Numerical models for management of anabaena circinalis

Abstract

The control of nuisance species of cyanobacteria in reservoirs is a critical issue for the internationalwater industry, as these organisms can produce toxins and compounds that taint potablewater with unpleasant tastes and odours. To assist with effective management of cyanobacterial growth, numerical models that are either site specific or universally applicable can be employed. An artificially destratified reservoir was modelled with the coupled hydrodynamicecological numerical model DYRESM-CAEDYM. The validation site was Myponga Reservoir, South Australia, a highly managed drinking water supply reservoir. Chemical dosing (CuSO4) and artificial mixing via an aerator and two raft-mounted mechanical surface mixers (hereafter referred to as surface mixers) are used at Myponga to manage the growth of the scum-forming cyanobacterium Anabaena circinalis. The dominant phytoplankton community was adequately modelled, and combinations of the various management options were simulated whereupon informed operational strategies could be implemented. Without any intervention, permanent stratification would occur and the growth of Anabaena circinalis would peak above 3 μg L−1, producing geosmin that would exceed the taste and odour threshold (10 ng L−1); with the individual use of the aerator or surface mixers, growth of Anabaena circinalis was significantly reduced to below 1 μg L−1.