First complete map of brain activity; Queen ant lays eggs of another species; The perils of scrolling while on the toilet

0:28This episode is brought to you by Accenture. Welcome to the World, the Universe and Us, the weekly news podcast from New Scientist. I'm Dr Rowan Hooper. And I'm Dr Penny Sarchet. On today's show we hear about an ant species where the queen ant lays eggs of another species. And we're also going to hear about the perils of using your smartphone on the toilet. And it's not that you might drop it in the can. It's not that particular peril, no. OK, I'm looking forward to that. We're going to start with news that scientists have made the first complete activity map of a mammalian brain. Yes, this is a big study.

1:00It took place across a dozen labs and it has revealed unprecedented insights into how decisions are made. The big takeaway from this study is that they found that decision making is a process that's distributed across the brain rather than in a particular area, which is what they were expecting. And then also this study may even hint at the roots of intuition, that feeling that you know something. And there's a really interesting sort of element here about what that might then mean for developing artificial consciousness. Alexandra Thompson is here. Alex, first of all, we should say this is a mouse, not a human.

1:34But didn't we talk really recently about the first map of a mouse brain? Yes. So in April, we reported on the largest mammalian brain map ever. But actually, that only captured the neuronal activity of one cubic millimeter of a mouse's visual cortex, which felt like a really big deal a few months ago. But now we've made quite the leap. And now we have a map of the activity of the neurons in an entire mouse's brain. OK, so let's get into that activity. What are we talking about? Well, neuroscientists have long wanted to capture the activity, which is how the different neuronal connections fire and sort of interact with each other.

2:11But doing that across the whole brain is really challenging at the level of individual neurons. Because we often talk about X part of the brain does Y function. And to some extent, that is the case. We see after somebody has a brain injury, like a stroke, their speech or movement might be affected and not much else. So we say, well, it's because part of the brain that affects speech is affected. I always think about that poor soul. He was a railway worker in the US in the 1800s. You probably recognize him from the textbooks. He had an iron rod go up and then out the other side of his brain. And his personality really changed, but his memory and otherwise cognitively was OK.

2:47So then we made all these leaps about very specific regions of the brain with specific functions. But when it comes to sort of complex behaviours, decision making, reward seeking, neuroscience have always suspected it's probably a bit more of a kind of dance going on within the brain than that. So it's not an easy thing to prove. There's a limit to how many neurons an electrode can record from and how many electrodes can be implanted in a single brain. So to overcome this, researchers across 12 labs have joined forces and they each ran the same experiments, but they recorded from slightly different areas of a mouse's brain with some overlap to make sure they were getting consistent data.

3:26And putting it all together, for more than 650 ,000 neurons, they produced the first brain-wide activity map of a complex behaviour in a mammal. Wow, so they've recorded activity in detail from all these different individual areas as they split it up across the labs, and then they've brought all of that together. I want to know what was the complex behaviour that they were making the mouse do when they were having a look at this? So at each lab, mice were trained to move a stripy target towards the middle of a screen and they moved it using a tiny Lego steering wheel. So they were driving, it was a Lego driving school.

4:04Exactly, yes. That's not what I would have thought when they said it's a complex behaviour, it's a Lego car. A maze or finding some food in a puzzle box or something. And the Lego steering wheel was just right for a mouse, isn't it? Yeah. Like that tiny little wheel. Yes. So when the stripes were high contrast, the target stood out clearly. But when the contrast diminished, the target was barely visible and the mice had to rely on their prior knowledge to answer correctly in order to receive a reward. So biases were built into the experiment to allow them to build these prior expectations. So the target might appear on the left 80 % of the time and on the right 20 % of the time.

4:42and then those biases were flipped to vice versa and the mice had to adapt their expectations accordingly. So this is really clever, isn't it? Because it allows the scientists to see what's happening in the brain when the mouse is deciding what to do, when maybe it's harder and they can't really see. And because it's in the same scenario, repeated many times and recorded in all different parts of the brain, they get an overview of that whole decision-making process. Yeah, and put it together to create this map. So the resulting map revealed that activity associated with decision-making isn't confined to a single region, but it is distributed across the brain.

5:19So here's Alexandre Pouget at the University of Geneva to tell you more. OK, there are really two major findings in our study. The first one is that decisions are not made by a few specialized areas in your cortex, as one would have been tempted to conclude from previous studies. Instead, by looking at the whole brain, we see that decisions emerge as a result of communication within a vast network of brain regions, both cortical or subcortical. And for subcortical, it's a bit surprising because, you know, decisions are often associated with cognition, which is supposed to be more in cortex. But in any case, what we see that processing in the brain is only very loosely hierarchical.

6:03It's instead extremely distributed. And the second major finding is about where expectations are included in the brain. In our experiments, animals have to decide whether a stimulus, a visual stimulus, appears on the right or on the left side of the screen. Importantly, the stimulus does not appear half the time on the right, half the time on the left. Instead, on a block of trials, it would be 80 % of the time on the right. And then on the next block of trials, it would be 80 % on the left. And what we found is that animals can use this information to guide their decisions and to improve their performance.

6:38Then the big question for us was, where is this information encoded? And what we found is that it is encoded all over the brain, even in very early sensory areas. In other words, we don't process sensory information in isolation. We do so in the context of what we expect. And we know from computational theories that it's a very powerful form of computation, but also one that's tricky. It's tricky because if you don't do the computation well, it's easy to end up trusting your expectation too much and to effectively take your dream for reality. In cognition, this is known as a confirmation bias, and we know that indeed in some situations, humans are indeed subject to such biases.

7:23so don't take your dream for reality what he's saying there good advice i mean that is where they're getting to the the roots of intuition that you mentioned earlier um when he says that the mouse can use prior information to try and gauge what's happening yeah that's it and and data previously collected does support that decision related signals build long before an action is taken. So Puget says that even before the experiment began, when the mice were just still, already signals associated with the next decision were apparent. And that build as the target appeared on the screen, reached a threshold, and then it moved to the steering wheel.

8:03It's a little bit speculative, but he even says this might correspond with what we call intuition. Wow. It makes me wonder if there's some mice like Lewis Hamilton mice, who are more intuitive, can take the corners better and, you know, much faster drivers, better drivers. I think they were just bog standard lab mice. I did see a similar study with rats that had been put in a sort of, it wasn't a Lego car, but it was a car that they could drive. And sometimes as the researchers sort of got the car into position, the rat would start to rev the engine of the car, anticipating that it was going to land and they could drive off to this food item they had to collect.

8:40So that's not quite intuition, but it is definitely anticipation. So the team behind this work suggests that their results and also their collaborative approach, the way they did it, could advance understanding of conditions like autism and how the brain works there. Mouse models of autism suggest these animals have difficulties updating their prior expectations when new information comes along. And that's kind of what happened in this experiment. Right. And I asked one of the team members of it, as you say, they're from around the world. And this is Anne Churchland. She's professor of neurobiology at the University of California, Los Angeles.

9:15Our hope is that even though mice and humans are different in a lot of ways, that there are commonalities amongst human and mouse circuits that mean that our work studying the mouse brain will be relevant to understand conditions like schizophrenia and autism in the human brain. And fortunately, there are mouse models of some of these disorders. For example, in my lab, we are studying norexin 1-alpha model of autism spectrum disorder. And by understanding how that genetic change alters neural circuits, we hope that we will gain insight into the changes that occur with these autism-related genes that we hope will one day inform treatments of autism spectrum disorder.

9:54And I wondered what this work might say about consciousness, given that it shows distributed activity pattern across the brain. Yeah, I wondered that exactly. And it made me think of different theories of consciousness and how they try to work out what's happening. And, you know, it's a big collaboration. Alex Pouget made the good point that consciousness isn't well defined mathematically. That's what always bothers me. It's really it's a philosophical problem too much. So it's very hard for them to say things about it. But I also spoke with Matteo Carandini. He's professor of visual neuroscience at University College London.

10:30And here's what he said. This is really interesting. The word consciousness is used in at least three ways with three different meanings. Philosophers use them in terms of how does feel to feel? How does it feel to see red? And that's I think they use the word qualia for that. and anesthesiologists on the other side say is the patient conscious or unconscious so if you can perceive things you're conscious and if you're out you're unconscious and i guess maybe an intermediate definition of consciousness is if you see something do you know that you saw it if you see that the light is red do you know that it's red and therefore you stop and one one way to measure that is to see if people stop at the red light right and um and this is essentially what we're asking our mice to do we ask them is something on the left or on the right do you see it on the left or do you see it on the right and if they indicate that they see them we know that that's their perception so perception is a form of consciousness you're you're you perceive something and it's some people think that's that's a definition of being conscious of something is if you know you're seeing it.

11:41So that's exactly what our study is about. We ask under what conditions mice see something and tell us that they see it, which indicates they're conscious of seeing it, and under which conditions they don't see it or they make mistakes. So that's the basis of our study and that's the topic of our study, but it's a kind of an intermediate concept of consciousness. The more philosophical one, how does it feel to see something, which I don't think we have a handle on. And the more practical one that anaesthesiologists use, is the patient conscious or not? So we are getting at an aspect of consciousness here, an intermediate concept of consciousness.

12:22Intermediate concept, yeah. And I think that's what's so encouraging, is that, as I say, for so long, it's been a philosophical dilemma, a problem for people to wrestle with. But this is an experimental way of getting at this problem. And it's very hard. Often it's been hard to find experimental ways. So might this tell us something about generating artificial consciousness? Yeah, well, so I also wondered that. And I asked Anne Churchland also about this. And then about how this work on the brain that she's been doing make decisions that could influence how we think or how we make LLMs, large language models, or AI more generally, that has some sort of conscious attributes.

13:03And here she is on that. One insight that my hunch is that we might gain might have to do with the relationship between cognition and movements. So AI systems, for the most part, don't move around the way that humans do. And much of our neural circuitry originally evolved to help us move around in the world, to gather food, to find mates, and so on. My hypothesis is that a lot of the circuits for higher cognitive functions like decision making ultimately hijacked some of the circuits that were used for movement. And that might relate to why there are so many movement signatures in decision making areas.

13:38So I don't know that the answer will be to make LLMs and AI systems that move around. But I do think that by understanding how circuits that were originally evolved for movement ultimately turned into circuits for cognition could be really interesting for helping us to develop better AI systems in the future. And that's such an interesting point, isn't it? That systems that evolved for movement and like simple decisions, making simple decisions, they may have ended up being co-opted in higher mammals into these higher levels of consciousness. It's evolution for you. Time for a short break to talk about New Scientist Live, Penny.

14:13Yay! Yes, that's coming up in October. It's the world's greatest festival of science here in London. It's on October 18th to 20th at the Excel Centre and it's streaming online. What are you looking forward to, Penny? Very much the live recording of this podcast. Of us? Yeah. You're looking forward to us doing our thing. I'm looking forward to that. I'm also chatting with Chris Packham live on stage. What a legend. I'm really looking forward to that. We're hearing about his favourite animals. Lots of cool stuff. I'm also looking forward to hearing from gut health expert Megan Rossi talking about prebiotics and all that kind of thing.

14:45There's loads of other fascinating talks as well. Search New Scientist Live and get your tickets now. Okay, how's this for weird? Entomologists have discovered that there is a species of ant where the queen lays eggs of a species that is not her own. That is weird. I'll give you that. It's pretty damn weird. So some of the eggs laid by Iberian harvester ant queens contain males of another species, the builder harvester ant, so Iberian and builder. And these males that she has, they go on to father all the workers of her colony. I mean, that is nuts even for ants. Yeah, exactly. We're just saying something.

15:22Jonathan Remigier at the University of Montpellier says, this statement sounds really, really crazy, like impossible. And yet he discovered it's true. And Sam Wong is here to explain this to us. Sam, what is going on? So Remigier was studying these Iberian harvester ants and he noticed that all of the workers in the colony were hybrids. So half of their DNA came from a different species, the builder-harvester ant. I'm going to call them by their Latin names, Ibericus and Structor. So the most obvious explanation is that Ibericus queens are mating with Structor males. And this kind of thing does happen in some ant species.

15:57No one's quite sure why, but there are two sort of possible explanations. One is that it's to do with hybrid vigor. So the idea that sometimes hybrids of two species can be fitter than other species. So I think there's some examples with like dogs, hybrids between two dog breeds. are. They live longer than the other ones. And the other possibility is that it might resolve this problem that Ibericus has and a few other ant species, where Ibericus queens, when they mate with Ibericus males, all of the offspring become queens. So that's due to a genetic work that's going to ensure that it propagates itself, but it's bad for the colony because they need to have workers.

16:31So maybe breeding with another species is a way for them to create workers. So what was the study here? What were the researchers doing? So they noticed that some of the males in the colony, they looked like Structor males. They didn't look like the typical Ibericus males, which are hairy and the Structor ones are less hairy. And that's particularly strange because these Ibericus colonies occur all over the Mediterranean, even in places where there are no Structor colonies. So where are these Structor males coming from? They tested the DNA of these males and they did confirm that most of their DNA does match the Structor ant species.

17:07but they found that the mitochondrial DNA which is this tiny bit of DNA in the cell that always comes from the mother that was a match for Ibericus so it seemed like these structural males are being produced by Ibericus queens so to test this they brought some colonies into the lab and this is quite a Herculean effort because when they bring colonies into the lab for some reason they just hardly ever produce males so they ended up bringing 50 colonies into the lab and monitoring them for two years until they finally managed to get males being born. 50 ant colonies. That's a lot of ants. Can you imagine the logistics?

17:42Okay, so they made this heroic experiment, 50 ant colonies. But when males of the other species were born in the lab, then what was going on? Was there sperm storage there? Because we know that ants and all social insect queens can store sperm for years. So was that what's going on? but this is sperm from another species. Yes, so having these colonies in the lab did allow them to confirm that these Ibericus queens were producing eggs that turned into structural males. And ants can do this because, as you say, the queens, they can store sperm in this organ called the spermatheca. So what must have happened is that the queens had mated with a structural male, stored his sperm, but they're producing two different kinds of males.

18:29They're producing Ibericus males as well as the structural males. and then some of their offspring are hybrids between the two. So the hybrids are the workers and then they're producing structural males as clones and those males are for her daughters to mate with. So they're clones because, like you said, the maternal DNA is only the mitochondrial. So that means the egg cell was from the queen but there was none of the usual sexual shuffling of chromosomes. They literally just inserted the entire male sperm DNA into this egg. Like she's cloning this other species. Yes, that's right. So she has mated with a structural male.

19:08She's cloning that. She's producing clones of the structural male. And then her offspring that become queens, they are mating with these other structural males, which are their brothers from another species. Absolutely mad. So I wanted to talk to a specialist about this. So I reached out to Jessica Purcell. and she's a professor of entomology at the University of California, Riverside. And I did get hold of her, but she wasn't there. She was in Alaska on an ant sampling trip. Ants in Alaska. Oh, yeah. Ants are everywhere. But she still managed to send this comment. This was a pretty astonishing discovery, but if it was going to happen in any species, I'm not surprised that it was found in an ant.

19:52The reason I wasn't surprised is because there's a couple of other really weird and wonderful ant mating systems. that kind of seem like they're precursors for this type of system. The first one is called sperm parasitism, and this is where ant queens will mate with males of a different species in order to produce workers. So the workers of these species are always hybrids between the maternal species and the species that was mated with just to produce workers. And the reason that this is called sperm parasitism is because workers don't generally reproduce, so this is a dead end for the male genotype.

20:26They're just in the workers, and then they don't have offspring. The other system that is sort of combined into this newly discovered system is a case where males are actually cloned by females. So what happens is the male and female will mate. That mating in these several species, which include electric ants and longhorn crazy ants, the mating will produce workers from fertilized eggs. But then in some cases, the sperm will fertilize an egg and then the maternal complement in the nucleus, so the maternal DNA in the nucleus, will be eliminated and that egg will develop into a cloned male. So it'll be an exact copy of the original male, aside from inheriting the maternal mitochondrial DNA, which is outside of the nucleus.

21:13And meanwhile, in those systems, the females also clone themselves, or at least some of the systems, the females also clone themselves so you end up with a system where the queens and the males are clones and as a result they can meet with each other without inbreeding because they're from distinct lineages even though the queen produces both daughters and sons so that is nuts isn't it um queens producing all these different different species and the males are clones um and the authors call this new system xenoparity. Xeno meaning alien and paris meaning birth. So giving birth to aliens or giving birth to the other more correctly.

21:53And here's Jessica again. In the xenoparis system, the queens are giving birth both to or laying eggs that will develop into males of her own species. So they inherit the nuclear genome from the mother. When they mate with those males of their own species, the fertilized eggs become future queens. They're also mating with these males that were originally derived from a different species. And in those cases, for fertilized eggs that retain both the maternal and the paternal copies of the genome, those will become workers. And then some of the eggs are purging the maternal complement of the nuclear genome, and those will become males of this other species.

22:33So that's why the authors called this Xenoparis. These queens are able to give birth lay eggs that will both be their own species and males of a different species. Now, I think the take home message from that is that apart from ants are even more nuts than we thought, is that reproductive strategies, they are more diverse than we can imagine because no one expected this. And aren't ants brilliant? Do you have a favourite ant species? So I'm ashamed to say, I don't particularly, but I did. You don't have to be ashamed to say. Well, you know, I love to nerd out about animals, but I did once ask the late, great E.O.

23:06Wilson, ant legend, what his favourite species was. And he told me it was the Thormatomyrmex. And that's a rare kind of ant that has pitchfork-like jaw appendages that they wondered for ages why. And it turned out it's for eating very spiky millipedes. So there you go. That's his favourite species. That's what he said, anyway. E.O. Wilson's favourite ant species. Sam, have you got one? Yeah, I'm going to say weaver ants. We did a story on them recently. They're one of the most effective teams in the natural world. So usually when you have a team, each individual puts in less effort and they're not very efficient at working together.

23:41But these ants, they roll up leaves to build their nests. And individually, they're pretty strong. They can pull 60 times their body weight. But when you have a group of them working as a team, then each one is pulling 100 times its body weight. Wow. That's pretty amazing. My favourite is the leafcutter ant. Well, many species of that, but they're brilliant. and to see them stripping down a bush or a tree and feeding the leaves to a fungus that they're growing underground. The first farmers. Yeah, ant farmers. Amazing, but very hard. Hard to make a call. Shall we open this up and call for listeners and viewers?

24:15I think that's a great idea. Tell us your favourite ant, please, everybody. Do comment. Favourite ant species, we want to know. Now, there's been a lot of debate recently about smartphones, particularly their effects on mental health in young people. But a study out this week found a link between smartphones and a health issue that might surprise you, and that's hemorrhoids. Michael, you reported on this for us. Of course you did. I think I was sitting next to you when you first heard about this story. How on earth do smartphones increase the risk of hemorrhoids or piles, as they're sometimes saying?

24:50Yeah, it's a bit of an unexpected one, isn't it? So it's all about how long we spend sitting on a toilet. so this small study looked at people who were taking their smartphones into the loo or not and they they basically found that if you take your smartphone into the loo you spend a lot more time sitting on the toilet and that translates in in their view into a 50 percent increased risk of getting hemorrhoids it's a pretty big increase in the risk yeah i always thought it was it was straining that was the day not sitting on the can for ages but straining was the thing you shouldn't Yeah, and when I looked around, I found loads of sources, even the NHS, they all say it's straining.

25:26But this study, anyway, this one study, it found absolutely no link between straining and the risk of hemorrhoids. And some other studies have actually linked hemorrhoids with diarrhea rather than constipation, which Jorster suggests it's not really to do with straining. I mean, basically, it turns out we just don't know that much about hemorrhoids and what causes them. and there's very little evidence and a lot of the studies were done a long time ago and weren't necessarily very good. And that's what inspired this latest study. So gastroenterologist Trisha Patricia said she sort of was sort of always told her patients, you know, don't spend too long on the loo.