(Cercopithecus Ascanius), Blue Monkeys (C. Mitis), and Hybrids

Home , Guenon, Theropithecus

COAT COLOR VARIATION BETWEEN RED-TAILED MONKEYS

(CERCOPITHECUS ASCANIUS), BLUE MONKEYS (C. MITIS), AND HYBRIDS

(C. ASCANIUS x C. MITIS) IN GOMBE NATIONAL PARK, TANZANIA

Elizabeth Tapanes

A Thesis Submitted to the Faculty of

The Dorothy F. Schmidt College of Arts and Letters

In Partial Fulfillment of the Requirements for the Degree of

Master of Arts

Florida Atlantic University

Boca Raton, FL

May 2016

ACKNOWLEDGEMENTS

I would like to thank my advisor, Kate M. Detwiler, for her support, guidance, and faith. This thesis would not have been possible without her willingness and enthusiasm to invite me into the Gombe Hybrid Monkey Project. I would like to thank my committee members, Drs. Douglas Broadfield and Andrew Halloran, for allowing me to pursue quite an ambitious masters thesis. I thank Florida Atlantic University for funding that made this project possible: Seed Grant (DOR-FY14), Technology Fee Grant

(B09-377), Graduate Research and Inquiry Program Grant (GRIP), and a Morrow

Research Fellowship granted by the Department of Anthropology. I also thank Sigma Xi for a Grant-in-Aid of Research.

I would like to thank the Gombe Stream Research Center, Tanzania National

Parks Authority, Tanzania Wildlife Research Institute, and Tanzania Commission for

Science and Technology for giving me permission to conduct research in Gombe

National Park, Tanzania. I thank Dr. Anthony Collins for his on the ground support, dedication, and enthusiasm for the project. My time in Gombe would not have been as memorable or successful without him. I thank Gombe Hybrid Monkey Project’s field assistants, Mary Nkoranigwa and Maneno Mpongo, for their help with my seemingly endless quest for great photos. I also thank Felix Angwella, for his help identifying and learning the female guenons of Mkenke Group A. Additionally, I thank Dr. Deuss

Mjungu for his help with research logistics.

iv To all my colleagues and friends who have sat through draft edits and color method brainstorms, I thank you profusely. Specifically, I thank Daniel Alempijevic for his push and insight into putting the phenotypic hybrid index (PHI) on a 0 to 1 scale. I would like to thank Amelia Villasenor and Dr. Robert O’Malley for their guidance on how to be a great field anthropologist. I also thank Amelia for her continual patience and help while I developed and tested methods for capturing color (which was not a fun process!). I extend a thank you to Andrew Bernard for accelerating my learning curve in regards to using a dSLR, proper photo data management, and taking great primate photographs. This was an integral part of the success of my thesis. Lastly, I thank

Christian Rodriguez for the countless hours spent rehashing the properties of light and color with me. These conversations have been paramount to the evolution of my thoughts on capturing color in arboreal primates.

Most importantly, I would like to thank my family for their never-ending support and love. I thank my mother, Yrelys Tapanes, for always pushing me to follow my dreams, and my father, Ramon Tapanes, for giving me the courage to do so. I thank my sisters, Yrelys Tapanes and Alexa Gonzalez, for being amazing sounding boards throughout my life and for not forgetting I existed while I was in Tanzania. I thank

Jeffrey Gonzalez and Christina Martinez, who both adopted me as an extended family member more than 15 years ago and have been invested in my success since day one.

This thesis would not have been possible without the incredible social network the six of them provide me with, and my words will forever be insufficient to express my deep gratitude.

ABSTRACT

Author: Elizabeth Tapanes

Title: Coat Color Variation Between Red-tailed Monkeys (Cercopithecus ascanius), Blue Monkeys (C. mitis), and Hybrids (C. ascanius x C. mitis) in Gombe National Park, Tanzania

Institution: Florida Atlantic University

Thesis Advisor: Dr. Kate M. Detwiler

Degree: Master of Arts

Year: 2016

Cercopithecus monkeys are a species-rich radiation where interspecific mating leads to novel phenotypes due to pelage color and pattern diversity within the genus. The goals of this thesis were to (1) test a new method for studying color objectively in wild arboreal primates, and (2) apply a phenotypic hybrid index (PHI) to known individuals of a hybrid zone between C. ascanius and C. mitis in Gombe National Park, Tanzania through the use of digital photography. I scored seven pelage character states as 0 (C. mitis), 0.25 (mitis-like), 0.50 (intermediate), 0.75 (ascanius-like), or 1 (C. ascanius).

Photos indicate most phenotypic hybrids express a white nose spot, but all other regions of pelage color and pattern are variable, and an assortment of hybrid phenotypes are seen at Gombe. Results indicate it is currently not possible to extend parameters for assessing color objectively with RGB values, but numerical non-RGB methods show promise.

DEDICATION

I dedicate this thesis to my parents, who left Cuba more than forty years ago so their children could have better lives, and who taught me through example that anything is

possible with enough perseverance and hard work.

I love you.

COAT COLOR VARIATION BETWEEN RED-TAILED MONKEYS

(CERCOPITHECUS ASCANIUS), BLUE MONKEYS (C. MITIS), AND HYBRIDS

(C. ASCANIUS x C. MITIS) IN GOMBE NATIONAL PARK, TANZANIA

TABLES ...... X

FIGURES ...... XII

INTRODUCTION ...... 1

CHAPTER ONE: BACKGROUND AND RESEARCH QUESTIONS ...... 4

Background ...... 4

African Cercopithecus Monkeys...... 5

Hybridization within Cercopithecus Monkeys ...... 7

Primate Coloration ...... 9

Study Site ...... 12

Gombe National Park, Tanzania ...... 12

Research Questions ...... 13

CHAPTER TWO: METHODS ...... 15

Developing and Testing an Extension to Established Color Methods ...... 15

Equipment and settings ...... 15

Method Description ...... 16

viii

Linearization and equalization ...... 17

Method validation ...... 18

Generating a Phenotypic Hybrid Index ...... 19

Camera guidelines for color photography ...... 19

Generating average phenotypic hybrid indices ...... 19

Analysis ...... 20

CHAPTER THREE: EXTENDED SEQUENTIAL METHOD ...... 24

Method Validation: Linearization and Equalization ...... 24

Accuracy and Precision ...... 25

CHAPTER FOUR: PHENOTYPIC HYBRID INDEX OF C. MITIS DOGGETTI, C.

ASCANIUS SCHMIDTI, AND HYBRIDS RESULTS ...... 28

Quantitative Pelage Variation of Guenons in Mkenke Group A ...... 28

Quantitative Pelage Variation of Guenons outside Mkenke Group A ...... 34

Qualitative Pelage Variation of Guenons at Gombe ...... 37

C. mitis doggetti pelage coloration ...... 37

C. ascanius schmidti pelage color variation...... 39

C. ascanius x C. mitis pelage coloration ...... 41

CHAPTER 5: DISCUSSION AND CONCLUSION ...... 46

Assessing Color Variation in Wild Arboreal Primates ...... 46

Hybrid Coloration within Gombe National Park ...... 48

FURTHER THOUGHTS ...... 51

APPENDICES………...... 53

A - IACUC Approval ...... 54

REFERENCES……………...... 55

TABLES

Table I. Phenotypic Characteristics for Hybrid Scoring using Pelage Morphology...... 22

Table II. Statistical Comparisons between Test and Control Squares for the Extended

Color Methods at the 10-minute Extension ...... 26

Table III. Statistical Comparisons between Test and Control Squares for an Extension to

Color Methods, at the 10-minute Extension in Different Placement Areas ...... 26

Table IV. Statistical Comparisons between Test and Control Squares for an Extension

to Color Methods, at the 10-minute Extension in Different Placement Areas and

Under Varying Canopy Conditions ...... 27

Table V. Scored Phenotypic Characteristics for Guenons of

Mkenke Group A (MkGrpA) ...... 30

Table VI. Scored Phenotypic Characteristics for Guenons

Outside Mkenke Group A (MkGrpA)...... 35

FIGURES

Fig 1. Field assistant holds up a color checker...... 17

Fig. 2. Relative frequency of phenotypes, using three hybrid categories

for MkGrpA ...... …..29

Fig. 3. Relative frequency of phenotypes, using expanded hybrid categories

for MkGrpA ...... 29

Fig. 4. C. m. doggetti facial coloration ...... 38

Fig. 5. C. m. doggetti dorsal surface and tail coloration ...... 38

Fig. 6. C. ascanius schmidti red tail with black distal tip, and

red dorsal flank coloration ...... 39

Fig. 7. C. ascanius schmidti black outline of cheek band and white ventral surface ...... 40

Fig. 8. C. ascanius schmidti facial coloration ...... 40

Fig. 9. C. m. doggetti x C. a. schmidti (B-BH) phenotype

from lower Mkenke valley … ...... ….41

Fig. 10. C. m. doggetti x C. a. schmidti (B-H) facial coloration from MkGrpA ...... 42

Fig. 11. C. m. doggetti x C. a. schmidti (I-H) facial coloration ...... 43

Fig. 12. C. m. doggetti x C. a. schmidti (I-H) tail coloration ...... 44

Fig. 13. C. m. doggetti x C. a. schmidti (R-H) facial coloration ...... 45

Fig. 14. C. m. doggetti x C. a. schmidti (R-RH) facial coloration ...... 45

xii

INTRODUCTION

Primates, including Homo sapiens, exhibit variation in both skin and pelage color.

The general assumption is that mammalian color is an adaptive trait used for camouflage, thermoregulation, and/or communication [Bradley and Mundy, 2008; Caro, 2013; 2005].

Objective assessment of color variation proves difficult though, because humans perceive color differently and because of variations in light that affect visual color assessment

[Bergman and Beehner, 2008]. Thus, it is important to analyze color by using methods that remove human subjectivity and which allow for statistical analysis [Bergman and

Beehner, 2008; Endler, 1990; Stevens et al., 2007]. Digital photography is non-invasive and has been used successfully to measure color objectively in many primate taxa. Two methods, the adjacent method and the sequential method, are currently used by primatologists and involve calibrating a photo against a known color standard to obtain

Red-Green-Blue (RGB) values.

The purpose of this research project was to quantitatively examine a new color method and the phenotypic diversity within the hybrid zone at Gombe between C. ascanius and C. mitis. All research was conducted in collaboration with Dr. Kate M.

Detwiler of Florida Atlantic University, who has been studying the Cercopithecus hybrid zone at the Gombe Stream Research Center intermittently since 1994 [Detwiler, 2002].

The population is currently phenotypically described through in-field qualitative

assessments using high-powered binoculars [Detwiler, 2010]. This project is the 1st attempt to characterize pelage diversity between known individuals in the Gombe hybrid zone and the first attempt at testing a method to analyze color objectively in the population.

My primary goal was to test methods to study pelage color using RGB values.

Closely related arboreal species (i.e. Cercopithicus, Eulemur, and callitricids) often manifest striking pelage color and pattern variation [Bradley and Mundy, 2008]. The problem remains that it is still difficult to study color objectively in wild arboreal primates [Allen and Higham, 2015; Allen et al., 2014]. This is likely attributed to the fact that arboreal primates live in high canopies and are fast moving, and even in flexible circumstances, methods require a color card in the same location as the primate within a two-minute window [Bergman and Beehner, 2008].

My secondary goal was to quantify pelage phenotypic diversity for known individuals in the hybrid zone between C. mitis and C. ascanius. This thesis modifies a phenotypic hybrid index (PHI) previously created for the population at Gombe [Detwiler,

2010] through the use of digital photography. It provides an average PHI score for all known individuals in Mkenke Group A (a habituated group consisting of C. ascanius, C. mitis, and hybrids). Additionally, I applied PHI scoring to various individuals living in other groups or as solitary males.

The results from the phenotypic hybrid index provide a preliminary analysis of pelage phenotypic variation within and between C. ascanius, C. mitis, and C. ascanius x

C. mitis at Gombe. However, the results from testing an extension to currently accepted color methods suggests it is not currently possible to capture RGB values from wild arboreal Cercopithecus monkeys. Other avenues may be available to apply numerical scoring to pelage color. This thesis is organized into five chapters. Chapter one provides an overview on Cercopithecus African monkeys, hybridization within Cercopithecus monkeys, primate coloration, gives a description of the study location, and concludes with a list of research questions I sought to answer. Chapter two provides detailed methods I employed to answer the proposed questions. Chapter three delves into the results of the tests for an extended method. Chapter four provides results of the application of the phenotypic hybrid index (PHI) at Gombe. Chapter five provides a discussion on the implications of the results, provides recommendations to future arboreal color studies, and discusses the importance of studying phenotypic diversity in the

Gombe Hybrid zone.

CHAPTER ONE: BACKGROUND AND RESEARCH QUESTIONS

Background

Spectrophotometry and digital photography are used to analyze color objectively.

The former measures the distribution of wavelengths (including ultraviolet wavelengths) reflected via a digital device [Zuk and Decruyenaenaere, 1994]. To use this method, the lens has to be placed a few inches from an animal and supplied a constant and consistent illuminant source [Bergman and Beehner, 2008]. Digital photography is used to capture phenotypic diversity in the natural world in a non-invasive manner. Photographs are calibrated with a known color standard and then Red-Green-Blue (RGB) values that represent true color of a specific patch of pelage or skin are obtained during post- processing.

Researchers currently accept two methods to analyze color via digital photography

– the adjacent method and the sequential method. The two methods have been used extensively and successfully to study museum skins [Kamilar et al., 2013; Kamilar and

Bradley, 2011; Kamilar, 2009], terrestrial primates [Bergman and Beehner, 2008;

Higham et al., 2013], and captive primates [Allen et al., 2014; Allen and Higham, 2013].

The adjacent method requires placing the color standard in the same frame as your subject. The sequential method requires the photo of the primate and color standard to be taken two minutes apart [Bergman and Beehner, 2008]. The problem remains that

studying color objectively in wild arboreal primates remains out of reach

[Allen and Higham, 2014]. This is likely because many arboreal primates live in high canopies and are usually fast moving. Devising robust methods to capture color objectively in arboreal primates is an important step towards understanding the evolution of pelage color in a variety of primate radiations, but specifically in ones that are 1) highly diverse, and/or that 2) exhibit patterns of interspecific mating.

African Cercopithecus Monkeys

Cercopithecus monkeys, also known as guenons, represent a large, diverse, mostly arboreal, and colorful taxonomic group. Preferred habitats consist of woodland, mangrove forests, and swamp forests. Typical home range extends from 3 ha to over 52 km2; and distance traveled per day varies between 500 m to over 4300 m [Butynski,

2002]. In general, Cercopithecus monkeys prefer fruits, but are opportunistic omnivores, consuming leaves, insects, and occasionally vertebrate prey [Chapman et al., 2002;

Gautier-Hion 1988].

Both C. mitis and C. ascanius live in one-male multi female groups, where females exhibit philopatry and solicit sex from males with behavioral cues such as head flagging, puckering, and presenting during estrus [Detwiler, 2010; Lawes et al., 2013]. In both species, less than 20% of matings occur with transient males [Cords, 1988;

Struhsaker, 1981]. For C. mitis, group numbers can range from 12 to 70 individuals

[Wolfheim, 1988], but for C. ascanius the range is from 10 to 35 individuals [Brown,

2013; Cords, 1987; 1984]. C. mitis and C. ascanius exhibit sexual dimorphism, with the males weighing substantially larger than the females (C. mitis: (F) 4231g (M) 7350g; C. ascanius: (F) 3300g (M) 4212g) [Harvey et al., 1987, Napier, 1981].

Polyspecific associations, also known as mixed-species associations, occur with regularity between C. ascanius and C. mitis. Both species likely occupy different ecological niches. By definition, no two species can ever occupy the same ecological niche although they may interact in the same ecological guild [Beaudrot et al., 2013;

Gause, 1934]. C. mitis occupies evergreen, semidecidious, and dry montane forests with a preference for high elevations up to 3300 meters. C. ascanius occupies gallery, swamp, acacia woodland, and dry montane forests and prefer to not exceed elevations of 2000 meters [Wolfheim, 1983]. C. mitis cover a larger area and travel further to find food when living in groups with C. ascanius [Gautier-Hion, 1987]. When associations do occur, the less abundant species joins the most abundant [Struhsaker, 1981].

The phylogeny of Cercopithecus monkeys remains largely debated and challenging to resolve [Butynski, 2002]. Although the first guenon is dated to the

Pliocene, most of the lineages evolved within the last 1 million years [Leakey, 1988]. The general consensus is that the radiation is likely still undergoing active speciation

[Gautier-Hion et al., 1988]. The most recent taxonomy for the guenons lists four genera,

23 species, and 55 sub-species [Grubb et al., 2003]. Guenons have been characterized as descendants of a terrestrial Cercopithecine ancestor who become arboreal secondarily

[Rollinson and Martin, 1991]. There are six recognized species groups: the cephus group, the mitis group, the mona group, the neglectus group, the diana group, and the hamlyni

group. C. mitis is currently classified under the mitis-group and C. ascanius classified under the cephus-group [Grubb et al., 2003]. Recent genetic analysis links cephus and mitis in one clade. Although the relationship is phylogenetic, researchers argue reticulate episodes (i.e. hybridization) are likely not entirely absent from the Cercopithecus group

[Tosi et al., 2005].

Hybridization within Cercopithecus Monkeys

Hybridization is defined as interbreeding between individuals from genetically distinct populations [Barton and Hewitt, 1985; Rhymer and Simberloff, 1966]. There are two modes of hybridization. Parapatric hybridization occurs when two populations interbreed at the interface where they meet and this creates a narrow hybrid zone.

Sympatric hybridization occurs when two populations interbreed throughout overlapping ranges [Woodruff, 1973]. Within African Cercopithecine monkeys, many taxa are known to hybridize but parapatric hybridization is the most common form [Jolly, 2001;

Woodruff, 1973]. Historically, hybridization has been regarded as evolutionarily maladaptive, but the recent general consensus is that hybridization can lead to speciation and/or novel morphological forms [Arnold, 1997; Dowling and Secor, 1997; Grant and

Grant, 1992].

Cercopithecus monkeys often form mixed species groups, and hybridization, although rare, has been reported throughout Africa [Detwiler et al., 2005; Detwiler,

2002]. There are 25 locations where hybrid Cercopithecus monkeys have been observed, and two known persistent hybrid zones [Detwiler, 2002]. One hybrid zone is parapatric,

and occurs between C. m. albogularis and C. m. stuhlmanni in Central Northern Tanzania

[Booth, 1968; Butynski and De Jong, 2012; Groves, 2001]. The second hybrid zone is sympatric, and occurs between C. m. doggetti and C. a. schmidti in Gombe National Park,

Tanzania [Detwiler, 2010; 2002; Detwiler et al., 2005]. These hybrid zones serve as valuable natural laboratories to study the process and outcomes of hybridization, such as phenotypic diversity, within the Cercopithecus radiation.

The hybrid zone at Gombe National Park, Tanzania offers a unique opportunity to study hybridization because it is the only known sympatric primate hybrid zone. Hybrids have been documented in the park for the past 60 years [Detwiler, 2002; Goodall, 2986;

2000; Stanford, 1998], but evidence shows hybridization in the park has been occurring for over 200 years and likely predates the earliest record of deforestation (approximately

1880) [Detwiler, 2010; Alin et al., 2002]. The hybrid zone is at least 15% phenotypically hybrid and hybrid females are fertile and produce offspring [Detwiler, 2010]. Hybrids are found throughout the park in groups with varying phenotypic composition, and this suggests introgression (i.e. hybridization and backcrossing) is bidirectional [Detwiler,

2002]. Backcrossing is defined as a hybrid mating with a parental phenotype.

Studies of C. ascanius x. C. mitis hybrids at other sites argue that hybrids are larger than both parental species and exhibit duller pelage coloration. Within the Kibale forest, historical records of hybrids indicate their pelage color is intermediate [Struhsaker et al., 1988]. Fertility of first generation (F1) and second generation (F2) hybrid females has been recorded at several locations [Detwiler, 2002; Struhsaker et al., 1988], however, the fertility of male hybrids is currently unknown. At Gombe National Park, females

mate with males of all phenotypes, and often give birth to offspring with varying pelage color and pattern expression.

Cercopithecus facial coloration is popularly hypothesized to serve as a mechanism for recognizing conspecifics and to avoid hybridization [Allen et al., 2014;

Kingdon, 1988]. An alternative hypothesis is that facial coloration helps individuals convey spatial information to coordinate movements in dense vegetation [Kingdon,

1988]. Dorsal surface coloration has loosely proven to follow Gloger’s rule, where darker hair on the dorsal surface correlates to higher evapotranspiration (AET) over geographic ranges [Kamilar and Bradley, 2011]. Studying color objectively in Cercopithecus hybrid zones is of high importance to understanding color evolution within the genus.

Primate Coloration

Primates exhibit variation in both skin and pelage coloration. It is generally hypothesized that color serves adaptive functions such as concealment, communication, and physiological regulation [Bradley and Mundy, 2008; Caro, 2005]. Although morphological change is not sufficient to signal species status, in the absence of genetic information, color morphs are often assumed species varieties. In general, color evolution of body parts (i.e. dorsal and ventral surfaces) are hypothesized to function for camouflage. In contrast, small patches on the face and tail can signal sexual and interspecific communication [Caro, 2005].

In the Old World monkeys, a diversity of facial colors and patterns express themselves in the mandrills (Mandrillus sphinx), mangabeys (Lephocebus and

Cercocebus), and guenons. This facial coloration is hypothesized to aid with communication – notably for sexual communication and to recognize conspecifics [Allen and Higham, 2015; Allen et al., 2014; Higham et al., 2013; Santana, 2013; Kingdon,

1988]. Skin coloration may also aid in sexual communication. For example, vervet monkey (Chlorocebus sabaeus) testicle coloration is often related to dominance within a hierarchy, with a deeper blue being characteristic of a high-ranking male [Cramer et al.,

2013]. Similarly, gelada baboon (Theropithecus gelada) leader males express a redder chest patch than other males in the group [Bergman et al., 2009].

Social factors (i.e. propensity to live in gregarious groups or form polyspecific associations) and ecological factors (i.e. living in tropical, dense African humid environs) are correlated with complex facial patterns in Old World monkeys [Santana et al., 2013].

Neotropical primates, in contrast, are hypothesized to have evolved complex facial patterns when living in smaller groups and in closer vicinity to the equator [Santana et al.,

2012]. This may be explained through the knowledge that Old World monkeys rely more heavily on facial communication [Dobson, 2009; Kingdon, 2007; 1992; 1980]. ‘Flagging’ head movements of guenon communication, for example, may be aided by complex pelage colors and patterns on the face [Bradley and Mundy, 2008; Kingdon, 1992; 1980].

Although it is possible these complex signals have evolved to enhance individual recognition in larger groups [Santana, 2013], the general assumption for the

Cercopithecus radiation is that they function as species-recognition signals [Allen et al.,

2014].

Physiological functions such as regulation of body temperature and glare reduction from the sun may be another adaptive property of pelage color. For Homo

sapiens, variation in pigmentation serves as an adjustment to UV radiation. It is hypothesized this occurred through independent genetic pathways which were under positive selection [Joblanski and Chaplin, 2010]. However, for other mammals including non-human primates, the mechanism behind such functions remains unclear [Caro,

2005].

Pelage color and pattern may also serve as concealment from predators, and is divided into pattern blending, countershade, disruptive coloration, and background matching [Bradley and Mundy, 2008; Caro, 2005]. Pattern blending remains vastly unexplored, but may occur with some regularity within nocturnal primate species

[Bradley and Mundy, 2008]. Disruptive coloration, similarly unexplored, refers to irregular lines marking or breaking an outline of a body [Stevens et al., 2006]. This is commonly seen in black-and-white colobus monkeys (Colobus angolenesis and Colobus guereza). Countershade, dorsal-ventral color differences, is common in primates. It has been attributed to affects such as positional behavior, body size, and phylogeny [Kamilar and Bradley, 2011; Kamilar, 2009]. Countershade has been hypothesized to aid in concealment by making an animal appear more flattened and inconspicuous.

Primatologists have not yet explored the effects of background matching on primates.

However, complex phenotypes are not just dependent on functional hypotheses.

Developmental and genetic factors may also contribute to an observed phenotype

[Santana et al., 2012]. If this holds, the assumption that females use visual signals to recognize conspecific mates may be fallacious, especially for hybrid swarms. Genetic compatibility may be more important than recognizing conspecifics [Saetre, 2013]. This may be specifically true for hybrid populations, where the extent of hybridization is

expected to change over time. In fact, phenotypic hybrid index correlates high with genotypic hybrid index within a population in initial stages of hybridization, but as a population becomes further admixed, that correlation weakens [Bergman et al., 2008].

Moreover, genetic drift may be largely responsible for color variation within a group

[Peres et al., 1996]. Therefore, as the model for color evolution is further developed, an important area of study will be the role of color in hybrid populations.

Study Site

Gombe National Park, Tanzania

Gombe National Park (GNP) in Tanzania (440S, 2938E) is a hilly, and narrow wooded area that extends 16 km along Lake Tanganyika on the eastern shore. There are thirteen major valleys, which drain down the Albertine Rift escarpment into Lake

Tanganyika. Gombe experiences two seasons: a rainy season from November to May and a dry season from June to October. The landscape has supported extensive primate research over the past 50+ years. The forest is home to the Cercopithecus hybrid zone and a variety of other primates, including the chimpanzee (Pan troglodytes), olive baboon

(Papio anubis), red colobus (Colobus badius), vervet monkey (Chlorocebus aethiops), and needle-clawed bushbaby (Euoticus elegantulus) [Goodall, 1986].

There are four types of phenotypic Cercopithecus groups that occur within

Gombe: B groups (C. mitis), R groups (C. ascanius), RH groups (C. ascanius + hybrids), and RBH groups (C. ascanius, C. mitis, + hybrids). All group types are found throughout

the park, but RBH groups occur at a higher density in the central valleys, which border a southern uninhabitable range for Cercopithecus [Detwiler, 2010].

The habitat at Gombe was likely once connected to habitat forests in Burundi and

Rwanda [Goodall, 1986], but today Gombe is heavily fragmented and has seen a 64% reduction in forest cover outside the park [Pintea et al., 2011]. However, forest cover inside the park has significantly increased since 1974, likely due to prescribed forest fires

[Pintea et al., 2011]. This signals that the internal ecology of Gombe may have remained intact regardless of anthropogenic changes [Detwiler, 2010].

Research Questions

This thesis fall into two categories: extending current accepted methods in color analysis and analyzing the variation in pelage color within the hybrid zone at Gombe

National Park, Tanzania. In order to answer the first question, I tested variations of a new method in the field with a digital camera and an Xrite Colorchecker. To answer the second question, I applied a phenotypic hybrid index as a means to understand pelage diversity in the hybrid zone through the use of photographs of known individuals. My research questions were:

I. Color methods questions

1) Can objective color methods, employing RGB values, be developed to study

wild arboreal primates?

2) Can alternative color methods, using numerical non-RGB values, be

developed to study wild arboreal primates?

II. Hybrid zone pelage variation questions

1) Are there discrete forms of hybrid phenotypes at Gombe?

2) In a mixed-species hybrid group (RBH), are all phenotypes expressed?

3) In a mixed-species hybrid group (RBH), how are pelage phenotypes

distributed?

For this project, I applied an average PHI as a numerical non-RGB measure of pelage diversity within Gombe. I could not apply methods to extract RGB color values.

Future research will be able to couple PHI scores and Genotypic Hybrid Indices (GHI) for all known individuals in the population.

CHAPTER TWO: METHODS

Developing and Testing an Extension to Established Color Methods

Equipment and settings

I took photos for testing the method using a digital single-lens-reflex camera:

Canon EOS 70D mounted with a Canon ES 100-400mm F4.5-5.6L IS USM zoom lens.

For photographing the color charts, I kept the zoom distance at 400mm to avoid chromatic aberration (i.e. color distortion that may occur with a zoom lens). I attached a

Tiffen UV filter to the camera to reduce lens flare. I also made sure to place the color standard in the center of the frame to reduce lens distortions that may occur at the periphery of images [Galbany et al., 2015].

I calibrated test photos using an X-rite Colorchecker Classic, which contains 24 squares of nominal and known reflectance values. I captured all images in RAW and saved them in CR2 16-bit format, and later converted them to Tagged Image File Format

(TIFF) 16-bit format using Adobe Photoshop Lightroom 5. I used aperture priority with a constant low aperture (f/8) to avoid clipping [Stevens et al., 2007]. I took subsequent photos in manual mode, replicating previous photo settings. I set white balance to

‘cloudy’ for all photos to account for the tests made under varying canopy coverage conditions. I photographed all color cards against a dull background, which typically consisted of green foliage.

Method Description

I tested variations of an extended method of the sequential method developed by

Bergman and Beehner [2008] for objective color assessments in arboreal wild primates.

The first variation involved testing the ability to extend the time limit beyond two minutes, to ten minutes. The second analysis involved examining color accuracy between the test and control charts when they are not placed in the same locations and are 10 minutes apart. The final tests involved examining color accuracy when the control and test chart were 10-minutes apart, not in the same location, and under three different canopy coverage conditions: no coverage, partial coverage, and full coverage. I always took photos at midday (when the sun was at it’s apex) under full sun. This was typically between hour 12 and 16. I took all photos within a zero-degree to thirty-degree relation to the color-chart. Distance varied horizontally within three meters to six meters.

For tests that involved placing the control and test charts in different locations, I attempted to mimic lighting conditions on the second chart by replicating canopy cover and monitoring the RGB histogram output. To account for canopy cover in the forest, I replicated light conditions by approximating percentage of sky visible to me within a five-meter radius. I took ten photos under each of three canopy conditions: (1) full coverage, 0-10% sky visible in 5 meter radius; (2) partial coverage, 15-85% sky visible in a 5 meter radius; and (3) no coverage, 90-100% sky visible in a 5 meter radius. I made subjective spherical assessments instead of vertical assessments [Korhonen et al., 2006].

When placing the color checker under similar lighting conditions, I always directed a field assistant to where the color chart should be placed and they held it up (Fig. 1).

After transferring photos to a computer, I organized and sorted them in Adobe

Photoshop Lightroom 5. I discarded any photos with evident clipping. I analyzed photos in Adobe Photoshop CS5 using the inCamera 4.0.1 filter plug-in, a plug-in that is specified to values on the Xrite Colorchecker Classic. I set the light source value to D50 to correspond to light at the equator.

Linearization and equalization

To test for linearity and equality, I followed methods used to validate the sequential method [Bergman and Beehner, 2008]. The only modification I performed was multiple

linearization and equalization tests at various stages of testing. First, I tested linearization and equalization of the Canon 70D mounted with a Canon ES 100-400mm F4.5-5.6L IS

USM lens at an extended time mark of 10 minutes, instead of the accepted two minutes. I also tested the response of color accuracy when the test and control chart were placed in different locations, at the 10-minute time mark. Lastly, I tested the extension to ten minutes, in different locations, and under three different canopy conditions.

To test for linearity, I used linear regression to assess the relationship between the red, green, and blue color channels following the steps outlined by Bergman and Beehner

[2008]. I examined the relationship for the 6 grey squares for all photographs of a) at the

10-minute mark in the same location (3 color channels x 10 photos = 30 regressions), b) at the 10-minute mark in a different location (3 color channels x 10 photos = 30 regressions), and c) at the 10-minute mark in a different location under varying canopy scenarios (3 color channels x 3 canopy coverage conditions x 10 photos = 90 regressions). In total, I performed 150 regressions to test for linearity.

Method validation

I validated the extended sequential method using two color charts following the method validation established by Bergman and Beehner [2008]. I selected eight squares of color for method validation to mimic the color continuum present in the pelage expression of C. ascanius, C. mitis, and their hybrids: white (19), neutral 8 (20), neutral 5

(22), black (24), orange yellow (24), yellow (16), red (15), and brown (1). Values were not normally distributed; therefore, I performed a Wilcoxon-signed rank test to test

between medians of the control and test charts under all scenarios.

Generating a Phenotypic Hybrid Index

Camera guidelines for color photography

I followed guidelines for camera settings as described for established color methods

[Stevens et al., 2009]. I used the same camera set up I used to test the method. Although I did not color calibrate the photos of guenons in this thesis, I took steps to avoid color distortion between images. I took all photos in RAW to retain as much color information as possible. I took all photographs under aperture-priority, in an attempt to avoid clipping. In the same fashion of the test chart conditions, I took all photos at a low f-stop

(either f/5.6 or f/8.0) and set white balance to cloudy. I allowed the camera to set the ISO and shutter speeds automatically, with the maximum limit for ISO set to 12000. I used the RGB histogram function to review photos and certify that they were properly exposed. Due to the saturation that may occur when animals are against blue sky, I avoided using photos of animals against sky. Instead, I only used photos where animals were situated in front of dull backgrounds, typically leaf foliage. In contrast to photos for the color method, guenon photos were taken at various times of the day, under various canopy coverage conditions, and sometimes on cloudy or overcast days. I took photos of guenons at various focal lengths between 100 and 400mm.

Generating average phenotypic hybrid indices

I scored guenons under the guidelines of an adapted phenotypic hybrid index (PHI)

from Detwiler [2010]. I modified the definition of character states and switched to a 0 to

1 scale. I scored individuals (N = 55) for pelage traits that were C. ascanius, ascanius- like, intermediate, mitis-like, or C. mitis (Table I). I scored seven pelage color character states, including: the crown, inferior crown patch, check patches, nasal area, ventral surface, dorsal surface, and tail. I scored the pelage character states between 0 (C. mitis) and 1 (C. ascanius). I assigned hybrid categories to scores of 0.25 (mitis-like), 0.50

(intermediate), and 0.75 (ascanius-like). The computed average PHI took pelage color into consideration from a photographic record of each individual. Each record contained approximately 40 photos that contained adequate reference to multiple body parts.

In total, I computed an AVG PHI for individuals (n = 45) in Mkenke Group A that included all adults, all known juveniles size two or above, and resident or transient males.

I also computed an AVG PHI for a handful of individuals (n = 10) living outside of

Mkenke Group A. This sample included individuals from other RBH groups in lower

Mkenke, an R group in Niasanga/Kalande, two B groups in Mitumba, and solitary males who frequented research headquarters.

Analysis

In the field, I lumped individuals into one of three categories: C. ascanius (R), C mitis (B), or hybrid (H). For the monkeys in Mkenke Group A, I analyzed the relative frequency of phenotypic composition of individuals I sampled when individuals are placed in these three broad categories. I also computed the relative frequency of phenotypes in Mkenke Group A when the hybrid categories are expanded to: R-RH, R-H,

I-H, B-H, and B-BH. The groups were divided as: B (score: 0), B-BH (score: 0.01 -

0.19), B-H (score: 0.20 – 0.39), I-H (score: 0.40 – 0.59), R-H (score: 0.60 – 0.79), R-RH

(score: 0.80 – 0.99), and R (score: 1). This is not representative of group composition because not all juveniles are included in the analysis. For individuals outside Mkenke

Group A, I computed an AVG PHI in an attempt to sample a wider variety of phenotypic combinations at Gombe. I did not calculate relative frequency of phenotypes for these individuals because of a low sample size.

Table I. Phenotypic Characteristics for Hybrid Scoring using Pelage Morphology

C. mitis (0) Intermediate C. ascanius (1)

Crown black (crown color black with no or very reduced patch with triangular patch with sharply defined triangular continues down to little yellow, orange ticking yellow, orange ticking yellow, orange patch (yellow, orange nape) ticking ticking)

Inferior Crown Patch sharply defined reduced, variably no diadem narrow black frontal and diadem, light grey diadem present, variably colored diadem, or no temporal bands brown (buff speckling) colored diadem and splayed black frontal band 22 Cheek Patches high rounded, speckled grey with no boundary or light grey to dirty white dirty white broad creamy white and grey very reduced dark grey with variably sized dark with variably sized narrow black boundary boundary grey boundary black boundary

Nasal Area black, dark grey white spot, variably shaped, sharply outlined by contrasting color entirely surrounding spot

Ventral Surface black, grey grey dirty-white, pale grey dirty white white

Table I: continued on page 23

TABLE I: continued from page 22 C. mitis (0) Intermediate C. ascanius (1)

Dorsal Surface sharply defined black black interscapular band, no black interscapular brown with yellow-orange ticking interscapular band, rest brown or reddish brown band, brown or rest light grey brown reddish brown with buff speckling Tail proximinal 70% like mitis-like but with some reduced orange in ascanius-like but dorsal and ventral back, distal 30% reddish-brown in mid- mid-section of dorsal paler orange on surfaces (except base) black, ventral surface section of dorsal surface and ventral surface dorsal and ventral reddish orange, ventral dark reddish brown surfaces at base grey, distal

black

CHAPTER THREE: EXTENDED SEQUENTIAL METHOD

Method Validation: Linearization and Equalization

All permutations showed a high degree of linearity, but the Canon at the 10-minute mark produced slightly better results. The R2 range of these tests was 0.9866-0.9941, with a mean ± SD of 0.9893 ± 0.0003 (n = 30). For the Canon at the 10-minute mark with a difference in chart placement, the R2 range was 0.9776-0.9986, with a mean ± SD of 0.9944

± 0.0010 (n = 30). For all canopy conditions, the R2 range was 0.9366-0.9987, with a mean

± SD of 0.9884 ± 0.0016 (n = 90).

The Canon at the 10-minute mark demonstrated the highest degree of RGB equalization. Out of a maximum possible difference of 255, the absolute value of the difference was in the range of 0-9 (0.00% - 2.75%), with a mean ± SD difference of

1.883 ± 0.1164. The percentage of RGB values that were within 5% of each other was

93%. For the Canon at the 10-minute mark with a different chart placement, the absolute difference was in the range of 0-14 (0.00% - 5.49%), with a mean ± SD difference of

3.635 ± 0.2086. The percentage of RGB values that were within 5% of each other was

80%. At all canopy coverage conditions, the absolute difference was in the range of 0-25

(0.00% - 9.80%), with a mean ± SD difference of 4.60 ± 0.175. The percentage of RGB values that were within 5% of each other was 79%.

Accuracy and Precision

For the method extension to ten minutes, I took measurements from the control chart for various squares (white, neutral 8, neutral 5, black, orange yellow, yellow, red, and brown). I choose these squares to test accuracy and precision because they span the spectrum of pelage color outputs expressed in the hybrid zone. Results from Wilcoxon- signed ranks tests indicate that the neutral 5, black, and red swatches do not respond well in terms of color accuracy under these extended conditions (Table II). In contrast, the white, neutral 8, orange yellow, yellow, and brown swatches do represent color accurately to these extensions (Table II).

I also analyzed how the extended method to ten minutes under different placements would respond to color. Similarly, I took measurements from both the control and test charts in order to compare means using the Wilcoxon-signed ranks test. Results indicate most color swatches do not relay color information accurately. The only color square that did not show a significant difference between the test and control chart was the yellow square (Table III).

To assess how the method would respond under variations in canopy cover, I combined the control and test values for the different canopy scenarios (10 photos x 3 conditions = 30 photos). I also compared means using the Wilcoxon-signed ranks test.

Results here also indicate most color squares do not relay color information accurately.

The only squares that relay color information accurately are orange yellow and red (Table

IV).

Table II. Statistical Comparisons between Test and Control Squares for the Extended Color Method at the 10-minute Extension

Square W Z d.f. P

White 7.5 -1.472 9 0.141

Neutral 8 12 -0.338 9 0.735

Neutral 5 0 -2.803 9 0.005

Black 0 -2.803 9 0.005

Orange Yellow 9 -1.886 9 0.059

Yellow 19 -0.415 9 0.678

Red 0 -2.803 9 0.005

Brown 14 -1.007 9 0.314

Table III. Statistical Comparisons between Test and Control Squares for an Extension to Color Methods, at the 10-minute Extension in Different Placement Areas

Square W Z d.f. P

White 0 -2.803 9 0.005

Neutral 8 0 -2.803 9 0.005

Neutral 5 0 -2.803 9 0.005

Black 0 -2.805 9 0.005

Orange Yellow 0 -2.293 9 0.022

Yellow 19 -0.866 9 0.386

Red 0 -2.803 9 0.005

Brown 0 -2.803 9 0.005

Table IV. Statistical Comparisons between Test and Control Squares for an Extension to Color Methods, at the 10-minute Extension in Different Placement Areas and Under Varying Canopy Conditions

Square W Z d.f. P*

White 0 -4.762 29 0

Neutral 8 0 -4.268 29 0

Neutral 5 0 -4.395 29 0

Black 0 -3.165 29 0.002

Orange Yellow 5 -0.257 29 0.797

Yellow 19 -3.034 29 0.002

Red 0 -0.566 29 0.572

Brown 0 -3.342 29 0.001

CHAPTER FOUR: PHENOTYPIC HYBRID INDEX OF C. MITIS DOGGETTI, C.

ASCANIUS SCHMIDTI, AND HYBRIDS RESULTS

I took approximately 12,000 photos during my field season from July 1st to

November 4th of 2014. I extracted a sub-set of high-resolution images for the purpose of this thesis (N = ~2000) to calculate a Phenotypic Hybrid Index (PHI) for all known individuals in Gombe National Park, Tanzania. Using three basic phenotypic categories, it is clear that C. ascanius individuals exceed other phenotypes in quantity within

Mkenke Group A. However, the composition of phenotypic diversity in the hybrids is lost (Fig. 2).

Quantitative Pelage Variation of Guenons in Mkenke Group A

Using basic phenotype scores renders Mkenke Group A, a habituated and known

RBH group, as 20% B, 29% H, and 51% R. With the addition of developing a robust

AVG PHI, I was able to expand those categories as established by Detwiler [2010] to include B-BH, B-H, IH, R-H, and R-RH. I confirm that the phenotypic characteristic of

Mkenke Group A is skewed towards ascanius-like, and a more robust statistical picture begins to form in regards to phenotypic outcomes for the RBH group (Fig. 3). In this regard, of the individuals sampled, the PHI renders the group: 21% B, 2% B-BH, 2% B-

H, 7% I-H, 11% R-H, 7% R-RH, and 51% R. These results indicate a propensity of the hybrid phenotypes in Mkenke Group A to lean towards the ascanius-phenotype. None of

these numbers represent actual group composition; they function as a representation of gross phenotypic diversity in the sampled RBH group.

Fig. 2. Relative frequency of phenotypes, using three hybrid categories for MkGrpA.

Fig. 3. Relative frequency of phenotypes, using expanded hybrid categories, for MkGrpA.

Table V. Scored Phenotypic Characteristics for Guenons of Mkenke Group A (MkGrpA)

ID Field PH CR DD CP NS VS DS T PHI QUAL

Females of Mkenke Group A

Carson B 0 0 0 0 0 0 0 0 B

Dove B 0 0 0 0 0 0 0 0 B

Jenny B 0 0 0 0 0 0 0 0 B

Amy R 1 1 1 1 1 1 1 1 R

Bea R 1 1 1 1 1 1 1 1 R

Delaya R 1 1 1 1 1 1 1 1 R

Elsa R 1 1 1 1 1 1 1 1 R

Flower R 1 1 1 1 1 1 1 1 R

Hope R 1 1 1 1 1 1 1 1 R

Ire R 1 1 1 1 1 1 1 1 R

Joly R 1 1 1 1 1 1 1 1 R

Table V: continued on page 31

TABLE V: continued from page 30 ID Field PH CR DD CP NS VS DS T PHI QUAL

Females of Mkenke Group A (continued)

Liza R 1 1 1 1 1 1 1 1 R

Pinda mkia R 1 1 1 1 1 1 1 1 R

Ruby R 1 1 1 1 1 1 1 1 R

Sara R 1 1 1 1 1 1 1 1 R

Sande R 1 1 1 1 1 1 1 1 R

Savannah R 1 1 1 1 1 1 1 1 R

Venus R 1 1 1 1 1 1 1 1 R

Viola R 1 1 1 1 1 1 1 1 R

Zalia B 0 0 0 0 .75 0 0 0.11 B-BH

Channel H .50 1 .75 1 1 1 .75 .86 R-RH

Gopiss H .25 .25 .50 1 .75 .50 .50 .54 I-H

Jackie H .25 .25 .50 .75 .75 1 .50 .57 I-H

TABLE V: continued on page 32

TABLE V: continued from page 31 ID Field PH CR DD CP NS VS DS T PHI QUAL

Females of Mkenke Group A (continued)

Mercy H .50 .50 .50 1 .75 1 .25 .64 R-H

Pieces H .50 .75 .50 1 1 1 .25 .71 R-H

Tatu H .50 .75 .75 1 1 1 .75 .82 R-RH

Uma H .50 1 .50 1 .50 / .50 .67 R-H

Zoo H .50 .75 .75 1 .75 1 .75 .78 R-H

Juveniles of Mkenke Group A

Andy B 0 0 0 0 0 0 0 0 B

Pablo B 0 0 0 0 0 0 0 0 B

Tisa B 0 0 0 0 0 0 0 0 B

BJ11 B 0 0 0 0 0 0 0 0 B

Kasulu R 1 1 1 1 1 1 1 1 R

Jinja R 1 1 1 1 1 1 1 1 R

TABLE V: continued on page 33

TABLE V: continued from page 32 ID Field PH CR DD CP NS VS DS T PHI QUAL

RJ11 R 1 1 1 1 1 1 1 1 R

RJ12 R 1 1 1 1 1 1 1 1 R

RJ31 R 1 1 1 1 1 1 1 1 R

Jimmy H .25 .25 .25 1 .25 .50 .25 .40 I-H

Juveniles of Mkenke Group A (continued)

Tarzan H .75 .75 .75 .75 1 1 1 .86 R-RH

33 1

HJ2 H 0 .25 .25 1 / .25 / .35 B-H

Males of Mkenke Group A

George B 0 0 0 0 0 0 0 0 B

Simba B 0 0 0 0 0 0 0 0 B

Blue R 1 1 1 1 1 1 1 1 R

Eric R 1 1 1 1 1 1 1 1 R

Zeus H .50 .75 .75 1 1 1 .25 .75 R-H

Quantitative Pelage Variation of Guenons outside Mkenke Group A

I took opportunistic photos outside of Mkenke Group A, including two trips to

Mitumba valley in the north and one trip to Niasanga and Kalande valleys in the South in an attempt to sample C. ascanius and C. mitis pelage variation in males ranging outside of the central valleys. I only sampled adult females in neighboring groups when I had a sufficient sample of photographs to differentiate the guenon as an individual. Although they are distinct from each other, I did not give them names because our field assistants are unfamiliar with them. I generated an AVG PHI for all distinct individuals (Table VI).

Table VI. Scored Phenotypic Characteristics for Guenons Outside Mkenke Group A (MkGrpA)

ID Field PH CR DD CP NS VS DS T PH QUAL

Females in lower Mkenke, Groups B or C

HF1 H .50 .25 .75 1 .50 .25 .25 .50 I-H

HF2 H 0 0 0 1 0 0 0 .14 B-BH

HF3 H .25 .50 .25 1 0 .25 .25 .36 B-H

Males in lower Mkenke, Group B or C

KP R 1 1 1 1 1 1 1 1 R

Solitary males in research camp

BM1 B

RM1 R 1 1 1 1 1 1 1 1 R

RM2 R 1 1 1 1 1 1 1 1 R

Males in Mitumba Valley

BM2 B 0 0 0 0 0 0 0 0 B

TABLE VI: continued on page 36

TABLE VI: continued from page 35 ID Field PH CR DD CP NS VS DS T PHI QUAL

BM3 B 0 0 0 0 0 0 0 0 B

Males in Niasanga/Kalande Valley

RM3 R 1 1 1 1 1 1 1 1 R

Qualitative Pelage Variation of Guenons at Gombe

C. mitis doggetti pelage coloration

C. mitis doggetti has a countershade between their ventral and dorsal surfaces of

1.69 [Kamilar, 2009]. The ventral surface is black or grey, and the dorsal surface is light grey or brown or with buff specking. The dorsal surface also has a black interscapular band that extends from the dorsal neck area and encapsulates are upper arms [Groves,

2002]. Limbs are also black, but legs may be tickled grey. The tail blends from an off- black color on the proximal end (upper 70%), to a full black color at the distal end (lower

30%) (Fig. 5). Some individual tails have a reddish brown hue on the ventral surface of their tails. The nasal area is black or dark grey, with low contrast and no clear or visible

“spot”. However, individuals sometimes have white buccal hairs that encapsulate the nasal region. The cheek patches are speckled grey with a high rounded shape. The diadem is present, and sharply defined by a light grey brown color with buff specking

(Fig. 4). The crown coloration is black, and it continues down to the nape; it forms an interscapular band that extends dorsally (Table 1).

C. ascanius schmidti pelage color variation

C. ascanius schmidti individuals have low variability in their pelage color and pattern across populations. Monkeys exhibit a countershade value between their ventral and dorsal surfaces of 2.87 [Kamilar, 2009]. The ventral surface is white and the dorsal surface is brown with orange ticking (Fig. 6). The check hair color may vary [Colyn,

1988]. At Gombe, the nose-spot is variably sized, sometimes heart-shaped, and always white. Individuals also exhibit black sharply defined black band that surrounds the cheeks

(Fig. 7). The crown is sharply defined with yellow orange ticking, and the inferior crown patch is in the shape of a black frontal band (Fig. 8). The dorsal and ventral surfaces of the tail are reddish brown; the base is black on the distal end. The interscapular hair banding patterns consist of two-three pairs of red/black bands [Groves, 2002].

C. ascanius x C. mitis pelage coloration

There are three variations of hybrid phenotypes recognized between C. ascanius and C. mitis previously identified by Detwiler [2010]. Hybrid phenotypes seem to express a random patterning and color variation between individuals. Most hybrid phenotypes express the C. ascanius white nose-spot but others do not. Generally, the loss of the white nose-spot appears the further into the mitis-phenotype an individual’s coat color becomes.

There are two mitis-like hybrid phenotypes, and although a variety of combinations can express themselves in the phenotype, the most telling sign is the presence of a white-nose spot, even if faint (Fig. 9). Some individuals may lack a white

nose-spot, and instead may express a hybrid-like or ascanius-like ventral surface. In contrast to B-BH individuals, B-H monkeys exhibit a C. mitis phenotype with a stronger inclination towards C. ascanius. This may be exhibited expressed as a C. mitis phenotype with a fully C. ascanius white nose-spot and variably colored diadem (Fig. 10).

An intermediate hybrid exhibits a phenotype directly in-between a mitis-like and ascanius-like hybrid with coloration patterns of both parental species present. In this respect, a multitude of color and pattern combinations may contribute to the observed phenotype. Individuals usually have some traits that score full C. mitis, some that score full C. ascanius, and the rest land somewhere on the spectrum between. For example, an individual may express a white C. ascanius nose spot, a diadem that is variably colored,

and cheek patches that are light grey with an undefined boundary (Fig 11). Tail coloration can vary between mitis-like, ascanius-like, or intermediate (i.e. Fig. 12).

There are two ascanius-like hybrids including R-H and an R-RH. Similar to the mitis-like hybrid, these individuals look C. ascanius except they their pelage colors are less pronounced (i.e. creamy or off-white cheek patches) (Figs. 13 and 14). The tail may be pale-orange in contrast to the reddish-orange of C. ascanius. Individuals in these hybrid categories always express a white nose-spot.

CHAPTER 5: DISCUSSION AND CONCLUSION

Assessing Color Variation in Wild Arboreal Primates

After preliminary testing of basic extensions to established color methods, results indicate that color methods that examine RGB values cannot be developed at this time

(Fig. 2-4) for wild arboreal monkeys. Even though tests show a fair to good amount of linearization and equalization, the accuracy and precision make the method irreproducible. During the most conservative testing, at the 10-minute mark, placing both the control and test chart in the same location under full sun – some of the color squares

(neutral 5, black, and red) had a significant amount of variation between both charts.

Without such a method in place, the problem of not being able to statistically analyze color variation in wild arboreal primates remains. However, alternative color methods that employ numerical non-RGB values can be developed to add statistical meaning to the study of color in wild arboreal primates. Methods that capture basic quantitative pelage diversity may even be more feasible for field primatologists.

One avenue of further research is the implementation of color scoring using digital photography, similar to the methods used in this thesis to generative phenotypic hybrid indices. Although the phenotypic hybrid index is modeled loosely around previous primate hybrid studies [Bergman and Beehner, 2004; Delmore et al., 2011; Nagel, 1973;

Phillips-Conroy and Jolly, 1986] where morphological trait scoring is important, I argue this can be expanded beyond hybrid primates. Color scoring has already been used successfully to study pelage coloration in wild western gorillas (Gorilla gorilla) where

using a color card was not a feasible solution [Breuer et al., 2007]. In a similar fashion, a scoring system has also been used for coat and tail condition evaluation successfully

[Berg et al., 2009] in wild Lemur catta. Therefore, a color scoring system may prove to be an accurate estimate of pelage color variation when the adjacent or sequential methods cannot be used.

Future research should focus on the diversity of variation that such a scoring system can generate, and it may be useful to link pelage scores to color chart swatches.

This is easily tested in museum collections, where photos can be captured under standardized lighting conditions with a color card. Subsequently, these same animals can be subjected to a less objective “matching” scoring system. The wider implications are that field primatologists should be able to compare the pelage of wild arboreal monkeys across populations, and seemingly, “hybrid” or “new” morphs could be quantitatively described in comparison to parental phenotypes or founder species populations. Efforts should be made by field primatologists to compile living digital collections of phenotypic diversity, where phenotype is measured numerically and recorded with other allometry data points.

For researchers that aim to garner an even more robust understanding of pelage color variation within or across a species, additional methods may be available. A robust pelage morphological analysis may also involve examining hair banding patterns and hair texture – information that can be obtained from museum specimens. Although it is highly unlikely all researchers will find a desire to assess pelage morphology on such a large scale, the field will garner a new understanding of pelage variation by moving away from

qualitative color descriptions and towards methods that allow for statistical testing – especially for highly diverse radiations such as Cercopithecus primates.

Hybrid Coloration within Gombe National Park

The data from phenotypic hybrid scoring for pelage morphology show C. mitis x

C. ascanius individuals at Gombe comprise an assortment of phenotypes. No discrete hybrid phenotypes arise. Instead, there appears to be a continuum that is expressed between C. ascanius and C. mitis. Using photographs, the five previously established hybrid categories by Detwiler [2005] were easily teased out, showing results similar to using high-powered binoculars in the field. The added benefit of using a phenotypic hybrid index post-production through photographs is the ability to compare between groups in the park and against hybrid and non-hybrid populations outside of the park.

High-resolution photographs may also serve as guidelines coupled to textual PHI descriptions for scoring hybrid individuals between C. mitis doggetti and C. ascanius schmidti.

The results indicate that Mkenke Group A, an RBH group, shows a positive skew towards an ascanius-like phenotype, with 23 individuals scoring C. ascanius and the majority of hybrids falling in the I-H, R-H, or R-RH categories. Only one juvenile

(unknown sex) scored B-H and one adult female in the group scored B-BH. Previous assessments of C. mitis x C. ascanius hybrids at Kibale argued individuals were intermediate between both parental forms, but with a mitis-like larger body size and less distinctive color pattern [Struhsaker et al., 1988]. One explanation behind the

incongruence between Gombe and Kibale assessment of hybrid phenotypes involves sample size – we were able to sample 55 guenons at Gombe and 15 phenotypic hybrids.

Another likely explanation is that without a phenotypic hybrid index or color scoring methods, certain traits and variations become lost as the reliance in mostly qualitative descriptions is subjective and does not allow robust comparisons across populations.

Furthermore, with only one fully sampled group, it is difficult to decipher what phenotypic distribution exists in the population, because other hybrid groups may prove to be mitis-like, or intermediate.

All pelage phenotypes were expressed in the mixed-species hybrid group Mkenke

Group A. It is unclear if this will play out in all RBH groups. From the preliminary assessments, most phenotypic hybrids express the C. ascanius white nose spot. Even a mitis-like B-BH may show faint signs of the nose-spot. This evidence suggests that there is possibly an adaptive significance to the C. ascanius nose-spot or that it is a dominant trait. Interestingly, a hybrid between Cercopithecus mitis albogularis x Chlorocebus pygerythrus in Kenya expressed a pale grey nose-spot, absent in both parental species.

The authors suggest that the trait may have been present in a common ancestor of

Cercopithecus and Chlorocebus (c. 8.1 mya) [de Jong and Butynski, 2010]. Functional significance, and thus adaptive value, may be deduced from direct evidence of a receiver’s response to the visual signal and further testing is needed for C. ascanius and species with similar facial characteristics, such as, the greater spot-nosed monkey (C. nictitans) and the lesser spot-nosed monkey (C. petaurista).

The polytypic nature of C. mitis may play a key role in the phenotypic variation of the Gombe hybrid zone. The phylogeny of C. mitis is highly debatable and most

researchers recognize 16 sub-species or more [Groves, 2001; 2005; Grubb, 2001; Grubb et al., 2003; Kingdon, 1997]. Within C. mitis, a large variation of pelage color is also seen. Currently the variation is explained as no more than species differences; however, this is complicated by the propensity of C. mitis to hybridize and create novel forms – sometimes with traits that do not resemble either parental species. Within C. mitis doggetti alone, the likely parental species at Gombe, there is a substantial amount of pelage variation in the dorsal flank.

At Gombe, significant variation in dorsal flank coloration is seen and varies from brown to grey. Interestingly, two Y-DNA haplotypes have been found in Gombe (i.e. the

“doggetti/stuhlmanni” haplotype and the “Lomami River” haplotype). Two waves of hybridization that brought emigrating C. mitis males are hypothesized to have occurred at

Gombe [Detwiler, 2010]. Phenotypic and genetic data thus suggests the possibility that two phenotypically distinct C. mitis emmigrations may be responsible for the some of the observed phenotypic diversity in the park. Where did these C. mitis males originate?

Genetic evidence from the north of Gombe (Nyungwe National Park, Rwanda) and to the south of Gombe (Mahale National Park, Tanzania) shows these two populations of C. m. doggetti are more similar to C. m. stuhlmanni than to each other [Detwiler, 2010]. More genetic sampling from the surrounding areas, including the Democratic Republic of

Congo, is needed.

FURTHER THOUGHTS

All phenotypes occur at Gombe – including full C. mitis and C. ascanius parental forms. Interestingly, all C. mitis individuals at Gombe show mixed genetic ancestry.

Specifically, all individuals express one of two C. ascanius mitochondrial haplotypes

(ascanius haplotype 8 or ascanius haplotype 9). Within Kibale, hybrid females of C. mitis x C. ascanius backcrossed with parental males, and none of the resulting offspring were distinguishable from the parental species [Struhsaker et al., 1988]. Gombe represents a different situation because hybridization has likely been occurring in the population >

200 years [Detwiler, 2010]. How long it takes for parental forms to re-emerge in the

Gombe population is currently unclear.

Without a way to analyze RGB coloration in the Gombe hybrid zone at this time,

I suggest expanding robust phenotypes by coupling PHI with other data. Two important non-invasive phenotypic measurements have been unexplored in this population: photogrammetry and vocal signal analysis. For many primate species, including large- bodied Gorilla gorilla gorilla and small-bodied Procolobus rufomitratus, photogrammetric techniques have been used successfully [Galbany et al., 2015; Rothman et al., 2008]. C. ascanius schmidti and C. mitis doggetti are known to differ substantially in body size [Estes, 1991], and results from the hybrid body size in the population may serve as an additional measurement of phenotypic variability between parental species and hybrids. Loud calls are often a reliable indicator of species in most primates [Gautier et al., 1988; Fuller, 2013; Macedonia and Taylor, 1985]. Evidence now links hybrid calls

in howlers to an “intermediate call” between both parental species [Cortéz-Ortiz et al.,

2015; Kitchen et al., 2014]. In order to gain a full understanding of the outcomes of hybridization at Gombe, a robust phenotypic study coupled to additional genetic data is needed.

Furthermore, behavioral data currently indicates that at a basal level, all phenotypes (R, B, and H) mate promiscuously [Detwiler, 2002]. I suggest that future studies examine mating patterns while accounting for expanded hybrid phenotypes (B-

BH, B-H, I-H, R-H, and R-RH) instead of the three basic sub-divisions. If evidence of promiscuous mating continues, this may signal a breakdown of species-specific visual signals. Although the current hypothesis is that guenons use visual signals (facial color and pattern) to recognize conspecific mates [Allen et al., 2014; Allen and Higham, 2015], females may not need to categorize mates as conspecifics or heterospecifics to reproduce successfully. Individuals only need to find mates with whom they are genetically compatible [Saetre, 2013]. For a radiation that is hypothesized to still be undergoing active speciation [Gautier-Hion et al., 1988], these visual signals may still be evolutionarily and thus functionally plastic.

APPENDICES

53 A – IACUC Approval

REFERENCES

Alberts SC, Altmann J. 2001. Immigration and hybridization patterns of yellow and Anubis baboons in and around amoseli, Kenya. American Journal of Primatology 53:139-154.

Allen WL, Stevens M, Higham JP. 2014. Character displacement of Cercopithecini primate visual signals. Nature Communications doi:10.1038/ncomms5266.

Allen WL, Higham JP. 2015. Assessing the potential information content of multicomponent visual signals: a machine learning approach. Proceedings of the Royal Society B 282:20142284.

Alin SR, O’Reilly CM, Cohen AS, Dettman DL, Palacios-Fest MR, McKee BA. 2002. Effects of land-use change on aquatic biodiversity: a view from the paleorecord at Lake Tanganyika, East Africa. Geology 30:1143-1146.

Arnold ML. 1997. Natural hybridization and evolution. Oxford: Oxford University Press.

Barton NH, Hewitt GM. 1985. Analysis of hybrid zones. Annual review of ecology, evolution, and systematics 16:113-148.

Beaudrot L, Stuebig MJ, Meijaard E, et al. 2013. Interspecific interactions between primates, birds, bats, and squirrels may affect community composition on Borneo. American Journal of Primatology 75:170-185.

Bergman TJ, Beehner JC. 2008. A simple method for measuring colour in wild animals: validation and use on chest patch colour in geladas (Theropithecus gelada). Biological Journal of the Linnean Society 94:23-240.

Bergman TJ, Beehner JC. 2004. Social system of a hybrid baboon group (Papio anubis x P. hamadrryas). International Journal of Primatology 25:1313-1330.

Berg W, Jolly A, Rambeloarivony H, Andrianome V, Rasamimanana H. 2009. A scoring system for coat and tail condition in ringtailed lemurs, Lemur catta. American Journal of Primatology 71:183-190.

Bergman TJ, Phillips-Conroy JE, Jolly CJ. 2008. Behavioral variation and reproductive success of male baboons (Papio abunis x Papio hamadryas) in a hybrid social group. American Journal of Primatology 70:136-147.

Bernstein IS. 1966. Naturally occurring primate hybrid. Science 154:1559-1560.

Booth AH. 1968. Taxonomic studies of Cercopithecus mitis Wolf (East Africa). National Geographic Society Research Reports (1963 projects):37-51.

Bradley BJ, Mundy NI. 2008. The primate palette: the evolution of primate coloration. Evolutionary Anthropology 17:97-111.

Breuer T, Robbins MM, Boesch C. 2007. Using photogrammetry and color scoring to assess sexual dimorphism in wild western gorillas (Gorilla gorilla). American Journal of Physical Anthropology 134:369-382.

Brown M. Food and range defence in group-living primates. Animal Behaviour 85:807- 816.

Butynski TM, De Jong Y. 2012. Survey of the primates of Loita Hilla, Kenya. Unpublished report to Primate Conservation Inc., Rhode Island, USA.

Butynski TM. 2002. The guenons: an overview of diversity and taxonomy. In: Glenn ME, Cords M, editors. The guenons: Diversity and adaptation in African monkeys. New York: Kluwer Academic/Plenum. p 3-13.

Caro T. 2013. The colours of extant mammals. Seminars in Cell and Developmental Biology. 24:542-552.

Caro T. 2005. The adaptive significance of coloration in mammals. BioScience 55:125- 136.

Chapman CA, Chapman LJ, Cords M, et al. 2002. Variation in the diets of Cercopithecus species: differences within forests, among forests, and across species. In: Glenn ME, Cords M, editors. The guenons: Diversity and adaptation in African monkeys. New York: Kluwer Academic/Plenum. p 325-350.

Colyn MM. 1988. Distribution of guenons in the Zaire-Lualaba-Lomami river system. In: Gautier-Hion A, Bourliere F, Gautier JP, Kingdon J, editors. A primate radiation: Evolutionary biology of the African guenons. Cambridge: Cambridge University Press. p 104-124.

Cords M. 1988. Mating systems of forest guenons: a preliminary review. In: Gautier- Hion A, Bourliere F, Gautier JP, Kingdon J, editors. A primate radiation: Evolutionary biology of the African guenons. Cambridge: Cambridge University Press. p 323-329.

Cords M. 1987. Mixed-species association of Cercopithecus monkeys in the Kakamega forest, Kenya. Berkeley and Los Angeles: University of California Press. p 118.

Cords M. 1984. Mating patterns and social structure in redtail monkeys (Cercopithecus ascanius). Zeitschrift für Tierpsychologie 64:313-329.

Cortés-Ortiz L, Agostini I, Aguiar LM, et al. 2015. Hybridization in howler monkeys: current understanding and future directions. In: Kowalewski M, Garber P, Cortés- Ortiz L, Urbani B, Youlatos D, editors. Howler monkeys: Adaptive radiation, systematics, and morphology. New York: Springer. p 107-131.

Delmore KE, Louis EE Jr, Johnson SE. 2011. Morphological characterization of a brown lemur hybrid zone (Eulemur rufifrons x E. cinereiceps). American Journal of Physical Anthropology 145:55-66.

Detwiler KM. 2010. Natural hybridization between Cercopithecus mitis x. C. ascanius in Gombe National Park, Tanzania (doctoral dissertation). Retrieved from ProQuest Dissertations and Theses Global. (3396649)

Detwiler KM, Burrell AS, Jolly CJ. 2005. Conservation implications of hybridization in African Cercopithecine monkeys. International Journal of Primatology 26:661- 684.

Detwiler KM. 2002. Hybridization between red-tailed monkeys (Cercopithecus ascanius) and blue monkeys (C. mitis) in East African Forests. In: Glenn ME, Cords M, editors. The guenons: Diversity and adaptation in African monkeys. New York: Kluwer Academic Press/Plenum. p 79-97.

Dobson SD. 2009. Socioecological coorelates of facial mobility in nonhuman anthropoids. American Journal of Physical Anthropology 139:413-420.

Dowling TE, Secor CL. 1997. The role of hybridization and introgression in the diversification of animals. Annual review of ecology, evolution, and systematics 28:593-619.

Estes RD. Guenons, mangabeys, and baboons: subfamily Cercopithecinae. In: Estes RD, editor. The Behavior guide to African mammals including hoofed mammals, carnivores, primates. Berkeley and Los Angeles: University of California Press. p 489-519.

Endler J. 1990. On the measurement and classification of colour in studies of animal colour patterns. Biological Journal of the Linnean Society 41:315-252.

Fuller JLR. 2013. Diversity of form, content, and function in the vocal signals of adult male blue monkeys (Cercopithecus mitis stuhlmanni): an evolutionary approach to understanding a signal repertoire (doctoral dissertation). Retrieved from ProQuest Dissertations and Theses Global. (3546577)

Galbany J, Stoinski TS, Disier A, et al. 2015. Validation of two independent photogrammetric techniques for determining body measurements of gorillas. American Journal of Primatology. Doi:1002/ajp/22511.

Gause GF. 1934. The struggle for existence. Baltimore, MD: Williams and Wilkins.

Gautier-Hion A. 1988. The diet and dietary habits of forest guenons. In: Gautier-Hion A, Bourliere F, Gautier JP, editors. A primate radiation: Evolutionary biology of the African guenons. Cambridge: Cambridge University Press. p 257-283.

Gautier JP. 1988. Interspecific affinities among guenons as deduced from vocalizations. In: Gautier-Hion A, Bourliere F, Gautier JP, editors. A primate radiation: Evolutionary biology of the African guenons. Cambridge: Cambridge University Press. p 194-226.

Goodall J. 1986. The chimpanzees of Gombe: patterns of behavior. Cambridge: Harvard University Press.

Goodall J. 2000. Africa in my blood: An autobiography in letters – the early years. Boston: Houghton Mifflin.

Grant PR, Grant BR. 1992. Hybridization of bird species. Science 256:193-197.

Groves CP. 2002. Primate Taxonomy. Washington, DC: Smithsonian Institution Press.

Grubb P, Butynski TM, Oates JF, Beader SK, Disotell TR, Groves CP, Struhsaker TT. 2003. Assesment of the diversity of the African primates. International Journal of Primatology 24:1301-1357.

Harvey PH, Martin RD, Clutton-Brock TH. 1987. Life histories in comparative perspective. In: Smuts BB, Cheney DL, Seyfarth RM, Wragham RW, Struhsaker TT, editors. Primate Societies. Chicago: University of Chicago Press. p 181-196.

Higham JP, Pfefferle D, Heistermann M, Maestripieri D, Stevens M. 2013. Signaling in multiple modalities in male rhesus macaques: sex skin coloration and barks in relation to androgen levels, social status, and mating behavior. Behavioral Ecology and Sociobiology 67:1457-1469.

Joblonski NG, Chaplin G. 2010. Human skin pigmentation as an adaptation to UV radiation. Proceedings of the National Academy of Science 107:8962-8968.

Jolly CJ. 2001. A proper study of mankind: analogies from the papionin monkeys and their implications for human evolution. Yearbook of Physical Anthropology 44:177-204.

Kamilar JM, Heesy CP, Bradley BJ. 2013. Did trichromatic color vision and red hair color coevolve in primates? American Journal of Primatology 75:740-751.

Kamilar JM, Bradley BJ. 2011. Interspecific variation in primate coat colour supports Gloger’s rule. Journal of Biogeography 38:2270-2277.

Kamilar JM. 2009. Interspecific variation in primate countershading: effects of activity pattern, body mass, and phylogeny. International Journal of Primatology 30:877- 891.

Kingdon J. 1992. Facial patterns as signals and masks. In: S Jones, MORE, editors. The Cambridge Encyclopedia of Human Evolution. p 161-167.

Kingdon J. 2007. Primate visual signals in noisy environments. Folia primatologica 78:389-404.

Kingdon J. 1988. What are the face patterns and do they contribute to reproductive isolation in the guenons? In: Gautier-Hion A, Bourliere F, Gautier JP, J Kingdon, editors. A primate radiation: Evolutionary biology of the African guenons. Cambridge: Cambridge University Press. p 227-245.

Kingdon J. 1980. The role of visual signals and face patterns in African forest monkeys (guenons) of the genus Cercopithecus. Transactions of the Zoological Society of London 35:425-475.

Kitchen DM, de Cunha RGT, Holzmann I, Oliveira DAG. 2015. Function of loud calls in howler monkeys. In: Kowalewski M, Garber P, Cortés-Ortiz L, Urbani B, Youlatos D, editors. Howler monkeys: Adaptive radiation, systematics, and morphology. New York: Springer. P 369-399.

Leakey M. 1988. Fossil evidence for the evolution of guenons. In: A. Gautier-Hion, F. Bourlière, JP Gautier, J Kingdon, editors. A primate radiation: Evolutionary biology of the African guenons. Cambridge: Cambridge University Press. p 7-12.

Lehman N, Eisenhawer AH, K. Mech LD, Peterson RO, Gogan PJP, Wayne RK. 1991. Introgression of coyote mitochondrial DNA into sympatric North American gray wolf populations. Evolution 45:104-119.

Macedonia JM, Taylor LL. 1985. Subspecific divergence in a loud call of the ruffed lemur (Varecia variegata). American Journal of Primatology 9:295-304.

Nagel U. 1973. A comparision of anubis baboons, hamadryas baboons and their hybrids at a species border in Ethiopia. Folia primatologica 19:104-165.

Napier PH. 1981. Catalogue of primates in the British Museum (Natural History). Part II: Family Cercopithecidae, Subfamily Cercopithecinae. London: Bristish Museum (Natuarl History).

Peres CA, Patton JL, daSilva MNF. 1996. Riverine barriers and gene flow in Amazonian saddle-back tamarins. Folia primatologica 67:113-124.

Phillips-Conroy JE, Jolly CJ. 1986. Changes in the structure of the baboon hybrid zone in the Awash National Park, Ethiopia. American Journal of Physical Anthropology 71:337-350.

Pintea L, Pusey AE, Wilson ML, et al. 2011. Long-term ecological changes affecting the chimpanzees of Gombe National Park, Tanzania. In: Plumptre AJ, editor. The ecological impact of long-term changes in Africa’s rift valley. New York: Nova Science Publishers. p 194-210.

Randler C. 2002. Avian hybridization, mixed pairing and female choice. Animal Behavior 63:103-119.

Rhymer JM, Simberloff D. 1996. Extinction by hybridization and introgression. Annual Review of Ecology, Evolution, and Systematics 27:83-109.

Rollison J, Martin RD. 1981. Comparative aspects of primate locomotion, with special reference to arboreal cercopithecines. Symposia of the Zoological Society of London: 48:377.

Rothman JM, Chapman CA, Twinomugisha D, et al. 2008. Measuring physical traits of primates remotely: the use of parallel lasers. American Journal of Primatology 70:1-5.

Saetre GP. 2013. Hybridization is important in evolution, but is speciation? Journal of Evolutionary Biology 26:256-258.

Santana SE, Alfaro JL, Noonan A, Alfaro ME. 2013. Adaptive response to sociality and ecology drives the diversification of facial colour patterns in catarrhines. Nature Communications 4:2765.

Santana SE, Alfaro JL, Alfaro ME. 2012. Adaptive evolution of facial colour patterns in Neotropical primates. Proceedings of the Royal Society B: Biological Sciences 279:2204-2211.

Stanford C. 1998. Chimpanzee and red colobus: the ecology of predator and prey. Cambridge: Harvard University Press.

Struhsaker TT, Butynski TM, Lwanga JS. 1988. Hybridization between redtail (Cercopithecus ascanius schmidti) and blue (C. mitis stuhlmanni) monkeys in the Kibale Forest, Uganda. In: Gautier-Hion A, Bourliere F, Gautier J, Kingdon J, editors. A primate radiation: Evolutionary biology of the African guenons. Cambridge: Cambridge University Press. p 477-497.

Stevens M, Cuthill IC, Windsor AMM, Walker HJ. 2006. Disruptive contrast in animal camouflage. Proceedings of Royal Society B: Biological Sciences 273:2433-2438.

Stevens M, Stoddard MC, Higham JP. 2009. Studying primate color: towards visual system-dependent methods. International Journal of Primatology 30:893-917.

Stevens M, Párraga C, Cuthill I, Partridge J, Troscianko T. 2007. Using digital photography to study animal coloration. Biological Journal of the Linnean Society 90:211-237.

Tosi AJ, Detwiler KM, Disotell TR. 2005. X-chromosomal window in the evolutionary history of the guenons (Primates: Cercopithecini). Molecular Phylogenetics and Evolution 36:58-66.

Wirtz P. 1999. Mother species – father species: unidirectional hybridization in animals with female choice. Animal Behavior 58:1-12.

Wolfheim JH. 1983. Primates of the world: Distribution, abundance, and conservation. Seattle: University of Chicago Press. N p.

Woodruff DS. 1973. Natural hybridization and hybrid zones. Systematic Zoology 22:213-217.

Zuk M, Decruyenaere J. 1994. Measuring individual variation in colour: a comparison of two techniques. Biological Journal of the Linnean Society 53:165-173.