A test of sexual cannibalism models, using a sit-and-wait predator

Sexual cannibalism, the capture and consumption of one sex by the other resulting from sexual interactions, provides an intriguing set of conditional fitness payoffs. Usually these payoffs will only benefit the cannibal, but under certain circumstances advantages might accrue to the cannibalized individual as well. Here I test three models of sexual cannibalism with data from a sit-and-wait predator, the crab spider Misumena vatia , whose lifestyle differs from the species for which these models were generated. The models include both precopulatory mechanisms (no gametes exchanged) and postcopulatory ones (gametes exchanged) perceived to generate adult fitness payoffs and situations in which cannibalism is not adaptive in the adult stage (aggressive spillover). In M. vatia , females sometimes cannibalize males before mating can occur. Precopulatory cannibalism is unlikely to reward female M. vatia significantly because of the diminutive size of the males, but it could be part of a broader syndrome that provides the females with large food payoffs earlier in the life cycle. Although the frequency of cannibalism is low (3.8–7.6%), this level suffices to be an important selective factor, a point accentuated by the extremely cautious behaviour of the males toward the females. I then compare the traits associated with these models to the species for which they were designed and certain others for which adequate data also exist. None of the sets of responses to these traits closely resembles each other in the different species. These results suggest that sexual cannibalism has developed in response to a heterogeneous collection of lifestyles and phylogenetic constraints.  © 2004 The Linnean Society of London, Biological Journal of the Linnean Society , 2004, 81 , 427–437.