Informing the management of harmful algal blooms in estuaries

Global change pressures negatively impact the water quality of estuaries, often culminating in the proliferation of eutrophic symptoms that hinder the provision of important ecosystem services. The formation of harmful algal blooms (HABs) is a common ecosystem response in estuaries prone to eutrophication, typically resulting in adverse impacts on the quality and biotic functionality of the effected environment. Heterosigma akashiwo is one of the most problematic and widespread HAB species. It has been documented to cause severe economic losses in the fishery and aquaculture sector, with documented impacts reported in 17 countries. Recurrent high-biomass HABs comprising H. akashiwo are a feature of the permanently eutrophic Sundays and Swartkops estuaries in South Africa. In these case study ecosystems, these HABs have been shown to negatively impact the development of fish larvae and damage the gills of adult fish, while also being responsible for causing mass fish mortality events . This study investigated these sites and developed a predictive tool for H. akashiwo blooms to assist in water quality and catchment management. In-situ monitoring data were used to build an artificial neural network model to indicate which abiotic variables can be used to predict the presence or absence of H. akashiwo in each respective estuary. Additional in-situ data were then collected following a natural freshwater pulse event to determine whether (1) these hydrological resetting events serve to displace H. akashiwo populations and (2) how long it takes for populations to reestablish in the estuaries. The neural network model found that H. akashiwo presence was influenced by inorganic nutrient and light availability, while pH also played a role. For the Sundays Estuary, the presence of H. akashiwo was closely linked to depth and salinity, while dissolved silica had the largest influence in the Swartkops Estuary despite not being a nutrient required by H. akashiwo. Additionally, freshwater pulse events were largely ineffective as a long-term removal mechanism for H. akashiwo, with high-biomass HABs of the species returning within 4-5 months of the initial flood peak in both estuaries. Therefore, management of estuaries cannot rely solely on the natural flushing of an estuary to remove HABs. Instead, catchment management and water quality improvement are needed to reduce nutrient inputs and ensure freshwater inflow variability.