To help local, rural planners make decisions on which Best Management Practices (BMPs) are most appropriate for the control of Non-Point Sources of Pollution (NPSP) within their jurisdiction, a method is developed for unbiased site-specific evaluation of BMPs appropriateness. An appropriate BMP is functional, practical and cost effective.

The method developed is a rational method suitable for the general evaluation of agricultural BMPs to control NPSP in rural watersheds. The method is functional, practical, and cost effective because it uses existing data and personnel wherever and whenever possible.

Unlike other NPSP modeling methods, this method lends itself to the selective application of BMPs throughout the watershed based on its actual land use and physical composition. This means, BMP's can be applied to the entire watershed uniformly or applied to a subset of the watershed based on selection criteria such as field slope, soil type, or existing cover crop. In addition, a combination of selection criteria can be used, and a combination of BMPs. This flexibility in selecting various BMPs and BMP combinations, allows planners to pose many different schemes for controlling NPSP within the watershed. The results of the schemes can be compared to determine which scheme is the most appropriate for use within the specific watershed.

The method developed relies heavily on the United States Department of Agriculture's Soil Conservation Service's (SCS) Universal Soil Loss Equation (USLE) and uses data already being collected by the SCS and the Agricultural Stabilization and Conservation Service (ASCS). The method does require the creation of Hydrologic Response Units (HRUs) comprised of unique combinations or soil type and land use areas.

To demonstrate the method's ability to select an effective BMPs for a watershed, a series of BMP scenarios were evaluated. The BMP scenarios were divided into three categories: a uniform application of a BMP throughout the watershed; a selective application of a BMP throughout the watershed and; the selective application of a set of BMP's throughout the watershed. Using a series of trial runs, the benefits of using the method were demonstrated. For example, similar results were obtained using a cropping BMP of "Corn, grain, soybean and cover" throughout the sample, and a ''no-till" BMP on 3 percent of the sample. This type of information was used by economic modelers to help target precious BMP implementation dollars.

As an indirect result of the method's HRU development, it was possible to test the independence of each of the USLE factors within the Nansemond/Chuckatuck watershed. The results of the Duncan's Multiple Range Test shows an interdependence between the "soil type and the cover", "overland flow versus cover", and "overland flow versus soil type".