Scientists solve millennia-old mystery about the argonaut octopus

By Ed Yong | May 18, 2010 7:00 pm

The argonauts are a group of octopuses unlike any other. The females secrete a thin, white, brittle shell called the paper nautilus. Nestled with their arms tucked inside this beautiful, translucent home, they drift through the open ocean while other octopus species crawl along the sea floor. The shell is often described as an egg-case, but octopus specialists Julian Finn and Mark Norman have discovered that it has another function – it’s an organic ballast tank.

An argonaut uses its shell to trap air from the surface and dives to a depth where the encased gas perfectly counteracts its own weight, allowing it to bob effortlessly without rising or sinking. Finn and Norman filmed and photographed live animals in the act of trapping their air bubbles, solving a mystery that has been debated for millennia.

Scientists have long wondered about the purpose of the argonaut’s paper nautilus. No less a thinker than Aristotle put forward a hypothesis. In 300 BC, he suggested that the female octopus uses its shell as a boat, floating on the ocean surface and using her tentacles as oars and sails. Despite a total lack of evidence for this ‘sailing hypothesis’, it was later championed thousands of years later by Jules Verne, who wrote about sailing argonauts in Twenty Thousand Leagues Under the Sea.

Since 1923 and the work of Adolf Naef, the shell has been viewed as a container for the argonaut’s eggs. After mating with a male (who is around 8 times smaller and 600 times lighter), the female secretes the papery shell using the tips of two large tentacles. She lays her eggs within the structure before snuggling inside herself. Besides her eggs, her only housemate is one of the male’s arms – the hectocotylus. The arm doubled as a penis, snapped off during sex and stays inside the female’s body.


Besides the female, her eggs and her disembodied sperm package, the paper nautiluses often contain pockets of air. Naef viewed these as a problem. According to him, the unintended pockets eventually trap argonauts at the sea surface and cost them their lives. That would certainly explain the mass argonaut strandings that are sometimes found, but Naef didn’t have any evidence to back up his claims. Others have speculated that the air bubbles were caused by aeration devices in aquariums and are only seen in captive argonauts. Yet others have suggested that the animals deliberately use the air pockets to maintain their buoyancy but until now, that’s been mere speculation.

Into this debate came Finn and Norman. Their names may be familiar to regular readers – they have discovered the smash-hit octopus that carries coconut shells as a suit of armour, dolphin chefs that can prepare a cuttlefish meal, and the awesome mimic octopus. As with these earlier discoveries, their work on argonauts was based on observations of wild animals. They rescued three greater argonauts (Argonauta argo) from nets in the Sea of Japan, released them into Okidomari Harbour and filmed them as they adjusted to their freedom. It’s their beautiful video that graces the top of this post.

All of the females were checked before their release to make sure that they had no air already trapped in their shells. Without this air, they were in danger of sinking and had trouble keeping their shells upright. All three animals fixed this problem in the same way.

Each one used its their funnel to jet to the ocean surface and bob the top of its shell in the overlying air. The shell has a couple of apertures at the top, which allows the argonaut to gulp in air, sealing it inside with a quick flick of two of its arms. Having sealed away this pocket, it points its funnel upwards, rolling the shell away from the water surface and forcing itself downwards. At the depth where this compressed bubble cancels out its weight, the argonaut levels off and starts swimming.

Naef was clearly wrong. The air isn’t life-threatening or even unintended – the argonaut deliberately introduces it and has total control over it. Once the animals dived again, Finn and Norman grabbed them and rotated them through 360 degrees – not a single bubble emerged. “To my delight the argonauts immediately put to rest decades of conflicting opinions, demonstrating their expert ability at obtaining and managing surface-acquired air,” says Finn.

Argonaut_diveThis neutral buoyancy is a big boon for animals that live in the open ocean, because they don’t have to expend energy on keeping their place in the water column. Other cephalopods use a combination of fins, jets of water and, in the case of the actual nautilus, chambered shells. The argonauts are the only species known to use bubbles, but it’s clearly an efficient tactic. Finn and Norman observed that once they had trapped their air pockets and reached the right depth, they could swim fast enough to outpace a human diver.

By rocking at the surface, the argonaut can also trap a sizeable volume of air, which, in turn, allows it to reach a greater depth before becoming neutrally buoyant. Finn and Norman think that this may allow these unusual octopuses to avoid the surface layers of the ocean, where they would be vulnerable to birds and other top-level hunters.

This penchant for deeper waters may also explain why this behaviour has never been seen before, even though argonauts have featured in aquariums. They simply weren’t kept in tanks that were deep enough. The animals created air pockets as they would in the wild but without the ability to dive to the right depth, the air just brought them back to the surface again.

NautilusAs a buoyancy aid, the argonaut’s paper nautilus is superficially similar to the much harder shell of its namesake, the chambered nautiluses (right). These animals also use shells with trapped air, but theirs are permanently stuck to their bodies and divided internally into many gas-filled chambers. The two groups – nautiluses and argonauts – are only distant relatives, but they have both arrived at similar ways of controlling their buoyancy.

The argonaut’s solution is undoubtedly simpler and more flexible, but the nautilus’s sturdier shell prevents increasing water pressure from compressing the trapped air too much.  As a result, the nautilus can dive far deeper than the argonaut, to a depth of 750 metres.

Finn and Norman’s study may have solved a longstanding argonaut mystery but there’s still much to learn about these enigmatic and beautiful animals. Even though people have known about them since Ancient Greece, their behaviour, distribution and biology are still shrouded in secrecy. To find out more, Finn and Norman are conducting a survey reviewing Australia’s argonauts, and they’ve set up a website with details about how you could help them in their Argosearch.


Reference: Proc Roy Soc B

Photos: Video and bottom photo by Yasushi Okumura, Japan Underwater Films; all other photos by Julian Finn

More on octopuses:


Twitter.jpg Facebook.jpg Feed.jpg Book.jpg


Comments (15)

  1. Tk

    Could someone please explain basic physics to me? :-) What determines the depth “where the encased gas perfectly counteracts [the argonaut’s] own weight”? I presume it’s where the density of the argonaut (including her air pocket) equals that of the surrounding water. But (a) how does the water density vary with depth? I thought water is incompressible, but maybe in the ocean large pressures, varying salinity and temperature make a difference? And (b) how does the compression of the argonaut’s air pocket affect the equation? Why can the nautilus dive deeper than the argonaut, when its air is compressed less? Surely less compression = more volume = lower density = floats more? I think there’s something I’m not understanding.

  2. Nathan Myers

    As the argonaut goes deeper, the air pocket compresses and gets smaller, denser, and less buoyant. Really, the argonaut gets pushed farther into its shell.

    If the nautilus shell can hold back some pressure, it means it has more latitude (as it were) to go up and down without losing its neutral buoyancy.

    The argonaut is stuck at one level. Above, it has to keep jetting not to bob to the surface; below, to keep from sinking.

  3. Chris M.

    It is absolutely an unstable equilibrium, and it would be very interesting to see whether some level of planning is going on, on the part of the argonaut, in targeting their intended depth. Good weather versus a squall, for example, might make different depths desirable. But yeah, there are some pretty serious drawbacks to this, primarily getting back down from the surface. They’re putting quite a lot of effort into “climbing” down the pressure gradient. Still, pretty clearly the added efficiency once they’re down there makes it worth it!

    (Aside: Your links to the prior octopus and the dolphin chef look to have an extra date tag in them)

  4. Alvin Alejandrino

    To address some of Tk’s questions: The air above us at sea level is 1atm and is equal to 1.0332 kg/sqcm. In the ocean or other large body of water, this pressure increases every 10m (33ft). The deeper you go, the more pressure the “atmosphere” exerts on you. This is the physics divers deal with when they SCUBA. They have a tank full of air and is balanced with the weight of the individual and accessory lead weights. Closer to the surface buoyancy is hard to maintain, but it is easier at deeper depths. Unfortunately for the paper nautilus, their shell is thin and fragile and will get crushed with increasing depths. The chambered nautilus’ hard shell allows for deeper diving. I hope this helps!

    Great study!

  5. This is really cool, but I’m most curious about the shape of the female’s shell. First, I wonder how thin it is. Second, I wonder why it’s shaped so much like a nautilus shell.Does it “grow” in the same manner? Does it grow as she grows? Can she be detached from the shell?

    I wonder if this is an evolutionary novelty or some kind of reversal?

  6. Torbjörn Larsson, OM

    Tk, you got the most part of the physics, I would say. The comments should help with the rest. What hasn’t been commented on is that real liquids aren’t ideally incompressible, so density increases somewhat with pressure:

    “For each atmosphere increase in pressure, the volume of water would decrease 46.4 parts per million.”

    Oh, and this might be confusing: “In the ocean or other large body of water, this pressure increases [~ 1 atm] every 10m (33ft).”

    This is why divers have to make sure to open their eustachian tube and regulate depth change rates, or the pressure differential would eventually be harmful, with gas embolism occurring (“the bends”) et cetera.

  7. Chris

    Arterial gas embolism isn’t the same as the bends. They are both related to dissolved gas in the blood and can affect divers but generally getting an embolism while diving is much more serious…as in dead. The bends are rarely very serious for recreational divers and very rarely lead to death.

  8. Jan

    Arterial gas embolism is related to dissolved gas in the blood in one case: if the diver has a patent foramen ovale, an unsealed hole in the heart between the venous and arterial sides of the heart.

    An infant in a womb does not breathe – his blood is oxygenated through the umbilical and there is no circulation of blood through the lungs. However, it’s the infant’s heart that pumps the blood around and to enable the blood to pass from the venous to the arterial side, a hole exists between the two sides. After the birth this hole usually closes. However, in a certain percentage of adults the hole never closes.

    If a diver ascends too quickly, the dissolved nitrogen in the tissues and blood forms bubbles. The bubbles in tissues cause the bends, or DCI (decompression illness). The bubbles in the blood, however, are trapped in the arterioles of the lungs and gas then diffuses out into the lungs and is breathed out. But if the hole is not closed, some of these bubbles pass directly into the arterial part and cause the arterial blood embolism.

    The other major cause is pulmonary barotrauma caused by holding breath while ascending after breathing compressed gas. In essence the overpressure of the gas in the lungs literaly tears the lungs open from the inside out. This can have several effects, among them pneumothorax (a collapse of the lung), air under your skin (you literally get a bubble under your skin, ususally in the neck area) and again arterial gas embolism, where the gas is forced through the torn venoles in the lungs into the arterial circulation and from there into the tissues and, most importantly, the brain. And this is the greatest danger – parts of brain get cut off from circulation and you have very serious problems. It takes a surprisingly small overpressure to do damage: breathing air at just 1.2m below surface and holding your breath can be enough.

    A good place to read about these things is

    Concerning the buoyancy of divers: The neoprene divesuits compress as you go deeper and therefore become less buoyant. In the old days, before the BCDs (buoyancy control devices – essentialy air bladders that you war in a form a jacket and that can be filled with air from a scuba tank and emptied into the water), the divers would weigh themselves with the ammount of lead that would depend on their planned depth. Usually they had to swim down to about half the planned depth and the rest was just letting the gravity do its work. On the way up they again had to swim to about half way up and then the buoyancy would complete the job. Nowadays BCDs are used to compensate the changes of buoyancy and maintain neutral buoyancy at any depth.

  9. I love this comment thread. Write on octopuses, get small treatises on air bubbles. Awesome.

  10. Tk

    Thanks, everyone; your comments really clear things up for me. I kept thinking the argonaut was somehow supposed to sink naturally to the desired depth and stay in a stable equilibrium — which would be nice and relaxing for her, but doesn’t physically make sense.

  11. Jan

    Oh and one more interesting fact about controling buoyancy in nature. Sperm whales (short for spermaceti whales) contain spermaceti organ above their skulls. Spermaceti is a waxy substance that has a melting point somewhere between the water temperature and the whale body temperature.

    Before diving the spermaceti is cooled with water, solidifying it, thus increasing density and decreasing buoyancy, making it easier to dive. During the dive the substance slowly heats up, decreasing density and increasing buoyancy, making it easier to surface.

  12. Andy Somerville

    Several of your links back to other discovermagazine blog posts are broken. It looks like a second date is encoded in the url for some reason.

  13. To answer Zach’s questions briefly:

    1) The shell walls are quite thin, less than a mm thick and easily broken. They are much thinner and weaker than those of the true nautilus.

    2) Despite the resemblance, the shell differs from that of nautilioids and all other molluscs both in terms of structure (it is calcitic rather than aragonitic) and is secreted not by the mantle as in other molluscs but by specialized organs on the arm. For these reasons it is regarded as an evolutionary novelty and not a reversal. The shell is not homologous with the ancestral molluscan shell as are the internal and external shells of other cephalopods.

    3) The shell of the Argonaut is only secreted by mature (fully grown) females and the females can survive outside the shell.

    The evolutionary origins of the Argonaut shell have inspired quite a lot of, often fanciful, speculation but so far as I know is still a complete mystery (perhaps Finn has some more secrets to reveal about this though!)

  14. Mr Know It All

    OK I kind of forget all the right names and terms…. so I will illustrate the principle of boyancy and sinking into the abyss.

    OK a submarine can be essentially thought of as an inverted jam jar…. and with a certain amount of mass, with a certain volume of air inside it, it will descend to a certain depth and “float” there’.

    However if the mass increases or the jar is forced to go deeper, the pressure causes the air inside the inverted jam jar to decrease in volume, thus decreasing it’s displacement, thereby lowering it’s boyancy.

    It’s the same reason that submarines can get into a situation that if they dive really fast to depth or some other situation occurs – they get into a position that no matter how much air they blow into the tanks to increase the volume of air and thus increase the boyancy of the craft – that the air keeps getting compressed into a smaller and smaller volume – and the sub in certain situations such as loss of motive power etc., will continue to sink until it implodes or hits bottom.

    This is what the Agonaut is doing when it descends down past neutral boyancy – and it rides at depth – and I assume that it mostly operates within a narrow band of negative to neutral boyancy most of the time.

    I do not know enough about the animal, but I would speculate that it probably could expand the airbubble in a diaphramatic way, to change the level or depth of boyancy, where neutrality occours, and that it may not technically be operating at a negative boyancy, in the region where negative boyancy commences.

  15. Jan

    The Argonaut does not have the same problem as the submarine. First of all, the Argonaut, from what I’ve gathered from the article, has a passive buoyancy system: it loads the air it at the surface and then keeps the air bubble inside the shell without any means of adding air (it can release it, but that just decreases buoyancy). It does not have any diaphragms that could actively change the volume of air. Even if it did, the shell would have to be very strong for the diaphragm to work and we know that not to be the case.

    While I don’t know the depths at which the Argonaut rides, let’s assume that it takes on enough air for it to be neutrally buoyant at 10 metres, i.e., 2 bar absolute pressure. Since the pressure at the surface is 1 bar absolute pressure, the air inside the shell is compressed to one half of its volume at 10 m. The absolute change in buoyancy depends on the change of the volume of air, so let’s assume that it takes on 16 millilitres of air, ammounting to 16 grams of buoyancy at the surface (you’ll see why such an “odd” number in a moment).

    If it’s neutrally buoyant at 10 metres, where the air is compressed to 8 millilitres, at the surface its shell contains 16 mililitres of air, giving it 8 grams of extra buoyancy compared to the 8 mililitres at 10 metres. That means it will have to provide 8 grams of vertical motive force at the surface to commence the dive.

    Now let’s assume that it becomes distracted by a beautiful Argonaut of the opposite sex, forgets to swim, and sinks to 30 metres. At that depth the absolute pressure is 4 bar and the volume of air is now 4 mililitres, which ammounts to 4 grams of negative buoyancy (a reduction from 8 to 4 millilitres of air). Since we know that it can produce 8 grams of vertical motive force to dive, it can easily provide half that ammount to be able to swim back from 30 to 10 metres.

    If it sinks to 70 metres, the absolute pressure is 8 bar, the volume reduces to 2 mililitres, and it has to provide 6 grams of force – again not a problem. If it sinks to 150 metres, the absolute pressure is 16 bar, the volume reduces to 1 millilitre and it has to provide 7 grams of vertical force – again not a problem. In fact, it can descend to the bottom of the Mariana Trench and still only has to provide at most 8 grams of vertical force to enable it to rise to the surface again, because in the limit of infinite pressure, the air is compressed to zero millilitres, thus making it 8 grams negatively buoyant.

    An if you assume it normally rides at a depth of 30 metres, 4 bars of absolute pressure, the force needed to ascend from any depth deeper than 30 metres is at most one quarter of the force it needs to descend to 30 metres.

    While I don’t know the details of balast tanks in a submarine, I suspect the real problem there is the rate of change of buoyancy. There’s a limit to how fast the water can enter and leave the tanks through the holes in the hull but, more importantly, how quickly the air can be pumped into the tanks through the pipes. What I think happens is that after a certain change of depth you simply cannot pump air quickly enough into the tanks to offset the input of water and the bubble of air in the tanks keeps getting smaller and smaller.

    There’s a similar situation in scuba diving. When you ascend, you have to vent air from the BCD to keep the volume of air inside the BCD more or less constant, compensating only for the increasing volume of the neoprene and the slowly decreasing mass of air in the tank – a difference of around 4 kilograms from the start to the end of the dive for a 15 litre tank pumped to 200 bar.

    If you’re do not start venting soon enough after the start of ascent, you get into the situation where the outflow of air from the BCD is too low to account for the increasing volume caused by the decreasing pressure and you get a so-called runaway ascent, a Very Bad Thing that can lead to DCI. Exactly the same situation occurs with a drysuit, that contains an even larger ammount of air, and where it’s even more crucial to keep it balanced. A drysuit diver in an uncontrolled ascent can breach the surface of the water almost like a jumping Great White shark.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Not Exactly Rocket Science

Dive into the awe-inspiring, beautiful and quirky world of science news with award-winning writer Ed Yong. No previous experience required.

See More

Collapse bottom bar