Removal efficiency of a constructed wetland in presence and absence of macrophyte aboveground parts.

MAINE, M. A.; HADAD H. R.; SANCHEZ, G. C.; CAFFARATTI, S. E.; PEDRO, M. C.; DI LUCA, G. A.; MUFARREGE, M. M.

Praga

Simposio; “Joint meeting of society of wetland scientists, wetpol and wetland and biogeochemistry symposium”; 2011

Wetpol

A 2,000 m2 free water surface wetland was constructed to treat wastewater containing metals (Cr, Ni, Zn and Fe) and nutrients from a tool factory in Argentina. Industrial wastewater and sewage were treated together. The emergent macrophyte Typha domingensis became the dominant species and covered 80 % of the wetland surface since February 2005. However, in June 2009, aboveground parts of plants were depredated by capybaras (South American amphibious rodents), living on the same site where is placed the constructed wetland. The wetland looked like a non-vegetated pond. However, the plant roots and rhizomes were not damaged. In November 2009, to allow the recovery of the vegetation, a simple perimeter protection system was installed to stop the approaching of animals. Subsequently, T. domingensis showed a luxuriant growth that was also enhanced by the growth season and the nutrient from the sewage, reaching a cover of 60 % after 30 days. We compared the removal efficiency of the system planted with T. domingensis during normal operation (February 2005-May 2009) and the same system without aboveground parts of plants, during the eventual problem (June-November 2009). Monthly, samplings of the influent and effluent, sediment and macrophytes were performed. The CW efficiently retained contaminants. There were not significant differences between the performance of the wetland operating in normal conditions and the wetland operating with temporal absence of plant aerial parts. The mean removal percents, expressed as: % during normal operation / %during the event, were    48.1/53.0 % Ni, 74.8/82.4 % Fe, 81.7/86.5 % NO3- and 82.0/86.4 % NO2-, 67.8/65.0 % BOD and 68.3/64.4 % COD. Regarding Cr, the mean removal presented differences between normal operation (49.0 %) and the stage without aboveground parts (66.5 %). Soluble reactive phosphorus (SRP) and NH4+ were not efficiently removed in both periods; probably due to dissolved oxygen (DO) concentration was low at the inlet and at the outlet, in roughly half of the samplings. Along the study, Cr, Ni, Zn and total phosphorus (TP) concentrations in the inlet sediment were significantly higher than that of the outlet sediment, demonstrating that they are efficiently retained in the wetland. During the normal stage, metal concentrations were higher in belowground than in aboveground parts, suggesting scarce translocation. Cr, Ni and Zn concentrations in belowground parts were higher during the period of plants without aboveground parts in comparison with the concentrations measured during the normal operation. Higher Cr, Ni and Zn concentrations were attained simultaneously in belowground parts of T. domingensis during the period of plants without aerial parts (1.2-1.6, 1.0-1.6 and 0.15-0.3 mg/g d.w., respectively) than in the normal period (0.015-0.979, 0.005-1.04, 0.022-0.319 mg g-1 d.w., respectively). Selective allocation to belowground parts in absence of aboveground parts is hypothesized. These results demonstrated that faced to an eventual problem, this wetland was capable to maintain its efficiency and to recover its vegetation. Our results proved the importance of the rhizospheric system in the dynamics of the wetland and the robustness of these treatment systems.