The Natural Selection of Sexual Cannibalism

Sexual cannibalism, in which a male is eaten by his mate following copulation, is expected to convey a selective advantage to the male under certain conditions. As shown quantitatively in our model the phenomenon is expected when (1) a male can mate only a few times in his lifetime and (2) the cannibalism significantly increases the number and/or viability of eggs fertilized by his own sperm. The expected number of male matings appears to be the more important of these two conditions. The few available observations of sexual cannibalism in several species of arthropods agree qualitatively with the predictions of our model. The same parameters of the model also explain why sexual cannibalism is generally rare among vertebrates. Phenomena selecting for sexual cannibalism appear more closely related to paternal investment strategies than to ecological factors associated with other forms of cannibalism. Because few field studies of invertebrates have followed the mating history of marked males, we cannot test our model quantitatively with existing data. We suggest specific field observations which would provide quantitative estimates of the expected number of male matings. In addition we describe possible laboratory experiments which would give more precise measurements of the increase in fecundity of the cannibalized male.