Denitrification in constructed wetlands can be very important in the treatment of swine lagoon effluent when land application areas are limited. The objectives of this investigation were to determine (i) the denitrification enzyme activity (DEA) in the marsh sediments of marshpond-marsh (MPM) constructed wetlands, (ii) changes in DEA with additions of carbon and nitrate, and (iii) the response of DEA to different wastewater N loading rates. Swine wastewater was applied to six MPM wetlands located at North Carolina A&T State University, Greensboro, NC, at rates of 4 to 35 kg N ha 21 d 21 . Soil samples were obtained from the top 25 mm of the marsh sections on four dates for determination of DEA via the acetylene blockage method (blocked at N 2 O). Headspace N 2 O was measured via gas chromatography. In the control treatment, they ranged from 0.06 to 1.13 and 0.16 to 0.79 mg N 2 O-N kg 21 soil hr 21 in the first and second marshes, respectively. In both marshes, the DEA rate was significantly increased with the addition of nitrate but not by glucose, indicating that nitrate was a clear limiting factor for denitrification. The DEA in both the control and the amended treatments increased dramatically with increased wastewater N loading, and the increases were generally more pronounced in the first marsh. The DEA values produced in the absence of acetylene blockage did not increase with wastewater N loading rate. Denitrification enzyme activity levels in the marsh sections of the MPM were generally consistent with a highly denitrifying environment. C ONSTRUCTED WETLANDS have been investigated and used for treatment of animal waste in the USA and around the world for over 10 yr (Cathcart et al., 1994;McCaskey et al., 1994). The understanding of their treatment processes, designs, and operations have advanced significantly (Kadlec and Knight, 1995;Knight et al., 2000;Hunt et al., 2002;Stone et al., 2002). Constructed wetlands have been found to generally provide somewhat limited P removal capacities but significant N removal capacities (Knight et al., 2000;Hunt et al., 2002). The effectiveness of N removal has varied considerably with wetland design and loading rates (Poach et al., 2003(Poach et al., , 2004b. Their research indicated that the continuous marsh wetlands were able to treat higher loads of N than the MPM wetlands. Furthermore, the continuous marshes were subject to less ammonia volatilization than the MPM wetlands (Poach et al., 2002(Poach et al., , 2004a.While soil accumulation and plant uptake of N were important at low N loading rates , they were responsible for smaller portions of the removal at higher N loading rates ($10 kg ha 21 d
21) . At the higher loading rates, denitrification became the apparent predominant pathway for N removal. Hunt et al. (2003) reported high levels of DEA in continuous marsh wetlands used to treat swine wastewater. These wetlands had sloped bottoms, and denitrification was higher at the upper slope area where oxygen was more available to facilitate nitrification of ...