Conflict management by hamadryas baboons (Papio hamadryas hamadryas) during crowding a tension-reduction strategy.код для вставкиСкачать
American Journal of Primatology 68:993–1006 (2006) RESEARCH ARTICLE Conflict Management by Hamadryas Baboons (Papio hamadryas hamadryas) During Crowding: A Tension-Reduction Strategy P.G. JUDGE1,2, N.S. GRIFFATON2, AND A.M. FINCKE2 1 Psychology Department, Bucknell University, Lewisburg, Pennsylvania 2 Program in Animal Behavior, Bucknell University, Lewisburg, Pennsylvania Primates change a variety of behavioral responses during short-term exposure to crowding. Under crowded conditions, rates of aggression, submissive behavior, and affiliative behavior may increase or decrease. Different patterns of change among these three categories of response have been interpreted as various coping ‘‘strategies’’ for managing the increased risk of conflict under crowded conditions. Grooming is of particular interest because this behavior is known to have a calming influence on the recipient and could be used to manage conflict under tense situations. A captive group of nine hamadryas baboons (Papio hamadryas hamadryas) composed of two harems was observed under short-term crowding to determine whether this species adopts conflict-management strategies similar to those described in other primates. The aggression, submission, affiliation, and displacement activities of the six adults in the group (two males and four females) were recorded, and behaviors in their small indoor quarters was compared with baseline behaviors in the outdoor section of their enclosure, which had over 10 times more space. Repeated-measures analysis of variance (ANOVA) was used to compare behavioral rates during crowding with baseline rates recorded immediately after crowding and during matched controls collected at the same time of day as the crowding sessions. Aggression and submission rates did not change significantly across conditions. Huddling together and proximity increased during crowding, and females increased grooming of their harem male during crowding. Displacement activities (e.g., pacing and self-grooming) increased during crowding, but scratching, an indicator of anxiety in primates, did not. The pattern of behavior exhibited by this group conforms to an active ‘‘tension-reduction’’ strategy in which animals successfully reduce the higher risk of aggression during crowding. Am. J. Primatol. 68:993–1006, 2006. c 2006 Wiley-Liss, Inc. Key words: aggression; crowding; grooming; conflict management; Papio hamadryas hamadryas Correspondence to: Peter G. Judge, Psychology Department and Program in Animal Behavior, Bucknell University, Lewisburg, PA 17837. E-mail: email@example.com Received 28 June 2005; revised 8 December 2005; revision accepted 2 January 2006 DOI 10.1002/ajp.20290 Published online 4 August 2006 in Wiley InterScience (www.interscience.wiley.com). r 2006 Wiley-Liss, Inc. 994 / Judge et al. INTRODUCTION Primates have developed behavioral mechanisms to manage conflict within their groups in order to reduce the costs of aggressive competition for resources and gain the benefits of group living. Reconciliation, in which opponents make friendly contact shortly after a fight, is one such mechanism. This response has been shown to restore disrupted relationships, reduce the likelihood of continued aggression, and reduce tension caused by the conflict (see review by Aureli et al. ). While reconciliation is an effective means of diminishing the costs of aggression after it has occurred, primates may also take proactive measures to reduce the risk of aggressive outbreaks. For example, previous studies have shown that grooming and other forms of conciliatory and tension-reducing social behavior increase in captive primate groups during the period immediately prior to a regularly scheduled feeding [de Waal, 1987; Koyama & Dunbar, 1996; Mayagoitia et al., 1993]. The authors concluded that the animals anticipated aggressive conflict and affiliated with likely competitors to manage conflict and gain tolerance during feeding. Studying primates living under various levels of crowding may also be a means of investigating whether animals modify their behavior to manage conflict. Historically, the increased competition for resources under crowded conditions has been thought to produce high rates of harmful unrestrained aggression (see review by Judge ). However, de Waal  argued that this view was oversimplified for primates because they have developed many behavioral mechanisms to manage conflicts in their groups. Instead, he proposed a ‘‘coping model’’ and suggested that primates living under crowded conditions would increase the rate at which they performed conflict-reducing responses in order to minimize the amount of aggression. de Waal  based the model on a study of a captive chimpanzee (Pan troglodytes) group that was restricted to its indoor quarters for several months each winter [Nieuwenhuijsen & de Waal, 1982]. Mild forms of aggression (e.g., stylized threats) increased under crowding, while unrestrained heavy aggression (e.g., bites and chases) did not. Submissive displays to appease potential aggressors also increased under crowding. de Waal  particularly emphasized an increase in allogrooming under crowded conditions because this behavior is known to have a calming influence on the recipient [Aureli et al., 1999; Boccia et al., 1989] and is considered to be a tensionreduction mechanism [Schino et al., 1988; Terry, 1970]. At high density, animals may groom more as a proactive means of reducing social tension. Judge and de Waal  tested the coping model on different rhesus macaque (Macaca mulatta) groups that had lived for generations under a wide range of population densities, and the results provided some support for the model. As density increased, the animals groomed each other more. In addition, females directed the increases in grooming toward nonkin females. Since the most serious aggression often occurs between matrilines in matrifocal societies, such as those of rhesus macaques [Ehardt & Bernstein, 1986; Gygax et al., 1997; Samuels & Henrickson, 1983], the authors suggested that the animals may have been strategically directing their grooming toward individuals that posed the highest risk for escalated aggression. As predicted, submissive behavior (particularly the bared-teeth display, which is a formal indicator of subordinate status in macaques [de Waal & Luttrell, 1985]) increased as density increased. Mild aggression increased with higher density, and, in contrast to the model, heavy aggression also increased among adult females. Although the animals appeared to be adopting a conflict reduction strategy through increased grooming, the tactic was not entirely effective in reducing aggression. Am. J. Primatol. DOI 10.1002/ajp Baboon Conflict Management and Crowding / 995 Similar studies conducted to assess coping responses under crowded conditions have found that animals appear to use a variety of behavioral ‘‘strategies’’ to manage conflict. A strategy can be defined as a specific pattern of change among particular behavioral categories that imply a particular function. For example, the increases in grooming and submission accompanied by no increase in heavy aggression found by Nieuwenhuijsen and de Waal  might be described as a ‘‘tension-reduction’’ strategy in which animals take proactive measures that are effective in reducing conflict. However, a ‘‘conflict avoidance’’ strategy was reported for a group of rhesus monkeys during a short-term crowding experiment in which density was manipulated by a temporary reduction of the available space [Judge & de Waal, 1993]. Mild aggression increased to threaten animals away, but there was no increase in heavy aggression. The animals avoided each other more often and increased their performance of formal submissive bared-teeth displays. They also increased huddling with their kin and remained immobile. Finally, contrary to a tension-reduction strategy, social grooming significantly decreased during crowding. The pattern of change across many behavioral categories appeared to serve the function of avoiding any interactions, and thus conflicts, during crowding. Virtually the same pattern of behavioral changes was observed in long-tail macaques (Macaca fascicularis) during a similar short-term crowding experiment [Aureli et al., 1995], which perhaps justifies the labeling of this pattern as a distinctive strategy. A somewhat different strategy was described for chimpanzees that were subjected to shortterm crowding [Aureli & de Waal, 1997]. The chimpanzees exhibited significant decreases in aggression, submission, and grooming. The pattern of change was interpreted as an ‘‘inhibition strategy’’ in which animals suppressed all behavior to reduce the possibility of conflict. The strategies mentioned above support the concept of a coping model, since the primates altered their behavior under crowded conditions in ways that may have managed conflict. The described avoidance and inhibition strategies were responses to crowding that did not reflect more proactive measures taken to reduce conflict, such as increases in grooming, that de Waal  originally proposed. de Waal suggested that the behavioral changes under high density predicted by a tension-reduction strategy might take long periods to develop and could only be observed in groups that had lived together under crowded conditions for many years, possibly even generations. During periods of shortterm crowding, animals may not have time to develop the more-active strategy and instead take immediate measures to reduce conflict by avoiding others and inhibiting interaction. Further, it may not be warranted to speak of particular unique strategies without conducting more tests on more species. Investigators may be observing idiosyncratic changes that are dependent on the species studied, the experimental conditions, and the length of crowding. The purpose of this study was to conduct a short-term crowding experiment on a captive group of hamadryas baboons (Papio hamadryas), a species that had not been tested in a crowding experiment, to determine whether they exhibited patterns of behavioral change that conformed to any of the strategies mentioned above. In addition, we examined whether the unique social system of hamadryas baboons would affect conflict management between particular partners. Hamadryas baboons live in multi-tiered societies in which the most basic social unit is a harem consisting of a male leader and his females [Kummer, 1984]. Males exert a rather tyrannical presence over their females, and offspring and females are coerced into bonding strongly with their male. Several harems associate to form clans in which males are generally intolerant of each other but cooperate to Am. J. Primatol. DOI 10.1002/ajp 996 / Judge et al. compete against other clans. Since the observed hamadryas group was composed of two harems, we tested for differential changes in behavior between males and their harem females during crowding. Previous studies of conflict management have indicated that animals strategically direct conflict reduction behavior toward the individuals with whom they are most likely to come into conflict. Since hamadryas females are likely to receive aggression from their harem males [Abegglen, 1984], we predicted that conflict-reduction behavior between these partners would increase during crowding. Finally, displacement activities, such as scratching, are positively correlated to anxiety in nonhuman primates [Maestripieri et al., 1992; Schino et al., 1996], and they have been used productively as behavioral indicators of emotion in studies of nonhuman primates [e.g., Aureli et al., 1989; Castles & Whiten, 1998; Judge & Mullen, 2005]. In the short-term crowding study of chimpanzees mentioned above [Aureli & de Waal, 1997], animals increased their rates of scratching and yawning under crowding, indicating an increase in tension despite the behavioral inhibition strategy employed. We recorded rates of displacement activities in this study to determine whether the crowding manipulation affected emotional states or the performance of conflict management responses controlled levels of social tension. MATERIALS AND METHODS Subjects and Housing The subjects of this study were a group of nine hamadryas baboons. The group was established in 1968 at Bucknell University in Lewisburg, Pennsylvania, and consisted of two adult males, four adult females, and three of their juvenile offspring (3–14 months old). The adult females were born into the group from a common ancestor. The adult males were introduced 2.5 years prior to the study. After the introduction of the males, two harems were established in which each male acquired two different adult females. Dominance/submission interactions indicated that one male was clearly dominant over the other. The group history differed from wild groups in that all females were related and were raised together since birth. In the wild, daughters typically disperse, join males from non-natal harems, and establish long-term relationships with the other females in the new harem. Thus, the females in this study may have been more strongly bonded than those in naturally-formed groups, which may have increased their tolerance levels during crowding. However, the females were fairly distantly related. The kinship classes represented among them were sister, niece, cousin, and second cousin, and there were no mother–daughter pairs. In addition, in all cases the female ancestor(s) that linked two individuals in the matriline were deceased, which further reduced familial relations. The enclosure consisted of a 9 11 4.5 m outdoor compound with adjoining 9 6 2.25 m indoor quarters. The outdoor section was constructed from concrete block and chain link fencing, and contained a gravel substrate. Perches, climbing structures, and swings were provided to promote species-typical locomotion, and enrichment devices were continually available. The indoor quarters were constructed from concrete block, chain link fencing, and a poured concrete floor. The indoor quarters consisted of three interconnected 3 6 2.5 m compartments, each of which had a doorway that led to the outdoor compound. Food and water were available ad libitum. Am. J. Primatol. DOI 10.1002/ajp Baboon Conflict Management and Crowding / 997 Data Collection Procedures Rates of behavior were assessed while the animals were in their outdoor compound (baseline condition) and compared with rates of behavior that occurred when the animals were locked into one of the three interconnecting compartments of their indoor quarters (crowding condition). Considering the twodimensional floor space available, the crowded condition represented a 5.5-fold reduction in space. Considering the three-dimensional space available, crowding represented an 11-fold reduction in space. The animals were accustomed to being locked inside for maintenance and repairs to their outdoor compound. They would also be locked into their heated indoor quarters for weeks or months at a time each winter due to cold weather. They usually had access to all three of the interconnected indoor compartments during these periods, but would frequently be locked into a single compartment for cleaning of the other two. Even though the animals typically had access to both the indoor and outdoor sections of their enclosure on a day-to-day basis, we considered the use of just the outdoor compound as a valid baseline because the group preferred the outdoors, and all members would usually stay outside when the doors to their indoor quarters were open. In addition, the animals were locked outside on a daily basis for indoor cleaning and would remain locked outside for most of the day while the indoor quarters dried. Immediately after the animals were locked inside, we conducted a 15-min habituation period with the observer standing at an indoor observation post. The observation post afforded an unrestricted view of the animals through a 1.9 1.0 m doorway constructed of 2.5-cm-wide metal bars. After habituation the observer would conduct a 10-min focal sample on each of the six adults in a randomized order. Immediately after the data were collected the animals were released back into their outdoor compound, which resulted in approximately 75 min of crowding. Eleven such crowding sessions were conducted over a 34-day period, for a total of 1.83 hr of crowding data for each animal. Crowding sessions were conducted in the early afternoon at approximately 1300 hr. Two types of baseline data were collected while the animals were locked outside in their outdoor compound: time-matched baseline and post-crowding baseline. Time-matched baseline observations were conducted in the outdoor compound at the same time of day as the crowding sessions (1300 hr). Time matching was included in the design to control for possible changes in behavior related to circadian changes in activity throughout the day. The 11 time-matched baseline sessions were typically interspersed between crowding days on an alternating basis. Exceptions occurred due to inclement weather, since we sometimes postponed scheduled baseline sessions to control for weather and temperature differences between baseline and crowding sessions. Post-crowding baseline observations were conducted immediately after crowding sessions when the animals were released into their outdoor compound. During each type of baseline session, a 10-min focal animal sample was collected on each adult animal in a randomized order. Only nine post-crowding sessions were conducted, because heavy rains occurred following two crowding sessions. The animals tended to aggregate under shelters and become immobile during heavy rain, so we considered observations under these conditions uncharacteristic of baseline levels of behavior, and possibly a confound when compared to behavior during crowding due to the enforced closeness produced by the rain. Thus, the postconflict baseline data were based on 1.5 hr of data for each focal subject. We collected both time-matched and post-crowding baseline data to conduct a more Am. J. Primatol. DOI 10.1002/ajp 998 / Judge et al. comprehensive test of the effects of crowding. If crowding alone affected behavior, we would expect similar rates of behavior during both baseline conditions. On the other hand, if being crowded influenced interactions immediately afterward, this aftereffect of crowding would produce differences in the two baseline conditions. Behavioral responses collected during focal animal samples were grouped into four general categories: aggression, submission, affiliation, and displacement. Aggression was divided into heavy aggression (bite and chase) and mild aggression (threat and rough behavior). The definitions used to score individual behavioral responses are listed in Table I. Submissive behavior included hindquarter presentation, avoidance, screams, crouching, and baring teeth. We analyzed only submissive responses that occurred outside of aggressive contexts (i.e., not following the aggressive responses scored) because these would provide an index of submission levels that was relatively independent of the levels of aggression that occurred during crowding. We analyzed hindquarter presentation separately because this response has been identified as a specialized gesture used to preempt aggression in baboons [Hausfater & Takacs, 1987], and as such may have had additional significance during crowding. Affiliative behavior included grooms, huddles, and proximity. Grooms and huddles were analyzed separately because these responses have been shown to vary independently with particular behavioral strategies associated with crowding. Proximity was analyzed as a manipulation check to verify that animals were actually crowded while locked inside, since proximity would be expected to increase under conditions of higher density. Displacement activities were recorded as another manipulation check because these responses have been shown to be indicators of anxiety and tension [Maestripieri et al., 1992]. We assumed that if animals were experiencing the subjective experience of crowding, these emotional indicators would increase during crowding. The displacement activities scored were pace, scratch, selfgroom, and yawn. Scratching was analyzed separately because this response has received the most corroboration as an anxiety indicator [Schino et al., 1996]. The other displacement activities were pooled for analysis, which is a common practice in studies of primate emotion [e.g., Castles & Whiten, 1998]. Analysis We used repeated-measures analyses of variance (ANOVAs) to test for differences in the mean hourly rates of behavior performed and received by subjects across the three observation conditions. When we analyzed the sex of partners (males with females or females with females), we conducted separate ANOVAs for each combination rather than a single two-way ANOVA (sex of partner condition) in order to maximize the small sample size for the male–female test (n 5 6) compared to the female–female test (n 5 4). Post hoc tests following significant ANOVAs were conducted using paired t-tests. All tests were conducted as two-tailed tests with the alpha level set at Po0.05. In two tests, Mauchly’s test indicated that the sphericity assumption was violated, and the Greenhouse-Geisser correction was used to assess the probability of obtaining those results. RESULTS Rates of aggression did not differ across the three conditions (F2,10 5 0.36, P 5 0.71; Fig. 1a). Too few instances of heavy aggression (bites and chases) occurred for meaningful analysis (n 5 6), but these were distributed uniformly across conditions (Fig. 1a). Since most aggression was initiated by or received Am. J. Primatol. DOI 10.1002/ajp Baboon Conflict Management and Crowding / 999 TABLE I. Definitions of Behavioral Responses Scored Response Definition I. Aggressive behavior Bite A strong grip of the skin/limb of another with the teeth which is almost always accompanied by a scream. Chase Rapid pursuit past the location the recipient maintained at the start of the interaction. Threat A variety of interactions containing any of the facial, vocal, or motor components listed below either alone or more typically in combination. Facial gestures could include mouth open wide, thrusting the head toward the target, glaring, and raised eyebrows displaying the white portions of the eyelids. Vocalizations could include loud low-pitched guttural grunts or screams. Motor components could include charges and lunges toward another animal that did not exceed the recepient’s location at the time the action began. Threats could also include manual contact. Rough behavior Mild physical contact that did not contain any of the facial or vocal components involved in a threat. Rough behavior included a variety of contacts such as grabbing, kicking, pulling, pushing, poking, slapping, pulling hair, butting, and shoving. II. Submissive behavior Avoid Walking or running more than one step from another animal within 15 seconds of an approach. Bare-teeth A facial expression in which the teeth were bared by tightly pulling back the muscles of the face. Crouch Lowering the chest and/or head to a position just above the ground by bending the forelimbs and/or hind limbs. Present Orientation of the hindquarters toward another animal, usually accompanied by lowering the forelimbs, lifting of the tail, or looking back over the shoulder. Scream A loud high-pitched vocalization. III. Affiliative behavior Groom Manipulation, brushing or licking of the fur of another animal with one or both hands for at least 15 seconds. Grooming bouts ended with interuptions of more than 15 seconds or by moving a distance of over 0.5 m. Huddle Stationary contact, other than grooming, with one or more other individuals for more than 15 seconds. Bouts ended with interruptions of more than 15 seconds or by moving a distance of over 0.5 m. Proximity Approach to within 0.5 m of another animal for at least 15 seconds without performing any other category of affiliative behavior. IV. Displacement activities Pace Sustained locomotion of at least 15 seconds in a repetitive pattern within a prescribed area. Scratch Quick repetitive touches on any part of the body with the hand or foot. Self-groom Grooming as defined above but directed toward oneself. Self-groom also included nail biting and wound licking. Bouts were punctuated by interruptions of more than 15 seconds. Yawn Opening of the jaws with a downward thrust of the lower mandible. Am. J. Primatol. DOI 10.1002/ajp 1000 / Judge et al. a. Aggression b. Submission Mild 8 3 Heavy Mean hourly rate Mean hourly rate 6 2 1 4 2 0 0 Baseline Baseline Crowding Post-crowding c. Hindquarter presentation Crowding Post-crowding d. Grooming 5 5 Alpha male 4 Others Mean hourly rate Mean hourly rate 4 3 2 1 2 1 0 0 Baseline Baseline Crowding Post-crowding e. Huddling f. Proximity 20 40 15 30 Mean hourly rate Mean hourly rate 3 10 5 Crowding Post-crowding 20 10 0 0 Baseline Crowding Post-crowding Baseline Crowding Post-crowding Fig. 1. Mean hourly rates (1SE) of the six categories of social behavior recorded under the three observation conditions. For aggression the proportion of heavy vs. mild aggression is indicated. For hindquarter presentation the proportion with the alpha male vs. other group members is indicated. Am. J. Primatol. DOI 10.1002/ajp Baboon Conflict Management and Crowding / 1001 from the alpha male (82.1%: 46 of 56 instances), the lack of change across conditions occurred because the alpha male did not react to crowding with aggression. Most submissive responses (avoid, scream, crouch, and bare teeth) involved cases in which one animal avoided another (92%: 113 of 123), and submissive responses did not differ across conditions (F2,10 5 1.74, P 5 0.22; Fig. 1b). The rate at which animals presented their hindquarters to other animals increased during crowding, although the result was not statistically significant (F2,10 5 2.25, P 5 0.16; Fig. 1c). Further, most of these hindquarter presentations were made to the alpha male (91%: 58 of 64; Fig. 1c). Grooming increased during crowding relative to both baseline conditions (Fig. 1d), but the difference across conditions was not statistically significant (F2,10 5 2.11, P 5 0.24). However, an analysis of the sex of grooming partners indicated that grooming changed across conditions between males and females (F2,10 5 5.64, P 5 0.023) and not between females (F2,6 5 0.49, P 5 0.63; Fig. 2). Male–female grooming rates were significantly greater during crowding than during baseline conditions (t(5) 5 3.06, P 5 0.03), but none of the other conditions differed significantly from each other (baseline/post-crowding: t(5) 5 2.44, P 5 0.06; crowding/post-crowding: t(5) 5 1.23, P 5 0.28). Since females initiated all male–female grooming and directed almost all of that behavior toward their harem male (96%: 48 of 50 cases), the increase in grooming during crowding was the result of the females’ increased grooming of their harem male. It is interesting to note that the two exceptions of females grooming their own male were cases in which a female in the harem of the lower-ranking male groomed the alpha male during crowding. Huddling significantly differed across crowding conditions (F2,10 5 17.42, P 5 0.001), with more huddling occurring during crowding (Fig. 1e). Partner analyses indicated that the adult males did not huddle with each other throughout the study, and when they huddled with females they only huddled with those in their harems. Although there was a general increase in huddling between males and females during crowding (Fig. 3), the change across conditions was not statistically significant (F2,10 5 1.64, P 5 0.24). Differences in huddling between adult females differed significantly across conditions (F2,6 5 15.40, P 5 0.03), with significantly more huddling occurring during crowding than Mean hourly rate per partner 1.5 1 Female-female Female-male 0.5 0 Baseline Crowding Post-crowding Fig. 2. Mean hourly rates per partner (1SE) of grooming under the three observation conditions, indicating rates of female–female and female–male grooming. Am. J. Primatol. DOI 10.1002/ajp 1002 / Judge et al. during baseline or post-crowding, and no difference between baseline and postcrowding (crowding/baseline: t(3) 5 3.15, P 5 0.05; crowding/post-crowding: t(3) 5 5.22, P 5 0.02; baseline/post-crowding: t(3) 5 2.29, P 5 0.11; Fig. 3). The rate at which animals came into proximity to each other differed significantly across crowding conditions (F2,10 5 21.24, Po0.001; Fig. 1f), with significantly more proximity during crowding than during baseline or post-crowding, and no difference between baseline and post-crowding (crowding/baseline: t(5) 5 5.30, P 5 0.004; crowding/post-crowding: t(5) 5 4.58, P 5 0.006; baseline/post-crowding: t(5) 5 0.89, P 5 0.42). Thus, the spatial manipulation did appear to produce the desired crowding effect. Scratching appeared to be unaffected by crowding (F2,10 5 0.74, P 5 0.50; Fig. 4a). However, the pooled rate of the other displacement activities (pace, selfgroom, and yawn) significantly differed across conditions (F2,10 5 11.94, P 5 0.003), with significantly more displacement activities occurring during crowding than during baseline or post-crowding, and no difference between baseline and post-crowding (crowding/baseline: t(5) 5 4.57, P 5 0.006; crowding/ post-crowding: t(5) 5 3.43, P 5 0.02; baseline/post-crowding: t(5) 5 0.94, P 5 0.39; Fig. 4b). Mean hourly rate per partner 2.5 2 1.5 Female-female 1 Female-male 0.5 0 Baseline Crowding Post-crowding Fig. 3. Mean hourly rates per partner (1SE) of huddling under the three observation conditions, indicating rates of female–female and female–male huddling. b. Displacement activities a. Scratch 30 8 6 Mean hourly rate Mean hourly rate 25 20 15 10 4 2 5 0 0 Baseline Crowding Post-crowding Baseline Crowding Post-crowding Fig. 4. Mean hourly rates (1SE) of (a) scratching and (b) other displacement activities under the three observation conditions. Am. J. Primatol. DOI 10.1002/ajp Baboon Conflict Management and Crowding / 1003 DISCUSSION In contrast to similar studies on macaques and baboons [Alexander & Roth, 1971; Aureli et al., 1995; Elton & Anderson, 1977; Judge & de Waal, 1993], the hamadryas baboons in this study showed no difference in aggression across crowding conditions. If anything, inspection of the means indicated a nonsignificant decrease in absolute levels of aggression during crowding (Fig. 1a). In addition, there was no change in levels of general submissive responses (e.g., avoid), which have often been shown to increase during short-term crowding [Alexander & Roth, 1971; Elton & Anderson, 1977; Judge & de Waal, 1993; Nieuwenhuijsen & de Waal, 1982]. Hindquarter presentation, a specialized gesture used to preempt aggression in baboons [Hausfater & Takacs, 1987], did not change significantly across conditions (P 5 0.16), but the frequency with which animals directed this display toward the alpha male nearly doubled during crowding (Fig. 1c). A similar increase in a formal submissive display, the bared-teeth display of macaques, has been observed during crowding in rhesus macaques [Judge & de Waal, 1993], and females have been shown to selectively increase the response to adult males at higher density [Judge & de Waal, 1997]. Crowding appeared to have little influence on aggressive and submissive behaviors, and perhaps motivated the animals to formally indicate their submissive status to the alpha male. Females increased huddling with other females during crowding–a response observed in similar studies [Aureli et al., 1995; Judge & de Waal, 1993]. Such increases in friendly behavior during tense situations may help relieve social tension [de Waal, 1989]. Perhaps most significantly, adult females groomed their harem males more often during crowding. Grooming is used to reduce social tension in primate groups [Schino et al., 1988; Terry, 1970] and it has a calming influence on the recipient [Aureli et al., 1999; Boccia et al., 1989]. Further, male harem leaders are the most aggressive adults in hamadryas baboon groups, particularly in captivity [Kummer & Kurt, 1965], and they are despotic rulers that are likely to aggressively herd their females and punish them with speciestypical neck bites [Abegglen, 1984]. By placating their leader males with grooming, females may reduce the increased possibility of aggression from males during crowding. Calming the males may also reduce the chances of the males fighting with each other, in effect mediating potential conflict that might have negative repercussions for the females themselves. As in previous studies of conflict management [de Waal, 1987; Judge & de Waal, 1997; Koyama & Dunbar, 1996; Mayagoitia et al., 1993], the animals appeared to strategically direct affiliative behavior toward the animals with which they were most likely to come into conflict. The overall pattern of change observed across all behavior categories could be interpreted as a ‘‘tension-reduction’’ strategy similar to the proactive conflict reduction strategy originally proposed by de Waal  but never observed in its entirety. The animals may have increased grooming to calm others, and increased friendly affiliative behavior to reduce social tension. The result was to manage aggression and keep it at the same levels as in noncrowded conditions. Not only is this study the first to observe some of the behavioral changes predicted to occur with the use of a proactive conflict-reduction strategy, it is also unique because the changes occurred during a short-term crowding study. Although de Waal  suggested that animals might have to live for long periods at high density before they develop Am. J. Primatol. DOI 10.1002/ajp 1004 / Judge et al. this style of behavior, the baboons in this study adopted the behavioral pattern during brief bouts of crowding. Perhaps the predictions also apply to short-term situations, or the animals were so accustomed to being locked into their indoor quarters over the years that they learned to utilize the strategy quickly and opportunistically. An alternate explanation for the lack of change in aggression could be that the 11-fold reduction in space produced by the crowding manipulation was not extreme enough to produce the increases in aggression observed in similar studies. The lack of increase in scratching rates during crowding may lend support to this explanation because scratching is an indicator of emotional arousal in nonhuman primates [Maestripieri et al., 1992; Schino et al., 1996], and one could argue that if scratching did not increase, the manipulation did not produce the subjective experience of crowding. On the other hand, the number of times the animals came into proximity significantly increased during crowding, indicating that they were physically closer together. Further, displacement activities such as yawning, self-grooming, and pacing increased during crowding, indicating some form of emotional reaction to the manipulation. Finally, one could also argue that scratching rates remained low during crowding because the behavioral mechanisms employed by the animals during crowding were effective enough to reduce social tension. Another indication that the crowding manipulation was valid is that rates of social behavior during crowding were always either slightly or markedly different from those during the two different types of baseline observations, which were always remarkably similar to each other (Fig. 1). If the crowding manipulation did not have an influence, and we were effectively sampling baseline behavior during all three conditions, one would not expect such consistency with the behavioral rate during crowding always being the unique extreme compared to the baselines. The close similarity of the two baseline conditions also indicates that there was no ‘‘aftereffect of crowding’’ such that the animals would have behaved differently during the post-crowding condition compared to the time-matched baseline on days when they were not crowded. The species-typical behavior of hamadryas baboons should also be considered in interpretations of the current results. Harems are closely knit units in which the females are strongly bonded to their males. Females are typically coerced into staying in close proximity to their male [Abegglen, 1984], thus producing what might be considered enforced crowding around him. The tension-reduction strategy that was adopted during short-term crowding could be similar to behavioral modifications the females have developed during tense situations near their males. The tendencies of hamadryas to gather at night in large groups on the face of a sleeping cliff, and for harems to share small outcroppings [Kummer, 1984] may also predispose this species to make rapid behavioral adjustments to temporary crowding that are effective in controlling or inhibiting aggression. Nevertheless, the results support the concept of the coping model [de Waal, 1989], which predicts that primates will develop behavioral responses to manage conflict in their groups, and increase performance of the responses when they are at risk for increased aggression. In addition, most conflict research and studies of the mechanisms developed by primates to control aggression and live in cohesive social groups have focused on events following aggression. The results of this study suggest that animals anticipate aggressive disruptions and use conflictreduction behavior routinely outside of aggressive contexts to manage aggression before it begins. Am. J. Primatol. DOI 10.1002/ajp Baboon Conflict Management and Crowding / 1005 ACKNOWLEDGMENTS We thank Mary Gavitt for providing technical assistance. The primate facility was supported by Bucknell University. REFERENCES Abegglen J-J. 1984. On socialization in hamadryas baboons. London: Associated University Press. 207p. Alexander BK, Roth EM. 1971. The effects of acute crowding on aggressive behavior of Japanese monkeys. Behaviour 39: 73–89. Aureli F, van Schaik CP, van Hooff JARAM. 1989. Functional aspects of reconciliation among captive long-tailed macaques (Macaca fascicularis). Am J Primatol 19: 39–51. Aureli F, van Panthaleon van Eck CJ, Veenema HC. 1995. Long-tailed macaques avoid conflicts during short-term crowding. Aggress Behav 21:113–122. Aureli F, de Waal FBM. 1997. Inhibition of social behavior in chimpanzees under highdensity conditions. Am J Primatol 41: 213–228. Aureli F, Preston SD, de Waal FBM. 1999. Heart rate responses to social interactions in free-moving rhesus macaques (Macaca mulatta): a pilot study. J Comp Psychol 113: 59–65. Aureli F, Cords M, van Schaik CP. 2002. Conflict resolution following aggression: a predictive framework. Anim Behav 64: 325–343. Boccia ML, Reite M, Laudenslager M. 1989. On the physiology of grooming in a pigtail macaque. Physiol Behav 45:667–670. Castles DL, Whiten A. 1998. Post-conflict behaviour of wild olive baboons. II. Stress and self-directed behaviour. Ethology 104: 148–160. de Waal FBM. 1987. Tension regulation and nonreproductive functions of sex in captive bonobos (Pan paniscus). Nat Geogr Res 3: 318–335. de Waal FBM. 1989. The myth of a simple relation between space and aggression in captive primates. Zoo Biol Suppl 1: 141–148. de Waal FBM, Luttrell LM. 1985. The formal hierarchy of rhesus monkeys: an investigation of the bared-teeth display. Am J Primatol 9:73–85. Ehardt CA, Bernstein IS. 1986. Matrilineal overthrows in rhesus monkey groups. Int J Primatol 7:157–181. Elton RH, Anderson BV. 1977. The social behavior of a group of baboons (Papio anubis) under artificial crowding. Primates 18:225–234. Gygax L, Harley N, Kummer H. 1997. A matrilineal overthrow with destructive aggression in Macaca fascicularis. Primates 38:149–158. Hausfater G, Takacs D. 1987. Structure and function of hindquarter presentations in yellow baboons (Papio cynocephalus). Ethology 74:297–319. Judge PG, de Waal FBM. 1993. Conflict avoidance among rhesus monkeys: coping with short-term crowding. Anim Behav 46: 221–232. Judge PG, de Waal FBM. 1997. Rhesus monkey behaviour under diverse population densities: Coping with long-term crowding. Anim Behav 54:643–662. Judge PG. 2000. Coping with crowded conditions. In: Aureli F, de Waal FBM, editors. Natural conflict resolution. Berkeley, CA: University of California Press. p 129–154. Judge PG, Mullen SH. 2005. Quadratic postconflict affiliation among bystanders in a hamadryas baboon group. Anim Behav 69: 1345–1355. Koyama NF, Dunbar RIM. 1996. Anticipation of conflict by chimpanzees. Primates 37: 79–86. Kummer H, Kurt F. 1965. A comparison of social behavior in captive and wild hamadryas baboons. In: Vagtborg H, editor. The baboon in medical research. Austin, TX: University of Texas Press. p 65–80. Kummer H. 1984. From laboratory to desert and back: a social system of hamadryas baboons. Anim Behav 32:965–971. Maestripieri D, Schino G, Aureli F, Troisi A. 1992. A modest proposal: displacement activities as an indicator of emotions in primates. Anim Behav 44:967–979. Mayagoitia L, Santillan-Doherty AM, Lopez-Vergara L, Mondragon-Caballos R. 1993. Affiliation tactics prior to a period of competition in captive groups of stumptail macaques. Ethol Ecol Evol 5: 435–446. Nieuwenhuijsen K, de Waal FBM. 1982. Effects of spatial crowding on social behavior in a chimpanzee colony. Zoo Biol 1: 5–28. Am. J. Primatol. DOI 10.1002/ajp 1006 / Judge et al. Samuels A, Henrickson RV. 1983. Outbreak of severe aggression in captive Macaca mulatta. Am J Primatol 5:277–281. Schino G, Scucchi S, Maestripieri D, Turillazzi PG. 1988. Allogrooming as a tension-reduction mechanism: a behavioral approach. Am J Primatol 16:43–50. Am. J. Primatol. DOI 10.1002/ajp Schino G, Perretta G, Taglioni AM, Monaco V, Troisi A. 1996. Primate displacement activities as an ethopharmacological model of anxiety. Anxiety 2:186–191. Terry R. 1970. Primate grooming as a tension reduction mechanism. J Psychol 76: 129–136.