Ep 146: Neuroscience, naturally (with Nachum Ulanovsky)

How can neuroscience be  more naturalistic ? Will lessons from behavioral ecology and evolution have value?

In this episode, we talk with Nachum Ulanovsky, Professor of Neuroscience and Head of the Zuckerman Center for Learning, Memory & Cognition at the Weizmann Institute of Science and author of the book Natural Neuroscience: Toward a Systems Neuroscience of Natural Behaviors. We talk with Nachum about his book, which makes the case that if we want to understand the brain and its effects on behavior, we have to do better about studying it ina natural environment. He also talks about new technologies and cross-disciplinary perspectives that could provide better insight into basic brain function but also new ideas for how to treat or manage neurobehavioral diseases.

Cover art by Brianna Longo

Check out substack for bonus content!

  • Marty Martin  0:00  

    Hey, Cam, I wanted to talk to you today about an obscure topic, something we've never touched before on the show.


    Cameron Ghalambor  0:06  

    Oh, really, what's that?


    Marty Martin  0:08  

    Plasticity


    Cameron Ghalambor  0:09  

    Oh, come on, that's probably the most common thing we talk about, even when we're not recording for the podcast. 


    Marty Martin  0:09  

    Oh, yeah. Okay, fine. You got me. But I do want to approach the idea from a new place.


    Cameron Ghalambor  0:20  

    Okay, go.


    Marty Martin  0:21  

    I want to talk about non adaptive plasticity.


    Cameron Ghalambor  0:24  

    Alright, now we're talking. It's not a form of plasticity we've discussed much, but it is always worth considering. When is plasticity adaptive and when is it not? What makes you interested in talking about it right now? 


    Marty Martin  0:38  

    Well, I think it's a perfect frame for a chat today with Nachum Ulanovsky, a professor of neuroscience at the Weizmann Institute in Israel. Nachum's new book Natural Neuroscience is an appeal to the field, and fans of neuroscience and behavior broadly, to do neuroscience in more natural conditions. 


    Cameron Ghalambor  0:54  

    Yeah, so even though I'm not a neuroscientist, I really enjoy the book and how he contrasted traditional approaches to classical problems like how bats navigate in space or how rodents process and integrate sensory information with modern approaches that involve more ecological realism. Listeners will hear about this more in a minute. But what do all these things have to do with non-adaptive plasticity?


    Marty Martin  1:19  

    Well, although I don't think he mentioned maladaptive or non-adaptive plasticity in the book, I came to feel that what he was saying was that laboratory biologists might largely have been studying non-adaptive plasticity, or what one could call hopeful monsters.


    Cameron Ghalambor  1:36  

    Okay, wait a second. Hopeful monsters is a totally unrelated term here. A hopeful monster, was actually coined by Richard Goldschmidt to basically look at the effects of rare mutations of large effect, kind of in a context of like an X- man mutant, where a new organismal variant, by chance is beneficial. So it's a hopeful monster, and it goes on to succeed and thrive. Goldschmidt offered the counter to that as the hopeless monster, which he thought would obviously be more common. Big mutations are always apt to be more problematic and more deleterious than beneficial. So if I'm going to continue with a movie analogy here, think about Jeff Goldblum's character in The Fly, not exactly a beneficial phenotype in the end.


    Marty Martin  2:23  

    Okay, yeah, and I guess that's not the hopeful monster I mean. In the case of lab neuroscience, immunology, physiology and maybe medicine generally, I think we might have been too hopeful that the beasts we study in controlled conditions are representative enough of the things we want to understand. For instance, it's always been assumed that controlling the environment or genetic variation in our study organisms is the way to go.


    Cameron Ghalambor  2:49  

    Well, yeah. I mean, shouldn't it be? How can you do an experiment if you don't control for some aspects of variation? To contrast, say, your treatment group to your control group, if you let everything be sort of variable, you would be really hard to make any inferences. Isn't that the bedrock of the scientific method?


    Marty Martin  3:07  

    Well, yeah, sure, but that assumes that the context in which the experiment is done is itself representative of the conditions that the system would have experienced.


    Cameron Ghalambor  3:17  

    Okay, you're losing me here. Give me an example.


    Marty Martin  3:19  

    Okay, sure. Let's do one with your guppies. Let's say you wanted to know what role one brain region played in guppy choice of mates. It'd be very hard to study that in the wild, as you might never catch the individuals twice to which you applied some treatment, maybe a drug to block a receptor or light up the brain region once it was activated by an interaction with a member of the other sex?


    Cameron Ghalambor  3:38  

    Right so that's why we do these kinds of studies in laboratory aquariums, and that way we can be sure to capture all of the animals while also keeping food, climate, photo period and everything else that might affect the fish constant.


    Marty Martin  3:51  

    Well, what happens to the phenotypes we measure when we don't conduct our studies in the context that resemble where those systems evolved? Maybe control is sometimes the wrong frame. Perhaps we can't just standardize conditions, perhaps we have to standardize but also do our comparisons in natural-ish conditions. So take lab mice. We typically feed them all they want to eat of a nutritious, but not necessarily naturally representative diet, and then we house them in social groups more suitable to our working with them than the preferences they'd make in the wild. All the while, we keep them in the temperatures that are comfortable for us but colder than their preferences. We also never let them be exposed to predators, and we go to great lengths to make sure that they're never infected by anything, by keeping their cages way more sterile than their nests would ever be.


    Cameron Ghalambor  4:38  

    Right. That's also called good animal husbandry. 


    Marty Martin  4:40  

    Well the point is almost the same as Nachum about neuroscience and nature, if we study our model organisms and conditions that are well controlled and seem humane or convenient for us, we could be studying a different kind of hopeful monster, in the sense that the traits that our lab mice manifest are actually examples of non-adaptive plasticity, trait forms that are released in environments  different from the ones the genotypes involved in. In nature, responses to natural forms of food, light, temperature, social cues, those would all produce very different phenotypes that are released by lab conditions. So to be blunt and perhaps overstated, we might have spent a lot of money and time studying complex phenotypes that would never exist because they never evolved as such in nature.


    Cameron Ghalambor  5:23  

    Yeah, no, I think that's a good point, but it does presume a few things. First, perhaps natural and lab environments aren't really that different. And second, perhaps some traits are not as plastic as you're making them out to be. Or, more likely, there's a lot of genotype by environment interactions where lab adapted mouse genotypes will respond to the lab environment in one way, and that might differ from wild-type genotypes in their natural environment.


    Marty Martin  5:49  

    And that's fair, but I wonder if it holds for all traits. Wouldn't it be great if we tested it, or at least addressed this potential problem? Just a few years back, NIH mandated that we study male and female rodents, because sexes differ so much. Wouldn't it also make sense then that we tried to study living systems in contexts more similar to the ones in which they evolved? To me, the issue is that most biological processes entail extremely complicated regulation, so in this light, it probably doesn't take major changes in resources, stress or the like, to change the phenotypic outcomes. My money is on that we're often studying non-adaptive, plastic responses, not the evolved and healthy variation that we hope.


    Cameron Ghalambor  6:27  

    Okay, yeah, I agree with you, and I think what you're saying is very similar to what Nachum is saying in his Natural Neuroscience book, that you know, we have to be more naturalistic in the lab. And I think that's a fair point, although I am skeptical that this will be embraced by the biological community, because these kinds of experiments are really hard to do.


    Marty Martin  6:47  

    Yeah, you're probably right, but I can't hope. And with books like Natural Neuroscience, maybe we'll start to see changes like that in the future.


    Cameron Ghalambor  6:54  

    Let's hope so.


    Marty Martin  6:57  

    Before we get started today with knock 'em a few things first, if you are in Knoxville, Tennessee on April 16, come and see Cam and me interview Dan Simberloff about his new book on invasive species at The Laurel Theater.


    Cameron Ghalambor  7:08  

    Admission is free, and you'll even have a chance to ask Dan questions about his work, which could appear on the podcast in the form of a conversation. Write to us at info at big biology dot org if you need more info.


    Marty Martin  7:21  

    And finally, don't forget to subscribe to Big Biology at big biology dot substack dot com. Molly Caroline, Brianna Clayton, Steve and new intern Cass rely on your support to keep the show coming.


    Cameron Ghalambor  7:34  

    We offer several subscription tiers, from monthly to yearly to group subscriptions. We really hope PIs department heads, school teachers and others can purchase these subscriptions and share them with their students and with other up and coming biologists. 


    Marty Martin  7:50  

    Sign up for a subscription and share access with your trainees, families and friends. And now on to the show. 


    Cameron Ghalambor  8:02  

    I'm Cameron Ghalambor.


    Marty Martin  8:03  

    I'm Marty Martin.


    Cameron Ghalambor  8:05  

    And you're listening to Big Biology.


    Marty Martin  8:19  

    Nachum Ulanovsky thank you so much for joining us on Big Biology today. 


    Nachum Ulanovsky  8:23  

    Thank you. 


    Marty Martin  8:23  

    So we invited you on to talk about your new book, Natural Neuroscience, but before we get to the contents of the book, tell us its origin story. How did you come to write this book in particular?


    Nachum Ulanovsky  8:33  

    So I'm a neuroscientist studying the brain, and I was thinking for a long time about the importance of studying the brain closer to a natural behavior of animals and humans. And more than a decade ago, I was collecting all sorts of ideas and thoughts that, you know, one day I might write a book about this, but this was like, you know, when I retire, it was not a concrete plan. And then I gave a talk at MIT Press about a decade ago, and then a few weeks later, I get an email from an editor at MIT Press saying, you know, I've seen your talk. It was very interesting, and in particular, caught my eye, this notion of natural neuroscience. Did you ever thought to write a book about it? So I told him, Well, actually I did. But you know, from here to there, it became a real project, rather than one day I will do it. 


    Marty Martin  9:23  

    Okay. You could have chosen lots of different topics, and writing a book is no small tax. So what is it about natural neuroscience that was so compelling that you made the decision, okay, let's go for it?


    Nachum Ulanovsky  9:35  

    So I think it's really the passion. Having worked on it for a number of years, I now know that to take , to write a book, is a lot of work. Is a lot more work than writing a paper, or even a number of papers, both in length and in scope and in the overall, overall approach. And to get this through, you have to be passionate about it. And I think I'm. I'm really passionate. I was, and I still am very passionate. So I'm very passionate about the importance of studying the brain closer to the natural behavior of animals and humans. You know, in neuroscience, you know, I come from physics background.


    Marty Martin  10:16  

    Okay


    Nachum Ulanovsky  10:16  

    And so I've been trained in the importance of highly controlled experiments, separating factors one by one. But when it comes to complex systems like the brain,  behavior, I think it's important on one hand, this highly reductionist approach, but it also has its strong limitations, in particular in brain research. So I'm really passionate about the importance of going there.


    Marty Martin  10:38  

    Okay, good. And we're going to come back to those ideas in a second too. 


    Cameron Ghalambor  10:43  

    I found your book very interesting. And so I'm not a neuroscientist. I'm more of a evolutionary ecologist working with natural populations of animals. And so on one hand, I guess I found it surprising that this, there was this sort of need to start thinking about natural neuroscience, as opposed to, I guess, maybe the traditional way of doing neuroscience, which is very lab-based and very removed from what happens in, in, in nature and and so part of me, my first reaction was like, Well, yeah, why is this? Why is this controversial? Like, why? Doesn't everybody already think this way? But, but I, you know, as I read, I could see that this was not really the case. And so I guess my first question for you is, you know is who was your audience? Were you really targeting other neuroscientists as, like, the primary audience for this book?


    Nachum Ulanovsky  11:52  

    Yeah, yes, the book is really targeting primarily other neuroscientists. I really. The goal, from my perspective, is to convince other neuroscientists, be it, you know, senior neuroscientist or or students who just enter the field about the importance of studying the brain closer to the natural behavior of animals and humans. And your question captures very nicely the disconnect that exists between, you know, ecology, zoology, you know, all these fields where for a lot of a lot of people, it's like, obvious, of course, you need to study the, you know, natural behavior of animals. That's the obvious thing, because that comes from, like, a very different tradition, whereas neuroscience comes more from the tradition of either physiology or psychology. It's dominated historically by, you know, traditions of physiology and psychology, both of which are highly, highly reductionist sciences. And so a typical neuroscience experiment is not doing things outdoors or not even in complex, naturalistic settings indoors. It's highly, highly reductionist and controlled, which, again, has its powers. I totally appreciate this, and I talked about this in the book, but, but it is not enough. If we really want to understand how the brain operates in the real world and generates behavior and generates interactions with environment, generates interactions with other organisms. We need to to let go a little bit of the over controlled setups. This is exactly this difference between the fields of ecology on one hand and neuroscience on the other hand.


    Cameron Ghalambor  13:26  

    And so for our audience members who are not neuroscientists, and also for myself, like if I if I envision kind of traditional neuroscience research, I would imagine a controlled lab environment, perhaps, you know, with with mice or other sort of model organisms, and I might expose the organism, the animal, to some kind of a stimulus, perhaps. And I may quickly dissect out the brain and find a particular region of the brain that I'm looking at, you know, doing maybe single cell RNA to look at patterns of gene expression. Or I have another vision of an animal with a bunch of wires coming out of its head where it's exposed to a stimulus. And then you're looking at, you know, what regions of the brain get sort of lit up with, you know, neuron activity. How would you take that into the field, like that? I think that would be like, the first obvious question is, like, is that even possible like to move from these highly controlled lab situations to being able to do anything similar in the field?


    Nachum Ulanovsky  14:39  

    Okay, so this is a good point to make the point that even though in my own research, and we can talk about this a little bit later, we have been actually taking things to the field, recording neural activity in bats, in our case, that are flying and navigating outdoors on a remote oceanic island. So really doing electrophysiological, neurological recordings outdoors, that's like an extreme of a spectrum, but that's, I'm not claiming that everybody should be doing that, but even inside the lab, what you were just describing, it depends how you do it. So, for example, with the neural recordings, traditionally, let's say, until the year 2000 or so, most neuroscience experiments were done in anesthetized animals. You would present sensory stimuli, be it visual or auditory, like sounds or or images or not even images, like natural images, but oriented bars or dots, and you'll see how a neuron in the brain of an anesthetized animal responds. Or you'll present pure tones or broadband noises, sounds that are very different than the typical sounds in the real world, and see how neurons in the brain respond to those. And the animal was not doing nothing. It was anesthetized. The brain is in a very different state. 


    Nachum Ulanovsky  15:52  

    Then, you know, in the last 20 years or so, people start in, you know, going away from the anesthesia and letting animals be awake. But even then, a lot of the experiments to this day are in head-fixed animals. They cannot really move. And, yes, you allow them to do some behaviors, but these behaviors are highly restricted because the animal is head fixed. A lot of it is for technical reasons, because you want to do imaging of the brain with big microscopes that really require head fixation, just that you can do the recording or the manipulation, or whatever it is. But, you know, letting animals, even letting animals run around in a small box in the lab, that's already pretty natural compared to a lot of experiments. That's already not bad, and that's relatively recent, that's relatively recent. So I mean, doing outdoors, that's already the extreme of the extreme. I'm saying, even in the lab, doing allowing animals to move around, let alone interact with one other animal in the box. That's already how a lot more natural than a lot of experiments are done to this day in neuroscience. So that's why there's a big, big gap between what you imagine in ecology and what's actually done in most neuroscience labs. 


    Nachum Ulanovsky  16:58  

    Yeah, it's surprising.


    Marty Martin  17:01  

    So I want to put a fine point on this, because, like some listeners, I think could hear what we're talking about and say, Well, you know, obviously you can't probe the brain of animals in any deep, sophisticated way if they're out moving around the world. That's just maybe, you know, right now, science fiction, and might remain so for a long time. So there's a quote from early in the book, and I don't remember exactly what chapter I wanted to have you respond to. You said: "We may have been studying the neural mechanisms of the wrong behaviors". So I mean, maybe articulate exactly what you mean by wrong there, because it's, it's not just the sort of we should do this better, because it's not natural. But what does it mean? What do you mean when you say it might be wrong?


    Nachum Ulanovsky  17:44  

    Right, so this sentence comes in the context of, when I talk about the heavy emphasis on mechanisms. You know, in biology, in general, molecular biology, like modern biology and neuroscience, there's a strong emphasis on understanding mechanisms. So you want to dig deeper and deeper into the mechanisms. You want to look at the neurons and the synapses and the molecules, and very deeply into the mechanisms that generate, let's say, a behavior. But because doing this mechanistic to do this mechanistic studies, you often need to have highly simplified behaviors. You can repeat them over and over, or you need to head-fixate, because there's for certain methods. So this emphasis on mechanistic understanding, I agree with it. At the end, you know, it's a matter of, what is it understanding? What does it mean to understand the brain you want, the mechanistic understanding you want to dig deeper. But then the other hand, there's the question of, what is it about the brain that you're understanding? The idea, you know, we, at least from my perspective, I want to understand, and I think many neuroscientists want to understand how the brain generates behavior, like real world behavior in the real world, how the brain generates cognition, perception, understanding, all of these high cognitive functions. This is what drives me and many neuroscientists. 


    Nachum Ulanovsky  18:59  

    Now, of course, you reduce this in the lab to simplified situations. You don't do decision if you're interested in how do you do decision making? You don't study the brain of humans as they invest in the stock market and make decisions about what to sell and what to buy. But instead, you do some simple, two alternative forced choice experiment, where they do something a lot simpler, right? So I'm okay with simplifying. That's part of what we do in science, but, but if we simplify too much, then this might be too far removed from behavior in the real world, which is very context dependent, very rich, very complex. And then we may invest decades, we as a field, may invest decades in studying the detailed mechanisms of behavior that, at the end of the day, bear very little relevance to the actual behavior that interests us, which is the actual motivation of why we came to study the problem in the first place. So this is what I meant here.


    Cameron Ghalambor  19:56  

    So can you give a, do you have like a favorite example of, you know, a case where we might be studying the neural mechanisms of the wrong behavior. 


    Nachum Ulanovsky  20:07  

    Yeah. So, so I can  give, you know, there are hundreds of examples in the book, but I can start with, with the first 300.


    Marty Martin  20:16  

    Go


    Nachum Ulanovsky  20:18  

    So we start with, with perception. The simple thing, which is, how the brain responds, not even behavior, but how the brain responds to stimuli in the real world. So it turns out that the original, you know, there were scientists called Hubel and Wiesel that studied how the visual response to visual stimuli that got the Nobel Prize for it. And these were very classical studies that essentially showed that the early stages in the visual system respond to simple things like oriented bars and dots, and then later you respond to some, you know, corners. And then even later, there are neurons in the brain that respond to complex things like faces or visual sceneries. 


    Nachum Ulanovsky  21:00  

    Now it turns out that if you that the whole idea was to the idea of using these simplified stimuli was that, if you will, you see how the brain responds to, let's say, the primary visual cortex responds to simple oriented oriented bars, or to simple stimuli like that. This will help us explain how the same neuron responds to complex scenery to like a real visual scenery, a real image of the of the world. People have done that— compared how the same neuron responds to visual the simple visual features versus complex image, and the predictive power is often less than 10%.  Meaning the notion just doesn't work, you know, the you cannot predict very well how a neuron will respond to a complex image, from the response of the same neuron to the simple stimuli. And likewise  in the auditory domain, the responses of a neuron to simple sounds often very poorly predict how the same neuron responds to a complex stimulus, like song of a bird, Jimi Hendrix playing the guitar, or anything like that. So that is, you know, you can be sort of positive about it and say: "Well, you know, we know that the brain is non linear and things are complicated. So of course, things will not work so well. And you could be a bit more negative and say this whole program failed, because it certainly does not explain well what the neurons are doing." So this is an example from just sensory side. 


    Nachum Ulanovsky  22:30  

    Now add to this that, also on the sensory side, but now we add a little bit of behavioral aspect, what's called active sensing. So most of the sensory systems in humans and animals are active in the sense that we do not perceive the world visually like a camera. We move our eyes around, and it turns out that we essentially see with our fovea, with the high resolution part of our eye, we see small pieces of the world sequentially, and then we piece this together to get an image of the world. Likewise, mice or rats move their whiskeys around to touch the world. We sniff the world. So we essentially get the world in a pulses of odor, not in a continuous manner. And so do rodents or dogs. Bats emit echolocation calls that are pulsatile and get echoes in pulses. And it turned out that if you start compare, now typically neuroscience experiment of, let's say this whisking system in the rats or the mice, is done by passively in a head-fixed or even anesthetized animal passively moving these whiskers. Turns out that the response of the same neuron, if you passively move the whisker, or you allow the animal to move the whisker and touch things, the response to the neurons is completely different. So now you added very simple behavior, like letting the animal do its normal active sensing of the world. This is before complex cognition, or moving around, or anything like that. The neuron, the brain, looks completely different. The responses the neurons are completely different.


    Nachum Ulanovsky  23:56  

    So these are, these are examples. Now you can think of decision making. Decision making in the real world is often continuous. Let's say, when I navigate through the world, you can think of it as sequential decision making. I will forage for food as an animal I all the time need to decide, do I veer left or do I veer right? Do I stop to eat or continue to search? This is continuous in time and continuous in the space of possible choices. But typically, neuroscience studies of decision making are done in a two alternative forced choice way, where an animal or human has two choices. I can choose either the right card or the green card or the red card, go right or go left. I have two choices. I have to make them. This is discrete in time, discrete in the binary in the space of possibilities, I have just two options. That's again, very far removed from real world decision making or foraging. And turns out that people have shown that this makes it, behaviorally, makes a big difference of how animals and humans behave. These are two examples, and I can give many, many more of how real world behavior differs from these highly simplified behaviors in the lab.


    Cameron Ghalambor  25:02  

    So to follow up, there was another sort of very provocative statement that you made that also sort of really stuck out for me, which was that you wrote that mainstream neuroscience has largely neglected what brains have evolved for. And your claim is that you know it that this is what guides behavior in the real world, in complex natural environments, and that that really, like, really just was a another one of these very surprising things to read that, you know, is this really the case that neuroscience has not taken into, you know, as an evolutionary biologist, that just seemed kind of shocking. And I'm curious if you could expand a little bit more about exactly how this sort of disconnect plays out in terms of neuroscience research.


    Nachum Ulanovsky  25:59  

    Right. So there is a disconnect, and it's exactly the part of thing I'm trying to bridge in the book, so in Ecology and Evolutionary Biology. So you know, Nico Tinbergen, Nobel laureate for his discoveries on animal behavior. And he had these sort of four famous four questions, what kinds of questions you need to ask to understand a biological problem. But this could be about behavior, about any, any biological problem. And he was talking about, you know, the current state of affairs and the past, but also about, you know, mechanistic, or how things are done by the system, versus why, or the or the reason the purpose of the system. Neuroscience is very mechanistic. It tries to explain how things are done, but not what for. What's the evolutionary consequence? Et cetera. Whereas evolutionary biology or behavioral ecology is totally focused on the other end, on the why the purpose of things. And in the book exactly, I try to bridge these ways of thinking and show specific examples of how you might, you might try to try to connect this. But these are really very two different disciplines, not just in the topics, but really in the way of thinking. So yes, whereas an evolutionary biologist or an ecologist often thinks about evolution, what a certain behavior or certain trait of an animal, what, you know, what advantages it gives it evolutionarily. Neuroscientists very rarely think about it. 


    Nachum Ulanovsky  26:45  

    And, you know, I know, when I talk to students, I often go to a poster at the Society for Neuroscience conference, and I see that they found some neurons that respond to, you know, I don't know, to a transition from some fast moving responses of the whiskers to no response. So I asked him, "What does it mean? What is it? Could it be, for example, be used for neuron that signals, you know, getting out of a tunnel of a rodent. Because when you get out of a tunnel, inside the tunnel, you have a high speed touch by the whiskers of the tunnel edges, and as you go out out of the tunnel, into the open space, suddenly you don't get this touch anymore. So this the neuron that's sensitive to such a transition.I'm just making up one example. This example is actually a specific case that I had, but there are many such cases where I ask them: "Can this be a neuron that signals that sort of a behavior?" And they're like, the students are surprised and say: "Oh, this is interesting." They just don't think about it this way. What's the function of this neuron? How does this relate to the natural behavior of animals? Because, you know, the animals in their experiments don't normally go in tunnels in the lab and go out of tunnels. They don't think about it this way. They have a head fixed animal. They stimulate the whiskers in a passive manner and see some responses. But connect it to the natural behavior is not something that students are trained for or typically think about, and I think that's a shame. We should be thinking more about it as a field. .


    Nachum Ulanovsky  27:25  

    Well, you definitely have Cam, and I in agreement that lots more evolutionary biology and ecology should be offered to any biology student. So let's, let's, let's go back to the scope of the book, and I love, I really love the structure of your chapter one, where you listed these sort of 12 different reasons that we need a natural neuroscience. And I don't think we're going to get to all 12 of those, but maybe that encourages people to buy the book. I want to start with the first one, because I bet that's one of the more important ones. And to paraphrase, you said that the brain "cares" about natural behaviors. We've been talking about that a little bit, but can you sort of say a little bit more about why you chose that to be your first major reason for doing natural neuroscience?


    Nachum Ulanovsky  28:40  

    Yeah. So that's the first is really the empirical evidence, right? Because many of the other reasons are in principle, why, in principle, we should care about this, like evolution, etc, but this reason is the key because, and most of the book focuses on surveying hundreds of examples of specific studies that were already published and that show that neurons and the brain in general cares or acts very differently for natural versus unnatural stimulus. So I gave a couple of examples on the sensory side, or on active sensing, or on decision making, on social on the social side. For example, we have shown ourselves, in bats, but there are some related studies also in rodents, that if you put bats in a group and let them interact, then in brain areas that normally are not thought of being social, suddenly you find neurons that are responsive to not just the presence versus absence of other animals, but to the positions of other animals, their identity, so who is it, the sex, their affiliation, the social hierarchy. So suddenly, in a rich social situation, you see very strong social signals in the brain that you wouldn't see otherwise. So this is, you know, just one example of many. 


    Nachum Ulanovsky  30:47  

    You see in the olfactory system, for example, that neurons are very strongly responsive to this. I mentioned this already in the active sensing examples, but neurons are strongly tuned to this inhalation, to this act of bringing the odor inside, the inside the nose. Again, a very simple behavior, but the fact that neurons are really tuned to this act of inhalation showed that the actual behavior matters, whether you have this active the sniffing process or not. There are many examples of, let's say, for example, of auditory neurons or visual neurons that show that there are, on the converse side, that as I said you they don't predict very well. It's hard to predict the responses of neurons to natural scenes or natural sounds from responses to simple sounds. But conversely, you have neurons, evidence for neurons that are tuned very strongly to sort of natural structure of sound and to natural sound, so that, or images as well in the visual system. So there's just a lot of examples, specific examples, of neurons that act very differently between a natural versus a unnatural stimuli and behaviors. And the book really surveys hundreds of such, examples. So I started because, you know, I'm an experimental scientist, and I'd rather start with the data and, you know, leave the, you know, the discussion, the argumentation, for later.


    Marty Martin  32:09  

    For later. Yeah, that makes sense. I guess, you know, it's the evolutionary ecologist in me, although I should say, I spent three years in a neuroscience lab as a postdoc. So I have had some time in that area. I used to go to the Society for Neuroscience meetings, and I've seen a lot of material that way. I wonder how far you're advocating to push this "natural" idea right? Because there's, sort of, we've been talking about natural as the sort of, there's the lab, and then there's the there's natural, right? But an evolutionary ecologist, you know, we really focus on the diversity of what natural means, and so the time of the year and the place in the world, and the sort of priorities in the context of reproductive fitness. I mean, there's a lot, I think there's a hierarchy of natural that might even inform more how to do natural neuroscience. Do you think that that's just too far to ever really happen? Are you aspiring towards, I think, what Cam and I would call like an integrative neuroscience, where there's sort of a thread that runs from the evolutionary ecology, way of thinking about brains all the way down to the molecular details? 


    Nachum Ulanovsky  33:20  

    So first of all, yes, I think that'll be wonderful if we can bridge all these levels of description. And I think in some particular systems and some behaviors, it can be possible. But I agree with you that defining natural is not easy, because you can think of at least three different ways of defining it. One, natural means something that is important for the survival of the animal. Another natural could be something that's important specifically for that animal, subjectively, which does not necessarily correlate with the first one. And third are natural are just things that are commonly encountered in the environment of the of the animal. 


    Nachum Ulanovsky  33:55  

     Now the thing is, you know, for a city rat, the sounds of car honks and cell phone ringtones. Maybe are quite natural, because in the last in the last 20 years, that's what rats have been hearing in New York City. Or in a lab, a laboratory rat or mouse, the sound of the air conditioning and of the animal care technicians coming to bring them food,  that's what they've been encountering lots of from the statistical perspective, that's natural for them. Does it add evolutionary consequences? None whatsoever, because whether they hear or don't hear the animal technician approaching, they will get the food, they will reproduce, they will survive. There's zero evolutionary consequence or survival consequence. But it's common. That's what they hear. They don't hear the forest sound. So it's, it's a hard question, what is exactly natural? I tried to discuss that to some degree, but I don't think there is a, there's a good answer there. But I do, I think that definitely, at least in as much as you know, there's some things that we will all agree are relevant under all definitions, for example, natural communication sounds. You know, using natural communication sounds as stimuli is important evolutionary, because it's the important evolutionarily, they're important for individual animals, and they're encountered a lot in their environment. So that's like one example of things that are under whatever definition you're using. It's a good thing to use, or social behaviors, or navigation, exploration, are behaviors that are important under all definitions, because the animal do it a lot, et cetera. So I think you can come up with certain behaviors and certain types of stimuli that will be natural no matter what, and those are the ones that we should focus on.


    Marty Martin  35:34  

    Yeah, yeah. Agreed. Agreed. I guess I mean, just to sort of try to make a little bit clear the evolutionary perspective, if you take the baboons that, you know, Robert Sapolsky had worked on for the longest time, there's hierarchies in that species, right, such that what it means to be dominant and subordinate is going to profoundly influence everything that it means to be reproductively successful as those individuals. So how the brain operates, right, the perception of stimuli as stressors, as adversity, as opportunity. I mean, there's this extra level of not just natural, but natural relative to the individual given its context. So it just becomes more and more exciting, but more and more complicated. And then, you know, you can push this forever, and then you might become paralyzed with the complexity.


    Nachum Ulanovsky  36:21  

    Right so, and I totally agree with you, and, in fact, you know, I don't want to sound too negative, a lot of, in social neuroscience, for example, people have been asking these questions. So there have been studies where people have asked exactly about dominance hierarchies. But the way this is typically done is done usually not in baboons, but in mice, because they're the standard animal models. And you need a reproducible assay for dominance in mice. So the way they do it is by letting them it's what's called the tube test. They let two mice crawl opposite each other, or run opposite each other inside a tube and push each other, and whoever is sort of the winner in this competition is considered to be more dominant. And you know, it has its advantages because highly reproducible, and you put the same two individuals over and over the usually the dominant will win over and over. So it has some good features. And you can also say, you know, mice are burrowing animals, so for them to meet another mouse in a burrow that's actually pretty naturalistic. Outside in the wild, this can happen, so you know, this is not bad. On the other hand, it turns out that when you let groups of mice run in a complex, naturalistic environment in the lab, but a more complex environment with four or five mice, and they are chasing each other, which is aggressive towards each other, which is a more naturalistic way of assessing dominance, then the correlation between dominance as a test under these group conditions and these classical dyadic pair wise tube test is low. The correlation is not zero, but it's low. It's not great, so. But what's the conclusion? I don't think we should drop entirely the tube test, because it has this power of being reproducible and controlled etc. But I think we should do both. We should test the same neuron to do the same manipulation, yeah, both under the the tube test condition and under these more naturalistic situations. And a lot of people are going now in this direction in when studying mice or other animals. So you know, people are appreciating these aspects. But I agree with you this, that the stress of a dominance hierarchy they have many effects on the on the animal, all over the brain.


    Cameron Ghalambor  38:31  

    Yeah so kind of to continue a little bit with this thread and to go down your list on your table, if we kind of combine the your kind of second and third point, they both deal with complexity. Kind of, I think what, what you and Marty were just sort of discussing, but the complexity comes from two, two different sides. The brain itself is a very complex organ, and the environment is also complex. So when complexity meets complexity, this is the reality of what and how animals are dealing with their environments. And yet, as you've been pointing out,  we've tried to strip away a lot of this complexity. So how are you advocating for us to deal with this? Because on one hand, it sounds like an opportunity, and at the other hand, it also sounds like, you know, overwhelming complexity that requires the levels of control that you know, traditional neuroscience has used.


    Nachum Ulanovsky  39:40  

    Yeah, so well, first of all, understanding the brain is a hard problem, and this is one of them. This is one of the many reasons. Other reasons, of course, it's multi-level so you know, from molecules and genes to cells to networks to systems, etc. So it's a very multi-level system. But also the thing, the complexity makes it inherently difficult. And by complexity I mean that the brain, but also the world with which the brain interacts, is a network of factors. So let's put it in concrete terms. You have a network of brain areas that interact with each other, and neurons are a network which interact with each other. And this network organization, by definition, makes it hard to use the standard logic of reductionist biology, where we looked at sufficiency of necessity, that the way you typically think about, classically about cracking biological problems, you're saying: "Okay, I have a, you know, a factor A that I suspect affects factor B. So let's, you know, tune up A and see whether this tunes up B, and this is like the sufficiency, or let's shut down a and see whether B doesn't function anymore. And this is like the necessity." But if you can go from A to B through C or D or other trajectories, like in a network, then this logic doesn't hold anymore. And this is important, because the whole logic of using or a big part of the logic of using controlled experiments, is I say I want to understand only, you know, let's say decision making. So I strip out movement, and I strip out sensory variability, I head fix the animals, and I strip out attention. I try to get rid of all these other factors. But when I get all these other factors, I mean, there are brain areas that are responsible for these factors, or that are involved in these factors. Or when I clamp down and I remove, essentially the contribution of these factors, it's not that they disappear, just I actually change the activity of brain areas that are supposed to have these top-down cognitive signals. I now sort of got rid of them, and that does affect the brain areas that I care about. So you cannot really, the idea that I will clamp down or get rid of certain factors that don't affect anymore, no, they do affect via their absence. And we know there are many studies that show that even relatively low brain areas like sensory systems are affected by top-down signals like attention, intention, planning. All of these cognitive signals do modulate, very strongly, responses of single neurons, of networks of neurons in even perceptual sensory areas. But if we try to get rid of that when, essentially, we got rid of a very important input into these brain areas, so you can't really get away from that. And this is a problem of any complex system. It's true in ecology. It's true even in complex physical systems, there's a classic, classical style papers in physics of complex systems to talk about these things, about embracing complexity there as well. So what I suggest is, you know, to combine, I think the best way for neuroscience is to combine these controlled experiments that do allow to look at one factor at a time, but also to acknowledge that this is, you know, we're, to some degree, misleading ourselves, and then do the more the full naturalistic experiments where we allow much more richer environments, much more richer interactions, and then compare what the neurons are doing under two conditions. And I think this is the best way to go forward in neuroscience.


    Marty Martin  43:11  

    Yeah. So Cam will tell you that one of my favorite things to do is to take complexity and make it complex. And it was great to read that your fourth item did exactly the same thing. You know, Cam asked you about the world and the brain being complicated, and that interface complexity meets complexity. You were alluding to, sort of loops within the brain, interactions within the brain. But of course, your fourth point is that the world in the brain are bi-directionally connected through these loops. So, I mean, can you talk about that? But maybe, clearly, we will continue to need experiments. We'll continue to do experiments. There's value in that. We've been doing it for hundreds of years. Why would we change? But on the other end of trying to get our heads around this complexity meets complexity meets complexity, is this just sort of large descriptive studies in natural context and trying to map patterns, or how, how do we confront this outcomes of complexity?


    Nachum Ulanovsky  44:12  

    So let me give you a concrete example, maybe from our experiment. I think because it will help, it will help elucidate or understand how you can still make sense out of a complex experiment, even though you'd think that it's hopeless. So we've done, you know, experiments in groups of interacting, groups of interacting bats where there was, there was no experimental manipulation. So a typical experiment. So, you know, we are interested in one of the branches of the work in my own lab, is understanding, you know, social interaction between animals. So you can do it in a dyadic way. So we had one experiment where we had two bats, an observer and a demonstrator, where the demonstrator was flying from a certain location to one of two other locations and then coming back, and then after it came back, the observer was flying to the same location as the other one. So it was like a mimicry task. So this is a highly controlled experiment which separates, in time, your behavior and my behavior, and also it's two alternative you can only go to one or the other location. So this is the more classical experiments in neuroscience, even though it's more complex because it has two animals, but it's still relatively controlled. And we were able, by this controlled experiment, we were able to show that neuron represent the position of the other animal, not just of my own position, but also the position of the other animal. 


    Nachum Ulanovsky  45:32  

    But then we did another experiment, which is a lot less controlled, where we had a group of bats, five to ten animals, both males and females, that live 24/7 in a room and can do whatever they want. There was no was no task whatsoever, and we're recording the neurons in their brain. And you'd say this is hopeless. How can we ever make any sense out of that? But it turns out that animals just don't just behave randomly. They repeat certain behaviors over and over. They fight over and over again. So we can align on these fighting moments or on these positive interactions like all- grooming and to these various behavioral events and see how the neurons respond to repeated instantiations of the same event. It turns out a lot of the neurons respond very nicely and very reproducibly to these social interactions. Even though I haven't imposed anything on the animal. They do it off their own volition, but they do it repeatedly. And likewise, they fly in certain ways, repeated patterns. So we can ask how the neurons represent the position of the bat to repeated flight behaviors, etc. So we can utilize the repeatability of the bats' behavior to make sense, to make sense of what the neurons are doing. So, so even though the behavior is complex, it's not completely random, which plays to our advantage.


    Marty Martin  46:50  

    Yeah, yeah. I like that. I like that a lot. Yeah. I mean, it's always we can't do experiments for everything, but what else do we do? You know? And I like this, sort of letting the natural behavior guide how to look at the system. I mean, it's, it's not the same kind of inference that one would gain from manipulation in the classic sense. But still, I mean, what else are you going to do? I agree it's informative. 


    Nachum Ulanovsky  47:14  

    And so you have, I have in the book, is sort of a diagram that looks at the various fields of neuroscience, or subfields of neuroscience, on these two axes of control versus naturalness. And essentially, you'd think that they it's like a one-dimensional axis. It's either I'm more controlled or I'm more natural. But it's not necessarily the case. It's more of a two-dimensional plot. So it's true that there is a trade-off there, but they can also be behaviors that are both highly natural, but also highly controlled in the sense that they are highly  repeatable. For example, bird song is one of my favorite examples. The song of certain bird species is, I mean, they sing, can sing hundreds of times a day, so they repeat certain behavior over and over again with a precision of a few milliseconds, a few thousandths of a second. So it's a highly, highly repeatable behavior, a lot more repeatable than any you know, mouse laboratory behavior, yet it is done of their own volition, and it's totally natural. So you can be both natural and controlled and repeatable. So and you can also do manipulations in this experiment, you can suddenly put a sound—and people have been doing that—you can present a sound in the middle of a song and see how that affects the singing of the birds and what the neurons are doing. You can cool down the brain or or inactivate areas of the brain with light, with what's called optogenetics, all these cool new methods in biology, in neuroscience. And that's wonderful. That's wonderful so, but you can be both natural and controlled to at least to some degree, and I think those kinds of behaviors are what we should focus on.


    Cameron Ghalambor  49:01  

    I'd like to maybe change gears a little bit and talk about plasticity. So partly out of selfish reasons, because this is one of the areas that I'm very interested in, research wise. But I think it's also important to make a distinction between how neuroscientists think about plasticity because it it's part of this complexity that we've been talking about and how evolutionary ecologists think about plasticity. So for an evolutionary ecologist, the concept of phenotypic plasticity is that  for a given individual, for a given genotype, the phenotype is dependent on the environment. So if you know the genotype and you know the the environment, you can predict the phenotype. So this is a sort of a predictable change. So you know that if a certain genotype is exposed to a certain environmental stimulus, or like certain environments, say, during development, a daphnia exposed to the pheremones of a predator, will develop a spike, a morphological spike that presumably is adaptive, that protects it against the predator. But neuroplasticity, or synaptic plasticity, I think, has a different definition that sort of deviates. So, so can you define plasticity as a neuroscientist would use it, and then maybe we can talk about similarities and differences that you know might allow for crosstalk between the disciplines? 


    Nachum Ulanovsky  50:40  

    Yeah. So first of all, people don't really neuroscience. People don't think of like evolutionary timescales,  not even timescales of the lifetime of an animal. Plasticity is usually on a timescale of either, you know, a few hours or a few days as you learn some task, for example, a mouse or any animal needs to learn something. You know, performing some task, needs to learn to associate certain images with certain behaviors or certain sounds of certain behaviors. So, you know, as the animal learns, there's plasticity or change in the behavior of the animal, and, concurrent with that, there are changes in synapses, so synaptic plasticity, and the idea is to try to to link the two. Now that's a good scenario. Often people study, you know, just enough, just the synapse level, or just the behavioral level, and don't connect the two, but, but a lot of studies do connect the two. And that's you know, what people normally when they talk about plasticity in neuroscience, this is it. It's learning, and how you learn something, and this becomes a memory. And how is that reflected in synaptic plasticity? It's not really about no lifetime of an animal or evolutionary time scales at all. 


    Cameron Ghalambor  51:55  

    Yeah, although so like physiologists, I think endocrinologists have also kind of come to start using a moreso, I would say, evolutionary perspective on plasticity. So, for example, we know that organisms have the capacity to acclimate. So this is a type of reversible plasticity. So seasonally, when it gets cold, you know, there are some biochemical, physiological changes that might make the organism more cold-adapted, and then in summer, those changes are reversed. So if you learn a task and you see the sort of plastic, neuroplastic changes, is there some analogy that can be thought of there that these are reversible sort of changes that could I'm trying hard to make connections.


    Nachum Ulanovsky  52:49  

    Yeah, no, no. So I can these connections are very, very rare, but they can be made. So I can give you an example from a study that actually my postdoc, Saikat Ray, has done, but not with me in his sort of previous lab. One of the studies that he did, he worked on the truth and shrews. Turns out that Etruscan shrews, every year shrink, their brain shrinks and expands, you know, seasonally. And in particular, their somatosensory cortex. This is the area that has to do with the sensing, with the whiskers they have, this huge whiskers that they use to sense the world by touch. And this one, this area changes both its volume, just gross volume, but also its neuronal tuning properties in a seasonal manner. So, yes, you can ask this in neuroscience in a way that links to, you know, the way an evolutionary biologist would think about plasticity, but these sorts of studies are very rare. I mean, there's literally a handful of studies of that sort. And I think exactly the conversation that we are having now is exactly the kinds of conversations that neuroscientists should be having to talk to ecologists, talk to ecologists, because this really inspires you. 


    Nachum Ulanovsky  54:00  

    I can give you an example from, sorry, it's not about plasticity, but an example why it's important for neuroscientists to talk to ecologists. So, you know, when I started my labs, some almost 20 years ago, I one of the first types of experiments that we did is to record, to look at how the brain represents three-dimensional space. Because we work in brain areas related to navigation, where there are neurons called place cells that represent my position in the world, and neurons called head direction cells represent the direction the world, and grid cells that represent sort of distances in the world. So you have sort of cognitive maps, compasses and distance meters in the brain. And those have been studied for many decades, in rodents primarily. And you know, the discoverers, Keith and the Mosers got the Nobel Prize for it. So it's a well established field in neuroscience. These are typically done in empty boxes, in animals, in rodents running in an open, two-dimensional environment. So one of the first questions that I asked was, how is three-dimensional space represented in the brain? Bats are an animal model, it's great animal model for them. They're mammals, so their brain is very similar to ours and to rodents, but they fly in three dimensions. So we've developed wireless neural recording technologies, very tiny devices that allow recording in freely flying bats. And we found that a lot of these neurons, or all of these neurons, exist also in two dimensional space. And there are some similarities to what you find in rodents two dimensions, and some differences. So we had a whole series of papers about that. 


    Nachum Ulanovsky  55:26  

    But then I remember when I was showing these first studies from the lab in conferences. I was invited to give a talk at the Ecological Society of America, which is not the typical conference I go to, but I went to that conference and other more like behavioral ecology type of conferences that very few neuroscientists go to. And I was presenting this, this sort of study that was done in bats flying in a five by six by three meter room, a lot bigger than the one meter boxes where you typically study rodents. So I was thought: "Oh, we're so great. We're recording bats in a five by six meter room. We're amazing." And then I come to the ecologists and they say: "Well, it's very interesting. It's wireless, it's natural behavior, nice, but dog bats normally navigate like kilometers outdoors, don't they? I mean a few meters." They were so under-impressed. Like every neuroscientist I talked to about this, was highly impressed, but the ecologists were highly under-impressed. They say you should record out in the wild. I mean, they don't know even how difficult. Nobody's recorded in the brains in the wild, their technical difficulties, etc. But, you know, they put a mirror in front of you, and you look at the mirror and say: "Well, you know, they have a point. Real behavior in the wild is even more extreme than that." So, and this encouraged me, over the years, to try to tackle this experimentally. So we've built these long tunnels, sort of big tent like structures that are hundreds of meters long, that allow us to study how the brain represents spaces are even larger than these five meter rooms, like 200 meter long tunnels, and we found that the way space is represented in these really large environments is completely different than the classical textbook picture. And more recently, we've been recording in bats flying outdoors on remote, uninhabited islands off the coast of Zanzibar. And we found very interesting findings there as well. So, but this was a lot of this was motivated by talking to the ecologists who you know, you know, give you the feedback on how they see things. So I really think this is first of all important to talk to people from other fields, because you get at least inspired in what you do. But also, technology matters here, because a lot of these things really were not possible without the technological advances.


    Cameron Ghalambor  57:44  

    Well, the Mosers are here. I'm at the same we're at the same university. So I maybe need to have some more conversations with them. So kind of continuing on this vein, I want to get back to this idea of plasticity, because I think, I think there's a there's also a deeper connection that I'm not sure was, you know, articulated in the book, but I think naturally comes out of the book, sort of implicitly. And this is a topic that Marty and I have talked a lot about, and we've talked about it in the context of immunology, but in a very sort of parallel, analogous way to, I think, the way you're thinking about it with neuroscience. And the idea is that you know back to more of the evolutionary sort of definition of plasticity, that that these phenotypic responses are dependent on on the environment, but they're also products of evolutionary history. So you know the the ability of the daphnia to produce this like spike when it's exposed to the the chemical cues of a of a potential predator, have likely evolved in response to natural selection. And in the natural context, we can see that you know this clearly, how this, something like this, could evolve. 


    Cameron Ghalambor  59:05  

    But then you know, if you were to take the daphnia out of its natural environment and put it into an artificial test tube and and try to make sense of this same behavior, the same phenotypic response during development, it might be totally may not make any sense at all. And organisms in general, when they're placed in novel environments, exhibit these sorts of, you know, what an evolutionary biologist would call plastic responses, but these plastic responses may be highly maladaptive in terms of, you know, the consequences of the physiological response, the immune response, the neural response.And I think that that, you know, a lot of the examples that you give throughout the book, you know, I'm looking at these thinking like, well, yes, these are examples of maladaptive plasticity. Of, you know, you've put the animal in an environment that it has not evolved in, and it's giving a response. You're measuring something, but it's very far removed from what we would think of as the the, you know, the response that would have evolved under natural selection. So does doesthat make sense? Or, my is this going off on a tangent? A little too much? 


    Nachum Ulanovsky  1:00:27  

    No, no, it makes sense. First of all, I think I mentioned in the book there what sometimes people call the the measurement fallacy, the fact that you can measure something doesn't mean that it's meaningful, right? So, yes, you can measure neural responses under any condition that you will put an animal in the question is, how much it teaches you about the brain in the real world? And yeah, so some of the things I discussed in the book, in the first chapter is about, you know, stress. So some of the over controlled experiments in neuroscience have to do with stress, have to do with plasticity, but in the bad, not in the your sense, but in the you know, you change what the brain is doing by over-constraining the animals. And if you're not, if you want to, if that's not what you care about, then that's a problem, etc. So I definitely think that there could be some of the experiments in neuroscience could be maladaptive in this sense. 


    Nachum Ulanovsky  1:01:20  

    And again, I think that you're right, that a lot of the things that we and others have been finding by allowing the animal to behave more more naturally, you know, you find interesting new things. But there could be two reasons for this, so let me give you a concrete example. So you know, we've shown that in the bats, when you introduce a navigational goal, then in the hippocampus, in this brain area that represents where I am in space, you suddenly start seeing that there are neurons that represents also the direction and distance to the navigational goal, the vector to the goal. Now the reason you people didn't see it before when letting animals run in empty boxes is that there was no navigational goal. So you cannot measure even a response to something that is not there in the experiment. This is, you know, one problem. The other problem, the more fundamental one that you are mentioning, is maybe by putting animals in situations that don't have goals. This response is, you know, over the evolution of laboratory mice in the last hundred years disappeared. 


    Nachum Ulanovsky  1:02:25  

    Now the good news, so I think it's more, it's more the former and not the latter, because there are many examples of studies. I mentioned some of them in the book that when you, even in laboratory animals that have, you know, "evolved", so to speak, in laboratory conditions for the last hundred years, you put, for example, you take rats and put or mice, and put them in a very big environment with natural earth, with dirt that they can dig in, and suddenly they start digging normally looking tunnels, tunnel systems, like in the wild. For me, that's amazing, and that's amazing, and that's actually very good for neuroscience, because it means that you can take the rat out of the wild, but you cannot take the wild out of the rat, even though these rats have been living in captivity for a hundred years. So that's hundreds of generations that they haven't seen, you know, natural dirt at all. You introduce it for the first time, after a hundred years, they know what to do. So that's actually very positive. It means that a lot of the natural traits are still there. 


    Nachum Ulanovsky  1:03:32  

    On the other hand, some traits, by looking at laboratory animals, and I also discussed some of that in the book, have disappeared. So for example, a lot of social traits. For example, female mice in the lab. In laboratory of female mice are very non aggressive, lot less aggressive than the males in the wild, female mice are as aggressive as the males. But a reason the laboratory ones are non-aggressive is because when the early breeders in the 1920s were breeding mice as pets, essentially, and those were the ones that, you know, the most common mouse strain in the world is C 57 black six mouse. You know where it comes from? C 57 black six. It's maternal lineage.came from female number 57 in a farm of a particular lady was growing mice as pets in the 1920s it's one female mouse that gave rise to 


    Marty Martin  1:04:26  

    One mouse 


    Nachum Ulanovsky  1:04:27  

    One mouse, particular female number 57 it gave rise to this whole mouse that is used for everything, biomedical studies, drug studies, neuroscience, everything. One mouse gave rise to it all. And of course, when they were breeding these mice, they were trying to get rid of females being nasty to each other because or attacking the pups. And it's bad for you as a breeder, the bad pop bad bad pet. If they kill each other, right? You want them to be nice to each other. But, you know, in the real world, are not always nice to each other. So, you know, you get rid of certain social behaviors this way, and this has been well documented. So certain traits stay there, certain traits disappear.


    Marty Martin  1:05:06  

    So one more question that's sort of along these lines, instead of the C 57 black six problem you work on, I guess you would call bats non-model organisms, as you know, Cam and I mostly work on non model organisms too, and especially for those species, we worry about stress. So given that, you know, Nagel said eloquently, it's a very different thing to be a bat than it is to be a human, how do you think about the influence of stress, given that your perception of stressors as a human are not going to be the bat's way of thinking about things?


    Nachum Ulanovsky  1:05:40  

    Yeah. So first of all, these particular bat species that we are studying, the Egyptian fruit bats, are living in cities. They're very used to humans. They're very common, which is good. They're they're, you know, by their nature, they're not so stressed by humans. Other bat species that are more shy, more afraid of humans, would be more of a problem to study. These particular bats are used to humans in general. Get along with humans very well. So that's already a good thing. Also, when you catch them right out of the wild, and I do the catching myself, so you know, you they're fine. They're really fine in this respect. But suddenly, over the years, we have noticed that when we you know, the more naturalistic our experiments became, the less stress the animals were. So in the first studies that we've done, where we had them crawl in the two-dimensional box, very similar to the rodent studies, that was hard on them, even though the face is very tiny. And so you think, oh, what's the problem? They can just cover this very tiny space. It shouldn't be a problem, but this was very stressful for them. Okay, we collected data and we published studies on that, but that was, you know, it was not easy on them. But as we increase those spaces to the five or six meters room, and then to these hundreds of meters of tunnels and the outdoors, the bigger the space and the more animals are involved, it's easier for them. They are lessed stress. It's a more naturalistic, even though, even if it's in the lab, there are less stress. So I think the idea of going to more naturalistic behaviors as a scientific endeavor also makes them less stressed, and the results therefore more valid. So I think this is these two directions align with each other, yeah.


    Marty Martin  1:07:15  

    So the more natural I mean, I'm hearing sort of more natural context to the extent possible. You give them more space, you let them be a bat, as opposed to forcing them to be the organism. That's convenient for you, right? Which is oftentimes what we've done. But I guess so to go back to the Nagle idea, we're just not bats. How can we ever really know? I mean, when you say that the bats aren't stressed or they're behaving different, presumably, than they did before, and that's probably representative of less stress, but even when they're behaving normally, you know, we're still not bats. It's still hard to, you know, the audio environment is a very different thing for a human than a bat. So how do you think about that?


    Nachum Ulanovsky  1:07:50  

    First of all, you're right, but we're also not mice, and we're also not many other species. 


    Marty Martin  1:07:56  

    Yeah


    Nachum Ulanovsky  1:07:56  

    So this is important to appreciate, because this even the sensory world of a sort of sensory world of a bat, specifically the Egyptian fruit bats that I study, have excellent vision, much better than rats or mice. So they are very visual, and so this actually makes them a bit more similar to us, but they also have echolocation that's a very alien sensory system compared to humans, to most humans, even though blind humans do use echolocation. But if you think about a mouse that has a very low resolution vision, so it's like you putting glasses that are, you know, five dioptres away from what you normally would put, you know, you'd see the world very blurry. This is our mouse sees the world very different than how we humans perceive the world. And they move their whiskers around in a periodic fashion. Also, that's not how we sense the world. So there are many differences also in how mice sense the world. They hear a different frequency radius that we do. They hear ultrasound, just like the bats we don't hear ultrasound. They are sensitive to smells that we are not sensitive to. So, you know, every species has its own idiosyncrasies, and it's important when you're using a certain species or studying a certain species, to understand the brain, or any aspect in biology, it's important to know, as a student of these, of these creatures, to know about these creatures, how they perceive the world and to appreciate it. So maybe subjectively, it's hard for me to know, to understand how it is, what is like to be a bat in English words, but, but I think it's true for any, any, any other species, except maybe the monkeys that are very similar to us. So we, but we need to, really to to understand the behaviors of the animals as well as we could, and put ourselves in their legs or wings, so to speak.


    Marty Martin  1:09:39  

    This is a fascinating topic. I absolutely love the book and there's so many things to hit, but let's spend a little bit of time on this future and how you see natural neuroscience developing as a field. In chapter eight, which is about the future outlook, you emphasize that there's value to bridging neuroscience, ethology and behavioral ecology. So what does that mean? How is that going to look and maybe, what are the first steps to making that happen?


    Nachum Ulanovsky  1:10:02  

    Yeah, so there are two aspects, I think maybe I'll very briefly just mention the technological aspects, and then let's talk about the conceptual aspects. The technological aspects, I think, is important to mention, because it's not that neuro, if you look, you know, even the early 50s, you know David Hubel. I mentioned Hubel and Wiesel. David Hubel talked already in 1950s about the importance of studying freely moving animals. He has this very unknown paper, but where he talks about this, and he has the paper that I could describe about this today. It's a paper from 1957 but back then, it was really hard, technically, to record neural activity in freely moving animals. And the revolution that happened in the last decade or two decades in neuroscience is that it became possible to record neural activity in freely moving animals, more and more freely, either by animals that are connected with the cable and you record their activity with electrophysiology or with imaging, with miniscopes, miniatures microscopes. Now there are these wireless systems that we've contributed to developing these wireless electrophysiology devices, these neural loggers, but there are also wireless microscopes that are tiny and can run on the head of a mouse or rat. So you can record the activity of the brain in animals performing naturally, which allows them to cover spaces, to go under things, to interact socially. Because, you know, we think of connecting an animal with all the animals with cables. If you try to record in social neuroscience experiments, all the cable will get entangled. So going wireless is very important. Also manipulating the brain. There are these beautiful optogenetic techniques that are now that allow to do manipulate the brain in a freely behaving animal, also sometimes wireless or with a cable. And also there are these amazing revolutions. There's this amazing revolution that uses machine learning to analyze the behavior of an animal. That is also amazing because classically, the ethologists, physiologists were creating etograms, which essentially delineate the behaviors of an animal. It was done manually by scoring and saying, Oh, now the animal does this, and that animal does that. That's prohibitive if you want to have hours and hours or many weeks of video data from an animal doing it manual is really impossible if you do things at scale with dozens of animals. But with these machine learning methods, turns out that these neural networks can really do this for you. They can segment the behaviors of the animal and annotate them, and this really allows you to let the animal now do the behavior more naturally. So these three revolutions, those have been the basis for some of the recent studies that I've surveyed in my book. But also I think they should be accelerated even more in the future, and all these technological advances will allow to do, you know, more fascinating experiments on the on the technical side. So this is one thing that that is very important. And I discuss some of that in the book. 


    Nachum Ulanovsky  1:12:54  

    Then there's the conceptual aspect. And I think we're really, as I mentioned in the beginning, we really have a disconnect in the kinds of questions that we're asking in neuroscience. For example in , you know, animal navigation, so migration of an animal. Let's think of migrating birds. You can think of many levels to ask the question you can ask, how does he now? How does the bird know and navigate in terms of space? But there's also, how does it decide on that, on the timing, the timing of the birds leaving, the bird leaving to migrate, you know, from the north to the south, the south and north, you know. So there's the evolutionary aspects, or the fitness aspects. If I leave, if I go north on my northern migratory route, if I leave too early, I might still get to the summer grounds. I might be hit on some the last winter storms and perish in the way. But if I get too late, too late, then the best nesting spots in my summer nesting area have been taken already by others. So I cannot leave too early and I cannot leave too late. There are consequences to both of fitness and evolution, etc. So that's what drives this on the this is our behavioral ecologist would think about it. But then, from a neuroscience perspective, there are the mechanisms. How do I, you know? How does the brain decide, rather, which circuits in the brain decide, make this decision of when to leave? And I think if you think about it this way, you can think of about both the consequences, but also about the mechanistic side and make progress.


    Nachum Ulanovsky  1:14:23  

    Likewise, you know, in echolocating bats, people have been presenting simple sounds, these pure tones and broadband noises and recording in the auditory cortex and didn't really make could maybe make sense, sense of what these audito recordings of bats was doing, but once they started using the naturalistic sounds of their echolocation. It turned out that the auditory recordings of bats contains neurons that is sensitive to the time delay between the pulse and the echo. So essentially, neurons that measure the distance to the target and neurons that measure the Doppler, the speed of the target, by the Doppler shift, the frequency shift of the returning echo, etc. So you it was a top-down approach, you know, something about the behavior, what the animal is trying to achieve, and then you use that to study the brain. So this is an example of how you can relate the field of behavioral e cology that has been studied for many decades to neuroscience in a way that wasn't really done before.


    Cameron Ghalambor  1:15:18  

    Yeah, no, I think that that kind of integrative approach holds huge, huge potential. I have one last question about this sort of, you know, integration, or integrative approach, and looking to the future. And it wasn't something that I saw, at least I couldn't remember distinctly in the book, which is about variation among individuals. So among, you know, in behavioral ecology and evolutionary biology, there's a lot of emphasis on the variation among individuals, because this is the variation that selection can act on, that can lead to, you know, adaptive solutions to problems. And, you know, I think, to some degree, behavioral ecologists are interested in topics like personality. How represented is that in traditional neuroscience? And is, and is this also going to be part of a naturalistic neuroscience moving into the future.


    Nachum Ulanovsky  1:16:23  

    So it's not, in traditional neuroscience, it's not very well represented, but I think it's becoming more and more appreciated, the issue of individual variability, and I think you can ask very fascinating questions about that. So historically, I remember, you know, during my postdoc, I was amazed when I was reading both studies that are coming from psychology and coming from zoology, let's say, of learning of various tasks, they even make the graphs completely differently. In psychology or neuroscience, you'd make what's called the learning curve, which is, you know, your average animal, average mouse, or whatever animal it is, monkey or human. So you'll have the average and some error bars around it showing how the performance improves over time. So that's the standard learning curve. But what underlies when you plot average in the standard deviation around that, what underlies this is that you assume that there is such a thing as your average mouse or human, and everything around this is noise, but that's not noise, that's individual variability. Whereas in zoology, they'll, you know, study ten animals, you achieve the individual learning curve of each of the ten individual animals. So it's like even the way things are plotted are differently and reflect these very different traditions in this field. But I think now in neuroscience, people are starting to appreciate that. You know, animals have personalities. And again, the important part is that the more if you allow more naturalistic behaviors, again, by naturalistic, I don't mean outdoors, I mean more naturalistic in the lab, then you can expose more of the individual variability. By over-restraining, if you head fix an animal and barely allow it to move or do anything, then they'll all be sort of a bit similar to each other, but the more you allow them to move around and to be more natural in their behaviors, then the more you expose individual their ability. And then you can start asking questions about for example, let's say the social aspect. If I'm an animal that likes to interact, you know, with with you Marty, but not with somebody else. Then do I have neurons in my brain that are representing you, but not somebody else, etc? And the answer is yes, we found exactly such neurons that represent my preferred individuals, the one I interact with with the most, which is exactly related to individual variability. But I think also in other other labs are more and more starting to think about this, both in neuroscience and in psychology. And I think you can ask questions animals that, let's say navigation, move more in straight lines or move more in curved lines, will the neurons in their brain be a bit different? I think you can, you can think of interesting questions and interesting experiments that utilize this individual variability and try to look at it at the at the neural level. I think it's very important. And there is huge individual variability.


    Marty Martin  1:19:09  

    Interesting. So I have one more question myself, and it sort of follows from this one, I'm not really an immunologist, I'm not really an endocrinologist. I just like lots of things, but most of my lab's research tends to be about the immune system and steroid hormones. When I'm reading your book, it doesn't feel like your appeal would only be useful to neuroscience. I mean, I think, wouldn't you argue that most dimensions of physiology could benefit from the same mindset?


    Nachum Ulanovsky  1:19:34  

     Yeah, I totally agree. I think outside of neuroscience, also looking at more naturalistic situations is important in biology. In general, biology has been, in my mind, a bit too over reductionist, again, because of the immense success of molecular biology, that is very reductionist. But I think, you know, more integrative biology is important. And you know, when I gave my book at the time to colleagues to read, some of them told me you should talk more broadly about other aspects, and also about molecular neuroscience that I do cover in the book a little bit, but not, not enough, for sure. But you know, you have to put a line somewhere, you know, my expertise is more in systems neuroscience and behavioral neuroscience, so I covered more of these areas that I understand about more, and did not delve into these other domains. But I totally agree that also looking at molecular neuroscience, at immunology, at many other aspects, under more naturalistic situations, could reveal very interesting aspects about the brain. In fact, I was half a year ago. I was in this sort of embo workshop on non-standard model organisms in biology, biomedicine. This was not about neuroscience. I was one of the only neuroscientists there people, there were immunologists, there were people study all sorts of aspects of biology, but in, often in non-standard organisms, and importantly, often in their naturalistic behaviors. And you see very interesting effects there, on immune systems, etc. So I definitely think this applies much more broadly to do biology.


    Cameron Ghalambor  1:21:01  

    Yeah, I agree. Well, I think we could go on for a lot longer, but we like to end usually by giving you an opportunity to kind of, if there was anything more that you'd like to say that we didn't get to cover, that you you really want the audience to sort of take away from our conversation, give you one last chance to speak on this. 


    Nachum Ulanovsky  1:21:24  

    Yeah, I think  one aspect that I haven't mentioned, I think it's important to say. So we focused on some of the examples from, you know, from my own work on bats. But I want to emphasize that, you know, I also have a sort of a graph like that in the book about the aspects of diversification of behavior versus diversification of species. So traditional neuroscience focuses on sort of standard, controlled, you know, standard species and standard behaviors, or simplified behaviors. And you can, you know, think of these, the importance of diversifying species, studying, you know, more non-standard species, which is definitely something I do personally. But, for me, more important is the other axis of diversifying behaviors. Even if you study and the book mostly focuses on standard species, on mice, on humans, on monkeys, etc. I think even if you study standard species, mice, Drosophila, humans, you can look at more naturalistic behaviors, and you'll learn a lot. So it's you don't necessarily. It's does not. It's not a book that focuses only on non-standard species. It focuses on all species. Essentially, no matter what the species that you study, if you are going to look at more naturalistic behaviors, you'll find something surprising and interesting about the brain. And this is my take-home message to people who, especially to students, to the young generation, students and postdoc who go into neuroscience and think about career choices. You know, studying natural behaviors is unique, is not not many labs do it. It gives you an interesting niche that where you can be unique, and it gives, really, the opportunity to make totally novel discoveries. And that's sort of regardless of the technological aspects and the methodological aspects, you know, which cool, new, you know, imaging or optogenetic method you want to use, that's fine, but do it on animals doing materialistic behaviors. I think that's the way to crack the brain. .


    Marty Martin  1:23:20  

    Yeah, I like that. That's sort of short term low hanging fruit, but quite promising low hanging fruit.


    Cameron Ghalambor  1:23:24  

    Yeah that is a good message to end on, yeah, yeah.


    Marty Martin  1:23:28  

    Yeah yeah. Well, thank you so much, Nachum. We really appreciate the chat. We wish you the best with the book. Like we said, it was a it was a fantastic read. A lot of fun to get to meet you.


    Cameron Ghalambor  1:23:36  

    Yeah really appreciate it. Thanks so much. 


    Nachum Ulanovsky  1:23:40  

    Thank you so much. Thank you guys.


    Cameron Ghalambor  1:23:59  

    Thanks for listening to this episode. If you like what you hear, let us know via Bluesky, Twitter, Facebook, Instagram, LinkedIn, Threads or leave a review wherever you get your podcast, and if you don't like something, we'd love to know that too. All feedback is good feedback. 


    Marty Martin  1:24:14  

    Thanks to Steve Lane, who manages the website, and Molly Magid for producing the episode.


    Cameron Ghalambor  1:24:18  

    Thanks also to Caroline Merriman and Cass Biles for help with social media, Brianna Longo, who produces our awesome cover images, and Clayton Glasgow, who blogs about topics covered in the main show. Check out his work on our sub stack page. 


    Marty Martin  1:24:32  

    Thanks to the College of Public Health at the University of South Florida, our Substack and Patreon subscribers and the National Science Foundation for support.


    Cameron Ghalambor  1:24:41  

    Music on the episode is from Podington Bear and Teiren Costello.

Big Biology