American Journal of Primatology 71:206–213 (2009)

RESEARCH ARTICLE

Social Organization of White-Headed Langurs (Trachypithecus leucocephalus)in the Nongguan Karst Hills, Guangxi, China

TONG JIN�, DE-ZHI WANG, QING ZHAO, LI-JIE YIN, DA-GONG QIN, WEN-ZHONG RAN, AND WEN-SHI PANGiant Panda and Wildlife Conservation Research Center, School of Life Sciences, Peking University, Beijing,People’s Republic of China

The number of males per group is the most variable aspect of primate social organization and is oftenrelated to the monopolizability of females, which is mainly determined by the number of females pergroup and their reproductive synchrony. Colobines show both inter-specific and intra-specific variationsin the number of males per group. Compared with other colobine species, little is known about the socialorganization of white-headed langur (Trachypithecus leucocephalus), despite its endangered status andunusual limestone habitat. As a part of a long-term study of the white-headed langurs in the NongguanKarst Hills, Guangxi, China, we quantitatively investigated their social organization by analyzingcensus data from 1998 to 2003. The population censuses revealed that the predominant socialorganization of bisexual groups was the one-male group, similar to a previous report on this species andmany other Asian colobines. In such groups, one adult male associated with 5.1 adult females, 0.1 sub-adult males, 2.6 juveniles and 2.9 infants on average, with a mean group size of 11.7 individuals. Inaddition, three multi-male groups were recorded, consisting of 2–3 adult males, 1–5 adult females, 0–2sub-adult males, 0–7 juveniles and 0–2 infants. They did not contain more adult females than the one-male groups and were unstable in group membership. The langurs outside bisexual groups wereorganized into small nonreproductive groups or lived as solitaries. The nonreproductive groupsaveraged 1.3 adult males, 1.3 sub-adult males and 2.6 juveniles. Juvenile females were present in suchgroups on 52.4% of all occasions. As predicted by the monopolization model, the prevalence of the one-male pattern in this species may mainly be attributed to the small number of females in the group. Thepossible reasons for the occurrence of multi-male groups and the presence of juvenile females innonreproductive groups are also discussed. Am. J. Primatol. 71:206–213, 2009. r 2008 Wiley-Liss, Inc.

Key words: Asian colobine; one-male group; group composition; monopolizability of females;group size

INTRODUCTION

The white-headed langur (Trachypithecus leuco-cephalus) is among the most endangered primatespecies in the world and endemic to China. They arerestricted to the limestone hills in a triangular area ofsouthern Guangxi province (221060–221420N,1071–1081E), with a total number of 700–800 indivi-duals mainly found in the Nongguan Karst Hills inChongzuo County and the Nonglin Karst Hills in FusuiCounty [Ran, 2003; Wang & Jin, 2004; Wang et al.,2005]. As this species was first described by Tan [1957],debates about its taxonomic status have continued onwhether it is an independent species [Eudey, 1987; Lu& Li, 1991], or a subspecies of the Francois’ langur(T. francoisi) [Li & Ma, 1980; Shen & Li, 1982; Wanget al., 1997]. Recently, Groves [2001] listed six closelyrelated langur taxa into the T. francoisi group, andrecognized the white-headed langur as a subspecies ofthe golden-headed langur (T. poliocephalus) in Viet-nam. This opinion was agreed by other researchers[Brandon-Jones et al., 2004]

Although its precise taxonomic status remainsunresolved, study of the white-headed langur isimportant because of its endangered status and itsspecial limestone habitat. Their unusual habitatrequirements may lead to differences in behavioralecology from other colobine species. Moreover,knowledge on their socioecology will be useful forconservation purpose. However, only limited infor-mation is available on this species. Previous studiesare mainly conducted in the Nonglin Karst Hills in

Published online 10 November 2008 in Wiley InterScience (www.interscience.wiley.com).

DOI 10.1002/ajp.20637

Received 30 July 2008; revised 10 October 2008; revisionaccepted 11 October 2008

Contract grant sponsors: Peking University; Margot MarshBiodiversity Foundation; Conservation International.

�Correspondence to: Tong Jin, Giant Panda and WildlifeConservation Research Center, School of Life Sciences, PekingUniversity, Beijing 100871, People’s Republic of China.E-mail: jintong80@hotmail.com

rr 2008 Wiley-Liss, Inc.

Fusui County, focusing primarily on activity budget,locomotion, feeding ecology and habitat quality[Huang & Li, 2005; Huang et al., 2002, 2003, 2007;Li, 1993; Li & Rogers, 2004a, 2005a, b, 2006; Li et al.,2003]. Information on the social organization of thisspecies has been limited to one study of the Nonglinpopulation [Li & Rogers, 2004b]. Based on briefobservations, they suggested that the principalmating system of white-headed langurs was one-male, but one multi-male group was found in thesame population. This study provides basic under-standing on the species’ social organization, butlacks any quantitative data and is unable to explainthe intra-specific variation in the number of adultmales per group. Compared with the Nonglinpopulation, little is known about the langurs in theNongguan Karst Hills.

To understand the social system of the white-headed langur, a long-term project was conducted inthe Nongguan Karst Hills in Chongzuo County,beginning in December 1996. As a part of thiscomprehensive study, in this article, we analyze theresults from regular population censuses conductedbetween 1998 and 2003. From a comparativeperspective, we aim to present quantitative data onthe social organization and group composition of the

little-studied white-headed langurs in the NongguanKarst Hills on a long-term scale.

METHODS

Study Site

The white-headed langurs were studied inthe Nongguan Karst Hills (221150–221170N,1071290–1071320E, 150–430 m above sea level), a22 km2 area within the Chongzuo Precious AnimalNature Reserve, Guangxi, China (Fig. 1). The sitereceives an average annual rainfall of 1,152 mm,which occurs in a seasonal pattern: a dry season fromNovember through March with r50 mm monthlyrainfall and a rainy season from May to August withZ150 mm monthly rainfall. The mean temperatureis 22.81C with a maximum of 28.71C in July and aminimum of 15.01C in January [Chongzuo WeatherBureau, 1998–2004, unpublished data].

The study site is typical karst, with steeplimestone hills rising over the flatlands, providing arelatively safe habitat for the langurs. The habitat ishighly fragmented by human activity. Almost all theflatlands and valleys around the hills have beencultivated by the local people to grow sugarcane, rice,corn and peanuts, whereas most of the original

Fig. 1. Study site of the Nongguan Karst Hills in Chongzuo County, Guangxi, China.

Am. J. Primatol.

Social Organization of White-Headed Langurs / 207

tropical monsoon forest on the hills and slopes hasbeen replaced by secondary shrub vegetation owingto firewood collection by villagers. The langursinhabit the limestone hills and rarely come down tothe flatlands, so crop raiding and provisioning arenever seen. No other primate species are sympatricwithin the study area. Only a few potential predatorspecies have survived in the study area owing tohuman activity, such as leopard cat (Felis bengalen-sis), yellow-throated marten (Martes flavigula) andgolden eagle (Aquila chrysaetos). Locally extinct aretiger (Panthera tigris), leopard (P. pardus), andclouded leopard (Neofelis nebulosa). Poaching wasthe main threat to the langur population in the past,but has been prohibited in the study area since 1998[Wang & Jin, 2004].

Age/Sex Classification

During population censuses, we categorized thelangurs into six age/sex classes: adult males, adultfemales, sub-adult males, juvenile males, juvenilefemales and infants.

The langurs show little sexual dimorphism inbody size, so we identified the sexes mainly by thecolor and shape of the perineal region. In males, thisregion is covered by black hair extending back to thefused ischial callosities, whereas in females thistriangular region is hairless and skin-colored, whichclearly contrasts with the surrounding black hairand can be clearly seen from a distance. This methodwas applicable in sex determination in the adults,sub-adults and juveniles during the census work.However, it was hard to apply to the infants, becauseof the slight contrast between the pale skin color andtheir yellow hair. Thus, distinguishing sex amonginfants was much more difficult but could still beachieved after repeated observations, by the pendu-lous penis in male infants.

Age/sex classes were assigned mainly by bodysize and pelage color pattern. Adult males have fullhead–body length and shoulder width. They areeasily discriminated by the discernable penis whenfacing observers. Adult females have a somewhatsmaller head–body length and shoulder width thanadult males, and can be distinguished from nullipar-ous juvenile females by elongated nipples, whichimply recent suckling. Sub-adult males are a littlesmaller and much more slender than the adults,usually with a visible glans and protruding snoutbecause of developing canine teeth. Juveniles aremuch smaller (less than two-thirds that of the adult)and have more white patches on their shoulders,limbs, feet and tail than the adults and sub-adults.Infants are distinguished by the differential pelagecolor and suckling behavior. Langurs are born withflamboyant orange pelage and pale skin. After about3 months, the coat begins to grow darker and theskin turns black. The pelage becomes yellow–brown

at about 1 year and gradually changes into the adultblack-and-white pattern after 1.5–2 year, at whichtime they are weaned and reclassified as juveniles[Weaning age: 19–21 months, Zhao et al., 2008].

Individuals were identified by age/sex class, coatcolor pattern, face/body shape and injuries. Adultfemales were also identified by the black blotches onthe skin-colored area at the perineum. Infants werealso recognized by the strong bond with theirmothers. Note that all the individual identificationswere done with great caution. Several physical andbehavioral traits were always combined to identifythe animals, usually within a given social orgeographic context. And individual identificationswere conducted by experienced observers.

Data Collection and Analysis

The langur population in Nongguan has beenunder observation since December 1996. Preliminaryobservations were conducted in a random way tocollect basic information from December 1996 toJanuary 1997 and March–May 1997 [Zhang, 1998].Since January 1998, this population has beensystematically monitored through regular populationcensuses and focal group observations by variousresearchers from Peking University. From 1998 to2003, each year at least one population census wascarried out to assess the age/sex composition of asmany groups as possible. The censuses were inten-sively conducted in the northern part of theNongguan Karst Hills (about 10 km2, Fig. 1).Because of the terrain and the unhabituated natureof the langurs, observations were normally made at adistance of 50–200 m through binoculars (LEICATRINOVID 8� 42BN) or a spotting scope (LEICAAPO-TELEVID77 20–60� ). This study compliedwith the protocols approved by the Guangxi ForestryAdministration and adhered to the legal require-ments of China.

Most langurs live in groups and select smallcaves or cracks on the cliffs as sleeping sites. Eachgroup has several fixed sleeping sites and leaves thesite in the early morning and moves back at dusk.The sleeping sites are easily recognized by thereddish-brown excrement markings on the rocksurface below them [Huang & Li, 2005; Huanget al., 2003; Wang et al., 2005]. During the popula-tion censuses, we looked for such sleeping sites alongthe trails and searched for the langur groups aroundthe sleeping sites. When a group was detected, dataconcerning group size and age/sex composition wererepeatedly recorded. Each sleeping site was thor-oughly searched at least twice. Groups were recog-nized by identified individuals, relatively stablegroup composition and home range.

In this study, we chose results of the censusesconducted between January and March from 1998 to2003 to calculate the percentage of each group type,

Am. J. Primatol.

208 / Jin et al.

the distribution of adult males among differentgroup types, and age/sex composition of each grouptype. During these months, the vegetation wassparsest and the visibility was highest, thus theresults were most accurate. Note that the samegroup may occur in the results of several censusesbut were at different phases in group composition, sothey were treated as different ones during the courseof calculation.

Kruskal–Wallis tests were used to compare themean group size and the mean number of adultfemales in one-male groups across time. Statisticaltests were two-tailed, with Po0.05 taken as thethreshold for significance (SPSS 15.0).

RESULTS

Social Organization at Group Level

According to censuses from 1998 to 2003, mostof the langurs in Nongguan were group-living,whereas some males lived as solitaries. Group-livinglangurs commonly lived in bisexual groups ornonreproductive groups (Table I). Nonreproductivegroups referred to the groups without adult females.In the censuses of 1998 and 1999, 92.3 and 85.7% ofthe bisexual groups were one-male groups, while oneand two multi-male groups were recorded, respec-tively. Thereafter, bisexual groups were exclusivelyone-male. Bisexual groups were accompanied bynonreproductive groups in all the censuses exceptin 1999. Overall, 72.4% of the group-living adultmales in the population lived in one-male groups,5.7% in multi-male groups and 21.9% in nonrepro-ductive groups.

Besides the group-living langurs, solitary maleswere occasionally encountered during the censuses.Owing to the difficulty in individual recognition ofsolitary males, we only recorded the number ofsightings, and this number to a great extent

depended on the duration of the censuses. Therelatively small number of the sightings of solitarymales in 2003 was attributed to the short censusperiod (Table I).

Social Composition of One-Male Groups

The age/sex composition of one-male groups ineach census from 1998 to 2003 is shown in Table II.There were no differences either in the mean groupsize (Kruskal–Wallis test: w2 5 2.597, df 5 5,P40.05), or the mean number of adult females(Kruskal–Wallis test: w2 5 6.199, df 5 5, P40.05)across time. On average, a one-male group consistedof one adult male, 5.1 adult females, 0.1 sub-adultmale, 2.6 juveniles and 2.9 infants, with a meangroup size of 11.7 individuals.

Social Composition of Multi-Male Groups

Three instances of multi-male groups were alsorecorded in the population over the course of thestudy: Group GYN, Group GY, and Group SHY. Inthe census of 1998, group GYN was documented tohave three adult males, three adult females and twonew-born infants. This group was first identified inJanuary 1997, with four adult males and three adultfemales [Zhang, 1998]. During the course of thecensus in 1998, two new-born infants were recordedin January and then one adult male disappearedfrom the group in March. When we revisited thegroup in the census of 1999, it had become a one-male group with one adult male, three adult femalesand two infants. Because the males were notindividually identified, the process from multi-maleto one-male was unknown.

In the census of 1999, group GY was recordedwith two adult males, one adult female, one sub-adult male, four juvenile males and three juvenilefemales. Group GY formed in April 1998, by several

TABLE I. Social Organization of the White-Headed Langurs in Nongguan Karst Hills

Bisexual group

One-male Multi-maleNonreproductive

group Total

YearWorkday

# ofgroups

# ofindivs.

# ofgroups

# ofindivs.

# ofgroups

# ofindivs.

# ofgroups

# ofindivs. Size AM/AFa

Sightingsof solitary

males

1998 65 12 132 1 8 4 17 17 157 9.2 1:3.0 101999 59 12 137 2 21 0 0 14 158 11.3 1:4.3 142000 45 14 181 0 0 3 15 17 196 11.5 1:3.5 152001 59 17 204 0 0 3 19 20 223 11.2 1:4.7 172002 61 17 204 0 0 5 25 22 229 10.4 1:3.1 172003 29 17 180 0 0 6 23 23 213 9.3 1:3.7 3

N, number of individuals; AM, adult male; AF, adult female.asolitary males were excluded from calculation because the actual number was unknown, making the sex ratio more female-biased than the real populationadult sex ratio.

Am. J. Primatol.

Social Organization of White-Headed Langurs / 209

individuals who were evicted from a one-male groupafter male take-over and at least six individuals ofunknown history [Zhang, 2002]. It was under focalobservation from August 1998 to July 2000. Thegroup was relatively stable in group compositionuntil the only adult female died in November 1999,and the group turned into a nonreproductive groupwith three juvenile females [Zhang, 2002].

Another multi-male group, group SHY, was alsorecorded in the census of 1999. At that time, thegroup consisted of two adult males, five adultfemales, two sub-adult males and one new-borninfant. But the infant had disappeared when werevisited this group in April. In August, group SHYsplit into a one-male group, which consisted of oneadult male and five adult females, and a nonrepro-ductive group comprising three adult males [Panet al., unpublished data].

Overall, a multi-male group comprised 2.370.6adult males, 3.072.0 adult females, 1.071.0 sub-adult male, 2.374.0 juveniles and 1.071.0 infant,with a size of 9.771.5 individuals. The multi-malegroups did not have more adult females than the one-male groups.

Social Composition of NonreproductiveGroups

The nonreproductive groups averaged 1.3 adultmales, 1.3 sub-adult males and 2.6 juveniles, with amean size of 5.2 individuals (Table II). The size ofnonreproductive groups was smaller than that ofbisexual groups. In general, a nonreproductive groupconsisted of adult, sub-adult, and juvenile males.Sometimes a nonreproductive group only had duos ortrios of adult and/or sub-adult males. However, up to

three juvenile females were present in such groupson 11 occasions (52.4% of all occasions). Among theindividuals found in nonreproductive groups, 82.6%were males of all age classes except infants and17.4% were juvenile females.

As previously noted, group GY became such agroup with three juvenile females in November 1999.In July 2000, when the three juvenile females beganestrous, extra-group males continuously harassedthe group. Finally, an invading male took over thegroup and evicted all the male members of the group[Zhang, 2002].

DISCUSSION

In Asian colobines, the modal pattern of bisexualgroups is the one-male group [Kirkpatrick, 2007;Newton & Dunbar, 1994]. A previous report on thewhite-headed langurs in the Nonglin Karst Hills inFusui County suggested that the principal matingsystem of this species is one-male [Li & Rogers,2004b]. In the Nongguan Karst Hills, repeatedpopulation censuses revealed that the predominantsocial organization of bisexual groups was also theone-male group, and this pattern was stable on along-term scale. A one-male group typically con-tained 5.1 adult females, which is similar to thelangurs in Fusui County [3–9 females, Li & Rogers,2004b].

Former analyses identified monopolizability offemales as the key factor responsible for the distinc-tion between one-male and multi-male groups[Clutton-Brock & Harvey, 1977; Emlen & Oring,1977; van Schaik & van Hooff, 1983; Wrangham,1980]. The number of females in a group and theirdegree of reproductive synchrony have been

TABLE II. Age/sex Composition of One-Male Groups and Nonreproductive Groups in Nongguan Karst Hills

Mean7SD (range)

One-male group Nonreproductive group

YearAdultfemale

Sub-adultmale Juvenile Infant Group size

Adultmale

Sub-adultmale

Juvenilemale

Juvenilefemale Group size

1998 5.572.3(1–9)

0 2.072.3(0–6)

2.572.2(0–4)

11.075.1(3–18)

2.070.0(2–2)

0 1.370.5(1–2)

1.070.8(0–2)

4.371.0(3–5)

1999 5.272.4(1–9)

0 2.872.4(0–6)

2.572.2(0–6)

11.475.0(4–19)

0 0 0 0 0

2000 5.272.1(1–9)

0 2.773.1(0–9)

4.072.4(0–8)

12.976.2(5–24)

2.370.6(2–3)

0.370.6(0–1)

1.372.3(0–4)

1.071.7(0–3)

5.074.4(2–10)

2001 4.872.5(2–11)

0.170.5(0–2)

3.173.2(0–10)

3.172.4(0–8)

12.077.3(4–29)

1.070.0(1–1)

2.770.6(2–3)

1.371.2(0–2)

1.371.5(0–3)

6.372.5(4–9)

2002 5.172.4(2–12)

0.270.8(0–3)

2.972.6(0–8)

2.871.9(0–6)

12.075.3(5–26)

0.870.4(0–1)

1.871.1(0–2)

1.471.3(0–3)

1.071.2(0–3)

5.072.4(3–9)

2003 4.873.5(1–14)

0.270.8(0–3)

1.972.4(0–7)

2.672.6(0–9)

10.676.9(4–30)

0.870.4(0–1)

1.771.0(0–3)

2.573.4(0–9)

0.570.8(0–2)

5.573.9(2–13)

Overall 5.172.6(1–14)

0.170.5(0–3)

2.672.7(0–10)

2.972.3(0–9)

11.776.0(3–30)

1.370.7(0–3)

1.371.2(0–3)

1.772.1(0–9)

0.971.1(0–3)

5.272.8(2–13)

Am. J. Primatol.

210 / Jin et al.

identified as important determinants of malemonopolization potential [Andelman, 1986; Dunbar,1988; Mitani et al., 1996; Nunn, 1999; Ridley, 1986].For Asian colobines, there are both inter-specific andintra-specific variations in the number of males pergroup. In the hanuman langurs (Semnopithecusentellus), data for 23 populations show that thepercentage of one-male groups is, indeed, related tothe number of females per group and breedingseasonality [Srivastava & Dunbar, 1996]. Theoreti-cally, a single hanuman langur male is able tomonopolize up to 12 adult females [Newton, 1988].However, in other langur subgenera, although multi-male groups are regularly observed, the monopoliz-ability of females is always high because the numberof females per groups is on average smaller and birthseasonality is less pronounced than the hanumanlangurs [Sterck & van Hooff, 2000]. In the case of thewhite-headed langurs, although the number offemales per group varied considerably (range: 1–14),the typical one-male group only consisted of 5.1 adultfemales, which was small enough to be defensible.Thus, the prevalence of the one-male groups in thisspecies is probably owing to the relatively smallnumber of females in the group, as predicted by themonopolization model. But in one case, an adult malewas able to monopolize as many as 14 adult females,so additional information about the degree ofreproductive synchrony in this species is still neededto understand the maximum limit of the number offemales per group that can be monopolized by onemale.

Although the modal pattern of bisexual groupsin this species is the one-male group, three multi-male groups were recorded in our study. This intra-specific variation in the number of males per groupwas also documented in the langurs in Fusui County[Li & Rogers, 2004b]. But owing to the shortobservation period, the only multi-male group inthe Nonglin Karst Hills was not followed longenough to determine whether it was a reproductivegroup or just temporary phases in a demographicdevelopment. In the Nongguan population, themulti-male groups did not contain more femalesthan the one-male groups, suggesting that they arenot a strategy to maintain access to more females. Inmany other colobine species, the multi-male groupsmay either be age-graded or temporarily occurduring male replacement [reviewed by Sterck &van Hooff, 2000]. In our study, groups GY and SHYwere associated with male take-over and all threemulti-male groups finally became one-male groups ornonreproductive groups, indicating their transitoryand unstable nature. The only multi-male groupreported in the Nonglin Karst Hills also formed aftera male take-over [Li & Rogers, 2004b]. Age-gradedmulti-male groups may also have the chance tohappen in this population, because sub-adult malesoccurred in the one-male groups on five occasions,

implying the relatively long male tenure in thisspecies and the possibility for young males maturingin their natal groups. Therefore, as in most otherlangur species, the multi-male group in this species isa temporary adaptive strategy in response to com-plex demographic processes.

Other than the bisexual groups, solitary malesand nonreproductive groups mainly consisting ofmales were also described in white-headed langurs,indicating that the males are the ‘‘dispersing sex’’, asin many other Asian colobines [Kirkpatrick, 2007;Newton & Dunbar, 1994]. However, unlike the all-male bands described in the Nonglin population,juvenile females were present in nonreproductivegroups on 11 occasions in Nongguan, indicating thatimmature female dispersal occurs in this species,although they only accounted for a small portion ofall the individuals in such groups. Similar situationshave been described in other colobine species, suchas Nasalis larvatus, T. auratus, T. johnii andPresbytis thomasi [Steenbeek et al., 2000; Yeager &Kool, 2000]. This kind of group may be formed by theimmatures of both sexes expelled from one-malegroups by the new resident male after take-over, asin the cases of the purple-faced langur (T. vetulus)[Rudran, 1973] and the Thomas’s langur (P. thoma-si) [Steenbeek et al., 2000], or by all-male bandsjoined by large juvenile females, who were presumedto have left their natal groups to avoid inbreeding, asin the case of the proboscis monkey (N. larvatus)[Murai, 2004]. In this species, the nonreproductivegroups were elsewhere referred to as male-immaturegroups and formed by the ousted male and all hismale offspring after male take-over [Zhao & Pan,2006]. In some cases, older female infants wereforced to wean and also left with the ousted male toavoid infanticide [Wang, 2004]. As in group GY,when these females approached sexual maturity, thegroup became unstable because of continuous maleharassment and finally turned into a one-male groupby male take-over.

In conclusion, we suggest that the prevalence ofone-male groups in this species may be mainlyattributed to the small number of females in thegroup. Low predation risk, within-group feedingcompetition and infanticide might also contributeto small group size in this langur population, butmore data are needed to test these hypotheses.

ACKNOWLEDGMENTS

For contribution to the data collection of thislong-term study we are very grateful to Lei Zhang,Yingyi Zhang and Lizhong Zhu. Our thanks go to thelocal guide, Mr. Qihai Chen and the staff inChongzuo Eco-park for assistance in the field. Thelocal government of Guangxi Province and ChongzuoCounty supported the field work. Thanks also go tothe Weather Bureau of Chongzuo County for sharing

Am. J. Primatol.

Social Organization of White-Headed Langurs / 211

the climate data. We thank Dr. Carola Borries, Prof.Kunio Watanabe and Prof. Terry Harrison forhelpful comments and discussion. Special thanksare extending to the editors and two anonymousreviewers for their valuable suggestions on editingthe manuscript. We gratefully acknowledge ourcolleagues in the Giant Panda and Wildlife Con-servation Research Center for their thoughtfuldiscussions on the subject. Thanks go to Dr. Iain C.Bruce for revising the language. The study wasfinancially supported by the 985 Project of PekingUniversity, the Margot Marsh Biodiversity Founda-tion, and Conservation International. This studycomplied with the protocols approved by the GuangxiForestry Administration and adhered to the legalrequirements of China.

REFERENCES

Andelman SJ. 1986. Ecologcial and social determinants ofcercopithecine mating patterns. In: Rubenstein DI,Wrangham RW, editors. Ecological aspects of social evolu-tion. Princeton: Princeton University Press. p 210–216.

Brandon-Jones D, Eudey AA, Geissmann T, Groves CP,Melnick DJ, Morales JC, Shekelle M, Stewart CB. 2004.Asian primate classification. Int J Primatol 25:97–164.

Clutton-Brock TH, Harvey P. 1977. Primate ecology and socialorganization. J Zool 183:1–39.

Dunbar RIM. 1988. Primate social systems. In: Lazarus J,editor. London & Sydney: Croom Helm.

Emlen ST, Oring LW. 1977. Ecology, sexual selection, and theevolution of mating systems. Science 197:215–223.

Eudey AA. 1987. Action plan for Asia primate conservation.IUCN—the world conservation union.

Groves CP. 2001. Primate taxonomy. Washington, DC:Smithsonian Institution Press.

Huang CM, Li YB. 2005. How does the white-headed langur(Trachypithecus leucocephalus) adapt locomotor behavior toits unique limestone hill habitat? Primates 46:261–267.

Huang CM, Wei FW, Li M, Quan GQ, Li HH. 2002. Currentstatus and conservation of white-headed langur (Trachy-pithecus leucocephalus) in China. Biol Conserv 104:221–225.

Huang CM, Wei FW, Li M, Li YB, Sun RY. 2003. Sleeping caveselection, activity pattern and time budget of white-headedlangurs. Int J Primatol 24:813–824.

Huang CM, Wu H, Zhou QH, Li YB, Cai XW. 2007. Feedingstrategy of Francois’ langur and white-headed langur atFusui, China. Am J Primatol 70:320–326.

Kirkpatrick CR. 2007. The Asian colobines: diversity amongleaf-eating monkeys. In: Campbell CJ, Fuentes A,MacKinnon KC, Panger M, Beader S, editors. Primates inperspective. Oxford: Oxford University Press. p 186–200.

Li ZY. 1993. Preliminary investigation of the habitats ofPresbytis francoisi and Presbytis leucocephalus, with noteson the activity pattern of Presbytis leucocephalus. FoliaPrimatol 60:83–93.

Li ZX, Ma SL. 1980. The taxonomy of white-headed langur.Acta Zootaxon Sin 5:440–442.

Li ZY, Rogers ME. 2004a. Habitat quality and activity budgetsof white-headed langurs in Fusui, China. Int J Primatol25:41–54.

Li ZY, Rogers ME. 2004b. Social organization of white-headedlangurs Trachypithecus leucocephalus in Fusui, China. FoliaPrimatol 75:97–100.

Li ZY, Rogers ME. 2005a. Are limestone hills a refuge oressential habitat for white-headed langurs in Fusui, China?Int J Primatol 26:437–452.

Li ZY, Rogers ME. 2005b. Habitat quality and range use ofwhite-headed langurs in Fusui, China. Folia Primatol76:185–195.

Li ZY, Rogers ME. 2006. Food items consumed by white-headed langurs in Fusui, China. Int J Primatol27:1551–1567.

Li ZY, Wei Y, Rogers ME. 2003. Food choice of white-headedlangurs in Fusui, China. Int J Primatol 24:1189–1205.

Lu LR, Li ZY. 1991. On the taxonomy of the Presbytis francoisileucocephalus. J Guangxi Normal Univ 11:12–16.

Mitani JC, Cros-Louis J, Manson JH. 1996. Number of malesin primate groups: comparative tests of competing hypoth-eses. Am J Primatol 38:315–332.

Murai T. 2004. Social behaviors of all-male proboscis monkeyswhen joined by females. Ecol Res 19:451–454.

Newton PN. 1988. The variable social organizationof Hanuman langurs (Presbytis entellus), infanticide,and the monopolization of females. Int J Primatol 9:59–77.

Newton PN, Dunbar RIM. 1994. Colobine monkey society. In:Davies AG, Oates JF, editors. Colobine monkeys: theirecology, behavior and evolution. Cambridge: CambridgeUniversity Press. p 311–346.

Nunn CL. 1999. The number of males in primate social groups:a comparative test of the socioecological model. Behav EcolSociobiol 46:1–13.

Ran WZ. 2003. The current status of wild-rangedwhite-headed langur: distribution, number, habitat use,and PVA. Ph.D. dissertation. Beijing, China: PekingUniversity. 94p.

Ridley M. 1986. The number of males in primate troop. AnimBehav 34:1848–1858.

Rudran R. 1973. Adult male replacement in one-male troops ofpurple-faced langurs (Presbytis senex senex) and its effect onpopulation structure. Folia Primatol 19:166–192.

Shen LT, Li HH. 1982. White-headed leaf monkey in Guangxi.J Guangxi Normal Univ 3:71–80.

Srivastava A, Dunbar RIM. 1996. The mating system ofHanuman langurs: a problem in optimal foraging. BehavEcol Sociobiol 39:219–226.

Steenbeek R, Sterck EHM, de Vries H, van Hooff JARAM.2000. Costs and benefits of the one-male, age-graded and all-male phases in wild Thomas’s langur groups. In: KappelerPM, editor. Primate males: causes and consequences ofvariation in group composition. Cambridge: CambridgeUniversity Press. p 130–145.

Sterck EHM, van Hooff JARAM. 2000. The number ofmales in langur groups: monopolizability of females ordemographic processes? In: Kappeler PM, editor. Primatemales: causes and consequences of variation in groupcomposition. Cambridge: Cambridge University Press.p 120–129.

Tan PC. 1957. Rare catches by Chinese animal collectors. ZooLife 12:61–63.

van Schaik CP, van Hooff JARAM. 1983. On the ultimatecauses of primate social systems. Behaviour 85:91–117.

Wang DZ. 2004. Social structure unit of free-ranging white-headed langur (Trachypithecus leucocephalus): the researchon family groups in Nongguan Karst Hills, Guangxi, China.Ph.D. dissertation. Beijing, China: Peking University.102p.

Wang DZ, Jin T. 2004. Current status of the white-headedlangur (Trachypithecus poliocephalus leucocephalus):distribution, population structure and conservation. In:Nadler T, Streicher U, Long HT, editors. Conservation ofprimates in Vietnam. Hanoi: Frankfurt Zoological Society.p 115–116.

Wang W, Forstner MRJ, Zhang YP, Liu ZM, Wei Y, Huang HQ,

Hu HG, Xie YX, Wu DH, Melnick DJ. 1997. A phylogeny ofChinese leaf monkeys using mitochondrial ND3-ND4 gene

sequences. Int J Primatol 18:305–320.

Am. J. Primatol.

212 / Jin et al.

Wang DZ, Ran WZ, Jin T, Pan WS. 2005. Population censusof the white-headed langur (Trachypithecus leucocephalus)at Longrui Karst Hills, Guangxi, China. Primates46:219–222.

Wrangham RW. 1980. An ecological model of female-bondedprimate groups. Behaviour 75:262–300.

Yeager CP, Kool K. 2000. The behavioral ecology of Asiancolobines. In: Whitehead PF, Jolly CJ, editors. Old worldmonkeys. Cambridge: Cambridge University Press. p 496–514.

Zhang L. 1998. Preliminary research on population and socialorganization of the white-headed langur (Presbytis leucoce-phalus) in BanLi xiang, ChongZuo county, Guangxi

province. M.Sc. dissertation. Beijing, China: Peking Uni-versity. 29p.

Zhang YY. 2002. Social organization of the free-ranging white-headed langurs (Trachypithecus francoisi leucocephalus).Ph.D. dissertation. Beijing, China: Peking University.65p.

Zhao Q, Pan WS. 2006. Male-immature interactions seem todepend on group composition in white-headed langur(Trachypithecus leucocephalus). Acta Ethologica 9:91–94.

Zhao Q, Tan CL, Pan WS. 2008. Weaning age, infant care, andbehavioral development in Trachypithecus leucocephalus.Int J Primatol; DOI 10.1007/s10764-008-9255-8.

Am. J. Primatol.

Social Organization of White-Headed Langurs / 213