ABSTRACT
Infanticide, the killing of unweaned or parentally dependent infants, is one of a variety of behavioral and physiological mechanisms that animals and humans use to modify the sex ratio of their offspring (Hrdy, 1979; Johansson, Chapter 23, this volume). From both theoretical and empirical perspectives, offspring sex-ratio manipulation is currently a major focus of research by behavioral biologists, evolutionary ecolo gists, sociobiologists, demographers, and anthropologists. However, pa rental manipulation of offspring sex-ratio distribution is, in fact, a spe cial case of the more general problem of sex allocation in plants and animals, and it is this latter issue that constitutes the topic of this chapter. As such, the present review is intended as a primer for readers interested in recent theoretical developments concerning the evolution of sex allocation as well as for those interested in the strategy and findings of recent investigations of sex-ratio manipulation in animals. This review begins by discussing, in general terms, the use of natural selection as an explanatory principle. In attempting to understand, for example, the distribution of separate versus combined sexes (i.e., dioecy versus hermaphroditism), theoretical predictions (e.g., When should an organism be a hermaphrodite?) are sought in terms of the reproductive consequences of possible alternative states or the transi tion between states. In short, it is asked when or under what environ
mental, social, or life-history conditions natural selection favors one or the other form of sexuality. The theory is tested by arranging experi ments, or geographic or taxonomic comparisons, to see if selection acts as one thinks it does or if a particular form of sexuality is matched to predicted environmental conditions. Thus, it is necessary to under stand sexuality in terms of the ultimate causes (the why questions) rather than the proximate mechanisms (the physiologic how questions). Seeking answers to the way in which nature is structured, in terms of why questions, is used here as selection thinking.
