A tail of moths and bats: a novel anti-predator trait

February 18, 2015 • 2:15 pm

The Saturniidae is a large family of moths (ca. 2300 species), and is also a family of large moths, including some of the world’s biggest and most beautiful.

Here’s one of them: the luna moth (Actias luna); this one’s a female. Notice two features of this insect: the long “tails” on the rear wing, and the eyespots on the same wings. There’s also one feature you can’t see: like mayflies, the adults hatch without mouths, so they can’t eat. They live only to mate and produce the next generation, and they die within a week after breaking out of the cocoon.


It’s been known for a while (from experiments) that these eyespots serve to distract predators, mostly birds who peck at them and, while possibly inflicting damage on the wing, also avoid pecking at the vital head and thorax, giving the pecked moth a chance to escape. There’s an obvious selective advantage in that. (In some lepidopterans the eyespots may also frighten away predators by resembling the eyes of scary birds like owls.)

But what about those long tails? A new paper in Proceedings of the National Academy of Sciences (US) by Jesse Barber and colleagues (reference below; free abstract and pdf) suggests that they help distract predatory bats.  Like most moths, saturniids are largely nocturnal, and so they’re subject to predation by animals that can’t see very well: bats! Bats hunt by sonar, and the authors tested the hypothesis that the moths’ “tails”, which spin and wiggle as they fly, are attractive to bats, who attack them and leave the vital body alone. They suspected that the tails had evolved to help the moths escape predation.

In short, the authors put luna moths together with big brown bats (Eptesicus fuscus) in a darkened and soundproofed room, and filmed what happened with infrared cameras (and listened with ultrasensitive microphones). They used eight bats and 162 luna moths, as well as “control” moths. In each trial, a bat was put in the room with one intact luna moth, one luna moth with its tail “ablated” (science jargon for “cut off”), as well as a control pyralid moth.  For convience in filming, moths were tethered to the ceiling with a length of monofilament line. Luna moths and big brown bats don’t live in the same place in nature, so the bats were hunting moths that they’d never before “seen.”

Here are the results from the paper, showing that the tails made a very big difference in the probability of being captured (you can also see five movies on the site showing the bats’ hunting behavior). My emphasis in the following:

Bats captured 34.5% (number of moths presented; n = 87) of tailed luna moths, and 81.3% (n = 75) of moths without tails were caught—a 46.8% survival advantage for hindwing tails (Fig. 1). Mixed-effects logistic regression revealed a moth without tails is 8.7 times [confidence interval (CI) = 2.1–35.3] more likely to be captured than a moth with tails. During each foraging session, we presented a bat with one intact luna moth, one luna moth with its tails ablated, and 1–2 pyralid moths as controls. Control moths were captured 97.5% (n = 136) of the time.

What happened is that bats aimed at the intact moths’ tails about half of the time, and when they did so they were only 4% successful at catching the moth. But when they aimed at the ablated moths’ bodies, they were as successful at killing it as they were when targeting the bodies of intact moths (72%).

While these results are strongly suggestive that the tails might have evolved to help elude bats, there are some caveats.

1. Maybe the tailed moths were harder to catch simply because they were bigger than the snipped moths. Big brown bats catch moths by enfolding them in a wing, and then bringing the tail up to scoop the moth into the mouth. Maybe the tailed moths are simply harder to catch that way, and so the tails evolved not to distract bats, but to make the moths harder to scoop up. To test that, the authors also presented the bats with saturniid moths that don’t have tails (Antheraea polyphemus), but are even bigger in body+ wing area than are luna moths. These controls were indeed caught less often than snipped luna moths, suggesting that bigger surface area does matter, but the effect wasn’t big enough to explain the difference between snipped and intact moths. However, the probability that the tails had an effect beyond just increasing size wasn’t impressive: the p value (probability that tails didn’t have an real effect but one that appeared by chance) was marginally significant—less than 5%. That’s significant statistically, but not terribly impressive.

2. Maybe there are other selective pressures that explain tails in moths. One suggestion is that they help the moths fly better. The authors analyzed flight behavior in snipped versus unsnipped moths and found that only the frequency of wingbeat changed—from 10 to 11 Hz (whatever that is)—but they claim that no feature of flight changed that could help the moth evade predators,

Well, okay, that’s something. What about sexual selection, though? Could the long tails be attractive to females, males, or both? The authors consider this unlikely because males and females have similar sized tails (sexually selected traits often differ in size or elaboration between the sexes, with males usually showing the larger or more elaborate trait). Further, the moths apparently mate at night, and don’t use visual cues for mating. In some species of moths, though, males do have longer tails than females, but the authors say that this might be due not to sexual selection but predation: males spend more time flying than do females, for they’re looking for the females, and thus are under more selection to avoid predation. That sounds a bit like special pleading.

All in all, though, I think the authors have a reasonable case that the tails deter bat predation, though it would be nice to do the experiment with other species of tailed moths. Further, the low significance of the results, especially in tests of the body-size hypothesis, is a bit worrisome. Still, the results are better than simply looking at the moths and making up an adaptive story without testing it! This was, after all, a very hard experiment to do.

Finally, the authors wanted to know how many time long tails in moths have evolved independently. They made a phylogeny (family tree) of representatives from 7 moth families using five nuclear genes and one mitochondrial gene, and then superimposed on that the tail sizes of the moths. The phylogenetic tree + tail size diagram shown below clearly indicates that long tails evolved independently in this group at least four times (the gray shaded bits of the tree). I’ve put the original figure caption below so you can see what the colors mean:

Screen Shot 2015-02-18 at 1.17.32 PM
ML molecular phylogeny of saturniid moths showing multiple independent origins of hindwing tails. Filled black circles indicate origin of tails. Open circles indicate losses. Branch colors indicate length of hindwing tail from absent (blue) to >50 mm (red), based on Phytools continuous character evolution analyses. Numbers by branches are bootstrap values. Gray shading denotes groups that have spatulate tails and contain species with tail lengths greater than 37.5 mm (the average for A. luna, n = 10). The images of saturniid moths used in these experiments are labeled: (A) A. luna and (B) A. polypheumus. Bold type and asterisks denote species that have tails longer than 37.5 mm. In combination with our bat–moth interaction data, this phylogeny suggests that tails serving a clear anti-bat function have evolved 4 times. Three additional origins of very short tails, of uncertain function, are also apparent.

The classic bromide of scientific conclusions holds here: “More work needs to be done.” Nevertheless, we know a bit more than we did before, and this result may lead others to look for additional and non-obvious antipredator strategies of lepidopterans. As we always say in the Darwin game: “Evolution is cleverer than you are.”


Barber, J. R., B. C. Leavell, A. L. Keener, J. W. Breinholt, B. A. Chadwell, C. J. W. McClure, G. M. Hill, and A. Y. Kawahara. 2015 Moth tails divert bat attack: Evolution of acoustic deflection. Proc. Nat. Acad. Sci. USA, early edition, 10.1073/pnas.1421926112

h/t: William

39 thoughts on “A tail of moths and bats: a novel anti-predator trait

  1. Great to be testing these ideas. But a tethered moth seems terribly artificial. I used to tether big saturniids so they would attract mates that I could catch and look at, and they really don’t fly naturally that way.

    I once flushed one of the longest-tailed of all saturniids, Copiopteryx, in daylight in Costa Rica, so I got to see it fly. It made the most marvellous loop to approach a roosting spot from below. The tails may have given it the ability to make this very steep, very rapid change in flight direction. That ability would be great for avoiding bats as well. More study is needed!

  2. Yes. Hz=Hertz, cycles per second, usually used by engineers when talking about electricity. Here it means beats/second.

  3. 2. Maybe there are other selective pressures that explain tails in moths.

    I think this must be part of it: AIUI, ‘swallowtail’ wing shapes are common in a lot of daytime butterflies, too.

    Imagine if we took that circular figure and use the gray shading to denote nocturnal species (instead of species with long tails). If the saturniids are all nocturnal as Jerry says, then the whole chart would be gray…and that doesn’t much indicate that nocturnalism selects for longer tails. OTOH if some of these species are diurnal or crepuscular, it would be very interesting to do that shading and see if the presence of tails correlates with nocturnalism.

    1. The tail sort of resembles a spider which relates back to this morning’s post on the spider-tailed snake. Looks like birds just can’t resist an 8-legged treat.

    2. Those tails are definitely antenna mimics. Some hairstreak butterflies even rub their hindwings together slowly to make the two “antennae” move as if they were sensing something–they cross each other, so movement of the wing makes them move in counterintuitive ways as if they were alive.

  4. For convenience in filming, moths were tethered to the ceiling with a length of monofilament line.

    And very frustrating to all concerned, I’m sure.

  5. How interesting!
    Now, a lot of questions come to mind.
    1) Idea #2, that tails may help (some) moths fly better seems plausible to me, especially as you point out that the tails twist a bit in flight. I have a secret hobby of making paper airplanes (having never grown up). I discovered many years ago that a ‘flying wing’ design of a gliding plane is very unstable, but if you simply prepare the plane so to cut out long tails on the hind edge, and twist them about 90 degrees, that they can be remarkably stable in flight. The tails seem to provide drag to keep them flying straight, and the twisty motion acts as a kind of auto-correction system for helping the plane stay straight. So maybe the tails help these moths control their flight better, and control means they are more nimble, hence their ability to elude bats.

    I wonder if one could add artificial tails to tailless moths and see how they fare in the bat test.

    2) The tails might be pseudo-antennae, directing a bird to peck that area if they find a moth at rest. Longer tails might direct the pecking to be farther away. The tails on swallowtail butterfly wings seem to be for this reason. So maybe.

    3) I am not sure, but it could be that tails in these moths is a primitive trait, and so they did not evolve several Xs but instead were lost several Xs.

    1. Point 3 is interesting, though it does not seem the family tree supports that hypothesis: the development of longer tails is pegged with filled-in circles, and the conventional moths tend to branch separately from those nodes. Interestingly, the long-tailed moths seem to be “closer” to the Ur moth than most short-tailed ones and with fewer branches, so it seems a long tail is an old adaptation that sticks (only two open-circle nodes signifying tail loss, one for a shorter tail and but one for complete tail loss).

    2. Your point 1 is true, long tails give greater stability, but it remains to be seen if ‘stability’ is a desired, from an evolutionary pov., trait. Why is, say, the F 16 such a good fighter jet? Because it is basically unstable, leading -in conjunction with its light weight- to great manoeuvrability (don’t get me started on the F35 disaster here).
      Ramphorynchids, with their long tails, were basically stable, the later pterosaurs replacing them, with a short tail, were much less. Let us note, however, that the stable pterosaurs did well for dozens of millions of years, so stability may not always be that much of a drawback, particularly (I guess) if no ‘unstable’ competition is around.

  6. Hz or hertz is the international measure of frequency, named after Hertz who was the first that observed EM waves as such.

    Hz = 1/s, so 10 hertz is a cycle time of 0.1 s, the wing beats making a base infrasound that is inaudible for humans (and bats) but not elephants.

    1. If elephants, under similar controlled conditions, catch a greater percentage luna moths than bats, then a rate of wing beats below bat hearing has been shown to have a survival advantage.

      Actias luna’s tails, therefore, evolved to reduce the frequency of their wings’ beats.

      1. “a rate of wing beats below bat hearing has been shown to have a survival advantage.”

        I think that (admittedly frivolous) hypothetical is worng.

        Bats echolocate by listening to their own beeps being reflected back to them, not (as I understand it) by listening to noises generated by their prey.
        In fact I think a rate of wing beats that almost matched bat ‘beeps’ – and hearing – would be quite advantageous in that it would interfere with the bat’s beep frequencies and confuse the echoes for the bats. (It would also be impossible to flap that fast, of course).

    2. That is amusing to imagine, an elephant annoyed by the sound of a swooping moth, much as we are annoyed by the sound of a mosquito buzzing our ears!

  7. I used to think moths were creepy and dirty. They are in my top five of all animals. Watching them is watching true brilliance.

    ¡Todos granizan polillas! Great post.

  8. It’s interesting that the fancier moths are of the non-eating, mothless variety – though I think most moths are of that sort. Is there no creature drab or annoying enough that its life cycle, morphology and evolution is not absolutely fascinating and sufficient to fill volumes? It doesn’t appear to be the case. I find it fascinating – and maybe a little poignant – that, in the case of these guys, all of that specialization for flying and that elaborate display for survival has developed for a form that mates and dies inside of a week. It seems like the path of least resistance would be for a caterpillar to simply evolve the ability to procreate and be done with it. You’d think that alone would drive a stake through the heart of “intelligent” design, but then that’s not an idea that has anything to do with critical observation, has it?

    1. “It seems like the path of least resistance would be for a caterpillar to simply evolve the ability to procreate…”
      I suspect once the dimorphism appears the creature is locked in. No backing out.
      It is interesting to think that the short duration of the mating phase might be a constraint on the rate of evolution. After all, if you spend a lifetime of many months or years of evading predators you’d think the many opportunities for culling would drive more rapid change. How many times can a moth be eaten in a few days of glory?

      1. It seems that with a lot of insects the adult is just a way to make more caterpillars/grubs, if that is not a daft way of putting the bleedin’ obvios. I think here of cicadas, various beetles that spend years underground etc.

    2. “Is there no creature drab or annoying enough that its life cycle, morphology and evolution is not absolutely fascinating and sufficient to fill volumes?” –

  9. I notice that the tails on the moths’ wings are not in a flat plane along with the rest of the wing – they look to me like they are slightly twisted. So I expect that when the moth is flying, the tails will set up some aerodynamic fluttering.

    Maybe that sets up a sound or vibration that the bats hear and concentrate on, attacking the tail rather than the juicy plump body?

    1. Indeed, unless those tails are unusually rigid, I’d expect them to flutter like ribbons in any sort of airflow. And besides being a distraction to the bat, they could also disrupt the bat’s echolocation.

  10. Bats captured 34.5% (number of moths presented; n = 87) of tailed luna moths, and 81.3% (n = 75) of moths without tails were caught — a 46.8% survival advantage for hindwing tails

    This arithmetic seems bogus to me. More than 65% of tailed moths survived, compared to less than 20% of tailless moths. That’s much more than a 46% increase in survival rate; it’s more like a 225% increase.

    That said, there’s a third caveat that must be addressed: these moths are adapted for tailed flight. Removing the tails means their flight reflexes are all wrong for their modified wing shape, and maybe that impaired their ability to escape the bats. It’s like amputating the little toes of Olympic sprinters and then assuming they’ll perform as well as if they’d been born with four toes.

  11. A couple of thought – without reading the above comments so forgive me if I repeat what someone else has said –
    1/ Could the longger wings just be a feature that has not been selected against? If advantageous to have the longer wings one would have expected all or many more moths to evolve that? Why do some Hyalophora have strange shaped forewings?
    2/ What is the advantage of experimenting with a bat that does not noramally occur in the same area & predate on these moths? Is the wing shape an advantage when the normal predator is around? I would expect the bats to have compensated for this if they evolved alongside each other…

  12. > The classic bromide of scientific conclusions holds here: “More work needs to be done.”

    Funny, you never see a paper ending “All questions were answered, further work is unnecessary.”

  13. Had the pleasure of seeing and getting a picture of one of these great moths on the side of my house in NC. I submitted the pic Jerry. Got a nice response, but alas, never made it to Readers Wildlife Photos 🙂

    An interesting read. It’s great to learn more about this moth’s features! Thanks!

  14. “They live only to mate and produce the next generation, and they die within a week after breaking out of the cocoon.”

    Well then, I’m sure these creatures will be especially gratified to be told by Christians that God approves of them spending that week getting their naughty on!

Leave a Comment

Your email address will not be published. Required fields are marked *