GROUP
ENCOUNTERS
AGONISM,
IN WILD
GIBBONS
AND THE
AFFILIATION,
INFANTICIDE
(HYLOBATES
LAR):
CONCEPT
OF
by
ULRICH REICHARD1) and VOLKER SOMMER 2,3,4)
(1Center for Conservation Biology, Mahidol University, Rama VI Road, Bangkok 10400,
Thailand; 2Department of Anthropology, University College London, Gower Street,
London WC1E 6BT, England, UK)
(Acc. 30-VI-1997)
Summary
1. Gibbons are the least studied apes and traditionally thought to live in nuclear families of
2-6 individuals including a pair of breeding adults who maintain lifelong, sexually monogamous relationships and vigorously defend territories against neighbours. The present paper
challenges this view.
2. During a long-term study on white-handed gibbons (Hylobates lar) in Thailand's Khao
Yai rainforest, 162 encounters were recorded between 3 habituated and 8 non-habituated
groups. Encounters characterized 9% of the activity day (mean 8.31 h). Those between
habituated groups were nine times more frequent (0.7/day) and lasted significantly longer
(median 70 min) than encounters with non-habituated neighbours (median 38 min). It
was also found that gibbon group home ranges (mean 24 ha) overlap extensively (64%)
with neighbours', all of whom were met. However, most previous studies centered on
single groups surrounded by unhabituated neighbours. This produced underreporting of the
importance of inter-group interactions, particularly the affiliative aspects observed presently.
3. Encounters included many elements which seem to have a 'defensive' function. Chases
occurred during 61%, contact aggression during 8-9%; each adult and subadult male chased
3)
Corresponding author, address: Department of Anthropology, University College London,
Gower Street, London WC1E 6BT, UK.
4) We thank Warren Y. Brockelman, Center for Conservation
Biology, Mahidol University,
Bangkok, for support and encouragement. The National Park Division of the Royal Forestry
Department and the National Research Council of Thailand, Bangkok, kindly granted research permissions. Fieldwork and data analysis was sponsored by the German Academic
Exchange Service (UR) and a Heisenberg-Fellowship of the German Research Council (VS;
So 218-3/1). We thank Guy Cowlishaw and the anonymous reviewer Carel van Schaik for
valuable comments.
1136
all others and was chased by all others. Moreover, encounters occurred often in or near
food trees and rates peaked during the early morning when ripe fruit were most abundant.
However, a seasonal correlation between the rate of encounters and (crude) measures of
resource availability could not be detected.
4. The study indicates that gibbon groups are structured by female resource-defense and
male mate-defense because adult females led 76% of all travels into and out of fruit trees,
whereas males moved to the front as soon as neighbours were encountered. Male-male
interactions constituted 90% of all inter-group chases. This pattern is compatible with the
idea that conflicts over food sources (which can be shared) will rarely provoke contact
aggression. Control of mate access, on the other hand, has a much higher relative value
for males. These explanations are more parsimonious than the traditional concepts of
'pairbonding' and 'territoriality'.
5. Contrary to earlier assumptions, encounters were nevertheless not always dominated by
aggression. Groups fed, traveled or rested together for prolonged times (35% of encounters).
Inter-group play between immatures was likewise common (21% of encounters; 55% dyadic,
45% triadic and quartetic play). Same-aged play partners are not present in a gibbon's natal
group. However, gibbon youngsters clearly preferred age-mates during inter-group play
which may indicate that play is ontogenetically important.
6. The adult female of one group copulated not only with her pair-mate, but also with
two neighbouring males. The overall proportion of extra-pair-copulations (EPC) was 12%
and they occurred during a period when the female conceived. Encounters thus provide
opportunities for alternative mating strategies. However, philandering males are faced with
the dilemma to lower their mate guard which creates a risk of cuckoldry for themselves.
This could explain why subadults are often tolerated in natal groups beyond sexual maturity,
because they assisted the resident adult male during agonistic encounters. Moreover, females
gave solo great calls during a quarter of all encounters. These calls increase the costs of
philandering for the paired male (who cannot answer without giving away that his female
is unguarded) and may at the same time attract neighbouring males.
7. For the first time, close spatial proximity and body contact between intruding adult males
and infants of neighbouring groups are reported, including play (during 6% of all inter-group
play sessions). Immatures were at times relaxed but at other times frightened in the presence
of neighbouring males. A near-zero mortality of infants at Khao Yai shows that infanticide
is absent or at least not a regular occurrence. EPCs and a long period of pre-conception
copulations could be strategies of females to confuse paternity and forestall infanticide. In
any case, the non-monogamous mating pattern makes it likely that kin-relationships extend
well into neighbouring groups. A reduced level of inter-group competition and aggression
is therefore not surprising.
Keywords: gibbon, territoriality, monogamy, infanticide.
Introduction
Aggression between groups is seen as a major selective factor in human
evolution (Aiello & Dunbar, 1993). Nevertheless,
studies of
behavioural
1137
on social interactions within groups. Thus,
primates focus overwhelmingly
the causes for friendly or agonistic group interactions are poorly understood (review in Cheney, 1987). A taxon traditionally seen as a model for
aggressive territorial group encounters are the gibbons, the least studied
Hominoidea.
Our data illustrate
that the concept of territorial defense as
encounters is too narrow.
We also dis-
the driving force behind group
cuss evidence for the theory that protection from infanticide fostered the
evolution of monogamy in gibbons (van Schaik & Dunbar, 1990) and perhaps permanent male-female associations amongst primates in general (van
Schaik & Kappeler, 1993).
Gibbons
are arboreal apes confined to South-East Asia,
(Hylobatidae)
Northwest-India
and Bangladesh (overviews on ecology and behaviour in
Preuschoft et al., 1984; Tuttle, 1986; Leighton, 1987). They live in groups
of usually 2-6 individuals
which mostly contain a single adult female, a
Gibbon females give birth about every
single adult male and immatures.
third year. Once an infant becomes independent
with about 2 years, females may cycle several months before a new conception.
Until recently,
were thought to form permanent pairs (e.g. Mc Cann, 1933; CarTenaza & Hamilton,
1940;
1971; Ellefson, 1974; Tilson, 1979;
penter,
Brockelman
& Srikosamatara,
1984). However, this view of mandatory
nuclear families with a socially and sexually monogamous
pair has been
of extra-pair copulations and partner changes
challenged by observations
gibbons
1994a, b; Reichard,
1995). Gibbon groups occupy rela(e.g. Palombit,
tively stable home ranges including an exclusively used core area which
many researchers consider to be a territory. Suitable habitats are usually
saturated with groups (e.g. Mitani, 1990), creating a tight mosaic pattern
of closely interlocking home ranges. Edges overlap and meetings between
neighbours are common.
Traditionally,
group relations in gibbons are thought to deal only with
territorial defense. However, actual knowledge about group encounters is
scarce, for several reasons. (a) Groups often flee silently from observers
which is facilitated by their small group size and arboreality. (b) Recording
complex events taking place high up in the canopy is difficult. (c) Most earlier observers habituated at best a single group. (d) Life-histories
and thus
kin-relations
of individuals
in adjacent groups are unknown.
We solved
these problems at least partly because our data derive from the longest
1138
field study on hylobatids,
spanning about 17 years and involving several
well-habituated
groups (cf. Brockelman et al., in press).
neighbouring
Material and methods
The study site was located at 730-870 m elevation inside the primary rain forest of the
Khao Yai National Park, Thailand (2,168 km2; 101°22' E, 14°26' N; 130 aerial km NE of
Bangkok). A population of white-handed gibbons (Hylobates lar) immediately west of park
headquarters in the Mo Singto area has been studied by various observers since 1979 (e.g.
Raemaekers & Raemaekers, 1984a, b; Treesucon, 1984; Brockelman, 1985; Whitington,
1990; Reichard, 1996). About half a dozen trees in the study area were illegally exploited for
an incense raw-product over the last decade, but otherwise the gibbons were not threatened
by poaching or forest destruction.
The authors collected data during the following periods; UR: Oct89-Jan90 (55 days,
150 h), Jan92-May93 (259 days, - 1600 h), Oct93 (N 15 days), Oct94 (N 15 days),
days); VS: Sep94-Ju196 (110 days, ?550 h). Observations focused on
May-Jun96
three adjacent gibbon groups (A, B, C) whereas neighbours were studied opportunistically
during encounters with the main study groups and surveys (Fig. 1). All these gibbons were
individually identifiable. Qualitative information presented in this paper may originate from
any of these observation periods. However, quantitative analysis is restricted to Jan92May93 when UR observed groups A, B and C during 250 days for a total of 1537 h. The
core data set stems from a one-year period of focal-animal sampling (Altmann, 1974) on
groups A, B, C (May92-Apr93, 175 days, 1209 h; see below, Table 1 ). Datawere collected
using a battery-powered, hand-held-size computer (Psion Organiser LZ 64) connected to
a bar-code reader pen. Codes were displayed on a score-board which was carried by the
observer in front of the belly (cf. Reichard, 1996). Fast sequences were at times tape
recorded. Most observation days started at dawn and ended around 16:30 after the apes
had settled in their sleeping trees. Data on group encounters were recorded ad libitum. An
encounter was defined to start when a group approached another within 50 m and it ended
when groups had again distanced themselves further away than 50 m.
Fig. 1. Home ranges of main study groups (A, B, C) and neighbours, Khao Yai National
Park, Thailand (1992-93).
1139
1140
1141
Table 2 lists the composition of the main study groups which were well habituated to
gibbon researchers (group A since about 1981, groups B and C since May92). Gibbons
are arboreal apes. Nevertheless, they do on rare occasions descend to the ground (e.g. VS
observed about half a dozen ground contacts of usually less then I min). However, they
sometimes traveled in heights of less then 5 m right above observers. Visibility into the
canopy was variable but often very good, particularely during the dry winter. A marked and
measured trail system, often along elephant paths, transversed the study area. Day ranges
of gibbon groups were reconstructed on maps with grid cells of 50 x 50 m.
Gibbons are peculiar amongst Hominoidea because of their vocalizations (review in
Haimoff, 1984; Raemaekers et al., 1984; Cowlishaw, 1992) which can be heard from
1-2 km away. Males and females sing loud duets, mostly from 07:00-11:00. Duets consist
of a warm-up phase by both individuals, followed by a high-pitched female 'great call'
and a male 'reply/coda'; the sequence may or may not be repeated many times. Male
gibbons also sing 'solos', mostly around dawn. A typical call during encounters is the
'male encounter vocalization', a series of low-key hoots. Other observers termed the calls
'conflict vocalization' (Brockelman & Srikosamatara, 1984) or 'conflict hoo(t)s' (Ellefson,
1974; Gittins, 1980). A more neutral term is preferred in this paper to avoid implications
about the calls' function.
Results
Home range
The following
use and group cohesion
section describes
general features of group life to provide the
which the role of encounters has to be viewed.
necessary backdrop against
Home ranges of groups A, B and C as measured from Aug-Dec92 (Neuareas bedenberger, 1993) comprised on average 24 ha. Non-overlapping
tween the 3 groups constituted 71 % (group A, 64.3% of 24.5 ha; group B,
74.4%
of 21.5 ha; group C, 73.6% of 26.5 ha). However, all groups had
and overlap existed with all of them. The length of borders
amongst groups A, B and C and the proportion of overlap was used to calculate the expected overlap with other neighbours, yielding a mean of 63.7%.
5-6 neighbours
only about one third of a group's range was exclusively used (cf.
use of overlap areas occurred,
Fig. 1). Both successive and simultaneous
the latter of which being the basis for encounters.
Adult females tended
Thus,
to enter and leave food sources
as the first group member whereas adult
males tended to travel behind them (Table 3; binomial test, zenter = -6.83,
= 39, p < 0.001). This progression
x = 15, p < 0.001, zleave = -3.67, x
order clearly contrasts the pattern during group encounters (male in front,
female in the background;
see below). The decision to treat 'groups' as
1142
'units' that encounter each other is justifiable because members maintained
close spatial proximity throughout the day. (An exception was SMC2 who
sometimes left for days to unknown places. He was then considered as
not being with the group.) The median distance to the nearest neighbour
arms' reach and 10 m (Fig. 2). Distances between the female and infant in group B, which was carried for most of the time were
excluded from the sample. Only during 0.8% of the time was any group
was between
(nearly always a subadult or adult male) found to be further away
than 50 m from others. But such periods of peripherization
were usually
short (< 45 min). In almost all cases did groups reunite if neighbours were
member
encountered.
Fig. 2. Intra-group spatial distributions. Based on distances to the nearest neighbour of focal
animals sampled in 15-min intervals, with equal distribution over daytime and individuals.
TABLE 3. Progression
orders
of adults while entering
trees
and leaving fruit
1143
Frequency
and duration
of group encounters
of group encounters, 86% were observed
Out of 21 possible combinations
(Fig. 3). The habituated main study groups A, B, C met all their neighbours. It is thus very likely that every group encountered
all neighbours
and that the few non-recorded
combinations
were a mere consequence
of
too limited observation
time. Encounters
between
the habituated
groups accounted for 77% of all 162 recorded meetings. Habituated groups met significantly more frequently with each other (every 11.l hours) than with nonhabituated neighbours
U-test; U = 0.01;
(every 100.0 h; Mann-Whitney
NI,2 = 8 ; p < 0.001). Meetings of group A with habituated groups lasted
significantly longer than those with non-habituated
groups (Mann-Whitney
U-test; U = 251.5, Nj = 71, N2 = 14, p < 0.01; Table 4). The overall
median of complete encounters was 70 min (Fig. 4). Encounter durations
did not differ between the study groups (Kruskal-Wallis
one-way analysis
of variance, H = 4.161; Ni = 36, N2 = 35, N3 = 14; p > 0.05). Clearly,
detailed information
on inter-group
relations can only be deduced from
meetings between individuals
further analysis is restricted.
tolerant
to observers,
to which most of the
Encounters
usually followed if two groups were aware of each other
distances of 100-150 m. In only 10% of all cases when groups
had come within 50 m did they not approach further below 20 m. En-
within
counters
occurred
every
1.7 days during
98 full-day
follows
of the main
Fig. 3. Encounters between 11gibbon groups at Khao Yai (N = 162; Jan92-May93).
1144
TABLE 4. Duration
of encounters
of group A
Fig. 4. Durations of 84 encounters observed from onset to end in main study groups A,
B, C. 71 encounters were between A, B, and C and 14 encounters involved non-habituated
groups which, however, did not detect the observer by all likelihood.
during 41.8% of these days,
during 35.7% (every 2.8 days), two during 17.3% (every
5.8 days), three during 3.1 % (every 32.7 days), four and the maximum of
five during 1.0% each (every 102.0 days). It was often (but not always)
study groups.
one encounter
No encounter
was recorded
1145
the same group that was met repeatedly during the same day. Overall,
main study groups met neighbours every 1.4 days (A = 1.3 days, B =
days, C = 2.0 days; Table 1). Monthly rates did not differ between
U-test, A&B: U = 26; Nl = 9; N2 = 6; p >
groups (Mann-Whitney
the
1.3
the
0.9;
U = 49; Nl = 9; N2 = 11; p > 0.9; B&C: U = 32.5;
Nl = 6;
N2 = 11; p > 0.9). Thus, group size (A, N = 4; B, N = 5; C, N = 6)
The average acseemed to have no influence on encounter frequencies.
A&C:
tivity day of gibbons of groups A, B and C lasted for 8 h 31 min from
for N = 163 records of focal individ06:16-14:47
(range 05:27-07:36
uals except infants leaving night-trees;
range 13:15-16:17 for N = 270
Encounters
occurred at all but the last
of entering night-trees).
hour of activity (Fig. 5) but were unevenly distributed throughout the day
test; K-S = 1.897, Nl = 20, N2 = 20, p = 0.001).
(Kolmogoroff-Smimov
Rates peaked between 06:30 and 07:59 and again shortly after 09:00. Half
records
began during the first third of the day and only about
12% during the last third. Groups spent 9% of their daytime activity in
encounters with no differences between groups (A, 8 h 53 min, Nl [leave
1.3 encounters/day,
sleeping tree] = 67; Ne [enter sleeping tree] = 119,
=
10.1 %; B, 8 h 40 min, NI = 32; Ne = 43, 1.3 encounaverage 70 min
ters/day, average 67 min = 9.9%; C, 7 h 58 min Ni = 72; Ne = 109, 2.0
of all encounters
encounters/day,
average
64 min = 6.7%).
Fig. 5. Diurnal distribution of encounters in main study groups A, B, C (N = 82). Based
on 98 follows from night tree to night tree. Encounters were scored only once at the
beginning; those with unknown onset (N = 4) were scored to the nearest hour.
1146
Fig. 6. Behaviour recorded during 80 encounters observed from onset to end between
main study groups A, B, C. The occurrence of each category was scored only once. For
definitions see text.
and affiliation
Aggression
during
encounters
were grouped in nine categories rated as either 'neutral',
'agThe
section
describes
these
beor 'affiliative'
6).
following
(Fig.
Behaviours
onistic'
haviours
and quantifies
them with regard to age-sex
classes.
Vocalization
were accompanied by calls. Male solos were heard
tenth
encounter, paired adult males and females duetted during
during every
every third, and females gave solo great calls during every fifth, which,
at times, elicited responses from neighbouring
females.
During encoun=
82 encounters,
N = 115 duets;
ters, 20% of duets were heard (N
Almost
all encounters
more than double as many than expected from the proporMay92-Apr93),
The low-key hoot-sequences
of male
tion of time devoted to encounters.
encounter vocalizations
were completely restricted to this context and often
The calls were usually uttered by adult males and
throughout.
of
all meetings.
90%
However, the two subadult males
during
of group C also produced them and even juveniles joined in on rare occasions. Often, a single male called, whereas at other times, two or more
The calls generally seemed to signal
called simultaneously
or alternating.
persisted
occurred
or tension. However, they did not necessarily culmiarousal, excitement
nate in physical aggression but could lead to more affiliative interactions
(see below).
1147
Sit opposite
of all encounters members of different groups sat in
three-quarters
adjacent trees and looked or stared at each other, typically accompanied by
male encounter vocalizations.
The gibbons would brachiate back and forth
During
without effectively
static affair which
reducing distance.
Sitting opposite was a relatively
in extreme cases lasted for hours, at most interrupted
by brief chases or play (see below). Adult males sat opposite in 98% of
all episodes (N = 207), whereas male-female
opposition accounted for
only 2%. Subadults of either sex and even juveniles were involved in a
few episodes, mostly in combination
with other individuals.
Opposition
between adult females was not observed. It is important to stress that adult
males were almost always at the front of encounters whereas adult females
remained almost always in the background - a pattern very different from
non-encounter
contexts (cf. Table 3).
Co-move/co-feed
tolerated each other in the same or directly adjacent trees during
third
encounter.
every
They cautiously entered a food source (almost always a tree with a large canopy) and fed at opposite sides of the crown
Groups
while constantly monitoring the neighbours' movements.
Co-feeding lasted
between 5 and 60 min. The gibbons would move in and out of the trees to
rest, play or groom calmly. Groups did not mingle except for immatures
during play (see below). Inter-group grooming was not observed. At times,
groups traveled parallel for 100-400 m. The latter pattern deviated clearly
from the common interaction radius of approximately
40 m for encounters.
Chase and contact
aggression
behaviours (vocalization,
sitting opposite, co-feeding) often
Gibbons chased each other during 61 % of
agonistic interactions.
Rather neutral
preceded
all encounters
but in only 8% did this lead to contact aggression. Male-male
interactions constituted 90% of all chases (N = 126) and 73% of all contact
aggression (N = 15). It thus seemed as if females had more difficulties in
escaping contact aggression because male-female victimization constituted
Females clearly
only 6% of all chases but 27% of all contact aggression.
received more frequent aggression by males than vice versa.
The rare
occurrence of female-male
chases (4%) did not reflect dyadic interactions
but cases in which a male together with a female chased a neighbour.
1148
Moreover, in a few cases, the subadult female of group B was seen to
chase other individuals
alone, but these instances shortly before her
- could not be
The
exact nature of
from
the
quantified.
emigration
group
aggressive physical contact was not clearly observable but included more
or less heavy grappling. Loud screams could be heard during about every
tenth encounter, both from males (2 encounters) and females (5 encounters).
It is likely (but not certain) that the few observed blood-drawing
injuries
were inflicted during encounters.
Agonistic exchanges were not the only possible outcome of neutral initial behaviour.
Instead,
fourth
encounter.
every
affiliative
interactions
were observed
during about
Play
developed during every fifth encounter.
chases and wrestling) were indisPlay categories (mostly rough-and-tumble
tinguishable from within-group play and play-partners rarely showed signs
are available for 80 completely and
of fear. Data on partner combinations
Inter-group
play between
immatures
16 partially recorded encounters between groups A, B, C, N and E during
which 52 play sessions were noted. 54.5% included dyadic play, 33.3%
play, 9.1 % dyadic and triadic play, and 3.0% triadic and quartetic
All
immatures of at least the main study groups played. Inter-group
play.
play involved 2 infant, 6 juvenile, 2 subadult and 2 adult partners in the foltriadic
IMA/JFB 1 x, IMA/IFB/JFB
1 x, IMA/JMA/JFB
lowing 18 combinations:
or JFB 1 x , IMA/JMA/IFB/JFB
1 x , JMA/JFB 11 x,
6 x, IMA/JMA/IFB
JMA/IFB/JFB
13 x,
5 x, JMA/JMC
1 x , IMA/JMC 2 x, IMA/JMA/JMC
JMA/JMC/SMC
1 x , JFB/JME 1 x ,
1 1 x , JMA/JFN 1 x , IMA/JMA/JMH
1x.
SFB/JME
JFB/AMB/JME
JFB/JME/AME
1
x
, SFB/AME
1
x
,
1x,
Groups A, B and C each contained an infant and a juvenile. However, the
immatures clearly preferred partners of the same age: dyadic play between
members of different age-classes constituted only 8% of all observed comas opposed to 54% for partners of the same age-class. Moreover,
subadults and adults participated
rarely in inter-group play whereas they
often played within their own groups (subadults adult males adult females).
binations
The negative correlation between age and play frequency obviously is more
rigorous in inter-group than intra-group play. This is illustrated by the development of JMA, JFB and JMC. When they had become subadults during
the 1994-96 study period, they still played regularly in their natal groups
1149
but not as much any more during
or tolerance
displayed indifference
Adults mostly
inter-group encounters.
towards immatures from neighbouring
groups which 'came over' for play. Only once was group A's adult female
observed to intervene and threaten group B's juvenile female who played
with group A's juvenile male, perhaps because the B-individual
had apfast.
The
her
too
immatures
shifted
their
proached
simply
activity a few
meters.
Approach
Adults and subadults
each other during every fourth encounter
as close as 5 m without any party retreating. Males initiated most proximity
(N = 37). They never approached each other but targeted only females
(91 %) with whom they even came a few times into body contact. In rare
cases, females
approached
approached
males (9%).
copulation (EPC)
were
recorded on 45 days during a 12-month window (May92Matings
and
included
72 in-pair copulations (IPC) in groups A and C and 8
Apr93)
extra-pair copulations of the adult female of group A observed during three
Extra-pair
days (7 times with the adult male of group C, once with the adult male of
group B; for details see Reichard, 1995). Two encounters with EPCs of
and 134 min on
AFA and AMC lasted 58 min on 27Jan93 ( copulation)
over a 20 min period). The resident male AMA
1 lMar93 (6 copulations
retreated
in both cases from his pair-mate and became
males of group C. AMC used these occassions
involved
with the
to approach AFA.
for
m
and
She hesitated
only 5-10
copulated soon. Her
Her
infant son (IMA) sometimes clang to her belly while she copulated.
son
sometimes
but
also
interest
in
the
showed
juvenile
squealed
mating
subadult
at first but retreated
within a few meters. Encounter vocalizations
of
couple and approached
AMA were heard from about 100 m away. AMC passed AMA when the
males retreated to their groups but little aggression was observed. The third
with an EPC (this time between AFA and AMB) occurred on
and
lasted for 70 min. The resident male AMA probably detected
25Apr93
the EPC and vigorously chased the neighbouring
male AMB over some
encounter
30 m away from the female. However, no aggression against his pair-mate
was observed. Undisturbed IPCs observed by UR (Oct89-Apr93)
lasted on
=
61
s
51
7-203
N
Undisturbed
EPCs
s,
(median
s,
17).
average
range
1150
during the same period had a shorter average of 46 s (median 23 s, range
3-179 s, N = 7). We believe that fear of detection by the pair-mate or
hightened sexual excitement of partners shortened copulations, although the
difference is statistically not significant (Mann-Whitney
U-test; U = 37,
N¡ = 17, Nz = 7, p < 0.1).
Close contacts
Adult
males
immatures
between
adult males and neighbouring
immatures
at times established
close spatial contact with neighbouring
The immatures, particularly infants, appeared
and at other times frightened.
Such incidences have not
and vice versa.
at times relaxed
previously been described and three episodes shall therefore be documented
in some detail. They mostly involved two sons of AFA (Aran, born in
Sep90, and Akira, born in Oct93).
Case a (observer: UR) 240ct92
A relaxed encounter between groups A and C since 09:48 includes inter-group play of
immatures. At 10:38, AMA briefly chases AMC who retaliates, aided by SMC2. AMA
then disappears out of sight. Soon after, AFA, JMA and IMA (Aran) are seen only about
3 m above ground. AMC and JMC approach slowly. Suddenly, an unidentified member
of group A is briefly chased to the ground. A few more chases follow until 11:10 when
AMC approaches AFA suddenly within arm's reach. AFA moves slowly about 4 m down
and IMA establishes body contact with his mother. JMA places himself between AFA and
AMC, only about 3 m away from the male. AMC passes JMA at II: II and dangles on one
arm right in front of AFA, nervously swinging the other. AFA and IMA retreat 2 m. The
infant then approaches AMC. Both individuals hang on their arms. For several seconds,
AMC gently engulfes IMA with both legs. IMA emits soft vocalizations, similar to those
heard during play. AFA does not intercept the contact. JMC also approaches closer. IMA
has already moved back towards AFA when the female suddenly screams loudly. JMC
and AMC quickly retreat. AFA screams once more, although less intense, when AMC
approaches again. This time, she and JMC withdraw immediately for 15 m. Approaches
and withdrawals continue. In the meantime, AMA has approached AFC to within 10 m.
The rather relaxed atmosphere changes when AMA chases SMC2. AFC then goes after the
neighbouring male, aided by AMC who returns to his group. The encounter ends at 11 :26.
Case b (observer: VS) 05Sep95
Group A meets group C at 12:04. AFA and IMA (Akira) remain in the background; the
two-year old infant has been weaned over the last weeks. AMA, SMA and JMA confront
AMC and SMC. Encounter vocalizations are heard while the males sit opposite to each
other, interrupted by short jumping displays. At 12:16, SMA flees quickly while loud
screams are heard, probably from JMA. The juvenile was probably attacked (and bitten?)
by a male from group C. AFA nervously rushes towards JMA, leaving her infant 20 m
behind. The males from group A retreat further at 12:23, slowly pursued by AMC. At
1151
12:28, AMA has disappeared from the scene; he probably moved again forward towards
the center of group C. AMC, on the other hand, sits only 5 m away from SMA. At 12:33,
SMA has positioned himself between AFA and IMA, 5 m away from each. AFA and IMA
are thus 10 m apart. AMC, however, sits besides IMA, almost in body contact (0.5 m).
The adult male appears self-confident although not aggressive. Both immatures seem to
be frightened, crouch and press themselves against the substrate. At 12:41, AFA lunges
at AMC who flees immediately. After some warm-up notes, AFA emits a single great
call at 12:45. AMA, about 50 m away near group C, does not respond. SMA joins the
confrontation again at 12:58, uttering encounter vocalizations. However, he remains well
behind AMA who marks the front towards AMC, SMC and JMC. The parties retreat at
13:24.
Case c (observer: VS) 07Sep95
Group A meets group C at about 13:45. AMA leads, uttering encounter vocalizations.
SMA is 20 m behind, AFA, JMA and IMA (Akira) about 30 m. The adult male of group
C appears all of sudden at 13:50 and positions himself about 1 m away from AFA; JMA
sits I m behind the female, IMA about 3-4 m sideways. AMC, AFA and JMA seem to
search for arthropods on branches. AMC slaps mildly at AFA 13:58, a gesture resembling
an invitation for play. One minute later the infant screams loudly, falls 12 m and hits hard
on a trail. AMC is in the spot the infant occupied before. This suggests that the adult male
jumped towards the infant which was either knocked out of the tree or dropped himself. AFA
screams, lunges at AMC and drives him away. The infant on the ground remains motionless
for seconds, then appears dizzy but ultimately climbs up again, using small trees and vines.
JMA descends a short distance while monitoring his younger brother's ascent. The infant
is diarrhetic but suffered no visible injury. AFA shows no intent to retrieve or comfort her
infant; instead, she feeds on leaves. AMA remained out of sight since AMC appeared but
his continuous vocalizations indicate a close encounter with group C 30-50 m away. At
14:20, AMC moves again into arms' reach of AFA, IMA and JMA who sit close together,
while SMA is about 10 m in the back. The immatures appear frightened, crouch and press
themselves against branches. AMC brachiates to and fro, addressing gestures at JMA which
resemble invitations for play. JMA does not react and AFA does not intervene. From about
14:28 onwards, AMC brachiates 1-5 m beside and above AFA, JMA, IMA and SMA. AFA
seems calm and does not retrieve her infant. AMC has no erection (indicating a lack of
immediate sexual interest). He utters encounter vocalizations and mild solo elements. AMA
does the same 30-50 m away. The situation appears tense at 14:35 when AFA takes her
infant at her breast. AMC retreats at 14:58 whereupon the four members of group A relax
and start to forage. AMA's encounter vocalization can still be distantly heard. He joins his
slowly traveling group at 15:25.
'Winners'
of encounters?
The question whether agonistic encounters have an 'outcome' can be tackled by analysing chasing-interactions
among males (Fig. 7). All adult
and subadult males participated in all encounters (except for SMC2 who at
times disappeared for days). All males chased and were chased by all other
1152
Fig. 7. Chase-interactions between neighbouring adult and subadult males of main study
groups A, B, C during 95 encounters.
non-group males. Relative rates varied greatly but the males of group C
clearly chased more often (combined ratios: chasing/being chased 2.1) than
0.8), yielding a
they were chased (combined ratios: being chased/chasing
'win/lose-ratio'
of 2.6.
The adult male of group A (ratios:
0.9 and 2.4)
1153
won only with a probability of 0.4, the adult male of group B (ratios: 0.8
and 1.3) with 0.6. Still, AMB appeared to be weakest because he achieved
no positive score against any other male whereas AMA's scores were superior to AMB and SMC1.
The subadults
probably added strength to group
C and may have facilitated AMC's numerous extra-pair copulations with
AFA by keeping AMA busy (see above). Nevertheless,
dominance relationships amongst males were not stable because wins, losses and draws
could vary from day to day. No male was always dominant or submissive.
For example, even AMB evicted whole group C single-handedly
out of a
fruit
This
tree.
male
with
the
lowest
'win-ratio'
also
disputed
copulated
with AFA despite AMA's superior chasing scores (not withstanding
that
AMC, who outcompeted
Seasonal
AMA, faired still better).
pattern
Encounter
rates were not equally distributed throughout the year (Kolmogotest; K-S = 0.86, N = 11, p < 0.05) but
goodness-of-fit
were highest from Mar-Jun (Fig. 8a). We tried to relate this to climatic
roff-Smimov
and social factors.
Food abundance
If encounters
relate to the resource
defense, they should be less frequent
Food availability is probably positively corwhich, in turn, is roughly a function of rainfall
when food is more abundant.
related
with plant biomass
the
two preceding months. In fact, encounters were highest at the
during
end of the dry season of 1992 which lasted unusually long (Fig. 8a). The
first heavy rains were during May92. Food abundance was probably higher
throughout the monsoon and for some months thereafter when more fruits
had ripened. The abundance might have again dropped during Mar-Apr93
leading to a renewed increase in territorial defense. However, a statistically
sound correlation - let alone a significant one - between rain during the
previous month and encounter rates could not be detected (r = -0.013).
Sexual
activity
If encounters reflect efforts for extra-pair copulations they should correlate
positively with high female sexual activity. However, this was not the case
The first peak in encounter rates in May-Jun92 coincided
(r = -0.250).
1154
Fig. 8. Monthly distribution of climatic and behavioural parameters (May92-Apr93).
(a) Rates of encounters (N = 123) of groups A, B, C with any other group and rainfall (as measured during observation days of the preceding month by UR and prorated for
full months). (b) Rates of copulations (N = 82). Based on 1843 h of non-overlapping
observations of groups A, B, C by three researchers (Ulrich Reichard, 1209 h; Bjarne
Klausen, 300 h; Jorg Neudenberger, 334 h. For details, see Reichard, 1995, Table 2. The
adult female of group B did not copulate). (c) Proportions of agonistic, neutral and affiliative behaviours recorded during 59 encounters observed from onset to end between main
study groups A, B, C. One-zero sample. For categories of neutral, agonistic and affiliative
behaviour see Fig. 6.
1155
with a lull in sexual
(when AFA conceived
copulation rates.
activity (Fig. 8b) and only the second around Mar93
her next infant, bom 160ct93) coincided with high
profiles during encounters
agonisitc encounters could reflect defense
Behavioural
Rather
of food or sexually acshould
inter-group interactions
more affiliative
Alternatively
with low encounter rates. However, agonistic and affiliative components of encounters did not vary dramatically in their relative proportions
throughout the year (Fig. 8c).
tive mates.
coincide
Discussion
The following section will discuss the implications of the present findings
on the traditional models that gibbons (a) form nuclear families with a pair
of adults who maintain a lifelong, monogamous
relationship and (b) that
We will also
reflect aggressive defense of territories.
group encounters
review the assumption that infanticide risks may have been a major force
in shaping the social system of gibbons.
Habituation
and observation
biases
in gibbons (Table 5) are marred by methodand biases particularly
because observations
usually
The
surrounded
unhabituated
by
neighbours.
groups
Previous
data on encounters
ological
centered
difficulties
on single
study showed
that this can lead to shorter encounter lengths. The
at Khao Yai yielded the longest recorded duration of
any gibbon encounter (231 min) and an average length (70 min) matched
of unhabituated
groups bias also
by hardly any other study. Observations
present
excellent
habituation
behaviours such as approaches and
against subtle and less attention-drawing
a
Moreover, play,
'low-priority behaviour' is terminated at the
copulations.
Absent records are thus almost certainly artifacts of
slightest disturbance.
is minimized durobservations.
We are confident that such underreporting
ing the present
study.
1156
1157
1158
1159
1160
Resource-defense
by females
and mate-defense
by males
considered
to be territorial and sexually
are traditionally
Earlier studies suggested that range defense establishes both
monogamous.
et al., 1974; Tenaza, 1975; Chivers, 1977).
characteristics
(Brockelman
it
out
that
was pointed
Later,
territory integrity is likely established
by
se
whereas
is
about
intra-sexual
brought
by
aggresmonogamy
groups per
Gibbon
pairs
& Srikosamatara,
sion (Brockelman
1984). Playback experiments simulating range intrusion supported this notion (e.g. Mitani, 1984, 1985; Raemaekers & Raemaekers,
1985). A territory can be defined as 'an area
within the home range occupied more or less exclusively by an animal
al., 1982: 246). This is applicable, al(Lincoln et
though only one-third of the groups' ranges did not overlap. Other studies
have most likely yielded lower overlaps because they did not habituate
or group
of animals'
areas of primate home-ranges
the neighbours
(cf. Table 5). Non-overlap
observation
time. Detailed analysis
become also smaller with increasing
of the range ecology in gibbons is beyond the scope of the present paper.
we believe that it is not necessary to invoke the rather hyNevertheless,
of gibbons but that more
pothetical construct of an intrinsic 'territoriality'
We
are
available.
attempt them by specifying
interpretations
parsimonious
influence travel patterns and inter-group relations in
how sex-differences
of female mammals is limited by food and that
gibbons. The reproduction
of males by breeding opportunities
(Trivers, 1972; cf. review in Dunbar,
of
1988). The spatial dispersion
gibbon females is probably the main factor preventing regular polygyny. It is not well understood why females do
not aggregate but food distribution is likely a crucial factor. Range sizes
probably reflect minimum areas needed to maintain a single female and
offspring, and males simply map themselves onto the spatial distribution
of females.
In fact, females at Khao Yai led most travels into and from
fruit trees whereas males tended to follow (76.0% of all cases). Sexual
in gibbons would then reflect that males are typically unmonomorphism
several females although the reason for this remains
able to monopolize
obscure (van Schaik & Dunbar, 1990). Larger body size could increase
but not without substantial costs such as (a)
intra-sexual competitiveness
the need to find more food and (b) somewhere else than in terminal branches
exploited by females, as well as (c) increased difficulties to guard mates
1161
against lighter competitors
and into smaller branches.
who could follow females
easier during
travel
Moreover, there are indications for a temporal clumping of fertile females.
Births in Khao Yai tend to be more common between August
and November.
fertilize
Such
This hampers roaming strategies which aim to guard and
a single female before moving to the next (cm Orians, 1969).
constraints
prevent
overt
polygyny
without,
strictly monogamous
breeding system. Loopholes
males to pursue alternative reproductive
strategies
vided by group encounters.
encounters
Nevertheless,
however, imposing a
for both males and feare for example proshould be understood
to originally reflect conflicts over space on part of the females (resourcedefense) and disputes over access to reproductive partners on parts of the
males (mate-defense).
Gibbon songs are likely to serve the same function
(Cowlishaw,
1992).
The concept
of resource
defense
range defense has traditionally been thought to be at the core of
in gibbons. The following section highlights revisions
group relationships
and extensions for this notion. The idea that resource-defense
by females
Aggressive
and mate-defense
1. Encounters
by males structures
ranging
leads to several predictions.
should be agonistic
aggression (cf. Fig. 6), thus confirming earlier studies
because
(cf. Table 5). Absence of contact aggression is not contradictory
animals may prefer to reduce the risk of injury or defeat and instead try
Almost
all included
to deter opponents (cf. Maynard Smith & Price, 1973). Fighting abilities
can be conveyed through encounter vocalizations,
solos and locomotion
Situations in which groups sit opposite to each other, co-feed
displays.
and co-travel
seem to reflect stand-offs
during which lines are drawn and
monitored.
moves of the opponents are carefully
Sex-differences
characterised
the agonistic components
of encounters
van Schaik et al., 1992): Male gibbons were
(similar to arboreal langurs;
at the front and females almost always held back. Females never interacted directly
of occasional
whereas
males chased
each other often and up to the point
A
contact aggression.
particular food item has a relatively
small value because it can be shared and also be found elsewhere.
Food
1162
In
is thus rarely worth intense agonistic confrontation
between females.
both already posaddition, females may be in a 'clear enemy' scenario
sess ranges and would not gain by aquiring another (Cowlishaw,
1992).
between females is also not necessarily expected beDirect confrontation
Males thus act in
by males also tails food defense.
certain ways as 'hired guns'.
Access to a mate, on the other hand, has
cannot be
a much higher relative value for a male because conceptions
cause
mate defense
are not abundant. Male-male competition should
shared and opportunities
resemble contests.
Moreover, other males pose a potential risk
to offspring (see below). Blood-drawing
injuries were not observed but it
male
that
at
least
seems likely
amongst males. Interestingly,
they occur,
therefore
gibbon
previous
penter,
canines
specimen
showed more often signs of
M/F
of torn ears 25/3%, Car(ratio
37/28%, Schultz, 1944; severely damaged
shot for museum
collections
than those of females
injuries
1940; healed
42/20%,
2. Encounters
fractures
Frisch,
1963).
should be more common
in the early morning
when ripe
fruit are more abundant
are highly selective feeders which prefer ripe fruit and re-visit rich
sources on consecutive days (Gittins & Raemaekers,
1980; Raemaekers &
is
most
valuable
A
fruit
source
Chivers, 1980).
during the early momboth
ing when it has not yet been depleted by diurnal food competitors,
Gibbons
The present study supports the prediction because
intra- and inter-specific.
encounter rates peaked between 06:30 and 07:59 and declined thereafter
(cf. Fig. 5) - a pattern similar to previous studies. The idea that gibbon
females lead their groups on the best possible foraging routes is also supoccurred in or near food trees,
ported by the fact that many encounters
1977, 1978). However, early
especially big figs (see also Raemaekers,
morning encounters could also simply be elicited by groups realizing they
This would not
are close to one another when members start vocalizing.
an
related
to
food
resources.
require
explanation
should be more frequent and more aggressive when food is
short (resource-defense)
or when females are fertile (mate-defense)
3. Encounters
were unevenly distributed
throughout the year (cf. Fig. 8a).
High rates correlated only partially with increased female receptivity. Food
availability could not be measured directly. If rainfall is used as a proxy, a
Encounters
1163
reduced
availability of food was also found to be at the most partially responsible for high encounter rates (cf. Fig. 8b). Moreover, the proportions
of agonistic and affiliative components during encounters were fairly constant throughout the year (cm Fig. 8c). This suggests that resource defense
is as best not the only function
4. Groups
of encounters.
should develop clear dominance
respected dominance relationships
Mutually
and submissive
individuals
relations
are beneficial
to both dominant
because
they regulate access to resources in acintrinsic resource holding potential, thus reducing the constant need to contest them aggressively (see e.g. Ellis, 1995).
A corresponding
avoidance mechanism (Mitani & Rodman, 1979) would
cordance
reduce
with an individual's
the amount
of time and energy groups have to invest in disputes.
that fed on
of
Sympatric groups
siamang and lar gibbons in Karu/Malaysia
the same food sources might have maintained such a relationship because
they met only every fifth day during the 119-days study - a lack of encounters
probably brought about by the clear dominance of the siamang
1978). Gibbon groups can easily avoid each other, esgroup (Raemaekers,
calls could serve as spacing mechanism.
Thus,
pecially since long-distance
it is not clear why lar gibbons at Khao Yai met so frequently (once every
1.4 days) and for so long (>1 h on average) when newcomers should in
fact either evict a feeding group quickly or leave the overlap area. Instead,
encounters account for a tenth of the activity day - a pattern which seem
to be gibbon-typical
(cf. Table 5). Moreover, a clear-cut dominance hiersubadult and adult males, won
archy lacked because the main 'fighters',
and lost in every possible combination
of chasing interactions (cf. Fig. 7).
There
because
is some indication
group C two subadult males.
that relative strength was linked
who had the best 'winning-probability'
to group size,
- contained
Several factors might be responsible that the last prediction is not fulfilled. First, the fighting abilities of gibbon males might be so similar that
clear-cut dominance
hierarchies
can simply not develop.
Second, most
This might just not be
groups contain only a single breeding female.
male whereas
worth enough to risk an all-out fight with a neighbouring
male-male competition is much more fierce in species where females live
together
and thus enable males to establish
polygynous
groups (e.g. Som-
1164
mer, 1987). Third, there may be short-term variations in the group-specific
values of a contested resource.
For example, the importance of certain
in
fruits
an overlapping
area is reduced if they are abundant in the core
area of the own range. Fourth, gibbons might in various ways benefit from
group encounters per se so that agonistic resource defense is at least supThe next section will deal with
plemented with various social functions.
this aspect,
Affiliative
which has received
social functions
virtually
no attention
till date.
of encounters
The literature
on gibbons emphasizes agonistic and neglects affiliative asOne reason is certainly that records of non-habituated
pects of encounters.
will
be
biased
towards attention-drawing
behaviours such as engroups
and chases.
vocalizations
Moreover, affiliative interactions take
time to develop and gibbons will often quickly flee from an observer. It is
also likely that the concept of gibbons as a territorial, rigidly monogamous
or to downplay
species led observers to overlook affiliative interactions
counter
In any case, about two thirds of ranges at Khao Yai overlapped
and affiliative interactions between members of different groups occurred
during about a quarter of all encounters.
them.
Inter-group play between youngsters was regularly seen. The functions
of play are still not well understood
(cf. Martin & Caro, 1985) but it is
and
that
a
mixture
of
immediate
delayed benefits is at stake such
likely
as the training of locomotory and social skills. Gibbons space births over
2-4 years (Brockelman
& Srikosamatara,
1984; Leighton, 1987; Palombit,
1992) so that immatures have no like-aged play mates in their natal group.
This does not necessarily
lead to a lack of play because immatures of
different age (infant, juvenile, subadult) may be present in the group; adults
will also join immatures for play, at times for extended periods of 90 min
(unpubl. data). Still, inter-group play is the only opportunity to find playpartners of the same age and sex. Such individuals are often preferred if
access
is not limited (e.g. chimpanzees,
Mendoza-Granados
& Sommer,
1995). Play with like-aged partners was also the preferred combination
of inter-group play in the gibbons perhaps because this allows to test and
further develop own abilities most adequately.
However, inter-group play was not limited to same-sex, like-aged partners. The existence of other combinations
might have to do with another
1165
beneficial
to monitor reproductive opaspect of close group interactions:
the
formation
of
a
new
Firstly,
pair amongst grown-up former play
A model for this is a former male of group
partners can be facilitated.
tions.
A who dispersed in the summer of 1990. He was rediscovered
in Mar93
when he had formed a new group (R) with group B's subadult female who
had dispersed in Aug92 (cf. Table 2). Secondly, encounters also enable
mature individuals to closely monitor the physical fitness of
reproductively
same-sex competitors
as well as potential mates. Pair formation in a new
have reached
range is probably a rare option where gibbon populations
capacity. In this case, the challenge of a breeding adult and sub1984;
sequent takeover of a range and mate (Treesucon & Raemaekers,
more
et
in
should
be
the
comBrockelman
Palombit, 1994b;
al.,
press.)
mon strategy, especially for dispersing subadults (Brockelman et al., 1973,
carrying
1974; Mac Kinnon & Mac Kinnon, 1977). To monitor reproduction options
from the safety of a known range is in any case less risky than random
search for mates in unfamiliar areas (cf. Mitani, 1990).
Encounters
terminate
to compare their current mate's quality with
In fact, gibbons have been observed to
neighbours.
and re-pair (Chivers & Raemaekers,
1980;
partnerships
allow
adult and subadult
current
adults
A less dramatic deviation from life-long monogamy
Palombit,
1994b).
are extra-pair
(EPC). Outside Khao Yai, they have been doccopulations
umented at only one other Hylobatidae
study site, i.e. amongst siamangs
1994a), but this is almost certainly an artifact of insufficient data
Half of the paired adults on which the present study focused engaged in EPCs; they occurred on 9% of all days with sexual behaviour and
12% of all copulations of the female of group A. The female
constituted
(Palombit,
collection.
gave birth in the third week of Oct93. Because gestation in white-handed
1981;
gibbons lasts 27-32 weeks (Ardito, 1976; Kollias & Kawakami,
Geissmann,
1991), conception occurred most likely between the first week
of Mar93 and mid Apr93. Thus, the female was probably fertile during at
least the second EPC-encounter.
During the 1994-1996 study period, EPCs
other
individuals
were
observed
from feof
(unpubl. data). DNA-analyses
ces to determine the degree of extra-pair paternity in the study population
& Sommer, in prep.).
males to leave more offspring. However, a male
has to lower his mate-guard in order to philander towards a neighbouring
are under way (Reichard
EPCs can help gibbon
1166
female.
Therefore, he has to balance the pursue of EPC opportunities with
the necessity to insure paternity of his own mate's offspring. Khao Yai data
clearly show that male gibbons follow females much more often than vice
versa, and that mates actively maintain spatial proximity. During encounters, on the other hand, they position themselves usually between their mate
and the neighbouring
male(s), obviously to guard their mates. Subadults
can probably provide crucial assistance in this respect (e.g. H. syndactylus,
1974; H. lar, Ellefson, 1974; H. agilis, Gittins, 1980; H. klossii,
Tilson, 1981; Fig. 7) which could explain why they are often tolerated in
gibbon groups beyond sexual maturity (cf. Table 2). Subadults could benefit from such assistance by gaining experience in same-sex competition,
Chivers,
by evaluating the chance to replace a neighbouring
cause they increase the likelihood that full-siblings
resident male, and beare bom into their natal
group.
Male interest
in neighbouring
females is reflected by the 'friendly' approaches observed during more than one quarter of all encounters. Females
await the outcome of male mate-guarding
gambles rather passively because
91 % of all approaches
vice versa.
were directed
from males to females and only the
active
Nevertheless,
strategic moves of females are
indicated by the occurrence of female great calls during almost one quarter
of all encounters.
Great calls are traditionally thought to be an integral part
of pair-duets, designed
out the neighbourhood
to broadcast
remainder
the existence of a 'pair bond' throughin
Cowlishaw,
(review
1992). Most of these solo
female calls occur just when their pair-mate is not close by. A philandering
male replying from far would give away the degree to which his female
is momentarily
Solo
unguarded and increase the risk of being cuckolded.
female great calls could thus serve a two-fold function: They increase the
costs of philandering for the paired male and at the same time attract neighbouring males. That two adult gibbons are found together does therefore
not necessarily reflect an emotional 'bond' but "following a more parsimonious explanation,
could also result from site attachment combined with
1992, p. 206).
aggression in both sexes" (Anzenberger,
An adaptation to increase the chance of reproduction
with a mate of
superior genetic quality (Fisher, 1930) might entice females to engage in
EPCs. Such options are necessarily restricted to fertile periods between
intrasexual
births.
Their
extent
is unknown
but females
may copulate
over at least
1167
one year before they conceive (Reichard, 1995). High-quality mate selection is somewhat questionable because the respective female (AFA) mated
not only with the probably
'strongest'
neighbour (AMC) but also with
the 'weaker'
one (AMB). On the other hand, any male that outsmarts
a paired male and copulates with a neighbour female might be de facto
a 'high-quality'
male. Moreover, AMB was older and had thus already
'demonstrated'
fitness.
The potential
risk of infanticide
More likely is a link between EPCs and the risk of infanticide which exists
in a variety of species and for a variety of reasons (Hausfater & Hrdy,
1984; Parmigiani & vom Saal, 1994). It has recently been interpreted as
the major selective force behind the evolution of monogamy in gibbons
insofar
as pair-males are thought to provide protection to offspring sired
female (van Schaik & Dunbar, 1990). Infanticide could occur
with 'their'
as a result of (a) temporary invasion by a neighbour male or (b) a pair-male
Gibbons are equipped with sharp and long canines and are
replacement.
certainly capable of killing conspecifics as evidenced by (a) a reported case
of lethal aggression (Palombit, 1993), (b) an anecdote about a captive gibbon male who killed a newborn capped langur (Presbytis [Semnopithecus]
pileatus) kept with his mother in the same enclosure (Anderson, 1893, cit.
in Zuckerman,
1981) and (c) an anecdote according to which gibbons presumably killed a large bird (chicken; Newkirk, 1973). Infant killing might
be advantageous
for several reasons.
Killing an unrelated offspring will
benefit a male if it increases
of siring the victimized mother's next infant. The death of an unweaned infant will typically shorten
The death of a weaned
the temporary sterility associated with nursing.
of
could
benefit
future
the
killer
offspring
progeny
by curtailing maternal
investment
into another
his likelihood
male's
offspring which may continue far beyond
nursing, e.g. through bridging gaps in the canopy with the own body, or
grooming. Killing older unrelated offspring also reduces the competition a
future own offspring would face for resources such as food, safe sleeping
trees or mates. Finally, the killer of an infant or juvenile might more easily
replace the current adult male resident if the female withdraws her support
from a mate who failed to protect
her progeny
(Hood,
1994).
1168
has not been observed in any gibbon species but to do so - would be difficult
anyway. Lethal aggression is a
especially in the wild
brief event which reduces the chance to witness it, even in a largely groundInfanticide
dwelling species such as Indian grey langurs (Sommer, 1987). Infanticide
under certain social condiis thus often inferred from infant disappearance
rare at Khao Yai.
tions. However, such indirect evidence is conspicuously
At least 9 gibbons were bom since 1979 into the main focal groups A,
B and C, and all survived till date except for one which disappeared as a
from neighal., in press). Infant disappearances
juvenile (Brockelman et
bouring groups which are less well monitored are rare, too. Infanticide
could at best be responsible for this remainder of a near-zero mortality but
other natural causes of death are, of course, as likely.
In species such as gorillas, group encounters are associated with a high
risk of infanticide
(Watts, 1989). In the Khao Yai gibbons, contact aggression occurred during 7.5% of all encounters, and 27% of them were
directed by males against females. However, it was impossible to decide if
targeted only the females or if aggressive male-female interactions
developed because females tried to intervene on behalf of their offspring
which were usually close by. Given that encounters occur every 1.4 days,
attacks
can thus expect a male attack every 71 days. This would translate to roughly 10 attacks during a two-year nursing period. Still, there
is no direct evidence that gibbon male attacks on females are linked to
a female
Firstly, the attacks could represent efforts to drive neighbours
from
food
sources or they could reflect attempts to coerce unwillaway
ing females into sexual activity. Secondly and more important: Our study
physical contact between inreports for the first time that non-aggressive
infanticide.
fants and neighbour males can occur, including about 6% of all inter-group
play sessions. This is surprising given that adult males can certainly harm
immatures.
Neighbour males could have also easily killed the infant in at
least two of the close-encounter
cases narrated
above.
Such non-aggressive
In fact, AMC had engaged
suggest that paternity is confused.
in EPCs during the period when the neighbouring
infant (Akira) was conceived. Female EPCs and the long period of pre-conception
copulations
could thus reflect counter-strategies
to forestall infanticide.
For example,
interactions
amongst
grey langurs,
infanticide
is common
in populations
with one-male
1169
breeding
structures
but much rarer in multi-male
groups
where paternity
certainty is low (e.g. Sommer, 1987).
We can speculate further that such conditions
favour even a slightly
of
A
male
can
degree
mate-guarding.
accept a certain proportion
of EPCs if his own risk of not siring an offspring with his female is
lower than the combined benefit from his own efforts to achieve EPCs and
lowered
the additional protection from infanticide his (likely) offspring gains as a
result of EPCs of his pair-mate. This trade-off might explain the somewhat
puzzling observation described above that males do sometimes not return
even if they hear or see that a neighbour male is close to their female.
Nevertheless,
neighbour males are not always perceived as friendly by
immatures
or their mothers, because immatures can show extreme fear
while mothers
tend to 'defend'
them. Thus, some ambiguity towards neighthe
issue of confused paternity. It can also not
despite
that case (c) described above represented a genuine effort of
bour males remains
be excluded
the neighbour male to harm the infant. In any case, the infant fell far to
the ground. Falls can cause at least fractures in gibbons (Schultz, 1944).
Thus, it remains unclear if gibbon males pose a threat to infants but it
seems premature to exclude the possibility.
Similarly, the absence of observations on predation does not necessarily indicate that there is no risk
of predation but may well reflect effective predator-avoidance
mechanisms
(Dunbar, 1988).
In sum, encounters between gibbon groups appear likely to have roots
in disputes over resources and mates but they also provide opportunities to
pursue a variety of reproductive
options, particularly sexual contacts with
more then one partner. These mating pattern are likely to create groups
where young are not always full-siblings but were kin-ship relations extend
et al.,
well into neighbour groups. Short dispersal distances (Brockelman
in press) reinforce this effect. A positive feedback is thus started, which
reduces the level of competition between groups and further promotes the
kind
of affiliative
inter-group
relationships
observed
during
the present
study.
References
Aiello, L.C. & Dunbar, R.I.M. (1993). Neocortex size, group size, and the evolution of
language. - Curr. Anthropol. 34, p. 184-193.
1170
Altmann, J. (1974). Observational study of behavior: sampling methods. - Behaviour 49,
p. 227-265.
Anzenberger, G. (1992). Monogamous social systems and paternity in primates. - In: Paternity in primates: Genetic tests and theories (R.D. Martin, A.F. Dixon & E.J. Wickings,
eds). Karger, Basel, p. 203-224.
Ardito, G. (1976). Check-list of the data in the gestation length of primates. - J. Hum.
Evol. 5, p. 213-222.
Brockelman, W.Y. (1985). A gibbon pelt (Hylobates lar entelloides) from Khao Yai National
Park, Saraburi Province, Thailand. - Nat. Hist. Bull. Siam Soc. 33, p. 55-57.
-& Srikosamatara, S. (1984). Maintenance and evolution of social structure in gibbons. - In: The lesser apes: Evolutionary and behavioural biology (H. Preuschoft,
D.J. Chivers, W.Y. Brockelman & N. Creel, eds). Edinburgh University Press, Edinburgh, p. 298-323.
- -,
Ross, B.A. & Pantuwatana, S. (1973). Social correlates of reproductive success in the
gibbon colony on Ko Klet Kaeo, Thailand. - Am. J. Phys. Anthropol. 38, p. 637-640.
& --,(1974). Social interaction of adult gibbons (Hylobates lar) in an
experimental colony. - In: Gibbon and siamang, vol. 3, natural history, social behavior, reproduction, vocalisations, prehension (D. Rumbaugh, ed.). Karger, Basel,
p. 157-175.
- -,
Reichard, U., Treesucon, U., Raemaekers, J.J. (in press). Dispersal, pair formation
and social structure in gibbons (Hylobates lar).
Carpenter, C.R. (1940). A field study in Siam of the behavior and social relations of the
gibbon (Hylobates lar). - Comp. Psychol. Monogr. 84, p. 1-212.
Chen, N. (1995). Ecology of the black-crested gibbon (Hylobates concolor) in the Ailao
Mt. Reserve, Yunnan, China. - M.Sc. thesis, Mahidol University, Bangkok.
Cheney, D.L. (1987). Interactions and relationships between groups. - In: Primate societies
(B.B. Smuts, D.L. Cheney, R.M. Seyfarth, R.W. Wrangham & T.T. Struhsaker, eds).
University of Chicago Press, Chicago, p. 267-281.
Chivers, D.J. (1971). The malayan siamang. - Malay. Nat. J. 24, p. 78-86.
The siamang in Malaya: A field study of a primate in a tropical forest.
- - (1974).
- In: Contributions to primatology, vol. 4 (H. Kuhn, C.R. Luckett, C.R. Noback,
A.H. Schultz, D. Starck & F.S. Szalay, eds). Karger, Basel, p. 1-335.
-(1976). Communication within and between family-groups of siamang. - Behaviour
57, p. 116-135.
-(1977). The lesser apes. - In: Primate conservation (H.S.H. Prince Rainier &
G.H. Bourne, eds). Academic Press, New York, p. 539-598.
-& Raemaekers, J.J. (1980). Long-term changes in behaviour. - In: Malayan forest
primates: Ten years' study in tropical rain forest (D.J. Chivers, ed.). Plenum Press,
New York, p. 209-260.
Cowlishaw, G. (1992). Song function in gibbons. - Behaviour 121, p. 131-153.
Dunbar, R.I.M. (1988). Primate social systems. - Croom Helm, London.
Ellefson, J.O. (1968). Territorial behavior in the common white-handed gibbon (Hyl. lar
Linn.). - In: Primates: Studies in adaptation and variability (P.C. Jay, ed.). Holt
Richert & Winston, New York, p. 180-199.
1171
--
(1974). A natural history of white-handed gibbons in the Malayan Peninsular. - In:
Gibbon and siamang, vol. 3, natural history, social behavior, reproduction, vocalizations, prehension (D.M. Rumbaugh, ed.). Karger, Basel, p. 1-136.
Ellis, L. (1995). Dominance and reproductive success among nonhuman animals: A crossspecies comparison. - Ethol. Sociobiol. 16, p. 257-333.
Fisher, R.A. (1930). The genetical theory of natural selection. - Clarendon Press, Oxford.
Frisch, J.E. (1963). Sex-differences in the canines of the gibbon Hylobates lar. - Primates 4, p. 1-10.
Geissmann, T. (1991). Reassessment of age of sexual maturity in gibbons (Hylobates ssp.).
- Am. J. Primatol. 23, p. 11-22.
Gittins, S.P. (1979). The behaviour and ecology of the agile gibbon (Hylobates agilis). Ph.D. thesis, University of Cambridge, Cambridge.
-(1980). Territorial behavior in the agile gibbon. - Int. J. Primatol. 1, p. 381-399.
-(1984). Territorial advertisement and defence in gibbons. - In: The lesser apes:
Evolutionary and behavioural biology (H. Preuschoft, D.J. Chivers, W.Y. Brockelman
& N. Creel, eds). Edinburgh University Press, Edinburgh, p. 420-424.
-& Raemaekers, J.J. (1980). Siamang, lar, and agile gibbons. - In: Malayan forest
primates: Ten years' study in tropical rain forest (D.J. Chivers, ed.). Plenum Press,
New York, p. 63-105.
-& Tilson, R.L. (1984). Notes on the ecology and behaviour of the hoolock gibbon. - In: The lesser apes: Evolutionary and behavioural biology (H. Preuschoft,
D.J. Chivers, W.Y. Brockelman & N. Creel, eds). Edinburgh University Press, Edinburgh, p. 258-266.
Haimoff, E.H. (1984). Acoustic and organizational features of gibbon songs. - In: The
lesser apes: Evolutionary and behavioural biology (H. Preuschoft, D.J. Chivers,
W.Y. Brockelman & N. Creel, eds). Edinburgh University Press, Edinburgh, p. 333353.
Hausfater, G. & Hrdy, S.B. (1984). Infanticide. Comparative and evolutionary perspectives.
- Aldine, New York.
Hood, L.C. (1994). Infanticide among ringtailed lemurs (Lemur katta) at Berenty Reserve,
Madagascar. - Am. J. Primatol. 33, p. 65-69.
Kappeler, M. (1981). The javan silvery gibbon (Hylobates lar moloch). Part I & II. - Ph.D.
thesis, Universitat Basel, Basel.
-(1984). Vocal bouts and territorial maintenance in the moloch gibbon. - In: The
lesser apes: Evolutionary and behavioural biology (H. Preuschoft, D.J. Chivers,
W.Y. Brockelman & N. Creel, eds). Edinburgh University Press, Edinburgh, p. 376389.
Kollias, G.V. & Kawakami, T.G. (1981). Factors contributing to the successful captive
reproduction of white-handed gibbons (Hylobates lar). - Am. Ass. Zoo Vet. Ann.
Proc. 1981, p. 45-47.
Leighton, D.R. (1987). Gibbons: Territoriality and monogamy. - In: Primate societies
(B.B. Smuts, D.L. Cheney, R.M. Seyfarth, R.W. Wrangham & T.T. Struhsaker, eds).
University of Chicago Press, Chicago, p. 135-145.
Lincoln, R.J., Boxshall, G.A. & Clark, P.F. (1982). A dictionary of ecology, evolution and
systematics. - Cambridge University Press, Cambridge.
1172
Mac Kinnon, J.R. & Mac Kinnon, K.S. (1977). The formation of a new gibbon group. Primates 18, p. 701-708.
Martin, P. & Caro, T.M. (1985). On the functions of play and its role in behavioral development. - In: Advances in the study of behavior, vol. 15 (J.S. Rosenblatt, C. Beer,
M.-C. Busnel & P.J.B. Slater, eds). Academic Press, New York, p. 59-103.
Maynard Smith, J. & Price, G.R. (1973). The logic of animal conflict. - Nature 246,
p. 15-18.
Mc Cann, C. (1933). Notes on the colouration and habits of the white-browed gibbon or
hoolock (Hylobates hoolock Harl.). - Joum. Bombay Nat. Hist. Soc. 36, p. 395-405.
Mendoza-Granados, D. & Sommer, V. (1995). Play in chimpanzees of the Arnhem Zoo:
Self-serving compromises. - Primates 36, p. 57-68.
Mitani, J.C. (1984). The behavioural regulation of monogamy in gibbons (Hylobates muelleri). - Behav. Ecol. Sociobiol. 15, p. 225-229.
-(1985). Gibbon song duets and intergroup spacing. - Behaviour 92, p. 59-95.
-(1990). Demography of agile gibbons (Hylobates agilis). - Int. J. Primatol. 11,
p. 411-424.
-& Rodman, P.S. (1979). Territoriality: The relation of ranging patterns and home
range size to defendability, with an analysis of territoriality among primate species.
- Behav. Ecol. Sociobiol. 5, p. 241-251.
Neudenberger, J. (1993). Monogamie als Paarungssystem: Eine Fallstudie am
Weißhandgibbon (Hylobates lar) im Khao Yai Nationalpark, Thailand. - MSc thesis,
Universität Gottingen, Gottingen.
Newkirk, J.B., III. (1973). A possible case of predation in the gibbon. - Primates 14,
p. 301-304.
Orians, G.H. (1969). On the evolution of mating systems in birds an mammals. - Am.
Nat. 103, p. 589-603.
Palombit, R.A. (1992). Pair bonds and monogamy in wild siamang (Hylobates syndactylus) and white-handed gibbon (Hylobates lar) in northern Sumatra. - Ph.D. thesis,
University of California, Davis.
-(1993). Lethal territorial aggression in a white-handed gibbon. - Am. J. Primatol.
31, p. 311-318.
-(1994a). Extra-pair copulations in a monogamous ape. - Anim. Behav. 47, p. 721723.
-(1994b). Dynamic pair bonds in hylobatids: Implications regarding monogamous
social systems. - Behaviour 128, p. 65-101.
Parmigiani, S. & vom Saal, F. (1994). Infanticide and parental care. - Harwood Academic
Publishers, London.
Preuschoft, H., Chivers, D.J., Brockelman, W.Y. & Creel, N. (1984). The lesser apes:
Evolutionary and behavioural biology. - Edinburgh University Press, Edinburgh.
Raemaekers, J.J. (1977). Gibbons and trees: Comparative ecology of the siamang and lar
gibbons. - Ph.D. thesis, University of Cambridge, Cambridge.
-(1978). Competition for food between lesser apes. - In: Recent advances in primatology, vol. 1, behaviour (D.J. Chivers & J. Herbert, eds). Academic Press, New
York, p. 327-330.
1173
--
& Chivers, D.J. (1980). Socio-ecology of Malayan forest primates. - In: Malayan
forest primates: Ten years' study in tropical rain forest (D.J. Chivers, ed.). Plenum
Press, New York, p. 279-315.
-& Raemaekers, P.M. (1984a). The Ooaa duet of the gibbon (Hylobates lar): A group
call which triggers other groups to respond in kind. - Folia Primatol. 42, p. 209-215.
-&-(1984b): Vocal interaction between two male gibbons, Hylobates lar. Nat. Hist. Bull. Siam Soc. 32, p. 95-106.
-&-(1985). Field playback of loud calls to gibbons (Hylobates lar): Territorial,
sex-specific and species-specificresponses. - Anim. Behav. 33, p. 481-493.
& Haimoff, E.H. (1984). Loud calls of the gibbon (Hylobates lar): Repertoir,
- -, - organisation and context. - Behaviour 91, p. 146-189.
Reichard, U. (1995). Extra-pair copulations in a monogamous gibbon (Hylobates lar). Ethology 100, p. 99-112.
-(1996). Sozial- und Fortpflanzungsverhalten von Weißhandgibbons (Hylobates lar):
Eine Freilandstudie im thailandischen Khao Yai Regenwald. - Cuvillier, Göttingen.
van Schaik, C.P., Assink, P.R. & Salafsky, N. (1992). Territorial behavior in Southeast Asian
langurs: Resource defence or mate defence? - Am. J. Primatol. 26, p. 233-242.
-& Dunbar, R.I.M. (1990). The evolution of monogamy in large primates: A new
hypothesis and some crucial tests. - Behaviour 115, p. 30-62.
-& Kappeler, P.M. (1993). Life history, activity period and lemur social systems. In : Lemur social systems and their ecological basis (P.M. Kappeler & J. Ganzhom,
eds). Plenum, New York & London, p. 241-260.
Schultz, A.H. (1944). Age changes and variability in gibbons. A morphological study on a
population sample of a man-like ape. - Am. J. Phys. Anthropol. 2, p. 1-129.
Sheeran, L.K. (1993). A preliminary study of the behavior and socio-ecology of black
gibbons (Hylobates concolor) in Yunnan Province, People's Republic of China. Ph.D. thesis, Ohio State University, Ohio.
Sommer, V. (1987). Infanticide among free-ranging langurs (Presbytis entellus), at Jodhpur (Rajasthan/India): Recent observations and a reconsideration of hypotheses. Primates 28, p. 163-197.
Srikosamatara, S. (1980). Ecology and behaviour of the pileated gibbon (Hylobates pileatus)
in Khao Soi Dao Wildlife Sanctuary, Thailand. - M.Sc. thesis, Mahidol University,
Bangkok.
Tenaza, R.R. (1975). Territory and monogamy among Kloss' gibbons (Hylobates klossii) in
Siberut island, Indonesia. - Folia Primatol. 24, p. 60-80.
-(1976). Songs, choruses and countersinging of Kloss' gibbons (Hylobates klossii) in
Siberut island, Indonesia. - Z. Tierpsychol. 40, p. 37-52.
-& Hamilton, W.J. III. (1971). Preliminary observations of the Mentawai islands
gibbon, Hylobates klossii. - Folia Primatol. 15, p. 201-211.
Tilson, R.L. (1979). On the behavior of hoolock gibbons (Hyobates hoolock) during different
seasons in Assam, India. - Journ. Bombay Nat. Hist. Soc. 76, p. 1-16.
-(1981). Family formation strategies of Kloss's gibbons. - Folia Primatol. 35, p. 259287.
Treesucon, U. (1984). Social development of young gibbons (Hylobates lar) in Khao Yai
National Park, Thailand. - M.Sc. thesis, Mahidol University, Bangkok.
1174
--
& Raemaekers, J.J. (1984). Group formation in gibbon through displacement of an
adult. - Int. J. Primatol. 5, p. 387.
Trivers, R.L. (1972). Parental investment and sexual selection. - In: Sexual selection and
the descent of man 1871-1971 (B. Campbell, ed.). Aldine Press, Chicago, p. 136-179.
Tuttle, R.H. (1986). Apes of the world. - Noyes Publications, New Jersey.
Watts, D. (1989). Infanticide in mountain gorillas: New cases and a reconsideration of the
evidence. - Ethology 81, p. 1-18.
Whitington, C. (1990). Seed dispersal by white-handed gibbons (Hylobates lar) in Khao
Yai National Park, Thailand. - M.Sc. thesis, Mahidol University, Bangkok.
Whitten, A.J. (1980). The kloss gibbon in Siberut rain forest. - Ph.D. thesis, University
of Cambridge, Cambridge.
Zuckerman, S. (1981). The social life of monkeys and apes. Re-issue of the 1932 edition
together with a postscript. - Routledge & Kegan Paul Ltd., London.