Transcript

How Lizards use Signalling

Audio

You can also listen to the interview [MP3 19MB].

iTunes

Visit this channel at La Trobe on iTunesU.

Transcript

Matt Smith:

Welcome to a La Trobe University Podcast. I would be your host, Matt Smith, but if I were a lizard and I waved at you like this, or I moved my head at you like this, or I shook my tail at you like this, then it might make for a rather ordinary podcast. But you probably wouldn't understand what I'm saying. One person who could understand what I'm saying is Dr. Richard Peters from the Department of Zoology, and he would translate my movements like this.

Richard Peters:

Go away!

Matt Smith:

Richard, what could I possibly be saying if I'm making movements like that? Is it as simple as go away?

Richard Peters:

At our level of understanding at the moment, yes. These signals are used in territoriality, rival males or a male has established a territory, and if any intruder or another male wanders into his territory, he will start signalling to that male basically to indicate he's there, "This is my territory. You best move on." The intruder can then decide to move away, may give a signal of his own that suggests "Yes, okay, I see you. Just leave me alone. I'm on my way through."

Alternatively, that male could turn and perform his own signal as part of a challenge to that resident, and then they can exchange signals for a period of time. Typically, that's where it ends. One of them decides to go. But of course, if it doesn't resolve the problem, the males will come together, and physical fighting, although rare, can be the endpoint. Eventually, there's a winner and a loser, and the winner either retains his territory or takes over the previous owner's territory and that territory gives him access to females whose territories overlap.

Matt Smith:

What sort of signal movements are we talking about here? How diverse are they?

Richard Peters:

In terms of lizards, we see movement-based signalling in most of the dragon lizards. They're the agamid lizards, bearded dragons, water dragons, jacky lizards, which I study, and some of the other smaller agamid lizards as well.

We also see movement-based signals in skinks, although much simpler. So in skinks, for example, some of the small skinks you might see in your garden. If you sit and watch them for a little bit, they do tiny little head bobs. That's very simple motor patterns.

The dragon lizards, for instance, the jacky lizard that I study, their signal involves tail flicking initially, and I believe that serves an alerting function. It's simply there to try and get the attention of the intruder that's wandered through the territory. After a period of time tail flicking, they then move to full legs wipes, that's what we've called them, where they just simply move the foreleg backward and forward. Then they'll do push-ups and we've termed the final motor pattern a body rock, which is basically jumping off the substrate or doing a little bit of wrap. So that's quite a complex display, and then you get variations upon that across the lizards.

In other parts of the world, some animal, the anolis lizards have extendable throat fans, which are often colourful, and you see a lot of that happening in other parts of the world where the lizard will extend its throat fan, which usually is brightly coloured and then retract it. And that forms part of their signal as well.

Matt Smith:

So how has signalling evolved this way in lizards? Why do we get this sort of diversity?

Richard Peters:

That's a good question and that's actually the topic of my research. If I take a step back and talk about animal signalling in general, the prevailing model of signal evolution, the sensory drive model, explains signal structure in terms of the sensory system of the receiver, the environment in which the signal is performed and other aspects of context.

For sound, we know that the particular habitat will influence the way sound is transmitted, and animals vary their sound according their habitat. They also vary their sounds depending on what other sounds are in the environment at the same time. So if there's lots of noise, there might be other species or even noise produced by humans, we know animals will adjust their signal to ensure that their signal is heard, in many cases, above that of the noise, or at least at different frequencies.

I'm really looking for the same kind of level of explanation for movement-based signals. We don't know as much, partly because it's quite technical. We haven't had the technologies, or at least the know how, to unravel the particular factors that influence design. But it seems quite obvious now that the signals have to be detectable against plant movement. And so plant movement seems to be a major contributing factor. So the signals we see have to be of a movement type that is different to plant motion.

Matt Smith:

When a lizard detects a signal from another lizard, are they detecting more than just a signal? Does it go by what they see as well or what they smell?

Richard Peters:

Typically, for long-range signalling, it's the visual cue that they see. It might be a motion cue. That motion cue then might draw the attention of the receiver and they might be static visual cues like colour or patterning. So we see a lot of lizards even Australia that have motion signals but also have bright-coloured patches. It might be on their chest and so their signal might expose that part of the body.

Olfactory communication certainly occurs in skinks and a number of lizards. With the dragon lizards, there is some, not reliance, but use of chemical signals for territoriality. I note that when lizards see an intruder or if you present a rival to a resident, they sometimes will lick the substrate, and the hypothesis there is that they're looking for chemical cues in the substrate. Whether the emitter of that chemical cue is done so deliberately, which would make it a signal in order to influence the receiver or not, is an open question.

Matt Smith:

What techniques have you been using to do this research?

Richard Peters:

I combined experimental work where I'm interested in pairing up an intruder and a resident. So I'll house animals in enclosures, allow them to establish themselves as the resident, and then introduce another lizard. So I use live interactions. I've also used video playback techniques in which I present a lizard displaying on video, present that on a high-definition monitor to a resident male, just to get them to react to that displaying male.

In order to generate some of the video that I've used, I also make use of 3D animations. So I can then precisely define the signal I'm interested in. So in those kinds of experiments, I'm interested in what kind of signal, what structure of signal is most effective. So I can manipulate, for example, with tail flicking, how fast the tail moves, what kind of amplitude it does, and how long it signals.

So that's experimental approaches I've used. I've also been recording and using motion analysis techniques to try and tease apart the velocity information or the movement information in the signal. I film lizards displaying and I film plants, and then I use the motion analysis algorithms to identify the spatio-temporal information that's in that particular scene, whether it be a signal or plant.

Matt Smith:

Does the speed of a tail flick make a difference then?

Richard Peters:

In jacky lizards, it doesn't. So one of the hypotheses was that in noisy environments, you might expect the jacky lizards to signal faster. We've seen that in anolis lizards in the Caribbean. When plant motion is quite strong, they seem to signal faster. But jacky lizards don't. They don't try and signal fast in noisy conditions. They actually signal for longer.

And one of the advantages of using a tail structure for long-duration signalling is that presumably it's quite inexpensive. So there's not a lot of energetic requirements to signal for longer. The reason there is the difference between jacky lizards and anolis lizards I put down to the context of their environment. Jacky lizards signal right up close to the plants, whereas anolis lizards are generally quite separated from plants. And if you do some calculations, you can actually explain mathematically why it doesn't make sense to try and signal faster if you're right up against plants.

Matt Smith:

So that sounds like the way that a lizard has developed maybe evolutionarily has been really affected by the environment that they live in.

Richard Peters:

Most certainly, yeah. Environment in general will have constrained the kinds of signals we see, and the environmental conditions at the time of signalling will certainly influence the energetic or the timing of the signals of moment-to-moment variations.

Matt Smith:

So if you change the environment that the animal's in, will it adapt to it?

Richard Peters:

Yes, yes, definitely. I've demonstrated that in jacky lizards where I manipulated the wind conditions and therefore the plant motion. And I have lizards signalling in common when the conditions, and yes, there was clear evidence that in fact all of the animals tested changed their signal to longer-duration tail flicking. And incidentally, to intermittent movement, so start-stop signalling, and that's something that motion vision systems are particularly sensitive to, so the onset and offset of motion.

Matt Smith:

Now you said that you've been using 3D imaging and also using an image of a lizard to see what reaction to get from a real lizard. It sounds like a high-tech version of putting a mirror image in front of it. How does it react? Does it react the same way as if that's a real lizard?

Richard Peters:

They show clear evidence that they react to video presentations the same way they would to a live example. Whether they actually see it as a real lizard or as an intruder, I really don't know the answer. But, with video playback techniques, I can present a male displaying and they will display back. I can present an animated cricket and they would jump at the screen to try and eat the cricket. I've also presented looming raptors flying towards a camera and then making it big on the screen, and they flee from that.

And colleagues have used video animation to animate a death adder, and death adders have caudal luring where they move their tail like a worm, which is used to attract small mammals and maybe lizards. And when they use that with the jacky lizards, they also jump towards that as though it was a prey item.

So if you look at how they're responding to the different video classes, then everything is appropriate to that class. What they interpret the video to mean I guess is an open question, and it's up to us to design experiments.

Matt Smith:

What direction are your studies going to take you now?

Richard Peters:

The main area is really to unravel the kind of evolutionary forces that influence a signal's design. And a colleague of mine, Jan Hemi, at the ANU has come up with a strategy in which we're essentially evolving the signals. So we start out with a signal and then we present that to a population of lizards. The lizards respond or not. We record what they do, whether they respond, how quickly. And we have a whole set of signals. And then we use that to rate the signals.

The signals are defined basically as binary strings in a simulation environment. And we use genetic algorithms to then pair up, cross the two signals and mutate the signals to get the next generation. We just get this process happening in which we gradually see a signal evolve for the particular context in which we presented in.

And then we start to play around with the context. We might change plant motion. We can impose costs on signalling. So, don't signal for too long because that might attract a predator, so we impose a cost on the signal. So this is a simulation environment using genetic algorithms.

Matt Smith:

It sounds like you're trying to teach lizards in some ways.

Richard Peters:

It's not actually teaching because there's no learning, there's no reward and there's no punishment. So we're simply presenting some stimuli, seeing if they respond, doing some background, back office stuff, presenting a new set of stimuli, seeing if they respond.

So we're not trying to lead them anywhere. We're presenting them with a whole series of things and asking for their response. All their signals are created using animation. So we have a binary string that defines how a tail, we're using tail flicks, because that's the cue that animals use to attract attention. So all these are defined using binary strings that can be translated in an animation environment. I'm also following a different line in which I'm taking everything into an animation environment.

If I want to understand how plant motion affects signal design, I'm actually going to recreate plant habitats. I'm doing this in collaboration with Tom Chandler at Monash. And we're recreating virtual habitats and we can animate plant movement. So then we can ask really important systematic questions about the types of environmental conditions. How does the light environment affect things? What about the position of the receiver? Pretend we're a predator and we're looking from an elevated position further away. What do they see?

So we can ask all types of questions in an animation environment so we have minimal impact on animals except for the initial stage. And we can actually unravel more of the mysteries of movement-based signalling.

Matt Smith:

That's all the time we've got for the La Trobe University Podcast today. If you have any questions, comments or feedback, you can send us an email at podcast@latrobe.edu.au. Dr. Richard Peters, thank you for your time today.

Richard Peters:

Thanks, Matt.