Killing Farmed Fishes Humanely Remains An Unsolved Problem

Aquaculture has now surpassed wild capture fisheries as the world’s largest producer of aquatic animals by biomass, yet welfare protections for farmed fishes have lagged far behind those for land-based farmed animals. Standards in many countries require that terrestrial farmed animals be rendered unconscious before slaughter and remain that way until death, but no such universal requirement exists for fishes.

The vast majority of farmed fishes globally are killed without any stunning at all, using methods such as ice slurry and asphyxiation that cause prolonged and painful deaths. Even where stunning is used, its reliability varies widely, and there are no validated species-specific tools for confirming unconsciousness.

This review aimed to critically evaluate the main stunning methods used in commercial aquaculture, with a focus on Europe, against a set of key welfare concerns, identifying each method’s strengths and weaknesses and highlighting where further research and development are needed.

The researchers used a two-part approach. First, they conducted a literature search covering 1990 to 2025 to identify the key welfare concerns at each stage of the slaughter process: pre-stunning, stunning induction, and loss of consciousness. A total of 71 papers and reports informed this stage. Second, they searched for studies published between 2015 and 2025 to evaluate five commercially used stunning methods against the key welfare concerns. The methods reviewed included:

• In-air electrical stunning (35 articles): Fishes are removed from the water and passed through a stunner that delivers a high-voltage electrical current through their brain
• In-water electrical stunning (22 articles): Fishes pass through a pipe filled with water through which an electrical current runs, intended to stun them
• Carbon dioxide stunning (17 articles): Fishes are placed in a tank of water saturated with carbon dioxide gas until they lose consciousness
• Percussive stunning (23 articles): Fishes are removed from the water and struck on the head with a club, hammer, or captive bolt device, intended to cause immediate brain failure
• Gas mixture stunning (10 articles): Fishes are placed in a tank of water saturated with some combination of gases, like nitrogen and argon, until they lose consciousness

For each method, the researchers assessed how likely each welfare concern was to occur in practice, how severe the impact would be if it did, and how strong the available evidence was.

Pre-Stunning Welfare Concerns

Before any stunning occurs, farmed fishes are typically subjected to crowding, handling, and removal from the water. These stressors compound one another, causing physical injuries, elevated stress hormones, and fear responses. Air exposure is especially harmful: one analysis found that fishes likely experience disabling or excruciating levels of negative affective states even during brief air exposure. These pre-slaughter stressors apply across all five methods reviewed and represent a baseline welfare problem that no current approach has fully resolved. They can also affect how reliably unconsciousness is induced and maintained once the stunning method is applied.

Welfare Concerns At The Time Of Stunning

During the induction phase, when fishes are first exposed to the stunning method before losing consciousness, methods diverge sharply in their welfare impact. Gas methods show some of the most serious concerns. Fishes exposed to carbon dioxide-saturated water show violent escape attempts, and carbon dioxide is known to directly excite pain receptors. Critically, gas methods don’t induce immediate unconsciousness; it can take several minutes for fishes to lose consciousness, during which they show clear behavioral signs of distress. Alternative gas mixtures are even less studied, but early evidence points to similarly aversive reactions.

Electrical and percussive methods have the potential to induce unconsciousness rapidly and without the acute distress seen with gas methods, but only when applied precisely and with appropriate parameters. For electrical methods, effectiveness is highly sensitive to fish size, species, orientation, skin condition, and equipment settings. For percussive methods, delivered by hand or a mechanical device, missed blows are common (and painful), and the problem worsens as fishes grow larger or vary in size within a batch.

The researchers found that all five methods carry a medium-to-high risk of mis-stuns, where fishes aren’t immediately rendered unconscious. The welfare impact of a mis-stun is high: a fish who hasn’t been successfully stunned may be subjected to the killing procedure while still fully conscious and capable of experiencing pain.

Welfare Concerns Around Loss Of Consciousness

Even successful stuns often don’t last. The review found a consistent and serious risk of fishes regaining consciousness before death with electrical methods in particular. Recovery is more likely when killing doesn’t immediately follow stunning, or when the killing step itself takes time.

A cross-cutting problem compounds all of this: the behavioral indicators used in commercial settings to assess unconsciousness, such as eye-roll reflexes, righting reflex, and gill-cover movements, are often unreliable. In multiple studies, fishes showed behavioral signs of unconsciousness while brain activity measurements indicated they were still capable of experiencing pain. The more reliable measure, electroencephalography, isn’t yet practical for commercial use, meaning operators may routinely misjudge whether fishes are truly unconscious.

The review has several important limitations. Similar to a previous review on stunning in wild-caught fishes, the vast majority of included studies were conducted at research scale; only four assessed stunning under actual commercial conditions. Research has also focused heavily on a small number of commercially dominant species, leaving many farmed fish species virtually unstudied. Additionally, most studies evaluate stressors in isolation rather than accounting for their cumulative impact — fishes who are injured during crowding, for example, may have a reduced capacity to cope with subsequent procedures. Operator skill and adherence to protocols, which can substantially affect outcomes, are also rarely captured in controlled studies.

The researchers conclude that no current stunning method performs adequately across all key welfare concerns. Each approach involves genuine risks of pain, fear, and injury — before, during, or after the stun — and none has been validated as reliable under real-world commercial conditions. They call for immediate action, including mandating the use of stunning in fish farming, phasing out methods shown to cause suffering such as gas stunning, improving operator training, and establishing monitoring and auditing frameworks. They also highlight the need for longer-term investment in commercial-scale research and more reliable consciousness indicators.

For animal advocates, this review provides a powerful evidence base on multiple fronts. The welfare science woven throughout the paper makes a compelling case that farmed fishes suffer at slaughter on a massive scale, and that current industry practices routinely fail to protect them even when humane slaughter is the stated goal. The researchers’ explicit statement that voluntary industry action is insufficient to drive meaningful change, and their call for mandatory regulation, financial incentives to switch to better methods, and enforceable monitoring, gives advocates clear policy targets to press for.

At a deeper level, evidence that even the best available stunning methods are unreliable in commercial settings invites advocates to question whether humane slaughter is achievable in practice, and to argue for reducing the scale of aquaculture altogether as the only certain route to addressing this suffering.

This summary was drafted by a large language model (LLM) and closely edited by our Research Library Manager for clarity and accuracy. As per our AI policy, Faunalytics only uses LLMs to summarize very long reports (~50+ pages) that are not appropriate to assign to volunteers, studies that contain graphic descriptions of animal cruelty or animal industries, and research on niche topics. We remain committed to bringing you reliable data, which is why any AI-generated work will always be reviewed by a human.