particular kind of bird of paradise, known as the King of Saxony.
These feathers, which are several times longer than the original owner 's body and stem from just above his eye, are like a car 's antenna sporting dozens of square blue pennants. Because they are molted once a year, do not grow until the bird of paradise is four years old, and are much in demand among local tribesmen, the plumes must be very hard for the bowerbird to acquire: Once
acquired they must be guarded against other jealous male bowerbirds anxious to steal them for their own bowers: So, in the words of Jared Diamond, a female bowerbird who finds a male that has decorated his bower with King of Saxony plumes knows 'that she has located a dominant male who is terrific at finding or stealing rare objects and defeating would-be thieves:'
So much for the bowerbird: What about the bird of paradise itself, the rightful owner of the plumes? The fact that he survived long enough to grow plumes, grew longer ones than any other male nearby, and kept them in good condition would be an equally reliable indicator of his genetic quality. But it reminds us of the thing that most puzzled Darwin and got the whole debate started: If the point of the plumes is to indicate his quality, might not the plumes themselves affect his quality? After all, every tribesman in New Guinea is out to get him, and every hawk will find him easier to spot: He may have indicated that he is good at surviving, but his chances of survival are now lower for having the plumes: They are a handicap. How can a system of females choosing males that are good at surviving encumber those males with handicaps to survival?
It is a good question with a paradoxical answer, for which we owe a debt of thanks to Amotz Zahavi, a mercurial Israeli scientist: He saw in 1975 that the more a peacock 's tail or a bird of paradise' s plumes handicapped the male, the more honest the signal was that he sent the female. She could be assured by the very fact of his survival that the long-tailed male in front of her had been through a trial and passed: He had survived despite being handicapped: The more costly the handicap, the better it was as a signal of his genetic quality; therefore, peacocks ' tails would evolve faster if they were handicaps than if they were not. This is the reverse of Fisher ' s prediction that peacocks ' tails should gradually cease evolving once they become severe handicaps:'°
It is an appealing—and familiar—thought: When a Maasai warrior killed a fierce beast to prove himself to a potential mate, he was running the risk of being killed but was also showing that he had the necessary courage to defend a herd of cattle. Zahavi 's
' handicap' was only a version of such initiation rituals, yet it was attacked from all sides, and the consensus was that he was wrong.
THE PEACOCK ' S TALE
:::
The most telling argument against it was that the sons would inherit the handicap as well as the good genes, so they would be encumbered to the same degree as they were endowed. They would be no better off than if they were unencumbered and unsexy.'
In recent years, however, Zahavi has been vindicated. Mathematical models proved that he might be right and his critics wrong.' His vindicators have added to his theory two subtleties that lend it special relevance to the Good-gene theory of sexual selection. The first is that handicaps might (perhaps must) not only affect survival and reflect quality but also do so in a graduated way; the weaker the male, the harder it would be to produce or maintain a tail of a given length. And indeed, experiments on swallows have shown that birds promoted above their station, by being given longer tail streamers than they grew naturally, could not the next time grow as long a tail as before; carrying the extra handicap had taken its toll: 39 The second is that the handicapping ornament might be designed so as best to reveal deficiency: After all, life would be a lot easier for swans if they were not white, as anybody who has tried swimming in a lake in a wedding dress would know.
Swans do not become white until they are a few years old and ready to breed; perhaps being whiter than white proves to a skeptical swan that its suitor can spare the time from feeding to clean his plumage.
The vindication of Zahavi played a critical role in reignit-ing the debate between Fishererians and Good-geners. Until that happened, Good-gene theories could work only if the ornaments they resulted in were not encumbrances to the males: Thus, a male might advertise the quality of his genes, but to do so at a high cost to himself would be counterproductive unless there were a sexy-son effect.
LOUSY MALES
The handicap theory now comes face-to-face with the central conundrum of sexual selection: This is the lek paradox: that peahens are constantly skimming off the cream of the genetic cream by
::: 150 :::
choosing only the very few best males to mate with, and as a result, within a very few generations, no variety is left to choose from: The good-gene assertion that mutations are likely to make ornaments and displays less effective provides a partial answer, but it is not a persuasive one: After all, it argues only for not choosing the worst rather than for choosing the best:
Only the Red Queen can solve our dilemma. What sexual selection theory seems to have concluded is that females are constantly running (by being so selective) but are staying in the same place (having no variety to select from). When we find that, we should be on the lookout for some ever-changing enemy, some arms-race rival. It is here that we meet Bill Hamilton again: We last encountered him when discussing the idea that sex itself is an essential part of the battle against disease. If the main purpose of sex is to grant your descendants immunity from parasites, then it follows directly that it makes sense to seek a mate with parasite-resistance genes: AIDS has reminded us all too forcibly of the value of choosing a healthy sexual partner, but similar logic applies to all diseases and parasites: In 1982, Hamilton and a colleague, Marlene Zuk (now at the University of California at Riverside), suggested that parasites might hold the key to the lek paradox and to gaudy colors and peacocks ' tails, for parasites and their hosts are continually changing their genetic locks and keys to outwit each other. The more common a particular strain of host is in one generation, the more common the strain of parasite is that can overcome its defenses in the next. And vice versa: Whatever strain of host is most resistant to the prevalent strain of parasite will itself be the prevalent strain of host in the next generation. Thus, the most disease resistant male might often turn out to be the descendant of the least resistant one in a previous generation: The lek paradox is thus solved at a stroke: By choosing the healthiest male in each generation, females will be picking a different set of genes each time and never run out of genetic variety to select from.'°
The Hamilton-Zuk parasite theory was bold enough, but
*
the two scientists did not stop there. They looked up the data for 1 09 species of bird and found that the most brightly colored
'
