Integrating phytoplankton phenology, traits, and model-data fusion to advance bloom prediction

While there is a diversity of approaches for modeling phytoplankton blooms, their accuracy in predicting the onset and manifestation of a bloom is still lagging behind what is needed to support effective management. We outline a framework that integrates trait theory and ecosystem modeling to improve bloom prediction. This framework builds on the concept that the phenology of blooms is determined by the dynamic interaction between the environment and traits within the phytoplankton community. Phytoplankton groups exhibit a collection of traits that govern the interplay of processes that ultimately control the phases of bloom initiation, maintenance, and collapse. An example of process-trait mapping is used to demonstrate a more consistent approach to bloom model parameterization that allows better alignment with models and laboratory- and ecosystem-scale datasets. Further approaches linking statistical-mechanistic models to trait parameter databases are discussed as a way to help optimize models to better simulate bloom phenology and allow them to support a wider range of management needs.