Sexual Selection

Living longer does not count; leaving more offspring does

One thing first: what evolution scores is never lifespan, it is how many offspring you leave that go on to breed. A variant that costs you two years but doubles your descendants spreads generation by generation — even while it makes life harder every single day. So "it cannot run fast" is not a counterexample by itself: if what is lost there is earned back elsewhere, the account balances.

The long-tailed widowbird — a black bird of the African grasslands — lays that account out very plainly. A male's tail feathers can be half a metre long, and flight is a laboured bobbing affair; the female's tail is as short as a sparrow's.

In 1982 Andersson caught thirty-odd males that already held territories and split them at random into four groups. One group had its tail feathers cut short. One had those cut feathers glued onto their own tails, making them longer. The remaining two were controls — one cut and reglued back to its original length (to rule out the cutting and gluing itself), one merely caught and released. He put them back, and some time later counted how many nests in each male's territory held eggs or chicks. The lengthened group had roughly twice as many as the controls, the shortened group about half, and the two controls came out much the same.

One detail matters more than the result itself: males with altered tails held their territories exactly as before — the ones who used to win still won, the ones who used to lose still lost. What changed was whose ground the females settled on.

Here a very easy mistake needs blocking: it is not that a long tail evolved "in order to attract females". Nobody remodels themselves towards a goal. What actually happens is this — in every generation males differ in tail length, the slightly longer ones leave more offspring on average, and so the average tail of the next generation shifts up a fraction. Repeat for a few hundred thousand generations.

Who does the choosing is not a fixed role — it follows who invests more

In that experiment the females did the choosing. Why is it usually that way round? The difference starts at the level of cells: eggs are large, expensive, and few; sperm are small, cheap, and produced by the tens of millions. The investment that follows is usually lopsided the same way — one side carries the eggs or the young, and the same side often does the incubating and feeding.

The consequence is concrete: the side that invests more loses a whole breeding season by choosing badly, so it is choosy; the side that invests less is short not of mates but of opportunities, so it competes. The two kits you see in nature — the showy one (trains, calls) and the fighting one (horns, canines, sheer size) — fall broadly on either side of that line.

The point is that this is not an innate script for the sexes. It tracks investment. In seahorses the male takes the eggs into a pouch and carries them; in jacanas and phalaropes the females are the brighter ones, holding territory and mating with several males while the males incubate and raise the chicks. Whichever side invests more becomes the choosy one. Flip the cost and the whole set of roles flips with it.

A word about how strong the evidence is. The founding experiment for this framework was Bateman's 1948 work on fruit flies; when it was repeated in 2012, his way of counting offspring turned out to carry a systematic bias, and done his way the original conclusion does not come out. Redone with a corrected method, selection still read as stronger on the male side. This one is still going back and forth. The asymmetry itself rests on a great deal of independent evidence — but using "males are all promiscuous, females all choosy" as a universal law is simply wrong.

Why ornaments get absurdly expensive: two explanations, still not separated

Females choosing explains why the tail gets longer. It does not explain why it gets ridiculous. There are two quite different mechanisms, and each can push it that far on its own.

The first is called runaway. Its crux is that the preference is heritable too. Suppose a female's tendency to settle more often near long-tailed males has a genetic component, and so does male tail length — then a female who picks a long-tailed male has sons with longer tails and daughters carrying the same preference. The two things are tied together in the same brood, and each generation pushes both at once. The loop accelerates itself, and can drive an ornament far past anything useful, until the cost of staying alive pins it down. In the 1980s the route was worked through mathematically: in the models it really does run, and at no point does the long tail need to be any good for survival.

The second is costly signalling, and its crux is usually stated backwards: a signal is credible not because it is expensive, but because it is more expensive for poor-quality individuals. A weak male would also like to hang out a big tail; the trouble is that the same inch of feather takes far more out of him. Something everyone can afford carries no information; it is precisely because some cannot afford it that affording it says anything. In 1990 that intuition was turned into a rigorous model: only when the cost rises as individual quality falls does this self-injuring form of advertising stay stable.

The trouble is that in the field the two are hard to pull apart — and the most famous case, the peafowl, argues with itself. In 1994 Petrie assigned peahens to peacocks at random (bypassing the females' own choosing), reared the offspring under identical conditions, and found that fathers with more eyespots had offspring that grew faster and survived better. Yet in 2008 another group watched a feral peafowl population in Japan for seven straight years and found peahens showing no preference at all for males with more eyespots.

This is not settled. Most people's view now is that the two mechanisms are not either-or; they often run at the same time.

In one mating, the two sides have different optimal outcomes

An assumption should be dismantled here: so far this reads as if the sexes were working together on one job, one choosing and one being chosen. Often they are not.

Along with sperm, male fruit-fly seminal fluid carries a set of proteins. Inside the female they speed up her egg-laying, stop her accepting other males for a while, and disable sperm left by a previous male. For that male, each of those converts directly into his own offspring.

A set of experiments in 1995 isolated the effect of those secretions: males that were genetically unable to make sperm but still secreted the proteins were used for mating, and exposure to the secretions alone was enough to make females die younger — the more exposure, the sooner.

Do not read this as the male "wanting" to harm his mate. He wants nothing; the shortened life is most likely a side effect. But as long as those proteins win his sperm the competition, they get passed on, even with the whole cost landing on the other party. Selection counts descendants; it does not count who got hurt.

The other side is being pushed at the same time: females more tolerant of those secretions leave more offspring, so both sides escalate round after round. Water striders turn this tug-of-war into something you can see — males grow grasping hooks on their forelegs, females grow matching spines and shields on the abdomen, and comparing across species shows that the more extreme one structure is, the more extreme the other. The net result is often that nobody gains, and both sides have paid.