(Pan Troglodytes Verus) and Resource Scarcity in a Savanna Environment at Fongoli, Senegal Stacy M

Home , Jenny (orangutan), Western chimpanzee

Iowa State University Capstones, Theses and Graduate Theses and Dissertations Dissertations

2014 Multilevel analysis of the foraging decisions of western chimpanzees (Pan troglodytes verus) and resource scarcity in a savanna environment at Fongoli, Senegal Stacy M. Lindshield Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd Part of the Behavior and Ethology Commons, and the Biology Commons

Recommended Citation Lindshield, Stacy M., "Multilevel analysis of the foraging decisions of western chimpanzees (Pan troglodytes verus) and resource scarcity in a savanna environment at Fongoli, Senegal" (2014). Graduate Theses and Dissertations. 14235. https://lib.dr.iastate.edu/etd/14235

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. i

Multilevel analysis of the foraging decisions of western chimpanzees (Pan troglodytes verus) and resource scarcity in a savanna environment at Fongoli, Senegal

Stacy M. Lindshield

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Major: Ecology and Evolutionary Biology

Program of Study Committee: Jill Pruetz, Major Professor Brent Danielson David Vleck Stan Harpole Dean Adams

Iowa State University

Ames, Iowa

2014

TABLE OF CONTENTS

Page

LIST OF FIGURES ...... v

LIST OF TABLES ...... vi

ACKNOWLEDGEMENTS ...... vii

ABSTRACT ...... viii

CHAPTER I INTRODUCTION: CHIMPANZEE FORAGING BEHAVIOR...... 1

Synopsis of field research on chimpanzees ...... 2 A multi-level research approach to savanna chimpanzee foraging behavior ...... 4 Level one: Diet and food selection ...... 6 Level two: Habitat quality and selection ...... 14 Level three: Pan biogeography...... 19 Dissertation structure...... 23 Literature cited ...... 23

CHAPTER II SAVANNA CHIMPANZEES: A BRIEF REVIEW ...... 42

Defining savannas ...... 42 Savanna chimpanzees across Africa ...... 45 Savanna chimpanzees in Senegal ...... 50 Food selection in savanna environments ...... 51 Habitat selection in savanna environments ...... 52 Summary ...... 55 Literature cited ...... 56

CHAPTER III NUTRITIONAL AND SOCIAL COMPONENTS OF FORAGING STRATEGY IN WESTERN CHIMPANZEES (PAN TROGLODYTES VERUS) AT FONGOLI, SENEGAL ...... 70

Abstract ...... 71 Introduction ...... 72 Methods ...... 75 Study location and period ...... 75 Study subjects and behavior sampling...... 75 Food sampling ...... 77 Food availability ...... 78 Nutritional chemistry ...... 78 Nutrient intake ...... 79 iii

Statistical procedures ...... 80 Research ethics ...... 81 Results ...... 81 Nutritional composition and availability of foods ...... 81 Relationships between food profitability and food choice ...... 83 Effects of estrus females, social rank, and party size on nutrient intake ...... 84 Discussion ...... 85 Significance of time, energy, and macronutrients on food selection...... 85 Male strategies to balance nutritional and social gains ...... 86 Savanna chimpanzee nutritional ecology ...... 87 Acknowledgements ...... 89 Literature cited ...... 89

CHAPTER IV FEEDING IN FEAR: HABITAT SELECTION AND PREDATOR- SENSITIVE FORAGING IN WESTERN CHIMPANZEES (PAN TROGLODYTES VERUS) ...... 107

Abstract ...... 108 Introduction ...... 109 Methods ...... 114 Study area and study subjects ...... 114 Food patches ...... 117 Behavioral follows ...... 120 Food sample collection and chemical analyses ...... 122 Statistical analyses ...... 124 Ethical standards ...... 124 Results ...... 125 Characteristics of real and potential food patches ...... 125 Baobab fruit counts ...... 126 Nutritional composition of baobab fruit ...... 127 Foraging party size and composition ...... 128 Male foragers’ associations with canopy cover and surface water ...... 128 Male foragers’ associations with human activity ...... 129 Discussion ...... 130 Antipredator foraging strategy...... 130 Foraging in a savanna mosaic ...... 132 Significance to savanna chimpanzee ecology and conservation ...... 133 Acknowledgements ...... 134 Literature cited ...... 134

CHAPTER V ON THE MARGINS OF SAVANNA CHIMPANZEE (PAN TROGLODYTES VERUS) HABITAT IN SENEGAL ...... 150

Abstract ...... 151 Introduction ...... 152 Materials and methods ...... 155 iv

Study sites ...... 155 Landscape cover and physiognomy ...... 156 Results ...... 157 Defining the upper latitudinal limit of Pan troglodytes in present-day Senegal ...... 157 Describing landscapes within and outside the Pan troglodytes geographical range ..... 157 Discussion ...... 158 Acknowledgements ...... 161 Literature cited ...... 162

CHAPTER VI CONCLUSION ...... 174

Take home messages ...... 174 Level one: Diet and food selection ...... 174 Level two: Habitat quality and selection ...... 176 Level three: Savanna chimpanzee biogeography ...... 179 New directions in savanna chimpanzee foraging ecology ...... 180 Nutritional ecology ...... 180 Predation-sensitive foraging ...... 181 Savanna chimpanzee biogeography ...... 181 Anthropological and practical applications ...... 182 The thermoregulatory hypothesis of bipedalism ...... 182 Crop-raiding management ...... 182 Literature cited ...... 184

APPENDIX SUPPLEMENTARY MATERIALS FOR CHAPTER IV ...... 188

LIST OF FIGURES

Page

Figure 1.1. Geographical distribution map for Pan troglodytes ...... 41

Figure 2.1. Mean annual recipitation and temperature at chimpanzee sites ...... 67

Figure 2.2. Precipitation and temperature seasonality at chimpanzee sites ...... 68

Figure 2.3. Tree cover at chimpanzee sites...... 69

Figure 3.1. Qualitative food availability scores by food type ...... 103

Figure 3.2. Daily metabolizable energy intake ...... 104

Figure 3.3 Food profitability by food part ...... 105

Figure 3.4. Food profitability by frequency of consumption ...... 106

Figure 4.1. Photographs of habitat types during the early dry season...... 145

Figure 4.2. Habitat characteristics of food patches ...... 146

Figure 4.3. Baobab fruit patch depletion across time ...... 147

Figure 4.4. Nutritional values of unripe and ripe baobab fruit ...... 148

Figure 4.5. Food intake with antipredator behavior ...... 149

Figure 5.1. Map of chimpanzee range in Senegal and study sites ...... 170

Figure 5.2. Flow chart of image processing procedure ...... 171

Figure 5.3. Flow chart of image processing procedure ...... 172

Figure 5.4. Woody vegetation cover and physiognomy among study sites...... 173

Figure 6.1. Conceptual summary of the three levels of analysis in this study ...... 187

LIST OF TABLES

Page Table 1.1 Taxonomic classification for Pan troglodytes ...... 40

Table 2.1. Climate and vegetation characteristics at long-term chimpanzee sites ...... 65

Table 2.2. Summary of publications on savanna chimpanzees ...... 66

Table 3.1. Adult male chimpanzee study subjects at Fongoli...... 96

Table 3.2. Nutritional values of foods in this study ...... 97

Table 3.3. Average nutritional values of ripe fruit mesocarp among Pan sites ...... 101

Table 3.4. Effects of party size and composition on daily nutrient intake ...... 102

Table 4.1. Effects of environmental and social variables on feeding ...... 143

Table 4.2. Nutritional characteristics of baobab fruit and variation among trees ...... 144

Table 5.1. Summary of chimpanzee biogeography studies...... 167

Table 5.2. Details of satellite imagery from this study ...... 168

Table 5.3. Summary of landscape metrics used in this study ...... 169

Table 5.4. Two-way ANOVA of landscape differences among sites ...... 169

vii

ACKNOWLEDGEMENTS I am deeply grateful to my advisers, mentors, collaborators, assistants, funders, study subjects, family, and friends for their important contributions to this project. I thank Jill

Pruetz for giving me the opportunity to conduct field research at El Zota Biological Field

Station in Costa Rica and at Fongoli in Senegal. Jill has been a wonderful mentor, role model, and supporter. I also appreciate the intellectual and academic advice of Brent

Danielson throughout my master’s and doctoral programs. It was also a pleasure to have

Dave Vleck, Stan Harpole, and Dean Adams on my dissertation program committee and I thank them for their contributions to this project. Thanks to Jessica Rothman for inviting me to her nutrition lab at CUNY-Hunter College to run the nutritional analyses, and to Jenny

Paltan for guidance in the lab.

Numerous other people have made important intellectual, technical, or logistical contributions to this project, including: Landing Badji, Kelly Boyer-Ontl, Maja Gašperšič,

Assane Goudiaby, Charles Hurburgh, Martha Jeffrey, Paul Lasley, Debra McKay, Papa

Ibnou Ndiaye, Sylvia Ortmann, Erik Otarola-Castillo, Gray Tappan, and Erin Wessling.

Also, my sincere thanks go to Ulises Villalobos-Flores, Michel Sahdiako, Dondo Kante and

Nene, Anna Olson, El-Hadji Sakho, Waly Camara, Mody Camara and Sira, and Mboule

Camara and Sira. A warm thanks goes to Morty, Beverly, Josh, and Emanuel Marshack for their kindness and generosity. Finally, I am thankful for my family and friends for their support and encouragement, especially my parents, Lori McKinney, David Lindshield, and

Daniel McKinney.

viii

ABSTRACT Savannas are the hottest, driest, and most open environments occupied by wild chimpanzees (Pan troglodytes). Chimpanzee subsistence strategies are poorly understood in these habitats and, thus, current knowledge primarily resides within the theoretical domain.

To address this gap, we empirically test food and habitat selection hypotheses by examining the foraging decisions of chimpanzees in a savanna mosaic environment at Fongoli, Senegal.

The foraging behavior of Fongoli chimpanzees was examined in relation to the macronutrient composition of their foods. As predicted under an energy maximizing strategy, individuals often selected foods that were energy-rich and easy to consume. However, this strategy was a poor predictor for some important foods. Fongoli chimpanzees may select lower quality foods at times to minimize risk of heat stress. At the level of habitat selection, this study asked how the foraging behavior of adult male Fongoli chimpanzees changed with predation risk. We tested for this sensitivity by measuring food intake among relatively risky and safe habitats. Elevated risk of predation did not fully deter adult males from feeding in these habitats, but during such visits they ingested more ripe fruit, an energy-rich food. The third level of analysis addresses landscapes-scale habitat selection processes. Our analysis merges findings on Fongoli chimpanzee food and habitat selection with information on the species’ distribution and remotely-sensed land cover to evaluate relationships between landscapes and the species’ range. We show that accessibility to drinking water sources, anthropogenic habitat disturbance, and habitat physiognomy are associated with the species’ distribution in southeastern Senegal. This study highlights the importance of concurrently examining chimpanzee foraging behavior at several levels to understand interconnected factors that shape their subsistence strategies. 1

CHAPTER I INTRODUCTION: CHIMPANZEE FORAGING BEHAVIOR A population of about 500 chimpanzees (Pan troglodytes verus) occupy the savanna- woodlands of the West African country of Senegal (Carter et al. 2003; Ndiaye 2011). The study of these apes’ foraging behavior is critical to improving efforts to conserve this population threatened with extinction, given that foraging is intrinsically linked to individuals’ survival and reproduction and that such behavior in savanna-dwelling chimpanzees is poorly understood. Moreover, savanna chimpanzee foraging behavior is significant to the field of evolutionary anthropology. Study of our closest living relatives, members of the genus Pan (chimpanzees and bonobos, Pan paniscus), can inform interpretations of foraging strategies in extinct members of the subfamily Homininae, particularly regarding early, Pan-like hominins associated with mosaic savanna environments, such as Ardipithecus ramidus (reviewed in Domingues-Rodriguo 2014) and the last common ancestor (LCA) between the human and chimpanzee/bonobo lineage

(McGrew 2010a). Aside from conservation and evolutionary approaches, the study of savanna chimpanzee foraging behavior stands alone as a worthwhile academic pursuit

(Goodall 1986), given that new research routinely illuminates the behavioral diversity that characterizes this species (Boesch et al. 2002).

In this vein, I investigate the food quest in an omnivorous and social African ape, the chimpanzee, on the savannas in southeastern Senegal, with the goal of advancing our understanding of these apes’ nutritional ecology and habitat selection in open-country environments. The main research objectives of this study include:

1. Identifying the foraging strategies of adult male chimpanzees at Fongoli, Senegal, at

the intersection of their food selection behaviors and the nutritional values of their

foods (Chapter 3).

2. Determining how variation in savanna habitat quality at Fongoli impacts the foraging

decisions of chimpanzees (Chapter 4).

3. Combining the above findings on food and habitat selection with land cover

information to evaluate determinants of chimpanzee biogeography in southeastern

Senegal (Chapter 5).

To contextualize these goals, this chapter provides a brief synopsis of research on wild chimpanzees as well as essential background information on food selection, habitat selection and biogeography of chimpanzees in Senegal.

Synopsis of field research on chimpanzees Species descriptions for Pan troglodytes first arose in the late 17th century (reviewed in Marks 2002 and Schwarz 1934) and, until recently, three subspecies were widely recognized based on morphological traits and geographic distributions (Hill 1969). Table 1.1 provides a general taxonomy for chimpanzees, following Groves (2001). Presently, between four and six subspecies are recognized based on phylogenetic approaches (Gonder et al.

2006, 2011; but see Fischer et al. 2006). Western chimpanzees range from Ghana to Senegal and are the most genetically differentiated of all the subspecies (Gonder et al. 2011). The

Dahomey gap, a vast savanna corridor stretching across Benin, Togo and Ghana interrupts a belt of equatorial rain forest and separates western chimpanzees from all other subspecies; however, the gap was not a geographic barrier until a few thousand years ago (Salzmann and

Hoelzmann 2005), well after P. t. verus was genetically differentiated (Fischer et al. 2006,

Gonder et al. 2011). The Nigerian subspecies (P. t. ellioti), identified after the introduction of 3 phylogenetic analyses based on genetics (?), ranges in Nigeria and Cameroon (Gonder et al.

1997). Populations in central and eastern Africa are genetically less differentiated, leading to several phylogenetic classifications (reviewed by Fischer et al. 2006 and Gonder et al. 2011).

The designation of two subspecies here, central (P. t. troglodytes) and eastern (P. t. schweinfurthii) chimpanzees, is widely recognized among primatologists (Groves 2001).

Central chimpanzees range from Cameroon to Congo and the Central African Republic, and the eastern subspecies can be found from the Democratic Republic of Congo to as far east as

Uganda and Tanzania. Figure 1.1 illustrates the geographical distribution of these four subspecies.

Chimpanzees are most closely related to bonobos, followed by humans (Homo sapiens), gorillas (Gorilla gorilla), and orangutans (Pongo pygmaeus), respectively (Ruvolo et al. 1994). Chimpanzees have been studied in captivity since the 1920s and were commonly used in laboratory settings for biomedical, behavioral and aerospace research due to their morphological and behavioral similarities to Homo sapiens (Altevogt et al. 2011). Although scientists have long recognized that chimpanzees and humans shared a relatively recent last common ancestor (reviewed in Marks 2002), molecular studies have verified this high degree of genetic relatedness (Ruvolo et al. 1994, TCSAC 2005).

Our closely shared evolutionary history with chimpanzees motivates many scientists to study their behavior, biology and cognition (Goddall 1986, Hoppius 1760, Yerkes 1925).

Beginning in the 1920’s, the psychologist Robert Yerkes began to disseminate his studies of captive chimpanzee behavior and later established the first breeding facility in the United

States (reviewed in Nystrom and Ashmore 2008). In 1930, Yerkes sent Henry Nissen to

Guinea to conduct a short study on wild chimpanzees (reviewed in Goodall 1994). Our 4 understanding of wild chimpanzee behavior tremendously improved after Jane Goodall and the late Toshisada Nishida (1941-2011) began their concurrent and independent field studies at Gombe and Mahale along the shore of Lake Tanganikya in Tanzania in the early 1960s

(Goodall 1986, Nishida 2012). These researchers successfully habituated chimpanzees via provisioning, meaning that individuals over time grew accustomed to the presence of researchers and generally tolerated and ignored them. The habituation process enabled

Goodall and Nishida to follow individuals at close range and describe their daily lives in detail. This approach led to groundbreaking discoveries, such that chimpanzees display a fission-fusion social system (Nishida 1968), use tools (van Lawick-Goodall 1970), hunt vertebrate prey and share meat (van Lawick-Goodall 1968), and are capable of intraspecific lethal aggression (Mason and Wrangham 1991). Research at Gombe and Mahale continue to this day, and were followed by a suite of additional long-term research sites that currently investigate the behavior of habituated chimpanzees across Africa (Figure 1.1). Largely due to the morphological and behavioral similarities between humans and chimpanzees, many of these long-term studies have focused social behavior and tool use (Boesch and Boesch-

Achermann 2000, Goodall 1986, Matsuzawa et al. 2011; McGrew 2001, Nishida 2012).

Meanwhile, studies on chimpanzee ecology have lagged behind (Gilby and Wrangham 2007;

Gilby et al. 2013; Emery-Thompson et al. 2007; Emery-Thompson and Wrangham 2008;

McGrew et al. 1981; Wrangham 1975).

A multi-level research approach to savanna chimpanzee foraging behavior The quest for food profoundly impacts the daily routines of chimpanzees, as individuals engage in feeding and traveling for approximately one-half of daylight hours

(Lehmann et al. 2007). The mosaic of savanna habitats imposes foraging challenges: for instance, Fongoli chimpanzees spend a disproportionate amount of time in forest habitats but 5 often forage in open habitats, especially woodland (Pruetz and Bertolani 2009), where feeding trees are scattered (Bogart and Pruetz 2011) and ambient temperature is relatively hot

(Pruetz 2007). It is not clear how the nutritional quality of foods, and food availability (i.e., abundance and distribution, sensu Pruetz 2009), on spatially and temporally complex savanna landscapes impact the daily foraging decisions of chimpanzees. Does caloric yield per time explain the majority of food choices at Fongoli, as predicted by energy rate maximization (Stephens and Krebs 1986)? Are savanna foods lower in energy or higher in antifeedants when compared to the diets of chimpanzees that occupy more forested environments? These are some of the questions I address in the nutritional ecology section of this dissertation (Chapter 3). Although the nutritional values of foods, and the time needed to process them, is expected to greatly influence foraging decisions, a suite of other forces impact foraging choices, namely the availability of other key resources such as surface water and shade (Chapter 4), mating behavior as well as dominance hierarchy formation and maintenance (Chapters 3), and the presence of predators or competitors that incur risk of injury or death to chimpanzees (Chapter 4). Collectively, these forces are expected to determine habitat suitability for savanna chimpanzees that occupy the margin of the species' range in Senegal (Chapter 5). For these reasons, I use a multi-level approach to study savanna chimpanzees, at the levels of: (1) diet quality and food selection at Fongoli, (2) habitat quality and selection at Fongoli, and (3) the species' geographical distribution in southeastern Senegal in relation to landscape composition and physiognomy. In the following sections, I provide a primer on diet and food selection, habitat quality and selection, and biogeography of chimpanzees as background information for chapters three, four and five, respectively. 6

Level one: Diet and food selection

Chimpanzees are omnivorous in that they consume plant and animal tissues

(reviewed in McGrew 2010a). Chimpanzees primarily consume fleshy fruits (e.g., mesocarp when they are available (Watts et al. 2012a). Other than fruit, chimpanzees most often consume seeds, leaves, flowers, pith/stems, bark/cambium, insects, small mammals, honey, soil, and resin (Bogart and Pruetz 2011: Table 9, Goodall 1986, Watts et al. 2012a: Table 2).

The best known aspects of this diverse diet involve the significance of ripe fruit, animal prey, and lean season foods in shaping the ecology and evolution of chimpanzees.

Ripe fruit specialists. Historically, primatologists have labeled chimpanzees as frugivores or ripe fruit specialists (Wrangham et al. 1998) to indicate that fruit predominates their omnivorous diet. This pattern has been verified at every long-term study site with habituated chimpanzees (reviewed in Watts et al. 2012a). Ripe fruit is a high quality food: it is usually easy to pluck fruits from the branch of a woody plant for immediate consumption, and fruit is rich in sugars that the body can rapidly convert into energy (e.g., fructose, glucose, sucrose).

Furthermore, wild fruits are often high in fiber, which aids in the passage of digesta through the gastro-intestinal (GI) tract (Milton and Demment 1988) and provides a source of energy via hindgut fermentation (Conklin and Wrangham 1994; Lambert and Fellner 2011). In addition to carbohydrates, wild fruit often contains sufficient amounts of protein to meet daily nutritional requirements (Conklin-Brittain et al. 1998), provides preformed water to the consumer (source), and is sometimes rich in lipids (Hohmann et al. 2010). Furthermore, fruit provides essential micronutrients, such as ascorbic acid (source). Although fruit is probably the most important food for chimpanzees, an exclusively frugivorous diet may lead to nutrient deficiencies (Tennie et al. 2009). Supplementing the diet with alternative foods, like 7 leafy material and animal prey, provides essential micronutrients that can be rare in fruits, such as essential amino acids and vitamin B12 (Oftedal 1991; Tennie et al. 2009). Moreover, fruit is a patchily distributed food in time and space (Wrangham 1980), thus, most chimpanzee communities encounter fruit scarcity on a seasonal basis (Taï: N'guessan et al.

2009; Fongoli: Pruetz 2006; Kanyawara, Kibale: Wrangham et al. 1991; but see Budongo:

Newton-Fisher 1999; Ngogo, Kibale: Watts et al. 2012a). During these periods, chimpanzees tend to expand dietary breadth to include lower quality foods, such as pith and bark (Isabirye-

Basuta 1989; Wrangham et al. 1991).

Fallback foods. Plant tissues other than fruit form a crucial component of the diet, especially when ripe fruit is rare. Foods with relatively low energy return rates tend to be lower in calories, higher in indigestible fiber or plant secondary compounds, take longer to ingest, or involve some combination of these features (Stephens et al. 2007). Among chimpanzees, these foods tend to include stems, pith, and bark; chimpanzees are said to "fallback" on these lower quality foods when ripe fruit is relatively scarce (Wrangham et al. 1991). This association between fruit availability, and dietary breadth expansion and contraction

(Isabirye-Basuta 1989), fits expectations of optimal food selection in a patchy environment

(MacArthur and Pianka 1966) and the "expanding specialist" foraging strategy (Heller 1980), where individuals specialize on a limited range of high value foods (e.g., ripe fruit), but shift to consuming a broader range of food items, including lower value foods, upon encountering scarcity of the former food type. In addition to “fallback” foods, abundant, lower quality foods that are a reliable source of calories can be viewed as “staple” foods. Foods that fall into this “staple” category for chimpanzees include fruits of genus Ficus (Kanyawara, 8

Kibale: Conklin and Wrangham 1994, Budongo: Newton-Fisher 1999) and termites (Fongoli:

Bogart and Pruetz 2011).

Animal prey. Chimpanzees at all long-term study sites consume vertebrate and invertebrate animal prey (Ngogo, Kibale: Watts et al. 2012a; Gombe: Goodall 1986, Budongo: Newton-

Fischer 1999, Bossou: Matsuzawa et al. 2011; Hockings et al. 2012; Fongoli: Pruetz 2006,

Taï: Boesch and Boesch-Achermann 2000, Mahale: Nishida 2012). The proportion of animal prey in the diet can vary substantially among individuals within a community (Goodall

1986), and among different chimpanzee communities (McGrew 2014). Animal prey usually comprises less than 14% of all feeding observations, and less than 7% in the case of vertebrate prey only (reviewed in Pruetz 2006). Although animal foods form a relatively small portion of the diet, these foods are highly concentrated in protein and fat as well as essential micronutrients, such as vitamin B12, vitamin A, and sodium (Deblauwe and

Janssens 2008; O'Malley and Power 2012; Tennie et al. 2009).

In relation to plant foods, most animal prey are difficult to capture. Chimpanzees have responded to the challenge of accessing social insects with protective nests (e.g.,

Macrotermes) or painful and stings and bites (e.g., Dorylus) with the use of extractive foraging tools (McGrew 2001). In a similar vein, chimpanzees use stick tools to capture small mammalian prey taking refuge in tree cavities (Pruetz and Bertolani 2007). More often, chimpanzees hunt by chasing or ambushing vertebrate prey (Stanford 1996).

The top animal prey in chimpanzees' diets include termites (e.g., Macrotermes), ants

(e.g., Dorylus) and monkeys (e.g., Piliocolobus), but the intensity of animal prey consumption varies between study groups and can be associated with factors such as food availability (Stanford 1996) and prey behavior (Bogart and Pruetz 2011). In some cases, 9 chimpanzees exhibit seasonal patterns in animal prey consumption related to food availability. For instance, the hunting of bushpigs at Mahale coincides with the annual birth pulse (Takahata et al. 1984). However, myriad other factors may influence the decision to capture animal prey. In the case of insectivory at Fongoli, there is a positive correlation between termite fishing and temperature, perhaps because termites are more difficult to capture during hotter temperatures as they descend lower in the nest to seek cooler microclimates (Bogart and Pruetz 2011). Moreover, ripe fruit availability can be high during this time; thus, fruit scarcity does not explain intensive insectivory at Fongoli. These findings highlight that the relationship between animal prey consumption and food availability is complex. More research is needed to assess the nutritional significance of animal foods to chimpanzee diets.

Water intake. Chimpanzees gain water primarily through absorption in the intestinal tract but also through metabolism, and they lose water through evaporation (perspiration, respiration), defecation, and urination (Bourne GH and Golarz de Bourne 1972). Chimpanzees primarily intake water from preformed and free sources (Robbins 1993), but the impact of metabolic water intake, or the water that is gained through oxidative processes during metabolism, is poorly understood in Pan (Bourne GH and Golarz de Bourne 1972). Free water is obtained from sources such as streams, and puddles on the ground or within tree crevices (Matsusaka et al. 2006). Preformed water describes the water molecules present in plant and animal tissues, which is ingested while feeding. Preformed water intake is not a neutral byproduct of consuming watery tissues; it is possible to maintain water balance with exclusively preformed water sources (Papio hamadryas: Zurovsky and Shklonik 1993), thus enabling individuals to travel long distances from free water sources and potentially avoid competitors 10 and predators at these locations. Although little is known about water balance in chimpanzees, kidney structure is similar between Pan and Homo; however, chimpanzees have fewer sweat glands (Bourne GH and Golarz de Bourne 1972). Observations at Fongoli indicate that, in arid environments, chimpanzees need to drink free water on a near-daily basis during the dry season (Pruetz, unpublished data), but more research is needed to determine the relative importance and contributions of preformed, free and metabolic water to daily intake.

Evolutionary ecology of diet. Chimpanzees are unusual among mammals in that they do not fit the typical pattern of decreasing diet quality with increasing body mass (Conklin-Brittain et al. 1998; Milton 1999). They maintain a relatively high quality diet through omnivory, with an emphasis on ripe fruit. Morphological differences in gastrointestinal (GI) tracts between the Hominidae is attributed to differences in dietary strategies (Aiello and Wheeler

1995; Chivers and Hladik 1980, Milton 1999). The alimentary canal of Pan reflects a long, dietary history of omnivory and frugivory, as well as their family’s (Hominidae) position on the phylogenetic tree. In relation to orangutans and gorillas, the GI tract is smaller in chimpanzees (Chivers and Hladik 1980; Stevens 1988), but a comparison with humans shows that chimpanzees have proportionally larger and more sacculated hindguts (Stevens

1988). Thus, a major difference between Homo and Pan is the greater importance of hindgut fermentation for the latter genus.

At the top of the alimentary canal, morphological features of the jaw indicate that processing large quantities of plant material has been an important feeding behavior over evolutionary time scales in chimpanzees. Except for humans, Hominoids have large skeletal and muscular jaw structures, which function to masticate large food boli at the top of the 11 alimentary canal (Ungar 2007). The following anecdote illustrates the enormous volume of the chimpanzee mouth: Bandit, an adult male chimpanzee at the Michael E. Keeling Center in Bastrop, Texas was said to routinely place three, medium-size oranges (Citrus sinensis) entirely in his mouth at the same time (Pruetz, personal communication). Although chimpanzees consume hard foods, such as bark, their relatively thin tooth enamel indicates that the need to routinely masticate tough or hard foods has not been a major selective pressure for chimpanzees (Vogel et al. 2008).

Chimpanzees are equipped to digest plant and animal tissues, especially ripe fruit.

Digesta transit rate is estimated at 38 hours on average (Milton and Demment 1988). As with most primates, chimpanzees have a simple stomach. The volumes of the stomach, and small and large intestines are 1335, 1391 and 2893 cm3, respectively (averaged values taken from

Table 7 in Chivers and Hladik 1980). The voluminous hindgut of the chimpanzee functions to pass large quantities of fibrous digesta, and digest additional fibrous material with the aid of symbiotic microorganisms (Lambert and Fellner 2011). Fragments of indigestible material, such as lignin, chitin, keratin, and bone, is routinely found in the fecal material of chimpanzees (Phillips and McGrew 2013). Because chimpanzees often do not destroy the seeds of many plant species, and they consume large quantities of ripe fruit, chimpanzees are endozoochorous seed dispersers within the ecological community (Lambert and Garber

1998).

Chimpanzees' sense of taste is similar to our own (Hellekant et al. 1997). Field primatologists have recognized this similarity for many decades (reviewed in Nishida et al.

2000). Insipid (i.e., bland), slightly bitter, slightly sweet, or sweet tastes most often characterize foods in the diet of Mahale chimpanzees (Nishida et al. 2000). Taste 12 experiments with captive chimpanzees show that individuals strongly prefer sweet and slightly sweet solutions over insipid ones (Remis 2006). Given that sweet foods, such as ripe fruit and honey, are high in energy, it is likely that a preference for sweet foods can be advantageous for chimpanzees. One taste difference between Pan and Homo is that the former have a higher tolerance for bitter plant foods (Hladik and Simmens 1997, Nishida et al. 2000). Bitter tasting foods contain alkaloids (Aniszewski 2007) or saponins (Shahidi

1995), groups of plant secondary compounds that are antinutritive at sufficiently high concentrations, but may be beneficial to animal health at lower concentrations.

Secondary compounds are synthesized by plants for protection against disease and excessive herbivory (Wink 2010), and are widely believed to impact primate foraging strategies. Chimpanzees consume a variety of secondary compounds in plant tissues

(Hohmann et al. 2010, Reynolds et al. 1998; Wrangham et al. 1998). In an experimental study, Remis (2006) presented chimpanzees with solutions that varied in concentrations of fructose (sweet) and tannic acid (astringent), a secondary compound; chimpanzees preferred fructose solutions with tannic acid to water, and equivalently selected fructose only and fructose-tannic acid solutions, when tannic acid concentration was weak in the latter solution.

These findings show that chimpanzees not only tolerate, but in some cases prefer, sweet foods containing tannic acid up to a certain threshold.

Although chimpanzees routinely ingest unripe and semi-ripe fruits, which often contain higher levels of plant secondary compounds to deter seed predation (Wrangham et al.

1998), they are likely more limited by secondary compounds than Old World Monkeys

(Superfamily Cercopithecoidea) (reviewed by Nishida et al. 2000). In a landmark study,

Wrangham and co-workers (1998) evaluated dietary overlap and antifeedant properties 13 among chimpanzees and sympatric frugivores in the subfamily Cercopithecinae; they found that, overall, chimpanzees consume more ripe fruit and fewer antifeedants than monkeys.

These findings suggest that chimpanzees lack physiological (e.g., proline-rich salivary proteins to counteract antinutritive properties of tannins in hamadryas baboons: Mau et al.

2011) or behavioral (e.g., charcoal consumption in Zanzibar red colobus: Struhsaker et al.

1997) traits allowing them to ingest higher levels of plant secondary compounds, or that chimpanzees are consistently more efficient at harvesting ripe fruit than sympatric frugivorous monkeys, or both. However, the significance of plant secondary compounds on food selection behavior is poorly understood. In addition to competing over fruit, chimpanzees have a predator-prey relationship with monkeys. Because chimpanzees hunt

Cercopithecine monkeys, the may avoid ripe fruit patches when chimpanzees are detected.

Thus, chimpanzees are effective, and at times fearsome, competitors in frugivore communities.

Collectively, these observations highlight the significance of ripe fruit in the diets of chimpanzees and their role as frugivores in ecological communities. In summary, chimpanzees gain the bulk of calories and macronutrients from fruit, but myriad plant and animal foods comprise their diets. In addition to intra- and inter-specific feeding competition, chimpanzee may adjust their food selection behavior in response to plant and animal prey defenses, such as plant secondary compounds and protective nests, respectively. The majority of information on chimpanzee foraging strategies was collected at relatively moist, forested environments, such as Kibale National Park, Uganda (Isabirye-Basuta 1989; Watts et al.

2012a, 2012b) and Taï National Forest in Ivory Coast (N’guessan et al. 2009). In contrast, 14 more research on this behavior in relatively arid environments, such as the savanna- woodland mosaic at Fongoli, Senegal, is needed and discussed in chapter two.

Level two: Habitat quality and selection

Chimpanzees are semi-arboreal, semi-terrestrial quadrupeds (Doran 1996; Hunt 1994) that are primarily found in African equatorial rainforest (Junker et al. 2012, Lehmann and

Dunbar 2009, McGrew 2010a). In general, chimpanzees spend most of their time in or nearby trees because they provide food (Matsuzawa et al. 2011), sleeping sites (Stewart et al.

2007), protection from predators (Pruetz et al. 2008), shade (Pruetz and Bertolani 2009), and are associated with the presence of free water in the case of gallery or riparian forest.

Chimpanzees exhibit several morphological adaptations to arboreality. They are obligate, knuckle-walking quadrupeds with relatively long arms and short legs; their long arms and vertically-oriented shoulder girdle facilitate suspensory locomotion, while a flexible ankle joint and opposable hallux are useful for climbing vertical substrates and gripping branches, respectively (Hunt 1994). Clearly, chimpanzees are well-equipped to live in trees; however, on average they spend about half of their time on the ground (Table 16.1 in

Doran 1996). Furthermore, although chimpanzees are quadrupedal, on occasion they engage in short bouts of bipedal posture and locomotion while in the trees or on the ground to forage, display, and scan their surroundings (Doran 1996; Hunt 1994; Stanford 2006). Thus, while they have retained several morphological features that reflect an arboreal lifestyle, chimpanzees also exhibit a high degree of flexibility in substrate use, locomotor and postural behavior.

Habitat use is strongly influenced by the availability of feeding trees. Most plant food items are from trees, although shrubs, lianas or herbaceous life forms are also important 15

(Matsuzawa et al. 2011, McGrew et al. 1988). Furthermore, several types of animal prey can be found in trees, such as honey, honey comb and brood (McGrew 2014), weaver and carpenter ants (Nishida 1973) and small mammals that nest in tree cavities (Pruetz and

Bertolani 2007). Individuals foraging with the community’s home range, which collectively encompasses the suite of feeding trees and other plants visited within and between years

(Nishida 2012), appear to have sophisticated mental maps and often move in a straight-line fashion between feeding trees (Normand and Boesch 2009).

All great apes construct sleeping nests, a strategy thought to improve comfort while sleeping (reviewed in Stewart et al. 2007). These sleeping structures, called nests, are usually fashioned from branches, leaves and twigs of trees. Although gorillas usually sleep on the ground in nests constructed from herbaceous plants (Willie et al. 2014), chimpanzees (Hicks et al. 2014) and orangutans (van Casteren et al. 2012) primarily sleep in trees. On occasion, they may sleep on the ground in nests made of herbaceous vegetation or saplings (Tagg et al.

2013). This dependency on trees for sleeping is widely believed to be, in part, an antipredator strategy for chimpanzees: resting on branches may increase one's ability to more quickly detect predators, and the tree canopy provides vertical and horizontal escape routes (reviewed in Stewart and Pruetz 2013).

Predation. Most animals have predator avoidance strategies, as there is strong selection for behaviors that minimize risk of depredation for individuals. Such strategies are diverse across the animal kingdom, including but not limited to crypticity, mobbing, and gregariousness

(reviewed in Lima and Dill 1990). Predation is widely believed to play a role in the formation and maintenance of social groups within the Order Primates (reviewed in Miller

2002). Some primates species have elaborate vocal communication systems to deter 16 predation (e.g., Cercopithecus nictitans: Arnold and Zuberbühler 2008; Chlorocebus aethiops: Seyfarth et al. 1980), or are successful at mobbing and attacking predators (e.g.,

Cebus capucinus: Digweed et al. 2005; Colobus badius: Boesch 1994). Great apes have few known natural predators because of their large body sizes and arboreal habits, which raises a question about the significance of predation to chimpanzee society (McGrew 2010b).

Cases of predation involving great ape prey primarily involve big cat predators.

Leopards (Panthera pardus) are the most pervasive carnivore in these cases, responsible for killing chimpanzees (Boesch 2009; Nakazawa et al. 2013), gorillas (Fay et al. 1995) and bonobos (D’Amour et al. 2006), and suspected of killing orangutans (Kanamori et al. 2012).

In addition, chimpanzees have fallen victim to lions (Panthera leo: Nishida 2012) and tigers

(Panthera tigris) might depredate orangutans (Rijksen 1978). Thus, large cats impose a risk of predation to chimpanzees where they are sympatric. Other potential predators additionally include hyenas (Crocuta crocuta) and wild hunting dogs (Lycaon pictus) (Pruetz et al. 2008), pythons (Python sebae) (Pruetz and LaDuke, in prep; Schel et al. 2013), and crowned eagles

(Stephanoaetus coronatus) (McGraw et al. 2006).

Overall, clear cases of predation are rare, which may be explained by low rates of predation, human interference in predator-ape dynamics (sensu Isbell and Young 1993), or both. However, we can examine the impact of predation risk on chimpanzees with indirect evidence, such as nesting and alarm calling behaviors. Goodall (1986) documented two types of antipredator vocalizations in the chimpanzees of Gombe: a “waaa-bark” call that is often given when an individual detects an imminent threat, and the “huu”, which is often a softer call to attention that may also indicate curiosity towards an unusual item. An experimental study by Schel and co-workers (2013) at Budongo found that waaa-barks and “alarm-huus” 17 functioned as a way to alert other community members to potential risk, whereas the “soft- huu” vocalization was generally a rapid reaction to an immediate threat and was less important in conveying predation risk to conspecifics.

Nesting site selection provides further evidence of antipredator strategy. In a comparison of nesting patterns at the sites of Assirik and Fongoli in Senegal, where predation risk from large carnivores is relatively higher and lower, respectively, Assirik chimpanzees tended to nest at greater heights in tree crowns and nested more often in gallery forest, where average tree height is greater (Pruetz et al. 2008). Nesting at greater heights is widely believed to give prey species more opportunity to flee from terrestrial predators. In addition, nesting in closer proximity to other community members may minimize risk through increasing predator detection. As predicted, chimpanzees at Assirik and Ugalla, Tanzania who encounter higher predation risk, nested in a clustered fashion relative to Fongoli chimpanzees (Pruetz et al. 2008; Stewart and Pruetz 2013).

Habitat selection in Pan-Homo sympatry. Humans are predators of (Hicks et al. 2010;

Willcox and Nambu 2007), prey for (reviewed in Hockings et al. 2009) and competitors with

(Halloran et al. 2013; Hockings and Sousa 2013) chimpanzees. Chimpanzees are listed as

Endangered (A4d ver 3.1) by the International Union for Conservation of Nature’s Red List of Threatened Species (Oates et al. 2008), and populations across Africa are currently declining. Their endangered status is widely attributed to the fact that we tend to “win” competitive and predatory interactions with chimpanzees.

In recent history, humans have directly or indirectly killed more chimpanzees than all other predators combined due to hunting and converting chimpanzee habitat to cropland, settlements, mines and roads, for example (Campbell et al. 2008; Kortlandt 1983; Sugiyama 18 and Koman 1992). Accidental or intentional snaring, infant poaching for the pet trade, capturing animals for the biomedical and entertainment industries, retaliation for crop or livestock raiding, bushmeat hunting and disease have led to significant reductions in chimpanzee population sizes (reviewed in Caldecott and Miles 2005 and Stiles et al. 2013).

Thus, chimpanzees generally avoid people, unless they are “naïve” (Morgan and Sanz 2003) or habituated for study (e.g., Goodall 1986). Conversely, chimpanzees have injured or killed people, primarily infants and children (Goldberg et al. 2008; Goodall 1986; Hockings et al.

2009); while these events are rare, they demonstrate a bidirectional predator-prey relationship between these two species.

Humans and chimpanzees compete for natural resources, namely food, water and land. Interspecific competition can negatively impact individual fitness, and is thus widely recognized as a selective force for niche divergence (Hutchinson 1957; Kricher 2011; Lotkaa

1932). Feeding competition with humans is demonstrated by dietary overlap in wild foods

(Pacheco et al. 2012, Sugiyama and Koman 1992), cultivars (Hockings et al. 2007, 2010;

Hockings and Sousa 2013), overlap in mammalian prey species (source) and, on rare occasions, livestock (Gašperšič and Pruetz 2011). Horticultural and agricultural land use can lead to crop raiding behavior in chimpanzees, but this behavior varies in locales where chimpanzees are sympatric with humans. For instance, western chimpanzees at Bossou,

Guinea routinely consume cultivars, such as Papaya carica fruit and Zea maize pith

(Hockings et al. 2007) while this behavior has only been seen once in nine years at Fongoli

(Pruetz, unpublished data). Crop raiding has been attributed to scarcity of wild food resources (McLennen 2013), indicating that non-crop raiding communities are able to find alternate food sources thus preventing human-chimpanzee conflict over cultivars. 19

There is little doubt that human activities profoundly impact chimpanzee populations across Africa. Humans have complex, ecological relationships with chimpanzees that involve competitive and predator-prey interactions. Although the intensive and likely unsustainable use of natural resources by humans within chimpanzee territories has led to several studies that assess conservation threats and develop policy recommendations to reduce conflict and halt chimpanzee population declines (see Stiles et al. 2013), less is known about how competitive and predator-prey relationships between humans and chimpanzees influences the daily lives of chimpanzees, especially in regards to foraging behavior. Chapters two and four address these gaps in greater detail.

Level three: Pan biogeography

Geographical distribution of Pan. The range of the common chimpanzee Pan troglodytes extends from approximately 14°N to 7°S and 15°W to 32°E (Figure 1.1), and from 0 to 2,800 m above sea level (Oates et al. 2008). Of course, there are numerous unoccupied areas within this range due to human land use, and inhospitable natural landforms such as the Dahomey

Gap (Butynski 2001). In general, the species distribution is bounded by extensive open- country savanna landscapes as well as extensive crop and range lands in these drier areas

(Caldecott and Kapos 2005), except that the western edge that is formed by the Atlantic ocean, and the south-central margin is bounded by the Congo River. Bonobos (Pan paniscus), who are only found in the Democratic Republic of Congo, range from the south side of the Congo river to the Kasai/Sankuru river system that is located at roughly latitude

5°S (Lacambra et al. 2005).

Determinants of chimpanzee biogeography. Some of the earliest studies of chimpanzees sought to explain their geographical distribution. For instance, Yerkes (1943:17) suggested 20 that chimpanzees are limited by “temperature and food supply”. However, the generality of this statement leaves much to the imagination. Explanations for chimpanzee biogeography fall into three general categories: extent of forest, geographic barriers and anthropogenic disturbance. Within this framework, hypotheses set forth to explain Pan biogeography involve five, non-mutually exclusive factors, namely forest cover, food and water availability, predation risk, body mass and social group size.

Forest cover and proximity is positively associated with chimpanzee biogeography

(Junker et al. 2010; Lehman and Dunbar 2009; Torres et al. 2010). In savanna environments around the margins of the species’ geographic distribution, chimpanzees are thought to be limited by woody vegetation cover because, in these environments, most dietary items are located in woodland or forest habitats (McGrew et al. 1981; Pruetz 2006) and these habitats provide refuge from predators (Lehmann and Dunbar 2009). Furthermore, accessibility of free water during the dry season, which should be linked to forest availability, is another factor that may explain their distribution in open-country habitats (Kortlandt 1972). McGrew and co-authors (1988) propose that savanna chimpanzees require that at least 3% of their home range consist of evergreen forest, given that this figure was the forest cover estimate for Assirik chimpanzees. Similarly, Lehmann and Dunbar (2009) suggest that, aside from extreme outliers like Assirik and Fongoli, chimpanzees should have at least a third of their home range covered by forest; further, they argue that areas with less forest cover are likely to be challenging for chimpanzees to occupy due to factors such as body mass, foraging time constraints, living in groups, and elevated predation risk in open habitats. They conclude that chimpanzees are unlikely to maintain communities in areas with low forest cover because their relatively high-quality diet and large body mass leads to unsustainably high feeding 21 competition within communities and foraging parties, and that solitary foraging is too risky for chimpanzees in open habitats with large predators such as leopards and lions; it is for these reasons that the species’ geographical distribution is bounded by savanna habitats

(Lehmann and Dunbar 2009). Although intriguing, this time budget model performs poorly with chimpanzees at Fongoli and Assirik, who exhibit species-typical fission-fusion patterns and range in areas with extremely low forest cover (McGrew et al. 1988; Pruetz and

Bertolani 2009; Tutin et al. 1983). Although risk of predation from large carnivores is generally low at Fongoli (Pruetz et al. 2008), feeding competition does not deter social foraging and, in fact, Fongoli chimpanzees appear to be more cohesive than many forest- dwelling communities (Pruetz and Bertolani 2009). Meanwhile, Assirik chimpanzees are sympatric with several large predators, including lions, leopards, hyenas and wild hunting dogs. More research is needed to understand the forces that shape chimpanzee biogeography in savanna habitats.

Several lines of evidence indicate that savanna habitats are marginal for chimpanzees

(see above). In addition, genetic diversity is greatest in central African rainforest populations, indicating that the species originated in these areas (Fischer et al. 2006). Consequently, chimpanzee biogeography is often associated with the refugia hypothesis, or the idea that the species’ range expanded and contracted synchronously with forest habitat during warming and cooling cycles in the Pleistocene epoch (approximately 2 mya - 11 kya BP) (reviewed by

Mayr and O'Hara 1986). However, the only study to explicitly test the refugia hypothesis found that it was only weakly supported by genetic evidence in eastern chimpanzees; more likely, gene flow occurred among populations in various forest refugia, probably from individuals dispersing through savanna, woodland, and riparian habitats in spite of 22 fluctuations in forest cover over time (Goldberg 1998). Unfortunately, the chimpanzee fossil record contributes rather little to our understanding of chimpanzee biogeography, as only one fossil site has been discovered. The dental remains of at least one individual dating to approximately 500 kya were uncovered in deposits indicative of a lake surrounded by wooded habitat in the East African Rift Valley (McBrearty and Jablonski 2005), east of the current species distribution. Prior to this important fossil discovery, it was hypothesized that open habitats within the rift valley barred chimpanzees from accessing coastal forests along the eastern seaboard (Kortlandt 1983). Studies of genetic diversity and fossils have answered many questions about chimpanzee biogeography and habitat selection, but many more remain. For instance, we do not know how survival and reproduction varies between forest and savanna populations. Thus, it is not yet possible to determine if, overall, savanna habitats lead to population “sinks” and forests promote population “sources” (Pulliam 1988). In any case, woody habitats are a critical resource for chimpanzees (Lehmann and Dunbar 2009;

McGrew et al. 1988; Pruetz and Bertolani 2009). However, few studies have examined relationships between chimpanzee biogeography and woody habitat composition and physiognomy (Torres et al. 2010). The next step is to use a landscape approach (sensu

Dunning et al. 1992) to identify patterns in savanna landscapes that are associated with the species distribution.

In recent history, habitat loss has impacted Pan biogeography, with range contraction across Africa being the dominant trend. Torres and co-workers (2010) found that chimpanzees were strongly associated with forest and woodland in Guinea, and that these habitats had considerably decreased between 1986 and 2003. Similarly, chimpanzee population size and forest cover have dramatically declined between 1989 and 2007 in the 23

Ivory Coast (Campbell et al. 2008). On a broader geographic scale, Junker and co-workers

(2010) found a decrease in suitable habitat across the species’ range associated with anthropogenic disturbance. Given that chimpanzee conservation is in crisis, a better understanding of their biogeography is critically needed. Chapter five elaborates on this issue and presents new findings that contribute to our understanding of chimpanzee biogeography in Senegal.

Dissertation structure This dissertation strives to fill gaps in our understanding of the behavioral ecology of savanna-dwelling chimpanzees. In chapter two, I discuss the significance of savanna chimpanzee behavior and ecology to the field of evolutionary anthropology. Chapter three evaluates the foraging decisions of Fongoli chimpanzees in light of the nutritional composition of dietary items as well as the abundance and distribution of foods within the chimpanzees' home range. Next, in chapter four I examine foraging decisions more broadly through evaluating individuals' food choices in relation to the habitat context in which they are made. Building on the information gleaned in the chapters three and four, chapter five assesses characteristics of real and potential chimpanzee habitats in southeastern Senegal in order to target key factors that shape the species' geographical distribution. Finally, in chapter six I summarize key findings, their relevance to anthropologists, ecologists and conservationists, and propose new research directions for the study of savanna chimpanzee foraging behavior.

Literature cited Aiello LC, and Wheeler P. 1995. The expensive-tissue hypothesis: the brain and the digestive

system in human and primate evolution. Curr Anthropol 36:199-221.

Altevogt BM, Pankevich PE, Shelton-Davenport MK, Kahn JP, Editors. 2011. Committee on

the use of chimpanzees in biomedical and behavioral research. National Research

Council.

Aniszewski T. 2007. Alkaloids - secrets of life: alkaloids chemistry, biological significance,

applications and ecological role. Amsterdam: Elsevier.

Arnold K, Zuberbühler K. 2008. Meaningful call combinations in a non-human primate. Curr

Biol 18:R202.

Boesch. 1994. Chimpanzee-red colobus monkeys: a predator-prey system. Anim Behav

47:1135-1148.

Boesch C. 2009. The real chimpanzee. New York: Cambridge University Press.

Boesch C. Boesch-Achermann H. 2000. The chimpanzees of the Taї forest. Oxford: Oxford

University Press.

Boesch C, Hohmann G, and Marchant LF, editors. 2002. Behavioural diversity in

chimpanzees and bonobos. Cambridge: Cambridge University Press. p 35-51.

Bogart S, Pruetz J. 2011. Insectivory of savanna chimpanzees (Pan troglodytes verus) at

Fongoli, Senegal. Am J Phys Anthropol 145:11-20.

Bourne GH, Golarz de Bourne, MN. 1972. The histology and histochemistry of the

chimpanzee tissues and organs. In: Bourne GH, editor. The chimpanzee vol. 5.

Baltimore, MA: University Park Press. p 2-76. 25

Butynski TM. 2001. Africa's great apes. In: Beck BB, Stoinski TS, Hutchings M, Maple TL,

Norton B, Rowan A, Stevens EF, and Arluke A, editors. Great apes and humans: the

ethics of coexistence. Washington D.C.: Smithsonian Institution Press. p 3-56.

Caldecott J, and Miles L, editors. 2005. World Atlas of Great Apes and their Conservation.

Berkeley, CA: UNEP-WCMC, University of California Press.

Campbell G, Kuehl H, Kouame PNG, Boesch C. 2008. Alarming decline of West African

chimpanzees in Côte d’Ivoire. Curr Biol 18: R903

Carter J, Ndiaye S, Pruetz JD, and McGrew WC. 2003. Senegal. In: Kormos R, Boesch C,

Bakarr M, and Butynski TM, editors. West African chimpanzees. Gland: World

Conservation Union. p 31-39.

Chivers DJ, Hladik CM. 1980. Morphology of gastrointestinal tract in primates: comparison

with other mammals in relation to diet. J Morphol 166:337-386.

Conklin NL, Wrangham RW. 1994. The value of figs to a hind-gut fermenting frugivore - a

nutritional analysis. Biochem Syst Ecol 22:137-151.

Conklin-Brittain NL, Wrangham RW, and Hunt KD. 1998. Dietary response of chimpanzees

and cercopithecines to seasonal variation in fruit abundance. II. Macronutrients. Int J

Primatol 19:971-998.

D’Amour DE, Hohmann G, Fruth B. 2006. Evidence of leopard predation on bonobos. Folia

Primatol 77:212-217.

Deblauwe I, and Janssens GPJ. 2008. New insights in insect prey choice by chimpanzees and

gorillas in Southeast Cameroon: The role of nutritional value. Am J Phys Anthropol

135:42-55. 26

Digweed SM, Fedigan LM, Rendall D. 2005. Variable specificity in the anti-predator

vocalizations and behaviour of the white-faced capuchin, Cebus capucinus.

Behaviour 142:997-1021.

Dominguez-Rodrigo M. 2014. Is the "savanna hypothesis" a dead concept for explaining the

emergence of the earliest hominins? Curr Anthropol 55:59-81.

Doran DM. 1996. Comparative positional behavior of the African apes. In: McGrew WC,

Marchant LF, and Nishida T, editors. Great ape societies. Cambridge: Cambridge

University Press. p 213-224.

Dunning J, Danielson BJ, Pulliam HR. 1992 Ecological processes that affect populations in

complex landscapes. Oikos 65:169-175.

Emery-Thompson M, and Wrangham RW. 2008. Diet and reproductive function in wild

female chimpanzees (Pan troglodytes schweinfurthii) at Kibale National Park,

Uganda. Am J Phys Anthropol 135:171-181.

Emery Thompson M, Kahlenberg SM, Gilby IC, and Wrangham RW. 2007. Core area

quality is associated with variance in reproductive success among female

chimpanzees at Kibale National Park. Anim Behav 73:501-512.

Fay JM, Carroll R, Kerbis Peterhans JC, Harris D. 1995. Leopard attack on and consumption

of gorillas in the Central African Republic. J Hum Evol 29:93-99.

Fischer A, Pollack J, Thalmann O, Nickel B, Pääbo S. 2006. Demographic history and

genetic differentiation in apes. Curr Biol 16:1133-1138

Gašperšič M. 2008. Fongoli chimpanzees (Pan troglodytes verus) and baobab (Adansonia

digitata) fruits: modeling the evolution of hominin material culture on percussive 27

technology. Ph.D. Dissertation, Ljubljana Graduate School of the Humanities,

Ljubljana, Slovenia.

Gilby I, and Wrangham R. 2007. Risk-prone hunting by chimpanzees (Pan troglodytes

schweinfurthii) increases during periods of high diet quality. Behav Ecol Sociobiol

61:1771-1779.

Gilby IC, Wilson ML, Pusey AE. 2013. Ecology rather than psychology explains co-

occurrence of predation and border patrols in male chimpanzees. Anim Behav 86:61-

74.

Goldberg TL. 1998. Biogeographic predictors of genetic diversity in populations of eastern

african chimpanzees (Pan troglodytes schweinfurthi). Int J Primatol 19:237-254.

Goldberg TL, Gillespie TR, Rwego IB. 2008. Health and disease in the people, primates, and

domestic animals of Kibale National Park: implications for conservation. In Science

and Conservation in African Forests: The Benefits of Long-Term Research, edited by

R. Wrangham and E. Ross. Cambridge, Cambridge University Press: pages 75-87.

Gonder MK, Oates JF, Disotell TR, Forstner MRJ, Morales JC, Melnick DJ. 1997. A new

West African chimpanzee subspecies? Nature 388:337.

Gonder MK, Disotell TR, Oates JF. 2006. New genetic evidence on the evolution of

chimpanzee populations and implications for taxonomy. Int J Primatol 27(4):1103-

1127.

Gonder MK, Locatelli S, Ghobrial L, Mitchell MW, Kujawski JT, Lankester FJ, Stewart CB,

Tishkoff SA. 2011. Evidence from Cameroon reveals differences in the genetic

structure and histories of chimpanzee populations. PNAS 108(12):4766-4771. 28

Goodall J. 1986. The chimpanzees of Gombe. Cambridge, MA: The Belknap Press of

Harvard University Press.

Goodall J. 1994. Foreword. In: Wrangham RW, McGrew WC, de Waal FBM, and Heltne

PG, editors. Chimpanzee cultures. Cambridge, MA: Harvard University Press.

Groves C. 2001. Primate taxonomy. Smithsonian Institution Press, Washington, DC, USA.

Halloran AR, Cloutier CT, Baisesay P. 2013. A previously undiscovered group of

chimpanzees (Pan troglodytes verus) is observed living in the Tonkolili district of

Sierra Leone. Am J Primatol 75:519-523.

Hellekant G, Ninomiya Y, and Danilova V. 1997. Taste in chimpanzees II: single chorda

tympani fibers. Physiol Behav 61:829-841.

Heller R. 1980. On optimal diet in a patchy environment. Theor Pop Biol 17:201-214.

Hicks T, Darby L, Hart J, Swinkels J, January N, Menken S. 2010. Trade in orphans and

bushmeat threatens one of the Democratic Republic of the Congo's most important

populations of eastern chimpanzees (Pan troglodytes schwinfurthii). Afr Primates

7:1-18.

Hicks TC, Tranquilli S, Kuehl H, Campbell G, Swinkels J, Darby L, Boesch C, Hart J,

Menken SBJ. 2014. Absence of evidence is not evidence of absence: discovery of a

large, continuous population of Pan troglodytes schwinfurthii in the Central Uele

region of northern DRC. Biol Conserv 171:107-113.

Hill WCO. 1969. The discovery of the chimpanzee. In: Bourne GH, editor. The chimpanzee.

vol. 1. Basel: Karger. p 2-49.

Hladik CM, Simmens B. 1997. Taste perception and feeding behavior in nonhuman primates

and human populations. Evol Anthropol 58-71. 29

Hockings KJ, Sousa C. 2013. Differential utilization of cashew—a low-conflict crop—by

sympatric humans and chimpanzees. Oryx 46:375-381

Hockings KJ, Humle T, Andersen JR, Biro D, Sousa C, Ohashi G, Matsuzawa T. 2007.

Chimpanzees share forbidden fruit. PLoS ONE 2: 886.

Hockings KJ, Anderson JR, and Matsuzawa T. 2009. Use of wild and cultivated foods by

chimpanzees at Bossou, Republic of Guinea: feeding dynamics in a human-

influenced environment. Am J Primatol 71:636-646.

Hockings KJ, Anderson JR, and Matsuzawa T. 2010. Flexible feeding on cultivated

underground storage organs by rainforest-dwelling chimpanzees at Bossou, West

Africa. J Hum Evol 58:227-233.

Hockings KJ, Anderson JR, and Matsuzawa T. 2012. Socioecological adaptations by

chimpanzees, Pan troglodytes verus, inhabiting an anthropogenically impacted

habitat. Anim Behav 83:801-810.

Hohmann G, Potts K, N'Guessan A, Fowler A, Mundry R, Ganzhorn JU, Ortmann S. 2010.

Plant foods consumed by Pan: exploring the variation of nutritional ecology across

Africa. Am J Phys Anthropol 141:476-485.

Hoppius CE. 1760. Anthropomorpha. Ph.D. Dissertation. Uppsala University.

Hunt K. 1994. The evolution of human bipedality: ecology and functional morphology. J

Hum Evol 26:183-202.

Hutchinson GE. 1957. Concluding remarks. Cold Spring Harbor Symposia Quant Biol

22:415-427. 30

Isabirye-Basuta G. 1989. Feeding ecology of chimpanzees in the Kibale Forest, Uganda. . In:

Heltne P, Marquardt L, editors. Understanding chimpanzees. Cambridge, MA:

Harvard University Press. p 116-127.

Junker J, Blake S, Boesch C, Campbell G, Toit L, Duvall C, Ekobo A, Etoga G, Galat-Luong

A, Gamys J, Ganas-Swaray J, Gatti S, Ghiurghi A, Granier N, Hart J, Head J,

Herbinger I, Hicks TC, Huijbregts B, Imong IS, Kuempel N, Lahm S, Lindsell J,

Maisels F, McLennan M, Martinez L, Morgan B, Morgan D, Mulindahabi F, Mundry

R, N'Goran KP, Normand E, Ntongho A, Okon DT, Petre CA, Plumptre A, Rainey H,

Regnaut S, Sanz C, Stokes E, Tondossama A, Tranquilli S, Sunderland-Groves J,

Walsh P, Warren Y, Williamson EA, Kuehl HS. 2012. Recent decline in suitable

environmental conditions for African great apes. Divers Distribut 18: 1077-1091.

Kanamori T, Kuze N, Bernard H, Malim TP, Kohshima S. 2012. Fatality of a wild Bornean

orangutan (Pongo pygmaeus morio): behavior and death of a wounded juvenile in

Danum Valley, North Borneo. Primates 53:221-226.

Kortlandt A. 1972. New perspectives on ape and human evolution. Amsterdam: Stitchting

voor Psychobiologie.

Kortlandt A. 1983. Marginal habitats of chimpanzees. J Hum Evol 12:231-278.

Kricher J. 2011. Tropical ecology. Princeton: Princeton University Press.

Lacambra C, Thompson J, Furuichi T, Vervaecke H, Stevens J. 2005. Bonobo (Pan

paniscus). In: Caldecott J, and Miles L, editors. The world atlas of great apes and

their conservation. Berkeley, CA: University of California Press. p 82-96.

Lambert JE, Garber PA. 1998. Evolutionary and ecological implications of primate seed

dispersal. Am J Primatol 45:9-28. 31

Lambert JE, Fellner V. 2012. In vitro fermentation of dietary carbohydrates consumed by

African apes and monkeys: preliminary results for interpreting microbial and

digestive strategy. Int J Primatol 33:262-281.

Lehmann J, and Dunbar R. 2009. Implications of body mass and predation for ape social

system and biogeographical distribution. Oikos 118:379-390

Lehmann J, Korstjens A, and Dunbar R. 2007. Fission–fusion social systems as a strategy for

coping with ecological constraints: a primate case. Evol Ecol 21:613-634.

Lima SL, and Dill LM. 1990. Behavioral decisions made under the risk of predation - a

review and prospectus. Can J Zool 68:619-640.

Lotkaa J. 1932. The growth of mixed populations, two species competing for a common food

supply. J Wash Acad Sci 22: 461469.

MacArthur RH, and Pianka ER. 1966. On optimal use of a patchy environment. Am Nat

100:603-609.

Marks J. 2002. What it means to be 98% chimpanzee: apes, people, and their genes.

Berkeley, CA: University of California Press.

Mason JH, Wrangham RW. 1991. Intergroup aggression in chimpanzees and humans. Curr

Anthropol 32: 369-390.

Matsusaka T, Nishie H, Shimada M, Kutsukake N, Zamma K, Nakamura M, and Nishida T.

2006. Tool-use for drinking water by immature chimpanzees of Mahale: prevalence

of an unessential behavior. Primates 47:113-122.

Matsuzawa T, Sugiyama Y, Humle T, eds. 2011. The chimpanzees of Bossou and Nimba.

Tokyo: Springer. 32

Mau M, Martinho de Almeida A, Varela Coelho A, Südenkum KH. 2011. First identification

of tannin-binding proteins in saliva of Papio hamadryas using MS/MS mass

spectrometry. Am J Primatol 73:896-902.

Mayr E, and O'Hara RJ. 1986. The biogeographic evidence supporting the Pleistocene forest

refuge hypothesis. Evol 40:55-67.

McBrearty S, and Jablonski NG. 2005. First fossil chimpanzee. Nature 437(7055):105-108.

McGraw WS, Cooke C, and Shultz S. 2006. Primate remains from African crowned eagle

(Stephanoaetus coronatus) nests in Ivory Coast's Tai Forest: implications for primate

predation and early hominid taphonomy in South Africa. Am J Phys Anthropol

131:151-165.

McGrew W. 2010a. In search of the last common ancestor: new findings on wild

chimpanzees. Phil Trans Roy Soc B 365:3267-3276.

McGrew W. 2010b. Christophe Boesch: The real chimpanzees. Sex strategies in the forest.

Primates 51:189-190.

McGrew WC. 2014. The 'other faunivory' revisited: insectivory in human and non-human

primates and the evolution of human diet. J Hum Evol 71:4-11.

McGrew WC, Baldwin P, and Tutin CEG. 1981. Chimpanzees in a hot, dry and open habitat:

Mt. Assirik, Senegal, West Africa. J Hum Evol 10:217-244.

McGrew WC, Baldwin PJ, and Tutin CEG. 1988. Diet of wild chimpanzees (Pan troglodytes

verus) at Mt. Assirik, Senegal: I. Composition. Am J Primatol 16:213-226.

McLennan MR. 2013. Diet and feeding ecology of chimpanzees (Pan troglodytes) in Bulindi,

Uganda: foraging strategies at the forest-farm interface. Int J Primatol 34:585-614. 33

Miller L, editor. 2002. Eat or be eaten: predator sensitive foraging among primates.

Cambridge, UK: Cambridge University Press.

Milton K. 1999. A hypothesis to explain the role of meat-eating in human evolution. Evol

Anthropol 8:11-21

Milton K, Demment MW. 1988. Digestion and passage kinetics of chimpanzees fed high and

low fiber diets and comparison with human data. J Nutr 118:1082-1088.

Morgan D, Sanz C. 2003. Naive encounters with chimpanzees in the Goualougo Triangle,

Republic of Congo. Int J Primatol 24: 369-381.

Nakazawa N, Hanamura S, Inoue E, Nakatsukasa M, Nakamura M. 2013. A leopard ate a

chimpanzee: first evidence from East Africa. J Hum Evol 65:334-337.

Ndiaye S. 2011. Conservation du Chimpanzé au Sénégal. Etat des Connaissances et

Réactualisation du Plan d'Actions. Report for USAID/Wula Nafaa.

Newton-Fisher NE. 1999. The diet of chimpanzees in the Budongo Forest Reserve, Uganda.

Afr J Ecol 37:344-354.

N'guessan AK, Ortmann S, and Boesch C. 2009. Daily energy balance and protein gain

among Pan troglodytes verus in the Taї National Park, Côte d'Ivoire. Int J Primatol

30:481-496.

Nishida T. 1968. The social group of wild chimpanzees in the Mahali Mountains. Primates

9:167-224.

Nishida T. 1973. The ant-gathering behaviour by the use of tools among wild chimpanzees of

the Mahali mountains. J Hum Evol 2:357-370.

Nishida T. 2012. Chimpanzees of the lakeshore: natural history and culture at Mahale.

Cambridge: Cambridge University Press. 34

Nishida T, Ohigashi H, Koshimizu K. 2000. Tastes of chimpanzee plant foods. Curr

Anthropol 41:431-438.

Normand E, Boesch C. 2009. Sophisticated Euclidean maps in forest chimpanzees. Anim

Behav 77:1195-1201.

Nystrom, P Ashmore P. 2008. The life of primates. Upper Saddle River, NJ: Pearson Prentice

Hall.

Oates JF, Tutin CEG, Humle T, Wilson ML, Baillie JEM, Balmforth Z, Blom A, Boesch C,

Cox D, Davenport T, Dunn A, Dupain J, Duvall C, Ellis CM, Farmer KH, Gatti S,

Greengrass E, Hart J, Herbinger I, Hicks C, Hunt KD, Kamenya S, Maisels F, Mitani

JC, Moore J, Morgan BJ, Morgan DB, Nakamura M, Nixon S, Plumptre AJ,

Reynolds V, Stokes EJ, Walsh PD. 2008. Pan troglodytes Retrieved 10 November

2010, from IUCN 2010. IUCN Red List of Threatened Species. Version 2010.4.

Oftedal OT, Whiten A, Southgate DAT, and Soest PV. 1991. The nutritional consequences of

foraging in primates: the relationship of nutrient intakes to nutrient requirements Phil

Trans Roy Soc B 334(1270):161-170.

O'Malley RC, and Power ML. 2012. Nutritional composition of actual and potential insect

prey for the Kasekela chimpanzees of Gombe National Park, Tanzania. Am J Phys

Anthropol 149:493-503.

Pacheco L. Fraixeda S, Ferández-Llamazares Á, Estela N, Moinee R, Guallar F. 2012.

Prospectives on sustainable resource conservation in community nature reserves: a

case study from Senegal. Sustainability 4:3158-3179.

Phillips C, McGrew WC. 2013. Identifying species in chimpanzee (Pan troglodytes) feces: a

methodological lost cause? Int J Primatol 34:792-807. 35

Pruetz JD. 2006. Feeding ecology of savanna chimpanzees (Pan troglodytes verus) at

Fongoli, Senegal. In: Hohmann G, Robbins MM, and Boesch C, editors. Feeding

ecology in apes and other primates. Cambridge, UK: Cambridge University Press. p

161-182.

Pruetz J. 2007. Evidence of cave use by savanna chimpanzees (Pan troglodytes verus) at

Fongoli, Senegal: implications for thermoregulatory behavior. Primates 48:316-319.

Pruetz JD. 2009. The socioecology of adult female patas monkeys and vervets in Kenya.

Upper Saddle River, NJ: Pearson/Prentice Hall.

Pruetz JD, and Bertolani P. 2007. Savanna chimpanzees, Pan troglodytes verus, hunt with

tools. Curr Biol 17:1-6.

Pruetz JD, and Bertolani P. 2009. Chimpanzee (Pan troglodytes verus) behavioral responses

to stresses associated with living in a savanna-mosaic environment: implications for

hominin adaptations to open habitats. Paleoanthropology:252-262.

Pruetz JD, Fulton SJ, Marchant LF, McGrew WC, Schiel M, and Waller M. 2008. Arboreal

nesting as anti-predator adaptation by savanna chimpanzees (Pan troglodytes verus)

in southeastern Senegal. Am J Primatol 70:393-401.

Pulliam HR. 1988. Sources, sinks, and population regulation. Am Nat 132:652-661.

Remis MJ. 2006. The role of taste in food selection by African apes: implications for niche

separation and overlap in tropical forests. Primates 47:56-64.

Reynolds V, Plumptre AJ, Greenham J, and Harborne J. 1998. Condensed tannins and sugars

in the diet of chimpanzees (Pan troglodytes schweinfurthii) in the Budongo Forest,

Uganda. Oecologia 115:331-336. 36

Rijksen HD (1978) A field study on Sumatran orangutans (Pongo pygmaeus abelii, Lesson

827), In: Veenman H, and Zonen BV, editors. Ecology, behavior and conservation.

Wageningen, The Netherlands.

Ruvolo M, Pan D, Zehr S, Goldberg T, Disotell TR, von Dornum M. 1994. Gene trees and

hominoid phylogeny. PNAS 91:8900-8904.

Salzmann U, Hoelzmann P. 2005. The Dahomey Gap: an abrupt climatically induced rain

forest fragmentation in West Africa during the late Holocene. The Holocene 15(2):

190-199.

Schel AM, Townsend SW, Machanda Z, Zuberbuhler K, Slocombe KE. 2013. Chimpanzee

alarm call production meets key criteria for intentionality. PLoSOne 8(10).

Schwarz E. 1934. On the local races of the chimpanzee. Ann Mag Nat Hist 13:576-583.

Seyfarth RM, Cheney DL, Marler P. 1980. Vervet monkey alarm calls: semantic

communication in a free-ranging primate. Anim.Behav. 28:1070–94

Shahidi F, editor. 1995. Antinutrients and phytochemicals in food. Washington DC:

American Chemical Society.

Stanford CB. 1996. The hunting ecology of wild chimpanzees: implications for the

evolutionary ecology of pliocene hominids. Am Anthropol 98:96-113.

Stanford CB. 2006. Arboreal bipedalism in wild chimpanzees: Implications for the evolution

of hominid posture and locomotion. Am J Phys Anthropol 129:225-231.

Stephens DW, and Krebs JR. 1986. Foraging theory. Princeton: Princeton University Press.

Stephens DW, Brown JS, and Ydenberg RC, editors. 2007. Foraging: behavior and ecology.

Chicago: University of Chicago Press. 37

Stevens CE. 1988. Comparative physiology of the vertebrate digestive system. Cambridge,

UK: Cambridge University Press.

Stewart FA, Pruetz JD. 2013. Do chimpanzee nests serve an anti-predatory function? Am J

Primatol 75:593-604.

Stewart F, Pruetz JD, Hansell MH. 2007. Do chimpanzees build comfortable nests. Am J

Primatol 69:1-10.

Stiles D, Redmond I, Cress D, Nellemann C, Formo RK. 2013. Stolen apes – the illicit trade

in chimpanzees, gorillas, bonobos and orangutans. A rapid response assessment.

United Nations Environment Programme, GRID-Arendal.

Struhsaker TT, Cooney DO, Siex KS. 1997. Charcoal consumption by Zanzibar red colobus

monkeys: its function and its ecological and demographic consequences. Int J

Primatol 18(1):61-72.

Sugiyama Y, Koman J. 1992. The flora of Bossou: its utilization by chimpanzees and

humans. Afr Study Monographs 13:127-169.

Tagg N, Willie J, Petre CA, Haggis O. 2013. Ground night nesting in chimpanzees: new

insights from central chimpanzees (Pan troglodytes troglodytes) in south-east

Cameroon. Folia Primatol 84:362-393.

Takahata Y, Hasegawa T, Nishida T. 1984. Chimpanzee predation in the Mahale mouontains

from August 1979 to May 1982. Int J Primatol 5:213-233.

The Chimpanzee Sequencing and Analysis Consortium (TCSAC). 2005. Initial sequence of

the chimpanzee genome and comparison with the human genome. Nature 43:69-87. 38

Tennie C, Gilby I, and Mundry R. 2009. The meat-scrap hypothesis: small quantities of meat

may promote cooperative hunting in wild chimpanzees (Pan troglodytes). Behav Ecol

Sociobiol 63:421-431.

Torres J, Brito JC, Vasconcelos MJ, Catarino L, Gonçalves J, and Honrado J. 2010.

Ensemble models of habitat suitability relate chimpanzee (Pan troglodytes)

conservation to forest and landscape dynamics in Western Africa. Biol Conserv

143:416-425.

Tutin CEG, McGrew WC, Baldwin PJ. 1983. Social organization of savanna-dwelling

chimpanzees, Pan troglodytes verus, at Mt. Assirik, Senegal. Primates 24:154-173.

UNEP-WCMC, IUCN. 2008. Pan troglodytes. IUCN Red List of Threatened Species.

Version 2013.2. IUCN.

Ungar PS, editor. 2007. Evolution of the human diet. Oxford: Oxford University Press. van Casteren A, Sellers WI, Thorpe SKS, Coward S, Crompton RH, Myatt JP, Roland Ennos

A. 2012. Nest-building orangutans demonstrate engineering know-how to produce

safe, comfortable beds. PNAS 109:6873-6877. van Lawick-Goodall J. 1968. The behaviour of free-living chimpanzees in the Gombe stream

reserve. Anim Behav Monographs 1:161-311 van Lawick-Goodall J. 1970. Tool-using primates and other vertebrates. In: Lehrman DS,

Hinde RA, Shaw E, editors. Advances in the study of behavior, vol. 3. New York:

Academic Press. p 195-249.

Watts DP, Potts KB, Lwanga JS, Mitani JC. 2012a. Diet of chimpanzees (Pan troglodytes

schweinfurthii) at Ngogo, Kibale National Park, Uganda, 1. Diet composition and

diversity. Am J Primatol 74:114-129. 39

Watts DP, Potts KB, Lwanga JS, Mitani JC. 2012b. Diet of chimpanzees (Pan troglodytes

schweinfurthii) at Ngogo, Kibale National Park, Uganda, 2. Temporal variation and

fallback foods. Am J Primatol 73:1-15.

Willcox AS, Nambu DM. 2007. Wildlife hunting practices and bushmeat dynamics of the

Banyangi and Mbo people of southwest Cameroon. Biol Conserv 134: 251-261.

Willie J, Tagg N, Petre CA, Pereboom Z, Lens L. 2014. Plant selection for nest building by

western lowland gorillas in Cameroon. Primates 55:41-49.

Wink M. 2010. Introduction: biochemistry, physiology, and ecological functions of

secondary metabolites. Ann Plant Reviews: 40:1-19.

Wrangham RW. 1975. The behavioural ecology of chimpanzees in Gombe National Park,

Tanzania. Ph.D. Dissertation, University of Cambridge, Cambridge.

Wrangham RW. 1980. An ecological model of female-bonded primate groups. Behaviour

75:262-300.

Wrangham RW, Conklin NL, Chapman CA, Hunt KD. 1991. The significance of fibrous

foods for the Kibale forest chimpanzees. Phil Trans Roy Soc B 334:171-178.

Wrangham RW, Conklin-Brittain NL, and Hunt KD. 1998. Dietary response of chimpanzees

and cercopithecines to seasonal variation in fruit abundance. I. Antifeedants. Int J

Primatol 19:949-970.

Yerkes RM. 1925. Almost human. New York: Century.

Yerkes RM. 1943. The Yale laboratories of primate biology. The Scientific Monthly 56:287-

290.

Zurovsky Y, Shklonik A. 1993. Water economy and body fluid distribution in the hamadryas

baboon (Papio hamadryas). J Therm Biol 18:153-157. 40

Figure 1.1. Geographical distribution map for Pan troglodytes according to subspecies and locations for long-term studies of habituated chimpanzees. Sources: UNEP-WCMC and IUCN

(2008), Fongoli (Pruetz 2006), Bossou (Matsuzawa et al. 2011), Taï (Boesch 2009), Budongo

(Reynolds et al. 1998), Kanyawara (Wrangham et al. 1998), Ngogo (Watts et al. 2012a), Gombe

(Gilby and Wrangham 2007) and Mahale (Nishida 2012). 42

CHAPTER II SAVANNA CHIMPANZEES: A BRIEF REVIEW The study of savanna chimpanzees is of significance to evolutionary anthropology because of the long held belief that savanna environments played an important role in the evolution of genus Homo. Specifically, it is hypothesized that early hominins left an arboreal lifestyle behind to extensively range in open-country habitats to search for sustenance (reviewed in Dominguez-Rodrigo 2014). Assuming that chimpanzees are the closest extant model of these early hominins (e.g., Ardipithecus, Australopithecus) in terms of morphology, physiology, and behavior (reviewed in McGrew 2010), anthropological primatologists that investigate how savanna chimpanzees survive in these arid environments may provide new insights on the behavior and ecology of the last common ancestor between humans and chimpanzees.

Defining savannas Most taxa within the Order Primates inhabit tropical and subtropical forest habitats, while relatively few primate species’ ranges include desert, and tropical or subtropical savannas

(Bourlière 1985). Chimpanzees today predominately occupy a belt of tropical forest in equatorial

Africa; however, they also range in woodlands and mosaics of grassland, woodland and forest, as well as agro-forest mosaics in human-dominated landscapes (Junker et al. 2012; Lehmann and

Dunbar 2009; McGrew 2010). Despite five decades of research on wild chimpanzees (Boesch and Boesch-Achermann 2000; Goodall 1986; Nishida 2012; Reynolds 2005), we know relatively little about the behavior and ecology of chimpanzee communities that inhabit African savannas with little forest cover (Hunt and McGrew 2002; Moore 1996; Pruetz 2006).

The term ‘savanna’ broadly refers to a range of vegetation types, including woodlands and mosaics of grasslands and woodlands (Dominguez-Rodrigo 2014). Key features of savannas include deciduous, woody trees and shrubs among a dominant, herbaceous understory and an 43

abundance of C4 plants. Savannas range from open grasslands with few, scattered woody shrubs and trees, to stands of large, and mostly deciduous trees with herbaceous understories (Shorrocks

2007: Figure 1.5). Heretofore, I use the term savanna to describe this complex mosaic of habitats. Abiotic factors shaping savanna vegetation include relatively high temperatures and extended periods of little or no rainfall, although additional factors such as soil fertility, fire, and faunal communities also affect savanna vegetation structure (Shorrocks 2007).

The species' present geographical distribution excludes relatively hotter, drier, and less- wooded habitats such as subdesert scrub and desert (Lehmann and Dunbar 2009; McGrew et al.

1981). Therefore, savannas are the hottest, driest, and most open environments that support chimpanzee populations. Chimpanzees occupy a range of vegetation types, from savanna mosaics to dense evergreen forests. There is some ambiguity over classifying the vegetation types for long-term chimpanzee study sites. For instance, Gombe and Mahale have been described as open woodland-savanna sites (Boesch 2009; Boesch and Boesch 1989). In contrast,

Stanford (1996) writes that “Gombe is not a savanna and Gombe chimpanzees are primarily forest chimpanzees. [….] the distinction between rainforest chimpanzees and “savanna chimpanzees” [Tai (forest) versus Mahale and Gombe (savanna) by Boesch and Boesch] is not an accurate characterization of these two populations.” In support of Stanford’s claim, Nishida

(2012) describes a similar forest habitat for his long-term studies on the M and K chimpanzee communities at Mahale National Park.

To evaluate these competing claims, I reviewed climate and vegetation for each long- term study site shown in Figure 1.1. I compared precipitation, precipitation seasonality, annual mean temperature, temperature seasonality, forest cover, and vegetation classes among the eight sites (Table 2.1). Climate data from the WorldClim database are extrapolations from long-term 44

(1950-2000) weather stations (Hijmans et al. 2005) were used in this analysis because comparable records from each chimpanzee site were not available. Temperature and precipitation

‘bioclimate’ factors (Hijmans et al. 2005) were selected because of their established associations with vegetation structure (Shorrocks 2007). Percent tree cover was collected from the Moderate

Resolution Imaging Spectroradiometer (MODIS) sensor and the vegetation continuous fields database (DiMiceli et al. 2011). I placed a 5 km buffer around each site location and randomly selected 50 points within this buffer to generate an average and standard deviation for percent tree cover at each study site (Table 2.1). Although method may include tree cover outside of chimpanzee home ranges at each site, water bodies were excluded. The final environmental descriptor of these sites was vegetation class, listed as either predominantly ‘savanna’ or ‘forest’, based on descriptions in the literature (Table 2.1).

Mahale had the lowest mean annual precipitation, followed by Fongoli and Gombe, but

Fongoli had the highest mean annual temperature (Figure 2.1, Table 2.1). Moreover, Fongoli was extremely variable in precipitation and temperature relative to all other study sites (Figure 2.2).

In addition to being very seasonal and hot, Fongoli had the least amount of tree cover (Figure

2.3, Table 2.1). In terms of evergreen tree cover, findings from pedestrian surveys converge on an estimate of 2-4% in southeastern Senegal (Bogart and Pruetz 2011; McGrew et al. 1981;

Pruetz and Bertolani 2009). Higher evergreen forest cover at Gombe and Mahale is likely related to the mountainous terrain and geographic proximity to Lake Tanganyika. For instance, Lake

Tanganyika and the Mahale Mountains form an orographic lift, leading to lower precipitation variation and a large swath of dense forest (Nishida 2012). Although annual precipitation at

Mahale and Gombe is slightly lower than Fongoli, precipitation seasonality is higher at the latter site; this indicates that water is more limited at Fongoli on a seasonal basis. If we are to classify 45

Mahale and Gombe as savanna sites, then additionally we should recognize the extreme savanna climate and vegetation at Fongoli. Furthermore, classifying chimpanzees as savanna-type or forest-type neglects variation within and among sites that may be important to the behavior and ecology of chimpanzees. Thus, diligence in defining and using these terms is encouraged.

Savanna chimpanzees across Africa Anthropological primatologists have sought to study chimpanzees in extremely hot, dry and open savanna environments since the 1960’s because of the long-standing idea that savannas played a major role in the evolution of bipedalism in humans (reviewed in Nishida 2012). This

‘savanna hypothesis’ (Dominguez-Rodrigo 2014 and references therein), combined with the referential modeling approach to reconstruct the behaviors of extinct hominins and the LCA

(McGrew 2010; Moore 1996) motivated several research teams to establish study sites in extremely hot and open savannas with the goal of habituating study subjects for long-term study

(Hunt and McGrew 2002; Nishida 2012). Locations for such research programs are listed in

Table 2.2. The Fongoli study is unique among these programs in that it is the only ‘extreme’ savanna site where habituated chimpanzees are studied. Habituation is a tradition in anthropological primatology because this approach enables researchers to collect detailed observations of behavior (Altmann 1974). While ethical considerations about habituation need to be carefully weighed (reviewed in Gruen et al. 2013), the data gleaned from protected, habituated study subjects are a rich in detail; thus, this approach is widely practiced and observations of habituated individuals form the empirical core of knowledge on chimpanzees

(Chapter 1). Ongoing habituation efforts at Toro-Semliki (Payne et al. 2008) indicate that in the not-too-distant-future it will be possible to make comparisons in behavior among more savanna chimpanzee communities. 46

Behavioral studies of unhabituated savanna chimpanzees have traditionally relied on indirect observations gleaned from sleeping nests, feeding remains, hair, tracks, or dung, as well as brief encounters with chimpanzees (diet: McGrew et al. 1988; ranging: Baldwin et al. 1982; sleeping-site selection: Hernandez-Aguilar et al. 2013; Ndiaye et al. 2013; Samson and Hunt

2012; Stewart et al. 2011; social organization: McGrew et al. 2004; Tutin et al. 1983; tool use:

Hernandez-Aguilar et al. 2007; McGrew et al. 2003). In addition, technological advancements have enabled researchers to ask increasingly sophisticated questions about unhabituated study subjects. For instance, camera traps are increasingly used to estimate population structure and habitat use (Boyer-Ontl and Pruetz 2014). Although the base of knowledge on chimpanzees occupying extremely open, hot and dry savannas has considerably increased in the past decade

(Table 2.2), there has been significantly more research on chimpanzees occupying wetter, cooler and more forested environments. This gap can be quantified through comparing the search results of peer-reviewed journal articles. I used the Web of ScienceTM database in December

2014 to compare article frequencies with savanna chimpanzee locations listed in Table 2.2

[search parameters: (pan troglodytes AND (assirik OR bafing OR dindefelo OR faleme OR fongoli OR kasakati OR semliki OR ugalla))] relative to the long-term study sites given in Table

2.1 [search parameters: (pan troglodytes AND (bossou OR budongo OR gombe OR kanyawara

OR mahale OR ngogo OR tai)]. Of the 1,010 located records, 9% of articles (n=86) involved our group of ‘savanna’ chimpanzees while 91% (n=924) could be attributed to ‘forest’ chimpanzees.

This disparity in knowledge may have important implications for our general understanding of the species, as well as for their conservation and use as referential models for human evolution studies. 47

Twenty years have passed since Strier (1994) published a benchmark paper dispelling the

“myth of the typical primate”, or the misconception that most primates conform to a model of social organization where males disperse, females have close bonds maintained through kinship, and social relationships are hierarchial. Instead, field studies on habituated primates have demonstrated a diverse array of social systems among primates (Strier 1994: Figure 2). Since then, a major vein of anthropological primatology has been devoted to documenting behavioral variation within and among primate species (reviewed in Jones 2005). Chimpanzee researchers have also contributed to this trend (Boesch et al. 1994; Whiten et al. 1999), with Boesch (2009:7) emphasizing that chimpanzees are characterized by this high behavioral diversity:

Foremost, we have learned that chimpanzees are much more diverse than originally thought, to the point where it is becoming more and more arbitrary to talk about ‘the chimpanzee’. We should describe them rather as the ‘Taï chimpanzee’ or the ‘Gombe chimpanzee’, just as we talk about the Inuits, the Touareg or the !Kung Bushmen in recognizing the diversity that exists in these traditional hunter-gatherer societies. Sociality of the sexes, vocalization, tool use, hunting behaviours, diet and cultural traits, all have been shown to differ profoundly between different chimpanzee populations.

Documenting behavioral diversity has been integral to savanna chimpanzee studies. Although research on open-country chimpanzees confirmed the presence of several species-typical behaviors, such as sleeping nest construction (reviewed in McGrew 2010), important differences have been reported as well.

Several authors have posited that ‘extreme’ savanna environments are more stressful for chimpanzees than relatively cooler, wetter and more forested regions (Baldwin et al. 1982;

McGrew et al. 1981; Moore 1996; Pruetz and Bertolani 2009; but see Kortlandt 1983). These chimpanzees adjust their foraging and ranging strategies to cope with seasonal resource scarcity, particularly concerning surface water (Pruetz and Bertolani 2009) and sleeping trees (Ndiaye et al. 2013) during dry seasons. In addition, food abundance (e.g., biomass kg ha-1) may be lower in 48 savannas environments. Extreme savanna chimpanzees have a lower diversity of plant and animal food species (Pruetz 2006; Watts et al. 2012) and, therefore, have fewer real and potential food types on the menu. In addition, chimpanzees in extreme savannas feed primarily on C3 plants (Sponheimer et al. 2006), but these plants are more widely scattered in savannas relative to forests (Shorrocks 2007) and the majority of plant food species are found in woodland habitats, rather than grasslands or forests (Fongoli: Pruetz 2006). These lines of evidence suggest that foods are less abundant in savanna environments. However, no study has compared biomass estimates between forest and savanna sites and, thus, support for this hypothesis is limited to indirect evidence. With this in mind, chimpanzee foraging behavior at several hot, dry, and open sites supports the hypothesis that plant and animal foods are less abundant there.

Chimpanzees living in extreme savannas appear to use extractive foraging techniques to overcome seasonal bottlenecks in food and water availability. Chimpanzees at Fongoli use spear- like tools to hunt small mammals (Pruetz and Bertolani 2007), while digging tools and feeding remains at Ugalla, Tanzania indicate that chimpanzees there excavate and consume geophytes

(Hernandez-Aguilar et al. 2007). Moreover, savanna chimpanzees at Fongoli, Assirik, and Toro-

Semliki, Uganda dig shallow ‘wells’ in streambeds by hand to access or filter drinking water

(Galat et al. 2008; Pruetz, unpublished data; Hunt and McGrew 2002). These unusual behaviors are not entirely unique to savanna chimpanzees, as forest-dwelling chimpanzees in Togo excavate wild tubers by hand for water (Lanjouw 2002), and Bossou chimpanzees raid cassava

(Manihot esculenta) crops (Hockings et al. 2010). In these instances, water and food scarcity, respectively, were associated with geophyte foraging. Furthermore, tool-assisted hunting of small mammals has been recorded a few times at Mahale, but whether or not this hunting strategy is associated with food availability is unknown (Nakamura and Itoh 2008). In addition, 49 tool-assisted hunting is rare at Mahale and common at Fongoli (Pruetz et al., in prep). Although chimpanzees living in environments other than extreme savannas also encounter food scarcity and can respond with extractive foraging strategies (Yamakoshi 1998), savanna chimpanzees are unified in that individuals face periods of resource scarcity that are associated with low annual precipitation, precipitation seasonality, and low forest cover, which result in major differences in the abundance and distribution of water, and probably food as well. Savanna resource scarcity likely explains these unusual extractive foraging behaviors, but more research is needed to elucidate these relationships.

Thermoregulatory behaviors also appear to distinguish savanna chimpanzees from conspecifics in more forested environments. Such activities include resting in caves, which are cooler than ambient temperatures in surrounding habitats (Pruetz 2007), soaking in water holes, foraging during moon-lit nights, and resting more during diurnal hours (Pruetz and Bertolani

2009). These behaviors likely minimize individuals’ energy expenditure or prevent heat stress.

Evidence that extreme savannas are marginal for chimpanzees also may be found in ranging behaviors and population densities (individuals km-2). If food and water is indeed more widely scattered and scarce in these environments, we can predict that chimpanzees must range over larger areas in search of food. And, if larger home ranges are required to meet individuals’ nutritive needs, then we can also expect to see that population densities are lower in savannas.

Trends in home range areas and population density estimates match these expectations. The

Fongoli chimpanzee community has a home range estimate of 85 km2 (Skinner and Pruetz 2012) and average community size of 35 individuals (Pruetz and Bertolani 2009). In relation to chimpanzees at Budongo, Gombe, Kibale, Mahale, and Tai (Morgan et al. 2006: Table 5 and references therein), the Fongoli community has the largest home range area (Average = 24 km2, 50

SD = 23) and lowest population density at 0.4 individuals km-2 (Average = 2.3 individuals km-2,

SD = 1.44).

Converging lines of evidence indicate that, all else being equal, savanna chimpanzees cope with intense levels of resource scarcity, namely periodic shortages in water and sleeping trees as well as widely scattered foods. This evidence strongly suggests that extreme savannas are marginal habitats for chimpanzees. However, due to the historic challenges of studying savanna chimpanzees across Africa, we have limited knowledge about how these resource limitation processes impact their ecology and behavior.

Savanna chimpanzees in Senegal All chimpanzees in Senegal belong to the West African subspecies (P. t. verus).

Senegalese chimpanzees range in the southeastern section of the country, which falls within the political boundaries of the Kedougou and Tambacounda regions. It is estimated that between 300 and 500 chimpanzees occupy Senegal (Carter et al. 2003; Ndiaye 2011).

Senegal's chimpanzees occupy savanna mosaic environments within a Sudano-Guinean vegetative zone. This is a transitional zone between the Sudanian and Guinean vegetation belts

(Aubréville 1950). The long axis of each belt runs from east to west, with no major orographic features interrupting each belt, whereas vegetation gradually changes along a latitudinal gradient and is correlated with duration of the dry season and annual precipitation (Gautier and Spichiger

2004). The Sudanian vegetation belt is positioned north of the Guinean belt. Guinean vegetation has more evergreen and liane species, and is characterized by a relatively higher diversity of plant species, whereas Sudanian vegetation has more deciduous and fire-adapted species. The transitional zone reflects a gradual and interdigitated merging of these two belts. In addition to climate, the savanna mosaic is shaped by cultivation, fire, soil conditions, and grazing (Gautier and Spichiger 2004; Tappan et al. 2004). 51

Senegalese chimpanzee research initiated in the 1960-70s with surveys and natural history observations at Niokolo-Koba National Park (Brewer 1978; De Bournonville 1967).

Chimpanzees within Niokolo-Koba, especially in the Assirik area, were the subject of short- term, intermittent research from 1976 to 2012 (Table 2.2). The Sterling Assirik Primate Project

(SAPP) was the most prolific research group during this period (1976-1979). The SAPP team demonstrated that savanna chimpanzees have relatively large home ranges compared to forest dwelling apes and that they rely extensively on small patches of gallery forest (Baldwin et al.

1982). Prior to the new millennia, there were few studies of chimpanzees beyond the park’s borders (Ndiaye 1999). Research outside of Niokolo-Koba has significantly increased in recent years (reviewed in Ndiaye 2011), mostly as a result of successes at the Fongoli Savanna

Chimpanzee Project (Table 2.2). Although the recent growth is savanna chimpanzee research is encouraging, many gaps in our knowledge remain. Thus, the following sections on food and habitat selection highlight important areas for new research.

Food selection in savanna environments In many ways, foods for ‘extreme’ savanna chimpanzees follow species' typical patterns.

Ripe fruit features prominently in their diet and the contributions of other food items, such as pith, bark, and meat fall within the range of variation for other study communities (Pruetz 2006).

However, there are important differences too. Diet diversity is much lower (Hunt and McGrew

2002; Pruetz 2006), which may be related to lower C3 plant species richness and a higher prevalence of low-quality foods, such as cellulose- and silica-laden grasses, leaves with thick layers of wax on the plant cuticle to minimize water loss, and plants with toxic levels of secondary compounds to prevent herbivory or disease (Hohmann et al. 2010). Given that savanna chimpanzees appear to have fewer palatable foods on the menu, unusual responses to these foraging challenges might be expected. 52

At Fongoli, the only extreme savanna site where we can examine food selection behavior in habituated chimpanzees, three unusual foraging behaviors have recently come to light. First,

Fongoli chimpanzees commonly hunt mammalian prey (bushbabies: Galago senegalensis) with stick tools (Pruetz and Bertolani 2007; Pruetz et al., in prep). Prior to this discovery, it was widely believed that only humans routinely hunted other animals with weapons (Pruetz and

Bertolani 2007). It has been hypothesized that this form of ‘spear’ use arose as a strategy to increase hunting success in an environment where monkeys, the most common mammalian prey, were less abundant or more difficult to capture (Pruetz and Bertolani 2007). Second, Fongoli chimpanzees invest much more time foraging for termites than other study communities with habituated study subjects (Bogart and Pruetz 2011), providing strong evidence that termites form a relatively larger proportion of nutrient intake. Third, individuals commonly re-ingest seeds on a seasonal basis (Bertolani and Pruetz 2011). Seed reingestion describes the process where an individual defecates, usually into their hand, extracts intact seeds from the fecal material, removes the softened seed coat, and ingests the seed kernel. Although seed reingestion behavior occurs elsewhere (Gombe: Goodall 1986; Semliki-Toro: Payne et al. 2008), Fongoli chimpanzees appear to reingest seeds more often than any other habituated chimpanzee study community. Each of these behaviors, intensive tool-assisted hunting, termitivory and seed- reingestion, is thought to be linked to resource availability and a maximizing energy return rate strategy. However, this claim lacks direct support in the form of quantitative estimates of the nutritional quality of foods as well as energy return rates for dietary items. The next step is to examine the nutritional context in association with these unusual foraging behaviors. This study specifically addresses the nutritional importance of insects (Macrotermes subhyalinus) and seed reingestion (mature seeds of Adansonia digitata) (Chapter 3). 53

Habitat selection in savanna environments Pan troglodytes occupies grasslands to some degree; however, open-country chimpanzees depend on medium to large sized trees, primarily found in woodland and evergreen forest habitats, for food, sleeping sites, water and shade (McGrew et al. 1988; McGrew 2010).

Shade as a high-value resource is probably not conspicuous for forest-dwelling chimpanzee communities because it is abundant in these habitats; however, cover from solar radiation can be rare in savanna landscapes. For instance, while shade and cloud cover is abundant during the wet season at Fongoli (Pruetz, unpublished data), there are few places to block exposure to solar radiation during the lengthy dry season.

Savanna chimpanzees may flexibly respond to seasonal differences in shade cover by shifting activity levels and habitat use. Taking refuge from the heat in shady locations and minimizing energy expenditure during the hottest times of day appears to be a thermoregulatory strategy of Fongoli chimpanzees (Pruetz and Bertolani 2009). Minimizing energy expenditure in this way may prevent heat stress and dehydration. There is no evidence to suggest that they possess physiological adaptations to water conservation that can be found in several arid-adapted mammals, such as highly concentrated urine (desert rodents: Pannabecker 2013) or dry feces

(springhares: Peinke and Brown 1999). Although we do not know how seasonality influences water balance in chimpanzees, studies of other nonhuman primates indicates that water scarcity has been a selective pressure in primate evolution. For instance, grey mouse lemurs (Microcebus murinus) increase water efficiency in drier habitats by going into daily torpor (Schmid and

Speakman 2009), and water efficiency appears to vary among closely related baboon species

(Papio hamadryas and P. anubis) in arid environments (Moritz et al. 2012).

Chimpanzees physiologically respond to thermally stressful conditions by increasing body temperature, sweating and panting (Elmadjian 1963; Hiley 1976). Chimpanzees are similar 54 to humans in distribution of apocrine and eccrine sweat glands (Montagna and Yun 1963), but it is not known how sweat gland densities compare (Sandel 2013). Although there are no apparent physiological adaptations to water conservation in savanna versus forest chimpanzees, this problem has yet to be evaluated. Seeking shade as a means to thermoregulate is a common strategy among mammals in arid environments (Cain et al. 2008; Hetem et al. 2012) and captive chimpanzees (Duncan and Pillay 2013). Furthermore, forest dwelling chimpanzees also exhibit thermoregulatory behaviors, namely resting and spending more time on the ground during peak daily temperatures (Kosheleff and Anderson 2009). Meanwhile, Fongoli chimpanzees differ in that they appear to spend more time resting and less time feeding during daylight hours, especially during the dry season, than chimpanzees at Gombe and Kibale, locations with more forest cover (Pruetz and Bertolani 2009). Thus, it is likely that ‘extreme’ savanna chimpanzee are under greater pressure to minimize water loss and risk of heat stress. Although shade-seeking behavior largely explains the preference for forest habitats at Fongoli, water availability also plays a role during the dry season, as evergreen trees oftentimes signify the presence of free water at springs and rivers (Pruetz and Bertolani 2009).

Savanna chimpanzees may exhibit stronger seasonal patterns in habitat use in relation to their forest counterparts. Although food seasonality, especially ripe fruit availability, impacts habitat use and ranging behavior for chimpanzees in general (Doran 1997; Nishida 2012;

Tweheyo et al. 2004), seasonal differences in habitat use is likely to be striking between wet and dry seasons for savanna chimpanzees, as these communities must additionally cope with variation in water and shade resource availability. Behavioral observations at Fongoli support this claim: during dry season months, individuals foraged more often in the early morning hours, when temperature was relatively cool; moreover, they spent much less time in forested habitats, 55 and more time in woodland habitats, during wet season months (Pruetz and Bertolani 2009).

These patterns indicate that Fongoli chimpanzees are released from heat stress and dehydration constraints during the wet months when temperatures are cooler and water is abundant within their home range. Moreover, Pruetz (2006) found that the vast majority of plant food species at

Fongoli were located in woodland rather than evergreen forest, which indicates that Fongoli chimpanzees rely on woodland habitat for sustenance. Furthermore, some grasses, climbers and shrubs located in grassland habitats are important food sources, especially during the wet season.

Savanna chimpanzees may exhibit shifting habitat preferences within savanna mosaic environments that reflect resource distribution and abundance dynamics both within and between seasons. However, these observations are based on the amount of time that individuals feed and travel within each habitat. Although an activity budget approach (Pruetz and Bertolani 2009) is a powerful tool for sampling behavior, time is not always representative of nutrient intake: foods can substantially differ in energy return rates due to macronutrient composition (Emerson and

Brown 2012), antifeedant properties (Rothman et al. 2008) and degree of difficulty processing food with jaws, hands and, in some cases, tools (O'Malley and Power 2014). Although it is difficult to evaluate differences in foraging behavior among habitats with the time budget method, approaches that consider energy return rates among habitat types can remedy this problem (Brown 1988; Stephens et al. 2007) (see Chapter 4).

Summary This chapter deconstructed the significance of the savanna and forest labels that we apply to wild chimpanzee communities, and explicated the complexities and assumptions underlying these descriptive terms. The analysis of forest cover and climate factors showed that chimpanzee habitats range along a savanna to forest continuum, instead of a savanna/forest dichotomy. In addition, this review highlights the rarity of savanna-dwelling chimpanzees as well as our 56 understanding of their behavior and ecology as presented in the primary literature. The subsequent chapters address some of these gaps in relation to food selection (Chapter 3), habitat use (Chapter 4), and how these factors may interact to define the species’ distribution in Senegal

(Chapter 5).

Literature cited Altmann J. 1974. Observational study of behavior - sampling methods. Behaviour 49(3-4):227-

267.

Aubréville A. 1950. Flore forestière Soudane-Guinéenne. Paris: Société d'éditions

Géographiques, Maritimes et Colonials.

Baldwin PJ, McGrew WC, and Tutin CEG. 1982. Wide-ranging chimpanzees at Mt Assirik,

Senegal. International Journal of Primatology 3(4):367-385.

Bertolani P, and Pruetz J. 2011. Seed reingestion in savannah chimpanzees (Pan troglodytes

verus) at Fongoli, Senegal. International Journal of Primatology:1-10.

Boesch C. 2009. The real chimpanzee. New York: Cambridge University Press. 196 p.

Boesch C, and Boesch C. 1989. Hunting behavior of wild chimpanzees in the Taї National Park.

American Journal of Physical Anthropology 78:547-573.

Boesch C, and Boesch-Achermann H. 2000. The chimpanzees of the Taї forest. Oxford: Oxford

University Press. 316 p.

Boesch C, Marchesi P, Marchesi N, Fruth B, and Frédéric J. 1994. Is nut cracking in wild

chimpanzees a cultural behaviour? Journal of Human Evolution 26:325-338.

Bogart SL, and Pruetz JD. 2011. Insectivory of savanna chimpanzees (Pan troglodytes verus) at

Fongoli, Senegal. American Journal of Physical Anthropology 145(1):11-20.

Bourlière F. 1985. Primate communities: Their structure and role in tropical ecosystems.

International Journal of Primatology 6(1):1-26. 57

Boyer-Ontl K, and Pruetz J. 2014. Giving the forest eyes: the benefits of using camera traps to

study unhabituated chimpanzees (Pan troglodytes verus) in Southeastern Senegal.

International Journal of Primatology 35(5):881-894.

Brewer S. 1978. The chimpanzees of Mt. Assirik. New York: Knopf.

Brown JS. 1988. Patch use as an indicator of habitat preference, predation risk, and competition.

Behavioral Ecology and Sociobiology 22(1):37-47.

Cain JW, Jansen BD, Wilson RR, and Krausman PR. 2008. Potential thermoregulatory

advantages of shade use by desert bighorn sheep. Journal of Arid Environments

72(8):1518-1525.

Carter J, Ndiaye S, Pruetz JD, and McGrew WC. 2003. Senegal. In: Kormos R, Boesch C,

Bakarr M, and Butynski TM, editors. West African chimpanzees. Gland: World

Conservation Union. p 31-39.

De Bournonville D. 1967. Contribution a l'etude du chimpanzee en Republique de Guinee.

Bulletin de l'Institute Fondamental d'Afrique Noire. p 1188-1269.

DiMiceli C, Carroll M, Sohlberg R, Huang C, Hansen M, and Townshend J. 2011. Annual

Global Automated MODIS Vegetation Continuous Fields (MOD44B) at 250 m Spatial

Resolution for Data Year 2010, Collection 5 Percent Tree Cover.

Dominguez-Rodrigo M. 2014. Is the "savanna hypothesis" a dead concept for explaining the

emergence of the earliest hominins? Current Anthropology 55(1):59-81.

Doran D. 1997. Influence of seasonality on activity patterns, feeding behavior, ranging, and

grouping patterns in Tai chimpanzees. International Journal of Primatology 18(2):183-

206. 58

Duncan LM, and Pillay N. 2013. Shade as a thermoregulatory resource for captive chimpanzees.

Journal of Thermal Biology 38(4):169-177.

Elmadjian F. 1963. Biochemical study of large primate response to severe environmental

stressors. Holloman Air Force Base, New Mexico: 6571st Aeromedical Research

Laboratory.

Emerson SE, and Brown JS. 2012. Using giving-up densities to test for dietary preferences in

primates: an example with samango monkeys (Cercopithecus (nictitans) mitis

erythrarchus). International Journal of Primatology 33(6):1420-1438.

Galat G, Galat-Luong A, and Nizinski G. 2008. Our cousins chimpanzees and baboons face

global warming by digging wells to filtrate drinking water. International Water Resources

Association World Water Congress. Montpellier, France.

Gautier L, and Spichiger R. 2004. The forest-savanna transition in West Africa. In: Poorter L,

Bongers F, Kouamé F, and Hawthorne W, editors. Biodiversity of West African forests:

an ecological atlas of woody plant species. Cambridge, MA: CABI Publishing. p 33-40.

Goodall J. 1986. The chimpanzees of Gombe. Cambridge, MA: The Belknap Press of Harvard

University Press. 674 p.

Gruen L, Fultz A, and Pruetz J. 2013. Ethical issues in african great ape field studies. ILAR

Journal 54(1):24-32.

Hernandez-Aguilar RA, Moore J, and Pickering TR. 2007. Savanna chimpanzees use tools to

harvest the underground storage organs of plants. Proceedings of the National Academy

of Sciences 104(49):19210-19213. 59

Hernandez-Aguilar RA, Moore J, and Stanford CB. 2013. Chimpanzee nesting patterns in

savanna habitat: Environmental influences and preferences. American Journal of

Primatology 75(10):979-994.

Hetem RS, Strauss WM, Fick LG, Maloney SK, Meyer LCR, Shobrak M, Fuller A, and Mitchell

D. 2012. Activity re-assignment and microclimate selection of free-living Arabian oryx:

responses that could minimise the effects of climate change on homeostasis? Zoology

115(6):411-416.

Hijmans RJ, Cameron SE, Parra JL, Jones PG, and Jarvis A. 2005. Very high resolution

interpolated climate surfaces for global land areas. International Journal of Climatology

25:1965-1978.

Hiley PG. 1976. Thermoregulatory responses of galago (Galago crassicaudatus), baboon (Papio

cynocephalus) and chimpanzee (Pan satyrus) to heat stress. Journal of Physiology-

London 254(3):657-671.

Hockings KJ, Anderson JR, and Matsuzawa T. 2010. Flexible feeding on cultivated underground

storage organs by rainforest-dwelling chimpanzees at Bossou, West Africa. Journal of

Human Evolution 58(3):227-233.

Hohmann G, Potts K, Antoine N'Guessan A, Andrew Fowler A, Roger Mundry M, Joerg U.

Ganzhorn J, and Sylvia Ortmann S. 2010. Plant foods consumed by Pan: exploring the

variation of nutritional ecology across Africa. American Journal of Physical

Anthropology 141(3):476-485.

Hunt KD, and McGrew WC. 2002. Chimpanzees in the dry habitats of Assirik, Senegal and

Semliki Wildlife Reserve, Uganda. In: Boesch C, Hohmann G, and Marchant LF, editors. 60

Behavioural diversity in chimpanzees and bonobos. Cambridge: Cambridge University

Press. p 35-51.

Jones C. 2005. Behavioral flexibility in primates: causes and consequences. Chicago, IL:

University of Chicago.

Junker J, Blake S, Boesch C, Campbell G, Toit Ld, Duvall C, Ekobo A, Etoga G, Galat-Luong

A, Gamys J et al. . 2012. Recent decline in suitable environmental conditions for African

great apes. Diversity and Distributions 18(11):1077-1091.

Kortlandt A. 1983. Marginal habitats of chimpanzees. Journal of Human Evolution 12:231-278.

Kosheleff VP, and Anderson CNK. 2009. Temperature's influence on the activity budget,

terrestriality, and sun exposure of chimpanzees in the Budongo Forest, Uganda.

American Journal of Physical Anthropology 139(2):172-181.

Lanjouw A. 2002. Behavioural adaptations to water scarcity in Tongo chimpanzees. In: Boesch

C, Hohmann G, and Marchant LF, editors. Behavioural diversity in chimpanzees and

bonobos. Cambridge: Cambridge University Press. p 52-60.

Lehmann J, and Dunbar R. 2009. Implications of body mass and predation for ape social system

and biogeographical distribution. Oikos 118(3):379-390.

Matsuzawa T, Sugiyama Y, Humle T, eds. 2011. The chimpanzees of Bossou and Nimba.

Tokyo: Springer.

McGrew WC. 2010. In search of the last common ancestor: new findings on wild chimpanzees.

3267-3276 p.

McGrew WC, Baldwin P, and Tutin CEG. 1981. Chimpanzees in a hot, dry and open habitat: Mt.

Assirik, Senegal, West Africa. Journal of Human Evolution 10:217-244. 61

McGrew WC, Baldwin PJ, Marchant LF, Pruetz JD, Scott SE, and Tutin CEG. 2003.

Ethoarchaeology and elementary technology of unhabituated wild chimpanzees at

Assirik, Senegal, west Africa. Paleoanthropology 1:1-20.

McGrew WC, Baldwin PJ, and Tutin CEG. 1988. Diet of wild chimpanzees (Pan troglodytes

verus) at Mt. Assirik, Senegal: I. Composition. American Journal of Primatology

16(3):213-226.

McGrew WC, Ensminger AL, Marchant LF, Pruetz JD, and Vigilant L. 2004. Genotyping aids

field study of unhabituated wild chimpanzees. American Journal of Primatology

63(2):87-93.

Montagna W, and Yun JS. 1963. Skin of primates. XV. Skin of chimpanzee (Pan satyrus).

American Journal of Physical Anthropology 21(2):189-203.

Moore J. 1996. Savanna chimpanzees, referential models and the last common ancestor. In:

McGrew W, Marchant, LF, Nishida, T, editor. Great ape societies. Cambridge:

Cambridge University Press. p 275-292.

Morgan D, Sanz C, Onononga JR, and Strindberg S. 2006. Ape abundance and habitat use in the

Goualougo Triangle, Republic of Congo. International Journal of Primatology 27(1):147-

179.

Moritz GL, Fourie N, Yeakel JD, Phillips-Conroy JE, Jolly CJ, Koch PL, and Dominy NJ. 2012.

Baboons, water, and the ecology of oxygen stable isotopes in an arid hybrid zone.

Physiological and Biochemical Zoology 85(5):421-430.

Nakamura M, and Itoh N. 2008. Hunting with tools by Mahale chimpanzees. Pan Africa News

15(1):3-6. 62

Ndiaye P. 1999. Biogeographie et elements d'ecologie du chimpanzee, Pan troglodytes, au

Senegal. Dakar: Universitie Cheikh Anta Diop.

Ndiaye P, Galat G, Galat-Luong A, and Nizinski G. 2013. Note on the seasonal use of lowland

and highland habitats by the West African Chimpanzee Pan troglodytes verus (Schwarz,

1934) (Primates: Hominidae): Implications for its conservation. Journal of Threatened

Taxa 5(2):3697-3700.

Ndiaye S. 2011. Conservation du chimpanze au Senegal: etat des connaissances et reactualisation

du plan d'actions. USAID-Wula Nafaa.

Nishida T. 2012. Chimpanzees of the lakeshore: natural history and culture at Mahale.

Cambridge: Cambridge University Press.

O'Malley RC, and Power ML. 2014. The energetic and nutritional yields from insectivory for

Kasekela chimpanzees. Journal of Human Evolution 71:46-58.

Payne C, Webster T, and Hunt K. 2008. Coprophagy by the semi-habituated chimpanzees of

Semliki, Uganda. Pan Africa News 15(2).

Peinke DM, and Brown CR. 1999. Osmoregulation and water balance in the springhare (Pedetes

capensis). Journal of Comparative Physiology B-Biochemical Systemic and

Environmental Physiology 169(1):1-10.

Pruetz J. 2007. Evidence of cave use by savanna chimpanzees (Pan troglodytes verus) at

Fongoli, Senegal: implications for thermoregulatory behavior. Primates 48(4):316-319.

Pruetz JD. 2006. Feeding ecology of savanna chimpanzees (Pan troglodytes verus) at Fongoli,

Senegal. In: Hohmann G, Robbins MM, and Boesch C, editors. Feeding ecology in apes

and other primates. Cambridge, UK: Cambridge University Press. p 161-182. 63

Pruetz JD, and Bertolani P. 2007. Savanna chimpanzees, Pan troglodytes verus, hunt with tools.

Current Biology 17:1-6.

Pruetz JD, and Bertolani P. 2009. Chimpanzee (Pan troglodytes verus) behavioral responses to

stresses associated with living in a savanna-mosaic environment: implications for

hominin adaptations to open habitats. Paleoanthropology:252-262.

Reynolds V. 2005. The chimpanzees of the Budongo forest. Oxford: Oxford University Press.

Rothman JM, Chapman CA, and Pell AN. 2008. Fiber-bound nitrogen in gorilla diets:

implications for estimating dietary protein intake of primates. American Journal of

Primatology 70(7):690-694.

Samson DR, and Hunt KD. 2012. A thermodynamic comparison of arboreal and terrestrial

sleeping sites for dry-habitat chimpanzees (Pan troglodytes schweinfurthii) at the Toro-

Semliki Wildlife Reserve, Uganda. American Journal of Primatology 74(9):811-818.

Sandel AA. 2013. Brief communication: hair density and body mass in mammals and the

evolution of human hairlessness. American Journal of Physical Anthropology

152(1):145-150.

Schmid J, and Speakman JR. 2009. Torpor and energetic consequences in free-ranging grey

mouse lemurs (Microcebus murinus): a comparison of dry and wet forests.

Naturwissenschaften 96(5):609-620.

Shorrocks B. 2007. The biology of African savannahs. Oxford: Oxford University Press.

Skinner MF, and Pruetz JD. 2012. Reconstruction of periodicity of repetitive linear enamel

hypoplasia from perikymata counts on imbricational enamel among dry-adapted

chimpanzees (Pan troglodytes verus) from Fongoli, Senegal. American Journal of

Physical Anthropology 149(3):468-482. 64

Stephens DW, Brown JS, and Ydenberg RC, editors. 2007. Foraging: behavior and ecology.

Chicago: University of Chicago Press.

Stewart FA, Piel AK, and McGrew WC. 2011. Living archaeology: Artefacts of specific nest site

fidelity in wild chimpanzees. Journal of Human Evolution 61(4):388-395.

Strier K. 1994. Myth of the typical primate. Yearbook of Physical Anthropology 37:233-271.

Tappan GG, Sall M, Wood EC, and Cushing M. 2004. Ecoregions and land cover trends in

Senegal. Journal of Arid Environments 59(3):427-462.

Tutin CEG, McGrew WC, and Baldwin PJ. 1983. Social organization of savanna dwelling

chimpanzees Pan troglodytes verus at Mount Assirik Senegal. Primates 24(2):154-173.

Tweheyo M, Lye KA, and Weladji RB. 2004. Chimpanzee diet and habitat selection in the

Budongo Forest Reserve, Uganda. Forest Ecology and Management 188(1-3):267-278.

Watts DP, Potts KB, Lwanga JS, and Mitani JC. 2012. Diet of chimpanzees (Pan troglodytes

schweinfurthii) at Ngogo, Kibale National Park, Uganda, 1. Diet composition and

diversity. American Journal of Primatology 74(2):114-129.

Whiten A, Goodall J, McGrew WC, Nishida T, Reynolds V, Sugiyama Y, Tutin CEG,

Wrangham RW, and Boesch C. 1999. Cultures in chimpanzees. Nature 399(6737):682-

685.

Wrangham RW, Rogers ME, and I-Basuta G. 1993. Ape food density in the ground layer in

Kibale Forest, Uganda. African Journal of Ecology 31:49-57.

Yamakoshi G. 1998. Dietary responses to fruit scarcity of wild chimpanzees at Bossou, Guinea:

Possible implications for ecological importance of tool use. American Journal of Physical

Anthropology 106(3):283-295. 65

Figure 2.1. Relationship between annual precipitation (mm) and annual mean temperature (Cº) among long-term study sites with habituated chimpanzees. 68

Figure 2.2. Relationship between precipitation seasonality (CV of mean monthly precipitation) and temperature seasonality (SD of monthly mean temperature) among long-term study sites with habituated chimpanzees.

Figure 2.3. Relationship between percent tree cover and simple qualitative descriptions of habitat. Differences in the two rows correspond to conflicting, colloquial usage of ‘savanna’ and

‘forest’ terms. Boesch (2009) broadly classifies Mahale and Gombe as ‘savanna’ sites, contra

Stanford (1996) and Nishida (2012). The ‘forest’ site with the lowest tree cover is Bossou, which is likely related to a high degree of deforestation in this area (Matsuzawa et al. 2011). 70

CHAPTER III NUTRITIONAL AND SOCIAL COMPONENTS OF FORAGING STRATEGY IN WESTERN CHIMPANZEES (PAN TROGLODYTES VERUS) AT FONGOLI, SENEGAL Manuscript will be submitted to the American Journal of Primatology

STACY M. LINDSHIELD,1,2* JESSICA M. ROTHMAN,3,4 SYLVIA ORTMANN,5 AND JILL

D. PRUETZ1

1Department of Anthropology, Iowa State University, Ames, IA 50011

2Ecology and Evolutionary Biology Program, Iowa State University

3Department of Anthropology, Hunter College of the City University of New York, New York, NY

4New York Consortium in Evolutionary Primatology, New York, NY

5Leibniz Institute of Zoo and Wildlife Research, Berlin, Germany

Contract grant sponsor: The Leakey Foundation, The Rufford Small Grant Foundation

*Correspondence to:

S.M. Lindshield

324 Curtiss Hall

Department of Anthropology

Iowa State University

Ames, IA 50011

E-mail: [email protected]

Abstract Nutritional ecology is of central importance to understanding the social foraging strategies of chimpanzees (Pan troglodytes). However, limited data on the nutritional values of foods in the diets of wild chimpanzees, as well as estimates of daily energy and macronutrient intakes, have hindered efforts to evaluate food selection theory with empirical observations of chimpanzee foraging behavior. The aim of this study is to examine food selection behavior at the interface of diet quality and social relationships in adult male chimpanzees (P.t. verus) at Fongoli, Senegal.

We conducted focal animal follows on adult males between October 2011 to January 2012 (N=8 males, and September 2012 to February 2013 (N=11 males) to observe food processing behavior and estimate daily food intakes. We sampled foods from 34 plant and three insect species to estimate the nutritional composition of each food in terms of energy in kJ and macronutrients

(e.g., fat, protein, carbohydrates). From these data, we calculated food profitability scores, which approximate a caloric yield per unit of time (kJ hr-1) for each food. Highly profitable foods included fleshy fruits from Adansonia, Tamarindus, and Ficus trees. Young leaves and flowers scored slightly lower in profitability. Bamboo (Oxytenanthera abyssinica) pith and termites

(Macrotermes subhyalinus) were considerably lower in profitability but commonly eaten, perhaps as a strategy to balance nutrients or to form a dietary staple when consumed together.

Adult males preferred highly profitable foods when given the choice between two foods of differing values. In addition, higher-ranking adult males consumed more water-soluble carbohydrates each day, which indicates that they use their social status in the community to get priority access to sweet, ripe fruits. Collectively, these results demonstrate that profitability is one of many factors involved in food choice.

Key words: savanna chimpanzee; Senegal; nutritional ecology; energy intake; insectivory

Introduction Nutritional ecology is of central importance to understanding the social foraging strategies of great apes. While there is considerable interest in nutritional aspects of primate feeding ecology [Ganas et al., 2009; Glander, 1977; Heesen et al., 2013; Irwin et al., 2014;

Milton, 1981; Rothman et al., 2008], relatively few studies have quantitatively assessed the nutritional values of foods in the diets of wild chimpanzees (Pan troglodytes) [Conklin-Brittain et al., 1998; Deblauwe and Janssens, 2008; Hohmann and Fruth, 2008; O'Malley and Power,

2012; Reynolds et al., 1998; Takemoto, 2003; Wrangham et al., 1998], or quantified daily energy

(kJ) and macronutrient (protein, fat, carbohydrate) intakes [Conklin et al., 2006; N'guessan et al.,

2009]. This problem limits our ability to empirically test food selection theory [Raubenheimer et al., 2009; Stephens and Krebs, 1986] in chimpanzees. We address this issue through examining food selection behavior at the interface of diet quality and social relationships in adult male chimpanzees at Fongoli, Senegal.

The optimal diet framework is a commonly used approach to assess food selection behavior [Emlen, 1966; MacArthur and Pianka, 1966; Stephens et al., 2007; Stephens and Krebs,

1986]. The basic framework predicts that individuals select foods that maximize nutritional gains, such as energy in kJ, over time as a means to maximize survival and reproduction

[Stephens and Krebs, 1986]. Although additional factors are thought to influence food selection in chimpanzees, such as food abundance [Heller, 1980], plant toxins [reviewed in Felton et al.,

2009], and predation risk [Lindshield et al., in prep], higher survival rates are linked to food selection behaviors that maximize energy return rates [Papio cynocephalus: Altmann, 1998].

Therefore, it is likely that, in many cases, chimpanzees select foods based on “caloric yield per time” [Emlen 1966:611]. 73

Chimpanzees are omnivorous in that they consume plant and animal tissues [McGrew,

2014], but they specialize on ripe fruit when it is available [reviewed in Watts et al., 2012].

Although fruit is important for chimpanzees, supplementing this diet with alternative foods, like leafy material and animal prey, may provide essential nutrients that are rare or absent in fruits

[i.e., nutrient balancing, reviewed in Felton et al., 2009]. Chimpanzees likely target vertebrate prey for protein, fat, and essential vitamins (e.g., vitamin B12) [Tennie et al., 2009]. Insects (e.g., ants and termites) may also be targeted for fat and protein, but their high mineral content is thought to be particularly important [O'Malley and Power, 2014]. In addition to nutrient balancing, the expansion and contraction of dietary breadth is associated with ripe fruit abundance [Isabirye-Basuta, 1989]. When chimpanzees experience periods of ripe fruit scarcity, they tend to focus on lower calorie foods, such as pith and bark [i.e., “fallback” foods, Marshall et al., 2009]. Chimpanzees may also depend on relatively common foods with adequate macronutrient levels to meet daily requirements (i.e., “staples”), such as Ficus fruits [Conklin and Wrangham, 1994; Newton-Fisher, 1999] and incorporate ephemeral foods with higher energy return rates based on their abundance and distribution. While these dietary trends apply to many chimpanzee communities, a common theme from these dietary studies is that food quality

[Hohmann et al., 2010] and food availability [Newton-Fisher, 1999] varies among sites to a degree that makes generalizations about chimpanzee diets a challenge. For this reason, it is important to study the nutritional ecology of chimpanzees across a range of environments.

Arid, savanna environments are hypothesized to produce foods that differ in nutritional quality from wetter, cooler regions occupied by chimpanzees, due to potential differences in the production of plant secondary compounds that minimize herbivory [Hohmann et al., 2010] to the abundance and distribution of foods, with savanna environments having fewer, large fruiting 74 trees as well as few herbaceous food options during the dry season. This study is the first to document the nutritional values of chimpanzee foods from an arid, savanna environment. Thus, one objective is to compare Fongoli diet quality to estimates from Pan study sites in relatively wet and cool environments. The second objective of this study is to test the hypothesis that adult male chimpanzees at Fongoli use an energy maximization foraging strategy. To accomplish this goal, we evaluate their food choices in relation each food’s energy return rate. Moreover, we aim to identify potential staple or fallback foods at Fongoli that enable chimpanzees here to survive in an extreme, savanna environment when energy-rich fruits are relatively scarce.

While the nutritional value of food is likely a key determinant of food selection, additional factors, such as group social dynamics, may influence food choices. Accessibility to key foods or feeding sites can be influenced by an individual’s social status while they are foraging with other community members. For instance, dominant individuals may supplant subordinate individuals from feeding locations [Goodall, 1986] or range in areas with higher food availability and quality [Emery Thompson et al., 2007]. In these cases, subordinates can minimize agonistic interactions with higher-ranking individuals by avoiding them. However, individuals that experience nutritional losses by avoiding conflict may need to compensate for these losses with alternate foraging strategies, such as selecting feeding away sites from dominant individuals [Riedel et al., 2011]. Therefore, we expect that chimpanzees will exhibit flexible feeding strategies that balance nutritional gains with the social complexities of group living. In this vein, our final objective is to assess how daily energy and macronutrient intakes vary with male social rank, party size, and party composition.

Methods Study location and period

The Fongoli study site (12°39ʹ N, 12°13ʹ W) is situated in the Kedougou region of southeastern Senegal. The vegetation at Fongoli is a patchy network of grasslands and woodlands with small intrusions of gallery and ecotone forest [for detailed descriptions of the study site, see Pruetz, 2006; Pruetz and Bertolani, 2009; Lindshield et al., in prep]. Behavioral observations of West African chimpanzees (P. t. verus) and samples of their dietary items were collected during two study periods, October 2011 to January 2012 and September 2012 to

February 2013.

Study subjects and behavior sampling

The Fongoli chimpanzee community had 33 to 34 individuals during the study period, and this number fluctuated due to births, deaths, and migrations. We conducted systematic behavioral follows of adult males (2011-2012: N=10, 2012-2013: N=11), which were well- habituated to research personnel. All individuals were uniquely recognized by their physical appearances and the social ranks of adult males were often identified by patterns in submissive vocalizations [Goodall, 1986]. However, due to shifts within the social hierarchy, including a change in the alpha male position and the addition of three young males, this hierarchy was in flux during the study period. Thus, we categorized males as having a high, middle, or low social rank. Opportunistic data were collected on adult females, juveniles and infants [for more information on this protocol, see Pruetz and Kante, 2010; Lindshield et al., in prep].

The foraging behavior of adult males was the focus of behavioral observations. Foraging behavior was defined as the process of searching for, capturing, and consuming food [Stephens and Krebs, 1986]. Travel was included in this definition, except when individuals covered ground for non-foraging, social purposes, such as short distance movement while grooming, 76 displaying, or mating [Nishida et al., 2010]. We aimed to uniformly sample all males each month using traditional behavioral sampling methods [Altmann, 1974]. Each day, we continuously recorded the foraging behavior of one adult male. Our goal was to locate a focal male at dawn and sample his behavior until he constructed a nest at dusk. If the focal male was lost and could not be relocated within 20 minutes during this period, we switched to another focal male. Less than five percent of sampling days involved a change in focal subjects. When the focal individual was foraging, changes in behavior were estimated to the nearest second. In rare instances, activity bouts were rounded to the nearest 30 seconds. The act of consuming food was separately timed from food capturing in instances when consuming and capturing did not simultaneously occur. Pauses in feeding behavior that exceeded 10 seconds were removed from estimates of feeding duration. We identified food processing strategies through direct observations of feeding as well as inspections of feeding and fecal remains. In addition, we recorded the food part(s) consumed (e.g., stem pith, mesocarp) and phenological state of plant food items (e.g., unripe, semi-ripe or ripe fruit). Whenever possible, we recorded the number of food items consumed per feeding patch visit. We used feeding rates, defined as the number of food items (e.g., fruit, inflorescence) consumed per unit of time, to estimate food intake during a feeding bout in cases when the focal male was out of view for a portion of that bout.

To gauge social components of food selection [Goodall, 1986], we monitored focal male foraging behavior in relation to foraging party size and composition, as well as female reproductive state and male social ranking. We recorded a total daily party size estimate, defined as all individuals seen together on a single day [Boesch, 1996; Furuichi, 2009]. In addition, we recorded the identities of all individuals within our view at 30 minute intervals. We estimated the estrus/anestrus state of females each day by ranking the swelling size of females’ anogenital 77 region on a four-point scale, where a score of zero is completely flaccid and three is maximally tumescent. Finally, we recorded all instances of copulations, agonistic interactions involving the focal male and cases of social rank reversals (i.e., dyadic changes in pant-grunt direction).

Food sampling

Fresh food samples were collected during the study period for nutritional analyses. Our sampling goal was to collect 20-30 g of dried sample per food type for this initial survey of diet quality. In addition, we extensively sampled ripe fruit (N=105) of the baobab tree because this item was a top food for six of the 14 months represented in this study period [see Lindshield et al., in prep]. For all other foods, one or two samples represented each food item. Each food sample was collected within a few days of observing a focal subject consuming that food.

Samples were collected from the same feeding plant, unless there was insufficient food material remaining at that source. In the latter case, we sampled from as many different individual plants as possible in order to capture intra-food variation [Chapman et al., 2003]. Foods were sampled in the same manner as they are eaten by focal subjects, that is, from the same phenological stage.

Ants and termites were collected at nests and placed in frozen storage within two hours of collection. Honey was sampled by a skilled collector without the aid of smoke or fire. In addition, plant parts and social insects not selected by Fongoli chimpanzees were also sampled to compare the nutrient quality of potential and real foods at Fongoli.

Fresh samples were immediately brought back to camp for measuring and drying. The entire and edible portion of foods were measured for wet and field-dried mass in grams. In addition, we measured the length and circumference or width of food parts [Ortmann et al.,

2006]. Samples were field dried in a solar dehydrator or gas oven (Coleman® camp oven) out of direct sunlight at a temperature range of 40 to 50° C until reaching a constant weight [Conklin- 78

Brittain et al., 1998]. After drying, the samples were stored in dark, air-tight containers with a food-safe desiccant. Mold or insect damaged samples were discarded.

Food availability

We used a belt transect (10-m width, 3.4-km long) to monitor food trees each month for the phenological states of fruits (ripe, unripe), flowers (bud, bloom), and leaves (young, mature, old). In addition, we assigned simple, qualitative abundance and spatial distribution scores to each food species, where each food was classified as scarce, abundant, or hyperabundant, and clumped or scattered (Pruetz 2009)

Nutritional chemistry

Standard wet-chemistry methods in primate nutritional ecology were used to estimate macronutrients and antifeedants in food items [Ortmann et al., 2006; Rothman et al., 2012]. Food samples were milled with a Wiley® Mini-Mill using a number 20 delivery unit (0.85 mm mesh screen) at the Grain Quality Laboratory at Iowa State University (ISU) in Ames, Iowa, USA. A portion of the samples were analyzed at the Leibniz Institute for Zoo and Wildlife Research

(IZW) in Berlin, Germany (N=13) and the remaining samples were analyzed at the CUNY-

Hunter College Primate Nutritional Ecology Laboratory in New York City, New York, USA.

Dry matter was estimated by determining moisture loss from the wet and dry mass field measurements combined with oven drying samples at 105° ± 2° C overnight in the laboratory

[Rothman et al., 2012]. We used the detergent fiber approach was used to estimate the fibrous components of food [Van Soest 1994]. Neutral detergent fiber represents the fraction of total structural carbohydrates (hemicellulose, cellulose, lignin) as well as silica, tannins, and cutins in a food. Acid detergent fiber estimates the fraction of NDF that is cellulose, lignin, silica, and acid-insoluble nitrogen, while acid detergent lignin represents the ligneous fraction. Crude fat 79 was estimated by using an ether extraction method [AOCS, 2005]. Crude protein and acid detergent insoluble protein were estimated with the dumas combustion method, where the nitrogen estimate is multiplied by 6.25 [Licitra et al., 1996]. Samples were ashed in a muffle furnace overnight at 550° C as a crude estimate of inorganic matter [Rothman et al., 2012] at

IZW or the Swine and Poultry Nutrition Lab at ISU. Total nonstructural carbohydrates were estimated through subtracting the sum of neutral detergent fiber (NDF), crude protein, crude fat and crude ash from 100 [Goering and Van Soest, 1970]. We calculated energy content in kilojoules per gram (Equation 1) based on the expected energy content of protein, carbohydrate, and fat [Conklin and Wrangham, 1994; Heesen et al., 2013], including the estimated fraction of

NDF converted to energy through hind-gut fermentation [54.3% digestibility based on a 34%

NDF diet, Milton and Demment, 1988].

푇푁퐶 퐶푃 퐹퐴푇 푁퐷퐹 Eq. 1 푀퐸 = (% 0.164 ) + (% 0.164) + (% 0.3766) + (0.164 × (% 0.543)) 퐷푀 퐷푀 퐷푀 퐷푀

The phenol-sulfuric acid assay was used to estimate water-soluble carbohydrates with a sucrose standard [DuBois et al. 1956]. Starch, sucrose, D-glucose, and D-fructose were measured enzymatically with the UV-method (R-Biopharm). The presence or absence of condensed tannins (CT) was assessed with the acid-butanol assay [Waterman and Mole, 1994]. We did not assess %CT per food because of inadequate sample sizes to develop specific condensed tannin extractions and standard curves per food part [Rothman et al., 2009]. Not all chemistry procedures were performed on all samples. All quantitative nutritional values are presented on a dry matter basis.

Nutrient intake

For our nutrient intake calculations, we only included days where the first foraging bout was reliably observed after dawn (0630-0700 hours) and where out-of-sight feeding events 80 comprised less than five percent of all feeding observations (N=19). Daily nutrient intake was calculated as the sum of consumed food items observed during a single day multiplied by their respective dry weights and chemical compositions. Average feeding rates were used to estimate the nutrient intake for a given feeding bout in cases when the total number of ingested food units

(e.g., leaf, flower, fruit) could not be counted [O'Malley and Power, 2014]. In cases where fruit development stage was not identified or where focal individuals consumed fruits in multiple stages during the same feeding bout, we used the average nutrient values for combined ripe, semi-ripe, and unripe fruits. Of the 594 total feeding observations used to calculate nutrient intakes, the food species or food item could not be identified for 2.4% of these observations.

When the food item but not the food species was recorded (e.g., young leaves of an unknown species), we used the average nutritional values of that food item in our total sample of foods as a proxy for that feeding observation. Neither the food item nor food species was identified in

0.6% of feeding observations. For these cases, an overall average of the nutritional values of all food items was substituted for the unknown food. To test the hypothesis that focal individuals selected foods based on energy return rates [sensu Stephens and Krebs, 1986], we examined the food choices made by focal individuals each day in relation to food “profitability”. Food profitability is the metabolizable energy (kJ g-1) of a food divided by the amount of time needed to process and ingest that food.

Statistical procedures

Non-parametric tests were used to examine group differences and correlations among variables. We used Fisher’s exact test compare food choices to changes in food profitability.

Spearman rank correlation was used to assess how daily nutrient intakes changed with male social rank (lower, middle, upper), daily party size, and the presence of sexually receptive 81 females. In addition, we considered how nutrient intakes differed among individual males with the Kruskal-Wallis test. Finally, we used a generalized linear mixed model (GLMM) approach to test for multivariate associations between food choice, nutritional composition, and social factors. The response variable was food choice (eaten, yes/no) for each day. The explanatory variables were food profitability, daily party size, and number of estrus females in the daily party. Focal study subject and sample day were included as random effects. The significance level was set at α=0.05. Statistical analyses were performed in R (R Core Development Team,

2014).

Research ethics

This project was authorized by the République du Sénégal, Ministère de l’Environnement et du Développement Durable, Direction des Eaux, Forêts, Chasses et de la Conservation des

Sols. The data collection protocol, which included non-invasive behavioral observations of habituated chimpanzees, was approved by the Institutional Animal Care and Use Committee at

Iowa State University. The United States Department of Agriculture authorized the import of plant and insect food samples. Our research adheres to the American Society of Primatologists’

Principles for the Ethical Treatment of Non-Human Primates and Code of Conduct for Protecting the Health of Wild Primates.

Results Nutritional composition and availability of foods

We were in contact with focal individuals for 677 hours, and continuously recorded 524 hours of focal individual behavior (Table 3.1). Focal individuals consumed 46 food items from

38 species during the study periods, including 34 plant, three insect, and one mammal species

(Table 3.2). Within all feeding observations in our sample (438 observations in 63 days), 6.8% of food items could not be identified to genus or species. Of these cases, about half were classified 82 according to food type (leaf, fruit, pith, gum). Thus, some foods may have been overlooked in this study, but our sample is broadly representative of diet at Fongoli [Bogart and Pruetz, 2011;

Pruetz 2006]. In addition to energy and macronutrients, we surveyed real and potential dietary items for the presence of condensed tannins (CT) to determine if Fongoli chimpanzees avoided foods with chemical compounds that inhibit protein digestibility. Condensed tannins were often eaten by focal individuals, as 81.2% of surveyed foods (N=33) tested positive. Of this sample,

100% of pith (N=2), flower (N=2), and leaves (N=4), 87.5% of fleshy fruits (N=16), and 50% of seed (N=6) foods contained CT. In terms of availability, most foods fit a clumped-scarce pattern, followed by scattered-abundant, scattered-hyperabundant, scattered-scarce, and clumped- abundant in that order (Figure 3.1).

We examined the nutritional values of fruits (fleshy mesocarp or endocarp) at Fongoli to comparable data at four, long-term Pan study sites to test for differences in food quality at

Fongoli that may be related to the savanna environment. These four sites include Taï National

Park in Ivory Coast (P.t. verus), Gashaka-Gumti National Park, Nigeria (P.t. ellioti), the Ngogo community at Kibale National Park, Uganda (P.t. schweinfurthii), and Salonga National Park in the Democratic Republic of Congo (P. paniscus) [Hohmann et al., 2006; Hohmann et al., 2010].

Fongoli had the highest average values for starch and simple sugars, and lowest values for neutral detergent fiber (Table 3.3); otherwise, Fongoli fruits were broadly similar to the four sites

[Hohmann et al., 2010].

We combined information on the nutritional value of food with observations of foraging behavior to assess environmental and social factors related to food choice. Adult males selected an average of seven different foods each day (range: 2-14). To estimate total daily nutrient intakes from these observations, we summed the number of food items consumed per feeding 83 bout. When it was not possible to count food items for an entire bout, we estimated nutrient intakes from food-specific ingestion rates (dry matter g hr-1). We recorded feeding duration and calculated feeding rate for 609 partial or complete feeding bouts where the focal individual captured, processed, and ingested food items in clear view. On average, 9804 kJ of metabolizable energy (Figure 3.2, range: 2878-19157), 81.91 g of crude protein (range: 32.17-220.93), 48.3 g of crude fat (range: 4.40-167.06), and 172.44 g of water-soluble carbohydrates (range: 42.61-

261.55) were consumed each day by focal subjects (N=19).

Relationships between food profitability and food choice

To evaluate the hypothesis that adult male foraging strategies include energy return rate maximization (ERRM), we compared food choices to food profitability rankings (metabolizable energy in kJ per gram / ingestion rate per gram). Gum was the most profitable food type, closely followed by fruit, leaves and flowers. Insects and pith had substantially lower values, but we did not conduct significance tests due to small sample sizes (Figure 3.3). Next, we compared these rankings to the number of observation days per field season when focal individuals consumed each food, using this frequency of consumption to estimate the relative importance of each food to the diets of adult males. Four plant species contributed the most to daily energy intake during the study period: Adansonia digitata (unripe fruit and seed combined, ripe fruit endocarp, and reingestion of mature seeds), Tamarindus indica (unripe fruit and seed), and Ficus ingens and F. umbellata (unripe, semi-ripe, and ripe figs, combined) (Figure 3.4). In addition, Macrotermes subhyalinus (soldiers, workers) and Oxytenanthera abyssinica pith were commonly eaten but had relatively low food profitability scores. We further explored the ERRM strategy in adult males by testing the prediction that individuals preferred more profitable foods when simultaneously given the choice of two or more foods. We examined food selection when two or more foods of 84 different profit ranks were consumed within the same feeding tree during the same food patch visit. We recorded 20 instances where an individual switched between ripe and unripe fruit of the same species, Adansonia digitata. Focal individuals more often consumed ripe fruit (endocarp only) first, which has a higher energy profit ranking, and then switched to unripe fruit and seed

(ranked 3rd and 15th, respectively, of 34 foods; Fisher’s Exact Test: P<0.001).

Effects of estrus females, social rank, and party size on nutrient intake

To gauge the impact of social dynamics on foraging, we evaluated individual and rank differences in daily metabolizable energy and macronutrient intakes (N=19) for adult males

(N=8). In addition, we assessed these daily intakes relative to total party size and the presence of estrus females in the community each day. Focal individuals did not differ in daily energy and macronutrient intakes (Table 3.4). Therefore, we pooled these intake estimates to evaluate other social factors. The social ranking of an adult male was not associated with daily energy, protein, or fat intakes (Table 3.4), but higher-ranked males consumed more water-soluble carbohydrates

(ρ=-0.51, P<0.05). Daily party size did not explain daily energy, protein, or sugar intakes (Table

3.4), but fat intake decreased substantially as party size increased (ρ=-0.60, P<0.01). In addition, adult males decreased daily energy intake in the presence of females with maximally-tumescent sexual swellings (Spearman: ρ= -0.42, P<0.05; Table 3.4).

Lastly, we used a GLMM to test whether or not choice of food part/species consumed

(Table 3.2) was explained by profitability value, daily party size, and number of estrus females in the daily party. For each food choice observation, we only included food items that should have been available within their home range, as determined by monthly phenological records on the vegetation belt transect. None of these factors were good predictors of food choice (profitability 85 estimate = 0.04, SE = 0.12, z = 0.32, P = 0.75; daily party size estimate = 0.04, SE = 0.03, z =

0.16, P = 0.17; estrus females estimate = 0.30, SE = 0.27, z = 1.08, P = 0.28).

Discussion Significance of time, energy, and macronutrients on food selection

This analysis considered the roles of nutritional quality, primarily macronutrient values, as well as social factors on food selection behavior in adult male chimpanzees. We found that both of these domains shaped food choices at Fongoli, which supports the idea that chimpanzee foraging strategies are characteristically socio-nutritional (Tennie et al. 2009). As predicted by basic optimal food selection models [Stephens and Krebs, 1986], adult male chimpanzees at

Fongoli preferred foods with higher profitability values when two foods of differing values were simultaneously available. However, food profitability was not a strong predictor of food choice in our larger model that considered all foods consumed during the study period.

Foods from five plant (Adansonia, Ficus, Oxytenanthera, Tamarindus, Strychnos) and one animal (Macrotermes) were commonly consumed during the study period (Figure 3.4).

Adansonia, Ficus and Tamarindus fruits had the highest profitability ranks and, thus, were top foods in terms of daily energy and macronutrient intakes. Macrotermes (termites) and

Oxytenanthera (bamboo pith) contributed fewer calories to daily food intake but were nonetheless important because focal subjects frequently consumed them. Oxytenanthera was

“wadged” by Fongoli chimpanzees, meaning that individuals masticated the pith to extract the easily digestible portion (i.e., sucrose, in this case), and then discarded the lignous bolus. Thus, bamboo pith was a common source of simple sugars. In contrast, Macrotermes was a common source of protein and minerals. Both termites and bamboo are abundant at Fongoli and eaten during most or all months of the year [Bogart and Pruetz, 2011; Pruetz, 2006; Pruetz and 86

Bertolani, 2009; Pruetz, unpublished data; this study]. Although these two foods ranked very low on the food profitability spectrum, they may be nutritionally complementary to each other when routinely eaten. In other words, pairing termites and bamboo pith may produce a staple food source. We recommend that future studies use a nutrient balancing approach to explore this idea.

Male strategies to balance nutritional and social gains

Our observations indicate that adult males at Fongoli require approximately 9804 kJ of metabolizable energy per day, over the long term, to meet or exceed daily energy requirements.

One caveat to this interpretation is that our study period occurred when a relatively large, abundant, and high-energy food, baobab (Adansonia digitata) fruit, was in season. A longer study that includes more periods of fruit scarcity may find that average daily energy intake at

Fongoli is lower than current estimates. We found a wide range of daily metabolizable energy intake estimates for adult males (range: 2878-19157 kJ); however, both outliers in our sample

(Figure 3.2) were from a single, young adult male that had recently entered the social hierarchy.

It may be that individuals experience higher variation in day to day nutrient intakes when they experience periods of social instability, such as changing social rank, but this idea requires more research. In spite of high variation, social rank class and individual differences were not associated with daily energy intake. The only social factor that explained daily energy intake was the presence of estrus females, with intakes substantially declining when more than one sexually receptive female was present each day.

Given that chimpanzees are not egalitarian, we expected to see differences in food intake among social ranks, with higher ranking individuals having better access to highly profitable foods. Although we found no relationship between social rank and food intake when examining energy intake, differences appeared at the macronutrient level. Higher ranking males ingested 87 more water-soluble carbohydrates (i.e., simple sugars) each day, which indicates that these males used their social status to gain access to sweeter foods. Given that daily energy intake did not change with rank, we can predict that lower ranking males compensate for secondary access to sugary foods through seeking out alternatives with more fat or protein, for example. However, this increase in protein or fat intake for lower ranking males was not observed (Table 3.4).

Another possibility is that lower ranking males are consuming more starchy foods, while higher ranking individuals get first pick of the sweeter foods. Unfortunately, small sample sizes prevent us from making this comparison. We recommend more research on this topic, as males may employ a range of macronutrient intake strategies based on social rank to meet their nutritional goals.

Savanna chimpanzee nutritional ecology

This study is the first to quantify the nutritional values of chimpanzee foods in the arid savanna-woodlands demarcating the northwestern boundary of western chimpanzees. After more than a decade of study at Fongoli, Senegal, 92 food items from 75 plant species have been identified [Pruetz, unpublished data]. This estimate is well below the average of 136 plant species consumed at forested sites [Table 6.2: Pruetz, 2006], but the lower number of food species fits the trend of decreasing species richness with increasing latitude [Rosenzweig, 1995].

Despite this relatively narrow breadth of plant food species, overall diet composition at Fongoli fits the typical chimpanzee pattern [Table 6.1: Pruetz, 2006]. Our analysis included the subset of dietary items (42 plant foods from 34 species) that were eaten by Fongoli chimpanzees from

September to February. This period falls between the middle of the wet season to the middle of the dry season [Pruetz and Bertolani, 2009]. 88

Hohmann and co-workers [2010] found substantial variation in average macronutrient values for fruits across four Pan study sites. Our study contributes a fifth site to the Pan comparative nutritional dataset and supports the finding that variation in fruit nutritional quality among sites is considerably high. Although average values at Fongoli for most macronutrients were within the range of variation (Table 3.3), average starch and simple sugars were relatively high. Of interest to this finding is that the majority of the fruits in our sample (63%) were also consumed by people in Senegal [Johnson-Fulton and Pruetz, 2004] and tasted sweet [S.L., personal observation]. Moreover, the seeds of most of these fruits are dispersed, rather than predated on, by chimpanzees. Perhaps fleshy-fruit trees that are scattered in savanna-woodlands are higher in sugar than similar trees in more forested areas to reward wide-ranging seed dispersers, such as chimpanzees. This idea should be tested by comparing the sugar content of more fleshy fruit tree species, as well as their abundance and distribution among Pan study sites.

We also showed that Fongoli chimpanzees commonly consume foods with condensed tannins, as has been demonstrated at several other chimpanzee sites [Hohmann et al., 2010; Reynolds et al.,

1998; Wrangham et al. 1998]. The extent to which CT inhibits protein digestion in chimpanzees is poorly understood; however, the presence of CT, even at relatively high concentrations, often does not deter feeding [Remis, 2002; Reynolds et al., 1998]. Until we have a better understanding of how CT impact chimpanzee digestive physiology, it will be difficult to determine if and how chimpanzees select foods based on CT content.

At a fundamental level, the findings presented here show that male chimpanzees at

Fongoli are energy maximizers, and suggest that nutrient mixing is a means to exploit abundant, low profitability foods. Moreover, it is likely that individuals can chose from multiple social foraging strategies that prioritize certain macronutrients to meet their daily energy requirements. 89

Finally, maximizing energy return rates is less important, on a temporary basis, when sexually receptive females are available. We recommend that future studies deepen the investigation into these socio-nutritional strategies.

Acknowledgements We thank the République du Sénégal, Ministère de l’Environnement et du

Développement Durable, Direction des Eaux, Forêts, Chasses et de la Conservation des Sols

(MEDD-EFCCS) for granting us permission to study chimpanzees in Senegal. This research was sponsored by the Leakey Foundation, Rufford Small Grant Foundation, ISU Department of

Anthropology, CUNY-Hunter Primate Nutritional Ecology Lab, and ISU Grain Quality Lab. Our study complies with laws concerning the export and import of plant and insect tissues that are regulated by MEDD-EFCCS and the United States Department of Agriculture. The ISU

Institutional Animal Care and Use Committee approved this research. SL thanks the following individuals: Dave Vleck, Stan Harpole, Dean Adams, Brent Danielson, Charles Hurburgh, Ulises

Villalobos-Flores, Michel Sahdiako, Dondo Kante, Anna Olson, El-Hadji Sakho, Jenny Paltan,

Papa Ibnou Ndiaye, Landing Badji, Waly Camara, Mody Camara, Assane Goudiaby, the Cheikh

Anta Diop University Herbarium, Josh Marshack, Morty and Beverly Marshack, Martha Jeffrey, and the ISU Swine and Poultry Nutrition Lab.

Literature cited Altmann J. 1974. Observational study of behavior - sampling methods. Behaviour 49(3-4):227-

267.

Altmann SA. 1998. Foraging for survival. Chicago: University of Chicago Press.

AOCS. 2005. Approved procedure AM 5-04: rapid determination of oil/fat utilizing high

temperature solvent extraction. Urbana, IL. 90

Boesch C. 1996. Social grouping in Taї chimpanzees. In: McGrew WC, Marchant LF, Nishida T,

editors. Great ape societies. Cambridge: Cambridge University Press. p 101-113.

Bogart SL, Pruetz JD. 2011. Insectivory of savanna chimpanzees (Pan troglodytes verus) at

Fongoli, Senegal. American Journal of Physical Anthropology 145(1):11-20.

Chapman CA, Chapman LJ, Rode KD, Hauck EM, McDowell LR. 2003. Variation in the

nutritional value of primate foods: among trees, time periods, and areas. International

Journal of Primatology 24(2):317-333.

Conklin NL, Knott CD, Wrangham RW. 2006. Energy intake by wild chimpanzees and

orangutans: methodological considerations and a preliminary comparison. In: Hohmann

G, Robbins MM, Boesch C, editors. Feeding ecology in apes and other primates.

Cambridge: Cambridge University Press. p 445-471.

Conklin NL, Wrangham RW. 1994. The value of figs to a hind-gut fermenting frugivore - a

nutritional analysis. Biochemical Systematics and Ecology 22(2):137-151.

Conklin-Brittain NL, Wrangham RW, Hunt KD. 1998. Dietary response of chimpanzees and

cercopithecines to seasonal variation in fruit abundance. II. Macronutrients. International

Journal of Primatology 19(6):971-998.

Deblauwe I, Janssens GPJ. 2008. New insights in insect prey choice by chimpanzees and gorillas

in Southeast Cameroon: The role of nutritional value. American Journal of Physical

Anthropology 135(1):42-55.

Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. 1956. Colorimetric method for determination of sugars and related substances. Analytical Chemistry, 28, 350–

356. 91

Emery Thompson M, Kahlenberg SM, Gilby IC, Wrangham RW. 2007. Core area quality is

associated with variance in reproductive success among female chimpanzees at Kibale

National Park. Animal Behaviour 73(3):501-512.

Emlen JM. 1966. The role of time and energy in food preference. The American Naturalist

100(916):611-617.

Felton A, Felton A, Lindenmayer D, Foley W. 2009. Nutritional goals of wild primates.

Functional Ecology 23(1):70-78.

Furuichi T. 2009. Factors underlying party size differences between chimpanzees and bonobos: a

review and hypotheses for future study. Primates 50(3):197-209.

Ganas J, Ortmann S, Robbins MM. 2009. Food choices of the mountain gorilla in Bwindi

Impenetrable National Park, Uganda: the influence of nutrients, phenolics and

availability. Journal of Tropical Ecology 25:123-134.

Glander KE. 1977. Poison in a monkey’s Garden of Eden. Natural History 86(3):34-41.

Goering HK, Van Soest PJ. 1970. Forage fiber analysis: United States Department of

Agriculture.

Goodall J. 1986. The chimpanzees of Gombe. Cambridge, MA: The Belknap Press of Harvard

University Press. 674 p.

Heesen M, Rogahn S, Ostner J, Schulke O. 2013. Food abundance affects energy intake and

reproduction in frugivorous female Assamese macaques. Behavioral Ecology and

Sociobiology 67(7):1053-1066.

Heller R. 1980. On optimal diet in a patchy environment. Theoretical Population Biology

17(2):201-214. 92

Hohmann G, Fowler A, Sommer V, Ortmann S. 2006. Frugivory and gregariousness of Salonga

bonobos and Gashaka chimpanzees: the influence of abundance and nutritional quality of

fruit. In: Hohmann G, Robbins MM, Boesch C, editors. Feeding ecology in apes and

other primates. Cambridge, UK: Cambridge University Press. p 123-159.

Hohmann G, Fruth B. 2008. New records on prey capture and meat eating by bonobos at Lui

Kotale, Salonga National Park, Democratic Republic of Congo. Folia Primatologica

79:103-110.

Hohmann G, Potts K, N'Guessan A, Fowler A, Mundry R, Ganzhorn JU, Ortmann S. 2010. Plant

foods consumed by Pan: exploring the variation of nutritional ecology across Africa.

American Journal of Physical Anthropology 141(3):476-485.

Irwin MT, Raharison J-L, Raubenheimer D, Chapman CA, Rothman JM. 2014. Nutritional

correlates of the "lean season": effects of seasonality and frugivory on the nutritional

ecology of diademed sifakas. American Journal of Physical Anthropology 153(1):78-91.

Isabirye-Basuta G. 1989. Feeding ecology of chimpanzees in the Kibale Forest, Uganda. . In:

Heltne P, Marquardt L, editors. Understanding chimpanzees. Cambridge, MA: Harvard

University Press. p 116-127.

Licitra G, Hernandez TM, VanSoest PJ. 1996. Standardization of procedures for nitrogen

fractionation of ruminant feeds. Animal Feed Science and Technology 57(4):347-358.

MacArthur RH, Pianka ER. 1966. On optimal use of a patchy environment. The American

Naturalist 100(916):603-609.

Marshall AJ, Boyko CM, Feilen KL, Boyko RH, Leighton M. 2009. Defining fallback foods and

assessing their importance in primate ecology and evolution. American Journal of

Physical Anthropology 140(4):603-614. 93

McGrew WC. 2014. The 'other faunivory' revisited: Insectivory in human and non-human

primates and the evolution of human diet. Journal of Human Evolution 71:4-11.

Milton K. 1981. Food choice and digestive strategies of two sympatric primate species.

American Naturalist 117(4):496-505.

Milton K, Demment MW. 1988. Digestion and passage kinetics of chimpanzees fed high and

low-fiber diets and comparison with human data. Journal of Nutrition 118(9):1082-1088.

N'guessan AK, Ortmann S, Boesch C. 2009. Daily energy balance and protein gain among Pan

troglodytes verus in the Taї National Park, Côte d'Ivoire. International Journal of

Primatology 30:481-496.

Newton-Fisher NE. 1999. The diet of chimpanzees in the Budongo Forest Reserve, Uganda.

African Journal of Ecology 37(3):344-354.

Nishida T, Zamma K, Matsusaka T, Inaba A, McGrew WC. 2010. Chimpanzee behavior in the

wild. Tokyo: Springer.

O'Malley RC, Power ML. 2012. Nutritional composition of actual and potential insect prey for

the Kasekela chimpanzees of Gombe National Park, Tanzania. American Journal of

Physical Anthropology 149(4):493-503.

O'Malley RC, Power ML. 2014. The energetic and nutritional yields from insectivory for

Kasekela chimpanzees. Journal of Human Evolution 71:46-58.

Ortmann S, Bradley BJ, Stolter C, Ganzhorn JU. 2006. Estimating the quality and composition

of wild animal diets - a critical survey of methods. In: Hohmann G, Robbins MM, Boesch

C, editors. Feeding ecology in apes and other primates. Cambridge, U.K.: Cambridge

University Press. p 397-420. 94

Pruetz JD. 2009. The socioecology of adult female patas monkeys and vervets in Kenya. Upper

Saddle River, NJ: Pearson/Prentice Hall. 164 p.

Pruetz J, Kante D. 2010. Successful return of a wild infant chimpanzee (Pan troglodytes verus)

to its natal group after capture by poachers. African Primates 7(1):35-41.

Pruetz JD. 2006. Feeding ecology of savanna chimpanzees (Pan troglodytes verus) at Fongoli,

Senegal. In: Hohmann G, Robbins MM, Boesch C, editors. Feeding ecology in apes and

other primates. Cambridge, UK: Cambridge University Press. p 161-182.

Pruetz JD, Bertolani P. 2009. Chimpanzee (Pan troglodytes verus) behavioral responses to

stresses associated with living in a savanna-mosaic environment: implications for

hominin adaptations to open habitats. Paleoanthropology:252-262.

R Core Team. 2014. R: A Language and Environment for Statistical Computing. Vienna,

Austria: R Foundation for Statistical Computing. http://www.R-project.org.

Raubenheimer D, Simpson SJ, Mayntz D. 2009. Nutrition, ecology and nutritional ecology:

toward an integrated framework. Functional Ecology 23:4-16.

Remis MJ. 2002. Food preferences among captive western gorillas (Gorilla gorilla gorilla) and

chimpanzees (Pan troglodytes). International Journal of Primatology 23(2):231-249.

Reynolds V, Plumptre AJ, Greenham J, Harborne J. 1998. Condensed tannins and sugars in the

diet of chimpanzees (Pan troglodytes schweinfurthii) in the Budongo Forest, Uganda.

Oecologia 115(3):331-336.

Riedel J, Franz M, Boesch C. 2011. How feeding competition determines female chimpanzee

gregariousness and ranging in the Taї National Park, Cote d'Ivoire. American Journal of

Primatology 73(4):305-313. 95

Rosenzweig ML. 1995. Species diversity in space and time. Cambridge: Cambridge University

Press.

Rothman J, Dierenfeld E, Hintz H, Pell A. 2008. Nutritional quality of gorilla diets:

consequences of age, sex, and season. Oecologia 155(1):111-122.

Rothman JM, Chapman CA, Van Soest PJ. 2012. Methods in primate nutritional ecology: a

user's guide. International Journal of Primatology 33(3):542-566.

Rothman JM, Dusinberre K, Pell AN. 2009. Condensed tannins in the diets of primates: a matter

of methods? American Journal of Primatology 71(1):70-76.

Stephens D, Brown J, Ydenberg R, editors. 2007. Foraging: behavior and ecology. Chicago, IL:

The University of Chicago Press. 576 p.

Stephens DW, Krebs JR. 1986. Foraging theory. Princeton: Princeton University Press. 247 p.

Takemoto H. 2003. Phytochemical determination for leaf food choice by wild chimpanzees in

Guinea, Bossou. Journal of Chemical Ecology 29(11):2551-2573.

Tennie C, Gilby I, Mundry R. 2009. The meat-scrap hypothesis: small quantities of meat may

promote cooperative hunting in wild chimpanzees (Pan troglodytes). Behavioral Ecology

and Sociobiology 63(3):421-431.

Watts DP, Potts KB, Lwanga JS, Mitani JC. 2012. Diet of chimpanzees (Pan troglodytes

schweinfurthii) at Ngogo, Kibale National Park, Uganda, 1. diet composition and

diversity. American Journal of Primatology 74(2):114-129.

Wrangham RW, Conklin-Brittain NL, Hunt KD. 1998. Dietary response of chimpanzees and

cercopithecines to seasonal variation in fruit abundance. I. Antifeedants. International

Journal of Primatology 19(6):949-970.

100

101

102

103

Figure. 3.1. Histograms of qualitative food availability scores by food part. Availability classes include: clumped-abundant (ca), clumped-scarc (cs), scattered-abundant (sa), scattered-hyperabundant

(sh), and scattered-scarce (ss). 104

Figure. 3.2. Histogram of daily metabolizable energy intake (kJ) for adult male study subjects. 105

Figure 3.3 Food profitability (log ME kJ hr-1) by food part.

106

Figure. 3.4. The relationship between food profitability (metabolizable energy per gram/ingestion rate per gram, or E/T) and the number of days per season that focal individuals consumed each food. Each point represents a unique food species and part. Symbols represent fruit (+), gum (), leaves (), pith (◊), flowers (), and insects (○). The dashed lines create quadrats that group foods into four categories: (1) high E/T and commonly eaten, (2) high E/T and rarely eaten, (3) low E/T and commonly eaten, and (4) low E/T and rarely eaten. All foods in the upper right quadrant were ripe or unripe fruits (rf: ripe fruit, uf: unripe fruit, sr: seed reingestion (Bertolani and Pruetz 2011)). Note that pith and insect foods had relatively low profitability scores, but Macrotermes and Oxytenanthera were commonly consumed.

107

CHAPTER IV FEEDING IN FEAR: HABITAT SELECTION AND PREDATOR-SENSITIVE FORAGING IN WESTERN CHIMPANZEES (PAN TROGLODYTES VERUS) Manuscript will be submitted to the American Journal of Physical Anthropology

Stacy M. Lindshield,1,2* Brent J. Danielson,3 Jessica M. Rothman,4,5 and Jill D. Pruetz1

1Department of Anthropology, Iowa State University, Ames, IA 50011

2Ecology and Evolutionary Biology Program, Iowa State University

3Department of Ecology, Evolution and Organismal Biology, Iowa State University

4Department of Anthropology, Hunter College of the City University of New York, New York, NY

5New York Consortium in Evolutionary Primatology, New York, NY

Grant sponsor: The Leakey Foundation, The Rufford Foundation

*Correspondence to:

S.M. Lindshield

324 Curtiss Hall

Department of Anthropology

Iowa State University

Ames, IA 50011

E-mail: [email protected]

Key words Savanna chimpanzee, Senegal, risk-sensitive foraging, nutrient intake, feeding rate

108

Abstract Adaptations that lower the probability of predation are widespread in mammalian prey species, as an act of successful depredation instantaneously eliminates any chance that prey will subsequently survive or reproduce. However, strategies that minimize predation risk are poorly understood in chimpanzees (Pan troglodytes), one of the best-studied non-human primates.

Chimpanzees encounter few large carnivores capable of successfully capturing and killing these large apes and, thus, predation events are rarely observed. To address this problem, we assessed antipredator strategy in western chimpanzees (P. t. verus) using humans as model predators in a savanna mosaic environment. Our objective was to evaluate relationships between foraging, habitat selection, and antipredator behaviors of adult males (N=11) at Fongoli, Senegal.

Proximity of food sources to anthropogenic landmarks were compared to food intake, measured in grams of dry matter, and behavioral indicators of perceived predation risk. In addition, we compared foraging behavior to habitat and surface water availability. Fongoli chimpanzees displayed antipredator behavior towards people, but this perceived risk did not deter subjects from visiting sources of high-quality food; instead, males consumed more food during feeding bouts when antipredator behavior occurred. In addition, male subjects increased feeding rate in open-canopy habitats, which likely offsets the higher metabolic cost of foraging in relatively exposed areas. Although people rarely hunt chimpanzees in Senegal, we show that chimpanzees perceive them as predators and adjust their foraging behavior in response to this risk. Our findings indicate that the risk of predator attack may affect a forager’s decision to increase or decrease food intake. From a conservation perspective, this result enriches our understanding of how food choice and habitat type selection factors into human-chimpanzee conflict.

109

Introduction Chimpanzees occupying savanna biomes or open-country landscapes face seasonal periods of abiotic and biotic resource scarcity associated with precipitation and temperature seasonality. These pulses of resource abundance and scarcity between wet and dry periods are associated with differences in feeding (Bogart and Pruetz 2011; Hernandez-Aguilar et al. 2007;

Pruetz and Bertolani 2007), ranging (Baldwin et al. 1982), social (Pruetz and Lindshield 2012;

Pruetz and Bertolani 2009), and sleeping behaviors (Ndiaye et al. 2013) as well as population carrying capacity (Baldwin et al. 1982) in relation to environments with less pronounced seasonality patterns. Although links between savanna environments and chimpanzee behavior have been widely recognized, the mechanisms and processes that explain for these patterns are rarely demonstrated in the empirical domain, largely due to the lack of habituated subjects until recently. Furthermore, across their geographical range, most savanna chimpanzees face strong competition for land and other resources with people that have long occupied and intensively used savanna environments for hunting, gathering, horticulture, pastoralism, and mineral extraction (Boyer, 2011; Kortland, 1983; Reid and Ellis, 1995). In addition, anthropogenic fire regimes, which shape savanna biomes (Higgins et al., 2007), may indirectly affect chimpanzees by changing the abundance and distribution of chimpanzees’ plant and animal foods. It is reasonable to assume that, when in sympatry, savanna chimpanzees have enduring ecological relationships with people that are primarily characterized by competitive and predatory interactions, and that such relationships have a long history. Thus, the chimpanzee home range consists of a multi-layered resource patchwork, which includes sources of food, water, shade, sleeping sites, and refuge from predators. However, how savanna chimpanzee behavior is shaped by this dynamic, resource mosaic remains unclear. 110

Habitat selection, the decision making process where individuals select resources within and between a range of available habitat types (e.g., gallery forest, woodland, grassland), is known to vary among chimpanzee communities that occupy different zones of the savanna to evergreen forest continuum. On the extremes, rainforest chimpanzees occupy evergreen and semi-deciduous woody plant communities (Boesch and Boesch-Achermann 2000; Morgan et al.

2006), whereas savanna chimpanzees primarily encounter deciduous woodland and grassland habitats but intensively use small areas of evergreen gallery forest with a dense canopy cover along water courses within these savanna biomes (Duvall 2000; Hernandez-Aguilar 2009;

McGrew et al. 1981; Pruetz and Bertolani 2009). Chimpanzees in southeastern Senegal occupy a savanna biome characterized by a relatively arid and hot climate with about 2-3% of land covered by small patches of evergreen forest (McGrew et al. 1981; Pruetz and Bertolani 2009).

These forest patches are critical sources of shade, sleeping sites, and food, and are often in close proximity to surface water. Although chimpanzees at Fongoli, Senegal spend a disproportionately long time in forested habitats, particularly when at rest (Pruetz and Bertolani

2009), most plant foods are rooted in woodlands consisting of small to medium deciduous trees and a thick, grassy understory (Pruetz 2006). In woodland habitat, individuals must range across a relative large area to acquire critical resources (Baldwin et al. 1982; Skinner and Pruetz 2012).

Thus, the process of habitat selection in chimpanzees is sensitive to the availability of food, water, shade, and sleeping sites in savanna habitat mosaics.

Habitats often vary in resource abundance and distribution, as well as predation risk and metabolic cost to foragers (Brown 1988; Rieucau et al. 2009). These characteristics, which fall under the umbrella term, habitat quality, can predictably influence foraging decisions (i.e., the process of searching for, handing and consuming food, following Stephens and Krebs 1986). 111

Few studies have investigated the role of habitat quality on chimpanzee foraging behavior

(Bogart and Pruetz 2011; Chancellor et al. 2012; Gilby et al. 2006; Head et al. 2012; Pascual-

Garrido et al. 2013; Pruetz and Bertolani 2009) and, to the best of our knowledge, no study has shown how food intake (dry mass ingested) varies with habitat quality. However, several predictable relationships between food intake and habitat quality occur in other animals, such as rodents, ungulates and birds (Brown 1999; Makin et al. 2012; Olsson and Molokwu 2007;

Rieucau et al. 2009). First, food intake may covary with food and water availability among habitats. For instance, foragers in food-rich habitats tend to leave more food behind than those in food-poor habitats (reviewed in Brown and Kotler 2007; Olsson and Molokwu 2007), likely because foragers have ample opportunity to revisit food sources in the former but not the latter case. Similarly, in water-limited environments, drinking water is a complementary resource, leading individuals to ingest more food when it is nearby (Corvus coronoides; Kotler et al.,

1998; seed-eating savanna birds; Molokwu et al., 2010). Second, habitats can incur different energetic costs on foragers and, thus, impact how long an individual occupies and feeds within a habitat. For example, white-tailed deer (Odocoileus virginianus) consume less food in habitats that require greater energy expenditure to regulate body temperature (Rieucau et al. 2009). A third component of habitat quality, predation risk, tends to leave a distinct foraging signature: individuals consume food at a higher rate of return in risky habitats to lower the probability of encountering a predator (reviewed in Lima and Dill 1990).

Adjusting foraging behavior in response to predation risk, also called predator-sensitive foraging (PSF, Sinclair and Arcese 1995), is one of a suite of behaviors that may comprise an antipredator strategy. Collectively, this set of behavioral adjustments minimizes predation risk for a given individual, group, or population (reviewed in Miller 2002). Conspicuous antipredator 112 behaviors include but are not limited to vigilance, which is an overall alertness to one’s surroundings, vocalizations that convey risk or immediate danger (i.e., alarm calls), fleeing, hiding, or mobbing predators. In contrast, some antipredator behaviors, such as PSF, are subtle: individuals may cluster into ephemeral and relatively large foraging parties while using high- value, high-risk resources (Ateles belzebuth; Link and Di Fiore 2013), preferentially use less risky habitats within a territory or home range (Cercopithecus mitis; Coleman and Hill 2014), consume less food in risky habitats (Chlorocebus aethiops; Makin et al. 2012), or take more risks when food is scarce (Connochaetes taurinus; Sinclair and Arcese 1995).

The impact of predation risk on chimpanzee behavior is poorly understood (reviewed in

Stewart and Pruetz 2013). Chimpanzees have few predators, which is likely related to their relatively large body mass (Miller 2002), arboreality (Hernandez-Aguilar et al. 2013), sociality, and predator mobbing ability (Boesch 2009). Documented predation events involve leopards

(Panthera pardus) (Boesch 2009; Nakazawa et al. 2013) and lions (Panthera leo) (Nishida

2012). However, such events are rarely witnessed, thus limiting our understanding of predator- prey dynamics. Local extirpation of large carnivores at several long-term study sites partially explains the scarcity of predator attacks on chimpanzees (Boesch 2009; Pruetz et al. 2008). To complicate matters, chimpanzees may compete with large carnivores for the same prey species, such as red colobus monkey (Colobus badius) (Zuberbühler and Jenny 2002), that is, this web of ecological relationships is multidimensional. For these reasons, the relative importance of predation on chimpanzee behavior can be controversial, and may vary among chimpanzee communities (McGrew 2010). However, we have an opportunity to evaluate chimpanzee antipredator strategies with people as model predators. 113

Although human societies vary in subsistence and land use behaviors, historic (Kay

2007) and contemporary (Effiom et al. 2013) evidence indicates that Homo sapiens are keystone predators. Humans impose a risk of depredation to chimpanzees in many locations where they are in sympatry (Halloran et al. 2013; Hicks et al. 2010; Pruetz and Kante 2010; but see Morgan and Sanz, 2003). In some cases, people do not hunt chimpanzees, but the latter exhibit signs of anxiety while in close proximity to people or anthropogenic landmarks, such as cropland

(Hockings et al. 2007). Although wild chimpanzees are known to kill people, primarily infants and children (Hockings et al. 2009), it is widely acknowledged that chimpanzees are more often prey than predator in human-chimpanzee relationships. Chimpanzees may exhibit nervousness, alarm calling, fleeing, or hiding in response to human activity (Hicks et al. 2012; Hockings et al.

2007). Moreover, crop-raiding indicates that chimpanzees are willing to enter risky areas to forage for high-energy foods, and that chimpanzees compete with humans for food (Chancellor et al. 2012; Hockings et al. 2012; Pienkowski et al. 1998; Tweheyo et al. 2005). Hockings and co-workers (2012) used a time-budget approach (Martin and Bateson 1986) to show that chimpanzees at Bossou, Guinea travelled and fed more often on days when foraging parties crop- raided. These findings are noteworthy because, under a basic food selection model (sensu

Stephens and Krebs 1986), traveling and feeding time are predicted to decline on days when individuals consume cultivars, as crop foods tend to be energy rich (kJ per gram) and large in size (mass in grams) relative to wild foods. Moreover, crop raiding is risky, and predator- sensitive foraging usually predicts that feeding decreases in risky habitats. Although the time budget approach used by Hockings and co-workers (2012) to demonstrate the relationship between crop-raiding and foraging is an established and credible method of sampling feeding behavior (Altmann 1974), it alone does not account for variation in nutritional composition 114 among foods to enable standardized comparisons of nutrient intakes and rates (Chivers 1998;

Nakagawa 2009; Schülke et al. 2006). What is missing from our understanding of chimpanzee antipredator strategy, and of habitat selection more broadly, is an assessment of feeding rate and nutrient intake responses to variation in habitat quality. This problem prompted us to evaluate

Fongoli chimpanzees’ foraging choices in light of predation risk and nutritional gains.

The objective of this study is to elucidate the process of habitat selection in savanna chimpanzees by examining foraging behavior vis-à-vis habitat quality using a multi-layered, conceptual model of resources and risks as perceived by the Fongoli chimpanzees. The first is a base layer of essential resources (food, water, shade) within the home range. The second, comingling layer is risk, or the potential for encountering unfamiliar humans. On the basis of foraging theory (Stephens et al. 2007) and empirical patterns on habitat use and quality for savanna chimpanzees, we hypothesize that individuals will respond to variation in habitat quality in a way that balances metabolic costs and predation risks with nutritional gains. Testable predictions of this hypothesis include: (1) increasing food intake in open-canopy habitats to compensate for the higher metabolic costs of foraging, (2) preferentially using food patches closer to surface water sources, (3) decreasing food intake in risky habitats due to the trade-off between vigilance and feeding, and (4) increasing foraging party size in risky habitats to reduce the probability of predator attacks. Study outcomes will elucidate chimpanzee antipredator strategies as well as foraging mechanisms of habitat selection for chimpanzees that occupy extreme, savanna environments.

Methods Study area and study subjects

The Fongoli study site (12°39ʹ N, 12°13ʹ W) is located in southeastern Senegal in the transitional Sudano-Guinean vegetation belt (White 1983). The dry season is six consecutive 115 months long; approximately 900 mm of precipitation falls between the months of June and

September, while May and October are transitional periods (Ba et al. 1997). Mean daily temperature is 28.4 C, and mean maximum daily temperature exceeds 40 C in the late dry season (Pruetz and Bertolani 2009). The savanna-woodland vegetation is a mosaic of grassland, woodland, and gallery and ecotone forest habitats (Pruetz 2006). Grasslands and woodlands have open tree canopies during the dry season, whereas evergreen species in gallery and ecotone forests provide a closed tree canopy year round (Figure 4.1). Botanical descriptions of vegetation types are provided elsewhere (Baldwin 1979, McGrew et al. 1981). Anthropogenic factors that shape the Fongoli landscape include permanent and seasonal settlements, horticulture, roads, foot trails, annual bush fires, wood collecting for timber and fuel, cattle ranging, sheep herding, and artisanal gold mining (Pruetz and LaDuke 2010, Pruetz and Kante 2010, Ndiaye 2011). About

4% of the chimpanzees' home range consists of cropland (Pruetz and Bertolani 2009). A thin layer of top soil and prominent bedrock material, composed primarily of laterite, characterize much of the landscape; alluvial soil is the dominant soil class for areas in crop rotation (Tappan et al. 2004). In southeastern Senegal, relatively poor soil quality limits cropland expansion

(Tappan et al. 2004).

We used two approaches to studying habitat selection in Fongoli chimpanzees: direct observations of feeding events by habituated study subjects using focal individual sampling

(Altmann 1974) and indirect observations of feeding behavior by measuring the depletion of food patches over time (Houle et al. 2006). Fongoli chimpanzee foraging behavior was assessed during two, annual baobab fruiting seasons, from November 2011 to January 2012 and

November 2012 to February 2013. During this time frame, between 33 and 34 individual chimpanzees were in the Fongoli study community, and all individuals were uniquely 116 recognizable in physical appearance. Community size fluctuated during this study due to migrations, births, and deaths. Adult males (2011-2012: N=10, 2012-2013: N=11) were the focus of this study, in part (see below) because adult males rarely displayed signals of nervousness or anxiety as a consequence of our physical presence around them each day (i.e., they were habituated to research personnel). The social ranks of most adult males were identified by patterns in submissive vocalizations (e.g., pant-grunts; Goodall 1986). However, due to frequent changes within the dominance hierarchy, including an alpha-male change and the introduction of three young males into the adult male hierarchy, we could not always identify a linear social hierarchy during the study period, i.e., there were periods of instability. Therefore, individuals were classified as having a high, middle, or low social ranking. Opportunistic data were collected on adult females, juveniles, and infants rather than systematically sampling behavior with continuous or instantaneous recording methods (sensu Altmann 1974). We did not use focal individual sampling for females and young because some females displayed signs of nervousness and agitation in response to our presence, which may be related to concern for the safety of their offspring (see Pruetz and Kante 2010). While the risk of poaching infants is rare at Fongoli (one recorded instance in 13 years; Pruetz and Kante 2010), minimizing contact between researchers and adult females may further decrease this risk. However, most females tolerate researchers when other group members are nearby, enabling us to record association patterns and female reproductive states. Although several instances of presumed intercommunity conflicts have occurred over the past decade, chimpanzees from neighboring groups were not observed during this study.

117

Food patches

Chimpanzees in general have a large dietary breadth. In their review of plant foods of habituated chimpanzees at long-term studies, Watts and co-workers (2012) reported that chimpanzees consume approximately 156 to 271 plant foods from 102 to156 plant species; presumably, each food has a unique nutritional profile. In the face of this diversity, combined with the expectation that the nutritional composition of foods influences foraging decisions

(Stephens and Krebs 1986), we simplified our research design by focusing on a single dietary item; we examined individuals’ habitat use while foraging for fruit of the baobab tree (Adansonia digitata), which is one of the top foods in their diets (Pruetz 2006). Although Fongoli chimpanzees are known to consume ripe baobab fruit as late as May, they intensively consume this food between the months of November and January, which coincides with the early to middle dry season at Fongoli. Baobab is a deciduous tree, with leaf abscission occurring during the early dry season for most trees (Wickens and Lowe 2008). Consequently, this lack of foliage enhances our ability to estimate food intake by focal individuals. Furthermore, baobab trees occupy a range of open- and closed-canopy habitats, and occur near and far from high-risk areas and drinking water sources at Fongoli. The baobab is used by relatively few frugivorous species due to the presence of a hard, woody exocarp that encases the fruit pulp and seeds. Chimpanzees, baboons (Papio hamadryas papio), and humans are the principal consumers, followed by vervet monkeys (Chlorocebus aethiops), and squirrels (Heliosciurus gambianus). Often, it is possible to identify consumer species from the remains of foraging activity, because each consumer may leave a suite of distinctive fruit processing signatures (Gašperšič 2008). Chimpanzees consume the pulp and seed kernel of unripe and ripe fruit. They tend to smash open fruits on hard substrates (e.g., anvils), use their canines as levers to tear apart the exocarp. This levering action 118 leaves large breaks and splits in the exocarp. In addition, chimpanzees discard the seed coat in unripe and semi-ripe fruits (Gašperšič 2008). Baboons often leave distinctive canine tooth marks on fruits that are bitten apart piecemeal.

Prior to the onset of the baobab fruiting season in 2011, we located baobab food patches at Fongoli with reconnaissance survey methods and records from previous field seasons (Pruetz, unpublished data). Two-thirds of patches at Fongoli were located with this approach, and the remaining patches were located during both fruiting study seasons. During the survey, we recorded the following features for each patch: (1) vegetation type, sensu McGrew and co- workers (1981), (2) luminescence at breast height or immediately above the herbaceous canopy with a light meter, (3) presence of human landmarks (cropland, mines, villages, roads) and surface water sources (permanent and seasonal), and (4) accessibility of tree crown to chimpanzees, as judged by trunk morphology and proximity to neighboring trees. We photographed the 200 patches located during the survey for archival purposes. The geographic coordinates of all patches were recorded with a global positioning system (GPS; e.g., Trimble®

GeoExplorer® 2005 Series) and visualized with mobile geographic information systems (GIS) software (ESRI®ArcPad). In addition, our team recorded the location of roads, water sources, fields and villages with GPS throughout the study period.

The tree canopy dramatically changes between wet and dry seasons in woodlands and grasslands at Fongoli, as evergreen species are rare. Following leaf abscission during the dry season, the tree canopy in these habitats is primarily open and with little shade cover for chimpanzees. Given that the baobab fruit season occurs during the dry season, we classified the vegetative context of each patch as open- or closed-canopy habitat. Furthermore, in transitional- canopy habitats, chimpanzees can move between open- and closed-canopy habitats during the 119 same feeding bout. For instance, we routinely observed individuals to collect fruit in an open- canopy habitat, then rapidly travel to a nearby closed-canopy habitat to consume that fruit. In order to account for food patches in these transitional areas, we used ArcMap (ESRI®) to add a

50 m buffer around all patches in our sample, and examined the vegetation within these buffers with survey notes or 1 to 4 m resolution multi-spectral satellite imagery, or both when possible.

Patches with gallery or ecotone forest (i.e., sources of shade during the dry season) within this 50 m radius were classified as closed-canopy habitat. Next, we used 0.5 m resolution pan-chromatic satellite imagery collected in 2010 to map active and fallow fields as well as rivers and streams that were missed during the field season. Euclidean distances (m) to water sources and fields were estimated for each patch with the Near tool in ArcMap.

All food patches located within the minimum convex polygon estimate (MCP; Moorcroft and Lewis 2006) of the Fongoli chimpanzees’ home range were included in this study. While new areas of their home range have been recorded in recent years, these discoveries are now rare after nine years of study post-habituation. Consequently, our survey area may be a slight underestimate, excluding only locations that are rarely used. Due to time constraints, we did not use strip transects (Buckland et al. 1993) throughout the entire home range of Fongoli chimpanzees. Furthermore, it was not possible to locate all baobab trees using satellite imagery alone. Therefore, our survey underrepresents areas that Fongoli chimpanzees rarely, if ever, use within the MCP home range. However, our surveys covered areas that are frequently used by the

Fongoli chimpanzees as well as cropland and villages that are in close proximity to intensively used areas of their home range. Baobab trees that were rooted within villages were excluded from the survey because, to the best of our knowledge, Fongoli chimpanzees do not travel within the villages' perimeters. In rare instances, two baobab trees that were adjacently rooted and had 120 overlapping tree crowns were classified as one patch. Food patches that did not bear fruit during the study period were excluded from the analysis of habitat selection (Season one: N=2, season two: N=12). Three trees were believed to be inaccessible to chimpanzees due to the lack of entrance routes to the tree crown and were not visited by focal individuals during the study period. Thus, they were eliminated from further analyses.

Behavioral follows

The foraging behavior of adult males in the Fongoli study community was the focus of direct behavioral observations. Travel was considered part of the foraging activity, except for cases clearly relating to non-foraging social interactions, such as short-distance movement involved in displaying, grooming, and mating (Nishida et al. 2010). Our goal was to uniformly sample all males each month. We used standardized and widely accepted behavioral sampling methods (Altmann 1974). Each day, we followed a single, adult male from dawn until dusk and continuously recorded his foraging activity, estimated to the nearest second. In rare instances, activity bouts were rounded to the nearest 30 seconds. Our goal was to locate a focal male before he left his sleeping nest at dawn and follow him until he constructed his sleeping nest at dusk. If the focal male was lost and could not be located among other individuals within 20 minutes, the observer switched to another focal male. This switching protocol was implemented in less than nine per cent of focal individual follows (N=4). When a focal male was capturing and consuming food, we closely observed his food processing behavior with the aid of binoculars, and verified these observations by inspecting feeding traces and fecal remains. Furthermore, the phenological state of plant food items (e.g., unripe, semi-ripe or ripe fruit) and the food parts consumed, such as endocarp or seed kernel, were recorded. The act of consuming food was separately timed from food capturing in instances when consuming and capturing were not concurrent. Pauses in 121 feeding behavior that exceeded ten seconds were removed from estimates of feeding duration.

Whenever possible, we recorded the number of food items consumed per food patch visit. We used feeding rates, defined as the number of food items consumed per unit of time, to estimate food intake (dry matter in grams) in cases when the focal male was out of view during a portion of a food patch visit. All instances of antipredator behavior, including vigilance, alarm-calling, fleeing and hiding, as well as the source of the disturbance (e.g., person, bicycle, dog, cattle, unknown), were recorded throughout the day. We tracked the location of focal males and food patch visits with a GPS. Geographic coordinates were collected at 60-second intervals while we were on the move, except during periods of poor signal reception. Concurrently, a secondary set of geographic coordinates was collected for focal males at 15-minute intervals by a field guide with a second GPS unit. Secondary geographic coordinates were used in cases of primary GPS unit failure.

In order to gauge social aspects of habitat selection (Goodall 1986), we considered the influence of female reproductive state, male social rank, and subgroup size and composition on adult male foraging behavior. We recorded a daily party size estimate each day, defined as all individuals seen together on a single day (Boesch 1996, Furichi 2009). In addition, we recorded the identities of individuals that visited the same food patch as the focal subject while that focal animal was in the food tree. The reproductive state of adult females impacts male fission-fusion behavior, where males are more likely to join larger parties when females in estrus are present

(Hashimoto et al. 2001). Thus, we estimated the estrus/anestrus state of females each day by ranking the swelling size of females’ anogenital region on a four-point scale, where a score of zero is completely flaccid and three is maximally tumescent (Pruetz and Lindshield 2011). We 122 recorded all instances of copulations, agonistic interactions involving the focal male, and cases of social rank reversals (i.e., dyadic changes in pant-grunt direction).

Although the focal individual approach provides a rich set of behavioral data, it requires a large sampling effort per study subject. In order to complement our focal data, we simultaneously collected indirect evidence of food patch visits through monitoring baobab fruit depletion over time among patches. In the days immediately following a food patch visit, we would revisit these patches and visually count the number of edible fruits in or below the tree.

Counts of fruits within the tree crown were conducted by a single researcher with the highest accuracy and precision, as determined by inter-observer reliability testing (Martin and Bateson

1986). Feeding remains were collected and deposited in a refuse pile in order to facilitate future counts at each patch. Fallen fruits, which are routinely eaten by chimpanzees, were scored with a unique number and left in situ. Consumer species were identified by established patterns in baobab processing behavior (Gašperšič 2008), but in some cases the consumers were unknown.

Feeding traces left by identified, non-chimpanzee consumers were excluded from the analysis.

This approach was inspired by the giving-up density concept (Brown 1988, Houle et al. 2006) except that we could not control or know the total amount of fruit available in all patches prior to all visits by all study subjects, nor could we immediately assess fruit patch depletion in cases where individuals revisited patches before our team conducted fruit counts. Despite these shortcomings, we were able to accrue a large sample of indirect feeding observations (N=605).

Food sample collection and chemical analyses

Given that domesticated and wild plant foods vary in their nutritional profiles, even within the same plant species and part (Chapman et al. 2002), we considered the impact of such variation among food patches on habitat selection. In addition, we evaluated the potential effect 123 of soil quality on nutritive quality of fruit (Appendix). Baobab fruits were collected among and within patches. The entire and edible (endocarp and seed) portions of fruits were measured for wet and field-dried mass. In addition, we also measured the length and circumference of whole, fresh fruits. Each sample came from a single, baobab fruit. Samples were field dried in a solar dehydrator (Privette 2005) or gas oven (Coleman® camp oven) at a drying temperature range of

40 to 50° C until they reached a constant weight (Conklin-Britain et al. 1994). After drying, the samples were stored in dark, air-tight containers with a food-safe desiccant. No baobab fruit samples were damaged by mold or insects. Samples were brought to the laboratory for oven drying at 105° ± 2° C overnight to remove adsorbed moisture (Rothman et al. 2012). For nutritional analyses, one sample consisted of the endocarp of a single ripe fruit. Samples were milled with a Wiley® Mini-Mill using a number 20 delivery unit (0.85 mm mesh screen) at the

Grain Quality Laboratory at Iowa State University in Ames, Iowa. Samples were analyzed for macronutrient and condensed tannin composition at the CUNY-Hunter College Primate

Nutritional Ecology Laboratory in New York City, New York, following standard laboratory procedures (Ortmann et al. 2006, Rothman et al. 2012) and presented on a percent of dry matter basis. The full suite of chemistry procedures was not performed for all samples. We estimated neutral detergent fiber, acid detergent fiber, and acid detergent lignin to assess the fiber fractions of hemicellulose, cellulose, and lignin (Van Soest 1994). Crude fat was estimated with the ether extraction method (AOCS 2005). Crude protein and acid detergent insoluble protein were estimated with the Dumas combustion method multiplied by 6.25 (Licitra et al. 1996, Rothman et al. 2008). Samples were ashed in a muffle furnace overnight at 550° C as a crude estimate of inorganic matter (Rothman et al. 2012). Total nonstructural carbohydrates were estimated through subtracting the sum of neutral detergent fiber, crude protein, crude fat, and crude ash 124 from 100, where the remainder is an estimate of total nonstructural carbohydrates (Van Soest et al. 1991). The phenol-sulfuric acid assay was used to estimate water-soluble carbohydrates with a sucrose standard (DuBois et al. 1956). Sucrose, D-glucose, and D-fructose were measured enzymatically with the UV method (R-Biopharm, Darmstadt, Germany) for a subset of samples.

In addition, total starch was estimated with the amyloglucosidase and α-amylase method

(Megazyme total starch assay; McLeary et al. 1997). The presence or absence of condensed tannins were surveyed with the acid-butanol assay (Waterman and Mole 1994). An initial screening for condensed tannins qualitatively tested positive. Consequently, we created a baobab-specific standard curve (Rothman et al. 2009) to estimate the fraction of condensed tannins.

Statistical analyses

We examined data for normality and used data transformation procedures to approximate the normal distribution (Sokal and Rohlf 2012) prior to using parametric tests. When data violated the assumptions of normality or large sample sizes for parametric testing, we used nonparametric rank correlation tests and evaluated group differences with Kruskal-Wallis or chi- square tests. The significance level was set at α=0.05. Variation in the nutritional composition of baobab ripe fruit among patches was assessed with the Fmax-test (Sokal and Rohlf 2012).

Statistical procedures were conducted in R version 3.0.2 (R Core Team 2014).

Ethical standards

Our research project was approved by the République du Sénégal, Ministère de l’Environnement et du Développement Durable, Direction des Eaux, Forêts, Chasses et de la

Conservation des Sols. The data collection protocol was approved by the Institutional Animal

Care and Use Committee at Iowa State University. The United States Department of Agriculture 125 approved the import of Adansonia digitata samples. Our research conforms to the code of ethics of the American Association of Physical Anthropologists.

Results Characteristics of real and potential food patches

We identified 304 baobab patches within the Fongoli study community’s home range that bore fruit during the study period. Baobab trees were spatially clustered, as the observed mean distance between nearest neighbors was smaller than expected (expected = 287 m, observed =

156 m, nearest neighbor ratio = 0.57, Z = -14.31, P < 0.0001). Baobab trees were highly aggregated at ten locations within the home range, and Fongoli chimpanzees intensively used the food patches within these baobab clusters. Therefore, we included an additional patch characteristic, known as cluster density, or the number of baobab patches within a set unit of area. We used the half-lengths of these ten clusters to scale the area for this cluster density estimate. The average half-length of the Euclidean distance between the two trees on either end of a single cluster was 390 m (Range=102-760 m). We calculated cluster density as the number of patches within a 195 m radius of each patch using the point density tool in ArcGIS. The mean cluster density was 3.9 patches (Range=1-13). As expected, illuminance (klux) was significantly higher in open-canopy habitats, where there is less shade cover (Kruskal-Wallis χ2 = 24.76, df =

1, P < 0.00001). Most trees were located in open (73%) rather than closed (27%) canopy habitats

(Figure 4.2). However, baobabs were located within or near closed-canopy habitats more than would be expected by chance (Pearson χ2 = 74.54, P < 0.0001) when considering that closed- canopy habitats cover approximate 2% of the Fongoli chimpanzees’ home range (Pruetz and

Bertolani 2009). Furthermore, patches in closed-canopy habitats were closer to surface water than open-canopy patches (ANOVA, F1,85 = 8.02, P < 0.006). These patterns are likely explained by the prevalence of baobab trees along the edges of ravines near gallery forest. Although 126 surface water is a scarce resource at Fongoli (Pruetz and Bertolani, 2009), the study occurred during the early dry season when several, ephemeral streams had flowing water or shallow pools.

On average, baobab patches were 763 m (SD = 604) from surface water during the study period

(Figure 4.2). About 13% of all baobab patches (N = 39) were within 50 m of roads, cropland, artisanal gold mines or villages and, on average, patches were 518 m (SD = 415) from one of these anthropogenic landmarks (Figure 4.2). Three artisanal gold mines (two active, one abandoned in 2011 before the study period) and two active mining facilities (e.g., storage shed) were located during the study period.

Baobab fruit counts

We monitored 247 patches for fruits and feeding traces during the two field seasons.

Within this sample, we documented feeding remains at 74% of these patches (N=184) but counted edible fruits in the patch canopy and on the ground for all patches in this sample. The remaining 57 surveyed baobab patches were not monitored for lack of time. We took a subsample (N=368) of the fruit count measurements, where each patch was represented no more than once per baobab season. When multiple measurements were taken within a season on the same patch, the last patch measurement of the season was included in this subsample. There was no difference in fruit counts between seasons (Kruskal-Wallis χ2 = 0.22, df=1, P>0.05) but fruit counts changed within a season (Kruskal-Wallis χ2 = 19.64, df=2, P<0.001) as expected when consumers deplete patches over time. Therefore, we divided the baobab season into early, middle, and late periods, with each period lasting between 19 (late) and 21 (early, middle) days

(Figure 4.3). Overall, fewer edible fruits remained in a patch as baobab cluster density increased

(Kendall’s τ =-0.13, z=-2.98, P=0.001) and this trend remained constant across all periods of the baobab season (Table 4.1). Overall, more fruits remained in patches closer to anthropogenic 127 landmarks (Kendall’s τ =-0.19, z=-2.98, P<0.0001), but only during the middle and late season

(Table 4.1). The number of fruits in a patch was lower in closed-canopy habitats during the late baobab season only (Kruskal-Wallis χ2 = 9.87, df=1, P<0.01), that is, patches depleted faster in closed-canopy habitats. Although patches in closed-canopy patches were often closer to surface water than open-canopy patches, there was no correlation between fruit counts and proximity to surface water (Table 4.1).

Nutritional composition of baobab fruit

Baobab fruit dominated the diet of focal male subjects during the study period.

Individuals fed from a median of five food species per day during the baobab season (range: 2-

10), but spent 67% of feeding time on baobab fruit endocarp and seed. None of the Fongoli chimpanzees consumed cultivars during the study period, which supports previous findings that crop-raiding by Fongoli chimpanzees is absent or rare (Pruetz and Lindshield 2011). Fruit endocarp (N=105) was relatively high in water-soluble carbohydrates (WSC, mean=54.2% on a dry matter (DM) basis) but low in crude protein (CP, 3.3% DM) and crude fat (0.9% DM). Seed kernels (N=4) were complementary to endocarp, as they were high in CP (38.5% DM) and crude fat (29.4% DM). Small sample sizes for ripe seed kernels (N=3), and unripe endocarp (N=1) and kernel (N=1) limits statistical comparisons. However, a qualitative comparison shows that moisture content was the greatest difference between unripe and ripe fruit (Figure 4.4). There was a wide range of variation in the nutritive quality of ripe fruit endocarp. The fractions of fiber, fat, protein, ash, and condensed tannins varied little among ripe fruits in terms of standard deviation from the mean, but higher standard deviations were observed for carbohydrates (Table

4.2). The average coefficient of variation for all nutrients and antifeedants was 31.6% (range 3.6-

78.0%) for ripe fruit endocarp (Table 4.2). To further explore this variation, we compared the 128

variance in endocarp nutritional quality with the Fmax-test from nine patches that had three to four fruit samples representing each tree; variance was homogeneous among patches (Table 4.2).

Furthermore, the nutritional composition of ripe fruit from patches with feeding traces did not differ from those patches without trace evidence of feeding activity (Table 4.2). We found a significant association between soil quality and fruit nutritive values, but these values were not associated with feeding traces (Appendix).

Foraging party size and composition

Focal male study subjects visited 62 different baobab patches during 537 contact hours with chimpanzee parties between December 2011 to January 2012 and November 2012 to

January 2013. Daily party sizes and the numbers of estrus females observed in a single day were not correlated to baobab fruit intake (Table 4.1), but were strongly correlated to each other

(Spearman’s ρ= 0.61, P<0.001), which supports the hypothesis that female estrus state is a strong predictor of fission-fusion dynamics in chimpanzee societies (Anderson et al. 2002; Hashimoto et al. 2001; Mitani et al. 2002). Although these measures of daily party size and composition were not linked to baobab feeding behavior, the size and composition of foraging parties during feeding bouts were associated with focal individuals’ food intake (grams of dry matter ingested) and intake rate per patch visit. Focal individuals increased food intake with the number of other adult males in a food patch (Spearman’s ρ= 0.37, P<0.01), while the number of females and dependent offspring in a patch was correlated to a decrease in ingestion rate (Spearman’s ρ= -

0.35, P<0.01).

Male foragers’ associations with canopy cover and surface water

Male subjects visited open-canopy patches (71%) more often than closed-canopy patches

(29%), and these visits mirrored the baobab population within each respective habitat. Focal 129 subjects visited patches in closed-canopy habitats more often than expected after adjusting for the rarity of closed-canopy habitat within their home range (Pearson χ2 = 17.96, P < 0.0001), supporting previous findings that savanna chimpanzees selectively use closed-canopy habitats

(Pruetz and Bertolani 2009). Furthermore, individuals increased ingestion rate while using open- canopy patches (Kruskal-Wallis χ2 = 6.35, df=1, P=0.01) but food intake during a feeding bout did not change (Table 4.1). Surface water was an important source of water intake for Fongoli chimpanzees during the baobab season, as focal individuals drank from rivers, streams, or puddles on 89% days when the presence or absence of drinking was reliably recorded (N=27 days). Food intake decreased with increasing distance to water sources (Spearman’s ρ= -0.23,

P<0.05), but there was no change in feeding rate and no association between water proximity and unripe or ripe fruit consumption (Table 4.1).

Male foragers’ associations with human activity

Focal individuals did not completely avoid patches in close proximity to villages, fields, mines, and roads, but antipredator behavior (Kruskal-Wallis χ2 = 9.23, df = 1, P < 0.01) and the number of males in a foraging party (Spearman’s ρ= -0.26, P<0.05) both increased when focal individuals were closer to these anthropogenic features. Moreover, focal subjects only visited patches rooted within active fields and mines on days when people were absent (N=4). Fongoli chimpanzees were frequently vigilant when non-project personnel were detected, and they exhibited antipredator behavior towards humans (67.3% of events) more often than all other animals combined (Pearson χ2 = 6.93, P < 0.01), including other Fongoli chimpanzees (9.7%), cattle (7.7%), dogs (2%), birds (2%), or a combination of non-chimpanzee animals, such as sheep and humans (7.7%). Food intake varied unexpectedly with antipredator behavior (Figure

4.5). The amount of fruit ingested per patch visit increased for cases when focal subjects 130 displayed antipredator behavior (Kruskal-Wallis χ2 = 6.58, df = 1, P = 0.01), but fruit ingestion rate did not change (Kruskal-Wallis χ2 = .063, df = 1, P > 0.05). This finding was explained by the concomitant increase in ripe fruit intake (Kruskal-Wallis χ2 = 4.19, df = 1, P < 0.05) (Figure

4.5). In addition, antipredator behavior varied with each male’s social ranking in the community.

High-ranking males exhibited less vigilance than mid-ranking males (Fisher’s exact test,

P<0.05), but there was no difference between high and low (Fisher’s exact test, P>0.05), or middle and low (Fisher’s exact test, P>0.05) ranked males.

Discussion Antipredator foraging strategy

Earlier studies have shown that chimpanzees take risks to capture high value foods. For instance, Hockings and co-workers (2007) demonstrate that Bossou chimpanzees in Guinea routinely forage for cultivars, such as papaya (Carica papaya), and that this crop-raiding behavior is associated with increasing time spent feeding. However, this research neglects the impact of predation risk on patterns of consumption involving wild plant foods and does not quantify the impact of risk on food intake. Overall, wild plant foods should differ from cultivars in abundance, spatial distribution, nutritional quality, and degree of risk experienced by foragers.

In the present study, we assessed the impact of predation risk on adult males’ foraging decisions for a key, non-cultivated food source, Adansonia digitata. In hypothesizing that Fongoli chimpanzee foraging is sensitive to predation risk, we predicted that food trees near anthropogenic features were visited less often. When individuals visited patches near these features, we expected that individuals would attempt to minimize the probability of predator attack by increasing foraging party size and increasing food intake rate.

Several of our findings support the predation-sensitive foraging hypothesis (Sinclair and

Arcese 1995). The number of fruits remaining in a patch was negatively correlated with patch 131 proximity to anthropogenic landmarks (roads, fields, villages, gold mines), which indicates that, overall, chimpanzees prefer to feed in baobab patches where there is a lower rate of human encounters. Furthermore, the number of males in a foraging party increased during visits to patches nearby these landmarks. Although there was no change in food intake with distance to anthropogenic features, food intake and ripe fruit consumption increased when chimpanzees exhibited antipredator responses towards people. This finding indicates that adult males at

Fongoli are more likely to forage in riskier habitats when the nutritional reward is high.

Furthermore, reducing feeding competition with females and dependent offspring may be another incentive for males to exploit riskier patches. To wit, their feeding rates decreased when more females and young, but not males or estrus females, were in a food patch.

These results complement previous work in showing that chimpanzees may invest more time, rather than less, in foraging for risky but high-quality food sources (Hockings et al. 2012), and additionally demonstrates that this strategy applies to wild foods distributed throughout the

Fongoli community’s home range. Unexpectedly, our findings show that feeding rate remained relatively constant between higher and lower risk areas. Thus, Fongoli males increased food intake through increasing feeding duration at times when predation risk was elevated. This is the first study to demonstrate a positive nutritional response to predation risk in great apes that involves wild food sources. Furthermore, despite a local taboo against hunting chimpanzees and despite complete habituation of study subjects towards project personnel, subjects often exhibited antipredator behavior towards non-project personnel. A natural interpretation of these results is that Fongoli chimpanzees are wary of non-project personnel, but they do not often abandon foraging efforts because the risk of imminent, mortal danger is relatively low. Rather, chimpanzees may maximize foraging investment when experiencing “moderate” levels of risk 132 from human activity. Given that chimpanzees show antipredator responses to people and alter their food intake upon detecting non-project personnel, we suspect that chimpanzees visit food trees in high-risk areas because nutritional benefits outweigh the immediate risks. One limitation is worth addressing, however. While the hypothesis of predation-sensitive foraging was qualitatively supported, we could not test for a relationship between food intake and the presence of project personnel; it is possible that focal subjects change their behavior in the absence of observers (sensu Crofoot et al. 2010). Therefore, future research should consider the role of observers of chimpanzees’ risk-taking activities while foraging.

Foraging in a savanna mosaic

Differences in resource abundance and distribution among vegetative habitats impact home range use for chimpanzees occupying extreme savanna mosaic environments. For example, Pruetz and Bertolani (2009) showed that Fongoli chimpanzees spend most of their time in evergreen forest habitats, especially during the dry season when shade and water are scarce, but forage in woodlands for the majority of their plant food sources. This study assesses the role of habitat on adult males’ decisions to forage among Adansonia digitata trees rooted within a mosaic of savanna habitats. We predicted that Fongoli chimpanzees would respond to differences in habitat quality within this mosaic by increasing food intake in open-canopy habitats to offset the higher metabolic costs of foraging, and, due to water limitation, preferentially using food patches closer to surface water.

Chimpanzees visited baobab trees in closed-canopy habitats more than expected based on their availability, and this finding is likely explained by the tendency for baobab trees to occupy closed habitats within the Fongoli chimpanzees’ home range. As expected, male focal subjects increased food intake rate (dry mass hr-1) in open-canopy habitats. Open habitats expose 133 chimpanzees to higher levels of solar radiation during the dry season and are significantly higher in temperature, overall (Pruetz 2007). Thus, heat stress and dehydration are more likely to occur with increasing use of open-canopy habitats. Although food intake rate per feeding bout was higher in open habitats, total food intake (dry mass) per bout between open- and closed-canopy habitats was similar. These findings support that hypothesis that Fongoli chimpanzees increase feeding efficiency in open habitats to minimize metabolic costs. It may be the case that feeding is more relaxed in closed-canopy habitats where shade cover is more abundant. In a similar vein, food intake decreased with increasing distance from drinking water sources. Thus, Fongoli chimpanzees appear to factor travel distance to surface water into their foraging routes during the early to middle dry season.

Significance to savanna chimpanzee ecology and conservation

This study advances research on the link between foraging and antipredator strategy in chimpanzees. Given that Fongoli chimpanzees are predator-sensitive foragers around people that usually do not actively hunt them, we suggest that future studies consider how chimpanzees’ foraging decisions change when faced with different predator strategies (sensu Lone et al. 2014), either from sympatric human hunters or large carnivores such as lions, leopards, or spotted hyenas (Crocuta crocuta). At a more applied level, this finding has important implications for chimpanzee conservation and human-wildlife conflict. First, Fongoli chimpanzees forage for wild foods in areas within or immediately adjacent to anthropogenic features, thus bringing them into close contact with people, crops, and livestock. Increasing habitat encroachment, especially with the continuing trend of gold mining in southeastern Senegal, or decreasing fear of people could ultimately intensify conflict with local residents at Fongoli through aggression or incipient crop-raiding, as has been suggested elsewhere (McLennan 2013; McLennan and Hill 2010). 134

Second, the overlap in land use between people and chimpanzees may elevate the risk of disease transmission (Zommers et al. 2013), as Fongoli chimpanzees are at risk of fecal-oral transmission of pathogens, especially at artisanal gold mines where disease prevention strategies, such as latrines, are not widely used.

Acknowledgements We thank the Government of Sénégal, Ministère de l’Environnement et du

Développement Durable, Direction des Eaux, Forêts, Chasses et de la Conservation des Sols for research authorization. Our data collection protocol was approved by the Institutional Animal

Care and Use Committee at Iowa State University and conforms to the code of ethics of the

American Association of Physical Anthropologists. We are grateful for the financial support received from the Leakey Foundation, Rufford Foundation, Iowa State University and CUNY-

Hunter College. Vital research assistance was provided by Ulises Villalobos-Flores, Michel

Sadiakho, Anna Olson, El Hadji Sakho, Waly Camara, Mody Camara, Dondo Kante and Jenny

Paltan. For research facilitation, we thank Sylvia Ortmann, Robin McNeely, Gray Tappan,

Charles Hurburgh, Martha Jeffrey, Morty Marshack and Beverly Marshack.

Literature cited Altmann J. 1974. Observational study of behavior - sampling methods. Behaviour 49:227-267.

Anderson DP, Nordheim EV, Boesch C, and Moermond TC. 2002. Factors influencing fission-

fusion grouping in chimpanzees in the Taï National Park, Côte d’Ivoire. In: Boesch C,

Hohmann G, and Marchant LF, editors. Behavioural diversity in chimpanzees and bonobos.

Cambridge: Cambridge University Press. p 90-101.

AOCS. 2005. Approved procedure AM 5-04: rapid determination of oil/fat utilizing high

temperature solvent extraction. Urbana, IL. 135

Baldwin PJ, McGrew WC, Tutin CEG. 1982. Wide-ranging chimpanzees at Mt Assirik, Senegal.

Int J Primatol 3:367-385.

Boesch C. 2009. The real chimpanzee. New York: Cambridge University Press.

Boesch C, Boesch-Achermann H. 2000. The chimpanzees of the Taї forest. Oxford: Oxford

University Press.

Bogart SL, Pruetz JD. 2011. Insectivory of savanna chimpanzees (Pan troglodytes verus) at

Fongoli, Senegal. Am J Phys Anthropol 145:11-20.

Boyer KM. 2011. Chimpanzee conservation in light of impending iron ore mining project in

southeastern Senegal. Ann Arbor: Iowa State University.

Brown JS. 1988. Patch use as an indicator of habitat preference, predation risk, and competition.

Behav Ecol Sociobiol 22:37-47.

Brown JS. 1999. Vigilance, patch use and habitat selection: foraging under predation risk. Evol

Ecol Research 1:49-71.

Brown JS, and Kotler BP. 2007. Foraging and the ecology of fear. In: Stephens DW, Brown JS,

and Ydenberg RC, editors. Foraging: behavior and ecology. Chicago: University of Chicago

Press. p 437-480.

Chancellor RL, Rundus AS, Nyandwi S. 2012. The influence of seasonal variation on

chimpanzee (Pan troglodytes schweinfurthii) fallback food consumption, nest group size,

and habitat use in Gishwati, a montane rain forest fragment in Rwanda. Int J Primatol

33:115-133.

Chivers DJ. 1998. Measuring food intake in wild animals: primates. P Nutr Soc 57:321-332.

Coleman BT, Hill RA. 2014. Living in a landscape of fear: the impact of predation, resource

availability and habitat structure on primate range use. Anim Behav 88:165-173. 136

Crofoot MC, Lambert TD, Kays R, Wikelski MC. 2010. Does watching a monkey change its

behaviour? Quantifying observer effects in habituated wild primates using automated

radiotelemetry. Anim Behav 80:475-480.

Dominguez-Rodrigo M. 2014. Is the "savanna hypothesis" a dead concept for explaining the

emergence of the earliest hominins? Curr Anthropol 55:59-81.

Duvall CS. 2000. Important habitat for chimpanzees in Mali. Afr Stud Monogr 21:173-203.

Effiom EO, Nuñez-Iturri G, Smith HG, Ottosson U, and Olsson O. 2013. Bushmeat hunting

changes regeneration of African rainforests. Proceedings of the Royal Society B: Biological

Sciences 280(1759).

Gilby IC, Eberly LE, Pintea L, Pusey AE. 2006. Ecological and social influences on the hunting

behaviour of wild chimpanzees, Pan troglodytes schweinfurthii. Anim Behav 72:169-180.

Halloran AR, Cloutier CT, Sesay PB. 2013. A previously undiscovered group of chimpanzees

(Pan troglodytes verus) is observed living in the Tonkolili District of Sierra Leone. Am J

Primatol 75:519-523.

Hashimoto C, Furuichi T, and Tashiro Y. 2001. What Factors Affect the Size of Chimpanzee

Parties in the Kalinzu Forest, Uganda? Examination of Fruit Abundance and Number of

Estrous Females. International Journal of Primatology 22(6):947-959.

Head JS, Robbins MM, Mundry R, Makaga L, Boesch C. 2012. Remote video-camera traps

measure habitat use and competitive exclusion among sympatric chimpanzee, gorilla and

elephant in Loango National Park, Gabon. J Trop Ecol 28:571-583.

Hernandez-Aguilar RA. 2009. Chimpanzee nest distribution and site reuse in a dry habitat:

implications for early hominin ranging. J Hum Evol 57:350-364. 137

Hernandez-Aguilar RA, Moore J, Pickering TR. 2007. Savanna chimpanzees use tools to harvest

the underground storage organs of plants. PNAS 104:19210-19213.

Hernandez-Aguilar RA, Moore J, Stanford CB. 2013. Chimpanzee nesting patterns in savanna

habitat: environmental influences and preferences. Am J Primatol 75:979-994.

Hicks T, Darby L, Hart J, Swinkels J, January N, Menken S. 2010. Trade in orphans and

bushmeat threatens one of the Democratic Republic of the Congo's most important

populations of eastern chimpanzees (Pan troglodytes schwinfurthii). Afr Primates 7:1-18.

Hicks TC, Roessingh P, Menken SBJ. 2012. Reactions of Bili-Uele chimpanzees to humans in

relation to their distance from roads and villages. Am J Primatol 74:721-733.

Higgins SI, Bond WJ, February EC, Bronn A, Euston-Brown DIW, Enslin B, Govender N,

Rademan L, O'Regan S, Potgieter ALF, Scheiter S, Sowry R, Trollope L, Trollope WSW.

2007. Effects of four decades of fire manipulation on woody vegetation structure in savanna.

Ecology 88:1119-1125.

Hockings KJ, Anderson JR, Matsuzawa T. 2012. Socioecological adaptations by chimpanzees,

Pan troglodytes verus, inhabiting an anthropogenically impacted habitat. Anim Behav

83:801-810.

Hockings KJ, Humle T, Andersen JR, Biro D, Sousa C, Ohashi G, Matsuzawa T. 2007.

Chimpanzees share forbidden fruit. PLoS ONE 2:886.

Hockings KJ, Yamakoshi G, Kabasawa A, Matsuzawa T. 2009. Attacks on local persons by

chimpanzees in Bossou, Republic of Guinea: long-term perspectives. Am J Primatol 71:1-

10. 138

Kay CE. 2007. Were native people keystone predators? A continuous-time analysis of wildlife

observations made by Lewis and Clark in 1804-1806. Canadian Field-Naturalist 121(1):1-

16.

Kortland A. 1983. Marginal habitats of chimpanzees. J Hum Evol 12:231-278.

Kotler BP, Dickman CR, Brown JS. 1998. The effects of water on patch use by two Simpson

Desert granivores (Corvus coronoides and Pseudomys hermannsburgensis). Aust J Ecol

23:574-578.

Lima SL, Dill LM. 1990. Behavioral decisions made under the risk of predation - a review and

prospectus. Can J Zool 68:619-640.

Link A, Di Fiore A. 2013. Effects of predation risk on the grouping patterns of white-bellied

spider monkeys (Ateles belzebuth belzebuth) in western Amazonia. Am J Phys Anthropol

150:579-590.

Makin DF, Payne HFP, Kerley GIH, Shrader AM. 2012. Foraging in a 3-D world: how does

predation risk affect space use of vervet monkeys? J Mammal 93:422-428.

Martin P, Bateson P. 1986. Measuring behaviour: an introductory guide. Cambridge, UK:

Cambridge University Press.

McGrew W. 2010. Christophe Boesch: The real chimpanzees. Sex strategies in the forest.

Primates 51:189-190.

McGrew WC, Baldwin P, Tutin CEG. 1981. Chimpanzees in a hot, dry and open habitat: Mt.

Assirik, Senegal, West Africa. J Hum Evol 10:217-244.

McLennan MR. 2013. Diet and feeding ecology of chimpanzees (Pan troglodytes) in Bulindi,

Uganda: foraging strategies at the forest-farm interface. Int J Primatol 34:585-614. 139

McLennan MR, Hill CM. 2010. Chimpanzee responses to researchers in a disturbed forest–farm

mosaic at Bulindi, western Uganda. Am J Primatol 72:907-918.

Mitani JC, Watts DP, and Lwanga JS. 2002. Ecological and social correlates of chimpanzee

party size and competition. In: Boesch C, Hohmann G, and Marchant LF, editors.

Behavioural diversity in chimpanzees and bonobos. Cambridge: Cambridge University

Press. p 102-111.

Molokwu MN, Nilsson JA, Ottosson U, Olsson O. 2010. Effects of season, water and predation

risk on patch use by birds on the African savannah. Oecologia 164:637-645.

Moorcroft PR, Lewis MA. 2006. Mechanistic home range analysis. Princeton, NJ: Princeton

University Press.

Morgan D, Sanz C. 2003. Naive encounters with chimpanzees in the Goualougo Triangle,

Republic of Congo. Int J Primatol 24:369-381.

Morgan D, Sanz C, Onononga JR, Strindberg S. 2006. Ape abundance and habitat use in the

Goualougo Triangle, Republic of Congo. Int J Primatol 27:147-179.

Nakagawa N. 2009. Feeding rate as valuable information in primate feeding ecology. Primates

50:131-141.

Nakazawa N, Hanamura S, Inoue E, Nakatsukasa M, Nakamura M. 2013. A leopard ate a

chimpanzee: first evidence from East Africa. J Hum Evol 65:334-337.

Ndiaye P, Galat G, Galat-Luong A, Nizinski G. 2013. Note on the seasonal use of lowland and

highland habitats by the West African Chimpanzee Pan troglodytes verus (Schwarz, 1934)

(Primates: Hominidae): Implications for its conservation. J Threatened Taxa 5:3697-3700.

Nishida T. 2012. Chimpanzees of the lakeshore: natural history and culture at Mahale.

Cambridge: Cambridge University Press. 140

Nishida T, Zamma K, Matsusaka T, Inaba A, McGrew WC. 2010. Chimpanzee behavior in the

wild. Tokyo: Springer.

Olsson O, Molokwu MN. 2007. On the missed opportunity cost, GUD, and estimating

environmental quality. Isr J Ecol Evol 53:263-278.

Pascual-Garrido A, Umaru B, Allon O, Sommer V. 2013. Apes finding ants: predator-prey

dynamics in a chimpanzee habitat in Nigeria. Am J Primatol 75:1231-1244.

Pienkowski MW, Watkinson AR, Kerby G, Naughton-Treves L, Treves A, Chapman C,

Wrangham R. 1998. Temporal patterns of crop-raiding by primates: linking food availability

in croplands and adjacent forest. J Appl Ecol 35:596-606.

Pruetz J. 2007. Evidence of cave use by savanna chimpanzees (Pan troglodytes verus) at

Fongoli, Senegal: implications for thermoregulatory behavior. Primates 48:316-319.

Pruetz J, Kante D. 2010. Successful return of a wild infant chimpanzee (Pan troglodytes verus)

to its natal group after capture by poachers. Afr Primates 7:35-41.

Pruetz J, Lindshield S. 2012. Plant-food and tool transfer among savanna chimpanzees at

Fongoli, Senegal. Primates 53:133-145.

Pruetz JD. 2006. Feeding ecology of savanna chimpanzees (Pan troglodytes verus) at Fongoli,

Senegal. In: Hohmann G, Robbins MM, Boesch C, editors. Feeding ecology in apes and

other primates. Cambridge, UK: Cambridge University Press. p 161-182.

Pruetz JD, and Bertolani P. 2009. Chimpanzee (Pan troglodytes verus) behavioral responses to

stresses associated with living in a savanna-mosaic environment: implications for hominin

adaptations to open habitats. Paleoanthropology:252-262. 141

Pruetz JD, Fulton SJ, Marchant LF, McGrew WC, Schiel M, Waller M. 2008. Arboreal nesting

as anti-predator adaptation by savanna chimpanzees (Pan troglodytes verus) in southeastern

Senegal. Am J Primatol 70:393-401.

R Core Team. 2014. R: A Language and Environment for Statistical Computing. Vienna,

Austria: R Foundation for Statistical Computing. http://www.R-project.org.

Reid RS, Ellis JE. 1995. Impacts of pastoralists on woodlands in south Turkana, Kenya:

livestock- mediated tree recruitment. Ecol Appl 5:978-992.

Rieucau G, Vickery WL, Doucet GJ. 2009. A patch use model to separate effects of foraging

costs on giving-up densities: an experiment with white-tailed deer (Odocoileus virginianus).

Behav Ecol Sociobiol 63:891-897.

Rothman JM, Chapman CA, Van Soest PJ. 2012. Methods in primate nutritional ecology: a

user's guide. Int J Primatol 33:542-566.

Schülke O, Chalise MK, Koenig A. 2006. The importance of ingestion rates for estimating food

quality and energy intake. Am J Primatol 68:951-965.

Sinclair ARE, and Arcese P. 1995. Population Consequences of Predation-Sensitive Foraging:

The Serengeti Wildebeest. Ecology 76(3):882-891.

Skinner MF, Pruetz JD. 2012. Reconstruction of periodicity of repetitive linear enamel

hypoplasia from perikymata counts on imbricational enamel among dry-adapted

chimpanzees (Pan troglodytes verus) from Fongoli, Senegal. Am J Phys Anthropol 149:468-

482.

Sokal R, and Rohlf F. 2012. Biometry. New York: W.H. Freeman and Company.

Stephens D, Brown J, Ydenberg R, editors. 2007. Foraging: behavior and ecology. Chicago, IL:

The University of Chicago Press. 142

Stephens DW, Krebs JR. 1986. Foraging theory. Princeton: Princeton University Press.

Stewart FA, Pruetz JD. 2013. Do Chimpanzee Nests Serve an Anti-Predatory Function? Am J

Primatol 75:593-604.

Tweheyo M, Hill CM, Obua J. 2005. Patterns of crop raiding by primates around the Budongo

Forest Reserve, Uganda. Wildlife Biol 11:237-247.

Watts DP, Potts KB, Lwanga JS, Mitani JC. 2012. Diet of chimpanzees (Pan troglodytes

schweinfurthii) at Ngogo, Kibale National Park, Uganda, 1. diet composition and diversity.

Am J Primatol 74:114-129.

Zommers Z, Macdonald DW, Johnson PJ, Gillespie TR. 2013. Impact of human activities on

chimpanzee ground use and parasitism (Pan troglodytes). Conserv Letters 6:264-273.

Zuberbühler K, Jenny D. 2002. Leopard predation and primate evolution. J Hum Evol 43:873-

886. 143

144

145

Figure. 4.1. Habitat types during the early dry season (December 2011) at Fongoli: (a) grassland,

(b) cultivated field, (c) woodland, and (d) gallery forest. 146

Figure. 4.2. Baobab patches in relation to canopy cover (open circle = open canopy, filled circle

= closed canopy), and patch proximity to nearest water source and anthropogenic landmark. 147

Figure. 4.3. Histogram of fruits (log N) remaining within a baobab patch for the early, middle, and late periods of the fruiting season. 148

Figure. 4.4. Qualitative comparison of unripe and ripe baobab fruit nutritional chemistry. All nutritive values are expressed as a % of dry matter.

149

Figure. 4.5. Food intake changes with antipredator behavior during a feeding bout. There is no difference in (a) food intake rate or (b) unripe fruit intake, but (c) overall fruit intake and (d) ripe fruit intake significantly increases with antipredator behavior. 150

CHAPTER V ON THE MARGINS OF SAVANNA CHIMPANZEE (PAN TROGLODYTES VERUS) HABITAT IN SENEGAL Manuscript will be submitted to the Journal of Biogeography

Stacy M. Lindshield1*, Kelly Boyer Ontl1, Maja Gašperšič2, Assane Goudiaby3, Papa Ibnou

Ndiaye4, Gray Tappan5, Erin Wessling6, Jill D. Pruetz1

1Department of Anthropology, Iowa State University, Ames, IA 50011, USA

2 Bandafassi Chimpanzee Conservation Project, Bandafassi, Senegal

3Institut des Sciences de l'Environnement, Faculté des Sciences et Techniques, Cheikh Anta

Diop University (UCAD), Dakar, Senegal,

4Département de Biologie Animale, UCAD, B.P 5005, Dakar-Fann, Senegal

5US Geological Survey (USGS)/Earth Resources Observation Systems (EROS) Data Center,

Sioux Falls, SD 57198, USA

6Department of Primatology, Max Planck Institute for Evolutionary Anthropology, Deutscher

Platz 6, D-04103 Leipzig, Germany

Correspondence: Stacy M. Lindshield, Department of Anthropology, Iowa State University, 324

Curtiss Hall, Ames, IA 50011, USA. Email: [email protected]

Short running head: Defining savanna chimpanzee habitat

151

Abstract Aim Proposed defining features of chimpanzee (Pan troglodytes) biogeography are sometimes overly simplistic and contradictory, yet we rely on these interpretations to shape conservation and management planning for this endangered ape. We address this problem through comparing landscapes on the margins of the species’ northwestern-most distribution for the cover and spatial configuration of critical habitat, with the goal of elucidating key determinates of chimpanzee biogeography.

Location The Bakel and Kedougou regions of southeastern Senegal.

Methods We combined reconnaissance surveys and interviews with published accounts to estimate the range of P. t. verus in Senegal and select study sites around this margin. High- resolution, multispectral satellite images were used to classify trees and shrubs at each site.

Landscapes within and outside of the species’ range were compared for woody vegetation cover and physiognomy, river density, and road density with ANOVA models.

Results We present an improved geographical distribution map for Senegal’s chimpanzee population that increases the northern extent of their range in West Africa. Our comparison of savanna landscapes around this margin showed that the physiognomy and coverage of shrubs and trees with green, foliage substantially changed across seasons and with chimpanzee occupancy on savanna landscapes. During the dry season, chimpanzees occupied areas with less vegetation cover, smaller patches of vegetation, and this vegetation was less clumped. However, river density was also higher in these areas and fewer roads transected their home-ranges.

Main conclusions Ligneous vegetation was associated with chimpanzee site occupancy in an unexpected way: higher cover and spatial aggregation occurred at ‘absent’. These findings show that percent cover is a poor predictor of chimpanzee biogeography at this scale of analysis. River 152 density was higher and road density was lower at chimpanzee ‘present’ sites, indicating that water scarcity and anthropogenic disturbance constrain their northern limit in Senegal.

Keywords high-resolution, biogeography, landscape, species’ distribution, forest cover, physiognomy, great ape, conservation

Introduction Chimpanzees (Pan troglodytes) are usually associated with tropical rain forests

(Lehmann & Dunbar, 2009; Junker et al., 2012) but lesser-known, savanna-dwelling chimpanzees can inhabit landscapes with a small fraction of evergreen forest, only 2-3% in some cases, in West and East Africa (Itani, 1979; McGrew et al., 1988; Pruetz & Bertolani, 2009). In a broad sense, chimpanzee biogeography is determined by forest cover and anthropogenic disturbance (Junker et al., 2012), but little attention has been given to the significance of landscapes, or entities comprised of abiotic features such as hydrogeological landforms and biotic communities that vary in abundances and spatial configurations (Kortlandt, 1983; Torres et al., 2010). On a conceptual level, landscapes are characterized by habitat physiognomy and composition. Physiognomy describes the physical structure and arrangement of habitats within the environment, while composition refers to the proportional relationships of habitat types

(Dunning et al., 1992). Using a landscape approach should improve our understanding of savanna chimpanzee biogeography because habitat physiognomy and composition are fundamental to species’ geographical distributions (Dunning et al., 1992), the availability of key resources for chimpanzees, namely water, shade, and food, can be scarce and widely dispersed on savanna landscapes during the dry season, and savanna environments appear to represent the maximum threshold tolerance of temperature and aridity for chimpanzees (McGrew et al., 1981;

Pruetz & Bertolani, 2009). From an applied perspective, a clearer understanding of ecological processes that explain chimpanzee ranging aid in the conservation and management of this 153 endangered species. We address this problem through assessing landscape features on the margins of the chimpanzee range to explain their occurrence, or lack thereof, in these arid environments.

Much emphasis has been placed on the importance of evergreen forest cover to chimpanzee biogeography (McGrew et al., 1988; Duvall, 2000; Lehmann & Dunbar, 2009;

Torres et al., 2010; Junker et al., 2012), and rightly so because most chimpanzees across Africa occupy areas with one-third or more forest cover (Lehmann & Dunbar, 2009). This vegetative habitat is often a source of food (Pruetz, 2006; Watts et al., 2012) and locations to construct nightly sleeping structures (Stewart & Pruetz, 2013). In savanna landscapes where evergreen forest is scarce, other significant aspects of this habitat type are readily apparent, as chimpanzees may rest under the cover of shade within these forest patches for the majority of the diurnal period to minimize heat stress (Fongoli, Senegal: Pruetz & Bertolani, 2009). In addition, these forest patches can be associated with permanent surface water sources (Baldwin et al., 1982;

Pruetz & Bertolani, 2009). However, deciduous trees and shrubs excluded in measures of evergreen forest cover (DeFries et al., 2000) are also critical chimpanzee habitat in savanna landscapes (Fongoli: Pruetz, 2006; Ugalla, Tanzania: Hernandez-Aguilar, 2009). Case in point,

Pruetz (2006) found that most plant foods consumed by chimpanzees at Fongoli were from ligneous plants located in deciduous woodland habitat. For this reason, we need not limit our view of important chimpanzee habitat to evergreen forest. Rather, trees and shrubs in general are important components of the chimpanzee niche.

Several proposed explanations for the species’ geographical distribution in West African savanna environments involve evergreen forest, as well as surface water availability and anthropogenic disturbance (Table 1). Although these factors are central to most hypotheses given 154 in Table 1, in his exhaustive description biogeographic factors, Kortlandt (1983) suggests that mosaic patterns in savanna plant communities (sensu Aubréville, 1950) may be causally linked to chimpanzee site occupancy. It is important to consider this view in the Shield ecoregion of southeastern Senegal, which gets its name from a Precambrian shield of volcanic and sedimentary rocks, which are often visible at the surface (Tappan et al., 2004). This ecoregion, which largely overlaps with the chimpanzee range, is characterized by a high degree of vegetative habitat heterogeneity, leading to a mosaic of plant communities and high plant species richness (Tappan et al., 2004). What is missing from our knowledge of savanna chimpanzee biogeography is a quantitative evaluation of habitat physiognomy. To address this gap, we investigate landscape composition and physigonomy on the margins of the species’ range in

Senegal.

The aim of this study is to elucidate landscape-scale relationships between chimpanzees and savanna habitat. To this end, we consider the hypothesis that ligneous plant physiognomy is important to the chimpanzee niche in arid environments. To evaluate this hypothesis, we predict that tree and shrub cover is higher and spatially-clumped on savanna landscapes within the chimpanzee range. The underlying logic of these predictions is grounded in empirical observations of habitat use by chimpanzees at Fongoli, Senegal. Given that most of their foods are from ligneous plants, there may be a positive correlation between woody vegetation cover and food availability. The second branch of reasoning involves the spatial configuration of this habitat. Aggregations of trees and shrubs may be indicative of suitable sleeping sites, shade cover, and surface water for this social great ape. Moreover, Fongoli chimpanzees appear to prefer clusters of feeding trees (Lindshield et al., in prep), perhaps as a way to increase foraging efficiency or reduce intraspecific feeding competition. 155

Materials and methods Study Sites

Selecting land areas for our analysis of savanna habitat composition and physiognomy required an accurate estimation of the western chimpanzee (Pan troglodytes verus) range in

Senegal. We incorporated two approaches to identify this range: field surveys and literature review (Galat-Luong et al., 1999; Carter et al., 2003; Ndiaye, 2011). The former was necessary because recent survey work (unpublished data from Boyer Ontl and Wessling) showed that maps of the species’ range underestimated its northern extent in Senegal (Figure 5.1).

Chimpanzee ‘present’ locations. The geographic locations of chimpanzee ‘communities’

(Goodall 1968, i.e., ‘groups’ or ‘unit-groups’) were identified through direct and indirect observations (Figure 5.1). Community ranging areas (km2) were estimated from direct observations of chimpanzees at the Fongoli site accustomed to the presence of researchers. At all other sites, brief sightings of unhabituated chimpanzees, and their feeding traces and nightly sleeping nests (Assirik: Baldwin et al., 1982; Pruetz et al., 2012; Kharakhena: Boyer-Ontl &

Pruetz, 2014, Boyer-Ontl, unpublished data; Keremekono and Makhana: Wessling, unpublished data; Nathia: Gašperšič, unpublished data) were used to approximate community ranging areas.

Ranging observations at the Fongoli site has shown that home-range area estimates increase with years of study (Pruetz, 2006; Skinner & Pruetz, 2012) and that home-range use fluctuates among years (Pruetz, unpublished data). To account for the possibility of underestimating home-range areas, we added a 1-km buffer to each area estimate.

Chimpanzee ‘absent’ locations. We found few reported chimpanzee ‘absences’ in the published literature (Kortlandt, 1983), a common problem in biogeographical research (Phillips et al.,

2006). Therefore, we identified five, potentially-devoid locations based on first-hand knowledge of the area (A. Goudiaby, personal observation), and ground-truthed these locations through 156 interviews with local residents and reconnaissance surveys of evergreen forests. For each new survey, we followed best-practice reconnaissance survey and interviewing techniques (Kühl et al., 2008). To define the area of each bona fide, absent site, we used the home-range area estimate at Fongoli (127 km2, with buffer) as a standard and positioned each site in an area with few settlements.

Landscape cover and physiognomy

To assess the impact of landscape factors on savanna chimpanzee biogeography, we compared ligneous vegetation, rivers, and roads between occupied and unoccupied sites. Woody plants were identified with high-resolution (2-4 m), multi-spectral satellite imagery from the

USGS-EROS archives (Table 2). Trees and shrubs were identified by RGB values (e.g., green, foliar tissue), and the size and shape of stems and leaves. Each site was covered in a lattice of 1- km2 cells using ArcMap 10.2 (ESRI, 2011), then ten cells were selected from each site according to a randomization sampling procedure (Figure 5.2). Trees and shrubs were grouped separately from all other landscape features with the binary and threshold tools in ImageJ (Abramoff et al.,

2004) (Figures 5.2 and 5.3). Randomly selected images were discarded when they contained cropland, roads, or settlements, as well as factors (e.g., dense, herbaceous vegetation indistinguishable from woody vegetation, cloud shadows) that prevented the accurate identification of trees and shrubs. After ligneous vegetation was in binary format (Figure 5.3), cover and configuration were summarized with landscape metrics in FRAGSTATS (McGarigal et al., 2012), including percentage of landscape cover (PLAND), largest patch index (LPI), aggregation index (AI), and perimeter-area fractal dimension (PAFRAC) (Table 3). In addition, river and road GIS layers (DIVA-GIS) were used to calculate average density (rivers km-2, roads km-2) by counting these features within ten, randomly-selected 1-km2 cells per site in ArcMap. 157

ANOVA models were used to compare these features among ‘present’ and ‘absent’ sites. We used the logit transformation to approximate the Gaussian distribution for PLAND, LPI, and AI, and the level of significance was set at α = 0.05. Statistical analyses were performed in R (R

Core Development Team, 2014).

Results Defining the upper latitudinal limit of Pan troglodytes in present-day Senegal

We ground-truthed five locations north of the estimated species range to verify their geographical boundary in Senegal (Figure 5.1). The absence of chimpanzees at Koussane was confirmed through an opportunistic survey (personal communication to A.G.). Pedestrian surveys and interviews with local residents showed that chimpanzees did not occupy the sites of

Goudiry Foulbe, Dianke Makam, Oumbare, and Bantanani. One informant reported that chimpanzees were extirpated from the Oumbare area at least ten years ago; no historic sightings were reported at the other sites. In addition, residents at each village stated that chimpanzees are not targeted for hunting because of a religious taboo. This finding supports earlier accounts of an indigenous ban on the killing of chimpanzees (Clavette, 2005) and indicates that this belief is widespread in southeastern Senegal. Informants at each village described a problem of water scarcity during the dry season. In January 2014, the middle part of the dry season, we found no flowing water within the gallery forests at these four sites.

Describing landscapes within and outside the Pan troglodytes geographical range

Landscape features, and ligneous vegetation cover and physiognomy were associated with the species’ range in Senegal. The densities of roads and rivers were higher (t = 5.627, d.f. =

43, P < 0.0001) and lower (t = -5.27, d.f. = 93.67, P < 0.00001), respectively, at chimpanzee

‘absent’ sites. We used a two-way ANOVA to examine relationships between ligneous vegetation, seasonality, and chimpanzee occupancy. Seasonality and occupancy were associated 158

with percent vegetation cover (Fseason = 28.81, d.f. = 1, P < 0.0001, Foccupancy = 4.957, d.f. = 1, P

= 0.03) and largest patch index (Fseason = 24.148, d.f. = 1, P < 0.001, Foccupancy = 7.058, d.f. = 1, P

< 0.01), while seasonality alone was associated with the aggregation index (Fseason = 4.135, d.f. =

1, P = 0.04) for ligneous vegetation (Table 4). We subsequently used a t-test to compare each landscape metric between chimpanzee ‘present’ and ‘absent’ locations within a dry or wet season

(Figure 5.4). Within wet season months, woody vegetation cover and physiognomy were not associated with chimpanzee occupancy. Perimeter-area fractal dimension (t = -1.101, d.f. =

34.298, P = 0.279) and aggregation index (t = 1.936, d.f. = 37.331, P = 0.06) were poor predictors of chimpanzee occupancy during the dry season, but aggregation approached significance. However, percent cover (t = 3.096, d.f. = 25.066, P < 0.01) and largest patch index

(t = 3.269, d.f. = 28.454, P < 0.01) for ligneous vegetation were unexpectedly lower at ‘present’ sites.

Discussion We present a revised range map for P. t. verus in Senegal that incorporates new occupancy locations and provides verified, chimpanzee ‘absence’ locations immediately north of this boundary (Figure 5.1). It is not assumed that chimpanzees occupy all areas within this range

(sensu Kortlandt, 1983), but it is unlikely that they currently occupy much higher latitudes in

Senegal. In addition, species’ range fluidity likely occurs due to short- and long-term environmental change. For instance, our interviews indicate that chimpanzees were extirpated from the Oumbare site within the last two decades. Thus, the recent historic range of Senegalese chimpanzees almost certainly included higher latitudes.

Although previous research has eluded to the importance of habitat physiognomy to savanna chimpanzee biogeography (Kortlandt, 1983), no study had explicitly quantified these spatial patterns. Through assessing ligneous vegetation physiognomy on the margins of the 159 chimpanzee range in Senegal, we discovered patterns in woody vegetation cover and physiognomy disputing the assumption that this habitat is more abundant and aggregated at sites occupied by chimpanzees. In fact, green trees and shrubs were equivalently abundant between sites in wet season months (Figure 5.4) and, contrary to expectations, less abundant and clumped at chimpanzee ‘present’ sites during the dry season. Consequently, we ought to exercise caution while interpreting the functional significance of woody vegetation cover for chimpanzees. For instance, Kortlandt (1983) argues that proportions of habitat cover (e.g., minimum threshold of forest cover) are poor predictors of chimpanzee biogeography; our landscape-scaled findings support this claim and indicate that it may be too simplistic a measure at this level of analysis

(but see Lehmann & Dunbar (2009) at larger spatial scales). Rather than percent habitat cover,

Kortlandt (1983) suggests that chimpanzees require highly diverse plant species communities. In this view, high species richness and vegetative habitat mosaics support the high diversity of plant foods that form the chimpanzee diet. A next step on this path is to identify relationships between plant community composition and physiognomy around the margins of the chimpanzee range.

One limitation to our analysis of ligneous plants is that foliage growth and photosynthesis may be temporally and spatially asynchronous. For instance, a late start or end to the rainy season may delay leaf growth or abscission, respectively. In addition, anthropogenic bush fires can stimulate growth in some plant species during the dry season (Grice, 1997). Although fire is routinely used to ‘clean’ these land areas in Senegal, and the first fires ignite shortly after herbaceous vegetation senescence, the process of incinerating this vegetation persists until the end of the dry season (Mbow et al., 2003). A future assessment of the impact of bushfire on green, woody vegetation during these dry periods may elucidate these differences in habitat cover and aggregation between chimpanzee ‘present’ and ‘absent’ sites. 160

While the comparison of habitat cover and physiognomy uncovered new layers of complexity, our findings support the idea that chimpanzees in arid environments are limited by surface water. The fact that river density was significant higher at sites occupied by chimpanzees indicates that chimpanzees would be less likely to encounter surface water were they to occupy these areas. One caveat to this interpretation is that discrete sources of permanent surface water in the Shield region, such as small, natural springs, can be difficult to accurately identify from satellite imagery alone (Pruetz, unpublished data). Thus, an accurate assessment of all surface water within an area requires exhaustive field surveying and interviewing. Although interviews with local residents, reconnaissance walks at ‘absent’ sites, and the assessment of river densities indicates that water scarcity is an important factor, more research is needed to determine if the drinking water supply north of the current species’ range is insufficient for chimpanzees.

Another attributable factor of the species’ boundary is more intensive, human land use at higher latitudes. Tappan and co-workers (2004) reported that human population densities within our full study area have been historically low due to its great distance from the capital city of

Dakar and poor soils for cultivation. However, they also documented an increase in charcoal production and agriculture at the higher latitudes of our study area between the 1980s and 1990s

(Tappan et al., 2004). Our finding that road density is significantly higher in these areas supports their conclusion and indicates that competition with humans over shared resources in part defines the northern limit of chimpanzees. This study also supports previous sociocultural research on human perspectives of chimpanzees; the collective weight of evidence shows that chimpanzees rarely fall victim to local hunters in Senegal. However, a recent increase in human population density in southeastern Senegal (ANSD/SRSD Kedougou, 2011), due in part to growth in the 161 gold mining industry, has important implications for the future state of the chimpanzee range in

Senegal.

In recent history, habitat loss has impacted chimpanzee ranges across Africa, with range area reduction being the dominant trend. Torres and co-workers (2010) found that chimpanzee habitat in Guinea had considerably decreased between 1986 and 2003. Similarly, chimpanzee population size and forest cover dramatically declined between 1989 and 2007 in the Ivory Coast

(Campbell et al., 2008). On a broader geographic scale, Junker and co-workers (2010) found a decrease in suitable habitat across the species’ range associated with anthropogenic disturbance.

Although this study is unusual in expanding the known geographical range of chimpanzees in

Senegal (Figure 5.1), this growth is almost certainly the result of increasing research effort into northern locations, rather than an expansion of chimpanzees into these areas. Senegal’s chimpanzee populations are at risk of declining due to habitat loss, primarily from human activity. The current, key drivers of habitat loss include mining and grazing (Ndiaye, 2011). The rapid growth of mining towns and the southern expansion of livestock grazing land are causally linked to increasing deforestation near chimpanzee communities (Fongoli: Pruetz, 2014;

Kharakhena: Boyer-Ontl, unpublished data). Furthermore, the impact of anthropogenic environmental change may be exacerbated by climate change, as southeastern Senegal has experienced higher average temperatures and lower average annual precipitations since the

1960s, and this trend is expected to continue over the next two decades (Funk et al., 2012).

These problems highlight the importance of identifying causal links between environment and geographical distribution for this endangered African ape.

Acknowledgements We thank the Republic of Senegal for granting us authorization to conduct research. S.L. thanks Brent Danielson, Stan Harpole, Dave Vleck, Dean Adams, and Bienvenu Sambou for 162 their advice and encouragement on this project. Mike Ervin and Kirk Maloney provided valuable assistance and feedback on an earlier version of this manuscript. S.L. also thanks Dondo Kante and Karfa Kante for their assistance in the field.

Literature cited Abramoff, M.D., Magalhaes, P.J. & Ram, S.J. (2004) Image Processing with ImageJ.

Biophotonics International, 11, 36-42.

ANSD/SRSD Kedougou (2011) Situation economique et sociale regionale 2010. In. Agence

Nationale de la Statistique et de la Démographie, Service Régional de la Statistique et de

la Démographie de Kedougou, Dakar, Senegal.

Aubréville, A. (1950) Flore forestière soudane-guinéenne. Société d'éditions géographiques,

maritimes et colonials, Paris.

Baldwin, P.J., McGrew, W.C. & Tutin, C.E.G. (1982) Wide-ranging chimpanzees at Mt Assirik,

Senegal. International Journal of Primatology, 3, 367-385.

Boyer-Ontl, K. & Pruetz, J. (2014) Giving the Forest Eyes: The Benefits of Using Camera Traps

to Study Unhabituated Chimpanzees (Pan troglodytes verus) in Southeastern Senegal.

International Journal of Primatology, 35, 881-894.

Campbell, G., Kuehl, H., N'Goran Kouamé, P. & Boesch, C. (2008) Alarming decline of West

African chimpanzees in Côte d'Ivoire. Current Biology, 18, R903-R904.

Carter, J., Ndiaye, S., Pruetz, J.D. & McGrew, W.C. (2003) Senegal. West African chimpanzees

(ed. by R. Kormos, C. Boesch, M. Bakarr and T.M. Butynski), pp. 31-39. World

Conservation Union, Gland.

Clavette, K. (2005) Conservation Through Folklore: Ethnoprimatology in Southeastern Senegal.

University of Northern Arizona, 163

DeFries, R., Hansen, M., Townshend, J., Janetos, A. & Loveland, T. (2000) A new global 1km

data set of percent tree cover derived from remote sensing. Global Change Biology, 6,

247-254.

Dunning, J.B., Danielson, B.J. & Pulliam, H.R. (1992) Ecological Processes That Affect

Populations in Complex Landscapes. Oikos, 65, 169-175.

Duvall, C.S. (2000) Important habitat for chimpanzees in Mali. African Study Monographs, 21,

173-203.

ESRI (2011) ArcGIS Desktop: Release 10. In. Environmental Systems Research Institute,

Redlands, CA.

Funk, C., Rowland, J., Adoum, A., Eilerts, G., Verdin, J. & White, L. (2012) A climate trend

analysis of Senegal. In: USGS Numbered Series Fact Sheet, Reston, VA.

Galat-Luong, A., Galat, G., Ndiaye, P.I. & Keita, Y. (1999) Fragmentation de la distribution et

statut actuel du chimpanzee Pan troglodytes verus en limite d'aire de répartition au

Senegal. African Primates, 4, 71-72.

Grice, A.C. (1997) Post-fire regrowth and survival of the invasive tropical shrubs Cryptostegia

grandiflora and Ziziphus mauritiana. Australian Journal of Ecology, 22, 49-55.

He, H.S., DeZonia, B.E., & Mladenoff, D.J. (2000) An aggregation index (AI) to quantify spatial

patterns of landscapes. Landscape Ecology, 15, 591-601.

Hernandez-Aguilar, R.A. (2009) Chimpanzee nest distribution and site reuse in a dry habitat:

implications for early hominin ranging. Journal of Human Evolution, 57, 350-364.

Itani, J. (1979) Distribution and adaptation of chimpanzees in an arid area. The great apes:

perspectives on human evolution (ed. by D.A. Hamburg and E.R. Mccrown), pp. 55-71.

Phoenix Publishing Services, San Francisco, CA. 164

Junker, J., Blake, S., Boesch, C., Campbell, G., Toit, L.d., Duvall, C., Ekobo, A., Etoga, G.,

Galat-Luong, A., Gamys, J., Ganas-Swaray, J., Gatti, S., Ghiurghi, A., Granier, N., Hart,

J., Head, J., Herbinger, I., Hicks, T.C., Huijbregts, B., Imong, I.S., Kuempel, N., Lahm,

S., Lindsell, J., Maisels, F., McLennan, M., Martinez, L., Morgan, B., Morgan, D.,

Mulindahabi, F., Mundry, R., N'Goran, K.P., Normand, E., Ntongho, A., Okon, D.T.,

Petre, C.-A., Plumptre, A., Rainey, H., Regnaut, S., Sanz, C., Stokes, E., Tondossama,

A., Tranquilli, S., Sunderland-Groves, J., Walsh, P., Warren, Y., Williamson, E.A. &

Kuehl, H.S. (2012) Recent decline in suitable environmental conditions for African great

apes. Diversity and Distributions, 18, 1077-1091.

Kortlandtt, A. (1983) Marginal habitats of chimpanzees. Journal of Human Evolution, 12, 231-

278.

Krummel, J.R., Gardner, R.H., Sugihara, G., O’Neill, R.V. & Coleman, P.R. (1987) Landscape

patterns in disturbed environments. Oikos, 48, 321-324.

Kühl, H.M., F., Ancrenaz, M. & Williamson, E.A. (2008) Best Practice Guidelines for Surveys

and Monitoring of Great Ape Populations. In, p. 32, Gland, Switzerland.

Lehmann, J. & Dunbar, R. (2009) Implications of body mass and predation for ape social system

and biogeographical distribution. Oikos, 118, 379-390.

Mbow, C., Sambou, B., Ba, A.T. & Goudiaby, A. (2003) Vegetation and fire readiness in main

morphological units of Niokolo Koba National Park (Southeast Senegal). Geografisk

Tidsskrift-Danish Journal of Geography, 103, 55-62.

McGarigal, K., Cushman, S.A. & Ene, E. (2012) FRAGSTATS v4: Spatial Pattern Analysis

Program for Categorical and Continuous Maps. In, University of Massachusetts,

Amherst. 165

McGrew, W.C., Baldwin, P. & Tutin, C.E.G. (1981) Chimpanzees in a hot, dry and open habitat:

Mt. Assirik, Senegal, West Africa. Journal of Human Evolution, 10, 217-244.

McGrew, W.C., Baldwin, P.J. & Tutin, C.E.G. (1988) Diet of wild chimpanzees (Pan troglodytes

verus) at Mt. Assirik, Senegal: I. Composition. American Journal of Primatology, 16,

213-226.

Ndiaye, S. (2011) Conservation du chimpanze au Senegal: etat des connaissances et

reactualisation du plan d'actions. In. USAID-Wula Nafaa

Phillips, S.J., Anderson, R.P. & Schapire, R.E. (2006) Maximum entropy modeling of species

geographic distributions. Ecological Modelling, 190, 231-259.

Pruetz, J.D. (2006) Feeding ecology of savanna chimpanzees (Pan troglodytes verus) at Fongoli,

Senegal. Feeding ecology in apes and other primates (ed. by G. Hohmann, M.M.

Robbins and C. Boesch), pp. 161-182. Cambridge University Press, Cambridge, UK.

Pruetz, J.D. (2014) Studying Apes in a Human Landscape. Primate Ethnographies (ed. by K.

Strier), pp. 228-237. Pearson, Upper Saddle River, NJ.

Pruetz, J.D. & Bertolani, P. (2009) Chimpanzee (Pan troglodytes verus) behavioral responses to

stresses associated with living in a savanna-mosaic environment: implications for

hominin adaptations to open habitats. Paleoanthropology, 252-262.

Pruetz, J.D., Ballahira, R., Camara, W., Lindshield, S., Marshack, J.L., Olson, A., Sahdiako, M.

& Villalobos-Flores, U. (2012) Update on the Assirik chimpanzee (Pan troglodytes

verus) population in Niokolo Koba National Park, Senegal. Pan Africa News, 19

Skinner, M.F. & Pruetz, J.D. (2012) Reconstruction of periodicity of repetitive linear enamel

hypoplasia from perikymata counts on imbricational enamel among dry-adapted 166

chimpanzees (Pan troglodytes verus) from Fongoli, Senegal. American Journal of

Physical Anthropology, 149, 468-482.

Stewart, F.A. & Pruetz, J.D. (2013) Do Chimpanzee Nests Serve an Anti-Predatory Function?

American Journal of Primatology, 75, 593-604.

Tappan, G.G., Sall, M., Wood, E.C. & Cushing, M. (2004) Ecoregions and land cover trends in

Senegal. Journal of Arid Environments, 59, 427-462.

Team, R.D.C. (2014) R: a language and environment for statistical computing. R Foundation for

Statistical Computing.

Torres, J., Brito, J.C., Vasconcelos, M.J., Catarino, L., Gonçalves, J. & Honrado, J. (2010)

Ensemble models of habitat suitability relate chimpanzee (Pan troglodytes) conservation

to forest and landscape dynamics in Western Africa. Biological Conservation, 143, 416-

425.

Watts, D.P., Potts, K.B., Lwanga, J.S. & Mitani, J.C. (2012) Diet of chimpanzees (Pan

troglodytes schweinfurthii) at Ngogo, Kibale National Park, Uganda, 1. diet composition

and diversity. American Journal of Primatology, 74, 114-129.

167

168

169

170

Figure 5.1. Map of study sites where chimpanzees are absent (▲) and present (●) and the revised species’ range for Senegal.

171

Figure 5.2. Flow chart of image processing procedure. 172

Figure 5.3. Landscape sampling areas (1-km2) showing trees and shrubs (black) in a binary format. White areas are ‘background’, including herbaceous vegetation, soil, and bedrock.

Quadrants clockwise from upper left: dry season ‘absent’ sites, dry season ‘present’ sites, wet season ‘present’ sites, wet season ‘absent’ sites. 173

Figure 5.4. Comparison of ligneous vegetation cover and physiognomy among study sites and between dry (white fill) and wet (rose fill) seasons. Significance values were calculated from t- tests (* P < 0.01). 174

CHAPTER VI CONCLUSION December, 2011: After spending much of the day resting and socializing in the ravine at Tukantaba, where intermittent water puddles could be found, the party made their way towards the fields of Maragoundi. A few days beforehand, a farming family had left their home here for the season after harvesting the yearly crop of millet and peanut. These barren fields were peppered with forbs, fig saplings, and the occasional baobab or tamarind tree. The party walked straight through these fields, using a trail system that they would normally avoid when the farming community occupied these areas. Bilbo and the others stopped at the fig saplings to eat tender, new leaves. Next, Bilbo scaled a large baobab tree in the middle of the field – one that the community had avoided during the baobab fruiting season, until today. He inspected several, large fruits and pulled one off of the tree. All of a sudden, Tia began to shriek excitedly at the discovery of ripe figs across the field. Bilbo immediately dropped his fruit, clambered down the tree, and quickly moved towards Tia. The rest of the party followed suit and they collectively gorged on figs for the rest of the day. The next morning, Siberuit picked up Bilbo’s discarded baobab fruit and had a meal of it.

The central goal of this study was to identify new links between the foraging behavior of savanna chimpanzees and habitats at three levels: food resources, resource availability in a habitat mosaic, and landscapes of habitat mosaics. Each non-mutually exclusive tier represents an important facet of the savanna biome for chimpanzees (Figure 6.1). The aim of this chapter is to highlight key take-home messages from each level and the connections between levels, to recommend new research that strengthens these veins, and to provide practical and research applications of our study outcomes.

Take home messages Level one: Diet and food selection

High-energy, high-sugar fruits are a preferred food. It is widely acknowledged that chimpanzees consume a wide variety of plant and animal foods but focus on ripe fruit, and Fongoli chimpanzees are no exception (Pruetz 2006). Furthermore, when two foods of similar accessibility but different metabolizable energy (ME) and simple-sugar intake rates (kJ hr-1 or g 175 hr-1 respectively) were present within a food patch (unripe and ripe baobab fruit), Fongoli chimpanzees preferentially selected ripe fruit, the higher energy, higher sugar food. In addition, dry matter intake rate (g hr-1) was higher in ripe fruit (Chapter 4). We could not test for ‘energy’ or ‘sweet’ preferences, as a controlled, experimental research design was not feasible for this field study. As discussed in Chapter 1, taste preference in captive chimpanzees shows that sweetness is a powerful predictor of choice (Remis 2006). Furthermore, chimpanzees are not easily deterred by astringent compounds when sugar content is relatively high (Remis 2006), as is the case for several fruits consumed by Fongoli chimpanzees (Chapter 3, Table x). Although more research is needed to test the relationship between food preference, nutritional composition, and rate of food intake, these findings show that Fongoli chimpanzees prefer foods with high nutritional rewards.

Gum, leaves, and flowers at Fongoli are good sources of metabolizable energy. In the field of primate feeding ecology, fruit is the archetypal energy-rich food, while leaves and other plant foods are commonly viewed as ancillary to fruit (Chapter 1). Our comparison of ME intake rate found that gum, not fruit, ranked highest among all foods (Figure 3.3). However, this gum was infrequently consumed during the study period, perhaps because each tree had few exuding scars. After gum, fruit was highest in ME intake rate (kJ hr-1), and flowers and young leaves were not far behind (Figure 3.3). These similarities at Fongoli support earlier claims that young leaves (Amato and Garber 2014) and flowers (Lappan 2009) can be underappreciated as high- quality food items for primates. Moreover, this finding shows that commonly used measures of food availability in chimpanzee studies that are exclusionary of flowers and leaves, such as the fruit availability index (Koops et al. 2014), neglect nutritionally important foods. Finally, the 176 significant overlap in ME intake rate among most plant foods in this study may partially explain why ME intake rate was a poor predictor of food choice in our sample.

Termite and bamboo pith foods are commonly consumed but uncommonly low in metabolizable energy. The foods most frequently consumed by Fongoli chimpanzees were relatively high in

ME intake rate, save for termites and bamboo pith (Figure 3.4). These foods are rich in protein and simple sugars, respectively, on a percent of dry matter basis. However, dry matter intake rates for each one were exceptionally low, which resulted in low energy intake rates. Many chimpanzee communities consume insects and piths (O'Malley and Power 2014; Pruetz 2006;

Watts et al. 2012), but Fongoli chimpanzees are unusual in that these foods are commonly eaten across most months of the year and when foods with higher ME intake rates are available.

Provided that daily protein requirements are met with other foods, such as fruit and leaves, chimpanzees may target termites for essential micronutrients (e.g., vitamin B12) that can be limited in other foods (Deblauwe and Janssens 2008). Although this supplementary hypothesis

(Rothman et al. 2014) has not been tested at Fongoli, the high frequency of termite consumption suggests that these insects are complementary to the diet, rather than supplementary (Bogart and

Pruetz 2011). That is to say, termites might be commonly consumed because they provides several, essential micronutrients or protein, or both (Rothman et al. 2014). From a nutritional perspective, bamboo pith consumption is perplexing because Fongoli chimpanzees select this food when higher energy alternatives are available in their home range. In other words, it does not fit the ‘fallback’ food label (Chapter 1). With this in mind, the larger habitat context associated with these foods is an important factor to consider (Chapter 4). Termites and bamboo are abundant at Fongoli and chimpanzees do not need to climb up trees to reach either of them.

In addition, bamboo is often associated with shade cover and surface water. In a savanna 177 environment where heat stress is a risk, chimpanzees may target these foods as part of an energy minimizing strategy (Bogart and Pruetz 2011; Pruetz 2006).

Level two: Habitat quality and selection

Fongoli chimpanzees consume less food at patches far from drinking water. Water is a scarce resource for savanna chimpanzees during months with low precipitation, or none at all.

Consequently, their foraging behavior is sensitive to water availability (Chapter 1). We hypothesized that visiting patches located far from surface water sources would incur higher metabolic costs due to the increasing risk of dehydration. In support of this idea, we found that food intake decreased with increasing distance from surface water sources. Furthermore, food intake rate did not change with distance to water, which shows that less time was allocated to feeding in these peripheral areas.

Chimpanzees eat faster in open-canopy habitats at Fongoli. Feeding in open-canopy habitats exposes chimpanzees to higher solar radiation. Closed-canopy habitats at Fongoli are associated with evergreen trees and lianas that provide a refuge from light and heat. Meanwhile, shade is a scarce resource within open-canopy habitats during the dry season. Even during the wet season, open-canopy habitats receive more light at a chimpanzee’s eye-level, but this difference is more pronounced following leaf abscission for tree and shrubs. Given this relationship, chimpanzees should be more susceptible to heat stress in open-canopy habitats and, ergo, foraging in these habitats almost certainly incurs higher metabolic costs. For these reasons, we predicted that study subjects would forage more efficiently in these areas by increasing feeding rate. While our results did not show an overall difference in food intake (dry matter ingested in grams) between patches in open- versus closed-canopy habitats, rate of intake (dry matter gm hr-1) was significantly higher while chimpanzees were in the open. 178

This association between intake rate and habitat supports the metabolic cost hypothesis

(Chapter 4); however, these results should be interpreted with caution because potential differences in food availability and predation risk between open- and closed-canopy habitats provide alternate explanations. For instance, food may be scarcer in open-canopy habitats, leading to greater uncertainty in determining when and where a productive food patch is located

(Brown 1988). Bearing in mind that most plant foods are associated with woodland habitats and open-canopy habitats are dominant at Fongoli (Pruetz 2006, Pruetz and Bertolani 2009), it is unlikely that total food mass was higher in closed-canopy habitats during the study period.

However, a comparison of food biomass (kg ha-1) among habitat types is needed to better evaluate this competing explanation. Alternatively, chimpanzees may be more exposed and vulnerable to predation in open-canopy habitats, thus leading study subjects to increase food intake rate while visiting these risky patches (Brown 1988). However, this interpretation was not supported by our test for predation-sensitive foraging (see next section). For these reasons, the metabolic cost hypothesis is the better explanation.

Ripe fruit is worth the risk. While there is a taboo against killing chimpanzees in southeastern

Senegal and Fongoli chimpanzees are habituated to the presence of researchers, antipredator behavior was commonly directed towards people. These responses mostly consisted of vigilance and alarm calls. However, this perceived risk of predation did not prevent chimpanzees from visiting food patches in high risk areas; that is to say, locations close to villages, fields, roads, and mines. When study subjects visited risky patches, we expected them to increase rate of food intake or decrease food intake to minimize their risk of exposure to people; these expectations were not met. Instead, feeding rate remained constant but food intake increased, meaning that study subjects remained in risky food patches for longer intervals of time. In addition, average 179 ripe fruit intake per food patch visit increased with risk. These findings indicate that focal subjects visited these food patches when ripe fruit was relatively abundant there. Earlier we discussed that Fongoli chimpanzees prefer ripe over unripe baobab fruit (Chapter 3); this relationship between food preference and risk-taking is a link between food and habitat levels of analysis (Figure 6.1).

Level three: Savanna chimpanzee biogeography

Water scarcity may limit the northern boundary of the species’ range in Senegal. Our comparison of rivers km-2 on landscapes occupied and unoccupied by chimpanzees showed that river density was significantly lower at unoccupied sites. Moreover, informants living nearby these landscapes reported very low water availability during the dry season at unoccupied sites.

It appears that unoccupied sites have less surface water during the dry season and this may explain the absence of chimpanzees there. Also, as mentioned above, we found that Fongoli chimpanzees decreased food intake with increasing distance from surface water. It may be that surface water sources are too scarce at unoccupied sites to support the energetic needs of chimpanzees (Figure 6.1). However, a challenge to this interpretation is that our methodology may have overlooked small, permanent water sources that are difficult to identify from satellite imagery. More information on permanent sources of surface water is required to support our initial interpretation.

Higher anthropogenic disturbance at higher latitudes of eastern Senegal may restrict chimpanzees to the south. While anthropogenic disturbance is pervasive throughout southeastern

Senegal, our comparison of roads, a proxy of habitat loss and degradation, demonstrated that there were fewer of them at sites with chimpanzees. Although additional landscape features, such as settlements, cropland, and mines should also be important indicators of habitat loss, roads 180 have been significant predictors of chimpanzee occurrence throughout the species’ range from

West to East Africa (Junker et al. 2012). Furthermore, Tappan and co-workers (2004) identified increasing habitat loss due to a rise in charcoal production and agriculture in the same vicinity as our northern-most, unoccupied sites. These intersecting lines of evidence support the idea that competition with people over natural resources has shaped the northern margin of the chimpanzee range in Senegal.

Minimum percent forest cover is a poor predictor of chimpanzee occupancy in Senegal. While road and river densities were good predictors of chimpanzee occupancy at our landscape level of analysis, minimum percent forest cover was not. Contrary to expectations, during dry months when most foliage should be associated with evergreen species, we found that tree and shrub cover, as well as largest contiguous patch of ligneous vegetation, was higher at sites unoccupied by chimpanzees. Thus, the use of a minimum forest cover threshold (Lehmann and Dunbar 2009;

McGrew et al. 1988), without the aid of other landscape features (see above sections), may not be a reliable indicator of chimpanzee site occupancy.

New directions in savanna chimpanzee foraging ecology Nutritional ecology

The nutritional analysis in Chapter 3 provides an initial survey of macronutrients and condensed tannins in plant and insect foods at Fongoli. In the examination of food selection and metabolizable energy (ME) intake rates (kJ hr-1), we found that Fongoli chimpanzees were selecting foods with similarly high ME intake rates, save for stem piths and insects. More research is needed to elucidate the importance of these low-energy foods at Fongoli. For instance, a survey of essential micronutrients is needed to test the supplementary versus complementary hypotheses for insectivory (Rothman et al. 2014). Moreover, comparing the energetic demands of foraging for termites and bamboo on the ground versus tree climbing for 181 fruit may clarify the impact of energy expenditure on food choice. Another research avenue to consider is comparative work with captive chimpanzees. Major obstacles to understanding the foraging strategies of wild chimpanzees include confounding variables that are difficult or impossible to control, small sample sizes, and ethical concerns with artificial food experiments

(Gruen et al. 2013). Observing volunteer participants in food selection experiments may reduce or eliminate confounding factors, such as the taste preference trials with chimpanzees and gorillas at the San Francisco Zoo (Remis 2006). Coordinated and complementary studies between wild and captive study subjects may greatly advance the field of primate nutritional ecology.

Predation-sensitive foraging

This study demonstrated that Fongoli chimpanzees perceived humans as predators.

Community members adjusted their foraging behavior to avoid direct contact with them, but entered risky habitats when nutritional rewards were high. A step forward to elucidating the ecology of fear is to evaluate chimpanzee feeding behavior under different types of predation pressure. For instance, Fongoli chimpanzees may have avoided risky patches, or consumed very little at these locations, had there been a higher risk of death. Future investigations should test these predictions with chimpanzees that forage in the presence of known predators, such as leopards or lions.

Savanna chimpanzee biogeography

This study used a landscape approach with remote sensing data to identify several important predictors of the chimpanzee range in Senegal. Woody vegetation physiognomy was associated with the species’ distribution but this relationship needs to be further explored. We recommend that future studies build on this study by comparing permanent water sources, plant 182 species communities, and habitat profiles (e.g., tree height, crown volume), as well as real and potential foods for nutritional composition and abundance. With these data in hand we may test the prediction that unoccupied sites lack adequate resources to sustain chimpanzee communities.

Alternatively, we may be observing an ‘empty’ savanna (sensu Redford 1992), where predation or intense competition with humans defined the northern range of Senegalese chimpanzees, or at least contributed to it. Answering these questions should improve chimpanzee distribution models and, more importantly, enrich our understanding of the savanna chimpanzee niche.

Anthropological and practical applications The thermoregulatory hypothesis of bipedalism

Paleoanthropologists that use savanna chimpanzee models to inform behavioral reconstructions of early hominins (Chapter 2) should take note of our findings on the significance of surface water. Given that Fongoli chimpanzees appear to adjust their feeding in relation to water proximity (Chapter 4), and the species’ range in Senegal is associated with watercourses (Chapter 5), it is clear that water scarcity constrains their foraging and ranging behavior. Albeit that drinking water has long been recognized as a limited resource for savanna chimpanzees (Kortlandt 1983), this study is the first to document a decrease in food intake

(grams of dry matter) with increasing distance from drinking water. In paleoenvironments similar to Fongoli, traits in early hominins that reduced the risk of heat stress and dehydration would have conferred a selective advantage. According to Wheeler (1991), obligate bipedalism was an effective strategy for thermoregulation in hot environments, as less surface area of the body would have been exposed to solar radiation. This, in turn, should have decreased sweating and water intake requirements. While there are challenges to the ape referential model paradigm

(reviewed in Sayers and Lovejoy 2008), the Fongoli site enables anthropologists to observe ape 183 strategies concerning water scarcity, an environmental pressure considered by many to be important in hominin evolution.

Crop-raiding management

Conservationists and land-managers that work on crop-raiding problems ought to consider our findings on predation-sensitive foraging. Fongoli chimpanzees are not considered agricultural pests, in the traditional sense, because they do not consume major crop foods in the area: millet, corn, and peanut. While crop-raiding is not a serious problem for the Fongoli

Savanna Chimpanzee Project, there is minor conflict over valued foods, such as honey, mango, and fruit of the Tamarindus indica tree. To our knowledge, this shared resource use has not led to cases of lethal retaliation and ‘crop’ losses have been relatively low. However, attitudes may change if economic damage becomes more significant.

One unusual aspect of Fongoli chimpanzee behavior is that they forage within and around cropland for wild food items. While they frequently visited food patches on the margins of fields, visits to food patches within cultivated land usually coincided with periods of minimal human activity (Chapter 4), probably to avoid direct interactions. These findings agree with reports of nocturnal crop-raiding behavior by Sebitoli chimpanzees at Kibale National Park, Uganda (Krief et al. 2014), and indicates that chimpanzees may respond to competition with humans through partitioning resource use along a temporal axis. Secondly, although Fongoli chimpanzees fear most local people, they are willing to accept some degree of risk to consume preferred food items. Our results suggest that Fongoli chimpanzees visit these risky areas when relatively high nutritional rewards can be gained. This response to wild food sources underscores the challenges of crop-raiding management, as cropland tends to yield highly aggregated patches of energy-rich foods. However, the case of cropland use at Fongoli demonstrates that chimpanzees can bypass 184 these foods when high-energy, non-cultivated fruits and leaves are planted in and around agricultural lands. Thirdly, in some cases Fongoli chimpanzees and farmers seem to have a mutualistic relationship around cultivated areas. Vervet monkeys, baboons, and patas monkeys are killed at Fongoli because they frequently raid crops. In addition, chimpanzees hunt these primate ‘pest’ species. Therefore, it is likely that monkeys avoid cropland when chimpanzees are detected nearby. Land managers and conservationists might take advantage of these relationships to encourage tolerance of chimpanzees around croplands.

The study of early hominins and human-wildlife conflict are part of a larger body of knowledge that benefits from a better understanding of savanna chimpanzee behavior and ecology. While savanna chimpanzee studies have recently flourished (Chapter 2), factors such as poor public and institutional support for natural history research, political instability in chimpanzee range countries, and disease outbreaks are some of the challenges that primatologists face while developing new field research programs on savanna chimpanzees. Exacerbating this problem is the downward trend of chimpanzee population sizes across the species’ geographical distribution. To continue research on savanna chimpanzees, such as the new directions for savanna chimpanzee ecology outlined above, researchers should advocate and contribute to chimpanzee conservation efforts before it is too late.

Literature cited Amato KR, and Garber PA. 2014. Nutrition and foraging strategies of the black howler monkey

(Alouatta pigra) in Palenque National Park, Mexico. American Journal of Primatology

76(8):774-787.

Bogart SL, and Pruetz JD. 2011. Insectivory of savanna chimpanzees (Pan troglodytes verus) at

Fongoli, Senegal. American Journal of Physical Anthropology 145(1):11-20. 185

Brown JS. 1988. Patch use as an indicator of habitat preference, predation risk, and competition.

Behavioral Ecology and Sociobiology 22(1):37-47.

Deblauwe I, and Janssens GPJ. 2008. New insights in insect prey choice by chimpanzees and

gorillas in Southeast Cameroon: The role of nutritional value. American Journal of

Physical Anthropology 135(1):42-55.

Gruen L, Fultz A, and Pruetz J. 2013. Ethical issues in african great ape field studies. ILAR

Journal 54(1):24-32.

Junker J, Blake S, Boesch C, Campbell G, Toit Ld, Duvall C, Ekobo A, Etoga G, Galat-Luong

A, Gamys J, Ganas-Swaray J, Gatti S, Ghiurghi A, Granier N, Hart J, Head J, Herbinger

I, Hicks TC, Huijbregts B, Imong IS, Kuempel N, Lahm S, Lindsell J, Maisels F,

McLennan M, Martinez L, Morgan B, Morgan D, Mulindahabi F, Mundry R, N'Goran

KP, Normand E, Ntongho A, Okon DT, Petre C-A, Plumptre A, Rainey H, Regnaut S,

Sanz C, Stokes E, Tondossama A, Tranquilli S, Sunderland-Groves J, Walsh P, Warren

Y, Williamson EA and HS Kuehl. 2012. Recent decline in suitable environmental

conditions for African great apes. Diversity and Distributions 18(11):1077-1091.

Koops K, Schöning C, McGrew WC, and Matsuzawa T. 2014. Chimpanzees prey on army ants

at Seringbara, Nimba Mountains, Guinea: Predation patterns and tool use characteristics.

American Journal of Primatology.

Kortlandt A. 1983. Marginal habitats of chimpanzees. Journal of Human Evolution 12:231-278.

Krief S, Cibot M, Bortolamiol S, Seguya A, Krief J-M, and Masi S. 2014. Wild chimpanzees on

the edge: nocturnal activities in croplands. PLoS ONE 9(10):e109925.

Lappan S. 2009. Flowers are an important food for small apes in southern Sumatra. American

Journal of Primatology 71(8):624-635. 186

Lehmann J, and Dunbar R. 2009. Implications of body mass and predation for ape social system

and biogeographical distribution. Oikos 118(3):379-390.

McGrew WC, Baldwin PJ, and Tutin CEG. 1988. Diet of wild chimpanzees (Pan troglodytes

verus) at Mt Assirik, Senegal. 1. Composition. American Journal of Primatology

16(3):213-226.

O'Malley RC, and Power ML. 2014. The energetic and nutritional yields from insectivory for

Kasekela chimpanzees. Journal of Human Evolution 71:46-58.

Pruetz JD. 2006. Feeding ecology of savanna chimpanzees (Pan troglodytes verus) at Fongoli,

Senegal. In: Hohmann G, Robbins MM, and Boesch C, editors. Feeding ecology in apes

and other primates. Cambridge, UK: Cambridge University Press. p 161-182.

Redford KH. 1992. The empty forest. Bioscience 42(6):412-422.

Remis MJ. 2006. The role of taste in food selection by African apes: implications for niche

separation and overlap in tropical forests. Primates 47:56-64.

Rothman JM, Raubenheimer D, Bryer MAH, Takahashi M, and Gilbert CC. 2014. Nutritional

contributions of insects to primate diets: Implications for primate evolution. Journal of

Human Evolution 71(0):59-69.

Sayers K, and Lovejoy CO. 2008. The chimpanzee has no clothes. Current Anthropology

49(1):87-114.

Tappan GG, Sall M, Wood EC, and Cushing M. 2004. Ecoregions and land cover trends in

Senegal. Journal of Arid Environments 59(3):427-462.

Watts DP, Potts KB, Lwanga JS, and Mitani JC. 2012. Diet of chimpanzees (Pan troglodytes

schweinfurthii) at Ngogo, Kibale National Park, Uganda, 1. diet composition and

diversity. American Journal of Primatology 74(2):114-129. 187

Wheeler PE. 1991. The influence of bipedalism on the energy and water budgets of early

hominids. Journal of Human Evolution 21(2):117-136.

Figure 6.1. Conceptual summary of the three levels of analysis in this study. Ecological and social factors described or tested, or both, in this study are listed for each level. Brackets and far- side boxes highlight key links between levels that are described in this chapter. This summary is not exhaustive. 188

APPENDIX SUPPLEMENTARY MATERIALS FOR CHAPTER IV Methods Plant-soil interactions may lead to intraspecific variation in the nutritive quality of plant tissues at detectable levels for mammalian herbivores (Arnold et al. 2014; O’Reilly-Wapstra et al. 2005). To account for this effect on our study subjects’ choice of food patch visits, we evaluated relationships between baobab fruit nutritive values and Fongoli chimpanzee patch visits to soil classes. This was a post-hoc analysis, where soil class was determined from satellite imagery, field notes, and SL’s knowledge of the terrain. Three, broad categories were used to classify soils: (1) silty, alluvial-like soils found near streams and rivers, forests, or cultivated fields (active or fallow), (2) laterite pan (bedrock) exposed at the surface, found near ravines and associated with gallery or ecotone forest vegetation, and (3) gravelly soils usually associated with woodland and grassland vegetation. We tested for a relationship between soil class and nutritive quality with a multivariate analysis of variance (MANOVA), and compared nutritive values among soil classes with a Kruskal-Wallis test with multiple comparisons. Given that laterite pan and gravelly soils were associated with closed and open canopy habitat types, respectively, we also tested for a relationship between canopy cover and baobab fruit nutritive values with a MANOVA test. Also, we used a MANOVA to test for a relationship between nutritive values and the presence of feeding traces (yes/no) at baobab food patches. The significance value was set at α=0.05.

Results There was a significant, multivariate relationship between nutritive values and soil class

(F=2.32, r2=0.05, P=0.01). Neutral detergent fiber (NDF; an estimate of total cellulose, hemicellulose, and lignin) and condensed tannins (CT) were significantly different but all other 189 nutrients were equivalent among soil classes. NDF was significantly higher in laterite pan relative to silty soils (observed difference = 19.93, critical difference = 18.94), but equivalent among gravel-pan (observed difference = 9.66, critical difference = 17.14) and gravel-silt

(observed difference = 10.26, critical difference = 16.32) class comparisons. In addition, CT was lower in silty soils relative to gravel (observed difference = 17.00, critical difference = 16.32) and pan (observed difference = 23.26, critical difference = 18.94) soils, but gravel and pan were not different from one another (observed difference = 6.26, critical difference = 17.14). While soil class was a significant predictor of nutritive quality, canopy cover was not (F=1.73, r2=0.02,

P=0.12). Furthermore, the nutritive value was not a reliable indicator of food patch visits

(F=0.59, r2=0.01, P=0.84).

Discussion Ripe fruits from trees (i.e., food patches) rooted in silty soils were lower in NDF and CT composition, but these differences did not appear to influence study subjects’ visits to food patches. Instead of nutritive quality within ripe fruit, the difference between ripe and unripe fruit quality was a better predictor of food choice (Chapter 4). Furthermore, the habitat context of food patches, such as proximity to anthropogenic landmarks and canopy structure were more important predictors of food choice (Chapter 4).

The lower levels of NDF and CT in silty soils may be related to higher soil fertility

(O’Reilly-Wapstra et al. 2005). Silty soils were associated with flood plains and areas of cultivation at Fongoli. The process of fluvial sedimentation can enrich soils with organic material, particularly in flood plain areas (Pinay et al. 2002). While cultivated areas are often associated with flood plains at Fongoli, soils in these areas may be further enriched through traditional soil fertilization practices (Harris 2002). Cattle, donkeys, and goats deposit nitrogen- rich urine and feces in these areas while foraging on peanut leaves and the senesced stems of 190 corn and millet following the harvest season. More research is needed to elucidate the relationships between baobab fruit quality and soil fertility.

Literature cited Arnold SG, Anderson TM, and Holdo RM. 2014. Edaphic, nutritive, and species assemblage

differences between hotspots and matrix vegetation: two African case studies. Biotropica

46(4):387-394.

Harris F. 2002. Management of manure in farming systems in semi-arid West Africa.

Experimental Agriculture 38(02):131-148.

O’Reilly-Wapstra J, Potts B, McArthur C, and Davies N. 2005. Effects of nutrient variability on

the genetic-based resistance of Eucalyptus globulus to a mammalian herbivore and on

plant defensive chemistry. Oecologia 142(4):597-605.

Pinay G, Clement JC, and Naiman RJ. 2002. Basic principles and ecological consequences of

changing water regimes on nitrogen cycling in fluvial systems. Environmental

Management 30(4):481-491.