Generalized linear models to assess environmental drivers of paralytic shellfish toxin blooms (Southeast Tasmania, Australia)

Abstract

Paralytic shellfish toxin (PST) events are often difficult to predict because synergistic processes, involving multiple temporal and spatial scales, are responsible for their development. In this study, we consider a suite of factors, from local to regional processes, driving paralytic shellfish toxin (PST) events associated with Alexandrium catenella on the east coast of Tasmania (Australia). PST concentrations in shellfish were compared against environmental variables, including daily wind speed and direction, satellite chlorophyll-a, sea surface temperature and river discharge. Relationships were tested using generalized linear models (GLMs). GLM coefficients confirm that high PST events were significantly favored by cooler temperatures and higher chlorophyll-a concentrations. PST events were seasonal, occurring between August and November and showed a significant relationship with sustained coastal upwelling and wind transition events after one to three months of predominantly upwelling winds. Satellite imagery shows evidence of cross shelf transport moderating PST events. Overall, the results suggest that a combination of factors (currents, upwelling conditions and changes in sea surface temperature) work together to initiate and moderate PST events. The use of GLMs provides a useful tool for understanding synergistic relationships. These techniques are of value for developing reliable predictive models to forecast and manage the effects of harmful and costly PST events.