Cuttlefish Hunting Behavior

From February 2014 to February 2015, I collaborated with the Marine Biology Lab at Woods Hole, USA, to study cuttlefish, an invertebrate marine animal related to octopus and squid, from the perspective of neuroscience. My primary goal was to learn about their behaviour, how to take care of them, and how to modify our lab’s behaviour experiment setups for rats to work with cuttlefish. I took care of cuttlefish in all stages of their life cycle and even developed unique relationships with individuals.

Before I started any experiments, I spent most of my days feeding the cuttlefish hatchlings and cleaning their home tubs:

Taking care of the cuttlefish on a daily basis taught me a lot about their behaviour, especially their hunting behaviour. I tried out many different experimental setups to study their hunting behaviour more closely, and learned a lot about arduinos and cameras and how to film sea creatures before I came up with the final design of the Cuttle Shuttle experiment.

Many thanks to Kendra Buresch, Stephen Senft, Alex Schnell, Andrew Carvey, Arthur Petron, Troy McInerney, Kelsey Cramer, Corinne Cramer, Andrea Rummell, George Bell, Alan Kuzirian, Barbara Burbank, Lyda Harris, Dan Calzarette, and Roger T Hanlon for their support and assistance at Woods Hole during this project. Additional thanks to Adam Kampff and the Intelligent Systems lab for enabling and supporting this collaboration.



The active camouflage of coleoid cephalopods (cuttlefish, octopuses, and squids) presents an opportunity to non-invasively study whole organism behaviour and single-unit neural activity simultaneously, a possibility which is reviving the attention of neuroscience researchers in the 21st century [4, 8]. A rich body of literature, reaching back before the 1900s, describes the remarkable diversity of cuttlefish body patterns in great qualitative detail. However, to date there are no unambiguous mathematical definitions of these body patterns [3, 9]. In particular, there are very few quantitative characterisations of body pattern changes of cuttlefish outside the context of camouflage and background matching. This paper presents a low-cost and open-source method for reliably evoking “acute” (as opposed to “stable”, as defined by [2]), event-locked, and ethologically relevant changes to body pattern in captive Sepia officinalis, and offers a first step toward mathematically characterising these rapidly transient body patterns, in the hopes of accelerating the re-diversification of model species used in the study of how nervous systems control behaviour [6, 10]. In light of recent advances in highly accessible computer vision tools and techniques for tracking non-captive behaviour of a large diversity of species [1, 5, 7], the methodology presented here has great promise for use not only in the lab but also in the wild.

[1] J. M. Graving, D. Chae, H. Naik, L. Li, B. Koger, B. R. Costelloe, and I. D.Couzin. Deepposekit, a software toolkit for fast and robust animal poseestimation using deep learning.eLife, 8:e47994, 2019.
[2] R. T. Hanlon and J. B. Messenger. Adaptive coloration in young cuttlefish (sepia officinalis l.): the morphology and development of body patterns and their relation to behaviour. Philosophical Transactions of the Royal Society of London. B, Biological Sciences, 1988.
[3] E. J. Kelman, D. Osorio, and R. J. Baddeley. A review of cuttlefish camouflage and object recognition and evidence for depth perception. Journal of Experimental Biology, 2008.
[4] A. Laan, T. Gutnick, M. J. Kuba, and G. Laurent. Behavioral analysis of cuttlefish traveling waves and its implications for neural control. Current Biology, 2014.
[5] G. Lopes, N. Bonacchi, J. Fraz˜ao, J. P. Neto, B. V. Atallah, S. Soares, L. Moreira, S. Matias, P. M. Itskov, P. A. Correia, R. E. Medina, L. Calcaterra, E. Dreosti, J. J. Paton, and A. R. Kampff. Bonsai: an event-based framework for processing and controlling data streams. Frontiers in neuroinformatics, 2015.
[6] P. R. Manger, J. Cort, N. Ebrahim, A. Goodman, J. Henning, M. Karolia, S.-L. Rodrigues, and G. ˇStrkalj. Is 21st century neuroscience too focussed on the rat/mouse model of brain function and dysfunction? Frontiers in Neuroanatomy, 2008.
[7] T. Nath, A. Mathis, A. C. Chen, A. Patel, M. Bethge, and M. W. Mathis. Using deeplabcut for 3d markerless pose estimation across species and behaviors. Nature Protocols, 2019.
[8] S. Reiter, P. H¨ulsdunk, T. Woo, M. A. Lauterbach, J. S. Eberle, L. A. Akay, A. Longo, J. Meier-Credo, F. Kretschmer, J. D. Langer, M. Kaschube, and G. Laurent. Elucidating the control and development of skin patterning in cuttlefish. Nature, 2018.
[9] M. Stevens and S. Merilaita. Defining disruptive coloration and distinguishing its functions. Philosophical Transactions of The Royal Society, 2008.
[10] M. M. Yartsev. The emperor’s new wardrobe: Rebalancing diversity of animal models in neuroscience research. Science, 2017.

Paper Figures

Below are the latest drafts of the figures included in the academic paper I’ve written about this project, titled “The Cuttle Shuttle: reliably evoking acute, event-locked, and ethologically relevant changes to body pattern in captive Sepia officinalis” (full-text available soon on Bioarxiv!):

Diagram: Camouflage and Signaling sequences associated with Prey Capture
Figure 1: Diagram of body pattern changes associated with prey capture. Based on personal communication from R.T. Hanlon and anecdotal evidence from other field biologists. This study focuses on the stage immediately after a prey capture event, circled in red in the diagram above. Illustration credit for Prey Capture Event: Jennifer Deutscher.
Schematic drawing of the Cuttle Shuttle experimental setup
Figure 2: The Cuttle Shuttle “hunting box”. An acrylic box (43 x 43 x 81 cm) with “robotic prey” on the left, perched on top of the wall of the hunting box, and “home base” on the right, where cuttlefish can hide and feel safe while acclimating to the hunting box. Expremental setup rendered in Google SketchUp and labeled using Adobe Illustrator by Danbee Kim.
Video pre-processing workflow
Figure 3. Example of the process of cropping and aligning a single frame of a Cuttle Shuttle hunting session video recording. Tentacle shot clips were temporally aligned to the “tentacles go ballistic” (TGB) moment and included the 3 seconds before TGB and 3 seconds after TGB. Manual alignment of all tentacle shot clips was done using Final Cut Pro and Adobe Premiere Pro by D. Kim.
Video analysis workflow
Figure 4. Workflow for transforming cropped and aligned images of the mantle body pattern during hunting into a numerical timeseries. Body pattern images processed with Bonsai ( and figure generated using Adobe Illustrator by D. Kim.
Table: Accuracy of tentacle shots
Table 1. Accuracy of seizure via tentacle shot while hunting the robotic prey, for all animals throughout the entire experimental protocol.
Visual Ethogram of body pattern changes during hunting behaviour in captive *Sepia officinalis*
Figure 5. Still-image ethogram of the changes to body pattern on the mantle of one hunting Sepia officinalis. The top row shows all tentacle shots that resulted in a catch, and the bottom row shows all tentacle shots that resulted in a miss. Each matrix of images constructed from cropped and aligned video clips of tentacle shots. Figure generated using Adobe Illustrator by D. Kim.
Characterising the 'tentacle shot pattern' using an 'edginess' score (as measured by OpenCV algorithm Canny Edge Detector)
Figure 6. A: Raw scores from individual animals were z-scored then pooled (N of animals = 5, total N of catches = 59, total N of misses = 81). B. A shuffle test for significance (number of shuffles = 20000) at every timebucket was used to calculate at which time point the mean values describing “catch” tentacle shots became significantly (p<0.05, pointwise; corrected for global p<0.05 at upper bound = 99.994 and lower bound = 0.006) different from the mean values describing “miss” tentacle shots. Plots generated in Python and figure assembled using Adobe Illustrator by D. Kim.

Supplemental Figures

Supplemental Figure 1: The Experimental Tank
Fig. S1: The Experimental tank, showing the hunting box, with hardware setup controlling the robotic prey, inside the circular holding tank lined with LED lighting, with overhead Point Grey FlyCap2 camera. Photo by D. Kim; figure generated in Adobe Illustrator by D. Kim.
Supplemental Figure 2: Overhead view of the hunting box.
Fig. S2: Overhead view of hunting box inside the experimental tank. This image is a screenshot from the session videos recorded by the overhead PointGrey FlyCap2 camera. Figure generated in Adobe Illustrator by D. Kim.

Video Summary

Research questions

In this experiment, cuttlefish must hunt for their food 4 days out of 7; their prey is a piece of shrimp at the end of an arduino-controlled skewer. We wanted to ask the following questions:

  • Can cuttlefish hunt artificial prey? If so, how complex of an artificial prey can cuttlefish hunt?

  • Does the cuttlefish display a reliable sequence of body patterns when it first notices the prey?

  • Does the cuttlefish display a reliable sequence of body patterns when it catches its prey?

  • What are the firing dynamics of cuttlefish tentacles?

Want to get involved?

To contribute to this project, please visit the Cuttle Shuttle project page at Every Mind Online.

Experimental Protocol

The most current draft of my experimental protocol.


Videos of the experimental sessions can be found online here:


Hunting Behaviour Ethogram

Below are acrylic models I made to show the main phases of a cuttlefish hunt:

Cuttlefish Hunting Behaviour: At Rest
When a cuttlefish is ‘at rest’, or just hanging out, all of their arms lie flat over their mouth. Their eyes are usually on the sides of their head, in a position optimal for having a 360 degree view all around them. The body pattern on the cuttlefish’s mantle, or back, is optimised to camouflage the cuttlefish and blend into its surroundings.
Cuttlefish Hunting Behaviour: Attention
When a cuttlefish first notices food or prey, it will turn to face the food or prey, and often it will raise its two middle arms.
Cuttlefish Hunting Behaviour: Positioning
As stealth hunters, cuttlefish sneak closer to their prey or let their prey come closer to them, as the cuttlefish hide by camouflaging themselves or burrowing into sand on the ocean floor. The two middle arms often remain raised in front of their face. When the cuttlefish is about one body length away from its food or prey, the next four arms create a barrel to guide and aim their tentacles, the tips of which can become visible during this phase.
Cuttlefish Hunting Behaviour: Tentacles Go Ballistic
When the cuttlefish is ready, it will throw its tentacles towards the food or prey, a moment that we call ‘tentacles go ballistic’ (TGB). The middle two arms are usually still raised, and the next four arms are still formed into a ‘barrel’ to aim the tentacles. During this phase, the bottom 2 arms spread out, as if to create a stabilizing tripod for the moment when tentacles go ballistic. The TGB moment also marks the appearance of the ‘tentacle shot pattern’, a unique, highly fractalated body pattern that appears only in this moment and not at any other time in the cuttlefish’s behavioural repetoire.

Video analysis

Further information about analysis work on the Cuttle Shuttle video dataset can be found here.


Presentations and Videos

Leave a Comment