Zahavi's handicap theory

We now turn to a second theory, in which the costliness of the male character is positively useful to the female. It is Zahavi's handicap theory. ('Handicap' is a character that reduces survival.) Suppose that some males have genes that confer higher fitness ('good genes') than do other males (who have 'bad genes'). According to Zahavi, a handicap may indicate that the male has 'good genes'. The argument runs like this.

Consider two females, one who mates at random, and one who mates preferentially with handicapped males.

• Indiscriminate females

If a female mates at random, her mates will have good and bad genes in the same proportions as the good and bad genes have in the whole population; if half the males in the population have good genes and half have bad, then 50% of her mates will have good genes and 50% bad. Now suppose that some of the males in the population possess a handicap - a character that reduces their survival.

• Choosy females

If only males with good genes can survive possessing a handicap, a female who mates preferentially with handicapped males will only mate with males with good genes. The choice will be favored by selection if the advantage through the superior genes outweighs the cost of the handicap: then the net quality of the choosy female's offspring will be higher than those of the randomly mating female.

The handicap thus acts as an indicator of genetic quality: but why does the indicator have to be costly?

The reason is that the cost guarantees that the indicator will be reliable. A male's genetic quality does not come written on him: it has to be inferred, and if females inferred it from an an inexpensive signal, there would be selection on males to cheat. If females preferentially mated with males who merely said "I have good genes" (or something analogous to saying this) and rejected those that said "I have poor genes", mutant males who said the former independently of their true genetic quality would be favored. Words (and their analogues) are cheap. But if the criterion favored by females is costly, as growing a long and ostentatious tail is, then selection will less automatically favor cheats. In particular, if the cost of growing a handicap is less for a truly high quality male than for a low quality male, handicaps will be grown only by high quality males and will be reliable signals for females to use.