and regional natural controls that govern vulnerability to contamination;
• provide checks on single-resource protection to avoid unintended environmental impacts and to fully account for contaminant sources, such as from ground water and the atmosphere; and
• include an effective mix of national protection strategies, local watershed efforts, and state water management.

The scientific understanding of the nation's ever-changing water-quality answers several important questions and raises many new questions at the same time. It is important that the nation continue to move forward, bringing these meaningful water-quality insights to the forefront of ongoing state and national policy debates and using today's knowledge to make the best possible decisions today, while also striving to improve the data and scientific understanding needed for future water-quality decisions.

• Improved information. Effective water-resource management requires improved information on the actual patterns of contaminant occurrence and exposure. Water-quality standards and guidelines need to address the many complexities in contaminant occurrence, such as mixtures, breakdown products, and seasonal pulses. Monitoring programs should also reflect existing environmental conditions. They cannot rely on one or two samples per year and should be designed to simultaneously measure streamflow and multiple compounds in one sample.
• Chemical use and application. Peducing chemical use and improving application efficiency are undoubtedly some of the most effective ways to reduce contaminant levels in streams and ground water in urban and agricultural settings. Improved tracking of chemical use is needed to definitively attribute specific pollutants to different sources in nonpoint runoff and to support management actions.
• Targeted actions. Not all water resources are at equal risk of contamination. Controlling nonpoint source pollution may, therefore, require targeted and thoughtful actions based on local and regional vulnerability rather than uniform treatment of contaminant sources. This approach requires combined knowledge of chemical use, contaminant occurrence, and local and regional cause-and-effect controls by natural features.
• A multiresource approach. Environmental compartments interact with one another. A policy must fully account for all contaminant sources and must consider consequences that are far beyond the immediate impacts on the resources it was designed to protect.
• Local, state, and national collaboration. Local watershed management or contaminant controls by themselves are not sufficient to address critical nonpoint source issues, such as in the Gulf of Mexico. Policies must take into account local watersheds and the larger water resource networks that connect them, resulting in an effective mix of national protection strategies and priorities with local and state watershed management. Success depends on collaboration among local, state, interstate, and international institutions.

National Water-Quality Assessment (NAWQA) sampling of streams and shallow ground water in urban areas shows that

• Concentrations of total phosphorus are generally higher in urban streams than in other settings, commonly exceeding the U.S. Environmental Protection Agency's desired goal to control excessive plant and algae growth.
• Insecticides occur at higher frequencies, and usually at higher concentrations, in urban streams than in agricultural streams. Most common are diazinon, carbaryl, chlorpyrifos, and malathion.
• Urban streams have the highest frequencies of occurrence of DDT (dichlorodiphenyltrichloroethane), chlordane, and dieldrin in fish and sediment and the highest concentrations of chlordane and dieldrin. DDT is an insecticide that was commonly used in the United States until the early 1970s to control mosquitoes and other insects. Chlordane and aldrin (the parent compound that breaks down to dieldrin) were used widely until the late 1980s to control termites. Despite downward trends in some areas, these persistent organochlorine insecticides are still found at elevated levels in bed sediment and fish in urban streams throughout the United States.
• Volatile organic compounds (VOCs), used in plastics, cleaning solvents, gasoline, and industrial operations, occur widely in urban ground water throughout the United States. At least one VOC was detected in 47 percent of wells sampled in urban areas. The four most frequently detected of the 60 measured VOC compounds are the industrial solvents trichloroethene and tetrachloroethene, the gasoline additive methyl tert-butyl ether (MTBE), and trichloromethane (also known as chloroform), which is a solvent and a byproduct of the disinfection of drinking water.
• Concentrations of selected trace elements are also elevated in populated urban settings, most likely caused by emissions from industrial and municipal activities and the widespread use of motor vehicles. For example, streambed-sediment and reservoir-sediment samples collected from the Chattahoochee River Basin and analyzed for total lead and zinc concentrations indicate that population density, which is strongly related to traffic density, is a predictor of lead and zinc concentrations in the environment.

Legend for Chart:

B - Streams Urban areas
C - Streams Agricultural areas
D - Streams Undeveloped areas
E - Shallow Ground Water Urban areas
F - Shallow Ground Water Agricultural areas

             A                     B             C           D
                                   E             F

Nitrogen                      medium        medium-high   low
                              medium        high

Phosphorus                    medium-high   medium-high   low
                              low           low

Herbicides                    medium        low-high      no data
                              medium        medium-high

Currently used insecticides   medium-high   low-medium    no data
                              low-medium    low-medium

Historically used             medium-high   low-high      low
insecticides                  low-high      low-high

NOTE: Relative levels of nutrient and pesticide contamination are
closely linked to land use and to the amounts and types of
chemicals used in each setting.

SOURCE: U.S. Geological Survey, "The Quality of Our Nation's
Waters--Nutrients and Pesticides," U.S. Geological Survey
Circular 1225 (Reston, Va., 1999), 6.

Legend for Chart:

B - Fish
C - Streams
D - Shallow ground water
E - Major rivers
F - Major aquifers

           A          B    C    D      E    F

Agricultural areas    85   92   59
Urban areas          100   99   49
Mixed land use        96              100   33

SOURCE: U.S. Geological Survey, "The Quality of Our Nation's
Waters--Nutrients and Pesticides," U.S. Geological Survey 1225
(Reston, VA., 1999), 58.

MAP: Figure 2. Amount of total nitrogen originating in watersheds

(n1.) Clean Water Action Plan-The First Year. The Future, a report collectively written by the U.S. Army Corps of Engineers, Department of the Interior, EPA, National Oceanic and Atmospheric Administration, and U.S. Department of Agriculture, (Washington, D.C., 1999), 3.

(n2.) D. S. Knopman and R. A. Smith, "Twenty Years of the Clean Water Act: Has U.S. Water Quality Improved?" Environment, January/February 1993, 16-20, 34-41.

(n3.) U.S. General Accounting Office, Report to the Chairman, Subcommittee on Water Resources and Environment: Water Quality-Key EPA and State Decisions Limited by Inconsistent and Incomplete Data, GAO/RCED-00-54 (Washington, D.C., 2000), 6.

(n4.) In 1991, the U.S. Congress appropriated funds for the U.S. Geological Survey (USGS) to begin the National Water-Quality Assessment (NAWQA) Program to better understand the spatial extent of water quality, how water quality changes with time, and how human activities and natural factors affect water quality across the United States. The program began investigations in 20 major river basins and aquifer systems in 1991 and phased in work in more than 30 additional basins by 1997. Insights presented here are based on results from the first set of 20 investigations. All references to water-quality conditions in these basins are based on data and reports available on the USGS-NAWQA web site, http://water.usgs.gov/nawqa. The primary reference is "The Quality of Our Nation's Waters-Nutrients and Pesticides," U.S. Geological Survey Circular 1225 (Reston, Va., 1999).

(n5.) L. D. Zynjuk and B. F. Majedi, "January 1996 Floods Deliver Large Loads of Nutrients and Sediment to the Chesapeake Bay," U.S. Geological Survey Fact Sheet FS-140-96 (Baltimore, Md., 1996).

(n6.) L. L. Goodbred et al., Reconnaissance of 17B-Estradiol, 11-Ketotestosterone, Vitellogenin, and Gonad Histopathology in Common Carp of United States Streams-Potential for Contaminant-Induced Endocrine Disruption, U.S. Geological Survey Open-File Report 96-627 (Sacramento, Calif., 1996).

(n7.) "Urban" primarily represents residential land use, typically with low to medium population densities, as reported in K. J. Hitt, Refining 1970s Land-Use Data with 1990 Population Data to Indicate New Residential Development, U.S. Geological Survey Water-Resources Investigations Report 94-4250 (Reston, Va., 1994).

(n8.) A. H. Welch, S. A. Watkins, D. R. Helsel, and M. J. Focazio, "Arsenic in Ground-Water Resources of the United States," U.S. Geological Survey Fact Sheet FS-063-00 (Denver, Colo., 2000).

(n9.) R. B. Alexander, R. A. Smith, and G. E. Schwarz, "Effect of Stream Channel Size on the Delivery of Nitrogen to the Gulf of Mexico," Nature, 17 February 2000, 761.

(n10.) S. J. Ryker and A. K. Williamson, "Pesticides in Public Supply Wells of Washington State," U.S. Geological Survey Fact Sheet FS-122-96 (Tacoma, Washington, 1996).

(n11.) R. Johnson, "MTBE-To What Extent Will Past Releases Contaminate Community Water Supply Wells?" Environmental Science & Technology, 1 May 2000, 7.

(n12.) L. J. Bachman, B. D. Lindsey, J. Brakebill, and D. S. Powars, Ground-Water Discharge and Base Flow Nitrate Loads on Nontidal Streams, and Their Relation to a Hydrogeomorphic Classification of the Chesapeake Bay Watershed, Middle Atlantic Coast, U.S. Geological Survey Water-Resources Investigations Report 98-4059 (Baltimore, Md., 1998).

(n13.) D. C. Fisher and M. Oppenheimer, "Atmospheric Nitrogen Deposition and the Chesapeake Bay Estuary," Ambio 20, no. 3-4 (1991): 102-08.

(n14.) U.S. Geological Survey, "Pesticides in the Atmosphere," U.S. Geological Survey Fact Sheet FS-152-95 (Sacramento, Calif., 1995).

(n15.) N. N. Rabulais et al., "Nutrient Changes in the Mississippi River and System Responses on the Adjacent Continental Shelf," Estuaries 19 (1996): 386-407.

~~~~~~~~

By Robert M. Hirsch; Timothy L. Miller and Pixie A. Hamilton

Robert M. Hirsch is associate director for water at the U.S. Geological Survey (USGS). Timothy L. Miller is chief of the National Water-Quality Assessment Program at USGS. Pixie A. Hamilton is a USGS staff hydrologist and communications specialist. The authors may be contacted through Hirsch at the U.S. Geological Survey, 409 National Center, Reston, VA 20192 (telephone: 703-648-5215; e-mail: rhirsch@usgs.gov). This article is largely based on a report (USGS Circular 1225) from the USGS National Water-Quality Assessment Program, available on-line at htpp://water.usgs.gov/pubs/circ/circ1225. Visit the USGS web site at http://www.usgs.gov/ to directly access data, maps, and reports. This article is in the public domain.