Common cuttlefish learn to master the “prawn-in-the-tube” test by observing conspecifics

Eduardo Sampaio shared his newest findings on social learning in the mostly solitary common cuttlefish, Sepia officinalis, with Cultured Scene. The whole study has recently been published in Animal Cognition.

The Paper

Cultured Scene (CS): What is the key finding of your study?

Eduardo Sampaio (ES): Our main finding was that a mostly solitary species such as the common cuttlefish, Sepia officinalis, is capable of learning by observing conspecifics. After similar findings on Octopus vulgaris, this seems to make the cephalopods one of the only invertebrate groups where social learning is documented, if we exclude eusocial species (e.g. ants and bees). Moreover, we showed that cuttlefish are capable of complex learning at a very young age (five days post hatching), while still undergoing neural maturation.

CS: What implications (if any) for social learning capacities during development can be drawn from your study?

ES: From our study, it perhaps becomes apparent that the ability to understand the properties of objects by watching conspecifics interacting with those objects (in other words, affordance learning) can come soon in the ontogeny of these animals. It would be interesting to explore if this is the case for other animals as well, both on invertebrate and vertebrate branches, since we know that, for example the capabilities for object permanence, increase through the developmental Piagetian stages in humans.

CS: Why is this topic important, and how do you feel it relates to social learning and cultural evolution more broadly?

ES: I think it is relevant to our understanding on the distribution of complex cognition and social abilities across the tree of life, and to the current consensus that complex cognition arose multiple times throughout evolution in different pathways. As a previous point, it is noteworthy that while vertebrate neural systems follow a centralized blueprint, in cephalopods, the neural arrangement is much more distributed, with roughly 2/3 of their neurons being located outside the brain. And what our study further confirms is exactly that regardless of the configuration of neural substrates, both vertebrates and invertebrates developed similar mechanisms to incorporate social information, most likely so that the observing individual can increase its fitness in nature.

The process

CS: How did you arrive at the idea for the study and was there a specific reason for using cuttlefish as your study organism?

ES: From the get-go I was interested in exploring the cognitive abilities of cephalopods, which I have always found as severely understudied considering the level of intelligence they display. A test that I saw performed multiple times in the literature was the prawn-in-the-tube procedure, where you present a glass tube containing prey to the cuttlefish, and the subject must learn to inhibit predatory behaviour. This test measures associative learning through the incorporation of multimodal sensorial (tactile and visual) cues to recognize the existence of a glass tube in front of the prey, that makes it inaccessible. There are numerous experiments performed with small tweaks aiming to disentangle the mechanisms behind learning, and since it is a somewhat complex process, I wondered what would happen if I just added another arena with a cuttlefish that could observe the test subject (demonstrator), and how would that cuttlefish (observer) fare, if subsequently tested.

CS: What were the key changes regarding your experimental design compared to previous studies?

ES: Well, the rationale was simple and straightforward, as I describe before. However, before submitting to Animal Cognition, we had submitted the paper to another journal. At that stage, we had only performed part of our now-published experimental design, and we received some suggestions from two reviewers on additional experimental conditions to better control for several factors. We found these suggestions very helpful and incorporated more conditions into our experimental design, since they did help explain our results better, and also explore which mechanisms of learning were occurring.

CS: What were the best and worst aspects of data collection – any funny stories? Any specifics about working with cuttlefish?

ES: If you have ever seen videos of cuttlefish hunting, they are pretty cool. But cuttlefishes “hunting” a glass tube, striking it with their tentacles and trying to pull it with their arms… they are hilarious! So that was definitely the funniest part of the work. As for the worst, I would have to say the “wasted” hours between trials. During the 12h of photoperiod available, we would spend around 8h doing experiments everyday. We had to change something in the setup (e.g. insert or remove tube with prey) every 10 minutes, but that time slot in between was not enough for anyone to work on something else, so we would end up just standing or sitting in front of the door behind where the experiments were taking place. Everyone had a different way of spending that time. I listened to, and got to know, a lot of new music (which is great), but it still was very boring. Of course that when you finally get the results in the end, it makes everything worthwhile.

Short video sequence of the social learning experiment using the “prawn-in-the-tube” procedure with juvenile cuttlefish (observer at top and demonstrator at bottom). Credit: Eduardo Sampaio.


CS: How did you manage the writing process? Was it straight forward, or were there challenges?

ES: I really enjoy statistical analysis and writing, so that was not a problem for me. Also, as it was the first time that I wrote a paper focusing more on behavior, it was a great opportunity to read on a lot of interesting topics, and a collection of nice books and papers I had on the side since starting the PhD, and had not found a chance to read yet. I really enjoyed writing the paper.

CS: How was the peer review process? 

ES: So after we added the extra conditions, we tried re-submitting the new version to that same journal but the editor wouldn’t have it. We then ended up submitting where it is now published, and the review process was very straightforward. We got a bit delayed in the process due to the ongoing pandemic, but in the end it went smoothly.

What’s next?

CS: Will you be following up on this research? What questions interest you next, based on your findings? 

ES: My PhD focuses on exploring the cognitive abilities of cephalopods. So after showing that they are capable of complex forms of learning through conspecifics, I am now interested in checking how they fare with heterospecific cues. Octopuses and fishes hunt collaboratively in nature, so this provides a really unique system to evaluate social learning, interspecific communication, and several other complex processes, at an interspecific level. I have been doing this project for the last 2 years (we were awarded a National Geographic grant for it), and hopefully will be able to start publishing some results during this year. I am very excited with what we are seeing.

CS: As early career researchers, we’re always learning. Is there anything you’d do differently in future, based on your experiences conducting this study?

ES: I would have definitely done all experimental conditions from the get go! But sometimes it is only when you have someone weigh in that is not directly involved in the experimental design, that some flaws become apparent, and they point you to the missing pieces. I guess it is also part of the academia process, and it is important to make your results as robust as possible!

CS: Finally – what do you think are some of the big questions / challenges facing the field of cultural evolution and social learning?

ES: Undoubtably the extent to which we see it in the tree of life is still a main question in the field, particularly when we are talking about cultural transmission. It is a more specific aspect that has seldomly been studied outside of primates and mammals (and perhaps some species of birds, more recently). Since social learning and cultural transmission provide such a huge advantage to animals that do it (since they can avoid trial-and-error and incurring in potentially high costs), it would be expected that these features are present in a wide range of animals. I may be a bit biased due to my model system of preference, but I believe that the study of species that diverge from the vertebrate tree branch is the fastest way to learn (or at least provide us with a more comprehensive view) about the necessary physical substrates and environments that enable the emergence of these cognitively-challenging processes, and the mechanisms that underlie them.

CS: Thank you, Eduardo, for sharing your research and experience as an ECR with Cultured Scene!

The publication:

Sampaio, E., Ramos, C.S., Bernardino, B.L.M. et al. (2020) Neurally underdeveloped cuttlefish newborns exhibit social learning. Animal Cognition.

Eduardo Sampaio has a MSc in marine biology and ecology, and is now a PhD candidate in Biology (Ethology) at the Faculdade de Ciências da Universidade de Lisboa (Portugal) and the Max Planck Institute of Animal Behavior (Germany). His research focuses on exploring the cognitive complexity of cephalopod neural systems, and on understanding how social cognition is used during interspecific hunts with multiple fish species. He is also a National Geographic Explorer (see:, loves diving, and recording natural history. (Photo credit: Martim Seco)

Visit Eduardo’s website:

Follow Eduardo on Twitter: