Experimental design and samples collection
As previously described, an important point was the necessity to evaluate the removal efficiency of D. polymorpha independently from any other settling process, which would remove the metals adsorbed on suspended solids. The
filtering action of D. polymorpha was first evaluated on the effluent outflowing secondary settling.
However, since this effluent had a very low COD (10 mg/L) and,consequently, a low suspended solid concentration (on which a relevant amount of MEs is normally adsorbed) [39,40], the following tests were performed with three other different wastewater mixtures, previously filtered through a 1 mm mesh bag filter to remove coarse matter. This allowed us to evaluate the filtration efficiency of D. polymorpha on wastewater with pollutin load and different amounts of suspended particulate, also taking into account that this bivalve selects particles for food with a diameter ranging between 15 and 40 mm [41]. The mixtures used in the tests, in addition to 100% outlet, are the following: 25% inlet/
75% outlet, 50% inlet/50% outlet and 100% inlet (wastewater incoming at WWTP). The ME removal evidence from wastewater were carried out through the measurement of their concentrations in the water samples taken from the pilot-plant with bivalves inside; at the same time, control tests were conducted into the pilot-plant without adhering animals. All tests were performed in triplicate. The ME removal progress was monitored for 4 h, by sampling the wastewaters every 30 min, which enabled to obtain the removal slope for each ME. We chose to evaluate the ME removal within 4 h, taking into account that the treated wastewaters remain in the Milano-Nosedo WWTP for about 24 h; thus,the selected time seemed to be a fair compromise in view of integrating the conventional treatment with limited dimensional requirements. To check the practicability of such assumptions, we carried out further final tests in single for a period of 24 h, taking
only two samples, one at the beginning and one at the end of the tests. The tests were conducted with an initial
flow rate corresponding to 3500 L/h, which would imply 18 min contact time, recirculating the effluent in the pilot-plant 84 times to obtain an overall 24 h contact time. After each test, the entire pilot-plant was washed with tap water, to avoid memory-effects related to the previous tests. For this reason, to minimize this problem, as well as to decrease the bivalve stress, the test schedule started with the most diluted waste (100% outlet) and gradually increased its concentration until 100% inlet. We monitored the wastewater temperature both at the beginning and at the end of each test in
order to take into account its possible interference with the filtration activity of zebra mussels. The wastewater temperature within the pilot-plant during the spring season ranged from 14 to 24 C, comparable with the optimal values for D. polymorpha filtration activity (10–20 C) [42]; we can thus exclude any negative interference of temperature on the
filtration-removal process. Samples were taken from the pilot-plant at the selected times by the use of a 250 mL plastic bottles, acidified with 1% of HNO3 and stored at 4 C at dark until analysis.