Ep 132: Feel the heat (with Jancitha Ellers, Enrico Rezende, & Wilco Verberk)

Ep 132: Feel the heat (with Jancitha Ellers, Enrico Rezende, & Wilco Verberk)

Big Biology

Download

Jancitha Ellers

Wilco Verberk

Enrico Rezende

How do scientists measure thermal tolerance and predict organismal responses in the wild? What kind of other data is needed to make predictive models better at helping us understand species responses to high temperatures?

In this special episode, roving podcaster Cameron Ghalambor went on the road to the University of Granada in Spain where he spoke about his own research in the symposium, Predictive Ecology in a Warming World. While there, Cam was inspired to get a few of the other experts into a room to talk about the broader field of predictive ecology, and this episode is the result. Guests on the episode include Jancitha Ellers, Professor at Vrije University of Amsterdam, Enrico Rezende, Associate Professor at Pontifical Catholic University of Chile, and Wilco Verberk, Associate Professor at Radboud University. Cam and colleagues discuss the methods and tools they use to measure heat tolerance in insects, fish, and other ectothermic animals and how collecting and sharing trait data is important to inform and implement predictive models.

Cover art: Keating Shahmehri. 

• Episode Transcript

  Cameron Ghalambor  0:00  

  Music.

  Marty Martin  0:05  

  So Cam, what's exciting these days in your part of the world?

  Cameron Ghalambor  0:08  

   Well, the Nordic world ski championship just ended here in Trondheim. It was kind of like hosting the Winter Olympics. Tens of thousands of people were here to watch cross country skiing, ski jumping. It was very cool.

  Marty Martin  0:21  

  Ah, from here in warm Florida, where everything is about spring training baseball, I'm imagining the snow covered landscape with people skiing through the forest. But, you know, I can't help but envision the opening of that old TV show, the wild world of sports, where the ski jump guy crashes and the announcer refers to the agony of defeat.

  Cameron Ghalambor  0:40  

  I actually think of the same thing every time I see a ski jump. However, the snow covered landscape, that's not very accurate.

  Marty Martin  0:49  

  What do you mean? Don't you come from the land of ice and snow, from the midnight sun, where the hot springs flow?

  Cameron Ghalambor  0:55  

  Ah, Wild World of Sports and Led Zeppelin references. I love it. I think we're showing our age. Alright, well, that snowy landscape may have been true in the past, but things are becoming much less predictable. During the ski championships, it rained most days and made for miserable conditions. In fact, the Nordic center where the competition took place has been piling up snow for the last year so it could use it during the competition.

  Marty Martin  1:21  

  Wait, that's crazy, piling up snow for a year?

  Cameron Ghalambor  1:24  

  Yeah, there was this huge mound that was covered with sawdust to prevent it from melting. I assume they used some of that old snow to cover bare patches and keep the competitors from having to ski on grass and rocks.

  Marty Martin  1:36  

  Okay, that sounds like good planning. I'm sure some climate model was also used that could give some predictions on what the conditions would be that time of year.

  Cameron Ghalambor  1:43  

   Yes, I'm totally envious of these kinds of climate models, because you can run them forwards and backwards and parameterize them as many ways as you want, and see how good they are and how they change under different emission scenarios.

  Marty Martin  1:57  

  For example, in a recent paper in Science Advances, led by Jared Wolfe and colleagues, they developed a model that found that for every one degree Celsius increase in temperature, there was a 63% decrease in survival of Amazonian birds during the dry season. So this means that over the past 50 years, bird populations in undisturbed sites in the tropics had been in major decline.

  Cameron Ghalambor  2:20  

  And as an introduction to today's episode, I was recently at a conference in Spain called Predictive Ecology for a Warming World, and the focus of this meeting was exactly that problem. What kinds of tools and methods do we have available to us to make predictions about what the biological impacts of climate warming will be?

  Marty Martin  2:38  

  Today's episode will be a field report from this conference and a discussion with some of the other biologists working on this difficult question. Our roving science communicator and co host Cam Gallambor is on assignment for this episode where he speaks with professors, Jacintha Ellers from Free University of Amsterdam in the Netherlands, Enrico Rezende from the Pontifical Catholic University of Chile, and Wilco Verberk from Radboud University, also in the Netherlands.

  Cameron Ghalambor  3:03  

  We like to say that Big Biology gives listeners the opportunity to listen in on conversations with biologists asking big questions and doing cutting edge research. And in today's episode, that conversation is happening in a small dorm room. So we apologize for a few glitches in the audio track.

  Marty Martin  3:21  

  And we should say at the outset that we're only getting a small window into the diverse approaches biologists are using to study these topics for a global biogeochemical perspective over deep time. Check out Episode 126 with Steven Porder.

  Cameron Ghalambor  3:37  

  Or from a paleo biology perspective, episode 114 with Geerat Vermeij.

  Marty Martin  3:43  

  A lot of Cam's conversation with Jacintha, Enrico and Wilco dealt with more methodological challenges, like, what's the best way to measure the heat tolerance of an organism? How does temperature interact with other forms of stress? What mechanisms are involved? How do species interactions alter the outcome, and what other traits besides heat tolerance should we be measuring?

  Cameron Ghalambor  4:02  

  If you're interested in these questions from a research perspective, we hope you pick up some useful insights from people using these approaches,

  Marty Martin  4:09  

  Or you can just be a fly on the wall eavesdropping on the conversation like I was in this episode. And

  Cameron Ghalambor  4:14  

  If you don't like this type of field reporting, all the blame falls on me. But if you'd like to hear more of these types of conversations let us know. Marty and I are always debating whether we should try to record conversations with the people we interact with when we go to conferences.

  Marty Martin  4:28  

  And speaking of hearing from you, we want to say a big thank you to everyone who's become a paid subscriber to Big Biology.

  Cameron Ghalambor  4:36  

  While we still face financial challenges in keeping the podcast going, we've seen the number of subscribers steadily grow over the past few months, and that is super encouraging. If we stay on this trajectory and keep growing our subscriptions, we'll be able to continue making Big Biology into the future.

  Marty Martin  4:53  

  So if you haven't subscribed but you enjoy the podcast, please consider becoming a subscriber at bigbiology.substack.com for the cost of a cappuccino per month, you can help us in our mission to share with you the biggest ideas in biology. And if you can't afford a subscription, just write to us and let us know. We'll give you one for free. I'm Cameron galenbour And I'm Marty Martin,

  Cameron Ghalambor  5:13  

  And you're listening to Big Biology.

  Cameron Ghalambor  5:26  

  Enrico, Jacintha, and Wilco, thanks so much for joining us today on Big Biology.

  Wilco Verberk  5:31  

  Yeah, great. Great to be here. Looking forward to the chat.

  Enrico Rezende  5:35  

  Thank you. Cameron. It's a pleasure to be here.

  Jacintha Ellers  5:37  

  Yeah, I'm looking forward to our discussion.

  Cameron Ghalambor  5:39  

  Great. So we're all here in Granada, Spain for an international symposium on the topic of Predictive Ecology in a Warming World. This has been a really great meeting so far. And we usually don't do these kinds of live events, but this is an attempt to try to share with our audience some of the topics that have been covered at the meeting by the speakers, and also to talk about some of the challenges of making ecology actually a more predictive science. So I have the pleasure of knowing all three of you. But can you give a little each, give a little brief introduction about some of your research interests, and you know what brought you to this conference here in Spain?

  Wilco Verberk  6:21  

  Sure, I'll go ahead and start. So I'm a freshwater ecologist, and I'm really interested in how traits work, environmental factors like temperature and oxygen, body size, and I guess my main reason for studying this is that species are not distributed randomly in nature, but they occur in regular, predictable patterns, at least most of the time. And my interest is really in trying to explain all of these patterns for all of these species. And that's, well, we need, we need tools to get there. So that's what brought me here.

  Cameron Ghalambor  6:56  

  Great, Jacintha?

  Jacintha Ellers  6:57  

  Yeah, I'm more of an evolutionary ecologist. So as an evolutionary biologist, I of course look at traits, because traits are the things that change in evolution, right? But traits are also, I think, the linking pin between ecology and evolution, because they determine how species interact with their environment. And so when the environment changes, we want to know which traits are actually doing well in the new environment, but also if these traits can adapt to new environments, right? So in this symposium where we're looking at ecology in a changing world, I think the evolutionary perspective is really indispensable.

  Enrico Rezende  7:35  

  Yes, and well, I'm an animal physiologist by formation. And actually I think many instances, I don't necessarily belong in an ecological discussion, because I don't have any explicit formation in ecology as such. I was very interested in understanding how animals work. Eventually, that brought me to evolutionary physiology, so how these physiological systems evolve. And later in my career, I started wondering, if we have like this understanding of physiological knowledge, how can we use it for predictive purposes? And that's how I somehow entered into, you know, ecological predictions. But I don't consider myself an ecologist by any means.

  Cameron Ghalambor  8:09  

  Okay, well, that's good, you know. I think we're going to benefit from having sort of a diversity of perspectives. And I think, you know, one of the goals of this meeting and the challenges that that we face is how we can use ecological, evolutionary physiological knowledge to make better predictions about what will happen in the future. And this is really interesting, because, like climate scientists, they run these big, complex global circulation models, and they can do a pretty good job of predicting how much the earth will warm, you know, given different scenarios of carbon dioxide emissions. And you know, these go into the IPCC reports for climate change, and they are very important for guiding policy. But when it comes to ecology, you know, if we were to ask most people you know at this meeting, you know, can you predict which individual species are going to be going extinct in the next 20 years, or what communities will be most impacted, or what ecosystems you know are going to do, I think you'd be very hard pressed to, you know, get a very concrete answer, like the way the climate scientists you know are able to do it. So this is a really big question, but why is it that ecology doesn't do such a good job of making these predictions?

  Enrico Rezende  9:32  

  Well, I think the first thing is that the ecological systems are highly complex. And perhaps one problem that we have, if we compare like ecology as a science against like physical science, is that we have problems determining the scale, the temporal, the spatial scale that things are happening. And so just imagine, like gravitational theory, with where people would actually try to understand how, you know, a planet should move around. It's very easy to define what a planet is, and you know the motions. And so eventually it's easy to isolate that system and study it. But for us in ecology, we're just at the very beginning of the question, right? So at which level are we studying it? What's the temporal scale of that? It becomes extremely difficult just to pinpoint the question, let alone what the answer might be.

  Jacintha Ellers  10:12  

  Well and on top of that, of course, there are so many different species, each with their own characteristics. So if you compare it to physics, for example, that can make really good, well fitting models. You have a number of different particles. But of course, the number of different species infinitely larger than the number of particles in physics. And worse than that, the species are actually changing by evolution. So also the characteristics of imagine that you would have physics where the characteristics of the particles would change over time, I think then people understand why it's so much more difficult to make ecological predictions.

  Wilco Verberk  10:46  

  Yeah I agree. I guess we do not know all the rules that are governing species, environment relationships, the rules keep changing. And yeah, so physics, they have a much easier job, I guess, than we do.

  Cameron Ghalambor  10:58  

  I know this is, I find that really funny, because, you know, the the physical sciences, you know, sometimes referred to as "the hard sciences", and then, you know, ecology is supposed to be this, like soft science, you know, it gets a bad rap sometimes, but, it's very much more difficult. I think, the orders of magnitude of these kinds of interactions that you were mentioning, of like, different species, they evolve. They're plastic. They interact with each other, and not in you know, I mean, there's some general rules, like a predator eats its prey, but within that there, there's so much more diversity in the way things interact that I find it, yeah, sometimes really overwhelming.

  Wilco Verberk  11:42  

  Yeah. It's like with the planetary movements that Jacintha was talking about, you know, the body, or the mass of the object, you can work out all the other things. Would be nice if we could just measure the body size at the weight of each animal, and then we'd be done. But we have lots and lots of different traits, like foraging, whether it can disperse. And then they also have different stages, so larval stages, adult stages, and they all do different things. So yeah, lots of things.

  Cameron Ghalambor  12:08  

  Given, you know, level of complexity, you know, we can't measure every bit to make things a little bit more tractable, we often tend to focus on traits that we think are really important and in the face of a warming world, let's start with the most obvious one. What the temperature tolerance of a given a given organism? This is a question thermal biologists have been interested in for an extremely long time, and just a few episodes ago, we spoke with Ray Huey about some of the history and applications of thermal biology, and the thermal performance curve and and how it's become a very important tool in trying to think about how species will respond to to a warming world. So, you know, to various degrees, we've been interested in estimating the thermal tolerance of a given organism. So let's talk about some approaches, and because I think our listeners are probably going to be very surprised, it's not as straightforward of a measurement as you might think. So. Yeah, do one of you want to start by just giving a 30,000 foot overview of some of the different methods that have been used?

  Enrico Rezende  13:15  

  Well, I can, I can talk a little bit about my personal experience, because I was never a thermal tolerance person to begin with, and eventually I just got involved in this area. And suddenly, you know, in many discussions, people talk about, you know, as I'm somehow some sort of Bulldog, you know, pushing one particular ideology. But, but it was a very interesting development to understand not only what thermal tolerance is or how it might be quantified, and the difference between what a measurement is, how we define a trait, and so forth, for me, at least, it was a personal growth and to get into more philosophical realms of how we do science.

  Enrico Rezende  13:49  

  So I think, for instance, the reason why I started with this was that people were measuring critical thermal limits. So essentially, they were thinking about temperature tolerance as a particular temperature that animals or organs in general could tolerate. And the problem that we had was that if we try to quantify it in different ways, we would get different results. And so it was a very, you know, at the beginning, it was a very easy, simple methodological question, right? Like, why do we measure it, or try to measure the same trait in two different ways, and we get two different results? And there was some philosophical issues behind that. Some people would argue it's two different traits, but then if it happens to be two or three or multiple different traits, then we cannot validate them, we cannot compare the different measurements. So we got into this big conundrum where the discussion was, we're confounding what we're actually measuring with how we are trying to measure it and eventually try to somehow get to the bottom of this. And it became some sort of philosophical discussion.

  Wilco Verberk  14:45  

  So maybe, maybe good to clarify at this point that how we actually measure these things. So I started measuring these in 2009. I looked up some papers, and I thought, well, this is apparently how you do it. So you basically take an animal, you put it in some kind of a heating environment. So you ramp up the temperature, and at some point, the animal is going to be a bit stressed, and it will keel over, as they sometimes do, and at that point, you just know, down the temperature, and then you say, well, that is it's heat tolerance. But as Enrico pointed out, there's different ways of doing that. You can start at a given temperature, can ramp them up quickly, or you can keep them at a certain temperature and measure the time. And so it is indeed not as straightforward, but I thought it would be good to mention at least how we do these things.

  Jacintha Ellers  15:32  

  Yeah, for me, that was also really a reason to, at some point, not wanting to do these experiments anymore, because if you don't have a unified way of measuring, a unified method, then, yeah, what do the data say that you get right? And so now I think we've moved forward on that part, and we'll probably talk more later about that, but for me, it was a reason to also dive more into the mechanisms, right, look at the physiology underlying these temperature responses. So for example, the phospholipid composition of membranes, right? If you, if you increase the temperature, then entropy within the membranes increases, and so they become more leaky. And what the organism does, it changes the types of phospholipids that are in the membrane to be better able to withstand that high temperature and keep the cell content inside the cell. And so I thought: "Well, maybe if we look at the mechanism, we can understand the sort of underlying, unifying response that would then help us solve these methodological problems." Well, of course, that turned out to be too simplistic, but I think it does contribute, and it does indicate that, yeah, that's a sort of universal methodological, biochemical response.

  Cameron Ghalambor  16:40  

  Yeah I think that's a really good point. Because when we talk about mechanism for like, non-ecologists or non physiologists, when they think about like heat stress killing an organism, you know, it's not necessarily the heat itself that is the cause. You know, the mechanism can be either, as you were mentioning, changes in the cell membrane that makes membrane more leaky and causes a loss of equilibrium. It could be proteins denaturing. It could also be running out of oxygen in aquatic systems. And I don't know maybe Wilco, you could talk a little bit about that, because I know you're-

  Wilco Verberk  17:15  

  Yeah, and indeed, that's what got me into thermal tolerance. So I saw this paper that apparently in the marine world, this was well established. So you heat up the animal, but it's not temperature per se that is killing them. It's actually because at increased temperatures, these animals need more and more oxygen, and at some point, capacity to extract and supply oxygen to their tissues, and ultimately, the mitochondria is going to be limiting. And so I thought, this is a neat theory. I'll just use that, and then I'll develop it in my own little niche of freshwater insects, which too few people care about, that, I have to say. And so we worked on this theory, and it worked beautifully. We got critical thermal tolerances of about 30, 32  degrees. And if you give them more oxygen, they could elevate these tolerances, if you reduce the oxygen levels in the water, they would succumb at lower temperatures. But it also highlights one of the problems, that 30 degrees is not a realistic temperature. These animals will never face those kinds of temperatures. And so that got me interested also in what else is happening. And to some extent, the protocol, the experimental which we measure these that we just briefly touched upon, is not realistic. But if we would do realistic experiments, they wouldn't be finished in half a day or a working day. And so the experiments, the absolute values we get, are maybe not directly relatable to what's happening in the field, and that's also that the gap that we need to bridge, and we talked about it at this conference.

  Cameron Ghalambor  18:47  

  Yeah, so I want to come back to Enrico, because you and Wilco both raised a good point, which is that we often think about the critical thermal limit as a single temperature as a point, and these thermal limits in when we when we measure them, they they are very, very high, often much higher than any temperatures that organisms would normally experience. And yet, we also know from observation in the field that many organisms actually die from heat waves at much, much cooler temperatures. And so the alternative is to move away from maybe thinking about these critical thermal limits as a single point, to thinking about it more as a function, something that is a function of time and the level of exposure. And so maybe Enrico, you could talk a little bit about this concept of the thermal death time and then maybe a little bit out the history of how, how it came about, and how, how you're using it in your work?

  Enrico Rezende  19:49  

  Oh yeah, absolutely. And I think before addressing this question, just to you know, put some things in perspective, it's interesting to think of why we talk about thermal tolerance nowadays so much, right? And seems like obvious that it because global warming is important, and so we're all concerned about temperature as one particular variable, but it's also a place where, let's put it, life history traits and physiological traits, they meet, right? And so you have physiological function and organismal function, but also whatever happens with thermal tolerance translates immediately to something that's really important for demographics, which is mortality. So it's a good meeting point for like physiologists and ecologists and evolutionary biologists.

  Enrico Rezende  20:28  

  But now coming back to the question, essentially, the main problem was that at a given time, for practical purposes, it was very easy to measure or to estimate thermal tolerance, as this, you know, threshold temperature, where below that temperature animals would survive, and above that critical temperature, animals would collapse. And what we realized is that actually theoretical model, the concepts that we worked with, was flawed, because thermal stress is a cumulative effect, and so that means that actually accumulating stress over time and eventually you have to include the temporal component in the equation to make things work, and that's when and just to put things in perspective, what Wilco mentioned that he was observing critical limits that were way too high and that animals didn't experience in the field. One way that you can actually make the connection between what we measure in the lab and field observations was understanding that actually, animals might not be exposed to these acute challenges in nature, but they might be experiencing, you know, lower temperatures, but for a longer period of time. And so we eventually, we came up with this idea. Well, it wasn't our idea, we rescued it from the literature that you have, what we call the thermal death time curves, right? And essentially, what it shows is a predictable relationship between the amount of temperature that a given organism can experience and the amount of time, the duration of that thermal stress, and it falls into a very simple relationship between the temperature and the logarithm of time, which for many of us, it makes life, you know, even though a little bit more complicated, it still remains quite easy and tangible, right? And so in that sense, it's being it's a little bit more complex, but at the very least it became way more realistic.

  Wilco Verberk  22:02  

  Yeah and we do get some credits now from the mathematicians, because we have to understand these logarithms as well. So that's good

  Cameron Ghalambor  22:10  

  And so and just as kind of then summarize, what that means is that at lower temperatures, if they're, you know, only mildly stressful, an organism can, can stay at those temperatures for a very long time, and the accumulation of the thermal stress is sort of minimal, but as temperatures increase, the amount of time that an organism can persist and stay at those temperatures becomes lower and lower, and eventually, at very high temperatures, you can tolerate them for a very short time and you end up dying pretty quickly. Is that capture the relationship?

  Enrico Rezende  22:45  

  Oh yeah, absolutely. I think we can use its basic common sense, right? Essentially, what we've been doing, let's, let's put it wrong, is we've been putting animals in a sauna bath, and eventually would say they were collapsing at very high temperatures. And we'd use these temperatures to infer what might be happening in the field, but at humans, we can tolerate a sauna bath of sometimes 80, 90 degrees for, you know, not that long of a period, but first, you know, a few minutes. And so that would essentially tell us that we could tolerate global warming and these temperatures in the field, which we know it's not true, right? So it's essentially the same problem. And the only thing that actually happened during the past decade was that we were able to formalize that into a more quantitative model, and the model seemed to work quite well,

  Jacintha Ellers  23:27  

  Yeah but it's not only that if we formalize it into a model, I think the real step forward by thinking about these thermal death curves, rather than just the single CT maxes, right the maximum thermal limit, is that we really made one step forward in unifying the field and all these disparate measurements that we couldn't unify, which actually divided the field. We now have a sort of an overarching theory, it's almost like universal law, right? That now connects these disparate measurements. And I think that's the real step forward. It also helps us to predict better how species are doing under warming conditions. But I think the fundamental advance is that we understand better how it works, right? And so as a field, we really moved forward through this.

  Cameron Ghalambor  24:17  

  So if we were sort of using a single point, like a CT Max estimate to come up with a an index of, like, say, vulnerability to warming, we would compare, you know, a single temperature across different species. But if we're then thinking about it more has this sort of linear response, then is the comparison how steep is the slope and what is the kind of the intercept of the slope is that what becomes the unit of comparison across different species?

  Wilco Verberk  24:49  

  Yeah, yeah basically the thermal that time curve gives us a way to convert one unit into another. So if, if the animal can tolerate maybe 35 degrees for an hour, they can actually tolerate. 30 degrees by so many time and we can use the relationship to convert these units. And the other thing is, it helps us bridge this gap between what we measure in the laboratory and the complexity in the field. Because in the field, the temperatures are not constant and they are fluctuating. But with our laboratory stuff, we can parameterize the model. We can work out the mortality at any time scale. And so in the field, we can then have these fluctuating temperatures, and for each of those different bins of temperatures, we can work out the mortality. And mathematically, we can easily accumulate mortality percentages by just multiplying them.

  Cameron Ghalambor  25:36  

  Yeah. So, so then I think that kind of leads into this concept of a three dimensional thermal landscape. And so what does this thermal landscape look like when we try to kind of communicate it to like a policy maker?

  Enrico Rezende  25:53  

  It's fairly simple in conceptual terms, right? Essentially, you have a combination of the temperature that you're experiencing, the amount of time, and then it translates into a probability that you're going to collapse, or a probability that you might survive or tolerate that stress. And so the third dimension, in this case, it's survival probability. And then you obtain that sort of, you know, like that 3D curve, that 3D landscape, where you have the combination of intensity and duration of a stress, and then how it translates into the possibility that you can actually tolerate it.

  Enrico Rezende  26:22  

  But let me just I'd like to take advantage of one thing that Jacintha was mentioning, which I think it's very nice that. Just because of this change, the switch in the way you perceive a problem, one thing that personally was a big shock was just suddenly how well the model would fit the data. And so many times when we were using and analyzing the data with the critical thermal limits. And we had these r squares just to put, you know, like explain variance, you know, a small fraction, sometimes, point 2.3, and we were actually very pleased, because we considered that we were explaining something quite well. By the time we switched to, you know, the thermal death time curves, or the tolerance landscapes, we started having our squares of, you know, in the order of point eight, sometimes point nine or point 95 which actually the interesting thing was that the unexplained variance, it wasn't because of noise in the data, it was actually our own ignorance of how we would conceptualize a given problem and try to make that connection between our conceptual understanding of the world and the empirical data that might support or be against it.

  Cameron Ghalambor  27:18  

  Yeah, that's super exciting. And I mean, I think my own background as an evolutionary ecologist, but, you know, increasingly being attracted to these big problems associated with global change, seeing this transition has been, you know, really exciting. I think it's a really interesting time. But Jacintha you also mentioned mechanism, and when we start to think about the problem of thermal tolerance  as a function of the level of exposure and the amount of time, then, does that give us a little bit more insight into where we should be thinking about the the mechanisms and how heat stress accumulates over time?

  Jacintha Ellers  28:00  

  Yeah, I think we can use that concept that we now know that both the time of exposure and the severity of the stress really interacts to determine survival or not. That is really pointing us in the same direction, because now we can see also, for example, if we apply a lower stress, whether we see the same physiological responses, but perhaps a sort of weakened response, right? And I think it would be really interesting to apply the same sort of thermal death also to a sort of physiological damage model, where we can then see whether we can understand this accumulating damage by seeing what happens at the physiological level. And I actually think that the whole concept of thermal deaths curves is not only applicable to thermal stress, but I'm really inspired now to also try to do it, for example, to desiccation stress, right? There's another stress that we know that's probably going to increase under global warming. And we also see there, if we submit animals to a really severe desiccation, that they succumb really quickly, and if we expose them to just a mild desiccation that they can withstand it much longer. And so I think that same sort of accumulating damage is present there as well. And of course, there may be different physiological mechanisms underlying that, right? So we still have to look into that.

  Cameron Ghalambor  29:26  

  I'm super excited by that as well, as you know, but I don't know if this is more of a practical or a sort of philosophical issue, but you know, like Wilco was talking about oxygen limitation as one mechanism dehydration, interacting with temperature, proteins, cell membranes. There may be a diversity of mechanisms. But in the end, if we're thinking about like making predictions about, you know, consequences for global warming is the is knowing that mechanism critical, or in the end would you say that really what matters is just probability of survival?

  Wilco Verberk  30:04  

  Yeah, I think there's a couple of answers here. So backtracking a little bit to what Jacintha said, stress, any stress, it's stress intensity and stress duration. So any biological problem that animals might face is accumulating over time. We need to include this temporal dimension. And the second thing that you alluded to, do we need to know the mechanism? Yes and no. If, if we could measure for each and every species the thermal tolerance landscape, fine. We don't need to know what is actually driving it. But at some point, in order to make these predictions for when animals are gonna collapse. But we cannot measure each and any every animal. So it's good to know the mechanism to come because then we can maybe extrapolate.

  Wilco Verberk  30:50  

  So, for example, if oxygen is a problem, then we know that the aquatic animals are going to be more vulnerable when water quality is going to be impaired. If desiccation is a problem, we know that under arid conditions, warming might have stronger effects, because it's going to compound this problem of desiccation. So I think it really depends on what you want to do with these kind of models. And if the goal is to predict whether or not an animal is going to survive, then knowing this relationship is enough if we want to extrapolate and get some ideas for other species, then we also need to know the mechanisms.

  Jacintha Ellers  31:25  

  I would even maybe make that a bit stronger, because I think now that we know the model, right, the thermal landscape model, we want to know whether the mechanism can validate and confirm this model. But we don't need mechanism to make the predictions. And I think it's even better, because like Wilco was saying, we can't measure this thermal landscape for every species, right? That would be too much. But since we know the general model now, we can also identify what sort of proxies we can use, easy to measure proxies with which we can that we may be able to apply to many more species, right? And then we're using that proper proxy, because it comes from this understanding model. Using these proper proxies, we can, for example, compare the thermal tolerances of species in a community, right? And that we can then use to make predictions. So I think having this more inclusive model really helps us to not have to measure the mechanisms anymore, and be able to simplify our measurements and thus measure much, many more species.

  Enrico Rezende  32:29  

  I totally agree. I think I remember, at a given time, we're talking at this conference about emergent properties, and a very good example is kinetic energy of molecules and temperature, right? I mean, even the idea of temperature is actually an emergent property. It's a regularity, but ultimately what we have is dynamics of motion of different molecules, and that gives you a given temperature. And so the idea of, for instance, with a thermal tolerance landscape, is that apparently this function, it predicts quite well what might happen in terms of mortality, but that doesn't mean that you have only a single mechanism underlying that. And maybe an interesting question is, is natural selection shaping the interaction between different traits to maintain that particular function? You know, let's put it this way, kind of smooth, right? I mean, it seems like just a regular, linear function. Maybe it's not. Maybe there are break points in different places, and maybe the break points may appear, might appear in different organisms. But so understanding the mechanism, it's definitely good, but at the same time, we need some shortcuts, right to start making predictions from things that you know come from the interaction between proteins and membranes and so forth, up to what we really need, which is demography, survival, tolerance and distribution ranges, for instance, right? And so I remember you telling me that when we talk about emergency emergent properties, the concept underlying it is that they're emergent because we don't understand how they emerge, right? But eventually we might get there as well, but in the meantime, because we need these shortcuts for predictive purposes, let's go ahead and use them.

  Cameron Ghalambor  33:56  

  Yeah, yeah. That's really good. So I think that also, you know, you mentioned natural selection and, and I think we'll, we should talk about, like, how these thermal tolerance curves might evolve. But before we go there, you know, I'm, I'm very interested in phenotypic plasticity. And I'm curious about, you know, how much plasticity, or in maybe physiological terms, the capacity for acclimation or acclimatization can shift these thermal tolerances. We read in the literature a lot that, you know, organisms have great capacity for acclimating to different environments, physiologically and this may be a very important buffer in the face of climate warming. But there have been also many reports that at least for measures like CT max. So there's a very well known paper by Alex Gunderson and John Stillman that found, based on a review of the literature, that there's actually not that much plasticity in CT Max, and  they're kind of fixed. As we move, maybe towards these thinking about the tolerance as a function of time and exposure. Is there evidence that there might be more plasticity than we previously thought there might be?

  Wilco Verberk  35:18  

  Yeah, we did a study recently, and it's very interesting, because if you plot these thermal death time curves for animals acclimated to one temperature or another, we see the curve shift. Obviously, the warm acclimated animals can tolerate these heat stress for longer, but the increase is not the same across the temporal scale. So at these very short temporal scales, the effect of plasticity is actually not that large, and that is actually the range of time where most of the studies that this paper, that you mentioned, by Alex and Jonathon, that's the range of time that they have the most data on. But in ecological terms, we are talking about longer time scales, less extreme stresses, and there we actually see a much bigger effect of acclimation. And so we did some some models predictions that even if we have, according to the literature, only a small benefit of acclimation. If you think about this nonlinear effect of time and temperature on the mortality, this small difference can make a huge difference on these longer time scales. So we had animals that were acclimated to 10 degrees and 20 degrees, and according to the model, the 10 degree animals would all die in a regular summer. And then if we made those predictions for the 20 degrees so the warm acclimated animals, they would all live. So so even this small effect of acclimation was basically the difference between all animals dying or all animals living. So we are heavily underestimating this effect of plasticity, in my opinion.

  Enrico Rezende  36:51  

  I'll play devil's advocate here. Okay, I think because I agree that the experimental results show that maybe we may be underestimating plasticity, but big global patterns such as latitude trends and species turnover tell you that evolution is quite important, right? And so we don't have like these, you know, super plastic or maybe invasive species, or a few of them, but we don't have these super plastic organisms that might be able to tolerate or survive in a tropical environment and also in a temperate environment, unless we might be dealing with something like competition that might be explaining these trends. So I think, on the one hand, we're underestimating the potential, you know, importance of plasticity. Maybe animals are more plastic than what we've been describing before. But on the other hand, in the long term, we've seen that, you know, plasticity doesn't seem to be enough to overcome species differences, right?

  Enrico Rezende  37:39  

  And so we're in this balance right now, that because of climate change and increasing temperatures very rapidly, we must rely on plastic responses, on the one hand, so to ameliorate the potential impacts of these, you know, this increased stress and and the big problems that empirically, we just don't know how, you know, how plastic these traits are and how much these species could actually respond to them. The other thing, for instance, is that, you know, many species are changing distributions, and they're changing also their phenology. So just claiming that you know tolerance plasticity would be enough. Apparently, as far as we can tell from empiric observations, maybe not, but definitely it's a factor that we should consider in the equation.

  Wilco Verberk  38:16  

  Yeah, no, I agree. So plasticity is having bigger effects. But of course, organisms are already using this at the edges of their ranges, so any further increases, they are still gonna suffer from that. Totally agree. I don't see

  Enrico Rezende  38:29  

  They're not mutually exclusive.

  Wilco Verberk  38:32  

  Indeed, yeah.

  Jacintha Ellers  38:33  

  and of course, we should consider that in the field, these species are not just by themselves, right? Often the niches they actually take up in the field are not their fundamental niches. So whether they can or cannot tolerate these conditions, they're their realized niches, and they are determined by competition from other species, displacement by other species. So I think it's really important to also look at these thermal tolerances and the role of plasticity, not only on a single species level, but within a community. If you have in a community, multiple species, right, you have to prey and predators, and the top predators and primary producers, how do their thermal tolerances and their plasticities compare to each other? Right? Because if the predator, for example, I work a lot on the green bean in the Netherlands, we find these top predators that are super plastic, right? You find them when it's hailing and snowing and they're active. But you also find it when it's a sunny day and when we measure the ground temperature, it's about 40 or 50 degrees and they're still active. So they are, I think, super organisms, but if they can survive, but their prey dies after just one heat wave, then, of course, there's screwed too. Y

  Jacintha Ellers  38:43  

  Yeah, that's a that gets back what we were talking about in the beginning, with some of the complexity in predicting, you know what, what might happen under a warming world.

  Enrico Rezende  39:55  

  Yeah, I just wanted to complement that, going outside of plasticity and going back a little bit. To the tolerance landscape. Just a very minor detail, but it has a massive importance, right? Is the fact that suddenly we're talking about survival probabilities. And that means that, you know, maybe temperature is affecting population such a way that it's almost imperceptible on the short scale. So maybe you have, like, a 1% mortality on a daily basis, but then it accumulates over time. So that means that, you know, over months, that population is not sustainable in time, but we won't perceive it if we keep using those type of tolerance estimates where we're thinking about things like 50% mortality, which is actually, you know, LT 50 in the experiments where we have, like, mean, you know, average values. A 50% mortality in a given experiment is absolutely massive, massive, right? And you cannot possibly translate that, that you got in an experiment under very standardized conditions, up to the field where actually the cumulative effects are going to happen over a long, long period of time. And so I think, you know, going to this probabilistic component is very important. And then, of course, plasticity will affect all these things. But I think it was important to, you know, emphasize that, you know, including this probability as a response variable is not a minor detail. It's actually very important if you want to understand how a thermal stress might affect the world,

  Cameron Ghalambor  41:10  

  Yeah so you know when, let's say there's a heat wave that lasts for, you know, let's say seven days, there's a cumulative build up of mortality in the population, but that's also a source of selection. It's a, you know, an episode of natural selection. And from sort of an evolutionary perspective, we would then say that, well, you know, those individuals' genotypes in the population that are least tolerant would be the first ones to sort of disappear from the population, and the survivors, you know, would be sort of a non-random subset that would presumably have slightly higher tolerance. What do we know about the heritability for some of these estimates for heat tolerance? Is there sufficient genetic variation? I know one of my colleagues, Frederik Jutfelt, has been doing artificial selection experiments in zebrafish on CT Max, and they can get a little bit of an evolutionary response, but not that much. And it may be because they're looking at CT Max as opposed to a thermal death time, for example. So, what do we know about heritability or, you know, additive genetic variance for heat tolerance?

  Enrico Rezende  42:23  

  That's a really good question. So one of the reasons why I started working in this field was because people were saying that the type of measurement would affect the heritability estimates, and essentially they were using critical thermal limits. And the point that we're making is, the fastest you use the ramping the higher the phenotypic variance, and supposedly the genetic variance that you had underlying it. But the opposite, the alternative view would be that if you don't ramp things at all, you wouldn't have any variance at all, because temperature is not varying. So that was really a methodological artifact. And so by now, the short answer is, we don't know. It's difficult to know exactly what the heritability of the function would be if we're working with this function, but there's some circumstantial evidence, for instance, that Drosophila, I mean, Ray Huey;s impact you show that, you know, climate change is impacting the genetic composition of the different populations in such a way that, apparently, they're evolving response to, you know, changing thermal environments. And we've seen that a heat wave has changed the genetic composition of Drosophila populations in Spain, for instance, immediately. I mean, right, there were calculations of the genetic polymer physics before and after the heat wave, and they changed according to the predictions that you had these chromosomal inversions that were associated with warm tolerance, and they were selected for during the heat wave. So we know that there is a genetic basis for heat tolerance. We know that it should be, in principle, heritable, but it's very difficult to estimate, on a species to species base, or even at populations, population basis, how heritable it is, and what's the adaptive potential of heat tolerance in that sense.

  Wilco Verberk  42:37  

  Yeah, I've seen this work from Frederik, and it's super interesting. And they don't see a huge effect, but they do see an impact of selection. And to remind you, this is selecting on part of the TDT curve where there's actually not that much variation going on, because we see much more variation at these longer timescales and slightly milder levels of heat stress, and yet you already see things happening. And what's also interesting in that study is that there are correlated things. So the animals that are selected to be more tolerant, they also display faster growth, they have more eggs, they have larger eggs, they even have improved cold tolerance. So, so you're not selecting on one trait, just selecting all the whole suite of trades. And I would just be so interested to see, how does that change the whole TDT curve? So is that also because a TDT curve has a slope and an intercept, and if you're only selecting on, say, the slope and not on the intercept, or the other way around, what is happening to that? Are there correlated responses or not? Well, if I can give a tip to Frederik, maybe try, try, look at the TDT landscape for your selected. I think that would be super interesting.

  Cameron Ghalambor  45:01  

  That might be a good transition point because you brought up all these other traits, you know, if we recognize that not all responses to climate change are only limited to temperature tolerance, and so that's only one aspect of vulnerability. And so maybe we can transition to kind of this, this other topic that's come up during the meeting here, which is what kind of falls under this umbrella term of trait-based approaches. And so can one of you define what, what our trace trait based approaches and and beyond just looking at temperature alone, how we incorporate information on different kinds of traits, and how we can relate that information to making better estimates about prediction, about what happens when environments change?

  Jacintha Ellers  46:04  

  Yeah, trait-based ecology is pretty hot at the moment. And I think the main reason why that is so is that well before we were looking mostly at species identity, right? And there are so many different species you already alluded to, that at start of our conversation, that that is just a dazzling number. And if you then want to make generalizable predictions, it's really difficult, because every species is different. Well, that's one way of looking at it, but you could also look at it the other way. Species share many characteristics, and so these characteristics are what we call traits. And if you have species that have the same traits, you may at least hypothesize that they also respond to environmental change in the same way. So basically, we use traits to be able to say something about the response of species to the environment. And then we, rather than looking at the species identity, we look at their single characteristics, or a whole syndrome of characteristics that may then allow us to generalize their responses to  the environment. And of course, with that, you do an enormous dimension reduction of the complexity of natural diversity.

  Wilco Verberk  47:14  

  So in my courses, I teach also about this, and I make this reference to Animal Farm, and I say, well, all species are equal, but some are more equal than others. The problem is that not all of the students read Animal Farm anymore, but that's a different matter. But it does capture this point, that you can actually combine species that are functionally more equivalent, and then you don't need to look at each and every species, you can just look at a few functional groups.

  Cameron Ghalambor  47:41  

  So, so what would be some of the common traits that would go into a trait-based approach? What would what are we? What are we talking about here? What are the important traits that are shared across species that have been shown to be very informative?

  Jacintha Ellers  47:58  

  Yeah, so I'm an animal ecologist, so I mostly think about animal traits, but you could do the same approach for plants, actually. But thinking about animals, of course, there are several traits that every animal has, right something like a metabolic rate, a development time, a reproductive rate, or fecundity, all of these traits we can measure on all animals, and with that, I think they will have the largest value in reducing the complexity of our whole ecological systems. But then there are other traits that are more specific to animals, right, dispersal ability, well, for some animals that will be walking speed, but others are able to fly. And then, of course, the dispersibility is immediately larger. And so I think we want to look at traits that we know have a sort of relationship with fitness, performance, whatever you want to call it, because that those we call functional traits, but, well, that definition is a bit controversial, but those traits are really important in, of course, predicting and explaining the response to changing environments

  Cameron Ghalambor  48:57  

  And so in reducing the complexity and our ability to compare across species. So with this sort of trait-based approach, then would you group species under larger umbrella terms like these species have high dispersal, these species, and so they're, they're kind of in a category. These are predators, these are herbivores, these are small things, big things. How do you take those individual traits associated with species and then get rid of the species to kind of come up with these, these more general terms?

  Jacintha Ellers  49:36  

  Yeah there are different levels at which you can do it. So doing it at the species level and having categories, that's, of course, a relatively easy way, right? It doesn't require a lot of measurements to do that. But I think actually, we would benefit a lot more from having a more precise way of recording and storing information about species traits. So personally, I think. We should really get traits measured on individuals, right, because we've been talking about, also in the context of the temperature, we've been talking about species traits, but as an evolutionary biologist, I think it's the most important thing that we have is actually individual variation within species, because that is one of the basic requirements for evolution.  And also, if you think about responses to environment, of course, it's not a species as a whole that's responding. It may be that part of the population dies out, but since there is individual variation in traits that determine the response, other individuals may survive, right? So what we really would like to do, in my ideal world, we would have for individual based trait measurements, right? So not categorical, but quantitative trait measurements. And we're working now on a way to store these trait measurements in a huge database so that everybody could then ask their own question, right? And answer their own question to it, and if for your question, it's necessary to really compile these traits values into categories, then you can do that. But if you want to answer your question in a much more detailed way, then we have these individual trait measurements for different species, for different traits, which could really provide us with much more nuanced answers to questions.

  Wilco Verberk  51:23  

  Yeah, I totally agree. We need this trait information. So there's a lot of monitoring schemes, measuring different animals, citizen science approaches. People just like to see animals plants and report that. And all of that data is just taxonomy based. So once we actually know what these species signify, because each species has its own niche, so that the species is there tells you something about the surroundings. So basically, all of these species are like thermometers or indicators that we can use, if only we know what they actually indicate beyond their own presence. And you ask about, what traits do we use? And it really depends on the question. So if you want to know whether there is a fragmentation problem in your landscape, you want to focus on dispersal capacity and see if well connected areas have more animals that are very poor dispersers. But if you want to know something about whether habitats are disturbed or not, you might want to see whether species that have synchronized life cycles, that focus on predictable habitats are there. And there's also approaches whether species are more susceptible to toxic and so at some point, you can just highlight a group, all of that information into sensitive or non-sensitive species, and then you can work out, and basically have a fingerprint, by the species, of whether there's a problem acting in a certain environment, so, so, but, but trades give you that kind of information, but in order to do that, we need trade databases.

  Enrico Rezende  52:52  

  But I'd like to point out that ideally, there's an assumption underlying that, you know, like, essentially, traits should indicate that sort of information ,should provide that sort of information. I think sometimes we just take for granted what we're measuring, because we think by doing it on a standardized fashion, it works out properly, but maybe it's not providing the information that we would like to obtain, right? And so there's always that conceptual leap between what we would like to measure and actually how we do it. And I think a good example would be, for instance, body size. We talk about body size as this abstract thing, but then when it comes to quantify it, we generally put something on a balance, right, and we weigh it. But body mass and body size are two different things. It's just that, you know, in general, body mass is a very good indicator of body size, but maybe in some instances it might not be, right? I mean, the many times people are testing allometric laws, and they say that, you know, they don't behave as predicted. And part of the reason is because there might be a change in shape. And, you know, it's like the changing shape is not being determined by body mass itself. And so I think it's very important, because underlying all that we've been talking about, there is this translation between, you know, concepts and whatever we can actually go and measure. And a lot of what has been historically determined as what we call "traits", in many instances, they've been chosen for logistical purposes, right? I mean, they're simply easy to quantify, but that doesn't mean that they're actually quantifying the things that we would like to quantify, right? I mean, that doesn't mean that they're actually doing a good job. And critical thermal limits is a good example of that, right? I mean, essentially, there was this big conundrum and this big discussion, and ultimately, it's like, because suddenly we believe that, because we couldn't quantify something easily, that's the trait, and no it's a proxy for a trait.

  Cameron Ghalambor  54:35  

  Yeah, so, and you know, we were talking earlier with thermal tolerance, about the importance of plasticity, acclimation, and evolution for these trait-based approaches, are they able to incorporate the capacity for that type of dynamics in the sort of I could imagine, like, like taking a snapshot, a picture of the combination of traits that an individual has. But it how, how context dependent is that? Like if you measure the same individual in a very different community, a very different environment, are those combination of traits relatively stable, or can they change a lot depending on the context in which you measure them?

  Wilco Verberk  55:16  

  Yeah, they can change. But I guess, yeah, we have this conceptual leap, but there's also a hierarchy to traits. So if we think about dispersal capacity, that's the thing we really want to get at. And so what people have been measuring is things like wing lengthen insects or the flight muscles, whether they are there or not. And all of those things combined give you an estimate of dispersal, but especially within aquatic insects, we see also more. So some within a population, you have animals that are capable of dispersing, they are long winged, and they have well developed flight muscles. But at the same time, we also have short winged, or even without wing, wingless animals, which are the same species, and sometimes there's a genetic basis, sometimes we don't know, but it does mean that the traits that are being expressed is actually dependent on the environment. So for instance, with some of these water bugs, you can actually pinpoint the age at which the age of the pond that they're in, because if it's a very recent pond, they will only have long winged morphs because they had to recolonize. But over time, the short wing morph becomes more and more abundant, presumably because there's a trade off. If you don't have to spend energy on dispersal, there's more energy on reproduction. And so they will, over time, accumulate those short wing more so the ratio between long and short wing can then be used as a measure of age.

  Jacintha Ellers  56:42  

  To me, that problem of context dependency, it actually has two aspects, right? A wanted and an unwanted one. And let's start with the unwanted one. We all know that the methodological circumstances under which we do these measurements, yeah, they can really determine the value you get from it, right? And so if we are going to have these trait databases, it's of crucial importance to have a lot of metadata. So metadata that actually describes what your measurement was under, and what conditions it was done, oxygen, temperature, drought conditions, what sort of machines did you use? What was the history of that individual? Was it a male or female? Well, I can go on for a long time so that that's actually the metadata are actually the data about your measurements, right, about your data.

  Wilco Verberk  57:29  

  The context

  Jacintha Ellers  57:30  

  And the context, because we know that it can introduce a lot of variance, and I call that sort of unwanted variance. But that's not entirely true, because we can derive relationships from that, for example, about plasticity, right? If we have all that metadata, we can compare. Well, do we, in general, get different measurements, higher values or lower values if we keep them under cold or warm conditions? And so that can also help you. Well, then we want to take out all that unwanted variation, and we can do that if we have the metadata, and then compare yeah sort of standardized measurements and trait values. And then we come to the second part that we want to understand if regions differ in their environments, do we then see that these trait associations or trait syndromes, do they start to change right, either by species sorting right, some species just can't make it into that environment, or, of course, by natural selection? think

  Enrico Rezende  58:20  

  But it's important to highlight all the time that a trait, or whatever we define as a trait, we define them as scientists, and we measure and quantify them, and it's important to take that into context. And I'll give you an example. As a physiologist, we talk about metabolic rate, but actually we measure things like maximum or resting metabolic rate, because metabolic rate itself as a trait is changing all the time, right? So it's extremely flexible, and the same should be applicable. For instance, when we talk about individual traits, and let's say, you know, body size, it's like, yes, you know, it's like, right now, I weigh 80 kilos, but you know, my body size changed through time since development, so the whole trajectory might determine what the trait looks like at the very end of it, right? And so it's a convenient approximation, in the sense that sometimes we have to look at reality as a snapshot, but in many instances, that snapshot is actually also the result of historical process of things that are happening in time. So even though, on the one hand, it's fundamental to work on the traits, and we don't have other possibilities to do so, it's important also not to forget that actually traits remain expressions of the interaction between, you know, genotypes and environment throughout time. It's always varying through time.

  Wilco Verberk  58:49  

  Yeah and it's basically getting to noise to signal. So of course, if you just measure a trait, and you measure it maybe in different ways, or active or passive metabolism, you get variation, and you want to isolate a signal, especially if you want to put species on the same line. And there's really two ways of going about this. One is having strict standards and enforcing that. Well, we tried that route, but scientists are people too, and they all measure their own organisms in their own different ways. And so I think the alternative is what Jacintha mentioned this to incorporate this variation in ways in which we measured it. So just embrace creation and use all the metadata. And without the metadata, we would never have arrived at the thermal tolerance landscape, because we would only have the CT max values and no time on anything associated with that. So only with the metadata could we actually go back to those dates and say, "Look, time matters. "And if we include time, we get a much clearer signal to the noise.

  Cameron Ghalambor  1:00:27  

  Yeah. So I think that's, that's, that's super helpful. And I'm imagining, you know, if I was a graduate student working on a project, I would find this insight, you know, super, super helpful. So, so there's, there's this kind of like Jacintha mentioned this sort of undesirable and desirable kind of aspects to the to the context in which the traits are measured. But we also talked about that, you know, the individual species, or, you know, in the individuals themselves are embedded in these communities, in these ecosystems, and they're interacting with each other. How do we incorporate that data? Because I'm a little bit a little bit confused on whether the traits give us insight into the species interactions and the functional roles they play in communities, but then those interactions are also potentially changing the trait values that they're part of that kind of environment interaction that the individuals have. So how do we conceptualize that? How do we, how do we deal with the trait based approach in this community ecosystem level? Because that's also something we would want to potentially make predictions about given what we know about the functions that the species or the individual organisms have?

  Enrico Rezende  1:01:48  

  I think there's a, when you have essentially different traits imply that the different organisms or species have some different niches, right? They behave differently within an ecological community. And sometimes this is perhaps, imagine, for example, you know, trophic niche, and we say that some animals are herbivores, and others carnivores, sayings that because they evolved different strategies, and it was a gradual change, but eventually what you have is two different species that have very different diets and behave very differently in the environment. So we could think about the same thing when it comes in a more like gradual differences, right? Animals with more or less tolerance, animals and, you know, move faster, less fast, and so forth. So there is a component of trait-based ecology, in the sense that it allows you to make some predictions, and depending on the trait, the predictions could be better or worse, right? I mean, we know that a tiger is not going to necessarily be, you know, eating lettuce. On the other hand, we may not know exactly, for instance, how variation in body size across different carnivores might impact the way of functioning within a given community. But ultimately, in terms of conceptually, that's a little bit of the goal, right? That's what we hopefully use that knowledge.

  Jacintha Ellers  1:02:52  

  Well, I think for predictive ecology, we really need to think about traits. I think the point has been made a couple of times already that we cannot predict individual species unless we have so much information about this species. Well, that's just not possible for all species, right? That point's made. So we, I think the traits are really important for predictive ecology, but how do we want to do that, right? And so if we know, for example, the responses from our larger trait database, which traits are doing well under certain climatic conditions. That could be one of the projections or the forecast that we could make for a region that's changing in climate. And then we could say: "Look, we particularly expect species with those types of traits to do well or to increase in the communities that we see there." And then we may not be able to identify exactly which species they are, but we do know which characteristics these species have. And I think we can even go one step further than that. In addition to having this, the traits that directly respond to the these conditions. So maybe they're more heat tolerant or more drought tolerant, or something like that, right, which would be directly selected for by the environmental conditions. If those turn out to be, for example, much smaller species, then we could also say something about the effect they have on the ecosystem functioning. Because if we can identify the relationship between these trait, know that if natural selection or species sorting really pulls on one trait that the other traits are sort of following, and they have different effects on the ecosystem function. So I'm just giving this example of drought resistant species. Maybe they are a lot maybe they also have different effect. For example, if there are soil dwelling, decomposition rate of that ecosystem may change, right? And we can predict these things if we know the relationship between the traits. And that's why it's really important to have this information on how traits relate to each other, not think about individuals or species in an isolated fashion, but really in this integrated phenotypes.

  Cameron Ghalambor  1:04:57  

  Yeah. I think that's super helpful. You've both hinted about the importance of databases, and so I have to ask about this share data database.

  Wilco Verberk  1:05:10  

  That's right.

  Cameron Ghalambor  1:05:10  

  Share trait database, yeah, I think both of you, when you gave your talks at the meeting, had advertising for this database with your t-shirt. So can you tell us a little bit about this database that you're curating, and use this opportunity as kind of an advertisement to share with our listeners about what you're doing with this database and what's going on with it?

  Jacintha Ellers  1:05:35  

  Yeah, we are in the process, we've really made our database now more robust and more robust form where it can really incorporate a lot of data. It's now a relational database. I'm not sure whether the listeners will know what it is, but it's a type of database that makes it much more error resistant, right? So it's very flexible and so what we really need now is actually that the people think, yes, I have collected trait data, so I should put it in this database. I realize that that's a big effort for people, right? So we have also worked on the user interface to make that as easy as possible. But it will still well cost time, that's for sure.

  Cameron Ghalambor  1:06:17  

  So there are these metadata: what kind of equipment did you use? What conditions under which you made these measurements, etc. Are those metadata standards freely available? I would imagine that instead of like after the data has been collected, if you were aware of what those metadata would be, before you started collecting the data, you would be able to provide some sort of best practices in terms of, you know, like keeping track of who collected the data, or what the temperature was outside when I was measuring the, you know, metabolic rate, or what equipment I was using. Do you have those sort of available to people

  Wilco Verberk  1:06:57  

  Oh it's all available because it's all open and the idea of standardization and send us a field to adhere to, but ultimately incorporating as much of the metadata is valuable. And the idea that we have in the hopefully not too distant future is that this is a part of publishing science. Publishing the data will be as important as the narrative and the story that we put

  Enrico Rezende  1:07:19  

  It works for Gene Bank, right? They mentioned that, essentially, if you're gonna upload some genetic sequences, you have to upload the sequences to gene bank as part of the process writing. You don't only publish the, you know, the science the publication, but you have to provide the sequences as well.

  Wilco Verberk  1:07:33  

  So wouldn't it be nice if we have that kind of implementation in our way of publishing? And I guess the other aspect of it is the time that you invest as a researcher, once you've done all these things and your paper is finally accepted, and you then have to also do this additional work. So what is the benefit? And one of the main things that we came up with is to get citations and recognition for that work. So now, if you do a meta analysis or a synthesis across, I don't know, a hundred studies, the journals are not going to accept you citing those hundred papers as part of the real paper, because that's just going to take up a lot of space. But that's actually the direction in which we need to go, because then, yeah, what is the currency in science, prestigiousness? Well, we track citations for better or worse, and so if anybody who has contributed primary data, they should get a citation each and every time their work is being used. It's as simple as that.

  Cameron Ghalambor  1:08:35  

  Yeah, I think with maybe people don't fully appreciate because it's a really fundamental shift in the way we think about how we do science. You know, the final product is often the paper, and the interpretation of the results and the raw data themselves are not sort of part of historically the final product. And I think what, what I'm hearing is that, you know, a shift in thinking that you've, you've generated this data set, and yes, it's part of this particular study, but you put in a lot of effort to get those data and that there's value in that beyond your individual study. Especially in, you know, now moving to the time of data science and these big analyses,

  Wilco Verberk  1:09:26  

  Also because of general AI. So I can envision a future where you said, this is the data we collect. Can you write a story about it? And sure, we'll do that. So the value of whatever story you write about your data is going to be less and less, and less, and it's going to be more and more about the type of questions and the type of analysis that we want to do in order to progress in science, and not so much about individual case studies per se, you know.

  Cameron Ghalambor  1:09:54  

  Yeah, and I mean, again, I'm old enough that, you know, I kind of grew up in a time where people were very protective of the data. Like I worked very hard to collect these data. These are my data. I don't want to share with anybody. I've had I've known many people like that, and I I feel like that type two is sort of going away now, because, because of data archiving, because of the standards we have for reproducibility and things like that, are becoming more and more part of, like, the standards of what we expect.

  Wilco Verberk  1:10:23  

  I have the same idea, but maybe I'm biased, or we both are biased, because as we progress in our careers, there's less and less data that we collect ourselves, so we don't feel as attached maybe to the data anymore. But I do see the same movement, and I think it's great.

  Jacintha Ellers  1:10:39  

  Well and people have to, I mean, like you say, the journals require data deposition now, it needs to be open. And so I think we really should work together with repositories such as dryad or so, so the people don't have to do it as an additional thing. But if share trait already provides them with a file with all the metadata there, also your submission to dry, it will be more easy, right? Or perhaps we can, in the end, integrate it. And so I think the whole point is that people should realize how valuable their data are. The data really deserves a better treatment than being put in some sort of cupboard where actually nobody looks at it anymore.

  Cameron Ghalambor  1:11:20  

  Right, right. So one last question. I mean, I love the idea You listed a relatively few number of traits that share trade is focused on. Why the limitation? Or do you see an expansion in the number of traits? Because, I mean, I could, I could imagine so many other traits that people might want to collect and contribute to the database.

  Jacintha Ellers  1:11:45  

  The short answer is, you have to start somewhere, and that's what we did. Like I said, those are universal traits, so then we wouldn't run into trouble that some species would have those, and others not. But we are expanding. So we're working together with existing databases to try to incorporate other data and also with people doing new meta analysis or research networks, so that we can expand the number of trades. Yeah and  the type of traits that we actually put in.

  Wilco Verberk  1:12:13  

  Yeah and the way we set it up, makes that possible, because now it's a relational database, so we have development time. But if you break down development time. It's really just age at stage. So what is the age as an egg, and what is the age when it's hatched, and what is the age when it's an adult? But if you then add the size at those time points, you have growth, so you have a new trait. So the relational structure makes it very easy to add things. And now we are in the process of adding thermal fertility. I can see how we can add water loss rates to accommodate maybe your work. Growth is an easy one to include, so that's not really a limit to what we want to do. It's just a matter of resources and time.

  Enrico Rezende  1:12:53  

  And I think it's important that we have to remember that science is a collective effort, right? And so the fact that people were very protective of their own data. It was a time when collecting the data, analyzing it and getting to conclusions was something that was possible within those confinements, right? But nowadays we want to come to like big scale patterns where not a single scientist can actually do it. And the other thing that we have to recognize is that some people may be very good at analyzing data, and some people may be very good at collecting data. Right, I mean, if you go back to the history of science, for instance, and you go to gravitational theory, Newton wouldn't get like the motion of planets if Kepler didn't describe very, very well how the planets would move in the first place, right? So data, at a given time, it was absolutely fundamental to come up with a theory. It's just that suddenly we trivialize the fact of collecting data, it's, it seemed like a trivial activity, and analyzing it was like more sophisticated and difficult. But that doesn't have to be the case, right? And so I think people should be more generous with their data in general, not only because we're, you know, getting older and, you know, we don't get so personal about it, but it's really understanding your own limitations as an individual, right? It's like none of us can actually do, accomplish all the tests that's necessary for, you know, to produce new knowledge. And so we have to acknowledge and embrace this sort of collective effort. And that implies, on the one hand, a lot of generosity, and the other, the appropriate recognition for, you know, everybody's effort.

  Cameron Ghalambor  1:14:15  

  Well, I think that's a really good sort of philosophical way to kind of wrap it up. So I just really want to thank all three of you for joining me today, here and talking about these really important ideas. We usually end by giving our guests an opportunity to say something that maybe you wanted to talk about, but we didn't have a chance to cover. Is there anything you want to add that we didn't get a chance to talk about?

  Wilco Verberk  1:14:43  

  Well, maybe just following on from what Enrico said. I think we all have pieces to a puzzle. We all have different skills, and I've been in sometimes with situations where there was scientific discussions, where it was more about defending points of view, rather than bridging these. But since we're all intelligent people, and we all have these different pieces to the puzzle. I think the yeah, the first step is to really integrate those to progress towards a predictive field, and I see a great potential for the future. Yeah, cool.

  Jacintha Ellers  1:15:14  

  My last words are really simple. You should all go to sharetrait.org and submit your data. Please do your data is worth it.

  Enrico Rezende  1:15:28  

  I would just like to thank the opportunity, Cam, for having us and just say how nice it's been to have met you guys, and hopefully we'll do it on a regular basis. Thanks so much, Cam.

  Wilco Verberk  1:15:38  

  Yeah, we'll have the second predictive ecology workshop soon, hopefully.

  Enrico Rezende  1:15:42  

  Well, fingers crossed, man.

  Cameron Ghalambor  1:15:43  

  Yeah, okay, well, thank you very much.

  Cameron Ghalambor  1:15:59  

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

  Marty Martin  1:16:13  

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

  Cameron Ghalambor  1:16:17  

  Thanks also to interns, Dayna de la Cruz, Caroline Merriman and Brady Quinn for helping with this episode. Keating Shahmehri produces our awesome cover art.

  Marty Martin  1:16:25  

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

  Cameron Ghalambor  1:16:32  

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