Geochemical and Paleontological Evidence

Home , Gosslingia, Greenland Gap Group, Ibyka, Leclercqia (plant), Pseudobornia, Wattieza

THE EARLIEST FOREST AND ASSOCIATED WILDFIRES LINKED TO MARINE

ANOXIA AND MASS EXTINCTIONS DURING THE LATE DEVONIAN:

GEOCHEMICAL AND PALEONTOLOGICAL EVIDENCE

MAN LU

YUEHAN LU, COMMITTEE CHAIR TAKEHITO IKEJIRI REBECCA TOTTEN MINZONI KIMBERLY GENAREAU RICHARD CARROLL JACK PASHIN

A DISSERTATION

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Geological Sciences in the Graduate School of The University of Alabama

TUSCALOOSA, ALABAMA

2020

ABSTRACT

The diversification and radiation of vascular plants during the Devonian is a critical life event in geological history. The overarching goal of this dissertation is to reconstruct the evolution patterns of vascular plants through the Devonian and their impacts on terrestrial and marine environments. In Project I, I presented data from microscopic and geochemical analyses of the Upper Devonian Chattanooga Shale (Famennian Stage) in northeastern

Alabama, USA. I found plant residues, molecular biomarkers and inorganic geochemical proxies increased throughout the section, suggesting that the southern Appalachian Basin, a region representing the southernmost Euramerica, became increasingly forested during the

Late Devonian. Furthermore, the geochemical results were combined with a synthesis of vascular plant fossil records, showing a rapid southward progression of afforestation and pedogenesis along the Acadian landmass during the Late Devonian. In Project II, I established an ultra-high-resolution profile of an Upper Kellwasser (UKW) extinction interval from the Chattanooga Shale of Tennessee, USA. Through analyses of multiple paleoenvironmental proxies, I observed periodic, short-lived marine anoxia during UKW coinciding with variations in marine primary productivity, terrestrial nutrient inputs and sea level. My results suggest that anoxic episodes were caused by pulsed inputs of terrigenous organic matter that were, in turn, regulated by sea-level variations. Results from time-series analysis of Ti/Al ratios profile through Late Frasnian–Early Famennian strata demonstrates that obliquity mediated the cycle of sea-level changes, providing the first evidence that ii

recurring, episodic environmental stresses on marine organisms during the UKW were paced by astronomical forcing. In Project III, I synthesized global fire occurrences based on three paleo-wildfire proxies—fossil charcoals, inertinites, and pyrogenic PAHs. Additionally, I performed a case study of reconstructing wildfire activities during the Late Devonian based on inertinites and pyrogenic PAHs abundances in the Chattanooga Shale of Tennessee. The results show that the wildfires increased dramatically and expanded rapidly across the

Euramerica from the Frasnian to Famennian. I further analyzed the dispersal range, species, and key morphological features of vascular plants during the Devonian. The concurrent spatiotemporal expansion in wildfires and early tees through the Late Devonian suggest a rise in forest fires fueled by Archaeopteris.

iii

DEDICATION

To my parents, to Xinguang

LIST OF ABBREVIATIONS AND SYMBOLS

F–F Frasnian–Famennian

UKW Upper Kellwasser

MN 13 Montagne Noire 13 sequence

TOC Total organic carbon

TP Total phosphorus

δ13C Carbon isotope

EF Enrichment factor

XRF X-ray fluorescence

XRD X-ray diffraction

CIA Chemical Index of Alteration

CPA Chemical Proxy of Alteration

SiO2 Silicon dioxide

Al2O3 Aluminium oxide

Ti Titanium

Al Aluminum

Zr Zirconium

Mo Molybdenum

U Uranium

DCM Dichloromethane v

MeOH Methanol normal alkane n-alkane

TAR Terrigenous-to-aquatic ratio

PAHs Polycyclic aromatic hydrocarbons

PZE Photic zone euxinia

DBF Dibenzofuran

C29/C30H C29/C30 αβ hopane

Py Pyrene

BaA Banzo[a]anthrene

BeP Benzo[a/e]pyrene

BF Benzo[b/k/j]fluoranthene

MN Methylnaphthalene

DMN Dimethylnaphthalene

TMN Trimethylnaphthalene

TeMN Tetramethylnaphthalene

PMN Pentamethylnaphthalene

MP Methylphenanthrene

EP Ethylphenanthrene

DMP Dimethylnaphthrene

TMP Trimethylnaphthrene

N Naphtharene

Fl Fluoranthene

Chry Chrysene

O2 Oxygen

CO2 Carbon dioxide pO2 Partial pressure of oxygen

HF Hydrofluoric acid

HNO3 Nitric acid

HClO4 Perchloric acid

HCl Hydrochloric acid

‰ Per-mille

% Percentage mL/L Milliliter per liter ng/μL Nanogram/microliter

ºC Degree Celsius min Minute

μg/g Microgram per gram m/z Mass-to-charge ratio m Meter cm Centimeter mm Millimeter

μm Micrometer v/v Volume per volume

MTM Multi-taper method

vii

COCO Correlation coefficient

FFT Fast Fourier transform

CL% Confidence Level kyr Thousand years

Ma Million years ago

viii

ACKNOWLEDGMENTS

I would like to thank everyone who helped me complete the dissertation research.

Especially, I want to thank my academic advisor, Dr. YueHan Lu, for providing me with a chance to work with her. This dissertation is impossible to be completed without her help, support, and patient guidance. She brought me to the field of geochemistry and taught me how to do research, to be an independent thinker and researcher. She taught me how to organize research projects, how to make concise and readable presentations and writings, and even how to pronounce single English words correctly. Her knowledge and noble qualities will be lifetime wealth for me. I also want to thank another great mentor, Dr. Takehito Ikejiri.

I learned a great deal of field research geology from him. I also appreciate that he is always willing to take a lot of time to discuss science with me, and these discussions have allowed enriching this dissertation. I also want to thank Dr. Lu and Dr. Ikejiri for their help in my life.

Here, I would like to extend my sincere gratitude to my other committee members: Dr.

Richard Carroll, Dr. Rebecca Minzoni, Dr. Kimberly Genareau and Dr. Jack Pashin for their commitment to serving on my committee and stimulating many constructive discussions. I also want to thank my coauthors: Dr. Yongge Sun, Dr. Thomas Algeo, Dayang Sun, Nicholas

Hogancamp, Dr. Qihang Wu, Dr. Ibrahim Çemen and Dr. Elliot Blair for their insightful comments and revision that significantly improved the quality the manuscripts derived from my dissertation. I am very grateful to the Department of Geological Sciences for providing me the research opportunity and the teaching assistantships. ix

I also want to thank all the people I met here. Especially, I would like to thank my former and current research group members and my friends Huijing Fang and Xiaoting Liu, for their great help in my research. Special thanks go to Shuo Chen and Yihuai Lou. They are always willing to help me when I have difficulty, and their friendship and companionship brings me a lot of joy in my Ph.D. life.

I would like to express my gratitude and deepest love to my parents. I would never finish a Ph.D. without their endless love, encouragement, and support. Your supports make my life meaningful. I hope I have made you proud. I want to especially thank my fiancé,

Xinguang Wang. You are always there, cheering me up and standing by me through the good and bad times.

Finally, I would like to thank my cousin Yunyun, who encouraged me to pursue a Ph.D.

Thank you for showing up in my life. I will miss you forever.

CONTENTS

ABSTRACT ...... ii

DEDICATION ...... iv

LIST OF ABBREVIATIONS AND SYMBOLS ...... v

ACKNOWLEDGMENTS ...... ix

LIST OF TABLES ...... xv

LIST OF FIGURES ...... xvi

CHAPTER 1: INTRODUCTION ...... 1

References ...... 7

CHAPTER 2: GEOCHEMICAL EVIDENCE OF FIRST FORESTATION IN THE SOUTHERNMOST EURAMERICA FROM UPPER DEVONIAN (FAMENNIAN) BLACK SHALES ...... 11

2.1 Abstract ...... 11

2.2 Introduction ...... 12

2.3 Material and methods ...... 15

2.3.1 Samples ...... 15

2.3.2 X-ray diffraction (XRD), X-ray fluorescence (XRF) and Scanning Electron Microscope (SEM) ...... 16

2.3.3 Trace element ...... 16

2.3.4 Organic petrography ...... 17

2.3.5 TOC and stable carbon isotope of TOC ...... 17

2.3.6 Biomarker quantification and compound-specific stable carbon isotope ...... 18

2.4 Results ...... 20

2.4.1 Geological background ...... 20

2.4.2 Biostratigraphy ...... 21

2.4.3 Plant residue ...... 23

2.4.4 Bulk and molecular characteristics of organic matter...... 26

2.4.5 Inorganic geochemical characteristics of the Chattanooga Shale ...... 28

2.5 Discussion ...... 30

2.6 Acknowledgment ...... 40

2.7 References ...... 42

2.8 Appendix I ...... 52

CHAPTER 3: PERIODIC MARINE EUXINIA AND TERRESTRIAL ORGANIC MATTER INPUTS LINKED TO ASTRONOMICAL FORCING DURING THE FRASNIAN–FAMENNIAN MASS EXTINCTION ...... 82

3.1 Abstract ...... 82

3.2 Introduction ...... 83

3.3 Geological setting ...... 85

3.4 Methods...... 86

3.4.1 Sample collection ...... 86

3.4.2 Palynology ...... 87

3.4.3 Bulk geochemistry ...... 87

3.4.5 Biomarker analysis...... 89

3.4.6 Time-series analysis ...... 90

xii

3.5 Results ...... 92

3.5.1 Organic matter characteristics of the UKW interval ...... 92

3.5.2. Major and trace elements of the UKW interval ...... 94

3.5.3. Time-series analysis ...... 96

3.6 Discussion ...... 99

3.7 Acknowledgments...... 107

3.8 References ...... 108

3.9 Appendix II ...... 115

CHAPTER 4: THE RISE OF FOREST STIMULATED WILDFIRES IN THE EURAMERICA DURING THE DEVONIAN: PALEONTOLOGICAL AND GEOCHEMICAL EVIDENCE ...... 128

4.1 Abstract ...... 128

4.2 Introduction ...... 129

4.3 Method and materials ...... 131

4.3.1 Literature synthesis of wildfire records and Devonian vascular plants ...... 131

4.3.2 Case study of Late Devonian wildfire reconstruction in the southern Euramerica using organic geochemical proxies ...... 132

4.4 Results ...... 136

4.4.1 Devonian wildfires in the Euramerica in time and space ...... 136

4.4.2 Geochemical evidence of wildfires from the Chattanooga Shale in central Tennessee ...... 140

4.4.3 Diversification and key character evolution of vascular plants in the Euramerica during the Devonian ...... 143

4.5 Discussion ...... 148

4.5.1 Pyrogenic origins of inertinite and PAHs ...... 148 xiii

4.5.2 Spatiotemporal expansion of wildfires in the Euramerica ...... 150

4.5.3 Causes for increased fires during the Famennian ...... 151

4.5.4 Fuel accumulation and fire adaptations during the Late Devonian ...... 153

4.5.5 Implications to marine biotic crisis ...... 158

4.6 Conclusions ...... 159

4.7 References ...... 161

4.8 Appendix III ...... 168

CHAPTER 5: CONCLUSIONS ...... 230

References ...... 234

xiv

LIST OF TABLES Table S2.1. A summary of stratigraphic and paleogeographic occurrences of Devonian land forests in the southcentral Euramerica...... 60

Table S2.2. Occurrence of land plants with wood tissues from Devonian strata of the central Euramerica...... 62

Table S3.1. Pearson’s correlations among the proxies indicating redox conditions, marine productivity, terrestrial inputs, sea-level changes and weathering intensity measured from the Upper Kellwasser interval of the Chattanooga Shale, central Tennessee...... 125

Table 4.1. A summary of early wildfire record from the Silurian to the end of the Devonian...... 138

Table 4.2. The Pearson’s correlations between concentrations of pyrogenic PAHs relative to C27 normal alkane (n-C27) and inertinite/(inertinite+vitrinite) ratios from the Upper Devonian Chattanooga Shale in central Tennessee...... 149

Table 4.3. Species-level diversity of vascular plants from the Euramerica through the Devonian...... 155

Table S4.1. Overview of published records of Devonian wildfire records (i.e., fossil charcoal, inertinite maceral and pyrogenic PAHs) from the Euramerica...... 170

Table S4.2. Maximum observed aerial axis diameter of vascular plant fossils in Euramerica through the Devonian...... 176

Table S4.3. Devonian vascular plant fossils from the Euramerica examined for leaf-size analysis...... 196

LIST OF FIGURES Figure 2.1. Late Devonian paleogeography...... 15

Figure 2.2. Stratigraphic column and conodont occurrence of the Upper Devonian Chattanooga Shale in northeastern Alabama...... 21

Figure 2.3. Scanning electron microscope (SEM) images of land plant remains from the Upper Devonian Chattanooga Shale in northeastern Alabama...... 24

Figure 2.4. Variations in inorganic and palynological proxies across the Upper Devonian Chattanooga Shale in northeastern Alabama...... 25

Figure 2.5. Variations in bulk and molecular organic proxies across the Upper Devonian Chattanooga Shale in northeastern Alabama...... 27

Figure 2.6. δ13C values of individual normal alkanes of two representative samples from the Upper Devonian Chattanooga Shale in northeastern Alabama...... 33

Figure 2.7. Cross plots of terrestrial plant proxies versus continental weathering proxies from the Upper Devonian Chattanooga Shale in northeastern Alabama...... 36

Figure 2.8. Spatiotemporal dispersal pattern of land forests in south-central Euramerican landmass during the Devonian...... 41

Figure S2.1. m/z 85 mass chromatogram of a representative sample of the Upper Devonian Chattanooga Shale in northeast Alabama...... 53

Figure S2.3. m/z 123 chromatograms showing the distribution of bicyclic sesquiterpenoids for representative samples from the upper unit (a) and the low unit (b) of the Chattanooga Shale in northeast Alabama...... 55

Figure S2.4. m/z 217 and 218 chromatograms showing the distribution of steranes for a representative sample from the Chattanooga Shale in northeast Alabama...... 55

Figure S2.5. m/z 191 chromatogram showing the distribution of hopanes for xvi

a representative sample from the Chattanooga Shale in northeast Alabama...... 56

Figure S2.6. Scanning electron microscope (SEM) images of conodont P1 element recovered from the Chattanooga Shale in northeast Alabama and occurrence distance above base of shale...... 57

Figure S2.7. Energy-dispersive X-ray spectroscopy (EDS) spectra of land plant remains from the Upper Devonian Chattanooga Shale in northeastern Alabama...... 58

Figure S2.8. Photomicrographs (white reflected light, oil immersion) of samples from the lower unit (a) and the upper unit (b) of the Chattanooga Shale in northeast Alabama...... 59

Figure S2.9. SEM images of mineral crystals from the lower unit (a) and the upper unit (b) of the Chattanooga Shale in northeast Alabama...... 59

Figure S2.10. Wood tissue-remained plant fossils from the Chattanooga Shale of the Southern Appalachian Basin...... 75

Figure 3.1. Geographic and stratigraphic context of the Upper Kellwasser (UKW) interval in the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee...... 86

13 Figure 3.2. High-resolution profiles of bulk and molecular proxies (δ Corg, n-C17+19 alkanes, C27 steranes) from the Upper Kellwasser (UKW) interval of the Chestnut Mound outcrop, Chattanooga Shale, central Tennessee...... 92

Figure 3.3. High-resolution profiles of inorganic and organic proxies for continental weathering and terrestrial plant inputs from the Upper Kellwasser (UKW) interval of the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee...... 93

Figure 3.4. High-resolution profiles of inorganic and organic proxies for sea-level changes from the Upper Kellwasser (UKW) interval of the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee...... 95

Figure 3.5. High-resolution profiles of organic and inorganic proxies for marine redox conditions from the Upper Kellwasser (UKW) interval of the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee...... 95

xvii

Figure S3.1. Inorganic and organic proxies collected at 25 cm intervals across the Frasnian–Famennian interval of the Chattanooga Shale, central Tennessee...... 115

Figure S3.2. Mass chromatogram (m/z 57) of aliphatic fraction from extracts of one typical sample from Upper Kellwasser (UKW) interval of the Chattanooga Shale in central Tennessee...... 116

Figure S3.3. Partial Mass chromatogram (m/z 217) showing sterane distributions in one typical sample from the Upper Kellwasser (UKW) interval of the Chattanooga Shale in central Tennessee...... 117

Figure S3.4. Partial mass chromatogram (m/z 191) showing hopane distributions in one typical sample from the Upper Kellwasser (UKW) interval of the Chattanooga Shale in central Tennessee...... 118

Figure S3.5. Partial Mass chromatogram (m/z 123) of one typical sample from the Upper Kellwasser (UKW) interval of the Chattanooga Shale in central Tennessee...... 119

Figure S3.6. Total ion current chromatogram of aromatic fraction from extracts of one typical sample from the Upper Kellwasser (UKW) interval of the Chattanooga Shale, central Tennessee...... 120

Figure S3.7. Mass chromatogram (m/z) 133 of one typical sample from the Upper Kellwasser (UKW) interval of the Chattanooga Shale, central Tennessee, showing the distribution of aryl isoprenoids...... 121

Figure S3.9. Raw Ti/Al data series and 47-kyr tuned Ti/Al data series of the Upper Frasnian–Lower Famennian interval of the Chattanooga Shale, central Tennessee...... 123

Figure S3.10. Palynomorphs observed from the Upper Kellwasser (UKW) interval of the Chattanooga Shale, central Tennessee...... 124

Figure 4.1. Distribution of paleo-wildfire records—fossil charcoals, inertinite maceral and pyrogenic PAHs during the Early, Middle and Late Devonian

xviii

(Frasnian–Famennian) ...... 133

Figure 4.2. Stratigraphic variations in wildfire records across the Frasnian–Famennian interval of the Upper Devonian Chattanooga Shale in central Tennessee...... 135

Figure 4.3. Wildfire occurrences, atmospheric evolution and vascular plant evolution during the Devonian...... 137

Figure 4.4. Total ion current chromatogram of aromatic fraction from extracts of one typical sample from the Upper Devonian Chattanooga Shale, central Tennessee...... 142

Figure 4.5. PAH diagnostic ratios derived from the Frasnian–Famennian interval of the Upper Devonian Chattanooga Shale in central Tennessee...... 142

Figure 4.6. Change in percentage of pyogenic PAHs with different ring count across the Frasnian–Famennian interval of the Upper Devonian Chattanooga Shale in central Tennessee...... 143

Figure 4.7. Maximum axial diameter (a), leaf length (b) and width (c) of Devonian vascular plants from the Euramerica...... 144

Figure 4.8. Average values of maximum axial diameter, leaf length and width of Devonian vascular plants from the Euramerica...... 147

Figure S4.1. Photomicrographs (white reflected light, oil immersion) of samples from the Frasnian (top) and Famennian (bottom) interval of the Chattanooga Shale in central Tennessee...... 168

Figure S4.2. Concentrations of pyrogenic PAHs and C27 normal alkane (n-C27) across the Frasnian–Famennian interval of the Upper Devonian Chattanooga Shale in central Tennessee...... 169

xix

CHAPTER 1:

INTRODUCTION

The Late Devonian (383–359 Ma) is a time of prolonged environmental instability with a series of key events in the history of terrestrial and marine life. On land, the Late Devonian first witnessed an explosive diversification of land plants characterized by increases in size, complexity, and dispersal range (Knoll et al., 1984; Driese and Mora, 2001; Algeo and

Scheckler, 2010). Along with the diversification of basal vascular plants, the earliest forest fist evolved during the Late Devonian (Stein et al., 2012; Morris et al., 2015; Lu et al., 2019).

The rise and expansion of early forests during the Late Devonian has been hypothesized to be mobilized tremendous soils and associated nutrient inputs into shallow marine environments that profoundly altered them and caused catastrophic perturbation of global marine ecosystems (Algeo and Scheckler, 1998; Tulipani et al., 2015; Moreno et al., 2018). To understand the role of the early forests that played in shaping terrestrial and marine ecosystems, it is important to reconstruct occurrences and dispersal patterns and processes of plant taxa and associated soils during the Devonian. Current knowledge on the geographic distribution of Devonian vascular plants is largely based on micro- and mostly macro-sized fossils (e.g., Richardson and McGregor, 1986; Scheckler, 1986; Wang et al., 2015) and, therefore, there is a great limitation to access data controlled by the degree of fossil preservation. Thus, an innovative method to trace the signatures of Devonian plants and soils

is needed. The Late Devonian is a period of extensive organic matter (OM) deposition along the continental margins (Ulmishek and Klemme, 1990), an important site for terrestrial organic matter burial. Thus, I suspect that widely distributed organic-rich shales offer new insight into the temporal and spatial patterns of afforestation and associated pedogenesis during the Devonian.

The Late Devonian is also known for shallow marine biotic crises of varying magnitudes. The most severe extinction is thought to occur around the Frasnian–Famennian

(F–F) boundary, which could cause up to 22–35% extinctions of marine genera. The Late

Devonian mass extinction event is ranked as one of the five most devastating mass extinctions during the Phanerozoic (Bambach et al., 2004). The ultimate causes that triggered the Late Devonian mass extinction are still controversial. In addition to early forest radiation

(Algeo and Scheckler, 1998), other ultimate causes proposed include bolide impacts

(Sandberg et al., 2002), sea-level fluctuations (Copper 2002), climate change (Joachimski and

Buggisch, 2002), volcanism (Racki et al., 2018) and tectonism (Averbuch et al., 2005).

Several marine anoxia events occurred around the F–F boundary, among which the Upper

Kellwasser (UKW) event is suggested to be coeval with the largest extinction pulse (House,

2002), which is characterized by widespread deposition of organic-rich horizons (Joachimski and Buggsch, 1993; Bond et al., 2004; Riboulleau et al., 2018). Although earlier studies favor the hypothesis that marine anoxia was persistent around the F–F boundary (Ettensohn 1992), more recent studies have tended to suggest short-lived, periodic anoxia in global shallow marine environments (Carmichael et al., 2014; Brown and Kenig, 2004; Boyer et al., 2014;

Haddad et al., 2016, 2018). This recent view agrees with the paleontological perspective that

recurring environmental stress depressed the origination of marine organisms and led to biodiversity loss eventually during the F–F extinction (Bambach et al., 2004; Boyer et al.,

2014). However, the frequency and duration of anoxic episodes, as well as mechanisms that initiated and regulated the frequency are still unknown.

As an important disturbance in a range of terrestrial ecosystems today, forest fires are suggested to have burned throughout the Earth’s history since first terrestrial plants started to provide a fuel source and added oxygen (Scott, 2000; Glasspool et al., 2004). Fire occurrences and intensity are largely controlled by types of vegetation available.

Furthermore, vascular plant evolution coupled with fluctuations in atmospheric O2 level are expected to alter wildfire occurrences through the Late Devonian (Scott and Glasspool,

2006). However, the spatiotemporal pattern of Devonian wildfires is largely lacking.

Furthermore, fires can also influence vegetation evolution. For example, fires can act as a filter of species traits and hence alter the structure and composition of flora (Pausas and

Keeley, 2009). Global or regional patterns of vegetation and fire have been linked to the evolution of leaf properties and plant architecture during the Late Paleozoic and Mesozoic

(e.g., Keeley and Rundel, 2005; Belcher et al., 2010; Bond and Scott, 2010; Karp et al.,

2018). However, the relationship between Devonian plnat taxa and the wildfire record is still poorly understood.

The overreaching goal of this dissertation is to integrate geochemical and paleontological data to reconstruct the corresponding evolutionary patterns and processes of early vascular plants through the Devonian and investigate their significance in shaping terrestrial and marine ecosystems. The dissertation consists of five chapters, an

introductionary chapter (Chapter 1), three manuscripts for peer-reviewed publications

(Chapter 2, Chapter 3, and Chapter 4), and an overall conclusive chapter (Chapter 5).

Chapter 2, entitled “Geochemical Evidence of First Forestation in the Southernmost

Euramerica from Upper Devonian (Famennian) Black Shales”, coauthored by YueHan Lu,

Takehito Ikejiri, Nicholas Hogancamp, Yongge Sun, Qihang Wu, Richard Carroll, Ibrahim

Cemen, and Jack Pashin, was published in Scientific Reports. This chapter focuses on the dispersal pattern of forests and associated petrogenesis during the Late Devonian. The study is motivated by the limited knowledge of the dispersal pattern of forests and associated soils in the southernmost Euramerica during the Late Devonian. I collected samples from one complete section of the Chattanooga Shale in northeastern Alabama, U.S.A., which was situated in the southernmost Euramerica during the Late Devonian. I identified the signature of forests and soils preserved in the studied section of the Famennian unfossiliferous black shales by employing a comprehensive suite of geochemical tracers (e.g., mineral composition, trace metals, stable carbon isotope ratio, vitrinite and inertinite macerals, and biomarker assemblages). To examine the afforestation and pedogenesis across the Euramerica during the Late Devonian, I conducted a synthesis of Devonian vascular plant fossils. By combining my geochemical data with the fossil records of vascular plants, I suggest a rapid southward dispersal pattern of forest in Euramerica through the Late Devonian.

Chapter 3, entitled “Periodic Marine Euxinia and Terrestrial Organic Matter Inputs

Linked to Astronomical Forcing during the Frasnian–Famennian Mass Extinction”, is coauthored with YueHan Lu, Takehito Ikejiri, Dayang Sun, Richard Carroll, Elliot Blair,

Thomas Algeo and Yongge Sun and under review for Proceedings of the National Academy

of Sciences (in March, 2020). This study was carried out between 2017 and 2019 and presents ultra-high-resolution, multi-proxy geochemical proxies from the Upper Kellwasser (UKW) interval of the Chattanooga Shale in central Tennessee, U.S.A. Geochemical indices for marine deoxygenation (mid-chain aryl isoprenoids, Mo enrichment), marine primary

13 productivity (δ Corg, short-chain normal alkanes, and C27 steranes), terrestrial plant inputs

(long-chain normal alkanes, norabietane, and diagnostic aromatic compounds), and sea-level changes (C29/C30 αβ hopane, Zr/Al, and chemical index of alteration) were examined to reconstruct the development pattern and trigger of the marine anoxia. To probe the linkage between these environmental changes and astronomical forcing, a time-series analysis of geochemical proxies was performed on high-resolution Ti/Al proxy throughout the Frasnian interval of the studied section. This study provides the first evidence for the influences of astronomical forcing on marine anoxia and contemporaneous terrestrial and marine environmental changes during the UKW extinction.

Chapter 4, entitled “The Rise of Forest Stimulated Wildfires in the Euramerica during the Devonian: Paleontological and Geochemical Evidence”, coauthored with YueHan Lu,

Richard Carroll, Yongge Sun, and Takehito Ikejiri. A part of this chapter is planned to be submitted to a special issue of Palaeogeography, Palaeoclimatology, Palaeoecology, “the

Gaia Files: Co-Evolution of Land Plants and Climate at Geological Time Scales”, by the end of March, 2020. This chapter is motived by our limited knowledge of the spatiotemporal occurrence of the Devonian wildfires. The study was carried out between 2016 and 2019. I performed a detailed literature review on paleowildfire occurrences based on the observations of fossil charcoals, inertinite macerals, and pyrogenic polycyclic aromatic hydrocarbons

(PAHs) and earliest trees (geographic dispersal, species, and plant morphologies) through the entire Devonian. I also conducted a case study of reconstructing temporal changes in wildfire activities through the Late Devonian in the southern Euramerica based on inertinite macerals and pyrogenic PAHs detected from one outcrop of the Upper Devonian (Frasnian to

Famennian) Chattanooga Shale in Tennessee, U.S.A. I observed a rapid rise in occurrences and quick geographic expansion of wildfires across the Euramerica from the Frasnian to

Famennian. Such elevated wildfire occurrences may be tied to rapid dispersal and diversification of early trees, represented by Archaeopteris. I also found that axial diameter and leaf size showed stepwise evolutionary increase through the Late Devonian in parallel with the rise of wildfires, which could be an adaptative feature that trees evolved in response to the fires. My results are among the first to link the evolution and identify spatiotemporal correspondence of wildfires and vascular plants during the Devonian.

In Chapter 5, the conclusions from each chapter are summarized, with a special emphasis given to explain how rises and expansion of the earliest forest altered Earth’s environment and ecosystems during the Late Devonian.

References Algeo, T. J. & Scheckler, S. E. 1998. Terrestrial-marine teleconnections in the Devonian: links between the evolution of land plants, weathering processes, and marine anoxic events. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 353, 113–130.

Algeo, T. J. & Scheckler, S. E. 2010. Land plant evolution and weathering rate changes in the Devonian. Journal of Earth Science, 21, 75–78.

Averbuch, O., Tribovillard, N., Devleeschouwer, X., Riquier, L., Mistiaen, B. & Van Vliet- Lanoe, B. 2005. Mountain building—enhanced continental weathering and organic carbon burial as major causes for climatic cooling at the Frasnian–Famennian boundary (c. 376 Ma)? Terra Nova, 17, 25–34.

Bambach, R. K., Knoll, A. H. & Wang, S. C. 2004. Origination, extinction, and mass depletions of marine diversity. Paleobiology, 30, 522–542.

Belcher, C. M., Mander, L., Rein, G., Jervis, F. X., Haworth, M., Hesselbo, S. P., Glasspool, I. J. & Mcelwain, J. C. 2010. Increased fire activity at the Triassic/Jurassic boundary in Greenland due to climate-driven floral change. Nature Geoscience, 3, 426–429.

Bond, D., Wignall, P. B. & Racki, G. 2004. Extent and duration of marine anoxia during the Frasnian–Famennian (Late Devonian) mass extinction in Poland, Germany, Austria and France. Geological Magazine, 141, 173–193.

Bond, W. J. & Scott, A. C. 2010. Fire and the spread of flowering plants in the Cretaceous. New Phytologist, 188, 1137–1150.

Boyer, D. L., Haddad, E. E. & Seeger, E. S. 2014. The last gasp: trace fossils track deoxygenation leading into the Frasnian–Famennian Extinction Event. Palaios, 29, 646– 651.

Brown, T. C. & Kenig, F. 2004. Water column structure during deposition of Middle Devonian–Lower Mississippian black and green/gray shales of the Illinois and Michigan Basins: a biomarker approach. Palaeogeography, Palaeoclimatology, Palaeoecology, 215, 59–85.

Carmichael, S. K., Waters, J. A., Suttner, T. J., Kido, E. & Dereuil, A. A. 2014. A new model for the Kellwasser Anoxia Events (Late Devonian): Shallow water anoxia in an open oceanic setting in the Central Asian Orogenic Belt. Palaeogeography, Palaeoclimatology, Palaeoecology, 399, 394–403.

Copper, P. 2002. Reef development at the Frasnian/Famennian mass extinction boundary. Palaeogeography, Palaeoclimatology, Palaeoecology, 181, 27–65.

Driese, S. G. & Mora, C. I. 2001. ‘Diversification of Siluro-Devonian plant traces in paleosols and influence on estimates of paleoatmospheric CO2 levels’ in Gensel, P. G. & Edwards, D. (eds.) Plants Invade the Land: Evolutionary and Environmental Perspectives. New York, Columbia University Press, 237–253.

Ettensohn, F. R. 1992. Controls on the origin of the Devonian-Mississippian oil and gas shales, east-central United States. Fuel, 71, 1487–1492.

Glasspool, I. J., Edwards, D. & Axe, L. 2004. Charcoal in the Silurian as evidence for the earliest wildfire. Geology, 32, 381.

Haddad, E. E., Boyer, D. L., Droser, M. L., Lee, B. K., Lyons, T. W. & Love, G. D. 2018. Ichnofabrics and chemostratigraphy argue against persistent anoxia during the Upper Kellwasser Event in New York State. Palaeogeography, Palaeoclimatology, Palaeoecology, 490, 178–190.

Haddad, E. E., Tuite, M. L., Martinez, A. M., Williford, K., Boyer, D. L., Droser, M. L. & Love, G. D. 2016. Lipid biomarker stratigraphic records through the Late Devonian Frasnian/Famennian boundary: comparison of high-and low-latitude epicontinental marine settings. Organic Geochemistry, 98, 38–53.

House, M. R. 2002. Strength, timing, setting and cause of mid-Palaeozoic extinctions. Palaeogeography, Palaeoclimatology, Palaeoecology, 181, 5–25.

Joachimski, M. M. & Buggisch, W. 1993. Anoxic events in the late Frasnian—Causes of the Frasnian–Famennian faunal crisis? Geology, 21, 675–678.

Joachimski, M. M. & Buggisch, W. 2002. Conodont apatite δ18O signatures indicate climatic cooling as a trigger of the Late Devonian mass extinction. Geology, 30, 711–714.

Karp, A. T., Behrensmeyer, A. K. & Freeman, K. H. 2018. Grassland fire ecology has roots in the late Miocene. Proceedings of the National Academy of Sciences, 115, 12130–12135.

Keeley, J. E. & Rundel, P. W. 2005. Fire and the Miocene expansion of C4 grasslands. Ecology Letters, 8, 683–690.

Knoll, A. H., Niklas, K. J., Gensel, P. G. & Tiffney, B. H. 1984. Character diversification and patterns of evolution in early vascular plants. Paleobiology, 10, 34–47.

Lu, M., Lu, Y., Ikejiri, T., Hogancamp, N., Sun, Y., Wu, Q., Carroll, R., Çemen, I. & Pashin, J. 2019. Geochemical evidence of First Forestation in the southernmost euramerica from Upper Devonian (Famennian) Black shales. Scientific Reports, 9, 1–15.

Morris, J. L., Leake, J. R., Stein, W. E., Berry, C. M., Marshall, J. E. A., Wellman, C. H., Milton, J. A., Hillier, S., Mannolini, F., Quirk, J., Beerling, D. J. & Lomax, B. 2015. Investigating Devonian trees as geo-engineers of past climates: linking palaeosols to palaeobotany and experimental geobiology. Palaeontology, 58, 787–801.

Pausas, J. G. & Keeley, J. E. 2009. A burning story: the role of fire in the history of life. Bioscience, 59, 593–601.

Racki, G., Rakociński, M., Marynowski, L. & Wignall, P. B. 2018. Mercury enrichments and the Frasnian–Famennian biotic crisis: A volcanic trigger proved? Geology, 46, 543–546.

Riboulleau, A., Spina, A., Vecoli, M., Riquier, L., Quijada, M., Tribovillard, N. & Averbuch, O. 2018. Organic matter deposition in the Ghadames Basin (Libya) during the Late Devonian—A multidisciplinary approach. Palaeogeography, Palaeoclimatology, Palaeoecology, 497, 37–51.

Richardson, J. B. & Mcgregor, D. C. 1986. Silurian and Devonian spore zones of the Old Red Sandstone continent and adjacent regions, Geological Survey of Canada, 354, pl–1.

Scheckler, S. E. 1986. Geology, floristics and paleoecology of Late Devonian coal swamps from Appalachian Laurentia (USA). Annales de la Société géologique de Belgique.

Scott, A. C. 2000. The Pre-Quaternary history of fire. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 281–329.

Scott, A. C. & Glasspool, I. J. 2006. The diversification of Paleozoic fire systems and fluctuations in atmospheric oxygen concentration. Proceedings of the National Academy of Sciences, 103, 10861–5.

Stein, W. E., Berry, C. M., Hernick, L. V. & Mannolini, F. 2012. Surprisingly complex community discovered in the mid-Devonian fossil forest at Gilboa. Nature, 483, 78–81.

Tulipani, S., Grice, K., Greenwood, P. F., Haines, P. W., Sauer, P. E., Schimmelmann, A., Summons, R. E., Foster, C. B., Böttcher, M. E., Playton, T. & Schwark, L. 2015. Changes of palaeoenvironmental conditions recorded in Late Devonian reef systems from the Canning Basin, Western Australia: A biomarker and stable isotope approach. Gondwana Research, 28, 1500–1515.

Ulmishek, G. F. & Klemme, H. D. 1990. Depositional controls, distribution, and effectiveness of world petroleum source rocks. United States Geological Survey Bulletin, Report #B, 59.

Wang, D., Liu, L., Guo, Y., Xue, J. & Meng, M. 2015. A Late Devonian fertile organ with seed plant affinities from China. Scientific Reports, 5, 10736

CHAPTER 2:

GEOCHEMICAL EVIDENCE OF FIRST FORESTATION IN THE SOUTHERNMOST

EURAMERICA FROM UPPER DEVONIAN (FAMENNIAN) BLACK SHALES

2.1 Abstract

The global dispersal of forests and soils has been proposed as a cause for the Late

Devonian mass extinctions of marine organisms, but detailed spatiotemporal records of forests and soils at that time remain lacking. I present data from microscopic and geochemical analyses of the Upper Devonian Chattanooga Shale (Famennian Stage). Plant residues (microfossils, vitrinite and inertinite) and biomarkers derived from terrestrial plants and wildfire occur throughout the stratigraphic section, suggesting widespread forest in the southern Appalachian Basin, a region with no macro plant fossil record during the

Famennian. Inorganic geochemical results, as shown by increasing values of SiO2/Al2O3,

Ti/Al, Zr/Al, and the Chemical Index of Alteration (CIA) upon time sequence, suggest enhanced continental weathering that may be attributed to the invasion of barren lands by rooted land plants. My geochemical data collectively provide the oldest evidence of the influences of land plants from the southernmost Appalachian Basin. My synthesis of vascular plant fossil record shows a more rapid process of afforestation and pedogenesis across southcentral Euramerica during the Frasnian and Famennian than previously documented.

Together, these results lead us to propose a new hypothesis that global floral dispersal had progressed southward along the Acadian landmass rapidly during the Late Devonian. 11

2.2 Introduction

The Late Devonian is known for the rapid and global radiation of early forests and soils such as spodosols and alfisols (Retallack, 1997; Algeo et al., 1995; Le Hir et al., 2011). The development of land plants and soils is hypothesized to have been either a trigger or consequence of a series of global changes in the lithosphere (e.g., increased weathering and erosion), hydrosphere (e.g., anoxic oceans, global transgression and regression), atmosphere

(e.g., global changes in O2 and CO2), and biosphere (i.e., mass extinctions of marine life) during the Middle to Late Devonian (Algeo and Scheckler, 1998; Berner, 1997; Turgeon et al., 2007; Błażejowski et al., 2015; Brom et al., 2018; Haworth et al., 2018; Percival et al.,

2018). One compelling hypothesis is that forest radiation mobilized a tremendous amount of soils and associated nutrients (nitrogen and phosphorus) to coastal oceans for the first time in

Earth’s history and led to dysoxic/anoxic oceans globally (Algeo and Scheckler, 1998; Algeo and Scheckler, 2010; Kaiho et al., 2013b; Tulipani et al., 2015). Testing this hypothesis, however, requires data on stratigraphic occurrences of soils and plants in specific paleogeographic areas.

Current knowledge on the paleogeographic distribution of Devonian forests is largely based on macrofossils, such as tree trunks, stems, leaves, and roots, as well as some microfossils such as spores (Richardson and McGregor, 1986; Scheckler, 1986b; Scheckler,

1986a; Wang et al., 2015). The oldest tree stems and stumps in the Euramerica are reported from the uppermost Givetian (Middle Devonian) strata of New York (i.e., Gilboa Park and

Cairo) representing the central Euramerican landmass (Stein et al., 2007) (Fig. 2.1). To date, the record of Devonian trees and shrubs assigned to pteridophytes, which were likely the

primary component for the first forest, is very limited from the southern Appalachian Basin along the southern Acadian landmass. Only a few uncertain remains have been reported as small wood fragments of possible Callixylon (Conant and Swanson, 1961) and Foerstia

(Schopf and Schwietering, 1970) from Tennessee. In contrast, some tree or shrub fossils are known from the northern Appalachian Basin (e.g., New York, Pennsylvania, West Virginia) and the Baltica and Avalonia landmasses (e.g., United Kingdom, Belgium) (Long, 1961;

Galtier and Meyer-Berthaud, 1996; Matten et al., 1980; Berry and Fairon-Demaret, 2001;

Streel et al., 2000; Scheckler et al., 2006; Prestianni and Gerrienne, 2010). This gap in the paleogeographic occurrence between the northern and southern parts of the Appalachian

Basin implies that forests originated from the central Euramerica in the late Middle Devonian and dispersed southward during the Late Devonian. Better understanding of the spatiotemporal occurrence of land plants in the southern Appalachian Basin will provide a better understanding of the dispersal pattern of the early land forests.

To date, fluvial sedimentary sequences are the most frequently reported units containing tree and shrub fossils (lycopsids, cladoxylopsids, progymnosperms, and possibly stem spermatophytes) and soils in the Appalachian Basin (Retallack, 1997; Cressler, 2006; Stein et al., 2007; Meyer-Berthaud et al., 2010). Those sandstone, siltstone, and limestone deposits are geographically distributed along the eastern side of the basin (close to the modern-day

Appalachian Mountains) (Woodrow et al., 1988). By comparison, very few tree fossils have been reported from the extensive Upper Devonian black shale deposits further offshore from the western margin of the Appalachian Basin, including the Ohio Shale, New Albany Shale,

Cleveland Shale, Huron-Dunkirk shales, Millsboro Shale, and Chattanooga Shale. Typically,

the preservation of land plant fossils in offshore marine deposits is not as common as that in nearshore environments (e.g., fluvial, deltaic), because plant remains can be easily broken into pieces and decomposed through taphonomic processes. However, if information on the distribution of early forests could be obtained from the abundant and regionally extensive

Upper Devonian marine black shales, it would significantly increase the amount of data on the occurrence of land plants available to geologists from the Paleozoic rock record. Such efforts would allow for the reconstruction of a far more spatially and stratigraphically detailed record of afforestation than is possible using the rarely preserved fluvial deposits alone.

In the present study, I present identifiable signatures of forests and soils preserved in unfossiliferous black shales in the southernmost Appalachian Basin. I further demonstrate the potential of using these signatures to generate a new understanding of the dispersal patterns of Famennian forest (land plant) and pedogenesis along the southern Acadian Orogen. I analyzed a complete section of the Chattanooga Shale in northeastern Alabama (Fig. 2.1).

The Chattanooga Shale and other Upper Devonian black shale units in the Appalachian Basin are interpreted to have accumulated in a basin-like depositional environment further offshore than equivalent sandstone or siltstone dominated formations (i.e., alluvial plain or basin margin-like environment) (Woodrow et al., 1988; Schieber, 1994). Plant macrofossils such as stems, leaves, stumps, and roots are nearly absent in the Chattanooga Shale, apart from a few brief notes from central Tennessee (Conant and Swanson, 1961). Using microscopic investigation and a comprehensive set of geochemical analyses (e.g., inertinite and vitrinite, mineral composition, trace metals, stable carbon isotope ratio, and biomarker assemblages), I

investigate whether early forests had left detectable signals in the Chattanooga Shale of

Alabama. The multiple geochemical-tracer approach I use here can overcome the preservation limit of macrofossils, and thus extend our knowledge of the spatiotemporal pattern of the early forest radiation in the southernmost Euramerican continent. My data can set the foundation for new hypotheses regarding how afforestation progressed during the Late

Devonian.

Figure 2.1. Late Devonian paleogeography. Left: the global scale; right: south-central Euramerica. The yellow star indicates the study site (the present location of Alabama), which represents the southern part of the Chattanooga Sea (dashed elliptical area). Maps are adapted from original map (360 Ma) from Global Paleogeography and Tectonics in Deep Time Series by Ron Blakey [© 2016 Colorado Plateau Geosystems Inc.].

2.3 Material and methods

2.3.1 Samples

Rock samples were collected from 45 layers at a 25-cm interval, which covered all identified or visible lithological changes in the Chattanooga Shale section (Fig. 2.2).

Weathered rocks (i.e., generally turning to light gray) were avoided for sampling, but freshly exposed rocks (i.e., darker color) were chosen. Prior to geochemical analysis, samples were

washed in sonicating baths sequentially using ultra-pure carbon-free water, dichloromethane, and hexane.

2.3.2 X-ray diffraction (XRD), X-ray fluorescence (XRF) and Scanning Electron Microscope

(SEM)

Fifteen samples collected at intervals of 25–50 cm were selected for X-ray diffraction analysis on a Bruker D8 Advance XRD. Mineral identification was based on diffraction patterns using the DIFFRACPLUSEVA 4.0 library (Bruker AXS), and the abundances of different mineral components were determined using the Rietveld Method (Rietveld, 1969).

Total element contents were determined by X-ray fluorescence spectrometry (Philips

PANalytical PW2424, Netherland) at the ALS Chemex Lab, Ltd (Guangzhou, China). Before the analysis, the powdered samples were dissolved using lithium metaborate mixed with lithium nitrate and heated at 1050 ºC for an hour. The mixtures were then transferred into a platinum mould and analyzed by XRF spectrometry.

Selected samples were fixed on stubs and coated with carbon for further examination of minerals and microfossils using a JOEL SEM (JSM-6010PLUS/LA). The SEM magnification was set to 500X to 3500X, depending on the size of the particles. The elemental composition of the samples was analyzed on a JEOL 7000 FE SEM equipped with EDX at the Central

Analytical Facility, The University of Alabama.

2.3.3 Trace element

Measurements of trace elements were made using a PerkinElmer Elan9000 element inductively coupled plasma mass spectrometry (ICP-MS) at the ALS Chemex Lab, Ltd

(Guangzhou, China). Ground samples were prepared using a four-acid (HF, HNO3, HClO4,

HCl) digestion method (Ouyang et al., 2016). Analytical precision for all elements is better than 7%, and accuracy was evaluated relative to international reference materials, including

GBM398-4c, GBM908-10, MRGeo08 and OGGeo08.

2.3.4 Organic petrography

Samples at an interval of 50 cm were selected for organic petrography analysis. 10g of each sample was demineralized using cold 10% HCl for 24 hours and then cold 48% HF for

48 hours. The samples were then treated with hot Schultz’s solution and sodium hydroxide, followed by a water rinse until a neutral pH was achieved. The residues of the samples were embedded in epoxy resin, polished, and observed using reflectance microscopy under a

Nikon Microphot microscope. The samples were examined under immersion oil using a ×40 objective lens, and the abundances of vitrinite and inertinite particles were point-counted

(500 points). All samples were analyzed in duplicate and mean values of the results are reported in this study.

2.3.5 TOC and stable carbon isotope of TOC

Samples at intervals of 25–100 cm were analyzed for total organic carbon (TOC) and stable isotopes of total organic carbon (δ13C). Samples were ground into 100–200 mesh powder, and approximately 10 mg of powdered samples were placed into tin capsules.

Samples were treated with 5% sulfurous acid to remove carbonate and then dried overnight in an oven at 50 ºC. The samples were analyzed on a Micro Cube elemental analyzer (Elementar

Analysensysteme GmbH, Hanau, Germany) interfaced to a PDZ Europa 20-20 isotope ratio mass spectrometer (The Sercon Ltd., Cheshire, UK) at The University of California Davis

Stable Isotope Facility (California, USA). The analytical precision was <0.2 ‰ based on

internal standards (including nylon, bovine liver, peach leaves, and glutamic acid) calibrated against NIST Standard Reference Materials (USGS-40, USGS-41). My data were reported as

δ13C values (‰) relative to V-PDB.

2.3.6 Biomarker quantification and compound-specific stable carbon isotope

Samples at an interval of 25–100 cm were selected for biomarker analysis at the Organic

Geochemistry Laboratory, Department of Geological Sciences, University of Alabama.

Duplicate samples were analyzed in every five samples, and solvent blanks were taken through the whole procedure for each run. Approximately 5 g powdered samples were ultrasonically extracted (20 minutes) three times with a mixture of 18 mL dichloromethane

(DCM) and 2 mL methanol. Blanks (i.e., only the solvent mixture) were analyzed every 5 samples. In order to remove sulfur, short copper turnings were added to the extracts during the extraction process (20 ºC) and overnight storage (−20 ºC). The extracts were then concentrated to a volume of ca. 1 mL with a gentle ultrahigh purity (UHP) nitrogen stream using a Zymark Turbo Vap LV Evaporator, and the concentrates were then transferred into

GC vials. The extracts were further blown dry gently, diluted with 300–350 μL of hexane, and run on an Agilent 7890B gas chromatograph interfaced with an Agilent 5977A mass selective detector (MSD). The MSD was operated at a full scanning mode in the mass range of m/z 50–700 at 2.3 scans per second at ionizing electron energy of 70 eV. A fused silica capillary column (Agilent Technologies: 30m × 0.32 mm, DB-5, 0.25 μm) was used with helium as the carrier gas at a rate of 0.9 mL/min. Sample injection was operated in a pulsed splitless mode at 320 ºC. The oven temperature was set at 60 ºC, held for 1 minute, and increased at a rate of 6 ºC/min to 325 ºC, held for 20 minutes. External standards are a

mixture of C7–C40 saturated n-alkanes (Sigma Aldrich 49453-U, St. Louis, Missouri) and a

PM-610 PAH (Ultra Scientific, North Kingstown, Rhode Island) for quantifying aliphatic and aromatic compounds, respectively. The concentration was calculated using a five-point, peak area vs. concentration calibration curve constructed from standard mixtures with known concentrations (concentration from 0.1 to 20 ng/μL). Compound concentrations were reported in values normalized to TOC contents (μg/g TOC) or the relative percentages.

Selected samples were separated into aliphatic and aromatic fractions after the precipitation of asphaltenes. The de-asphalted extracts were then separated into saturate, aromatic, and polar fractions hexane, benzene, and methanol, respectively.

For compound-specific carbon isotopes of normal alkanes, saturated hydrocarbon fractions were further separated into n-alkanes and branched/cyclic alkanes by urea adduction

(Sun et al., 2005). The δ13C values of n-alkanes were measured in duplicate on a Thermo

Fisher Trace GC Ultra coupled with a Thermo Fisher MAT-253 mass spectrometer. The GC was fitted with a 60 m × 0.32 mm i.d. A DB-1MS fused silica capillary column with a film thickness of 0.25 μm leading directly into the combustion furnace was used. The GC oven temperature was programmed from 50 ºC (1 min) at 1.5 ºC /min to 125 ºC, then increased to

300 ºC at 5 ºC /min, and finally held at 300 ºC for 30 min. Helium was used as carrier gas.

The isotopic values were calibrated against the reference gas and were reported in the usual

“del” notation relative to VPDB. The precision of the measurements was typically < 0.5‰.

The accuracy of the instrument was evaluated two to three times daily via analyzing a mixture of n-alkanes with known δ13C values acquired from Indiana University, USA.

2.4 Results

2.4.1 Geological background

The outcrop of the Chattanooga Shale for this study is located in DeKalb County, northeastern Alabama (Fig. 2.1) (Material and Methods). The Chattanooga Shale is exposed with disconformable boundaries above the Upper Silurian Red Mountain Sandstone and below the Lower Mississippian Maury Shale. The Chattanooga Shale in this outcrop is 11.3 m thick and is subdivided here into a lower and upper unit based on lithological characteristics (Fig. 2.2). The lower unit is 4.3 m thick and characterized by thinly laminated, pyritic, fissile rocks composed of layers of interbedded gray and black shale. The upper unit is 7.0 m thick and is composed of dark gray to black, silty, blocky shales. Similar stratigraphic features have been reported from other Chattanooga Shale sections, entirely or in part, in northeastern Alabama (Hass, 1956; Pashin et al., 2011).

No macrofossils (visible to the naked eye) have been reported from the Chattanooga

Shale in Alabama, aside from a single bivalve shell from a different nearby section (personal observation in 2015). Detailed field work indicates that the studied outcrop is largely unfossiliferous (i.e., no macrofossils were observed with the naked eyes), confirming previous studies. In contrast, macrofossils of marine invertebrates (e.g., brachiopods, crinoids, corals) are relatively common in the Chattanooga Shale in Tennessee (Schieber,

1994; Schopf and Schwietering, 1970). A few possible plant remains have also been reported from Tennessee, but, to date, nothing that has been identified to a specific taxon has been documented from Alabama or Tennessee (Conant and Swanson, 1961).

Figure 2.2. Stratigraphic column and conodont occurrence of the Upper Devonian Chattanooga Shale in northeastern Alabama. The site is located in Fort Payne, northeastern Alabama. The conodont biozone is based on Over (2007) and Li and Schieber (2015).

2.4.2 Biostratigraphy

The phosphatic tooth-like remains of conodonts can be found in almost all marine deposits from the Lower Paleozoic to the Triassic (Sweet, 1988). For this reason, conodonts are the primary biostratigraphic fossils in most Paleozoic stages, including the Upper

Devonian (Over, 2007; Hansma et al., 2015; Hillbun et al., 2015; Day and Witzke, 2017;

Spalletta et al., 2017). Conodonts were identified on bedding plane breaks throughout the studied outcrop to provide a biostratigraphic framework to constrain the timing of

geochemical variation. The zonation names used in this section are from the updated

Famennian zonation of Spalletta et al. (2017). For clarity and for easier comparison to previous studies (e.g., Over, 2007), the older zonation names are also included in parenthesis.

Most conodonts observed in the studied outcrop of the Chattanooga Shale are external molds (Appendix I Fig. S2.6). Some phosphatic remains of conodont elements are present, but are often weathered, resulting in a white, partially dissolved element. Conodonts are most abundant in the lower unit, within the lowermost 1–2 m interval (Fig. 2.2). The absence of

Palmateolepis glabra unca and P. lobicornis combined with the presence of P. superlobata in the 0–0.75 m interval indicates Palmateolepis termini Zone (Middle crepida Zone) in age.

P1 elements of P. lobicornis, P. minuta, P. superlobata, and P. glabra unca were identified from the 1.25 m and 1.75 m levels, suggesting that this interval is Palmatolepis glabra prima

Zone (Upper crepida Zone) in age (Over, 2007; Spalletta et al., 2017).

Conodonts were found through the study section. Although those remains tend to be scattered or poorly preserved in the upper unit (except well-preserved Palmateolepis perlobata at the 7.5 m-level), relatively abundant materials were recovered from the lower unit (esp., in the 0 m to 5 m interval from the base) (Fig. 2.2). Tasmanitid algal cysts are abundant from 5.5 to 7.5 m, and much more silty bedding planes were observed from 7.5 m to the top. Some skeletal fragments were observed under a Scanning Electron Microscope

(SEM) that may be derived from brachiopods, bivalves, gastropods, and/or probable spicules of sponges. The abundant Famennian conodonts observed within basal 2 m of the outcrop show that the studied section comprises only the Famennian Gassaway Member, and that the older Frasnian age Dowelltown Member is missing. Furthermore, the two lithological units

identified in the study section resemble the lower and middle units of the Gassaway Member of the Chattanooga Shale in central Tennessee described previously (Schieber, 1998; Conant and Swanson, 1961).

2.4.3 Plant residue

Plant residue was identified under microscopy. Those include fragments of leaves, branches, roots, and spores. Tubular or irregularly shaped, carbon-rich fragments are more common in the lower unit (Fig. 2.3; Appendix I Fig. S2.7). Those tubular fragments display an overall curved shape with a smooth cortex. Spore-like particles and woody fragments were only observed in the upper unit. One well-preserved piece is elongated with a dimension of

30 μm in length and 10 μm in width and likely represents the remains of a spore. Spore-like particles have a rounded or elliptical shape with a bulged surface and are likely derived from trilete spore. Woody fragments are typically stick-like in shape, have a smooth surface, and have dimensions of 20–70 μm in length and about 10 μm in width. Some fragments with xylem- or phloem-like structures indicate functionally conducting wood tissues.

Organic petrographic analyses show that more than 50% of macerals are amorphous organic matter (Appendix I Fig. S2.8) that is generally considered to be a degradation product of organic materials of mainly marine origin (Taylor et al., 1998; Pickel et al., 2017). Figured components are mainly vitrinite and inertinite, which are commonly used as an indicator of plant residue in both fluvial and marine strata (Nichols and Jones, 1992; Falcon-Lang, 1998;

Jahren et al., 2001; Marynowski and Filipiak, 2007; Petersen et al., 2013; Rimmer et al.,

2015). In my samples, the majority of inertinite fragments are of high reflectance and sharply angular in shape (Fig. 2.4) that are suggested as typical features resulting from plant

combustion (Taylor et al., 1998; Scott and Glasspool, 2007). The relative abundance of vitrinite+inertinite is hence used to represent terrestrial inputs. My data show a significantly increasing trend from the lower unit (mean ± standard deviation = 15.6 ± 5.4%) to the upper unit (24.5 ± 6.4%) (Mann-Whitney U test: P = 0.026).

Figure 2.3. Scanning electron microscope (SEM) images of land plant remains from the Upper Devonian Chattanooga Shale in northeastern Alabama. a Tubular particle from the lower unit. b Deformed trilete spore from the upper unit. c Piece of wood fragment. d Wood fragment with conducting tissues inside. The associated energy-dispersive X-ray spectroscopy (EDS) spectra are presented in Appendix I Figure S2.7.

Figure 2.4. Variations in inorganic and palynological proxies across the Upper Devonian Chattanooga Shale in northeastern Alabama. Light grey color marks the upper unit of the section. Box plots show a comparison of upper vs. lower unit, and asterisks indicate that significance is detected between the lower and upper units (α = 0.05). The higher values of SiO2/Al2O3 and vitrinite+ inertinite in the upper unit reflect enhanced terrestrial plant inputs accompanied by high siliciclastic input. Correspondingly, the weathering indices (CIA and CPA) values are higher in the upper unit, reflecting enhanced weathering of land materials. The higher Ti/Aland Zr/Al values in the upper unit suggest increased contributions of heavy detrital sediment inputs, and the higher TP contents in the upper unit likely indicate a higher marine primary productivity.

2.4.4 Bulk and molecular characteristics of organic matter

Total organic carbon (TOC) content of the rocks in the section range from 2.8% to

13.7% (Fig. 2.5). The TOC of black shales averages 9.6 ± 3.3% in the lower unit, which is significantly higher than that in the upper unit (6.6 ± 1.3%) (Mann-Whitney U test: P =

0.017). The δ13C values of TOC fluctuate between −29.9 and −27.9 ‰ with a significantly high value in the upper unit (−28.7 ± 0.4‰) than in the lower unit (−29.5 ± 0.5‰) (Mann-

Whitney U test: P = 0.01).

The distribution of normal alkanes shows a carbon range from n-C13 to n-C32 with an enrichment in low molecular weight n-alkanes (n-C15–n-C19: 45 ± 9.4 % of total n-alkanes), which is typical for marine black shales(Sikes et al., 2009; Brown and Kenig, 2004;

Riboulleau et al., 2018). The δ13C values of n-alkanes of selected samples demonstrate that short-chain n-alkanes are on the average of 0.71‰ more enriched in 13C than long-chain n- alkanes (Fig. 2.6). This suggests different biological origins of the short and long-chain n- alkanes. The terrigenous-to-aquatic ratio (TAR), defined as (n-C27 + n-C29 + n-C31)/(n-C15 + n-C17 + n-C19), has been widely used to quantify terrestrial versus aquatic source contributions in sedimentary organic matter (Bourbonniere and Meyers, 1996; Fabbri et al.,

2005; Lu and Meyers, 2009; Silva et al., 2012). The TAR values range between 0.07 and 0.75 and average 0.27 ± 0.20 (Fig. 2.5a). The upper unit has TAR values (0.43 ± 0.23) that are significantly higher than the lower unit (0.16 ± 0.07) (Mann-Whitney U test: P = 0.005).

Figure 2.5. Variations in bulk and molecular organic proxies across the Upper Devonian Chattanooga Shale in northeastern Alabama. Light grey color marks the lower unit. Box plots show a comparison of upper vs. lower unit, and asterisks indicate that significance is detected between the lower and upper units (α = 0.05). TOC contents are lower in the upper unit, reflecting the dilution by detrital, inorganic materials. Higher δ13C values in the upper unit may reflect increased phytoplankton growth. Normal alkane parameters all show a significantly higher contribution of higher land plant-derived organic matter in the upper unit, and PAHs from higher plants or plant combustion are present in all samples, indicating the contributions of organic matter from higher plants throughout the Famennian.

A range of polycyclic aromatic hydrocarbons (PAHs) were identified, and the terrestrial plant-derived compounds are presented in Fig. 2.5b and Appendix I Fig. S2.2.

Retene, perylene, 1,7-dimethylphenanthrene, 1,2,5,6-tetramethylnaphthalene, and 1- methylphenanthren are compounds that potentially indicate the contributions of organic matter from land plants to aquatic sediments (Laflamme and Hites, 1978; Armstroff et al.,

2006; Grice et al., 2009; Romero-Sarmiento et al., 2011b; Tulipani et al., 2015). Although these compounds were detected in all samples, their concentrations do not show a statistically significant difference between the lower and upper units. The relative concentrations of perylene and the three alkyl PAHs are overall higher in the upper unit, which are 1.1 ± 0.3% and 4.6 ± 0.7% relative to total PAHs, respectively, than those in the lower unit, which are

0.3 ± 0.4 and 4.5 ± 0.4%, respectively (Fig. 2.5b). The relative concentrations of retene are overall stable in both the upper (0.9 ± 0.3%) and lower (0.9 ± 0.2%) units. PAHs that have been previously used to reconstruct land plant burning and forest fires in geological history were detected in all samples. These include pyrene, benzo(a)pyrene, benzo(e)pyrene and coronene (Appendix I Fig. S2.2) (Tulipani et al., 2015; Cesar and Grice, 2017). The relative concentrations of these combustion-derived PAHs show no significant change from the lower to the upper unit (Mann-Whitney U test: P = 1.000).

2.4.5 Inorganic geochemical characteristics of the Chattanooga Shale

Based on X-ray fluorescence (XRF) analyses, the samples contain SiO2 (64.8 ± 12.8 wt. %), Al2O3 (11.3 ± 1.2 wt.%), K2O (2.9 ± 0.9 wt.%), Na2O (0.6 ± 0.2 wt.%) and CaO (0.5

± 0.5 wt.%). The Chemical Index of Alteration (CIA = [Al2O3/(Al2O3 + Na2O + CaO + K2O)]

× 100) and the Chemical Proxy of Alteration (CPA = [Al2O3/(Al2O3 + Na2O)] × 100) were

calculated as previously described (Nesbitt and Young, 1982; Buggle et al., 2011). These two proxies provide a quantitative measure for silicate rock and soil weathering, since alumina becomes increasingly enriched throughout the weathering process and sodium, calcium and potassium are more preferentially removed (Buggle et al., 2011; Nesbitt and Young, 1982).

CIA values differ significantly between the two stratigraphic units (Mann-Whitney U test:

P=0.008), showing a lower mean value (67.5 ± 5.0) in the lower unit than the upper unit (73.1

± 4.3) (Fig. 2.4). Similarly, CPA values show a significantly lower value (91.6 ± 1.1) in the lower unit than in the upper unit (93.3 ± 1.7, Mann-Whitney U test P = 0.006) (Fig. 2.4).

X-ray diffraction (XRD) analyses show that quartz and clay minerals are the main components in the Chattanooga Shale section. Quartz is the most abundant mineral; it varies in a large range, however, from 6.6% to 90.5% (57.9 ± 20.2%). Clay minerals are the next abundant, ranging from 12.0% to 92.2% (36.3 ± 21.6%), and chlorite and illite are the two most abundant minerals. The quartz-to-clay ratio increases significantly from 1.7 ± 1.0 in the lower unit to 3.6 ± 2.4 in the upper unit (Mann-Whitney U test: P = 0.189). This pattern is in agreement with the XRF data that show an increasing trend in the ratios of SiO2 to Al2O3, from 5.1 ± 1.1 in the lower unit to 7.2 ± 1.3 in the upper unit (Mann-Whitney U test: P =

0.001) (Fig. 2.4). Crystal forms examined under SEM show that the quartz component is primarily made up of detrital grains (Appendix I Fig. S2.9), instead of originating from cysts of green algae Tasmanites, which instead appear as flattened organic spheres lacking early quartz cement. Therefore, the variation in SiO2 primarily reflects changes in the relative amount of terrigenous materials. On the other hand, Al-normalized concentrations of Ti and

Zr, Ti/Al and Zr/Al ratios in the lower unit are (0.07±0.01 and (2.19 ± 0.27) × 10-3

respectively) significantly lower relative to the values for the upper unit (0.09±0.01 and (3.28

± 0.11) × 10-3, respectively; Mann-Whitney U test: P ≤0.003) (Fig. 2.4). Ti and Zr are thought to be contributed by high-density minerals such as zircon, rutile, sphene, and ilmenite

(Calvert and Pedersen, 2007). This increasing trend in the studied section suggests a change in mineral assemblages during the deposition of the Chattanooga Shale.

2.5 Discussion

Although black shales in eastern North America may not preserve abundant macrofossils of Devonian trees and shrubs, my geochemical and microscopic data show that the signatures of afforestation can be identified in offshore environments. Here, we present three main lines of geochemical evidence that demonstrate land plants contributed organic and inorganic terrestrial material to offshore environments throughout the deposition of the

Chattanooga Shale. First, plant body parts and combustion residue (e.g., wood pieces, spores, vitrinite, inertinite) throughout the entire Alabama section, provide the most direct, visible evidence of the land plant contribution to the Chattanooga Shale (Fig. 2.3; Appendix I Figs.

S2.7 and S2.8). Vitrinite is thought to be derived from wood tissues (Taylor et al., 1998;

Kennedy et al., 2013; Killops and Killops, 2013), and inertinite represents highly oxidized materials generated from slow oxidation of organic matter or rapid oxidation during wildfires

(Jones and Chaloner, 1991, Scott, 2002, Uhl and Kerp, 2003, Scott and Glasspool, 2007). In the studied section, vitrinite and inertinite show an overall increasing trend from the lower to upper unit (Fig. 2.4). Similarly, (Rimmer et al., 2015) also reported this pattern of increases in inertinite from the uppermost Devonian terrestrial and marine rocks (including black

shales) in the northern Appalachian Basin, and it was interpreted to be a result of an increasing occurrence of wildfires.

The second line of geochemical evidence is based on biomarkers, including PAHs and normal alkanes. PAH compounds including pyrene, benzo(a)pyrene, benzo(e)pyrene and coronene (Fig. 2.5b) possibly indicate land plant burning. These PAH compounds have been used to indicate wildfire events throughout the Phanerozoic from a diverse type of rocks and sediments, including Devonian marine sedimentary rocks (Marynowski and Filipiak, 2007;

Kaiho et al., 2013a; Tulipani et al., 2015; Riboulleau et al., 2018). For example, benzo(a)pyrene, benzo(e)pyrene, pyrene and coronene co-occurring with inertinite were reported from Upper Devonian marine rocks in Poland as the evidence of paleo-wildfires in the eastern Avalonia (Marynowski and Filipiak, 2007). In addition to combustion-related

PAHs, a range of compounds that may indicate the occurrence of terrestrial plant material are present in my samples, including retene, perylene, long chain n-alkanes, 1,7- dimethylphenanthrene, 1,2,5,6-tertramethylnaphathalene, and 1-methylphenanthrene (Fig.

2.5). Retene is structurally similar to abietane that is derived from the conifer biomarker abietic acid. Although the oldest macrofossil record of conifers was reported from the Late

Carboniferous (Scott, 1974), the earliest tracheophytes may also produce the conifer biomarkers (Romero-Sarmiento et al., 2011b). The occurrence of retene in ancient rocks has been considered to be strong evidence for the contributions of early terrestrial higher plant

(Armstroff et al., 2006; Menor-Salván et al., 2010; Romero-Sarmiento et al., 2011a; Romero-

Sarmiento et al., 2010). Perylene is believed to originate from the activity of wood-degrading fungi (Grice et al., 2009; Jiang et al., 2000). It is frequently found in sediments and crude oils

dating back to the Mesozoic but appears to be largely absent in marine sediments lacking terrestrial input and samples deposited before the rise of vascular plants (Grice et al., 2009).

Previous studies have used perylene in Devonian marine formations to reflect organic matter contributions from terrestrial higher plants (e.g., Romero-Sarmiento et al., 2011b; Tulipani et al., 2015; Song et al., 2015). In my samples, retene and perylene were detected throughout the studied section, supporting the presence of organic matter from terrestrial higher plants to the Chattanooga Shale of Alabama (Fig. 2.5b).

Other biomarkers that are less source-specific but may also indicate terrestrial plants were also evaluated. Traditionally, short-chain (C15 – C19) n-alkanes in aquatic sediments have been used to represent contributions from algae and microorganisms (Gelpi et al., 1970;

Meyers and Ishiwatari, 1993; Blokker et al., 2001), whereas long-chain alkanes (≥n-C27) are thought to originate primarily from terrestrial vascular plants (Eglinton and Hamilton, 1967;

Rieley et al., 1991). Compound-specific stable carbon isotopes of n-alkanes can further differentiate the biological sources of short versus long-chain n-alkanes (Freeman et al.,

1990; Collister et al., 1994; Sikes et al., 2009; Hockun et al., 2016). Based on this assumption, TAR was applied to represent organic matter contributions of land plants relative to marine microorganisms, and it shows an increasing trend from the lower to the upper unit in the Chattanooga Shale (Fig. 2.5a). The assumption that long chain and short-chain n- alkanes originate from different biological origins is supported by the observation that the

δ13C values of long-chain n-alkanes are more depleted than the short-chain counterparts (Fig.

2.6). It also needs to be noted that although long-chain n-alkanes are among the most widely utilized biomarkers for terrestrial higher land plants (Eglinton and Hamilton, 1967; Chevalier

et al., 2015; Yandoka et al., 2015), mosses and the non-marine microalgae, Botryococcus braunii, also produce these compounds (Lichtfouse et al., 1994; Bush and McInerney, 2013).

Despite this source ambiguity, the TAR ratios in my samples show a strong covariation with the abundance of inertinite and vitrinite (Pearson’s r=0.702, P=0.004), supporting the idea that TAR can be used as a proxy of variation in land plant input to the Chattanooga Shale.

Additionally, several PAH compounds, 1,7-dimethylphenanthrene, 1,2,5,6- tertramethylnaphathalene, and 1-methylphenanthrene, are present in my samples (Fig. 2.5;

Appendix I Fig. S2.2). These compounds are generally not considered as source-specific terrestrial plant biomarkers as they may be derived from aromatization of organic matter of various biological origins. Terpenoids structures that are prevalent among land plants is one likely source (Pichersky and Raguso, 2016), and these compounds have been previously used to indicate land plant input into marine sediments (e.g., Romero-Sarmiento et al., 2011b;

Chattopadhyay and Dutta, 2014; Armstroff et al., 2006).

Figure 2.6. δ13C values of individual normal alkanes of two representative samples from the Upper Devonian Chattanooga Shale in northeastern Alabama. The left and right panels show samples from the upper and lower units, respectively. The δ13C values of short-chain n- alkanes are more positive than those of long-chain n-alkanes, supporting my interpretation that the two homologues represent different source organisms. Error bars represent standard deviation of replicate measurements. 33

The third line of evidence lies in inorganic geochemical proxies, which show that continental weathering became more intense during the deposition of the Chattanooga Shale.

SiO2/Al2O3, CIA, and CPA all show an increasing trend from the lower to upper unit (Fig.

2.4). SiO2/Al2O3 is a useful indicator for changes of detrital input into marine environments

(Caplan and Bustin, 1996; Rimmer et al., 2004). CIA and CPA calculated from marine sediments have been widely used to evaluate the chemical weathering intensity of source areas and rocks (e.g., Visser and Young, 1990; Dingle and Lavelle, 1998; Shen et al., 2012;

Fathy et al., 2017). Both CIA and CPA are positively correlated with SiO2/Al2O3 (CIA vs.

SiO2/Al2O3: Pearson’s r=0.662, P=0.001; CPA vs. SiO2/Al2O3: Pearson’s r=0.621, P=004), and their increasing trends indicate an increase in terrigenous quartz input accompanied by the intensification of weathering on land during the deposition of the Chattanooga Shale.

Correspondingly, TOC concentrations in the upper unit are lower, probably reflecting an increasing dilution of in situ produced marine organic material caused by increasing amount of continental clastic material. It needs to be acknowledged that marine productivity declines can also lead to the TOC concentration reduction, but this interpretation does not agree with the higher δ13C values and total phosphorus in the upper unit (Figs. 2.4 and 2.5).

Additionally, the upward increases in Ti/Al and Zr/Al suggest that heavier, coarse minerals were deposited over time. This reflects a stronger force of mobilizing allocthonous minerals

(Morton and Hallsworth, 1999; Hubert, 1962) and further confirms the increased contribution of terrigenous sediments. The Ti/Al and Zr/Al ratios have also been used in other Upper

Devonian marine sedimentary sequences to indicate the relative contribution of heavy

minerals and the strength of material transportation from land to sea (Sageman et al., 2003;

Rimmer et al., 2004; Pujol et al., 2006; Riboulleau et al., 2018; Lash, 2017).

The enhancement in continental weathering during the deposition of the Chattanooga

Shale may be caused by a climatic shift to warmer and wetter conditions, yet this explanation contradicts previous data suggesting that the early and middle Famennian climate became cooler and drier globally and near the study area (Streel et al., 2000; Joachimski et al., 2009).

More likely, the intensified continental weathering is due to land plant invasion onto unvegetated, barren lands. The early development and invasion of rooted land plants to barren lands have been suggested to accelerate physical and chemical weathering of bedrocks through the Devonian (Algeo and Scheckler, 1998; Algeo et al., 2001). The roots of land plants evolved from being small (1–3 mm in diameter and up to 30 cm long), having limited geochemical effects on soils during the Early Devonian (Driese et al., 1997; Gensel et al.,

2001; Hao et al., 2010; Kenrick and Strullu-Derrien, 2014), to being large (> 2.5 cm in diameter), deep (reaching > 1 m in depth), and effective in breaking down rocks during the

Late Devonian (Algeo et al., 2001; Algeo and Scheckler, 1998; Driese et al., 1997; Retallack,

1985). My data show significant positive correlations between the proxies of terrestrial plant abundance (TAR, vitrinite and inertinite) and the proxies of continental input and weathering

(SiO2/Al2O3, CIA and CPA) (Fig. 2.7), providing further evidence supporting the interrelated connections among land plants, continental weathering and soil development. The combustion and land plant related PAH compounds show a more scattered pattern (Fig. 2.5) and do not correlate significantly with the weathering proxies, but their low concentrations make reliable quantifications difficult. Nevertheless, their occurrences throughout the studied

section strongly support that land plants were widespread in the southern Acadian land during the Late Devonian.

Figure 2.7. Cross plots of terrestrial plant proxies versus continental weathering proxies from the Upper Devonian Chattanooga Shale in northeastern Alabama. Pearson’s P and r values are presented. Red lines denote linear regression lines. The significant postive correlations between the indicators for plant inputs and weathering intensity suggest that early land plants likely intensified continental weathering.

Combining geochemical analyses, including multiple organic and inorganic geochemical proxies, my data provide the first evidence of afforestation on the southern Acadian land and the associated changes in land-ocean biogeochemical linkages during the Famennian. The occurrences of microfossils (wood fragments and spores) and biomarkers indicate that forests were present during the Famennian time on the southern Acadian land — a paleogeographic region and time that is largely absent of plant records based on conventional investigations of fossils. Several proxies (vitrinite and inertinite, TAR, retene, perylene, chemical weathering indices) further demonstrate that terrestrial plants became an increasing source of organic matter that likely intensified continental weathering and better mobilized clastic materials during the deposition of the Gassaway Member of the Chattanooga Shale.

Current knowledge of the paleogeographic distribution of Devonian forests and associated soils is primarily based on paleobiogeographic occurrences of vascular plants in fluvio-deltaic sandstone and siltstone successions (Arnold, 1939; Matten, 1974; Scheckler,

1986 a,b; Cressler, 2006; Morris et al., 2015) (Appendix I Table S2.1). My data from black shales, therefore, make an important addition to the scarce records of paleogeographic occurrences of the early forest and soil formations during the Late Devonian by presenting clear evidence of afforestation and the associated input to marine sediments in a paleogeographic area with no previously known records (i.e., the southern Acadian Orogen).

Because upper Devonian black shale units are geographically distributed in a large area from the northernmost to southernmost margins of the Appalachian Basin (Woodrow et al., 1988), they overcome the limitation due to the poor preservation of terrestrial deposits and can place the record of afforestation within a detailed biostratigraphic framework. Although the

geochemical and microscopic data do not provide diagnostic characteristics to identify specific plant taxa, accumulated information on biostratigraphic and paleobiogeographic occurrences of vascular plant fossils (e.g., Beck, 1978; Cressler, 2006; Morris et al., 2015) can offer a reasonable clue. In southcentral Euramerica along the Acadian landmass, plants with wood tissues, assigned to first shrubs or trees, appeared and soon diversified during the latest Middle to the end of the Famennian (Fig. 2.8; Appendix I Table S2.1). Because woody tissue (taller and robust stem, megallophyles, and deeper roots) is thought to be advantageous for adapting to or invading drier and more inland environments, those species are thought to be a major contributor of the earliest forests in this paleogeographic region (Morris et al.,

2015; Meyer-Berthaud et al., 2010).

To place my geochemical data in the context of the spatiotemporal evolution of the

Devonian forest, we synthesize the fossil record of vascular plants (Eutracheophytes) based on the published data, museum specimens, and the Paleobiology Database

(https://paleobiodb.org/). Figure 2.8 shows a summary of the generic- and species-level occurrence across time (three Age- or Stage-based intervals) and space (state-province- country), which characterizes the Devonian afforestation and pedogenesis (raw data of the figure available in Appendix I Table S2.1). Stratigraphically, the oldest trees and shrubs

(within euphyllophytes) appeared in the Givetian (later Middle Devonian); only a few basal eutracheophtyes, such as the trimerophytes, are known from Emisian-Eifelian strata of the northern Acadian-southern Caledonian orogens (indicated by states/region/country in dark grey on the map of Fig. 2.8). Although cladoxylopsids were the most dominant group in the

Middle Devonian, anuerophytales eventually took over the niches through the Givetian–

Frasnian transition. Paleogeographically, the two main tree-shrub bearing groups (taxa assigned to Cladoxylopsida and the progymnosperm Aneurophytopsida) were restricted in the northern Acadian region by the end of the Frasnian (see the stratigraphic occurrence with a cladogram in Fig. 2.8 and Appendix I Table S2.2). However, through the Famennian Age, the single taxon of Archaeopteridopsida, Archaeopteris–Callixylon, had spread from 2 to 12 states/provinces. This pattern is unlikely to result from differential preservation in space and time because of two reasons. First, plant fossils have been observed in formations composed of various rock types (e.g., sandstone, siltstone, shale, see Appendix I Table S2.1), suggesting that vascular plant fossil preservation in the Appalachian Basin is not selective for rock types.

Second, I observe temporal changes in plant fossils for a given formation with little change in lithology. For example, Frasnian shales (e.g., those lower members of the Chattanooga Shale, the Fynn Creek Formation) in the Southern Appalachian have no known wood fossils, but

Famennian shales of the Chattanooga Shale contain some well-preserved large tree fossils in central Tennessee (i.e., in the Gassaway Member) (Appendix I Fig. S2.10). Rather than reflecting preservation differences, these stratigraphic and paleogeographic data reveal two important trends of floral turnover that occurred on a large scale through the Frasnian–

Famennian transition near the southcentral Euramerica landmass. First, the global floral turnover had progressed southward (see the red arrow of ‘south path’ in Fig. 2.8). Second, this global southward dispersal had progressed in a relatively short time during the Frasnian–

Famennian transition. My geochemical data provide the first strong evidence for the southern end of this southward path. I further hypothesize that this transition could have extended further into the southern American landmass of Gondwana as part of global-scale

afforestation, progressing from north to south during the Famennian (e.g., Stein et al., 2007;

Hammond and Berry, 2005) especially if the physical landbridge between Euramerica and

Gondwana existed by the end of the Devonian (Golonka and Gawęda, 2012). The widespread, yet under-utilized, unfossiliferous Devonian black shales may hold the key to test this hypothesis by filling temporal and spatial gaps in the global path of afforestation and pedogenesis.

2.6 Acknowledgment

The study of the Chattanooga Shale from Alabama was partially funded by the National

Science Foundation (NSF EAR-1255724), Gulf Coast Association of Geological Societies

(GCAGS) Grant, Newton/Winefordner Scholarship from the Geological Survey of Alabama, the University of Alabama Graduate School Research and Travel Support Fund, the UA

Department of Geological Sciences W. Gary Hooks Geological Sciences Advisory Board

Fund, and the A.S. Johnson Travel Fund. I also want to thank Kim Genareau for allowing me to access to the SEM Lab.

Figure 2.8. Spatiotemporal dispersal pattern of land forests in south-central Euramerican landmass during the Devonian. Vascular plant occurrence is summarized by taxa (specific- level), time (Emsian to Famennian ages), and space (state, province, and/or country). Circle size indicates stratigraphic distribution: the older record with a larger symbol (see the legend on left bottom). Three gradients in states/region/country (dark grey, intermediate, and light grey) represent dispersal time-sequence: the older with the dark to the younger with the light color), showing a southern dispersal pattern (the ‘south path’ arrow). A simplified cladogram of higher-level euphyllophytes (selected tree and shrub taxa) shows a diversity pattern through the time (i.e., Famennian expansion). Raw data are available in Appendix I Table S2.2. 41

2.7 References Algeo, T. J., Berner, R. A., Maynard, J. B. & Scheckler, S. E. 1995. Late Devonian oceanic anoxic events and biotic crises:“rooted” in the evolution of vascular land plants. GSA today, 5, 45–66.

Algeo, T. J. & Scheckler, S. E. 1998. Terrestrial-marine teleconnections in the Devonian: links between the evolution of land plants, weathering processes, and marine anoxic events. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 353, 113–130.

Algeo, T. J. & Scheckler, S. E. 2010. Land plant evolution and weathering rate changes in the Devonian. Journal of Earth Science, 21, 75–78.

Algeo, T. J., Scheckler, S. E. & Maynard, J. B. 2001. Effects of the Middle to Late Devonian spread of vascular land plants on weathering regimes, marine biotas, and global climate. Plants invade the land: evolutionary and environmental perspectives. Columbia University Press, New York, 213–236.

Armstroff, A., Wilkes, H., Schwarzbauer, J., Littke, R. & Horsfield, B. 2006. Aromatic hydrocarbon biomarkers in terrestrial organic matter of Devonian to Permian age. Palaeogeography, Palaeoclimatology, Palaeoecology, 240, 253–274.

Arnold, C. A. 1939. Observations on fossil plants from the Devonian of eastern North America. IV. Plant remains from the Catskill Delta deposits of northern Pennsylvania and southern New York. Contributions from the Museum of Paleontology, University of Michigan, 5, 271–314.

Beck, C. B. 1978. Periastron reticulatum Unger and Aerocortex kentuckiensis, n. gen. et sp., from the New Albany Shale of Kentucky. American Journal of Botany, 221–235.

Berner, R. A. 1997. The rise of plants and their effect on weathering and atmospheric CO2. Science, 276, 544–546.

Berry, C. M. & Fairon-demaret, M. 2001. The middle Devonian flora revisited. Plants invade the land: evolutionary and environmental perspectives. Columbia University Press, New York, 120–139.

Błażejowski, B., Gieszcz, P., Brett, C. E. & Binkowski, M. 2015. A moment from before 365 Ma frozen in time and space. Scientific Reports, 5, 14191.

Blokker, P., Van bergen, P., Pancost, R., Collinson, M. E., De leeuw, J. W. & Damste, J. S. S. 2001. The chemical structure of Gloeocapsomorpha prisca microfossils: implications for their origin. Geochimica et Cosmochimica Acta, 65, 885–900. 42

Bourbonniere, R. A. & Meyers, P. A. 1996. Sedimentary geolipid records of historical changes in the watersheds and productivities of Lakes Ontario and Erie. Limnology and Oceanography, 41, 352–359.

Brom, K. R., Salamon, M. A. & Gorzelak, P. 2018. Body-size increase in crinoids following the end-Devonian mass extinction. Scientific Reports, 8, 9606.

Buggle, B., Glaser, B., Hambach, U., Gerasimenko, N. & Marković, S. 2011. An evaluation of geochemical weathering indices in loess–paleosol studies. Quaternary International, 240, 12–21.

Bush, R. T. & Mcinerney, F. A. 2013. Leaf wax n-alkane distributions in and across modern plants: implications for paleoecology and chemotaxonomy. Geochimica et Cosmochimica Acta, 117, 161–179.

Calvert, S. & Pedersen, T. 2007. Chapter fourteen elemental proxies for palaeoclimatic and palaeoceanographic variability in marine sediments: interpretation and application. Developments in Marine Geology, 1, 567–644.

Caplan, M. L. & Bustin, R. M. 1996. Factors governing organic matter accumulation and preservation in a marine petroleum source rock from the Upper Devonian to Lower Carboniferous Exshaw Formation, Alberta. Bulletin of Canadian Petroleum Geology, 44, 474–494.

Cesar, J. & Grice, K. 2017. δ13C of polycyclic aromatic hydrocarbons to establish the facies variations in a fluvial deltaic Triassic record (Dampier sub-Basin, Western Australia). Organic Geochemistry, 107, 59–68.

Chattopadhyay, A. & Dutta, S. 2014. Higher plant biomarker signatures of Early Eocene sediments of North Eastern India. Marine and Petroleum Geology, 57, 51–67.

Chevalier, N., Savoye, N., Dubois, S., Lama, M. L., David, V., Lecroart, P., Le Ménach, K. & Budzinski, H. 2015. Precise indices based on n-alkane distribution for quantifying sources of sedimentary organic matter in coastal systems. Organic Geochemistry, 88, 69–77.

Collister, J. W., Rieley, G., Stern, B., Eglinton, G. & Fry, B. 1994. Compound-specific δ13C analyses of leaf lipids from plants with differing carbon dioxide metabolisms. Organic Geochemistry, 21, 619–627.

Conant, L. C. & Swanson, V. E. 1961. Chattanooga Shale and related rocks of central Tennessee and nearby areas. Geological Survey Professional Paper, 357.

Cressler, W. L. 2006. Plant paleoecology of the Late Devonian Red Hill locality, north- central Pennsylvania, an Archaeopteris-dominated wetland plant community and early tetrapod site. Geological Society of America Special Papers, 399, 79–102.

Day, J. & Witzke, B. J. 2017. Upper Devonian biostratigraphy, event stratigraphy, and Late Frasnian Kellwasser extinction bioevents in the Iowa Basin: western Euramerica. Stratigraphy & Timescales, 2, 243–332.

Dingle, R. & Lavelle, M. 1998. Late Cretaceous–Cenozoic climatic variations of the northern Antarctic Peninsula: new geochemical evidence and review. Palaeogeography, Palaeoclimatology, Palaeoecology, 141, 215–232.

Driese, S. G., Mora, C. I. & Elick, J. M. 1997. Morphology and taphonomy of root and stump casts of the earliest trees (Middle to Late Devonian), Pennsylvania and New York, USA. Palaios, 524–537.

Eglinton, G. & Hamilton, R. J. 1967. Leaf epicuticular waxes. Science, 156, 1322–1335.

Fabbri, D., Sangiorgi, F. & Vassura, I. 2005. Pyrolysis–GC–MS to trace terrigenous organic matter in marine sediments: a comparison between pyrolytic and lipid markers in the Adriatic Sea. Analytica Chimica Acta, 530, 253–261.

Falcon-Lang, H., 1998. The impact of wildfire on an Early Carboniferous coastal environment, North Mayo, Ireland. Palaeogeography Palaeoclimatology Palaeoecology, 139, 121–138.

Fathy, D., Wagreich, M., Zaki, R., Mohamed, R. S. & Gier, S. 2017. Geochemical fingerprinting of maastrichtian oil shales from the central eastern desert, Egypt: Implications for provenance, tectonic setting, and source area weathering. Geological Journal, 53, 2597–2612

Freeman, K. H., Boreham, C. J., Summons, R. E. & Hayes, J.M. 1990. Evidence from GC- MS carbon-isotopic measurements for multiple origins of sedimentary hydrocarbons. Nature, 353, 254–256.

Galtier, J. & Meyer-Berthaud, B. 1996. The early seed plant Tristichia tripos (Unger) comb. nov. from the Lower Carboniferous of Saalfeld, Thuringia. Review of Palaeobotany and Palynology, 93, 299–315.

Gelpi, E., Schneider, H., Mann, J. & Oro, J. 1970. Hydrocarbons of geochemical significance in microscopic algae. Phytochemistry, 9, 603–612.

Gensel, P.G., Kotyk, M.E. & Basinger, J.F., 2001. Morphology of Above-and Below-Ground Structures in Early Devonian (Pragian–Emsian) Plants. Plants invade the land: evolutionary and environmental perspectives, p.83.

Golonka, J. & Gawęda, A. 2012. Plate tectonic evolution of the southern margin of Laurussia in the Paleozoic. Tectonics recent advances. InTech, 261–282.

Grice, K., Lu, H., Atahan, P., Asif, M., Hallmann, C., Greenwood, P., Maslen, E., Tulipani, S., Williford, K. & Dodson, J. 2009. New insights into the origin of perylene in geological samples. Geochimica Et Cosmochimica Acta, 73, 6531–6543.

Hammond, S. E. & Berry, C. M. 2005. A new species of Tetraxylopteris (Aneurophytales) from the Devonian of Venezuela. Botanical Journal of the Linnean Society, 148, 275– 303.

Hansma, J., Tohver, E., Yan, M., Trinajstic, K., Roelofs, B., Peek, S., Slotznick, S. P., Kirschvink, J., Playton, T. & Haines, P. 2015. Late Devonian carbonate magnetostratigraphy from the Oscar and Horse Spring Ranges, Lennard Shelf, Canning Basin, Western Australia. Earth and Planetary Science Letters, 409, 232– 242.

Hao, S., Xue, J., Guo, D. & Wang, D. 2010. Earliest rooting system and root: shoot ratio from a new Zosterophyllum plant. New Phytologist, 185, 217–225.

Hass, W. H. 1956. Age and correlation of the Chattanooga Shale and the Maury Formation. USGS Professional Paper, 286, 2330–7102

Haworth, M., Belcher, C. M., Killi, D., Dewhirst, R. A., Materassi, A., Raschi, A. & Centritto, M. 2018. Impaired photosynthesis and increased leaf construction costs may induce floral stress during episodes of global warming over macroevolutionary timescales. Scientific Reports, 8, 6206.

Hillbun, K., Playton, T. E., Tohver, E., Ratcliffe, K., Trinajstic, K., Roelofs, B., Caulfield- kerney, S., Wray, D., Haines, P. & Hocking, R. 2015. Upper Kellwasser carbon isotope excursion pre-dates the F–F boundary in the Upper Devonian Lennard Shelf carbonate system, Canning Basin, Western Australia. Palaeogeography, Palaeoclimatology, Palaeoecology, 438, 180–190. 45

Hockun, K., Mollenhauer, G., Ho, S. L., Hefter, J., Ohlendorf, C., Zolitschka, B., Mayr, C., Lücke, A. & Schefuß, E. 2016. Using distributions and stable isotopes of n-alkanes to disentangle organic matter contributions to sediments of Laguna Potrok Aike, Argentina. Organic Geochemistry, 102, 110–119.

Hubert, J. F. 1962. A zircon-tourmaline-rutile maturity index and the interdependence of the composition of heavy mineral assemblages with the gross composition and texture of sandstones. Journal of Sedimentary Research, 32, 440–450.

Jahren, A. H., Arens, N. C., Sarmiento, G., Guerrero, J. & Amundson, R. 2001. Terrestrial record of methane hydrate dissociation in the Early Cretaceous. Geology, 29, 159– 162.

Jiang, C., Alexander, R., Kagi, R. I. & Murray, A. P. 2000. Origin of perylene in ancient sediments and its geological significance. Organic Geochemistry, 31, 1545–1559.

Joachimski, M., Breisig, S., Buggisch, W., Talent, J., Mawson, R., Gereke, M., Morrow, J., Day, J. & Weddige, K. 2009. Devonian climate and reef evolution: insights from oxygen isotopes in apatite. Earth and Planetary Science Letters, 284, 599–609.

Jones, T. P. & Chaloner, W. G. 1991. Fossil charcoal, its recognition and palaeoatmospheric significance. Palaeogeography, Palaeoclimatology, Palaeoecology, 97, 39–50.

Kaiho, K., Yatsu, S., Oba, M., Gorjan, P., CASIER, J. G. & IKEDA, M. 2013b. A forest fire and soil erosion event during the Late Devonian mass extinction. Palaeogeography Palaeoclimatology Palaeoecology, 392, 272–280.

Kennedy, K. L., Gibling, M. R., Eble, C. F., Gastaldo, R. A., Gensel, P. G., Werner- zwanziger, U. & Wilson, R. A. 2013. Lower Devonian coaly shales of northern New Brunswick, Canada: plant accumulations in the early stages of Terrestrial colonization. Journal of Sedimentary Research, 83, 1202–1215.

Kenrick, P. & Strullu-derrien, C. 2014. The origin and early evolution of roots. Plant physiology, 166, 570–580.

Killops, S.D. & Killops, V.J., 2013. Introduction to Organic Geochemistry. John Wiley & Sons.

Laflamme, R. E. & Hites, R. A. 1978. The global distribution of polycyclic aromatic hydrocarbons in recent sediments. Geochimica et Cosmochimica Acta, 42, 289–303.

Lash, G. G. 2017. A multiproxy analysis of the Frasnian–Famennian transition in western New York State, USA. Palaeogeography, Palaeoclimatology, Palaeoecology, 473, 108–122.

Le Hir, G., Donnadieu, Y., Goddéris, Y., Meyer-Berthaud, B., Ramstein, G. & Blakey, R.C., 2011. The climate change caused by the land plant invasion in the Devonian. Earth and Planetary Science Letters, 310, 203–212.

Lichtfouse, E., Derenne, S., Mariotti, A. & Largeau, C. 1994. Possible algal origin of long- chain odd n-alkanes in immature sediments as revealed by distributions and carbon isotope ratios. Organic Geochemistry, 22, 1023–1027.

Long, A. G. 1961. XIII.—On the Structure of Deltasperma fouldenense gen. et sp. nov., and Camptosperma berniciense gen. et sp. nov., petrified seeds from the Calciferous Sandstone Series of Berwickshire. Earth and Environmental Science Transactions of The Royal Society of Edinburgh, 64, 281–295.

Lu, Y. & Meyers, P. A. 2009. Sediment lipid biomarkers as recorders of the contamination and cultural eutrophication of Lake Erie, 1909–2003. Organic Geochemistry, 40, 912–921.

Marynowski, L. & Filipiak, P. 2007. Water column euxinia and wildfire evidence during deposition of the Upper Famennian Hangenberg event horizon from the Holy Cross Mountains (central Poland). Geological Magazine, 144, 569–595.

Matten, L. C. 1974. The Givetian flora from Cairo, New York: Rhacophyton, Triloboxylon and Cladoxylon. Botanical Journal of the Linnean Society, 68, 303-318.

Matten, L. C., Lacey, W. S., May, B. I. & Lucas, R. C. 1980. A megafossil flora from the uppermost Devonian near Ballyheigue, Co. Kerry, Ireland. Review of Palaeobotany and Palynology, 29, 241–251.

Menor-Salván, C., Tornos, F., Fernández-Remolar, D. & Amils, R. 2010. Association between catastrophic paleovegetation changes during Devonian–Carboniferous boundary and the formation of giant massive sulfide deposits. Earth and Planetary Science Letters, 299, 398–408.

Meyer-Berthaud, B., Soria, A. & Decombeix, A.L. 2010. The land plant cover in the Devonian: a reassessment of the evolution of the tree habit. Geological Society, London, Special Publications, 339, 59–70.

Meyers, P. A. & Ishiwatari, R. 1993. Lacustrine organic geochemistry—an overview of indicators of organic matter sources and diagenesis in lake sediments. Organic Geochemistry, 20, 867–900. 47

Morris, J.L., Leake, J.R., Stein, W.E., Berry, C.M., Marshall, J.E., Wellman, C.H., Milton, J.A., Hillier, S., Mannolini, F., Quirk, J, & Beerling, D.J., 2015. Investigating Devonian trees as geo-engineers of past climates: linking palaeosols to palaeobotany and experimental geobiology. Palaeontology, 58, 787–801.

Morton, A. C. & Hallsworth, C. R. 1999. Processes controlling the composition of heavy mineral assemblages in sandstones. Sedimentary Geology, 124, 3–29.

Nesbitt, H. & Young, G.M., 1982. Early Proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature, 299, 715.

Nichols, G. & Jones, T. 1992. Fusain in Carboniferous shallow marine sediments, Donegal, Ireland: the sedimentological effects of wildfire. Sedimentology, 39, 487–502.

Ouyang, T., Li, M., Zhao, X., Zhu, Z., Tian, C., Qiu, Y., Peng, X. & Hu, Q. 2016. Sensitivity of Sediment Magnetic Records to Climate Change during Holocene for the Northern South China Sea. Frontiers in Earth Science, 4, 54.

Over, D. J. 2007. Conodont biostratigraphy of the Chattanooga Shale, Middle and Upper Devonian, southern Appalachian Basin, eastern United States. Journal of Paleontology, 81, 1194–1217.

Pashin, J. C., Kopaska-Merkel, D., Arnold, A. & Mcintyre, M. 2011. Geological foundation for production of natural gas from diverse shale formations. Research Partnership to Secure Energy for America (RPSEA), Final Report, 7122, 156.

Percival, L., Davies, J., Schaltegger, U., De Vleeschouwer, D., Da Silva, A.C. & Föllmi, K. 2018. Precisely dating the Frasnian–Famennian boundary: implications for the cause of the Late Devonian mass extinction. Scientific Reports, 8, 9578.

Petersen, H. I., Schovsbo, N. H. & Nielsen, A. T. 2013. Reflectance measurements of zooclasts and solid bitumen in Lower Paleozoic shales, southern Scandinavia: correlation to vitrinite reflectance. International Journal of Coal Geology, 114, 1–18.

Pichersky, E. & Raguso, R. A. 2016. Why do plants produce so many terpenoid compounds? New Phytologist. 220, 692–702.

Pickel, W., Kus, J., Flores, D., Kalaitzidis, S., Christanis, K., Cardott, B.J., Misz-Kennan, M., Rodrigues, S., Hentschel, A., Hamor-Vido, M. & Crosdale, P., 2017. Classification of liptinite–ICCP System 1994. International Journal of Coal Geology, 169, 40–61.

Prestianni, C. & Gerrienne, P. 2010. Early seed plant radiation: an ecological hypothesis. Geological Society, London, Special Publications, 339, 71–80. 48

Pujol, F., Berner, Z. & Stüben, D. 2006. Palaeoenvironmental changes at the Frasnian/Famennian boundary in key European sections: Chemostratigraphic constraints. Palaeogeography, Palaeoclimatology, Palaeoecology, 240, 120–145.

Retallack, G. J. 1997. Early forest soils and their role in Devonian global change. Science, 276, 583–585.

Retallack, G.J. & Lawson, J.D., 1985. Fossil soils as grounds for interpreting the advent of large plants and animals on land. Philosophical Transactions of the Royal Society of London. B, Biological Sciences, 309, 105–142.

Riboulleau, A., Spina, A., Vecoli, M., Riquier, L., Quijada, M., Tribovillard, N. & Averbuch, O., 2018. Organic matter deposition in the Ghadames Basin (Libya) during the Late Devonian—a multidisciplinary approach. Palaeogeography, Palaeoclimatology, Palaeoecology, 497, 37-51.

Richardson, J.B., 1986. Silurian and Devonian spore zones of the Old Red Sandstone continent and adjacent regions. Geological Survey of Canada, Bulletin, 354, l–21.

Rieley, G., Collier, R.J., Jones, D.M., Eglinton, G., Eakin, P.A. & Fallick, A.E., 1991. Sources of sedimentary lipids deduced from stable carbon-isotope analyses of individual compounds. Nature, 352, 425–427.

Rietveld, H. 1969. A profile refinement method for nuclear and magnetic structures. Journal of Applied Crystallography, 2, 65–71.

Rimmer, S. M., Hawkins, S. J., Scott, A. C. & Cressler, W. L. 2015. The rise of fire: Fossil charcoal in late Devonian marine shales as an indicator of expanding terrestrial ecosystems, fire, and atmospheric change. American Journal of Science, 315, 713– 733.

Rimmer, S. M., Thompson, J. A., Goodnight, S. A. & Robl, T. L. 2004. Multiple controls on the preservation of organic matter in Devonian–Mississippian marine black shales: geochemical and petrographic evidence. Palaeogeography, Palaeoclimatology, Palaeoecology, 215, 125–154.

Romero-Sarmiento, M.F., Riboulleau, A., Vecoli, M., Laggoun-Défarge, F. & Versteegh, G. J. M. 2011a. Aliphatic and aromatic biomarkers from Carboniferous coal deposits at Dunbar (East Lothian, Scotland): Palaeobotanical and palaeoenvironmental significance. Palaeogeography, Palaeoclimatology, Palaeoecology, 309, 309–326.

Romero-Sarmiento, M.F., Riboulleau, A., Vecoli, M. & Versteegh, G.J.M. 2011b. Aliphatic and aromatic biomarkers from Gondwanan sediments of Late Ordovician to Early 49

Devonian age: An early terrestrialization approach. Organic Geochemistry, 42, 605– 617.

Romero-Sarmiento, M.F., Riboulleau, A., Vecoli, M. & Versteegh, G.J.M. 2010. Occurrence of retene in upper Silurian–lower Devonian sediments from North Africa: Origin and implications. Organic Geochemistry, 41, 302–306.

Sageman, B.B., Murphy, A.E., Werne, J.P., Ver Straeten, C.A., Hollander, D.J. & Lyons, T.W., 2003. A tale of shales: the relative roles of production, decomposition, and dilution in the accumulation of organic-rich strata, Middle–Upper Devonian, Appalachian basin. Chemical Geology, 195, 229–273.

Scheckler, S. E. 1986a. Floras of the Devonian-Mississippian transition. Studies in Geology, Notes for a Short Course, 15, 81–96.

Scheckler, S. E. 1986b. Geology, floristics and paleoecology of Late Devonian coal swamps from Appalachian Laurentia (USA). Annales de la Société géologique de Belgique, 109, 209–222

Scheckler, S. E., Skog, J. E. & Banksc, H. P. 2006. Langoxylon asterochlaenoideum Stockmans: anatomy and relationships of a fern-like plant from the Middle Devonian of Belgium. Review of Palaeobotany and Palynology, 142, 193–217.

Schieber, J. 1994. Evidence for high-energy events and shallow-water deposition in the Chattanooga Shale, Devonian, central Tennessee, USA. Sedimentary Geology, 93, 193–208.

Schieber, J. 1998. Developing a sequence stratigraphic framework for the Late Devonian Chattanooga Shale of the southeastern USA: relevance for the Bakken Shale. Williston Basin Symposium.

Schopf, J. M. & Schwietering, J. F. 1970. Foerstia zone of the Ohio and Chattanooga shales.

Scott, A. 1974. The earliest conifer. Nature, 251, 707.

Scott, A. C. & Glasspool, I. J. 2007. Observations and experiments on the origin and formation of inertinite group macerals. International Journal of Coal Geology, 70, 53–66.

Scott, L. 2002. Microscopic charcoal in sediments: Quaternary fire history of the grassland and savanna regions in South Africa. Journal of Quaternary Science, 17, 77–86.

Shen, J., Algeo, T., Zhou, L., Feng, Q., Yu, J. & Ellwood, B. 2012. Volcanic perturbations of the marine environment in South China preceding the latest Permian mass extinction and their biotic effects. Geobiology, 10, 82–103.

Sikes, E. L., Uhle, M. E., Nodder, S. D. & Howard, M. E. 2009. Sources of organic matter in a coastal marine environment: evidence from n-alkanes and their δ13C distributions in the Hauraki Gulf, New Zealand. Marine Chemistry, 113, 149–163.

Silva, T. R., Lopes, S. R., Spörl, G., Knoppers, B. A. & Azevedo, D. A. 2012. Source characterization using molecular distribution and stable carbon isotopic composition of n-alkanes in sediment cores from the tropical Mundaú–Manguaba estuarine–lagoon system, Brazil. Organic Geochemistry, 53, 25–33.

Song, D., He, D., Qi, X., Wang, S. & Li, D. 2015. Occurrence and composition of solid bitumens from the Bulonggoer Devonian paleo-oil reservoir, North Xinjiang, China. Organic Geochemistry, 83, 1–15.

Spalletta, C., Perri, M. C., Over, D. J. & Corradini, C. 2017. Famennian (Upper Devonian) conodont zonation: revised global standard. Bulletin of Geosciences, 92, 31–57.

Stein, W. E., Mannolini, F., Hernick, L. V., Landing, E. & Berry, C. M. 2007. Giant cladoxylopsid trees resolve the enigma of the Earth's earliest forest stumps at Gilboa. Nature, 446, 904–908.

Streel, M., Caputo, M. V., Loboziak, S. & Melo, J. H. G. 2000. Late Frasnian–Famennian climates based on palynomorph analyses and the question of the Late Devonian glaciations. Earth-Science Reviews, 52, 121–173.

Sun, Y., Chen, Z., Xu, S. & Cai, P. 2005. Stable carbon and hydrogen isotopic fractionation of individual n-alkanes accompanying biodegradation: evidence from a group of progressively biodegraded oils. Organic Geochemistry, 36, 225–238.

Sweet, W. 1988. A quantitative conodont biostratigraphy for the Lower Triassic. Senckenbergiana Lethaea, 69, 253–273.

Taylor, G.H., Teichmüller, M., Davis, A.C.F.K., Diessel, C.F.K., Littke, R. and Robert, P., 1998. Organic petrology.

Turgeon, S. C., Creaser, R. A. & Algeo, T. J. 2007. Re–Os depositional ages and seawater Os estimates for the Frasnian–Famennian boundary: implications for weathering rates, land plant evolution, and extinction mechanisms. Earth and Planetary Science Letters, 261, 649–661.

Uhl, D. & Kerp, H. 2003. Wildfires in the Late Palaeozoic of Central Europe–The Zechstein (Upper Permian) of NW-Hesse (Germany). Palaeogeography, Palaeoclimatology, Palaeoecology, 199, 1–15.

Visser, J. N. & Young, G. M. 1990. Major element geochemistry and paleoclimatology of the Permo-Carboniferous glacigene Dwyka Formation and postglacial mudrocks in southern Africa. Palaeogeography, Palaeoclimatology, Palaeoecology, 81, 49–57.

Wang, D., Liu, L., Guo, Y., Xue, J. & Meng, M. 2015. A Late Devonian fertile organ with seed plant affinities from China. Scientific Reports, 5, 10736.

Woodrow, D.L., Dennison, J.M., Ettensohn, F.R., Sevon, W.T. & Kirchgasser, W.T., 1988. ‘Middle and Upper Devonian stratigraphy and paleogeography of the central and southern Appalachians and eastern Midcontinent, USA’ in McMillan, N. J., Embry, A. F. & Glass, G. J. (eds) Devonian of the World: Proceddings of the Second International Symposium on the Devonian System, Canada Society of Petroleum Geologists Memoir 14, 1, 277–301.

Yandoka, B.M.S., Abdullah, W.H., Abubakar, M.B., Hakimi, M.H. & Adegoke, A.K. 2015. Geochemical characterisation of Early Cretaceous lacustrine sediments of Bima Formation, Yola Sub-basin, Northern Benue Trough, NE Nigeria: Organic matter input, preservation, paleoenvironment and palaeoclimatic conditions. Marine and Petroleum Geology, 61, 82–94.

2.8 Appendix I

2.8.1 Non-plant Lipid Biomarkers

Notes on branched alkanes: Monomethylalkanes (MMAs) were identified in the range of

C12–C20, with the dominance of 7-, 5-, 4-, 2- and 3-methyl- isomers (Fig. S2.1). These compounds are generally thought to represent the inputs of bacteria (Gelpi et al., 1970; Shiea et al., 1990; Lu et al., 2003).

Figure S2. 1. m/z 85 mass chromatogram of a representative sample of the Upper Devonian Chattanooga Shale in northeast Alabama. Red square represents normal alkane. Blue dot represents isoprenoid.

Notes on PAHs: Abundant constituents include phenanthrene and alkylated phenanthrenes, alkylated naphthalenes, chrysene, alkylated chrysene, and perylene (Fig. S2.2). Among them,

1,7-dimethylphenanthrene, 1,2,5,6-tetramethylnaphthalene, 1-methylphenanthrene, retene

and perylene are indicators for contributions from land plants. Pyrene, benzo(a)pyrene, benzo(e)pyrene and coronene are indicators of plant combustion.

Figure S2. 2. Total ion current (TIC), partial m/z 202, m/z 219/234, m/z 252 and m/z 300/324 chromatograms for one representative sample of the Upper Devonian Chattanooga Shale in northeast Alabama. Black squares represent methylchrysene isomers. Triangles represent the methylphenanthrene isomers, and circles represent dimethylphenanthrene isomers.

Notes on terpenoids: Bicyclic sesquiterpenoids (m/z =123), steranes (m/z=218,217) and hopanes (m/z =191) were also detected (Figs. S2.3, S2.4 and S2.5). These compounds were not quantified because of their low abundances.

Figure S2. 3. m/z 123 chromatograms showing the distribution of bicyclic sesquiterpenoids for representative samples from the upper unit (a) and the low unit (b) of the Chattanooga Shale in northeast Alabama.

Figure S2. 4. m/z 217 and 218 chromatograms showing the distribution of steranes for a representative sample from the Chattanooga Shale in northeast Alabama.

Figure S2. 5. m/z 191 chromatogram showing the distribution of hopanes for a representative sample from the Chattanooga Shale in northeast Alabama.

2.8.2 Microfossils

Figure S2. 6. Scanning electron microscope (SEM) images of conodont P1 element recovered from the Chattanooga Shale in northeast Alabama and occurrence distance above base of shale. Scale bar corresponds to 0.25 mm. a: Palmatolepis superlobata, 1.25m; b: Palmatolepis lobicornis, 1.75m; c: Palmatolepis lobicornis, 1.25m; d: Palmatoelpis lobicornis, 1.75m; e: Palmatolepis superlobata, 1.00m; f: Palmatoelpis minuta, 1.25m; g: Palmatolepis glabra unca, 1.70m; h: Palmatolepis glabra ssp, 1.00m; i: Palmatolepis glabra unca, 1.25m.

Figure S2. 7. Energy-dispersive X-ray spectroscopy (EDS) spectra of land plant remains from the Upper Devonian Chattanooga Shale in northeastern Alabama. The associated SEM images are in Fig. 2.3 (corresponding to the labels 1a–2d).

Figure S2. 8. Photomicrographs (white reflected light, oil immersion) of samples from the lower unit (a) and the upper unit (b) of the Chattanooga Shale in northeast Alabama. V=Vitrinite; I=Inertinite.

2.8.3 Mineralogy

Figure S2. 9. SEM images of mineral crystals from the lower unit (a) and the upper unit (b) of the Chattanooga Shale in northeast Alabama. Q=quartz; P= pyrite; C= clay minerals. Those quartz crystals are eroded and turned into subangular grains and extensively coated by clay minerals, indicating that quartz in my samples are primarily detrital grains.

2.8.4 Macro-size Plant Fossils

Table S2. 1. A summary of stratigraphic and paleogeographic occurrences of Devonian land forests in the southcentral Euramerica. Vascular plant taxa (Eutracheophytes) with wood tissues, such as trees and shrub are selected. Same data are used in Fig. 2.8. Raw data are listed in Table S2.2. A. Generic and specific counts occurred in major taxonomic groups. Lycopsida Cladoxylopsids Anuerophytales Archaeopterids genera species genera species genera species genera species Famennian 3 3 4 4 1 1 1(2) 14 Frasnian 1 1 3 3 3 3 1(2) 13 Middle Devonian 0 0 9 16 1 4 2 2 Total 4 4 13 21 1 6 1(2) 20

Spermatophytes Equisetophyta Lepidocarpopsida genera species genera species genera species Famennian 3 3 2 2 1 1 Frasnian 0 0 1 2 0 0 Middle Devonian 0 0 0 0 0 0 Total 3 3 3 4 1 1

B. Generic and specific counts occurred in each region (state/province for US/Canada and country for Europe) Lycopsida Cladoxylopsids Anuerophytales Archaeopterids Famennian 3 2 2 11 Frasnian 1 1 1 7 Middle Devonian 5 5 4 1

Spermatophytes Equisetophyta Lepidocarpopsida Famennian 3 2 1 Frasnian 0 1 0 Middle Devonian 0 0 0

Table S2. 2. Occurrence of land plants with wood tissues from Devonian strata of the central Euramerica. Same data are used in Fig. 2.8. Abbreviations for paleogeographic region and state/province; CAN: Canada; US: United States; BE: Belgium; GER: Germany; CZE: Czech Republic; NOR: Norway App Basin: Appalachian Basin; NY: New York; PA: Pennsylvania; WV: West Virginia. An asterisk symbol (*) in the Formation column indicates a black shale unit.

EMSIAN (Early Devonian) Higher taxa 1 Higher taxa Genus Species Paleogeographic Country State/ Formation References 2 region Province Granoff et al.,

Lycopodiophytes Radiatopses Pertica varia Avalonia CAN Quebec Battery Point 1976 New Andrews et al.,

Lycopodiophytes Radiatopses Pertica dalhousii Avalonia CAN Brunswick Campbellton 1975 Battery Granoff et al.,

Lycopodiophytes Radiatopses Trimerophyton robustius Avalonia CAN Quebec Point(?) 1976 Andrews et al.,

Lycopodiophytes Radiatopses Pertica quadrifaria Avalonia US Maine Trout Valley 1975 Andrews et al.,

Lycopodiophytes Radiatopses Pertica quadrifaria Avalonia US Maine Trout Valley 1975

Givetian (Middle Devonian) Higher taxa 1 Higher taxa 2 Genus Species Paleo- Country State/ Formation References geographic Province region Cladoxylopsida Pseudosporochnales Calamophyton bicephalum App Basin US NY Ashokan Krausel and Weyland, 1926, Kräusel and Weyland, 1929, Leclercq, 1969, Schweitzer, 1973 Cladoxylopsida Pseudosporochnales Calamophyton bicephalum Avalonia BE ? ? Leclercq, 1969 Cladoxylopsida Pseudosporochnales Calamophyton primaevum Avalonia(?) GER Elberfeld Honseler Leclercq and Schichten Andrews, 1960 Cladoxylopsida Pseudosporochnales Calamophyton primaevum Avalonia BE Brabant, Bois de Leclercq, 1969 Ronquieres Bordeaux Cladoxylopsida Pseudosporochnales Cladoxylon hueberi App Basin US NY Kiskatom Stein and Hueber, 1989 Cladoxylopsida Pseudosporochnales Cladoxylon scoparium Avalonia GER ? Honseler Krausel and Schichten Weyland, 1926 Cladoxylopsida Pseudosporochnales Cladoxylon scoparium Avalonia GER Wuppertals Brandenberg Krausel and Schichten Weyland, 1926 Cladoxylopsida Pseudosporochnales Cladoxylon sp. App Basin US NY Kiskatom Matten, 1974, Scheckler, 1975 Cladoxylopsida Pseudosporochnales Wattieza givetiana Avalonia BE Brabant Fromelennes Stockmans, 1968 Cladoxylopsida Pseudosporochnales Wattieza sp. App Basin US NY Oneonta Stein et al., 2007 Cladoxylopsida Pseudosporochnales Rellimia thomsonii App Basin US NY Panther Dannenhoffer and

Mountain Bonamo, 1989 Cladoxylopsida? Eospermatopteris erianus App Basin US NY Moscow Stein et al., 2012 Cladoxylopsida? Eospermatopteris sp. App Basin US NY Kiskatom Stein et al., 2012 Cladoxylopsida Pseudosporochnales Hyenia banksii App Basin US NY Bellvale Arnold, 1941 Sandstone Cladoxylopsida Pseudosporochnales Hyenia elegans Avalonia BE Brabant Bois de Fairon-Demaret Bordeaux and Berry, 2000 Cladoxylopsida Pseudosporochnales Hyenia elegans Avalonia(?) GER Germany Honseler Fairon-Demaret Schichten and Berry, 2000 Cladoxylopsida Pseudosporochnales Hyenia vogtii Baltica NOR Svalbard Wide Bay Fairon-Demaret and Berry, 2000 Cladoxylopsida Pseudosporochnales Lorophyton goense Avalonia BE ? Pepinster Fairon-Demaret and Li, 1993 Cladoxylopsida Pseudosporochnales Pseudosporochnu chlupaci Baltica CZE ? Roblin Obrhel, 1961 s Schichten; Kacak Schichten; Srbsko- Schichten Cladoxylopsida Pseudosporochnales Pseudosporochnu krejcii Avalonia BE Brabant Fromelennes Stockmans, 1968 s Cladoxylopsida Pseudosporochnales Pseudosporochnu krejcii Avalonia BE Ronquieres Bois de Stockmans, 1968 s Bordeaux Cladoxylopsida Pseudosporochnales Pseudosporochnu nodosus Avalonia BE Belgium shale and Leclercq and s sandstone Banks, 1962, Berry and Edwards, 1997

Cladoxylopsida Pseudosporochnales Pseudosporochnu nodosus Avalonia BE Ronquieres Bois de Berry and Edwards, s Bordeaux 1997 Cladoxylopsida Pseudosporochnales Pseudosporochnu sp. Avalonia GER Hohen Venns Friiesenrather Neumann- s Schichten Mahlkau, 1965 Cladoxylopsida Pseudosporochnales Pseudosporochnu sp. App Basin US NY Kiskatom Stein and Hueber, s 1989 Matten, 1974 Cladoxylopsida Pseudosporochnales Pseudosporochnu sp. App Basin US NY Ashokan Stein and Hueber, s 1989, Matten, 1974 Cladoxylopsida Pseudosporochnales Pseudosporochnu verticilatus Baltica CZE Bohemia Roblin Obrhel, 1961 s Schichten, Kacak Schichten, Srbsko- Schichten Cladoxylopsida Pseudosporochnales Xenocladia medullosina App Basin US NY Tully Pyrite Arnold, 1952 Cladoxylopsida Pseudosporochnales Xenocladia medullosina App Basin US NY Ludlowville Arnold, 1952 Progymnosperm Aneurophytopsida Aneurophyton bohemicum Baltica CZE Bohemia Roblin Obrhel, 1961 Schichten, Kacak Schichten, Srbsko- Schichten Progymnosperm Aneurophytopsida Aneurophyton bohemicum Avalonia GER Elberfeld Honseler Obrhel, 1961 Schichten Progymnosperm Aneurophytopsida Aneurophyton furcatum Avalonia BE Brabant Fromelennes Stockmans, 1968

Progymnosperm Aneurophytopsida Aneurophyton germanicum Avalonia BE Belgium Evieux? Lessuise and Fairon-Demaret, 1980 Progymnosperm Aneurophytopsida Aneurophyton germanicum Avalonia GER Germany Honseler Scott, 1926 Schichten Kräusel and Weyland, 1932 Progymnosperm Aneurophytopsida Aneurophyton germanicum Avalonia Germany Wuppertal Brandenberg Krausel and Schichten Weyland, 1926 Progymnosperm Aneurophytopsida Aneurophyton germanicum Avalonia Germany Wuppertal formations at Schweitzer and Wuppertal- Matten, 1983 Elberfeld Locality Progymnosperm Aneurophytopsida Aneurophyton hallii App Basin US NY Ludlowville Arnold, 1940 Progymnosperm Aneurophytopsida Aneurophyton hallii App Basin US NY Tully Pyrite Arnold, 1940 Progymnosperm Aneurophytopsida Aneurophyton sp. App Basin US NY Moscow Stein, 2002 PBDB 28414 Progymnosperm Archaeopteridopsida Archaeopteris obtuse App Basin US NY Bellvale Arnold, 1941 Sandstone Progymnosperm Archaeopteridopsida Callixylon petryi App Basin US NY Sherbure Arnold, 1941 Sandstone Progymnosperm Archaeopteridopsida Triloboxylon arnoldii App Basin US NY Kiskatom Matten, 1974

Frasnian (Late Devonian) Higher taxa Higher taxa 2 Genus Species Paleo- Country State/ Formatio References 1 geographic Province n region Cladoxylopsida Pseudosporochnales Calamophyton primaevum Avalonia BE Brabant, Bois de Leclercq and Andrews, Ronquieres Bordeaux 1960, Bonamo and Banks, 1966 Cladoxylopsida Pseudosporochnales Hyenia elegans Avalonia BE Brabant Bois de Leclercq and Andrews, Bordeaux 1960, Bonamo and Banks, 1966 Cladoxylopsida Pseudosporochnales Pseudosporochnus krejcii Avalonia BE Ronquieres Bois de Leclercq and Andrews, Bordeaux 1960, Bonamo and Banks, 1966 Cladoxylopsida Pseudosporochnales Pseudosporochnus nodosus Avalonia BE Ronquieres Bois de Leclercq and Andrews, Bordeaux 1960, Bonamo and Banks, 1966 Lycopsida isoetalean Lepidosigillaria whitei Acadian US NY ? Grierson and Banks, 1963 Equisetophyta Sphenopsida Calamospora atava Acadian US Maryland Foreknobs Curry, 1975 Equisetophyta Sphenopsida Calamospora nigrata Acadian US Maryland Foreknobs Curry, 1975 Cladoxylopsida Pseudosporochnales Cladoxylon dawsoni Acadian US NY Genundew Beck, 1953 a Limestone Cladoxylopsida? Eospermatopteris sp. Acadian US NY Oneonta, Stein et al., 2012 Stony Clove Cladoxylopsida Pseudosporochnales Rhymokalon trichium Acadian US NY Oneonta Scheckler, 1975

Progymnosperm Aneurophytopsida Aneurophyton germanicum Acadian US NY Delaware Serlin and Banks, 1978 River Flags Progymnosperm Aneurophytopsida Aneurophyton hallii Acadian US NY Ithaca Stein, 2002 Progymnosperm Aneurophytopsida Aneurophyton rachides Acadian US NY Oneonta Stein, 2002 Progymnosperm Archaeopteridopsida Archaeopteris fissilis N. CAN Nunavut black-gray Stein, 2002 Euramerica shale* Progymnosperm Archaeopteridopsida Archaeopteris gaspiensis Avalonia CAN Quebec Hugh Dawson, 1882 Miller Cliffs Progymnosperm Archaeopteridopsida Archaeopteris halliana Avalonia CAN Quebec Escuminac Dawson, 1882 Progymnosperm Archaeopteridopsida Archaeopteris halliana Acadian US NY Chemung; Carluccio et al., 1966, Katzberg Traverse and Schuyler, 1994 Progymnosperm Archaeopteridopsida Archaeopteris halliana Acadian US PA Chemung Arnold, 1939 Progymnosperm Archaeopteridopsida Archaeopteris hibernica Acadian US NY Katzberg Carluccio et al., 1966 Progymnosperm Archaeopteridopsida Archaeopteris jacksoni Avalonia CAN Quebec Dawson, 1882, Chaloner and Pettitt, 1964 Progymnosperm Archaeopteridopsida Archaeopteris macilenta Acadian US NY Chemung; Carluccio et al., 1966, Katzberg; Arnold, 1939, Beck, 1960 Oneonta; Stony Clove Progymnosperm Archaeopteridopsida Archaeopteris minor Acadian US PA Chemung Arnold, 1939; Progymnosperm Archaeopteridopsida Archaeopteris obtusa N. CAN Nunavut black-gray Stein, 2002 Euramerica shale

Progymnosperm Archaeopteridopsida Archaeopteris obtusa Avalonia CAN Quebec Escuminac Andrews et al., 1975 ; Hugh Miller Cliffs Progymnosperm Archaeopteridopsida Archaeopteris obtusa Acadian US NY Katzberg Carluccio et al., 1966 Progymnosperm Archaeopteridopsida Archaeopteris rogersi Acadian US PA Chemung Progymnosperm Archaeopteridopsida Archaeopteris sphenophyll Acadian US NY Delaware Carluccio et al., 1966, ifolia River Arnold, 1939 Flags; Katzberg Progymnosperm Archaeopteridopsida Archaeopteris sp. N. CAN Alberta Yahatinda Scheckler, 1978 Euramerica Progymnosperm Archaeopteridopsida Archaeopteris sp. Avalonia CAN Quebec Escuminac Dawson, 1882 Progymnosperm Archaeopteridopsida Archaeopteris sp. Acadian US NY Oneonta Carluccio et al., 1966 Progymnosperm Archaeopteridopsida Archaeopteris sp. Acadian US NY Enfield; Stein, 2002 Honesdale ; Katsberg; Walton; Stony Clove; West Hall (Beers Hill) Progymnosperm Archaeopteridopsida Archaeopteris sp. Acadian US PA Chemung; Traverse and Schuyler, (lower) 1994

Pocono Progymnosperm Archaeopteridopsida Callixylon newberryi Michigan US Michigan Antrim Arnold, 1934 Basin Shale Progymnosperm Archaeopteridopsida Callixylon newberryi Ohio Basin US Ohio Olentangy Baker, 1942 Shale* Progymnosperm Archaeopteridopsida Callixylon petryi Acadian US NY Geneseel Carluccio et al., 1966, Oneonta Beck, 1953 Progymnosperm Archaeopteridopsida Callixylon zalesskyi Acadian US NY Katsberg; Carluccio et al., 1966, Oneonta Beck, 1960 Progymnosperm Archaeopteridopsida Callixylon sp. N. CAN Nunavut ? Andrews et al., 1965 Euramerica Progymnosperm Archaeopteridopsida Callixylon sp. Acadian US NY Genessee Beck, 1952, Arnold, 1930 (Genunde wa Limestone ) Progymnosperm Archaeopteridopsida Callixylon sp. Acadian US NY Katzberg Carluccio et al., 1966 Progymnosperm Archaeopteridopsida Callixylon sp. Acadian US NY Stony Stein, 2002 PBD 6895 Clove Progymnosperm Archaeopteridopsida Callixylon sp. Acadian US NY Oneonta Carluccio et al., 1966 Progymnosperm Archaeopteridopsida Callixylon sp. App Basin US Virginia Foreknobs Skog, 1983 Pinophyta Pinopsida Cordaites angustifolia Avalonia CAN Quebec ? Dawson, 1882

Famennian (Late Devonian) Higher taxa 1 Higher taxa 2 Genus Species Paleo- Country State/ Formation References geographic Province region Equisetophyta Equisetopsida Eviostachya hoegii Avalonia BE Namur Evieux Stockmans, 1948 Equisetophyta Equisetopsida Eviostachya sp. App Basin US WV Hampshire Scheckler, 1986 Equisetophyta Equisetopsida Sphenophyllum subtenerrimu Avalonia BE Liege Fairon-Demaret, m 1996 Equisetophyta Equisetopsida Sphenophyllum subtenerrimu App Basin US WV Hampshire Scheckler, 1986 m Lycopsida Lycophytes Cyclostigma sp. Acadian US PA Meyer-Berthaud and Decombeix, 2010 Lycopsida Isoetales Lepidodendrop sp. Acadian US PA Catskill Cressler, 2006 sis Lycopsida Lycopsida Otzinachsonia beerboweri Acadian US PA Catskill Cressler, 2005 Isoetophytina Lepidocarpopsida Jurinodendron brevifolium Avalonia BE Namur Evieux Doweld, 2001 Cladoxylopsida Pseudosporochnales Cladoxylon sp. Illinois Basin US Indiana New Albany Read and Shale* Campbell, 1939; Read, 1947 Cladoxylopsida Pseudosporochnales Hierogramma jeffreyi App Basin US Kentucky New Albany Cressler, 2006, (= Shale*; Cross and Cladoxylon?) Hampshire Hoskins, 1951 Cladoxylopsida Pseudosporochnales Pietzschia polyupsilon Illinois Basin US Indiana New Albany Read and Shale* Campbell, 1939 Cladoxylopsida Pseudosporochnales Pietzschia polyupsilon App Basin US Kentucky New Albany Read and

Shale* Campbell, 1939 Cladoxylopsida Pseudosporochnales Polyxylon elegans Illinois Basin US Indiana New Albany Read and Shale* Campbell, 1939 Progymnosperm Aneurophytopsida Aneurophyton olnense Avalonia BE Namur Evieux Stockmans, 1948 Progymnosperm Aneurophytopsida Aneurophyton olnense App Basin US WV Hampshire Scheckler, 1986 Progymnosperm Archaeopteridopsida Archaeopteris eastmanii App Basin US Kentucky New Albany Read, 1936 Shale* Progymnosperm Archaeopteridopsida Archaeopteris fissilis N. Euramerica CAN Nunavut sandstone/sha Arnold, 1930; le Beck, 1952 Progymnosperm Archaeopteridopsida Archaeopteris halliana Acadian US NY Chemung Arnold, 1939 Progymnosperm Archaeopteridopsida Archaeopteris halliana Acadian US PA Chemung Arnold, 1939 Progymnosperm Archaeopteridopsida Archaeopteris halliana App Basin US WV Hampshire Phillips et al., 1972 Progymnosperm Archaeopteridopsida Archaeopteris latifolia Acadian n US PA Oswayo Arnold, 1939 Sandstone Progymnosperm Archaeopteridopsida Archaeopteris macilenta App Basin US WV Hampshire Scheckler, 1986 Progymnosperm Archaeopteridopsida Archaeopteris minor Acadian US PA Chery Ridge Arnold, 1939 Red Shale; Chemung Progymnosperm Archaeopteridopsida Archaeopteris obtusa N. Euramerica CAN Nunavut sandstone/sha Andrews et al., le 1965 Progymnosperm Archaeopteridopsida Archaeopteris obusa App Basin US WV Hampshire Scheckler, 1986 Progymnosperm Archaeopteridopsida Archaeopteris obusa App Basin US Virginia Hampshire Scheckler, 1986 Progymnosperm Archaeopteridopsida Archaeopteris roemeriana Avalonia BE Namur Evieux Fairon-Demaret Province and Leponce, 2001

Progymnosperm Archaeopteridopsida Archaeopteris roemeriana Avalonia BE Namur Fairon-Demaret and Leponce, 2001 Progymnosperm Archaeopteridopsida Archaeopteris roemeriana Acadian US PA Pocono Arnold, 1939 Progymnosperm Archaeopteridopsida Archaeopteris rogersi Acadian US PA Chemung Arnold, 1939 Progymnosperm Archaeopteridopsida Archaeopteris sphenophyllif App Basin US WV Hampshire Scheckler, 1986 olia Progymnosperm Archaeopteridopsida Archaeopteris sp. Acadian US NY Cattaraugus; Arnold, 1939, Conneaut Stein, 2002 (PBD (Oswayo) 29005) Progymnosperm Archaeopteridopsida Archaeopteris sp. Acadian US PA Catskill Stein, 2002 Progymnosperm Archaeopteridopsida Archaeopteris sp. App Basin US WV Hampshire Cornet et al., 1976 Progymnosperm Archaeopteridopsida Callixylon brownii Illinois Basin US Indiana New Albany Read and Shale* Campbell, 1939 Progymnosperm Archaeopteridopsida Callixylon brownii App Basin US Kentucky New Albany Stein, 2002 Shale* Progymnosperm Archaeopteridopsida Callixylon clevelandensi Ohio Basin US Ohio Ohio Black Chitaley, 1992 s Shale Progymnosperm Archaeopteridopsida Callixylon erianum Acadian US NY Gowanda Arnold, 1935 shale Progymnosperm Archaeopteridopsida Callixylon erianum App Basin US WV Hampshire Scheckler, 1986 Progymnosperm Archaeopteridopsida Callixylon newberryi Illinois Basin US Indiana New Albany Stubblefield et al., Shale* 1985 (Blackiston) Progymnosperm Archaeopteridopsida Callixylon newberryi Illinois Basin US Indiana New Albany Read and Shale* Campbell, 1939

Progymnosperm Archaeopteridopsida Callixylon trifilievi N. Euramerica CAN Alberta Beaverhill Campbell, 1963 Lake Progymnosperm Archaeopteridopsida Callixylon sp. N. Euramerica CAN Nunavut black gray Andrews et al., shale 1965 Progymnosperm Archaeopteridopsida Callixylon sp. Acadian US PA Oswayo Arnold, 1939 Sandstone; Pocono; Cuba silt shale (Chadakoin) Progymnosperm Archaeopteridopsida Callixylon sp. App Basin US Virginia Hampshire Phillips et al., 1972 Progymnosperm Archaeopteridopsida Callixylon sp. App Basin US WV Hampshire Scheckler, 1986 Progymnosperm Archaeopteridopsida Callixylon(?) App Basin US Tennessee Chattanooga Prestianni et al., 2013 Lycopsida Lycophytes Cyclostigma Acadian US PA Meyer-Berthaud and Decombeix, 2010 Gymnospermopsida spermatophytes Aglosperma quadripartita Acadian US PA Catskill Prestianni et al., 2013 Gymnospermopsida spermatophytes Aglosperma bertrandii Avalonia BE ? Prestianni et al., 2013 Gymnospermopsida spermatophytes Aporoxylon ? Avalonia(?) GER Thuringia Mintz et al., 2010 Gymnospermopsida spermatophytes Araucarites ? Acadian US NY Albany Mintz et al., 2010 Shale*

Figure S2. 10. Wood tissue-remained plant fossils from the Chattanooga Shale of the Southern Appalachian Basin.

References

Andrews, H. N., Gensel, P. G. & Kasper, A. E. 1975. A new fossil plant of probable intermediate affinities (Trimerophyte–Progymnosperm). Canadian Journal of Botany, 53, 1719–1728.

Andrews, H. N., Phillips, T. L. & Radforth, N. W. 1965. Paleobotanical studies in Arctic Canada: I. Archaeopteris from Ellesmere Island. Canadian Journal of Botany, 43, 545– 556.

Arnold, C. A. 1930. Bark structure of Callixylon. Botanical Gazette, 90, 427–431.

Arnold, C. A. 1934. The so-called branch impressions of Callixylon newberryi (Dn) Elkins and Wieland and the conditions of their preservation. The Journal of Geology, 42, 71–76.

Arnold, C. A. 1935. Some new forms and new occurrences of fossil plants from the Middle and Upper Devonian of New York State. Bulletin of the Buffalo Society of Natural Sciences, 17, 1–12.

Arnold, C. A. 1939. Observations on fossil plants from the Devonian of eastern North America. IV. Plant remains from the Catskill Delta deposits of northern Pennsylvania and southern New York. Contributions from Museum of Paleontology University of Michigan, 5, 271–314.

Arnold, C. A. 1940. Structure and relationships of some Middle Devonian plants from western New York. American Journal of Botany, 27, 57–63.

Arnold, C. A. 1941. Observations on fossil plants from the Devonian of Eastern North America V Hyenia banksii, sp. nov. Contributions from the Museum of Paleontology, University of Michigan, 6, 53–57.

Arnold, C. A. 1952. Observations on fossil plants from the Devonian of eastern North America. VI. Xenocladia medullosina Arnold. Contributions from Museum of Paleontology University of Michigan, 9, 297–309.

Baker, R. C. 1942. The age and fossils of the Olentangy Shale of central Ohio. American Journal of Science, 240, 137–143.

Beck, C. B. 1952. Studies in the Genus Callixylon. Ph.D. Thesis, Cornell University.

Beck, C. B. 1953. A new root species of Callixylon. American Journal of Botany, 40, 226– 233.

Beck, C. B. 1960. The identity of Archaeopteris and Callixylon. Brittonia, 12, 351–368. 76

Berry, C. M. & Edwards, D. 1997. A new species of the lycopsid Gilboaphyton Arnold from the Devonian of Venezuela and New York State, with a revision of the closely related genus Archaeosigillaria Kidston. Review of Palaeobotany and Palynology, 96, 47–70.

Bonamo, P. M. & Banks, H. P. 1966. Calamophyton in the Middle Devonian of New York State. American Journal of Botany, 53, 778–791.

Campbell, J. 1963. Callixylon from the Upper Devonian of northwestern Alberta. American Journal of Botany 50, 648–652.

Carluccio, L. M., Hueber, F. M. & Banks, H. P. 1966. Archaeopteris macilenta, anatomy and morphology of its frond. American Journal of Botany, 53, 719–730.

Chaloner, W. & Pettitt, J. 1964. A seed megaspore from the Devonian of Canada. Palaeontology, 7, 29–36.

Chitaley, S. 1992. On the occurrence of Prototaxites in the Cleveland Black Shale of Ohio, USA. Review of Palaeobotany and Palynology, 72, 257–271.

Cornet, B., Phillips, T. L. & Andrews, H. N. 1976. The morphology and variation in Rhacophyton ceratangium from the Upper Devonian and its bearing on frond evolution. Palaeontographica Abteilung B, 158, 105–129.

Cressler, W. L. & Pfefferkorn, H. W. 2005. A Late Devonian isoetalean lycopsid, Otzinachsonia beerboweri, gen. et sp. nov., from north-central Pennsylvania, USA. American Journal of Botany, 92, 1131–1140.

Cressler, W. L. 2006. Plant paleoecology of the Late Devonian Red Hill locality, north- central Pennsylvania, an Archaeopteris–dominated wetland plant community and early tetrapod site. Geological Society of America Special Papers, 399, 79–102.

Cross, A. T. & Hoskins, J. H. 1951. Paleobotany of the Devonian–Mississippian black shales. Journal of Paleontology, 713–728.

Curry, R. P. 1975. Miospores from the Upper Devonian (Frasnian) Greenland Gap Group, Allegheny Front, Maryland, West Virginia and Virginia, USA. Review of Palaeobotany and Palynology, 20, 119–131.

Dannenhoffer, J. M. & Bonamo, P. M. 1989. Rellimia thomsonii from the Givetian of New York: secondary growth in three orders of branching. American Journal of Botany, 76, 1312–1325.

Doweld, A. 2001. Jurinodendron—a new replacement name for Cyclostigma S. Haughton ex O. Heer, 1871 (Lycopodiophyta). Paleontological Journal, 35, 218–221.

Dawson, J. W. 1882. The Fossil Plants of the Erian (Devonian) and Upper Silurian Formations of Canada. Geological Survey of Canada, Montreal, 95–140.

Fairon–Demaret, M. 1996. Dorinnotheca streelii Fairon–Demaret, gen. et sp. nov., a new early seed plant from the upper Famennian of Belgium. Review of Palaeobotany and Palynology, 93, 217–233.

Fairon–Demaret, M. & Berry, C. M. 2000. A reconsideration of Hyenia elegans Kräusel et Weyland and Hyeniacomplexa'Leclercq: two Middle Devonian cladoxylopsids from western Europe. International journal of plant sciences, 161, 473–494.

Fairon–Demaret, M. & Leponce, I. 2001. Leaf dimorphism in Archaeopteris roemeriana (Progymnosperm): further early fossil evidence of shoot dorsiventrality. American journal of botany, 88, 729–735.

Fairon–Demaret, M. & Li, C. S. 1993. Lorophyton goense gen. et sp. nov. from the Lower Givetian of Belgium and a discussion of the Middle Devonian Cladoxylopsida. Review of Palaeobotany and Palynology, 77, 1–22.

Granoff, J. A., Gensel, P. G. & Andrews, H. N. 1976. A new species of Pertica from the Devonian of eastern Canada. Palaeontographica Abteilung B, 155, 119–128.

Grierson, J. D. & Banks, H. P. 1963. Lycopods of the Devonian of New York state, Paleontological Research Institution.

Krausel, R. & Weyland, H. 1926. Beitrage zur Kenntnis der Devonflora II. Abh. Senckenb. Naturforsch. Gesell, 40, 115–155.

Kräusel, R. & Weyland, H. 1929. Beiträge zur Kenntnis der devonflora. III. Abh Senckenb Natforsch Ges, 41, 315–360.

Kräusel, V. R. & Weyland, H. 1932. Pflanzenreste aus dem Devon III. Uber Hyenia Nath. Senckenbergiana, 14, 274–280.

Leclercq, S. 1969. Calamophyton primaevum: the complex morphology of its fertile appendage. American Journal of Botany, 56, 773–781.

Leclercq, S. & Andrews, H. N. 1960. Calamophyton bicephalum, a new species from the Middle Devonian of Belgium. Annals of the Missouri Botanical Garden, 1–23.

Leclercq, S. & Banks, H. P. 1962. Pseudosporochnus nodosus sp. nov., a Middle Devonian plant with cladoxylalean affinities. Palaeontographica Abteilung B, 1–34.

Lessuise, A. & Fairon-Demaret, M. 1980. Le gisement à plantes de Niaster (Aywaille, Belgique): repère biostratigraphique nouveau aux abords de la limite Couvinien- Givetien. Annales de la Société géologique de Belgique, 103, 157–181.

Lu, H., Peng, P. A. & Sun, Y. 2003. Molecular and stable carbon isotopic composition of monomethylalkanes from one oil sand sample: source implications. Organic Geochemistry, 34, 745–754.

Matten, L. C. 1974. The Givetian flora from Cairo, New York: Rhacophyton, Triloboxylon and Cladoxylon. Botanical Journal of the Linnean Society, 68, 303–318.

Meyer-Berthaud, B., Soria, A. & Decombeix, A.–L. 2010. The land plant cover in the Devonian: a reassessment of the evolution of the tree habit. Geological Society, London, Special Publications, 339, 59–70.

Mintz, J. S., Driese, S. G. & White, J. D. 2010. Environmental and Ecological Variability of Middle Devonian (Givetian) Forests in Appalachian Basin Paleosols, New York, United States. Palaios, 25, 85–96.

Neumann-Mahlkau, P. 1965. Neue Pflanzenfunde und ihre Bedeutung für die Grenze Unterdevon/Mitteldevon am Nordwestabfall des Hohen Venn. Fortschr. Geol. Rheinld. u. Westf, 9, 877–882.

Obrhel, J. 1961. Die Flora der Srbsko-Schichten (Givet) des mittelböhmischen Devons. Sbornik Ustredniho ustavu geologickeho, 26, 7–46.

Phillips, T. L., Andrews, H. N. & Gensel, P. G. 1972. Two heterosporous species of Archaeopteris from the Upper Devonian of West Virginia. Palaeontographica Abteilung B, 139, 47–71.

Prestianni, C., Hilton, J. & Cressler, W. 2013. Were All Devonian Seeds Cupulate? A Reinvestigation of Pseudosporogonites hallei, Xenotheca bertrandii, and Aglosperma spp. International Journal of Plant Sciences, 174, 832–851.

Read, C. B. 1936. The flora of the New Albany shale: Part 1, Diichnia kentuckiensis, a new representative of the Calamopityeae. Report No. 2330–7102, 149–161.

Read, C. B. 1947. Pennsylvanian floral zones and floral provinces. The Journal of Geology, 55, 271–279.

Read, C. B. & Campbell, G. 1939. Preliminary account of the New Albany Shale flora. American Midland Naturalist, 21, 435–453.

Scheckler, S. E. 1975. Rhymokalon, a new plant with cladoxylalean anatomy from the Upper Devonian of New York State. Canadian Journal of Botany, 53, 25–38.

Scheckler, S. E. 1978. Ontogeny of progymnosperms. II. Shoots of upper Devonian Archaeopteridales. Canadian Journal of Botany, 56, 3136–3170.

Scheckler, S. E. 1986, Geology, floristics and paleoecology of Late Devonian coal swamps from Appalachian Laurentia (USA). Annales de la Société géologique de Belgique.

Schweitzer, H. J. 1973. Die Mitteldevon—Flora von Lindlar (Rheinland). 4. Filicinae––– Calamophyton primaevum Kräusel & Weyland. Palaeontographica Abteilung B, 117– 150.

Schweitzer, H. J. & Matten, L. C. 1983. Aneurophyton germanicum and Protopteridium thomsonii from the Middle Devonian of Germany. Palaeontographica Abteilung B, 184, 64–106.

Scott, D. 1926. New discoveries in the Middle Devonian Flora of Germany. New Phytologist, 25, 373–379.

Serlin, B. S. & Banks, H. P. 1978. Morphology and Anatomy of Aneurophyton, A Progymnosperm from the Late Devonian of New York. Palaeontographica Americana, 8, 343–359.

Skog, J. E. 1983. Callixylon wood from the Greenland Gap Group (Upper Devonian) of southwestern Virginia. Virginia journal of science, 34, 58–64.

Stein, W. E. 2002. SUNY–Binghamton Paleobiology Collection. (From Paleobiology Database/Fossilworks; https://paleobiodb.org/) (July 2017).

Stein, W. E., Berry, C. M., Hernick, L. V. & Mannolini, F. 2012. Surprisingly complex community discovered in the mid–Devonian fossil forest at Gilboa. Nature, 483, 78–81.

Stein, W. E. & Hueber, F. M. 1989. The anatomy of Pseudosporochnus: P. hueberi from the Devonian of New York. Review of Palaeobotany and Palynology, 60, 311–359.

Stein, W. E., Mannolini, F., Hernick, L. V., Landing, E. & Berry, C. M. 2007. Giant cladoxylopsid trees resolve the enigma of the Earth's earliest forest stumps at Gilboa. Nature, 446, 904–908.

Stockmans, F. 1968. Végétaux mésodévoniens récoltés aux confins du massif du Brabant: Belgique.

Stockmans, F. 1948. Végétaux du Dévonien supérieur de la Belgique. Musée royal d'histoire naturelle de Belgique, 110, 3–85.

Stubblefield, S. P., Taylor, T. N. & Beck, C. B. 1985. Studies of paleozoic fungi. IV. Wood– decaying fungi in Callixylon newberryi from the upper Devonian. American Journal of Botany, 1765–1774.

Traverse, A. & Schuyler, A. 1994. Palynostratigraphy of the Catskill and part of the Chemung Magnafacies, southern New York State, USA. Courier Forschungsinstitut Senckenberg, 169, 261–274.

CHAPTER 3:

PERIODIC MARINE EUXINIA AND TERRESTRIAL ORGANIC MATTER INPUTS

LINKED TO ASTRONOMICAL FORCING DURING THE FRASNIAN–FAMENNIAN

MASS EXTINCTION

3.1 Abstract

Marine anoxia is considered as the direct cause of the Frasnian–Famennian mass extinction, yet the frequency and causality of anoxia remain controversial. Here, we demonstrate the significance of astronomical forcing in pacing the episodic deposition of marine anoxia during the Frasnian–Famennian mass extinction by establishing a high- resolution, multi-proxy geochemical profile of an Upper Kellwasser (UKW) interval from the

Chattanooga Shale of Tennessee, USA. Geochemical indices for marine anoxia/euxinia, marine primary productivity, terrestrial plant inputs, and sea-level changes show synchronous fluctuations paced by orbital obliquity during the UKW interval. My data suggest a scenario in which obliquity regulated sea-level variations and pulses of terrestrial plant/soil organic matter to the ocean that, in turn, mediated the frequency of algal blooms and marine euxinia.

This study provides the first unambiguous evidence linking the periodicity of marine euxinia and contemporaneous terrestrial and marine environmental changes to obliquity, suggesting that astronomical forcing was a key modulator of environmental stresses during the Frasnian–

Famennian mass extinction.

3.2 Introduction

The Frasnian–Famennian (F–F) boundary biotic crisis, featuring the extinction of 22–

35% of marine genera, was one of the five most devastating mass extinctions during the

Phanerozoic (Bambach et al., 2004). Various ultimate causes proposed for this biocrisis remain intensively debated, including bolide impacts (Sandberg et al., 2002), sea-level fluctuations (Copper, 2002), climate change (Joachimski and Buggisch, 2002), volcanism

(Racki et al., 2018), tectonism (Averbuch et al., 2005), and land plant evolution (Algeo and

Scheckler, 1998). However, there is a general consensus that marine anoxia was the immediate, direct cause based on widespread organic-rich sediments contemporaneous with the marine diversity loss across the F–F boundary (e.g., Joachimski and Buggisch, 1993;

Bond et al., 2013; Carmichael et al., 2014). Although early studies favored the hypothesis that stable deoxygenation persisted in late Frasnian seas (e.g., Ettensohn, 1992), more recent geochemical and sedimentological data have shown that anoxic/euxinic events were short- lived, occurring periodically in shallow epeiric seas during the Late Devonian (Brown and

Kenig, 2004; Boyer et al., 2014; Carmichael et al., 2014; George et al., 2014; Haddad et al.,

2018, 2016; Lash, 2017). The pulsed nature of anoxia is highly intriguing and agrees with the paleontological perspective that the biodiversity loss during the F–F extinction was due to recurring environmental stress (Bambach et al., 2004; Boyer et al., 2014). However, the frequency and duration of anoxic episodes, as well as mechanisms that initiated and regulated the frequency, have not been determined (Carmichael et al., 2019).

The F–F boundary was characterized by a series of extinction events, among which the

Upper Kellwasser (UKW) event during the latest Frasnian represented the largest pulse of

diversity loss in shallow marine ecosystems (House, 2002). The nature and potential causes of the UKW extinction event have been widely investigated in North America across the northern Appalachian, Michigan, Illinois, Great and Iowa basins to western Canada (e.g.,

Brown and Kenig, 2004; Duan et al., 2010; Formolo et al., 2014; Whalen et al., 2015;

Haddad et al., 2018; De Vleeschouwer et al., 2017; Bond et al., 2004). Here, I determine the frequency and mechanism of cyclic variation in marine anoxia during the UKW event by presenting a geochemical profile from the uppermost Frasnian strata of a Chattanooga Shale outcrop in the southern Appalachian Basin, North America. I reconstructed an ultra-high- resolution (1-cm spacing), multi-proxy marine environmental record during the UKW interval including redox conditions, primary productivity, terrestrial inputs, and sea-level changes. I further probed the linkage between these environmental changes and astronomical forcing during the UKW interval. Although astronomical forcing has been identified as a key driver regulating variation in lithofacies and geochemical proxies throughout the Devonian

(e.g., Filer, 2002; Chen and Tucker, 2003; Elrick et al., 2009; De Vleeschouwer et al., 2017,

2013; Liu et al., 2019), the role of astronomical forcing during mass extinction events remains an open question (De Vleeschouwer et al., 2017). A recent study (i.e., De

Vleeschouwer et al., 2017), for the first time, reported that δ13C chemostratigraphy of the

UKW interval was paced by obliquity, yet the δ13C of bulk carbon pools is an integrative proxy that does not provide source- or process-specific information to identify regulating mechanisms. The results of this study provide the first evidence that astronomical forcing regulated not only the pace of marine euxinia but also the fluctuations of terrestrial and marine biogeochemical drivers that led to the euxinia during the UKW interval. This finding

offers new insights into the extrinsic causal mechanism of the F–F mass extinction, that is, which mechanism is more likely to have responded to astronomical forcing in the process of creating recurring lethal marine environmental conditions.

3.3 Geological setting

The Chattanooga Shale was deposited in an epicontinental sea from the early Frasnian to late Famennian (Li and Schieber, 2015). The studied section of the Chattanooga Shale is located in Chestnut Mound, central Tennessee (36.207838, −85.834365, Fig. 3.1a). A detailed description of its sedimentological and lithological features can be found in Li and Schieber

(2015). This 7.25-m-thick outcrop exposes both the Frasnian-age Dowelltown Member and

Famennian-age Gassaway Member of the Chattanooga Shale (Conant and Swanson, 1961)

(Fig. 3.1b). The UKW interval in this section was identified based on: (i) the occurrence of the latest Frasnian conodonts (i.e., Palmatolepis linguiformis) as previously described by

Over (2007), (ii) the presence of a volcanic ash bed (i.e., the Center Hill Ash) (Over, 2002) near the top of the MN 13 Zone, and (iii) a pronounced positive excursion (> 2‰) of stable carbon isotopic values of organic matter within the uppermost Frasnian MN 13 Zone, followed by a negative shift through the lower Famennian strata (Fig. 3.1b). A similar

13 positive excursion in δ Corg has been widely observed in UKW deposits around the world

(e.g., Stephens and Sumner, 2003; Whalen et al., 2015; De Vleeschouwer et al., 2017).

Sample collection was done at 1-cm intervals yielding a total of 35 samples within the

UKW interval, and at 25-cm intervals yielding a total of 30 samples for the entire exposed

Frasnian to Famennian section. The results below describe the ultra-high-resolution, 1cm- spaced data from the UKW interval, and the supplemental section presents those from the

entire exposed section (Appendix II Fig. S3.1).

Figure 3.1. Geographic and stratigraphic context of the Upper Kellwasser (UKW) interval in the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee. a Late Devonian global paleogeographic map (360 Ma) (top) and enlarged view of south-central Euramerica (bottom) (adapted from Ron Blakey © 2016 Colorado Plateau Geosystems Inc.). Red star indicates the study site, and yellow line depicts the paleo-Equator. AB: Appalachian Basin; 13 MB: Michigan Basin; IB: Illinois Basin. b Stratigraphic log, δ Corg profile (25 cm spacing), and XRF-derived Ti/Al ratios (2–3 cm spacing) for the Chestnut Mound outcrop of the Chattanooga Shale. Lithostratigraphy is modified after Li and Schieber (2015), and the conodont biostratigraphy is from Over (2007). Grey bar highlights the UKW interval.

3.4 Methods

3.4.1 Sample collection

Thirty samples were collected at an interval of ~25 cm throughout the 7.25-m-thick section and additional thirty-five samples were collected at an interval of 1 cm from the

UKW interval. In addition, 193 samples were collected at 2–3 cm intervals from the base to 4 m of the Chattanooga Shale. Surface and weathered rocks were removed before sampling.

Prior to geochemical analyses, the surface of each sample was further removed in the lab by knife, and only freshly exposed pieces were used for subsequent analyses. All samples were washed using ultra-pure carbon-free water.

3.4.2 Palynology

Twenty-three samples were selected for palynology analysis. About 5 g of crushed sample was initially treated in HCl (10%) overnight, followed by digestion in HF (48%) for at least 48 hours. Residues were then treated with Schultz’s solution (HNO3+KClO3) for 30 mins for oxidization, followed by the addition of sodium hydroxide until reaching a neutral pH. The palynomorphs were examined using a Nikon Microphot transmitted light microscope with ×20 (dry) and ×100 (oil immersion) objectives. The quantity of the palynomorphs of each slide was counted (300 points) under a ×20 objective. Three palynological slides were analyzed for each sample, and the average value from the three slides was reported.

3.4.3 Bulk geochemistry

13 Prior to the analysis of total organic carbon (TOC) and organic carbon isotope (δ Corg), all samples were ground into 100–200 mesh powder. Approximately 10 mg of black shale or

20 mg of gray-greenish shale sample was placed into a tin capsule and decarbonated with 5% sulfurous acid. All samples were dried overnight in an oven at 50 ºC. The samples were then submitted to the University of California Davis Stable Isotope Facility (California, USA) and analyzed on a Micro Cube elemental analyzer (Elementar Analysensysteme GmbH, Hanau,

Germany) interfaced to a PDZ Europa 20–20 isotope ratio mass spectrometer (The Sercon

13 Ltd., Cheshire, UK). δ Corg values (‰) were calibrated against NIST Standard Reference

Materials (USGS-40, USGS-41) and reported relative to V-PDB. The analytical precision was less than 0.2 ‰ based on internal standards that included nylon, bovine liver, peach leaves, and glutamic acid.

Major and trace element concentrations of 65 samples were measured at the ALS

Chemex Lab, Ltd (Guangzhou, China). For major elements, about 5 g sample was prepared by fusion with lithium metaborate mixed with lithium nitrate and heated at 1,050 ºC for an hour. The mixtures were then placed on a platinum mould and then analyzed by an X-ray fluorescence (XRF) spectrometry (Philips PANalytical PW2424, Netherland). Analytical precision for all major elements was better than 5%. From the XRF data, CIA was calculated as CIA = Al2O3/(Al2O3+K2O+Na2O+CaO*) × 100 (Nesbitt and Young, 1982), where CaO* represents phosphate corrected values following McLennan (1993). For trace elements, 2–5 mg of grounded sample was treated using a three-acid (HF-HNO3-HClO4) digestion method.

The residue was then leached with HCl. The analysis of the solution was conducted using a

PerkinElmer Elan9000 element inductively coupled plasma mass spectrometry (ICP-MS).

The analytical precision for all trace elements was better than 7%. Enrichment factors (EFs) of Mo or U were calculated as (Mo or U/Al of a sample) / (Mo or U/Al of the average shale), where the average values of shale were from Wedepohl (1971).

I also measured bulk elemental compositions of all collected samples using the Bruker tracer IIISD hand-held XRF under vacuum. These data were used only for the time series analysis described below. Three measurements were taken on a fresh surface of each sample with source energy of 15keV (no filter, 25µA) and 180-seconds count time. Element

intensities were obtained by measuring the peak area of raw X-ray spectra using the ARTAX software. The ratio between titanium and aluminum counts (Ti/Al) was calculated.

3.4.5 Biomarker analysis

Sixty-five samples were processed for molecular biomarker analyses. About 20 grams of powdered sample material were Soxhlet-extracted with a mixture of Dichloromethane (DCM)

/Methanol (MeOH) (97:3, v/v) for at least 72 hours. Activated copper was added to remove the elemental sulfur. The extracts were concentrated using a rotary evaporator and transferred into a pre-weighed 4mL vial. Each sample was subsequently separated into aliphatic, aromatic, and polar fractions by silica gel column chromatography with a mixture of activated silica gel and alumina (3:1, v/v). The aliphatic fraction was eluted with petroleum ether (4 column volumes) followed by the aromatic fraction eluted with benzene and the polar fraction eluted with MeOH. Each of these fractions was collected and dried via rotary evaporation and then diluted with 1 mL of hexane. The saturated fraction of each sample was then separated into straight alkane and branched/cyclic alkane (b/c) fractions through urea adduction following Xu and Sun (2005).

The aliphatic fraction was quantified on an Agilent 7890A gas chromatograph (GC) equipped with a flame ionization detector (FID) and a DB-1MS capillary column (60 m×0.32 mm, 0.25 μm film thickness). The initial oven temperature was 60 ℃, maintained for 2 min, and increased at 4 ℃/min to 295 ℃, maintained for 30 min. Nitrogen was the carrier gas at a flow rate of 1.0 ml/min. The inlet temperature was set at 295 ℃ and FID temperature was

300 ℃. Normal alkanes were quantified by their peak areas relative to an internal standard

(n-C24D50) added prior to GC-FID measurement. The branched/cyclic alkane and aromatic

fractions were analyzed on an Agilent 7890B GC-5977A mass spectrometer (MS) equipped with a DB-1MS capillary column (60 m×0.32 mm, 0.25 μm film thickness). Using helium as the carrier gas (1.0 ml/min), the GC-oven was heated from 80 ℃ with a 2 min hold to 220 ℃ at 3 ℃/min and further increased to 300 ℃ at 2 ℃/min with a 30 min hold. The source was operated in 70 eV electron impact mode at 230 ℃. Compounds were identified by mass spectra and relative retention times. Selected ion monitoring (SIM) was used for quantifying hopanes (m/z 191) and steranes (m/z=217) in branched/cyclic alkane fraction as well as aryl isoprenoids (m/z=133) in aromatic fraction. For the quantification of aromatic fractions, known amounts of phenanthrene-D10 and dibenzothiophene-D8 were added prior to GC-MS measurement.

3.4.6 Time-series analysis

The detection and interpretation of major Milankovitch cycles were conducted on the

193 samples collected at 2–3 cm intervals from the Upper Frasnian to Lower Famennian interval (4 m thickness). Prior to the analyses, the Ti/Al data were interpolated linearly to a constant 2-cm resolution and subsequently detrended by subtracting a linear trend using the

“detrending” toolbox in Acycle v2.1 software (Li et al., 2019).

The time-series analysis of processed Ti/Al data was performed in Acycle v2.1(Li et al.,

2019). Multi-taper method (MTM) (Thomson, 1982) was used with three 2π tapers and a time-bandwidth product of 2. The MTM power spectra of detrended Ti/Al data were tested against robust first-order autoregressive “AR (1)” red-noise models (Mann and Lees, 1996) at

90%, 95%, and 99% confidence levels with a 20% median filter length and linear fitting.

Correlation coefficient (COCO) analysis was carried out to evaluate the most possible

sedimentation rate for the studied section, through comparison between power spectra of proxy series and specific astronomical forcing targets (Li et al., 2018b). The targeted theoretical astronomical frequencies included main obliquity and precession frequencies

(1/21.3 kyr, 1/20.2 kyr, 1/17.4 kyr and 1/34.4kyr) at 370 Ma (Waltham, 2015) and three orbital eccentricity frequencies (1/405 kyr, 1/128kyr and 1/95 kyr) (Laskar et al., 2004). The sedimentation rates tested by COCO spanned from 0.05 cm/kyr to 2 cm/kyr with a step of 0.2 cm/kyr. This range of sedimentation rate enclosed the long-term average sedimentation rate determined based on conodonts. According to the Devonian time scale (Kaufmann, 2006), the linguiformis biozone (upper MN 13 biozone) has a duration of ~500 kyr and the MN 13 Zone of the studied section spans ~0.8 m (Over, 2007), giving a range of long-term mean sedimentation rate of ~0.2 cm/kyr. The significance of the sedimentation rate was tested using

Monte Carlo simulation with 5000 iterations. Based on the sedimentation rates constrained by the COCO analysis, major cycles recognized in the power spectrum were converted from the stratigraphic to time domain and isolated by Gaussian band pass filtering (Kodama and

Hinnov, 2014). The filtered 405-kyr and ~34-kyr cycles were extracted by Gaussian filtering with passbands of 0.89±0.4 cycles/m and 11.8±0.4 cycles/m, respectively.

The interpreted 405-kyr eccentricity related sedimentary cycles were used to convert data from the stratigraphic to time domain. Prior to subsequent analyses, the 405-kyr tuned data series were linearly interpolated to a uniform sample rate of 7.2 kyr. To track the presence of eccentricity, obliquity, and precession cycles in the time domain, the 2π MTM analysis was performed on 405 kyr tuned data with “AR(1)” model at 90%, 95% and 99% confidence levels with a 20% median filter length and linear fitting. Evolutionary Fast

Fourier Transform (FFT) analysis (Kodama and Hinnov, 2014) was performed on 405-kyr tuned data to depict spectral features and corresponding cyclicity related to variable sediment accumulation rates.

3.5 Results

3.5.1 Organic matter characteristics of the UKW interval

13 Total organic carbon (TOC) contents range from 0.2 to 5.0 %, and δ Corg values vary from −30.8 to −23.2 ‰. Both proxies show high variability through the UKW interval (Fig.

3.2). Normal alkanes show a carbon number range from C11 to C36 with a peak at n-C17 or n-

C19 and no odd-over-even predominance from n-C23 to n-C31 (Appendix II Fig. S3.2).

3 Concentrations of n-C17+19 and n-C27+29+31 alkanes range from 0.2 to 6.1×10 μg/g rock and

0.06 to 1.4×103 μg/g, respectively, and they show frequent fluctuations throughout the UKW interval (Figs. 3.2 and 3.3).

13 Figure 3.2. High-resolution profiles of bulk and molecular proxies (δ Corg, n-C17+19 alkanes, C27 steranes) from the Upper Kellwasser (UKW) interval (the grey bar in Fig. 3.1b) of the Chestnut Mound outcrop, Chattanooga Shale, central Tennessee. Blue dots show the enlarged 13 fluctuations in δ Corg, normal C17+19 alkanes (n-C17+19 alkanes), C27 steranes, and the corresponding scales are shown above. 92

Figure 3.3. High-resolution profiles of inorganic and organic proxies for continental weathering and terrestrial plant inputs from the Upper Kellwasser (UKW) interval (the grey bar in Fig. 3.1b) of the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee. Blue dots highlight the enlarged fluctuations in CIA and normal C27+29+31 alkanes (n-C27+29+31 alkanes), and the corresponding scales are shown above. CIA: chemical index of alteration. DBF: dibenzofuran. CIA and DBF indicate changes in chemical weathering intensity and associated soil erosion. Normal C27+29+31 alkanes, norabietane, retene, tetrahydroretene are biomarkers for land plants.

Abundant steranoids and hopanoids were identified in all samples. Diasteranes and steranes are both dominated by C29, while C27 and C28 steranes are present at relatively low abundances (Appendix II Fig. S3.3). The concentrations of C27, C28, and C29 steranes are highly variable, ranging from 3.7 μg/g to 54.4 μg/g, 5.6 μg/g to 1.1×103 μg/g, and 6.3 μg/g to

904.4 μg/g, respectively. Hopanoids from the UKW interval range from C27 to C35 with a peak at C30 (Appendix II Fig. S3.4). The C29/C30 αβ hopane (C29/C30H) ratio varies from 0.6 to 0.8 (Fig. 3.4). Compared to pentacyclic triterpenoids, the abundance of tricyclic terpenoids is low. Norabietane was identified in all samples, and it ranges between 3.7 and 54.4 μg/g

(Fig. 3.3, Appendix II Fig. S3.5). Other land-derived biomarker compounds that were detected include retene, tetrahydroretene, and dibenzofuran (DBF) in the aromatic fraction

(Appendix II Fig. S3.6), ranging from 0.3 to 13.6 μg/g, 1.1 to 20.6 μg/g, and 0 to 6.3 μg/g, respectively (Fig. 3.3). Aryl isoprenoids ranging from C13 to C32 were detected in the aromatic fractions of all samples (Appendix II Fig. S3.7), with the concentration of intermediate-chain aryl isoprenoids (C18–20) fluctuating from 10.0 μg/g to 331.2 μg/g (Fig.

3.5).

3.5.2. Major and trace elements of the UKW interval

The samples contain Al2O3 (14.32–19.85 wt.%), K2O (4.33–9.65 wt.%), Na2O (0.28–

0.64 wt.%), CaO (0.04–3.71 wt.%), and P2O5 (0.07–0.27 wt.%). Chemical index of alteration

(CIA) varies within the range of 70.0 to 81.7 with frequent fluctuations (Fig. 3.3).

Enrichment factors (EFs) of redox-sensitive trace elements vary from 1.8 to 59.1 for Mo and

2.3 to 6.4 for U. The EFs of these two elements covary through the UKW interval, as shown by a significant positive correlation (Pearson’s r=0.689, P=0.000). Zr/Al fluctuates between

1.9 ×10-3 and 4.0×10-3, and X-ray fluorescence (XRF)-derived Ti/Al ratio ranges between 2.8 and 5.1 (Fig. 3.4).

Figure 3.4. High-resolution profiles of inorganic and organic proxies for sea-level changes from the Upper Kellwasser (UKW) interval (the grey bar in Fig. 3.1b) of the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee.

Figure 3.5. High-resolution profiles of organic and inorganic proxies for marine redox conditions from the Upper Kellwasser (UKW) interval (the grey bar in Fig. 3.1b) of the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee. Blue dots show the enlarged fluctuations in C18–20 aryl isoprenoids, and the corresponding scale is shown above.

3.5.3. Time-series analysis

Time-series analysis was conducted on a high-resolution XRF-derived Ti/Al record collected from the Upper Frasnian to Lower Famennian interval (4 m in thickness) (Fig.

3.1b). Multi-taper method (MTM) results in the stratigraphic domain reveal three major clusters of significant spectral peaks above the 99% confidence level: at ~0.9, ~12, ~22–23 cycles/m (Fig. 3.6a). Correlation coefficient (COCO) analysis shows sedimentation rates of

0.27–0.39 cm/kyr and 1.16–1.32 cm/kyr with a correlation coefficient greater than 0.4 and the significance level of null hypothesis (H0, no astronomical forcing) lower than 1% (Fig. 3.7a, b). Given the ~0.2 cm/kyr sedimentation rate constrained by biostratigraphy and the largest total number of astronomical parameters (Fig. 3.7c), 0.27–0.39 cm/kyr is the most probable sedimentation rate range with an average sedimentation rate of 0.3 cm/kyr. Over this sedimentation rate range, the significant peaks at ~0.9, ~12, and ~22–23 cycles/m correspond to ~405 kyr long-eccentricity cycle (Laskar et al., 2004), ~34 obliquity cycle and ~17 precession cycle respectively (Appendix II Fig. S3.8) (Waltham, 2015).

In order to construct an astronomical timescale, the data were converted from the stratigraphic domain to the time domain. Since the 405-kyr long-eccentricity cycle is the strongest cycle observed (Fig. 3.6a), the average value of the period associated with this cycle was tracked and used as a metronome for this purpose (Appendix II Fig. S3.9). The MTM analysis of the 405-kyr tuned Ti/Al record provides further evidence for the presence of long- eccentricity (405 kyr), obliquity (~34 kyr), and precession cycles (~21 and ~17 kyr) (Fig.

3.6b).

Figure 3.6. Time-series analysis of high-resolution XRF-derived Ti/Al record from a 4-meter Upper Frasnian to Lower Famennian interval of the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee. a Results of 2π multiple-taper method (MTM) analysis in the stratigraphic domain. Numbers represent significant peaks shown in cycles/m; these peaks exceed the 99% confidence level (CL) for the AR(1) red-noise model. b Results of 2π MTM analysis of 405-kyr tuned Ti/Al series. Numbers represent significant peaks shown in cycles/kyr; these peaks exceed the 99% CL for the AR(1) red-noise model. c Detrended Ti/Al data series (grey), filtered 405-kyr eccentricity cycles (blue) and the 34-kyr obliquity cycles (red). d Evolutionary Fast Fourier transform (FFT) spectrogram of the 405-tuned Ti/Al series. E: 405-kyr eccentricity, O: ~34-kyr obliquity, P: ~21-kyr precession, p: ~17-kyr precession.

The bandpass filter generated visible astronomical cyclicities in the stratigraphic domain

(Fig. 3.6c). Ti/Al ratios and other geochemical proxies for marine deoxygenation, marine productivity, terrestrial plant and soil inputs, continental weathering, and sea-level change all 97

show 2–6 cm spaced fluctuations through the UKW interval, suggesting that small-scale oscillations should correspond to obliquity and precession cycles rather than to long- eccentricity cycle. I then applied the evolutionary Fast Fourier Transform analysis (FFT) on the 405-kyr tuned Ti/Al record. The FFT results show a strong and good alignment of the obliquity cycle yet sporadic and weaker signals of precession cycles (Fig. 3.6d).

Figure 3.7. Correlation coefficient (COCO) analysis of the Ti/Al record from a 4-meter Upper Frasnian to Lower Famennian interval of the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee. a Correlation coefficient over a range of sedimentation rate; numbers indicate probable sediment rates with high correlation coefficients. b Null hypothesis (H0) of no astronomical forcing tests show the sedimentation rates of 0.27–0.39 cm/kyr and 1.08–1.36 cm/kyr at the significance level of less than 1%. c Number of contributing astronomical parameters in tested sedimentation rate. The astronomical parameters are from Laskar et al. (2004) and Waltham (2015).

3.6 Discussion

Episodic anoxia/euxinia has been well documented from other UKW sections worldwide

(e.g., in North America, Australia, Bolivia, China, Libya, and Poland) based on geochemical

(Brown and Kenig, 2004; Carmichael et al., 2014; Haddad et al., 2018, 2016; Riboulleau et al., 2018; Spaak et al., 2018), petrographic (Murphy et al., 2000), and palynological/microfossil data (Boyer et al., 2014; Song et al., 2019). The intermediate-chain aryl isoprenoid (C18-20) and MoEF data are consistent with these previous findings. Aryl isoprenoids detected in the studied samples can be attributed to carotenoids specific to photosynthetic green sulfur bacteria, whose presence is an indicator of photic zone euxinia

(PZE) (Grice et al., 2005). The frequent fluctuations of intermediate-chain aryl isoprenoids

(C18-20) clearly show sulfide-rich, PZE conditions that developed episodically during the

UKW. This interpretation is consistent with MoEF, which is a redox-sensitive proxy that has been widely used to assess the degree of deoxygenation at the bottom of the Late Devonian seas (e.g., Formolo et al., 2014; Lash, 2017). MoEF also shows frequent fluctuations and covaries with C18-20 aryl isoprenoids through the UKW interval (Pearson’s r = 0.338, P =

0.047).

13 Three proxies were used to reconstruct marine productivity, i.e., δ Corg, C17 and C19 n-

13 alkanes, and C27 steranes. High δ Corg values can indicate elevated marine primary productivity (George et al., 2014) or better organic matter preservation (Kump and Arthur,

13 1999). In the study section, a positive excursion of δ Corg was observed within the UKW interval (Fig. 3.1b). Similar δ13C excursions have been globally observed in both organic and inorganic carbon pools during the UKW and interpreted as a result of higher

photoautotrophic production (Joachimski et al., 2001). C17 and C19 n-alkanes in low-thermal- maturity rocks can also indicate primary productivity by algae, phytoplankton, and photosynthetic bacteria (Han et al., 1968). C27 steranes are a more specific productivity

13 biomarker that represents the abundance of red algae (Kodner et al., 2008). δ Corg, n-C17+19 alkanes, and C27 steranes all show frequent fluctuations that suggest episodic enhancement of marine productivity through the UKW interval (Fig. 3.2). Significantly, increases in these proxies coincided with the development of PZE, as evidenced by positive correlations with redox proxies (i.e., C18-20 aryl isoprenoids, MoEF) (Appendix II Table S3.1), supporting the top-down eutrophication model that excessive primary productivity led to anoxia/euxinia in the Late Devonian seas (Carmichael et al., 2019 and references therein).

What is the cause of these episodic increases in marine productivity during the UKW? It is revealed by positive correlations between marine productivity-associated proxies and terrestrially sourced biomarkers that reflect changes in land vegetation and soils (Fig. 3.8,

Appendix II Table S3.1), suggesting that nutrient inputs from land plants and soils were responsible for eutrophication. All terrestrial biomarkers show frequent fluctuations during the UKW, indicating multiple pulses of land-derived organic matter inputs (Fig. 3.3).

Norabietane, retene, and tetrahydroretene are generally interpreted as compounds from abietic acid produced by trees of the conifer family (Lu et al., 2013), and the presence of abietane-type compounds in older rocks is considered an indicator of early tracheophytes

(Scott, 1974; Lu et al., 2019). Long-chain n-alkanes (n-C27+29+31 alkanes) are commonly used to indicate vascular plant contributions to sediments and rocks with the caveat that some moss and non-marine algae can also be a potential source (Eglinton and Hamilton, 1967;

100

Bush and McInerney, 2013). DBF originates from dehydrated polysaccharides often produced through microbial decay in soils (Sephton et al., 2005), and it is thus an indirect proxy for terrestrial plant inputs for reflecting soil erosion intensity in the source area

(Sephton et al., 2005). In studied samples, both long-chain n-alkanes and DBF covary with abietane-type compounds (Pearson’s P<0.05; Appendix II Table S3.1), confirming that all of these compounds are indicative of terrestrial plant and soil inputs. In addition to biomarkers, terrestrial palynomorphs (trilete spores and phytoclasts) are present throughout the entire exposed section of the Chattanooga Shale and exhibit increases in abundance within the

UKW interval, substantiating the prevalence of tracheophytes in the Late Devonian southern

Appalachian Basin region and the interpretations based on terrestrial plant biomarkers

(Appendix II Figs. S3.1 and S3.10). Although the importance of terrestrial nutrient inputs has been frequently discussed in the development of Late Devonian marine anoxia (Algeo and

Scheckler, 1998; Lash, 2017), my data are the first showing synchronous, high-resolution variations between PZE and pulses of terrestrial organic inputs (Fig. 3.8; Appendix II Table

S3.1).

What is the cause of the pulsed inputs of terrestrial nutrients during the UKW interval?

These pulses coincided with periods of intensified continental weathering and sea-level fall.

CIA has been widely used as an index for chemical weathering intensity of source areas for marine sediments (e.g., Zhao and Zheng, 2015). Ti/Al and Zr/Al ratios, on the other hand, can indicate physical weathering intensity, with higher values indicating a larger proportion of silt-fraction heavy minerals whose transport could indicate sea-level changes (Lash, 2017;

Riboulleau et al., 2018). In my samples, CIA covaries positively with Ti/Al and Zr/Al (Fig.

101

3.8, Appendix II Table S3.1), and their temporal variations are similar to those of terrestrial plant biomarkers and productivity- and redox-associated proxies (Fig. 3.8, Appendix II Table

S3.1). These patterns suggest that enhancement of both physical and chemical weathering played a key role in mediating terrestrial inputs and, thus, the frequency of marine productivity blooms and PZE. I also employed another sea-level indicator, C29/C30H, for which higher values suggest deeper, carbonate-rich depositional environments and lower values for shallower, sandstone- or shale-rich environments (Clark and Philp, 1989).

C29/C30H is significantly correlated with Ti/Al and Zr/Al (Fig. 3.8, Appendix II Table S3.1), substantiating the sea-level interpretations based on Ti/Al and Zr/Al. All three proxies show fluctuating values through the UKW interval that indicate high-frequency sea-level changes

(Fig. 3.4). Although no consensus has been reached regarding Late Devonian sea-level changes at a global scale, lithological evidence from Western Australia (Stephens and

Sumner, 2003), southern China (Chen and Tucker, 2003), Poland (De Vleeschouwer et al.,

2017) and the eastern USA (Filer, 2002) indicates frequent fluctuations around the F–F boundary. Falling sea levels can enhance continental weathering by exposing larger land surfaces and causing fluvial downcutting and entrenchment (Chen and Tucker, 2003;

DiPietro, 2013).

102

Figure 3.8. Cross-plots of representative inorganic and organic paleoenvironmental proxies from the Upper Kellwasser (UKW) interval of the Chestnut Mound outcrop of the Chattanooga Shale, central Tennessee. Pearson’s P and r values are shown. Blue lines denote the best-fit linear regression line. CIA: Chemical Index of Alteration. DBF: dibenzofuran. The bivariate correlations of all paleoenvironmental proxies measured in this study are presented in Appendix II Table S2.

103

Time-series analysis shows the prominent influence of obliquity in modulating sea-level changes during the Late Frasnian to Early Famennian (Fig. 3.6). De Vleeschouwer et al.

(2017) were the first reporting a strong obliquity signal in six globally distributed δ13C profiles of the UKW interval. These authors suggested that the dominance of obliquity during the UKW was due to a prolonged period of low eccentricity, and they noted a similar astronomical configuration during the Cretaceous Oceanic Anoxic Event-2 (OAE-2), implying a close link between astronomical forcing and global-scale organic carbon burial.

The mechanism underlying this link, however, was not identified. Based on multiple proxies that are source- or process-specific, my results demonstrate that obliquity directly controlled the timing and extent of marine anoxia through biogeochemical linkages between terrestrial and marine ecosystems via a chain process that involved sea-level fall, intensification of physical and chemical weathering, terrestrial organic matter and nutrient inputs, and marine eutrophication.

Although obliquity has the largest influence on high-latitude regions, it can influence

(sub)tropical records remotely through the waxing and waning of ice sheets and associated glacio-eustatic responses. Specifically, obliquity nodes inhibit summer ice melting and favor ice sheet growth and subsequent eustatic fall (Boulila et al., 2011). The obliquity-forced sea- level changes observed in the present study may have been caused by the presence of small and ephemeral ice sheets on Gondwana. Although no physical evidence of continental ice sheets has been reported during the Frasnian to date, some stratigraphic and palynologic data support the presence of the Frasnian ice sheets (Streel et al., 2000; Filer, 2002). Alternatively, obliquity-forced sea-level oscillations may be attributable to changes in continental aquifer

104

storage, which can generate up to 5 m of eustatic variation (Jacobs and Sahagian, 1993). High obliquity accelerates heat and moisture transfer from low to high latitudes that can increase groundwater storage at mid- and high-latitudes and subsequently cause global regression.

Conversely, low obliquity attenuates hydrological cycles and can lead to global transgression

(Wendler et al., 2016). Obliquity-driven aquifer eustasy has been suggested for other largely ice-free geological epochs, such as the Early Triassic (e.g., Li et al., 2018a) and mid-

Cretaceous (e.g., Wendler et al., 2016), and my data are among the first suggesting aquifer eustasy during the Late Devonian.

By integrating all geochemical proxies, my dataset demonstrates, for the first time, that astronomical forcing regulated the intensity of terrestrial-marine biogeochemical linkages and thereby the extent of marine anoxia during the UKW. These findings offer new clues on the ultimate causal mechanism of the UKW mass extinction. Progressive volcanism was hypothesized to be responsible for the pulsed nature of marine anoxia during the Late

Devonian (Racki et al., 2018). However, the frequency of volcanic activity is unlikely to be connected to orbital controls. In contrast, another widely cited mechanism—the rapid evolution of land plants and the establishment of early forests (Algeo and Scheckler, 1998)— fits better with the obliquity-controlled marine anoxia scenario. Archaeopteris, the dominant arborescent lignophyte taxon during the Late Devonian, began to expand during the early

Frasnian, and archaeopterid forests were widely dispersed across Euramerica by the late

Frasnian (Lu et al., 2019). The initial radiation of forests is likely to have significantly altered terrestrial weathering patterns and liberated massive amounts of nutrients that were washed from continents into oceans (Algeo and Scheckler, 1998). My data of terrestrial biomarkers

105

provide strong, direct evidence that early afforestation played a crucial role in the development of Late Devonian marine anoxic events (Fig. 3.8; Appendix II Table S3.1), and the synchronous across-proxy oscillations show that obliquity can mediate terrestrial nutrient fluxes via sea-level fluctuations. The present study area, i.e., southernmost Euramerica, was likely invaded by early forests before the Famennian (Lu et al., 2019).

I observe that marine euxinia occurred in cycles mediated by astronomical forcing, which agrees well with the paleontological view that the biodiversity loss during the F–F mass extinction was due to unstable marine environments with recurring lethal conditions.

Under such conditions, marine life was unable to adapt to repeated environmental stresses and hence exhibited an extremely low rate of species origination (Bambach et al., 2004;

Boyer et al., 2014). Furthermore, my dataset offers new insights into both the direct, immediate and ultimate causal mechanisms of pulsed anoxic episodes during the UKW. As for the direct cause, my data support the “top-down” eutrophication model that emphasizes the importance of terrestrial nutrient inputs, as opposed to the “bottom-up” model arguing that the formation of black shales was due to the upwelling of anoxia deep waters (e.g.,

Carmichael et al., 2019). As for the ultimate cause, the role of astronomic forcing as a pacemaker of anoxic events favors the hypothesis that early forest radiation was ultimately responsible for the F–F boundary mass extinction (Algeo and Scheckler, 1998). As such, I propose a scenario where land plant evolution and eustatic sea-level were coupled and led to the UKW extinction. This scenario serves as a plausible candidate for a globally applicable mechanism unrestricted by local tectonic setting, volcanic activity, or basin geometry, and it also agrees with a previous hypothesis that complex multicausal factors were involved. My

106

findings, however, are the first to demonstrate astronomically mediated cyclicity in not only marine anoxia but also contemporaneous terrestrial and oceanic environmental drivers that collectively led to recurring lethal environmental conditions during one of the big five biotic crises in the Earth’s history.

3.7 Acknowledgments

I would like to acknowledge the support from the Gulf Coast Association of Geological

Societies and NSF EAR-1255724. I also want to thank Xinguang Wang, Huijing Fang and

Xiaoting Liu for field assistance.

107

3.8 References Algeo, T. J. & Scheckler, S. E. 1998. Terrestrial-marine teleconnections in the Devonian: links between the evolution of land plants, weathering processes, and marine anoxic events. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 353, 113–130.

Averbuch, O., Tribovillard, N., Devleeschouwer, X., Riquier, L., Mistiaen, B. & Van Vliet- Lanoe, B. 2005. Mountain building‐enhanced continental weathering and organic carbon burial as major causes for climatic cooling at the Frasnian–Famennian boundary (c. 376 Ma)? Terra Nova, 17, 25–34.

Bambach, R. K., Knoll, A. H. & Wang, S. C. 2004. Origination, extinction, and mass depletions of marine diversity. Paleobiology, 30, 522–542.

Bond, D. P., Zatoń, M., Wignall, P. B. & Marynowski, L. 2013. Evidence for shallow‐water ‘Upper Kellwasser’anoxia in the Frasnian–Famennian reefs of Alberta, Canada. Lethaia, 46, 355–368.

Boulila, S., Galbrun, B., Miller, K. G., Pekar, S. F., Browning, J. V., Laskar, J. & Wright, J. D. 2011. On the origin of Cenozoic and Mesozoic “third-order” eustatic sequences. Earth-Science Reviews, 109, 94–112.

Boyer, D. L., Haddad, E. E. & Seeger, E. S. 2014. The last gasp: trace fossils track deoxygenation leading into the Frasnian–Famennian Extinction Event. Palaios, 29, 646– 651.

Bush, R. T. & Mcinerney, F. A. 2013. Leaf wax n-alkane distributions in and across modern plants: implications for paleoecology and chemotaxonomy. Geochimica et Cosmochimica Acta, 117, 161–179.

Carmichael, S. K., Waters, J. A., Königsof, P., Suttner, T. J. & Kido, E. 2019. Paleogeography and paleoenvironments of the Late Devonian Kellwasser event: A review of its sedimentological and geochemical expression. Global and Planetary Change, 102984. 108

Chen, D. & Tucker, M. E. 2003. The Frasnian–Famennian mass extinction: insights from high-resolution sequence stratigraphy and cyclostratigraphy in South China. Palaeogeography, Palaeoclimatology, Palaeoecology, 193, 87–111.

Clark, J. & Philp, R. 1989. Geochemical characterization of evaporite and carbonate depositional environments and correlation of associated crude oils in the Black Creek Basin, Alberta. Bulletin of Canadian Petroleum Geology, 37, 401–416.

Conant, L. C. & Swanson, V. E. 1961. Chattanooga Shale and related rocks of central Tennessee and nearby areas. Geological Survey Professional Paper, 357.

Copper, P. 2002. Reef development at the Frasnian/Famennian mass extinction boundary. Palaeogeography, Palaeoclimatology, Palaeoecology, 181, 27–65.

De Vleeschouwer, D., Da Silva, A. C., Sinnesael, M., Chen, D., Day, J. E., Whalen, M. T., Guo, Z. & Claeys, P. 2017. Timing and pacing of the Late Devonian mass extinction event regulated by eccentricity and obliquity. Nature Communications, 8, 2268.

De Vleeschouwer, D., Rakociński, M., Racki, G., Bond, D. P., Sobień, K. & Claeys, P. 2013. The astronomical rhythm of late-Devonian climate change (Kowala section, Holy Cross Mountains, Poland). Earth and Planetary Science Letters, 365, 25–37.

Dipietro, J. A. 2013. Landscape evolution in the United States: an introduction to the geography, geology, and natural history, Wltham, M A, Elsevier.

Duan, Y., Severmann, S., Anbar, A. D., Lyons, T. W., Gordon, G. W. & Sageman, B. B. 2010. Isotopic evidence for Fe cycling and repartitioning in ancient oxygen-deficient settings: Examples from black shales of the mid-to-late Devonian Appalachian basin. Earth and Planetary Science Letters, 290, 244–253.

Eglinton, G. & Hamilton, R. J. 1967. Leaf epicuticular waxes. Science, 156, 1322–1335.

Elrick, M., Berkyová, S., Klapper, G., Sharp, Z., Joachimski, M. & Frýda, J. 2009. Stratigraphic and oxygen isotope evidence for My-scale glaciation driving eustasy in the Early–Middle Devonian greenhouse world. Palaeogeography, Palaeoclimatology, Palaeoecology, 276, 170–181.

109

Ettensohn, F. R. 1992. Controls on the origin of the Devonian-Mississippian oil and gas shales, east-central United States. Fuel, 71, 1487–1492.

Filer, J. K. 2002. Late Frasnian sedimentation cycles in the Appalachian basin—possible evidence for high frequency eustatic sea-level changes. Sedimentary Geology, 154, 31– 52.

Formolo, M. J., Riedinger, N. & Gill, B. C. 2014. Geochemical evidence for euxinia during the Late Devonian extinction events in the Michigan Basin (USA). Palaeogeography, Palaeoclimatology, Palaeoecology, 414, 146–154.

George, A. D., Chow, N. & Trinajstic, K. M. 2014. Oxic facies and the Late Devonian mass extinction, Canning Basin, Australia. Geology, 42, 327–330.

Grice, K., Cao, C., Love, G. D., Böttcher, M. E., Twitchett, R. J., Grosjean, E., Summons, R. E., Turgeon, S. C., Dunning, W. & Jin, Y. 2005. Photic zone euxinia during the Permian- Triassic superanoxic event. Science, 307, 706–709.

Han, J., Mccarthy, E., Calvin, M. & Benn, M. 1968. Hydrocarbon constituents of the blue- green algae Nostoc muscorum, Anacystis nidulans, Phormidium luridium and Chlorogloea fritschii. Journal of the Chemical Society C: Organic, 2785–2791.

House, M. R. 2002. Strength, timing, setting and cause of mid-Palaeozoic extinctions. Palaeogeography, Palaeoclimatology, Palaeoecology, 181, 5–25.

Jacobs, D. K. & Sahagian, D. L. 1993. Climate-induced fluctuations in sea level during non- glacial times. Nature, 361, 710.

Joachimski, M. M. & Buggisch, W. 1993. Anoxic events in the late Frasnian—Causes of the Frasnian-Famennian faunal crisis? Geology, 21, 675–678.

Joachimski, M. M. & Buggisch, W. 2002. Conodont apatite δ18O signatures indicate climatic cooling as a trigger of the Late Devonian mass extinction. Geology, 30, 711–714.

110

Joachimski, M. M., Ostertag-Henning, C., Pancost, R. D., Strauss, H., Freeman, K. H., Littke, R., Damste, J. S. S. & Racki, G. 2001. Water column anoxia, enhanced productivity and concomitant changes in δ13C and δ34S across the Frasnian–Famennian boundary (Kowala—Holy Cross Mountains/Poland). Chemical Geology, 175, 109–131.

Kaufmann, B. 2006. Calibrating the Devonian Time Scale: a synthesis of U–Pb ID–TIMS ages and conodont stratigraphy. Earth-Science Reviews, 76, 175–190.

Kodama, K. P. & Hinnov, L. A. 2014. Rock magnetic cyclostratigraphy, Wiley-Blackwell; Chichester.

Kodner, R. B., Pearson, A., Summons, R. E. & Knoll, A. H. 2008. Sterols in red and green algae: quantification, phylogeny, and relevance for the interpretation of geologic steranes. Geobiology, 6, 411–420.

Kump, L. R. & Arthur, M. A. 1999. Interpreting carbon-isotope excursions: carbonates and organic matter. Chemical Geology, 161, 181–198.

Lash, G. G. 2017. A multiproxy analysis of the Frasnian–Famennian transition in western New York State, USA. Palaeogeography, Palaeoclimatology, Palaeoecology, 473, 108– 122.

Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A. & Levrard, B. 2004. A long- term numerical solution for the insolation quantities of the Earth. Astronomy & Astrophysics, 428, 261–285.

Li, M., Hinnov, L. & Kump, L. 2019. Acycle: Time-series analysis software for paleoclimate research and education. Computers & Geosciences, 127, 12–22.

Li, M., Hinnov, L. A., Huang, C. & Ogg, J. G. 2018a. Sedimentary noise and sea levels linked to land–ocean water exchange and obliquity forcing. Nature Communications, 9, 1004.

Li, M., Kump, L. R., Hinnov, L. A. & Mann, M. E. 2018b. Tracking variable sedimentation rates and astronomical forcing in Phanerozoic paleoclimate proxy series with evolutionary correlation coefficients and hypothesis testing. Earth and Planetary Science Letters, 501, 165–179.

Li, Y. & Schieber, J. 2015. On the origin of a phosphate enriched interval in the Chattanooga Shale (Upper Devonian) of Tennessee—A combined sedimentologic, petrographic, and geochemical study. Sedimentary Geology, 329, 40–61.

Liu, J., Algeo, T. J., Jaminski, J., Kuhn, T. & Joachimski, M. M. 2019. Evaluation of high- frequency paleoenvironmental variation using an optimized cyclostratigraphic 111

framework: Example for C-S-Fe analysis of Devonian-Mississippian black shales (Central Appalachian Basin, U.S.A.). Chemical Geology, 525, 303–320.

Lu, Y., Hautevelle, Y. & Michels, R. 2013. Determination of the molecular signature of fossil conifers by experimental palaeochemotaxonomy-Part 1: The Araucariaceae family. Biogeosciences, 10, 1943–1962.

Mann, M. E. & Lees, J. M. 1996. Robust estimation of background noise and signal detection in climatic time series. Climatic Change, 33, 409–445.

Mclennan, S. M. 1993. Weathering and global denudation. The Journal of Geology, 101, 295–303.

Murphy, A. E., Sageman, B. B., Hollander, D. J., Lyons, T. W. & Brett, C. E. 2000. Black shale deposition and faunal overturn in the Devonian Appalachian Basin: clastic starvation, seasonal water‐column mixing, and efficient biolimiting nutrient recycling. Paleoceanography, 15, 280–291.

Nesbitt, H. W. & Young, G. 1982. Early Proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature, 299, 715.

Over, D. J. 2002. The Frasnian/Famennian boundary in central and eastern United States. Palaeogeography, Palaeoclimatology, Palaeoecology, 181, 153–169.

Over, D. J. 2007. Conodont biostratigraphy of the Chattanooga Shale, Middle and Upper Devonian, southern Appalachian Basin, eastern United States. Journal of Paleontology, 81, 1194–1217.

Racki, G., Rakociński, M., Marynowski, L. & Wignall, P. B. 2018. Mercury enrichments and the Frasnian-Famennian biotic crisis: A volcanic trigger proved? Geology, 46, 543–546.

Sandberg, C. A., Morrow, J. R. & Ziegler, W. 2002. Late Devonian sea-level changes, catastrophic events, and mass extinctions. Special Papers-Geological Society Of America, 473–488.

112

Scott, A. 1974. The earliest conifer. Nature, 251, 707.

Sephton, M. A., Looy, C. V., Brinkhuis, H., Wignall, P. B., De Leeuw, J. W. & Visscher, H. 2005. Catastrophic soil erosion during the end-Permian biotic crisis. Geology, 33, 941– 944.

Song, J., Crasquin, S. & Gong, Y. 2019. Late Devonian benthic ostracods from western Junggar, NW China: Implications for palaeoenvironmental reconstruction. Geological Journal, 54, 91–100.

Spaak, G., Edwards, D. S., Allen, H. J., Grotheer, H., Summons, R. E., Coolen, M. J. & Grice, K. 2018. Extent and persistence of photic zone euxinia in Middle–Late Devonian seas–Insights from the Canning Basin and implications for petroleum source rock formation. Marine and Petroleum Geology, 93, 33–56.

Stephens, N. P. & Sumner, D. Y. 2003. Late Devonian carbon isotope stratigraphy and sea level fluctuations, Canning Basin, Western Australia. Palaeogeography, Palaeoclimatology, Palaeoecology, 191, 203–219.

Thomson, D. J. 1982. Spectrum estimation and harmonic analysis. Proceedings of the IEEE, 70, 1055–1096.

Waltham, D. 2015. Milankovitch period uncertainties and their impact on cyclostratigraphy. Journal of Sedimentary Research, 85, 990–998.

Wedepohl, K. 1971. Environmental influences on the chemical composition of shales and clays. Physics and Chemistry of the Earth, 8, 307–333.

Wendler, J. E., Wendler, I., Vogt, C. & Kuss, J. 2016. Link between cyclic eustatic sea-level change and continental weathering: Evidence for aquifer-eustasy in the Cretaceous. Palaeogeography, Palaeoclimatology, Palaeoecology, 441, 430–437.

Whalen, M. T., Śliwiński, M. G., Payne, J. H., Day, J. E. J., Chen, D. & Da Silva, A. C. 2015. Chemostratigraphy and magnetic susceptibility of the Late Devonian Frasnian– Famennian transition in western Canada and southern China: implications for carbon and nutrient cycling and mass extinction. Geological Society, London, Special Publications, 414, 37–72.

113

Xu, S. & Sun, Y. 2005. An improved method for the micro-separation of straight chain and branched/cyclic alkanes: Urea inclusion paper layer chromatography. Organic Geochemistry, 36, 1334–1338.

Zhao, M. Y. & Zheng, Y. F. 2015. The intensity of chemical weathering: Geochemical constraints from marine detrital sediments of Triassic age in South China. Chemical Geology, 391, 111–122.

114

3.9 Appendix II

Figure S3. 1. Inorganic and organic proxies collected at 25 cm intervals across the Frasnian– Famennian interval of the Chattanooga Shale, central Tennessee. a Proxies for continental weathering and terrestrial inputs. CIA: chemical index of alteration. DBF: dibenzofuran. b Proxies for marine primary productivity (n-C17+19 alkanes, C27 steranes, marine palynomorph abundances). (c) Proxies for marine redox conditions (C18-20 aryl isoprenoids and MoEF) and sea-level changes (C29/C30 αβ hopane). EF: enrichment factor.

115

Figure S3. 2. Mass chromatogram (m/z 57) of aliphatic fraction from extracts of one typical sample from Upper Kellwasser (UKW) interval (the grey bar in Fig. 3.1b) of the Chattanooga Shale in central Tennessee, showing the distribution of n-alkanes, regular isoprenoids and branched alkanes. Number above symbols denote carbon number.

116

Figure S3. 3. Partial Mass chromatogram (m/z 217) showing sterane distributions in one typical sample from the Upper Kellwasser (UKW) interval (the grey bar in Fig. 3.1b) of the Chattanooga Shale in central Tennessee. Number above symbols denote carbon number.

117

Figure S3. 4. Partial mass chromatogram (m/z 191) showing hopane distributions in one typical sample from the Upper Kellwasser (UKW) interval (the grey bar in Fig. 3.1b) of the Chattanooga Shale in central Tennessee. Number above symbols denote carbon number.

118

Figure S3. 5. Partial Mass chromatogram (m/z 123) of one typical sample from the Upper Kellwasser (UKW) interval (the grey bar in Fig. 3.1b) of the Chattanooga Shale in central Tennessee. Number above symbols denote carbon number.

119

Figure S3. 6. Total ion current chromatogram of aromatic fraction from extracts of one typical sample from the Upper Kellwasser (UKW) interval (the grey bar in Fig. 3.1b) of the Chattanooga Shale, central Tennessee. Abbreviations: MN-methylnaphthalene; DMN+TMN+TeMN: dimethyl-, trimethyl- and tetramethylnaphthalene; PMN: pentamethylnaphthalene; MP: methylphenanthrene; EP+DMP+TMP: ethyl-, dimethyl- and trimethylphenanthrene.

120

Figure S3. 7. Mass chromatogram (m/z) 133 of one typical sample from the Upper Kellwasser (UKW) interval (the grey bar in Fig. 1b) of the Chattanooga Shale, central Tennessee, showing the distribution of aryl isoprenoids. Compounds with base peak m/z 134 (open circle) are tentatively identified as 3,4,5-methylsubstituted aryl isoprenoids, and compounds with base peak m/z 133 (closed circle) are tentatively identified as 2,3,6- methylsubstituted aryl isoprenoids.

121

Time-series analysis results

Figure S3. 8. Astronomical forcing targets compared with 2π multiple-taper method (MTM) power spectra of detrended Ti/Al series of the Upper Frasnian–Lower Famennian interval of the Chattanooga Shale in central Tennessee over tested sedimentation rate range of 0.27–0.39 cm/kyr. a Astronomical forcing targets from Laskar (2004) and Waltham (2015). Numbers represent target cycles in kyr/cycle. b 2π MTM power spectra of detrended Ti/Al series with significant peaks (>99% confidence level) labeled. Numbers are shown in cycles/m.

122

Figure S3. 9. Raw Ti/Al data series and 47-kyr tuned Ti/Al data series of the Upper Frasnian– Lower Famennian interval of the Chattanooga Shale, central Tennessee. a Detrended Ti/Al data series in stratigraphic domain. b 405-kyr tuned Ti/Al data series in time domain.

123

Palynology analysis result

Figure S3. 10. Palynomorphs observed from the Upper Kellwasser (UKW) interval (the grey bar in Fig. 1b) of the Chattanooga Shale, central Tennessee. a: prasinophycean cysts (Tasmanite sp.); b: trilete spore; c: chitinozoa; d: phytoclast (traceid) and prasinophycean cyst).

124

13 Table S3. 1. Pearson’s correlations among the proxies indicating redox conditions (C18–20 aryl isoprenoids, MoEF), marine productivity (δ Corg, n-C17+19 alkanes, C27 steranes), terrestrial inputs (n-C27+29+31 alkanes, norabietane, retene, tetrahydroretene, dibenzofuran (DBF)), sea-level changes (C29/C30H (C29/C30 αβ hopane), Ti/Al, Zr/Al) and weathering intensity (CIA) measured from the Upper Kellwasser interval (the grey bar in Fig. 3.1b) of the Chattanooga Shale, central Tennessee. C 18–20 n- aryl 13 n-C17+19 C27 Tetra- C29/C30 MoEF δ Corg C27+29+31 norabietane retene DBF Zr/Al CIA Ti/Al isopren alkanes steranes hydroretene H alkanes oids r 1 0.338* -0.124 0.468** 0.335* 0.659** 0.076 0.728** 0.668** 0.743** -0.17 0.235 0.231 0.143 C18–20 aryl P 0.047 0.476 0.005 0.049 0 0.665 0 0 0 0.328 0.174 0.181 0.413 isoprenoids N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r 0.338* 1 -0.379* 0.105 0.141 0.179 0.205 0.488** 0.526** 0.541** 0.404* 0.158 0.081 -0.238 MoEF P 0.047 0.025 0.549 0.418 0.305 0.237 0.003 0.001 0.001 0.016 0.365 0.642 0.169 N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r -0.124 -0.379* 1 -0.001 -0.056 0.003 -0.064 -0.197 -0.223 -0.236 -0.224 -0.099 0.217 0.192 13 δ Corg P 0.476 0.025 0.994 0.748 0.986 0.717 0.257 0.199 0.179 0.196 0.572 0.21 0.269 N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r 0.468** 0.105 -0.001 1 0.749** 0.895** 0.734** 0.572** 0.536** .617** -.387* 0.215 0.252 0.205 n-C17+19 P 0.005 0.549 0.994 0 0 0 0 0.001 0 0.022 0.215 0.144 0.237 alkanes N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r 0.335* 0.141 -0.056 0.749** 1 0.764** 0.700** 0.494** 0.432** 0.547** -0.13 0.467** 0.471** 0.217 C27 steranes P 0.049 0.418 0.748 0 0 0 0.003 0.01 0.001 0.455 0.005 0.004 0.21 N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r 0.659** 0.179 0.003 0.895** 0.764** 1 0.510** 0.789** 0.722** 0.748** -0.425 .457** 0.436** 0.363* n-C27+29+31 P 0 0.305 0.986 0 0 0.002 0 0 0 0.011 0.006 0.009 0.032 alkanes N 35 35 35 35 35 35 35 35 35 34 35 35 35 35

125

r 0.076 0.205 -0.064 0.734** 0.700** 0.510** 1 0.206 0.226 0.337 0.092 0.011 0.193 -0.18 norabietane P 0.665 0.237 0.717 0 0 0.002 0.234 0.192 0.052 0.599 0.95 0.266 0.301 N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r 0.728** 0.488** -0.197 0.572** 0.494** 0.789** 0.206 1 0.982** 0.818** -0.226 .404* 0.473** 0.303 retene P 0 0.003 0.257 0 0.003 0 0.234 0 0 0.192 0.016 0.004 0.077 N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r 0.668** 0.526** -0.223 0.536** 0.432** 0.722** 0.226 0.982** 1 .748** -0.156 0.327 0.465** 0.226 tetra- P 0 0.001 0.199 0.001 0.01 0 0.192 0 0 0.371 0.055 0.005 0.191 hydroretene N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r 0.743** 0.541** -0.236 0.617** 0.547** 0.748** 0.337 0.818** 0.748** 1 -0.13 .341* 0.321 0.239 DBF P 0 0.001 0.179 0 0.001 0 0.052 0 0 0.462 0.048 0.064 0.174 N 34 34 34 34 34 34 34 34 34 34 34 34 34 34 r -0.17 0.404* -0.224 -0.387* -0.13 -0.425* 0.092 -0.226 -0.156 -0.13 1 -0.096 -0.061 -0.534** C29/C30H P 0.328 0.016 0.196 0.022 0.455 0.011 0.599 0.192 0.371 0.462 0.583 0.729 0.001 N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r 0.235 0.158 -0.099 0.215 0.467** 0.457** 0.011 0.404* 0.327 0.341* -0.096 1 0.423* 0.409* Zr/Al P 0.174 0.365 0.572 0.215 0.005 0.006 0.95 0.016 0.055 0.048 0.583 0.011 0.015 N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r 0.231 0.081 0.217 0.252 0.471** 0.436** 0.193 0.473** 0.465** 0.321 -0.061 0.423* 1 0.388* CIA P 0.181 0.642 0.21 0.144 0.004 0.009 0.266 0.004 0.005 0.064 0.729 0.011 0.021 N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 r 0.143 -0.238 0.192 0.205 0.217 0.363* -0.18 0.303 0.226 0.239 -.534** 0.409* 0.388* 1 Ti/Al P 0.413 0.169 0.269 0.237 0.21 0.032 0.301 0.077 0.191 0.174 0.001 0.015 0.021 N 35 35 35 35 35 35 35 35 35 34 35 35 35 35 *. Correlation is significant at the 0.05 level. **. Correlation is significant at the 0.01 level.

126

References

Laskar, J., Robutel, P., Joutel, F., & Gastineau, M. 2004. A long-term numerical solution for the insolation quantities of the Earth. Astronomy & Astrophysics, 428, 261–285 (2004)

Waltham, D. 2015. Milankovitch period uncertainties and their impact on cyclostratigraphy. Journal of Sedimentary Research, 85, 990–998 (2015).

127

CHAPTER 4:

THE RISE OF FOREST STIMULATED WILDFIRES IN THE EURAMERICA DURING

THE DEVONIAN: PALEONTOLOGICAL AND GEOCHEMICAL EVIDENCE

4.1 Abstract

Fires play a crucial role in the elemental cycles and energy flows through modern and ancient ecosystems. The Devonian period represents the earliest stage of forest expansion and wildfire evolution, yet the spatiotemporal extent of wildfires remains poorly constrained.

Here, I evaluated the spatiotemporal evolution of wildfires and underlying mechanisms in the

Euramerica throughout the Devonian based on a literature synthesis of occurrences of wildfire indicators (fossil charcoals, inertinite macerals and pyrogenic polycyclic aromatic hydrocarbons (PAHs)) and early vascular plant diversity (number of species) and morphology

(axial diameter, leaf size) in Euramerica. I also performed a case study of reconstructing paleo-wildfires through the Late Devonian via analyzing inertinite maceral and pyrogenic

PAHs from an outcrop of the Chattanooga Shale in the southern Euramerica. Results from the literature synthesis reveal a rapid spatiotemporal expansion in fire occurrences across the

Euramerica during the Late Devonian. Inertinite maceral and pyrogenic PAHs showed an increasing trend for the Famennian strata of the outcrop, suggesting a marked and continuous increase in wildfires during the Late Devonian. Axial diameter, leaf size and species number of early trees showed a stepwise increase through the Late Devonian in parallel with the rise of wildfire events. Collectively, my data suggest that the expansion and diversification of 128

early trees during the Late Devonian may have stimulated wildfires by providing extensive fuels, and wildfires, in turn, may have aided the expansion of large trees and forests via selecting fire-adapted traits.

4.2 Introduction

Since the first appearance during the late Silurian, wildfire has been a significant disturbance and evolution driver to varied biomes and ecosystems (Bond and Keeley, 2005;

Bowman et al., 2009). Paleo-wildfires vary spatiotemporally throughout the Earth’s history, and their occurrences are thought to be mediated by vegetation and atmospheric oxygen levels (pO2) (Bowman et al., 2009). The Devonian time is known for the explosive diversification of vascular plants characterized by increases in size, complexity and dispersal range (Knoll, 1984; Driese and Mora, 2001; Algeo and Scheckler, 2010). This rapid evolution of early vascular plants can increase pO2 (Dahl et al., 2010; Lenton et al., 2016) and fuel availability, thereby stimulating wildfires through time and space. However, the spatiotemporal occurrences of Devonian wildfires, as well as the role that early plant evolution played in wildfire occurrences, remain poorly constrained.

My objectives are to reconstruct the spatiotemporal evolution of wildfires during the

Devonian and probe the evolutionary linkages between wildfires and vascular plants.

Although various vascular plants, including lycopsids, cladoxylaleans, lignophytes and early seed plants, were suggested as fuel sources for Devonian wildfires (Rowe and Jones, 2000), the evolutionary features that were directly associated with fire frequency are unknown.

Indeed, wildfire evidence and plant fossil records are usually presented independently, with only a scarce number of studies that have evaluated both simultaneously (e.g., Edwards and

129

Axe, 2004; Kennedy et al., 2013; Song et al., 2015). In the present study, I conducted a comprehensive literature synthesis to establish the spatiotemporal correspondence between wildfires and some key evolutionary features of early vascular plants (species, leaf morphology) in the Euramerica during the Devonian.

Furthermore, I performed a case study of reconstructing the Late Devonian paleo- wildfire history in the southern Euramerica using geochemical proxies. I analyzed inertinite macerals and pyrogenic PAHs from a Chattanooga Shale outcrop in central Tennessee, USA.

Inertinite maceral group, often used interchangeably with charcoals (Scott, 2010, 2000;

Hudspith et al., 2012), is directly related to wildfire activity (Glasspool et al., 2015). Inertinite maceral is now a general term for macroscopic ancient charcoal (Tanner and Lucas, 2016) and can be identified as cellular or arc-shaped fragments with relatively high reflectance under a reflective microscope. Pyrogenic PAHs are an array of compounds produced by the combustion and they have been widely accepted as biomarkers for wildfires through the geological history including the Devonian period (Marynowski and Filipiak, 2007;

Marynowski and Simoneit, 2009; Kaiho et al., 2013; Tulipani et al., 2015; Karp et al., 2018).

These compounds are resistant to degradation and hence are particularly useful in the absence of fossil charcoals (Scott, 2009).

This study is among the first to link the evolution of wildfires and vascular plants in the

Euramerica during the Devonian. Although it has been long recognized that fires act as an important disturbance to modern forests, their ecological impacts over a long period remain poorly understood. My findings show a rapid rise in the frequency and geographic extent of wildfires that can be linked to the diversification and spread of early trees during the Late

130

Devonian, suggesting the significance of forest-fire interactions in shaping ancient ecosystems over the geological history. The rapidly rising wildfires during the Late Devonian are analogous to modern fires that are getting more frequent and intense under the influence of human activities and climate change, and my results, therefore, offer insights into long- term consequences of fires on modern forest ecosystems.

4.3 Method and materials

4.3.1 Literature synthesis of wildfire records and Devonian vascular plants

Three wildfire indicators, those are fossil charcoals, inertinite macerals and pyrogenic

PAHs, were summarized from studies published to date to quantify wildfire occurrences during the Devonian. Wildfire evidence reported by different studies from the same or equivalent strata in the same state/province was counted as one occurrence. Temporal changes in wildfire occurrences were evaluated across seven stages of the Devonian —

Lochkovian, Pragian, Emsian, Eifelian, Givetian, Frasnian and Famennian.

In order to understand evolution patterns of Devonian vascular plants in Euramerica, I gathered data on species, maximum axial diameter, and leaf size from literature. Vascular plant species from the Lochkovian to Famennian stage of the Euramerica were mostly extracted from the Devonian plant database in Lu et al. (2019), with additional data added from more recently published works. For the calculation of diversity at the species level, we excluded occurrences of sporomorph taxa and records in open nomenclature (i.e., “sp.”,

“spp.”, “?”).

For the assessment of evolution in morphological features of vascular plants in the

Euramerica from the Lochkovian to Famennian, two approaches were adopted. The first was

131

to compile the known dimensional data of fossil plants, including maximum axial diameter, leaf length, and leaf width. The second approach was to measure the main axis diameter and leaf size based on published photographs of fossil species using an image analysis software

(ImageJ). Since compound leaves did not evolve until the Late Devonian and were associated with only a small number of vascular plants, the leaf morphology analysis included only simple leaves, which are a common feature of Devonian vascular plants. Maximum leaf length and width were defined as the largest intact dimension of the lamina or terminal nonlaminated branch. The lamina here corresponded to a pinnule or the entire leaf in a simple form. Rather than lamina, most Devonian vascular plants exhibited tree-dimensional lateral ultimate vegetative appendages, and they have been variously referred to as “dichotomous ultimate appendages”, “branch-leaf complexes” or “proto-leaves”.

4.3.2 Case study of Late Devonian wildfire reconstruction in the southern Euramerica using organic geochemical proxies

4.3.2.1 Sampling sites

The studied outcrop is located in the Chestnut Mound, central Tennessee (Fig. 4.1). The

Late Devonian Chattanooga Shale in central Tennessee was deposited in the epicontinental marine environment from the Frasnian to Famennian (de Witt et al., 1993; Schieber, 1998).

The Chattanooga Shale of the studied section is 7.25 cm in thickness, overlies the Ordovician

Leipers Formation, and is overlain by the Mississippian Maury Shale (Li and Schieber, 2015)

(Fig. 4.2a). The Frasnian–Famennian (F–F) boundary in this section was identified based on:

(i) lithological differences between the Dowelltown Member of the Frasnian age (thin layers of green-grayish shales interbedded with black shales) and the Gassaway Member of the

132

Famennian age (mostly black shales), (ii) the latest Frasnian conodont zone (i.e., MN 13

Zone) appearing near the top of the Dowelltown Member (Over, 2007), and (iii) the presence of a volcanic ash bed near the top of the Dowelltown Member that is correlative to the

Central Hill Ash occurring right below the F–F boundary in the northern Appalachian Basin

(Over, 2002). The F–F boundary is approximately 2.8 m above the base of the Chattanooga

Shale in the studied outcrop.

Figure 4.1. Distribution of paleo-wildfire records—fossil charcoals, inertinite maceral and pyrogenic PAHs during the Early, Middle and Late Devonian (Frasnian–Famennian). Details see Table 4.3 and Appendix Table S4.1. The base map is modified from Lu et al. (2019).

Thirty samples were collected at 25 cm intervals from the Chattanooga Shale of the studied outcrop. Weathered surface rocks (i.e., generally light gray) were removed and freshly exposed rocks (i.e., darker grey to black) were collected. Prior to geochemical analyses, fragments of weathered rock surfaces were further removed using knives in the lab.

Samples were then thoroughly washed using ultra-pure carbon-free water.

4.3.2.2 Organic petrography

Samples at an interval of 50 cm were selected for organic petrography analysis. Six 133

grams of each sample were demineralized in 10% HCl for 24 hours followed by 48% HF for

48 hours. The samples were then treated with a hot Schultz’s solution and sodium hydroxide, followed by a water rinse until a neutral pH was achieved. The residues were embedded in epoxy resin, polished, and observed using reflectance microscopy under a Nikon Microphot microscope. The samples were examined under immersion oil through a ×40 objective len, and the abundances of vitrinite and inertinite particles were point-counted (600 points).

4.3.2.3 Biomarker analysis

Samples at an interval of 25 cm were analyzed for molecular biomarkers. About 20 grams of powdered sample material was weighed and Soxhlet-extracted with a mixture of

Dichloromethane (DCM) /Methanol (MeOH) (97:3, v/v) for at least 72 hours. Activated copper was added to the extracts to remove the elemental sulfur. The extracted lipid was concentrated using a rotary evaporator and transferred into a pre-weighed 4mL vial. Each extract was subsequently separated into aliphatic, aromatic, and polar fractions via silica gel column chromatography (activated silica gel and alumina, 3:1 by volume). The aliphatic, aromatic, and polar fraction were eluted in sequence with petroleum ether, benzene, and methanol, respectively (4 column volume for each solvent). Each fraction was dried via rotary evaporation and then diluted with 1 mL of hexane.

The aliphatic fraction was analyzed using an Agilent 7890A gas chromatograph (GC) system coupled to a flame ionization detector (FID) equipped with a DB-1MS capillary column (60 m×0.32 mm, 0.25 μm film thickness). The oven temperature was increased at

4 ℃/min from 60 ℃, held for 2 minutes, and then to 295 ℃, held for 30 minutes. The carrier gas was nitrogen at a flow rate of 1.0 ml/min. The temperature of the inlet and FID was set at

134

295 ℃ and 300 ℃, respectively. Individual normal alkane was quantified by comparing the peak area to that of the internal standard n-C24D50. Aromatic hydrocarbons containing PAHs were analyzed on an Agilent 7890B GC - 5977A mass spectrometer equipped with a DB-

1MS capillary column (60 m×0.32 mm, 0.25 μm film thickness). Using helium as the carrier gas (1.0 ml/min), GC-oven was heated from 80 ℃ with a 2 min hold, then increased at

3 ℃/min to 220 ℃, and finally at 2 ℃/min to 300 ℃ with a 30 min hold. The source was operated in 70 eV electron impact mode at 230 ℃. Compounds were identified by comparing mass spectra and relative retention time to those reported by previous studies. Compounds were quantified using peak areas compared to known amounts of phenanthrene-D10 and dibenzothiophene-D8 that were added to the samples prior to GC-MS measurement.

Figure 4.2. Stratigraphic variations in wildfire records across the Frasnian–Famennian interval of the Upper Devonian Chattanooga Shale in central Tennessee. a Generalized biostratigraphy and stratigraphy of studied section of the Chattanooga Shale (modified based on Li and Schieber, 2015). b Pyrogenic PAHs concentrations relative to C27 normal alkane (n-C27) and abundance of inertinite (Inert) relative to sum of inertinite and vitrinite (V). Py: pyrene; BaA: benzo[a]anthrene; BeP: benzo[a/e]pyrene; BF: benzo[b/k/j]fluoranthene. 135

4.4 Results

4.4.1 Devonian wildfires in the Euramerica in time and space

Through the Devonian, a total of 46 occurrences of wildfires were reported all over the world and 26 occurrences were from the Euramerica (Figs. 4.1 and 4.3a; Table 4.1). In general, global or Euramerica wildfires remained low and below the average (6.6 global occurrences and 4.3 Euramerica occurrences per stage) from the Early Devonian to Middle

Devonian. The wildfire occurrences, however, increased rapidly from the late Frasnian to

Famennian. In Euramerica, wildfire occurrences decreased through the Early Devonian

(Lochkovian–Emsian) and remained stable through the middle Devonian to Frasnian, followed by a rapid rise during the Famennian.

Paleogeographically, the majority of wildfire events during each stage was reported from the Euramerica, ranging from 50% to 100% of the global occurrences (Figs. 4.1 and 4.3a).

The evidence of the earliest wildfires (fossil charcoals) in the Euramerica was found from the

Upper Silurian strata in England (Table 4.1). During the Early–Middle Devonian, the wildfire occurrences were limited in the Avalonia (i.e., present United Kingdom, Germany, Belgium) and northern Laurentia (i.e., the present location of northeastern Canada). From the Frasnian to Famennian, wildfires became more widespread across the Euramerica, with pronounced increases reported from the Famennian strata in the southern Euramerica (i.e., present

U.S.A.). In particular, abundant wildfire occurrences were observed from the Late Devonian strata in the north-central Appalachian Basin (i.e., New York, Pennsylvania and West

Virginia), northern Euramerica. Fewer evidence was available for the southern U.S.A., yet the detection of pyrogenic PAHs from the Woodford Shale in Oklahoma and inertinite macerals

136

and pyrogenic PAHs from the Chattanooga Shale in Alabama (Lu et al., 2019) suggest that the wildfires had occurred in southernmost Euramerica during the Famennian (Table 4.1).

Figure 4.3. Wildfire occurrences, atmospheric evolution and vascular plant evolution during the Devonian. a Fire occurrences in Euramerica. b Comparison between modeled reconstruction of changes in atmospheric pO2. c Vascular diversity at species level in Euramerica.

137

Table 4.1. A summary of early wildfire record from the Silurian to the end of the Devonian. Data based on Appendix III Table S4.1. Period Stage Fire Fire Paleogeography Depositional Rock type Occurrence Plant types occurrence indicators setting with plant fossils Silurian Pridoli 1 fossil Euramerica marine siltstone 1 rhyniophytoid charcoal (Avalonia) Early Lochkovian 5 fossil Euramerica marine, siltstone, 2 rhyniophytoid; Devonian charcoal, (Avalonia); terrestrial sandstone trilete spores; pyrogenic northeastern and cryoptospores PAHs Gondwana; mudstone Siberia Early Pragian 4 fossil Euramerica marine, carbonate, 0 NA Devonian charcoal, (Avalonia, terrestrial chert, inertinite Laurentia); shales, coaly macerals Siberia shale, sandstone, mudstone and gneiss Early Emsian 4 fossil Euramerica terrestrial gneiss, coal 0 NA Devonian charcoal, (Avalonia, and coaly inertinite Laurentia) shale macerals Middle Eifelian 1 inertinite Siberia terrestrial coal 0 NA Devonian macerals Middle Givetian 5 inertinite Euramerica marine, coal, 1 lycopsids, Devonian macerals, (Baltica, terrestrial mudstone zosterophylls, pyrogenic Laurentia), and fern like plants PAHs China, Siberia sandstone and 138

progymnosperm Late Frasnian 9 inertinite Euramerica marine, shale, coal, 1 spores Devonian macerals, (Baltica, terrestrial mudstone, pyrogenic Avalonia, limestone PAHs Laurentia), and marl northwestern Gondwana, Siberia Late Famennian 18 fossil Euramerica marine, shale, 6 rhyniophytoids, Devonian charcoal, (Baltica, terrestrial sandstone, lycopsids, inerinite Avalonia, siltstone, progymnosperm, macerals, Laurentia), mudstone, early seed plants pyrogenic Gondwana, coal and PAHs Siberia coaly shale

139

Only ten occurrences of wildfire indicators were presented together with land plant fossils (Table 4.1). The associated plant assemblages contained herbaceous and basal vascular plants, including trimerophytes, zosterophyllopsids, rhyniophytoids and lycopsids. Notably, many Famennian charcoals co-occurred with abundant axes, leaves and spores that were attributed to trees of Archaeopteris (Table 4.1).

4.4.2 Geochemical evidence of wildfires from the Chattanooga Shale in central Tennessee

Organic petrographic analysis provides information on organic matter origin. Macerals from the Chestnut Mound outcrop of the Chattanooga Shale were dominated by liptinite that varied from 50% to 70% (Appendix III Fig. S4.1). Inertinite contents were low, varying between 1% and 6%, and vitrinite contents varied from 5% and 33%. The relative abundance of inertinite to the summed abundance of inertinite and vitrinite fluctuated in a narrow range of 10.7%–14.9% (mean ± standard deviation =12.7±1.7%) in the Frasnian interval and a wide range of 11.5%–25% (20.0±4.2%) in the Famennian interval (Fig. 4.2). The values were significantly higher in the Famennian interval than in the Frasnian interval (Mann-Whitney U test P=0.03).

Changes in biomarker composition and concentrations from the Frasnian to Famennian interval were assessed. The concentrations of C27 normal alkane (n-C27) in the Frasnian interval were significantly higher than in the Famennian interval (Mann-Whitney U test

P=0.000) (Appendix III Fig. S4.2). A range of PAHs were found in both Frasnian and

Famennian intervals, including naphthalene (N), fluoranthene (Fl), pyrene (Py), benzo[a]anthrene (BaA), chrysene (Chry), benzo[b/k/j]fluoranthene (BF), benzo[a/e]pyrene

(BeP) and benzo[g,h,i]perylene (BgP) (Fig. 4.4). The concentrations of these compounds

140

relative to rock all showed significantly higher values in the Frasnian interval than in the

Famennian interval (Mann-Whitney U test P=0.000–0.006) (Appendix III Fig. S4.2). The concentrations of Py, BaA, BeP and BF were normalized to n-C27 to indicate more occurrences of wildfires. The Py/(n-C27+Py), BaA/(n-C27+BaA), BeP /(n-C27 + BeP), and

−4 BF/(n-C27+BF) show similar variations through the section, and they vary between 1.8×10 –

3.2×10−3 (1.3±0.9×10−3), 5.9×10−5–1.4×10−1 (5.6±4.6×10−2), 6.7×10−4–2.5×10−2

(8.1±7.1×10−3) and 1.9×10−3–2.5×10−2 (1.3±0.8×10−2), respectively, for the Frasnian interval, and between 0.00–0.04 (0.01±0.01), 0.03–0.59 (0.26±0.15), 0.00–0.18 (0.06±0.05) and 0.01–0.32 (0.12±0.09) respectively, for the Famennian interval. All ratios showed significantly higher values in the Famennian than in the Frasnian (Mann-Whitney U test

P=0.00) (Fig. 4.2). In addition, a series of diagnostic ratios were calculated following Yunker et al. (2011) to assess the source of PAHs. The ratios of Fl/(Fl+Py), BaA/(BaA+Chry), and

BF/(BF+BeP) ranged between 0.5–1.0 (0.6±0.1), 0.9–1.0 (1.0±0.0) and 0.4–0.9 (0.8±0.1), respectively, through the entire section (Fig. 4.5).

The relative abundance of PAHs of different rings that can indicate combustion sources was calculated (Fig. 4.6). The most abundant compounds in both Frasnian and Famennian intervals were 4-ring PAHs (i.e., Fl, Py, BaA, Chry) (>50%) followed by 5-ring PAHs (i.e.,

BF, BeP). The percentages of 2-ring (i.e., N), 3-ring (i.e., Fl) and 4-ring PHAs averaged

3.5±2.1%, 2.9±2.0% and 67.4±6.8% respectively, in the Frasnian interval and they decreased to 2.4±0.8%, 2.2±0.6% and 63.4±3.2%, respectively, in the Famennian. In contrast, percent 5- ring and 6-ring PAHs averaged 26.1±5.9% and 2.3±0.8%, respectively, in the Frasnian and increased to 28.5±4.1% and 3.1±4.3%, respectively, in the Famennian.

141

Figure 4.4. Total ion current chromatogram of aromatic fraction from extracts of one typical sample from the Upper Devonian Chattanooga Shale, central Tennessee. Red dots indicate pyrogenic PAHs. MN: methylnaphthalene; DMN+TMN+TeMN: dimethyl-, trimethyl- and tetramethylnaphthalene; PMN: pentamethylnaphthalene; MP: methylphenanthrene; EP+DMP+TMP: ethyl-, dimethyl- and trimethylphenanthrene.

Figure 4.5. PAH diagnostic ratios derived from the Frasnian–Famennian interval of the Upper Devonian Chattanooga Shale in central Tennessee. The red dash line represents the value of 0.7. Fl: fluoranthene; Py: pyrene; BaA: benzo[a]anthrene; Chry: chrysene; BF: benzo[b/k/j]fluoranthene; BeP: benzo[a/e]pyrene. 142

Figure 4.6. Change in percentage of pyogenic PAHs with different ring count across the Frasnian–Famennian interval of the Upper Devonian Chattanooga Shale in central Tennessee.

4.4.3 Diversification and key character evolution of vascular plants in the Euramerica during the Devonian

The plant taxa from the Devonian of Euramerica are compiled and tabulated with a high- rank classification (Table 4.3). An overall increase in species-level diversity occurred during the Devonian (Fig. 4.3c). Paratrachaeophytes (i.e., Cooksonioids and Rhyniophytina), psilophytophyta (e.g., Cladoxylopsida), and stem lycopodiopsida (e.g., Zosterophyllopsida) are the three major components of the Early–Middle Devonian floras in the Euramerica (Fig.

4.3c, Table 4.3). The Middle/Late Devonian (i.e., Givetian/Frasnian) boundary was characterized by a decline in shrubby, primitive plants, such as Zosterophyllopsids, followed by explosive diversification of lignophytes, lycopodiopsida and pteridophyte ferns through the Late Devonian (Fig. 4.3c, Table 4.3). Lignophytes first appeared during the Emsian and rose to dominate the Middle and Late Devonian forests (Fig. 4.3c, Table 4.3). The most

143

marked diversification of lignophytes occurred in the Famennian, when they increased ~1.5 times in the number of species.

Figure 4.7. Maximum axial diameter (a), leaf length (b) and width (c) of Devonian vascular plants from the Euramerica. Data of maximum axial diameter and leaf size are based on Appendix III Table S4.2 and S4.3, respectively.

144

The maximum diameter of main axes of vascular plants from the Euramerica increased through the Devonian (Fig. 4.7a and 4.8a). Among the Early Devonian lycopsids,

Drepanophycus had the widest axis up to 4.2 cm (Appendix III Table 4.2 and references therein). The axis further increased up to 6 cm in Protolepidodendron during the Middle

Devonian and to ~40cm in Leptophloeum during the Late Devonian (Appendix III Table S4.2 and references therein). For euphyllophytes, the axial diameters were less than 10 mm during the Middle Devonian (Fig. 4.7a). The largest diameter of the Middle Devonian euphyllophytes was reported in cladoxylopsid Eospermatopteris erianus trees, reaching as wide as 106 cm (Fig. 4.7a, Appendix III Table 4.2 and references therein). The mean maximum diameter increased dramatically through the Late Devonian with a rise of

Archaeopteris, a major genus of lignophytes (Fig. 4.7a and 4.8a). The trunks of Archaeopteris may have reached 150 cm in diameter (Fig. 4.7a, Appendix III Table 4.2 and references therein).

The leaf size of vascular plants in Euramerica increased through the Devonian, as shown by an increasing trend in the mean and maximum leaf length and width (Fig. 4.7b, c and 4.8 b,c). The length and width of lycopsids, which typically bore microphylls, remained at consistently low values through the Middle to early Late Devonian (Frasnian) (length≤3 cm, width≤0.4 cm) but increased rapidly from the Frasnian to Famennian (Figs. 4.7b, c). The outliner was Enigmophyton superbum, a stem lycopsid that was reported only from the Upper

Middle (or Lower Upper) Devonian strata in Norway, and it bore leaves up to 16 cm long and

12 cm wide (Appendix III Table S4.3 and references therein). With the exception of the single occurrence of Enigmophyton, the longest vegetative leaves of Devonian lycopsids were

145

recorded in Famennian Jurinodendron (=Cyclostigma) kiltorkense and had a length up to 15 cm (Appendix III Table S4.3 and references therein). For euphyllophytes, maximum and average leaf length and width increased stepwise through the Devonian. The longest and widest leaves were reported in Eviostachya hoegii from the Famennian, showing dissected leaves up to 7.5 cm long and 3.8 cm wide (Appendix III Table S4.3 and references therein).

During the Early–Middle Devonian, early euphyllophytes, represented by Psilophyton, usually had three-dimensionally dichotomous sterile ultimate appendages rather than laminate leaves (Appendix III Table S4.3), and they were treated as megaphyll precursors

(Beerling and Fleming, 2007). Leaf morphology diversified mostly during the Late Devonian, when fan- or wedge-shaped leaves evolved in the lignophytes (Appendix III Table 4.3). Some fragmentary laminate structures from the Emsian and Givetian strata in Norway, named

Platyphyllum and Enigmophyton, may represent the oldest fan- or wedge-shaped leaves in the

Euramerica, yet their morphology and affinities have been called into question (Appendix III

Table S4.3).

146

Figure 4.8. Average values of maximum axial diameter, leaf length and width of Devonian vascular plants from the Euramerica. The bar represents the 95% CL. Data of axial diameter and leaf size are based on Appendix III Table S4.2 and S4.3, respectively.

147

4.5 Discussion

4.5.1 Pyrogenic origins of inertinite and PAHs

I used inertinite macerals and PAHs as geochemical proxies of wildfires in the case study. These compounds are products of high-temperature alteration of hydrocarbons

(Arinobu et al., 1999) or land plant and fossil fuel combustion (Murchison and Raymond,

1989; Grice et al., 2007, Marynowski and Filipiak, 2007, Yunker et al., 2011, Kaiho et al.,

2013; Tulipani et al., 2015). However, the studied section showed consistently low maturity indicated by Tmax being ≤438℃ that excluded the influences of maturity and volcanisms on the abundances of the inertinites and pyrogenic PAHs. The diagnostic ratios of methylated and nonmethylated PAHs suggest that PAHs were derived predominantly from pyrogenic sources but not petrogenic sources. More specifically, BF/(BF+BeP) and Fl/(Fl+Py) ratios being greater than 0.7 and 0.5, respectively, and BaA/(BaA+Chry) ratios <1.0 indicate a pyrogenic origin of PAHs (Yunder et al., 2011). All but two samples had BF/(BF+BeP) larger than 0.7, and all samples had Fl/(Fl+Py) ratios >0.5 and BaA/(BaA+Chry) ratios <1.0, suggesting biomass burning, rather than fossil fuel burning, as the primary source (Fig. 4.5).

Furthermore, the significant correlations between inertinite/(inertinite+vitrinite) and the concentrations of pyrogenic PAHs (normalized by n-C27) support a common source (Table

4.2) support a common source and indicate that inertinite also originated from biomass burning.

148

Py/(n-C27+Py) BaA/(n-C27+BaA) BeP/(n-C27 + BeP) BF/(n-C27+BF) Inert./(Inert.+V) ** ** ** * Py/(n-C27+Py) r 1 0.890 0.948 0.941 0.550 P 0.000 0.000 0.000 0.027 N 29 29 29 29 16

** ** ** ** BaA/(n-C27+BaA) r 0.890 1 0.877 0.973 0.825 P 0.000 0.000 0.000 0.000 N 29 30 30 30 16

** ** ** * BeP/(n-C27 + BeP) r 0.948 0.877 1 0.939 0.567 P 0.000 0.000 0.000 0.022 N 29 30 30 30 16

** ** ** ** BF/(n-C27+BF) r 0.941 0.973 0.939 1 0.770 P 0.000 0.000 0.000 0.000 N 29 30 30 30 16 inertinite/(inertinite+vi r 0.550* 0.825** 0.567* 0.770** 1 trinite) P 0.027 0.000 0.022 0.000 N 16 16 16 16 16 **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level. Py: pyrene; BaA: benzo[a]anthrene; BF: benzo[b/k/j]fluoranthene; BeP: benzo[a/e]pyrene; Inert.: inertinite; V: vitrinite 149

The concentrations of pyrogenic compounds in sedimentary sequences can also be influenced by the transportation and preservation of these compounds. From this perspective, both the absolute concentrations of pyrogenic PAHs and n-C27 may result from changes in riverine inputs to marine environment. In the studied section of the Chattanooga Shale, the normalized PAH concentrations relative to n-C27 did not correlate with the absolute concentration of pyrogenic PAHs (relative to the mass of the sediments), suggesting the normalized concentrations were unlikely to be due to changes in the strength of transportation and/or preservation but reflected changes in wildfires.

4.5.2 Spatiotemporal expansion of wildfires in the Euramerica

My synthesis of the global distribution of wildfire indicators through the Devonian illustrates a rapid expansion of wildfires across the Euramerica through the Famennian (Fig.

4.1). This trend is unlikely to have resulted from rock types and depositional environments because wildfire evidence has been reported from different types of rocks deposited in a wide range of depositional settings (Table 4.1).

My geochemical data from the Chattanooga Shale in the southern Appalachian Basin are consistent with the broad global pattern and show elevated occurrences of wildfires in the southern Euramerica during the Late Devonian. Although no macro-charcoal fossils were found, a considerable amount of inertinite macerals appeared in my samples. The abundance of inertinite relative to vitrinite has been commonly used to indicate fire occurrences, where vitrinite is thought to originate from unburned woody tissues (Taylor et al., 1998). In the studied section, an increase in inertinite/(vitrinite+inertinite) during the Famennian suggests elevated occurrences of wildfires at that time (Fig. 4.2). Further evidence for wildfires arose

150

from changes in PAHs (i.e., Fl, Py, BaA, BF and BeP) identified throughout the studied section of the Chattanooga Shale The abundance of pyrogenic PAHs relative to land plant biomarkers also indicate variations in fire occurrences (Karp et al., 2018; Denis et al., 2017).

In my samples, PAHs/ n-C27 showed consistently low values in the Frasnian interval but an increasing trend through the Famennian interval (Fig. 4.2), suggesting increasing fire occurrences across the F–F boundary.

4.5.3 Causes for increased fires during the Famennian

Fire dynamics depends on and reflects three factors—ignition source, oxygen, and fuel availability (Scott, 2000). Lighting acts as the most common source of ignition for modern wildfires not caused by humans. However, it is unlikely that lightning intensity changed significantly over the Phanerozoic (Glasspool and Scott, 2010). Moreover, lightning usually occurs frequently, as indicated by the high frequency of modern global lightning strikes

(Christian et al., 2003). Fulgurite occurrences (fossil record of lightning) suggest that lightning is unlikely the limiting factor for paleo-wildfire occurrences (Scott and Jones, 1991;

Glasspool et al., 2015). More recent research on modern ecosystems also questions the role of lightening in wildfire ignition and suggests that the number of lightning strikes is not directly related to fire area and intensity (Bistinas et al., 2014).

Excluding the ignition source leaves two other factors to consider for the rise in wildfire during the Late Devonian—pO2 and fuel availability. Throughout the Paleozoic, atmospheric oxygen has dominant influences on the occurrence of fire (Scott and Glasspool, 2006). In turn, wildfire proxies (e.g., inertinite macerals) have been used to reconstruct past pO2 (e.g.,

Belcher and McElwain, 2008; Shen et al., 2011; Glasspool et al., 2015; Rimmer et al., 2015;

151

Baker et al., 2017). Atmospheric pO2 is presumed to have changed dramatically during the

Devonian, based on a number of quantitative models (e.g., Bergman et al., 2004; Berner,

2009, 2006, 2003; Algeo and Ingall, 2007; Glasspool et al., 2015). These models calculate pO2 based on carbon cycling and mass balance (e.g., Berner et al., 2003), isotopic data (e.g.,

Berner, 2006), and the inertinite content of Devonian coals (Glasspool and Scott, 2010;

Glasspool et al., 2015). Although large uncertainties and disagreement exist, most of the models show an overall increasing trend in pO2 through the Late Devonian (Fig. 4.3b).

Recent models estimated that the Famennian oxygen levels exceeded 17 % and peaked at

20% by the end of the Late Devonian (e.g., Berner 2009; Glasspool et al., 2015). Combustion experimentation suggested that wildfires would be completely suppressed if pO2 dropped below 16.5 % but become uncontrollable on a global scale if pO2 exceeded 35% (Belcher et al., 2010). Thus, sufficiently high atmospheric oxygen concentrations during the Late

Devonian provide suitable conditions for increased fires. However, the occurrence of wildfire events during the Famennian was not geographically evenly distributed, with the majority reported from the Famennian localities in the Euramerica (Fig. 4.1). This geographic pattern may reflect collection biases to some extent, yet it may also suggest wildfire responded to local environments with specific flora assemblages that determined fuel availability.

The expansion of wildfires corresponded to the spatiotemporal evolution in vascular plants during the Devonian. Temporally speaking, fires were scarce during the Early–Middle

Devonian and started to increase during the Late Devonian when trees and tree-like plants began to diversify and spread (Fig. 4.3c). Early trees originating in the Givetian included two main groups, i.e., Archaeopteris trees and cladoxylopsid trees (Meyer-Berthaud et al., 2010).

152

It was suggested that the Archaeopteris-dominated forest was established by the Late

Frasnian and spread rapidly through the Famennian (Scheckler et al., 2006; Lu et al., 2019).

Spatially, fire records during the Early–Middle Devonian were only reported from the north- central Euramerica, and they spanned to the entire Euramerica during the Famennian. This pattern coincided with the southward dispersal of early trees (Lu et al., 2019). Furthermore, large trees, represented by Archaeopteris, would have provided fuels needed to sustain more active fires during the Famennian. My geochemical data from the southern Euramerica provide further support to this interpretation. Specifically, PAHs with more rings indicate burning at higher temperatures (Schmeltz and Hoffmann, 1976; Marynowski and Simoneit,

2009). The enhanced proportion of more complex PAHs observed in the Famennian interval suggested high-temperature burning that is associated with crown fires of standing trees (Fig.

4.6). In modern forests, ground and surface fires that burn shrub-dominated vegetation and forest litters have temperatures less than 350℃, whereas fires spreading to the crowns have a combustion temperature higher than 600℃ (Scott and Jones, 1994; Scott, 2000).

4.5.4 Fuel accumulation and fire adaptations during the Late Devonian

The data synthesis also shows that early trees underwent notable diversifications and major morphological innovations during the Late Devonian (Figs. 4.3c and 4.7, Table 4.3), which could play an important role in modifying the susceptibility of biomass to fires. Early trees evolved high volumes of wood, indicated by increasing axial diameter, that can serve as a fuel source, such as arborescent lycopods exhibiting thick cortex during the Late Devonian

(Fig. 4.7a) (Meyer-Berthaud et al., 1999, Meyer-Berthaud and Decombeix, 2009). The widespread appearance of laminate megaphylls and leafy branches during the Late Devonian

153

created additional new fuels (Appendix III Table S4.3). In contrast, early vascular plants from the Early through Middle Devonian, with several exceptions (e.g., Platyphyllum), lack any true megaphyllous leaves (Osborne et al., 2004). The laminated leaves first appeared widely in lignophytes with the evolutionary radiation of Archaeopteris from the Frasnian, and soon the leaf diversification of the pteridosperms followed during the Famennian (Appendix III

Table S4.3). Leafy branches evolved by early trees might serve as a ladder fuel and facilitate the spreading of crown fries among plants (Davies and Belcher, 2013). The Late Devonian trees, represented by Archaeopteris, are thought to become deciduous (Meyer-Berthaud et al.,

1999; Cressler, 2006) that could have built up fuels on the ground and promoted regular surface fires at that time. In modern fire ecosystems, leaf litter and low-lying shrubs are most prone to ignition and fires (Van Wagtendonk, 2006; Scott, 2010). Furthermore, arborescent and tall plants had flattened branch systems and leaves that could provide a canopy for understory vegetation in the Late Devonian, which was dominated by fern-like plants (e.g.,

Rhacophyton) (Greb et al., 2006). In addition to evolutionary innovations, the colonization of early trees in drier extrabasinal and even upland settings (Retallack and Huang, 2011) could have created more fire-prone ecosystems.

From the Frasnian to Famennian, the lycopods and pteridosperms became diversified

(Fig. 6c, Table 4.3), likely as a result of high environmental disturbance. Most of these plants were cheaply constructed and grew quickly after fires (DiMichele, 2014), and hence they could provide rapidly recycled fuels for surface fires (Glasspool et al., 2015). Among these early plants, many ferns exhibit clonal growth (DiMichele and Phillips, 2002; Greb et al.,

2006) that allow them to survive frequent forest fires (Callaghan et al., 1992).

154

Table 4.3. Species-level diversity of vascular plants from the Euramerica through the Devonian. Period Stage Para. basal stem Lyc. Lyc. Pter. Psil. Lign. Total Euphyll. species No. % No. % No. % No. % No. % No. % No. %

Early Loch. 20 69.0 0 0.0 7 24.1 1 3.4 0 0.0 1 3.4 0 0.0 29 Devonian

Early Prag. 10 16.9 1 1.7 27 45.8 7 11.9 6 10.2 8 13.6 0 0.0 59 Devonian

Early Em. 6 6.7 2 2.2 32 36.0 13 14.6 11 12.4 23 25.8 2 2.2 89 Devonian

Middle Eif. 2 5.4 0 0.0 4 10.8 6 16.2 12 32.4 7 18.9 6 16.2 37 Devonian

Middle Giv. 0 0.0 0 0.0 5 6.8 13 17.6 28 37.8 11 14.9 17 23.0 74 Devonian

Late Fras. 2 4.0 0 0.0 1 2.0 6 12.0 16 32.0 1 2.0 24 48.0 50 Devonian

155

Late Fam. 0 0.0 0 0.0 2 2.6 15 19.5 20 26.0 2 2.6 38 49.4 77 Devonian

Para.: paratracheophyte; Euphyll.: Euphyllophytes; Lyc.: lycopodiopsida; Pter.: pteridophyte; Psil.: psilophyte; Lign.: lignophytes. Loch: Lochkovian; Prag.: Pragian; Em.: Emsian; Eif.: Eifelian; Giv.: Givetian; Fras.: Frasnian; Fam.: Famennian. No.: Species number; %: percentage of species number of a vascular plant group relative to total vascular plant species.

156

Vegetation structure and growth strategy are important modulators of fire occurrences and propagation (Scott et al., 2014), and fires, in turn, play a role in the maintenance or evolution of vegetation structure (Glasspool et al., 2015 and refs therein). Correspondingly, a number of traits evolved by plants can be considered advantageous in fire-prone ecosystem or biome.

In modern forests, trees of 1–2 m in height can significantly reduce mortality in fires

(Scott et al., 2014) and may be favored in a fire-prone environment. Since it is difficult to determine the height of a fossil tree based on stem fragments alone, the maximum axial diameter has been used to speculate overall plant height (e.g., Niklas, 1994; Retallack and

Huang, 2011; Xue et al., 2018). My data increases in the maximum axial diameter from the

Frasnian to Famennian, indicting trees and tree-sized plants became taller as wildfires became more frequent (Fig. 7a), suggesting a possible role that wildfires played in selecting tall trees. High trees are associated with a large gap between the ground surface and branched crown that could have prevented fires moving up through ladder fuels. Other studies also pointed out that branch abscission being another evolutional feature reducing the accumulation of ladder fuels and severe crown fire fires (e.g., Looy, 2013).

As the most flammable part of a plant (Davies and Belcher, 2013), leaf morphology and size are known to impact litter flammability and thereby fire behavior (De Magalhaes and

Schwilk, 2012). My data show that leaf blades became longer and wider, and leaf areas became greater from the Frasnian to Famennian (Fig. 4.8b, c). Concurrently, undoubtable fan- or wedged-shaped leaves first appeared in Archaeopteris trees during the Late Devonian

(Appendix III Table S4.3). The more frequent wildfires during the Late Devonian may

157

facilitate the evolution of these leaf features. Relative to small needle-like leaves, these large broad leaves had a high surface area to volume ratios and therefore were less flammable

(Belcher et al., 2010; Scott et al., 2010). Previous studies of modern ecosystem also reported that species that increased after fires had larger leaves (Slik et al., 2010).

4.5.5 Implications to marine biotic crisis

The biogeochemical significance of extensive fires during the Late Devonian could reach beyond terrestrial ecosystems and have important implications for marine ecosystems.

There were multiple extinction pulses of shallow marine organisms through the Late

Devonian due to widespread development of marine anoxia (Joachimski and Buggisch, 1993;

Bond et al., 2004). Enhanced nutrient fluxes from the terrestrial landscape due to vascular plant evolution has been proposed as a possible causal mechanism of marine anoxia

(Marynowski and Filipiak, 2007; Kaiho et al., 2013). My findings suggest a possible role of enhanced wildfires during the Late Devonian played in the mass extinctions. Wildfires could enhance terrestrial nutrient fluxes via burning vegetation and litter layers, increasing surface runoff, and magnifying soil erosion and nutrient mobilization (Kump, 1988; Shakesby and

Doerr, 2006). The massive kill of reef ecosystems during the Late Devonian mass extinction events could result from dramatic increases in turbidity due to raid increases in soil particles and charcoals in coastal seas after massive fires.

On the other hand, massive fires may also facilitate the end of marine anoxia and the recovery of marine ecosystems. My data suggest a rapid rise of wildfires right after the F–F boundary, which could be due to a positive fire-pO2 feedback loop. Following massive wildfires in the Famennian, tremendous amounts of charcoal fragments would accumulate on

158

land or be delivered into the oceans, leading to widespread deposition of organic-rich rocks.

This enhancement in organic carbon burial would increase the buildup of oxygen in the atmosphere (Fig. 4.3b) (Berner, 2004; Canfield, 2005). Rises in pO2 can accelerate the termination of marine anoxia through ocean ventilation (Handoh and Lenton, 2003).

Stimulated forest fires concurrent with the rapid evolution of vascular plants during the Late

Devonian, therefore, may have strengthened the terrestrial-marine biogeochemical linkages that generated far-reaching ecological consequences in Earth’s system.

4.6 Conclusions

In the present study, I reconstructed the fire history on the Euramerica during the

Devonian and discussed the causal mechanisms and ecological implications of the spatiotemporal evolution of wildfires. I compiled the published occurrences of three wildfire proxies (i.e., fossil charcoal, inertinite macerals and combustion-derived PAHs) around the world and found a rapid rise in wildfire occurrences that are concurrent with rapid geographic expansion across the Euramerica during the Famennian. My geochemical data from the

Chattanooga Shale in the southern Euramerica confirmed more frequent wildfires through the

Famennian. The rapid spread of wildfires in the Euramerica during the Famennian may be tied to the expansion and diversification of early trees, represented by Archaeopteris. My synthesis on the evolution leaf morphological features of Devonian vascular plants showed large increases in the size of tree leaves and tree bodies through the Late Devonian, which could be an adaptive feature selected by more frequent and intensive wildfires. My results are among the first to establish the spatiotemporal correspondence between wildfires and vascular plants during the Devonian. It is also notable that a remarkable rise in wildfires

159

occurred across the F–F mass extinction boundary, which may suggest that wildfires aggravated marine anoxia by accelerating terrestrial runoffs to coastal seas, highlighting the significance of wildfires in modulating terrestrial-marine biogeochemical linkages in geological history.

160

4.7 References Algeo, T. J. & Ingall, E. 2007. Sedimentary Corg: P ratios, paleocean ventilation, and Phanerozoic atmospheric pO2. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 130–155.

Algeo, T. J. & Scheckler, S. E. 2010. Land plant evolution and weathering rate changes in the Devonian. Journal of Earth Science, 21, 75–78.

Arinobu, T., Ishiwatari, R., Kaiho, K. & Lamolda, M. A. 1999. Spike of pyrosynthetic polycyclic aromatic hydrocarbons associated with an abrupt decrease in δ13C of a terrestrial biomarker at the Cretaceous-Tertiary boundary at Caravaca, Spain. Geology, 27, 723–726.

Baker, S. J., Hesselbo, S. P., Lenton, T. M., Duarte, L. V. & Belcher, C. M. 2017. Charcoal evidence that rising atmospheric oxygen terminated Early Jurassic ocean anoxia. Nature communications, 8, 1–7.

Beerling, D. J. & Fleming, A. J. 2007. Zimmermann's telome theory of megaphyll leaf evolution: a molecular and cellular critique. Current opinion in plant biology, 10, 4–12.

Belcher, C. M. & Mcelwain, J. C. 2008. Limits for combustion in low O2 redefine paleoatmospheric predictions for the Mesozoic. Science, 321, 1197–200.

Bergman, N. M., Lenton, T. M. & Watson, A. J. 2004. Copse: a new model of biogeochemical cycling over Phanerozoic time. American Journal of Science, 304, 397– 437.

Berner, R. A. 2004. The Phanerozoic carbon cycle: CO2 and O2, Oxford University Press on Demand.

Berner, R. A. 2006. Geocarbsulf: A combined model for Phanerozoic atmospheric O2 and CO2. Geochimica et Cosmochimica Acta, 70, 5653–5664.

Berner, R. A. 2009. Phanerozoic atmospheric oxygen: New results using the GEOCARBSULF model. American Journal of Science, 309, 603–606.

Berner, R. A., Beerling, D. J., Dudley, R., Robinson, J. M. & Wildman Jr, R. A. 2003. Phanerozoic atmospheric oxygen. Annual Review of Earth and Planetary Sciences, 31, 105–134. 161

Bistinas, I., Harrison, S., Prentice, I. & Pereira, J. 2014. Causal relationships vs. emergent patterns in the global controls of fire frequency. Biogeosciences, 11, 5087–5101.

Bond, W. J. & Keeley, J. E. 2005. Fire as a global ‘herbivore’: the ecology and evolution of flammable ecosystems. Trends in ecology & evolution, 20, 387–394.

Bond, D., Wignall, P. B. and Racki, G., 2004. Extent and duration of marine anoxia during the Frasnian–Famennian (Late Devonian) mass extinction in Poland, Germany, Austria and France. Geological Magazine, 141, 173–193.

Bowman, D. M., Balch, J. K., Artaxo, P., Bond, W. J., Carlson, J. M., Cochrane, M. A., D’Antonio, C. M., Defries, R. S., Doyle, J. C. & Harrison, S. P. 2009. Fire in the Earth system. Science, 324, 481–484.

Callaghan, T. V., Carlsson, B. Å., Jónsdóttir, I. S., Svensson, B. M. & Jonasson, S. 1992. Clonal plants and environmental change: introduction to the proceedings and summary. Oikos, 63, 341–347.

Canfield, D. E. 2005. The early history of atmospheric oxygen: homage to Robert M. Garrels. Annual Review of Earth and Planetary Sciences, 33, 1–36.

Christian, H. J., Blakeslee, R. J., Boccippio, D. J., Boeck, W. L., Buechler, D. E., Driscoll, K. T., Goodman, S. J., Hall, J. M., Koshak, W. J. & Mach, D. M. 2003. Global frequency and distribution of lightning as observed from space by the Optical Transient Detector. Journal of Geophysical Research: Atmospheres, 108, ACL 4–1-ACL 4–15.

Dahl, T. W., Hammarlund, E. U., Anbar, A. D., Bond, D. P., Gill, B. C., Gordon, G. W., Knoll, A. H., Nielsen, A. T., Schovsbo, N. H. & Canfield, D. E. 2010. Devonian rise in atmospheric oxygen correlated to the radiations of terrestrial plants and large predatory fish. Proceedings of the National Academy of Sciences, 107, 17911–17915.

Davies, G. M. 2013. ‘Understanding fire regimes and the ecological effects of fire’, in Belcher, C. M. (ed) Fire Phenomena and the Earth System: an Interdisciplinary Guide to Fire Science. Wiley Blackwell, Chichester, 97–124.

De Magalhaes, R. M. & Schwilk, D. W. 2012. Leaf traits and litter flammability: evidence for non‐additive mixture effects in a temperate forest. Journal of Ecology, 100, 1153–1163.

162

De Witt, W., Roen, J. & Wallace, L. 1993. Stratigraphy of Devonian black shales and associated rocks in the Appalachian Basin, United States Geological Survey Bulletin, 1909-B, B1–B47.

Denis, E. H., Pedentchouk, N., Schouten, S., Pagani, M. & Freeman, K. H. 2017. Fire and ecosystem change in the Arctic across the Paleocene-Eocene Thermal Maximum. Earth and Planetary Science Letters, 467, 149–156.

Dimichele, W. A. 2014. Wetland-dryland vegetational dynamics in the Pennsylvanian ice age tropics. International Journal of Plant Sciences, 175, 123–164.

Dimichele, W. A. & Phillips, T. L. 2002. The ecology of Paleozoic ferns. Review of Palaeobotany and Palynology, 119, 143–159.

Driese, S. G. & Mora, C. I. 2001. ‘Diversification of Siluro-Devonian plant traces in paleosols and influence on estimates of paleoatmospheric CO2 levels’. In Gensel, P.G., and Edwards, D. (eds.) Plants Invade the Land: Evolutionary and Environmental Perspectives. Columbia University Press, New York 237–253.

Edwards, D. & Axe, L. 2004. Anatomical Evidence in the Detection of the Earliest Wildfires. Palaios, 19, 113–128.

Glasspool, I. J. & Scott, A. C. 2010. Phanerozoic concentrations of atmospheric oxygen reconstructed from sedimentary charcoal. Nature Geoscience, 3, 627–630.

Glasspool, I. J., Scott, A. C., Waltham, D., Pronina, N. & Shao, L. 2015. The impact of fire on the Late Paleozoic Earth system. Front Plant Science, 6, 756.

Greb, S. F., Dimichele, W. A. & Gastaldo, R. A. 2006. Evolution and importance of wetlands in earth history. Geological Society of America Special Papers, 399, 1–40.

Grice, K., Nabbefeld, B. & Maslen, E. 2007. Source and significance of selected polycyclic aromatic hydrocarbons in sediments (Hovea-3 well, Perth Basin, Western Australia) spanning the Permian–Triassic boundary. Organic Geochemistry, 38, 1795–1803.

Handoh, I. C. & Lenton, T. M. 2003. Periodic mid-Cretaceous oceanic anoxic events linked by oscillations of the phosphorus and oxygen biogeochemical cycles. Global Biogeochemical Cycles, 17.

Hudspith, V., Scott, A., Collinson, M., Pronina, N. & Beeley, T. 2012. Evaluating the extent to which wildfire history can be interpreted from inertinite distribution in coal pillars: An example from the Late Permian, Kuznetsk Basin, Russia. International Journal of Coal Geology, 89, 13–25.

163

Joachimski, M. M. and Buggisch, W., 1993. Anoxic events in the late Frasnian—Causes of the Frasnian–Famennian faunal crisis?. Geology, 21, 675-678.

Kaiho, K., Yatsu, S., Oba, M., Gorjan, P., Casier, J.-G. & Ikeda, M. 2013. A forest fire and soil erosion event during the Late Devonian mass extinction. Palaeogeography, Palaeoclimatology, Palaeoecology, 392, 272–280.

Karp, A. T., Behrensmeyer, A. K. & Freeman, K. H. 2018. Grassland fire ecology has roots in the late Miocene. Proceedings of the National Academy of Sciences, 115, 12130– 12135.

Kennedy, K. L., Gibling, M. R., Eble, C. F., Gastaldo, R. A., Gensel, P. G., Werner- Zwanziger, U. & Wilson, R. A. 2013. Lower Devonian coaly shales of northern New Brunswick, Canada: plant accumulations in the early stages of Terrestrial colonization. Journal of Sedimentary Research, 83, 1202–1215.

Knoll, H. 1984. Character Diversification and Patterns of Evolution in Early Vascular Plants. Paleobiology, 10, 34–47.

Kump, L. 1988. Terrestrial feedback in atmospheric oxygen regulation by fire and phosphorus. Nature, 335, 152–154.

Lenton, T. M., Dahl, T. W., Daines, S. J., Mills, B. J., Ozaki, K., Saltzman, M. R. & Porada, P. 2016. Earliest land plants created modern levels of atmospheric oxygen. Proceedings of the National Academy of Sciences, 113, 9704–9709.

Li, Y. & Schieber, J. 2015. On the origin of a phosphate enriched interval in the Chattanooga Shale (Upper Devonian) of Tennessee-A combined sedimentologic, petrographic, and geochemical study. Sedimentary Geology, 329, 40–61.

Looy, C. V. 2013. Natural history of a plant trait: branch-system abscission in Paleozoic conifers and its environmental, autecological, and ecosystem implications in a fire-prone world. Paleobiology, 39, 235–252.

164

Marynowski, L. & Simoneit, B. R. T. 2009. Widespread Upper Triassic to Lower Jurassic Wildfire Records from Poland: Evidence from Charcoal and Pyrolytic Polycyclic Aromatic Hydrocarbons. Palaios, 24, 785–798.

Meyer-Berthaud, B. & Decombeix, A.-L. 2009. Evolution of earliest trees: The Devonian strategies. Comptes Rendus Palevol, 8, 155–165.

Meyer-Berthaud, B., Scheckler, S. E. & Wendt, J. 1999. Archaeopteris is the earliest known modern tree. Nature, 398, 700–701.

Meyer-Berthaud, B., Soria, A. & Decombeix, A. L. 2010. The land plant cover in the Devonian: a reassessment of the evolution of the tree habit. Geological Society, London, Special Publications, 339, 59–70.

Murchison, D. G. & Raymond, A. C. 1989. Igneous activity and organic maturation in the Midland Valley of Scotland. International Journal of Coal Geology, 14, 47–82.

Niklas, K. J. 1994. Predicting the height of fossil plant remains: an allometric approach to an old problem. American Journal of Botany, 81, 1235–1242.

Osborne, C., Beerling, D., Lomax, B. & Chaloner, W. 2004. Biophysical constraints on the origin of leaves inferred from the fossil record. Proceedings of the National Academy of Sciences, 101, 10360–10362.

Over 2007. Conodont biostratigraphy of the Chattanooga Shale, Middle and Upper Devonian, Southern Appalachian basin, Eastern United States. Journal of Paleontology, 81, 1194– 1217.

Over, D. J. 2002. The Frasnian/Famennian boundary in central and eastern United States. Palaeogeography, Palaeoclimatology, Palaeoecology, 181, 153–169.

Retallack, G. J. & Huang, C. 2011. Ecology and evolution of Devonian trees in New York, USA. Palaeogeography, Palaeoclimatology, Palaeoecology, 299, 110–128.

Rowe, N. P. & Jones, T. P. 2000. Devonian charcoal. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 331–338.

Schieber, J. Developing a sequence stratigraphic framework for the Late Devonian Chattanooga Shale of the southeastern USA: relevance for the Bakken Shale. 1998. Williston Basin Symposium.

Schmeltz, I. & Hoffmann, D. 1976. Formation of polynuclear aromatic hydrocarbons from combustion of organic matter. Carcinogenesis, 1, 225–239.

Scott, A. C. 2000. The Pre-Quaternary history of fire. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 281–329.

Scott, A. C. 2009. ‘Forest fire in the fossil record’, in Cerdà, A. & Robichaud, P. R. (eds.) Fire effects on soils and restoration strategies. Science Publishers, New Hampshire, 177–196.

Scott, A. C. 2010. Charcoal recognition, taphonomy and uses in palaeoenvironmental analysis. Palaeogeography, Palaeoclimatology, Palaeoecology, 291, 11–39.

Scott, A. C., Bowman, D. M., Bond, W. J., Pyne, S. J. & Alexander, M. E. 2014. Fire on earth: an introduction, Wiley-Blackwell, Oxford.

Scott, A. C. & Jones, T. P. 1991. Fossil charcoal: a plant-fossil record preserved by fire. Geology Today, 7, 214–216. Scott, A. C. & Jones, T. P. 1994. The nature and influence of fire in Carboniferous ecosystems. Palaeogeography, Palaeoclimatology, Palaeoecology, 106, 91–112.

Scott, A. C., Kenig, F., Plotnick, R. E., Glasspool, I. J., Chaloner, W. G. & Eble, C. F. 2010. Evidence of multiple late Bashkirian to early Moscovian (Pennsylvanian) fire events preserved in contemporaneous cave fills. Palaeogeography, Palaeoclimatology, Palaeoecology, 291, 72–84.

Shakesby, R. & Doerr, S. 2006. Wildfire as a hydrological and geomorphological agent. Earth-Science Reviews, 74, 269–307.

Shen, S. Z., Crowley, J. L., Wang, Y., Bowring, S. A., Erwin, D. H., Sadler, P. M., Cao, C.- Q., Rothman, D. H., Henderson, C. M. & Ramezani, J. 2011. Calibrating the end- Permian mass extinction. science, 334, 1367–1372.

166

Slik, J. F., Breman, F. C., Bernard, C., Van Beek, M., Cannon, C. H., Eichhorn, K. A. & Sidiyasa, K. 2010. Fire as a selective force in a Bornean tropical everwet forest. Oecologia, 164, 841–849.

Tanner, L. H. & Lucas, S. G. 2016. Stratigraphic distribution and significance of a 15 million–year record of fusain in the Upper Triassic Chinle Group, southwestern USA. Palaeogeography, Palaeoclimatology, Palaeoecology, 461, 261–271.

Taylor, G., Teichmüller, M., Davis, A., Diessel, C., Littke, R. & Robert, P. 1998. Organic petrology, Gebruder Borntraeger, Berlin.

Van Wagtendonk, J. W. 2006. ‘Fire as a physical process’ in Sugihara, N. G., van Wagtendonk, J. W., Fites-Kaufman, J., Shaffer, K. E. & Thode, A. E. (eds) Fire in California’s ecosystems. University of California Press, Berkeley, 58–74.

Xue, J., Huang, P., Wang, D., Xiong, C., Liu, L. & Basinger, J. F. 2018. Silurian-Devonian terrestrial revolution in South China: taxonomy, diversity, and character evolution of vascular plants in a paleogeographically isolated, low-latitude region. Earth-Science Reviews, 180, 92–125.

Yunker, M. B., Macdonald, R. W., Snowdon, L. R. & Fowler, B. R. 2011. Alkane and PAH biomarkers as tracers of terrigenous organic carbon in Arctic Ocean sediments. Organic Geochemistry, 42, 1109–1146.

167

4.8 Appendix III

Figure S4. 1. Photomicrographs (white reflected light, oil immersion) of samples from the Frasnian (top) and Famennian (bottom) interval of the Chattanooga Shale in central Tennessee. V=Vitrinite; I=Inertinite.

168

Figure S4. 2. Concentrations of pyrogenic PAHs and C27 normal alkane (n-C27) across the Frasnian–Famennian interval of the Upper Devonian Chattanooga Shale in central Tennessee.

169

Table S4. 1. Overview of published records of Devonian wildfire records (i.e., fossil charcoal, inertinite maceral and pyrogenic PAHs) from the Euramerica. Stage Paleowildfire Paleo- Formatio Present Location Paleo- Rock Fossil Evidence References indicator continental n environment Type for Vegetation types Region (in situ) Lochkovian fossil Euramerica- Ditton Shropshire, terrestrial siltstone bryophytes and Glasspool et charcoal Avalonia Fm. England, United (wetland- rhyniophytoids al., 2004 Kingdom fluvial) Lochkovian fossil Euramerica- Ditton South Wales, terrestrial sandsto Nematasketum Burgess and charcoal Avalonia Fm., United Kingdom ne diversiforme Edwards, 1988, Ludlow Edwards and Bone Axe, 2004 Bed, Panylan Mudston e Lochkovian pyrogenic northwestern Tadrart Ghadamis Basin, shallow sandsto bryophyte- Romero- PAHs Gondwana Fm. Tunisia marine- ne and derived Sarmiento et continental mudston cryptospores al., 2011 e and tracheophyte- derived trilete spores Lochkovian– fossil Euramerica- Chortkiv Podolia, Ukraine marine carbonat Nematasketum Filipiak and Early Pragian charcoal Baltica and e– spp. Szaniawski, Ivanye silicicla 2016 Fms. stic deposits

170

Late fossil Euramerica- Rhynie Aberdeenshire, terrestrial chert NA Hill et al., 1997 Lochkovian– charcoal Avalonia Chert Scotland, United (hot spring lends in Pragian Bed Kingdom system) shales and sandsto nes Pragian inertinite Euramerica- Val New Brunswick, terrestrial coaly plant debris Kennedy et al., macerals Laurentia d’Amour Canada (wetland- shale 2013 Fm. fluvial) and mudston e Pragian– fossil Euramerica- NA Germany NA gneiss NA Pflug and Emsian charcoal Avalonia Prössl, 1991a, b Emsian inertinite Euramerica- NA Bad terrestrial coal NA Wollenweber et macerals Avalonia Munstereifel, al., 2006 Deutsche Eislauf- Union, Germany Emsian inertinite Euramerica- L'Anse- Tar Point, Gaspe, terrestrial coal NA Glasspool and macerals Laurentia a-Brillant Canada Scott, 2010 Coal, York River Fm. Emsian inertinite Euramerica- Campbell New Brunswick, terrestrial coaly lycopsids, Kennedy et al., macerals Laurentia ton Fm. Canada (wetland) shale trimerophytes, 2013 zosterophyllopsids and 171

rhyniaceans. Eifelian- inertinite Siberia Barzas Kuznetsk, terrestrial coal NA Ammosov, Givetian macerals Coal Barzas, Russia 1964, Volkova, 1994 Givetian inertinite South China Luquan Luquan, Yunnan, terrestrial coal NA Dai et al., 2006 macerals Fm. China Givetian pyrogenic Tarim Hujiesite Hefeng Basin, marine- carbona lycopsids, Song et al., PAHs Fm. Xinjiang, China terrigenous ceous zosterophylls, 2015 alterating mudston fern-like plants settings es and and pro- sandsto gymnosperms nes Late inertinite Euramerica- Mimer Spitsbergen, terrestrial coal NA Diessel, 2010, Givetian– macerals Baltica Valley Norway Wollenweber et Middle Fm. al., 2006 Frasnian Late inertinite Euramerica- Weathera Melville Island, terrestrial coal spores Goodarzi and Givetian– maceral Laurentia ll and Nunavut, Canada (swamp Goodbody, Middle Hecla lakes) 1990 Frasnian Bay Fm. Early– inertinite Euramerica- Beverley Melville Island, terrestrial coal NA Gentzis and Middle maceral Laurentia Inlet Fm. Nunavut, Canada (delta plain) Goodarzi, Frasnian 1991, Goodarzi et al., 1989 Late pyrogenic Euramerica- Sinsin Belgium marine limeston NA Kaiho et al., Frasnian– PAHs Avalonia Section (shelf) e, marl, 2013 Famennian and mudston 172

e Late inertinite Euramerica- Hanover Appalachian marine shale NA Liu et al., 2018 Frasnian– maceral Laurentia and Basin, New York, Famennian Dunkirk USA Fms. Late pyrogenic Euramerica- Woodfor Oklahoma, USA marine shale NA Jones, 2017 Frasnian- PAHs Laurentia d Fm Famennian Frasnian– pyrogneic Gondwana Sadler Canning Basin, marine mudston NA Tulipani et al., Famennian PAHs and Gogo Australia e/shale 2015 Fms. Frasnian– inertinite Baltica NA Timan region, terrestrial coal NA Volkova, 1994 Famennian maceral Russia Famennian pyrogenic Gondwana Aouinet Ghadames Basin, marine shale tracheophyte- Riboulleau et PAHs Ouenine Libya derived trilete al., 2018 spores Early–the inertinite Euramerica- New Kentucky, USA marine shale NA Rimmer et al., Latest maceral Laurentia Albany 2004 Famennian Shale Famennian inertinite Euramerica- Ohio Ohio, USA marine shale NA Rimmer et al., maceral Laurentia Shale 2015 Famennian inertinite Euramerica- Hampshi West Virginia, marine shale NA Rimmer et al., macerals Laurentia re Fm USA 2015 Famennian fossil Euramerica- Duncann Red Hill, terrestrial sandsto Archaeopteris, Cressler, 2006 charcoal Laurentia on Mbr, Pennsylvania, (floodplain ne, Rhacophyton, Catskill USA and delta) siltstone lycopsids, Fm. and Gillespiea and mudston cupulate and 173

e acupulate gymnosperms late inertinite Euramerica- Tunheim Bear Island, terrestrial coal and NA Glasspool et Famennian Baltica Mbr., Bjornoya, (fluvial) coaly al., 2015, Roedvika Norway shales Michelsen, Fm. interbed 1991 ded with sandsto ne and mudston e late pyrogenic Euramerica- Hangen- Holy Cross marine shale NA Marynowski Famennian PAHs and Avalonia berg Mountain, Poland and Filipiak, fossil black 2007 charcoal shale late fossil Euramerica- Hangen- Oese, Sauerland, marine sandsto Archaeopteris Rowe and Famennian charcoal Avalonia berg Germany (offshore) ne and lycopsid Jones, 2000 Sand- microphylls and stone sporophylls late fossil Euramerica- Knoppen Bergisch terrestrial siltstone early ferns and Fairon-Demaret Famennian charcoal Avalonia bissen Gladbach- gymnosperms and Hartkopf- Fm. PafTrath Fröder, 2004 Syncline, Rhenish Massif, Germany late fossil Euramerica- Evieux three quarries, terrestrial sandsto remains Prestianni et Famennian charcoal Avalonia Fm., Belgium (lagoon- ne attributed to al., 2010 174

Condroz fluvial) Callixylon Gp. Zallessky, leaf remains of Sphenocyclopter idium belgicum, Sphenopteris mourlonii, S. modavensis; sporangial remains of Rhcophyton late pyrogenic Euramerica- Sapolno Pomerannian marine shale NA Matyja et al., Famennian PAHs Avalonia Calcareo Basin, Poland 2015 us Shale

175

Table S4. 2. Maximum observed aerial axis diameter of vascular plant fossils in Euramerica through the Devonian. Stage Higher taxa Genus Species Maximum References axial width (cm) Lochkovian Paratracheophytes- Concavatheca banksii 0.18 Habgood et al., 2002 Cooksonioides Lochkovian Paratracheophytes- Cooksonia caledonica 0.18 Edwards, 1970a Cooksonioides Lochkovian Paratracheophytes- Cooksonia hemisphaerica 0.15 Edwards, 1979 Cooksonioides Lochkovian Paratracheophytes- Cooksonia pertoni 0.099 Lang, 1937 Cooksonioides Lochkovian Paratracheophytes- Cooksonia cambrensis 0.091 Edwards, 1979 Cooksonioides Lochkovian Paratracheophytes- Tortilicaulis offaeus 0.01 Edwards et al., 1994 Rhyniophytian Lochkovian Paratracheophytes- Tortilicaulis transwalliensis 0.04 Fanning et al., 1992 Rhyniophytian Lochkovian Paratracheophytes- Uskiella reticulata 0.05 Fanning et al., 1992 Rhyniophytian Lochkovian Paratracheophytes- Tarrantia salopensis 0.09 Shute and Edwards, Rhyniophytian 1989 Lochkovian Paratracheophytes- Monnowella bennettii 0.15 Morris and Edwards, Rhyniophytian 2014 Lochkovian Paratracheophytes- Salopella allenii 0.2 Edwards and Rhyniophytian Richardson, 1974a Lochkovian Paratracheophytes- Taeniocrada ? spitzbergensis 0.2 Høeg, 1942 Rhyniophytian 176

Lochkovian Paratracheophytes- Taeniocrada decheniana 1.5 Taylor, 1986 Rhyniophytian Lochkovian Psilophytophyta Psilophytites gileppensis 0.38 Gerrienne, 1992 Lochkovian Psilophytophyta Psilophyton princeps 1 Edwards et al., 1989, Hueber and Grierson, 1961 Lochkovian Lycopodiopsida Sengelia radicans 3.5 Matsunaga and Tomescu, 2017 Lochkovian stem Lycopodiopsida Zosterophyllum fertile 0.15 Leclercq, 1942 Lochkovian stem Lycopodiopsida Craswallia haegensis 0.18 Morris and Edwards, 2014 Lochkovian stem Lycopodiopsida Zosterophyllum rhenanum 0.2 Krausel and Weyland, 1932 Lochkovian stem Lycopodiopsida Zosterophyllum myretonianum 0.3 Edwards, 1975 Lochkovian stem Lycopodiopsida Gosslingia breconensis 0.4 Edwards, 1970b Pragian Paratracheophytes- Cooksonia hemisphaerica 0.15 Edwards, 1979, Lang, Cooksonioides 1937 Pragian Paratracheophytes- Rhynia gwynne- 0.3 Edwards, 1980b Rhyniophytian vaughanii Pragian Paratracheophytes- Uskiella spargens 0.31 Shute and Edwards, Rhyniophytian 1989 Pragian Paratracheophytes- Sennicaulis hippocrepiformis 0.6 Edwards, 1981 Rhyniophytian Pragian Paratracheophytes- Stockmansella langii 0.7 Fairon-Demaret, Rhyniophytian Taeniocrada 1985, Fairon- Demaret, 1986b Pragian Paratracheophytes- Huvenia kleui 1.5 Hass and Remy, 1991 Rhyniophytian 177

Pragian Paratracheophytes- Taeniocrada decheniana 1.5 Höeg, 1937 Rhyniophytian Pragian Paratracheophytes- Taeniocrada dubia 2 Hotton et al., 2001 Rhyniophytian Pragian basal Euphyllophyte Armoricaphyton chateaupannense 0.25 Gerrienne and Gensel, 2016 Pragian Pteridophyta/Monilophyta Hostinella strictissima 0.2 Høeg, 1942 Pragian Pteridophyta/Monilophyta Hostinella heardii 0.3 Edwards, 1980a Pragian Pteridophyta/Monilophyta Hostinella racheneuri 0.3 Ledoux-Marcelle, 1927 Pragian Pteridophyta/Monilophyta Estinnophyton wahnbachense 1.1 Fairon-Demaret, 1979 Pragian Psilophytophyta Psilophyton dapsile 0.2 Boyer and Stein, 2008, Kasper et al., 1974 Pragian Psilophytophyta Sartilmania jabachensis 0.2 Fairon-Demaret, 1986a Pragian Psilophytophyta Dawsonites arcuatus 0.3 Halle, 1916 Pragian Psilophytophyta Psilophytites gileppensis 0.38 Gerrienne, 1992 Pragian Psilophytophyta Psilophyton princeps 1 Edwards et al., 1989, Hueber and Grierson, 1961 Pragian Psilophytophyta Aphyllopteris robusta 1 Brauer, 1980 Pragian Psilophytophyta Psilophyton sp. 0.3 Høeg, 1942 Pragian Lycopodiopsida Asteroxylon mackiei 1 Kidston and Lang, 1920 Pragian Lycopodiopsida Drepanophycus gaspianus 1 Stockmans, 1939 Pragian Lycopodiopsida Protolepidodendron wahnbachense 1.1 Fairon-Demaret, 178

=Estinnophyton 1979 Pragian Lycopodiopsida Sengelia radicans 3.5 Matsunaga and Tomescu, 2017 Pragian Lycopodiopsida Leclercqia andrewsii 3.5 Gensel and Kasper, 2005 Pragian Lycopodiopsida Drepanophycus spinaeformis 4.2 Fairon-Demaret, 1978b, Gensel and Kasper, 2005, Li et al., 2000, Rayner, 1984, Stubblefield and Banks, 1978 Pragian stem Lycopodiopsida Distichophytum sp. 0.12 Kotyk et al., 2002 Pragian stem Lycopodiopsida Zosterophyllum artesianum 0.15 Danzé-Corsin, 1956, Edwards, 2006 Pragian stem Lycopodiopsida Danziella artesiana 0.15 Edwards, 2006 Pragian stem Lycopodiopsida Zosterophyllum arcticum 0.15 Høeg, 1942 Pragian stem Lycopodiopsida Zosterophyllum fertile 0.15 Leclercq, 1942 Pragian stem Lycopodiopsida Zosterophyllum llanoveranum 0.15 Croft and Lang, 1942, Edwards, 1969a Pragian stem Lycopodiopsida Distichophytum mucronatum 0.2 Kotyk et al., 2002, Mägdefrau, 1938 Pragian stem Lycopodiopsida Zosterophyllum? australianum 0.22 Croft and Lang, 1942, Edwards, 1969b Pragian stem Lycopodiopsida Nothia aphylla 0.25 Kerp et al., 2001 Pragian stem Lycopodiopsida Trichopherophyton teuchansii 0.25 Lyon and Edwards, 1991, Taylor et al., 2009 Pragian stem Lycopodiopsida Yarravia minor 0.25 Danze-Corsin, 1956 179

Pragian stem Lycopodiopsida Zosterophyllum myretonianum 0.3 Edwards, 1975 Pragian stem Lycopodiopsida Gosslingia breconensis 0.4 Edwards, 1970b, Gerrienne, 1990 Pragian stem Lycopodiopsida Krithodeophyton croftii 0.43 Edwards, 1968 Pragian stem Lycopodiopsida Sawdonia ornata 0.45 Hueber, 1971, Rayner, 1983 Pragian stem Lycopodiopsida Ventarura lyonii 0.72 Powell et al., 1999 Pragian stem Lycopodiopsida Thrinkophyton formosum 0.82 Kenrick and Edwards, 1988 Pragian stem Lycopodiopsida Deheubarthia splendens 1.1 Edwards et al., 1989 Pragian stem Lycopodiopsida Bathurstia denticulata 1.2 Kotyk and Basinger, 2000 Pragian stem Lycopodiopsida Zosterophyllum deciduum 1.25 Gerrienne, 1988 Emsian Paratracheophytes- Cooksonia hemisphaerica 0.15 Edwards, 1979, Lang, Cooksonioides 1937 Emsian Paratracheophytes- Eogaspesiea gracilis 0.05 Daber, 1960 Rhyniophytian Emsian Paratracheophytes- Eddianna gaspiana 0.2 Pfeiler and Tomescu, Rhyniophytian 2018 Emsian Paratracheophytes- Rhynia gwynne- 0.3 Edwards, 1980b Rhyniophytian vaughanii Emsian Paratracheophytes- Huvenia sp. 0.5 Hotton et al., 2001 Rhyniophytian Emsian Paratracheophytes- Stockmansella langii 0.7 Fairon-Demaret, Rhyniophytian Taeniocrada 1985, Fairon- Demaret, 1986b Emsian Paratracheophytes- Huvenia elongata 0.9 Schultka, 1991, Rhyniophytian Pfeiler and Tomescu, 180

2018 Emsian Paratracheophytes- Taeniocrada decheniana 1.5 Höeg 1937 Rhyniophytian Emsian Paratracheophytes- Taeniocrada dubia 2 Hotton et al., 2001 Rhyniophytian Emsian basal Euphyllophyte Armoricaphyton chateaupannense 0.25 Gerrienne and Gensel, 2016 Emsian basal Euphyllophyte Franhueberia gerriennei 0.19 Hoffman and Tomescu, 2013 Emsian Lignophytes Chaleuria cirrosa 1 Andrews et al., 1974 Emsian Lignophytes Oocampsa catheta 1.6 Andrews et al., 1975 Emsian Pteridophyta/Monilophyta Tursuidea paniculata 0.08 Schweitzer, 1987 Emsian Pteridophyta/Monilophyta Hostinella strictissima 0.2 Høeg, 1942 Emsian Pteridophyta/Monilophyta Foozia minuta 0.4 Gerrienne 1992 Emsian Pteridophyta/Monilophyta Hyenia ramosa 0.5 Høeg, 1945 Emsian Pteridophyta/Monilophyta Kenricrana bivena 0.761 Toledo, 2018 Emsian Pteridophyta/Monilophyta Pertica dalhousii 1 Doran et al., 1978 Emsian Pteridophyta/Monilophyta Pertica varia 1.4 Granoff et al., 1976 Emsian Pteridophyta/Monilophyta Pertica quadrifaria 1.5 Kasper and Andrews, 1972 Emsian Pteridophyta/Monilophyta Estinnophyton gracile 4.5 Fairon-Demaret, 1978a Emsian Psilophytophyta Psilophyton parvulum 0.11 Gerrienne, 1995 Emsian Psilophytophyta Psilophyton dapsile 0.2 Boyer and Stein, 2008, Kasper et al., 1974 Emsian Psilophytophyta Psilophyton genseliae 0.2 Gerrienne, 1997 Emsian Psilophytophyta Sartilmania jabachensis 0.2 Fairon-Demaret, 181

1986a Emsian Psilophytophyta Dawsonites arcuatus 0.3 Halle, 1916 Emsian Psilophytophyta Psilophyton crenulatum 0.35 Doran, 1980 Emsian Psilophytophyta Psilophyton microspinosum 0.4 Kasper et al., 1974 Emsian Psilophytophyta Psilophyton wyomingense 0.5 Dorf, 1933 Emsian Psilophytophyta Psilophyton szaferi 0.6 Zdebska, 1986 Emsian Psilophytophyta Psilophyton goldschmidtii 0.7 Halle, 1916 Margophyton Emsian Psilophytophyta Psilophyton rectissimum 0.7 Høeg, 1945 Emsian Psilophytophyta Psilophyton arcuatum 0.8 Schweitzer, 1980 Emsian Psilophytophyta Psilophyton forbesii 0.9 Andrews et al., 1968 Emsian Psilophytophyta Psilophyton princeps 1 Edwards et al., 1989, Hueber and Grierson, 1961 Emsian Psilophytophyta Aphyllopteris robusta 1 Brauer, 1980 Emsian Psilophytophyta Bitelaria dubjanski 1.2 Johnson, 1992 Emsian Psilophytophyta Psilophyton dawsonii 3.5 Banks et al., 1975 Emsian Lycopodiopsida Leclercqia complexa 0.7 Gensel and Albright, 2006 Emsian Lycopodiopsida Kaulangiophyton akantha 0.9 Gensel et al., 1969 Emsian Lycopodiopsida Asteroxylon mackiei 1 Kidston and Lang, 1920 Emsian Lycopodiopsida Protolepidodendron wahnbachense 1.1 Fairon-Demaret, =Estinnophyton 1979 Emsian Lycopodiopsida Sengelia minor 1.41 Xu et al., 2013 Emsian Lycopodiopsida Sengelia devonica 2 Matsunaga and Tomescu, 2017, Schweitzer and 182

Giesen, 1980 Emsian Lycopodiopsida Sugambrophyton pilgeri 2.5 Schmidt, 1954 Emsian Lycopodiopsida Drepanophycus gaspianus 3 Stockmans, 1939 Emsian Lycopodiopsida Baragwanathia abitibiensis 3.2 Hueber, 1983 Emsian Lycopodiopsida Leclercqia andrewsii 3.5 Gensel and Kasper, 2005 Emsian Lycopodiopsida Drepanophycus spinaeformis 4.2 Fairon-Demaret, 1978b, Gensel and Kasper, 2005, Li et al., 2000, Rayner, 1984, Stubblefield and Banks, 1978 Emsian stem Lycopodiopsida Zosterophyllum artesianum 0.12 Danzé-Corsin, 1956 Emsian stem Lycopodiopsida Deuterophyton stockmansii 0.14 Gerrienne, 1998 Emsian stem Lycopodiopsida Renalia hueberi 0.15 Gensel, 1976 Emsian stem Lycopodiopsida Danziella artesiana 0.15 Edwards, 2006 Emsian stem Lycopodiopsida Zosterophyllum fertile 0.15 Leclercq, 1942 Emsian stem Lycopodiopsida Gosferia curvata 0.17 Gerrienne, 1999 Emsian stem Lycopodiopsida Distichophytum mucronatum 0.2 Kotyk et al., 2002, Mägdefrau, 1938 Emsian stem Lycopodiopsida Rebuchia or ovata 0.2 Hueber, 1970, Tanner, Distichophytum? 1984 Emsian stem Lycopodiopsida Konioria andrychoviensis 0.25 Zdebska, 1982 Emsian stem Lycopodiopsida Nothia aphylla 0.25 Kerp et al., 2001 Emsian stem Lycopodiopsida Trichopherophyton teuchansii 0.25 Lyon and Edwards, 1991, Taylor et al., 2009 Emsian stem Lycopodiopsida Yarravia minor 0.25 Danzé-Corsin, 1956 183

Emsian stem Lycopodiopsida Zosterophyllum spectabile 0.25 Schweitzer, 1979 Emsian stem Lycopodiopsida Odonax borealis 0.27 Gerrienne, 1996 Emsian stem Lycopodiopsida Crenaticaulis verruculosus 0.3 Banks and Davis, 1969 Emsian stem Lycopodiopsida Zosterophyllum myretonianum 0.3 Edwards, 1975 Emsian stem Lycopodiopsida Zosterophyllum sp. 0.3 Edwards and Richardson, 1974b Emsian stem Lycopodiopsida Anisophyton gothani 0.38 Xue, 2012 Emsian stem Lycopodiopsida Zosterophyllum divaricatum 0.4 Gensel, 1982a Emsian stem Lycopodiopsida Sawdonia ornata 0.45 Hueber, 1971, Rayner, 1983 Emsian stem Lycopodiopsida Rebuchia or mucronata 0.47 Hueber, 1970, Tanner, Distichophytum 1984 Emsian stem Lycopodiopsida Oricilla bilinearis 0.5 Gensel, 1982b Emsian stem Lycopodiopsida Sawdonia spinosissima 0.5 Schweitzer, 1982 Emsian stem Lycopodiopsida Ventarura lyonii 0.72 Powell et al., 1999 Emsian stem Lycopodiopsida Sawdonia acanthotheca 0.8 Gensel et al., 1975 Emsian stem Lycopodiopsida Forania plegiospinosa 1 Jensen and Gensel, 2013 Emsian stem Lycopodiopsida Zosterophyllum deciduum 1.25 Gerrienne, 1988 Emsian stem Lycopodiopsida Renalia graberti 6.5 Schweitzer, 1980 Eifelian Paratracheophytes- Cooksonia hemisphaerica 0.15 Edwards, 1979, Lang, Cooksonioides 1937 Eifelian Paratracheophytes- Stockmansella remyi 0.27 Schultka and Hass, Rhyniophytian Taeniocrada 1997 Eifelian Paratracheophytes- Taeniocrada dubia 2 Hotton et al., 2001 Rhyniophytian Eifelian Lignophytes Aneurophyton germanicum 0.5 Banks, 1944, 184

Grierson and Banks, 1963, Serlin and Banks, 1978, Stein et al., 2007 Eifelian Lignophytes Aneurophyton furcatum 0.6 Berry, 2009 Eifelian Pteridophyta/Monilophyta Hyenia sphenophylloides 0.04 Høeg, 1945, Høeg, 1935 Eifelian Pteridophyta/Monilophyta Hostinella crispa 0.1 Arnold, 1939 Eifelian Pteridophyta/Monilophyta Hyenia? rhizoides 0.4 Bonamo and Banks, 1966a Eifelian Pteridophyta/Monilophyta Pertica quadrifaria 1.5 Kasper and Andrews, 1972 Eifelian Pteridophyta/Monilophyta Calamophyton primaevum 2 Leclercq, 1969 Eifelian Pteridophyta/Monilophyta Estinnophyton gracile 4.5 Fairon-Demaret, 1978a Eifelian Pteridophyta/Monilophyta Duisbergia mirabilis 30 Giesen and Berry, 2013 Eifelian Psilophytophyta Dawsonites ellenae 0.1 Höeg, 1937, Høeg, =Psilophyton 1935 Eifelian Psilophytophyta Psilophyton dapsile 0.2 Boyer and Stein, 2008, Kasper et al., 1974 Eifelian Psilophytophyta Psilophyton microspinosum 0.4 Kasper et al., 1974 Eifelian Psilophytophyta Psilophyton kraeuselii 0.5 Obrhel, 1959 Eifelian Psilophytophyta Psilophyton forbesii 0.9 Andrews et al., 1968 Eifelian Psilophytophyta Psilophyton princeps 1 Edwards et al., 1989, Hueber and Grierson, 1961 185

Eifelian Lycopodiopsida Asteroxylon elberfeldense 0.05 Kräusel and Weyland, 1926 Eifelian Lycopodiopsida Leclercqia complexa 0.7 Gensel and Albright, 2006 Eifelian Lycopodiopsida Kaulangiophyton akantha 0.9 Gensel et al., 1969 Eifelian Lycopodiopsida Colpodexylon trifurcatum 2.5 Banks, 1944 Eifelian Lycopodiopsida Leclercqia andrewsii 3.5 Gensel and Kasper, 2005 Eifelian Lycopodiopsida Drepanophycus gaspianus 5 Stockmans, 1939 Eifelian stem Lycopodiopsida Hicklingia edwardii 0.14 Edwards, 1976 Eifelian stem Lycopodiopsida Eocladoxylon thomsoni 2.5 Leclercq and Protopteridium Bonamo, 1973 Givetian Paratracheophytes- Cooksonia hemisphaerica 0.15 LEdwards, 1979, Cooksonioides Lang, 1937 Givetian Lignophytes Actinopodium nathorstii 0.2 Høeg, 1942 Givetian Lignophytes Actinopodium banksii 0.25 Matten, 1968 Givetian Lignophytes Aneurophyton germanicum 0.5 Schweitzer and LC, 1982 Givetian Lignophytes Cairoa lamanekii 0.8 Matten, 1973 Givetian Lignophytes Reimannia aldenense 0.95 Arnold, 1935, Stein1982Arnold, 1935, Stein, 1982 Givetian Lignophytes Aneurophyton bohemicum 1 Krausel and Weyland, 1923 Givetian Lignophytes Svalbardia avelinesiana 1 Stockmans, 1968 Givetian Lignophytes Svalbardia polymorpha 1.5 Høeg, 1942 Givetian Lignophytes Triloboxylon hallii 2 Scheckler and Banks, Aneurophyton 1971 186

Givetian Lignophytes Rellimia thomsonii 2.5 Bonamo, 1977, Leclercq and Bonamo, 1973 Givetian Lignophytes Callixylon trifilievi 60 Snigirevskaya, 1984 in Retallack 2001Retallack and Huang, 2011, Snigirevskaya, 1984 Givetian Pteridophyta/Monilophyta Hyenia sphenophylloides 0.04 Høeg, 1945, Høeg, 1935 Givetian Pteridophyta/Monilophyta Wattieza givetiana 0.1 Berry, 2009, Berry, 2000 Givetian Pteridophyta/Monilophyta Hostinella crispa 0.1 Arnold, 1939 Givetian Pteridophyta/Monilophyta Hyenia banksii 0.4 Arnold, 1941 Givetian Pteridophyta/Monilophyta Hyenia? rhizoides 0.4 Bonamo and Banks, 1966Bonamo and Banks, 1966a Givetian Pteridophyta/Monilophyta Sphenopteris brabantica 0.4 Berry, 2009 Givetian Pteridophyta/Monilophyta Hyenia vogtii 0.9 Berry, 2005 Givetian Pteridophyta/Monilophyta Avelinesia antiqua 1 Berry, 2009 Givetian Pteridophyta/Monilophyta Brabantophyton runcariense 1.45 Gerrienne and Meyer- Berthaud, 2006, Momont et al., 2016 Givetian Pteridophyta/Monilophyta Cladoxylon hueberi 1.5 Stein and Hueber, Pseudosporochnus 1989 Givetian Pteridophyta/Monilophyta Lorophyton goense 1.5 Fairon-Demaret and Li, 1993 Givetian Pteridophyta/Monilophyta Calamophyton bicephalum 2 Bonamo and Banks, 187

1966a, Leclercq and Andrews, 1960 Givetian Pteridophyta/Monilophyta Calamophyton primaevum 2 Leclercq, 1969 Givetian Pteridophyta/Monilophyta Xenocladia medullosina 1 Arnold, 1952 Givetian Pteridophyta/Monilophyta Protolepidodendropsis pulchra 1 Berry and Marshall, 2015, Schweitzer, 1965 Givetian Pteridophyta/Monilophyta Pseudosporochnus chlupaci 13 Obrhel, 1959 Givetian Pteridophyta/Monilophyta Cephalopteris? praecox 30 Schweitzer, 1968 =Protocephalopteris Givetian Pteridophyta/Monilophyta Eospermatopteris erianus 106.7 Boyer, 1995 Givetian Psilophytophyta Dawsonites ellenae 0.1 Höeg, 1937, Høeg, =Psilophyton 1935 Givetian Psilophytophyta Psilophyton grande origianlly 0.1 Penhallow, 1893 grandis Givetian Psilophytophyta Dawsonites roskiliensis 0.2 Chaloner, 1972 Givetian Psilophytophyta Psilophyton kraeuselii 0.5 Obrhel, 1959 Givetian Psilophytophyta Psilophyton arcticum 0.8 Høeg, 1942 Givetian Psilophytophyta Ibyka amphikoma 1.5 Skog and Banks, 1973 Givetian Lycopodiopsida Asteroxylon elberfeldense 0.05 Kräusel and Weyland, 1926 Givetian Lycopodiopsida Leclercqia scolopendra 0.39 Benca et al., 2014 Givetian Lycopodiopsida Protolepidodendron gilboense 0.45 Grierson and Banks, 1963 Givetian Lycopodiopsida Actinoxylon banksii 0.5 Matten, 1968 Givetian Lycopodiopsida Leclercqia complexa 0.7 Gensel and Albright, 2006 188

Givetian Lycopodiopsida Protolepidodendron scharianum 0.75 Fairon-Demaret, 1980, Jurina, 2009 Givetian Lycopodiopsida Drepanophycus colophyllus 0.914 Grierson and Banks, 1983 Givetian Lycopodiopsida Clwydia formerly vanuxemi 1.2 Fairon-Demaret and Archaeosigillaria Banks, 1978 Givetian Lycopodiopsida Gilboaphyton oldringiae 1.2 Arnold, 1937 formally goldringiae Givetian Lycopodiopsida Gilboaphyton griersonii 1.45 Berry and Edwards, 1997 Givetian Lycopodiopsida Colpodexylon deatsii 1.5 Banks, 1944 Givetian Lycopodiopsida Protolepidodendron primaevum 6 Leclercq, 1969 Givetian stem Lycopodiopsida Zosterophyllum bohemicum 0.45 Obrhel, 1959 Givetian stem Lycopodiopsida Barinophyton norvegicum 0.5 Schweitzer and Pectinophyton Giesen, 2008 Givetian stem Lycopodiopsida Enigmophyton superbum 0.5 Høeg, 1942 Givetian stem Lycopodiopsida Serrulacaulis furcatus 1.2 Berry and Edwards, 1994, Hueber and Banks, 1979 Givetian stem Lycopodiopsida Eocladoxylon thomsoni 2.5 Leclercq and Protopteridium Bonamo, 1973 Frasnian Paratracheophytes- Taeniocrada lesquereuxii 0.45 Taylor, 1986 Rhyniophytian Frasnian Paratracheophytes- Taeniocrada stilesvillensis 0.5 Taylor, 1986 Rhyniophytian Frasnian Lignophytes Svalbardia banksii 0.25 Matten, 1981 Frasnian Lignophytes Stenokoleos bifidus 0.5 Matten and Banks, 189

1969 Frasnian Lignophytes Aneurophyton germanicum 0.5 Schweitzer and Matten, 1982 Frasnian Lignophytes Triloboxylon ashlandicum 1.3 Matten and Banks, 1966 Frasnian Lignophytes Triloboxylon hallii 2 Scheckler and Banks, Aneurophyton 1971 Frasnian Lignophytes Cordaites sp. 2.5 Retallack, 2011 Frasnian Lignophytes Callixylon trifilievi 150 Taylor et al., 2009 Frasnian Lignophytes Archaeopteris fissilis 150 Taylor et al., 2009 Frasnian Lignophytes Archaeopteris gaspiensis 150 Taylor et al., 2009 Frasnian Lignophytes Archaeopteris halliana 150 Taylor et al., 2009 Frasnian Lignophytes Archaeopteris hibernica 150 Taylor et al., 2009 Frasnian Lignophytes Archaeopteris jacksoni 150 Taylor et al., 2009 Frasnian Lignophytes Archaeopteris magnacensis 150 Taylor et al., 2009 Frasnian Lignophytes Archaeopteris obtusa 150 Taylor et al., 2009 Frasnian Lignophytes Archaeopteris sphenophyllifolia 150 Taylor et al., 2009 Frasnian Lignophytes Callixylon newberryi 150 Taylor et al., 2009 Frasnian Lignophytes Callixylon petryi 150 Taylor et al., 2009 Frasnian Lignophytes Callixylon schmidtii 150 Taylor et al., 2009 Frasnian Lignophytes Callixylon zalesskyi 150 Taylor et al., 2009 Frasnian Pteridophyta/Monilophyta Ellesmeris sphenopteroides 0.9 Hill et al., 1997 Frasnian Pteridophyta/Monilophyta Avelinesia antiqua 1 Berry, 2009 Frasnian Pteridophyta/Monilophyta Brabantophyton runcariense 1.45 Gerrienne and Meyer- Berthaud, 2006, Momont et al., 2016 Frasnian Pteridophyta/Monilophyta Cladoxylon Rotoxylon dawsoni 1.5 Cordi and Stein, 2005, Read, 1935 190

Frasnian Pteridophyta/Monilophyta cf.Calamites sp. 1.5 Montero and Dieguez, 2010 Frasnian Pteridophyta/Monilophyta Calamophyton primaevum 2 Leclercq, 1969 Frasnian Pteridophyta/Monilophyta Cephalopteris mirabilis 2 Schweitzer, 2006 Frasnian Pteridophyta/Monilophyta Rhacophyton condrusorum 2 Andrews and Phillips, 1968, Crépin, 1875 Frasnian Pteridophyta/Monilophyta Rhacophyton zygopteroides 2 Leclercq, 1951 Frasnian Pteridophyta/Monilophyta Prosseria grandis 2.5 Read, 1953 Frasnian Pteridophyta/Monilophyta Rhymokalon trichium 2.6 Scheckler, 1975b Frasnian Pteridophyta/Monilophyta Protolepidodendropsis pulchra 1 Berry and Marshall, 2015, Schweitzer, 1965 Frasnian Pteridophyta/Monilophyta Pseudobornia ursina 60 Schweitzer, 1967, 2006 Frasnian Pteridophyta/Monilophyta Eospermatopteris erianus 106.7 Boyer, 1995 Frasnian Psilophytophyta Aphyllopteris delawarensis 0.2 Arnold, 1939 Frasnian Lycopodiopsida Clwydia formerly vanuxemi 1.2 Fairon-Demaret and Archaeosigillaria Banks, 1978 Frasnian Lycopodiopsida Phytokneme rhodona 3 Andrews et al., 1971, Matten, 1989 Frasnian Lycopodiopsida Sublepidodendron antecedens 7 Gothan and Zimmermann, 1936 Frasnian Lycopodiopsida Sublepidodendron mirabile 7 Wang et al., 2003 Frasnian Lycopodiopsida Lepidosigillaria whitei 38.5 Berry and Marshall, 2015, Grierson and Banks, 1963, Pigg, 2001 Frasnian stem Lycopodiopsida Serrulacaulis furcatus 1.2 Berry and Edwards, 191

1994, Hueber and Banks, 1979 Famennian Lignophytes Sphenopteridium keilhauii 0.25 Nathorst, 1902 Famennian Lignophytes Stenokoleos simplex 0.35 Beck, 1960 Famennian Lignophytes Moresnetia zalesskyi 0.5 Fairon-Demaret and Scheckler, 1987 Famennian Lignophytes Elkinsia polymorpha 0.9 Serbet and Rothwell, 1992 Famennian Lignophytes Reimannia indianensis 2.5 Read and Campbell, 1939 Famennian Lignophytes Aporoxylon primigenium 5 Decombeix et al., 2005 Famennian Lignophytes Archaeopteris fimbriata 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris gaspiensis 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris halliana 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris hibernica 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris hitchcockii 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris intermedia 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris jacksoni 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris latifolia 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris macilenta 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris magnacensis 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris minor 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris obtusa 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris roemeriana 150 Taylor et al., 2009 Famennian Lignophytes Archaeopteris sphenophyllifolia 150 Taylor et al., 2009 Famennian Lignophytes Callixylon arnordii 150 Taylor et al., 2009 Famennian Lignophytes Callixylon beckii 150 Taylor et al., 2009 192

Famennian Lignophytes Callixylon brownii 150 Taylor et al., 2009 Famennian Lignophytes Callixylon clevelandensis 150 Taylor et al., 2009 Famennian Lignophytes Callixylon erianum 150 Taylor et al., 2009 Famennian Lignophytes Callixylon henkei 150 Taylor et al., 2009 Famennian Lignophytes Callixylon huronense 150 Taylor et al., 2009 =huronensis Famennian Lignophytes Callixylon newberryi 150 Taylor et al., 2009 Famennian Lignophytes Callixylon oweni =owenii 150 Taylor et al., 2009 Famennian Lignophytes Callixylon petryi 150 Taylor et al., 2009 Famennian Lignophytes Callixylon schmidtii 150 Taylor et al., 2009 Famennian Lignophytes Callixylon whiteanum 150 Taylor et al., 2009 Famennian Pteridophyta/Monilophyta Gillespiea randolphensis 0.11 Erwin and Rothwell, 1989 Famennian Pteridophyta/Monilophyta Eviostachya hoegii 0.3 Stockmans, 1948, Wang, 1993 Famennian Pteridophyta/Monilophyta Sphenophyllum subtenerrimum 0.4 Nathorst, 1902, =tenerrumum Schweitzer, 2006 Schweitzer 2nana6 Famennian Pteridophyta/Monilophyta Polyxylon elegans 0.8 Read and Camp, 1939 Famennian Pteridophyta/Monilophyta cf.Calamites sp. 1.5 Montero and Dieguez, 2010 Famennian Pteridophyta/Monilophyta Cephalopteris mirabilis 2 Schweitzer, 2006 Famennian Pteridophyta/Monilophyta Rhacophyton ceratangium 2 Dittrich et al., 1983 Famennian Pteridophyta/Monilophyta Rhacophyton condrusorum 2 Andrews and Phillips, 1968, Crépin, 1875 Famennian Pteridophyta/Monilophyta Rhacophyton incertum 2 Andrews and Phillips, 1968 193

Famennian Pteridophyta/Monilophyta Rhacophyton mirabilis 2 Schweitzer, 2006 Famennian Pteridophyta/Monilophyta Rhacophyton zygopteroides 2 Leclercq, 1951 Famennian Pteridophyta/Monilophyta Pietzschia polyupsilon 2.5 Read and Camp, 1939; Soria and Meyer-Berthaud, 2003 Famennian Pteridophyta/Monilophyta Melvillipteris quadriseriata 6 Xue and Basinger, 2016 Famennian Pteridophyta/Monilophyta Pseudobornia ursina 60 Schweitzer, Schweitzer, 1967 Famennian Lycopodiopsida Asteroxylon setchelli 0.5 Read and Campbell, 1939 Famennian Lycopodiopsida Protolepidodendron microphyllum 0.8 Read and Campbell, 1939 Famennian Lycopodiopsida Lycopogenia callicyrta 2 Read, 1936 Famennian Lycopodiopsida Phytokneme rhodona 3 Andrews et al., 1971, Matten, 1989 Famennian Lycopodiopsida Clevelandodendron ohioensis 3 Chitaley and Pigg, 1996 Famennian Lycopodiopsida Wexfordia hookense 3 Matten, 1989; Klavins, 2004 Famennian Lycopodiopsida Sublepidodendron antecedens 7 Wang et al., 2003 Famennian Lycopodiopsida Sublepidodendron mirabile 7 Wang et al., 2003 Famennian Lycopodiopsida Otzinachsonia beerboweri 10.3 Cressler and Pfefferkorn, 2005 Famennian Lycopodiopsida Cyclostigma/Jurinoden kiltorkense 15 Stockmans, 1948, dron Wang, 1993 Famennian Lycopodiopsida Leptophloeum rhombicum 40 Wang et al., 2005 194

Famennian stem Lycopodiopsida Barinophyton citrulliforme 0.85 Brauer, 1980

195

Table S4. 3. Devonian vascular plant fossils from the Euramerica examined for leaf-size analysis. Stage Higher taxa genus species simpl simpl Shape Reference e leaf e leaf length width cm cm Lochkovian Lycopodiopsida Sengelia radicans 0.7 - Planar, fan-/wedge- Matsunaga and shape Tomescu, 2017 Pragian Lycopodiopsida Asteroxylon mackiei 0.78 0.44 Planar, linear/strap- Kidston and like Lang, 1920 Pragian Lycopodiopsida Drepanophycus spinaeformis 2 0.3 Not planar Fairon- Demaret, 1978b, Gensel and Kasper, 2005, Li et al., 2000, Rayner, 1984, Stubblefield and Banks, 1978 Pragian Pteridophyta/Monilophyt Estinnophyton wahnbachense 1 - Not planar Fairon- a Demaret, 1979 Pragian Lycopodiopsida Leclercqia andrewsii 0.5 0.29 Planar, cuneate Gensel and Kasper, 2005 Pragian Lycopodiopsida Protolepidodendron wahnbachense 1 - Not planar Fairon- =Estinnophyton Demaret, 1979 Pragian Psilophytophyta Psilophyton dapsile - 0.1 Not planar Andrews et al., 1977, Banks et al., 1975 196

Pragian Lycopodiopsida Sengelia radicans 0.7 - Planar, linear/strap- Matsunaga and like Tomescu, 2017 Emsian Lycopodiopsida Asteroxylon mackiei 0.78 0.44 Not planar Kidston and Lang, 1920 Emsian Lycopodiopsida Baragwanathia abitibiensis 2.5 0.2 Planar, linear/strap- Hueber, 1983 like Emsian Lycopodiopsida Drepanophycus spinaeformis 2 0.3 Not planar Fairon- Demaret, 1978b, Gensel and Kasper, 2005, Li et al., 2000, Rayner, 1984 Emsian Pteridophyta/Monilophyt Estinnophyton gracile 0.7 - Not planar Fairon- a Demaret, 1978a Emsian Pteridophyta/Monilophyt Foozia minuta 1.7 0.09 Not planar Osborne et al., a 2004 Emsian Lycopodiopsida Leclercqia andrewsii 0.5 0.29 Planar, cuneate Gensel and Kasper, 2005 Emsian Lycopodiopsida Leclercqia complexa 0.65 0.18 Planar, cuneate Gensel and Albright, 2006 Emsian Pteridophyta/Monilophyt Pertica quadrifaria 0.04 0.3 Not planar Kasper and a Andrews, 1972 Emsian Lycopodiopsida Protolepidodendron wahnbachense 1 - Not planar Fairon- =Estinnophyton Demaret, 1979 Emsian Psilophytophyta Psilophyton charientos - 0.03 Not planar Gensel, 1979 Emsian Psilophytophyta Psilophyton dapsile - 0.1 Not planar Andrews et al., 1977 197

Emsian Psilophytophyta Psilophyton dawsonii - 0.1 Not planar Banks et al., 1975 Emsian Psilophytophyta Psilophyton forbesii - 0.2 Not planar Andrews et al., 1977, Robertson et al., 2013 Emsian Psilophytophyta Psilophyton microspinosum - 0.1 Not planar Andrews et al., 1977 Emsian Psilophytophyta Psilophyton princeps - 0.08 Not planar Osborne et al., 2004 Emsian Lycopodiopsida Sengelia devonica 0.45 - Planar, linear/strap- Matsunaga and like Tomescu, 2017, Schweitzer and Giesen, 1980 Emsian Lycopodiopsida Sengelia minor 0.6 0.11 Planar, linear/strap- Xu et al., 2013 like Eifelian Lignophytes Aneurophyton furcatum - 0.02 Not planar Osborne et al., 2004 Eifelian Lignophytes Archaeopteris latifolia - - Planar, ovate Moreno- Sánchez, 2004 Eifelian Pteridophyta/Monilophyt Calamophyton primaevum 1.6 0.2 Not planar Leclercq, 1969 a Eifelian unknown cf. Platyphyllum sp. 2 1.5 Planar, fan-/wedge- Høeg, 1942 shape Eifelian Lycopodiopsida Colpodexylon trifurcatum 3 0.2 Not planar Banks, 1944 Eifelian Pteridophyta/Monilophyt Duisbergia mirabilis 5 1.79 Planar, fan-/wedge- Giesen and a shaped Berry, 2013 Eifelian Pteridophyta/Monilophyt Estinnophyton gracile 0.7 Not planar Fairon- 198

a Demaret, 1978a, Fairon- Demaret, 1979 Eifelian Pteridophyta/Monilophyt Hyenia sphenophylloide 0.25 0.09 Not planar Høeg, 1945, a s Høeg, 1935 Eifelian Lycopodiopsida Leclercqia andrewsii 0.5 0.29 Planar, cuneate Gensel and Kasper, 2005 Eifelian Lycopodiopsida Leclercqia complexa 0.65 0.18 Planar, cuneate Gensel and Albright, 2006 Eifelian Pteridophyta/Monilophyt Pertica quadrifaria 0.04 0.3 Not planar Kasper and a Andrews, 1972 Eifelian unknown Platyphyllum peachii 3.5 3.5 Planar, fan-/wedge- Høeg, 1942 shape Eifelian unknown Platyphyllum pusillum 2.2 1.4 Planar, fan-/wedge- Høeg, 1942 shape Eifelian unknown Platyphyllum sp. 7.62 - Planar, fan-/wedge- Høeg, 1942 shape Eifelian Pteridophyta/Monilophyt Pseudosporochnus nodosus - 1.1 Not planar Berry and a Fairon- Demaret, 2002 Eifelian Psilophytophyta Psilophyton dapsile - 0.1 Not planar Andrews et al., 1977, Banks et al., 1975 Eifelian Psilophytophyta Psilophyton forbesii - 0.2 Not planar Andrews et al., 1977 Eifelian Psilophytophyta Psilophyton microspinosum - 0.1 Not planar Andrews et al., 1977 Eifelian Psilophytophyta Psilophyton princeps - 0.08 Not planar Osborne et al., 199

2004 Eifelian Lignophytes Svalbardia scotica 2 0.3 Planar, cuneiform Chaloner, 1972 Givetian Lignophytes Actinopodium banksii - 0.1 Not planar Matten, 1968 Givetian Lignophytes Aneurophyton germanicum - 0.1 Not planar Osborne et al., 2004 Givetian Lignophytes Archaeopteris obtusa 1.8 3.5 Planar, obovate Moreno- Sánchez, 2004 Givetian Lignophytes Archaeopteris latifolia - - Planar, ovate Moreno- Sánchez, 2004 Givetian Pteridophyta/Monilophyt Calamophyton bicephalum 1.5 0.05 Not planar Bonamo and a Banks, 1966b, Leclercq and Andrews, 1960 Givetian Pteridophyta/Monilophyt Calamophyton primaevum 1.6 0.2 Not planar Leclercq, 1969 a Givetian Pteridophyta/Monilophyt Cephalopteris? praecox - 0.2 Planar, oblanceolate Schweitzer, a =Protocephalopteris 1968 Givetian unknown cf. Platyphyllum sp. 2 1.5 Planar, fan-/wedge- Høeg, 1942 shape Givetian Lycopodiopsida Clwydia formerly vanuxemi 0.6 0.15 Not planar Fairon-Demaret Archaeosigillaria and Banks, 1978 Givetian Lycopodiopsida Colpodexylon deatsii 3 0.08 Not planar Banks, 1944 Givetian Lycopodiopsida Drepanophycus colophyllus - 0.18 Not planar Grierson and Banks, 1983 Givetian Lycopodiopsida Drepanophycus colophyllus 0.26 - Not planar Grierson and Banks, 1983 Givetian stem Lycopodiopsida Enigmophyton superbum 16 12 Planar, fan-/wedge- Høeg, 1942 200

shape Givetian Pteridophyta/Monilophyt Eospermatopteris erianus - - Planar, fern-like fonds Boyer, 1995 a Givetian Lycopodiopsida Gilboaphyton griersonii 0.9 0.149 Planar Berry and Edwards, 1997 Givetian Lycopodiopsida Gilboaphyton oldringiae 0.08 - Planar Arnold, 1937 formally goldringiae Givetian Lignophytes Ginkgophytopsis sp. 4 3 Planar, fan-/wedge- Canright, 1970 shape Givetian Pteridophyta/Monilophyt Hostinella globasa - 0.05 Not planar Osborne et al., a 2004 Givetian Pteridophyta/Monilophyt Hostinella racemosa - 0.2 Not planar Osborne et al., a 2004 Givetian Pteridophyta/Monilophyt Hyenia banksii 1.1 0.05 Not planar Arnold, 1940 a Givetian Pteridophyta/Monilophyt Hyenia elegans - 0.1 Not planar Fairon-Demaret a and Berry, 2000 Givetian Pteridophyta/Monilophyt Hyenia sphenophylloide 0.25 0.09 Not planar Høeg, 1945 a s Givetian Pteridophyta/Monilophyt Hyenia vogtii 1.25 Not planar Berry, 2005 a Givetian Psilophytophyta Ibyka amphikoma 0.1 0.05 Not planar Skog and Banks, 1973 Givetian Lycopodiopsida Leclercqia complexa 0.65 0.18 Planar, cuneate Gensel and Albright, 2006 Givetian Lycopodiopsida Leclercqia scolopendra 0.13 0.248 Planar, cuneate Benca et al., 2014 201

Givetian Pteridophyta/Monilophyt Lorophyton goense 4.5 0.45 Not planar Fairon-Demaret a and Li, 1993 Givetian unknown Platyphyllum peachii 3.5 3.5 Planar, fan-/wedge- Høeg, 1942 shape Givetian unknown Platyphyllum pusillum 2.2 1.4 Planar, fan-/wedge- Høeg, 1942 shape Givetian unknown Platyphyllum sp. 7.62 - Planar, fan-/wedge- Høeg, 1942 shape Givetian Lycopodiopsida Protolepidodendron gilboense 0.6 - Not planar Boyer, 1995 Givetian Lycopodiopsida Protolepidodendron scharianum 0.6 - Planar, Fairon- lanceolate/oblanceolat Demaret, 1980 e Givetian Pteridophyta/Monilophyt Pseudosporochnus nodosus - 1.1 Not planar Leclercq and a Banks, 1962 Givetian Lignophytes Rellimia thomsonii - 0.06 Not planar Osborne et al., 2004 Givetian Pteridophyta/Monilophyt Sphenopteris brabantica 1 0.6 Planar, fan-/wedge- Berry, 2009 a shaped Givetian Lignophytes Svalbardia avelinesiana - 0.25 Not planar Osborne et al., 2004 Givetian Lignophytes Svalbardia boyi - 0.21 Not planar Osborne et al., 2004 Givetian Lignophytes Svalbardia polymorpha 2.5 0.05 Not planar Osborne et al., 2004 Givetian Lignophytes Svalbardia scotica 2 0.3 Not planar Allen and Marshall, 1986 Givetian Lignophytes Triloboxylon hallii 0.05 0.05 Not planar Arnold, 1940, Aneurophyton Stein and 202

Beck, 1983 Givetian Pteridophyta/Monilophyt Wattieza givetiana 1.3 - Not planar Berry, 2009 a Frasnian Lignophytes Actinopodium nathorstii - - Not planar Høeg, 1942 Frasnian Lignophytes Aneurophyton germanicum - 0.07 Not planar Osborne et al., 2004 Frasnian Lignophytes Archaeopteris fissilis 1 1 Not planar Carluccio et al., 1966, Osborne et al., 2004 Frasnian Lignophytes Archaeopteris gaspiensis - 1.2 Planar, fan-/wedge- Osborne et al., shape 2004 Frasnian Lignophytes Archaeopteris halliana 1.5 0.55 Planar, cuneate Beck, 1971, GUO and WANG, 2009, Moreno- Sánchez, 2004 Frasnian Lignophytes Archaeopteris hibernica 2 2 Planar Carluccio et al., 1966, Moreno- Sánchez, 2004 Frasnian Lignophytes Archaeopteris jacksoni 1.5 1.05 Planar, fan-/wedge- Moreno- shape Sánchez, 2004, Osborne et al., 2004 Frasnian Lignophytes Archaeopteris macilenta 2.5 2 Planar, Arnold, 1936, obovate/cuneate Carluccio et al., 1966 Frasnian Lignophytes Archaeopteris minor 1 0.4 Planar, obovate Arnold, 1939, Moreno- 203

Sánchez, 2004 Frasnian Lignophytes Archaeopteris obtusa 1.8 3.5 Planar, obovate Osborne et al., 2004 Frasnian Lignophytes Archaeopteris rogersi 1.5 - Planar, fan-/wedge- Smith and shape White, 1905 Frasnian Lignophytes Archaeopteris sphenophyllifoli 2 - Planar, fan-/wedge- Arnold, 1936 a shape Frasnian Pteridophyta/Monilophyt Calamophyton primaevum 1.6 0.2 Not planar Leclercq, 1969 a Frasnian Pteridophyta/Monilophyt Cephalopteris mirabilis 2 1.5 Planar, oblanceolate Schweitzer, a 2006 Frasnian Lycopodiopsida Clwydia vanuxemi 0.6 0.15 Not Planar Fairon-Demaret Archaeosigillaria and Banks, 1978 Frasnian Lignophytes Eddya sullivanensis 6 2.69 Planar, fan-/wedge- Beck, 1967 shape Frasnian Pteridophyta/Monilophyt Ellesmeris sphenopteroides 0.9 0.7 Planar, cuneate and Osborne et al., a lobate 2004 Frasnian Pteridophyta/Monilophyt Hyenia elegans - 0.1 Not planar Osborne et al., a 2004 Frasnian Lycopodiopsida Lepidosigillaria whitei 3 - Not Planar Berry and Marshall, 2015 Frasnian Pteridophyta/Monilophyt Prosseria grandis - 0.6 Planar, linear Read, 1953 a Frasnian Pteridophyta/Monilophyt Pseudobornia ursina 6 1 Planar, obcuneate to Schweitzer, a linear 2006 Frasnian Pteridophyta/Monilophyt Pseudosporochnus nodosus - 1.1 Not planar Osborne et al., a 2004 204

Frasnian Pteridophyta/Monilophyt Rhacophyton condrusorum - 0.04 Not planar Osborne et al., a 2004 Frasnian Pteridophyta/Monilophyt Rhacophyton zygopteroides 1 1 Not planar Leclercq, 1951 a Frasnian Lignophytes Svalbardia banksii 3 2 Not planar Matten, 1981 Frasnian Lignophytes Triloboxylon ashlandicum - 0.1 Not planar Scheckler, 1975a Frasnian Lignophytes Triloboxylon hallii 0.05 0.05 Planar, ovate Arnold, 1940, Aneurophyton Stein and Beck, 1983 Famennian Lignophytes Aneurophyton olnense - 0.03 Not planar Osborne et al., 2004 Famennian Lignophytes Archaeopteris fimbriata 1.5 1 Planar, fan-/wedge- Carluccio et al., shape 1966, Moreno- Sánchez, 2004 Famennian Lignophytes Archaeopteris gaspiensis - 1.2 Planar, fan-/wedge- Osborne et al., shape 2004 Famennian Lignophytes Archaeopteris halliana 1.5 0.55 Planar, fan-/wedge- Carluccio et al., shape 1966; Guo and Wang, 2009 Famennian Lignophytes Archaeopteris hibernica 2 2 Planar, ovate Moreno- Sánchez, 2004, Osborne et al., 2004 Famennian Lignophytes Archaeopteris hitchcockii 0.24 0.06 Planar Smith and White, 1905 Famennian Lignophytes Archaeopteris jacksoni 1.5 1.05 Planar Moreno- Sánchez, 2004, 205

Osborne et al., 2004 Famennian Lignophytes Archaeopteris latifolia 1 1.13 Planar Osborne et al., 2004 Famennian Lignophytes Archaeopteris macilenta 2.5 2 Planar Arnold, 1936, Carluccio et al., 1966 Famennian Lignophytes Archaeopteris minor 1 0.4 Planar Arnold, 1939, Moreno- Sánchez, 2004 Famennian Lignophytes Archaeopteris obtusa 1.8 3.5 Planar Osborne et al., 2004 Famennian Lignophytes Archaeopteris roemeriana 2.65 1.45 Planar, spatulate Kenrick and Fairon- Demaret, 1991 Famennian Lignophytes Archaeopteris sphenophyllifoli 2 - Planar, fan-/wedge- Osborne et al., a shape 2004 Famennian Lycopodiopsida Asteroxylon setchelli - - Planar, linear/strap- Read and like Campbell, 1939 Famennian Pteridophyta/Monilophyt Cephalopteris mirabilis 2 1.79 Planar, oblanceolate Giesen and a Berry, 2013, Osborne et al., 2004 Famennian Lycopodiopsida Clevelandodendron ohioensis 0.45 0.13 Planar, linear Chitaley and Pigg, 1996 Famennian Lycopodiopsida Cyclostigma kiltorkense 15 - Planar, linear Chaloner, 1968, Jurinodendron Haughton, 1860 Famennian Pteridophyta/Monilophyt Eviostachya hoegii 7.5 3.8 Planar, fan-/wedge- SStockmans, 206

a shape 1948, Wang, 1993 Famennian Lignophytes Laceya hibernica - 0.21 Not planar Klavins and Matten, 1996 Famennian unknown Platyphyllum brownianum 15 3.4 Planar, cuneate and Høeg, 1942 lobate Famennian unknown Platyphyllum sp. 1.9 5.09 Planar, fan- /wedge- Høeg, 1942 shape Famennian unknown Platyphyllum sp. - 1.3 Planar, fan-/wedge- Høeg, 1942 shape Famennian Pteridophyta/Monilophyt Pseudobornia ursina 6 1 Planar, obcuneate Schweitzer, a 2006 Famennian Pteridophyta/Monilophyt Rhacophyton ceratangium 2.5 0.22 Planar, linear Cornet et al., a 1976, Osborne et al., 2004 Famennian Pteridophyta/Monilophyt Rhacophyton condrusorum - 0.04 Planar, linear Osborne et al., a 2004 Famennian Pteridophyta/Monilophyt Rhacophyton zygopteroides 1 1 Planar, linear Leclercq, 1951, a Osborne et al., 2004 Famennian Lignophytes Sphenocyclopteridiu belgicum - 0.51 Planar Osborne et al., m 2004 Famennian Pteridophyta/Monilophyt Sphenophyllum subtenerrimum 0.5 0.92 Planar, fan-/wedge- Nathorst, 1902, a =tenerrumum shape Schweitzer, 2006 Famennian Lignophytes Sphenopteridium keilhauii - 0.13 Not planar Osborne et al., 2004 Famennian Lignophytes Sphenopteridium norbergi - 0.16 Planar Osborne et al., 207

2004 Famennian Lignophytes Sphenopteridium rigidum - 0.25 Planar Osborne et al., 2004 Famennian Pteridophyta/Monilophyt Sphenopteris falccida - 0.17 Planar, fern-like fonds Osborne et al., a 2004 Famennian Pteridophyta/Monilophyt Sphenopteris hookeri - 0.1 Planar, fern-like fonds Osborne et al., a 2004 Famennian Pteridophyta/Monilophyt Sphenopteris maillieuxi - 0.06 Planar, fern-like fonds Osborne et al., a 2004 Famennian Pteridophyta/Monilophyt Sphenopteris mourloni - 0.46 Planar, fern-like fonds Osborne et al., a 2004 Famennian Pteridophyta/Monilophyt Sphenopteris schimpeliana - 0.06 Planar, fern-like fonds Osborne et al., a 2004

208

References

Allen, K. & Marshall, J. 1986. Svalbardia and the 'Corduroy' plant from the Devonian of the Shetland islands, Scotland. Special Paper in Palaeontology, 35, 7–20.

Ammosov, I. 1964. Petrographic composition of coals in the USSR and some of its changes. International Geology Review, 6, 1798–1804.

Andrews, H. N., Gensel, P. & Forbes, W. 1974. An apparently heterosporous plant from the Middle Devonian of New Brunswick. Paleontology, 17, 387–408.

Andrews, H. N., Gensel, P. G. & Kasper, A. E. 1975. A new fossil plant of probable intermediate affinities Trimerophyte–Progymnosperm. Canadian Journal of Botany, 53, 1719–1728.

Andrews, H. N., Kasper, A. & Mencher, E. 1968. Psilophyton forbesii, a new Devonian plant from northern Maine. Bulletin of the Torrey Botanical Club, 1–11.

Andrews, H. N., Kasper, A. E., Forbes, W. H., Gensel, P. G. & Chaloner, W. G. 1977. Early Devonian flora of the Trout Valley formation of northern Maine. Review of Palaeobotany and Palynology, 23, 255–285.

Andrews, H. N. & Phillips, T. L. 1968. Rhacophyton from the Upper Devonian of West Virginia. Botanical Journal of the Linnean Society, 61, 37–64.

Andrews, H. N., Read, C. B. & Mamay, S. H. 1971. A Devonian lycopod stem with well- preserved cortical tissues. Palaeontology, 14, 1–9.

Arnold, C. A. 1935. Some new forms and new occurrences of fossil plants from the Middle and Upper Devonian of New York State. Bulletin of the Buffalo Society of Natural Sciences, 17, 1–12.

Arnold, C. A. 1936. Observations on fossil plants from the Devonian of eastern North America. II. Archaeopteris macilenta and A. sphenophyllifolia of Lesquereux. Contributions from Museum of Paleontology, University of Michigan, 5, 49–56.

Arnold, C. A. 1937. Observations on fossil plants from the Devonian of eastern North America. III. Gilboaphyton goldringiae, gen. et sp. nov., from the Hamilton of eastern New York. Contributions from the Museum of Paleontology, University of Michigan, 5, 76.

Arnold, C. A. 1939. Observations on fossil plants from the Devonian of eastern North America. IV. Plant remains from the Catskill Delta deposits of northern Pennsylvania

209

and southern New York. Contributions from Museum of Paleontology University of Michigan, 5, 271–314.

Arnold, C. A. 1940. Structure and relationships of some Middle Devonian plants from western New York. American Journal of Botany, 27, 57–63.

Banks, H., Leclercq, S. & Hueber, F. 1975. Anatomyand morphology of Psilophyton dawsonii, sp. n. from theLate Lower Devonian of Quebec Gaspe, and Ontario, . Palaeontographic Americana, 8, 77–127.

Banks, H. P. 1944. A new Devonian lycopod genus from southeastern New York. American Journal of Botany, 649–659.

Banks, H. P. & Davis, M. 1969. Crenaticaulis, a new genus of Devonian plants allied to Zosterophyllum, and its bearing on the classification of early land plants. American Journal of Botany, 56, 436–449.

Beck, C. B. 1960. Studies of New Albany Shale plants. I. Stenokoleos simplex comb. nov. American Journal of Botany, 47, 115–124.

Beck, C. B. 1967. Eddya sullivanensis, gen. et sp. nov., a plant of gymnospermic morphology from the Upper Devonian of New York. Palaeontographica Abteilung B, 1–22.

Beck, C. B. 1971. On the anatomy and morphology of lateral branch systems of Archaeopteris. American Journal of Botany, 58, 758–784.

Benca, J. P., Carlisle, M. H., Bergen, S. & Strömberg, C. A. 2014. Applying morphometrics to early land plant systematics: A new Leclercqia Lycopsida species from Washington State, USA. American journal of botany, 101, 510–520.

Berry, C. 2009. Short Note: The Middle Devonian plant collections of Francois Stockmans reconsidered. Geologica belgica, 12, 25–30.

Berry, C. M. 2000. A reconsideration of Wattieza Stockmans here attributed to Cladoxylopsida based on a new species from the Devonian of Venezuela. Review of Palaeobotany and Palynology, 112, 125–146. 210

Berry, C. M. 2005. ‘Hyenia’vogtii Høeg from the Middle Devonian of Spitsbergen—Its morphology and systematic position. Review of Palaeobotany and Palynology, 135, 109–116.

Berry, C. M. & Edwards, D. 1994. New data on the morphology and anatomy of the Devonian zosterophyll Serrulacaulis Hueber and Banks from Venezuela. Review of Palaeobotany and Palynology, 81, 141–150.

Berry, C. M. & Fairon-Demaret, M. 2002. The architecture of Pseudosporochnus nodosus Leclercq et Banks: a Middle Devonian cladoxylopsid from Belgium. International Journal of Plant Sciences, 163, 699–713.

Berry, C. M. & Marshall, J. E. 2015. Lycopsid forests in the early Late Devonian paleoequatorial zone of Svalbard. Geology, 43, 1043–1046.

Bonamo, P. M. 1977. Rellimia thomsonii Progymnospermopsida from the Middle Devonian of New York State. American Journal of Botany, 64, 1272–1285.

Bonamo, P. M. & Banks, H. P. 1966a. Calamophyton in the Middle Devonian of New York State. American Journal of Botany, 53, 778–791.

Bonamo, P. M. & Banks, H. P. 1966b. Calamophyton in the Middle Devonian of New York State. American Journal of Botany, 778–791.

Boyer, J. 1995. Reexamination of Eospermatopteris eriana Dawson Goldring from the upper Middle Devonian = Givetian flora at Gilboa, New York. M.S. Thesis, Southern Illinois University, Carbondale.

Boyer, J. S. & Stein, W. E. 2008. A Morphometric Analysis of an Early Euphyllophyte, Psilophyton dapsile. Ph.D. Thesis, The State University of New York.

Brauer, D. F. 1980. Barinophyton citrulliforme Barinophytales incertae sedis, Barinophytaceae from the Upper Devonian of Pennsylvania. American Journal of Botany, 67, 1186–1206.

Burgess, N. & Edwards, D. 1988. A new Palaeozoic plant closely allied to Prototaxites Dawson. Botanical Journal of the Linnean Society, 97, 189–203.

211

Canright, J. E. 1970. Spores and associated macrofossils from the Devonian of Arizona. Geoscience and Man, 1, 83–88.

Carluccio, L. M., Hueber, F. M. & Banks, H. P. 1966. Archaeopteris macilenta, anatomy and morphology of its frond. American Journal of Botany, 53, 719–730.

Chaloner, W. 1968. The cone of Cyclostigma kiltorkense Haughton, from the Upper Devonian of Ireland. Botanical Journal of the Linnean Society, 61, 25–36.

Chaloner, W. G. 1972. Devonian plants from Fair isle, Scotland. Review of Palaeobotany and Palynology, 14, 49–61.

Chitaley, S. & Pigg, K. B. 1996. Clevelandodendron ohioensis, gen. et sp. nov., a slender upright lycopsid from the Late Devonian Cleveland Shale of Ohio. American Journal of Botany, 83, 781–789.

Cordi, J. & Stein, W. E. 2005. The anatomy of Rotoxylon dawsonii comb. nov. Cladoxylon dawsonii from the Upper Devonian of New York State. International journal of plant sciences, 166, 1029–1045.

Crépin, F. 1875. Observations sur quelques plantes fossiles des dépôts dévoniens rapportés par Dumont à l'étage quartzoschisteux inférieur de son système Eifelien. Bulletin de la Société Royale de Botanique de Belgique/Bulletin van de Koninklijke Belgische Botanische Vereniging, 214–230.

Cressler, W. L. & Pfefferkorn, H. W. 2005. A Late Devonian isoetalean lycopsid, Otzinachsonia beerboweri, gen. et sp. nov., from north-central Pennsylvania, USA. American journal of botany, 92, 1131–1140.

Croft, W. N. & Lang, W. H. 1942. The Lower Devonian flora of the Senni beds of Monmouthshire and Breconshire. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 231, 131–163.

Daber, R. 1960. Eogaspesiea gracilis n.g.n. sp. Geologie, 4, 418–425.

212

Dai, S., Han, D. & Chou, C. L. 2006. Petrography and geochemistry of the Middle Devonian coal from Luquan, Yunnan Province, China. Fuel, 85, 456–464.

Danzé-Corsin, P. 1956. Contribution à l'étude des flores dévoniennes du Nord de la France. II. Flore éodévonienne de Rébreuve. Annales de la Société géologique de Belgique, 76, 24–50.

Decombeix, A. L., Meyer-Berthaud, B., Rowe, N. & Galtier, J. 2005. Diversity of large woody lignophytes preceding the extinction of Archaeopteris: new data from the middle Tournaisian of Thuringia Germany. Review of Palaeobotany and Palynology, 137, 69– 82.

Diessel, C. F. 2010. The stratigraphic distribution of inertinite. International Journal of Coal Geology, 81, 251–268.

Dittrich, H. S., Matten, L. C. & Phillips, T. L. 1983. Anatomy of Rhacophyton ceratangium from the upper devonian Famenian of west Virginia. Review of palaeobotany and palynology, 40, 127–147.

Doran, J. B. 1980. A new species of Psilophyton from the Lower Devonian of northern New Brunswick, Canada. Canadian Journal of Botany, 58, 2241–2262.

Doran, J. B., Gensel, P. G. & Andrews, H. N. 1978. New occurrences of trimerophytes from the Devonian of eastern Canada. Canadian Journal of Botany, 56, 3052–3068.

Dorf, E. 1933. A new occurence of the oldest known terrestrial vegetation, from Beartooth Butte, Wyoming. Botanical Gazette, 95, 240–257.

Edwards, D. 1968. A new plant from the lower Old Red Sandstone of south Wales. Palaeontology, 11, 683–690.

Edwards, D. 1969a. Further observations on Zosterophyllum llanoveranum from the Lower Devonian of South Wales. American Journal of Botany, 56, 201–210.

Edwards, D. 1969b. Zosterophyllum from the lower old red sandstone of South Wales. New Phytologist, 68, 923–931.

Edwards, D. 1970a. Fertile Rhyniophyta from the Lower Devonian of Britain. Palaeontology, 13, 451–461.

Edwards, D. 1970b. Further observations on the Lower Devonian plant, Gosslingia breconensis Heard. Philosophical Transactions of the Royal Society of London. B, Biological Sciences, 258, 225–243.

213

Edwards, D. 1975. 11.—Some Observations on the Fertile Parts of Zosterophyllum myretonianum Penhallow from the Lower Old Red Sandstone of Scotland. Earth and Environmental Science Transactions of The Royal Society of Edinburgh, 69, 251–265.

Edwards, D. 1976. The systematic position of Hicklingia edwardii Kidston and Lang. New Phytologist, 76, 173–181.

Edwards, D. 1979. A late Silurian flora from the Lower Old Red Sandstone of south-west Dyfed. Palaeontology, 22, 23–52.

Edwards, D. 1980a. Studies on lower devonian petrifactions from Britain. 1. Pyritised axes of Hostinella from the brecon beacons quarry, Powys, South Wales. Review of Palaeobotany and Palynology, 29.

Edwards, D. 1981. Studies on Lower Devonian petrifactions from Britain. 2. Sennicaulis, a new form genus for sterile axes based on pyrite and limonite petrifactions from the Senni Beds. Review of Palaeobotany and Palynology, 32, 207–226.

Edwards, D. 2006. Danziella artesiana, a new name for Zosterophyllum artesianum from the Lower Devonian of Artois, northern France. Review of Palaeobotany and Palynology, 142, 93–101.

Edwards, D. & Axe, L. 2004. Anatomical Evidence in the Detection of the Earliest Wildfires. Palaios, 19, 113–128.

Edwards, D., Fanning, U. & Richardson, J. 1994. Lower Devonian coalified sporangia from Shropshire: Salopella Edwards & Richardson and Tortilicaulis Edwards. Botanical Journal of the Linnean Society, 116, 89–110.

Edwards, D., Kenrick, P. & Carluccio, L. M. 1989. A reconsideration of cf. Psilophyton princeps Croft & Lang, 1942, a zosterophyll widespread in the Lower Old Red Sandstone of South Wales. Botanical journal of the Linnean Society, 100, 293–318.

Edwards, D. & Richardson, J. B. 1974a. Lower Devonian Dittonian plants from the Welsh borderland. Palaeontology, 17, 1.

Edwards, D. & Richardson, J. B. 1974b. Lower Devonian Dittonian plants from the Welsh borderland. Palaeontology, 17, 311–324.

Edwards, D. S. 1980b. Evidence for the sporophytic status of the Lower Devonian plant Rhynia gwynne-vaughanii Kidston and Lang. Review of Palaeobotany and Palynology, 29, 177–188.

214

Erwin, D. M. & Rothwell, G. W. 1989. Gillespiea randolphensis gen. et sp. nov.Stauropteridales, from the Upper Devonian of West Virginia. Canadian journal of botany, 67, 3063–3077.

Fairon-Demaret, M. 1978a. Estinnophyton gracile gen, et sp. nov., A new name for specimens previously determined Protolepidodendron wahnbachense Kräusel and Wayland from the Siegenian of Belgium. Bulletin de la Classe Des Sciences Academie Royale de Belgique, 64, 597–610.

Fairon-Demaret, M. 1978b. Observations nouvelles sur les axes végétatifs de Drepanophycus spinaeformis Göppert de l'Emsien Inférieur des “Nouvelles Carrières” de Dave, Belgique. Review of Palaeobotany and Palynology, 26, 9–20.

Fairon-Demaret, M. 1979. Estinnophyton wahnbachense Kräusel et Weyland comb. nov., une plante remarquable du Siegenien d'Allemagne. Review of Palaeobotany and Palynology, 28, 145–160.

Fairon-Demaret, M. 1980. A propos des spécimens déterminés Protolepidodendron scharianum par Kräusel et Weyland, 1932. Review of Palaeobotany and Palynology, 29, 201–220.

Fairon-Demaret, M. 1985. Fossil plants from the Emsian of Sart Tilman, Belgium. I. Stockmansia langii Stockmans comb. nov. Review of Palaeobotany and Palynology, 44, 243–260.

Fairon-Demaret, M. 1986a. Les plantes emsiennes du sart tilman Belgique. II. Sartilmania jabachensis Kräusel et Weyland comb. nov. Review of palaeobotany and palynology, 47, 225–239.

Fairon-Demaret, M. 1986b. Stockmansella, a new name for Stockmansia Fairon-Demaret fossil. Taxon, 35.

Fairon-Demaret, M. & Banks, H. P. 1978. Leaves of Archaeosigillaria vanuxemii, a Devonian lycopod from New York. American Journal of Botany, 65, 246–249.

Fairon-Demaret, M. & Berry, C. M. 2000. A reconsideration of Hyenia elegans Kräusel et Weyland and Hyeniacomplexa'Leclercq: two Middle Devonian cladoxylopsids from western Europe. International journal of plant sciences, 161, 473–494.

Fairon-Demaret, M. & Hartkopf-Fröder, C. 2004. Late Famennian plant mesofossils from the Refrath 1 Borehole Bergisch Gladbach–Paffrath Syncline; Ardennes–Rhenish Massif, Germany. Courier Forschungsinstitut Senckenberg, 251, 89–121.

215

Fairon-Demaret, M. & Li, C. S. 1993. Lorophyton goense gen. et sp. nov. from the Lower Givetian of Belgium and a discussion of the Middle Devonian Cladoxylopsida. Review of Palaeobotany and Palynology, 77, 1–22.

Fairon-Demaret, M. & Scheckler, S. E. 1987. Typification and redescription of Moresnetia zalesskyi Stockmans, 1948, an early seed plant from the Upper Famennian of Belgium. Bulletin de l'Institut Royal des Sciences naturelles de Belgique, 183–199.

Fanning, U., Edwards, D. & Richardson, J. 1992. A diverse assemblage of early land plants from the Lower Devonian of the Welsh Borderland. Botanical Journal of the Linnean Society, 109, 161–188.

Filipiak, P. & Szaniawski, H. 2016. Nematophytes from the Lower Devonian of Podolia, Ukraine. Review of Palaeobotany and Palynology, 224, 109–120.

Gensel, P., Kasper, A. & Andrews, H. N. 1969. Kaulangiophyton, a new genus of plants from the Devonian of Maine. Bulletin of the Torrey Botanical Club, 265–276.

Gensel, P. G. 1976. Renalia hueberi, a new plant from the Lower Devonian of Gaspé. Review of Palaeobotany and Palynology, 22, 19–37.

Gensel, P. G. 1979. Two Psilophyton species from the Lower Devonian of eastern Canada with a discussion of morphological variation within the genus. Palaeontographica Abteilung B, 81–99.

Gensel, P. G. 1982a. A new species of Zosterophyllum from the Early Devonian of New Brunswick. American Journal of Botany, 69, 651–669.

Gensel, P. G. 1982b. Oricilla, a new genus referable to the Zosterophyllophytes from the late Early Devonian of northern New Brunswick. Review of Palaeobotany and Palynology, 37, 345–359.

Gensel, P. G. & Albright, V. M. 2006. Leclercqia complexa from the Early Devonian Emsian of northern New Brunswick, Canada. Review of Palaeobotany and Palynology, 142, 103–121.

Gensel, P. G., Andrews, H. N. & Forbes, W. H. 1975. A new species of Sawdonia with notes on the origin of microphylls and lateral sporangia. Botanical Gazette, 136, 50–62.

Gensel, P. G. & Kasper, A. E. 2005. A new species of the Devonian lycopod genus, Leclercqia, from the Emsian of New Brunswick, Canada. Review of Palaeobotany and Palynology, 137, 105–123.

216

Gentzis, T. & Goodarzi, F. 1991. Thermal maturity and hydrocarbon potential of the sedimentary succession from the Hecla field in Sverdrup Basin, Arctic Canada. International journal of coal geology, 19, 483–517.

Gerrienne, P. 1988. Early Devonian plant remains from Marchin north of Dinant Synclinorium, Belgium, I. Zosterophyllum deciduum sp. nov. Review of Palaeobotany and Palynology, 55, 317–335.

Gerrienne, P. 1990. Quelques spécimens de Gosslingia breconensis Heard Heard, 1927 dans le Dévonien inférieur de Wihéries bord nord du synclinorium de Dinant, Belgique. Annales de la Société géologique de Belgique.

Gerrienne, P. 1992. Psilophytites gileppensis nov. sp., une plante épineuse du Dévonien inférieur du Bord Sud du Synclinorium de Verviers Belgique. Geobios, 25, 439–447.

Gerrienne, P. 1995. Les fossiles végétaux du Dévonieninférieur de Marchin bord nord du Synclinorium de Dinant, Belgique. III. Psilophyton parvulum nov. sp. Geobios, 28, 131– 144.

Gerrienne, P. 1996. Lower Devonian plant remains from Marchin northern margin of Dinant Synclinorium, Belgium. IV. Odonax borealis gen. et sp. nov. Review of Palaeobotany and Palynology, 93, 89–106.

Gerrienne, P. 1997. The fossil plants from the Lower Devonian of Marchin northern margin of Dinant Synclinorium, Belgium. V. Psilophyton genseliae sp. nov., with hypotheses on the origin of Trimerophytina. Review of Palaeobotany and Palynology, 98, 303–324.

Gerrienne, P. 1998. Deuterophyton stockmansii gen. et sp. nov. du Dévonien inférieur de Belgique et identification d'un nouveau groupe de plantes vasculaires primitives. Comptes Rendus de l'Académie des Sciences-Series IIA-Earth and Planetary Science, 326, 369–375.

Gerrienne, P. 1999. Gosferia, a new name for Forgesia fossil plants. Taxon, 48, 61–61.

Gerrienne, P. & Gensel, P. G. 2016. New data about anatomy, branching, and inferred growth patterns in the Early Devonian plant Armoricaphyton chateaupannense, Montjean-sur- Loire, France. Review of Palaeobotany and Palynology, 224, 38–53.

Gerrienne, P. & Meyer-Berthaud, B. 2006. The permineralized lignophyte remains from Ronquières middle Devonian, Belgium: preliminary results.

Giesen, P. & Berry, C. M. 2013. Reconstruction and growth of the early tree Calamophyton Pseudosporochnales, Cladoxylopsida based on exceptionally complete specimens from

217

Lindlar, Germany Mid-Devonian: organic connection of Calamophyton branches and Duisbergia trunks. International Journal of Plant Sciences, 174, 665–686.

Glasspool, I. J., Edwards, D. & Axe, L. 2004. Charcoal in the Silurian as evidence for the earliest wildfire. Geology, 32, 381.

Glasspool, I. J. & Scott, A. C. 2010. Phanerozoic concentrations of atmospheric oxygen reconstructed from sedimentary charcoal. Nature Geoscience, 3, 627–630.

Glasspool, I. J., Scott, A. C., Waltham, D., Pronina, N. & Shao, L. 2015. The impact of fire on the Late Paleozoic Earth system. Front Plant Sci, 6, 756.

Goodarzi, F., Brooks, P. & Embry, A. 1989. Regional maturity as determined by organic petrography and geochemistry of the Schei Point Group Triassic in the western Sverdrup Basin, Canadian Arctic Archipelago. Marine and Petroleum Geology, 6, 290–302.

Goodarzi, F. & Goodbody, Q. 1990. Nature and depositional environment of Devonian coals from western Melville Island, Arctic Canada. International Journal of Coal Geology, 14, 175–196.

Gothan, W. & Zimmermann, F. 1936. Weiteres über die alt-oberdevonische Flora von Bögendorf-Libichau bei Waldenburg. Jahrbuch der Preußischen Geologischen Landesanstalt für, 57, 487–506.

Granoff, J. A., Gensel, P. G. & Andrews, H. N. 1976. A new species of Pertica from the Devonian of eastern Canada. Palaeontographica Abteilung B, 155, 119–128.

Grierson, J. D. & Banks, H. P. 1963. Lycopods of the Devonian of New York state, Paleontological Research Institution.

Grierson, J. D. & Banks, H. P. 1983. A new genus of lycopods from the Devonian of New York State. Botanical Journal of the Linnean Society, 86, 81–101.

Guo, Y. & Wang, D. 2009. Archaeopteris halliana from the Late Devonian Famennian of Anhui Province, China. Acta Geologica Sinica‐English Edition, 83, 479–491.

Habgood, K., Edwards, D. & Axe, L. 2002. New perspectives on Cooksonia from the Lower Devonian of the Welsh Borderland. Botanical Journal of the Linnean Society, 139, 339– 359.

Halle, T. G. 1916. Lower Devonian plants from Röragen in Norway, Almquist & Wiksells boktryckeri-a.-b.

218

Hass, H. & Remy, W. 1991. Huvenia kleui nov. gen., nov. spec.: ein Vertreter der Rhyniaceae aus dem Höheren Siegen des Rheinischen Schiefergebirges. Argumenta Palaeobotanica, 8, 141–168.

Haughton, S. 1860. XLVI.—On Cyclostigma, a new genus of fossil plants from the Old Red Sandstone of Kiltorcan, co. Kilkenny; and on the general law of phyllotaxis in the natural orders Lycopodiaceæ, Equisetaceæ, Filices, &c. Annals and Magazine of Natural History, 5, 433–445.

Hill, S. A., Scheckler, S. E. & Basinger, J. F. 1997a. Ellesmeris sphenopteroides, gen. et sp. nov., a new zygopterid fern from the Upper Devonian Frasnian of Ellesmere, NWT, Arctic Canada. American journal of botany, 85–103.

Høeg, O. 1945. Contributions to the Devonian flora of western Norway, Ill. Nor. Geol. Tidsskr, 25, 183–192.

Höeg, O. A. 1937. The Devonian floras and their bearing upon the origin of vascular plants. The Botanical Review, 3, 563–592.

Høeg, O. A. 1935. Further contributions to the Middle Devonian flora of western Norway, Norsk geologisk tidsskrift.

Høeg, O. A. 1942a. The Downtonian and Devonian flora of Spitsbergen. Svalbard Ishavs- Undersök, Skrifter, 83, 1–228.

Hoffman, L. A. & Tomescu, A. M. 2013. An early origin of secondary growth: Franhueberia gerriennei gen. et sp. nov. from the Lower Devonian of Gaspe Quebec, Canada. Am J Bot, 100, 754–763.

Hotton, C., Hueber, F., Griffing, D., Bridge, J., Gensel, P. & Edwards, D. 2001. Early terrestrial plant environments: an example from the Emsian of Gaspé, Canada. Plants invade the land: evolutionary and environmental perspectives, 179–212.

Hueber, F. & Grierson, J. 1961. On the occurrence of Psilophyton princeps in the early Upper Devonian of New York. American Journal of Botany, 48, 473–479.

Hueber, F. M. 1970. Rebuchia: a new name for Bucheria Dorf. Taxon, 19, 822

Hueber, F. M. 1971. Sawdonia ornata: a new name for Psilophyton princeps var. ornatum. Taxon, 20, 641–642.

Hueber, F. M. 1983. A new species of Baragwanathia from the Sextant Formation Emsian northern Ontario, Canada. Botanical Journal of the Linnean Society, 86, 57–79.

219

Hueber, F. M. & Banks, H. P. 1979. Serrulacaulis furcatus gen. sp. nov., a new zosterophyll from the lower upper Devonian of New York state. Review of Palaeobotany and Palynology, 28, 169–189.

Jensen, D. P. & Gensel, P. G. 2013. Forania plegiospinosa, gen. et sp. nov.: a zosterophyll from the Early Devonian of New Brunswick, Canada, with a novel emergence type. International Journal of Plant Sciences, 174, 687–701.

Johnson, N. G. 1992. The cuticular epithelium of Bitelaria, an Early Devonian vascular plant from New Brunswick, Canada. International journal of plant sciences, 153, 646–657.

Jones, L. C. 2017. An Integrated Analysis of Sequence Stratigraphy, Petroleum Geochemistry, and Devonian Mass Extinction Events in the Woodford Shale, Southern Oklahoma.

Jurina, A. 2009. Middle Devonian lycophyte Protolepidodendron scharianum: History of study and description of the holotype. Paleontological Journal, 43, 1253–1261.

Kasper, A. E. & Andrews, H. N. 1972. Pertica, a new genus of Devonian plants from northern Maine. American Journal of Botany, 59, 897–911.

Kasper, A. E., Andrews, H. N. & Forbes, W. H. 1974. New fertile species of Psilophyton from the Devonian of Maine. American Journal of Botany, 61, 339–359.

Kenrick, P. & Edwards, D. 1988. A new zosterophyll from a recently discovered exposure of the Lower Devonian Senni Beds in Dyfed, Wales. Botanical journal of the Linnean Society, 98, 97–115.

Kenrick, P. & Fairon-Demaret, M. 1991. Archaeopteris roemeriana Göppert sensu Stockmans, 1948 from the Upper Famennian of Belgium: anatomy and leaf polymorphism. Bulletin de l’Institute Royal des Sciences Naturelles de Belgique, Sciences de la Terre, 61, 179–195.

220

Kerp, H., Hass, H. & Mosbrugger, V. 2001. New data on Nothia aphylla Lyon 1964 ex El- Saadawy et Lacey 1979, a poorly known plant from the Lower Devonian Rhynie Chert. Plants invade the land: evolutionary and environmental perspectives, 52–82.

Kidston, R. & Lang, W. H. 1920. On old red sandstone plants showing structure, from the Rhynie chert bed, Aberdeenshire. Part III. Asteroxylon mackiei, Kidston and Lang. Earth and Environmental Science Transactions of the Royal Society of Edinburgh, 52, 643– 680.

Klavins, S. D. 2004. Re-interpretation of Wexfordia hookense from the Upper Devonian of Ireland as an arborescent lycophyte. Botanical Journal of the Linnean Society, 144, 275– 287.

Klavins, S. D. & Matten, L. C. 1996. Reconstruction of the frond of Laceya hibernica, a lyginopterid pteridosperm from the uppermost Devonian of Ireland. Review of Palaeobotany and Palynology, 93, 253–268.

Kotyk, M. E. & Basinger, J. F. 2000. The Early Devonian Pragian zosterophyll Bathurstia denticulata Hueber. Canadian Journal of Botany, 78, 193–207.

Kotyk, M. E., Basinger, J. F., Gensel, P. G. & De Freitas, T. A. 2002. Morphologically complex plant macrofossils from the Late Silurian of Arctic Canada. American Journal of Botany, 89, 1004–1013.

Krausel, R. & Weyland, H. 1932. Pflanzenreste aus dem Devon II. Senckenbergiana, 14, 185–190.

Kräusel, R. & Weyland, H. 1926. Beiträge zur Kenntnis der devonflora. Abh. Senckenb. Naturforsch. Gesell., 2, 115–155.

Lang, W. H. 1937. On the plant-remains from the Downtonian of England and Wales. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 227, 245–291.

Leclercq, S. 1942. Quelques plantes fossiles recueillies dans le Dévonien inférieur des environs de Nonceveux Bordure orientale du bassin de Dinant. Annales de la Société Géologique de Belgique, 65, 193–211.

Leclercq, S. 1951. Étude morphologique et anatomique d'une fougère du Dévonien supérieur: le Rhacophyton zygopteroides nov. sp. Annales de la Société géologique de Belgique.

Leclercq, S. 1969. Calamophyton primaevum: the complex morphology of its fertile appendage. American Journal of Botany, 773–781.

221

Leclercq, S. & Andrews, H. N. 1960. Calamophyton bicephalum, a new species from the Middle Devonian of Belgium. Annals of the Missouri Botanical Garden, 1–23.

Leclercq, S. & Banks, H. P. 1962. Pseudosporochnus nodosus sp. nov., a Middle Devonian plant with cladoxylalean affinities. Palaeontographica Abteilung B, 110, 1–34.

Leclercq, S. & Bonamo, P. M. 1973. Rellimia thomsonii, a new name for Milleria Protopteridium thomsonii Lang 1926 emend. Leclercq and Bonamo 1971. Taxon, 435– 437.

Ledoux-Marcelle, H. 1927. Sur les flores du Dévonien de la Belgique. Bulletin de la Société Belge de Géologie, de Paléontologie et d’Hydrologie, 37, 19–30.

Li, C. S., Hueber, F. & Hotton, C. 2000. A neotype for Drepanophycus spinaeformis Göppert 1852. Canadian Journal of Botany, 78, 889–902.

Liu, Z., Selby, D., Hackley, P. C. & Over, D. J. 2018. Fossil charcoal as evidence for intense wildfire at the Frasnian-Famennian boundary early Late Devonian and implications for atmospheric oxygen level rise. The Soceity of Organic Petrology Annual Conference 2018.

Lyon, A. & Edwards, D. 1991. The first zosterophyll from the Lower Devonian Rhynie chert, Aberdeenshire. Earth and Environmental Science Transactions of the Royal Society of Edinburgh, 82, 324–332.

Mägdefrau, K. 1938. Eine Halophyten-Flora aus dem Unterdevon des Harzes. Beihefte zum Botanischen Centralblatt, 58, 243–251.

Marynowski, L. & Filipiak, P. 2007. Water column euxinia and wildfire evidence during deposition of the Upper Famennian Hangenberg event horizon from the Holy Cross Mountains central Poland. Geological Magazine, 144, 569–595.

Matsunaga, K. K. & Tomescu, A. M. 2017. An organismal concept for Sengelia radicans gen. et sp. nov.—morphology and natural history of an Early Devonian lycophyte. Annals of botany, 119, 1097–1113.

Matten, L. C. 1968. Actinoxylon banksii gen. et sp. nov.: a progymnosperm from the Middle Devonian of New York. American Journal of Botany, 55, 773–782.

Matten, L. C. 1973. The Cairo flora Givetian from eastern New York. I. Reimannia, terete axes, and Cairoa lamanekii gen. et sp. n. American Journal of Botany, 60, 619–630.

Matten, L. C. 1981. Svalbardia banksii sp. nov. from the upper Devonian Frasnian of New York State. American Journal of Botany, 68, 1383–1391. 222

Matten, L. C. 1989. A petrified lycopod from the uppermost Devonian of Hook Head, County Wexford, Ireland. Botanical Gazette, 150, 323–336.

Matten, L. C. & Banks, H. P. 1966. Triloboxylon ashlandicum gen. and sp. n. from the Upper Devonian of New York. American Journal of Botany, 53, 1020–1028.

Matten, L. C. & Banks, H. P. 1969. Stenokoleos bifidus sp. n. in the Upper Devonian of New York State. American Journal of Botany, 56, 880–891.

Matyja, H., Sobień, K., Marynowski, L., Stempień-Sałek, M. & Małkowski, K. 2015. The expression of the Hangenberg Event latest Devonian in a relatively shallow-marine succession Pomeranian Basin, Poland: the results of a multi-proxy investigation. Geological Magazine, 152, 400–428.

Michelsen 1991. A regional study on coals from Svalbard organic facies, maturity and thermal history. Bulletin de la Société Géologique de France, 162, 385–397.

Momont, N., Gerrienne, P. & Prestianni, C. 2016. Brabantophyton, a new genus with stenokolealean affinities from a Middle to earliest Upper Devonian locality from Belgium. Review of Palaeobotany and Palynology, 227, 77–96.

Montero, A. & Dieguez, C. 2010. Devonian floral assemblages and plant megafossils from the Iberian Peninsula: a review. Review of Palaeobotany and Palynology, 162, 231–238.

Moreno-Sánchez, M. Graphic approach for morphometric analysis of Archaeopteris leaves. Annales de paléontologie, 2004. Elsevier, 161–173.

Morris, J. & Edwards, D. 2014. An analysis of vegetational change in the Lower Devonian: new data from the Lochkovian of the Welsh Borderland, UK. Review of palaeobotany and palynology, 211, 28–54.

Nathorst, A. G. 1902. Zur Oberdevonischen flora der Bären-insel, Kungl. boktryckeriet, PA Nordstedt & söner.

Obrhel, J. 1959. Neue Pflanzenfunde in den Srbsko-Schichten Mitteldevon. Vestnik Ustredniho Ustavu Geologickeho, 34, 384–389.

Penhallow, D. P. 1893. Notes on Erian Devonian Plants from New York and Pennsylvania: Notes on Nematophyton Crassum, US Government Printing Office. 223

Pfeiler, K. C. & Tomescu, A. M. 2018. An Early Devonian permineralized rhyniopsid from the Battery Point Formation of Gaspé Canada. Botanical Journal of the Linnean Society, 187, 292–302.

Pflug, H. & Prössl, K. 1991a. Palynology in the pilot hole of the Continental Deep Drilling Program: results and implications. KTB report, 91, 239–258.

Pflug, H. & Prössl, K. 1991b. Palynostratigraphical and paleobotanical studies in the pilot hole of the German continental deep drilling program:. results and implications. Scientific drilling, 2, 13–33.

Pigg, K. B. 2001. Isoetalean lycopsid evolution: from the Devonian to the present. American Fern Journal, 91, 99–114.

Powell, C. L., Edwards, D. & Trewin, N. H. 1999. A new vascular plant from the Lower Devonian Windyfield chert, Rhynie, NE Scotland. Earth and Environmental Science Transactions of the Royal Society of Edinburgh, 90, 331–349.

Prestianni, C., Decombeix, A.-L., Thorez, J., Fokan, D. & Gerrienne, P. 2010. Famennian charcoal of Belgium. Palaeogeography, Palaeoclimatology, Palaeoecology, 291, 60–71.

Rayner, R. 1983. New observations on Sawdonia ornata from Scotland. Earth and Environmental Science Transactions of the Royal Society of Edinburgh, 74, 79–93.

Rayner, R. 1984. New finds of Drepanophycus spinaeformis Göppert from the Lower Devonian of Scotland. Earth and Environmental Science Transactions of The Royal Society of Edinburgh, 75, 353–363.

Read, C. B. 1935. An occurrence of the genus Cladoxylon Unger, in North America. Journal of the Washington Academy of Sciences, 25, 493–497.

Read, C. B. 1936. A Devonian flora from Kentucky. Journal of Paleontology, 215–227.

Read, C. B. 1953. Prosseria grandis, a new genus and new species from the Upper Devonian of New York. Journal of the Washington Academy of Sciences, 43, 13–16.

Read, C. B. & Campbell, G. 1939. Preliminary account of the New Albany Shale flora. American Midland Naturalist, 21, 435–453.

Retallack, G. J. 2011. Woodland hypothesis for Devonian tetrapod evolution. The Journal of Geology, 119, 235–258.

224

Retallack, G. J. & Huang, C. 2011. Ecology and evolution of Devonian trees in New York, USA. Palaeogeography, Palaeoclimatology, Palaeoecology, 299, 110–128.

Riboulleau, A., Spina, A., Vecoli, M., Riquier, L., Quijada, M., Tribovillard, N. & Averbuch, O. 2018. Organic matter deposition in the Ghadames Basin Libya during the Late Devonian—A multidisciplinary approach. Palaeogeography, Palaeoclimatology, Palaeoecology, 497, 37–51.

Robertson, D. S., Lewis, W. M., Sheehan, P. M. & Toon, O. B. 2013. K-Pg extinction: Reevaluation of the heat-fire hypothesis. Journal of Geophysical Research: Biogeosciences, 118, 329–336.

Romero-Sarmiento, M.-F., Riboulleau, A., Vecoli, M. & Versteegh, G. J.-M. 2011. Aliphatic and aromatic biomarkers from Gondwanan sediments of Late Ordovician to Early Devonian age: An early terrestrialization approach. Organic Geochemistry, 42, 605– 617.

Rowe, N. P. & Jones, T. P. 2000. Devonian charcoal. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 331–338.

Scheckler, S. E. 1975a. A fertile axis of Triloboxylon ashlandicum, a progymnosperm from the Upper Devonian of New York. American Journal of Botany, 62, 923–934.

Scheckler, S. E. 1975b. Rhymokalon, a new plant with cladoxylalean anatomy from the Upper Devonian of New York State. Canadian Journal of Botany, 53, 25–38.

Scheckler, S. E. & Banks, H. P. 1971. Anatomy and relationships of some Devonian progymnosperms from New York. American Journal of Botany, 58, 737–751.

Schmidt, W. 1954. Pflanzen-Reste aus der Tonschiefer-Gruppe Unteres Siegen des Siegerlandes. I. Sugambrophyton Pilgeri n. G., N. sp., eine Protolepidodendracee aus den Hamberg-Schichten. Palaeontographica Abteilung B, 1–22.

225

Schultka, S. & Hass, H. 1997. Stockmansella remyi sp. nov. from the Eifelian: new aspects in the Rhyniaceae sensu Hass et Remy, 1991. Review of Palaeobotany and Palynology, 97, 381–393.

Schweitzer, H. J. 1965. Über Bergeria mimerensis und Protolepidodendropsis pulchra aus dem Devon Westspitzbergens. Palaeontographica Abteilung B, 117–138.

Schweitzer, H. J. 1967. Die Oberdevon-Flora der Bäreninsel. 1. Pseudobornia ursina Nathorst. Palaeontographica Abteilung B, 116–137.

Schweitzer, H. J. 1968. Pflanzcnrcstc aus dcm Devon Nord-wcstspitzbergcns. Palaeontographica B, 123, 43–75.

Schweitzer, H. J. 1979. Die Zosterophyllaceae des rheinischen Unterdevons. Bonner Paläobotanische Mitteilungen, 3, 1–32.

Schweitzer, H. J. 1980. Die Gattungen Renalia Gensel und Psilophyton Dawson im Unterdevon des Rheinlandes. Bonner Paläobotanische Mitteilungen, 6, 1–34.

Schweitzer, H. J. 1982. Neue Pflanzenfunde im Unterdevon der Eifel Deutschland. Senckenbergiana Lethaea, 80, 371–395.

Schweitzer, H. J. 2006. Die Oberdevon-Flora der Bäreninsel-5. Gesamtübersicht. Palaeontographica Abteilung B, 1–191.

Schweitzer, H. J. & Giesen, P. 2008. Weitere Pflanzenfunde im Mitteldevon von Wuppertal im Bergischen Land Westdeutschland. Palaeontographica Abteilung B, 101–140.

Schweitzer, H. J. & Lc, M. 1982. Aneurophyton germanicum and Protopteridium thomsonii from the Middle Devonian of Germany.

Schweitzer, H. J. & Matten, L. 1982. Aneurophyton germanicum and Protopteridium thomsonii from the Middle Devonian of Germany. Palaeontographica, 184B, 65–106.

Schweitzer, H. J. 1987. Tursuidea paniculata nov. gen et spec.Trimerophytales aus dem Unterdevon der Eifel. Bonner Paläobotanische Mitteilungen, 12, 1–28.

Schweitzer, H. J. & Giesen, P. 1980. Über Taeniophyton inopinatum, Protolycopodites devonicus und Cladoxylon scoparium aus dem Mitteldevon von Wuppertal. Palaeontographica Abteilung B Paläophytologie, 173, 1–25. Serbet, R. & Rothwell, G. W. 1992. Characterizing the most primitive seed ferns. I. A reconstruction of Elkinsia polymorpha. International Journal of Plant Sciences, 153, 602–621.

226

Serlin, B. S. & Banks, H. P. 1978. Morphology and Anatomy of Aneurophyton, A Progymnosperm from the Late Devonian of New York. Palaeontographica Americana, 8, 343–359.

Shute, C. & Edwards, D. 1989. A new rhyniopsid with novel sporangium organization from the Lower Devonian of South Wales. Botanical Journal of the Linnean Society, 100, 111–137.

Skog, J. & Banks, H. 1973. Ibyka amphikoma, gen. et sp. n., a new protoarticulate precursor from the late Middle Devonian of New York State. American Journal of Botany, 60, 366–380.

Smith, G. O. & White, D. 1905. The geology of the Perry Basin in southeastern Maine. United States Geological Survey Professional Paper, 35, 1–92.

Snigirevskaya, I. S. 1984. Kornev'ie sistemy arkheopterisov'kh v verkhnem debone Donbassa (Root systems of archaeopterids from the Upper Devonian of Donbas). Ezegodnik Vsesusnogo Paleontologicheskogo Obsvesta Leningrad, 27, 28–41.

Soria, A. & Meyer-Berthaud, B. 2003. Occurrence of whorled organotaxis in the cladoxylopsid Pietzschia polyupsilon Read and Campbell Lower Carboniferous, USA. Review of Palaeobotany and Palynology, 124, 29–49.

Stein, W. 1982. The Devonian plant Reimannia, with a discussion of the class Progymnospermopsida. Palaeontology, 25, 605–622.

Stein , W. E. & Beck, C. B. 1983. Triloboxylon arnoldii from the Middle Devonian of western New York. Contributions from the Museum of Paleontology, the University of Michigan, 26, 257-288.

Stein, W. E. & Hueber, F. M. 1989. The anatomy of Pseudosporochnus: P. hueberi from the Devonian of New York. Review of palaeobotany and palynology, 60, 311–359.

Stein, W. E., Mannolini, F., Hernick, L. V., Landing, E. & Berry, C. M. 2007. Giant cladoxylopsid trees resolve the enigma of the Earth's earliest forest stumps at Gilboa. Nature, 446, 904–908.

Stockmans, F. 1939. Note préliminaire sur des empreintes végétales du Dévonien inférieur de la Belgique. Bulletin du Musée royal d'histoire naturelle de Belgique, 15, 1–6.

227

Stockmans, F. 1948. Vegetaux du Devonien Superieur de la Belgique. Musée royal d'histoire naturelle de Belgique, 110, 1–85.

Stockmans, F. 1968. Végétaux mésodévoniens récoltés aux confins du massif du Brabant: Belgique. Institut Royal Des Sciences Naturelles De Belgique, Memoire, 159, 3–49.

Stubblefield, S. & Banks, H. P. 1978. The cuticle of Drepanophycus spinaeformis, a long‐ ranging Devonian lycopod from New York and eastern Canada. American Journal of Botany, 65, 110–118.

Tanner, W. 1984. A fossil flora from the Beartooth Butte Formation of northern Wyoming. Ph.D. Dissertaton, Southern Illinois University, Carbondale.

Taylor, D. W. 1986. Anatomical and morphological study of a new species of Taeniocrada, a Devonian tracheophyte from New York State. Review of palaeobotany and palynology, 47, 63–87.

Taylor, E. L., Taylor, T. N. & Krings, M. 2009. Paleobotany: the biology and evolution of fossil plants, Elsevier, Amsterdam, Netherlands.

Toledo, S. 2018. A new anatomically-preserved plant from the lower Devonian of Quebec Canada: implications for Euphyllophyte phylogeny and early evolution of structural complexity. M.S. Thesis. Humboldt State University.

Tulipani, S., Grice, K., Greenwood, P. F., Haines, P. W., Sauer, P. E., Schimmelmann, A., Summons, R. E., Foster, C. B., Böttcher, M. E. & Playton, T. 2015. Changes of palaeoenvironmental conditions recorded in Late Devonian reef systems from the Canning Basin, Western Australia: A biomarker and stable isotope approach. Gondwana Research, 28, 1500–1515.

Volkova, I. B. 1994. Nature and composition of the Devonian coals of Russia. Energy & fuels, 8, 1489–1493.

Wang, Q., Geng, B. Y. & Dilcher, D. L. 2005. New perspective on the architecture of the Late Devonian arborescent lycopsid Leptophloeum rhombicum Leptophloeaceae. American Journal of Botany, 92, 83–91.

Wang, Q., Hao, S.-G., Wang, D.-M., Wang, Y. & Denk, T. 2003. A Late Devonian arborescent lycopsid Sublepidodendron songziense Chen emend.Sublepidodendraceae Kräusel et Weyland 1949 from China, with a revision of the genus Sublepidodendron Nathorst Hirmer 1927. Review of Palaeobotany and Palynology, 127, 269–305.

Wang, Y. 1993. First discovery of Eviostachya hoegii Stockmans from Wutung formation in China. Gǔshēngwùxué bào, 32, 430–440. 228

Wollenweber, J., Schwarzbauer, J., Littke, R., Wilkes, H., Armstroff, A. & Horsfield, B. 2006. Characterisation of non-extractable macromolecular organic matter in Palaeozoic coals. Palaeogeography, palaeoclimatology, palaeoecology, 240, 275–304.

Xu, H. H., Feng, J., Jiang, Q. & Wang, Y. 2013. Report of Drepanophycus Göppert L ycopsida from the Middle Devonian of Xinjiang, China. Journal of Systematics and Evolution, 51, 765–772.

Xue, J.-Z. & Basinger, J. F. 2016. Melvillipteris quadriseriata gen. et sp. nov., a new plant assigned to Rhacophytales from the Upper Devonian Famennian of Arctic Canada. Geological Magazine, 153, 601–617.

Xue, J. 2012. Lochkovian plants from the Xitun Formation of Yunnan, China, and their palaeophytogeographical significance. Geological Magazine, 149, 333–344.

Zdebska, D. 1982. A new zosterophyll from the Lower Devonian of Poland. Paleontology, 25, 256.

Zdebska, D. 1986. Psilophyton szaferi sp. nor. from the Lower Devonian of the Holy Cross Mountains. Poland. Acta Societatis Botanicorum Poloniae, 55, 315–324.

229

CHAPTER 5:

CONCLUSIONS

In this dissertation, I combined geochemical and paleontological approaches and reconstructed a series of changes in Late Devonian terrestrial and marine environments corresponding to the rise of the earliest forest. My results demonstrate that Devonian marine black shales deposited in shallow epicontinental seas along the Appalachian Basin serve as an invaluable archive of changes for both marine and terrestrial environments. In marine realms,

I present evidence that the early forest evolution served as a direct cause of severe marine anoxia that contributed to mass extinction. In terrestrial settings, I present evidence linking the earliest afforestation on the Earth to increased wildfire activities. The main results of this dissertation are summarized as follows.

In Chapter 2 (published in Scientific Reports: Lu et al., 2019), I reconstructed the afforestation pattern on the southern Acadian land during the Famennian based on the following evidence: (i) the occurrences of microfossils (wood fragments and spores) and plant-derived biomarkers in the Upper Devonian (Famennian stage) Chattanooga Shale in northern Alabama that indicate the presence of forests during the Famennian time on the southern Acadian land, and (ii) increasing values of vitrinite and inertinite content, TAR, retene, perylene and chemical weathering indices that demonstrate that terrestrial plants became an increasing source of organic matter during the deposition of the Gassaway

Member of the Chattanooga Shale. By combining geochemical data with a synthesis of 230

Devonian vascular plant fossil records in the Euramerica, I observed a more rapid southward dispersal of early forest in Euramerica through the Famennian than previously documented. I further hypothesized that the southern Appalachian Basin might serve as a bridge through which the earliest forest extended further into the southern American landmass of Gondwana as part of global-scale afforestation progressing from north to south during the Famennian.

Significantly, different from previous reconstruction based on fossil record in terrestrial sandstone and siltstone (e.g., Arnold, 1939; Matten, 1974; Scheckler, 1986, Cressler, 2006;

Morris et al., 2015), my study suggested that the widespread, yet under-utilized, unfossiliferous Devonian marine black shales may hold the key to reconstruct the process of the earliest afforestation.

In Chapter 3 (under review for Proceedings of the National Academy of Sciences, in

March, 2020), I found that the expanding terrestrial vegetation and establishment of early forests in the Euramerica by the late Frasnian might serve as an ultimate cause of the UKW extinction. Following the findings of Chapter 1, I put forward that that massive amounts of nutrients, caused by the spread of the earliest forest through enhanced weathering coupled with eustatic sea-level change, were washed into shallow marine and further contributed to eutrophication and widespread marine anoxia during the UKW extinction. Furthermore, I observed that marine euxinia (i.e., anoxia and sulfidic) occurred episodically, arguing the persistent marine anoxia in cycles mediated by astronomical forcing. During the UKW extinction, astronomic forcing served as a pacemaker led to fluctuations of sea level and then pulses of terrestrial inputs into marine, causing the development of marine anoxia at the beat of obliquity (~34 kyr) (Whalen, 2015). Such finding agrees well with the paleontological

231

view that the stepwise extinction during the F–F transition was a consequence of lack of origination attributed to unstable marine environments with recurring lethal conditions

(Bambach et al., 2004; Boyer et al., 2014). My findings are the first to demonstrate astronomically mediated cyclicity in not only marine anoxia but also contemporaneous terrestrial and oceanic environmental drivers that collectively led to recurring lethal environmental conditions during one of the big five biotic crises in the Earth’s history.

In Chapter 4 (under preparation for an invited submission to Palaeogeography,

Palaeoclimatology, Palaeoecology), I found that the diversification and expansion for the earliest forest led to spatiotemporal changes in fire activities through the Devonian. At first, I reconstructed a continuous temporal change of fire activities based on abundances of inertinite contents and pyrogenic PAHs preserved in one section of the Upper Devonian marine shales in the southern Euramerica. I observed consistently low abundances of wildfire proxies in the Frasnian interval yet a sharp increase through the Famennian interval, suggesting a rapid rise in fire activities across the F–F boundary in the Euramerica. To reconstruct the global change of fire activities through the Devonian, I combined my geochemical and petrographic data with a compilation of published occurrences of three wildfire proxies (i.e., fossil charcoal, inertinite macerals and combustion-derived PAHs), suggesting a global temporal rise of wildfire activities and rapid expansion across the

Euramerica during the Famennian. Furthermore, I investigated the theoretical linkage between fire and plant evolution during the Late Devonian by compiling and analyzing an extensive dataset of species of vascular plants. I found that the earliest trees, represented by

Archaeopteris, may serve as a major fuel causing a temporal rise in fire activities and a quick

232

southward dispersal of wildfires through the Euramerica in the Famennian. Additionally, I tested trends in entire plant size, indicated by maximum axial diameter, and leaf size, indicated by leaf maximum length and width. I found a dramatic rise in the tree body and size of leaves across the F–F boundary correlated with enhanced fire activities, suggesting that these traits may evolve as an adaption to high fire environments. Finally, the expansion of fires during the Late Devonian would also have effects on marine ecosystem over varying durations. By combining the results of the Chapter 3 and 4, I suggested that the rise of wildfires accompanied the expansion of the earliest forest would have led to increased terrestrial inputs into marine through the post-fire increased erosion and facilitated the development of marine anoxia in a short-term. Over a long-term, the expansion of wildfires would have caused an enhancement of carbon burial, which would, in turn, led to an increase of atmospheric O2 level and further terminated marine anoxia.

233

References Arnold, C. A. 1939. Observations on fossil plants from the Devonian of eastern North America. IV. Plant remains from the Catskill Delta deposits of northern Pennsylvania and southern New York. Contributions from Museum of Paleontology University of Michigan, 5, 271–314.

Bambach, R. K., Knoll, A. H. & Wang, S. C. 2004. Origination, extinction, and mass depletions of marine diversity. Paleobiology, 30, 522–542.

Boyer, D. L., Haddad, E. E. & Seeger, E. S. 2014. The last gasp: trace fossils track deoxygenation leading into the Frasnian–Famennian Extinction Event. Palaios, 29, 646– 651.

Cressler, W. L. 2006. Plant paleoecology of the Late Devonian Red Hill locality, north-central Pennsylvania, an Archaeopteris-dominated wetland plant community and early tetrapod site. Geological Society of America Special Papers, 399, 79–102.

Matten, L. C. 1974. The Givetian flora from Cairo, New York: Rhacophyton, Triloboxylon and Cladoxylon. Botanical Journal of the Linnean Society, 68, 303–318.

Scheckler, S. E. 1986. Floras of the Devonian-Mississippian transition. Studies in Geology, Notes for a Short Course, 15, 81–96.

Waltham, D. 2015. Milankovitch period uncertainties and their impact on cyclostratigraphy. Journal of Sedimentary Research, 85, 990–998.

234