THE IMPACT OF HABITAT QUALITY ON FEMALE RED COLOBUS MONKEY (Procolobus rufomitratus ) REPRODUCTION

BY

KRISTA MARIE MILICH

DISSERTATION

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Anthropology in the Graduate College of the University of Illinois at Urbana-Champaign, 2012

Urbana, Illinois

Doctoral Committee:

Associate Professor Rebecca M. Stumpf, Chair Professor Janice M. Bahr Professor Colin A. Chapman Professor Paul A. Garber Assistant Professor Laura L. Shackelford

ABSTRACT

This study examines the relationship between habitat quality and reproduction in female red colobus monkeys in Kibale National Park, Uganda. Because energetic constraints impact reproductive function in female primates, this dissertation includes data on feeding ecology, activity budgets, mating behaviors, ketone levels, and reproductive hormone concentrations.

This multi-faceted approach tests a model in which habitat quality impacts diet and behaviors, which ultimately effect reproductive success. By including four research groups, two in logged areas and two in old-growth areas, this research adds to our understanding of the reproductive behaviors and physiology of red colobus monkeys as well as a comparative perspective for understanding intraspecific differences for primates living in different quality habitats. Red colobus monkeys in logged areas eat more tree species, smaller amounts of each of these species, and include novel foods in their diet compared to red colobus living in old growth forest. They also spend more time feeding than individuals in old-growth areas. Female red colobus monkeys in logged areas are more constrained in their mating behaviors than females in old growth forests. Specifically, they have lower overall copulation rates and focus mating efforts during conceptive periods; whereas, females in old-growth areas mate more often, including during non- conceptive periods. Females in logged areas also have a shorter duration of maximal tumescence. This pattern supports predictions of the cost-of-sexual-attraction hypothesis, where females in poor quality habitats adjust mating behaviors to maximize the likelihood of conception while minimizing their energetic output. However, this energetic strategy may come at a cost of reducing the ability for females to use nonconceptive sex as a reproductive strategy to confuse paternity, avoid infanticide, and enhance male support. Despite differences in habitat quality, females in logged and old-growth areas do not differ as expected in their reproductive

ii hormone concentrations, ketone levels, or degree of reproductive seasonality. Phenotypic and behavioral plasticity in diet and energy expenditure may allow female red colobus monkeys to maintain reproductive function in lower quality habitats that are stable and protected. The ability to adjust to previously logged habitats that are now protected has important implications for conservation management strategies, such as prioritizing areas for protection, and provides insight into how extant primates, including humans, adjust to environmental change.

iii

For Gary Macko, my uncle and most enthusiastic supporter.

iv ACKNOWLEDGMENTS

It is only appropriate to start a dissertation on females with an acknowledgment of all the strong, inspiring, supportive, loving, funny, talented, intelligent women in my life. The first, of course, is my mother. Karen Macko-Milich taught me how to care for people, how to overcome obstacles, how to be independent and capable. She always supported me in all of my academic endeavors, even when she did not understand my choices or worried about my safety. I owe all of my professional success to her. My sister, Jennifer Milich, who was the quintessential older sister—teaching me how to be cool, making me tough, answering all the questions I couldn’t ask my mom, putting up with my antics, and making sure I didn’t make too many stupid mistakes along the way. I followed in her footsteps in many ways, including my love of animals, my vegetarian diet, and my bachelors in zoology. My grandmother, my Baba, Josephine Macko.

While I did not inherit her love of shopping, her gorgeous blue eyes, or her adorable ability to be a proper lady, I have benefited from her tender care, her sense of humor, and her perseverance.

She kept me supplied with fried egg sandwiches and haluski on all my trips home and brightened my days with weekly phone calls to check on me no matter what city I lived in. My wonderful aunts – Aunt Cathy, Aunt Lana, Aunt Pat, and Aunt Susie – who filled my childhood with many happy memories. For both my sister and me, these women have been there to celebrate our greatest successes and to sympathize with our greatest losses. My rockstar cousins, Kim, Kelly, and Jess, who were so cool and so exciting and who I idolized growing up. I believe that even to this day, Kim’s early advice on how to manage my big hair is what has kept me from looking entirely like a crazy scientist. Kelly was my first and best pen pal and a welcome confident.

Jessica made me laugh endlessly every time I saw her, and to me, she was larger than life. And my Grandma Milich. I know she would be very proud of me for this accomplishment.

v My academic development has been equally inspired by wonderful women. My advisor,

Rebecca Stumpf, who mentored me through my moments of excitement and frustration. When I applied to graduate schools, I wanted to combine my interest in conservation with my interest in reproductive biology. Becky helped me to design a project that did just that. In my first semester of graduate school, Becky’s class on Behavioral Endocrinology provided me with the basis for my dissertation research. Her intellectual contribution not only strengthened my dissertation and graduate career, but has also made a lasting impression that will shape my approach to future research. I hope to be able to emulate how kindly and gracefully Becky deals with difficult situations. At times, she shared her personal stories of navigating the field of primatology as a woman, which was exactly the type of insight I was hoping to gain from having a female advisor. My lab advisor and close mentor, Janice Bahr, has provided me endless support in every possible way during my graduate career. She welcomed me into her lab and provided me with whatever I needed for my research. She taught me reproductive biology and lab methods in endocrinology, but she also taught me how to be an academic and a mentor.

Outside of the lab, she made me a part of her family and was always there for me. When I was hobbling around on crutches in the snow, she would pick me up and take me wherever I needed to go. She has been like a mother to me. My committee member, Laura Shackelford, who provided me with valuable feedback on my dissertation and invaluable moral support during my graduate career. I am always searching for positive female role models in my field—those that are successful in their careers and happy in their lives. Laura is exactly that type of role model and she gives me hope for my life in this field.

My friends, who are like family to me, include equally inspiring female figures. Amanda

Byers, Angie Puertas, Beth Kostelnick, and Sue Wismar – childhood pals that still remain my

vi closest girlfriends. Beth is my oldest friend and still so important in my life, Angie is the sweetest person I know and I appreciate the hope she has for me, Amanda and I can spend hours in deep discussion, and Sue is hilarious and has gone great lengths to help me out (like driving 5 hours to pick me up when I’ve been sick). Sara Critchfield is a kindred spirit who has taken on many adventures with me from our first moments as trees in a high school play to caring for me when I had typhoid in India. Daniela Atanasovski is an inspiration as a successful female nuclear engineer. Her family, most especially Marijana and Vesna, have been a lovely second family to me and help me to engage with my Eastern European roots. All of these friends are near and dear to my heart and have offered me shelter and encouragement along the way.

Emily Otali and Sarah Paige were my two closest friends during field work. They are both strong, intelligent women who were always willing to help deal with logistics for my project when I was not able. More importantly, they offered much needed moral support.

I must now turn my focus to the amazing male role models in my life. The person I attribute my career as a primatologist to is my grandfather, PapPap, Joseph Macko. From a young age, he showed me the beauty of the natural world – taking me fishing, letting me pick tomatoes from the garden, and spending time outside. His career as an amateur archaeologist

(his large collection of Native American artifacts made him something of a local celebrity) and his love of nature somehow combined in me to spark the fire of a biological anthropologist. In many more ways than this, PapPap left a lasting impression on me and influenced the path of my life. Each of my uncles has played a very important role in my life. As a biology teacher, my

Uncle Jim was never short of things to teach me about biology and the natural world. He is an outdoorsman through and through and he still provides me with endless amounts of delicious, fresh vegetables, opportunities to wander around in the forest with him, and interesting insight

vii into the biological world. My Uncle Ron is without a doubt the most influential person in giving me the opportunity to travel the world. I am fortunate that he is a talented international airline captain and that he is generous with his family passes that allow me to go on many adventures around the world, including to see primates in distant habitats. He (and my Aunt Lana) have been so supportive in this way. My Uncle Gary who was more excited about my career than anyone else I know. At times, he was even more excited about it than me, and I learned things about my field from the stories he would tell me about wildlife documentaries he watched. He was always on the look out to see a place I had traveled to or a person I had worked with. And though I know he always hated to see me go, often giving me a hasty goodbye so that I would not see him cry, he never tried to stop me. And my Uncle Ed who is the definition of tough guy, but can still show his soft side. He’s always welcoming and wonderful and even reluctantly tells me he loves me when we say goodbye – putting his tough-guy image in serious jeopardy.

I have had very strong male mentors in my academic career. Colin Chapman and Paul

Garber are both committee members who have contributed greatly to my graduate education.

Both have provided me with significant feedback on my proposals for research funding and my dissertation chapters. Colin has offered me endless support as a researcher at Kibale, including both theoretical feedback to understand my research results and logistical help with storing samples, getting paperwork processed, and much more. He always takes time to provide me with feedback and is extremely kind with his time. He has patiently helped me work through difficult obstacles when I have felt overwhelmed or as if I’ve hit a roadblock. His positive attitude and friendly guidance has been indispensible. Paul has been very influential through my entire graduate career. During my first semester, I took his seminar where he taught me to think critically about research. I attribute my critical thinking skills to him, in particular. He also was

viii willing to meet with me to discuss my research at each stop of the process leading up to my defense. My undergraduate advisor, John Mitani, was a great influence on my scientific career.

I have fond memories of discussing homosexual behaviors in gorillas with him as I worked to complete my honors thesis. Richard Adelman was my boss as a research assistant in the

Gerontology Department all four years at Michigan. He not only provided me with an amazing research opportunity that would shape my research interests for the rest of my life, but he also provided much needed mentorship on how to be an academic professional. He and his wife,

Lynn, provided me with warm hospitality throughout my undergraduate degree.

There are truly too many people to name in these acknowledgments, but I will try my best. Eric Radke has been a best friend and life saver since high school. He has always taken the time to listen to my hardships and cheer me up when I’m down, celebrate my victories, and goof off with me all along the way. I owe him much more than I can express here. Graduate school has provided me with many wonderful friends. Scott Williams and Milena Shattuck became lifelong friends as we tackled both big and small hurdles together in our graduate careers.

Ashley Stinespring-Harris and Nathan Harris are both exceptional people and an adorable couple who opened their home to me and provided me with a truly wonderful atmosphere for my final months in Urbana-Champaign. Jamie Arjona is a hilarious, wild lady. Nicoletta Righini is amazing and talented and beautiful. Renee Archibald is a dance genius – I can’t fathom how she does what she does, but I’m so glad I get to hang out with her and glance into her world. Phil

Slater has been a true friend and I hope to see him in East Africa in the future. Carl Lehnen is a great friend and scholar. Carolina Sternberg, Rose Kaczmarowski, and Jean Stover are interesting and vibrant women who I am happy to call my friends. Alexander Georgiev is a field friend who has become a life-long friend and collaborator. Thanks to Don Darga for his funny

ix emails and his endless proofreading and file converting. And to Jacklyn Ramsey and Amanda

Ellwanger who learned how to be field primatologists with me. And Jeff Cruder and Chris

Costigan who kept me laughing and kept my computer working. Thank you also to good friends

Young Ji Kim, Britni Day, Wilbur Chang, and Steve Dannemiller. I have many great UIUC friends from Anthropology – Petra Jelinek, Christine Deloff, Liz Mallott, Amanda Butler,

Melissa Raguet-Schofield, Kati Fay, Melissa Baltus, Sarah Baires, Talia Melber, Kari Zobler,

Jenn Baldwin, Mark Grabowski, Greg Blomquist, and Martin Kowalewski; and from biology and ecology – Kate Laskowski, Greg Spyreas, John Andrews, James Walsh, and Katie Amato.

Elena Pedroso is a wonderful office mate and friend. Annie Newell-Fugate made my many hours in the lab much more enjoyable. And I appreciate the rest of my Animal Science family –

Andrea Braundmeier, Lindsey Brunett, Vanessa Peters, Fae Koohestani, Brad Diagnea, Chanaka

Rabel, Tahir Abbas, JiaJia Bi, Malavika Adur, Sergio Machado, David Miller, and Josh Smith.

All of those who have taken the time to mentor me – Katie MacKinnon, Charles

Roseman, Ripan Malhi, Lyle Konigsberg, Amy Lu, Tom Frank, Steve Leigh, Tom Gillespie,

Ken Paige, and Romana Nowak. Jennifer Sevin was an amazing mentor during my Smithsonian days and is still a great resource in the field of conservation. Melissa Emery Thompson and Toni

Ziegler have been great for all my hormone questions. Thank you to the women who really got things done – Liz Spears, Julia Spitz, Karla Harmon, HiDee Ekstrom, and Nancy Henry.

Thank you to the wonderful people of Urbana-Champaign who have been a pleasure to interact with. Most especially, thanks to all of my friends from Quality who provided me with a comfortable spot to prop up my broken foot and revise my dissertation – Neil, Aaron, Terry,

Mark, Adrian, Kris, Eric, Jason, Emma, Jessica, and Emily. You have certainly made my

x presentations and articles more entertaining and my life here a lot of fun. And thank you, Julian, for getting me through the last push.

I was fortunate to work with many wonderful assistants both in Uganda and in the USA.

Josie Chambers was a great companion in the forest and quickly became a good friend. I have enjoyed watching her advance in the field and grow as a young scientist. Rachel Petersen has been a delight in the lab and a pleasure to mentor. Thank you, also, to Heather Lopez, Hasib

Neaz, Jacob Schanks, Elsa Holden, and Maggie Malinowski. Kugonza Moses has been my most trusted field assistant and has been such an asset to my project. Sabiiti Richard, Atwijiuze Seeze,

Koojo John, Busobozi Richard, Bangirana Wilfred, Ongwang Jimmy, and Mutebi Michael were all outstanding field assistants and indispensible on this project. This research would not have been possible without their hard work. Thank you to my other friends from the field –Margret

Abwooli, Alice Akiiki, Jessica Rothman, Mutegeki Richard, Omeja Patrick, Twinmugisha

Dennis, Kato Innocent, Dumba Charles, John and Lydia Kasenene, Jerry Lwanga, Elizabeth

Ross, Richard Wrangham, Martin Muller, Aaron Sandel, Caroline Deimel, and Pawel Fedurek.

Thank you to all of my funding sources – National Science Foundation, LSB Leakey

Foundation, Primate Conservation Inc, The Explorers Club, The Sophie Danforth Conservation

Fund, Primate Action Fund, Idea Wild, UIUC Graduate College, Columbus Zoo Conservation

Fund, and American Society of Primatologists. Thank you to the Uganda Wildlife Authority and

Uganda National Council for Science and Technology for research permission. Thank you to the

Makerere University Biological Field Station for administrative support and a friendly place to conduct this research. Finally, and most importantly, thank you to the people of Uganda who welcomed me so warmly into their country. A piece of my heart will always stay in Uganda. I hope to see you all again very soon, my friends…

xi TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION …………………………………………………………... 1

CHAPTER 2: FEMALE RED COLOBUS MONKEYS MAINTAIN AN IDEAL FREE DISTRIBUTION THROUGH NOVEL FORAGING STRATEGIES IN LOGGED FORESTS …………………………………………………... 25

CHAPTER 3: TIMING IS EVERYTHING: EXPANDING THE COST-OF-SEXUAL-ATTRACTION HYPOTHESIS ……………………………………. 68

CHAPTER 4: FEMALE RED COLOBUS MONKEYS MAINTAIN REPRODUCTIVE FUNCTION IN DEGRADED HABITATS …………………………... 92

CHAPTER 5: CONCLUSION ………………………………………………………….…. 131

APPENDIX A: FEMALE ESTRADIOL AND PROGESTERONE GRAPHS …………… 139

xii

CHAPTER 1

INTRODUCTION

Impetus for this study

This project was undertaken to examine the impact of anthropogenic forest disturbance on the reproduction of an endangered species. Numerous studies have documented the impact of habitat quality on nonhuman primates in general (Atlmann & Muruthi 1988; Boyle & Smith

2010; Chapman et al. 2006; Eley et al. 1989; Isbell & Young 1993; Iwamoto & Dunbar 1983;

Marsh 2003; Ratsimbazafy 2007; Silva and Ferrari 2009; Singh & Vivanthe 1990; Wong &

Sicotte 2007; and Watts 1988) and particularly on their reproduction (Bercovitch 2001; Emery

Thompson et al. 2007; Knott 1999; Lipson 2001; Wrangham 2002). These studies illustrate the variety of responses that primates can have to different environments; therefore, it is important to continue to examine primate responses to habitat changes and determine proper approaches to conservation efforts. Furthermore, understanding the flexibility of primates to adjust to changing environments provides important insight into understanding how extant primates, especially humans, dealt with environmental changes throughout evolution and the relevance of different strategies on reproductive success.

The specific research design of this project was based on a shorter study that I completed from 2007-2008. During that study, I analyzed reproductive hormone concentrations of fecal samples from adult male and female red colobus in six different red colobus monkey groups.

Three of these groups lived around the Makerere University Biological Field Station at

Kanyawara in Kibale National Park, Uganda, and three of them lived in small, unprotected forest fragments outside of the park. The results of this study showed that habitat quality was not associated with significant differences in reproductive hormone concentrations in male red

1 colobus monkeys, but that female reproductive hormones were significantly lower in the fragments than in the park. Based on these results, I decided to focus this dissertation research on adult female red colobus monkeys.

While conducting field research in 2007, it became clear that fragments presented theoretical and logistical problems for a larger study. The fragments were all surrounded by agricultural fields where pesticides and other chemicals were being used. Agricultural products like these are often endocrine disruptors (e.g. atrazine, Hayes 2005), which can cause reproductive problems. The fragments also often had domestic animals in them, including cattle and dogs. Dogs sometimes eat red colobus and often bark at them, which can lead to additional stress for these groups. The cattle carry parasites that are known to infect red colobus (Gillespie

2006). All of these are confounding factors that make it difficult to examine the impact of forest structure on reproduction. Furthermore, the groups in the fragments were not habituated and were hard to follow and identify. I felt it was unethical to habituate animals that lived in such close proximity to humans for many reasons, including that they would be at increased risk of being killed by humans or their dogs and that the humans may then be burdened by increased crop raiding by the red colobus. For these reasons, I decided to not study red colobus in forest fragments for my dissertation.

Kibale Forest was not established as a National Park until 1994. Prior to this, parts of the forest were commercially logged. Thus, there are habitats of differing qualities within the park.

I selected groups that live well within the park (either in the old-growth area or the heavily logged area) to study. Because these groups did not live in close proximity to humans, I felt comfortable habituating them. Furthermore, because they all lived well within the protected park

2 boundaries, there was little concern about confounding factors such as exposure to endocrine disruptors and domestic animals.

Therefore, this dissertation examines four groups of red colobus monkeys living in either heavily logged or old-growth areas of Kibale National Park. What follows in this chapter is a discussion of habitat quality on non-human primate foods, behaviors, and reproduction and an introduction to the red colobus of Kibale National Park.

Background

Continued reproductive success is vital to the conservation of any species. However, anthropogenic activity can impact mammalian reproductive success through a number of means, including increased exposure to chemicals that act as endocrine disruptors, increased exposure to parasites, and decreased access to food. To better understand how anthropogenic deforestation and fragmentation affect reproductive success, changes in food availability, parasite load, energy balance, and other sources of stress must be evaluated. Additionally, baseline stress levels and reproductive function should be examined. Here, I review our current understanding of factors that influence reproductive success and discuss valid ways to analyze these factors in non-human primates. In particular, I will highlight our current knowledge of red colobus ( Procolobus rufomitratus ) reproduction.

Because primates have relatively long gestation periods and extended periods of lactation, interbirth intervals are relatively long and are difficult to document unless a group is studied for many consecutive years. Some non-human primates require decades of data before reproductive success can be determined. Thus, other methods for evaluating individual fitness and reproductive function are essential. Hormones are a valuable source for assessing both stress and reproductive function in mammals.

3 Stress-induced changes in hormone concentrations may lower a male mammal’s reproductive fitness. In both rats (Ge et al. 1997) and olive baboons (Sapolsky 1985), increased glucocorticoid concentrations were associated with a decrease in testosterone concentrations.

Testosterone and other androgens are essential for male sexual development and fertility

(reviewed in Dohle et al. 2003). However, few studies have found a correlation between stress and reduced male fertility, which suggests that male reproduction is not as affected by stress as it is for females.

Male reproductive success is limited by the number of mates, while female reproductive success is limited by access to resources. Therefore, females are more likely to be heavily impacted reproductively by the loss of resources associated with deforestation. In an evaluation of the reproductive ecology of old world monkeys, Bercovitch (2001) found that resource availability had a larger impact on female sexual maturation than on male sexual maturation. For example, testosterone concentrations in male chimpanzees at Kanyawara did not decrease in response to low food availability (Muller and Wrangham 2005). On the other hand, nutritional status is one major influence on female reproductive success. Female mammals need sufficient energy balance in order to conceive, maintain pregnancy, and support lactation. Higher reproductive success is influenced by the age of menarche, interbirth intervals, and breeding season length, all of which are affected by nutrition. For this reason, studies of decreased reproductive success within primate populations due to environmental stressors should focus on factors that impact female reproduction.

Here, I review the evidence linking deforestation to decreased reproductive fitness.

Specifically, I examine how food availability, parasite and disease exposure, toxicants in the environment, and stress hormones impact reproductive hormone concentrations and reproductive

4 success. Understanding these relationships is a vital step toward conserving species that are impacted by deforestation, and could provide important insight into understanding the current state of red colobus populations in Kibale National Park. Furthermore, this background will set the stage for examining adaptive responses of primates facing ecological stressors.

Habitat disturbance and food availability

Deforestation can significantly impact primate populations, especially because of changes in food availability. Deforestation often causes food stress for wild animal populations, and food stress is associated with longer inter-birth intervals and higher fetal mortality (Altmann and

Alberts 2003; Knott 1999; Emery Thompson et al. 2007). Reproductive hormones, leptin levels, and ketones have all been used to measure food stress and reproductive problems.

Deforestation is associated with lower food availability (Chapman et al. 2006; Felton et al. 2003; Roa & van Schaik 1997). For example, during a three year study of forest fragments in

Uganda, Chapman et al. (2006) found that, on average, logging activities led to a 29.5% decrease in the DBH (diameter at breast height) of available food trees. Orangutan preferred food trees were the most frequently logged species in Gulung Panung National Park, Indonesia (Felton et al. 2003). These logging activities lead to fewer and more dispersed trees (Felton et al. 2003).

Ultimately, this decrease in food trees was associated with a decline in the number of orangutans.

Surveys of night nests show that there were significantly fewer nests per km 2 in logged areas than in unlogged areas (Felton et al. 2003). Knott (1999) found that orangutan interbirth intervals were positively influenced by peaks in fruit abundance; therefore, the logging of large food trees could have serious consequences for female reproductive success.

More dispersed trees can also lead to an increase in energetic costs associated with logged forests. In a study of Sumatran orangutans, Rao and van Schaik (1997) found that

5 orangutans in the logged forest adopted more energetically expensive locomotor behavior and spent more time traveling. Females, especially ones with young, have the most difficulty traveling through the logged forests (Rao and van Schaik 1997).

A similar pattern has been seen in colobus habitats where logging occurs. Colobus living in forest fragments fed on mature leaves significantly more often than colobus living in the continuous forest (Chapman et al. 2004). Feeding on mature leaves rather than the young leaves, which are their preferred food), is one indication that deforestation is causing a decrease in preferred food sources. Additionally, Chapman et al. (2006) found that red colobus in fragments rested more than individuals in the continuous forest. This finding suggests that they try to reduce their energy expenditure because of a deficit in energy available for them to consume.

Habitat disturbance and behavioral changes

Behavioral differences associated with habitat quality have been documented in several species of non-human primates (baboons, Atlmann & Muruthi 1988; Eley et al. 1989; Iwamoto

& Dunbar 1983; bearded saki monkeys, Boyle & Smith 2010; Silva and Ferrari 2009; black and white colobus monkeys, Wong & Sicotte 2007; bonnet macaques, Singh & Vivanthe 1990; gorillas, Watts 1988; howler monkeys, Marsh 2003; ruffed lemurs, Ratsimbazafy 2007; and vervets, Isbell & Young 1993). Different species appear to use different strategies, but the overall pattern suggests that primates adjust energy input and output in response to habitat differences. For example, bearded saki monkeys and black and white colobus monkeys both show a pattern of spending more time resting and less time traveling in degraded areas (Boyle &

Smith 2010; Wong & Sicotte 2007). Ruffed lemurs also decrease their time spent traveling in degraded areas (Ratsimbazafy 2007). Increased time resting in degraded habitats may be a way to conserve energy. However, some primates show the opposite pattern of increased time spent

6 resting in high quality compared to low quality habitats (e.g.baboons (Iwamoto & Dunbar 1983) and gorillas (Watts 1988)). Spending more time resting in high quality habitats may be the result of having plenty of food and requiring less time spent foraging and feeding. In support of this, baboons (Altmann & Muruthi 1988; Eley et al. 1989; Iwamoto & Dunbar 1983), gorillas (Watts

1988), and bonnet macaques (Singh & Vivanthe 1990) all spent less time foraging and feeding in high quality areas.

These studies suggest a trade-off between time spent feeding and time spent on other activities. Increased time spent foraging is negatively correlated with resting (Menon & Poirier

1996) and social activities (Gillespie and Chapman 2001) and positively correlated with eating lower quality foods (Menon & Poirier 1996; Onderdonk & Chapman 2000). To continue to obtain enough energy, individuals in degraded habitats may need to spend more time getting the necessary food. They may also have to utilize different types of resources. Ruffed lemurs living in degraded habitats diversified their diet compared to those in better habitats (Ratsimbazafy

2007). Similarly, howler monkeys show differences in resting and foraging time that seem to allow them to succeed in heavily degraded areas (Marsh 2003, Rodriquez-Luna et al. 2003;

Silver & Marsh 2003; Estrada & Coates-Estrada 1996). Individuals in disturbed areas may also try to adjust their energy output by reducing any unnecessary behaviors. For example, bearded saki monkeys engaged in fewer social behaviors in disturbed habitats (Silva & Ferrari 2009).

These studies suggest that wild primate species may adjust their behaviors under variable ecological conditions.

It is important to note that some studies have not found any differences in activity budgets across different habitats (Chapman, Wasserman, & Gillespie 2006; Estrada et al. 1999;

Onderdonk and Chapman 2000). For example, Chapman, Wasserman, and Gillespie (2006)

7 found that differences in activity budgets (time spent feeding, traveling, and resting) of red colobus monkeys did not differ based on habitat type, but instead based on group size. For example, a large group in an unlogged area traveled more than both a small group in an unlogged area and a mid-sized group in a logged area. Similarly, the large group fed less than the small group in the unlogged area while the logged group was intermediate. The factors affecting differences between individuals living in logged vs. unlogged areas may have been clearer if the study groups were of similar size. In taxa as socially complex and diverse as non-human primates, ecological conditions will certainly not be the only determining factor for how individuals spend their time, and this must be taken into consideration when examining the impact of habitat quality on activity budgets. Inconsistencies in the relationship between ecology and activity budgets point to the importance of social structures and group dynamics across different primate populations.

Food stress and reproductive success

Food stress can negatively impact reproductive function. Nutritional status is one major influence on female reproductive success since reproduction requires additional energy (Wade and Schneider 1992). In a study of female golden lion tamarins ( Leontopithecus rosalia), reproductive females expended more energy than nonreproductive females during the reproductive period (Miller et al. 2006). Better nutrition increases reproductive fitness through earlier menarche, shorter interbirth intervals, and longer breeding seasons due to better nutrition

(Knott 1999; Tardif et al. 2005).

One way to measure nutritional status is through ketones. Ketones are produced by the body when nutritional intake is low and then excreted in the urine (McGarry & Foster 1980;

Robinson & Williamson 1980). Knott (1998) found that ketones were only present in orangutan

8 urine during severe fruit shortages. During the high fruit availability period and the beginning of the low fruit availability period, no ketones were found in orangutan urine samples (Knott 1999).

It appears, then, that fat metabolism occurs as a result of food stress (Knott 1999). Therefore, ketones could be used to determine if food stress is affecting individuals living in logged areas.

Ketones could be useful to show a correlation between lower food availability, nutritional stress, and population declines. In this way, ketones would add significantly to studies such as

Chapman et al. (2006) where anthropogenic fragmentation of red colobus habitats is associated with a decline in population size. Between 2000 and 2003, Chapman et al. (2006) found that the population of red colobus monkeys living in Kibale fragments decreased from 127 to 102. The ratio of infants to adult females decreased from 0.31 to 0.19, which suggests that these population decreases may be a result of reproductive suppression or higher rates of infant mortality. Additionally, the number of subadults diminished fourfold from 2000 to 2003. In this same timespan, the basal area of trees in the fragments decreased from 9,002 m 2 to 5,293 m 2, and the basal area of food trees decreased form 915 m 2 to 535 m 2. Thus, a decrease in food availability was correlated with a decrease in population size and infant numbers in Kibale red colobus monkeys.

Trivers (1972) and Wrangham (1979) both argue that food is the primary limitation on reproductive success in female mammals. Chronic energetic stress leads to decreased reproductive success over a lifetime via slower growth and maturation, longer interbirth intervals, and lower concentrations of ovarian hormones (Lipson 2001). The first mechanism for which food availability can impact reproductive success is conception. Sherry (2002) found that the conceptive periods of Kibale chimpanzees were correlated with fruit availability. Thus, in these chimpanzees, females need to have access to plenty of food during this critical period in

9 the reproductive cycle (Sherry 2002). A decrease in food availability could impair a female’s ability to conceive. In a study of chimpanzees in Kibale National Park, females living in areas of

“more abundant or higher-quality food” sources had “significantly higher levels of urinary oestrogen and progesterone conjugates” than females living in other areas (Emery Thompson et al. 2007). Consequently, these higher concentrations of reproductive hormones resulted in much shorter interbirth intervals and higher infant survival (Emery Thompson et al. 2007).

The next mechanism for which low food availability impacts reproductive success is pregnancy maintenance. Tardif et al. (2005) found that energy restrictions led to a rapid loss of placental and fetal function. Therefore, food stress makes it difficult or impossible for a female to carry her pregnancy to term. To have a successful pregnancy, females must have access to a sufficient level of “oxidizable metabolic fuel” (Wade and Schneider 1992). A decrease in reproductive success certainly seems like a logical reaction to depleted food resources since the body responds to stress by shutting down any unnecessary body systems (e.g. digestive and reproductive) (Sapolsky 1994). Similar results were found in studies of moose living in

Yellowstone National Park. A decline in moose population size was found to be the result of lower pregnancy rates (Berger et al. 1999). This problem is seen in populations living in predator-free habitats, which suggests that high levels of food competition restrict population size (Berger et al. 1999).

A third mechanism impacting reproductive success is interbirth intervals. Altmann and

Alberts (2003) found that when Amboseli baboons had more food available to them, they had a significant decrease in their interbirth interval, equating to about a 25% reduction for this population of baboons. Interbirth intervals are often shorter in captivity, as well, where animals are being provisioned with plenty of food.

10 The fourth way in which food availability can impact reproductive fitness is in growth and survival of offspring. Altmann and Alberts (2005) found that food availability was the primary source of body size variation in immature individuals in a wild population of savannah baboons ( Papio cynocephalus ). In comparisons of food-enhanced populations and non- provisioned populations, Altmann and Alberts (2005) found that these two different conditions resulted in significant differences in infant survival, growth rate, and maturation. Food limitations in the wild make it difficult, if not impossible, for small-for-age juvenile baboons to have compensatory growth spurts (Altmann and Alberts 2005). Thus, even if a female is able to carry her pregnancy to term, her reproductive fitness could still be negatively affected if her infant is not able to survive because of food stress.

Females can help regulate their energy balance by adjusting their food intake or adjusting their energy expenditure. For example, pregnant and lactating primates may decrease time foraging and increase resting time to meet their increased energy needs (Miller et al. 2006).

Wild golden lion tamarin females spend more time foraging during the nonreproductive period and more time being stationary in the reproductive period (Miller et al. 2006). Miller et al.

(2006) suggest that this reproductive strategy involves storing up extra energy before the reproductive period that can then be conserved and used over the reproductive period. However, in disturbed habitats, individuals are not able to make this adjustment as easily because they either do not have access to additional food resources to increase their energy intake, have to travel further distances (thus, expending extra energy) to reach their food resources, or a combination of both.

Ellison (1990) found that a positive energy balance (the relationship between nutritional intake and energy expenditure) has a significant impact on ovarian function. A slight increase in

11 body weight leads to higher concentrations of ovarian hormones, which leads to a higher rate of conception (Ellison 1990). This has also been seen in female orangutans where weight gain is associated with increased urinary hormone concentrations (Masters and Markham 1991).

Conversely, fluctuating energy balance is associated with suppressed ovarian function and decreased fecundity (Knott 2001).

Knott (1999) found that estrone conjugate values were significantly lower in wild than captive orangutans, particularly during periods of food shortages. This evidence suggests that differences in nutritional status result in changes in ovarian hormone concentrations. As previously discussed, these hormone concentrations are directly related to reproductive success.

Reproductive behavior also correlates with these findings. All matings in wild orangutans occur during periods of high fruit and flower availability (Knott 1999). There was a significant, positive relationship between energy balance in orangutans and the availability of ripe fruit

(Knott 1999). Plus, orangutan interbirth intervals were influenced by peaks in fruit abundance

(Knott 1999). Knott (1999) hypothesized that a reduction in fat stores may lead to lengthened interbirth intervals. Similarly, food stress in marmosets led to termination of the pregnancy

(Tardif et al. 2005). These findings suggest that low food availability in non-human primate habitats lead to decreased concentrations of ovarian hormones, longer interbirth intervals and ultimately lower reproductive fitness.

In humans, energy balance has a significant impact on ovarian hormones (Ellison et al.

1993). High estradiol concentrations were associated with a high success rate of implantation in in-vitro fertilization patients (Akman et al 2002). Similarly, fetal loss was associated with significantly lower fecal estrogen concentrations (compared to concentrations of successful pregnancies) during the trimester of loss (Beehner et al. 2006). Lenton et al. (1988) found that

12 elevated progesterone concentrations were also necessary for successful implantation rates.

Progesterone can also be impacted by changes in energy balance. An increase in energy expenditure was correlated with a decrease in progesterone concentrations in women who lost weight, but not in women who gained weight (Jasienska 2001). These studies show that energy intake (food availability) has an impact on ovarian hormones, which subsequently have an effect on reproductive success.

Similar results are seen in non-human primate studies. Strier and Ziegler (2005) found that there was a minimum estradiol threshold concentration that was needed to resume cycling postpartum in female northern muriquis ( Brachyteles hypoxanthus ). Estrogen also appears to be a very important hormone for maintaining pregnancy. In captive studies where estrogen was experimentally suppressed, 50% of the pregnancies ended in miscarriages (Albrecht et al. 2000).

Therefore, if ecological stressors cause estrogen to be low, it could significantly impact the reproductive success of female primates. Evidence from a diversity of primate species

(including humans) in both the Old and New World show the interconnectedness of deforestation, nutritional status, and reproduction..

Red colobus reproduction

Red colobus are a valuable case study for how ecological stressors impact non-human primate populations. Growing human populations have caused deforestation to infringe on wild primate habitats. It is uncertain how and if logged or fragmented areas can support healthy wild non-human primate populations. Red colobus are highly endangered (Struhsaker and Leland

1987), and understanding their reproduction is imperative to their successful conservation.

Red colobus at Kibale National Park, Uganda live in habitats of varying qualities. Kibale

Forest has designated undisturbed park land alongside forest fragments that have faced varying

13 degrees and types of anthropogenic activity (Onderdonk & Chapman 2000). Unlike other species of primate in the area that seem to be unaffected by the disturbance (i.e. black and white colobus), human activity has impacted red colobus populations (Gillespie and Chapman 2006).

Red colobus reproductive behavior, physiology, and hormone profiles have not been well studied. Assumptions about copulation behaviors and sexual swellings are based on a limited number of behavioral studies, but have not been validated with hormonal data. The red colobus of Kibale National Park do not appear to have a birth season, but do show a pattern of bimodal birth peaks that occur during the rainy months (Strusaker and Leland 1987). Struhsaker (1975), however, does not find a relationship between the months of mounting and the months of rainfall. He also finds that there is a lack of seasonality in food availability, which may explain the aseasonality of copulations. In other locations, such as Gambia, researchers have documented a strict birth season (Starin 1991). In addition, almost no copulations during the periods of lowest rainfall and temperatures (Starin 1991), suggesting a clearer mating season.

Though there is limited data on colobus interbirth intervals, a two year study of Colobus badius suggests that the interbirth interval is approximately 2 years (25.5 ± 5.1 months)

(Struhsaker and Leland 1987). Similarly, the interbirth interval for successful births was 29.4 months in Gambian red colobus (Starin 1991). In the closely related black and white colobus

(Colobus guereza ) of Kibale National Park, the median interbirth interval is 21.5 months (Harris and Monfort 2006).

The gestation length in red colobus is also not known, but Struhsaker (1975) suggests that it is probably between 4.5 and 5.5 months. These estimates are similar to the findings for C. guereza in Kibale National Park, which have a gestation length of approximately 158 days

(Harris and Monfort 2006). These findings for African colobines are much shorter than the

14 results for an Asian colobine species (Hanuman Langur), which has an average gestation length of 211.6 ± 3.4 days (Ziegler et al. 2000). The average cycle length is 27.7 (Starin 1991). This is similar to the 24 day cycle seen in C. guereza (Harris and Monfort 2006).

The age of red colobus sexual maturity is between 35.4 and 58 months for males and between 38 and 46 months for females in Kibale National Park (Struhsaker and Leland 1987). In

Gambia, the average age of female sexual maturity was 34.25 (n=4) (Starin 1991). Unlike in other colobines, female red colobus transfer from their natal groups (Strushaker and Leland

1979, 1987; Krostjens 2001). Red colobus groups are patrilinear and there is a loose hierarchy for the males (Struhsaker and Leland 1979, 1987; Korstjens 2001). Dominant males copulate most (at least 80%), and copulating pairs are regularly harassed by other adult males in the group

(Struhsaker and Leland 1979).

Female red colobus have sexual swellings that are assumed to be correlated with ovulation (Struhsaker and Leland 1987; Starin 1991). However, these sexual swellings have also been seen in pregnant females (Struhsaker and Leland 1985; Starin 1991). Because these swellings are never seen in females with newborn infants, Struhsaker (1975) argues that these swellings may be under estrogenic control. The size and color of these swellings differ between females, and it is the increase in size of the swelling (not the color) that Struhsaker (1975) argues is an indicator for ovulation. In addition to red colobus, there are only 2 other colobine species that develop sexual swellings (Newton and Dunbar 1994). Newton and Dunbar (1994) argue that red colobus are also one of the three colobines where females rarely solicit sex. This nearly unique pattern of sexual swellings and no female solicitation (olive colobus are the only other colobine species to also demonstrate these qualities [Newton and Dunbar 1994]) may suggest that paternity confusion is important in this species. Further studies comparing behavioral estrus

15 and hormone concentrations would help to clarify when females are ovulating and if they are mating selectively during this period.

While Newton and Dunbar (1994) argue that female red colobus do not solicit sex, other researchers have found that they do. The red colobus in Gambia performed audio-visual displays that Starin (1991) argues are “an advertisement for receptivity and an invitation to copulation.”

These displays could be used to attract males outside of ovulation, and thus, confuse paternity or they could actually correspond with ovulation. These audio-visual displays have not been seen in any other population of red colobus (Starin 1991), but new studies of the red colobus at Kibale may reveal additional mating behaviors that have not yet been recorded in this population.

Red colobus have been studied at multiple locations for many years. In this way, there is a wealth of information on many aspects of the red colobus monkey. However, much of the information we know for reproductive parameters is estimated from behavioral observations.

This dissertation combines behavioral observation with laboratory analyses of fecal and urine samples to provide a better understanding of both red colobus reproduction and their response to habitat quality.

Chapter descriptions

This dissertation consists of three semi-autonomous article-style chapters that are flanked by an Introduction (this chapter) and a Conclusion (Chapter 5). Each chapter consists of its own background section that provides more in-depth background for each topic than what is provided here. Additionally, each chapter has its own methods section, so there is not a separate methods chapter.

Chapter 2: Female red colobus monkeys maintain an ideal free distribution through novel foraging strategies in logged forests

16 In this chapter, I provide an overview of red colobus feeding ecology in logged and old- growth areas of Kibale National Park. I use published data on tree densities in different parts of the forest to show variation in forest structure and feeding tree availability. I also use published data on red colobus densities in each area of the forest to illustrate that red colobus densities do not differ between habitats. I present data on behavioral observations of what red colobus are eating in Kibale National Park and measurements on red colobus trees that were consumed during the study. I add to our understanding of the variation of red colobus foods in each area by combining density data with measurements of feeding trees and I show how red colobus in logged areas eat more diverse diets utilizing more tree species than red colobus in old-growth areas. I also show that red colobus in logged areas spend more time feeding and less time doing all other activities. I argue that red colobus are able to maintain similar densities in logged and unlogged areas because individuals in logged areas adjust their diet and behaviors to offset energetic costs of living in degraded habitats.

Chapter 3: Timing is everything: expanding the cost-of-sexual-attraction hypothesis

This chapter examines differences in the reproductive behaviors of female red colobus monkeys. Following the cost-of-sexual-attraction hypothesis proposed by Wrangham (2002), I compare the copulation rates, timing of mating behaviors, and length of genital swellings for females in logged and old-growth areas. I argue that females in logged areas are constrained in their sexual behaviors. They copulate less often and at specific times in their cycle much more so than females in old-growth areas who are more variable and less constrained in their reproductive behaviors. While this pattern allows females in logged areas to maintain reproductive function, it limits their abilities to use reproductive strategies that may ultimately be very important for reproductive success.

17 Chapter 4: Female red colobus monkeys maintain reproductive function even in degraded habitats

This chapter is dedicated to analyses of fecal and urine samples for female red colobus monkeys in logged and old-growth areas. Estradiol and progesterone are used as indicators of reproductive function. These hormones were extracted from fecal samples and analyzed with radioimmuoassays. Ketones are used as an indicator of nutritional stress. Ketones were measured in urine samples using a chemstrip. Females in logged and old-growth areas do not show significant differences in reproductive hormone concentrations or ketone levels. These results indicate that females in logged areas are not nutritionally stressed and are able to maintain their reproductive function. Additional studies on reproductive success are essential to understand if reproductive hormone concentrations can be used as a proximate measure of reproductive fitness.

Chapter 5: Conclusion

In this chapter, I summarize my findings and discuss the relevance of my research to the fields of biological anthropology and conservation biology. I explain how the results of this study can help determine proper conservation management plans and provide insight into how females adjust to changing environments throughout evolution. I also suggest avenues for future research.

18 Bibliography

Akman MA, Erden HF, Bener F, Liu JE, and Bahceci M. 2002. Can luteal phase estradiol levels predict the pregnancy outcome in in vitro fertilization cycles of good responders whose excess embryos yield blastocysts? Fertility and Sterility 77(3): 638-639.

Albrecht ED, Aberdeen GW, and Pepe GJ. 2000. The role of estrogen in the maintenance of primate pregnancy. American Journal Obstetrics and Gynecology 182(2): 432-438.

Altmann J and Alberts SC. 2003. Variability in reproductive success viewed from a life-history perspective in baboons. American Journal of Human Biology 15:401-409.

Altmann J and Muruthi P. 1988. Differences in daily life between semi-provisioned and wild- feeding baboons. American Journal of Primatology 15: 213–221.

Altmann J, Ward Lynch J, Nguyen N, Alberts SC, and Gesquiere LR. 2005. Life-history correlates of steroid concentrations in wild peripartum baboons. American Journal of Primatology 64: 95-106.

Bales KL, French JA, Hostetler CM, and Dietz JM. 2005. Social and reproductive factors affecting cortisol levels in wild female golden lion tamarins (Leontopithecus rosalia ). American Journal of Primatology 67: 25-35.

Beehner JC, Nguyen N, Wango EO, Alberts SC, Altmann J. 2006. The endocrinology of pregnancy and fetal loss in wild baboons. Hormones and Behavior 49: 688-699.

Bercovitch FB. 2001. Reproductive ecology of old world monkeys. In Ellison PT (Ed) Reproductive Ecology and Human Evolution (New York, Aldine de Gruyter):369-396.

Berger J, Ward Testa J, Roffe T, Monfort SL. 1999. Conservation Endocrinology: a noninvasive tool to understand relationships between carnivore colonization and ecological carrying capacity. Conservation Biology 13 (5): 980-989.

Boyle, S.A., Smith, A.T., 2010. Behavioral modifications in northern bearded saki monkeys (Chiropotes satanas chiropotes) in forest fragments of central Amazonia. Primates 51, 43–51.

Chapman CA, Chapman LJ, Naughton-Treves L, Lawes MJ, and McDowell LR. 2004. Predicting folivorous primate abundance: validation of a nutritional model. American Journal of Primatology 62: 55-59.

Chapman CA, Wasserman MD, Gillespie TR, Speirs ML, Lawes MJ, Saj TL, and Ziegler TE. 2006. Do food availability, parasitism, and stress have synergistic effects on red colobus populations living in forest fragments? American Journal of Physical Anthropology.

19 Coop RL and Holmes PH. 1996. Nutritional and parasite interaction. International Journal of Parasitology 26: 951-962.

Cordeiro Sousa MB, Sales Da Rosha Albuquerque AC, Da Silva Albuquerque F, Araujo A, Emilia Yamamoto M, and De Fatima Arruda M. 2005. Behavioral strategies and hormonal profiles of dominant and subordinate common marmoset ( Callithrix jacchus ) females in wild monogamous groups. American Journal of Primatology 67: 37-50.

Daszak P, Cunningham AA, Hyatt AD. 2000. Emerging infectious diseases in wildlife—threats to biodiversity and human health. Science 287: 443-449.

Eley RM, Strum SC, Muchemi G, and Reid GDF. 1989. Nutrition, body condition, activity patterns and parasitism of free-ranging baboons ( Papio anubis ) in Kenya. American Journal of Primatology 18: 209–219.

Ellison PT. 1990. Human ovarian function and reproductive ecology: New hypotheses. American Anthropologist 92 (4): 933-952.

Emery Thompson M. 2005. Reproductive endocrinology of wild female chimpanzees ( Pan troglodytes schweinfurthii ): Methodological considerations and the role of hormones in sex and conception. American Journal of Primatology 67: 137-158.

Emery Thompson M, Kahlenberg SM, Gilby IC, and Wrangham RW. 2007. Core area quality is associated with variance in reproductive success among female chimpanzees at Kibale National Park. Animal Behaviour 73:501-512.

Estrada A., Anzures A.D. and Coates-Estrada R. 1999. Tropical rain forest fragmentation, howler monkeys ( Alouatta palliata ) and dung beetles at Los Tuxtlas, Mexico. American Journal of Primatology 48: 253–262.

Felton AM, Engstrom LM, Felton A, and Knott CD. 2003. Orangutan population density, forest structure and fruit availability in hand-logged and unlogged peat swamp forests in West Kalimantan, Indonesia. Biological Conservation 114: 91-101.

Foley, C.A.H., Papageorge, S. and Wasser, S.K. 2001. Non-invasive stress and reproductive measures of social and ecological pressures in free-ranging African elephants ( Loxodonta african a). Conservation Biology 15: 1134-1142.

Ge, R.S., Hardy, D.O., Catterall, J.F., and Hardy, M.P. 1997. Developmental changes in glucocorticoid receptor and 11-hydroxysteroid dehydrogenase oxidative and reductive activities in rat Leydig cells. Endocrinology 138: 5089-5095.

Gesquiere LR, Altmann J, Khan MZ, Couret J, Yu JC, Endres CS, Alberts SC, and Wango EO. 2005. Coming of age: Steroid hormones of wild immature baboons ( Papio cynocephalus ). American Journal of Primatology 67: 83-100.

20 Gillespie TR. 2006. Non-invasive assessment of gastro-intestinal parasite infections in free- ranging primates. International Journal of Primatology 27:1129-1143.

Gillespie TR and Chapman CA. 2001. Determinants of group size in the red colobus monkey (Procolobus badius ): an evaluation of the generality of the ecological-constraints model. Behavioral Ecology and Sociobiology 50: 329-338.

Gillespie, T.R. and Chapman, C.A. (2006). Prediction of parasite infection dynamics in primate metapopulations based on attributes of forest fragmentation. Conservation Biology 20 (2), 441–448.

Gillespie TR, Chapman CA, and Goldberg TL. 2005. Primates and the Ecology of their Infectious Diseases: How will Anthropogenic Change Affect Host-Parasite Interactions? Evolutionary Anthropology 14: 134-144.

Gregory RD and Hudson PJ. 2000. Population biology: parasites take control. Nature 406: 33- 34.

Harris TR and Monfort SL. 2006. Mating behavior and endocrine profiles of wild black and white colobus monkeys ( Colobus guereza ): Toward an understanding of their life history and mating system. American Journal of Primatology : 68: 383-396.

Hayes TB. 2005. Welcome to the revolution: Integrative biology and assessing the impact of endocrine disruptors on environmental public health. Integrative and Comparative Biology 45:321-329.

Huffman MA. 1997. Current evidence for self-medication in primates: A multidisciplinary perspective. Yearbook of Physical Anthropology 46: 171-200.

Isbell, L.A. & Young, T.P. ( 1 9 9 3 ) . Human presence reduces leopard predation in a population of vervet monkeys Cercopithecus aethiops. Anon. Behav ., 45, 1233-5.

Iwamoto T & Dunbar RIM. 1983. Thermoregulation, habitat quality and the behavioural ecology of gelada baboons. Journal of Animal Ecology 52: 357-366.

Jasienska G. 2001. Why energy expenditure causes reproductive suppression in women: An evolutionary and bioenergetic perspective. In Ellison PT (Ed) Reproductive Ecology and Human Evolution (New York, Aldine de Gruyter):59-84.

Kemnitz JW, Sapolsky RM, Altmann J, Muruthi P, Mott GE, Stefanick ML. 2002. Effects of food availability on serum insulin and lipid concentrations in free-ranging baboons. American Journal of Primatology 57: 13-19.

Knott, CD. 1999. Reproductive, physiological and behavioral responses of orangutans in Borneo to fluctuations in food availability. PhD dissertation. Harvard University.

21 Knott C. 2001. Female reproductive ecology of the apes: Implications for human evolution. In Ellison PT (Ed) Reproductive Ecology and Human Evolution (New York, Aldine de Gruyter): 429-463.

Knott CD. 2005. Radioimmunoassay of estrone conjugates form urine dried on filter paper. American Journal of Primatology 67: 121-135.

Korjsten AH. 2001. The mob, the secret sorority, and the phantoms: an analysis of the socio- ecological strategies of the three colobines at Tai. PhD dissertation. University of Utrecht, Utrecht.

Lenton EA, Gelsthorp CH, and Harper R. 1988. Measurement of progesterone in saliva: assessment of the normal fertile range using spontaneous conception cycles. Clinical Endocrionology 28(6): 637-646.

Lipson SF. 2001. Metabolism, Maturation, and Ovarian Function. In Ellison PT (Ed) Reproductive Ecology and Human Evolution (New York, Aldine de Gruyter):235-248.

Marsh, L. K. 2003. Primates in fragments: ecology and conservation . Kluwer Academic/Plenum Publishers, New York.

Masters AM and Markham RJ. 1991. Assessing reproductive status in Orang-utans by using urinary estrone. Zoo Biology 10: 197-207.

McGarry JD and Foster DW. 1980. Regulation of hepatic fatty acid oxidation and ketone body production. Annual Reviews of Biochemistry 49: 395-420.

Menon S, Poirier FE. 1996. Lion-tailed macaques ( Macaca silenus ) in a disturbed forest fragment: activity patterns and time budget. International Journal Primatology 17:969– 98.

Miller KE, Bales KL, Ramos JH, and Dietz JM. 2006. Energy intake, energy expenditure, and reproductive costs of female wild golden lion tamarins ( Leontopithecus rosalia ). American Journal of Primatology 68: 1037-1058.

Muller MN and Wrangham RW. 2005. Testosterone and energetics in wild cimpanzees ( Pan troglodytes schweinfurthii ). American Journal of Primatology 66: 119-130.

Newton PN and Dunbar RIM. 1994. Colobine monkey Society. In: Davies AG and Oates JF (eds). Colobine Monkeys: Their Ecology, Behavior, and Evolution . Cambridge University Press, Cambridge: 311-346.

Onderdonk DA and Chapman CA. 2000. Coping with forest fragmentation: The primates of Kibale National Park, Uganda. International Journal of Primatology 21: 587-611.

22 Pedersen AB, Altizer S, Poss M, Cunningham AA, and Nunn CL. 2005. Patterns of host specificity and transmission among parasites of wild primates. International Journal for Parasitology 35: 647-657.

Pie MR, Rosengaus RB, and Traniello JFA. 2004. Nest architecture, activity pattern, worker density and the dynamics of disease transmission in social insects. Journal of Theoretical Biology 226: 45-51.

Ratsimbazafy J. 2007. Diet Composition, foraging, and feeding behavior in relation to habitat disturbance: implications for the adaptability of ruffed lemurs ( Varecia v. editorium ) in Manombo forest, Madagascar. In Gould L and Sauther ML (eds) Lemurs: Ecology and Adaptation . Springer, New York: 403-422.

Roa M and van Schaik CP. 1997. The beharioral ecology of Sumatran orangutans in logged and unlogged forest. Tropical Biodiversity 4(2): 173-185.

Robinson AM and Williamson DH. 1980. Physiological roles of ketone bodies as substrates and signals in mammalian tissues. Physiological Reviews 60 (1): 143-187.

Sapolsky, RM. 1985. Stress-induced suppression of testicular function in the wild baboon: role of glucocorticoids. Endocrinology 116: 2273-2278.

Sapolsky, R.M. 1994. Why zebras don’t get ulcers: A guide to stress, stress-related diseases, and coping. W.H. Freeman, New York.

Sherry DS. 2002. Reproductive seasonality in chimpanzees and humans: Ultimate and proximate factors. PhD dissertation. Harvard University.

Silva, S. S. B., & Ferrari, S. F. (2009). Behavior patterns of southern bearded sakis Chiropotes satanas) in the fragmented landscape of eastern Brazilian Amazonia. American Journal of Primatology , 71, 1–7.

Silver SC and Marsh LK. 2003. Dietary flexibility, behavioral plasticity, and survival in fragments: Lessons from translocated howlers. In Marsh LK (ed.) Primates in fragments: ecology and conservation . Kluwer Academic/Plenum Publishing, New York: 251-265.

Singh M and Vinathe S. 1990. Inter-population differences in the time budgets of bonnet monkeys ( Macaca radiata ). Primates 31(4): 589-596.

Starin ED. 1991. Socioecology of the red colobus monkey in the Gambia with particular reference to female-male differences and transfer patters. PhD thesis. City University of New York, New York.

Strier KB and Ziegler TE. 2005. Variation in the resumption of cycling and conception by fecal androgen and estradiol levels in female northern muriquis ( Brachyteles hypoxanthus ). American Journal of Primatology 67: 69-81.

23

Struhsaker TT. 1975. The Red Colobus Monkey . University of Chicago Press, Chicago.

Struhsaker TT and Leland L. 1979. Socioecology of five sympatric monkey species in the Kibale Forest, Uganda. Advances in the Study of Behavior 9: 159-228.

Struhsaker TT and Leland L. 1987. Colobines: infanticide by adult males. In: Smuts BB, Cheyney DL, Syefarth RM, Wrangham RW, and Struhsaker TT (eds). Primate Societies . University of Chicago Press, Chicago: 83-97.

Tardif SD, Ziegler TE, Power M, and Layne DG. 2005. Endocrine changes in full-term pregnancies and pregnancy loss due to energy restriction in the common marmoset (Callithrix jacchus ). The Journal of Clinical Endocrinology and Metabolism 90(1): 335- 339.

Wade GN and Schneider JE. 1992. Metabolic fuels and reproduction in female mammals. Neuroscience and Behavioral Reviews 16: 235-272.

Wasser, S.K., K. Bevis, G. King, E. Hanson. 1997. Noninvasive physiological measures of disturbance in the Northern Spotted Owl. Conservation Biology 11: 1019-1022.

Watts DP. 1988. Environmental influences on mountain gorilla time budgets. American Journal of Primatology 15: 295–312.

Wrangham, RW. 2002. The cost of sexual attraction: is there a trade-off in female Pan between sex appeal and received coercion? In Boesch C, Hohmann G, and Marchant LF (eds.) Behavioural Diversity in Chimpanzees and Bonobos. University Press, Cambridge: 204- 215.

Wildt DE and Wemmer C. 1999. Sex and wildlife: the role of reproductive science in conservation. Biodiversity and Conservation 8: 956-976.

Wong S. N. P. and Sicotte P. (2007). Activity budget and ranging patterns of Colobus vellerosus in forest fragments in Central Ghana. Folia Primatol. 78: 245–254.

Ziegler T, Hodges K, Winkler P, and Heistermann M. 2000. Hormonal correlates of reproductive seasonality in wild female Hanuman langurs (P resbytis entellus ). American Journal of Primatology 51:119-134.

Ziegler TE and Wittwer DJ. 2005. Fecal steroid research in the filed and laboratory: Improved methods for storage, transport, processing, and analysis. American Journal of Primatology 67: 159-174.

24 CHAPTER 2

FEMALE RED COLOBUS MONKEYS MAINTAIN AN IDEAL FREE DISTRIBUTION THROUGH NOVEL FORAGING STRATEGIES IN LOGGED FORESTS

Abstract

Logging activities can have very long lasting effects on both forest structure and primate populations. Logging within Kibale National Park, Uganda, is thought to have led to reduced primate food availability initially and up to the present. Following the predictions of the ideal free distribution theory, primate densities are expected to decrease in areas of lower resource availability so that the available resources per individual would be equal in logged and old- growth areas. However, red colobus monkeys ( Procolobus rufomitratus ) occur at similar densities in both logged and old-growth areas of Kibale National Park. This suggests that either the ecological differences between the two areas are not sufficient to impact population densities of red colobus or that red colobus are using different foraging strategies in heavily logged areas to maintain their densities. To test this, I examined four groups of red colobus monkeys, two in logged and two in old-growth areas of Kibale National Park, Uganda, and compared their feeding behavior as well as feeding tree size and productivity in both conditions. I predicted that females in logged areas consume a wider diversity of plant species, have fewer feeding bouts on each food species, and spend a higher percentage of their activity budget on feeding than females in old-growth areas. Furthermore, I predicted that food trees would be smaller and produce fewer food items than in old-growth areas. Individuals in logged areas ate a greater number of plant species and spent more time feeding than females in old-growth areas. Changes in food availability likely account for some of these differences as some common red colobus foods were targeted as high value timber species during logging. By expanding their diet, females in logged areas effectively increased the resources available to them, contributing to their ability to

25 maintain similar densities to females in the old-growth areas. Information on dietary flexibility in response to disturbance can help guide conservation of endangered red colobus monkeys that are declining due to habitat destruction.

Introduction

Primate diets are often highly variable in terms of the resources consumed (Garber 1987).

Environmental changes, such as deforestation or climate change, produce even greater variability in primate diets (Chapman & Chapman 1990). This has important implications for applying traditional foraging theories to primate behavioral ecology (Chapman 1988; Garber 1987; Janson

& Chapman 1999). For example, optimal foraging theory was originally proposed by

MacArthur and Pianka (1966) to explain the energy that predators gain from capturing prey compared to the energy that is expended in acquiring prey. This theory has been extrapolated to discuss efficient foraging strategies in a variety of animals. However, because primates have such variable diets and are rarely dietary specialists (i.e., one food type makes up the majority of the diet), feeding strategies are more complicated than typically considered by the optimal foraging theory (Garber 1987). For example, capuchin monkeys ( Cebus olivaceus ) eat plant materials (leaves, roots, and fruits), invertebrates, and vertebrates and consume each of these items in different amounts during different seasons (Robinson 1986). Thus, explaining such dietary type and seasonal differences through an optimal foraging perspective is challenging.

Extremely variable primate diets also make it difficult to understand the spatial distributions of populations using classic ecological models. For example, the ideal free distribution theory was proposed by Fretwell and Lucas (1969) to describe how bird populations distribute themselves among feeding patches and specifically predicts that the number of individuals in an area will be proportional to the foraging gains. Under the ideal free distribution

26 model, a decrease in resources (foraging gains) should lead to a decrease in consumer numbers

(Pulliam & Caraco 1984). However, this prediction is not always upheld (Cowlishaw 1997) and other factors, such as predators, territoriality, and disease, can impact the distribution of consumers. Furthermore, since primates can adopt variable diets, individuals may adopt new foraging strategies in areas with poor resource availability allowing them to obtain consistent foraging gains by alternative means. For example, Nakagawa (1990) argued that Japanese monkeys sometimes fed in less crowded patches of lower quality food rather than on higher quality foods in crowded patches because the net benefit for these two may be equal.

Pigliucci (2001) argues that phenotypic and behavioral plasticity may be an adaptive strategy in all habitats, but especially in a stressed environment, such as one that has experienced anthropogenic disturbance. Such plasticity might be the means by which animal populations are able to achieve an ideal free distribution. For example, it has frequently been suggested that members of the same primate species living in different forest types physiologically and behaviorally adjust to regulate their energy balance in response to their environment (Altmann &

Muruthi 1988 (baboons); Iwamoto & Dunbar 1983 (geladas); Boyle & Smith 2010 (sakis); Silva and Ferrari 2009 (sakis); Wong & Sicotte 2007 (colobus); Singh & Vivanthe 1990 (bonnet macaques); Watts 1988 (mountain gorillas); Ratsimbazafy 2007 (lemurs); and Wrangham 2002

(chimpanzees)), which appears to correspond to differences in their distribution. Dietary flexibility and behavioral plasticity are thought to be particularly important for surviving in degraded habitats (Chapman et al. in press ; Marsh 2003).

Ecological pressures from living in different quality habitats, including variation in tree density, resource availability, temperature, and moisture, can lead to intraspecific and within- population variation in diet composition, activity budgets, and group size. These differences can

27 ultimately lead to variation in social structure (reviewed in Chapman & Rothman 2009). For example, having larger groups or smaller groups may be related to increased predation risk or reduced food availability, respectively. It is still unclear what impact the environment vs. phylogenetic inertia has on dietary, behavioral, and social structure variation in different groups of a single primate population. Examining between-group behavioral variation compared to the overall expected norm for a population is important for understanding behavioral plasticity

(Komers 1997) and on small spatial scales, phylogeny is likely not an important issue, allowing an evaluation of how ecological factors influence animals.

Chapman et al. (2010) have shown a significantly lower cumulative diameter at breast height (DBH) of red colobus ( Procolobus rufomitratus ) food trees in heavily logged areas of

Kibale National Park, Uganda (8,000 cm) compared to old-growth areas of the park (11,000 cm).

Rather than the open areas of the logged areas regenerating with young trees, the area has been colonized primarily by a particular shrub ( Acanthus pubescens ) that directly competes with regenerating trees for space and attracts elephants that subsequently destroy young saplings

(Lawes & Chapman 2006; Paul et al. 2004). These differences in DBH correlate with differences in crown size (Catchpole & Wheeler 1992; Enquist et al 1998; Enquist & Niklas

2001) and more specifically food availability (leaf biomass) for red colobus monkeys (Chapman et al 1992; Snaith & Chapman 2008). Thus, red colobus in heavily logged areas are expected to have lower resource availability compared to red colobus in old-growth areas of Kibale.

Based on the ideal free distribution, population densities of red colobus should be lower in heavily logged areas. Instead, there are no significant differences in red colobus densities between heavily logged and old-growth areas of the park (Chapman et al 2010; Struhsaker 1997;

Struhsaker 2010). This is the conclusion reached by two researchers with over 50 years of

28 experience on primate densities in Kibale (Chapman, personal communication). Furthermore,

Chapman et al. (2010) reported on over 25 years of census data including 238 transects spanning

1104 km of Kibale that were repeatedly measured over the years, often by the same observers, and did not find density differences between the different forestry compartments. These findings are not consistent with the ideal free distribution model. This suggests that either the ecological differences between the two areas are not sufficient to affect population densities of red colobus, or that red colobus use different strategies in heavily logged areas to maintain their densities.

Here, I examine dietary and behavioral differences in female red colobus living in different areas of Kibale National Park, Uganda, and test if females in heavily logged areas have a plastic behavioral response including having different diet compositions and different activity budgets.

These changes may allow them to maintain the positive energy balance necessary for reproduction equivalent to females in old-growth areas. Under this model, individuals in degraded environments should show behavioral plasticity in feeding behaviors and should expand their diet to include a greater variety of foods than usually seen for red colobus, which would effectively increase their resource availability and thus be able to support an equivalent biomass of individuals under the ideal free distribution. Specifically, I contrast two red colobus groups in logged and two in old-growth forest to test whether females 1) eat the same plant species and plant parts, 2) eat these foods at the same frequency, and 3) spend different amounts of time feeding. Females were predicted to eat a more diverse diet in terms of plant species in the logged areas compared to the old-growth areas. As the result of having a more diverse diet, females in logged areas were expected to eat smaller quantities of common red colobus foods.

Additionally, females in the logged areas were expected to spend more time feeding than females in old-growth areas. Alternatively, if the reason population densities remain similar between

29 logged and old-growth areas is because there is not significant enough ecological differences between the two habitats, I expect to see either female red colobus in the logged areas utilizing similar strategies to acquire similar resources or variation across all groups rather than differences that fall specifically between logged and old-growth areas, but not within. The same species of trees are available in both areas, so if logging was not extensive enough to reduce food availability then females in logged areas should eat the same diet as in the old-growth areas. By expanding on the ideal free distribution model to include behavioral plasticity, females are expected to diversify their diet in areas where common resources are low, which would effectively increase the availability of resources, and result in groups of equivalent size

(biomass) in both logged and old-growth forests.

Methods

Study site

Research was conducted in the forest near Makerere University Biological Field Station

(MUBFS) in Kibale National Park, Uganda (hereafter Kibale; 795 km 2; 0°13'- 0°41'N, 30°19'-

30°32'E). Kibale is a mid-altitude, moist evergreen forest that receives 1696 mm of rainfall annually (1990-2011; Chapman unpublished data; Chapman and Lambert 2000; Stampone et al.

2011). Kibale provides a valuable setting for examining the impact of habitat disturbance on foraging strategies in primate, since both the forest and the primate populations have been studied for over 40 years. Forestry compartment K-30 is old-growth forest that was never commercially logged (Struhsaker 2010, Chapman et al. 2010), while K-15 was heavily logged and 50% of the trees were cut or indirectly killed (Kasenene 1987; Skorupa 1988; Struhsaker

1997; Chapman & Chapman 1997; Chapman and Chapman 2004; Figure 1). The K-13 area was heavily logged and poisoned with arboricide (Oates 1999). These previously logged areas are

30 still highly degraded and have not recovered (Skorupa 1988; Gebo & Chapman 1995). Logging activities have resulted in large gaps in the canopy that persist today. Four tree species

(Aningeria altissima, Lovoa swynnertonii, Strombosia scheffleri, and Parinari excelsa ) consisting of approximately 25% of red colobus feeding time in old-growth forest at Kibale

(Struhsaker 1978, 1997), are also considered high value timber and were heavily targeted during logging operations.

Study animals

Red colobus ( Procolobus rufomitratus ) are endangered folivorous primates that have been studied at Kibale for over 40 years (Struhsaker 1997; Chapman et al. 2000, 2005,

2006a&b). They live in multi-male, multi-female groups where females disperse out of their natal groups. Reproductive demographics for red colobus can be found in Table 1. One advantage to using red colobus as a model for this study is the detailed information already documented on how environmental stressors can impact wild red colobus populations. For example, red colobus living in forest fragments outside Kibale National Park have experienced an 83% population decline from 2000 to 2010 (N=11 fragments, mean size= 3.7 ha), primarily as a result of forest clearing (Chapman et al. 2006b, Chapman et al. in press ). Such a population decline clearly indicates that this population is stressed and provides valuable information on how a stressed population will behave and what its diet would be like. In these fragments colobus (N=19 groups) ate more mature leaves than colobus in old-growth habitats (N=2 groups)

(14% mature leaves and 65% young leaves in fragments vs. 4-6% mature leaves and 78- 84% young leaves in old-growth areas; Chapman et al. 2004). The increased reliance on mature leaves, considered a fallback food (Davies 1994), suggest that red colobus living in anthropogenically disturbed forest experience a decline in food availability. Females may be

31 particularly susceptible to habitat changes, and resultant changes in food availability as indicated by the fact that in forest fragments (N=11) around Kibale, the ratio of red colobus infants to adult females decreased from 0.31 to 0.19 from 2000 to 2003 (Chapman et al. 2006a). Hunting is not a significant cause of this decline as the local human population does not eat red colobus, and chimpanzees are rarely, if ever, seen in the forest fragments. Nevertheless, red colobus in forest fragments show signs of physiological stress with the individuals in these areas having increased cortisol concentrations (Chapman et al. 2006). The lowest cortisol concentration for individuals in the fragments was still higher than the highest concentration for individuals in the continuous forest). This information allows us to make predictions on how the red colobus in logged forest should respond if they are food limited, and if not,whether they maintain an ideal free distribution by other means.

Research design

Four groups of red colobus were simultaneously followed and observed; 2 groups ranged in old-growth areas of K-30 and two groups ranged in the heavily logged areas of K-13 and K-

15. These groups had very similar infant to adult female ratios (Table 2.) 0.45 logged group 1,

0.43 logged group 2, 0.47 old-growth group 1, 0.59 old-growth group 2), further supporting the published data on similarities in red colobus densities across habitat types. The second old- growth group was the largest of all of the research groups, which may account for the higher infant to adult female ratio. Each focal group was followed from 8:00 am to 4:00 pm six days a week. Prior to the start of data collection, I noted which females appeared to have infants or juveniles and then selected females from that subset that were no longer lactating to be focal females. Five focal females (adult, parous females without infants) were selected in each group and observed on a rotating basis until they gave birth, at which time a new adult female without

32 an infant entered the rotation, so that at each point in time, data were obtained from 5 females in each group. Sample sizes of focal females differs slightly among groups because additional females were added when prior focal females gave birth. Females were identified by distinguishing characteristics, such as kinks in their tails, differences in coloration, or scars. On the rare occasion that a focal female was out of sight for longer than one hour (N=29 times), another focal female was followed for the remainder of the day. Instantaneous focal animal behavioral samples were collected on the focal female every ten minutes. The female’s activities

(feeding, traveling, resting, mating, and grooming), the identity and distance of her nearest neighbor, the activity of the nearest neighbor, and any other individuals within a 5 m proximity were recorded.

Feeding observations included the plant species and plant part (e.g., young leaf, leaf petiole) consumed. Each day, two feeding trees were selected and measured for each group: the first tree the focal female was seen feeding in the morning and the second tree that she was seen feeding in after noon were tagged as feeding trees and scored. A visual estimate of the percentage (0, 25, 50, or 100%) of fruit, young leaves, mature leaves, and flowers was recorded for each food tree. Diameter at breast height (DBH) and crown size (visual estimates of length, width, and height) of these trees were also recorded.

To get a clearer estimate of the biomass of food for red colobus in each area, we created a proxy for biomass by multiplying the average crown size of specific species by the density of trees. To get a specific biomass for each type of food, we also multiplied this number by the percent of the crown that is made up of any particular food type (e.g. young leaves). Thus, I combined the crown volume and young leaf production of the top ten food trees measured with data on densities from Chapman (2002).

33 All statistical analyses were first checked with a Levene’s Test for Equality of Variances

(Levene 1960) and further statistical tests were chosen based on these results (see below).

Levels of significance were set at 0.05. SPSS was used to analyze feeding tree and feeding observation results. Plant species and parts consumed by each group, as well as the two daily feeding trees for each group were compared using a univariate ANOVA. DBH, crown size, and food scores of measured food trees were compared using an independent sample T-test. Activity budgets were calculated for each female by dividing the number of observations of each behavior

(e.g. feeding, traveling, resting, or grooming) by the total number of observations of all behaviours for that female. Data for female behaviors in logged vs. old-growth areas were of unequal variances so they were compared in SAS with Wilcoxon Rank-Sum tests.

Results

Randomly selected feeding tree comparisons

Data from the two feeding trees selected per group per day resulted in 51 different plant species across all groups (Old-growth Group 1 = 20 species, Old-growth Group 2 = 24, Logged

Group 1 = 29, Logged Group 2 = 30). The most common species that were tagged as feeding trees among all four groups were: Albizia grandibractata (4.7% in old-growth, 8.5% in logged) ,

Bosqueia phoberos (9.5% in old-growth, 0% in logged) , Celtis africana (9.9% in old-growth,

3.6% in logged) , Celtis durandii (6.9% in old-growth, 5.2% in logged) , Dombeya mukoke (9.9% in old-growth, 0.7% in logged) , Funtumia latifolia (3.5% in old-growth, 11.5% in logged) ,

Markhamia lutea (11.6% in old-growth, 3.9% in logged) , Milletia dura (8.2% in old-growth,

7.9% in logged) , Prunus africana (0.4% in old-growth, 5.2% in logged) , and Strombosia scheffleri (14.2% in old-growth, 8.9% in logged) (Figure 2). These constituted 73-88% of the

34 feeding trees for groups living in the old-growth area, but only 37-48% of the feeding trees of those in the logged area (Table 3).

Individuals in the logged area ate a more diverse diet (29-30 species vs. 20-24 in old- growth areas), thus, each species made up a smaller percentage of their overall diet (Figure 2).

Each feeding tree species made up 12% or less of the feeding trees selected by individuals in logged areas; whereas, each tree species made up 16-20% in the old-growth area. Based on recorded feeding trees for females, the dietary composition differed significantly between logged and old-growth areas (F=4.695; df=12; p<0.001).

Feeding tree measurements

Strombosia was the only feeding tree that was significantly smaller (based on DBH) in the logged than old-growth areas(Figure 3a) (t=-3.430, df=53, n logged =24, n old-growth =31, p=0.001, xlogged =38.5104, x old-growth =68.9113, stdev logged =22.00241, stdev old-growth =38.81331). Strombosia also had a smaller crown area (length + heigth +width) in the logged areas (Figure 3b) (t=-4.302, df=48.799, p<0.000, x logged =27.52, x old-growth =38.84, stdev logged =6.923, stdev old-growth =12.354).

Albizia was the only tree species that was significantly larger in DBH (t=2.567, df=34, p=0.015) and crown size (t=3.39, df=35, p=0.002) in the logged areas (N=26, x=47, stdev=22.814 and x=33.92, stdev = 12.076, respectively) compared to the old-growth areas (N=10, x=27.5, stdev=11.336 and x=19.91, stdev=9.762, respectively). These data only explain the differences between feeding trees that have been selected by red colobus, not the overall differences in food availability between different habitats in Kibale. However, previous studies (Chapman et al

2010; Snaith and Chapman 2008; Struhsaker 1997) provide evidence of the overall differences in tree densities and food availability between logged and old-growth areas.

35 There were no differences in the amount of flowers or young leaves present on the colobus selected feeding trees (n logged =132, n old-growth =107, sample sizes are the same for all feeding tree analyses; flowers: t=1.827, df=236.482, p=0.069, x logged =7.01, x old-growth =3.27, stdev logged =17.829, stdev old-growth =13.772; young leaves: t=-0.505, df=206.018, p=0.614, xlogged =41.86, x old-growth =43.22, stdev logged =18.661, stdev old-growth =22.405). There were significantly more fruits and mature leaves on trees in the logged vs old-growth areas (fruits: t=3.788, df=200.179, p<0.000, x logged =7.01, x old-growth =1.64, stdev logged =14.261, stdev old-growth

=7.097; mature leaves: t=2.053, df=211.622, p=0.041, x logged =56.25, x old-growth =49.30, stdev logged =24.002, stdev old-growth =27.570). These results suggest that trees in colobus selected logged forest contained more fruit and mature leaves and equal amounts of young leaves and flowers than trees in old-growth forest. However, this does not represent the overall differences in food availability between the logged and old-growth areas, but rather differences in food availability on selected red colobus feeding trees. Because there are fewer trees in the logged areas (Kasenene 1987; Skorupa 1988; Struhsaker 1997; Chapman & Chapman 1997; Chapman and Chapman 2004; Snaith & Chapman 2008; Chapman et al. 2010), the amount of young leaves and flowers is actually expected to be lower in logged areas.

The feeding tree measurements are from trees that were selected by the monkeys to feed in each day, but do not provide an accurate description of what foods are available in each area of the forest. We expect monkeys to select trees with available foods, so it is not surprising that we do not see large differences in the feeding trees between logged and old-growth areas.

Furthermore, these trees were specifically measured and scored for productivity when the monkeys were feeding in them, not at other points during the season

36 By combining a biomass estimate based on my feeding tree measurements with published data on tree densities, differences in the biomass of red colobus foods between logged and old- growth areas become more apparent. Out of the top 10 food trees, 5 tree species had measurable densities in both logged and old-growth areas (based on Chapman 2002). Not only were

Strombosia trees smaller, they also occurred at a lower density in logged areas (1 tree/hectare compared to 12.5 trees/hectare). Density multiplied by crown volume and % young leaves results in only about 8.6 units of biomass for Strombosia foods in logged areas compared to

117.5 in old-growth areas. Using this same method, Celtis durandii biomass is 482.5 in logged and 747.2 in old-growth, Markhamia and Funtumia biomasses are similar in logged (345.4 for

Markhamia and 189.0 for Funtumia ) and old-growth areas (360.6 for Markhamia and 168.92 for

Funtumia ), and Celtis africana biomass is actually higher in logged areas (105.4) compared to old-growth areas (54.0).

Feeding behaviors

Overall, female red colobus were observed eating 62 different species of plants. Groups in the logged areas differed from groups in the old-growth areas in the number of species they ate and the frequency of eating these species (Tables 4 & 5; F=5.084; df=13; p<0.000).

Strombosia was the only plant species within the top ten most commonly eaten foods across all 4 groups. Females in the logged groups ate a more diverse diet (42 and 43 different species) whereas females in the old-growth groups fed on fewer tree species (29 and 34). Females in the logged groups also ate smaller amounts of each species in their diet. The ten top tree species accounted for 61.6% and 57.9% in the Logged 1 and 2 groups diet (Figure 4a,b ). In contrast, the top ten species of made up 72.6% and 69.7% of the Old-growth Group 1 and 2 diet, respectively

(Figures 4c,d).

37 Both groups in the logged area ate many of the same foods (Table 4). In the logged areas, the two study groups’ food preferences overlapped considerably (28 out of 30 species). In the old-growth area, groups overlapped in by 24 out of 30 plant species. However, all 4 groups only had similar preferences for top feeding trees 50% of the time with only half of the top 30 trees contributing to 90% of the diet for any one group. (Table 5). The food species eaten most often in each of the old-growth groups made up approximately 16% of the diet for that group

(15.7% in Old-growth Group 1 and16.7% in Old-growth Group 2). However, the top food species in the logged areas only made up 9.3% (Logged Group 1) and 10.8% (Logged Group 2).

Across all groups, leaves were the most common plant part eaten (55% -75of the diet), followed by leaf petioles (15-30% of the diet). However, groups differed in the variety of plant parts eaten. Within the old-growth area, neither group was seen eating fruit; however, both groups in the logged area ate fruits (making up 1-5% of the diet). Individuals in the logged areas ate both fruit and leaves from a variety of fig species ( Ficus spp .), making up 9.5% of feeding observations, while fig species only accounted for 0.25% of feeding observations for individuals living in the old-growth areas. Another source of fruit were Strombosia trees. Strombosia was one of the top 10 food trees in all four groups; however, only the groups in the logged areas ate the fruits from this species (though it constituted less than 1% of the diet). There were not major seasonal differences in the plant parts red colobus ate (Table 6).

Activity Budgets

Considering logged and old-growth habitats together, female red colobus spend most of their time feeding and resting, with traveling being the third most common activity. All grooming behaviors make up a small percentage of the overall activity budget in all groups (less than 8%; Table 6). Females had different activity budgets in logged (N=15 individuals) vs. old-

38 growth (N=12 individuals) areas (Figure 7). Females in logged areas fed more often than in old- growth areas ( Z=-3.15, x logged =47.65, stdevl ogged =9.5, x old growth =32, stdev old growth =7.67; p=0.0021). Resting was observed less (Z=3.7816, p=0.0004) in logged areas than in old-growth areas. Traveling was observed significantly less (Z=2.22, p=0.0180) in logged areas than in old- growth areas. Allogrooming was observed significantly less (Z=3.05, p=0.0030) in logged areas than in old-growth areas.

Discussion

The ideal free distribution model predicts that population densities and group sizes decrease in relation to decreased food availability. A decrease in food availability and preferred food resources is the most obvious result of deforestation. Logging activities lead to anthropomorphic changes to primate habitats and can impact wild primate populations by changing the availability, density, abundance, and distribution of plant species. For example, redtail monkey ( Cercopithecus ascanius ) food in Kibale National Park was three times less abundant in heavily logged (7.3 +/- 6.1 m 2/ha) than old-growth areas (21.9 +/- 2.2 m 2/ha) (Rode et al. 2006). These differences in food resulted in lower population densities, smaller group size, fewer polyspecific associations, and more home range overlap for redtails in the heavily logged areas than old-growth areas (Rode et al. 2006).

Red colobus densities differ across Africa, and habitat quality has been associated with these differences (Table 7; Gatinot 1975; Galat Luong & Galat 2005). In this study however, while the density of trees differ in both previously logged and old-growth areas, the red colobus monkeys of Kibale National Park occur at the same densities. This goes against the predictions of the ideal free distribution model and thus requires explanation. It could be possible that the logging activities in Kibale were not sufficient to impact red colobus monkeys or it could be that

39 red colobus demonstrate behavioral plasticity to adjust to living in these degraded habitats (Table

9). Or individuals may be able to adjust their behavior and expand their diet to utilize novel resources in altered habitats. This would effectively expand the resource base in a given area and still uphold the standard that the number of individuals in an area is proportional to the number of resources, as suggested by the ideal free distribution.

Results of this study of feeding strategies of female red colobus in degraded and nondegraded habitats suggest that female behavioral plasticity allows them to maintain their densities. This study found that red colobus in degraded habitats altered their diet compared to groups living in old-growth areas by eating a higher number of plant species, increasing their time spent feeding, and eating foods that have not been recorded in the red colobus diet at Kibale previously (despite being available). By expanding their diet to include a broader range of foods, females in logged areas could effectively increase the resources available to them.

Using behavioral flexibility to maintain densities was observed in one previous study of red colobus monkeys. Galat Luong and Galat (2005) found that over a 30 year period of logging, the forest was reduced by 50%, which resulted in a 30% reduction in species diversity in the forest. However, the red colobus ( Procolobus badius temmincki ) only decreased approximately 17% from 600 to 500 (Galat Luong & Galat 2005). Galat Luong and Galat

(2005) argued that there were 5 major behavioral changes in the 30 year period that may have allowed for coping with the habitat degradation: 1) frugivory, 2) terrestriality, 3) polyspecific associations with green monkeys, 4) frequenting open habitats more, and 5) using mangrove forest for refuge and foraging. Through these behavioral changes, the monkeys were able to adjust to the degraded habitats and maintain densities that did not suffer large decreases. The

40 current study shows similar patterns of behavioral changes for individuals living in degraded areas.

Results of this study’s old-growth populations support previously published work on the feeding ecology of red colobus monkeys in old-growth areas of Kibale National Park (Struhsaker

2010), which found that 15 plant species made up 82.6% of the diet of one group of red colobus monkeys (compared to 13 and 14 plant species making up 83% of the diet for old-growth groups in this study) and that the most commonly eaten food species made up 15.4% of the diet

(Struhsaker and Leland -sympatric) similar to this study. However, the logged groups in this study ate a greater diversity of species in the diet and the most common food species made up lower smaller percentages of the diet for these groups. The plant species observed eaten in the old-growth area are similar to that reported in previous studies (see Struhsaker 2010), including

Celtis africana, Celtis durandii, Markhamia, Strombosia scheffleri, Albizia grandibracteata,

Funtumia latifoilia, Aningeria altissima, Newtonia buchananii , and Prunus africana (Chapman

& Chapman 1999; Chapman, Wasserman, & Gillespie 2006; Struhsaker 2010). However,

Newtonia buchananii and Prunus africana were not commonly eaten by either of the groups in the old-growth area. The current study is only on adult females, which may explain some differences in the results compared to previously published studies on all age/sex classes.

Lovoa and Fagaropsis were determined to be unimportant to red colobus diets in previous studies (Struhsaker 2010). Similarly, in this study, both species were rarely or never observed to be eaten in the old-growth areas. However, Fagaropsis was one of the top species in the diet of females living in logged areas. Also of note, Carapa leaves were eaten by both groups in the logged areas, but never eaten in the old-growth areas. This species has not been observed

41 to be eaten by red colobus at Kibale previously. These results suggest that females in logged areas diversify their diet and utilizing new food sources.

Overlap in high-value timber species and red colobus food sources can make logging activities especially problematic. Strombosia was one of the highest valued timber species that was targeted during logging in Kibale Forest. The remaining Strombosia in logged areas have smaller DBH and crown size than in the old-growth areas. Smaller Strombosia tree sizes in logged areas may explain why Chapman and colleagues (2006) found that Strombosia constituted a smaller part of the red colobus diet in logged areas. Larger Strombosia trees in the old-growth areas likely provide more food for the monkeys to eat.

A high level of overlap exists between the types of food being eaten in different groups within the same habitat; however, these consistencies were lost when looking across both logged and old-growth habitats. Females in logged habitats ate a greater number of species and smaller amounts of each of these species and also spent more time feeding overall. Because these feeding behaviors differed between habitat types more than within, this suggests that logging has impacted the feeding ecology of red colobus monkeys in Kibale. Females in old-growth areas ate foods typical for red colobus whereas females in logged areas utilized a wider range of resources in smaller amounts. A relatively small amount of feeding time on some food species in the logged area may be a strategy to utilize the resources that are available to maintain nutrient intake when preferred foods are not available. A similar pattern was found in ruffed lemurs, where individuals living in degraded habitats diversified their diet to include different types of fruits compared to those in pristine habitats (Ratsimbazafy 2007). Also, increases in populations densities of mantled howler monkeys Alouatta palliata ) in forest fragments of less than 10 ha in

42 size were associated with more diverse diets and incorporating new foods compared to groups living at lower densities in larger forest areas (Cristobal-Azkarate & Arroyo-Rodriquez 2007).

Previous studies also have shown that individuals in disturbed habitats have different activity budgets than individuals in undisturbed areas (Altmann & Muruthi 1988; Boyle & Smith

2010; Eley et al. 1989; Isbell & Young 1993; Iwamoto & Dunbar 1983; Marsh 2003;

Ratsimbazafy 2007; Silva and Ferrari 2009; Singh & Vivanthe 1990; Watts 1988; and Wong &

Sicotte 2007). For example, baboons (Altmann & Muruthi 1988; Eley et al. 1989; Iwamoto &

Dunbar 1983), gorillas (Watts 1988), and bonnet macaques (Singh & Vivanthe 1990) spent more time foraging and feeding in degraded areas than individuals living in high quality habitats.

Female red colobus monkeys in logged and old-growth areas of Kibale National Park also differed in their behaviors, despite similar overall behavioral patterns . As predicted, females in logged areas spent significantly more time feeding than females in unlogged areas. This increase in feeding time may be the result of having lower quality foods. Females in logged areas ate a significantly different diet than females in unlogged areas. Thus, females in logged areas may spend more time feeding to maintain energy intake despite different resource availability, suggesting they may be using an energy-minimizing foraging strategy suggested by Milton

(1980).

Chapman, Wasserman, & Gillespie (2006) found that Olea welwitschii densities were the same between logged and old-growth areas despite being targeted during logging operations.

Whereas red colobus in old-growth areas rarely ate this species (3% of feeding observations), groups in the logged areas fed on it relatively frequently and had it as one of their top 15 foods in both the wet and dry season. Olea are known for growing larger in open areas. Of the Olea feeding trees recorded in this study (N=10), the ones in the logged areas had higher DBH and

43 crown size than those in the old-growth areas. Though based on a small sample size, this may explain why Olea trees are eaten in logged areas and not old-growth areas. Olea appear to have the opposite response to logging than Strombosia , so while Strombosia trees were smaller and may not provide as much food (hence the lower rates of feeding on this species in the logged area), Olea seem to thrive and provide more food for red colobus in the logged areas.

Nonetheless, it is unlikely that this is a high quality food source considering that multiple studies of red colobus do not show that this is a preferred food.

In addition to eating different species of plants, red colobus in the logged areas ate different plant parts. Individuals living in the logged areas ate fruitwhereas individuals in the old-growth areas never did. This is a similar behavioral change seen in Galat Luong and Galat

(2005) in which frugivory was one of the top 5 behavioral changes observed in red colobus over

30 years of habitat degradation. The gut morphology of red colobus is adapted to processing foliage, and the acidity of fruit can impact the ability to successfully digest foods (Danish et al.

2006). The consequence may be insufficient food resources to maintain population size and negative impacts on the health and physiology of logged individuals, resulting in population size decreases. Eating foods that are not commonly eaten by red colobus in other parts of Kibale may make it possible for individuals in logged areas to maintain their densities. This study provides evidence that red colobus differ in both their activity budgets and diet composition, which supports the idea that red colobus have a plastic response to living in disturbed areas.

Additional studies examining the nutrient content of the foods being eaten and impacts on the physiology and overall health of the red colobus in degraded habitats are essential for further understanding the impact of habitat disturbance on this species.

44 Aside from dietary changes, differences in habitat quality may also impact species on a variety of other levels (e.g. diminished reproduction, starvation, lower immunity, lower survival, etc). Previous studies have shown that primates living in logged areas of Kibale National Park have decreased reproductive fitness, smaller body size, and occur in lower densities. Olupot

(2000) found that the body mass of mangabeys ( Lophocebus albigena ) living in logged areas was significantly lower than the body mass of mangabeys living in old-growth areas of the park

(approximately 8.5 kg on average compared to approximately 9.5 kg on average). Similarly, chimpanzees living in areas of the park with “more abundant or higher-quality food” sources had higher reproductive hormone concentrations (estrogen and progestin conjugates), shorter interbirth intervals (e.g. 31.8 vs 55.9 months), and higher infant survival (e.g. 15.4 vs 7.6 years) than females living in lower quality areas (Emery Thompson et al. 2007).

Understanding how wild populations respond to degraded habitats is important for conservation management plans, for understanding interspecific variation, and for hypothesizing about the role of the environment in past speciation events (e.g see Soule et al 2005; Kennedy and Grey (1993) By building on previous studies examining habitat quality and primate food availability within and around Kibale National Park, Uganda (e.g. Chapman et al. 2004, 2006;

Emery Thompson et al. 2007), these results can help to inform conservation policy and management practices for this ecosystem. Specifically, these results suggest that red colobus are able to adjust to degraded habitats when there are alternative resources for them to utilize and when the habitat is stable. This contributes to our understanding of why red colobus in logged areas of Kibale are able to maintain densities similar to those in old-growth areas, but red colobus in forest fragments are not. These results also support the benefits of protecting forested land, even if it has been degraded previously, andsuggest that restoration efforts in Kibale

45 National Park should be focused on the needs of other species. Kibale is home to 13 species of primates and restoration efforts cannot address the specific needs of each individual species.

Identifying what animals are flexible in different environments is key to determining what species should be targeted for conservation efforts.

The current study suggests that individuals in logged areas adjust their feeding behaviors in response to habitat changes, which has important implications for both evolutionary theory

(understanding how species deal with environmental change) and conservation practices

(prioritizing protection of different forest areas). Additional studies which examine the nutrient content of the foods being eaten in each area, evaluations of trees species available in each area, and behavioral observations of differences in time spent on specific activities will help to illuminate the role that food availability has on these feeding differences. An understanding of the impact of logging on endangered primate populations is essential for management practices and can provide insight into the response of species to ecological changes throughout evolution.

46 Work Cited

Altmann J and Muruthi P. 1988. Differences in daily life between semi-provisioned and wild- feeding baboons. American Journal of Primatology 15: 213–221.

Boyle, S.A., Smith, A.T., 2010. Behavioral modifications in northern bearded saki monkeys (Chiropotes satanas chiropotes) in forest fragments of central Amazonia. Primates 51, 43–51.

Chapman CA and Chapman LJ. 1990. Dietary variability in primate populations. Primates 31 (1): 121-128.

Chapman CA. 1988. Patterns of foraging and range use by three species of neotropical primates. Primates 29 (2): 177-194.

Chapman CA and Chapman LJ. 1997. Forest regeneration in logged and unlogged forests of Kibale National Park, Uganda. Biotropica 29: 386-412.

Chapman, C.A., R. Wrangham, and L.J. Chapman. 1995. Ecological constraints on group size: An analysis of spider monkey and chimpanzee subgroups. Behavioural Ecology and Sociobiology 36:59-70.

Chapman, C.A. and L.J. Chapman. 2000. Determinants of group size in social primates: The importance of travel costs. On the Move: How and Why Animals Travel in Groups, Pp. 24-42. S. Boinski and P.A. Garber (eds). University of Chicago Press, Chicago.

Chapman CA, Balcomb SR, Gillespie TR, Skorupa JP, and Struhsaker TT. 2000. Long-term effects of logging on African primate communities: a 28-year comparison from Kibale National Park, Uganda. Conservation Biology 14: 207-217.

Chapman CA and Chapman LJ. 1999. Implications of small scale variation in ecological conditions for the diet and density of red colobus monkeys. Primates 40:215-232.

Chapman CA and Chapman LJ. 2002. Foraging challenges of red colobus monkeys: influence of nutrients and secondary compounds. Comparative Biochemistry and Physiology Part A 133: 861-875.

Chapman CA and Chapman LJ. 2004. Unfavorable successional pathways and the conservation value of logged tropical forest. Biodiversity and Conservation 13: 2089-2105.

Chapman CA, Chapman LJ, Wrangham R, Isabirye G, and Ben-David K. 1997. Spatial and temporal variability in the structure of a tropical forest. African Journal of Ecology 35:287–302.

47 Chapman CA, Chapman LJ, Naughton-Treves L, Lawes MJ, and McDowell LR. 2004. Predicting folivorous primate abundance: validation of a nutritional model. American Journal of Primatology 62: 55-59.

Chapman CA and Lambert JE. 2000. Habitat alteration and the conservation of African primates: A case study of Kibale National Park, Uganda. American Journal of Primatology 50: 169-186.

Chapman CA, Struhsaker TT, and Lambert JE. 2005. Thirty years of research in Kibale National Park, Uganda, reveals a complex picture for conservation. International Journal of Primatology 26:539-555.

Chapman CA, Wasserman MD, and Gillespie TR. 2006. Behavioral patterns of colobus in logged and unlogged forests: The conservation value of harvested forests. In: Newton- Fisher NE, Notman H, Paterson JD, and Reynolds V (eds.) Primates of Western Uganda . Springer, New York: 373-390.

Chapman CA, Wasserman MD, Gillespie TR, Speirs ML, Lawes MJ, Saj TL, and Ziegler TE. 2006b. Do food availability, parasitism, and stress have synergistic effects on red colobus populations living in forest fragments? American Journal of Physical Anthropology 131:525-534 .

Chapman CA, Struhsaker TT, Skorupa JP, Snaith TV, and Rothman JM. 2010. Understanding long-term primate community dynamics: implications of forest change. Ecological Applications 20 (1): 179-191.

Catchpole WR. and Wheeler CJ. 1992. Estimating plant biomass: a review of techniques. Australian Journal of Ecology 17(2): 121-131.

Cowlishaw G. 1997. Trade-offs between foraging and predation risk determine habitat use in a desert baboon population. Animal Behaviour 53: 667-686.

Chapman, C.A., R.R. Ghai, R. Reyna-Hurtado, A.L. Jacob, S.M. Koojo, J.M. Rothman, D. Twinomugisha, M.D. Wasserman, and T.L. Goldberg. In press. Going, going, gone: A 15-year history of the decline in abundance of primates in forest fragments. In: L.K. Marsh and C.A. Chapman. Primates in fragments: complexity and resilience. Springer Press.

Danish L, Chapman CA, Hall MB, Rode KD, and O’Driscoll Worman C. 2006. The role of sugar in diet selection in redtail and red colobus monkeys. In: Hohman G, Robbins MM, Boesch C, editors. Feeding Ecology in Apes and Other Primates . Cambridge, UK: Cambridge University Press. p. 473-487.

Chapman, C.A. and Rothman, J.M. 2009. Within-species differences in primate social structure: Evolution of plasticity and phylogenetic constraints. Primates 50:12-22.

48 Davies AG. 1994. Colobine populations. In: Davies AG and Oates JF, editors. Colobine monkeys . Cambridge University Press, Cambridge. p 285-310.

Eley RM, Strum SC, Muchemi G, and Reid GDF. 1989. Nutrition, body condition, activity patterns and parasitism of free-ranging baboons ( Papio anubis ) in Kenya. American Journal of Primatology 18: 209–219.

Emery Thompson M, Kahlenberg SM, Gilby IC, and Wrangham RW. 2007. Core area quality is associated with variance in reproductive success among female chimpanzees at Kibale National Park. Animal Behaviour 73: 501-512.

Enquist BJ, Brown JH, West GB. 1998. Allometric scaling of plant energetics and population density. Nature . 395:163–165.

Enquist BJ, Niklas KJ. 2001. Invariant scaling relations across tree dominated communities. Nature . 410:655–660.

Felton AM, Engstrom LM, Felton A, and Knott CD. 2003. Orangutan population density, forest structure and fruit availability in hand-logged and unlogged peat swamp forests in West Kalimantan, Indonesia. Biological Conservation 114:91-101.

Fretwell SD & Lucas HL. 1969. On territorial behavior and other factors influencing habitat distribution in birds. Acta biotheoretica , 19(1), 16-36.

Garber PA. 1987. Foraging strategies among living primates. Annual review of Anthropology, 339-364.

Gatinot BL. 1975. Ecologie d'un Colobe bai (Colobus badius temmincki, Kuhl 1820) dans un milieu marginal au Senégal. Doctoral Dissertation. L’Universite de Paris.

Galat Luong A & Galat G. 2005. Conservation and survival adaptations of Temminck’s red colobus ( Procolobus badius temmincki ), in Senegal. International Journal of Primatology 26 (3): 585-603.

Gebo DL and Chapman CA. 1995. Habitat, annual, and seasonal effects on positional behavior in red colobus monkeys. American Journal of Physical Anthropology 96: 73-82.

Gillespie TR and Chapman CA. 2006. Prediction of parasite infection dynamics in primate metapopulations based on attributes of forest fragmentation. Conservation Biology 20:441-448.

Isbell, L.A. & Young, T.P. ( 1 9 9 3 ) . Human presence reduces leopard predation in a population of vervet monkeys Cercopithecus aethiops. Anon. Behav ., 45, 1233-5.

Iwamoto T & Dunbar RIM. 1983. Thermoregulation, habitat quality and the behavioural ecology of gelada baboons. Journal of Animal Ecology 52: 357-366.

49

Kasenene JM. 1987. The influence of mechanized selective logging, felling intensity and gap- size on the regeneration of a tropical moist forest in the Kibale Forest Reserve, Uganda. Ph.D. Dissertation. Michigan State University, Michigan.

Komers PE. 1997. Behavioural plasticity in variable environments. Canadian Journal of Zoology, 75(2), 161-169.

Lawes MJ & Chapman CA. 2006. Does the herb Acanthus pubescens and/or elephants suppress tree regeneration in disturbed Afrotropical forest?. Forest ecology and management , 221(1), 278-284. MacArthur RH & Pianka ER. 1966. On optimal use of a patchy environment. American Naturalist , 603-609.

Marsh, L. K. 2003. Primates in fragments: ecology and conservation . Kluwer Academic/Plenum Publishers, New York.

Martin P and Bateson P. 1986. Measuring behaviour: An introductory guide . Cambridge University Press, Cambridge.

Milton K. 1980. The foraging strategy of howler monkeys: a study in primate economics . Columbia University Press, New York.

Nakagawa N. 1990. Decisions on time allocation to different food patches by Japanese monkeys (Macaca fuscata). Primates 31(4): 459-468.

Oates JF. 1994. The natural history of African colobines. In: Davies AG and Oates JF (eds.) Colobine Monkeys: Their Ecology, Behavior, and Evolution . Cambridge University Press, Cambridge: 75-128.

Oates JF. 1999. Myth and reality in the rainforest: how conservation strategies are failing in West Africa. University of California Press, Berkeley.

Olupot W. 2000. Mass differences among male mangabey monkeys inhabiting logged and unlogged forest compartments. Conservation Biology 14:833-843.

Onderdonk DA and Chapman CA. 2000. Coping with forest fragmentation: The primates of Kibale National Park, Uganda. International Journal of Primatology 21:587-611.

Pigliucci M. 2001. Phenotypic Plasticity: Beyond Nature and Nurture. Johns Hopkins University Press, Baltimore.

Pulliam HR, Caraco T. 1984. Living in groups: is there an optimal group size? In: Krebs JR, Davies NB (eds) Behavioural ecology. An evolutionary approach, 2 edn. Blackwell, Oxford: 122-147

50 Ratsimbazafy J. 2007. Diet Composition, foraging, and feeding behavior in relation to habitat disturbance: implications for the adaptability of ruffed lemurs ( Varecia v. editorium ) in Manombo forest, Madagascar. In Gould L and Sauther ML (eds) Lemurs: Ecology and Adaptation . Springer, New York: 403-422.

Robinson JG. 1986. Seasonal variation in use of time and space by the wedge-capped capuchin monkey, Cebus olivaceus: implications for foraging theory. Washington, DC: Smithsonian Institution Press.

Rode KD, Chapman CA, McDowell LR, and Stickler C. 2006. Nutritional correlates of population density across habitats and logging intensities in redtail monkeys (Cercopithecus ascanius ). Biotropica 38:625-634.

Silva, S. S. B., & Ferrari, S. F. (2009). Behavior patterns of southern bearded sakis Chiropotes satanas) in the fragmented landscape of eastern Brazilian Amazonia. American Journal of Primatology , 71, 1–7.

Singh M and Vinathe S. 1990. Inter-population differences in the time budgets of bonnet monkeys ( Macaca radiata ). Primates 31(4): 589-596.

Skorupa JP. 1988. The effect of selective timber harvesting on rain forest primates in Kibale Forest, Uganda. PhD Thesis. University of California, Davis.

Skorupa JP and Kasenene JM. 1984. Tropical forest management: can rates of natural treefalls help guide us? Oryx , 18:96-101.

Snaith TV and Chapman CA. 2008. Red colobus monkeys display alternative behavioral responses to the costs of scramble competition. Behavioral Ecology 19: 1289-1296.

Stampone MD, Hartter J, Chapman CA, and Ryan SJ. 2011. Trends and variability in localized precipitation around Kibale National Park, Uganda, Africa. Research Journal of Environmental and Earth Sciences 3 (1): 238-253.

Struhsaker TT. 1978. Food habits of five monkey species in the Kibale Forest, Uganda. In Chivers, D. J., and Herbert, J. (eds.), Recent Advances in Primatology, Vol. 1, Behaviour, Academic Press, New York, pp. 225–248.

Struhsaker TT. 1997. Ecology of an African Rain Forest: Logging in Kibale and the conflict between conservation and exploitation . University Press of Florida, Gainesville.

Struhsaker TT. 1975. The Red Colobus Monkey . University of Chicago Press, Chicago.

Struhsaker TT. 2010. The Red Colobus Monkeys . Oxford University Press, Oxford.

51 Struhsaker TT and Leland L. 1979. Socioecology of five sympatric monkey species in the Kibale Forest, Uganda. Advances in the Study of Behavior 9:159-228.

Struhsaker TT and Leland L. 1987. Colobines: infanticide by adult males. In: Smuts BB, Cheney DL, Seyfarth RM, Wrangham RW, and Struhsaker TT (eds.) Primate Societies . University of Chicago Press, Chicago: 83-97.

Watts, DD.P. 1988. Environmental influences on mountain gorilla time budgets. American Journal of Primatology 15: 295–312. , 1988

Wrangham, RW. 2002. The cost of sexual attraction: is there a trade-off in female Pan between sex appeal and received coercion? In Boesch C, Hohmann G, and Marchant LF (eds.) Behavioural Diversity in Chimpanzees and Bonobos. University Press, Cambridge: 204- 215.

Wong S. N. P. and Sicotte P. (2007). Activity budget and ranging patterns of Colobus vellerosus in forest fragments in Central Ghana. Folia Primatol. 78: 245–254.

52 K-13

K-15

K-30

Figure 1: Map of the logging history at the Kanyawara Field Site in Kibale National Park, Uganda (adapted from Struhsaker 1975). K-13 was heavily logged and poisoned with an arborcide, K-15 was heavily logged, and K-30 was never commercially logged.

53 Table 1: Demographic information for red colobus at Kibale National Park, Uganda and Abuko Nature Reserve, The Gambia ( 1Struhsaker and Leland 1987; 2 Starin 1991; 3 Struhsaker 1975).

Trait Kibale National Park Abuko Nature Reserve Interbirth interval 25.5 ± 5.1 months 1 29.4 months 2 Gestation length 4.5 – 5.5 months 3 5.25 months 2 Menstrual cycle length Unknown 27.7 days 2 Age at sexual maturity (males) 35.4 - 58 months 1 > 28 months 2 Age at sexual maturity (females) 38 - 46 months 1 34.25 months 2

Table 2: Group compositions for the four research groups.

Logged Logged Old-growth Old-growth Group 1 Group 2 Group 1 Group 2 Adult Males 8 4 9 11 Adult Females 22 14 15 22 Subadults 3 3 2 4 Juveniles 8 10 1 23 Infants 10 6 7 13 Total 51 37 34 73 Infant/Adult Female Ratio 0.45 0.43 0.47 0.59

54

Old-growth

Albizia grandibractata.

Bosqueia phoberos

Celtis africana

Celtis durandii

Group 1 Dombeya mukoke Group 1 Funtumia latifolia

Markhamia lutea

Milletia dura

Prunus africana

Strombosia scheffleri

Other

Group 2 Group 2 (p<0.000

Figure 2: Diversity of feeding trees in the logged and old-growth areas. The top 10 species that red colobus selected as feeding trees in each area are color coded and all other species are labeled “other” and assigned the beige color. The two groups in the logged areas had a much more diverse range of feeding trees and each of these species made up a relatively small portion of the overall census. In the old-growth areas, the majority of the diet consisted of a few key preferred food species.

55

Table 3: The most commonly eaten tree species based on twice daily feeding tree records (percentage that each species made up of overall feeding tree records for each group). Feeding trees were tagged and measured in the same seasons for all groups, as all study groups were followed simultaneously.

Old- Old-growth Tree Species Logged 1 Logged 2 growth 1 2 Albizia grandibractata 11 7 7 0 Bosqueia phoberos 0 0 14 8 Celtis africana 0 0 16 7 Celtis durandii 0 0 7 7 Dombeya mukoke 0 0 8 12 Funtumia latifolia 11 12 7 0 Markhamia lutea 0 0 13 11 Milletia dura 7 8 8 8 Prunus africana 12 0 0 0 Strombosia scheffleri 7 10 8 20 Other 52 63 12 27

56 Table 4: The most commonly eaten foods for each group. This table represents the species making up 90% of the diet in each group (as noted by X). Old- Old- growth 1 growth 2 Logged 1 Logged 2 Accasia X X Albizia grandibractata X X X X Aningeria altissima X X X X Bosqueia phoberos X X Celtis africana X X X X Celtis durandii X X X X Cordia X X X Diospyrus abyssinica X X Dombeya mukoke X X Ehretia cymosa X X Eucalyptus X Ficus dawei X X Ficus natalensis X X Fagaropsis X X X Funtumia latifolia X X X X Hypocratia plubmea X X X X Markhamia lutea X X X X Milletia dura X X X X Mimusops bagshawei X X X X Myrianthus X X Newtonia buchannii X X X X Olea welwitschii X X X Parinari escelsa X X X X Premna angolensis X X Prunus africana X X X X Sapium ellipticum X X X Strombosia scheffleri X X X X Sycrostaches X X X Symphonia X X X X Urella X X X X

57 Table 5: The tree species that were observed being eaten most commonly by female red colobus (percentage that each species made up of overall observed feeding records for each group during wet and dry seasons). Feeding observations were all collected simultaneously for all groups.

Season: Dry Wet Habitat: Old-growth Logged Old-growth Logged Species 1 2 1 2 1 2 1 2 Accasia 0.00 0.00 2.85 5.66 0.00 0.43 14.38 13.94 Albizia grandibractata 4.78 1.97 12.56 6.37 4.07 2.28 0.67 6.12 Aningeria 4.78 7.30 0.84 3.07 3.25 1.28 2.69 0.00 Bosqueia phoberos 21.53 5.72 0.00 0.00 5.69 5.42 0.00 0.00 Celtis africana 3.83 1.78 1.17 3.77 14.63 3.14 1.88 4.84 Celtis durandii 8.13 6.51 3.18 5.19 3.25 6.85 1.75 4.13 Dombeya mukoe 4.78 4.34 0.00 0.00 8.94 8.27 0.00 0.00 Funtumia latifolia 1.91 7.89 1.68 8.73 4.07 4.14 5.78 7.11 Markhamia lutea 5.74 9.47 4.36 0.94 15.45 10.27 3.49 3.27 Milletia dura 1.91 2.56 6.20 2.36 6.50 6.70 5.38 7.68 Mimusopes bagshawei 4.78 7.50 0.00 9.91 1.63 4.14 2.55 2.84 Olea welwitschii 5.26 0.00 3.52 4.72 0.00 0.00 6.45 3.84 Prunus africana 0.96 0.99 13.57 2.36 0.00 0.29 5.65 4.41 Stronbosia scheffleri 11.00 18.54 4.36 0.00 5.69 15.41 4.84 6.54 Urella 0.00 0.39 9.05 0.94 1.63 2.14 8.33 3.70

58

Figure 3a: Difference in DBH of tree species in logged (yellow) and old-growth (green) areas.

59

Figure 3b: Difference in crown size between tree species that were fed on by females in logged (yellow) and old-growth (green) areas.

60 Top 10 foods for Logged Group 1

Accasia Albizia grandibractata Ehretia cymosa Funtumia latifolia Milletia dura Newtonia buchannii Olea welwitschii Prunus africana Strombosia scheffleri Urella Other

Figure 4a : The top 10 plant species consumed make up 61.6% of Logged Group 1’s diet

61 Top 10 foods for Logged 2

Accasia Albizia grandibractata Celtis africana Celtis durandii Funtumia latifolia Milletia dura Mimusops bagshawei Olea welwitschii Strombosia scheffleri Symphonia Other

Figure 4b: The top 10 plant species consumed make up 57.9% of Logged Group 2’s diet.

62

Top 10 foods for Old-growth Group 1

Albizia grandibractata Aningeria altissima Bosqueia phoberos Celtis africana Celtis durandii Dombeya mukoke Hypocratia plubmea Markhamia lutea Strombosia scheffleri Sycrostaches Other

Figure 4c: The top 10 food species consumed make up 72.6% of Old-growth Group 1’s diet.

63

Top 10 foods for Old-growth Group 2

Bosqueia phoberos Celtis durandii Dombeya mukoke Funtumia latifolia Markhamia lutea Milletia dura Mimusops bagshawei Parinari escelsa Premna angolensis Strombosia scheffleri Other

Figure 4d: The top 10 food species consumed make up 69.7% of Old-growth Group 2’s diet.

Table 6: Percentage of the plant parts eaten for each group in the wet and dry seasons. All groups were observed simultaneously.

Old- Old- Old- Old- Log Log growth growth Log Log growth growth 1 2 1 2 1 2 1 2 Dry Dry Dry Dry Wet Wet Wet Wet Bark 4.04 9.79 9.62 4.94 2.34 4.88 4.07 0.72 Flowers 1.01 0.24 2.88 2.96 1.43 1.58 2.44 0.29 Fruits 1.18 4.30 0.00 0.00 1.04 5.60 0.00 0.00 Leaves 75.08 64.92 55.29 48.81 75.68 66.00 54.47 62.27 Petioles 15.99 15.04 25.48 33.60 15.73 14.92 28.46 27.12 Pith 1.01 1.19 1.44 0.59 0.91 2.87 0.81 4.45 Seeds 1.52 4.30 5.29 8.10 2.47 4.16 9.76 3.01

64 Table 7: Activity budgets of female red colobus in the logged and old-growth areas of Kibale National Park. All observations were collected simultaneously for all groups.

Activity budgets of focal females (percentage of overall activity budget) Activity Logged (%) Old-growth (%) Struhsaker Snaith & Struhsaker 1975 (%) Chapman 2010 (%) 2008 (%) Feeding 46 35 44.5 40-51 30-50 Resting 34 39 34.8 25-30 30-40 Traveling 14 18 9.2 16-29 7-11 Allogroom 4 6 5.3 5-10 2-8 Self-groom 2 2 NA NA NA

65 Activity budget 70

60

50

40 Old-growthUnlogged Logged 30

20 Percent of Total Activity Budget

10

0 FeedFeed Travel Rest Rest Travel Allogroom Allogroom

Figure 5: Activity budgets of red colobus females in logged and old-growth areas of Kibale National Park. Compared to females in old-growth areas (N=12 individuals, x=32, stdev=7.67), females in logged areas (N=15 individuals, x=47.65, stdev=9.5) spent more time feeding (Z=- 3.15, p=0.0021) and less time resting (Z=3.7816, p=0.0004), traveling (Z=2.22, p=0.0180), and allogrooming (Z=3.05, p=0.0030).

66

Table 8: Densities of red colobus monkeys in different areas of Africa.

Species Density Group Habitat Citation

Size

P. temminckii 105 individuals/km 2 32.08 Open Forest Gatinot 1975

P. temminckii 433 individuals/km 2 32.08 Dense Rivervine Gatinot 1975 Forest P. temminckii 225 individuals/km 2 26.5 Mosaic (largely Starin 1991 undisturbed) P. badius 112 individuals/km 2 52.25 Old-growth Calculated by Struhsaker 2010 from Korstjens 2001 P. tephrosceles 160 individuals/km 2 40 Old-growth Calculated from Chapman 2010 P. tephrosceles 160 individuals/km 2 40 Heavily logged Calculated from Chapman 2010

Table 9: Hypotheses for why densities do not differ between the logged and unlogged areas with expected observations.

Reason for no differences in Expected Observation densities between logged and old- growth areas .

Logging did not cause significant Option 1: Individuals eat the same plant species, food differences in habitat quality or types (i.e. leaves), and spend the same amount of time food availability for red colobus feeding in all four groups. monkeys. Option 2: There is variation in feeding strategies across all groups rather than similarities within a given habitat type, but differences across habitats. Behavioral plasticity allows Individuals in each area (i.e. logged vs old-growth) eat individuals to adjust to degraded similar plant species and food types, but this pattern habitats. differs between the different habitats. Individuals in each habitat type have similar activity budgets, but they differ from individuals in the other habitat.

67 CHAPTER 3

TIMING IS EVERYTHING: EXPANDING THE COST-OF-SEXUAL-ATTRACTION HYPOTHESIS

Abstract

Anthropogenic disturbances present challenges to animals and lead to behavioral changes. Behavioral plasticity may be an important strategy to adjust to degraded habitats. This study examines how ecological conditions impact female red colobus ( Procolobus rufomitratus ) reproduction in Kibale National Park, Uganda. Wrangham (2002) proposed the “cost of sexual attraction” hypothesis to explain the relationship between ecology and female reproduction in

Pan . Here I expand on and for the first time test this hypothesis in a folivorous, non-fission- fusion species, the red colobus monkey. I compared 4 groups of red colobus monkey

(Procolobus rufomitratus ) (two in historically logged areas of the park and two in old-growth areas) to examine differences in female reproductive behaviors and physiologies. I predicted that females living in logged areas would 1) have genital swellings for a shorter period of time, 2) mate less frequently, and 3) constrain mating behaviors more to periods of maximal genital tumescence compared to females in old-growth areas. Results indicate that females in logged areas were fully inflated for a significantly shorter amount of time (P=0.0468), copulated significantly less frequently (P=0.0043), and exhibited mating behaviors when fully inflated significantly more (P=0.0117) than females in old-growth areas. Female behavioral plasticity in logged areas results in behavioral and physiological changes that allow them to maintain reproductive function under environmental constraints.

Introduction

Behavioral plasticity is thought to be an important adaptation for individuals responding to selective pressures of the environment. One form of behavioral plasticity, flexible mating

68 behaviors, may be particularly important for females who are often subordinate to males. Female primates are notable in that their mating is often not confined to ovulation, but occurs over an extended timeframe (Hrdy & Whitten 1987; Hayssen et al 1993; Loy 1987; Struhsaker 1997).

These non-reproductive mating behaviors have complex origins, including as important social mediators. Female primates use a variety of strategies to maximize their reproductive success

(Stumpf et al. 2011). For example, females benefit from mating with multiple males over the course of their cycle, but selectively mating with preferred males during periovulatory periods

(Stumpf & Boesch 2005). Mating with a variety of males throughout the cycle is an important strategy to avoid infanticide (Hrdy 1979; Palombit et al. 2000; Smuts & Smuts 1993).

Behavioral plasticity may be especially important for individuals dealing with energetic constraints in different environments (Marsh 2003, Rodriquez-Luna et al. 2003; Silver & Marsh

2003; Estrada & Coates-Estrada 1996). Ecological conditions are found to correlate with reproductive fitness in a number of species (Altmann & Alberts 2003; Bercovitch 2001; Berger et al. 1999; Emery Thompson et al. 2007; Knott 1999; Lipson 2001). Maintaining a positive energy balance is particularly important for females because it correlates positively with ovarian function, maintains pregnancy, and supports lactation (Ellison 1990; Emery Thompson &

Wrangham 2008; Miller et al. 2006; Sherry 2002; Tardif et al. 2005; Wade & Schneider 1992).

Certain species may use specific strategies to deal with the energetic demands of reproduction, such as seasonally breeding sifakas ( Propithecus verreauxi verreauxi ) that give birth during periods of low food availability and wean during periods of the highest food availability (Lewis

& Kappeler 2005). Even within a species, environmental variability can cause variation in life history traits from the population mean (Boyd 2000). Behavioral plasticity may be a vital way for females to cope with living in degraded habitats and maintain reproductive function.

69 Wrangham’s (2002) cost-of-sexual-attraction hypothesis broadly argued that ecological conditions impact female reproductive physiology and sexual behaviors. He specifically proposed this relationship for the genus Pan and compared published data for eastern and western chimpanzees and bonobos. Two important characteristics of Pan are that females have genital swellings that become maximally tumescent around ovulation and that the members of this genus live in fission-fusion societies. Wrangham (2002) argued that maximally tumescent females attract males, which effectively increases their subgroup size and that the cost of larger subgrouping increases feeding competition. Thus, Wrangham (2002) argued that females in low quality habitats cannot afford to spend time in large parties, so fewer reproductive cycles and limited time for mating would be advantageous to females. This pattern should result in high mating rates over fewer cycles and could be seen through shorter periods of maximal tumescence and increased daily copulation rates.

Wrangham (2002) tested his hypothesis with published data, but miscalculated some reproductive parameters of western chimpanzees (see Deschner & Boesch 2007). Although western chimpanzees still had more estrous cycles compared to East African chimpanzees, expectations for the cost-of-sexual-attraction hypothesis were complicated by variation in individual female reproductive parameters, e.g. parous vs. nulliparous females (Deschner &

Boesch 2007).

The cost-of-sexual-attraction hypothesis warrants further investigation, in particular, with data collected specifically to tests this hypothesis. Here I expand on Wrangham’s cost-of-sexual- attraction model and provide the first systematic test of its assumptions. In this study, I examine four groups of red colobus ( Procolobus rufomitratus ) (including two occupying previously logged forest and two occupying old-growth forest) to determine whether females living in

70 degraded habitats minimize the cost-of-sexual-attraction by limiting their mating behavior to the time when they are most likely to conceive. Specifically, I test three main hypotheses: 1) females in logged areas have a shorter duration of maximal tumescence than females in old- growth areas, 2) females in disturbed areas constrain their mating more to maximal tumescence compared to females in the old-growth area, and 3) females in the logged areas mate less than the females in the old-growth areas.

Methods

Research was conducted in the forest near Makerere University Biological Field Station

(MUBFS) in Kibale National Park, Uganda (hereafter Kibale; 795 km 2; 0°13'- 0°41'N, 30°19'-

30°32'E). Kibale NP is a mid-altitude, moist evergreen forest that receives 1696 mm of rainfall annually (1990-2011; Chapman unpublished data; Chapman and Lambert 2000; Stampone et al.

2011). Kibale provides a valuable setting for examining the impact of habitat disturbance on foraging strategies in primate. Forestry compartment K-30 is old-growth forest that was never logged (Struhsaker 2010, Chapman et al. 2010), while K-15 was heavily logged and 50% of the trees were cut or indirectly killed (Kasenene 1987; Skorupa 1988; Struhsaker 1997; Chapman &

Chapman 1997; Chapman and Chapman 2004; Figure 6). The K-13 area was heavily logged and poisoned with arboricide (Oates 1999). These previously logged areas are still highly degraded and have large gaps in the canopy that persist today.

Red colobus monkeys are an important species for investigating primate responses to habitat disturbances because they are highly endangered (Struhsaker & Leland 1987) and affected by ecological stressors, such as living in small forest fragments outside of Kibale

National Park (Chapman et al. 2006; Gillespie & Chapman 2006). Moreover, multiple groups can be studied in a relatively small area, making this a valuable species for comparative work.

71 Red colobus groups are patrilinear and males form a loose dominance hierarchy (Korstjens 2001;

Struhsaker & Leland 1979, 1987). Unlike other colobines, female red colobus transfer from their natal groups (Korstjens 2001; Struhsaker & Leland 1979, 1987). Most of what is known about red colobus reproductive parameters comes from Kibale and Abuko (Struhsaker 2010), where both sites show similarities in interbirth intervals, menstrual cycle length, and age at sexual maturity (Table 10). Female red colobus have perineal swellings that fluctuate in size over the course of their cycle (Struhsaker 1975; Struhsaker & Leland 1987). Copulations occur through the year (Struhsaker 1997) and females mate with multiple males throughout their cycle, including when they are not tumescent (Starin 1991; Struhsaker 1975, 2010; Struhsaker &

Leland 1985).

Data were collected on four red colobus groups: two groups in the previously logged areas of K-15 and K-13 (Figure 6) and two groups in the area that was never commercially logged, K-30. Each of the 4 focal groups was followed by the research team from 8:00 am to

4:00 pm six days a week (Table 11). Five focal females (adult, parous females without infants) were selected in each group and observed on a rotating basis until they gave birth, at which time

I would choose a new adult female without an infant to enter the rotation so that at each point in time, data were obtained from 5 females from each group. (Note: Because target females changed depending on whether they gave birth during the study, the sample size of focal females for each group differs.) Instantaneous focal animal samples were collected on focal females every 10 minutes. The female’s activities (including feeding, traveling, resting, and grooming), the identity of her nearest neighbor, the distance to her nearest neighbor, the activity of the nearest neighbor, and any other individuals within 5m of her were recorded. Additionally, all occurrence data were collected for all mating behaviors for the focal females being targeted each

72 day. The anogenital region of each focal female was characterized daily based on Starin’s

(1991) classifications: inflating, fully swollen, deflating, and flat. A total of 27 focal females (15 logged, 12 old-growth) were observed over approximately 4600 hours of observation time.

Copulation rates were calculated by dividing the number of copulations by the number of observation days for each female. Results were standardized by the number of adult males per group and log transformed. Females in logged areas were then compared to females in old- growth areas using an ANCOVA. Sexual behaviors (copulations, genital presentations to males, genital inspections by males, and grooming males) were tallied during each phase of genital swelling: inflating, fully inflated, deflating, and flat. The number of behaviors seen in each phase was then divided by the total number of mating behaviors seen in the female to calculate what percentage of mating behaviors occurred during each genital swelling phase. The frequency of mating behaviors during each phase was then compared between logged and old- growth areas using a Wilcoxon Rank-Sum test. The number of days that female were inflating, deflating, fully inflated, and flat were recorded and the lengths of each of these phases was compared using a T-test. All statistical tests were run in SAS and significance values were set at

0.05.

Results

Sexual swelling lengths differed between females living in logged and old-growth areas of the forest (Figure 7). Females in logged areas had significantly shorter durations of maximally tumescence (p=0.0468, t=-1.9, DF = 8) compared to females in old-growth areas.

Females in logged and old-growth areas did not differ significantly in the amount of time they were inflating (p=0.8099; t=0.25; DF=12), deflating (p=0.8701; t=0.17; DF=4), and flat

(p=0.3742; t=-1.00; DF=4).

73 In both forest conditions, copulations occurred through the year and sexual behaviors most frequently occurred during maximal swelling. However, females in logged areas copulated less frequently overall and were less variable in their copulation frequencies than females in old- growth areas (t=-3.64, p=0.0043, x logged =0.70, stdev logged =0.29, x old-growth =2.47, stdev old- growth =1.60; Figure 8). However, on days when females did copulate, logged area females copulated at higher frequencies compared to females in old-growth areas. The rate of copulations per day when copulations were observed was 1.9 for females in logged areas and 1.5 for females in old-growth areas. Furthermore, the majority (73%) of sexual behaviors occurred when logged females were fully swollen. In comparison, only 46% of sexual behaviors of females in old-growth areas occurred during maximal swelling (Z=-2.45, n=11 logged ; stdev logged

=0.37, n=11 umlogged ; stdev umlogged = 0.36, p= 0.0117). However, mating behaviors during periods of deflating were higher for females in old-growth areas than for females in logged areas

(Z=1.77; p=0.0458; n=22; stdev =0.26). Furthermore, there was a trend (Z=1.62; p=0.0600: n=22; stdev =0.39) suggesting that females in old-growth areas have a higher percentage of mating behaviors than females in logged areas when genital swellings were inflating. Females in logged and old-growth areas showed no difference in frequency of mating behaviors when flat

(no tumescence) (Z=0.8164, p=0.2117; Figure 9). Neither females in logged or old-growth areas used copulation calls, as has been documented in other populations of red colobus monkeys (i.e.

Starin 1991).

Discussion

This study provides evidence for plasticity in primate reproductive strategies in the face of ecological constraints and the first systematic test of the cost of sexual attraction hypothesis.

Female red colobus in logged areas had a significantly shorter duration of maximal swelling

74 compared to females in old-growth areas. Females in logged areas also copulated significantly less frequently than females in old-growth areas. However, mating bouts in logged areas occurred more often during periods of maximal swelling than they did in old-growth areas.

Consequently, although females in the logged areas mated less overall, during periods of maximal swelling, they mated more frequently than females in the old-growth areas. Females in old-growth areas tended to mate during periods of inflating and deflating more often than females in the logged areas, and thus were less constrained than logged females who limited their mating to maximal tumescence.

This pattern supports Wrangham’s (2002) cost-of-sexual-attraction hypothesis, in which females minimize their energy expenditure by reducing sexual attractivity and limiting copulation to conception periods. This strategy may be key to reproducing in degraded habitats.

An example of a female from the logged area that conceived (Figure 10) illustrates how females in logged ares focus mating bouts during ovulatory periods after a peak in estradiol when genital swellings are fully tumescent. The logged female example shows that she mated multiple times at this specific time period; however, not at other points of her cycle. The example female from the old-growth area, on the other hand, copulated during her periovulatory period and on other days outside of this period (Figure 10).

Behavioral flexibility and phenotypic plasticity (the differential expression of a trait in response to environmental influences (Schlichting & Pigliucci, 1998)) are particularly important when faced with changing environmental conditions (Pigliucci 2001). The environment impacts the expression of various possible phenotypic outcomes influencing individual physiology and behavior, allowing animals to cope with the energetic constraints imposed by changes to their

75 habitat. These behavioral and physiological changes may be sufficient to maintain a positive energy balance and reproductive hormone concentrations needed for reproductive function.

Evidence for plasticity in primate populations is relevant for understanding variation within the Primate Order and primate evolution. Behavioral plasticity was originally conceptualized as a lack of specialization and was associated with generalists (Klopfer &

MacArthur 1960); however, growing evidence supports that plasticity is a strategy for both specialists and generalists (Greenberg 1990; Morse 1980). Primates are often categorized as specialized for their environments; however, the existence of a species across multiple habitat types and evidence for their behavioral plasticity suggests that primates have a tolerance for and an adaptive potential to live in various ecosystems (Jones 2005; Strier 2011) and respond to changing environments (Morse 1980). The interaction between the genotype and environment likely results in different phenotypes in these various habitats (i.e. phenotypic plasticity).

While phenotypic plasticity may aid he patterns observed in this study may have repercussions for females in logged vs. old-growth areas. Reproduction is costly for females, especially for primates where females invest considerable time and energy into successfully rearing offspring (Stumpf et al. 2010). Female primates utilize a variety of reproductive strategies to cope with both social and ecological challenges, including infanticide avoidance

(Hrdy 1979, van Noordwijk & van Schaik 2000; Wolff & MacDonald 2004; Heistermann et al.

2001), cryptic female choice (van Noordwijk & van Schaik 2000; Wolff & MacDonald 2004), indirect benefits, such as friendship (Palombit 2009), and combinations of each (Stumpf and

Boesch 2005). If females in old-growth areas are less constrained by the duration of estrus, they can utilize strategies for paternal confusion and infanticide avoidance more flexibly, including mating outside of periovulatory periods. However, under adverse ecological conditions, females

76 may not have the energetic surplus necessary for a flexible mating strategy and thus have to hedge their bets via more temporally constrained mating. For example, females more constrained by the costs of grouping in logged areas mainly mated during peri-ovulation and thus may not be able to employ mixed reproductive strategies as extensively. Long-term fitness consequences could result for females in degraded areas if ecological stressors constrain a female’s ability to flexibly utilize these strategies.

One caveat to these findings is that groups in the logged areas were not studied both before and after logging occurred. Therefore, it is difficult to say conclusively that behavioral and physiological differences are the result of logging pressures. Natural habitat variation within

Kibale National Park may also affect the differences seen in this study. However, previous studies of feeding behaviors and food availability suggest that logging may be the key component driving the behavioral differences between the research groups. Increases in feeding time are positively correlated with eating lower quality foods (Menon & Poirier 1996;

Onderdonk & Chapman 2000). Also, logging records show that preferred red colobus foods were key targets during logging operations (Skorupa, 1988; Struhsaker, 1997), leaving red colobus in these areas with diminished access to these resources and lower quality foods.

Therefore, it is likely that anthropogenic changes to these habitats are a major factor influencing the observed differences.

Regardless of the cause, phenotypic flexibility in itself illustrates the ways in which female red colobus adjust to living in different quality habitats and indicates the adaptive potential of this species. Jones (2005; p. 138) argues that “the worldwide biodiversity crisis increases the significance of research on all aspects of behavioral flexibility, plasticity, and habitat disturbance.” The results presented here support previous studies (e.g. Marsh 2003),

77 which have found that behavioral and dietary flexibility are important to the survival of wild primate populations in degraded habitats. By examining behavioral responses to degraded habitats, this study provides important insight into conservation strategies, particularly the importance of protecting areas, even those subject to prior anthropogenic disturbance, as animals may hold the potential to adjust to the ecological challenges of these areas. More broadly, the red colobus also provide important insight into understanding how females may have adjusted to environmental change throughout evolution. Humans, in particular, evolved during periods of extreme environmental change, and our dietary and behavioral flexibility may have been key to maintaining reproductive function in our own evolutionary history.

78 Work Cited

Altmann J and Alberts SC. 2003. Variability in reproductive success viewed from a life-history perspective in baboons. American Journal of Human Biology 15:401-409.

Altmann J and Muruthi P. 1988. Differences in daily life between semi-provisioned and wild- feeding baboons. American Journal of Primatology 15: 213–221.

Bercovitch FB. 2001. Reproductive ecology of old world monkeys. In: Ellison PT (ed.) Reproductive Ecology and Human Evolution . Aldine de Gruyter, New York: 369-396.

Berger J, Ward Testa J, Roffe T, Monfort SL. 1999. Conservation Endocrinology: a noninvasive tool to understand relationships between carnivore colonization and ecological carrying capacity. Conservation Biology 13 (5): 980-989.

Boyd IL. 2000. State-dependent fertility in pinnipeds: contrasting capital and income breeders. Functional Ecology 14: 623-630.

Boyle, S.A., Smith, A.T., 2010. Behavioral modifications in northern bearded saki monkeys (Chiropotes satanas chiropotes) in forest fragments of central Amazonia. Primates 51, 43–51.

Chapman CA and Chapman LJ. 1997. Forest regeneration in logged and unlogged forests of Kibale National Park, Uganda. Biotropica 29: 386-412.

Chapman CA and Chapman LJ. 2004. Unfavorable successional pathways and the conservation value of logged tropical forest. Biodiversity and Conservation 13: 2089-2105.

Chapman CA, Chapman LJ, Wrangham R, Isabirye G, and Ben-David K. 1997. Spatial and temporal variability in the structure of a tropical forest. African Journal of Ecology 35 (4): 287–302.

Chapman CA and Lambert JE. 2000. Habitat alteration and the conservation of African primates: A case study of Kibale National Park, Uganda. American Journal of Primatology 50: 169-186.

Chapman CA, Wasserman MD, and Gillespie TR. 2006. Behavioral patterns of colobus in logged and unlogged forests: The conservation value of harvested forests. In: Newton- Fisher NE, Notman H, Paterson JD, and Reynolds V (eds.) Primates of Western Uganda . Springer, New York: 373-390.

Chapman CA, Wasserman MD, Gillespie TR, Speirs ML, Lawes MJ, Saj TL, and Ziegler TE. 2006. Do food availability, parasitism, and stress have synergistic effects on red colobus populations living in forest fragments? American Journal of Physical Anthropology 131: 525-534 .

79 Chapman CA, Struhsaker TT, Skorupa JP, Snaith TV, and Rothman JM. 2010. Understanding long-term primate community dynamics: implications of forest change. Ecological Applications 20 (1): 179-191.

Deschner T and Boesch C. 2007. Can the patterns of sexual swelling cycles in female Tai chimpanzees be explained by the cost-of-sexual-attraction hypothesis? International Journal of Primatology 28: 389-406.

Eley RM, Strum SC, Muchemi G, and Reid GDF. 1989. Nutrition, body condition, activity patterns and parasitism of free-ranging baboons ( Papio anubis ) in Kenya. American Journal of Primatology 18: 209–219.

Ellison PT. 1990. Human ovarian function and reproductive ecology: New hypotheses. American Anthropologist 92 (4): 933-952.

Estrada A., Anzures A.D. and Coates-Estrada R. 1999. Tropical rain forest fragmentation, howler monkeys ( Alouatta palliata ) and dung beetles at Los Tuxtlas, Mexico. American Journal of Primatology 48: 253–262.

Estrada A and Coates-Estrada R. 1996. Tropical rain forest fragmentation and wild populations of primates at Los Tuxtlas. International Journal of Primatology 5: 759-783.

Emery Thompson M, Kahlenberg SM, Gilby IC, and Wrangham RW. 2007. Core area quality is associated with variance in reproductive success among female chimpanzees at Kibale National Park. Animal Behaviour 73: 501-512.

Emery Thompson M and Wrangham RW. 2008. Diet and reproductive function in wild female chimpanzees ( Pan troglodytes schweinfurthii ) at Kibale National Park, Uganda. American Journal of Physical Anthropology 135: 171-181.

Foley CAH, Papageorge S, and Wasser SK. 2001. Non-invasive stress and reproductive measures of social and ecological pressures in free-ranging African elephants ( Loxodonta african a). Conservation Biology 15: 1134-1142.

Gillespie TR and Chapman CA. 2001. Determinants of group size in the red colobus monkey (Procolobus badius ): an evaluation of the generality of the ecological-constraints model. Behavioral Ecology and Sociobiology 50: 329-338.

Gillespie TR and Chapman CA. 2006. Prediction of parasite infection dynamics in primate metapopulations based on attributes of forest fragmentation. Conservation Biology 20 (2): 441-448.

Greenberg R. 1990. Ecological plasticity, neophobia, and resource use in birds. Studies in Avian Biology 13: 431–437

80 Hayssen V, Van Tienhoven A, & Van Tienhoven A. 1993. Asdell's patterns of mammalian reproduction: a compendium of species-specific data . Ithaca, NY: Comstock Publishing Associates.

Hrdy SB. 1979. Infanticide among animals: a review, classification, and examination of the implications for the reproductive strategies of females. Ethol Sociobiol 1:13-40.

Hrdy SB & Whitten PL. 1987. Patterning of sexual activity. In Primate Societies (Smuts BB, Cheney DL, Seyfarth RM, Wrangham RW, & Struhsaker TT, eds). University of Chicago Press, Chicago. Pp. 370-384.

Isbell, L.A. & Young, T.P. ( 1 9 9 3 ) . Human presence reduces leopard predation in a population of vervet monkeys Cercopithecus aethiops. Anon. Behav ., 45, 1233-5.

Iwamoto T & Dunbar RIM. 1983. Thermoregulation, habitat quality and the behavioural ecology of gelada baboons. Journal of Animal Ecology 52: 357-366.

Jones CB. 2005. Behavioral flexibility in primates: causes and consequences . Springer, New York.

Kasenene JM. 1987. The influence of mechanized selective logging, felling intensity and gap- size on the regeneration of a tropical moist forest in the Kibale Forest Reserve, Uganda. Ph.D. Dissertation. Michigan State University, Michigan.

Korstjens AH. 2001. The Mob, the Secret Sorority, and the Phantoms: An analysis of the socio-ecological strategies of the three colobines of Ta ї. PhD thesis, Utrecht University.

Knott CD. 1999. Reproductive, physiological and behavioral responses of orangutans in Borneo to fluctuations in food availability. PhD dissertation. Harvard University.

Klopfer PH and MacArthur RH . 1960 Niche size and faunal diversity American Naturalist 94: 293-300.

Lewis RJ and Kappeler PM. 2005. Are Kirindy Sifaka capital or income breeders? It depends. American Journal of Primatology 67:365-369.

Lipson SF. 2001. Metabolism, Maturation, and Ovarian Function. In: Ellison PT (ed.) Reproductive Ecology and Human Evolution . Aldine de Gruyter, New York: 235-248.

Loy, J. 1987. The sexual behavior of African monkeys and the question of estrus. E. Zucker, ed. Monographs in Primatology , Vol. 10. New York: Alan R. Liss, Inc.

Marsh, L. K. 2003. Primates in fragments: ecology and conservation . Kluwer Academic/Plenum Publishers, New York.

81 Martin P and Bateson P. 1986. Measuring behaviour: An introductory guide . Cambridge University Press, Cambridge.

Miller KE, Bales KL, Ramos JH, and Dietz JM. 2006. Energy intake, energy expenditure, and reproductive costs of female wild golden lion tamarins ( Leontopithecus rosalia ). American Journal of Primatology 68: 1037-1058.

Menon S, Poirier FE. 1996. Lion-tailed macaques ( Macaca silenus ) in a disturbed forest fragment: activity patterns and time budget. International Journal Primatology 17:969– 98.

Milton K. 1980. The foraging strategy of howler monkeys: a study in primate economics . Columbia University Press, New York.

Morse DH. 1980. Behavioral mechanisms in ecology. Harvard University Press, Cambridge.

Oates JF. 1994. The natural history of African colobines. In: Davies AG and Oates JF (eds.) Colobine Monkeys: Their Ecology, Behavior, and Evolution . Cambridge University Press, Cambridge: 75-128.

Oates JF. 1999. Myth and reality in the rainforest: how conservation strategies are failing in West Africa. University of California Press, Berkeley.

Onderdonk DA and Chapman CA. 2000. Coping with forest fragmentation: The primates of Kibale National Park, Uganda. International Journal of Primatology 21: 587-611.

Palombit R, Cheney DL, Fischer J, Johnson S, Rendall D, Seyfarth RM, Silk JB. 2000. Male infanticide and defense of infants in chacma baboons. In: van Schaik CP, Janson CH, editors. Infanticide by males and its implications. Cambridge: Cambridge University Press. p 123-153.

Palombit RA. 2009. ‘‘Friendship’’ with males: a female counterstrategy to infanticide in Chacma baboons of the Okavango Delta. In: Muller MN, Wrangham RW, editors. Sexual coercion in primates and humans . Cambridge: Harvard University Press: 377–409.

Pigliucci M. 2001. Phenotypic Plasticity: Beyond Nature and Nurture. Johns Hopkins University Press, Baltimore.

Ratsimbazafy J. 2007. Diet Composition, foraging, and feeding behavior in relation to habitat disturbance: implications for the adaptability of ruffed lemurs ( Varecia v. editorium ) in Manombo forest, Madagascar. In Gould L and Sauther ML (eds) Lemurs: Ecology and Adaptation . Springer, New York: 403-422.

Rodríguez-Luna E, Domínguez-Domínguez LE, Morales-Mávil JE, Martínez-Morales M (2003). Foraging strategy changes in a Alouatta palliata mexicana troop released on an

82 island. In Primates in Fragments: Ecology and Conservation (Marsh LK, ed.), pp 229– 250. New York, Kluwer Academic/Plenum Publishers.

Ross, C. & Jones, K. E. 1999. Socioecology and the evolution of primate reproductive rates. In: Comparative Primate Socioecology (Ed. by P. C. Lee), pp. 73e110. Cambridge: Cambridge University Press.

Schlichting CD and Pigliucci M. 1998. Phenotypic Evolution: A Reaction Norm Perspective . Sinaur Associates, Sunderland.

Sherry DS. 2002. Reproductive seasonality in chimpanzees and humans: Ultimate and proximate factors. PhD dissertation. Harvard University.

Silver SC and Marsh LK. 2003. Dietary flexibility, behavioral plasticity, and survival in fragments: Lessons from translocated howlers. In Marsh LK (ed.) Primates in fragments: ecology and conservation . Kluwer Academic/Plenum Publishing, New York: 251-265.

Silva, S. S. B., & Ferrari, S. F. (2009). Behavior patterns of southern bearded sakis Chiropotes satanas) in the fragmented landscape of eastern Brazilian Amazonia. American Journal of Primatology , 71, 1–7.

Singh M and Vinathe S. 1990. Inter-population differences in the time budgets of bonnet monkeys ( Macaca radiata ). Primates 31(4): 589-596.

Skorupa JP. 1988. The effect of selective timber harvesting on rain forest primates in Kibale Forest, Uganda. PhD Thesis. University of California, Davis.

Smuts BB, Smuts RW. 1993. Male aggression and sexual coercion of females in nonhuman primates and other mammals: evidence and theoretical implications. Adv Study Behav 22:1-63.

Struhsaker TT. 1978. Food habits of five monkey species in the Kibale Forest, Uganda. In Chivers, D. J., and Herbert, J. (eds.), Recent Advances in Primatology, Vol. 1, Behaviour, Academic Press, New York, pp. 225–248.

Struhsaker TT. 1997. Ecology of an African Rain Forest: Logging in Kibale and the conflict between conservation and exploitation. University Press of Florida, Gainesville.

Struhsaker TT. 2010. The Red Colobus Monkeys . Oxford University Press, Oxford.

Struhsaker TT. 1975. The Red Colobus Monkey . University of Chicago Press, Chicago.

Struhsaker TT and Leland L. 1979. Socioecology of five sympatric monkey species in the Kibale Forest, Uganda. Advances in the Study of Behavior 9: 159-228.

83 Struhsaker TT and Leland L. 1985. Infanticide in a patrilineal society of red colobus monkeys. Zeitschrift fuer Tierpsychologie 69(2): 89-132.

Struhsaker TT and Leland L. 1987. Colobines: infanticide by adult males. In: Smuts BB, Cheney DL, Seyfarth RM, Wrangham RW, and Struhsaker TT (eds.) Primate Societies . University of Chicago Press, Chicago: 83-97.

Starin ED. 1991. Socioecology of the red colobus monkey in The Gambia with particular reference to female-male differences and transfer patterns. PhD dissertation. The City University of New York.

Snaith TV and Chapman CA. 2008. Red colobus monkeys display alternative behavioral responses to the costs of scramble competition. Behavioral Ecology 19: 1289-1296.

Stampone MD, Hartter J, Chapman CA, and Ryan SJ. 2011. Trends and variability in localized precipitation around Kibale National Park, Uganda, Africa. Research Journal of Environmental and Earth Sciences 3 (1): 238-253.

Strier KB. 2011. Distinguished Primatologist Lecture. Midwest Primate Interest Group. Kent, OH.

Stumpf RM, Boesch C. 2005. Does promiscuous mating preclude female choice? Female sexual strategies and mate preference in chimpanzees of the Taï Forest, Côte d'Ivoire. Behav Ecol Sociobiol 57:511-524.

Stumpf RM, Martinez-Mota R, Milich KM, Righini N, and Shattuck MR. Sexual Conflict in Primates. Evolutionary Anthropology 20: 62-75.

Tardif SD, Ziegler TE, Power M, and Layne DG. 2005. Endocrine changes in full-term pregnancies and pregnancy loss due to energy restriction in the common marmoset (Callithrix jacchus ). The Journal of Clinical Endocrinology and Metabolism 90 (1): 335- 339. van Noordwijk MA and van Schaik CP. 2000. Reproductive patterns in eutherian mammals: adaptations against infanticide. Infanticide by males and its implications. Cambridge: Cambridge University Press: 322-360.

Wade GN and Schneider JE. 1992. Metabolic fuels and reproduction in female mammals. Neuroscience and Behavioral Reviews 16: 235-272.

Watts DP. 1988. Environmental influences on mountain gorilla time budgets. American Journal of Primatology 15: 295–312.

Wolff JO and Macdonald DW. 2004. Promiscuous females protect their offspring. Trends in Ecology and Evolution 19: 127–134.

84 Wrangham, RW. 2002. The cost of sexual attraction: is there a trade-off in female Pan between sex appeal and received coercion? In Boesch C, Hohmann G, and Marchant LF (eds.) Behavioural Diversity in Chimpanzees and Bonobos. University Press, Cambridge: 204- 215.

Wong S. N. P. and Sicotte P. (2007). Activity budget and ranging patterns of Colobus vellerosus in forest fragments in Central Ghana. Folia Primatol. 78: 245–254.

85 Table 10: Demographic information for red colobus at Kibale National Park, Uganda and Abuko Nature Reserve, The Gambia (1Struhsaker and Leland 1987; 2 Starin 1991; 3 Struhsaker 1975). Trait Kibale National Park Abuko Nature Reserve Interbirth interval 25.5 ± 5.1 months 1 29.4 months 2 Gestation length 4.5 – 5.5 months 3 5.25 months 2 Menstrual cycle length Unknown 27.7 days 2 Age at sexual maturity (males) 35.4 - 58 months 1 > 28 months 2 Age at sexual maturity (females) 38 - 46 months 1 34.25 months 2

Table 11: Group compositions for the four research groups. Logged Logged Old-growth Old-growth Group 1 Group 2 Group 1 Group 2 Adult Males 8 4 9 11 Adult Females 22 14 15 22 Subadults 3 3 2 4 Juveniles 8 10 1 23 Infants 10 6 7 13 Total 51 37 34 73 Infant/Adult Female Ratio 0.45 0.43 0.47 0.59

86

K-13

K-15

K-30

Figure 6: Map of the logging history at the Kanyawara Field Site in Kibale National Park, Uganda (modified from Struhsaker 1975). K-13 was heavily logged and poisoned with an arborcide, K-15 was heavily logged, and K-30 was never commercially logged.

87 MatingLength Patters of Genital vs. Habitat Swellings Quality

70

60

50

40 Logged Old-growthUnlogged 30 Length in Length Days

Duration 20

10

0 Inflating Fully Deflating Flat

Figure 7: Females in logged areas had significantly shorter periods of time being maximally tumescent (p=0.0468, t=-1.9, DF = 8) compared to females in old-growth areas, but not in the amount of time they were inflating (p=0.8099; t=0.25; DF=12), deflating (p=0.8701; t=0.17; DF=4), and flat (p=0.3742; t=-1.00; DF=4).

88

Copulation Rates copulations/observation days/males copulations/observation # copulations/# of observation days of observation # copulations/#

ANCOVA Logged Old-growthUnlogged SAS p=0.0043, t=-3.64

Figure 8: Copulation rates of females living in logged and old-growth areas of Kibale National Park. Females in logged areas copulated less frequently overall and were less variable in their copulation frequencies than females in old-growth areas (t=-3.64, p=0.0043, x logged =0.70, stdev logged =0.29, x old-growth =2.47, stdev old-growth =1.60; Figure 3).

89 Mating Behaviors During Cycle

100

90

80

70

60 Logged 50 Old-growthUnlogged 40

30

20 % Mating % Behaviors/Female

10 Percent of Total Mating Behaviors

0 Inflating Fully inflated Deflating Flat

Figure 9: Mating behaviors in relation to extent of perineal swelling in red colobus females living in logged and old-growth areas of Kibale National Park. Females in logged areas performed mating behaviors significantly more often than females in old-growth areas when fully inflated swelling (Z=-2.45, n=11 logged ; stdev logged =0.37, n=11 umlogged ; stdev umlogged = 0.36, p= 0.0117). However, mating behaviors during periods of deflating were higher for females in old-growth areas than for females in logged areas (Z=1.77; p=0.0458; n=22; stdev =0.26), and there was a trend (Z=1.62; p=0.0600: n=22; stdev =0.39) suggesting that females in old-growth areas have a higher percentage of mating behaviors than females in logged areas when genital swellings were inflating. There was no significance difference in mating when females were flat.

90 Logged Female Fully tumescent

16 2500 14 2000 12

10 1500 P4 8 E2 6 1000 4 500

2 Estradiol dry (pg/g wt) Progesterone dry(ng/g wt) 0 0

9 9 9 9 0 0 09 09 09 - - - - - t-09 -Oct-09-Oct-09Oct-0 Oct-0 3-Oct-096-Oct-099-Oct 8 1 4- 7- 15-Sep18-Sep21-Sep24-Sep27-Sep-0930-Sep-09 12-Oct-0915-Oct-091 2 2 2 30-Oc 2-Nov-095-Nov-098-Nov-09 Birth in early April

Old-growthUnlogged Female Fully tumescent 8 700

7 600 6 500 5 400 P4 4 300 E2 3 200 2

1 100 Estradiol dry (pg/g wt) Progesterone dry (ng/g wt) 0 0

9 9 9 9 9 9 0 0 0 0 0 0 09 09 10 10 v-0 v-0 c-0 c-0 c-0 c-0 -1 -1 -1 -1 -1 -1 - an an an eb- -Dec- -Dec- -Dec-09 -Dec-09 -Dec-09 -J -J -F 6-No9-No2 5-Dec-098 4 0 3-De6 9-De 1 4-J 7 0-Jan3-Jan-106-Jan9-Jan-102-Jan5-Jan-108-Jan1-Jan-103 2 2 11-De1 17-De2 2 2 2 1 1 1 1 2 2 2 3 Birth in late June 2010

Figure 10: An example of a female in the logged area (top) that restricts mating to periovulatory periods after an increase in estradiol and maximum genital tumescence, and an example of a female in the old-growth area (bottom) where the female mates during the periovulatory period and at other points in the cycle.

91 CHAPTER 4

FEMALE RED COLOBUS MONKEYS MAINTAIN REPRODUCTIVE FUNCTION IN DEGRADED HABITATS

Abstract

This study examines the relationship between ecological conditions and reproductive function in female red colobus monkeys ( Procolobus rufomitratus ) in Kibale National Park,

Uganda. Whereas many non-human primate studies suggest that ecological stressors are associated wtith decreased female reproductive success, studies on humans indicate that females living under different ecological conditions can adapt to those conditions and adjust their reproduction accordingly (Borgerhoff Mulder, 1992; Vitzthum, 2001). In this study, I test differences in reproductive hormones, ketones, and seasonality between female red colobus monkeys living in logged and old-growth areas. The red colobus monkeys of Kibale National

Park are important subjects for investigating primate responses to habitat disturbances because they have had to adjust to forest that was severely logged approximately 40 years ago.. Here, I examine female red colobus monkeys in logged and old-growth areas of Kibale National Park,

Uganda to test for differences in reproductive hormone concentrations and ketone levels. Fecal and urine samples from females (n = 20) were analyzed for reproductive hormone concentrations and ketone levels. Results indicate that estradiol and progesterone concentrations and ketone levels are similar between habitats and do not show seasonal differences. Female dietary, activity, and mating adjustments may explain the ability of females to maintain reproductive function in heavily logged areas.

92 Introduction

Female mammals need a positive energy balance (the relationship between nutritional intake and energy expenditure) to conceive, maintain a pregnancy, and support lactation (Miller et al.

2006; Sherry 2002; Tardif et al. 2005; Wade & Schneider 1992). Previous studies indicate that habitat quality is positively correlated with reproductive fitness (Berger et al. 1999; Emery

Thompson et al. 2007; Knott 1999; Lipson 2001; Ross & Jones 1999; Tardif et al. 2005).

Habitat disturbance can have a negative influence on food availability, which can impact energy balance for individuals living in these areas and can ultimately affect reproduction. Animals living in degraded habitats may be unable to maintain the energy balance needed to produce sufficient hormone concentrations at each stage, thus compromising reproduction. The negative impact of ecological stressors on reproductive hormone concentrations has been documented in a number of species including elephants (Foley et al. 2001), old world monkeys (Bercovitch 2001), and chimpanzees (Emery Thompson et al. 2007).

Trivers (1972) and Wrangham (1979) argue that food is the primary limitation on reproductive success in female mammals. Ecological stressors (e.g. logging and climate change) are particularly hazardous for female reproduction because of the high energetic costs associated with pregnancy and lactation (Bercovitch 2001; Muller & Wrangham 2005). Chronic energetic stress leads to decreased reproductive success over a lifetime via slower growth and maturation, longer interbirth intervals, and lower levels of ovarian hormones (Lipson 2001).

Estrogen and progestin concentrations regulate female reproductive cycles and influence the timing of ovulation. Estrogen peaks at the end of the follicular phase, signaling impending ovulation. During the luteal phase, progesterone levels rise to prepare the uterus for implantation

(Campbell et al. 1999). Adequate nutrition availability is necessary for triggering the hormonal

93 feedback loop that signals cycling to begin (Ellison 1990; Strier & Ziegler 2005). Without sufficient hormone concentrations, female reproductive function is inhibited (Ellison 1990;

Emery Thompson et al. 2007; Strier & Ziegler 2005). For example, Strier and Ziegler (2005) found that female northern muriquis ( Brachyteles hypoxanthus ) required a minimum estradiol threshold to resume cycling postpartum.

Ellison (1990) argues that a positive energy balance has a significant impact on ovarian function in women. A slight increase in body weight corresponds with higher levels of ovarian hormones and a higher rate of conception (Ellison 1990). This pattern has also been seen in female orangutans, where weight gain is associated with increased urinary hormone concentrations (Masters & Markham 1991). Conversely, estrone conjugate values were lower in orangutan females during periods of food shortages (Knott 1999). In orangutans, a fluctuating energy balance was associated with suppressed ovarian function and decreased fecundity (Knott

2001). Ziegler et al. (2000) also found that seasonality in ovulatory function in female Hanuman langurs ( Presbytis entellus ) was correlated with ecological factors, including food availability.

Moreover, female chimpanzees living in areas of Kibale National Park with better food sources had significantly higher levels of urinary oestrogen and progesterone conjugates than females living in lower quality areas (Emery Thompson et al. 2007). These higher levels of reproductive hormones corresponded to shorter interbirth intervals and higher infant survival (Emery

Thompson et al. 2007).

Ketones are another physiological marker that can be used as an indicator of ecological stress and to understand better the relationship between habitat and reproduction. Ketones are produced by the body when nutritional intake is low and then excreted into the urine (McGarry

& Foster 1980; Robinson & Williamson 1980). Specifically, fatty acids are converted to ketone

94 bodies in the livers of “starved” or food stressed individuals (Robinson & Williamson 1980).

Knott (1999) found that ketones were only present in orangutan urine during severe fruit shortages when fat reserves were being metabolized to produce energy.

In addition to physiological changes, variation in mating behaviors and ovarian cycles may be more restricted among animals in lower quality habitats. For example, behavioral research suggests that red colobus monkeys of Kibale National Park do not have a birth season, but show a pattern of bimodal birth peaks during the rainy months (Struhsaker & Leland 1987).

Struhsaker (1975) also found a lack of seasonality in food availability in this study population.

In contrast, in other regions with different environmental conditions (i.e. increased predation and mosaic land cover), such as the Abuko Nature Reserve in The Gambia, red colobus monkeys have a strict birth season (Starin 1991) and almost never copulate during the periods of lowest rainfall and temperatures (Starin 1991). Thus, individuals living in lower quality habitats may have restricted breeding seasons that are influenced by periods of highest food availability.

Empirical tests of the relationship between habitat quality and ovarian function are lacking and necessary for understanding how animals respond to increasing habitat disturbance.

This study examines the relationship between ecological conditions and reproductive function

(as indicated by estradiol and progesterone concentrations) and energetic status (as indicated by ketone levels) in 4 groups of female red colobus monkeys ( Procolobus rufomitratus ) living in logged and and old-growth forest in Kibale National Park, Uganda. Specifically, I tested three predictions: 1) females in logged areas have lower estradiol and progesterone concentrations than females in old-growth areas; 2) ketone levels, a physiological indicator of ecological stress, are significantly higher for females in logged areas than in old-growth areas; and 3) females in

95 logged areas show more seasonality in reproductive behaviors and birth peaks than females in old-growth areas.

The alternative hypotheses for these predictions are that females in logged and old- growth areas 1) do not have different fecal reproductive hormone concentrations, 2) do not have different urinary ketone levels, and 3) do not differ in level of seasonality. If females do not differ in these various parameters, it would suggest that females in logged areas are able to maintain reproduction under these conditions. This ability may be the result of adjusting their diet and energy expenditure (through behavioral changes) to maintain the positive energy balance needed for reproductive function. Behavioral plasticity may be especially important for individuals dealing with energetic constraints in different environments (Marsh 2003, Rodriquez-

Luna et al. 2003; Silver & Marsh 2003; Estrada & Coates-Estrada 1996). Behavioral flexibility and phenotypic plasticity (the differential expression of a trait in response to environmental influences (Schlichting & Pigliucci, 1998)) are particularly important when faced with changing environmental conditions (Pigliucci 2001).

Methods

Site and subjects

Research was conducted in the forest surrounding the Makerere University Biological

Field Station in Kanyawara, Kibale National Park, Uganda (0°13'- 0°41'N, 30°19'- 30°32'E), a

795 km 2protected area located in western Uganda just north of the equator (Oates 1994). Kibale

National Park is a mid-altitude, moist evergreen forest, which receives approximately 1741 mm of rain during two annual rainy seasons (Chapman & Lambert 2000; Chapman et al. 1997;

Struhsaker 1975). Kibale is home to 13 species of primates including 1 species of ape, 3 prosimians, and 9 monkeys (Onderdonk & Chapman 2000).

96 Kibale National Park, Uganda, provides an effective natural laboratory in which to test the relationship between forest disturbance and primate hormone concentrations. Kibale contains varying levels of deforestation throughout the park and the surrounding land, including areas that have been previously logged and areas that have never been logged (Chapman and Chapman

2004). Moreover, the ecology of Kibale has been closely documented over several decades (e.g.

Chapman and Chapman 2004).

Red colobus monkeys are ideal subjects for research on the impact of habitat on reproductive function for several reasons. Multiple groups can be studied within a relatively small geographic area encompassing a range of ecological conditions, including logged and old- growth habitats. Red colobus monkeys are highly endangered (Struhsaker & Leland 1987) and are clearly affected by ecological stressors (Chapman et al. 2006, 2007; Gillespie & Chapman

2006). For example, between 2000 and 2003, Chapman et al. (2006) found that the population of red colobus monkeys living in forest fragments (small forest patches experiencing ongoing disturbances) decreased 20% (from 127 to 102 individuals).

One reason for the documented decline in red colobus monkeys in the Kibale forest fragments may be due to poor nutrition and the corresponding decrease in reproductive function.

On average, logging activities led to a 29.5% decrease in the diameter at breast height (DBH) of available food trees in these fragments (Chapman et al. 2006). This decrease in available foods corresponded to a difference in colobus monkey diet. Colobus in the fragments ate more mature leaves (rather than the preferred food source of young leaves) than colobus in unlogged habitats

(Chapman et al. 2004).The ratio of infants to adult female red colobus monkeys decreased from

0.31 to 0.19 in fragments (Chapman et al. 2006), suggesting that these population decreases may be a result of reproductive suppression or higher rates of infant mortality. The impact of habitat

97 disturbance on reproductive function is also supported by changes in reproductive hormone profiles. Females in the forest fragments had significantly lower reproductive hormones than females living in the protected park area (Milich et al. 2008).

Female red colobus have perineal swellings that fluctuate in size over the course of their cycle (Struhsaker 1975; Struhsaker & Leland 1987). These swellings can also be present during pregnancy, making them an unreliable indicator of cycling. Additionally, females mate across their cycle, including when they are not swollen (Struhsaker 2010). Whereas many of the physiological markers associated with ovulation, pregnancy, and lactation have not yet been studied, estimated lengths of reproductive parameters have been suggested from behavioral observations (Table 12).

This study focuses on 4 groups of habituated, individually identified red colobus monkeys – two groups living in logged areas and two groups living in old-growth areas of Kibale

National Park (Table 13). The logged groups ranged between the areas K-15 and K-13 (see map in Figure 11). K-15 was heavily logged from September 1968 to April 1969, resulting in the destruction of approximately 50% of the trees in this compartment (Chapman & Chapman 1997;

Kasenene 1987; Skorupa 1988). K-13 was heavily logged and also poisoned with an arborcide.

The old-growth groups lived in K-30, which has never been commercially harvested. The heavily logged areas have been left with large gaps in the canopy; whereas the unlogged area has more continuous canopy cover (Chapman & Chapman 1997).

This study specifically focuses on adult parous, cycling female red colobus monkeys living in these four research groups. In the months prior to the start of data collection, I noted which females appeared to have infants or juveniles and then selected focal females from that

98 subset that were no longer lactating. This method ensures that females are of a reproductive age and should be cycling.

Behavioral observations and sample collection

Each focal group was followed from 8:00 am to 4:00 pm six days a week. Focal females were followed on a rotating basis in each group. Five focal females were selected in each group and observed on a rotating basis until they gave birth, at which time I would choose a new adult female without an infant to enter the rotation so that at each point in time, data were obtained from 5 females from each group. (Note: Because females changed depending on if they gave birth, the sample size of focal females for each group differs.) Urine and fecal samples were collected opportunistically from focal females in each group and subsequently frozen. The hormone data represent 20 focal females and approximately 400 samples obtained between

August 2009 and May 2010.

Hormone Extraction and Analyses

To test the first hypothesis that reproductive hormone concentrations would differ between habitat types, progesterone and estradiol concentrations were measured . Fecal hormones are active, which means that they are able to enter cells and bind to the steroid receptor to bring about a physiological response. Using fecal samples with highly specific antibodies in a radioimmunoassay (RIA) allowed me to measure estradiol and progesterone concentrations directly instead of their metabolites.

To extract the hormones in the field, fecal samples were thawed, thoroughly mixed, and divided. 0.50g of each fecal sample was dried and then weighed again to obtain a dry weight.

Another 0.50g of each fecal sample was taken for hormone extraction using Alltech Prevail C18

Maxi-Clean Cartridges (300mg). Samples were first weighed, then mixed with 95% ethanol and

99 citrate buffer for 24 hours, centrifuged, and supernatant was passed through the hormone cartridge at a rate of 4ml/min.

At the University of Illinois Animal Sciences Lab, the cartridges were flushed with 2ml of methanol, dried down, and reconstituted with 2mls of PBS-gel. Radioimmunoassays (RIA) were used to analyze concentrations of progesterone and estradiol in all samples. An aliquot of the reconstituted fecal sample was incubated with either 3H-progesterone or 3H-estradiol, PBS-gel, and progesterone or estradiol antibody overnight. The next day, 3H-steroid not bound to the antibody was removed using charcoal-dextron. Each tube was centrifuged so that the supernatant with only the bound hormones could be removed. This supernatant was mixed with scintillation fluid and counted in a scintillation counter to determine the concentration of either estradiol or progesterone in each sample (following Bahr et al. 1983).

Estradiol and progesterone concentrations were calculated and reported as picograms (pg) and nanograms (ng) per grams of dry feces, respectively. Samples were separated into “cycling” and “pregnant” categories based on back calculations from when infants were born. However, this procedure does not account for stillbirths or fetal loss during pregnancy. Luteal phases were determined by changes in progesterone concentrations. The number of luteal phases was then used to calculate the number of cycles each month.

To control for uneven sampling and expected fluctuations in hormones over the reproductive cycle, I calculated a z-score for progesterone and estradiol concentrations (following Emery

Thompson 2005, Emery Thompson et al. 2007). An average z-score was also then calculated for each female. This method is appropriate for dealing with opportunistically collected samples that result in an irregular distribution of samples for each female over time (Emery Thompson et

100 al. 2007; Emery Thompson 2005). Z-scores were compared using both a One Way ANOVA and a Two Way ANOVA in SPSS. All significance levels were set at 0.05.

Ketone measurements

To test the second hypothesis that ketone levels are significantly higher for females in logged areas than in old-growth areas, urinary ketone levels were measured. In the field, urine was pipetted onto urinalysis strips (Boehringer Mannheim Chemstrip 10 w/ SG) to measure ketone bodies. Ketone results were recorded as positive or negative based on color. A negative ketone result indicates that an animal is not nutritionally stressed, whereas a positive ketone result suggests the opposite. Differences in ketone levels for each female were compared using a

Fisher’s Exact Test in SPSS.

Seasonality

To test whether females in logged areas show more seasonality in reproductive behaviors and birth peaks than females in old-growth areas, the dates of copulations and the birth of new infants were compared to assess variation in reproductive behaviors and birth peaks (an indicator of seasonality) between logged areas and old-growth areas.

Results

Hypothesis 1: females in logged areas have lower estradiol and progesterone concentrations than females in old-growth areas. Females red colobus monkeys show considerable inter-individual variation in their reproductive hormone concentrations.

Representative date from red colobus monkeys are presented in Figures 12-15. Females in both logged and old-growth areas show changes in estradiol and progesterone concentrations over the course of the reproductive cycle and during pregnancy. For females that conceived, there was a peak in estradiol prior to the estimated date of conception. The estradiol peak was usually

101 followed by an elevation in progesterone concentrations. Birth was associated with a decrease in both hormones. Prior to conception, females in both logged and old-growth areas had one luteal phase or cycle each month.

Contrary to the prediction, females in logged areas did not have significantly lower reproductive hormone concentrations compared to females in old-growth areas (Figures 16 &

17). Female hormone concentrations varied considerably both within and between individuals in each habitat. When comparing the average z-score of individual females between logged and old-growth areas with a One-way ANOVA, there were no significant differences between females living in logged areas and females living in old-growth areas for progesterone (Figure

18; p=0.58; n=20; F= 0.317; df=1) or estradiol (Figure 19; p=0.97; n=20; F=3.074; df=1).

Because there was only a sample size of 20 individuals, I also ran a One-way ANOVA to test the differences for all samples for each female, instead of using an average. (Because the samples are not independent, this method is not equivalent to having a larger sample size of individual females; however, it can provide some insight into how a larger sample size may impact the results.) Again, progesterone concentrations were not significantly different between females in the logged areas and females in the old-growth areas (Figure 20; p=0.102; n=300; F=2.698; df=1); however, estradiol was significantly higher in the logged area than in the old-growth area

(Figure 21; p=0.006; n=296; F=7.522; df=1). Similarly, when I used a Two-way ANOVA to compare samples from logged vs. old-growth areas and separated females that are cycling and pregnant, there was no significant difference in logged and old-growth areas for progesterone concentrations (p=0.135; n=300; F =2.24; df=1), but estradiol concentrations were significantly higher in logged areas than in old-growth areas (p=0.008; n=296; F=7.031; df=1).

102 Hypothesis 2: ketone levels are higher for females in logged areas than in old-growth areas. Contrary to the prediction, ketone levels did not differ between females living in logged and old-growth areas (Figure 22). Results indicated no significant difference in the ketone levels of individual females between logged (N=11) and old-growth areas (N=9) (Fisher’s Exact Test, df = 1, p=0.579).

Hypothesis 3: females in logged areas show more seasonality in reproductive behaviors and birth peaks than females in old-growth areas. There is no clear seasonality in this species.

Females in logged areas and old-growth areas copulated during every month of the year and did not show any seasonal variation when mating behaviors occurred. Because primates mate for reasons other than reproduction, potential seasonal differences in conception and birth may be more prevalent. However, females also conceived and gave birth throughout the year and showed no birth peak in either the logged or old-growth areas. Thus, the mating behaviors and birth records indicate that there is no seasonal pattern for Kibale red colobus in either forest type.

Though more years of birth records are needed to make an accurate assessment of birthrate, the preliminary results from this study suggest that birth rates do not differ between logged and old- growth areas. Females in logged areas had a birthrate of 64.3% and females in old-growth areas had a birthrate of 66.7%.

Discussion

The goal of this study was to examine differences in reproductive function (as indicated by estradiol and progesterone concentrations), seasonal variation in reproduction (based on behavioral observations and birth records), and energetic status (as indicated by ketone levels) in female red colobus monkeys living in logged and old-growth areas of Kibale National Park.

Results indicated that contrary to predictions, females in logged areas did not have lower

103 estradiol and progesterone concentrations, higher ketones, and increased seasonality in mating and birth than females in old-growth areas,.

Females in logged and old-growth areas have normal reproductive hormone profiles that change as expected throughout the cycle and during pregnancy, including a peak in estradiol prior to ovulation and an elevation of progesterone and estradiol during pregnancy. Estradiol and progesterone decrease drastically at the time of birth. These patterns are consistent with expected patterns of hormonal changes in old world monkeys. Menstrual cycle length, as predicted from luteal phase counts, is consistent with other colobus species and estimated cycle lengths of red colobus monkeys from behavioral observations (see Starin 1991).

As expected, hormone concentrations fluctuate greatly over the course of the cycle and during different reproductive phases. Females in both logged and old-growth areas have considerable inter-individual variation. However, females in logged areas do not show significant decreases in reproductive hormone concentrations, as predicted. When comparing individual female hormone concentrations, there are no differences between individuals in the logged and old-growth; however, when analyzing the individual samples, fecal samples from red colobus monkeys in the logged areas had significantly higher estradiol concentrations than those in the old-growth forest. There was still no difference in progesterone.

It is important to reiterate that this study was comparative in nature, resulting in irregular sample collection for each individual female. Females in the logged area may have higher estradiol concentrations; however, it is not clear if this is an actual pattern or sampling bias.

Estradiol, in particular, may be sensitive to sampling error because of the normal cyclicity of estradiol throughout the female reproductive cycle. Estradiol peaks prior to ovulation, making consistent sampling extremely important to capture this peak. If samples are not collected at the

104 peak, a female’s maximum estradiol value may not reflect the actual maximum estradiol concentration for that female, but a point on the way up or down from the peak. Because of the comparative nature of this study, each female was not sampled on a regular basis and estradiol peaks may have been missed. This issue could be particularly the case for the old-growth area, as logistical problems resulted in a smaller sample size for this area.

Alternatively, potentially lower estradiol concentrations in old-growth areas may be the result of dietary differences. As illustrated in Chapter 2, females ate different plant species in the logged and old-growth areas. These foods as digesta may have different passage time in the gastrointestinal tract of the red colobus monkeys. There are a large number of studies in animals and humans documenting this suggestion (e.g. Goldin et al 1981; Ehle et al 1982). Gut passage time of the digesta influences the concentration of estradiol in blood, urine and feces. In the gut, conjugated estrogens ( also androgens but not progestins) are deconjugated, then adsorbed by the gut epithelium where estrogens are reconjugated and then enter the hepatic portal vessel and are transported to the liver (Figure 23). The rate at which the digesta moves through the digestive tract will determine the efficiency of the deconjugation of the conjugated steroids. If this step is efficient, i.e. digesta moving slowly, then the majority of conjugated estrogens are deconjugated and the majority (98%) of the steroids in the feces are active steroids, e.g. estradiol. However, if the deconjugation is less efficient which could be caused by increased rate of movement of digesta, then there is an increase in conjugated estrogens and a decrease of active estrogens in the feces (Goldin et al 1981; Rouff 1988). For example, Goldin et al. (1981) found fecal estrogen concentrations differed between vegetarian and non-vegetarian women because of differences in digestion rate. This phenomenon does not impact progesterone concentrations because progesterone is not conjugated and recycled as are estrogens, so it is not affected by the rate

105 digesta moves through the gut. Because progesterone does not go through the same conjugation and deconjugation cycle, it is not affected by differences in gut passage time, which helps explain why estradiol results in this study are different, but progesterone results are not.

Recent studies in red colobus monkeys also suggest that some red colobus foods possess phytoestrogens. Wasserman et al. (2012) tested phytoestrogens in common red colobus foods.

Whereas not all plants tested in that study overlapped with plants commonly seen eaten in this study, the only foods that did overlap and tested positive for phytoestrogens were plant species that were commonly eaten in the logged areas and not the old-growth areas (Table 14).

Wasserman et al. ( in press ) also found that an increase in the consumption of plants high in phytoestrogens resulted in a significantly correlated increase in fecal estradiol concentrations in male red colobus monkeys. It may be that the higher estradiol concentrations are not the direct result of eating phytoestrogens, but instead that these foods are also associated with slower gut passage time. Whereas preliminary, these results suggest an alternative explanation for the differences in estradiol concentrations between logged and old-growth groups.

Dietary differences did not correlate with differences in nutritional status, as indicated by urinary ketone concentrations. Females in logged and old-growth areas did not have significant differences in ketone levels. These findings suggest that females in logged areas do not have higher energetic stress than females in old-growth areas.

Females mated consistently over the course of the study and produced babies throughout all parts of the year. These results suggest that female red colobus in logged and old-growth areas of Kibale National Park do not show a seasonal reproductive pattern as is seen at other sites

(i.e. Starin 1991) or suggested by previous studies of Kibale red colobus (Struhsaker and Leland

1987)

106 Habitat disturbance can have a major impact on female reproductive success. Females must find ways to cope with energetic constraints in degraded habitats. This research builds on previous studies of female reproductive fitness (e.g. Altmann & Alberts 2003; Bercovitch 2001;

Ellison 2003; Foley et al. 2001; Miller et al. 2006; Wade & Schneider 1992) by examining how female reproductive function responds to environmental differences. Female red colobus monkeys living in previously logged and old-growth areas of Kibale National Park, Uganda differ in feeding ecology (see Chapter 2), activity budgets (see Chapter 2), mating behaviors (see

Chapter 3), and genital swelling lengths (see Chapter 3). Despite these differences, female red colobus monkeys do not show the expected differences in ketone levels, seasonality, or reproductive hormone profiles. Theses findings suggest that the dietary and behavioral changes

(Ch 2 and 3) allow females in degraded habitats to maximize their energy balance and continue to maintain reproductive function.

Since becoming a National Park in 1994, Kibale Forest has not had ongoing disturbance.

Thus, tree composition in logged areas have been relatively stable and the red colobus monkeys have had approximately 40 years to adjust since the logging occurred. Reproductive hormone concentrations in logged vs old-growth areas suggest that females living in previously logged areas of the park may have adjusted to living in that disturbed environment. Phenotypic plasticity may allow red colobus to adjust in this environment where they are not facing ongoing anthropogenic activity. Recent studies in humans have focused on how females living under different ecological conditions can adapt to those conditions and adjust their reproduction accordingly (Borgerhoff Mulder 1992; Vitzthum 2001).

Vitzthum (2001) argues that there are “population-specific optimal conditions” as opposed to one norm for optimal conditions for female reproduction. It is then this change from the

107 population-specific optimal condition that impacts reproductive function (Vitzthum 2001).

Frisancho (1993) has argued that conditions during development have an organizational effect to adapt the body to those specific conditions. Thus, the ecological condition in which an individual develops shapes the phenotypic outcome of that individual for that specific environment. The variation in optimal condition between groups is an example of how phenotypic plasticity and developmental reaction norms apply to reproductive variables (e.g. Lee & Kappeler 2003). The developmental reaction norm includes the various possible phenotypic outcomes for a trait and examines how the environment impacts the expression of these phenotypes (Schlichting &

Pigliucci, 1998). Pigliucci (2001) argues that phenotypic plasticity may be an adaptive strategy to impact individual fitness, especially in a stressed environment. Therefore, a female who developed in a lower quality habitat may be able to reproduce (albeit potentially at a tradeoff for lower overall fitness); whereas, a female that developed in a higher quality environment may delay reproduction during short-term ecological stress to invest in future offspring.

Whereas no differences were found in ovarian function, further studies are also needed to determine whether reproductive success differed between habitats. Changes in food availability and hormone concentrations can also impact pregnancy and interbirth intervals. In captive baboon studies where estrogens were experimentally suppressed, 50% of the pregnancies ended in miscarriages (Albrecht et al. 2000). Tardif et al. (2005) found that food stress in marmosets led to pregnancy termination. The opportunistic nature of the sample collection for this study makes it difficult to identify failed pregnancies. Altmann and Alberts (2003) documented that

Amboseli baboons with increased food availability had a significant decrease in their interbirth interval (approximately 25% reduction).Similarly, orangutan interbirth intervals decreased during peaks in fruit abundance (Knott 1999). Knott (1999) hypothesized that a reduction in fat

108 stores may lead to lengthened interbirth intervals. Emery Thompson et al. (2007) also found that higher concentrations of reproductive hormones in female chimpanzees living in high quality habitats corresponded to shorter interbirth intervals and higher infant survival in Kibale National

Park. Thus, despite having similar reproductive function, female red colobus in logged areas may have lower reproductive success because of differences in interbirth intervals and infant survival. Future studies should investigate differences in the success of pregnancies and interbirth intervals between logged and old-growth areas.

The ability for females in once-disturbed areas of a protected park to adjust to the ecological conditions contrasts notably to the apparent inability of females in unprotected forest fragments to adjust to those conditions (Chapman et al. 2006, 2007; Milich et al. 2008). Females in forest fragments face ongoing and more severe disturbances (Chapman et al. 2007), which could trigger stress responses that interfere with reproduction. Individuals are able to adjust to predictable stressors; however, unpredictable stress can have devastating effects on the health of individuals.

These results point to the need to protect wild habitats, especially for endangered species.

It is impossible and unethical to protect all wild land, especially in countries where people are in urgent need of resources. Thus, management of wild areas requires considerable trade-offs between local human populations and wildlife. Consequently it is important to prioritize habitats most relevant to species survival. The results reported here suggest the need to protect sizeable areas of land, even if they have been previously disturbed.

The current data provide strong evidence that behavioral and phenotypic plasticity can allow female red colobus monkeys to adjust to certain levels of habitat degradation provided the conditions are stable. These findings are critical for strategizing how to manage wild

109 populations of this endangered species. Furthermore, the similarities to human populations provide important insight into understanding how primates have adjusted to changing environments throughout evolution.

Whereas the results of this study are promising, there were limitations to the data set.

The comparative nature of this work allowed for broad, robust inter-population comparisons, but limited potential for intra and inter-individual comparisons. Future studies should focus on following individuals on a daily basis, to obtain more fine-grained data on hormone fluctuations.

Focusing on more detailed information for a few females in several groups would help to clarify some of the variation found in this data set. Regular sampling would especially help to determine if the potential differences in maximum estradiol concentrations are a result of actual differences in estradiol or the product of a sampling bias. Additional sampling would also give a clearer view of the hormone profiles for females during cycling, conception, pregnancy, and parturition for more detailed comparisons and for assessing changes in hormone profiles during these different stages. Finally, determining actual reproductive success via long-term data on interbirth intervals, infant survival, and age to sexual maturity is essential for assessing the effectiveness of using fecal estradiol and progesterone concentrations as indicators of reproductive function and contributing to our understanding of the impact of habitat disturbance on female red colobus reproduction.

110 Work Cited

Albrecht ED, Aberdeen GW, and Pepe GJ. 2000. The role of estrogen in the maintenance of primate pregnancy. American Journal Obstetrics and Gynecology 182(2): 432-438.

Altmann J and Alberts SC. 2003. Variability in reproductive success viewed from a life-history perspective in baboons. American Journal of Human Biology 15:401-409.

Bahr JM, Wang SC, Huang MY, and Calvo FO. 1983. Steroid concentrations in isolated theca and granulose layers of preovulatory follicles during the ovulatory cycle of the domestic hen. Biology of Reproduction 29: 325-334.

Bercovitch FB. 2001. Reproductive ecology of old world monkeys. In: Ellison PT (ed.) Reproductive Ecology and Human Evolution . Aldine de Gruyter, New York: 369-396.

Berger J, Ward Testa J, Roffe T, Monfort SL. 1999. Conservation Endocrinology: a noninvasive tool to understand relationships between carnivore colonization and ecological carrying capacity. Conservation Biology 13 (5): 980-989.

Borgerhoff Mulder M. 1992. Reproductive Decisions. In: Smith EA, Winterhalder B (eds.) Evolutionary Ecology and Human Behaviour . Aldine de Gruyter, New York: 339-374.

Campbell NA, Reece JB, and Mitchel LG. 1999. Biology. Benjamin-Cummings, Menlo Park, CA.

Chapman CA and Lambert JE. 2000. Habitat alteration and the conservation of African primates: A case study of Kibale National Park, Uganda. American Journal of Primatology 50: 169-186.

Chapman CA, Chapman LJ, Wrangham R, Isabirye G, and Ben-David K. 1997. Spatial and temporal variability in the structure of a tropical forest. African Journal of Ecology 35:287–302.

Chapman CA and Chapman LJ. 2004. Unfavorable successional pathways and the conservation value of logged tropical forest. Biodiversity and Conservation 13: 2089-2105.

Chapman CA, Naughton-Treves L, Lawes MJ, Wasserman MD, and Gillespie TR. 2007. Population declines of colobus in western Uganda and conservation value of forest fragments. International Journal of Primatology 28 (3): 513-528.

Chapman CA and Chapman LJ. 1997. Forest regeneration in logged and unlogged forests of Kibale National Park, Uganda. Biotropica 29: 386-412.

Chapman CA, Wasserman MD, and Gillespie TR. 2006. Behavioral patterns of colobus in logged and unlogged forests: The conservation value of harvested forests. In: Newton-

111 Fisher NE, Notman H, Paterson JD, and Reynolds V (eds.) Primates of Western Uganda . Springer, New York: 373-390.

Chapman CA, Wasserman MD, Gillespie TR, Speirs ML, Lawes MJ, Saj TL, and Ziegler TE. 2006b. Do food availability, parasitism, and stress have synergistic effects on red colobus populations living in forest fragments? American Journal of Physical Anthropology 131:525-534 .

Chapman CA, Chapman LJ, Naughton-Treves L, Lawes MJ, and McDowell LR. 2004. Predicting folivorous primate abundance: validation of a nutritional model. American Journal of Primatology 62: 55-59.

Ehle FR, Jeraci JL, Robertson JB, and van Soest PJ. 1982. The influence of dietary fiber on digestability, rate of passage and gastrointestinal fermentation in pigs. Journal of Animal Science 55: 1071.

Ellison PT. 2003. Energetics and Reproductive Effort. American Journal of Human Biology 15:342-351.

Ellison PT. 1990. Human ovarian function and reproductive ecology: New hypotheses. American Anthropologist 92 (4): 933-952.

Emery Thompson M, Kahlenberg SM, Gilby IC, and Wrangham RW. 2007. Core area quality is associated with variance in reproductive success among female chimpanzees at Kibale National Park. Animal Behaviour 73: 501-512.

Emery Thompson M. 2005. Reproductive endocrinology of wild female chimpanzees ( Pan troglodytes schweinfurthii ): Methodological considerations and the role of hormones in sex and conception. American Journal of Primatology 67: 137-158.

Foley CAH, Papageorge S, and Wasser SK. 2001. Non-invasive stress and reproductive measures of social and ecological pressures in free-ranging African elephants ( Loxodonta african a). Conservation Biology 15: 1134-1142.

Frisancho RA. 1993. Human Adaptation and Accommodation . University of Michigan Press, Ann Arbor.

Gillespie TR and Chapman CA. 2006. Prediction of parasite infection dynamics in primate metapopulations based on attributes of forest fragmentation. Conservation Biology 20 (2): 441-448.

Goldin BR, Adlercreutz H, Dwyer JT, Swenson L, Warram JH, and Gorbach SL. 1981. Effect of diet on excretion of estrogens in pre- and postmenopausal women. Cancer Research 41: 3771-3773.

112 Kasenene JM. 1987. The influence of mechanized selective logging, felling intensity and gap- size on the regeneration of a tropical moist forest in the Kibale Forest Reserve, Uganda. Ph.D. Dissertation. Michigan State University, Michigan.

Knott CD. 1999. Reproductive, physiological and behavioral responses of orangutans in Borneo to fluctuations in food availability. PhD dissertation. Harvard University.

Knott C. 2001. Female reproductive ecology of the apes: Implications for human evolution. In: Ellison PT (ed.) Reproductive Ecology and Human Evolution . Aldine de Gruyter, New York: 429-463.

Lee PC and Kappeler PM. 2003. Socioecological correlates of phenotypic plasticity of primate life histories. In: Kappeler PM and Pereira ME (eds.) Primate life histories and socioecology . The University of Chicago Press, Chicago: 41-65.

Lipson SF. 2001. Metabolism, Maturation, and Ovarian Function. In: Ellison PT (ed.) Reproductive Ecology and Human Evolution . Aldine de Gruyter, New York: 235-248.

Masters AM and Markham RJ. 1991. Assessing reproductive status in Orang-utans by using urinary estrone. Zoo Biology 10: 197-207.

McGarry JD and Foster DW. 1980. Regulation of hepatic fatty acid oxidation and ketone body production. Annual Reviews of Biochemistry 49: 395-420.

Milich KM, Stumpf RM, and Bahr JM. 2008. The effects of deforestation on stress and reproductive hormone levels in red colobus monkeys ( Piliocolobus tephrosceles ) (abstract). American Journal of Physical Anthropology 135 (S46): 155.

Miller KE, Bales KL, Ramos JH, and Dietz JM. 2006. Energy intake, energy expenditure, and reproductive costs of female wild golden lion tamarins ( Leontopithecus rosalia ). American Journal of Primatology 68: 1037-1058.

Muller MN and Wrangham RW. 2005. Testosterone and energetics in wild cimpanzees ( Pan troglodytes schweinfurthii ). American Journal of Primatology 66: 119-130.

Oates JF. 1994. The natural history of African colobines. In: Davies AG and Oates JF (eds.) Colobine Monkeys: Their Ecology, Behavior, and Evoloution . Cambridge University Press, Cambridge: 75-128.

Oates JF. 1999. Myth and reality in the rainforest: how conservation strategies are failing in West Africa. University of California Press, Berkeley.

Onderdonk DA and Chapman CA. 2000. Coping with forest fragmentation: The primates of Kibale National Park, Uganda. International Journal of Primatology 21:587-611.

113 Pigliucci M. 2001. Phenotypic Plasticity: Beyond Nature and Nurture. Johns Hopkins University Press, Baltimore.

Robinson AM and Williamson DH. 1980. Physiological roles of ketone bodies as substrates and signals in mammalian tissues. Physiological Reviews 60 (1): 143-187.

Ross, C. & Jones, K. E. 1999. Socioecology and the evolution of primate reproductive rates. In: Comparative Primate Socioecology (Ed. by P. C. Lee), pp. 73e110. Cambridge: Cambridge University Press.

Ruoff W. L. 1988. Enterohepatic circulation of estrogens in the domestic pig, Sus scrofa . PhD Dissertation. University of Illinois.

Starin ED. 1991. Socioecology of the red colobus monkey in The Gambia with particular reference to female-male differences and transfer patterns. PhD dissertation. The City University of New York.

Schlichting CD and Pigliucci M. 1998. Phenotypic Evolution: A Reaction Norm Perspective . Sinaur Associates, Sunderland.

Sherry DS. 2002. Reproductive seasonality in chimpanzees and humans: Ultimate and proximate factors. PhD dissertation. Harvard University.

Skorupa JP. 1988. The effect of selective timber harvesting on rain forest primates in Kibale Forest, Uganda. PhD Thesis. University of California, Davis.

Strier KB and Ziegler TE. 2005. Variation in the resumption of cycling and conception by fecal androgen and estradiol levels in female northern muriquis ( Brachyteles hypoxanthus ). American Journal of Primatology 67: 69-81.

Struhsaker TT 2010. The red colobus monkeys: variation in demography, behavior, andecology of endangered species. Oxford University Press, New York, NY.

Struhsaker TT. 1975. The Red Colobus Monkey . University of Chicago Press, Chicago.

Struhsaker TT and Leland L. 1987. Colobines: infanticide by adult males. In: Smuts BB, Cheney DL, Seyfarth RM, Wrangham RW, and Struhsaker TT (eds.) Primate Societies . University of Chicago Press, Chicago: 83-97.

Tardif SD, Ziegler TE, Power M, and Layne DG. 2005. Endocrine changes in full-term pregnancies and pregnancy loss due to energy restriction in the common marmoset (Callithrix jacchus ). The Journal of Clinical Endocrinology and Metabolism 90 (1): 335- 339.

Trivers RL. 1972. Parental investment and sexual selection. In: Campbell B (ed.) Sexual selection and the descent of man. Aldine, Chicago.

114

Vitzthum VJ. 2001. Why Not So Great Is Still Good Enough. In: Ellison PT (ed.) Reproductive Ecology and Human Evolution . Aldine de Gruyter, New York: 179-202.

Wade GN and Schneider JE. 1992. Metabolic fuels and reproduction in female mammals. Neuroscience and Behavioral Reviews 16: 235-272.

Wasserman MD, Taylor-Gutt A, Rothman JM, Chapman CA, Milton K, and Leitman DC. 2012. Estrogenic plant foods of red colobus monkeys and mountain gorillas in Uganda. American Journal of Physical Anthropology 148: 88-97.

Wasserman MD, Chapman CA, Milton K, Gogarten JF, Wittwer DJ, and Ziegler TE. In press. Estrogenic plant consumption predicts red colobus monkey (Procolobus rufomitratus) hormonal state and behavior. Hormones and Behavior .

Wrangham, RW. 1979. On the evolution of ape social systems. Social Science Information 18: 335-368.

Ziegler T, Hodges K, Winkler P, and Heistermann M. 2000. Hormonal correlates of reproductive seasonality in wild female Hanuman langurs ( Presbytis entellus ). American Journal of Primatology 51: 119-134.

115 Table 12:Reproductive traitsfor red colobus at Kibale National Park, Uganda and Abuko Nature Reserve, The Gambia ( 1Struhsaker and Leland 1987; 2 Starin 1991; 3 Struhsaker 1975). Kibale National Park Abuko Nature Reserve Interbirth interval 25.5 ± 5.1 months 1 29.4 months 2 Gestation length 4.5 – 5.5 months 3 5.25 months 2 Menstrual cycle length Unknown 27.7 days 2 Age at sexual maturity (males) 35.4 - 58 months 1 > 28 months 2 Age at sexual maturity (females) 38 - 46 months 1 34.25 months 2

Table 13:Group compositions of the four research groups. Logged Logged Old-growth Old-growth Group 1 Group 2 Group 1 Group 2 Adult Males 8 4 9 11 Adult Females 22 14 15 22 Subadults 3 3 2 4 Juveniles 8 10 1 23 Infants 10 6 7 13 Total 51 37 34 73 Infant/Adult Female Ratio 0.45 0.43 0.47 0.59

116 K-13

K-15

K-30

Figure 11: Map of the logging history at the Kanyawara Field Site in Kibale National Park, Uganda (adapted from Struhsaker 1975). K-13 was heavily logged and poisoned with an arborcide, K-15 was heavily logged, and K-30 was never commercially logged.

117

Old-growth Female 3 Cycling & Pregnant

25 1400

1200 20 Conception 1000 15 800

600 10

400

5 200 Estradiol Concentrations (pg/g dry wt) dry (pg/g Concentrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 9 9 9 9 0 0 0 0 -0 09 -09 -0 0 -09 -0 -0 -10 1 -1 -10 -10 t v- c n n-10 b r r ct-09 a b- a O Oct- Ja -Sep - -Nov -No -De -Dec - -J -Fe -Fe -M -Ma 5 5- 4 4-Dec 4 3-Jan-1 3 2 4 15-Sep-092 15 25-Oc 14 24-Nov-09 1 24 13 2 12 22-Feb-1 14-Mar-1024

P4 (ng/g) E2 (pg/g)

Figure 12: Progesterone and estradiol concentrations for a female in the old-growth area. This female became pregnant in mid to late January.

118 Old-growth Female 5 Pregnant Birth 16 12000

14 10000 12 Conception 8000 10

8 6000

6 4000

4 dry wt) (pg/g Estradiol Progesterone (ng/g dry dry wt) (ng/g Progesterone 2000 2

0 0

9 0 09 09 09 10 10 10

-Dec-09 5-Oct-09 4-Nov-09 4-Dec-09 3-Jan-10 2-Feb-10 4-Mar-1 15-Sep-25-Sep-09 15-Oct-0925-Oct-0 14-Nov-24-Nov- 14-Dec-0924 13-Jan-1023-Jan-10 12-Feb-22-Feb- 14-Mar-1024-Mar-

P4 E2

Figure 13: Progesterone and estradiol concentrations for a pregnant female in the old-growth area. This female gave birth in mid March.

119 Logged Female 4 Cycling & Pregnant Birth Conception 16 4500

14 4000

3500 12 3000 10 2500 8 2000 6 1500 4 1000

2 500 Estradiol Concentrations (pg/g dry wt) Estradiol Concentrations Progesterone Concentrations (ng/g dry wt) Concentrations Progesterone 0 0

9 9 9 9 9 9 9 0 0 0 0 0 0 0 0 0 -0 -0 -0 0 1 1 10 1 10 -1 10 v v v - - - - - r r- r-1 o o c c-09 n n n b a a Oct- Oct- e e a Ja Ja e - - -No -N -N -D -D - -Ma -M -M 6-Oct-096 6 5 5 5 5 5 4-J 4 3-Feb- 5 5 5 16-Sep-0926-Sep- 1 2 1 2 1 25-Dec-09 14- 2 13-Feb-123-F 1 2

P4 (ng/g) E2 (pg/g)

Figure 14: Progesterone and estradiol concentrations for a female in the logged area. This female became pregnant in late October.

120 Logged Female 7 Pregnant Conception Birth

12 2000

1800 10 1600

1400 8 1200

6 1000

800 4 600

400 2

200 wt) dry (pg/g Concentrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 9 9 9 9 9 9 9 9 0 0 0 -09 0 0 10 -0 -0 v- v- -0 - r-10 r-1 p- p- ct ct ov o o a a e e O O Jan-10 Jan eb-10 S S -Oct-09 - - N -N -N -Dec F -M -M 5- 5- 5 4- 4 4 4 3-Jan-103- 3- 2- 4-Mar-10 1 2 15 25 1 2 14-Dec-024-Dec-0 1 2 12-Feb-1022-Feb-10 14 24

P4 (ng/g) E2 (pg/g)

Figure 15: Progesterone and estradiol concentrations for a pregnant female in the logged area. This female gave birth at the end of March.

121 Progesterone Concentrations

6.00

5.00

4.00

3.00

2.00 Progesterone Concentrations (ng/g dry weight) dry (ng/g Concentrations Progesterone 1.00

0.00 Unlogged AVG Logged AVG

Figure 16: Raw data for progesterone concentrations in logged and old-growth areas show that females in both areas have similar means and standard deviations of this hormone.

122 Estradiol Concentrations

3000.00

2500.00

2000.00

1500.00 E2

1000.00 Estradiol Concentrations (pg/g dry wt) dry (pg/g Concentrations Estradiol

500.00

0.00 Unlogged AVG Logged AVG

Figure 17: Raw estradiol data shows that females in old-growth areas have a lower mean estradiol concentration and less variation than in logged areas where there is a higher mean and more variation between females and samples.

123

Figure 18: This figure shows z-scores of progesterone concentrations for individual females compared with a One-way ANOVA in SPSS (p=0.58; n=20; F= 0.317; df=1). Progesterone concentrations did not differ significantly between female red colobus living in logged and old-growth areas of Kibale National Park, Uganda.

124

Figure 19: There is no significant difference in the z-scores of fecal estradiol concentrations of individual females when compared with a One-way ANOVAin SPSS (p=0.97; n=20; F=3.074; df=1).

125

Figure 20: Fecal progesterone concentrations did not differ between logged and old-growth areas when individual sample (p=0.102; n=300; F=2.698; df=1) were analyzed with a one-way ANOVA in SPSS.

126

Figure 21: Fecal estradiol concentrations were significantly higher in samples from logged areas than old-growth areas when comparing individual samples in a one-way ANOVA (p=0.006; n=296; F=7.522; df=1).

127 Figure 22: Ketone concentrations for urine samples collected from female red colobus living in logged and old-growth areas. The percentage of negative and positive results for individual female ketone results were compared using a Fisher’s Exact Test and were not significantly different (p = 0.579).

128 Blood Heart General Liver circulation

St-SO 4 St St-G B i l e s y d ne Gallbladder u id c K t Intestine Urine Feces (Conjugated St-SO 4 Bacteria St Steroids) H St-G (98% free steroids) e p a St St-G,St-SO t 4 ic P o r ta l V e in

s e Indicates ri a v movement of O steroids Adrenal glands Figure 23: Steroid conjugation in the gut. Gut passage time of different foods can impact estradiol concentrations. For this reason, animals that eat different diets may have different estradiol concentrations.

129

Table 14 : Commonly eaten tree species by red colobus monkeys and their phytoestrogen properties as indicated by Wasserman et al. (2012). X indicates foods commonly eaten by females in logged and old-growth areas of Kibale National Park. Lavender highlighting indicates plant species that did not test positive for phytoestrogens and red indicates plants that tested positive for phytoestrogens.

Logged Unlogged Accasia X Albizia grandibractata X Bosqueia phoberos X Celtis africana X Celtis durandii X X Dombea mukoke X Funtumia latifolia X X Hypocratia X Markhamia lutea X Milletia dura X Mimusops X X Newtonia buchannii X Olea welwitschii X Prunus africana X Strombosia scheffleri X X Urella X Ficus natalensis X

130 CHAPTER 5

CONCLUSION

Overview of concepts

The impact of ecological stressors on wild populations of nonhuman primates has been documented in a number of species (Atlmann & Muruthi 1988; Boyle & Smith 2010; Chapman et al. 2006; Eley et al. 1989; Isbell & Young 1993; Iwamoto & Dunbar 1983; Marsh 2003;

Ratsimbazafy 2007; Silva and Ferrari 2009; Singh & Vivanthe 1990; Wong & Sicotte 2007; and

Watts 1988). Female reproduction, in particular, is susceptible to changes in environmental conditions (Bercovitch 2001; Emery Thompson et al. 2007; Knott 1999; Lipson 2001;

Wrangham 2002). These studies demonstrate a reproductive cost to living in degraded areas.

However, recent studies of humans show that females can adjust to different conditions to maintain reproductive function (Borgerhoff Mulder 1992; Vitzthum 2001). Thus, female red colobus monkeys living in degraded areas may be able to adjust as well.

Females may use an energy saving strategy (Milton 1980) to reproduce successfully in logged habitats. By expanding their diet, spending more time feeding, and focusing mating efforts during conceptive periods, females may be able to maintain reproductive function.

However, females in degraded areas may also be limited in their ability to use reproductive strategies that help with overall reproductive fitness (e.g. infanticide avoidance, Hrdy 1979).

Summary and synthesis of findings

In this dissertation, I examine behavioral and physiological differences in female red colobus monkeys living in previously heavily logged and old-growth areas of Kibale National

Park, Uganda. Red colobus are an endangered species that provide a useful model for understanding the ability of wild animals to adjust to degraded habitats, and Kibale provides an

131 excellent semi-controlled natural setting to examine the usefulness of protecting previously logged habitats for wild animal populations. In Chapter 1, I review published literature on primate responses to habitat differences, with a particular focus on reproductive function. I also introduce the red colobus of Kibale National Park.

Chapter 2 focuses on the diets and activity budgets of red colobus in logged and old- growth areas. Using published data on logging in Kibale Forest and red colobus populations densities, I show that while logged areas differ in tree composition and biomass, red colobus densities do not. I compare the diets of red colobus living in each area to see if the two groups in the logged area show a different pattern than the two groups in the old-growth areas. I find that groups in either area of the forest are consistent with each other in the number of species they eat and the amount of time they spend eating these species, but they show differences from the groups living in the other area of the forest. In other words, groups in the logged or old-growth areas are more similar to each other than to groups in the opposite area of the forest.

Specifically, red colobus in the logged area eat a more diverse diet and eat smaller portions of each food species than individuals in the old-growth area. Furthermore, female red colobus in logged areas spend more time feeding and less time doing all other activities than female red colobus in old-growth areas. I conclude that females in logged areas have a plastic behavioral response to living in degraded habitats that allows them to expand their diet to include other foods (effectively increasing the resources available to them), which allows them to maintain the same density as red colobus in the old-growth areas.

In Chapter 3, I describe the sexual swelling length, frequency of copulations, and timing of mating behaviors for females in logged and old-growth areas. Females in logged areas have shorter periods of time having fully inflated genital swellings compared to females in old-growth

132 areas. Additionally, they focus most of their mating bouts in this period of maximum tumescence even more than females in old-growth areas. Females in old-growth areas have higher copulation rates overall, but are less constrained to only mating during periods of maximum tumescence. Following Wrangham (2002), I argue that females in the logged areas minimize energetic costs associated with reproduction and focus mating bouts around periods of ovulation to maintain successful reproduction. Females in the old-growth areas are less constrained in their mating behaviors and are thus able to utilize different reproductive strategies that involve having sex outside of conceptive periods. This may lead to higher overall reproductive success over a lifetime for females in old-growth areas.

Chapter 4 is dedicated to presenting the results from fecal and urine analyses. Urine samples were analyzed for ketone levels (an indicator of nutritional stress) and fecal samples were analyzed for reproductive hormone concentrations. Ketone levels did not differ between logged and old-growth areas. Progesterone and estradiol concentrations did not differ as expected in logged and unlogged areas.

Together, these findings suggest that female red colobus in logged areas of Kibale

National Park have a plastic response to the environment that allows them to maintain reproductive function in a degraded environment. Considering the logged areas of the park are now protected and stable, females are able to adjust to these conditions. This is in stark contrast to the reduced reproductive function of females in forest fragments as indicated by reproductive hormones (Milich et al. 2008) and low infant to adult female ratios (Chapman et al. 2006).

Though the logged areas of the forest are degraded compared to the old-growth areas, they do not face ongoing disturbance. Furthermore, primates in Kibale National Park do not face pressure from human hunting (Struhsaker 1997). This means that they live in a stable

133 environment that might provide them space to adjust to these changes. Potts (2011) suggested that this lack of hunting was the reason that the Kibale chimpanzee population is stable.

These findings indicate the importance of protecting primate habitats. Even degraded areas may have some conservation value. However, these results should not be used as an indication that logging does not impact wild primate populations. First, the extent to which logged areas of Kibale impact reproductive success of red colobus is still unclear. Additionally, red colobus may be more flexible than other species and may be able to adjust to habitat degradation better than other endangered animals. Finally, Kibale may present a special condition in which other ecological stressors, such as hunting, are not present, which allows red colobus to adjust to previously logged areas.

Wildt and Wemmer (1999) argue that the role of reproductive research in conservation is to determine the best strategy for managing wild areas. This study expands our understanding of the complex ways in which animals can cope with environmental changes. These results are not only important for conservation, but also provide valuable insight into understanding how extant primate species, including humans, have dealt with environmental change throughout evolution.

Future Directions

This dissertation contributes to our understanding of the diet, behaviors, and hormone profiles of Kibale red colobus monkeys. Furthermore, it provides evidence of intraspecific differences related to habitat quality. Behavioral and phenotypic plasticity allow individuals in logged areas to adjust to habitat changes and maintain reproductive function. However, it is still unclear if females in logged and old-growth areas have the same reproductive success.

Additional long-term studies documenting the births and infant survival of offspring in logged

134 and old-growth areas are essential to determine if there is a fitness difference between the two areas and also if fecal hormones are a reliable indicator of reproductive success.

Further studies on the reproduction of red colobus monkeys should also prioritize daily fecal and urine samples to examine hormonal fluctuations on a regular basis. These data will provide us with the baseline knowledge needed to understand specific changes in hormone concentrations over the reproductive cycle. That would also help to clarify pieces of this dataset, such as times when it is unclear if the female was pregnant and lost the fetus or was cycling regularly.

More specific hormonal studies will also help to explain the long periods of maximal tumescence and explain the role that hormones play in regulating sexual swellings. This is particularly important for red colobus monkeys that show great variation in sexual swelling size and length and occurrence during non-ovulatory periods, such as pregnancy. Knowing how these swelling relate to the reproductive cycle is very important for understanding reproductive strategies in this species.

Future studies should continue to elucidate the impact that habitat disturbance has on ovarian hormones, as understanding the relationship between ecological stressors and reproductive success is critical for conserving this highly endangered species. Combining feeding ecology with reproductive endocrinology will provide insight into how non-invasively collected fecal samples may provide inconsistent hormonal data because of variation in digestion time. Determining gut passage time of foods and how that correlates with estradiol concentrations would be invaluable to the field of primatology.

The role of males in reproductive outcomes is also a valid topic for future research.

Females in logged and old-growth areas show differences in mating behaviors, but it is unclear

135 what role males play in this. Do males in logged areas also have fewer mating behaviors than males in old-growth areas? Is there more of a reproductive skew in one habitat or the other?

These questions would help to determine if there are different male reproductive strategies as a result of anthropogenic modifications to primate habitats.

Lastly, there is still much to understand about the dietary differences between red colobus living in different habitats. Nutritional analyses of red colobus foods could help explain if individuals are acquiring the same nutrients from different foods. This could account for the differences in diet, but lack of differences in ketone levels seen between logged and old-growth areas.

The research presented here expands our knowledge of red colobus reproduction, behaviors, and diet. It also provides important comparative data on one species living in different habitats. Many of the results were surprising and counter to the predictions, which helps to provide a new understanding of these systems. Thus, these results provide fodder for numerous additional studies that will continue to shape our understanding of wild primate populations and their response to the environment.

136 Works Cited

Altmann J and Muruthi P. 1988. Differences in daily life between semi-provisioned and wild-feeding baboons. American Journal of Primatology 15: 213–221.

Bercovitch FB. 2001. Reproductive ecology of old world monkeys. In: Ellison PT (ed.) Reproductive Ecology and Human Evolution . Aldine de Gruyter, New York: 369-396.

Borgerhoff Mulder M. 1992. Reproductive Decisions. In: Smith EA, Winterhalder B (eds.) Evolutionary Ecology and Human Behaviour . Aldine de Gruyter, New York: 339-374.

Boyle, S.A., Smith, A.T., 2010. Behavioral modifications in northern bearded saki monkeys (Chiropotes satanas chiropotes ) in forest fragments of central Amazonia. Primates 51, 43–51.

Chapman CA, Wasserman MD, Gillespie TR, Speirs ML, Lawes MJ, Saj TL, and Ziegler TE. 2006. Do food availability, parasitism, and stress have synergistic effects on red colobus populations living in forest fragments? American Journal of Physical Anthropology 131: 525-534 .

Eley, R. M., Strum, S. C., Muchemi, G., and Reid, G. D. F. (1989). Nutrition, body condition, activity patterns and parasitism of free-ranging baboons ( Papio anubis ) in Kenya. American Journal of Primatology 18: 209–219.

Emery Thompson M, Kahlenberg SM, Gilby IC, and Wrangham RW. 2007. Core area quality is associated with variance in reproductive success among female chimpanzees at Kibale National Park. Animal Behaviour 73: 501-512.

Hrdy SB. 1979. Infanticide among animals: a review, classification, and examination of the implications for the reproductive strategies of females. Ethol Sociobiol 1:13-40.

Isbell, L.A. & Young, T.P. ( 1 9 9 3 ) . Human presence reduces leopard predation in a population of vervet monkeys Cercopithecus aethiops. Anon. Behav ., 45, 1233-5.

Iwamoto T & Dunbar RIM. 1983. Thermoregulation, habitat quality and the behavioural ecology of gelada baboons. Journal of Animal Ecology 52: 357-366.

Knott CD. 1999. Reproductive, physiological and behavioral responses of orangutans in Borneo to fluctuations in food availability. PhD dissertation. Harvard University.

Lipson SF. 2001. Metabolism, Maturation, and Ovarian Function. In: Ellison PT (ed.) Reproductive Ecology and Human Evolution . Aldine de Gruyter, New York: 235-248.

Marsh, L. K. 2003. Primates in fragments: ecology and conservation . Kluwer Academic/Plenum Publishers, New York.

137 Milich KM, Stumpf RM, and Bahr JM. 2008. The effects of deforestation on stress and reproductive hormone levels in red colobus monkeys ( Piliocolobus tephrosceles ) (abstract). American Journal of Physical Anthropology 135 (S46): 155.

Milton K. 1980. The foraging strategy of howler monkeys: a study in primate economics . Columbia University Press, New York.

Potts KB. 2011. The long-term impact of timber harvesting on the resource base of chimpanzees in Kibale National Park, Uganda. Biotropica 43 (2): 256-264.

Ratsimbazafy J. 2007. Diet Composition, foraging, and feeding behavior in relation to habitat disturbance: implications for the adaptability of ruffed lemurs ( Varecia v. editorium ) in Manombo forest, Madagascar. In Gould L and Sauther ML (eds) Lemurs: Ecology and Adaptation . Springer, New York: 403-422.

Silva, S. S. B., & Ferrari, S. F. (2009). Behavior patterns of southern bearded sakis Chiropotes satanas) in the fragmented landscape of eastern Brazilian Amazonia. American Journal of Primatology , 71, 1–7.

Singh M and Vinathe S. 1990. Inter-population differences in the time budgets of bonnet monkeys ( Macaca radiata ). Primates 31(4): 589-596.

Struhsaker TT. 1997. Ecology of an African Rain Forest: Logging in Kibale and the conflict between conservation and exploitation. University Press of Florida, Gainesville.

Vitzthum VJ. 2001. Why Not So Great Is Still Good Enough. In: Ellison PT (ed.) Reproductive Ecology and Human Evolution . Aldine de Gruyter, New York: 179-202.

Watts, D.P. Environmental influences on mountain gorilla time budgets. American Journal of Primatology 15: 295–312, 1988

Wildt DE and Wemmer C. 1999. Sex and wildlife: the role of reproductive science in conservation. Biodiversity and Conservation 8: 956-976.

Wong S. N. P. and Sicotte P. (2007). Activity budget and ranging patterns of Colobus vellerosus in forest fragments in Central Ghana. Folia Primatol. 78: 245–254.

Wrangham, RW. 2002. The cost of sexual attraction: is there a trade-off in female Pan between sex appeal and received coercion? In Boesch C, Hohmann G, and Marchant LF (eds.) Behavioural Diversity in Chimpanzees and Bonobos. University Press, Cambridge: 204- 215.

138 APPENDIX A

FEMALE ESTRADIOL AND PROGESTERONE GRAPHS

Old-growth Female 1 Cycling

14 800

12 700

600 10 500 8 400 6 300 4 200

2 100 Estradiol Concentrations (pg/g dry wt)

Progesterone Concentrations (ng/g dry wt) 0 0

9 9 9 9 9 9 9 9 9 9 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 - - t- t- t- v- v- v------r- r- r- p p c c c o o o c c c n n n b b b a a a e e O O O e e e Ja Ja Ja e e e -S -S - - - -N -N -N -D -D -D - - - -F -F -F -M -M -M 7 7 7 7 7 6 6 6 6 6 6 5 5 5 4 4 4 6 6 6 1 2 1 2 1 2 1 2 1 2 1 2 1 2

P4 (ng/g) E2 (pg/g)

Figure 24: Progesterone and estradiol concentrations for a cycling female in the old-growth area. This female did not become pregnant during the duration of sample collection.

139 Old-growth Female 2 Cycling

6 1000

900 5 800

700 4 600

3 500

400 2 300

200 1

100 wt) dry (pg/g Concentrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 9 9 9 9 9 0 0 09 09 09 0 10 10 10 10 10 10 t-09 -09 -0 -0 -0 - r- r- p-0 p-0 c ct- ct ov v- v- n- n- n-1 b-1 b- ar-10 e e ec ec- ec Ja e -O -N -No -No -D -D -Ja -Ja -Fe -Feb -M -Ma -Ma 5-S 5-O 5-O 4 4-D 3 2-F 2 4 4 4 1 25-S 15 2 14 24 14 24 13 23- 12 2 1 2

P4 (ng/g) E2 (pg/g)

Figure 25: Progesterone and estradiol concentrations for a cycling female in the old-growth area. This female did not become pregnant during the duration of sample collection.

140 Old-growth Female 4 Cycling & Pregnant

16 1400 Conception 14 1200

12 1000

10 800 8 600 6

400 4

2 200 Estradiol Concentrations (pg/g dry wt) dry (pg/g Concentrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 9 9 9 9 0 0 0 0 0 0 0 -09 09 1 1 1 1 - -09 t-0 t- t- v v-09 - - - -10 r- p p c o c c-0 b b b ar- a e Oc e e e -S -Se - -N -De -D -Jan-10 -F -F -Fe M -M 5 5 5-Oc 5 4 4 4 3 2 2 2 4- 4 1 2 15-O 2 14-No 24-Nov-09 14-Dec-092 13-Jan-1023-Jan-10 1 2 14-Mar-102

P4 (ng/g) E2 (pg/g)

Figure26: Progesterone and estradiol concentrations for a female in the old-growth area. This female became pregnant in early to mid February.

141 Old-growth Female 6 Pregnant Birth 8 350

7 300

6 250

5 200 4 150 3

100 2

1 50 Estradiol Concentrations (pg/g dry wt) dry (pg/g Concentrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 9 9 9 9 0 0 0 0 -09 -09 -09 -0 -0 -1 10 -10 -10 -10 - t t t n n-10 - r r c a a Ja -Sep -Oc -Nov -Nov -Dec-0 -Dec-0 - -M -Ma 5 5-O 5 4 4-Nov-094-Dec-09 3-Jan-1 2-Feb 2-Feb 4 15-Sep-092 15-Oc 2 14 2 14 24 13 23-J 1 22-Feb-1 14-Mar-1024

P4 E2

Figure 27: Progesterone and estradiol concentrations for a pregnant female in the old-growth area. This female gave birth in late October.

142 Logged Female 1 Cycling

20 1600

18 1400 16 1200 14 1000 12

10 800

8 600 6 400 4 200

2 wt) dry (pg/g Concentrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 0 09 09 09 -09 10 10 10 v -09 -09 - r-10 -10 -10 p- p-09 ct- ct- ov ov-0 c c-09 n-10 n- b-10 ar ar e ec-09 e a Jan- e eb-1 -Se -N D -Feb -F 5-Oct-095-O 5-O 4-No 4- 4-De 3-J 3- 3-Ja 2 4-Ma 4-M 4-M 15 25-S 1 2 14-N 24 1 24-D 1 2 12-F 22 1 2

P4 (ng/g) E2 (pg/g)

Figure 28: Progesterone and estradiol concentrations for a cycling female in the logged area. This female did not become pregnant during the time of sample collection.

143 Logged Female 2 Cycling

14 7000

12 6000

10 5000

8 4000

6 3000

4 2000

2 1000 Estradiol Concentrations (pg/g dry wt) dry (pg/g Concentrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 9 9 9 0 0 09 -09 -0 09 09 -09 09 -10 10 -10 -10 -10 -1 p- p-0 ov- ov-0 c c- n ar-1 ar-10 Oct-0 an- eb -Se -Se -Nov -N Dec- -Feb Mar -M M 5- 4 4-N 4- 4-De 4-De 3-Ja 3-J 3-Jan 2 2-F 4- 15 25 15-Oct 25-Oct-09 1 24 1 2 1 2 12-Feb-102 14 24-

P4 (ng/g) E2 (pg/g)

Figure 29: Progesterone and estradiol concentrations for a cycling female in the logged area. This female did not become pregnant during the time of sample collection.

144 Logged Female 3 Cycling

14 2500

12 2000

10

1500 8

6 1000

4

500 2 Estradiol Concnetrations (pg/g dry wt) dry (pg/g Concnetrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 9 9 9 9 0 0 0 0 -0 09 -09 -0 0 -09 -0 -0 -10 1 -1 -10 -10 t v- c n n-10 b r r ct-09 a b- a O Oct- Ja -Sep - -Nov -No -De -Dec - -J -Fe -Fe -M -Ma 5 5- 4 4-Dec 4 3-Jan-1 3 2 4 15-Sep-092 15 25-Oc 14 24-Nov-09 1 24 13 2 12 22-Feb-1 14-Mar-1024

P4 (ng/g) E2 (pg/g)

Figure 30: Progesterone and estradiol concentrations for a cycling female in the logged area. This female did not become pregnant during the time of sample collection.

145 Logged Female 5 Cycling & Pregnant Conception 25 2500

20 2000

15 1500

10 1000

5 500 Estradiol Concentrations (pg/g dry wt) dry (pg/g Concentrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 9 9 9 9 0 0 0 -0 -09 -09 -0 0 -0 -1 1 -10 10 -10 p t t v- v-09 n-10 b-10 b r an a a J Jan- -Oc -Nov -No No -Dec -Dec-0 - - -Fe -Fe -Mar- 5-Oct-09 5 4 4- 4 3 2 4-Mar-10 4-M 15-Sep-0925-Se 15-Oc 2 14 2 14 24-Dec-09 13 23-J 12-Feb-122 14 2

P4 (ng/g) E2 (pg/g)

Figure 31: Progesterone and estradiol concentrations for a female in the logged area. This female became pregnant in mid November.

146 Logged Female 6 Cycling & Pregnant

9 3500

8 3000 7 Conception 2500 6

5 2000

4 1500

3 1000 2 500 1 Estradiol Concentrations (pg/g dry wt) dry (pg/g Concentrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 0 09 09 09 -09 10 10 10 v -09 -09 - r-10 -10 -10 p- p-09 ct- ct- ov ov-0 c c-09 n-10 n- b-10 ar ar e ec-09 e a Jan- e eb-1 -Se -N D -Feb -F 5-Oct-095-O 5-O 4-No 4- 4-De 3-J 3- 3-Ja 2 4-Ma 4-M 4-M 15 25-S 1 2 14-N 24 1 24-D 1 2 12-F 22 1 2

P4 (ng/g) E2 (pg/g)

Figure 32: Progesterone and estradiol concentrations for a female in the logged area. This female became pregnant in mid December.

147 Logged Female 7 Pregnant Conception Birth

12 2000

1800 10 1600

1400 8 1200

6 1000

800 4 600

400 2

200 wt) dry (pg/g Concentrations Estradiol Progesterone Concentrations (ng/g dry wt) dry (ng/g Concentrations Progesterone 0 0

9 9 9 9 9 9 9 9 9 0 0 0 -09 0 0 10 -0 -0 v- v- -0 - r-10 r-1 p- p- ct ct ov o o a a e e O O Jan-10 Jan eb-10 S S -Oct-09 - - N -N -N -Dec F -M -M 5- 5- 5 4- 4 4 4 3-Jan-103- 3- 2- 4-Mar-10 1 2 15 25 1 2 14-Dec-024-Dec-0 1 2 12-Feb-1022-Feb-10 14 24

P4 (ng/g) E2 (pg/g)

Figure 33: Progesterone and estradiol concentrations for a pregnant female in the logged area. This female gave birth at the end of March.

148