Where ant colonies keep their brains | Deborah Gordon

Welcome back to "From Our Neurons to Yours," a podcast from the Wu Tsai Neurosciences Institute at Stanford University.

In this episode, we explore the collective intelligence of ant colonies with Deborah Gordon, a professor of biology at Stanford, an expert on ant behavior, and author of a new book, The Ecology of Collective Behavior.

We discuss how ant colonies operate without centralized control, relying on simple local interactions, such as antennal contact, to coordinate their behavior. Gordon explains how studying ant colonies can provide insights into the workings of the human brain, highlighting parallels between different types of collective behavior in ants and the modular functions of the brain.

Listen to the episode to learn more about the intelligence of ant colonies and the implications for neuroscience.

Links
Dr. Gordon's research website
What ants teach us about the brain, cancer and the Internet (TED talk)
An ant colony has memories that its individual members don’t have (Aeon)
The Queen does not rule (Aeon)
Local links run the world (Aeon)
The collective wisdom of ants (Scientific American)
Deborah Gordon: Why Don't Ants Need A Leader? (NPR)
What Do Ants Know That We Don't? (WIRED)

Episode Credits
This episode was produced by Michael Osborne, with production assistance by Morgan Honaker, and hosted by Nicholas Weiler. Cover art by Aimee Garza.

Thanks for listening! If you're enjoying our show, please take a moment to give us a review on your podcast app of choice and share this episode with your friends. That's how we grow as a show and bring the stories of the frontiers of neuroscience to a wider audience.

Learn more about the Wu Tsai Neurosciences Institute at Stanford and follow us on Twitter, Facebook, and LinkedIn.

Transcript

Nicholas Weiler:

This is From Our Neurons to Yours, a podcast from the Wu Tsai Neurosciences Institute at Stanford University. On this show, we crisscross scientific disciplines to bring you to the frontiers of brain science. I'm your host, Nicholas Weiler.

Imagine yourself hovering in midair about a foot above the ground looking down on an anthill. One ant has found a chunk of apple under a bush, and soon orderly lines of workers march from the nest entrance and back to collect this windfall. Other ants appear to be gathering bits of leaf matter for the colony, perhaps to build underground nests. Others detect a spider getting too near the nest, and quickly converge on it, driving it away.

At first glance, it looks like a highly organized operation. But if you watch for a while and focus in on some of the individual ants, you may notice something odd. While the group as a whole is going about its business with remarkable efficiency, the individual ants look, well, sort of lost. Some of those ants collecting food appear to have forgotten which direction they're supposed to be going. Another has lost the trail, and appears to be wandering aimlessly a few inches away. Another's way is blocked by a leaf. It seems unsure what to do next.

So what we're observing here is actually something of a puzzle. As a whole, ants are an extremely successful group, with well over 10,000 species thriving in a wide array of environments around the world. They're able to quickly find food and mobilize in large numbers to feed colonies of thousands or tens of thousands of individuals. Workers can switch flexibly between different jobs as needed, switching from foraging to defense to caring for young. But individual ants, I probably wouldn't hire them.

So how could individual ants be so hopeless and bumbling, but thousands of ants acting collectively are performing complex behaviors that add up to something like intelligence? Particularly given that there's no centralized control over an ant colony's activities. So what does this have to do with neuroscience? Well, in many ways, understanding the behavior of ant colonies could teach us about the way billions of relatively simple neurons work together in our brains.

Deborah Gordon:

An ant colony is like a brain in that there's no central control. So just as no neuron tells the other neurons what to do, so no ant tells the other ants what to do.

Nicholas Weiler:

This is Deborah Gordon. She's a professor of biology at Stanford, a world-renowned expert on ant behavior, and author of a new book called “The Ecology of Collective Behavior.”

Deborah Gordon:

Ant colonies work through a network of very simple local interactions. Most ants can't see and they operate by smell. And one of the interactions that's very important for coordinating the behavior of an ant colony is a simple an antennal contact. Ants smell with their antennae, and when one ant touches another, it smells the other. And individual ants use their recent experience of antennal contacts to decide what to do. The whole colony operates through a network of these very brief olfactory interactions where one ant smells another.

Nicholas Weiler:

So they're touching their antennae, and that's what drives the decision making. Do you have some examples?

Deborah Gordon:

So the harvester ants that I study in the desert use the rate of antennal contact to regulate their foraging activity. An outgoing forger doesn't leave the nest until it gets enough contact with returning forgers with food. An individual ant assesses its rate of contact using the same kind of excitable dynamics that neurons use. So just as a neuron uses the rate at which it's stimulated by other neurons to decide whether to fire, so an ant uses the rate at which it's stimulated by other ants to decide whether to go out.

Nicholas Weiler:

Are there ways in which the way that we've come to understand how the brain works informs your work studying ant colonies?

Deborah Gordon:

Ants are an amazing system to think about collective behavior because there are so many different kinds of ants, and they've evolved in so many different kinds of environments. And so they provide an opportunity to think about how we see different kinds of collective behavior to solve different kinds of ecological problems.

In the same way brains perform many different functions, and those different functions are under different kinds of ecological constraints. Some kinds of tasks have to be done very quickly, like vision happens really fast. Memory happens much more slowly. So those are different kinds of functions that brains perform that have evolved to use local interactions among neurons differently. Some are more modular, so are concentrated in a particular part of the brain. Some are more diffuse, and that reflects the different kinds of conditions that different brain functions respond to.

For example, the ants in your kitchen that respond immediately as soon as you drop a crumb, and have a very opportunistic way of recruiting and taking advantage of resources when they show up, they're using a different kind of collective behavior than the ants that, for example, forage for very stable resources, that stay there day after day. Like the harvester ants eat seeds that just hang around on the desert floor for days, so they don't need to recruit quickly, and they have a different kind of collective behavior that is more thorough, but slower. We can see parallels in other natural systems to the way that different kinds of collective behavior are organized to respond to different kinds of environmental constraints.

Nicholas Weiler:

And you have a book coming out soon on just this topic. Can you tell me a little bit about how you expand on this thesis about looking at different types of environments that drive different flavors of collective behavior?

Deborah Gordon:

So the book lays out some hypotheses about how collective behavior evolves to fit the way that environments change, the stability of the environment, how likely there is to be a change in some situation, how risky it is, and also the cost of operating in that environment. So in some environments, it takes a lot of work to get whatever the collective behavior is obtaining, and the distribution of resources in the environment, whether the system has to respond to events that could happen anytime anywhere or occur in bursts. So those are different kinds of environmental conditions of the rate at which collective behavior operates, the kind of feedback that it uses, and the modularity, the way that interactions are structured are all likely to correspond to the way that the environment changes.

Nicholas Weiler:

How interesting. Yeah. And of course this show is about neuroscience. And so as you're talking about these things, I just keep seeing parallels again and again. But it feels like a good reminder to neuroscientists that the brain does many things. The brain works in many ways in response to... There's not just one answer about how these systems work. Was there anything that particularly surprised you that you learned as you were researching and developing this book?

Deborah Gordon:

I think that there is a lot of interesting work going on now about the modularity of the connections that are associated with different kinds of brain activity. Perception, for example, in the visual system tends to be very located in particular functional areas of the brain, and it operates quite quickly. Whereas cognitive activity tends to be much more diffuse. And that probably reflects how quickly a system can respond, for example, to visual stimuli. I guess a simple way to say it is that we see faster than we think.

Nicholas Weiler:

Yeah. Our visual systems are like the opportunistic ants in our kitchen, snatching up any light rays that happen to come through.

Deborah Gordon:

The visual system has to be fast. In fact, the part of the system that deals with keeping an image fixed in your perception when the head turns has to be incredibly quick to compensate for the speed of movement of your head. So that's an example of part of the brain that has to work extremely quickly.

Nicholas Weiler:

This is something that you brought up a little bit earlier that I wanted to come back to. In neuroscience, there's this philosophical conundrum about the ghost in the machine, people call it sometimes. We have the feeling that we are at the controls in our brains, that there's a unitary central control of our actions. But it's not really true, right? Experience and our behavior is the product of the collective action of billions of nerve cells, in much the same way that the behavior of an ant colony is the product of millions, hundreds of millions, billions of ants, I assume, depending on the species. There was an article you wrote a couple of years back about how a colony of ants over many generations can know things. It can have memories that no individual ant contains. Which strikes me to being similar to our own experiences being the owners and operators of a brain.

Deborah Gordon:

Well, that's the question. Are we really the operators of our brains?

Nicholas Weiler:

Right. Well, I'd love to hear just a couple more examples to help drive home the many ways in which different ant species use relatively simple algorithms, interactions between different ants, to produce what we might call intelligent behavior at the colony level. What are some of the examples that you think showcase the emergence of intelligent behavior at the collective level?

Deborah Gordon:

One way to compare different kinds of collective behavior that's associated with how quickly the environment changes is to think about two kinds of ants that I study. One is harvester ants in the desert who eat seeds that are scattered around in the desert and don't change very quickly. And another species of arboreal ants in the canopy of the tropical forest in Mexico called turtle ants. And they eat resources that are very ephemeral, like nectar on flowers. It shows up for a day, and then it's gone. The harvester ants have a kind of centralized system where ants go out of the nest, search around, and it's not until an ant comes back with food that it helps to stimulate another ant to go out, but there's no spatial information. Ants come in. The rate at which they come in from anywhere stimulates ants to go out. An ant coming in from one place can stimulate an ant to go out someplace else.

Nicholas Weiler:

And the logic of the ants there is if a lot of forages are coming back with food, there must be more food.

Deborah Gordon:

That's right. Because on a day when there's been a lot of wind, and so a lot of seeds have blown in, there are seeds everywhere. On the day after it's rained, and some of the top layer of the soil has been washed up, then they're more seeds everywhere. So if ants are finding seeds in one direction, ants are likely to find seeds in another, and there's no need for any local information. And so the only information exchanged is back at the central nest.

For the turtle ants who are moving around in complicated vegetation and are looking for ephemeral resources, there's a lot of spatial information. So an ant doesn't cross a particular junction where one vine crosses another in the tangle of vegetation. An ant doesn't cross a particular junction unless it's been reinforced by pheromone from another ant. Their search is very local and very modular, which actually corresponds to the distribution of the resources. Because if there's a flower growing on this stem, there's likely to be another flower on a stem nearby. So they are foraging in trees which have a big trunk, and they have all the flowers up at the top, at the edge of the canopy.

The ants that are looking for ephemeral resources in a very complicated spatial environment are using a lot of local connections. And the distribution of their connections is in little modules or clusters. So that's a contrast between the distribution of resources and the speed at which they have to operate, and it leads to a very different structure of connections among the ants.

Nicholas Weiler:

As you were describing, the system of the turtle ants not crossing junctions between vines unless they're reinforced, it made me wonder, we've been talking about individual ants as sort of interchangeable components in the complex system that is the colony. I know that in some ant species, there is specialization where you've got different types of ants that have different functions. But it made me wonder if there's also variation in the ants, so that some ants are a little bit more bold and they're more likely to be the first ant out. Maybe they are less constrained by the crossing the vine pheromone rule because some ant has to find the next flower. Do you see that ants have different behaviors in terms of how constrained they are by some of these algorithms?

Deborah Gordon:

In any ant colony, there are ants performing a certain task at any time, but ants can switch tasks from one to another. So they're flexible in the allocation of ants to tasks. So an ant could be a nest worker today. And if more ants are needed to do foraging, it could become a forager tomorrow. Then there is variation among ants in how they do those tasks. I haven't noticed that some turtle answer more likely to explore than others. It seems to be mostly constrained by how much pheromone there is. So if an ant comes up to a junction and there isn't enough pheromone to tell it which way to go, then it starts to wander around from node to node until maybe it finds a direction with more pheromone.

There probably are differences among ants in how they respond to the same local cues. If you watch ants at all, you end up wanting to help them because it seems that the ant that you're watching just can't seem to get it together to do the obvious thing that you think it should be doing. There's a lot of just bumbling around when you look closely at ants. But in the aggregate, somehow they get a lot of things done, and they're very successful as a group, although individually don't look very competent.

Nicholas Weiler:

I see reflections of that in human society as well from time to time.

Deborah Gordon:

Yes.

Nicholas Weiler:

It's remarkable that we get anything done. The last thing I wanted to ask you about, you had a TED talk back in 2014 where you made a provocative comparison, not only between ant colonies in the brain, which we've discussed, but also the way that cancer cells operate in the body. These are all examples of collective behavior, lots of individuals acting together without centralized authority. Is there something deeper that you feel unites these different examples? Or are they just sort of useful analogies to help us understand what's going on in an ant colony?

Deborah Gordon:

I think that we see a lot of parallel trends in different natural systems for the ways that interactions are used to generate collective behavior in particular environments. There's been a lot of recent work in cancer showing how important the local microenvironment is for predicting the behavior of cancer cells and their evolution.

Nicholas Weiler:

And so the cancer cells are also not very, you wouldn't call a cancer cell intelligent, but together the cancer cells are responding to their environment and interacting with each other to help them grow at the expense of the host.

Deborah Gordon:

Yes, cancer is a form of collective behavior.

Nicholas Weiler:

And then the final thing I want to ask is where do collective behaviors come from? But I realize that's a huge question.

Deborah Gordon:

Well, collective behavior evolves through the evolution of the local interactions among participants. Natural selection shapes how neurons interact with each other or how ants interact with each other because the outcome has a function. Natural selection shapes how neurons interact because how brains function matters for the survival of the organism.

Nicholas Weiler:

Well, that's so fascinating. I love thinking about this, actually, thinking about this has made me enjoy watching ants even when they're invading my house, just because, as you say, they are so fascinating and how they're getting things done is remarkable. Well, thank you so much for coming on the show. It's been a pleasure talking with you.

Deborah Gordon:

Thanks very much.

Nicholas Weiler:

Thanks again to our guest, Debra Gordon. We'll link to more information about her work in the show notes. 

We're very excited for season two of from our neurons to yours. If you are too, please take a moment to give us a review on your podcast app of choice and share this episode with your friends. That's how we grow as a show and bring the stories of the frontiers of neuroscience to a wider audience. 

This episode was produced by Michael Osborne with assistance from Morgan Honaker. I'm your host, Nicholas Weiler. See you next time.