Much has been written recently in the popular press about agriculture and its negative or harmful affects on the environment. Problems in North Carolina associated with confined pork production facilities and degradation of streams adjacent to these facilities seems to have affected the perception of agriculture across the Southeast if not the entire nation. The Clinton Administration's 1998 Clean Water Initiative and recent South Carolina law requires revision and increased enforcement of effluent guidelines and rules for all animal feeding operations, in part because of the perception that these operations contribute significantly to non-point source pollution. Because of these and numerous other reasons, the authors wished to determine if animal feeding operations contribute significantly to the non-point source pollution problem in South Carolina.

This component of the overall agricultural perceptions study attempted to determine whether or not spatial relationships existed between areas of the state where large numbers of agricultural animals are concentrated and water quality by watersheds. The authors also examined the human population in the same context. The study area was limited primarily to the state of South Carolina, though some of the watersheds analyzed crossed political boundaries into North Carolina and Georgia.

All project data were analyzed using Arc/Info and ArcView geographic information systems (GIS) software. The following datasets were either used in the final analysis or used to determine which datasets were necessary for the project. The 1992 Agricultural Census was utilized to obtain total numbers of cattle, total numbers of hogs and total numbers of chicken farms. The data for each category was broken down by zip code in order to provide the finest spatial resolution (highest level of detail) possible for the state. Stream or hydrological data was aggregated to 8 digit hydrologic units thereby forming 34 separate watersheds. Aggregating hydrologic data was necessary in order to reduce the data volume associated with the thousands of streams in the state.

The map titled "Index of Watershed Indicators" depicts the watersheds used in this study. Initially the project team used US Environmental Protection Agency (EPA) water quality data tagged to the 8 digit hydrologic units. However, it was determined the EPA data was generalized to include all forms of contaminants and would not accurately portray potential water contamination by agricultural animals. The project team decided to use selected South Carolina Department of Health and Environmental Control (DHEC) Clean Water Act303(d) data as the indicator of water quality impairment.

Data used from within the Department included the following:

• Water chemistry data from 700 surface water and sediment monitoring sites
• Approximately 500 shellfish growing monitoring sites
• Fish, oyster, and crab tissue monitoring data
• Stream macroinvertebrate assessments
• Discharge Monitoring Report data
• Lake eutrophication monitoring data
• Environmental Surveillance Oversight Program (Savannah River Site)
• Nonpoint Source (319) data from State Nonpoint Source Management Plans
• 304(l) Short List
• State Watershed Water Quality Management Strategies
• State Shellfish Restoration Committee
• Special studies or general knowledge

After considering the potentially relevant indicators of stream impairment, the project team decided that fecal coliform bacteria (an indicator of levels of animal waste pathogens) and dissolved oxygen (an indicator of levels of oxygen-demanding pollutants) would be the measures most closely linked to animal agriculture.

Phosphorus was not used as an indicator because of the very rare incidence of elevated levels of phosphorus in South Carolina streams. Nitrogen was not used because recent work completed by the US Geological Survey's (USGS) National Water Quality Assessment Program has shown for at least three of the five major river basins in South Carolina, non-point source nitrogen loads are significantly heavier from the atmosphere than from manure and fertilizer. Other measures like metals and chemical pollutants were not used because it is thought that they would not be indicative of animal agriculture land use. Measures of impaired use collected by DHEC but not used in this study include phosphorous, cesium, chromium, copper, hydrogen ion concentration, macroinvertebrates, mercury, contaminated sediments, pesticides, strontium and zinc.

If any stream within a watershed was considered impaired for use because of fecal coliform or dissolved oxygen, the entire watershed was labeled as impaired by that particular condition. Impaired streams as defined by DHEC fall under four categories; 1) Impaired for Recreational Use (Swimming), 2) Impaired for Recreational Use (Fish Consumption), 3) Impaired for Recreational Use (Shellfish Harvesting) and 4) Impaired for Aquatic Life Use. Again, streams were considered impaired if fecal coliform or dissolved oxygen impaired any of the four use categories. Once the impaired watersheds were identified, the different animal population numbers as well as the human population numbers were overlaid to determine spatial relationships between water quality degradation and animal concentrations.

Results

The final analysis produced nine maps of new information regarding South Carolina's waterways and animal agriculture. The first map, as mentioned earlier, is the "Index of Watershed Indicators" containing the EPA water quality data. Again, the data represented here is too generalized for in depth analysis but does show generally good water quality for the majority of watersheds in South Carolina. The remaining maps can be categorized into two broad groupings; animals and humans related to dissolved oxygen impairment and animals and humans related to fecal coliform bacteria impairment.

Dissolved Oxygen

When examining the dissolved oxygen maps one finds the majority of problems in the coastal watersheds as well as the Lower Catawba, Lower Pee Dee and Enoree watersheds. The Lower Catawba (Charlotte, North Carolina), the Lower Pee Dee (Florence, South Carolina) and Enoree (Greenville/Spartanburg, South Carolina) are all associated with urban areas as are the coastal watersheds. Notable exceptions to this pattern are the watersheds adjacent to the Columbia, South Carolina metropolitan area and the Aiken, South Carolina/Augusta, Georgia areas. One explanation of this pattern is that many of the older coastal urban areas have problems with combined sewer overflows, where sanitary wastewater lines and stormwater lines run together and their overflow empties into streams. Newer systems inland better separate and contain the sanitary sewer system and the stormwater sewer system. Other dissolved oxygen problems are thought to be associated with urban runoff.

What is clear from these maps is that the spatial relationship between impaired streams for dissolved oxygen and concentrations of hogs, cattle and chicken farms is not strong. Whereas, the spatial relationship between the human population and impairment for dissolved oxygen appears to be significant.

Fecal Coliform Bacteria

The maps showing impaired waters from fecal coliform bacteria revealed a different pattern of problems than the dissolved oxygen maps. Once again, many of the same watersheds associated with or adjacent to urban areas turned out to contain fecal coliform problems. The total number of watersheds with problems increased, with the notable exception of watersheds downstream from the Columbia, South Carolina metropolitan area. The spatial correlations between hogs and chicken farms and water use impairment from fecal coliform bacteria appear very low. However, the spatial relationship between cattle and fecal coliform bacteria appears to be significant. In addition the relationship between human population and fecal coliform bacteria impairment appears to be significant, especially in the upstate watersheds. It should be noted that there are many more humans in these upstate watersheds than cattle, so pure numbers would make the relationship stronger for human populations.

Implications

It has been reported recently that agriculture impairs 70% of US rivers and streams and that animal feeding operations are a significant contributor. The maps produced from this study indicate something to the contrary, at least in South Carolina. In South Carolina, it is the urban areas that are most closely linked to dissolved oxygen and fecal coliform bacteria problems.

President Clinton's Clean Water Initiative and new DHEC regulations primarily aim to reduce non-point source pollution from animal feeding operations. Data from EPA, USGS, and US Department of Agriculture (USDA) clearly show the potential for problems with animal feedlot runoff is exclusive to a relatively small number of states, not including South Carolina. The authors suggest that federal dollars potentially appropriated through the Clean Water Initiative to South Carolina might be better spent enforcing existing regulations on industries and municipalities, rather than enforcing new regulations on agriculture enterprises to correct problems that do not seem to exist.

A working paper of the South Carolina Water Resources Center at the Strom Thurmond Institute, Clemson University. August, 1998.