Mandrills learn two-day time intervals in a naturalistic foraging situation.

Primates display high efficiency in finding food in complex environments. Knowledge that many plant species produce fruit simultaneously, can help primates to anticipate fruit finding at the start of fruiting seasons. Knowledge of elapsed time can help primates decide when to revisit food trees to find ripened fruit and to return before competitors find these fruits. To investigate whether mandrills are able to learn time intervals of recurring food, we recorded the foraging choices of captive mandrills in a group setting. We used a procedure with renewable food rewards that could be searched for: carrots and grapes, hidden underground in specific places with different renewal intervals (2 and 5 days, respectively). We monitored the first choice of location for individuals, if other individuals had not already searched at the same location, to exclude possible effects of individuals following others rather than relying on memory. Throughout the study, the mandrills became increasingly likely to first search at carrot locations on carrot days, while the probability of them searching at carrot locations decreased on days without carrot. Due to model instability, our results were inconclusive about an effect of grape days on the choice of the mandrills. Cues provided by conspecifics indicating the availability of simultaneously emerging food rewards did not affect the choice of the mandrills. We conclude that mandrills can take into account elapsed time in a foraging context. Thereby, this study indicates how mandrills can use temporal cognitive abilities to overcome temporal challenges of food-finding in a group setting.

Natural infection of free-ranging mandrills (Mandrillus sphinx) by enteroviruses and astroviruses in southern Gabon.

Enteroviruses (Picornaviridae) and astroviruses (Astroviridae) cause various diseases in humans and animals, including in non-human primates (NHPs). Some enteroviruses and astroviruses detected in NHPs are genetically related to those infecting humans, indicating the occurrence of interspecies transmissions. In this study, we screened 200 fecal samples of 56 free-ranging mandrills (Mandrillus sphinx) by nested reverse transcription-PCR with primers targeting the VP1 and RdRp genes, to evaluate the diversity of enterovirus and astrovirus infection, respectively, and the associated zoonotic risk. Overall, ten samples from six mandrills were enterovirus-positive (5%), and three samples from three mandrills were astrovirus-positive (1.5%). This is the first evidence of astrovirus infection in mandrills. Phylogenetic analyses based on the VP1 sequences revealed that all ten enterovirus sequences were part of the species Enterovirus J, suggesting low zoonotic risk. Phylogenetic analysis of the three astrovirus sequences showed that they all belonged to the Mamastrovirus genus. Two astrovirus sequences were highly divergent from all human astrovirus sequences (63.4-73% nucleotide identity), while one sequence (AstV-5) suggested cross-species transmission from humans to mandrills. Additional studies are needed to better characterize the identified astroviruses and to confirm whether mandrills are host of astroviruses than can be transmitted to humans.

The dynamics of grooming interactions: maintenance of partner choice and the consequences of demographic variation for female mandrills.

A large body of evidence suggests that female Old World monkeys maintain selective long-term grooming interactions with fitness benefits. The last two decades have produced evidence that the regulation of social interactions among primates can be, in part, explained by the Biological Markets theory, with grooming behaviour as the focus of these studies. Grooming facilitates bonding between individuals, constituting an essential part of the regulation of social relationships among female cercopithecids. In contrast to the well-studied baboons (Papio spp), knowledge about the nature of grooming interactions and their regulation is generally lacking for the large, terrestrial species of mandrills (Mandrillus sphinx). We used a combination of social network analysis tools and well-established methods for assessing partner diversity and reciprocity to characterise grooming networks, partner choice and patterns of trade (be groomed, give grooming) among females in a captive group of mandrills, both within and across two separate observation periods. Our results suggest that, even though the relatively stable conditions of captivity allowed the studied females to maintain selective grooming interactions across time, small scale demographic changes affected the grooming dynamics of the group in accordance with the expectations of the Biological Markets theory. In particular, the maturation and consequent integration of a high ranking female into the group's grooming network from one period to the next resulted in a more pronounced effect of rank on the regulation of grooming interactions. In addition, the influence of the maturation of a dependent infant on the grooming interactions of his mother were evident between periods. Our results also demonstrate that grooming networks are dynamic and that high ranking individuals are not necessarily the most central in grooming networks. Finally, we discuss the potential of social network analysis to identify cases of social exclusion and its consequences for captive management.

Androgens in a female primate: Relationships with reproductive status, age, dominance rank, fetal sex and secondary sexual color.

A comprehensive understanding of the role of androgens in reproduction, behavior and morphology requires the examination of female, as well as male, hormone profiles. However, we know far less about the biological significance of androgens in females than in males. We investigated the relationships between fecal androgen (immunoreactive testosterone) levels and reproductive status, age, dominance rank, fetal sex and a secondary sexual trait (facial color) in semi-free-ranging female mandrills (Mandrillus sphinx), using samples collected from 19 reproductively mature females over 13months. Fecal androgens varied with reproductive status, being highest during gestation. Fecal androgens began to increase at 3months of gestation, and peaked at 5months. This pattern is more similar to that found in a platyrrhine than in other cercopithecine species, suggesting that such patterns are not necessarily phylogenetically constrained. Fecal androgens did not vary systematically with rank, in contrast to the relationship we have reported for male mandrills, and in line with sex differences in how rank is acquired and maintained. Offspring sex was unrelated to fecal androgens, either prior to conception or during gestation, contrasting with studies of other primate species. Mean facial color was positively related to mean fecal androgens across females, reflecting the same relationship in male mandrills. However, the relationship between color and androgens was negative within females. Future studies of the relationship between female androgens and social behavior, reproduction and secondary sexual traits will help to elucidate the factors underlying the similarities and differences found between the sexes and among studies.

Ten days in the life of a mandrill horde in the Lopé Reserve, Gabon.

Mandrills have long been known to form large aggregations in the wild, but it has proved difficult to obtain detailed information on the socioecology of these groups. An unusually large (>600) horde of mandrills was followed for ten days during the 1995 dry season in Central Gabon, and data were collected on group composition and ecology while the mandrills were in an area of forest-savanna mosaic habitat in the north of the Lopé Reserve. Three separate counts of most of the group showed that fully coloured "fatted" adult males were present throughout the horde at a mean ratio to other individuals of 1:21. Paler "non-fatted" adult and sub-adult males were also distributed throughout. Mandrill diet over the ten days consisted mainly of insects, seeds from forest trees, and leaves or stems of understory herbaceous plants. Feeding was extremely selective, with most food items consumed in a much higher proportion than would be predicted from their relative availability. Ranging data also showed that the mandrills foraged preferentially in certain forest types within the forest-savanna mosaic, namely in Marantaceae and Rocky Forest. It is suggested that one reason why mandrills pass through gallery forests and forest-savanna mosaic in the dry season in the Lopé Reserve is because they find fruit there from preferred species, which are no longer fruiting in the main forest block, thus allowing them to maintain the fruit component of their diet at a time of fruit shortage. © 1996 Wiley-Liss, Inc.