Free The Panda’s Thumb by Stephen Jay Gould Page B

fifty females. Populations made predominantly of females should, by their more rapid rates of reproduction, win any evolutionary race with populations that maintain equality in numbers between the sexes.
    What appeared obvious is therefore rendered problematical and the question remains: why do most sexual species contain approximately equal numbers of males and females? The answer, according to most evolutionary biologists, lies in a recognition that Darwin’s theory of natural selection speaks only of struggle among individuals for reproductive success. It contains no statement about the good of populations, species, or ecosystems. The argument for ninety females and ten males was framed in terms of advantages for populations as a whole—the usual, congenial, and dead wrong, way in which most people think of evolution. If evolution worked for the good of populations as a whole, then sexual species would contain relatively few males.
    The observed equality of males and females, in the face of obvious advantages for female predominance if evolution worked upon groups, stands as one of our most elegant demonstrations that Darwin was right—natural selection works by the struggle of individuals to maximize their own reproductive success. The Darwinian argument was first framed by the great British mathematical biologist R.A. Fisher. Suppose, Fisher argued, that either sex began to predominate. Let us say, for example, that fewer males than females are born. Males now begin to leave more offspring than females since their opportunities for mating increase as they become rarer—that is, they impregnate more than one female on average. Thus, if any genetic factors influence the relative proportion of males born to a parent (and such factors do exist), then parents with a genetic inclination to produce males will gain a Darwinian advantage—they will produce more than an average number of grandchildren thanks to the superior reproductive success of their predominantly male offspring. Thus, genes that favor the production of males will spread and male births will rise in frequency. But, this advantage for males fades out as male births increase and it disappears entirely when males equal females in number. Since the same argument works in reverse to favor female births when females are rare, the sex ratio is driven by Darwinian processes to its equilibrium value of one to one.
    But how would a biologist go about testing Fisher’s theory of sex ratio? Ironically, the species that confirm its predictions are no great help beyond the initial observation. Once we frame the basic argument and determine that the species we know best have approximately equal numbers of males and females, what do we achieve by finding that the next thousand species are similarly ordered? Sure, it all fits, but we do not gain an equal amount of confidence each time we add a new species. Perhaps the one to one ratio exists for another reason?
    To test Fisher’s theory, we must look for exceptions. We must seek unusual situations in which the premises of Fisher’s theory are not met—situations that lead to a specific prediction about how sex ratio should depart from one to one. If change of premises leads to a definite and successful prediction of altered outcome, then we have an independent test that strongly boosts our confidence. This method is embodied in the old proverb that “the exception proves the rule,” although many people misunderstand the proverb because it embodies the less common meaning of “prove.” Prove comes from the Latin probare— to test or to try. Its usual, modern meaning refers to final and convincing demonstration and the motto would seem to say that exceptions establish indubitable validity. But in another sense, closer to its root, “prove” (as in “proving ground” or printer’s “proof”) is more like its cognate “probe”—a test or