By Janet M. Denis and Joel D. Blomquist
The Great Valley Carbonate subunit of the Potomac River Basin is a major agricultural region, where high nitrate concentrations in streams from agriculture and other land uses are a concern. Important agricultural sources of nitrogen include commercial fertilizer and manure. Nitrate concentrations in 25 streams in the Great Valley Carbonate subunit of the Potomac River Basin ranged from 0.78 to 9.0 mg/L (milligrams per liter) as nitrogen because of variations in geology and land use in each basin. Agricultural land use covers about 75 percent of the Great Valley Carbonate subunit. The median nitrate concentration in cropland (6.6 mg/L) is significantly greater than the median nitrate concentration in pasture (2.6 mg/L). Nitrate concentrations in streams increase significantly from the southern part of the Great Valley Carbonate subunit in Virginia to the northern part in Maryland and Pennsylvania, and reflect the percentage of land used for crop production in a county. Nitrate concentrations in all 25 streams were below the U.S. Environmental Protection Agency (USEPA) maximum contaminant level of 10 mg/L for drinking water.
The Great Valley Carbonate subunit covers 15 percent of the Potomac River Basin and is a major agricultural region transecting parts of Maryland, Pennsylvania, Virginia, and West Virginia. High nitrate concentrations in streams from agriculture and other land uses are a concern through out this region. Nitrate can cause human health problems, degrade local stream conditions, and cause nitrogen enrichment downstream in the Potomac River and Chesapeake Bay. Twenty-five stream sites were sampled in September 1993 to assess the occurrence of nitrate in streams throughout the Great Valley Carbonate subunit as part of the U.S. Geological Survey's National Water Quality Assessment program (NAWQA). The objective of the NAWQA program is to determine the status and trends in water quality of streams and rivers throughout the Nation. The Potomac River Basin is one of 60 NAWQA study units assessing water-quality conditions in major rivers and aquifer systems,
The Great Valley, a subprovince of the Valley and Ridge physiographic province, is bounded by the Blue Ridge Mountains on the east, and Great North Mountain on the west, and is interrupted by Massanutten Mountain in Virginia. For water-quality assessment purposes, the NAWQA program's Potomac River Basin study unit subdivided the Great Valley into two subunits with two distinctive rock types: carbonate (limestone and dolomite rock), and noncarbonate (shale and sandstone rock) (fig. 1). The carbonate subunit covers about 2,220 mi2 (square miles) or 70 percent of the Great Valley, and is characterized by rolling hills, with caverns and sinkholes present in the valleys. The major streams in the Great Valley include the North Fork Shenandoah River, South Fork Shenandoah River, the mainstem Shenandoah River, Opequon Creek, Conococheague Creek, and Antietam Creek (fig. 1).
Land use in the carbonate subunit is predominantly agricultural (75 percent). Forests cover 15 percent of the subunit and are located primarily on hillsides and along some stream buffers. Urban areas are located throughout the Great Valley, but cover only 10 percent of the subunit (fig. 2).
Between September 7 and 15, 1993, water-quality samples were collected from 25 streams in the Great Valley Carbonate subunit during base-flow conditions (table 1). The 25 sampling sites were selected on the basis of four criteria to permit comparison among sites. All sampling sites were located within the Great Valley Carbonate subunit and were distributed across the subunit for spatial representation. Sites were selected with basin areas of approximately 10 mi2. The basins were selected to represent agricultural, urban, and forested land uses, which are present throughout the subunit. Only sites unaffected by known point sources were selected so the effects of nonpoint sources from agricultural and urban areas could be assessed. Water samples were collected and analyzed for five forms of nitrogen including nitrite plus nitrate and nitrite, and were reported in milligrams per liter (mg/L) as nitrogen. Concentrations of nitrite were negligible at all sites. Nitrate concentration is calculated by subtracting dissolved nitrite as nitrogen from dissolved nitrite plus nitrate as nitrogen.
Figure 1. The Great Valley Carbonate subunit consists of limestone and dolomite rock that covers about 70 percent of the Great Valley. Twenty-five stream sites were sampled in the carbonate subunit between September 7 and 15, 1993, to assess the occurrence of nitrate in small streams as part of the Potomac River Basin NAWQA study.
Base-flow conditions result when ground-water discharge is the only contributing source of water to streams. Ground-water discharge is generally greater in streams underlain by carbonate rock because ground-water storage and transmissivity are usually greater in carbonate rock than in the surrounding noncarbonate rock. Ground-water inflow has been found to contribute about two-thirds of the annual streamflow in basins underlain by carbonate rock in Washington County, Md. (Duigon and Dine, 1991). Although the 25 stream sites were sampled in basins underlain by carbonate rock, many of the streams are underlain by noncarbonate rock along the fringes of the basin boundaries. Long-term streamflow data from Marsh Run at Grimes, Md. (site 19) indicate that the streamflow conditions during the sampling period were similar to 7-day minimum flows that recur about every 2 years. The streamflow during sampling was 3.2 ft /s (cubic feet per second), only 6 percent greater than typical low, base-flow conditions.
Figure 2. Land use in the Great Valley Carbonate subunit is about 75-percent agricultural, 15-percent forested, and 10-percent urban. Agricultural land uses appear to affect nitrate concentrations in all 25 streams sampled in the Great Valley. (1973 land-use data from U.S. Geological Survey, Geographic Information Retrieval and Analysis System).
Table 1. Selected Characteristics of 25 stream sites in the Great Valley Carbonate Subunit.
NITRATE CONCENTRATIONS IN STREAMS
Nitrate concentrations in streams in the Great Valley Carbonate subunit ranged widely because of variations in geology and land use in each basin. Nitrate concentrations ranged from 0.78 to 9.0 mg/L (milligrams per liter) as nitrogen with a median concentration of 3.6 mg/L (fig. 3). Agricultural land uses (cropland and pasture) appear to affect nitrate concentrations in all 25 streams. Streams draining predominantly cropland basins have significantly higher concentrations of nitrate than streams draining predominantly pasture basins. Nitrate concentrations show a strong geographic pattern in the subunit because of variations in cropland and pasture.
Figure 3. Nitrate concentrations in streams ranged widely in the Great Valley Carbonate subunit during base-flow conditions. The median nitrate concentration in cropland (6.6 milligrams per liter) is significantly greater than the median concentration in pasture (2.6 milligrams per liter). All 25 stream nitrate concentrations were below the U.S. Environmental Protection Agency maximum contaminant level for drinking water.
The wide range in nitrate concentrations in the Great Valley is attributed to the variations in the type of rock and land use in each basin. In general, basins with higher percentages of carbonate rock and cropland also had higher nitrate concentrations, whereas basins with higher percentages of noncarbonate rock were predominantly forested and had the lowest nitrate concentrations. For example, the lowest nitrate concentration (0.78 mg/L) was detected at Happy Creek at Crosby Stadium at Front Royal, Va. (site 11), where 89 percent of the basin is underlain by crystalline rock and approximately 51 percent is forested. This sampling site is located in the carbonate subunit in an urban setting. Toms Brook at Toms Brook, Va. (site 10), is underlain by about 50-percent carbonate rock and is predominantly forested. The nitrate concentration at Toms Brook (2.1 mg/L) was less than nitrate concentrations at nearby sites underlain by greater percentages of carbonate rock. In contrast to these two streams, the highest nitrate concentration detected in the Great Valley Carbonate subunit of the Potomac River Basin was 9.0 mg/L at Falling Spring at Chambersburg, Pa. (site 25), a basin underlain entirely by carbonate rock with approximately 51-percent cropland.
Relation of Nitrate Concentrations to Land Use
Nitrate concentrations are related to the predominant land use of the basins contributing to each site. High concentrations of nitrate were detected at sites that have a large percentage of cropland in the basin. Cropland and pasture are the most widespread land uses in the subunit and may affect nitrate concentrations in most basins. Nitrate concentrations in streams draining pastured basins ranged from 1.0 to 3.8 mg/L and were less than concentrations in cropland basins which ranged from 4.7 to 9.0 mg/L (fig. 3). Nitrate concentrations ranged from 2.0 to 6.1 mg/L in four urban basins and 0.78 to 3.6 mg/L in three forested basins.
The basins were classified by the predominant land use: pasture, cropland, urban, or forest (table 1). Land-use data were compiled from four sources: (1) U.S. Geological Survey Geographic Information Retrieval and Analysis System; (2) Maryland Office of Planning; (3) Virginia Geographic Information System; and (4) Natural Resources Conservation Service, Jefferson County, W.Va., field office. Land-use classifications for cropland and pasture were unavailable for the entire Great Valley Carbonate subunit; therefore, aerial photographs from the National Aerial Photography Program were used to classify land use in parts of eight basins.
Land used for crop production had a greater effect on nitrate in the 25 basins than other land uses. The highest nitrate concentrations in streams measured in the Great Valley Carbonate subunit were in cropland basins (fig. 3). The median nitrate concentration in cropland (6.6 mg/L) is significantly greater than the median concentration in pasture (2.6 mg/L). A Kruskall-Wallis test shows that the probability of finding similar concentrations in cropland and pasture is small (probability = 0.0004). Sources of nitrogen to cropland and pasture differ greatly as commercial fertilizers and manure are applied to cropland in greater amounts than pasture.
Although urban land uses are predominant in four basins within the subunit, agricultural land uses appear to be the overriding cause of high nitrate concentrations. In urban basins where pasture is dominant, nitrate concentrations in streams are similar to concentrations found in streams in pastured basins. For example, Lewis Creek at Staunton, Va. (site 2), is 49-percent urban and about 29-percent pasture. The nitrate concentration at this site (2.0 mg/L) is less than the median concentration (2.6 mg/L) found at pastured sites. Conversely, nitrate concentrations in urban basins where cropland is dominant are similar to concentrations in cropland basins. The nitrate concentration at Hamilton Creek at Hagerstown, Md., (site 20), is 6.1 mg/L, and is similar to concentrations found in cropland basins. Land use at site 20 is three-fourths urban and only about one-fourth cropland.
Geographic Distribution of Nitrate Concentrations
Nitrate concentrations in streams increase significantly from the southern part of the Great Valley Carbonate subunit in Virginia to the northern part in Maryland and Pennsylvania (fig. 4). This geographic pattern is related to variations in agricultural land uses within the subunit. The distribution of cropland in the Great Valley is shown in figure 4. The percentage of area devoted to cropland ranges from less than 28 percent in Augusta and Warren Counties, Va., to greater than 57 percent in Washington County, Md., and Franklin County, Pa. These cropland percentages were obtained from the U.S. Department of Commerce, 1989 and Census of Agriculture, 1987.
Figure 4. Nitrate concentrations in streams increase significantly from the southern part of the Great Valley Carbonate subunit in Virginia to the northern part in Maryland and Pennsylvania, and reflect the patterns of the percentage of land used for crop production in a county.
A strong correlation was determined between the percentage of cropland and concentration of nitrate. The geographic distribution of nitrate concentrations reflects the percentage of land used for crop production in a county. In the southern part of the study area, 11 sites have nitrate concentrations less than 3 mg/L. Moderately high nitrate concentrations ranging from 3.0 to 6.0 mg/L were found at seven sites in the north-central part of the valley. Nitrate concentrations higher than 6.0 mg/L were found at seven sites in the northern part of the Great Valley. The highest nitrate concentration was found at Falling Spring at Chambersburg, Pa. (site 25), in Franklin County, Pa.
NITRATE LOADS AND YIELDS IN STREAMS
Nitrate loads and yields in the Great Valley Carbonate subunit during base-flow conditions are important due to high nitrate concentrations and proportionally high streamflow in this part of the Potomac River Basin. During base flow, the 25 streams carried an instantaneous nitrate load of 3,212 lb/d (pounds per day). This load is equal to 28 percent of the nitrate load measured in the Potomac River at Washington, D.C. on September 14, 1993, yet the 25 sites drain only about 2 percent of the Upper Potomac River Basin. Some of the nitrate measured in these streams probably never reaches Washington, D.C., because of denitrification and biological uptake; nevertheless it is clear that streams in the Great Valley Carbonate subunit contribute a disproportionately high share of the nitrate measured in the Potomac River at Washington, D.C. The amount of nitrate transported varied from basin to basin and was affected by basin size, nitrate concentration, and stream-flow. Nitrate yields, expressed as loads divided by drainage area, are used to assess differences in nitrate loads among the basins of different sizes. The geographic distribution of nitrate yields in the subunit is similar to the distribution of nitrate concentrations with higher yields in the northern part of the subunit where cropland is predominant.
Nitrate yields ranged wisely in the subunit (0.04 to 171 (lb/d)/mi2), because of variations in nitrate concentrations and streamflow. The distribution of nitrate yields at the 25 sites in six groups is shown in figure 5. Sixteen of the 25 sites had nitrate yields less than 8.0 (lb/d)/mi2, and all sites except for Falling Spring (site 25) had nitrate yields less than 20 (lb/d)/ mi2. High streamflow from springs contributed to large nitrate yield at two sites. The greatest streamflow (33 ft3/s) and nitrate concentration (9.0 mg/L) were detected at Falling Spring at Chambersburg, Pa. (site 25), and produced the greatest nitrate yield of the 25 sites. The combined effects of high stream-flow (16 ft3/s) and low nitrate concentration (1.0 mg/L) at Mossy Creek near Spring Creek, Va. (site 5), produced a nitrate yield of about 5.7 (lb/d)/mi2. Springs are common in the subunit and may obtain water through rock fractures, solution channels, and converging ground-water-flow paths that extend beyond the surface drainage basin boundary; therefore, streams with large springflow may transport nitrate from a larger undefined basin area, and may have nitrate yields reflecting a much larger basin size. Low streamflow at Conococheague Creek Tributary at Fayetteville, Pa. (site 24), produced a nitrate yield of 1.5 (lb/d)/mi2, although this stream had the second highest nitrate concentration of 7.0 mg/L.
POTENTIAL EFFECTS OF NITRATE IN STREAMS
Nitrate concentrations in the 25 streams sampled in the Great Valley Carbonate subunit were all below the U.S. Environmental Protection Agency (USEPA) maximum contaminant level for drinking water. Nitrate concentrations in excess of 10 mg/L are known to cause methemogiobinemia or "blue baby" syndrome in infants. Because of this and other potential human health risks, the USEPA has set the maximum contaminant level for nitrate at 10 mg/L as nitrogen. Surface water is not regulated in this manner because surface water is not usually consumed without treatment.
Figure 5. Base-flow nitrate yields were less than 20 pounds per day per square mile at 24 of 25 sites in the Great Valley Carbonate subunit. Falling Spring at Chambersburg, Pa. had the greatest nitrate yield due to the combined effects of the highest streamflow and highest nitrate concentration.
High nitrate concentrations can cause eutrophication, a condition that can affect aquatic life in the local streams as well as in the Potomac River and the Chesapeake Bay. Nonpoint sources of nutrients from agricultural land uses in the Great Valley Carbonate subunit can potentially supply large quantities of nutrients through streamflow and ground-water inflow resulting in large nutrient accumulations downstream and extensive eutrophication (Johnson and others, 1993). Eutrophication is caused by excessive nutrients, primarily nitrate and phosphate, that foster excessive algal growth, and the result is a large excess of algal biomass that will eventually die and sink to the bottom of the river or estuary. These events are often referred to as "algal blooms." Subsequent algal decay causes the destruction of habitat and the depletion of dissolved oxygen, which usually results in the disappearance of intolerant aquatic insect species and fish.
During base-flow conditions, the Great Valley Carbonate subunit contributes a disproportionately large amount of nitrogen to the Potomac River. The 25 streams measured in this study drain about 2 percent of the Upper Potomac River Basin, yet carried a base-flow nitrate load equal to 28 percent of the nitrate load in the Potomac River at Washington, D.C. In order to effectively reduce base-flow nitrate loads to the Potomac River and Chesapeake Bay, management programs must continue to address agricultural land-use practices in the Great Valley Carbonate subunit.
The authors would like to thank the report team: James M. Gerhart, Robert A. Hainly, Gary T. Fisher, Eugene C. Hayes, G. Jean Hyatt, Valerie M. Gaine, and Jonathan J. Dillow. The authors also want to thank John W. Brakebill for drafting the figures and editing the text. The team provided valuable data sources in a timely fashion, technical and editorial suggestions, and insight during the planning and reviewing phases of this report.
For further information contact:
Phone: (410) 238-4200