It may be costly or even impossible for females to sample many different males before settling on a mate. For example, a female pied flycatcher visits on average only 3.8 males before choosing one. When females compete vigorously among themselves for a limited number of especially desirable males, their options may be even more restricted. As a result, females can hardly be expected to make a very informed choice; or, at least, their "choice" of a mate may be largely a matter of settling for whatever they can get. However we look at it, depending on how many additional males they encounter after they are paired up, some females may be likely to discover a male who is more desirable than the one with whom they find themselves. Under these conditions, "till death do us part" does not make a whole lot of sense. More likely is a strategy of "having your cake and eating it, too."
Think of it as a kind of one-way ratchet, whereby females, after accept ing an initial mating, will mate again, but only if, by doing so, they are "ratcheting up," improving the genetic situation of their offspring. The tac tic would be to mate with a seemingly good male-one who meets the minimum criteria of being of the right species, the right sex, and basically adequate-then remain available to mate with a better one, if he shows up. In a species of salamander, the European smooth newts, females pick up the sperm packets of males with particularly large head crests. In one experiment, females were exposed to males varying in the size of their crests, separated by 20 days. The females mated the first time, then had the choice of remating a second time or continuing to lay eggs fertilized by the sperm of their first mate. In this situation, most females only remated if the male to whom they were exposed the second time had a larger crest than the one whose sperm they had initially accepted. (Crestfallen, in such cases, is a severe condition indeed.)
—David P. Barash, The Myth of Monogamy, p. 67