South Carolina Case Study
Principal Investigator: Megan Mehaffey
States, territories and tribal lands are required to assess the environmental status of their waters on a biennial schedule under section 305 of the Clean Water Act. Part of the assessment includes identification, listing and prioritization of the water bodies that do not meet water quality standards, which is referred to as the 303(d) list. States and other jurisdictions are required to develop Total Maximum Daily Loads (TMDL) for the water bodies on the 303(d) list. A TMDL is a pollution budget for a water body that estimates the amount of pollutant a water body can receive and still meet the designated water quality standards. More information about the TMDL process can be found at: (http://www.scdhec.net/eqc/water/html/tmdl.html), a majority (~65%) of sites were placed on the 303(d) list due to fecal coliform. Fecal coliform are bacteria that usually reside in the digestive tract of warm blooded animals, and are used as a surrogate for the presence of pathogens in water bodies. High concentrations of fecal coliform are indicative of nonpoint source (NPS) pollution - pollution that cannot be tracked to effluent emanating from a pipe discharging into a water body. The nonpoint source nature of the South Carolina's fecal coliform problem indicated that land-use characteristics of the watershed were likely a contributing factor, and that quantitative evaluation of the relationship between land use and fecal coliform pollution might reveal and underlying geographic aspect to the problem.
Logistic regression was use to evaluate the quantitative relationship between land use and exceedence of fecal coliform water quality standards. Logistic regression is a statistical tool for estimating the likelihood of a response (e.g,. 0 = fecal coliform standard not exceeded; 1 = fecal coliform standard exceeded). Measures of land-use characteristics were estimated for all watersheds that included a monitoring site on the 303(d) list, whether it was listed for exceeding fecal coliform or some other water quality standard. Five measures of land-use characteristics were selected for evaluation. They were: 1) percentage urban, 2) percentage agriculture, 3) percentage of agriculture on slopes greater than nine percent, 4) percentage natural vegetation, and 5) percentage of the total streamlength in the watershed with natural vegetation.
Model results indicated a significant relationship between land-use characteristics and likelihood of exceedence of the fecal coliform water quality standard. Model concordance, a measure of goodness-of-fit, was 72%. Significant explanatory variables included: 1) percentage urban, 2) percentage of agriculture on slopes greater than nine percent, and 3) percentage of the total streamlength in the watershed with natural vegetation. An implication of the results is that there is likely to be an underlying geography to the fecal coliform problem in South Carolina. For example, the likelihood of exceedence of fecal coliform standards increases as the percentage urban and steep-slope (>9%) agriculture increases in the watershed.
The project has been completed. Follow-on work may include refinement of watershed delineation and re-evaluation of quantitative relationships using updated 303(d) list. We used the 303(d) list for 1998, but the 303(d) lists for 2000 and 2002 are now available.
303(d) list: South Carolina Department of Health and Environmental Control (SCDHEC), Bureau of Water, 1998 data for 303(d) listings, http://www.scdhec.net/eqc/water/index.html.
Land cover: U.S. Multi-Resolution Land Characteristics Consortium (MRLC), National Land Cover Data (NLCD) ca. 1992, http://www.epa.gov/mrlc
Elevation: National Elevation Dataset (NED), http://gisdata.usgs.net/NED
Streams: National Hydrography Dataset (NHD), http://nhd.usgs.gov/
The project has been completed. Follow-on work may be undertaken focusing on model refinement using updated 303(d) data, and updated watershed delineations.
Smith, J.H., Wickham, J.D., Norton, D.J., Wade, T.G., and Jones, K.B. 2001. Utilization of landscape indicators to model pathogen impaired waters. Journal of the American Water Resources Association, 37:1-10.
Jennings, D.B., Jarnigan, S.T., and Ebert, D.W. in press. A modeling approach for estimating watershed impervious surface area from National Land Cover Data 92. Landscape Ecology.
Jones, K.B., Neale, A.C., Nash, M.S., Van Remortel, R.D., Wickham, J.D., Riitters, K.H. and O'Neill, R.V. 2001. Predicting nutrient and sediment loadings to streams from landscape metrics: a multiple watershed study from the United States mid-Atlantic Region. Landscape Ecology, 16:301-312.
Vogelmann, J.E., Howard, SM., Yang, L., Larson, C.R., Wylie, B.K., and Van Driel, N. 2001. Completion of the 1990s National Land Cover Data Set for the Conterminous United States from Landsat Thematic Mapper data and ancillary data sources. Photogrammetric Engineering and Remote Sensing, 67:650-662.
Wickham, J.D., Riitters, K.H., O'Neill, R.V., Reckhow, K.H., Wade, T.G., and Jones, K.B. 2000. Land cover as a framework for assessing risk of water pollution. Journal of the American Water Resources Association, 36(6):1-6.