Ep 112: The Entangled Organism (with Sonia Sultan)
Does biological plasticity have a cost? Are there evolutionary consequences of plasticity and of organisms acting on their environments?
In this episode, we talk with Sonia Sultan, the Alan M. Dachs Professor of Science in the Department of Biology at Wesleyan University. Sonia has spent her career studying the interplay between organisms and their environment. Specifically, she studies how environmental conditions influence the development of organisms, and when and how these developmental trajectories can include niche construction. Sonia refers to this complexity as “entanglement,” genes and environments working together to alter phenotypic expression, but then variation in phenotypic expression feeds back on environments and genes to alter evolution. She put this perspective to the test in a recent paper coauthored with Mike Wade in the Journal Evolution & Development. In it, Sonia and Mike took a new perspective on the Price equation, finding that niche construction and other forms of organismal agency can change evolutionary outcomes.
Cover photo: Keating Shahmehri
-
Cameron Ghalambor 0:06
Marty, how's your semester wrapping up?
Marty Martin 0:08
I'm just finishing up my grading on our last exam and evolutionary medicine, final grades are due tomorrow.
Cameron Ghalambor 0:14
Me too. I'm in the middle of grading final exams for genetics.
Marty Martin 0:18
And how's that going?
Cameron Ghalambor 0:19
It's going okay. But I've been thinking a lot about how we teach certain topics and how we study them.
Marty Martin 0:25
Give me an example.
Cameron Ghalambor 0:26
Well, in quantitative genetics, we teach the total phenotypic variation in a population can be broken down into that, which is due to the heritable component, think genes, and that which is due to the environment. And that which is due to the interaction between genes and the environment.
Marty Martin 0:44
Ah yes, the fundamental formula of quantitative genetics. But what's wrong with it?
Cameron Ghalambor 0:48
There's nothing really wrong with it. I don't have a fundamental problem with the approach. But in natural populations, it's just such a boring way of viewing the environment's effects on the phenotype.
Marty Martin 0:59
Ah, I see where you're going. The environment isn't just a source of phenotypic variation and noise. It can also cause predictable changes in phenotypes in one of our favorite biological ideas, phenotypic plasticity.
Cameron Ghalambor 1:10
Exactly. For example, the environmental term explains why daphnia develops spines when exposed to predator cues, why plants grow larger leaves in the shade, or why animals can adjust their physiology and acclimate to seasonal changes in temperature.
Marty Martin 1:27
We will recognize that these examples of adaptive plasticity partly a rise because natural selection excellent genetic variation for different forms of plasticity in populations. But incorporating this plasticity into models of evolutionary change, that remains a big challenge.
Cameron Ghalambor 1:42
One problem is that the organism can also modify its environment via this plasticity. This interplay between the environment causing phenotypic variation and organisms changing their environment can impact the strength of selection and lead to unexpected evolutionary dynamics.
Marty Martin 2:00
The most popular evolutionary models tend to ignore these feedbacks.
Cameron Ghalambor 2:04
Or assume their effects are too small to matter.
Marty Martin 2:06
And instead just focus on how a trait changes from one generation to the next based on the strength of selection and how heritable the trait is.
Cameron Ghalambor 2:13
The catch is that fitness, heritability, and the strength of selection are all very context dependent.
Marty Martin 2:20
So in stable or controlled environments, our traditional models work really well.
Cameron Ghalambor 2:24
But in the messy natural world, they can start to break down, especially when the environment that the parents' generation experiences is different from the environment experienced by their offspring.
Marty Martin 2:35
Our guest today is Sonia Sultan, who is the Alan M. Dachs, Professor of science in the Department of Biology at Wesleyan University. Sonia spent her career thinking about the interplay between genetic and environmental factors and evolution.
Cameron Ghalambor 2:47
We talked to Sonia about her views on plasticity and how they've changed over time. We also focus on a recent paper she published with Mike Wade in the Journal Evolution and Development, where they incorporate how organisms modify their environments into the Price equation.
Marty Martin 3:03
The Price equation, one of the most commonly used models of evolution, has two terms. The first term models a population's response to selection by estimating the relationship or covariance between trait values and fitness.
Cameron Ghalambor 3:16
The second term in the Price equation deals with all the complicating forms of the environment, including how environmental change impacts fitness, but also genetic forces including epistasis. This second term has typically been ignored or assumed to be constant.
Marty Martin 3:32
In the paper, Sonia and Mike asked what the implications are if we reverse our thinking and hold the first term constant, and then explore different scenarios for how organisms might adaptively change or select the environments they occupy.
Cameron Ghalambor 3:45
They found that when there is genetic variation for traits that modify the environment, the feedback between organisms and their environments can actually accelerate adaptation.
Marty Martin 3:54
While their model represents a first attempt to incorporate some of the complexity that has long been recognized in natural populations, it provides a really exciting way to start thinking about old ideas in new ways.
Cameron Ghalambor 4:05
Before we get started with Sonia, please remember, we're a nonprofit, and the Big Biology coffers are getting dangerously low. We want to keep making the show, but we need your help.
Marty Martin 4:16
To help us you can become a patron at www.patreon.com slash bigbio. There, you could donate 1, 2, 3, 10, 25 or even $50 a month. We're also revising our Patreon tier benefits and we'll share more info about those soon.
Cameron Ghalambor 4:33
Or if you prefer, go for a one time donation and on our webpage, www dot big biology.org. All amounts are welcome.
Marty Martin 4:42
And if the holiday season has you strapped for cash, no worries, support us by telling a friend about the show or just talking about us on your social media feeds.
Cameron Ghalambor 4:49
Or if you're looking for holiday gifts, buy that big biology superfan, a t-shirt or a throw pillow from our red bubble store. You can access that store on our homepage.
Marty Martin 4:59
Remember, Big biology team members Molly, our producer, Dayna, our social media expert, and Keating, our artist, are all supported by you to keep the show going, please give if you can.
Cameron Ghalambor 5:09
I'm Cameron Ghalambor.
Marty Martin 5:10
And I'm Marty Martin.
Cameron Ghalambor 5:11
And you're listening to Big Biology.
Cameron Ghalambor 5:21
Sonia Sultan, thanks so much for joining us today on Big Biology. We're really looking forward to talking to you today about your research your perspectives on phenotypic plasticity, niche construction and the implications for adaptive evolution.
Marty Martin 5:36
So to begin, let's start off with one of our favorite topics and one of your favorite topics- plasticity. How did you start thinking about plasticity and evolution? I think you were doing it long before it was sort of fashionable. Where did that come from? And what was your inspiration?
Sonia Sultan 5:52
I was plasticity when plasticity wasn't cool? Yes, that is very true. So honestly, my inspiration came in part because when I began graduate school, which I did, because I was just sort of passionately curious about organisms, and especially plants. When I began graduate school, I had very little background in science, because instead of majoring in biology, I had studied history and philosophy of science, mostly because I am a stubborn, rebellious person. And I do not like memorizing stuff out of textbooks. And after spending a week as a bio major, I thought: "No, this is not for me, blank this, I'm going to do something interesting."
Sonia Sultan 6:38
So I arrived to graduate school, at Harvard, without much in the way of training. And you know, that was very difficult in a lot of ways, I had a lot to learn. On the other hand, it left me free in a certain way to start with questions rather than to already have been given answers to them. So you know, I would walk to the bio labs, and you know, there's weeds growing along the sidewalk as I walk, and I would see the same species growing in the sun that looks a certain way. And then a foot away from that, there's another individual of the species growing in a little shade, for some reason under a little tree. And it looks very different. And I thought, like, well, what is that? Because I was learning and had learned, you know, the very basics, you know, Ernst Mayr's book had been my textbook, as an undergrad, and everybody tells you, you know, you've got the genes for sun phenotypes and the genes for shade phenotypes and they're different genes, and that's how you get different phenotypes. And I knew enough to say: "Well, look, these things are a foot away from each other, they're not going to be different populations. What's going on here?" And I would I would ask a couple of my professors. And they would say very kindly, which is the way people spoke to female graduate students in those days, "You're confused. Let me explain this to you." And they would explain it. And I would think: "Well, no," because as I said, rebellious and stubborn.
Sonia Sultan 8:05
So I started reading more and thinking more. And I was both relieved, and also slightly disappointed to find that someone had thought about this before me. And that was Anthony Bradshaw, who had published, you know, many years before, a paper on phenotypic plasticity and plants- A beautiful paper, which I actually had the privilege of asking him about, subsequently. And where he told me that, you know, he had written that paper, and no one had touched it for like two decades, no one ever cited it, because people did not know what to make of that concept in the, you know, the heady heights of the modern synthesis and the sense of power that people have gained in the 1960s and 70s, with the idea that once we know the genes, we've got the whole story. So that's a very long answer to your question, I guess.
Sonia Sultan 8:56
So I became interested in how, what's this other thing that's going on? How does it fit our understanding of the evolutionary process and the role of genes? What is the role of the environment? And how come no one ever talks about it? If this is real, is this property of flexibility or plasticity genuinely influences phenotypes? That's gonna affect natural selection? So how does this work? What does this mean for the evolutionary process? Is it less deterministic than everyone is telling me? Is there some play in that process as a result of what these individual organisms are doing?
Cameron Ghalambor 9:35
Yeah, so you've published extensively on plasticity, and I think a lot of your ideas kind of came together in your 2015 book, Organism, Environment and Ecological Development. I have my copy here. And so as I read the book, it was really interesting because I could really see sort of this kind of arc in your thinking, and I could detect sort of some changes maybe in some of your early work versus kind of the more recent work and how you thought about plasticity. And I know this is a bit of an open-ended question. But I'm curious, you know, if as you reflect back on your thinking, are there points along that path that were like pivotal that, you know, you kind of influenced you and made you think about plasticity's role in evolution a little bit more differently?
Sonia Sultan 10:26
Yeah, thank you for that question. When I started, there was an existing framework, which, of course, was from quantitative genetics, which is the idea of genotype-environment interaction. That was a way to sort of plant your feet and start thinking about how the response of an organism to different potential environments could be incorporated into an understanding of natural selection, or let's say the evolutionary process more broadly. So G by E interaction as an idea, first of all, it identifies two categories, the G and the E, they're, they're completely autonomous, completely separate from each other. And by looking at an interaction between those two things, you can conceptualize the environmental response pattern, what's called the norm of reaction, you can conceptualize that as a property of the genotype. In other words, you can say a genotype has evolved to express phenotype A in environment A and phenotype B in environment B. And you can still sort of suck that response pattern inside the genome, you could say it is an internal, self contained program for response. And that is how everybody thought about it, then, I believe, certainly most people, that is how I thought about it.
Sonia Sultan 11:46
So what that does is it lets you use the tools and the models for thinking about sort of conventional evolutionary change. But you substitute for the idea that a gene simply determines a trait, a certain state of a trait, you substitute for that the idea that the gene determines a pattern of response. And when you do that, a lot of problems go away, because you can stay inside the framework of thinking that a gene for that thing, like a gene for that response, is still a sort of autonomous actor, biologically speaking. So that idea the plasticity is probably the genome is something that I assumed, and it made it possible to talk to other evolutionary biologists, essentially, about plasticity, and allowed people to think of plasticity as kind of a special case, it's like, Wow, so that's cool. Some things are plastic, okay. You know, people are okay with that. So we kind of push the boundary and a little bit, a lot of people do working on that study, probably with Steve Stearns, at very end of the 1970s, or really starting with Anthony Bradshaw, you know, people kind of push the boundary and it's like okay, this is acceptable evolutionary biology. We're still in the zone, where we understand this is how it works. The genes are what's important.
Sonia Sultan 13:02
For a good while, I studied in my lab, we studied the response of individual plants, of different genotypes, where we create genetic copies of a set of genotypes, in our case drawn from natural populations, because I'm interested in how these properties evolve in the real world. We would do experiments where we vary the conditions in which individuals develop and see what they do. And it's a way of just kind of looking at what's the repertoire of response, like how far can one genotype go? And that was super fascinating. And I did that for a good while, and then it became clear that another interesting question, was the question whether the different environments of that generation carried over to influence the characteristics of the next generation? In other words, are there inherited effects of these different environmental conditions that these plants were reared in, and that at the time when I started doing it, which was in the very beginning of my postdoc work, very unconventional. I sent a paper for example, I sent the paper to Am Nat. From an experimental and analytical point of view, the paper was fine. There's nothing wrong with this work. It was strong work. And honestly, I think they had no idea what I was doing. And they just, they're just like, "No." I mean, the rejection was like, "No." They didn't even have it reviewed. So I'm like, "Wow, like this, I've hit a nerve here. What? Why? This is just what the plants are doing."
Sonia Sultan 14:32
So anyway, so of course, I love to make people upset. So I love to make them like, you know, cling and freak out and ask new questions. So I dug into this transgenerational stuff, and after a few years, I had a few very good, at the time, undergraduates, one of them now finishing his PhD at Columbia. And they did a little project in the greenhouse where we had plants that we had reopened sun and in shade and they were growing the offspring of these plants to see if they showed any differences. And yes, they did. And we were looking at the data, they had been plotting the data by looking at the average effects of the shade parent environment, versus the sun parent environment, on the seedlings. And I said, "What if we plot this as a normal reaction?" Where we because we had grown the seedlings also in both sun and in shade, what if we plot the seedling norm of reaction that is the seedlings pattern of response for given genotype for both sun and shade. And we plot the norm of reaction when the parent was growing in the sun. And we plot the norm of reaction when the parent was grown in the shade. And, and because this is all inside very, very highly inbred genetic lines, we were looking at a plot where we had a genotype, and two different environments that the individuals had grown in, these are the seedlings. And instead of one norm of reaction, there were two, because the norm of reaction was very strongly influenced by the parents' environment. And I looked at that data plot and I said, well, the norm of reaction is not a property of the genome. There was no other way to look at it.
Sonia Sultan 16:15
Okay, so at that point, what's happening is something different. Previously, the understanding, my understanding was that an individual's response to its environment was something that was built into its genetic heritage, its DNA sequence. That result showed that instead, the norm of reaction, of a given genotype, is itself influenced by something else, in this case by inherited environmental information. So at that point, it started to make more sense to me to think about the expression of a phenotype, the way an individual develops, as an active kind of regulatory process that's happening in that individual, as a result of a bunch of information: the genetic factors that are built into it, the environmental influences it is receiving as it develops in the moment, in the present. And other information it has inherited through could be through, you know, cytoplasmic inheritance from what we call maternal effects, which are when the composition of a seed or an egg is different, depending on the the maternal individual's environment, that's nothing new, we've known about that for, you know, a hundred years. Or what we now know to be epigenetic effects, that is to say, molecular changes to the DNA or to the will say that hereditary material, you know, to the DNA, methylation marks or chromatin changes, packing three 3D changes, small RNAs that are going to turn stuff on and off, all of those factors are inherited, along with the gene sequence, right? And they affect, you know, not what the genes are, but what the genes do. So if you think about an individual, a developing individual, it's got all that information. And it needs to integrate, or let's say it's evolved to integrate that information in some way, presumably, in a way that in the past contributed to its success. So at that point, what you're looking at is the evolution of developmental and regulatory systems that are hugely context dependent, complex, responsive, flexible systems. And that is a different picture, that is simply a different picture than the picture view carrying around of, you know, alleles, big A and a little a.
Marty Martin 18:44
Yeah, would you say, Sonia, so like my charitable read of this, everything that you've said, you know, you're talking about molecular epigeneticists. That's a big part of what my lab does, so I'm fully on board with thinking that way. But would you say that some part of the holding on in traditional evolutionary biology, I mean, it comes from success, right? That has been an incredibly powerful way to understand basic evolutionary biology. We have made unquestionable progress, but we made practical progress too, right? I mean, in terms of our food supply, the way that we do medicine, there's a lot of utility in that mindset. So success begets success. That's probably a big part of the explanation of why we are where we are, not to say that it's a full explanation, as you well articulated. But to what extent do you think that we remain in that same place just because those efforts have been successful?
Sonia Sultan 19:35
Yeah. So part of the answer is, yes, certainly. In evolutionary biology, our understanding of the role of genes is based on insights to genetics, circa 1924. You know, Sir Ronald, yes amazing statistician. Great. You know, I have no problem with what happened a hundred years ago, I think it's fantastic. I respect a lot of those early evolutionary biologists. Thinking about genes as sort of individual determinants of outcomes is a very, very powerful approach for certain systems, particularly systems where the environment is controlled.
Sonia Sultan 20:16
So for example, you mentioned crops, the environment in which, let's say modern engineered versions of crop varieties are super successful and high yield are highly controlled environments where there is a lot of sun, a lot of water, typically through irrigation, and a lot of nitrogen, through fertilizers, chemical fertilizers. The use of fertilizer has increased whatever it is a thousand percent in the last 20 years, creating all kinds of problems. So because we control the environment, then these manipulations can result in an organism that is very successful at making a certain desired phenotype in that environment. Yes.
Sonia Sultan 21:02
The medical side of it, I think, is a very different question. That would be a long conversation in itself.
Marty Martin 21:07
Next time.
Sonia Sultan 21:10
Yeah. So yes, success begets success. And that approach works when the environment is controlled. Whether that understanding of the role of genetic variation is applicable to the natural world, that is to say, to the way organisms actually evolved, and to what is actually going to happen to them out there is a very different question. And in fact, what we're discovering, I think, is that approach the idea, for example, in conservation biology, the idea that conservation genetics, that looking at genetic variation in itself, is all we need to do to be able to predict success and failure of taxa? It's not working. And instead, I think people are much more interested right now, on understanding the plasticity side of it, right? Because right now, things are happening so fast in the way we're in changing the habitats, everything on the planet, and whatever habitat it is, no one's undoing that. So every habitat is changing, right now, very fast, and every organism has to either evolve or blink out. There's no third option. So right now, I think what we're seeing is that people are asking how much plasticity is in these systems to persist long enough to evolve?
Cameron Ghalambor 22:29
So I have two questions. But I think one comment, which is, you know, I'm completely in agreement with both of your comments, but, but I also feel like, you know, part of the success, you know, the foundations of evolutionary biology in population and quantitative genetics, and the mathematical rigor is obviously a bit of a an abstraction, you know, it has to be right, because of the complexity of trying to model things that are highly context dependent. I mean, I think it's, it's interesting, because, you know, at some level, I think Fisher was very dismissive of plasticity. And maybe there are some also, without going into the details of Fisher's political views that maybe went along with that. But you know, I think, Wright and Haldane were keenly aware of the variation and then, you know, there have been these attempts, Waddington, Schmalhausen and, and others, you know, more recently, also Mary Jane West-Eberhard bringing a lot of these ideas of trying to kind of incorporate the complexity. So, I think, it's not an easy problem to solve, but at least being aware of it, I think, is a good one. So yeah, I think that was my one comment.
Cameron Ghalambor 23:48
Why would you say it's not an easy problem to solve? Which problem?
Cameron Ghalambor 23:50
I think, in general, the context dependency, and the dynamics of organisms as complex systems that have a lot of redundancy and are taking in information and transforming it into physiological, developmental, behavioral kinds of responses. Marty and I argue a lot about is that, you know, for me, I think there's a lot of cases where these responses aren't necessarily adaptive, but that also gives an opportunity for selection to act. And so you know, where does adaptive plasticity come from? It's not just necessarily inherent part of, you know, genomes, but it is something that's subject to selection, and so it comes from the variation within populations and, and I think one thing that we just don't really spend enough time talking about is, what is it that maintains the variation?
Sonia Sultan 24:53
One of the things that's been really interesting for me is that as I've been talking about things, in terms of not just interaction but entanglement, that is to say, as I've been grappling with the difficulty of recognizing how complex these processes are, and how complex the causal relationships are, I find myself fortunate to be talking more and more to philosophers of science. Most scientists think philosophy of science is irrelevant to what they do. But for me, I've come up against some things that I don't really have, that I need more language for, as a scientist. And one of those things has to do with how we think about causation.
Sonia Sultan 25:41
So you know, the way the scientific method really is framed, and the way most of our explained explanations are framed, is with the idea that there's a kind of a one way arrow from a cause to a certain effect. That is, for example, the idea behind a knockout strategy for looking at the effects of genes, to go back to this success breeds success, idea, right? And that, of course, is what Dawkins built so dramatically into his view of the world by proposing that everything that every aspect of the phenotype, the beaver building its dam is an extended phenotype coming out of genes, like coming out in the nucleus, big arrow, right? And and it's a huge relief to be able to say, Oh, yes, all of this complexity, all of this context dependent expression, all of that we can trace it all back to a gene, potentially a selfish gene, or perhaps a shy gene or an outgoing gene, or a gene with some other attribute, but in any case, a gene. In fact, of course, that's not the only way effects take effect. There's also like feedbacks, which are arrows that loop back, and even more disturbingly than that, there's reciprocal causation, which is the kind of causation I talked about in that book, Organism and Environment because that is a very hard thing to think about. The environment affects what the organism does and is. The organism affects the environment does and is. And you can't pull them apart, they co construct as Lewington so beautifully said they build each other. So interestingly, a very famous evolutionary biologist told me that Richard Dawkins hates reciprocal causation. Now, in my view, you cannot hate a type of causation. You don't get to do that. You have to look at the world, you have to look at the world, right? And the world is doing that.
Marty Martin 27:48
Well, we just had this conversation a couple of days ago, the episode we released with Kevin Mitchell, he calls it "spiral causation."
Sonia Sultan 27:54
That's nice.
Marty Martin 27:55
That's Kevin's creativity. But I want to come back to plasticity. Because you know Cam alluded to adaptive plasticity, you intimated how adaptive plasticity you know, that exists, shade, sun plants, all of those kinds of things, I think we can understand the adaptive part. But one of the pervasive ideas in that space has been maybe their cost of plasticity, right? This is why everything isn't infinitely plastic. Where do you think the literature now stands on costs? Have we measured them? Do they exist? Have we been too simplistic in the ways that we've looked for them? Or what do you think?
Sonia Sultan 28:34
So I've always had trouble understanding why we should assume there's a cost of plasticity. And here's the thing, everything else people study in evolutionary biology is like an adaptive trait. People have no trouble with adaptive traits. We don't have to say, for example, it's impossible for something to be infinitely adaptive, or there's a cost to being adaptive, we just say it's like, yeah, selection gets to do that. Selection gets to produce an adaptive trait. But plasticity is different. Right? So plasticity, it can't be perfect. Well, of course not. No adaptation is perfect. Obviously, because adaptation isn't perfect because there's all kinds of other constraints. You know only certain things are available to be the factors, you know.
Marty Martin 29:22
Good enough, right? I mean, that was Darwin's idea: survival of the fittest good enough, it doesn't have to be perfect.
Sonia Sultan 29:28
Yeah, you know, evolutionary poker game, you just stay in the game. So I always thought that was very weird, that we had to say, "Well, we know plasticity can't be perfect," you know, because then if plasticity was perfect, what's the rest of that thought? Unless the thought is then then everything is different than we thought it was. Well, okay. Everything is different than we thought. I have no problem with that, because everything is plastic, in the sense that no trait is fixed. Yes, there is time trades changes expression through time, through the lifetime of an organism. Traits change on all different timescales. Some of them change in the scale of like milliseconds, and some of them change over scales years, depending on the organism. And there is no such thing as you know, the fixed trait state, the particular expression of a trait in an individual will be conditioned by all other kinds of things, the body size of the individual, the dietary status of the individual. So, it's like people use a different lens to think about plasticity, as if it's a special case.
Sonia Sultan 30:33
In my work, what became very clear is that plasticity is development. In other words, those two things are the same thing. Plasticity is not a separate attribute. Development is plasticity. That being said, that particular pattern of expression, how much the expression of a trait varies in time in an individual, or how much it varies. In different environmental circumstances, it can be a very small amount of variation, or a very big amount of variation, depending on the trait and the genetic system, and the environmental access or the environmental differences. Yes, I get that sometimes you could say, a particular trait might be more plastic and one individual or in one genotype than another, or one trait might show more plasticity in general than another trait. You know, head number, this hardly any plasticity for that. Because, yeah, those individuals generally don't actually become alive, they end at an earlier point in their development. So okay, I get that. There is variation in the expression of variation, right? But to say the plasticity is a special case of development, I think misses something very fundamental, which is that every trait is influenced in its expression, to some extent, by environmental conditions. And we know that to the very deepest level, because genes are differently expressed in different environments. And even the most mainstream, you know, very, let's say, conventionally-minded evolutionary biologist, there's no disagreeing with this empirically. There is no disagreeing with this point. This is what we did not understand in 1924, or indeed, in 1956. Once we understand that gene expression is environmentally context dependent. Our understanding of phenotypic expression needs to change to reflect that. And that means that development is inherently plastic, to some extent, depending on the organism and the trait.
Cameron Ghalambor 32:44
Yeah, in fact, Sonia, I'm not sure if you remember, many years ago, we met at a conference, and I asked you about the cost of plasticity. I said something very flippantly, like, you know "Well, if there's some costs associated with plasticity, and then you stopped me. And you said, along the lines of kind of what you just said, "Why would you expect there to be a cost?" And I had to really think about it for a while. And I, you know, I was part of a paper back in 2015, with Courtney Murren and Carl Schlichting, and others where we reviewed the literature, and there's very little evidence for these costs. So thank you for putting me on the right path many years ago.
Sonia Sultan 33:24
I remember that conversation because we both just written a very similar paper about like, limits to adaptive norms reaction or something like that. It was like 2007, or 2008. Right?
Sonia Sultan 33:35
Yeah yeah yeah.
Sonia Sultan 33:38
I remember. So yeah. So in any case, since you asked, where the literature stands, is, people have done very careful experiments looking for a cost. You can find a cost to the expression of plasticity, just like you could find, I mean, sometimes you can find some kind of energetic, or materials cost to the expression of any phenotype. You know, you want to build a big shell take something yet to build, it takes stuff you can't use for something else. However, in terms of some kind of a, you know, the idea that there's some kind of a fitness cost to the capacity for plastic expression, no one's ever found evidence of that, to my knowledge, and I think that's because we shouldn't have been looking for it in the first place. So that's an example of a kind of a metaphysical quantity. As scientists, we're supposed to look for stuff. That's real, something we can observe firsthand empirically, rather than something that we imagined to be there. So we don't look for you know, humors anymore in the human body or, you know, we look for the stuff we can measure, basically. And we can't measure that thing because it has a metaphysical quality assumed that it has to be there because if it's not there, we have to think differently.
Marty Martin 35:05
Let's turn to the what we can measure, maybe what we should measure. Because I find it really intriguing... Well "should," you know, and that's going to dictate to everyone who gets to make those decisions. But you wrote this paper recently with Mike Wade, that Cam and I really enjoyed. Took us a little time, I don't want to try to sell on anybody that this is going to be an easy read, devote some time to thinking hard about this, but it's a fantastic paper. It's entitled Niche Construction and The Environmental Term of the Price equation, how natural selection changes when organisms alter their environments. So maybe say a little bit about niche construction, We've danced around it, but we've not really connected it to plasticity. But how does misconstruction a type of plastic response? Or is it a fundamentally different phenomenon than the plasticity we've been talking about?
Sonia Sultan 35:54
So the phrase has a long history, it was devised initially by Richard Lewinton ten decades ago, and more recently was fleshed out in a book by John Odling-Smee, Kevin Laland, and Mark Feldman. And the idea of niche construction is to focus on the ways that organisms change their environments. And like any very general term, there's a number of different takes on it. I will tell you how Mike and I used that idea to generate this model. We wanted to focus on the ways that organisms influence their own environments. That's the most general definition of niche construction. It's the idea that organisms by being in their environments, they affect their environments. The other part of the idea is that because they do that, there's a feedback. Because organisms change their environments, they change their own selective circumstances. And that gives organisms a role in their own evolution. That's kind of the heart of the niche construction idea, and I think that's the reason it's controversial for a lot of people is because it gives organisms that active role in evolution.
Sonia Sultan 37:04
So what we do in the paper is we define the construction in the following way. And many people have written very beautifully about one or the other aspect of niche construction. And we tried to bring them together by including three categories of niche construction trait. The first is the way organisms can choose to be in one environment over another. So for example, in a very hot place, in the middle of a sunny day, a lizard will choose to go in the shadow of a rock instead of on top of the rock. Environments have spatial variation, and organisms, you know, very often choose to locate themselves in a more favorable patch of the environment they're in. Plants do that through dispersal, and by altering the angle of their leaves. So that's one thing is to choose a better environment. Another aspect of niche construction is to actually change the external environment in some measurable way. So for example, if a fish digs a burrow, there's lots of burrowing fish or you know, or a bird builds a nest or a mammal builds a little burrow that's insulated. These are ways of changing the environment to be more favorable in terms of temperature extremes. So those are two well recognized aspects of niche construction. And the third one, which I was very determined to include is plasticity. And here's how I think about plasticity as niche construction because normal people think of those separate categories.
Marty Martin 38:37
That was the first time I'd ever seen it put that way, but it makes sense.
Sonia Sultan 38:40
Yeah, well, so that's the thing, it doesn't make sense. So when organisms respond to an environmental challenge, through adaptive plasticity, the consequence of that is to change the way the organism experiences the environment that's there. So for example, so a plant in a nutrient poor soil, it cannot add nutrients to the soil. But what it can do is increase its allocation to its roots, and make those roots longer and thinner to more efficiently gain access to the nutrients that are there. And by doing that, it effectively experiences an environment where there's more nutrient ions. So or an animal that expresses defense plasticity, it's not taking the predators out of its environment. But by creating some kind of a toxin or building a shell that a predator cannot chew through, it's experiencing that same environment in which there are predators as one that is predator free. So that is a more subtle kind of niche construction. And in my book, I talk about it as experiential niche construction. It means the organism changes its own body in a way that improves the environment it experiences and that's a harder one to think about.
Sonia Sultan 39:58
So in this paper, we wanted to look at all three of those things because those are all ways that organisms in their own lifetimes as individuals, change their environment. And by doing so, if you think at the population level, because this is population genetics, if you want to think about evolutionary change at the population level, change the distribution of environments that the population experiences. And if they do it in a directional way, which means if these are adaptive niche constructing traits, as they generally are, they change the environmental distribution of the population to one in which the frequency of the favorable environment increases. And as a result of that, the mean fitness of the population goes up. So that is a kind of a key insight, which is that organisms change the environmental distribution in a directional way, which is precisely what is not present in standard evolutionary model, starting with Fisher and in the Price equation, where we continue to assume that the second term of the Price equation, which is the term which is structurally there, to allow for evolutionary effect of the environment, that second term is assumed to be zero, because we do not consider the change in the frequency of environments that organisms themselves enact.
Marty Martin 41:21
Okay? And without a chalkboard, it's going to be really different to draw out exactly, I mean, even write out what the Price equation is. But can you say a little bit about how it typically is used. And then just, you know, generally the novelty of what you guys did with this part of the equation, that's usually zero?
Sonia Sultan 41:39
Well, so the Price equation is a way of structuring the elements or the factors and the process of adaptive evolution. Basically, following Fisher to say there's part of the process which results from change in allele frequencies, depending on how much additive genetic variation is in the scenario, that's the first term. And then there's the second term, which is the potential effect of change in the environment, because it is recognized that the environment could also have an impact on fitness. Right? So it's a kind of an update, the Price equation is a kind of an updating of, I would say, Fisher's initial way of thinking about the sources of population change. And both of them focus completely on the genetic variance part, which is the first term of the Price equation. And indeed, one thing I think it's really interesting is that by definition, so one of the things we quote in this paper, a recent review about fundamental evolutionary theories. And in that paper, by David Queller, a very distinguished population genetics thinker, he says, basically, by definition, to be a fundamental evolutionary theorem, it is essential to focus on the first term. Okay, so in other words, we define evolutionary biology then as only considering the additive genetic variation as a source of evolutionary change. Okay, so we by definition, only allow ourselves to look at genetic variation, and in particular, additive genetic variation as a potential engine for selection.
Sonia Sultan 43:23
So we ask the question in this paper, what if we don't do that? What if we include something we all know is there? There's no question empirically, that organisms, plants and animals and other organisms do these things? There's no question. Right? So we know this is going on, what if we figure out a way to include it in an evolutionary model? And so what this model does, which I think is very cool, so what what standard models do starting with Fisher is to hold the second term constant by assuming negligible or even zero, and vary the first term, that is that we vary the additive genetic variants in a system. And we watch how the increase in additive genetic variants causes faster selection and greater selective change. That's how we've always done this.
Sonia Sultan 44:14
In this model, we do the opposite. You know, models or thought experiments, we do the opposite experiment, we hold the additive genetic variants constant, and we vary the environment part. And we do that by introducing what Mike calls the genetic possibility of a niche constructing trait. And we do it for each of those three types of trait. And those models are very, they're very similar. And they show that one thing, which is not surprising, which is that when you allow organisms to do this kind of adaptive niche construction, the population mean fitness increases, as the frequency of the more favorable environment goes up. And that means that selection will favor the traits associated with that. So adaptiveness contracting traits are expected to evolve and then makes sense because they have evolved. So that's good. So far, so good. So that's-
Marty Martin 45:07
Recreating reality. That's a useful model. That's what you shoot for.
Sonia Sultan 45:11
Hallelujah. Yes, that's always nice, right? So that was the expected result. The unexpected result was what happened when we asked the question, if that's happening, if a niche constructing trait is evolving, and it is increasing the frequency of a more favorable environment, how does that affect evolution on the rest of the genome? Because the environmental distribution is changing for that organism, not just for one gene, right? So when we ask that question, this really interesting thing happens, especially when we also recognize that the effect of most genes on fitness shows genotype environment interaction. In other words, the effect of a gene on fitness will not be the same in every environment, there are different possible patterns of variation in fitness impact, depending on the environment, not for all of them. But for, let's say two out of the three main patterns, most prevalent patterns. In fact, when the environmental distribution changes in this way, the additive genetic variation for those traits increases. The additive genetic variation for other traits, not an entrant striking trait increases, because the population is being pushed toward that environmental distribution. And that means that when a niche constructing trait is evolving under selection, because of its impact on the environmental distribution that has higher fitness, it increases the additive genetic variance for traits that will therefore accelerate their own adaptive evolution to the niche constructed environment. That was not expected.
Marty Martin 46:54
So that part, I mean, I can sort of think of two possible mechanisms whereby that happens. So which one do you think or? No, it is? Is it that selection? is sort of changing the allele frequencies? Or is it more than it's releasing latent genetic variation?
Sonia Sultan 47:10
It's both
Marty Martin 47:11
It's both, okay.
Cameron Ghalambor 47:12
Or is it pleiotropic effects?
Marty Martin 47:15
Well, you probably did get into pleiotropy. I can imagine that makes things ridiculous.
Sonia Sultan 47:20
Well, actually, Mike is going to he's hoping to work on a more thorough version of the model are completely flush out the model, including looking at linkage disequilibrium and pleiotropy and so on. So he's, I think he's gonna go in that direction, probably. But so that's what's so cool is that it does both of those things, it increases the additive genetic variance that is present, that is to say that genetic variants that is available to selection, because there is fitness, F by E in natural systems. So that's kind of the revealed part of it, the Cam is mentioning. And once that happens, it increases the selective response. So the evolution of a niche restricting trait acts to accelerate the evolution on other traits. And it increases the narrow sense heritability, because it pushes the environmental frequency away from intermediate frequencies of say, to alternative environmental states towards a higher and higher frequency as the favored environmental state. Just as selection increasingly substitutes, one allele for another selection of an entrance, directing trade increasingly substitutes one environment for another.
Sonia Sultan 48:32
There's an amazing symmetry in the system when you allow it to be there, rather than assuming an unrealistic negligible value for environmental change. So what we discovered in this model, which is super interesting to both of us, is that when you allow for what we know organisms actually do, the evolutionary dynamics of a very simple, big A, little A type population genetic system, are transformed. Here's the reason I am so excited about this result. What that means is that in the very language of simplified, you know, reductionist population genetic theory, in that very language, we have shown that the theory is missing something important. In its own terms, and the terms of additive genetic variance , it is missing something that has a big impact on the outcome.
Cameron Ghalambor 49:30
Yeah, so I'd like to kind of circle back to this environmental term in the price equation, and, and also just, maybe for listeners and, and to, for myself to kind of clarify, I think, one of the reasons, the main reason probably why historically, it's held constant is this assumption, I think, within even like the standard quantitative genetic model. When we think about the environmental effects on phenotypic variants, we think of those as typically being random or directionless. So you had mentioned, for example, the plants that respond differently to sun and shade. And I think the assumption when we look at these models is often that like, if you take a bunch of genotypes, and you distribute them randomly in sun and shade, a lot of the phenotypic variation that you see, that's environmentally induced, is uncorrelated with the underlying genetic variation.
Cameron Ghalambor 50:40
And I think what's been missing in a lot of these models, not just the Price equation, but even like if we think about the Lande Arnold equation is that if we incorporate predictable change in the phenotype as a function of the environment, then you have the environment acting in a dual way. Which is one to generate variation in a particular way. And then it also acts as a source of selection. So that and I think, I think what's critical, then is that there is genetic variation for the plasticity, which is also what G by E is. Not all genotypes in the population, respond to the environment in the same way. So there is this variation there. And that's so interesting to me, because we have been largely obsessed with estimating heritability. And to do that, it's very context dependent. And we say that, you know, we want to keep the environment constant. But we have this other term, the G by E, which is the genetic variation of plasticity, that rarely gets measured, even though we know, organisms experience variation in the wild. So-
Sonia Sultan 51:59
it's like the one sentence in a genetics textbook that says the environment is also important. It's true,
Cameron Ghalambor 52:04
It's true, it's true. But this also, the sort of dual role of the environment, and the strength of selection is another thing that I was trying to wrap my head around when I was reading the paper, which is that, you know, it's fairly intuitive to me that, you know, an allele that increases fitness is going to increase in frequency in the population. That follows. But when an organism either matches its phenotype to its environment, either by moving to a favorable environment or changing that environment, it's essentially reducing the mismatch between the phenotype and the environment. And the more you reduce the mismatch, the weaker selection becomes. And so that was the part of the model that I was struggling with was that if you take these niche construction traits as being adaptive, and helping, you know, the population in these different environmental scenarios, that should actually weaken selection. And that should actually slow the process way down.
Sonia Sultan 53:15
Yes, of course, just as for additive genetic variance, the additive genetic variance gets eaten up by the process of selection it's precisely the same.
Sonia Sultan 53:24
Yeah
Sonia Sultan 53:24
I mean, you're not troubled by that on the other side of the equation.
Cameron Ghalambor 53:27
No, but I was trying to understand there had to be some kind of covariance between the two sides of the equation. And by holding one side constant, I was struggling to see how, you know, holding one side constant could respond, it wasn't clear to me what the linkages were.
Sonia Sultan 53:51
Yeah, so there's an exception in there the magnitude of the overall environmental component compared to the overall magnitude of the additive genetic component? The answer to that is a little bit in the weeds in the sense that that's a quantitative question. That's not a question about the way the model works. Every model has that limit, which is that just a model that has a genotype factor, and a G by E factor has the same thing. So you have to try to reconcile how they both changing and that's kind of an algebraic problem.
Sonia Sultan 54:23
But I will also say, the problem for me is more that this model, we're continuing, we're continuing to talk about genotype environment interaction is of those are two separate things that we can conveniently combine in a very controlled context. But the fact that you cannot take the environment out of the organism and you cannot take the organism out of the environment, that conceptually I think is much more challenging, especially when you recognize that the boundary between those two things changes during the lifetime of organism, during the lifetime of a population, and during the lifetime of a habitat or an ecological community. And the one thing we know about life, and what makes life different than rocks is the change. We know that is real. And we know that change is happening on all kinds of timescales, all the time. And that's what's so distinctive about the processes that go on and living things. They are processes that are constantly like morphing, and in flux, and responding and regenerating and rebuilding, and that's what makes it life. And we have been ignoring that for a hundred years in evolutionary biology. It's the thing that makes us fascinated by organisms. And it's the very thing we have defined out of our models. You know, what's really been exciting to me in this era of molecular biology and molecular epigenetics, and, you know, very sophisticated experimental tools and statistical tools, what's really been exciting is to try to meet the challenge of bringing life back into evolutionary biology and bringing organisms back, not as vehicles for their genes. That's absurd. To bring them back as living systems. And that has been super interesting. I'm not yet bored. After doing this for many years, it's still as interesting as it was on day one.
Marty Martin 56:37
Yeah. I mean, what you're saying about our ability to do more with it than we used to do, that's what makes it really exciting to me, because like, you know, in fairness, Fisher had to invent statistics to do what they were doing.
Sonia Sultan 56:48
Absolutely right.
Marty Martin 56:49
So now we have computers that, you know, he never would have really thought of, there's the kinds of things that we can do that just was not even close to on the table.
Sonia Sultan 56:56
Yes. And we can study gene expression. So, you know, for the first 15 years of my career, the last thing I wanted to talk about was genes, because I did not see gene sequence as teaching me what I needed to know about what was going on. Now-
Marty Martin 57:12
Expression!
Sonia Sultan 57:13
I mean, yes, let me study that, you know, I mean, that's just we're in a whole different world of understanding what genes do and how they do it, and the role they play in these processes. And I think evolutionary biologists should get on board.
Marty Martin 57:31
So I've got to bring in the word that we've yet to say. But it has been the point of many, many, many a discussion between Cam and me. You wrote another paper with Denis Walsh and Armin Moczek on agency. So this is another one of these concepts like niche construction, and early when we started talking plasticity, I think we probably are talking about agency. But what does agency mean to you now? And is this that part of evolutionary biology, biologists doing their thing without life, is agency part of what's been missing?
Sonia Sultan 58:12
That's a surprisingly hard question, I think. But thank you for asking it. So I think one difficulty is that the word carries a lot of baggage. And that baggage is not useful for that term in science, or let's say, in biology. Because if you want to ask a question about the agency of organisms in their own development, and evolution, people are suspicious that you're putting something supernatural in there, or at the very least some kind of, you know, imputing freewill or deliberate intention to, you know, a squid or an alga. And I, for one do not wish to do that. That is not my goal at all. The agency idea, yes, I think the agency idea is to try to shift our emphasis, let's say. So our emphasis in much of biology and certainly in evolutionary biology, our emphasis has been very single mindedly on gene sequence, on DNA sequence. And that has, I think, painted us into a lot of corners, including the corner of what happens now when we want to know, can birds speed up their egg laying if spring arrives earlier, right? And I think there is a version of the concept of agency that can help us there.
Sonia Sultan 59:35
And that's what we tried to talk about in that paper you mentioned, which is to shift our emphasis away from DNA sequence as such, as a self contained explanation for variation. Let's say why some people might get a disease and others not get that disease. That's the goal of GWAS studies is to look from very simple correlations, having a gene, having a phenotype. That is the very conflation of correlation and causation we have been warned not to do. That's what GWAS studies are built to do. And they do it because the sample sizes are often so big that there is immense statistical power to detect those statistical correlations. That's an approach which is based on a very, very simplistic understanding of where variation comes from.
Sonia Sultan 1:00:27
And what we propose in that paper is to bring in an idea of agency biological agency, not human agency, to recognize that there are response systems, let's say in the causal space, between the DNA sequence and the actual physiology, and morphology and structures of the organism. There's a big space in there, where the organism actually takes shape. And if you focus, instead of DNA sequence, if you focus on the processes that are influenced by environmental conditions, let's say, and by inherited influences of previous environments, through epigenetics, and through maternal effects. And if you include how the system that is operating in that causal space, how that system works. So for example, certain mutations may change a regulatory pathway and many others will not change that pathway, because the pathway has its own resilience. If you focus on those resilient and flexible systems that are the ones that, you know, through which an organism actually takes shape and develops its physiology, then you're focusing on something where there's a lot more action, there's a lot more that is genuinely relevant to the outcome. And you will learn more, you will expand your explanatory framework for understanding the causes of the outcomes.
Sonia Sultan 1:02:04
So for example, if we take on board that complex set of influences that is part of the regulatory and developmental system, then we can ask questions that we're now starting to ask like: If someone has certain genes, how does their exposure to estrogen affect their chance of developing breast cancer? Or if a person's maternal grandfather was in a community with very poor nutrition, how does that affect this person's likelihood to develop diabetes and obesity? And the remarkable thing is when we ask those questions where we are bringing in all of those other complex possible factors, very often the answer is those things are super important. Those environmental and inherited environmental influences that we ignore, when we do something like just look at the DNA sequence, those things are incredibly important, and they are filling in gaps in our understanding that we have been perplexed by. If genes are the way to explain human traits, why do these immense GWAS studies explain 6% of the variation in body weight? That's nuts., The answer is that's not where we should be looking. So the point is maybe to look elsewhere and agency is a kind of a shorthand for focusing on this part of biology, that sort of response system part of biology.
Cameron Ghalambor 1:03:36
Yeah. So Marty and I argue a lot about agency and part of it is my uncomfortableness with the baggage that you mentioned. But also, you know, I guess I have two issues where I struggle, because on one hand, I completely agree and recognize biological agency occurs, you know. There's no doubt that living systems develop, function, respond to their environments, in ways that are adaptive. But my skepticism comes partly from how I've seen agency depicted and for me, when I think about a system that exhibits agential behavior, for example, to me, that is a very natural outcome of natural selection. That selection acts on the integration of development and physiology and behavior and everything. And, you know, over time, any organism or system that didn't act in an agential way, has probably long gone extinct, because that's not a very good strategy. But I've seen agency depicted also as something that seems to occur outside of the realm of natural selection, that it's somehow something different a property of complex system that sort of can occur or outside of being a product of selection. And I see this I think in kind of the more philosophical literature of how agency is left out. How do you see agency in relation to selection?
Sonia Sultan 1:05:13
So, for me, everything in organisms is in organisms because it evolved, period. Selection is just arithmetic. There is no getting away from selection. Selection is always happening. The question for me is not whether selection explains what we got, it's what explains the stuff selection sees, that selection acts on. So you know, I totally agree with you. I am a biologist, I do not bring any metaphysics to what I do, and to my understanding of organisms. And everything to me that we see in organisms is there simply because of past evolution, selection, and the other stuff that's also in there because of evolution, right? So I have no problem with that. And I, I agree that the word is loaded in an unfortunate way. But I, I use it only in this very specific sense. And what I asked people is give us a better word. And it's very hard. So I went to a wonderful workshop organized by Sir Pat, the much missed Sir Pat Bateson, 20 years ago now. And the name of this workshop was "The Active Role of the Organism in its own Evolution." It's too long, man, it's too long. So I agree with you, the word is a mess, the term means a million different things to a million different people.
Sonia Sultan 1:06:42
Some years ago, Evelyn Fox Keller, recently deceased, came to Wesleyan where I teach. And she gathered, she had just written a beautiful book called The Century of the Gene, which a lot of people hate, which I personally love. And she gathered together in a room, she invited all the people in my university, in molecular biology and physics and biology and neuroscience, and all the people who were interested to come to this conversation about genes. And she asked each person, each of us to define the word. And after, you know, 15 people, everyone's looking at each other, like, "Oh, my god," because every person in the room defined it differently. And finally, one of the molecular biologists is getting really, really, really uncomfortable says: "We don't need that same definition. We all know what we mean by it." I don't really want to have to say that because we don't all know what I mean by it, in this case. But it is true that we work as biologists with all kinds of terms that we define differently and understand differently, and that pluralism might be necessary. I don't know if it's okay, but it might be inevitable.
Cameron Ghalambor 1:08:02
Okay, so I have to, but I have to ask the follow up question, because in order for a term, even if we define it differently, to be useful, you know, we have to measure something. And, you know, one reason why I really love the paper that you and Mike Wade wrote is because, you know, there has been this debate and divide about, you know, do we need an extended evolutionary synthesis or not? And a lot of the arguments that niche construction has been left out, or plasticity has been left out, I think, have been largely philosophical. And I haven't seen many attempts to say, well, here's how you incorporate it. And so I thought that, you know, the modification of the Price equation is exactly what we want. Because that sort of says, you know, there doesn't need to be a divide, we all can use the same sort of model. But when it comes to agency, what I want to know is, you know, if I'm studying fish, or insects or birds, what do I measure?
Sonia Sultan 1:09:06
One thing that we do in that paper, which could be useful is, you know, we very clearly articulate these three different aspects of agency. The first two can certainly be studied experimentally. So, for example, you can set up experimental evolution of populations. And in one population, you allow individuals, let's say you're working with lizards, you allow them to choose their basking spots freely, which they will do to regulate their temperature. So you've got you know, I don't know what heat lamps or something in your lab setup or mesocosm of some kind. And then in another population, you don't let them do that. So you can ask the question in terms of the fitness of the individuals and therefore of the population, what happens when you do not let them niche construct, you do not let them choose their habitat patch, and what happens when you do, Probably lizards are not the best way to do this, I don't know, I'll think of a better system.
Marty Martin 1:10:03
Fruit flies.
Sonia Sultan 1:10:05
Okay.
Marty Martin 1:10:07
Well I mean it on some level, it sounds like you're talking about Martha Munoz and Ray Huey and other people that have studied the Bogert effect. Right. I mean, it wasn't an accident that I said drosophila here.
Sonia Sultan 1:10:17
Ray Huey. Yes, yes, absolutely. Right. I mean, honestly, I think in some ways, there's a great deal of work that has been done. Not in these terms, but that we can use. So you know, when I wrote that book, I cited something like eleven hundred data papers, not by me, this is all work by other people that is presented in a completely different framework. But if you know, if you look at it in this way, you can get information about this stuff. So there's all kinds of wonderful evolutionary ecology and all kinds of work out there to use, where people may have done these things without thinking that that's what they were doing. The other thing you could definitely do is study cases where organisms, the effect on the external environment, then feeds back and you know, famously, the people who studied what they call eco-evolutionary feedbacks are, in fact, studying niche construction.
Cameron Ghalambor 1:11:13
Yeah, when we talk about niche construction, we often assume, you know, our minds usually go to, you know, the beaver building the dam, which changes its environment. But in many cases, and getting even back to Fisher and the fundamental theorem, one big way in which populations impact the environment is by increasing a number and increasing in density. And density dependence is sometimes a pretty nasty consequence of how organisms change their environment. So it's not always the case, that modification of the habitat and changing of the habitat that organisms are in is always beneficial, sometimes it can have, you know, negative consequences.
Sonia Sultan 1:12:04
Absolutely, it will always have that negative consequence of resource depletion. And then that negative consequence will have other consequences, for example, on other co-occurring species that will come back to the target species in you know, either positive or negative way. I mean, this is what's so difficult, I think, is that once you allow for these sort of external effects, once you bring in the engagement of the organism with its environment, you open up a whole bunch of more diffuse causal processes, both going out and coming back in. And there are, of course, positive and negative effects. I mean, I think you're very right to point out that there is no assumption that every effect is adaptive. No, I mean, there should never be that assumption for any, for any aspect of evolution, right.
Sonia Sultan 1:12:54
And, you know, one of the things that I think is confined our thinking in the past is to try to separate adaptations from constraints. In my view, everything I've evolved is always both, it's always constrained. And it is also always adaptive, and so far as selection, you can't get away from selection. We've tried to decompose biology, I mean, naturally enough, we try to make things simple for ourselves. We've tried to decompose biological factors for a long time, including separating genes and environment, separating out these different categories as if they are completely autonomous. And you know, at this point, the time has come to recognize that they are engaged with each other in very complex ways that do spiral, as Marty said before, they do spiral through time because there's feedbacks over time, and that is the spiral. So it's very much harder to think about that process. And I don't know exactly what tools we will ultimately require to do that.
Sonia Sultan 1:13:57
So one of the words that I now find myself having to use to discuss the way that genes and environment jointly inform phenotypic expression is to replace the word interaction with the word "entanglement." And the entanglement reflects the multiple generations of influence that all come in, and the fact that those environmental influences are genetically conditioned. In other words, those two factors, environment and the gene, are not independent at any point, because the impact of the environment is always genetically contingent, and the impact of the genes are always environmental and contingent, and there are many, many layers of that contingent causation packed into each of us as individuals, and that is not simple G by E. And I don't know how we will turn that more complicated and more true recognition about biological causation, I don't know exactly how we're going to turn it into the kind of models, we like to have to make predictions. What I do know is that we better start studying what organisms are doing in their environments, because they are in trouble. And we need to start to figure some of this stuff out. And just focusing on DNA sequence is not going to get us there.
Marty Martin 1:15:32
Yeah, agreed. And you know, one thing we should mention, we do have the rise of AI is coming along. So we can use artificial life potentially, or intelligence, at least to help with the real thing, if they don't destroy us first. And maybe they can help us solve what it means to be alive. Thank you so much for your time we've been going on for a long time. Well, we always like to do with our guests is just give you space to say anything or a raise any point that we didn't prompt you to do. So what else would you like to say?
Sonia Sultan 1:15:59
One thing I do want to say is that one change I have made, is I don't talk about plasticity as a field or as an area of interest anymore. I try to use the term eco devo, which is ecological development. And my goal in doing that is to emphasize that development is plasticity. And that development always reflects the environment. The thing I don't like about the term plasticity is that it does define a special case or a separate category. And I actually think that's wrong. I think what is accurate, what is biologically accurate, is to simply say that development is inherently plastic, because that's what we know now about how things work.
Cameron Ghalambor 1:16:44
But what about physiology?
Marty Martin 1:16:46
Yeah, well, that's what I would say, I call it physiology.
Sonia Sultan 1:16:49
Agreed. So I didn't make up the term. It came from a developmental biologist, Scott Gilbert. But I've talked to Mary Jane West-Eberhard about it, and she uses the term development to mean everything about the expression of biology, everything about the expression of the phenotype. And if you look in the Oxford English Dictionary, there are two ways to use that term. One of them means development in the narrower sense, like development of structures, and the other one means it's like this is what comes.
Marty Martin 1:17:21
It's like the realization through time, no matter what the trait is.
Sonia Sultan 1:17:24
Yes. And that's how she uses the term development. And that is how Scott used it, I believe, in Eco Devo. So I'm okay with it. But I get you that there has to be that clarification.
Marty Martin 1:17:35
Yeah. And we left out behavior. When we say physiology, we leave out behavior or pick some other organism or phenomenon, other people are unhappy. So yeah, development is nice for that reason.
Sonia Sultan 1:17:43
Yeah, exactly. And then, but I think it's wise, you know, if you're doing something in writing, I think that should be sent explicitly. It's not just about like morphology and structure. The other thing is about this idea. Yes, you're absolutely right, Cam that the point of coming up with anything new and science should be to provide some ideas for doing different experiments, and measuring different things and thinking about things differently. All of those. I do believe that an agency perspective as articulated in that paper does that. It points to some things that could be studied. But one of the things that is always part of what you mentioned about whether or not to extend the evolutionary synthesis is this people who don't want to extend the synthesis or who are completely satisfied with what we have, or what we've had, what we have had for the last 50 years, say we already have all this stuff we already know about this, we know about plasticity, we know about, you know, extra genetic inheritance, whatever you want to call it, we've got a covered. So this is unnecessary, it is unnecessary to change anything.
Sonia Sultan 1:18:55
And the other people argue, and I'm one of the people who argue this, that there's a difference to what is made central to our explanations and what is made peripheral. And the only thing that is central in classical evolutionary approaches is the so-called, you know, the genetical basis of natural selection. The only thing that is central is the DNA sequence. And it does change you as a scientist, and it does change your experiments. When you say I'm going to make this peripheral phenomenon, like environmental inheritance, or plastic expression of phenotypes. I'm going to push that thing to the center, and I'm going to ask the question, if this is important, what would that mean? Or how can I determine if this thing is important? And when that is done, when people focus their experimental approaches. So standard experimental approaches prevent us from seeing those things because for example, if you do everything in one controlled environment, you cannot see a role of the environment, whether it's there or not, you will never know.
Sonia Sultan 1:20:06
So if instead you say, I'm going to ask the question in such a way that we'll see that thing, if it's there, I'm going to bring environmental variation into my experimental design that changes what you do when you ask different questions, you have the possibility of learning different stuff. And Dawkins himself knows that because if you read the introduction, as I'm sure you both have, to The Extended Phenotype, he says, This is not something you measure. This is a way of thinking. And we are scientists, and we have ways of thinking. And those ways of thinking determine what we expect to find, and therefore how we design our experiments and what questions we ask. And if we change our ways of thinking, we will ask some new questions and do some new experiments. And that will be good.
Marty Martin 1:20:53
Yes, totally agreed. What a wonderful way to wrap up. Thank you so much on you. This has really been fascinating. We really appreciate it.
Marty Martin 1:21:00
Really great to talk to you.
Sonia Sultan 1:21:02
My pleasure.
Marty Martin 1:21:20
Thanks for listening. If you like what you hear, let us know via X, Facebook, Instagram, or just leave a review wherever you get your podcasts. And if you don't, we'd love to know that too. Write to us at info at bigbiology.org/
Cameron Ghalambor 1:21:31
Thanks to Steve Lane who manages the website and Molly Magid for producing the episode.
Marty Martin 1:21:36
Thanks to Dayna De :a Cruz for her amazing social media work. And Keating Shahmehri produces our fantastic cover art.
Cameron Ghalambor 1:21:41
Thanks also to the College of Public Health at the University of South Florida and the National Science Foundation for support.
Marty Martin 1:21:47
Music on the episode is from Podington Bear and Tieren Costello.