SIMULATED FLOOD OF A SMALL CONSTRUCTED FLOODPLAIN WETLAND IN VIRGINIA: EVENT-SCALE POLLUTANT ATTENUATION

Abstract. A 0.2 ha constructed floodplain wetland in northern Virginia was artificially flooded using a controlled hydrograph to simulate an overbank event in the fall and spring. The objectives were to (1) evaluate the event-scale nutrient attenuation capacity of a constructed floodplain wetland, (2) identify spatial variability in nutrient concentrations throughout the wetland during simulated overbank flow conditions, and (3) identify temporal variability in nutrient removal between the fall and spring events and within a controlled flood hydrograph. TSS was the constituent with the greatest attenuation in terms of percent removed (73% spring, 69% fall) and total mass removed, while NH 4 + had the greatest percent removal (54% spring, 58% fall) of the measured nutrient species. Greater removal of TP relative to PO 4 - suggests that particulate settling of sorbed P was the driving mechanism in P removal. Faster attenuation rates for NH 4 + and TN in the spring resulted in larger mass removal than in the fall. Linear regression of mass flux was used to determine pollutant removal rates between the wetland inlet and outlet. Adjusting paired inlet-outlet data using an offset for residence time removed some variability in percent removal, indicating that sampling protocols must be selected carefully in order to provide the most accurate results. Spatial variability of dissolved fractions of N and P varied (relative standard deviation between 12% and 47%) in both seasons, while TSS variability was higher in the fall (above 130%) and relatively lower in the spring (58%). Overall, these results suggest that construction or restoration of floodplain wetlands may be an effective way to manage stormwater nutrient and sediment at concentrations commonly found in streams affected by land use change due to urbanization and agriculture.