Mechanisms controlling nitrogen removal in agricultural headwater streams

The export of nitrogen from agricultural watersheds in the Midwestern United States is a large environmental and human health concern. Denitrification, a natural nitrogen removal process, is thought to be higher in low order streams compared to larger streams and rivers. The objective of my dissertation was to determine what controls denitrification, examine hydraulics that affect nitrogen transport, and assess nitrogen removal in agricultural headwater streams with varying riparian land use. Paired 50-meter reaches (one reach with agricultural/urban land use and the other with forested land use) were identified along five headwater streams in the Upper Sugar Creek Watershed in Northeast Ohio, USA (40°51'42''N, 81°50'29''W).; Three sites were examined to determine if temperature, nitrate, organic carbon, or pH affect denitrification. Organic carbon amendments and pH had no effect on denitrification while temperature and nitrate did. Denitrification followed Michaelis-Menten type kinetics, and the calculated km value (1.0 mg NO3--N L-1) was below the median in-stream NO3- concentrations (5.2 - 17.4 mg NO3--N L-1), suggesting denitrification was nitrate saturated.; Solute transport variables were measured in all ten reaches on seven different occasions (n = 70). Relative transient storage was high but similar in ag/urban and forested reaches (2.2 +/- 2.5 and 1.5 +/- 1.3, respectively). Longitudinal dispersion coefficient (ANOVA; F1,69 = 17.087; p < 0.01) and dispersivity (ANOVA; F1,69 = 8.157; p < 0.01) were higher in ag/urban reaches. Overall, hydraulics in ag/urban sites were more dynamic than the forested sites, most likely a result of inundated riparian grasses in the ag/urban reaches.; At all ten reaches, nitrogen removal metrics and stream characteristics were measured during spring, summer, and fall. Although denitrification rates were high (< 0.1 - 17.2 mg N m-2 hr-1), annual nitrogen removal was low. The ag/urban and forested reaches had similar removal capacities despite forested reaches having higher hydraulic residence times (ANOVA; F1,59 = 8.362, p = 0.005). Temperature, in-stream nitrate concentrations, and relative transient storage were identified as variables controlling nitrogen removal. In-stream nitrogen removal was low and riparian land use had no effect on removal capacity because of high in-stream nitrate.