(Anderson and Dixson, 2002) vary between monogamous and polygynous mating systems as a result of selection for sperm competition. For instance, seasonally breeding rhesus macaques have 60 percent larger testes than do the closely related and similarly sized nonseasonally breeding pigtailed macaques, a result of the greater mating competition during seasonal breeding. In addition, rhesus males have 60 percent larger abdominal fatfolds than pigtail males (Muhlenbein et al., 2002), suggesting that substantial energy reserves are needed to offset the energetic costs of male-male competition, among male rhesus.

Such findings are powerful evidence for the impact of selection on sperm production and delivery across species. However, the relatively small testicular to body size ratio exhibited by humans suggests that sperm competition in human males has played only a small role in human evolution (Dixson, 1998; Smith, 1984). Furthermore, the low percentage of normal sperm in human ejaculates is comparable to that of gorillas and very different from that of chimpanzees (Small, 1988), suggesting that the human male reproductive strategy has more in common with preinsemination mate guarding among gorillas than the postinsemination sperm competition of the chimpanzee. Thus, the fact that the size of male ejaculates has been shown to vary with time since exposure to their partners is less evidence for sperm competition in humans (Baker and Bellis, 1989) per se than it is for the importance of mate guarding.

However, in the case of humans, the role of paternal investment in offspring means that selection on males includes not only behavior leading to sex, and aspects of the male reproductive tract associated with insemination, but also behaviors related to pair bonding and the provisioning of offspring. Anecdotal evidence that testosterone increases in males exposed to mates after a prolonged absence (Bribiescas, 2001) suggests a role for testosterone in sexual behavior and pair bonding (see Young, this volume). Recent reports of lower testosterone levels in fathers of young children than in nonfathers (Berg and Wynne-Edwards, 2001; Storey et al., 2000) suggest that changes in testosterone may play a role in shifting behavior from mating to parenting.

With that brief orientation toward the critical elements of male reproductive strategies among humans, I turn to the central focus of this paper, the role of reproductive maturation in the development of male sexual behavior, pair bonding, and parenting.

Here I argue that this process can be considered to have two major dimensions: 1) sexual behavior that is related to reproductive maturation and somatic growth and that is organized by testosterone as part of the reproductive axis, and 2) social behavior that is related to early childhood experience and to the adrenal axis, which produces cortisol and dehydroepiandrosterone and its sulfate DHEA/S. Briefly put, maturation of the