Because of the necessity of lactation, mammalian mothers must perform at least a minimum amount of infant care. In cooperatively breeding species, other group members aid in all other aspects of infant care. However, some mothers continue to carry and nurse their infants more than others. The golden lion tamarin, Leontopithecus rosalia, is a small, communally breeding primate in the family Callitrichidae. We studied hormonal, individual, historical and social factors hypothesized to contribute to variation in levels of maternal care. We used neonatal weight as a measure of prenatal care, and carrying and nursing as measures of postnatal care. Greater neonatal weight was associated with smaller litter size, lower prepartum levels of oestrogen conjugates, and higher prepartum cortisol levels. Higher rates of carrying during weeks 2 and 3 were associated with higher maternal weight, larger litter size and smaller numbers of helpers per infant. Higher rates of nursing in weeks 2 and 3 were predicted by smaller group size and provisioning of the mother. The most important factors affecting postnatal maternal care were maternal weight, group size, litter size and provisioning status of the mother. Thus, females that display higher levels of maternal care do so either because they have to (they have fewer helpers) or because they can (they are in better condition). © 2002 The Association for the Study of Animal Behaviour.

We examined the effects of several variables on the number of live births in multiparous females in a wild population of golden lion tamarins (Leontopithecus rosalia). Independent variables included the number of infants born to a female the previous breeding season, the number of infants weaned the previous breeding season, the female's age and body mass, the number of adult males and helpers in the group, and the inbreeding coefficient of the offspring. We also tested the hypothesis that trapping and chemical immobilization during pregnancy affected the number of live births. Female body mass was the only statistically significant predictor of the number of live-born infants in the current season when both first and second peaks were included. Characteristics that predicted higher numbers of infants in the first peak of a season were the number of infants born the previous season and the body mass of the female. The greater the number of infants born the previous season, the greater the number of infants born in the first peak of the current season. Factors positively correlated with the number of live births in the second peak within a season included the number of infants born the previous season, as well as the number of available helpers. Due to sample size constraints, the analysis of litters in the second peak did not include body mass of the female as a variable. Inbreeding and handling did not affect the number of live births. We found no evidence that current reproduction negatively impacts future reproduction in this species. We also found no evidence for an age-related reduction in fertility.

Callitrichid females are often described as energetically constrained. We examined the energy budgets of 10 female wild golden lion tamarins (GLTs, Leontopithecus rosalia) in an effort to understand how energy intake and expenditure might influence physical condition and therefore reproductive performance. We used focal animal sampling to record behavioral data and conducted energy analyses of foods consumed by GLTs to estimate intake and expenditure. We used two-tailed Wilcoxon signed-rank tests to compare intake in the reproductive vs. nonreproductive period and expenditure in the reproductive vs. nonreproductive period. Energy intake decreased during the reproductive period compared to the nonreproductive period. While total expenditure did not vary significantly across the two periods, females spent more time and therefore expended significantly more energy engaged in energetically inexpensive behaviors (i.e., sleeping or being stationary) during the reproductive period compared to the nonreproductive period. We suggest that reproductive female GLTs may adopt a reproductive strategy that includes high intake prior to pregnancy and lactation, and energy conservation during pregnancy and lactation.

This study examines factors affecting androgen and cortisol levels in wild, male golden lion tamarins (Leontopithecus rosalia). Golden lion tamarins are a cooperatively breeding species in which groups often contain two potentially breeding males. Brothers frequently emigrate together and develop a clear dominance hierarchy, but interactions between them are primarily affiliative. Duos in which the males are not related are less stable. In addition, reproductive skew theory predicts that dominant males will be less likely to share reproduction with related subordinates. As such, we predicted that both androgens and cortisol would be higher in subordinate males unrelated to the dominant male. We also predicted that androgens in breeding males would be higher during the mating season than the birth/infant care season, as per Wingfield's "challenge hypothesis" (1990). Fecal samples were collected from 24 males in 14 social groups and assayed by enzyme immunoassay. Androgen levels were higher in breeding males during the mating season, thus supporting the challenge hypothesis. However, while subordinate males unrelated to the dominant male had significantly lower androgens than any other group, cortisol levels were not correspondingly higher. These results suggest that unrelated subordinate males show measurable reproductive suppression and may use strategies such as infantilization to avert aggression from dominant males.

The steroid hormone cortisol has been associated with different levels of "stress" as well as different reproductive conditions in many primates. In callitrichids, cortisol has more often been reflective of female reproductive status than of chronic stress. In this study, we addressed the hypothesis that wild golden lion tamarin (Leontopithecus rosalia) females, whose social structure is characterized by low aggression and high social support, would not show rank ("stress")-related differences in glucocorticoids but would show reproductive changes. We collected 710 fecal samples from 22 adult females in Poço das Antas Reserve, Brazil, and nearby reintroduction areas, and assayed them for cortisol. Differences in cortisol levels were found between different reproductive conditions. Females in the first trimester of pregnancy had lower cortisol levels than nonpregnant females, although we did not differentiate between basal and ovulating levels in nonpregnant females. Cortisol rose in the third trimester of pregnancy. Primiparous females had a higher rise in the third trimester than multiparous females. No differences in cortisol levels were found among dominant females, ovulatory subordinate females, or anovulatory subordinate females. These results are similar to those obtained in other studies of callitrichid females. The lack of differences in cortisol excretion between dominants and subordinates is likely due to the low levels of overt aggression and the high level of social support available to subordinate females.