Primates (2003) 44:203–216 DOI 10.1007/s10329-002-0012-x O R I GI N A L A R T IC L E Daniel Stahl Æ Werner Kaumanns Food competition in captive female sooty mangabeys (Cercocebus torquatus atys) Received: 5 March 2002 / Accepted: 18 October 2002 / Published online: 15 April 2003
Japan Monkey Centre and Springer-Verlag 2003 Abstract We studied the social and foraging behavior of two captive groups of sooty mangabeys under two different spatial food situations. These food conditions were clumped (food was placed in a box) and dispersed (food was dispersed over the entire enclosure). In each group five adult females and two adult males were observed. As a criterion for food competition, individual differences in the relative food intake were used. Adult female mangabeys had a linear, stable, and unidirectional dominance hierarchy. Access to food was rank dependent among females only under clumped food distribution, as current models of the evolution of primate social systems predict. However, feeding success appeared to be mediated not by female but by male agonistic behavior toward females. High-ranking females received relatively less aggression from males and could, therefore, stay and feed longer in the feeding area. Male tolerance of higher-ranking females seems to mediate female feeding success under restricted food resources. The establishment of a special relationship with a high-ranking male might, therefore, be a strategy to get better access to food. This study demonstrates that female competition for access to food should not be analyzed separately from male influences on females and suggests that a more integral role of males in socioecological models of the evolution of primate social systems should be considered. Keywords Food competition Æ Dominance Æ Social tolerance Æ Mangabeys D. Stahl (&) Æ W. Kaumanns Deutsches Primatenzentrum Göttingen, Germany D. Stahl Yerkes Regional Primate Research Center, Atlanta, GA, USA Present address: D. Stahl Department of Primatology, Max Planck Institute for Evolutionary Anthropology, Deutscher Platz 6, 04103 Leipzig, Germany E-mail: stahl@eva.mpg.de Introduction Competition is generally regarded as one of the main components of natural selection (Keller and Lloyd 1992). In most animals, the sexes compete for different key resources that limit their reproductive success (Trivers 1972). Therefore, Wrangham (1979) suggested for group-living primates linking ecological factors primarily to female–female rather than to male–male or male–female relationships. Wrangham (1980) introduced a socioecological model on the evolution of social systems in primates that assumes a causal relationship between diet and its distribution in space and time, dominance system, and migration pattern. This model was extended by van Schaik (1989). In this model, predation pressure is regarded as the ultimate factor of group life per se and food competition among females especially within groups as the main ultimate force of the inner structure of the group (van Schaik 1989; van Hooff and van Schaik 1992). Van Schaik (1989) distinguishes between two types of competition: scramble and contest. He defines food competition as of the scramble type if individuals cannot effectively exclude others from a common resource. The resource tends to be shared equally (e.g. leaves for mountain gorilla, Gorilla gorilla beringei: Watts 1988), and food intake success and therefore reproductive success negatively depend primarily on group size. Van Schaik argues that aggression over food and coalitions and alliances between females are not effective in improving access to food. Hence, during the course of evolution, females should develop individualistic and egalitarian dominance hierarchies, and agonistic support should be rare and not kin dependent. [cf. non-femalebonded (NFB) social system: Wrangham 1980]. Contest-type food competition occurs if the distribution of food resources [especially fruit trees; e.g. longtailed macaques (Macaca fascicularis): van Schaik and van Noordwijk 1988] allows some animals to exclude others from obtaining a greater share of the resource 204 than otherwise. Aggressive behavior gives rise to differences in access to food. Highly differentiated social structures of the female-bonded (FB) type (cf. Wrangham 1980) should evolve: females should form coalitions with kin to increase their power to monopolize a food resource, resulting in a linear, stable, unidirectional, and matrilineal dominance hierarchy within the group. Female reproductive success should be positively influenced by female rank within the group. Van Schaik’s socioecological model withstood evaluation by field studies in many cases (Sterck et al. 1997; for a review of alternative models: Matsumura 2001). However, most primate field studies are of a descriptive nature and can therefore only provide evidence for the influence of ecological factors on social structures. Intraspecific phenotypic variation of food competition and other patterns of social behavior do not simply depend on the ecological factor ‘‘food distribution’’ but also on demographic factors and confounding variables, for example, predation pressure (van Schaik 1989; Barton et al. 1996). Thus methodological difficulties arise in the study of the causal link between food distribution and social system in group-living primates, and controlled experimental studies are needed as an important extension to descriptive field studies (van Hooff and van Schaik 1992). A quantitative experimental study on food competition under strictly controlled conditions is almost impossible in the wild. For this reason we chose an experimental study on food competition with captive primates to test hypotheses derived from the van Schaik model. If food competition is a deciding factor in the evolution of social systems, then the resulting basic competitive structures should be found in captive studies as well. A recent field study by Range and Noë (2002) showed remarkable similarities of dominance structures and of female social relationships of wild sooty mangabeys compared to the social system described by captive studies. The van Schaik model will be used to predict how individuals of a group of primates behave in their daily life under different environmental conditions. The comparison between expected and observed behavior allows a discussion about the general acceptance of the socioecological model for the evolution of social structure. Van Schaik’s socioecological model as well as a recent extension by Sterck et al. (1997) did not consider the influence of males on females’ feeding success. Although van Hooff and van Schaik (1992) discussed the possibility that males might supply females with food by defending territories or food patches, almost all studies focusing on female food competition in primates only analyzed female–female relationships. Furthermore, adult males are always present in groups and compete for food as well (cf. Clutton-Brock 1977). Males might affect females’ feeding success (e.g. Janson 1988) and should be considered as a confounding variable. For these reasons, we analyzed the adult males’ behavior as well. As research subjects for a study on female social relationships we chose sooty mangabeys because of their rather unusual social system. Sooty mangabeys live in West Africa, ranging from Liberia to Senegal (Schwartz 1921; Struhsaker 1971). They live in multimale–multi-female groups with males considerably larger than females. Group size ranges from approximately 15 (Sierra Leone: Harding 1984) up to 90 animals (Ivory Coast: Bergemüller 1998). Their large (approximately 8 km2) home ranges overlap considerably (Bergemüller 1998). Sooty mangabeys are frugivourous, and agonistic interactions over food items were observed in the field (Bergemüller 1998; Range and Noë 2002). Mangabeys are in general described as FB species (van Hooff 1988). Research on captive sooty mangabeys has shown that there is indeed a stable, linear, unidirectional hierarchy among females as expected for a FB species (Gust and Gordon 1994; Stahl 1998; Stahl and Kaumanns 1999). A stable, linear hierarchy with unidirectional dominance relationships among adult female sooty mangabeys was also observed in a field study by Range and Noë (2002). Yet, captive sooty mangabeys do not exhibit a matrilineally based social structure as shown in FB species: affiliative (Ehardt 1988), aiding (Gust and Gordon 1993), reconciliation (Gust and Gordon 1993), and tolerant (Stahl 1996) behavior as well as dominance rank (Gust and Gordon 1994) among adult females are not kin dependent. Furthermore, aiding among females is low compared to female rhesus monkeys, a ‘‘classic’’ FB species (Gust and Gordon 1993). Sooty mangabeys, therefore, show behavioral features that are typical for an NFB species as well. Since the social system of sooty mangabeys cannot be classified clearly as either FB or NFB, its study can provide a promising tool to assess the predictive power of van Schaik’s model. In this experimental study, food intake and social behavior of two groups of sooty mangabeys were investigated under two different conditions. The first feeding condition allowed each individual access to food simultaneously and was expected to induce mainly scramble-type competition. The second feeding condition allowed only a few animals to feed at the same time at the food resource and was expected to induce contesttype competition. It is hypothesized that if sooty mangabeys are adapted to contest food competition sensu van Schaik (1989), females should compete for access to food under clumped food distribution. A female’s social position should affect access to food under clumped but not under dispersed food distribution. The differential feeding success among females should be mediated by direct competition between females. Changes in the variance of feeding success should be positively associated with rates of aggressive and submissive behavior (Janson and van Schaik 1988; van Schaik 1989). If sooty mangabeys are adapted to an envi- 205 ronment in which mainly scramble food competition has played an important role in shaping the species’ social structure, then the range of behavior expressed in this study should fall within some species-specific phylogenetic constraints. Therefore, if mangabeys are adapted to scramble competition, there should be no effect of social position on access to food. In a species adapted to scramble competition we expect that agonistic interactions should not be associated with the type of food distribution, as it was shown in captive studies with NFB primate species (Presbytis obscurus: Kerscher 1991; Papio hamadryas: Gore 1993; Zinner 1993). During the experimental period, the animals were observed during two different feeding conditions, clumped and dispersed. Both groups were accustomed to feeding under clumped and dispersed food distributions during a preliminary experimental test phase, which lasted 3 weeks. Regular feeding by the caretakers also included rather clumped (pellets were distributed outside the fence for a length of about 3–4 m) and dispersed (fruits were evenly distributed over most of the compound) feeding conditions. Under both feeding regimes the animals were observed daily, twice during feeding (morning and afternoon feeding sessions) and once outside feeding times. Non-feeding sessions took place 15 min after the end of the morning feeding observations. It was during this time period that animals were the most active outside of feeding sessions. Feeding conditions Methods Study groups Two social groups of sooty mangabeys (Cercocebus torquatus atys) were observed in the outdoor enclosures at the Field Station of the Yerkes Regional Primate Research Center in Lawrenceville near Atlanta, Georgia, USA. All mangabeys were descendants of a group of 27 animals (Bernstein 1971). The groups were living in compounds of the same size and with similar arrangements. The outdoor enclosures in which all observations took place had a surface area of 225 m2 (15x15 m). Normally the animals had access to water ad libitum, commercial monkey chow twice (morning and afternoon), and fruit once (afternoon) daily. The group size of 21 and 23 animals, respectively, and their multi-male–multi-female group composition resembled small to medium-sized natural mangabey groups. At the onset of the study, group S1 consisted of two adult males, five adult females, four subadult males, two juvenile males, four juvenile females, and three infants. One infant was born during the study. Group S3 was made up of two adult males, six adult females, two subadult males, one subadult to adult female, two juvenile to subadult females, three juvenile males, two juvenile females, and one infant (see Table 1 for details). Four infants were born during the study. The assignment to age classes was based on Gust (1994). Table 1 Focal animals of groups S1 and S3: shown are sex, age in years at onset of the study, rank in its sex class, average weight (in kilograms) used for estimating basal metabolism rate (BMR), estimated daily BMR (kilojoules per day), and mean correction value for females due to reproductive state for each of the four observations blocks (Cl clumped. Di dispersed food distribution; 1 and 2 first and second observation period under type of distribution) Experimental design Animal (code) Sex Age in years Under ‘‘clumped’’ food distribution, food was placed in a box (62·18·10 cm, length x width x height) fixed outside the fence. Only a few animals (up to four to five adult animals) could eat at the same time directly at the food box. Under ‘‘dispersed’’ food distribution, food was evenly distributed in most parts of the outdoor enclosure. A caretaker distributed the pellets from an observation tower while the first author started recording the behavioral data. A pilot study showed that under dispersed distribution animals were distributed all over the compound, moving around and feeding in a more or less undisturbed way (Stahl 1998). A monopolization of food by one or more animals was not observable. Only standard primate pellets were fed during the food competition experiments. The use of pellets allowed an accurate assessment of the amount an animal ate. Food competition was accomplished by restricting food to the amount they actually needed and could eat within an hour or less. The amount of food was determined during a test phase. This amount of food was enough to ensure that all animals did get enough food but forced the animals to feed and compete within the 1-h observation period. All food was eaten within this time period. The well-being of the animals was of first priority. Every weekday an independent observer, who knew the groups well, visually checked all animals’ state of health. Furthermore, all animals of a group were weighed once before and during a clumped food distribution session to ensure that no animal lost weight. Rank Average weight (kg) BMR (kJ/day) Correction value for BMR Cl 1 Di 1 Cl 2 Di 2 Group S1 Sgo M Rgo M Wco F Veo F Ilo F Blo F Fdo F 8 8 14 10 4 4 13 M1 M2 F1 F2 F3 F4 F5 13 13 7.55 8.5 6.5 6.8 7.3 1,965 1,965 1,307 1,429 1,168 1,209 1,275 1.25 1.15 1.5 1.44 1 1.5 1.15 1.5 1.5 1 1.5 1.15 1.5 1.5 1 1.5 1.15 1.5 1.5 1 Group S3 Pfo M Jgo M Mdo F Cho F Ugo F Zko F Zgo F 10 9 15 9 9 5 9 M1 M2 F1 F2 F3 F4 F5 13 11.9 9.0 9.1 9.6 7.7 8.4 1,965 1,833 1,491 1,498 1,559 1,327 1,410 1.25 1.25 1.15 1 1.4 1.42 1.34 1 1 1.4 1.5 1.5 1 1 1.4 1.5 1.5 1 1 1.5 206 Observation procedures Data analysis Focal animal sampling (Altmann 1974) Focal animal and instantaneous scan samples were not included in the analyses if a female was in peak estrous, to minimize the effects of estrous (Gust and Gordon 1991). During most of the time of the study the females were either pregnant or nursing their offspring and only few estrous (16 peak swellings from ten females) were observed. A separate analysis was, therefore, not possible. Focal animal sampling was used to collect behavioral data. In each group seven animals (two adult males and five adult females, see Table 1) were in focus. One adult female of group S3 was blind and therefore not included as a focal animal. During a focal observation an animal was observed for 10 min. Its behavior was recorded continuously. Observations concentrated on an individual’s foraging and food intake behavior and its agonistic relations and grooming behavior with all other group members. Under clumped food distribution it was recorded how long the animals stayed within the ‘‘feeding area’’ (area within a radius of 2 m around the food box) and within touching distance (<0.5 m) of the ‘‘food box’’. In addition, the duration the focal animal stayed together with other focal animals at the food box was determined. The different types of aggressive and submissive behavioral interactions were categorized into aggressive and submissive behavior. Submissive behavior included avoidance behavior. The following behaviors were recorded: – – Behaviors associated with ‘‘aggression’’: threat stare, head bob, lunge, displace, charge, chase, slap (hit), grab, pin to ground, tail bite, other bites, fight Behaviors associated with ‘‘submission’’: avoid, move away, submissive present, tongue flickering, fear grimace, redirect threat, crouch, flee Most definitions were taken from unpublished manuscripts by D. Gust and I. Bernstein; a detailed description of the behavioral definitions is published in Stahl (1998). For each focal animal the mean active aggressive and submissive behavior toward (a) focal adult males and (b) focal adult females during a 10-min focal observation was calculated. Interactions with other non-adult animals of the group were recorded and analyzed but are not presented here (Stahl 1998). Observation schedules During each of the three daily observations blocks (morning feeding, non-feeding, and afternoon feeding sessions) six 10-min focal animal samplings were done. The order of the focal animals was changed randomly during each observation block. Under each feeding condition the animals were observed twice about 6 weeks in a row. Within a 60-min feeding session every focal animal was observed in a 10-min focal animal sampling at each period of time (0–10th min, 11–20th min, ..., 51st–60th min) at least three times and therefore in each 10-min-block under each feeding and non-feeding condition at least six times. Each animal was observed for at least 6 hours during each morning feeding, non-feeding, and afternoon feeding session under each food distribution type. Animals of group S1 were observed 120 days during a 7-month period, and animals of group S3 were observed 116 days during a 6-month period. Under both feeding conditions animals were observed twice in rotating order to distribute the effect of season equally. Instantaneous scan sampling (Altmann 1974) During feeding times the entire group was rapidly scanned (about 30 s) at the end of each 10-min focal animal observation by marking the positions of the animals in a copied drawing of the enclosure. The positional behavior data were used to determine proximity patterns between males and females. Following Zinner et al. (1997) we used the relative number of certain positional behaviors (e.g. staying in the feeding area with an adult male) after a 10-min observation block as a measure of duration for the previous 10-min observation block. Dominance relationships Dominance relationships between focal animals were assessed from the distribution of dyadic aggressive and submissive interactions. Data of agonistic interactions were obtained from focal animal sampling and additionally from ad libitum sampling. Ad libitum sampling was done opportunistically outside of the focal animal sampling observations. As in the literature (Bernstein 1976, Gust and Gordon 1994), a linear dominance hierarchy with males dominant over females was easily established. Among the adult animals there were decided and temporarily stable dominance relations within each possible dyad independent of feeding context. Dominance relationships could be deduced by either aggressive or submissive interactions. No coalitions or alliances against adult males were observed. For further details on dominance relationships see Stahl (1998) and Stahl and Kaumanns (1999). Individual energy gain Animals were fed entire pellets or parts of pellets. By counting the numbers of fully eaten pellets and the number of bites of a pellet, the ingested energy gain could be estimated. Pellets weighted on average 1.85 g (dry weight) and contained 36 kJ of digestive energy (information from the producer, Harlan Teklad, Madison, Wis., USA). In a preliminary test phase, bite size in relation to time was estimated. Average bite size differed between the sexes and feeding conditions and decreased over feeding time. The decrease of bite size over time was caused by animals having to feed more and more on fragments of the original pellets. Therefore, energy gain had to be calculated for each observation block for males and females during the two feeding situations separately. Bite size was estimated by counting the number of bites per fully eaten pellet. Fragment size was estimated by collecting and weighing pellet fragments at different periods of times. Energy gain was calculated for a 10-min block by the following formula: hX ingested energy ¼ ðwhole eaten pelletsÞ i X þ ðbites
bite sizeÞ
36 kJ Individual feeding success Feeding success can be estimated by comparing the energetic gain with the energetic need of an individual. Individual energetic need differs between animals and depends on several factors, such as body weight, activity, reproductive status (for females), thermoregulation, and growth. Therefore, to compare feeding success between individuals, a relative measurement of energetic need must be used. For a comparison of feeding success the energy gain rate of an animal was set in relation to its estimated basal metabolism rate (BMR). The BMR is the energy rate of a resting, not digesting body within its temperature optimum, which is necessary to maintain its function. In many species BMR is a linear function of the logarithm of body weight (Kleiber 1961), which was confirmed for non-folivorous prosimian primates (Müller 1985; Ross 1992). 207 Among females, BMR was adjusted for their reproductive state. Basal metabolism rate was increased during the latter stages of pregnancy by 25% and during lactation by 50% (Lee and Bowman 1995). If a female was only occasionally lactating we assumed an increase of 15%. A pilot study following the methods of Coehlo et al. (1976) showed that differences in energetic need due to different activity budgets were negligible in this captive environment as was also seen in a captive study on hamadryas baboons (calculated from data of Zinner 1993). Feeding success was measured by the amount of energy an animal gained during a period of time in relation to its adjusted BMR: relative energy gainðtÞ ¼ energy ingested ðkJ=tÞ adjusted BMR ðkJ=24hÞ where t is time period in hours, w is body weight of an animal in kilograms, c is the correction value for reproductive state of females: 1.25 if pregnant, 1.5 if lactating, 1.15 if lactation is phasing out, and adjusted BMR=287·w0.75·c in kilojoules per 24 h. The body weight is the mean of the weight measured before and near the end of the study during the observations. Body weight correction value for reproductive state of females can be seen in Table 1. Spatial male–female relationships Male–female spatial relationships were analyzed during the first 20 min of a feeding session under clumped food distribution. To describe spatial male–female relationships, instantaneous scan samples of positional behavior after the first and second 10-min focal animal sampling were used to determine interindividual distances between adult females and both adult males. The distances within a dyad were classified into four distance categories (Table 2). Afterward, the relative number of scans in each category was calculated. Because a single score is easier to use for analyses, a proximity index for each female–male dyad was calculated. Following the method of Perry (1997), the relative number of incidents of each distance category was multiplied with a weighting factor. The weighting factor was derived by the reciprocal of the area each distance class covered. This method weights closer proximity more heavily and takes into account the exponentially increasing chance of being seen or being a target of a behavioral act with increasing proximity. Almost no agonistic behavior occurred between animals at distances greater than 2 m during this observation period under clumped food distribution. Therefore these distances were regarded as a neutral distance and neglected. The method described by Perry for wild arboreal capuchin monkeys was adjusted for the mainly two-dimensional structure of the enclosure and for the terrestrial behavior of sooty mangabeys by using the reciprocal of the area and not of a concentric sphere. The area of the first three categories was calculated by the following equation: areacovered ¼pru2 prl2 where ru is the radius of the upper limit of category n, and rl is the radius of the lower limit of category n. For a better readability of the data the reciprocal of the area was multiplied by p and then used as the relative weighting factor. weighting factorcategory x ¼ 1=areacategory x
p The proximity index P between two animals was then calculated by P ¼n1
4þn2
1:33þn3
0:33þn4
0 where n is the proportion of samples in the relevant category. Statistics Statistical analyses were carried out by using the program STATISTICA for Windows 5.0, (StatSoft, Inc., 1996). Each focal animal was observed under two experimental conditions to investigate the effect of food distribution on its behavior. The same focal animal was observed at a given focal observation period at least six times. We averaged these replicated observation scores to obtain a single score for each individual under each experimental condition for further analyses (Sokal and Rohlf 1995). Two-factorial analyses of variances (ANOVAs) with a repeated measure design were used for the analyses of focal females (n=10) to describe the influence of two independent variables, food distribution and group affiliation (Sokal and Rohlf 1995). Group affiliation was included as an independent variable to reduce unexplained error variance. Correlations were done with Pearson product–moment correlation tests. For correlations involving members of both groups, the data were standardized by the mean of each group. Only relative data were therefore used for these correlations. Because of their higher power, parametric tests were preferred to non-parametric tests, although the relatively small sample size makes it difficult to evaluate the assumptions of such parametric tests. However, there were no outliers in the data set and data were unimodally distributed and not seriously skewed. On that condition the used parametric tests are robust against deviations against their theoretical assumptions (Sokal and Rohlf 1995). Statistical analyses including the males were done with nonparametric tests because of the small sample size (n=4). Analyses with dependent variables were carried out with Wilcoxon matched pairs tests and sex differences were analyzed with the Mann– Whitney U test (Siegel and Castellan 1988). Because there was no effect of time of day on relative energy gain of the animals, the mean of morning and afternoon feeding observations was used in the following analyses. Under dispersed food distribution almost all food was eaten within 20 min. Therefore, we used this 20-min period as a baseline of scramble competition for further analyses and divided the 60-min observation session of feeding under clumped food distribution into three 20min observation blocks (0–20 min, 21–40 min, and 41–60 min). During a feeding session under clumped food distribution, males and females obtained on average during the first two 20-min blocks more than 95% of their total food intake. Therefore, the last 20min block under clumped food distribution was ignored in the analyses (see Stahl 1998 for details). Means are reported with standard errors throughout this article. Only two-tailed tests were used. A null hypothesis was rejected at an a-level of 0.05 with the exception of dependent samples of males. Here, a null hypothesis was rejected at an a-level of 0.07 because of the small sample size. Table 2 Categories for interindividual distances between adult females and adult males Category Interindividual distance Area (m2) Weighting factor (1/area*p) 1 £ 0.5 m (arm’s reach) 0.5 m to £ 1 m 1 m to £ 2 m >2 m 0.785 4 2.36 9.42 1.33 0.33 0 2 3 4 Results The overall female mean relative energy gain within 60-min feeding sessions did not differ between the two feeding conditions (F(1,8)=0.046, P=0.836, n=10). Animals were foraging longer under clumped (‡50 min) than under dispersed food distribution where basically all food was eaten within 20 min. 208 Access to food Figure 1 shows the relationship between female rank and food intake success, measured as relative energy gain, under clumped and dispersed food distribution. Under clumped food distribution, high-ranking females gained more relative energy during the first 20 min of feeding sessions than lower-ranking females (Fig. 1a, Table 3). Lower-ranking females could not completely compensate for their deficit at a later time, resulting in a tendency for high-ranking females to be more successful in obtaining relative energy during a total 60-min feeding session (Fig. 1b, Table 3). Similarly, high-ranking females spent more time within the feeding area during the first 20 min under clumped food distribution (r=0.64, P<0.05, n=10; Fig. 2a), and time spend within the feeding area was positively associated with relative energy gain (r=0.84, P=0.005, n=10; Fig. 2b). There was no relationship between time spent in the feeding area and rank during the second or third 20-min block (21st–40th min: r=)0.315, P=0.375; 41st)60th min: r=0.018, P= 0.960, n=10) and overall, high-ranking females tended to spend more time within the feeding area during a complete feeding session under clumped food distribution (r=)0.587, P=0.074, n=10). Under dispersed food distribution there was at no time a correlation between female rank and relative energy gain (Fig. 1c, Table 3) or time spent in the feeding area (0–20th min: r=)0.098, P=0.787, n=10). Agonistic behavior Although female feeding success correlated positively with rank only during the first 20 min under clumped food distribution, there was no difference in aggressive and submissive interaction rates among the females compared with under dispersed food distribution or compared with the second 20 min under clumped food distribution (Figs. 3, 4). Only during the first 20 min under clumped food distribution were rates of male aggression toward females substantially higher than rates of aggression between females. During this time, males showed the highest levels of aggressive behavior toward females compared to other observation periods under clumped and dispersed food distribution (Fig. 3). Similar patterns were observed for female submissive behavior rates toward males (Fig. 4). During feeding under both food distributions no coalitions between females were observed. Table 3 Correlation between female rank and standardized relative energy gain during different observation blocks under clumped and dispersed food distributions (Pearson product–moment correlation, n=10) Period Fig. 1a–c Correlation between female rank and relative energy gain during the first and second 20-min block under clumped food distribution (a, b) and the first 20 min under dispersed food distribution (c). Relative energy gain is standardized by the mean of each group (n=10) Clumped food distribution Total (0–60th min) 0–20th min 21–40th min 41st–60th min Dispersed food distribution 0)20th min 0–10th min r P –0.61 –0.664 –0.244 0.111 0.059 0.036 0.497 0.760 0.380 0.310 0.276 0.386 209 Fig. 3a, b Female (a) and male (b) mean aggressive behavior toward adult females under clumped (cl) and dispersed (di) food distribution during the first (cl 0–20) and second (cl 21–40) 20-min block under clumped food distribution and the first 20-min block (di 0–20) under dispersed food distribution. Asterisk *P<0.05 (P<0.07 for dependent samples with males). Statistical test for female and male aggressive behavior as follows. Clumped 0–20th min versus dispersed 0–20th min: males (Wilcoxon test): T=0, Z=1.83, P<0.07, n=4; females (two-way ANOVA with repeated measurements): Fig. 2 a Correlation between female rank and standardized time spent in the feeding area during first 20 min under clumped food distribution. b Correlation between standardized time spent in the feeding area and standardized relative energy gain during first 20 min under clumped food distribution. Data are standardized by the mean of each group (n=10) Factor df F P-level Group Distribution Group·Distribution 1,8 1,8 1,8 2.949 0.251 1.042 0.124 0.630 0.337 and clumped 0–20th min versus clumped 21st–40th min: males (Wilcoxon test): T=0, Z=1.83, P<0.07, n=4; females (two-way ANOVA with repeated measurements): Factor df F P-level Group Distribution Group·Distribution 1,8 1,8 1,8 1.004 0.311 0.003 0.346 0.592 0.955 Feeding area Under clumped food distribution females’ mean stay in the feeding area during the first and the second 20-min block did not change (Fig. 5). Males stayed longer during the first 20-min block than during the second 20-min block under clumped food distribution (Wilcoxon matched pairs test: Z=1.826, P<0.07, n=4). During this time period males spent more time than females within the feeding area (Mann–Whitney U test: Z=)2.404, P=0.016, nmales=4, nfemales=10). At least one of the two males stayed within the feeding area during more than 80% (group S1: 80.9%; group S3: 82.2%) of the time (Fig. 5) and here mainly directly at the food box. During the remainder of this time period, usually at least one of the males was close by the feeding area. During this first 20-min block, females spent most of their time within the feeding area with at least one male (Fig. 6; two-way ANOVA with repeated measurements: F(1,8)=11.735, P<0.009). During the second 20-min block, males were less often within the feeding area, and females usually fed when males were not present in the feeding area (Fig. 6; two-way ANOVA with repeated measurements: F(1,8)=25.685, P<0.001). Male–female relations Direct competition between the sexes occurred, therefore, mainly at the beginning of the feeding session. During this time period, high-ranking females stayed in closer proximity to the males (r=)0.78, P=0.007; Fig. 7a). In both groups, it was mainly the dominant female who sat and fed especially with the alpha male directly at the food box (Fig. 8a, b). Taking proximity into account, high-ranking females experienced less aggression from and showed less submissive behavior toward males than did low-ranking females (aggression received: r=0.87, P=0.001; Fig. 7b; submission: r=0.78, P=0.008; Fig. 7c). Females who stayed closer to the males stayed longer within the feeding area (Fig. 9a; Pearson’s correlation: r=0.940, P<0.001) and gained more relative energy (Fig. 9b; r=0.867, P<0.001). Proximity to males thus accompanies females’ access to food under clumped food distribution. 210 Grooming relations Females did not groom males longer during non-feeding times following feeding sessions under clumped compared to under dispersed food distribution (Table 4). High-ranking females groomed the a- and b-males longer than lower-ranking females independent of the food distribution (Table 5). Males hardly groomed females during non-feeding observations under both food distributions (Table 4). The dominant female of both groups was in no case recipient of the most grooming by males. Discussion In this study, access to food among adult female mangabeys was rank dependent under clumped but not under dispersed food distribution, as the model of van Schaik (1989) predicts for species’ everyday behavior adapted to contest food competition and as has been shown for many FB species (Sterck et al. 1997): Highranking female mangabeys stayed longer within the feeding area and closer to the preferred food site than lower-ranking females, thereby obtaining more food. Experimental studies on two captive groups of NFB species did not reveal better feeding success of highranking females under either clumped or dispersed food distribution (Presbytis obscurus: Kerscher 1991; Papio hamadryas: Gore 1993; Zinner 1993). Contrary to the predictions derived from the van Schaik model, females’ differential feeding success at the restricted food resource was not accompanied by direct agonistic competition between females: there was no difference in female aggressive and submissive behavior between feeding under clumped and dispersed food distribution. If females were trying to restrict other females’ access to food, agonism levels should be higher under a clumped than under a dispersed food distribution, as has been described in several FB species (e.g. Cebus capuchinus: Phillips 1995a; Cercopithecus (Chlorocebus) aethiops: Pruetz and Isbell 2000; Macaca fuscata: Mori 1977; Saito 1996; M. mulatta: Belzung and Anderson 1986; Brennan and Anderson 1988; M. radiata: Boccia et al. 1988; P. anubis: Barton and Whitten 1993). Similarly to our study, in two captive studies of NFB species, agonism rates between females did not differ during feeding under clumped and dispersed food distribution (Presbytis obscurus: Kerscher 1991; Papio hamadryas: Zinner 1993). Females of both species maintained their scramble competitive social system under artificial clumping as expected for NFB species adapted only to non-monopolizable food resources. In a field study, Sterck (1995) showed that the displacement rates between females of the NFB Thomas langurs were similarly high to those of the sympatric, FB long-tailed macaques (see also Sterck and Steenbeck 1997). However, aggression rates in female langurs did not differ Fig. 4a, b Female submissive behavior toward females (a) and males (b) during the first (cl 0–20) and second (cl 21–40) 20-min block under clumped food distribution and the first 20-min block (di 0–20) under dispersed food distribution. Asterisks **P<0.01. Statistical tests for female submissive behavior as follows (two-way ANOVAs with repeated measurements). Clumped 0–20th min versus dispersed 0–20th min: Factor Toward females Group Distribution Group·Distribution Toward males Group Distribution Group·Distribution df F P-level 1,8 1,8 1,8 4.480 3.190 5.786 0.067 0.112 0.043 1,8 1,8 1,8 0.029 16.91 0.127 0.868 0.003 0.731 and clumped 0–20th min versus clumped 21st–40th min: Factor Toward females Group Distribution Group·Distribution Toward males Group Distribution Group·Distribution df F P-level 1,8 1,8 1,8 2.403 3.412 0.389 0.150 0.102 0.550 1,8 1,8 1,8 0.398 17.941 0.081 0.547 0.003 0.783 when feeding on leaves compared with feeding on fruits, which were expected to induce scramble and contest competition, respectively. The aggressive and submissive behavior rates between female sooty mangabeys under clumped compared with under dispersed food distribution, therefore, resembled those of several NFB species. The comparison of agonistic behavior between two time periods under clumped food distribution showed that aggressive and submissive behavior rates between females were not higher during the first time period when feeding success was rank related and individual variance was larger. Because mean time that females stayed within the feeding area during the two time periods did not differ, higher rates of aggressive and submissive interactions were expected during the first time period if females are adapted to contest competition (cf. Janson and van Schaik 1988; van Schaik 1989). 211 Fig. 5 Mean time (and standard error) females spent in the feeding area in comparison to total time at least one male stayed within the feeding area during the 0–20th min and the 21st–40th min observation block under clumped food distribution. Statistical test for the effect group membership and 20-min blocks on time females stayed in the feeding area during feeding session under clumped food distribution as follows (two-way ANOVA with repeated measurements): Factor df F P-level Group 20-min block Group·20-min block 1,8 1,8 1,8 0.720 0.280 0.026 0.421 0.611 0.875 Contest competition can be mediated not only by overt aggressive and submissive interactions but also by spatial deployment mechanisms, especially in systems with decided dominance relations. Subordinate animals can avoid overt costly conflicts over restricted resources (food or safe positions) by occupying locations away from the dominant animals and the contested resource as has been described for several FB species (e.g. Cebus apella: Janson 1990; C. olivaceus: Robinson 1981; Cercocipethecus aethiops: Whitten 1983; Macaca fascicularis: van Noordwijk and van Schaik 1987; M. fuscata: Saito 1996; M. mulatta: Belzung and Anderson 1986; Papio anubis: Barton 1993; P. cynocephalus: Collins 1984). However, only in one study by Whitten (1983 on C. aethiops) was rank-dependent access to clumped food patches maintained by spatial avoidance without a previous increase of aggression. In our study, food was presented at only one place in a spatially restricted environment. Low-ranking animals could not avoid conflicts by feeding at a different, lessprofitable or less-safe patch, as observed by Whitten (1983) in female vervet monkeys or by Saito (1996) in Japanese macaques) or by even foraging outside the main group (e.g. M. fascicularis: van Noordwijk and van Schaik 1987). Furthermore, in this study the amount of food was restricted to the amount they actually needed (no surplus). Because both mangabey groups were feeding more during ad libitum feeding than during the experimental phase, low-ranking animals could not avoid dominant animals at the food resource by feeding at a later time of day when higher-ranking animals were fin- Fig. 6 Mean relative number of scans females of group S1 and S3 spent in the feeding area with or without a male present in the feeding area during the 0–20th-min and the 21st–40th-min observation block of a feeding session under clumped food distribution (n=10). Asterisks **P<0.01; ***P<0.001. Right y-axis Mean time spent per 10 min estimated from instantaneous scan samples (seconds/10 min). Statistical test for effect of group membership and presence of males in the feeding area on time females spent in the feeding area, with and without males in the feeding area (two-way ANOVA with repeated measurements) as follows. Clumped 0–20th-min block: Factor df F P-level Group Males present Group·Males present 1,8 1,8 1,8 0.016 11.735 0.048 0.903 0.009 0.832 and clumped 21st–40th-min block: Factor df F P-level Group Males present Group·Males present 1,8 1,8 1,8 0.291 25.685 0.127 0.604 <0.001 0.731 ished feeding. This has been seen, for example, in vervet or capuchin monkeys (Whitten 1983; Phillips 1995b). It is, therefore, unlikely that the differences in females’ feeding success at the restricted food resource were mediated by subtle competitive mechanisms among the female mangabeys without an initial increase of aggression, especially since the same females did not avoid the high level of aggression of the larger males during the same time period. A comparison of wild and provisioned Japanese macaques by Hill and Okayasu (1995, 1996) showed that provisioning groups with food induces closer proximity among females. Closer proximity increases the chances of having potential allies around. Alliances were, therefore, more profitable and occurred more often than among females in wild groups. The authors assumed that the youngest ascendancy hierarchy in female macaques might be a result of a high level of opportunity for effective alliances in provisioned or captive groups. In our study both groups of mangabeys should therefore show more female–female alliances under clumped food distribution if females are adapted to contest 212 Fig. 8a, b Time individual females of group S1 (a) and group S3 (b) stayed simultaneously with the a- or b-male at the food box during the first 20 min of a feeding session under clumped food distribution (curves fitted by method of least square smoothing procedure). Male 1 a-male; male 2 b-male Fig. 7a–c First 20-min block under clumped food distribution: correlation between female rank and (a) standardized proximity index toward both males, (b) aggression received from males/ proximity index under clumped food distribution, and (c) submission shown toward males/proximity index competition. However, during feeding under both food distributions there were no female alliances observed. Concluding, different competitive feeding situations among female mangabeys did not result in changes of the agonistic level. Access to food in female mangabeys within a social group seems not to be mediated by direct competition between females as is the case in FB species. Although there is a strict dominance hierarchy among the female mangabeys, the relationships between them appear to be egalitarian in the context of food competition as defined by de Waal and Luttrell (1989). Competition for food between the female mangabeys can be described as mainly of the non-contest type and cannot account for the rank-related differences in feeding success. The van Schaik model does not describe this competitive social system. Male influence on female feeding success Rank-related feeding success among the studied adult females, however, was associated with a high level of male aggression and occupation of the food resource by one of the males. During this time period under restricted feeding conditions males were most aggressive toward adult females (and other non-adult groups) 213 Table 5 Correlation between female rank and time spent grooming the a- and b-male, respectively, of each group during non-feeding times under clumped and dispersed food distributions (Pearson product–moment correlation, n=10; time standardized by the groups’ mean) Females grooming a-male b-male Fig. 9a, b Correlation between standardized proximity index toward both males and (a) standardized time females spent in the feeding area and (b) standardized female relative energy gain during first 20 min under clumped food distribution. Time data are standardized by the mean of each group (n=10) Table 4 Mean duration (and standard error) of female–male and male–female grooming sessions during non-feeding observations under clumped and dispersed food distribution (seconds/10 min per individual) Grooming session Females groominga a-male b-male Females groomed by a-male (S1) b-male (S1) a-male (S3) b-male (S3) Clumped food distribution Dispersed food distribution 19.7 (11.67) 27.6 (11.61) 26.3 (9.82) 23.6 (10.95) 1.6 (1.57) 1.0 (0.66) 5.2 (2.39) 0 4.3 (1.95) 0.4 (0.27) 4.6 (1.81) 0 a Statistical test for females grooming the a- and b-male (nonfeeding observations under clumped vs. under dispersed food distribution; two-way ANOVA with repeated measurements): factor Group: df=2,7, F=0.861, P=0.124; factor Distribution: df=2,7, F=0.255, P=0.630; factor Group·Distribution: df=2,7, F=0.332, P=0.337 and, correspondingly, females’ rates of submissive interactions toward males were highest. Similarly, male aggressive behavior was highest during the time periods under restricted food resources when they stayed longest at the food resource and obtained the most food. Clumped food distribution Dispersed food distribution r P r P –0.791 –0.638 0.006 0.047 –0.645 –0.670 0.044 0.034 Females were, therefore, more affected by male than by female direct agonism under restricted food resources. Males restricted but did not fully monopolize access to food. Females’ access to food seems to be restricted differentially by the males: high-ranking females were less aggressed by the males and showed less submissive behavior toward them. Because of closer proximity to males, these females stayed longer within the feeding area and closer to the food box, which allowed better access to food. Female sooty mangabeys showed predictable dominance effects in feeding success similar to FB species but differential access to the food resource and feeding success among females seem to be determined by a lower competitive tendency of dominant males toward higherranking females. Males selectively expressed ‘‘social tolerance’’ (sensu de Waal 1986, 1989) toward higherranking females. In addition to dominance, social tolerance can determine feeding success and according to de Waal (1986, 1989) social tolerance opens an alternative route for subordinate animals to gain access to food or other resources. Being tolerated by a particular dominant male might give a female higher payoffs than trying to compete with other females. On an ultimate level, it could be a strategy for female primates to become tolerated by dominant males instead of competing with other females as van Schaik assumes in his model. In both groups, the dominant female showed strong associations especially with the dominant male. Among the females, the dominant female almost exclusively stayed and fed together with the dominant male at the food box. This strong association can be regarded as a ‘‘special relationship’’ (cf. Smuts 1985). Therefore, among female mangabeys in this study the special value of being dominant is that her close association with the male can provide her priority of access to food over adult females, the second-ranking male, and all other members of the group. Male–female associations were observed in wild sooty mangabeys (Range 1998). Although sooty mangabeys in the wild do not exhibit such restricted food resources as presented in this study, they feed on patches, which are unlikely to accommodate the whole group (Range and Noë 2002). A female’s association with a dominant male 214 can, therefore, be an advantage for better access to a food patch and for feeding unhindered within it. Most primate studies on food competition have focused only on female–female interactions and have not taken male influences on differences in females’ feeding success into account. However, a comparable pattern of competitive behavior in clumped food situations as shown in this study was observed among wild brown and white-capped capuchin monkeys (Cebus apella: Janson 1984, 1988, 1990; C. capucinus: Fedigan 1993; Rose 1994). While foraging, males were more aggressive than females toward females, and males’ agonisms adversely affected females’ foraging success. The dominant male restricted access to the best feeding sites where only the dominant female and her offspring were tolerated. In some clumped high-quality food patches, whole fruit trees were monopolized by this pattern (C. apella: Janson 1988, 1990). In several other primate species, it was also noted that adult females sitting close to their male ‘‘friend’’ (sensu Smuts 1985) had better access to restricted food resources than higher-ranking animals (M. fuscata: Takahata 1982; P. anubis: Smuts 1985; M. mulatta: Chapais 1986). Recently, the influence of adult male primates on female social relationships has been incorporated into the van Schaik model (van Schaik 1996; Sterck et al. 1997) but only male sexual coercion (Smuts and Smuts 1993), including male infanticidal behavior, was considered as a major ultimate factor. Sterck et al. (1997) suggested that in NFB species the threat of infanticide by new immigrating males may force females to leave the groups (e.g. mountain gorillas: Watts 1990; Thomas langurs: Sterck and Steenbeck 1997). Male infanticidal behavior after joining a new group or after a rank reversal may further explain the formations of male– female special relationships after mating because of male protection of the infant (van Schaik and Dunbar 1990; Smuts and Smuts 1993; van Schaik 1996). This study demonstrates another possible effect of males on female social relationships that is not considered in van Schaik’s model: the establishment of a special relationship with a high-ranking male can be a strategy to get better access to food. Competition for access to restricted food resources would be mediated via the males and not by direct competition between the females. This aspect should be considered in future (field) studies and may need to be integrated into current models of the evolution of social systems in primates. Infanticide or predation pressure is likely to act on the establishment of this social system as well. A female (and her offspring) may further benefit from a male’s close presence by getting better access to safe positions against predators or by protection against potential infanticidal males (as was observed by Busse and Gordon (1983, 1984) in captive sooty mangabeys). This would further enhance the power of male–female relationships. Helping and protecting his most likely infant as well as establishing and maintaining access to mating partners would explain why a male should have an interest in such a relationship. If access to food (or safe position) is not mediated by direct competition between females as proposed by van Schaik but by the differential tolerance of males toward females during foraging, females’ strategies to improve feeding success should be aimed at establishing and maintaining a relationship with a male. Grooming is generally regarded as a means to establish affiliative contact and to maintain social relationships (Boyd and Silk 1997). In this study, females groomed males more often than vice versa during non-feeding times and highranking females groomed the two males in each group more often than did the low-ranking females. Therefore, female mangabeys, especially high-ranking ones, seem to invest in a good relationship with males. The effect of males’ presence in groups on females’ competitive system has been hardly looked at, possibly because of the popular view that females compete mainly for food (Hemelrijk and Luteijn 1998). However, there are growing indications and observations that female primates compete for males (e.g. Kummer 1968; Smuts 1987; Zinner et al. 1994; Küster and Paul 1996; Hemelrijk and Luteijn 1998; Palombit et al. 2001). In this study it was occasionally observed that a higher-ranking female displaced another female who was grooming a male at that time and groomed the male herself afterward. Due to the small number of observations during focal animal sampling of such agonistic interactions, a quantitative analyses was not possible. Disturbing the grooming interaction might be used to prevent the lower-ranking female from establishing a relationship with a dominant male. A quantitative analysis of grooming competition especially during times of instability of adult and subadult female dominance relationships (Gust 1995) or introductions of new males would be of special interest to get further information on female competition over access to males. Summarizing, this study suggests researchers should consider a further role of males in current socioecological models of the evolution of primate social systems. It demonstrates that future primate studies on female competition for access to food should not be analyzed separately from male influences on females. This aspect is neglected in most studies on food competition in female primates. Acknowledgements We thank Debbie Gust and Tom Gordon for making it possible to study the mangabeys at the field station of the Yerkes Regional Primate Research Center. Many thanks are owed to both for their support and help. We are grateful to the caretakers of the field station for their valuable help and cooperation, and to Filippo Aureli, Joachim Burghardt, Joanna Fietz, Len Thomas, Dietmar Zinner, and three anonymous referees for their helpful comments on the manuscript. Last we wish to thank Prof. H-J Kuhn for his support. This study was supported by the Deutscher Akademischer Austauschdienst (512 402 575 3) and Deutsche Forschungsgemeinschaft (KU131/13-1). Animal maintenance was supported by a Yerkes NIH base grant. 215 References Altmann J (1974) Observational study of behavior: sampling methods. Behaviour 49:227–267 Barton RA (1993) Sociospatial mechanisms of feeding competition among female olive baboons, Papio anubis. Anim Behav 46:777–789 Barton RA, Whitten A (1993) Feeding competition among female olive baboons. Anim Behav 46:777–789 Barton RA, Byrne RW, Whitten A (1996) Ecology, feeding competition and social structure in baboons. Behav Ecol Sociobiol 38:321–329 Belzung C, Anderson JR (1986) Social rank and responses to feeding competition in rhesus monkeys. Behav Process 12:307– 316 Bergemüller R (1998) Nahrungsökologie der Rauchgrauen Mangabe. Ein Schlüsssel zu sozialen Organisation? Unpublished diploma thesis, Universität Erlangen Bernstein IS (1971) The influence of introductory techniques on the formation of captive mangabeys groups. Primates 12:33–44 Bernstein IS (1976) Activity patterns in a sooty mangabey group. Folia Primatol 26:185–206 Boccia ML, Laudensberger M, Reite A (1988) Food distribution, dominance, and aggressive behaviors in bonnet macaques. Am J Primatol 16:123–130 Boyd R, Silk JB (1997) How humans evolved. Norton, New York Brennan J, Anderson JR (1988) Varying responses to feeding competition in a group of rhesus monkeys (Macaca mulatta). Primates 29(3):353–366 Busse CD, Gordon TP (1983) Attacks on neonates by a male mangabey (Cercocebus atys). Am J Primatol 5:345–356 Busse CD, Gordon TP (1984) Infant carrying by adult male mangabeys (Cercocebus atys). Am J Primatol 6:133–142 Chapais B (1986) Why do adult male and female rhesus monkeys affiliate during the birth season? In: Rawlins MG, Kessler MJ (eds) The Cayo Santiago macaques. History, behavior and biology. State University of New York Press, Albany, pp 173– 200 Clutton-Brock TH (1977) Some aspects of intraspecific variation in feeding and ranging behaviour in primates. In: Clutton-Brock TH (ed) Primate ecology: studies of feeding ecology and ranging behaviour in lemurs, monkeys and apes. Academic Press, London, pp 539–556 Coelho AM Jr, Bramblett CA, Quick LS, Bramblett SS (1976) Resource availability and population density in primates: a socio-bioenergetic analysis of the energy budgets of Guatemalan howler and spider monkeys. Primates 17:63–80 Collins DA (1984) Spatial structure of chacma baboon groups. Int J Primatol 5:247–261 de Waal FBM (1986) The integration of dominance and social bonding in primates. Q Rev Biol 61:459–479 de Waal FBM (1989) Dominance ‘‘style’’ and primate social organization. In: Standen V, Foley RA (eds) Comparative socioecology. The behavioural ecology of human and other mammals. Blackwell Scientific, Oxford, pp 243–264 de Waal FBM, Luttrel LM (1989) Towards a comparative socioecology of the genus Macaca: different dominance styles in rhesus and stumptail macaques. Am J Primatol 19:83– 109 Ehardt CL (1988) Absence of strongly kin-preferential behavior by adult female sooty mangabey (Cercocebus atys). J Phys Anthropol 76:233–243 Fedigan LM (1993) Sex differences and intersexual relations in adult white-faced capuchins (Cebus capucinus). Int J Primatol 14(6):853–875 Gore MA (1993) Effects of food distribution on foraging competition in rhesus monkeys, Macaca mulatta, and hamadryas baboons, Papio hamadryas. Anim Behav 45:773–786 Gust DA (1994) Alpha-male sooty mangabeys differentiate between females’ fertile and their postconception maximal swellings. Int J Primatol 15(2):289–301 Gust DA (1995) Moving up the dominance hierarchy in young sooty mangabeys. Anim Behav 50:15–21 Gust DA, Gordon TP (1991) Male age and reproductive behavior in sooty mangabeys, Cercocebus torquatus atys. Anim Behav 41:277–283 Gust DA, Gordon TP (1993) Conflict resolution in sooty mangabeys. Anim Behav 46:685–694 Gust DA, Gordon TP (1994) The absence of a matrilineally based dominance system in sooty mangabeys, Cercocebus torquatus atys. Anim Behav 47:589–594 Harding RSO (1984) Primates of the Kilimi area, northwest Sierra Leone. Folia Primatol 42:96–114 Hemelrijk CK, Luteijn M (1998) Philopatry, male presence and grooming reciprocation among female primates: a comparative perspective. Behav Ecol Sociobiol 42:207–215 Hill DA, Okayasu N (1995) Absence of ‘‘youngest ascendancy’’ in the dominance relations of sisters in wild Japanese macaques (Macaca fuscata yajui). Behaviour 132:367–379 Hill DA, Okayasu N (1996) Determinants of dominance among female macaques: nepotism, demography and danger. In: Fa JE, Lindburgh DG (eds) Evolution and ecology of macaque societies. Cambridge University Press, Cambridge, pp 459–472 Hooff JARAM van (1988) Meerkatzenartige. In: Grzimek G (ed) Grzimek’s Enzyklopädie—Säugetiere, vol 2. Kindler, Munich, pp 208–285 Hooff JARAM van, Schaik CP van (1992) Cooperation in competition: the ecology of primate bonds. In: Harcourt AH, de Waal FBM (eds) Coalitions and alliances in humans and other animals. Oxford University Press, Oxford, pp 357–389 Janson CH (1984) Female choice and mating system of the brown capuchin monkey Cebus apella (Primates: Cebidae). Z Tierpsychol 65:172–200 Janson CH (1988) Food competition in wild capuchin monkeys (Cebus apella): quantitative effects of group size and tree productivity. Behaviour 105:53–76 Janson CH (1990) Social correlates of individual spatial choice in foraging groups of brown capuchin monkeys, Cebus apella. Anim Behav 40:910–921 Janson CH, Schaik CP van (1988) Recognizing the many faces of primate food competition: methods. Behaviour 105:165–186 Keller EF, Lloyd EA (eds) (1992) Keywords in evolutionary biology. Harvard University Press, Cambridge Kerscher R (1991) Sozial- und Nahrungsaufnahmeverhalten unter zwei Futterverteilungen untersucht an einer Brillenlangurengruppe (Presbytis obscura) des Zoologischen Gartens Wuppertal. Unpublished diploma thesis, Universität Erlangen, Nuremberg Kleiber M (1961) The fire of life: an introduction to animal energetics. Wiley, New York Kummer H (1968) Social organisation of hamadryas baboons. A field study. Chicago University Press, Chicago Küster J, Paul A (1996) Female–female competition and male mate choice in Barbary macaques (Macaca sylvanus). Behaviour 133(9–10):763–790 Lee PC, Bowman JE (1995) Influence of ecology and energetics on primate mothers and infants. In: Price CR, Martin RD, Skuse D (eds) Motherhood in human and nonhuman primates. Karger, Basel, pp 47–58 Matsumura S (2001) The myth of depotism and nepotism: dominance and kinship in matrilineal societies of macaques. In: Matsuzawa T (ed) Primate origins of human cognition and behavior. Springer, Tokyo Berlin Heidelberg, pp 441–462 Mori A (1977) Intra-troop spacing mechanism of wild Japanese monkeys of the Koshima troop. Primates 18(2):331–357 Müller EF (1985) Basal metabolism rate in primates: the possible role of phylogenetic and ecological factors. Comp Biochem Physiol A 81(4):707–711 Noordwijk MA van, Schaik CP van (1987) Competition among female long-tailed macaques, Macaca fascicularis. Anim Behav 35:577–589 Palombit RA, Cheney DL, Seyfarth RM (2001) Female–female competition for male ‘‘friends’’ in wild chacma baboons, Papio cynocephalus ursinus. Anim Behav 61:1159–1171 216 Perry S (1997) Male–female social relationships in wild white-faced capuchin monkeys, Cebus capucinus. Behaviour 134:477–510 Phillips KA (1995a) Foraging-related agonism in capuchin monkeys (Cebus capucinus). Folia Primatol 65:159–162 Phillips KA (1995b) Resource patch size and flexible foraging in white-faced capuchins (Cebus capucinus). Int J Primatol 16:509– 519 Pruetz JD, Isbell LA (2000) Correlations of food distribution and patch size with agonistic interactions in female vervets (Chlorocebus aethiops) and patas monkeys (Erythrocebus patas) living in simple habitats. Behav Ecol Sociobiol 49:38–47 Range F (1998) Sozialsystem und Konkurrenztyp weiblicher Rauchgrauer Mangaben (Cercocebus torquatus atys). Unpublished diploma thesis, Universität Bayreuth Range F, Noë R (2002) Familiarity and dominance relations among female sooty mangabeys in the Ta National Park. Am J Primatol 56:147–153 Robinson JG (1981) Spatial structure in foraging groups of wedgecapped capuchin monkeys, Cebus nigrivittatus. Anim Behav 29:1036–1056 Rose LM (1994) Benefits and costs of resident males to females in white-faced capuchins (Cebus capucinus). Am J Primatol 32:235–248 Ross C (1992) Basal metabolism rate, body weight and diet in primates: an evaluation of evidence. Folia Primatol 58(1):7–23 Saito C (1996) Dominance and feeding success in female Japanese macaques, Macaca fuscata: effects of food patch size and interpatch distance. Anim Behav 51:967–980 Schaik CP van (1989) The ecology of social relationships among female primates. In: Standen V, Fole RA (eds) Comparative socioecology, the behavioural ecology of human and other mammals. Blackwell, Oxford, pp 195–218 Schaik CP van (1996) Social evolution in primates: the role of ecological factors and male behaviour. Proc Br Acad 88:9–31 Schaik CP van, Dunbar RIM (1990) The evolution of monogamy in large primates: a new hypothesis and some crucial tests. Behaviour 115:30–62 Schaik CP van, Noordwijk MA van (1988) Scramble and contest among female long-tailed macaques in a Sumatran rain forest. Behaviour 105:77–98 Schwartz E (1921) The species of the genus Cercocebus. Ann Mag Nat Hist 10:664–670 Siegel S, Castellan NJ Jr (1988) Nonparametric statistics for the behavioral sciences, 2nd edn. McGraw-Hill, Singapore Smuts BB (1985) Sex and friendship in baboons. Aldine, New York Smuts BB (1987) Sexual competition and mate choice. In: Smuts BB, Cheney DL, Seyfarth RM, Wrangham RW, Struhsaker TT (eds) Primate societies. Chicago University Press, Chicago, pp 385–399 Smuts BB, Smuts RW (1993) Male aggression and sexual coercion of females in nonhuman primates and other mammals: evidence and theoretical implications. Adv Study Behav 22:1–63 Sokal RR, Rohlf FJ (1995) Biometry, 5th edn. Freeman, San Francisco Stahl D (1996) Social tolerance in captive female mangabeys (Cercocebus torquatus atys). Primate Rep 44:45–46 Stahl D (1998) Food competition in captive female mangabeys (Cercocebus torquatus atys). (PhD thesis, Universität Tübingen) Cuvillier, Göttingen Stahl D, Kaumanns W (1999) Female dominance hierarchies in captive sooty mangabeys (Cercocebus torquatus atys). Primate Rep 55:39–52 Sterck EHM (1995) Female, food, and fights. PhD thesis, University of Utrecht Sterck EHM, Steenbeck R (1997) Female dominance relationships and food competition in the sympatric Thomas langur and long-tailed macaque. Behaviour 134:749–774 Sterck EHM, Watts DP, Schaik CP van (1997) The evolution of female social relationships in nonhuman primates. Behav Ecol Sociobiol 41:291–309 Struhsaker TT (1971) Notes on Cercocebus atys atys in Senegal, West Africa. Mammalia 35:343–344 Takahata Y (1982) Social relations between adult males and females of Japanese monkeys in the Arashiyama B troop. Primates 23:1–23 Trivers RL (1972) Parental investment and sexual selection. In: Campbell B (ed) Sexual selection and the descent of man. Aldine, Chicago, pp 136–179 Watts DP (1988) Environmental influence on mountain gorilla time budgets. Am J Primatol 15(3):195–211 Watts DP (1990) Ecology of gorillas and its relation to female transfer in mountain gorillas. Int J Primatol 11:21–45 Whitten PL (1983) Diet and dominance among female vervet monkeys (Cercocepithecus aethiops). Am J Primatol 5: 139–159 Wrangham RW (1979) On the evolution of ape social systems. Social Sci Inf 18:334–368 Wrangham RW (1980) An ecological model of female-bonded primate groups. Behaviour 75:262–300 Zinner D (1993) Nahrungskonkurrenz bei Mantelpavianen. Eine experimentelle Studie. (PhD thesis, Universität Göttingen) Shaker, Aachen Zinner D, Schwibbe MH, Kaumanns W (1994) Cycle synchrony and probability of conceptions in female hamadryas baboons (Papio hamadryas). Behav Ecol Sociobiol 135(3):175–183 Zinner D, Hindahl J, Schwibbe M (1997) Effects of temporal sampling patterns of all-occurrence recording in behavioural studies. Ethology 103(3):236–246