Domestication

How did wild plants and animals become the crops and livestock we now depend on?

The usual picture has an early farmer choosing: penning the calmest animals, saving seed from the best plants, patiently shaping a wild species into a useful one. It is a satisfying story, and it credits someone with a plan. But it asks us to believe that people with no notion of heredity ran a breeding programme whose payoff arrived centuries after their deaths. What if almost none of it was chosen?

Start with wheat, because the mechanism there is unusually clean. A wild cereal must scatter its seed to reproduce, so its ear is built to shatter — the stalk holding the grain is brittle and breaks as the seed ripens. Now picture a gatherer cutting stands of it with a sickle and carrying the bundles home. Which grains arrive back at the settlement? Only the ones still attached. The shattering majority fell in the field.

Follow that one more step. Some of what is carried home gets sown next season — so, without anyone intending it, the harvest acted as a filter and the sowing made the filter heritable. Repeat annually and the non-shattering variant, a rare mutant that is useless in the wild because it cannot disperse, climbs toward fixation. The archaeobotanical record in Southwest Asia is consistent with exactly this: the shift to non-shattering ears took on the order of a thousand to two thousand years, far too slow for deliberate breeding and about right for an unconscious filter.

The other crop traits follow the same logic. Seed that germinates promptly beats dormant seed in a tilled field; large seeds push seedlings through deep-sown soil; grain that ripens all at once is the grain a single harvest collects. Nobody had to want any of this.

Now try animals, where the striking observation is different. Compare dogs, pigs, cattle, sheep, rabbits, horses — unrelated lineages, domesticated in different places for different purposes. They share a cluster of changes: shortened snouts, smaller teeth, floppy ears, curly tails, patchy white depigmentation, reduced brain size, prolonged juvenile behaviour, more frequent and less seasonal breeding. Why would selecting cattle for milk and dogs for hunting produce the same odd cosmetic traits? Nobody was selecting for a floppy ear.

That cluster — the domestication syndrome — suggests the traits are not separately selected but come along with something else. The cleanest test of that came from Dmitry Belyaev's fox experiment, begun in 1959 in Novosibirsk. Silver foxes were selected on one criterion only: how readily they approached a human hand. Nothing about ears, coats or skulls entered the decision. Within a few dozen generations the tame line showed piebald patches, floppy ears in kits, curlier tails, altered breeding timing and lowered stress-hormone levels — the syndrome, arriving as a package behind a single behavioural selection.

Why should that be? The leading explanation is a hypothesis and should be labelled as one: that mild reductions in neural crest cells — an embryonic population contributing to adrenal tissue, pigment cells, and the cartilage of face and ear — could lower fear responses and, as a side effect, alter pigment, ears and jaws together. The fox work itself has been criticised too: the founding animals came from Canadian fur farms already under some selection for handling, and how consistently the classic traits appeared has been questioned.

One last turn: who was selecting whom? For the animals that came to us first, the pressure was human proximity, not human intent — wolves and wildcats that tolerated scavenging around settlements ate better and left more offspring, so the population selected itself for tolerating us before anyone penned anything. And we changed in return: adult lactose tolerance spread in dairying populations within the last several thousand years, and extra copies of starch-digesting genes accompany a cereal diet in both humans and dogs. Domestication is not something one species did to others.

Two refinements tie the parts together. First, "unintentional" does not mean "no selection". The selection was real, strong and consistent; what was missing was foresight, which is the one ingredient natural selection has never needed either. Deliberate breeding did arrive, and produced the dramatic divergence of the last few centuries — but it worked on species already domesticated by the slower process.

Second, the syndrome is a genuine empirical pattern with a contested explanation. That unrelated species converge on it is not evidence of some force in human proximity; it is evidence that the same pressure was applied to all of them — reduced fear of humans — and that this trait appears developmentally entangled with others. The convergence lives in the pressure, not in shared ancestry.

Finally, resist reading domestication as a favour done to grateful species. Domesticated lineages have generally lost the ability to survive without us — maize cannot disperse its own seed at all — which is dependence, not reward.

• The three pathways — commensal, prey, and directed — and why dogs and horses arrived by different routes.
• Why so few large mammals were ever domesticated, out of the many that were available.
• What the neural crest hypothesis predicts that could actually falsify it.

Every term the collection defines is gathered in the glossary.