A commentary on the modelling of the causal linkages among nutrient loading, harmful algal blooms, and hypoxia patterns in Lake Erie

In this study, our primary aim is to evaluate the capacity of past and current modelling efforts to depict the causal relationships between major water quality indicators (e.g., chlorophyll a, harmful algal blooms, dissolved oxygen) and nutrient loading in Lake Erie. We first conduct a review of nearly all the modelling projects documented in the pertinent literature, and then evaluate the performance of six of these models applied over the past thirty years. We examine the strengths and weaknesses of the different modelling strategies, their adequacy in representing the processes underlying plankton dynamics, and their ability to reproduce the spatiotemporal variability in hypoxia or harmful algal blooms. Our analysis shows that these models have mainly offered heuristic tools to examine different ecological hypotheses and dictate future data collection efforts. Our study critically discusses the most appropriate next steps to improve the reproduction of the spatiotemporal patterns of major phytoplankton groups, e.g., cyanobacteria, the functional role of dreissenid mussels, and the relative importance of diagenesis processes on the manifestation of hypoxia in Lake Erie. Finally, we advocate the standpoint that a single “correct” strategy does not exist, and therefore we should strive for a synthesis of multiple modelling approaches which can contribute to an integrative view on the functioning of the system.