Introduction : A direct relation
between population density and aggression has
been proposed to explain the results of
Calhoun's (1962) classic study of crowding in
rodents. Calhoun reported that as population
density increases, aggressive interactions
increase as well. However, recent studies on
nonhuman primates do not support a simple
relation between available space and the level
of aggression. These studies avoid potentially
confounding variables, such as novelty of the
social and/or physical environment. When
well-established groups of primates are crowded
into familiar areas, they do not uniformly react
by increasing aggressive behavior. De Waal
proposed an alternative to Calhoun's so-called
density/aggression model. He argued that
gregarious primates exhibit behavioral
flexibility and social adjustment to
environmental change. This resulted in a coping
model according to which aggression may increase
somewhat under higher density but severe forms
of aggression will be limited. Recent findings
suggest that at least two coping strategies may
be used depending on the duration of the
high-density condition.
Confinement of a well-established group of
macaques into a familiar but smaller than usual
area for a limited amount of time (i.e., from a
few hours up to a few days) results in an
increase in submissive behavior and huddling
with companions and a reduction in allogrooming,
play, and exploration of the environment. This
short-term response to high population density
has been interpreted as a conflict-avoidance
strategy aimed at reducing the risk of
aggression by decreasing the overall level of
movement and social activity in an environment
in which forced proximity and limited escape
opportunities increase the likelihood of
conflict. Long-term behavioral adjustment to
high population density, on the other hand,
develops when a group lives under high-density
conditions for many years or generations. Under
such conditions primates seem to adopt a more
active tension-reduction strategy by increasing
affiliative interactions, such as
allogrooming to relieve tension due to
high density and therefore reduce the risk of
aggression.
Most studies on the effects of high
population density have been carried out on Old
World monkeys, particularly macaques. To our
knowledge, only one published study has
systematically investigated the effects of
population density on the behavior of a
well-established group of chimpanzees (Pan
troglodytes). This study examined the responses
of the Arnhem colony of chimpanzees to temporary
confinement in their indoor enclosure during
winters. This condition falls somewhere between
short- and long-term confinement since the
high-density condition lasted for about 6 months
per year. Some of the behavioral changes under
this medium-term high-density condition resemble
the long-term responses of macaques to high
density, namely an increase in allogrooming and
only a slight increase in aggression.
Thus far, no systematic investigation of the
short-term effects of high population density in
chimpanzees has been carried out. The present
study was designed to fill this gap as part of a
large-scale project that aims to compare the
behavioral adjustment of chimpanzees to
environmental change with that of other
primates. Five well-established chimpanzee
groups of various composition were periodically
confined for short periods into a familiar area.
A variety of social interactions was recorded
under both the high-density and control
conditions. We also measured the rates
of
self-grooming, self-scratching,
and yawning under the two conditions, as
these behavior patterns may index social stress
or anxiety. [...]
Discussion : This study suggests that
chimpanzees cope with a short-term increase in
population density. They reduce their overall
activity level, as suggested by a decrease in
time spent allogrooming and a lower frequency of
pant-grunt greetings. Most importantly, they
reduce their aggressive behavior along with
hooting and bluff displays that can lead to
charges against other group members. Hence, the
response of chimpanzees to the temporary
reduction of interindividual distance seems to
be an overall inhibition of activity, including
agonistic behavior. This result is quite in
contrast to that predicted by the
density/aggression model to explain increased
aggression under high-density conditions. This
difference from the "rodent pattern" is in line
with other recent studies on primates that
report the absence of a direct effect of high
social density on aggressive behavior (see
Introduction). Our study, however, goes further
than that and actually demonstrates a decrease
of agonistic behavior under high-density
conditions when several potentially confounding
variables are taken into account.
Some other studies have reported a decrease
of aggression under higher density. However, as
Erwin (1979) pointed out, factors other than
density could explain this result. For example,
the decrease in aggression among females in
pigtail macaques when they had access to only
one room of their two-roora enclosure was better
explained by the fact that the male was more
efficient in controlling female aggression when
all the females were in the same room. In our
study, all groups except one had access to more
than one room under the high-density condition,
so it is unlikely that the same explanation
applies. Interestingly, the one group (i.e.,
group B) that experienced a housing condition
similar to the pigtail macaques in Anderson
(1977) study was the only one that did not
reduce the rate of agonistic behavior under the
high-density condition (note that no instance of
bluff-charge and aggression was observed in this
group).
The present study avoided group manipulations
and unfamiliar environments to create a
high-density condition. The high-density and
control conditions did not fundamentally differ
in the cage furnishing, feeding routines, level
of human activity, and opportunities for most
chimpanzees to visually separate themselves from
others. In addition, there were no differences
in ambient temperature, levels of neighbor
vocalization, and number of females showing
maximal genital swelling between the
high-density and control conditions. Of possible
confounding variables, the only one that
presents itself is that only under the control
condition had the chimpanzees access to the
outdoor section of their runs. This could have
allowed them to spend time observing the outside
surroundings. If this difference had any impact
on their social behavior, we would have expected
a reduced rate of interaction with group members
in the control condition because more time would
have been devoted in observing the more pleasant
surroundings (instead of the concrete walls of
the inside). The opposite, however, was actually
found: social activity was higher under the
control condition. We are inclined, therefore,
to interpret the differences in social behavior
between the two conditions as a short-term
response to the different amounts of available
space.
What is the evidence that our chimpanzees
actively inhibited agonistic behavior? Baker and
Aureli found that chimpanzees housed under
various conditions (including the subjects of
the present study) were more likely to hoot,
bluff-display, bluff-charge, and initiate
aggression when neighboring groups produced loud
vocalizations or other noisy display. The
proportion of time in which minimally one
neighbor chimpanzee vocalized was not
différent under the highdensity and
control conditions (see Methods), probably
because of the large number of groups housed
within auditory reach of one another at the
Primate Center. The similarity in the level of
neighbor vocalizations between the two
conditions combined with the reduced rate of
agonistie behavior under the high-density
condition suggests that agonistic responses to
neighbor vocalizations were inhibited. Our
chimpanzees seem to actively suppress their
tendency to respond to vocalizations from other
groups in order to avoid hooting and bluff
displays that could lead to bluff charges and
aggression in an environment in which spacing
opportunities were sharply reduced. This view is
supported by a temporal analysis of the
responses to neighbor vocalizations of
chimpanzees living for long periods under two
différent conditions. The chimpanzees
living under higher density hooted or
bluff-displayed less frequently in immediate
response to neighbor vocalizations.
Further support for the notion of inhibition
of agonistic behavior under higher density comes
from a comparison of the findings of this study
with those reported by Nieuwenhuijsen and de
Waal for the Arnhem colony. The inhibition of
aggression and related behavior may perhaps
occur only under extremely high density. The
density in the high-density and control
conditions of our study was, respectively, ten
and four times higher than that of the Arnhem
chimpanzees in their indoor enclosure. In
contrast with our study, the Arnhem chimpanzees
showed an inerease of aggression under the
high-density condition. There are undoubtedly
various factors, such as group size, enclosure
complexity, and duration of the high-density
condition, which might have contributed to the
différence in the changes of agonistic
behavior under higher density between the two
studies. It is possible, however, that the large
différence in density between the two
studies may account for some of the changes.
Higher density may increase somewhat the
frequency of aggression when movements are not
too constrained by the presence of others, as
suggested by the study in Arnhem. However, when
interindividual distances are dramatically
reduced, such as under the conditions of the
present study, all social interactions,
including agonistic interactions, may be
suppressed.
The view of an actual inhibition of
aggression under extreme density conditions is
strongly supported by comparisons of aggression
rates recorded in the two studies. Using exactly
the same definition of aggression, the rates of
Arnhem chimpanzees in both the indoor quarters
and the outdoor large island were at least ten
times higher than those of Yerkes chimpanzees in
both the high-density and control
conditions.
The behavior of the adult chimpanzees of the
present study bas common féatures with
the short-term responses of macaques to higher
density. Chimpanzees seem to go a step further
than macaques, however, in the reduction of
social interactions. Whereas macaques manage to
control severe forms of aggression but generally
fail to limit mild forms, chimpanzees seem to
successfülly reduce all forms of agonistic
behavior, even behavior such as hooting that is
only a potential for direct charges or
aggression.
The conclusion that the suppression of
agonistic behavior by chimpanzees goes beyond
what bas been reported for macaques is further
supported by our findings on pant-grunt
greetings. Whereas an increase in submissive
display results from high density in macaques,
the opposite was found in the present study. The
increase in submissive display in macaques bas
been interpreted as part of a strategy to
appease dominants and avoid escalation of
aggression under circumstances in which
interindividual distances are reduced. Among
chimpanzees, such an appeasement strategy is
less needed, as aggression and bluff charges are
actually less common under higher density. The
decrease of pant-grunt greetings can be
interpreted, therefore, as a part of a general
tendency to lower overall activity, including
locomotion, to avoid potential conflict when
space is reduced.
The reduction in allogrooming is a common
short-terra response to high density for
chimpanzees (this study) and macaques discussed
the decrease in allogrooming as part of a
general decrease in overall activity. In an
environment where interindividual distances are
reduced, the risk of receiving aggression by
others while approaching potential grooming
partners increases.
Macaques also decrease activity such as play
which might accidentally lead one into proximity
with other group members. The chimpanzees of
this study, however, did not show such a
decrease. Actually, juveniles increased their
level of playing under high density. This
difference may be related to interspecific
différence in the degree of tolerance of
adults toward juveniles playing or in the
development of coping strategies, but data from
the Arnhem colony suggest that the relation
between play and density may be more
complex.
In fact, the medium-term response of the
Arnhem juveniles to higher density was a
reduction of the time spent in social play from
31% to 22%. Surprisingly, these values are
similar to those found in the present study
under the high-density condition (26%). A
possible explanation is that juvenile play may
be strongly affected by adult activity levels.
When interindividual distance is reduced and
social interactions between adults increase as
during winters at the Arnhem Zoo, juveniles may
play less often to avoid getting in the way of
potentially aggressive adults. This situation
may reach an extreme under density conditions
similar to those of our control condition in
which time spent in social play is only 14%.
However, when adult activity is dramatically
reduced, as the decrease of social interactions
under the high-density condition of the present
study suggests, juveniles can restore their play
habit by probably finding areas where there are
no stationary adults.
The effectiveness of chimpanzees in coping
with a short-term increase in density should not
lead us to conclude that the chimpanzees are
relaxed under this condition. The exuberant
pandemonium of the Arnhem chimpanzees in
response to their release from their winter
quarters each spring described by de Waal is
indicative of their preference for the larger
outdoor enclosure. Our data on rough
scratching and yawning, probably the most
reliable behavioral indicators of anxiety in
chimpanzees, support this view. The increase in
the rate of these activities under the
high-density condition strongly suggests that
social tension increases when interindividual
distances are reduced. The lower rate of gentle
scratching under the high-density condition may
be simply due to inaccuracy of data collection
when the observer had to monitor more
individuals for a rather inconspicuous behavior
(see Data Collection). Another possibility is
that gentle scratching indexes uncertainty
related to interaction between group nierabers;
its decrease under the high-density condition
might then reflect the lower level of social
interaction under this condition.
The existence of tensions under high density
suggests that conflict-avoidance strategies are
probably effective only in the short run. At
least two conflictavoidance strategies seem to
be used to cope with the short-terni effects of
high density: Macaques use an appeasement
strategy, whereas chimpanzees avoid
conflîcts through an inhibition strategy.
Both strategies, however, do not deal with the
increase in social tension, only with the
potentially negative consequences of high
density. If high density is maintained for long
periods, this situation probably cannot be
sustained, and the strategy may need to change.
The change should consist of an increase in
social interactions, such as allogrooming,
leading to an actual reduction of social
tension. Based on the limited data available,
chimpanzees seeni quicker than macaques in
adopting such a tension-reduction strategy or
affiliation strategy. In fact, the Arnhem
chimpanzees increased their frequency of
allogrooming as a medium-term. response to high
density, whereas under comparable conditions a
group of stumptail macaques (Macaca arctoides)
did not show any change in allogrooming. If the
high-density condition lasts longer, however,
macaques too adopt an affiliation strategy.
Studies on various macaque groups that have
lived under their respective conditions for many
years found higher rates of allogrooming in
groups living under the higher density.
Conclusions :
Chimpanzees suppress agonistic behavior as a
short-term response to high density.
Allogrooming and pant-grunt greetings are
reduced in adult chimpanzees as a short-term
response to high density, whereas juvenile play
increases.
Social tension increases under high
density, as reflected in elevated rates of rough
scratching and yawning.
The changes in behavior listed under
conclusions 1 and 2 may be interpreted as an
inhibition strategy that reduces the risk of
open conflict but does not eliminate social
tensions (see conclusion 3).
Deputte BL
Study of yawning in two species of
Cercopithecidae, Cercocebus albigena albigena
gray and Macaca fascicularis raffles: Research
on causal and functional factors; A
consideration of socio-bioenergetic factors.
Thèse presentée devant
l'Universite de Rennes, pour obtenir le titre de
Docteur en Troisième Cycle, February
1978.
Deputte BL
Revue sur le comportement de bâillement
chez les vertébrés Bull interne
société française pour
l'étude du comportement animal., 1,
26-35, 1974
Deputte BL,
Fontenelle A Menace et bâillement chez
Macaca Fascicularis: intérêt de
l'étude électomyographique
comparée.Biology of behaviour
1980,5,47-54
Deputte BL
Ethological study of yawning in primates
Ethology 98, 221-245, 1994
