Nitrate (NO3) loadings from stormwater runoff promote eutrophication in surface waters. Low Impact Development (LID) is a type of best management practice aimed at restoring the hydrologic function of watersheds and removing contaminants before they are discharged into ground and surface waters. Also known as rain gardens, a bioretention system is a LID technology that is capable of increasing infliltration, reducing runoff rates and removing pollutants. They can be planted with visually appealing vegetation, which plays a role in nutrient uptake. A modified bioretention system incorporates a submerged internal water storage zone (IWSZ) that includes an electron donor to support denitrification. Modified (or denitrifying) bioretention systems have been shown to be capable of converting NO3 in stormwater runoff to nitrogen gas through denitrification; however, design guidelines are lacking for these systems, particularly under Florida-specific hydrologic conditions. The experimental portion of this research investigated the performance of denitrifying bioretention systems with varying IWSZ medium types, IWSZ depths, hydraulic loading rates and antecedent dry conditions (ADCs). Microcosm studies were performed to compare denitrification rates using wood chips, gravel, sand, and mixtures of wood chips with sand or gravel media. The microcosm study revealed that carbon-containing media, acclimated media and lower initial dissolved oxygen concentrations will enhance NO3 removal rates. The gravel-wood medium was observed to have high NO3 removal rates and low final dissolved organic carbon concentrations compared to the other media types. The gravel-wood medium was selected for subsequent storm event and tracer studies, which incorporated three completely submerged columns with varying depths. Even though the columns were operated under equivalent detention times, greater NO3 removal efficiencies were observed in the taller compared to the shorter columns. Tracer studies revealed this phenomenon was attributed to the improved hydraulic performance in the taller compared to shorter columns. In addition, greater NO3 removal efficiencies were observed with an increase in ADCs, where ADCs were positively correlated with dissolved organic carbon concentrations. Data from the experimental portion of this study, additional hydraulic modeling development for the unsaturated layer and unsaturated layer data from other studies were combined to create nitrogen loading model for modified bioretention systems. The processes incorporated into the IWSZ model include denitrification, dispersion, organic media hydrolysis, oxygen inhibition, bio-available organic carbon limitation and Total Kjeldahl Nitrogen (TKN) leaching. For the hydraulic component, a unifying equation was developed to approximate unsaturated and saturated flow rates. The hydraulic modeling results indicate that during ADCs, greater storage capacities are available in taller compared to shorter IWSZs Data from another study was used to develop a pseudo-nitrification model for the unsaturated layer. A hypothetical case study was then conducted with SWMM-5 software to evaluate nitrogen loadings from various modified bioretention system designs that have equal IWSZ volumes. The results indicate that bioretention systems with taller IWSZs remove greater NO3 loadings, which was likely due to the greater hydraulic performance in the taller compared to shorter IWSZ designs. However, the systems with the shorter IWSZs removed greater TKN and total nitrogen loadings due to the larger unsaturated layer volumes in the shorter IWSZ designs.

Turfgrass management activities on golf courses have been identified as a possible source of Chesapeake Bay nutrient pollution. Total Maximum Daily Load goals are in place to reduce nutrient amounts entering the Bay. Dissertation investigations include (1) the role of golf course turfgrass management in nutrient deposition or attenuation in local streams, (2) estimations of total nitrogen (N) discharging to the watershed from stream outlet points as a function of land use and watershed area, and (3) other factors potentially affecting water quality on golf courses, including soil characteristics and use of best management practices (BMPs). Total N, nitrate-N, ammonium-N, phosphate-phosphorus (P), streamwater temperature, specific conductance (SpC), pH and dissolved oxygen (DO) were sampled at 12-14 golf course stream sites in the James River and Roanoke River watersheds during baseflow conditions. Discharge was determined at outflow locations. Unit-area loads (UALs) were calculated from monitoring data. These UALs were then compared to UALs from Chesapeake Bay Watershed Model land use acreages and simulated loads for corresponding watershed segments. Virginia golf course superintendents were also surveyed to determine BMP use. No consistent impairment trends were detected for streamwater temperature, SpC, pH, or DO at any of the sites. Outflow NO3-N was below the 10 mg L-1 EPA drinking water standard. However, some sites may be at increased risk for benthic impairment with total N concentrations >2 mg L-1, as suggested by VADEQ. Significant increases in nitrate-N at OUT locations were measured at four sites, whereas decreases were measured at two sites. Ammonium-N significantly decreased at two sites. Golf course N UALs calculated from baseflow monitoring were lower than or similar to UALs estimated for forested areas in the associated watershed segment at seven out of the 12 sites. Golf course UALs ranged from 1.3-87 kg N ha-1 yr-1. Twenty-one of 32 surveyed BMPs had an adoption rate ≥50% among survey respondents. In most cases, presence of golf courses generally does not appear to significantly degrade baseflow water quality of streams in this study. Management level appears to be an influencing factor on water quality and concerns may be heightened in urban areas. / Ph. D.