Best Management Practices (BMPs) are applied in agricultural and urban areas to reduce export of anthropogenic N and P to downstream lakes and estuaries. While the effectiveness of BMPs has usually been tested at the plot scale, effects of BMPs on water quality are not often directly measured at the watershed scale. Furthermore, tests at the watershed scale may show little improvement in water quality despite additions of BMPs within the watershed. In this case, BMP efficiency may be overestimated or insufficient numbers of BMPs may be implemented. We are working with farmers in three experimental watersheds and one control watershed in the Choptank Basin on the Mid-Atlantic Coastal Plain of the US to test whether effects of additional BMPs beyond those already present can be detected at the field scale where the BMPs are implemented, at intermediate points in the watershed, and at the outlet of the watershed. At the farm scale we sample groundwater, tile lines, and ditches monthly where BMPs have been installed, at the intermediate scale we sample small streams monthly, and at the watershed scale we sample baseflow monthly and stormflow 8-10 times per year for analysis of N, P, and sediment.
Results/Conclusions
We recruited ~2/3 of the farmers in our watersheds, and we work with these farmers to increase BMP adoption. Original BMP coverage has doubled, and these BMPs are primarily cover crops, drainage control structures, denitrifying bioreactors, and gypsum applications. Several years of monitoring have shown that cover crops reduce N export from our agricultural fields, with little effect on baseflow P concentrations. Drainage control structures on tile lines raise the water table, which reduces N export from tile lines; however, in hydric soils the effect is marginal, and we have observed P mobilization. Bioreactors divert stream or ditch water through wood chip chambers with high microbial N demand and N reduction efficiencies >90% for water passing through the reactor; however, many storms create flows exceeding the bioreactor’s capacity, and the overall efficiency including storm bypass is ~25%. At the watershed scale, two experimental watersheds show reductions in N, but the third watershed has a low cooperation rate by farmers, very sandy well-drained soils, few additional BMPs, and increasing N at the watershed outlet. These results show that some BMPs have detectable effects at multiple spatial scales, but BMPs also need to be based on anthropogenic inputs and watershed characteristics.