Chimpanzee (Pan troglodytes verus) Behavioral Responses to Stresses Associated With Living in a Savanna-Mosaic Environment: Implications for Hominin Adaptations to Open Habitats

Anthropologists have long been interested in the behavioral ecology of nonhuman primates living in savannas given what we know of early hominin environments. As expected, chimpanzees in the Fongoli community in southeastern Sénégal show a unique suite of behavioral adaptations to stresses associated with their savanna habitat. While Fongoli chimpanzees are species-typical in certain regards, such as including ripe fruit in the diet during all months of the year, they also adjust their behavior to the particular stresses of this dry, hot and open environment. These behaviors include using caves as shelters during the dry season, soaking in pools of water during the hot, early rainy season, and traveling and foraging at night during maximum phases of the moon. Adult males of this 35-member community serve as focal subjects in a long-term study of the ecology and behavior of chimpanzees in a savanna-mosaic environment. Here, we report on Fongoli chimpanzee activity budgets, grouping behavior, and habitat use during the dry versus wet season based on over 2500 hours of observation from March 2005–July 2006. Findings support the hypothesis that ecological pressures associated with a savanna environment significantly affect great ape behavior. The Fongoli chimpanzees’ large home range (>65km2) is sometimes used cyclically, with the community traveling as one large party, in contrast to the typical chimpanzee fission-fusion pattern. Combined with data on temperature in the various habitats within the savanna mosaic, results show that Fongoli chimpanzees minimize energy expenditure during the hottest months and at the hottest time of day by resting more and traveling less, in addition to selectively using small patches of closed-canopy habitats, such as gallery forest. They move significantly more during early hours of the hot, dry season specifically and range in smaller parties at this time compared to during the wet season. The stresses associated with a savanna-mosaic environment and chimpanzees’ behavioral adjustments to them have important implications for understanding early hominin behavior in similar environments. The “Energetic Studies in Hominin Evolution” Symposium, Paleoanthropology Meetings, Philadelphia, PA, 27— 28 March, 2007; symposium papers guest edited by Karen Steudel-Numbers (University of Wisconsin) and Cara Wall-Scheffler (Seattle Pacific University). INTRODUCTION E hypotheses have long been used to explain significant changes in hominin evolution (Dart 1925; Darwin 1871; Jolly 1970; Robinson 1954). In particular, the ‘savanna hypothesis’ has been used to explain such distinctive hominin features as bipedalism, tool use, and increased brain size (see Potts 1998 for review). Given their close kinship with humans, as well as the unique habitat in which they live, chimpanzees living in savanna-woodland environments have long been of interest to anthropologists (Suzuki 1969; Kano 1972; McGrew et al. 1981; Moore 1992; Sept 2002). Although the savanna hypothesis has been largely abandoned in its original form, a move from a closed to a more open environment over the course of hominin evolution remains a significant change that likely acted as an important selective pressure regarding the behavioral ecology of early hominins. With more detailed information on paleoenvironments associated with possible basal hominins such as Sahelanthropus (Brunet et al. 2002; Vignaud et al. 2002) and the Australopithecines in general (White et al. 2006), for example, anthropologists recognize the complexity of the environment inhabited by early hominins and the mosaic nature of the habitats specifically. Anthropologists have shifted from a tendency to associate major anatomical and behavioral changes in early hominins with a grassland savanna environment to the recognition that riverine and open and closed woodlands are perhaps more characteristic of these hominins’ habitats (Potts 1998). This has resulted in a revised form of the savanna hypothesis, which focuses on the woodland-forest environment PaleoAnthropology 2009: 252−262 . © 2009 PaleoAnthropology Society. All rights reserved. ISSN 1545-0031 doi:10.4207/PA.2009.ART33 JIll D. PRUeTz Department of Anthropology, 324 Curtiss, Iowa State University, Ames, IA 50011, USA; [email protected] PACo BeRTolANI Leverhulme Centre for Human Evolutionary Studies, The Henry Wellcome Building, Cambridge University, Fitzwilliam Street, Cambridge, CB2 1QH, UNITED KINGDOM; [email protected] Savanna Chimpanzees and Hominin Behavior • 253 as the critical feature influencing hominin evolution (Potts 1998). This has important implications for reconstructing early hominin socioecology and makes the consideration of chimpanzees living in woodland-savanna mosaics pertinent to related questions. A savanna-woodland habitat is thought to be associated with selective pressures operating on early hominin evolution, and hypotheses proposed to explain key adaptations in hominins look to environmental factors. However, Potts (1998) examined such habitat-specific hypotheses in light of the environmental records at key hominin localities and concluded that they do not adequately explain the key adaptive changes in the hominin lineage. Nonetheless, inferences continue to be made regarding the influence of open environments on early hominin behavior (Reed 1997; Spencer 1997; Vrba 1985; and, see Foley 1999) largely because ecology remains a key component to be considered in hominin evolution. The issue is far from resolved. Given the difficulties in understanding the behavior of extinct hominins, extant apes of the Family Hominidae provide an opportunity to assess the effects of ecological variables on the behavior of our closest living relatives in similar habitats. However, little evidence exists regarding these habitats’ effects on living hominins, chiefly because apes in this environment are notoriously difficult to habituate to researcher presence (McGrew et al. 2003). Although extensive data are available regarding the ways in which monkeys, such as baboons (Papionini), adapt and adjust to open environments (Dunbar 1998; Hill and Dunbar 2002; Kamilar 2006), the Fongoli site in southeastern Sénégal provides the first data on a habituated community of chimpanzees living in such an environment. Chimpanzees in the Fongoli community in southeastern Sénégal show a unique suite of behavioral adaptations to stresses associated with their savanna environment. While chimpanzees at Fongoli are species-typical in certain regards, such as including ripe fruit in the diet during all months of the year (Pruetz 2006), they also adjust their behavior to the particular stresses of this dry, hot and open environment. Fongoli chimpanzees exhibit unique behaviors such as using caves as shelters during the dry season (Pruetz 2007), soaking in pools of water during the early rainy season, and moving and foraging at night during maximum phases of the moon in the dry season. They also exploit protein sources, such as prosimian primates, that chimpanzees at more forested sites ignore, and they use tools to hunt these prosimians (Pruetz and Bertolani 2007). Detailed behavioral data on our closest living relative that allows us to understand the responses of apes in a savanna mosaic can add to the comprehensive body of data used to make hypotheses about the behavior of early hominins in open, hot and dry habitats. Comparing chimpanzees in such a habitat to those living in more forested areas gives us insight into how apes respond to the selective pressures associated with such an environment (Moore 1996) and which variables should be considered most heavily in scenarios of early hominin evolution. METHODS STUDY SITE Chimpanzees in Sénégal inhabit the area known as the Mandingue Plateau, which also constitutes part of southwestern Mali and northeastern Guinea (Carter et al. 2003). This area defines the northern and geographical limits of chimpanzees’ range in West Africa. Due to drought during the 1970s and 1980s in the Sahel region of Africa, the intertropical convergence zone has moved southward to encompass part of Sénégal (Giannini et al. 2003; zeng 2003), perhaps increasing the stresses acting on chimpanzees in this habitat. In southeastern Sénégal, savanna chimpanzees coexist with humans belonging to the Bédik, Malinké, Bassari, Diahanke, and Puhlar groups. Humans here compete with chimpanzees over wild plant foods, such as the fruit of Saba senegalensis (Pruetz and Knutsen 2003). Humans do not hunt chimpanzees for food, although chimpanzees are periodically targeted for the capture of infants that can be sold as pets (Pruetz and Kante in prep.). This region of southeastern Sénégal is Sudanian savanna and Guinean woodland, which can be envisioned as a savanna-woodland mosaic. Rainfall in this area averages 900mm annually (Ba et al. 1997). In southeastern Sénégal, the rainy season is from June through September. May and october may be considered transitional months, although virtually no rain falls in May in some years. November through April constitute the dry season, when rainfall does not occur. At this time, almost all trees lose their leaves, and natural and human-induced bush fires sweep through the area. other animal species at the site include those associated with open environments, such as patas monkeys (Erythrocebus patas) and oribi (Oribi oribi), as well as species that typically use closed canopy habitats more extensively, such as green monkeys (Cercopithecus aethiops sabeus) and bushbuck (Tragelaphus scriptus). The Fongoli site (12o40’ N, 12o13’ W) is at the junction of the Sudanian and Sudano-Guinean vegetation belts (Figure 1). The chimpanzee home range is estimated, minimally, to be 65km2, which is from two to six times as large as ranges of chimpanzee communities living in more forested areas (Basabose 2005; Chapman and Wrangham 1993; Herbinger et al. 2001; Newton-Fisher 2003). The home range is predominantly open woodland and grassland, with small patches of gallery forest and seasonally cultivated fields (Pruetz 2006; Figure 2). Trees average around 10m in height in woodland habitat, while grassland includes few scattered trees (Figure 3; Pruetz et al. 2008). Small patches of gallery forest contain taller trees, and the canopy is often contiguous. Woodland, grassland, and bamboo woodland (Figure 4) are characterized by a predominantly grass understory, while gallery forest understory is characterized by a diversity of herbaceous-level plants. Fields may be completely cleared or characterized by some degree of regrowth of vegetation (Figure 5). estimations of habitat types that constitute the Fongoli chimpanzees’ home range were derived from transects oriented N-S (3km) and e-W (3km), which were placed at the approximate center of the Fongoli study site, plus one additional randomly-placed transect 1km 254 • PaleoAnthropology 2009 long. Each 100m along transects, a circular plot of a 5m radius was sampled for specific vegetative characteristics and classified as gallery forest, woodland (including bamboo woodland), field, or grassland. Temperature data were collected using remote data loggers (Pruetz 2007). From 2001 to 2004, HoboTM temperature loggers were placed in woodland, grassland, and gallery forest habitats so as to be fully shaded and were set to record ambient temperatures each 6–10 minutes (Pruetz 2007). Although battery failure resulted in data loss for some habitats in some years, temperatures were collected over at least one annual cycle for each of the different habitats (Pruetz 2007). Data were analyzed with BoxcarTM software, excluding nighttime temperatures (1801 to 0559 hours) in statistical analyses (Pruetz 2007). Temperature differences were analyzed with t-tests, using mean monthly temperatures in the different habitats as dependent variables (Pruetz 2007). Data on rainfall were collected using a rain gauge located at Fongoli village, within the chimpanzees’ home range. Rainfall is presented as monthly averages based on data collected over the course of 3.5 years at Fongoli, from June 2005 through December 2008. STUDY SUBJECTS The Fongoli community contained 35 individuals during the study, which included 10 adult males, seven adult females and 18 immature chimpanzees. Nine of ten adult males served as focal subjects as part of a long-term study of the ecology and behavior of chimpanzees in a savannamosaic environment. one aged adult male was excluded due to his poor sight and hearing and resultant timidity around observers. Both focal animal and scan sampling with interval recording methods were used to collect data on activities such as resting, feeding, traveling, grooming, and other social behaviors as well as habitat use. Data on activity were collected during more than 2500 contact hours with chimpanzees from March 2005 through March 2006. Mean values for individual males were analyzed according to dry versus wet season and by hour so as to avoid problems with pooling of data and to control for seasonal effects and time of day, respectively. Additionally, we report on chimpanzee grouping behavior during dry versus wet seasons based on over 3000 hours of data collected from March 2005–July 2006. Party size is defined as all individuals encountered in a single day, following Boesch (1996). Figure 1. Map of Senegal and Fongoli study site (figure modified and used with permission from Projet Niokolo Badiar FED Number 4213/RED, DPNS, DNFF, ORTOM). Savanna Chimpanzees and Hominin Behavior • 255