A Strange Beauty: Understanding the Octopus


PBS LearningMedia and April Qian

The various mazes the Root Lab used to study octopus behavior and preferences.


Professor Tom Root in the biology department spoke on Friday, March 2, to pods of students and professors from a plethora of scientific disciplines, including biology, neuroscience and psychology, about his lab’s prolific work on the California Two-Spot Octopus (Octopus bimaculoide), a species considered to be the white mouse of cephalopod research.

The title of the talk was “Strange Beauty,” a reference to a line in a poem by Arthur Clement Hilton. Titled “Octopus,” the poem intentionally caricatures the “eight-limbed and eight-handed” creature into a cunning marine monster, an entertaining but somewhat anthropomorphic characterization that has limited scientific accuracy.

From a neurobiological perspective, octopuses are of particular interest because the lobes and regions of their brain are remarkably similar to those in mammals, so any insights gleaned from research on octopuses could potentially allow for better understanding of the mammalian and human nervous system.

Research on cephalopod intelligence boomed after an influential 1992 study published in Science Magazine by researchers Fiorito and Scotto on observational learning in Octopus vulgaris, which focused on whether octopuses could learn by observing the actions and corresponding consequences of other octopuses.

Using positive and negative reinforcement associated with white and red balls, respectively, the researchers reported that the octopus learned from the example of an octopus on the other side of the glass to choose the correct ball, despite having been exposed to neither positive nor negative reinforcement associated with either ball.

Unfortunately, though this paper initially inspired much excitement in the field, subsequent follow-up attempts by other researchers, including that of the Root lab in 2008, to reproduce Fiorito and Scotto’s results were found unsuccessful, and the octopuses used in the original study were found to have been raised in unusually competitive and forced social environments, rendering the results unreproducible and flawed.

The Root lab then switched gears to the feeding behavior of Octopus bimaculoide, focusing on their association of preferred conditions or food with visual, auditory and chemical stimuli.  The work of alumna Alexa Warburton ’10 expounded upon the ability of octopuses to learn to associate a visual sign such as “x” or “o” to preferred or non-preferred conditions in a variety of paradigms. In just one summer, the accuracy rate of octopuses finding their preferred dark chamber in a T-maze increased from 50 percent to over 80 percent.

Other paradigms such as the round maze and the Y-maze were also used to test how quickly and accurately the octopuses associated different images, sounds, motion and chemicals with their preferred conditions or source of food.

Later researchers in the Root lab tested the same visual stimuli in the tank that the octopuses inhabited to reduce the impact of other variables such as the observer effect. Throughout the past decade, the work of various students in the Root lab showed that octopuses preferred dimly lighted areas and tended to attack prey more often in the presence of red colors, contrasting colors, polarized light and a chemical called proline that is often used in the fishing industry.

Cece Wheeler ’19 started working in the octopus lab this spring.

“Twice a week I go into the lab and tie a fiddler crab to a piece of fishing wire, and then lower that into the octopus tank to see if I can get one of the octopuses to come out overtop different types of substrates that I’ve laid down,” she said.  “As [the octopuses] move onto the substrate, they camouflage themselves to it, typically in a uniform, mottled or disturbance pattern in response to the different patterns of substrate.”

A future area of research for the Root lab lies in the development of camouflage in baby octopuses, which often begins immediately after birth.

“The lab has done some past work with learning, but I think the consensus now is that it’s hard to judge how an animal is making decisions without understanding the basics of how they are receiving and interpreting information from the world around them,” Wheeler said. “Right now I’m starting work with camouflage behaviors. I want to know more about what environmental cues initiate different camouflage patterns. There are a lot of different aspects to an octopus’ surroundings, and I’m curious as to which ones it is responding to when it camouflages itself.”

Instead of rationalizing octopus behavior through a human lens, such an approach attempts to understand the decision-making process of octopuses through examining their perception of the world.

“All of the our animals are juveniles right now, so they aren’t full grown, and some are slightly bigger than others,” Wheeler said. “They also like to hide in flower pots, behind bricks, or under pieces of coral, so it’s kind of tricky figuring out how to lure them out effectively. I also record everything on a video camera so I can go back and look at the footage later. Hopefully I will be able to categorize camouflage responses to different substrate patterns, and come up with an analysis of that data this spring.”