If there ever were two truths universally acknowledged, they would be that

a) mermaids are half fish

and

b) you get omega-3 from fish.

So could we turn mermaids into omega-3 supplements for humans?1 At face value, this seems like a no-brainer. Mermaids should make, at the very least, a half-decent source of omega-3. So why have I written a 3000-word essay about this?

Let’s talk about the bottom half

According to Wikipedia, “in folklore, a mermaid is an aquatic creature with the head and upper body of a female human and the tail of a fish.” Since it’s the tail that is supposedly made of fish and would therefore contain the omega-3, that’s the part we need to dissect.

Many fish have tails, but so do many aquatic mammals such as dolphins, whales, and seals, and the tails are not the same. Fish tails move sideways while mammal tails move up and down, similarly to what human legs do when swimming in butterfly stroke. Fish tend to be vertically flat while mammalian tails are flat in the same way as two human legs held closely together - which is essentially the origin of aquatic mammals, having evolved from land animals with legs.

What I was curious about is, is one of the designs better? More efficient? More optimal? LLMs converge on the same answer: Neither is better, but they win at different things. Somewhat obviously, sideways tail movement is good for moving side-to-side, and vertical tail movement is good for going up and down. Mammals need to visit the surface to breathe, so “going up” is a pretty important use case. But sideways maneuverability and staying in the same horizontal plane are generally much more useful for a fish. When it comes to raw top speeds, tuna and dolphins are equally optimised, and long distance endurance isn’t determined by tail design either.

Now, consider the mermaid. In nearly every visual depiction (except for some ancient ones that have two tails, or human legs AND a tail, which I will dismiss as nonsense), the tail of the mermaid is shown to bend similarly to how bound-together human legs would bend. Under and behind, not to the side. In some cases, she appears to have a butt or even knees underneath the scales. All clear signs of a mammal-like skeletal structure, and of a spine intended for dorsoventral rather than lateral flexion.

Further reasoning follows oxygen intake. While I can’t say for certain whether an aquatic half-human strictly speaking needs air to breathe, two pieces of evidence point in that direction:

- The human half is the top one, and that’s where I’d expect to find lungs. Also, you need lungs for singing and screeching, both of which mermaids are known for.

- I’ve never encountered a story or a legend involving mermaids where they stayed under the sea the whole time. To the contrary, they seem to mostly hang out, or aspire to hang out, out of the water.

And if mermaids do indeed breathe air, then they need to be able to go up on the surface, and tail optimised for such a movement is clearly the way to go. To me, that’s enough to conclude that the mermaid musculoskeletal design has to be a mammalian one.

Now, you could argue that mermaid tails are nearly always covered in scales, which fish have but mammals don’t, or that mermaid tails often have fish-like fins. And I would counter-argue that scales and fins could simply be a cosmetic or minimally-functional adaptation, while attaching a fish tail to a mammalian spine at the waist would require a fundamental change in architecture. And besides, there are examples of much more dramatic cross-class trait analogies. We associate beaks and eggs with birds, but the platypus has both. Wings are usually found in birds and insects, but bats have them too. And when it comes to swimming, it’s not only the aquatic mammals that do well — penguins swim faster than many fish. So I won’t accept a few scales as proof that mermaids aren’t mammals.

Who even is a mammal

Us humans seem to have a strong need to sort all of biology into neat boxes, while all of biology stubbornly resists being sorted. When it comes to classifying fauna, one way to do it is to consider ancestry, along the lines of “everything in the evolutionary tree after this arbitrary node is a mammal.” From that perspective, what would a mammal-fish hybrid even mean? A true cross between such different animals would simply not happen after hundreds of millions of years of evolution. The only evolutionary options for something that looks like a human-fish hybrid would either be a fish that evolved to be partially like a human, or vice versa. I don’t claim to be a biologist, an evolutionary scientist, or indeed any kind of an expert on this topic, but I’d be shocked if human cognition appeared in any of the fish classes before aquatic mobility occurred in the Homo tree. So a mammal adopting some fishlike traits is the vastly more likely explanation.

There is a different definition for “mammal”, which is descriptive rather than evolutionary. There are many characteristics that are typical to mammals, though none of the items on the list are definitive proof either way — each trait either appears also in non-mammals, or does not appear in all mammals. This approach is not definitive but it’s the best we’ve got, so here is a (non-exhaustive) list of traits commonly associated with mammals, and my assessment of whether they would apply to mermaids:

- Mammals have hair. -> Check. Never seen a bald mermaid.

- Mammals have mammary glands for breastfeeding. -> Uhh, check. Never assume, but check. In fact, you can check almost any medieval mermaid picture.

- Three middle ear bones, a dentary-squamosal jaw joint, heterodont teeth. -> I wouldn’t be able to tell you exactly what these are, but it sounds like they exist above the waist, so… check.

- Neocortex. -> Ditto above the waist. Check.

- Endothermy, or warm-bloodedness. -> I don’t know. Seems plausible either way.

- Giving birth to live young. -> This is interesting. Let’s talk about this, right after the nipples.

Where little mermaids come from

Giving birth to live offspring is way less of a mammalian trait than what I was taught at school. Of course there is the platypus, a magnificent creature that refuses to conform to any societal expectations by laying eggs and lactating while looking like a duck and being hairy and having echolocation and also somehow being venomous, but there are also plenty of non-mammals, such as some snakes, sharks, and toads, that have developed live birth. Marine mammals like dolphins, seals, and whales give birth to live young, and if you’re curious, you can do what I did and google “dolphin genitalia” to discover exactly how they keep their reproductive parts hidden when not in use. These creatures are modest, and extremely hydrodynamic when not aroused.

In other words, there are unfortunately no definitive visual clues that would tell us what kind of a reproductive system mermaids might have. Mermaids could lay eggs if it was convenient to them, and that still wouldn’t make them fish. Or they could give birth to fully formed merbabies, and they still might not be mammals. Instead of live birth, a better characteristic to put on the mammalian checklist would be whether the eggs are fertilised internally or externally. If mermaids wanted to establish themselves as mammals, the one thing they should avoid is fertilising their eggs outside of the body — that would be suspiciously fishlike. Unless they were doing IVF, which you probably can’t invent without a mammalian neocortex.

In short, we don’t know enough about mermaid reproduction to definitively judge them one way or another. But since the tail orientation, the majority of the classic mammalian traits, and the sheer evolutionary likelihood point in one direction, and there is no clear evidence to the contrary, I assume their clearly mammalian lower halves probably contain mammalian reproductive systems.

Conclusion: Mermaids are mammals, not half fish.

But… but… magic!

Okay, so let’s say you’re not convinced, simply for the reason that in a world where mermaids exist, there is already some kind of magic happening, and if there’s magic, it can produce anything it wants, even mammal-fish hybrids. I’ll humour you.

Even if we imagine we are in a world with magic that can do anything, we can still observe the effects of said magic. We might not understand how the magic works, but we can see what it produces. So… show me the evidence that mermaids are half fish! Show me the story that describes mermaids performing a sideways movement pattern consistent with a mammalian spine fused with a fish spine, where both spines work at the same time moving in different directions. Or show me any reputable depiction of a mermaid releasing a batch of eggs hoping a merman will come around to fertilise them.

I would guess your main evidence would be that “people in that world say mermaids are half fish.”I’m also not going to accept that as proof, because people are notoriously bad at classifying unusual creatures. In our own world, some people used to think bestiality could result in human-animal hybrids2, ancient Indians allegedly believed that dogs and tigers could produce hybrids,3 and even today many people think corals are plants instead of animals. It’s clear that people can’t be trusted to correctly identify the biology of an unusual creature, but rather they will gladly call something a hybrid if they can’t figure out which category it belongs in.

In fact, dolphins were considered to belong in the same category as fish until about 2000 years ago, when Aristotle classified them differently. In history, many languages used the same word for dolphins and fish, so clearly it didn’t matter which direction their tail moves or how they reproduce. If anything had a tail like that, it would have been called a fish tail. Furthermore, the earliest known depictions of merpeople as half-human half-fish are another 1500-2000 years older than Aristotle! Of course, if you only have one word for all sea creatures with tails, that’s what you think mermaids are made of. Heck, the mermaids might have been confused about this themselves if they spoke ancient Mesopotamian.4 But that’s a limitation of language and understanding, not of biology, whether magical or not.

You could also argue that, like in the picture shown above, some of the older sources did actually depict mermaid tails in ways that look much more like fish anatomy than marine mammal anatomy. You might say that’s not just people being ignorant of the difference between fish and dolphins, and that these pictures should be considered evidence that mermaids are exactly what people say they are: half fish. Or you might simply conclude that if mermaids were real, and everyone thought they were half fish, then that’s what they would be, and even if it were implausible, magic would make it possible.

And… honestly, I can’t prove you wrong. But if I accepted your line of argumentation, then this whole article would be pointless, because the answer to everything would be “we don’t know, because it’s magic, so it could be anything.” Since I’ve decided to take this seriously, I have to work with probabilities and plausibilities. And on that basis, the mermaid is a mammal.

So how much omega-3 is there in mammals?

I thought I was done. I thought it would be enough to establish that mermaids are not half fish, and we’d know to look for omega-3 elsewhere.

First, let’s make the question a bit more precise. Whether a creature is a good source of omega-3 supplements for humans is affected by at least the following:

- How much of the creature’s fat is omega-3: This determines whether it should be considered as a potential source at all.

- How much fat is in the creature: This influences commercial viability.

- What types of omega-3 the creature contains: Long-chain omega-3s (EPA, DHA, DPA) are probably what you want to supplement with.5

- How high in the food chain the creature is: Oil made from predators contains more accumulated pollutants.

To form an educated opinion on mermaids, I had to answer the extremely normal question of “Can you make omega-3 out of humans?” and followed roughly the following path of discovery:

- Omega-3 doesn’t just magically appear in fish because of their inherent fishiness. Fish have a lot of it because they eat algae, or they eat other fish that eat algae.

- However, being a fish does not qualify a creature to be a great omega-3 source, since the share of both fat and of omega-3 varies greatly between different species and diets.

- Marine mammals like seals and dolphins are extremely rich in omega-3, given that they are basically on the same diet as fish. There are seal oil products sold as omega-3 supplements, although it’s uncommon.

- However, mammals have a lower share of omega-3 of their total fat, even when on an all-fish diet, than the most omega-3-rich fish. While marine mammals have more fat than fish do, and therefore contain more omega-3 pound-for-pound, seal oil has a much lower share of omega-3 than standard fish oil.

- Humans, with our regular diet, would be a pitiful source of omega-3. We simply have too much of other fats competing for the same internal resources for processing and storing.

- If a human ate a diet of mainly seaweed and fish, their body would store much more omega-3. Let’s call that person a “marine-diet human.” Claude thinks such a diet would likely be sustainable, assuming you eat all parts of the fish, including liver, roe, skin, and small bones, and assuming you eat at least some fatty fish.

- It’s not clear if humans and marine mammals have a similar fat metabolism and whether they would store fats in the same way if on a similar diet. But it seems there are enough similarities to assume they could at least be in the same ballpark.

- To be truly comparable to a seal, and to be an abundant source of omega-3, the marine-diet human would additionally have to have a significant body fat percentage. Seals can have as high as 50% body fat.

- Human fat tissue needs a long time for a full turnover, meaning that someone aspiring to be an omega-3-rich human would have to be a marine-diet human for a year or more before their adipose tissue had the final omega-3-maximised composition.

The following table6 illustrates the differences between different species and diets in omega-3 concentration:

Since I landed on the conclusion that a long-time marine-diet human would have roughly the same share of omega-3 as a seal, that’s what I will also assume a mammalian mermaid would have. That brings us to the final list of open questions that will determine the answer to the original question of whether mermaids would be a good source of omega-3:

- Are they mammals?

- If they are not mammals, which fish are they made of?

- What is their body fat percentage?

- Do they eat an all-marine diet rich in omega-3?

- How high are they in the food chain?

For your convenience, I have organised the answers into a literal mermaid diagram:

As you can see, if mermaids were half-anchovy or similar, they would be a fantastically abundant source of omega-3. Not only would mermaid oil (made from the tail only) have a high level of omega-3, you would get so much of it per mermaid. Even with conservative estimations, half of a mermaid would contain over 2000 times as much fish as a single Peruvian anchovy.7 Talk about economies of scale.

But here comes the finding I was absolutely not expecting when starting to write this piece: In the specific edge case that a mermaid actually is half fish, but the fish tail is very low in fat, it would actually be the human half that contains more omega-3. And the exact opposite could be true if mermaids are mammals; the tail could contain a thicker layer of fat than the human half. But I admit this is pure speculation.

I could spend more time resolving the open questions, as well as dig into the commercial viability of a mermaid oil business, but my patience with this topic is running out. Here’s my final opinion: Mermaids are mammals, top and bottom, and they probably are a decent source of omega-3 anyway. And if I’m wrong and they are half fish, then all bets are off and you should test their fat composition before committing to a large-scale commercial operation.

Omega-3: Essential fat your body can’t make on its own but needs for various purposes, which is why many people take it as a supplement. This article doesn’t explain why it’s useful, so you’re either going to have to take my word for it, or do some reading first.

It’s a bit unclear to me how ancient Mesopotamian worked; the same word could have applied to fish as well as all marine creatures as a category.

There’s much more complexity to the different subtypes of omega-3, which I will not go into here, but I will encourage you to read about if you are curious. For example, salmon has less overall long-chain omega-3s, but a higher share of DHA, which is needed for brain and eye function in humans.

Peruvian anchovies are the most commonly used fish in fish oil production. Fish weight is estimated based on the length, using the formula W = a · Lᵇ. Values specific to the Peruvian anchovy were found here under “Bayesian length-weight” and result in about 16g of weight per 14-cm anchovy. For mermaids, I assumed a weight of 70 kg and that 50% of it (35 kg) sits in the fish half. 35kg divided by 16g results in the mermaid tail weighing 2178.5 times as much as a single anchovy.