Painful fish deaths? You might be guilty!


Painful fish deaths? You might be guilty!

Humanely destroying a fish is an inevitable part of keeping them in any numbers. Shocked by some advice being touted online, Nathan Hill takes a closer look at how to do it - and how not to.

Somehow I chanced upon a forum discussing methods of killing a sick fish. That’s not to say that I’d stumbled across some den of aquatic serial killers, talking out their twisted fantasies. But given some of the methods suggested, I might be forgiven for thinking that I had.

It’s inevitable that for many of us, we may be forced to play our hand, and take a fish’s life in order to alleviate suffering. Accidents and aggression happen, diseases occur, and we might be called upon to provide the final service for the pets in our care.

At the least, we can ensure that after years of enjoyment, we don’t turn our fish’s last moments into some horrific splatterhouse movie, with fumbling hands and inappropriate techniques.

We could argue that it’s not even down to us to decide the point of release for our fish. The issue of euthanasia is contentious enough in human circles already, people divisive in their views. It’s no surprise then, that there are those who would let nature take its course, and not intervene in the death of a fish.

There are others who would give benefit of the doubt until the very end. I’ll admit, if I’d spent £250 on a fish that somehow ended up with ravaged skin, absent fins and was gushing blood from its gills, the mercenary element of me inside might be asking ‘do you really want to do this, you know… just in case?’

I am keen to alleviate suffering in any fish under my control, and given the duration of my experience in aquatics I have a reasonable intuition of when a fish has reached a point of no return.

For the individual, it is their call whether or not they think a fish may be saved. I do not envy the inexperienced keeper here, but with reaction time on the Internet as it stands, and with so many experts now readily available, advice can be sought at the first hint of trouble.

Herein lies a problem. It is often difficult for the newcomer to separate the terrible advice from the good. Whatever the hindrances, always seek a second (or third, or fourth) opinion before culling any livestock, as it is so very easy to misdiagnose.

I honestly wish I could offer a clear-cut method of assessment that could be used with any fish, to see if humane killing is justified or needed, but I cannot. The subject, including the individual behaviour of a fish, is too complex to sum up in a simple list of do and don’t rules. But if anyone out there has a fool-proof, tested approach then I want to hear it.

Once the decision has been made to kill a fish, the suitable methods that should be employed are much, much less open to debate, and there are some approaches that any empathic, compassionate individual would want to avoid. Yet still, so many years after they have been debunked, these inappropriate techniques insist on rearing their torturous heads. It’s akin to the old ‘fish are fine in bowls’ or ‘fish growing to the size of the tank’ debates. We know they’re wrong, but by Jove they are persistent.

Thankfully, I’ve been in a position to speak to some authorities on the subject, and those on the cutting edge of humane destruction methods, so I’ve accessed some of the most up to date thinking, and can hopefully offer a little clarity on what goes – and what certainly doesn’t.


Hugely contentious, chilling may have a role to play in the humane killing of some fish. It does need some clarification, though, because I want it on the record that you’re not okay to take an awake, alert fish and throw it in the freezer.

Ice has been used to chill water to 2 – 4°C, and small, tropical species up to 5cm long added to this if they need to be killed. The idea here is that cool water has a soporific quality to it that will send small fish into a stupor, and eventually brings about a lethal shock, most likely through osmoregulatory functions being disrupted.

Dr. Lynne Sneddon of the Institute of Integrative Biology directs my attention to the work of Monte Matthews and Zoltan Varga, who have published papers on the subject of chill euthanasia. She notes, “ice must not be allowed to touch the body to prevent ice crystals forming.” It is considered that if the body develops crystals inside before becoming unconscious, then this could be pain inducing.

Correct temperature is essential. Some usage of too-cold ice slurry has produced extreme physical and behavioural reactions, and indicated a long and protracted death.

However, done correctly and with small fish, Dr. Sneddon informs me that, “some authors report that fish gill movements cease after 10 seconds so it’s quick and it’s less aversive than anaesthesia.”

One method of dispatch is described whereby the fish is cooled until losing consciousness (which can be approximately ten to twenty minutes of ice water exposure) followed by maceration. That’s not something that the home aquarist can do easily, without specialist equipment to literally destroy the body instantly.

It should also be noted that some authors report that cooling of a fish reduces nerve conduction, but may not eliminate it. Ergo, a cooled fish may lose locomotion, but it’s not necessarily assured that the fish cannot feel anything as it drifts off.

Dr. Peter Burgess has some reservations regarding the cooling approach. He draws my attention to Ross and Ross’s book, Anaesthetic and sedative techniques for aquatic animals, where he points out that "slow cooling is thought to give very incomplete analgesia," and suspects that this method has the potential to cause pain.

Carbon dioxide

This one has come up a few times, most frequently from US websites and forums where I can only presume it is still widely used. Methods cited include the addition of a fish to a closed chamber of water, and methodically bubbling CO2 gas through to drive out oxygen, using carbonated water, or even dropping Alka Seltzer into a container of water with fish.

Alka Seltzer was often cited as a fish anaesthetic in scientific research from the early 1980s, which compared its efficiency against the anaesthetic MS 222, but things have moved a long way in 30 years.

Dr. Sneddon has published papers on this very subject and informs me, “CO2 has been banned in the EU for slaughter of fish farms because CO2 causes asphyxiation and a peripheral acidosis which stimulates nociceptors giving rise to pain.

“So Alka Seltzer is bad, but CO2 of fish in farms is allowed if it is an emergency. The Norwegians are still using it.”

Though the use of CO2 is still considered by some to be a more efficient and faster method of killing than, say, injection, it does create uncomfortable conditions for fish, resulting in increased respiration, and acidosis caused by the formation of carbonic acid.

Given the other methods of euthanasia available to us, there’s no reason for anyone to need to consider CO2 as a hobbyist. Besides which, lacking experience in the use of CO2 as a humane killer, the amateur fishkeeper could potentially get some part of the procedure very wrong, drawing out a slow and painful death of acidic burning.

Boiling/immersion in boiling water

I’m surprised to see this one doing the rounds. For my part, I can find no scientific paper that has ever been carried out regarding boiling live fish as a euthanasia method, but the general consensus is that the method is not instantaneous, and likely to cause the most intense nociceptive reaction of any method of dispatch.

Until a paper appears verifying its feasible use as a method, don’t consider this as a potential way of painlessly ending a fish’s life.

Anaesthetic overdose

What should be the most humane method of dispatch is not without some contention.

As aquarists, we have two options for anaesthetic overdose: Eugenol, as found in clove oil, or Aqua-Sed. Some commercial products such as Koi Calm are clove oil/eugenol based, and readily available. If unable to find an aquatic retailer stocking it, a chemist will usually carry clove oil for toothache.

Aqua-Sed is 2-phenoxyethanol, and available to the general public. At low doses it is used to sedate a fish, and at higher doses can be used to kill a fish quickly and efficiently. A dosage of AquaSed at four pumps per litre is enough to ensure lethality, and the fish should then be left in the solution long enough to ensure that it is dead. The actual timescale will vary from fish to fish, the trend being that larger fish may take longer to perish from an overdose. Alternatively, a deeply anaesthetised fish can then be finished with decapitation and pithing.

Anaesthetics are not without their risks, or potential to cause distress. Many fish show an adverse reaction to being placed in a body of water containing the prescription only sedative, MS 222. The problem with this chemical, and some others like it, is that it is inherently acidic, and a sample of water containing a lethal dose may easily have a pH of 5.0 or lower. Taking a fish and putting it straight into this could easily cause acidosis, which is, of course, painful.

If using an anaesthetic, it should be accompanied by buffering of the water it is going in to, if needed.

Clove oil, containing the ingredient eugenol, is good but has minor issues, in that it isn’t always readily mixed with water, preferring to remain as clusters of oily droplets. In an ideal world, it could be made into a stock solution ready for use, mixed with ethanol, which would allow easy application. But we then face the issue of adverse reactions to ethanol.

Current clove oil methods involve adding a dose of 400mg/l to water, and premixing it so that all of the oil has emulsified. The fish is then transferred to this water, after which it will lose consciousness within seconds, and breathing will quickly cease. Eventually the fish will succumb to hypoxia and die, though like other methods this can vary in timescale from fish to fish. It is recommended that the fish be left in the solution for at least two hours.

William Wildgoose’s BSAVA Manual of Ornamental Fish Health mentions clove oil as a technique for killing fish at 10 drops per litre as a minimal dose, noting that clove oil can be properly dissolved by mixing with a little hot water first.

Dr. Peter Burgess suggests the fish should be left in anaesthetic solution “at least one hour, and if in doubt, overnight. If the fish shows activity after 20-30 minutes or so, then add extra anaesthetic.” He adds that the person performing the task should also monitor closely for signs of movement, and in the case of very small fish, a magnifying glass may be required to be certain.

Anaesthetic is a method preferred by many public aquaria, under the guidance of a vet, and Dave Wolfenden of The Blue Planet informs me that “an overdose of MS 222 needs to be carried out carefully so that welfare isn’t compromised – so careful buffering needs to be performed to prevent drastic acidosis.”

For really large fish, he tells me, “ typically use Pentobarbitone or another drug injected intravenously. Euthanising a three metre long shark is no easy feat.”

Interestingly, in the world of research, a fish being dispatched needs to be carried out away from sight of its former tankmates.

Other considerations need to be made, such as the stress involved in catching the fish in the first place to transfer it. Although an overdose when properly used is currently considered the most humane method, it may be the case that extreme stress involved in capture may push you to move to a faster, instantaneous method such as decapitation and pithing.

Blunt trauma/concussion/decapitation

Unless you’re squeamish, this may be one of the best methods to quickly dispatch a dying fish.

For larger specimens, a blow needs to be made rendering the fish unconscious at the least, and causing outright brain destruction at best. It is essential that the fish is correctly dispatched before it can regain consciousness, and the first strike will need to be decisive. Too weak a blow, and you will only cause pain and more suffering.

If you are in any way unsure of how to dispatch a larger fish in one blow, then this method is not for you. An active, wriggling fish can be a hard target, and a last minute twitch can be the difference between success and failure.

Concussion and head trauma can be used after the successful administration of anaesthetic, which can make the task considerably easier, and give some leeway in the event of a failed first hit.

Smaller fish (the majority of or aquarium residents) can be decapitated, as long as the brain is subsequently destroyed. This will involve pithing, which requires that a sharp object is placed into the brain cavity ensuring its destruction. If you’re not familiar with techniques to pith a fish, seek guidance beforehand.

Do not decapitate without pithing. Although death will follow, the detached head of a fish may still be both alive and aware for several minutes after.


This is likely one of the most common methods employed by casual keepers of things like simple goldfish setups, and little interest in the intellectual side of aquatics. Novelty aquarists have flushed dying fish for as long as we’ve had toilets, and most likely will for many more decades to come.

Flushing a fish is subjecting it to drastic changes in water quality and temperature, and why people have this idea that the moment it passes the U bend it suddenly dies shall remain for me one of life’s eternal mysteries.

Don’t assume the fish even dies in the sewage workings. It’s not unknown for coldwater species to successfully make their way through the system, and then get into native waterways. But even if they don’t, drowning in other people’s slurry is an unacceptable method of death.

Oh, and as an aside, if you flush a fish, then you may well have just broken the law.


I’m not sure I even need to explain this one. Anyone who has found a fish, dried on the floor, clearly having been there for some time, but still moving its gills will understand why this approach isn’t going to be in any way acceptable to end the suffering of a fish. Just no.

Suffocation of a fish is a slow and painful process, involving gill damage, muscular pain, escape responses and eventual organ failure. None of us would appreciate going that way.

Breaking the spine

Fish are a lot more robust to some things than you might believe. We’ve all seen films with an assassin breaking a person’s neck, and then he falls to the floor dead (the victim, not the assassin). Well, yes and no. Spinal damage paralyses the body, but it doesn’t guarantee immediate death.

The same applies in fish. Your poorly Oscar may stop flinching once you bend that head back until the vertebrae disconnect, but just because it’s not moving anymore, doesn’t mean it’s not still experiencing a (somewhat abysmal) life.

Spinal breaks won’t work without pithing of the brain. No pith, no break. Besides, given our alternative, more humane options, spine breaking is all a bit excessive to a peaceful end. It simply increases the potential for pain, without making it easy to follow up with a fast method of dispatch.

Other methods

As well as the above, there are a couple of methods available to some other animals that have been explored in fish, but may be either impractical, or inapplicable to aquatic life forms.

One technique involves microwave radiation focused directly on key parts of the brain. However, this would require specially designed equipment, and its effectiveness in water is questionable and has not been adequately explored.

I want to make it absolutely clear before this ever becomes a fishkeeping myth that microwaving a fish is most certainly not a humane method of destroying it. I only mention it because I know how frequently information can be seized upon from scientific abstracts, and peddled out of all context as factual. I’m attacking this myth before it even gains hold.

Electrocution is a method sometimes used in the aquaculture world, though it is not only wildly dangerous to try to do in the confines of a tank, but is also doubtful as a flawless method of euthanasia. Some fish that have been electrocuted have twisted and contorted their bodies, breaking bones and snapping their own vertebrae, only to then regain consciousness (in some cases not even losing it) some time after the event. It is very hard to get right, and requires specialist equipment.

So what should we do?

From the authorities I have spoken with, the almost unanimously favoured approach to humane killing is anaesthetic overdose, with concussion and brain destruction in (very close) second place.

Fish nutritionist and researcher Dr. Donna Snellgrove sums up what I consider the best option when she tells me: “The easiest route for fish keepers to humanely kill a non-recovering sick fish would be by an overdose of anaesthetic and then destruction of the brain, using some sharp scissors/scalpel/skewer/large needle, appropriate for the size of fish.

“This would be most straight forward, least squeamish and most humane method.”

This double-action approach to destruction is widely advocated across the board, and as Dr. Snellgrove quite rightly clarifies, “In essence the fish is killed twice - once to kill the fish and once to confirm death.”

If in doubt...

I have images now of terrified aquarists picturing themselves in horrendous situations involving a dying fish.

Thankfully the reality of this hobby is that we’re in it to keep fish alive, not to routinely kill them, and so for the typical hobbyist, the requirement to intervene in a fish’s ultimate fate is slim, and most of us will never have to do this.

Hopefully, most aquarists will have a final coup-de-grace method on standby. If you don’t, there’s no better reason to splash out on a small bottle of clove oil than panic avoidance if the time ever comes. If you find yourself with a tormented, fishy soul calling upon you to do that final, kind act, you’ll be able to show you really do care about your fish – all the way to the bitter end.

Why not take out a subscription to Practical Fishkeeping magazine? See our latest subscription offer.

Don't forget that PFK is now available to download on the iPad.