Keywords

Introduction

More than 50 years have passed since studies on wild chimpanzees began in the 1960s. These studies revealed many aspects of chimpanzee sociality, such as their high degree of fission-fusion dynamics and female-biased dispersal patterns (Nishida 1968; Nishida and Kawanaka 1972; Pusey 1979; Goodall 1986). Many studies have focused on social relationships among males, reporting male dominance and strong male-male bonds (Goodall 1986; Nishida and Hiraiwa-Hasegawa 1987; Nishida and Hosaka 1996; Boesch and Boesch-Achermann 2000; Watts 2000; Arnold and Whiten 2003, Fig. 5.1). Male chimpanzees associate more with males than females do with females (Nishida and Hiraiwa-Hasegawa 1987; Newton-Fisher 1999). Grooming between males is more common than between males and females or between females, and reciprocity in male grooming and interchange of grooming for agonistic support have been reported (Goodall 1986; Takahata 1990b; Watts 2002; Arnold and Whiten 2003). In addition, pairs or trios of top-ranking males sometimes engage in cooperative mate guarding (Watts 1998). Duffy et al. (2007) reported that the alpha male selectively tolerates his allies mating in exchange for support in conflicts. Male chimpanzees maintain long-lasting and equitable social bonds whose formation is affected by maternal kinship and the quality of social relationships (Mitani 2009).

Fig. 5.1
figure 1

Affinitive relationships among male chimpanzees. Male chimpanzees often share meat among themselves

Full size image

In contrast, few studies have focused on social relationships among female chimpanzees, possibly because social relationships among male chimpanzees are more prominent and because it is difficult to collect sufficient data on females since they tend to range alone with their offspring. In the 2000s, however, several researchers started to study chimpanzee female-female relationships at various long-term study sites.

In this chapter, we examine the characteristics of chimpanzee sociality compared with that of bonobos. After reviewing the life history traits of female chimpanzees, we examine the patterns of female chimpanzee ranging and association by reviewing previous studies and analyzing our own data on the ranging patterns of chimpanzees in the Kalinzu Forest, Uganda, and bonobos at Wamba, Democratic Republic of the Congo. Finally, we discuss the flexibility in female chimpanzee social association and life history, in relation to the male-philopatric female-dispersal social structure.

Study Groups

We observed the M group of chimpanzees living in the Kalinzu Forest Reserve, which is located in western Uganda (0°17′S 30°07′ E, 0°17′ E) (Howard 1991; Hashimoto et al. 2001; Hashimoto and Furuichi 2006). In this chapter, we analyzed data from a period between September 2004 and August 2005. In this period, M group included 19 adult males, 4 adolescent males, 18 adult females, 5 adolescent females, and 17 immature individuals.

We observed the E1 group of wild bonobos living at Wamba, DR Congo (0°11′8″ N, 22°37′58″ E) (Furuichi 1987; Kano 1992; Furuichi et al. 2012) from November 2003 to October 2004. During this period, E1 group included 10 adult males, 5 adult females, 2 adolescent females, and 8 immature individuals.

During the chimpanzee study period, we followed an encountered party and recorded members of the party using the 1-h party method (Hashimoto et al. 2001). We recorded all individuals in sight at the beginning of each 1-h segment and continued recording individuals that appeared in sight until the end of that hour. While tracking a party, we recorded their position at 1-min intervals using a global positioning system (GPS) receiver. We obtained the position of the party in each hour by averaging the data of each minute within the hour.

For the bonobo study, we followed a party of bonobos and recorded its members using the same 1-h party method. We also recorded the position of the party at 30-min intervals using a GPS receiver.

Life History of Female Chimpanzees

Chimpanzees and bonobos live in multi-male/multi-female communities (Nishida 1968; Goodall 1986; Kano 1992). Their dispersal patterns are different from the typical mammalian dispersal pattern. Females in most mammal species tend to remain in their natal group for their whole life, whereas males leave their natal group at sexual maturity (Pusey 1987; Pusey and Packer 1987). However, male philopatry and female dispersal are the predominant patterns for chimpanzees and bonobos (Nishida and Kawanaka 1972; Goodall 1986; Kano 1992).

Female chimpanzees show their first maximal swelling in late adolescence (9–12 years old: Goodall 1986; Pusey 1990; Stumpf et al. 2009), by which time they begin to increase their distance from their mother (Pusey 1990; Stumpf et al. 2009). After about 1 year from the first maximal swelling, females emigrate from their natal group (Stumpf et al. 2009).

Although females generally leave their natal group during adolescence, the likelihood of female transfer varies between sites. In M group of Mahale Mountains National Park, Tanzania (Nishida et al. 2003), the north community of Taï National Park, Côte d’Ivoire (Boesch and Boesch-Achermann 2000), and Kanyawara community in the Kibale National Park, Uganda (Stumpf et al. 2009), the majority of females transfer to another group. However, in Kasakela community of Gombe National Park, Tanzania, only about half of all females transfer (Pusey et al. 1997). At Bossou, all individuals except one male disappeared before reaching 15 years old (Sugiyama 2004). Sugiyama (2004) suggested that at least some males and females emigrated from Bossou, although no emigrant from Bossou, male or female, has been confirmed alive in the adjacent chimpanzee habitat because observation there is difficult.

Little is known about female chimpanzees after they leave their natal group until their immigration into a new group. There is only a small difference between age at emigration from the natal group and age at immigration into a new group. Females emigrate from the natal group at 11 to 12 years (Kanyawara, 12.6 years old (Stumpf et al. 2009); Mahale, 11 years old (Nishida et al. 1990); Taï, 11.42 years old (Boesch and Boesch-Achermann 2000)) and immigrate into a new group at 10–13 years (Fig. 5.2; Gombe, 10–11 years old (Goodall 1986); Kanyawara, 13 years old (Stumpf et al. 2009); Taï, 11 years old (Boesch and Boesch-Achermann 2000)). When immigrant females appear in the new group, they are often in estrus (Nishida 1979; Pusey 1979; Goodall 1986).

Fig. 5.2
figure 2

Females with infant often associate together

Full size image

Although only nulliparous females generally transfer, parous females may also transfer. Females at Bossou leave their natal group after giving birth for the first time (Sugiyama 2004). Nishida et al. (1985) reported that parous females of K group immigrated into M group at Mahale when K group became extinct. Emery Thompson et al. (2006) reported the immigration of more than five parous females with offspring into the study community at Budongo (in the study of chimpanzees, the terms “community” and “group” or “unit group” have the same meaning, as Van Elsacker et al. (1995) have explained in relation to bonobos).

There is a period of adolescent sterility for 2–3 years between the first maximal swelling and the first conception (Goodall 1986; Nishida et al. 1990; Wallis 1997; Boesch and Boesch-Achermann 2000). Females give birth for the first time when they are around 13–15 years old (Nishida et al. 1990; Wallis 1997; Boesch and Boesch-Achermann 2000). Females at Bossou start to give birth much earlier than females at other sites; the mean primiparous age at Bossou is 10.6 years (Sugiyama and Fujita 2011).

There is no evidence that menopause is a characteristic of chimpanzee life histories. Emery Thompson et al. (2007b) analyzed demographic data from six long-term chimpanzee study sites and compared fertility and mortality patterns with those of humans. They found that healthy female chimpanzees maintain high birth rates late into life.

Social Relationships Among Females

Female chimpanzees have fewer affiliative and agonistic social interactions than male chimpanzees. Grooming between females is less frequent than that of male-male dyads (Gombe: Goodall 1986; Mahale: Takahata 1990a, b; Ngogo: Watts 2000; Budongo: Arnold and Whiten 2003; Taï: Boesch and Boesch-Achermann 2000). Agonistic interactions among adult females occur infrequently (Gombe: Goodall 1986; Mahale: Nishida 1989; Kanyawara: Muller 2002; Taï: Wittig and Boesch 2003).

Linear hierarchies have not been documented at many study sites (Gombe: Goodall 1986; Williams et al. 2002; Mahale: Nishida 1989; Ngogo: Wakefield 2008) except at Taï (Wittig and Boesch 2003), although female dominance relationships are ordered in broad rank categories and alpha females can be identified (Gombe: Pusey et al. 1997; Kanyawara: Wrangham et al. 1992).

Characteristics of Females’ Use of Space

Three models have been proposed to describe chimpanzee social organization (Fig. 5.3; Wrangham 1979; Boesch 1991). The male-only community model proposes that only male chimpanzees belong to the community and females are distributed evenly across the habitat in overlapping core areas independent of the males’ range. The male-bonded community model assumes that females settle within the male-defended range but use individually distinct home ranges to minimize feeding competition (Wrangham 1979; Smuts and Smuts 1993). A third model proposes that both male and female chimpanzees occupy the same home range (Itani and Suzuki 1967). Wrangham (1979) referred to this third model as the “classic model,” and Boesch (1991, 1996) termed it the “bisexually bonded community model.”

Fig. 5.3
figure 3

The three models of fission-fusion society in chimpanzees (blue lines for males and red lines for females) originally proposed by Wrangham (1979) and modified by Boesch (1991), Williams et al. (2002), and Lehmann and Boesch (2005). (a) Male-only community model, (b) male-bonded community model, and (c) bisexually bonded community model

Full size image

Studies of chimpanzees in Gombe have supported the male-bonded community model. Wrangham (1979) found that Gombe females occupied core areas dispersed within the community range, smaller than those of males in Gombe. He suggested that Gombe females ranged according to the male-only community model because three anestrous mothers occasionally associated peacefully with males from two communities. Pusey (1980) reported observations that males show unusually violent aggression to females in border areas as evidence for the male-bonded community model rather than the male-only community model. A recent study using a 10-year dataset showed that females spend most of their time in small overlapping core areas within the community range, supporting the male-bonded community model (Williams et al. 2002). Most core areas at Gombe cluster into two neighborhoods, north and south. Dominant females have higher site fidelity and smaller core areas than subordinates, and new immigrants use areas away from dominant females (Murray et al. 2007). High-ranking females use small core areas even during periods of food scarcity, while middle- and low-ranking females use much larger core areas. Because high-ranking females have more reproductive success than the other females (Pusey et al. 1997; Jones et al. 2010), there may be competition among females for areas of good quality (Murray et al. 2007).

Several other studies of chimpanzees at other sites in East Africa support the male-bonded community model. Chapman and Wrangham (1993) revealed that chimpanzee females of Kanyawara community have smaller core areas within the male home range and are dispersed relative to males within the defended area. Emery Thompson et al. (2007a) analyzed a 9-year dataset and found that the ranging patterns of Kanyawara chimpanzees agreed with the male-bonded community model. Like Gombe chimpanzees, Kanyawara females use the community range in a differentiated manner, forming distinct northern, central, and southern clusters. Females in neighborhoods containing more preferred foods have high reproductive success. There ought to be competition over where female chimpanzees settle within the home range, although dominance interactions are less frequent among females than among males (Emery Thompson et al. 2007a). Kahlenberg et al. (2008) examined aggression and dominance relationships among Kanyawara females using 10 years of data. The results supported Emery Thompson et al. (2007a) by showing that females occupying core areas high in foraging quality ranked highly overall and higher than expected for their ages, whereas females occupying low-quality core areas were lower ranking and ranked lower than expected for their ages.

Results from a study by Hasegawa (1990) in Mahale also supported the male-bonded community model, indicating that males used the community home range more evenly than females, who used some areas more preferentially.

In contrast, a recent study of chimpanzees of the Sonso Community in the Budongo Forest, Uganda, showed that female ranges fell within the area of male ranges and that there was no significant difference between male and female core areas, supporting the bisexually bonded community model (Fawcett 2000). Studies of West African chimpanzees also support the bisexually bonded community model. Boesch (1991, 1996) showed that female chimpanzees in Taï range extensively over the whole home range like males, irrespective of their estrous cycle. Recent analysis of a 5-year dataset on the ranging patterns of Taï chimpanzees also suggests that both male and female individual home ranges and core areas greatly overlap (Lehmann and Boesch 2005). A similar pattern was also reported for chimpanzees in Bossou, Guinea (Sugiyama 1988; Sakura 1994; Ohashi, personal communication).

We examined the ranging pattern of M group chimpanzees in the Kalinzu Forest, Uganda. Like Budongo and West African chimpanzees, the ranging pattern in Kalinzu supported the bisexually bonded community model (Fig. 5.4). Both male and female chimpanzees used almost the entire area of the group home range, although females used slightly smaller areas than males (95 % range for males, 8.8 ± 1.0 (SD) km2, n = 19; for females, 6.3 ± 0.8, n = 12; Mann-Whitney U = 8.0, p < 0.001). Unlike Gombe and Kanyawara chimpanzees, Kalinzu females did not form clusters and most females used most of the community home range.

Fig. 5.4
figure 4

Ranging patterns of M group chimpanzees in the Kalinzu Forest. Each line represents the ranging pattern of an individual (blue lines for males and red lines for females). (a) Ranging patterns of males, and (b) ranging patterns of females, (c) ranging patterns of males and females. We estimated individual areas (95 and 50 % of usage) using the fixed kernel density estimation method with Hawth’s Tools and ArcGIS 9. We set the single parameter smoothing factor (h) for the kernel density estimation at 650

Full size image

In summary, early studies suggested that chimpanzees in East Africa followed a male-bonded community model and that chimpanzees in West Africa followed a bisexually bonded community model. However, current evidence suggests that female ranging patterns vary among East African chimpanzees. However, one feature is common to all chimpanzee study sites: male ranges are larger than female ranges, which are incorporated into the male ranges.

There has been only one study of ranging patterns in bonobos, which showed that males and females have very similar ranging patterns (Waller 2011). We examined the ranging patterns of bonobos of E1 group at Wamba, DR Congo. Because there is a human village in the center of the E1 group range, each individual had two or three core areas (Fig. 5.5). The ranging patterns of most individuals except three (Nord, Jacky, and Yuki) were similar (Fig. 5.5). These three individuals probably joined E1 group from extinct groups (Hashimoto et al. 2008). Although the ranging patterns of these three immigrants were slightly different, there was no difference in the size of ranging areas between males and females (95 % range for males, 55.4 ± 1.8 (SD) km2, n = 19; for females, 55.2 ± 2.5, n = 6; Mann-Whitney U = 26, n.s.; Fig. 5.5). That bonobos at Wamba have a much larger ranging area than chimpanzees at Kalinzu might be partly explained by the fact that neighboring groups disappeared during the warfare at Wamba (chapter 6). Our findings agree with Waller’s (2011) results, suggesting that the bisexually bonded community model explains bonobo behavior at these two study sites.

Fig. 5.5
figure 5

Ranging patterns of bonobos of E1 group at Wamba. Each line represents the ranging pattern of an individual (blue lines for males and red lines for females). (a) Ranging patterns of males, (b) ranging patterns of females, (c) ranging patterns of males and females, and (d) ranging patterns of individuals that transferred into E1 group from an extinct group. We estimated individual areas (95 and 50 % of usage) using the fixed kernel density estimation method with Hawth’s Tools and ArcGIS 9.x. We set the single parameter smoothing factor (h) for the kernel density estimation at 650

Full size image

Association Among Females

Based on association patterns (in most studies, “association” is defined as two individuals observed in the same party), previous studies have revealed that males are more gregarious than females, although female gregariousness varies (Wrangham and Smuts 1980; Nishida et al. 1990; Pepper et al. 1999; Lehmann and Boesch 2005).

Goodall (1986) reported that males were more gregarious than females at Gombe, and Wrangham and Smuts (1980) found that females spent more than twice as much time alone as males. Females in the same neighborhood (in which a subset of females have overlapping core ranging areas in the same general part of their community’s range) had levels of association stronger than the average female-female association. Low-ranking females associated more with other low-ranking females than with dominant females (Murray et al. 2006).

Pepper et al. (1999) found that anestrous chimpanzee females showed low gregariousness in Ngogo community in Kibale National Park, Uganda, whereas adult males and estrous females associated with each other significantly more than expected. Unlike at Gombe, anestrous females at Ngogo preferred each other as party members. Although Ngogo females were less gregarious than males, they spent a mean of 64 % of their time in association with ≥1 other female (Wakefield 2008). Females spent as much time ranging in all-female parties as they did alone and exhibited association preferences that extended beyond the dyadic level. They formed distinct association clusters termed “cliques,” within which affiliative interactions occurred more than expected for female-female dyads (Wakefield 2008, 2013).

In the Sonso community at Budongo, female-female associations are weaker than male-male associations (Fawcett 2000). However, some female dyads showed levels of association stronger than the median level for male-male dyads. In addition, associations between females were stronger during rich food periods than during poor food periods. Associations of female chimpanzees were also weaker than those of males in the Kanyawara community at Kibale (Wrangham et al. 1992). Gilby and Wrangham (2008) analyzed 10 years of Kanyawara chimpanzee data and found the same results. A comparison of periods with different alpha males showed that the strength of association changed more frequently among males than among females; female association patterns appeared to be more a consequence of individual ranging behavior rather than a correlate of cooperation (Gilby and Wrangham 2008). Emery Thompson and Wrangham (2006) compared the gregariousness of chimpanzees in the Sonso and Kanyawara communities. In both communities, males were more gregarious than females, and this tendency was stronger when the party size was larger. The sex difference was more pronounced in the Kanyawara community, probably because peripheral females were more gregarious in the Sonso community than in the Kanyawara community.

Huffman (1990) compared the association index between old/prime males and female chimpanzees at Mahale. Old and prime adult males associated more with other males, and old and prime adult females associated more with young and prime adult females than with individuals of the opposite sex.

For West African chimpanzees, Boesch (1996) found a high percentage of mixed parties (parties involving both males and females) at Taï. Although female-female associations were about one third less strong than male-male associations, their associations were much higher than those in other chimpanzee populations (Wittig and Boesch 2003). Female-female association did not occur randomly: 84 % of adult females had at least one preferred partner, and 78 % of these preferences were maintained for at least three of four consecutive years (Lehmann and Boesch 2009).

We examined dyadic association patterns among Kalinzu chimpanzees. Because estrous females tend to associate with males (Hashimoto et al. 2001) and because the previous studies analyzed only anestrous female association patterns, we used a dataset from anestrous females. We calculated the dyadic association index (DAI, Nishida 1968; Martin and Bateson 2007): DAIAB = ∑A B/(∑A + ∑B − ∑AB), where ∑A is the amount of time that only A was observed in a party, ∑B is the amount of time individual B was seen in a party, and ∑AB is the amount of time individuals A and B were in the same party. Cluster analyses were performed employing the unweighted pair group method using arithmetic averages (UPGMA, Sneath and Sokal 1973). The association patterns of the whole community are shown in Fig. 5.6, which shows separate clusters for males and females. Although most females used almost the entire group range, levels of association among females were low (Fig. 5.7a). The mean dyadic association index (DAI) between males (0.093 ± 0.040 (SD), n = 231) was significantly larger than that between males and females (0.020 ± 0.019, n = 484) and between females (0.023 ± 0.040, n = 231). There was no significant difference in the DAIs between males and females and those between females. One pair of females (ga-mn) formed the closest association in the group (Fig. 5.6). When we observed them for the first time in 2001, both of them were immature individuals (ga was an early adolescent (Goodall 1986) and mn was a juvenile) and ranged together without their mothers. Therefore, we presumed them to be sisters coming from another community. Our results therefore confirm that association among chimpanzee females is in general weaker than that among males. However, the level of female-female association varies considerably among sites.

Fig. 5.6
figure 6

Dendrogram based on the DAI of chimpanzees in Kalinzu drawn by UPGMA method. The names of females are written in small letters and those of males are written in capital letters

Full size image
Fig. 5.7
figure 7

Comparison of DAIs for (a) chimpanzees of M group in Kalinzu and (b) bonobos of E1 group in Wamba. A significant difference among dyads of different sex combinations was found for chimpanzees (one way ANOVA, df = 2, F = 482.8, p < 0.01), but not for bonobos (df = 2, F = 0.02, n.s.). (*Pairwise comparisons showed significant differences in the DAI between chimpanzee male-male and male-female dyads and between male-male and female-female dyads (Fisher’s LSD test, p < 0.01))

Full size image

Although there are only a few studies on bonobo associations, the pattern is different from that in chimpanzees. A review by Furuichi (2009) revealed that the mean relative party size (i.e., the percentage of individuals in the party from all individuals in the group: Boesch 1996) tends to be larger in bonobos than in chimpanzees (Wamba: Kuroda 1979; Furuichi 1989; Mulavwa et al. 2008; Lomako: White 1988; Hohmann and Fruth 2002; Hohmann et al. 2006). Although there was no significant difference, the presence ratio (i.e., the proportion of observation days on which each individual was observed in a mixed party) was slightly higher between females than between males in E1 group at Wamba during a period in which bonobos were food provisioned (90.9 % for males and 92.1 % for females) (Furuichi 1989). This result was not due to the provisioning, as a study conducted long after the cessation of provisioning also found that the relative party size was slightly higher for females than for males (Mulavwa et al. 2008).

At Lomako, DR. Congo, White (1988) found that, on average, there were more females than males in parties in the Hedrons, Rangers, and Blobs communities. Association among females was the highest among all combinations, and cluster analysis showed that most subgroups included both males and females (White and Burgman 1990). Hohmann and Fruth (2002) analyzed dyadic association in the Eyengo community and found that dyadic association between community members (association time >25 %) occurred most frequently between females, followed by mixed-sex dyads and male-male dyads. Moreover, long-term association occurred between male-female dyads, while most female-female association did not last longer than one field season (Hohmann et al. 1999). Similarly, female-female pairs were more cohesive (based on the tendency of pairs of animals to be close together, measured by Jacobs’ index) than male-male pairs or male-female pairs in the Eyengo and Splinter communities (Waller 2011).

We examined dyadic associations in the E1 group at Wamba (Fig. 5.8) using the same method we used for Kalinzu chimpanzees. Cluster analysis using the UPGMA method showed the strongest associations between the alpha male (TW) and some females. Remnants of a disappearing group who joined the E1 group in 2004 (ND, yk, jk) formed a separate cluster. A young adolescent female (me) that joined E1 group recently was most distant. The DAI was considerably higher than that in chimpanzees, and there was no significant difference among DAIs for male-male, male-female, and female-female dyads (Fig. 5.7b). Although Furuichi (Furuichi 1989) showed that some mother-adult son pairs showed a high frequency of association, we did not find this tendency. The mother of the alpha male TW had already died. Although we assumed that ki and NB were mother and son from their facial characteristics and DNA analysis (Hashimoto et al. 2008), they did not show a close association during this period.

Fig. 5.8
figure 8

Dendrogram based on the DAI of bonobos at Wamba drawn by the UPGMA method. The names of females are written in small letters and those of males are written in capital letters

Full size image

In summary, the association between female bonobos is stronger overall than that between males or that between males and females (Fig. 5.9).

Fig. 5.9
figure 9

Close association in female bonobos. Many females came together around the female whose fingers were caught by snares

Full size image

Discussion

It has generally been assumed that East African female chimpanzees do not form close social relationships with one another. When we compared association data between our study groups, female chimpanzees were indeed less gregarious than female bonobos, which is consistent with previous studies (Stumpf 2011; Watts 2012). However, recent studies based on long-term data collection from several sites revealed a large variation in female-female associations across East African chimpanzees. While females in some communities do not form close associations, such as at Gombe, Kanyawara, and Kalinzu (Fig. 5.10; Goodall 1986; Wrangham et al. 1992; Emery Thompson and Wrangham 2006; this study), females in some communities, such as Ngogo and Sonso (Fawcett 2000; Emery Thompson and Wrangham 2006; Wakefield 2013), form close and consistent associations with at least some other females. Although many studies of chimpanzee social behavior have focused on male-male or male-female relationships, more studies on female-female relationships are needed to fully understand the social structure of chimpanzee communities.

Fig. 5.10
figure 10

Mother and her offspring. Female chimpanzees tend to forage in their mother’s party

Full size image

There seems to be a large variation in female chimpanzee dispersal patterns. Although it is well established that chimpanzees have a male-philopatric social organization, the proportion of female transfers varies among communities. While the majority of females leave their natal group during adolescence and transfer to other groups at sites like Mahale and Taï (Boesch and Boesch-Achermann 2000; Nishida et al. 2003; Stumpf 2011), nearly half of females remain in their natal group at Gombe (Pusey et al. 1997). Even transfer of parous females with offspring occurs at some sites (Mahale: Nishida et al. 1990, 2003: Sonso: Emery Thompson et al. 2006; Bossou: Sugiyama and Fujita 2011). Such variation in female transfer may be related to a range of reasons that change over time, such as inbreeding avoidance (Pusey 1990), availability of energy for dispersal (Stumpf et al. 2009), and the presence or absence of neighboring groups (Sugiyama 2004).

There is also great variation in female ranging and association patterns across communities, which seems to be related to social and/or ecological conditions. At Gombe and Kanyawara, females seem to compete with each other over high-quality areas with good food availability, resulting in higher-ranking females having higher-quality core areas and higher reproductive success than lower-ranking females (Emery Thompson et al. 2007a; Murray et al. 2007). At Sonso and Ngogo, food resources are abundant and stable (Chapman et al. 1999; Newton-Fisher et al. 2000), and female chimpanzees do not show either extensive overlap in space use or competition over core areas (Fawcett 2000; Wakefield 2008, 2013). Moreover, because of the huge size of the Ngogo community, associated travel costs may place an upper limit on party size, and female chimpanzees form cliques to maintain consistent social relationships among limited numbers of individuals to minimize any costs of gregariousness (Wakefield 2013). By contrast, traveling alone may be too risky for females at Taï because chimpanzees frequently encounter leopards and suffer a higher risk of predation than at other sites (Boesch 1991; Boesch and Boesch-Achermann 2000). This leads to higher gregariousness in Taï females.

Social factors may also affect association patterns. The size of the main study community at Taï decreased from 54 individuals in 1999 to a minimum of 34 individuals in 2006, and the number of females decreased from 19 to 11. Following this decline, party size and dyadic association between females decreased (Wittiger and Boesch 2013). By contrast, changes in male social dynamics did not affect female gregariousness at Kanyawara (Gilby and Wrangham 2008). A comparison of periods with different alpha males revealed that changes in the strength of association occurred more frequently among males, while patterns of female association were less sensitive to changes in the male hierarchy. Overall, these findings may indicate that associations among males and among females tend to be influenced by different social factors.

In female-philopatric primate species, variation in association among females seems to be limited compared with that among males. In Japanese macaques, for example, grooming cliques of female kin are found in many of the groups studied, whether of large or small size, provisioned or wild, or living in different environmental conditions (Mori 1975; Furuichi 1984). By contrast, the association among male Japanese monkeys, and their emigration rate, is generally different across groups. For example, in Takasakiyama, some males stay in the natal troops and obtain high dominance rank, while all males in Yakushima and Kikazan leave their natal group (Sprague et al. 1998). While female monkeys confine their grooming to kin in most of the populations studied, male-male grooming relationships are different among populations. For example, males frequently groom one another in Yakushima-M troop, while males rarely groom one another in Yakushima-Ko troop, from which the Yakushima-M troop fissioned (Furuichi 1985).

It should also be noted that both one-male groups and multi-male groups are found in the same species according to ecological or reproductive conditions. For example, in Hanuman langurs, both one-male groups and multi-male groups have been observed (Sugiyama 1964; Newton 1988). Patas monkeys and forest guenons usually form a one-male group throughout the year, and extra-group males enter the troop during some conceptive seasons to copulate with females (Cords 2000; Carlson and Isbell 2001). This suggests that the males of female-philopatric species, which typically do not have ties with kin, can change their association with other males rather freely according to the social and ecological conditions.

The variation in female chimpanzee association patterns may be understood in the same context. Because they do not in general depend on kin-based ties, females can more freely manage their associations according to the social and ecological conditions, by changing their associations with members of other groups (or communities) and transferring between groups. Wrangham (1979) reported that three anestrous mothers occasionally associated peacefully with males of both the northern and southern communities at Gombe. At Kalinzu, we observed two young adult females move back and forth between neighboring communities and associate peacefully with members of both communities. Nishida et al. (1985) reported female transfer during the process of group extinction. When the number of adult males in K group dropped to only two, many cycling females of K group began to additionally associate with the males of M group. When only one adult male remained in K group, all the fertile, cycling females of K group began to associate mostly with M group. Such varying patterns of female association and transfer may suggest that the “group” is not as important an entity for female chimpanzees as it is for males.

Bonobos seem to show even higher flexibility in female association with members of other groups or communities. Research at Wamba shows that when two groups encounter each other, not only do they forage together for several days, but females also mate with males of the other group (Idani 1990). Although DNA analysis showed that most infants were sired by resident males of the Eyengo community in Lomako (Gerloff et al. 1999), there is a possibility that relatedness among resident males is weaker in bonobos due to such extra-group copulation. It is interesting that bonobo males also show more flexible associations with males of other groups than do chimpanzee males. Hohmann (2001) reported that two strange males entered the Eyengo community and one of them stayed there for at least one year. At Wamba, we observed several cases in which adult males disappeared from the study group for several months, eventually coming back to the original group. Although those males were observed sometimes ranging alone, it is possible that they temporarily visited other groups (chapter 6). Moreover, Hashimoto et al. (2008) reported permanent transfer of male bonobos at Wamba. After the likely extinction of neighboring groups during warfare, two adult males and two adult females with dependent infants joined the E1 study group and became stable members (Hashimoto et al. 2008).

Variations in female transfer and association are also found in eastern gorillas (Gorilla beringei). Some females leave their natal group before sexual maturity, while others give birth in their natal group (Robbins et al. 2009; Robbins & Robbins, chapter 4). The pattern of intergroup transfer varies according to sites, subspecies, and the period of observations (Yamagiwa et al. 2011). Studies of associations among female chimpanzees have increased in recent years, but the number of study groups and study areas remains limited to fully understand the extent and implications of this variation. Furthermore, our knowledge of female bonobo associations is derived from only a few study groups. We need more long-term data from different sites for chimpanzees and bonobos to obtain a full picture of associations for both sexes and to understand the adaptive meaning of the variation of such associations in Pan species.

Conclusion

Previous studies reported that female East African chimpanzees spend most of their time in small overlapping core areas within the community range. In this chapter, we examined the ranging and association patterns of chimpanzees and bonobos at a variety of sites, including new data from chimpanzees of the Kalinzu Forest and bonobos at Wamba. We found that male chimpanzees utilize the whole community range and that the home range of chimpanzee males is larger than that of females. However, the ranging pattern of female chimpanzees varies among study sites. Female chimpanzees at Taï, Sonso, and Kalinzu utilize a whole community range, supporting the bisexually bonded community model. The ranging patterns of female chimpanzees at Gombe and Kanyawara support the male-bonded community model. The differences in female ranging patterns are not explained by subspecies differences. Our findings show that both male and female bonobos utilize the whole group range, supporting the bisexually bonded community model. The association patterns of female chimpanzees also vary among study sites, although females are less gregarious than males at all sites. At Gombe, Kanyawara, and Kalinzu, association between female chimpanzees is very weak. At Taï, Ngogo, and Sonso, female chimpanzees have specific female partners with whom they associate more than average. These differences in association patterns are not explained by subspecies differences either. Associations between female bonobos are stronger than those between male bonobos. The variation in female chimpanzee patterns of association and intergroup transfer show that, as they are free from kin ties in a male-philopatric society, they can change their behavior according to various factors, such as ecological and/or social conditions.