Nonpoint phosphorus (P) pollution of surface and shallow groundwater is a global environmental problem, particularly in regions where animal agriculture is concentrated. In the Coastal Plain region of Delaware, over application of manures and P fertilizers have saturated soils with respect to P. One practice commonly recommended by advisory agencies to control nonpoint source (NPS) P loss from agricultural land is the vegetated filter strip (VFS). Vegetative filter strip effectiveness at mitigating NPS P pollution is usually based on quantity of stormwater infiltrated and sediment settled out of the water column. When soils are saturated with respect to P, VFS effectiveness at preventing dissolved P losses by enhancing infiltration and retaining dissolved P moving laterally through subsoil is less certain. It is therefore the goal of this research to address three key P management questions. First (Chapter 2) is to determine methods to increase soil P sorption capacity that can be included in VFS design, based on processes that control P sorption in the acid sandy Mid Atlantic Coastal Plain soils. Second (Chapter 3), is to monitor dissolved and total P flow in surface runoff and groundwater flow through VFS to determine the relative importance, and mechanisms of dissolved P transport relative to total P. Data from this study will be used both to determine how VFS function in the Coastal Plain landscape, as well as to validate the P components of the Riparian Ecosystem Management Model (REMM). Finally, in Chapter 4 the P components of REMM will be validated for use in the Mid-Atlantic Coastal Plain region.

Eight Conservation Reserve Enhancement Program VFS were selected for evaluation, representing the major agricultural soils and variability in VFS design observed throughout Delaware. Soils were collected from VFS and adjacent agricultural fields at three depths, 0-15, 15-30, and 30-45 cm, and were characterized by routine analysis of the University of Delaware soil testing lab, as well as for soil adsorption/desorption properties. Deep tillage of the soil profile was then simulated by mixing equal portions of 0-15, 15-30, and 30-45 cm depth soils, and mixed soils were analyzed for P sorption characteristics. VFS establishment had little effect on the soil P source component of P loss, including soil test P (Mehlich-3 P, M3) and soil P saturation (measured as the M3-P saturation ratio, PSR) relative to adjacent agricultural fields. Simulated deep tillage of VFS soils to 45 cm decreased M3-P in mixed soils as much as 62%, compared to M3-P at the 0-15 cm soil depth, and achieved maximum soil PSR reductions of 70% as compared to 0-15 cm soil prior to mixing.

Vegetated Filter Strips were then established in two fields at the University of Delaware Research and Education Center, in Georgetown, DE. VFS in each field were further split into two plots, one deep tilled using a chisel plow and the other disk tilled prior to planting. Groundwater wells were installed and groundwater samples were collected weekly. Surface runoff was collected within 24 hours of each storm event. Water samples were analyzed for total P (TP), dissolved reactive P (DRP) nitrate (NO₃^--N) and ammonia (NH₄⁺-N). Chisel plowing was not effective at redistributing P from surface soils through the soil profile. Average surface runoff P concentrations (TP and DRP) were as much as an order of magnitude greater than groundwater P. Groundwater P concentration spiked shortly after storm events, and then quickly declined to background levels, though groundwater DRP concentration was rarely less than 0.1 mg P L^-1.

Phosphorus sorption and desorption equations were modified in REMM to reflect properties important to acid sandy Coastal Plain soils. Based on strong coefficients of determination, changing these equations would be expected to improve prediction of P transport through vegetated filter strips. Simulation data were significantly different from measured data, though simulated and measured data were similar with respect to environmental relevance.

Vegetated filter strips are an important practice to mitigate agricultural NPS pollution. In P saturated soils, dissolved P transport in subsurface flow remains an environmental concern. Further evaluation of methods to thoroughly mix the soil profile prior to planting should be investigated to improve retention of dissolved P by VFS. Models such as REMM can be helpful in designing VFS and improving placement within the Coastal Plain Landscape.