Evolving backward - natural selection and the shedding of non-useful physical traits
Sept, 1998 by Jared Diamond

The blind mole rat -- a fat, furry sausage of a rodent with tiny legs and no eyes visible -- is helping evolutionary biologists figure out how the human lost his tail.

Most discussions of biological evolution concern the acquisition of useful traits. As Darwin showed, the process involves natural selection: an individual animal that possesses some useful characteristic is more likely than another animal to survive and leave behind offspring with similar advantages. But evolution is also about losing traits as well as acquiring them, about discarding old features that have become less useful. The long list of ancestral traits that were lost or reduced in the course of human evolution includes tails, body hair, wisdom teeth, the ability to synthesize vitamin C, the size of our teeth and appendix, the thickness of our skulls, and the bony browridges over our eyes. Likewise, snakes lost their legs, whales lost most of their sense of smell, and the dodo and many other birds on remote, predator-free islands lost their power of flight. The process is common enough to be unsurprising. Yet it pays to remember that such traits took remarkable luck as well as many thousands or millions of years to arise in the first place. Why would any animal cavalierly dispense with such hard-won victories? What evolutionary force drove their loss?

It is not at all obvious how natural selection could provide the answer. What harm would it really do us to sport a tail? Why do we survive better without one? A tail wouldn't get seriously in our way: we could occasionally use it for balance or gesturing, otherwise curl it up or hold it aside, and let it protrude through a rear-facing fly in the back of our trousers or skirt. That would hardly constitute a disadvantage. So why did every single one of us end Up without a tail? Why would it be such a tragedy if we had somewhat larger teeth, a larger appendix, or a thicker skull?

Those turn out to be difficult questions. Darwin proposed two alternative solutions but was unable to decide between them, and biologists continue to debate them today. They are most easily understood by considering the corresponding two explanations for why a derelict building, once abandoned, will probably no longer be standing 20 years later.

One of the explanations invokes normal wear and tear, which goes unrepaired when a building (or a gene for a biological trait) falls into disuse. When the owner stops repairing and maintaining a building, windows that crack from time to time will no longer be fixed, rain and wind and thieves will come in, the building will rust and crumble, and eventually it will fall down by itself. The biological analogue of wear and tear is that genes occasionally experience mutations, most of them harmful. (When you blindly poke a complex device like a piano or a gene, you are much more likely to damage it than to improve it.) As long as a biological structure is in use, natural selection tends to eliminate individuals bearing harmful mutations of the structure because the individuals are thereby placed at a disadvantage. But if the structure is useless anyway, mutations dismantling it would be selectively neutral and would do the animal no harm. Eventually, enough mutations will accumulate to prevent the structure from being built at all. For instance, as long as our proto-ape ancestors stayed in trees, mutations dismantling the tail meant bad news for the animal and were selected out. But once ancestral apes were spending much of their time on the ground, mutants with shrunken tails were at no disadvantage, and eventually mutants whose tails were shrunk to nothing became the norm.

The other explanation can be described as motivated wrecking for profit. A derelict building's owner may call in a wrecker because he is having to pay property tax without receiving any rental income, and he would like to use the space differently so as to earn money, even if only as a parking lot. In the biological realm, any physiological structure occupies space in an animal's body and costs energy to make and maintain. But an animal has only finite amounts of energy and body space. Hence mutations that eliminate the useless structure are not merely neutral but positively beneficial, freeing up space and energy that could instead be devoted to some useful new structure. For instance, a mutant proto-ape that lost its tail could have devoted the freed-up energy, protoplasm, and space to a larger brain. Hypothetically, the proto-ape got more benefits from its newly enlarged brain than from its old lost tail.

Darwin's two competing theories make for good cocktail party conversation. But they would be nothing more than that if we couldn't figure out how to test them. To resolve this debate, we'll examine the story of a remarkable Israeli animal that has nearly lost its eyes, and of a remarkable Israeli scientist who figured out why the animal did so.

The scientist is a delightful, inexhaustibly energetic, and sparldingly creative evolutionary biologist at the University of Haifa named Eviatar Nevo and nicknamed Eibi. He has studied virtually every interesting animal and plant species that he sees in Israel, ranging from algae, ants, barley, beetles, and crickets to salamanders, shrews, snails, wheat, and worms. He juggles more research projects simultaneously and publishes more papers each year than the entire scientific establishments of numerous countries. In addition to testing his ideas himself, he has a gift for enlisting other scientists with specialized talents to collaborate with him. When I finally met Eibi in person in 1992, within minutes of my walking into his office he was proposing to send me Israeli rodents by airmail so that I could apply my experience of digestive physiology in a collaborative study.

In 1948, when the state of Israel was founded, Eibi was living in a kibbutz near the Lebanon border. While plowing vegetable gardens with his mule, he kept encountering domelike mounds more than five feet long and one foot high, which scared the poor mule. His curiosity aroused, Eibi dug open one of the mounds, to find inside a spiraling labyrinth of tunnels, storage chambers full of edible plant parts, a nest chamber, and a toilet chamber full of the droppings of some small animal. Three pups in the nest identified the architect of this miniature city as a rodent known as the blind mole rat, or Spalax ehrenbergi.

Most people would have dismissed Spalax as ugly and nasty because it bites any fellow Spalax and any human who handles it. Eibi being Eibi, though, he considered the rat beautiful and fascinating. In 1952, as a student at the Hebrew University in Jerusalem, he began to study it, inevitably attracting dozens and dozens of collaborators. He started publishing articles on every aspect of blind mole rats: anatomy, behavior, biochemistry, genetics, evolution, physiology. When I last took count, Eibi's publications on mole rats numbered 240.

Spalax resembles a muddy-colored, furry, fat sausage about eight inches long, because at first glance one may not notice any marks or appendages: easily overlooked short legs, no tail, and not the slightest hint of eyes. Once you have figured out which end is the head (the end on which a mouth opens to show teeth that bite you), you realize that all the rest of the head except for the mouth has an uninterrupted covering of fur. To us humans, accustomed to vertebrates sporting eyes and other hallmarks of a face, the expressionless sausage end that constitutes Spalax's head is disconcerting. But that strange appearance makes perfect sense functionally because blind mole rats spend most of their lives in underground tunnels, where eyes would be of even less use to them than a tail would be to us. If some Teenage Mutant Ninja humans should in the future evolve into a population burrowing underground, they would also be best off with reduced eyes, ears, and nose, a bulldozer-like head, and arms resembling shovels.

The blind mole rats use their sharp, protruding incisor teeth to excavate tunnels and to bite off edible plant parts which they store in pantry chambers. Instead of shoveling dirt outside the burrow with their small legs, they push it out with their broad, flat heads, like animated bulldozers. Although the animals are solitary for most of the year, they communicate by pounding their heads on the ceiling of the burrow, tapping out messages in a complex, Morse code-like pattern and thumping duets back and forth with mole rats in other burrows. "Singers" do not hear the other voices in the duets with their ears. Instead they sense the thumps' seismic vibrations by touch, much as Californians feel rather than hear earthquakes. If you pound on the roof of a Spalax tunnel, the occupant may thump back in response. During the breeding season, rather than seeking a mate above ground like most other mammals, the male mole rat digs around in search of a female's tunnel. If he is lucky enough to break into the tunnel of a receptive female instead of an angry rival male, mating may ensue, after which the male leaves and the female rears the pups unassisted.

Blind mole rats are not entirely eyeless, although they certainly look that way from the outside. Tiny eyes lie concealed beneath the fur and skin of their heads. Spalax is extreme in its eye reduction but not unique. Hundreds of other species of subterranean mammals--rodents like North America's pocket gopher, moles, and even a marsupial--have much smaller eyes than aboveground mammals. Those small-eyed burrowing mammals belong to three orders, 11 families, and 50 genera, proving that reduction of eye size underground has evolved independently many times.

The small eyes of most of those mammals protrude visibly through the skin, just like our own, but a few species besides Spalax have buried them under the skin--notably, Australia's marsupial "mole," a distant relative of kangaroos, whose resemblance to the moles of the rest of the world constitutes a famous example of convergent evolution. Subterranean mammals are not the only mammals with limited use for eyes: echolocating nocturnal bats, aquatic shrews, and blind river dolphins in the muddy Amazon River have small eyes as well. Subterranean, cave-dwelling, or muddy-water species belonging to many other groups of animals, including reptiles, amphibians, fish, and insects, also have diminished or even entirely vanished eyes.

Dissection shows Spalax's eyes to be literally the size of pinheads--0.7 millimeter (one thirty-sixth of an inch) in diameter, 99 percent smaller than those of an aboveground rodent of the same body size. But that tiny size is the least of the problems limiting the eyes' usefulness for imaging the world. Besides being completely covered by skin and fur, each eye is embedded in a thick gland; its lens is shriveled and shot through with blood vessels; and it has no pupil, nor eye muscles to move or focus it. Not surprisingly, if you flash a light at Spalax, it exhibits no reaction whatsoever. In addition, electrophysiological recordings of its brain show no response to a flash of light, as they would in you or me or a rat--and as they do if you thump at Spalax (because the mole rat can sense vibrations).

Nevertheless, Spalax's eyes have not evolved completely out of existence. The eye has a tiny retina (the light-sensitive lining of the eye). That retina has cones (the eye structures we use for color vision in bright light) containing a light-sensitive protein, called opsin, which is very similar to the light-sensitive proteins of human and mouse eyes--though Spalax has only 4 percent as much opsin as a mouse. Its retina is connected through 823 ganglion cells--99 percent fewer than in a similar-size hamster or rat--to a tiny optic nerve with less than 1,000 nerve fibers. What, if anything, is that tiny eye doing, since the animal shows no reaction to light flashes?

It's easy for us humans to forget that even our own eyes carry out two distinct functions. The obvious one is to form precise images of our changing environment by means of a lens, pupil, eye muscles, and point-to-point connections between the retina and the brain's visual cortex (the part of the brain that processes visual images). The other function is to perceive day-night cycles in overall light level and thereby to control our photoperiodic responses--our daily and seasonal cycles of sleep, activity, reproduction, and other physiological functions. Those physiological cycles depend on the hormone called melatonin. The cycles become disrupted when we fly across time zones, and are gradually reset when our eyes detect the shift in light cycle of the new time zone and tell our body to shift our cycle of melatonin release. That's why many travelers today try to accelerate the resetting by taking melatonin pills around the time of travel.

SPALAX DOES EXHIBIT PHOTOPERIODIC responses, even though it fails to move in response to light flashes. It has 24-hour activity cycles. If you simulate a plane trip by shifting the 24-hour lighting schedule in a laboratory, a caged Spalax shifts its activity cycle appropriately. If you simulate winter by giving it a 24-hour schedule with a longer dark period and shorter light period, it increases its physiological resistance to cold. Those physiological responses disappear in Spalax from which the tiny eyes have been surgically removed. Evidently Spalax's eyes, which do contain melatonin, are still being used to perceive changes in diffuse light levels and to control photoperiodic responses. On reflection, that's not astonishing: some light can still get through fur and skin. While Spalax's buried eyes cannot form a focused image, they suffice to recognize the difference between light and dark whenever Spalax sticks its head out of its tunnel to bulldoze soil to the outside.

How has such drastic yet precise reduction of Spalax's eyes been accomplished? The answer came from Eibi Nevo's collaborations with French scientists Howard Cooper and Marc Herbin, who studied the microanatomy of Spalax's eyes and brain, and his work with German scientists Reinhold Necker and Gerd Rehkamper, who studied brain electrophysiology.

In the course of Spalax evolution, the original eye inherited from sighted rodent ancestors was not dismantled by a wrecker's ball nor shrunk indiscriminately but was reconfigured very selectively. The eye and brain structures that sighted animals use for forming images, such as the retina and the dorsal lateral geniculate nucleus, are smaller by 90 to 99 percent in Spalax compared with those of sighted rodents of the same body size. But the brain structure involved in detecting diffuse light levels and controlling the photoperiodic response--the suprachiasmatic nucleus--is as large in Spalax as in sighted rodents. The part of the brain cortex that in humans or sighted rats would be devoted to vision and image formation, the occipital cortex, is devoted to somatosensory processing in Spalax--that is, to the body's sense of touch. Spalax's somatosensory cortex has thereby expanded to be almost twice as large as that of a similar-size sighted rat.

Cooper, Herbin, and Nevo did some calculations to understand why natural selection led to such a selective dismantling of the Spalax eye. (Think of the selective process--or Mother Nature--as a building owner who calculates property taxes, janitorial costs, and repair costs for a parking garage and an underoccupied multistory office building and decides to leave the garage and the building's ground floor but to raze the building's upper stories.) First, the collaborating scientists estimated what fraction of its whole energy budget a sighted rat otherwise similar to Spalax would lavish on its visual system. While the brain accounts for only about 1.3 percent of Spalax's body weight, it uses an astonishing 20 percent of the animal's oxygen and glucose, because the brain is one of the most expensive tissues of the body to operate. The visual cortex occupies one-tenth of the brain of a sighted rat, so, all other things being equal, it should consume one-tenth of 20 percent, or 2 percent, of the whole animal's metabolism. (In reality, all other things are not equal: among the parts of the brain, the visual cortex has nearly the highest oxygen consumption; the retina is even more expensive, using 100 times as much oxygen per gram of tissue as the rest of the body on average.) Sighted animals devote most of their visual system to image formation, needing very little for photoperiodic responses. Hence by getting rid of most of the image-forming part of the visual system, Spalax saves over 2 percent of its whole energy budget.

At first, 2 percent doesn't impress us as a big savings. A millionaire might not get excited if his financial adviser showed him ways to save 2 percent of his annual expenditures and increase his already enormous discretionary income. But a clerk with an annual salary of $30,000 and almost no discretionary income might well be thrilled: a 2 percent savings would permit him to repair a car otherwise sitting unused. Among animals, Spalax's oxygen consumption and energy budget re4semble those of a clerk, not a millionaire. Oxygen levels in its tunnels are low because of very poor air circulation, so it has to economize on what little oxygen it can get. It already has to spend an inordinate amount of its energy budget just getting around, because tunneling a certain distance underground costs hundreds or thousands of times more energy than does walking the same distance above ground. Spalax doesn't dare burn a drop of energy above the minimum lest it overheat in its tunnels, where there is no wind to cool its body and the atmosphere is too damp for sweat to evaporate efficiently. Hence Spalax and other subterranean mammals keep their basal metabolic rate--the minimum metabolic rate that they incur just lying still and doing nothing--considerably below that of aboveground mammals, the profligate millionaires of the rodent world. To be able to save 2 percent every day from a spartan energy budget is a fantastic windfall for Spalax: energy to convert into milk for rearing pups, into searching for food, or into tunnel construction to locate a mate.

Thus, energy savings are one potent reason that the reduction of Spalax's visual system produced a huge advantage through natural selection. But there is another reason as well: Spalax has better uses for its brain space than to waste it on useless image formation underground. Consider what other senses it really needs for a life spent in unlit tunnels: it needs an exquisite sense of touch over its head and body so that it can feel its way through tunnels and dig headfirst in the darkness. It also needs that sense of touch to communicate with mole rats in other burrows by detecting their thumps. It needs a good sense of hearing so that it can respond appropriately when another Spalax tunnels into its burrow calling out a threat or a courtship song. It needs a keen sense of smell to detect edible plants that it cannot see and to evaluate other mole rats by the odor of their urine. And it also has a remarkable system of magnetic compass orientation, permitting it to navigate without wrong turns through its labyrinth of tunnels, whether patrolling for intruders or fleeing from snakes.

All these sensory modalities require processing space in the Spalax brain, just as vision takes up 10 percent of the brain in sighted mammals. These multiple modalities place a big burden on underground rodents, who need several times as much brain space as sighted rodents do. Getting rid of the visual cortex and turning its space over to the sense of touch is one way Spalax copes with the problem of how to navigate blind without its expensive brain becoming even larger.

IN SHORT, WHEN SPALAX ENTERED A DARK world, its visual system didn't just rust away as mutations slowly accumulated, dismantling the system unopposed by natural selection. Instead natural selection favored mutants that quickly dumped their visual system over a cliff, because such mutants saved energy and space. To be more precise, natural selection favored mutants that jettisoned the visual system piece by piece, depending on how useful each part was.

This message that Eibi Nevo extracted from mole rats has a much broader significance. It furnishes a model that may apply in other species, resolving Darwin's puzzle about how natural selection destroys as well as creates. Let's take just one specific example out of millions of potential examples: Why did our ancestors lose their tails? As I noted at the outset, it might seem at first as if a tail would do us no real harm, even if it rarely did any good either.

But the way to think of your hypothetical tails, all you tailless Discover readers, is as a lost-opportunity cost. Picture the tail of a monkey, scaled up to your size. It would weigh quite a few pounds. At minimum, those additional pounds could increase your risk of having a heart attack; at worst, they could get caught in the lawn mower. Then ask yourself about the sacrifices that you would have to make elsewhere in your body to free up protoplasm for a tail without adding to your body weight or volume. Fewer pounds of brain? Then you would revert to being a big speechless monkey, foraging in the jungle for insects and eating them raw. Fewer pounds of muscle? Then you would be a wimp, easy prey for lions. Fewer pounds of kidney? Then please tell me how you propose to make enough urine. No, tails are fortunately gone forever, as far as we humans are concerned. Our ancestors didn't use them, they did lose them, and we profited.