Edwards’ Dodo, painted by Roelant Savery in 1626.

When Animals Shed Their Wings

Worker ants can sprout wings—but don’t. Ant and termite queens destroy theirs after mating. Many island birds evolve into flightlessness. Can evolution explain why?

Richard Dawkins
Richard Dawkins
7 min read

And why the sea is boiling hot—
And whether pigs have wings

    ~ Lewis Carroll, Through the Looking-Glass

The sea is not boiling hot, though one day (about five billion years hence), it will be. And pigs certainly don’t have wings, but it’s actually not a silly question to ask why not. It’s a kind of jokey approach to a more general question: “If such and such is so great, why don’t all animals have such and such? Why don’t all animals, even pigs, have wings?”

Many biologists would say, “It’s because the necessary genetic variation to evolve wings was never available for natural selection to work on. The right mutations didn’t arise, and perhaps couldn’t because pig embryology is simply not geared to sprout little projections that might eventually grow into wings.” I am perhaps eccentric among biologists in not leaping immediately to that answer. I would add a combination of the following three answers: “Because wings wouldn’t be useful to them; because wings would be a handicap in their particular way of life; and because even if wings might be useful to them, the usefulness would be outweighed by the economic costs.” The fact that wings are not always a good thing is demonstrated by those animals whose ancestors used to have wings but who have given them up.

Worker ants don’t have wings. They walk everywhere. Well, perhaps “run” is a better word. The ancestors of ants were winged wasps, so modern ants have lost their wings over evolutionary time. But we don’t have to go back that far. Nowhere near. The worker ant’s immediate parents, her mother and her father, both had wings. Every worker ant is a sterile female fully equipped with the genes of a queen, and would sprout wings if reared differently, as queens are. The potential for wings is, so to speak, coiled up in the genes of all ants, but in workers it doesn’t burst forth.

There must be something wrong with having wings, otherwise worker ants would realise their undoubted genetic ability to grow them. The pluses and minuses for and against wings must be pretty finely balanced if a female sometimes grows them and sometimes doesn’t.

An illustration by Jana Lenzova, showing a queen ant shedding her wings.

Queens need their wings to found a new nest far from their original home nest. Wings also enable young queens to meet winged males not from their own nest. Workers, since they don’t reproduce, have neither of these two needs. They typically spend a great deal of their time underground, crawling through confined spaces. Perhaps wings would get in the way in the cramped corridors, galleries, and chambers of an underground nest. This possibility is vividly coloured by the fact that a queen ant, having mated for the only time in her life and then having flown to a suitable place to found her new underground nest, loses her wings. In some species, she bites them off, in others she rips them off with her legs.

To bite off your own wings is pretty drastic testimony that wings aren’t always desirable. They’ve served their purpose on the mating flight and the search for a new nest site. Surplus to requirements and probably an active hindrance underground, they are thrown away. Or eaten.

Admittedly, worker ants don’t spend all their time underground. They scuttle about foraging for food which they bring back to the nest. Even if wings are a handicap underground, mightn’t it still be a good idea to keep them so the workers could forage fast like their wasp ancestors?

An illustration by Jana Lenzova, showing ants co-operating to drag a millipede.

Well, wasps may be faster than ants but consider this: Foraging ants often drag home to their nest great lumps of food heavier than themselves: a whole beetle, for instance. They couldn’t fly with such a burden. Often, they collaborate in teams to drag even larger prey. Teams of army ants have even been seen dragging a whole scorpion along. Where wasps and bees forage over large distances for small parcels of food, ants specialise in food that is relatively close to home and which can be too large to carry in flight.

Even without a full cargo, flying is very energy-intensive. As we’ll see later, wasp flight muscles are little reciprocating engines, and they burn a lot of sugary aviation fuel. Wings themselves must cost something to grow. Any limb has to be made of materials that enter the body as food, and four wings for every one of the thousands of workers in a nest would not be cheap to grow. They’d be a heavy drain on the colony’s economic resources. Probably all these considerations tipped the workers’ balance towards not growing wings.

Termites are very different from ants in some ways, not in others. When I was a child in Africa, we called them “white ants,” but they aren’t ants, not even close. Where ants are related to wasps and bees, termites are closer to cockroaches. In their evolution, they independently converged towards an ant-like way of life from their cockroach-like beginnings, as ants evolved from their wasp-like beginnings. But there are important differences between the two outcomes.

Where worker ants, bees, and wasps are always sterile females, worker termites are sterile males as well as sterile females. But they are like ants in that the workers are wingless while the reproductive females and males (queens and kings) have wings, which they use for the same purpose as winged ants. And winged termites swarm in a similar way to ants—rather spectacularly at certain times of the year. I had childhood friends in Africa who, when the winged “white ants” were swarming, used to rush about stuffing them into their mouths—and, toasted, they were a local delicacy.

As with ants, and presumably for the same reasons (termites typically spend even more time in enclosed spaces than ants), queen termites shed their wings after the mating flight. Indeed, they turn into grotesquely swollen shapes, for whom the very idea of wings would seem like a joke. The head, thorax, and legs are unmistakably those of an insect, but the abdomen is a massively bloated, fat, white bag of eggs. The queen is just a walking egg factory—actually not even a walking one, as she is too fat to walk. She’ll churn out more than 100 million eggs during her long life.

Left: A winged termite queen. Right: The termite queen as wingless egg factory.

No birds bite their wings off. It’s hard to even imagine. The only remotely similar example I can think of among vertebrates is autotomy of the tail. From the Greek for self-cutting, autotomy is the shedding of the tail, or part of it, when a predator has caught it. It’s a useful trick that has arisen many times independently in lizards and amphibians. But never in birds. Unlike queen ants, no bird autotomises its wings.

Over evolutionary time, however, plenty of birds have gradually shrunk their wings, or even lost them altogether. Especially on islands—where more than 60 species of birds today (many more if you count extinct species) are known to have become flightless: among them geese, ducks, parrots, falcons, cranes, and more than 30 species of rail, including the tiny Inaccessible Island rail of Tristan da Cunha.

Why do island birds lose the power of flight over evolutionary time? Flightless birds are often found on islands too remote to have been reached by mammal predators or competitors. The lack of mammals has two effects. Firstly, birds, having arrived on wings, are able to take over the ways of life that would normally be filled by mammals; ways of life that don’t require wings. The role of large mammals in New Zealand was filled by the now extinct flightless moas. Kiwis behave like medium-sized mammals. And the role of small mammals in New Zealand is (or was) filled by a flightless wren, the Stephens Island wren (recently extinct), and by flightless insects, giant crickets called wētās. All are descended from winged ancestors.

Secondly, given that there are no mammal predators on their island, birds “discover” that wings aren’t necessary to escape being eaten. This is, presumably, the story for the dodos of Mauritius, and related flightless birds on neighbouring islands, descended from flying pigeons of some kind.

I put “discover” in quotation marks for a reason. Obviously, those ancestral pigeons, newly touched down in Mauritius or Rodriguez, didn’t look around and say, “Oh goody, no predators, let’s all shrink our wings.” What really happened over many generations is that those individuals who happened to have genes for slightly smaller wings than average were more successful. Probably because they saved on the economic costs of growing them. They, therefore, could afford to rear more children, who inherited the slightly reduced wings. And so, as the generations went by, the wings steadily shrank.

At the same time the bodies of the pigeons got larger. You could see this as diverting to other parts of the body resources saved through not needing to grow and service wings. Flying consumes plenty of energy, and diverting all that energy into other things, including increased size, makes a lot of sense.

Excerpted from Flights of Fancy: Defying Gravity by Design and Evolution, by Richard Dawkins and illustrator Jana Lenzova, published by Apollo. Copyright © 2022 by Richard Dawkins & Jana Lenzová.

Evolutionary BiologyevolutionScience

Richard Dawkins Twitter

Richard Dawkins is an evolutionary biologist and best-selling author. He is an emeritus fellow of New College, Oxford, and Professor for Public Understanding of Science at Oxford from 1995 to 2008.