Tests were carried out under controlled conditions in the Experimental Plant of Viville (Arlon, Belgium) to enhance the purification of urban wastewater by "natural" means. The results demonstrate the need to structure treatment systems in a series of different artificial ecosystems (or a Hierarchical Mosaic of Artificial Ecosystems -MHEA in French). The first two levels we used were made up of an unplanted aquatic ecosystem (stabilization pond) followed by a semi-aquatic ecosystem planted with Typha lat(folia L. in which the water flows over the substrate. At a flow rate of 4 m2/PE (l PE = 150 l/day of typical urban wastewaters in Belgian rural zones), this first stage substantially reduces suspended solids (SS), COD and BOD 5, a significant amount of tot-N and tot-P, and reduces pathogens by 100-fold. Further, the system is easy to manage (sludge is eliminated in the first stage and biomass is collected in the second stage) and the treatment system does not clog up.Nevertheless, real and sustainable environmental protection demands even higher performance rates, and these first two stages, both in terms of design and dimension, can only be considered as a satisfactory part of a MHEA system. Artificial aquatic, semi-aquatic, and terrestrial ecosystems were systematically compared at the third and fourth stage of the system to increase the overall removal efficiency.The most complete and efficient system in our tests (i.e., the one that provides the most successful primary (SS), secondary (COD and BOD 5) and tertiary (N and P) treatment and the best pathogens removal rates) was made up of 3 sequential series of ecosystems: an aquatic ecosystem whose flow went into a planted Typha latifolia system (surface water flow), that flowed into a terrestrial ecosystem planted with ,41nus glutinosa (L.) Gaertn (vertical subsurface water flow). A total surface area (stages 1-4) of 8 m2/PE ensured a high performance level whose outflow conformed to the strictest European norms.

No abstract