Using qPCR and high-resolution sensor data to model a multi-species Pseudo-nitzschia (Bacillariophyceae) bloom in southeastern Australia

•Report on a multispecies bloom of Pseudo-nitzschia in SE Australia.•P. cf. cuspidata was the only species to produced domoic acid.•Successful development of a novel qPCR assay to detect P. pseudodelicatissima complex Clade I.•Modeling revealed temperature and/or salinity more predictive of cell abundance than rainfall.•A combination of tools (qPCR, sensor) can provide a more early warning of Pseudo-nitzschia blooms. Harmful algal blooms, including those caused by the toxic diatom Pseudo-nitzschia, can have significant impacts on human health, ecosystem functioning and ultimately food security. In the current study we characterized a bloom of species of Pseudo-nitzschia that occurred in a south-eastern Australian oyster-growing estuary in 2019. Using light microscopy, combined with molecular (ITS/5.8S and LSU D1-D3 rDNA regions) and toxicological evidence, we observed the bloom to consist of multiple species of Pseudo-nitzschia including P. cf. cuspidata, P. hasleana, P. fraudulenta and P. multiseries, with P. cf. cuspidata being the only species that produced domoic acid (3.1 pg DA per cell). As several species of Pseudo-nitzschia co-occurred, only one of which produced DA, we developed a rapid, sensitive and efficient quantitative real-time polymerase chain reaction (qPCR) assay to detect only species belonging to the P. pseudodelicatissima complex Clade I, to which P. cf. cuspidata belongs, and this indicated that P. cuspidata or closely related strains may have dominated the Pseudo-nitzschia community at this time. Finally, using high resolution water temperature and salinity sensor data, we modeled the relationship between light microscopy determined abundance of P. delicatissima group and environmental variables (temperature, salinity, rainfall) at two sites within the estuary. A total of eight General Linear Models (GLMs) explaining between 9 and 54% of the deviance suggested that the temperature (increasing) and/or salinity (decreasing) data were generally more predictive of high cell concentrations than the rainfall data at both sites, and that overall, cell concentrations were more predictive at the more oceanic site than the more upstream site, using this method. We conclude that the combination of rapid molecular methods such as qPCR and real-time sensor data modeling, can provide a more rapid and effective early warning of harmful algal blooms of species of Pseudo-nitzschia, resulting in more beneficial regulatory and management outcomes.