sign in sign up
<<
Home , Bacillus licheniformis, Caihong, Cathayornis, Cephalopod ink, Chamaeleo, Chiroxiphia

THE EVOLUTIONARY HISTORY AND PRESERVATION OF AND

MELANOSOMES

A Dissertation

Presented to

The Graduate Faculty of The University of Akron

In Partial Fulfillment

of the Requirements for the Degree

Doctor of Philosophy

Jennifer A. Peteya

August, 2018

THE EVOLUTIONARY HISTORY AND PRESERVATION OF MELANINS AND

MELANOSOMES

Jennifer A. Peteya

Dissertation

Approved: Accepted:

______Advisor Program Director, Integrated Bioscience Robert Joel Duff Hazel A. Barton

______Co-Advisor/Committee Member Interim Dean of the College Matthew D. Shawkey Linda Subich

______Committee Member Dean of the Graduate School Liliana D’Alba Chand Midha

______Committee Member Date Julia A. Clarke

______Committee Member John M. Senko

ii ABSTRACT

Melanins are a of ubiquitous that provide not only coloration to the organisms in which they produced, but also serve as a mechanism for thermoregulation, protection from ultraviolet radiation and cytotoxicity induced by free radicals, as well as numerous other beneficial roles. In most , melanins are housed in called melanosomes. morphology correlates with -based coloration in and , but melanosome diversity is limited in basal .

Approximately 10 ago, the study of melanosomes, melanin, and coloration was extended to the record by the reinterpretation of microbodies preserved in fossil initially interpreted as as remains of melanosomes. However, despite countless lines of evidence for this hypothesis that have been described since, some authors argue that the bacterial hypothesis is equally parsimonious.

This dissertation will address some of the concerns of purveyors of the bacterial hypothesis and discuss the evolutionary history of melanosomes and melanins. We first tested the hypothesis that striped fossil feathers are the result of bacterial preservation through taphonomic experimentation. We found that modern keratinolytic bacteria preferentially colonize unpigmented stripes in modern feathers over melanized stripes, so it is unlikely that this hypothesis was supported in ancient ecosystems. We also found that bacteria and biofilms preserved in association with fossil integument are dissimilar to microbodies interpreted as fossil melanosomes.

iii We began the second part of this dissertation by describing an enantiornithine that exhibits a mix of juvenile skeletal characteristics and sexual ornaments.

Melanosomes preserved in its feathers are similar in morphology and arrangement to those in modern iridescent feathers, which are predominantly used to attract mates. We also described melanosomes associated with the fossilized skin of amphibians and lampreys, which do not exhibit diverse melanosome morphologies. Convergent shifts in melanosome diversity in mammals and pennaraptoran may have occurred due to changes in physiology between these two groups and their ancestors. Finally, we reviewed the functions of melanins in modern organisms and evidence for these functions in the fossil record. Based on their modern ubiquity, we hypothesized that melanin may have evolved early in the history of .

iv DEDICATION

This work is dedicated to the former instructors that encouraged and inspired me to continue pursuing : my mom, who taught me how to look for when

I was about three years old; my dad, who decided to spend vacations looking for fossils and going to museums in Nova Scotia; my second and third grade teachers, Judy

Chester and Dorothy Ramsey, who embraced my love of paleontology and encouraged me to continue; Connie Hubbard, my high school Science Experimental Research

Program (SERP) instructor, who let me do a project on dinosaurs rather than a typical experimental project; Dr. Lee Gray, my soft rock geology professor who mentored me throughout college and encouraged me to apply to do research in ; Joel Collins, my undergraduate drawing instructor who never got tired of seeing dinosaurs or trilobites in my work; my graduate professor, Dr. Bill Ausich, who encouraged me to press on throughout my master’s studies; and finally my master’s advisor, Dr. Loren

Babcock.

v ACKNOWLEDGEMENTS

First, I would like to thank my former advisor, Dr. Matthew Shawkey, for his patience, encouragement, help, and everything else he has given me for the past five years. I would also like to thank Dr. R. Joel Duff for taking over the role as my advisor since Dr. Shawkey moved to Belgium and my other committee members, Dr. Liliana

D’Alba, Dr. Julia Clarke, and Dr. John Senko. Dr. D’Alba not only helped Nick Justyn and me with the statistical analysis in Chapter 2, but has also served as a source of encouragement and inspiration throughout my dissertation. Similarly, Dr. Clarke also served as a source of inspiration and was especially instrumental in the production of

Chapter 4.

Additionally, I would like to thank my current and former labmates, Nick Justyn,

Dr. Brani Igic, Dr. Bill Hsiung, Dr. Ming Xiao, Asritha Nallapaneni, Dr. Chad Eliason,

Dr. Rafael Maia, Weiyao Li, Xiaozhou Yang, Mario Echeverri, Dr. Daphne Fechyr-

Lippens, and Jiuzhou Zhao. I served as Nick’s mentor for his undergraduate honors thesis, throughout which we co-wrote Chapter 2, and has helped with Chapter 6. He has been a great friend for the past four years and has helped me through some of the toughest times throughout the completion of this project. Chad and Rafael taught me the basics of R programming, Brani and Chad both helped me with statistical analyses, Ming helped me learn how to use Origin Pro and Igor Pro, and each one of my labmates has helped with editing my manuscripts, bouncing ideas, and moral support.

vi I would also like to thank several other people instrumental in the completion of my dissertation. Dr. Quanguo Li from the University of Geosciences and Dr. Ke-

Qin Gao from Peking University provided specimens and guidance for Chapters 4 and 5.

Dr. Zhorro Nikolov, Dr. Bojie Wang, Thomas Quick, and Dr. Richard Londraville provided access to and training on the Raman spectrometer, field-emission SEM, environmental SEM, and spectrophotometer, respectively. Dr. Kevin Abbasi from Case

Western Reserve University provided the ToF-SIMS analysis for Chapter 3. Scott

Thomas, Matt Bos, and Dr. Peter Niewiarowski provided modern amphibian and lizard samples for Chapter 5. The late Dr. Edward Burtt provided licheniformis for the

Chapter 2 experiment and for comparative Raman spectroscopy. Prasad Raut provided carbon samples and Dean Pearson from the Pioneer Trails Regional Museum provided the Platanus specimen. Drs. Zhiheng Li, Xia Wang, Sally Thomas, and an anonymous reviewer provided criticism and feedback for Chapter 4, which is published in Palaeontology. Anonymous reviewers also improved Chapter 2, which is published in

The Auk. Drs. Loren Babcock, Baoyu Jiang, and Fangchen Zhou provided helpful discussion and criticism for Chapter 3.

Lastly, I would like to thank my family members for their continued support throughout my pursuit of science, including my parents, Karen and Dennis Peteya, and my husband, Dan Lopez. I would especially like to thank my sister, Dr. Stephanie Peteya, who after completing her undergraduate degree at the University of Akron suggested I contact Dr. Shawkey about working on fossil for my Ph. D.

vii TABLE OF CONTENTS

Page

LIST OF TABLES ...... xi

LIST OF FIGURES ...... xii

CHAPTER

I. INTRODUCTION ...... 1

II. PREFERENTIAL ATTACHMENT AND COLONIZATION OF THE KERATINOLYTIC BACTERIUM BACILLUS LICHENIFORMIS ON BLACK AND STRIPED FEATHERS...... 6

Introduction ...... 6

Material and Methods ...... 8

Results ...... 14

Discussion ...... 15

III. MICROBIAL BIOFILMS PRESERVED WITH FOSSILIZED INTEGUMENT ARE DISTINCT FROM FOSSIL MELANOSOMES ...... 20

Introduction ...... 20

Material and Methods ...... 22

Results ...... 26

Discussion ...... 33

IV. THE AND COLORATION OF AN ENANTIORNITHINE BIRD FROM THE EARLY OF CHINA ...... 38

viii Introduction ...... 38

Material and Methods ...... 39

Systematic Paleontology ...... 43

Description ...... 47

Discussion ...... 63

V. DIVERSIFICATION OF MELANOSOME SHAPE IS LINKED TO PHYSIOLOGICAL CHANGES BETWEEN VERTEBRATE TAXA ...... 68

Introduction ...... 68

Material and Methods ...... 71

Results ...... 75

Discussion ...... 80

VI. WHAT WAS THE ORIGINAL FUNCTION OF MELANIN? A REVIEW OF THE MODERN FUNCTIONS OF MELANINS AND EVIDENCE FROM THE FOSSIL RECORD ...... 87

The Types of Melanin ...... 90

Melanins are Thermoregulators ...... 92

Melanins Contribute to Coloration ...... 95

Melanins are Protectors ...... 101

When Did the Functions of Melanins Evolve? ...... 110

Conclusions ...... 119

VII. CONCLUSION ...... 121

REFERENCES ...... 123

APPENDICES ...... 147

APPENDIX A. CHAPTER 2 SUPPLEMENT ...... 148

APPENDIX B. CHAPTER 3 SUPPLEMENT ...... 151

ix APPENDIX C. CHAPTER 4 SUPPLEMENT ...... 163

APPENDIX D. CHAPTER 5 SUPPLEMENT ...... 170

x LIST OF TABLES

Table Page

5.1. Measurements of melanosome morphology in extant and fossil (indicated) skin samples ...... 76

6.1. Summary of the functions of melanins ...... 111

xi LIST OF FIGURES

Figure Page

2.1. Colonization of Bacillus licheniformis on a white- and black-striped Eurasian Three- toed ...... 12

2.2. Histograms comparing bacterial count and cell density between feather regions and location on the feather ...... 15

3.1. Biofilm structures preserved in Haikouella ...... 27

3.2. SEM images of microbodies and biofilm preserved on Haikouella matrix sample 13 ...... 28

3.3. Biofilm structures preserved in LSJC from the Jiulongshan Formation, Daohugou Locality, , China ...... 29

3.4. Biofilm preserved on c.f. from China ...... 30

3.5. Chemical characterization of the exceptionally preserved Haikouella integument ....32

4.1. Ventral view of the new bohaiornithid specimen, (CUGB P1202, primary slab) ...... 45

4.2. The new bohaiornithid specimen, (CUGB P1202, counter slab ...... 46

4.3. The of CUGB P1202 ...... 50

4.4. The pectoral girdle in CUGB P1202 ...... 51

4.5. The of CUGB P1202 ...... 53

4.6. The pelvic girdle and hind limbs of CUGB P1202 ...... 56

4.7. Crown and neck feathers as well as preserved melanosome morphologies in the new bohaiornithid ...... 58

4.8. Details of the remiges and rectrices in the counterpart slab of CUGB P1202 ...... 59

4.9. Right and body contour feathers on the primary slab of CUGB P1202 ...... 60

xii 4.10. Statistical analysis and Raman spectra for fossil and extant samples ...... 62

5.1. SEM images of melanosomes from the skin of fossil and extant agnathans and amphibians ...... 77

5.2. Raman spectra generated from exceptionally preserved skin from fossil amphibian (top) and lamprey (bottom) samples with spectra from modern Necturus and Petromyzon melanin () for comparison ...... 78

5.3. SEM images of small, pill-shaped melanosomes from black-colored skin of A, a wing and B, the ring of a -breasted grosbeak ...... 80

5.4. Length of the long axis and aspect ratio of melanosomes in skin, feathers, and of extant vertebrates ...... 82

5.5. Diversity of melanosome morphology from modern and fossil integument across Vertebrata ...... 83

xiii CHAPTER 1

INTRODUCTION

Pigments can provide numerous selective advantages for the organisms that produce them. Not only do they confer color to the organism, which itself may effectively it from potential predators or serve as a mechanism of communication between individuals, but some pigments also scavenge free radicals, assist with thermoprotection or thermoregulation, shield an organism from ultraviolet radiation, or serve a suite of other functions that help it eat, survive, and reproduce. The most ubiquitous of these pigments are melanins.

Melanins are a class of pigments that includes a series of complex polymers that generally produce dark . They occur in four main chemical variants, although due to their complexity and general insolubility the specific chemical structure of all melanins is unknown. Eumelanin and pheomelanin are both derived from the breakdown of the tyrosine by . Eumelanin is composed of 5,6-dihydroxyindole-2- carboxylic acid (DHICA) and 5,6-dihydroxindole (DHI) moieties and produces black or colors. Pheomelanin consists of a backbone of benzothiazine and generates red- brown or buff colors (Wakamatsu and Ito 2002). Neuromelanin is produced from in the brain of and some other (Zecca et al. 1996; Tribl et al.

1 2009). Although its production pathway and chemical structure are unknown, neuromelanin granules are thought to consist of a core of pheomelanin surrounded by eumelanin, metals, , and (Zucca et al. 2017). Allomelanins, or “other melanins,” include a suite of complex biopolymers that predominantly generated from polyphenols. The most common allomelanin is pyomelanin, which is produced in bacteria and fungi via the polymerization of homogentisic acid (Turick et al. 2010).

Eumelanin and pheomelanin are the dominant melanins produced by animals. In most animals, they are housed in organelles called melanosomes that are produced by cells called melanophores or . Melanosomes in bird feathers and exhibit high morphological variability that is correlated with the color they produce

(Li et al. 2010, 2012, 2014). Red-brown feathers and hairs contain small, nearly spherical melanosomes composed predominantly of pheomelanin while eumelanin dominates the elongate, pill-shaped melanosomes that comprise black feathers and hairs (Li et al. 2010).

Birds exhibit further morphological disparity because they produce a wider range of vibrant colors than mammals, including iridescent and non-iridescent structural colors

(e.g. matte or ), via the nanoscale arrangement of melanosomes and other materials of differing refractive indices, including and air. Melanosomes in iridescent feathers occur in a variety of shapes, including high aspect ratio melanosomes arranged end-to-end in layers, flattened, plate-like, and hollow melanosomes that further increase the contrast of the feather (Prum 2006; Eliason et al. 2013).

Recently studies of coloration have been pushed back to extinct organisms. Early studies of pill-shaped microbodies discovered in exceptionally preserved integument assumed that they were bacteria associated with decay or biofilms

2 that surrounded the carcass and shielded it from decay (e.g. Davis and Briggs, 1995).

However, due to morphological similarities between these fossil microbodies and modern melanosomes and their limitation to the integument itself, Vinther et al. (2008) reinterpreted them as preserved melanosomes. This reinterpretation spawned a series of inquiries into fossil color and the of melanins and melanosomes. In 2010,

Zhang et al. and Li et al. released nearly simultaneous publications illustrating that fossil melanosomes could be used to reconstruct the original melanin-based colors of extinct dinosaurs based on morphological similarities between the preserved melanosomes in their feathers and the color-correlated melanosomes in the feathers of extant birds. The new field of fossil color further expanded when Li et al. (2012) described hyper- elongated melanosomes with an end-to-end arrangement preserved in , suggesting that this had at least weakly-iridescent feathers in life. The melanosome affinity of these microbodies was corroborated by the detection of melanin itself in exceptionally preserved integument, including both eumelanin (Wogelius et al.

2011; Glass et al. 2013; Lindgren et al. 2014, 2015a, 2015b) and more recently pheomelanin (Colleary et al. 2015; Brown et al. 2017).

Despite abundant evidence suggesting that these described microbodies are indeed melanosomes, some authors argue that the bacteria hypothesis has not efficiently been ruled out (Moyer et al. 2014: Lindgren et al. 2015a). I therefore begin this dissertation by addressing some of the concerns presented by these authors. The first chapter focuses on the claim that feather-degrading (keratinolytic) bacteria may have preferred to feed on melanized regions of feathers – a hypothesis which Moyer et al.

(2014) uses to explain the persistence of fossil microbodies in specific regions of feathers

3 (i.e. dark stripes) and not others (i.e. stripes of an originally striped feather). This chapter focuses on an experiment completed in collaboration with Nicholas Justyn in which we inoculated black- and white-striped feathers from the modern three-toed woodpecker Picoides tridactylus to test if bacteria preferentially colonized the black or white stripes of the feather. In the second chapter, I present three examples of putative fossil biofilms associated with exceptionally preserved integument from a early chordate, a Middle salamander, and an bird. The aim of this chapter is to show that while mineralized biofilms and bacteria do persist in the fossil record and can be associated with preserved integument, they can also be distinguished from fossil melanosomes based on their location, the presence of mitotic fissures, and associated biofilm structures.

The next three chapters involve the study of fossil color and melanosomes. In

Chapter 4, I describe a new specimen of bohaiornithid bird from China in collaboration with Drs. Julia Clarke, Li Quanguo, and Gao Ke-Qin. This new specimen retains characteristics of a juvenile , but also shows evidence of adult plumage, such as ornamental tail feathers and iridescent plumage suggested by hyper-elongated, end-to-end oriented melanosomes. It therefore adds to growing evidence that enantiornithines were sexually mature prior to adulthood. We then examine the hypothesis presented by Li et al.

(2014) that high melanosome diversity in hairs and feathers may have evolved due to convergent changes in the melanocortin system, which controls numerous aspects of vertebrate physiology, between basal amniotes and mammals and pennaraptoran dinosaurs in Chapter 5. We test this hypothesis and two others by extending the study into extant anamniotes and Middle Jurassic and Early Cretaceous salamanders and

4 lampreys from China. Lastly, in Chapter 6 I review the functions of melanins in modern organisms and the reported evidence for their modern functions in the fossil record in an attempt to elucidate melanin’s original function and when it likely evolved. My dissertation research contributes to our understanding of the taphonomy (preservational history) of exceptionally preserved organisms, the development and behaviors of an extinct group of birds, and the of the most ubiquitous in the modern world.

5 CHAPTER 2

PREFERENTIAL ATTACHMENT AND COLONIZATION OF THE

KERATINOLYTIC BACTERIUM BACILLUS LICHENIFORMIS ON BLACK AND

WHITE STRIPED FEATHERS

Introduction

Feathers serve numerous essential functions for birds, from flight, signaling, and camouflage to thermoregulation (see for review McGraw 2006). Abrasion or degradation of feathers can reduce their effectiveness at carrying out these integral functions. In particular, flight feathers weakened by physical abrasion can limit a bird’s ability to escape predators or catch prey (Bonser 1995), and degradation of the keratin cortex by keratinolytic bacteria can potentially alter the observed color of a feather and, in turn, the bird’s reproductive success (Shawkey et al. 2007; Gunderson et al. 2009). For these reasons, the maintenance of feathers to preserve their structural and visual integrity is an essential behavior that can directly affect fitness. Molting, preening (with uropygial oil), dusting, and sunning are all active methods that birds use to combat keratinolytic bacteria that may colonize and degrade their feathers (Hillgarth and Wingfield 1997; Shawkey et al. 2003; Ruiz-Rodriguez et al. 2009; Kent and Burtt 2016).

6 Some pigments, such as melanin and psittacofulvins, deposited in feathers have previously been shown to aid in resistance to degradation by keratinolytic bacteria

(Goldstein et al. 2004; Gunderson et al. 2008; Burtt et al. 2010; Ruiz-de-Castañeda et al.

2012), but the mechanism for this resistance is currently unknown. Keratinolytic bacteria, such as Bacillus licheniformis, degrade the β-keratin matrix of a feather by secreting the hydrolytic enzyme keratinase (Santos et al. 1996; Ruiz-Rodriguez et al. 2009). Black feathers containing a high concentration of eumelanin (a ubiquitous chemical variant of melanin that produces black and brown colors) are more resistant in vitro to bacterial degradation than white feathers by B. licheniformis (Goldstein et al. 2004). Similarly, melanized areas within individual feathers are more resistant to bacterial degradation than unmelanized areas (Ruiz-de-Castañeda et al. 2012). Melanin may enhance resistance by thickening the feather cortex or by making the keratin matrix more difficult to break down (Goldstein et al. 2004; Gunderson et al. 2008). Additionally, melanin and melanosomes are known to inhibit bacterial and fungal growth in mammal integument

(MacKintosh 2001) and may directly affect keratinolytic bacteria in birds (Gunderson et al. 2008).

Whether melanized feathers are more difficult to degrade because melanin enhances their toughness or their antimicrobial activity is still unclear because all experimental studies thus far have only examined feathers after degradation is complete.

Attachment is the first step in degradation process, and thus some natural materials have evolved defenses such as nanospheres on the surface of eggshells that prevent bacterial adhesion (D’Alba et al. 2014, 2016). Melanin’s known antimicrobial properties

(MacKintosh 2001) suggest that it may serve a similar function in feathers, but this has

7 not previously been tested. Here we test this hypothesis by experimentally inoculating

Bacillus licheniformis onto striped feathers in vitro, predicting that the bacteria will preferentially attach to and colonize unmelanized white stripes over melanized black stripes. We also discuss potential mechanisms that may affect bacterial attachment to feathers and assess the implications for pre-burial taphonomic processes that may have occurred prior to the exceptional preservation of fossil feathers.

Material and Methods

Feather Preparation

Forty-four black and white striped feathers were plucked from the back of study skins of male and female Eurasian Three-toed (Picoides tridactylus) from the University of Akron collections. The feathers were cut to approximately equal sizes

(0.5 cm) so that 22 of the samples had a white region at the apical end and black at the base and 22 had a black region at the apical end and white at the base. This was done to avoid any differences in keratin thickness, such as the wider circumference of keratin at the base as opposed to the tip of the feather, and to make the size of white and black regions similar in size. We took pictures of each feather using a binocular microscope

(Leica, model S8AP0) as references for the location of melanized regions on each feather.

While previous authors have autoclaved feathers used in bacterial degradation experiments (Grande et al. 2004), autoclaving may degrade feather keratin and influence the activity of feather-degrading bacteria (Gunderson et al. 2008). We instead removed

8 excess dust and other particles by washing the feathers in deionized water and then cleaned them with 100% ethanol. Our subsequent ESEM analyses of control (sham inoculated; see below) feathers did not contain any bacteria, confirming that this method of sterilization was successful in removing any potential microbes from the surface of the feather as well as between the barbs and barbules.

Bacteria Preparation

We used the keratinolytic bacterium Bacillus licheniformis as a model organism for the study of preferential colonization of striped feathers. B. licheniformis is the most common feather-degrading bacterium, at times representing over 80% to 90% of a feather’s bacterial load (Burtt and Ichida 1999). We grew B. licheniformis strain 138B

(Burtt and Ichida 1999), on sterile tryptic soy agar (TSA) source plates for 48 hours.

Twenty sealable test tubes containing 2 mL each of tryptic soy broth (TSB) were inoculated with 20 µl of bacteria from the source plates and incubated overnight in a shaking incubator at 37 °C and 180 rpm. 2 mL of fresh TSB was mixed into each of the

20 inoculated tubes using a vortex mixer to create a 1:1 B. licheniformis to TSB solution, and were placed back in the shaking incubator for four hours using the same settings.

This four-hour period for bacterial growth falls within the projected optimal growth period for B. licheniformis (Frankena et al. 1985) and the bacteria should feed more readily on feathers during this period. The four-hour tubes were emptied into a sterile graduated cylinder and combined with equal parts 0.9% saline solution to create a homogenous 1:1 mixture. The saline was added to increase the likelihood that the bacteria would attach to and degrade the feathers rather than remain in solution in the

9 nutrient-rich TSB. The absorbance of this mixture was tested using a 96 well spectrophotometer at a 530 nm of light with a SpectraMax Plus 384

Microplate Reader. We obtained an absorbance value of 0.805, which correlated to approximately 6400 colonies of B. licheniformis based on our previously established standard (Table A.1).

Experiment

Forty of the sampled feathers (20 with white at the apical end and 20 with black at the apical end) were sterilized and placed into separate sterile glass test tubes and inoculated with 2 mL of the 1:1 B. licheniformis and saline solution. The samples were placed in the shaking incubator for 18 hours at 37 °C and 180 rpm. Previous studies reported that incubation times longer than 18 hours led to the total degradation of feathers

(e.g. Goldstein et al. 2004; Grande et al. 2004; Ruiz-de-Castañeda et al. 2012), while times shorter than 18 hours led to minimal attachment (Ramnani et al. 2005). We then washed the feathers using a 5-stage ethanol rinse to remove any biofilm and unattached bacteria from the feathers, leaving only bacteria that were securely adhered (Dunne

2002). The feathers were washed at 25%, 50%, 75%, and 100% ethanol for five minutes each, then 100% ethanol for 10 minutes. One sample with a white base was inadvertently contaminated during the serial washing process and was therefore excluded from analysis. Samples were dried in an incubator at 37 °C, then mounted with carbon tape on aluminum stubs and sputter-coated for 3 minutes with -palladium using a Polaron

E5000 Sputter-Coater (Quorum Technologies) for Environmental Scanning Electron

Microscopy.

10 We prepared two additional sets of uninoculated control feathers, each consisting of two feathers with opposite color , that were sterilized using the same methods as the experimental feathers. “Wet” control feathers were submerged in 2 mL of a 1:1

TSB and then 0.9% sterile saline solution mix. The samples were then incubated for 18 hours at 37 °C and 180 rpm and washed using the same five-stage ethanol wash procedure as experimental samples. The “dry” control samples were cut and washed in

100% ethanol, then directly placed on aluminum stubs and sputter-coated.

Electron Microscopy

We used an FEI Quanta 200 Environmental Scanning

(ESEM) at The University of Akron. Each image was taken under high vacuum at a standard working distance of 10 mm, magnification of 12,000 x, spot size of 4.5, and voltage of 30 kV. The feather samples were chosen at random and imaged blindly so that the melanized and unmelanized regions could not be distinguished from one another during sampling. Four images were taken of the rachis of each sample. One image each was taken from the base, lower central section, upper central section, and the apical end of the rachis to ensure of both melanized and unmelanized sections on each feather. Three images were taken at the basal, middle, and apical sections of two barbs, one each on the right and left halves of the feather (Figure 2.1). We chose barbs that ran the full length of the feather with the assumption that these barbs would contain both melanized and unmelanized sections. Sampling locations within each region were haphazardly selected. Additionally, we took ESEM images at lower magnification at each

11 sampling site to later cross-reference with images for color identification (see below).

Figure 2.1. Colonization of Bacillus licheniformis on a white- and black-striped Eurasian Three-toed

Woodpecker feather. (A) Light microscope image detailing the environmental scanning electron microscope (ESEM) imaging locations on 1 of the 39 feathers analyzed. Red circles indicate sampling locations in unmelanized regions, and blue circles indicate sampling locations in melanized regions. (B–D)

ESEM images corresponding to similarly labeled sampling locations on the feather; B and C show B. licheniformis cells (arrow), whereas D shows no bacterial cells. Scale bars=5 µm.

12 Analysis

We counted the number of bacteria in each high magnification ESEM image and then determined the color of each sampling location by comparing the low magnification

ESEM images to optical images of each feather. Individual barbules were counted for each ESEM and microscope image to determine the precise location and color of each sampled area. We also selected the total surface area of the feather available for attachment in each image using Image J software (National Institutes of Health; available for download at http://rsb.info.nih.gov/nih-image/index.html) and calculated density as the number of bacteria per micron2. We measured bacterial density to account for the increased surface area in the lower and middle sections of the feathers relative to the surface area of single barbs near the apical end of the feather.

To test whether bacteria preferentially attached to certain colors or locations on the feathers, we used a zero-inflated Poisson model in R for bacterial cell count (R

Development Core Team 2007). The Poisson model was zero-inflated to account for the large number of sampling sites that did not have attached bacterial cells. Sampling location was separated into three general regions: 1) top, which included the apical one- third of each feather; 2) middle, which included the central one-third of the feather; and

3) bottom, which included the basal one-third of the feather. We also recorded whether bacteria were present in the rachis or the barbs of each feather. Our models included bacterial density as response variable and sampling location, region color and feather structure (rachis or barb) as explaining factors.

Results

13

Color and location significantly affected the number of bacterial cells attached to the feather. Of the 390 total feather regions imaged, 206 were white and 184 were black.

Overall, approximately twice as many white regions (67) had attached bacterial cells than black regions (35) and there were approximately twice as many individual B. licheniformis attached to the white stripes (1569 cells) as to black stripes (879 cells;

Figure 2.2). B. licheniformis preferentially colonized white over black stripes (z = 3.2, p

= 0.001). Bacterial count was also significantly lower toward the base (n=156 images) than at the middle (n=78 images, z = -2.9, p = 0.004) and top of the feather (n = 156 images, z = -2.7, p = 0.006). Density calculations of each section indicated that the base of the feather supported fewer bacteria per µm2 (average density: Top: 0.0172 +/- 0.0041,

Middle: 0.0168 +/- 0.0047, Base: 0.0112 +/- 0.0044). Additionally, there was no difference between the numbers of bacteria that attached to the barbs of each feather versus the rachis (z = 1.1; p = 0.27).

14

Figure 2.2. Histograms comparing bacterial cell count and cell density between black and white feather regions and location on the feather.

Discussion

Previous studies have shown that keratinolytic bacteria degrade unmelanized white feathers more quickly than melanized black feathers (Goldstein et al. 2004;

Gunderson et al. 2008; Ruiz-de-Castañeda et al. 2014). However, the mechanisms by which this occurs and the roles that melanin plays are unclear. Here, we show that the common keratinolytic bacterium B. licheniformis preferentially attaches to and colonizes

15 unmelanized white stripes of a single feather nearly twice as often as melanized black stripes in vitro. Indeed, B. licheniformis attached to nearly twice as many unmelanized regions, with twice the number of individual cells as unmelanized regions. It is possible that the bacteria initially colonized both feather regions equally, then reproduced more quickly in the white regions due to the lack of melanin and therefore greater availability of keratin. However, given our short incubation time and scant evidence of keratin degradation, it is more likely that the bacteria initially preferentially colonized the unmelanized regions. We also used density of bacteria to ensure that this effect was not due to changing surface area over the feather (i.e. thinner feather tips as opposed to a wider feather base). Our results show that B. licheniformis is more likely to colonize the tips of a feather rather than the base. However, in vivo colonization may differ from our experimental results, as feathers often contain a greater diversity of bacteria than one single (Burtt and Ichida 2004; Shawkey et al. 2005; Kent and Burtt 2016).

Additionally, our feathers were experimentally inoculated in solution rather than through contact with soil, as would occur in vivo (Burtt and Ichida 1999). We therefore cannot rule out the possibility that the range of other bacteria found on feathers of birds in their natural environment and differing inoculation methods may affect the colonization and degradation of the feathers (Czirják et al. 2013).

Preferential colonization on unmelanized areas may partially explain the greater degree of bacterial degradation of unmelanized versus melanized feathers (Santos et al.

1996; Ruiz-Rodriguez et al. 2009). If bacteria attach more quickly and easily to unmelanized feathers, then they should degrade them more quickly. The presence of melanin may thus reduce the number of keratinolytic bacteria that adhere to feathers,

16 whether attached to live birds or molted, through several possible mechanisms. Melanin may indirectly affect the ability of a bacterium to attach to a feather, such as by increasing the roughness of the keratin surface. Nanoscale differences in roughness can have a significant negative impact on bacterial adherence and their ability to form biofilms (Mitik-Dineva et al. 2009; Singh et al. 2011). However, we saw no obvious difference in roughness between the different feather locations or between black and white sections of the feather, although these regions exhibit some texture. The presence of melanosomes may also reduce colonization by making feather keratin thicker and harder, or by influencing the surface chemistry. This is accomplished by increasing the number of disulfide bonds present, which also makes these bonds more difficult to break down by bacterial keratinases during degradation (Goldstein et al. 2004; Ramnani et al.

2005).

Melanin may also directly interact with bacteria. For example, because melanosomes are lysosomal organelles (Raposo et al. 2002), melanized feather regions may contain lysosomal enzymes that could interfere with bacterial processes and make feathers less suitable for bacterial proliferation (Burkhart and Burkhart 2005). Melanin may also inhibit degradation by actively binding keratinases secreted by bacteria and limiting their ability to break down feather keratin (Gunderson et al. 2008). Indeed, there is evidence that melanin decreases the effect of bacterial proteases on fungi, including those of Bacillus subtilis (Kuo and Alexander 1967). Melanin may thus help slow degradation through prevention of both bacterial colonization and growth. Additionally, melanin may also directly affect a bacterial cell’s ability to attach to and degrade a feather via electrostatic repulsion. Melanin is highly negatively charged and melanization

17 of fungal cells has been shown to increase their negative charge, making them more resistant to phagocytosis by, for example, (Nosanchuk and Casadevall

1997). Melanization of feathers may similarly increase the negative charge of the keratin matrix, particularly if the melanosomes are located close to the surface of the feather (i.e. just below the outer keratin cortex), as in glossy or iridescent feathers (Maia et al. 2010).

It is, however, unlikely that these direct methods could affect the attachment or colonization of bacteria on the surface of the feather prior to degradation because the bacteria are not in direct contact with the melanin or melanosomes. These methods may, however, affect growth of bacteria after attachment. Whether direct or indirect, all of these potential mechanisms could limit bacterial colonization and degradation of fathers regardless of the feather’s environment (i.e. attached to a live bird or molted). Further experiments should test these mechanistic hypotheses.

Our experiment has additional implications for the pre-burial taphonomy of fossilized feathers. Although the preservation of feathers in terrestrial konservat lagerstätten – deposits with soft tissue preservation – is not uncommon (Davis and Briggs

1999; Schweitzer 2011), our understanding of feather taphonomy is limited. Previous studies have shown that melanin (Colleary et al. 2015; Lindgren et al. 2015) and melanosomes (Chapter 4; Vinther et al. 2008; Clarke et al. 2010; Li et al. 2010, 2012,

2014; Carney et al. 2012; Field et al. 2013; Huang et al. 2016) are preserved in fossil feathers, but few have examined the potential effects of bacteria on pre-burial degradation of feathers (see Moyer et al. 2014). Indeed, feather-degrading bacteria were likely involved in the degradation of theropod dinosaur integument. Our results and those of others (Goldstein et al. 2004; Gunderson et al. 2008; Ruiz-de Castañeda et al. 2012)

18 suggest that eumelanin inhibits the colonization and degradation of feather keratin by bacteria. White feather stripes or whole white feathers therefore would have been more susceptible to bacterial colonization and degradation than melanized feathers prior to burial.

We have shown that keratinolytic bacteria preferentially colonize unmelanized white regions of striped feathers over melanized regions. The processes that occur prior to the degradation of a feather by keratinolytic bacteria are largely uncharacterized.

Examining these processes is essential to understanding feather degradation by keratinolytic bacteria, the mechanism by which melanin inhibits bacterial degradation, avian ecology, and the pre-burial processes which may have affected the preservation of feathers in the fossil record.

19 CHAPTER 3

MICROBIAL BIOFILMS PRESERVED WITH FOSSILIZED CHORDATE

INTEGUMENT ARE DISTINCT FROM FOSSIL MELANOSOMES

Introduction

Deposits of exceptional preservation of non-biomineralized tissues, called

Konservat-Lagerstätten, are extraordinarily rare in the fossil record. In these rare examples, burial in fine sediments and subsequent authigenic mineralization of integument and viscera must have occurred rapidly, as taphonomic experiments have shown that these tissues can degrade within hours to weeks after depending on tissue composition and decay environment (Briggs and Kear 1993; Briggs et al. 1993).

Rapid preservation of tissues and soft-bodied organisms in known Konservat-

Lagerstätten occurred via pyritization (e.g. Chengjiang Biota; Gabbott et al. 2004; Zhu et al. 2005), phosphatization (e.g. the Orsten fauna; Maas et al. 2006) or replication by clay minerals and carbonization of original organic material (e.g. most Burgess Shale- localities; Orr et al. 1998; Butterfield 2003; Gaines et al. 2008). Taphonomic interpretations of fossils from these lagerstätten are complicated because they often present multiple preservational modes, commonly between different tissue types in one

20 organism (e.g. cuticle, muscle, gut structures; Butterfield 2002) or between different taxa

(Zhu et al. 2005; Parry et al. 2018). In many cases, the preservation of soft tissues was mediated by microbial biofilms. Microbial halos surrounding a decaying carcass have been shown experimentally to change the chemistry of the surrounding sediment

(Borkow and Babcock 2003), induce of soft tissues (Briggs and Kear

1993), and reduce decay rate (Briggs 2003).

It was therefore logical for researchers to assume that microbodies found in the integument of exceptionally preserved extinct animals were fossilized bacteria (e.g. Davis and Briggs 1995). However, due to their restriction to the fossilized integument and morphological similarities to melanin-containing organelles, called melanosomes, in modern integument, Vinther et al. (2008) reinterpreted these microbodies to be fossil melanosomes rather than bacterial cells. Since this reinterpretation, numerous other lines of evidence have been discovered for the melanosome affinity of these preserved microbodies (reviewed in Vinther 2015). For example, microbodies preserved in the feathers of extinct pennaraptoran dinosaurs exhibit similar diversity to the color- correlated melanosomes in modern bird feathers, which has led to the interpretation of melanin-based colors in these extinct animals (e.g. Li et al. 2010, 2012; Zhang et al.

2010). In contrast, melanosome morphologies in modern non-mammalian, non- pennaraptoran vertebrate skin are far less variable and extinct examples of these taxa exhibit a similar lack of morphological disparity (Chapter 5; Li et al. 2014). The diversity of microbody size and shape is therefore linked to vertebrate . This association would not be expected if these microbodies were indeed preserved bacteria, since modern and fossil bacteria occur in an even greater range of morphological diversity. Spiral and

21 coccoid-type bacteria, for example, are preserved in association with muscle tissue in the horseshoe crab Mesolimulus from the Upper Jurassic Nusplingen Plattenkalk in Germany

(Briggs et al. 2005). Melanins themselves have also been chemically characterized in association with preserved microbodies in fossil integument using several different chemical techniques, including both black/brown eumelanin (e.g. Lindgren et al. 2014,

2015a, 2015b) and red-brown pheomelanin (Colleary et al. 2015; Brown et al. 2017), which also suggests a melanosome affinity.

Despite abundant evidence that microbodies that are preserved in association with fossil integument are indeed melanosomes, some authors argue that the original microbial hypothesis is equally parsimonious because they are similar in size and shape to modern bacteria (Moyer et al. 2014; Lindgren et al. 2015a). Although both physical and chemical evidence are not consistent with this hypothesis, it remains important to distinguish between preserved biofilms and melanosomes. Here, we present evidence of biofilms preserved in association with fossil integument from the early Cambrian, Middle Jurassic, and Early Cretaceous of China and compare the biofilm structures, including mitotic bacterial cells, to reported fossil melanosomes. The preserved biofilm structures that we report and fossil melanosomes are easily distinguishable both morphologically and chemically. Our study has additional implications for microbially-mediated preservation of early in the Chengjiang and non-biomineralized tissues in vertebrates in the

Daohugou, and Jehol lagerstätten.

Material and Methods

22 Specimens and Sampling

We sampled fossil chordate specimens from China with exceptionally-preserved integument for imaging and chemical analysis, including the skin of four specimens of the early Cambrian basal chordate Haikouella from the Maotianshan Shale, Mafang section, Haikou, Yunnan Province, the skin of an undescribed Middle Jurassic

Salamander (LSJC) from the Jiulongshan Formation, Daohugou Locality, Inner

Mongolia, and the feathers of an undescribed specimen of the Early Cretaceous ornithurine bird c.f. Hongshanornis (ORN) from China. Samples approximately 1 mm in length were removed from the exceptionally-preserved tissues and matrix of each of the fossils following standard techniques (Li et al. 2010). Two samples were removed from each sampling locality for LSJC and Hongshanornis: one for electron microscopy and a second from the same area of the fossil for chemical characterization. Only one sample from each sampling location was removed from the Haikouella fossils. None of the fossils were exposed to glues or other chemicals prior to analysis.

Scanning Electron Microscopy and EDX Analysis

We fixed the samples to aluminum stubs using carbon tape and imaged them using a JSM-7401 Field Emission Scanning Electron Microscope (JEOL Solutions for

Innovation). Hongshanornis and salamander samples were coated in gold-palladium using a Polaron E5000 Sputter-Coater (Quorum Technologies) for three minutes, then imaged at a working distance of 7.0 mm at 7 kV. Due to limited sampling, Haikouella samples were not sputter-coated prior to imaging and the SEM was operated at 1 kV and a working distance of 8 mm for these samples. One uncoated Haikouella skin sample was

23 also analyzed using Energy-dispersive X-ray spectroscopy (EDX) with a working distance of 8.0 mm at 15 kV.

We obtained length and width measurements of microbodies imaged in

Haikouella matrix sample 13 using ImageJ, then calculated the length:width (aspect) ratio for comparison to measurements for modern and fossil melanosomes. We did not measure microbodies that had transverse indents across the center because their morphology is not indicative of the size of individual cells.

Raman Spectroscopy

We used Raman spectroscopy to chemically characterize samples from each of the fossils. We completed our Raman analysis of the fossil tissue and matrix samples using a LabRAM High Resolution Raman microscope (HORIBA Scientific). We used a

50 mW 532 nm excitation laser through a 10% filter and a 50x objective in at least three different areas on each sample to test repeatability of the spectra acquired. Spectra were collected using a five second exposure time x one acquisition. Since fossils from the

Chengjiang biota are typically preserved with iron oxides and framboidal pyrite (Gabbott et al. 2004; Zhu et al. 2005), we also analyzed pyrite and iron oxide minerals (hematite and goethite) for comparison with the Haikouella samples. To test for the presence of preserved eumelanin, we also compared the results to Raman spectra for eumelanin samples extracted from modern bird feathers. Raman spectroscopy has been used previously for the detection of eumelanin in the fossil record, but this method may be susceptible to carbon contamination, as carbonaceous fossils can have similar peaks to eumelanin (Chapter 4). We therefore compared the samples to Raman spectra from

24 carbon-based materials, including graphite, carbon black, coal, and fossil plant samples

(see Chapter 4). Lastly, to compare the Raman spectra from the fossils to modern bacterial biofilm, we analyzed the common keratinolytic feather bacterium Bacillus licheniformis. The bacteria were grown on a sterile white nape feather of the Mallard

Anas platyrhynchos overnight at 37˚C, then placed on a glass slide and subjected to

Raman analysis. The center of the resultant biofilm was analyzed using a 1% filter. All samples were checked for burning between trials and spectra from burned areas were removed from analysis. All Raman spectra were graphed in Origin Pro 8.5 (OriginLab) and peaks were fitted using a Gaussian distribution in IgorPro6 (Wavemetrics).

Time-of-Flight Secondary Ion Mass Spectrometry

We used Time-of-Flight Secondary Ion Mass Spectrometry (ToF-SIMS) as an additional method for characterizing sample chemistry for Haikouella. Three fossil tissue samples (5, 6, 10) and two matrix samples (3, 13) were compared to a Sepia melanin standard and eumelanin extracted from mallard (Anas platyrhynchos) feathers. The Sepia melanin powder was pressed onto indium foil on top of tape while the mallard melanin and Haikouella samples were placed directly onto copper tape. Each sample was scanned using an unbunched pulsed gallium ion beam with a beam size of 500 nm on a

PHI TRIFT V nanoTOF (Ulvac-Phi, Inc.). We used an ultra-high vacuum of -6 torr for 10 minutes with a field of view of 200 µm x 200 µm. We detected secondary ions using both positive and negative polarity. Mass calibration was conducted using aluminum. We analyzed 38 peaks based on theoretical masses for eumelanin (Lindgren et al. 2015).

Relative intensity was calculated by dividing the observed intensity for each sample at a

25 given theoretical mass by the sum of all intensities for that mass, then z-scores were calculated based on the mean and standard deviation of the intensities of each theoretical mass for all samples (Table A.2; Colleary et al. 2015). A principal component analysis was performed on the resulting relative intensities using R (R Development Core Team

2007).

Results

Fossil Microstructures

SEM of the Haikouella samples revealed that both the fossil integument and matrix surfaces were predominantly covered in clay minerals. In several areas, the clay is overlain by thin, flattened branching structures or thick amorphous structures that often exhibit a wrinkled texture and branched outlines that appear to be stuck to the clay matrix

(Figure 3.1). Discernable microbodies are preserved on the surface of an amorphous structure in Haikouella matrix sample 13 (Figure 3.2). These microbodies have an average length of 1.01 µm, width of 0.69 µm, and 1.49 aspect ratio. At least 35% of them are indented transversely across their center. In most examples, this indentation is shallow, but there are also examples in which it is so deep that two distinct microbodies can be discerned (Figure 3.2D).

26

Figure 3.1. Biofilm structures preserved in Haikouella. A-C, sampling maps for specimens 1-3, respectively, of Haikouella from the Maotianshan Shale, Mafang section, Yunnan Province, China; D-G,

SEM images of the biofilm; D, wrinkled amorphous structure interpreted as extracellular polymeric substances; E, outgassing bubbles; F-G, examples of clay mineral matrix overlain by a hyphal network.

Scale bars = 10 µm for D and F, 1 µm for E and G.

27

Figure 3.2. SEM images of microbodies and biofilm preserved on Haikouella matrix sample 13. A, zoomed out image of the amorphous biofilm structure. B, zoomed in on microbodies in A. C and D, individual microbodies with arrows pointing to mitotic fissures. Scale bars = 10 µm for A, 1 µm for B-D.

Much of LSJC is covered in an amorphous material. The tail sample (Sample 1) exhibits mounds of thick, ropy material that are connected by a thick branching network

(Figure 3.3). Some branches have a bead-like microstructure (Figure 3.3D). The feather samples from Hongshanornis also show thick branching structures, but these are stick- like rather than flattened (Figure 3.4), as they are in Haikouella and LSJC. The surface of the Hongshanornis feather samples is littered with thick amorphous structures (Figure

3.4C) and much of the matrix sample is covered in an amorphous coating overlain by occasional thin branching structures (Figure 3.4D).

28

Figure 3.3. Biofilm structures preserved in LSJC from the Jiulongshan Formation, Daohugou Locality,

Inner Mongolia, China. A, sampling map; B-D, SEM images of biofilm structures associated with the exceptionally preserved skin of LSJC. Scale bars = 100 µm for B, 10 µm for C, and 1 µm for D.

29

Figure 3.4. Biofilm preserved on c.f. Hongshanornis from China. A, sampling map; B-D, SEM images of preserved biofilm structures. B, branching stick-like structures and web-like structure (arrow) on feather sample 1; C, amorphous structure interpreted as extracellular polymeric substances on feather sample 2; D, matrix sample covered in an amorphous film and thin branching structures. Scale bars = 1 µm for B, 10 µm for C and D.

Chemical Analyses

Raman analysis of the Haikouella integument samples shows two main peaks

(Figure 3.5). A short, broad peak is present at approximately 1355 cm-1 and a taller, narrow peak at approximately 1592 cm-1 in eight of the integument samples (Specimen 1, samples 2, 4, and 5; Specimen 2, samples 6 and 7; Specimen 3, samples 9 and 10;

30 Specimen 4, sample 1; Figure A.1). Peak location and peak morphology in these fossil samples contrast those from eumelanin extracted from modern bird feathers and proposed fossil eumelanin samples (Chapter 4). Eumelanin is characterized by two broad Raman peaks, including a short peak at approximately 1370 cm-1 and a taller peak at approximately 1570 cm-1 (Galván 2013). The 1570 cm-1 peak is usually broader than the sharp 1592 cm-1 peak in the Haikouella fossil samples. Instead, Haikouella fossil peaks resemble those in carbon samples (Figure 3.5). Carbon black, coal, and graphite have

Raman peaks at 1361 and 1591 cm-1, 1363 and 1587.5 cm-1, and 1354 and 1584 cm-1, respectively (Figure A.2). Peak morphology resembles that of coal, which due to its rough surface texture exhibits more fluorescence (i.e. a more intense baseline with fuzzier peaks) than the graphite sample. EDX analysis also suggested the presence of carbon in a

Haikouella fossil sample and neither EDX nor ToF-SIMS showed the presence of eumelanin in the fossil samples. Additionally, the Raman peaks do not resemble pyrite nor the iron oxide minerals (hematite and goethite) tested (Figure A.2). Thus, we interpret the Raman peaks as carbon peaks rather than eumelanin. Two of the matrix samples also present potential carbon peaks. Sample 11 exhibits similar peaks to the integument samples while sample 13 exhibits a narrow peak around 1390 cm-1 that may be indicative of the presence of carbon on the sample surface (Figure 3.5b). However, fluorescence is so high in sample 13 that the presence of other Raman peaks cannot be determined.

Raman analysis of LSJC and Hongshanornis integument samples did not reveal discernable Raman peaks in this region (Figure A.3).

31

Figure 3.5. Chemical characterization of the exceptionally preserved Haikouella integument. A, a comparison of Raman spectra from Haikouella integument sample 9, coal, eumelanin extracted from a mallard (Anas platyrhynchos) feather, and a biofilm consisting of the modern feather-degrading bacterium

Bacillus licheniformis showing differences between the major carbon peaks and eumelanin; B, EDX analysis of the Haikouella integument from Specimen 4, sample 1.

EDX analysis of the Haikouella samples reveal that they are composed of Si, Al,

O, K, Mg, C, F, and Fe. The presence of Si, Al, and K is consistent with previous EDX analyses of the aluminosilicate-rich clay matrix of other fossils from the Chengjiang lagerstätte (Zhu et al. 2005). Iron oxide may also be present in the clay matrix or as a diagenetic product of original pyrite material, which is responsible for the preservation of many Chengjiang fossils (Gabbott et al. 2004; Zhu et al. 2005), but we did not find evidence of framboids or euhedral crystals during our SEM analysis. There were also no

Raman peaks for iron oxide species or pyrite, although this could be because Raman usually only detects the most Raman active molecule (i.e. molecules heavy in non-polar bonds, such as C-C). ToF-SIMS mass spectra corroborate the presence of Si, Al, O, K, C,

32 Fe, and Mg and a lack of melanin preservation in both the Haikouella fossil and matrix samples (Figure A.4-A.6).

Discussion

The structures preserved both in the fossil integument samples and in the matrix of the Middle Jurassic salamander LSJC, the Early Cretaceous ornithurine bird

Hongshanornis, and four specimens of the early Cambrian chordate Haikouella resemble modern microbial biofilms that consist of a mix of bacteria and fungi. Amorphous structures that occur as both coverings over the fossil matrix and as isolated masses in all three taxa appear to be the mineralized remains of extracellular polymeric substances

(EPS; Figure 3.1-3.4). Tears and folds within the EPS and isolated masses of EPS may be indicative of dehydration prior to mineralization. Branching stick-like and rope-like structures present in all three taxa are probably the mineralized remains of fungal hyphae

(Borkow and Babcock 2003). We also interpret the microbodies preserved on Haikouella matrix sample 13 as bacteria. The rounded shape of these microbodies is similar to that of modern coccoidal bacteria and none of the chemical analyses performed on this sample detected preserved melanin. These microbodies are also preserved in association with

EPS on a matrix sample rather than within fossil integument. Furthermore, transverse indents in approximately 35% of the cells appear to be mitotic fissures associated with the late (i.e. anaphase and telophase) stages of mitosis (Figure 3.2C-D). The bead-like microstructures associated with possible hyphae and EPS in LSJC may likewise be diagenetically mineralized bacterial cells (Figure 3.3D). Additionally, chemical analyses

33 of all fossil samples suggest that melanin is not preserved (Figure 3.5, A.1-A.3, A.5)

Chemical results for Haikouella suggest that preservation may be due to microbially- mediated replication by Fe-rich aluminosilicate minerals and carbonized organic material

(Gabbott et al. 2004; Zhu et al. 2005).

It is unlikely that the bacteria and biofilms that we have described are modern contaminants. Thin biofilm structures like the web-like EPS in ORN and near-vertically- oriented hyphal structures and EPS did not collapse in the strong vacuum of the field- emission scanning electron microscope, no matter whether these structures were sputter- coated, as in LSJC and ORN, or uncoated, as in Haikouella. The samples were not fixed with glutaraldehyde or other preservational substances prior to analysis. Raman spectroscopy was also performed prior to SEM analysis on the same samples for

Haikouella. When viewed under SEM, the biofilm structures did not show evidence of damage from laser bombardment (e.g. collapsed structures, holes, or other evidence of burning), as might be expected for an unmineralized biofilm. Additionally, the Raman results from the Haikouella bacteria more closely resemble the carbon control samples than the modern Bacillus licheniformis biofilm and the narrow peak around 1390 cm-1 in the matrix sample with biofilm preservation (13) is absent in the modern biofilm (Figure

3.5, A.1).

Microbial mats are commonly associated with exceptional preservation of soft parts in Konservat-Lagerstätten around the world, including the Chengjiang and Jehol lagerstätten. Biofilms are hypothesized to have enveloped the carcasses, thus shielding them from decay organisms, altering the geochemical environment within the biofilm structure, and promoting early diagenetic mineralization of non-biomineralized structures

34 (Briggs 2003). Zhu et al. (2005) reported pyrite framboids and microspherules composed of pyrite in the Chengjiang worm Maotianshania and the Waptia, as well as microspherules composed of apatite in the worm Palaeoscolex with a possible bacterial affinity. They also report organic coatings on some of the same specimens as well as

Anomalocaris and suggest that diagenetic mineralization of Chengjiang animals was microbially mediated (Zhu et al. 2005). While fungal hyphae similar to those proposed here have not been described in Chengjiang fossils, microbial halos composed of similar bacterial and fungal structures, including probable cells, EPS, and hyphae, comprise pyrite crusts and concretions surrounding fossils of the Middle Alden Pyrite

Bed (Ledyard Shale Member, Ludlow Formation, New York; Borkow and Babcock

2003).

To our knowledge, our study represents the first example of biofilm structures associated with the Daohugou lagerstätte. The presence of preserved microbial structures in LSJC suggests that exceptional preservation of Daohugou fossils may at least in part have been microbially mediated. Preservation of melanosomes and melanin in other salamanders from the same locality suggests that melanin also plays a role in skin preservation in the Jiulongshan Formation (Chapter 5). We did not find evidence of melanin or melanosome preservation in LSJC, although we did not section the samples to test if they may be preserved bel the biofilm. In the Jehol lagerstätte, microbial mats may be responsible for thin laminations, mound-like structures, and wrinkles in fine-grained lacustrine sediments in the (Fürsich et al. 2007; Hethke et al. 2013).

Putative microbial biofilms and pyrite framboids have been described in association with

Jehol insect fossils (Wang et al. 2012).

35 The structures that we have described and interpreted as fossil microbial biofilms and bacteria differ morphologically from microbodies interpreted as melanosomes. First, biofilm structures, including bacterial cells, are found in the matrix as well as the surrounding sediment. Putative fossil melanosomes are limited to the fossil itself. Second, while the coccoid shape of the bacteria preserved in Haikouella matrix sample 13 may appear superficially similar to that of the sub-spherical melanosomes in red-brown feathers and hairs (Li et al. 2010) and the skin of basal and non-amniote vertebrates (Li et al. 2014; Chapter 5), they are 1.5-2x the average size of these melanosomes in modern animals and their fossilized counterparts. Furthermore, more than 35% of the bacteria preserve mitotic fissures, which are not present in reported examples fossil melanosomes nor do modern melanosomes undergo mitosis. The fossil bacteria are also preserved atop an amorphous biofilm structure, whereas fossil melanosomes are preserved as indentations in the matrix or as three-dimensional microbodies associated with the matrix of fossil integument (e.g. Vinther et al. 2008,

2010; Clarke et al. 2010; Li et al. 2010, 2012, 2014; Zhang et al. 2010; Chapter 4; Hu et al. 2018), eye (Tanaka et al. 2014, 2017; Clements et al. 2016), or ink samples (Glass et al. 2012). Lastly, none of the chemical analyses presented here detected melanin preserved in any of the Haikouella or LSJC samples. Instead, carbon, iron oxides, and aluminosilicates are present in the Haikouella samples. Eumelanin has been detected numerous times in association with the preservation of fossil melanosomes in integument and as well as melanin granules in fossil ink, predominantly with ToF-SIMS

(Lindgren et al. 2014, 2015a, 2015b; Colleary et al. 2015; Clements et al. 2016), but also

Raman spectroscopy (Chapter 4), trace metal mapping (Wogelius et al. 2011), and high-

36 performance liquid chromatography, XPS, and FTIR (Glass et al. 2012). Based on previous evidence (reviewed in Vinther 2015), evidence for biofilm and bacterial preservation in the fossils we have described, and the above comparisons, the hypothesis that microbodies preserved in association with fossil integument, eyes, and ink are microbial remains is unsupported and the melanosome hypothesis remains the most parsimonious.

We have described evidence of fossil biofilm structures preserved in association with early Cambrian, Middle Jurassic, and Early Cretaceous chordates. Like other reported examples, these fossil biofilms resemble modern biofilms composed of a mix of fungal structures and bacteria but differ both structurally and chemically from reported examples of fossil melanosomes. Additionally, the biofilms are found in both the fossil samples and the surrounding matrix, whereas melanosome preservation is strictly limited to preserved tissues that are melanic in modern animals. We therefore contend that while bacteria and their associated biofilms may be preserved in the fossil record, they can be easily differentiated from preserved melanosomes.

37 CHAPTER 4

THE PLUMAGE AND COLORATION OF AN ENANTIORNITHINE BIRD FROM

THE EARLY CRETACEOUS OF CHINA

Introduction

The exceptionally-preserved dinosaurs from the Lower Cretaceous Jehol lagerstätte of northeastern China have greatly contributed to our understanding of avian evolution and ecology. The Jehol Biota preserves a highly diverse assemblage of avian and non-avian dinosaurs, in many cases including the integument, allowing for inferences concerning avian flight (Zhou 2004; Zhou et al. 2008; Norell and Xu 2005; Clarke and

Middleton 2008; Heers and Dial 2012; Xu et al. 2014) and coloration (Li et al. 2010,

2012; Huang et al., 2016). Here, we describe a new bohaiornithid bird specimen from the

Early Cretaceous with remarkably preserved feathers and present, to our knowledge, the first reconstruction of the color of an enantiornithine bird.

Bohaiornithidae, including Bohaiornis, Longusunguis, Parabohaiornis, Shenqiornis,

Sulcavis, and Zhouornis, is the most speciose recognized of enantiornithine birds.

This derived group, of which Shenqiornis is the oldest, spanned across at least five million years (120-125 Ma; Wang et al. 2014a). While potential diets and dominant

38 habitats have been inferred for some bohaiornithids, little is known about their ecology

(Li et al. 2014a; Wang et al. 2014a). Feather coloration is relevant to numerous aspects of extant avian ecology, including , camouflage, and other inter- and intraspecific communications (McGraw 2006). Poor preservation of the plumage in previously described Bohaiornithidae means that nothing is known of their plumage color. CUGB P1202 is a young individual that retains the best preservation of plumage of any bohaiornithid, including preserved melanosomes with chemically preserved melanin, exceptionally detailed primary remiges, an , elongate crown feathers, contour feathers of the nape and body, numerous short rectrices, and paired elongate, rachis- dominated rectrices.

Material and Methods

Specimen and Sampling

Specimen CUGB P1202 (CUGB, China University of Geosciences, ), consists of part and counterpart. It was collected from the Early Cretaceous Jiufotang

Formation, near Lamadong Village, Jianchang County, Province, China. and feather morphology was measured from high-resolution digital images using ImageJ

(available for download at http://rsbweb.nih.gov/ij/). Using previously described techniques (Li et al. 2010), we removed samples measuring approximately 1 mm in length from four different exceptionally-preserved feathers of the CUGB P1202 primary slab (Figure A.7) for scanning electron microscopy (SEM) and Raman Spectroscopy. We removed one sample for SEM and a separate, adjacent sample for Raman spectroscopy,

39 from feathers of the crown, the nape, the right wing next to the ulna, and contour feathers next to the distal right tibia and proximal . As a control, we also sampled the -colored matrix in which the fossil is preserved, white-colored matrix above the fossil, and white-colored matrix between the and the preserved feathers. Sampling locations within the white matrix include near the cranial and caudal cervical vertebrae, below the right dentary, and above the left humeral head on the counter slab (Figure A.7).

As in the case of the fossil samples, we took separate samples from each targeted area of the matrix for Raman spectroscopy and SEM. X-ray computed tomography images of the specimen were acquired using a High-Resolution X-Ray CT scanner (Nikon XT H 320

LC) at The China University of Geosciences. We generated 80 slices with an image resolution of 2000 x 2000 along the coronal axis of the primary slab using a voltage of 180 kV, a current of 118 μA, no filter, and a 0.075211 mm voxel size.

Scanning Electron Microscopy

Before imaging, we placed the fossil and matrix samples on carbon tape and sputter-coated them for three minutes in gold-palladium using a Polaron E5000 Sputter-

Coater (Quorum Technologies). We used a JSM-7401 Field Emission Scanning Electron

Microscope (JEOL Solutions for Innovation) to image the samples at 7 kV and a 7 mm working distance.

Based on chemical and morphological similarities to melanosomes in extant bird feathers (see below), elongate three-dimensional granules and impressions preserved in the feathers of CUGB P1202 are interpreted as melanosomes. We measured the size and shape of melanosomes of both preservational modes and compared these parameters to

40 melanosome morphology in extant organisms using previously described techniques

(Clarke et al. 2010; Li et al. 2010, 2012, 2014b). Lengths (long axis) and widths (short axis) of complete three-dimensionally preserved melanosomes and melanosome impressions were measured from SEM images using ImageJ. These values were then used to calculate the aspect (length:width) ratio of each measured melanosome.

We used quadratic canonical discriminate analysis (CDA) in JMP Pro 11 (SAS

Institute Inc.) to predict the color of CUGB P1202. This standard method has previously been used to predict the color of extant avian feathers with 82% accuracy (Li et al. 2012).

The size (length and width) and shape (aspect ratio) of melanosomes preserved in CUGB

P1202 were compared to those from extant bird feathers, which can be separated into five categories based on color and morphology: black, brown, grey, iridescent, and specialized “-type” melanosomes (Li et al. 2012). Experimental maturation of feathers suggests that melanosomes undergo up to 10 to 20% taphonomic shrinkage

(McNamara et al. 2013; Colleary et al. 2015). We therefore increased the length and width values of three-dimensionally preserved melanosomes in CUGB P1202 (sample

62; Figure A.8) by 10% and 20% to incorporate this range of potential taphonomic reduction in size.

Raman Spectroscopy

We used Raman spectroscopy to assess the presence of melanin in each sample.

This is a standard technique for qualitatively assessing the chemical structure of a sample by generating and recording vibrations of molecular bonds on the sample surface

(Colthup et al. 1990). Raman spectra were generated using a LabRAM High Resolution

41 Raman microscope (HORIBA Scientific) with a 50 mW 532 nm excitation laser through a 50x objective. We used a wafer of pure silicon to calibrate the instrument, then photobleached each sample using a 10% filter for 10 minutes prior to data collection.

Photobleaching is common technique in which a highly-fluorescent sample is irradiated by a laser with a low wavelength (usually 532 nm) prior to data collection to decrease fluorescent signals (Golcuk et al. 2006). Raman spectra were collected at least three times from each sample to account for repeatability of Raman peaks using a grating of 1200 lines per mm, a 400 µm pinhole, a 100 μm slit , and a wavenumber range of 300-

2500 cm-1. A five second exposure time x 1 acquisition was used for each collection.

These results were compared to spectra of melanin extracted from extant avian feathers: iridescent feathers of a mallard (Anas platyrhynchos) and a wild

(Meleagris gallopavo), black feathers of a (Gallus gallus) and a Red-winged

Blackbird (Agelaius phoeniceus), and brown feathers from a Cooper’s Hawk ( cooperii) and a House Wren (Troglodytes aedon). Melanin was extracted from the β- keratin matrix using a standard Proteinase-K-based technique (Liu et al. 2003).

We collected spectra from an unpigmented remige from the Sulfur-crested

Cockatoo Cacatua galerita as a standard for B-keratin and from carbon black as a control for the potential effect of carbon on the surface of the fossil samples. We also removed 2 mm samples from the cuticle preserved on the basal midrib of a Platanus leaf (PTRM

#20639) from the , Mud Buttes (PTRM #P88002),

Bowman County, North Dakota, and a carbonized area on a lycopsid root

Stigmaria (CMNH P-21773) from the Joggins Formation, Joggins, Nova Scotia, for

42 comparison to the CUGB P1202 samples as negative controls for melanin. We further collected spectra from a colony of the common feather-degrading (keratinolytic) bacterium Bacillus licheniformis to test the potential affinity of preserved granules in

CUGB P1202 to microbes (see Moyer et al., 2014). The colony of B. licheniformis was transferred from tryptic soy agar plates to a sterile unpigmented Mallard nape feather and grown overnight at 37 °C. Spectra from extant samples, carbon black, fossil plants, and bacteria were generated in a similar manner to the fossil samples, although they were not photobleached to avoid burning the samples. Both fossil and extant samples were examined for burning directly after spectra were collected. If a sample was burned, the resulting spectra were discarded. We used Origin Pro (OriginLab) to normalize the spectra for comparison and Igor Pro (Wavemetrics) to conduct peak fitting for each spectrum using a Gaussian distribution.

Institutional abbreviations. BMNHC, Beijing Museum of Natural History,

Beijing, China; CMNH, Cleveland Museum of Natural History, Cleveland, Ohio; CUGB,

China University of Geosciences, Beijing, China; IVPP, Institute of Vertebrate

Paleontology and Paleoanthropology, Beijing, China; PTRM, Pioneer Trails Regional

Museum, Bowman, North Dakota.

Systematic Paleontology

AVES Linnaeus, 1758

AVIALE Gauthier, 1986 sensu Gauthier and deQuieroz, 2001

ORNITHORACES Chiappe, 1995

43 Walker, 1981

BOHAIORNITHIDAE Wang et al., 2014a

Gen et sp. indet.

Specimen. CUGB P1202 consists of a nearly complete articulated specimen preserved in part and counterpart with well-preserved melanized feathers (Figure 4.1,

4.2). The matrix is dark grey in color and white material surrounds the skeleton. The bones and preserved feathers are black in color. Most of the skeletal elements are preserved in the primary slab (Figure 4.1a). With the exception of the skull, which is preserved in oblique lateral view, the skeleton is preserved in ventral view.

Ontogenetic status. The specimen is interpreted to be a sub-adult individual.

While it has a pair of elongate tail feathers that suggest that the animal had reach sexual maturity at the time of death, some pitting is visible on the periosteal surface of the tibia, suggesting that it had not yet reached skeletal maturity (Chiappe et al. 2007). Fusion of the distal carpals to the metacarpals and of the proximal tarsals to the tibia is absent while these states are variably present in other parts of the Bohaiornithidae (fused in the

Zhouornis holotype and Bohaiornis IVPP V17963; unfused in the Bohaiornis holotype,

Shenqiornis, Sulcavis, Zhouornis BMNHC Ph756, Parabohaiornis, and Longusunguis) and more basal avialans. CUGB P1202 is approximately half the size of inferred adult bohaiornithid specimens, including the Zhouornis holotype and the referred specimen of

Bohaiornis (Zhang et al. 2013; Li et al. 2014a) and metacarpal III is only about 61% of the size of Parabohaiornis V18691, the previous smallest bohaiornithid (45% Zhouornis,

49% Sulcavis, 51% Shenqiornis, 56% Longusunguis and Bohaiornis IVPP V17963;

44 Wang et al. 2014a). CUGB P1202 is therefore the smallest and potentially the youngest bohaiornithid specimen to date.

Figure 4.1. Ventral view of the new bohaiornithid specimen, (CUGB P1202, primary slab). A, photograph of the primary slab; and B, interpretive drawing. Anatomical abbreviations: co=; cv=cervical vertebrae; fe=femur; fu=; hu=humerus; il=; mcI-III=metacarpals I-III; mt=metatarsals; pd I-

IV=pedal digits I-IV; phI-1=first phalanx of I; phII/III-1/2=first/second phalanx of digit II/III; pu=pubis; ra=; rad=radiale; mcI-III=metacarpals I-III; py=; sc=; sk=skull; sml=semilunate carpal; st=sternum; sy=synsacrum; th=thoracic vertebrae; ti=tibia; ul=ulna; uln=ulnare.

Scale=1 cm.

45

Figure 4.2. The new bohaiornithid specimen, (CUGB P1202, counter slab).

Occurrence. Early Cretaceous Jiufotang Formation, near Lamadong Village,

Jianchang County, Liaoning Province, China.

Referral to the Bohaiornithidae. Based on Chiappe (2002) and Chiappe and

Walker (2002), synapomorphies of Enantiornithes that are recognizable in CUGB P1202 include a dorsal margin of the furcula that is narrower than the ventral margin, a well- developed furcular apophysis, a concave central margin of the humeral head, metacarpal

III extending further distally than metacarpal II, a deep intercondylar sulcus of the

46 , and metatarsals II and III wider than metatarsal IV. CUGB P1202 is further referred to Bohaiornithidae based on the following shared synapomorphies (Wang et al.

2014a): large caudally-recurved teeth that are robust and sub-conical in shape the rostral- most of which is smaller than the caudal teeth, a mediolaterally-expanded sternal midline, caudally-expanded lateral trabeculae of the sternum, blunt omally-expanded tips of the furcula, an elongate, robust scapular acromion process, and an elongate pedal digit III ungual that measures > 40% (60%) of the length of the tarsometatarsus and is the longest pedal digit.

Preservation. CUGB P1202 consists of a nearly complete skeleton with preserved melanized feathers. A white “halo” surrounds and intermingles with skeletal elements and obscures the proximal ends of feathers. A similar “halo” has previously been described in Microraptor sp. and suggested to be the result of changes in matrix chemistry due to water being trapped between the feathers and the body (Hone et al.

2010). In some specimens (e.g., Hone et al. 2010) the halo appears to have a gradational contact with the preserved feathers and often occurs in cracks in the rock itself. In CUGB

P1202, this contact is more abrupt and the white material does not occur in cracks in the matrix.

Description

The skull is preserved in left oblique lateral view. The premaxilla contains four robust teeth that begin in the rostral portion of the premaxilla and several neurovascular foramina. Like other bohaiornithids (Wang et al. 2014a), the rostral-most tooth is slightly

47 smaller in size than the other teeth and all of the teeth are unserrated. The nasal process of the premaxilla is elongate and contacts the nasal. The nasal is broad and slightly displaced. The is damaged caudally. Four robust, conical teeth are preserved in the cranial portion of the maxilla. The maxilla has a relatively long nasal process and has a slender dorsal process that is obscured dorsally by matrix. The frontal has been crushed cranially and obscures the orbit. The jugal is displaced caudally. Like Bohaiornis, the jugal is elongate and strap-like, but lacks the slight ventral recurvature that is characteristic of Bohaiornis (Li et al. 2014a). The jugal bears an elongate caudodorsal process that projects dorsally in a manner similar to Shenqiornis and Bohaiornis (Wang et al. 2010; Li et al. 2014a).

The dentaries are straight and unfused rostrally. The rostral portion of the left dentary has been displaced caudodorsally so that it lies directly ventral to the caudal premaxilla. The right dentary shows a visible Meckel’s groove, which like Bohaiornis,

Longusunguis, and Parabohaiornis (Li et al. 2014a; Wang et al. 2014a) does not extend to its rostral tip. Both dentaries contain at least three robust, slightly recurved, conical teeth in the rostral portion, each of which lack serrations. CUGB P1202 has fewer dentary teeth than any other bohaiornithid. Dental morphology is opposed to , which has many small teeth that lack recurvature (Zhou et al. 2008). Like many other enantiornithines (Chiappe and Walker 2002), the dentary slopes caudoventrally. Caudal elements of the , including the surangular, angular, articular, and lateral are displaced caudally. The surangular has a slightly concave dorsal margin. A process on the dorsal surface of the surangular, interpreted as the lateral mandibular process by

Wang et al. (2014a), is not preserved on the left surangular, but may be visible on the

48 right surangular when viewed in the x-ray computed tomography image (Figure 4.3d). A lateral condyle is preserved on the articular and there may be a short retroarticular process, although the articular is incompletely preserved caudally.

At least eight articulated cervical vertebrae are preserved in ventral view. The cervical vertebrae are approximately as wide as they are long and the cranial cervical vertebrae are weakly heterocoelous. As in other enantiornithines, the centra of the cranial cervical vertebrae bear a small ventral process (Chiappe and Walker 2002). The thoracic series is poorly preserved. Three thoracic vertebrae are preserved on the surface of the part slab and at least two more are revealed by CT scanning. The centra bear deep lateral depressions. Ribs and gastralia are disarticulated and preservation is split between the part slab and the counterpart slab. Most of the exposed ribs are preserved as casts. Six gastralia are preserved on the counterpart slab. Little of the synsacrum is visible on the surface of the part slab. The sacrum is comprised of at least seven fused sacral vertebrae.

It bears a thin, shallow central groove similar to that of Zhouornis (Zhang et al. 2014). At least three disarticulated free caudal vertebrae are preserved. As in other basally divergent avialans, two of the caudal vertebrae retain elongate transverse processes. An elongate rod-like pygostyle comprised of at least four vertebrae is displaced caudally from the synsacrum. The pygostyle is approximately 13.06 mm long and is proximately slightly expanded into a T-shape showing the incorporation of laterally projected transverse processes. Distally, the pygostyle gradually constricts in a manner similar to other bohaiornithids (Wang et al. 2014a). The distal end is cracked and is displaced caudally from the rest of the pygostyle.

49

Figure 4.3. The skull in CUGB P1202. A, photograph of the skull; and B, interpretive drawing; C, details of the dentaries and teeth; D, CT scan of the skull. Anatomical abbreviations: cv=cervical vertebrae; fr=frontal; ju=jugal; lan=left angular; lar=left articular; lde=left dentary; llc=left lateral condyle; lmp=lateral mandibular process; lsa=left surangular; ma=maxilla; na=nasal; pa=parietal; pm=premaxilla; rde=right dentary; rsa=right surangular. Scale = 1 cm for A, C, and D; 0.5 cm for B.

Much of the sternum is preserved in dorsal view on the counter slab (Figure 4.4a-b). Only a small portion of the cranial end of the sternum and the distal left caudolateral process are preserved on the part slab in ventral view (Figure 4.4c). The cranial section is poorly preserved and reveals little about the cranial morphology of the sternum. The counter slab preserves two elongate caudolateral processes (lateral trabeculae) and two shorter medial trabeculae. The left lateral trabecula is broken and displaced caudolaterally from the rest of the sternum. The distal ends of the lateral trabeculae are caudally-expanded, as in other bohaiornithids. Similar medial and lateral expansions have been

50

Figure 4.4. The pectoral girdle in CUGB P1202. A, photograph of the sternum preserved in dorsal view on the counter slab; B, interpretive drawing of the sternum preserved in dorsal view on the counter slab; C, photograph; and D, CT scan of the primary slab, preserved in ventral view. Anatomical abbreviations: ac=acromion process; ap=furcular apophysis; co=coracoid; cv=cervical vertebrae; fu=furcula; gf=glenoid fossa; lt=lateral trabecula; rb=ribs; sc=scapula; sf=n. supracoracoideus foramen; sm=sternal midline; st=sternum. Scale=1 cm for A, B, and C; 15 mm for D. reported for Zhouornis (Zhang et al. 2013) and Parabohaiornis (Wang et al. 2014a). The right caudolateral process is displaced from the rest of the sternum and the left caudolateral process is broken distally. The caudal process is subequal in length to the caudolateral processes and the central portion is obscured by a large crack in the rock. A weakly-keeled sternal midline extends to approximately the caudolateral process, which

51 expands caudally into a weak T-shape. A similar morphology has been reported for the referred specimen of Bohaiornis (Li et al. 2014a).

Both are preserved in ventral view on the primary slab (Figure 4.4).

They are also visible in dorsal view the CT images (Figure 4.4d). The acrocoracoid process is slightly medially hooked, as in Bohaiornis (Li et al. 2014a). An enclosed N. supracoracoideus foramen is visible in the CT images (Figure 4.4d). The lateral margin of the coracoid is straight to slightly concave. The sternal margin is expanded and slightly concave, as in the referred specimen of Bohaiornis (Li et al. 2014a). The furcula is robust and generally Y-shaped (Figure 4.4c). Its apophysis is long and thin and the furcular rami bear a slight ventral depression as in other enantiornithines. The proximal ends of the scapulae of CUGB P1202 are visible on the surface of the specimen. The acromion process is greatly extended, robust, and generally rectangular in shape with a slight omal expansion, as in Bohaiornis. The scapular blades are strap-like (Figure 4.4d).

The is approximately 80% the total length of the hind . The humerus is slightly longer than the ulna and is preserved in cranial view (Figure 4.5). Generally, the humerus bears a slight sigmoidal shape. The left humeral head is not preserved, but the right humeral head is visibly strap-like and nearly flat. This morphology is similar to other bohaiornithids, but opposed to the globose humeral head in Pengornis and

Xiangornis (Zhou et al. 2008; Hu et al. 2012). The deltopectoral crest is broad with a convex dorsal margin and extends approximately one-fourth of the length of the humerus.

In Sulcavis, the deltopectoral crest extends greater than one-third of the humerus

(O’Connor et al. 2013). The deltopectoral crest extends one-third of the humeral length in

52 other bohaiornithids. The distal terminus of the deltopectoral crest is less abrupt than in

Shenqiornis and Parabohaiornis (Wang et al. 2010, 2014a).

Figure 4.5. The forelimbs of CUGB P1202. A, left wing; B, CT scan of the caudal right humerus; and C, right wing. Anatomical abbreviations: I/II/III=digits I-III; ci=capital incisure; dc=deltopectoral crest; hh=humeral head; hu=humerus; mcI/II/III=metacarpals I-III; ra=radius; rad=radiale; rg=radial groove; sml=semilunate; uln=ulnare; vt=ventral tubercle. Scale=1 cm for A and C; 15 mm for B.

The right radius and ulna are cracked and damaged. The ulna is significantly shorter than the humerus, which is contrary to the longer ulna preserved in Sulcavis, the

Zhouornis holotype, and Longusunguis (O’Connor et al. 2013; Zhang et al. 2013; Wang et al. 2014a). The humerus and ulna are subequal in length in Shenqiornis, Bohaiornis, and Parabohaiornis (Wang et al. 2010, 2014a; Li et al. 2014a). The left ulna is better preserved and less than twice the width of the radius, unlike Bohaiornis (Li et al. 2014a).

The cranial surface of the radius retains a longitudinal groove that is also present in

Bohaiornis, Parabohaiornis, Longusunguis, Zhouornis (Wang et al. 2014a; Zhang et al.

2014), and other Enantiornithes (Chiappe and Walker 2002).

The radiale, ulnare, and semilunate carpals are preserved in ventral view on both sides of the specimen and are not fused. The ulnare is -shaped, as in other

53 enantiornithines. Metacarpal I is unfused and is short with a broadly rounded cranial margin. Metacarpals II and III appears to be fused proximally but not distally, although damage obscures much of the distal ends of the metacarpals on both sides. Like many other enantiornithines, metacarpal III is longer and extends further distally than metacarpal II. Metacarpal III is slightly curved. Phalanx I-1 is slender and straight. It articulates with a small ungual. Phalanx II-1 is longer and wider than phalanx II-2. Its margins are slightly concave and the general shape is rectangular. Phalanx III-1 is shorter than the other phalanges and does not appear to articulate with an ungual. It is proximally robust and narrows distally. The unguals are small and slightly recurved, as in many other enantiornithines (e.g. other bohaiornithids, , ).

The pelvic girdle is incompletely preserved in ventral view on the primary slab and in dorsal view on the counterpart slab, but much of it is visible in CT scans (Figure

4.6a-c). Small portions of bone, such as the shaft of the left pubis and partial distal end of the right pubis, and outlines of the pelvic bones are preserved in counterpart. The ilia are preserved on either side of the synsacrum, although the left ilium is better preserved. Like

Parabohaiornis (Wang et al. 2014a), the ilia are not fused to the sacrum. The proximal ends of the pubes are flared and overlie the proximal femora. The shafts of the pubes are elongate, narrow, and bowed slightly. The pubic foot bears a slight distal expansion, as in other bohaiornithids.

Bones of the hind limbs are also generally poorly preserved. The left femur is slightly bowed cranially. At the distal end of the left femur the medial and lateral appear to be separated by a shallow popliteal fossa. Pitting on the periosteal surface of the proximal end of the left tibia suggests that CUGB P1202 is a not an adult

54 individual (Figure 4.6d). The proximal tarsals are preserved on the right hind limb

(Figure 4.6e-g). The proximal end of the ascending process of the astragalus is not completely fused to the tibia. In CT images the distal tarsals are visibly not fused to the metatarsals (Figure 4.6f).

Breakage obscures the distal ends of the metatarsals on the left, but these are well- preserved on the right side (Figure 4.6f, g). Metatarsal II is broader than metatarsal IV, as in other enantiornithines (Chiappe and Walker 2002; Zhou et al. 2007) and some basal avialans (e.g. Asparavis ukhaana; Clarke and Norell 2002). Metatarsal II and metatarsal

IV are subequal in distal extent. The trochlea of metatarsals II and IV are curved slightly plantarly, as in Bohaiornis, Sulcavis, Zhouornis, and Parabohaiornis (Hu et al. 2011;

O’Connor et al. 2013; Zhang et al. 2013; Wang et al. 2014a). Metatarsal I is broad with a slight j-shape.

Pedal digits on both sides are cracked and incompletely preserved. Molds of the original positions of these bones are, however, preserved on the counterpart slab. Based on these molds, digit II is more robust than the other digits, as in other bohaiornithids

(Wang et al. 2014a). Digit III is longer than the other digits and ungual III is longer and less highly recurved than all other unguals. This morphology is similarly present in other bohaiornithids (Wang et al. 2014a). A keratinous sheath is preserved on all unguals in both part and counterpart. The elongate pedal unguals in CUGB P1202 are also consistent with the interpretation of bohaiornithids as predominately arboreal birds (Wang et al.

2014a).

55

Figure 4.6. The pelvic girdle and hind limbs of CUGB P1202. A, photograph of the pelvic girdle preserved on the primary slab; B, CT scan; C, photograph of the pelvic girdle preserved in dorsal view on the counterpart slab; D, periosteal pitting of the left tibia; E, right tarsus; F, CT scan of the right hind limb; and

G, photograph of the right hind limb. A, B, D-G are in ventral view. Anatomical Abbreviations:

I/II/III/IV=pedal digits I-IV; ast=astragalus; ca=free caudal vertebrae; fe=femur; ga=gastralia; il=ilium; mt

I-IV=metatarsals I-IV; sy=synsacrum; ti=tibia; th=thoracic vertebrae; pt=proximal tarsals; pu=pubis.

Scale=1 cm for A, B, and C; 5 mm for D, E, F, and G.

Plumage

Melanized feathers are preserved throughout the body of CUGB P1202. Feathers are poorly preserved proximally and the articulation between the bones and the feathers is masked by a white matrix “halo” that surrounds the skeleton. Distally, feathers are better preserved and individual barbs can often be distinguished. Preserved feathers include pennaceous body and crown contour feathers, facial feathers, primary and secondary wing feathers, alular feathers on the left side, short rectrices, and two elongate central

56 rectrices. Feathers may occur on the hind limbs, but these are indistinguishable from the covert feathers of the tail. Feathers do not occur on the pedal digits, unlike some maniraptoran dinosaurs (e.g. huxleyi; Hu et al. 2009).

CUGB P1202 retains long crown feathers near the caudal tip of the rostrum

(Figure 4.7). Similarly, elongate crown feathers occur in Microraptor, which have been shown to be the result of diagenetic compression of typical skull feathering rather than a projected crest (Li et al. 2012). Tiny, narrow feathers are preserved within the on the primary slab (Figure 4.7c).

As preserved, the are folded in toward the body and are not as darkly- colored as the crown, body contour feathers, and short rectrices (Figure 4.8). An alula is visible on the right side of the body (Figure 4.8c, d). The presence of an alula suggests that like some other enantiornithines CUGB P1202 was capable of maneuvered flight.

This feature is not preserved in any other bohaiornithid, but has been reported for other enantiornithines, including , Eoenantiornis, and Grabauornis (Sanz et al.

1996; Zhou et al. 2005; Dalsätt et al. 2014). Asymmetrically-veined primary feathers are visible and the distal tips of three of these feathers are distinguishable distally on the right side in counterpart (Figure 4.8). Four primaries can be distinguished on in the left wing.

The distal ends of other primary feathers are either not preserved or are obscured by a crack in the matrix. The longest of these feathers is approximately twice the combined length of the humerus and ulna. Two long, veined covert feathers with distinct barbs are preserved projecting from the right side. Individual covert feathers are otherwise not visible on the wings.

57

Figure 4.7. Crown and neck feathers as well as preserved melanosome morphologies in the new bohaiornithid. A, crown feathers; B, inset, details of loose feathers preserved near the crown; C, hair-like feathers (arrows) within the antorbital fenestra on the primary slab; D, molds of elongate melanosomes preserved near the base of the neck (triangle); and E, elongate three-dimensional melanosomes preserved in crown feathers (circle). Triangle and circle shapes represent melanosome sampling sites from equivalent points on the primary slab. Scale=1 cm for B; 0.25 cm for C; 1 µm for D and E.

Feathers near the tail include at least twelve short (approx. 20.02 mm), narrow feathers that surround the cranial caudal vertebrae and pygostyle (Figure 4.8e). The proximal 86.82 mm of two elongate rectrices are preserved missing their distal tips.

These two rectrices appear to be rachis-dominated and lack distinguishable barbs, as in

Confuciusornis, , and Bohaiornis (Chiappe et al. 1999; Zhang and Zhou

2000; Hu et al. 2011).

58

Figure 4.8. Details of the remiges and rectrices in the counterpart slab of CUGB P1202. A, the left wing; B, close-up image of the first primary wing feather showing individual barbs and the rachis, outlined in image

A; C, the right wing; D, details of the alula and coverts preserved on the right wing, outlined in image C; and E, short plumaceous rectrices and paired elongate rachis-dominated rectrices. Abbreviations: al=alula; cov=covert feathers; ra=rachis. Scale=1 cm for A, c; 0.25 cm for B.

Elongate granules are preserved both three-dimensionally and as impressions in exceptionally-preserved crown, nape, and body contour feathers from CUGB P1202

(Figure 4.6, 4.9; samples 62-64). We interpret these granules as melanosomes because their size, shape, and chemistry are similar to melanosomes from extant avian feathers.

The melanosomes are highly elongate, have high aspect ratios, and many show a degree of organization similar to melanosomes in extant bird feathers (Figure 4.6e; Table A.5).

These melanosomes are within the size and shape (Table A.5; mean length and aspect ratio, respectively, of sample 62=1159.23 nm and 7.13; sample 63=1727.36 nm and 5.74; sample 64=1628.83 nm and 7.11) range of only melanosomes from extant iridescent

59 feathers (751.54-1739.82 nm mean length, 3.80-8.57 mean aspect ratio; Li et al. 2012).

Quadratic canonical discriminant analysis predicted the color of these preserved feathers as iridescent with greater than 94% probabilities (sample 62=99.86%; sample

63=94.17%; sample 64=100%; Figure 4.10a). These results were not significantly affected by hypothesized 10% to 20% taphonomic shrinkage of three-dimensionally preserved melanosomes, as the probability of iridescent melanosomes in sample 62 remained above 99.82% (Figure A.8; McNamara et al. 2013; Colleary et al. 2015). The sample taken from the right wing (sample 65) may preserve degraded melanosomes

(Figure 4.9c), but these are unmeasurable and therefore excluded from the CDA analysis.

Matrix samples do not preserve similar textures or microbodies (Figure A.9).

Figure 4.9. Right wing and body contour feathers on the primary slab of CUGB P1202. A, photograph of the right wing and long contour feathers next to the tibia; B, molds of elongate melanosomes preserved near in contour feathers near the tibia (triangle); and C, potential degraded melanosomes preserved in the right wing (circle). Triangle and circle shapes represent melanosome sampling sites. Scale=1 cm for A; 1

µm for B; 10 µm for C.

60 Raman spectroscopy suggests that melanin may be preserved in CUGB P1202 feathers, confirming the affinity of the preserved microbodies to melanosomes. Extracted eumelanin from black, brown, and iridescent extant bird feathers shows two broad peaks: a smaller Raman peak between 1372 and 1386 cm-1 and a taller peak between 1574 and

1585 cm-1. Spectra from all four fossil samples show two broad peaks: a taller peak between 1566 and 1574 cm-1 and a shorter peak between 1353 and 1364 cm-1 (Figure

4.10a). These peaks are similar in both location and peak morphology to those generated by extracted eumelanin from extant black, brown, and iridescent feathers (Figure 4.10a).

Eumelanin has previously been reported to have two similar peaks in studies of extracted eumelanin from extant avian feathers and mammalian hair (Galván et al. 2013a, b), synthetic eumelanin (Perna et al. 2013), and eumelanin from Sepia officinalis ink (Huang et al. 2004; Centeno and Shamir 2008). Although all melanosomes likely contain some ratio of phaeomelanin to eumelanin (McGraw 2006; Liu et al. 2014), we could not determine the presence or absence of phaeomelanin within these samples. We also compared the peaks generated by the fossil to carbon black, which produced peaks with a similar wavenumber to the fossil spectra, but with a different peak morphology. Carbon peaks are thinner than fossil and extant eumelanin peaks and the 1360 cm-1 peak produced by carbon has a similar intensity to the 1590 cm-1 peak (Figure A.9a).

Eumelanin spectra are also not present in Bacillus licheniformis, the unpigmented white feather sample, nor the matrix samples (Figure A.9). However, the Platanus and

Stigmaria samples presented Raman spectra that are similar to the fossil samples in both peak morphology and location (Figure A.10). We therefore caution that Raman spectroscopy alone is not a definitive test for the presence of melanin in fossil samples.

61 However, coupled with morphological evidence of preserved melanosomes from scanning electron microscopy and a eumelanin signal that is limited to the preserved plumage, Raman spectroscopy serves as an effective indicator for the preservation of eumelanin in CUGB P1202.

Figure 4.10. Statistical analysis and Raman spectra for fossil and extant samples. A, Raman spectra for extracted melanin from extant birds (top) and CUGB P1202 samples (bottom). The black dashed line shows the position of a shorter peak at approximately 1370 cm-1. The grey dashed line denotes the position of a taller peak at approximately 1570 cm-1; and B, quadratic canonical discriminant analysis of melanosome morphology. Open circles and numbers refer to CUGB P1202 samples. Colored points refer to average feather melanosome morphologies from extant avian species (Database S1 in Li et al. 2012). Black points are black, grey points are grey, brown points are red-brown, points are iridescent, and blue points are penguin-type melanosome morphologies. Circles represent the 95% confidence interval.

Abbreviations: Irid=iridescent; Peng=penguin-type; RWB=Red-winged Blackbird.

62 Discussion

CUGB P1202 can be differentiated from other bohaiornithids by a suite of skeletal characteristics, such as a strap-like jugal bar that lacks ventral recurvature

(present in Bohaiornis and Longusunguis) and has a dorsally-deflected caudal end (dorsal deflection absent in Zhouornis and Parabohaiornis) that is not forked (forked in

Longusunguis and Shenqiornis; sensu Wang et al. 2014a); a deltopectoral crest that extends approximately one-fourth of the length of the humerus (greater than one-third in

Sulcavis; approximately one-third of the length of the humerus in other bohaiornithids) with a gradual distal terminus (abrupt in Sulcavis, Shenqiornis, and Parabohaiornis); a humerus that is longer than the ulna (subequal in length in Bohaiornis, Parabohaiornis, and Shenqiornis; humerus shorter in Sulcavis, the Zhouornis holotype, and

Longusunguis). However, because ontogenetic variation in these features has not been addressed, referral to a specific species in the Bohaiornithidae is not presently supported.

The new specimen presents a combination of juvenile skeletal features and display-linked plumage characteristics that inform the role of sexual selection in

Enantiornithes. Currently, there is comparatively little evidence available for ontogenetic shifts in morphology within the Bohaiornithidae and CUGB P1202 is the smallest and possibly the youngest bohaiornithid yet described. The holotype of Zhouornis and the referred specimen of Bohaiornis are inferred to be adult specimens based on their large size and complete fusion of compound bones, including the carpometacarpus, tibiotarsus, and tarsometatarsus (Li et al. 2014a; Wang et al. 2014a). CUGB P1202 is approximately half the size of both adult bohaiornithid specimens and only about 75% of the size of the

63 previous smallest specimen (Parabohaiornis V 18691) based on the length of the long bones. Like other subadult bohaiornithids, including Shenqiornis, Sulcavis,

Longusunguis, Parabohaiornis, and Zhouornis BMNHC Ph 756 (Xuri et al. 2010;

O’Connor et al. 2013; Wang et al. 2014a; Zhang et al. 2014), CUGB P1202 lacks fusion of compound elements – a state which is considered to be indicative of subadult ontogenetic stage (e.g., Chiappe et al. 2007). Unlike these other subadult bohaiornithids,

CUGB P1202 retains some periosteal pitting that would be characteristic of a markedly juvenile animal, although the large, well-developed sternum and plumage indicates that

CUGB P1202 was volant and sexually mature (Chiappe et al. 2007).

With a mixture of juvenile and subadult skeletal characteristics, CUGB P1202 preserves feather morphologies indicative of sexual maturity. The new specimen possesses well-developed pennaceous remiges, which have been proposed to have developed long before skeletal maturity in juvenile enantiornithines (Chiappe et al. 2007; de Souza Carvalho et al. 2015). The tail plumage is generally similar to that of other bohaiornithids and Eoenantiornis in that short rectricial feathers appear dominantly plumaceous rather than pennaceous. However, CUGB P1202 and the Bohaiornis holotype (Hu et al. 2011) are the only known bohaiornithids in which paired and elongate, rachis-dominated rectrices are preserved, although neither preserves the distal tips of these feathers. In other basal birds, this condition is variable. and the primitive enantiornithine Protopteryx similarly preserve paired rectrices with thick rachises (Chiappe et al. 1999; Zhang and Zhou 2000, Clarke et al., 2006) while some pengornithids preserve elongate, fully pennaceous rectrices (Hu et al. 2014, 2015; Wang et al. 2014b) or are proposed to exhibit fan-shaped morphologies (O’Connor et al. 2015).

64 The rachis-dominated morphology has been interpreted as “ribbon-like” or “streamer- like” by some authors (e.g. Chiappe et al. 1999; Li et al. 2012) while others suggest that since these feathers seem to undergo little post-burial deformation, they were more likely stiff structures (Wang et al. 2014b). The elongate tail feathers preserved in nearly three- dimensions in Cratoavis support a rigid morphology (de Souza Carvalho et al. 2015). In extant birds, elongate tail feathers are often used as sexual ornaments. For example, elongate tail feathers in male widowbirds and -tufted malachite sunbirds have been shown to be more favorable to females than shorter rectrices (Andersson 1982, 1992;

Evans and Hatchwell 1992). The highly-modified rectrices preserved in the new specimen and in other extinct maniraptoran dinosaurs may have similarly functioned in intraspecific communications, specifically ornamentation (Clarke et al. 2006; Li et al.

2012; Wang et al. 2014b; de Souza Carvalho et al. 2015). The presence of elongate rectrices has been proposed to be potentially limited to males in some basal birds. has been hypothesized for Confuciusornis due to the presence of paired rectrices on some specimens and lack of these rectrices on others (Hou et al. 1999;

Chiappe et al. 1999, 2008) and by the proposed presence of medullary bone in a female

Confuciusornis that lacks elongate rectrices and its absence in purported males with these feathers (Chinsamy et al. 2013). Additional sampling is necessary, but it is possible that the presence of similar paired streamers in CUGB P1202 and Bohaiornis is associated with sexual dimorphism in bohaiornithids.

The morphology of melanosomes preserved in the feathers of CUGB P1202 are consistent with a role for color in sexual communication in enantiornithines. The melanosomes are highly elongate and there is a slight degree of parallel organization of

65 the melanosomes similar to the weakly-iridescent melanosomes of some extant bird feathers (Figure 4.6; Li et al. 2012). In extant feathers, is produced by the coherent scattering of light through nanostructures of differing refractive indices, specifically layers of β-keratin, melanin, and air. A layer of elongate melanosomes below these nanostructural layers serves to absorb incident light and saturate the color that is produced (Prum 2006). Highly elongate melanosomes associated with iridescence have previously been reported for fossil maniraptoran dinosaur feathers (Vinther et al. 2010; Li et al. 2012), although CUGB P1202 is the first enantiornithine reported to preserve iridescent melanosome morphologies.

Like previously-reported fossil iridescent feathers (Vinther et al. 2010; Li et al.

2012), the specific of the iridescent feathers preserved in the new specimen cannot be determined due to the lack of evidence of preserved keratin. In extant birds, iridescent plumage serves a primary function in intraspecific communications, namely sexual selection and display (Doucet and Meadows 2009), and may serve as an honest signal for the condition of a potential mate (McGraw et al. 2002). There are, however, other potential functions of iridescence that have yet to be thoroughly characterized, such as startling potential prey or predator animals, other interspecific communications, , or mechanical strengthening of the feather structure (Doucet and Meadows 2009; Maia et al. 2011). These other functions therefore cannot be ruled out for the plumage of CUGB

P1202, although coupled with ornamental plumage a role for sexual signaling seems likely.

The combination of immature skeletal characteristics and ornaments in CUGB

P1202 is indicative of the importance of sexual selection in basal enantiornithines. In

66 extant avian taxa, skeletal maturity typically occurs long before sexual maturity and males sometimes delay the production of sexual ornaments (e.g. Long-tailed

Chiroxiphia linearis; Arévelo and Heeb 2005). However, bone of non-avian maniraptoran dinosaurs and the presence of ovarian tissues in skeletally-immature fossil birds, including some enantiornithines and basal birds such as and

Confuciusornis, suggest that like extant crocodiles these animals reached sexual maturity long before skeletal maturity (Erickson et al. 2007; Chinsamy et al. 2013; Zheng et al.

2013; O’Connor et al. 2014). CUGB P1202 retains immature skeletal characteristics, including periosteal pitting of the long bones and lack of fusion of compound bones, while also preserving likely sexual ornaments, such as the elongate paired rectrices and iridescent plumage. CUGB P1202 had therefore reached sexual maturity before full developmental maturity.

67 CHAPTER 5

DIVERSIFICATION OF MELANOSOME SHAPE IS LINKED TO PHYSIOLOGICAL

CHANGES BETWEEN VERTEBRATE TAXA

Introduction

Modern vertebrates possess a diverse array of colors, many of which are either produced or aided in production by the ubiquitous pigment melanin. Melanins can be separated into two primary chemical varieties: eumelanin that produces black and brown and pheomelanin that produces rusty (rufous) red and some buff yellow colors (Liu et al.

2014). In structurally-colored feathers, melanin absorbs incoherently-scattered white light to saturate colors produced by the interaction of light between materials of differing refractive indices (e.g. keratin and air spaces within the feather structure; Prum 2006;

Shawkey and Hill 2006). Eumelanin also serves protective functions, such as antimicrobial defense (Chapter 2; Kuo and Alexander 1967; Nosanchuk and Casadevall

2006; Gunderson et al. 2008), material strengthening (Bonser 1995), and photoprotection

(Colombo et al. 2011), as well as thermoregulation (Watt, 1969) and chelation of metals

(Hong and Simon 2007).

68 Animal melanin is housed in organelles called melanosomes in skin, hair, and feathers, and the remains of both melanosomes and melanin have been characterized in the fossil record. Melanosomes have been identified in exceptionally preserved integument, including the skin of and mosasaurs (Lindgren et al., 2014), turtles, and lizards, pycnofibers, protofeathers from non-pennaraptoran dinosaurs (Li et al., 2014), mammal hairs (Colleary et al. 2015), as well as numerous examples of feathers from birds (Chapter 4; Vinther et al. 2008, 2010; Clarke et al. 2010;

Zheng et al. 2017) and other pennaraptoran dinosaurs (Zhang et al., 2010; Li et al., 2010;

2012). The correlation between melanosome morphology and color in mammals and pennaraptoran dinosaurs has enabled reconstruction of original melanin-based colors in fossil taxa (Li et al. 2010; 2012; 2014; Carney et al. 2012; Colleary et al. 2015).

However, low morphological diversity of melanosomes in fossil and extant lepidosaurs, non-pennaraptoran , and testudines prevent color reconstructions for fossilized representatives of these taxa (Li et al. 2014).

Three hypotheses have been proposed for the origin of the convergent increase in melanosome morphology in bird and mammal integument. Li et al. (2014) suggested that the increase in melanosome morphologies may have been caused by convergent shifts in the melanocortin system that pleiotropically links many aspects of vertebrate physiology, including melanization, , and immunity (melanocortin hypothesis). Shifts in this system that produce, for example, higher metabolic rates, may have pleiotropically led to an increase in melanosome diversity. Support for this hypothesis was found in a more recent study of extant avian melanosome diversity, which recovered convergently lower melanosome diversity in large-bodied paleognath birds (Eliason et al. 2016). By

69 contrast, Vinther (2015), proposed that the convergent shifts were more likely driven by the loss in mammals and feathered dinosaurs of non-melanic (e.g. xanthophores and iridophores) – pigment-producing cells that provide vibrant colors in many basal amniotes, amphibians, and fish. He proposed that the diversity of melanosomes in birds and mammals was driven by the selective advantage of more diverse melanic colors after the loss of chromatophores (color hypothesis). A final hypothesis is that the convergent diversification in melanosome morphology was due to the evolution of epidermal extensions (i.e. hairs and feathers), perhaps due to developmental constraints, since diverse morphologies have only been reported in hairs and feathers (integument hypothesis; Li et al. 2014).

Our aim here is to shed light on these hypotheses by expanding studies of melanosome morphology in integument across Vertebrata. Most studies of melanosomes preserved in fossil vertebrate integument have been limited to amniotes (however, see

Colleary et al. 2015; Gabbott et al. 2016; McNamara et al. 2016) and extant sampling in birds and mammals has only included epidermal extensions. Sampling further down the vertebrate is needed to determine if anamniotes share patterns of melanosome diversity with basal or derived amniotes. We therefore examine melanosomes preserved in the compressed remains of the skin of Middle- salamander and Early Cretaceous lamprey fossils from China. We compare the morphologies of these fossil melanosomes to melanosomes from the skin of a suite of modern non-amniote vertebrates and a database of amniote melanosomes (Li et al. 2014).

A lack of morphological diversity in ectothermic anamniotes, especially those without a significant non-melanic -based color palette (e.g. lampreys; Kimura et al.

70 2014), would support the melanocortin hypothesis (Li et al. 2014). By contrast, greater melanosome diversity in anamniotes with few non-melanic chromatophores would support the color hypothesis. Additionally, we characterize melanosome morphology in darkly-colored mammal and bird skin to test if they retain the limited ovoid morphologies of basal amniote skin, which would support the integument hypothesis. We identify broad patterns in melanosome evolution across vertebrates as well as support for a physiological driver of the convergent increases in melanosome diversity in hair and feathers.

Materials and Methods

Sample Preparation

We embedded skin samples from modern amphibians, fish (including both

Osteichthyes and Chondrichthyes), agnathans, birds, and mammals in resin using described techniques (Appendix D; Shawkey et al. 2003). We placed the samples in 200 proof ethyl alcohol for 24 hours, then in a succession of 15, 50, 70, and 100% Epon solutions with 24 hours on a bench-top shaker between each step. Samples were then placed in molds and allowed to cure for 16 hours at 60 ̊C. Once hardened, the embedded samples were cut to expose internal structures using a Leica EM Trim2 high speed milling system, then cut into 3500 nm tangential thin sections using a diamond knife on a

Leica UC-6 ultramicrotome.

We collected samples from the part (IMS1) and counterpart (IMS2) of a neotenic salamander and a frog (IMF) from the Middle Jurassic Jiulongshan Formation, Daohugou

71 Locality, Inner Mongolia, China, and from two salamanders (LSJA, LSJB) and one juvenile salamander (LSJUS) from the Middle-Late Jurassic ,

Dashishan Locality, Liaoning Province, China (Figures A.11, A.12). Additionally, we sampled two undescribed lampreys (CUGB P1404 and CUGB P1405) from the Lower

Cretacious Yixian Formation, exposed at Fenging, Hebei Province, China. All fossils are reposited at the China University of Geosciences and Peking University and were not exposed to glues or other chemicals prior to sampling. Amphibian fossils consist of nearly-complete articulated with associated skin preservation. IMS1 and IMS2 consist of a nearly complete skeleton and excellent preservation of soft tissues, including skin around the torso, tail, and proximal limbs, and gill structures (Figure A.11). IMF lacks preservation of the skull and the skin is poorly-preserved around the torso and hindlimbs (Figure A.13). Preservation of the skin in LSJA and LSJB is poor and limited to the torso and tail and the torso, respectively. Integument is preserved in the neck, torso, and tail regions of LSJUS (Figure A.12). CUGB P1404 is nearly complete and exhibits well-preserved integument, including the caudal and gill pouches on the right latero- ventral side. CUGB P1405 exhibits poorer preservation of the skin and lacks preservation of much of the head and tail regions. Samples were removed from the buccal opening and caudal fin of CUGB P1404 and the dorsal surface of CUGB P1405.

Sampling of the fossil integument was conducted in a similar manner to Li et al.

(2010). We used a small probe to remove two 1 mm x 1 mm samples from each fossil and attached the samples integument-side-up to carbon tape. One sample from each sampling locality on all fossils was used for scanning electron microscopy (SEM) and the

72 other for chemical analysis. We removed one matrix sample from each fossil proximate to the preserved integument as negative controls.

Raman Spectroscopy

To test the melanosome affinity of preserved microbodies in the fossils, we used one sample collected from each sampling locality for chemical analysis. We tested these samples using Raman spectroscopy, which has previously been employed to detect fossil melanin with some caveats (Chapter 4). Both fossil integument and matrix samples were placed onto a glass slide and analyzed with a 532 nm () excitation laser through a

50x objective on a LabRAM High Resolution Raman microscope (Horiba Scientific) using an extended wavenumber range of 300-2500 cm-1, a grating of 1200 lines per mm, a 100 µm slit aperture, and a 400 µm pinhole. The samples were irradiated for ten minutes using a 10% filter before analysis to reduce fluorescent signals (Golcuk et al.

2006). We then collected spectra from at least two different locations on each sample using 10%, 25%, and 50% filters sequentially to test repeatability of the spectra. We used a five second exposure time x one acquisition for each collected spectrum. The Raman spectrometer was calibrated using a wafer of pure silicon.

For comparison, we cut 2 mm cross sections from the skin of the common mudpuppy (Necturus) and the sea lamprey (Petromyzon), which were then dried for 24 hours in ethanol. We aimed the Raman laser at melanic regions using the 50x objective and a 1% filter to reduce the chance of burning the samples. For both fossil and extant samples, if a sample was burned, the resulting spectra were discarded and the sampling location was moved to an unburned area. We compared all samples to spectra previously

73 reported for melanin extracted from modern bird feathers, carbon control samples, and bacteria (Chapter 4). We normalized and graphed the spectra in Origin Pro (OriginLab) and used IgorPro (Wavemetrics) to fit the peaks with a Gaussian distribution and a linear baseline.

Scanning Electron Microscopy and Morphological Analysis

Prior to imaging, fossil and extant samples were mounted on stubs with carbon tape and coated with gold-palladium for three minutes in a Polaron E5000 Sputter-Coater

(Quorum Technologies). We viewed the samples using a JSM-7401 Field Emission

Scanning Electron Microscope (JEOL Solutions for Innovation) operated at 7 kV with a working distance of 7.0 mm. Samples were imaged at 3,000 to 5,500x magnification.

We quantified melanosome morphology by measuring the shape and size of melanosomes from the SEM images using ImageJ (available for download at http://rsbweb.nih.gov/ij/) in a manner similar to previous work (Li et al. 2010; 2012;

Clarke et al. 2010). We measured the linear long (length) and short (width) axes of complete melanosomes. Length and width values were used to calculate the aspect ratio

(length:width) of each measured melanosome.

We used a two-sample Student’s T-test assuming unequal variance to compare length, width, and aspect ratio of melanosomes preserved in amphibian fossils to extant amphibian melanosomes. These analyses were performed using the original melanosome measurements taken from the fossil samples as well as length and width measurements for melanosomes preserved in three-dimensions increased 10% and 20% to account for the proposed diagenetic shrinkage in size (McNamara et al 2013; Colleary et al. 2015).

74 These changes do not affect aspect ratio. Melanosome molds may not have undergone significant taphonomic alteration and thus these measurements were not altered (Vinther

2015). Length, width, and aspect ratio of melanosomes from modern non-amniote vertebrates were compared to data previously collected for extant amniotes (Li et al.

2014) – to which we added sixteen additional lepidosaur species – by analysis of variance

(ANOVA; aov function) and Tukey’s honest significant difference (HSD; TukeyHSD function) in the R programming language (R Development Core Team). We also used a two-sample Student’s T-test assuming unequal variance to compare melanosome morphology between modern mammal and bird skin samples and those from other integument samples, including black and brown hairs and feathers, respectively, and anamniote, lepidosaur, testudine, and crocodylomorph skin (Li et al. 2014).

Results

The Jiulongshan and Tiaojishan fossils are exceptionally-well preserved, including articulated skeletons and skin. SEM of the integument preserved in lamprey and amphibian fossils revealed small, oblate microbodies preserved in all but two of the fossils sampled (Figures 5.1, A.11-A.13). These microbodies were preserved in three dimensions in IMS1 and its counterpart IMS2. Microbodies were preserved as external molds in LSJA, LSJB, LSJUS, and CUGB P1405. Microbodies preserved in both the fossil agnathan and amphibians are small (average length less than 1 µm) and oblate in shape (aspect ratio less than 2; Table 5.1).

75 Table 5.1. Measurements of melanosome morphology in extant and fossil (indicated) skin samples. Average measurements are plus or minus standard error and range measurements include minimum and maximum values.

Taxon Measurement Averages (nm) Range (nm) Length 542.3 ± 7.8 259.6 - 894.6 Agnatha Aspect Ratio 1.3 ± 0.01 1.0 - 2.2 Length 540.8 ± 10.1 358.5 - 768.0 Fossil Agnatha Aspect Ratio 1.5 ± 0.0 1.0 - 2.7 Length 530.4 ± 3.0 275.6 - 887.0 Fish Aspect Ratio 1.3 ± 0.0 1.0 - 2.7 Length 591.0 ± 5.8 259.9 - 1600.4 Amphibian Aspect Ratio 1.3 ± 0.0 1.0 - 2.3 Fossil Length 561.2 ± 6.8 276.7 - 1066.7 Amphibian Aspect Ratio 1.6 ± 0.0 1.1 - 2.8 Length 601.042 ± 13.7 431.1 - 872.9 Mammal Aspect Ratio 2.0 ± 0.06 1.4 - 3.0 Length 654.7 ± 17.6 314.3 - 1107.3 Bird Aspect Ratio 2.3 ± 0.07 1.1 - 4.3

76

Figure 5.1. SEM images of melanosomes from the skin of fossil and extant agnathans and amphibians. A, black Petromyzon skin; B, skin preserved in CUGB P14-5; C, black Necturus skin; D, skin preserved in

IMS. Scale bars = 1 µm.

The presence of microbodies in the fossil samples correlated with potential eumelanin Raman signatures. Modern eumelanin samples, such as melanin extracted from modern feathers (Chapter 4), have a short Raman peak at approximately 1370 cm-1 and a taller peak at approximately 1580 cm-1. Raman spectra from modern Necturus and

Petromyzon samples showed two similar peaks: a shorter peak with wavenumbers 1382 and 1360 cm-1, respectively, and a taller peak at 1578 and 1583 cm-1, respectively.

Similar peaks were also generated from six of the fossil integument samples, including all salamander fossils and CUGB P1405 (Figure 5.2; Table A.7). Fossil

77

Figure 5.2. Raman spectra generated from exceptionally preserved skin from fossil amphibian (top) and lamprey (bottom) samples with spectra from modern Necturus and Petromyzon melanin (red) for comparison. Insets are SEM images of melanosomes preserved in IMS (top) and CUGB P1405 (bottom). a.u. = arbitrary units. Scale bars = 1 µm. amphibian spectra included a shorter peak between 1366 and 1381 cm-1 and a taller peak between 1575 and 1583 cm-1. The shorter peak occurred at 1361 cm-1 and the taller peak occurred at 1570 cm-1 in spectra from CUGB P1405. None of the matrix samples nor the integument samples from IMF and CUGB P1404 had Raman peaks (Figures A.13, A.14).

We surveyed melanosome morphology of modern anamniotes, including representatives from Agnatha (Petromyzon and Myxine), Chondrichthyes and

Osteichthyes (collectively referred to as “fish”), and Amphibia. Representatives from

78 each taxon had sub-spherical melanosomes with length measurements less than 1 µm and aspect ratios less than 2 (Figure 5.1; Table A.8). Skin color in agnathans, fish, and amphibians is not correlated with melanosome morphology (Figure A.15), like non- mammalian, non-avian amniote skin (Li et al. 2014). Fossil amphibian microbodies were similar in size (length p-value = 0.93, width p-value = 0.15) and shape (aspect ratio p- value = 0.05) to melanosomes from modern amphibian skin. Increasing the length and width measurements for IMS by 10% and 20% did not significantly affect T-test results

(length p-value: 10% = 0.95, 20% = 0.83; width p-value: 10% = 0.18, 20% = 0.23, aspect ratio remained unchanged). Due to our small sample size, fossil microbodies and modern agnathan melanosomes could not be statistically analyzed, although preserved microbodies in CUGB P1405 do not fall outside of the size and shape range of melanosomes in Petromyzon or Myxine (Table A.8, A.9).

Additionally, we measured melanosomes from darkly-colored skin from modern mammals and birds to test if elongate melanosomes in these taxa are limited to hairs and feathers, respectively. Melanosome morphology in bird and mammal skin was significantly longer and thinner than those of modern anamniotes, lepidosaurs, testudines, and crocodylomorphs (length and aspect ratio p-value < 0.05). Melanosomes from avian and mammalian skin samples were small, but pill-shaped (Figure 5.3; Table A.7). Avian skin melanosomes were also smaller and wider than melanosomes in black feathers

(average length = 1002.3 nm and average aspect ratio = 3.7 for black feathers; p-value <

0.05), but longer and thinner than those in brown feathers (average length = 654.7 nm and

497.5 nm, respectively; average aspect ratio = 2.3 and 1.8, respectively; p-value < 0.05;

Li et al. 2012). Mammal skin melanosomes were similar in morphology to brown hair

79 samples, although slightly longer (average length = 601.0 nm and 697.4 nm, respectively; average aspect ratio = 2.0 and 1.9, respectively; p-value = 0.06 for length and 0.14 for aspect ratio; Figure 5.3; Table 5.1; Li et al. 2014).

Figure 5.3. SEM images of small, pill-shaped melanosomes from the black-colored skin of A, a bat wing and B, the eye ring of a rose-breasted grosbeak. Scale bars = 1 µm.

Discussion

Amphibian and agnathan fossils from Jurassic and Cretaceous beds in northern

China retain exceptionally preserved skin with oblate microbodies that are consistent with modern melanosomes both in morphology and Raman signature (see Appendix D).

These microbodies are consistently small (less than 1 µm in length) and sub-round in shape (aspect ratio less than 2), similar to melanosomes in modern anamniote skin and consistent with melanosome measurements reported previously for non-pennaraptoran and non-avian amniotes (Li et al. 2014). Our findings are consistent with previous measurements of melanosomes from the patterned skin of the Carboniferous lamprey

Mayomyzon (Gabbott et al. 2016) and Cenozoic frogs (Colleary et al. 2015).

80 Melanosome morphology varies greatly in pennaraptoran dinosaur feathers and mammal hairs (Figure 5.4; Li et al. 2014). Modern bird feathers and mammal hairs produce melanosome morphologies ranging from small, sub-spherical melanosomes that produce red-brown and buff-yellow colors to elongate, high-aspect ratio melanosomes that produce black and brown colors (Li et al. 2014). Birds further produce a fantastic array of vibrant structural colors by incorporating extremely-high aspect ratio melanosomes arranged in layers or hollow rod-shaped, flattened, plate-like, or hollow platelet-type melanosomes into the feather structure (Prum 2006; Eliason et al. 2013;

Maia et al. 2013). In contrast, melanosomes in both fossil and extant anamniotes (this study) and non-pennaraptoran, non-mammalian amniotes are small and sub-spherical in morphology, with little variation (Figures 5.1, 5.4), including those from extinct such as non-pennaraptoran dinosaurs (Li et al. 2014), ichthyosaurs, and mosasaurs

(Figure 5.5; Lindgren et al. 2014). Similarly, small and oblate melanosomes are present in samples removed from modern squid skin and otherwise pigmented squamate skin

(Figure A.16). Thus, the small, sub-spherical morphology was likely the basal melanosome morphology for integument with at least two diversification events in a mammalian ancestor and a pennaraptoran ancestor (Li et al. 2014).

81

Figure 5.4. Length of the long axis and aspect ratio of melanosomes in the skin, feathers, and hair of extant vertebrates. Black, brown, and gray colors are representative of the color of the integument. Agnathan skin was not analyzed by color due to small sample size. Iridescent and penguin-type avian feathers are represented by purple and blue, respectively. Similar boxplots (two-sided Tukey HSD, p < 0.05) are represented by the same letter (u, v, w, x, y, z). Data for feathers, hairs, lepidosaurs, testudines, and crocodylomorphs is adapted from Li et al. (2014) with additional input from the current study.

82

Figure 5.5. Diversity of melanosome morphology from modern and fossil integument across Vertebrata.

All fossil bird, fossil mammal, and modern melanosomes are colored based on integument color and are representative of the average melanosome morphology for that color. For fossils, these colors are predicted based on melanosome size and shape. Penguin-type melanosomes are characterized by blue shapes and melanosomes in iridescent feathers by purple. Green color represents fossil melanosomes for which original color has not been predicted and their shape is based on the shortest and longest melanosomes described for that taxon. Cartoon melanosomes are to scale. Melanosome measurements are based on

Tables A.8-A.10 and the references therein, Prum (2006), and Li et al. (2012, 2014). Asterisk indicates loss of the ability to produce chromatophores

83 The melanocortin hypothesis proposes that convergent changes in melanosome size and shape in mammals and pennaraptoran dinosaurs may be related to changes in the melanocortin system of these animals (Li et al., 2014; Eliason et al. 2016). The melanocortin system is a neuronal pathway that consists of five posttranslational bioactive peptides, called melanocortins. Two of these melanocortins, ACTH and αMSH induce the expression of a -specific structural , PMEL17 (also known as gp100 or ; Kwon 1993), which is responsible for the formation of a fibrillar matrix within melanosomes, which in turn leads to their elongation (Kushimoto et al.

2001). The absence or downregulation of PMEL17 during melanogenesis leads to round melanosomes shapes (Furumura et al. 1998). Although PMEL 17 transcripts are expressed in several tissues in vertebrates (Brouwenstijn et al., 1997), the expression of the PMEL17 protein in epidermal melanocytes has been found exclusively in mammals

(Berson et al. 2001; Raposo et al. 2001) and birds (Kerje et al. 2004) coinciding with the presence of elongate melanosomes in these clades. Identifying the steps in the translational regulation of the PMEL17 protein that has led to the diversification of melanosomes shape in birds and mammals remains a fertile area for future investigation.

When melanocortins bind to their receptors, they regulate basic functions such as sexual activity, aggression, energy homeostasis, inflammatory response, and melanogenesis (see for review Roulin et al., 2011). Each of these phenotypic traits may be pleiotropically linked, causing the covariation of melanin-based coloration and some of the physiological and behavioral traits that are regulated by the melanocortin system, such as metabolism (Ducrest et al., 2008; Emaresi et al., 2013). Indeed, flightless large- bodied paleognaths, which have a lower metabolic rate than other birds, exhibit relatively

84 lower melanosome diversity than neognaths, but greater diversity than outgroup taxa

(Eliason et al. 2016). Lower melanosome diversity is limited to vertebrates with a lower metabolic rate, including the anamniotes studied here.

Alternatively, the integument and color hypotheses propose, respectively, that increased melanosome diversity in birds and mammals may have been driven by the evolution of novel integument types (i.e. hair and feathers) and/or the loss of non-melanic chromatophores in these taxa (Oliphant et al. 1992, Vinther 2015). Chromatophores are pigment-containing cells that produce the vibrant and diverse colors of fish, amphibians, lepidosaurs, testudines, and crocodylomorphs (Parker, 1930), but are not present in the integument of birds and mammals. Thus, it is argued that the evolution of diverse melanosomes may have enabled birds and mammals to produce diverse colors without chromatophores. These hypotheses are not supported. First, birds maintain the capacity to produce every type of chromatophore in the iris (Oliphant et al. 1992), so they have not completely lost the ability to produce them. It may be that they cannot be deposited in feathers, however. Second, selection for brighter colors via organization of high-aspect ratio melanosomes may have led in part to diversification of melanosome morphology in feathers, but mammals also have high melanosome shape variation and retain dull- colored hairs (Li et al., 2014). Third, only two chromatophore types have been reported for lampreys – melanophores and iridophores (Wright and McBurney 1992) – and they do not possess the genes to produce other chromatophores (Kimura et al. 2014), but they still have low diversity of melanosome morphologies. Finally, in contrast to the proposal that high melanosome diversity is limited to hairs and feathers, possibly because of constraints during development, we have found that darkly-colored mammal and bird

85 skin possess pill-shaped melanosomes (Figure 5.3) rather than the sub-spherical morphologies of ectothermic animal skin. Additionally, both elongate and sub-spherical melanosome morphologies are produced in the retinal pigment epithelium in the eyes of most vertebrates, including basal vertebrates such as lampreys and other fish (e.g.

Burgoyne et al. 2015), although elongate morphologies are absent in the ocelli of amphioxus (Figure A.17). Thus, diversity of melanosome shapes does not seem to be associated with either the lack of chromatophores or presence of feathers or hairs.

The low morphological disparity of melanosomes in extinct and extant anamniotes is shared with extant ectothermic amniotes (Figure 5.5). This is contrary to the high morphological diversity of melanosomes in pennaraptoran dinosaurs and mammals. These data support the hypothesized convergent shift in the melanocortin system of birds and mammals proposed by Li et al. (2014) and suggest that small, sub- spherical melanosomes morphologies may have been the primitive state for animal integument. Future research should focus on the genetic basis of these shifts and expand the study of melanin evolution to the entire tree of life.

86 CHAPTER 6

WHAT WAS THE ORIGINAL FUNCTION OF MELANIN? A REVIEW OF THE

MODERN FUNCTIONS OF MELANINS AND EVIDENCE FROM THE FOSSIL

RECORD

Introduction

All organisms, no matter their size or complexity, must possess tools that help them combat numerous pressures in to eat, survive, and pass their genes on to future generations. Environmental toxicity, parasitism, disease, radiation, and numerous other chemical or environmental pressures can all be detrimental to survival for all forms of life. Organisms have therefore evolved multiple mechanisms in an attempt to ensure their survival and ability to reproduce.

Among the most ubiquitous and versatile of these mechanisms for survival are melanins – a generic class of heterogeneous and complex biopolymers that occur throughout the natural world. Melanins occur four main chemical variants, including allomelanin, eumelanin, pheomelanin, and neuromelanin, that are produced endogenously by organisms ranging from single-celled bacteria and protists to fungi, plants, and

87 animals. They serve a multitude of functions for these organisms, most of which involve melanin’s reactive properties that allow them to bind to toxic metals and free radicals that may cause DNA or other cellular damage. Additionally, some melanins also function in thermoregulation and coloration, producing both neutral colors such as black or brown that function in camouflage and vibrant structural colors that function in sexual display.

However, the functions of melanins are not mutually exclusive and they are able to serve multiple roles at once for the organism in which they are produced.

Given their ubiquity, it is likely that melanins have played essential roles in organisms throughout time. The complexity of the polymeric structure of melanins, especially eumelanin, makes them difficult to break down due to their insolubility.

Therefore, their preservation potential in the fossil record should be relatively high.

Indeed, multiple taphonomic experiments have demonstrated that eumelanin in animals and the organelles in which it is housed, called melanosomes, are able to survive the high temperatures and pressures associated with the fossilization process (Glass et al. 2013;

McNamara et al. 2014; Colleary et al. 2015). Eumelanin and pheomelanin have been chemically characterized in the fossil record using multiple techniques, such as time-of- flight secondary ion mass spectrometry (ToF-SIMS; Lindgren et al. 2014; 2015a, 2015b;

2017; Clements et al. 2016; Brown et al. 2017), Raman spectroscopy (Chapter 4), trace metal mapping (Wogelius et al. 2011), and high-performance liquid chromatography

(HPLC; Tanaka et al. 2014). Additionally, melanosomes have been found using scanning electron microscopy (SEM) in many exceptionally preserved organisms, leading to inferences about their coloration and behaviors (Chapter 4, 5; Clarke et al. 2010; Li et al.

88 2010, 2012; Hu et al. 2018), physiology (Li et al. 2014; Vinther 2015), and phylogenetic relationships (Clements et al. 2016).

Currently, the earliest known occurrence of melanin and melanosomes in the fossil record is in the exceptionally preserved eyes of the 307 million--old

Tullimonstrum (Clements et al. 2016), although melanin probably evolved long before

Tullimonstrum swam the Upper Carboniferous oceans. The authors hypothesized that the alternation of spherical and rod-shaped eumelanin-filled melanosomes in the eyes of

Tullimonstrum were probably used in a similar manner to modern eye melanosomes – to screen the photoreceptors of the eyes from aberrant light and aid in the animal’s visual acuity. However, probably not the first occurrence of melanin nor its original function, both of which are currently unknown. While the presence of melanin in older fossils has yet to be determined, it may be possible to elucidate the original function of melanin and when it may have evolved by looking at its functions in modern organisms.

Here we review the roles melanins serve for modern organisms. We begin by differentiating between the different chemical variants of melanin, which is important for understanding their functions and evolution. We then review the three major functions of melanins in modern organisms in descending order of when they may have evolved, beginning with thermoregulation, then coloration, and finally protection. Lastly, we examine the known evidence for each of these functions in the fossil record and try to elucidate when melanin may have originally evolved, what its original function was, and when the modern functions diverged.

The Types of Melanin

89

To effectively assess the functions of melanins in modern organisms and elucidate the initial function of the first melanin, we must first describe what we mean by the term

“melanin.” Historically, “melanin” has been used as a blanket term to describe dark pigments, but it includes a ubiquitous group of complex biopolymers produced that are generally insoluble by conventional means (Riley 1997; Solano 2014). Melanins consist of four main chemical variants, including eumelanin, pheomelanin, neuromelanin, and allomelanins. Eumelanin and pheomelanin are produced via a tyrosine-derived pathway driven by the enzyme tyrosinase. Put simply, tyrosinase converts tyrosine to dopaquinone via L-DOPA, then dopaquinone produces either pheomelanin or eumelanin, depending on the presence of certain amino acids. Pheomelanin production depends on the presence of cysteine and consists of a benzothiazine backbone. Eumelanin is produced when cysteine is absent and consists of a backbone of 5,6-dihydroxyindole-2-carboxylic acid (DHICA) and 5,6-dihydroxindole (DHI). Both eumelanin and pheomelanin are often referred to as

“animal melanins” and pheomelanin production does indeed appear to be limited to animals, although eumelanin occurs in nearly all major taxa. In animals, these two types of melanin are produced by and co-occur in melanosomes (Simon et al. 2008), which are membrane-bound organelles produced by cells called melanocytes or melanophores. It is important to distinguish melanosomes from melanin granules, which occur in ink are not membrane-bound.

Neuromelanin is produced by specialized neurons in the substantia nigra and, in lesser quantities, the locus coeruleus of the brain, giving these regions their dark color

(Zecca et al. 1996; Tribl et al. 2009). The production pathway, chemical structure, and

90 hormonal and genetic controls are unknown for neuromelanin. Neuromelanin melanosomes are composed of an outer layer of eumelanin generated by the oxidation of dopamine and a core of pheomelanin produced via cysteine’s integration with dopamine, but they lack the proposed stacked arrangement of the eumelanin structure. Metal ions, lipids, and proteins are also contained within the neuromelanin melanosome structure

(reviewed in Zucca et al. 2017). substantia nigra contain the highest concentrations of neuromelanin and although it does occur in smaller quantities in the brains of some non-human , neuromelanin production in other vertebrates is contentious (reviewed in Fedorow et al. 2005). It is possible that its presence in other organisms has yet to be assessed due to the focus of most neuromelanin research on its link to Parkinson’s disease.

Finally, allomelanins, or “other melanins,” are usually dark brown or black melanins unlike eumelanin and pheomelanin most are derived from the oxidation of polyphenols rather than tyrosine and their chemical structure lacks nitrogen (Fedorow et al. 2005). Allomelanins include a variety of lesser-known melanins, such as those derived from tetrahydroxynaphthalene (THN) and dihydroxynaphthalene (DHN) in fungi, catechol-derived melanins in plants, and hexahydroxyperylenequinone (HPQ) melanins in bacteria. The most common allomelanin is pyomelanin, which is predominantly produced by bacteria and fungi. Pyomelanin is generated via a tyrosine or phenylalanine- derived pathway that leads to the overproduction of homogentisic acid (HGA), which is externally polymerized then absorbed by the or (Turick et al.

2010). Pyomelanin (alkapton) can be excreted in the urine of humans with certain metabolic disorders, such as alkaptonuria caused by an accumulation of HGA in the

91 body’s tissues due to loss-of-function in the homogentisate 1,2 dioxygenase gene (Fernández-Cañón, 1996; Beltrán-Valero et al. 1999).

Melanins are Thermoregulators

A defining characteristic of melanins is that they broadly absorb across the ultraviolet (UV) and visible light spectrum, which allows them to contribute to thermoregulation and coloration. Melanins function as thermoregulators by absorbing UV radiation and converting that energy to heat via photon-photon conversion (McGinness and Proctor, 1973). Although some microorganisms incorporate melanin as a mechanism for thermoprotection (e.g. Cryptococcus neoformans; Rosas and Casadevall, 1997), a true thermoregulatory function for melanins is limited to animals. Eumelanin is especially effective because it absorbs across a wider range of light, and, as reviewed below, thus produces darker colors than pheomelanin. The thermal melanism hypothesis suggests that the internal temperature of darker individuals will increase more rapidly than lighter individuals, assuming that they are similar in body size (reviewed in Clusella-Trullas et al. 2007). In contrast, lighter colors should confer such advantages to individuals in warmer environments because they would need to spend less time dissipating excess heat.

The function of eumelanin for thermoregulation is most extensively studied in insects. Lepidopterans ( and moths) often serve as model organisms for studies of melanic thermoregulation because they bask in the sunlight to reach an optimum body temperature of between 30-40 ˚C that is necessary for flight (Clusella-Trullas et al. 2007).

92 Watt (1968), for example, noticed that Colias butterflies in cooler temperatures tended to have darker hindwings than those in warmer temperatures. Darker hindwings correlated with increased flight duration in higher elevations compared to lighter conspecifics in the same region. Conversely individuals with lighter wings suffered less from overheating at lower elevations than their darker counterparts (Watt 1968). Darker animals in cooler environments should therefore be more successful at evading predators, finding mates, or protecting their territory since they need to spend less time reaching their thermal optimum than lighter individuals in the same environment. In support of this relationship,

Ellers and Boggs (2004) found that when the hindwings of female Colias butterflies were artificially darkened, maturation rate significantly increased in cooler temperatures relative to naturally yellow-colored animals. This suggests a correlation between the decreased time to reach optimum temperature caused by wing darkening and reproductive success. There can, however, be tradeoffs to increased melanism for thermoregulation in . For instance, Hegna et al. (2013) found that altitude- based melanization for thermoregulation is inversely correlated with the efficacy of aposematic (warning) coloration in the wood tiger moth Parasemia plantaginis.

Thermal melanism may provide similar advantages to ectothermic vertebrates.

Many lizards are able to rapidly change color by dispersing (darken) or concentrating

(lighten) melanosomes within the melanocytes in their skin. While rapid color change is commonly used as a mechanism for camouflage, as discussed below, it may also be advantageous for thermoregulation. Smith et al. (2016a), for instance, found that wild- caught central bearded dragons (Pogona vitticeps) rapidly darkened and exhibited lower light reflectance (i.e. greater absorption) in their dorsal regions when exposed to a cooler

93 temperature (15 ˚C). In contrast, dragons exposed to a warmer temperature (40 ˚C) lightened their dorsal coloration and consequently increased reflectance (i.e. lower absorption). Given that these color changes were limited to the dorsal surfaces of the animals, as reflectance on the ventral side remained static, and the observation of similar behaviors in the field (Smith et al. 2016b), rapid color change here is likely a thermoregulatory response rather than physiological (Smith et al. 2016a). However, the same authors also found that color change in bearded dragons in the field was more commonly used for camouflage against dark or light backgrounds, although its use for thermoregulation was still significant (Smith et al. 2016b). Melanin’s use for thermoregulation may therefore be situational, as the need for melanin-based coloration for camouflage or signaling may sometimes outweigh the benefits of its thermoregulatory function.

Melanin’s thermoregulatory function in endotherms is poorly studied and proposed examples are often secondary to other functions, such as signaling. Walsberg et al. (1978) tested heat load on dark versus lightly-colored pigeon plumage. They found that solar radiation in calm conditions did not penetrate past 25% of darkly colored plumage, whereas it penetrated up to 80% of lightly-colored plumage and may warm the bird. When wind speed was taken into account, however, they found greater heat loads on darkly-colored plumage than lighter plumage (Walsberg et al. 1978; Wolf and Walsberg,

2000). This may explain why desert birds tend to have melanized plumage, but their colors and color patterns are also used for camouflage and protection from ultraviolet radiation (McGraw 2006). Additionally, the temperature of the darkly-melanized ear tufts of bearded vulture Gypaetus barbatus barbatus study skins has been shown to increase

94 after exposure to a heat source. They may therefore warm air near the ear region of the animal in life (Margalida et al. 2008), although other potential functions are unknown and, as the authors admit, heat penetration through these feathers was not measured.

Melanins Contribute to Coloration

Coloration is one of the first things we see when we look at organisms and many colors in the natural world are produced in some way by melanin. While other melanins may contribute to the color of an organism (e.g. DHN melanin confers a grey-green color to the conidia of the fungus Aspergillus fumigatus; Langfelder et al. 2003), pheomelanin and eumelanin are the only melanins known to be utilized specifically for color-related functions. Both of these melanins contribute to a vast expanse of diverse colors that serve a range of functions, such as camouflage and signaling, in nearly all major animal taxa by acting as a pigment or by contributing to the nanostructure of the integument to produce .

Pigment

Pigments generate colors by selectively absorbing certain of light.

Eumelanin produces black and dark brown colors by uniformly absorbing light across all visible and ultraviolet wavelengths (Fox 1976), potentially due to its greater number of indole quinones compared to other melanins. In contrast, pheomelanin has fewer carbonyl groups than eumelanin and incorporates sulfur into its structure via

95 benzothiazoles (Riley 1997). Pheomelanin therefore produces rusty-brown and buff yellow colors by incompletely absorbing visible light.

In vertebrates, eumelanin and pheomelanin are housed together in varying concentrations in melanosomes (Ito and Fujita 1985; McGraw et al. 2005; Simon et al.

2008), and it is the eumelanin-pheomelanin ratio that determines the melanin-based color of the integument (McGraw 2006). Birds and mammals exhibit diverse melanosome morphologies that correlate with the color of their feathers and hairs, respectively.

Elongate, pill-shaped melanosomes with greater concentrations of eumelanin are found in black feathers and hairs while red-brown or buff yellow feathers and hairs that contain higher concentrations of pheomelanin are housed in smaller (less than 1 µm), oblate melanosomes (Li et al. 2010, 2014). Melanosome diversity is greater in feathers because many birds also produce a variety of vibrant, structurally-colored feathers as well (see below). However, vertebrates with lower metabolic rates produce melanosomes with less morphological variation. The skin of fish, amphibians, lepidosaurs, testudines, and crocodylomorphs houses small, oblate melanosomes similar to those in red-brown feathers or hairs no matter the color of the integument (Chapter 5; Li et al. 2014). While melanins in most insects are diffused throughout the cuticle rather than housed in organelles (Shawkey et al. 2009; Stavenga et al. 2012), melanosomes present in other invertebrate integument, such as cephalopod skin (Chapter 5) and cuticles (Hsiung et al. 2017), are similarly small and rounded.

As pigments, pheomelanin and eumelanin serve numerous color-related functions, including effective camouflage (cryptic coloration) in a variety of environments.

Melanin’s function for camouflage can be static, often involving color patterns that help

96 to disrupt the animal’s outline and effectively hide it from potential predators.

Background crypsis is a common example that involves the animal’s integument being colored in the same manner as its environment. Desert mammals, for instance, are commonly sand-colored, which is produced by intermixing red-brown pheomelanic hairs and darker eumelanic hairs. , which involves the darkening of the dorsal side of the animal that is exposed to light and lightening of the less-exposed ventral side, helps to hide the three-dimensionality of an animal from potential predators or prey

(Allen et al. 2012). This method of camouflage is commonly used by insects, , and many vertebrate taxa. Similarly, complex, high contrast color patterning can be used to disrupt an animal’s outline. While birds are able to incorporate a variety of feather pigments, such as garnered from their diet and more specialized pigments (e.g. psittacofulvins (Psittaciformes), porphyrins (Musophagidae,

Strigiformes, Caprimulgidae, and Otididae), and spheniscins (Sphenisciformes)), melanins are the only pigments that can be used to produce within-feather color patterns such as stripes or spots (McGraw 2006). Some birds (e.g. and nightjars) use complex within-feather color patterns, like spots and thin stripes, to disrupt their outline against similarly-colored tree bark, effectively concealing themselves using as well as background matching.

Melanin-based camouflage can also be dynamic, ranging from rapid (on the scale of seconds or minutes) to seasonal color changes to adapt to environmental fluctuations in color. Many fish, amphibians, lepidosaurs, cephalopods, and are able to rapidly change their color to camouflage against a dark or light background. In vertebrates, this color-changing system is regulated by the melanocortin system, which

97 controls hormones responsible for dispersing and concentrating melanin. Melanocyte- stimulating hormones (MSH) disperse melanosomes throughout the dermal melanocytes, effectively making the animal darker in order to camouflage it against a dark background.

To adapt to a light background, melanocyte-concentrating hormone is released to concentrate melanosomes in the cell, making the animal appear lighter (Sugimoto 2002;

Logan et al. 2006; Shiraki et al. 2010). In cephalopods, rapid color change via the movement of melanophores and other chromatophores in the skin is under neuromuscular control rather than hormonal control (Messenger 2001). Other animals, including birds and mammals in which melanosomes are housed in keratinized epidermal extensions (i.e. feathers and hairs, respectively), cannot rapidly change color. Instead, some mammals and birds shed or molt darker or lighter coats or feathers to camouflage against environments that change seasonally.

Melanins as pigments also play important roles for both inter- and intraspecific communication. As discussed below, many of these functions involve brightly-colored patterns produced by nanostructures, other pigments, or a combination of pigments, including pheomelanin and eumelanin, and nanostructures. Melanic colors can enhance the visual effect of these brighter colors. The high-contrast red (a mixture of carotenoids, eumelanin, and to a lesser extent pheomelanin) and yellow (dominantly pheomelanin) epaulets against a black eumelanic feather background of male red-winged blackbirds

(Agelaius phoeniceus), for example, may function as aggressive signals or status symbols

(McGraw et al. 2004).

Structural Coloration

98 Although melanin’s role in coloration is commonly associated with making non- iridescent pigmentary colors, melanin can also play an essential role in the production of a vibrant array of iridescent and non-iridescent structural colors in animals. These structural colors are produced by the interaction of light between materials of differing refractive indices. Eumelanin has a high refractive index, which ranges from 1.7-1.8 depending on the wavelength of light, that is greater than that of air (1.0) or the integument in which melanin is commonly deposited, such as beta keratin (1.54-1.58) in feathers or (1.53-1.59) in scales (Leertouwer et al. 2011; Stavenga et al.

2015). The difference in refractive indices, or refractive index contrast, between these nanostructures causes light to be refracted and coherently scattered at certain wavelengths, thus producing the observed color (Stavenga et al. 2015). This is often achieved by organizing the deposited melanin in a specific arrangement.

The organization and spacing of melanosomes in bird feathers can influence whether its color is matte black, glossy black, or iridescent. Matte black feathers are composed of randomly dispersed melanosomes. In contrast, glossy black feathers have quasi-ordered melanosomes near the feather’s outer cortex while the outer cortex of iridescent feathers is comprised of highly organized melanosome layers (Maia and

Shawkey 2010). Melanosomes can be organized in a variety of ways, such as in a single layer (Maia et al. 2009), multiple layers (Zi et al. 2003), or arranged in a two-dimensional lattice (Eliason and Shawkey 2012), to produce striking displays. Additionally, birds can produce a variety of melanosome shapes that contribute to iridescence, including layers of flattened, plate-like, or extremely elongated melanosomes (Prum 2006). Specialized hollow melanosomes in and hollow platelet-type melanosomes in

99 incorporate air into the center of the melanosome, which further increases the refractive index contrast (Greenewalt et al. 1960; Eliason et al. 2013). Both eumelanin and pheomelanin are useful for the production of iridescent structural colors, giving birds the ability to produce a wide range of iridescent colors (Prum 2006; Xiao et al. 2014). In addition to producing iridescent colors, melanin also plays a crucial role in the production of structurally-generated non-iridescent plumage color in birds. A basal layer of melanosomes deposited beneath feather keratin and air nanostructures absorbs incoherently scattered light, which produces non-iridescent blue in the Steller’s Jay

Cyanocitta stelleri (Shawkey and Hill 2006). Structural color in bird feathers is predominantly used for sexual selection, although its role in other aspects of avian ecology has yet to be thoroughly investigated (Li et al. 2012).

Insects likewise produce melanin-enhanced structural colors, as exemplified in damselflies, tiger , leaf beetles, jewel beetles, and buprestid beetles (Schultz and

Rankin 1985; Vukusic et al. 2004; Noyes et al. 2007; Stavenga et al. 2012). Melanin here is concentrated in layers rather than arrangements of organelles, but it similarly acts as a high refractive index medium capable of refracting light, and its absorbance may also factor into the color production (Stavenga et al. 2012).

Melanins are Protectors

Melanins serve multiple protective functions for the organisms in which they are housed. Their dark color, for example, is useful for deterring predators or shielding organisms from UV damage. Many of melanins’ protective functions are attributed to

100 their stable free radical structure, which allows them to cross-link with proteins to strengthen materials and bind to cytotoxic metals and highly reactive free radicals, including reactive oxygen species.

Melanins are Used for Anti-predator Defense

Although many organisms have some defensive mechanism to ward off predators, the anti-predator defense function for melanin is limited to coleoid cephalopods (i.e. , , and squid). Most coleoids have an internal sac that generates darkly-colored ink composed of a mix of melanin, mucous, and other constituents. In the cuttlefish Sepia officinalis, the ink is composed of granules of pure eumelanin (Glass et al. 2012) released by the rupture of melanosomes produced by the rough and trans-Golgi network of matured ink gland cells (Palumbo 2003). Coleoid ink is used in two main ways to deter predators, depending on how it is released from the animal: as a visual barrier or as a distraction. As a visual barrier, the cephalopod releases eumelanin as a dark cloud or “smoke screen” (sensu Caldwell 2005) that inhibits the visual of the attacker, allowing the cephalopod to swim away. Cephalopods can also release a mixture of melanin granules and , creating a dark blob that acts as a decoy while it escapes (Caldwell 2005).

Melanins Strengthen Materials

Melanins provide resistance to structural damage for the cell walls of fungi, animal integument, and plants. This added strength is likely due to the large size of these polymers and their ability to cross-link with proteins (Riley 1992). In the fungus

101 Cryptococcus neoformans, the incorporation of melanin in the cell wall may strengthen it, protecting the cell wall from structural damage caused by extreme heat and cold (Rosas and Casadevall, 1997). In animals, the melanization of the jaws of the bloodworm

Glycera dibranchiata and the bill of the European vulgaris has been shown to increase the mechanical strength of these mouthparts (Bonser and Witter, 1993;

Moses et al. 2006). The role of eumelanin in strengthening feather keratin, however, is contentious. In 1995, Bonser calculated hardness of melanized and unmelanized rachises of remiges (tail feathers) from the Willow Ptarmigan (Lagopus lagopus) and found that melanized portions of the rachis were harder than the unmelanized portions. However,

Butler and Johnson (2004) reassessed eumelanin’s strengthening properties in feathers by testing not only barb melanization, but also the location of the barbs along the rachis and their cross-sectional area in the osprey Pandion haliaetus. They found that when only melanization is taken into account, the melanized parts of the feather are indeed tougher than unmelanized parts, but these differences are negligible when barb location along the feather is considered (Butler and Johnson 2004).

Melanins Confer Resistance to Microbes and Other Parasites

Melanization in vertebrates is strongly linked to immunity via the melanocortin system. This neuronal pathway regulates a number of physiological and behavioral traits, including inflammation, sexual activity, stress response, pigmentation, energy homeostasis, and aggression. The melanocortin system consists of five posttranslational bioactive peptides, called melanocortins, that are derived from the proopiomelanocortin

(POMC) gene as well as their receptors (reviewed in Roulin et al. 2011). In particular,

102 melanogenesis and immunity are linked via the melanocortin receptor MC1. When α-

MSH binds to the MC1 receptor, eumelanogenesis as well as immune functions (Luger et al. 2003). Indeed, these physiological traits appear to covary in vitro. Darker, eumelanic female barn owls, for example, rear chicks that exhibit stronger immunological responses against ectoparasitic flies (Roulin et al. 2001, 2004) and other antigens (Roulin et al.

2000). Pheomelanin, however, is produced by agouti-signaling protein (ASP) binding to the MC1 receptor, which inhibits the binding of α-MSH to that receptor. This in turn causes the downregulation of eumelanin production (Suzuki et al. 1997), but its effect on

α-MSH-induced immune responses is unknown (Luger et al. 2003).

Melanized integument itself appears to confer microbial resistance. In feathers, eumelanin has been shown to resist colonization and degradation by harmful bacteria.

Keratinases secreted by feather parasites, such as the common soil bacterium Bacillus licheniformis, can damage the beta-keratin structure of the feathers, which in turn makes it more difficult for birds to fly or attract a mate (Shawkey et al. 2007; Gunderson et al.

2009). Melanized stripes of black- and white-striped three-toed woodpecker (Picoides tridactylus) feathers were found to be more resistant to attachment and colonization by B. licheniformis than unmelanized regions in vitro than white stripes on the same feather, even when taking into account stripe location on the feather (Justyn et al. 2017).

Similarly, and perhaps directly related, melanized feathers and feather regions have been found to be more resistant to bacterial degradation than their unmelanized counterparts

(Goldstein et al. 2004, 2008; Ruiz de Castañeda et al. 2012). Given their black color, all of these feathers are assumed to be heavily eumelanized rather than pheomelanized.

Currently, the mechanisms behind eumelanin’s protection of feather keratin from

103 bacterial attachment and degradation are unknown and require further investigation. This protective function could be indirect, such as through thickening of the feather keratin

(Goldstein et al. 2004) or by increasing surface roughness (e.g. surface roughness of glass surfaces inhibits biofilm production, Mitik-Dineva et al. 2009), although differences in feather roughness are not apparent between melanized and unmelanized regions of the same feather (Justyn et al. 2017). Direct mechanisms, such as by eumelanin binding the bacterial keratinases (Gunderson et al. 2008) or increasing the feather’s electrostatic charge (as in fungi, Nosanchuck and Casadevall 1997), are also possible if the melanin is close to the feather surface, as in iridescent and glossy feathers (Maia et al. 2010).

Many animals and plants release melanins to wound sites as an immune response.

Wound melanization has been studied extensively in arthropods and may be ancestral, as it occurs in most arthropod lineages, including insects (Ashida and Brey 1995; Nakhleh et al. 2017), chelicerates, myriapods, (Bryan-Walker et al. 2007), springtails, and even some otherwise unpigmented cave-adapted species (Bilandžija et al. 2017).

Although Bilandžija et al. (2017) suggests that wound melanization has been lost in the horseshoe crab Limulus polyphemus and isopod crustaceans, like other arthropods isopods melanize and encapsulate pathogens and infected tissues to inhibit their growth and proliferation (Koehler and Poulin, 2010). These invaders are ultimately killed by melanin and its cytotoxic quinone precursors produced via the prophenoloxidase activating system in these animals (Söderhäl and Cerenius 1998; Cerenius and Söderhäl

2004). The main melanin utilized in immune response in arthropods is eumelanin produced using dopamine as a (Sugumaran 2002), although pheomelanin has recently been characterized in a few insect species using Raman spectroscopy (Galván et

104 al. 2015; Jorge García et al. 2016; Polidori et al. 2017). However, pheomelanin’s known role in insects is thus far limited to coloration (Whitton and Coates 2017). Similarly, when plants are cut they cover the wound site in a brown pigment in order to protect it from microbes. Allomelanins produced through a polyphenol pathway result in the browning that occurs soon after we cut fruits and vegetables (Solano, 2014).

Melanized hyphal walls of some fungi are resistant to lysis by other microorganisms. The non-pathogenic fungus Aspergillus nidulans synthesizes DOPA- melanin (Gonçalves et al. 2012) and non-melanic cells are more susceptible to bacterial enzymatic lysis than melanized cells in vitro (Kuo and Alexander 1967). Bloomfield and

Alexander (1967) found that catechol melanins in Rhizocotonia solani (Chen et al. 2015) and uncharacterized melanins or melanin-like substances in the spicules of Aspergillus phoenicis, sclerotium of Sclerotium rolfsii, and hyphal wall of Cladosporium sp. were similarly resistant to microbial lysis. Non-melanized regions of these fungal cells (e.g. the hyphae of A. phoenicis and S. rolfsii) were easily degraded (Bloomfield and Alexander

1967). The mechanism behind the resistence of fungal melanins to microbial hydrolytic enzymes is unknown and little studied. This resistence could be due to the binding of the hydrolytic enzymes by the melanins, the strength of bonds between the cell wall polysaccharides and the complex polymeric structure of melanins, or the steric effects of the melanins (Jacobson 2000). Resistance to lysis would also be advantageous for melanized pathogenic microorganisms.

Melanins Contribute to Microbial Virulence

105 While melanins are commonly used to deter parasitism by microorganisms, microbial melanins can also contribute to their virulence. In plant pathogens, the melanization of the appressoria of several pathogenic fungi aids in their penetration of and attachment to their host. Magnaporthe grisea, a pathogenic fungus that causes diseases such as grey leaf spot disease and rice blast disease in grasses, deposits a thick layer of DHN melanin in its cell wall to aid in the buildup of hydrostatic pressure required for hyphal penetration of host (Howard and Ferrari 1989; Chumley and Valent

1990). Pigmentless mutants of M. grisea are less successful than melanized forms or are non-pathogenic (Howard and Ferrari 1989).

Animal pathogens use melanins for protection against the host’s immune defenses. In vitro experiments have shown that both the DHN melanin and pyomelanin produced by the pathogenic fungus Aspergillus fumigatus protect it from reactive oxygen intermediates released as an immune defense by the host (Schmaler-Ripcke et al. 2009).

Cryptococcus neoformans, which can cause meningoencephalitis in immunocompromised patients or cryptococcosis in the lungs, produces DOPA-melanin

(eumelanin; Nosanchuk et al. 2015) when infecting a host (Casadevall et al. 2000), although pigment-production in C. neoformans is substrate-dependent (Karkowska-

Kuleta et al. 2009). Melanized C. neoformans are protected from reactive oxygen species

(Jacobson and Tinnell, 1993; Wang and Casadevall 1994), macrophages (Wang et al.

1995), and microbicidal peptides released by the host (Doering et al. 1999) as well as antibiotics (reviewed in Casadevall et al. 2000). Melanins produced by some pathogenic bacteria similarly function in helping them resist host immunity and increase their virulence. Proteus mirabilis, which causes infections in the urinary tract, and

106 Burkholderia cepacia, which causes lung infections, both produce dark pigments that are assumed to be pyomelanin (Turick et al. 2010). Pigments in both of these bacterial species protect cells against reactive oxygen species (Agodi et al. 1996; Zughaier et al.

1999), both of which may be released by the host as immune defenses (Nosanchuk and

Casadevall 2003; Plonka and Grabacka 2006). Similarly, pyomelanin produced by the plant pathogen Ralstonia solanacearum and Pseudomonas aeruginosa protects these virulent bacteria from peroxides and host-induced oxidative stress (Rodríguez-Rojas et al.

2009; Ahmed et al. 2016). Melanin’s negative charge may also protect melanized pathogenic cells from phagocytosis by host cells (Nosanchuck and Casadevall 1997).

Melanins Screen Ultraviolet Radiation

Ultraviolet radiation can have detrimental effects on many organisms, ranging from DNA photodamage to skin cancer. UVA (320-400 nm) causes the production of reactive oxygen species that damage DNA and DNA-protein crosslinks while UVB (280-

320 nm) is absorbed by and directly damages DNA (Brenner and Hearing 2008). UVC

(200-280 nm) is absorbed by the earth’s ozone layer and does not naturally affect life on

Earth, although UVC are often used for sterilization. Because melanins, especially eumelanin, broadly absorb across the UV spectrum, they serve as effective screens against UV radiation when incorporated into the organism’s integument or cell wall.

Melanins scatter and absorb UV and bind reactive oxygen species resulting from UVA exposure, thereby reducing the amount that enters and damages the organism (Sulaimon and Kitchell 2003; Brenner and Hearing 2008). Franco-Belussi et al. (2016) suggested that visceral eumelanin in frogs may also protect internal organs from UV damage after

107 noticing an increase in melanization and genotoxicity after long periods of external exposure to UV. However, they did not test the amount of UV radiation that penetrated the skin and these effects instead may have been caused by exposure of these tissues to reactive oxygen species rather than direct UV exposure (Mason et al. 1960; Sulaimon and

Kitchell 2003).

In microorganisms, melanin deposited in the cell wall or cell membrane can also serve a photoprotective function. Wang and Casadevall (1994) found that DOPA- melanized C. neoformans were less susceptible to irradiation by UVC than unmelanized forms. HPQ-melanin produced by Streptomyces griseus similarly protected melanized wild-type forms better than unmelanized cells in vitro (Funa et al. 2005). UV resistance has also been demonstrated in pigmented mutants of the natural pesticide Bacillus thuringiensis, which produce a dark melanin similar to animal eumelanin (Patel et al.

1996). Melanins here may not only absorb harmful UV waves, but also serve as an antioxidant to sequester the reactive oxygen species produced by UVA (Belozerskaya et al. 2017).

Melanins React with Cytotoxic Molecules

The binding properties of melanins are arguably their most important asset to most modern organisms. Melanins are themselves stable and complex free radicals that are able to chelate metals and bind reactive oxygen species (ROS) and other free radicals

(Mason et al. 1960; McGraw 2003; 2005), thereby protecting tissues, DNA, mitochondria, proteins, and numerous other macromolecules from their cytotoxic effects.

108 Pyomelanin, pheomelanin, neuromelanin, and eumelanin are all highly reactive in the presence of ROS, making these melanins effective antioxidants. Eumelanin and pheomelanin are able to bind or capture electrons from ROS and reactive nitrogen species because they contain reducing o-hydroqunione and oxidizing o-quinone functional groups (Borovansky 1996; McGraw 2005). In vertebrate eyes, the eumelanin in the retinal pigment epithelium helps prevent age-related macular degeneration by protecting the photoreceptors from ROS produced by UVA (Wang et al. 2006). Visceral eumelanin in frogs sequesters cytotoxic superoxide anions (Germia et al. 1984). As discussed above, pyomelanins and other allomelanins in virulent fungal and bacterial species protect the cells against ROS released by the host’s .

However, the production of melanins as antioxidants can be a double-edged sword. The over-production of both eumelanin and pheomelanin in animal skin as a response to increases in UVR results in the production of semiquinoid-type free radicals and reactive oxygen species (Solano 2014). While eumelanin may be large enough to scavenge these free radicals, pheomelanin may not be able to sequester the free radicals and ROS involved in its production, resulting in UV sensitivity and increased risk of skin cancer in humans rather than photoprotection (Wenczl et al. 1998; Brenner and Hearing

2008).

Melanins are also able to bind to transition metal ions, such as iron, that can cause cell damage when accumulated in high concentrations. Oftentimes cell damage is caused by oxidative stress resulting from the catabolism of oxidative reactions by these reactive metals (Stoths and Bagchi 1995). Neuromelanin is perhaps the most studied type of melanin for its ability to chelate metals because if its link to Parkinson’s disease in

109 humans. While its functions for the brain are not fully understood, neuromelanin in the substantia nigra of the human brain may chelate reactive iron and other metals that could cause neurodegradation via oxidative stress (Zecca et al. 1993; Bolzoni et al. 2002).

Neuromelanin granules serve as a major iron-storage site in the substantia nigra (Zecca et al 2001), likely as a result of both the presence of eumelanin as well as L-ferritin – a protein used for extended iron storage in tissues – in their structure (Tribl et al. 2009).

However, when overwhelmed neuromelanin and iron could cause neurotoxicity and neuroinflammation associated with Parkinson’s disease (reviewed in Zucca et al. 2017).

When Did the Functions of Melanins Evolve?

We have reviewed the modern functions of melanins in organisms ranging from bacteria and fungi to plants and animals (summarized in Table 6.1). The question now is, what may have been the first function of melanin and when did the different functions evolve? In this section, we will review the current evidence for melanin’s functions in the fossil record and attempt to elucidate the earliest function.

110 Table 6.1. Summary of the functions of melanins.

Diffuse, Taxon Tissue Type Producing Structure/Cell Melanin Type melanosome, Melanin Function References ? Rodríguez-Rojas et al. (2009); Turick et al. Bacteria Cell wall Pyomelanin Diffuse Virulence, antioxidant (2010); Ahmed et al. (2016) Allomelanin, Bacteria Cell wall Diffuse UV protection Patel et al. (1996); Funa et al. (2005) eumelanin? Allomelanin, Bloomfield and Alexander (1967); Kuo and Fungi Cell wall pyomelanin, Diffuse Immunity, antioxidant Alexander (1967); Casadevall et al. (2000); eumelanin Nosanchuk et al. (2015) Howard and Ferrari (1989); Chumley and Fungi Appressoria Allomelanin Diffuse Virulence Valent (1990) Fungi Cell wall Allomelanin Diffuse UV protection Wang and Casadevall (1994) Plantae Ovary, seed Allomelanin Diffuse Immunity Solano (2014) Animalia, Integument Melanophore Eumelanin Melanosome Coloration Messenger (2001) Cephalopoda Immunity, coloration, Clusella-Trullas et al. (2007); Shawkey et al. Animalia, Insecta Cuticle Eumelanin Diffuse thermoregulation (2009); Stavenga et al. (2012) Animalia, Cuticle Unknown Eumelanin Melanosome Unknown Hsiung et al. (2015; 2017) Araneae Animalia, Eye Unknown Pheomelanin Unknown UV protection Speiser et al. (2014) Polyplacophora Pigmented epithelium of Animalia, Eye/ the retina, iris, and Eumelanin Melanosome Antioxidant, UV protection Liu et al. (2005) Cephalochordata choroid Animalia, Integument Melanocyte/melanophore Pheomelanin Melanosome Coloration, antioxidant Borovansky (1996); McGraw (2005) Vertebrata Coloration, UV protection, Animalia, immunity, antioxidant, Integument Melanocyte/melanophore Eumelanin Melanosome McGraw (2006) Vertebrata thermoregulation, material strengthening Animalia, Neurons of the substantia Zucca et al. (1993; 2001; 2017); Bolzoni et Brain Neuromelanin Melanosome Antioxidant, metal chelation Primates nigra and locus coeruleus al. (2002)

Granules Animalia, Ink gland cells Eumelanin (derived from Anti-predator defense Palumbo (2003) Coleoidea melanosomes)

Did Melanin First Evolve for Thermoregulation?

There is currently no direct evidence for a thermoregulatory function for melanins in the fossil record. A thermoregulatory function would be difficult to infer given that this function is commonly related or secondary to other roles that melanins play in modern animals. However, Smithwick et al. (2017) suggested that an alternative function for evidence of countershading in , which they hypothesized primarily functioned for camouflage, could be thermoregulation.

The evolutionary history of melanins’ roles as thermoregulators is likewise difficult to assess. This function is limited to cutaneous melanins – namely eumelanin – in animals. Although melanins have been shown to enable microbes to survive in extreme

111 conditions in vitro, the incorporation of melanin into their cell wall may serve to strengthen it against the mechanical stresses of a freeze-thaw cycle rather than converting light into heat (Rosas and Casadevall, 1997). Melanin’s function in these microorganisms is therefore thermoprotective rather than thermoregulatory. Thermoregulation probably presented a challenge when animals first began to inhabit terrestrial environments. Given that the ability to rapidly disperse or aggregate dermal melanosomes is common between fish and amphibians, early amphibians may have likewise used rapid color change to help them adapt to life on land.

Did Melanin First Evolve as a Colorant?

The functions of eumelanin and pheomelanin for coloration are the most researched functions for melanins in extinct organisms. Melanins preserved in fossil integument are hypothesized to have served color-related functions for at least 160 million years. In 2008, Vinther et al. reinterpreted micron-scale microbodies preserved in fossil feathers as melanosomes rather than bacteria. Since this discovery, melanosomes have been found in numerous exceptionally-preserved fossil integument and, in some cases, the original melanin-based color of these animals could be determined based on their morphology. Melanosomes preserved in fossilized feathers of pennaraptoran dinosaurs – a clade that includes birds and related non-avian dinosaurs – have similar morphological disparity to melanosomes in modern bird feathers (Li et al. 2010, 2012;

Zhang et al. 2010). We can therefore assume that these diverse morphologies also correlated to the melanin-based color of the feathers of these extinct animals as they do in modern feathers. Anchiornis huxleyi, a small paravian dinosaur from the Late Jurassic of

112 China, retains spots and spangles of melanin-based color patterning of the feathers preserved throughout its body that can be observed on a macro scale. The size and shape of preserved melanosomes in A. huxlei suggest that the animal was mostly grey in color with spots of white (assumed for regions that lack preserved melanosomes) and black in its wings and a red-brown crest of feathers on its head. These colors are hypothesized to have served a display or camouflage function for the animal, as similar markings do in modern birds (Li et al. 2010). In 2012, Li et al. discovered some dinosaurs had structurally-colored feathers when they found that the 123 million-year old dromaeosaurid Microraptor had highly-elongate, arranged melanosomes indicative of at least weakly iridescent feathers hypothesized to have functioned in sexual display.

Recently, the production of melanin-enhanced structural colors was pushed back to 160 million years ago by the discovery of the Chinese paravian dinosaur juji.

Arrangements of flat, platelet-type melanosomes preserved around the head of C. juji suggest that these feathers were highly iridescent in life, as iridescent feathers in modern swifts and songbirds are colored using similar melanosome arrangements and morphologies. These iridescent feathers in Caihong also likely served as sexual ornaments (Hu et al. 2018). However, the original hue of the feathers preserved in

Microraptor, Caihong, and other examples of structurally-colored dinosaurs (e.g. bohaiornithid bird CUGB P1202; Chapter 4) cannot be assessed because the original configuration of keratin, air spaces, and melanosomes is not preserved.

Melanin’s function as a colorant probably evolved long before Caihong and

Anchiornis, given that this function is widely distributed throughout Animalia rather than restricted to birds and their extinct relatives. However, the study of fossil color is

113 currently based on the size and shape of preserved melanosomes and these parameters are only known to vary and correspond to melanin-based color in mammals and pennaraptoran dinosaurs. Melanosome morphology does not vary based on color in modern or fossil fish, amphibian, testudine, crocodylomorph, or lepidosaur skin. Instead, these melanosomes are relatively small (usually less than 1 µm in diameter) and nearly spherical no matter the color of the integument (Chapter 5; Li et al. 2014). Morphological variability also appears to be limited in other extinct taxa, including non-pennaraptoran dinosaurs such as Sinosauropteryx prima (Zheng et al. 2010) and , (Li et al. 2014), and ichthyosaurs (Lindgren et al. 2014), although it could be argued that these and other extinct taxa have not yet been sampled extensively. If it is possible to reconstruct the colors of animals without color-correlated morphological variability of their melanosomes, and therefore elucidate the functions of those colors, we will need to rely on other methods. However, reconstructing color based on chemistry alone is difficult. Eumelanin and pheomelanin are the only melanins known to be used for color-related functions and, until recently (Colleary et al. 2015; Brown et al. 2017), eumelanin was the only melanin detected in the fossil record. Even so, the ratio of these two melanins to confer different colors in fossil animals have yet to be determined. Other factors, such as input from other pigments (e.g. carotenoids), nanostructures (e.g. keratin and air spaces in feathers determine the hue of structural colors), or other chromatophores

(e.g. red xanthophores or iridescent iridophores in fish, amphibian, and lepidosaur skin) also currently cannot be determined because they have not been confidently characterized extinct taxa.

114 The question of when melanins as colorants evolved may be answered by looking into the evolution of eyes and color patterns. Melanins as colorants are only useful if the color is visible (i.e. the perceiver has eyes that can perceive differences in coloration) and, given the ubiquity of melanins in organisms that lack eyes, their function for coloration is probably derived. Fossil photoreceptors required for have been reported for an enantiornithine bird (Tanaka et al. 2017), suggesting that these birds were able to see the shiny colors they produced (inferred from Chapter 4; Li et al. 2014), and a

300 million-year-old acanthodian fish (Tanaka et al. 2014). However, the earliest eyes are known from the , when many of the major animal body plans as well as the first predators to use vision for capturing prey appeared approximately 540 million years ago. Trilobites and anomalocaridids, for example, had acute visual systems characterized by compound eyes composed of numerous image-forming lenses (Paterson et al. 2011). While it is unknown if these animals also possessed color vision, they at least may have been able to detect differences between different color tones. Indeed, some of the earliest examples of color patterns – a good place to look for the oldest examples of melanin or other pigments used for color-related functions – are from the

Cambrian. For example, the Middle Cambrian trilobite Anomocare vittata preserves dark- light stripes on its pygidium (tail; Raymond 1922). Examples of patterned trilobites, ranging from pygidial bands and spots (Williams 1933; Wells 1942; Teichert 1944;

Schoenemann et al. 2014; Vinther 2015) to a fully-spotted (McRoberts et al.

2013), occur throughout the Paleozoic. Other examples of early color patterning include nanostructures, indicative of structural coloration, preserved in the sclerites of the enigmatic Wiwaxia currogata, head shield of the arthropod Marrella

115 splendens, and setae of the annelid Canadia spinosa from the Middle Cambrian Burgess

Shale (Parker 1998).

The mechanisms behind trilobite color markings have yet to be elucidated and they have not been tested for the presence of preserved melanin or other pigments. While trilobites have no modern ancestors, eumelanin and melanosomes do occur in the cuticles of their modern relatives, the chelicerates (Hsiung et al. 2015, 2017), and may be ancestral to Arthropoda (Bilandžija et al. 2017). However, chelicerates and other arthropods also use other pigments for dark coloration (e.g. ommochromes). What the presence of color patterns on trilobites and structural coloration in Cambrian organisms suggests is that some early animals could see these color patterns. Trilobite spots and stripes are hypothesized to have had a role in camouflage (McRoberts et al. 2013;

Schoenemann et al. 2014) and it is possible that this was the earliest role for pigments as colorants. Healed bite marks preserved on the of some trilobites suggest that they were indeed attacked by top Cambrian predators like Anomalocaris (Babcock 1993;

Nedin 1999). Additionally, trilobite sclerites are preserved in the alimentary canals of some Cambrian animals, such as Utahcaris and Sydneya, as well as large

(Bruton 1981; Conway Morris and Robison 1988), although it is unclear if these animals scavenged or predated the trilobites (Babcock 2003). Signaling functions for trilobite color patterns, however, should not be ruled out given their complex visual systems.

Future studies should determine if melanin or other pigments are preserved in these early examples of color patterns, which may push pigment preservation and its function for coloration back to the Paleozoic and possibly the Cambrian Explosion.

116 Did Melanin Evolve as a Protector?

Few examples of melanin preserved in the fossil record have been hypothesized to have served protective functions. Melanin and corresponding granules found in 160 million-year-old cephalopod ink sacs probably served a similar antipredatory function as modern cephalopod ink (Glass et al. 2012). Additionally, melanosomes of the retinal pigment epithelium have been described in the exceptionally preserved eyes of an acanthodian fish and the enigmatic Tullimonstrum, suggesting that photoprotective and antioxidant functions for eumelanin are at least 307 million years old (Tanaka et al. 2014;

Clements et al. 2016). Most other examples of melanin and melanosomes in the fossil record have been hypothesized to function in primarily for coloration. This does not, however, mean that coloration was melanin’s only role, as many of the functions we have described are not mutually exclusive. For example, as reviewed above, eumelanin in feathers not only confers their color, but also deters bacterial colonization and degradation of the feather keratin (Ruiz de Castañeda et al. 2012; Justyn et al. 2017).

Feather bacterial load may have similarly affected extinct dinosaurs and the eumelanin preserved in fossil feathers, such as in Anchiornis (Lindgren et al. 2015) and bohaiornithid CUGB P1202 (Chapter 4), may have protected the feathers from degradation while also contributing to their color. It is also possible that eumelanin may have strengthened the feather keratin or served as a scavenger of free radicals as it does in many modern birds.

Generally, the protective functions of modern melanins not only span across all major taxa in which melanins are produced, but also across each of the chemical variants of melanin, suggesting that melanin is incredibly old and its first function was probably

117 protective. Given its limitation to coleoid cephalopods, eumelanin’s functions as a predator deterrent is probably derived. Most of the other protective functions that melanins serve are related to their stable free radical structure. For example, some authors

(Geremia et al. 1984; Różanowska et a. 1999) attribute much of melanins’ function in UV protection to their ability sequester the reactive oxygen species produced by UVA that can damage DNA. Eumelanin’s function in material strengthening is probably related to its ability to cross-link with proteins (Riley 1992). Melanins may also bind to enzymes to protect a host from microbial proliferation or to protect cells from lysis (Kuo and

Alexander 1967; Chen et al. 2015). Furthermore, melanins can bind to cytotoxic reactive molecules such as metals, ROS, or other free radicals to protect virulent microbes from their host’s immune system or to protect cells and vital macromolecules from damage.

Given the numerous protective advantages that melanins provide organisms ranging from bacteria to fungi, plants, and animals, the first melanin and its first protective function likely evolved relatively early in the .

We therefore need to look at the stresses on life in the oceans.

Although putative microfossils have been found in rocks that may be as old as 4.28 billion years from the Nuvvuagittuq belt in Quebec (Dodd et al. 2017), the earliest direct evidence for life on Earth are carbonaceous microorganisms preserved in the 3.47 billion- year-old Chert of Western (Schopf 1993). Isotopic analyses indicate that these some of these early microorganisms were photosynthetic while others were methane produces or consumers (Schopf et al. 2017). Evidence of microbial mats, called stromatolites, from the 3.7 billion-year-old Isua supracrustal belt of southwest Greenland suggest that photosynthesis evolved even earlier in Earth’s history (Nutman et al. 2016).

118 Much of the oxygen produced by these early photosynthetic microorganisms, or through abiotic processes (François 1986; Holland 2002; however, see Konhauser et al. 2002,

2007), was sequestered by metals, namely dissolved ferrous iron from original pyrite, and the resultant iron oxides formed the banded iron formations. Over time, these metals became saturated with oxygen and approximately 2.3 billion years ago free oxygen was released into the atmosphere and shallow ocean in what is known as the Great Oxidation

Event (Bekker et al. 2004), although the atmospheric accumulation of free oxygen may have begun earlier (Anbar et al. 2007). Cytotoxicity induced by ROS would have been detrimental to early organisms and melanin, given its ability to bind ROS, may have provided significant survival advantages to organisms that were able to produce it. The ability to bind reactive oxygen species may therefore have been the first function of early melanin. Alternatively, the reactive metals that sequestered free oxygen produced by cyanobacteria may also have reacted with and damaged microbial DNA and other vital macromolecules. Melanin’s first function could therefore have been to chelate potentially cytotoxic metals. Regardless, microorganisms that were able to produce melanin were probably less susceptible to cytotoxicity induced by reactive molecules than unmelanized individuals.

Conclusions

We have reviewed the diverse functions of the different chemical variants of melanin in modern organisms and the evidence for these functions in the fossil record with the goal of elucidating melanin’s earliest function and when it may have evolved.

119 Thus far the only melanins known to preserve in the fossil record are eumelanin and pheomelanin, although this does not necessarily mean that other melanins may not preserve nor that the oldest known example of fossil melanin (Upper Carboniferous,

Clements et al. 2016) was its earliest occurrence. Melanins are veritable Swiss army knives among modern macromolecules, as they serve a great multitude of functions for nearly every major taxon. The first melanin therefore likely evolved early in the history of life and we hypothesize that it may have been a mechanism for microbial survival against cytotoxic metals and reactive oxygen in the Precambrian oceans. However, given the differences in chemical structure between the different melanin variants and their presence in Bacteria, but not Archaea, and differential presence within Animalia, the possibility that melanins had multiple origins should be explored further.

The new field of fossil color has greatly expanded our understanding of extinct organisms in just 10 years and thus remains a fruitful avenue for future research. Future studies should look into Paleozoic color patterns, especially in the Cambrian, to determine if melanin is preserved and when its involvement in coloration may have begun. Maturation experiments on pyomelanin have yet to be completed nor has it been detected in fossils. Given its complexity and presence in many modern microbes it may persist in fossil microorganisms. Future work should also look at the potential for melanin preservation in Precambrian microorganisms.

120 CHAPTER 7

CONCLUSION

Throughout my dissertation research we have addressed questions concerning melanosome preservation in the fossil record as well as the evolution of melanins and melanosomes. We first addressed some of the concerns of those in favor of a microbial affinity for microbodies preserved in fossil feathers, skin, and hairs (Moyer et al. 2014;

Lindgren et al. 2015a). Through experimental analysis, we found that modern keratinolytic bacteria are less likely to colonize melanized stripes on feathers than unpigmented stripes. Bacterial load may have similarly affected extinct pennaraptoran dinosaur feathers, suggesting that contrary to Moyer et al. (2014) the preservation of microbodies in certain regions of fossil feathers and not others was not due to the nutritional benefits of original melanin. Additionally, we described microstructures associated with both the fossilized integument and matrix of three exceptionally preserved chordates, including the Lower Cambrian chordate Haikouella, that resemble modern microbial biofilms. We showed both morphologically and chemically that biofilm structures and fossilized bacteria differ markedly from the microbodies that many authors have interpreted as melanosomes.

121 We then described a subadult enantiornithine bird with both juvenile and adult characteristics, including iridescent plumage and elongate tail feathers that may have been used to attract a mate. Our work on melanosome morphology in modern and fossil anamniotes supports the hypothesis presented by Li et al. (2014) that convergent increases in melanosome diversity in mammals and pennaraptoran dinosaurs is associated with physiological changes in these two clades. Finally, we reviewed the functions melanins provide for modern organisms and hypothesize that the first function of melanin may have evolved in the Precambrian to combat metal or oxygen cytotoxicity. Our work has implications for both the preservational and evolutionary history of the modern world’s most ubiquitous pigment.

122 REFERENCES

Ahmed, S., Lee, S. Y., Kong, H. G., Jo, E. J., Choi, H. K., Khan, R., and Lee, S. 2016. Genetic determinants for pyomelanin production and its protective effect against oxidative stress in Ralstonia solanacearum. PLoS One, 11: doi:10.1371/journal.pone.0160845.

Agodi, A., Stefani, S., Corsaro, C., Campanile, F., Gribaldo, S., and Sichel, G. 1996. Study of a melanic pigment of Proteus mirabilis. Research in Microbiology, 147:167-174.

Allen, W. L., Baddeley, R., Cuthill, I. C., and Scott-Samuel, N. E. 2012. A quantitative test of the predicted relationship between countershading and environment. The American Naturalist, 180: doi:10.1086/668011.

Anbar, A. D., Duan, Y., Lyons, T. W., Arnold, G. L., Kendall, B., Creaser, R. A., Kaufman, A. J., Gordon, G. W., Scott, C., Garvin, J., and Buick, R. 2007. A whiff of oxygen before the ? Science, 317:1903-1906.

Andersson, M. 1982. Female choice selects for extreme tail length in a widowbird. Nature, 299:818-820.

Andersson, S. 1992. Female preference for long tails in lekking Jackson’s widowbirds: experimental evidence. Animal Behavior, 43:379-388.

Arévelo, J. E. and Heeb, P. 2005. Ontogeny of sexual dimorphism in the Long-tailed Manakin Chiroxiphia linearis: long maturation of display trait morphology. Ibis, 147:697-705.

Ashida, M., and Brey, P. T. 1995. Role of the integument in insect defense: pro-phenol oxidase cascade in the cuticular matrix. Proceedings of the National Academy of Sciences of the of America, 92:10698-10702.

Babcock, L. E. 1993. Trilobite malformations and the fossil record of behavioral asymmetry. Journal of Paleontology, 67:217-229.

123 Babcock, L. E. 2003. Trilobites in Paleozoic predator-prey systems, and their role in reorganization of early Paleozoic ecosystems. In Predator-Prey Interactions in the Fossil Record (eds. P. H. Kelley, M. Kowalewski, and T. Hansen), pp. 55-92. New York, NY: Kluwer Academic.

Barden, H. E., Wogelius, R. A., Li, D., Manning, P. L., Edwards, N. P. and van Dongen, B. E. 2011. Morphological and geochemical evidence of eumelanin preservation in the feathers of the Early Cretaceous bird, yumenensis. PLoS One, 6:1- 9.

Bekker, A., Holland, H. D., Wang, P. L., Rumble III, W. D., Stein, H. J., Hannah, J. L., Coetzee, L. L., and Beukes, N. J. 2004. Dating the rise of atmospheric oxygen. Nature, 427:117-120.

Belozerskaya, T. A., Gessler, N. N., and Aver’yanov, A. A. 2017. Melanin pigments of fungi. In Fungal Metabolites (eds. J. Mérillon and K. G. Ramawat), pp. 263-291. Switzerland: Springer International Publishing.

Beltrán-Valero de Bernabé, D., Jimenez, F. J., Aquaron, R., and Rodríguez de Córdoba, S. 1999. Analysis of alkaptonuria (AKU) mutations and polymorphisms reveals that the CCC sequence motif is a mutational hot spot in the homogentisate 1,2 dioxygenase gene (HGO). American Journal of Human Genetics, 64:1316-1322.

Berson, J. F., Harper, D. C., Tenza, D., Raposo, G., and Marks, M.S. 2001. Pmel17 initiates premelanosome morphogenesis within multivesicular bodies. Molecular Biology of the Cell, 12:3451-3464.

Bilandžija, H., Laslo, M., Porter, M. L., and Fong, D. W. 2017. Melanization in response to wounding is ancestral in arthropods and conserved in albino cave species. Scientific Reports, 7: DOI:10.1038.s41598-017-17471-2.

Bloomfield, B. J., and Alexander, M. 1967. Melanins and resistance of fungi to lysis. Journal of Bacteriology, 93:1276-1280.

Bonser, R. H. C. 1995. Melanin and the abrasion resistance of feathers. The Condor, 97:590-591.

Bonser, R. H. C., and Witter, M. S. 1993. Indentation hardness of the bill keratin of the European Starling. The Condor, 95:736-738.

Borovansky, J. 1996. Free radical activity of melanins and related substances: biochemical and pathobiochemical aspects. Sbornik Lekarsky, 97:590-591.

124 Brenner, M., and Hearing, V. J. 2008. The protective role of melanin against UV damage in human skin. Photochemistry and Photobiology, 64:539-549.

Browestijn, N., Slager, E. H., Bakker, A. B. H., Schreurs, M. W. J., Van der Spek, C. W., Adema, G. J., Schrier, P. I., and Figdor, C. G. 1997. Transcription of the gene encoding melanoma-associated antigen gp100 in tissues and cell lines other than those of the melanocyte lineage. British Journal of Cancer, 76:1562-1566.

Brown, C. M., Henderson, D. M., Vinther, J., Fletcher, I., Sistiaga, A., Herrera, J., and Summons, R. E. 2017. An exceptionally preserved three-dimensional armored dinosaur reveals insights into coloration and Cretaceous predator-prey . Current Biology, 27:1-8.

Bryan-Walker, K., Leung, T. L. F., and Poulin, R. 2007. Local of immunity against trematode parasite in marine amphipod populations. Marine Biology, 152:687-695.

Burgoyne, T., O’Connor, M. N., Seabra, M. C., Cutler, D. F., and Futter, C. E. 2015. Regulation of melanosome number, shape and movement in the zebrafish retinal pigment epithelium by OA1 and PMEL. Journal of Cell Science, 128:1400-1407.

Burkhart, C. G. and C. N. Burkhart. (2005). The mole theory: primary function of melanocytes and melanin may be antimicrobial defense and immunomodulation (not solar protection). International Journal of Dermatology 44:340-342.

Burtt, E. H., and J. M. Ichida. (1999). Occurrence of feather-degrading in the plumage of birds. The Auk 116:364-372.

Burtt, E. H., and J. M. Ichida. (2004). Gloger’s rule, feather-degrading bacteria, and color variation among Song Sparrows. The Condor 106:681-686.

Burtt, E. H., M. R. Schroeder, L. A. Smith, J. E. Sroka, and K. J. McGraw. (2010). Colourful feathers resist bacterial degradation. Biology Letters 7:214-216.

Bush, W. D., Garguilo, J., Zucca, F. A., Albertini, A., Zecca, L., Edwards, G. S., Nemanich, R. J., and Simon, J. D. 2006. The surface oxidation potential of human neuromelanin reveals a spherical architecture with a pheomelanin core and a eumelanin surface. PNAS, 103:14785-14789.

Butler, M., and Johnson, A. S. 2004. Are melanized feather barbs stronger? The Journal of Experimental Biology, 207:285-293.

Caldwell, R. L. 2005. An observation of inking behavior protecting adult bocki from predation by green turtle (Chelonia mydas) hatchlings. Pacific Science, 59:69-72.

125 Capozzi, V., Perna, G., Gallone, A., Biagi, P. F., Carmone, P., Fratello, A., Guida, G., Zanna, P., and Cicero, R. 2005. Raman and optical spectroscopy of eumelanin films. Journal of Molecular Structure, 744:717-721.

Carney, R. M., Vinther, J., Shawkey, M. D., D’Alba, L., and Ackermann, J. 2012. New evidence on the colour and nature of the isolated feather. Nature Communications 3:1-6.

Casadevall, A., Rosas, A. L., and Nosanchuk, J. D. 2000. Melanin and virulence in Cryptococcus neoformans. Current Opinion in Microbiology, 3:354-358.

Centeno, S., and Shamir, J. 2008. Surface enhanced Raman scattering (SERS) and FTIR characterization of the sepia melanin pigment used in works of art. Journal of Molecular Structure, 873:149-159.

Cerenius, L., and Söderhäl, K. 2004. The prophenoloxidase-activating system in invertebrates. Immunological Reviews, 198:116-126.

Chen, J., Wang, S., Zhu, M., and Zhou, E. 2015. Isolation and characterization of a melanin from Rhizoctonia solani, the causal agent of rice sheath blight. European Journal of Plant Pathology, 142:281-290.

Chiappe, L. M. 1995. The phylogenetic position of the Cretaceous birds of : Enantiornithes and deferrariisi. Courier Forschungsinstitut Senckenberg, 181:55-63.

Chiappe, L. M., Marugán-Lobón, J., Ji, S. and Zhou, Z., 2008. Life history of a basal bird: of the Early Cretaceous Confuciusornis. Biology Letters, 4:719-723.

Chiappe, L. M., Ji, S., and Ji, Q. 2007. Juvenile birds from the Early Cretaceous of China: implications for enantiornithine ontogeny. American Museum Novitates 3594:1- 46.

Chiappe, L. M., Ji, S., Ji, Q. and Norell, M.A. 1999. Anatomy and of the (: Aves) from the Late of northeastern China. Bulletin of the American Museum of Natural History 242:1-89.

Chiappe, L. M., and Walker, C. A. 2002. Skeletal morphology and systematics of the Cretaceous enantiornithines (Ornithoraces: Enantiornithes). In Chiappe, L. M. and Witmer, L. M. (eds.). Mesozoic Birds Above the Heads of Dinosaurs. University of California Press, 240-67, 520 pp.

Chinsamy, A., Chiappe, L. M., Marugán-Lobón, J., Chunling, G. and Fengjiao, Z. 2013. Gender identification of the Mesozoic bird Confuciusornis sanctus. Nature Communications:doi:10.1038/ncomms2377.

126

Chumley, R. G., and Valent, B. 1990. Genetic analysis of melanin-deficient, nonpathogenic mutants of Magnaporthe grisea. Molecular Plant-microbe Interactions, 3:135-143.

Clarke, J. A., Zhou, Z. and Zhang, F. 2006. Insight into the evolution of avian flight from a new clade of Early Cretaceous ornithurines from China and the morphology of grabau. Journal of Anatomy, 208, 287-308.

Clarke, J. A., and Middleton, K. 2008. Mosaicism, modules, and the : results from a Bayesian approach to the study of morphological evolution using discrete character data. Systematic Biology, 57:185-201.

Clarke, J. A., Ksepka, D. T., Salas-Gismondi, R., Altamirano, A. J., Shawkey, M. D., D’Alba, L., Vinther, J., Devries, T. J., and Baby, P. 2010. Fossil evidence for evolution of the shape and color of penguin feathers. Science, 330:954-957.

Clarke, J. A., and Norell, M. A. 2002. The morphology and phylogenetic position of Asparavis ukhaana from the Late Cretaceous of Mongolia. American Museum Novitates, 3387:1-46.

Clements, T., Dolocan, A., Martin, P., Purnell, M. A., Vinther, J., and Gabbott, S. E. 2016. The eyes of Tullimonstrum reveal a vertebrate affinity. Nature 532, 500- 503.

Clusella Trullas, S., van Wyk, J. H., and Spotila, J. R. 2007. Thermal melanism in ectotherms. Journal of Thermal Biology, 32:235-245.

Colleary, C., A. Dolocan, J. Gardner, S. Singh, M. Wuttke, R. Rabenstein, J. Habersetzer, S. Schaal, M. Feseha, M. Clemens, B. F. Jacobs, E. D. Currano, L. L. Jacobs, R. L. Sylvestersen, S. E. Gabbot, and J. Vinther. (2015). Chemical, experimental, and morphological evidence for diagenetically altered melanin in exceptionally preserved fossils. PNAS. doi:10.01073/pnas.1509831112.

Colthup, N. B., Daly, L. H., and Wiberley, S. E. 1990. Introduction to Infrared and Raman Spectroscopy, 3rd Edition. Academic Press, New York, NY, 547 pp.

Colombo, S., Berlin, I., Delmas, V., and Larue, L. 2011. Classical and nonclassical melanocytes in vertebrates. In Melanins and Melanosomes: Biosynthesis, Biogenesis, Physiological, and Pathological Functions (eds. J. Borovanský and P. A. Riley), pp. 21-61. Weinheim, Germany: Wiley-Blackwell.

Czirják, G. Á., P. L. Pap, C. I. Vágási, M. Giraudeau, C. Mureşan, P. Mirleau, and P. Heeb. (2013). Preen gland removal increases plumage bacterial load but not that of feather-degrading bacteria. Naturwissenschaften 100:145-151.

127 D'Alba, L., D. N. Jones, H. T. Badawy, C. M. Eliason, and M. D. Shawkey. (2014). Antimicrobial properties of a nanostructured eggshell from a compost-nesting bird. Journal of Experimental Biology 217:1116-1121.

D'Alba, L., R. Maia, M. E. Hauber, and M. D. Shawkey. (2016). The evolution of eggshell cuticle in relation to nesting ecology. Proceedings of the Royal Society B 283. doi.org/10.1098/rspb.2016.0687

Dalsätt, J., Ericson, P. G. P. and Zhou, Z. 2014. A new Enantiornithes (Aves) from the Early Cretaceous of China. Acta Geologica Sinica, 88:1034-1040.

Davis, P. G., and D. E. G. Briggs. (1999). Fossilization of feathers. Geology 23:783-786.

De Souza Carvalho, I., Novas, F. E., AGNOLÍN, F. L., ISASI, M. P., FREITAS, F. I., and ANDRADE, J. A. 2015. A Mesozoic bird from Gondwana preserving feathers. Nature communications, 6:doi:10.1038/ncomms814.

Dodd, M. S., Papineau, D., Grenne, T., Slack, J. F., Rittner, M., Pirajno, F., O’Neil, J., and Little, C. T. S. 2017. Evidence for early life in Earth’s oldest hydrothermal vent precipitates. Nature, 543:60-64.

Doering, T. L., Nosanchuk, J. D., Roberts, W. K., and Casadevall, A. 1999. Melanin as a potential cryptococcal defence against microbicidial proteins. Medical Mycology, 37:175-181.

Doucet, S. M., and Meadows, M. G. 2009. Iridescence: a functional perspective. Journal of the Royal Society Interface, 6:S115-S132.

Ducrest, A., Keller, L., and Roulin, A. 2008. Pleiotropy in the melanocortin system, coloration and behavioral syndromes. Trends in Ecology and Evolution, 23:502- 510.

Dunne, W. M. (2002). Bacterial adhesion: seen any good biofilms lately? Clinical Microbiology Reviews 15:155-166.

Durrer, H. 1986. Colouration. In Bereiter-Hahn, M., Matolsky, A. G. and Richards, K. S. (eds.). Biology of the Integument 2: Vertebrates. Springer-Varlag, 239-47, 855 pp.

Eliason, C. M., Bitton, P., and Shawkey, M. D. 2013. How hollow melanosomes affect iridescent colour production in birds. Proceedings of the Royal Society B, 280: doi: 10.1098/rspb.2013.1505.

Eliason, C. M., and Shawkey, M. D. 2012. A photonic heterostructure produces diverse iridescent colours in duck wing patches. Journal of The Royal Society Interface, 9: 2279-2289.

128

Eliason, C. M., Shawkey, M. D., and Clarke, J. A. 2016. Evolutionary shifts in the melanin-based color system of birds. Evolution, 70:445-455.

Ellers, J., and Boggs, C. L. 2004. Functional ecological implications of intraspecific differences in wing melanization in Colias butterflies. Biological Journal of the Linnean Society, 82:79-87.

Emaresi, G., Ducrest, A., Bize, P., Richter, H., Simon, C., and Roulin, A. 2013. Pleiotropy in the melanocortin system: expression levels of this system are associated with melanogenesis and pigmentation in the tawny (Strix aluco). Molecular Biology, 22:4915-4930.

Erickson, G. M., Curry Rogers, K., Varricchio, D. J., Norell, M. A., and Xu, X. 2007. Growth patterns in brooding dinosaurs reveals the timing of sexual maturity in non-avian dinosaurs and genesis of the avian condition. Biology Letters, 3:558- 561.

Evans, M. R., and Hatchwell, B. J. 1992. An experimental study of male adornment in the scarlet-tufted malachite sunbird: II. The role of the elongated tail in mate choice and experimental evidence for a handicap. Behavioral Ecology and Sociobiology, 29:421-427.

Evans, M. R., and Thomas, A. L. R. 1992. The aerodynamic and mechanical effects of elongated tails in the scarlet-tufted malachite sunbird: measuring the cost of a handicap. Animal Behavior, 43:337-347.

Fedorow, H., Tribl, F., Halliday, G., Gerlach, M., Riederer, P., and Double, K. L. 2005. Neuromelanin in human dopamine neurons: comparison with peripheral melanins and relevance to Parkinson’s disease. Progress in Neurobiology, 75:109-124.

Fernández-Cañón, J. M., Granadino, B., Beltrán-Valero de Bernabé, D., Renedo, M., Fernández-Ruiz, E., Peñalva, M. A., and Rodríguez de Córdoba, S. 1996. The molecular basis of alkaptonuria. Nature Genetics, 14:19-24.

Field, D. J., D’Alba, L., Vinther, J., Webb, S. M., Gearty, W., and Shawkey, M. D. 2013. Melanin concentration gradients in modern and fossil feathers. PLoS One, 8:1-6.

Fox, D. L. (1976). Animal biochromes and structural colours: physical, chemical, distributional & physiological features of coloured bodies in the animal world. Univ of California Press.

Franco-Belussi, L., Nilsson Sköld, H., and de Oliveira, C. 2016. Internal pigment cells respond to external UV radiation in frogs. Journal of Experimental Biology, 219:1378-1383.

129 François, L. M. 1986. Extensive deposition of banded iron formations was possible without photosynthesis. Nature, 320:352-354.

Frankena, J., H. W. van Verseveld, and A. H. Stouthamer. (1985). A continuous culture study of the bioenergetic aspects of growth and production of exocellular protease in Bacillus licheniformis. Applied Microbiology and Biotechnology 22:169-176.

Funa, N., Funabashi, M., Ohnishi, Y., and Horinouchi, S. 2005. Biosynthesis of hexahydroxyperylenequinone melanin via oxidative aryl coupling by cytochrome P-450 in Streptomyces griseus. Journal of Bacteriology, 187:8149-8155.

Furumura, M., Sakai, C., Potterf, S. B., Vieira, W. D., Barsh, G. S., and Hearing, V. J. 1998. Characterization of genes modulated during pheomelanogenesis using differential display. Proceedings of the National Academy of Sciences of the United States of America, 95:7374-7378.

Gabbott, S. E., Donaghue, P. C. J., Sansom, R. S., Vinther, J., Dolocan, A., and Purnell, M. A. 2016. Pigmented anatomy in Carboniferous cyclostomes and the evolution of the vertebrate eye. Proceedings of the Royal Society B, 283: doi:10.1098/rspb.2016.1151.

Galván, I., Jorge, A., Edelaar, P., and Wakamatsu, K. 2015. Insects synthesize pheomelanin. Pigment Cell Research, 28:599-602.

Galván, I., Jorge, A., Ito, K., Tabuchi, K., Solano, F., and Wakamatsu, K. 2013b. Raman spectroscopy as a non-invasive technique for the quantification of melanins in feathers and hairs. Pigment Cell & Melanoma Research, 26:917-923.

Galván, I., Jorge, A., Solano, F. and Wakamatsu, K. 2013a. Vibrational characterization of phaeomelanin and trichochrome F by Raman spectroscopy. Spectrochemica Acta Part A: Molecular and Biomolecular Spectroscopy, 110:55-59.

Gauthier, J. 1986. Saurischian and the . Memoirs of the California Academy of Science, 8:1-55.

Gauthier, J., and De Queiroz, K. 2001. Feathered dinosaurs, flying dinosaurs, crown dinosaurs, and the name “Aves.” In Gauthier, J. and Gall, L. F. New Perspectives on the Origin and Evolution of Birds: Proceedings of the International Symposium in Honor of John H. Ostrom. Peabody Museum of Natural History, Yale University, 7-41, 613 pp.

Geremia, E., Corsaro, C., Bonomo, R., Giardinelli, R., Pappalardo, P., Vanella, A., and Sichel, G. 1984. Eumelanins as free radicals trap and superoxide dismutase activities in Amphibia. Comparative Biochemistry and Physiology, 79B:67-69.

130 Glass, K., Ito, S., Wilby, P. R., Sota, T., Nakamura, A., Bowers, C. R., Vinther, J., Dutta, S., Summons, R., Briggs, D. E. G., Wakamatsu, K., and Simon, J. D. 2012. Direct chemical evidence for eumelanin pigment from the Jurassic Period. PNAS, 109:10218-10223.

Glass, K., Ito, S., Wilby, P. R., Sota, T., Nakamura, A., Bowers, C. R., Miller, K. E., Dutta, S., Summons, R. E., Briggs, D. E. G., Wakamatsu, K., and Simon, J. D. 2013. Impact of diagenesis and maturation on the survival of eumelanin in the fossil record. Organic Geochemistry, 64:29-37.

Golcuk, K., Mandair, G. S., Callender, A. F., Saher, N., Kohn, D. H., and Morris, M. D. 2006. Is photobleaching necessary for Raman imaging of bone tissue using a green laser? Biochemica et Biophysica Acta, 1758:868-873.

Goldstein, G., Flory, K. R., Browne, B. A., Majid, S., Ichida, J. M., and Burtt, E. H. 2004. Bacterial degradation of black and white feathers. The Auk, 121:656-659.

Gonçalves, R. C. R., Lisboa, H. C. F., and Pombeiro-Sponchiado, S. R. 2012. Characterization of melanin pigment produced by Aspergillus nidulans. World Journal of Microbiology and Biotechnology, 28:1467-1474.

Grande, J. M., J. J. Negro, and M. J. Torres, M. J. (2004). The evolution of bird plumage colouration: a role for feather-degradation bacteria? Ardeola 51:375-383.

Greenewalt, C. H., Brandt, W., and Friel, D. D. 1960. Iridescent colors of feathers. JOSA, 50:1005-1013.

Gunderson, A. R., M. H. Forsyth, and J. P. Swaddle. (2009). Evidence that plumage bacteria influence feather coloration and body condition of eastern Sialia sialis. Journal of Avian Biology, 40:440-447.

Gunderson, A. R., A. M. Frame, J. P. Swaddle, and M. H. Forsyth. (2008). Resistance of melanized feathers to bacterial degradation: is it really so black and white? Journal of Avian Biology, 39:539-545.

Hegna, R. H., Nokelainen, O., Hegna, J. R., and Mappes, J. 2013. To quiver or to shiver: increased melanization benefits thermoregulation, but reduces warning signal efficacy in the wood tiger moth. Proceedings of the Royal Society B, 280: DOI:10.1098/rspb.2012.2812.

Heers, A. M., and Dial, K. P. 2012. From extant to extinct: locomotor ontogeny and the evolution of avian flight. Trends in Ecology & Evolution, 27:296-305.

Hillgarth, N., and J. C. Wingfield. (1997). Testosterone and immunosuppression in vertebrates: implications for parasite-mediated sexual selection. In Parasites and Pathogens (N. E. Beckage, Editor). Chapman & Hall, NY, USA.

131

Holland, H. D. 2002. Volcanic gases, black smokers, and the Great Oxidation Event. Geochimica et Cosmochimica Acta, 66:3811-3826.

Hone, D. W. E., Tischlinger, H., Xu, X. and Zhang, F. 2010. The extent of the preserved feathers on the four-winged dinosaur Microraptor gui under ultraviolet light. PLoS One, 5:1-7.

Hong, L., and Simon, J. D. 2007. Current understanding of the binding sites, capacity, affinity, and biological significance of metals in melanin. Journal of Physical Chemistry B, 111:7938-7947.

Howard, R. J., and Ferrari, M. A. 1989. Role of melanin in appressorium function. Experimental Mycology, 13:403-418.

Hsiung, B., Blackledge, T. A., and Shawkey, M. D. 2015. do have melanin after all. Journal of Experimental Biology, 218:3632-3625.

Hsiung, B., Justyn, N. M., Blackledge, T. A., and Shawkey, M. D. 2017. Spiders have rich pigmentary and structural colour palettes. Journal of Experimental Biology, 220:1975-1983.

Hou, L., Martin, L. D., Zhou, Z., Feduccia, A. and Zhang, F. 1999. A diapsid skull in a new species of primitive bird Confuciusornis. Nature, 399:679-682.

Hu, D., Clarke, J. A., Eliason, C. M., Qiu, R., Li, Q., Shawkey, M. D., Zhao, C., D’Alba, L., Jiang, J., and Xu, X. 2018. A bony-crested Jurassic dinosaur with evidence of iridescent plumage highlights complexity in early paravian evolution. Nature Communications, 9: DOI:10.1038/s4167-017-02515-y.

Hu, D., Hou, L., Zhang, L. and Xu, X. 2009. A pre-Archaeopteryx troodontid theropod from China with long feathers on the metatarsus. Nature, 461:640-643.

Hu, D., Li, L., Hou, L. and Xu, X. 2011. A new enantiornithine bird from the Lower Cretaceous of western Liaoning, China. Journal of , 31:154-161.

Hu, D., Xu, X., Hou, L. and Sullivan, C. 2012. A new enantiornithine bird from the Lower Cretaceous of western Liaoning, China, and its implications for early avian evolution. Journal of Vertebrate Paleontology, 32:639-645.

Hu, H., O’Connor, J. K. and Zhou, Z. 2015. A new species of Pengornithidae (Aves: Enantiornithes) from the Lower Cretaceous of China suggests a specialized scansorial habitat previously unknown in early birds. PLoS One: DOI:10.1371/journal.pone.0126791.

132 Hu, H., Zhou, Z. and O’Connor, J. 2014. A subadult specimen of Pengornis and character evolution in Enantiornithes. Vertebrata PalAsiatica, 52:77-97.

Huang, J., Wang, X., Hu, Y., Liu, J., Peteya. J. A., and Clarke, J. A. 2016. A new ornithurines from the Early Cretaceous of China sheds light on the evolution of early ecological cranial diversity in birds. PeerJ:doi:10.7717/peerj.1765.

Huang, Z., Lui, H., Chen, X. K., Alajlan, A., McLean, D. I. and Zeng, H. 2004. Raman spectroscopy of in vivo cutaneous melanin. Journal of Biomedical Optics, 9:1198- 1205.

Ito, S., & Fujita, K. (1985). Microanalysis of eumelanin and pheomelanin in hair and melanomas by chemical degradation and liquid chromatography. Analytical biochemistry, 144(2), 527-536.

Ivins, B. E., and Holmes, R. K. 1981. Factors affecting pheomelanin production by melanin-producing (mel) mutant of Vibrio cholerae. Infection and Immunity, 34:895-899.

Jacobson, E. S. 2000. Pathogenic roles for fungal melanins. Clinical Microbiology Reviews, 13:4708-4717.

Jacobson, E. S., and Tinnell, S. B. 1993. Antioxidant function of fungal melanin. Journal of Bacteriology, 175:7102-7104.

Jorge García, A., Polidori, C., Nieves-Aldrey, J. L. 2016. Pheomelanin in the secondary sexual characters of male parasitoid wasps (Hymenoptera: Pteromalidae). Arthropod Structure & Development, 45:311-319.

Karkowska-Kuleta, J., Rapala-Kozik, M., and Kozik, A. 2009. Fungi pathogenic to humans: molecular bases of virulence of Candida albicans, Cryptococcus neoformans, and Aspergillus fumigatus. Acta Biochemical Polonica, 56:211-224.

Kent, C. M., and E. H. Burtt. (2016). Feather-degrading bacilli in the plumage of wild birds: prevalence and relation to feather wear. The Auk 133:583-592.

Kerje, S., Sharma, P., Gunnarsson, U., Kim, H., Bagchi, S., Fredriksson, R., Schütz, K., Jensen, P., Heijne, G. V.,Okimoto, R., and Andersson, L. 2004. The Dominant white, Dun and Smoky color variants in chicken are associated with insertion/deletion polymorphisms in the PMEL17gene. Genetics, 168:1507–1518.

Kimura, T., Nagao, Y., Hashimoto, H., Yamamoto-Shiraishi, Y., Yamamoto, S., Yabe, T., Takada, S., Kinoshita, M., Kuroiwa, A., and Naruse, K. 2014. Leucophores are similar to xanthophores in their specification and differentiation processes in medaka. Proceedings of the National Academy of Sciences of the United States of America, 111:7343-7348.

133

Koehler, A. V., and Poulin, R. 2010. Host partitioning by parasites in an intertidal community. Journal of Parasitology, 96:862-868.

Konhauser, K. O., Amskold, L., Lalonde, S. V., Posth, N. R., Kappler, A., and Anbar, A. 2007. Decoupling photochemical Fe(II) oxidation from shallow-water BIF deposition. Earth and Planetary Science Letters, 258:87-100.

Konhauser, K. O., Hamade, T., Raiswell, R., Morris, R. C., Ferris, F. G., Southam, G., and Canfield, D. E. 2002. Could bacteria have formed the Precambrian banded iron formations? Geology, 30:1079-1082.

Kuo, M. J., and M. Alexander. 1967. Inhibition of lysis of fungi by melanins. Journal of Bacteriology 94:624-629.

Kushimoto, T., Basrur, V., Valencia, J., Matsunaga, J., Vieira, W.D., Ferrans, V. J., Muller, J., Appella, E. and Hearing, V. J. 2001. A model for melanosome biogenesis based on the purification and analysis of early melanosomes. Proceedings of the National Academy of Sciences of the United States of America, 98:10698-10703.

Kwon, B. S. 1993. Pigmentation genes: the tyrosinase gene family and the pmel 17 gene family. Journal of investigative dermatology, 100:S134-S140.

Langfelder, K., Streibel, M., Jahn, B., Haase, G., and Brakhage, A. A. 2003. Biosynthesis of fungal melanins and their importance for human pathogenic fungi. Fungal Genetics and Biology, 38:143-158.

Leertouwer, H. L., Wilts, B. D., & Stavenga, D. G. (2011). Refractive index and dispersion of butterfly chitin and bird keratin measured by polarizing interference microscopy. Optics Express, 19(24), 24061-24066.

Li, Q., Clarke, J. A., Gao, K., Zhou, C., Meng, Q., Li, D., D’Alba, L., and Shawkey, M. D. 2014a. Melanosome evolution indicates a key physiological shift within feathered dinosaurs. Nature, 507:350-353.

Li, Q., Gao, K., Vinther, J., Shawkey, M. D., Clarke, J. A., D’Alba, L., Meng, Q., Briggs, D. E. G., and Prum, R. O. 2010. Plumage color patterns of an extinct dinosaur. Science, 327:1369-1372.

Li, Q., Gao, K., Meng, Q., Clarke, J. A., Shawkey, M. D., D’Alba, L., Pei, R., Ellison, M., Norell, M. A., and Vinther, J. 2012. Reconstruction of Microraptor and the evolution of iridescent plumage. Science, 335:1215-1219.

Li, Z., Zhou, Z., Wang, M. and Clarke, J. A. 2014b. A new specimen of large-bodied basal enantiornithine Bohaiornis from the early Cretaceous of China and the

134 inference of feeding ecology in Mesozoic birds. Journal of Paleontology, 88:99- 108.

Lindgren, J., Kuriyama, T., Madsen, H., Sjövall, P., Zheng, W., Uvdal, P., Engdahl, A., Moyer, A. E., Gren, J. A., Kamezaki, N., Ueno, S., and Schweitzer, M. H. 2017. Biochemistry and adaptive colouration of an exceptionally preserved juvenile fossil sea turtle. Scientific Reports, 7: DOI:10.1038/s41598-017-13187-5.

Lindgren, J., Moyer, A., Schweitzer, M. H., Sjövall, P., Uvdal, P., Nilsson, D. E., Heimdal, J., Engdahl, A., Gren, J. A., Schultz, B. P., and Kear, B. P. 2015a. Interpreting melanin-based coloration through deep time: a critical review. Proceedings of the Royal Society B, 282: DOI: 10.1098/rspb.2015.0614.

Lindgren, J., Sjövall, P., Carney, R. M., Cincotta, A., Uvdal, P., Hutcheson, S. W., Gustafsson, O., Lefèvre, U., Escuillié, F., Heimdal, J., Engdahl, A., Gren, J. A., Kear, B. P., Wakamatsu, K., Yans, J., and Godefroit, P. 2015b. Molecular composition and ultrastructure of Jurassic paravian feathers. Scientific Reports, 5: Doi:10.1038/srep13520.

Lindgren, J., Sjövall, P., Carney, R. M., Uvdal, P., Gren, J. A., Dyke, G., Schultz, B. P., Shawkey, M. D., Barnes, K. R., and Polcyn, M. J. 2014. Skin pigmentation provides evidence of convergent melanism in extinct marine . Nature, 506:484-488.

Linnaeus, C. 1758. Systema natura per regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. Ed. 10, tom. 1-3 Holmiae: Impensis L. Salvii.

Liu, S. Y., Shawkey, M. D., Parkinson, D., Troy, T. P., and Ahmed, M. 2014. Elucidation of the chemical composition of avian melanin. Royal Society of Chemistry Advances, 4:40396-40399.

Liu, Y., Hong, L., Wakamatsu, K., Ito, S., Adhyaru, B. B., Cheng, C., Bowers, C. R., and Simon, J. D. 2005. Comparisons of the structural and chemical properties of melanosomes isolated from retinal pigment epithelium, iris, and choroid of newborn and mature bovine eyes. Photochemistry and Photobiology, 81:510-516.

Liu, Y., Kempf, V. R., Nofsinger, J. B., Weinerty, E. E., Rudnicki, M., Wakamatsu, K., Ito, S. and Simon, J. D. 2003. Comparison of the structural and physical properties of human hair eumelanin following enzymatic or acid/base extraction. Pigment Cell Research, 16:355-365.

Logan, D. W., Burn, S. F., and Jackson, I. J. 2006. Regulation of pigmentation in zebrafish melanophores. Pigment Cell Research, 19:206-213.

135 Luger, T. A., Scholzen, T. E., Brzoska, T., and Böhm, M. 2003. New insights into the functions of α-MSH and related peptides in the immune system. Annals of the New York Academy of Sciences, 994:133-140.

Mackintosh, J.A. (2001). The antimicrobial properties of melanocytes, melanosomes and melanin and the evolution of black skin. Journal of Theoretical Biology 211:101- 113.

Maia, R., Caetano, J. V. O., Báo, S. N., & Macedo, R. H. (2009). Iridescent structural colour production in male blue-black grassquit feather barbules: the role of keratin and melanin. Journal of the Royal Society Interface, 6:S203-S211.

Maia, R., L. D'Alba, and M. D. Shawkey. (2010). What makes a feather shine? A nanostructural basis for glossy black colours in feathers. Proceedings of the Royal Society of London B. doi:10.1098/rspb.2010.1637.

Maia, R., Macedo, R. H. F. and Shawkey, M. D. 2012. Nanostructural self-assembly of iridescent feather barbules through depletion attraction of melanosomes during keratinization. Journal of the Royal Society Interface, 9:734-743.

Maia, R., Rubenstein, D. R., and Shawkey, M. D. 2013. Key ornamental innovations facilitate diversification in an avian radiation. PNAS, 110:10687-10692.

Mason, H. S., Ingram, D. J. E., and Allen, B. 1960. The free radical property of melanins. Archives of Biochemistry and Biophysics, 86:225-230.

McGinness, J., and Proctor, P. H. 1973. The importance of the fact that melanin is black. Journal of Theoretical Biology, 39:677-678.

McGraw, K. J. 2003. Melanins, metals, and mate quality. Oikos, 102:402-406.

McGraw, K. J. 2005. The antioxidant function of many animal pigments: are there consistent health benefits of sexually selected colourants? Animal Behavior, 69:757-764.

McGraw, K. J., 2006a. Mechanics of -based coloration. In Bird Coloration, Vol. 1 (eds. G. E. Hill and K. J. McGraw), pp. 177-242. Cambridge, MA: Harvard University Press.

McGraw, K. J. 2006b. Mechanics of melanin-based coloration. In Bird Coloration, Vol. 1 (eds. G. E. Hill and K. J. McGraw), pp. 243-294. Cambridge, MA: Harvard University Press.

McGraw, K. J., Mackillop, E. A., Dale, J. and Hauber, M. E. 2002. Different colors reveal different information: how nutritional stress affects the expression of

136 melanin and structurally based ornamental plumage. The Journal of Experimental Biology, 205:3747-3755.

McGraw, K. J., Safran, R. J., & Wakamatsu, K. 2005. How feather colour reflects its melanin content. Functional Ecology, 19:816-821.

McGraw, K. J., Wakamatsu, K., Clark, A. B., & Yasukawa, K. 2004. Red‐winged blackbirds Agelaius phoeniceus use carotenoid and melanin pigments to color their epaulets. Journal of Avian Biology, 35:543-550.

McNamara, M. E., Briggs, D. E. G., Orr, P. J., Field, D. J., and Wang, Z. 2013. Experimental maturation of feathers: implications for reconstructions of fossil feather colour. Biology Letters, 9:1-6.

McNamara, M. E., van Dongen, B. E., Lockyer, N. P., Bull, I. D., and Orr, P. J. 2016. Fossilization of melanosomes via sulfurization. Palaeontology, 59:337-350.

McRoberts, C. A., Hegna, T. A., Burke, J. J., Stice, M. L., Mize, S. K., and Martin, M. J. 2013. Original spotted patterns on Middle Devonian phacopid trilobites from western and central New York. Geology, 41:607-610.

Messenger, J. B. 2001. Cephalopod chromatophores: neurobiology and natural history. Biological Reviews, 76:473-528.

Meunier, J., Pinto, S. F., Burri, R., & Roulin, A. 2011. Eumelanin-based coloration and fitness parameters in birds: a meta-analysis. Behavioral Ecology and Sociobiology, 65:559-567.

Mitik-Dineva, N., J. Wang, V. K. Truong, P. Stoddart, F. Malherbe, R. J. Crawford, and E. P. Ivanova. (2009). Escherichia coli, Pseudomonas aeruginosa, and Staphylococcus aureus attachment patterns on glass surfaces with nanoscale roughness. Current Microbiology 58:268-273.

Moses, D. N., Mattoni, M. A., Slack, N. L., Waite, J. H., and Zok, F. W. 2006. Role of melanin in mechanical properties of Glycera jaws. Acta Biominerialia, 2:521-530.

Moyer, A. E., Zheng, W., Johnson, E. A., Lamanna, M. C., Li, D., Lacovara, K. J., and Schweitzer, M. H. 2014. Melanosomes or microbes: testing an alternative hypothesis for the origin of microbodies in fossil feathers. Scientific Reports, 4:1- 9.

Nakhleh, J., El Moussawi, L., and Osta, M. A. 2017. The melanization response in insect immunity. Advances in Insect Physiology, 52:83-109.

Nedin, C. 1999. Anomalocaris predation on nonmineralized and mineralized trilobites. Geology, 27:987-990.

137

Norell, M. A. and Xu, X. 2005. Feathered dinosaurs. Annual Review of Earth and Planetary Sciences, 33:277-299.

Nosanchuk, J. D., and A. Casadevall. 1997. Cellular charge of Cryptococcus neoformans: contributions from the capsular polysaccharide, melanin, and monoclonal antibody binding. Infection and Immunity 65:1836-1841.

Nosanchuk, J. D., and Casadevall, A. 2003. The contribution of melanin to microbial pathogenesis. Cellular Microbiology, 5:203-223.

Nosanchuk, J. D., and Casadevall, A. 2006. Impact of melanin on microbial virulence and clinical resistance to antimicrobial compounds. Antimicrobial Agents and Chemotherapy, 50:3519-3528.

Nosanchuk, J. D., Stark, R. E., and Casadevall, A. 2015. Fungal melanin: what do we know about structure? Frontier in Microbiology, 6: DOI:10.3389/fmicb.2015.01463.

Noyes, J. A., Vukusic, P., and Hooper, I. R. 2007. Experimental method for reliably establishing the refractive index of buprestid exocuticle. Optics express, 15:4351-4358.

Nutman, A. P. Bennett, V. C., Friend, C. R. L., van Kranendonk, M. J., and Chivas, A. R. 2016. Rapid emergence of life shown by discovery of 3,700-million-year-old microbial structures. Nature, 537:535-538.

O’Connor, J. K., Wang, M., Zheng, X., Wang, X. and Zhou, Z. 2014. The histology of two female Early Cretaceous birds. Vertebrata PalAsiatica, 52:112-128.

O’Connor, J. K., Wang, X., Zheng, X., Hu, H., Zhang, X. and Zhou, Z. 2015. An enantiornithine with a fan-shaped tail, and the evolution of the rectricial complex in early birds. Current Biology, 26:114-119.

O’Connor, J. K., Zhang, Y., Chiappe, L. M., Meng, Q., Li, Q., and Di, L. 2013. A new enantiornithine from the Yixian Formation with the first recognized avian enamel specialization. Journal of Vertebrate Paleontology, 33:1-12.

Oliphant, L. W., Hudon, J., and Bagnara, J. T. 1992. Pigment cell refugia in homeotherms—the unique evolutionary position of the iris. Pigment Cell & Melanoma Research, 5:367-371.

Palumbo, A. 2003. Melanogenesis in the ink gland of Sepia officinalis. Pigment Cell Research, 16:517-522.

138 Parker, A. R. 1998. Colour in Burgess Shale animals and the effect of light on evolution in the Cambrian. Proceedings of the Royal Society of London B, 265:967-972.

Parker, G. H. 1930. Chromatophores. Biological Reviews, 5:59-90.

Patel, K. R., Wyman, J. A., Patel, K. A., and Burden, B. J. 1996. A mutant of Bacillus thuringiensis producing a dark-brown pigment with increased UV and insecticidal activity. Journal of Invertebrate Pathology, 67:120-124.

Paterson, J. R., García-Bellido, D. C., Lee, M. S. Y., Brock, G. A., Jago, J. B., and Edgecombe, G. D. 2011. Acute vision in the giant Cambrian predator Anomalocaris and the origin of compound eyes. Nature, 480:237-240.

Perna, G., Lasalvia, M., Gallo, C., Quartucci, G. and Capozzi, V. 2013. Vibrational characterization of synthetic eumelanin by means of Raman and Surface Enhanced Raman Scattering. The Open Surface Science Journal, 5:1-8.

Plonka, P. M., and Grabacka, M. 2006. Melanin synthesis in microorganisms – biotechnological and medical aspects. Acta Biochemica Polonica, 3:429-443.

Polidori, C., Jorge, A., and Ornosa, C. 2017. Eumelanin and pheomelanin are predominant pigments in bumblebee (Apidae: Bombus) pubescence. PeerJ: DOI 10.7717/peerj.3300.

Prum, R. O. 2006. Anatomy, physics, and evolution of structural colors. In Bird Coloration, Vol. 1 (eds. G. E. Hill and K. J. McGraw), pp. 295-353. Cambridge, MA: Harvard University Press.

R Development Core Team. 2007. R: a language for data analysis and graphics. Vienna, AUT. http://cran.r-project.org/.

Ramnani, P., R. Singh, and R. Gupta. (2005). Kerationlytic potential of Bacillus licheniformis RG1: structural and biochemical mechanism of feather degradation. Canadian Journal of Microbiology 51:191-196.

Raposo, G. and M. S. Marks, (2002). The dark side of ‐related organelles: specialization of the endocytic pathway for melanosome biogenesis. Traffic 3:237-248.

Raposo, G., Tenza, D., Murphy, D. M., Berson, J. F., and Marks, M. S. 2001. Distinct protein sorting and localization to premelanosomes, melanosomes, and in pigmented melanocyte cells. Journal of , 152:809-823.

Raymond, P. E. 1922. A trilobite retaining color-markings. American Journal of Science, 4:461-464.

139 Riley, P. A. 1997. Melanin. The International Journal of Biochemistry & Cell Biology, 29:1235-1239.

Robertson, J. 1986. Amorphous carbon. Advances in Physics, 35, 317-374.

Rodríguez-Rojas, A., Mena, A., Martín, S., Borrell, N., Oliver, A., and Blásquez, J. 2009. Inactivation of the hmgA gene of Pseudomonas aeruginosa leads to pyomelanin hyperproduction, stress resistance, and increased persistence in chronic lung infection. Microbiology, 155:1050-1057.

Rosas, A. L., and Casadevall, A. 1997. Melanization affects susceptibility of Cryptococcus neoformans to heat and cold. FEMS Microbiology Letters, 153:265-272.

Roulin, A., and Ducrest, A. 2011. Association between melanism, physiology and behavior: a role for the melanocortin system. European Journal of Pharmacology, 660:226-233.

Różanowska, M., Sarna, T., Land, E. J., and Truscott, T. G. 1999. Free radical scavenging properties of melanin interaction of eu- and pheo-melanin models with reducing and oxidizing radicals. Free Radical Biology & Medicine, 26:518- 525.

Ruiz-de-Castañeda, R., E. H. Burtt, S. Gonzalez-Braojos, and J. Moreno. (2012). Bacterial degradability of an intrafeather unmelanized ornament: a role for feather-degrading bacteria in sexual selection? Biological Journal of the Linnean Society 105:409-419.

Ruiz-Rodriguez, M., E. Valdivia, J. J. Soler, M. Martin-Vivaldi, A. M. Martin-Platero, and M. Martinez-Bueno. (2009). Symbiotic bacteria living in the hoopoe’s prevent feather degradation. The Journal of Experimental Biology 212:3621-3626.

Santos, R. M., A. A. Firmino, C. M. De Sa, and C. R. Felix. (1996). Keratinolytic activity of Aspergillus fumigatus fresenius. Current Microbiology 33:364-370.

Sanz, J. L., Chiappe, L. M., Pérez-Moreno, B. P., Buscalioni, A. D., Maratalla, J. J., Ortega, F. and Poyato-Ariza, F. J. 1996. An Early Cretaceous bird from and its implications for the evolution of avian flight. Nature, 382:442-445.

Schoenemann, B., Clarkson, E. N. K., and Ryck, U. 2014. Colour patterns in Devonian trilobites. The Open Geology Journal, 8:113-117.

Schopf, J. W. 1993. Microfossils of the Early Archean Apex Chert: new evidence of the antiquity of life. Science, 260:640-646.

140 Schopf, J. W., Kitajima, K., Spicuzza, M. J., Kudryavtsev, A. B., and Valley, J. W. 2017. SIMS analyses of microfossils document their taxon-correlated carbon isotope compositions. PNAS, 115:53-58.

Schultz T. D., and Rankin, M. A. 1985. The ultrastructure of the epicuticular interference reflectors of tiger beetles (Cicindela). Journal of Experimental Biology, 117:87- 110.

Schweitzer, M. H. 2011. Soft tissue preservation in terrestrial Mesozoic vertebrates. Annual Review of Earth and Planetary Sciences 39:187-216.

Shawkey, M. D., Estes, A. M., Siefferman, L. M., and Hill, G. E. 2003. Nanostructure predicts intraspecific variation in ultraviolet-blue plumage colour. Proceedings of the Royal Society B, 270:1455-1460.

Shawkey, M. D., and Hill, G. E. 2006. Significance of a basal melanin layer to production of non-iridescent structural plumage color: evidence from an amelanotic Steller's jay (Cyanocitta stelleri). Journal of Experimental Biology, 209:1245-1250.

Shawkey, M. D., Mills, K. L, Dale, C., and Hill, G. E. 2005. Microbial diversity of wild bird feathers revealed through culture-based and culture-independent techniques. Microbial Ecology, 50:40-47.

Shawkey, M. D., Morehouse, N. I., and Vukusic, P. 2009. A protean palette: color materials and mixing in birds and butterflies. Journal of the Royal Society Interface, 6:S221-S231.

Shawkey, M. D., Pillai, S. R., and Hill, G. E. 2003. Chemical warfare? Effects of uropygial oil on feather‐degrading bacteria. Journal of Avian Biology, 34:345- 349.

Shawkey, M. D., Pillai, S. R., Hill, G. E., Siefferman, L. M., and Roberts, S. 2007. Bacteria as an agent for change in structural plumage color: correlational and experimental evidence. The American Naturalist, 169:S112-S121.

Shiraki, T., Kojima, D., and Fukada, Y. 2010. Light-induced body color change in developing zebrafish. Photochemical & Photobiological Sciences, 9:1498-1504.

Simon, J. D., Hong, L., and Peles, D. N. 2008. Insights into melanosomes and melanin from some interesting spatial and temporal properties. Journal of Physical Chemistry B, 112:13201-13217.

Singh, A. V., Vyas, V., Patil, R., Sharma, V., Scopelliti, P. E., Bongiorno, G., Podestà, A., Lenardi, C., Gade, W. N., and Milani, P. 2011. Quantitative characterization of the influence of the nanoscale morphology of nanostructured surfaces on

141 bacterial adhesion and biofilm formation. PLoS One, 6: DOI:10.1371/journal.pone.0025029.

Smith, K. R., Cadena, V., Endler, J. A., Kearney, M. R., Porter, W. P., and Stuart-Fox, D. 2016b. Color change for thermoregulation versus camouflage in free-ranging lizards. The American Naturalist, 188:668-678.

Smith, K. R., Cadena, V., Endler, J. A., Porter, W. P., Kearney, M. R., and Stuart-Fox, D. 2016a. Colour change on different body regions provides thermal and signalling advantages in bearded dragon lizards. Proceedings of the Royal Society B, 283: DOI/10.1098/rspb.2016.0626.

Smithwick, F. M., Nicholls, R., Cuthill, I. C., Vinther, J. 2017. Countershading and stripes in the theropod dinosaur Sinosauropteryx reveal heterogeneous habitats in the Early Cretaceous Jehol Biota. Current Biology, 27:3337-3343.

Smyth, J. R., Porter, J. W. and Bohren, B. B. 1951. A study of pigments from red, brown, and buff feathers and hair. Physiological Zoology, 24:205-216.

Söderhäll, K., and Cerenius, L. 1998. Role of the prophenoloxidase-activating system in invertebrate immunity. Current Opinion in Immunology, 10:23-28.

Solano, F. 2014. Melanins: skin pigments and much more – types, structural models, biological functions, and formation routes. New Journal of Science, 2014: DOI:10.1155/2014/498276.

Speiser, D. I., DeMartini, D. G., and Oakley, T. H. 2014. The shell-eyes of the chiton Acanthopleura granulata (, Polyplacophora) use melanin as a screening pigment. Journal of Natural History, 48:2899-2911.

Stavenga, D. G., Leertouwer, H. L., Hariyama, T., De Raedt, H. A., and Wilts, B. D. 2012. Sexual of the damselfly Calopteryx japonica caused by a melanin-chitin multilayer in the male wing veins. PloS One, 7: e49743.

Stavenga, D. G., Leertouwer, H. L., Osorio, D. C., & Wilts, B. D. (2015). High refractive index of melanin in shiny occipital feathers of a bird of paradise. Light: Science & Applications, 4(1), e243.

Stohs, S. J., and Bagchi, D. 1995. Oxidative mechanisms in the toxicity of metal ions. Free Radical Biology & Medicine, 18:321-336.

Sugimoto, M. 2002. Morphological color changes in fish: regulation of pigment cell density and morphology. Microscopy Research and Technique, 58:496-503.

Sugumaran, M. 2002. Comparative biochemistry of eumelanogenesis and the protective roles of phenoloxidase and melanin in insects. Pigment Cell Research, 15:2-9.

142

Sulaimon, S. S., and Kitchell, B. E. 2003. The biology of melanocytes. Veterinary Dermatology, 14:57-65.

Suzuki, I., Tada, A., Ollmann, M. M., Barsh, G. S., Lamoreux, M. L., Hearing, V. J., Nordlund, J. J., and Abdel-Malek, Z. A. 1997. Agouti signaling protein inhibits melanogenesis and the response of human melanocytes to alpha-melanotropin. Journal of Investigative Dermatology, 108:838-842.

Tanaka, G., Parker, A. R., Hasegawa, Y., Siveter, D.J., Yamamoto, R., Miyashita, K., Takahashi, Y., Ito, S., Wakamatsu, K., Mukuda, T., Matsuura, M., Tomikawa, K., Furutani, M., Suzuki, K., and Maeda, H. 2014. Mineralized rods and cones suggest colour vision in a 300 Myr-old fossil fish. Nature Communications, 5:1-6.

Tanaka, G., Zhou, B., Zhang, Y., Siveter, D. J., and Parker, A. R. 2017. Rods and cones in an enantiornithine bird eye from the Early Cretaceous Jehol Biota. Heliyon, 3: DOI:10.1016/j.heliyon.2017.e00479.

Teichert, C. 1944. trilobites from Western Australia. Journal of Paleontology, 18:455-463.

Tribl, F., Asan, E., Arzberger, T., Tatschner, T., Langenfield, E., Meyer, H. E., Bringmann, G., Riederer, P., Gerlach, M., and Marcus, K. 2009. Identification of L-ferritin in Neuromelanin granules of the human substantia nigra. Molecular & Cellular Proteomics, 8:1832-1838.

Turick, C. E., Knox, A. S., Becnel, J. M., Ekechukwu, A. A., and Milliken, C. E. 2010. Properties and function of pyomelanin. In Biopolymers (ed. M. Elnashar), pp. 449-472. Rijeka: InTech.

Vinther, J. 2015. A guide to the field of palaeo colour. Bioessays 37:643-656.

Vither, J., Broggs, D. E. G., Prum, R. O. and Saranathan, V. 2008. The colour of fossil feathers. Biology Letters, 4:522-525.

Vinther, J., Briggs. D. E. G., Clarke, J., Mayr, G. and Prum, R. O. 2010. Structural coloration in a fossil feather. Biology Letters, 6:128-131.

Vinther, J., Nicholls, R., Lautenschlager, S., Pittman, M., Kaye, T. G., Rayfield, E., Mayr, G., Cuthill, I. C. 2016. 3D camouflage in an ornithischian dinosaur. Current Biology, 26:2456-2462.

Vitek, N. S., Vinther, J., Schiffbauer, J. D., Briggs, D. E. G. and Prum, R. O. 2013. Exceptional three-dimensional preservation and coloration of an originally iridescent fossil feather from the Middle Messel Oil Shale. Paläontologische Zeitschrift, 87:493-503.

143

Vukusic, P., Wootton, R. J., and Sambles J. R. 2004. Remarkable iridescence in the hindwings of the damselfly Neurobasis chinensis chinensis (Linnaeus) (Zygoptera: Calopterygidae). Proceedings of the Royal Society B, 271:595-601.

Walker, C. A. 1981. New subclass of birds from the Cretaceous of South America. Nature, 292:51-53.

Walsberg, G. E., Campbell, G. S., and King, J. R. 1978. Animal coat color and radiative heat gain: a re-evaluation. Journal of Comparative Physiology, 126:211-222.

Wang, M., Zhou, Z., O’Connor, J. K. and Zelenkov, N. V. 2014a. A new diverse enantiornithine family (Bohaiornithidae fam. nov.) from the Lower Cretaceous of China with information from two new species. Vertebrata PalAsiatica, 52:31-76.

Wang, X., O’Connor, J. K., Zhao, B., Chiappe, L. M., Gao, C. and Cheng, X. 2010. New species of Enantiornithes (Aves: Ornithoraces) from the Qiaotou Formation in Northern Hebei, China. Acta Geologica Sinica, 84:247-256.

Wang, X., O’Connor, J. K., Zheng, X., Wang, M., Hu, H. and Zhou, Z. 2014b. Insights into the evolution of rachis dominated tail feathers from a new basal enantiornithine (Aves: Ornithoraces). Biological Journal of the Linnean Society, 113, 805-819.

Wang, Y., Aisen, P., and Casadevall, A. 1995. Cryptococcus neoformans melanin and virulence: mechanism of action. Infection and Immunity, 63:3131-3136.

Wang, Y., and Casadevall, A. 1994b. Decreased susceptibility of melanized Cryptococcus neoformans to UV light. Applied and Environmental Microbiology, 60:3864-3866.

Wang, Y., and Casadevall, A. 1994a. Susceptibility of melanized and nonmelanized Cryptococcus neoformans to nitrogen- and oxygen-derived oxidants. Infection and Immunity, 62:3004-3007.

Wang, Z., Dillon, J., and Gaillard, E. R. 2006. Antioxidant properties of melanin in retinal pigment epithelial cells. Photochemitry and Photobiology, 82:474-479.

Watt, W. B. 1969. Adaptive significance of pigment polymorphisms in Colias butterflies, II. Thermoregulation and photoperiodically controlled melanin variation in Colias eurytheme. Zoology 63:767-774.

Wells, J. W. 1942. Supposed color-markings in trilobites from Ohio. American Journal of Science, 240:710-713.

144 Wenczl, E., Van der Schans, G. P., Roza, L., Kolb, R. M., Timmerman, A. J., Smit, N. P. M., Pavel, S., and Schothorst, A. A. 1998. (Pheo)melanin photosensitizes UVA- induced DNA damage in cultured human melanocytes. The Journal of Investigative Dermatology, 111:678-682.

Williams, J. S. 1933. A color on a new Mississippian trilobite. American Journal of Science, 20:61-64.

Williamson, P. R., Wakamatsu, K., and Ito, S. 1998. Melanin biosynthesis in Cryptococcus neoformans. Journal of Bacteriology, 180:1570-1572.

Whitten, M. M. A., and Coates, C. J. 2017. Re-evaluation of insect melanogenesis research: views from the dark side.

Wogelius, R. A., Manning, P. L., Barden, H. E., Edwards, N. P., Webb, S. M., Sellers, W. I., Taylor, K. G., Larson, P. L., Dodson, P., You, H., Da-qing, L., and Bergmann, U. 2011. Trace metals as biomarkers for eumelanin pigment in the fossil record. Science, 333:1622-1626.

Wright, G. M., and McBurney, K. M. 1992. Changes in the ventral dermis and development of iridophores in the anadromous sea lamprey, Petromyzon marinus, during metamorphosis: an ultrastructural study. Histology and Histopathology, 7:237-250.

Wolf, B. O., and Walsberg, G. E. 2000. The role of plumage in heat transfer processes of birds. American Zoologist, 40:575-584.

Xiao, M., Dhinojwala, A., and Shawkey, M. 2014. Nanostructural basis of -like iridescence in chalcoptera feathers. Optics express, 22:14625-14636.

Xu, X., Zhou, Z., Dudley, R., Mackem, S., Chuong, C., Erickson, G. M. and Varrichio, D. J., 2014. An integrative approach to understanding bird origins. Science, 346: 1253293.

Zecca, L., Shima, T., Stroppolo, A., Goj, C., Battiston, G. A., Gerbasi, R., Sarna, T., and Swartz, H. M. 1996. Interaction of neuromelanin and iron in substantia nigra and other areas of the human brain. Neuroscience, 73:407-415.

Zecca, L., and Swartz, H. M. 1993. Total paramagnetic metals in human substantia nigra and its neuromelanin. Journal of Neural Transmission, 5:203-213.

Zhang, F., Kearns, S. L., Orr, P. J., Benton, M. J., Zhou, Z., Johnson, D., Xu, X., and Wang, X. 2010. Fossilized melanosomes and the colour of Cretaceous dinosaurs and birds. Nature 463:1075-1078.

145 Zhang, F., and Zhou, Z. 2000. A primitive enantiornithine bird and the origin of feathers. Science, 290:1955-1959.

Zhang, Y., O’Connor, J., DI, L., Qingjin, M., Sigurdsen, T. and Chiappe, L. M. 2014. New information on the anatomy of the Chinese Early Cretaceous Bohaiornithidae (Aves: Enantiornithes) from a subadult specimen of Zhouornis hani. PeerJ, 2:1-19.

Zheng, X., O’Connor, J., Huchzermeyer, F., Wang, X., Wang, Y., Wang, M. and Zhou, Z., 2013. Preservation of ovarian follicles reveals early evolution of avian reproductive behavior. Nature, 495:507-511.

Zheng, X., O’Connor, J. K., Wang, X., Pan, Y., Wang, Y., Wang, M., and Zhou, Z., 2017. Exceptional preservation of soft tissue in a new specimen of and its biological implications. National Science Review: doi: 10.1093/nsr/nwx004.

Zhou, Z. 2004. The origin and early evolution of birds: discoveries, disputes, and perspectives from fossil evidence. Naturwissenchaften, 91:455-471.

Zou, Z., Chiappe, L. M. and Zhang, F., 2005. Anatomy of the Early Cretaceous bird Eoenantiornis buhleri (Aves: Enantiornithes) from China. Canadian Journal of Earth Sciences, 42:1331-1338.

Zhou, Z., Clarke, J., and Zhang, F. 2008. Insight into diversity, body size and morphological evolution from the largest Early Cretaceous enantiornithine bird. Journal of Anatomy, 212:565-577.

Zhou, Z. and Zhang, F. 2007. Mesozoic birds of China – a synoptic review. Frontiers of Biology in China, 2:1-14.

Zi, J., Yu, X., Li, Y., Hu, X., Xu, C., Wang, X., and Fu, R. 2003. Coloration strategies in peacock feathers. Proceedings of the National Academy of Sciences, 100:12576- 12578.

Zucca, F. A., Segura-Aguilar, J., Ferrari, E., Muñoz, P., Paris, I., Sulzer, D., Sarna, T., Casella, L., and Zecca, L. 2017. Interactions of iron, dopamine and neuromelanin pathways in brain aging and Parkinson’s disease. Progress in Neurobiology, 155:96-119.

Zughaier, S. M., Ryley, H. C., and Jackson, S. K. 1999. A melanin pigment purified from an epidemic strain of Burkholderia cepacia attenuates monocyte respiratory burst activity by scavenging superoxide anion. Infection and Immunity, 67:908-913.

146

APPENDICES

147 APPENDIX A

CHAPTER 2 SUPPLEMENT

Additional Methods

Prior to experimentation, we calculated absorbance values based on serial dilutions of

Bacillus licheniformis to determine the number of colony-forming units (CFU) that inoculated our feather samples. We incubated 0.01 mL of bacteria in 2 mL of sterile tryptic soy broth (TSB) overnight at 37 °C. From this source tube, we took 2 mL bacteria and mixed it with 2 mL TSB (1:1 solution). The 1:1 bacteria-saline solution was incubated for four hours at 37 °C, which falls within the optimal growth period, and therefore optimal feeding period, for B. licheniformis (Frankena et al. 1985). After four hours, the 1:1 tube was removed from the incubator and mixed thoroughly with a Vortex-

Genie mixer. We created 10 dilutions with two replicates each using a homemade 0.9% saline solution. We made the saline solution using deionized water and a fine sea salt lacking iodide and anti-caking agent, which were mixed thoroughly with heat until dissolved, then autoclaved for 15 minutes at 121°C. The first dilution included 2 mL of the bacterial solution and 2 mL of sterile saline (1:1). 2 mL of this solution was then mixed with 2 mL of saline in a second tube, which generated a 1 saline to 2 bacteria

148 solution (1:2). The process was repeated for the remaining tubes, generating dilutions of

1:1, 1:2, 1:4, 1:8, 1:16, 1:32, 1:64, 1:128, 1:256, and 1:512.

To calculate the absorbance values, 600 µL of each dilution was placed in sterile, untreated 96-well microplates for two replicates (300 µL per well). Each tube was mixed prior to placement within the wells using the Vortex-Genie mixer. Absorbance values were calculated using a 96 well SpectraMax Plus 384 Microplate Reader with a basic endpoint procedure and a wavelength of 530 nm. Absorbance values between the two replicates were averaged.

We used a spread-plating technique to count the bacteria. Twenty sterile plates with tryptic soy agar (TSA) were prepared prior to the experiment. To remain consistent between the absorbance calculations and the plate counting, we took 300 µL from each dilution and placed it in the center of each respective plate. Each dilution had two replicate plates. The solution was spread evenly across the surface of each plate using a metal cell spreader and a rotating disk. We heat-sterilized the spreader and let it cool, then placed it on the and rotated the plate counter-clockwise for one full rotation, then sterilized the spreader again and rotated the plate clockwise for one full rotation. This process was repeated for each plate, then the plates were placed in an incubator for 2.5 hours at 37 °C. The number of CFU were counted for each plate separately by NMJ and JAP under a dissecting microscope. These results were then averaged for each dilution (Table A.1). The number of CFU for the 1:2 dilution was adjusted by dividing the 8224 CFU present in the 1:1 solution by two to account for lack of bacterial growth on one of the two replicate plates.

149 Table A.1. The absorbance values and number of colony-forming units of Bacillus licheniformis per dilution.

Colony- Dilution forming Absorbance units Approx. 1:1 0.908 8224 1:2 4112* 0.52 1:4 2036 0.349 1:8 1298 0.163 1:16 1086 0.109 1:32 830 0.073 1:64 331 0.057 1:128 332 0.057 1:256 2 0.05 1:512 0 0.048 *Denotes adjusted value (original value was 1348 CFU).

150 APPENDIX B

CHAPTER 3 SUPPLEMENT

Table A.2. Observed masses and standardized relative intensities for ToF-SIMS spectra of ions associated with the eumelanin structure (Lindgren et al. 2015) compared between a Sepia officinalis eumelanin standard, eumelanin extracted from a modern mallard (Anas platyrhynchos) feather, Haikouella fossil

Haik5, 6, and 10), and Haikouella matrix (Haik1 and 13) samples.

Theoretical Observed Mass Relative Intensity Ion Mass Sepia Mallard Haik3 Haik5 Haik6 Haik10 Haik13 Sepia Mallard Haik3 Haik5 Haik6 Haik10 Haik13

C4 48.000 48.000 48.000 48.000 48.000 48.000 48.000 48.000 2.074 0.372 -0.082 -0.370 -0.431 -0.723 -0.840

C4H 49.008 49.000 49.000 49.000 49.000 49.000 49.000 49.000 2.077 0.302 -0.081 -0.270 -0.374 -0.766 -0.889

C3N 50.003 50.000 50.000 50.000 50.000 50.000 50.000 50.000 2.127 0.282 -0.137 -0.409 -0.444 -0.654 -0.764

C4HO 65.003 65.000 65.000 65.000 65.000 65.000 65.000 65.000 1.662 0.809 0.087 0.020 -0.386 -0.984 -1.207

C3NO 65.998 66.000 66.000 66.000 66.000 66.000 66.000 66.000 1.880 0.738 -0.042 -0.293 -0.528 -0.795 -0.960

C6 72.000 72.000 72.000 72.000 72.000 72.000 72.000 72.000 1.963 0.542 -0.056 -0.202 -0.484 -0.782 -0.981

C6H 73.008 73.000 73.000 73.000 73.000 73.000 73.000 73.000 1.034 0.532 0.236 1.095 -0.346 -1.176 -1.375

C5N 74.003 74.000 74.000 74.000 74.000 74.000 74.000 74.000 1.683 0.752 -0.097 0.228 -0.395 -0.963 -1.207

C7 84.000 84.000 84.000 84.000 84.000 84.000 84.000 84.000 1.238 1.593 -0.276 -0.326 -0.518 -0.774 -0.938

C7H 85.008 85.000 85.000 85.000 85.000 85.000 85.000 85.000 0.881 1.577 0.188 0.015 -0.431 -0.985 -1.244

C6N 86.003 86.000 86.000 86.000 86.000 86.000 86.000 86.000 1.091 1.632 -0.082 -0.242 -0.505 -0.824 -1.071

C6HN 87.011 87.000 87.000 87.000 87.000 87.000 87.000 87.000 1.014 1.535 0.080 0.019 -0.495 -0.924 -1.229

C6H2N 88.019 88.000 88.000 88.000 88.000 88.000 88.000 88.000 1.171 0.812 0.598 0.393 -0.464 -1.040 -1.471

C2H3NO3 89.011 89.000 89.000 89.000 89.000 89.000 89.000 89.000 0.124 -0.042 0.190 1.931 -0.110 -0.950 -1.142

C5NO 89.998 90.000 90.000 90.000 90.000 90.000 90.000 90.000 1.331 0.751 0.331 0.513 -0.472 -1.055 -1.399

C8 96.000 96.000 96.000 96.000 96.000 96.000 96.000 96.000 0.483 1.295 0.784 0.453 -0.620 -1.115 -1.281

C8H 97.008 97.000 97.000 97.000 97.000 97.000 97.000 97.000 0.498 1.799 0.383 -0.057 -0.662 -0.939 -1.022

C7N 98.003 98.000 98.000 98.000 98.000 98.000 98.000 98.000 1.217 1.608 -0.216 -0.344 -0.565 -0.802 -0.897

C9 108.000 108.000 108.000 108.000 108.000 108.000 108.000 108.000 1.596 1.222 -0.245 -0.297 -0.535 -0.794 -0.946

C9H 109.008 109.000 109.000 109.000 109.000 109.000 109.000 109.000 1.634 1.161 -0.171 -0.289 -0.602 -0.813 -0.920

C8N 110.003 110.000 110.000 110.000 110.000 110.000 110.000 110.000 1.154 1.686 -0.283 -0.413 -0.554 -0.760 -0.831

C8HN 111.011 111.000 111.000 111.000 111.000 111.000 111.000 111.000 1.053 1.765 -0.252 -0.476 -0.558 -0.735 -0.797

C8H2N 112.019 112.000 112.000 112.000 112.000 112.000 112.000 112.000 0.940 1.799 -0.058 -0.473 -0.612 -0.741 -0.854

C4H3NO3 113.011 113.000 113.000 113.000 113.000 113.000 113.000 113.000 0.726 1.953 -0.257 -0.336 -0.554 -0.736 -0.795

C7NO 113.998 114.000 114.000 114.000 114.000 114.000 114.000 114.000 1.073 1.719 -0.201 -0.349 -0.571 -0.775 -0.897

C10 120.000 120.000 120.000 120.000 120.000 120.000 120.000 120.000 1.329 0.821 0.284 0.414 -0.330 -1.139 -1.379

C10H 121.008 121.000 121.000 121.000 121.000 121.000 121.000 121.000 1.705 0.320 0.348 0.268 -0.337 -1.030 -1.274

C9N 122.003 122.000 122.000 122.000 122.000 122.000 122.000 122.000 1.223 1.526 -0.074 -0.203 -0.595 -0.860 -1.017

C11 132.000 132.000 132.000 132.000 132.000 132.000 132.000 132.000 1.518 0.891 -0.036 0.363 -0.509 -1.000 -1.228

C11H 133.008 133.000 133.000 133.000 133.000 133.000 133.000 133.000 1.585 0.666 -0.101 0.566 -0.511 -0.998 -1.207

C10N 134.003 134.000 134.000 134.000 134.000 134.000 134.000 134.000 1.648 0.888 -0.116 0.179 -0.513 -0.963 -1.122

C10HN 135.011 135.000 135.000 135.000 135.000 135.000 135.000 135.000 0.869 0.292 0.501 1.129 -0.103 -1.259 -1.428

C10H2N 136.019 136.000 136.000 136.000 136.000 136.000 136.000 136.000 1.376 1.135 0.147 0.082 -0.490 -1.032 -1.218

C6H3NO3 137.011 137.000 137.000 137.000 137.000 137.000 137.000 137.000 0.602 0.771 0.659 0.893 -0.187 -1.247 -1.491

C9NO 137.998 138.000 138.000 138.000 138.000 138.000 138.000 138.000 1.539 1.256 -0.149 -0.253 -0.624 -0.816 -0.954

C12 144.000 144.000 144.000 144.000 144.000 144.000 144.000 144.000 1.806 0.952 -0.221 -0.314 -0.592 -0.784 -0.848

C12H 145.008 145.000 145.000 145.000 145.000 145.000 145.000 145.000 1.690 0.945 -0.104 -0.009 -0.569 -0.915 -1.038 C11N 146.003 146.000 146.000 146.000 146.000 146.000 146.000 146.000 1.731 0.962 -0.106 -0.162 -0.592 -0.846 -0.986

151

Figure A.1. Raman spectra from each of the Haikouella samples.

152

Figure A.2: Raman spectra of pyrite, hematite, goethite, and carbon-based materials.

Figure A.3. Raman spectra of LSJC and Hongshanornis samples.

153 154 155 156

Figure A.4. Negative polarity ToF-SIMS results for three Haikouella fossil samples (5, 6, 10), two

Haikouella matrix samples (3, 13), and eumelanin samples from Sepia officinalis ink and mallard (Anas platyrhynchos) feathers. Blue “+” indicate theoretical masses for eumelanin presented by Lindgren et al.

(2015).

157 158 159 160

Figure A.5. Positive polarity ToF-SIMS results for three Haikouella fossil samples (5, 6, 10), two

Haikouella matrix samples (3, 13), and eumelanin samples from Sepia officinalis ink and mallard (Anas platyrhynchos) feathers. PDMS = Polydimethylsiloxane, a possible plastic contaminant.

161

Figure A.6. Principal component plot of ToF-SIMS data for modern eumelanin samples (Sepia and mallard), three Haikouella integument samples (Haik 5, 6, and 10), and two Haikouella matrix samples

(Haik 3 and 13). Modern samples are indicated by blue dots and fossil samples by black dots. PC1 accounted for 85.5% of the variance while PC2 accounted for 8.1%.

162 APPENDIX C

CHAPTER 4 SUPPLEMENT

Table A.3. Measurements of the new bohaiornithid specimen (CUGB P1202). Measurements are in millimeters.

Skull length 26.51 Skull height 13.56 Antorbital length 8.21 Antorbital height 3.00 Pygostyle length 13.06 Coracoid length (left) 19.35 Coracoid distal width (left) 10.06 Furcula length 26.22 Furcula proximal width 11.40 Furcula, hypocleidium length 4.16 Humerus length (right) 33.36 Humerus, midshaft width (left) 3.18 Ulna length (left) 27.56 Ulna, midshaft width (left) 2.73 Radius length (left) 23.88 Radius, midshaft width (left) 1.80 Metacarpal I length (right) 3.37 Metacarpal II length (right) 11.76 Metacarpal III length (right) 11.89 First phalanx of manual digit I length (right) 6.35 First phalanx of manual digit II length (left) 7.16 Second phalanx of manual digit II length (left) 4.84 Femur length > 25.65 Tibiotarsus length (right) 35.61 Tarsometatarsus length (right) 17.09 Pedal digit I-1 length (right) 4.74 Pedal digit I ungual length (counterpart, left) 10.02 Pedal digit II-1 length 7.16 Pedal digit II-2 length (left) 5.19 Pedal digit II ungual length (counterpart, right) 10.22 Pedal digit III-1 length (counterpart, right) 9.61 Pedal digit III-2 length (counterpart, left) 6.64 Pedal digit III-3 length (counterpart, right) 7.72 Pedal digit III ungual length (counterpart, right) 10.27 Pedal digit IV ungual length (counterpart, right) 6.88 Crown feather (counterpart) 19.54 Alula (counterpart right) 12.80 First primary remige (counterpart, right) 129.82 Elongate rectrice (counterpart, right) > 86.82 Elongate rectrice total width (counterpart, right) 2.27 Elongate rectrice midshaft rachis width (counterpart, right) 1.21 Short rectrice (counterpart, right) 20.02 Body contour feather (counterpart, right) 32.78

163 A B 6 2 i1 6 3h1 g 6 1 9 6 5

6 4 70

6 8 Figure A.7. Melanosome sampling sites on CUGB P1202. A, primary slab; and B, counter slab. Solid lines represent samples taken from feathers and dashed lines represent matrix samples.

164

Fig A.8. Quadratic canonical discriminate analysis that includes raw melanosome measurements for CUGB

P1202 as well as measurements rescaled up to account for proposed taphonomic shrinkage of 10% (62,

10%) to 20% (62, 20%). Open circles and numbers refer to average measurements for CUGB P1202 samples and sampling sites from Fig. S1. Colored points refer to extant feather melanosome morphologies

(average values) from the Li et al. (2012) database. Circles represent the 95% confidence interval.

165 Table A.4. Eigenvalues, canonical likelihood scores, and standardized scoring coefficients for the quadratic canonical discriminant analysis estimating color in CUGB P1202 samples (Figure A.7; Table A.5) and extant bird feathers (from Li et al. 2012).

% Aspect Canonical Canonical Length Diameter Aspect Eigenvalue Variation Length Diameter Ratio Axis Correlation CV CV Ratio Explained Skew Canonical 1 1.89 61.94 0.81 0.36 -0.04 -0.01 0.28 0.83 0.07 Canonical 2 0.77 25.23 0.66 0.80 0.52 0.09 -0.14 -0.50 -0.29 Canonical 3 0.23 7.56 0.43 0.44 0.01 -0.16 0.76 -0.34 0.80 Canonical 4 0.16 5.27 0.37 -1.16 -0.44 1.74 0.35 1.16 0.19

Table A.5. Number of complete melanosomes measured (n) and average measurements of melanosomes preserved in each CUGB P1202 sample (Figure A.7) plus or minus standard deviation.

Average Length Average Width Average Aspect Sample n (nm) (nm) Ratio Crown (Sample 62) 21 1159.23 ± 325.65 168.41 ± 41.17 7.13 ± 2.09 Nape (Sample 63) 21 1721.36 ± 345.70 330.68 ± 99.94 5.74 ± 2.24 Body Contour (Sample 64) 4 1628.83 ± 140.22 243.91 ± 63.48 7.11 ± 2.30

166 A B c

D E F

Figure A.9. Raman spectra and SEM images compared to CUGB P1202 feather samples. A, Raman spectra generated by extant eumelanin extracted from a brown Cooper’s Hawk feather compared to spectra from

CUGB P1202 and carbon black to control for the potential effects of carbon on the surface of the fossil; B,

Raman spectra produced by keratin from an unpigmented primary remige from the Sulfur-crested

Cockatoo; C, Raman spectra generated by the keratinolytic bacterium Bacillus licheniformis (strain 138B); and D-F, Raman spectra (D) and SEM images (E and F) for matrix samples from CUGB P1202. For D, numbers and letters in legend indicate sampling numbers for CUGB P1202.

167

Figure A.10. Comparison of Raman spectra generated by fossil plants and CUGB P1202. A, Stigmaria

(CMNH P-21773) from the Carboniferous Joggins Formation, Joggins, Nova Scotia; B, Platanus (PTRM

#20639) from the Late Cretaceous Hell Creek Formation, Mud Buttes, North Dakota; C, Raman spectra for

Stigmaria, Platanus, and CUGB P1202. Arrows indicate sampling locations. Scale bars = 1 cm.

168 Table A.6. Peak fitting for Raman spectra generated by melanosomes in extant feathers and fossil feathers preserved in CUGB P1202 (top) and Platanus, Stigmaria, carbon black, feather keratin from an unpigmented primary remige from the Sulfur-crested Cockatoo, and the keratinolytic bacterium Bacillus licheniformis (strain 138B).

Raman Peaks Sample Color 1 2 3

Chicken Black 1378.52 1578.80 Red-winged Blackbird Black 1194.45 1385.90 1584.22 Mallard Iridescent 1166.35 1372.28 1576.94 Wild Turkey Iridescent 1157.54 1374.42 1577.92 Cooper's Hawk Brown 1166.51 1373.92 1577.06 House Wren Brown 1128.57 1384.21 1574.98 Black Bird Sample 62 Fossil 1143.75 1363.74 1574.59 Black Bird Sample 63 Fossil 1132.47 1360.58 1572.33 Black Bird Sample 64 Fossil 1142.33 1353.83 1571.23 Black Bird Sample 65 Fossil 1354.78 1566.72

Sample Raman Peaks Platanus 1151.30 1351.69 1578.68 Stigmaria 1369.49 1584.15 Carbon Black 1360.57 1590.91 Feather Keratin 514.12 1003.42 1127.33 1146.01 1206.87 1244.22 1454.63 1665.68 Bacillus licheniformis 1006.05 1543.03 1665.28 2028.89 2169.47 2302.65

169 APPENDIX D

CHAPTER 5 SUPPLEMENT

Supplementary Methods

Modern sampling was focused on melanized integument from agnathans, fish

(including Osteichthyes and Chondrychthes), amphibians, and lepidosaurs. Melanic color in these animals is limited to brown, black, and grey integument, as other colors, including iridescence, are produced by other chromatophores or differing arrangements of chromatophores in the skin. All modern samples were cut from the dorsal surface of the animal, although sampling location on the animal’s body varied depending on the type or degree of melanization. Perch, menhaden, hagfish, skate, stingray, dogfish, catfish, sea robin, sea lamprey, mudpuppy, and grass frog samples were preserved in

Carosafe and acquired from Carolina Biological Supply Company. Hagfish, and congo eel samples were acquired from Ward’s Science and preserved in formalin. and anole samples, which were frozen prior to sampling, were provided by the Niewiarowski

Lab at the University of Akron. The Niewiarowski Lab also provided fresh eastern newt and spotted salamander carcasses collected from The University of Akron Bath Nature

Preserve Field Station near Bath, Ohio. Both carcasses were in excellent condition and showed no signs of decay. Additional lepidosaur samples were acquired from the

170 Smithsonian Institution National Museum of Natural History. Squid, walleye, haddock,

Atlantic , swordfish, Lake Smelt, mackeral, and yellowtail snapper were purchased fresh from meat counters at local grocery stores in the Akron and Canton,

Ohio, area. All other fish and amphibian samples were acquired from The University of

Akron biological collections and were preserved in formaldehyde. These differing preservational methods did not appear to affect melanosome size or shape, which remained consistent across all preservation styles.

We also examined skin samples from red and green lepidosaurs to show that melanosomes in samples colored by other pigments or chromatophores were similar in size and shape to those in skin samples in which the dominant pigment type is melanin

(Figure A.16; Table A.9). Melanosome morphology did not deviate from the patterns observed in melanic skin samples. Melanosomes from darkly-colored squid skin also exhibited small, rounded morphologies not unlike those of basal vertebrate skin samples

(Figure A.16; Table A.9). Red and green lepidosaur and squid samples were not included in the ANOVA analysis.

To test for the presence of elongate melanosomes in modern bird and mammal skin, we sampled black foot scales and bare skin from the black eye rings of a grey catbird and rose-breasted grosbeak, black foot scales from a cedar waxwing, and black skin from the nose and foot of a groundhog, nose of a shrew, and wing of a bat. The bat was acquired from Ward’s Science while the other mammal and bird samples were from window-killed or vehicle-struck corpses that showed no signs of decay. These corpses were frozen after collection.

171 We also removed an eye from the lamprey carcass to test for the presence of elongate melanosomes in the retinal pigment epithelium, as has been documented in other basal vertebrates (Burgoyne et al. 2015; Clements et al. 2016). The eye was fixed in 2% glutaraldehyde overnight prior to being embedded in Epon using the same procedure as the skin samples.

Supplementary Results

We interpret the microbodies preserved in the fossil integument of amphibians and agnathans from the Jiulongshan and Tiaojishan Formations as melanosomes. We base this interpretation on similarities in size and shape to melanosomes previously described in the fossil record (Li et al 2014) and chemical and morphological similarities to extant melanosomes. Morphological variation of microbodies preserved in fossil integument remains consistent with melanosome diversity in modern taxa. Melanosomes preserved in the skin of basal amniote and anamniote fossils consistently retain the low morphological diversity of their modern counterparts. Similarly, melanosomes preserved in the feathers of fossil pennaraptoran dinosaurs are consistently diverse in morphology, as in modern bird feathers. These patterns would not be expected if the preserved microbodies were microbial in origin, as bacteria occur in a wide variety of morphologies, which would be expected to preserve across vertebrate taxa rather than remain taxon-specific (i.e. only small, oblate microbodies are preserved in fossil lampreys, amphibians, and non-mammalian, non-pennaraptoran archosaurs while diverse microbodies are preserved in pennaraptoran dinosaurs and mammals). Additionally,

172 preserved microbodies and potential Raman signals for eumelanin are limited to the fossilized integument and are not present in the matrix samples. The microbodies do not show evidence of mitotic fissures, as would be expected for bacterial cells, nor is there other evidence of a preserved biofilm, such as hyphae (e.g., Borkow and Babcock 2003).

Raman spectroscopy can be influenced by contaminants, including carbon present on the surface of compression fossils such as these, which can have similar spectra to eumelanin (Chapter 4). However, despite the deep black carbon compression-type preservation of LL01, it lacks evidence of preserved melanosomes and a Raman signal for carbon or eumelanin. While not as well-preserved, IMF similarly lacks melanosome preservation and Raman peaks for carbon or eumelanin (Figure S3). Given this strict relationship between microbody preservation and a Raman signal for eumelanin, we interpret our Raman results as evidence of the presence of eumelanin rather than carbon contamination. We did not find evidence for the preservation of pheomelanin in any of the fossil specimens, although pheomelanin may be less prevalent in anamniotes and basal amniotes than in mammals and birds. For example, only a few studies have shown evidence for pheomelanin in modern frogs (Wolnicka-Blubisz et al 2012) and turtles

(Roulin et al. 2013) using electron paramagnetic resonance (EPR) and high-performance liquid chromatography (HPLC). We have been unsuccessful in detecting pheomelanin in modern vertebrate samples using 532 nm and 785 nm Raman lasers. However, Raman peaks for pheomelanin have been characterized in synthetic samples (Galván et al. 2013) and modern feathers (Galván et al. 2017) using Raman spectroscopy, which are not present in Raman spectra for the fossil samples presented here.

173

Figure A.11. Sampling maps of salamander IMS from the Jiulongshan Formation, preserved in part (A) and counterpart (C), and corresponding SEM images of the preserved integument (B and D). Arrows point to preserved melanosomes. Scale bars = 1 µm.

174

Figure A.12. Sampling maps for salamanders of the Tiaojishan Formation, including LSJA (A), LSJB (C), and LSJUS (E), and corresponding SEM images of their integument (B, D, F). Arrows point to preserved melanosomes. Scale bars = 1 µm.

175

Figure A.13. Melanosomes preserved in undescribed lamprey LL02 (A) and fossil samples that were negative for melanosome and melanin preservation (B-E). A) SEM image of melanosomes preserved in

LL02; B) SEM image of a sample removed from the buccal opening of lamprey LL01; C) frog IMF and an example SEM image of its integument (D). E) Raman spectra of LL01 and IMF compared to modern eumelanin spectra from the sea lamprey Petromyzon and the common mudpuppy Necturus. Arrow points to preserved melanosomes in (A). Scale bars = 1 µm.

176

Figure A.14. Raman spectra from the matrix of each fossil sampled.

177

Figure A.15. A comparison of extant melanosome morphology from agnathan, fish, and amphibian skin by integument color. A) sea lamprey. B) hagfish. C) mackerel. D) perch. E) skate. F) common mudpuppy. G) grass frog. H) red-backed salamander. Scale bars = 1 µm.

178

Figure A.16. SEM images of melanosomes from additional lepidosaur and squid skin samples. A, green- colored skin from the dull day gecko; B, red-colored skin from the Tokay gecko; C and D, darkly-colored squid skin. Scale bars = 1 µm.

179

Figure A.17. SEM images of eye and ocelli melanosomes in the sea lamprey Petromyzon marinus and amphioxus. A and B, retinal pigment epithelium melanosomes in the eye of a sea lamprey (Petromyzon marinus), including A, sub-spherical melanosome morphologies, and B, elongate melanosome morphologies. C, sub-spherical melanosomes in amphioxus ocelli. Scale bars = 1 µm.

180

Figure A.18. Width of melanosomes in skin, feathers, and hair of extant lampreys, amphibians, lepidosaurs, testudines, and archosaurs, birds, and mammals. Black, brown, and gray colors are representative of the color of the integument. Agnathan skin was not analyzed by color due to small sample size. Iridescent and penguin-type avian feathers are represented by purple and blue, respectively. Similar boxplots are represented by the same letter (v, w, x, y, z). Data for feathers, hairs, lepidosaurs, testudines, and crocodylomorphs adapted from Li et al. (2014) with additional input from the current study (lepidosaurs).

181 Table A.7. Peak fitting for fossil and extant Raman Spectra.

Sample Peak Fossil IMS1 1374.81 1576.67 IMS2 1374.85 1577.42 IMS3 1367.56 1576.22 LSJA2 1366.74 1575.47 LSJB1 1380.49 1582.29 LSJUS 1369.67 1575.94 LL02 1360.85 1569.98 Extant Necturus 1382.20 1577.88 Petromyzon 1360.23 1583.45

Table A.8. Average melanosome measurements for extant anamniotes, lepidosaurs, testudines, and crocodylomorphs.

Mean Length Mean Aspect Color and Taxon (nm) Ratio Agnatha 514.56 ± 32.34 1.31 ± 0.02 Black Fish 508.33 ± 19.71 1.28 ± 0.02 Brown Fish 513.94 ± 22.58 1.27 ± 0.04 Grey Fish 556.60 ± 23.30 1.44 ± 0.07 Black Amphibian 688.99 ± 52.73 1.37 ± 0.04 Brown Amphibian 477.42 ± 37.73 1.29 ± 0.04 Grey Amphibian 489.05 ± 33.37 1.25 ± 0.03

*Black Lepidosaur, Testudine, Crocodile 527.65 ± 23.38 1.62 ± 0.08

*Brown Lepidosaur, Testudine, Crocodile 546.14 ± 17.01 1.63 ± 0.07

*Grey Lepidosaur, Testudine, Crocodile 586.72 ± 32.86 1.96 ± 0.07 *Includes measurements from the present study and Li et al. (2014)

182 Table A.9. Means for length, width, and aspect ratio of melanosomes in the integument of fossil and extant squid, agnathans, fish, amphibians, lepidosaurs, mammals, and birds sampled in this study.

Length Width Aspect Species Common Name Color Integument (nm) (nm) Ratio Mesomyzon (LL02) N/A Fossil Skin 542.21 375.32 1.48 IMSJ N/A Fossil Skin 550.23 326.17 1.71 LSJA N/A Fossil Skin 541.57 361.07 1.52 LSJB N/A Fossil Skin 409.01 307.10 1.35 LSJUS N/A Fossil Skin 778.65 470.34 1.69 Squid Fossil Skin 533.22 432.66 1.24 Petromyzon marinus Sea Lamprey Black Skin 546.90 426.26 1.29 Myxine Hagfish Brown Skin 482.23 368.62 1.32 Skate Black Skin 526.63 376.03 1.42 Skate Grey Skin 615.97 388.84 1.61 Stingray Grey Skin 525.57 382.53 1.38 Stingray Black Skin 495.91 395.69 1.25 Dogfish Grey Skin 570.94 385.38 1.49 Sander vitreus Walleye Black Skin 477.12 395.49 1.22 Melanogrammus aeglefinus Haddock Black Skin 442.79 358.39 1.25 Salmo salar Atlantic Salmon Black Skin 469.07 367.65 1.29 Xiphias Swordfish Black Skin 419.28 349.57 1.20 Lake Smelt Black Skin 517.80 417.85 1.25 Spiny Boxfish Brown Skin 541.84 418.36 1.32 Mackerel Black Skin 567.63 443.93 1.29 Ocyurus chrysurus Yellowtail Snapper Brown Skin 469.23 391.78 1.20 Catfish Grey Skin 513.92 413.50 1.26 Perch Brown Skin 530.74 415.26 1.29 Sea Robin Black Skin 534.35 431.78 1.24 Menhaden Black Skin 632.70 466.97 1.36 Anaxyrus terrestris Southern Toad Brown Skin 496.52 352.31 1.42 Anaxyrus americanus American Toad Brown Skin 393.18 309.96 1.28 Anaxyrus woodhousii Woodhouse's Toad Black Skin 556.42 462.65 1.20 Lithobates heckscheri River Frog Brown Skin 582.00 476.28 1.23 Lithobates forreri Grass Frog Brown Skin 458.68 332.75 1.38 Small-mouth Ambystoma texanum Salamander Grey Skin 455.68 360.27 1.27 Ambystoma tigrinum Tiger Salamander Black Skin 722.46 534.57 1.36 Ambystoma maculatum Spotted (metamorph) Salamander Black Skin 595.22 419.39 1.42 Common Necturus maculosus Mudpuppy Black Skin 660.29 532.27 1.25 Red-backed Plethodon cinereus Salamander Grey Skin 522.42 428.45 1.22 Four-toed Hemidactylium scutatum Salamander Brown Skin 358.13 305.55 1.19 Siren lacertina Greater Siren Black Skin 981.03 657.04 1.52

183 Long-tailed Eurycea longicauda Salamander Brown Skin 576.00 466.47 1.24 Amphiuma tridactylum Congo Eel Black Skin 635.39 437.17 1.45 Notophthalmus viridescens Eastern Newt Black Skin 672.10 496.76 1.36 Common Wall Tarentola mauritanica Gecko Black Skin 459.14 320.12 1.45 Marbled Gecko Black Skin 446.57 353.43 1.27 Flying Gecko Black Skin 554.12 407.26 1.36 dubia Dull Day Gecko Green Skin 480.93 383.87 1.26 Gekko gecko Tokay Gecko Green Skin 465.23 321.85 1.48 Gekko gecko Tokay Gecko Red Skin 587.26 462.56 1.28 Phelsuma lineata Lined Day Gecko Red Skin 552.96 459.17 1.22 Anole Green Skin 615.72 453.75 1.37 Yellow-bellied Sea Pelamis platurus Brown Skin 510.48 422.42 1.22 Northern Water Nerodia sipedon sipedon Snake Black Skin 623.05 449.75 1.41 Picado's Jumping Bothrops picadoi Pitviper Brown Skin 520.19 316.15 1.66 Scaphiodontophis Common venustissimus Neckband Snake Black Skin 745.94 584.79 1.29 Eastern Hognose Heterodon platirhinos Snake Black Skin 468.46 361.69 1.32 Boa constrictor Boa Constrictor Black Skin 358.44 275.63 1.32 Slender Glass Ophisaurus attenuatus Lizard Black Skin 358.37 280.28 1.29 Jackson's Chamaeleo jacksonii Skin 603.85 418.87 1.46 Spiny Softshell Apalone spinifera Turtle Brown Skin 495.85 371.92 1.34 Northern Map Graptemys geographica Turtle Brown Skin 586.74 405.53 1.46 Common Box Terrapene carolina Turtle Brown Skin 398.18 311.79 1.31 Clemmys guttata Spotted Turtle Black Skin 590.03 431.33 1.38 Common Musk Sternotherus odoratus Turtle Brown Skin 519.62 396.69 1.32 Common Snapping Chelydra serpentina Turtle Brown Skin 479.89 341.66 1.44 Bat Black Skin 459.11 229.15 2.05 Shrew Black Skin 581.15 305.08 1.93 Marmota monax Groundhog Black Skin 657.09 291.69 2.31 Dumetella carolinensis Grey Catbird Black Skin 721.86 316.84 2.35 Rose-breasted Pheucticus ludovicianus Grosbeak Black Skin 598.42 250.97 2.42 Dumetella carolinensis Grey Catbird Black Scale 525.678 256.544 2.14 Rose-breasted Pheucticus ludovicianus Grosbeak Black Scale 676.15 304.20 2.23 Bombycilla cedrorum Cedar Waxwing Black Scale 653.86 378.47 1.70

184

Table A.10. Summary of melanosome measurements from fossil taxa used to calculate melanosome sizes for Figure 5.5. Fossil pennaraptoran dinosaurs for which color has not been assessed are not included. All measurements are as-preserved and do not account for proposed taphonomic changes to three- dimensionally-preserved melanosomes. *measurements as reported in Li et al. 2014. +measured using

ImageJ from published SEM images, as specific melanosome size and shape measurements were not reported or were not reported for each predicted color.

Integument Predicted Higher Taxon Species (Specimen) Reference Length Width Ratio

Agnatha Mesomyzon/LL02 Skin N/A This study 542.2 375.3 1.5

Mayomyzon piechoensis Gabbott et al. Agnatha (ROMV56800b) Skin N/A 2016+ 695.4 488.0 1.4 Amphibia IMS Skin N/A This study 550.2 326.2 1.7 Amphibia LSJA Skin N/A This study 541.6 361.1 1.5 Amphibia LSJB Skin N/A This study 409.0 307.1 1.4 Amphibia LSJUS Skin N/A This study 778.7 470.3 1.7

Pelobates (PW 2005- Colleary et al. Amphibia 5034-LS_GDKE) Skin N/A 2015+ 577.4 415.7 1.4

Undescribed frog (MU Colleary et al. Amphibia 41-13) Skin N/A 2015+ 793.4 505.8 1.6

Paleobatrachus (SMF- Colleary et al. Amphibia ME 11390a) Skin N/A 2015+ 676.3 404.1 1.7

Undescribed frog McNamara et Amphibia (MNCN 63663) Skin N/A al. 2016+ 1145.6 753.3 1.5

Xianglong zhaoi (PMOL Lepidosaur 000666) Skin N/A Li et al. 2014 644.6 358.0 1.9

Yabeinosaurus sp. Lepidosaur (PKUP V1059) Skin N/A Li et al. 2014 675.4 387.4 1.8

Undescribed mosasaur Lindgren et al. Lepidosaur (SMU 76532) Skin N/A 2014 504.9 323.0 1.6

Undescribed turtle Testudines (PKUP V1070) Skin N/A Li et al. 2014 599.4 401.4 1.5

Undescribed turtle Lindgren et al. Testudines (FUM-N_1450) Skin N/A 2014 812.6 467.8 1.7

Tasbacka danica Lindgren et al. Testudines (MHM-K2) Skin N/A 2017 816.4 505.3 1.7

Undescribed Lindgren et al. Ichthyopterygia (YORYM 1993.338) Skin N/A 2014 778.5 483.9 1.6

Palaechiropteryx (SMF- Colleary et al. Mammalia ME 11406a) Hair Brown 2015+ 573.3 318.3 1.9

Hassianycteris (SMF- Colleary et al. Mammalia ME 11407b) Hair Brown 2015+ 587.6 316.4 1.9

185 Undescribed pterosaur c.f. Jianchangnathus robustus (BMNHC Pterosauria PH000988) Pycnofiber N/A Li et al. 2014 544.1 319.5 1.7

Undescribed pterosaur Pterosauria (PMOL AP00022) Pycnofiber N/A Li et al. 2014 1069 488.5 2.3 Psittacosaurus lujiatunensis (PKUP V1050) Skin N/A Li et al. 2014 482.5 279.3 1.7 Psittacosaurus lujiatunensis (PKUP Ornithischia V1051) Skin N/A Li et al. 2014 546.3 389.7 1.4 Non- pennaraptoran Beipiaosauris (BMNHC Theropoda PH000911) Protofeather N/A Li et al. 2014 744.4 436.1 1.8 Non- pennaraptoran Sinosauropteryx (IVPP Zhang et al. Theropoda V14202) Protofeather N/A 2010* 702.6 299.5 2.4

Non-avian Microraptor (BMHNC PH881) Feather Iridescent Li et al. 2012* 1060 195.5 5.6

Non-avian zoui Pennaraptora (PMOL AD00020) Feather Black Li et al. 2014 1039 332.1 3.3

Non-avian Anchiornis huxleyi Pennaraptora (BMNHC PH828) Feather Black Li et al. 2010 772.1 156.7 5.1

Non-avian Anchiornis huxleyi Pennaraptora (BMNHC PH828) Feather Brown Li et al. 2010 624.2 427.3 1.5

Non-avian Anchiornis huxleyi Pennaraptora (BMNHC PH828) Feather Grey Li et al. 2010 737.5 201.3 3.7

Non-avian (IVPP Zhang et al. Pennaraptora V12811) Feather Black 2010+ 921.1 346.5 2.7

Non-avian Sinornithosaurus (IVPP Zhang et al. Pennaraptora V12811) Feather Brown 2010+ 703.4 595.8 1.2 Non- Unnamed ornithurine enantiornithine (CUGB P1202) Feather Iridescent Chapter 4 1459 249.1 6.5 Non- Archaeopteryx ornithurine lithographica Carney et al. Avialae (MB.Av.100) Feather Black 2012* 1006 262.7 3.9 Non- Undescribed ornithurine enantiornithine (CUGB Avialae P1201) Feather Iridescent Li et al. 2014 1533 182.0 9.2 Non- Undescribed ornithurine enantiornithine (CUGB Avialae G20120001) Feather Iridescent Li et al. 2014 1759 397.2 5.3 Non- ornithurine Confuciusornis sanctus Zhang et al. Avialae (IVPPV13171) Feather Black 2010+ 766.8 267.1 3.0 Non- ornithurine Confuciusornis sanctus Zhang et al. Avialae (IVPPV13171) Feather Brown 2010+ 466.2 342.8 1.4 Non- ornithurine Eoconfuciusornis indet. Zheng et al. Avialae (STM7-144) Feather Black 2017+ 1181 284.2 4.1 Non- ornithurine Eoconfuciusornis indet. Zheng et al. Avialae (STM7-144) Feather Brown 2017+ 446.5 313.9 1.4

186 Non- ornithurine Eoconfuciusornis indet. Zheng et al. Avialae (STM7-144) Feather Grey 2017+ 824.8 323.2 2.6

Inkayacu paracasensis Clarke et al. (MUSM 1444) Feather Brown 2010 451.8 198.4 2.3

Inkayacu paracasensis Clarke et al. Ornithurae (MUSM 1444) Feather Grey 2010 661.1 207.8 3.3

Undescribed ornithurine Ornithurae (CUGB G20100053) Feather Iridescent Li et al. 2014 1710 389.1 5.1

Gansus yumenensis(MGSF317 Barden et al. Ornithurae and MGSF318) Feather Black 2011 1660.0 490.0 3.4

Changzuiornis ahgmi Huang et al. Ornithurae (AGB5840) Feather Black 2016 1113 286.9 4.1

Isolated feather (SMF Vinther et al. Ornithurae ME 3850) Feather Iridescent 2010+ 885.7 193.2 4.7

187