Nitrogen retention in mature constructed wetlands

Abstract

Constructed wetlands are created for the purpose of treating anthropogenic discharges, such as agricultural and urban runoff to reduce the potential of undesirable effects in receiving waters. Leikvollbekken is a mature constructed wetland located in north-west of Store Stokkavatnet in Stavanger Municipality. The wetland is a two-pond free water surface system constructed with the aim to reduce excessive nitrogen and phosphorus in the pelvis before reaching Store Stokkavatnet. The main focus of this thesis was to monitor the constructed wetland with respect to nitrogen. Nitrogen removal in the wetland was believed to occur through biological assimilation and dissimilation. The hypothesis was that the degree of retention would depend on flow, resulting in a positive retention over time. With some exceptions, weekly water samples were collected in the period October 2018 to May 2019 in inlet, mid-pond and outlet of the wetland. Flow in and out of the wetland was measured with an interval of 15 minutes during the thesis period by an integrated flow meter at site. Grab samples from Store Stokkavatn and Madlabekken were included in a period to compare concentrations with water in Leikvollbekken. In addition, a storm event was included to investigate the effect of high hydraulic loadings on influent concentrations and concentrations in the wetland. Water samples were analyzed for total nitrogen, nitrate and ammonium. Total and fixed suspended solids, pH, conductivity, alkalinity and color were included to investigate if any significant correlations existed with nitrogen concentrations in the samples taken. The overall results showed higher concentrations in Leikvollbekken compared to Store Stokkavatn and Madlabekken. A positive retention of 9.4 % TN (68.6 kg), 6.1 % NO3- (33.8 kg) and 87.1% NH4+ (7.2 kg) was observed during the period studied. Highest concentrations were observed during the storm event. An interesting observation was the increasing concentrations of TN, NO3- and NH4+ in mid-pond and outlet samples in April and during the storm event, indicating contribution from additional sources than the inlet water. No significant correlation between average flow per sampling and total nitrogen was found. Moreover, no connection was found between nitrogen compounds and the additional parameters except conductivity and color.