Evaluation of Nitrogen Removal Potential in a Reed Filter Bed System by the 15N Tracer Method

Abstract

The mechanisms of nutrient removal in a sand-filled reed filter bed (RFB) system involve filtration, sedimentation, plant uptake and microbial degradation in the soil-water interface and plant rhizosphere where the nitrification-denitrification process occurs. Information concerning these mechanisms, however, is limited, especially regarding the denitrification rate in a RFB system. With this in mind, the denitrification rates were directly measured using the 15N tracer method. In addition, the seasonal changes of the denitrification rates were investigated. Using a bench-scale RFB system and sand filter system without reeds, denitrification rates ranging from 0.0 to 5.5 (average 2.27) and 0.0 to 3.4 (average 1.48) g-N m2 d, respectively, were obtained. These values indicate that about 60 and 80% of inflow nitrogen loading was removed by denitrification. It was also found that 25% of inflow nitrogen loading in the RFB system was removed by plant uptake. The higher denitrification rate in the RFB system suggested that the microbial activity in the reed rhizosphere is of key importance for the nitrogen removal efficiency during the stationary phase of the reed's growth curve. As for their associated seasonal changes, the denitrification rates were proportional to air temperature, so that the denitrification rates were lowest during the winter. Although the denitrification rate in the RFB system was only 0.23 to 0.68 g-N/m2 d at 6.2 to 11.2 °C in December 1998, the increase of air temperature to 26.2 °C using a temperature-controlled room remarkably enhanced the denitrification rate to 2.95 g-N/m2 d. This result confirmed the possibility that the RFB system could maintain denitrification activity by controlling air temperature.