Inputs from several sources of nitrogen and phosphorus were compiled to compare the magnitude of inputs among the various sources, to identify areas in the basin with higher nitrogen and phosphorus inputs, and to compare inputs with instream loads. Two distinct types of input estimates are presented: land-phase inputs (mass applied to the land surface of the watershed) and stream inputs (mass discharged directly to the stream channel). Estimates of land-phase inputs cannot be compared directly to estimates of stream inputs or instream loads.
Sources of nutrients to watersheds and to streams and reservoirs include both point sources (wastewater and stormwater discharge, combined sewer overflows) and nonpoint sources (atmospheric deposition, fertilizer application, livestock waste, urban runoff, failing septic systems, contaminated ground water, and natural sources). Inputs from wastewater discharge, atmospheric deposition, fertilizer application, and livestock waste were quantified for 11 sites on the tributary streams and reservoirs and for the major hydrologic units in the LTEN River Basin (tables 3, 4, 5, and 6). Inputs from wastewater discharge are direct stream inputs, and inputs from nonpoint sources are land-phase inputs. Data from 1992 were used to estimate inputs where possible because more data were available in the data sets for sources, land use, and instream loads during this period than in any other period. Inputs from the other listed sources are discussed in the section "Additional Sources of Nitrogen and Phosphorus" but are not quantified in this report.
Sites on the main stem of the Tennessee River were excluded from the analysis of inputs because the main stem of the lower Tennessee River carries considerable load from the upper Tennessee (UTEN) River Basin, and the intent of this report is to interpret sources and transport only in the LTEN River Basin. Treece and Johnson (1997) and Johnson and Treece (1998) describe nitrogen and phosphorus sources, yields, and trends for the upper Tennessee River Basin. In addition, the many large impoundments along the main stem in both the upper and lower parts of the Tennessee River alter the instream transport of nutrients and sediment, confounding direct comparison of inputs with instream loads for the main stem sites.
Point sources discharge directly to the stream channel from a discrete location and include municipal and industrial wastewater discharges, municipal and industrial stormwater discharges, and sanitary and combined sewer overflows. Data were not available to estimate nutrient inputs from stormwater discharges and sewer overflows; therefore, the only point-source inputs estimated in this report are municipal and industrial wastewater discharge.
In 1992, an estimated 730 municipal and industrial facilities discharged wastewater into streams and reservoirs in the LTEN River Basin. Nitrogen and phosphorus inputs were estimated for the individual wastewater discharges (fig. 4) using methods described in Appendix C and summarized in table 7. These estimates were then summed for the tributary basins and the major hydrologic units. The input estimates were normalized by watershed drainage area to allow comparisons among basins and hydrologic units.
Estimated inputs of nitrogen from wastewater discharge for selected tributary monitoring basins (table 3) ranged from 0 (ton/mi2)/yr (site 11, Flint Creek near Falkville) to 0.61 (ton/mi2)/yr (site 1, Clarks River at Almo), and of phosphorus (table 4), from 0 (ton/mi2)/yr (site 11, Flint Creek near Falkville) to 0.14 (ton/mi2)/yr (site 1, Clarks River at Almo). Estimated inputs of nitrogen for major hydrologic units (table 5) ranged from 0.021 (tons/mi2)/yr (hydrologic unit 06020004, Sequatchie River basin) to 5.5 (tons/mi2)/yr (hydrologic unit 06040006, area contributing to below Kentucky Dam) and of phosphorus (table 6), from 0.005 (ton/mi2)/yr (hydrologic unit 06020004, Sequatchie River basin) to 0.11 (ton/mi2)/yr (hydrologic unit 06030002, area contributing to Wheeler Reservoir).
Nonpoint-source inputs to a watershed have diffuse source areas, ranging from a few square miles to the entire watershed area. Only a part of the input from these sources reaches the stream channel; the remainder accumulates within the watershed or is lost through processes such as crop harvest and export and denitrification. Nonpoint-source inputs of nutrients estimated in this report (atmospheric deposition, fertilizer application, and livestock waste) primarily are related to human activities.
More than 3.2 million tons of nitrogen is deposited in the United States each year from atmospheric deposition (Puckett, 1994). The combustion of fossil fuels such as coal and oil is the major source of nitrogen in atmospheric deposition. Atmospheric deposition of nitrogen may be in a wet form as rain, snow, hail, fog, and freezing rain, or in a dry form as particulates, gases, and droplets. Atmospheric deposition of phosphorus is not considered a significant source of phosphorus to watersheds in general and is not estimated for this report; however, atmospheric deposition may contribute significant amounts of phosphorus in some locales (Harned, 1995).
The methods used for developing estimates of inputs of nitrogen from atmospheric deposition are described in Appendix C and summarized in table 7. Estimated inputs of nitrogen for selected tributary monitoring basins (table 3) ranged from 1.6 (tons/mi2)/yr (site 1, Clarks River at Almo) to 2.0 (tons/mi2)/yr (site 12, Town Creek near Geraldine). Estimated inputs of nitrogen for major hydrologic units (table 5) covered this same range; from 1.6 (tons/mi2)/yr (areas contributing to lower and upper Kentucky Reservoir and to below Kentucky Dam, hydrologic units 06040001, 06040005, and 06040006, respectively) to 2.0 (tons/mi2)/yr (area contributing to Guntersville Reservoir, hydrologic unit 06030001). These estimates fall near the upper part of the range [0.9-1.8 (tons/mi2)/yr] reported in a national assessment (Puckett, 1994).
Agricultural lands are associated with major nutrient sources and sinks at the land surface. Major sources include fertilizer application to crop, nitrogen fixation by leguminous crops, and livestock waste released from feedlots and husbandry operations (fig. 6). Part of the nutrient mass applied to cropland (in both inorganic fertilizer and livestock manure) is utilized by plants and incorporated into plant biomass. This mass, along with the amount of nitrogen assimilated through biological fixation, is removed from the basin when crop plants are harvested and exported. The amount of applied nutrient not removed from the basin by crop uptake and harvest may be transported, overland or in the subsurface, to the stream channel. To obtain a general estimate of the balance of sources and sinks available to enter the overland and subsurface transport phase, the estimate of crop uptake is subtracted from the sum of estimates of fertilizer application, nitrogen fixation, and livestock waste. However, this estimate does not reflect other processes, such as accumulation or denitrification in the soil, which may represent major components of the nutrient budget in agricultural lands.
Commercial nitrogen fertilizer is applied as either ammonia or nitrate. Part of the nitrogen is absorbed by the growing crop, part is released in gaseous form to the atmosphere, and part remains as nitrate in the soil. Nitrate is soluble in water and readily leached from soils, allowing the rapid transport of nitrate into streams and ground-water systems. Phosphorus fertilizer is commonly applied as phosphate, which readily adheres to clay particles and is relatively insoluble in water. Soil erosion and transport is, therefore, the primary process by which significant amounts of particulate phosphate travel to streams.
The methods used to quantify inputs of nitrogen and phosphorus from fertilizer application are described in Appendix C and summarized in table 7. Input estimates of nitrogen for selected tributary monitoring basins (table 3) ranged from 1.5 (tons/mi2)/yr (site 13, Sequatchie River at Valley Road) to 23 (tons/mi2)/yr (site 1, Clarks River at Almo), and of phosphorus (table 4), from 0.35 (tons/mi2)/yr (site 11, Flint Creek near Falkville) to 5.1 (tons/mi2)/yr (site 1, Clarks River at Almo). Estimated inputs of nitrogen (table 5) for major hydrologic units ranged from 1.1 (tons/mi2)/yr (area contributing to Nickajack Reservoir, hydrologic unit 06020001) to 10 (tons/mi2)/yr (area contributing to below Kentucky Dam, hydrologic unit 06040006), and of phosphorus (table 6), from 0.21 (tons/mi2)/yr (Bear Creek basin, hydrologic unit 06030006) to 2.3 (tons/mi2)/yr (area contributing to below Kentucky Dam, hydrologic unit 06040006). The spatial distribution of nitrogen and phosphorus inputs generally corresponds to the distribution of cultivated land (figs. 3 and 5).
Leguminous crops, such as soybeans, absorb atmospheric nitrogen from Rhizobium bacteria which infect their roots (nitrogen fixation). The mass of nitrogen fixed by crops through biological fixation was estimated because this mass is part of the balance between applied fertilizer, livestock manure, and crop uptake, but is not an expected input to water bodies.
The methods used to quantify inputs of nitrogen from nitrogen fixation are described in Appendix C and summarized in table 7. Estimated inputs of nitrogen for selected tributary monitoring basins (table 3) ranged from 0.30 (ton/mi2)/yr (site 2, Buffalo River near Flat Woods, and site 13, Sequatchie River at Valley Road) to 7.6 (tons/mi2)/yr (site 1, Clarks River at Almo). Input estimates for the major hydrologic units (table 5) ranged from 0.14 (tons/mi2)/yr (area contributing to Nickajack Reservoir, hydrologic unit 06020001) to 4.0 (tons/mi2)/yr (area contributing to below Kentucky Dam, hydrologic unit 06040006).
The nutrient mass applied to cropland as fertilizer, along with the mass fixed biologically, is partly removed from the land surface when crops are harvested. Part of the plant remains on the land as residue after harvest; therefore, the mass removed as harvested crop is somewhat less than crop uptake. The methods used to quantify nutrient removed as harvested crop are described in Appendix C, and summarized in table 7. Estimates are reported as negative values because harvest represents removal of nitrogen. Crop uptake estimates of nitrogen for selected tributary monitoring basins (table 3) ranged from -1.7 (tons/mi2)/yr (site 13, Sequatchie River at Valley Road) to -24 (tons/mi2)/yr (site 1, Clarks River at Almo), and of phosphorus (table 4), from -0.19 (tons/mi2)/yr (site 11, Flint Creek near Falkville and site 13, Sequatchie River at Valley Road) to -3.2 (tons/mi2)/yr (site 1, Clarks River at Almo). Crop uptake estimates of nitrogen for major hydrologic units (table 5) ranged from -0.83 (tons/mi2)/yr (Bear Creek basin, hydrologic unit 06030006) to -12 (tons/mi2)/yr (area contributing to below Kentucky Dam, hydrologic unit 06040006), and of phosphorus (table 6), from -0.10 (tons/mi2)/yr (Bear Creek basin, hydrologic unit 06030006) to -1.5 (tons/mi2)/yr (area contributing to below Kentucky Dam, hydrologic unit 06040006).
Nationwide, approximately 7 billion farm animals generate millions of tons of manure containing some 6.5 million tons of nitrogen and 2 million tons of phosphorus each year. Organic nitrogen and urea in the manure are converted to ammonia (part of which volatilizes) and ultimately to nitrate (Mueller and Helsel, 1996). Most organic phosphorus is converted to phosphate, which adheres to soil particles and may become mobile through soil erosion. Confined animal feeding operations, which concentrate animals, feed, and manure on a small land area, have a greater potential to contribute nutrients to surface runoff and ground water. Manure produced by these operations may be applied to pasture land and crop land, becoming available for either crop uptake or losses to the environment (fig. 6).
The methods used to quantify inputs of nitrogen and phosphorus from livestock waste are described in Appendix C and summarized in table 7. Estimated inputs of nitrogen for selected tributary monitoring basins (table 3) ranged from 2.0 (tons/mi2)/yr (site 2, Buffalo River near Flat Woods) to 13 (tons/mi2)/yr (site 12, Town Creek near Geraldine), and of phosphorus (table 4), from 0.62 (tons/mi2)/yr (site 2, Buffalo River near Flat Woods) to 4.3 (tons/mi2)/yr (site 12, Town Creek near Geraldine). Input estimates of nitrogen (table 5) for major hydrologic units ranged from 1.5 (tons/mi2)/yr (area contributing to lower Kentucky Reservoir, hydrologic unit 06040005) to 7.0 (tons/mi2)/yr (area contributing to Guntersville Reservoir, hydrologic unit 06030001), and of phosphorus (table 6), from 0.52 (tons/mi2)/yr (area contributing to lower Kentucky Reservoir, hydrologic unit 06040005) to 2.2 (tons/mi2)/yr (area contributing to Guntersville Reservoir, hydrologic unit 06030001). The spatial distribution of nitrogen and phosphorus input generally corresponds to the distribution of pasture land (figs. 3 and 5).
Several other sources of nutrients have been identified, but quantitative information for these sources is not available. Urban runoff and combined sewage overflow are potentially large sources of nutrients from urban areas, but may be comparatively small sources in the LTEN River Basin because of the relatively small urban component in the basin. Nutrient contribution from natural sources, such as weathering and erosion of geologic materials in the watershed, may be significant, but is difficult to quantify.
Nutrient contribution from leachate from failing septic systems may be significant in residential areas experiencing high rates of failure. A general estimate of nutrient loads in surface runoff from these areas (presented in this section) is derived from runoff-monitoring data from a residential area near Scottsboro, Alabama, which experienced a high rate (60 percent) of septic-system failure during the period 1983-85 (Sagona, 1988). Surface-runoff loads of total nitrogen and total phosphorus during 1985 in this area were 1.9 and 0.34 (lbs/acre)/storm event, respectively. These estimates were adjusted to account for other nutrient sources, such as lawn fertilizer, by subtracting the estimate for average runoff load from an adjacent residential area with a lower rate of septic-system failure (45 percent).
Annual estimates of unit-area surface-runoff load from failing septic systems were calculated by multiplying the adjusted average runoff load by the average number of runoff events per year for the area (Steurer and Nold, 1986) as 42 (tons/mi2)/yr nitrogen, and 6.7 (tons/mi2)/yr phosphorus. These estimates may be conservative due to sampling bias: the period of runoff monitoring at the high-failure-rate site did not include the wet season, when leachate amounts are highest; and the presence of some septic-system leachate in the samples from the adjacent, lower-failure-rate site probably caused the adjustment for background to be too large.
The estimates from the residential area near Scottsboro are at least an order of magnitude higher than estimates of unit-area input from wastewater discharge for most of the tributary basins (tables 3 and 4). However, these rates cannot be compared directly because the septic system loading rates apply to local areas with high failure rate, which may represent only a small part of the watershed area, whereas the wastewater discharge loading rates apply to the entire watershed. Because watershed-wide data on septic-system failure rates for unsewered areas were not available, extrapolation of these rates to produce basin estimates of nitrogen and phosphorus inputs, similar to the other sources in this report, was not possible.
| Next | Title Page | Table of Contents | List of Illustrations | List of Tables | Conversion Factors and Vertical Datum | Glossary |