This is the first of a number of posts on “classic papers” in our new series called Journal Pub .
Our first topic is mating systems, and I have the pleasure of summarizing and commenting on Angus Bateman’s study of sexual selection in Drosophila. In 1948, although ¾ of a century had passed since Darwin published Sexual selection and the descent of man, Bateman (1948) remarked that the evidence that sexual selection explained sexual differences remained circumstantial, and that there was considerable debate concerning the importance of mate competition in producing secondary sex characters. For example, Huxley (1938) argued that monogamous birds with striking secondary sexual differences seemed to display mostly after pair formation (and therefore presumably not in the context of contests at all).
What we now know about the many possible episodes for sexual selection (including for example, the potential for postcopulatory sperm competition and female choice) colours our impression of Huxley’s objections, but at the time Bateman’s empirical approach was probably the only sensible response: could he demonstrate that males and females differ in the potential to gain fitness through mating?
He conducted a classic experiment in which he housed an equal number (either 3 or 5 of each sex) of virgin male and female Drosophila melanogaster together for three or four days, and collected the offspring produced within the fly vials during this time. Because each of his flies carried a unique dominant marker, he was able to unambiguously assign all offspring to their parents, and therefore to retroactively work out the reproductive successes of males and females in his experiment. He first noted that males had much higher variation in fitness than females did: there were more males who had no fitness at all, and a few males had dramatically high fitness. (Note that he did a lot of technical work to make sure that his observations of differences in variance across the sexes were real, and not due to errors in experimental execution or measurement.)
Furthermore, he showed that the relationship between mate number and fitness was much stronger in males than it was in females. The figure below (stolen from his paper) has been reproduced many times to illustrate this key result.
It’s worth noting that the differences between sexes was much stronger in his last pair of experimental blocks (Series 5 and 6, on the right) than in his first four blocks. The reasons for this discrepancy are not clear, but Bateman speculated that his first four blocks suffered from poor vigour; if weak males could not transfer enough sperm to fertilize all of a female’s eggs, several inseminations would be needed.
The key difference between the sexes therefore was that males can gain a lot from remating, but females usually much less so. If this difference in selection on mating was representative of the general situation in animals, that would explain (as Darwin had suggested) why males so often are the showier, more heavily armed sex: they have more to gain from contests over sex than females do.
In discussing this paper on Tuesday, several people mentioned the very recent (and controversial) publication of work that reanalysed (Snyder & Gowaty 2007) or replicated (Gowaty et al., 2012) Bateman’s classic experiments, criticizing many of the conclusions Bateman had drawn. None of us felt sufficiently well prepared to discuss these in any depth, but perhaps they will be interesting for further reading and discussion another day.