To assess the elimination potential of slow sand filters for cyanobacterial hepatotoxins (microcystins), two full-scale experiments were conducted using the German Federal Environment Agency's experimental field in Berlin, Germany. One experiment was carried out with dissolved microcystins extracted from a cyanobacterial bloom on one of Berlin's lakes, dosed as short-term, single-pulse application. The other experiment simulated natural conditions more closely, with a longer-term exposure of the filter to living cyanobacterial cells (collected from the same lake) so that most toxins were initially contained inside the cells. The microcystins were detected by ELISA and HPLC/photodiode array detector and subsequently identified by MALDI-TOF MS. The experiment with dissolved microcystins yielded very high elimination rates (>95%) inside the filter bed attributed to biodegradation, whereas retardation by adsorption was low. The obtained half-lives for the microcystins detected by ELISA were about 1 h. The second experiment, which was with mostly cell-bound microcystins, showed similar results during the first days after application of cyanobacteria (elimination >85%). As the population declined in late autumn, the proportion of extracellular to cell-bound microcystins increased. At the same time the elimination rates declined to values <60%. This decline is most likely attributable to retarded biodegradation at temperatures of <4 degrees C. Altogether the results of the experiments show that under moderate temperatures, with an intact schmutzdecke (biofilm) with previous contact with microcystins, slow sand filtration is an effective treatment for eliminating microcystins from drinking water.
Copyright 2002 Wiley Periodicals, Inc.