An assessment of the nutrient stripping function of two constructed wetlands in the Swan-Canning Estuary

Abstract

The use of constructed wetlands and wet detention basins has proven to be highly effective in removing pollutants from industrial discharges and stormwater runoff throughout the world. This is attributed to design of the key treatment components in a constructed wetland, catchment source characteristics and climatic conditions. A disproportionate amount of research and monitoring effort has gone into constructed wetlands due to their cost effectiveness and ability to optimize multiple benefits. In Western Australia, several wetland monitoring studies on the role of constructed wetlands especially in Swan-Canning estuary have been done, but often do not address their design efficiencies in stormwater treatment. Two wetlands (Liege St and Tom Bateman wetland) constructed for nutrient stripping proximal to the Swan-Canning estuary have been monitored for two years. Liege St wetland was constructed to reduce the nutrients reaching the Canning River directly and improve the amenity value of the area. Similarly, Tom Bateman wetland was constructed to reduce nutrients of the Banister Creek catchment draining into the Canning River as well as for stormwater management and habitat use. Physicochemical and biological indicators were used to assess the nutrient stripping efficiency of the wetlands. In some cases, data from previous studies were used to determine the health and viability of the selected wetland sites. The limnological indicators used included; dissolved oxygen, pH, water temperature, electrical conductivity and nutrient levels. The biological included; bacteria, nutrients and chlorophyll in periphyton, macroinvertebrates and diatoms. Differences in the community structure of periphyton, macroinvertebrates and water quality were found from the inlet to the outlet in both Liege St and Tom Bateman wetlands.Despite the poor water quality, Liege St wetland exhibited significant nutrient removal efficiencies for TP while Tom Bateman wetland had very high removal efficiency for TN. The TP removal in Liege St wetland was attributed to the design of key treatment components which included a gross pollutant trap, concrete lined sedimentation pond, vegetated sumplands, weirs and clay lining for the wetland bed. In contrast, Tom Bateman wetland lacked the above key treatment components. Additionally, the wetland experienced short-circuiting especially during high flow periods. The high TN removal in Tom Bateman wetland was attributed to assimilation by plants and micro-organisms especially by the dense growth of Potamogeton crispus observed on the wetland floor and the non- biological transformation processes such as volatilisation, sorption and sedimentation. The poor water quality of the inflow in both wetlands was attributed to catchment characteristics which were not fully investigated in this study. In an attempt to improve the nutrient stripping function of Liege St and Tom Bateman wetland, changes to the wetland design and routine maintenance were suggested for Tom Bateman and Liege St wetland respectively. Also the use of the Swan-Canning Cleanup Programe (SCCP) water quality targets as opposed to the ANZECC trigger values in water quality assessments in constructed wetlands in the Swan-Canning estuary is suggested among others.