Maternal Effort, Food Quality, and Cortisol Variation During Lactation in Propithecus Coquereli in Northwestern Madagascar

Home , Silky sifaka, Woolly lemur

Maternal Effort, Food Quality, and Cortisol Variation During Lactation in Propithecus coquereli in Northwestern Madagascar

Abigail C. Ross

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Anthropology University of Toronto

Maternal Effort, Food Quality, and Cortisol Variation During Lactation in Propithecus coquereli in Northwestern Madagascar

Abigail C. Ross

Doctor of Philosophy

Department of Anthropology University of Toronto

2017 Abstract

The duration and quality of the infant care mothers provide is paramount to investigating life history theory. Maternal care is the principal determinant of infant survival and future reproductive success. Lactation is the most energetically expensive activity for mammals, in turn causing lactating females to compensate for drastically greater energy requirements. Lemur reproduction occurs under strict seasonal parameters to cope with harsh climatic conditions. I evaluated maternal behavioral care-giving effort towards infants over 26 postnatal weeks in

Coquerel’s sifaka (Propithecus coquereli) (n=10 infants, n=10 lactating females). Secondly, I evaluated the nutritional quality of foods consumed exclusively by lactating females (n=10).

Lastly, I examined stress responses across sex and reproductive classes in earlier (weeks 1-12) versus later lactation (weeks 13-24) (n=10 lactating females, n=19 adult males, n=8 non-lactating adult females). I conducted fieldwork over two consecutive birth seasons (2010 and 2011) in

Ankarafantsika National Park located in northwestern Madagascar. Earlier lactation occurs during the austral winter, and coincides with the seasonally driest time of the year. I quantified maternal care-giving by measuring infant transport position, carrier identity, and bodily contact.

Nutritional food quality was measured by protein, fiber, energy (n=123), and mineral content

(n=119). I quantified stress responses to determine how sex and reproductive classes respond to seasonal pressures and lactation by measuring cortisol variation in two lactation stages in

lactating females (n=180), adult males, (n=133), and non-lactating adult females (n=62). Infants were more altricial relative to other lemurs of similar body size. Allomaternal care was documented, although mothers were the primary infant transporters. Lactating females most frequently selected foods high in non-protein energy, and regularly selected foods high in available protein and fiber. Lactating females had lower cortisol relative to adult males that approached statistical significance, and significantly lower cortisol than non-lactating adult females. Adult males had significantly higher cortisol during the earlier lactation stage in comparison to the later stage. Understanding the behavioral, nutritional, and endocrinological responses of lactating females provides a comprehensive view of how maternal energetics, infant care, and infant development are dynamically performing during the most energetically constraining time of year in a stochastic environment.

iii

Dedication

To my exceptional father, Bill Ross (1950-2016)—a natural born world shaker.

Acknowledgments

“Words are like eggs: when hatched, they have wings.” “Ny teny toy ny atody: raha foy manana elatra.” ~ Malagasy proverb

Stumbling through all the theoretical, logistical, and statistical tribulations of a PhD program has been a tremendous and an exhilarating challenge. The country of Madagascar is an enchanting and mythical place, though can be frustrating and difficult to traverse. When I was covered in bug bites, weary from chasing sifakas, longing for a hot shower, and ready to jump on the next taxi brousse out of Ampijoroa to feast on a cheeseburger and an ice-cold beer; I would experience a magnificently beautiful thing in the forest, hear booming laugher in the village, or smell the ylang-ylang blossoms before dawn. This is Madagascar to me: the grittiest, yet most rewarding place I have ever been. The memory fills me with a deep, quiet satisfaction. Madagascar can be a sorrowful place, plagued by extreme poverty and disease, ravaged by political and economic instability, and the relentless destruction of its environment. But, it is also filled with a penetrating, vibrant richness and depth. I feel a tugging from within when I think of Madagascar, its people, and its lemurs. It will bring me back. Thank you to the Coquerel’s sifaka groups that let me join and share in their world: Citron (village of Ampijoroa), Fito (seven), Iva (low), Kambana (twins), Mainty (black), Rambo (tail), Vaovao (news), Volo (hair), and Zaza (baby). Your lessons are invaluable to me, both as a scientist and human, and I continue to learn from our time together in the forest.

More organizations and people in Madagascar contributed to this dissertation than I ever could have imagined. I thank the Madagascar Institute for the Conservation of Tropical Environments (MICET) for their logistical assistance ranging from acquiring export permits to finding a FedEx in Antananarivo where I could ship my biological samples. MICET welcomed me amiably into their country and provided tremendous backing throughout my project. My dissertation would not have been possible without their expertise. I thank the director of MICET, Benjamin Andriamihaja, for his gracious support. I also thank Tiana, Benji, Jean, Aja, Iton, Silvan, and David at MICET for their assistance and logistical support.

I sincerely thank my field assistant, Ravalohery Fara Nomena, for her diligent commitment to the project for 14 months. Nomena and her family welcomed me into their home in Itaosy. Nomena proved invaluable in countless ways and I feel incredibly lucky to have worked with her. Nomena is much more than an excellent field assistant; she is my colleague, close friend, interpreter, and cultural liaison. I admire her patience and skill collecting behavioural data. Nomena’s witty humor never failed to make me laugh and I always enjoyed our conversations in the forest. I owe the success of the project to her perseverance and look forward to working alongside her in the future.

Thank you to Randrianjaka (Njaka) Frankin for his devotion to the project as my park guide for 14 months. Njaka has an amazing ability to find sifakas in the forest and went far beyond the call of duty to help both Nomena and myself. He was never late once during the entire project. In fact, I remember Njaka calling my name from outside the tent to wake me up after sleeping through my alarm on several occasions. Njaka’s wife was pregnant with their second child and I recall communicating to him through a series of hand gestures and terrible Malagasy-English to take time off when the baby was born. Nomena chimed in to clarify and Njaka just smiled, shaking his head. Months later, he received a phone call while we were following a sifaka group in the forest to say that his wife was in labor. He told me that he would finish his work for the day and then go see his wife in Ambohimanga, the adjacent village to Ampijoroa. He was back working the next morning. He climbed the trees to collect my plant samples and learned the local tree names from his father, Zama, and other park guides. Chapter 3 would not have been possible with him. Njaka takes tremendous pride in everything he does, including my project and I will forever be grateful for his hard work.

Thank you to Ang and her family for providing nutritious food every day for Nomena, Njaka, and myself. Ang’s kind smile greeted us each morning with hot coffee gave us the energy we needed before heading into the forest. Ang gave us hardboiled eggs for breakfast every Sunday and even delivered potage to my tent when I sick.

Thank you to Madagascar National Parks, Ankarafantsika National Park, Ministere de l’Environnement et des Forets, Ecole Normale Supérieure, Parc Zoologique et Biologique de Tsimbazaza, and Missouri Botanical Gardens. These organizations granted me permission to

conduct research in their country and helped me navigate through Madagascar. The assistance of each organization was paramount to the execution of my project on the ground. A special thank you to Lalao Andriamahefarivo, Herisoa Manjakahery, and Faranirina Lantoarisoa at Missouri Botanical Gardens in Antananarivo for their help exporting my plant samples.

I sincerely thank Rakotondradona Remi, Director of Ankarafantsika National Park, for his support of my project. Thank you to Razaiarimanana Jacqueline and Rakotoarimanana Justin for their assistance and daily problem-solving skills. Ankarafantsika National Park was a wonderful place to work and call home because of the support from each of you. Thank you, Travis Steffens, Keriann McGoogan, Kim Valenta, Sharon Kessler, Danielle Levesque, Alida Hasiniaina, and Razafitsalama Mamy for your friendship and remarkable support in the field. We shared many incredible experiences together in a short time, and I am fortunate to have gained lifelong friends and colleagues.

Thank you to my supervisor Dr. Shawn Lehman for his academic and personal support during the last eight years. Shawn offered guidance and help when I needed it, but perhaps most importantly, he gave me the freedom to construct and implement my own project. Shawn taught me how to be an independent field scientist and trusted in my abilities even when I did not. His confidence in me instilled confidence in myself. Shawn went to bat for me on numerous occasions; his loyalty and dedication to his students is exceptional.

Thank you to the Smithsonian Conservation Biology Institute, Conservation Ecology Center, Nutrition Lab, National Zoological Park for allowing me to visit your lab for a summer. An enormous and gracious thank you to Dr. Mike Power. I first met Mike in 2007 when he graciously invited me to his lab to assay Goeldi’s monkey milk. I have been under his wing since that time. I have gained an incredible arsenal of knowledge from his expertise. Mike is an exceptional scientist and an extraordinary person to learn from. His tenacity is inspiring and his enthusiasm for science is infectious. Mike enjoys chatting with students, and continually blows my mind with his impromptu remarks about nutritional science. I am brimming with new ideas, reevaluating old thoughts, and delving into new topics of inquiry after every one of our conversations.

vii

I thank Dr. Joyce Parga for her assistance constructing my ethogram and extremely helpful suggestions for collecting behavioural data. I found Joyce’s approachability and encouragement so warm when I entered the Ph.D. program. Her courses and frequent emails tremendously expanded my knowledge of lemur reproduction and were essential to the formation of chapter 2. She continues to be a great mentor. I wish her happiness and success in Los Angeles.

Thank you to my committee members: Drs. Julie Teichroeb, Michael Schillaci, Rebecca Stumpf, and Becky Raboy for their revisions and thoughtful insight from the conception of my project to its conclusion. You each challenged the parts of my dissertation that needed more attention. Your contributions significantly improved all my chapters, my ability to think as a scientist, and will make future publications more successful.

Thank you, Michael, Jakubasz and Michael Maslanka at the Department of Nutrition, Smithsonian National Zoological Park. Your assistance started in Madagascar during the shipment of my samples and continued through my lab visit. I truly value my experience at the nutrition lab and gained an instrumental skill set because of your knowledge and thoughtful teaching. Thank you to Dr. Christina Petzinger for your assistance with my plant samples. You were a great resource for me in the lab and taught me a great deal about nutritional assays. A colossal thank you to the nutrition lab interns: Dr. Cari Lewis, Nicole Johnson, Jessica Cooper, Katie Murtough, and DaeKyu Lee. Your hard work and countless hours spent in the lab were profoundly appreciated. The meticulousness and perseverance with which you approached my nutritional assays extended far beyond your responsibilities and demonstrated true passion.

Thank you to Dr. Toni Ziegler, the Wisconsin National Primate Research Center, and Assay Services for acquiring my import permits, opening your lab to me, and conducting my cortisol assays. A very special thank you to Dan Wittwer for teaching me how to run the cortisol assay from start to finish. My first day in the lab Dan asked if I knew the difference between a beaker and a graduated cylinder; and when I responded, “Uh,” Dan knew he had his hands full. Dan is a delightful, easygoing teacher and an absolute pleasure to learn from. His sarcastic humor sustained me while hand grinding 412 fecal samples. He was determined to teach me how to successfully use a multichannel pipette, even after several failed attempts. Toni and Dan

viii

continued to answer questions long after I left the lab. Their immense dedication to the success of my cortisol assays will always be remembered.

Thank you to my incredible parents, Bill Ross (1950-2016) and Sue McCausland, for their unyielding love and support. You both taught me to reach for the stars, and to always follow my heart. I am lucky enough to still believe that anything is possible because of your parenting. Thank you for letting me play in the mud and eat boxelder bugs. I would not have discovered my passion for animals and the outdoors without your guidance. Dad, one of the last things you told me was how proud you were. I miss you always. I know as I write these words that your advice to me now would be: take it easy, and just be happy.

Erik, you are an endless source of love and support. Your steadfastness and perceptivity in all things lifts me. Thank you. From going to the Sadie Hawkins dance senior year of high school together, it took travelling half away around the world for us to find each other again.

Thank you to my little sister, Caitlin Ross, for always having my back. You are the best teacher I know. I hope that I can motivate and inspire students in the way you do. Thank you to my grandparents, Ed and Renee Ross (1921-2015). Your support over the years enabled me to pursue my passions. I am forever grateful. Thank you to Bob Lund for always lending an ear. Your esteemed insight has been, and continues to be, a vital influence in my individual and professional development. Thank you to Ron and Vickie Lund for your support, encouragement, and kind words that prevailed even during the toughest of times. Thank you to my dear friends Lisa Fischer, Melisa Smith, Jenny Kalousek, Ellie Pedersen, and Kate Nash for believing in me, providing much needed comic relief, and giving me perspective. I am lucky to have each of you in my life. Thank you to my three furry felines, Miel “Big Boy Roy” (2004-2014), Macko, and Baloo for their companionship and affection during the many hours spent on the computer.

I sincerely thank Drs. Leila Porter, Dan Gebo, and Mitch Irwin. Leila was my M.A. supervisor and Dan was on my thesis committee at Northern Illinois University. Leila and Dan prepared me for the rigors of a Ph.D. program, and continued to lend an ear when I moved back home to Rockford, Illinois to write my dissertation. Leila’s gentle way of teaching opened the discipline of Biological Anthropology to me. She is a remarkable person and mentor. Leila helped me

pinpoint my research interest in maternal-infant relationships. Her patient and nurturing way of engaging with students while being a compelling and scrupulous professor is an unmatched combination of traits I have encountered in academia. I truly admire Leila’s teaching style and it is my aim to exemplify what I learned from her moving forward in my career. Dan’s challenging courses were invigorating and innovative and developed my critical thinking and writing skills. His leadership encouraged me to apply for Ph.D. programs. Thank you, Dan, for offering to listen to apprehensions I had during my writing stage. Leila and Dan were both integral to my development as a graduate student and transition into becoming a professional primatologist. Thank you to Mitch for being my lemur mentor close to home. Your thoughts and guidance have been immensely valuable.

Thank you to Bill Buhl, my incredible high school biology teacher who first sparked my interest in science. Bill invited me to join his class trip to the Galápagos Islands when I was an undergraduate living in Madison. It was on that trip where I decided I would major in Anthropology. I am indebted to Bill for his encouragement throughout the years.

Thank you to my funding agencies for their generous contributions, including the Department of Anthropology and the Department Graduate Fellowships and Awards Committee Research Funds- University of Toronto, School of Graduate Studies Research Travel Grant- University of Toronto, Primate Conservation, Inc. Research Grant, American Society of Primatologists Conservation Committee Small Grant, The Explorers Club Exploration Fund, Edward and Renee Ross Charitable Foundation, Department of Anthropology Fellowship- University of Toronto, Faculty of Arts and Sciences Fellowship- University of Toronto, and support from S. Lehman's NSERC.

Thank you to the University of Toronto for providing a vivacious community in which to learn and an exceptional, enriching academic atmosphere. Thank you to the statistical consulting service at Northern Illinois University for their patience and mathematical wizardry with chapters 2 and 4. Thank you to digital cartographers, Erik Lund and Charlie Lunn, of Rockford Map Publishers for their ardent interest in nature that inspired them to create all the maps used in my dissertation. The maps were an integral component to my project and will be continue to be valuable to future research in Ankarafantsika National Park.

Table of Contents

Abstract ii

Dedication iv

Acknowledgments v

Table of Contents xi

Chapter 1: Introduction 1

1.1. Influences on animal life histories 1

1.1.2. Inherited environmental effects model 1

1.1.3. Maternal and allomaternal care 2

1.1.3.1. Lactation 4

1.2. Strepsirhines 5

1.2.1. Strepsirhine phylogeny 5

1.2.2. Conservation research justification 6

1.2.3. Propithecus spp. socioecology 7

1.2.3.1. Infant mortality 8

1.2.3.2. Seasonality in dry deciduous forests 9

1.3. Maternal effects model 9

1.4. Dissertation objective 10

1.4.1. Research hypothesis 11

1.5. Maternal behavioral care-giving background 11

1.5.1. Female social dominance 11

1.5.2. Basal metabolic rates during reproduction 12

1.5.3. Infant growth rates 12

1.5.4. Maternal body mass 13

1.5.5. Lactation and phenotypic plasticity 14

1.5.6. Infant behavior research justification 14

1.6. Maternal behavioral care-giving predictions 15

1.7. Nutritional food quality background 15

1.7.1. Propithecus spp. diet 16

1.7.2. Protein-to-fiber ratios 17

1.7.3. Gross energy 18

1.8. Nutritional food quality predictions 18

1.9. Stress response background 19

1.10. Temporal cortisol variation predictions 19

1.11. Dissertation overview 20

1.12. References 24

Chapter 2: Maternal Effort in Propithecus coquereli in Ankarafantsika National Park, Northwestern Madagascar 30

2.1. Abstract 30

2.2. Introduction 31

2.3. Methods 36

2.3.1. Study site and species 36

2.3.2. Data collection 38

2.3.3. Sampling methods 39

2.3.4. Data analysis 40

2.4. Results 41

2.4.1. Infant transport position: Ventral 41

2.4.2. Infant transport position: Dorsal 42 xii

2.4.3. Infant transport position: Independent 42

2.4.4. Comparison of infant transport positions from 1-26 weeks postnatal 43

2.4.5. Infant carriers 43

2.5. Discussion 46

2.6. Author note on publication 53

2.7. References 54

Chapter 3: Nutritional Food Quality of Foods Exclusively Selected by Propithecus coquereli Lactating Females 87

3.1. Abstract 87

3.2. Introduction 88

3.3. Methods 95

3.3.1. Study site and species 95

3.3.2. Botanical and biotic variable collection 96

3.3.3. Botanical processing and preservation 97

3.3.4. Macronutrient and micronutrient assays 98

3.3.5. Laboratory drying and grinding 98

3.3.6. Crude protein determination 99

3.3.7. Neutral detergent fiber and acid detergent fiber determination 99

3.3.8. Total mineral (ash) 100

3.3.9. Gross energy determination 100

3.3.10. Data analysis 100

3.4. Results 101

3.4.1. High available protein foods: Cluster 1 101

3.4.2. High fiber foods: Cluster 2 101

xiii

3.4.3. High non-protein gross energy foods: Cluster 3 101

3.4.4. Mineral (ash) content 101

3.4.5. Frequency of ingestion index 102

3.5. Discussion 102

3.5.1. High available protein foods: Cluster 1 104

3.5.2. High non-protein gross energy foods: Cluster 3 106

3.5.3. High fiber foods: Cluster 2 108

3.6. References 111

Chapter 4: Cortisol Variation Across Sex and Reproductive Classes in Propithecus coquereli 131

4.1. Abstract 131

4.2. Introduction 131

4.3. Methods 138

4.3.1. Study site and species 138

4.3.2. Fecal collection and preservation 140

4.3.3. Dried feces sample preparation 141

4.3.4. Cortisol steroid extraction and recovery 141

4.3.5. Cortisol steroid validation 141

4.3.6. Enzyme immunoassays (EIA) 142

4.3.7. Data analysis 142

4.4. Results 143

4.4.1. Lactating females 143

4.4.2. Adult males 144

4.4.3. Non-lactating adult females 144

4.4.4. Comparison across sex/reproductive classes 144 xiv

4.5. Discussion 145

4.6. References 152

4.7. Appendix 170

Chapter 5: Conclusions and Future Directions 180

5.1. Theoretical significance 180

5.2. Summary of findings 181

5.3. Future directions of study 189

5.4. Conservation implications 191

5.5. Evolutionary implications 195

5.6. References 197

List of Tables

Chapter 1: Introduction 1

Table 1.1. Gross milk composition in mammals 21

Table 1.2. Gross milk composition in primates 22

Chapter 2: Maternal Effort in Propithecus coquereli in Ankarafantsika National Park, Northwestern Madagascar 30

Table 2.1. P. coquereli group size and composition including infants season 1 62

Table 2.2. P. coquereli group size and composition including infants season 2 63

Table 2.3. P. coquereli focal data collection columns 64

Table 2.4. P. coquereli ethogram 65

Table 2.5. P. coquereli behavioral focal hours 67

Table 2.6. P. coquereli infant birth months 68

Table 2.7. Mixed effects linear regression comparison of infant position 69

Table 2.8. Mixed effects linear regression comparison of infant carrier identity 70

Table 2.9. Poisson regression considering number of occurrences mothers initiated and broke infant contact 71

Table 2.10. Poisson regression considering number of occurrences infants initiated and broke mother contact 72

Chapter 3: Nutritional Food Quality of Foods Exclusively Selected by Propithecus coquereli Lactating Females 87

Table 3.1. Identification of assayed foods selected by lactating P. coquereli 116

Table 3.2. Available protein, neutral detergent fiber (NDF), acid detergent fiber (ADF), non-protein gross energy (NPGE), and mineral (ash) values of foods selected by lactating P. coquereli 120

xvi

Table 3.3. Nutrient profile of high available protein foods consumed by lactating P. coquereli 124

Table 3.4. Nutrient profile of high neutral detergent fiber (NDF) foods consumed by lactating P. coquereli 125

Table 3.5. Nutrient profile of high acid detergent fiber (ADF) foods consumed by lactating P. coquereli 126

Table 3.6. Nutrient profile of high non-protein gross energy foods consumed by lactating P. coquereli 127

Table 3.7. Nutrient profile of minerals (ash) in foods consumed by lactating P. coquereli 128

Table 3.8. Frequency of ingestion index (FOI) 129

Chapter 4: Cortisol Variation Across Sex and Reproductive Classes in Propithecus coquereli 131

Table 4.1. P. coquereli fecal samples collected by individual sex/reproductive class and data collection season 156

Table 4.2. Comparison of cortisol in P. coquereli by individual sex/reproductive class and lactation phase from 1-24 weeks postnatal 157

Table 4.3. Comparison of cortisol in P. coquereli by individual sex/reproductive class from 1-24 weeks postnatal 158

Table 4.4. Mixed effects linear regression of cortisol in P. coquereli by individual sex/reproductive class and lactation phase from 1-24 weeks postnatal 159

Table 4.5. Mixed effects linear regression estimates of cortisol in P. coquereli by individual sex/reproductive class and lactation phase from 1-24 weeks postnatal 160

Table 4.6. Comparison of P. coquereli fecal cortisol and P. verreauxi fecal corticosterone 161

Chapter 5: Conclusions and Future Directions 180

xvii

List of Figures

Chapter 1: Introduction 1

Figure 1.1. Inherited environmental effects model: components of offspring phenotype 23

Chapter 2: Maternal Effort in Propithecus coquereli in Ankarafantsika National Park, Northwestern Madagascar 30

Figure 2.1. Ankarafantsika National Park, northwestern Madagascar 73

Figure 2.2. Percentage of time P. coquereli infants spent ventrally on all carriers 74

Figure 2.3. Percentage of time P. coquereli infants spent dorsally on all carriers 75

Figure 2.4. Percentage of time P. coquereli infants spent independently 76

Figure 2.5. Percentage of time P. coquereli infants spent ventrally, dorsally, and independently 77

Figure 2.6. Percentage of time P. coquereli infants spent carried by mothers 78

Figure 2.7. Percentage of time P. coquereli infants spent carried by adult males 79

Figure 2.8. Percentage of time P. coquereli infants spent carried by adult females 80

Figure 2.9. Percentage of time P. coquereli infants were transported by mothers, adult males, adult females, and independent 81

Figure 2.10. Occurrences of infant contact initiated by P. coquereli mothers 82

Figure 2.11. Occurrences of infant contact broken by P. coquereli mothers 83

Figure 2.12. Occurrences of mother contact initiated by P. coquereli infants 84

Figure 2.13. Occurrences of mother contact broken by P. coquereli infants 85

xviii

Figure 2.14. Occurrences of infant contact initiated/broken by P. coquereli mothers and mother contact initiated/broken by infants 86

Chapter 3: Nutritional Food Quality of Foods Exclusively Selected by Propithecus coquereli Lactating Females 87

Figure 3.1. Ankarafantsika National Park, northwestern Madagascar 130

Chapter 4: Cortisol Variation Across Sex and Reproductive Classes in Propithecus coquereli 131

Figure 4.1. The hypothalamic-pituitary-adrenal (HPA) axis, negative feedback response to chronic and acute stress, and effects of stressors on bodily processes 162

Figure 4.2. The biological response of animals to stress 163

Figure 4.3. Ankarafantsika National Park, northwestern Madagascar 164

Figure 4.4. Cortisol concentrations in P. coquereli lactating females 165

Figure 4.5. Cortisol concentrations in P. coquereli adult males 166

Figure 4.6. Cortisol concentrations in P. coquereli adult non-lactating females 167

Figure 4.7. Average cortisol in P. coquereli by sex/reproductive class 168

Figure 4.8. Cortisol comparison across sex/reproductive classes and lactation phases 169

Chapter 5: Conclusions and Future Directions 180

xix

List of Appendices

Appendix 1

P. coquereli fecal field collection and cortisol concentrations 170

Chapter 1 Introduction

“All who live under the sky are woven together like one big mat.” “Tsihy be lambanana ny ambanilantra.” ~ Malagasy proverb

1.1. Influences on animal life histories

Life history studies investigate individual variation in genotypic and phenotypic fitness through adaptations and constraints (Stearns 1992). Darwinian natural selection causes evolutionary changes in populations and produces adaptations. Adaptations allow an organism to survive and reproduce in its environment. “An organismal character constitutes an adaptation if it performs a function that is of utility to the organisms possessing it and if the character evolved by natural selection for that particular function” (Larson and Losos 1996, p. 187). Constraint is defined as a phylogenetically dependent state where each developmental stage hinges on the previous stage (Oster and Alberch 1982). Non-human primate life history models are predominately constructed using climatically stable or cyclic environments and have not considered the evolutionary responses or adaptive processes at work in erratic environments like Madagascar, where unpredictable intra- and inter-annual rainfall results in irregular and highly seasonal food abundance and distribution (Dewar and Richard 2007; Dunham et al. 2011; Wright 1999). Consequently, life history models are not designed for stochastic environments and maternal behavioral and ecological responses to stochastic environments are not represented in contemporary life history theory.

1.1.2. Inherited environmental effects model Inherited environment effects are defined as, “components of an offspring’s phenotype that are derived from the parent, apart from nuclear genes” (Rossiter 1996, p. 451). Thus, inherited environmental effects result from a combination of intrinsic and extrinsic factors, including parental environment, and interactions between parental environment and parental genotype (Rossiter 1996). Inherited environmental effects are represented in abiotic, nutritional, and a variety of other ecological characteristics occurring in the parental environment (Rossiter 1996). For example, food quality and other seasonal features (e.g., hormonal stress responses) are

1 2

environmental variables that contribute to inherited environmental effects (Rossiter 1996). These variables are permanent components of species’ environments, and thereby reflective of ecological and evolutionary processes at work (Rossiter 1996). Distinguishing inherited genotypic effects from observable phenotypic effects is problematic because a multitude of interrelated factors selectively act on offspring success (Lacey 1998; Rossiter 1996). The inherited environmental effects model is derived from a genetics model of inherited environmental effects, with the inclusion of offspring environmental variability (Eisen and Saxton 1983), and further addition of variability within the parental environment; thus comprising a comprehensive model of the genotypic and phenotypic effects selecting on offspring phenotype (Rossiter 1996) (Figure 1.1). My dissertation primarily addresses parental performance phenotype (source 4; Figure 1.1) (measured by maternal behavioral care-giving effort and temporal variation in cortisol concentrations across sex and reproductive classes); and parental environment (source 5, Figure 1.1) (measured by the nutritional food quality of foods consumed by lactating females). Inherited environment effects select on offspring phenotype by the, “contribution of the parental performance phenotype to offspring phenotype due to parental performance genotype; contribution of the parental environment; interaction between parental and offspring environment; interaction between parental environment and offspring genotype; and covariance between parental performance genes expressed in the parental and subsequent generations” (Rossiter 1996, p. 454-55) (see sources 4-8, Figure 1.1).

1.1.3. Maternal and allomaternal care Maternal care behavior is defined as the amount of time mothers engage in infant care-giving, and the quality and duration of this care (Pryce 1995). Parental investment in their offspring is a function of balancing maternal reproductive costs while maximizing infant care and survivability without impeding future female reproductive success (Trivers 1974). The typical, albeit not universal, mammalian parental investment pattern shows a positive correlation between body mass and reproductive function, which is often used to quantify maternal health (Clutton-Brock 1991; Dobson and Michener 1995; Lewis and Kappeler 2005b; Richard et al. 2000). Other proxies used to determine maternal health include hormonal stress responses (see Abbott et al. 2003; Bales et al. 2005; Beehner and McCann 2008; Boonstra et al. 1998; Brockman et al. 1998; Brockman et al. 2009; Busch and Hayward 2009; Casolini et al. 1997; Creel et al. 2002; Gould et al. 2005; Ostner et al. 2008; Saltzman and Abbott 2009; Setchell et al. 2008), and the nutritional

quality of foods consumed by mothers also reflect maternal health and are effective temporal measures of how mammals respond to their environments (see Burgess and Chapman 2005; Chapman et al. 2003; Ganzhorn 2002; Gould et al. 2011; Milton 1979; Norscia et al. 2006; Sauther and Cuozzo 2009). Mammalian maternal care is unique in that milk is the exclusive nutritional source available to infants until the introduction of solid foods, thereby requiring mothers to participate in care-giving for varying, though comparatively long, durations relative to other animal classes (Clutton-Brock 1991).

Maternal care is augmented with allomaternal care in some primate taxa (Hrdy 1976). Allomaternal care is an assortment of different behaviors used by non-mothers, in part, to reduce the energetic burden on lactating females (Ross and MacLarnon 2000). These behaviors can be divided in three broad categories: infant transport, babysitting, and energy transfer (reviewed in Tecot et al. 2013). The presence of allomaternal care is prevalent in the order Primates, though the form and frequency of allomaternal care is decidedly variable across taxa (Ross and MacLarnon 2000). Allomaternal caregivers are defined as nonbreeding (e.g., juveniles) or reproductive individuals that decide to expend energy on infants other than their own (Solomon and French 1997). Allomaternal care is typically thought to be adaptive to both mothers and caregivers, but it also poses risks. For example, infants may be injured due to mishandling by inexperienced caregivers, nursing time may be reduced, or mothers may decrease time spent foraging to increase vigilance while other caregivers are watching infants (reviewed in Tecot et al. 2013). Given these potential risks, the benefits of allomaternal care must be considered high for species to partake and include: shorter interbirth intervals, faster infant growth, improved predator protection for infants, thermoregulation, and increased breeding opportunities for males (reviewed in Tecot et al. 2013). Allomaternal may be particularly advantageous in difficult environments such as Madagascar (Wright 1999), where alleviating energetic stress may have immediate benefits for mothers and infants in times of extreme environmental instability. There is evidence of allomaternal care in numerous lemur taxa, though longitudinal data is severely lacking, including: fat-tailed dwarf lemur (Cheirogaleus medius), black lemur (Eulemur macaco), mongoose lemur (Eulemur mongoz), red-bellied lemur (Eulemur rubriventer), eastern lesser bamboo lemur (Hapalemur griseus), ring-tailed lemur (Lemur catta), silky sifaka (Propithecus candidus), diademed sifaka (Propithecus diadema), golden-crowned sifaka (Propithecus tattersalli), Verreaux’s sifaka (Propithecus verreauxi), and Coquerel’s sifaka

(Propithecus coquereli). In contrast to previous studies focused on haplorhines, there was no relationship found between allomaternal care and faster infant growth or shorter interbirth intervals in a comparative phylogenetic study of 23 lemur species (Tecot et al. 2012). Infant parking and nesting were related to faster life histories, as parking and nesting both positively correlated with fetal and postnatal growth rates (Tecot et al. 2012). The authors of this study suggest that lemurs may have already reached maximal growth rates, and decreased interbirth intervals may not increase rates of reproduction given that lemurs are strict seasonal breeders (Wright 1999, Tecot et al. 2012). It is likely that ecological variables (e.g., food quality and abundance) influence the expression of allocare in lemurs.

1.1.3.1. Lactation Lactating mothers represent a unique study opportunity because lactation is the single most energetically expensive activity in which mammalian mothers engage (Tardif 1994). In the case of some primates, the twofold cost of carrying infants while simultaneously lactating further amplifies this already high energetic cost (Bales et al. 2000; Tardif 1994). The exact selective pressures involved in the origin and evolution of mammalian lactation have not been identified (Sellen 2009), although the emergence of lactation as an exclusively mammalian strategy has undoubtedly caused the emergence of life history traits and evolved parental-infant relationships unique to mammals (reviewed in Dall and Boyd 2004). Within mammals, the Primates Taxonomic Order is distinguished by exceptionally extended life history traits including prolonged gestation length, reduced litter size, and delayed infant independence. Primate lactation strategies are characterized by extended length and frequency of nursing, and large milk volume (Oftedal 1984; Power et al. 2002). Milk is composed of dry matter (total solids), fat, protein, sugars, and ash (minerals). Early lactation is distinguished by changing milk composition prior to mid-lactation, and includes colostrum and transitional milk (Oftedal and Iverson 1995). Mid-lactation is operationally defined as the period of maximal lactation performance, and is more energetically constraining than early and late lactation (Iverson and Oftedal 1995). Late lactation is characterized by declining milk yields and mixed feeding, where infants are simultaneously consuming solid foods and milk (Iverson and Oftedal 1995). The nutrient composition of mammalian milk is exceedingly variable (Table 1.1), with primates generally producing relatively low quality, dilute milks in comparison to other mammals (Oftedal 1984; Power et al. 2002; Tardif et al. 2001; Tilden and Oftedal 1997). Dilute milk has

high water content, thereby having high sugar and low fat and energy content (M. Power, pers. comm.). Fat is the most variable nutrient, ranging from 0.2% in black rhinos (Diceros bicornis) to 60% in hooded seals (Cystophora cristata) (reviewed in Iverson and Oftedal 1995).

Milk nutrient composition is influenced by divergent infant behavioral care strategies (Tilden and Oftedal 1997). Dilute milk is typically found in genera with continual nipple access with infants that suckle frequently in comparison to higher fat, energy dense milks found in genera with less frequent nipple access (Ben Shaul 1962; Oftedal and Iverson 1995). Even though primates typically have dilute milk, there is significant variation in milk nutrient composition between closely related genera, most notably between lemurs and lorises, that is unrelated to body mass (Table 1.2) (Tilden and Oftedal 1997). Infants have restricted nipple access when they are “parked” and left alone in comparison to infants that are continually carried and therefore ingest a greater milk volume per suckling bout while suckling less frequently, in turn requiring more energy dense milk opposed to carried infants that ingest smaller volumes of dilute milk more frequently (Table 1.2) (Tilden and Oftedal 1997). Infants habitually nurse (and have easier access to nipples) in the ventral position compared to the dorsal position (reviewed in Tecot et al. 2013), though this measure cannot be used as a reliable measure of milk intake since young infants may not suckle while in the ventral position (Cameron 1996).

1.2. Strepsirhines 1.2.1. Strepsirhine phylogeny Madagascar reached its current geographic position 430km east of Mozambique approximately 120 million years ago (MYA), and separated from the Indian subcontinent between 80-90 MYA (Masters et al. 2006). The majority of wildlife in Madagascar evolved under geographically isolated conditions, thereby resulting in exceptionally high numbers of endemic taxa. Estimates of species richness reveal that 92% of vascular plants, excluding ferns, and 84% of land vertebrates are endemic to the island (Goodman and Benstead 2005). The Primates Order is divided into two taxonomic suborders, the Strepsirhini and Haplorhini. Phylogenetic evidence demonstrates the Strepsirhini and Haplorhini clades spilt 87 MYA (Perelman et al. 2011). Within Strepsirhini, the ancestors of Lemuriformes (Malagasy lemurs) / Chiromyiformes (Malagasy aye-aye) and Lorisiformes (lorises, galagos, pottos) spilt 68.7 MYA, with the origins

of these infraorders estimated at 58.6 MYA and 40.3 MYA, respectively (Perelman et al. 2011). There are four taxonomic families of Lemuriformes found singularly in Madagascar: Cheirogaleidae, Lepilemuridae, Lemuridae, and Indriidae. Lemuriformes first evolved 38.6 MYA, and remain partially phylogenetically unresolved due to current taxonomic debates at the subspecific level (Perelman et al. 2011). Chiromyiformes are represented by the single genus Daubentonia, and diverged from a common Lemuriformes ancestor (Perelman et al. 2011). Considering the four families within Lemuriformes, the Lemuridae emerged first, and subsequently the Indriidae, with a monophyletic lineage spilt that occurred 32.9 MYA, and resulted in the formation of the two sister lineages Lepilemuridae and Cheirogaleidae (Perelman et al. 2011). Indriidae contains three genera: Indri, Avahi, and Propithecus. Within Propithecus, the diademed sifaka (Propithecus diadema) is the eastern geographic type and Verreaux’s sifaka (Propithecus verreauxi) is the western geographic type, with the P. verreauxi clade subsequently splitting into verreauxi-deckeni-coronatus in southwestern Madagascar and the coquereli- tattersalli clade in the northwest (Mayor et al. 2004; Pastorini et al. 2001). Previous studies of Coquerel’s sifaka (Propithecus coquereli) in Ankarafantsika National Park (ANP) (Richard 1974; Richard 1976; Richard 1978; Richard 1985; Richard 1987) classified it as P. verreauxi prior to the taxonomic elevation of P. coquereli to a separate species (Mayor et al. 2004).

1.2.2. Conservation research justification There has been an 80% reduction of forested habitats during the last 50 years in Madagascar (Harper et al. 2007). Dry deciduous forests are one of the most degraded biomes in the world as a result of large-scale logging operations (Ganzhorn et al. 2001). Only 3% of dry deciduous forest cover remains in Madagascar and ANP is one of the largest existing deciduous forest blocks (Ganzhorn et al. 2001; Smith 1997). Unregulated use of natural resources in the forms of slash-and-burn agriculture, hunting, logging, fuel wood collection, and seasonal dry forest burning for cattle pasture coupled with rapid population growth are the primary causes of deforestation in Madagascar (Kull 2000).

P. coquereli is classified as an endangered species on the International Union for Conservation of Nature (IUCN) Red List. This lemur is confined to two protected areas in northwest Madagascar, the Bora Special Reserve and ANP. The highest P. coquereli densities occurring in the dry deciduous forests of Ankarafantsika region (Mittermeier 2010). An estimated population

of approximately 47,000 P. coquereli currently live in ANP (Kun-Rodrigues et al. 2014). Density estimates range from 5-100 individuals/km2, with habitat quality (i.e., negative effects of roads and forest edges) as the principal factor for this high variability (Kun-Rodrigues et al. 2014). An estimated 5 individuals/km2 presently exist in Ampijoroa (Kun-Rodrigues et al. 2014), in comparison to an estimated 60-75 individuals/km2 in the 1980s (Albignac 1981). Ampijoroa is a village within ANP and includes its surrounding regions. This is a rapid decrease of more than 90% of the P. coquereli population in Ampijoroa during the last 30 years (Kun- Rodrigues et al. 2014). Hunting pressure on P. coquereli have increased dramatically during recent years in ANP because their relatively large body size produces greater protein yields relative to other sympatric lemurs (Gerardo and Goodman 2003). Adult P. coquereli weigh between 3.7-4.3kg, with a head-body length of 42-50cm, and total length of 93-110cm (Kappeler 1991; Ravosa et al. 1993; Tattersall 1982). Additionally, human migration influxes have decreased protection previously provided by region-specific food taboos (Gerardo and Goodman 2003).

1.2.3. Propithecus spp. socioecology Propithecus spp. social groups are flexible in terms of composition and size, with the average size ranging between five to six individuals (Kubzdela 1997; Lewis and van Schaik 2007; Richard 1974; Richard 1985). Larger numbers of Propithecus females in social groups increases feeding competition while negatively influencing reproduction (Kubzdela 1997). Social relationships between males in adjacent groups is the primary factor contributing to their reproductive access to females, rather than relationships between males and females (Young et al. 1990). Females encourage subordinate males to enter social groups to increase mating choice, vigilance, and facilitate intergroup hostility (Richard 1985).

“Bet-hedging” is a biological strategy that assists in mothers achieving the balance between self- preservation and providing infant care (Stearns 1976). “Bet-hedging” specifically refers to a collection of life history traits where higher infant mortality rates, often more typical in unstable environments, drive females to reproductively invest less per gestation event (Stearns 1976). Females live longer in these environments, and by doing so have increased rates of reproduction over a longer duration, thereby increasing the probability that some infants will survive unstable environmental conditions and eventually reproduce (Stearns 1976; Stearns 1992). Propithecus

spp. females are considered exemplary “bet-hedgers” by displaying unusually “slow” life histories (Richard et al. 2002). P. verreauxi females only reach sexual maturity after five years, relatively late in life, and often do not reproduce until six years of age, with females reproducing and living longer than expected compared to mammals of similar body size (Richard 1976; Richard et al. 2002). The energy conservation hypothesis emerged from Stearn’s (1972) work and proposes that less postnatal maternal investment is present in lemurs relative to other nonhuman primates, resulting in lemur mothers reserving energetic resources for subsequent births due to high infant mortality (Jolly 1966; see Wright 1999).

1.2.3.1. Infant mortality Infant and juvenile primates typically experience higher mortality rates relative to breeding adults as they become less dependent on mothers and more self-reliant; for example, heightened predation risk and foraging incompetence are two risk factors contributing to higher mortality rates in young primates (Janson and van Schaik 2002). Ring-tailed lemurs (Lemur catta) and P. verreauxi infants incur higher mortalities within the first postnatal year in comparison to gregarious anthropoids of similar body size (reviewed in Gould et al. 2003). Only 52% of P. verreauxi infants survive the first postnatal year and fatalities occur primarily during the wet season or shortly after birth, although no significant seasonal pattern has been detected (Richard et al. 2002). This is a high infant mortality rate relative to 29% in white-faced capuchins (Cebus capucinus) (Fedigan et al. 1996) or 16-19% in red howler monkeys (Alouatta seniculus) (Crockett and Rudran 1987), especially given the low reproductive output in Propithecus social groups of one infant per female every two years (Richard et al. 2002). L. catta infant mortality increased from 52% in a non-drought year to 80% during a drought year and 20% of all adult females in Beza-Mahafaly Reserve perished during this time, demonstrating that increased environmental stress on mothers contributes to high lemur infant mortality (Gould et al. 1999; Gould et al. 2003). The L. catta population in Beza-Mahafaly Reserve recovered quickly post- drought as L. catta females reproduce annually, have high annual birth rates (.80-.86), and reach sexual maturity earlier (between two to three years of age) than Propithecus spp. (Gould et al. 1999; Gould et al. 2003).

1.2.3.2. Seasonality in dry deciduous forests Propithecus spp. infants in western dry forests are born during the dry season months in the austral winter (June-August) and weaned during the subsequent wet season (January-February) (Lewis and Kappeler 2005b; Young et al. 1990). Two strategies have been proposed to explain the variation seen in the timing of reproduction in primates. The first strategy suggests that females conceive during intervals of high or low food availability, and subsequently resources are the most abundant during the second half of lactation (van Schaik and van Noordwijk 1985), which is the most energetically expensive component of reproduction (Oftedal 1993; Tardif et al. 2001; Trivers 1974). In the second strategy, females conceive during a peak in food abundance and store energy reserves until mid-lactation, which occurs during periods of decreased food availability (van Schaik and van Noordwijk 1985). P. verreauxi females cannot employ the second strategy given that they lose 18% of their body mass during the dry season and hence cannot store sufficient fat reserves for periods of decreased seasonal food availability (Lewis and Kappeler 2005b). P. verreauxi follows the first reproductive strategy, where the birth season coincides with periods of reduced food availability, and the most energetically costly portion of lactation corresponds with greater food availability (Lewis and Kappeler 2005b; Young et al. 1990). This trade-off allows mothers to have the greatest access to food resources when experiencing the greatest level of energetic cost. Subsequently, early infant development occurs when resources are the scarcest and infants become independent from their mothers during periods of high food availability. Given differences in reproductive physiology, males and lactating females likely physiologically respond differently to seasonal constraints. Males also face energetic constraints during this time due to reduced food availability. Additionally, the risk of infanticide increases when dependent infants are present, and resident males must protect infants from immigrant males during lactation. Consequently, these factors may induce an elevated stress response in males during lactation. In contrast, lactating female may experience a reduced stress response relative to adult males to more successfully care for infants, though may also have an overall greater stress response during earlier lactation opposed to later lactation.

1.3. Maternal effects model

The maternal effects model postulates that the quality of the physical environment and care- giving a mother provides influences the development and fitness of her offspring (see Bernardo

1996; Cheverud and Wolf 2009; Galloway 2005; Inchausti and Ginzburg 1998; Kirkpatrick and Lande 1989; Lacey 1998; Maestripieri 2009; Maestripieri and Mateo 2009; Mateo 2009; McGinley et al. 1987; McLaren 1981; Mousseau and Fox 1998; Rossiter 1994; Willham 1972). Until recently, consideration of maternal effects had been largely absent from life history studies (Roosenburg 1996), which were primarily evaluated from a genetic, inherited perspective instead of examining the role of acquired environmental effects in maternal performance and subsequent effects on offspring (Bernardo 1996; Cheverud 1984; Rossiter 1998). Assessing maternal effects quantifies individual differences across generations over time, in turn measuring the rapidity and intensity of natural selection (Mateo 2009; Stearns 1976). Postnatal maternal effects directly influence offspring outside of the womb, whereas prezygotic (e.g., offspring genotype) and prenatal (e.g., placenta) maternal effects are mediated through the mother’s physiology (Rossiter 1998). Documenting significant changes in maternal care-giving behavior without solely capturing individual maternal behavioral variation has been successful in recent mammalian studies (Champagne et al. 2003; Fairbanks and McGuire 1995), which is of particular importance given the small sample sizes characteristic of many field studies. Studies of postnatal maternal effects provide data on intergenerational phenotypic plasticity since mothers quickly respond to environmental cues thereby influencing their immediate offspring, in turn influencing the subsequent generation (Bernardo 1996). Maternal effects are more relevant in mammalian evolutionary dynamics relative to any other taxa due to the extensive care-giving provided by mothers that influences offspring even after weaning (Reinhold 2002).

1.4. Dissertation objective

My study examines maternal behavior, food nutrient composition, and stress responses in P. coquereli, an endangered lemur species belonging to the taxonomic family Indriidae, that inhabits a localized geographic range in the tropical dry deciduous forests in northwestern Madagascar. The objective of my dissertation is to quantify P. coquereli maternal care-giving behavioral effort from birth until six and a half months postnatal concurrently with measuring food nutrient values and identifying nutrient selection profiles in lactating females and temporal cortisol variation in lactating females compared to other group members (adult males and non- lactating adult females). I will demonstrate how mothers balance their nutritional needs and physiological responses with providing infant care during the seasonally energetically depletive

lactation period. This will advance current discussions in evolutionary anthropology on maternal care-giving and its potential effects on offspring development.

1.4.1. Research hypothesis The inherited environmental effects (Rossiter 1996) and maternal effects models (Bernardo 1996) served as the theoretical foundation from which I formulated the hypothesis that I tested. My dissertation tests the following hypothesis and eleven related predictions established a priori:

H1: P. coquereli mothers experience greater postnatal reproductive stress, which is measured by behavioral care-giving effort, nutritional food quality, and cortisol stress responses (see predictions below) during early/earlier-mid lactation (designated as 1-12 weeks postnatal) (May- August) in comparison to later-mid/late lactation (designated as 13-26 weeks postnatal) (September-December) due to the increased energetic costs of lactation while simultaneously caring for dependent infants during the driest seasonal months in the austral winter when food quality is low.

H0: There is no difference in the postnatal reproductive stress P. coquereli mothers experience during early/earlier-mid lactation (designated as 1-12 weeks postnatal) (May-August) in comparison to later-mid/late lactation (designated as 13-26 weeks postnatal) (September- December).

1.5. Maternal behavioral care-giving background 1.5.1. Female social dominance Propithecus spp. live in matrifocal groups and females outrank males in feeding priority (Richard 1974; Richard 1985). Female social dominance is a result of the high energetic cost of reproduction (Jolly 1984) and females often synchronize reproduction with the most favorable environmental conditions (Lewis and Kappeler 2005b; Young et al. 1990). Gestation length is approximately 163 days and females give birth to single offspring (Petter-Rousseaux 1962). One infant is born per female, with an interbirth interval of 24 months which is only reduced to one year if the neonate dies, indicating that reproduction is too energetically expensive to occur annually (Richard et al. 2002). All-parental behavior occurs (Richard 1974; Richard 1985), but in a captive study on a single P. coquereli infant, the infant spent the most time on mother’s

relative to other group members, and the least time with immigrant males and subordinate females (Grieser 1992).

1.5.2. Basal metabolic rates during reproduction Females in species expressing female social dominance, particularly in Propithecus spp., metabolically invest more in neonates per day of gestation in comparison to male dominant or co-dominant strepsirhines (Young et al. 1990). Basal metabolic rate (BMR) measures the rate at which energy is depleted by an organism, excluding the influence of environmental or behavioral effects. BMRs are more difficult to quantify in primates than in other mammals because primates respond aversively to stressful conditions and capture of wild primates is necessary in order to measure BMR via oxygen consumption (Genoud 2002). The positive relationship between lemur BMR and prenatal maternal investment (measured as average daily maternal energy output) supports the prediction that lemurs have high prenatal investment in neonates for their BMR (Young et al. 1990). Thus, high prenatal investment appears to have contributed to the evolution of female dominance in lemurs, but other selective pressures also influenced its emergence since female dominance is not a characteristic exclusive to taxa with low BMRs (Young et al. 1990). Body mass is the primary determinant of mammalian BMRs, but there is also considerable mass-independent variability, demonstrating that physiological and environmental adaptations clearly play a role in accurately assessing BMRs and their rate of change under variable conditions (Genoud 2002). Haplorhines typically have higher BMRs than strepsirhines (reviewed in Harcourt 2008). Lemurs have low BMRs relative to haplorhines that elevate during reproduction (Richard and Nicoll 1987). The unpredictable climate in Madagascar (see Dewar and Richard 2007) has been used as the principal explanation for low BMRs in lemurs, but low BMRs are more widespread among African strepsirhines than previously thought and further studies are needed to test whether lemurs do indeed have a lower relative BMR for their body mass than other strepsirhines (Harcourt 2008).

1.5.3. Infant growth rates Strepsirhines produce neonates with smaller body mass and rapid postnatal growth rates of shorter durations in comparison to haplorhines (Leigh and Terranova 1998; Leutenegger 1973). Lemurs are generally sexually monomorphic in body size and the selective pressures

contributing to the lack of dimorphism are both ontogenetic and environmental (Leigh and Terranova 1998). There is a lack of sex difference in most lemur postnatal growth rates, a trait referred to as bimaturism (Leigh and Terranova 1998, but see Tennenhouse 2016). The similar growth rate between sexes is unexpected given high rates of intermale competition and seasonal breeding in lemurs (Kappeler 1993; Mass et al. 2009; Richard 1991). The extent of bimaturism can be influenced by seasonality, photoperiod sensitivity of ontogeny, and female reproductive synchrony (Leigh and Terranova 1998). Ontogenetic adaptations to seasonal food abundance and quality is the primary restraint on bimaturism because male growth is limited as a result of decreased food abundance and quality during the dry season (Leigh and Terranova 1998). Lemur infants are still relatively altricial despite rapid postnatal growth (Jolly 1984). Accelerated postnatal growth is related to the rapid or concentrated milk nutrient transfer that is needed to wean infants more quickly (Power et al. 2002).

1.5.4. Maternal body mass Numerous parental and environmental sources considered within the inherited environmental effects model are responsible for the quality and duration of prezygotic, prenatal, and postnatal investment in offspring. Reproductive females must acquire and retain sufficient resources to sustain pregnancies and subsequently provide behavioral and nutritional care of infants. Maternal body mass is clearly a vital indicator of pre-and postnatal infant growth and thus is pertinent to infant survivorship and future fecundity (see Altmann and Alberts 2005; Leigh and Terranova 1998; Tardif and Bales 2004). Interestingly, the duration of lactation in large-bodied mammals is weakly correlated with both maternal and neonatal body mass and appears more dependent on overall maternal health (Lee et al. 1991). Lactation duration cannot be attributed to a single source, but rather results from a combination of individually specific physiological and seasonally contingent variables. Weaning age is reflective of environmental quality, where earlier weaning can result from a mother’s inability to sustain lactation due to low environmental quality (Lee et al. 1991). Conversely, high milk nutrient transfer resulting from greater food quality can cause weaning weight to be reached more rapidly than in lower quality environments, in turn facilitating earlier weaning (Lee et al. 1991).

1.5.5. Lactation and phenotypic plasticity The Callitrichidae are a classic primatological example of coping with the energetic costs of lactation with phenotypic plasticity. Phenotypic plasticity is operationally defined as the genotypic capability to produce alternative morphological, physiological, or behavioral states in response to the surrounding environment (West-Eberhard 1989). Callitrichids, excluding Goeldi’s monkey (Callimico goeldii), experience severe energetic constraints due to biannual twin births, and combat this cost with reduced maternal behavioral care and high rates of paternal care, where fathers are the primary carriers from the time infants are only a few days old (Tardif et al. 1993). In contrast, C. goeldii underwent a reduction in litter size which enables mothers to delay the onset of paternal carriage and extend weaning age, thus demonstrating divergent behavioral strategies in closely related callitrichid species (Ross et al. 2010). Wied’s marmoset (Callithrix kuhlii) mothers that conceived early in their postpartum phase significantly reduced behavioral investment in infants, by carrying less and rejecting infants more frequently to conserve energy for the subsequent litter (Fite et al. 2005). In regards to this study, paternal care in P. coquereli has only been measured in a captive setting (Bastian and Brockman 2007), and longitudinal data on the quality, duration, and variability of allomaternal care in wild. P. coquereli infant carriage have not been available until my dissertation.

1.5.6. Infant behavior research justification Although behavioral development in Propithecus spp. infants has been examined in prior studies, circumspection must be exercised due to the lack of longitudinal studies, noncontinuous data, or small sample sizes in two wild studies (n=3 in Jolly 1966; n=9 in Richard 1976) and two captive studies (n=1 in Eaglen and Boskoff 1978; n=2 in Grieser 1992). Propithecus spp. infants are carried until approximately six months after birth, coinciding with the weaning process (Jolly 1966; Richard 1976). Although P. verreauxi infants began to taste solid foods one to two weeks following birth, infants often tasted foods not consumed by adults and consumed foods exclusively eaten by group members by five months postnatal (Richard 1976). Unexplained variation exists in documenting infant development. Sequences of independent locomotive behavior were first observed at fifteen days (Richard 1976) in comparison to two months postnatal (Jolly 1966).

1.6. Maternal behavioral care-giving predictions

The maternal effects model (Bernardo 1996) and the numerous supported energetic constraints associated with lactation serve as the theoretical underpinnings for my predictions on maternal- infant relationships. I predict that the duration, frequency, and type of P. coquereli maternal infant carriage will decrease as infants increase in age, as evidenced by strepsirhine growth rates. Infants become less dependent on their mothers over time for their transport and nutritional needs. I predict P. coquereli mothers will be the principal infant behavioral care-givers as evidenced by high prenatal maternal investment that has contributed to the evolution of female feeding priority and dominance in some lemur species, thus enabling mothers primary access to higher quality food resources. I propose that lactating females sufficiently offset the high energetic costs of lactation and infant care without comprising their current or future reproductive success. P1. Maternal behavioral care-giving effort will decrease as P. coquereli infants increase in age as measured by infant transport position (ventral, dorsal, independent). Infants will spend the greatest duration in the ventral position, followed by the dorsal position, and the least duration independently from carriers from birth until weaning (twenty-six weeks postnatal). P2. P. coquereli mothers will provide the majority of infant behavioral care-giving relative to adult males and non-mother adult females as measured by the frequency and duration of infant carriage from birth until weaning. P3. Infant contact initiated by P. coquereli mothers will decrease while infant contact broken by mothers will increase as infants age. P4. Contact with mothers initiated by P. coquereli infants will decrease while contact with mothers broken by infants will increase as infants age.

1.7. Nutritional food quality background

Food selection, foraging ability, and thermoregulation are influenced by food quality, thereby contributing to the amount of available energy mothers can invest in infant care-giving while maintaining their own nutritional needs (Lee and Bowman 1995; Stearns 1977). Few studies have examined nutritional food quality specifically during lactation in non-human primates. Lactation increases the quantity of amino acids needed to energetically sustain females, in turn significantly increasing protein requirements (Jessop 1997). Non-human primate nutritional

models are often rooted in human nutritional models, and therefore do not accurately represent protein sources available to wild primates (Oftedal 1991). Protein, the primary nutrient involved in mammalian reproductive function, requirements increase by more than 1/3 during early lactation in humans (Oftedal 1984; Cameron 1996; Tilden and Oftedal 1997; Power et al. 2002). Protein requirement guidelines are primarily based on captive animals that experience reduced activity, fewer nutritional demands, and environmental stress in comparison to wild animals. There is a lack of agreement over the amount of protein needed to sustain wild mammals during gestation and lactation, in part because the laboratory and statistical significance of protein is a measure of total protein, which does not accurately reflect the biological significance of the food item since the amount of available protein is unknown (Gogarten et al. 2012; Oftedal 1991).

1.7.1. Propithecus spp. diet Propithecus spp. have highly seasonal diets and are primarily folivore-frugivores, with flowers and bark being consumed in smaller quantities during certain times of the year (Hemingway 1998; Irwin 2008b; Lewis and Kappeler 2005a; McGoogan 2011; Norscia et al. 2006; Richard 1974). In terms of digestive anatomy, P. coquereli has a relatively long, spiraled colon and large, sacculated caecum (Campbell et al. 2004), and are classified as caeco-colic fermenters since either the caecum or colon is the primary fermentation chamber (Lambert 1998). This fermentation type decreases the amount of protein available prior to digestion and therefore, best suits taxa that consume digestible compounds (Alexander 1993). The large quantities of leaves consumed by P. coquereli are higher in insoluble fiber than other plant parts (Lambert 1998). P. verreauxi select foods primarily based on nutritional quality throughout the year, with protein and sugar consumption the highest when more fruit and flowers are consumed during wet season months (Norscia et al. 2006). The greatest diversity in foods consumed by P. diadema occurred during the calendar months when leaf consumption was greatest, suggesting that Propithecus spp. can afford to be more selective when better quality foods are seasonally available; either because higher quality foods fulfill nutritional requirements more quickly than lesser quality foods, or to avoid secondary compounds that can inhibit digestion or be toxic in larger quantities (Irwin 2008b). P. diadema consume markedly more fruits and seeds during the wet season, and shift to consuming greater quantities of leaf buds and flowers during the dry season (Irwin 2008b). P. coquereli follows a similar pattern of feeding predominantly on mature leaves and dormant buds during the dry season and shift to young leaves, fruit, and flowers during the wet

season (Richard 1974). P. coquereli consumed bark throughout the dry season, but did not consume dead wood; the opposite pattern was found during the wet season, although insectivorous behavior was not observed while groups consumed bark (Richard 1974). A more recent long-term study of the same species found that both dead and alive wood was consumed throughout the year (McGoogan 2011). This same study found that P. coquereli adult females spent significantly more time feeding than adult males, though this may be due to the group sex ratio of one male to four females (McGoogan 2011). Plant parts consumed by P. coquereli are spatially auto correlated and likely related to the synchronous patterns of fruiting and flowering characteristic of Madagascar (McGoogan 2011).

1.7.2. Protein-to-fiber ratios Protein-to-fiber ratios often determine leaf choice, with primates preferentially consuming leaves with a high protein-to-fiber ratio (Milton 1979). Trees along habitat edges receive more sunlight than trees closer to the forest interior, and can result in leaves that occur on habitat edges having higher protein-to-fiber ratios (Ganzhorn 1995a). Protein-to-fiber ratios and food consumption have been primarily examined in New World monkeys (Milton 1979) and colobines (Chapman et al. 2004; Gogarten et al. 2012), and have recently been evaluated in L. catta (Gould et al. 2011) and P. coquereli (McGoogan 2011). P. coquereli groups occupying home ranges closes to habitat edges showed no difference in the nutritional quality of foods consumed relative to forest interior groups (McGoogan 2011). All groups consumed foods high in crude protein and high protein-to-fiber ratios (McGoogan 2011). Gestating and lactating females forage less and consume leaves higher in protein relative to other group members to conserve metabolic expenditures (in callitrichids, Goldizen 1987; Price 1991; in lemurs, Sauther 1994; Vasey 2002). Conversely, Gould et al. (2011) found no differences in nutrient intake between gestating and lactating L. catta in comparison to conspecifics. Instead, the study found that early gestating females spent a greater amount of time feeding than during the period of early/mid-lactation (Gould et al. 2011). Young leaves consumed by colobines were found to have more overall protein, were consumed more frequently, had higher protein-to-fiber ratios, and were more digestible than mature leaves (Chapman et al. 2004).

1.7.3. Gross energy The term gross energy refers to the measurement of energy released after heat combustion in a laboratory setting, and is exclusively dependent on the nutrient composition of a food (Maynard and Loosli 1965b). Accordingly, a gross energy value is analogous to the energetic density of a food (Hinde and Milligan 2011), with high fat foods having higher gross energy values than carbohydrates, and high protein foods typically having moderate gross energy values (Maynard and Loosli 1965b). Metabolizable energy is a term used to describe the energy remaining after digestion and absorption, excluding indigestible material. Digestion is the chemical breakdown of food into usable compounds. Absorption is the transfer of these digested compounds across the gastrointestinal tract into the bloodstream. Metabolizable energy measures the absolute energy consumed by an animal, whereas gross energy measures the total caloric content of the food item. Thus, considering metabolizable and gross energy in tandem more accurately assesses the biological significance of foods, rather than considering either energy measure independently. All L. catta group members consume foods with significantly higher gross energy during early gestation than during early/mid-lactation, but no sex differences were found, which is likely due to extreme climatic events like cyclones (Gould et al. 2011). An alternative explanation is that males experience nutritional duress after mating competition, and require foods with higher nutritional value during the early gestation period (Gould et al. 2011). P. coquereli spent most of their annual activity budget consuming foods high in gross energy (Acacia spp.) and high protein-to-fiber ratios (McGoogan 2011). The nutritional quality of foods selected by lactating P. coquereli females was unknown prior to my dissertation.

1.8. Nutritional food quality predictions

I predict that P. coquereli mothers will select relatively high quality foods, which I define as foods high in available protein, low in fiber, high in non-protein gross energy, or high in minerals. My prediction is based on the theoretical groundwork of optimal foraging theory (Stephens and Krebs 1986), nitrogen maximization (Mattson 1980), fiber limitations (Milton 1979), and energy maximization models (Schoener 1971). P4. P. coquereli lactating females will select foods high in available protein. P5. P. coquereli lactating females will select foods high in minerals. P6. P. coquereli lactating females will select foods low in fiber.

P7. P. coquereli lactating females will select foods high in non-protein energy.

1.9. Stress response background

Variation in human maternal care is influenced by endocrinological fluctuations that determine a mother's attachment to her infant (Fleming et al. 1997). Socioendocrinology studies indicate that hormonal fluctuations are strongly associated with seasonality, sociality, reproductive function, and offspring development (e.g., Beehner and McCann 2008; Boonstra et al. 1998; Brockman et al. 2009; Gould et al. 2005). Glucocorticoids are steroid hormones that regulate glucose metabolism in the adrenal cortex, and are released into the bloodstream in response to acute stressors. Stress does not necessarily incite negative responses since energy is made available at critical moments, thereby improving biological fitness (Keay et al. 2006). Cortisol is a glucocorticoid that suppresses the immune system under chronic conditions. Distinguishing between stress and distress is essential when evaluating stress responses. This distinction can be measured as the biological cost of the stress to the animal, and is present under acute and chronic circumstances (Moberg 2000). It is not the amount of stress an animal experiences that is quantified when glucocorticoids such as cortisol are measured, but more accurately the stress response of the animal to its environment during a particular time (Busch and Hayward 2009). Climatic instability, food quality and availability, reproduction, infant presence, and predation pressure all produce elevated glucocorticoid levels in non-human primates (e.g., Abbott et al. 2003; Bales et al. 2005; Brockman et al. 2009; Fichtel et al. 2007a; Girard-Buttoz et al. 2009; Gould et al. 2005; Ostner et al. 2008; Rangel-Negrín et al. 2009; Saltzman and Abbott 2009; Setchell et al. 2008; Ziegler 2000). For example, P. verreauxi males experience an elevated stress response when infants are present in comparison to when infants are absent from groups (Brockman et al. 2009). Infant presence coincides with elevated infanticide risk during the lactation period from immigrant males while infants are still dependent on their mothers for milk (Brockman et al. 2009). Temporal variation in P. coquereli stress responses within or between sexes and reproductive classes have not been studied prior to my dissertation.

1.10. Temporal cortisol variation predictions

My predictions are formulated from the cortisol-adaptation (Bonier et al. 2009b) and brood-value hypotheses (Heidinger et al. 2006). I predict P. coquereli mothers and adult males will have

significantly higher fecal cortisol levels during early/earlier-mid lactation, coinciding with the peak of the annual dry season, in comparison to later-mid/late lactation when food quality and abundance begins to improve. Infant births occur during the peak of the austral winter when food quality in the forest is the lowest, risk from immigrate males is intensified, infants are the most altricial, and lactation constraints are the highest. Since Propithecus spp. have a low reproductive output, this makes the reproductive value of infants relatively high (see Heidinger et al. 2006) in comparison to species that produce more offspring during the course of a lifetime. This, in turn, incites a decreased stress response for mothers to successfully care for their infants and prepare for future pregnancies. I predict that P. coquereli mothers will have significantly lower fecal cortisol levels relative to adult males and non-lactating adult females. P9. P. coquereli lactating females will have significantly higher fecal cortisol levels during early/earlier-mid lactation relative to later-mid/late lactation. P10. P. coquereli adult males will have significantly higher fecal cortisol levels during early/earlier-mid lactation relative to later-mid/late lactation. P11. P. coquereli lactating females will have significantly lower fecal cortisol levels relative to adult males and non-lactating adult females during both lactation phases.

1.11. Dissertation overview

Chapters 2 through 4 are presented as independent manuscripts. Chapter 2 assesses P. coquereli maternal behavioral care-giving effort in their infants from birth until 26 weeks postnatal. Chapter 3 examines nutritional food quality and the biological significance of foods selected by lactating females. Chapter 4 investigates temporal variation in cortisol across different sex and reproductive classes during lactation. Chapter 5 draws conclusions and offers insights from the preceding four chapters and provides directions for future research.

Table 1.1. Gross milk composition in mammals Genus & species N DM CP Fat Sugar Ash (%) (%) (%) (%) (%)

Hooded seala 15 69.8 4.9 61.0 1.0 n/a (Cystophora cristata)

Domestic doga 25 22.7 7.5 9.5 3.8 1.1 (Canis familiaris)

Humanb n/a 12.4 0.9 3.8 7.0 0.02 (Homo sapiens)

Common marmosetc 46 13.9 2.7 3.5 7.4 n/a (Callithrix jacchus)

Ring-tailed lemurd 1 10.9 2.0 1.8 8.1 n/a (Lemur catta)

DM dry matter, CP crude protein aData from Oftedal and Iverson (1995) bData from Hambraeus (1984) cData from Power et al. (2002) dData from Tilden and Oftedal (1997)

Table 1.2. Gross milk composition in primates Genus & species N DM CP Fat Sugar GE C/P Female (%) (%) (%) (%) (kcal/g) body mass (kg) Black-and-white ruffed 5a 14.0a 4.2a 1.3a 7.7a 0.84a P 3.5c lemur (Varecia variegata) Common brown lemur 6a 9.6a 1.3a 0.9a 8.5a 0.49a C 2.3c (Eulemur fulvus) Sunda slow loris 4a 16.3a 3.9a 7.0a 6.6a 1.1a P 0.6c (Nycticebus coucang) Bolivian squirrel monkey 16b 16.6b 3.7b 5.0b 6.9b 0.91b C 0.7c (Saimiri boliviensis) DM dry matter, CP crude protein, GE gross energy, C/P carry or park infants aData from Tilden and Oftedal (1997) bData from Milligan et al. (2007) cData from Smith and Cheverud (2002)

Figure 1.1. Inherited environmental effects model: components of offspring phenotype

The components of offspring phenotype expressed in time t, deriving from the direct contribution of nuclear genes by one parent, a time-lagged presentation of the parental environment (Em), a time-lagged expression of parental performance genes and their interactions with the parental environment to produce the parental performance phenotype (Pm), plus the offspring’s own environment (Eo). For simplicity of presentation, G indicates additive genetic effects with dominance and epistasis assumed to be negligible. The numbered sources indicate possible routes of contribution to the offspring phenotype; Source 8 is any genetic covariance (cov) between genes expressed in two generations such as covGm Go or cov(GmEo)(G0Eo) (figure and description taken from Rossiter 1996)

1.12. References

Abbott D, Keverne E, Bercovitch F, Shively C, Mendoza S, Saltzman W, Snowdon C, Ziegler T, Banjevic M, and Garland Jr. T, Sapolsky, RM. 2003. Are subordinates always stressed? A comparative analysis of rank differences in cortisol levels among primates. Hormones and Behavior 43:67-82. Albignac R. 1981. Lemurine social and territorial organization in a north-western Malagasy forest (restricted area of Ampijoroa). In: Chiarelli A, and Corruccini R, editors. Primate Behavior and Sociobiology. New York: Springer-Verlag. p 25-29. Alexander R. 1993. The relative merits of foregut and hindgut fermentation. Journal of Zoology: Proceedings of the Zoological Society of London 231:391-401. Altmann J, and Alberts S. 2005. Growth rates in a wild primate population: ecological influences and maternal effects. Behavioral Ecology and Sociobiology 57:490-501. Bales K, Dietz J, Baker A, Miller K, and Tardif S. 2000. Effects of allocare-givers on fitness of infants and parents in callitrichid primates. Folia Primatologica 71:27-38. Bales K, French J, Hostetler C, and Dietz J. 2005. Social and reproductive factors affecting cortisol levels in wild female golden lion tamarins (Leontopithecus rosalia). American Journal of Primatology 67:25-35. Beehner J, and McCann C. 2008. Seasonal and altitudinal effects on glucocorticoid metabolites in a wild primate (Theropithecus gelada). Physiology & Behavior 95:508-514. Ben Shaul D. 1962. The composition of the milk of wild animals. Journal of Dairy Science 4:333-342. Bernardo J. 1996. Maternal effects in animal ecology. American Zoologist 36:83-105. Bonier F, Moore I, Martin P, and Robertson R. 2009. The relationship between fitness and baseline glucocorticoids in a passerine bird. General and Comparative Endocrinology 163:208-213. Boonstra R, Hik D, Singleton GR, and Tinnikov A. 1998. The impact of predator-induced stress on the snowshoe hare cycle. Ecological Monographs 68:371-394. Brockman D, Whitten P, Richard A, and Schneider A. 1998. Reproduction in free-ranging male Propithecus verreauxi: the hormonal correlates of mating and aggression. American Journal of Physical Anthropology 105:137-151. Brockman DK, Cobden AK, and Whitten PL. 2009. Birth season glucocorticoids are related to the presence of infants in sifaka (Propithecus verreauxi). Proceedings of the Royal Biological Society B 276:1855-1863. Burgess M, and Chapman C. 2005. Tree leaf chemical characters: selective pressures by folivorous primates and invertebrates. African Journal of Ecology 43:242-250. Busch D, and Hayward L. 2009. Stress in a conservation context: A discussion of glucocorticoid actions and how levels change with conservation-relevant variables. Biological Conservation 142:2844-2853. Cameron JL. 1996. Regulation of reproductive hormone secretion in primates by short-term changes in nutrition. Reviews of Reproduction 1:117-126. Campbell J, Williams C, and Eisemann J. 2004. Use of total dietary fiber across four lemur species (Propithecus verreauxi coquereli, Hapalemur griseus griseus, Varecia variegata, and Eulemur fulvus): does fiber type affect digestive efficiency? American Journal of Primatology 64:323-335. Casolini P, Cigliana G, Alemá G, Ruggieri V, Angelucci L, and Catalani A. 1997. Effect of increased maternal corticosterone during lactation on hippocampal corticosteroid receptors, stress response and learning in offspring in the early stages of life. Neuroscience 79:1005-1012. Champagne F, Francis D, Mar A, and Meaney M. 2003. Variations in maternal care in the rat as a mediating influence for the effects of environment on development. Physiology and Behavior 79:359-371 Chapman CA, Chapman L, Naughton-Treves L, Lawes MJ, and McDowell LR. 2004. Predicting a folivorous primate abundance: validation of a nutritional model. American Journal of Primatology 62:55-69. Chapman CA, Chapman LJ, Rode KD, Hauck EM, and McDowell LR. 2003. Variation in the nutritional value of primate foods: among trees, time periods, and areas. International Journal of Primatology 24:317-333. Cheverud J. 1984. Evolution by kin selection: a quantitative genetic model illustrated by maternal performance in mice. Evolution 38:766-777. Cheverud JM, and Wolf JB. 2009. The genetics and evolutionary consequences of maternal effects. In: Maestripieri D, and Mateo JM, editors. Maternal Effects in Mammals. Chicago: University of Chicago Press Clutton-Brock T. 1991. Parental care in birds and mammals. In: Krebs J, and Clutton-Brock T, editors. The Evolution of Parental Care. Princeton Princeton New Jersey Press. p 131-152. Creel S, Fox J, Hardy A, Sands J, Garrott R, and Peterson R. 2002. Snowmobile activity and glucocorticoid stress responses in wild wolves and elk. Conservation Biology 16:809-814. Crockett C, and Rudran R. 1987. Red howler monkey birth data II: interannual, habitat, and sex comparisons. American Journal of Primatology 13:369-384. 24 25

Dall S, and Boyd I. 2004. Evolution of mammals: lactation helps mothers to cope with unreliable food supplies. Proceedings of the Royal Society B 271:2049-2057. Dewar R, and Richard A. 2007. Evolution in the hypervariable environment of Madagascar. Proceedings of the National Academy of Sciences of the USA 104:13723-13727. Dobson F, and Michener G. 1995. Maternal traits and reproduction in richardson's ground squirrels. Ecology 76:851- 862. Dunham AE, Erhart EM, and Wright PC. 2011. Global climate cycles and cyclones: consequences for rainfall patterns and lemur reproduction in southeastern Madagacar. Global Change Biology 17:219-227. Eaglen R, and Boskoff K. 1978. The birth and early development of a captive sifaka, Propithecus verreauxi coquereli. Folia Primatologica 30:206-219. Eisen E, and Saxton A. 1983. Genotype by environment interactions and genetic correlations involving two environmental factors. Theoretical and Applied Genetics 67:75-86. Fairbanks L, and McGuire M. 1995. Maternal condition and the quality of maternal care in vervet monkeys. Behaviour 132:733-754. Fedigan L, Rose L, and Avila R. 1996. See how they grow: tracking capuchin monkey (Cebus capucinus) populations in a regenerating Costa Rican dry forest. In: Norconk M, editor. Adaptive Radiations of Neotropical Primates New York: Plenum. p 289-307. Fichtel C, Kraus C, Ganswindt A, and Heistermann M. 2007. Influence of reproductive season and rank on fecal glucocorticoid levels in free-ranging male Verreaux's sifaka (Propithecus verreauxi). Hormones and Behavior 51:640-648. Fite J, Patera K, French J, Rukstalis M, Hopkins E, and Ross C. 2005. Opportunistic mothers: female marmosets (Callithrix kuhlii) reduce their investment in offspring when they have to, and when they can. Journal of Human Evolution 49:122-142. Fleming S, Ruble D, Krieger H, and PY W. 1997. Hormonal and experimental correlates of maternal responsiveness during pregnancy and the puerperium in human mothers. Hormones and Behavior 31:145-158. Galloway L. 2005. Maternal effects provide phenotypic adaptation to local environmental conditions. New Phytologist 166:93-100. Ganzhorn J. 1995a. Low-level forest disturbance effects on primary production, leaf chemistry, and lemur populations. Ecology 76:2084-2096. Ganzhorn J. 2002. Distribution of a folivorous lemur in relation to seasonally varying food resources: integrating quantitative and qualitative aspects of food characteristics. Oecologia 131:427-435. Ganzhorn J, Lowry P, Schatz G, and Sommer S. 2001. The biodiversity of Madagascar: One of the world's hottest hotspots on its way out. Oryx 37:115-118. Genoud M. 2002. Comparative studies of basal metabolism in primates. Evolutionary Anthropology 11:108-111. Gerardo G, and Goodman SM. 2003. Hunting of protected animals in the Parc National d'Ankarafantsika, northwestern Madagascar. Oryx 37:115-118. Girard-Buttoz C, Heistermann M, Krummel S, and Engelhardt A. 2009. Seasonal and social influences on fecal androgen and glucocorticoid excretion in male long-tailed macaques (Macaca fascicularis). Physiology & Behavior 98:168-175. Gogarten J, Guzman M, Chapman C, Jacob A, Omeja P, and Rothman J. 2012. What is the predictive power of the colobine protein-to-fiber model and its conservation value? Tropical Conservation Science 5:381-393. Goldizen A. 1987. Tamarins and marmosets: communal care of offspring. In: Smuts B, Cheney D, Seyfarth R, and Wrangham R, editors. Primate Societies Chicago: University of Chicago Press. p 34-43. Goodman S, and Benstead J. 2005. Updated estimates of biotic diversity and endemism for Madagascar. Oryx 39:73-77. Gould L, Power ML, Ellwanger N, and Rambeloarivony H. 2011. Feeding behavior and nutrient intake in spiny forest- dwelling ring-tailed lemurs (Lemur catta) during early gestation and early to mid-lactation periods: compensating in a harsh environment. American Journal of Physical Anthropology 145:469-479. Gould L, Sussman RW, and Sauther ML. 1999. Natural disasters and primate populations: the effects of a 2-year drought on a naturally occurring population of ring-tailed lemurs (Lemur catta) in southwestern Madagascar. International Journal of Primatology 20:69-84. Gould L, Sussman RW, and Sauther ML. 2003. Demographic and life-history patterns in a population of ring-tailed lemurs (Lemur catta) at Beza Mahafaly Reserve, Madagascar: A 15-year perspective. American Journal of Physical Anthropology 120:182-194.

Gould L, Ziegler T, and Wittwer D. 2005. Effects of reproductive and social variables on fecal glucocorticoid levels in a sample of adult male ring-tailed lemurs (Lemur catta) at the Beza Mahafaly Reserve, Madagascar. American Journal of Primatology 67:5-23. Grieser B. 1992. Infant development and parental care in two species of sifakas. Primates 33:305-314. Harcourt A. 2008. Are lemurs' low basal metabolic rates an adaptation to Madagascar's unpredictable climate? Primates 49:292-294. Harper G, Steininger M, Tucker C, Juhn D, and Hawkins F. 2007. Fifty years of deforestation and forest fragmentation in Madagascar. Environmental Conservation 34:325-333. Heidinger B, Nisbet I, and Ketterson E. 2006. Older parents are less responsive to a stressor in a long-lived seabird: a mechanism for increased reproductive performance with age? Proceedings of the Royal Biological Society B 273:2227-2231. Hemingway C. 1998. Selectivity and variability in the diet of Milne-Edwards’ Sifakas (Propithecus diadema edwardsi): Implications for folivory and seed-Eating. American Journal of Primatology 19:355-357. Hinde K, and Milligan L. 2011. Primate milk: proximate mechanisms and ultimate perspectives. Evolutionary Anthropology 20:9-23. Hrdy S. 1976. Care and exploitation of nonhuman primate infants by conspecifics other than the mothers. In: Rosenblatt J, Hinde R, Shaw E, Beer C, editors. Advances in the Study of Behavior. New York: Academic, p 101-158. Inchausti P, and Ginzburg LR. 1998. Small mammals cycles in northern Europe: patterns and evidence for maternal effect hypothesis. Journal of Animal Ecology 67:180-194. Irwin M. 2008b. Feeding ecology of Propithecus diadema in forest fragments and continuous forest. International Journal of Primatology 29:95-115. Iverson S, and Oftedal O. 1995. Philogenetic and ecological variation in the fatty acid composition of milks. In: Jensen RG, editor. Handbook of Milk Composition. New York: Academic Press. Janson C, and van Schaik C. 2002. Ecological risk aversion in juvenile primates: slow and steady wins the race In: Pereira M, and Fairbanks L, editors. Juvenile Primates: Life history, Development, and Behavior. Chicago: The University of Chicago Press p57-74. Jessop N. 1997. Protein metabolism during lactation. Proceedings of the Nutrition Society 56:169-175. Jolly A. 1966. Propithecus verreauxi verreauxi. Lemur Behavior: A Madgascar Field Study. Chicago: University of Chicago Press. p 20-68. Jolly A. 1984. The puzzle of female feeding priority. In: Small M, editor. Female Primates: Studies by Women Primatologists. New York: Alan Liss, INC. Kappeler P. 1993. Sexual selection and lemur social systems. In: Kappeler P, and Ganzhorn J, editors. Lemur Social Systems and their Ecological Basis. New York: Plenum Press. p 223-240. Kappeler PM. 1991. Patterns of sexual dimorphism in body weight among prosimian primates. Folia Primatologica 57:132-146. Keay J, Jatinder Singh B, Gaunt M, and Kaur T. 2006. Fecal glucocorticoids and their metabolites as indicators of stress in various mammalian species: A literature review. Journal of Zoo and Wildlife Medicine 37:234- 244. Kirkpatrick M, and Lande R. 1989. The evolution of maternal characters. Evolution 43:485-503 Kubzdela K. 1997. Sociodemography in diurnal primates: The effects of group size and female dominance rank on intra-group spatial distribution, feeding competition, female reproductive success, and female dispersal patterns in white sifaka (Propithecus verreauxi verreauxi) [dissertation]. Chicago: University of Chicago. Kull C. 2000. Deforestation, erosion, and fire: degradation myths in the environmental history of Madagascar. Environment and History 6:1-28. Kun-Rodrigues C, Salmona J, Besolo A, Rasolondraibe E, Rabarivola C, Marques TA, and Chikhi L. 2014. New density estimates of a threatened sifaka species (Propithecus coquereli) in Ankarafantsika National Park. American Journal of Primatology 76:515-528. Lacey E. 1998. What is an adaptive environmentally induced parental effect? In: Mousseau T, and Fox C, editors. Maternal Effects As Adaptations. Oxford: Oxford University Press. p 54-66. Lambert J. 1998. Primate digestion: interactions among anatomy, physiology, and feeding ecology. Evolutionary Anthropology 7:8-20. Larson A, and Losos J. 1996. Phylogenetic Systematics of Adaptation San Diego: Academic Press. 507 p. Lee P, Majluf P, and Gordon I. 1991. Growth, weaning and maternal investment from a comparative perspective. Zoology 225:99-114. Lee PC, and Bowman JE. 1995. Influence of ecology and energetics on primate mothers and infants. Motherhood in Human and Nonhuman Primates. Basel: Karger. p 47-58.

Leigh S, and Terranova C. 1998. Comparative perspectives on bimaturism, ontogeny, and dimorphism in lemurid primates. International Journal of Primatology 19:723-749. Leutenegger W. 1973. Maternal-fetal weight relationships in primates. Folia Primatologica 20:280-293. Lewis R, and van Schaik C. 2007. Bimorphism in male Verreaux's sifaka in the Kirindy Forest of Madagascar. International Journal of Primatology 28:159-182. Lewis RJ, and Kappeler PM. 2005a. Are Kirindy sifaka capital or income breeders? It depends. American Journal of Primatology 67:365-369. Lewis RJ, and Kappeler PM. 2005b. Seasonality, body condition, and timing of reproduction in Propithecus verreauxi verreauxi in the Kirindy Forest. American Journal of Primatology 67:347-364. Maestripieri D. 2009. Maternal influences on offspring growth, reproduction, and behavior in primates. In: Maestripieri D, and Mateo J, editors. Maternal Effects in Mammals. Chicago: University Chicago Press. p 256-291. Maestripieri D, and Mateo J. 2009. The role of maternal effects in mammalian evolution and adaptation In: Maestripieri D, and Mateo J, editors. Maternal Effects in Mammals. Chicago: University of Chicago Press. p 1-10. Mass V, Heistermann M, and Kappeler PM. 2009. Mate-guarding as a male reproductive tactic in Propithecus verreauxi. International Journal of Primatology 30:389-409. Masters J, de Wit M, and Asher R. 2006. Reconciling the origins of Africa, India, and Madagascar with vertebrate dispersal scenarios. Folia Primatologica 77:99-418. Mateo J. 2009. Maternal influences on development, social relationships, and survival behaviors In: Maestripieri D, and Mateo J, editors. Maternal Effects in Mammals. Chicago: Chicago University Press. p 133-158. Mattson W. 1980. Herbivory in relation to plant nitrogen content. Annual Review of Ecology and Systematics 11:119-161. Maynard L, and Loosli J. 1965. Bioenergetics. In: Maynard L, and Loosli J, editors. Animal Nutrition New York: McGraw-Hill. p 29-47. Mayor M, Sommer J, Houck M, Zaonarivelo J, Wright P, Ingram C, Engel S, and Louis EJ. 2004. Specific Status of Propithecus spp. International Journal of Primatology 25:875-900. McGinley M, Temme D, and Geber M. 1987. Parental investment in offspring in variable environments: theoretical and empirical considerations. American Naturalist 130:370-398. McGoogan K. 2011. Edge effects on the behaviour and ecology of Propithecus coquereli in Northwest Madagascar [dissertation]. Available from TSpace Repository, 1807/31861: University of Toronto 225 p. McLaren A. 1981. Analysis of maternal effects on development in mammals. Journal of Reproduction and Fertility 62:591-596. Milton K. 1979. Factors influencing leaf choice by howler monkeys: a test of some hypotheses of food selection by generalist herbivores. American Naturalist 114:362-378. Mittermeier R, Louis Jr., E, Richardson, M, Schwitzer, C, Langrand, O, Rylands, A, Hawkins, F, Rajaobelina, S, Ratsimbazafy, J, Rasoloarison, R, Roos, C, Kappeler, P, Mackinnon, J. 2010. Indriidae. Lemurs of Madagascar. Third ed. Bogotá, Colombia: Conservation International. p 481-597. Moberg G. 2000. Biological response to stress: implications for animal welfare. In: Moberg G, and Mench J, editors. The Biology of Animal Stress: Basic Principles and Implications for Animal Welfare Cambridge, MA: CABI Publishing p1-21. Mousseau TA, and Fox CW. 1998. The adaptive significance of maternal effects. TREE 13(10):403-407. Norscia I, Carrai V, and Borgognini-Tarli SM. 2006. Influence of dry season and food quality and quantity on behavior and feeding strategy of Propithecus verreauxi in Kirindy, Madagascar. International Journal of Primatology 27:1001-1022. Oftedal O. 1984. Milk composition, milk yield and energy output at peak lactation: a comparative review. Symposium Zoological Society of London 51:33-85. Oftedal O. 1991. The nutritional consequences of foraging in primates: the relationship of nutrient intakes to nutrient requirements Philosophical Transactions of the Royal Society B 334:161-170. Oftedal O. 1993. The adapation of milk secretion to the constraints of fasting in bears, seals, and baleen whales. Journal of Dairy Science 76:3234-3246. Oftedal O, and Iverson S. 1995. Comparative analysis of nonhuman milks: Phylogenetic variation in the gross composition of milks. In: Jensen RG, editor. Handbook of Milk Composition. New York: Academic Press. Oster G, and Alberch P. 1982. Evolution and bifurcation of developmental programs. Evolution 36:444-459. Ostner J, Kappeler P, and Heistermann M. 2008. Androgen and glucocorticoid levels reflect seasonally occurring social challenges in male redfronted lemus (Eulemur fulvus rufus). Behavioral Ecology and Sociobiology 62:627-638.

Pastorini J, Forstner M, and Martin R. 2001. Phylogenetic history of sifakas (Propithecus: Lemuriformes) derived from mtDNA sequences. American Journal of Primatology 53:1-17. Perelman P, Johnson W, Roos C, Seuànez H, Horvath J, Moreira M, Kessing B, Pontius J, Roelke M, Rumpler Y et al. 2011. A molecular phylogeny of living primates. PLoS Genetics 7:1-17. Petter-Rousseaux A. 1962. Studies on the biology of reproduction in lower primates. Mammalia 26:1-88. Power M, Oftedal O, and Tardif S. 2002. Does the milk of callitrichid monkeys differ from that of larger anthropoids? American Journal of Primatology 56:117-127. Price E. 1991. Competition to carry infants in captive families of cotton-top tamarins (Saguinus oedipus). Behaviour 118:66-88. Pryce CR. 1995. Determinants of motherhood in human and nonhuman primates. Motherhood in Human and Nonhuman Primates. Basel: Karger. p 1-15. Rangel-Negrín A, Alfaro JL, Valdez RA, Romano MC, and Serio-Silva JC. 2009. Stress in Yucatan spider monkeys: effects of environmental conditions on fecal cortisol levels in wild and captive populations. Animal Conservation 12:496-502. Ravosa M, Meyers D, and Glander K. 1993. Relative growth of the limbs and trunk in sifakas: heterochronic, ecological, and functional considerations. American Journal of Physical Anthropology 92:499-520. Reinhold K. 2002. Maternal effects and the evolution of behavioural and morphological characters: a literature review indicates importance of extended maternal care. Journal of Heredity 93:400-405. Richard A. 1974. Intra-specific variation in the social organization and ecology of Propithecus verreauxi. Folia Primatologica 22:178-207. Richard A. 1976. Preliminary observations on the birth and development of Propithecus verreauxi to the age of six months. Primates 17:357-366. Richard A. 1978. Behavioral variation: case study of a Malagasy prosimian: Lewisburg Bucknell University Press. Richard A. 1985. Social boundaries in a Malagasy prosimian, the sifaka (Propithecus verreauxi). International Journal of Primatology 6:553-568. Richard A. 1987. Malagasy prosimians: female dominance. In: Smut B, Cheney D, Seyfarth R, Wrangham R, and Struhsaker T, editors. Primate Societies. Chicago: University of Chicago Press. p 25-33. Richard A. 1991. Aggressive competition between males, female-controlled polygyny and sexual monomorphism in a Malagasy primate, Propithecus verreauxi. Journal of Human Evolution 22:395-406. Richard A, Dewar R, Schwartz M, and Ratsirarson J. 2000. Mass change, environmental variability and female fertility in wild Propithecus verreauxi. Journal of Human Evolution 39:381-391. Richard A, Dewar R, Schwartz M, and Ratsirarson J. 2002. Life in the slow lane? Demography and life histories of male and female sifaka (Propithecus verreauxi verreauxi). The Zoological Society of London 256:421-436. Richard A, and Nicoll M. 1987. Female social dominance and basal metabolism in a Malagasy primate, Propithecus verreauxi. American Journal of Physical Anthropology 12:309-314. Roosenburg WM. 1996. Maternal condition and nest site choice: an alternative for the maintenance of environmental sex determination? American Zoologist 36:157-168. Ross A, Porter L, Power M, and Sodaro V. 2010. Maternal care and infant development in Callimico goeldii and Callithrix jacchus. Primates 51:315-325. Ross C, MacLarnon A. 2000. The evolution of non-maternal care in anthropoid primates: a test of the hypotheses. Folia Primatologica 71:93-113. Rossiter M. 1994. Maternal effects hypothesis of herbivore outbreak. Bioscience 44:752-763. Rossiter M. 1996. Incidence and consequences of inherited environmental effects. Annual Review of Ecology and Systematics 27:451-476. Rossiter M. 1998. The role of environmental variation in parental effects expression. In: Mousseau T, and Fox C, editors. Maternal Effects As Adaptations. Oxford: Oxford University Press. p 112-134. Saltzman W, and Abbott D. 2009. Effects of elevated circulating cortisol concentrations on maternal behavior in common marmoset monkeys (Callithrix jacchus). Psychoneuroendocrinology 34:1222-1234. Sauther M. 1994. Wild plant use by pregnant and lactating ringtailed lemurs, with implications for early hominid foraging. Eating on the Wild Side Tucson: University of Arizona Press. p 240-256. Sauther M, and Cuozzo F. 2009. The impact of fallback foods on wild ring-tailed lemur biology: a comparison of intact and anthropogenically disturbed habitats. American Journal of Physical Anthropology 140:671-686. Schoener T. 1971. Theory of feeding strategies. Annual Review of Ecology and Systematics 2:369-404. Sellen D. 2009. Evolution of human lactation and complementary feeding: implications for understanding contemporary cross-cultural variation In: Goldberg G, Prentice A, Prentice A, Filteau S, and Simondon K, editors. Breast-Feeding: Early Influences on Later Health. Dordrecht: Springer Science.

Setchell J, Smith T, Wickings E, and Knapp L. 2008. Factors affecting fecal glucocorticoid levels in semi-free- ranging female mandrills (Mandrillus sphinx). American Journal of Primatology 70:1023-1032. Smith A. 1997. Deforestation, fragmentation, and reserve design in western Madagascar. In: Lawrence W, and Bierregaard O, editors. Tropical Forest Remnants, Ecology, Management, and Conservation of Fragmented Communities Chicago: University of Chicago Press. p 415-441. Solomon NG, JA French. 1997. The study of mammalian cooperative breeding In: Solomon NG, French JA, editors. Cooperative Breeding in Mammals. Cambridge: Cambridge University Press p 1-10. Stearns SC. 1976. Life-history tactics: a review of the ideas. The Quarterly Review of Biology 51:3-47. Stearns SC. 1977. The evolution of life history traits: a critique of the theory and a review of the data. Annual Review of Ecology and Systematics 8:145-171. Stearns SC. 1992. Evolutionary Explanations. The Evolution of Life Histories. New York: Oxford University Press. Stephens D, and Krebs J. 1986. Foraging Theory. Princeton, NJ: Princeton University Press. 247 p. Tardif S. 1994. Relative energetic cost of infant care in small-bodied neotropical primates and its relation to infant- care patterns. American Journal of Primatology 34:133-143. Tardif S, and Bales K. 2004. Relations among birth condition, maternal condition, and postnatal growth in captive common marmoset monkeys (Callithrix jacchus). American Journal of Primatology 62:83-94. Tardif S, Harrison M, and Simek M. 1993. Communal care in marmosets and tamarins: relation to energetics, ecology, and social organization. In: Rylands A, editor. Marmosets and Tamarins: Systematics, Behaviour, and Ecology. Oxford: Oxford University Press. Tardif S, Power M, Oftedal O, Power R, and Layne D. 2001. Lactation, maternal behavior and infant growth in common marmoset monkeys (Callithrix jacchus): effects of maternal size and litter size. Behavioral Ecology and Sociobiology 51:17-25. Tattersall I. 1982. The Primates of Madagascar. New York: Colombia University Press. Tecot SR, Baden AL, Romine NK, and Kamilar JM. 2012. Infant parking and nesting, not allomaternal care, influence Malagasy primate life histories. Behavioral Ecology and Sociobiology 10:1375-1386. Tecot S, Baden A, Romine N, and Kamilar J. 2013. Reproductive strategies and infant care in the Malagasy primates. In: Clancy KB, Hinde K, and Rutherford J, editors. Building Babies. New York: Springer Science+Business Media. Tilden C, and Oftedal O. 1997. Milk composition reflects patterns of maternal care in prosimian primates. American Journal of Primatology 41:195-211. Trivers RL. 1974. Parent-offspring conflict. American Zoologist 14:249-264. van Schaik C, and van Noordwijk MA. 1985. Interannual variability in fruit abundance and the reproductive seasonality in Sumatran long-tailed macaques (Macaca fascicularis). Journal of Zoology, London 206:533- 549. Vasey N. 2002. Niche separation in Varecia variegata rubra and Eulemur fulvus albifrons: II. Intraspecific patterns. American Journal of Physical Anthropology 118:169-183. West-Eberhard M. 1989. Phenotypic plasticity and the origins of diversity. Annual Review of Ecology and Systematics 20:249-278. Willham R. 1972. The role of maternal effects in animal breeding: III. biometrical aspects of maternal effects in mammals. Journal of Animal Science 35:1288-1293. Wright P. 1999. Lemur traits and Madagascar ecology: coping with an island environment. Yearbook of Physical Anthropology 42:31-72. Young A, Richard A, and Aiello L. 1990. Female dominance and maternal investment in strepsirhine primates. American Naturalist 135:473-488. Ziegler T. 2000. Hormones associated with non-maternal infant care: a review of mammalian and avian studies. Folia Primatologica 71:6-21.

Chapter 2 Maternal Effort in Propithecus coquereli in Ankarafantsika National Park, Northwestern Madagascar

“Behave like the chameleon: look forward and observe behind.” “Mitondra tena tahaka ny tanalahy: mijery eo aloha sy mandinika ny aoriana.” ~ Malagasy proverb

2.1. Abstract

Lactation is the most energetically expensive activity in which mammals participate. Lactating females face energetic constraints absent in non-lactating conspecifics and in turn, must compensate for substantially higher energy requirements during the lactation process. Extended durations of infant transport are widespread throughout the taxonomic order Primates while rare considering the class Mammalia. Infant transport is also an energetically costly mammalian activity seen predominately in primates, yet few primatologists examined this behavior in detail. I evaluated Coquerel’s sifaka (Propithecus coquereli) maternal behavioral care-giving effort in Ankarafantsika National Park, northwestern Madagascar from 1-26 weeks postnatal (n=10 mothers with infants) and compared it to non-mothers (n=19 adult males, n=8 adult females without infants). Maternal behavioral care-giving and the quality of the environment mothers provide directly influences infant development and future reproductive success. The volatile climate and extreme seasonality characteristic of Madagascar coupled with the associated costs of lactation and infant transport gives rise to a large energetic challenge faced by P. coquereli mothers. I measured infant transport position (ventral, dorsal, independent), the duration of infant transport by carriers (mothers, adult males, adult females), and the frequency of infant bodily contact between mothers and their infants for 678 focal hours of observation over two consecutive birth seasons (2010 and 2011). Infant transport position was used as a proxy for development. Infants spent significantly more time in the ventral transport position than either dorsally or independently. Mothers were the primary infant transporters. Adult males and females that were not mothers both participated in infant transport, but for significantly less time than mothers. Infants initiated and broke bodily contact with mothers more frequently than mothers initiated and broke contact with their infants. P. coquereli infants continued to be transported 26% of the time by the 26th postnatal week. This clearly

30 31 demonstrates that P. coquereli infants are more altricial relative to other lemurs of comparable body size and dependent on their mothers for longer durations than suggested by previous studies. P. coquereli mothers were the primary infant behavioral care-givers and must employ alternative behavioral, physiological, or nutritional strategies to offset the associated energetic costs of lactating while transporting infants in a harsh environment.

2.2. Introduction

Mammalian parental care is long in duration and infants are more altricial than in other animal classes (Clutton-Brock 1991). Nonetheless, long-term infant transport is uncommon in most mammals while frequent in primates (Ross 2001a). Lactation is the single most energetically costly activity for mammals (see Oftedal 1984), followed by infant transport, though lactation is considerably more costly than transport (Altmann and Samuels 1992; Dewey 1997; Nievergelt and Martin 1999; Tardif et al. 1993). Lactation increases daily energy expenditures up to 150% in mammals, with mean caloric intake during lactation increasing 66-188% (reviewed in Gittleman and Thomspon 1988; Lee et al. 1991). This striking rise forces lactating females to compensate for substantially higher energy requirements by utilizing stored energy (e.g., fat reserves), increasing energy consumption (e.g., food), reducing energy invested in physiological states (e.g., basal metabolic rate), or reducing energy in behavioral activities (e.g., infant transport).

The maternal effects model proposes that the care mothers provide their offspring determines how offspring respond to their physical environment (e.g., Bernardo 1996). Maternal care includes all prezygotic, prenatal, and postnatal investment in offspring that is either expressed through a genetic, physiological, behavioral, or environmental condition (Rossiter 1996; Wade 1998). Mammalian research on maternal effects has primarily focused on ungulates (e.g., Cameron and Linklater 2000) and within the non-human primates, the cercopithecines (e.g., Altmann 1988). Lactation is divided into two phases when infants: (1) are exclusively dependent on milk nutrients and (2) rely on a combination of milk nutrients and solid food (Langer 2008). Maternal effects both directly and indirectly effect offspring phenotype, the observable traits of an organism, with a combination of parental and environmental sources selecting on offspring phenotype. These parental and environmental sources are either genotypic or phenotypic and are collectively referred to as the inherited environmental effects model (Eisen and Saxton 1983;

Rossiter 1996). Thus, maternal effects are a component of the inherited environmental effects model. Nongenetic maternal effects influence individual adaptations to environmental conditions and contribute to evolutionary changes in resident populations (reviewed in Maestripieri and Mateo 2009).

Environmental quality is a source included within the inherited environmental effects model that plays a significant role in the structure, quality and duration of maternal-infant relationships (Rossiter 1996). The dynamic relationship between the expression of parental effects and environmental quality exists irrespective of the additional variables contributing to offspring phenotypic expression, is species dependent, and seasonally variable (Rossiter 1996; Rossiter 1998). For instance, the quality and duration of parental care invested in human offspring is influenced by environmental risk factors, such as famine and warfare (Quinlan 2007; Quinlan et al. 2003). When these human risk factors increase, maternal care investment decreases (Quinlan 2007). Human mothers with low socioeconomic status, and thus more restricted access to resources, have been found to wean high-risk infants (defined as infants with lower birth weights in poorer physical condition) more rapidly than healthy infants to maximize the potential reproductive value of future healthy infants (Bereczkei 2001). This life history trade-off supports that mothers do not invest care in high-risk infants because they offer little inclusive fitness in comparison to healthy infants that have a greater chance of survival and eventually producing their own offspring. These studies demonstrate that human parental care patterns are directly affected by environmental quality and contingent on resource stability.

The quality of maternal care infant receive may partly reflect infant sex. The Trivers-Willard (T- W) model is a classic life history model which predicts maternal health is the primary determinant of sex ratio variation, and that individual females will produce biased birth sex ratios (BSRs) favoring the sex benefiting most from increased parental care (Trivers and Willard 1973). The T-W model has not always been supported when applied to primate taxa (reviewed in Richard et al. 1991), thereby demonstrating that the evolutionary factors selecting biased BSRs cannot be solely attributed to maternal health and care-giving, but must expanded to consider additional beneficial elements. The local resource competition hypothesis (LRC) is a competing hypothesis to the T-W model proposing that BSRs will favor the dispersing sex due to intrafemale competition in female philopatric groups, which is most palpable when resource

33 competition is high and unstable (Hamilton 1967; Silk 1984). Another competing model to the T- W model is the local resource enhancement (LRE) model, which suggests that BSRs in cooperatively breeding species will favor the more effective helper sex to maximize infant care- giving (Pen and Weissing 2000).

Resource instability in mammals besides humans has been shown to negatively influence maternal care. For example, variability in Long-Evans rat (Rattus norvegicus) litter size, sex, and weaning weight does not alter individual maternal care behavior across multiple litters when resource access is constant and only changes when resources are unstable (Champagne et al. 2003). Bonnet macaque (Macaca radiata) mother-infant relationships are more negatively influenced by resource unpredictability than a constant state of decreased food availability (Rosenblum and Andrews 1994). In the same study, captive infants that experienced a combination of high (food was sparse and patchily distributed) and low (food was abundant and readily distributed) foraging demands, had fewer positive interactions with their mothers and displayed greater emotional distress than infants raised exclusively in the lower quality, higher foraging demand environment where food was sparse and patchily distributed (Rosenblum and Andrews 1994). Thus, resource instability affects gestational development and is also a driving factor in the quality of postnatal care mothers provide.

In contrast to the T-W model, the LRC and LRE models focus on broader scale genetic and ecological processes rather than exclusively on maternal body condition during conception to determine the causes of preferred BSRs (Silk and Brown 2008). A review of BSRs in 102 primate species found BSRs are not related to the magnitude of sexual dimorphism present in dimorphic species, and thus BSRs do not reflect differences in the costs of behavioral care-giving between male and female offspring since BSRs would be expected to be skewed in favor of the less costly sex (most typically females due to their smaller body size) (Silk and Brown 2008). The same study found that female dispersing species had female-biased BSRs (Silk and Brown 2008). Cooperatively breeding species had male-biased BSRs, which followed predictions, given that males are typically more invested in behavioral care-giving than females in cooperatively breeding species (Silk and Brown 2008). These findings support predictions from the LRC and LRE models and demonstrate that examining BSRs on a population level in comparison to individualized responses recommended by the T-W model will assist in isolating

34 the causes of preferred BSRs and encourage interspecific comparisons.

Mammalian mothers are typically the primary infant care-givers since infants rely on mothers for milk, but allomaternal care is present within mammals, and is more common, albeit highly variable, within primates than earlier understood (see Bales et al. 2000; Chism 2000; Fite et al. 2005; Goldizen 1987; Gould 1992; Morland 1990; Roberts et al. 2001; Ross and MacLarnon 2000; Schradin and Anzenberger 2001a; Schradin and Anzenberger 2001b; Zahed et al. 2008). Allomaternal care is a collection of care behaviors including infant transport, infant guarding, babysitting, and energy transfer by non-mothers (reviewed in Tecot et al. 2013) used, in part, to reduce the energetic burden on lactating females (see Ross and MacLarnon 2000 Table 2.1 for a list of proximate, adaptive, and nonadaptive determinants of allocare). Infant carrying is the second most energetically costly activity within this list (Altmann and Samuels 1992) and is associated with elevated nutritional and predation risks (Schradin and Anzenberger 2001a). Lemur allomaternal studies are alarmingly few (reviewed in Tecot et al. 2013, but see Baden 2011), due partly to lower documented incidences within lemurs, and fewer lemur postnatal behavioral studies in comparison to callitrichids and cercopithecines. There is no relationship between allomaternal care and offspring sex in golden lion tamarins (Leontopithecus rosalia), but there was a significant relationship between the number of allocare-care givers and the number of surviving infants (Bales et al. 2002). Callitrichid infants grow rapidly and solid food provisioning by group members is critical to infant survivability (Bales et al. 2002). Verreaux’s sifaka (Propithecus verreauxi) tertiary (exclusively reproductive individuals) sex ratios are nearly equal, in contrast to many haplorhines that have female-biased tertiary sex ratios (Richard et al. 1991). Interestingly, the P. verreauxi secondary (at birth) sex ratio is skewed, which suggests a disparity in sex survivability that results from higher mortality rates in reproductive females (Richard et al. 1991). Longitudinal data in the wild on carrier identity and duration of P. coquereli infant transport have not been available until my dissertation.

Lemur reproductive events occur under fixed seasonal parameters as a response to restricted food availability and quality (see Tecot 2010; Wright 1999). Madagascar is characterized by erratic rainfall patterns that have caused the evolution of unique tree phenology and unpredictability in food abundance and distribution (Dewar and Richard 2007), including the Mahajanga province where data for my dissertation were collected. Thus, fruiting is restricted to a brief number of

35 calendar months (reviewed in Dewar and Richard 2007), thereby requiring P. coquereli to rely on alternative, less energy dense resources (e.g., leaves) during the majority of the year (McGoogan 2011). P. verreauxi males and females undergo seasonal changes in body mass, weighing the least during the middle to end of the dry season, with females experiencing greater changes in body mass than males (Richard et al. 2000). P. verreauxi females with higher body mass during the mating season had a greater number of births the subsequent season opposed to females that weighed significantly less and gave birth to fewer infants (Richard et al. 2000).

P. verreauxi reproductive females experience punctuated intervals of high mortality because of the high energetic costs of reproduction together with extreme seasonality on Madagascar (Richard et al. 1991). Cyclones and rainfall levels have adverse effects on Milne-Edwards’ sifaka (Propithecus edwardsi) reproductive rates (Richard et al. 2002). The number of P. edwardsi infants per female per year surviving to one postnatal year is negatively related to cyclone presence during gestation (Dunham et al. 2011). The number of drought months infants experience during the first postnatal year are also negatively associated with survival, demonstrating that lemur reproductive success is contingent on climatic events (Dunham et al. 2011). Maternal dental senescence and rainfall influences P. edwardsi infant survival, with older mothers requiring more daily precipitation to sustain dependent infants during lactation (King et al. 2005). Lactation also increases demands to water balance (Gittleman and Thomspon 1988), a cost that is especially pronounced in species living in xeric habitats (Soholt 1977) such as P. coquereli.

Propithecus spp. have unsealed vulvas in contrast to other Malagasy strepsirhines and exhibit concealed ovulation, a character present in many primate genera (Brockman et al. 1998). Females undergo single 0.50 to 0.96-hour estrus periods during the three-month breeding season (Brockman and Whitten 1996). Due to the short duration of estrus, Many female strepsirhines display estrous asynchrony within the seasonal synchrony of estrus, which increases mate choice by reducing female-female competition for mates, although males also successfully mate guard females and harass mating pairs irrespective of social status (Sauther 1991). Propithecus spp. polyandrous mating may be a strategy to confuse paternity, thereby decreasing the risk of infanticide (Wright 1995; but see Erhart and Overdorff 1999). However, monoandrous and polyandrous matings both result in conception and thus the function of polyandry is not a

36 strategy exclusively used to decrease infanticide (Brockman and Whitten 1996). A 10 to 15-day elevation in estradiol distinguish behavioral estrus and increased progesterone levels determine conception one to three days’ post-estrus, with estradiol levels characteristic of the gestational phrase present between 42 to 45 days after conception (Brockman and Whitten 1996).

Maternal behavioral care-giving effort is a nongenetic maternal effect that I define as the frequency and duration of infant transport by P. coquereli mothers in comparison to adult males and non-mother adult females. My dissertation tests the hypothesis that P. coquereli mothers experience greater postnatal reproductive stress, which is measured by behavioral care-giving effort, nutritional food quality, and cortisol stress responses during early/earlier-mid lactation (designated as 1-12 weeks postnatal) (May-August) in comparison to later-mid/late lactation (designated as 13-26 weeks postnatal) (September-December) due to the increased energetic costs of lactation while simultaneously caring for dependent infants during the driest seasonal months in the austral winter when food quality is low. This chapter examines P. coquereli maternal behavioral care-giving effort by addressing four predictions under this hypothesis. P1. Maternal behavioral care-giving effort will decrease as P. coquereli infants increase in age as measured by infant transport position (ventral, dorsal, independent). Infants will spend the greatest duration of time in the ventral position, followed by the dorsal position, and the least duration independent from carriers from birth until weaning (twenty-six weeks postnatal). P2. P. coquereli mothers will provide the majority of infant behavioral care-giving effort relative to adult males and non-mother adult females as measured by the frequency and duration of infant carriage from birth until weaning. P3. Infant contact initiated by P. coquereli mothers will decrease while infant contact broken by mothers will increase as infants age. P4. Contact with mothers initiated by P. coquereli infants will decrease while contact with mothers broken by infants will increase as infants age.

2.3. Methods

2.3.1. Study site and species I conducted this study from Ampijoroa Forestry Station in Ankarafantsika National Park (ANP) located in the Mahajanga Province in northwestern Madagascar (Figure 2.1). The

Ankarafantsika region (135,000 ha) was first established as two protected areas in 1927 and recognized as a national park in 2003. Ampijoroa is situated in the southwestern portion of ANP. The Marovoay area (38,000 ha) found within the Ankarafantsika region is the second largest producer of rice, the primary subsistence crop, in Madagascar (Alonso and Hannah 2002). The GPS coordinates of the research base camp are 16°18’31” South, 46°48’49” East and is 88m above sea level. ANP is characterized as a dry deciduous forest with an exceptionally pronounced dry season (Alonso and Hannah 2002; Du Puy and Moat 1996). Forested areas surrounding Ampijoroa are experiencing anthropogenic disturbance from slash and burn agriculture, fire, relatively high volumes of human traffic, unregulated presence and herding of domestic cattle, bushmeat hunting, and hole digging for Dioscorea maciba tuber extraction (Alonso and Hannah 2002; Crowley et al. 2012; Gerardo and Goodman 2003). The underlying geological formation in the Ankarafantsika region is composed of sandstones and the study area sits atop a sandstone plateau between 310-340 m above sea level (Du Puy and Moat 1996; Lourenço and Goodman 2006). Soils are either red, speckled or white, with red soil containing the highest water content and white sand the lowest (Crowley et al. 2012). Red soil most often occurs in the forest edge and in the savannah itself, though is also present in the forest interior and is presumably more nutrient dense due to its higher water content than forest interior quartz white sands (Crowley et al. 2012). Many trees grow in nutrient poor, acidic white sands and a thick layer of loose sand is present on the soil surface as a result of sandstone erosion (Du Puy and Moat 1996; Lourenço and Goodman 2006). Flora are speciose and the forest understory is moderately thick with sparse leaf litter (Lourenço and Goodman 2006) Annual precipitation in ANP ranges from 1100-1600 mm, with the majority of rainfall occurring in January and February and a period of extreme desiccation from May to September in which there is very little rainfall (Rendigs et al. 2003). Average daily temperatures range from 16°C during the dry season to 37°C in the wet season (Rendigs et al. 2003). Vertebrates have varied adaptive responses to cope with the austral winter including torpor or hibernation and flora are resilient to desiccation (Lourenço and Goodman 2006; Rendigs et al. 2003).

38 brown mouse lemur (Microcebus ravelobensis) (Alonso and Hannah 2002). The International Union for Conservation of Nature estimates that the P. coquereli has experience rapid population declines of more than 50% during the last 30 years primarily due to habitat loss (Andrainarivo 2008, accessed 17 November 2013) and more recently, hunting for human consumption (Gerardo and Goodman 2003). The slow life histories exemplified by Propithecus spp. increases extinction risk (Purvis et al. 2000) and amplifies the devastating impact of hunting given Propithecus spp. do not reproduce until later in life, have interbirth intervals of every other year, and only one infant is present per social group (Pochron et al. 2004; Richard et al. 1991). An estimated population of ~47,000 P. coquereli currently live in ANP (Kun-Rodrigues et al. 2014). Density estimates range from 5-100 individuals/km2, with habitat quality (i.e., negative effects of roads and forest edges) as the principal factor for this high variability (Kun-Rodrigues et al. 2014). An estimated 5 individuals/km2 presently exist in Ampijoroa (Kun-Rodrigues et al. 2014), in comparison to an estimated 60-75 individuals/km2 in the 1980s (Albignac 1981). This is a rapid decrease of more than 90% of the P. coquereli population in Ampijoroa (Kun- Rodrigues et al. 2014).

This chapter examines the duration and frequency of P. coquereli maternal behavioral effort from infant birth (referred to as week 1 postnatal) until 26 weeks postnatal. Infant transport position, infant carrier identity, and infant bodily contact between mothers and their infants were used as the measures of maternal behavioral effort and compared with non-mothers.

2.3.2. Data collection Data were collected for a total of 14 months between June-December 2010 and 2011. A previously established research trail system in Jardin Botanique A (JBA)(see Rendigs et al. 2003 for a detailed site description) and the tourist trails identified as the “Coquereli circuit” were used to initially locate P. coquereli groups. I, along with 1-2 research assistants, collected data on ten habituated P. coquereli groups. The principal investigator trained assistants in focal animal sampling techniques and inter-observer reliability diagnostic tests were performed with assistants once per month for the duration of this study. Inter-observer reliability was performed a total of five times approximately every three months and was calculated using a Kappa coefficient as the measure of agreement between observers following Lehner (1996), where Po = observed proportion of agreements and Pc = chance proportion of agreements:

Kappa = (Po - Pc)

(1 - Pc)

The Kappa statistic was applied separately to infant position (85% agreement between observers), infant carrier (80% agreement between observers), and infant bodily contact between mothers and infants (87% agreement between observers).

Seven P. coquereli groups were studied in season 1 and three groups were studied in season 2 for 26 consecutive weeks. Attempts to locate additional groups with infants in season 2 were unsuccessful and likely due to Propithecus spp. interbirth interval length of every other year. An 8th group was located in season 1, but the infant disappeared during the 3rd observation week and was presumed dead. Thus, the infant mortality rate during this study was only 9%. Early/earlier- mid lactation was designated from May to August and later-mid/late lactation from September to December. P. coquereli groups were considered habituated when no alarm calling occurred after researcher presence was detected. Phenotypic variations (e.g., tail length and shape, facial features) were used to differentiate between individuals. Group composition remained constant in all groups throughout both seasons (Tables 2.1, 2.2). A total of 47 individuals composed the ten groups (mothers n=10, infants n=10, adult males n=19, adult female, non-mothers n=8). The mean group size including infants for seasons 1 and 2 were 4.9 ± 0.9 and 4.3 ± 0.6, respectively for a mean group size of 4.6 ± 0.8 individuals.

2.3.3. Sampling methods Each P. coquereli group was followed for one day per week using ten minute continuous focal sampling (Altmann 1974) to collect infant carrier and transport position data for 6 hours beginning at dawn after groups were located in their sleeping trees. All-occurrence sampling (Altmann 1974) was used to collect data on bodily contact between mothers and infants. The number of occurrences where mothers initiated and broke bodily contact with infants and the numbers of occurrences where infants initiated and broke bodily contact with mothers (adopted from Bardi et al. 2003; Hinde and Atkinson 1970) were recorded simultaneously with infant carrier and position (see Tables 2.3, 2.4 for focal data columns and ethogram). The focal animal was alternated weekly between mothers and infants for a total of 678 focal hours (see Table 2.5

40 for focal hours by group and individual). Adult females were established as mothers when infants were first seen attached on the nipple prior to the first focal follow by the principle investigator. Nursing behavior was not recorded because suckling duration cannot be distinguished from time spent on the mother (Cameron 1998; Cameron et al. 1999).

Only P. coquereli groups with infants that were a maximum of four weeks postnatal were followed and only one infant was present in all groups (see Table 2.6 for infant birth months). The same group, Zaza, was followed during seasons 1 and 2 as the season 1 infant was absent from the group and presumed dead at the beginning of season 2 when a newborn infant was present. Groups were first located when no infants were present. These same groups were rechecked weekly until an infant was present (n=9). For example, Group Rambo was first located with no infant present on 05/28/2011, subsequently located on 06/06/2011 with no infant present, and relocated on 06/11/11 with an infant present. Therefore, the infant was a maximum of 5 days old. The number of days that elapsed between the initial and subsequent group locations determined maximum infant age. Age was estimated based on speaking with local guides and comparing relative body size to the other study infants (n=9) by taking photographs of each infant in the instance where the infant was present when the group was first located (n=1).

2.3.4. Data analysis Box and whisker plots illustrating infant transport position and carrier identities were constructed by converting the total weekly focal times (hours: minutes: seconds) for each group and converting it to a percentage. For example, if an infant spent 3:00:00 (the total continuous focal time) ventrally, this was converted by the following formula:

(=hour [3:00:00] +minute [3:00:00]/60+second [3:00:00]/3600).

This value was then converted by the following formula: (=3.00/3*100) to yield a percentage, in this case 100%. Box and whisker plots were created from the average percentage of time for each postnatal week, weekly minimum, first quartile, second quartile (median), third quartile, and weekly maximum. These values were used to create boxes from the 25th, 50th, and 75th percentiles. Whiskers were created from subtracting weekly minimums from the first quartile and the third quartile from maximums. Data were not available on infant transport by adult

41 females in week 26 as no adult females were present in observed groups.

Mixed effects linear regression models in SAS® Version 9.2 were applied to determine the relationship between the independent variable, infant age (1-26 weeks postnatal) and the dependent variable, carrier position (ventral, dorsal, or independent). Mixed effect linear regression models use structured covariance models to show dependence among observations within individuals (SAS/STAT® 2010), and were selected for their robusticity to missing values. Data were log transformed prior to statistical analyses and met the assumption of normality. The covariance structure is applied in repeated measure designs to control for potential observational dependence between individuals. All effects in this model were fixed. Mixed effects linear regression models determined the relationship between infant age and the dependent variable, time spent on carriers (mothers, adult males, adult females). A Poisson regression using a PROC GENMOD procedure designed to fit generalized linear models measured the significance between infant age and the number of occurrences mothers initiated and broke bodily contact with infants, and the numbers of occurrences infants initiated and broke bodily contact with mothers. All analyses were conducted in SAS® Version 9.2. Maps were created in ArcGIS® software version 10.0.

2.4. Results 2.4.1. Infant transport position: Ventral There was a negative relationship between P. coquereli infant age and percentage of time spent ventrally (Figure 2.2). Infants (n=10) were almost exclusively transported ventrally from weeks 1-5 postnatal by all carriers (n=37), with a median above 90% until week 6. The average was 96.69% in week 1 and 6.56% in week 26. Infants consistently spent the most time ventrally during 1-5 weeks postnatal, with the least amount of variation during weeks 1-3 postnatal. The minimum percentage of time infants spent ventrally increased in week 6 relative to weeks 1-5. Weeks 7-15 showed a relatively consistent decrease in time spent ventrally, with weeks 10 and 11 showing greater overall variation than preceding weeks. The mean in week 16 was slightly higher than week 15, but the overall range is comparable. Considerable variation was present during weeks 7-18. Weeks 17-26 showed consistent decreases in time spent ventrally. Min/max variation decreased beginning in week 22.

2.4.2. Infant transport position: Dorsal There was a positive relationship between P. coquereli infant age and percentage of time spent dorsally on all carriers until week 19 (Figure 2.3). Infants were not transported dorsally in weeks 1-2 postnatal. The average was 0.00% in week 1 and 22.21% in week 26. Infants were transported dorsally in very low percentages (<10%) in weeks 4-6, though with a much greater maximum in week 6 (35.08%). Weeks 7-10 showed a gradual increase in time spent dorsally. The max/min in weeks 11-12 were comparatively high and weeks 13-17 followed a positive trend, though the median in week 14 is less than week 13, the values fell within the expected range. Weeks 15-18 showed relatively high variability in means and mins/maxs. The greatest median percentage of time spent dorsally occurred in week 19 (48.40%). Weeks 20-22 had comparable ranges, and time spent dorsally time began to steadily decline in week 23, though considerable variation was present until week 25.

2.4.3. Infant transport position: Independent There was a positive relationship between P. coquereli infant age and the percentage of time spent independently (Figure 2.4). Infants spent no time independently in weeks 1-2. The average was 0% in week 1 and 68.42% in week 26. Time spent independently remained <10% until the maximum in weeks 6 and 7 reached 10.02% in the third quartile. Weeks 7-11 revealed a steady increase in time spent independently. Weeks 12 and 13 showed a slight decrease and the ranges are smaller. The week 14 min/max was 8.53% and 41.42%, the greatest range when all proceeding weeks were considered. Weeks 15-21 displayed gradual increases, and week 17 showed a very minimum (0.09%). Week 22 showed less variation between the minimum and all quartiles since occurred in week 5, with a comparatively high maximum (67.27%). Weeks 23-25 demonstrated a rapid increase in time spent independently from carriers, though mins/maxs maintained relatively high variation. Quartile 3 reached 71.90% in week 24, dropped to 62.47% in week 25, and increased to 75.89% in week 26.

2.4.4. Comparison of infant transport positions from 1-26 weeks postnatal Mixed effects linear regression models considered the time infants spent in ventral, dorsal, or independent positions. There was a significant relationship between duration of being carried

43 and infant position. Infants spent significantly more time on a carrier’s ventrum (least square mean [LSM]=4.87) (standard error [SE]=0.061), p<0.001) (Table 2.7). Infants spent intermediate time in the dorsal position (LSM=4.26, SE=0.062, p<0.001), as time spent in the dorsal position was significantly less than the ventral position, and was greater than being independent. Infants spent the least time independently from birth to 26 weeks postnatal (LSM=3.97, SE=0.062, p<0.001).

Infants spent the greatest percentage of time being transported ventrally when all weeks of postnatal development were considered (Figure 2.5). The average percentage of time spent ventrally was 50.84%. The average time spent dorsally was 26.10%. The average spent independently was 20.93%. The median percentage of time was 45.81%. The greatest variation occurred in ventral transport and ranged from 6.56-98.56%. The median percentage of time infants spent dorsally was 28.66%. The minimum for dorsal transport was 0.00% and maximum was 51.78%. Infants spent the least time independent from carriers, with a median of 17.23%. The minimum for independent infants was 0.00% and the maximum was 68.42%. The least amount of variation occurred when infants were independent.

2.4.5. Infant carriers A mixed effects linear regression model considered the duration infants were transported by P. coquereli mothers in relationship to other group members in addition to carrier identity (mothers, adult males, non-mother adult females). LSM revealed a significant relationship between the duration of maternal infant carriage (LSM=4.68, SE=0.062, p<0.001) in comparison to carriage by adult males and adult females that were not mothers (LSM=3.94, SE=0.062, p<0.001) from birth until 26 weeks postnatal (Table 2.8). P. coquereli mothers carried infants for significantly longer durations than adult males and adult females without infants.

There was a negative relationship between infant age and the percentage of time infants spent being transported by mothers (Figure 2.6). Infants spent a very high percentage of time (>80%) on mothers from weeks 1-9. The average percentage of time spent on mothers was 90.52% in week 1 and 22.47% in week 26. The median percentage of time was 90.52% in week 1 and 26.42% in week 26. The least overall variation occurred in weeks 1-6 and week 26. Weeks 10-

21 displayed relatively consistent ranges, with a gradual negative trend that emerged over these weeks. The average percentage of time spent on mothers was 77.54% in week 10 and 70.44% in week 21. The median percentage of time spent on mother was 70.49% in week 10 and 75.06% in week 21. The min/max was 59.77% and 91.78% in week 10, and 50.55% and 87.62% in week 21. The minimum (32.73%) was low relative to quartiles 1-3 in week 22. Considerably high mins/maxs were present in weeks 23-25, though both averages and medians steadily declined, with a relatively large decrease in week 25.

Infant transport by adult males was relatively low in weeks 1-26 postnatal (Figure 2.7). Males transported infants <10% of the time in all quartiles. Maximums in weeks 4, 6, 7, 12, and 13 were >10%. The average percentage of time males transported infants was the highest (6.17%) in week 1. Weeks 2-26 all have averages and medians <5%, weeks 4 (5.23%) and 13 (5.15%) excluded. There was considerable maximum variation in weeks 4, 6, 7, and 13. Much less variation in minimums was present when all weeks were considered.

Overall, adult females without their own infants transported infants the least amount of time relative to mothers and adult males (Figure 2.8). Adult females transported infants <6% of the time in all quartiles from weeks 1-25 postnatal. Weeks 5, 11, 15, 16, 17, 19, and 20 maximums were >6%. The average time adult females transported infants was lower than adult males in all weeks aside from 15-17, 19, 20, 22, 23, 24, and 25. Medians in weeks 15-17, 19 were less in adult females relative to males. Medians in weeks 20, 24, and 25 were greater in adult females than in males. Medians in weeks 22 and 23 were equal in adult females and males.

Mothers transported infants for the greatest percentage of time from weeks 1-26 postnatal (Figure 2.9). Mothers transported infants for an average of 74.20%, adult males for an average of 2.22%, and adult female non-mothers for an average of 1.25%. Infants spent an average of 20.97% independent from carriers. 28.76% was the minimum amount of time infants were transported by mothers. The minimum infants were transported by adult males, female non- mothers, and spent independently was 0.00%. The median for maternal transport was 77.41%, 1.82% in adult males, 0.61% in adult females, and 16.70% independent from carriers. Mother transport quartiles ranged from 68.27-88.15%, 1.10-3.04% in adult males, and 0.36-2.27% in adult female non-mothers. The maximum percentage of time infants spent transported by

45 mothers was 96.54%, 6.17% by adult males, 4.52% by adult females, and 68.42% independent from carriers.

The generalized linear models examining the number of occurrences mothers initiated and broke contact with infants from 1-26 weeks postnatal showed a negative parameter estimate (PE) (PE=- 0.324, SE=0.115, p<0.004) (Table 2.9). The negative slope indicated that the number of occurrences where mothers initiated infant contact significantly decreased and the frequency mothers broke infant contact increased with infant age. Mothers did not initiate contact with infants in weeks 1, 18, 19, 21-26 (Figure 2.10). The greatest frequencies occurred in weeks 2, 4, 6-9, and 12-13. The greatest frequency average of 1.00 occurred in week 8. Weekly medians were 0.00 (week 6=0.50, excluded). Frequencies did not exceed 1.00 when quartiles in all weeks were considered. Minimums ranged from 0.00-0.00 and maximums ranged from 0.00-7.00 in week 8. Mothers did not break infant contact in weeks 1, 3, 4-12, and 15-17 (Figure 2.11). The greatest frequencies occurred in weeks 20, 23-26. The greatest frequency average of 2.17 occurred in week 24. Weekly medians were 0.00 excluding 1.00 in weeks 25-26. Frequencies did not exceed 1.50 when quartiles in all weeks were considered. Minimums ranged from 0.00- 0.00 and maximums ranged from 0.00-15.00 in week 14.

The generalized linear models examining the number of occurrences infants initiated and broke contact with mothers from 1-26 weeks postnatal did not agree with prediction 4. The number of occurrences infants initiated and broke contact with mothers was significantly positive when a negative relationship was predicted (PE=0.008, SE=0.0002, p<.0001) (Table 2.10). The number of occurrences infants initiated contact with their mothers did not decrease as the number of occurrences infants broke contact with their mothers increased. Instead, a nearly identical relationship emerged between the frequencies infants initiated and broken contact (Figures 2.12, 2.13). The number of occurrences infants initiated contact increased as the number of occurrences infants broke contact with their mothers also increased from weeks 1-11. Considerable variation was present from weeks 12-19, though both data categories were nearly identical in their frequencies. Frequencies decreased weekly from weeks 20-26 in both data categories.

Infants initiated contact with mothers in all weeks (week 1 excluded) (Figure 2.12). The greatest

46 frequency average of 26.11 occurred in week 20. The greatest median frequency of 28.5 occurred in week 16. Frequencies did not exceed 13.75 when quartiles in all weeks were considered. Minimums ranged from 0.00-12.00 in week 19, while maximums ranged from 0.00- 56.00 in weeks 12 and 14.

Infants broke contact with mothers in all weeks (Figure 2.13). The greatest frequency average of 29.1 occurred in week 14. The greatest median frequency of 31.5 occurred in week 16. Frequencies did not exceed 10.25 when quartiles in all weeks were considered. Minimums ranged from 0.00-11.00 in week 19, while maximums ranged from 1.00-69.00 in week 14. Mothers initiated and broke infant contact less frequently than infants from 1-26 weeks postnatal (Figure 2.14). The number of average occurrences mothers initiated and broke contact was 2.58 ± 3.26 (n=10) and 2.88 ± 4.66 (n=10), respectively. The average number of occurrences infants initiated and broke contact was 132.42 ± 94.27 (n=10) and 137.73 ± 97.69 (n=10), respectively. Quartiles ranged from 0.00-4.00 in mother initiated contact and broken contact. Quartiles ranged from 37.75 to 221.25 in infant initiated contact and 45.75-232.50 in broken contact with mothers. The median for mother initiated contact was 1.00 and 0.00 for contact broken by mothers. The median for infant initiated contact was 149.50 and 148.50 for contact broken by infants. The minimum was 0.00 for mother initiated and broken contact. The minimum was 0.00 for infant initiated contact and 1.00 for broken contact with mothers. The maximum was 10.00 for mother initiated contact and 15.00 for broken contact. The maximum was 280.00 for infant initiated contact and 291.00 for broken contact with mothers.

2.5. Discussion

In this chapter, I evaluated four predictions related to P. coquereli maternal behavioral care- giving. P1. Maternal behavioral care-giving effort will decrease as infants increase in age as measured by infant position. Infants will spend the greatest duration in the ventral position, followed by the dorsal position, and the least duration independently. P2. Mothers will provide the majority of infant behavioral care-giving effort relative to adult males and non-mother adult females as measured by the frequency and duration of infant carriage.

P3. Infant contact initiated by P. coquereli mothers will decrease while infant contact broken by mothers will increase as infants age. P4. Contact with mothers initiated by P. coquereli infants will decrease while contact with mothers broken by infants will increase as infants age.

Predictions 1-3 were statistically supported. Prediction 4 was not statistically supported. This demonstrates P. coquereli infant transport positions were dependent on infant age, mothers provided significantly more infant transport compared to other group members, and the frequency of infant contact initiated by mothers decreased while the frequency of infant body contact broken by mothers increased as infants aged. Thus, P. coquereli mothers were the primary infant transporters and mothers decreased care-giving effort as infants aged.

P. coquereli mothers simultaneously lactate and carry infants for significantly more time than other group members while concurrently experiencing seasonal constraints. Adult Propithecus spp. exemplify slow life histories (Richard et al. 2002) and I propose infants follow this same pattern. P. coquereli infants spent significantly more time in the ventral position from 1-26 weeks postnatal than either dorsally or independently. Infants were less altricial as they aged, though infants remained dependent on carriers for relatively long durations during the 26th postnatal week. Infants were more ontogenetically constrained by independence than either ventral or dorsal position. Infants frequently returned to ventral and dorsal positions throughout the duration of this study, but engaged in independent locomotion more gradually. During early infant development, this study recorded a single infant as independent in week 3 and in week 4, 5 infants were observed independent. Infants spent the most consistent durations in the ventral or dorsal position during early development. Intermediate behavioral development is defined by more variation and exploratory in terms of locomotive development. Falling risk is decreased in the ventral position since carriers are more able to regulate infant movements opposed to the dorsal position where carriers have less bodily control over infants. Nipple access is facilitated when infants are ventral relative to dorsal, but infants may reach around mothers to access nipples located in armpits while in the dorsal position. Carriers could be more vigilant of ventrally facing infants and offer behavioral care with greater ease when infants were visible. Thus, the dorsal position is intermediate and provides a step towards independence.

The transition from locomotive dependence to independence is longer in P. coquereli infants than originally thought. Infants were carried ventrally more than 90% of the time until week 7 and still carried 26% of the time in week 26. These are exceptionally high percentages of time given infants were previously reported as largely independent even earlier in development (see Richard 1974; Richard 1976). A preliminary study on four P. coquereli infants suggested infants were carried only ventrally until 3-4 weeks postnatal and subsequently were carried dorsally by their mothers and other group members (Richard 1976). Contra to Richard (1976), another study determined infants were dorsal intermittently by 3 months postnatal (Jolly 1966). Infants were previously reported to only cling ventrally until 7 weeks postnatal (Richard 1976). Contra to Richard (1976), infants continued to be carried ventrally throughout the duration of this study (1- 26 weeks postnatal).

P. coquereli infants are typically considered weaned and largely independent between 20-24 weeks postnatal (Richard 1974; Richard 1976). While P. coquereli infants develop quickly relative to haplorhines, infants have slower maturation that cannot be exclusively attributed body size in comparison to other lemurs. Eaglen and Boskoff (1978) first proposed P. coquereli captive infants develop more slowly than other lemurs of comparable body size, based on extended infant carriage and delayed independence relative to Varecia spp., Lemur catta (ring- tailed lemurs), and E. fulvus. Black-and-white ruffed lemurs (Varecia variegata) are of similar body size, but infants develop much more rapidly and are fully mobile by 3-4 months postnatal (Baden 2011; Morland 1990). In the earliest study, the first observation of P. verreauxi infant independence occurred during the second postnatal month (Jolly 1966), but was recorded in P. coquereli at only 15 days postnatal in another study (Richard 1976). In striking contrast to P. coquereli, where infants were dorsal less than 15% of the time until week 9, and the maximum time spent independently was 77% in week 26, L. catta infants frequently rode dorsally during the second postnatal week (Sussman 1977) and were primarily independent by 16 weeks postnatal (Gould 1992).

I documented that P. coquereli mothers provided significantly more infant behavioral care- giving in comparison to all other group members. I emphasize that although mothers were the primary transporters, adult males (<10% in all quartiles) and female non-mothers (<6% in all quartiles) still participated in transport (Figures 2.9, 2.10, 2.11). There was considerable weekly

49 variation in allomaternal transport and determining the factors causing this variation will provide insight into the advantages of infant carrying by non-mothers. P. coquereli mothers did not utilize other group members for extended durations of infant transport in contrast to other primate species where allomaternal care assists in offsetting energetic costs mothers (Schradin and Anzenberger 2001a; Schradin and Anzenberger 2001b; Tardif 1994; Tardif et al. 1986).

Lactation is the most energetically costly activity engaged in by mammals and increases daily energy expenditures up to 150% (Gittleman and Thomspon 1988; Lee et al. 1991). This is a tremendous increase without considering the adverse effects of the dry season on P. coquereli mothers. For example, P. verreauxi females lose 18% of their body mass during the dry season (Lewis and Kappeler 2005b). Callitrichids twin litters are exceptionally heavy for carriers, weighing between 15-25% of a female’s body weight at birth (Kleiman 1977). Assuming a P. coquereli infants weighs ~100g (Ballentine 2013) and the adult carrier weighs ~4kg (Mittermeier 2010), a typical infant weighs only ~2.5% of a carrier’s body weight at birth. In the future, comparing P. coquereli birth weights and growth rates to those of other lemurs of similar body size would help establish the direct energetic costs of infant transport.

The importance of infants receiving sufficient milk nutrients during early infant development likely contributes to mothers acting as primary carriers. Infants spent the greatest durations (>80%) on mothers from weeks 1-9 postnatal, which suggests infants are most dependent on mothers during the first 9 weeks of life. P. coquereli and diademed sifaka (Propithecus diadema) infants spent the most time on mothers followed by the dominant male and high ranking females (Grieser 1992). In the same study, infants spent the least time on immigrant males and subordinate females (Grieser 1992). Carrier patterns did not fluctuate as the infant developed from four to ten weeks of age, but the P. coquereli infant spent more time off the mother than P. diadema infants from weeks 4-10 (Grieser 1992). The same study concludes that P. coquereli develops more slowly than P. diadema (Grieser 1992). Milk continued to be the primary nutrition source during the second postnatal month in P. coquereli infants, and differences between the time spent nursing and feeding was no longer significant beginning in month three (Grieser 1992). However, caution must be exercised in quantifying nursing time as the amount of time infants time spend on the mother’s ventrum is not reflective of milk or nutrient transfer, but more accurately measures nipple access (Cameron 1998; Cameron et al.

1999). P. coquereli infants were seen attached on mother’s nipples from the dorsal position, suggesting that infants have nipple access from both infant transport positions (A. Ross, pers. comm.). Future studies of milk nutrient composition will assist in determining the nutritional quality infants receive during early infant development.

The relative absence of bimaturism in lemurs in combination with rapid postnatal growth over short durations compared to body size demonstrates the critical role seasonality plays in lemur reproduction and ontogeny since sexual dimorphism would be expected in taxa with high intermale competition like many lemur species (Leigh and Terranova 1998). P. coquereli mothers provide the majority of nutritional and behavioral investment in infants while living in a highly unpredictable environment (see Dewar and Richard 2007), characterized by rainfall variability, long periods of fruit scarcity (Wright 1999), poor quality soils (Crowley et al. 2012; Du Puy and Moat 1996; Lourenço and Goodman 2006), and small tree crowns (Balko and Underwood 2005; McGoogan 2011). The mid-lactation period is the most energetically strenuous lactation period (Oftedal 1984; Oftedal and Iverson 1995) and in P. coquereli coincides with the annual period of extreme desiccation from May-September (Rendigs et al. 2003). Thus, mothers experience the most energetically costly physiological state while acting as the primary infant transporters concurrently with the fewest available resources in the austral winter.

Oral transport or infant parking is the ancestral condition of primate infant transport with the derived condition being infants clinging to the fur of carriers (Kappeler 1998; Ross 2001b). Both ancestral and derived characters are present in extant lemurs (Tecot et al. 2013), with Propithecus spp. following the derived condition (Richard 1976). Carrying infants is energetically expensive and decreases foraging efficiency and feeding time in carriers while increasing predation risk (Goldizen 1987; Sánchez et al. 1999; Schradin and Anzenberger 2001a; Schradin and Anzenberger 2001b; Tardif 1994; Tardif and Bales 1997; Tardif et al. 1993). In cotton-top tamarins (Saguinus oedipus), a species where allomaternal care is extensive, males lose up to 11% of their body weight when infants are most dependent during the mother’s periovulatory period (Achenbach and Snowdon 2002; Sánchez et al. 1999). No relationship was found between feeding observations and weight loss, suggesting that carrying as a form of transport itself is primarily responsible for weight reduction in cotton-top tamarin carriers (Sánchez et al. 1999). Common marmoset (Callithrix jacchus) carriers leapt significantly shorter

51 distances (17%) while increasing basal metabolic rates by 20% when saddled with infants (Schradin and Anzenberger 2001a). Decreases in the rapidity and travel distance presumably increase susceptibility to predators, though direct measures in the wild are not available (Schradin and Anzenberger 2001a).

Propithecus spp. are characterized by extremely slow life histories (Richard et al. 2002). More than 50% of Propithecus spp. infants die within their first postnatal year (Kappeler et al. 2009; Richard et al. 2002). P. verreauxi infant deaths occur primarily during the wet season or shortly after birth, although no significant pattern has been detected (Richard et al. 1991). The mortality rate in this study was 9% in the first 6.5 postnatal months, substantially lower than previous studies. Thus, P. coquereli mothers are successful carriers. The energy conservation hypothesis proposes that less postnatal maternal investment is present in lemurs relative to other primates, and that mothers may reserve energetic resources for subsequent births due to high rates of infant mortality (Wright 1999). In P. coquereli, if mothers reduce postnatal investment as is predicted by the energy conservation hypothesis (Wright 1999), it is not in expressed in infant transport. Twenty-four month interbirth intervals may help mothers recover energetically to allow extended infant transport. The energy conservation hypothesis supports the bet-hedging reproductive strategy characteristic of Propithecus spp., where females invest less per infant and have comparatively slow life histories (Richard et al. 2002). It is imperative for mothers to respond quickly to environmental changes albeit seasonal or precipitous to achieve a balance between energetic costs to themselves while maximizing infant care-giving effort and preparing for future pregnancies.

Litter size contributes to low rates of P. coquereli allomaternal care since only a single infant is present per group (Richard and Dewar 1991), in comparison to V. variegata where multiple infants are typical and allomaternal care is frequent (Baden 2011; Morland 1990). Communal care in V. variegata is biased towards kin and female affiliates and has been shown to improve infant survivorship as well as maternal energy balance (Baden 2011). Conspecifics of different age and sex classes, including non-father adult males, carried L. catta infants (Gould 1992). P. coquereli males participate in infant care in varying frequencies and durations, but overall male infant care occurs at low rates (Bastian and Brockman 2007; Eaglen and Boskoff 1978; Grieser 1992; Richard 1976; This study). Jolly (1966) observed male interest in infants, though

52 longitudinal contact rates are not available. Bastian and Brockman (2007) found that captive fathers provided 38.5% of infant care, with older males grooming and carrying infants more frequently than younger males. Infant transport by non-mothers is opportunistic and likely related to other socioecological factors instead of energetic necessity characteristic of callitrichids. Thus, mothers do not cope with difficult environmental conditions by passing off the energetic burden of carrying infants to conspecifics, and instead employ other behavioral, physiological, or nutritional strategies to combat an extreme climate. V. variegata typically has 2-3 infants per litter (Baden 2011), thereby energetically requiring allomaternal care. Female social dominance is one explanation as to why allomaternal transport occurs at significantly lower rates than maternal transport in P. coquereli. Female feeding priority may assist with mothers sustaining their own energy requirements while transporting infants. Reduced basal metabolic rates may assist mothers with offsetting the high costs of lactation, transport, and a harsh environment.

The occurrence of bodily contact between mothers and infants is a direct measure of interactions between P. coquereli mothers and infants. Mothers initiated and broke infant contact less frequently than infants from 1-26 weeks postnatal. The highest frequencies of infant initiated contact by mothers occurred during early infant development, suggesting that mothers prefer to maintain closer proximity opposed to later development. Infants choose contact frequency with mothers. Infants initiated contact at the highest frequencies during intermediate development. Infants are reluctant to leave mothers during this time and is indicative of parent-offspring conflict (Trivers 1974). Mothers are exceptionally tolerant of infant bodily contact until late infant development; and even during this time mothers broke contact with infants at comparatively lower rates than infants broke contact with their mothers.

2.6. Author note on publication

A modified version of Chapter 2 of this dissertation was previously published. Ross A, and Lehman S. 2016. Infant transport and mother–infant contact from 1 to 26 weeks postnatal in Coquerel’s sifaka (Propithecus coquereli) in northwestern Madagascar. American Journal of Primatology 78:646-658.

2.7. References

Abbott D, Keverne E, Bercovitch F, Shively C, Mendoza S, Saltzman W, Snowdon C, Ziegler T, Banjevic M, and Garland Jr. T, Sapolsky, RM. 2003. Are subordinates always stressed? A comparative analysis of rank differences in cortisol levels among primates. Hormones and Behavior 43:67-82. Achenbach G, and Snowdon C. 2002. Costs of care-giving: weight loss in captive adult male cotton-top tamarins (Saguinus oedipus) following the birth of infants. International Journal of Primatology 23:179-189. Albignac R. 1981. Lemurine social and territorial organization in a north-western Malagasy forest (restricted area of Ampijoroa). In: Chiarelli A, and Corruccini R, editors. Primate Behavior and Sociobiology. New York: Springer-Verlag. p 25-29. Alexander R. 1993. The relative merits of foregut and hindgut fermentation. Journal of Zoology: Proceedings of the Zoological Society of London 231:391-401. Alonso L, and Hannah L. 2002. Introduction to the Reserve Naturelle Integrale and Reserve Forestiere d’Ankarafantsika and to the rapid assessment program. In: Schulenberg T, Radilofe S, and Missa O, editors. A biological assessment of the Reserve Naturelle Integrale d’Ankarafantsika, Madagascar. Washington DC: Conservation International p 28-62. Altmann J. 1974. Observational study of behavior: sampling methods. Behaviour 49:227-267. Altmann J. 1988. Determinants of reproductive success in savannah baboons. In: Clutton-Brock T, editor. Reproductive Success. Chicago: University of Chicago Press p 403-418. Altmann J, and Alberts S. 2005. Growth rates in a wild primate population: ecological influences and maternal effects. Behavioral Ecology and Sociobiology 57:490-501. Altmann J, and Samuels A. 1992. Costs of maternal care: infant-carrying in baboons. Behavioural Ecology and Sociobiology 29:391-398. Andrainarivo C, Andriaholinirina, V.N., Feistner, A., Felix, T., Ganzhorn, J., Garbutt, N., Golden, C., Konstant, B., Louis Jr., E., Meyers, D., Mittermeier, R.A., Perieras, A., Princee, F., Rabarivola, J.C., Rakotosamimanana, B., Rasamimanana, H., Ratsimbazafy, J., Raveloarinoro, G., Razafimanantsoa, A., Rumpler, Y., Schwitzer, C., Thalmann, U., Wilmé, L. & Wright, P. 2008. IUCN Red List of Threatened Species: Propithecus coquereli. 17 November 2013 ed. Baden A. 2011. Communal infant care in black-and-white ruffed lemurs (Varecia variegata) [dissertation]. Available from ProQuest, 3490436: Stony Brook University 252 p. Bales K, Dietz J, Baker A, Miller K, and Tardif S. 2000. Effects of allocare-givers on fitness of infants and parents in callitrichid primates. Folia Primatologica 71:27-38. Bales K, French J, and Dietz J. 2002. Explaining variation in maternal care in a cooperatively breeding mammal. Animal Behaviour 63:453-461. Bales K, French J, Hostetler C, and Dietz J. 2005. Social and reproductive factors affecting cortisol levels in wild female golden lion tamarins (Leontopithecus rosalia). American Journal of Primatology 67:25-35. Balko E, and Underwood H. 2005. Effects of forest structure and composition in food availability for Varecia variegata at Ranomafana National Park, Madagascar American Journal of Primatology 66:45-70. Ballentine J. 2013. Maryland Zoo Welcomes Newborn Sifaka Baby. Press release: The Maryland Zoo in Baltimore p 1-2. Bardi M, Shimizu K, and Borgognini-Tarli S. 2003. Mother-infant relationships and maternal estrogen metabolites changes in macaques (Macaca fuscata, M. mulatta). Primates 44:91-98. Bastian M, and Brockman D. 2007. Paternal care in Propithecus verreauxi coquereli. International Journal of Primatology 28:305-313. Beehner J, and McCann C. 2008. Seasonal and altitudinal effects on glucocorticoid metabolites in a wild primate (Theropithecus gelada). Physiology & Behavior 95:508-514. Ben Shaul D. 1962. The composition of the milk of wild animals. Journal of Dairy Science 4:333-342 Bereczkei T. 2001. Maternal trade-off in treating high-risk children. Evolution and Human Behavior 22:197-212. Bernardo J. 1996. Maternal effects in animal ecology. American Zoologist 36:83-10Boonstra R, Hik D, Singleton GR, and Tinnikov A. 1998. The impact of predator-induced stress on the snowshoe hare cycle. Ecological Monographs 68:371-394. Brockman D, Whitten P, Richard A, and Schneider A. 1998. Reproduction in free-ranging male Propithecus verreauxi: the hormonal correlates of mating and aggression. American Journal of Physical Anthropology 105:137-151. Brockman DK, Cobden AK, and Whitten PL. 2009. Birth season glucocorticoids are related to the presence of infants in sifaka (Propithecus verreauxi). Proceedings of the Royal Biological Society B 276:1855-1863. Burgess M, and Chapman C. 2005. Tree leaf chemical characters: selective pressures by folivorous primates and invertebrates African Journal of Ecology 43:242-250 54 55

Busch D, and Hayward L. 2009. Stress in a conservation context: A discussion of glucocorticoid actions and how levels change with conservation-relevant variables. Biological Conservation 142:2844-2853. Cameron E. 1998. Is suckling behaviour a useful predictor of milk intake? A review. Animal Behaviour 56:521-532. Cameron E, and Linklater W. 2000. Individual mares bias investment in sons and daughters in relation to their condition. Animal Behaviour 60:359-367. Cameron E, Stafford K, Linklater W, and Veltman C. 1999. Suckling behaviour does not measure milk intake in horses. Animal Behaviour 57:673-678. Cameron JL. 1996. Regulation of reproductive hormone secretion in primates by short-term changes in nutrition Reviews of Reproduction 1:117-126. Campbell J, Williams C, and Eisemann J. 2004. Use of total dietary fiber across four lemur species (Propithecus verreauxi coquereli, Hapalemur griseus griseus, Varecia variegata, and Eulemur fulvus): does fiber type affect digestive efficiency? American Journal of Primatology 64:323-335. Casolini P, Cigliana G, Alemá G, Ruggieri V, Angelucci L, and Catalani A. 1997. Effect of increased maternal corticosterone during lactation on hippocampal corticosteroid receptors, stress response and learning in offspring in the early stages of life. Neuroscience 79:1005-1012. Champagne F, Francis D, Mar A, and Meaney M. 2003. Variations in maternal care in the rat as a mediating influence for the effects of environment on development. Physiology and Behavior 79:359-371 Chapman CA, Chapman L, Naughton-Treves L, Lawes MJ, and McDowell LR. 2004. Predicting a folivorous primate abundance: validation of a nutritional model American Journal of Primatology 62:55-69. Chapman CA, Chapman LJ, Rode KD, Hauck EM, and McDowell LR. 2003. Variation in the nutritional value of primate foods: among trees, time periods, and areas. International Journal of Primatology 24:317-333. Cheverud J. 1984. Evolution by kin selection: a quantitative genetic model illustrated by maternal performance in mice. Evolution 38:766-777. Cheverud JM, and Wolf JB. 2009. The genetics and evolutionary consequences of maternal effects. In: Maestripieri D, and Mateo JM, editors. Maternal Effects in Mammals. Chicago: University of Chicago Press Chism J. 2000. Allocare patterns among cercopithecines. Folia Primatologica 71:55-66. Clutton-Brock T. 1991. Parental care in birds and mammals. In: Krebs J, and Clutton-Brock T, editors. The Evolution of Parental Care. Princeton Princeton New Jersey Press. p 131-152. Creel S, Fox J, Hardy A, Sands J, Garrott R, and Peterson R. 2002. Snowmobile activity and glucocorticoid stress responses in wild wolves and elk. Conservation Biology 16:809-814. Crockett C, and Rudran R. 1987. Red howler monkey birth data II: interannual, habitat, and sex comparisons. American Journal of Primatology 13:369-384. Crowley B, McGoogan K, and Lehman S. 2012. Edge effects on foliar stable isotope values in a Madagascan tropical dry forest. Public Library of Science ONE 7:1-9. Dall S, and Boyd I. 2004. Evolution of mammals: lactation helps mothers to cope with unreliable food supplies. Proceedings of the Royal Society B 271:2049-2057. Dewar R, and Richard A. 2007. Evolution in the hypervariable environment of Madagascar. Proceedings of the National Academy of Sciences of the USA 104:13723-13727. Dewey K. 1997. Energy and protein requirements during lactation. Annual Review of Nutrition 17:19-36. Dobson F, and Michener G. 1995. Maternal traits and reproduction in richardson's ground squirrels. Ecology 76:851- 862. Du Puy D, and Moat J. 1996. A refined classification of the primary vegetation of Madagascar based on the underlying geology: using GIS to map its distribution and assess its conservation status. In: Lourenço W, editor. Proceedings of the international symposium on the biogeography of Madagascar Paris: Editions de l'Orstrom p205-218. Dunham AE, Erhart EM, and Wright PC. 2011. Global climate cycles and cyclones: consequences for rainfall patterns and lemur reproduction in southeastern Madagacar Global Change Biology 17:219-227. Eaglen R, and Boskoff K. 1978. The birth and early development of a captive sifaka, Propithecus verreauxi coquereli. Folia Primatologica 30:206-219. Eisen E, and Saxton A. 1983. Genotype by environment interactions and genetic correlations involving two environmental factors. Theoretical and Applied Genetics 67:75-86. Fairbanks L, and McGuire M. 1995. Maternal condition and the quality of maternal care in vervet monkeys. Behaviour 132:733-754. Fedigan L, Rose L, and Avila R. 1996. See how they grow: tracking capuchin monkey (Cebus capucinus) populations in a regenerating Costa Rican dry forest. In: Norconk M, editor. Adaptive radiations of neotropical primates New York: Plenum. p 289-307.

Fichtel C, Kraus C, Ganswindt A, and Heistermann M. 2007. Influence of reproductive season and rank on fecal glucocorticoid levels in free-ranging male Verreaux's sifaka (Propithecus verreauxi). Hormones and Behavior 51:640-648. Fite J, Patera K, French J, Rukstalis M, Hopkins E, and Ross C. 2005. Opportunistic mothers: female marmosets (Callithrix kuhlii) reduce their investment in offspring when they have to, and when they can. Journal of Human Evolution 49:122-142. Fleming S, Ruble D, Krieger H, and PY W. 1997. Hormonal and experimental correlates of maternal responsiveness during pregnancy and the puerperium in human mothers. Hormones and Behavior 31:145-158. Galloway L. 2005. Maternal effects provide phenotypic adaptation to local environmental conditions. New Phytologist 166:93-100. Ganzhorn J. 1995a. Low-level forest disturbance effects on primary production, leaf chemistry, and lemur populations. Ecology 76:2084-2096. Ganzhorn J. 2002. Distribution of a folivorous lemur in relation to seasonally varying food resources: integrating quantitative and qualitative aspects of food characteristics. Oecologia 131:427-435. Ganzhorn J, Lowry P, Schatz G, and Sommer S. 2001. The biodiversity of Madagascar: One of the world's hottest hotspots on its way out. Oryx 37:115-118. Genoud M. 2002. Comparative studies of basal metabolism in primates. Evolutionary Anthropology 11:108-111. Gerardo G, and Goodman SM. 2003. Hunting of protected animals in the Parc National d'Ankarafantsika, northwestern Madagascar. Oryx 37:115-118. Girard-Buttoz C, Heistermann M, Krummel S, and Engelhardt A. 2009. Seasonal and social influences on fecal androgen and glucocorticoid excretion in male long-tailed macaques (Macaca fascicularis). Physiology & Behavior 98:168-175. Gittleman J, and Thomspon S. 1988. Energy allocation in mammalian reproduction. American Zoologist 28:863- 875. Gogarten J, Guzman M, Chapman C, Jacob A, Omeja P, and Rothman J. 2012. What is the predictive power of the colobine protein-to-fiber model and its conservation value? Tropical Conservation Science 5:381-393. Goldizen A. 1987. Tamarins and marmosets: communal care of offspring. In: Smuts B, Cheney D, Seyfarth R, and Wrangham R, editors. Primate Societies Chicago: University of Chicago Press. p 34-43. Goodman M, Porter C, Czelusniak J, Page S, Schneider H, Shoshani J, Gunnell G, and Groves C. 1998. Toward a phylogenetic classification of Primates based on DNA evidence complemented by fossil evidence. Molecular Phylogenetics and Evolution 9:585-598. Goodman S, and Benstead J. 2005. Updated estimates of biotic diversity and endemism for Madagascar. Oryx 39:73-77. Gould L. 1992. Alloparental care in free-ranging Lemur catta at Berrenty Reserve, Madagascar. Folia Primatologica 58:72-83. Gould L, Power ML, Ellwanger N, and Rambeloarivony H. 2011. Feeding behavior and nutrient intake in spiny forest- dwelling ring-tailed lemurs (Lemur catta) during early gestation and early to mid-lactation periods: compensating in a harsh environment. American Journal of Physical Anthropology 145:469-479. Gould L, Sussman RW, and Sauther ML. 1999. Natural disasters and primate populations: the effects of a 2-year drought on a naturally occurring population of ring-tailed lemurs (Lemur catta) in southwestern Madagascar. International Journal of Primatology 20:69-84. Gould L, Sussman RW, and Sauther ML. 2003. Demographic and life-history patterns in a population of ring-tailed lemurs (Lemur catta) at Beza Mahafaly Reserve, Madagascar: A 15-year perspective. American Journal of Physical Anthropology 120:182-194. Gould L, Ziegler T, and Wittwer D. 2005. Effects of reproductive and social variables on fecal glucocorticoid levels in a sample of adult male ring-tailed lemurs (Lemur catta) at the Beza Mahafaly Reserve, Madagascar. American Journal of Primatology 67:5-23. Grieser B. 1992. Infant development and parental care in two species of sifakas. Primates 33:305-314. Hambraeus L. 1984. Human milk composition. Nutrition Abstracts and Reviews in Clinical Nutrition 54:221-236. Hamilton W. 1967. Extraordinary sex ratios. Science 156:477-488. Hanson A, and Sodaro V. 2002. Ethogram Menu: Species: Goeldi's Monkey (Callimico goeldii). http://www.lpzoo.com/ethograms/ethoframes.html. Harcourt A. 2008. Are lemurs' low basal metabolic rates an adaptation to Madagascar's unpredictable climate? Primates 49:292-294. Harper G, Steininger M, Tucker C, Juhn D, and Hawkins F. 2007. Fifty years of deforestation and forest fragmentation in Madagascar. Environmental Conservation 34:325-333.

Hemingway C. 1998. Selectivity and variability in the diet of Milne-Edwards’ Sifakas (Propithecus diadema edwardsi): Implications for folivory and seed-Eating. American Journal of Primatology 19:355-357. Hinde K, and Milligan L. 2011. Primate milk: proximate mechanisms and ultimate perspectives. Evolutionary Anthropology 20:9-23. Hinde R, and Atkinson S. 1970. Assessing the roles of social partners in maintaining mutual proximity, as exemplified by mother-infant relations in rhesus monkeys. Animal Behaviour 18:169-176. Inchausti P, and Ginzburg LR. 1998. Small mammals cycles in northern Europe: patterns and evidence for maternal effect hypothesis. Journal of Animal Ecology 67:180-194. Irwin M. 2008b. Feeding ecology of Propithecus diadema in forest fragments and continuous forest. International Journal of Primatology 29:95-115. Iverson S, and Oftedal O. 1995. Philogenetic and ecological variation in the fatty acid composition of milks. In: Jensen RG, editor. Handbook of Milk Composition. New York: Academic Press. Janson C, and van Schaik C. 2002. Ecological risk aversion in juvenile primates: slow and steady wins the race In: Pereira M, and Fairbanks L, editors. Juvenile Primates: Life history, Development, and Behavior. Chicago: The University of Chicago Press p 57-74. Jessop N. 1997. Protein metabolism during lactation. Proceedings of the Nutrition Society 56:169-175. Jolly A. 1966. Propithecus verreauxi verreauxi. Lemur Behavior: A Madgascar Field Study. Chicago: University of Chicago Press. p 20-68. Jolly A. 1984. The puzzle of female feeding priority. In: Small M, editor. Female Primates: Studies by Women Primatologists. New York: Alan Liss, INC. Kappeler P. 1993. Sexual selection and lemur social systems. In: Kappeler P, and Ganzhorn J, editors. Lemur Social Systems and their Ecological Basis. New York: Plenum Press. p 223-240. Kappeler P, Mass V, and Port M. 2009. Even adult sex ratios in lemurs: potential costs and benefits of subordinate males in verreaux’s sifaka (Propithecus verreauxi) in the Kirindy Forest CFPF, Madagascar. American Journal of Physical Anthropology 140:487-497. Kappeler PM. 1991. Patterns of sexual dimorphism in body weight among prosimian primates. Folia Primatologica 57:132-146. Kappeler PM. 1998. Nests, tree holes, and the evolution of primate life histories. American Journal of Primatology 46:7-33. Keay J, Jatinder Singh B, Gaunt M, and Kaur T. 2006. Fecal glucocorticoids and their metabolites as indicators of stress in various mammalian species: A literature review. Journal of Zoo and Wildlife Medicine 37:234- 244. King S, Arrigo-Nelson S, Pochron S, Semprebon G, Godfrey L, Wright PC, and Jernvall J. 2005. Dental senescence in a long-lived primate links infant survival to rainfall. Proceedings of the National Academy of Sciences of the USA 10:16579-16583. Kirkpatrick M, and Lande R. 1989. The evolution of maternal characters. Evolution 43:485-503 Kleiman D. 1977. Monogamy in mammals. The Quarterly Review of Biology 52:39-69. Kubzdela K. 1997. Sociodemography in diurnal primates: The effects of group size and female dominance rank on intra-group spatial distribution, feeding competition, female reproductive success, and female dispersal patterns in white sifaka (Propithecus verreauxi verreauxi) [dissertation]. Chicago: University of Chicago. Kull C. 2000. Deforestation, erosion, and fire: degradation myths in the environmental history of Madagascar Environment and History 6:1-28. Kun-Rodrigues C, Salmona J, Besolo A, Rasolondraibe E, Rabarivola C, Marques TA, and Chikhi L. 2014. New density estimates of a threatened sifaka species (Propithecus coquereli) in Ankarafantsika National Park. American Journal of Primatology 76:515-528. Lacey E. 1998. What is an adaptive environmentally induced parental effect? In: Mousseau T, and Fox C, editors. Maternal Effects As Adaptations. Oxford: Oxford University Press. p 54-66. Lambert J. 1998. Primate digestion: interactions among anatomy, physiology, and feeding ecology. Evolutionary Anthropology 7:8-20. Langer P. 2008. The phases of maternal investment in eutherian mammals. Zoology 111:148-162. Larson A, and Losos J. 1996. Phylogenetic systematics of adaptation San Diego: Academic Press. 507 p. Lee P, Majluf P, and Gordon I. 1991. Growth, weaning and maternal investment from a comparative perspective. Zoology 225:99-114. Lee PC, and Bowman JE. 1995. Influence of ecology and energetics on primate mothers and infants. Motherhood in Human and Nonhuman Primates. Basel: Karger. p 47-58. Lehner P. 1996. Handbook of Ethlogical Methods: Cambridge University Press. 672 p.

Leigh S, and Terranova C. 1998. Comparative perspectives on bimaturism, ontogeny, and dimorphism in lemurid primates. International Journal of Primatology 19:723-749. Leutenegger W. 1973. Maternal-fetal weight relationships in primates. Folia Primatologica 20:280-293. Lewis R, and van Schaik C. 2007. Bimorphism in male Verreaux's sifaka in the Kirindy Forest of Madagascar. International Journal of Primatology 28:159-182. Lewis RJ, and Kappeler PM. 2005a. Are Kirindy sifaka capital or income breeders? It depends. American Journal of Primatology 67:365-369. Lewis RJ, and Kappeler PM. 2005b. Seasonality, body condition, and timing of reproduction in Propithecus verreauxi verreauxi in the Kirindy Forest. American Journal of Primatology 67:347-364. Lourenço W, and Goodman S. 2006. Notes on the postembryonic development and ecology of Grosphus hirtus Kraepelin, 1901 (Scorpiones, Buthidae) from the Parc National d'Ankarafantsika, northwest Madagascar. Zoologischer Anzeiger 244:181-185. Maestripieri D. 2009. Maternal influences on offspring growth, reproduction, and behavior in primates. In: Maestripieri D, and Mateo J, editors. Maternal Effects in Mammals. Chicago: University Chicago Press. p 256-291. Maestripieri D, and Mateo J. 2009. The role of maternal effects in mammalian evolution and adaptation In: Maestripieri D, and Mateo J, editors. Maternal Effects in Mammals. Chicago University of Chicago Press. p 1-10. Mass V, Heistermann M, and Kappeler PM. 2009. Mate-guarding as a male reproductive tactic in Propithecus verreauxi. International Journal of Primatology 30:389-409. Masters J, de Wit M, and Asher R. 2006. Reconciling the origins of Africa, India, and Madagascar with vertebrate dispersal scenarios Folia Primatologica 77:99-418. Mateo J. 2009. Maternal influences on development, social relationships, and survival behaviors In: Maestripieri D, and Mateo J, editors. Maternal Effects in Mammals. Chicago: Chicago University Press. p 133-158. Maynard L, and Loosli J. 1965. Bioenergetics. In: Maynard L, and Loosli J, editors. Animal Nutrition New York: McGraw-Hill. p 29-47. Mayor M, Sommer J, Houck M, Zaonarivelo J, Wright P, Ingram C, Engel S, and Louis EJ. 2004. Specific Status of Propithecus spp.. International Journal of Primatology 25:875-900. McGinley M, Temme D, and Geber M. 1987. Parental investment in offspring in variable environments: theoretical and empirical considerations. American Naturalist 130:370-398. McGoogan K. 2011. Edge effects on the behaviour and ecology of Propithecus coquereli in Northwest Madagascar [dissertation]. Available from TSpace Repository, 1807/31861: University of Toronto 225 p. McLaren A. 1981. Analysis of maternal effects on development in mammals. Journal of Reproduction and Fertility 62:591-596. Milligan L, Gibson S, WIliams L, and Power M. 2007. The composition of milk from Bolivian squirrel monkeys (Saimiri boliviensis boliviensis). American Journal of Primatology 69:1-14. Milton K. 1979. Factors influencing leaf choice by howler monkeys: a test of some hypotheses of food selection by generalist herbivores. American Naturalist 114:362-378. Mittermeier R, Louis Jr., E, Richardson, M, Schwitzer, C, Langrand, O, Rylands, A, Hawkins, F, Rajaobelina, S, Ratsimbazafy, J, Rasoloarison, R, Roos, C, Kappeler, P, Mackinnon, J. 2010. Indriidae. Lemurs of Madagascar. Third ed. Bogotá, Colombia: Conservation International. p 481-597. Moberg G. 2000. Biological response to stress: implications for animal welfare. In: Moberg G, and Mench J, editors. The Biology of Animal Stress: Basic Principles and Implications for Animal Welfare Cambridge, MA: CABI Publishing p1-21. Morland H. 1990. Parental behavior and infant development in ruffed lemurs (Varecia variegata) in a northeast Madagascar rain forest. American Journal of Primatology 20:253-265. Mousseau TA, and Fox CW. 1998. The adaptive significance of maternal effects. TREE 13:403-407. Nievergelt C, and Martin R. 1999. Energy intake during reproduction in captive common marmosets (Callithrix jacchus). Physiology & Behavior 65:849-854. Norscia I, Carrai V, and Borgognini-Tarli SM. 2006. Influence of dry season and food quality and quantity on behavior and feeding strategy of Propithecus verreauxi in Kirindy, Madagascar. International Journal of Primatology 27:1001-1022. Oftedal O. 1984. Milk composition, milk yield and energy output at peak lactation: a comparative review. Symposium Zoological Society of London 51:33-85. Oftedal O. 1991. The nutritional consequences of foraging in primates: the relationship of nutrient intakes to nutrient requirements Philosophical Transactions of the Royal Society B 334:161-170.

Oftedal O. 1993. The adapation of milk secretion to the constraints of fasting in bears, seals, and baleen whales. Journal of Dairy Science 76:3234-3246. Oftedal O, and Iverson S. 1995. Comparative analysis of nonhuman milks: Phylogenetic variation in the gross composition of milks. In: Jensen RG, editor. Handbook of Milk Composition. New York: Academic Press. Oster G, and Alberch P. 1982. Evolution and bifurcation of developmental programs. Evolution 36:444-459. Ostner J, Kappeler P, and Heistermann M. 2008. Androgen and glucocorticoid levels reflect seasonally occurring social challenges in male redfronted lemus (Eulemur fulvus rufus). Behavioral Ecology and Sociobiology 62:627-638. Pastorini J, Forstner M, and Martin R. 2001. Phylogenetic history of sifakas (Propithecus: Lemuriformes) derived from mtDNA sequences. American Journal of Primatology 53:1-17. Pen I, and Weissing F. 2000. Sex-ratio optimization with helpers at the nest. Proceedings of the Royal Biological Society B 267:539-543. Petter-Rousseaux A. 1962. Studies on the biology of reproduction in lower primates. Mammalia 26:1-88. Pochron S, Tucker W, and Wright P. 2004. Demography, life history, and social structure in Propithecus diadema edwardsi from 1986–2000 in Ranomafana National Park, Madagascar. American Journal of Physical Anthropology 125:61-72. Power M, Oftedal O, and Tardif S. 2002. Does the milk of callitrichid monkeys differ from that of larger anthropoids? American Journal of Primatology 56:117-127. Price E. 1991. Competition to carry infants in captive families of cotton-top tamarins (Saguinus oedipus). Behaviour 118:66-88. Pryce CR. 1995. Determinants of motherhood in human and nonhuman primates. Motherhood in Human and Nonhuman Primates. Basel: Karger. p 1-15. Purvis A, Gittleman J, Cowlishaw G, and Mace G. 2000. Predicting extinction risk in declining species. Proceedings of the Royal Society B 267:1947-1952. Quinlan R. 2007. Human parental effort and environmental risk Proceedings of the Royal Society B 274:121-125. Quinlan R, Quilan MB, and Flinn MV. 2003. Parental investment and age at weaning in a Caribbean village. Evolution and Human Behavior 24:1-16. Rangel-Negrín A, Alfaro JL, Valdez RA, Romano MC, and Serio-Silva JC. 2009. Stress in Yucatan spider monkeys: effects of environmental conditions on fecal cortisol levels in wild and captive populations. Animal Conservation 12:496-502. Ravosa M, Meyers D, and Glander K. 1993. Relative growth of the limbs and trunk in sifakas: heterochronic, ecological, and functional considerations. American Journal of Physical Anthropology 92:499-520. Reinhold K. 2002. Maternal effects and the evolution of behavioural and morphological characters: a literature review indicates importance of extended maternal care. Journal of Heredity 93:400-405. Rendigs A, Radespiel U, Wrogemann D, and Zimmermann E. 2003. Relationship between microhabitat structure and distribution of mouse lemurs (Microcebus spp.) in northwestern Madagascar. International Journal of Primatology 24:47-64. Richard A. 1974. Intra-specific variation in the social organization and ecology of Propithecus verreauxi. Folia Primatologica 22:178-207. Richard A. 1976. Preliminary observations on the birth and development of Propithecus verreauxi to the age of six months. Primates 17:357-366. Richard A. 1978. Behavioral variation: case study of a Malagasy prosimian: Lewisburg Bucknell University Press. Richard A. 1985. Social boundaries in a Malagasy prosimian, the sifaka (Propithecus verreauxi). International Journal of Primatology 6:553-568. Richard A. 1987. Malagasy prosimians: female dominance. In: Smut B, Cheney D, Seyfarth R, Wrangham R, and Struhsaker T, editors. Primate Societies. Chicago: University of Chicago Press. p 25-33. Richard A. 1991. Aggressive competition between males, female-controlled polygyny and sexual monomorphism in a Malagasy primate, Propithecus verreauxi. Journal of Human Evolution 22:395-406. Richard A, and Dewar R. 1991. Lemur ecology. Annual Review of Ecology and Systematics 22:145-175. Richard A, Dewar R, Schwartz M, and Ratsirarson J. 2000. Mass change, environmental variability and female fertility in wild Propithecus verreauxi. Journal of Human Evolution 39:381-391. Richard A, Dewar R, Schwartz M, and Ratsirarson J. 2002. Life in the slow lane? Demography and life histories of male and female sifaka (Propithecus verreauxi verreauxi). The Zoological Society of London 256:421-436. Richard A, and Nicoll M. 1987. Female social dominance and basal metabolism in a Malagasy primate, Propithecus verreauxi. American Journal of Physical Anthropology 12:309-314.

Richard A, Rakotomanga P, and Schwartz M. 1991. Demography of Propithecus verreauxi at Beza Mahafaly, Madagascar: sex ratio, survival, and fertility, 1984-1988. American Journal of Physical Anthropology 84:307-322. Roberts RL, Jenkins KT, Lawler T, Wegner FH, Norcrossa JL, Bernhardsa DE, and Newmana JD. 2001. Prolactin levels are elevated after infant carrying in parentally inexperienced common marmosets. Physiology & Behavior 72:713-720. Roosenburg WM. 1996. Maternal condition and nest site choice: an alternative for the maintenance of environmental sex determination? American Zoologist 36:157-168. Rosenblum LA, and Andrews MW. 1994. Influences of environmental demand on maternal behavior and infant development. Acta Paediatr 397:57-63. Ross A, Porter L, Power M, and Sodaro V. 2010. Maternal care and infant development in Callimico goeldii and Callithrix jacchus. Primates 51:315-325. Ross C. 2001a. Life history, infant care strategies, and brain size in primates. In: Kappeler P, and Pereira M, editors. Primate Life Histories and Socioecology. Chicago: University of Chicago Press. p 266-284. Ross C. 2001b. Park or ride? The evolution of infant carrying in primates. International Journal of Primatology 22:749-771. Ross C, and MacLarnon A. 2000. The evolution of non-maternal care in anthropoid primates: a test of the hypotheses. Folia Primatologica 71:93-113. Rossiter M. 1994. Maternal effects hypothesis of herbivore outbreak. Bioscience 44:752-763. Rossiter M. 1996. Incidence and consequences of inherited environmental effects. Annual Review of Ecology and Systematics 27:451-476. Rossiter M. 1998. The role of environmental variation in parental effects expression. In: Mousseau T, and Fox C, editors. Maternal Effects As Adaptations. Oxford: Oxford University Press. p 112-134. Saltzman W, and Abbott D. 2009. Effects of elevated circulating cortisol concentrations on maternal behavior in common marmoset monkeys (Callithrix jacchus). Psychoneuroendocrinology 34:1222-1234. Sánchez S, Peláez F, Gil-Bürmann C, and Kaumanns W. 1999. Costs of infant-carrying in cotton-top tamarins (Saguinus oedipus). American Journal of Primatology 48:99-111. Sauther M. 1991. Reproductive behavior of free-ranging Lemur catta at Beza Mahafaly Special Reserve, Madagascar. American Journal of Physical Anthropology 84:463-477. Sauther M. 1994. Wild plant use by pregnant and lactating ringtailed lemurs, with implications for early hominid foraging. Eating on the Wild Side Tucson: University of Arizona Press. p 240-256. Sauther M, and Cuozzo F. 2009. The impact of fallback foods on wild ring-tailed lemur biology: a comparison of intact and anthropogenically disturbed habitats. American Journal of Physical Anthropology 140:671-686. Schradin C, and Anzenberger G. 2001a. Costs of infant carrying in common marmosets, Callithrix jacchus: an experimental analysis Animal Behaviour 62:289-295. Schradin C, and Anzenberger G. 2001b. Infant carrying in family groups of Goeldi's monkeys (Callimico goeldii). American Journal of Primatology 53:57-67. Sellen D. 2009. Evolution of human lactation and complementary feeding: implications for understanding contemporary cross-cultural variation In: Goldberg G, Prentice A, Prentice A, Filteau S, and Simondon K, editors. Breast-Feeding: Early Influences on Later Health. Dordrecht: Springer Science. Setchell J, Smith T, Wickings E, and Knapp L. 2008. Factors affecting fecal glucocorticoid levels in semi-free- ranging female mandrills (Mandrillus sphinx). American Journal of Primatology 70:1023-1032. Silk J. 1984. Local resource competition and the evolution of male-biased sex ratios. Journal of Theoretical Biology 108:203-213. Silk J, and Brown G. 2008. Local resource competition and local resource enhancement shape primate birth sex ratios. Proceedings of the Royal Society of London B 275:1761-1765. Smith A. 1997. Deforestation, fragmentation, and reserve design in western Madagascar. In: Lawrence W, and Bierregaard O, editors. Tropical Forest Remnants, Ecology, Management, and Conservation of Fragmented Communities Chicago: University of Chicago Press. p 415-441. Smith R, and Cheverud J. 2002. Scaling of sexual dimorphism in body mass: a phylogenetic analysis of Rensch's rule in primates International Journal of Primatology 23:1095-1135. Soholt L. 1977. Consumption of herbaceous vegetation and water during reproduction and development of Merriam's kangeroo rat, Dipodomys merriami. The American Midland Naturalist Journal 98:445-457. Stearns SC. 1976. Life-history tactics: a review of the ideas. The Quarterly Review of Biology 51:3-47. Stearns SC. 1977. The evolution of life history traits: a critique of the theory and a review of the data. Annual Review of Ecology and Systematics 8:145-171. Stearns SC. 1992. Evolutionary Explanations The Evolution of Life Histories New York: Oxford University Press.

Sussman RW. 1977. Socialization, social structure, and ecology of two sympatric species of Lemur In: Chevalier- Skolnikoff S, and Poirier F, editors. Primate bio-social development: Biological, social, and ecological determinants New York: Garland p 515-528. Tardif S. 1994. Relative energetic cost of infant care in small-bodied neotropical primates and its relation to infant- care patterns. American Journal of Primatology 34:133-143. Tardif S, and Bales K. 1997. Is infant-carrying a courtship strategy in callitrichid primates? Animal Behaviour 53:1001-1007. Tardif S, and Bales K. 2004. Relations among birth condition, maternal condition, and postnatal growth in captive common marmoset monkeys (Callithrix jacchus). American Journal of Primatology 62:83-94. Tardif S, Carson R, and Gangaware B. 1986. Comparison of infant care in family groups of the common marmoset (Callithrix jacchus) and the cotton-top tamarin (Saguinus oedipus). American Journal of Primatology 11:103-110. Tardif S, Harrison M, and Simek M. 1993. Communal care in marmosets and tamarins: relation to energetics, ecology, and social organization. In: Rylands A, editor. Marmosets and Tamarins: Systematics, Behaviour, and Ecology. Oxford: Oxford University Press. Tardif S, Power M, Oftedal O, Power R, and Layne D. 2001. Lactation, maternal behavior and infant growth in common marmoset monkeys (Callithrix jacchus): effects of maternal size and litter size. Behavioral Ecology and Sociobiology 51:17-25. Tattersall I. 1982. The Primates of Madagascar New York: Colombia University Press Tecot S. 2010. It's all in the timing: birth seasonality and infant survival in Eulemur rubriventer. International Journal of Primatology 31:715-735. Tecot S, Baden A, Romine N, and Kamilar J. 2013. Reproductive strategies and infant care in the Malagasy primates. In: Clancy KB, Hinde K, and Rutherford J, editors. Building Babies. New York: Springer Science+Business Media. Tilden C, and Oftedal O. 1997. Milk composition reflects patterns of maternal care in prosimian primates. American Journal of Primatology 41:195-211. Trivers RL. 1974. Parent-offspring conflict. American Zoologist 14:249-264. Trivers RL, and Willard D. 1973. Natural selection of parental ability to vary the sex of offspring. Science 179:90- 92. van Schaik C, and van Noordwijk MA. 1985. Interannual variability in fruit abundance and the reproductive seasonality in Sumatran long-tailed macaques (Macaca fascicularis). Journal of Zoology, London 206:533- 549. Vasey N. 2002. Niche separation in Varecia variegata rubra and Eulemur fulvus albifrons: II. Intraspecific patterns. American Journal of Physical Anthropology 118:169-183. Ventura R, and Buchanan-Smith H. 2003. Physical environmental effects on infant care and development in captive Callithrix jacchus. International Journal of Primatology 24:399-413. Vick LG, and Conley JM. 1976. An ethogram for Lemur fulvus. Primates 17:125-144. Wade M. 1998. The evolutionary genetics of maternal effects. In: Mousseau TA, and Fox CW, editors. Maternal Effects as Adaptations. Oxford Oxford University Press p5-21. West-Eberhard M. 1989. Phenotypic plasticity and the origins of diversity. Annual Review of Ecology and Systematics 20:249-278. Willham R. 1972. The role of maternal effects in animal breeding: III. biometrical aspects of maternal effects in mammals. Journal of Animal Science 35:1288-1293. Wright P. 1999. Lemur traits and Madagascar ecology: coping with an island environment. Yearbook of Physical Anthropology 42:31-72. Young A, Richard A, and Aiello L. 1990. Female dominance and maternal investment in strepsirhine primates. American Naturalist 135:473-488. Zahed S, Prudom S, Snowdon C, and Ziegler T. 2008. Male parenting and response to infant stimuli in the common marmoset (Callithrix jacchus). American Journal of Primatology 70:84-92. Ziegler T. 2000. Hormones associated with non-maternal infant care: a review of mammalian and avian studies. Folia Primatologica 71:6-21.

Table 2.1. P. coquereli group size and composition including infants season 1 (June-December 2010) Group Group size Adult males (n=14) Non-mother adult females (n=6) Citron 4 2 0 Fito 5 1 2 Kambana 6 2 2 Mainty 4 2 0 Vaovao 4 1 1 Volo 6 3 1 Zaza 5 3 0

Table 2.2. P. coquereli group size and composition including infants season 2 (June-December 2011) Group Group size Adult males (n=5) Non-mother adult females (n=2) Iva 4 1 1 Rambo 4 1 1 Zaza 5 3 0

Table 2.3. P. coquereli focal data collection columns Date Time Group Start time End time Actor State Receiver Infant carry start Infant carry stop Infant position Contact initiate Contact broke

Table 2.4. P. coquereli ethogram Modified from (°)(Vick and Conley 1976); (*) (Hanson and Sodaro 2002); ()(Eaglen and Boskoff 1978); ()(Richard 1976); ()(Tardif et al. 1986; Ventura and Buchanan-Smith 2003); (‡)(Hinde and Atkinson 1970) State behavior Action Event Aggression (AG) °Intergroup aggression (GA): actor lunges, bites, cuffs, or pushes another individual without attempting to access infant

Infant access (AA): actor (carrier) refuses infant access by lunging, biting, or cuffing prospective carrier

*Infant rid (IR): carrier attempts to dislodge infant by biting, rubbing it against substrate, or pulling on its limbs. May or may not result in broken contact Feed (FE) Leaves (LE): actor forages or consumes young, mature leaves, and/or leaf buds Fruit (FR): actor forages or consumes fruit

Flowers (FO): actor forages or consumes flowers and/or flower buds

Bark (BA): actor forages or consumes dead or alive bark

Solid food manipulation (SM): infant manipulates solid food in hands or puts item in mouth

() Non-discriminate (ND): food item not consumed simultaneously by conspecifics

() Discriminate (DI): food item consumed by conspecific adults simultaneously

State behavior Action Event Groom (GR) Actor repeatedly touches, combs, or licks another actor Rest (RE) Actor is inactive and is not engaged in feeding, travelling, or grooming Travel (TR) Actor locomotes from one substrate (e.g., tree, liana, forest floor) to another substrate Out of sight (OOS) Actor not visible *Infant carrier (IC) Actor supports a minimum of 50% of infant’s body weight Ventral (V) Infant clings to actor with its head facing either nipple

Dorsal (D) Infant clings to the back of the actor

Independent Infant is not clinging to a carrier with (IND) any limbs and its weight is unsupported by a carrier ‡Contact initiate Actor initiates infant carriage by moving toward the infant, (CI+actor) grooming, pulling, or picking up the infant resulting in carriage

‡Contact broken Actor attempts to break infant contact with an infant rid (CB+actor) (see ‘aggression’ state behavior for description) ‡Infant contact Independent infant initiates contact with actor initiate (CI+IN) by climbing or jumping on the actor or leaves one carrier for another, resulting in ventral or dorsal carriage

‡Infant contact Infant willingly (without infant rid) breaks contact broken (CB+IN) with carrier to move independently or be carried

Table 2.5. P. coquereli behavioral focal hours Group Mother Infant Citron 36 36 Fito 33 30 Iva 33 33 Kambana 36 30 Mainty 39 36 Rambo 39 36 Vaovao 24 21 Volo 27 36 Zaza (s1) 39 36 Zaza (s2) 39 39 Total 345 333

Table 2.6. P. coquereli infant birth months Group May June August Citron X Fito X Iva X Kambana X Mainty X Rambo X Vaovao X Volo X Zaza (s1) X Zaza (s2) X

Table 2.7. Mixed effects linear regression comparison of infant position (n=10) Least squares means Position Time SE Pr>|t|

Ventral 4.8690 0.06120 <.0001 Dorsal 4.2641 0.06152 <.0001

Independent 3.9655 0.06158 <.0001

Pairwise comparison of least Position effect Time SE Pr>|t| squares means Ventral versus dorsal -0.6049 0.02739 <.0001

Ventral versus -0.9036 0.02748 <.0001 independent Dorsal versus independent -0.2986 0.02816 <.0001

Table 2.8. Mixed effects linear regression comparison of infant carrier identity (n=10 mothers, n=10 infants, n=19 adult males, n=8 adult females) Least squares means Carrier Time SE Pr>|t| Non-mother 3.9392 0.06243 <.0001 Mother 4.6844 0.06151 <.0001 Pairwise comparison of Carrier effect Time SE Pr>|t| least squares means Non-mother versus mother -0.7452 0.02300 <.0001

Table 2.9. Poisson regression considering number of occurrences mothers initiated and broke infant contact (n=10 mothers, n=10 infants) Poisson regression Contact effect Time SE Pr>|t| Infant contact -0.3244 0.11480 0.0047 Weeks -0.0182 0.02370 0.4432 Intercept 1.5481 0.23830 <.0001

Table 2.10. Poisson regression considering number of occurrences infants initiated and broke mother contact (n=10 mothers, n=10 infants) Poisson regression Contact effect Time SE Pr>|t| Infant contact 0.0082 0.0002 <.0001 Weeks 0.0230 0.0034 <.0001 Intercept 3.1204 0.0693 <.0001

Figure 2.1. Ankarafantsika National Park, northwestern Madagascar

Figure 2.2. Percentage of time P. coquereli infants spent ventrally on all carriers (n=37)

Figure 2.3. Percentage of time P. coquereli infants spent dorsally on all carriers (n=37)

Figure 2.4. Percentage of time P. coquereli infants spent independently (n=10)

Figure 2.5. Percentage of time P. coquereli infants spent ventrally, dorsally, and independently (n=10 infants, n=37 carriers)

Figure 2.6. Percentage of time P. coquereli infants spent carried by mothers (n=10)

Figure 2.7. Percentage of time P. coquereli infants spent carried by adult males (n=19)

Figure 2.8. Percentage of time P. coquereli infants spent carried by adult females (n=8)

Figure 2.9. Percentage of time P. coquereli infants were transported by mothers (n=10), adult males (n=19), adult females (n=8), and independent (n=10)

Figure 2.10. Occurrences of infant contact initiated by P. coquereli mothers (n=10)

Figure 2.11. Occurrences of infant contact broken by P. coquereli mothers (n=10)

Figure 2.12. Occurrences of mother contact initiated by P. coquereli infants (n=10)

Figure 2.13. Occurrences of mother contact broken by P. coquereli infants (n=10)

Figure 2.14. Occurrences of infant contact initiated/broken by P. coquereli mothers and mother contact initiated/broken by infants (n=10)

Chapter 3 Nutritional Food Quality of Foods Exclusively Selected by Propithecus coquereli Lactating Females "Planting a tree when you are young, cooling under its shadow when you get old." “Tanora mamboly hazo, manan kihalofana rehefa antitra.” ~ Malagasy proverb

3.1. Abstract

The nutritional quality of wild primate foods can be exceedingly variable and may directly affect maternal and infant health during the energetically constraining lactation period. The demand for certain nutrients increases during lactation and females must meet these demands to nutritionally sustain themselves and dependent infants by prioritizing their nutritional parameters. Optimal foraging theory states that an animal will attempt to gain the greatest energetic benefit for the lowest energetic cost while foraging to maximize fitness. Extreme seasonality in Madagascar causes resource abundance and distribution to be highly unpredictable. Coquerel’s sifaka (Propithecus coquereli) females give birth during the lean season in the austral winter and infants are weaned during the subsequent wet season. Other field studies have found preferred foods are scarce during the austral winter, causing some lemur species to consume less abundant and lower quality foods. I collected foods selected by lactating P. coquereli (n=10) in Ankarafantsika National Park, northwestern Madagascar who were followed once per week for 93 focal hours on 31 calendar days over two consecutive birth seasons (2010 and 2011). I assayed the samples for nitrogen, neutral detergent fiber (NDF), acid detergent fiber (ADF), gross energy (n=123), and ash (n=119) to measure protein (calculated from nitrogen), fiber, energy, and mineral content. I calculated crude protein, available protein, and non-protein energy. I applied the nitrogen maximization, fiber limitations, and energy maximization models to foods consumed by lactating P. coquereli to establish nutritional quality and determine the biological significance of these foods. Lactating P. coquereli most frequently selected foods high in non-protein energy (calculated from the gross energy assay), followed by

87 88

foods high in available protein and lastly high fiber foods. Consuming foods high in non-protein energy gives lactating females increased access to energy by providing greater concentrations of metabolic fuel. Consuming higher energy density foods is most seasonally advantageous during the austral winter when food is the least abundant and females are lactating and encumbered with dependent infants. Protein requirements increase during lactation, thereby consuming foods high in available protein helps females nutritionally compensate for increased protein demand. Foods high in non-protein energy and available protein were relatively low in fiber, but curiously females still selected high fiber foods. Consuming high fiber foods may promote digestive efficiency and digestive health by providing fermentable substrate for the microbiota in the extensive lower gut found in Propithecus spp., or be a residual effect of consuming the foods seasonally available in the forest during the lean season. Understanding the seasonally distinct nutritional repertoire in lactating P. coquereli will quantify nutritional goals during this critical life history stage for both mothers and their infants, and is an important component to conservation planning for forest replanting in northwestern Madagascar.

3.2. Introduction

Animal bodies are composed of water and organic substances, minerals, blood, muscle and tissue that all are influenced by food nutrient composition (Maynard and Loosli 1965a). The relationships between food nutrient composition and an individual animal are dynamic and multi-dimensional since every component of an animal body is affected by the nutrients selected, consumed, digested, and subsequently absorbed. The three basic macronutrient categories are protein, fat, and carbohydrate. Micronutrients are needed in lesser quantities and include many different vitamins and minerals. “Nutrition involves various chemical and physiological activities which transform food elements into body elements” (Maynard and Loosli, p. 11). One effective way to examine the transformative interrelationship between food and body elements is to examine the nutritional quality of foods consumed by individual animals. Nutritional quality is defined as the number of units (i.e., k/cal) of a macronutrient or micronutrient present in a food and/or the number of units an animal receives from a food item (Maynard and Loosli 1965a). The quantities of macro and micronutrients required for nutritional success is both species and individual specific and dependent on body size, metabolism, digestive anatomy and physiology, and life history (Lambert 1998; Milton 1983).

Optimal foraging theory postulates that animals will develop foraging strategies to achieve a balance between energy consumed and energy expended (Stephens and Krebs 1986). This theory is deeply rooted in primate socioecological models (Sterck et al. 1997; Wrangham 1980) and continues to be applied widely across primatological field studies. Primate nutritional ecology is the study of “the interactions between the environment and a primate’s nutrient intake, and the individual’s resultant physiological state” (Felton et al. 2009, p. 70). Nutritional parameters are fluid and vary seasonally, thereby requiring primates to shift ecological or behavioral strategies to meet these goals. To achieve this, primates must seasonally identify, select, forage, consume, and digest particular foods. Thus, primates must prioritize nutritional parameters to meet their nutritional goals (Felton et al. 2009). Quantifying nutritional quality and the nutrient quantities available metabolically are exceptionally valuable measures in determining which nutritional parameters are accentuated during seasonal events, for example; lactation.

The enlarged lower gut characteristic of hindgut fermenters is composed of the caecum, a portion of the large intestine, and colon. The lower gut serves as a fermentation chamber where large populations of microbes are housed in the caecum. Digestion initially occurs in the stomach, followed by the caecum, and further acted on by microorganisms. From there, digesta is returned to the large intestine where fermentation and absorption continues. Indriids have enlarged caecums which aid in fiber digestion (Lambert 1998). Microbial fermentation breaks down celluloses and hemicelluloses into volatile fatty acids that are either digested or absorbed. Microbes found in the caecum are capable of fermenting fiber, in turn producing energy for indriids in the form of short-chain volatile fatty acids (primarily acetate, butyrate, and propionate), amino acids, vitamins, and a host of other bioactive molecules that may benefit the host (see Lin et al. 2012). Short-chain fatty acids are metabolites formed from complex carbohydrates formed by gut microbiota (Lin et al. 2012).

Five nutritional ecology models that all emphasize diet selection based on the acquisition of different nutrients have been applied to determine how primates achieve nutritional goals and include: nitrogen (as a proxy for protein) maximization (Mattson 1980), fiber limitations (Milton 1979), energy maximization (Schoener 1971), nutrient balancing (Raubenheimer and Simpson 2004), and the regulation of plant secondary metabolites (Freeland and Janzen 1974). The

nitrogen maximization model proposes that metabolizable nitrogen is a limited and difficult resource to acquire, particularly for folivores, due to shortages of amino acids and specific proteins (Mattson 1980). Lactation further increases the demand for amino acids needed to energetically sustain females (Jessop 1997). By extension, it is also difficult for these same animals to acquire essential minerals. Thereby it is advantageous for lactating females to consume food higher in minerals (e.g., potassium, sodium). Nitrogen is essential to metabolic processes and improves overall health by facilitating growth and reproduction (Mattson 1980). Thus, many herbivores evolved adaptations to maximize ambient nitrogen levels for protein synthesis (Mattson 1980). The nutritional costs of reproduction and growth determine nutrient intake in G. beringei, where adult female and juvenile mountain gorillas (Gorilla beringei) consume more dry matter and ingested more protein per kilogram of metabolic body mass relative to silverback males (Rothman et al. 2008).

The fiber limitations model proposes that a relatively high protein, low fiber diet is preferred in small-bodied folivores to capitalize on the metabolic and reproductive benefits of protein while minimizing difficult to digest fiber (Milton 1979). Dietary fiber is composed of structural carbohydrates (cellulose, hemicellulose), and lignin; which are all indigestible compounds in vertebrates (Moir 1965). Bacteria in the gastrointestinal tract have enzymes that assist with the detoxification of secondary plant compounds (reviewed in Milton 1979). The fiber limitations model postulates hindgut fermenters may prefer to eat less fibrous plants and plant parts (Milton 1979). The fiber limitations model proposes that examining specific animals in specific habitats using cost-benefit analysis of foraging behavior is the most productive approach given the multitude of factors influencing food selection (Milton 1979). Mantled howler monkeys (Alouatta palliata) do not have specialized gut morphology and spent 82% of their time consuming young versus mature leaves, with protein and fiber content as the primary determinants of leaf selection (Milton 1977). Mature leaves are often higher in fiber and protein- to-fiber ratios are a predictor of leaf selection in smaller bodied herbivores (Milton 1979).

The energy maximization model holds that feeding strategies are based on acquiring food energy to optimize diet, foraging space, foraging time, and group size (Schoener 1971). Achieving optimal conditions is dependent on an animal’s determination of the feeding variables needing to be maximized or minimized (i.e., time and energy expenditures required for food consumption),

their costs-benefits, and subsequently creating mathematical equations to quantify these relationships (Schoener 1971). Frugivorous Peruvian spider monkeys (Ateles chamek) preferentially consume fruits high in lipids and soluble carbohydrates (Di Fiore et al. 2008) and as a result, energy has been hypothesized as the food derivative driving food selection in frugivorous atelines (Di Fiore and Rodman 2001).

Ecological stoichiometry recognizes that the fundamental unit in ecological processes is species- specific regulatory physiology, wherein a multivariate approach to ecosystem exchanges is employed that considers energy flow and nutrient balancing versus traditional univariate models that primarily focus on either nitrogen or energy acquisition (Raubenheimer and Simpson 2004). A tenet of ecological stoichiometry is that nutrient exchanges between organisms and their environments, or nutrient budgets, are central to understanding an animal’s nutritional goals (reviewed in Raubenheimer and Simpson 2004). Nutrient budgets are mass-balance equations, where intake is partitioned into either retained (i.e., growth) or dissociated (i.e., lost in feces) elements (Raubenheimer and Simpson 2004). The geometric framework measures foods along a vector to determine its nutritional balance and was partially formulated from the principles of ecological stoichiometry (reviewed in Raubenheimer and Simpson 2004). In the geometric framework, “an organism’s nutritional relations are modeled as an n-dimensional nutrient space where each dimension represents a nutrient” (Raubenheimer and Simpson 2004, p. 1205). The nutrient balancing model suggests, “An animal’s current nutritional state with respect to the ingestion of relevant nutrients can be described as a point within the nutrient space, as can the state that would maximize fitness” (Raubenheimer and Simpson 2004, p. 1205). A food is deemed nutritionally balanced when it passes through the target intake, where fitness is maximized when activity, growth, or reproduction are improved (reviewed in Raubenheimer and Simpson 2004). A nutritionally imbalanced food does not meet optimal requirements for all nutrients simultaneously; instead certain foods are over-ingested while others are under-ingested (Raubenheimer and Simpson 2004).

Plants defend themselves from folivorous animals by producing defensive chemicals referred to as secondary compounds and animals must regulate their consumption to avoid untenable toxicity levels (Freeland and Janzen 1974). Varying concentrations of toxic substances are found in plants and as such, dosage effects are also variable dependent on the amount of secondary

compounds ingested by the animal (Freeland and Janzen 1974). White leadtree (Leucaena leucocephala) contains a toxic amino acid that instigates responses ranging from anestrous in lower concentrations to infertility when higher concentrations were consumed by female Sprague-Dawley rats (Hylin and Lichton 1965). Animals avoid secondary plant compounds by consuming plants or selecting the plant parts with fewer secondary compounds (Freeland and Janzen 1974). Thus, regulation of the secondary plant compounds model proposes that animals actively choose food resources based on cognitive capacity (Freeland and Janzen 1974). Detoxification of secondary plant compounds by microsomal enzymes in the endoplasmic reticula and gastrointestinal tract are physiological mechanisms used by animals to combat plant toxicity, and vary interindividually by body size, sex, age, and reproductive state (Freeland and Janzen 1974). Synthesizing chemicals or consuming foods (e.g., high in protein) that assist in neutralizing toxic compounds (e.g., decrease tannic acid toxicity) often diminishes the deleterious effects of consuming secondary plant compounds (Freeland and Janzen 1974). Nutrient transfer from mother to offspring begins in utero via the placenta in eutherian mammals and continues in the form of milk after birth. The placenta continues to affect infants postnatally: a classic example being low birth weight as an indicator of later ontogenetic, deleterious health issues (see Rutherford 2013). Placental efficiency is a term used to describe the neonatal weight divided by the placental weight, with placental efficiency typically increasing as pregnancy progresses (Myatt 2006). Maternal nutrition and behavior directly impacts placental growth, where nutrient restrictions have the ability to negatively influence placental efficiency depending on the timing and duration of nutrient restrictions (Rutherford 2013). For example, placental efficiency in vervet monkeys (Chlorocebus sabaeus) increases almost 2.5 times from mid- gestation until the last two weeks of gestation (Rutherford et al. 2010). Consequently, the nutritional quality of foods ingested by mothers during gestation and lactation directly affects maternal and infant health.

Lactation is the most energetically expensive activity in which female primates participate, with daily energy expenditures increasing up to 150% in mammals and mean caloric intake increasing 66-188% (reviewed in Gittleman and Thomspon 1988; Lee et al. 1991; Oftedal 1984). Protein is the primary nutrient involved in reproductive function and protein requirements increase by more than 1/3 in humans during early lactation (Cameron 1996; Oftedal 1991; Tilden and Oftedal 1997). Thus, there is a shift in nutritional parameters for lactating females given these amplified

energetic requirements. Primates require a minimum of 14% protein per dry matter basis for reproduction and between 7-11% for growth and development (Oftedal 1991). These values are primarily based on captive animals. There remains a lack of complementary data on wild primates. Leaves consumed by many primates are composed of 12-16% protein (Glander 1982). Thus, protein acquisition likely is not problematic for most herbivorous mammals aside from lactating females consuming leaves with high tannin content that can inhibit digestive uptake (Oftedal 1991). An increase in protein consumption has been documented in several lemur species during gestation and lactation (in Lemur catta, Gould et al. 2011, Sauther 1994; in Propithecus tattersalli, Meyers and Wright 1993; in Varecia rubra, Vasey 2005).

The nutritional quality of foods is highly variable and resources are not always interchangeable (see Conklin and Wrangham 1994; Gould et al. 2011; Simmen et al. 2013). This variability occurs both inter and intraspecifically and is even present in the same resource in a single tree. For example, there are significant differences between the nutrient composition of pulp and seed fractions in Ficus spp. consumed by chimpanzees (Pan troglodytes) (Conklin and Wrangham 1994). Neutral detergent fiber (NDF) is higher in seeds relative to pulp and interspecies variation is exceedingly high in NDF (23.5-65.4%) and crude protein (4.3-20.7%) (Conklin and Wrangham 1994). The protein-to-fiber ratio is often used to predict leaf choice in folivores, which predicts that foliage with high protein-to-fiber ratios will be the most preferred resource in primates (Milton 1979). Young leaves consumed by colobines were found to have more overall protein, were consumed more frequently than mature leaves, had higher protein-to-fiber ratios, and were more digestible (Chapman et al. 2004).

Trees in edge habitats often receive increased sunlight exposure relative to trees in the forest interior and, in turn, leaves located near habitat edges have greater protein-to-fiber ratios (Ganzhorn 1995a). This does not necessarily have a positive adaptive effects on animal or plant species since forest edges are less stable than the interior as they are more exposed to climatic oscillations like wind speed and increased temperature (Laurance et al. 2002). Significant rises in tree mortality along edges result from climatic and structural changes exceeding tree physiological thresholds (Ferreira and Laurance 1997; Oliveria et al. 1997). Canopy-gap dynamics are transformed and regenerating trees near the edge will favor secondary species relative to old-species as tree mortality increases (Laurance et al. 1998). No edge effects were

found on plant chemicals (McGoogan 2011), whereas a strong edge effect exclusively for nitrogen isotopes was present in Ankarafantsika National Park (ANP) (Crowley et al. 2012).

Madagascar is characterized by irregular rainfall patterns and extreme seasonality, which has led to the evolution of unique tree phenology and unpredictability in the abundance and distribution of food resources (Dewar and Richard 2007; Dunham et al. 2011). The majority of lemur species including P. coquereli give birth during the dry season when resources are of relatively low quality and wean infants during the wet season when resources are higher quality, thus allowing weanlings to benefit from increased food quality and availability as they achieve nutritive independence from their mothers (Jolly 1984; Wright 1999). This strategy even further increases energetic demands on lactating lemurs since females are behaviorally and nutritionally caring for infants during the most seasonally depletive time for resources. For example, P. verreauxi females lose an astounding 18% of their body mass while lactating during the dry season (Lewis and Kappeler 2005b). My dissertation tests the hypothesis that P. coquereli mothers experience greater postnatal reproductive stress, measured by behavioral care-giving effort, nutritional food quality, and cortisol stress responses, during early/earlier-mid lactation (designated as 1-12 weeks postnatal) (May-August) in comparison to later-mid/late lactation (designated as 13-26 weeks postnatal) (September-December) due to the increased energetic costs of lactation while simultaneously caring for dependent infants during the driest seasonal months in the austral winter when food quality is low. I examined the nutritional food quality and biological significance of foods selected by lactating P. coquereli with the application of the nitrogen maximization (Mattson 1980), fiber limitations (Milton 1979), and energy maximization (Schoener 1971) models contextualized within the theoretical framework of optimal foraging theory (Stephens and Krebs 1986). Protein, fiber, energy, and mineral food content were used as the units of measure to determine nutritional food quality and establish nutrient profiles during lactation. Given that food quality affects the health of an animal (see Conklin and Wrangham 1994), quantifying the nutrient composition of foods and subsequently, translating these values into what it means biologically for an individual animal, helps to determine why primates select particular foods in order to nutritionally sustain themselves and, for lactating females, their dependent offspring. Nitrogen maximization model P1. P. coquereli lactating females will select foods high in available protein.

P2. P. coquereli lactating females will select foods high in minerals. Fiber limitation model P3. P. coquereli lactating females will select foods low in fiber. Energy maximization model P4. P. coquereli lactating females will select foods high in non-protein energy.

3.3. Methods 3.3.1. Study site and species I conducted this study at the Ampijoroa Forestry Station in ANP located in the Mahajanga Province in northwestern Madagascar (Figure 3.1). The Ankarafantsika region (135,000 ha) was first established as two protected areas in 1927 and recognized as a national park in 2003. Ampijoroa is situated in the southwestern portion of ANP. The Marovoay area (38,000 ha) found within the Ankarafantsika region is the second largest producer of rice, the primary subsistence crop, in Madagascar (Alonso and Hannah 2002). The GPS coordinates of the research base camp are 16°18’31” South, 46°48’49” East and it sits 88m above sea level. ANP is characterized as a dry deciduous forest with an exceptionally pronounced dry season (Alonso and Hannah 2002; Du Puy and Moat 1996). Forested areas surrounding Ampijoroa are experiencing anthropogenic disturbance from slash and burn agriculture, fire, relatively high volumes of human traffic, unregulated presence and herding of domestic cattle, bushmeat hunting, and hole digging for Dioscorea maciba tuber extraction (Alonso and Hannah 2002; Crowley et al. 2012; Gerardo and Goodman 2003). The underlying geological formation in the Ankarafantsika region is composed of sandstones and the study area sits atop a sandstone plateau between 310-340 m above sea level (Du Puy and Moat 1996; Lourenço and Goodman 2006). Soils are either red, speckled or white, with red soil containing the highest water content and white sand the lowest (Crowley et al. 2012). Red soil most often occurs in the forest edge and in the savannah itself, though is also present in the forest interior and is presumably more nutrient dense due to its higher water content than forest interior quartz white sands (Crowley et al. 2012). Many tree species grow in nutrient poor, acidic white sands and a thick layer of loose sand is present on the soil surface as a result of sandstone erosion (Du Puy and Moat 1996; Lourenço and Goodman 2006). Flora are speciose and the forest understory is moderately thick with sparse leaf litter (Lourenço and Goodman 2006) Annual precipitation in ANP ranges from

1100-1600mm, with the majority of rainfall occurring in January and February and a period of extreme desiccation from May to September in which there is very little rainfall (Rendigs et al. 2003). Average daily temperatures range from 16°C during the dry season to 37°C in the wet season (Rendigs et al. 2003). Vertebrates have varied adaptive responses to cope with the austral winter including torpor or hibernation and flora are resilient to desiccation (Lourenço and Goodman 2006; Rendigs et al. 2003).

There are eight extant lemur species in ANP including: Coquerel’s sifaka (Propithecus coquereli), common brown lemur (Eulemur fulvus), mongoose lemur (Eulemur mongoz), western woolly lemur (Avahi occidentalis), lesser dwarf lemur (Cheirogaleus medius), Milne-Edwards' sportive lemur (Lepilemur edwardsi), Gray mouse lemur (Microcebus murinus), and the golden- brown mouse lemur (Microcebus ravelobensis) (Alonso and Hannah 2002). The International Union for Conservation of Nature estimates that the P. coquereli has experience rapid population declines of more than 50% during the last 30 years primarily due to habitat loss (Andrainarivo 2008, accessed 17 November 2013) and more recently, hunting for human consumption (Gerardo and Goodman 2003). The slow life histories exemplified by Propithecus spp. increases extinction risk (Purvis et al. 2000) and amplifies the devastating impact of hunting given Propithecus spp. wait until relatively later in life to reproduce, have interbirth intervals of every other year, and only one infant is present per social group (Pochron et al. 2004; Richard et al. 1991). An estimated population of ~47,000 P. coquereli currently live in ANP (Kun-Rodrigues et al. 2014). Density estimates range from 5-100 individuals/km2, with habitat quality (i.e., negative effects of roads and forest edges) as the principal factor for this high variability (Kun- Rodrigues et al. 2014). An estimated 5 individuals/km2 presently exist in Ampijoroa (Kun- Rodrigues et al. 2014), in comparison to an estimated 60-75 individuals/km2 in the 1980s (Albignac 1981). This is a rapid decrease of more than 90% of the P. coquereli population in Ampijoroa (Kun-Rodrigues et al. 2014) and is indicative of a population experiencing duress.

3.3.2. Botanical and biotic variable collection Data were collected for a total of 14 months between June-December 2010 and 2011 for 93 focal hours over 52 weeks (26 consecutive weeks/season). A previously established trail system in Jardin Botanique A (JBA) for research (see Rendigs et al. 2003 for a detailed site description) and the tourist trails identified as the “Coquereli circuit” were used to initially locate P. coquereli

groups. I, along with 1-2 research assistants, collected data on ten habituated P. coquereli groups. Each P. coquereli group was followed once per week using ten minute continuous focal sampling (Altmann 1974) for 6 hours beginning at dawn after groups were located in their sleeping trees. All trees and lianas from which P. coquereli mothers (n=10) fed from during the focal follow were marked with a GPS waypoint, identified with their vernacular name for subsequent Latin identification, and a botanical sample was collected immediately after P. coquereli groups left feeding locations.

Sample ID, social group the sample was collected from, the parts of the tree consumed by lactating females (e.g., fruit), diameter-at-breast-height (DBH), frequency of ingestion index (FOI), vernacular name, and scientific name were documented for each botanical sample (Table 3.1). I created the FOI index to measure relative food consumption by mothers where: number of occurrences food item was consumed during focal follow / total number of foods consumed during the focal follow. Botanical samples were individually stored in manila paper coin envelopes of various sizes depending on sample size until they were transported to a propane drying oven at the end of each focal follow. Herbarium specimens were collected for Latin name identifications collectively made by Parc Tsimbazaza, Ankarafantsika National Park, and Missouri Botanical Gardens. DBH was measured 1.4 m from the ground and rounded to the nearest 0.5 cm. DBH was calculated below the interference if a tree had branches near 1.4 m. If the tree spilt into several trunks, the DBH was calculated for each trunk separately then each measurement was multiplied by itself, summed, and square rooted to calculate the final DBH.

3.3.3. Botanical processing and preservation Samples were dried on-site in a propane oven at a maximum of 50°C using a water resistant max/min digital thermometer (HBE International Inc.) following (Chapman et al. 2004) until a constant weight was reached for a minimum of 48 hours. Botanical samples were weighed daily to determine dry weights and not exposed to direct sunlight during processing to limit post- collection changes in nutrient composition. Botanical samples were kept in their original collection envelopes and placed in 3M SCC Dri-Shield 2000 moisture barrier bags (65x50 cm) with silica gel. Dri-Shield moisture barrier bags were stored in plastic containers with silica gel in a concrete storage area.

3.3.4. Macronutrient and micronutrient assays Standard proximate macronutrient analyses were used to determine nutrient composition. Laboratory assays were conducted at the Nutrition Laboratory, Conservation Ecology Center, Smithsonian Conservation Biology Institute, in Washington D.C. from May to December 2012 under the direction of Dr. Michael Power. All assays followed the standard operating procedures drafted by the laboratory manager. Assays included: nitrogen (N) as an index for protein (n=123), neutral detergent fiber (NDF) (n=123), acid detergent fiber (ADF) (n=123), gross energy (GE) (kcal/g) (n=123), and ash as an index for total mineral content (n=119). We calculated crude protein (CP), available protein (AP), and non-protein gross energy (NPGE) from the assay values, where:

CP = % total protein1 AP = (total protein – [ADF*{6.25*N{ADF residue}])2 NPGE = (GE-5.86*total protein)3

The NDF, ADF, and ash procedures are gravimetric assays. Gravimetric analyses measure botanical sample weight prior to the assay and again after its completion to determine sample weight loss (M. Power, pers. comm.). This difference in mass was used to calculate the percentage of compound being assayed. I chose botanical samples based on the frequency of the food item as well as sample quality and dry weight. Nutritional sample ID, group ID, and parts selected are shown alongside available protein, NDF, ADF, nonprotein gross energy, and ash values (Table 3.2). Nitrogen determinations were assayed in triplicate (Jakubasz 2012c). ADF, NDF, gross energy, and ash determinations were assayed in duplicate (Jakubasz 2011; Jakubasz 2012a; Jakubasz 2012b; Jakubasz 2012d).

3.3.5. Laboratory drying and grinding The dry botanical samples collected in the field were re-dried at the laboratory at 55°C for a minimum of 48 hours and subsequently ground using a small or ED-5 intermediate Wiley Mill

1Jakubasz (2012c) 2Jakubasz (2012a) 3Jakubasz (2012b)

(Thomas Scientific, Swedesboro, NJ) to achieve a homogeneous subsample at The Smithsonian Conservation Biology Institute. Plant material passed either through a 0.38mm sieve (CHN procedure); 0.86mm sieve, or was ground with a ceramic mortar and pestle depending on consistency and sample size.

3.3.6. Crude protein determination Tin vials were filled with 3.0-7.0 mg of sample, re-dried at 60°C for 24 hours, folded using needle nosed micro forceps, and rolled using needle nosed curved forceps. Carbon (C), hydrogen (H), and nitrogen (N) content were measured using a combustion method to convert the sample to CO2, H2O, and N2 gases in a PerkinElmer 2400 Series II Analyzer (PerkinElmer, Waltham, MA). CP content was calculated from N content by multiplying by the N-CP conversion factor of 6.25g CP/g N (Maynard and Loosli 1965c). Samples were re-assayed if the % dry matter was less than 90% or greater than 104% to ensure accuracy.

3.3.7. Neutral detergent fiber and acid detergent fiber determination The ANKOM fiber procedure using an ANKOM Fiber 200 Analyzer or the Van Soest fiber procedure (Van Soest et al. 1991) were used for NDF and ADF determination, where:

NDF = hemicellulose + cellulose + lignin ADF = cellulose + lignin

The Van Soest procedure (Van Soest et al. 1991) was used exclusively on small quantity samples, as only a total of 0.4 g of sample was needed in comparison to the 2.0 g required for the ANKOM procedure. Six large quantity samples were assayed using both ANKOM and Van Soest procedures to detect any inter-procedure variability. No inter-procedure differences were present in NDF or ADF values. Samples were considered for re-assay if the standard deviation was greater than 2.0, or the coefficient of variation was greater than 2.5% between duplicates.

100

3.3.8. Total mineral (ash) Total mineral content of the samples was determined by ashing the samples in a muffle furnace. Crucibles were filled with 0.25-0.50g of sample and heated for six hours at 450°C.

3.3.9. Gross energy determination I also determined caloric content of samples. Caloric content (cal/g) was measured using adiabatic bomb calorimetry to measure the heat from sample combustion (Jakubasz 2012b). Pellets were formed from 0.25-0.75 g of sample and re-dried for one hour at 60°C. A Parr 1241 Adiabatic Calorimeter (Parr Instrument Company, Moline, IL) was used to measure the gross energy content (GE). Samples were considered for re-assay if duplicates varied by >200 cal/g.

3.3.10. Data analysis

All nutrient results are reported on a dry matter basis to control for the effect of water content. Principal component analysis (PCA) was conducted on the macronutrient values to reduce the number of parameters. Seven loadings were considered in the component score including: crude protein, available protein, gross energy, NDF, ADF, ash, and non-protein gross energy. PCA identified three axes (protein, fiber, non-protein gross energy) based on the loadings for each axis considered. Each food was then plotted by the values from the significant PCA axes. Next, a k- means cluster analysis (with k set to the number of PCA axes) determined if foods grouped together by nutrient composition by distributing the foods into the three PCA axes. K-means partitions objects into different clusters, wherein the objects contained in each cluster are more similar in comparison to the objects contained in another cluster (reviewed in Jain 2010; Legendre and Legendre 1998). Lastly, A one-way ANOVA on the nutrient parameters measured (e.g. crude protein, available protein etc.) with the k-means cluster groups set as the grouping variable was applied. Post hoc comparisons using a Bonferroni correction were performed to control for multiple comparisons and confirmed the ANOVA significance. All analyses were conducted in SPSS® Version 20.0.

101

3.4. Results 3.4.1. High available protein foods: Cluster 1 PCA identified three axes that explained more than 88% of nutrient variation. These axes were used to identify three distinct foods categories: protein, fiber, and non-protein energy. Cluster 1 (n=44) foods were highest in available protein (mean [x̅ ]=15.9%, standard error of the mean [SE]=0.6; Table 3.3). In contrast, high fiber foods (represented by cluster 2) were lowest in available protein (x̅ =4.3%, SE=0.5). Non-protein gross energy foods (represented by cluster 3) were intermediate, albeit relatively low, in available protein (x̅ =7.7%, SE=0.4).

3.4.2. High fiber foods: Cluster 2

Cluster 2 (n=33) foods were highest in NDF (x̅ =59.0%, SE=2.4) (Table 3.4). High available protein foods (cluster 1) were lowest in NDF (x̅ =29.5%, SE=1.5). This result was similar to high non-protein gross energy foods (cluster 3) (x̅ =30.7%, SE=1.9). High fiber foods (cluster 2) were highest in ADF (x̅ =45.3%, SE=2.4) (Table 3.5). High available protein foods (cluster 1) had the lowest ADF (x̅ =19.8%, SE=1.0). The result was similar for high non-protein gross energy foods (cluster 3) (x̅ =22.2%, SE=1.5). As expected, the %ADF were lower for all clusters relative to %NDF.

3.4.3. High non-protein gross energy foods: Cluster 3

Cluster 3 (n=46) foods were highest in non-protein gross energy (x̅ =4.3kcal/g, SE=0.06) (Table 3.6). High available protein foods (cluster 1) were the lowest in non-protein energy (x̅ =3.5kcal/g, SE=0.05). High fiber foods (cluster 2) were intermediate in non-protein energy (x̅ =4.0kcal/g, SE=0.07).

3.4.4. Mineral (ash) content Ash was not grouped into a food category cluster. High available protein foods (cluster 1) were highest in minerals (x̅ =5.9%, SE=0.3) (Table 3.7). High fiber foods (cluster 2) were intermediate in minerals (x̅ =4.8%, SE=0.7). High non-protein gross energy foods (cluster 3) were lowest in minerals (x̅ =4.0%, SE=0.2). A one-way between subject’s ANOVA was used to compare the

102

effect of the three different food nutrient types selected by lactating P. coquereli that were produced by the k-means cluster analysis. These included: available protein, NDF, ADF, and non-protein gross energy. Minerals did not cluster into a food nutrient type, but was included to determine its effect. There was a significant effect of the percentage of available protein (cluster 1 as established by the k-means cluster analysis) consumed by lactating P. coquereli at the p<0.05 level (F [2, 120] =122.52, p<0.001). There was a significant effect of the percentage of NDF (cluster 2 as established by the k-means cluster analysis) (F [2, 120] =68.18, p<0.001) consumed by lactating P. coquereli. There was a significant effect of the percentage of ADF (cluster 2 as established by the k-means cluster analysis) (F [2,120] =68.28, p<0.001) consumed by lactating P. coquereli. There was a significant effect of the amount (kcal/g) of non-protein gross energy (cluster 3 as established by the k-means cluster analysis) (F [2,120] =5.72, p<0.001) consumed by lactating P. coquereli. There was also a significant effect of the percentage of minerals (F [2,116] =7.05, p<0.001) consumed by lactating P. coquereli.

3.4.5. Frequency of ingestion index Based on the percentages reported in the relative FOI, foods high in non-protein gross energy (cluster 3), 54% ranked the highest with a score of 5 in the FOI (Tables 3.1, 3.8). In contrast, 42% of high fiber foods (cluster 2) foods ranked the highest, whereas only 5% of high available protein foods (cluster 1) foods ranked the highest (Tables 3.1, 3.8).

3.5. Discussion

Nutritional food quality is a “bottom-up” ecological process critical in evaluating the evolution of life history traits and social behavior (McCabe and Fedigan 2007). However, identifying the nutritional factors driving food selection in lactating Propithecus spp. is glaringly absent from primatological literature (but see, Irwin 2008b; Irwin et al. 2014; Simmen et al. 2013). I applied three nutritional models to expand on the theoretical framework of optimal foraging theory: nitrogen maximization, fiber limitations, and energy maximization to determine the quality of foods consumed by lactating P. coquereli and their biological significance. Nitrogen maximization model P1. P. coquereli lactating females will select foods high in available protein. P2. P. coquereli lactating females will select foods high in minerals.

103

Fiber limitation model P3. P. coquereli lactating females will select foods low in fiber. Energy maximization model P4. P. coquereli lactating females will select foods high in non-protein energy. Predictions 1 and 4 were statistically supported, in that lactating P. coquereli selected foods high in available protein (cluster 1) and non-protein energy (cluster 3). Prediction 3 was not supported, as lactating P. coquereli selected foods high in fiber (cluster 2). Prediction 2 was partially statistically supported, in that minerals were not clustered into a food category, but foods with significantly high mineral content were consumed. Thus, lactating P. coquereli selected three nutritionally distinct food types, as represented by each cluster, considering the foods seasonally available during the dry season. The most frequently selected foods were those high in non-protein energy (cluster 3). High fiber (cluster 2) and available protein foods (cluster 1) were regularly selected. Available protein (cluster 1) and non-protein energy resources (cluster 3) had lower fiber content, and correspondingly high protein-to-fiber ratios. Though the non-protein energy values within high available protein foods (cluster 1) (3.4 kcal/g) and high non-protein energy foods (cluster 3) (4.3 kcal/g) appeared relatively similar to high fiber foods (cluster 2) (4.0 kcal/g), the total amount of energy in fiber was not available to P. coquereli as it remained bound up within the fiber fraction. The FOI supports that lactating P. coquereli selected foods rich in non-protein energy at greater frequency than high available protein (cluster 1) and fiber resources (cluster 2) as 54% of non-protein energy foods (cluster 3) ranked the highest in the index.

I found evidence to support both nitrogen (Mattson 1980) and energy maximization (Schoener 1971) models, since lactating P. coquereli maximize protein intake while concomitantly minimizing fiber intake when consuming high available protein foods (cluster 1). The fiber limitation model (Milton 1979) was indirectly supported since high available protein foods were concurrently lowest in fiber, but high fiber foods were still regularly selected. Based on these findings, optimal foraging theory (Stephens and Krebs 1986) is supported since lactating P. coquereli acquired sufficient nutrients to balance energetic expenditures. In the future, considering nutrients in tandem as per the geometric framework in the nutrient balancing model (Raubenheimer and Simpson 2004) will assist in understanding how multiple nutrients interact with each other. Viewing an animal’s nutritional state and the nutrients it ingests as a dynamic,

104

open system will provide the most comprehensive view of its nutritional goals. Gestating and lactating females often consume leaves higher in protein relative to conspecifics to conserve metabolic expenditures (e.g., Goldizen 1987; Sauther 1994; Vasey 2002). Protein is critical in regulating reproductive function and consuming foods high in protein is advantageous since lactation increases the quantity of amino acids needed to energetically sustain females, thereby increasing protein requirements (Jessop 1997).

Resource seasonality led to the evolution of female feeding priority and weaning synchrony in lemurs (Jolly 1984). Reproductive females experience high metabolic requirements and Propithecus spp. females give birth to singletons every other year, with only one reproductive female and one infant present per group to cope with extreme seasonality and nutritional duress (Richard et al. 1991). P. coquereli infants are born during the austral winter (Lewis and Kappeler 2005b) and mothers are the primary infant carriers from birth until weaning (Ross and Lehman 2016). Consequently, mothers must cope with the energetic challenges of lactating and carrying infants while simultaneously experiencing the most annually constraining climatic conditions. A long-term study found that P. coquereli adult females spent significantly more time feeding than adult males (McGoogan 2011), which is one foraging strategy to manage these challenges. Conversely, Gould et al. (2011) found no sex differences in L. catta intake rates of the top five most frequently consumed foods during early gestation and early to mid-lactation, time spent feeding, or the specific food types consumed. Thus, determining the nutritional quality of food and its biological significance in lactating P. coquereli demonstrates how females nutritionally compensate to successfully behaviorally and nutritionally care for infants during the dry season.

3.5.1. High available protein foods: Cluster 1 Lactating P. coquereli selected high protein foods from 1-26 weeks postnatal over two consecutive birth seasons. The mean (15.9%) of high available protein foods exceeded minimum requirements for primate reproduction (14%) and infant development (7-11%) (Oftedal 1991). The available protein foods consumed by lactating P. coquereli ranged from 9.8-28.3%. This range is considerable given minimum requirements. Consuming foods with a wide range of high available protein is advantageous since high available protein foods concurrently have intermediate NDF and the lowest ADF relative to high fiber and high non-protein gross energy

105

foods. The distribution of red-tailed sportive lemurs (Lepilemur ruficaudatus) positively correlated with the spatial distribution of leaf protein during the wet season during lactation, though neither leaf availability or nutrient composition influenced L. ruficaudatus distribution in the dry season during the mating period (Ganzhorn 2002). L. ruficaudatus adjusted their home ranges when leaves were most abundant and optimized access to high quality food during lactation, signifying that lactation imposes greater constraints than food availability and quality during the dry season (Ganzhorn 2002). Lactating white-headed capuchins (Cebus capucinus) consumed 33% more grams of protein per hour than pregnant and cycling females, and ingested more grams of other nutrients including fat and sugar, indicating an overall increase in macronutrient consumption rather than a targeted increase of exclusively protein consumption during lactation (McCabe and Fedigan 2007). Protein regulation preceding the importance of non-protein energy suggests that acquiring short-term excess or deficits in carbohydrates and fats to obtain the necessary amount of protein are small relative to failure of acquiring sufficient protein (Cheng et al. 2008). P. coquereli experience heightened duress from annual lactation during the dry season. If lactation is the primary constraint and is an important factor driving food selection, selecting different nutrient types to fulfill the nutrients most depleted or most needed during lactation, such as maximizing protein as predicted by the nitrogen maximization model (Mattson 1980), is a seasonally dependent strategy.

High available protein foods also had the lowest non-protein gross energy, and interestingly, the highest mineral content. Lactating P. coquereli may experience seasonal mineral deficiencies and were observed licking cement tourist benches and returned to the same dead trees throughout the study to consume bark (A. Ross, pers. obs.). Consuming foods high in minerals may assist P. coquereli with offsetting mineral deficiencies during the dry season. Decaying bark holds little nutritional value, yet G. beringei have difficulty acquiring sufficient sodium in their environment and exploit decaying wood as a sodium source (Rothman et al. 2006a). Tropical, terrestrial plants contain low sodium concentrations and folivorous primates are most susceptible to sodium deficiencies (Rothman et al. 2006b). Bark consumed by lactating P. coquereli contains considerably higher mineral content than fruits and leaves (see Table 3.2). Geophagy was very rarely observed during the dry season (A. Ross, pers. obs.), and it is unlikely lactating P. coquereli use this strategy to obtain sufficient sodium. In the future, perchloric acid assays will

106

be applied to specifically isolate sodium content within total mineral content will determine if P. coquereli experience sodium deficiencies during lactation.

3.5.2. High non-protein gross energy foods: Cluster 3 Lactating P. coquereli frequently selected foods high in non-protein energy as predicted by the energy maximization model (Schoener 1971). Foods high in non-protein energy were intermediate in available protein, had the lowest NDF, intermediate ADF, and lowest mineral content. Reproductive females increase their chances of producing more total offspring that will survive until reproductive age by consuming foods rich in available nutrients (McCabe and Fedigan 2007). Non-protein energy is stored more easily than protein and can be used metabolically during negative energy balance; this is not possible for protein (Felton et al. 2009b). These foods provide lactating females with metabolic fuel, thereby giving mothers greater access to immediate energy and energy storage. This dietary pattern would be particularly advantageous during the energetically expensive lactation period. P. verreauxi infant mortality and fecundity are closely associated with female body condition in the dry season (Richard et al. 2000). P. verreauxi infants have a mortality rate of 48% within the first postnatal year, which is unusually high relative to primates of similar body size (Richard et al. 2002). Since body weight is often used as a proxy for measuring body condition in primates (see Leutenegger 1973; Lewis and Kappeler 2005b; Tardif et al. 2004), consuming high energy foods for both immediate and future use would help females maintain a healthy body condition to lactate, care for infants, and subsequently mate.

“Income breeders” must acquire nutrients from immediate food consumption while “capitol breeders” obtain resources from their own food stores (Jonsson 1997). Propithecus spp. employ a mixed strategy, where both food intake and stores in fat or other tissue are used by lactating females to transfer nutrients to their offspring, thereby increasing protein and energy requirements relative to non-reproductive conspecifics (Jonsson 1997; Lee 1996). Lactating C. capucinus ingested foods at significantly higher rates though captivatingly they did not receive more energy (kcal/hour) relative to gestating and cycling females (McCabe and Fedigan 2007). Increased energy intake in captive common marmosets (Callithrix jacchus) exclusively occurred in lactating females (Nievergelt and Martin 1999). Lactating C. jacchus increased their energy intake (kcal/kg0.75/24 hours) up to 100% though continued to lose weight (Nievergelt and Martin

107

1999). Callitrichid infants weigh between 10-15% of their mother’s body weight at birth and reach 25% by weaning (Hearn 1983). In contrast, captive P. coquereli weigh between 85-115g at birth (Ballentine 2013), and adults weigh approximately 4kg (reviewed in Mittermeier 2010). Thus, P. coquereli infants weigh between 2.1-2.9% of their mother’s body weight at birth. C. jacchus mother’s continued weight loss may be the result of caring for heavy infants as energy storage is more difficult in comparison to relatively smaller P. coquereli infants. I found that the infant P. coquereli mortality rate was 9% from birth until 6.5 months postnatal. Thus, lactating P. coquereli are receiving adequate nutrients to sustain themselves and dependent infants. Other factors, such as infant falls, are more likely contributing to high infant mortality rather than mothers not receiving adequate nutrition. Consuming high energy foods may contribute to rapid postnatal growth rates characteristic of strepsirrhines (see Leigh and Terranova 1998). In the future, measuring P. coquereli infant growth from birth until one year postnatal would determine how quickly infants grow.

Climatic events and habitat quality can have detrimental demographic effects on lemur populations. L. catta infant mortality and female survival during a drought reflected prior body condition (Gould et al. 1999). L. catta infant mortality increased from 52% to 80% during a drought year and 20% of adult females died (Gould et al. 1999; Gould et al. 2003). Rainfall patterns fell within expected ranges during the time I collected my data (Durrell 2011). Consuming high energy foods potentially prepares P. coquereli lactating females for droughts caused by cyclones since energy storage is most critical during periods of decreased food availability. It is clear that the dry season has negative life history outcomes in Propithecus spp., though unlike L. ruficaudatus, P. coquereli lactate during the dry season and consequently this tradeoff allows reproductive females access to higher quality resources during weaning, mating, and gestation. Irwin et al. (2014) found daily energy and decreased macronutrient intake in diademed sifaka (Propithecus diadema) during the dry season was not the result of nutrient composition, but rather the quantity of food ingested and subsequently absorbed. The home ranges of the two study groups contained both interior and edge habitat, which likely contributed to the nutrients available in the foods selected by P. diadema (Irwin et al. 2014). Additionally, the two adult females were lactating only during the first half of the study, which may have influenced food selection (Irwin et al. 2014). Primates lactate for longer durations in comparison to other mammals with considerable variation occurring in milk composition and quality across

108

primate species (Ben Shaul 1962; Hinde 2007; Milligan et al. 2007; Power et al. 2002; Tilden and Oftedal 1997). Lactation is a dynamic process, for instance there is substantial interspecific and interindividual variation in milk nutrient composition between early, mid and late lactation (see Buss et al. 1976; Hinde et al. 2009; Iverson and Oftedal 1995; Oftedal 1984; Oftedal and Iverson 1995; Power et al. 2002). This, along with my results, raises two questions: do Propithecus spp. males select nutritionally different foods or are they eating the same food types as gestating and lactating females? Secondly, do lactating females ingest food at higher rates than males and non-lactating females?

3.5.3. High fiber foods: Cluster 2 While available protein and non-protein energy foods have relatively low fiber content, lactating P. coquereli still selected high fiber foods. High fiber foods have the lowest available protein, and intermediate ash and non-protein gross energy content. It is unclear why lactating females consume high fiber foods; here I present three possibilities for future investigation. High fiber consumption may reflect digestive efficiency. Propithecus spp. are specialized hindgut fermenters (Lambert 1998). Hindgut fermenters incorporate unique gut morphology such as enlarged caecum’s and elongated colons. This specialized morphology could enable the digestion of fibrous materials while increasing nutrient extraction from difficult to digest resources. Lactating P. coquereli consume bark during the dry season (A. Ross, pers. obs.) and conspecifics have been observed consuming bark and decaying wood year-round (McGoogan 2011).

Lactating P. coquereli may select high fiber foods to promote digestive health or improve body condition. Females may benefit from fermentation and nutrient absorption overnight when they eat less and are less active if greater quantities of high fiber foods are consumed later in the day. P. coquereli may have specialized digestive adaptations for breaking down secondary plant compounds or possess high thresholds for these compounds (see Burgess and Chapman 2005). Tannins are plant polyphenols and secondary compounds that bind to proteins, inhibit protein uptake, and form insoluble complexes (Lambert 1998). Tannins have primarily been viewed as anti-nutritional agents and tannin consumption have been inadequately studied in primates. Tannin consumption in lactating females can result in protein deficiencies (Oftedal 1991). Preliminary unpublished data demonstrates that P. coquereli avoided tannin containing plants

109

(reviewed in Ganzhorn 1989), although P. verreauxi periparturient (occurring near the time of parturition) females selectively consumed plants with higher tannin content than conspecifics (Carrai et al. 2003). Tannins contain anti-abortive properties and are associated with weight gain and stimulate milk secretion (reviewed in Carrai et al. 2003). A combination of a specialized gastrointestinal system and the positive nutritive benefits gained from tannin consumption may offset the costs of consuming high fiber foods during lactation and assist in the preparation for future pregnancies. The study concluded P. verreauxi periparturient females self-medicate with tannin containing plants to improve reproductive fitness (Carrai et al. 2003). I argue that this conclusion is preliminary based on the lack of evidence demonstrating a clear relationship between the identification of preferred foods and the cognitive and ecological constraints determining how lactating Propithecus spp. choose resources.

High fiber consumption may be a residual effect of lactating P. coquereli unassumingly consuming the foods available in the forest during the dry season. P. diadema were more selective in food choices when food was abundant and of higher quality during the wet season (Irwin 2008b). High fiber foods may only be what are available for consumption during the dry season in a dry deciduous forest, thereby constraining females to select difficult to digest resources.

In sum, lactating P. coquereli frequently select foods high in non-protein energy, though consistently select high available protein and high fiber foods. I found support for the nitrogen and energy maximization nutritional models that can be contextualized within optimal foraging theory. While other primates utilize behavioral compensation such as male infant carrying (e.g., Lappan 2009; Tardif 1994) to offset lactation costs, lactating P. coquereli successfully acquire sufficient nutrients to help balance energetic costs in an unstable environment by food selection. Future studies quantifying basal metabolic rates and other adaptations stemming from the energy conservation hypothesis (see Wright 1999) will assist in determining how P. coquereli optimize their physical and physiological environments. Tree identification from this dissertation can be implemented as a conservation management tool to replant specific tree species selected by lactating P. coquereli. Ensuring P. coquereli mothers have continued access to resources selected during lactation, particularly in forest fragments where there are already fewer trees, will assist in protecting the health and survival of mothers and their infants. Future studies

110

comparing food intake rates and nutrient composition of foods within each lactational stage will help determine if early, mid, or late lactation influences macronutrient intake in Propithecus spp. Another opportunity for future study is examining digestive efficiency relative to energy availability in tandem with ingestion rates and nutrient composition Quantifying P. coquereli milk composition will establish a baseline for examining nutrient transfer from mothers to infants. Comparing P. coquereli milk to the dilute milks characteristic of other lemurs (Tilden and Oftedal 1997) is necessary to determine the evolution of lactation and milk quality in some of the most basal extant primate taxa. Lastly, determining the proximate and ultimate causes for food selection will reveal how Propithecus spp. are constrained in food selection seasonally. For example, discerning whether lactating females choose food from specific areas for specific reasons or whether selection is more random and opportunistic.

3.6. References

Albignac R. 1981. Lemurine social and territorial organization in a north-western Malagasy forest (restricted area of Ampijoroa). In: Chiarelli A, and Corruccini R, editors. Primate Behavior and Sociobiology. New York: Springer-Verlag. p 25-29. Alonso L, and Hannah L. 2002. Introduction to the Reserve Naturelle Integrale and Reserve Forestiere d’Ankarafantsika and to the rapid assessment program. In: Schulenberg T, Radilofe S, and Missa O, editors. A biological assessment of the Reserve Naturelle Integrale d’Ankarafantsika, Madagascar. Washington DC: Conservation International p 28-62. Altmann J. 1974. Observational study of behavior: sampling methods. Behaviour 49:227-267. Andrainarivo C, Andriaholinirina, V.N., Feistner, A., Felix, T., Ganzhorn, J., Garbutt, N., Golden, C., Konstant, B., Louis Jr., E., Meyers, D., Mittermeier, R.A., Perieras, A., Princee, F., Rabarivola, J.C., Rakotosamimanana, B., Rasamimanana, H., Ratsimbazafy, J., Raveloarinoro, G., Razafimanantsoa, A., Rumpler, Y., Schwitzer, C., Thalmann, U., Wilmé, L. & Wright, P. 2008. IUCN Red List of Threatened Species: Propithecus coquereli. 17 November 2013 ed. Ballentine J. 2013. Maryland Zoo Welcomes Newborn Sifaka Baby. Press release: The Maryland Zoo in Baltimore p 1-2. Ben Shaul D. 1962. The composition of the milk of wild animals. Journal of Dairy Science 4:333-342. Burgess M, and Chapman C. 2005. Tree leaf chemical characters: selective pressures by folivorous primates and invertebrates. African Journal of Ecology 43:242-250. Buss D, Cooper R, and Wallen K. 1976. Composition of lemur milk. Folia Primatologica 26:301-305. Cameron JL. 1996. Regulation of reproductive hormone secretion in primates by short-term changes in nutrition. Reviews of Reproduction 1:117-126. Carrai V, Borgognini-Tarli SM, Huffman MA, and Bardi M. 2003. Increase in tannin consumption by sifaka (Propithecus verreauxi verreauxi) females during the birth season: a case for self-medication in prosimians? Primates 44:61-66. Chapman CA, Chapman L, Naughton-Treves L, Lawes MJ, and McDowell LR. 2004. Predicting a folivorous primate abundance: validation of a nutritional model. American Journal of Primatology 62:55-69. Cheng K, Simpson S, and Raubenheimer D. 2008. A geometry of regulatory scaling. American Naturalist 172:681- 693. Conklin N, and Wrangham R. 1994. The value of figs to a hind-gut fermenting frugivores: a nutritional analysis. Biochemical Systematics & Ecology 22:137-151. Crowley B, McGoogan K, and Lehman S. 2012. Edge effects on foliar stable isotope values in a Madagascan tropical dry forest. Public Library of Science ONE 7:1-9. Dewar R, and Richard A. 2007. Evolution in the hypervariable environment of Madagascar. Proceedings of the National Academy of Sciences of the USA 104:13723-13727. Di Fiore A, Link A, and Dew J. 2008. Diets of wild spider monkeys. In: Campbell C, editor. Spider monkeys: behavior, ecology, and evolution of the genus Ateles. Cambridge (UK): Cambridge University Press. p 81- 137. Di Fiore A, and Rodman P. 2001. Time allocation patterns of lowland woolly monkeys (Lagothrix lagotricha poeppigii) in a neotropical terra firma forest. International Journal of Primatology 22:449-480. Du Puy D, and Moat J. 1996. A refined classification of the primary vegetation of Madagascar based on the underlying geology: using GIS to map its distribution and assess its conservation status. In: Lourenço W, editor. Proceedings of the international symposium on the biogeography of Madagascar Paris: Editions de l'Orstrom p 205-218. Dunham AE, Erhart EM, and Wright PC. 2011. Global climate cycles and cyclones: consequences for rainfall patterns and lemur reproduction in southeastern Madagacar Global Change Biology 17:219-227. Durrell. 2011. Durrell precipitation and temperature data. Ankarafantsika National Park, Madagascar. Durrell Wildlife Conservation Trust. Felton A, Felton A, Lindenmayer D, and Foley W. 2009. Nutritional goals of wild primates. Functional Ecology 23:70-78. Ferreira L, and Laurance W. 1997. Effects of forest fragmentation on mortality and damage of selected trees in central Amazonia. Conservation Biology 11:797-801. Freeland W, and Janzen D. 1974. Strategies in herbivory by mammals: the role of plant secondary compounds. American Naturalist 108:269-289. Ganzhorn J. 1989. Primate species separation in relation to secondary plant chemicals. Human Evolution 4:125-132.

111 112

Ganzhorn J. 1995a. Low-level forest disturbance effects on primary production, leaf chemistry, and lemur populations. Ecology 76:2084-2096. Ganzhorn J. 2002. Distribution of a folivorous lemur in relation to seasonally varying food resources: integrating quantitative and qualitative aspects of food characteristics. Oecologia 131:427-435. Gerardo G, and Goodman SM. 2003. Hunting of protected animals in the Parc National d'Ankarafantsika, northwestern Madagascar. Oryx 37:115-118. Gittleman J, and Thomspon S. 1988. Energy allocation in mammalian reproduction. American Zoologist 28:863- 875. Glander K. 1982. The impact of plant secondary compounds on primate feeding behavior. Yearbook of Physical Anthropology 25:1-18. Goldizen A. 1987. Tamarins and marmosets: communal care of offspring. In: Smuts B, Cheney D, Seyfarth R, and Wrangham R, editors. Primate Societies Chicago: University of Chicago Press. p 34-43. Gould L, Power ML, Ellwanger N, and Rambeloarivony H. 2011. Feeding behavior and nutrient intake in spiny forest- dwelling ring-tailed lemurs (Lemur catta) during early gestation and early to mid-lactation periods: compensating in a harsh environment. American Journal of Physical Anthropology 145:469-479. Gould L, Sussman RW, and Sauther ML. 1999. Natural disasters and primate populations: the effects of a 2-year drought on a naturally occurring population of ring-tailed lemurs (Lemur catta) in southwestern Madagascar International Journal of Primatology 20:69-84. Gould L, Sussman RW, and Sauther ML. 2003. Demographic and life-history patterns in a population of ring-tailed lemurs (Lemur catta) at Beza Mahafaly Reserve, Madagascar: A 15-year perspective American Journal of Physical Anthropology 120:182-194. Hearn J. 1983. The common marmoset (Callithrix jacchus). In: Hearn J, editor. Reproduction in New World Primates. Lancaster: MTP Press p183-215. Hinde K. 2007. Milk composition varies in relation to the presence and abundance of Balantidium coli in the mother in captive rhesus macaques (Macaca mulatta). American Journal of Primatology 69:625-634. Hinde K, Power M, and Oftedal O. 2009. Rhesus macaque milk: magnitude, sources, and consequences of individual variation over lactation. American Journal of Physical Anthropology 138:148-157. Hylin J, and Lichton I. 1965. Production of reversible infertility in rats by feeding mimosine. Biochemical Pharmacology 14:1167-1169. Irwin M. 2008b. Feeding ecology of Propithecus diadema in forest fragments and continuous forest. International Journal of Primatology 29:95-115. Irwin M, Raharison J, Raubenheimer D, Chapman C, and JM R. 2014. Nutritional correlates of the "lean season": effects of seasonality and frugivory on the nutritional ecology of diademed sifakas. American Journal of Physical Anthropology 153:78-91. Iverson S, and Oftedal O. 1995. Philogenetic and ecological variation in the fatty acid composition of milks. In: Jensen RG, editor. Handbook of Milk Composition. New York: Academic Press. Jain A. 2010. Data clustering: 50 years beyond K-means. Pattern Recognition Letters 31:651-666. Jakubasz M. 2011. Dry matter determination using forced air convection oven. Nutrition laboratory. Smithsonian National Zoological Park, Washington, D.C. Jakubasz M. 2012a. Acid detergent fiber procedure using Ankom 200 Analyzer. Nutrition Laboratory Manual Smithsonian National Zoological Park, Washington, D.C. Jakubasz M. 2012b. Bomb calorimetry using Parr 1241 Adiabatic Calorimeter. Nutrition Laboratory Manual Smithsonian National Zoological Park, Washington, D.C. Jakubasz M. 2012c. Determination of carbon, hydrogen, and nitrogen using PerkinElmer 2400 Series II Analyzer. Nutrition Laboratory Manual. Smithsonian National Zoological Park, Washington, D.C. Jakubasz M. 2012d. Neutral detergent fiber procedure using Ankom 200 Analyzer. Nutrition Laboratory. Smithsonian National Zoological Park, Washington, D.C. Jessop N. 1997. Protein metabolism during lactation. Proceedings of the Nutrition Society 56:169-175. Jolly A. 1984. The puzzle of female feeding priority. In: Small M, editor. Female Primates: Studies by Women Primatologists. New York: Alan Liss, INC. Jonsson K. 1997. Capital and income breeding as alternative tactics of resource use in reproduction. Oikos 78:57-66. Kun-Rodrigues C, Salmona J, Besolo A, Rasolondraibe E, Rabarivola C, Marques TA, and Chikhi L. 2014. New density estimates of a threatened sifaka species (Propithecus coquereli) in Ankarafantsika National Park. American Journal of Primatology 76:515-528. Lambert J. 1998. Primate digestion: interactions among anatomy, physiology, and feeding ecology. Evolutionary Anthropology 7:8-20.

113

Lappan S. 2009. The effects of lactation and infant care on adult energy budgets in wild siamangs (Symphalangus syndactylus). American Journal of Physical Anthropology 140:290-301. Laurance W, Ferreira L, Rankin-de Merona J, Laurance S, Hutchings R, and Lovejoy T. 1998. Effects of forest fragmentation on recruitment patterns in Amazonian tree communities Conservation Biology 12:460-464 Laurance WF, Lovejoy TE, Vasconcelos HL, Bruna EM, Didham RK, Stouffer PC, Gascon C, Bierregaard RO, Laurance SG, and Sampaio E. 2002. Ecosystem decay of Amazonian forest fragments: a 22 year investigation. Conservation Biology 16:605-618. Lee P, Majluf P, and Gordon I. 1991. Growth, weaning and maternal investment from a comparative perspective. Zoology 225:99-114. Lee PC. 1996. The meanings of weaning: growth, lactation, and life history. Evolutionary Anthropology 5:87-98. Legendre P, and Legendre L. 1998. Cluster analysis. In: Legendre P, and Legendre L, editors. Numerical Ecology. New York: Elsevier. p 303-383. Leigh S, and Terranova C. 1998. Comparative perspectives on bimaturism, ontogeny, and dimorphism in lemurid primates. International Journal of Primatology 19:723-749. Leutenegger W. 1973. Maternal-fetal weight relationships in primates. Folia Primatologica 20:280-293. Lewis RJ, and Kappeler PM. 2005. Seasonality, body condition, and timing of reproduction in Propithecus verreauxi verreauxi in the Kirindy Forest. American Journal of Primatology 67:347-364. Lin H, Frassetto A, Kowalik Jr EJ, Nawrocki A, Lu M, Kosinski J, Hubert J, Szeto D, Yao X, Forrest G, Marsh D. 2012. Butyrate and propionate protect against diet-induced obesity and regulate gut hormones via free fatty acid receptor 3-independent mechanisms. PLoS ONE 7: e35240. Lourenço W, and Goodman S. 2006. Notes on the postembryonic development and ecology of Grosphus hirtus Kraepelin, 1901 (Scorpiones, Buthidae) from the Parc National d'Ankarafantsika, northwest Madagascar. Zoologischer Anzeiger 244:181-185. Mattson W. 1980. Herbivory in relation to plant nitrogen content. Annual Review of Ecology and Systematics 11:119-161. Maynard L, and Loosli J. 1965a. The animal body and its food. In: Maynard L, and Loosli J, editors. Animal Nutrition New York: McGraw-Hill. p 11-27. Maynard L, and Loosli J. 1965b. The proteins and their metabolism. In: Maynard L, and Loosli J, editors. Animal Nutrition New York: McGraw-Hill. p 115-153. McCabe G, and Fedigan L. 2007. Effects of reproductive status on energy intake, ingestion rates, and dietary composition of female Cebus capucinus at Santa Rosa, Costa Rica. International Journal of Primatology 28:837-851. McGoogan K. 2011. Edge effects on the behaviour and ecology of Propithecus coquereli in Northwest Madagascar [dissertation]. Available from TSpace Repository, 1807/31861: University of Toronto 225 p. Meyers D, and Wright P. 1993. Resource tracking: food availability and Propithecus seasonal reproduction. In: Kappeler P, and Ganzhorn J, editors. Lemur Social Systems and their Ecological Basis. New York: Plenum Press. p 179-192. Milligan L, Gibson S, WIliams L, and Power M. 2007. The composition of milk from Bolivian squirrel monkeys (Saimiri boliviensis boliviensis). American Journal of Primatology 69:1-14. Milton K. 1977. The foraging strategy of the howler monkey (Alouatta palliata) in the tropical forest Barro Colorado Island, Panama [dissertation]. New York: New York University. Milton K. 1979. Factors influencing leaf choice by howler monkeys: a test of some hypotheses of food selection by generalist herbivores. American Naturalist 114:362-378. Milton K. 1983. Diet and primate evolution. Scientific American August:86-93. Mittermeier R, Louis Jr., E, Richardson, M, Schwitzer, C, Langrand, O, Rylands, A, Hawkins, F, Rajaobelina, S, Ratsimbazafy, J, Rasoloarison, R, Roos, C, Kappeler, P, Mackinnon, J. 2010. Indriidae. Lemurs of Madagascar. Third ed. Bogotá, Colombia: Conservation International. p 481-597. Moir R. 1965. The comparative physiology of ruminant-like animals. In: Dougherty R, editor. Physiology of Digestion in the Ruminant. Washington, D.C.: Butterworths. Myatt L. 2006. Placental adaptive responses and fetal programming Journal of Physiology 572:25-30. Nievergelt C, and Martin R. 1999. Energy intake during reproduction in captive common marmosets (Callithrix jacchus). Physiology & Behavior 65:849-854. Oftedal O. 1984. Milk composition, milk yield and energy output at peak lactation: a comparative review. Symposium Zoological Society of London 51:33-85. Oftedal O. 1991. The nutritional consequences of foraging in primates: the relationship of nutrient intakes to nutrient requirements. Philosophical Transactions of the Royal Society B 334:161-170.

114

Oftedal O, and Iverson S. 1995. Comparative analysis of nonhuman milks: Phylogenetic variation in the gross composition of milks. In: Jensen RG, editor. Handbook of Milk Composition. New York: Academic Press. Oliveria A, Marcio de Mello J, and Scolforo J. 1997. Effects of past disturbance and edges on tree community structure and dynamics within a fragment of tropical semideciduous forest in south-eastern Brazil over a five-year period (1987-1992). Plant Ecology 131:45-66. Pochron S, Tucker W, and Wright P. 2004. Demography, life history, and social structure in Propithecus diadema edwardsi from 1986–2000 in Ranomafana National Park, Madagascar. American Journal of Physical Anthropology 125:61-72. Power M, Oftedal O, and Tardif S. 2002. Does the milk of callitrichid monkeys differ from that of larger anthropoids? American Journal of Primatology 56:117-127. Purvis A, Gittleman J, Cowlishaw G, and Mace G. 2000. Predicting extinction risk in declining species. Proceedings of the Royal Society B 267:1947-1952. Raubenheimer D, and Simpson S. 2004. Organismal stoichiometry: quantifying non-independence among food components. Ecology 85:1203-1216. Rendigs A, Radespiel U, Wrogemann D, and Zimmermann E. 2003. Relationship between microhabitat structure and distribution of mouse lemurs (Microcebus spp.) in northwestern Madagascar. International Journal of Primatology 24:47-64. Richard A, Dewar R, Schwartz M, and Ratsirarson J. 2000. Mass change, environmental variability and female fertility in wild Propithecus verreauxi. Journal of Human Evolution 39:381-391. Richard A, Dewar R, Schwartz M, and Ratsirarson J. 2002. Life in the slow lane? Demography and life histories of male and female sifaka (Propithecus verreauxi verreauxi). The Zoological Society of London 256:421-436. Richard A, Rakotomanga P, and Schwartz M. 1991. Demography of Propithecus verreauxi at Beza Mahafaly, Madagascar: sex ratio, survival, and fertility, 1984-1988. American Journal of Physical Anthropology 84:307-322. Ross A, and Lehman S. 2016. Infant transport and mother–infant contact from 1 to 26 weeks postnatal in Coquerel’s sifaka (Propithecus coquereli) in northwestern Madagascar. American Journal of Primatology 78:646-658. Rothman J, Dierenfeld E, Molina D, Shaw A, Hintz H, and Pell A. 2006a. Nutritional chemistry of foods eaten by gorillas in Bwindi Impenetrable National Park, Uganda. American Journal of Primatology 68:675-691. Rothman J, Dierenfeld ES, Hintz H, and Pell A. 2008. Nutritional quality of gorilla diets: consequences of age, sex, and season. Oecologia 155:111-122. Rothman J, Van Soest P, and Pell A. 2006b. Decaying wood is a sodium source for mountain gorillas. Biology Letters 2:321-324. Rutherford J. 2013. The primate placenta as an agent of developmental and health trajectories across the life course In: Clancy KB, Hinde K, and Rutherford J, editors. Building Babies: Primate Development in Proximate and Ultimate Perspective, Developments in Primatology: Progress and Prospects. New York: Springer Science+Business Media p27-45. Rutherford J, Hurley P, Lawrence M, and Redmond Jr. D. 2010. Fetoplacental growth dynamics in the vervet monkey (Chlorocebus aethiops). Placenta 31:A29. Sauther M. 1994. Wild plant use by pregnant and lactating ringtailed lemurs, with implications for early hominid foraging. Eating on the Wild Side Tucson: University of Arizona Press. p 240-256. Schoener T. 1971. Theory of feeding strategies. Annual Review of Ecology and Systematics 2:369-404. Simmen B, Tarnaud L, Marez A, and Hladik A. 2013. Leaf chemistry as a predictor of primate biomass and the mediating role of food selection: a case study in a folivorous lemur (Propithecus verreauxi). American Journal of Primatology 76:563-575. Stephens D, and Krebs J. 1986. Foraging Theory. Princeton, NJ: Princeton University Press. 247 p. Sterck E, Watts D, and van Schaik C. 1997. The evolution of female social relationships in nonhuman primates. Behavioural Ecology and Sociobiology 41:291-309. Tardif S. 1994. Relative energetic cost of infant care in small-bodied neotropical primates and its relation to infant- care patterns. American Journal of Primatology 34:133-143. Tardif S, Power M, Layne D, Smucny D, and Ziegler T. 2004. Energy restriction initiated at different gestational ages has varying effects on maternal weight gain and pregnancy outcome in common marmoset monkeys (Callithrix jacchus). British Journal of Nutrition 92:841-849. Tilden C, and Oftedal O. 1997. Milk composition reflects patterns of maternal care in prosimian primates. American Journal of Primatology 41:195-211. Van Soest P, Robertson J, and Lewis B. 1991. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition Journal of Dairy Science 74:3583-3597.

115

Vasey N. 2002. Niche separation in Varecia variegata rubra and Eulemur fulvus albifrons: II. Intraspecific patterns. American Journal of Physical Anthropology 118:169-183. Vasey N. 2005. Activity budgets and activity rhythms in red ruffed lemurs (Varecia rubra) on the Masoala Peninsula, Madagascar: seasonality and reproductive energetics. American Journal of Primatology 66:23- 44. Wrangham R. 1980. An ecological model of female-bonded primate groups. Behaviour 75:262-300. Wright P. 1999. Lemur traits and Madagascar ecology: coping with an island environment. Yearbook of Physical Anthropology 42:31-72.

116

Table 3.1. Identification of assayed foods selected by lactating P. coquereli ID Parts DBH FOI Vernacular Family Genus/ selected species 2 Fruit 55.0 14% KESIKA VERBENACEAE Tectonia grandis 3 Fruit 57.0 15% KESIKA VERBENACEAE Tectonia grandis 4 Fruit 25.0 23% TSITIPAHA MORACEAE Treculia perrieri 6 Leaves 14.0 11% MANARY FABACEAE Dalbergia trichophylla 8 Bark - - UNKNOWN - - 9 Fruit 18.5 11% HAZOPIKA SAPOTACEAE Mimusops spp. 11 Leaves 16.5 15% DITIMENA ANACARDIACEAE Abrahamia ditimena 12 Leaves 0.5 10% KABIJALAHY RHAMNACEAE Bathiorhamnus spp. 14 Leaves 6.0 11% TSIMATIMANOTA CLUSIACEAE Mammea punctata 15 Leaves 7.0 11% KABIJALAHY RHAMNACEAE Bathiorhamnus spp. 16 Fruit Liana 20% PIRA APOCYNACEAE Landolphia gummifera 17 Fruit 9.0 15% HAZOPIKA SAPOTACEAE Mimusops spp. 18 Fruit 13.5 5% SELIALA MALVACEAE Grewia spp. 19 Fruit 11.0 8% TSIMATIMANOTA CLUSIACEAE Mammea punctata 20 Leaves 10.0 23% MIMOZA FABACEAE Bussea perrieri 21 Fruit Liana 11% PIRA APOCYNACEAE Landolphia gummifera 22 Bark 5.0 32% MAEVALAFIKA ROSACEAE Grangeria porisa 23 Leaves 22.0 7% MAEVALAFIKA ROSACEAE Grangeria porisa 24 Leaves & 55.5 14% MANGA ANACARDIACEAE Abrahamia spp. fruit 25 Fruit 3.0 11% VOATSIRINDRANA ANACARDIACEAE Sorindeia madagascariensis 26 Leaves 2.0 15% GODROA - - 28 Leaves 9.5 5% MAEVALAFIKA ROSACEAE Grangeria porisa 29 Leaves 36.5 14% MANGA ANACARDIACEAE Abrahamia spp. 30 Leaves 9.5 8% HAZOTSIFAKA OLEACEAE Norhonia spp. 31 Leaves 2.0 15% GODROA - - 32 Leaves 5.0 15% KITSAKITSANALA - - 33 Leaves 13.0 4% VOATSIKIDY - - 34 Leaf buds Liana 4% BONGAPISO - - & leaves 35 Fruit Liana 11% PIRA APOCYNACEAE Landolphia gummifera 36 Flowers Liana 19% VOASALAY - - 37 Leaves 4.5 6% DITIMENA ANACARDIACEAE Abrahamia ditimena 38 Leaves 3.0 11% NATOVAVY CLUSIACEAE Garcinia verrucosa 39 Leaves 12.0 18% GODROA - - 41 Leaves Liana 6% VAHAFISAKA FABACEAE Dalbergia bracteolata 42 Leaves 8.5 6% HAZOPIKA APOCYNACEAE/ Tabernaemontana SAPOTACEAE coffeoides/ Mimusops spp. 43 Leaves 8.0 6% TSIMATIMANOTA CLUSIACEAE Mammea punctata 44 Fruit 7.0 18% NATOVAVY CLUSIACEAE Garcinia verrucosa 45 Leaves 4.0 6% DITIMENA ANACARDIACEAE Abrahamia ditimena 46 Leaves 3.0 6% MAMOARAVINA CELASTRACEAE Polycardia libera 47 Leaves 7.5 12% KILILO MORACEAE Trilepisium madagascariensis

117

ID Parts DBH FOI Vernacular Family Genus/ selected species 48 Leaves 1.5 6% AMANINOMBY COMBRETACEAE Terminalia boivini 49 Leaf buds Liana 11% BONGAPISO - - 50 Leaves 8.0 6% KABIJALAHY RHAMNACEAE Bathiorhamnus spp. 51 Leaves 6.0 6% AMBALAHY ANNONACEAE Polyalthia spp. 52 Leaves Liana 11% MAEVALAFIKA ROSACEAE Grangeria porisa 53 Leaves 3.5 17% HAZOPIKA APOCYNACEAE/ Tabernaemontana SAPOTACEAE coffeoides/ Mimusops spp. 54 Leaves 17.0 8% DITIMENA ANACARDIACEAE Abrahamia ditimena 55 Leaves 6.5 18% MAEVALAFIKA ROSACEAE Grangeria porisa 56 Leaves 14.0 6% AMBALAHY ANNONACEAE Polyalthia spp. 57 Leaves 7.0 4% MATAMBELONA BURSERACEAE Commiphora spp. 59 Leaves - 6% NATOVAVY CLUSIACEAE Garcinia verrucosa 60 Leaf buds 5.5 5% MAEVALAFIKA ROSACEAE Grangeria porisa & leaves 61 Leaves 2.0 6% HAZOTSIFAKA OLEACEAE Norhonia spp. Leaf buds 25.0 4% VAKAKOA LOGANIACEAE Strychnos 62 & leaves madagascariensis 63 Leaves 3.5 8% MORAMENA OCHNACEAE Ochna ciliata 64 Leaf buds 8.0 8% MAEVALAFIKA ROSACEAE Grangeria porisa & leaves 65 Leaves 12.0 17% NATOVAVY CLUSIACEAE Garcinia verrucosa 66 Fruit 28.0 4% VALOMAMAY MELIACEAE Astrotrichylia asterotricha 67 Bark 7.5 4% DITIMENA ANACARDIACEAE Abrahamia ditimena 68 Fruit 4.0 4% PIRA APOCYNACEAE Landolphia gummifera 69 Leaves Liana - UNKNOWN - - 70 Leaves 15.0 15% KITSAKITSANALA - - 72 Leaves Liana - UNKNOWN - - 74 Leaves 5.5 18% HAZOPIKA APOCYNACEAE/ Tabernaemontana SAPOTACEAE coffeoides/ Mimusops spp. 76 Fruit 13.0 15% NATOVAVY CLUSIACEAE Garcinia verrucosa 78 Leaves & 5.0 - SELIVATO TILIACEAE Grewia ambongensis fruit Baillon 79 Leaves 4.0 4% ANDRIAMANAMORA MELIACEAE Malleastrum gracilis 80 Leaves 1.5 8% SAKOALA - - 81 Flowers 16.5 8% SAKOALA - - 82 Leaves 16.0 6% DITIMENA ANACARDIACEAE Abrahamia ditimena 83 Fruit 4.0 8% MAROAMPOTOTRA SAPINDACEAE Macphersonia gracilis 85 Leaves 3.5 11% TSIMATIMANOTA CLUSIACEAE Mammea punctata 86 Leaves Liana 16% BONGAPISO - - 87 Leaves 8.0 5% GODROA - - 89 Fruit 16.0 5% HAZONDRINGITRA RHOPALOCARPA- Rhopalocarpus CEAE similis 91 Leaves 5.0 5% MORAMENA OCHNACEAE Ochna ciliata 92 Leaves 17.5 5% HAZOPIKA APOCYNACEAE/ Tabernaemontana SAPOTACEAE coffeoides/ Mimusops spp. 93 Leaves Liana 17% BONGAPISO - -

118

ID Parts DBH FOI Vernacular Family Genus/ selected species 95 Leaves 2.5 - SELIVATO TILIACEAE Grewia ambongensis Baillon 96 Leaf buds 5.0 5% NATOVAVY CLUSIACEAE Garcinia verrucosa & leaves 97 Flowers 5.0 8% AMBALAHY ANNONACEAE Polyalthia spp. 98 Leaves 3.0 8% KITSAKITSANALA - - 99 Fruit 5.0 6% MAEVALAFIKA ROSACEAE Grangeria porisa 100 Fruit Liana 16% BONGAPISO - - 101 Fruit 11.5 6% DREYDAL - - 103 Leaves 5.0 23% KITSAKITSANALA - - 105 Fruit Liana 8% DREYDAL - - 106 Leaf buds Liana 19% FOTSIAVADIKA ANNONACEAE Monanthotaxis spp. 107 Fruit 5.0 10% TSIMATIMANOTA CLUSIACEAE Mammea punctata 108 Fruit 27.0 20% MANARY FABACEAE Dalbergia trichophylla 109 Fruit 2.0 11% SELIALA MALVACEAE Grewia spp. 111 Fruit 19.5 6% MANARY FABACEAE Dalbergia trichophylla 113 Leaves 3.0 5% HAZONJIA MYRTACEAE Unknown 114 Leaves 2.5 4% GODROA - - 115 Leaves 3.5 4% ALIBIZAHA FABACEAE Albizia mainaea 117 Leaves 5.0 6% HAZOMAFANA - - 118 Leaves 2.0 4% KITSAKITSANALA - - 119 Leaves Liana 6% VAHAFISAKA FABACEAE Dalbergia bracteolata 120 Fruit 6.0 11% VOATSIRINDRANA ANACARDIACEAE Sorindeia madagascariensis 121 Fruit 13.0 8% HAZOPIKA APOCYNACEAE/ Tabernaemontana SAPOTACEAE coffeoides/ Mimusops spp. 123 Leaves 5.0 18% SAKOALA - - 124 Fruit Liana 11% PIRA APOCYNACEAE Landolphia gummifera 125 Fruit Liana 6% DREYDAL - - 126 Fruit 6.0 32% TSIMATIMANOTA CLUSIACEAE Mammea punctata 127 Leaves 7.5 18% HAZOPIKA APOCYNACEAE/ Tabernaemontana SAPOTACEAE coffeoides/ Mimusops spp. 128 Fruit 6.0 4% NATOVAVY CLUSIACEAE Garcinia verrucosa 129 Fruit 11.0 18% NATOVAVY CLUSIACEAE Garcinia verrucosa 130 Fruit 37.0 23% TSITIPAHA MORACEAE Treculia perrieri 131 Leaves 15.5 23% MIMOZA FABACEAE Bussea perrieri 132 Flowers 22.5 44% VOATSIRINDRANA ANACARDIACEAE Sorindeia madagascariensis 134 Bark - 59% HAZOTSIFAKA OLEACEAE Norhonia spp. 135 Leaves & 20.0 9% KINININA - - bark 136 Flowers Liana - UNKNOWN ASCLEPIADACEAE Cynanchum spp. 139 Fruit 37.0 14% TSITIPAHA MORACEAE Treculia perrieri 140 Fruit Liana 59% DREYDAL - - 143 Leaf buds 7.0 8% MAEVALAFIKA ROSACEAE Grangeria porisa & fruit

119

ID Parts DBH FOI Vernacular Family Genus/ selected species 144 Fruit Liana 8% DREYDAL - - 145 Fruit Liana 19% FOTSIAVADIKA ANNONACEAE Monanthotaxis spp. 146 Flowers 5.0 8% VOATSIRINDRANA ANACARDIACEAE Sorindeia madagascariensis 156 Flowers - - PAMBA CAPPARIDACEAE Crateva exelsa 157 Leaves Liana 15% VAHAFISAKA FABACEAE Dalbergia bracteolata 159 Leaves 15.0 6% MATAMBELONA BURSERACEAE Commiphora spp. 161 Leaves 2.5 15% GODROA - -

120

Table 3.2. Available protein, neutral detergent fiber (NDF), acid detergent fiber (ADF), non-protein gross energy (NPGE), and mineral (ash) values of foods selected by lactating P. coquereli ID Parts Available NDF (%) ADF (%) NPGE Ash (%) selected protein (%) (0.1=10%) (0.1=10%) (kcal/g) (0.1=10%) (0.1=10%) 2 Fruit 0.05 0.7778 0.6413 4.10 4.50 3 Fruit 0.05 0.7912 0.6502 4.32 3.78 4 Fruit 0.10 0.3166 0.2557 3.55 9.95 6 Leaves 0.17 0.2018 0.1491 2.42 4.67 8 Bark 0.01 0.8115 0.6514 4.15 - 9 Fruit 0.02 0.7000 0.4604 4.28 1.65 11 Leaves 0.11 0.2603 0.2238 3.71 5.24 12 Leaves 0.10 0.4372 0.2971 4.41 5.24 14 Leaves 0.07 0.3757 0.2661 4.37 2.97 15 Leaves 0.10 0.4957 0.3394 4.19 3.26 16 Fruit 0.03 0.5148 0.3320 3.88 1.45 17 Fruit 0.03 0.6412 0.3913 4.35 2.19 18 Fruit 0.04 0.6779 0.5088 4.02 3.81 19 Fruit 0.03 0.2097 0.0934 4.77 2.11 20 Leaves 0.22 0.2396 0.1707 3.33 4.97 21 Fruit 0.03 0.5335 0.3685 3.94 2.54 22 Bark 0.01 0.7958 0.6852 3.59 19.37 23 Leaves 0.10 0.5647 0.3359 3.74 4.60 24 Leaves & 0.03 0.0860 0.0624 3.70 1.85 fruit 25 Fruit 0.05 0.1480 0.0840 3.81 2.93 26 Leaves 0.08 0.4283 0.3601 3.93 5.04 28 Leaves 0.09 0.4529 0.2777 3.76 13.10 29 Leaves 0.08 0.5040 0.3979 2.47 4.78 30 Leaves 0.06 0.4680 0.3422 4.60 3.59 31 Leaves 0.17 0.1559 0.1201 3.69 6.07 32 Leaves 0.21 0.2910 0.1787 3.06 4.85 33 Leaves 0.21 0.4482 0.2827 3.39 7.22 34 Leaf buds 0.28 0.4528 0.2996 2.83 5.81 & leaves 35 Fruit 0.03 0.4367 0.2927 3.99 2.37 36 Flowers 0.10 0.4618 0.3667 4.01 3.85 37 Leaves 0.05 0.4497 0.3663 4.15 4.83 38 Leaves 0.06 0.4345 0.3373 4.30 5.44 39 Leaves 0.08 0.2072 0.1571 3.99 7.80 41 Leaves 0.12 0.3264 0.2017 3.55 5.58 42 Leaves 0.12 0.1652 0.1311 3.50 4.71 43 Leaves 0.07 0.5198 0.3931 4.48 4.26 44 Fruit 0.06 0.2094 0.1341 4.20 3.50 45 Leaves 0.05 0.3341 0.3030 4.41 3.33

121

ID Parts Available NDF (%) ADF (%) NPGE Ash (%) selected protein (%) (0.1=10%) (0.1=10%) (kcal/g) (0.1=10%) (0.1=10%) 46 Leaves 0.12 0.2081 0.1622 4.20 6.28 47 Leaves 0.10 0.2790 0.1902 3.36 6.48 48 Leaves 0.08 0.5285 0.4373 3.02 3.33 49 Leaf buds 0.11 0.4373 0.2762 3.61 8.21 50 Leaves 0.13 0.4531 0.3550 4.61 3.52 51 Leaves 0.12 0.5156 0.2917 4.15 4.38 52 Leaves 0.03 0.5249 0.3502 3.87 2.15 53 Leaves 0.06 0.4149 0.3442 4.89 4.28 54 Leaves 0.07 0.2489 0.2065 4.15 3.89 55 Leaves 0.11 0.3894 0.2549 4.33 3.89 56 Leaves 0.12 0.4473 0.2718 3.73 4.00 57 Leaves 0.11 0.2362 0.2041 3.62 5.30 59 Leaves 0.05 0.4020 0.3121 4.57 3.40 60 Leaf buds 0.07 0.4691 0.3220 3.89 12.82 & leaves 61 Leaves 0.09 0.3974 0.3277 4.87 5.93 62 Leaf buds 0.18 0.1984 0.1275 3.73 4.80 & leaves 63 Leaves 0.17 0.2132 0.1524 3.63 3.66 64 Leaf buds 0.08 0.2750 0.1824 4.09 5.40 & leaves 65 Leaves 0.08 0.3812 0.3619 4.48 3.19 66 Fruit 0.08 0.1926 0.1455 5.07 5.19 67 Bark 0.00 0.7917 0.7517 4.53 3.42 68 Fruit 0.12 0.3642 0.2268 4.07 4.76 69 Leaves 0.20 0.2218 0.1208 3.63 7.19 70 Leaves 0.16 0.2110 0.1437 3.45 4.94 72 Leaves 0.15 0.3080 0.2572 3.77 10.03 74 Leaves 0.13 0.3480 0.1790 3.70 4.96 76 Fruit 0.06 0.1618 0.1122 3.96 3.14 78 Leaves & 0.17 0.2905 0.2266 4.11 6.50 fruit 79 Leaves 0.18 0.5016 0.3854 1.68 5.20 80 Leaves 0.09 0.1652 0.1453 4.08 3.27 81 Flowers 0.07 0.2485 0.2141 4.14 2.91 82 Leaves 0.09 0.2633 0.2174 4.31 4.62 83 Fruit 0.03 0.6153 0.3863 3.77 3.33 85 Leaves 0.05 0.6097 0.5012 4.55 3.42 86 Leaves 0.22 0.3499 0.2161 3.03 7.24 87 Leaves 0.13 0.3079 0.2007 3.88 5.22 89 Fruit 0.05 0.3079 0.1714 3.90 2.61 91 Leaves 0.12 0.4365 0.3054 3.78 4.04 92 Leaves 0.11 0.1607 0.1243 3.67 4.83

122

ID Parts Available NDF (%) ADF (%) NPGE Ash (%) selected protein (%) (0.1=10%) (0.1=10%) (kcal/g) (0.1=10%) (0.1=10%) 93 Leaves 0.13 0.4244 0.1355 3.27 10.79 95 Leaves 0.21 0.2538 0.1579 2.89 7.81 96 Leaf buds 0.08 0.3172 0.1876 4.72 4.35 & leaves 97 Flowers 0.19 0.2125 0.2504 3.48 5.24 98 Leaves 0.20 0.2000 0.1372 3.31 6.27 99 Fruit 0.07 0.4851 0.3973 4.39 3.50 100 Fruit 0.13 0.3577 0.1532 3.37 5.36 101 Fruit 0.03 0.5675 0.4473 4.27 - 103 Leaves 0.20 0.2773 0.1985 3.28 4.9437 105 Fruit 0.08 0.4329 0.3305 3.69 3.5251 106 Leaf buds 0.14 0.3620 0.2373 3.86 - 107 Fruit 0.02 0.2517 0.0469 4.01 1.78 108 Fruit 0.15 0.4370 0.3276 4.87 2.99 109 Fruit 0.03 0.8099 0.6503 4.33 2.70 111 Fruit 0.08 0.4026 0.3118 5.35 3.4757 113 Leaves 0.08 0.4589 0.3471 4.27 5.7855 114 Leaves 0.10 0.1920 0.1533 3.93 4.8023 115 Leaves 0.16 0.4024 0.1501 4.03 4.5498 117 Leaves 0.16 0.2734 0.2146 3.95 3.8205 118 Leaves 0.17 0.3233 0.2273 3.50 3.6684 119 Leaves 0.12 0.3340 0.2024 2.99 4.8840 120 Fruit 0.04 0.1882 0.1202 3.72 3.3172 121 Fruit 0.02 0.5837 0.3780 4.34 2.7742 123 Leaves 0.13 0.2159 0.1708 3.60 3.4582 124 Fruit 0.03 0.5799 0.3882 3.87 2.1172 125 Fruit 0.05 0.5034 0.3831 3.94 4.1322 126 Fruit 0.03 0.2118 0.1564 4.29 2.8614 127 Leaves 0.07 0.4350 0.3577 4.65 5.2676 128 Fruit 0.06 0.1988 0.1563 4.41 4.3091 129 Fruit 0.05 0.1884 0.1072 4.19 3.1287 130 Fruit 0.09 0.2846 0.2384 3.88 8.7026 131 Leaves 0.22 0.2190 0.1335 3.37 6.9366 132 Flowers 0.10 0.0780 0.0576 3.61 3.66 134 Bark -0.01 0.7122 0.6266 3.67 11.2234 135 Leaves & -0.02 0.8178 0.7268 4.34 2.6749 bark 136 Flowers 0.06 0.4287 0.3842 3.70 - 139 Fruit 0.12 0.3024 0.2581 3.60 11.1994 140 Fruit 0.06 0.4324 0.3714 3.99 3.7415 143 Leaf buds 0.07 0.5272 0.4157 4.27 3.82 & fruit 144 Fruit 0.10 0.3162 0.2262 4.29 2.45

123

ID Parts Available NDF (%) ADF (%) NPGE Ash (%) selected protein (%) (0.1=10%) (0.1=10%) (kcal/g) (0.1=10%) (0.1=10%) 145 Fruit 0.10 0.6352 0.2936 4.89 2.18 146 Flowers 0.09 0.0885 0.0724 4.35 3.04 156 Flowers 0.13 0.2680 0.1732 3.30 6.90 157 Leaves 0.18 0.2162 0.1367 3.52 4.18 159 Leaves 0.13 0.1065 0.0854 3.47 5.87 161 Leaves 0.19 0.1615 0.1255 3.25 6.33

124

Table 3.3. Nutrient profile of high available protein foods consumed by lactating P. coquereli (n=123) Cluster Category Mean (%) SE Lower Upper bound bound 95% CI 95% CI 1 Available 15.9 0.6 14.6 17.1 protein

2 Fiber 4.4 0.5 3.3 5.5

3 Non-protein 7.7 0.4 6.9 8.5 gross energy

125

Table 3.4. Nutrient profile of high neutral detergent fiber (NDF) foods consumed by lactating P. coquereli (n=123) Cluster Category Mean (%) SE Lower Upper bound 95% bound CI 95% CI 1 Available 29.5 1.5 26.4 32.5 protein

2 Fiber 59.0 2.4 54.2 63.8

3 Non-protein 30.7 1.9 26.8 34.6 gross energy

126

Table 3.5. Nutrient profile of high acid detergent fiber (ADF) foods consumed by lactating P. coquereli (n=123) Cluster Category Mean (%) SE Lower Upper bound 95% bound CI 95% CI 1 Available 19.8 1.0 17.9 21.7 protein

2 Fiber 45.3 2.4 40.4 50.1

3 Non-protein 22.2 1.5 19.2 25.3 gross energy

127

Table 3.6. Nutrient profile of high non-protein gross energy foods consumed by lactating P. coquereli (n=123) Cluster Category Mean SE Lower Upper (kcal/g) bound bound 95% CI 95% CI 1 Available 3.4 0.05 3.4 3.6 protein

2 Fiber 4.0 0.07 3.8 4.1

3 Non-protein 4.3 0.06 4.2 4.4 gross energy

128

Table 3.7. Nutrient profile of minerals (ash) in foods consumed by lactating P. coquereli (n=119) Cluster Category Mean (%) SE Lower Upper bound bound 95% CI 95% CI 1 Available 5.9 0.3 5.3 6.5 protein

2 Fiber 4.8 0.7 3.3 6.3

3 Non-protein 3.9 0.2 3.5 4.4 gross energy

129

Table 3.8. Frequency of ingestion index (FOI) (n=123) Cluster Category N Mean Highest FOI (score of 5) 1 Available 44 15.9± 0.6% 5% protein

2 Fiber 33 59.0±2.3% (NDF) 42% 45.3±2.4% (ADF)

3 Non-protein 46 4.3±0.6 kcal/g 54% gross energy

130

Figure 3.1. Ankarafantsika National Park, northwestern Madagascar

Chapter 4 Cortisol Variation Across Sex and Reproductive Classes in Propithecus coquereli

“Nothing is so full of victory as patience.” “Fa tsy misy fandresena tahaka ny faharetana.” ~ Malagasy proverb

4.1. Abstract

Acute and chronic stressors are omnipresent in all environments inhabited by animals and elucidate a multiplicity of physiological and behavioral adaptive responses. Though stress responses are adaptive in the short term, they pose serious health consequences when experienced over longer durations. The hypothalamic-pituitary-adrenal axis releases glucocorticoid steroids into the blood stream and energy reserves are activated to combat stressors. I evaluated cortisol in Coquerel’s sifaka (Propithecus coquereli) lactating females, adult males, and non-lactating females in early/earlier-mid relative to later-mid/late lactation. Analyses were based on 375 fecal samples collected over two consecutive birth seasons (2010 and 2011). Cortisol, a glucocorticoid, is a non-invasive, highly effective temporal measure of cortisol concentrations. I collected feces weekly from P. coquereli lactating females (n=180) after infants were first observed from 1-24 weeks postnatal, for comparison with adult males (n=133), and non-lactating adult females (n=62) in Ankarafantsika Park, northwestern Madagascar. Adult males had significantly higher cortisol during weeks 13-24 relative to weeks 1-12. Lactating females had lower cortisol relative to adult males and significantly lower cortisol than non-lactating females. Evaluating stress responses temporally assists in determining the costs of seasonal pressures across sex and reproductive classes during the critical stage of infant development in the highly unpredictable environment of Madagascar.

4.2. Introduction

Environmentally demanding conditions have led to evolved behavioral and physiological strategies to aid in coping with energetically challenging or depleting situations. Allostasis is the

131 132

regulation of internal bodily condition and energetic balance while enduring environmental changes to achieve homeostasis. The ambiguous term “stress” has been used to describe a suite of both biological and physiological processes and responses and thus, no universally clear definition currently exists (see Bonier et al. 2009a; Busch and Hayward 2009; Sheriff et al. 2011). For the purposes of continuity, the term stress is broadly defined as “the biological response elicited when an individual perceives a threat [defined as a stressor] to its homeostasis” (Moberg 2000, p. 1). Sources of stressors, defined as the specific threats to homeostasis, are environmental, physiological, or behavioral and may either be short-term or long-term depending on the situation the individual is experiencing. Fleeing from a predator is an example of an acute stressor, whereas extreme seasonal variation in rainfall is a chronic stressor. Individuals experience a stress response when a stressor is present (Moberg 2000). This stress response is the body’s attempt to reestablish homeostasis through behavioral and physiological mechanisms (Nelson 2000).

Glucocorticoids (GCs) are steroid hormones released from the adrenal cortex, anatomically positioned along the edge of the adrenal gland on top of the kidneys, and epinephrine is released from the adrenal medulla, the center portion of the adrenal cortex (Chrousos et al. 1995). Figure 4.1 elucidates the hypothalamic-pituitary-adrenal (HPA) axis after it has been activated by the neuroendocrine stress response and GCs are secreted to restore homeostasis. It also illustrates the negative feedback response to chronic stress, acute stress, and the effects of these stressors on bodily processes (Boonstra et al. 1998; Chrousos et al. 1995; Sheriff et al. 2011). The HPA axis is composed of the hypothalamic paraventricular nucleus (PVN), anterior pituitary gland, and adrenal cortex (Sheriff et al. 2011). The GCs, corticosterone and cortisol, are referred to as the stress hormones, though also play a principal role in metabolic function and regulation by affecting glucose release (Selye 1937). Cortisol is the specific metabolic and stress hormone present in non-human primates. The PVN is stimulated when a stressor is experienced, in turn causing the release of corticotrophin-releasing hormone along with other related hormones (Sheriff et al. 2011). It takes approximately 3-5 minutes to show measurable increases in GC concentrations after the PVN is stimulated (Sheriff et al. 2011).

Measuring variation in GCs can be used to quantify the stress responses of an individual, not the stress incurred by the individual (see Sheriff et al. 2011). This distinction is particularly crucial

133

when measuring how individuals cope with energetically demanding conditions since “stress is the recognition by the body of a stressor and therefore, the state of threatened homeostasis; stressors are threats against homeostasis, and adaptive responses are the body’s attempt to counteract the stressor and reestablish homeostasis” (Chrousos et al. 1988, p. 4). Cortisol mobilizes energy reserves, thus making its increased production advantageous in the short-term, but poses long-term pathological health risks (Brockman et al. 2009; Sapolsky 1992; Selye 1937). Acute physiological effects during a stress response include: increased immediate energy availability, increased oxygen intake, decreased blood flow to areas not critical for movement, enhanced memory and sensory function, and the inhibition of digestion, growth, immune function, reproduction, and pain perception (Nelson 2000). Thus, the terms “stress” and “stressor” do not inherently signify negative health consequences, but instead represent a spectrum of adaptive bodily responses utilized in response to stimuli. In contrast, chronic stress responses cause negative pathological health effects including: fatigue, hypertension, ulcers, reproductive suppression, impaired immune function, cancer, and accelerated neural degeneration during aging (Sapolsky 1992).

Ascertaining the biological cost of the stress to an individual determines whether the stress response negatively impacts the welfare of the animal (Keay et al. 2006; Moberg 2000). Figure 4.2 illustrates the three phases of the biological stress response: recognition of a threat to homeostasis, the stress response, and the consequences of stress (Moberg 1999). A stress response is initiated when the central nervous system identifies a threat to homeostasis. The organization of biological defenses begins after the threat is recognized and includes a combination of behavioral, autonomic, neuroendocrine, and immunological responses (Moberg 1999). The consequences of stress replace normal biological function with altered function, and when prolonged lead to a prepathological state and eventual development of pathologies (Moberg 1999). Neuroendocrine responses (e.g., cortisol concentrations) and the duration of stress responses are primary measures of biological cost to an animal (Keay et al. 2006). Distress along with pathological stress responses are experienced when coping with the stressor redirects resources away from other biological functions, thereby increasing an animal’s vulnerability to stressor induced pathologies (Moberg 2000). Thus, longitudinally evaluating cortisol offers the most comprehensive view in distinguishing between acute and chronic stress responses in non-human primates.

134

Several physiological indicators have been used to determine stress levels (e.g., heart rate), although measuring cortisol is one of the most accurate and widely applied methods. Cortisol may be extracted from a variety of biofluids including milk, saliva, urine, blood, and feces. Milk acts as a pathway for glucocorticoid signaling that can influence offspring growth and behavioral phenotype (Hinde et al. 2015). For example, milk cortisol is associated with infant temperament in captive rhesus macaques (Macaca mulatta), with higher maternal-origin GCs producing more nervous and less confident offspring (Hinde et al. 2015). Fecal collection is the least invasive method for quantifying cortisol in field studies (see Busch and Hayward 2009; Keay et al. 2006; Sheriff et al. 2011), as animals are not near during collection, handled, or anesthetized. Fecal glucocorticoids (fGCs) are a biometric measure even more accurate than blood cortisol in gauging stress responses since blood collection introduces methodological biases by artificially raising cortisol concentrations during animal capture, which are reflected in the interpretation of results (Keay et al. 2006). Blood cortisol concentrations vary according to circadian rhythms and normal pulsatile fluctuations, and subsequently measuring blood cortisol captures only a snapshot in time, while feces are a representation of pooled cortisol and signify a greater temporal aggregate of the stress response (Keay et al. 2006; Sheriff et al. 2011).

The cortisol-fitness (C-F) hypothesis proposes that cortisol levels negatively co-vary with reproductive fitness, where relatively higher levels of baseline cortisol represent individuals in worse physical condition, and thereby, reduced fitness in comparison to individuals with lower cortisol concentrations (Bonier et al. 2009a). GCs fluctuate daily, seasonally, and interindividually. The C-F hypothesis stems from the tenet that environmental challenges trigger greater cortisol secretion, in turn causing a reallocation of resources from normal behavioral activities to cope with the environmental challenge at hand (Bonier et al. 2009a). This hypothesis has been applied to the interpretation of cortisol profiles in a variety of vertebrate taxa and has a mixture of evidence supporting and disputing the hypothesis (reviewed in Bonier et al. 2009a). A review article found that 74% of studies considering a variety of mammalian, avian, and reptilian taxa found a significant relationship between cortisol and fitness, thus supporting the cortisol-fitness hypothesis (Bonier et al. 2009a). Though the study found high variability in the relationships between cortisol and fitness, and suggested caution when using cortisol as the primary measure of relative fitness (Bonier et al. 2009a). In contrast, fGC levels in red-bellied lemurs (Eulemur rubriventer) were significantly higher and more variable in undisturbed forest

135

than in disturbed forest, contrary to predictions (Tecot 2013). The same study found fGCs were highest during the convergence of parturition and food scarcity, with high fGCs prevailing until fruit became more abundant and infants reached independence (Tecot 2013).

The cortisol-adaptation (C-A) hypothesis is a theoretical reconsideration of the C-F hypothesis, with the addition of all challenges related to reproduction as threats to allostasis and by extension, homeostasis (Bonier et al. 2009a). The effects of seasonality, reproductive state, diet, and gender expressively influence stress responses (Keay et al. 2006). The incorporation of reproductive events provides a more comprehensive view of the how environmental pressures influence stress responses relative to the C-F hypothesis since seasonal events regulate mating, pregnancy, postnatal hormonal function, and infant behavioral care. The inclusion of these variables is paramount in examining life history theory, especially since many reproductive and seasonal events occur in tandem (e.g., parturition and food scarcity) (see Tecot 2013). The majority of field studies on stress responses have focused on either avian species (e.g., Bonier et al. 2009b; Bonier et al. 2011) or, in mammalian species, African wild dogs (Lycaon pictus), wolves (Canis lupus), elk (Cervus elaphus), snowshoe hares (Lepus americanus), and artic ground squirrels (Spermophilus plesius) (see Boonstra et al. 1998; Boonstra et al. 2001; Creel et al. 1996; Creel et al. 2002; Creel et al. 2009; Monfort et al. 1998). Studies of GC stress responses in non-human primates have chiefly been conducted on captive or semi-free ranging cercopithecines and callitrichids (reviewed in Abbott et al. 2003; e.g., da Silva Mota et al. 2006; Saltzman and Abbott 2009; Setchell et al. 2008; Ziegler 2000; Ziegler et al. 1995). Comparatively few studies have examined cortisol in wild strepsirhines (Brockman et al. 2009; Cavigetti 1999; Gould et al. 2005; Oster et al. 2008). A phylogenetic meta-analysis of 64 bird species determined that peak corticosterone was highest in birds that bred in higher altitudes as a result of greater climatic constraints in comparison to bird species that bred at lower altitudes (Bókony et al. 2009). Cortisol levels in male bluegill sunfish (Lepomis macrochirus) during the early breeding season were negatively correlated with reproductive success; males were more likely to abandon their nests, and males that had the highest baseline cortisol abandoned their nests before other males that also eventually abandoned nests (Bonier et al. 2009a). Fecal GCs were significantly higher in red deer (Cervus elaphus) during the cold winter months when resources were severely limited (Huber et al. 2003). Similarly, cortisol levels in male muriquis (Brachyteles arachnoides) were significantly higher during the dry season when mating occurs,

136

and less food resources are available (Strier et al. 1999). Applying the C-A hypothesis will achieve a more representative view of stress responses than the C-F hypothesis since cortisol levels are reflective of both ecological and reproductive pressures (Bonier et al. 2009a).

The brood-value hypothesis originated from avian species (see Carlisle 1982; Nur 1984), and posits stress responses are determined by current reproductive value, where stress response are predicted to be reduced when reproductive value is high to ensure propagative success (Heidinger et al. 2006). It also predicts that stress responses will be attenuated in the sex providing more infant care (O'Reilly and Wingfield 2001). Female primates are the sex typically more involved in care-giving, particularly mothers or related females. Surprisingly, the brood- value hypothesis has not been applied to primates, which generally epitomize slow life histories compared to other mammals. Time spent by mothers rearing broods of the common goldeneye (Bucephala clangula) was dependent on mortality already experienced by the brood, not the predicted size of the brood (Pöysä et al. 1997). In a meta-analysis of bird species, corticosterone levels during the period of parental care were higher in species with higher reproductive value (Bókony et al. 2009). The same study found weaker corticosterone responses in females when the pair showed female-biased parental care (Bókony et al. 2009).

Forests in Madagascar are energetically poor, characterized by slow tree growth, low fruit quality, and poor soil fertility (Wright 1999). Harsh climatic conditions together with droughts and cyclones contribute to high rates of infant mortality in lemurs including Propithecus spp. (Gould et al. 1999; Richard et al. 1991; Tecot 2010). Propithecus spp. are characterized by remarkably slow life histories relative to their body size including late sexual maturity, small litter size, and 2-year interbirth intervals (Richard et al. 2002). Given this slow life history pattern, the value of every reproductive event in Propithecus spp. is intrinsically high. P. coquereli mothers are the primary infant care-givers (Ross and Lehman 2016). As such, a decreased stress response during later infant development would be advantageous for mothers to allocate more resources to prepare for future pregnancies and increase overall fecundity. Contrary to expectations, male-male competition for females and increased agonism during the breeding season had no effect on males GCs in ring-tailed lemurs (Lemur catta) (Gould et al. 2005). L. catta males residing in groups with fewer males had decreased GCs and male group composition likely influences stress responses (Gould et al. 2005). The authors suggest that

137

future consideration understanding the nuances or potential overlap between life history stages (i.e., migration, breeding) and emergency life history stages (i.e., limited resources, habitat destruction) is an avenue of critical importance to understand physiological stress responses in different seasons and life history stages (Gould et al. 2005). GCs were significantly higher during both breeding and births seasons compared to other times of the year in male redfronted lemurs (Eulemur fulvus rufus) (Ostner et al. 2008). The authors suggest that heighted intra-group competition during reproduction and infanticide risk cause elevated GCs during the breeding and birth season, respectively. Aggression and GCs in Verreaux’s sifaka (Propithecus verreauxi) males were significantly higher during the breeding compared to the birth season, with dominant males displaying higher GCs in the breeding season relative to subordinate males (Fichtel et al. 2007b). The authors conclude that higher GCs in dominant males is the result of high energetic demands associated with male reproduction (Fichtel et al. 2007b). No previous studies have focused on weekly temporal cortisol variation in different reproductive/sex classes during lactation or tested the C-A and brood-value hypotheses in Propithecus spp. until this dissertation. Determining baseline cortisol concentrations during two lactation phases and successively examining temporal cortisol variation illuminates how the seasonal events of infant birth and development influence stress responses among different sex/reproductive classes of individuals.

I hypothesized that P. coquereli mothers experienced greater postnatal reproductive stress during early/earlier-mid lactation (designated as 1-12 weeks postnatal) (May-August) in comparison to later-mid/late lactation (designated as 13-24 weeks postnatal) (September-December) due to the increased energetic costs of lactation while simultaneously caring for dependent infants during the driest seasonal months in the austral winter when food quality is low. I tested this hypothesis by measuring P. coquereli behavioral care-giving effort, nutritional food quality, and cortisol stress responses. This chapter specifically examines temporal variation in P. coquereli lactating females, adult males, and non-lactating females by addressing three predictions posited from the C-A and brood value hypotheses. P1. P. coquereli lactating females will have significantly higher fecal cortisol levels during early/earlier-mid lactation relative to later-mid/late lactation. P2. P. coquereli adult males will have significantly higher fecal cortisol levels during early/earlier-mid lactation relative to later-mid/late lactation.

138

P3. P. coquereli lactating females will have significantly lower fecal cortisol levels relative to adult males and non-lactating adult females during both lactation phases.

4.3. Methods 4.3.1. Study site and species I conducted my study at the Ampijoroa Forestry Station in Ankarafantsika National Park (ANP), located in the Mahajanga Province in northwestern Madagascar (Figure 4.3). The Ankarafantsika region (135,000 ha) was first established as two protected areas in 1927 and recognized as a national park in 2003. Ampijoroa is situated in the southwestern portion of ANP. The Marovoay area (38,000 ha) found within the Ankarafantsika region is the second largest producer of rice, the primary subsistence crop, in Madagascar (Alonso and Hannah 2002). The GPS coordinates of the research base camp are 16°18’31” South, 46°48’49” East and sits 88 m above sea level. ANP is characterized as a dry deciduous forest with an exceptionally pronounced dry season (Alonso and Hannah 2002; Du Puy and Moat 1996). Forested areas surrounding Ampijoroa are experiencing anthropogenic disturbance from slash and burn agriculture, fire, relatively high volumes of human traffic, unregulated presence and herding of domestic cattle, bushmeat hunting, and hole digging for Dioscorea maciba (Family) tuber extraction (Alonso and Hannah 2002; Crowley et al. 2012; Gerardo and Goodman 2003). The underlying geological formation in the Ankarafantsika region is composed of sandstones and the study area sits atop a sandstone plateau between 310-340 m above sea level (Du Puy and Moat 1996; Lourenço and Goodman 2006). Soils are either red, speckled or white, with red soil containing the highest water content and white sand the lowest (Crowley et al. 2012). Red soil most often occurs in the forest edge and in the savannah itself, though is also present in the forest interior and is presumably more nutrient dense due to its higher water content than forest interior quartz white sands (Crowley et al. 2012). Many trees grow in nutrient poor, acidic white sands and a thick layer of loose sand is present on the soil surface as a result of sandstone erosion (Du Puy and Moat 1996; Lourenço and Goodman 2006). Flora are speciose and the forest understory is moderately thick with sparse leaf litter (Lourenço and Goodman 2006) Annual precipitation in ANP ranges from 1100-1600 mm, with the majority of rainfall occurring in January and February and a dry period from May to September in which there is very little rainfall (Rendigs

139

et al. 2003). Average daily temperatures range from 16 °C during the dry season to 37 °C in the wet season (Rendigs et al. 2003).

There are eight extant lemur species in ANP including: Coquerel’s sifaka (Propithecus coquereli), common brown lemur (Eulemur fulvus), mongoose lemur (Eulemur mongoz), western woolly lemur (Avahi occidentalis), lesser dwarf lemur (Cheirogaleus medius), Milne-Edwards' sportive lemur (Lepilemur edwardsi), Gray mouse lemur (Microcebus murinus), and the golden- brown mouse lemur (Microcebus ravelobensis) (Alonso and Hannah 2002). The International Union for Conservation of Nature estimates that the P. coquereli has experience rapid population declines of more than 50% during the last 30 years primarily due to habitat loss (Andrainarivo 2008, accessed 17 November 2013) and more recently, hunting for human consumption (Gerardo and Goodman 2003). The slow life histories of Propithecus spp. likely increases extinction risk (Purvis et al. 2000) and amplifies the devastating impact of hunting given Propithecus spp. wait until relatively later in life to reproduce, have interbirth intervals of every other year, and only one infant is present per social group (Pochron et al. 2004; Richard et al. 1991). An estimated population of ~47,000 P. coquereli currently live in ANP (Kun-Rodrigues et al. 2014). Density estimates range from 5-100 individuals/km2, with habitat quality (i.e., negative effects of roads and forest edges) as the principal factor relating to this high variability (Kun-Rodrigues et al. 2014). An estimated 5 individuals/km2 presently exist in Ampijoroa (Kun-Rodrigues et al. 2014), in comparison to an estimated 60-75 individuals/km2 in the 1980s (Albignac 1981). This is a rapid decrease of more than 90% of the P. coquereli population in Ampijoroa (Kun- Rodrigues et al. 2014).

This chapter examines temporal variation in fecal cortisol concentrations in P. coquereli mothers, adult males, and non-lactating females starting in June during the less resource abundant dry season until the resource abundant wet season in December. I first compare cortisol concentrations between P. coquereli mothers, adult males, and non-lactating females beginning with infant births and continuing until the weaning period. Secondly, I consider cortisol concentrations in these three classes of individuals in relationship to infant age. Groups were first located when no infants were present. These same groups were rechecked weekly until an infant was present (n=9). For example, Group Rambo was first located with no infant present on 05/28/2011, subsequently located on 06/06/2011 with no infant present, and relocated on

140

06/11/11 with an infant present. Therefore, the infant was a maximum of 5 days old. The number of days that elapsed between the initial and subsequent group locations determined maximum infant age. If a group was initially located with an infant present, infant age was estimated based on relative body size seen in other infants where precise birth date was known. Thus, the number of weeks postnatal (1-24) was a conservative estimate of infant age.

4.3.2. Fecal collection and preservation Fecal samples were collected weekly from P. coquereli lactating adult females, adult males, and non-lactating adult females in seven focal groups from July to December 2010 (season 1, n=221) and three focal groups from June to December 2011 (season 2, n=154) for a total sample size of 375 (Table 4.1). A total of 180 fecal samples were collected from lactating females, 133 from adult males, and 62 from non-lactating adult females. Samples were collected from non-lactating adult females from weeks 4-23 postnatal. Lactating females and adult males are representative of 1-24 weeks postnatal. Individual adult males and non-lactating females were not distinguished from each other during collection when more than one adult male and one non- lactating female were present in each group. Thus, it is possible that multiple samples were collected from the same individuals. Weekly fecal sample collections were uneven due to differences in the timing of infant births, no observed defecations during the focal period, and to ensure correct individual identification. The total amount of feces in each defecation was collected from individuals when P. coquereli groups began to move from their sleeping sites in the early morning. Feces were either caught in the air with a plastic sheet before hitting the ground or collected immediately following defecation from the ground if the individual was clearly identified and no other group members defecated simultaneously. Feces were placed in aluminum foil, flattened to increase surface area, and transported within a maximum of four hours to an enclosed propane gas oven heated to 83C and dried for 2-3 hours or until thoroughly dried (following Brockman and Whitten 1996). Temperature was measured using a water- resistant max/min digital thermometer (HBE International Inc.) and monitored every twenty minutes to avoid overheating and potential hormone degradation. Dried feces were individually stored in manila paper coin envelopes (14x8cm) in 3M SCC Dri-Shield 2000 moisture barrier bags (65x50cm) with silica gel. Dri-Shield moisture barrier bags were stored in a plastic container with silica gel.

141

4.3.3. Dried feces sample preparation Dried feces were pulverized using a ceramic mortar and pestle sterilized with ethyl alcohol (90%-100%) to create a homogeneous particle distribution. The pulverized sample was sifted through a 1 mm mesh filter and 0.1-0.2 g of feces were scooped with a sterile spatula into a polypropylene tube.

4.3.4. Cortisol steroid extraction and recovery Laboratory assays were conducted at the Wisconsin National Primate Research Center, Madison, Wisconsin, USA under of the supervision of Mr. Dan Wittwer and Dr. Toni Ziegler. I added 2.5 mL distilled H2O and 2.5 mL ethanol to the polypropylene tubes, vortexed for ten minutes, and centrifuged at 2000 RPM for ten minutes. The supernatant was poured into minivials and 1 mL was aliquoted into glass extraction tubes. A total of 4 mL of ethyl acetate was added to extract cortisol from the H2O/ethanol solution, vortexed for eight minutes, and centrifuged at 1000 RPM for three minutes. The top serum layer was aspirated and evaporated in a hot water bath. I added 1 mL of ethanol after evaporation and extracted samples were kept refrigerated until enzyme immunoassays. Cortisol extraction recovery rate is used to determine steroid extraction efficiency by measuring the procedural loss incurred during the extraction phase. A fecal sample was randomly selected to determine the recovery rate. Titrated cortisol was extracted with an established recovery rate of 75%, a recovery rate that falls within the range of previous studies (following Brockman and Whitten 1996). Cortisol concentrations were adjusted for this procedural loss prior to statistical analyses (see Appendix 1 for cortisol concentrations).

4.3.5. Cortisol steroid validation Validation was conducted to determine if circulating cortisol levels were subsequently reflected in the measured concentrations of fecal cortisol, and to ensure the solution volume did not influence cortisol values (reviewed in Touma and Palme 2005). Cortisol was validated successfully on the standard curve after charcoal stripping. Charcoal stripping acts through adsorption, where the solid binds with the cortisol. Approximately 20 µl charcoal treated feces were used in validation and subsequently added to the standard curve during enzyme immunoassays.

142

4.3.6. Enzyme immunoassays (EIA) Extracted cortisol, EIA buffer, six standards, low and high pools were brought to room temperature and vortexed. Sample (20 µl) was aliquoted into culture tubes and 20 µl charcoal stripped feces were added into standard tubes and evaporated in a hot water bath. The horseradish peroxidase (HRP) and EIA buffer solution were prepared by calculating volume based on the number of EIA plates (e.g., 25 mL EIA buffer and 20 µl HRP for two EIA plates). EIA buffer (50 µl) was added to culture tubes, 300 µl EIA buffer was added to the non-specific binding (NSB) EIA tubes, 50 µl standard was added to standard tubes, and 250 µl F:HRP was added to culture and standard tubes, vortexed, and multi-channel pipetted into a cortisol coated plate. The cortisol plate was shook for five minutes and incubated for two hours. The substrate

(25 mL citrate, 80 µl H2O2, and 250 µl ABTS [color changing solution] for two EIA plates was applied to the cortisol plate, shook for five minutes, and incubated for 45 minutes. The stop solution (12.5 mL hydrochloric acid, 30 µl ethylenediaminetetraacetic acid, and 15mL distilled water for two EIA plates) was added and the results read in a SpectraMax microplate reader. The cortisol plate equation is (Wisconsin National Primate Research Center, 2012):

[Absorption well – NSB absorption /

B0 absorption (100%, only HRP, no cortisol) – NSB absorption (0%)]

The low pool coefficient of variation was 15.46/5.17 and the high pool was 15.56/2.09, which fell within the expected range of inter and intra-assay variation. The same pools were used in 2011 and 2012. Cortisol concentrations were averaged prior to statistical analyses if more than one fecal sample per week was collected from the same individual.

4.3.7. Data analysis Box and whisker plots illustrating cortisol concentrations in nanograms per gram (ng/g) in lactating females, adult males, and non-lactating females from 1-24 weeks postnatal (4-23 in non-lactating females) were constructed weekly for each of the ten P. coquereli groups. Box and whisker plots were created from the weekly minimum, first quartile, second quartile (median), third quartile, and weekly maximum. These values were used to generate the 25th, 50th, and 75th

143

percentile boxes. Subtracting weekly minimums from the first quartile and the third quartile from the maximum created whiskers. Mixed effects linear regression models in SAS® Version 9.3 were applied to determine the relationships between between cortisol and sex/reproductive class, where:

Cortisol = intercept + group + individual sex/ reproductive class; lactation phase + error term

Mixed effect linear regression models were selected for their robusticity to missing cortisol values. Group ID and sex/reproductive (lactating females, adult males, non-lactating adult females) class were treated as the between-subject effects. Lactation phase (early/earlier-mid: weeks 1-12 postnatal; later-mid/late lactation: weeks 13-24 weeks postnatal) was treated as the within-subject effect. I used a repeated measure as there were multiple observations over time within the same sifaka group, sex/reproductive class, and lactation phase combination. Pseudoreplication was partially accounted for by the within-group and within-subject correlations in the mixed effect linear regression models, given that individuals from the same group are expected to be more correlated than individuals between groups. Since observations from the same individual were treated as observations from the same sifaka group, the within subject correlation helps to account for individual dependence. The assumption of normality using the residuals was satisfied (µ=0.08, Σ=53.98).

4.4. Results 4.4.1. Lactating females Mean cortisol in lactating P. coquereli (n=10) from weeks 1-12 postnatal was 185.65 ± 45.59 ng/g (n=81) and 189.60 ± 57.10 ng/g in weeks 13-24 (n=99) (Table 4.2). Mean cortisol considering all postnatal weeks was 187.83 ± 52.13 ng/g (n=180) (Table 4.3). Median cortisol ranged from 157.97 ng/g in week 5 to 280.51 ng/g in week 2 (Figure 4.4). Minimum cortisol was 152.78 ng/g and occurred in week 6 (Figure 4.4). Maximum cortisol was 280.51 ng/g and occurred in week 2 (Figure 4.4). A mixed effects linear regression model demonstrated that lactating females had higher cortisol during weeks 13-24 in comparison to weeks 1-12, but the result was not significant (p=0.56) (Tables 4.4, 4.5). Higher cortisol concentrations were present

144

in later-mid/late lactation given that the estimate between early/earlier-mid and later-mid\late lactation was negative (Tables 4.4, 4.5).

4.4.2. Adult males Mean cortisol in adult male P. coquereli (n=19) 1-12 weeks postnatal was 189.78 ± 59.66 ng/g (n=64) and 209.90 ± 57.64 ng/g in weeks 13-24 (n=69) (Table 4.2). Mean cortisol considering all postnatal weeks was 200.22 ± 59.26 ng/g (n=133) (Table 4.3). Median cortisol ranged from 121.48 ng/g in week 3 to 284.63 ng/g in week 5 (Figure 4.5). Minimum cortisol was 138.33 ng/g and occurred in week 3 (Figure 4.5). Maximum cortisol was 265.65 ng/g and occurred in week 5 (Figure 4.5). Adult males had significantly higher cortisol during weeks 13-24 relative to weeks 1-12 (p=0.04) (Tables 4.4, 4.5). Higher cortisol concentrations were present in later-mid/late lactation given that the estimate between early/earlier-mid and later-mid/late lactation was negative (Table 4.5).

4.4.3. Non-lactating adult females Mean cortisol in non-lactating adult female P. coquereli (n=8) from weeks 1-12 was 207.12 ± 75.03 ng/g (n=27) and 212.43 ± 64.43 ng/g in weeks 13-23 (n=35) (Table 4.2). Mean cortisol considering all postnatal weeks was 210.12 ± 68.70 ng/g (n=62) (Table 4.3). Median cortisol ranged from 155.13 ng/g in week 6 to 306.87 ng/g in week 23 (Figure 4.6). Minimum cortisol was 161.91 ng/g and occurred in week 6 (Figure 4.6). Maximum cortisol was 306.87 ng/g and occurred in week 23 (Figure 4.6). Non-lactating females had higher cortisol during later lactation, though this result was not significant (p=0.66) (Table 4.5).

4.4.4. Comparison across sex/reproductive classes Median cortisol in lactating females was 182.74 ng/g, 198.74 ng/g in adult males, and 204.20 ng/g in non-lactating females considering all weeks postnatal (Figure 4.7). Lactating females had lower cortisol at values that approached statistical significance during both lactation phases relative to adult males (p=0.07) (Table 4.5, Figure 4.8). Lactating females had significantly lower cortisol than non-lactating adult females in both lactation phases (p=0.01) (Table 4.5, Figure 4.8). Adult males had higher cortisol in weeks 13-24 compared to weeks 1-12 (Figure 4.8).

145

4.5. Discussion

Endocrinological processes provide an opportunity to acutely assess how individuals respond to environmental stressors. Glucocorticoids are released in response to stressors and improve short- term survival, but concomitantly shift resources away from reproduction and infant care-giving. Fecal hormone levels are reflective of steroid production rate, or cumulative secretion and elimination over several hours, and thus feces are less influenced by variation in hormonal secretions relative to plasma (reviewed in Touma and Palme 2005). Lactation is the most energetically constraining behavior mammals engage in, with daily energy expenditures escalating up to 150% (reviewed in Gittleman and Thomspon 1988). Lemur reproductive events occur under stringent seasonal parameters resulting from food scarcity during the dry season (see Wright 1999). Food scarcity coincides with the P. coquereli birth season, and consequently infants are weaned during the food abundant wet season (Richard 1976). Propithecus spp. infants have mortality rates of greater than 50% during the first postnatal year, yet factors contributing to high infant mortality have yet to be determined (Kappeler et al. 2009; Richard et al. 2002). This high rate coupled with the unusually slow life histories distinctive of Propithecus spp. makes examining temporal variation in stress responses during lactation an imperative conservation instrument while extensive forest degradation continues in Madagascar. Quantifying stress responses during an energetically depletive time provides insight as to how P. coquereli respond to the stressors associated with lactation, while simultaneously considering dry season effects in an extreme climate across different sex and reproductive classes of individuals. Longitudinal socioendocrinological studies emphasizing lactating females in Propithecus spp. have been noticeably absent until my dissertation. In this chapter, I evaluated three predictions constructed from the C-A and brood-value hypotheses to determine temporal variation of fecal cortisol in lactating females, adult males, and non-lactating adult females. P1. P. coquereli lactating females will have significantly higher fecal cortisol levels during early/earlier-mid lactation relative to later-mid/late lactation. P2. P. coquereli adult males will have significantly higher fecal cortisol levels during early/earlier-mid lactation relative to later-mid/late lactation. P3. P. coquereli lactating females will have significantly lower fecal cortisol levels relative to adult males and non-lactating adult females during both lactation phases.

146

Predictions 1 and 2 were not statistically supported. Prediction 3 was partially supported. Captivatingly, there is evidence to support the inverse relationship of predictions 1 and 2. For prediction 1, there was a trend towards the opposite expectation, though it was not significant: lactating females had higher cortisol during later-mid/late lactation (weeks 13-24) relative to early/earlier-mid lactation (weeks 1-12). For prediction 2, the inverse prediction was statistically significant. Adult males had higher cortisol during later-mid/late lactation relative to early/earlier-mid lactation. There was a strong trend nearing statistical significance for the first half of prediction 3. Lactating females had lower cortisol than adult males. The second half of prediction 3 was statistically significant. Lactating females had significantly lower cortisol relative to non-lactating adult females.

Parental-investment theory suggests that maternal and paternal effort is partially determined by the reproductive value of current offspring (Trivers 1972). The number of offspring and their survivability is contingent on the level of investment by parents. Limited resources force parents to make evolutionary choices and trade-offs to maintain their own fitness while provisioning their offspring (Stearns 1992). Propithecus spp. are distinguished by a suit of slow life history traits, including biennial interbirth intervals, only one infant birth per group, and high infant mortality rates (Richard et al. 2002). Accordingly, this low reproductive output makes the reproductive value of each infant exceedingly high, unless environmental variables prevent adequate parental care. Following with the brood-value hypothesis, where high reproductive value produces a decreased stress response, lactating P. coquereli had significantly lower cortisol than non-lactating females and a trend towards lower cortisol in comparison to adult males (Tables 4.2, 4.3). This dissertation is the first-time support for both the C-A and brood-value hypotheses have been substantiated in lemurs. A fundamental concept in stress physiology is that short-term stress responses are designed to improve fitness while chronic stress responses instigate a host of negative health consequences that inhibit growth, reproduction, digestion, and immune function (Sapolsky 2000). My findings provide a worthwhile juxtaposition to the energy conservation hypothesis (ECH), which suggests that Madagascar ecology and environmental seasonality have caused considerable energetic stress, especially in reproductive lemur females (Jolly 1966; Wright 1999). The ECH also predicts lemurs invest less in their offspring relative to other primates to conserve energy in a highly stochastic environment (Jolly

147

1966; see Wright 1999). The reduced stress response displayed by lactating females offsets the dual energetic responsibility of carrying infants for more than six months while concurrently lactating. A decreased stress response can also help prepare lactating females, who are most often the dominant female in the social group, for future pregnancies and increase overall fecundity.

Growth is an important indicator of health, and contributes to overall reproductive fitness (Clutton-Brock 1991). Growth can be examined both prenatally and postnatally. Propithecus is a genus that metabolically invests more in neonates per day than other strepsirhines (Young et al. 1990). High prenatal investment in lemurs is one of the pressures contributing to the evolution of female dominance (Young et al. 1990). Strepsirhines give birth to small neonates that grow extremely rapidly relative to haplorhines (Leigh and Terranova 1998; Leutenegger 1973). P. coquereli initially displays rapid postnatal growth, followed by a period of slow growth (Tennenhouse 2016). Weaning age is a strong predictor of the first growth phase (~2.2 years postnatal) in P. coquereli (Tennenhouse 2016). In males and females, the age at first reproduction was also found to be a significant predictor of first growth phase (Tennenhouse 2016). Cortisol concentrations influence relative body mass in a variety of primate taxa (reviewed in Tennenhouse 2016). Growing P. coquereli with elevated cortisol hair concentrations had relatively low body mass, and these same individuals had lower adult body masses in comparison to those with lower cortisol, and greater body mass during development and adulthood (Tennenhouse 2016). Thus, chronic elevation of cortisol during development influences adult body mass (Tennenhouse 2016). Postnatal growth is a critical time in it of itself that also intensely contributes to adult body mass in P. coquereli.

I argue that P. coquereli lactating females do not invest less, but instead heavily invest in the first six postnatal months opposed to later ontogenetic development. While lemurs may behaviorally invest less in terms of the total length of time infants are dependent on mothers compared to haplorhines, P. coquereli mothers do play an enormous role in infant care-giving relative to other group members during the first six months of life. P. coquereli infants are altricial during these first six months, especially considering their rapid postnatal growth. Mature black-and-white ruffed lemurs (Varecia variegata) weigh between 3.1-3.6 kg, whereas P. coquereli weigh between 3.7-4.3 kg (reviewed in Mittermeier 2010). Nonetheless, V. variegata infants develop

148

considerably more quickly, with infants reaching independence 3-4 months postnatal (Baden 2011; Morland 1990) in comparison to more than 6 months in P. coquereli (Ross and Lehman 2016). The reproductive value of each Propithecus spp. infant is greater in comparison to other lemurs of comparable body size with faster life histories. A decreased stress response is a physiological mechanism utilized by lactating P. coquereli that permits infants to receive care while allocating sufficient resources for self-preservation and future pregnancies.

Female social dominance is a common trait in Lemuriformes; however, the evolutionary history and proximate mechanisms of female social dominance remain ambiguous (reviewed in Petty and Drea 2015). The theory of mammalian sexual differentiation identities a suit of traits associated with female social dominance signifying hormonal masculinization including: enhanced body size, pronounced scent-marking, delayed puberty, and male-like external genitalia (reviewed in Petty and Drea). In non-human primates, female social dominance is thought to have evolved because of high reproductive costs (Jolly 1984; Richard 1974; Richard 1985), with several lemur genera exhibiting numerous traits associated with hormonal masculinization (reviewed in Petty and Drea 2015). When Eulemur spp. expressing female social dominance were compared to Eulemur spp. exhibiting egalitarian social patterns, the species expressing female social dominance also had masculinized androgen profiles (Petty an Drea 2015). These findings provide new evidence to support that female masculinization was the ancestral condition in strepsirhines (Petty and Drea 2015). Androgen production is metabolically costly and it remains unclear how pregnancy influences androgen production (reviewed in Petty and Drea 2015). Understanding long-term hormonal changes in the mother- fetal unit will help understanding the evolution of female social dominance. Propithecus spp. exhibit female social dominance, with females outranking males in feeding priority (Richard 1974), and subsequently gaining access to preferred or higher quality items. This is especially important during periods of decreased food abundance and quality in the dry, lactation season. Diminished access to higher quality foods could be a factor causing significantly higher cortisol during later-mid/late lactation relative to early/earlier-mid lactation P. coquereli adult males. Low food availability may also impact stress responses in lactating females since they exhibited higher cortisol during the later lactation stage. The seasonal period of extreme desiccation is May-September and January receives the highest monthly rainfall (Rendigs et al. 2003). The calendar months of June-August coincide with the early/earlier-mid lactation, and September-

149

December coincides with later-mid/late lactation. September is the month with the harshest seasonal conditions, and these difficult climatic conditions likely spill over into November before food abundance and availability begins to improve. P. coquereli males are constrained by female feeding priority throughout the year and an elevated stress response beginning during the most climatically severe month supports that seasonal environmental pressures impact stress responses. Another possibility is that males are beginning to prepare for the next breeding season in January-February and inter- and intra-group agonism and competition may begin to increase during later lactation. Female P. verreauxi display age and rank related asynchronous receptivity, where dominant females are the first group members to mate (Brockman 1999). The same study found that female-female competition coincided singularly with synchronous periovulatory receptivity, suggesting female competition over breeding males (Brockman 1999). Dominant female feeding priority over lower ranking females may be a factor contributing to significantly higher cortisol in non-lactating females in comparison to lactating females. Subordinate females were always non-mothers (A. Ross, pers. obs.), which may suggest hormonal suppression of ovulation in subordinates. Future investigations of group dynamics and fecal cortisol during gestation and lactation will illuminate why different sex and reproductive class respond differently to stressors.

The presence of infants was correlated with GCs in P. verreauxi males (Brockman et al. 2009) (see Table 4.6 for a comparison between P. coquereli and P. verreauxi). Thus, infant presence is may cause a stress response in Propithecus spp. males. The infant mortality rate during my data collection was 9% in the first 6.5 months postnatal. This is a striking contrast to a rate of more than 50% documented in P. verreauxi during the first twelve months of life (Kappeler et al. 2009; Richard et al. 2002). It is probable that infant mortality steeply increases from falls or predation after they achieve independence, and no longer rely on mothers for milk or transport. This rise in infant mortality could be related the significantly higher cortisol response in adult males during later lactation when infants spend greater durations off carriers. I propose that infant age is a factor contributing to P. coquereli male stress responses, with more altricial infants that spend the majority of time physically attached to their mothers, inciting a lesser stress response than older infants that spend more time off carriers. Males disperse from their natal groups, migration occurs during the birth season, with 60% of P. verreauxi groups experience male migration events (Brockman et al. 2001). Takeover rate is high, with 80% of

150

five groups experiencing takeovers during the birth season (Brockman et al. 2009). All infants died or disappeared after takeover events (Brockman et al. 2009). Whether greater stress responses are a function of male migration events, infant presence, infant age, or most likely a combination of all these stressors remains partially unclear, it is evident infants are vulnerable and that uncontrollable seasonal events contribute to elevated stress responses.

There may be a nutritional component to help facilitate a reduced stress response in lactating females. I found that P. coquereli mothers consumed foods high in non-protein energy, high in available protein, and high in fiber (Ross and Power, in prep). Thus, lactating P. coquereli acquire sufficient nutrients that do not inhibit infant care, and permit a reduced stress response. A high fiber diet increases gut transit time, thereby allowing for less time for the reabsorption of GCs (reviewed in Keay et al. 2006). Decreased intestinal reabsorption of steroid hormones may increase the cortisol metabolites present in feces. However, the severity and variability of GC reabsorption on cortisol concentrations is unclear. For example, seasonality and diet type were the two most significant effects in GC variation in Alaskan brown bears (Urus arctos horribilis), with a high fiber diet of berries yielding the lowest fecal GCs (Ohe von der et al. 2004). A high fiber diet increased fecal bulk in yellow baboons (Papio cynocephalus) and did not influence fecal concentrations of progestogen (Wasser et al. 1993). Lactating P. coquereli consumed high fiber foods yet had the lowest fecal cortisol, therefore decreased reabsorption does not appear to effect cortisol concentrations.

In conclusion, the birth season causes higher stress responses in P. coquereli adult males and non-lactating adult females than in lactating females. Despite severe inter-annual unpredictability in rainfall that has been suggested to drive a suite of adaptive lemur traits (Dewar and Richard 2007), and the energetically constraining period of lactation, lactating females still exhibited a lower stress response. It is important to point out that this does not signify a non-stress or even a low-stress state in lactating females. Rather, lactating females may be coping with stressors using alternative physiological or behavioral mechanisms to increase infant survival given the high reproductive value of every infant. Adult males had significantly higher cortisol during later-mid/late lactation relative to early/earlier-mid lactation. Non-lactating adult females had significantly higher cortisol relative to lactating females considering weeks 1-24 postnatal. While the findings were not significant, a trend following the significant adult male pattern

151

detected lactating females had higher cortisol during later-mid/late lactation relative to early/earlier-mid lactation. There was also a trend supporting lactating females had lower cortisol than adult males from weeks 1-24 postnatal.

The ubiquitous idea that lemurs invest diminutively in infants does not apply to P. coquereli infants from birth to six months postnatal. I argue that a decreased stress response in lactating females is a way for mothers to devote more resources towards infant care and future pregnancies. The elevated stress response in adult males later in lactation demonstrates that a multitude of seasonal and reproductive pressures are exerted during the birth season. These include: decreased preferred food access in the already arduous dry season, preparation for mating, migration, and infant age. My findings will contribute to parental-investment and life history theory. Future studies using fecal isotope analysis to quantitatively measure the weaning process in tandem with cortisol will help determine exact interbirth intervals and stress responses throughout weaning. My study will serve as a catalyst to investigate P. coquereli stress responses during the mating season and will be a useful comparison to birth season cortisol.

4.6. References

Abbott D, Keverne E, Bercovitch F, Shively C, Mendoza S, Saltzman W, Snowdon C, Ziegler T, Banjevic M, and Garland Jr. T, Sapolsky, RM. 2003. Are subordinates always stressed? A comparative analysis of rank differences in cortisol levels among primates. Hormones and Behavior 43:67-82. Albignac R. 1981. Lemurine social and territorial organization in a north-western Malagasy forest (restricted area of Ampijoroa). In: Chiarelli A, and Corruccini R, editors. Primate Behavior and Sociobiology. New York: Springer-Verlag. p 25-29. Alonso L, and Hannah L. 2002. Introduction to the Reserve Naturelle Integrale and Reserve Forestiere d’Ankarafantsika and to the rapid assessment program. In: Schulenberg T, Radilofe S, and Missa O, editors. A biological assessment of the Reserve Naturelle Integrale d’Ankarafantsika, Madagascar. Washington DC: Conservation International p28-62. Andrainarivo C, Andriaholinirina, V.N., Feistner, A., Felix, T., Ganzhorn, J., Garbutt, N., Golden, C., Konstant, B., Louis Jr., E., Meyers, D., Mittermeier, R.A., Perieras, A., Princee, F., Rabarivola, J.C., Rakotosamimanana, B., Rasamimanana, H., Ratsimbazafy, J., Raveloarinoro, G., Razafimanantsoa, A., Rumpler, Y., Schwitzer, C., Thalmann, U., Wilmé, L. & Wright, P. 2008. IUCN Red List of Threatened Species: Propithecus coquereli. 17 November 2013 ed. Baden A. 2011. Communal infant care in black-and-white ruffed lemurs (Varecia variegata) [dissertation]. Available from ProQuest, 3490436: Stony Brook University 252 p. Bókony V, Lendvai A, Liker A, Angelier A, Wingfield J, and Chastel O. 2009. Stress response and the value of reproduction: are birds prudent parents? American Naturalist 173:589-598. Bonier F, Martin P, Moore I, and Wingfield J. 2009a. Do baseline glucocorticoids predict fitness? Trends in Ecology and Evolution 24:634-642. Bonier F, Moore I, Martin P, and Robertson R. 2009b. The relationship between fitness and baseline glucocorticoids in a passerine bird. General and Comparative Endocrinology 163:208-213. Bonier F, Moore I, and Robertson R. 2011. The stress of parenthood? Increased glucocorticoids in birds with experimentally enlarged broods. Biology Letters 7:944-946. Boonstra R, Hik D, Singleton GR, and Tinnikov A. 1998. The impact of predator-induced stress on the snowshoe hare cycle. Ecological Monographs 68:371-394. Boonstra R, McColl CJ, and Karels TJ. 2001. Reproduction at all costs: the adaptive stress response of male arctic ground squirrels. Ecology 82:1930-1946. Brockman D. 1999. Reproductive behavior of female Propithecus verreauxi at Beza Mahafaly, Madagascar. International Journal of Primatology 20:375-398. Brockman D, and Whitten P. 1996. Reproduction and free-ranging Propithecus verreauxi: estrus and the relationship between multiple partner matings and fertilization. American Journal of Physical Anthropology 100:57-69. Brockman D, Whitten P, Richard R, and Benender B. 2001. Birth season testosterone levels in male Verreaux's sifaka, Propithecus verreauxi: insights into the sociodemographic factors mediating seasonal testicular function. Behavioral Ecology and Sociobiology 49:117-127. Brockman DK, Cobden AK, and Whitten PL. 2009. Birth season glucocorticoids are related to the presence of infants in sifaka (Propithecus verreauxi). Proceedings of the Royal Biological Society B 276:1855-1863. Busch D, and Hayward L. 2009. Stress in a conservation context: A discussion of glucocorticoid actions and how levels change with conservation-relevant variables. Biological Conservation 142:2844-2853. Carlisle T. 1982. Brood success in variable environments: implications for parental care allocation. Animal Behavior 30:824-836. Cavigetti S. 1999. Behavioural patterns associated with faecal cortisol levels in free-ranging female ring-tailed lemurs, Lemur catta. Animal Behaviour 57:935-944. Chrousos G, Loriaux D, and Gold P. 1988. The concept of stress and its historical development. In: Chrousos G, Loriaux D, and Gold P, editors. Mechanisms of Physical and Emotional Stress New York: Plenum Press. p 3-7. Chrousos G, McCarty R, Pacak K, Cizza G, Sternberg E, Gold P, and Kvetnansky R. 1995. Stress: Basic mechanisms and clinical implications: New York Academy of Sciences 771 p. Clutton-Brock T. 1991. Parental care in birds and mammals. In: Krebs J, and Clutton-Brock T, editors. The Evolution of Parental Care. Princeton Princeton New Jersey Press. p 131-152. Creel S, Creel N, Mills M, and Monfort S. 1996. Rank and reproduction in cooperatively breeding African wild

152 153

dogs: Behavioral and endocrine correlates. Behavioral Ecology and Sociobiology 8:298-306. Creel S, Fox J, Hardy A, Sands J, Garrott R, and Peterson R. 2002. Snowmobile activity and glucocorticoid stress responses in wild wolves and elk. Conservation Biology 16:809-814. Creel S, Jr. W, JA, and Christianson D. 2009. Glucocorticoid stress hormones and the effect of predation risk on elk reproduction. Proceedings of the National Academy of Sciences of the United States of America 106:12388- 12393. Crowley B, McGoogan K, and Lehman S. 2012. Edge effects on foliar stable isotope values in a Madagascan tropical dry forest. Public Library of Science ONE 7:1-9. da Silva Mota M, Franci C, and Cordeiro de Sousa M. 2006. Hormonal changes related to parental and alloparental care in common marmosets (Callithrix jacchus). Hormones and Behavior 49:293-302. Dewar R, and Richard A. 2007. Evolution in the hypervariable environment of Madagascar. Proceedings of the National Academy of Sciences of the USA 104:13723-13727. Du Puy D, and Moat J. 1996. A refined classification of the primary vegetation of Madagascar based on the underlying geology: using GIS to map its distribution and assess its conservation status. In: Lourenço W, editor. Proceedings of the international symposium on the biogeography of Madagascar Paris: Editions de l'Orstrom p 205-218. Fichtel C, Kraus C, Ganswindt A, and Heistermann M. 2007. Influence of reproductive season and rank on fecal glucocorticoid levels in free-ranging male Verreaux's sifakas (Propithecus verreauxi. Hormones and Behavior 51:640-648. Gerardo G, and Goodman SM. 2003. Hunting of protected animals in the Parc National d'Ankarafantsika, northwestern Madagascar. Oryx 37:115-118. Gittleman J, and Thomspon S. 1988. Energy allocation in mammalian reproduction. American Zoologist 28:863- 875. Gould L, Sussman RW, and Sauther ML. 1999. Natural disasters and primate populations: the effects of a 2-year drought on a naturally occurring population of ring-tailed lemurs (Lemur catta) in southwestern Madagascar. International Journal of Primatology 20:69-84. Gould L, Ziegler T, and Wittwer D. 2005. Effects of reproductive and social variables on fecal glucocorticoid levels in a sample of adult male ring-tailed lemurs (Lemur catta) at the Beza Mahafaly Reserve, Madagascar. American Journal of Primatology 67:5-23. Heidinger B, Nisbet I, and Ketterson E. 2006. Older parents are less responsive to a stressor in a long-lived seabird: a mechanism for increased reproductive performance with age? Proceedings of the Royal Biological Society B 273:2227-2231. Hinde K, Skibiel A, Foster A, Del Rosso L, Mendoza S, and Capitanio J. 2015. Cortisol in mother's milk across lactation reflects maternal life history and predicts infant temperament. Behavioral Ecology 26:269-281. Huber S, Palme R, and Arnold W. 2003. Effects of season, sex, and sample collection on concentrations of fecal cortisol metabolites in red deer (Cervus elaphus). General and Comparative Endocrinology 130:48-54. Jolly A. 1966. Propithecus verreauxi verreauxi. Lemur Behavior: A Madgascar Field Study. Chicago: University of Chicago Press. p 20-68. Jolly A. 1984. The puzzle of female feeding priority. In: Small M, editor. Female Primates: Studies by Women Primatologists. New York: Alan Liss, INC. Kappeler P, Mass V, and Port M. 2009. Even adult sex ratios in lemurs: potential costs and benefits of subordinate males in verreaux’s sifaka (Propithecus verreauxi) in the Kirindy Forest CFPF, Madagascar. American Journal of Physical Anthropology 140:487-497. Keay J, Jatinder Singh B, Gaunt M, and Kaur T. 2006. Fecal glucocorticoids and their metabolites as indicators of stress in various mammalian species: A literature review. Journal of Zoo and Wildlife Medicine 37:234- 244. Kun-Rodrigues C, Salmona J, Besolo A, Rasolondraibe E, Rabarivola C, Marques TA, and Chikhi L. 2014. New density estimates of a threatened sifaka species (Propithecus coquereli) in Ankarafantsika National Park. American Journal of Primatology 76:515-528. Leigh S, and Terranova C. 1998. Comparative perspectives on bimaturism, ontogeny, and dimorphism in lemurid primates. International Journal of Primatology 19:723-749. Leutenegger W. 1973. Maternal-fetal weight relationships in primates. Folia Primatologica 20:280-293. Lourenço W, and Goodman S. 2006. Notes on the postembryonic development and ecology of Grosphus hirtus Kraepelin, 1901 (Scorpiones, Buthidae) from the Parc National d'Ankarafantsika, northwest Madagascar. Zoologischer Anzeiger 244:181-185.

154

Mittermeier R, Louis Jr., E, Richardson, M, Schwitzer, C, Langrand, O, Rylands, A, Hawkins, F, Rajaobelina, S, Ratsimbazafy, J, Rasoloarison, R, Roos, C, Kappeler, P, Mackinnon, J. 2010. Indriidae. Lemurs of Madagascar. Third ed. Bogotá, Colombia: Conservation International. p 481-597. Moberg G. 1999. When does stress become distress? Laboratory Animals 28:22-26. Moberg G. 2000. Biological response to stress: implications for animal welfare. In: Moberg G, and Mench J, editors. The Biology of Animal Stress: Basic Principles and Implications for Animal Welfare Cambridge, MA: CABI Publishing p1-21. Monfort S, Mashburn K, Brewer B, and Creel S. 1998. Evaluating adrenal activity in African wild dogs (Lycaon pictus) by fecal corticosteroid analysis. Journal of Zoo and Wildlife Medicine 29:129-133. Morland H. 1990. Parental behavior and infant development in ruffed lemurs (Varecia variegata) in a northeast Madagascar rain forest. American Journal of Primatology 20:253-265. Nelson R. 2000. Stress. An Introduction to Behavioural Endocrinology. Sunderland, MA: Sinauer. p 557-591. Nur N. 1984. Feeding frequencies of nestling blue tits (Parus caerulecus): costs, benefits, and a model of optimal feeding frequency. Oecologia 65:125-137. O'Reilly K, and Wingfield J. 2001. Ecological factors underlying the adrenocortical response to capture stress in artic-breeding shorebirds. General and Comparative Endocrinology 124:1-11. Ohe von der C, Wasser S, Hunt K, and Servheen C. 2004. Factors assocaited with fecal glucocorticoids in Alaskan brown bears. Physiological and Biochemical Zoology 77:313-320. Oster J, Kappeler P, and Heistermann M. 2008. Androgen and glucocorticoid levels reflect seasonally occurring social challenges in male redfronted lemurs (Eulemur fulvus rufus). Behavioural Ecology and Sociobiology 62:624-638. Ostner J, Kappeler P, and Heistermann M. 2008. Androgen and glucocorticoid levels reflect seasonally occurring social challenges in male redfronted lemus (Eulemur fulvus rufus). Behavioral Ecology and Sociobiology 62:627-638. Petty J, and Drea C. 2015. Female rule in lemurs is ancestral and hormonally mediated. Scientific Reports 5:DOI:10.1038/srep09631. Pochron S, Tucker W, and Wright P. 2004. Demography, life history, and social structure in Propithecus diadema edwardsi from 1986–2000 in Ranomafana National Park, Madagascar. American Journal of Physical Anthropology 125:61-72. Pöysä H, Virtanen J, and Milonoff M. 1997. Common goldeneyes adjust maternal effort in relation to prior brood success and not current brood size. Behavioral Ecology and Sociobiology 40:101-106. Purvis A, Gittleman J, Cowlishaw G, and Mace G. 2000. Predicting extinction risk in declining species. Proceedings of the Royal Society B 267:1947-1952. Rendigs A, Radespiel U, Wrogemann D, and Zimmermann E. 2003. Relationship between microhabitat structure and distribution of mouse lemurs (Microcebus spp.) in northwestern Madagascar. International Journal of Primatology 24:47-64. Richard A. 1974. Intra-specific variation in the social organization and ecology of Propithecus verreauxi. Folia Primatologica 22:178-207. Richard A. 1976. Preliminary observations on the birth and development of Propithecus verreauxi to the age of six months. Primates 17:357-366. Richard A. 1985. Social boundaries in a Malagasy prosimian, the sifaka (Propithecus verreauxi). International Journal of Primatology 6:553-568. Richard A, Dewar R, Schwartz M, and Ratsirarson J. 2002. Life in the slow lane? Demography and life histories of male and female sifaka (Propithecus verreauxi verreauxi). The Zoological Society of London 256:421-436. Richard A, Rakotomanga P, and Schwartz M. 1991. Demography of Propithecus verreauxi at Beza Mahafaly, Madagascar: sex ratio, survival, and fertility, 1984-1988. American Journal of Physical Anthropology 84:307-322. Ross A, and Lehman S. 2016. Infant transport and mother–infant contact from 1 to 26 weeks postnatal in Coquerel’s sifaka (Propithecus coquereli) in northwestern Madagascar. American Journal of Primatology 78: 646-658. Ross A, and Power M. in prep. Food for thought: Lactating Coquerel’s sifaka (Propithecus coquereli) select high non-protein energy, high available protein, and high fiber foods. Saltzman W, and Abbott D. 2009. Effects of elevated circulating cortisol concentrations on maternal behavior in common marmoset monkeys (Callithrix jacchus). Psychoneuroendocrinology 34:1222-1234. Sapolsky RM. 1992. Neuroendocrinology of the stress-response In: Becker J, Breedlove S, and Crews D, editors. Behavioural Endocrinology. Cambridge, MA: MIT. p 287-324. Sapolsky RM. 2000. Stress hormones: good and bad. Neurobiology of Disease 7:540-542. Selye H. 1937. The significance of the adrenals for adaption. Science 85:247-248.

155

Setchell J, Smith T, Wickings E, and Knapp L. 2008. Factors affecting fecal glucocorticoid levels in semi-free- ranging female mandrills (Mandrillus sphinx). American Journal of Primatology 70:1023-1032. Sheriff M, Dantzer B, Delehanty B, Palme R, and Boonstra R. 2011. Measuring stress is wildlife: techniques for quantifying glucocorticoid. Oecologia 166:869-887. Stearns SC. 1992. Evolutionary Explanations The Evolution of Life Histories New York: Oxford University Press. Strier K, Ziegler T, and Wittwer D. 1999. Seasonal and social correlates of fecal testosterone and cortisol levels in wild male muriquis (Brachyteles arachnoides). Hormones and Behavior 35:125-134. Tecot S. 2010. It's all in the timing: birth seasonality and infant survival in Eulemur rubriventer. International Journal of Primatology 31:715-735. Tecot S. 2013. Variable energetic strategies in disturbed and undisturbed rain forests: Eulemur rubriventer fecal cortisol levels in south-eastern Madagascar. In: Masters J, Gamba M, and Génin F, editors. Leaping Ahead: Advances in Prosimian Biology. New York: Springer-Verlag. p 185-195. Tennenhouse E. 2016. Unique lemur traits: proximate and ultimate perspectives [dissertation]. University of Toronto 169 p. Touma C, and Palme R. 2005. Measuring fecal glucocorticoid metabolites in mammals and birds: the importance of validation. Annals New York Academy of Sciences 1046:54-74. Trivers RL. 1972. Parental investment and sexual selection. Chicago: Aldine. Wasser S, Thomas R, Nair P, Guidry C, Southers J, Lucas J, Wildt D, and Monfort S. 1993. Effects of dietary fibre on faecal steroid measurements in baboons (Papio cynocephalus cynocephalus). Reproduction 97:569-574. Wright P. 1999. Lemur traits and Madagascar ecology: coping with an island environment. Yearbook of Physical Anthropology 42:31-72. Young A, Richard A, and Aiello L. 1990. Female dominance and maternal investment in strepsirhine primates. American Naturalist 135:473-488. Ziegler T. 2000. Hormones associated with non-maternal infant care: a review of mammalian and avian studies. Folia Primatologica 71:6-21. Ziegler TE, Scheffler G, and Snowdon CT. 1995. The relationship of cortisol levels to social environment and reproductive functioning in female cotton-top tamarins, Saguinus oedipus. Hormones and Behavior 29:407- 424.

156

Table 4.1. P. coquereli fecal samples collected by individual sex/reproductive class and data collection season Class Season 1 (N) Season 2 (N) Seasons combined (N) Lactating females 116 64 180 Adult males 77 56 133 Non-lactating 28 34 62 adult females Total 221 154 375

157

Table 4.2. Comparison of cortisol in P. coquereli by individual sex/reproductive class and lactation phase from 1-24 weeks postnatal (weeks 4-23 in non-lactating adult females) Class Lactation N Mean SD Minimum Maximum phasea, b cortisol (ng/g) Lactating Weeks 1-12 81 185.65 45.59 102.31 364.89 females (LF) LF Weeks 13-24 99 189.60 57.10 65.90 484.28 Adult males Weeks 1-12 64 189.78 59.66 92.81 393.20 (AM) AM Weeks 13-24 69 209.90 57.64 69.23 349.72 Non-lactating Weeks 4-12 27 207.12 75.03 91.30 498.16 adult females (AF) AF Weeks 13-23 35 212.43 64.43 112.19 427.44 aEarly lactation designated as weeks 1-12 bLate lactation designated as weeks 13-24

158

Table 4.3. Comparison of cortisol in P. coquereli by individual sex/reproductive class from 1-24 weeks postnatal (weeks 4-23 in non-lactating adult females) Class N Mean cortisol SD Minimum Maximum (ng/g) Lactating 180 187.83 52.13 65.90 484.28 females Adult males 133 200.22 59.26 69.23 393.20 Non-lactating 62 210.12 68.70 91.30 498.16 adult females

159

Table 4.4. Mixed effects linear regression of cortisol in P. coquereli by individual sex/reproductive class and lactation phase from 1-24 weeks postnatal (weeks 4-23 in non- lactating adult females) Effect Group Sex/reproductive Estimate SE DF t value Pr>|t| class Intercept - - 236.36 10.95 16 21.58 <.0001 Group Citron - -13.19 11.48 16 -1.15 0.2677 Group Fito - -52.17 9.90 16 -5.27 <.0001 Group Kambana - -31.45 10.18 16 -3.09 0.007 Group Mainty - -33.93 9.81 16 -3.46 0.0032 Group Vaovao - -23.36 10.74 16 -2.18 0.0449 Group Volo - -54.55 9.72 16 -5.61 <.0001 Group Zaza - -30.36 10.05 16 -3.02 0.0081 Group Iva - -14.23 8.80 16 -1.62 0.1253 Group Rambo - -0.38 8.16 16 -0.05 0.9638 Group Zaza (s2) - 0.00 - - - - Sex/reproductive Lactating females -27.22 10.32 20 -2.64 0.0158 class Early lactationa Lactation phase Sex/reproductive Lactating females -22.24 10.11 20 -2.20 0.0398 class Late lactationb Lactation phase Sex/reproductive Adult males -25.53 10.84 20 -2.35 0.0289 class Early lactation Lactation phase Sex/reproductive Adult males -4.71 10.77 20 -0.44 0.6668 class Late lactation Lactation phase Sex/reproductive Non-lactating adult -6.38 14.18 20 -0.45 0.6578 class females Lactation phase Early lactation Sex/reproductive Non-lactating adult 0.00 - - - - class females Lactation phase Late lactation

Numerator Denominator Effect DF DF F value Pr>F Group 9 16 8.00 0.0002 Sex/reproductive 5 20 3.31 0.0245 class Lactation phase aEarly lactation designated as weeks 1-12 bLate lactation designated as weeks 13-24

160

Table 4.5. Mixed effects linear regression estimates of cortisol in P. coquereli by individual sex/reproductive class and lactation phase from 1-24 weeks postnatal (weeks 4-23 in non-lactating adult females) Class/lactation phase Estimate SE DF t value Pr>|t| Lactating females (LF) -4.97 8.37 20 -0.59 0.5590 Earlya vs. lateb lactation Adult males (AM) -20.83 9.65 20 -2.16 0.0432 Early vs. late lactation Non-lactating adult females -6.38 14.18 20 -0.45 0.6578 (AF) Early vs. late lactation LF vs. AM -9.61 5.02 20 -1.91 0.0701 LF vs. AF -21.54 7.04 20 -3.06 0.0062 aEarly lactation designated as weeks 1-12 bLate lactation designated as weeks 13-24

161

Table 4.6. Comparison of P. coquereli fecal cortisol and P. verreauxi fecal corticosterone Variable Species N Mean cortisol (ng/g) SD Lactating females P. coquereli 180 187.83 52.13 Adult males P. coquereli 133 200.22 59.26 Non-lactating adult females P. coquereli 62 210.12 68.7 Mean corticosterone (ng/g) Adult malesa P. verreauxi 42 59.63 32.95 Caregiving malesa P. verreauxi 30 69.85 40.16 Non-caregiving malesa P. verreauxi 43 70.03 37.65 Infants presenta P. verreauxi 73 71.39 37.56 Infants absenta P. verreauxi 56 53.92 36.04 aData from Brockman et al. (2009)

162

Figure 4.1. The hypothalamic-pituitary-adrenal (HPA) axis, negative feedback response to chronic and acute stress, and effects of stressors on bodily processes

Stressor

Hypothalamus

CRH

Pituitary Feedback Feedback under chronic under acute stress ACTH stress

Adrenals

Cortisol

Blood

Energy Immunity and mobilization Growth Reproductive Digestive inflammatory at the cost of suppression suppression suppression response energy storage suppression

Adapted from Boonstra et al. (1998)

163

Figure 4.2. The biological response of animals to stress

Stimulus

Recognition of a threat to homeostasis Central nervous system

Perception of stressor

Organization of biological defense

Biological response (behavioural, autonomic, neuroendocrine, immunological)

Stress response

Normal biological function

Altered biological function

Prepathological state Consequences of stress

Development of pathology

Adapted from Moberg (1999)

164

Figure 4.3. Ankarafantsika National Park, northwestern Madagascar

165

Figure 4.4. Cortisol concentrations in P. coquereli lactating females (n=10) 500 475 450 425 400 375 350 325

300

) g

/ 275

n (

250

o s

i 225

t r

o 200 C 175 150 125 100 75 50 25 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Weeks postnatal

166

Figure 4.5. Cortisol concentrations in P. coquereli adult males (n=19) 295

270

245

220

195

)

g /

g 170

(

l o

s 145

r o

C 120

-5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Weeks postnatal

167

Figure 4.6. Cortisol concentrations in P. coquereli adult non-lactating females (n=8) 345

320

295

270

245

220

)

g /

g 195

(

o s

i 170

r o

C 145

120

-5 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Weeks postnatal

168

Figure 4.7. Average cortisol in P. coquereli by sex/reproductive class 325

300

275

250

) g

/ 225

(

o s

i 200

o C 175

150

125

100 Lactating females Adult males Non-lactating adult females 1-24 weeks postnatal

169

Figure 4.8. Cortisol comparison across sex/reproductive classesa and lactation phasesb,c 320

300

280

260

240

220

200

180

160

140

120

Cortisol Cortisol (ng/g) 100

0 LF (1-12) LF (13-24) AM (1-12) AM (13-24) AF (4-12) AF (13-23)

Weeks postnatal aLF= lactating females, AM= adult males, AF= non-lactating adult females bEarly lactation designated as weeks 1-12 cLate lactation designated as weeks 13-24

170

4.7. Appendix

P. coquereli fecal field collection and cortisol concentrations Assay number Collection time Cortisola (ng/gram) Individual Infant age in weeks CIT001 6:28 130.08 MO 7 CIT002 10:43 155.01 AM 8 CIT003 12:19 96.93 AM 9 CIT004 10:34 173.55 MO 10 CIT005 5:58 187.03 MO 11 CIT006 6:35 166.24 AM 13 CIT007 8:47 150.81 MO 13 CIT008 7:11 291.15 AM 14 CIT009 7:18 252.36 AM 14 CIT010 7:21 256.23 MO 14 CIT011 7:01 349.72 AM 15 CIT012 7:15 484.28 MO 15 CIT013 7:15 185.64 MO 16 CIT014 8:28 155.44 MO 17 CIT015 6:49 160.45 MO 18 CIT016 5:30 133.64 AM 19 CIT017 7:36 183.84 MO 19 CIT018 7:21 155.31 AM 20 CIT019 8:42 258.31 MO 20 CIT020 6:10 214.25 MO 21 CIT021 6:15 233.03 AM 21 CIT022 8:43 171.41 AM 22 CIT023 7:08 193.89 MO 23 CIT024 5:08 247.32 AM 24 CIT025 6:37 207.57 MO 24 FIT001 9:36 178.04 MO 3 FIT002 7:45 154.49 AF 4 FIT003 8:01 138.03 AM 4 FIT004 7:36 216.59 MO 5 FIT005 8:05 189.77 AF 5 FIT006 7:25 164.01 MO 6 FIT007 8:33 213.47 AM 6 FIT008 7:22 148.51 AF 6 FIT009 7:00 150.76 MO 7 FIT010 7:16 168.91 AF 7 FIT011 7:36 142.33 AF 7 FIT012 8:38 168.85 MO 7

171

Assay number Collection time Cortisola (ng/gram) Individual Infant age in weeks FIT013 7:20 185.17 AM 8 FIT014 7:36 183.33 MO 8 FIT015 7:29 190.97 MO 9 FIT016 8:50 197.23 AM 9 FIT017 7:05 173.65 MO 10 FIT018 7:19 130.93 AF 10 FIT019 6:21 91.29 AF 11 FIT020 6:37 150.40 MO 11 FIT021 6:52 141.17 MO 12 FIT022 7:00 145.75 MO 13 FIT023 7:06 132.59 AF 13 FIT024 6:49 158.19 MO 14 FIT025 6:02 143.31 MO 15 FIT026 6:35 151.20 AM 15 FIT027 7:30 167.91 AF 15 FIT028 6:49 188.79 AF 16 FIT029 7:54 144.95 MO 16 FIT030 6:55 162.40 MO 17 FIT031 7:02 276.07 AF 18 FIT032 7:10 202.92 AM 18 FIT033 7:34 132.41 MO 18 FIT034 6:48 116.59 MO 19 FIT035 6:37 176.51 MO 20 FIT036 6:40 230.41 AM 20 FIT037 6:16 182.12 AM 20 FIT038 6:43 94.59 MO 20 FIT039 5:48 185.40 MO 21 FIT040 6:03 228.39 AF 21 IVA001 8:11 121.48 AM 3 IVA002 8:14 147.57 MO 3 IVA003 8:08 218.65 MO 4 IVA004 8:17 240.24 AF 4 IVA005 9:47 192.87 AM 4 IVA006 8:33 157.97 MO 5 IVA007 8:41 373.00 AM 5 IVA008 8:52 206.84 AF 5 IVA009 9:09 102.31 MO 6 IVA010 9:10 208.19 AF 6 IVA011 9:17 177.25 AM 6 IVA012 8:14 221.25 MO 7 IVA013 8:15 190.76 AF 7

172

Assay number Collection time Cortisola (ng/gram) Individual Infant age in weeks IVA014 7:32 159.28 AM 8 IVA015 7:57 175.89 MO 8 IVA016 8:06 196.85 MO 8 IVA017 8:13 214.20 AF 8 IVA018 8:59 185.59 MO 9 IVA019 6:44 177.31 MO 9 IVA020 6:57 202.28 AM 9 IVA021 7:11 281.16 AF 9 IVA022 7:07 215.67 MO 10 IVA023 7:09 300.56 AM 10 IVA024 7:33 267.45 AF 10 IVA025 7:02 237.84 MO 11 IVA026 7:33 224.68 AM 11 IVA027 6:37 173.55 MO 12 IVA028 6:47 160.47 AM 12 IVA029 7:22 219.71 AF 12 IVA030 6:26 195.17 AF 13 IVA031 6:27 224.31 MO 13 IVA032 6:31 177.01 AM 13 IVA033 8:08 143.95 MO 14 IVA034 8:17 268.21 AF 14 IVA035 8:25 252.95 AM 14 IVA036 6:16 202.48 MO 15 IVA037 6:40 209.56 AM 15 IVA038 7:16 278.53 AF 15 IVA039 6:39 222.44 AF 16 IVA040 8:40 205.63 MO 16 IVA041 6:20 181.11 MO 16 IVA042 7:02 275.93 AF 16 IVA043 8:53 254.31 AM 16 IVA044 5:50 323.28 MO 17 IVA045 8:01 251.85 AM 17 IVA046 6:52 195.55 MO 18 IVA047 7:13 192.12 AF 18 IVA048 5:39 222.05 MO 19 IVA049 5:58 171.79 AF 19 IVA050 8:06 169.25 AM 19 IVA051 6:20 125.55 AM 20 IVA052 6:39 134.73 MO 20 IVA053 8:11 123.87 AF 20 IVA054 6:15 188.55 MO 21

173

Assay number Collection time Cortisola (ng/gram) Individual Infant age in weeks KAM003 10:11 195.85 MO 7 KAM004 7:50 184.49 MO 9 KAM005 8:44 219.25 MO 9 KAM006 7:29 211.49 AM 9 KAM007 7:17 139.01 MO 10 KAM008 8:10 115.59 AM 10 KAM010 6:55 199.17 AM 11 KAM011 6:56 148.04 MO 11 KAM012 7:24 170.80 MO 12 KAM013 8:36 192.03 AM 12 KAM014 7:40 180.59 MO 13 KAM015 7:55 181.13 AM 13 KAM016 8:39 111.69 MO 14 KAM017 6:22 128.80 MO 14 KAM018 6:29 170.64 MO 15 KAM019 6:38 175.51 AM 15 KAM020 6:51 132.09 AF 15 KAM021 8:22 173.92 MO 16 KAM022 8:43 167.72 AF 16 KAM023 9:06 231.12 AM 16 KAM025 6:26 244.75 AF 17 KAM027 5:43 105.89 MO 18 KAM028 6:57 189.28 AF 18 KAM029 6:56 157.71 MO 19 KAM030 6:40 152.32 MO 20 KAM031 7:14 235.53 AF 20 KAM032 8:08 238.64 AM 20 KAM033 6:33 135.69 MO 21 KAM034 6:19 200.96 MO 22 KAM035 6:36 186.29 AF 23 KAM036 7:08 223.15 MO 23 KAM037 7:31 242.72 AM 23 KAM038 6:12 321.41 AM 24 KAM039 6:26 315.99 MO 24 MAI001 7:43 142.97 MO 5 MAI003 7:48 225.49 AM 6 MAI004 7:49 155.32 MO 6 MAI005 9:18 141.79 MO 7 MAI006 9:32 138.40 AM 7 MAI007 8:30 234.39 MO 8 MAI008 7:55 133.03 MO 8

174

Assay number Collection time Cortisola (ng/gram) Individual Infant age in weeks MAI009 8:37 178.83 AM 8 MAI010 7:47 173.32 MO 9 MAI011 7:58 192.91 AM 9 MAI012 7:25 253.60 MO 10 MAI013 7:29 265.56 AM 10 MAI014 7:24 279.15 MO 11 MAI015 7:30 192.65 AM 11 MAI016 7:15 138.41 AM 12 MAI017 6:40 131.88 MO 12 MAI018 7:16 170.52 MO 13 MAI019 7:54 180.03 AM 13 MAI020 6:05 266.40 AM 14 MAI021 6:08 186.24 MO 14 MAI022 6:43 168.93 MO 15 MAI023 7:11 309.71 AM 15 MAI025 8:38 183.05 MO 16 MAI026 6:11 79.95 MO 17 MAI027 6:36 135.96 AM 17 MAI028 5:38 171.45 MO 18 MAI029 5:53 130.09 AM 18 MAI030 6:21 165.07 AM 19 MAI031 6:26 174.41 MO 19 MAI032 6:28 208.33 MO 20 MAI033 6:00 175.75 MO 21 MAI034 7:05 231.33 AM 21 MAI035 7:54 241.00 MO 21 MAI036 9:13 79.48 MO 23 RAM001 7:46 144.07 MO 1 RAM002 7:55 138.57 AM 1 RAM003 6:43 196.12 MO 2 RAM004 7:01 145.39 AM 2 RAM005 7:15 141.67 MO 3 RAM006 7:21 92.81 AM 3 RAM007 7:10 214.69 MO 4 RAM008 8:48 167.35 AF 4 RAM009 6:50 498.16 AF 5 RAM010 8:01 284.63 AM 5 RAM011 8:35 224.41 MO 5 RAM012 6:40 181.57 MO 6 RAM013 6:51 161.75 AF 6 RAM014 7:46 132.45 AM 6

175

Assay number Collection time Cortisola (ng/gram) Individual Infant age in weeks RAM015 6:58 238.85 MO 7 RAM016 7:17 217.73 AM 7 RAM017 7:45 232.40 MO 8 RAM018 7:47 157.91 AM 8 RAM019 8:56 183.28 AF 8 RAM020 7:13 221.40 MO 9 RAM021 7:25 265.72 AF 9 RAM022 7:28 185.64 AM 9 RAM023 7:08 259.88 MO 10 RAM024 7:19 224.83 AF 10 RAM025 8:02 174.77 AF 11 RAM026 6:40 170.52 MO 12 RAM027 6:54 202.17 AM 12 RAM028 7:58 224.13 AF 12 RAM029 8:14 205.88 MO 13 RAM030 8:32 162.57 AF 13 RAM031 6:34 201.75 AF 14 RAM032 6:47 332.92 AM 14 RAM033 7:06 174.52 MO 14 RAM034 6:23 174.93 MO 15 RAM035 7:13 203.12 AM 15 RAM036 6:30 164.01 MO 16 RAM037 6:49 184.93 AF 16 RAM038 7:02 219.23 AM 16 RAM039 7:19 194.85 MO 17 RAM040 8:06 232.57 AF 17 RAM041 8:15 210.16 AM 17 RAM042 6:42 236.32 AM 18 RAM043 6:51 246.08 MO 18 RAM044 8:16 244.79 AF 18 RAM045 6:02 216.49 MO 19 RAM046 7:19 211.87 AF 19 RAM047 7:30 292.87 AM 19 RAM048 6:36 224.55 MO 20 RAM049 6:58 193.20 AM 20 RAM050 8:28 250.51 AF 20 RAM051 6:51 286.25 MO 21 RAM052 7:03 250.33 AM 21 RAM053 8:45 280.81 AF 21 RAM054 7:03 248.79 MO 22 RAM055 8:07 348.29 AF 22

176

Assay number Collection time Cortisola (ng/gram) Individual Infant age in weeks RAM056 5:59 249.41 MO 23 RAM057 6:12 427.44 AF 23 RAM058 7:54 213.37 AM 23 RAM059 6:40 173.07 AF 24 RAM060 6:48 273.84 MO 24 RAM061 6:54 283.59 AM 24 VAO001 9:08 210.12 MO 3 VAO002 7:25 181.43 MO 4 VAO003 7:26 128.68 AM 4 VAO004 7:39 139.33 AM 5 VAO005 7:57 178.88 AF 5 VAO006 8:26 162.47 MO 5 VAO007 7:03 171.81 MO 6 VAO008 7:08 116.25 AM 6 VAO009 7:19 129.19 AF 6 VAO010 8:42 188.84 AM 7 VAO011 8:01 185.28 MO 8 VAO012 5:50 198.44 MO 9 VAO013 7:11 162.52 AM 9 VAO014 7:54 165.17 MO 10 VAO015 5:54 265.79 AM 11 VAO016 6:03 258.07 AF 11 VAO017 6:17 189.55 MO 11 VAO018 6:17 360.29 MO 12 VAO019 6:29 256.84 AF 12 VAO020 6:30 264.65 AM 12 VAO021 5:42 274.23 AF 13 VAO022 5:49 225.87 MO 13 VAO023 5:58 153.07 AM 13 VAO024 5:36 223.23 MO 14 VAO025 5:45 178.16 AF 14 VAO026 5:47 275.80 MO 15 VAO027 7:00 124.71 MO 16 VAO028 5:48 154.00 AF 17 VAO029 5:50 136.77 AM 17 VAO030 5:57 128.93 MO 17 VOL001 7:27 132.69 MO 5 VOL002 8:13 130.88 MO 6 VOL003 9:20 176.60 AM 6 VOL004 7:21 152.39 AM 7 VOL005 7:31 145.24 MO 7

177

Assay number Collection time Cortisola (ng/gram) Individual Infant age in weeks VOL006 8:29 170.15 AF 7 VOL007 8:08 193.08 MO 8 VOL008 9:12 164.03 MO 8 VOL009 7:42 244.91 MO 9 VOL011 11:11 137.23 MO 9 VOL012 7:45 159.63 MO 10 VOL013 8:05 200.17 AM 10 VOL014 7:15 172.24 MO 10 VOL015 6:45 162.07 MO 11 VOL016 6:52 139.51 AM 11 VOL017 6:59 98.37 MO 12 VOL018 7:13 122.89 AM 12 VOL019 6:52 261.04 MO 12 VOL020 7:01 210.60 MO 13 VOL021 7:17 204.49 MO 13 VOL022 7:25 153.41 AM 13 VOL023 6:22 192.08 MO 14 VOL024 6:45 152.81 AM 14 VOL025 6:00 135.80 MO 15 VOL026 6:05 69.23 AM 15 VOL027 6:06 188.72 MO 16 VOL028 6:16 151.53 AM 16 VOL029 6:24 200.36 AF 16 VOL030 6:21 227.29 AM 17 VOL031 6:53 65.89 MO 17 VOL032 7:33 178.97 MO 18 VOL033 5:35 93.17 MO 19 VOL034 5:43 130.87 AM 19 VOL035 6:12 173.04 MO 19 VOL036 5:50 169.11 AF 20 VOL037 5:57 212.35 AM 20 VOL038 7:35 187.35 MO 20 VOL039 7:01 248.47 MO 21 VOL040 7:56 188.28 AM 21 VOL041 6:15 154.16 MO 22 VOL042 7:07 112.19 AF 22 ZAZ001 6:58 185.83 AM 6 ZAZ001 (s2) 8:52 153.77 AM 1 ZAZ002 9:36 202.76 MO 1 ZAZ003 8:17 180.52 AM 7 ZAZ003 8:26 364.89 MO 2

178

Assay number Collection time Cortisola (ng/gram) Individual Infant age in weeks ZAZ004 (s2) 8:36 199.36 AM 2 ZAZ005 10:20 142.59 MO 8 ZAZ005 (s2) 7:35 193.33 MO 3 ZAZ006 7:39 166.60 AM 9 ZAZ006 (s2) 7:56 200.71 AM 3 ZAZ007 7:43 160.21 MO 9 ZAZ007 (s2) 7:15 157.31 AM 4 ZAZ008 6:47 223.44 AM 10 ZAZ008 (s2) 11:08 183.19 MO 4 ZAZ009 7:58 221.25 MO 10 ZAZ009 (s2) 8:53 156.13 MO 5 ZAZ010 7:23 161.15 MO 11 ZAZ010 (s2) 7:41 223.08 AM 6 ZAZ011 8:43 255.76 AM 11 ZAZ011 (s2) 7:51 163.53 MO 6 ZAZ012 7:43 157.57 MO 12 ZAZ012 (s2) 8:31 190.43 AM 7 ZAZ013 7:55 238.32 AM 12 ZAZ013 (s2) 8:54 180.89 MO 7 ZAZ014 8:03 237.52 AM 12 ZAZ014 (s2) 8:24 168.89 AM 7 ZAZ015 7:33 139.29 AM 13 ZAZ015 (s2) 9:16 200.45 MO 7 ZAZ016 7:43 144.04 MO 13 ZAZ016 (s2) 7:57 175.44 MO 8 ZAZ017 6:23 221.59 AM 14 ZAZ017 (s2) 8:22 140.43 AM 8 ZAZ018 7:39 166.03 MO 14 ZAZ018 (s2) 8:41 123.63 AM 9 ZAZ019 7:21 188.95 MO 15 ZAZ019 (s2) 9:31 125.89 MO 9 ZAZ020 7:48 189.15 AM 15 ZAZ020 (s2) 7:50 265.71 AM 10 ZAZ021 7:03 178.77 AM 16 ZAZ021 (s2) 7:57 246.40 MO 10 ZAZ022 8:47 184.07 MO 16 ZAZ022 (s2) 7:16 330.81 MO 10 ZAZ023 6:51 196.01 MO 17 ZAZ023 (s2) 7:19 341.99 AM 10 ZAZ024 7:02 230.69 AM 17 ZAZ024 (s2) 6:20 393.20 AM 11

179

Assay number Collection time Cortisola (ng/gram) Individual Infant age in weeks ZAZ025 6:40 211.88 MO 18 ZAZ025 (s2) 7:08 250.95 MO 11 ZAZ026 6:52 200.63 MO 19 ZAZ026 (s2) 7:01 170.45 MO 12 ZAZ027 5:37 265.36 AM 20 ZAZ027 (s2) 8:47 257.57 AM 12 ZAZ028 5:53 133.47 MO 20 ZAZ028 (s2) 6:36 213.53 AM 13 ZAZ029 5:38 203.61 MO 21 ZAZ029 (s2) 9:09 156.88 MO 13 ZAZ030 6:09 151.36 MO 22 ZAZ030 (s2) 6:36 222.41 MO 13 ZAZ031 6:50 147.55 AM 22 ZAZ031 (s2) 7:07 206.04 AM 13 ZAZ032 6:51 164.93 MO 23 ZAZ032 (s2) 6:51 212.43 AM 14 ZAZ033 6:54 196.31 AM 24 ZAZ033 (s2) 6:59 206.60 MO 14 ZAZ034 7:37 198.47 MO 24 ZAZ034 (s2) 6:22 199.93 MO 14 ZAZ035 (s2) 6:26 210.35 AM 14 ZAZ036 (s2) 6:36 206.64 MO 15 ZAZ037 (s2) 6:40 236.01 AM 15 ZAZ038 (s2) 6:29 248.39 MO 16 ZAZ039 (s2) 7:29 322.61 AM 16 ZAZ040 (s2) 6:03 285.36 MO 17 ZAZ041 (s2) 6:40 298.77 AM 17 ZAZ042 (s2) 5:38 230.43 MO 18 ZAZ043 (s2) 5:40 169.95 AM 18 ZAZ044 (s2) 5:35 96.71 MO 19 ZAZ045 (s2) 6:10 171.63 AM 19 ZAZ046 (s2) 6:08 210.08 MO 20 ZAZ047 (s2) 7:03 277.44 AM 20 ZAZ048 (s2) 8:26 266.85 MO 21 ZAZ049 (s2) 9:34 185.04 AM 21 ZAZ050 (s2) 9:13 282.44 AM 22 ZAZ051 (s2) 9:29 188.41 MO 22 aWith recovery

Chapter 5 Conclusions and Future Directions

“Without the forest, there will be no more water; without water, there will be no more rice.”

“Raha tsy misy ala, tsy misy rano; raha tsy misy rano, tsy misy vary.” ~ Malagasy proverb

5.1. Theoretical significance

My dissertation examined Coquerel’s sifaka (Propithecus coquereli) maternal behavioral care-giving effort, the nutritional quality of foods selected by lactating females, and temporal fecal cortisol variation during lactation in lactating females, adult males, and non-lactating adult females in Ankarafantsika National Park located in northwestern Madagascar. My introduction was rooted within the theoretical foundation of the inherited environmental effects model (Eisen and Saxton 1983; Rossiter 1996). The inherited environmental effects model considered parental performance phenotype, which I measured as maternal behavioral care-giving and temporal cortisol variation; and parental environment, which I measured as food nutrient values. I considered each data chapter independently within three separately constructed theoretical frameworks that were all linked to the inherited environmental model (Eisen and Saxton 1983; Rossiter 1996). Chapter 2 was formulated based on the inherited environmental effects and maternal effects models (Bernardo 1996; Rossiter 1996). Chapter 3 was constructed from the theoretical groundwork of optimal foraging theory (Stephens and Krebs 1986), nitrogen maximization (Mattson 1980), fiber limitations (Milton 1979), and energy maximization models (Schoener 1971). Chapter 4 was devised from the cortisol- adaptation (Bonier et al. 2009b) and brood-value hypotheses (Heidinger et al. 2006).

Life history theory did not traditionally incorporate stochastic environments into contemporary mammalian models, and in turn has left a theoretical and empirical void in

180 181 the study of maternal-infant relationships, nutrient selectivity, and physiological responses in unstable environments such as Madagascar. My investigation of maternal behavioral effort was the first study to longitudinally quantify maternal-infant relationships in Propithecus spp. from birth until twenty-six weeks postnatal. I demonstrated that parental performance phenotype (maternal care and cortisol variation) and environment (food nutrients) play indispensable roles in the healthy development of infants. My infant transport and contact findings helped establish why maternal care is still important in a lesser studied mammal that is typically thought to not invest much time or energy in their infants. My dissertation helped determine how mothers responded to seasonal environments, while they transported infants, maintained infant contact, and prepared for future pregnancies. My investigation of nutritional food quality established three nutrient profiles selected by lactating P. coquereli. These results assisted in the identification of the nutritional and ecological underpinnings driving food selection. In the future, my results will aid in the development of feeding ecology models specific to the lactation period, and will contribute to the development of more comprehensive diets for captive sifakas. My exploration of temporal cortisol variation identified seasonal environmental stressors, examined their fluidity over time, and determined how different individual reproductive and sex classes experienced stress responses during the lactation period. My results established the biological cost of stress to individuals during lactation and promoted the application of the cortisol-adaptation and brood-values hypotheses in future studies of lemur socio- and environmental endocrinology. My dissertation contributed to understanding maternal energetics, infant development, and the exceptionally slow life histories representative of Propithecus spp.. Additionally, my research will assist in understanding the origin of female dominance in lemurs, and more comprehensively understand its contemporary manifestations.

5.2. Summary of findings

My dissertation tested one hypothesis and eleven a priori predictions methodologically grounded in numerous theoretical models and hypotheses. Predictions 1, 2, 3, 5, 7 were statistically supported. Predictions 8 and 11 were partially supported. Predictions 4, 6, 9, and 10 were not supported. Thus, there was mixed support for my hypothesis. More

182 research is needed to further our knowledge of maternal-infant relationships in harsh environments.

H1: P. coquereli mothers experience greater postnatal reproductive stress, which is measured by behavioral care-giving effort, nutritional food quality, and cortisol stress responses during early/earlier-mid lactation (designated as 1-12 weeks postnatal) (May- August) in comparison to later-mid/late lactation (designated as 13-26 weeks postnatal) (September-December) due to the increased energetic costs of lactation while simultaneously caring for dependent infants during the driest seasonal months in the austral winter when food quality is low.

P1. Maternal behavioral care-giving effort will decrease as P. coquereli infants increase in age as measured by infant transport position (ventral, dorsal, independent). Infants will spend the greatest duration in the ventral position, followed by the dorsal position, and the least duration independently from carriers from birth until weaning (twenty-six weeks postnatal). Behavioral data were recorded for 678 focal hours during 26 consecutive weeks over two birth seasons (n=10 infants, n=10 mothers, n=19 adult males, n=8 adult female non- mothers) in Ankarafantsika National Park, Madagascar. Each P. coquereli group was followed once per week using ten minute continuous focal sampling (Altmann 1974) to collect infant carrier and transport position data for 6 hours beginning at dawn after groups were located. All-occurrence sampling (Altmann 1974) was used to collect data on contact between mothers and infants. I measured the relationship between infant transport position and infant age with mixed effects linear regression models.

Infants spent significantly more time on a carrier’s ventrum than either dorsally or independently when all weeks of postnatal developed were evaluated. The average time infants spent in the ventral position was 96.69% in week 1 and 6.56% in week 26. Infants spent an intermediate amount of time in the dorsal position, as time spent dorsally was significantly less than time spent ventrally, and was greater than time spent independently. Infants were exclusively transported ventrally in weeks 1 and 2. The average time spent in the dorsal position was 22.21% in week 26. Infants spent the least

183 amount of time independent from carriers. Infants were not recorded as independent in weeks 1 and 2. The average time infants spent independently was 68.42% in week 26. Infants were transported in the ventral position more often, and thereby more dependent on their mothers in weeks 1-13 relative to 14-26.

Infant transport positions reflected infant age. I argued that transport position is indicative of ontogenetic development and P. coquereli infants were carried longer during development than previously suggested (Richard 1974; Richard 1976). I postulated that slower ontogenetic development allows mothers to carry infants for longer time periods, and demonstrated that infants were altricial for much longer than other comparably sized lemurs. It may even be that dry deciduous forest Propithecus spp. develop more slowly relative to Propithecus spp. occupying geographic areas with greater habitat quality. For example, humid forests in eastern Madagascar likely contain a greater abundance of foods high in energy dense nutrients (e.g., ripe fruit), which may impact milk quality and volume, in turn influence infant growth trajectories.

P2. P. coquereli mothers will provide the majority of infant behavioral care-giving relative to adult males and non-mother adult females as measured by the frequency and duration of infant carriage from birth until weaning (twenty-six weeks postnatal). I measured the relationship between infant carrier identity and infant age with mixed effects linear regression models using the methods stated above. Mothers carried infants for significantly longer periods than adult males and adult female non-mothers. Mothers carried infants for the longest duration in week 2 (98.94%). Time spent on mothers decreased as infants aged, and the lowest duration of transport (26.42%) occurred in week 26. Adult males carried infants for the longest period (6.17%) in week 1. Non-mother adult females transported infants the least time overall, and the longest duration (4.52%) occurred late in infant development during week 25.

Given the multitude of factors contributing to infant care and survival, I argued that the seemingly minor and variable durations of allocare should not be overlooked, since the energetic benefits from allocare to lactating females are unknown, and may in fact, be

184 substantial during the energetically constraining time of lactation combined with the seasonally challenging dry season. For example, allomaternal care and 2-year interbirth intervals may help offset direct energetic costs to mothers, in turn facilitating extended durations of infant transport by mothers. In contrast to other studies that documented very high infant mortality in Propithecus spp. (Kappeler et al. 2009; Richard et al. 2002), my study found low infant mortality (9%, n=11). I posited that maternal and allomaternal is tremendously successful in promoting high infant survivability during the first 26 postnatal weeks, and higher mortality occurs only after infants become fully independent from carriers. No predation events occurred during my study.

P3. Infant contact initiated by P. coquereli mothers will decrease while infant contact broken by mothers will increase as infants age. I measured the number of occurrences mothers initiated and broke contact with their infants using the methods outlined under prediction 1. I determined the frequency of mother-infant contact using generalized linear models with Poisson regression. Mothers initiated and broke contact less frequently compared to infants. On average, the occurrences mothers initiated and broke contact were 2.58  3.26 from 1-26 weeks postnatal and 2.88  4.66 from 1-26 weeks postnatal, respectively. As infants aged, mother-initiated contact significantly decreased as the frequency at which mothers broke contact with their infants increased. Mothers were very tolerant of infant presence until late in development, which was indicative of slow ontogenetic development and demonstrated that infants cultivate the transition to independence more than mothers.

P4. Contact with mothers initiated by P. coquereli infants will decrease while contact with mothers broken by infants will increase as infants age. I measured the number of occurrences infants initiated and broke contact with their mothers using the methods outlined under prediction 1. I determined the frequency of infant-mother contact using generalized linear models with Poisson regression. Contrary to my prediction, infants initiated and broke contact with their mothers at comparable frequencies. Infants initiated contact in all weeks aside from week 1. Infants broke contact in all 26 postnatal weeks. On average, the occurrences infants initiated and broke

185 contact were 132.42  94.27 from 1-26 weeks postnatal and 137.73  97.69 from 1-26 weeks postnatal, respectively. Infants broke contact significantly more often than mothers broke contact. It is infants, not mothers, that fostered the transition from dependence to independence.

P5. P. coquereli lactating females will select foods high in available protein following the nitrogen maximization model. I collected foods consumed by lactating females for a total of 93 focal hours, on 28 calendar days, spanned over 52 weeks in Ankarafantsika National Park. The parts of the tree consumed (e.g., fruit), diameter-at-breast-height, frequency of ingestion index (FOI), vernacular name, and scientific name were documented for each sample. Samples were dried on-site in a propane oven at a maximum of 50°C using a max/min digital thermometer until a constant weight was reached. Laboratory assays were conducted at the Nutrition Laboratory, Conservation Ecology Center, Smithsonian Conservation Biology Institute, in Washington D.C. Gravimetric analyses were used to determine nutrient composition. Assays included: nitrogen as an index for protein (n=123), neutral detergent fiber (n=123), acid detergent fiber (n=123), gross energy (kcal/g) (n=123), and ash (dry matter) as an index for total mineral content (n=119). I calculated crude protein, available protein, and non-protein gross energy from the assay values. A k-means cluster analysis was used to determine if food categories grouped together by nutrient composition. A one-way ANOVA with the k-means clusters set as the group variable was subsequently applied. Post hoc comparisons that used a Bonferroni correction were performed. I created a FOI index to measure relative food consumption by mothers. The FOI indexed was determined by the number of occurrences food item was consumed during focal follow / total number of foods consumed during focal follow. I subsequently scored food items 1-5 based on relative consumption. The FOI index was applied to k- means, and determined the relationship between each cluster and the frequency at which foods were consumed.

Lactating females selected foods high in available protein, and correspondingly high protein-to-fiber ratios, considering the foods that were seasonally available during the dry

186 season. The mean (15.9%) of high available protein foods exceeded minimum requirements for primate reproduction (14%) and infant development (7-11%) (Oftedal 1991). High available protein foods were low in fiber and non-protein gross energy. High available protein foods were high in minerals. Considering the foods seasonally available during the dry season, I argued that lactating P. coquereli selected three nutritionally distinct food types. I proposed that lactating P. coquereli successfully acquire sufficient nutrients to balance energetic costs in an unstable environment by food selection.

P6. P. coquereli lactating females will select foods high in minerals following the nitrogen maximization model. Following the methods outlined under prediction 5, I measured minerals in foods consumed by lactating females. Though minerals did not emerge as a selected food cluster, there was a significant effect of minerals consumed by lactating females. High available protein foods were highest in minerals. High fiber foods were intermediate in minerals. High non-protein gross energy foods were lowest in minerals. Lactating females may experience seasonal mineral deficiencies, and likely selected foods higher in minerals during the dry season. Isolating specific minerals (e.g., sodium) in the future will determine mineral deficiencies experienced by lactating females.

P7. P. coquereli lactating females will select foods low in fiber following the fiber limitations model. Following the methods outlined under prediction 5, I measured fiber content in foods consumed by lactating females. Contrary to my prediction, lactating females selected foods high in fiber. These foods were selected less often than foods high in non-protein gross energy, but more often than high available protein foods. High fiber foods were lowest in available protein, and foods high in non-protein gross energy and minerals were both intermediate in available protein. High fiber consumption may improve morphological digestive efficiency, promote digestive health, or simply be a residual effect from consuming seasonally available foods.

187

P8. P. coquereli lactating females will select foods high in non-protein gross energy following the energy maximization model. Following the methods outlined under prediction 5, I measured non-protein gross energy in foods consumed by lactating females. Lactating females most frequently selected foods high in non-protein gross energy. High non-protein gross energy foods were relatively low in available protein, fiber, and minerals. Non-protein energy foods provided metabolic fuel and are stored more easily than protein, which gave individuals greater access to immediate energy and energy storage. I proposed that high energy foods may help strepsirhine infants grow more quickly than haplorhine infants.

P9. P. coquereli lactating females will have significantly higher fecal cortisol levels during early/earlier-mid lactation relative to later-mid/late lactation. Fecal samples were collected weekly from lactating females (n=180 fecal samples), adult males (n=133 fecal samples), and non-lactating adult females (n=62 fecal samples) over 26 postnatal weeks in two consecutive birth seasons. Feces were collected in their entirety after P. coquereli groups began to move from their sleeping sites in the early morning. Feces were caught in the air with a plastic sheet or collected from the ground immediately following defecation, only if the individual was clearly identified and no other group members defecated simultaneously. Feces were transported within a maximum of four hours to a propane gas oven, heated to 83C , and dried thoroughly. Laboratory assays were conducted at the Wisconsin National Primate Research Center, Madison, WI. Cortisol was first extracted from samples, validated, and enzyme immunoassays were conducted. Mixed effects linear regression models measured the relationships within and between individual sex/reproductive class and cortisol variation.

Intriguingly contrary to my prediction, lactating females had higher cortisol during later- mid/late lactation (weeks 13-24) than early/earlier-mid lactating (weeks 1-12). Though this finding was not significant, a trend was detected. I argued that P. coquereli slow life history traits increased the reproductive value of every infant. Accordingly, a diminished stress response was most advantageous to increase the quality of long-term infant care. I proposed that P. coquereli invested in their infants during the first 6.5 months of life

188 more profoundly than previously thought, and that high metabolic investment documented in neonates (Young et al. 1990) transmuted to high behavioral investment in dependent infants. I posited that female dominance and feeding priority assisted lactating females with maintaining a relatively lower stress response during the first 12 postnatal weeks.

P10. P. coquereli adult males will have significantly higher fecal cortisol levels during early/earlier-mid lactation relative to later-mid/late lactation. Following the methods outlined under prediction 9, I measured fecal cortisol in adult males. The inverse of my prediction was statistically significant. Adults males had higher cortisol during later-mid/late lactation relative to early/earlier-mid lactation. I argued that P. coquereli males exhibit an elevated stress response as a result of seasonal constraints and decreased food quality/access, which is further amplified by female feeding priority. I also suggested that a higher stress response was caused by heightened agonism and competition as males began to prepare for the short breeding season. Lastly, I argued that infant age contributed to male stress responses, and older infants incited a greater stress response since they spent a significant portion of time off carriers and became more susceptible to falling, predators and infanticide.

P11. P. coquereli lactating females will have significantly lower fecal cortisol levels relative to adult males and non-lactating adult females during both lactation phases. Following the methods outlined under prediction 9, I compared fecal cortisol across sex/reproductive classes. There was a strong trend that approached statistical significance for the first half of my prediction. Lactating females had lower cortisol relative to adult males. The second half of this prediction was statistically significant. Lactating females had significantly lower cortisol relative to non-lactating adult females. I argued that dominant females, whom typically were the only females to give birth in Propithecus spp., had feeding priority over lower ranking females, which was a factor that likely contributed to significantly higher cortisol in non-lactating females.

189

5.3. Future directions of study

I would like to continue my research on maternal-infant relationships and energetics in Ankarafantsika National Park by expanding on several behavioral, nutritional, and physiological research questions raised by my dissertation. I will introduce new areas of research in morphometrics, basal metabolic rates (BMRs), milk composition, stable isotope analysis, and gut microbiomes. In the future, I will create a project that measures seasonal variation in lactating female body mass relative to other group members. Body mass is a critical measure of health and influences the level of care mothers provide infants in common marmosets (Callithrix jacchus) and other primates (see Tardif et al. 2001). Significant weight loss during the dry season has been documented in Propithecus verreauxi (see Lewis and Kappeler 2005a), but no P. coquereli weights have been recorded. Future evaluation of potential weight loss during the dry season will be an important index of maternal health during lactation. No birth weight data are available for wild P. coquereli infants. Measuring infant body mass monthly from birth through the weaning period that will quantify how quickly infants develop and be an important index for infant growth and health. Lactating female and infant weights can be collected simultaneously since infants are predominately carried by mothers.

Lemurs have low BMRs relative to haplorhines that become elevated during reproduction, though it remains unclear why or if lemurs have lower BSMs relative to other strepsirhines (Harcourt 2008; Muller 1985). BMRs have not been measured in P. coquereli. In the future, I would like to measure seasonal fluctuations in P. coquereli basal metabolic rates to help determine if low BMR is an adaptation to climatic instability in Madagascar as has been traditionally suggested (Richard and Dewar 1991). Quantifying BMRs will also help quantify the physiological effects of seasonality on P. coquereli. It will be a useful comparison to temporal cortisol variation from chapter four of my dissertation.

Lactation strategies determine lactation duration, suckling frequency, milk volume, and milk nutritional composition (Power et al. 2007). Primates generally produce low- quality, dilute milk compared to other mammals (Oftedal 1984), and milk quality

190 influences maternal care of infants and development in rhesus macaques (Macaca mulatta) (Hinde et al. 2009). In the future, I would like to collect P. coquereli milk samples during different lactation stages. Darting mothers with dependent infants certainly poses a risk to both individuals, but has been done successfully in other Propithecus spp. and is possible given that the dry deciduous forest in the Ankarafantsika National Park has high visibility especially during the dry season after the trees lose their leaves, the ability for researcher to get within close proximity of animals in the areas of the park frequented by tourists, and that infants tightly cling to their mothers (E. Louis, pers. comm.). Milk will be assayed for dry matter, crude protein (CP), fat, sugar, and gross energy (GE), and the percentage of GE from CP, fat, and sugar will be subsequently calculated from the assay following Power et al. (2007). Milk composition reflects maternal care-giving in a variety of strepsirhine genera including Varecia and Eulemur (Tilden and Oftedal 1997), and comparing the nutritional value of P. coquereli milk to other strepsirhines would assist in understanding maternal behavior and the divergent infant care strategies present across strepsirhine taxa. It would also demonstrate the quantities of nutrients P. coquereli infants receive from milk and could be paralleled to infant growth rates to see how milk nutrients influence growth trajectories.

Weaning is a process that gradually occurs over time, making it difficult to quantify since time spent on the nipple is not a reliable measure of nutrient transfer or time spent suckling (Cameron 1998; Cameron et al. 1999). A new technique using stable isotope analysis was applied to investigate the weaning process in François’ langur (Trachypithecus francoisi) by measuring stable carbon isotope ratios, nitrogen isotope ratios, and nitrogen content in feces in a mother-infant pair (Reitsema 2012). The study found that weaning went on for a considerably longer duration than was expected based on visual estimates in captive langurs. Solid food was introduced to infants at approximately two months of age, though nursing continued into the twelfth postnatal month (Reitsema 2012). In the future, applying fecal stable isotope analysis to precisely determine the duration of weaning in P. coquereli will help establish a timeline for interbirth intervals and demonstrate how quickly infants reach independence from their mothers. Quantifying stress responses during gestation and comparing those results to

191 stress responses during lactation would also be worthwhile.

Lastly, I would like to further investigate the potential benefits of high fiber consumption in lactating P. coquereli. Gut microbiomes are key components in nutrient processing and regulation. They also are crucial indicators of digestive efficiency and vary based on age and sex. For example, adult males, adult females, and juveniles all had distinct gut microbial communities, with adult females exhibiting increased microbes related to energy production and folate biosynthesis (Amato et al. 2014). Exploring variation in gut microbial communities in P. coquereli will help establish how lactating females and other group members meet the nutritional demands of different seasons by understanding their digestive health. Establishing gut microbiome profiles by assaying fecal samples for metabolite profiling and looking at metabolomes by mass spectrometry in P. coquereli will help identify trade-offs between energy requirements intake, and storage during reproduction (see Amato et al. 2014; Dufour and Sauther 2002). The questions demanding future attention are: do sifakas select high-fiber foods because they are available, and there is simply not enough high quality food available during the dry season; or is a specific quantity of high fiber food essential for maintaining gut health and proper digestive function? If low habitat quality is the driving factor causing high-fiber consumption, it would be expected that most captive sifakas would be healthy, given they do not experience the difficulty of acquiring high quality foods in the wild during the dry season. Indeed, if high fiber foods yield unknown benefits current captive sifaka diets warrant careful reexamination.

5.4. Conservation implications

The dynamic relationship between mothers and their infants determines the future reproductive success of many species. Understanding maternal-infant relationships is critical to the future conservation of all primate species. Yet, these relationships have been overlooked in evolutionary anthropology. Maternal care is the primary determinant of infant survival and fitness. In this way, mothers and their infants can be considered the fundamental unit of natural selection. Thus, species survival in the future hinges unequivocally on the relationships concerning mothers and their infants. Species will go

192 extinct without the presence of healthy infants that eventually go on to reproduce themselves. This has tremendous conservation implications for long-term species survival and validates the importance of examining the many facets of maternal-infant relationships. This relationship is especially vital in endangered species that exhibit slow life histories, as do all Propithecus spp.. The goal of my research is to demonstrate how maternal energetics, infant care, and development are compellingly interrelated; while contributing to the understanding of evolutionary dynamics at work in seasonal environments.

Madagascar is considered the single greatest conservation priority because of its exceptionally endemic wildlife and severe habitat conversion (Schwitzer et al. 2013). The cascading, primarily human driven changes to lemur habitats are having distressing ramifications. A combination of aerial photography and Landsat images demonstrate that forest cover in Madagascar decreased by 40% from the 1950s to c. 2000 (Harper et al. 2007). This same study found that areas of core forest (defined as > 1 km from a non- forest edge) were reduced nearly 80% during the same time span (Harper et al. 2007). The preferred habitat of many lemur species, including Propithecus spp., falls within the core forest range (McGoogan 2011). Recent geospatial analysis predicted that many lemur species will experience substantial reductions in habitat ranges exclusively attributed to climate change (i.e., models do not take into account sociopolitical factors, biotic interactions, etc.) (Brown and Yoder 2015). Appropriate lemur habitat does not exist in many of these geographic areas between the current and future predicted ranges, which will result in lemurs being unable to travel in search of new habitat (Brown and Yoder 2015). Lemur reproduction and infant survival have already been shown to be dependent on climatic variability (Dunham et al. 2011; Gould et al. 1999). Additional environmental pressures caused by climate change in habitats already highly compromised is of great concern, particularly for the health, survival, and relationship between mothers and their infants.

A political crisis has plagued Madagascar since a coup d’état occurred in 2009. Since then, the country has experienced a devastating increase in environmental crimes

193 including tremendous surges in hardwood timber extraction and illegal bushmeat hunting (Barrett and Ratsimbazafy 2009). The withdrawal of foreign aid along with continued political and socioeconomic instability have only intensified environmental crime and poverty, since little or no governmental regulation is present (Barrett and Ratsimbazafy 2009). Dry deciduous forests are one of the most degraded biomes in the world as a result of large-scale logging (Schwitzer et al. 2013). Less than 3% of the dry deciduous forest cover remains in Madagascar (Smith 1997), and the Ankarafantsika region is the largest remaining dry deciduous forest block (Ganzhorn et al. 2001; Schwitzer et al. 2013). The tree identification of the genera and/or species selected by lactating P. coquereli in chapter three of my dissertation can be implemented as a conservation management tool. The trees selected can be replanted in the Ankarafantsika region to help ensure P. coquereli have adequate resources during the dry season.

The hunting and consumption of bushmeat, the meat derived from wild animals, already has had calamitous ramifications throughout western and central Africa (Walsh et al. 2003). Increased zoonotic disease transmission (e.g., Ebola outbreaks) and predicted species extinctions have been directly correlated with bushmeat hunting and its consumption (Brashares et al. 2001; Chapman et al. 2005; Gillespie and Chapman 2008; Leroy et al. 2004; Rwego et al. 2008). For example, the indri (Indri indri) and diademed sifakas (Propithecus diadema) currently face extirpation in northwestern Madagascar (Jenkins et al. 2011). The demand for bushmeat has recently been on the rise in Madagascar, with disturbingly high consumption rates documented for protected species, including those listed as critically endangered by the IUCN (Jenkins et al. 2011; Randriamamonjy et al. 2015; Razafimanahaka et al. 2012; Schwitzer et al. 2014). Geographic cultural taboos that had previously prevented bushmeat consumption in Madagascar have become futile, and the demand for bushmeat continues to grow with the influx in human migration across the country (Jenkins et al. 2011). Madagascar is one of the poorest countries in the world, with more than 92% of the population living on less than $2 USD per day (World Bank 2013). This staggering statistic makes lemur conservation immensely difficult; however, the promotion of ecotourism, creation of protected areas managed by Malagasy citizens at a community level, and the expansion

194 of long-term research presence will assist in decreasing hunting while simultaneously alleviating poverty (Schwitzer et al. 2014).

Bushmeat is often critical food source for the rural poor in continental Africa and Madagascar (Gardner and Davies 2013; Jenkins et al. 2011; Randriamamonjy et al. 2015). Bushmeat markets are both local and commercial; however, a new market has emerged in recent years focused on selling lemur bushmeat as a luxury commodity to consumers in Madagascar (Barrett and Ratsimbazafy 2009; Schwitzer et al. 2014). The transformation from lemurs being hunted on a relatively subsistence-based, local scale to being exploited as a national urban delicacy is particularly concerning given the magnitude at which environmental crime presently occurs. Commercial exploitation is more environmentally harmful than subsistence hunting due to the demand for high meat volume, high commercial prices bushmeat fetches at markets, and infrastructure built in forested habitats to develop and maintain these operations (Alvard et al. 1996; Fa et al. 2003; Fa et al. 1995). Wealth is considered a primary factor influencing bushmeat consumption, as animal protein is typically more expensive than other food items (Fa et al. 2003). Nonetheless, this relationship is not straightforward since local preferences for wild versus domestic meat, access to guns/snares, and employment alternatives for hunters are crucial predictors of bushmeat consumption (Schenck et al. 2006; Wilkie and Carpenter 1999). Hunting bushmeat is less expensive than purchasing domestic meat from local markets, though recent studies agree bushmeat is less preferred by rural Malagasy to domestic meat and fish (Gardner and Davies 2013; Jenkins et al. 2011). It is important to note that this tendency may be changing in accordance with migrations and the trend of bushmeat being viewed as a commercial, luxury commodity. P. coquereli are particularly susceptible to hunting, as their large body size and diurnal activity pattern make easy targets for hunters. Additionally, slow life histories make it very difficult for P. coquereli populations to recover from hunting pressures. Introducing inexpensive, readily available protein substitutes to bushmeat that appeal to urban and rural residents alike, while enforcing wildlife and firearm laws are crucial steps urgently needed moving forward (Gardner and Davies 2013; Jenkins et al. 2011; Randriamamonjy et al. 2015; Razafimanahaka et al. 2012).

195

There are many feral dogs living in the forested areas surrounding the village of Ampijoroa. These dogs pose a direct threat to all flora and fauna in Ankarafantsika National Park. I personally witnessed a feral dog attack an adult male sifaka in October 2011. I recommend that all dogs residing in the park be spayed or neutered annually, at minimum. Local dogs could be trained as conservation detection dogs to protect all animals living in the park. Conservation detection dog organizations are currently having success in reducing poaching and illegal wildlife trafficking in other parts of the developing and developed world. Trained professionals should euthanize injured and ill feral dogs.

A three-year emergency plan was developed by the IUCN and collaborating scientists in 2013 for conserving lemurs throughout Madagascar (Schwitzer et al. 2013). Scientists with long-term research projects established in Ankarafantsika National Park outlined their site-specific strategy with recommendations to permanently establish forest wardens, forest agents, researchers, and conservationists in the Ankarafantsika region to promote long-term protection of the ecosystem (Radespiel and Razafindramanana 2013). Advanced training in biodiversity assessment and monitoring by park personnel is another key component to improve patrolling, anti-fire, and anti-poaching efforts in Ankarafantsika National Park (Radespiel and Razafindramanana 2013). Rapid assessment programs, especially in more remote areas of the park, occurring on an annual basis will help monitor species livelihoods (Radespiel and Razafindramanana 2013). Lastly, conservation education programs in villages, and regular meetings with village elders on conservation related issues throughout Ankarafantsika National Park is essential to the success of the lemur conservation plan (Radespiel and Razafindramanana 2013).

5.5. Evolutionary implications

Individual variation is the fundamental unit of natural selection, and maternal effort is the primary determinant of offspring fitness (Maestripieri and Mateo 2009). Maternal effects are most relevant in mammalian evolutionary dynamics, and expressly in primates, due to the extensive care-giving influencing offspring even after the weaning process (Reinhold

196

2002). However, evaluating maternal effects in an evolutionary and adaptive context has only been considered relatively recently in wild animals (Bernardo 1996). My dissertation examined how lactating P. coquereli behaviorally, nutritionally, and physiologically responded to a stochastic environment. Few studies have focused on the process of transitional feeding in strepsirhines, where infants receive nutrition from milk and solid foods (Sellen 2007). This will advance current discussion in evolutionary anthropology and biology on the importance of maternal care-giving.

The paleobiological record shows evidence of the adaptive significance of hominid maternal care-giving strategies in stochastic environments (Foley 1995). Hominids are specious, suggesting high variability in maternal care-giving strategies relative to modern humans (Anton 2007). Relatively large brain size is a defining characteristic in human evolution and is associated with high metabolic expenditure of maintaining this organ (reviewed in Foley 1995). Increased encephalisation is correlated with this increased energetic demand, thereby requiring mothers to invest in energetically expensive offspring (Foley 1995). My dissertation and future studies of extant lemur maternal-infant relationships will provide insights to early changes in pre-australopithecine maternal life histories and reconstructing ancestral social organization. My research will assist in understanding the emergence of derived human characters. For example, the emergence of bipedality and alloparental care were dependent on hominid weaning strategies (Sellen 2007).

Malagasy strepsirhines are the most basal extant representations of Eocene primates, such that Microcebus spp. are typically considered a living depiction of fossil taxa, as they are most closely representative extant genus to the proposed ancestral weight (10-15 g) of the very first primates that were shrew-sized (Gebo 2004). Some species of subfossil lemurs briefly coexisted with human populations until extinctions resulting from human colonization and overhunting (Burney et al. 2004). Behavioral studies of extant lemurs will serve as models for innovative theories on primate evolution and subfossil lemur extinctions in Madagascar. This study will contribute to the ongoing debate on the evolution of maternal care strategies in early primates and hominids.

5.6. References

2013. Madagascar: Measuring the Impact of the Political Crisis. World Bank, Washington DC: World Bank. Altmann J. 1974. Observational study of behavior: sampling methods. Behaviour 49:227-267. Alvard M, Robinson J, Redford K, and Kaplan H. 1996. The sustainability of subsistence hunting in the neotropics. Conservation Biology 11:977-982. Amato K, Leigh S, Kent A, Mackie R, Yeoman C, Stumpf R, Wilson B, Nelson K, White B, and Garber P. 2014. The role of gut microbes in satisfying the nutritional demands of adult and juvenile wild, black howler monkeys (Alouatta pigra). American Journal of Physical Anthropology 155:652- 664. Anton S. 2007. Climatic influences on the evolution of early Homo? Folia Primatologica 78:365-388. Barrett M, and Ratsimbazafy J. 2009. Luxury bushmeat trade threatens lemur conservation. Nature 461:470. Bernardo J. 1996. Maternal effects in animal ecology. American Zoologist 36:83-105. Bonier F, Moore I, Martin P, and Robertson R. 2009. The relationship between fitness and baseline glucocorticoids in a passerine bird. General and Comparative Endocrinology 163:208-213. Brashares J, Arcese P, and Sam M. 2001. Human demography and reserve size predict wildlife extinction in West Africa. Proceedings of The Royal Society B: Biological Sciences 268:2473-2478. Brown D, and Yoder A. 2015. Shifting ranges and conservation challenges for lemurs in the face of climate change. Ecology and Evolution 5:1131-1142. Burney D, Burney LP, Godfrey LR, Jungers W, Goodman SM, Wright H, and Jull A. 2004. A chronology for late prehistoric Madagascar. Journal of Human Evolution 47:25-63. Cameron E. 1998. Is suckling behaviour a useful predictor of milk intake? A review. Animal Behaviour 56:521-532. Cameron E, Stafford K, Linklater W, and Veltman C. 1999. Suckling behaviour does not measure milk intake in horses. Animal Behaviour 57:673-678. Chapman C, Gillespie T, and Goldberg T. 2005. Primates and the ecology of their infectious diseases: how will anthropogenic change affect host-parasite interactions? Evolutionary Anthropology 14:134- 144. Dufour D, and Sauther M. 2002. Comparative and evolutionary dimensions of the energetics of human pregnancy and lactation. American Journal of Human Biology 14:584-602. Dunham AE, Erhart EM, and Wright PC. 2011. Global climate cycles and cyclones: consequences for rainfall patterns and lemur reproduction in southeastern Madagacar. Global Change Biology 17:219-227. Eisen E, and Saxton A. 1983. Genotype by environment interactions and genetic correlations involving two environmental factors. Theoretical and Applied Genetics 67:75-86. Fa J, Currie D, and Meeuwig J. 2003. Bushmeat and food security in the Congo Basin: linkages between wildlife and people's future. Environmental Conservation 30:71-78. Fa J, Juste J, Perez del Val J, and Castroviejo J. 1995. Impact of market hunting on mammal species in Equatorial Guinea. Conservation Biology 9:1107-1115. Foley R. 1995. Evolution and adaptive significance of hominid maternal behaviour. In: Pryce C, Martin R, and Skuse D, editors. Motherhood in Human and Nonhuman Primates: Biosocial Determinants Basel Karger. Ganzhorn J, Lowry P, Schatz G, and Sommer S. 2001. The biodiversity of Madagascar: One of the world's hottest hotspots on its way out. Oryx 37:115-118. Gardner C, and Davies Z. 2013. Rural bushmeat consumption within multiple-use protected areas: qualitative evidence from southwest Madagascar. Human Ecology 42:21-34. Gebo D. 2004. A shrew-sized origin for primates. Yearbook of Physical Anthropology 47:40-62. Gillespie T, and Chapman C. 2008. Forest fragmentation, the decline of an endangered primate, and changes in host-parasite interactions relative to an unfragmented forest. American Journal of Primatology 70:222-230. Gould L, Sussman RW, and Sauther ML. 1999. Natural disasters and primate populations: the effects of a 2-year drought on a naturally occurring population of ring-tailed lemurs (Lemur catta) in

197 198

southwestern Madagascar. International Journal of Primatology 20:69-84. Harcourt A. 2008. Are lemurs' low basal metabolic rates an adaptation to Madagascar's unpredictable climate? Primates 49:292-294. Harper G, Steininger M, Tucker C, Juhn D, and Hawkins F. 2007. Fifty years of deforestation and forest fragmentation in Madagascar. Environmental Conservation 34:325-333. Heidinger B, Nisbet I, and Ketterson E. 2006. Older parents are less responsive to a stressor in a long-lived seabird: a mechanism for increased reproductive performance with age? Proceedings of the Royal Biological Society B 273:2227-2231. Hinde K, Power M, and Oftedal O. 2009. Rhesus macaque milk: magnitude, sources, and consequences of individual variation over lactation. American Journal of Physical Anthropology 138:148-157. Jenkins R, Keane A, Rakotoarivelo A, Rakotomboavonjy V, Randrianandrianina F, Razafimanahaka H, Ralaiarimalala S, and Jones J. 2011. Analysis of patterns of bushmeat consumption reveals extensive exploitation of protected species in eastern Madagascar. PLoS ONE 6:1-12. Kappeler P, Mass V, and Port M. 2009. Even adult sex ratios in lemurs: potential costs and benefits of subordinate males in verreaux’s sifaka (Propithecus verreauxi) in the Kirindy Forest CFPF, Madagascar. American Journal of Physical Anthropology 140:487-497. Leroy E, Telfer P, Kumulungui B, Yaba P, Rouquet P, Roques P, Gonzalez J, Ksiazek T, Rollin P, and Nerrienet E. 2004. A serological survey of Ebola virus infection in Central African nonhuman primates. Journal of Infectious Diseases(190):1895-1899. Lewis RJ, and Kappeler PM. 2005. Are Kirindy sifaka capital or income breeders? It depends. American Journal of Primatology 67:365-369. Maestripieri D, and Mateo J. 2009. The role of maternal effects in mammalian evolution and adaptation In: Maestripieri D, and Mateo J, editors. Maternal Effects in Mammals. Chicago University of Chicago Press. p 1-10. Mattson W. 1980. Herbivory in relation to plant nitrogen content. Annual Review of Ecology and Systematics 11:119-161. McGoogan K. 2011. Edge effects on the behaviour and ecology of Propithecus coquereli in Northwest Madagascar [dissertation]. Available from TSpace Repository, 1807/31861: University of Toronto 225 p. Milton K. 1979. Factors influencing leaf choice by howler monkeys: a test of some hypotheses of food selection by generalist herbivores. American Naturalist 114:362-378. Muller E. 1985. Basal metabolic rates in primates: the possible role of phylogenetic and ecological factors. Physiology 81:707-711. Oftedal O. 1984. Milk composition, milk yield and energy output at peak lactation: a comparative review. Symposium Zoological Society of London 51:33-85. Oftedal O. 1991. The nutritional consequences of foraging in primates: the relationship of nutrient intakes to nutrient requirements. Philosophical Transactions of the Royal Society B 334:161-170. Power M, Verona C, Ruiz-Miranda C, and Oftedal O. 2007. The composition of milk from free-living common marmosets (Callithrix jacchus) in Brazil. American Journal of Primatology 69:1-9. Radespiel U, and Razafindramanana J. 2013. Ankarafantsika National Park. In: Schwitzer C, Mittermeier R, Davies N, Johnson S, Ratsimbazafy J, Razafindramanana J, Louis Jr. E, and Rajaobelina E, editors. Lemurs of Madagascar: A Strategy for Their Conservation 2013-2016. Bristol, United Kingdom: IUCN SCC Primate Specialist Group, Bristol Conservation and Science Foundation, and Conservation International. Randriamamonjy V, Keane A, Razafimanahaka J, Jenkins RKB, and Jones JPG. 2015. Consumption of bushmeat around a major mine, and matched communities, in Madagascar. Biological Conservation 186:35-43. Razafimanahaka JH, Jenkins RKB, Andriafidison D, Randrianandriananina F, Rakotomboavonjy V, Keane A, and Jones JPG. 2012. Novel approach for quantifying illegal bushmeat consumption reveals high consumption of protected species in Madagascar. Oryx 46:584-592. Reinhold K. 2002. Maternal effects and the evolution of behavioural and morphological characters: a literature review indicates importance of extended maternal care. Journal of Heredity 93:400-405. Reitsema L. 2012. Introducing fecal isotope analysis in primate weaning studies. American Journal of Primatology 74:926-939. Richard A. 1974. Intra-specific variation in the social organization and ecology of Propithecus verreauxi. Folia Primatologica 22:178-207.

199

Richard A. 1976. Preliminary observations on the birth and development of Propithecus verreauxi to the age of six months. Primates 17:357-366. Richard A, and Dewar R. 1991. Lemur ecology. Annual Review of Ecology and Systematics 22:145-175. Richard A, Dewar R, Schwartz M, and Ratsirarson J. 2002. Life in the slow lane? Demography and life histories of male and female sifaka (Propithecus verreauxi verreauxi). The Zoological Society of London 256:421-436. Rossiter M. 1996. Incidence and consequences of inherited environmental effects. Annual Review of Ecology and Systematics 27:451-476. Rwego I, Isabirye-Basuta G, Gillespie T, and Goldberg T. 2008. Gastrointestinal bacterial transmission among humans, mountain gorillas, and livestock in Bwindi Impenetrable National Park, Uganda. Conservation Biology 22:1600-1607 Schenck M, Effa Nsame E, Starkey M, Wilkie D, Abernethy K, Telfer P, Godoy R, and Treves A. 2006. Why people eat bushmeat: results from two-choice, taste tests in Gabon, Central Africa. Human Ecology 34(3):433-445. Schoener T. 1971. Theory of feeding strategies. Annual Review of Ecology and Systematics 2:369-404. Schwitzer C, Mittermeier R, Davies N, Johnson S, Ratsimbazafy J, Razafindramanana J, Louis Jr. E, and Rajaobelina E. 2013. Lemurs of Madagascar: A Strategy for Their Conservation 2013-2016. Bristol, United Kingdom: IUCN SCC Primate Specialist Group, Bristol Conservation and Science Foundation, and Conservation International. Schwitzer C, Mittermeier R, Johnson S, Donati G, Irwin M, Peacock H, Ratsimbazafy J, Razafindramanana J, Louis EE, and Chikhi L. 2014. Averting lemur extinctions amid Madagascar's political crisis. Science 343: 842-843. Sellen D. 2007. Evolution of infant and young child feeding: implications for contemporary public health. The Annual Review of Nutrition 27:123-148. Smith A. 1997. Deforestation, fragmentation, and reserve design in western Madagascar. In: Lawrence W, and Bierregaard O, editors. Tropical Forest Remnants, Ecology, Management, and Conservation of Fragmented Communities Chicago: University of Chicago Press. p 415-441. Stephens D, and Krebs J. 1986. Foraging Theory. Princeton, NJ: Princeton University Press. 247 p. Tardif S, Power M, Oftedal O, Power R, and Layne D. 2001. Lactation, maternal behavior and infant growth in common marmoset monkeys (Callithrix jacchus): effects of maternal size and litter size. Behavioral Ecology and Sociobiology 51:17-25. Tilden C, and Oftedal O. 1997. Milk composition reflects patterns of maternal care in prosimian primates. American Journal of Primatology 41:195-211. Walsh PD, Abernethy KA, Bermejo M, Beyers R, De Wachter P, Akou ME, Huijbregts B, Mambounga DI, Toham AK, Kilbourn AM et al. . 2003. Catastrophic ape decline in western equatorial Africa. Nature 422:611-614. Wilkie D, and Carpenter J. 1999. Bushmeat hunting in the Congo basin: an assessment of impacts and options for mitigation. Biodiversity and Conservation 8:927-955. Young A, Richard A, and Aiello L. 1990. Female dominance and maternal investment in strepsirhine primates. American Naturalist 135:473-488.