Ammonium removal in constructed wetlands with recirculating subsurface flow: Removal rates and mechanisms

From June 1993 through February 1994, the removal of NH4-N was evaluated in constructed wetlands at the TVA constructed wetland research facility in Muscle Shoals, AL. The objectives were to determine rates for NH4-N removal and speculate on potential mechanisms for removal. Nine constructed wetland cells were used with approximate dimensions of 9.1 x 6.1 x 0.6 m(3) and a recirculating subsurface flow system in a gravel base. Treatments consisted of an unplanted (WO=control) and two polycultural planting schemes (P1=Scirpus acutus, Phragmites communis and Phalaris arundinacea; P2=Typha sp., Scirpus atrovirens georgianus and Scirpus cyperinus) replicated 3 times. Salt solutions were added and recirculated in each cell resulting in initial concentrations of 50 and 300 mg l(-1) of NK4-N and COD, respectively, when fully diluted with wetland water. Salts were added to wetlands approximately every 6 weeks with the first addition on June 1, 1993 and the last addition on February 9, 1994 for a total of 6 time periods (times I, II, III, IV, V and VI). The COD of the waters was removed at rates ranging from 5.5 to 10 g/m(2)/d during times I through IV with no discernible difference amongst the planting treatments. Wetland cells with PI were more efficient at removing NH4-N (1.1 g/m(2)/d) than P2 (0.6 g/m(2)/d) or WO (0.5 g/m(2)/d) at time I with differences decreasing by time IV (0.3 to 0.7 g/m(2)/d). During the winter (times V and VI), there were no differences in NH4-N removal amongst planting treatments with an average removal rate of 0.35 g/m(2)/d. There was a seasonal change in NH4-N removal in all the treatments, with the change most noticeable in the planted cells. The removal of NH4-N in WO was speculated to be due to a combination of sorption onto gravel, microbial assimilation, and nitrification at the air-water interface, The extra NH4-N removal in the planted cells diminished in the winter because the removal was most likely due to a combination of enhanced nitrification from O-2 transport and NH4-N uptake mediated by seasonal macrophyte growth.