A model study of the dynamics and controls on toxic cyanobacterial blooms in a temperate lagoon system

A 3-D coupled hydrodynamic-biogeochemical/ecological model was developed and validated for the coastal Gippsland Lakes system, Australia, to explore the relative importance of physical, chemical and biological controls on Nodularia blooms. Initially, winter/spring floods stimulated blooms of diatoms and dinoflagellates that were most accurately simulated when including salinity-dependent grazing. Subsequently, high carbon delivery to the sediment caused depleted bottom-water oxygen levels and large summer sediment phosphorus release. High summer phosphorus levels controlled the duration, size and severity of Nodularia blooms; however, temperature and salinity were the primary factors that determined bloom formation and location. A nutrient budget indicated that sediment phosphorus release, rather than catchment load, supplied most of the phosphorus required to support the summer Nodularia bloom. An assessment of sensitivity to model complexity identified that the model could only reproduce observed phosphorus release rates (and therefore Nodularia bloom size) when bioirrigation was implemented and zooplankton grazing excluded Nodularia. The model was additionally used to explore how different nutrient inputs and the effects of global warming and sea level rise could influence the dynamics of Nodularia blooms. Catchment nutrient load reduction was predicted to not significantly reduce the inorganic phosphorus available to Nodularia. This was because the sediment phosphate fluxes supplied about 60% of total phosphorus input to the lakes over the 2-year simulation period. Instead, nitrogen was more sensitive to catchment input. Nitrogen became more limited when the catchment load was reduced and it reduced the production rate of non-diazotrophic phytoplankton. Consequently, less phosphorus was assimilated by non-diazotrophic phytoplankton and lower concentrations of non-Nodularia phytoplankton also meant more abundant light available for Nodularia. On the other hand, more catchment nutrient inputs resulted in higher growth rates of non-diazotrophic phytoplankton. Although inorganic phosphorus concentrations also increased, the effect of enhanced shading by other fast growing phytoplankton ultimately limited the growth of Nodularia during the simulation period. In addition, increases in catchment nutrient load would certainly increase the phosphorus accumulation in the sediment. Unlike nitrogen, there is no chemical or biological removal reaction for phosphorus. As a result, nutrient load increase will potentially increase the severity of Nodularia blooms in the long term, although the bloom was less severe in the short term modelled results. This highlights the complexity associated with predicting toxic bloom formation and the different scales over which the driving processes occur. Scenarios were also run to simulate different entrance management options. Widening the existing entrance or construction of a second entrance might be able to prevent Nodularia blooms by increasing the salinity in the lakes more quickly after a flood. In contrast, model results have shown that global warming and sea level rise could enhance the sediment denitrification rate, increase the sediment phosphate fluxes, and provide larger suitable habitat for the growth of Nodularia in the Gippsland Lakes.