The Greatest Show on Earth

by Richard Dawkins

  What happens under domestication is that animals are artificially shielded from many of the risks that shorten the lives of wild animals. A pedigree dairy cow may yield prodigious quantities of milk, but its pendulously cumbersome udder would seriously impede it in any attempt to outrun a lion. Thoroughbred horses are superb runners and jumpers, but their legs are vulnerable to injury during races, especially over jumps, which suggests that artificial selection has pushed them into a zone that natural selection would not have tolerated. Moreover, Thoroughbreds thrive only on a rich diet supplied by humans. Whereas Britain’s native ponies, for example, flourish on pasture, racehorses don’t prosper unless they are fed a much richer diet of grains and supplements – which they would not find in the wild. Again, I’ll return to such matters in the arms race chapter.

  DOGS AGAIN

  Having finally reached the topic of natural selection, we can turn back to the example of dogs for some other important lessons. I said that they are domesticated wolves, but I need to qualify this in the light of a fascinating theory of the evolution of the dog, which has again been most clearly articulated by Raymond Coppinger. The idea is that the evolution of the dog was not just a matter of artificial selection. It was at least as much a case of wolves adapting to the ways of man by natural selection. Much of the initial domestication of the dog was self- domestication, mediated by natural, not artificial, selection. Long before we got our hands on the chisels in the artificial selection toolbox, natural selection had already sculpted wolves into self-domesticated ‘village dogs’ without any human intervention. Only later did humans adopt these village dogs and transmogrify them, separately and comprehensively, into the rainbow spectrum of breeds that today grace (if grace is the word) Crufts and similar pageants of canine achievement and beauty (if beauty is the word).

  Coppinger points out that when domestic animals break free and go feral for many generations, they usually revert to something close to their wild ancestor. We might expect feral dogs, therefore, to become rather wolf-like. But this doesn’t happen. Instead, dogs left to go feral seem to become the ubiquitous ‘village dogs’ – ‘pye-dogs’ – that hang around human settlements all over the third world. This encourages Coppinger’s belief that the dogs on which human breeders finally went to work were wolves no longer. They had already changed themselves into dogs: village dogs, pye-dogs, perhaps dingos.

  Real wolves are pack hunters. Village dogs are scavengers that frequent middens and rubbish dumps. Wolves scavenge too, but they are not temperamentally suited to scavenging human rubbish because of their long ‘flight distance’. If you see an animal feeding, you can measure its flight distance by seeing how close it will let you approach before fleeing. For any given species in any given situation, there will be an optimal flight distance, somewhere between too risky or foolhardy at the short end, and too flighty or risk-averse at the long end. Individuals that take off too late when danger threatens are more likely to be killed by that very danger. Less obviously, there is such a thing as taking off too soon. Individuals that are too flighty never get a square meal, because they run away at the first hint of danger on the horizon. It is easy for us to overlook the dangers of being too risk-averse. We are puzzled when we see zebras or antelopes calmly grazing in full view of lions, keeping no more than a wary eye on them. We are puzzled, because our own risk aversion (or that of our safari guide) keeps us firmly inside the Land Rover even though we have no reason to think there is a lion within miles. This is because we have nothing to set against our fear. We are going to get our square meals back at the safari lodge. Our wild ancestors would have had much more sympathy with the risk-taking zebras. Like the zebras, they had to balance the risk of being eaten against the risk of not eating. Sure, the lion might attack; but, depending on the size of your troop, the odds were that it would catch another member of it rather than you. And if you never ventured on to the feeding grounds, or down to the water hole, you’d die anyway, of hunger or thirst. It is the same lesson of economic trade-offs that we have already met, twice.*

  The bottom line of that digression is that the wild wolf, like any other animal, will have an optimal flight distance, nicely poised – and potentially flexible – between too bold and too flighty. Natural selection will work on the flight distance, moving it one way or the other along the continuum if conditions change over evolutionary time. If a plenteous new food source in the form of village rubbish dumps enters the world of wolves, that is going to shift the optimum point towards the shorter end of the flight distance continuum, in the direction of reluctance to flee when enjoying this new bounty.

  We can imagine wild wolves scavenging on a rubbish tip on the edge of a village. Most of them, fearful of men throwing stones and spears, have a very long flight distance. They sprint for the safety of the forest as soon as a human appears in the distance. But a few individuals, by genetic chance, happen to have a slightly shorter flight distance than the average. Their readiness to take slight risks – they are brave, shall we say, but not foolhardy – gains them more food than their more risk-averse rivals. As the generations go by, natural selection favours a shorter and shorter flight distance, until just before it reaches the point where the wolves really are endangered by stone-throwing humans. The optimum flight distance has shifted because of the newly available food source.

  Something like this evolutionary shortening of the flight distance was, in Coppinger’s view, the first step in the domestication of the dog, and it was achieved by natural selection, not artificial selection. Decreasing flight distance is a behavioural measure of what might be called increasing tameness. At this stage in the process, humans were not deliberately choosing the tamest individuals for breeding. At this early stage, the only interactions between humans and these incipient dogs were hostile. If wolves were becoming domesticated it was by self-domestication, not deliberate domestication by people. Deliberate domestication came later.

  We can get an idea of how tameness, or anything else, can be sculpted – naturally or artificially – by looking at a fascinating experiment of modern times, on the domestication of Russian silver foxes for use in the fur trade. It is doubly interesting because of the lessons it teaches us, over and above what Darwin knew, about the domestication process, about the ‘side-effects’ of selective breeding, and about the resemblance, which Darwin well understood, between artificial and natural selection.

  The silver fox is just a colour variant, valued for its beautiful fur, of the familiar red fox, Vulpes vulpes. The Russian geneticist Dimitri Belyaev was employed to run a fox fur farm in the 1950s. He was later sacked because his scientific genetics conflicted with the anti-scientific ideology of Lysenko, the charlatan biologist who managed to capture the ear of Stalin and so take over, and largely ruin, all of Soviet genetics and agriculture for some twenty years. Belyaev retained his love of foxes, and of true Lysenko-free genetics, and he was later able to resume his studies of both, as director of an Institute of Genetics in Siberia.

  Wild foxes are tricky to handle, and Belyaev set out deliberately to breed for tameness. Like any other animal or plant breeder of his time, his method was to exploit natural variation (no genetic engineering in those days) and choose, for breeding, those males and females that came closest to the ideal he was seeking. In selecting for tameness, Belyaev could have chosen, for breeding, those dogs and bitches that most appealed to him, or looked at him with the cutest facial expressions. That might well have had the desired effect on the tameness of future generations. More systematically than that, however, he used a measure that was pretty close to the ‘flight distance’ I just mentioned in connection with wild wolves, but adapted for cubs. Belyaev and his colleagues (and successors, for the experimental program continued after his death) subjected fox cubs to standardized tests in which an experimenter would offer a cub food by hand, while trying to stroke or fondle it. The cubs were classified into three classes. Class III cubs were those that fled from or bit the person. Cla
ss II cubs would allow themselves to be handled, but showed no positive responsiveness to the experimenters. Class I cubs, the tamest of all, positively approached the handlers, wagging their tails and whining. When the cubs grew up, the experimenters systematically bred only from this tamest class.

  After a mere six generations of this selective breeding for tameness, the foxes had changed so much that the experimenters felt obliged to name a new category, the ‘domesticated elite’ class, which were ‘eager to establish human contact, whimpering to attract attention and sniffing and licking experimenters like dogs’. At the beginning of the experiment, none of the foxes were in the elite class. After ten generations of breeding for tameness, 18 per cent were ‘elite’; after twenty generations, 35 per cent; and after thirty to thirty-five generations, ‘domesticated elite’ individuals constituted between 70 and 80 per cent of the experimental population.

  Such results are perhaps not too surprising, except for the astonishing magnitude and speed of the effect. Thirty-five generations would pass unnoticed on the geological timescale. Even more interesting, however, were the unexpected side-effects of the selective breeding for tameness. These were truly fascinating and genuinely unforeseen. Darwin, the dog-lover, would have been entranced. The tame foxes not only behaved like domestic dogs, they looked like them. They lost their foxy pelage and became piebald black and white, like Welsh collies. Their foxy prick ears were replaced by doggy floppy ears. Their tails turned up at the end like a dog’s, rather than down like a fox’s brush. The females came on heat every six months like a bitch, instead of every year like a vixen. According to Belyaev, they even sounded like dogs.

  Belayev with his foxes, as they turn tame – and doglike

  These dog-like features were side-effects. Belyaev and his team did not deliberately breed for them, only for tameness. Those other dog-like characteristics seemingly rode on the evolutionary coattails of the genes for tameness. To geneticists, this is not surprising. They recognize a widespread phenomenon called ‘pleiotropy’, whereby genes have more than one effect, seemingly unconnected. The stress is on the word ‘seemingly’. Embryonic development is a complicated business. As we learn more about the details, that ‘seemingly unconnected’ turns into ‘connected by a route that we now understand, but didn’t before’. Presumably genes for floppy ears and piebald coats are pleiotropically linked to genes for tameness, in foxes as well as in dogs. This illustrates a generally important point about evolution. When you notice a characteristic of an animal and ask what its Darwinian survival value is, you may be asking the wrong question. It could be that the characteristic you have picked out is not the one that matters. It may have ‘come along for the ride’, dragged along in evolution by some other characteristic to which it is pleiotropically linked.

  The evolution of the dog, then, if Coppinger is right, was not just a matter of artificial selection, but a complicated mixture of natural selection (which predominated in the early stages of domestication) and artificial selection (which came to the fore more recently). The transition would have been seamless, which again goes to emphasize the similarity – as Darwin recognized – between artificial and natural selection.

  FLOWERS AGAIN

  Let’s now, in the third of our warm-up forays into natural selection, move on to flowers and pollinators and see something of the power of natural selection to drive evolution. Pollination biology furnishes us with some pretty amazing facts, and the high point of wondrousness is reached in the orchids. No wonder Darwin was so keen on them; no wonder he wrote the book I have already mentioned, The Various Contrivances by which Orchids are Fertilised by Insects. Some orchids, such as the ‘magic bullet’ Madagascar ones we met earlier, give nectar, but others have found a way to bypass the costs of feeding pollinators, by tricking them instead. There are orchids that resemble female bees (or wasps or flies) well enough to fool males into attempting to copulate with them. To the extent that such mimics resemble females of one particular insect species, to that extent will males of those species serve as magic bullets, going from flower to flower of just the one orchid species. Even if the orchid resembles ‘any old bee’ rather than one species of bee, the bees that it fools will still be ‘fairly magic’ bullets. If you or I were to look closely at a fly orchid or a bee orchid (see colour page 5), we would be able to tell that it was not a real insect; but we would be fooled at a casual glance out of the corner of our eye. And even looking at it head-on, I would say the bee orchid in the picture (h) is pretty clearly more of a bumble-bee orchid than a honey-bee orchid. Insects have compound eyes, which are not so acute as our camera eyes, and the shapes and colours of insect-mimicking orchids, reinforced by seductive scents that mimic those of female insects, are more than capable of tricking males. By the way, it is quite probable that the mimicry is enhanced when seen in the ultraviolet range, from which we are cut off.

  The so-called spider orchid, Brassia (colour page 5 (k)), achieves pollination by a different kind of deception. The females of various species of solitary wasp (‘solitary’ because they don’t live socially in large nests like the familiar autumn pests, called yellowjackets by Americans) capture spiders, sting them to paralyse them, and lay their eggs on them as a living food supply for their larvae. Spider orchids resemble spiders sufficiently to fool female wasps into attempting to sting them. In the process they pick up pollinia – masses of pollen grains produced by the orchids. When they move on to try to sting another spider orchid, the pollinia are transferred. By the way, I can’t resist adding the exactly backwards case of the spider Epicadus heterogaster, which mimics an orchid. Insects come to the ‘flower’ in search of nectar, and are promptly eaten by it.

  Some of the most astonishing orchids that practise this seduction trick are to be found in Western Australia. Various species in the genus Drakaea are known as hammer orchids. Each species has a special relationship with a particular species of wasp of the type called thynnids. Part of the flower bears a crude resemblance to an insect, duping the male thynnid wasp into attempting to mate with it. So far in my description, Drakaea is not dramatically different from other insect-mimicking orchids. But Drakaea has a remarkable extra trick up its sleeve: the fake ‘wasp’ is borne on the end of a hinged ‘arm’, with a flexible ‘elbow’. You can clearly see the hinge in the picture (colour page 5 (g)). The fluttering movement of the wasp gripping the dummy wasp causes the ‘elbow’ to bend, and the wasp is dashed repeatedly back and forth like a hammer against the other side of the flower – let’s call it the anvil – where it keeps its sexual parts. The pollinia are dislodged and stick to the wasp, who eventually extricates himself and flies off, sadder but apparently no wiser: he goes on to repeat the performance on another hammer orchid, where he and the pollinia he bears are duly dashed against the anvil, so that his cargo finds its destined refuge on the female organs of the flower. I showed a film of this astounding performance in one of my Royal Institution Christmas Lectures for Children, and it can be seen in the recording of the lecture called ‘The Ultraviolet Garden’.

  In the same lecture I discussed the ‘bucket orchids’ of South America, which achieve pollination in an equally remarkable but rather different way. They too have specialized pollinators, not wasps but small bees, of the group called Euglossine. Again, these orchids provide no nectar. But the orchids don’t fool the bees into mating with them either. Instead, they provide a vital piece of assistance for male bees, without which the bees would be unable to attract real females.

  These little bees, which live only in South America, have a strange habit. They go to elaborate lengths to collect fragrant, or anyway smelly, substances, which they store in special containers attached to their enlarged hind legs. In different species these smelly substances can come from flowers, from dead wood, or even from faeces. It seems that they use the gathered perfumes to attract, or otherwise court, females. Many insects use particular scents to appeal to the opposite sex, and most of them manufacture the perfumes i
n special glands. Female silk moths, for example, attract males from an astonishingly long distance by releasing a unique scent, which they manufacture and which males detect – in minute traces from literally miles away – with their antennae. In the case of Euglossine bees, it is the males that use scent. And, unlike the female moths, they don’t synthesize their own perfume but use the smelly ingredients that they have collected, not as pure substances but as carefully concocted blends which they put together like expert perfumiers. Each species mixes a characteristic cocktail of substances gathered from various sources. And there are some species of Euglossine bee that positively need, for manufacturing their characteristic species scent, substances that are supplied only by flowers of particular species of the orchid genus Coryanthes – bucket orchids. The common name of Euglossine bees is ‘orchid bees’.

  What an intricate picture of mutual dependence. The orchids need the Euglossine bees, for the usual ‘magic bullet’ reasons. And the bees need the orchids, for the rather weirder reason that they can’t attract female bees without substances that are either impossible or at least too hard to find except through the good offices of bucket orchids. But the way in which pollination is achieved is even weirder still, and it superficially makes the bee look more like a victim than a cooperating partner.

  A male Euglossine bee is attracted to the orchids by the smell of the substances that he needs in order to manufacture his sexual perfumes. He alights on the rim of the bucket and starts to scrape the waxy perfume into the special scent pockets in his legs. But the rim of the bucket is slippery underfoot – and there’s a reason for this. The bee falls into the bucket, which is filled with liquid, in which he swims. He cannot climb up the slippery sides of the bucket. There is only one escape route, and this is a special bee-sized hole in the side of the bucket (not visible in the picture that appears on colour page 4). He is guided by ‘stepping stones’ to the hole and starts to crawl through it. It’s a tight fit, and it becomes even tighter as the ‘jaws’ (these you can see in the picture: they look like the chuck of a lathe or electric drill) contract and trap him. While he is held in their grip, they glue two pollinia to his back. The glue takes a while to set, after which the jaws again relax and release the bee, who flies off, complete with pollinia on his back. Still in search of the precious ingredients for his perfumery, the bee lands on another bucket orchid and the process repeats itself. This time, however, as the bee struggles through the hole in the bucket, the pollinia are scraped off, and they fertilize the stigma of this second orchid.