Cephalopods use their vision a lot; it’s a big part of how they orient themselves in the water, hunt, and recognize predators and each other. It makes sense, then, that they have particularly well-developed eyes. In fact, they are the only invertebrates to have camera eyes. Camera eyes are eyes that focus an image on the retina, which transmits the image to the brain where it can be used to help the animal get around, find food, and whatever else it needs to do. Like cameras, these eyes can focus on different things – they focus in a different way than the eyes of vertebrates, though. Vertebrate’s eyes focus by changing the shape of the lens, while the eyes of cephalopods focus by moving the lens closer to or farther from the retina.
Drawing of a cephalopod eye and the optic lobe in cross-section, from Young 1962
So, cephalopods have these big fancy eyes – once light hits the lens, it forms a well-focused image on the retina. What happens then? When light hits the retina, it causes specialized nerve cells called photoreceptors to send electrical signals through other cells in the retina and the optic nerve to the brain. Along the way, these signals are converted into information that the brain can use to put together a picture of the world. For example, this information allows the animal to identify objects, boundaries, and features of their environment – something that is helpful to the organism, as this lets it identify things like sources of food or predators that might hurt it.
I am going to talk more about cephalopod vision in a moment; to get there, though, I want to explain a few things about the way vision works in you and I. Early on in the process, the visual system starts to turn the raw image from the eye into information about the environment. To understand this process, let’s think about something familiar – a computer screen. A computer screen displays an image by changing the changing the amount and color of light that is projected by each pixel. You can look at this information in different ways: at the most basic level, you could look at a list of the brightness and color of each pixel. This information would be complete, in that it would represent everything on the screen faithfully, but it wouldn’t be very useful (unless you can remember a list of a million numbers and then imagine a picture based on them, which I sure as heck can’t.) You could write a program that would take this raw information and recognize different features that are being displayed on a screen; for example, you could write a program to recognize regions of the display that were high contrast, or exceptionally bright, or a specific color, or a specific shape. This would tell you a little more about what was on the screen, but it wouldn’t actually be very useful for most things; these sorts of tools would give you a list of statements like “there is a white rectangle in the center of the screen and there are small black objects within this square.” The way we analyze what’s on a computer screen takes it a few steps further – our eyes, of course, take in light from each pixel, but then our brains analyze it into features, shapes, objects, and patterns. These pictures of the environment somehow get mixed in with our memories and thoughts, and we can get an idea of what’s going on in the environment – for example, I can say “there is a word processer open on my computer screen, and there is an unfinished blog post in it.” We often think of the brain as the place where things like this happen, but in fact, our nervous systems start interpreting the information from the visual field before it gets to our brains.
As you can see in the little diagram below (you can click it for a bigger version), the retina, the part of the eye that sits at the back and detects light, is made up of a bunch of layers of cells. Out of all of these, the cells that actually sense light (the photoreceptors) sit way at the back. Light passes through all the other cell layers, is picked up by the rod and cone cells, and then information about the light is passed back through the layers of the retina until they reach a layer of cells called “ganglion cells”, which send the information to the brain through the optic nerve. It’s easy to see that there’s a lot more to the retina than just picking up light – all those other cell layers must be there for a reason, right?
It turns out that by the time information is sent from the retina to the brain, it has already been processed from raw information about light entering the eye into a simple picture of the environment. If you look at the electrical activity of neurons in the optic nerve, which carries information from the eye to the brain, you find that these cells don’t simply respond to points of light like photoreceptors do. They respond to more complex things, such as a dark spot on a light background, a light spot on a dark background, the edges of an object, or movement in a specific direction. Each ganglion cell (these are the cells that make up the optic nerve) is programmed to respond to features of a certain size and position in the visual field. Before information about the visual field even leaves our eyes, it is processed a few steps towards a useable picture of the environment.
(If you are interested in reading more involved stuff about retinal ganglion cells in mammals, check out this great link.)
In octopuses, the same sort of process must go on – after all, the information from the eye has to get passed on to the brain somehow, so that the brain can use it. This image processing doesn’t go on in the retina, though; it takes place in the brain itself.
J. Z. Young was a scientist who studied the octopus brain extensively, and he found that when he cut the optic nerve of the octopus the photoreceptor cells in the eye died – along with his other observations, this indicates the the photoreceptor cells in the octopus’s eye send information straight to the brain, with no other cells in between (in a vertebrate – a mouse for example – cutting the optic nerve would make retinal ganglion cells die, but would spare photoreceptors, because they don’t actually extend outside the retina.)
Young’s observations weren’t made in a vacuum – they were supported by other work as well. Two year before his paper on the optic lobe of the octopus, Edward MacNichol and Warner Love had published a study in which they made electrical recordings from a squid’s optic nerve as it left the eye and found that the signals it sent in response to light shining on the eye looked like those of a simple photorecptor rather than the more complex signals that you would find if you recorded from a vertebrate’s eye.
Dr. Young also noticed that the part of the brain that these cells send information to, the optic lobes, have layers on the outside of them that look like the image-processing layers of the retina, and probably have the same function: to take the raw information from the photoreceptors and start to identify features in the environment. In fact, he was so convinced that the outer layer of the octopus’s optic lobe was doing the same thing as our retinas that he referred to it as the “deep retina”.
So what happens to the information after that? One (relatively) easy way of asking this question is to figure out where the optic lobe sends connections to, and where it gets connections from. J. Z. Young’s work gives us answers to these questions – sort of. In his 1962 paper on the octopus optic lobe, he lists all the places that nerve fibers from the optic lobe end up, and it’s most of the brain! This makes sense, though, when you think about all the things that octopuses use visual information for – they use it to guide their movement through the water (so the optic lobe and the parts of the brain that control the movement of the arms need to communicate with each other); they use visual information to figure out what colors and/or patterns they should display on their skin (so the optic lobe and the chromatophore lobe, which controls the colored organs in an octopus’s skin, need to communicate with each other); the octopus uses visual information to learn where and what things are (so the optic lobe and the vertical lobe, which is involved in memory, have to communicate with each other) – the list goes on, but you get the idea. After it gets to the optic lobe, information from a cephalopod’s eye is sent all over the place, which makes its paths much more difficult to trace. You imagine can that as we follow the flow of information from the eye, it branches out like a tree, getting more and more spread out and finely divided as it moves through the brain to whatever its destinations are. While we may not be able to follow it this far at the moment, as we get better and better at figuring out how brains work and put more and more time into studying them, it seems likely that we’ll figure out what those destinations are.
This is a sometimes useful, but ultimately wrong way to think about how information from the senses makes its way through the brain – in truth, it has no real destination. The brain’s main job is to control behavior, and so there’s no place within the brain where information from the senses gets to and then stops. The brain takes in information from the senses, and uses it to guide behavior – in a sense, then, the final destination of the information any neuron sends into the brain is a behavior. One can imagine the same tree-like structure of signals but in reverse to describe the flow of information from the brain into a behavior; at its most basic level, a behavior is just a specific sequence of muscle contractions. However, if we follow the flow of information backwards through the nervous system, we’ll find more and more neurons that play a part in controlling that behavior. As we go further and further, each individual neuron, each bit of information, will presumably play a smaller part in the overall behavior. If we kept following the branching tree, we would eventually end up so far away from the starting point that it would be hard or even impossible to figure out how we had gotten there – just like we had picked a random cell in the brain and said “what does *this* cell do?” If it’s a cell that’s particularly close to a certain sensor (like the ear) or a certain behavior (like the movements of a dancer), we have a good shot at figuring out pretty much exactly what it does, but there’s still a big area in the middle where it is much harder to figure out what different groups of neurons are doing and what effect they have on behavior. This is as true for cephalopods as it is for humans, but is part and parcel of studying the brain – when we try to understand something so complex, we have to expect some big question marks to pop up.
Thanks for reading! See you next week!
MacNichol, E., & Love, W. (1960). Electrical Responses of the Retinal Nerve and Optic Ganglion of the Squid Science, 132 (3429), 737-738 DOI: 10.1126/science.132.3429.737
Young, J. (1962). The Retina of Cephalopods and Its Degeneration After Optic Nerve Section Philosophical Transactions of the Royal Society B: Biological Sciences, 245 (718), 1-18 DOI: 10.1098/rstb.1962.0004
Young, J. (1962). The Optic Lobes of Octopus vulgaris Philosophical Transactions of the Royal Society B: Biological Sciences, 245 (718), 19-58 DOI: 10.1098/rstb.1962.0005