International Journal of Environment and Climate Change

Harmful algal blooms represent one of the few climate-sensitive hazards capable of destroying an entire marine farm's standing biomass within days, yet the scientific literature that underpins farm protection remains fragmented across oceanography, fish pathology, sensor engineering and resource economics. This review critically examines the evidence connecting climate variability and change to bloom hazard in aquaculture regions, the quantification of farmed fish mortality, the cellular and physiological mechanisms through which exposure becomes death, and the effectiveness of early warning systems and farm-level adaptation. Peer-reviewed literature was identified through structured searching of open scholarly databases and indexes, supplemented by backward and forward citation searching, and appraised for design adequacy, measurement validity, ecological realism and consistency with independent evidence. Four conclusions emerge with differing degrees of confidence. Attribution of increased bloom frequency to climate forcing remains contested at the global scale because observational effort has expanded alongside aquaculture itself, although regional attribution is considerably stronger where hydroclimatic anomalies have been reconstructed independently of bloom records. Mortality quantification rests largely on cell-abundance thresholds whose transferability between species, sites and years is weak, and which conflate very different killing mechanisms. Mechanistic work has converged on reactive oxygen species, lipophilic and polyketide toxins, and mechanical gill injury as complementary rather than competing pathways, but laboratory potency rankings predict field outcomes poorly because exposure dose, water-column structure and prior gill condition are rarely measured concurrently. Early warning systems have matured technologically while remaining substantially under-evaluated against operational decision criteria, and most farm-level mitigation measures are supported by mechanistic plausibility and small-scale trials rather than by controlled comparison at commercial scale. Priorities include standardised exposure reporting, prospective forecast verification against farm decisions, and evaluation of mitigation under realistic bloom conditions.