Content uploaded by Carel P van Schaik
Author content
All content in this area was uploaded by Carel P van Schaik on Dec 18, 2018
Content may be subject to copyright.
235
least one sex nd the other. The mating system of
orangutans is of interest because it is based on an
interaction between female choice and male har-
assment and coercion, against a background of
16.1 Introduction
In species where both sexes are more or less soli-
tary, such as orangutans, mating requires that at
CHAPTER 16
Orangutan mating behaviour
and strategies
S. Suci Utami Atmoko, Tatang Mitra Setia, Benoît Goossens,
Sheena S. James, Cheryl D. Knott, Helen C. Morrogh-Bernard,
Carel P. van Schaik and Maria A. van Noordwijk
16-Wich-Chap16.indd 23516-Wich-Chap16.indd 235 7/7/2008 4:01:01 PM7/7/2008 4:01:01 PM
236 ORANGUTANS
male–male competition predominate (Schürmann
and van Hooff, 1986; Rodman and Mitani, 1987).
Between-male contest competition must clearly
have contributed to t he developme nt of th e ext reme
s ex ua l di mo r ph is m o f t h is sp e ci es . I n de ed , R od m an
and Mitani (1987) proposed it as the sole factor.
However, because females selectively approach
or allow themselves to be approached by anged
males when they are in their fertile periods, ‘female
choice’ must also play a critical role in maintaining
male physical characteristics.
Another unusual feature of the mating system
of orangutans is the occurrence of forced mat-
ings (‘rapes’) (Rijksen 1978; Galdikas 1979, 1985b, c;
Mitani 1985a, b; Fox 1998, 2002). Even though pre-
mating aggression and physical coercion to mate
are seen in other primates, orangutan females seem
to resist mating attempts by particular males much
more strongly than seen in other species. Mating
resistance, in which a female struggles and attempts
to prevent intromission while the male attempts to
restrain the female by grabbing and holding on
to her arms, legs and body, can result in a erce
physical ght between a male and a female (and
sometimes her youngest offspring). These inter-
actions may involve hitting, pushing and biting by
all involved and whimpering, squealing, grumph-
ing, kiss-squeaking and other vocalizations by the
fema le and/or her off sprin g. Th e level an d ti min g of
female cooperation and resistance is variable: some
male mating attempts may star t with resistance and
continue passively or even cooperatively, others
may start cooperatively but end with resistance
(Fox 1998), making classi cation complicated. Since
females resist mating attempts by particular males
even when no other males are present (and may not
even be within hearing distance) this behaviour
is unlikely to incite direct male–male competition
as suggested for other species with conspicuous
female mating resistance (e.g. Cox and LeBoeuf
1977; Boness et al. 1982) but seems to be an honest
re ection of female preference (Fox 1998, 2002). So
far, in all populations females have been observed
to engage in a wide range of mating interactions
from female-initiated active participation to ercely
contested, with many variations in between.
Perhaps the most unusual feature of orang utans
is the remarkable individual variation in the age
intense male–male competition, which has pro-
duced extreme sexual dimorphism and is accom-
panied by very unusual male bimaturism.
The aim of this chapter is to assess how these
three aspects of sexual selection (male competition,
female choice and mating con ict) interact to pro-
duce the mating behaviour of orangutans. After
a theoretical introduction and a review of what
is known about orangutan mating behaviour, we
present a compilation of basic data from various
sites in order to assess whether the broad picture
built up over the years is similar throughout the
orangutan’s geographic range.
Orangutans have a semi-solitary life style.
Individuals usually live alone in highly overlap-
ping home ranges, and at some sites they occasion-
ally aggregate in large fruit trees or groves (see
Chapter 17). During times of high fruit abundance,
these aggregations develop into travel bands, in
which individuals travel together in a coordinated
fashion (Sugardjito et al. 1987; Utami et al. 1997;
van Schaik 1999; Singleton and van Schaik 2002).
Consortships, a form of travel band in which a
male–female pair may range together in a coordi-
nated way for several days, weeks or even months
and engage in sexual behaviour, are seen at all
sites. Other individuals (females, un anged males
and adolescent individuals) may also associate
around a consort pair (Schürmann and van Hooff
1986), although fully mature, anged males never
associate together and usually behave antagonis-
tically towards each other (see Chapter 15).
Orangutans have three striking features for pri-
mates: (1) extreme sexual dimorphism in body size,
(2) forced matings, and (3) bimaturism among males.
As in many other dimorphic primate species, the
se xu al di mo rph is m i n bo dy s iz e ar is es bec au se m al es
continue to grow beyond the age at which females
stop growing. In orangutans, male growth con tinues
gradually until well after the twentieth year of age
(‘indeterminate growth’, Leigh and Shea [1995]). In
addition to larger body size, the male secondary sex-
ual characters (SSCs) consist of anged cheeks and a
throat sack, with which they can produce long calls
to advertise their presence (MacKinnon 1974; Rijksen
1978; Rodman 1984; Galdikas 1985b).
Hypotheses to explain sexual dimorphism
invoke sexual selection in which female choice or
16-Wich-Chap16.indd 23616-Wich-Chap16.indd 236 7/7/2008 4:01:04 PM7/7/2008 4:01:04 PM
ORANGUTAN MATING BEHAVIOUR AND STRATEGIES 237
This chapter considers the mating strategies of
males and females and their interactions, based on
eld studies in Ketambe (Sumatra), Suaq Balimbing
(Sumatra), Gunung Palung (West Kalimantan),
Tanjung Puting (Central Kalimantan), Sabangau
(Central Kalimantan), Tuanan (Central Kalimantan),
Kutai (Eas t Ka l i m a ntan) an d Ki n abat a ngan (S a bah).
Table 16.1 gives an overview of the database used
for this chapter.
16.2 Mating behaviour of orangutans
16.2.1 Ontogeny of mating behaviour
Mating behaviour is not unique to adult and ado-
lescent individuals. In Ketambe, infants as young
as two years of age, of either sex, already show
keen interest in sexual behaviour. For instance,
the infant male Yossa sometimes masturbated as
part of his play behaviour, sometimes using his
own hands or feet, once using a stick, and occa-
sionally even thrusting into his mother’s vagina
while she was hanging. Immatures also directed
sexual behaviour at each other, especially when
they were playing. For instance, 6-year-old female
Tati had 3-year-old Yossa masturbate her, and after
that she masturbated him. During rest, orangutan
mothers often licked their infant’s genitals to clean
them, and this may be the rst sexual experience
at which sexually mature males develop their
SSCs—a phenomenon called bimaturism (Uchida
1996; Maggioncalda et al. 1999, 2002). In some
males, at least in Sumatra, this development may
be delayed until the male is well over 30 years old
(Utami Atmoko 2000; Utami et al. 2002), some 20
years after reaching sexual maturity and at least
15 years after reaching the size of adult females.
Although these un anged or ‘arrested’ males lack
SSCs, they are fertile, sexually active and able to
sire offspring (Kingsley 1982; Maggioncalda et al.
1999, 2002; Utami Atmoko 2000; Utami, et al. 2002).
This bimaturism leads to the coexistence of two
adult, sexually mature morphs: anged and unf-
langed males.
The presence of an uncontested anged male
has been proposed as a key proximate factor main-
taining developmental arrest in un anged males,
at least in captivity (Kingsley 1982; Graham and
Nadler 1990; Maggioncalda et al. 1999, 2002; Utami
Atmoko 2000). However, this is unlikely to work
in the wild, where the two kinds of males inhabit
large ranges widely overlapping with those of
many others, anged and un anged. Utami et al.
(2002) could show that two early hypotheses for
the evolution of bimaturism were inconsistent
with the data. The data reviewed in this chapter
can be used to evaluate other hypotheses (see
discussion).
Table 16.1 Overview of the database used for this chapter
Site Period Researchers
Sumatra
Ketambe 1972–1978 H.D. Rijksen, C.L. Schürmann
1979–1996 Ketambe orangutan project Universitas Nasional
Jakarta—Utrecht University Netherlands
Suaq Balimbing 1994–1999 Suaq Balimbing orangutan project Universitas
Indonesia Jakar ta—Duke University USA
Borneo
Kinabatangan 2000–2003 M. Ancrenaz, S.S. James
Kutai NP 1981–1982 J. Mitani (1985b)
Sabangau 2003–2005 H. Morrogh-Bernard
Tuanan 2003–2006 Tuanan orangutan project Universitas Nasional
Jakarta—University of Zürich, Switzerland
Tanjung Puting 1971–1975 B.M.F. Galdikas (1979, 1981, 1985a, b)
Gunung Palung 1994–2003 C.D. Knott
16-Wich-Chap16.indd 23716-Wich-Chap16.indd 237 7/7/2008 4:01:04 PM7/7/2008 4:01:04 PM
238 ORANGUTANS
search for females. Thus the mating strategies of
the two different male morphs differ considerably
in the way they nd potential mates.
To see whether males have more copulations
with females who are potentially receptive than
with females in a non-reproductive stage (clearly
pregnant or nursing a dependent infant), we
counted the number of their copulations with
females in either stage. Figure 16.1 shows that, at
all sites, anged males copulate more with poten-
tially fertile females than with non-fertile females.
Un anged males show a weaker tendency, but
more detailed data are needed to detect whether
there are real differences between male morphs
and populations.
A consortship occurs when a male and a female
travel together and show coordination in their
behaviour for several days or weeks, up to a month,
during which interactions, including copulations,
may take place (MacKinnon 1974; Rijksen 1978;
Utami Atmoko 2000). Data from several sites sug-
gest that adult (parous) females prefer anged
males as consort partners (Galdikas 1979, 1981;
Mitani 1985b; Rodman and Mitani 1987; Mitra Setia
1995; Fox 2002). This is con rmed by the available
quantitative data for Ketambe, Sumatra (Fig. 16.2):
almost one half of the copulations by un anged
for young individuals. They also acquire early
knowledge of mating behaviour by watching their
mothers copulate with males. Especially un anged
males occasionally try to mate with mothers with
small infants; infants almost always respond to
these attempts by struggling with the male.
16.2.2 Male mating behaviour
Male–male competition has been cited as a major
determinant of orangutan social organization,
in which males compete for access to females
(Rodman 1973b, 1988; Rijken 1978; Galdikas 1979;
Mitani 1985a; Utami Atmoko 2000; Utami Atmoko
et al. in preparation; see Chapter 15). In support of
this argument, anged males are highly intoler-
ant of each other, and their ‘long calls’ function
as a spacing mechanism (Mitani 1985a; Mitra
Setia and van Schaik 2007; see Chapter 15). These
same long calls also function as a locator call to
females, and attract receptive females (Mitra Setia
and van Schaik 2007; see Chapter 17). Receptive
females may initiate association by approaching a
male’s long call, or accept his initiative to associate
(Schürmann and van Hooff 1986; see Chapter 17).
Un anged males, on the other hand, cannot attract
females through vocalizations and have to actively
100
N 164 N 353
***
N 55 N 15 N 42 N 24
80
60
40
% Reproductive females
20
0
ke Su
Sumatra Borneo
Tu Sa GP Kn
Flanged
Unflanged
Figure 16.1 Percentage of copulations by fl anged and unfl anged males with reproductive (sexually active) females in six dif ferent sites.
Numbers above columns indicate total number of copulations observed and stars indicate signifi cant difference between fl anged and
unfl anged males in Gadj -test: * P ⬍0.05; ** P ⬍0.01.
16-Wich-Chap16.indd 23816-Wich-Chap16.indd 238 7/7/2008 4:01:04 PM7/7/2008 4:01:04 PM
ORANGUTAN MATING BEHAVIOUR AND STRATEGIES 239
Utami Atmoko 2000). Un anged males were never
observed to have an ECC in either site.
In Ketambe, females were found more often
in the proximity of the dominant male than of
other anged males; their associations with him
lasted longer as well (Utami and Mitra Setia
1995; Mitra Setia 1995). Likewise, in Suaq, non-
dominant anged males were very rarely found
in consortship and had negligible mating success
(van Schaik 2004). However, Ketambe females
maintained a good relationship not only with a
dominant anged male but also with some unf-
langed males (Utami Atmoko 2000). Tolerance of
un anged males by anged males was sometimes
observed, even during consortships of a anged
male and a receptive female (cf. Rijksen 1978;
Galdikas 1985b, c; Mitani 1985b; van Schaik and
van Hooff 1996). In Ketambe, for example, it has
been observed that the anged male Jon allowed
the un anged male Boris to travel within 20 m dis-
tance for several days while he was in consort with
the adult female Yet (Utami Atmoko 2000; Utami
Atmoko et al. in preparation). In such situations
females have been seen to have ECCs with an unf-
langed male (Utami Atmoko 2000; Utami Atmoko
et al. in preparation).
16.2.3 Female mating behaviour
Female orangutans mate promiscuously, not only
with anged males other than the dominant male
but especially with un anged males (Rijksen 1978;
Galdikas 1985b, c; Rodman and Mitani 1987). Early
reports suggested that females mated coopera-
tively with the anged male that is resident within
her home range (Schürmann 1982; Galdikas 1985b;
Schürmann and van Hooff 1986), whereas unf-
langed males usually obtained matings by force
(Galdikas 1985b; Rijksen 1978; Rodman and Mitani
1987). However, occasionally, females demon-
strated the reverse pattern, displaying proceptiv-
ity toward some un anged males (Fox 1998; Utami
Atmoko 2000) and resisting mating attempts by
anged males (Rijksen 1978; Mitani 1985b; Knott
and Kahlenberg 2007).
Quantitative data from multiple sites (Fig. 16.3)
show more variation in mating patterns than previ-
ously thought: females in the Sumatran and at least
males are not linked to consortships (of at least 5
hours duration), roughly three times as much as
for anged males. The very limited sample from
Sabangau suggests the same pattern there as well.
The reason is that un anged males are frequently
unable to successfully mate-guard a female for a
long time, since the consortships are often involun-
tary, maintained only by the male, and an encoun-
ter with a more dominant male (either un anged
or anged) terminates the consortship. However,
even un anged males in Ketambe and Sabangau
achieve the majority of copulations in consortships
rather than during brief encounters. This suggests
that both male morphs attempt to mate-guard
when they get the chance, but anged males may
be more successful because they have no or fewer
competitors and they may be preferred company
by females who endure less sexual harassment in
their proximity (Fox 2002).
Even when males mate-guard a particular female
this relationship does not completely exclude oth-
ers: in Ketambe, Sabangau and Gunung Palung
anged males occasionally engage in extra-consort
copulations (ECC), that is, during a consortship
they copulate not only with their consort partner,
but also with another female (temporarily) present
in the same party. In Ketambe, most ECCs hap-
pened while the consort pair visited a large fruit
tree (most often g trees) in which other orang-
utans were already aggregated (Utami et al. 1997;
0.6
0
Flan
g
ed Unflan
g
ed
0.1
0.2
0.3
Proportion copulations
0.4
0.5
Consort
Outside
ECC
Figure 16.2 Distribution of copulations by fl anged and unfl anged
males over three different contexts (ECC, extra-consort copulation)
in Ketambe. Even though the two male morphs have a signifi cantly
different distribution over three categories (Gadj ⫽ 22.0 P ⬍0.001),
the proportion of copulations during consort vs other context did
not differ signifi cantly (Gadj ⫽ 2.07 NS).
16-Wich-Chap16.indd 23916-Wich-Chap16.indd 239 7/7/2008 4:01:05 PM7/7/2008 4:01:05 PM
240 ORANGUTANS
in the proximity of a anged male. Thus, despite a
vulnerability to harassment caused by their semi-
solitary lifestyle, orangutan females not only resist
particular mating attempts, they also can manipu-
late mating success of non-preferred males by
manoeuvering themselves under the protective
umbrella of a dominant male. This reasoning also
implies that the un anged males are in a ‘waiting-
room’ situation, making the best of the bad job.
In other words, the un anged stage is not a per-
manent alternative tactic but a transitional stage
(Utami Atmoko and van Hooff 2004).
16.3 Paternity
So far, only two studies described the genetic con-
sequences of the mating strategies of anged and
un anged males, one in Sumatra (Utami et al.
2002), and one in Borneo (Goossens et al. 2006b).
These studies used human-derived microsatellites
to estimate paternity, and showed that both unf-
langed and anged males are successful in siring
offspring. In Ketambe, paternity analysis of 11 off-
spring born over a 15-year period was carried out
and 6 out of 10 offspring could be attributed to 3
un anged males and 4 to 3 anged males in the
area at the time of conception of these offspring
(the father for the eleventh offspring could not be
two of the Bornean sites cooperate in the majority
of all copulation attempts. In both Sumatran and
most Bornean sites, except for Gunung Palung,
cooperative matings by females are signi cantly
more likely with anged males than with unf-
langed males (see Fig. 16.3). However, in most
Bornean sites, females do resist a higher propor-
tion of mating attempts by anged males than in
Sumatra, supporting the idea that females in both
Sumatra and Borneo have a preference for mating
with anged males. However, the differences in
female cooperation rates for mating with the two
male morphs are variable and for the Bornean
sites based on small numbers of females. Whether
a female resists a particular mating attempt prob-
ably depends not only on the status and morph of
the male, but also on other factors, such as female
parity and the relationship between the partners.
Females in association with a non-preferred
male can play an active role in ending the asso-
ciation, even when they cannot ‘outrun’ an agile
un anged male, who tends to intercept her when
she ees (e.g. van Schaik 2004). Females can travel
in the direction of recent long call by dominant
males and thus orchestrate a male–male encounter
(Utami and Mitra Setia 1995; Fox 2002). Fox (1998,
2002) found in Suaq a clear decrease in mating
attempts by un anged males when females stayed
100
N 164 N 353 N 43 N 15 N 42 N 52 N 179
*** *** *** ****
80
60
40
% Copulations cooperative
20
0
Ke Su
Sumatra Borneo
Tu Sa GP TP Ku
Flanged
Unflanged
Figure 16.3 Percentage of copulations by both fl anged and unfl anged males that are cooperative (i.e. non-resisted by females) in seven
sites. Numbers above columns indicate total number of copulations observed and stars indicate signifi cant difference between fl anged and
unfl anged males in Gadj -test: * P ⬍0.05; ** P ⬍0.01; P ⬍0.001.
16-Wich-Chap16.indd 24016-Wich-Chap16.indd 240 7/7/2008 4:01:05 PM7/7/2008 4:01:05 PM
ORANGUTAN MATING BEHAVIOUR AND STRATEGIES 241
These two latter points may well be related. At
Ketambe, Schürmann (1982) studied the socio-
sexual development of an adolescent female. He
noted a slow and very gradual process of develop-
ing relations between the maturing young female
and males. Young un anged males were the rst
to show interest in her and they formed volun-
tary consortships with her. The adolescent female,
however, showed a clear preference for the bigger,
anged males, in particular for the biggest one in
the area, who in turn only gradually developed an
interest in her. It took a long time, at least 5 years,
for the adolescent female to build up a relationship,
which involved consortship and mating with the
dominant local anged male using various kinds
of soliciting behaviour. It was not until the last
year before she conceived that this dominant local
anged male started to react to her elaborate pro-
ceptive behaviour with ‘male presenting’. Flanged
males appeared much more interested in older
parous females, who showed less pronounced pro-
ceptive behaviour (Schürmann, 1982). Observations
identi ed) (see Table 16.2). At least, the available
data show that most or all infants were sired dur-
ing voluntary consortships, even if these were not
with anged males.
The results from Ketambe suggest roughly equal
per capita siring success of un anged and anged
males (Utami et al. 2002), but this conclusion needs
some quali cation. First, six offspring were sired
by three (of at least six known) un anged males.
One of these un anged males sired three infants
and eventually developed SSCs and became the
dominant anged male in the area. Second, ve
of the ten offspring with an identi ed father were
from matrilines whose founding females were
ex-rehabilitants released at Ketambe in the 1970s
(see Rijksen 1978). Only one of these ve was sired
by a anged male and none by the then-dominant
anged male in the area. Third, at least four of the
six offspring sired by un an ged ma les were bo rn to
primiparous mothers, and only one anged male is
known to have sired a possibly rst-born offspring
(of a wild mother) in this sample.
Table 16.2 Overview of the data used for this chapter: sites, periods during which mating behaviour was studied, and major researchers
Ans
Pet
Chris
Her Eib
Puji
Yos Set
Gen Kel
Dominant male
Yea r bo rn
1975
1983
1987
1988 1991
1991
1992 1993
1997 1996
Jon P P * P PPP 1972–1990
Nur P P P * P * 1991–1995
Erik PAPPPP
Miki PPPPP
WP P
I P P P Old male?
Jan P1995 attempt?
Boris PP1995–?
Doba
Bobby P *
A2
Boris P * P PP * P * P PP
Bas PPPPP
Wiba P
X P *PPPP *
Dedi PP
Aldo P * P
See the Preface and Chapter 7 for descriptions of habitat. Identified sires for 10 infants born during study period in Ketambe. Infants (columns)
in bold are from wild local matrilines, others from rehabilitant matrilines. Males (rows) present in year of conception in study area; names in
capitals refer to flanged males, in lower case to unflanged males. P, present, P *, identified as sire.
16-Wich-Chap16.indd 24116-Wich-Chap16.indd 241 7/7/2008 4:01:05 PM7/7/2008 4:01:05 PM
242 ORANGUTANS
or aversion to particular males (Fox 2002). Thus,
females select their sexual partners and choose
when to cooperatively consort and mate. A close
social relationship built up during consorts seems
to be of importance for cooperative and probably
also for successful copulation. Since female orang-
utans show no visible oestrus signs, it must be the
changes in their behaviour and additionally olfac-
tory signals that induce the male to mate or at least
to cooperate (Schürmann 1982).
Genetic paternity studies in the eld (Utami
et al. 2002; Goossens et al. 2006b) con rm observa-
tions from zoos that un anged males are able to
sire offspring. However, even though more data
are needed, it appears un anged males were most
successful in siring offspring with primiparous
females. Other studies rely ing on behaviour to esti-
mate paternity suggested that the resident domin-
ant anged male fathered all or most offspring of
females within his home range (Rodman 1973b;
Galdikas 1985b; Schürmann and van Hooff 1986).
Even though all males are seen to mate (Galdikas
1978; Mitani 1985b; Utami et al. 2002) they appar-
ently differ in their timing relative to conception,
and in the Sumatran sites subordinate anged
males do not seem to be successful at all (this chap-
ter; van Schaik 2004).
The reproductive success of anged males is
made possible by the females’ preference. The
dominant anged male in an area may be able to
exclude other anged males from his immediate
ranging area, but he certainly does not exclude all
un anged males. Also, a dominant anged male
cannot prevent un anged males from formi ng con-
sortships with females in the same area. However,
his dominant position allows him to maintain a
consort, whereas un anged males may be forced
by bigger males to give up a consort.
The picture that emerges from these data indi-
cates an interaction between the three compo-
nents of sexual selection. Male–male competition
is intense, in particular among anged males.
Females prefer mating with the dominant anged
male, who is therefore likely to have the highest
siring success in the local population. They have
an aversion to mating with un anged males, espe-
cially when they are receptive, who must there-
fore resort to coercion and have a lower per capita
at Suaq (van Schaik, unpublished data) paint very
much the same picture, and concur with ndings
for chimpanzees (Muller et al. 2006), and primates
in general, that males prefer parous over nul-
liparous females (Anderson 1986). Rehabilitant
females, being better fed than their wild counter-
parts, may have matured faster (cf. Knott 2001), and
thus have become pregnant while still in the phase
of voluntary consortships with un anged males.
In conclusion, then, the high siring success of unf-
langed males may be due to a high proportion of
rstborns, especially to females of the ex-rehabili-
tant matrilines.
Furthermore it should be noted that three of
four infants sired by a anged male were offspring
of the local dominant and wild local females, the
fourth was the offspring of an ex-rehabilitant
mother. Since one to four other anged males were
present in the area during the conception periods,
this suggests a clear advantage to the dominant
anged male over non-dominant anged males in
achieving paternity.
Goossens et al. (2006b) performed a simi lar pater-
nity analysis in the 4-km2 study site in the lower
Kinabatangan oodplain. Parentage was analysed
for 16 individuals, but paternity could be assigned
for only ve offspring (compared to 10 in Utami
et al. 2002). One out of ve offspring could be attrib-
uted to an un anged male, the four others could
be attributed to anged males, but no additional
information on the mothers was available.
16.4 Discussion
16.4.1 Male–male competition and
female choice
Flanged males advertise their location by giving
long calls and can thus attract females to come to
them, whereas un anged males have to actively
travel through an area to locate potentially fertile
females. Overall, and certainly in Sumatra, the
majority of sexual interactions were coopera-
tive and occurred during a consort relationship.
However, females everywhere resist a proportion
of mating attempts. Since these resisted matings
generally occur in the absence of other males they
are assumed to re ect honest female preference for
16-Wich-Chap16.indd 24216-Wich-Chap16.indd 242 7/7/2008 4:01:05 PM7/7/2008 4:01:05 PM
ORANGUTAN MATING BEHAVIOUR AND STRATEGIES 243
but mainly with females not (yet) effectively mate-
guarded by anged males, for example because of
their unpredictable fertility (nulliparous females).
Thus, if the new interpretation of the Ketambe data
hold s (and if the observational estimates from most
other orangutan observers are correct), un anged
males have lower per capita success than the domi-
nant a ng ed ma le, but h ig he r s ucc es s t ha n t he ot he r
anged males. Paternity data from more sites and
of larger samples are needed to assess the differ-
ence in reproductive success within and between
the two male morphs, but at our present state of
knowledge the morphs do not seem to yield equal
reproductive success.
A second possibility is that the phenomenon of
developmental arrest represents a ‘waiting room
strategy’ in which anged males bide their time
until they can assume, without too much risk, the
highly contested role of anged male (Schürmann
and van Hooff 1986; van Hooff 1995). In the mean-
time they can make the best of a poor job by engag-
ing in sexual interactions, even though this offers
only limited reproductive success. This idea leaves
out one critical issue, however: most anged males
are spectacularly unsuccessful at achieving mat-
ings. To deal with this omission, van Schaik (2004)
argued that a anged male who is not dominant
over other anged males in the same area is worse
off than an un anged male in acquiring matings.
The large anged males have high costs of loco-
motion and maintenance and cannot inde nitely
maintain a consortship when the female does not
cooperate, as she does with a preferred male. If such
males give long calls to attract females they run a
high risk of a confrontation with a more dominant
anged male. Un anged males travel faster and
roam more widely (see Chapter 18), and can also
endure longer associations (e.g. Wich et al. 2006b).
They are therefore better able to gain some poten-
tially fert ile copulations through chance encounters
wit h fer ti le fem al es and by c lo sel y fo ll owi ng con so rt
pairs and engaging in sneak matings. Thus, only
when a male is likely to achieve dominance in an
area would the mat ing bene ts of SSCs outweigh its
costs. It is possible that a male may need to wait for
a long time for such an opportunity to arise, which
would explain the developmental arrest.
mating success than the dominant anged male.
Th us, f ema le prefe rence s se t the s cene f or the ma le–
male competition; if they had no preferences at all,
un anged males would be much more successful
than anged males due to their higher mobility.
At least in Sumatra, other anged males achieve
very few matings, and almost certainly sire very
few offspring. Whether this is also true for Borneo
needs to be assessed in future work.
16.4.2 Male bimaturism
Utami et al. (2002) discussed (and reject) two pre-
viously untested hypotheses, advanced to explain
male reproductive behaviour and bimaturism in
Pongo sp., and proposed a third hypothesis. The
‘range-guardian’ hypothesis (MacKinnon 1974)
as se rts t hat t he ang ed m ales ar e po st- rep rod uct ive
and defend a range in which they tolerate sexually
active un anged male relatives. This hypothesis
could be rejected because the un anged males
are not clearly related to the dominant anged
male, and the latter sires offspring. The ‘female
choice’ hypothesis asserts that anged males tol-
erate un anged males in their range because they
rely on female preference to favour anged males.
However, un anged males also sire offspring, and
all males compete heavily for access to females.
Utami et al.’s (2002) preferred hypothesis was that
the two male morphs represent coexisting alter-
native male reproduct ive strategies, ‘sitting, call-
ing, and waiting’ for anged males versus ‘going,
searching, and nding’ for un anged males.
While clearly consistent with the known data on
mating behaviour, this idea leaves unexplained
why un anged males may remain un anged for
so long.
The evolutionary strategy of delayed matura-
tion could then be understood in two ways. One
possibility is that the timing of the switch is due
to frequency-dependent bene ts. Then the two
strategies would be alternatives yielding equal
reproduct ive success in an equilibrium situation
(Galdikas 1985b; Maggioncalda et al. 1999; Utami
et al. 2002; Utami Atmoko and van Hooff 2004;
Knott and Kahlenberg 2007). However, current data
suggest than un anged males can sire offspring,
16-Wich-Chap16.indd 24316-Wich-Chap16.indd 243 7/7/2008 4:01:06 PM7/7/2008 4:01:06 PM
244 ORANGUTANS
Acknowledgements
We gratefully acknowledge the cooperation and
support of the Indonesian Institute of Science (LIPI,
Jakarta), the Indonesian Nature Conservation
Service (PHKA), the Economic Planning Unit (EPU,
Kuala Lumpur) and Sabah Wildlife Department
(SWD, Kota Kinabalu) for giving the permission to
perform research in the various research sites. We
also thank Fakultas Biologi Universitas Nasional
(Jakarta), Leuser International Foundation;
Sumatran Orang-utan Conservation Programme,
the Borneo Orangutan Survival Foundation and
its Mawas programme, Centre for International
Cooperation in Management of Tropical Peatland
(CIMTROP), the University of Palangkaraya, the
Universitas Tanjungpura (UNTAN), the NGO
HUTAN and their Kinabatangan Orangutan
Conservation Project (KOCP), and the Universiti
Malaysia Sabah (Kota Kinabalu) for their logis-
tical support. We thank WOTRO (the Netherlands
Foundation for the Advancement of Tropical
Research), Wildlife Conservation Society, L.S.B.
Leakey Foundation, A.H. Schultz Foundation,
University of Zürich, National Geographic Society,
Conservation, Food and Health Foundation,
Orangutan Conservancy, US Fish and Wildlife
Service, National Science Foundation, Wenner-
Gren Foundation, Harvard University, Cambridge
University, Primate Conservation Inc., Leverhulme
Trust, Royal Society and Darwin Initiative for the
Survival of Species for major nancial support. We
especially thank all the eld assistants, students
and other researchers who have collected data over
many years.
16.4.3 Island differences
Much of the above account is based on intensive
observations at Ketambe with additional data col-
lected at Suaq, whereas the data from the Bornean
sites is still less detailed. It is possible that some
aspects of sexual behaviour and male mating strat-
egies differ between Sumatra and Borneo, and
perhaps even within islands. A rst indication for
this is the observed difference in local ratios of
anged to un anged males between Sumatra and
Borneo (Delgado and van Schaik 2000): for every
anged male observed at Suaq or Ketambe, there
are about two un anged males, whereas for all
known Bornean sites the opposite holds, showing
roughly two known anged males for each known
un anged male. This observation suggests that
developmental arrest is certainly more prominent
among Sumatran orangutans, and perhaps even
limited to them. One possible explanation for this
is that dominant Sumatran anged males may be
more effective at mate guarding tha n their Bornean
counterparts because they can afford to maintain
longer consortships (Delgado and van Schaik 2000).
Thus non-dominant Sumatran anged males are at
a disadvantage compared to more agile un anged
males in acquiring matings. However, if Bornean
anged males cannot maintain long consorthips,
other anged males could sire offspring as well at
any given time at a particular site, whereas unf-
langed males will be virtually excluded (except
from nulliparous females). More work is needed to
establish any rm island differences, but on cur-
rent information some meaningful differences are
to be expected.
16-Wich-Chap16.indd 24416-Wich-Chap16.indd 244 7/7/2008 4:01:06 PM7/7/2008 4:01:06 PM
