First tool-using invertebrate: octopus with a mobile home

December 15, 2009 • 7:07 am

Yes, I know that P. Z. has already posted on it, but I think it deserves a bit fuller explication, for it’s the first reported instance of tool use in an invertebrate (reported in the news here, and at the journal Current Biology here). In this case the creature is the veined octopus, Amphioctopus marginatus, and the “tool” is a coconut shell that it uses as a protective home, just like a hermit crab uses another creature’s shell.  What makes this real “tool use,” as opposed to what hermit crabs do, is that the octopus carries the coconut shell around with it, like an Airstream, and assembles the two halves into a protective hideaway.  They’re smart little buggers. Here’s the (pardon me) meat of the story; I’ve put the direct links to the movies below:

Between 1999 and 2008, we undertook more than 500 diver hours (day and night) on subtidal soft-sediment substrates to 18 metres deep off the coasts of Northern Sulawesi and Bali in Indonesia. Over this period, we studied more than 20 individuals of the Veined Octopus, Amphioctopus marginatus (Figure 1). Octopuses were encountered in a range of behavioral states — emerged and active on the seafloor (Figure 1A); occupying empty gastropod shells, discarded coconut shell halves (Figure 1B) or other human refuse; or buried within the substrate (with or without accompanying shells; see Supplemental Movie S1 in the Supplemental Data available on-line with this issue). When flushed from shells by the observer, individuals quickly reoccupied the shells. On four occasions (three in Northern Sulawesi, one in Gilimanuk, Bali), individuals were observed to travel over considerable distances (up to 20 m) while carrying stacked coconut shell halves below their body (Figure 1C; Movie S1 ). For all instances of this behaviour, observing divers (JF, MN) remained static for up to 20 minutes at 1–2 metres from stationary octopuses, which emerged from the cover of one or two shells halves, arranged the shell(s) under the arm crown, and departed. Two shell-less octopuses were also observed to extract previously un-encountered coconut shells buried in the substrate, aided by jets of water to flush mud from shells (Movie S1 ).

To carry one or more shells, this octopus manipulates and arranges the shells so that the concave surfaces are uppermost, then extends its arms around the outside and walks using the arms as rigid limbs. We describe this lumbering octopedal gait as ‘stilt walking’ (see Movie S1 ). This unique and previously undescribed form of locomotion is ungainly and clearly less efficient than unencumbered locomotion (i.e. costly in terms of energy and increased predator risk compared with normal walking or the faster jet swimming escape; see Movie S2 ). While ‘stilt-walking’ the octopus gains no protective benefits from the shell(s) it is carrying as the head and body are fully exposed to potential predators. The only benefit is the potential future deployment of the shell(s) as a surface shelter (Figure 1B) or as a buried encapsulating lair (Movie S1 ).

The fact that the shell is carried for future use rather than as part of a specific task differentiates this behaviour from other examples of object manipulation by octopuses, such as rocks being used to barricade lair entrances [10]. Further evidence that this shell-carrying behaviour is an example of tool use comes from the requirement of the octopus to correctly assemble the separate parts (when transporting two shells) in order to create a single functioning tool.

One of the scientists describes his reaction to the behavior, and I include it here because only an Aussie scientist, God love them, would describe it this way:

“I was gobsmacked,” said Finn, a research biologist at the museum who specializes in cephalopods. “I mean, I’ve seen a lot of octopuses hiding in shells, but I’ve never seen one that grabs it up and jogs across the sea floor. I was trying hard not to laugh.”

Figure 1 (from journal): Veined octopus, Amphioctopus marginatus.  (A) Emerged on sand. (B) Using coconut shell halves assembled as shelter. (C) ‘Stilt-walking’ while carrying two stacked coconut shell halves (see Movie S1). Photos: M. Norman (A), R. Steene (B,C).

MOVIES:  There are two of them at the journal site. Click the links below.

Movie S1

Movie S2

Movie 2

h/t: otter

UPDATE:  Neil notes in a comment below that he’s summarized several other cases of putative tool use by invertebrates.


Finn, J. K., T. Tegenza, and M. D. Norman. 2009.  Defensive tool use in a coconut-carrying octopus. Current Biology 19:R1069-R1070

29 thoughts on “First tool-using invertebrate: octopus with a mobile home

  1. I don’t think I would call this tool use. I don’t think the difference from hermit crab shells (useful to the crab while walking, but not useful to the octopus till it stops walking) is particularly salient or telling. In the video I watched (the one on the BBC), the octopus do not bring separate shells together to form a larger lair, but used shells already next to one another (with no way from the film to tell how they came to be that way, although it wasn’t clear that the two halves were completely separated). If bringing multiple items together to construct a defendable place of retreat counts as tool use, then caddis fly larvae are also tool users.

    1. TOOL: 1 a : a handheld device that aids in accomplishing a task b (1) : the cutting or shaping part in a machine or machine tool

      Caddisflies don’t hold the stones in their hands! 🙂

      1. The operative meaning of “tool” is definitely a contentious subject. But at least the octopus doesn’t do the equivalent of what the larvae does – or rather, that equivalent of lair building can be seen in other octopuses according to the press release. This is different, certainly in the direction of tool rather as in building material. If it gets there is another matter.

        Also, the text here mentions the animals walking with stacked shells, so I don’t get the claim in “the octopus do not bring separate shells together”?

        The shells are next to each other because the octopus prepared them so AFAIU.

        Likely initially, especially as stacking implies dissimilar nut halves, so more likely different individuals.

        And certainly meanwhile. I do not bring bills together in my wallet each time I use them (together or not), they are already prepared so.

  2. My favourite comment from Finn was “I almost drowned laughing when I saw this the first time… I could tell it was going to do something, but I didn’t expect this – I didn’t expect it would pick up the shell and run away with it.”

    (from BBC).

    I’m also a bit confused about this being the first documented case of tool use (and, like Gregory, a bit confused at the definition of “tool use”). I thought octopuses were already known to throw things around their enclosures and shoot out lights when they got bored?

  3. I’ve seen video of an octopus lugging around a glass jug as a surrogate shell. I would think that would be the first recorded invertebrate use of tools, since I saw it on TV more than a year ago and presumably it was filmed somewhat before that.

    Unfortunately, I do not remember the program on which I saw it.

    1. Actually, on googling “Octopus bottle,” there’s dozens of videos of octopi using bottles, etc. as surrogate shells.

  4. I think, perhaps, what makes this more significant, is that the tool in question consists of multiple parts that the octopus has to stack before carrying them.

    Either way, very cool behavior. Those are some fascinating animals.

  5. appy polly loggies for OT comment, but are you going to be blogging about this: ?
    (Venditti, Meade, and Pagel on Red Queen and speciation)

    I’m very curious what you’d say… it seems to me that they’re going out on a limb, so to speak, but I am very far from an expert. For one thing, and remember, I’m not a biologist, I don’t get their distinction between “speciation factors” and “special driving forces.”

    1. This paper is deeply problematic, but needs a reply in a technical journal, I’m afraid. Going into the difficulties here would bore people to death!

  6. When I die I want my hollowed out skull thrown into the sea where these beasties live so they can use it as a lair. It will be the first intelligent thing to have occupied my headspace.

  7. Does this mean that nest-building in birds, and lodge-building in beavers is also a form of tool use?

    I would hope so because I can’t see the difference between those activities and the octopus’ activities.

    I imagine there is some disagreement among biologists about what constitutes tool use.

    Dragging half a coconut shell around doesn’t seem to be enough, although putting a coconut shell on top as a roof seems a little closer.

    But it’s not quite like chimps and some birds using grasses to extract insects from their nests.

    Perhaps it’s similar to nest and lodge building?

  8. How long before the Pacific North West Octopus becomes a reality and starts taking homes from homo sapiens? We must eat this species before it threatens us.

  9. Can it be long before the IDers apply the “purposeful arrangement of parts” argument as well as the “if you find a coconut shell house in the (underwater) sand, you know it had to have a designer” argument to this system??

  10. Given Hermit crabs, one wonders why the excitement. And I agree it is all about definitions. So, as to “tool use”: explain why what has been observed is anything different from claiming fish have discovered the amazing tool of SCUBA, because, yes, they do breathe underwater, and it is self-contained.

    Exactly what is “amazing” here? Of course they do cool things (that ink bit still has me in thrall). So what? In what what way is any of the described behaviour really amazing?

    Oh wait, hold the presses, beavers cut down trees and build dams. Oh, sorry, old story.

Leave a Reply