Mandrills are large African monkeys that live in the dense equatorial forest of Central Africa. They are often described as living in multi-level societies, similar to those found in hamadryas and geladas, with large multi-male, multi-female groups composed of smaller one-male, multi-female units.
However, observations at CIRMF quickly showed that this was not the case. Instead there is always one dominant male associated with the social group of females and their offspring, while other males vary in the extent to which they associate with the group (Setchell & Dixson 2001). The number of males in the group varies with the presence of receptive females.There is no evidence that the multi-male, multi-female splits into sub-groups with one male in each.
Female mandrills inherit their mother's dominance rank, with the youngest daughter ranking just below the mother (Setchell et al 2008). Females very rarely fight and female ranks in the CIRMF colony have changed very little, beyond births and deaths, since they were first recorded in the 1980s.
Male mandrills become peripheral to the social group around the age of 6 years spending time on the edge of their group or solitary. It seems likely that they would disperse in the wild, but in the CIRMF colony they either remain solitary or rejoin the group when they are full size (Setchell & Dixson 2002, Setchell et al 2006). Male ranks are far more dynamic than those of females and are contested physically, often resulting in serious injury.
The CIRMF mandrill colony houses by far the largest population of captive mandrills in the world. It facilitates integrated, multidisciplinary, analyses of morphology, endocrinology, pathogens, genetics, and social behaviour
Mandrills live in forested enclosures, within their natural habitat range. This naturalistic environment, the opportunity to capture the animals periodically for collection of morphological data, blood and other biological samples, and the availability of historical records and banked DNA and serum samples for most individuals provides a unique opportunity to study these little known animals long-term.
Stress, life history, and dental development in primates
This project, funded by a Leverhulme project grant, combines Jo Setchell’s long-term studies of mandrill life history and earlier work on mandrill teeth (Setchell & Wickings 2004) with Wendy Dirks' expertise in dental histology.
We are using teeth of mandrills with known life histories who have died a natural death to compare the growth lines in their teeth with important life history events to test whether these events are recorded in the dentition.
We aim to establish a method that can be applied to fossil teeth of extinct hominin species, to provide a better understanding of human life history evolution.
Monkeys and apes have traditionally been considered as microsmatic, and their chemical communication has thus been almost entirely neglected. However, mandrills have a sternal gland on their chest that they rub vigorously against trees.
We made the first chemical investigations of odour signals in an Afroeurasian monkey and showed that odour encodes sex, age and dominance rank in mandrills (Setchell et al 2010).
The combination of an odour profile that signals sex, age and rank with increased motivation to scent-mark and increased production of secretion in high-ranking males leads to a potent signal of the presence of a dominant, adult male with high testosterone levels. This may be particularly useful in the dense Central African rain-forest which mandrills inhabit.
Female mandrills show mate choice for males that are genetically dissimilar to themselves (Setchell et al 2010), which produces more genetically diverse offspring, with a stronger immune system. We showed that genetically similar mandrills have similar odour, suggesting that mandrills use odour to recognise kin and identify optimal mating partners (Setchell et al 2011).
Mate choice for major histocompatibility complex (MHC) genes is thought to give offspring a fitness advantage through disease resistance.
We found a link between mating outcome and MHC dissimilarity in mandrills (Setchell et al 2010). These results are the first to demonstrate mate choice for genetic dissimilarity in a species with high male reproductive skew, and suggest that MHC-associated mate choice can occur even where male–male competition is intense.
This led us to investigate how primates identify their ideal mate. Monkeys and apes have traditionally been considered as microsmatic, and the study of olfactory cues has thus been almost entirely neglected. However, our chemical investigations of odour signals in mandrills (the first such investigations in Afroeurasian monkeys) showed that odour encodes not only sex, age and dominance rank, but also MHC genotype and, crucially, MHC dissimilarity (Setchell et al 2013). Thus, odour provides a mechanism by which primates may detect their ‘optimal’ mate.
Zahavi's handicap theory of sexual selection predicts that exaggerated secondary sexual ornaments are condition dependent, and that only individuals of superior quality will be able to express costly ornamentation. In particular, Hamilton & Zuk's parasite-mediated sexual selection hypothesis suggests that ornaments reliably reflect an individual’s ability to resist parasites by revealing current health status.
The immunocompetence handicap hypothesis extends this model further, positing that testosterone-dependent ornaments signal the ability to cope with the immunosuppressive effects of testosterone. Under these models, members of the opposite sex should choose the most ornamented mate because these high quality mates provide fitness benefits, either directly, through avoidance of parasite transmission and increased investment in offspring or both, or indirectly, by passing on ‘good genes’ for vigour and health to offspring.
We examined links between facial coloration, parasites, immune status, endocrinology, and genotype in male mandrills. Red colour is related to testosterone (Setchell & Dixson 2001a, Setchell & Dixson 2001b, Setchell et al 2008), and to specific genotypes, suggesting that red may signal ‘good genes’ in mandrills (Setchell et al 2009). However, red is not related to parasitism, immune status or genetic diversity, challenging some of the main theoretical explanations for the evolution of sexually selected male ornaments in animals(Setchell et al 2009).
An alternative model suggests that the immunocompetence handicap hypothesis functions via a trade-off between glucocorticoid levels and the immune system. If subordinate males suffer elevated glucocorticoid levels due to stress, this may suppress their immune system, and prevent them from producing testosterone and, therefore, red colour. We conducted the first test of this stress-mediated hypothesis for the evolution of condition-dependent traits in mammals. We found no relationship between red colour and glucocorticoid levels, suggesting that glucocorticoids do not play a role in translating social conditions or physical health into ornament expression in mandrills (Setchell et al 2010).
Female ornaments have received far less attention from evolutionary biologists than those of males.
Female mandrills have two types of ornament: facial colour and sexual swellings.
Female facial colour varies from bright pink to almost entirely black in mandrills. Unlike in males, red is not related to female rank or to reproductive success. However, females get brighter with age, and when they are in the follicular phase of their menstrual cycle. Colour also peaks during lactation (Setchell et al. 2006b).
Lack of support for the "reliable indicator" hypothesis
Female mandrills have pink sexual swellings that increase in size during the follicular phase of the menstrual cycle and are attractive to males.
Sexuall swellings are hypothesised to indicate female quality and allow males to select high quality mates. However, we found that differences in sexual swelling size and colour between female mandrills do not reliably advertise female quality (Setchell & Wickings 2004, Setchell et al 2006) and that males do not allocate more mating effort to females with particular swelling characteristics (Setchell & Wickings 2004).
Little support for a very popular idea
Where females cycle at the same time, alpha males are unable to monopolise sexual access
However, where females cycle asynchronously, alpha males can monopolise access to each female in turn and acheive a much higher paternity rate
The notion that women synchronize their menstrual cycles when they live together has great popular appeal. We tested for cycle synchrony in mandrills, and found little evidence that females that associate with one another cycle together (Setchell et al 2011). We suggest that reports for women are more likely due to an evolved human tendency to detect patterns where there are none than to actual synchrony.
"When the animal is excited all the naked parts become much more vividly tinted” (Darwin, 1871, referring to the mandrill)
Dominant (left) and non-dominant (right) adult male mandrills at CIRMF
Dominant males are far more brightly colored than other male mandrills. Males that gain rank become more colourful while those that lose rank decrease in colour (Setchell & Dixson 2001, Setchell et al 2008). This dramatic colour is related to testosterone, suggesting that male red is a dynamic, honest signal of competitive ability and willingness to fight (Setchell & Dixson 2001, Setchell et al 2008). This confirms Darwin's observation, although the colour changes occur over months rather than immediately.
Males that are similar in colour threaten one another more, fight more and have more tense 'stand-off' encounters, while clear submission is more frequent where males are very different in colour. This suggests that males use relative red coloration to assess relative fighting ability, regulating the degree of costly, escalated conflict between well-armed males (Setchell 2005)
Red colour is related to specific immune genotypes, suggesting that it may be condition dependent, and that only individuals of superior quality will be able to express color fully, and supporting the Hamilton-Zuk hypothesis (Setchell et al 2009)
An exceptional case of colour change: male 18 before, during, and after his time as dominant male
Red as an honest signal of current androgen status, competitive ability and willingness to engage in fights
Dominant male mandrills are redder (Setchell & Dixson 2001, Setchell et al. 2008) and have longer canines (Leigh et al 2008) than subordinate males. They also produce more testosterone (Setchell et al. 2008) and show elevated cortisol levels (Setchell et al 2010) during the mating season.
Dominant males mate-guard females during the mating season and sire the great majority of offspring (Setchell et al 2005).
Males gaining top rank increase in red colour and testosterone (Setchell & Dixson 2001). Red coloration develops after a male has attained top rank (Setchell et al. 2008). Colour continues to develop with tenure as dominant male. Dominant males that lose their top rank decrease in both testosterone and colour, although post-dominant males are also more colourful than males have never been dominant.
Males use the relative brightness of their red coloration to facilitate the assessment of individual differences in fighting ability (Setchell & Wickings 2005).
Females prefer to mate with redder males, independent of their attraction to dominant males (Setchell 2005).
The interaction of male-male competition and mate choice, both favouring bright red coloration in males, may explain the exaggerated ornamentation in this species.
Close ties between fitness and weaponry in a male primate
Male mandrills have upper canines up to 5cm long.
Males aged 9-11 years have the longest canines after which canine size diminishes through breakage and wear. Some old males have very small, blunt canines (Leigh et al 2008).
Canine size is strongly correlated with male reproductive success. Canine size matches the curve of reproduction vs age very closely and sires have larger teeth than males that fail to reproduce (Leigh et al 2008).
These results are the first to demonstrate close ties between fitness and weaponry in male primates.
Moreover, mandrill canines are non-renewable, unlike other kinds of mammalian weaponry, notably horns, antlers, and body mass, which can be renewed either continually or periodically. Canine length may limit the male reproductive lifespan
Social modulation of testosterone
Testosterone is closely linked to reproduction in vertebrates, and influences male sexual behavior, aggression, displays and secondary sexual characters (armaments and ornaments). However, high levels of testosterone are costly and can reduce immune function
Testosterone in male mandrills is positively related to dominance rank, suggesting that males live in a permanently aggressive context in which they must actively maintain their dominance status (Setchell et al 2008)
Testosterone increases when receptive females are available and when male ranks are unstable, both situations where males compete intensely (Setchell et al 2008), supporting the "challenge hypothesis" that testosterone promotes aggression when it is benefical to males (Wingfield et al 1990 American Naturalist, 136, 829-846).
Females can still choose even in a highly sexually dimorphic species
Although male mandrills are much larger than females, females can still decide whom to mate with because their small size allows them to escape up trees. They exhibit choice by inviting some males to mate and refuse unwanted mating attempts by avoiding the male's approaches or simply by lying down (Setchell 2005).
Female mandrills show mate choice for dominant males and for red males (Setchell 2005). This reinforces the influence of male-male competition on male reproductive success and may explain the very high reproductive skew in mandrill males and their extraordinary appearance.
Females also choose mates for genetic diversity and compatible genes (Setchell et al 2010).
The first exploration of gamete selection in a large primate and the logistical difficulties presented by such a study
Female mandrills choose genetically dissimilar mates (Setchell et al 2010). This mate choice may occur before copulation, at which point it is relatively easy to measure by observing behaviour (Setchell 2005, Setchell & Wickings 2006). However, mate choice can also occur post-copulation, where females mate with multiple males. This is far harder to study under naturalistic conditions as it requires information on exactly who a female mates with and when, throughout her fertile period.
We developed a method that circumvented this problem. However, our results suggest that it will be extremely difficult to determine conclusively whether post-copulatory selection mechanisms exist, particularly if the effect sizes are small. Nevertheless, we found no evidence of strong selection for post-copulatory female choice for male genotype in mandrills (Setchell et al 2013).
Subordinate female mandrills mature later and have longer intervals between successive pregnancies than dominant females (Setchell et 2002, Setchell et al 2005). Dominant females, therefore, have more offspring over a lifetime than subordinate females. However, this effect is not mediated by rank effects on 'stress' hormones (Setchell et al 2008).
Males choose in a highly competitive species
Although they invest little in their offspring beyond mating, reproduction is still costly for male mandrills, in terms of time and energy invested and the risk of aggression from both other males and from females (Setchell et al 2006).
Accordingly, males show mate choice by preferentially allocating their mating effort to ‘higher quality’ females – multiparous and higher-ranking females (Setchell et al. 2006), that are more likely to conceive (Setchell et al. 2002; Setchell & Wickings 2004), and produce larger offspring when they do so (Setchell et al 2001).
Achieving top-rank in male mandrills translates into huge reproductive advantages
Dominant male mandrills sire the large majority of offspring while very few subordinate males reproduce (Setchell et al. 2005; Charpentier et al. 2005).
Mate-guarding is a good predictor of paternity, but consistently overestimates the reproductive success of the domiant male (Setchell et al. 2005). This suggests that other males sneak copulations and supports the limited-control model of reproductive skew.
Beyond the effect of dominance rank, reproductive success in male mandrills also increased with genetic diversity (Charpentier et al 2005), MHC diversity and MHC dissimliarity to the female (Setchell et al 2010).
Odour provides a mechanism by which primates may detect their ‘optimal’ mate
Male mandrill scent-marking
When selecting a mate, animals may choose a partner with 'good genes' that they will pass on to their offspring. However, as both partners pass on genetic material, they may also select mates with genes that are compatible with their own. In particular, they may choose partners with MHC (major histocompatibity complex) genes that complement their own.
We tested this in mandrills, and found that they show mate choice for both good genes (MHC diversity) and complementary genes (dissimilar MHC) (Setchell et al. 2010). This is surprising, given the strong influence of male dominance rank on male reproductive success (Setchell et al 2005).
This raises the question of how mandrills recognize and assess the MHC genotype of conspecifics. Visual signals like a bright red face cannot convey information regarding genetic similarity, as this depends on the genotype of the receiver as well as the signaller. One possiblity is that females use post-copulatory mechanisms to select sperm (Setchell et al 2013). An alternative is that they compare their own odour to that of potential mates. Mandrills of both sexes possess a sternal gland, that males rub vigorously against trees. When we compared genotypes with odour profiles we found that genetic similarity correlates with odour similarity, providing a mechanism by which mandrills may recognise kin and detect compatible mates (Setchell et al 2013).
Different reproductive priorities lead to very different life histories in males and females
Animals have finite resources to allocate to growth, maintenance and reproduction. Life history theory suggests that these allocation decisions, and the schedule and duration of key events across an individual's lifetime are shaped by natural selection to maximise fitness.
In sexually dimorphic, polygynous species like mandrills, life history theory predicts sex differences in age-specific reproductive output and mortality profiles, and greater variance in lifetime reproductive success in males than in females.
Male and female growth strategies and reproductive careers conform to these predictions in mandrills (Setchell et al 2005).
Although both males and females can reproduce at about 4 years, males continue to grow for a further 6 years resulting in a much larger body size that helps them to compete with other males (Setchell et al 2001). Their average age at first reproduction is 11.6 years in males, by which time females already have several offspring.
Females do not need to grow large to compete physically, so begin to reproduce much earlier than males. Pregnancy and laction limit females to a maximum of one offspring per year (Setchell et al 2005)
Average lifespan is shorter for males (14 years) than females (>22 years) (Setchell et al 2005)
All females of breeding age produce offspring; while only 1/3 of males sire. However, the reproductive output of a successful male is far more offspring than a female can bear in a lifetime (Setchell et al 2005).
One of the most sexually dimorphic and brightly coloured mammals
Male mandrills weigh more than three times as much as females (Setchell et al al 2001), making them the most sexually dimorphic primate and one of the most sexually dimorphic mammals. They are also one of the most colourful mammals.
Males are born slightly larger than females, but most of the adult size difference is achieved after weaning when male grow both faster and for longer than females (Setchell et al al 2001). Male red coloration and other secondary sexual traits begin to develop at 6 years, when males are already reproductively mature and can sire offspring but are far smaller than adult males. Males then take another four years to reach full adult body size at around 10 years (Setchell & Dixson 2002; Setchell et al. 2006a).
These differences in size and growth strategies relate to sex differences in reproductive priorities. Females reach adult body mass at 7 years, by which time they have had more than one baby. While all females reproduce, male reproduction is largely limited to dominant males so males must grow large to win fights. Males reach adult size at about 10 years.
Males also possess long canine teeth, on average 4.5 cm long. Males with larger teeth sire more offspring (Leigh et al 2006).
Mother's rank and experience affect offspring growth and fitness
Higher-ranking female mandrills have heavier babies than lower ranking females (Setchell et al 2001). These early advantages persist after weaning, when mothers are no longer directly investing in offspring.
Daughters inherit their mother's rank, and dominant females have their first baby on average 1.3 years earlier than lower ranking females (Setchell et al 2002).
Sons of higher-ranking females are more likely to survive to adulthood than sons of low-ranking females, mature faster than sons of lower-ranking mothers, and are larger as adults (Setchell et al 2006)
Older female mandrills also have heavier babies than younger mothers (Setchell et al 2001).
Social interactions can be an important source of stress
Hormones associated with stress, such as cortisol, allow animals to cope with acute events. However, long-term high levels of such hormones can have serious health consequences. For example, male mandrills with higher cortisol levels suffer more parasite infections, suggesting that glucocorticoids suppress the immune system (Setchell et al 2010).
We examined the relationship between social behaviour and cortisol in male and female mandrills.
The relationship between cortisol and male rank depends on the social environment. When the dominance hierarchy is stable clevels are higher in lower ranking males. However, when the hierarchy is unstable, higher ranking males have higher cortisol. These patterns are likely to be due to differences in male ability to predict and control the social environment during stable and unstable periods.
Cortisol is also higher in dominant males than in subordinates during mating periods, suggesting that dominant males are more stressed than subordinates during such periods (Setchell et al 2010).
We found no relationship between cortisol and rank in females (Setchell et al 2008). This suggests that lower reproductive success in low-ranking females (Setchell et 2002) is not due to chronic stress.
These long-term studies would have been impossible without the support of the CIRMF, Primate Centre Staff past and present (Audrey Morelli, Guy Dubreuil, Olivier Bourry, Pierre Rouquet, Bettina Sallé, Paul Bamba, Philippe Enganja, Anaïs Herbert, Barthelemy Ngoubangoye and many others), and the Laboratoire d’Analyses Medicales.
Similarly none of this work would have been possible without the help of collaborators: Alan F. Dixson introduced me to mandrills and CIRMF and supervised my PhD; E. Jean Wickings welcomed me to CIRMF and generously shared her long-term knowledge and data on the mandrill colony, her cars and even her house on occasion; Phyllis Lee examined my PhD and then taught me how to write; much of this work would not have happened without Marie J. E. Charpentier; Leslie A. Knapp introduced me to the wonders of the MHC and was my post-doc mentor for three years; Kristin Abbott conducted the MHC genotyping; Stefano Vaglio & I met at a conference, realised we were studying the same question and initiated an exciting and ongoing collaboration.
Thanks also to Steve Leigh and Robin Bernstein for friendship and collaboration on growth and development; the late and much missed Sosthene John and Benoit Goossens for help with sampling; Trish Reed and Issa-Ben Bedjabaga for parasite analyses; Tessa E. Smith for teaching me non-invasive EIA and welcoming me into her lab; the Centro di Servizi di Spettrometria di Massa, Universita` degli Studi di Firenze for odour analyses; Charlie; and, of course my families in Gabon and the UK.