Bubble curtains underperform commercial expectations against ichthyotoxic HABs under field conditions: hydrodynamic and cell motility-dependent constraints in Patagonian salmon aquaculture

Resumen: Bubble curtains are frequently used in Chilean salmon farms as a physical barrier against harmful algal blooms (HABs), yet their field efficacy under natural ichthyotoxic bloom conditions remains unclear and has not been evaluated at operational scales. This study presents the first integrated field assessment of bubble curtain performance during HAB events in the strongly stratified fjord systems of southern Chile, combining in situ cell abundance measurements, hydrographic profiling, laboratory characterisation of bubble size distributions, bubble plume hydrodynamic modelling, and a nationwide survey of 121 salmon farms. Short-term deployments during blooms of Heterosigma akashiwo and Prorocentrum micans yielded exclusion efficiencies of 22% and 0%, respectively, while long-term continuous operation during Pseudochattonella verruculosa and H. akashiwo blooms achieved only 40–53% containment, well below commercially reported values (>90%). Cross-correlation analysis between internal and external cell abundance in long-term time series confirmed that bloom dynamics on both sides of the curtain were fully synchronised, indicating that the barrier did not decouple the pen from external bloom progression. Hydrodynamic modelling showed that lateral currents of 0.1 and 0.2 m s⁻¹ deflected rising bubble plumes by 3.84 and 7.70 m respectively, compromising barrier continuity across the velocity range typical of Chilean farming sites, while field observations indicated that motile flagellates accumulated within the barrier through gyrotactic retention rather than being passively displaced. In the tested diffuser, increasing gas flow rate from 3 to 10 L min⁻¹ shifted mean bubble diameter from 754 to 916 µm and nearly doubled size variability, confirming that higher compressor output compromises plume coherence rather than improving exclusion. The farm survey showed that 93% of installations do not deploy bubble curtains against HABs, 43% have never evaluated their performance, and the current velocities optimal for salmon husbandry (0.2–0.25 m s⁻¹) coincide precisely with those at which curtain efficacy collapses. Meaningful improvement will require species-aware deployment protocols, diffuser redesign, and depth-resolved monitoring beyond surface cell counts.