PRIMATES, 11: 327-333, 1970

327

Coalitions among Gelada Baboons

CLAUD A. BRAMBLETT

University of Texas

ABSTRACT. Geladas spend an unusually large amount of time in quiet, mutual grooming in this captive sample, part of which is presumably due to the effects of captivity and part is a reflection of the gelada pattern of social interactions. The most frequent signals associated with threat or displacement (lid) occurs only at a rate of 5.3 times per hour. Herding and displacement activities are .important, but infrequent. One-male bisexual units are extremely cohesive. A male does not tolerate the positioning of another male between himself and any of his females, and the adults of the groups react antagonistically to each other's proximity. The solitary male is not groomed by other adults, but stays with Chief's group and backs Chief in interactions with the other group. Chief's group is a two-male unit in some ways, but with Demon being excluded from any activity or location which interferes with Chief's interactions with his group. The females in Echo's group exhibit a dominance hierarchy, as measured in displacement, or denial of desirable social space. Interactions between individuals of Echo's group reflect strong preferences between individuals, prohibitions toward certain relationships by more dominant individuals, and the ability to enlist the support of others in coalitions. Coalitions are simply another expression of social relationships that are more frequently expressed in grooming, proximity, and displacement.

Gelada baboons are described by CROOK (1966, 1968) as terrestrial quadrupeds which inhabit the highlands of central and northeast Ethiopia. Males are large (45 lb) with long brown capes and baboon-like canine teeth. Females are about half the body weight of males. Both sexes have a chest patch of colored skin, and females have ventral caruncles (beads) which signal their reproductive status. Geladas are organized in large seasonal social units called 'herds.' Counts for different herds range from 30 to 400 individuals. Within herds CROOK distinguishes four kinds of groups: A. bisexual groups (which he refers to as 'harems'). B. all-male groups. C. juvenile groups. D. infant play groups. Bisexual groups consistently are composed of one fully adult male, one or more adult females, and their juvenile offspring. The ratio of adult males to adult females in 14 one-male units tabulated by CROOK (1968) is 1:5.7. These bisexual groups seem to be reproductive units of the herd. My brief study examines coalitions among a laboratory group of gelada baboons by recording the frequency and social context of signals which they exchange. A

328

C . A . BRA~VmLETT

Groom ~ 14.74 Present for Groom 7.31 Brow Raise 5.37 Coalitions 4.40 Displace 3.27 Lipsmack a.93 Sex Present m 2.00 Brow Raise/Displace m 1.98 Pursuits ii 1.48 Lip Roll m 1,25 Mount I 0.90 Bead Inspect [ ] 0.66 Bead Present I 0.53 Heart I 0.38 I 0.31 Rear Embrace Fig. 1. hour.

Frequency of singals in gelada baboon sample. n=2,850 signals.

Fig. 2. Grooming pattern between the members of Echo's group. Each line represents 5% of the total groom score. Scores less than 5 % are not shown.

The figures express signals per

Fig. 3. Patterns of sitting in Echo's group. Neighborhood is recorded as proximity < 10 ft when animals are at rest. Each line represents 5 ~ of the total score. Scores less than 5 % are not shown.

Coalitions among Gelada Baboons

329

coalition is an alliance for joint action. Signals are structured behaviors which a sender orients toward a receiver. The intent of the study is to sample social communication and spacing between individuals to test the hypothesis that subgroupings, coalitions, reflect long term relationships that can be quantitatively detected. Three kinds of data are collected. First, signal context, sequence, orientation, and frequency are recorded during 2,000 second test periods. An arbitrary choice of the most frequent and readible signals is made to reduce the complexity of the data sheets. The sum of signal count test periods is 60 hours, sampled between June 9 and August 25, 1969. Secondly, neighborhood (proximity) is recorded on an individual sampling basis to determine patterns of sitting. The sum of the 30 second sample neighborhood intervals for each animal is 20 hours. Thirdly grooming activities are recorded in 1,000 second test sessions by orientation and duration (n= 21 hr). For the present analysis, only the adult data are presented. Maternal or infant-directed behaviors are omitted. THE STUDY SAMPLE

The study sample is composed of 14 geladas which are housed in a 100 x 100 ft open air compound at the Yerkes Regional Primate Center Field Station (Lawrenceville Georgia). They are organized into three social units, one solitary adult male (Demon) and two bisexual groups. The smaller bisexual group includes one adult male (Chief), two adult females (Lark and Kay), and Kay's five month old infant male (Oscar). Echo, another adult male, six adult females (Fiend, Grey, Hanna, Iris, Juju, Mattie) and two male infants comprise the second group. Juju's six month old infant (Nero) is the oldest subadult and Grey's two month old male (Quo) is the youngest infant in the compound. FREQUENCY OF SIGNALS The most frequent signals recorded in the groups are grooming exchanges (14.7 bouts per hour). Mattie, the least active adult in either bisexual group is involved in grooming interactions 15 ~ of the day. The most active adult (Grey) spends 40 of her time grooming or being groomed by another adult. The mean percent of time that individual adults engaged in grooming is 27 ~ (omitting Demon with his zero score). Relatively high frequencies of present for groom and sex present (Fig. 1) reflect the usage of these signals in the formation of grooming interactions. The second most frequent category of signals are those related to threat and displacement. Brow raise or 'lid' gesture is most frequent (5.37/hr) and is followed by displace movements, lipsmack (frequently given in response to a lid signal), pursuit, and lip roll. Lid, lipsmack, and lip roll are all utilized by males when herding females. Therefore their frequency does not have the same implications about patterns of dominance that one might find in a primate society which did not 'herd.' Prolonged or mutual embraces such as mount, bead inspect, bead present, heart, and rear embrace are much less frequent than signals related to grooming or displacement. Coalitions (as a general category) occur at approximately the frequency of the threat and displacement interactions.

330

C . A . BRAMBLETT

Demon Demon is an isolated male, although he follows Chief's group in its movements about the compound. Infants approach and interact with him. He grooms, carries, plays, and mounts them.

Chief's GROUP In Chief's group, females are more frequently involved in grooming activity than is Chief. Each grooms him about the same amount (13 ~ of total grooming). Kay receives twice as much grooming from both Chief and Lark receives. Chief mounts Kay almost three times as much as he mounts Lark during the study period. Kay is the nearest animal to Chief in spatial distribution (31 ~ of the sample compared to 25 ~o for Lark). Kay initiates the bead present gesture toward Lark more often (8:1) than Lark to Kay. Echo's GROUt, Echo's group is more complex due to its larger size (Fig. 2). Echo sits near, grooms, and is groomed primarily by Iris. Among the females, Iris grooms and is primarily groomed by Juju. Juju and Grey have intense proximity and grooming relationships. Mattie is relatively isolated and grooms mainly with Fiend. In the group, Grey seems to be the social magnet, but Juju actively displaces Mattie and Fiend if they approach Grey. Hanna is an ubiquitous groomer and she grooms every-one a little. Juju displaces her from Grey but she persists. In terms of grooming relationships Echo is not the center of the group, but is quite peripheral. His grooming pattern with Iris is not disrupted by estrus in other females. During the signal counts, he mounted Juju 15 times, Iris 5 times, Mattie and Hanna twice. Comparison of neighborhood is very informative (Fig. 3). Animals may sit near each other without interacting, but one would expect it to correspond to grooming patterns. Two noteworthy exceptions are present: 1. Hanna does not sit near Yuju. 2. Even though Hanna grooms Grey more than Juju, Juju is near Grey more than

Hanna. Juju gives most of the intragroup lid signals (45 ~). Mattie receives the greatest portion of these, followed by Hanna, Iris, and Fiend. Grey sends 18 ~ of the lid signals, a large portion of which are directed toward Juju. Mattie receives more than twice the number of lid signals that are received by any other female (33 ~o of those sent) while Grey receives the least, only 3 ~. It is noteworthy that Mattie receives 37 ~ of all lid signals which are followed by displacement. If displacement and pursuit are used to measure dominance, the following sequence occurs:

Echo Grey Juju Iris Hanna Fiend Mattie

Coalitions among Gelada Baboons

331

INTERACTIONS BETWEEN GROUPS

The most common interactions between groups are signals which are exchanged when one group displaces the other. Twenty-six percent of the total lid signals scored occur between individuals of separate groups. Echo was far more active than Chief (2.4:1) with Lark and Juju the most active females. Chief displaced Echo 70 times during the signals count tests. Echo displaced Chief only once. Most displacements are nonviolent although Echo is observed to slap Lark. Infrequently Echo is pursued by either Chief or Demon (5 times). COALITIONS

The most frequent coalition recorded is the bisexual group itself. Chief and Echo carefully herd or follow their females and keep the two groups physically distinct, often at opposing portions of the compound. No adult member of Echo's group has any measurable contact with Demon. Demon is closely associated with Chief. He follows Chief's group about the compound, carefully avoiding passing between Chief and his females. Demon and Chief act as a coalition against Echo, and with Demon's backing, Chief is able to displace Echo. Coalitions within Chief's group primarily represent interactions between groups (135:2 or 98 ~). In contrast, coalitions recorded within Echo's group are associated primarily with intragroup interactions (100:13 or 88 ~). There is a definite social context within Echo's group in which coalitions occur, or are most likely to occur. For example half of the coalitions which include Echo are situations in which he is assisting Mattie against the more dominant and more aggressive individuals of the group. Juju and Grey back up Echo in intergroup interactions. Grey and Juju aid each other in interactions directed at other members of Echo's group with the exception that Grey will support Hanna against Juju when Juju becomes too enthusiastic about displacing Hanna from sitting near or grooming Grey. The most common object of coalitions is Mattie, and Mattie participates least of all in forming coalitions with the other females. When enlisting aid rather than giving it, Echo enlists Juju and Grey to support him against members of Chief's group (4 of 4). Juju enlists Grey or Echo against the more subordinate individuals in 94~ (16 of 17) of her enlistments. Fifty-eight percent (10 of 17) of these interactions are directed against Mattie. Just as Juju is the most active of the females at sending the lid signal, she initiates 70 ~ of the intragroup enlist signals. Coalitions of females in Echo's harem do pursue him, although he appears 'dominant' with respect to the male role of protector and herder. Echo was pursued by Iris and Mattie, Fiend and Hanna, and twice by Hanna and Juju in situations where he was clearly subordinate in his response to the females.

DISCUSSION Although herding behaviors make gelada males conspicuous group members, these males are not central to social interactions between adult females. In contrast to appearance, the male is peripheral, in many ways similar to the patas (HALL,

332

C . A . BRAMBLETT

1965). The adult male gelada performs a well-defined role in the group. One facet of this role involves group spacing and cohesion. He reacts when a female is too far from the group. Distance is a relative condition and varies according to the social context and state of excitement. No adult male non-group member may block access between a male and any of the group females. No non-aggressive contact was observed in this study between bisexual groups. Herding behavior may be accompanied by a variety of signals, many of which are submissive, The male looks at and often approaches the offender. He may direct both the lid and lip r011 signals toward her. A brief pursuit of the offender by the male may occur. Thus gelada males have the appearance of dominance when they are herding. If the level of excitement is high, a frightened female may chase the adult male. In extreme situations, females seem to react as a class to chase the male. Such chases are both noisy and rapid. If a vigorous retreat does not pacify the females, Echo may stop the chase by fleeing up one of the leafless trees in the compound. A second facet of the male role is his priority to certain kinds of social space. Normally this does not require activity on the part of the male. Animal technicians and caretakers can approach the group when the animals are near the compound fence. The sudden appearance of food in this context stimulates flight by adjacent females. In a limited food situation, a male will approach and occupy the position of priority. Female behavior is structured by long-term relationships with other females and by levels of excitement. Most of the patterns of contact and avoidance are stable extensions of dyadic relationships which do not change during the three months of observation. The exception to this situation is recorded only in violent chases when individuals seem to react according to their involvement or excitement. Such situations are most often stimulated by a female. Lark often penetrates or approaches the area of Echo's group. Chief and Demon follow. Chief's group and Echo's group exchange threats; Echo herds his females and his group is displaced. When precipitating a displacement, Lark sometimes carries Oscar dorsally, both sending 'lid' signal. Contact between groups rarely occurs. If a displacement turns into a chase, Echo sometimes runs at Lark and cuffs her with an open hand before fleeing from Demon and Chief. One such interaction is described below: June 11, 1969. 9:45 A.M. Lark is carrying Oscar dorsally as she approaches Echo's group. Mattie starts the chase by charging Lark. Lark, Chief, and Demon respond. Lark cuffs ffu]u. Mattie and Juju threaten Lark, then everyone is moving and threatening. Lark, Chief, and Demon chase Echo's group. Hanna becomes the aggressor for a moment. Echo's group is chased three-fourths of the way around the compound. Kay seems timid and brings up the rear. Lark retains Oscar dorsally. Oscar is also sending the lid signal. The importance of females in the interaction should be clear, even though the males have the size and strength to inflict extreme physical damage. The complexity of interaction within the group is well illustrated by the following notes: July 4, 1969. 12:40 P.M. Echo's group runs inside the trailer to avoid Chief, Demon, and Lark. Mattie has Nero dorsally. Juju chases Mattie to retrieve Nero. Grey assists Juju. Hanna assists Juju. Iris acts as pacifier. Iris sexually presents to Ju]u and simultaneously

Coalitions among Gelada Baboons

333

lipsmacks at Mattie. Nero is playing. Juju plays with him. Echo does repeated headstands on the perch. He falls each time, sometimes doing backward flips. Echo finally leads the group out. Hanna is last to leave. Through all interactions Fiend avoids participation. Time elapsed between entry and exit is 1,600 seconds. Although maternal and infant behaviors are not included in this discussion, the previous examples illustrate the importance of the infant. In one sense infants are not group members. They move freely between bisexual groups and interact, sometimes fearfully, with Demon. In another sense they are extensions of their mothers and in some ways derive her influence: June 11, 1969. Sheet II. Demon is carrying Oscar dorsally. Demon walks through Echo's group and displaces Echo who is sitting in the midst of his females. Demon enlists Chief. Echo moves out. Echo's females ignore the interaction. Both Demon and Lark appear to use Oscar in this way when displacing Echo's group. It is unfortunate that distance and the observation window prevent scoring of vocal communications. Geladas are highly vocal, both in quiet resting postures and in violent interactions. Group movements are frequently initiated by females, Lark often responsible for movements of Echo's group. At other times one or more females in Echo's group may initiate movement simply by starting to walk away. Echo may then respond by herding the other females after them. The pattern of activity during the day is greatly affected by sun and temperature. An obvious expression of this is the positioning of animals in the shade in the afternoons. Shadows of the metal compound walls are most popular, the trailer is used primarily by Echo's group to escape Chief's group rather than the heat. Acknowledgements. I wish to thank Dr. IRWIN BERNSTEINand the staff of the Yerkes Regional Primate Research Center for assistance and encouragement in observation. The research is supported by N.I.H. Grant FR-00165 from the Animal Resources Branch. SHARON Bm~raBL~rr helped with tabulation of data, typing of the manuscript, and illustrations.

REFERENCES

CROOI~, J. H., 1966. Gelada herd structure and movement, a comparative report. Symp. ZooL Soc. Lond., 18: 237-258. & P. ALDRIcn-BLArd~, 1968. Ecological and behavioral contrasts between sympatric ground dwelling primates in Ethiopia. Folia primat., 8: 192-227. HALL, K. R. L., 1965. Behaviour and ecology of the wild patas monkey, Erythrocebus patas, in Uganda. J. of Zoology, 148: 15-87. --Received July 24, 1970; Accepted September 29, 1970. Author's Address: CLAUDA. BRAMBLETT,Department of Anthropology, University of Texas, Austin, Texas 78712, U.S.A.