Louis Lefebvre Department of Biology, McGitl University, 1205, av. Docteur Penfield, Montrgal, Qugbec,H3A 1B1 Canada Received 30 June 1984 and accepted 30 October 1984 Ke~words: Food sharing, early weaning, chimpanzees, golden-lion tamarins, diet. Parent-Offspring Food sharing: a Statistical Test of the Early Weaning Hypothesis Among non-human primates, Pan troglodytes' and Leontopithecusrosalia are reported to show frequent parent-offspringfood sharing. If sharing of solid foods could lead to early weaning, femalesin these specieswould reduce the energetic (nutrient transformation) and reproductive (detayed fertility) costs of lactation. Age at weaning in the two species is compared with the 95% confidencelimits of the regressions that relate weaning age to adult female body weight, gestation length and age of sexual maturity in 52 primate species, since correlations with these variables must be taken into account to assess the food sharing-early weaning hypothesis statistically. P. troglodytes and L. rosalia fall within the confidencelimits of the regressions in all cases but one, leading to rejectionof the hypothesis. Food learning, in association with diet breadth, appears to be better supported as an alternative function for parent-offspring food sharing. Journal of Human Evolution (1985) 14, 255-261 1. Introduction Despite its obvious benefits to the young, lactation involves several costs to a fkmale primate, mainly the energetic cost of transforming food and fat reserves into milk (Buss & Reed, 1970; McGrew, 1975) and the reproductive cost of delayed return to fertility through lactational amenorrhea (Nadler et al., 1981). According to current evolutionary theory, females should thus be expected to wean infants at a time that minimizes these costs while maximizing lifetime reproductive success, taking into account survival and growth of the offspring being weaned as well as future reproductive opportunities (Trivets, 1974). In contrast to lactation, direct provisioning of the young with solid foods involves neither the costs of metabolic transformation nor of delayed fertility. If the young are capable of processing solid foods, parent-offspring food sharing could thus be an efficient alternative to lactation. Among mammals, parent-offspring food sharing is present in carnivores, but is only reported to occur frequently in two genera of non-human primates: chimpanzees (Pan troglodytes: Goodall, 1968; McGrew, 1975; Silk, 1978; Pan paniscus: Kano, 1980) and golden-lion tamarins (Leontopithecus rosalia: Ditmars, 1933; Coimbra-Filho, 1969, cited by Hershkovitz, 1977; Suyder, 1974; Wilson, 1971; Brown & Mack, 1978). In tamarins, Brown & Mack (1978) suggest that giving solid food to infants removes the strain of lactation from the female, while McGrew (1975) argues in his discussion of sharing that early weaning is in a female chimpanzee's reproductive interests. However, any rigorous assessment of this hypothetical relationship between food sharing and early weaning can only be done on a relative basis, since many other variables are likely to affect weaning age besides food eaten by the infant. In mammals, weaning age is correlated with several developmental life history traits, as well as with adult body weight (Blaxter, 1971). These effects must therefore be removed before we conclude that chimpanzees and golden-lion tamarins wean earlier than expected due to parent-offspring food sharing. In this paper, effects of adult female body weight, gestation length and age of sexual maturity on weaning age in 52 primate species are first determined by least-squares regression, The weaning ages of P. troglodytes and L. rosalia (P. paniscus is not considered here because life history data for this species are not readily available) are then compared with the confidence limits of 0047-2484/85/030255 + 07 $03.00/0 9 1985 Academic Press Inc. (London) Limited