Investigating the applicability of the CEGAP model to predict the development of harmful algal blooms in the Klipvoor Dam

Abstract

Africa is a water scarce country depending primarily on reservoirs and lakes for socio-economic wellbeing. Most of these reservoirs are contaminated with nutrients making them either eutrophic or even hypertrophic. The algae blooms are common amongst these reservoirs. There is however a worrying trend of an increasing number of reservoirs with cyanobacterial blooms. Cyanobacterial blooms produce cyanotoxins which may result in human and animal deaths. Therefore, it is important to have a reliable tool that can predict emergence of these blooms for the South African reservoirs. The CEGAP model has been developed for hypertrophic reservoirs of South African to predict the emergence of Microcystis spp. blooms. The objective of this research was to test the applicability of the CEGAP model at Klipvoor Dam. Ten year data from the Department of Water and Sanitation database was used to assess the applicability and accuracy of the CEGAP model to predict occurrence of the cyanobacterial blooms for the Klipvoor Dam. The model was used to predict emergence of the Microcystis spp. blooms for real time, 14 days and 28 day forward predictions. The results show that the model was able to predict the emergence of the algae blooms for the real time and 14-day forward predictions. The model could not predict the emergence of blooms for 28-day forward predictions reliably. The model could predict the emergence of bloom events with reasonable accuracy but could not predict accurately the severity of the blooms. The results indicate that the model is able to predict the emergence of the Microcystis biomass but the severity of the biomass may be influenced by other algae species, which in certain instances may be larger in size and extent than the Microcystis species, and hence contribute a larger percentage of the biomass and chlorophyll-a concentration, even if they are low in terms of the species counts. It was therefore recommended that the models ability to predict the severity of the blooms and its ability to predict the emergence of other algae species blooms such as Ceratium spp. be investigated further.