Assessment tools were developed to address the need for a low cost, rapid method of quantifying the quality of agricultural best management practices (BMPs). Best management practices are either cost-shared, where some or all of the capital costs of the practice were subsidized with federal, state, or local funds, or non cost-shared, where the cost of the practice and its upkeep is paid for by the landowner or farm operator. Cost-share practices are required to comply with state standards, while non cost-share practices are not subject to any standards. For this study, BMP quality is defined as the adherence to design, site selection, implementation, and maintenance criteria relating to water quality as specified by state and federal agencies promoting BMP implementation.

The two objectives of this research were:

1. develop a set of assessment tools to quantify the quality of agricultural best management practices in a rapid low-cost manner, and
2. test the tools and determine if differences in quality exist between cost-share and non cost-share BMPs in the James River Basin of Virginia.

Assessment tools were developed for sixteen practices: alternative water systems, stream fencing, streambank stabilization, grass filter strips, wooded buffers, permanent vegetative cover on critically eroding areas, permanent vegetative cover on erodible cropland, reforestation of erodible crop and pasture land, animal waste storage facilities, grazing land protection systems, loafing lot management systems, late winter split application of nitrogen on small grains, protective cover for specialty crops, sidedress application of nitrogen on corn, small grain cover crops-fertilized and harvested, and small grain cover crops for nutrient management.

Assessment tools were developed using both Virginia BMP standards and expert knowledge. Virginia Department of Recreation and Conservation (DCR) and Virginia and national Natural Resource Conservation Service (NRCS) BMP standards were collected and sorted into the four quality component categories; design, site selection, implementation, and maintenance. Standards that pertained directly to a BMPs' potential to protect water quality were translated into question format. Multiple-choice or yes/no questions were used as often as possible to avoid potential bias and for ease of processing. Assessment tool development involved an iterative process that included input from a research team (university-based researchers) and an expert team (public and private sector professionals and practitioners responsible for BMP design and assessment).

One hundred and fifty-five cost-shared BMPs and 150 non cost-shared BMPs were assessed on 128 independent farms in the James River Basin of Virginia over a period of four months. The assessment tools were loaded onto a personal digital assistant (PDA), which facilitated data collection and eliminated the need for data transcription. Data collected on the PDA were uploaded periodically to a computer database. A digital camera was used to develop a photographic record of the assessed BMPs.

Best management practice quality scores were based on five-point scale, with one being the lowest quality score and five as the highest. Statistical analyses conducted on both the overall quality scores and the quality component scores, indicate that there is not a strong significant difference (p = 0.05) in quality between the cost-shared and non cost-shared BMPs assessed for this study. Statistically significant differences between cost-share and non cost-share practices did, however, exist. For the filter/buffer strips practices (grass filter strips and wooded buffers), the implementation quality component cost-share mean (3.35) and the non cost-share mean (3.88) were statistically different at the 0.05 level (p-value = 0.026). One other statistically significant difference was found. For stream fencing, the overall quality cost-share mean was 4.68 while the non cost-share mean was 4.20; the means are statistically different at the 0.05 level (p-value = 0.043).

Statistical analyses were performed to determine if age of practice, farm size, or Soil and Water Conservation District (SWCD) had effects on the BMP quality. No statistically significant differences (p = 0.05) were found relating to the age of an assessed BMP or farm size. One SWCD, the Robert E. Lee district, had a statistically significant difference in the design quality component means; cost-share mean = 4.21, non cost-share mean = 2.94 with a p-value of 0.048.

The statistically significant differences that were detected do not establish a clear trend; it appears that for the BMPs assessed here the qualities of cost-share and non cost-share practices are roughly equal. The fact that cost-share practices and non cost-share practices do appear to be roughly equal may be the result of education and outreach programs sponsored by Virginia's SWCDs and Virginia Cooperative Extension. Non cost-share practices may be of equal quality to cost-share practices because those implementing BMPs without the benefit of cost-share may have a greater stake (both financial and personal) in those practices performing well.

If no statistically significant difference in quality exists between cost-share and non cost-share practices, then non cost-share practices should be treated equally when accounting for BMPs in NPS pollution in watershed management and computer modeling. Currently, only cost-share practices are included in computer models, in part because these are the only practices tracked by the existing BMP establishment infrastructure. Estimating the numbers and distribution of non cost-share practices and incorporating them into NPS water quality modeling efforts will more accurately reflect the steps agricultural producers have and are taking to decrease the amount of NPS pollution reaching water bodies. Additionally, policy regarding NPS pollution and BMPs should reflect the apparent equal qualities of cost-share and non cost-share practices.

The assessment tools developed as a part of this study can potentially be applied to determine the quality of BMPs on basin or state-wide scales to give policy makers a better understanding of the practices and populations that the policies are created for. Moreover, BMP quality scores have the potential to be used as a surrogate measure for BMP performance. Further research recommendations include correlating BMP quality scores with BMP performance, wider scale testing of the tools, continued revision of the tools, and using the assessment tool scores to diagnose BMP quality problems.