Population Demography of Northern Muriquis (Brachyteles hypoxanthus) at the Estac¸ a˜ o Biolo´ gica de Caratinga/Reserva Particular do Patrimoˆ nio Natural-Felı`ciano Miguel Abdala, Minas Gerais, Brazil

AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 130:227䬢237 (2006)
Population Demography of Northern Muriquis
(Brachyteles hypoxanthus) at the Estac¸a˜o Biolo´gica
de Caratinga/Reserva Particular do Patrimoˆnio
Natural-Felı`ciano Miguel Abdala, Minas Gerais, Brazil
Karen B. Strier,1* Jean P. Boubli,2 Carla B. Possamai,3 and Se´rgio L. Mendes4
1Department of Anthropology, University of Wisconsin-Madison, Madison, Wisconsin 53706
2Conservation and Research for Endangered Species, Zoological Society of San Diego, San Diego, California, 92027
3Programa de Po´s-Graduac¸a˜o em Zoologia de Vertebrates, Pontifica Universidade Cato´lica de Minas Gerais,
Minas Gerais, Brazil, 30535-610
4Departamento de Cieˆncias Biolo´gicas, Universidade Federal do Espı´rito Santo, Vito´ria, Espı´rito Santo, 29043-065
KEY WORDS
northern muriqui; Brachyteles hypoxanthus; demography; infant mortality; sex
ratios
ABSTRACT
The 957-ha forest at the Estac¸a˜o Biolo´gica
among adults and subadults was female-biased (0.75), while
de Caratinga/Reserva Particular do Patrimoˆnio Natural-
that among immatures was male-biased (1.47). Consistent
Felı`ciano Miguel Abdala, in Minas Gerais, Brazil, supports
with expectations from mean interbirth intervals, 64.18% of
one of the largest known populations of the critically en-
adult females gave birth in 2003 and 2004. However, by
dangered northern muriqui (Brachyteles hypoxanthus). We
January 2005, only 52.31% of adult females were still carry-
combine long-term data on one group that has been moni-
ing infants <24 months of age due to unusually high infant
tored since 1982 with new data obtained on the other three
mortality. First-year survivorship among the 25 infants
groups since 2002 and 2003 to describe the demographic
born in the population in 2003 was only 76%, considerably
structure of this population, evaluate its potential for future
lower than previously documented in the longest-studied
growth, and predict how dispersal and competitive regimes
group. High female fecundity is indicative of a healthy pop-
may change in response to current demographic conditions.
ulation, but the current male bias in births will result in a
As of January 2005, the 226 individuals in the population
decline in the population growth rate within two decades,
were divided into four mixed-sex groups with 37–77 mem-
and may increase levels of male reproductive competition
bers, and an all-male unit whose eight males maintained
and alter dispersal patterns. Am J Phys Anthropol 130:
transient associations with two of the mixed-sex groups.
227–237, 2006.
V
C 2005 Wiley-Liss, Inc.
Although 51.77% of the population was female, the sex ratio
Long-term demographic data can provide insights into
The northern muriqui (Brachyteles hypoxanthus) is
the relationships among individual life histories, intra- and
one of the world’s most critically endangered primates
intergroup dynamics, and environmental variables that
(Konstant et al., 2002). It is endemic to the Atlantic For-
affect survival and reproduction (Altmann and Altmann,
est of southeastern Brazil, a biodiversity ‘‘hot spot’’ that
1979; Alberts and Altmann, 2003). Demographic data are
has been reduced to less than 9% of its original extent
also essential for evaluating the viability of small, isolated
due to severe deforestation and fragmentation (Myers
populations of endangered species (Strier, 2003a,b), and for
et al., 2000; Brooks and Rylands, 2003; Camaˆra, 2003).
distinguishing normal fluctuations that constitute demo-
graphic ‘‘noise’’ in large, healthy populations, but that are
warning signs of increased extinction risks in small, iso-
Grant sponsor: National Geographic Society; Grant sponsor: Mar-
lated populations (Lacy, 1993a,b). For example, although
got Marsh Biodiversity Foundation; Grant sponsor: Graduate School,
skewed sex ratios at birth may average out over evolution-
University of Wisconsin-Madison; Grant sponsor: Zoological Society
ary time (Silk and Brown, 2004), successive years of male-
of San Diego; Grant sponsor: Sustainable Development Program of
or female-biased birth cohorts can significantly impact lev-
Brazilian Biodiversity, Ministry for the Environment, Federal Gov-
els of reproductive competition, dispersal patterns, and
ernment of Brazil; Grant sponsor: Rufford Foundation; Grant spon-
sor: Primate Action Fund, Conservation International; Grant sponsor:
population growth rates as members of these cohorts
CAPES.
mature over ecological time (Strier, 2003c). The potential
impact of demography on behavior is particularly pro-
*Correspondence to: Karen B. Strier, Department of Anthropology,
nounced in species with slow life histories and correspond-
University of Wisconsin-Madison, 1180 Observatory Drive, Madison,
ingly long generations and lifespans (Mueller et al., 1991;
WI 53706. E-mail: kbstrier@wisc.edu
Cowlishaw and Dunbar, 2000). Stochastic fluctuations that
result in unfavorable demographic conditions can also
Received 6 May 2005; accepted 31 July 2005.
severely reduce the ability of small populations to recover
to minimal viable sizes (Reed et al., 2003), necessitating
DOI 10.1002/ajpa.20366
close monitoring of the demographic condition of popula-
Published online 19 December 2005 in Wiley InterScience
tions of endangered and critically endangered species.
(www.interscience.wiley.com).
V
C 2005 WILEY-LISS, INC.

---

228
K.B. STRIER ET AL.
Fig. 1. Map of EBC/RPPN-FMA. See text for precise coordinates. General core areas of each of four mixed-sex muriqui groups
are shown. Arrow and bold line indicate ridge top that divides Mata˜o and Jao´ sides of forest, and which is highest point in forest.
Map courtesy of Luiz G. Dias, provided by C.B. Possamai, updated here.
Although estimates of the number of northern muriquis
often used in comparative analyses of primate social evolu-
have increased from about 500 (Strier and Fonseca,
tion (e.g., Mitani et al., 1996; Nunn, 1999). Longitudinal
1996/1997) to nearly 1,000 individuals due to new discov-
data can thus provide important insights into the ways in
eries and research in other forests, all populations are
which primates adjust their behavior in response to local
isolated from one another, and none are known or esti-
demographic conditions.
mated to exceed 300 individuals (Strier et al, 2005). This
In this paper, we describe the demographic characteris-
is well below even the most conservative estimates of
tics of one population of northern muriquis that is now
minimum viable population size (Cowlishaw and Dun-
being monitored in its entirety. We review the ongoing
bar, 2000; Reed et al., 2003).
dynamic processes leading to the population’s current sta-
Northern muriquis, like their closest atelin relatives,
tus, and evaluate the population’s potential for growth.
woolly monkeys (Lagothrix spp.) and spider monkeys
We also discuss how changes in the number and composi-
(Ateles spp.), have slow life histories compared to Old
tion of social groups may affect dispersal patterns in both
World monkeys of similar size that may reflect phyloge-
females and males, and we develop predictions about how
netic effects (Strier, 1999a,b, 2003a; Nishimura, 2003).
current demographic conditions may affect future intra-
Life-history data from one of our study groups that has
and intergroup social and sexual behavior as individuals
been monitored since 1982 indicate that females typi-
in this population mature.
cally disperse from their natal groups at a median age of
6.12 years, prior to the onset of puberty (Strier and Zie-
gler, 2000), and give birth to their first infants some 2.75
METHODS
years later, when they are roughly 9 years old (Strier
Study site
et al., 2002a). The median birth interval following
infants that survive to weaning in their second year is
Our study site is the 957-ha forest at the Estac¸a˜o Bio-
36 months (n ¼ 63; updated from Strier et al., 2003),
lo´gica de Caratinga/Reserva Particular do Patrimoˆnio
and some females known to be at least in their late 20s
Natural-Felı`ciano Miguel Adbala (EBC)/RPPN FMA, lo-
and estimated to be in their mid-30s are still reproduc-
cated at North 198 440, South 418 490 in Minas Gerais,
tively active in this group. Their late age at first repro-
Brazil (see Strier and Boubli, in press). The forest is clas-
duction and long interbirth intervals make the small
sified as low, semideciduous montane forest (Rizzini,
populations of northern muriquis that persist more vul-
1979), mostly secondary, dominated by trees from the
nerable to unfavorable demographic conditions than pri-
Leguminosae, Euphorbiaceae, and Flacourticaceae fami-
mates with faster life histories, or those with comparably
lies (Boubli et al., 2003). It is surrounded by pasture and
slow life histories but living in larger populations.
plantations, and isolated from other forests that support
Demographic fluctuations can also affect the number,
muriquis, resulting in a closed population. Rainfall pat-
size, and composition of social groups, and therefore have
terns are strongly seasonal, with a distinct dry season
implications for levels of intragroup reproductive competi-
from May–September (Strier et al., 2001).
tion and patterns of intergroup transfer. Despite the sensi-
The landscape is defined by irregularly shaped hills
tivity of these variables to demographic fluctuations and
and valleys, with altitudes ranging from 400–640 m above
their corresponding potential to change significantly over
sea level (Strier, 1987). The highest point is found along a
time and to vary both between and within populations of
ridge that divides the forest approximately in half
the same species (Strier, 2003c; Jones, 2005), they are
(Fig. 1). Only two muriqui groups (Jao´ and Mata˜o) were
American Journal of Physical Anthropology—DOI 10.1002/ajpa

---

NORTHERN MURIQUI POPULATION DEMOGRAPHY
229
TABLE 1. Onset of population-wide monitoring
TABLE 2. Sightings of Mata˜o female emigrants in Jao´ and
other groups from 1999 census to June 2002
Systematic monitoring
Group
Initial counts
initiated
Sightings in Jao´
group as of
First sightings
Current
Mata˜o
June 1982
June 19831
Individual1
census, 1999
as of June 2002
group
Jao´
June 19822, August 19993
June 20024
Nadir
June 20015
June 2002
Brh
August 4, 1999
June 20023
Jao´
M2
January 19926, August 19993
June 20037
Cat
August 4–5, 19992
June 20022
Jao´
Den
August 3–5, 1999
June 20023
Nadir
1 Same age-sex classes of 22 members present in June 1982
Fab
August 4–5, 1999
June 20023
Jao´
were also present in June 1983, when individual identifications
Hls
August 3–4, 1999
June 20023
Jao´
were made.
Mrn
August 5, 1999
June 20022
Jao´
2 This group was known to be present previously (Valle et al.,
Nd
August 4–5, 19992
June 20022
Nadir
1984), and is therefore considered one of original groups in this
Pri
August 3, 1999
June 20023
Jao´
forest.
Rs
August 3–4, 1999
June 20025
Jao´
3 Cited during population-wide census (Strier et al., 1999).
Flr
June 20024
Jao´
4 Individual identifications with systematic monitoring initiated.
Brs
June 20022
Nadir
5 Initially part of Jao´ group, identifiable females were sighted
Nn
June 20025
Nadir
together during 1999 census. By June 2002, members of this
Kk
June 20026
Jao´
group routinely associated with one another, separate from
other members of Jao´ group.
1 Female emigrants from Mata˜o group who were sighted by
6 Initially considered a subgroup of Jao´ group that periodically
experienced observers in same group with other muriquis in Jao´
exploited Mata˜o group’s home range (Strier et al., 1993). Six
on days or months indicated.
males from Jao´ periodically made incursions into Mata˜o group
2 Carrying dependent infant or associated with 2–3-year-old
and copulated with Mata˜o females between 1988–2001 (Strier,
infant on first sighting or resighting.
1994, 1997).
3 Nulliparous at first sighting, but have subsequently given
7 Opportunistic monitoring was initiated in September 2002.
birth.
4 Nulliparous from first sighting to present.
5 Uncertain condition at first sighting or resighting, but have
present when systematic monitoring began in 1982,
subsequently given birth.
named after the Jao´ Valley located on the northern side of
6 Nulliparous at first sighting; last seen November 2002, and
this ridge, and the Mata˜o Valley on its southern side,
presumed dead.
respectively (Valle et al., 1984). Members of the Jao´ group
were observed in 40% of the Mata˜o group’s home range in
June 1983, but the first systematic survey of muriquis in
1983–1984 (Strier, 1987), but the Mata˜o group has never
the Jao´ area was conducted during a 5-day census in
been observed in the Jao´ Valley. Beginning in the late
August 1999 (Strier et al., 1999). In total, 73 individuals
1980s, the Mata˜o group began shifting its home range to
in Jao´ were sighted at that time, including nine natal
the south (Strier et al., 1993). About 40% of the Mata˜o
Mata˜o females who had previously transferred into this
group’s current home range overlaps with that of a third
group and were still recognizable to observers who had
group (Mata˜o 2, or M2), whose original members split
known them in their natal group (Table 2). In 2002, a
from the Jao´ group and whose occupancy of a part of the
trail system was developed in Jao´ Valley, and two experi-
Mata˜o group’s original home range may be responsible for
enced students who had worked with the Mata˜o project
the Mata˜o group’s shift to the south (Strier et al., 1993;
began the process of identifying all Jao´ muriquis. Age
Dias and Strier, 2003). The Jao´ group underwent a second
classes or ages to within 1 year for infants were esti-
split, resulting in a fourth group (Nadir). Systematic anal-
mated for Jao´ group members, based on visual physical
yses of home-range overlap among all of the muriqui
comparisons with Mata˜o group members of known age
groups in this forest are currently underway, and will be
(described below). In total, 13 natal Mata˜o females,
presented elsewhere (Boubli et al., 2005, unpublished
including the 9 females sighted during the census and 4
results).
who subsequently immigrated, were present in the Jao´
Subjects
group when systematic monitoring of this group was ini-
tiated in June 2002. Known females who were nullipar-
Northern muriquis can be individually recognized by
ous at the time, as evident from the visibly undeveloped
their natural markings and facial features, such as fur
condition of their nipples, and have since given birth,
color and patterning, ear shape, and face shape and pig-
provide definitive ages at first reproduction for dispers-
mentation. Their sex can usually be determined within a
ing females in this species (Strier et al., 2002a).
week or so of birth, based on the shape and positioning
At the outset of the Jao´ study in 2002, observers noted
of their genitalia. Positive identifications of all muriquis
that members of the Jao´ group tended to associate in
were obtained within a few months of the onset of sys-
distinct subgroups, with the familiar females from the
tematic monitoring of each of the four groups, after
Mata˜o group split by their associations with different
which all individuals were followed for a minimum of
subsets of Jao´ females and males. By 2003, these sub-
2 weeks per month (Table 1).
groups had split entirely, with the Jao´ group remaining
Demographic and life-history data have been main-
in its valley, and the newly formed Nadir group spending
tained on all members present in the Mata˜o group since
most of its time on the slopes and ridges to the south of
June 1982, when long-term studies were initiated. The
Jao´ Valley (Boubli et al., 2005, unpublished results). In
ages of all individuals born in the Mata˜o group since
addition, a subset of 14 adult and subadult males from
1982 are known, including those of natal Mata˜o females
Jao´ appeared to move between these two mixed-sex sub-
who dispersed and are now being monitored in the other
groups. By December 2004, one of these males had dis-
groups.
appeared and is presumed to have died, and 5 of the
The Jao´ group has been opportunistically observed
remaining 13 males were associating exclusively with
whenever any of its members were encountered since
the Nadir group (Boubli et al., 2005).
American Journal of Physical Anthropology—DOI 10.1002/ajpa

---

230
K.B. STRIER ET AL.
Systematic monitoring was initiated on the M2 group
their successful emigrations were confirmed. As a result,
in June 2003. This group was first observed in the late
we can only confirm the survival of 1 of 10 natal Mata˜o
1980s, when its original members were sighted repeat-
females who emigrated between 1987–1993 because of
edly in what had been the northern part of the Mata˜o
her highly distinctive pelage. However, 17 of 22 natal
group’s home range (Strier et al., 1993; see above). The
Mata˜o females who emigrated between 1994–2004 were
process by which the M2 group members fissioned from
positively confirmed in one of the other groups, and 16
the Jao´ group was similar to the pattern observed dur-
were still alive as of January 2005 (Table 2). Since June
ing the more recent establishment of the Nadir group,
2002, all natal dispersals from all four mixed-sex groups
and is described in greater detail below.
have been followed.
Adult females carrying new infants in the Mata˜o
group in June 1982 are estimated to have been at least
Birth records and age determinations
7–9 years old at the time, based on the minimum age at
Ages are known to within a few days in most cases,
first reproduction in this population (Strier and Ziegler,
and a few months in the remaining cases for all muri-
2000; Martins and Strier, 2004). Four of the six ‘‘orginal’’
quis born in each of the groups since the outset of sys-
mothers in this group are still alive, and are therefore
tematic monitoring (Table 1). We use the longitudinal
presumed to be 30 years. In addition, two visibly nul-
data of known ages of Mata˜o group members to calculate
liparous adult females who were also present in the
age at first reproduction for Mata˜o females after they
group in 1982 are still alive.
disperse, and to estimate the age or age classes of both
Only one of six adult males present in the Mata˜o
males and females of the other groups. Briefly, infants
group in 1982 is still alive. Two old juvenile males
are carried ventrally during their first 3–6 months, and
present in 1982 have also died. The two next oldest adult
dorsally until weaning, which occurs during their second
males in the Mata˜o group were both newborns in 1982,
year, when they are visibly larger and also spend more
and are therefore known to be about 22.5 years old as of
of their active time out of contact with their mothers
January 2005. The ages of all other younger males in
(Odalia Rı´moli, 1992). Infants continue to travel closely
this group are known to within a few days (n ¼ 34) or
and sleep with their mothers until a new sibling is born,
within 6 months (n ¼ 2), and no males have emigrated
usually during their third year. Throughout their fourth
from the Mata˜o group since the study began.
year, immature muriquis of both sexes are still visibly
smaller than adults and subadults in both body size and
Analyses
genitalia, and have not been observed to copulate. Thus,
immatures in each of the groups at the time monitoring
We use three sets of data to describe the population’s
began could be confidently classified as either dependent in-
demographic characteristics: 1) number of individuals in
fants (born from 2003–2004, and therefore <2 years old) or
each age-sex class; 2) percent of adult females who gave
independent immatures (born from 2000–2002, and there-
birth in 2003 and 2004 and are still carrying infants;
fore 2–5 years old) as of January 2005. Unless otherwise
and 3) first-year survivorship among the 2003 infant
noted, all individuals whose known or estimated births
cohort, the only cohort for which we have complete, pop-
occurred from 2000–2004 are considered ‘‘immature.’’
ulation-wide data. We distinguish between the popula-
By age 5 years, both males and females are visibly
tion-wide and group-specific values for these demo-
approaching or at their full adult size, and their genita-
graphic variables, and examine whether demographic
lia appear adult-like in both size and development. The
variables relate to group size and history. We also de-
minimum age at first copulation with ejaculate is 5 years
scribe the process of new group formation and how the
for natal Mata˜o males (Strier, 1996), similar to the age
increase in dispersal options created by these new groups
at first copulation of a Mata˜o female who remained and
may affect dispersal patterns.
reproduced in her natal group (Martins and Strier,
To evaluate the prospects for the future growth and
2004). Although most females do not give birth to their
viability of this population, we compare the population-
first infants until they are nearly 9 years old, they usu-
wide data on mortality and sex ratios with those com-
ally begin to copulate 1–2 years earlier (Strier and Zie-
piled from a previous population viability analysis (PVA),
gler, 2000; Strier et al., 2002a; see below). We consider 5
based on data from the Mata˜o group from 1982–1993
years to be the minimum age at sexual maturity in this
(Strier, 1993/1994; Rylands et al., 1998). We also consider
population, and classify both males and females >5 years
the effects of additional groups and sex ratios on behavior.
of age as subadults prior to the onset of copulation, and as
adults once they begin to copulate. Unless otherwise
RESULTS
noted, subadults and adults are grouped together in our
Population size and composition
analyses because of our inability to distinguish between
these age classes based on physical characteristics alone.
The 226 muriquis present in this population as of Jan-
Thirty-one of 36 natal Mata˜o females who survived to
uary 2005 were distributed among four mixed-sex groups
age 5 years are known to have emigrated, defined only
and one male unit with 8 members. Adults and sub-
after they are observed in association with members of
adults account for 67.26% of the total population, and
another group (updated from Strier and Ziegler, 2000).
66.05% of 218 individuals in the mixed-sex groups.
Muriqui facial features change gradually as they age,
Immatures account for nearly one-third of the population
with their all-black faces typically becoming increasingly
(32.74%), with independent immatures (17.70%) and
less pigmented, with distinctive patterns of pink and
dependent infants (15.04%) similarly represented.
white mottling that help us to recognize them individu-
The population sex ratio (males/females) is slightly
ally. However, these changes occur to varying degrees at
female-biased (0.93), with females accounting for 51.77%
variable rates, and it has been difficult to confidently re-
of the population. Although the sex ratio of adults and
identify females who emigrated during the early years of
subadults is female-biased (0.75), the sex ratios of inde-
the study but were not followed systematically after
pendent immatures (1.50) and infants (1.47) are male-
American Journal of Physical Anthropology—DOI 10.1002/ajpa

---

NORTHERN MURIQUI POPULATION DEMOGRAPHY
231
pendent infants, which accounts for 20.75% of the Jap´
group, or nearly twice the 11.69% representation of
infants in the Mata˜o group (mean ¼ 15.90 6 4.08%,
median ¼ 15.58%, n ¼ 4). Independent immatures
account for 15.69% of the Nadir group, to 21.62% of the
M2 group (mean ¼ 18.59 6 2.44%, median ¼ 18.52%,
n ¼ 4). There are no clear patterns between group age
structure and either group size or group longevity across
the four mixed-sex groups (Table 3).
Sex ratios in the mixed-sex groups are more variable
than their age class distributions (Table 3). Sex ratios
average 0.84 6 0.15 (median ¼ 0.83), and range from
female-biased in the M2 group (0.68) to near-parity in
the Mata˜o group (1.03). Sex ratios are also currently
female-biased in the Jao´ (0.77) and Nadir groups (0.89),
but will approach or exceed parity depending on how
members of the male unit ultimately distribute them-
selves between these two groups.
Fig. 2. Population sex ratio by age class. Based on popula-
Adult and subadult sex ratios are female-biased in all
tion as of January 2005. Bars at left show entire population
of the mixed-sex groups, ranging from 0.52 in the Jao´
independent of age class: All, adults/subadults (5 years), and
group to 0.85 in the M2 group (mean
immatures (<5 years). Bars at right show immatures further
¼ 0.66 6 0.16,
divided according to birth years.
median ¼ 0.64, n ¼ 4). Conversely, sex ratios among
immatures are male-biased in all but the M2 group,
where they are female-biased among both independent
biased. Overall, females account for 57.24% of adults
immatures (0.33) and dependent infants (0.67). The
and subadults, but only 40.54% of immatures (Fig. 2).
average sex ratio of independent immatures and infants
These changes appear to reflect a population-wide shift
combined in the four mixed-sex groups is 1.62 6 0.92
from female- to male-biased infant sex ratios (see below).
(median ¼ 1.81).
Forty-three of 67 sexually active females (64.18%) in
The percentage of sexually active females who gave
the population gave birth during the 2003 and 2004
birth in each group during the 2003 and 2004 birth sea-
birth seasons combined. Two of these females were old
sons ranged from 55.56–81.25%. The percentage of sur-
mothers from the Mata˜o group who disappeared and are
viving adult females still carrying their infants as of
presumed to have died with their infants, one born in
January 2005 declined in all four groups, though to
2003 and the other in 2004. Only 34 of 65 (52.31%) sur-
varying degrees (Table 4).
viving adult females were still carrying their infants as
First-year survivorship for infants born in 2003 was
of January 2005.
100% in both the M2 (n ¼ 4) and Nadir (n ¼ 5) groups,
Six of 25 infants born in 2003 died before reaching 12
but lower in both the Jao´ group (6/8 births, or 75%) and
months of age, including one of the infants that disap-
Mata˜o group (4/8 births, or 50%; 4/7 infants born to
peared with his old Mata˜o mother. First-year survival
mothers who survived, or 57.14%). Two of the three sur-
among the 2003 birth cohort was 76% when the 19 sur-
viving mothers who lost 2003 infants were among the
viving infants of all 25 mothers are included, and
oldest females in the Mata˜o group, whereas all four of
79.17% when only the 24 surviving mothers are in-
the surviving infants in this cohort had younger mothers.
cluded. Three of 18 infants born in 2004 had also died by
As of January 2005, the only mortality among the six
January 2005, including a second infant that disap-
2004 births in the Mata˜o group involved an infant who
peared with her old Mata˜o mother. One of the surviving
disappeared with her old mother. One of the two infants
mothers who lost 2003–2004 infants was an old Mata˜o
born in 2004 in the M2 group (50%), and one of the five
female whose infant died within a week of its birth,
2004 Nadir group infants (20%), have already died. No
before its sex could be confirmed. Of the 8 other first-
mortalities have yet occurred among the five infants
year mortalities in these cohorts combined, 6 were male,
born in the Jao´ group in 2004.
and 2 were female. Thus, population-wide sex ratios at
All of the first-year mortalities of 2003 infants of
birth were even more male-biased than among surviving
known sex in the Mata˜o group involved males. In the
infants (birth vs. survivor sex ratios, 2003 infants, 1.67
Jao´ group, by contrast, first-year survival of the 2003
vs. 1.38; 2004 infants, 1.57 vs. 1.12; and 2003–2004
cohort was higher among males (80%, n ¼ 5 births) than
infants combined, 1.62 vs. 1.25).
females (66.67%, n ¼ 3 births).
All 14 infants born in the Mata˜o and Jao´ groups in
Variation in group size and composition
2002 survived to 12 months of age. However, only 5 of
The size of mixed-sex groups averaged 54.5 6 16.60
the 7 Mata˜o yearlings (71.43%) and 6 of the 7 Jao´ year-
members (median ¼ 52, range ¼ 39–77, n ¼ 4). With 77
lings (85.71%) survived to 24 months of age. Although
members, the Mata˜o group is 31–50% larger than any of
we only have yearling survivorship data on 2 of the 4
the other three mixed-sex groups. Despite the twofold
groups, their mortality levels in 2003–2004 (when they
size difference between the smallest and largest mixed-
were 12–24 months old) were similar to those of infants
sex groups, the representation of different age classes in
born in 2003 during their first year of life.
these groups is quite similar (Table 3). The percentage of
adults and subadults ranges from a low of 60.38% in the
Process of new group formation
Jao´ group to a high of 70.13% in the Mata˜o group (mean ¼
65.17 6 3.99%, median ¼ 65.09%, n ¼ 4 groups), and
In contrast to the Mata˜o group, which has not fis-
appears to be inversely related to the representation of de-
sioned during the 22.5 years it has been monitored, the
American Journal of Physical Anthropology—DOI 10.1002/ajpa

---

232
K.B. STRIER ET AL.
TABLE 3. Group size and composition
Mata˜o
M2
Nadir
Jao´
Male unit
Population
Age class
N
%
N
%
N
%
N
%
N
%
N
%
Adults and subadults
54
70.13
24
64.86
34
66.67
32
60.38
8
100
152
67.26
2000–2002 Survivors1
14
18.18
8
21.62
8
15.69
10
18.87
40
17.70
2003–2004 Survivors1
9
11.69
5
13.51
9
17.65
11
20.75
34
15.04
Total
77
100.0
37
100.0
51
100.0
53
100.0
8
100
226
100.0
Sex ratio
N
SR
N
SR
N
SR
N
SR
N
SR
N
SR
Ad and subadult M
23
0.74
11
0.85
12
0.55
11
0.52
8
65
0.75
Ad and subadult F
31
13
22
21
87
2000–2002 M1
10
2.50
2
0.33
6
3.00
6
1.50
24
1.50
2000–2002 F1
4
6
2
4
16
2003–2004 M1
6
2.00
2
0.67
6
2.00
6
1.20
20
1.43
2003–2004 F1
3
3
3
5
14
All male
39
1.03
0.68
24
0.89
23
0.77
8
109
0.93
All female
38
27
30
117
1 Values shown based on survivors as of January 2005. See text for numbers of births in these cohorts and mortality in 2003–2004.
TABLE 4. Female reproduction in 2003–2004
N infants
N adult
N infants
N adult
Group
born1
females
%
surviving2
females2
%
Mata˜o
14
23
60.87
9
21
42.86
M2
6
10
60.00
5
10
50.00
Nadir
10
18
55.56
9
18
50.00
Jao´
13
16
81.25
11
16
68.75
Population
43
67
64.18
35
65
52.31
1 Births in 2003 and 2004 are combined. See text for annual breakdown.
2 Infants and adult females surviving as of January 2005. See text for annual breakdown of mortalities.
Jao´ group has undergone two distinct fissioning events:
and Nadir groups, where they also copulate (Tokuda
the first when the M2 group formed between 1988–1991
et al., unpublished results).
(Strier et al., 1993), and the second when the Nadir
group was beginning to form in 2002 (Boubli et al.,
Female dispersal
2005). In both cases, a distinct subset of adult females,
accompanied by dependent young and independent
Initially, all natal Mata˜o females who emigrated did so
immatures, began associating with one another with
into the Jao´ group, and all females who immigrated into
increasing exclusivity and spatial segregation from their
the Mata˜o group could only have been born in the Jao´
original groups, and a subset of adult males began to
group. Assuming that females transfer at a median age
maintain transient associations with different mixed-sex
of 6 years, the earliest a natal M2 group female could
groups (Fig. 3).
have transferred would have been in 1994. Nonetheless,
In 1988, when the M2 group was being established, 6–
the five females who migrated into the Mata˜o group in
8 adult and subadult males from the Jao´ group also
1989–1990 could have been among those who were born
began to make routine incursions into the Mata˜o group’s
in Jao´ and fissioned with their mothers to form the M2
home range, where they associated with Mata˜o females
group before dispersing into the Mata˜o group.
for days at a time and accounted for up to 12% of all cop-
Dispersal options for prepubescent females transfer-
ulations involving Mata˜o females in some years (Strier,
ring out of their natal groups have increased with the
1994, 1997). Their copulation frequencies with Mata˜o
number of groups in the forest (Fig. 3). One immigrant
females declined by 1997, and the last observed copula-
female associated sporadically with the Mata˜o and M2
tion between one of these males and a Mata˜o female was
groups from May 1998–October 1999, but by November
in September 2002. Five of the original transient males
1999 she had joined the M2 group, where she has
are still alive and are now permanently associated with
remained and reproduced. Two natal Mata˜o females
the M2 group. The other three transient males have not
were similarly sighted traveling with multiple groups
been sighted in any of the extant groups by observers
before settling together in the same one. The first of
who were familiar with them, and these males are there-
these females emigrated in December 2003, and was
fore presumed to have died.
sighted with the Nadir group, and then the Jao´ group,
When the Nadir group began to travel independently
and traveled briefly with the male unit before returning
of the Jao´ group, a subset of 14 adult and subadult
to the Nadir group, where she has remained. The second
males was also observed to maintain transient associa-
natal Mata˜o female emigrated in January 2004 to the
tions with both of these groups. By December 2004, at
Jao´ group, where she encountered her natal group col-
least five of the males were associating exclusively with
league and then accompanied her with the male unit
the Nadir group. One of the remaining males has disap-
until she also joined the Nadir group.
peared, and the 8 males now comprising the male unit
Of the 12 females who dispersed since June 2002, 5
still maintain transient associations with both the Jao´
emigrated from the Mata˜o group, 3 emigrated from the
American Journal of Physical Anthropology—DOI 10.1002/ajpa

---

NORTHERN MURIQUI POPULATION DEMOGRAPHY
233
maternally related brothers and most likely fathers also
reside.
Female age at first reproduction
Age at first reproduction for dispersing females can
only be calculated for six natal Mata˜o females who were
visibly nulliparous when population-wide monitoring
began, and have since given birth in their new groups.
Mean age at first reproduction for these six females is
9.88 6 0.97 years (median, 9.75; updated from Strier
et al., 2002a). Age at first reproduction for two females
who remained and reproduced in their natal Mata˜o
group was calculated at 7.25 years for one female, and
estimated at 7.5 years for the other (Martins and Strier,
2004). A third Mata˜o female has since been observed
copulating in her natal group at age 7 years, and she is
currently mating during what is her second mating sea-
son. She will be roughly 9 years of age if she reproduces
during the 2005 birth season, or similar to the median
Fig. 3. History of new group formation. First recognition of
age at which emigrant females have given birth to their
each group is shown by vertical bars, with M2 and Nadir groups
first infants in their new groups.
originating with members of Jao´ group. Solid horizontal lines
indicate first known female transfers (circles), with total fre-
The latency from natal group transfer to first repro-
quencies shown and coded according to their known natal
duction has averaged 3.86 6 0.69 years (median ¼ 4.0,
groups. Dotted lines indicate both cases in which subsets of
minimum–maximum ¼ 2.67–5.42 years) for the six
males (triangles, with original numbers shown) fissioned from
Mata˜o females whose complete reproductive histories
Jao´ group to form transient groups, as described in text. Males
after dispersing are known. This latency is longer, but
who ultimately joined a mixed-sex group are shown by arrows
within the range of variation, of females who gave birth
to those groups; question mark indicates one male suspected of
to their first infants between 1.58 to nearly 8 years after
having returned to his group of origin. *We cannot confirm
immigrating into the Mata˜o group (mean ¼ 2.34 6 6.3,
groups of origin (natal or otherwise) for 10 of 19 females who
median
immigrated into Mata˜o group after formation of M2 group and
¼ 2.38, n ¼ 8 precise dates; values are slightly
prior to onset of systematic monitoring of this group. **At least
higher, but still under 3 years, when approximate dates
four natal Mata˜o females who initially transferred into Jao´ sub-
of their immigrations or first births are included;
sequently fissioned with Nadir group at time of its formation; at
updated from Strier and Ziegler, 2000).
least two natal Mata˜o females may also have joined M2 group
after its formation. See text for additional discussion of Mata˜o
Demographic changes
emigrants currently being monitored.
Key demographic features of the population today dif-
fer from those of the Mata˜o group during its first decade
Jao´ and M2 groups each, and 1 emigrated from the
(Table 6). Although the percentage of females who repro-
Nadir group (Table 5). Twice as many females (n ¼ 5)
duced is comparable, the immature sex ratio has shifted
immigrated into the Mata˜o group than the 2.33 females
from female- to male-biased. In addition, infant mortal-
expected if the seven dispersing females from the other
ities during 0–1 and 1–2 years have risen for both males
three groups had distributed themselves equally among
and females.
their three non-natal, mixed-sex group options.
The total increase in size of the Mata˜o group, from 22
As in the past, all female emigrants were nulliparous
members in 1982 to 77 members in 2005, and the 54%
at the time they dispersed from their natal groups.
increase between 1993–2005, occurred under more favor-
Although our time depth on dispersal among the other
able demographic conditions than those that characterize
groups is still limited, there have been no cases in which
either this group or the population today. Extrapolating
parous females transferred into or out of the Mata˜o
from the maximum known density of 0.3 muriquis/ha
group since 1982. Females often, but not always, dis-
(Strier and da Fonseca, 1996/1997) to the 957 ha of for-
perse from their natal groups following intergroup en-
est available at the EBC today yields a total of 320 muri-
counters. Females also sometimes disperse in cohorts,
quis, or a maximum estimated population that is about
although not always with other females their own age.
42% larger than its current size.
Two of the five sets of maternal sisters (2–3 sisters each)
that dispersed from the Mata˜o group and are now being
DISCUSSION
followed reside together in the Jao´ and Nadir groups,
Demographic status
respectively. Two of the three maternal sisters of a third
set reside in the Jao´ group together, without their
The quadrupling of the EBC muriqui population over
younger sister, who joined the Nadir group. The other
the past 22.5 years has been accompanied by consis-
two sets of maternal sisters joined different groups from
tently high female fecundity evident across all four
one another, so that one of each set is now in the Jao´
mixed-sex groups. Consistent with their average 3-year
group and the other is now in the Nadir group. None of
birth intervals, nearly two-thirds of all sexually mature
the known daughters of natal Mata˜o females living in
females gave birth in 2003–2004. However, both in-
the other groups are old enough to disperse, so we do
creases in infant mortality and the increasingly male-
not yet know whether females will disperse into, or
biased immature sex ratios suggest that the high rate of
avoid, their own mothers’ natal groups where their
population growth that has characterized these muriquis
American Journal of Physical Anthropology—DOI 10.1002/ajpa

---

234
K.B. STRIER ET AL.
TABLE 5. Female transfers since June 2002
Natal group (N)
ID (date emigrated)
Sequence
Joined
Mata˜o (5)
Ric (March 2003)
?
Jao´
Ins (December 2003)
Nadir?Jao´?male unit
Nadir
Clr (January 2004)
Jao´?male unit
Nadir
Fd (May 2004)
Direct
Jao´
Rf (December 2004)
Direct
Nadir
Jao´ (3)
SS (November 2002)
?
M2
VN (March 2003)
?
Mata˜o
MD (February 2004)
Direct
Mata˜o
M2 (3)
TN (March 2004)
Direct
Mata˜o
MY (May 2004)
Direct
Mata˜o
KI (July 2004)
Direct
Mata˜o
Nadir (1)
Marg (November 2004)
Direct
M2
TABLE 6. Demographic comparisons of Mata˜o group over time
curred during this year. Analyses of long-term phenologi-
and present population
cal monitoring will permit us to evaluate whether muri-
Mata˜o,
Mata˜o,
Population,
qui food availability differed in 2003–2004 compared
Variable
1982–19931
2002–20042
2002–20042
to other years (Boubli et al., unpublished results). Pre-
dators were the suspected cause of mortality of two
Proportion of
0.366
0.667
0.618
13-month-old infants in the Mata˜o group previously
males at birth
(Printes et al., 1996), but we do not yet know if local
% females
31%
30.4%
32.1%
predator populations have increased in recent years.
reproducing
annually
Future studies of potential muriqui predator popula-
Mortality
tions, especially of felines such as ocelots, which have
0–1 years
0 (M), 0 (F)
0.5 (M), 0 (F)
0.27 (M), 0.11 (F)
been sighted in this forest, will provide additional in-
1–2 years
0 (M), 0 (F)
0 (M), 1.0 (F)
0 (M), 0.43 (F)3
sights into possible sources of infant mortality.
1
While more data are needed to evaluate whether
Data from Strier (1993/1994).
2
infant survivorship is declining, there is no doubt of the
Average of 2003 and 2004, except for mortality at age
shift from female- to male-biased sex ratios among
1–2 years which uses 2002 birth data.
3 Combined 2002 birth data from Mata˜o, Jao´, and Nadir groups
immatures. Female-biased infant sex ratios with high
only.
survivorship emerged as a key determinant of the steady
growth that previously characterized the Mata˜o group,
and simulations with varying infant sex ratios demon-
since 1982 will decline even if other variables, such as
strated a decline in the rate of increase within 20 years,
female fecundity, remain constant.
as infants matured to reproductive age (Strier, 1993/
We are cautious about overinterpreting what appears
1994).
to be a decline in infant survivorship, because our longi-
The recent demographic shift toward increased infant
tudinal perspective is limited to the survivorship of pre-
mortality might reflect density-dependent factors if the
vious cohorts of infants in the Mata˜o group (Strier, 1993/
actual carrying capacity of the EBC forest is lower than
1994; Strier et al., 2001), when it was smaller and repro-
current estimates based on maximum known densities
ductive females were younger than they are at present.
for muriquis. While there is no evidence that day ranges
Although both group size and maternal age effects on
have increased with group size, the Mata˜o group’s home
infant survivorship have been documented in other spe-
range is more than 300 ha, or roughly twice as large as
cies (e.g., Gould et al., 2003; Pochron et al., 2004), we
it was when the group was much smaller (Dias and
cannot evaluate these effects on muriqui survivorship to
Strier, 2003). Although we do not yet know whether
12 and 24 months because of the comparatively brief
their home range expansion has affected their habitat
period of time that 3 of the 4 groups in this population
quality or diets, it is consistent with a need for an
have been monitored. First-year survivorship of the 2003
increased food-supply area. This expansion, as well as
birth cohorts was higher in the two smaller and younger
the increase in the number of groups and high levels of
groups (M2 and Nadir), but the effects of maternal age
home-range overlap among them, may have implications
on infant mortality can only be inferred from the high
for accurately estimating the number of muriquis that
proportion of old Mata˜o group mothers who lost their
this forest can support. Moreover, even if the population
infants, because no comparable data on maternal ages in
can achieve its maximum estimated size of 320 muriquis,
the other groups are yet available. Moreover, while the
it would still be highly vulnerable to stochastic demo-
higher mortality among 2003 infants in the Mata˜o and
graphic events, such as those that may result in male-
Jao´ groups might suggest some effects of group size or
biased sex ratios. Ongoing efforts to increase the size of
longevity, there have already been mortalities among the
the forest through regeneration of corridors are neces-
2004 infants in both of the smaller and younger groups.
sary to ensure sufficient suitable habitat to sustain a
It is worth noting that although first-year survival of
population large enough to rebound from random fluctu-
the 2002 infants born in the two groups for which we
ations in demographic conditions (Strier and Boubli, in
have data was 100%, this cohort experienced comparable
press).
mortality when they were yearlings, as did the 2003
infants during their first year of life (i.e., in 2003–2004).
Behavioral consequences of demographic shifts
Although it may be only a coincidence, it is also possible
that ecological variables, such as subtle shifts in food
The increase in number of groups in the forest may
availability or an increase in predation pressure, oc-
reduce the risks of close inbreeding, affect dispersal pat-
American Journal of Physical Anthropology—DOI 10.1002/ajpa

---

NORTHERN MURIQUI POPULATION DEMOGRAPHY
235
terns, and increase levels of male-male competition
group over the past two decades coincided with favorable
(Strier et al., 1993). Previously, with only two groups,
demographic conditions in which adult sex ratios were
dispersing females had no option other than to transfer
strongly female-biased (Strier, 1992; Strier et al., 2002b).
into their mothers’ natal groups, where their maternally
As adult sex ratios become increasingly male-biased, lev-
and paternally related uncles would still reside (Strier,
els of male-male competition over access to receptive
2004a,b). With the additional dispersal options available
females might be expected to increase. Higher levels of
now, females can transfer into groups where they are
male competition could be alleviated if males transfer,
less closely related to potential mates.
either individually or in cohorts, into new groups with
It is not clear, however, why dispersing females appear
more favorable sex ratios than those in their natal
to prefer to join particular groups and avoid others, at
groups. Indeed, the high overlap between the newly
least among the mixed-sex groups that have established
established Nadir group and that of the male unit sug-
home ranges and more permanent associations compared
gests that males are monitoring female availability and
to the all-male unit. If females prefer smaller groups
adjusting their behavior in response to group sex ratios
over larger ones, they should avoid transferring into the
(Boubli et al., unpublished results).
Mata˜o group. Yet 5 of 7 females who emigrated from
Increased levels of male competition due to unfavora-
smaller groups since the onset of population-wide moni-
ble sex ratios might also result in changes in the social
toring transferred into the large Mata˜o group instead of
dynamics among males who remain in their natal groups
one of the other two smaller, non-natal mixed-sex groups
with fewer mating opportunities (Strier, 2003c). Just as
in the population. Indeed, the number of dispersing
high mortality among male olive baboons at Masai Mara,
females who joined the Mata˜o group was twice as high
Kenya, may have contributed to a relaxation of their
as expected if dispersal patterns had been random, sug-
hierarchical, agonistic interactions (Sapolsky and Share,
gesting that these females either prefer large groups or
2004), increasingly male-biased sex ratios in the EBC
perceive other advantages to membership in the Mata˜o
muriquis could reduce the tolerant, egalitarian relation-
group. Potential advantages might involve ecological var-
ships that have characterized their relationships to date
iables, such as the Mata˜o group’s larger home range or
(Strier et al., 2002b), and increase their levels of overt,
habitat quality, or social variables, such as the size or
agonistic competition. We have not observed any such
composition of the group, but evaluating these alterna-
changes in male social behavior yet; nor do we expect
tives will require continued monitoring to determine
that any changes will occur while adult sex ratios con-
whether female fitness varies across groups.
tinue to be female-biased. It is only when the current
The four mixed-sex groups that now occupy this forest
cohort of immatures reaches adulthood that adult sex
may also create new dispersal options for otherwise philo-
ratios will change and levels of male-male competition
patric males, without necessarily incurring new risks of
might increase. Changes in mating and conception pat-
close inbreeding with their daughters or sisters (Nishida
terns might also occur as young males, who are less
et al., 2003; Strier, 2004a). Moreover, male-biased sex
likely to achieve ejaculation than older males, begin to
ratios in their natal groups might lead male muriquis to
account for an increasing proportion of available mates
disperse into groups with more favorable sex ratios
(Possamai et al., 2005).
(Strier, 2004b), similar to the case of male natal-group
transfer reported in bonobos (Hohmann, 2001).
A resulting shift, from female dispersal to bisexual dis-
CONCLUSIONS
persal patterns, could ultimately impact the size and
Our study highlights two additional points that are
number of groups in the population if it continues to
widely applicable and have significant implications for
grow. In other primates with growing populations, sex-
interpretations of behavior patterns in other species.
biased dispersal regimes tend to lead to increases in
First, it emphasizes the importance of integrating behav-
group sizes, whereas bisexual dispersal can also lead to
ior, ecological, and demographic contexts, which can fluc-
increases in the number of groups, particularly if males
tuate over the course of an individual’s lifetime (Alt-
follow females (Strier, 2000). Both times in which the
mann and Altmann, 1979; Dunbar, 1979; Gould et al.,
Jao´ group fissioned, to form the M2 and Nadir groups,
2003; Strier, 2003a,c; Jones, 2005). Recent analyses sug-
subsets of Jao´ males initially maintained transient asso-
gest that demographic conditions involving high mortal-
ciations and mated with females in more than one group,
ity may limit the formation of large same-sex family
until they ultimately transferred permanently into each
groups in sifakas (Pochron et al., 2004), similar to the
of the newly established groups. Documenting the proc-
effects we suggest that male-biased sex ratios may have
esses by which subsets of males shift from philopatry, to
on philopatric male kin in northern muriquis.
transient associations, to membership in newly estab-
Second, behavioral responses to demographic fluctua-
lished groups will be critical to understanding both
tions, including the establishment of new groups and
intergroup dynamics and the ways in which muriquis in
dispersal patterns among them, can directly impact the
a growing population distribute themselves within this
genetic structure of populations (Cheney and Seyfarth,
habitat (Boubli et al., 2005, unpublished results). None-
1983; Jones, 2005). For example, although dispersing
theless, we emphasize a distinction between the persis-
females appear to avoid mating with philopatric males
tence of female dispersal, which is typical in this popula-
from their natal groups during intergroup encounters
tion, and the two times in which subgroups of Jao´ males
(Strier, 1997), bisexual dispersal into newly established
shifted their associations opportunistically when sub-
groups could confound these avoidance mechanisms if
groups of Jao´ females split to establish new groups.
closely related males and females transfer out of their
Although these males may be characterized as having
natal groups and into the same groups. Understanding
transferred out of their natal group, the process by which
how primates adjust their behavior in response to chang-
they did so appears to differ from dispersal by females.
ing demographic conditions is critical to evaluating the
It is also important to note that the low levels of overt
viability and evolutionary potential of isolated popula-
competition documented among males in the Mata˜o
tions.
American Journal of Physical Anthropology—DOI 10.1002/ajpa

---

236
K.B. STRIER ET AL.
ACKNOWLEDGMENTS
Forest of South America: biodiversity status, threats, and out-
look. Washington, DC: Island Press. p 31–42.
We thank Conselho Nacional de Desenvolvimento
Cheney DL, Seyfarth RM. 1983. Non-random dispersal in free-
Cientifico e Tecnolo´gico (CNPq), Instituto Brasileiro do
ranging vervet monkeys: social and genetic consequences. Am
Meio Ambiente e dos Recursos Naturais Renova´veis
Nat 122:392–412.
(IBAMA), and the Abdalla family for permission to con-
Cowlishaw G, Dunbar RIM. 2000. Primate conservation biology.
duct this research, E.M. Veado and J. Gomes for provid-
Chicago: University of Chicago Press.
Dias LG, Strier KB. 2003. Effects of group size on ranging pat-
ing essential logistical support, and Fundac¸a˜o Biodiversi-
terns in Brachyteles arachnoides hypoxanthus. Int J Primatol
tas and Conservation International Brazil for infrastruc-
24:209–221.
ture and administrative help. Long-term research on the
Dunbar RIM. 1979. Population demography, social organiza-
Mata˜o group has been possible thanks to sponsorship by
tion, and mating strategies. In: Bernstein IS, Smith EO, edi-
C. Valle, C. Ades, G.A.B. Fonseca, and S.L. Mendes, with
tors. Primate ecology and human origins: ecological influen-
contributions to demographic monitoring by E. Veado,
ces on social organization. New York: Garland STPM Press.
J. Gomes, F. Mendes, J. Rı´moli, A.O. Rı´moli, F. Neri, P.
p 65–88.
Coutinho, A. Carvalho, L. Oliveira, C. Nogueira, S. Neto,
Gould L, Sussman RW, Sauther ML. 2003. Demographic and
W. Teixeira, R. Printes, M. Maciel, C. Costa, A. Oliva, L.
life-history patterns in a population of ring-tailed lemurs
Dib, D. Carvalho, N. Bejar, C. Coelho, L. Dias, V.O. Gui-
(Lemur catta) at Beza Mahafaly Reserve, Madagascar: a 15-
year perspective. Am J Phys Anthropol 120:182–194.
maraes, W.P. Martins, J.C. da Silva, C. de Borba Possa-
Hohmann G. 2001. Association and social interactions between
mai, Regiane R. de Oliveira, F.P. Paim, M.F. Iurckk, K.
strangers and residents in bonobos (Pan paniscus). Primates
Tolentino, V. Souza, D. Guedes, and J. Fidelis. Research
42:91–99.
on the Jao´ group has been possible thanks to demo-
Jones CB. 2005. Behavioral flexibility in primates: causes and
graphic monitoring by C.P. Nogueira, V.O. Guimaraes, I.
consequences. New York: Springer.
Mourthe, M. Tokuda, and Leo G. Dias. Research on the
Konstant WR, Mittermeier RA, Rylands AB, Butynski TM,
M2 and Nadir groups has been possible thanks to R.L.
Eudey AA, Ganzhorn J, Kormos R. 2002. The world’s top 25
Young and C.P. Pereira. Past funding for the Mata˜o
most endangered primates—2002. Neotrop Primates 10:128–
study was provided by the National Science Foundation
131.
(BNS 8305322, BNS 8619442, and BNS 8959298), the
Lacy RC. 1993a. What is population (and habitat) viability anal-
ysis? Primate Conserv 14/15:27–33.
Fulbright Foundation, the Joseph Henry Fund of the
Lacy RC. 1993b. VORTEX: a computer simulation model for
National Academy of Sciences, Sigma Xi, the L.S.B. Lea-
population viability analysis. Wildlife Res 20:45–65.
key Foundation, the World Wildlife Fund, the Seacon
Martins WP, Strier KB. 2004. Age at first reproduction in philo-
Fund of the Chicago Zoological Society, the Liz Claiborne
patric female muriquis (Brachyteles arachnoides hypoxan-
and Art Ortenberg Foundation, the Scott Neotropic Fund
thus). Primates 45:63–67.
of the Lincoln Park Zoo, the National Geographic Soci-