Contributed Papers III: Species Interactions: Mutualism, Competition, and Deception

These interactions, mutualism competition session. And as we all know, beef is interactions underlie much of the living world as we know it. Um, and today we have a great ledger of, um, a variety of talks going from, um, classic pollination mutualistic interactions, all the way to equivalent Canary, arms, race, um, and competition. And so without much ado, I would like to invite our first presenter, Lucas Camacho. Who’s going to talk to us about co-evolutionary exploitive, uh, directions and how they can increase true disparity and modularity in mutualistic networks. I’m going to meet myself now and you’re on Lucas. Okay And thank you for the introduction I will share my screen now is everybody sees my presentation Nice. Um, let me just start here. Okay. So, um, um, thank you everyone for having me here today is a, is a great pleasure to talk to you all. And I’m here to talk a little about my master’s project, which I did in collaboration with several other researchers who helped me a lot in my first, um, academic experience So I’ll, we’ll today. Talk about a brief introduction about what was my master project. Then I will say about, uh, the equivalent model that I use to answer my questions, uh, questions about the frequency of exploitation and co-evolution and the structure of Naturex and exploitation And finally, I will give some, uh, final general conclusions. And I would like to start this talk, talking about this incredible and diverse interaction. That is the metabolisms that you are seeing, the slides, the pollination, the seed dispersal, and the premier conference directions. Of course, there are other several types of metabolisms by here in this talk, I will focus on these three types and we can, uh, material is all about the benefits. So we can describe it as interactions, which both of interacting species are benefited. But of course, in the toilets, it’s not only about benefits, it’s about cost as well. For example, we have the cost of seed and fruit prediction for attract possible seed. Dispersers. We have the cost of produce nectar and polling for attract possible pollinators. And we can think about in another course related to the trait matching, which is the similarity between the tracks traits, media mediating directions, for example, between these two species here, the bill of the hummingbird and the tip of the flower, and maybe this trait metric could lower the potential of all the interactions that these species could have in the community. So we can think this as a treat mentioned as a curse as well, and thinking about discuss, maybe metabolisms are prone to be explored by other species who doesn’t contribute to the balance of the curse and benefits. Uh, we have some examples in here, for example, this bee who not touch the reproductive structures of the flower, she makes a hole in the base of the tube, gets an actor and doesn’t pollinate the flower, or for example, some memos who does not only eat the fruits, but also eats the seeds of the plants or some examples of ants in, in say Karakia, which, uh, do not, um, repeal some hair before it’s efficiently. So we can take these interactions as interaction, which one species of between the benefits of metabolisms without providing benefits in exchange. Uh, we can take this interactions materialisms and exploitative interactions, uh, as a way of, uh, negative and positive effects. For example, that first example of the bee in the flower, we can describe it like this, a positive effect from the plants by in the B and the negative effect, uh, from the B in the plant. And we also know that, uh, there are not a pair of only pairwise interaction. The community is actually, there’s a lot of interaction going on at the same time. And so in different interactions. So putting in a network context, we’ll have something like this, a be parted network with materialism and split the team interactions happening at the same time. Uh, so thinking about that, we can describe positive and negative effects in interactions. And, you know, there’s a lot of interactions going on as we can describe the networks. And also that these interactions will, uh, um, change how species are reproduction and surviving. So we can see that this, these interactions could influence co-evolution the reciprocal evolutionary change between

direction, species and evolution will change traits and how species are interacting in the community. So we have a feedback between the college and the evolution here and networks changing evolution, evolution, co-evolution changing interactions. And think about this. We, my main question is how this explicative interactions that are showed to you will drive the evolutionary process in mortality CIC networks. Um, for that, I will explain a little about the equivalent model that I use to answer some of these questions Uh, this motto, I try to describe a single trait from a species, a population of the species. I, for example, here we have, uh, several individuals with different traits, uh, of species by species. I N I can get the average trait, which I call Z. And this is traits that I, this is the trait that I will describe in my model and how we’ll describe these traits in my motto. Well, uh, this trait, uh, depends on the interactions, the mutualistic and exploitative interactions, metabolisms in blue interpretative and Reggie here. We also have all other Selectric pressures, not directly related to these interactions that I call here. The environmental selection here in black and I also need is the equivalent model. So also taking into account the selective gradient ratings of selection, which is the minor change in traits, uh, changing the fitness, the average fitness of this species, I, and the additional, uh, achieved genetic variance of the species. So we have a model like this. We have our companies of interactions in here and this parameter fee, which, uh, is, uh, takes into account this, uh, additive genetic variance in the, uh, selection gradients And I also have some assumptions related to this model, which how the dynamics of materialism and exploitative interactions happens, or I will show you some of these right now, for example, in mutualisms we have these traits that deal with the hummingbirds and the tube of the flower. And I, here, I have the trait distribution of the community, and I expect that selection will favor trait matching for the revolution of metabolism. So in time I expect that this traits would become very similar, uh, first part active interactions It’s not, uh, it’s kind of, it’s a different, uh, we can consider the same traits in here and now are interaction in an exploitative way and expect that selection favors for the exporter species trait matching. So individuals who are, uh, resembles the, the victim will reproduce and survive more. And, but, but for the victim, we expect strict mismatch So individuals who are very different to the exploiter will survive and reproduce more So selection. We favor this, this type of dynamics, like an arms race dynamics. So we have the student depending on the interaction Okay. Uh, I have this revolution model here, but I also need to know, uh, who, which pieces are interacting in the community. For that I download some empirical networks from the internet. I downloaded eight networks of phentermine. It costs and directions HFC dispersed. So in eight of pollination from these two databases where the flight and I WDB, uh, it’s very, I’m very lucky to have the data available online for free for download So it’s very nice. So now I using my crew evolution model and using these empirical networks of mutualisms. I write a code, a computer program, and this code is available in each hub. If everyone would like to see in criticize me, you will help a lot with this code. I like magic. We’ll have simulations of how this means. Pieces, traits, change, changing time. This is what’s supposed to be a gift, but I think it’s not working how this mean treat means species traits of each species in the community will change due to the coalition, our process of materialism in exploitative interactions. So now I have my framework, then I start to answering my questions just a second, a water. All right Well, my first question is related to the frequency of exploitation and evolution. So we can think in a, in a network, we can have several frequency of exploit achieving directions For example, here we have an axis in the left side. We can see there’s, uh, just a few, but actually just one expert that even correction, all the other ones I’m materialisms. And as we move to this axis, we can see that the frequency are increasing There’s a lot. I believe that every interaction is, uh, right network is split the team interaction

So how these different frequencies are split, the team interaction will affect the equivalent urinary dynamics for that, or we’ll use a numerical simulations. I will use my 24 networks is all over the internet and their explore frequencies of 0.01 to one of exploitative interactions. And I will do 300, 2000 simulations per network in a total of seven, two 70 to 2002, those simulations and what I expect that due to the arms, race and dynamics that I show to you, I expect that in higher frequencies of split interactions, and there will be a higher trig disparity. So they mean distance of traits of species in that community will be higher, would be excused. It would be more different due to that arms race dynamics, but also I have a copilot in software matching happens in here because some, all their metabolism may be happy in the community, or there is that the, the, of the exploiter who steals a selection, favorite street matching for him. So do not expect only treat disparity, but I also expect a higher trait, cluster formation, which are species to raise. There are very similar to each other, like a group formation of traits. And after I did my simulations, I saw that in natural works with higher frequency of split interactions, there is a higher rate disparity for all the tree types of metabolisms tested. But when we saw the group formation, the number of clusters, there is a initial increase of district grocers due to the frequency of spread of team interactions, but there’s also a decrease. So we’re not expecting that. And what I see in my resource, and I think that’s the, the main contribution of this first question, and this may be, could be a possible explanation for low trade matching ecosystems. Uh, Patricia Delano shows in his amazing work that just different city spirits will seniors in here. So maybe what’s happened is not only trait matching due to maternalistic interactions or molding species traits, but also other types of interactions. For example, this negative effect from split the team interactions Also, it’s very, very curious that the same pattern, uh, could emerge due to different process. This incredible work by Lucas Hernandez and his collaborators show that when you link in space with gene flow, we have the same pattern of trait clusters formation. So maybe what’s happened there. There’s one more than one process, not only gene flow connect interactions, but also different interactions in the community are, are maintaining this, this kind of process And in finally, uh, I’m very curious about this linear of spreading giving directions in how we can test this in the wild. Now that we made some of America simulations, maybe it’s time that we can go to the natural communities and try to some way, uh, see how much of this, uh, how much of frequency of this split team interactions happens in natural communities. And how is this translated in the trade disparity of natural communities? Okay. Uh, we will move to my second question Now it’s about the structure of networks and its quotation. So we know that there is, um, two very known basic structures in mutualistic networks, uh, modeler and networks, which are more compartmentalizing in general, generally are less connected, is very common in and splints and networks. And we also have this nest of networks, which are highly connected generalist species, which specialist species interacting with those generators generate lists. So it’s common in pollination in citrus person. And we can think now that there is a company of trade disparity who is increasing three disparity in the simulations. So what happens if maybe these pieces are becoming so different, but so different that they cannot enter interacting anymore, like a treat by here, how the equivalent process of these different types of interaction, which they change the structure of the network. So for that, I will do an America simulations. Like when, again, with my 24 metrics, I will explore the frequencies of 0.01 to one of exploitative interactions again, and now will do my 3000 simulations per network and a total of 72 to 72,000. And what I expect is that, uh, due to this higher tweet disparity from the exploitative interaction squared will be mutualisms I expect that the naturalists

will become more modeler and less net, less nested in higher frequency frequency of the interactions. And after that, that I did my numerical simulations. I see that this pattern of increasing modularity in, in decreasing acid that you see in all three types of metabolisms, uh, in the plant networks, uh, dispatcher was not so strong maybe because this natural gas is already modular, but we can see here that just increasing modularity in decreasing nastiness, this black points are, uh, parallels in waste simulations, which I remove interactions randomly just to compare how is their share the lost of interactions due to evolutionary process and the random loss of interactions We see that there is a changing structure for both the black points and the color red phones, white dots, but, uh, maybe there is a strong, uh, how can I say changing structure when we consider not only, uh, lost of interactions, it randomly by loss of interactions due to trade disparate Lucas, you have four minutes left. Thank you. Um, now, uh, maybe this is a, uh, uh, this, uh, loss of interaction increasing modularity was the rally seat by cylinder as, and his collaborators, which showed that in Taiwanese tic networks where the loss of interactions will also increase the modularity and decrease method of antagonistic interactions So maybe what we see here is a path of mutualistic interactions. That’s works becoming more and more like the type of mystic Naturex. When we increase the split, the interactions in the evolution process happened, changing traits and species lost their interactions. And finally, I’m very curious about what will happen if we do not lose only interactions, but also will lose the species, how the quick station simulations will, uh, how the different questions or simulations will just wrap the next month, how much of the exploitative interactions, where you need to completely develop a network. So I’m very curious about that. So I will make some general conclusions about my, my work with, see that’s where the team interactions generate strict disparity and increase with the decrease of number of species streets, clusters in mutualistic networks. Also, we see that, uh, split the team interactions generates more modern networks, then random we move on through directions And finally, I’d like to thank, uh, all of my collaborators who helped me a lot in this process or the past and present members of the immuno labs. And I was going to pick you up for funding my master’s degree, and the department of ecology at speed, all this stuff over here to all of the Lamar. And thank you for an instant that gives me talking for more than five minutes and adopts. You can talk to me now or just send me an interview. Thanks a lot. Thanks Lucas. For the contesting talk Um, we definitely have a couple of quick minutes for some questions. And if you take that over, just give people a minute. If they’re typing questions either in Slack or chat, um, I’ll just relay Judy’s question. Um, I assume your model used very costly exploiters, but a lot of exploiters have fairly weak effects. Would it be possible to use your model to see if the effect of exploiters on co-evolution increases linearly with severity of exploiter effects? I could imagine alternately that there’s a threshold below. It would be there, there would be no effect and above it, there will be a high effect. Actually, my motto, a considers the effects of split, the team interactions is very similar. So what could be done is that changing how the dynamics of the model, uh, works. So in that, that the, the part of the dynamics in which derives for the teeth interactions and the treat matching mismatch, this could change how this, uh, if the name works in, put some, uh, symmetries, maybe how this interaction restraint, right? So maybe we can lower the effect of the type of split the interaction and see what happened. Of course, yeah. This, uh, assumption of my model they would affect are very similar. Yeah, that’s cool. Um, maybe we have time for one more question. Um, John’s asking, um, can you partition the effects of direct and indirect effects in your networks? Well, I think yes, but I didn’t do it. I am very curious about that too, because I think that, uh, if mutualistic interactions are so important for the

indirect effects, maybe we have a, uh, other force for the other side, maybe what holds these indirect effects in wireless networks, if we’re considering for the team interaction as well is the effect of exploiters. But, uh, I’m very curious about that, but I, I, I didn’t do it. Yes. Okay. Thanks for that, Lucas. Thanks, Casey. And there’s more questions that I’m sure you’ll be able to answer later on in the chat, um, or directly to the people Um, we’ll move on to our next speaker, just keeping an eye on the clock and we’ve got Joshua Mesa. Who’s gonna tell us about the evolution of increased tolerance in the invaded range of the yellow star puzzle. Um, Joshua, if you can share your screen. Yeah, it’s good It’s over to you now. All right. Thank you very much. Okay. So yeah, today I’m talking about the evolution of increased tolerance to and invaded range, yellow star thistle, and, uh, again, I’m miles, and this is work that was conducted in the lab of Katrina DeLuca Shh. Who was my, uh, post-doctoral advisor So, uh, starting out, um, this is with invasive species. And one thing that a couple of things that we know about invasive species are that they are a major component of global change, and they have a wider range of economic and societal effects and impacts, and they could cost us more than $120 billion in damages and in control efforts. Um, and they, and threatened all kinds of different industries, whether that’s agricultural forestry and fishing Uh, and they’re really a major problem and threat to people’s livelihoods, especially in developing countries. Uh, but we also know that they have a large ecological impact as well as they can. Out-compete native species and displace them, which can lead to species extinctions and a loss of biodiversity. And they can completely alter the community structures in which they invade, uh, which can degrade ecosystems and impact all kinds of natural cycles, such as the nutrient energy or water cycles and that area. Uh, so one of the major questions though, is what leads an introduced species or an exotic species to become invasive Uh, and there’s a lot of hypothesis that have been explored to account for invader success, um, and or invasiveness, and, uh, the rapid evolution of an invasive species in its traits is often invoked as a key to a successful invasion. And there’s abundant evidence that, uh, novel species interactions play a role, uh, whether that’s from the loss of competitors or mutualists, and, uh, changes in plant enemies and the herbivores. Uh, and it’s really essential to kind of understand these different factors so we can help facilitate that a successful control, um, of these different kinds of species understanding why they’re becoming evaded face of in these areas. Um, so native and invasive plants interact of course, with a wider range of, uh, different kinds of herbivores And those herbivores within the community can vary greatly. And it’s really often thought that invasive species are able to experience a relaxation of enemy pressure in their environment or in their new novel environment, such as like enemy escape so way they can focus less on plan defenses and more on increasing growth and reproductive potential. Um, however, there’s, there’s some studies that suggest that they are still actually susceptible to their novel generalists herbivores in their new environment So if we look over here from this one, uh, meta analysis, um, we have to see the cursor I’m going to assume that you can, we have plant origin, uh, which is we have native plants and exotic species coming into this, uh, you know, native area. And we have herbivore effect on the Y axis and a negative number here denotes that it’s a negative effect on the plant from an herbivore from these native generalist, uh, species that are chomping on it. So what we see here is that actually exotic plants have a, more of a detrimental effect, um, from these native generalists herbivores, which they’re just now encountering then to the plants that are already residing there, which is kind of interesting because, you know, we were one would expect that relaxation of enemy pressure, but in some instances it looks like there’s actually, uh, an increase

perhaps, or at least, you know, that they’re getting hammered more than native plants So that kind of begs the question what’s going on there and how, what is exactly evolving in these areas and how can they interact with these herbivores that they’re encountering? One of the major, uh, uh, defense traits, uh, for plants is of course, plant resistance. Uh, these are pretty typical what people know about, uh, their traits that prevent reduced tissue damage, and they can be structural traits such as spines and thorns or trashcans like over here Um, but then they can also be a chemical defense traits such as secondary metabolites, like glucose analytes, or tan and something that reduces metabolism or something like that versus toxic to an inset or mammal. Um, and most studies of invasive species in their natural enemies have focused on plant resistance However, um, there’s also so other types of interactions that, uh, plant heart report can have, and this is, or that a plant can have two herbivores and that’s namely point tolerance, which is a regrow strategy. And this is, uh, basically a compensation for any tissue, those lost through damage through bravery. And there’s been recent recognition that plant tolerance may actually play a role in facilitating invasions. Uh, so, but there’s still comparatively a few studies when you look at, you know, compare tolerance to resistance, and we think it could be an important given, uh, there’s so many novel journalists herbivores that are still damaging this plant. So maybe that there’s been an increase in plant tolerance and some of these invasive species allowing them to cope with that increase in damage. And it, this is just an example of, uh, you know, what plant, uh, tolerance or compensation can look like on the left part of this photo. We have a typical plant here. Um, something comes by like a deer or something eats, it jumps all the way down to the base rows, and then it can regrow. And in this case over here, it has more, uh, branching number and kind of has more real estate allowed for flowers and seed production. So in this particular case, this plant actually benefited from being, um, so that kind of thing, that tolerance can, it kind of exist on a continuum, and this is some data for a rabbit. Opsis some different eco types showing that, um, over here on the left hand side, on the X axis, we have, you know, three eco types here in these gray bars, and we have percent change in seed production when you compare a Gammage versus undamaged plants. And these three on the left actually do worse after being eaten, they’re able to regrow, but they still are not able to fully compensate for the damage that they occur incurred. Whereas in the middle of these four and the white bars, they’re able to fully compensate after damage and they’re able to produce just as much flour and seeds, um, as their undamaged counterparts And then over in the black bars here, we have a few eco types that are able to increase, um, their seed and fruit production after being damaged. So they overcompensate, so tolerance kind of existed and terror spectrum that we can look at. And our question really is how has plant tolerance evolved in an invasive range, land, pest species? So for our study, we used a yellow star thistle. It’s a very highly successful invader, um, especially in Western, uh, California and the Western United States, and it’s native to Eurasia, but again has been introduced to both South and North America and the invaders show increases in size. So if you look over here at this picture, we have a native, uh, eco type over here, and then from an eco type from the California invasion, and the plants are not only show an increase in biomass, you can tell it’s much larger, but they also have an increase in reproductive potential as well. And there’s some evidence of changes in how the immune response is working with, you know, some types of pathogens. Um, but one thing that we do know is that yellow start this interacts with hundreds of species across it’s invaded range Um, however, there’s a lack information on health plant tolerance may be changing. So that’s what we really wanted to look at. Um, yeah, so the question is how do native innovated populations, the yellow star to Seoul respond to damage? And for that, we took four invader range populations from California, and we also had four populations from Western Europe as well

Uh, so these are native ones, of course. And then for our design, we, uh, used to randomize block design and greenhouse. And once these plants, uh, you know, really, you know, they set the rosettes and then they started to elongate their inflorescence. Um, once they started bolting, we went in there, damaged them, remove the inflorescence. Once they reached about 10 centimeters, the height and crop it down to where only the rows that is left. Um, and we did this because in this area they’re often mode or grazed, uh, as a form of management and control in California Um, so that’s why we did this particular one instead of, you know, insect damage, like present damage. Um, so then we took both our control on damaged plants and damaged plants and allowed them to go and set flour. And then we went into, uh, take our biomass and reproductive fitness, uh, as well as permits And this was to assess both the vegetative and reproductive tolerance of these plants So what we have here is what we see is that if you look over here on our native populations on the left, uh, we have perpetual a number of flower number, uh, on the y-axis. We have our control plants that were undamaged from the native range, and they significantly, uh, do worse after being damaged. They under compensate. Um, so they’re not able to fully compensate and, and recover after being damaged. However, from plants from the invasive range are able to equally compensate. They’re able to just, you know, do just as well as their own damaged counterparts, um, after being damaged. And we see similar things with vegetative compensation, and this is from, uh, biomass measurements. Uh, so basically went in, threw them in the oven, uh, take the biomass and we see similar things, native, uh, populations are smaller than their, uh, you know, after being damaged, whereas an invader range we’re able to fully compensate So it does look like that the invaded range populations have evolved in increase in its, uh, tolerance or compensatory ability during this invasion. And one of the other things that we wanted to do, and I was interested in was kind of exploring any type of mechanism for this increase in tolerance. And one of the reasons I was really interested in this is because the lab that I was previously in, uh, we did a lot of work with, uh, plant compensation and understanding, uh, how, uh, what kind of mechanisms leads to increase compensatory and disability. And we did this in rapid Opsis And what we found was that this process of endo re duplication was able to was the causal mechanism for plant compensation. So in case you don’t know what indirect duplication is, it is a successive rounds of my ptosis without subsequent, uh, cell division. So after a couple of rounds, you go from like a to in, uh ploidy and after a couple of rounds, you can just keep going up until like 16 and you never divide afterwards. So you can very quickly go from like 10 to 80 chromosomes And this leads to different kinds of nucleus typic affects larger, uh, cells. And you can also have increases in gene expression and things like that as well, and increases in water transportation between cells. So, um, just a little bit of background on it. This is from a rabid opposites. We have, uh, this black, uh, circles here is an eco type Colombia, and this is able to increase our overcompensate after being damaged. And if you look over here at nuclear DNA content, after being damaged, it has a significant increase. Whereas this Landsberg eco type is an under compensator It doesn’t change has no ability to undergo into reduced vocation. And then if you expand that out with a bunch of different, uh, eco types, we still see a general relationship between the two. So we have changed in soul cycle value, which is our measure for endeavor duplication. So over here, it’s very high and up here as well. We have, uh, on the right-hand side, uh, increases in seed production after being damaged. So there’s a generalized relationship, and this was followed up with, um, uh, genetic manipulations actually showing that it was a causal, uh, cause and effect between the two. So this is an under compensating you could type is it is a different Columbia than the one I showed earlier, but, uh, this one after being damaged has a reduction in seed production. As you can see in the white bar over here on the left, as well as biomass And it actually has a negative cell cycle value after damage. So it does not appear to reduplicate. And whereas if you increase, uh, interview application, uh, uh, experimentally

via that’s overexpression of increased level pointing one, you not only increase indoor duplication levels, but you also change this plant from an under compensator to an equal compensator. So, um, yeah, that was kind of some background and why I was interested in this and some of the work that was done previously So we wanted to do the same thing and to do this, we use a flow cytometry and, uh, one of my undergrads helping me with this, that really focused on this was Jesse. Unfortunately don’t have a picture of him on my computer So we’ll just say, this is Jesse. He’s a very happy guy, um, and very, uh, you know, helpful for this whole thing. So anyways, you just put in yourself more or less, if you chop them up, uh, release the, uh, nucleus, you hit it with a light source and it measures for essence. So you can get like your different employee levels. Um, so we used a rabid offices for reference, just kind of showing you this is the kind of graph. So this goes all the way up to 32 in, and then taking, you know, now looking at, uh, you know, we’ll start this whole, uh, we just didn’t find anything. It just is, uh, two in and, uh, you know, we put a lot of work in and it just turns out that, um,, it just doesn’t end over duplicate. So that pretty much shows that Endover duplication is not, you know, the only thing that can lead to increases in compensation. Um, one of the things that we really want to look at is re allocation of resources from the roots and different things like that. Some of the work that’s been done in the past as well in other species. Um, but yeah, anyways, um, so we found, you know, again, just as a conclusion going over things, rapid evolution of invasive species is often invoked as a key to a successful invasion. Um, and plant tolerance, uh, may plan an understated but important role in invasions. And it could be really important for the types of management we use for these, uh, plants that are able to, uh, compensate because as I said earlier, yellow start those solos often mode or graze for management So that could really pose a problem if you go through mow it and it just pops up and it just has no fitness effects. It’s able to produce the same amount of flowers and seeds, and it can also have possible implications for long-term bio controls, if it increases in, uh, you know, competence, you know, that tolerance is what you’re shown. Um, and of course, indoor duplication was not the mechanism for this increase in tolerance, but, uh, there’s some, I, I found this paper that, that was interesting looking at the D the yellow surface was a deep root system, and it showed, they were seeing that reallocation of resources or what appeared to be re allocation from that root system makes them very similar genes, which, um, so again, they were experiencing these, uh, you know, difficulties and mowing, uh, to control it. So, yup. That’s all I got, um, some acknowledgements from everyone in the lab. And of course my, uh, excellent undergraduate research assistants, um, Jesse and Mikayla, Mikayla, his work is different, so it wasn’t here, but yeah, that’s all I got. So any questions and thank you. Thanks Joshua for that excellent stock. We have time for maybe one very quick question. Oh, I’m sorry. I didn’t know how it was going. I’ll pass on a question from Slack from Ellen Ellen. She says, cool. Study flowering time could be a tolerance mechanism, especially if mowing or a livery is uniform across individuals and time. So why did you choose to clip at 10 centimeters rather than it a single time? And how might the results be different if you’d done it the other way, if we Oh, okay Okay. I see. I see what you’re saying. Yeah, you can definitely. Um, I think, Hmm, we just wanted a uniform type of, um, more or less, uh, like we cause some that is true. Like some of them flowered earlier than others, and if we clip everything at the same time, sometimes we weren’t going to even have anything, but a Rosetta at that point, some of them were definitely taking longer to actually mature. Um, but that’s a good point and, you know, changes in phonology can definitely be a way to avoid any types of damage as well So that’s a whole other, you know, thing that you can look into, um, for avoidance of herbivory completely, instead of like tolerance or compensation instead, um, this is what we were interested in or what I was, so, yeah, but that’s a very good point. Yeah. I don’t think there’s a right or wrong way to do it. You’re you’re missing something whichever way you do it Yeah, definitely. Yeah. All right. Thanks

Jeff. Thanks, Casey. Um, just keeping the time to move on to our next speaker Joyce, and she will talk, tell us about the evolutionary inferences on the community organization of flower humming, but interactions. Um, Joyce, we can see your screen now, so it’s over to you. Can you please unmute yourself? Joyce, Joyce, we’re not able to hear you looks like there might be an issue with your audio settings. Sorry Um, okay. So I’m going to present this work that I developed together with, with. Can you please share your screen again? Okay Sorry. Um, so there are a lot of factors that may fleece the structure of communities. And one of those factors is pollination. So pollination mind place the community structure by several ways, for instance, promote and competition or facilitation and influence in this species distribution or SP resistance. And, uh, the way in which, uh, pollination will interfere on the structure of communities will depends on the degree of dependence of, uh, interactions and systems. Uh, the interactions tends to be very specialized, especially for some plant groups like any cornea and Ramili assays But when we look at the hummingbirds, um, more, uh, with more attention, we see that there are a lot of differences between the species that allow us to classify them into two ecological groups, the hermits and the no hermits, uh, this difference are related, especially with, uh, the morphology of, uh, the bills, uh, with the feeding behavior, uh, the degree of diversification and, uh, uh, geographic distribution. And in general, the hermits Herman Berg’s tends to be more specialized than the known hermits and, uh, the morphology of the humming birds and all the other species are associated with the morphology of the flowers. And when we have the straits matches, uh, this interactions appears as a modules in plant pollination at works. A module is when, uh, interaction, uh, tends to occur, uh, more often than others And these interactions are probably a result of a joint evolutionary history. And when we have a joint of originator history like evolution, the phylogenetic tree of the groups tends to mirror each other. Like what happens in golfers and lights, uh, opposite for, uh, pollination. We cannot expect, expect a perfect congruence, but we still can search for fellow genetic signal between the phylogenetic trees Uh, the signal suggests that the current association are a product of a joint evolutionary history Innovation sacks are shared similar ability, the pressures. Um, and so in this, uh, started with two research questions. The first one is if the genetic seasonal changes between communities pollinated, mainly by hermits and non hermits hummingbirds. And what we expect is to find that a higher co-fellows genetic signal for communities pollinated by hermits rather than a non hermits, because this is, this group is more specialized. And the second question is if the lack of Leatherneck CNO in communities, is it by these two group of hummingbirds changes with less chewed and altitudes and respect find a higher governance and ethical Stevie now in low lands and areas close to the trucks then for Highlands and areas far from the trucks, because hermits hummingbirds, the more specialized to come in birds are restricted to lowlands and areas

close to the trucks. So to answer this questions, we use a use a to Phillips genetic trees, one for plants, and one for him in birds. So in this, uh, uh, phylogeny, we can see the hermits here, the rabbit, they are represented by the Hermitage plate and they are older and less diverse. They’re diversified. Then the others, we also use to analysis. The first one is proposals approach to cough lodge in is, uh, also known as Packo and this analysis yelled at global, uh, goodness of feed statistic and the p-value and yelled the most important associations, meaning that they show us the associations with higher contribution to the signal. Uh, the second analysis is complimentary, eh, it’s, uh, uh, it’s called handle, uh, dental and partitions and yelled and normalize Gini coefficient. This G edges kind of saints ranges from zero to one, and as the version of proportional to the covalence genetic signal So when we have a fellow of St, closer to zero, we have a strong signal, which means to say that either every interaction contributes in a unique way to the, to the, uh, congruence So, um, every interaction is unique. And when we have a value close to one, we have a weak signal. That means to say that every interaction contributes in a similar way to the signal We do have a threshold. This value is 0.72 for my detail. Uh, please see the, the original paper on the analysis. The stretch holds 0.7 and above, below this value, we have a strong signal and above a weak signal, uh, the interaction data we use it, uh, what we organize it from a data paper on vertebrate pollination attends to the forest that will be spirited very soon. And we divided this data into eight lots to, to know, uh, ranges and five altitudinal ranges. Each one of these ranges is considered as a community. So we have eight large, two tunnel communities and five altitudinal communities, and we calculated their better diversity for, uh, among these communities. And also, uh, the coalition ethical signal. So we ran Parkway held on campus, uh, first, uh, considered the hummingbird group. So we ran this analysis for the whole data that we called metal web, and then we divide it into two groups, the hermits and the non hermits. And then we did the same for the five, uh, attitudinal communities and, uh, for the eight longitudinal rains And we ran, uh, this analysis a total of 33 times. And our results is this. Here, we have, uh, the phylogeny of, uh, plants on the left and for hummingbirds on the rights, the lines represents the associations blue lines represents associations with a higher contribution to the signal. And the gray lines are presents, uh, the associations with low or non contribution to the signal. And what we can see is that we did find a congruent psychological signal for all, uh, all cases. Uh, Jesse here, the P value was equal to zero, uh, which is expect to, since these interactions are very specialized. Uh, but when we look at the genus signal, Ginny value, uh, we see that we find strong punk ranks on the, for the metal web and for now in specially for non Harmatz webs. And we didn’t find a strong signal for the no hammer to that. Um, and this is also expected things, the hermits are more specialized and probably, uh, had a joint evolutionary history. Uh, we can see here that that is a group of plants with high contribution to the signal. These are the romantic bromeliads, and we also have a group of hummingbirds with high contribution to the signal. And these are the hermits and especially one species named regarding latch two-thirds. Uh, here we have the results of the diversity here. We have the better diversity, uh, then the turnover rates, which she represents the change in species composition, uh, among communities and the nastiness less than this, but, uh, indicates the loss of species among communities and not we, we had, it’s a higher value of better diversity in turnover and a low value of message for both planes and birds, which means to say that, um, the communities in Atlantic forest are highly diverse and rich, uh, regardless to the genetic signal

Uh, we remember that we hope it’s find, um, a higher co-fellows electric signal for area schools, which the tropics and this plots, we have the results, the Gini value for each one of the large student arrange, and the dots represents the, the communities. So here we have the MetAware app, the Hermitage web, and then Ohio Mitchell app, and the Phillips symbols represents the communities in which we found a congruence, a significant congruence between, uh, plants and hummingbirds. So, uh, we have to look for this point and with it fine, I strong and ethical signal for some communities So the GFL is below zero, 0.72, but in general, when we did find, uh, congruence, wasn’t the areas in the cyclical, in the tropical areas, in not in its civic, sort of tropical errands for the, uh, we found similar results for plants, but for birds, we found a low turnover rates and high Mastin’s rates, which is also expected since we have some loss of species with elevation. Um, the hermits are restricted to learn lines, and we hope to find that, uh,, uh, meaning that we hope to find a higher covalence kinetic signal in lowlands. And here we have a similar part with, with, uh, innovation. And, uh, we didn’t find a strong genetic signal for any case, but when we did find a congruent squats in the lowlands, well, comparing this to groups of hummingbirds So we found evidence is for, um, a joint evolutionary history for hermits hummingbirds. These hummingbirds are more specialized in mythology and chaplain behavior, and especially with, uh, Ramit bromeliads But when we, uh, we cannot use, uh, the philosophy, uh, the plant phylogenetic tree that we have, because it’s not dated, it’s not calibrated, but here we have, um, a phylogeny that’s calibrated and, uh, with the reconstruction of the pollination of illusion for the Burmese gases. And, uh, what we can see is that the orange of this group, uh, was, uh, uh, from 20, 20 million years ago, eh, and it’s Quincy to, with the orange of hummingbirds, more, especially the orange of that hummingbird with higher contribution to the signal. The half of the nephews was 15 million years ago, and it coincides with the orange of, uh, uh, the orange of bird pollination in the bromeliads. This is another evidence that maybe this, uh, group of hummingbirds heads, uh, a joint evolutionary history, they shared some similar oppression with these plants. Uh, we didn’t find the same results for, uh, the hermits, eh, and they are very generalist is humming birds, uh, in territory, in, uh, distribution, uh, uh, in deferential behavior and also in view morphology. So here we have some examples of hummingbirds with a great variation on bill morphology. Uh, and we also saw that a higher diversity and a high species turnover in the communities in a transcript pharmacy. Uh, this might influence the special distribution and the phenological for lab in the traits matching on this communities, which are dependent on the ecological context, might force the no hermits to behave in a more specialized way And then, uh, uh, organizing these communities in some places, uh, they volunteered inflationary context of the community communities also make forest this hummingbirds, uh, to behaving in a more specialized way. Uh, here we have an example of this species., it’s a known hermit time in birds, and it has a trap liner and a territorial is behavior, and it’s, uh, it’s restricted to the Northeast portion of the Atlantic forest. And when we look at the results that we found in this area, we did found a chunk of fanatical seasonal and this hummingbirds together with, uh, another hematoma in birds, uh, contributed, uh, more significantly to the Seguinum. So in this case, we might also have, uh, evidences for joint evolutionary history of these hummingbirds with, uh, the plants, especially the bromeliads, uh, in these areas. Some communities, you have five though. Okay. Thank you. Um, in conclusion, um, we found, uh, dances for couple of, uh, couple of evolutionary history for hermits

and plants. Uh, we then find the same for, uh, the no hermits hummingbirds, but dependent on the context that evolutionary ecological context of communities, these various generalized hummingbirds might behave in a more specialized way and then organize communities. And we also found evidence as for our conclusion ethicacy now interrupt courageous. And in all lanes, uh, this works have some caveats And one that I like to point out is that, uh, we used an uncut calibrated phylogenetic tree for plants, which makes comparisons very difficult to make And also, uh, it makes difficult to draw some conclusions. Um, I would like to thank the university of San Paolo, the for allowing this work to happen. The American society of naturalists in-between as Lamar for the opportunity to present, to show my work by a Colwell thirst, and some people that help with comments and data organization. Thank you very much. Thanks, Joyce. That was fantastic Talk and great timing. So we have plenty of time giving people a few seconds to type out their questions. Joyce John Thompson, as, uh, are there plant taxa that show a joint history with hermits and non hermits? I’m sorry. I didn’t quite understood. Yeah. Are there any, um, are there any plants that have a history with both hermits and non hermits? Uh, no only pro uh, permadeath Familia says Familia SSR, a very important group and attends to forest. And I think that they might, uh, uh, avoid, uh, to find CNO with other plants I would have to run this analysis without this group, because I think that it might show up, might show something, but, uh, with bromeliads we didn’t find for another. Oops And then there’s, uh, another question from Lucas Camacho on, um, Slack. He says, nice talk, Joyce. Um, do you know how using a calibrated tree will change your results? So thinking that maybe the evolutionary history of this species are very important, maybe using a calibrated tree will reinforce your main results about plant hummingbird co-evolution yeah, it would. You would also change a little bit of the analysis. So this analysis, the second one random top is a very recent, recent one, and it changes the trash hold it’s changes, uh, uh, the conclusions you can make. So it would be very, very different results if you had a calibrated three, but I believe that we it’s a very hard task. I don’t think that we have so some, because especially for our tentacle forest, it’s very diverse place, but yes, it will change the results probably. Yeah. I expect that, uh, we would be, we would find stronger phylogenetic signal. Okay. Any other questions? Right If there are no further questions, um, thanks again, Joyce for the talk and, and I will maybe invite our next speaker. Who’s Natasha demand Corp, and she’s going to, um, discuss the biases and our perception of plant pollinator networks, which is very intriguing title Um, Natasha, if you could share your screen yep. It’s up now and over to you. Okay. Thanks for the introduction, Sarah. Hello everybody My name is Natasha and today I will present some results of my PhD days that I did in France at the university of live about our perception of lampooning it, my course, with a comparison of visit and polling visa presentations of the same network. So it can be used to describe ecological communities and interactions among species, but in

the context of climate change and species loss, it is critical to understand the mechanisms that drive natural stability in order to preserve these maps. So it is important to have an exhaustive representation of the ecological network in this talk. I will with a major focus on wellbeing. So most part of networks are, uh, base, uh, and construct from the observation in the field direction between applied insect and a flowering plant, but some of these interactions will remain and observe and probably even a bias picture of the plant. So not tentative metal to study plan, putting a direction, not observing the field is the use of Poland since it’s a natural marker or natural activity, because you remain attached to the pollinator bodies after their visitation. So in these studies, I focus on treating scariest grassland in three different regions in France, uh, distributed along an environmental, but also diversity gradient because it’s seldom size display a higher, highest diversity. And in this size sample, the interactions in the month of July, 2016, which was one of the richest month of this year. So I’m old symbol insight. I focus on the females and I collected volume from the insect bodies to prepare two fullness minds, one collector, specialized structure, and one with the police Patridge on their bodies And I decided to say per 82 colon tides, because they bring two different formation. Deponent scattered is Depoali passively transported It gave a spectrum of pole or potential oldest flowers, which are originating from visit, uh, for other types of resources, such as nectar and disease DePaul, and also available for the preservation of that. While the pollen collected is deep polling activity collected by the coordinators on a selection of land species, you start in specialized structures and his department that they bring to the last two Fijians a lobby. This Poland is often packed and mix with saliva enzymes So this partially reduce the Poland viability, uh, to help housing even duplication of the bowling rains collector from the interest bodies. I collect Yonkers. So of oldest following species and our tree size, uh, which has an error of one actor each. And we prepare T three local colon atlases. So in this study we have clear policies do. First one is that you scattered Poland provides more information than the activity collected Poland. The second part is, is, is that which our communities are more likely to be understandable than poor communities. And it turns it both is this is that the addition of the polling information can lead to changes both in genetics infrastructures and the species role, especially for single species. So to, um, try to report to our first people to this. And we compare the refraction perks of from the juke polenta lives. And we didn’t find any significant difference in the number of realized beings in any of our sites. So even if we find a percentage of unique links, which was either for the Poland Scouts, right, as we expected. So since we didn’t find any difference between the Poland type, we merge information in one Poland based network And we prepare one Poland for each site. As other studies point out, when we add the polling information in a Poland based network, we have a higher number of species and compared to the visit based name, but we cannot compare metrics because they are strongly affected by the nectar dimensions. So to avoid this problem, I am create a simulated network where we fix the number of interactions per coordinators species, as in this example here, but we randomized the interaction there’s recording, uh, using information in department days now. So now that we have to, we can investigate where the additional reporting information can lead to changes internet, infrastructure, and spaces, and to do so. I use different indices or methodology for the nexus factor connections, which is the proportion realized this over once the age to index the index of networks, the sensation and our classroom methodology, which is developed

model and to investigate changes in species walls. I used these species positions, uh, in braider, other mentors that I use, but I will not present today. So despite the fact we found marked differences in both species richness, and also in the interaction goals for the visit based network and the Poland based network with the South on site displaying the highest diversity. We didn’t find any significant difference voting the age to indexing to us in our three sites When we compare the visit based network, which is the black bar here, we just see related networks, which is the light gray bar. And this was true for both in disease and outside So this means that the richer communities, communities are not, uh, under sample then for community during this, it changes in the next structure. And in Metro plus visitations, I use the black model methodology who finds the best grouping for plants and insects separate me and this grouping have all more genius connection patterns. Uh, and as in this example here, I have decided to choose my Northern side and we have three separate blocks for the plants and two blogs for them. So when I call the changes in the blog, clusterings, uh, I use an alluvial diagram, which is this type of diagram here where we have the visit network on the left and the pollen network on right. And the bottom, we have the insects pieces clusters and on the top of the plant spaces. So this, uh, diagrams I highlighted in color only displeases that actual change Um, so if we take this site here, we can observe that you block two mean uses nightclub. Most parts of TCS remain in the same book, just one or few spaces change. So despite this species rearrangement that we observe no resize, we didn’t observe any substantial changes in the natural structure. And these results was also confirmed by the NMI index, which is a normalized index, which was close to one, meaning there were almost a perfect congruence between the music network and deponent clusterization. So we don’t have any changes in electrical structure, but what about species position? But you introduced the concept of species. Patricia. I have introduced the concept for motive for based on board of CMLs at talk 2019. So if we imagine our methods as a legal house here, the Mounties, uh, different PCs with different colors and different styles that compose the house and the region, each motive, we have this position, uh, representing the species roles and, uh, in each multiple rehab, at least two position, one position for the one position, for instance. So the same species can appear in different position, different motives and the change of position, depending on the complexity of commodities, the number of spaces, but also by on their interactions. So by counting frequency with which species are putting different position, uh, we can quantify the species, which depends on direct interaction. So the interaction between the group of plants and insects, but also indirect interactions. So the interaction within each group, though, between pointing angels or between plants, which are, uh, that are competing for the same reason. So as we expected, we find that, uh, when we add on information, we had, uh, changes in the species, especially for senior common spaces. And this example here, we have, uh, people don’t plant species with only one partner and one interaction. And so do you position in the visit based are only one, one positions, which means that was he shown with only one pattern as in these three examples that have that, when we add the polling information we observe and we found that, uh, there is another interaction with these species, and this means that now is not the same pieces anymore, because now the Aquila has two barns and a new Parkland brings also new positions And so now we can find our spaces and to see more exposition in position, which two partners,

as in these two examples. Yeah. The, when we compare, uh, the visit position with all the simulated positions, we found that the distance between the visit causation and the simulator century was greater than the 95% of the old simulated, uh, distances. So all the simulators position, which means that there is a change in the species position and the species that, uh, this was also was not only observed for seniors on species, but also for more connected spaces, such as in August. Now the corner here, where once again, we have a great distance between the visits position and simulation central and differences, and just basis disposition was not due to the additional when your partner, but loss of a partner, which was not confirmed by default, then based in the PolyBase based necklace. And this results was also true for other sites in my study. Um, for example, here we have, uh, a proteinaceous species and another plant species that were observed in a connection in the visit based network, but not in department one. So this brings to changes in both species positions So to conclude, uh, our results show that recording interactions, uh, from visit observation does not provide a bias picture of the network Use opponent can provide a more competing formation at the species level, and this will be especially meaningful for study investigating the Polynesia 15 Venus, for example. So do you take a message of my study is that when we use VZ based observation, this is a faster and aggravate me to the Elegy to gain information, advantage interaction metrics, but I’m more detailed methodology such as DePaul can really improve our knowledge of the species home within ecological communities and with important implication also for conservation. So thank you for listening. And if you are interested in learning more about, or you can read our paper, which is published, hello, and thanks for having me. Yeah. It was a fantastic talk and you’re very good at keeping time. That’s the time now for discussion and questions and yeah, the, I suspect there’s a lot of people breathing a sigh of relief after seeing your results, relaxing, their visitation That works. You’re still very good. Just waiting to see if some questions come in here. Aloe has a question. So wait a minute for that. Okay. So Paula is asking, do you find changes in the role of highly connected species across the two net representations? Uh, yeah. Um, among all the differences in your role position, half of these differences were forcing its own species and the other half was more connected spaces. And I was expected for seeing its own species because we did additional poll and raise the number of the interaction at the Oles for more connected spaces. Sometimes they really, really increased their interaction with space there that we didn’t observing before interaction visit basement. So, um, it was a really interesting results. Cool. And then another question from John Thompson is how may, um, pollen amount per flower versus pollen distribution among flowers. Uh, so some pollinators deposit, much more pollen per visit than others. So

how, how could you incorporate that information? Yeah, so, um, I didn’t re quantified the amount of colon. Uh, I only add as, um, single interaction, uh, and what I did was to use a threshold Uh, so when I count less than five polling grains of once pieces, I didn’t consider consider that like a true interaction. Uh, this is the only difference that we made or try to make, but we didn’t really count, uh, the number of grains, but for different species, but only Costilla as an interaction or non-construction Yeah, that makes sense. And then Anna asks, um, she’s wondering what this means regarding the prevalence or importance of pollen or nectar fees. So how often do you find species that visited certain flowers, but never carried their pollen or their free pollen? Yeah, so I had this, uh, thing for several interaction. Um, so when I say bird did the information for the two type of polling, I also, they pray to when I have only an interaction in de and this was also the case or different spaces And the major difference was based on the tree shows that we decide to, to use. So even if, sometime there was one holding grain, we didn’t think that was enough to justify the interaction or just to confirm the interaction, but this could change depending on the time of the day, or also maybe sometime was just came out from their nest and maybe they even have time to collect department. I can make up this difference, any other questions? Okay We still have a few minutes for questions And while we’re waiting for people to pick up more up here’s one, how many individuals per pollinator species were collected to quantify pollen, plant species? How many, sorry, the pollinator. Yeah, exactly. Uh, I, this depend on our sampling for some, for some species, we only have one individuals and for the other species, we have a cyber or cyber need. You need needles and do was different. And depending on the side where we sample the insects, but yeah, we didn’t really, uh, I keep all the old, uh, insect as a carry both, uh, or like collected fall in their structure and then on their bodies. So this also reduce my, my sample. Uh, but I do see if there was a difference between the type of the two type of following one was one. So, okay. I don’t see any other questions, so thanks Tasha That was great. Thank you. Thanks for that, Natasha. Um, before we move on to our next speaker, I just want to remind everyone to maybe take a stretch because it’s a long evening ahead of us and we want you all to be completely with us. Um, we will have a break, um, at five 20, um, but we have a minute to stretch your neck arms, get up for a minute. All right Um, we will start moving on to our next speaker, David Pete, who’s going to tell us about the

soft filler center mayor, David, if you can share your screen. Excellent. All right. Thank you. Um, hello, my name is David PD. Uh, I am currently a first-year PhD student, um, at Brown university. And today I’ll be talking about, um, the work done in my undergrad institution, the university of North Carolina at chapel Hill during my time in the two day lab. And, uh, my talk is focused on the evolution of host preference into Sophos into Maya. But first I want to acknowledge and thank Dr. Daniel and Dr. Brandon Cooper for all their work and amazing mentorship that made this project possible. So ecological specialization is a process whereby related species evolved to utilize different subsets of the total niche space available to them Many groups of insects that rely on plants for food or breeding sites, display such a pattern and have evolved to specialize on a small fraction of the total plant species available to them, which is known as plant hosts, specialization. There’s a software Malana guests or species subgroup is a well-suited to study the ecological and genetic basis of hosts specialization. This group contains nine described species that have evolved over the last 10 to 15 million years to exploit a wide diversity of ecological niches, four of the species within the subgroup that are outlined in blue on this philosophy or soft flaw Malana gasser lands more Shyanna your Cuba are considered dietary journalists. Well, five other species outlined in orange. Uh, their software Tessie Arie says Cecilia or Raina, Santamaria and director Athar are thought to be dietary specialists. However, just off a test Sierra or Safa Santamaria have never been apparently studied with respect to their genetic or physiological basis of behavioral preference for their respective hosts. For this study, I will focus on the most recently discovered species and the Malana guesser subgroup or Sophos Santa Maria, which is endemic to the high altitude misfortune on the Island of stock to me, which is this little Island, um, shown by the cell.off the coast of Africa while it’s sister species or soft Lake Cuba is found in open and drier habitats at lower elevations on the Island of South may, as well as throughout continental Africa. The motivation for this study came from years of field collections. So the Island of South to me, where it has been suggested that to Sophos santamania breeds exclusively on the nontoxic endemic fig subspecies bike committee, a Cobre well, it’s just a species or soft way. Cuba is considered a dietary generalist. So with the knowledge that ecological specialization is widespread among insects, we want it to empirically test if their software Santamaria has specialized on the endemic fig species from an ecological engine number perspective. So from my talk today, our first to tell you about the ecological and behavioral work performed on the Island side to me, as well as our classical population genetic analysis to detect positive selection. So first ecological and behavioral States. So we first studied the theological structure of the Island of South tomato for fruiting trees and just awful abundance. Each month, we scored the amount of fruits produced by each tree. And the number of collected in flight traps. With this information, we aim to identify the proximate fruiting period of each fruiting tree, along with the relative abundance of Drosophila over a two year time span. Um, the endemic fig species is outlined in red shown here. I’ll be outlined in red for the rest of the presentation And I want to take a moment here to recognize our collaborator at Lucio materia, who made this part of our study possible. So here I am showing you, um, uh, the relative abundance on the Y axis, which is just the monthly contribution, uh, divided by the yearly total for each of the species, um, on the X axis is each month over the two year time span starting in January of 2018 and ending in, uh, December, 2019 And the first plot, uh, up here are the different fruiting tree species and the bottom plot or the different selfless species at different, um, elevations. And in case you’re unfamiliar with the notation, uh, F1 just stands for, uh, uh, your Cuba hybrid. Um, and so here, uh, the data’s kind of messy, but don’t worry, I’ll walk you through it. So here we see that as expected, most species tend to produce the most fruit during the long fruiting season from February to early June, which I’ll outlined here, and that some species will continue

to produce fruit during the shorter rainy season from October to December outlined here But what we are really concerned with is seeing if the relative abundance of JIRA software Santa Maria COVID varies with the relative abundance of the endemic fixed species. So I will clear all of the data of the species that we are, are not too interested in. And after running a Spearman’s rank correlation test, we in fact do learn that the relative abundance of the soft list into Mayo significantly covariates with the relative abundance of the endemic fig species. So with this knowledge, we next wanted to see if your softness intimate is perf preferentially breeding on the academic fixed species to test this. We gather fruit from each of the tree species and recorded the identity of each of the species that hatched from each fruit. So in this part, on the axis, I’m showing you the fruiting tree species and the Y axis the counts of the hatched flies observed. And the first thing you’ll notice is that, uh, I originally showed you a lot more fruiting tree species. Um, we excluded, um, fruiting tree species from this figure where the fruits were so small that no true SoFloat would even lay their eggs on it. So these are the only true fruit species that actually produced your software. And so after running a linear model, we find that the fruit species identity is a statistically significant predictor of breeding sites for each selfless species and post-hoc analysis confirmed center Sockless Santamaria, preferentially breeds on the academic big species, and that softly Cuba preferentially breeds on ficus from COSO and CityMD guava. With this information, we next one in the test of any differential fitness effects, if there are any differential fitness effects, we’re just off of Santa’s may have breeds on other fixed species to assess relative fitness. We performed a viability to say where we reared to Sophala on standard cornmeal, medium until they were larvae where they were then transplanted onto one of the fig species. And the number of flow rate that reached sexual maturity was recorded, which is shown here. So on the X axis is a sub substrates, which is cornea, which is a standard lab medium And then our three different fixed species found on the Island of South today. And the Y axis is the percent of Loray that reach sexual maturity. And so after running another linear model, we find that species by substrate interactions is a statistically significant predictor of the relative fitness and post-hoc analysis reveals that their selfless internet, as a highest relative fitness on the endemic fixed species, well just software, your kudos relative fitness is uniform across all fake species. The fact that your software, your Cuba does not differ in the relative fitness among fix species. It is rarely found breeding on the endemic fish species suggest that the higher propensity to breed on the endemic species is a derived trait found exclusively in the male lineage. Well, the physiological ability to properly complete the Bellmont on fake species, is it an ancestral trait to both lineages? Furthermore, we find that the selfless Santamaria almost exclusively breeds on the endemic fixed species while also exhibiting a decrease in fitness when artificially introduced to the other fix substrates, which suggests that your softness into Maya has experienced an adaptive evolutionary pressures resulting in the preference for and strict specialization to the demic fig species. So I’ve just showed you that the relative abundance of Sophos Santamaria covariates with the endemic fig species that are Sophos into map, preferentially seeks out the endemic fixed species, and that’s the software sentiment has a higher relative fitness on the endemic species fixed species, which provides our basis for the investigation of the genetic basis of hosts preference and your selfless hand-to-mouth. So chemo reception in insects is controlled by a diverse set of proteins. That includes a Dorn binding proteins, chemo, sensory proteins, or factory receptors, and accusatory receptors, adorn binding proteins and chemo, sensory proteins, or small soluble proteins express in the sensory organs of insects at five distinct hydrophobic adorance in Fairmont’s, which facilitate their transport to a factory and fused Satori receptors that initiate signal transduction in sensory neurons. This set of will factory genes on the slide has been implicated in a number of cases to drive host specialization in insects and are commonly thought to evolve through positive next natural selection. So knowing this, um, we wanted to estimate the relative evolutionary rates of genes. And so we calculated Omega values, which is the ratio of non synonymous to synonymous substitutions. We first compared

the average Omega value of each olfactory gene family, and compared it to the genomic background to see if any of factory gene families had significantly elevated Omega values. So here on the X axis, I’m showing you the different gene families on the Y axis are the mega values. And this is just for your selfless amps map. And unsurprisingly, we found significantly elevated Omega values for all olfactory gene families and are selfless And to map, uh, this next plot, uh, is the same X and Y axis, but for. And we also found significantly elevated Omega values for all of factory gene families, except for olfactory receptors. Um, I’m now showing you the two pots essentially concatenated onto one, however, despite seeing elevated Omega values compared to the genomic background for the majority of effector gene families in both species, we see that the average Omega value in your selfless map is significantly higher in all gene families, except for chemo, sensory proteins These results suggest that the average that on average for Sophos into Mia has experienced more adaptive evolutionary forces than to softly Cuba. And it’s consistent with our hypothesis. This is a post specialization and the just off of centimeter lineage. We also compare the number of unique genes exhibiting an Omega value of greater than one, which is indicative of positive selection in each species to assess if a factor change showed an enrichment of elevate Omega values compared to the rest of the genome. Uh, this is the same plot as before for just office and to man. And after running a Fisher’s exact test, we find more genes within a mega value of greater than one than we would expect by random chance for all of factory gene families, you yourself, placenta map showing you the same pot, but first to softly Cuba, um, we only observed this trend and adore binding proteins and chemo sensory. Again, these results suggest that on average, your selfless intimate has experienced more adaptive evolutionary forces than yourself like Cuba because of the biases and limitations though, to results that are reliant on only Omega values. We also perform two additional tests for selection, which I will not discuss in detail in his talk here, though, is an overview of all the direct tests for selection, with all factor gene families, with the most support across all measures, highlighted for each species. If we take a very strict interpretation where gene family must exhibit signs consistent with adaptive evolution across all tests, we find that adorn binding proteins and gustatory receptors have the strongest evidence for experiencing positive selection. There’s also Santa man, well, chemo, sensory proteins have the strongest evidence for experiencing positive section just off like Cuba and just silly. There are multiple occasions. We find significant evidence for positive selection and the same factor of gene families for both selfless Antonia, Andrew Safia Cuba, which could suggest that all factory genes on average or evolving more rapidly among the Sophos species, which also includes evolving for loss of function. Lastly, to actually identify candidate genes underlying the evolution of host specialization and just awful since ma’am we conducted a relatively standard genome scan for positive selection, which is outlined on this slide by comparing local levels of diversity to the overall background levels of diversity or positive section scan identified 243 tenet genes, five of which all factory genes were just shown here on this slide I’m showing you the five olfactory candidate genes from our genome scans. And what was really surprising is the fact that we did not see signatures of positive selection for any genes in the OBP 19 or OBP 22 gene clusters, which have been implicated to underlie host specialization and not only a number of their Sophos species, but also other insects. These results suggest that the evolution of host preference might be predisposed to involve a factory genes, but the specific genes underlying specialization differ from species to species So conclusions. So first, uh, I told you that the relative abundance of the Southwest centimeter coat varies with the relative impotence of intimate fixed species on the silent on the Island of September. I then showed you that the Southwest Sante may have preferential seeks out the endemic big species for breeding sites. I then showed you that the selfless, except it’s a higher relative fitness on the academic fig species. And lastly, I showed you that olfactory genes show signals of signals consistent with adaptive evolution, and there’s a selfless anti-male lineage. And lastly, my acknowledgements, I want to thank, uh, Dr. Daniel at UNC chapel Hill, Dr. Brennan Cooper. You proceeded Montana who assumed material, um, the members of the Tutsi lab and the members of the quartz of Centris lab for all of the help and feedback along the way that NIH for funding this work, uh, the

American society of naturalists, uh, for allowing me the opportunity to present this research I’m happy to take any questions. Thanks, David, for rebel time talk, we have five minutes or questions. That was great. David, uh, let me read Mark Nick peaks question. Uh, he says, nice talk, David. The differences in olfactory gene evolution is really cool, but they must have evolved because of the fitness differences you showed and those differences are driven by other traits. So what do you think causes the fitness differences among different host fruits? Yeah, so that’s a really interesting and it’s on a case-by-case basis. Um, so, uh, a really classic example is, um, this office, a Shalia, which, um, as specialized on this, um, fruit, which is toxic to other just awful species. Um, however, um, the intimate fixed species is non-toxic. So, um, nothing jumps out, um, to mind with this ex example, but you’re exactly right. This idea of specialization involves a whole host of other genes outside of diesel factory genes. Um, what we saw when we had elevated levels of, um, Omega in the rest of the genome, um, and just all placenta, may I compare to your sophomore year Cuba. Um, and then Peter, Ralph is asking on Slack, um, does the native fig live only on some parts of the Island or in some habitats? Yeah, so it’s actually only occurs at the, um, Altitude’s, um, which the software centimeter occurs, which was one of the, um, indicators for the study. However, in the hybrid zones, in the Midlands, on the Island of, um, the, uh, the software, your Cuba does have access to these endemic figs, just waiting on any other questions while we’re waiting. I just want to say that there’s some really interesting discussion happening on the snack Slack channel for this, um, session. So if you haven’t yet looked at that, it’d be a good time to have a look and thanks everyone for contributing. Yeah, I was saying you could also that Slack channel will still exist after the session ends. So you can ask me to answer questions there as well, too. Um, one other comment in the chat, um, were any of the genes that popped out as adapting, particularly surprising any plans on follow-ups for functional studies? Yeah So what in particular to me was, um, this gustatory receptor 97, a, uh, which is, um, co is a bitter taste signaling and, or soft the larvae. And so there’s bitter tastes signaling Um, it’s normally, um, a way to tell if your substrate is toxic or not, but this is puzzling because, um, um, the endemic fig is actually non-toxic. So at first we were wondering, well, why, why would this be one of the hits? But, um, we see that from my previous self, a centimeter has a higher relative fitness on the endemic fig species and actually, um, the better tasting signaling, not just necessarily for toxicity or not, but it can, um, differentiate between the fine scale cues of different chemical profiles of different, um, fig species. So, um, that might be one to, um, follow up on where we’re still kind of debating if we want to go down the functional route and see if we can actually knock out any of these genes or do any CRISPR

stuff. Yeah, that makes sense. That’s cool I’m just going to sneak in this one quick question. Um, how do you differentiate between the parental species from the hybrids at the larval stage? Okay. Um, so at the, I mean, and we couldn’t do it in the, um, in the field, but we didn’t actually need to do it in, in the field and the, in a lab, we could obviously do it based off of the, um, off of the known crosses, but in the field to differentiate the hybrids, the santamania Quba hybrids, their abdomen, uh, is Brown, um, compared to a regular, um, that kind of opaque color Awesome. Thanks. Thanks for that, David. Thanks, Casey. Um, we’ll now move to our next speaker, Sarah McPeak, who will tell us about the theories of natural selection on resource provisioning in pollination mutualism. Sarah, if you could share your screen sure thing, here we go. Um, oops. I can’t see the doc Here we go. All right. Can everyone hear me and see that? Great. Thanks. So, um, thanks everyone. Uh, my name is Sarah McPeak. I am a second year grad, I guess he didn’t have focused on change. Oh, Oh, sorry. Heard some feedback here. Um, but I’m a graduate student with Butch Brody at the university of Virginia And, um, it’s really been a goal of mine to stand up on stage at Sylmar and present some of my research for several years. And so I’m just going to pretend that I’m there instead of at home in my parents’ living room in fuzzy slippers. Um, so in this session we’ve heard a lot of different ways that we can think about these species interactions that we study And I’m going to present a different perspective, which is the perspective of consumer resource theory. And in consumer resource theory, we think about a lot of species interactions as taking place in an energy economy, and that all the species in these interactions are consumers and resources that are exchanging energy. And we can divide these interactions into two main categories based on the effect that one species foraging has on the other species. So in consumer resource antagonisms, um, the consumer species has a negative effect on the resource abundance, and it has a negative effect on its partner species fitness, but in consumer resource mutualisms while the consumer species depletes resource abundance, it actually increases the other species fitness. And the majority of theory work on natural selection in these interactions is focused on antagonisms. And I think part of this bias might be because in natural selection in interactions like antagonisms is direct selection on the one species is a direct result of another species is foraging. Whereas in a lot of consumer resource mutualisms such as pollination and seed dispersal selection, one species is actually an indirect result of another species foraging And to illustrate that a little bit more, um, let’s think about how selection is acting on these traits into interactions between a plant and an insect. So in an interaction like your memory, obviously the plant’s goal is not to be eaten. And so traits that will be selected for, um, are things that will affect, um, how the herbivore is able to consume its tissue. So defense traits, things like that, but in contrast, and a consumer resource mutualism like pollination, the goal is not simply to be eaten, but to be pollinated by the consumer in the process of foraging. And so a lot of plants produce nectar for pollinators, and that’s a resource and those traits will certainly impact the interaction, but so will all of other these traits that will influence the consumer’s behavior in a way that benefits the plant. So in, in these consumer research, mutualisms, the resource is actually just one means to that end. And that changes the way that we can think about natural selection on these interactions. But if we want to think about natural selection in a consumer resource framework, it’s really essential that we focus on the resources themselves because it’s these resource traits that will impact most directly the ecological dynamics of the interaction And we want to understand how that interaction between plants and pollinators will generate natural selection on these resource provisioning traits that will then feed back on the ecological dynamics. And so today we’re going to do that by building a simple plant pollinator interaction using a bit of math. So I want you to imagine a plant population in a field that has an abundance of R and M for the simplicity of the math. We’re about to do. We’re going to

make a few assumptions about this plant. We’re going to say that our species is hermaphroditic, that it’s self compatible, that all individuals only produce a single flower, and that nectar is the primary, um, resource that the plants are using to attract pollinators. So think of something that has a really, um, pretty basic floral design like this Magnolia in here, not a lot of the bells and whistles, um, have some flowers that attracts pollinators So they’re coming for the nectar. And we can begin by thinking about the ecology of the nectar itself because the resource has its own dynamics. And so if you imagine a flower as being essentially a cup for nectar, um, the Flint is going to produce nectar, and we’re going to assume that follows a logistic growth curve. And so what that means is that when the reservoir is completely empty, the plant will produce nectar at the fastest rate, which will cause the NPR. And as that reservoir fills up, that rate will decline until we reached the carrying capacity for this cup And that’s, zerv the maximum nectar reservoir of the plant, but there’s actually a second part to this equation because of course, pollinators are going to impact the amount of nectar that’s in the plants cup at any one time by doing this. And so we’re going to assume that the pollinator has, um, a saturating type two functional response on the plant’s nectar And what that means is that as the plant population produces a larger crop of nectar, the pollinators will flourish on that nectar at a faster rate, which will saturate at this constant a and that’s essentially because of course, tiny little insects can’t consume infinite nectar And we also think about how that nectar is going to impact the pollinators population as its resource, as its food. Um, and so in this equation, the pollinator has a conversion efficiency B, and that essentially describes how much nectar an individual pollinator used to consume in order to produce one offspring, one baby. And that’s going to scale with a functional response on the nectar. And the pollinator is also going to have a death rate of F. Now we can begin to think about the ecology of the plants, and we’ll start by thinking about how producing nectar affects a plant without pollinators So in the absence of pollinators, we’ll say that plant growth follows a simple logistic growth equation. Um, and that these two traits that we’re following XE, NPR and CRV are going to impact the plant population growth, and that’s going to have two effects. So first we’re going to say that producing the structures to make and hold nectar is costly, and that’s going to affect the plants intrinsic rate of increase as a quadratic function of the values of its trait. And second, we’re also going to say that there’s an basically an energy allocation cost for individuals of producing every drop of nectar. And so anytime you make a drop of nectar, you’re allocating energy away from other structures. That’s also going to have an impact on your grocery And so, um, when this population evolves without pollinators, they actually evolve to lose natural production altogether, because of course they’re paying these costs of producing these structures and making doctor, but receiving no benefit from pollinators. Now, if we add in the pollinators, we add in a third term to this equation that describes the benefit that pollinators are providing to plants. That’s going to positively impact their growth rate. And we’ll say that this benefit will increase as a pollinator population size increases because as there are more pollinators, there’ll be interacting more with the plants and in the process providing a greater benefit by fertilizing those individuals. But that’s also going to saturate because of course the plants don’t have into infinite adios to be fertilized. And so when the pollinators are present, the fitness of a plant with these metric traits actually becomes a stabilizing function. And the reason why that is, is because the total fitness of this plant can be partitioned into these two costs of producing nectar and producing the structures that make it, and also the benefit the plant is receiving from interacting with its pollinators. And then lastly, because we’re using the per capita fitness form of the plant population growth rate, um, we can easily convert that into the multivariate breeders equation for these two nectar traits that we’re following maximum nectar production rate and maximum reservoir volume. And we’ll assume that these two traits have an additive genetic basis and that they can evolve independently of one another. So now we’ve said this ecological interactions and the conditions for natural selection to happen, and we’re going to use this model today to ask how does pollinator ecology affect selection on nectar? And so to do this, um,

we’re going to plant identical, starting populations of these plants in the same antibiotic abundant environment with the same values of costs and benefits from nectar, um, but with different properties in their pollinators So imagine something, I think like this and what I’m going to be showing you in the next couple of slides is the come the evolutionary outcome of a plant that’s evolved under these different conditions of natural selection imposed by its pollinators. So what I’m showing you here is a trait service for one of our traits, the plants maximum your production rate, and each of the intersections along this graph describes the evolutionary outcome of one plant population. That’s evolved with a certain combination of pollinator properties And the two properties that we’re following here are the pollinators maximum metric consumption rate. So that’s basically how quickly the pollinators are visiting the plants and eating their nectar. And so on this side, we see that, um, pollinators have, um, very low nectar consumption rates. So they’re not visiting the plants as much. Whereas over here, they have high nectar consumption rates. They’re visiting them a lot and they’re eating a lot of nectar. And then on the second access here, we have maximum pollinator conversion efficiency or be parameter. And that measures how quickly the Polynese are converting the energy they consume into offspring. So on this part of the graph, we have pollinators that don’t need an, a lot of nectar in order to reproduce And on this side, we have pollinators that have a really large energy need, um, in order to reproduce. And what you’ll notice is that plants are evolving higher nectar provisioning actually for less efficient foragers and lower nectar provisioning for more efficient foragers. And if we look at our second trait, maximum natural reservoir or volume, which remember I told you can evolve independently, we see the exact same trend Now this kind of seems a little bit counter-intuitive because don’t you think that plants would want to feed more active pollinators more because they’ll get greater fitness benefits, but let’s think a little more carefully about what nectar does to pollinators. So the first thing we know is that feeding pollinators caused them to forge more. So an evolutionary increase in either one of these traits, the rate at which you make nectar, or the amount of vector that you hold will affect your standing nectar crop of the population. So how much nectar is in the population at one time as that increases, we know that pollinators will forage on the nectar at faster rates, thereby they’ll provide a greater fitness benefit to the plants because they’re interacting more often with pollinators. Um, and we, we can also say that plants will earn a greater fitness benefit from producing more nectar for inactive pollinators, thus causing them to add it, to interact more than they will for pollinators that are always really active and visiting the plants. But because we know that these traits carry costs, we also find that when the plant evolves to make more nectar, we see a slight decrease in its total population size. Um, and that’s compensated by the greater fitness benefit that it’s getting from interacting with the pollinators. So the population growth is balancing these costs and benefits of the traits. And if we want to think a little biologically about, um, how we could think about this possible pattern that we’re seeing in nature, think about traits that will affect your pollinators functional response. So I think about instantly about things like diet, breadth, and specialization, um, if your pollinator eats a lot of things besides a plant’s nectar, it might actually be selecting for that plant to make more nectar because the plant earns a higher benefit from causing that pollinator to visit it more Whereas if your pollinator forage almost exclusively on your nectar, that might actually be selecting for plants to evolve lower provisioning because they don’t, uh, essentially if they’re not needing to work harder in order to achieve a fitness benefit. Now, the second thing we know that nectar does to pollinators is that it actually causes them to reproduce more because in this model, we’re considering nectria to be a primary food source and you need food to reproduce. So again, the increasing these true traits is increasing the standing nectar crop, and that’s having the effect of providing more food to pollinators, thereby you have more pollinators and the plant earns a greater fitness benefit because it has more pollinator individuals to interact with. And so, um, what we see is that the plant is actually evolving to make the most nectar when its pollinator population is the smallest. And the reason why the pollinator population is small here is because the pollinator has a low conversion efficiency. So that pollinator, that means that pollinator needs more energy

in order to reproduce. And because of that, um, the planter balls up there, or sorry, the plant earns a higher fitness benefit, um, from evolving to increase its nectar production When there aren’t many Polly years around to begin with. And so essentially by producing more nectar, the plant population can boost its local pollinator abundance over evolutionary time. So you have five minutes. Thank you Let’s think about a couple of pollinator traits that might impact this evolutionary trajectory So things that are going to impact the numerical response and, um, something that Springs to mind to me is things like sociality. So, uh, if you think about the structure of social versus solitary species, social species share the burden of gathering the energy to grow their colonating the rest of their colony with all of the individuals in their population, we’re foraging, whereas solitary species are solely responsible for gathering the energy they need in order to reproduce. And so what we could predict right, this model is that social species might actually be selecting for lower nectar provisioning, um, than solitary species because solitary species require a larger any energy budget. And thus their population is bumped up by having more nectar from the plants and, um, Elizabeth Crowe, and did a really interesting study in 2013, where she increased the floral density in a meadow and saw how that affected the local pollinator populations in the following year we are And what you saw is that solitary populations had a huge increase the year after those, those floral resources were increased wrench He didn’t see the same thing, social species So that’s, that’s some that there might be this ecological mechanism to generate an evolutionary benefit from the plant. We can also think about things like body size, because we know that pollinators often visit plants that have more nectar if the pollinator has a larger body size. And that makes a lot of sense from the pollinators perspective. But what our model suggests is that POL plants might actually receive benefits from evolving to provide more nectar to those pollinators who are more needy, because it increases the pollinators local abundance, thereby providing more benefit to the plants. And so just to sum up really fast, um, I hope I’ve made you think a little bit about how natural selection on the provider’s resource dynamics is an indirect outcome of consumer foraging in these consumer resource Mutualisms like pollination and mutualist resources in pollination, like nectar have to effects on the provider’s fitness and their indirect effects because they’re both mediated through the consumer species. So first there’s the trait mediated, indirect effect we’re producing more net or increases the consumers for inching rate, thereby providing a greater pollination benefit to the plant And second, we have an abundance indirect effect we’re increasing, uh, nectar production increases the consumers population growth rate, thereby also providing a greater fitness benefit to the, and lastly, um, using frameworks like this that are based on traits can generate testable predictions about how these resource traits might evolve in nature. And if you study an interaction that’s not pollination, you can think a little bit about how this general framework could have applied to your system, because the main thing I really want you to get here is that, um, we can think about the ecological dynamics as defining the evolution of resource provision in consumer resource mutualisms. And we can also think about how that can feed back on the intellectual, sorry, uh, ecological interaction itself and the rest as with all patterns in nature is up to ecological context. And, um, with that I’ll thank my collaborators, uh, Margaret Keke and Judy Bronstein, some, a lot of folks who provided really great comments on writing this paper and developing this model. And thank you for listening if there’s time. Happy to take questions got time for one quick question One really quick question, Sarah. Um, it seems like if you’re referring to plants and pollinators, they’re probably going to have really similar generation times, um, like I have a hard time coming up with exceptions to that, but I can imagine other systems where there might do, how much do you think have any sinked up generation time matters or do you think like lag effects would affect what’s going on there? Oh, that’s really interesting. Um, we didn’t think explicitly about time in this interaction. We just thought about how those Ms. Selective conditions are going to change. Um, but lag, it could certainly affect the pace of change, but I think you’d still eventually come to the same outcome,

um, because the same conditions of natural selection were still present. They’re just shifted if that makes sense. Yeah, that makes sense. Thank you. Excellent. Thanks for that, Sarah. Thanks, Casey. Um, we’ll, we’ll move on to our next speaker last talk for the first half of this session, Alex and Alex will, um, tell us about resource investment trade off between growth and your production in two species of Detra and Alex. We can see your screen now. Great. Excellent. And over to you. Great. Thank you so much. Hi everyone I’m Alex carnage. I’m from the university of Arizona, and I’ll be talking about how organisms allocate their resources to life history functions, and mutualisms so over the course of an organism’s lifetime, they need to make choices about how to allocate their resources. Um, so this is usually split up between four main life history functions, maintenance, growth, defense, and reproduction. And so when they’re allocating those resources, they have sets of priority rules of which determine the order of allocation to the different life history functions. And we found that an organisms that have indeterminate growth maintenance has been seen to have the highest priority And this has been seen in water fleas, clams, butterflies, and spruce trees. So mutualisms are embedded in these life history functions So mutualisms are crucial for the growth, survival and reproduction of any species while also usually being costly when it comes to resource investment. So kind of my big question is how our mutual is mediated life history functions prioritized when it comes to these priority rules. So I’ll be focusing kind of with plants and how they allocate their resources, mostly because they’re settles. So they can’t move, which means they’re probably more vulnerable to local resource limitation and for plants, uh, they allocate their resources to growth and reproduction. Uh, so for growth, they invest resources to produce root stems and leaves. And then for reproduction, there are kind of three main categories where they can, uh, invest their resources. The first is seed production. So putting resources into producing pollen and, that will eventually become seeds pollination, where they put resources into flowers, uh, to attack, attract, and reward animal pollinators, and then seed dispersal, uh, where they put resources into, um, producing structures that attract and reward animal dispersers. Um, and with in this kind of general, uh, question about how, um, priority rules factor in mutualisms, I’m interested in how different life history strategies affect how mutualist mediated life history functions are prioritized. Um, right now you’re probably wondering how we find these priority rules And this is usually done by manipulating resource availability, um, either by doing starvation experiments or by having separate treatments where there are different levels of resources available. And then once you set up the, these treatments, you look for resource trade-offs between low priority and high priority functions over to kind of see, which are the most important and down. So a high priority life history function is one where when resources are reduced investment into that function, doesn’t change and the size or magnitude of that function doesn’t change. So for example, if we say that growth is a high priority, a life history function, when organisms are grown in an environment where they have access to fewer resources, uh, those organisms will be the same size and grow to be the same size as those that have more resources, a low priority life history function is one where investment is reduced when resources are reduced. So the size or magnitude of that function, uh, is decreased So, uh, if we used growth again, for example, uh, if growth is a low priority life history function, when organisms grow in an environment where there are fewer resources, they’ll grow to be smaller and fewer resources invested into growth, then organisms that have access to more resources And then when I’m showing these priority rules, I’m going to be using these resource allocation diagrams. So these are just kind of a toy conceptual figures to illustrate other priority of different life history functions. So the total length of the line is the amount of available resources that the organisms can use. And the subsections, uh, represent the amounts allocated to separate life history functions. So here the gray box is the box that represents the amount allocated to maintenance So moving on to my life history hypothesis, when we look at annual plants, so they grow over one season, they have one chance to reproduce I hypothesized that they should prioritize

investment into reproduction and create our producing reproductive structures at the cost of growth. So when we compare plants that are grown in a high resource environment, to those that are grown in a low resource environment and investment into growth should decrease, um, but investment to the different aspects of reproduction. So pollination seed production and seed dispersal should remain the same, um, moving on to perennial plants So those that grow over multiple seasons and have multiple chances to reproduce, um, I hypothesized they should make reproduction a lower priority than growth and maintenance So when plants are grown in a low resource environment with compared to those in a high resource environment, other investment into different aspects of reproduction, so pollination seed production and seed dispersal, uh, should decrease, whereas investment into maintenance and growth should remain the same. And then finally, um, I thought this is for plants that are able to, uh, reproduce autonomously. Uh, so that means that they have flowers that are able to be fertilized from the pollen, from their, uh, same flower So I hypothesized that species that are primarily autonomous, uh, should make pollination a lower priority because they don’t need to really attract and reward, uh, animal pollinators in order to produce seeds. So when, uh, plants that are primarily autonomous are grown in low resource environments, their investment into pollination should be reduced when compared to plants grown in high resource environments So to do this, I’m using species within the genus Ditra, uh, the Jimson weeds they’re located in the Solon AC the tomato family, and there are 18 species are really well resolved phylogeny. Uh, right now I’m just looking at two different species, but don’t worry, I’m planning on eventually expanding the number of species I’m, uh, studying to kind of make broader, uh, kind of claims about how life history affects resource allocation. So the two species I’m presenting on today are detour a discolor and detour variety. I deter, I just color is an annual species. Um, it’s relatively small of a nature when they have plenty of resources here in desert, like water, um, they might grow to maybe a meter across, and these plants are primarily autonomous So, um, they have their answers located just above their stigma. So before the flowers even open, those answers will dehisce and release pollen onto the stigma. Uh, essentially fertilizing the flower before it even opens on the other species to turn a righty eye is a perennial species. Um, it’s relatively large. So in nature, uh, here in Arizona, it might get to be between two and three meters across when it has plenty of water and this one can self Pawnee, but, uh, is not, uh, as primarily autonomous as the charges color, it still as the available to do it, or it still is able to do it. Uh, but doesn’t do it as the same high rates that discolor does. Both species are pollinated and eaten by hoc mods and their larva, and both have seeds with Elias homes attached. So Elias zones are nutrient rich fat bodies that attract and reward ant dispersers So these, both these plants are dispersed by Ana. Yeah. Uh, moving on, I grew plants in a greenhouse in sand meant two nutrient treatments. Uh, one is my high nutrient, so high resource treatment where I gave them a regular strength solution of hokum solution Hoagland solution is a nutrient solution that contains all the macro and micronutrients plants need in order to survive. And then I have my low nutrient of low resource treatment where I gave them a quarter strength solution of Hoagland solution. And then, um, the flowers, either hand pollinate or left alone and split into three different pollination treatments, because I just wanted to see the effects of pollination on seed production. And those are a self-treatment where I hand pollinated the flower with pollen from its own flower across treatment, where a hand pollinated, a flower with pollen from another plant that was blooming that night, and then an autonomous treatment where I left flowers alone to self pollinate. So just a quick comparison, um, and an example of how I’ll be presenting my data for the rest of the presentation on the left. I’ll always have the animal species deter, just color and on the right I’ll always have the perennial species deter righty eye. And then within the figures, I’ll always have the high nutrient treatment on the left and the low nutrient treatment on the right. And you can see in these pictures that plants, um, in both species that had access to fewer nutrient resources were generally smaller than plants that had a lot more resource. Uh, so Mo looking or moving into biomass and the plants investment

into growth. Um, our results are consistent with the annual hypothesis, but not the perennial hypothesis. So deter a discolor, the annual species when it had fewer resources and had lower above ground biomass, whereas it’s below ground bio-mass was not affected by the treatments I’m not shocked by the below ground biomass being the same, uh, determined does color be an annual species doesn’t produce that much below ground biomass to begin with. So the Turo just colors, investment into growth, at least for above-ground bio-mass was reduced and fewer resources, which is with our annual hypothesis deter. Our righty, I are perennial species when it had fewer resources reduced both above and below ground biomass. So reduced its investment into growth, which is not consistent with our perennial hypothesis, where we hypothesized that perennial plants would maintain their investment into growth. Moving on to investment, into pollination, uh, for Detroit discolor, our PR our annual species. Now all the measurements are consistent with our annual hypothesis and the are inconsistent with the autonomy hypothesis. So the number of days until first flour and the size of the flowers was the same in both pollination treatments, indicating that investment into pollination was maintained, but the total number of flowers produced was lower wood plants grew in an environment with fewer resources Um, so it’s not completely consistent with our annual hypothesis. That annual species would maintain investment into pollination It is inconsistent with our autonomy hypothesis where we said plants are primarily autonomous, which deter just color is, would decrease their investment into pollination, um, which is especially shocking because the flower size remains the same in both treatments And if this plant ostensibly doesn’t need to rely on animal pollinators, uh, to get their, uh, seed produced, it’s kind of strange that they keep their flowers to be the size So these are flowers about 16 centimeters long they’re relatively big flowers. Um, moving on to deter a variety eyes, pollination investment. This are perennial species. The results are inconsistent with the perennial hypothesis, but consistent with the autonomy hypothesis. So for a term variety, I, the number of days until first flower, the total number of flowers produced and the size, the flowers is the same in both nutrient treatments, indicating that investment into pollination is the same, um, rather than being a blower as we had a pot size with perennial species, but this is consistent with our Tagomi hypothesis. So is not as autonomous as a turdus color. So we hypothesized that they would maintain their investment into pollination, moving on to investment in seed production Um, our results are consistent with the annual and perennial hypothesis. So Detroit discolored, the annual species, uh, pollination treatment had no effect on the number of seeds produced and nutrient treatment had no effect of the numbers. Uh, number of seeds produced, which, um, is consistent with our hypothesis, that investment into seed production would be maintained an annual species. And then for deter RIT, I, the perennial species, uh, pollination treatment did not have an effect on seed production, but nutrient treatment did. Um, so when plants have access to fewer resources, they produce fewer seeds, which indicates a reduction, uh, in investment to seed production, which is consistent with our perennial hiphop. Yes, just to summarize the species really quickly So to turn this color annual species, when resources are reduced, uh, investment in growth is reduced Whereas investment in reproduction is maintained so that its investment into pollination and seed production, uh, is the same and growth has reduced, is consistent with our annual plant hypothesis and inconsistent with the autonomy hypothesis, uh, for deter RIT. I are perennial species when the resources were reduced, investment in growth and seed production were reduced and investment in pollination was maintained. So this is inconsistent with the perennial hypothesis where we said that perennial plants would maintain growth and reduce investment into pollination and seed production and other aspects of reproduction and consistent with the autonomy hypothesis that plants that aren’t as autonomous should maintain investment in pollination on just some take home messages that I want you to take from this presentation. The main one is that investment in some mutualism mediated life history functions, uh, decreases when resources are limiting. So resource availability does affect neutralism investment. And the pattern of those reductions are related to life history traits. So whether or not plant is an annual or a perennial or, um, whether it relies heavily on animal pollinators can affect a resource allocation. And just to

kind of all look forward, you may have noticed that I didn’t talk about us seed dispersal investment. Um, currently I’m measuring changes in investment to dispersal under resource limitation by measuring the mass of the seeds and Elias zones. And I will start, uh, measuring the caloric content of those Elias zones, um, in the future, in the near future. And I’m really excited to start looking at that. And I’m currently working on adding more species of deterrent to this experiment. So I can kind of make broader claims about how life history affects resource allocation. Um, as you can see here, these are the main six species that I’m interested in using, and they kind of, they are, there are a bunch of annual species, more perennial species, a different kind of autonomous autonomy and selfing rates between those as well as a group of deterrent species, uh, that have completely lost in Elia sown. So rather than being amped dispersed, they are, um, just kind of passively dispersed. I’m really interested in getting those, looking at those species to see the effects of losing, um, a mutualism on a allocation of those resources, to the remaining mutualisms. And with that, I’d like to thank a bunch of people. Uh, my advisor, Judy Braunstein, and my committee, as well as five members and especially my, uh, undergraduate researcher, Gretchen Gabler, who did a lot of seed measuring and counting. And with that, thank you for listening. Thanks so much for that, Alex. And we certainly have time for some questions. I’ll ask a question, Alex, while we’re waiting for people to type in questions. Um, I really liked was talking to him. I liked the, your conceptual model, these boxes, and whether they can fluctuate in size and stuff. Um, I’m wondering how it’s assuming all those, um, life histories processes are independent of one another, but you can imagine where they might not be where, um, like say if you invest a lot of energy and growth, that’s going to provide more resources because you gain more light or nitrogen or something, and that might give you more resources to invest in reproduction later. And I’m wondering if that matters or not to sort of, um, or, or if that has any effect on sort of your empirical results. Um, it definitely matters, especially, I mean, everything in this, um, is connected. So one of the things is the plant, the does color didn’t produce as many flowers, um, when to have fewer resources And that’s probably because it wasn’t able, it had reduced growth, so it was smaller So that there’s kind of a direct one of like that limits the amount of places you could put flowers, especially with the tourist species Usually they’ll kind of branch and make two leaves and then have a bud that turns into a flower. So you don’t have enough, these branching points you’re really restricted to the number of flowers you can re produce Um, which kind of is like a concrete thing that I’ve really noticed. And then for resource acquisition. Yeah, it certainly matters a lot of the stuff with kind of resource allocation has been with animals So I kind of tend to think like animals only have one mouth, whereas with plants, they can put in more resources to make more mouths, to get more resources. Um, kind of, uh, I’m trying to, sorry, I’m thinking language right here to be articulate, um, to kind of combat that, um, the way I’m kind of doing it is to really, um, really stress these plants Um, cause you could imagine if they have enough resources, they can put enough resources to make structures, to get more resources, but if you stress them enough, you’ll hit this point where if you put resources into resource acquisition, um, you don’t get a positive response. Like you’ll gain more resources at some point. Um, that makes sense that that’s the thing that’s perpetually going on in my mind, especially the plants in, they can do anything. Great. Um, a couple other questions on the chat, um, Sam Watson S have you thought about looking into how investment in root growth changes with of resources? Um, I was trying to grab that by looking at below ground biomass. Um, it’s certainly something I could look at and that I probably will look at, especially if deter righty. I didn’t really talk about it, but they do, uh, form this really large taproot, um, that they use to store resources for the next season. Um, so certainly it’s very important, um, with these I haven’t, I, so for perennial species, I didn’t look at the following season just because I wasn’t able to keep them in the greenhouse that long. Um, but there are probably some effects on root growth, especially perennial

species on maintaining, um, growth the following season. But I did see a decrease in below ground biomass. So they are saving resources from putting herbs. They are shunting resources there for the next year instead of putting them in what they have this year. So great And then one quick question from Elizabeth Moore is, have you thought about looking at other metrics or pollination investment, um, like insect attraction traits or, or odors, or might of nectar or other things like that rather than just flowers? Um, yes. So that’s kind of another aspect of my dissertation that I’m going to eventually be doing, um, is kind of growing these plants under these with different amounts of resources and then seeing, um, then bringing mods in and seeing whether they’re attracted them or not other, having fewer resources changes their level of attractiveness, um, to mods. Um, I did try to measure nectar in this experiment, but for some reason the plants weren’t producing vector in the greenhouse, I have no idea why, um, but I, I would, I’m going to grow more plants and hopefully get answers for those parts of pollination. Great, awesome. Excellent Um, thanks Alex for the great talk. And I would like to just take this opportunity to once again, thanks all of our speakers today. Who’ve had fantastic presentations and discussions as well. Thanks everyone for participating. We are officially on break So please take this opportunity to stretch, hydrate and refresh, and we’ll be back again at, um, what’s the time in the U S at 5:40 PM. Um, and join us again at that point to continue the second half of the session. Um, and there’s some very exciting talks in that session as well. So we look forward to seeing you again. All right, welcome back everyone And we’re going to start the next half of the species interaction session. Um, our first speaker is Karena basket and she has a very interesting talk title. So I had to go and look at her abstract. Um, she’s going to be telling us about flower color in wild snapdragons and hybrid zones. Karina, can you please share your screen with us? Yep. Is it looking good? Yep. That is perfect. I will mute myself and it’s over to you now. Okay, thanks. So, hi, I’m Korean, I’m a, post-doc in the Barton group at ISD Austria. And, uh, I wanted to thank the organizers for this virtual meeting It’s my fourth ASN meeting. And I hope that if, uh, anyone, if it’s their first, I guess a lot of people, it is that you might join next time that it’s in person, that we can all be together in person. Um, and so this talk is my Valentine to the Snapdragon flowers Roses are red, violets are blue. Oh, dragon flower. What do you do? And how do I answer that? Okay. Um, so a lot of us in evolutionary ecology, we’re working on the wonders, the puzzling wonders of nature. So we want to fit together these puzzle pieces of genotype and phenotype and fitness. And, um, I’m particularly interested in where phenotype and fitness intersect. So slight natural selection as the correlation between phenotyping fitness and then what are the ecological drivers of selection what’s causing that selection? And so in some systems, different parts of this puzzle are easier than others. And so for example, um, and these two minimalist sisters species are one is be pollinated. And when it’s hummingbird pollinated and they have very different flowers, it’s, I mean, you still have to do work to show it, but it’s a little bit more obvious from the get go what’s going on with ecology and phenotype And so we can fit that part together. And then the really challenging part at least back 20 years ago, when this work was done, was getting at the genetics and doing lots of, um, crosses of populations and GTL mapping to understand what are the genes underlying those, um, all these puzzle pieces. Whereas in the other extreme, there are cases like hybrid zones, which are natural experiments where all of that crossing has already been done. And, um, so in that case, uh, it’s much easier to understand what’s going on with the selection on phenotypes and maybe even genotypes, but the ecology can be totally mysterious. And that’s what I’ll be talking about here. So, um, but first, some more details about hybrid zones. So we have, um, two different species, or it can

also be two different populations, uh, that come together and exchange a lot of genes in these hybrid areas. Um, there’s a lot of gene flow except for traits or genes that are showing Clines. And so where you have on one side of the hybrid zone, one trait value, and on the other side, another trait value, and you can actually, um, estimate long-term natural selection, um, by looking at the client with, so how steep is it and looking at dispersal distance. And, um, so we already have then three pieces of our puzzle there. We have selection and phenotype and fitness, just like having a hybrid zone with trait cleanse. And, um, so I’m going to talk about a hybrid zone in inter Rainham. This is, uh, snapdragons that are native to the Spanish, Earl native to the Iberian peninsula And we work on them in the Spanish, Spanish Pyrenees. And so we have a hybrid zone here where we have these pair Patrick’s subspecies, or you might just want to call them populations, but they have different flower colors. And that’s their only distinguishing trait that anyone knows about. Um, and they usually live in different parts of the peer needs, but where they meet hybrid zones are readily formed. And, um, here we have an example where there’s a Valley that, uh, has two roads and upper road and a lower road, and they have the same kind of hybrid zone across both roads. And, um, probably these hybrid zones were reformed when people started making roads and railroads. Um, this one was recorded over 150 years ago, so it’s pretty stable. Um, yeah, I think that’s what I need to tell you about that. And, Oh, yes. And so then we have this really steep flower colored Cline and this hybrid zone it’s over only two kilometers And so we know that there’s really strong selection on fire color. Um, and we know the genetic basis because people have been using snapdragons as a model system to look at genetics of thought or color for a long time. And so you don’t have to memorize all these names or anything, but just that there’s one gene that controls whether there’s a yellow or not. And then another gene that controls, well, it’s actually two very tightly linked chain, but it kind of acts as one that controls whether there’s high, medium, or low magenta And so there are six different phenotypes here. Um, it’s a pretty simple genetic basis and that’s nice because we can actually use the phenotype to get the genotypes, so we don’t need to actually change the type them So that makes it pretty easy to do evolutionary genetics in the system Um, and so we have these, a lot of these pieces here, but what is the ecology here? What are the selective agents on flower color? Why doesn’t the magenta just March across the Valley and invade the yellow or vice versa, or why don’t they just co um, you know, overlap completely, you know, have hybrids all over the place and yellow and magenta all over the place. Um, and so that’s the question I am interested in getting it here. And of course, um, the first thing you might think of with flowers is that it must be pollinators Um, and indeed, uh, snapdragons are self incompatible and they’re pollen limited. So pollination is an important step in their life history Um, however, the major pollinators are constant across the hybrid zone. So we don’t have, you know, one B on this side of the Valley and another over here, they’re, they’re the same big bumblebees. And so, um, at least in that sense, there’s no differences. Um, and then also color is not a signal of different shape or reward quality. Um, it’s not, it doesn’t, as far as we can tell, it’s not associated with other traits, um, or scent or anything like that. And, um, there’s some genomic work I don’t really have time to get into showing that, uh, we think that the flower color regions of the genome are the only ones that are different between the subspecies in this area. So it seems like it’s the only thing that’s different So it’s not actually that clear that the pollinators should really care about power color if it’s not associated with different reward or shape And also furthermore, these awesome papers have come out in the last time a few years saying that there are lots of non pollinator agents of selection on floral traits and there’s this great meta analysis recently. Um, and one thing I really liked about the met analysis was they, uh, differentiated between different types of traits. And so they found that pollinators selection is weaker for traits, um, dealing with attraction compared to actually like the fit of the pollinator in the flower and flower color is of course an attractive trait So, um, so I think what all of this means

is that, um, when we’re asking this big question of what are the selective agents on flower color, actually, what we need to do is first ask the question, what is the ecology of flower color? And go really broad with that question, because we don’t know that yet with our hybrid zone, we have all these other pieces, but we don’t really know much about the ecology So, um, with that, I wanted to look at both biotic and abiotic factors of the environment So I’ll start with the biotic observational studies in the hybrid zone. Um, and so here, I’m asking why your flower color or hybrid status predicts interaction probability for several interactions, and I’m putting them into these categories of their potential effects on fitness. Although we have not measured their effects on fitness. Um, and so first pollination, um, and here we were looking at visitation and, um, and I would call that a direct effect on reproduction. Um, and then we, I don’t really have time to go into the natural history, all of these, unfortunately So after zoom through, but we have seed predations and specialized renews. So weevils, uh, forgive me, I put a question, right? Cause I’m not totally sure what these little levels and the flowers are doing, but they seem to be eating the audience, but I have trouble rearing them. Um, and then in more indirect effects on reproduction through for free, um, which would maybe affect pollinator visitation But I, that one, I wasn’t sure which category to put it in. Cause I think a lot of the times the Florida borders are also going for the pollen, they’re eating the pollen. So that might be more direct. Um, and the nectar robbing So the same Bombus trust that is pollinating is also robbing. And then we have some, uh, effects on plant size that could be if flower color is pleiotropic for leaf trades or other defenses traits through our Bovary or rust And so it’s a lot of interactions. And for each one we have measurements of it as a presence or an absence of the interaction on many flowers or fruits or plants, depending on what kind it is. Um, here’s how many, so some of these we have over a few years and, um, so out of when you add up all those years and sample sizes, it’s quite a few plants and flowers and these datasets were collected all the way from 2004 to this year. Um, so it’s been really great to, to ask people who’ve been working on this system, Hey, do you have any data about this stuff? And to convert it all to is to a similar format that we could analyze in the same way. And so I have to think all these people who’ve given me their data. Um, and so I’ll just show you a summary of their results. Cause I don’t have time to get into the details. Um, so here for each one, the symbol there’s a symbol for each year that it was measured. So for example, pollinator visitation was measured over four years and there was no effect of hybrid status. That’s why it’s an X and there’s a few little symbols I’ll explain in a second. Um, so, well first what you can see is a lot of Xs. So there’s a lot of, um, non patterns here where flower color and hybrid status don’t affect these biotic interactions or the probability of the interaction occurring. Um, and so I’ll go down this hybrid status column. So we see this is the most exciting one where, uh, being a hybrid, it means that there’s more for every, so it’s worse to be a hybrid in terms of, for every now, whether that really matters for the plant fitness. I don’t know, we’d have to look into that. Um, but I think what’s really interesting about that is this was one of the stronger effects and same with this, um, color effect here. And, um, what I think is cool is that even if we’re not sure what the effect is on the plans, we should probably expect that the one inside here that’s actually eating the flowers would be affected the most by the flower color. So I thought that was pretty neat that it, that those effects were stronger, um, for delivery riverine rest, we actually see that the parents, uh, do worse than the hybrid so that doesn’t help explain the maintenance of the hybrid zone. Um, and then these other color effects they’re inconsistent, but just essentially like being yellow is better for pollination and for, for seed predation It just depends on the year, whether magenta is better or worse and uh, for a very, in one year being Magento was associated with higher flurry. So, um, there’s a lot of inconsistency here and, uh, but I’m still gonna follow up with some more analysis of spatial patterns and try to look at whether it flower, color frequency, like local color neighborhood affects these interactions. So, um, getting to our April attic effects or factors, um, here, I focused in on drought tolerance and that’s because it’s linked to flower color polymorphisms and multiple systems such as this winning this period, um, and, and several others

And, um, that’s because anthocyanin pigments play a role in stress tolerance in general Um, and so, yeah, so, uh, but okay, so we don’t have a ecosystem with, uh, it’s not a desert, it’s not a Mediterranean system with annual severe drought, but, um, I do see that precipitation various tenfold over 30 years. So it seems like there are times pretty frequently when there’s much more and much less precipitation, so it could still be, um, an important selective factor in the area. And so I did a preliminary drought experiment, um, looking at, uh, so used allopatric crosses because somebody was growing them for something else. And I thought I’m just going to mess with these plants in the greenhouse and see what happens. Um, one thing I, I rather like about this experiment is the way that I’m doing the drought is, um, I water the controls when a few of them start wilting. So when the plants tell me that they need water, then I water the controls. And every other time I water the droughts. And so over course, several of these, um, treatments, they do end up quite stressed out and I ended up killing quite a few plants, like just the perfect amount, you know, like not all of them, but so now not like a dry down experiment where you kill all of them. Um, it killed like an intermediate amount and that was good. Um, and so I found actually that the F1 hybrids had poor survival, especially compared to the magenta parents. So that’s pretty cool and intriguing. Um, to me, this is more intriguing than the results of all the insects stuff. Um, but these are from allopatric populations. And so it could be that stuff other than flower color differs between these populations and is causing these effects. Um, whereas with our hybrid zone plans, they should be really similar genetic backgrounds and only for our colors different And so of course the next step is that we’re doing an experiment with hybrids and seeds This is happening right now in the greenhouse I don’t have results yet. They’ve only gone through one cycle of drought. Um, but at least I can show you cool pictures of anthocyanins, of being expressed on the underside of the leaves. We’re not totally sure why this is, um, why they did this in the greenhouse, because they don’t actually do that very often in the field. Uh, it doesn’t seem like a very stressful place, but we definitely see a lot of variation and the science. And so we’re excited to see whether that has anything to do with drought tolerance and whether that has anything to do with fire color. And, uh, this work is with Sean’s Dan Koski. Who’s done some work on drought tolerance and Minneapolis and Alex twister, who was a intern in our group. Um, what’s that, that’s your five minutes. Okay Thanks. Perfect. Um, so in conclusion for my poem, my Valentine, uh, if I asked the snapdragons, what do they do? I think they would say my functions are many. Yes, it is true. So much data you have yet hardly a clue So I feel like even though we do have credit quite a bit of data, um, I still don’t know very much about the ecology of flower color in this system. Um, but I’m really excited to see what happens with this drought experiment Um, if, if we see no effects there, then, then I’m not exactly sure what my next step will be. Um, I guess trying to look at the effects of slavery and seeing if that’s really associated with a lot of, um, pollen theft or not, um, or maybe digging in more to better measurements of pollination, cause these are pretty rough measurements for sure. Um, so yeah, thanks to lots and lots of field assistants And I just want to quick plug if you’re looking for a field manager or lab manager, um, let me know, because it is really great and she’s on the job market because we can’t, uh, we don’t have a full-time managers, so, um, he questions. Yeah Thank you. Thanks Corrina for this very colorful talk and good luck with your experiments We have time for a couple of quick questions and thanks for bringing the poetry Creighton essence. Very excellent. Um, uh, in the chat, uh, guest on ask, do you know how about, do you know how the whole genome or other genes not like with coloration vary along the claim? Yeah, I, um, here let me find that slide Yeah. So, um, this was a study where they do, they did an FST scan, um, along the genome And so they find that at the scale of the hybrid zone, if you just look a few kilometers apart and look at magenta and yellow that are kind of right outside the hybrid zone, you see that there’s really low differentiation across the genome, except for genomic regions that contain flower color genes. And then

if you go further away, like not even that far away, 25 kilometers or more, then there’s more and more differentiation. So, um, so there might be interesting things going on with local adaptation that are totally unrelated to flower color. Um, if you go further away, but in this area, it seems like the only differences are far color. And then related to that, uh, Sarah peek is asking, um, she’s interested in whether other traits besides flower color could be contributing to the biotic effects you observed in your studies. And do you have any thoughts on traits that are correlated with flower color that might affect animal interactions with hybrids, but perhaps might be more difficult to discern or measure? Yeah, I think in this system, um, I’m not sure. So there has been some stuff, or some people have looked at scent in the system and they haven’t found any differences in scent. Um, but I can’t remember whether that was with a hybrid zone plants or other populations. Um, but I think because of this kind of evidence with the genomic work, um, it seems like it’s probably just follower color here, but, uh, but of course there are other traits that are probably impacting the interactions with the, um, with the insects, like, you know, like nectar and things like that. It’s just that they’re not probably associated with flower color. Um, so it would be a good system to look at selection on, um, other trades, especially cause, um, we’re working here, we’re building a pedigree. And so we have some really interesting ways to measure fitness over long time using the pedigree So it would be great to like really dig into other traits and look at how those affect the biotic interactions. Cool. Great. Um, excellent talk, thanks for that Karena. Thanks, Casey. Um, next up we have, um, Anika Rose person and she’s gonna tell us talk about, she’s going to talk about a very hot topic, which is the effect of early snow melt on plant pollinator interactions in Alpine systems. Um, Anika, are you able to share your screen with us and I think you’re on mute. Yep. We can hear you. Okay, great Over to you now. Okay. Thanks so much. Um, here we go. Okay. Um, thanks for the introduction Yeah. My name is Ana Kerberos Pearson, and I’m really excited to share my work with you today. So I’ll be talking about the effects of early snow melt and topography on pollination in an Alpine subalpine ecosystem. So as you all know, mutualisms are crucial interactions across all different types of organisms. They are critical to the generation and also the maintenance of biodiversity and insect pollination is one of the most important mutualisms on the planet. Almost 90% of plant species rely on insect pollination to reproduce. And additionally about a third of human based food crops are, um, require, uh, insect, pollination. However, global climate change is having effects on all kinds of species interactions across the planet, um, including on, uh, pollination and climate change can impact pollination and all kinds of ways. I’m going to talk about three really briefly. Um, so, uh, warming can drive, uh, plants and also insects to change their spatial ranges in particular to shift upward in elevation. But this can cause a spatial shift between plants and pollinators Additionally, um, physiological changes can occur that, um, cause plants to, uh, uh, sorry to produce different types of nectar and volatile organic compounds that attract different groups of pollinators to them. Additionally, um, the phenological shifts, which I’m going to focus on today can lead to mismatches between plants and pollinators. And just to be clear, phonology is the timing of an organism’s life history events. So if, for instance, a plant’s phonology is advanced because of climate change and its pollinators phonology is not mirroring that same magnitude of advance. We may see some sort of mismatch between plants and their pollinators. So I’m interested broadly in how phenological shifts affect Alpine pollination Um, so we know that pollination and seed production influence plant demography and community dynamics in the Alpine. Um, and some previous work has shown that there are detrimental effects of early snowman on plant fecundity. Um, and this can be via exposure to late spring frosts

Um, however, this has been disputed a little bit recently. Um, although we do see a lot of like late, uh, late spring, early summer snow events, I saw a couple of those this summer that did have an impact on, on some flowers, not necessarily in my subplots, but it definitely is something that happens. Um, climate change or sorry, early snow melt can also affect plants via, um, extended summer where in plants that, uh, green up earlier on because of climate change, maybe experiencing a longer period of drought between snow melt in those late summer, monsoonal rains that help rehydrate the soil. However, we do know that topographic heterogeneity can help drive how communities respond to early snow melt and in the Alpine Tundra specifically, um, topographic heterogeneity can cause localized soil moisture gradients that can affect the length of the growing season and also the phonology of plants across the landscape Um, and so because of that, um, systems like the Alpine Tundra, which have a very, uh, heterogeneous typography may be buffered against the effects of climate change. And this is just another example from the summer where a tiny shrub caused a huge snow drift to form Um, and that lasted about about two weeks So even something as small as a shrub can affect, um, where snow accumulates in the Alpine. So overall I had two main questions Number one, how does early snow melt affect pollinator visitation rates in an Alpine subalpine ecosystem? And number two, how does snow melt early snow melt affects pollinator diversity in the system? So our study system was located in, um, at Nightwatch Ridge, which is, uh, just outside of Boulder, Colorado, just up the Hill in the Rocky mountains. And we had four sites there that were located in sub-alpine and Alpine environments that are from 34 to around 3,500 meters in elevation. So that’s just at an above tree line. And these sites are in, um, a diversity of topographies with slightly different elevations, but pretty different aspects. And, um, concave cities conduct studies. And at each of these sites, we had two 40 by 10 meter plots Um, one of which has experimentally advanced snow melt, and one with unmanipulated snow Mount, which I’ll talk more about in a second And in each, um, at each site we had five subplots in each, uh, treatment type, um, that were two by one meters in size. And these were also located within, um, diverse topographies within the, within the site itself. So it ended up looking a little bit like this. So we had a 40 by 10 meter, um, plot with experiments We advanced, um, snow melt out date, which I’ll be calling my early plot from now one and also a control plot without, um, without any type of experimental manipulation, just adjacent to the early snowmelt plot. And within those, we had the five, um, two by one meter subplots. So maybe wondering how we advance snow melt if you’re not familiar with, with Noah. Um, so a few years ago, uh, Karen Forrester, bill Bowman, and a bunch of other amazing folks initiated this project called the black sand project where we add black sand, um, to the surface of the snow in early spring, which decreases the elbow of the snow and causes snow to melt out up to two weeks earlier than in control sites. And this is effect, this has been shown to affect the phonology of some Alpine species and that work is ongoing So look out for a paper by Kiera Forrester in the next few months, but, um, as you can see from this image here, it does have a pretty severe effect on, on, um, snow melt about timing, at least in some sites So to look at, uh, pollinators in these different treatments, I performed at least 10, 15 minute, uh, pollinator observations in each subplot Um, and I performed these from when flowers open to when they sent us. And then I collected the first two morpho species of each visitor in recorded the identity of the following visitors. And then I also collected some information about the environment. So, um, I looked at the number of flowers, subplots, and also at the site wide scale. Um, I also recorded some weather and soil moisture data, and then to examine how topography affects, um, the interaction or affects the effect of climate change on pollination. We calculated, um, aspect slope and topographic position index, which effectively tells you how concave or convex that an area is. And this can be calculated at either a local or a more broad scale. Okay So before I get into the, the real data, I wanted to just share with you a kind of an overview of how effective the black sand treatment was on each of the sites. This is really preliminary, um, and super coarse data that I, um, created and we’ll have better, more specific metrics, um, in the next few months I hope. Um, and this is largely work here at Forester’s expertise

So this is just some basic stuff that I came up with But, um, so here on the Y axis, we have the difference between peak flowering of the early snowmelt plots and the controlled snowmelt plots in days. And on the X axis, we have the different sites. So you can see that, um, some sites like for instance, site number three here in the light green, um, there was a 24 day difference in peak flowering between the early and the, um, could control multiple sites. I mean, treatments. Um, whereas site two, there were, there was like very little difference or at least with this metric, no appreciable difference between the early and the control, um, snow Mount sites. Um, so I’ll be referring back to this, this difference in peak flowering, um, days as, as we go So, um, over the course of the summer, I was able to observe around 2000 total insect visitors and I collected around 300 of those. Um, so far I’ve identified about 95 morpho species, but I’d expect that number to increase as I get into the lab, um, and, and start identifying, um, past more of those species. And, uh, dip Tara was the most common order. We had a lot of circuity flies, um, but there were also quite a few lepidopteran, some ants, um, some coleoptera and also of course, bees. So just, um, to get into the data, I wanted to show this, uh, graph that shows the number of visits per observation, period, by the day of year. So on our wax is we have the visits per observation period on our X We have the day of year, and this is only at site number one. Um, and the treatments are going to be shown using color. So the blue dots and lines are representing the control subplot or the control treatment, and the red will be representing the early snow maltreatment So, um, just looking at this data initially, we can see that the early snow melt treatment appears to be peaking in insect, um, visitor abundance a little bit earlier than the control snow melt is it also appears to be peaking quite higher. And I’m not totally sure why that is, but we’re hoping to hoping to glean more about that as we go. Um, now I want to keep in mind that this is one of the sites where there were seven days between peak flowering between this early in the control snowmelt treatments. So keep that in mind. Um, and now we’ll show you all four sites and her visitation, um, changed over time in those sites. So across the sites, we can see, uh, somewhat similar patterns. So it’s site three We can see that, um, early snow visitation is peaking a little bit earlier than controls Normal visitation pattern might be reflected in site four, but it’s not super clear. Um, keeping in mind that again, these did have different, the, the treatment had different effects at these different sites. So to understand this in a more statistical way, um, I did a GLM E R and I looked at, um, a large number of different variables, environmental variables as all of these flower variables and the treatment effect. Um, and it turned out that the days between peak flowering. So the effect of the treatment on the sites was significantly affecting visitation rate, um, as was the number of flowers in each subplot as was the day of year’s interaction with the days between peak flowering. So this is not righteously surprising, and it’s something that we can kind of glean from, from his graphs, um, pretty in a straight forward way. Okay. So, and to understand, uh, pollinator diversity, I went through the same, um, kind of group of, of, uh, graphs and statistics. So here on our Y axis, we have, um, insect visitors, Shannon diversity, or Shannon entropy, and in our x-axis we have the day of year. And again, we’re showing, um, the control site or the control treatment in blue and the earliest nominal treatment in bread. And this is again, just for site one to start out. So here we don’t see that much of a pattern that’s caused by the treatment effect. Um, we do see that there’s kind of this UDA modal, um, shaped to the data where diversity is peaking at the middle of the summer, which isn’t surprising Um, and now, again, let’s remember that there were seven days between peak flowering in the site. However, it doesn’t seem to be affecting, uh, insect diversity, all that much. Um, so looking at all of the sites together, once again, we don’t really see that much of an effect of treatment. We may see a little bit of an effect site for, um, but we’ll see what the stats are that doesn’t actually come out Um, and typically we have this kind of unit model pattern where diversity peaks at the middle of the summer, except for insight three, which seems to be a bit of an outlier there and keeping in mind that the treatments, um, did affect these sites differently. So to look at these, um, how different environmental factors and the treatment effected, um, diversity, um, we performed a GLM ER again, and we found

that the day of year it’s quadratic term, um, both had a significant effect on, um, insect diversity, um, and also the number of subplot flowers significantly affected insect diversity and did so positively. Um, and then interestingly, we found that, um, local scale topographic position index, as well as broad scale topographic position index significantly affected pollinator diversity So again, topographic position index is a measure of, um, how convex or concave the landscape is. So it’s just a measure of like, I think, um, it can be associated with how much snow accumulates in an area. So that’s something we’re going to look more into in the future So, um, just to go a little bit more in depth and hopefully visualize a little bit better, how visitor abundance and visitor diversity is changing over time. I created this little animation. Um, so on the horizontal axis here, I’m going to be showing the number of insect, um, visitors and on the Y or the vertical access. We have the individual subplots, um, on the top here, we’re showing, um, the control treatment and on the bottom in the red we’re showing the advanced snowmelt treatment. And then we’re going to be looking at only site one first. And, um, it’s going to be changing over time. So as the gift goes on, you’ll see, um, how the summer progresses and the colors are going to be representing different morpho species of insects and their abundances Okay. So as you can see, there does appear to be an increase in visitation, uh, at earlier at an earlier period in the early snowmelt treatment than in the control snowmelt treatment And because we know that, um, from the GL Mer that, uh, insect visitor was sorry, the number of insect visitors were affected by flower number. Um, I also wanted to include, uh, how flower diversity and floral number was changing over time. So I’ve included a panel on the right here, um, that will show you the number and diversity of flowers as they change in the control in the earliest snowmelt plots over time. So here’s that animation. And again, these are changing with, um, or the colors are representing different species of flowers. Okay. Annika, thank you. So not surprisingly, the number of visitors is tracking the number of flowers as we can kind of see there. It’s not super clear, but, um, but there does appear to be a connection. And again, the control appears to be peaking later than the advanced and, um, just to show all four sites across the landscape, I’m just going to bring in right now. Um, we can see that there’s quite a bit of dynamism going on, right? Like this, um, site on the top right side to over here is flowering way earlier than site four down here. Um, and there’s, uh, quite a bit of overlap and the species that are occurring among these sites as well. So it’s a really dynamic, exciting system. Um, and so to conclude, um, we asked how does early snowmen affect visitation rate? Um, we found that early snowball and its interaction with time appear to affect, um, pollinator visitation rate. We found that, uh, the subplot flower number also affected visitation rate. And we also asked how early snowmelt effects visitor diversity. Um, and we found that the day of year and the number of subplot flowers does affect visitor diversity and that local and landscape scale topography also affect, um, appear to affect, uh, visitor diversity. So, um, it’s important to note that, um, that this is preliminary data, but I think that we can say that the earliest snowmelt did have a significant effect, at least on the insect abundance, um, that were visiting those flowers. And it, we can also conclude that it may be important to consider a topography when we’re assessing plant phonology and, um, and how pollinators will respond to shifts in snow melt, training driven by climate change. Um, so as I move forward, I’m working on a structural equation model that will look at, um, how topography, the timing of, um, snow melt out in floral landscape, all interact to affect a pollinator diversity and abundance. I’m also going to be doing a phylogenetic analysis of pollinators Um, we’re also analyzing the broad pollinator community. I put out a bunch of Bibles to collect, um, abroad, pollinator community across the summer. And, um, I also performed a pollinator exclusion experiment to look at how important pollination is to seed set within the different treatments and across the summer. So I have a bunch of awesome undergrads who were helping me with that remotely. Um, so yeah, thank you everybody. Thanks to my

funding sources and everybody at CU Boulder, especially Kara Forester who had been so instrumental to this work, um, and to my committee at UC Riverside in particular, Nicole Rafferty and Marco Stasi of it. So thanks everyone. Thanks Monica for the great talk. And we have time for a couple of quick questions. That’s great Hanukkah Um, one question in that chat is, does the addition of black sand affect the microclimate or temperature of a site beyond speeding of snow melt? For example, I could imagine the black sand increasing the soil temperature and, or decreasing soil moisture after snowbelt, which might affect plant phonology. Yeah, that’s a really good question. Um, I, we have done some work looking at how the, the black sand effects, um, micro climate, and we didn’t find any significant effect. So this is going to be a six year experiment. So each, each year we like a thin layer of the black sand on the surface of the snow. Um, and so the actual accumulation of black sand isn’t that much actually on the surface of the soil Um, but we are cutting it off after six years of black sand applications. So we don’t just get like a huge layer of, uh, blocks and building up and really strongly affecting micro climate Um, in additionally, I’d like to note that the, the box in is a nurse, so we found that it doesn’t have any impact on soil microbes or, um, soil fertility in any way. Did that answer your question? Yeah, that’s cool. That’s really clever. That’s cool. Yeah. Um, I cannot take credit for that, but as a lot of work today, they haul a lot of pounds, like tons and tons and tons of black sand up there every year. Any other quick questions? Oh, one more. Um, amazing talk. I’m wondering, how did you collect insects? I recently learned about the insect vacuum, which was actually a children’s toy, but serve as a useful tool for research. That sounds like a really cool idea. I’ve heard of those being used also to like exclude certain insects, um, kind of like the, the opposite, but yeah, I just use, um, netting just nets and I, I got to where I could identify more for species without netting and collecting things, um, but typically just netting and putting them in vials and freezing them and bring them back down the mountain with them. Awesome. Thanks Monica Thanks Casey. Now moving on to our next speaker and when you met, who is going to be talking about macro evolutionary and global patterns of interest specific variability in her herbivory William, are you able to share your screen with us? Yes. Can you see it? Yep. It’s working and we didn’t hear you just fine. So over to you now. Okay. All right. Hi everyone Thanks for tuning in. I’m really excited to share the first set of results from a new global collaboration called the herbivory variability network. And I’m looking forward to your feedback on these really fresh results So I actually, I think I’m going to turn off my video to save my bandwidth here. I don’t have a lot. Okay. You’re good. Yeah. All right Thanks. Okay One of the most common observations about plant or before interactions is variability, herbivorous everywhere, but like many ecological processes, it varies tremendously in space and time across ecosystems, species genotypes, and even organs within plants. This biological variation has inspired generations of ecologists and evolutionary biologists most work on herbivory However, has focused on understanding differences in mean damaged levels, asking what leads a plant to get a certain level of herbivory This beautifully simple framework has taught us most of what we know about plant herbivore biology, but it overlooks a key aspect of her delivery. And that is the causes and consequences of the variability itself herbivory is indeed variable, but some systems are more or less variable than others. For example, herbivores among buildings there on the right, the traditional focused on mean conditions would explain why a given leaf has a certain level of herbivory, but it overlooks the orange, the origins and consequences of the variability itself. We can also frame this from a more statistical perspective where distributions of herbivory among plants within a species not only have means they also have variability and skew We can ask what causes a species to have low variability, high variability, or even a high skew in these distributions theory theory

tells us that variability and skew are important for population dynamics, the outcomes of species interactions, rates of evolution, super spreading and host parasite dynamics, and many other key processes, despite the theoretical importance of variability in species interactions. And perhaps because of our tendency to focus on differences in between means, would be lack of basic descriptive data on how variability varies amongst systems A fundamental understanding of variability patterns could help us build and test major new hypotheses about her bravery as well as transform longstanding ones. For example, several major hypothesis deal with how plant or before interactions vary with latitude We know me and her very often decreases with latitude, but does this mean that variability increases with latitude because lesser Beverly at high latitudes means that our, that herbivory can be more okay. Another major set of herbivory hypotheses posits that being large and abundant makes plants more Epic, a parent herbivores suggesting they should read, receive herbivory more consistently, but most tests have instead focused on how me and herbivory differs between a parent and unapparent plants overlooking the variability itself to collect the basic data. We need to start addressing these questions, several colleagues, and I recently formed the herbivory variability network. We’re a global network of biologists working to understand, um, how and why plant herbivore interactions vary across the tree of life and around the world. Our central focus is yeah, it’s just some stuff, error or variability in their interactions with herbivores more broadly though, our goal is to advance how we think about variability in ecology and evolution in general. We’re about 18 months old and we already have more than 200 collaborators from a hundred plus institutions in 30 plus countries In the first phase of our project, we’re focused on collecting the observational data. We need to lay the groundwork for a macro ecological and macro evolutionary understanding of variability We have three initial questions is high variability, a common feature of plant or, or interactions What types of plants exhibit higher or lower variability and herbivore damage and what bionic and a biotic factors correlate with variability and or before damage. We’re interested in this, a two scales among plants within populations and among organs or leaves within plants this stage in our network, or simply quantifying variability in herbivory using a standardized protocol. The essence of the protocol is picking a species insight, randomly selecting 60 plants and quantifying herbivory at three scales for each plant. These include estimating whole plant percent herbivory proportion of leaves, damage and percent herbivory for 10 leaks. The key feature of our protocol that sets it apart from past work is the really high sampling intensity within populations, which allows us to estimate variability skew and other district distributional features so far, the network has completed 716 surveys on six continents, including 474 plants species from 135 plant family. We concluded our first phase of data collection in early November, but amazingly over just the last two months, these two outstanding scientists, Mariah Robinson, and Luke there have made so much progress wrangling our enormous dataset that we have a few preliminary results to share with you today. So I’m gonna use those to address four questions. How variable is herbivory and how does variability in herbivory very geographically with plant traits and across scales. I’m going to summarize variability using two metrics. The first is the genie index, which represents the inequality or variability of a distribution. It ranges from zero perfectly, even to one perfectly uneven it normalizes by the total amount of herbivory allowing comparisons across systems. The second metric is the residuals of the mean variance relationship, our data like most, um, biological data shows strong relationship between the mean and the variance. So we took the residuals of this and use them as a metric of the amount of variance in herbivory subtracting out the effect of dummies. Okay. First, just how variable is herbivory as expected based on just about every naturalist, subjective impressions of

her livery. It’s highly variable. Moreover, the variability itself varies among species This histogram shows the number of species with different genie values. The meeting species had a genie of about 0.56, and for reference a genie of 0.5 means about 10% of plants got more than 50% of all or before damage, which is substantially uneven. Many species were far above or below this, which gives us lots of important variation to explain, um, with the networks, okay, utilize the field patterns from these Gini values. So here’s a panel of 12 surveys I randomly chose from the data set. Each one has a histogram that shows the number of plants in the population with different levels of her beverage. And you can see high variability and long, right tails are clearly the norm for plant or before interactions. And this is significant because theory tells us that this shape has key implications for ecology and evolution, influencing population dynamics, selection, and race of evolution. Okay. Question two. How to plant traits influence variability in herbivory. We have lots of plant trait data in the pipeline, but right now what we have ready is plant size. Fortunately, a plant size is one of its most important characteristics and size plays a major role in a pleasingly simple hypothesis about what determines variability in herbivory that is that variability in the distribution of herbivory among plants depends or is hypothesized to depend on the ratio of herbivore meal size to plant size. When plants are small, a fixed meal size, either completely wipes out a plant or misses it entirely resulting in a variable distribution when plants are large, they essentially average over the variation in herbivory to a greater extent, leading to a homogeneous distribution This hypothesis is almost embarrassingly simple, but it’s a central though, perhaps underappreciated component of Phoenix plant, apparently hypothesis one of the most influential ideas in plant herbivore, ecology and evolution as predicted amazingly as predicted by the simple hypothesis, larger plans experienced less variability and greater homogeneity in her bravery in this figure. Each point is one survey from our dataset. The X axis shows the mean plant size in each survey. And the Y axis shows the genie variability index And you can see the strong negative relationship, small plants get hammered or not. Well, large plants seem to average out the variation in her bedroom. This relationship is just as strong, even if we use the mean variance residuals to subtract out differences in me and her Every, this means that this phenomenon is not driven by how much her Berry plants are getting on average. It’s really just about how that herbivory is distributed among plants This result has major implications for our understanding of the evolution of plant defense Yeah, for example, that plant defense should be more common for small plants so they can turn the defenses on if an urban word shows up, but not turn them on. If, if there’s no herbivore there and constituent defenses should be more common for large plants, which should always experience moderate level of herbivorous Question three does variability April’s here But today I’m going to focus on latitude latitude on ingredients have taught us an amazing amount about ecology and evolution. Several key latitudinal hypothesis assume differences in and heterogeneity between the tropics and higher latitudes Most empirical studies. However, focus on means of biotic interaction intensities overlooking the variability itself. Before I show you the variability results, I want to show you that like other large-scale studies, we found a meaningful, but noisy, negative latitudinal gradient in the mean level of herbivory. So her memory is higher in the tropics, but there’s also a ton of variation. Okay. Okay. Now, when we put variability in herbivory on the y-axis with the Gini index, we see that plants experience more variability in herbivory at higher latitudes And here’s the real zinger. I think the gradient invariability appears to be stronger than the gradient. In mean, herbivory suggesting that geographic variation and variability may be central to these macro scale ecological patterns. This means that low latitude plants do not just experience higher mean levels of herbivory. They also experience more consistent herbivorous and high latitude plans do not just experience lower me and her delivery They also experience it more variably. This phenomenon, I think could be a potential explanation for why plant defense levels are often found

not to follow the latitudinal gradient in her bravery high latitude plants may experience lower herbivory on average, but the infrequent very high levels of damage. They occasionally do experience may be enough for them to maintain high levels of defense. Despite the fact that they have lower Habib average, when we use the residuals of the mean variance relationship to account for the latitude North radiant and mean herbivory the variability gradient is still present, but weaker suggesting that part, but not all of the latitudinal variability gradient is related to different. So Jessica studied the mean and the fence and costs are okay. Last question. Number four. How does variability vary scales? Essentially, we wanted to know if variability patterns across plants would be similar to or different from variability patterns among leaves within plants. This could tell us something about the scale of plant herbivore interactions that really matters So that’s what this figure shows here. You can see, we found a strong, positive correlation between the variance among plants and the variance within plants. In other words, some systems have high variance at both scales Other systems have low variance at both scales This suggests to me, at least that the factors that promote variability in herbivorous transcend these scales. It also suggests that among plant variability is probably not just a function of plant genetic diversity, because then we would expect within plant variability to be unrelated to among plant variability. So there’s something else going on. Okay, I’m gonna leave you three very preliminary take home messages First, our data indicate that variability is an essential feature of planter before interactions. It’s too significant to be left out of how we think about the ecology and evolution of plants and herbivores incorporating it into the fields, major hypotheses and developing novel hypotheses centered on variability patterns, I think has the potential to advance our understanding of plant or before and directions. Second, focusing on variability revealed that plant size, one of the most fundamental plant tree, it’s one of the most fundamental things about being a planet, just how big they are plays an outsize role in shaping interactions between plants and other organisms. It just boggles my mind that something so simple as plant size could be so important for shaping these distributions. And perhaps, maybe it’s a good reminder of how important something as simple as geometry is for setting the stage for species interactions. Third, the factors that influence variability seems to transcend scales within and among plants, which is just that some systems are simply more variable than others. For reasons we hope to discover as we continue diving into our dataset, adding additional analyses, phylogenetic analysis, special explicit analysis analysis of verb before density, pathogens, reproductive damage, um, plant traits, and, uh, many of the other variables that we have. So stay tuned. I want to thank the earth bar planning group, which includes Mariah Robinson, Lee Dyer, Phil Han Brian, in a way nor Underwood and Susan Whitehead. We’ve also recently only been joined by these three outstanding scientists who are helping us to take our bar to the next level and make it a really lasting collaborative network. I also want to thank the members of our seven topical subgroups and a big thank you to our coordinators. Um, Mariah and Luke, we’ve done an amazing job with it. Just a massive amount of organizing wrangling people, emailing, developing protocols, managing data, and more. Thank you for listening. Thanks well for this fantastic talk and we have plenty of time for questions Question is coming very quickly time. Um, will Sarah McPeak answer? It’s just, uh, this is a huge, exciting collaboration. Um, I’m wondering if your team is collecting data on herbivore population and treat dynamics as well as those are the plants. I think it would be interesting to think about factors that impact herbivore populations that I need to, it could be affecting the patterns of variability we see across these different gradients. Yeah. Yeah. I totally agree. Um, a subset of the collaborators have collected or before data, um, especially those who are more intimate, logically inclined. Um, and so, um, yeah, we’re really interested to see how those, those patterns match up. Um, that’d be cool. Uh, and then Joey Bernard asked, uh, says, awesome talk. Does the amount of variability in herbivory vary with spatial scale and does plant size also vary with latitude? Yeah, so, Oh yeah, it was a good question Um, we have, um, the data collected in a special explicit manner, but we haven’t dived into those analyses yet, but I am really excited to see how the variability varies with facial scale. Um, that would be really exciting

Um, so stay tuned. And so it is, it is known that plants in general are larger at lower latitudes, closer to the equator. Um, and I actually forgot to look if there’s a signal of that in our data set. Um, so, um, that would be an important thing to check. Um, let me think. We don’t, I think that that signal I think is stronger for, especially for trees. Um, and we haven’t finished cleaning a lot of the larger mature tree data that we have. So most of that data was sort of up to trees up to maybe two or three meters tall. So we have, um, some tree species in there, but nothing like this, like maximum sized trees. No, no canopy trees right now Cool. Thank you. Any other questions? We’ve got a couple minutes. Okay. All right. I’m sure there’ll be more questions for will later on. And it’s now time to dive into the dissection and of this symposium session and with Adam stickers, stupids talk on the genomics of malarian mimicry. Um, Adam, are you able to share your screen with us? Yes, we can hear you well. So I’m going to mute myself and it’s over to you now. All right. Great. Thank you everyone for virtually attending my talk My name is Adam Sucre, and I’ll be talking about some work that I’ve done both as a PhD student in Kyle summer’s lab at East Carolina university, as well as a postdoc and Matthew McMannis his lab at university of New Hampshire where I’m currently stationed, but still virtually So if you travel throughout, um, throughout nature and you look, there are a huge variety of signals and modalities that signaling is do is done in nature. So on the top, we have two frogs that have visual signals in their skin. Um, on the bottom left, we have watermarks that are using tactiles signaling to garner meetings. We’re all familiar with how birds use auditory signals as a way to maintain territories and also acquire mates. And then also there are signals that occur in modalities that we probably don’t think very often of things like olfactory signaling. And I’m primarily interested in visual signaling and both of these frogs are using their integument or their skin to, as a kind of classic anti predator mechanism. The frog on my left is using it to blend into the background. Crypsis the ID being that predators can eat them. If a predator can’t find prey items, the frog on the right has pretty much the opposite end of the spectrum. As far as an anti predator mechanism goes to have these bright colors and patterns in their skin really contrasting with the background habitat. And it’s coupled with the presence of these, uh, defensive toxins in their skin. And this combination is known as ABO Semitism and AICPA sematic species are those that are both conspicuous and defended. And the really important thing about AICPA Semitism is that predators have to avoid, have to learn to avoid a particular phenotype. So they have to associate that appearance with a bad outcome for them. And because there’s this component of learning where predators are, each prejudice has to potentially learn to avoid that phenotype We’ve long thought that predators should select for mama morphism. So they should drive these populations and potentially species towards a single phenotype to kind of ease the burden of learning across species and with an especially within populations. But that’s not really what we see. If we look in nature, we see a lot of variation across species. For example, in this a phylogenetic tree of poison frog family, you can see there’s a huge diversity of color and pattern. We see a lot of variation that’s occurring across populations. For example, these two Ladybird beetle populations, they have discreet color morphs. So we have an example of polymorphism and we also see variation that occurs within populations that can be discreet, where you have kind of a con or not discreet, where you have a continuous

gradient, um, or it can be discreet where you have distinct color morphs or polymorphisms within a population. And so the important thing here is that HIPAA summit, Avis medic colors and patterns are really astonishingly variable, despite what we’ve kind of historically predicted. So I’m interested in recently, how is this variation produced? And I’ve been using this little poison frog, this endemic to Peru, its name is rented a ma imitator and that specific epithet imitator, um, refers to the fact that it’s a mimic. So as you travel throughout Peru, there are all of these allopatric mimicry complexes involving imitator. Imitator is the frog on the left and each of these pairs and all of these are kind of non-overlapping mimicry complexes where imitator has evolved to, to mimic these established non-generic species. And so imitator historically variegated into Peru, and then really cool thing here is that adaptive radiations are powerful tools for exploring evolutionary patterns. And so we set about, um, assembling an imitator genome. We assembled it with PAC bio data, polished it with Illumina data, scaffolded it with 10 X data and Nanopore data. And then eventually annotated the genome with maker. This slide has exactly 22 words on it. And I know because I counted, but it represents literally years of agony of my life. Why is it agony? Because the imitator gene has Gino is actually fairly large. It turns out it’s about 6.8 gigabase pairs in size. Um, our final assembly had a cons again, 50 kind of a measure of genome contiguity of 300,000 base pairs, which isn’t phenomenal, but we’re relatively pleased with given the size of the genome. And we also have really good gene and content. So we have 93% of the expected tetrapod orthologs as determined by a boost go analysis, which means that we’ve got really, really good gene and content, especially for a gene on this size. This is pretty analogous to a decent mammalian genome Okay. Now, why are we relatively happy with that? Somewhat mediocre contiguity it’s because poison frog genomes are actually replete and these repeat elements, they’re just everywhere in the genome. So an imitator, we have a 6.8 gigabase bear genome. Um, the repeats are mostly lines and unclassified elements, things that we’re still kind of working on on further classifying and gypsy and Marriner are present, but not overly abundant. Now, why am I highlighting these to gypsy and Mariner? Because in the other poison frog that’s been assembled and published with, what they did is they have a similar size genome, but these genes, this genome is really raw. Um, really just chock full of these two elements, marinara and gypsy almost 1.4 gigabase pairs of it, or like half of a human genome almost are these two repeat elements. So the cool thing here is that we have approximately the same size genome. We think between these two species, there are some pretty big differences in the, in the quality of these assemblies as they are right now, but we think that they have evolved very differently, um, which indicates that there’s, there’s very little ability for these frogs to stop repeat elements or, um, from proliferating, but they have a different evolutionary history, which is really cool to delve into in the near future. So now that we’ve just demo this genome, we then decided to leverage it, to look at, um, color and pattern projection, and we have these four mimetic morphs of imitator And so we really wanted to start there. And the cool thing here is that imitator, which is in this dashed box here and mimics these three common, generic species, two more so very obvious on the left Um, some Rezai on the top, right? And fantastic on the bottom, right? They have a different evolutionary history. So if you look at the phylogeny of written Mayo imitator here has these really short branch lengths with, which is indicative of that adaptive radiation that I talked about. Those species that imitator mimics have much longer branch links. So it indicates that those species were established prior to imitator radiating out. And so we think that there are different evolutionary histories, their imitator diverged more recently, and add verged onto established phenotypes that those models species had, which means that these two, these groups of species have different evolutionary histories. And that means that we can examine the production of these phenotypic variations from multiple

perspectives. So how do they evolve to have that same phenotype despite different evolutionary histories? Are they the same mechanisms or are they different? So it’s a nice powerful system for that. Now to start with, we took a gene expression approach to examine color and pattern phenotypes within the system, primarily because with a 6.8 gigabase bear genome, it’s really fundamentally impossible right now, um, fiscally as well as computationally storage wise to, to do these kinds of whole genome approaches. So we, we took a reduced representation approach and we examined these former medic Morrison imitator, uh, two more, very, almost the imitator mimics and then two fantastic and Marissa have convergent phenotypes with imitator. I’m sorry. They have the same phenotype as imitator imitator at verged. I used the wrong language. And what we did is we looked at gene expression in the skin from Polskin from really early on and tadpole life, when they’re pretty much brand new into the water all the way through early froglets dumb, or just after they met a morpho sounds to the water. And so, and during that time point, time period, they go from, um, not having any sense of pattern to all the way through, uh, assemblance of their adult color and pattern. So they’re putting down all those Chrometa fours and that pigmentation that they would have as an adult that then continues developing after they become a juvenile. And if we look at the gene expression data, what we see what neatly falls out is that imitator falls out as a Columbus side from very obvious, uh, which is on the left and Fantastica on the bottom, right? Um, so imitator is its own kind of separate grouping on this principle component analysis and this kind of recapitulates phylogeny very obvious and fantastic are much more closely related than imitator. And we also see, so that’s a small proportion of the variance, 13% principle kimono too, but the large part of it is that what we see is developmentally, there’s a lot of change in gene expression from young tadpoles all the way through young froglets. And so that is, that is the primary driver in our gene expression data. And when we were looking at elements of color and, or in particularly teaching of some idea, do things in, in, in other organisms, which is always a really good starting point when you’re in doing this kind of work, we saw that there are quite a few interesting genes that fell out a lot of them. Um, I’m going to highlight some of the cool ones There are quite a few that were in the melanin pathway. Things that, um, are involved in either producing Mylanta fours are involved in melanin synthesis and two really common ones from other literature, particularly mammalian literature is Milena Orton receptor one, and tyrosinase related protein, one, both of which play a role in melanin synthesis. And so we see that these genes are playing likely a pretty, pretty key role here in actual, um, pigmentation and color pattern elements. And so this isn’t particularly surprising given that these are our main players and a lot of other taxa, but it’s interesting that we’re seeing them here. Again, we also see a lot of evidence that targeting genes, which are eventually produced, um, yellow and orange pigmentation that gets deposited into chromatophores We see that these genes are playing a pretty important role, especially between color morphs within species. And so these three genes are especially interesting because some other work that we’ve done in other species and, um, one or two other published studies in poison frogs seem to continuously highlight genes in this pathway as being really important. These genes can consistently fall out as differentially expressed between color morphs or patches of skins. So we think that these are really, really important for producing actual pigment elements that produce different colors in these species. Um, and this is true also in a number of other taxes as well. So a lot of vertebrate lineages there’s, um, a number of Teradyne genes that are involved in producing coloration and to kind of corroborate this, um, some work by Andrew Rubio and the rest of us. This is currently in review. Um, he found doing some network analysis of different skin color patches that a lot of these genes that I just highlighted, followed as hub genes So those are genes that are particularly important in gene expression patterns, um, and are likely driving other elements of gene expression And so we see two of those, um, targeting genes that I highlighted PTs and SDH as well as one of those melanin genes MC one are the

Malana cord receptor. So this is nice independent evidence that we’re we’re on the right track here. And one really interesting thing that we haven’t delved too deeply into as far as doing any functional validation or anything like that is the possible role of keratin genes in producing different colors and patterns in poison frogs and potentially vertebrate as a whole, um, Evan to me, and a bunch of coauthors in a recent AMAT paper found that the actual platelet thickness of the, the guanine platelet layers within the Rita force, which are the core metaphors that are responsible for blue and white coloration, largely at least we thought, um, that the platelet thickness within those iridophores is actually a primary determinant of the color of frogs, kind of, regardless of what, um, what color these frogs actually are. Okay And so, because of this, we also see a suite of genes that are in the keratin family that are differentially expressed within our dataset, um, amongst these are care T1 and Cara T2 So just, um, standard keratin and family genes, and these genes have been seen as, um, important for pigmentation and different color production and a variety of other burden vertebrates It’s not super common that people see those, but we also see a number of other keratin family genes that are differentially expressed between color morphs and species like, or T 10 18. Um, and so we think that the keratin genes, because they’re being, they’re actually producing keratin in the epidermis differential expression, and that can actually change either the thickness of these into four layers or maybe their orientation. And so this is an interesting Avenue for kind of future work and functional validation that hasn’t really been delved into very deeply by anyone, but we think that this is going to be a really cool Avenue of inquiry. And so why did these frogs have such variable colors and patterns? Um, our gene expression data strongly hints at certain color genes being important, particularly, uh, melanin synthesis genes, as well as Teradyne genes, but we need some functional validation. Um, some of which is ongoing and we’re hoping to do a lot more soon pending funding. Um, obviously this is kind of the, the first pass at this and a loo like variation is definitely going to be, especially between these different populations and these different species. And so we have a number of projects I’m going to kind of look at that in a lot more plan And with that I’ll acknowledge, um, my current lab and, uh, funding agencies. And I thank you for coming to my talk virtually, and I’ll take any questions you may have. Thanks so much Adam for that talk. And we definitely have time for questions over to Casey. You can see what comes in here. Okay. Sneak one thing in totally loved the pictures on the frogs And they’re so cute. Thank you. That’s really why I’m here. There was a question. Um, great talk regarding the PE percent of the genome between species. Do you think the common ancestor lack to ease or as one replaced the other? This is a good question. And I don’t have the answer to that. So almost certainly there are a lot of transposable elements, um, just because frogs in general have a lot of them, probably at least that’s what the, a lot of the data seems to indicate, but, um, we don’t know really, which came first, the transposable elements or the poison frogs That makes sense. Uh, another question from Eva Fisher is yea working on beautiful frogs Um, my question is if there’s a species specific signature of skin, gene expression, but amazing color convergence, how do we begin to understand which gene expression patterns are the relevant ones? Do you think this means there are different solutions for generating the same color pattern? That’s a really good question. And that’s, that’s what we’re hoping to get at. I mean, obviously there’s there going to be species

specific patterns of gene expression based off of, you know, different evolutionary history race we think. Um, but they also have a relatively recent common ancestor and, you know, they’re all con generic. And so we think, we think that it’s likely that they’re shy. They’re just kind of utilizing the same mechanisms, um, across species, but we don’t, we don’t know that as of yet. Um, I’m sure that there’s a lot of cases where there are certain genes that are playing a role in pigmentation that are producing maybe the actual pigments, like terracing genes. And then in other cases it might be, you know, uh, I know they like variant that is producing different pattern elements in a similar manner. any other questions for anyone out there? okay Well I guess not. So I’ll stop sharing and give whoever is next, the chance to get ready Thank you so chance to look at that frog picture Yeah. I was happy to look at the frog as well, but thanks. Thanks for that, Adam. Great talk Great photos. Thanks, Casey. Um, we can probably start moving on to our next speaker, Bob wheat, and they are going to talk about the co-evolutionary arms races and the conditions for the maintenance of mutualism. Bob. You’re able to share your screen. That looks right. And if you can say something so we can check your sound, right? Oh, hello? Yep. We can hear you. Okay. Great All right. Thanks everybody for joining me Um, this has been a really great meeting so far, lots of really awesome presentations and discussions. Um, so today I’ll be talking about the conditions for the maintenance of mutualistic interactions and the face of co-evolutionary arms races. So, uh, typically, uh, in the past, when I’ve thought about, uh, co-evolutionary arms races, my mind typically went to these, uh, classic antagonistic examples, um, which seems that like most coalition arms races would be antagonistic. And so here’s a classic example, uh, where this garter snake is predating on the new and the new has evolved, uh, toxicity Um, and then the snake has kind of evolved as resistance to toxicity. And so there’s this exculpatory dynamic for a greater toxicity and resistance to that. Um, another classic example is between this, uh, Camilia weevil and the chameleon fruit, uh, the chameleon, we will, uh, bores into the, the pericarp of this Camelia fruit to have a posit, um, it’s larvae. And so there’s this co-evolutionary dynamic for greater mouth parts of the community, a weevil and thicker Perry carps of the chameleon fruit. However, um, these collusionary arms races are not restricted to antagonisms. We see them also, um, particularly and pollination based. Mutualisms such as that between, uh, this MOGUS drink is laundry justice fly and lap Rosa and SEPs, um, flower down in South Africa. And so we see this fly has this ginormous mouth part that it uses to insert into this narrow, but deep, uh, floral nectar tube to retrieve a tiny droplet of nectar at the base of that tube. And by being forced to insert a small part further into the flower, it, um, rubs against the reproductive organs of the flower, thereby, uh, transferring, um, Paul. And, and so the idea here is that, um, the flower is inducing selection on the fly for greater mouthparts that the fly has an easier time obtaining nectar at the base of that too. But at the same time, the fly is inducing selection on the flower for deeper nectar tubes. So that way, uh, the probability of pollination is increased. And so my question became whether or not, um, uh, the mutualism can break down, where are the conditions under which this mutualism can break down and the face of this coalition in your arms race, if the mouth part of that fly becomes, um, so exaggerated

beyond the floral tube of the flower and might not transfer pollen in might be better described as a nectar, see friends than a pollinator and vice versa. If that fly is hardwired to visit these flowers and the flowers evolve, uh, for all tube depth, that’s longer than the mouth part of these flies. Um, the flies might not ever be able to retrieve that nectar, but for us to pollinate those flowers, every visit. Um, and furthermore, I’m curious about, uh, whether, whether this outcome depends on the phenotypic interface. So there’s these two classic phenotypic interfaces trait differences, which is a bigger, is better type situation, which you might think that having a greater mouth part, always confers increased fitness, or having a deeper floral, an extra tube, always confers increased fitness. Um, but, uh, there’s potential that there’s actually some costs associated with this. That’s completely due to the interaction. We can see that this fly is actually having a little bit trouble, uh, trying to insert its mouth part into some of these flowers. And so there could be, um, some selection, some sort of stabilizing selection to keep the, the, the mouth part from, from not evolving too far past the flowers. Um, and so I took a completely theoretical approach to understand, um, uh, the consequences of these two, uh, phenotypic mechanisms. Um, and so the trait differences model can be summarized by, uh, this figure here. Uh, so we see the fly fitness w sub X on the left. Um, it increases as the mouth part length on the X axis, um, increases past the floral tube depth on the y-axis. And so the brighter colors here correspond to increased fitness and vice versa for the flower. Um, another potential mechanism to that can mediate. Um, the, this interaction is, is offset matching model. And in this case, it’s, instead of bigger is always better Uh, fitness is actually maximize at some finite offset from the partner trait value. And so we see, uh, for the flight fitness on the left, that the mouth part, uh, fitness is maximized when that mouse part length is this lower craze, Greek letter Delta greater than the floral tube depth of the flower. And vice-versa the floral tube depth of the flower confers maximum fitness for the flower when it is greater than the mouth part of the fly by this Greek letter Delta. Um, and so to understand, uh, whether or not this is a mutualism or an antagonism, we can summarize the interaction by the effects on growth rates. And so for the models, I’ll be considering, um, these growth rates and bar represents a population level growth rate, which is that average across all a trait values and that species, uh, decomposes additively into this intrinsic gross rate for that species and this capital, I, which represents the effect of the interaction on the growth rate of that species. And we can do that same decomposition for the partner species Y uh, and so if those capital eyes are in the top, right, then we see that the interaction can be considered a mutualism, but if either of them become negative, such as the top left or the bottom, right, then this mutualism has collapsed into basically a parasitism. Um, and so the question comes down to, uh, how do these interaction effects evolve? Um, so it turns out that under this model, these interaction effects are functions of mean trait values. And so really we can follow the question down to tracking the evolution of these mean trade values and the evolution of these cheat mean trait values classically, uh, can be quantified as this, uh, co-variance of growth rate and tr uh, and trait value And so what we need to do is we need to calculate, uh, these growth rates from those fitness functions. I showed you on those previous slides. Um, and so to do this, we could just take this expected lifetime reproductive output, which I had presented on those previous slides Um, we want to calculate these instantaneous rate of growth that are functions of trait values, and we can use, we can do that by taking this little rescale limit here. Um, and so, uh, before I dive into the models, I just want to provide a little overview of the important parameters involved with these models. So the capital B X and capital B Y, those are the strengths of Biotics selection, the selection induced by this interaction. Um, and those are going to appear in both the trait differences and the offset matching model, uh, that lowercase Greek letter Delta is the optimal offset, which occurs in this offset matching model and, uh, GX N G Y to note the amount of heritable variation and phenotypes and each of these species. Um, and there, you can also just think of them as the additive genetic variances And this will also appear in both the trait differences and outset matching models. Um,

so under the trait differences model, we get the dynamics accomplished. Narrator dynamics are, are very simple. In fact, they’re constant We see that the change, uh, in X-bar over time is just a product of the additive genetic variance of species X and the strength of Biotics selection on species X and similarly for species Y. So these trait values are just going to continue along at constant rates independently of each other. Um, even though the evolution is induced by this interaction Um, and, and so that as, uh, as the product of the outer genetic variance and strength, the Biotics selection differs between these species, the rates at which they evolve will also differ in contrast under the offset matching model. We see this additional green term appearing on the right here. And so in addition to the product of heritable variation and biotic selection, we see this offset matching term appearing in the green Um, and so for, uh, the evolution of mean trait and species X, if that, uh, X bar increases past a Y bar plus that optimal offset, then the rate of change blocks will become negative And so that creates a sort of a stabilizing mechanism where, uh, the evolution of species acts would actually slow down if we were to halt the evolution of species Y um, however, this still leads to indefinite escalate, Tori, uh, trait dynamics. Um, and so pictured here are the trajectories of mean traits for both the offset matching and the trade differences models. Uh, the trade differences models is pictured here with dash lines. And so we can see, of course, that they’re just trudging along at these constant rates and this capital Delta I have on the right, uh, pictures, the difference in mean traits between the two species, when you can see that that difference between the main traits of the two species continues to increase indefinitely as well In contrast, under the offset matching model, which is a solid curves in the middle there, um, the there’s still a indefinite trait, escalation, and both of the species. However, the difference between main traits of those species converges to finite value, which is determined by the optimal offset. Um, and so to get back to the question of, uh, what are the interaction effects and therefore, w when do, when does this mutualism break down, um, so we can count, we can calculate what those interaction effects decompose as under these models. Um, so the interaction effects for species X, uh, decomposes additively, and to the intrinsic benefit of the interaction, which is just a benefit, how much this species benefits from the interaction when the strength of Biotics selection is equal to zero. So when there’s no Biotics selection induced by the interaction, and then on the right, we have the product of the strength of Biotics selection and the difference, um, between me and traits and this hold similarly for species Y. And so based off of the previous slide, we can see that because the mean traits are going to continue to continue to the difference between the main traits is going, or is going to continue to escalate over time. Um, the eventually, uh, the, the interaction effects for one of these species is going to become negative. And so therefore the mutualism, uh, will eventually dissolve and to, uh, parasitism under this, uh, trait differences model. So on the left, uh, on the left plot here, uh, see the evolution of the interaction effects, uh, over time. And we see sure enough, the species Y and the dash line, uh, eventually the interaction in fact becomes negative And for species X, it continues to increase over time. And we can also calculate the rate at which just transition occurs, uh, which we have in the rate. And that’s pictured as the difference between the products of heritable variation and strengths of Biotics selection for the two species. And, uh, this figure would look symmetric if, uh, those products were the same for both species. Um, so in contrast, under the offset matching model, uh, we get a different term for the interaction effects, um, slightly more complicated. Um, so in the purple we still have the strength of biotic selection. Um, now in green, uh, reappears, this sort of optimal offset term, uh, squared in this, uh, uh, equation. Uh, but we also see the express variation. So this is the variation that you had go out and measure just by measuring the express phenotypes disease species, um, appearing and the red there, and, uh, this decomposition holds for both species. And so we could see that, uh, as the, the, the interest specific variation increases for either species, um, this mutualism is at risk of collapsing into a parasitism, uh, or, uh, AF the Optima, or if the difference between the main traits continues to increase. And this, um, mutualism

can also collapse into parasitism. However, we saw on a couple slides ago that the difference between me and traits under the offset matching model actually converges. Um, and so because it converges to a finite number that means for certain levels of biotic selection and certain levels of express variation, this mutualism can be, uh, preserved under the offset matching model. And so here’s two examples of that on the left is a interaction that was initiated as a mutualism. Um, so we have the interaction effects on the y-axis. And so we can see that the interaction effects are both positive in this situation. So it was initiated as a mutualism, and it was maintained as a mutualism, even though the main traits, meanwhile, are continuing to escalate indefinitely, um, on the right, we see a situation, uh, of a novel parasitism actually being transformed into a mutualism due to this offset matching mechanism. And so, uh, this is just a very simple little theoretical exploration. There’s some important caveats to that. Um, I didn’t include the effects of antibiotic stabilizing selection and, and that would prevent me and traits from, uh, escalating indefinitely Um, I ignored any effects of equal evolutionary feedback such as with the, uh, abundance dynamics of these species. And that could also modulate the strength of selection experienced by these species and also ignored, uh, any multi specific interactions that his interactions with three or more species. And, uh, this can also provide some conflicting sources of selection complicating the results. So then how do we interpret the results and light of all these restrictions? Uh, well, I would argue a similar interpretations of how we look at lab and field experiments that are controlling for external variables by controlling for, uh, a box stabilizing selection and multi specific interactions and et cetera. Uh, we can focus on the effects of the phenotypic interface, whether or not it’s the trait differences or the offset matching on the outcomes of these mutualistic co-evolutionary arms races And, uh, before I finish, I just want to, uh, discuss real quick some implications, uh, for mutualistic networks. And this actually ties back, uh, pretty well to a talk by, uh, Lucas Camacho. I hope I got that pronunciation, right. Um, that who opened up the session talking about, um, the role of exploitation and mutualistic networks. Um, and so, uh, going back to this, this system in South Africa, uh, this fly in the flower, um, the flight actually visits a Guild of about, uh, 20 other, uh, flowers. And in fact, it has been referred to as a Keystone pollinator of that region Um, uh, and it also visits flowers that seem, uh, not very well adapted to visits such as this Bobby on a thunbergii pictured here Uh, it seems that this fly is just totally robbing the nectar out of this flower without transferring, um, any pollen. And, um, then the question becomes if there’s sufficient variation in populations of, uh, Bobby on a thunbergii, uh, then there’s potential for, uh, for adaptation, for evolutionary adaptation in response to interactions with this fly And as, um, Bobby, Anna thunbergii adapts in response to interactions with this fly, this can have indirect co-evolutionary effects on, uh, other, uh, members of this pollination networks, such as these Melkite sunbirds, which are the primary pollinators for Bobbi Jonathan Bregy. So, uh, the question is whether or not this little fly can actually lead to the indirect evolution of elongated, uh, beak links and those Malik Heights on birds, which I think is very interesting and maybe sounds a little bit farfetched, but, uh, but perhaps not according to some recent work done, uh, by Paula Guillermo’s at all, um, which, uh, demonstrates that actually indirect effects can drive a very large amount of revolutionary change in mutualistic networks. Uh, so with that, I would like to thank American society of naturalists for hosting. This has been really wonderful meeting, um, Scott newsman, who, um, coauthored this work with me and was also my PhD advisor, um, the Institute for bioinformatics and evolutionary studies at the university of Idaho for providing a lot of support from you throughout my PhD and the NSF as well for providing, uh, support And with that, I’ll take any questions. Thanks, Bob. We definitely have time for questions and I see they’re already rolling. So Casey, over to you. Thanks, Bob. That was great Um, yeah, so Lucas Camacho is asking, um, great talk, very impressive work. We should talk more for sure. Um, did you have any expectation when you, when you insert your model in a three or more species system? And so when there’s more of a network of interactions, um, it w uh, I, I have some vague expectations. Um, I think it, my a, an interesting follow-up project

to this that I was thinking about is actually very similar to Lucas’s work. Uh, but instead of using this trait mismatching model using a, uh, the offset matching or a trait differences model and see how these, um, phenotypic interfaces can, uh, the, the frequency of different phenotypic interfaces, mediating, these interactions changes to structured mutualistic interactions Um, I think that, uh, for these escalatory type dynamics, we might see, uh, particularly for the trait differences, we would see probably increased nest illness, uh, because in the, in the trait differences, which is a bigger, is better type mechanism, the species with greater, uh, trait values are, can always, um, interact with species of lesser trait values. That makes sense. That’s cool. Bob, as you were describing that model, it occurred to me that it seems like there might be different fitness consequences for the plant in the insect. And you kind of touched on this at the end, too, that the fly can go to other flowers if it doesn’t match the Iris, but the Iris, I assume, can only be pollinated by that fly. Um, and so is that built into the model already? Like, does that affect the strength of selection on, on the plant in the fly? No, I think that would be included under multispecialty interactions, which I indefinitely, uh, controlled for in this. I basically forced these two, if they couldn’t visit each other, then they’re doomed to extinction. Cool. All right. Let’s thank Bob again for their top. And we can move on to, I would start getting next set up for his talk about the survival of brood parasitic young and how it’s contingent on host siblings Sounds very interesting. Um, Nate, are you able to share your screen with us? Um, let me see. Yep. Yeah, let me get it pulled up here. Nope. In your audiences. Fine. So moving on to you then. Sounds good. Um, thanks everybody for sticking around for my talk. Uh, I’m super excited to talk to you today about one of my dissertation chapters, um, looking at support for a Goldilocks principle of post optimality for brute parasites. And so, um, just to get started that I wanted to briefly talk about, um, just introducing people to brood parasitism, if, um, it’s new to you. So group Harrison was just a symbiotic relationship between parasites, um, and hosts, um, parasites lay their eggs in the nest of hosts and hosts care for those offspring. Um, and this can occur either within species, which is known as conspecific brood parasitism, or between species, which is hetero specific brute parasitism Um, this occurs in three main groups, it occurs, um, a species of catfish and it was the cuckoo catfish paper wasps, and then probably most, um, most well known is in birds, uh, which is the focus of my talk And so looking at, um, avian obligate, brood parasites, those that can, that don’t build their own nests and can only parasitize other species. Um, we see that, um, there are roughly 101 species, which has roughly 1% of all birds are obligate avian brood parasites. Um, there are at least, um, there are at least seven independent origins across five different orders. And these species range from specialists, the parasite only one or very few species all the way up to generalists, such as, um, the Brown, the cowbird, which is the focus of today’s talk that compares ties, uh, somewhere around 300 different species. And, um, brood parasites have far ranging effects on a number of bird species as they parasitize roughly 17% of all bird species. Uh, and they often reduce host species fitness, uh, because, uh, post parents end up taking care of the brood parasite either along with, or, uh, in spite of their own, um, offspring. And so, uh, for a brute Parris that a golf spring that doesn’t share any genetic relatedness to either its host parents or, um, the other nestlings in the nest after it hatches. Um, it really has two options for success that can either kill its nest mates, or it can share the nest with those nest mates. Um, and so within killing there’s a couple of different ways that brood parasites do this. They can either, uh, destroy, they can either, um, directly, uh, attack the host nestlings or host eggs Um, this has seen in honey guides, which use

a hook on the end of their bill to, um, tear into other nestlings or the eggs or through eviction of eggs and nestlings, which is commonly seen in Cuckoo’s. Um, Howard does on the other hand, our net sharing brood parasites, um, and so either some or all of the, um, the host nestlings of the, the host species survive in those nests. And so the question becomes is when is it beneficial to kill your nest mates versus to share the nest with your nest mates? Um, and what we see with this is that it’s really a provisioning trade-off. Um, and so brew parasites that are able to stimulate host parents on their own, um, tend to be more of your, um, your natural killers versus brood parasites that have, uh, more of a low stimulative ability, but, uh, tend to be more competitive, tend to share the nest with nest mates, um, as a way of soliciting provisions from the host parents than just out competing the host parents when they bring the food to the nest. And so we look at common, cuckoos common, cuckoos are highly speculative, and they’re begging, um, is actually considered a supernormal stimulus, uh, because it stimulates parents to provision at a rate of three grid, Reed, warbler, nestlings Um, but we see that there’s a trade-off with this in that, um, those when cuckoos are forced to be raised in this with, um, with hosts Nestle and great Reed warblers, uh, their growth rates actually drop, uh, and at least early in development. Um, and it’s four days in the checkbox on the graph on the right, they receive much less, um, provisions than what would be expected by random chance. And so this, this pattern is pretty well established for Cuckoo’s when we switch over to Hubbard’s On the other hand, what we see is that cupboard growth actually seems to be dependent on, um, plus brood size in a different way. So, whereas, um, cuckoos actually grew worse with, um, with host nestlings in the nest with them We see that too, at least in at least some point, um, Hubbard growth is, is dependent on, uh, host nestlings, also being in the nest. Um, but beyond that point, it doesn’t seem to be anymore. And so what it appears is happening is that there’s, there’s a bit of a Goldilocks principle with, um, host species And I should point out that each black.in the graph on the left represents a different species. So this is, uh, across species. Um, this is the pattern that’s been shown in, in the prior literature. And so, um, we see kind of this Goldilocks principle where, um, if the nest has too many nestlings or too few nestlings, um, it’s not as good for growth rates of, of the brood parasite as, um, kind of an intermediate amount of competition in the number of nestlings that the, the brute Parris that has to compete with. And so this also what the mechanistic, um, or what the mechanism behind this seems to be is that it translates to provisioning So caliber is in a sense seem to be using host nestlings, to bring in more food. And then once it comes to the nest with the graph on the right, what we can see is that coverage with two, uh, the arrays with two Phoebe chicks tend to receive a greater amount of the provisions that come to the nest, then singled coverage alone. And so it seems to be that there’s a provisioning mechanism, the covers themselves are not able to stimulate host parents to bring in the food without, um, host nestlings, but once the food is there, they tend to be able to out-compete the host nestlings for that food. And so getting to my study system, we wanted to see if this pattern of, um, this Goldilocks pattern of growth translates to survival, um, for brood parasites. And so within the nest boxes of planetary warblers in the Shawnee national forest in Southern Illinois, we were able to, um, manipulate brood sizes and, um, essentially weed out any effects of, um, predation or ecto parasitism And so we put Calvert guards in the nest boxes to avoid predation by other cupboards, um, that may be coming to essentially get the warblers to reset, um, by killing the brood so that they can turn around. And parasitize, um,

the brood, um, we use Raptor, we use wire guards around the nest boxes to, um, discourage Raptors from coming in and eating the nestlings And then all of our nest boxes are on greased, aluminum conduit, Pauls. So we prevented ground predation, um, from squirrels and raccoons and snakes. Uh, and then we changed nesting material on the fourth day after hatching to prevent, um, nest failure due to blowflies blow, fly ecto parasitism. And that seemed to be effective. I didn’t find blowflies after, after doing that. And so what I ended up doing is I essentially, um, at hatching, I modified, I manipulated the brood sizes, um, to foster worldliness, linked out to unpair, as enticing, as boxes in the system to produce, um, treatments of either zero host, nestlings to host nestlings or four hosts nestlings each with a Calvert in the nest box. Uh, and then I monitored these nest boxes all the way through development, which is for a Calbert. It takes about 10 days, um, to fledge after hatching. And so what we find using a quadratic regression is that coverage survival seems to be highly dependent on brood size, just as, um, patterns of growth rate where, and so we see that, um, survival for, um, seems to be best around, um, to, to host nestlings growing up in the nest with, um, with cowbirds. And so we see the survival, there is somewhere around 80%, the actual number of kind of ideal brute size is, um, 2.3 hosts nestlings. And then we see that on either side of that too, a huge drop-offs in the probability of survival with, um, four hosts nest things being much better than two hosts nestlings, um, but nowhere near as good as, um, as two host nestlings. And so we got really curious about this to, uh, kind of what else, what was driving this, this pattern of, of survival. And so we looked at it a little bit more higher resolution of this data, and what we were able to show is that it looks like the costs of, of suboptimal brewed sizes are paid very early in development for the cupboards. And so if we look in, in pink is the, um, the host, the zero host nestling treatment in green is the two hosts nestling treatment. And in blue is the four host nestling treatment. We can see that on day two and day three. Um, there are a huge drop-offs in the survival of, um, the zero host nestling treatments. Those cupboards did not do very well, but if they survived, um, today three, they were able to make it most of the rest of the way out. And I, I should point out that I cut off, um, the number of days post hash for this graph at eight. Um, just because cupboards can fledge on day nine and day 10 day nine or day 10 And so it wasn’t able to resolve whether or not, um, they had, they had fledged or died because the, uh, the adult warblers, um, sanitize the nest. And so they’ll pull the dead nestlings out of the nest. Um, so I just, I cut it off a day because I, they don’t seem to fledge by day eight. Um, and then the other thing we can see is that, um, the two in the forest, they, um, the two hosts nestling treatment in the four host nestling treatment stayed pretty competitive, um, for the first three days. And then after that, the four host nestling treatment drops, um, below the two hosts nestling treatment. And so we see that those costs tend to be paid pretty early in development, um, looking at a potential provisioning mechanism Um, we see that we do see that in the two hosts nestling treatment cupboards are fed much better than they are in the four or zero hosting close nestling treatment. Um, with four hosts nestlings, the thought is, is that competition obviously increases from two hosts, nestlings and SOCOM birds are just getting fewer feeds because there are more mouths to feed in that nest. And then in the zero treatment, um, for the provisioning rate, the thought is just that the cupboards don’t have the stimulated stimulative ability to call in provisions from the parents as they bag, um, looking at the proportion of provisions to it’s both in the two and four hosts nestling treatment as the zero host nestling treatment would be 100% of provisions coming to the nest would be of the caliber. Um, we see that both are above what would be expected by random

chance, but then in the two hosts nestling treatment, um, it is still higher and there is a significant difference between those two groups. And so this got us thinking, um, about the potential effect of, um, what cupboards might be doing to, uh, avoid being in a sub-optimal, um, breed size. And so we looked at this by comparing the two hosts nestling treatment to the four hosts missing treatment and the number of, uh, per sanitary warbler nestlings that were reduced out of the nest that just that died, um, throughout the nestling period, because we had controlled for every form of predation and ecto Parris doesn’t that we could think of. Um, and what we see is that the number of hosts nestlings reduced is much greater in the four treatment compared to the two treatment. And even when we look at this proportionally, um, still the, the proportion of host nestlings in the, for treatment that, um, likely starved out were greater than, um, what we saw in the two treatment. And so we think that this shows some pretty good evidence for, um, Calvary mediated brood reduction And this seems to be pretty consistent with adaptive niche construction by the cupboards to increase their survival, um, by reducing the number of hosts nestlings, from what suboptimal closer to what is more optimal, if you notice the number of re um, the average number reduced was around 1.6. Uh, and so this is pretty close to, if you take four minus 1.6, that’s 2.4. And what we showed with our probabilities is that the optimal brood size in a predominately warbler NASA’s around two, 3.3. Uh, and so in summary, uh, coverage period, the engineers did their own routes. We identified a variable, or we identified a survival cost to variable boot sizes. We showed that this cost is paid early in development. We demonstrated that the likely mechanism for optimal survival is provisioning, and we showed that cowbirds are responsible for reducing their brutes to a more optimal brute size. And so with that, I want to thank my PhD advisor, Mark Halber, uh, other coauthors, Wendy Shefsky, Rebecca Kilner and Derek Toleman along with, um, Jeff Hoover members of the cupboard lab and my other two female assistants, Katie Stenstrom, and Angela Pirie. Uh, I’d also like to thank the department of evolution, ecology, and behavior at the university of Illinois at Urbana champagne. Um, again, fellowship through the department of education and, uh, the virtual Sylmar conference for having me for this talk. And with that, I can take any questions. Excellent work, thanks so much for your talk. Great job, Nick. I’ll just, I’ll let you know if any questions show up here. Here we go from Sarah McPeak, um, says, great talk. Uh, you’re finding that Calbert in essence, success is dependent on host brute size is intriguing, and it makes me wonder whether this might also impose selection on parents’ choice of hosts. So, do we see any evidence of cowbird parents preferentially laying eggs and nests with moderate brood sizes? Or do you think the effects of Cabernets liens are having on their brood size post hatching would swamp out any selection on the parent’s behavior? So I answered the first half of that question first, um, cupboard Calbert is lay, um, eggs before clutch completion And so they don’t wait until a collage is complete to make that decision. It’s typically when there are three or four eggs already in the nest that they will, they will lay their egg. Um, but there is there’s wide variation that, so it is totally possible that, um, they could, I mean, I’ve had a, cowbird lay, an egg in a nest that had nothing in it. And then a female warbler came in and started incubating that egg. It was the strangest thing ever this summer. Um, I have no idea why that’s the case, but so they will lay kind of at, at any time. But most of the time it’s when they’re either three or four eggs in the nest. And then what was the second half of that question? I’m sorry Um, it’s kind of a continuation of that, but saying, do you think the effects of the effects Calbert nestlings are having on their brood size post hatching would swamp out any selection on the parent’s behavior? So I, yeah, I think it could, we haven’t done that yet, but that

is, that is a thought. Yeah. Thank you. There’s a couple other questions here. One from Andrew Henry says any idea why the different brood, parasite species converge on different solutions to the same problem. Yeah. So there, the brood parasites are pretty diverse. They kind of cover really a whole half of the avian phylogeny Um, and they’re pretty well spread out across that. So I think, yeah, it just probably stems from that would be my take on that. Um, and then one other question, uh, from Maria is, is there any evidence that cowbird nestlings compete more with some of the hosts nestlings than others? So I’m working on that right now, that data, um, is being analyzed by a fantastic team of undergraduates. Um, and so hopefully I’ll know more about that soon That’d be interesting to know any other questions that was really cool. Thanks for the excellent talk there, Nick. Um, lucky last speaker for the session, Joey Bernhard is our next speaker is going to talk about life in fluctuating environments. Joey, can you share your screen with us? Yeah, let me just get it going here. Yep. That looks like it’s working Just give me one sec. Okay. There we go. All right. I will mute myself and it’s over to you now. Okay. Yeah. Thank you, man. Great And thanks to you. Um, all the organizers for putting this conference together, it’s my first ASM meeting and it’s been really fun to participate and I’m really looking forward to the next couple of days. Um, so today, uh, Lyrica today, I’m going to talk to you about one of the defining features of life on earth, environmental variability, and the ways in which living systems across scales from cells to organisms, to populations and communities cope with environmental fluctuations in order to survive and persist. So nature is variable. And, uh, sorry, I’m just trying to get this note. Oh, there we go. Okay. Uh, nature is very well and almost every environment on earth fluctuates over time and space, right? Environmental conditions vary from minute to minute from day to day and from season to season. And these patterns of variability differ across biomes and they differ across the globe. And these patterns, variability are expected to change into the future, increasing in some places and decreasing in others. And so living systems exist in a variable world and ecosystems are often structured by regular variation in the environment in this, in this, uh, in this top panel here, what I’m showing you is, uh, an example of a late ecosystem, which is famously known for, um, seasonality and erotic variation, um, which influences primary production, the timings of peaks and production, as well as secondary production through spring blooms, fall blooms and transitions, increasing pressure. Variability can also come from biological processes. So in these bottom two panels, for example, population dynamics that are internal to one species or to the connection between two species as shown in this links here, example define a situation in which oscillations in populations create variability in the environment, uh, that other species and parts of the ecosystem must respond to and environments, we know environments fluctuate at all scales. So the figure that I’m showing you here on the left is a time series of sea surface temperature off the coast of Norway and the plot on the right shows, the same time series, but now visualized in the frequency domain. And what we can see from this figure is that organisms with different generation times, different body sizes and other traits will experience the same environment in different ways and

respond differently. And so we now know that environmental variability and variation in responses to changes in the environment are fundamental mechanisms that allow biological diversity to exist and to persist, right? So it’s the environment of homogenous over time and space. We might find a situation in which one species or a limited set of phenotypes dominates the system, right? They might harvest all the available resources and lead to dominance or competitive exclusion of other species. And much of our work in population and community college is actually about understanding the feedbacks between environmental variability and ecological structure. And yet at the organismal and population levels, environmental variability actually poses a challenge to physiological functioning. And this is because organisms usually have a range of environmental conditions over which they perform well, right? And then if the environment varies over time and moves away from these optimal conditions, performance or fitness declines. And so this raises the question then of how do living systems persist in fluctuating environments if environmental variability poses a challenge, um, what are the solutions to living in, in variable environments, whether they’re predictable or unpredictable And this is a really important question to answer, uh, because we expect patterns of environmental variability to change in the future. And I’ve just shown you that responses to variation are critical to biodiversity patterns. And so understanding how living systems persist in fluctuating conditions is really critical to predict, to predicting future ecological States. So what I’m going to do today is, uh, present to you a new framework that I had developed with, with co-authors, um, which describes a set of mechanisms that operate across biological that allow living systems persist in fluctuating environments And this framework, and this framework, we outlined three classes of mechanisms that allow living systems to persist in environmentally variable environments. Well, before I go into those classes of mechanisms, I want to remind you that classically in population ecology, we, we often think about populations as responding passively or acting as filters of environmental variation. So populations essentially integrate, um, patterns of variation and they essentially sort of smooth patterns of environmental, uh, environmental variables So on, uh, in these sets of plots here on the left hand side will be some, a time series of environmental variation shown in the gray and the biological response to that variation is shown in the black. So in contrast to passive responses are simply filtering out, uh, uh, environmental variation, uh, I mean an outline three classes of mechanisms, uh, but the differ and the first is, um, is a class of mechanisms that rely on feedbacks. So feedbacks are probably very familiar to all of us here. Um, they rely on organisms or other living systems cells, or maybe even populations, um, responding to changes in their own internal state. And the key kind of, um, defining feature of feedbacks is that since the system is, is sensing a change in its own internal state, it can only respond after the fluctuation has occurred And so inherent in feedback mechanisms is an inevitable time lag in the response. So this biological response is responding, uh, to the variation in gray, but with some delay in contrast, if feedback isn’t as maybe a slightly less familiar class of mechanisms that we call feed forward and feed forwards before mechanisms differ fundamentally from feedbacks in that they, they rely on sensing a change in the environment and a change in himself. And so, uh, often they’re, um, they’re based on some sort of cue or signal that the living system uses to anticipate future environmental conditions So what this bottom plot shows you is a scenario in which, um, the, the system is able to respond in time with the environmental fluctuation, because it’s cute on some other condition that happens in advance of that fluctuation, essentially allowing the organism or the, or some other living system to predict the future environment because of some correlation between a cue and an expected future environmental state. So this type of response or mechanism might seem a little bit counter-intuitive

because it, in some ways kind of defies causality Um, we, we often think about organisms, uh, not being able to predict the future, but only being able to respond to the present, but actually the reliance on cues essentially allows the system to pull the, pull the future into the present and act, uh, in the present in, um, in anticipation of future environmental condition. And this is, uh, we argue that this is beneficial because it allows the system to re uh, to prepare in advance. So the law, the third class of, uh, mechanisms that we described here as what we call general adaptive systems, and these are essentially a combination of feedback feed forward. Okay. So feedback. So, as I mentioned before, are familiar to many of us, um, they are reactive processes that respond to changes after conditions have deviated from a set point. So a classic example that might be familiar to many of us is Thermo regulation, um, in which, uh, in, in which the organism is responding to deviations in internal core body temperature, as that temperature deviates from set some set point, um, the system responds to that deviation either through the constriction or dilation of blood vessels to return it back to that set point. And this returned back to a set point, uh, enables, uh, internal conditions to maintain a relatively constant over time. Uh, and, and this is beneficial for physiological functioning. So the key features of feedbacks are that they rely on some internal stimulus or some deviation from some internal state. And they, they are by definition, reactive, meaning that they, um, they have to occur after the change is synced internally. And so there is some time delay contrast to feedback is feed forward and feed forward processes are proactive and they’re anticipatory and they, and so they allow organisms and other living systems to prepare themselves in advance for some future state. And so here, I’m illustrating feed forward, uh, with an, with an example of Diallo Vernon, vertical migration, in which, um, CocoaPods here experience a variable predation risk at the surface, uh, of, of, um, of the water. And this is because, uh, predators of Copa pods are visual predators. And so when it becomes light out every day, predation risk increases because the predators can see their COVID pod prey. And so these CocoaPods though, if they internalize some correlation between some environmental cue, like a change in light availability, uh, and they can, uh, they can then capitalize on a correlation between light availability and predation risk at the surface to move away from the predators, swim down to depth in advance of predation risk, being the highest at, uh, at the surface. So you see that, um, here there’s some external environments of cue represented in this little sun, uh, that hap that allows the cocoa pod to swim down in advance of this event or this time period of high predation risk at the surface And so feed forwards are anticipatory, and they rely on external cues like changes in photo period or light regimes, which are correlated with some future selective environment like predation at the, at the, at the surface But what’s critical to understanding feed forward is to understand that feed forwards rely on an internal model, which, uh, relates some environmental cue to an expected future condition. And this model controls the response of the organism to acute. And so that’s illustrated here, uh, in, in this middle panel. So a feed forward examples are everywhere. And one example that I really like a feed forward is, uh, the dropping of leaves by trees in the fall. And so here, uh, these, these trees are, are responding to a change in photo period that happens in advance of impending winter conditions. And that allows the trees to, to drop their leaves before, before winter comes in there, they are going to be exposed to potential frost damage. So dropping of leaves is just one example of a feed forward a of a feed forward mechanism that relies on this correlation between a change in photo period and some future winter conditions. So feedbacks and feed forwards, uh, often or rarely occur in isolation. And they’re most often they most

often occur together. So that example that I gave you a feed of the feed forward of the dial vertical migration in CocoaPods might also occur, uh, in parallel with the feedback where, um, where copepods can actually feed back. They can, they can swim away from creditors after they’ve physically detected them in their environment. And so, um, so these, these two types of mechanisms often occur in parallel, but it’s useful to distinguish them because we expect that we expect these two different classes of mechanisms to change, uh, differently as the environment changes. But the key thing here is that, uh, all of these mechanisms combine allow combined allow, uh, allow living systems like this Cobra pod, uh, here to essentially minimize his experience of environmental variation So if predation risk in the, in the surface and the Epic logic zone varies over time, by moving away from, uh, by moving away from the surface, these, these CocoaPods can actually minimize their experienced predation risk. And, um, and, and this would Chris, uh, enhances their, their, um, performance. So how do you feed forwards arise? Well, if you forwards arise as organisms exploited, repeated associations between correlated environmental variables with a time lag to anticipate change. So here’s an example that comes from Ecolab, which I love, which shows. Um, so if you envision, or if you imagine, uh, some sequence of, uh, uh, in VR events, in an environment that are, um, shown here in these blue bars, uh, these blue bars might seem fairly unpredictable There’s a very wide distribution of time lags between each successive blue event. Um, and same thing for these gray events. There’s also a, a very wide distribution of, of, um, timelines, meaning that, um, these two events seem somewhat unpredictable maybe from the perspective of the organism, but what you can see is that, uh, there’s a very short timeline between the gray and the blue meaning that the, of the blue event is almost always proceeded very closely by the great event And, uh, and, and so in the context of the gray event, the blue event is actually highly predictable. Uh, even if the, the organism in this case that you collect can not predict the blue event, as long as it can sense the gray event. It, um, it can, it, it knows, or it can predict that a, uh, blue event will occur, uh, soon after And so this could be an example, uh, an example of this is a change in oxygen and a change in temperature. Uh, Eco-Line might not be able to sense or predict a change in oxygen in and of itself, but if equal, I can sense the change in temperature and that temperature is, is correlated with the change in oxygen Then all the Eco-Line needs to do is to be able to change sense of change in temperature in order to, um, uh, to predict a change in oxygen. And this is actually what equal I used to switch between aerobic and anaerobic metabolism as they go from outside to inside the human gut. And so here, the internal model would, would be sort of encoded in the metabolic network of the equalize cell cells themselves And so the forward processes that rely on correlations between multiple environmental variables and cues are everywhere. It’s really easy to find examples in phonology, circadian are another example of feed forward or queue based processes, as well as the diurnal vertical migrations. I just showed you and feed forwards and feedbacks. Don’t only operate within individuals when communities act as collectives, then feedbacks and feed forwards can occur, operate across scales from individuals to populations all the way up to communities. And so this is an example of coring quorum, sensing bacteria, um, which, which use feed forward to cause collective changes in gene expression and behavior. And in this case, the internal model that enables feed forward might be the architecture of the quorum sensing network itself. And so more generally the ways in which living systems sense, communicate, anticipate, and respond to environmental fluctuations. We argue determines patterns of biodiversity, um, themselves And it’s important to understand these feed forward mechanisms because, uh, human impacts are actually altering the informational Amelia and the reliability of cues. So an example that I’m showing you here builds on this previous

previous example of trees dropping their leaves Um, we, we have observed that when trees are living beside street lights, which essentially decouple the correlation between photo period and temperature, they actually hang onto their leaves much longer. And, um, this exposes them to, to, uh, increased frost damage. So not only that changes and cues can have ecosystem level impacts, 10 minutes left. Okay. Yeah Thanks. Um, uh, so a, a change, this is an example of an, of an ecosystem from Alaska where, um, bears beat on both salmon and elderberries, and what’s happened here, uh, over the last few decades of environmental change is that, um, the salmon pray, which rely on social cues to show up in Alaska and the elderberries, which rely on some sort of temperature cue, um, they, they are shifting at different rates causing, um, causing synchrony in the availability of these two resources due to these differential cue patterns, which is actually reducing, um, reducing the, the time span of rich prey are available in the ecosystem and having major implications for the sort of energy pathways through this system So to conclude what I’ve shown you is that organisms and living systems across scales use a combination of feedback and feed forward mechanisms to survive and persist, and the ways in which living systems and the variety of ways in which these systems anticipate and respond to environmental fluctuations, underlies patterns of biodiversity, but human human activities are changing the availability and reliability of cues. And so if we want to predict biological responses to global change, we need to study the internal models that organisms use to predict future environments and how these models, or how these models might adapt to changing selective environments So with that, I would like to think my sources of funding, uh, and if there’s time, I’ll take any questions. Thanks for that jury Um, unfortunately you finished right on the dots, so we don’t have time for any questions, but I’m sure there’ll be questions coming in the Slack channel and the chat. Uh, in fact, if people could direct their questions to the Slack channel at this point would be great because, um, it’d be easier for Joey to answer your questions there instead of, um, in the chat, which we’ll know which will cease to exist once we close the session, um, before we, um, close, uh, what I wanted to say was it has been such a fantastic day of talks, such engaging and inspiring talks from all of our speakers, both in the second and the first session, um, today. And I would like to take this opportunity to thank everyone for putting all, all their oldest effort into this, to get, give us such an inspiring I’m coming out of it, very inspired and excited by all the signs I heard today. Um, and I’m really looking forward to continuing this in the sessions tomorrow. Um, on that note, we still have a very exciting plenary from Scott Edwards coming up at nine o’clock, um, basically now So I hope to see you all over there. Um, and