Environmental setting described in this report includes aspects of both natural and cultural settings of the lower Tennessee River Basin and their effects on water quality. Natural setting affects the distribution of physical properties, such as temperature, dissolved oxygen, and pH, and the occurrence and distribution of major and some trace inorganic constituents of surface and ground water. Cultural setting influences the quality of water that results from natural processes. Modification of the hydrologic system and other activities, such as dam construction, cultivation of land, and mining, increase the mobility of naturally occurring water-quality constituents and affect water quality. Human activities also introduce naturally occurring compounds where they are not normally present, or in amounts greater than would be present naturally (fertilizer use), and introduce compounds that do not occur naturally anywhere in the environment (synthetic organic compounds such as pesticides). In order to assess the effect of human-related activities on water quality, an estimate of the natural variability in water quality is needed.
Geology and physiography were the primary natural factors used to subdivide the lower Tennessee River Basin into nine subunits that represent areas of relative lithologic, and physical or geomorphic homogeneity (fig. 2). Subdivision of the study unit based on these natural factors provides a framework in which natural variability in water quality can be quantified and the effects of cultural factors on water quality can be assessed. This framework is similar to the ecoregion framework used by State and Federal agencies for evaluating and managing environmental and water resources (Bailey and others, 1994). Level III ecoregion boundaries were used to define the boundaries for the subunits and level IV ecoregion boundaries were used in Tennessee (Griffith and others, 1997). Geologic contacts were used instead of level IV ecoregion boundaries to define subunit boundaries in northern Alabama, Georgia, and Mississippi, because level IV ecoregion boundary delineations were unavailable in these areas. However, this substitution for the boundaries does not affect regional consistency of the delineation, because the boundaries for the level IV ecoregions in the lower Tennessee River Basin generally correspond to the boundaries of geologic units.
Subunits delineated in the lower Tennessee River Basin are the Coastal Plain, Transition, Western Highland Rim, Outer Nashville Basin, Inner Nashville Basin, Eastern Highland Rim, Plateau Escarpment and Valleys, Cumberland Plateau, and Valley and Ridge (fig. 2). The natural and cultural settings within these subunits is described in the following sections.
Geology and soils are the primary factors of the natural setting that affect the chemical composition of ground and surface water. The effect of the water-rock interaction on water quality or the amount of solids that dissolve from the rock is dependent on rock type and length of time water is in contact with rock. Soil properties, climate, and ground- and surface-water hydrology affect the movement of water in a watershed, and therefore affect the water-rock interaction.
Carbonate rocks underlie much of the lower Tennessee River Basin, but unconsolidated sediments and siliciclastic rocks crop out in the extreme western part and in the eastern parts and extreme southern parts of the study unit, respectively (fig. 3). Ordovician-age carbonates generally crop out near the center of the study unit (fig. 3), and are predominantly limestone with some thin shaly beds and bentonite layers. Phosphatic limestones are present in three Ordovician-age formations: the Hermitage Formation, the Bigby Limestone, and the Leipers Formation. Phosphate has been mined commercially in the Bigby Limestone and the Leipers Formation in the Nashville Basin (Miller, 1974). Ordovician phosphatic limestones and residuum in the study unit generally crop out in a north-south trending band in Maury and Giles Counties, Tennessee, and Limestone County, Alabama (fig. 4).
Most of the Ordovician-age rocks crop out in an area that is a topographic basin (fig. 3), but was originally a structural dome (Nashville Dome). Uplift and fracturing of this area during the late Paleozoic Era accelerated erosion of the younger, more resistant rocks overlying the Ordovician carbonates, and resulted in subsequent erosion of the Ordovician carbonates creating a topographic basin (fig. 3). The Ordovician carbonates are undeformed and relatively flat-lying to gently dipping; however, stresses related to doming formed joints in most of the area. A thin outcrop of Ordovician carbonate is present in the Sequatchie Valley, an eroded anticlinal valley (fig. 3), and minor outcrops are present in the extreme eastern part of the study unit. The Sequatchie Valley represents the western extent of deformation and folding associated with the Appalachian Mountains. The Ordovician carbonates are characterized by karst landforms, such as sinkholes, caves, and disappearing streams. These rocks are the predominant geologic units in the Outer and Inner Nashville Basin.
Devonian- and Silurian-age rocks crop out in a limited area in the western part of the study unit (fig. 3) and are predominantly carbonates. Phosphatic limestones and phosphatic residuum also are present in areas underlain by Devonian- and Silurian-age rocks (fig. 4); however, these deposits are generally scattered and isolated (Smith and Whitlatch, 1940). These rocks are relatively flat-lying to gently dipping away from the Nashville Dome. The Chattanooga Shale is the uppermost unit of the Devonian-age rocks (fig. 3). These rocks crop out primarily in the Western Highland Rim, with minor outcrops also in the Transition and Coastal Plain.
Mississippian-age carbonate rocks are the most areally extensive geologic unit in the study unit (fig. 3). The Mississippian-age carbonates are predominantly limestone, but some shaly beds and chert interbeds and nodules also are present. The amount of insoluble material (clay and chert) in the Mississippian-age carbonates is greater than that in the Ordovician-age carbonates, resulting in the development of a thicker regolith or residuum above bedrock than is present above the Ordovician-age carbonates. The Mississippian-age carbonates are relatively flat-lying to gently dipping away from the Nashville Dome, except in the areas near the Sequatchie Valley where folding has occurred. Parts of the study unit underlain by Mississippian-age carbonates are characterized by karst landforms. The Eastern and Western Highland Rim are underlain by these rocks, and these rocks outcrop in the valleys of the Plateau Escarpment and Valleys and Transition.
Pennsylvanian-age rocks located in the eastern and southern part of the study unit (fig. 3) generally are siliciclastic with some coal beds present. These rocks are characterized by repeating sequences of sandstone, shale, and coal. The rocks dip gently [20 to 50 feet per mile (ft/mi)] to the east (Brahana, Macy, and others, 1986). Generally, the Pennsylvanian-age rocks are undeformed, but fractures are present in many parts of these geologic units. These rocks are the predominant lithology in the Cumberland Plateau and also cap the hills in the Plateau Escarpment and Valleys.
Unconsolidated sand, silt, and clay of predominantly Late Cretaceous to Tertiary age are located at the western edge of the study unit (fig. 3). These sediments dip to the west at a rate of about 35 ft/mi (Brahana, Mulderink, and others, 1986) and thicken in the direction of dip. Quaternary alluvial deposits and loess are present locally. These sediments are primarily in the Coastal Plain and the Transition.
Soils develop from physical and chemical weathering of the underlying rocks or sediments: sandy, well-drained soils tend to develop in areas where sand or sandstone are the predominant lithology and clayey, poorly drained soils develop in areas where clays, shale, or shaly limestone are present. Other factors that affect the development of soils are topographic relief, climate, and plant growth.
Soil type has implications for ground- and surface-water quality. Well-drained, coarse textured soils have a higher relative infiltration rate and transmit precipitation or recharge to aquifers more readily than poorly drained soils. A higher infiltration rate increases the probability of contamination of ground-water aquifers from surface sources. A slower infiltration rate increases the potential for precipitation to become surface runoff, and increases the probability of surface contaminant transport to surface-water bodies.
About 60 percent of the area of the lower Tennessee River Basin is covered with soils with moderate infiltration rates. These soils overlie primarily the Ordovician- and Mississippian-age carbonate rocks and generally are in the Outer Nashville Basin, Western Highland Rim, and Transition subunits (fig. 5). Soils with slow infiltration rates cover about 38 percent of the study unit. These soils primarily overlie the sandstone and unconsolidated sand, but also overlie carbonate rock in the Inner Nashville Basin, Plateau Escarpment and Valleys, and Cumberland Plateau. The Eastern Highland Rim and Coastal Plain have a mixture of soils with moderate to slow infiltration rates, and in some places very slow infiltration rates, which cover about 2 percent of the area. Soils with high infiltration rates are present in less than 1 percent of the area and are located at the northern and eastern parts of the area in the Valley and Ridge and the Western Highland Rim. The remainder of the study unit is open water bodies.
Other properties of soil that have implications for water quality are the organic content of the soil, which affects the transport of pesticides and other organic compounds. Soil slope and erosion potential affects the movement of both soil and water-quality constituents associated with soil. The chemical composition of soil also affects water quality through processes such as cation exchange between water and soil.
The lower Tennessee River Basin includes parts of the Coastal Plain, Interior Low Plateaus, and Appalachian Plateaus Physiographic Provinces. The Coastal Plain Physiographic Province is located along the western edge of the study unit (fig. 1). It is an area of relatively low topographic relief with land-surface altitudes of about 650 ft above sea level.
The Interior Low Plateaus Physiographic Province covers most of the study unit (fig. 1) and is subdivided into the Nashville Basin and Highland Rim Physiographic Sections (fig. 2). The Nashville Basin, locally also called the Central Basin, has low to moderate relief with some small hills and knobs. Local relief of more than 100 ft is common (Fenneman, 1938). Land-surface altitudes generally are about 700 ft above sea level. The Highland Rim surrounds the Nashville Basin and is underlain predominantly by Mississippian carbonate rocks. The boundary between these two physiographic sections is an escarpment with about 300 ft of relief. Southwest and west of the Nashville Basin the Highland Rim generally has more relief and is more dissected than to the south and east. Land-surface altitudes range from 500 ft near the Tennessee River to 900 ft elsewhere. East of the Nashville Basin, the Highland Rim is a relatively even plateau dissected by a few entrenched meandering streams. Land-surface altitudes reach 1,300 ft, but generally are about 1,000 ft above sea level. Total relief in the Highland Rim is about 650 ft. Karst features such as caves, sinkholes, and springs are present in the Nashville Basin and the Highland Rim.
The Interior Low Plateaus are bounded to the south and east by the Cumberland Plateau Section of the Appalachian Plateaus Physiographic Province (figs. 1 and 2). A steep escarpment separates the Cumberland Plateau from the Highland Rim with the average altitude of the plateau about 1,000 ft above the Highland Rim. The Sequatchie Valley (fig. 3) lies about 500 ft lower than the adjacent upland parts of the Cumberland Plateau (Fenneman, 1938). East of the Sequatchie Valley is a plateau locally known as Walden Ridge or Sand Mountain, with an average land-surface altitude of about 1,300 ft and relief generally between 100 and 200 ft. In contrast, west of the Sequatchie Valley, the Cumberland Plateau is more dissected with land-surface altitudes ranging from about 1,000 to 1,600 ft with local relief as much as 300 ft.
Ecoregions represent areas in which factors (soils, vegetation, climate, geology, and physiography) affecting terrestrial and aquatic ecological systems are relatively uniform (Griffith and others, 1997). Ecoregion boundaries provide a framework for evaluating and managing environmental resources. Ninety-eight level III ecoregions have been defined across the nation (Bailey and others, 1994). The lower Tennessee River Basin includes parts of six level III ecoregions (fig. 6), which generally correspond to the physiographic provinces with some additional subdivision. Most of the study unit is in the Interior Plateau ecoregion, which corresponds to the Interior Low Plateaus Physiographic Province. The Southwestern Appalachians ecoregion corresponds to the Appalachian Plateaus Physiographic Province, and the Southeastern Plains is roughly coincident with the Coastal Plain Physiographic Province (fig. 6). The remaining level III ecoregions (Interior River Lowlands, Mississippi Valley Loess Plains, and Ridge and Valley) cover relatively small areas and are subdivisions within the physiographic provinces (fig. 6).
Subdivisions of level III ecoregions (level IV) decrease the heterogeneity within the larger-scale ecoregions. Level IV ecoregions are being defined on a state by state basis. This level of definition was not available for Alabama; however, level IV ecoregions had been delineated for Tennessee (Griffith and others, 1997) and were used for the subdivision of the study unit in that State.
Temperature is an important variable that influences the rate of most physical and chemical reactions. The quantity and quality of precipitation affect the quality of water in watersheds. The lower Tennessee River Basin has a temperate and warm, humid climate. Average temperature across the study unit ranges from 56 to 61 °F with an average annual temperature of about 59 °F (U.S. Department of Commerce, 1995). The warmest months are July and August and the coolest month is January. The difference between the average monthly high and low temperature for the year is about 40 °F.
Average annual precipitation in the lower Tennessee River Basin is about 56 inches (in.) (U.S. Department of Commerce, 1995). The average amount of precipitation for individual sites in the study unit ranges from about 50 in. in the western part of the study unit to about 60 in. in the eastern part of the study unit (fig. 7). The increase in precipitation from west to east generally corresponds to the increase in elevation from west to east. Average rainfall amounts are highest during November through May, with March generally being the wettest month (fig. 8). Average rainfall amounts are lowest June through October (fig. 8). The month in which rainfall is lowest varies from site to site. August through October usually is the driest part of the year (fig. 8).
The main stem of the Tennessee River is highly regulated with few free-flowing stream reaches. Six large reservoirs are located on the main stem (fig. 1), and many additional reservoirs are located on tributaries to the main stem. Most of these reservoirs were constructed from the 1930's to the 1970's for power generation, navigation, flood control, and for water supply. The largest tributaries to the Tennessee River are the Duck and Elk Rivers, which drain areas of 2,700 and 2,250 mi², respectively. Mean-annual discharge for the Tennessee River is about 35,900 cubic feet per second (ft³/s) at Chattanooga, Tennessee, and is about 65,600 ft³/s at the mouth of the Tennessee River at Paducah, Kentucky. The Duck and Elk Rivers combined contribute about 7,700 ft³/s, or only about 26 percent of the 29,700 ft³/s of flow that is gained in the study unit; most of the streamflow is contributed from river basins smaller than 1,000 mi².
Long-term streamflow data at selected sites illustrate the range of mean-annual discharges for large rivers in the lower Tennessee River Basin (fig. 9). The time series of mean-annual discharge was similar at all of the sites shown in figure 9. During the selected period (1930-95), the lowest mean-annual discharge on the Tennessee River occurred during 1988 (fig. 9) at Chattanooga and Savannah (23,000 ft³/s at Savannah). The largest mean-annual discharge during the selected period at Savannah was about 87,000 ft³/s in 1973. The 1970's were the "wettest" years at these sites with 6 of the 10 years having streamflow above the long-term mean-annual discharge. The 1980's were the "driest" 10-year period, with 3 years of streamflow well below the long-term mean-annual discharge (fig. 9).
The regulation of flow on the Tennessee River and the upper parts of the Elk and Duck Rivers and other tributaries moderates extremes that normally occur in streamflow. Streamflow data for the 1992 water year (October 1 through September 30), a year with near average streamflow (fig. 10), show the combined effects of basin size and reservoirs. Streamflow in the Duck River above Hurricane Mills (2,557 mi²) is minimally affected by Normandy Lake, which is located about 220 mi upstream; storm peaks generally are followed by smooth recessions in discharge (fig. 10). Streamflow in the Tennessee River at Savannah (33,140 mi²) is affected by several reservoirs upstream, primarily Pickwick Lake. Storm peaks on the hydrograph generally are broader and the corresponding recessions are somewhat slower than for the same storms affecting the Duck River (fig. 10). The hydrograph for Savannah also shows peaks and recessions resulting from upstream reservoir operations. Streamflow response to changes in releases from reservoirs generally are more rapid than response to storm events.
Reservoir operations also affect longer periods of streamflow in addition to short-term releases. Streamflow was generally lower than average during July through October in the Duck River as a result of less precipitation and runoff. However, streamflow at Savannah was near the mean-annual discharge between July and October (fig. 10). Near average streamflow during these months was maintained by releases from Pickwick Lake to generate power during the summer months.
Principal aquifers in the lower Tennessee River Basin are the Cretaceous sand, Pennsylvanian sandstone, Mississippian carbonate, and Ordovician carbonate aquifers (fig. 11). The areal extent of these aquifers in the study unit generally corresponds to the principal geologic units. Ground-water flow in the principal aquifers in the study unit generally is along relatively short flow paths through solution openings, bedding planes, joints, fractures in bedrock, and unconsolidated sediments. The amount of flow to a deeper, regional ground-water-flow system from the principal aquifers is relatively small.
Sediments of predominantly Cretaceous age, and minor Quaternary-age alluvium and Tertiary-age sediments, which consist of unconsolidated sand, silt, clay, and minor gravel constitute the Cretaceous sand aquifer (fig. 11), which is confined in most places. Recharge to the aquifer is from precipitation on sandy units in the outcrop areas. Ground water flows through the primary porosity of this aquifer. East of the Tennessee River, ground water in the Cretaceous sand aquifer moves along short flow paths and discharges within the Tennessee River Basin (Brahana, Mulderink, and others, 1986). Ground water west of the Tennessee River also discharges to streams in the study unit, but a component of ground water flows to the west, along regional flow paths, in the direction of the regional dip of the sediments. Ground-water gradients and velocities generally are lower in the Cretaceous sand aquifer than in other hydrogeologic settings in the study unit.
Sandstones and conglomerates are the primary water-bearing units that constitute the Pennsylvanian sandstone aquifer (fig. 11). These rocks generally have low permeability, and ground water flows primarily through fractures in these rocks (fig. 12). The number, size, and density of fractures decreases with depth, limiting the regional component of ground-water flow. Flow paths typically are short and ground water discharges to streams and springs. Recharge to the aquifers occurs where precipitation falls on outcrops of these units and where streambeds intersect these units (Brahana, Macy, and others, 1986). A shale confining unit between the Pennsylvanian sandstone aquifer and the underlying Mississippian carbonate aquifer retards the movement of ground water between these aquifers (fig. 12).
The Mississippian carbonate aquifer is the most areally extensive and productive aquifer in the lower Tennessee River Basin (fig. 11). Ground water in this aquifer flows in solution openings along bedding planes, joints, and fractures. These openings and the zone of dynamic ground-water circulation generally are at depths less than 300 ft below land surface (fig. 12). Recharge to the Mississippian carbonate aquifer is from precipitation percolating through the overlying regolith to the bedrock aquifer. The regolith can store large amounts of water that is slowly transmitted downward into the bedrock aquifer. Locally, the regolith may contain abundant gravel and is a productive aquifer. Generally, flow paths are short and ground-water gradients are relatively steep; ground water moves rapidly and discharges to streams and springs. The Mississippian carbonate aquifer is underlain by the Devonian-age Chattanooga Shale (fig. 3), a regional confining unit that effectively limits the movement of ground water between the Mississippian carbonate and the underlying Ordovician carbonate aquifers.
Ground water in the Ordovician carbonate aquifer (fig. 11) flows primarily in solution openings along bedding planes, joints, and fractures. These solution openings are most common and shallow ground-water flow occurs primarily within about 300 ft of land surface. Shaly beds and bentonite layers present in the Ordovician rocks retard the downward movement of ground water, but locally, joints in or erosion of these layers allows deeper circulation of ground water (fig. 12). The limestones that make up these aquifers are predominantly calcite, and contain little insoluble, soil-forming material and little or no regolith (generally less than 20 ft) overlies the aquifer. Flow paths are short and ground water moves rapidly and discharges to streams and springs.
Ground-water discharge to streams is a significant component of streamflow and affects the water quality of streams in the lower Tennessee River Basin. Water-quality properties that often are affected by ground-water discharge to streams include pH, temperature, dissolved oxygen, and concentrations of dissolved inorganic constituents. During dry periods of the year, ground-water discharge to springs and streams provides much or all of the flow in streams.
Hydrograph-separation and baseflow estimation methods were applied to streamflow data for selected streams in the lower Tennessee River Basin to estimate the contribution of ground water to streamflow. To cover a broad range of hydrologic conditions, surface-water discharge records of 10 or more years were selected for analysis; a few shorter periods of record also were used (table 1). The computer program RORA (Rutledge, 1998), which automates the Rorabaugh (1964) and Daniel (1976) method, estimates ground-water recharge using the recession-curve-displacement method. The computer program PART estimates baseflow using streamflow partitioning (Rutledge, 1998). Generally, the estimated ground-water discharge (PART) is less than the estimated recharge (RORA); this difference is attributed to riparian evapotranspiration if ground-water withdrawals are negligible (Rutledge, 1998). For the purposes of this report, the results of both programs are termed baseflow estimates and results from RORA represent an upper limit of the baseflow estimate. Both programs require the assumptions that the aquifers contributing baseflow are isotropic and uniform and that the streams and aquifers are not significantly affected by withdrawals, discharges, or impoundments. All of these assumptions are not met, particularly the first, in the selected stream basins and throughout much of the lower Tennessee River Basin; therefore, the results of this analysis are considered estimates.
Estimates of baseflow calculated by both programs (table 1) are considerably higher than estimates of net annual recharge for average flow years reported for the same streams (Hoos, 1990) calculated by the manual Rorabaugh-Daniel hydrograph-separation method. However, estimates of baseflow calculated by PART fall within the range reported for streams with watersheds underlain by the Pennsylvanian, Mississippian, and Ordovician aquifers using the manual baseflow estimation technique (Zurawski, 1978). Although estimates differ between the methods and studies (Zurawski, 1978; Hoos, 1990), the ranking of each principal aquifer setting based on estimates of baseflow is similar for all methods. Baseflow (ground-water discharge) is highest for basins underlain by the Mississippian aquifers and lowest in basins underlain by the Cretaceous sand aquifer (table 1). Baseflow in basins underlain by the Cretaceous sand aquifer is low, in part, because of loss of ground water to a regional flow system. Baseflow in basins underlain by the Ordovician carbonate aquifer and Pennsylvanian sandstone aquifer is about the same. Estimates for basins underlain by both Mississippian carbonate and Pennsylvanian sandstone aquifers were slightly higher than basins underlain by Pennsylvanian sandstone aquifer only (table 1).
Mean-annual runoff (inches of water discharged per unit area) in the study unit ranges from about 30 in/yr [about 2 (ft³/s)/mi²] in the area near Guntersville Lake to about 18 in. [about 1 (ft³/s)/mi²] in the northwestern part of the study unit (fig. 13). Runoff is affected by climate (temperature and precipitation), topography (slope), and soils. The decrease in runoff from east to west generally corresponds to a decrease in the amount of precipitation across the study unit (fig. 7). In the eastern part of the study unit, about 45 percent of rainfall becomes runoff and in the western part about 35 percent does. The movement of ground water along a regional flow path in the Coastal Plain may contribute to the decrease in runoff across the study unit.
Cultural setting includes human activities that affect natural surface- and ground-water quality. Human activities affect water quality through input of physical and chemical constituents to a water body or to land surface in a watershed, modifications to the hydrologic system, such as reservoirs and impoundments, and surface- and ground-water use. Inputs of constituents from human activities to a watershed can be characterized as one of two types, point or nonpoint sources. Nonpoint sources to surface and ground water may be a mix of cultural and natural sources and include runoff and infiltration of constituents from urban and agricultural areas and atmospheric deposition of constituents. Whereas the distribution and magnitude of point sources are generally well documented, the distribution and magnitude of nonpoint sources are more difficult to quantify. Land-use data are often used as a surrogate for the distribution of nonpoint sources in water-quality studies.
Reservoirs, water use, land-use data, and inputs of selected constituents in the lower Tennessee River Basin are described and quantified in the following sections. Inputs of nitrogen, phosphorus, and pesticides were estimated because occurrence and distribution of these constituents are largely influenced by human activities.
Reservoirs on the main stem of the Tennessee River and its tributaries are used for power generation, navigation, flood control, water supply, and recreation. Wilson Lake (fig. 1), completed in 1924, was the first reservoir constructed on the main stem in the study unit. Four additional reservoirs were completed on the main stem by 1944. Nickajack Lake, completed in 1967, was the last reservoir constructed on the main stem in the lower Tennessee River Basin. The Barkley-Kentucky Canal, which is used for navigation and power generation at Kentucky Lake, allows interbasin transfer of water between the lower Tennessee River Basin and the Cumberland River Basin. Interbasin transfer of water also occurs between the Mobile and lower Tennessee River Basins via the Tennessee-Tombigbee Waterway near Pickwick Lake.
Reservoirs affect water quality by acting as sinks for sediment, nutrients, metals, and hydrophobic organic compounds. Reservoirs also affect stream water temperature, dissolved oxygen concentrations, and concentrations of other water-quality constituents. Hydraulic residence time in reservoirs is an important factor in the effect of reservoirs on water quality. The average hydraulic residence time for main stem reservoirs ranges from about 4 to 35 days; whereas the average residence time of five tributary reservoirs maintained by the Tennessee Valley Authority ranges from about 10 to 500 days (E.A. Thornton, Tennessee Valley Authority, written commun., 1998). These tributary reservoirs are more likely to function as sinks for sediment and nutrients associated with sediment or for aquatic-plant uptake of nutrients because of longer residence times. The storage volume of the six reservoirs on the main stem of the Tennessee River is about 6.7 million acre-ft and the tributaries to the Tennessee River have a total storage volume of about 800,000 acre-ft (E.A. Thornton, Tennessee Valley Authority, written commun., 1998).
Total average water use in the lower Tennessee River Basin was about 5,380 million gallons per day (Mgal/d) in 1995 (U.S. Geological Survey, 1997). Thermoelectric use accounted for about 4,700 Mgal/d, and most of this water was returned to streams and reservoirs in the basin. The second largest water-use category was industrial water use, about 380 Mgal/d with about 350 Mgal/d of the total amount withdrawn from surface-water sources (fig. 14). Most of the thermoelectric and much of the industrial surface-water use occurs along the main stem of the Tennessee River. Public supply was the third largest water-use category, about 220 Mgal/d; most of this, about 160 Mgal/d, was withdrawn from surface-water sources. Water used for livestock amounted to about 30 Mgal/d and commercial water use was about 20 Mgal/d; most of the water for these uses was withdrawn from surface-water sources. Domestic, self-supplied water use was about 20 Mgal/d and was exclusively from ground-water sources.
The location and size of public-water supply surface-water withdrawals in the study unit are shown in figure 15a; withdrawal amounts grouped by subunit and water body are shown in figure 15b. The main stem of the Tennessee River supports the largest amount of public water supply (fig. 15b), including the largest surface-water public supply in the study unit, the City of Decatur, Alabama, which withdrew about 28 Mgal/d in 1995. The City of Huntsville, Alabama, withdrew about 22 Mgal/d in 1995 from the Tennessee River. The Duck and Elk Rivers and their tributaries supplied about 50 Mgal/d for public supply. Withdrawals from these streams are shown as mixed Eastern Highland Rim/Nashville Basin in figure 15b, because the drainage areas contributing to the streams at the withdrawal point include parts of both of these subunits. Withdrawals from the other tributaries in the study unit are associated with a single subunit, and accounted for about 46 Mgal/d combined.
The location and size of public-water supply ground-water withdrawals in the study unit are shown in figure 16a; withdrawal amounts of both public-supply and self-supplied domestic use by subunit are shown in figure 16b. Domestic and public supply ground-water use were highest in the Eastern Highland Rim subunit (fig. 16b), where about 40 Mgal/d was withdrawn from the Mississippian aquifers in 1995 (35 Mgal/d for public supply, 5 Mgal/d domestic use). The City of Huntsville, Alabama, located in this subunit, is the largest ground-water public supply user in the study unit (fig. 16a); ground water makes up 40 percent of the water used to supply the city's needs. About 14 Mgal/d was withdrawn from relatively shallow wells (maximum depth 125 ft) and a large spring. About 14 Mgal/d of ground water was withdrawn in both the Western Highland Rim and the Coastal Plain subunits (fig. 16b); domestic water use accounted for about 30 percent of the water used in the Western Highland Rim, but only about 15 percent of the water used in the Coastal Plain. Total ground-water use in each of the remaining subunits was about 5 Mgal/d or less in 1995. Ground-water withdrawals may induce recharge or reverse hydraulic gradients, which could have implications for water quality.
Wastewater and stormwater discharges are the primary point-source discharges to streams and rivers in the lower Tennessee River Basin. Wastewater discharge to streams in the lower Tennessee River Basin were estimated using effluent monitoring data reported to State agencies by permitted wastewater dischargers (J. Hughes, Tennessee Department of Environment and Conservation, written commun., 1998; M. Rief and T. Cleveland, Alabama Department of Environmental Management, written commun., 1998; V. Prather, Kentucky Department of Environmental Protection, written commun., 1998; G. Odom, Mississippi Department of Environmental Quality, written commun., 1998). Mean-daily flow data for 1992 or 1995 were available for 264 permitted dischargers (fig. 17), including all of the dischargers that were classified as major dischargers. In general, a municipal facility discharging more than 1 Mgal/d, or an industrial facility with specific process wastewater of concern, is considered to be a major discharger; others are classified as minor dischargers. Many additional wastewater dischargers are permitted, but because flow data were not available in digital format, they were not included in this estimate. These discharges represent a small percentage of the total wastewater discharged to streams in the lower Tennessee River Basin (S. Fishel, Tennessee Department of Environment and Conservation, written commun., 1998).
Facilities discharging 1 Mgal/d or more contributed about 90 percent of the 13,396 Mgal/d of wastewater effluent from the 264 dischargers. Most of the discharges (about 13,300 Mgal/d) were to the main stem of the Tennessee River (fig. 17). About 66 Mgal/d of wastewater was discharged to streams in the Elk River Basin and about 17 Mgal/d was discharged to streams in the Duck River Basin. Of the total amount of wastewater discharged, industrial and power generation (federal) facilities accounted for about 99 percent of the total effluent from permitted facilities in the study unit. Much of this water (about 88 percent) was used for cooling at power generation facilities and at manufacturing plants, and the predominant effects on water quality are on water temperature and dissolved oxygen concentrations. Municipal wastewater discharges totaled about 135 Mgal/d, with 21 Mgal/d of that (about 15 percent) discharged to the Elk and Duck River Basins.
Nutrient loads were estimated for the 264 dischargers with flow data in digital format. Generally, concentration data for total nitrogen and phosphorus in effluent were not reported, so average concentration values were used for the various types of treatment facilities (S. Fishel, Tennessee Department of Environment and Conservation, oral commun., 1998; National Oceanic and Atmospheric Administration, 1993). The estimated total nitrogen wastewater load discharged to streams throughout the lower Tennessee River Basin was about 13,430 tons in 1995. About 68 percent of the total nitrogen load was from facilities discharging a mean of 1 Mgal/d or more. Slightly more than half of the total nitrogen load was discharged directly into Wheeler Reservoir on the main stem. About 4 percent of the total nitrogen load was discharged into streams in the Elk and Duck River Basins combined. Although municipal facilities contributed about 1 percent of the flow from wastewater discharges, they contributed about 23 percent of the total nitrogen load in wastewater.
The estimated total phosphorus load from wastewater discharged within the study unit was about 770 tons in 1995. About 40 percent of the phosphorus load was discharged directly into Wheeler Reservoir on the main stem. About 14 percent of the total phosphorus load was discharged into the Elk and Duck River basins. Municipal wastewater effluent contributed about 93 percent of the total phosphorus load in wastewater.
About 1.5 million people (1995) reside in the study unit. Huntsville, Alabama (160,000), Chattanooga, Tennessee (152,000), and Decatur, Alabama (52,000), are the largest urban areas (U.S. Department of Commerce, 1997). Much of the population in the study unit is located in counties along the Tennessee River in northern Alabama (fig. 18), with much of this area lying within the Eastern Highland Rim. About 38 percent of the population is located in the Eastern Highland Rim (table 2), about 14 percent in the Coastal Plain, 14 percent in the Western Highland Rim, and 13 percent is located in the Inner and Outer Nashville Basin combined (table 2). The remaining 20 percent of the total population is distributed across the Plateau Escarpment and Valleys, Cumberland Plateau, Transition, and Valley and Ridge. Between 1980 and 1995 the population in the study unit grew by about 15 percent. The largest increases in population (up to 20 percent) between 1980 and 1995 occurred in the Inner and Outer Nashville Basin and the Eastern Highland Rim.
Land use in the lower Tennessee River Basin largely reflects the geology and physiography of the study unit and the distribution of people (fig. 19). About 51 percent of the study unit is forested land (table 2) based on 1992 land-use data. The amount of forested land by subunit in the lower Tennessee River Basin ranges from about 27 percent in the Eastern Highland Rim to 68 percent in the Plateau Escarpment and Valleys. Forested land is the predominant land cover in six of the nine subunits (table 2), with the largest forested areas located in the Western Highland Rim, Plateau Escarpment and Valleys, and Transition subunits, where topographic relief is the highest. Agricultural land (pasture and cultivated land) accounts for about 40 percent of the land use in the study unit (table 2); agriculture by subunit ranges from 25 percent of the Plateau Escarpment and Valleys to 63 percent of the Inner Nashville Basin (table 2). Pasture land accounts for about 72 to 97 percent of all agricultural land throughout the lower Tennessee River basin, and is the predominant land use in the Inner and Outer Nashville Basin and in the Eastern Highland Rim (table 2). Cultivated land generally is located in the Eastern Highland Rim, Cumberland Plateau, Inner Nashville Basin, and Coastal Plain (table 2 and fig. 19) where relief is lowest. The Eastern Highland Rim contains the largest percentage of cultivated land (16 percent) in the study unit (table 2); less than 6 percent of the land in other subunits is cultivated. Urban areas in the lower Tennessee River Basin represent 1 percent or less of the land use in most subunits (table 2) and about 1 percent of the study unit overall. Exceptions are the Valley and Ridge with about 11 percent urban land (the outlying area of Chattanooga, Tennessee), and the Eastern Highland Rim with about 3 percent urban land. The remaining 8 percent of the study unit includes open water and other land uses (mined land and wetlands).
The cultural activities related to land use most likely to have a widespread effect on water quality in the lower Tennessee River Basin are animal and rowcrop agriculture. County-level data for crop production and livestock population from the 1992 Census of Agriculture (U.S. Department of Commerce, 1994) were converted to estimates by subunit through a land-use weighting algorithm. This weighting algorithm apportions the county level data to each subunit based on the portion of cultivated land or pasture land, by county, encompassed within each subunit.This procedure required the assumption that different crop types and livestock were evenly distributed across their respective land use in each county.
Corn, soybeans, cotton, and wheat were the predominant crops grown in the lower Tennessee River Basin in 1992 (table 3). Corn acreage was the largest of the crop areas in 1992 and accounted for about 34 percent of the total harvested acreage of these crops (U.S. Department of Commerce, 1994). Soybean acreage accounted for about 32 percent, cotton about 23 percent, and wheat about 11 percent of the total harvested acreage of all four crops in 1992.
Harvested acreage for each crop type was normalized by the area of each subunit to illustrate the relative intensity of the agricultural activity by subunit in 1992 (fig. 20). The Eastern Highland Rim and the Coastal Plain ranked highest with respect to the number of cultivated acres per square mile. Cotton was grown primarily in the Eastern Highland Rim and was the largest crop both in total acreage in a subunit (table 3) and acreage per square mile (fig. 20). Corn and soybeans were grown in all of the subunits. The Coastal Plain, Eastern Highland Rim, and Inner Nashville Basin ranked highest in farming intensity for corn and soybeans and also supported the greatest amount of harvested acreage of wheat per square mile.
County-level 1992 livestock population data were used to estimate the number of livestock by subunit (table 3). These data also were normalized by the area of each subunit to illustrate the relative intensity of livestock production by subunit (fig. 21). About 42 million chickens, 0.9 million cattle, and 0.3 million hogs were raised in the study unit in 1992 (table 3). The Cumberland Plateau ranked highest with respect to the number of head per square mile (fig. 21). Chicken production was the largest livestock activity in the Cumberland Plateau, which supported about 8,000 chickens per square mile. The Valley and Ridge, Eastern Highland Rim, Inner Nashville Basin, and Plateau Escarpment and Valleys supported more than 2,000 chickens per square mile. The Inner and Outer Nashville Basin ranked first and second in cattle production and supported more than twice as many cows per square mile as did the remaining subunits. The Coastal Plain ranked first in hog production, with about twice as many hogs per square mile as the second highest-ranking subunit.
Nonpoint sources of the nutrients nitrogen and phosphorus include urban runoff, fertilizer application, failing septic tanks, livestock waste, nitrogen fixation, sediment and rock dissolution, and atmospheric deposition. Nutrient inputs to the study unit from atmospheric deposition, fertilizer application, livestock waste, and nitrogen fixation (from soybean crops) were estimated by subunit for 1992. The contribution of nutrients from failing septic tanks, urban runoff, and dissolution of rocks could not be estimated from available data.
Estimates of total nitrogen inputs from precipitation were derived from deposition rates for total nitrogen from five stations in the National Atmospheric Deposition Program/National Trends Network, a national network of precipitation chemistry monitoring stations operated in cooperation between State agricultural experiment stations, the U.S. Geological Survey, U.S. Department of Agriculture, and other governmental and private entities. Nitrogen deposition rates for subunits were estimated by calculating a weighted average deposition rate in 1992 from the five stations, based on their distance from the centroid of each subunit, and were multiplied by the subunit area to compute a total nitrogen input. Because phosphorus concentrations in precipitation generally are below the reporting level of the analytical method, atmospheric deposition is not expected to be a significant source of phosphorus (M.A. Nilles, U.S. Geological Survey, written commun., 1999); however, atmospheric deposition may contribute significant amounts of phosphorus in some regions (Harned, 1995).
Nitrogen and phosphorus inputs related to agricultural activities include fertilizer application, livestock waste, and nitrogen fixation. Estimates of inputs from fertilizer application were based on county sales data for 1991 and agricultural census data for 1987. The sources for these data and the methods of calculating county estimates are described in detail in Battaglin and Goolsby (1995). Estimates of inputs from livestock waste are based on the 1992 Census of Agriculture and estimates of nutrient content of livestock waste (U.S. Department of Commerce, 1994). Nitrogen fixation estimates were based on 1992 harvested acreage of soybeans, multiplied by a rate for nitrogen fixation (Craig and Kuenzler, 1983; U.S. Department of Commerce, 1994 ). County-level estimates of nutrient inputs from agricultural activities were converted to subunit-level estimates by weighting the area of cultivated and pasture land in a county and in each subunit.
Estimated nitrogen inputs to the study unit from nonpoint sources were about 195,000 tons (table 4). The nitrogen input to the study unit from agricultural activities (fertilizer, livestock waste, and nitrogen fixation) was about 161,000 tons in 1992. Atmospheric deposition of nitrogen totalled about 34,100 tons in 1992 (table 4) and contributed about 18 percent of the nitrogen input. Most of that input was contributed by livestock waste and fertilizer application; nitrogen fixation by soybeans accounted for about 8 percent of the nitrogen input. Although much of the nitrogen input from agricultural activities is taken up by crops and removed by harvest, some of this input may affect water quality, where inputs greatly exceed crop uptake.
The magnitude of nonpoint-source inputs varies across the lower Tennessee River Basin. The distribution of nitrogen input by subunit is shown, on a per unit area basis, in figure 22. The Eastern Highland Rim had the largest nitrogen input in the study unit (table 4) and also had the highest nitrogen input per unit area (fig. 22). Unit-area nitrogen inputs from livestock waste were the largest nonpoint source of nitrogen in the Cumberland Plateau, the Inner and Outer Nashville Basin, and the Valley and Ridge, and accounted for at least a quarter of the nitrogen input for all subunits. Unit-area nitrogen inputs from fertilizer application were the largest nonpoint source of nitrogen in the Eastern Highland Rim and Coastal Plain. Unit-area nitrogen inputs from atmospheric deposition were about the same for all subunits. In the Transition, Western Highland Rim, and Plateau Escarpment and Valleys, however, atmospheric deposition accounted for about 25 percent of the nitrogen input from nonpoint sources (fig. 22) because inputs from other sources are relatively low. Nitrogen fixation contributed the smallest amount of nitrogen in all subunits, but accounted for as much as about 16 percent of the nitrogen input in the Coastal Plain.
Inputs of phosphorus from livestock waste also were larger on a per unit area basis than inputs from fertilizer application for each subunit except the Eastern Highland Rim and Coastal Plain, where phosphorus input from fertilizer application was about the same or slightly more than from animal waste (fig. 22). About 23,500 tons of phosphorus were contributed by livestock waste and 14,500 tons of phosphorus by fertilizer application to the study unit in 1992 (table 4).
Pesticide use has increased tenfold since 1975, with 75 percent of the pesticide use related to agricultural production (Ware, 1989). Pesticides have become an important crop management tool for controlling insects, weeds, fungi, and bacteria, and their use has significantly increased crop production. Despite the increases in agricultural productivity and the associated economic benefits, a general concern exists about dispersing large quantities of these substances, which are designed to be toxic, into the environment.
The fate of pesticides in the environment is influenced by the method of application, physical and chemical properties of each pesticide, and many environmental factors. Pesticides are designed to degrade in a few days or weeks after application under ideal conditions. Movement of pesticides into ground-water or surface-water systems may reduce the rate at which pesticides degrade. Degradation products of pesticides can behave quite differently than parent compounds and may be more persistent or mobile in the environment.
Use of pesticides (herbicides, insecticides, and other pesticides) was estimated for each subunit to identify the primary compounds used, and the areal distribution of use in the lower Tennessee River Basin. For the purpose of this report, organic compounds used in crop production such as growth inhibitors and defoliants, are referred to as other pesticides. Estimates of pesticide use are based on reported harvested acreage of crops (U.S. Department of Commerce, 1994) and statewide application rates for a given compound (Gianessi and Anderson, 1995). These estimates, therefore, do not account for municipal and private use of pesticides or where local farming practices differ from general practices in the State.
About 3.7 million pounds of pesticides (active ingredient) was used in the lower Tennessee River Basin in 1992. Herbicides accounted for about two-thirds of total pesticide use (table 5). Eleven herbicides were used in amounts exceeding 100,000 pounds (table 5). Atrazine, a herbicide used primarily for corn production, was used in the largest amounts; followed by monosodium methanearsonate (MSMA), used primarily on cotton; 2,4-D used on corn and pasture land; and metalochlor used on corn and soybeans. Insecticide use in 1992 totalled about 0.9 million pounds (table 5) and accounted for about a quarter of the total pesticide use. Methyl parathion was used in the largest amounts, and was primarily applied to cotton. Thiodicarb, 1,3 dichloropropene, and aldicarb followed methyl parathion in amount used in 1992 and also were used primarily on cotton. Compounds termed other pesticides in table 5 include plant growth regulators (mancozeb and ethephon), defoliants (tribufos), fumigants (methyl bromide), and fungicides (sulfur). Methyl bromide was used in the largest amounts, primarily as a soil fumigant for tomato and tobacco cultivation. Tribufos and ethephon were used primarily on cotton and followed methyl bromide in 1992 estimated use (table 5).
Almost half of the total pesticides used in 1992 were applied to cropland in the Eastern Highland Rim, which had the largest unit-area pesticide use of all subunits, about 540 lb/mi² (table 5). Insecticide use in the Eastern Highland Rim represented 69 percent of the total insecticide use for the study unit and herbicide use there represented about 40 percent of the total. The majority of the insecticide used in the study unit was for cotton, which was primarily grown in the Eastern Highland Rim. Estimates of unit-area pesticide use for the Coastal Plain and Inner Nashville Basin were ranked second and third among the subunits; these ranks generally correlate with the relative percentage of cultivated land in each subunit (table 2).
| Next | Title Page | Table of Contents | List of Illustrations | List of Tables | Conversion Factors and Vertical Datum |