Fish-killing microalgal blooms are responsible for much greater global socio-economic impacts
than the well-studied HAB species causing seafood biotoxin contamination. Examples are the 1972 Chattonella marina bloom in the Seto Inland Sea, Japan (estimated USD 71M loss to yellowtail aquaculture), the 1988 Prymnesium polylepis bloom in the European Kattegat with broad marine ecosystem impacts, and the 2015/16 Pseudochattonella verruculosa bloom in Chile (USD 800M loss to salmon aquaculture).

Highly potent fish-killers include the globally distributed, taxonomically unrelated dinoflagellate genera Alexandrium, Karenia, Karlodinium and Margalefidinium, raphidophytes Chattonella and Heterosigma, dictyochophytes Pseudochattonella and Vicicitus, and haptophytes Chrysochromulina and Prymnesium. All these species have in common their propensity to produce lytic compounds that irreparably damage the sensitive gill tissues of fish which ultimately die from suffocation. Except for recent advances with Karlodinium (karlotoxins), Prymnesium (prymnesins), and Karenia brevisulcata (brevisulcenals), the precise mechanisms of how such microalgae kill finfish remain poorly under- stood. Reactive Oxygen Species can be a co-factor in ichthyotoxicity, notably with raphidophytes such as Chattonella. While some species are always ichthyotoxic, others such as Heterosigma, Pseudochat- tonella and Alexandrium catenella kill fish only under certain conditions or life stages. Broad scale ecosystem impacts from fish killing algae are less common with raphidophytes and dictyochophytes that require intimate cellular contact for harmful effects, compared to Karenia and Prymnesium where intracellular or excreted toxins are responsible.

Critical hurdles that limit progress in our understanding of ichthyotoxins and their control and miti- gation include: HABs at fish farms are not usually a research priority until a major bloom occurs; data sharing between industry and scientists is very limited; and there is a lack of standardized methods to detect ichthyotoxins in low concentrations dissolved in seawater. Currently, the RT fish-gill W1 (rain- bow trout epithelial gill cell line) and Chaetoceros Quantum Yield bioassays are the most promising candidates for international standardization and intercalibration for some HABs.

The abundance of HABs that will adversely impact or kill fish is of considerable interest to fish farm- ers, open-water fishers, and natural resource management authorities. However, this varies with
HAB strains and species, type and age of fish, but also local conditions of water temperature, salinity, turbulence and tidal flushing. Climate change also contributes to the unpredictability of fast fish killing blooms. Prevention, prediction and monitoring are no longer sufficient, but we actively need to pursue broad-scale tools to stop the blooms, for example by means of clay flocculation of algal biomass and/ or targeted mopping up of ichthyotoxins.

We review existing knowledge and provide a roadmap for scientists, aquaculturists and insurance companies to improve management of fish-killing algal blooms that put pressure on seafood security for an ever-increasing human population.