Greeting behaviour between male baboons: oestrous females, rivalry and negotiation

Anim. Behav., 1991, 41, 49-60

Greeting behaviour between male baboons: oestrous females, rivalry and negotiation

F E R N A N D O C O L M E N A R E S MRC Unit on the Development and Integration of Behaviour, University Sub-department of Animal

Behaviour, Madingley, Cambridge, U.K. and

Departamento de Psicologia Evolutiva, Universidad A utrnoma de Madrid, 28049 Madrid, Spain*

(Received21 July 1989; initial acceptance 19 October 1989; final acceptance 18 June 1990; MS number: 3435)

Abstract. Among hamadryas baboons, Papio hamadryas, but not savanna baboons, P. anubis, P. cynocephalus, P. ursinus, the most frequent interaction observed among males, especially adults, in the context of rivalry over females is greeting. This paper investigates its possible functions. Interactions of greeting (asymmetrical, symmetrical and unreciprocated), aggression and coalitions observed between 20 males, members of a large multi-harem colony of baboons (hamadryas, cynocephalus and their hybrids) housed in the Madrid Zoo, were recorded. Males in possession of sexually mature females, and especially those in control of oestrous females (possessors), tended to initiate and receive symmetrical greetings, tended not to be greeted back when they initiated a greeting and not to greet back when they received a greeting, and tended to be frequent targets of coalition solicitations. Greetings were always given by possessors to rivals. It is suggested that, in the context of rivalry over females, greeting could be a strategy of negotiation whereby possessors and their rivals can assess the situation, influence each other's roles in the relationship, and eventually resolve conflicts without having to fight. Qualitative evidence concerning developmental careers and contexts are also used to substantiate this interpretation. Differences between male hamadryas and savanna baboons in the use of strategies directed towards rivals and possessors in the context of competition over females are also discussed.

Kummer (1968) described a type of non-agonistic interaction between hamadryas baboons, Papio hamadryas, which he called 'notifying'. It occurs fairly frequently between adult males and is a type of greeting behaviour characterized by the sequence of approach/retreat patterns used by one or both males during the interaction, the form of the locomotion (the swinging gait), and, very frequently, the absence of physical contact when the animals come close to each other (Kummer 1968; Abegglen 1984; Colmenares 1990). Kummer (1968) noted that these interactions were exchanged mainly among harem leaders and, since they tended to occur in connection with changes in the relative positions of the adult males when they initiated, and during, the daily foraging march, he suggested that they played an important role in coordinating travelling. In fact, the term notifying refers to

*Address for correspondence.

its presumed function, i.e. the signalling of the direction of the march.

In a series of cage experiments conducted on the same species to investigate why non-harem males did not try to take females that had already bonded with another male (the possessor) by attacking him (inhibition hypothesis), Kummer and his co- workers found that notifying tended to be more frequent in the triadic situation of a male that possesses a female and a rival male, most of the notifying being directed by the possessor towards his rival (Kummer et al. 1974). This has been con- firmed in further experiments (Kummer et al. 1978; Muller 1980) and under free-ranging conditions (Abegglen 1984; see also Kummer et al. 1981). The hypothesis is that notifying behaviour tends to occur in the context of rivalry over females, and that it is more frequent in the direction of leader (possessor) notifies follower (rival) than vice versa.

In contrast to the hamadryas baboon, savanna baboons ( P. anubis, P. cynocephalus and P. ursinus)

0003-3472/91/010049 + 12 $03.00/0 �9 1991 The Association for the Study of Animal Behaviour 49

50 Animal Behaviour, 41, 1

have been studied in many different localities, and consortship and competition between males over oestrous females have been particularly well docu- mented and analysed, and yet notifying between males has not been reported (P. anubis: Popp 1978, Masai Mara, Kenya; Packer 1977, 1979; Ransom 1981, Gombe, Tanzania; Strum 1982; Bercovitch 1985, 1988; Manzolillo 1986, PHG troop, Gilgil; Smuts 1985, EC troop, Gilgil, Kenya; DeVore 1965; Hall & DeVore 1965, Nairobi, Kenya; P. ursinus: Saayman 1970, 1971; Stolzt & Saayman 1970, Transvaal; Seyfarth 1978, Mountain Zebra, South Africa; P. cynoeephalus: Hausfater 1975; Noe 1986; Hausfater & Takacs 1987, Amboseli, Kenya).

In a replication of Kummer's early triadic inhibition experiments (Kummer et al. 1974), this time involving hamadryas, anubis, hamadryas-like hybrid and anubis-like hybrid males, Muller (1980) found that only the hamadryas and hamadryas-like hybrid males displayed notifying. This has also been found in wild groups of hamadryas-like hybrids (Sugawara 1979) and in a large colony of hamadryas, cynocephalus and hybrid baboons in the Madrid Zoo (Pelaez 1982; Colmenares 1990). For the only other baboon species with a harem structure, the gelada baboon, Theropithecus gelada, Kummer (1975) also reported that, in enclosure experiments in which males competed over the possession of females, they exchanged interactions that resembled the pattern of notifying observed in hamadryas.

My own study of greeting interactions, including notifying, between 20 subadult and adult males in the Madrid colony of baboons (hamadryas, cynocephalus and their hybrids) showed that a male's reproductive status class (i.e. subadult, follower, new leader, prime leader, old leader and old follower) accounts for much of the variability observed in his rate of participation in three different types of greeting interactions, i.e. asymmetrical, symmetrical and non-reciprocal. Male reproductive (and social) status is correlated with the rate of greeting behaviour; the high-ranking males in control of females (prime and new leaders) are the most frequent initiators and recipients of, in particular, symmetrical and unreciprocated greetings, and, interestingly enough, of aggression as well (Colmenares 1990). The morphology of greeting interactions, particularly of the symmetrical and unreciprocated greetings, i.e. the three-stage sequence of movement patterns (approximation,

proximity and retreat), and the absence of physical contact, suggests the existence, at the behavioural level, of a certain conflict of tendencies between approach and retreat, which might also indicate the presence of underlying emotional states of aggression and fear (Colmenares 1990). The contexts in which both aggression and many greeting episodes take place, i.e. agonistic conflicts and competition over resources, suggest that these two categories of behaviour might share some causal factors and that greeting may be regarded as a quasi-aggressive behaviour rather than as affiliative or friendly behaviour, aimed at testing a potential or actual rival's tendencies during competition (Colmenares 1990). Finally, a previous analysis of greeting interactions among males during polyadic conflicts showed that greeting fulfilled the functions of seeking and giving reassurance, and recruiting as welt as inhibiting support (on behalf of opponents) from potential interveners (Colmenares & Rivero 1986).

In this paper I use the same database on greeting interactions as before (Colmenares 1990), but I address two different questions concerning specifically: (1) the possible function of greeting as a non-aggressive strategy to resolve conflicts of interest over access to a particular type of resource (i.e. oestrous females); and (2) the possible existence of species-specific differences (i.e. hamadryas versus cynocephalus baboons) in the use of two behavioural strategies (greeting and coalitions) during episodes of competition between males over females.

M A T E R I A L S AND M E T H O D S

Colony of Baboons

The Madrid colony of baboons originally con- sisted of 26 animals of both sexes, and included 12 mature and 14 immature individuals (see Colmenares 1990). I have studied the colony since its foundation in June 1972 so I have long-term records of births/deaths, kinship relations between individuals, reproductive parameters of females, and so on. With the exception of three individuals removed in 1974 (an adult male), 1976 (a juvenile male) and 1977 (an adult female), respectively, the colony was not manipulated until November 1982, when there were 91 individually recognizable subjects. Between 1983 and 1985 chosen individuals were taken out of the colony at various times, and sometimes re-introduced after a while,

Colmenares: Baboon greeting behaviour

Table I. Definition of status classes of males

51

Features Subadult Follower New Leader Prime Leader Old Leader Old Follower

Sexually mature No Yes Yes Yes Yes Yes Reproductive No No Yes Yes Yes No Harem holder No No Yes Yes Yes No Harem size - - - - Small Large Small - - Bond with

harem females - - - - Weak Strong Strong - -

and the effects of these practices on the social organization of the colony were studied. The colony was housed in a large compound consisting of an outdoor enclosure measuring 36 m long, 26 m wide, and 7 m deep, and an attached indoor area. The outdoor enclosure was terraced and contained several climbing structures, a water-filled moat and several islands.

Sample of Males

There were 20 males in the present study. The status classes used are defined in Table I. From the outset, the Madrid colony of baboons duplicated the types of social structure reported to exist in wild troops of hamadryas, cynocephalus and hybrid baboons (e.g. Sigg et al. 1982; Sugawara 1982; Smuts 1985). It contained cohesive harem units (hamadryas), rather loose subgroups (cynocephalus), and groupings of intermediate type (hybrids; Colmenares 1987).

Data Collection and Analysis

I recorded the data used in the analyses between July 1976 through March 1985, using a focal- animal sampling technique (Altmann 1974 ). During part of the study period examined here I also used concurrently a focal-subgroup sampling technique to record all occurrences of some behaviour patterns, namely, greeting, grooming and agonism, involving any of the males of the colony. Ad libitum observations of interactions between the males were recorded over the whole study. The quantitative study presented here is based on an analysis of 72 male coalition episodes, 1583 greeting interactions and 1039 aggressive bouts involving at least two males. Behavioural observations were tabulated in terms of frequency of occurrence per hour of observation of each dyadic combination of individual males (focal-individual samples). The

data were analysed with non-parametric statistical tests, following Zar (1984).

Types of Greeting Interactions

An interaction qualified as a greeting when, during the exchange, at least one of the interactants displayed any of the following behaviour patterns towards the partner: lipsmack, ear-flattening, grimace, geck, hindquarter presentation, touching or grasping of the partner's hindquarters and/or the genitalia and mounting. A greeting interaction in which both males exchanged greeting patterns was classified as a reciprocal greeting interaction. Sometimes, however, the recipient of a greeting pattern would not respond with greeting. He would either ignore the greeting approach of the other partner, or respond with aggression. In these instances, the greeting was categorized as unreciprocated. Reciprocated greeting interactions in which both males used similar behaviour patterns (e.g. lipsmacking) were termed symmetrical, and those involving an exchange of different patterns were termed asymmetrical (e.g. lipsmack versus grimace, mount versus present). There were several types of symmetrical greeting interactions but they were all characterized by the swinging gait of at least one of the participants, the approach/retreat pattern of the movement, and, usually, the absence of physical contact (Colmenares 1990). A symmetrical greeting as defined here is therefore equivalent to what Kummer (1968) termed a notifying interaction. I describe elsewhere (Colmenares 1990) the morphology of the different types of greeting interactions.

Behavioural Roles

(1) Initiator role in asymmetrical greeting (A1): when a male lipsmacked at and/or touched/grasped/ mounted another male and received grimace and/or present from him.


(2) Recipient role in asymmetrical greeting (A2): when a male was approached by another, received grimace or present and responded by lipsmacking and/or touching/grasping/mounting him.

(3) Initiator role in asymmetrical greeting (BI): when a male initiated the interaction and either grimaced and/or presented his hindquarters towards another male.

(4) Recipient role in asymmetrical greeting (B2): when a male was approached by another male and responded with either grimace or hindquarter presentation.

(5) Initiator role in symmetrical greeting (S 1): in a symmetrical greeting interaction, when a male initiated the interaction.

(6) Recipient role in symmetrical greeting ($2): in a symmetrical greeting interaction, when a male was not the initiator of the interaction.

(7) Initiator role in unreciprocated greeting (U 1): in an unreciprocated greeting interaction, when a male initiated a greeting interaction that was not responded to by a greeting pattern by the recipient.

(8) Recipient role in unreciprocated greeting (U2): in an unreciprocated greeting interaction, when a male was the recipient of a greeting pattern by another male and did not respond with greeting.

(9) Initiator role in aggressive interaction (AD): when a male initiated an aggressive episode with another mate.

(10) Recipient role in aggressive interaction (AR): when a male was the recipient of an aggressive episode initiated by another male.

Coalitions

Coalition episodes were agonistic interactions in which at least three males were involved. In this paper I do not analyse attempted coalitions, that is, solicitations that were not supported. The male promoting a coalition alternated soliciting behaviour directed towards a non-involved male with aggressive patterns directed towards the opponent. The most typical soliciting behaviour patterns were screaming, pant-bark and uh-uh roargrunt (e.g. Hall & DeVote 1965). The aggressive patterns included short charges against the opponent, normally followed by retreats, and intense screaming with tails raised. In this study, I distinguish three roles: Solicitor, Supporter and Target. The Solici- tor is the male who starts the coalition by soliciting support against his opponent, i.e. the Target. The Supporter is the male who intervenes in the agon-

istic interaction in support ofthe Solicitor against the Target.

R E S U L T S

Greeting and Aggression Interactions The frequency distribution with which the 10

behavioural roles were filled by 11 status classes of male dyads is presented in Figs 1 and 2. I decided to focus only on New Leader and Prime Leader dyads because the majority of the aggression and greeting interactions were accounted for by these two classes of sexually mature males (Colmenares 1990) and it seemed that the questions addressed in this paper (see Introduction and Discussion) could be pro- perly examined by looking at these two classes of dyads. There are 11 possible different combinations of male status classes of dyads: New Leader/Sub- adult, New Leader/Follower, New Leader/New Leader, New Leader/Prime Leader, New Leader/ Old Leader, New Leader/Old Follower, Prime Leader/Subadult, Prime Leader/Follower, Prime Leader/Prime Leader, Prime Leader/Old Leader and Prime Leader/Old Follower. The class dyad New Leader/Prime Leader appears in both Figures for clarity.

New Leader males interacted very infrequently, or not at all, with Subadult males (Fig. 1 a-c). The relationship between these two classes of males, as assessed by the behavioural measures examined here, is therefore difficult to define. With Followers they interacted very infrequently as well, but in this dyad class there was a better defined pattern: New Leaders were aggressors (AD) more often than recipients of aggression (AR), received grimace and/or presentation (B2) more often than vice versa (BI), and initiated a greeting with lipsmack (A1) more often than they were recipients of it (A2). Their most frequent interaction with males of their own status class was aggression (Fig. 1), although they also displayed frequent symmetrical greetings among themselves (Fig. 1). Their relationship with Prime Leader males was fairly well defined. There were frequent symmetrical greetings, rather more frequently initiated by Prime Leaders ($2), and they grimaced and/or presented (B1) to, and were approached by a lipsmacking (A2) Prime Leader more often than they were recipients and initiators of these interactions (B2 and A1, respectively). Their greeting interactions with Old Leaders were relatively frequent, and, among these, the most notable feature is that New Leaders very often

Colmenares: Baboon greeting behaviour 53

50

25

20

15

I0

5

0

H

, . V k . . _

0

50 i- (b)

25

2O

15

O ~

50

25

2O

15

I0

5

o

(c)

I:.'-:.:-:.:.i

" ~ " ' 3 ,

New Leader / New Leader / Subadult Fol lower

i.........-.-.-

::!:!:!:!:i:!~

New Leader / New Leader / New Leader/ New Leader / New Leader Prime Leader Old Leader Old Follower

Figure 1. Frequency (mean class rate) with which New Leader males played 10 different roles in greeting. (a) Asym- metrical greetings: II, Initiator role (A1); Fq, Recipient role (A2); [], Initiator role (B1); [], Recipient role (B2). (b) Symmetrical and unreciprocated greetings: III, Initiator role (S1); [], Recipient role ($2); ~ , Initiator role (U1); [], Recipient role (U2). (c) Aggressive interactions:[], Initiator role (AD); [3, Recipient role (AR). Rates were calculated taking the number of possible members in each class of male dyad into account.

refused to greet back (U2) to Old Leaders' greeting approaches.

Prime Leaders frequently directed aggression (AD) at, and initiated asymmetrical greetings in which they lipsmacked, and sometimes grasped or mounted (A1), Subadult males, and received, in return, frequent presentations/grimaces (B2) from them (Fig. 2). With Followers the role differen- tiation in asymmetrical greetings was even more

pronounced than, and in the same direction (Fig. 2) as, it was with Subadult males. Prime Leaders started directing symmetrical greeting to Followers (Fig. 2), i.e. the youngest but already sexually mature males. With New Leaders their most frequent type of interaction was symmetrical greeting, which was initiated more frequently by Prime Leaders (Fig. 2). Prime Leaders had more frequent asymmetrical greeting interactions with New


Leaders than with any of the other classes of males, and the imbalance was very pronounced (Fig. 2), the former more often lipsmacking (A1) where- as the latter more often presented (B2). In their relationships with males of their own class, Prime Leaders used symmetrical greetings (S1, $2), as the most frequent type of interaction (more than aggression, AD, Fig. 2). With Old Leaders there was still some symmetrical greeting and aggression interactions, which were more often initiated by the Prime Leaders (S 1 and AD, respectively). With Old Followers there was a kind of regression in that Prime Leaders interacted with them in a way that very much resembled the one they had with the males attaining sexual maturity, i.e. the Subadults: symmetrical greetings were very infrequent (Fig. 2) and asymmetrical greeting interactions with a clear imbalance of roles were rather frequent (A 1 and B2, Fig. 2) as were aggressive episodes (AD, Fig. 2).

Prime Leader males used asymmetrical greeting less often, symmetrical greeting more often, aggression less often, and reciprocated greeting more often, than New Leaders (Figs 1 and 2). The latter's pattern of distribution of asymmetrical greetings was less well defined than that of Prime Leaders, although between these two male classes the direction of the asymmetrical greeting relationship was very well defined: New Leaders presented and Prime Leaders lipsmacked and, sometimes, grasped the former's hindquarters (Figs 1 and 2). Prime Leaders had a similar relationship, defined in terms of frequency and role in an asymmetrical greeting interaction, with three different classes of males: Subadults, Followers and Old Followers (Fig. 2). In both classes of males, the direction of the symmetrical greetings correlated very well with the reproductive status of the target male: the higher the rank (Prime Leader, Old Leader, New Leader) the more likely he would be a target of a symmetrical greeting by a Prime Leader or a New Leader (Figs 1 and 2). This correlation was particularly strong in the case of Prime Leaders (Fig. 2). The tendency not to reciprocate greetings displayed by New Leaders was also remarkable (Fig. 1). Although they were the most aggressive of all males, New Leaders were more frequent recipients of aggression from Prime Leaders and Old Leaders than vice versa (Fig. 1).

Table II shows which greeting role was most frequently displayed in each status class of male dyads, and which male class played that role most often. The roles observed in asymmetrical greetings

(roles A and B) were typical of interactions with Subadult and Follower males. Lipsmacking (role A) was typical of Prime Leaders in Prime Leader/ Subadult dyads, and grimace/present (role B) was typical of Subadults and Followers in New Leader/Subadult, New Leader/Follower and Prime Leader/Follower dyads. Initiation of symmetrical greetings was typical of New Leader/New Leader, New Leader/Prime Leader, Prime Leader/Prime Leader and Prime Leader/Old Leader dyads, and except in the New Leader/New Leader dyads, Prime Leaders were the most frequent initiators. Finally, Old Followers tended not to reciprocate greetings initiated by New Leaders and Prime Leaders, and Old Followers displayed the same tendency when New Leaders addressed greetings at them.

Greetings, Coalitions and Females

So far, I have been using male dyad class rates as the units of analysis to investigate the directionality of the interactions between different status classes of males. In this section, I conduct a more detailed study and examine the correlations between a number of variables, now using individual rates as the units of analysis. The variables analysed were: initiator and recipient roles in symmetrical greeting (S I, $2), initiator and recipient roles in unreciprocated greeting (U1, U2), Solicitor (SL), Supporter (SP), Target (T) and FF, FF +. The behavioural variables (S1-T) represent the frequency with which each particular role was filled by males. FF represents the number of different females owned by each male; and FF § represents the number of females monopolized by a given male during their period of ovulation.

Males in possession of sexually mature females, or in control of females in oestrus, tended to initiate and receive symmetrical greetings (S 1/FF: r s = 0'76; S1/FF+: rs=0"73; S2/FF: rs=0"72; S2/FF+: rs= 0-72, N= 18, P<0'001), tended not to be greeted back when they initiated a greeting and not to greet back when they received a greeting (U1/FF: rs=0"52; U1/FF§ rs=0"54; U2/FF: rs=0'58; U2/FF+: rs=0"62, N= 15, P < 0'05), and tended to be frequent targets of coalitions (T/FF: rs= 0"62, P<0"05; T/FF§ rs=0'66, N= 16, P<0"01). Also, these males tended not to be Solicitors of coalitions (SL/FF: rs=-0.35, N = l l , NS; SL/FF§ rs= --0'38, N=9, ns), although they were sometimes Supporters (SP/FF: rs=0.56; SP/FF+: rs=0.44 ,


3O

f (a ]

25

15

I

olm , - m ,n _ q _ , .

2O

15

o

3o

L (b) 25

30 Ce)

25

2O

IO

Prime Leader/ Prime Leader/ Prime Leader/ Prime Leader/ Prime Leader/ Prime Leader/ Subodult Follower New Leader Prime Leader Old Leader Old Follower

Figure 2. Frequency (mean class rate) with which Prime Leader males played 10 roles in greeting. (a) Symmetrical and unreciprocated greetings: II, Initiator role (A1); O, Recipient role (A2); k~, Initiator role (B1); [~, Recipient role (B2). (b) Symmetrical and unreciprocated greetings: IB, Initiator role (S1); O, Recipient role ($2); [], Initiator role (U1); [], Recipient role (U2). (c) Aggressive interactions: [], Initiator role (AD); [3, Recipient role (AR). Rates were calculated taking the number of possible members in each class of male dyad into account.

N = 11, NS). Males who were frequent Targets of coalitions were rarely either Solicitors or Sup- porters (SL/T: r s = - 0 ' 5 0 , Ns; SP/T: rs=--0"30, N = 11, NS).

Greetings Towards Rivals Versus Non-rivals

The rate with which selected male possessors directed symmetrical greetings towards particular

males when they were rivals and when they were non-rivals were significantly different (binomial test, P = 0-007; Table III). When a male became a potential threat to another male's control over his female(s), the possessor directed frequent greetings to him and, in many cases, a high percentage of those greetings were not responded to by the rival (Table III).


Table II. Roles most frequently played by male classes in greeting interactions

Male class Male class dyad Greeting role

Subadult Follower New Leader Prime Leader Old Leader Old Follower Prime Leader Follower Prime Leader Prime Leader Old Follower

New Leader/Subadult New Leader/Follower New Leader/New Leader New Leader/Prime Leader New Leader/Old Leader New Leader/Old Follower Prime Leader/Subadult Prime Leader/Follower Prime Leader/Prime Leader Prime Leader/Old Leader Prime Leader/Old Follower

B: grimace, present B: grimace, present S: symmetrical greeting S: symmetrical greeting U: unreciprocated greeting U: unreciprocated greeting A: lipsmack, grasp, mount B: grimace, present S: symmetrical greeting S: symmetrical greeting U: unreciprocated greeting

Table HI. Rate at which individual males directed symmetrical greetings towards other particular males when they were rivals and when they were non-rivals, and the percentage of those greetings that were not reciprocated by the target males (in parentheses)

Male Target possessor male Rival Non-rival

Principe Mokambo 0.72 (33) 0.13 (0) Chacma Meridiano 2-17 (39) 0.10 (50) Farabn Pinocho 4.02 (17) 0.05 (0) Mokambo Fara6n 6.54 (100) 0.08 (0) Yango Mokambo 2.13 (37) 0.19 (0) Meridiano Yango 4.36 (11) 0.21 (19) Pinocho Fara6n 1.53 (88) 0 (0)

D I S C U S S I O N

Male Greeting and Rivalry over Females The results of this study agree with those of

Kummer and his co-workers carried out both in enclosure and wild conditions, and support their hypothesis that greeting (i.e. notifying in their ter- minology) between adult male hamadryas baboons occurs mainly in the context of rivalry over the access to and control of females, and that it is possessors who direct greeting towards their rivals (Kummer et al. 1974, 1978; Abegglen 1984).

Species Differences in Use of Greeting Interactions between males during consortships

have been particularly well studied in savanna baboons. In these interactions, consort males tend to display intense female-directed behaviour such as grooming and herding (Saayman 1971; Hausfater 1975; Muller 1980; Ransom 1981). In the case of

rivals, their harassing behaviour also includes coalition solicitations which can favour turnovers of consorts (Hall & DeVore 1965; Saayman 1971; Hausfater 1975; Packer 1977; Popp 1978; Ransom 1981; Bercovitch 1985; Smuts 1985; Noe 1986). The rival-directed behaviour that consort males normally exhibit consists of threatening gestures and aggressive charges, which often lead to physical combat between consort males and rivals (Hall & DeVore 1965; Saayman 1971; Hausfater 1975; Packer 1977; Popp 1978; Ransom 1981; Smuts 1985). When harassed by rival males, hamadryas consort males (possessors) also herd the females being contested (Kummer et al. 1974, 1978; Muller 1980; Abegglen 1984; personal observations). This strategy is displayed especially when possessors are confronted with uninhibited rivals (Kummer et al. 1978; personal observations). When harassment is more intense, possessors also respond by threatening or attacking their rivals. Interestingly enough, in contrast to what I observed in this study, hamadryas rivals have not been reported to solicit coalitions against possessors (Kummer 1968; Abegglen 1984; Stammbach 1987). The same seems to be true of geladas, the other harem-forming baboon (Dunbar 1984). Whether this reflects species-specific differences, an emergent property of a particular type of social system (e.g. harem structure), or troop-specific characteristics is not yet clear.

However, the most important difference, and the one that has received special attention in this paper, is that hamadryas male possessors have a third strategy, greeting, which is apparently very rarely observed between savanna males and which is directed towards rivals. Why savanna males do


not use this strategy is not clear. In a study of savanna baboons, P. anubis, Manzolillo (1986) found that, during movements between troops, the rate of greetings between males tended to increase, and that these interactions tended to be initiated by immigrants (i.e. rivals?) while residents (i.e. possessors?) responded to them. So, while the contexts of occurrence share some features, i.e. competition over females, Manzolillo's findings on who initiated the majority of the greetings seem to work exactly in the opposite direction: that is, in hamadryas, possessors greet rivals (Kummer et al. 1974, 1978; this study) while in Manzolillo's troop of anubis, newcomers greeted residents. In another study of anubis baboons, Ransom (1981, page 144) wrote: ' . . . Between attempted copulations extensive interactions frequently occurred between the two males, most often initiated by the consort male but also occasionally by the follower. Typically one male approached the other to initiate a series of tense but apparently reassuring interactions such as presents �9 This description fits better with the directionality of greeting observed in hamadryas males in similar contexts.

Greeting as an Alternative to Aggression

Why do hamadryas possessor males use greeting so frequently to interact with their rivals in the context of rivalry over females? And, more specifically, why do hamadryas possessors respond with greeting to their rivals' harassment? As noted earlier, they could attack and fight their rivals but this strategy is very costly because of the risk of physical injury and of losing the female(s), either to the winner or because, especially when the possessor-female bond is weak, she runs away during the combat. The highest risk from fighting is defeat which is normally irreversible (hamadryas: Abegglen 1984; geladas: Dunbar 1984). There seem to be, therefore, good reasons to expect possessors not to choose fighting as the first strategic choice to try to settle the conflict with their rivals. An obvious, related question is why rivals do not attack possessors straight away instead of harassing them. The answer to that seems clear: if access to females can be gained without getting involved in a fight, with the risk of being injured and defeated, then a non-fighting strategy should be favoured.

Greeting as an Assessment Strategy

It seems as if both possessors and rivals are confronted with a conflict: for the former, to keep

control of their female resources, for the latter, to take the females away from their current possessors, and both would also seem to gain if they could resolve the conflict without fighting. The best option for both parties would therefore be to use behaviour that could (1) dissuade their opponent, and/or (2) enable them to assess when, and if, fighting would be the best, or the only, strategy available. Rivals, having less to lose, use various forms of harassment which seem to be designed to unnerve the consorting males (Ransom 1981; Strum 1982; Smuts 1985). In some studies of savanna baboons, cases of consorting males aban- doning their females without fighting in response to harassment by other males have been reported (Bercovitch 1985, 1988; Smuts 1985). Possessors, at least in hamadryas baboons, use greeting, and I suggest here (see below) that this non-fighting strategy is an assessment tactic which provides the possessor with information of use both to take decisions and to negotiate roles in the conflict with their rivals. It must be noted that while useful to its performers (i.e. possessors) this tactic also seems to be equally informative to the recipients (i.e. the rivals).

Regulation of Social Relationships

Studying chimpanzees, Pan troglodytes, and macaques, Macaca mulatta, de Waal and his co- workers found that, after a conflict, former male opponents tended to greet one another (de Waal & Roosmalen 1979; de Waal & van Hooff 1981: de Waal & Yoshihara 1983). He termed these post-conflict greetings reconciliations (de Waal & Roosmalen 1979), and interpreted them as strategies of conflict resolution whereby individuals can mend disturbed social relationships (de Waal 1986a, b, 1987). These reconciliation interactions were more often initiated by the loser, and the winner would respond to them only if the loser showed submission, de Waal (1986b, 1987) called this conditional reassurance; the winner accepted the reconciliation only if both agreed on the complementarity (i.e. dominance/subordination) of their roles in the relationship.

In a study of savanna baboons, Strum (1982) found that newcomers and maturing males exhibited a set of characteristic behaviour patterns, including various forms of harassment and greeting towards resident males, whose goal appeared to be both an assessment of the recipients' social power and a method of influencing them. Strum argued


that the knowledge about the characteristics of resident males that such strategies seemed to

J

provide them with would be important to assess when and with whom to compete aggressively. In another study on savanna baboons, Smuts (1985, page 149) described greeting interactions between male partners in a coalition: ' . . . Each morning as soon as the baboons began to stir, (males) AA and BZ would seek out the other for relaxed, leisurely greeting . . . . Their reciprocal greeting, in which each male took turns adopting the conciliatory role of presenter, seemed to mirror their coalition relationship, in which each took turns helping his partner to acquire females from young males'. Smuts interpreted these greetings as a method of reaffirming the relationship between coalition partners. Manzolillo (1986) also interpreted the greeting behaviour of savanna baboons which was primarily directed by newcomers towards resident males as a way of assessing individuals.

Negotiation of Roles In his early field study on hamadryas baboons,

Kummer (1968) interpreted the frequent notifying (greeting) interactions observed between males as a decision process whereby leaders came to a sort of agreement about which way to travel. This hypothesis was further substantiated by findings in Stolba's (1979) study. Enclosure experiments with hamadryas and gelada baboons showed that, in the context of rivalry over females, possessors tended to greet rivals (Kummer et al. 1974, 1978; Kummer 1975; Muller 1980). As displayed in this context, Kummer and his co-workers have interpreted greetings as behaviour that possessors direct towards males in order to establish or re-establish a friendly relationship that is regressing towards stages of greater incompatibilty (i.e. a fight).

Any consideration of the social function of greeting must take into account the fact that greeting is elicited by different social causes (i.e. it happens in different social contexts) and therefore its meaning and social effects can vary accordingly. A prelimi- nary study of some of the contexts in which greeting was observed during triadic male interactions in the Madrid colony of baboons revealed the existence of three different categories: reconciliations (post- conflict greetings among former opponents), con- solations (greetings directed at an opponent by a non-involved third party) and aid-seeking (greetings directed by an opponent at a non-involved third party; Colmenares & Rivero 1986).

During play between juvenile males, play partners sometimes get very rough. Continuation of these play episodes often depends on the ability of each to keep the level of arousal below a certain threshold. This seems to be achieved by means of exchanges of asymmetrical greetings in which, for example, the male playing the dominant role in the play game agrees to adopt momentarily the subordinate role in the greeting exchange (personal observations). During the ontogeny of the relationship between a leader and his follower it is com- mon to see the following stages. (1) When the follower approaches the leader's females he almost invariably comes up to the leader and presents his hindquarters to him. Sometimes he even has to repeat this presenting several times before he gets a greeting response from the leader. If he does not display this asymmetrical greeting he may not be allowed to stay (this is an example of conditional reassurance as defined by de Waal, for example, 1987, in chimpanzees). (2) As the follower gets older, both the initiating and the non-responding roles in greeting start to be played more often by the leader. (3) When, finally, the follower starts challenging the leader, the latter directs frequent greeting towards the follower who typically refuses to greet back (see also Kummer 1968; Dunbar & Dunbar 1975; Mori 1979). When a follower, or other class of male, takes over females, he often directs greeting towards those males from whom he most seems to fear an attack. The greetings displayed in the contexts just mentioned, together with those displayed during the coordination of travelling, suggest that processes of negotiation might be involved during those exchanges (see also Dunbar 1988). In the context of rivalry over females, the possessor seems to be trying to assess, and influence, the attitude of the rival by directing greetings at him. The rival's frequent refusal to adopt a subordinate role (for example, by not presenting his hindquarters in response), can inform the possessor that the complementarity of roles existing previously in the relationship is no longer accepted, or is going to be challenged, by the rival. On the other hand, the frequency of greetings displayed by the possessor can be used by the rival as an indicator of the former's vulnerability. It seems that both males obtain information; however, exactly what it is about remains to be determined. Future studies should concentrate on an analysis of contexts and the intra- and inter-male sequences of behaviour in which greetings take place.


Conclusion

The concept of negotiation as applied to the specifc situation examined here implies: (1) conflict of interests between the two incumbent males; (2) attempts (by one or both parties) to resolve the conflict (at least initially) by non-aggressive means; (3) behavioural and cognitive tools to assess, and to influence, the other partner's motivation and intentions; and (4) an ability to accommodate one's initial goals to those of the partner if, during the course of the interactions, that turns out to be more advantageous (see Hinde 1985, for use of the concept in a more restricted context).

In the highly complex social life of primates, conflicts between individuals are settled by using both aggressive and non-agonistic strategies. The traditional neglect of the latter has provided a rather incomplete picture of the nature of conflict resolution mechanisms and has hindered our under- standing of how they work (see also de Waal 1989). From the perspective of interactions and relationships (e.g. Kummer 1979; Hinde 1983; Smuts 1985; de Waal 1987) adopted in this paper, I have empha- sized the interpretation of behaviour patterns in relation to the social context of occurrence, the properties of the social relationship in which their performers are involved, and how they themselves perceive them. Behaviour patterns are regarded as strategies that allow individuals to assess situ- ations and make appropriate tactical decisions accordingly.

A C K N O W L E D G M E N T S

I thank Robert Hinde, Michael Simpso n and John Chadwick-Jones for encouragement and helpful comments on an earlier draft of the manuscript. During writing I was supported in Madingley, Cambridge, by a post-doctoral Fleming Research Fellowship from the British Council (UK)/M.E.C. (Spain), and in Madrid, by a post-doctoral Fellow- ship (Programa de Reincorporaci6n) from the M.E.C. (Spain). I am grateful to the Directors of the Departments (Pat Bateson, Robert Hinde and Jose Luis Linaza, respectively) for the use of their facilities. I also thank the Directorate of the Madrid Zoo for providing facilities for the maintenance of the long-term research project on the baboon colony.

REFERENCES Abegglen, J. J. 1984. On Socialization in Hamadryas

Baboons. Cranbury, New Jersey: Associated University Presses.

Altmann, J. 1974. Observational study of behavior: sampling methods. Behaviour, 49, 227-265.

Bercovitch, F. B. 1985. Reproductive tactics in adult female and adult male olive baboons. Ph.D. thesis, University of California, Los Angeles.

Bercovitch, F. B. 1988. Coalitions, cooperation, and reproductive tactics among adult male baboons. Anita. Behav., 36, 1198-1209.

Colmenares, F. 1987. Ethological study of a social system of baboons (Papio spp.): socio-spatial behaviour. Primate Eye, 32, 26-27.

Colmenares, F. 1990. Greeting behaviour in male baboons, I: communication, reciprocity and symmetry. Behaviour, 113, 81-116.

Colmenares, F. & Rivero, H. 1986. A conceptual model for analysing social interactions in baboons: a pre- liminary report. In: Quantitative Models in Ethology (Ed. by P. Colgan & R. Zayan), pp. 63-80. Toulouse: Universit~ Paul Sabatier.

Dunbar, R. I. M. 1984. Reproductive Decisions. Princeton: Princeton University Press.

Dunbar, R. I. M. 1988. Primate Social Systems. London: Croom Helm.

Dunbar, R. I. M. & Dunbar, E. P. 1975. SocialDynamics of Gelada Baboons. Basel: S. Karger.

Hall, K. R. L. & DeVore, I. 1965. Baboon social behavior. In: Primate Behavior (Ed. by I. DeVore), pp. 53-110. New York: Holt, Rinehart & Winston.

Hausfater, G. 1975. Dominance and Reproduction in Baboons (Papio cynocephalus). Basel: S. Karger.

Hausfater, G. & Takacs, D. 1987. Structure and function of hindquarter presentations in yellow baboons (Papio cynocephalus). Ethology, 74, 297-319.

Hinde, R. A. 1983. A conceptual framework. In: Primate Social Relationships (Ed. by R. A. Hinde), pp. 1-7. Oxford: Blackwell.

Hinde, R. A. 1985. Expression and negotiation. In: The Development of Expressive Behaviour: Biology- Environment Interactions (Ed. by G. Zivin), pp. 103- 116. London: Academic Press.

Kummer, H. 1968. Social Organization of Hamadryas Baboons. Basel: S. Karger.

Kummer, H. 1975. Rules of dyad and group formation among captive gelada baboons (Theropithecus gelada). In: Proceedings of the Symposia of the 5th Congress of the International Primatological Society, 1974 (Ed. by S. Kondo, M. Kawai, M. Ehara & S. Kawamura), pp. 129-159. Tokyo: Japan Science Press.

Kummer, H. 1979. On the value of social relationships to nonhuman primates: a heuristic scheme. In: Human Ethology (Ed. by M. von Cranach, K. Foppa, W. Lepenies & D. Ploog), pp. 381 395. Cambridge: Cambridge University Press.

Kummer, H., Goetz, W. & Angst, W. 1974. Triadic differ- entiation: an inhibitory process protecting pair bonds in baboons. Behaviour, 49, 62-87.

Kummer, H., Abegglen, J. J., Bachmann, C. H., Falett, J. & Sigg, H. 1978. Grooming relationships and object competition among hamadryas baboons. In: Recent


Advances in Primatology. Vol 1. Behaviour (Ed. by D. J. Chivers & J. Herbert), pp. 31-38. London: Academic Press.

Kummer, H., Banaja, A. A., Abo-Khatwa, A. N. & Ghandour, A. M. 1981. A survey of hamadryas baboons in Saudi Arabia. Fauna of Saudi Arabia, 3, 441-471.

Manzolillo, D. L. 1986. Factors affecting intertroop transfer by adult male Papio anubis. In: Primate Ontogeny, Cognition and Social Behaviour (Ed. by J. G. Else & P. C. Lee), pp. 371-380. Cambridge: Cambridge University Press.

Mori, U. 1979. Individual relationships within a unit. In: Ecological and Sociological Studies of Gelada Baboons (Ed. by M. Kawai), pp. 93-124. Basel: S. Karger.

Muller, H. 1980. Variations of social behaviour in a baboon hybrid zone ( Papio anubis • Papio hamadryas) in Ethiopia. Ph.D. thesis, Zurich University.

Noe, R. 1986. Lasting alliances among adult male savannah baboons. In: Primate Ontogeny, Cognition and Social Behaviour (Ed. by J. G. Else & P. C. Lee), pp. 381-392. Cambridge: Cambridge University Press.

Packer, C. 1977. Reciprocal altruism in olive baboons. Nature, Lond., 265, 441-443.

Packer, C. 1979. Male dominance and reproductive activity in Papio anubis. Anita. Behav., 27, 37-45.

Pelaez, F. 1982. Greeting movements among adult males in a colony of baboons, Papio hamadryas, P. cynocephalus and their hybrids. Primates, 22, 233-244.

Popp, J. L. 1978. Male baboons and evolutionary principles. Ph.D. thesis, Harvard University.

Ransom, T. W. 1981. Beach Troop of the Gombe. East Brunswick, New Jersey: Associated University Presses.

Saayman, G. S. 1970. The menstrual cycle and sexual behaviour in a troop of free-ranging chacma baboons (Papio ursinus). Folia primatol., 12, 81-110.

Saayman, G. S. 1971. Behaviour of the adult males in a troop of free-ranging chacma baboons (Papio ursinus). Foliaprimatol., 15, 26-57.

Seyfarth, R. M. 1978. Social relationships among adult male and female baboons. II. Behaviour throughout the female reproductive cycle. Behaviour, 64, 227-247.

Sigg, H., Stolba, E., Abegglen, J. J. & Dasser, V. 1982.

Life history of hamadryas baboons: physical development, infant mortality, reproductive parameters, and family relationships. Primates, 23, 473-487.

Smuts, B. B. 1985. Sex and Friendship in Baboons. New York: Aldine.

Stammbach, E. 1987. Desert, forest, and montane baboons: multilevel societies. In: Primate Societies (Ed. by B. B. Smuts, D. L. Cheney, R. M. Seyfarth, R. W. Wrangham & T. T. Struhsaker), pp. 112 120. Chicago: University of Chicago Press.

Stolba, A. 1979. Entscheidungsfindung in Verbanden yon Papio hamadryas. Ph.D. thesis, University of Zurich.

Strum, S. C. 1982. Agonistic dominance in male baboons: an alternative view. Int. J. Primatol., 3, 175~202.

Sugawara, K. 1979. Sociological study of a wild group of hybrid baboons between Papio anubis and P. hamadryas in the Awash valley, Ethiopia. Primates, 20, 21-56.

Sugawara, K. 1982. Sociological comparison between two wild groups of anubis-hamadryas hybrid baboons. African Study Monogr., 2, 73 131.

de Waal, F. B. M. 1986a. Conflict resolution in monkeys and apes. In: Primates, the Road to Self-sustaining Populations (Ed. by K. Benirschke), pp. 341-350. New York: Springer-Verlag.

de Waal, F. B. M. 1986b. The integration of dominance and social bonding in primates. Q. Rev. Biol., 61, 459-479.

de Waal, F. B. M. 1987. Dynamics of social relationships. In: Primate Societies (Ed. by B. B. Smuts, D. L. Cheny, R. M. Seyfarth, R. W. Wrangham, & T. T. Struhsaker), pp. 413-429. Chicago: Chicago University Press.

de Waal, F. B. M. 1989. Peacemaking Among Primates. Cambridge, Massachusetts: Harvard University Press.

de Waal, F. B. M. & van Hooff, J. 1981. Side-directed communication and agonistic interactions in chimpanzees. Behaviour, 77, 164-198.

de Waal, F. B. M. & van Roosmalen, A. 1979. Reconcili- ation and consolation among chimpanzees. Behav. Ecol. Sociobiol., 5, 55~6.

de Waal, F. B. M. & Yoshihara, D. 1983. Reconciliation and redirected affection in rhesus monkeys. Behaviour, 85, 225-241.

Zar, J. H. 1984. BiostatisticalAnalysis. Englewood Cliffs, New Jersey: Prentice-Hall.

Greeting behaviour between male baboons: oestrous females, rivalry and negotiation

Documents