Feathers Down Under: Exploring the evolution of sexual signals in Australian Fairywrens

IAN OWENS: My name is Ian Owens. I’m the director of the Lab of Ornithology, and it’s a pleasure to welcome you all to the 2025 Mundinger Lecture. The lecture was established in honor of Paul Mundinger, who was a PhD student here at Cornell and in the Neurobiology and Behavior Department.

So I’m very grateful to the Mundinger family for supporting this lecture, which was always meant to be an opportunity to celebrate behavior and evolution, and also to bring people together to talk about that across different levels. So it’s lovely to be able to do that. Let me see if I can turn– I’m being talked to by myself. Well, it’s going to keep going to keep doing it.

Well, in that context, it’s a real pleasure to introduce, so to speak, tonight’s speaker, who is Mike Webster. How many of you will already know Mike. His pedigree name is the Robert Engel, Professor of Ornithology in Neurobiology and Behavior. And he’s of course, also the director of the Macaulay Lab at the Cornell Lab of Ornithology. He’s been in those roles for about 15 years.

He grew up in the Rocky Mountains of Colorado. Then from there fled to the Coast, went to UC San Diego, where he did his first degree, then came to Cornell. First brush with Cornell as a PhD student, where he studied Montezuma’s oropendola.

Then went off to Chicago for a period as a research associate and a lecturer, then to Buffalo, before settling for quite a long time at Washington State, where he went through the whole set of academic hierarchy from assistant professor all the way through to full professor, before being lured back again to Cornell– you had been warned, Mike– about 2009, to take up his current roles.

One of the advantages of having somebody speaking in this series, that is from our community is, you know them well. And to me, Mike is really a consummate organismal biologist. His interests span behavior, ecology, evolution, migration, and so on. And he uses a diverse array of techniques to study those. Field experiments, field observations, molecular approaches, phylogenetics, and museum specimens.

I suspect, what we’ll see today is how you bring that set of approaches together to really understand how the world works. Not just describing patterns, but understanding how it works. I’m very glad that he’s almost always used birds as his study organisms. Started off with New World blackbirds in those days, went on and did wood warblers. And I think today we’re going to hear about some of the Australasian work using fairywrens. So not a bad selection of species, I have to say.

In total, if you look at Mike’s CV or you look at his online presence, I think he’s published about 150 papers. That’s a lot of papers. And if you look at how they’ve been cited, lots of them have been cited to an incredible degree thousands of times, particularly some of those migration papers. What you don’t see, though, by just looking at an online Google Scholar account or something like that, is everything else that Mike has done.

So some of that includes the teaching he’s here done at Cornell where he has effectively taught the fundamental behavior courses for 15 years. So it’s an undertaking in itself to keep yourself fresh for all those years. He has also helped to guide, I would say, a very large number of graduate students over the years. I think acting as– I had to count these through on the CV, Mike, but I counted about– as primary advisor to students, about 30 students. And then helping on the committees of about another 75 students.

So that’s 100 young people that Mike’s acted as a mentor for, which I think a lot of people end up looking back and realizing the 150 papers is what gets you your promotions and gets you known around the world, but it’s those 100 students that you mentor that can have an enormous impact in the world. As well as the students, he’s also played a prominent role in professional societies.

He was editor of Animal Behavior, one of the premier journals in this field, for a period. And most recently he was president of the Professional Society for American Ornithologists, the AOS. And he took on that just towards the end of COVID, when all of these organizations were trying to figure out, how do we run ourselves as a professional society? So Mike was president at a time when you really had to think about that and really modernize how these societies run.

And then, of course, in addition to that could be a classic academic career. But in addition to all of that, he’s also been the director of the Macaulay Lab during a period of extraordinary growth, evolution, and change. Probably centered around the idea of digitizing collections, and what you can do with digital collections. But also, of course, associated with the development of Merlin and everything that’s done, now the most popular nature app in the world, I would guess. So that’s two careers, at least in one person.

OK, that’s enough hearing from me about Mike. The last thing I would just like to say is how much I know that myself and so many colleagues, students, and friends have enjoyed working with Mike. I’ve tried very hard to get bad stories about Mike. I’m afraid I’ve failed. So many people have said what an intelligent and collaborative person, you’ve been to work with, and how much they’ve enjoyed that.

Which is quite a testament to somebody that’s been here for 15 years and then a career before that. And for me, always somebody who was willing to step up and fundamentally interested, as I said before, in how the natural world works, and exciting people with that mission. Over to you, Mike.

MIKE WEBSTER: All right.

[APPLAUSE]

That was really nice.

[LAUGHTER]

Can you hear me? Can you hear me, OK? Good. I have a lot of hardware on, so I might at some point fall over from the weight. But if I don’t fall over, it’s fun to be here and tell you a little bit about some of the things that we’ve been doing over the years, and I emphasize some of the things. So– there’s a sound coming out of my computer if I sit over here.

So there are basically three different areas that I’ve been lucky enough to be working over the last several years. One is in evolutionary science. I’m very interested in the signals that animals use to communicate with each other, how those signals evolve, and what the evolutionary consequences of those signals are.

The other area is more environmental, ecological. And interested in how organisms deal with an unpredictable and changing world, and in particular migratory birds, and how they deal with the fact that they don’t know what the world is going to be like when they come from the south to the north. And how they deal with it. And how the fact that it’s actually a little bit warmer every year when they come back, how do they deal with that.

And then the third area where I’ve been working is library science, so the Macaulay Library. And again, as Ian was saying, that’s been very gratifying. It’s outside of the research that I do, but firmly embedded in just the world of birding and getting people excited about birds. And what I want to do today is talk mostly about this first thing, the evolutionary biology and the evolution of signals. But I’ll touch on these other two things a little bit. And it’s all about birds. And there are two things about birds that are important.

[BIRDS CHIRPING]

So the first thing is birds are loud, right? They make a lot of noise. They’re constantly singing. This is a recording from the Northeast here. And if you go out during the spring and listen to this dawn chorus, you hear all these birds talking to each other. They’re all singing, exchanging the news, catching up on what happened overnight, and just having a conversation that we can listen to and enjoy. So that’s one thing about birds.

The other is that birds are colorful. They’re very, very beautiful. And this, of course, is also how birds talk to each other. They have plumage signals. They have displays. They have visual communication signals. Ways of talking to each other that are outside of the auditory and in the visual realm. And it’s not just birds, other organisms do the same thing. Birds are just easy to study.

These jumping spiders, also very colorful, also have very elaborate displays and also make sounds.

So this is a male jumping spider courting a female. And he’s facing her– It’s a black and white film, but he’s facing her with a very colorful front. The elaborately-colored forearms. He’s stridulating. So the sounds that he’s making are actually going through the substrate, and they’re being picked up by the female through her legs.

So she doesn’t have the same kind of ears we have. She’s picking up the vibrations through the legs. And he’s courting her. And if I let this film go on, you’d see he gets more and more excited because she’s not leaving or trying to eat him. And not leaving and not being eaten is a sign of receptivity in female spiders. And he gets more excited. He gets closer. But I’m going to stop the film because it’s just it’s disappointing when he discovers that the female he’s courting is freeze dried.

[LAUGHTER]

Point being that visual and acoustic signals and other modalities too are really central to courtship. How males woo females, how animals talk to each other, and birds exemplify that. I’m slightly afraid of spiders, and so that’s why I study birds instead of spiders. And Darwin was interested in the evolution of these kinds of signals and just elaborateness of them, because they didn’t really seem to fit nicely into his ideas about natural selection.

Because brightly-colored things that make a lot of noise and draw attention to themselves, you would think would be selected against by predation and natural selection, things like that. And so he pondered over that a lot, and he developed this idea of sexual selection. And he proposed that these elaborate traits, these ornaments, evolved through this process of sexual selection and that can account for why we see them in nature.

He proposed that it usually leads to ornamentation in males, not females. And also that these sorts of traits in the process of sexual selection may be really important to the process of speciation, or how a single lineage splits into two or more lineages. And so a lot of my career has been focused on trying to understand the exceptions to these rules, and maybe the limits of these rules, and how good a general explanation is this idea.

And so to talk a little bit about that work, I want to go down under right, go to Australia, where we’ve done a lot of this research. And Australia, of course, is full of kangaroos, and echidnas, and koalas, and really cool mammals. It’s also home to a lot of really interesting birds, parrots, and emus, and birds of paradise, and things like that. But it’s also home to the most spectacular family of birds of all, the fairywrens.

And I can objectively say that they are the coolest group of animals ever because of the species names. I mean, it’s the splendid fairywren. It’s not the white-throated sparrow. It’s the splendid fairywren. It’s the superb wren. The variegated wren. The lovely wren. They’re the emperor. I mean, these are amazing birds.

And so we’ve been doing a lot of work and all but one of the species on our previous slide are birds that I’ve worked on or had the good luck and fortune to work on. But I want to talk mostly about our work with one species, one that has a relatively drab name, actually. And I want to tell you three stories.

So the first is a story about variable sexual signals within a population. Why different males look different from each other, and maybe sound different from each other. And the other, the second story is about variation between populations, and whether that variation in these signals is important or not to the process of speciation. And then finally, at the end, if I have time, we’ll talk a little bit about birds on the brink.

Oh, yeah. So the first two stories are all about the red-backed fairywren. Not as cool a name as most of the other fairywrens, but it’s still nicely descriptive. And this is a pair of red-backed fairywrens, a male and female, preening each other. They’re extremely cute. The male, very elaborately-colored and ornamented. The female cryptically-colored and brown.

You can see they’re very social. They often have multiple group members. Young birds often stay with mom and dad and help breed the next nest of young. So they’re cooperative breeders. And you can also see that we band the birds, so they have colored leg bands, so we know who’s who. We can identify individuals.

And we also find all their nests so we can look at reproductive success. We collect blood samples so that we can look at genetics. And we measure them. We do all sorts of horrible things to these poor birds.

[LAUGHTER]

So yeah, this slide just says what I was showing during that movie. The other important point, though, is the rates of extra-pair of paternity are extremely high in fairywrens. An extra-pair of paternity, what that means is that this is a socially monogamous species. So one male and one female pair with each other, raise their young together, and defend the territory together, and all that sort of thing.

But it turns out that a lot of the young in the nest are sired by some other male from a different territory. And so males and females are socially paired with each other, but they’re also mating with individuals from other territories.

And about half of the young in fairywrens, in these red-backed fairywrens, are actually sired by males outside of the territory. An extra-pair male. And because of that, the force of sexual selection can be very, very strong males. Some males can sire a lot of offspring and other males very few because of these extra-pair copulations.

OK, so story one, variable plumage within a population. So here’s a different bird. This is a wire-tailed manakin. Also beautiful, elaborate plumage, a nice cool little physical display that he gives to court females. But what’s interesting is not all adult males look like that. So there’s a variation in plumage color across males of the same population in these wire-tailed manikins. And it turns out in many, many other species too.

So we have wore-adult manikins, house finches, buntings, you name it. There’s tons and tons of species out there where the plumage signals that males use to talk to other birds, and especially to females, vary a lot. And that variation is interesting to me. And some of that variation is age-related. So young males tend to be less bright than older males, but it’s pretty loose.

So in house finches, for example, those two males are probably the same age, even though they look completely different from each other. It’s not an age-related difference. And in lazuli buntings, young males tend to be more drab than older males, but the variation in one-year-old males is incredible. Some of those males look like females.

The guy at the top is very cryptically-colored, whereas the guy at the bottom is another one-year-old male, but he looks a lot like an older male, very brightly-colored. So that’s the questions like, why do we see so much variation, especially given that the females like the bright ones? So why wouldn’t every male be brightly-colored, if that’s what females are attracted they’re attracted to?

And red-backed fairywrens are the same thing. Red-backed fairywrens, I said before, males are brightly-colored, attract females with that bright coloration. But not all males look like that. So this is another breeding pair of red-backed fairywrens where the male looks almost identical to the female. You can tell he’s a male because he’s a black bill.

Breeding males have black Bills, whether they’re brown or brightly-colored, and he has just a tinge of red in his shoulders. So that’s a male that looks very much like a female. And so why would this male opt to breed in cryptic brown plumage?

And it is age-related. Young males tend to be less brightly-colored than older males. But within one-year-old males, young males, you can see brightly-colored breeders, you can see dull breeders, and you can see non-breeding helpers. So you get all three types of males within an age class. So these are two 1-year-old males. And they’re both breeders.

And you can see one is dramatically brightly-colored and the other looks like a female. And there’s a whole bunch of anatomical and behavioral traits that go along with plumage color. So compared to red, black-ornamented males, dull males look different. They have smaller testes. They are very good dads. They feed their young at high rates compared to the red-black males.

They make guard their female very closely. They very rarely leave their territory. And they have low levels of testosterone, whereas red-backed males show the opposite pattern in all of those things. So it’s not just color, it’s a whole suite of things that go along with color that are different between these two types of males. And the other thing that’s really different is their reproductive success.

So if you look at how many young these males sire, and we did that genetically, so we actually know who’s siring who. You can see that helpers are very few young, which isn’t too surprising. They don’t have mates. Dull-breeding males do OK, but they’re not siring that many males. But brightly-colored males, whether they’re one-year-old or older, that’s the OM2 and OM1. Where’s a pointer?

Yeah, kind of. So this is a one-year-old bright male, and this is an older bright male. They’re siring a lot of offspring and we know that most of that is because of their success in extra-pair fertilization. So they are siring a lot of young with other females on other territories.

And we’ve done experiments with captive birds and shown that females actually prefer to associate with bright males over dull males. So females like bright coloration. So again, here’s this question. Why would some males not breed in bright color if they could?

And so it’s not just fairywrens. This is a general question across birds, across other taxa too. Why do these sexual signals vary if they’re so important in mating success? So you can think of peacocks. Males have these great big elaborate tails, with big eyes, and everything like that. And they use it to attract females. But some males have really big tails and some males have small tails. And so why that variation?

And it’s probably because of cost. It’s not just the benefits of having a big tail. It’s also the cost of having that big tail. I mean, this thing is massive. You got to carry it around, that probably costs energy. Might put you at risk. And so it’s very plausible to hypothesize or suppose that males in poor condition can’t bear those costs. Whereas males who are in good condition can. And so that’s what we’ve been–

One of the things we’ve been interested in is identifying the cost of having this sort of ornamentation. And we’ve considered physiological costs. We’ve thought about ecological costs. We’ve thought about social costs. We’ve thought about social costs. So with respect to physiological–

Oh, the other thing I wanted to say is in red-backed fairywrens– now going back to our species, away from peafowl– you can see that plumage is associated with condition of molt. So males who are molting into bright plumage are in better condition than males who are molting into dull plumage.

And of course, this is a correlation. So we also did a field experiment where we manipulated condition of one-year-old males, young males. And males who we manipulated be in low condition, basically molted into brown plumage, whereas those who are in good condition molted into bright coloration.

And then we also know that testosterone is important to this whole process. So testosterone, if you give a one-year-old male a testosterone implant months before breeding starts, he immediately molts and molts into red and black plumage. Whereas if instead, if you give him a blank control, he molts into brown plumage. And so testosterone is part of the mechanism that induces a male to adopt this bright coloration. And it’s also associated with condition.

And so we thought, well, maybe it’s because testosterone is so costly, the poor-condition males can’t molt into red black plumage, because testosterone really is toxic. It does a lot of things to us, including causes to burn more energy, be more active, so we’re more susceptible to predation. Suppresses the immune system. So testosterone does a lot of things that it’s very likely or plausible that males who are in poor condition just can’t bear those costs. So we looked at that, and we did a few things.

One is that we gave males an implant of GnRH gonadotropin-releasing hormone, which is what induces males to produce testosterone. And we did that to males who were dull plumage and in bright plumage, red-black versus brown. And you can see if you give a male GnRH, they are perfectly capable of increasing their testosterone. There’s nothing holding them back. They’re perfectly fine at having those high levels of testosterone.

And then we also performed an experiment where we gave helper males who have very low levels of testosterone the opportunity to breed by removing breeding males from a neighboring territory. The helpers go over and take over the breeding position. They occupy that breeding vacancy. And as soon as they do, their testosterone levels go up.

So these are males who have just moved from being a helper to being a breeder, and their testosterone goes way up and their bills turn black. Which is another testosterone-dependent trait. And if you pluck feathers off their back and let new feathers grow in, they grow in red. So these guys are perfectly capable of having high levels of T and producing red-black plumage, but they don’t. So we don’t think physiological costs are what it’s about.

So we thought well maybe ecological costs. If you’re a brightly-colored male bird sitting in the environment, you’re pretty conspicuous. Cryptic females and cryptic males are pretty inconspicuous. There are a lot of things out there that you don’t want to be seen by, including accipiters and other predators. And so another very plausible hypothesis is that these brightly-colored birds are at a predation risk, and only high-quality, males in good condition can bear that cost, that risk.

So we’ve looked at survival. I’m going to flick through this slide really fast. The basic upshot is there’s no survival difference between males of different coloration, or even between breeders and non-breeders. But of course, this is correlational. It could be confounded by things like condition.

So we also did an experiment which was a lot of fun. When I say, I should say, Kristal Cain, who’s a collaborator at University of Auckland, worked with us. And she came up with this great idea that, well, we could put out model birds. So she made these 3D-printed fairywrens with little tails on springs so the tails go back and forth, which is what fairy wrens do in nature. And she painted a bunch to look like– this is a different species, but it’s the same idea.

It’s bright males, dull females, and bright females. In some species of fairywrens, some females are brightly colored. And she put all these models out in nature. And had actually 2,500 replicates. So put these models all over the place, different habitats, different parts of Australia, so she could look at the effects of plumage coloration and habitat type.

And she put a bunch of cameras on many of them, not all of them, to see what happened. And some just got knocked over by kangaroos and things like that, but many were attacked by predators. So here’s a couple of kookaburras. Kookaburras, by the way, are pretty voracious predators. A couple of kookaburras just tearing apart one of those models. They did like, yum, tastes like 3D-printed stuff.

[LAUGHTER]

And so then the obvious prediction is that the predation rates on the brightly-colored birds, this one and this one, should be much higher than on the cryptically-colored bird or models. And actually, she didn’t find that at all. It was actually perfectly one third of the attacks were on bright colored males, one third on bright colored females, and one third on cryptic females. So zero evidence that having bright plumage carries a predation risk.

OK, so we don’t think ecological costs are very important here. So then we thought, well, what about social costs? These birds are despite their tiny cute size, very aggressive, and do attack each other to defend their territories pretty vigorously, things like that. So maybe plumage color attracts the aggression of other conspecifics in the population, and that’s a cost to being brightly-colored.

And we do know that in laboratory settings, bright males are dominant over brown males. And so we did this experiment where we put caged birds out on the territories of other birds. And the bird in the cage could be brown, or it could be red and black. A red and black or brown male. And when you do that, the red and black males receive a lot more aggression from the territory owner. The brown birds receive almost none, whether it’s a brown female or a brown male.

So this does support this idea that well, as being brightly-colored, causes other birds in the population to be aggressive to you. And so maybe if you’re not a high-quality bird, if you’re not in good condition, you don’t want to be brightly colored.

So Joe Schalchlin, a recent PhD student, followed up on that work. And he did this really cool experiment where he gave a testosterone implant to young males during the non-breeding season, so prior to breeding, and induced them to start molting. Like this guy, he’s starting to grow in black feathers. And then he had control birds that got a blank implant.

And he looked at aggression towards the controls and the brightly-colored males, or those who were molting as they’re just starting their molt. And what he saw was there was much more aggression toward young males who were molting in bright feathers compared to the controls. And so to the point of– this is anecdotal– but one case where a brother chased for an incredibly long time, his other brother, who was molting into bright plumage, chased him around the territory.

Eventually caught up with him, drove him into the ground and attacked him on the ground. He was like, how dare you turn bright. You’re not supposed to do that. I’m your older brother. I’m supposed to be the right one. And we never saw that experimental male again. He either died from the attack or left the territory, left the study site. So it’s like it can be pretty costly to receive that kind of conspecific aggression.

And the other thing that Joe did was looked at exactly which are the young males who molt into red-black plumage versus brown. And what he found was the males who molt into red-black plumage are those who pair with the female early in the non-breeding season. They start hanging out with the female, start getting positive feedback from that female.

And those are the ones that molts into red-black. They start molting once they pair with that female. And they also reduce their home range size and avoid interacting with other males. So it’s like, I have a female. I’m going to molt into bright plumage, but I’m going to avoid other males, so they don’t attack me.

So the upshot of all of that is that we have these males who have the option of being ornamented or not. The ones who are ornamented have high fitness, are in good condition. The guys who are brown have low fitness. They don’t have many offspring, and they’re in poor condition. And it looks like what’s feeding into that is social cues. How they’re being interacted with by other birds in the population.

Like, are females paying attention to me? Am I getting a lot of aggression from other adults? And that’s affecting testosterone levels along with their own condition. If testosterone levels go up, they become a bright male. If testosterone levels go down, they become a brown male. And we think that early life affects that as well. But the important point here is that these young males are adjusting their sexual phenotype, what kind of plumage they have to the social conditions.

If it looks like they’re going to be a breeder, they molt into bright plumage. If not, they stay brown because they’re probably on a track to be a helper and not a breeder. And they don’t want to bear the cost, the social cost, of having the red and black plumage. But some of these guys get a mate later, but they don’t turn red and black. And the reason for that is they’ve already molted. They can’t molt again.

So if you molt and molt into brown plumage and then you get a mate, then you’re stuck as a brown breeder, even though it’d would probably be better to be a red and black breeder because you could sire extra-pair young. And so it’s really like these guys are making the best of a bad job, and they adjust their behavior accordingly. If I’m a brown male, I don’t have the plumage that females like. So instead of even trying to mate, I stay at home. I take care of my kids, I guard my female, and I’m a good dad.

Whereas the other males are doing the opposite of everything. They’re investing in mating and in particular mating with females on other territories. So that’s the one story, how males adjust their plumage signals to fit social conditions. Now I want to expand out and think more about across populations and not within a population.

And again, going back to the spiders. It’s a very common observation that closely-related species differ most in these sexual signals. The thing that really distinguishes them in these spiders, and birds, or whatever, is the signals that males use to attract females. And Darwin noticed that and thought that, well, maybe this is why sexual selection and sexual signals are really central to the process of speciation.

And after Darwin– he didn’t really know what was going on there. But well, after him, a few theories were developed to explain what could be going on. One is the classic speciation model, where you have a single population of animals that gets split into two. We call that allopatry. So you have one population, then you have two with very little gene flow between them.

When there’s low gene flow, those two populations can diverge from each other genetically and become more and more different over time, so that if they come back together and hybridize, the parents are genetically incompatible and so their offspring don’t do well. And we call that post-zygotic isolation. So the divergence of genes leads to post-zygotic isolation. And that in turn leads to selection to not interbreed with the other population, and that’s when you get divergence of these signals. So they don’t pay attention to each other. So this is the classic speciation model.

And in contrast, the sexual selection model reverses a couple of those steps. You again have allopatry. You have a single population that get split into two. But what diverges in this model is the very signals that males use to attract females. And so if those diverge and in the populations come back together, they won’t interbreed, even though if they did, their offspring would be perfectly fine. They’re genetically compatible parents.

And so in this model, you get divergence of signals, which leads to prezygotic isolation. They don’t mate with each other even though they could. And then genetic incompatibilities develop much later.

So again, going back to our friends in Australia, red-backed fairywrens are pretty nice species for looking at these kinds of questions for a couple of reasons. One, is that there are two forms. So the birds along the top end, the north part of Australia, there’s one subspecies called cruentatus. And I’ll just call them the red subspecies. And males have a very crimson deep red back.

The other subspecies along the east coast is melanocephalus. And males of that subspecies have an orange back, not deep red. So there’s quite a difference in back coloration, even though it’s the red-backed fairywren. Some are crimsom-backed and some are orange-backed. And they are genetically distinct subspecies. Scott Edwards and June Lee did some nice work a while ago, showing that they probably diverged from each other back in the Pleistocene.

So we’ve been studying and doing various experiments with three different sites. One in Darwin, which is right in the middle of the red subspecies range. Another down in Brisbane, which is in the middle of the orange subspecies range. And then Herberton, is right in the zone where the two subspecies come together. So you get red and orange individuals, and a whole spectrum of colors there actually in Herberton.

The other reason that we think that plumage color is really important and could be interesting to look at in this speciation question, is that the males really show off their back to females during courtship. So I’m going to show this little film. There are a couple of males up here. One or both of them are extra-pair males. They don’t belong in that territory. And one’s going to start courting the female. And there’s a bunch of female and young birds down here. And so just watch this male start to court the female.

So he notices them. He’s like, whoa, look at those. And now he just lifts that whole back, all the feathers on his back, get lifted up. And now we’re going to look at it again, but in slow motion so you can really see it. So he’s coming down. And it’s like a dinner plate behind his head, a big bright red dinner plate that he’s showing the female. And you’ll see in a minute when he leaves, he kind of hops over them and actually does a somersault, so that red dinner plate stays pointed at them the whole time.

So it seems like they really want that female to see the color of their back. That’s the way I interpret that. The other thing that fairywrens do that implies that back coloration is important is that they bring flowers to females.

So this is truly the most endearing trait of a fairywren. [LAUSGHS] The slightly less endearing part is they bring flowers to extra-pair females, or very rarely, to their social mate.

[LAUGHTER]

That’s a different story. But this is a male fairywren getting ready to court an extra-pair of female, and he has a flower in his bill. And they’re very selective in the colors of flowers that they select. And red-backed fairywrens always or almost always select red things. There’s the yellow flower petals, there’s white, and everything, but they always go for red flower petals, or berries, or red insects, to show to the female. And we think that it’s kind of complementing that back coloration.

So if you think about this, if coloration is important to divergence between populations, then it’s sort of like if they’re not interbreeding with each other because they aren’t attracted to the different colors, to the different plumage colors, it’s sort of like a fence being put up between the two populations that impedes gene flow.

It’s not a physical fence, but it’s a behavioral fence, and it impedes gene flow between the two populations, so that over time, you expect genetic differences to build up between the two. So you get DNA that’s different between the two sides of the fence.

And so if we mapped genetic variation across the entire range of the species and also variation in these signals where the genetics change, if you go from one genome to another genome type, that should be right exactly where the signals also change. Because the signals are the fence that impede gene flow.

So you expect to have concordance between a map that shows how plumage color changes, and a map that shows how genetics change over the range of the species. And it could be instead of plumage color, it could be song or some other trait. But whatever it is, it should line up with the genetic differences.

So we wanted to look at that. And so we sampled birds at a whole bunch of different populations. These numbers are just the number of birds we sampled. And in each population, we collected back feathers from the males so we could bring the feathers back into the lab and quantify the coloration with the spectrophotometer.

So we could actually, instead of just us looking at it, we could quantify how red or orange the feather was. And then we also collected lots of blood samples so we could look at genetics across the same range.

And this, you don’t need to look at this. By the way, I just put these graphs in for people who like graphs. I know a lot of you don’t. So if you don’t like graphs, you can completely ignore it. But this just shows how alleles or gene variants change over the course of this transect we ran.

And it shows that, for example, at this genetic marker, this locus, this variant, this allele, is a very high frequency. And then you hit a point where it just drops off and goes to 0. So on one side of this divide, the frequency is very high. On the other side, it’s very low. And we could look at a whole bunch of genetic markers to see how well they line up with each other.

So we did that for plumage color. We looked at the reflectance, the coloration of the feathers, the back feathers of the males. And this is the cline, so it changes. They’re pretty red, pretty red, pretty red. And then you hit a point where the redness goes away, and they become orange. And that point, the midpoint of this cline, is this dotted line right here. So this is what I call the plumage break between the two subspecies.

Doing the same thing with genetics, we see the same thing. This black line is the average of a whole bunch of loci. So you’re having one genotype up here. All of a sudden it drops off, and you have the other genotype down here. And this center right here is actually not concordant at all with the plumage. This point right here where you switch from one genotype to the other is up in the Gulf of Carpentaria, right up there. So you have the genotype is changing way before the plumage changes geographically.

So just looking at this a slightly different way, maybe more intuitive. Here’s the red birds, cruentatus over here. Here’s the orange birds on the east coast. If you go across this transect from west to east, they’re one genome type. And then you hit this point between populations 5 and 6 where you switch to being the orange genotype. And if you put on the same figure where they switch from one plumage type to the other, it’s way over here between 11 and 12.

And so something is pushing red plumage from one subspecies into the genome, the genetic background of the other subspecies. That’s weird, right? And so to understand that– well, before we just wanted to understand that we thought, well, what about song? They don’t just show their plumage. They also sing a lot.

Maybe song shows the same thing. So let’s look at song. This is where Emmett Craig came in, who was a postdoc in my lab before she started running Project Feederwatch. And so she went all over Australia recording birds, getting their songs. She went to the Macaulay Library and dug out every recording we had there of red-backed fairywrens.

And these are a couple of songs so you can see there’s a spectrogram, time versus pitch. And you can just look and see the songs are different. These guys have a higher-pitched song than these guys. These guys have a longer song than these guys. This is what they sound like.

[BIRDS TWITTERING]

So not the most melodious thing, but kind of twittery and nice. And this is the other subspecies.

[BIRDS CHIRPING]

A little faster, shorter, higher pitched, and in particular, has these three to four introductory notes right here that the other populations don’t have. And so if I hear a song, and I hear those three notes, I know where that bird was from.

So Emma did all this recording and then analyzed the heck out of these songs. She measured all sorts of things on the spectrograms, threw it all into a big principal component analysis that we don’t need to worry about the details of. And basically was asking, what do these birds between the genetic break and the plumage break, what do they sound like?

And it turns out they sound like the birds down here in the east, the orange birds. These are the yellow dots or the yellow populations. The orange dots or the orange populations. The blue dots are these crimson birds up in the top end. And you can see the orange and the yellow dots completely overlap with each other. They sound more or less identical. Whereas the guys in the west, the red-backed birds, the crimson-backed birds sound different.

So just showing this in cartoon land. We have the crimson-backed guys, are genetically different. They have a different coloration and a different song from the orange guys on the east coast. And these birds up in Queensland, up in Cape York, are chimeras. They have the genes and song of one subspecies, but the black coloration of the other subspecies.

So that all means– the interpretation of that is, song is a good barrier to gene flow. Song is preventing genes from flowing from one subspecies to the other, but for some reason plumage leaks through. Plumage is getting through that barrier to gene flow. And to understand why that is, you need to look at the behavioral responses to the signals. So songs are a reproductive barrier, but plumage leaks through.

So Emma worked with Dan Baldassare, who’s a PhD student in the lab at the time, and they did a massive, really cool experiment where they did playbacks of either orange or red songs. So one song type versus the other, paired with lifelike models that showed the plumage of either orange-backed or red-backed males.

So you could have an orange male singing an orange song, a red male singing a red song, or combinations. You could have a red-backed model with the orange song. And basically asked the question, does the male who we’re doing the playback to care about the song or the plumage?

And they did every possible combination. So they had local song, local plumage. Local song, foreign plumage. Foreign plumage, local song. And they did it both in Darwin, where everybody’s red-backed, and also in Brisbane, where everybody’s orange-backed. And so it was this massive experiment with, as I said, lifelike models. These are lifelike models.

So they’re basically clumps of clay with feathers stuck on them. And most of the feathers come from just like a hobby shop. But the back feathers are fairywren back feathers. So they had either red or orange backs on these guys. And this is a closely-related species, the white-winged fairywren, which looks totally different. It’s blue and white, and that’s their heterospecific control.

And they did all these experiments at Darwin and Brisbane. Did over 500 playbacks. Again, it was a massive effort. And I really liked this experiment because I don’t have to show a graph for this one. So I’m not going to show a graph. I’m going to show a video of the response of a local male. So this is where Emma and Dan have taken a model and a speaker underneath the model.

And then the model, you can see has a red back. And this is in the orange population. This is in Brisbane. So in the orange population, here’s a model with a wrong-colored back, but they’re going to play the local song, the orange song. So orange population, red plumage orange song type. And you’re going to see the territorial male come in and respond. There’s the introductory notes. That’s the orange song.

[BIRD TWITTERING]

You can see he has an orange back.

[BIRD TWITTERING]

He’s not happy. There is an intruder on his territory singing the local song. How dare he do that. And now you can even put a– it doesn’t even look like the right species. It’s blue and white. But underneath this blue and white model, they played the local orange song. And this guy just goes bananas over, who are you on my territory singing that song you shouldn’t be singing? And they don’t respond like that to non-local song.

So this is very– OK. This is why I like to say fairywrens are 7 grams of pure fury.

[LAUGHTER]

But I could show you the actual graphs, but they all look like that. The local song brings out very strong, aggressive responses from the territorial male. Doesn’t matter very much what the plumage looks like. And the non-local song doesn’t really elicit as much of a response. So males are paying attention to song.

So males respond to song and of ignore plumage to a certain degree. I’m kind of glossing over some subtleties, but they mostly ignored it. Well, what about females? So, do females like orange or red? What’s going on?

Sarah Kalil is sitting right there, has done a nice preliminary analysis. She told me to caution that she’s not done, but it suggests that hue– this is redness going from orange to red here in the hybrid zone, where you have a lot of variation in color– affects lifetime reproductive success. So total number of offspring that a male sires over his lifetime. Redder males are siring more offspring. And most of that is due to extra-pair matings. Extra-pair young that they say.

So that’s correlational. So again, it’s good to do an experiment to really see what the causation is. So Dan did this really nice plumage manipulation experiment in the orange population in Brisbane, where he caught a bunch of males and took a red art marker and turned them redder. And the color of red that he did actually very closely mimicked the red of the other subspecies. And then he had a bunch of males that he captured and he colored with a clear marker.

Now, I asked Dan– I don’t understand. I still don’t understand– why artists use clear markers because they have no color. I don’t see the point. But anyway. But they do. And so Dan had these clear markers that he could use as a control. So he caught a bunch of other males, colored them with a clear marker, and they were still orange. And that’s a nice control for the manipulation.

And then he let them go. He just let them go do their business, live their lives. And then found all the nests, collected blood samples from the young, blood samples from all the adults in the population, did the genetics to figure out who had sired whom in that population. And it turns out that with respect to within-pair young– so how many young males produce in their own nests– the manipulation had no effect. So they all produced about the same number of young in their nests.

But extra-pair parentage differed dramatically. So males who had been reddened, had their backs turned red, sired twice as many extra-pair young as– or more than twice as many extra-pair young as the controls. And you put those together, that’s total reproductive success, and the reddened males did very well in terms of reproduction compared to the controls.

So this basically mimics a new mutant in the population that produces redder plumage, or an immigrant that comes in that has redder plumage. And it strongly suggests that an immigrant like that would have very high reproductive success.

So the upshot from both of these together, the plumage manipulation and the playback experiment, is that males respond strongly to song, but not to plumage so much. Which means that males who don’t sing the right song would have a very hard time establishing a territory and defending it.

And so foreign songs should be selected against. It’s like if you want to put a keep-out sign in your yard here in Ithaca, it probably wouldn’t do really good to have it in Lithuanian or something. You have to speak the local language in order to defend your yard.

But females appear to be paying attention to plumage color, and they really like red, and even in an orange population where females never seen a red male before. In these populations, where red is totally novel, they are like, wow, that red’s pretty. I want to mate with that guy. So there’s a preference for red. That means that redder males will have an advantage.

So together, this is all like song is a reproductive barrier to gene flow. But pre-existing female preferences seem to open a gate in that barrier and allow plumage color to go through. So then the next prediction is, well, that means we should be able to identify genes that affect plumage color that are moving through that gate. And so that’s where Sarah’s work– Sarah Khalil is a postdoc here at Cornell now.

And Sarah has been looking to see if she can find genes, alleles that show the same cline, the same change going across the transect as the plumage cline, and to differ from the genetic cline that’s shown by other genetic markers. And I’m going to just gloss over a whole lot of genomic work that she’s been doing to identify those genes that seem to.

But the upside is she identified a bunch of genes that are associated with red plumage coloration. She sequenced hundreds of birds across this transect for those loci. And she found a handful, I should say, that do seem to track.

So here’s one. It’s called FASN, which is fatty acid synthase, which very closely tracks the plumage cline. So the cline in this one gene tracks the plumage cline very closely and is quite different from the genetic cline. And genotype at this locus– I won’t describe this fully, but I’ll just say that the genotype at this locus– affects plumage color.

So males who are homozygous for one variant of the gene are red. Males who are homozygous for the other variant are orange. And so she found a few other loci as well that seemed to show the same pattern.

So this really strongly suggests that this gate that’s been opened by female preferences in the wall that divides the two subspecies that’s produced by song, that gate is allowing certain genes to get through. And the genes that are getting through are specifically those that affects the male back coloration.

And there’s been a lot of recent studies, very recent studies, like the last year or so that seem to be finding similar patterns. And we’re starting to think, well, maybe these visual signals in general, in birds are not a big barrier to gene flow. Other traits like song may be. But it seems like there’s a lot of cases where genes affecting plumage are moving from one subspecies to another, or even sometimes across species boundaries.

All right. So really quickly, the lessons from our red-back work is that males adjust their sexual phenotype to social context. That social costs are important to maintaining that variation. Sexual signals like song, some sexual signals form a reproductive barrier, but others do not. And in particular, pre-existing preferences can facilitate gene flow across these barriers.

So I just want to super quickly close with a totally different thing. Going back to birds on the brink, which kind of gets me back more into the lap of ornithology and the Macaulay Library. So this is the first species of fairywren that I worked on.

I started working on these guys when I was a postdoc. This is the splendid fairywren. I worked in South Australia on them. This is an amazingly detailed eBird map of the distribution of the splendid fairywren. So you can see in gory detail exactly where it’s abundant and where it’s not, thanks to eBird data. And if you stack a whole bunch of eBird data together and divide the years, you can actually look at how populations are trending in this species. Are they increasing or decreasing? And the picture is actually pretty grim.

Red in this map means populations are declining. Blue means populations are increasing. So for the splendid fairywren, it’s doing poorly through most of its range. The populations are declining pretty dramatically. Just down here in the southwest by Perth, they seem to be doing OK. And that is not an isolated story.

So bird populations across Australia have declined about 60% in the last few decades. Very similar pattern in North America, very similar pattern in India. And a lot of places where people have looked, birds are declining.

To reverse those trends, I would argue that three things are necessary. One is good research to figure out the cause of those declines. Another is conservation efforts and policy fueled by that research. So what we learned should guide how we try to reverse the trends. And then the third thing is public engagement. So you’re not going to save birds, or any other animal, or plant, if people don’t care about it. You have to have– the public really needs to be passionate about the animals that you want to save.

And that’s one of the guiding principles of the Lab of Ornithology. We do a lot of research. We do a lot of data analysis that feeds into guiding policy and conservation efforts. But we also put a lot of effort into public engagement and getting people to care about birds. And one of our main tools is Merlin, that Ian mentioned earlier, which has now been–

I should say, Merlin started as a bird identification app a few years ago. We added sound identification to it, Sound ID to it. And since then 2.7 billion birds have been identified with Merlin by voice. And if you go onto TikTok and gather TikTok videos of people talking about birds in Merlin, there’s hours and hours and hours of those TikTok videos. These are all just screen grabs from some TikToks that we found recently.

I could have put up 30 or 40 more. And the important point here is that this slide shatters the stereotype of who’s a birder. People care about birds. And we like to think that Merlin is bringing a lot of people to birds who never knew anything about birds, even NFL football players. So it’s the tool that we are trying to use to engage the broader public, and I think it’s working.

So here’s a graph of the number of people using Merlin across years, across time since it first started. This dotted line right here is when added Sound ID, and it exploded. It just went through the roof with people really getting excited about identifying birds and knowing something about birds in their backyard.

The problem, going back to Australia, is it doesn’t work in Australia. Only 10% of the birds in Australia are in the Merlin Sound ID model. And the reason for that is we don’t have enough recordings. You need about 100 or 150 recordings to train the algorithm to the sound model uses. And we don’t have enough of those recordings from some parts of the world. And embarrassingly to me, that includes Australia, where I’ve worked since 1992. I’ve been very ineffective.

So to try and change that, we are doing sound recording workshops at a lot of different places across the globe, including Australia. So here’s a couple of workshops that we did in Australia last year to try and teach people how to record birds, and how to upload those recordings to eBird so that they can go into the Macaulay Library, and we can use them to train the Merlin Sound ID model.

And so the hope is– and I’m going back next month in November Jay and Andrew and I will be back in Australia leading more workshops to try and build this network of recordists who are recording and uploading recordings of Australian birds. The goal being to help more and more people fall in love with birds in Australia, as well as North America, Asia, Africa, and South America, where have you. That’s the ultimate goal.

So with that, the last thing I want to say is to take about one minute, is that science is a team sport. Nothing, literally nothing I’ve talked about here today was me. It was a whole bunch of people who did a whole bunch of really cool and hard work. I just was standing near them while they did it. And that includes the folks at the Lab of O and the Macaulay Library who’ve been diligently making the Macaulay what it is, making Merlin what it is, making eBird what it is, so that people can upload things.

So here’s a couple of group shots from some fairly recent events. I have been privileged to work with people who are really passionate about what they do, and creative about what they do to make all of this happen. So I’m grateful for that.

And then likewise for the research, all of that research was done by many, many, many people over 20 years, over 20 years of research on red-backed fairywrens. I couldn’t fit them all onto this slide. So this is just a select few who really were major contributors to what I talked about tonight.

There’s probably an equal number who are off the sides of the screen that you can’t see who did a lot of really important work. And I do want to just call out and thank them first off, but also Emma Greig, who was a good friend and collaborator. I’m very grateful. OK. Recomposing myself.

Also a lot of funding agencies who helped make this happen. A lot of people who didn’t make it onto the last slide with all the photos, who I really appreciate their help. And thank you all for paying attention.

[APPLAUSE]

And I am happy to take a couple questions. I know we have to be out of here by 7:00. So we don’t have a ton of time for questions, but I’m happy to answer any that you have. Yes? Max, is that you?

AUDIENCE: A really fun talk. I’m curious, with the red-backed fairywrens, if there’s selection sexual selection for redded backs, then what causes the orange population to survive at all? Why does the red not completely introgress when–

AUDIENCE: Can you repeat the question?

MIKE WEBSTER: Yeah. I will repeat the question. The question is, why are there orange populations at all? Why haven’t the red-back coloration swept through the entire range? And part of the answer is that, maybe it will. Maybe we’re catching it in process. But the other thing is that there is a biogeographic barrier near where that plumage break occurs. It’s called the burdekin gap. It’s an area of dry– it’s pretty dry there.

And it’s a major biogeographic gap for a whole bunch of taxa. So it may be very hard for birds to disperse across that into the populations that are still orange. But we don’t know, that’s just a hypothesis.

AUDIENCE: Thank you.

MIKE WEBSTER: Sure. Yes?

AUDIENCE: I don’t know if I missed that, or related to the question. Is that the orange trait more ancestral than the red?

MIKE WEBSTER: A very good question. So Laurie’s this question is, is orange ancestral to crimson or red? And our hypothesis that is the case. So if it’s a pre-existing female preference for red, then that implies that there used to be orange in allopatry during the Pleistocene when the two when the populations weren’t interacting.

Mutations for redder plumage appeared in one population but not in the other. And now that they’re coming back together, the female preference for red allows the new red alleles to pass through. We haven’t tested what is the ancestral state genetically yet, but that would be the prediction. It’s a really good. Good observation. Brian?

AUDIENCE: Do the ornamented males and the unornamented males that are breeding both sing roughly a similar songs?

MIKE WEBSTER: They do. Yes, I just barely touched on song. We’ve not found any difference in song between them. We’ve also looked at the song centers in the brain, no obvious differences. Females also sing, and they do duet with males. And females do have a slightly different song, but not majorly. And they have slightly different song centers in the brain, but only slightly. The song differences are pretty minor within a population between different types of birds. Yes?

AUDIENCE: You talked about experimentally manipulating brown males to go into red plumage, but have you ever found that actually occurring in the wild? And also does it happen vice versa? Does a male with a more costly plumage–

MIKE WEBSTER: Ever go back to being brown?

AUDIENCE: –yeah.

MIKE WEBSTER: A great question. So the first part is, it happens all the time. So males who are brown early in life, if they survive, turn red-black. So it’s really, your first year, sometimes two years of life, is where the plasticity is. You can be brown or red and black. And males who survive eventually do get red and black, so that happens all the time. Do we ever see males who were red-black in one year go back to being brown? We’ve never seen that in our field notes.

We have a couple of weird observations where it seems like that’s happening, but it’s so rare that it’s either a super-rare event or we have mistakes in our field notes. It’s like we just wrote something down wrong. So I don’t think it happens very often, if at all. Which is a good question, why not? Yeah?

AUDIENCE: Are there are different levels of conspecific aggression between the red and orange populations?

MIKE WEBSTER: Yeah, that’s good. We did look at that. I’ve forgotten. I think the answer is yes. I can’t remember which direction it is. I think the more red populations in the north are more aggressive. On average their aggression scores are higher. But it might be the opposite. It might be the orange ones are more aggressive. I honestly can’t remember right now, but there seems to be something going on there with aggression. Yes?

AUDIENCE: If the brown versus colored phenotype is related to condition, do you see different rates of brightly-colored males in different years?

MIKE WEBSTER: Yeah. So that’s a great question. And Joe, he looked specifically at that. And so the question was, do you see in harsher years, drier years, does that affect the number of males that go into red-black plumage versus brown plumage? And the answer is for young males, yes. No, sorry. For older males, yes. But even stronger than that is when they molt. In harsh years, they’re molting really late.

One thing I didn’t say is, the older males are molting months before the breeding season actually starts. Young males don’t. They wait. But they can molt weeks, if not months ahead of breeding, and that’s strongly affected by environment. In harsh years, they don’t molt until it’s almost breeding time. In lush years where there’s a lot of food, they molt very early.

Young males, it’s not tied to environment. It’s much more closely tied to social conditions. Did they pair or did they not? And when they pair determines if they molt or not. And so, yeah. So for older males exactly. But for younger males, it’s like, do I have a mate or not? Glen?

AUDIENCE: Thinking back over 20, 30 years?

MIKE WEBSTER: 20 something. Yeah.

AUDIENCE: If you had to–

MIKE WEBSTER: I’m feeling incredibly old right now, by the way.

AUDIENCE: If you had to go back and do it again, what might you do differently? What did you do right? And what things might you–

[LAUGHS] Wow, how much time do we have?

[LAUGHTER]

So a brief answer is that I think I would have liked to figure out how to get at some of these questions a little more quickly than we did. There’s a lot of trial and error, and experimentation. So the experiments I showed you were the ones that worked, and there were a lot of experiments we tried that didn’t work.

So, for example, Joe tried another experiment. We just use a marker and turn brown males black and red. Didn’t work because it just didn’t work. And things like that. And in hindsight, when we look back and look at the ones that failed, it’s was like, oh yeah. Obviously that’s why that wouldn’t work. I wish we had thought that through better before we started.

So in hindsight, I would say very carefully think through the experiment and whether it’s likely to work or not, because a lot of them don’t. And the ones that don’t, often, oh, yeah, I can see why that didn’t work. And I should have figured that out before I even went to the effort. So that’s one small thing, but there are a lot of other things I could talk about.

IAN OWENS: OK, that sounds like a good moment before we go on to the small things. Thank you, Mike, once again, for the talk this evening and all you’ve brought to Cornell over the last 15 years. Thank you, Mike.

MIKE WEBSTER: Thank you all. Thanks.

[APPLAUSE]

End of transcript