Nodularin together with cell numbers of Nodularia spumigena were monitored during a toxic bloom of this cyanobacterium during the summer/autumn of 1994-1995 in Lakes Alexandrina and Albert, South Australia. There was a good correlation between cell numbers and toxin concentration over the course of the bloom at a number of locations. The correlation was poorer toward the end of the bloom, and may have reflected cell lysis and release of toxin to the surrounding water where it was rapidly degraded. The results suggested that the suitability of the lake water for drinking could be better determined by monitoring toxin content rather than cell numbers. Degradation of nodularin was studied and found to be more rapid in lake water that was supporting a bloom compared with water collected that had not experienced a bloom for several months. Half-lives were 24 and 54 h, respectively. There was no lag phase in water that was supporting a bloom, suggesting a relatively high population of microorganisms capable of degrading the toxin to be present. Toxin released to the lake water from lysing cells would therefore disappear very quickly. Toxin degradation was inhibited in the presence of copper ions. Copper sulphate treatment of blooms may therefore be doubly disadvantageous. Not only are cells lysed by copper sulphate addition, thereby releasing toxin to the surrounding water, but the natural microbial population capable of degrading the toxin may be destroyed, or the enzymes capable of toxin degradation inhibited. (C) 1997 by John Wiley & Sons, Inc.
