The Pennsylvania State University

The Graduate School

Department of Geosciences

PRODUCTION AND PRESERVATION OF ORGANIC AND FIRE-DERIVED

CARBON ACROSS THE PALEOCENE-EOCENE THERMAL MAXIMUM

A Dissertation in

Geosciences and Biogeochemistry

Elizabeth H. Denis

 2016 Elizabeth H. Denis

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

August 2016

The dissertation of Elizabeth H. Denis was reviewed and approved* by the following:

Katherine H. Freeman Evan Pugh University Professor of Geosciences Dissertation Advisor Chair of Committee

Lee R. Kump Professor of Geosciences Head of the Department of Geosciences

Elizabeth A. Hajek Assistant Professor of Geosciences

Margot W. Kaye Associate Professor of Forest Ecology

Michael A. Arthur Professor of Geosciences Associate Head for Graduate Programs and Research in Geosciences

*Signatures are on file in the Graduate School

iii

ABSTRACT

The storage and release of organic carbon from the biosphere are influenced by temperature and precipitation through changes in plant productivity and in oxidative loss, such as fire and microbial respiration. The long-term fate of soil organic carbon during global warming is important because soil carbon is the largest terrestrial organic carbon reservoir and soil can serve as a sink or a source for atmospheric CO2. Soil carbon degradation is multifaceted as different pools of organic carbon in soils (e.g., fresh biomass, refractory soil organic matter, and thermally mature fossil organic matter) have different reactivity. Fire, an important component of ecosystems at a range of spatial and temporal scales, affects vegetation distribution, the carbon cycle, and climate. Because there are several variables and mechanisms are complex, it is difficult to predict future and infer past changes in both soil degradation and fire activity based on climate and environmental conditions. Examining changes in soil organic carbon, climate, and fire during past warming events, such as the Paleocene-Eocene Thermal Maximum (PETM), should help elucidate climate-carbon cycle relationships, especially effects that are expressed over long durations (e.g., 100 – 10,000 years).

Abrupt global warming during the PETM dramatically altered vegetation and hydrologic patterns, and, likely, terrestrial organic carbon production and preservation. The PETM coincided with a negative carbon isotope excursion (CIE), signifying a large release of 13C-depleted carbon to the biosphere and a major perturbation to the carbon cycle. Bulk organic carbon isotopes

13 13 (δ Corg) are often used to identify the CIE, but in terrestrial sections the δ Corg CIE can be highly

13 variable and distorted. It has been suggested that δ Corg values were highly variable because of soil carbon degradation by microbes and allochthonous (pre-PETM) fossil carbon inputs.

Constraining the degree and extent of degradation is critical in identifying the 13C-depleted carbon source and understanding carbon cycling processes and possible underlying organic

iii

iv carbon destabilization mechanisms during the PETM.

At three Paleocene-Eocene fluvial sites in the western USA, my co-authors and I test the hypothesis that there were increased degradation (soil carbon loss) and refractory (allochthonous) carbon inputs during the PETM. Clay minerals stabilize organic carbon, but we hypothesize decreased clay content and changes in mineralogy destabilized organic carbon during the PETM.

If soil moisture was a control on soil organic carbon degradation, then sites with similar soil moistirue conditions would have a similar loss of organic carbon. Using polycyclic aromatic hydrocarbons (PAHs), combustion byproducts that are relatively resistant to degradation, as a proxy for intermediate refractory carbon helped to discern the relative preservation of different carbon pools in the soils. I developed a novel molecular metric of degradation by calculating the percent loss of PAHs relative to total organic carbon (TOC) to estimate the extent of organic carbon loss and proportion of refractory allochthonous carbon during the PETM.

All forms of soil carbon decreased during the PETM, and PAH concentrations decreased even more than TOC, which suggests a more refractory phase was present, such as allochthonous fossil carbon. Positive correlations between elemental oxide weight percents (e.g., Al2O3 and

TiO2) and TOC suggests organic carbon preservation was associated with clay minerals. Wetter sites had a greater percent loss of organic carbon during the PETM than drier sites. Reduced soil organic matter preservation during the PETM was due to a combination of increased temperatures

(which increased microbial decomposition rates), decreased clay content and changes in mineralogy (which inhibited stability of fresh carbon), and fluctuations in soil moisture (which destabilized older, refractory carbon). Soil carbon degradation, even of intermediately refractory carbon, was not just a local phenomenon and was regional, and potentially global, in scope.

In the marine sediments of the Arctic, where organic carbon was well-preserved during the PETM, we used PAHs as an indicator for fire and plant biomarkers, as well as published

v pollen data, to decipher the dynamics between fire, precipitation, and vegetation changes in the paleoecosystem. In modern ecosystems, climate influences fuel availability (e.g., vegetation), fuel flammability (e.g., precipitation and temperature), and ignitions (i.e., lightning). In the paleorecord, authors often invoke drier conditions as a cause of increased fire occurrence. During the PETM, Arctic sediments exhibit higher PAH concentrations, and they both increased relative to plant input and tracked the increase in angiosperms (inferred from plant biomarker ratios and pollen). Our results suggest wetter conditions, followed by increased temperature, favored angiosperms and enhanced fire occurrence. Like modern fire dynamics, shifts in past fire patterns reflect a balance of variability in precipitation and sufficiently flammable vegetation. Increased fire in a wetter Arctic suggests PETM precipitation was seasonal, or variable on a longer timescale, and that hotter temperatures and angiosperm-dominated forests further facilitated burning.

Overall, we used PAHs as a primary signal of production (i.e., fire occurrence) in marine sediments and as a secondary signal of preservation (e.g., organic carbon degradation) in ancient soils. Our results highlight that terrestrial organic carbon was better preserved in the marine section than the fluvial sections. Increased temperatures, decreased clay content, changes in mineralogy, and variations in soil moisture destabilized carbon on millennial timescales and, with sustained higher temperatures across the PETM (~150 thousand years), increased soil carbon degradation persisted for tens of thousands of years. As temperatures warmed and remained warmer than the Paleocene, soils served as a sustained source of CO2 to the atmosphere rather than a sink. Although CO2 released from microbial respiration enhanced the greenhouse warming, increased organic carbon preservation in the marine realm may have counteracted the increased carbon output from soils.

TABLE OF CONTENTS

List of Figures ...... ix

List of Tables ...... xii

Acknowledgements ...... xiv

Chapter 1 Introduction ...... 1

1.1. Importance of soil in the terrestrial carbon cycle ...... 1 1.2. Mechanisms responsible for soil organic carbon stabilization ...... 2 1.3. Potential environmental effects that could destabilize organic carbon ...... 4 1.4. Environmental change during the Paleocene-Eocene Thermal Maximum (PETM) ...... 7 1.5. Carbon isotope excursion marks the PETM ...... 7 1.6. Observed changes in terrestrial organic carbon ...... 9 1.7. Influences on the production and preservation of soil organic carbon ...... 10 1.7.1. Vegetation changes during the PETM ...... 10 1.7.2. Hydrologic changes during the PETM ...... 11 1.8. Pyrogenic carbon as a molecular tool ...... 12 1.9. Study objectives and chapter outlines ...... 14 1.10. Anticipated publications from this work ...... 16 1.11. Figures ...... 18 1.12. References ...... 24

Chapter 2 Decreased soil carbon in a warming world: Degraded pyrogenic carbon during the Paleocene-Eocene Thermal Maximum ...... 33

2.1. Abstract ...... 33 2.2. Introduction ...... 34 2.3. Study Sections ...... 38 2.4. Methods ...... 38 2.5. Results ...... 40 2.5.1. Basin Substation ...... 40 2.5.2. Polecat Bench ...... 40 2.6. Discussion ...... 41 2.7. Conclusion ...... 45 2.8. Figures ...... 45 2.9. References ...... 51

Chapter 3 Widespread soil carbon loss during the Paleocene-Eocene Thermal Maximum .... 57

3.1. Abstract ...... 57 3.2. Introduction ...... 58 3.3. Study Site ...... 64 vi

vii

3.4. Methods ...... 65 3.5. Results ...... 68 3.6. Discussion ...... 69 3.6.1. Organic carbon degradation and the refractory carbon component ...... 69 3.6.2. Degradation and the carbon isotope excursion ...... 71 3.6.3. Mechanisms that stabilized and destabilized organic carbon ...... 72 3.6.4. Implications ...... 75 3.7. Conclusion ...... 76 3.8. Figures ...... 77 3.9. References ...... 82

Chapter 4 Fire and ecosystem change in the Arctic across the Paleocene-Eocene Thermal Maximum ...... 89

4.1. Abstract ...... 89 4.2. Introduction ...... 90 4.3. Study Section ...... 95 4.4. Methods ...... 96 4.4.1. Samples ...... 96 4.4.2. Extraction and analysis ...... 96 4.4.3. Normalized plant biomarker abundance to terrestrial organic carbon inputs ...... 97 4.5. Results ...... 98 4.6. Discussion ...... 99 4.6.1. Relationship between PAHs and plant biomarkers ...... 99 4.6.2. Plant biomarkers and pollen: Percent of angiosperms relative to gymnosperms ...... 101 4.6.3. Terrestrial plant inputs ...... 102 4.7. Conclusions ...... 106 4.8. Figures ...... 107 4.9. References ...... 112

Chapter 5 Summary and Future Work ...... 118

5.1. Research Summary ...... 118 5.2. Future Work ...... 121 5.3. References ...... 123

Appendix A Chapter 2 Supplemental Material ...... 125

Study Section ...... 125 Basin Substation PETM Stratigraphy ...... 125 Methods ...... 126 PETM Sedimentation Rate Estimate ...... 128 Figures ...... 129 Tables ...... 134 References ...... 140

Appendix B Chapter 3 Supplemental Material ...... 142 vii

viii

Supplemental Text ...... 142 Figures ...... 143 Supplemental Table ...... 143 Data Tables ...... 145 References ...... 152

Appendix C Chapter 4 Supplemental Material ...... 153

Figures ...... 153 Data Tables ...... 155

Appendix D Carbon isotopes of polycyclic aromatic hydrocarbons (PAHs) ...... 164

Carbon isotopes of polycyclic aromatic hydrocarbons (PAH) to evaluate pyrogenic carbon source ...... 164 References ...... 165

Appendix E Colombia and Venezuala (Mar2x core) ...... 166

Introduction ...... 166 Methods ...... 166 Results ...... 167 Discussion ...... 167 Figures ...... 168 Data Tables ...... 170 References ...... 174

Appendix F Tanzania ...... 176

Introduction ...... 176 Methods ...... 176 Discussion ...... 177 Figures ...... 178 Data Tables ...... 179 References ...... 183

Appendix G Model of how increased soil respiration affects carbon reservoirs ...... 184

Model calculations ...... 184 Conditions and variables ...... 184 Assumptions ...... 185 Discussion ...... 185 Figures ...... 186 References ...... 187

viii

LIST OF FIGURES

Figure 1-1. A schematic of the terrestrial carbon cycle...... 18

Figure 1-2. Conceptual model of organic matter stabilization and degradation...... 19

Figure 1-3. Conceptual model of a soil particle with organic-mineral interactions...... 20

Figure 1-4. Model of organo-mineral interactions...... 21

Figure 1-5. Measurements of the negative carbon isotope excursion (CIE) in terrestrial sections at the Paleocene-Eocene Thermal Maximum...... 22

Figure 1-6. The carbon isotope excursion (CIE) magnitude of both marine (blue) and terrestrial (green) carbon archives is shown in box-and-whisker plots...... 23

Figure 1-7. Molecular structures of a selection of polycyclic aromatic hydrocarbons (PAHs)...... 24

Figure 2-1. Map of the Bighorn Basin, Wyoming...... 46

Figure 2-2. Core records of the PETM, Basin Substation, Wyoming...... 47

Figure 2-3. Core records of the PETM, Polecat Bench, Wyoming...... 48

Figure 2-4. Box and whisker plots of PAH data (pyrene and coronene) from Polecat Bench and Basin Substation cores...... 49

Figure 3-1. Map of study site modified from Foreman et al., (2012) with PCBa section marked (red star)...... 78

Figure 3-2. Straitigraphic profile of geochemical parameters...... 79

Figure 3-3. Box and whisker plots of total organic carbon (TOC) and total PAH by lithology...... 80

Figure 3-5. Total organic carbon correlations (TOC) with TiO2 Al2O3, K2O, Na2O, and SiO2...... 82

Figure 4-1. Location of IODP Hole 302-4A (star) in the paleogeographical context of the late Paleocene-early Eocene (from Weijers et al., 2007)...... 107 ix

Figure 4-2. Depth profile of geochemical data...... 108

Figure 4-3. Depth profile of molecular compounds and pollen...... 109

Figure 4-4. A) Cross plot of Benzofluoranthenes (BF) with Simonellite; B) Cross plot of BF with β-amyrin derivative...... 110

Figure 4-5. Cross plot of Simonellite with β-amyrin derivative...... 111

Figure 4-6. Correlation of percent of angiosperms (%Angiosperms) with temperature...... 111

Figure A-1. Bivariate plot of PAH solubility in water and temperature modified from and using data from May et al., (1978)...... 129

Figure A-2. Box and whisker plots of TOC data from Polecat Bench and Basin Substation cores and compiled TOC data from terrestrial PETM sites published in literature...... 130

Figure A-3. Cross plot of PAH concentrations and TOC from Basin Substation...... 131

Figure A-4. Cross plots of coronene concentrations and TOC from Basin Substation and Polecat Bench cores in terms of lithology...... 132

Figure A-5. Depth profile of the ratio of 6-ring PAH (coronene) to 4-ring PAH (pyrene) from Basin Substation and Polecat Bench...... 133

Figure B-1. Depth profile of the ratio of 6-ring PAH (coronene) to 4-ring PAH (pyrene) from Basin Substation, Polecat Bench, and Piceance Basin...... 143

Figure C-1. Depth profile from IODP Hole 302-4A of carbon isotope values, total organic carbon (TOC), and Coronene/Pyrene+Coronene ratio...... 153

Figure C-2. Cross plot of Coronene/Pyrene+Coronene ratio and %TOC...... 154

Figure C-3. Relative percentage of terrestrial palynomorphs by type: spores (square); angiosperms (black diamond), gymnosperms (white triangle) (data from Sluijs et al., (2006))...... 155

Figure E-1. Geographical map of study location in Colombia (Mar 2x core, yellow star) from Jaramillo et al. (2010)...... 168

Figure E-2. Depth profile of total organic carbon (TOC) and polycyclic aromatic hydrocarbons (PAHs) concentrations by ring size...... 169

Figure E-3. Depth profile of total organic carbon (TOC) and polycyclic aromatic hydrocarbon (PAHs) concentrations...... 169

Figure F-1. Geographical map of Tanzania study site (yellow star) from Handley et al. (2008)...... 178

Figure G-1. Schematic of box model...... 186

Figure G-2. Plot of modeled carbon in the soil reservoir with time...... 186

Figure G-3. Plot of modeled temperature with time...... 187

xii

LIST OF TABLES

Table A-1. Basin Substation percent loss calculations and allochthonous (fossil) carbon estimate...... 134

Table A-2. Polecat Bench percent loss calculations and allochthonous (fossil) carbon estimate...... 135

Table A-3. Grams of sediment extracted and PAH abundanes (μg/g TOC) for Basin Substation samples...... 136

Table A-4. Additional PAH abundanes (μg/g TOC) for Basin Substation samples...... 137

Table A-5. Basin Substation samples' lithology, grain size, and color from BBCP Science Team...... 138

Table A-6. δ13Corg, TOC, PAH values and dry weight of grams extracted for Polecat Bench samples...... 139

Table B-1. Fossil (allochthonous) carbon calculations: Piceance Basin, median values reported...... 144

Table B-2. δ13Corg and TOC values and dry weight of grams extracted for Piceance Basin samples...... 145

Table B-3. PAH concentrations (ng/g TOC) for all Piceance Basin samples ...... 147

Table B-4. Field descriptions for all samples collected in 2013 (PCBa-ED) ...... 149

Table B-5. Piceance Basin XRF elemental data from Foreman (2012)...... 151

Table C-1. Arctic TOCterrestrial and uncertainty with depth ...... 156

Table C-2. PAH/plant biomarker ratios with depth...... 157

Table C-3. Coronene/Pyrene ratio, total organic carbon (TOC), plant biomarker (ng/g TOCterrestrial) and pollen (number/g TOCterrestrial) concentrations...... 159

Table C-4. Arctic Pristane/Phytane ratios...... 161

Table C-5. Percent of angiosperms based on biomarkers and on pollen...... 162

Table E-1. Mar2x samples: grams of sediment extracted, TOC, and carbon isotope values...... 170

Table E-2. Mar2x sample lithology information...... 171

Table E-3. Mar2x PAH concentrations in μg/g TOC...... 172

xii

xiii

Table E-4. Additional Mar2x PAH concentrations in μg/g TOC ...... 173

Table F-1. Tanzania PAH concentrations (pg/μl injected onto GC-MS) ...... 179

Table F-2. Additional Tanzania PAH concentrations (pg/μl injected onto GC-MS) ...... 180

Table F-3. Tanzania sample depth, extracted weight, PAH ratios...... 181

Table F-4. PAH concentrations ng/g dry sediment...... 182

xiii

xiv

ACKNOWLEDGEMENTS

I am grateful for each person who has supported me to this day and forward. I can hardly begin to touch on all the small moments that came together to make this Ph.D. possible, and my life better, but I would like to acknowledge some highlights and key components.

Thank you to my wonderful advisor, Kate Freeman. I have learned, and continue to learn, so much from you. I appreciate all the time and effort you have put forth. In talking with you, I usually leave feeling better than when I started and on the occasion as necessary, you have helpfully grounded me, to remind me of priorities and what is realistic. You have been incredibility supportive and understanding with work and personally. I am glad to have you as my mentor and teacher.

Thank you to my committee members: Lee Kump (co-advisor), Liz Hajek, Margot Kaye, and previously, Jennifer Balch. I’ve appreciated your guidance and our insightful discussions. To

Chris House, thank you for your friendly discussion and interest in my work and beyond.

To Denny Walizer, thank you for so much help and support in the lab, with the instruments, and anything else technical that might need a super fix. And for all those times I asked you to walk over, since your presence itself often could make a problem disappear.

The work in this dissertation was made possible by many funding sources, most notably the Bighorn Basing Coring Project (NSF Grant No. EAR0958951 to K. Freeman) and the NSF

Graduate Research Fellowship Program (NSF Grant No. DGE1255832 to E. Denis). Additional support for E. Denis was provided by Shell Research Facilitation Award, Biogeochemistry

Research Dual-Degree Program Award, Charles E. Knopf, Sr., Memorial Scholarship, Hiroshi and Koya Ohmoto Graduate Fellowship, Geological Society of America (GSA) Student Research

Grant, American Geosciences Institute (AGI) Harriet Evelyn Wallace Fellowship, GSA

xiv

Northeastern Section Student Travel Grant, Paul D. Krynine Scholarship, ConocoPhillips

Graduate Student Fellowship, and Pennsylvania Space Grant Graduate Student Fellowship.

I am lucky to have many great collaborators who have helped from sample collection to shared data to many conversations. Chapter 2: Bianca Maibauer and Gabe Bowen (University of

Utah), Allie Baczynski (Penn State), Francesca McInerney (University of Adelaide), Phil Jardine

(Open University), Guy Harrington (University of Birmingham), Margaret Collinson (Royal

Holloway University of London), Claire Belcher (University of Exeter), Scott Wing (Smithsonian

Institution), and the Bighorn Basin Coring Project (BBCP) Science Team. Chapter 3: Brady

Foreman (Western Washington University). Chapter 4: Nikolai Pedentchouk (University of East

Anglia), Stefan Schouten (Royal Netherlands Institute for Search (NIOZ), Mark Pagani (Yale

University). Tanzania work: Marcus Badger and Rich Pancost (Bristol University). Thank you to

Laurie Eccles for all your help in lab and being so diligent and careful with my precious samples.

Thank you to Flo Ling and Peter Heaney for assisting me with a few test-run XRD analyses.

Thank you to all my labmates through the years. To those ‘elder’ grad students, thank you for all the knowledge, techniques, and methods you passed down, and for your patience as I learned and asked many questions: Heather Graham, Clay Magill, Kat Dawson, and Aaron

Diefendorf. To the post-docs that I could look up to and could count on for help and care: Allie

Baczynski, Sara Lincoln, Anna Henderson. To Laurence Bird, thank you for lab tips usually when it was most needed and for minding Research West while I minded the Deike lab and the GC-MS.

To the ‘younger’ grad students who kept me at the top of my game (and bolstered me with energy when I lacked it), thank you for sharing your ideas, asking questions, and for listening: Christine

Doman, Angela Chung, Laura Herren, Shelby Lyons, and Allison Karp. Not unlike labmates, my officemates through the years have made the work atmosphere more pleasant – thank you for being up for sharing a laugh at any moment and for many helpful conversations and inspirations.

xvi

Near and dear to my heart and mind are my friends and family. Your endless support has helped me in so many ways through the challenges and the excitement of grad school. A special thanks to Emily Woodward, Rosie Oakes, Leah Brandt, Ellen Chamberlin, Erica Pitcavage, Laura

Herren, Helen Gall, Julie Weitzman, Shauna Rainford, Allison Karp, Peter Ilhardt, Christine

Doman, Paul Grieve, Megan Carter, Joe Orlando, Ashlee Dere, Bob Nasuti, Emily Doyle and all my activity buddies (e.g., soccer, tennis, and the Penn State women’s club ice hockey team).

Thank you to my family for all your love and support: my parents Jan and Clyde and my brother Michael.

xvi

xvii

Have you heard the one about the Graduate Student, Computer, Gas Chromatograph (GC), and Mass Spectrometer in a lab?

What did the computer say to the GC and Mass Spec? “I detect some chemistry.” What did the Grad Student say to the Computer? “I’ll fix them up.” What did the Mass Spec say to the GC? “You’re so hot.” What did the GC say to the Mass Spec? “I want to separate.” What did the Mass Spec say to the GC? “Breaking up is hard to do.” What did the GC say to the Mass Spec? “I missed that, the column broke.” What did the Computer say to the GC and Mass Spec? “We have disconnected.” What did the Grad Student say to the Computer, GC, and Mass Spec? “PLEASE can you just work together.”

What did the advisor say to the grad student? “Stop making jokes and finish your dissertation.”

---Written by Elizabeth Denis; inspired by Christine Brandt; developed with the aid of Emily Woodward, Valerie Woodward, and Leah Brandt---

xvii

Chapter 1

Introduction

1.1. Importance of soil in the terrestrial carbon cycle

The main reservoirs of carbon on the Earth’s surface are oceans, sediments, the atmosphere, biomass, and soil (Killops and Killops, 2005; Sundquist and Visser, 2005). Soil carbon is the largest terrestrial organic carbon reservoir (1,500 Pg C in the top 1 m; 840 Pg C in the top 1-3 m (Jobbágy and Jackson, 2000)), followed by vegetation (500 Pg C) (Sundquist and

Visser, 2005) (Figure 1-1). Litterfall (60 Pg C/yr) delivers biomass carbon to soil, where it is broken down by microbes and lost by soil respiration (59.6 Pg C/yr) and erosion (0.4 Pg C/yr)

(Sundquist and Visser, 2005). Most soil carbon surveys inventory a fixed soil depth of 1 m (1500

Pg C total), with ~50% of carbon located in the top 20 cm (Jobbágy and Jackson, 2000). The residence time for organic carbon in soil is ~25-40 years (Oades, 1988; Jobbágy and Jackson,

2000; Sundquist and Visser, 2005), yet a fraction (~0.1% (Oades, 1988; Hedges and Oades,

1997)) of organic carbon is stabilized in the soil profile on longer timescales (thousands to tens of thousands of years) (Oades, 1988; Sundquist and Visser, 2005; Chaopricha and Marín-Spiotta,

2014) and preserved in the paleorecord through burial. Climate influences the preservation of organic carbon, but the effects on 100 – 10,000 year timescales is unclear.

Understanding the effects of climate on soil carbon is important for agricultural soil fertility (Ramankutty et al., 2002) and for assessing if soil will be a source (e.g., thawing permafrost (Schuur et al., 2008)) or sink of carbon during warming (Lal, 2004). It has been

proposed that soils can sequester anthropogenic CO2 (Lal, 2004), enhanced by application of biochar, pyrolyzed biomass that is similar to charcoal (Lehmann et al., 2011), but the duration of

2 such storage (e.g., 100 to 10,000 years) is uncertain. Warming could strengthen the oxidation and loss of organic carbon, by fire or soil respiration, which, in turn, would amplify warming through release of carbon to the atmosphere.

The long-term fate of soil organic carbon during global warming is difficult to predict.

Different pools of organic carbon in soils (e.g., fresh biomass and old, refractory, “stable” carbon) have different reactivity, but all are sensitive to temperature (Knorr et al., 2005). Both temperature and water availability influence the production and preservation of terrestrial organic carbon in the biosphere. The main source of carbon to soils, plant matter, increases with rainfall

(Huxman et al., 2004). The main mechanism of carbon loss from soils, microbial respiration, increases with temperature (Lloyd and Taylor, 1994; Knorr et al., 2005; Luo, 2007; Carvalhais et al., 2014; Karhu et al., 2014), but it also is strongly affected by soil moisture and oxygen concentrations (Leirós et al., 1999; Krull et al., 2003).

1.2. Mechanisms responsible for soil organic carbon stabilization

For organic matter to accumulate in soils, organic matter inputs must exceed loss by decomposition. Because organic matter is inherently unstable on the order of days to tens of years, organic accumulation requires some form of protection from degradation (Lorenz et al.,

2007). Soil organic carbon stability is influenced by environmental factors such as temperature and soil moisture, the chemical characteristics of the compounds, the accessibility of the compounds to microorganisms, and the characteristics of the minerals in the soil (e.g., the amount and type of clay minerals present) (Von Lützow et al., 2006; Schmidt et al., 2011) (Figure 1-2).

The three main types of protection mechanisms that inhibit organic carbon degradation in soils are the inherent chemical recalcitrance of the compound, the accessibility of the compound to degradation, and physical protection through interactions between other organic compounds and

3 minerals (Von Lützow et al., 2006; Von Lützow et al., 2008; Marschner et al., 2008)

Molecular recalcitrance is the resistance of organic matter to degradation due to the structure, chemical, and physical properties of the molecule (Von Lützow et al., 2006; Schmidt et al., 2011). For example, large fire-derived compounds with aromatic ring structures tend to be more stable in soil than other organic matter (Von Lützow et al., 2006; Schmidt et al., 2011).

Compounds that are small and water soluble are more available to organisms and degrading enzymes, whereas hydrophobic compounds are less accessible (Von Lützow et al., 2006; Kleber,

2010; Schmidt et al., 2011). Selective preservation based on inherent chemical compound characteristics is particularly important on the initial phase of degradation (e.g., years), but persistence on longer timescales (hundreds to tens of thousands of years) is not just due to structural characteristics but other stabilization mechanisms (Von Lützow et al., 2006; Lawrence et al., 2015).

Spatial inaccessibility of organic matter to microorganisms, enzymes, water, and oxygen inhibits carbon degradation. Within the soil, preferential flow paths of water can permit better nutrient and substrate supply compared to the whole soil matrix, especially as depth in the soil profile increases (on the order of meters) (Von Lützow et al., 2006). Organic matter forms aggregates with soil particles, particular with Fe- and Al- oxides or hydroxides (Figure 1-3). Soil biota, through microbial cells and excretions, can help bind particles and organic matter together.

Aggregation reduces the access of microorganisms and enzymes to the organic carbon and reduces diffusion of enzymes and oxygen (Von Lützow et al., 2006). When soil aggregates are disrupted by bioturbation, erosion, or water, organic matter can be more available for degradation

(Von Lützow et al., 2006; Schmidt et al., 2011).

Organic matter is stabilized by interaction with mineral surfaces and metal ions. Finer grained particles (e.g., clays) have more surface area for organic matter to attach to (Christensen,

2001; Von Lützow et al., 2006; Von Lützow et al., 2008) (Figure 1-4). In the presence of

4 polyvalent cations like Ca+2 and Mg+2 in neutral and basic soils and Al+3 and Fe+3 in acidic soils, organic compounds can be connected by cation bridges (Oades, 1988; Von Lützow et al., 2006;

Von Lützow et al., 2008). This helps form larger molecules that are insoluble and not accessible to microorganisms (Oades, 1988; Von Lützow et al., 2006; Von Lützow et al., 2008). It is estimated that ~90% of the organic matter preserved in marine sediments is associated with mineral surfaces (Keil et al., 1994), which implies that organic-mineral interactions can persist on

100 – 10,000 year timescales (Chaopricha and Marín-Spiotta, 2014), or longer.

Over thousands of years, sediments accumulate and bury the soil, which inhibits the degradation of organic matter by microbes and facilitates the preservation of the residual organic matter in the geologic record for millions of years. The higher the sedimentation rate, the less time organic matter is exposed to microbial degradation and the more likely it is preserved (Ibach,

1982). As sedimentary layers accumulate, bottom layers get compacted by the increasing pressure with burial, water content decreases, and minerals precipitate (e.g., calcite) to cement the sediment together as part of the process of lithification (i.e., conversion of unconsolidated sediments into a solid body of rock) (Killops and Killops, 2005). With burial, increasing temperatures restrict biological activity and the chemical bonds of the organic matter remain intact as long as temperatures are below (~60°C), usually a relatively shallow burial depth (<1 km) (Killops and Killops, 2005). Thermal maturation (higher temperatures) can further alter organic matter during burial to phases, such as kerogen, that are resilient to weathering and microbial degradation over millions of years.

1.3. Potential environmental effects that could destabilize organic carbon

Laboratory incubations and field experiments over days to years in agricultural and forest soils indicated that additions of fresh biomass (or substrates like glucose and carbohydrates) can

5 increase microbial activity (called the priming effect) and induce microbial decomposition of older, more stable carbon (Fontaine et al., 2003; Marschner et al., 2008; Guenet et al., 2012; Don et al., 2013). These authors proposed that additional substrate enhanced microbial growth and enzyme production, which expanded the ability or likelihood that some microorganisms metabolized less energy favorable compounds, especially in a nutrient limited system (Fontaine et al., 2003; Marschner et al., 2008; Guenet et al., 2012; Don et al., 2013). The priming effect may only be a short-term effect (e.g., years), but it can facilitate the degradation of older carbon (e.g.,

100 – 1,000 years age) stored in soils (Kuzyakov et al., 2000; Stockmann et al., 2013). Although the priming effect could explain an initial burst in microbial metabolism and decomposition of older carbon, sustained increased inputs of fresh biomass on centennial or millennial scales would likely result in a new steady state and ameliorate the effect of priming.

A shift in the microbial community structure can affect the type of compounds that are degraded and the extent of degradation (Schmidt et al., 2011; Pisani et al., 2016). For example, the exclusion of roots from mineral soils during a 20-year forest plot study led to a microbial shift toward fungi and increased actinomycetes, filamentous bacteria associated with the decomposition of more refractory compounds (Pisani et al., 2016). Other conditions that result in shifts in the microbial composition are changes in available nutrients, such as nitrogen (Agren et al., 2001; Schmidt et al., 2011).

Incubation experiments on grassland soils demonstrated that wet/dry cycles (on the order of months) increased microbial respiration 6-fold (Xiang et al., 2008). However, when the soils were kept constantly moist, respiration was low, except when the soils were dried and rewet, which was possibly an indication that respiration was limited based on the ability of microbes to access the organic carbon (Xiang et al., 2008). A follow-up study that tracked the age of the carbon respired using radiocarbon dating observed that wet/dry cycles not only facilitated total respiration, but that older carbon (100+ years) was degraded (Schimel et al., 2011). The drying

6 and rewetting of soils can destabilize the organic matter association with clay minerals (Oades,

1988). As clay mineral structures expand and contract with the addition and loss of water, organic matter desorbs from the clays, and becomes more accessible to microorganisms (Oades, 1988;

Schimel et al., 2011).

The extent of organic carbon stabilized is linked to the grain size and mineral composition of the soil matrix. Clays, including aluminosilicates and Al- and Fe- oxides, hydroxides, and oxyhydroxides, provide the majority of sorbent surface area in soil (Oades, 1988;

Sollins et al., 1996). Clay-size grains have greater surface area for organic-mineral interactions as opposed to sands (Mayer, 1994). In addition, the type of clay minerals present can affect the organic matter composition (Sollins et al., 1996; Von Lützow et al., 2006). On hydroxylated surfaces, the net surface charge varies depending on pH, while on smectites, illites, and other silicate layered clays, permanent charges develops during ion substitution within the crystal structure of the clay (Sollins et al., 1996). Soils with a higher clay content retain higher amounts of organic carbon, but it is unclear how differences in clay mineralogy affect organic carbon preservation (Oades, 1988; Sollins et al., 1996; Von Lützow et al., 2008).

Modern incubation and field studies of the effect of organic carbon inputs, temperature, and precipitation on soil organic carbon preservation are on the order of days to years and, rarely, tens of years (Pisani et al., 2016). Trends over longer time scales (100+ years) must be extrapolated from these studies, but the mechanisms observed in the modern may not operate on these longer timescales. With this in mind, studying past global warming events, such as the

Paleocene-Eocene Thermal Maximum (PETM), which was ~170 ky (Röhl et al., 2007;

McInerney and Wing, 2011), provides information about perturbations in carbon preservation on scales of thousands to tens of thousands of years. I hypothesize that decreased clay content in the soils (which inhibited stability of fresh carbon) and fluctuations in soil moisture (which destabilized older, refractory carbon), in conjunction with increased temperatures (which

7 increased microbial decomposition rate), contributed to a reduction in soil organic matter preservation during the PETM.

1.4. Environmental change during the Paleocene-Eocene Thermal Maximum (PETM)

The PETM was a geologically abrupt period of climate change that occurred about 55.5 million years ago (Westerhold & Röhl, 2012). This period is well studied and widely invoked as a geologic analog for modern CO2-driven climate change, even though modern climate change may be unprecedented in terms of carbon release rate (modern rate (~10 Pg C/yr) is 10 times faster than the PETM) (Cui et al., 2011; Zeebe and Zachos, 2013; Zeebe et al., 2016)(Cui et al., 2011;

Zeebe and Zachos, 2013; Zeebe et al., 2016). The hyperthermal event is marked by a negative carbon isotope excursion (CIE), signifying a major perturbation to the carbon cycle (McInerney and Wing, 2011). At least 3,000 Pg of 13C-depleted carbon were released into the atmosphere over ~10 ky and global temperatures rose ~5-8°C over ~170 ky (Wing et al., 2005; Zachos et al.,

2005; Röhl et al., 2007; Secord et al., 2010; Cui et al., 2011; McInerney and Wing, 2011).

1.5. Carbon isotope excursion marks the PETM

Based on a global compilation of isotope data, the carbon isotope excursion (CIE) had a generalized shape with three phases: a rapid “onset” of the excursion, a “body” of continual negative carbon isotope composition, and a “recovery” when the reservoir isotopic composition returned to pre-PETM values (Bowen et al., 2006; McInerney and Wing, 2011). The CIE onset was ~8-23 ky based on astronomical cyclostratigraphy and calculated sedimentation rates in terrestrial sections (Magioncalda et al., 2004; Aziz et al., 2008), although estimates have ranged from 13 years, based on carbon isotopes within 13 sedimentary layers at an Atlantic coastal plain

8 site (Wright and Schaller, 2013), to 6-35 ky based on astronomical cyclostratigraphy, extraterrestrial 3He fluxes, and modeling studies (Kennett and Stott, 1991; Bowen et al., 2006;

Murphy et al., 2010; Cui et al., 2011). Based on similar evidence, most authors estimated the body of the CIE was 80-160 ky (Norris and Röhl, 1999; Farley and Eltgroth, 2003; Röhl et al.,

2007; Aziz et al., 2008; Westerhold et al., 2009; Murphy et al., 2010), and the recovery to pre-

PETM values occurred within ~35 kyr according to 3He fluxes (Murphy et al., 2010).

Multiple suggested sources for 13C-depleted carbon include methane clathrates, thermogenic methane, and burning of peat and coal. Although the exact source is unknown

(McInerney and Wing, 2011), most likely, a combination of sources and feedbacks contributed to the large release of CO2 to the atmosphere (Bowen et al., 2015). Further addition of CO2 to the atmosphere by fire or soil respiration could have enhanced the atmospheric greenhouse.

The magnitude of the CIE varies among carbon archives (e.,g, terrestrial records include

13 soil carbonate, plant lipids, tooth enamel, and bulk soil organic matter (δ Corg ), and marine records include benthic and planktonic foraminifera, bulk carbonate, bulk marine organic matter, and algal lipids (Figure 1-5 and Figure 1-6; McInerney and Wing, 2011). Generally in terrestrial

13 sections the δ Corg CIE is among the most attenuated, and records are often highly variable (in

13 terms of variation in δ Corg values) and distorted (in terms of shape and magnitude relative to

13 other δ Corg excursions), even on local scales (Baczynski et al., 2013; Maibauer, 2013; Baczynski et al., 2016). Multiple lines of evidence from the Bighorn Basin, Wyoming (one of the study areas for this dissertation) revealed both increased organic carbon degradation and a relatively greater

13 proportion of allochthonous (pre-PETM) carbon diminished the δ Corg excursion (Baczynski et al., 2013; Bataille et al., 2013; Cotton et al., 2015; Baczynski et al., 2016). This can cause truncation of the CIE and underestimation of the length of the PETM interval (Baczynski et al.,

2013), and it can even obscure the CIE all together (Maibauer, 2013). Constraining the magnitude

9 and duration of the CIE is critical to identify the 13C-depleted carbon source and to understand carbon cycling processes during the PETM.

1.6. Observed changes in terrestrial organic carbon

Terrestrial sediments during the PETM tended to have low total organic carbon (%TOC), with values reported on the order of 0.01% – 1% and lower during the event than the late

Paleocene or subsequent early Eocene (Magioncalda et al., 2004; Clechenko et al., 2007;

Domingo et al., 2009; Foreman et al., 2012). The reduced TOC could reflect either a decrease in production, or in preservation, or some combination of both. Geochemical evidence, from soil carbonates and leaf wax n-alkanes, from the western USA suggests there was increased soil respiration during the PETM, and hence, decreased organic carbon preservation (Cotton et al.,

2015; Baczynski et al., 2016). In contrast, terrestrial carbon models and paleoclimate simulations of the Paleocene-Eocene suggested an increase in the amount of terrestrial carbon storage during the PETM (Beerling, 2000). The models took into account increases in CO2 and temperature based on a variety of paleoevidence and modern-day estimates, but no adjustments were made to constrain precipitation since the global pattern of any change in precipitation during the PETM was uncertain. The simulated environmental change exhibited spatially heterogeneous increases in vegetation and soil carbon storage and the greatest response was observed in areas that received the greatest amount of precipitation (Beerling, 2000).

10 1.7. Influences on the production and preservation of soil organic carbon

1.7.1. Vegetation changes during the PETM

During the PETM, vegetation and precipitation patterns were dramatically altered globally (Wing et al., 2005; Pagani et al., 2006b; Kraus and Riggins, 2007; Smith et al., 2007;

Smith et al., 2008; Jaramillo et al., 2010; Foreman et al., 2012; Handley et al., 2012; Wing and

Currano, 2013). For example, in the Bighorn Basin, Wyoming, where there has been extensive plant fossil research, flora shifted dramatically during the PETM (McInerney and Wing, 2011;

Wing and Currano, 2013). Based on plant fossil evidence, plants adapted to medium conditions of moisture, markedly conifers, decreased while thermophilic and dry-tolerant species, particularly

Fabaceae (legumes), surged in abundance (Wing and Currano, 2013). The PETM flora in the western USA was most similar to dry tropical forests. Floral response had regional variations, but were generally consistent with expansion of tropical flora to higher latitudes, such as was observed in the Bighorn Basin (Wing and Currano, 2013). There was little extinction during the event and after the event, most late Paleocene flora returned along with an addition of some immigrant species to the floral composition (Wing and Currano, 2013).

Modern CO2-enrichment studies (Nobel and Hartsock, 1986; Nobel and Cortazar, 1991;

Bowes, 1993) suggest plant biomass abundance could have increased during the high CO2 conditions of the PETM. In the studies, plants accumulated 30% more biomass when atmospheric

CO2 was doubled (pCO2 was increased from ~350 ppm to ~650 ppm) (Nobel and Hartsock,

1986; Nobel and Cortazar, 1991; Bowes, 1993). As a comparison for the PETM, these experiments are problematic for several reasons. In the experiments, other conditions, including moisture and temperature, were the same between the environmental chamber and the elevated-

CO2 chamber, yet both these conditions changed during the PETM. Precipitation conditions are

11 important to consider, too, because a modern global study of terrestrial ecosystems in multiple biomes observed that net primary productivity and standing biomass were proportional to mean annual precipitation (Huxman et al., 2004). Further, the CO2 fertilization studies were short-term experiments (< 2 years), observed limited vegetation types, and the baseline and elevated pCO2 were both lower than the estimated atmospheric pCO2 pre-PETM (600 – 2,800 ppm) and during the PETM (700 – 25,000 ppm) (Pagani et al., 2006a; McInerney and Wing, 2011). Despite the limitations of these experiments, they suggest that higher CO2 during the PETM may have driven increased plant biomass accumulation. Increased delivery of biomass to the soils may have helped organic carbon to accumulate in soils faster than it was decomposed or, aided by warmer temperatures, it may have enhanced microbial growth and facilitated the degradation of older carbon (e.g., 100 – 1,000 years age) stored in the soils (Fontaine et al., 2003; Marschner et al.,

2008; Guenet et al., 2012; Don et al., 2013)

1.7.2. Hydrologic changes during the PETM

Standing biomass, and thus net primary productivity, is proportional to rainfall (Huxman et al., 2004). Changes in precipitation can, therefore, affect both the production of organic carbon and as discussed earlier, the preservation of organic carbon (e.g., wet-dry cycles can destabilize soil carbon) (Oades, 1988; Xiang et al., 2008; Schimel et al., 2011). Increased atmospheric CO2 and corresponding global warming are expected to intensify the hydrologic cycle (Sloan et al.,

1995; Manabe, 1997; Bice and Marotzke, 2002). Increased CO2 in the atmosphere leads to warming of the troposphere, which results in increased absolute humidity in the air (Manabe,

1997). During the PETM, precipitation changes varied regionally and globally. For example, based on current interpretations, Colombia and Venezuela at low-latitudes were wetter or had no change in precipitation (Jaramillo et al., 2010) and Tanzania was drier (Handley et al., 2012).

12 However, there is conflicting evidence of drying or wetting in the tropics during the PETM from independent tests including animal physiology, plants, and molecular fossils (Huber, 2009;

Jaramillo et al., 2010). At mid-latitudes, Wyoming was drier or seasonally drier (Wing et al.,

2005; Kraus and Riggins, 2007; Smith et al., 2007; Smith et al., 2008) and Colorado was wetter or had more seasonal precipitation (Foreman, 2012; Foreman et al., 2012). At the onset of the

PETM there was transient drying in the Bighorn Basin, Wyoming based on fossil flora, trace fossils, and the chemical composition and physical morphology of paleosols (Wing et al., 2005;

Smith et al., 2008). At high-latitudes, the Arctic was wetter (Pagani et al., 2006b).

Pagani et al. (2006b) suggested more water was exported from the tropics to higher latitudes based on multiple lines of evidence. The isotopic composition of Arctic PETM precipitation was considerably deuterium-enriched compared to today, indicating reduced rainout along the poleward trajectory of tropical airmasses. Increased water supply to the Arctic was supported by the pervasiveness of low-salinity-tolerant dinocyst assemblages. If poleward water vapor transport increased, then precipitation likely decreased in the subtropics or mid-latitudes during the PETM (Pagani et al., 2006b). Globally, increased terrestrial kaolonite in marginal marine settings during the PETM suggest that there was increased run off and increased intensity and variability in rain events (Robert and Kennett, 1994; Crouch et al., 2003). How the changes in precipitation affected terrestrial productivity and organic matter storage is unclear, although modern terrestrial net primary productivity is correlated with mean annual precipitation (Huxman et al., 2004).

1.8. Pyrogenic carbon as a molecular tool

Pyrogenic carbon is a continuum of combustion products generated as solid residue or volatiles, ranging from slightly charred material to soot (Schmidt and Noack, 2000; Masiello,

13 2004; Bird et al., 2014). We use polycyclic aromatic hydrocarbons (PAHs) (Figure 1-7) as representative products of combustion and as a metric for degradation. Pyrogenic PAHs are produced during combustion as volatiles and in association with particles (Masclet et al., 1995) and are used as a marker for fire in the paleorecord (Ramdahl, 1983; Killops and Massoud, 1992;

Page et al., 1999; Yunker et al., 2002; Denis et al., 2012). Charcoal has a relatively long residence time in soils, estimated on the order of 500 – 10,000 years (Schmidt and Noack, 2000; Forbes et al., 2006; Von Lützow et al., 2006). Like charcoal (Schmidt and Noack, 2000), PAHs have an aromatic structure (Coulon et al., 2005; Johnsen et al., 2005; Von Lützow et al., 2006). PAH bioavailability and degradation rates vary with molecular size. For example, aqueous solubility of

PAHs decreases almost logarithmically with increasing molecular mass, and therefore, 5-7 ring

PAHs are significantly less soluble than 2-4 ring counterparts (May et al., 1978; Johnsen et al.,

2005). Because of the overall recalcitrant nature of the larger fire-derived compounds (Schmidt and Noack, 2000; Johnsen et al., 2005; Bird et al., 2014), we use PAHs with more than 4 rings as proxies for refractory carbon. Such refractory carbon is relatively resistant to microbial degradation on the order of hundreds to tens of thousands of years and can be preserved in the fossil record for millions of years.

Studies that used PAHs as an indicator for fire in the paleorecord are limited to the

Paleozoic and Mesozoic and focused on six time periods (often boundary events) - Late Devonian

(Marynowski and Filipiak, 2007; Kaiho et al., 2013), Pennsylvanian (Scott et al., 2010), Permian-

Triassic boundary (Shen et al., 2011), Triassic-Jurassic boundary (Jiang et al., 1998; Marynowski and Simoneit, 2009; van de Schootbrugge, 2010), Jurassic (Killops and Massoud, 1992;

Marynowski et al., 2011), and Cretaceous-Tertiary boundary (Venkatesan and Dahl, 1989;

Arinobu et al., 1999; Mita and Shimoyama, 1999). These paleorecord studies, of both continental and marine sediments, primarily use PAHs as a fire marker to reconstruct environmental conditions and relate fire occurrence to changes in climate or extinction events. Studies based on

14 PAHs in recent sediments (e.g., lakes, soils) are a useful resource. They give insights into to how well PAHs can record fire given the complexities of transport and preservation (Jiang et al., 2000;

Muri and Wakeham, 2009; Vergnoux et al., 2011; Denis et al., 2012). Based on modern studies, we know patterns are not simply a primary signal (e.g., of environment or fire), but also are a secondary signal impacted by conditions of transport and degradation rates.

PAHs produced during combustion are released as aerosols and transported with rising heat and smoke to greater heights in the atmosphere, where they can be transported long distances

(Baek et al., 1991). PAHs association with fine fraction particles, often <5 μm and the majority

<1 μm, facilitates their persistence in the atmosphere (Baek et al., 1991). PAHs are often physically removed from the atmosphere through wet or dry deposition (Baek et al., 1991).

For organic carbon in general, there is enhanced preservation in euxinic sediments as a result of reduced efficiency of anoxic organic matter decomposition processes (e.g., sulfate reduction) (Canfield, 1989). Sulfate reduction can degrade PAHs, however, from modern observations, only the lowest molecular PAHs (≤ 3-ring; e.g., napthalene and phenanthrene) were significantly degraded in sediments, whereas ≥ 4-ring PAHs (e.g., pyrene and benzo[a]pyrene ) were not degraded (Coates et al., 1997). Although the decrease in efficiency of anoxic decomposition compared to oxic decomposition is marginal, the proportional increase in preservation and overall carbon concentrations is much greater in sediments under euxinic conditions (Canfield, 1989).

1.9. Study objectives and chapter outlines

Chapters in this dissertation summarize my research at the Pennsylvania State University from 2010 to 2016. A common goal in this research is to understand how warming climate and

15 associated changes in the hydrologic cycle affect the production and preservation of terrestrial organic carbon.

Chapter 2 focuses on the source and fate of terrestrial organic carbon and pyrogenic carbon during the PETM in the Bighorn Basin, Wyoming. This chapter develops the novel use of larger PAHs as a proxy for intermediate refractory carbon to help discern the relative preservation of different carbon pools in soils. I develop a molecular compound metric of degradation by calculating the molecular percent loss based on changes in intermediately refractory compounds to estimate the extent of organic carbon loss and proportion of refractory allochthonous carbon during the PETM. Key questions addressed in this chapter are: How did organic carbon production and preservation on land change across the PETM? To what extent did degradation affect the preservation of terrestrial organic carbon and the continuum of organic carbon pools: primary biomass, intermediate refractory carbon and refractory “stable” carbon during the

PETM? What portion of the organic carbon preserved in the PETM record is refractory allochthonous carbon?

Chapter 3 analyzes the preservation of terrestrial organic carbon and pyrogenic carbon during the PETM in the Piceance Basin, Colorado. This study provides a regional perspective on the extent of degradation during the PETM. The Piceance Basin has a range of lithologies and provides an opportunity to investigate the mechanisms that decreased organic carbon preservation during the PETM. I delve into how soil organic carbon is stabilized and destabilized on timescales of 100 – 10,000 years. I compare results between the Wyoming and Colorado sites and use elemental oxide weight percents to examine the effects of temperature, clay content and mineralogy, and variations in soil moisture on carbon degradation. Key questions in this chapter are: What is the effect of changes in lithology on the abundance of various pools of organic carbon in the record? How aggressive was degradation of organic carbon during the PETM on a more regional scale? Were trends of both reduced soil carbon and increased proportion of

16 allochthonous carbon inputs a local phenomenon in Wyoming or was it a more widespread trend? What are the mechanisms that enhanced destabilization and degradation of organic matter? Was organic carbon stabilized by clay minerals? How did variations in soil moisture affect organic carbon preservation?

Chapter 4 examines the sources and preservation of terrestrial organic carbon in marine sediments in the Arctic. It provides an overview of hydrocarbon results from Lomonosov Ridge to understand the mechanisms that influenced the source and fate of organic and pyrogenic carbon in the marine realm. Better organic carbon preservation at this site provides an opportunity to examine the occurrence of fire in a paleoecosystem during global warming and the feedbacks between fire, precipitation, and vegetation changes. I discuss changes in fire occurrence and fire ecology across the PETM. Key questions that inspired this work are: Do marine sections preserve terrestrial organic carbon better than terrestrial sections? What is the relationship between precipitation, vegetation, and fire occurrence before, during and after the PETM?

1.10. Anticipated publications from this work

Each chapter constitutes an individual publishable unit. Chapters provide more detail than the anticipated publication. Publications arising from this work include:

Chapter 2: “Decreased soil carbon in a warming world: Degraded pyrogenic carbon

during the Paleocene-Eocene Thermal Maximum”, will be submitted to Geology by

Elizabeth H. Denis with co-authors Bianca J. Maibauer, Gabriel J. Bowen, Allison A.

Baczynski, Francesca A. McInerney, Phillip E. Jardine, Guy J. Harrington, Margaret E.

Collinson, Claire M. Belcher, Scott L. Wing, and Katherine H. Freeman.

17 Chapter 3: “Widespread soil carbon loss during the Paleocene-Eocene Thermal

Maximum”, will be submitted to Earth and Planetary Science Letters by Elizabeth H.

Denis with co-authors Brady Z. Foreman and Katherine H. Freeman.

Chapter 4: “Fire and ecosystem change in the Arctic across the Paleocene-Eocene

Thermal Maximum”, will be submitted to Earth and Planetary Science Letters by

Elizabeth H. Denis with co-authors Nikolai Pedentchouk, Stefan Schouten, and Katherine

H. Freeman.

18 1.11. Figures

Figure 1-1. A schematic of the terrestrial carbon cycle. Approximate amounts of carbon (Petragrams of carbon (Pg C)) in each reservoir are in black outlined boxes for each reservoir. Soil reservoir (1,500 Pg C) is the first 1 m of soil (often reported by authors), and ~840 Pg C is in the next 1-3 m of soil (Chaopricha and Marín-Spiotta, 2014). Arrows represent carbon fluxes (Pg C/yr). Imaged adapted from TecEco.com; values are from Sundquist and Visser (2003).

Figure 1-2. Conceptual model of organic matter stabilization and degradation. Depicts mechanisms that stabilize organic carbon and factors that affect degradation.

Figure 1-3. Conceptual model of a soil particle with organic-mineral interactions. Polycyclic aromatic hydrocarbons (PAHs) are adsorbed onto the surface of the soil organic matter and penetrate into soil particle cavities or diffuse into the organic soil fraction. Figure from Jonsson et al., (2007).

Figure 1-4. Model of organo-mineral interactions. Amphiphilic compounds accumulate on soil surfaces through electrocstatitic interactions between hydrophilic organic compounds and surface oxygens or hydroxylated surface-metal cations. Hydrophobic portions of the molecules are protected from water by another layer of amphiphillic compounds, which forms a bilayer. Cation bridges form between polar groups and covalent metal ions (e.g., Ca+2). Hydrophobic compounds are also attracted to the hydrophobic structures. Figure from Kleber et al., (2007).

Figure 1-5. Measurements of the negative carbon isotope excursion (CIE) in terrestrial sections at the Paleocene-Eocene Thermal Maximum. Each record is characterized by the pre-PETM average, the most negative data point within the CIE (or average at sample level), and the post-PETM average (green circles). The dashed lines are the generalized shape of the CIE based on a global compilation of records and connect the averages of the plotted values (black circles). Figure is modified from McInerney and Wing (2011).

Figure 1-6. The carbon isotope excursion (CIE) magnitude of both marine (blue) and terrestrial (green) carbon archives is shown in box-and-whisker plots. Data is based on a global compilation of CIE records. The horizontal line represents the median; the box identifies the upper and low quartile and contains 50% of the data; the whiskers identify the range; open circles are outliers. Figure is modified from McInerney and Wing (2011).

Figure 1-7. Molecular structures of a selection of polycyclic aromatic hydrocarbons (PAHs). In this work, a suite of 16 PAHs are quantified ranging from 3-ring PAH (e.g., phenanthrene) to 7-ring PAH (e.g., coronene).

1.12. References

Agren, G.I., Bosatta, E., and Magill, A.H., 2001, Combining theory and experiment to understand effects of inorganic nitrogen on litter decomposition: Oecologia, v. 128, no. 1, p. 94–98, doi: 10.1007/s004420100646. Arinobu, T., Ishiwatari, R., Kaiho, K., and 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, v. 27, no. 8, p. 723–726, doi: 10.1130/0091-7613(1999)027<0723:SOPPAH>2.3.CO. Aziz, H.A., Hilgen, F.J., van Luijk, G.M., Sluijs, A., Kraus, M.J., Pares, J.M., and Gingerich, P.D., 2008, Astronomical climate control on paleosol stacking patterns in the upper Paleocene-lower Eocene Willwood Formation, Bighorn Basin, Wyoming: Geology, v. 36, no. 7, p. 531–534, doi: 10.1130/G24734A.1. Baczynski, A.A., McInerney, F.A., Wing, S.L., Kraus, M.J., Bloch, J.I., Boyer, D.M., Secord, R., Morse, P.E., and Fricke, H.C., 2013, Chemostratigraphic implications of spatial variation in the Paleocene-Eocene Thermal Maximum carbon isotope excursion, SE Bighorn Basin, Wyoming: Geochemistry, Geophysics, Geosystems, v. 14, no. 10, p. 4133–4152, doi: 10.1002/ggge.20265.

25 Baczynski, A.A., McInerney, F.A., Wing, S.L., Kraus, M.J., Morse, P.E., Bloch, J.I., Chung, A.H., and Freeman, K.H., 2016, Distortion of carbon isotope excursion in bulk soil organic matter during the Paleocene-Eocene thermal maximum: Geological Society of America Bulletin, , no. Xx, p. B31389.1, doi: 10.1130/B31389.1. Baek, S.O., Field, R.A., Goldstone, M.E., Kirk, P.W., Lester, J.N., and Perry, R., 1991, A review of atmospheric polycyclic aromatic hydrocarbons: Sources, fate and behavior: Water, Air, and Soil Pollution, v. 60, no. 3-4, p. 279–300, doi: 10.1007/BF00282628. Bataille, C.P., Mastalerz, M., Tipple, B.J., and Bowen, G.J., 2013, Influence of provenance and preservation on the carbon isotope variations of dispersed organic matter in ancient floodplain sediments: Geochemistry, Geophysics, Geosystems, v. 14, no. 11, p. 4874– 4891, doi: 10.1002/ggge.20294. Bice, K.L., and Marotzke, J., 2002, Could changing ocean circulation have destabilized methane hydrate at the Paleocene/Eocene boundary? Paleoceanography, v. 17, no. 2, p. -, doi: Artn 1018\rDoi 10.1029/2001pa000678. Bird, M.I., Wynn, J.G., Saiz, G., Wurster, C.M., and McBeath, A., 2014, The Pyrogenic Carbon Cycle: Annual Review of Earth and Planetary Sciences, v. 43, p. 150223150959000, doi: 10.1146/annurev-earth-060614-105038. Bowen, G.J., Bralower, T.J., Delaney, M.L., Dickens, G.R., Kelly, D.C., Koch, P.L., Kump, L.R., Meng, J., Sloan, L.C., Thomas, E., Wing, S.L., and Zachos, J.C., 2006, Eocene Hyperthermal Event Offers Insight Into Greenhouse Warming: Eos, v. 87, no. 17, p. 2–4, doi: 10.1029/2001PA000678.Bowen. Bowen, G.J., Maibauer, B.J., Kraus, M.J., Röhl, U., Westerhold, T., Steimke, A., Gingerich, P.D., Wing, S.L., and Clyde, W.C., 2015, Two massive, rapid releases of carbon during the onset of the Palaeocene-Eocene thermal maximum: Nature Geosci, v. 8, no. 1, p. 44–47, doi: 10.1038/ngeo2316. Bowes, G., 1993, Facing the inevitable: Plants and increasing atmospheric C02: Annual Review of Plant Physiology and Plant Molecular Biology, v. 44, p. 309–332. Canfield, D.E., 1989, Sulfate reduction and oxic respiration in marine sediments: implications for organic carbon preservation in euxinic environments: Deep-Sea Research, v. 36, no. 1, p. 121–138. Carvalhais, N., Forkel, M., Khomik, M., Bellarby, J., Jung, M., Migliavacca, M., Μu, M., Saatchi, S., Santoro, M., Thurner, M., Weber, U., Ahrens, B., Beer, C., Cescatti, A., et al., 2014, Global covariation of carbon turnover times with climate in terrestrial ecosystems: Nature, v. 514, no. 7521, p. 213–217, doi: 10.1038/nature13731. Chaopricha, N.T., and Marín-Spiotta, E., 2014, Soil burial contributes to deep soil organic carbon storage: Soil Biology and Biochemistry, v. 69, p. 251–264, doi: 10.1016/j.soilbio.2013.11.011. Christensen, B.T., 2001, Physical fractionation of soil and structural and functional complexity in organic matter turnover: European Journal of Soil Science, v. 52, no. 3, p. 345–353, doi: 10.1046/j.1365-2389.2001.00417.x. Coates, J.D., Woodward, J., Allen, J., Philp, P., and Lovley, D.R., 1997, Anaerobic degradation of polycyclic aromatic hydrocarbons and alkanes in petroleum-contaminated marine harbour sediments: Applied and Environmental Microbiology, v. 63, no. 9, p. 3589–3593. Cotton, J.M., Sheldon, N.D., Hren, M.T., and Gallagher, T.M., 2015, Positive feedback drives carbon release from soils to atmosphere during Paleocene/Eocene warming: American Journal of Science, v. 315, no. 4, p. 337–361, doi: 10.2475/04.2015.03. Coulon, F., Pelletier, E., Gourhant, L., and Delille, D., 2005, Effects of nutrient and temperature on degradation of petroleum hydrocarbons in contaminated sub-Antarctic soil: Chemosphere, v. 58, no. 10, p. 1439–1448, doi: 10.1016/j.chemosphere.2004.10.007.

26 Crouch, E.M., Dickens, G.R., Brinkhuis, H., Aubry, M.P., Hollis, C.J., Rogers, K.M., and Visscher, H., 2003, The Apectodinium acme and terrestrial discharge during the Paleocene-Eocene thermal maximum: New palynological, geochemical and calcareous nannoplankton observations at Tawanui, New Zealand: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 194, no. 4, p. 387–403, doi: 10.1016/S0031- 0182(03)00334-1. Cui, Y., Kump, L.R., Ridgwell, A.J., Charles, A.J., Junium, C.K., Diefendorf, A.F., Freeman, K.H., Urban, N.M., and Harding, I.C., 2011, Slow release of fossil carbon during the Palaeocene–Eocene Thermal Maximum: Nature Geoscience, v. 4, no. 7, p. 481–485, doi: 10.1038/ngeo1179. Denis, E.H., Toney, J.L., Tarozo, R., Anderson, R.S., Roach, L.D., and Huang, Y., 2012, Polycyclic aromatic hydrocarbons (PAHs) in lake sediments record historic fire events: Validation using HPLC-fluorescence detection: Organic Geochemistry, v. 45, p. 7–17, doi: 10.1016/j.orggeochem.2012.01.005. Don, A., Rödenbeck, C., and Gleixner, G., 2013, Unexpected control of soil carbon turnover by soil carbon concentration: Environmental Chemistry Letters, v. 11, no. 4, p. 407–413, doi: 10.1007/s10311-013-0433-3. Farley, K.A., and Eltgroth, S.F., 2003, An alternative age model for the Paleocene-Eocene thermal maximum using extraterrestrial 3He: Earth and Planetary Science Letters, v. 208, no. 3-4, p. 135–148, doi: 10.1016/S0012-821X(03)00017-7. Fontaine, S., Mariotti, A., and Abbadie, L., 2003, The priming effect of organic matter: A question of microbial competition? Soil Biology and Biochemistry, v. 35, no. 6, p. 837– 843, doi: 10.1016/S0038-0717(03)00123-8. Forbes, M.S., Raison, R.J., and Skjemstad, J.O., 2006, Formation, transformation and transport of black carbon (charcoal) in terrestrial and aquatic ecosystems: Science of the Total Environment, v. 370, no. 1, p. 190–206, doi: 10.1016/j.scitotenv.2006.06.007. Foreman, B.Z., 2012, Fluvial Response to the Paleocene-Eocene Thermal Maximum in Western North America: University of Wyoming, 147 p. Foreman, B.Z., Heller, P.L., and Clementz, M.T., 2012, Fluvial response to abrupt global warming at the Palaeocene/Eocene boundary.: Nature, v. 491, no. 7422, p. 92–5, doi: 10.1038/nature11513. Guenet, B., Juarez, S., Bardoux, G., Abbadie, L., and Chenu, C., 2012, Evidence that stable C is as vulnerable to priming effect as is more labile C in soil: Soil Biology and Biochemistry, v. 52, p. 43–48, doi: 10.1016/j.soilbio.2012.04.001. Handley, L., O’Halloran, A., Pearson, P.N., Hawkins, E., Nicholas, C.J., Schouten, S., McMillan, I.K., and Pancost, R.D., 2012, Changes in the hydrological cycle in tropical East Africa during the Paleocene-Eocene Thermal Maximum: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 329-330, p. 10–21, doi: 10.1016/j.palaeo.2012.02.002. Hedges, J.I., and Oades, J.M., 1997, Comparative organic geochemistries of soils and marine sediments: Organic Geochemistry, v. 27, no. 7-8, p. 319–361, doi: 10.1016/S0146- 6380(97)00056-9. Huber, M., 2009, Climate change: Snakes tell a torrid tale.: Nature, v. 457, no. 7230, p. 669–671, doi: 10.1038/457669a. Huxman, T.E., Smith, M.D., Fay, P. a, Knapp, A.K., Shaw, M.R., Loik, M.E., Smith, S.D., Tissue, D.T., Zak, J.C., Weltzin, J.F., Pockman, W.T., Sala, O.E., Haddad, B.M., Harte, J., et al., 2004, Convergence across biomes to a common rain-use efficiency.: Nature, v. 429, no. 6992, p. 651–654, doi: 10.1038/nature02561.

27 Ibach, L.E.J., 1982, Relationship between sedimentation rate and total organic carbon in ancient marine sediments: The American Association of Petroleum Geogolists Bulletin, v. 66, no. 2, p. 170–188. Jaramillo, C., Ochoa, D., Contreras, L., Pagani, M., Carvajal-Ortiz, H., Pratt, L.M., Krishnan, S., Cardona, A., Romero, M., Quiroz, L., Rodriguez, G., Rueda, M.J., de la Parra, F., Morón, S., et al., 2010, Effects of rapid global warming at the Paleocene-Eocene boundary on neotropical vegetation: Science, v. 330, no. 6006, p. 957–961, doi: 10.1126/science.1193833. Jiang, C., Alexander, R., Kagi, R.I., and Murray, A.P., 2000, Origin of perylene in ancient sediments and its geological significance: Organic Geochemistry, v. 31, no. 12, p. 1545– 1559, doi: 10.1016/S0146-6380(00)00074-7. Jiang, C., Alexander, R., Kagi, R.I., and Murray, A.P., 1998, Polycyclic aromatic hydrocarbons in ancient sediments and their relationships to palaeoclimate: Organic Geochemistry, v. 29, no. 5-7 -7 pt 2, p. 1721–1735, doi: 10.1016/S0146-6380(98)00083-7. Jobbágy, E.G., and Jackson, R.B., 2000, The Vertical Distribution of Soil Organic Carbon and Its relation to climate and vegetation: Ecological Applications, v. 10, no. 2, p. 423–436, doi: 10.1890/1051-0761(2000)010[0423:TVDOSO]2.0.CO;2. Johnsen, A.R., Wick, L.Y., and Harms, H., 2005, Principles of microbial PAH-degradation in soil: Environmental Pollution, v. 133, no. 1, p. 71–84, doi: 10.1016/j.envpol.2004.04.015. Jonsson, S., Persson, Y., Frankki, S., van Bavel, B., Lundstedt, S., Haglund, P., and Tysklind, M., 2007, Degradation of polycyclic aromatic hydrocarbons (PAHs) in contaminated soils by Fenton’s reagent: A multivariate evaluation of the importance of soil characteristics and PAH properties: Journal of Hazardous Materials, v. 149, no. 1, p. 86–96, doi: 10.1016/j.jhazmat.2007.03.057. Kaiho, K., Yatsu, S., Oba, M., Gorjan, P., Casier, J.G., and Ikeda, M., 2013, A forest fire and soil erosion event during the Late Devonian mass extinction: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 392, p. 272–280, doi: 10.1016/j.palaeo.2013.09.008. Karhu, K., Auffret, M.D., Dungait, J. a J., Hopkins, D.W., Prosser, J.I., Singh, B.K., Subke, J.-A., Wookey, P. a, Agren, G.I., Sebastià, M.-T., Gouriveau, F., Bergkvist, G., Meir, P., Nottingham, A.T., et al., 2014, Temperature sensitivity of soil respiration rates enhanced by microbial community response.: Nature, v. 513, no. 7516, p. 81–4, doi: 10.1038/nature13604. Keil, R., Montluçon, D., Prahl, F., and Hedges, J., 1994, Sorptive preservation of labile organic matter in marine sediments: Nature, v. 370, p. 549–552, doi: 10.1038/370549a0. Kennett, J.P., and Stott, L.D., 1991, Abrupt deep-sea warming, palaeoceanographic changes and benthic extinctions at the end of the Palaeocene: Nature, v. 353, p. 225–229. Killops, S., and Killops, V., 2004, Introduction to Organic Geochemistry: Introduction to Organic Geochemistry, , no. c, p. 30–70, doi: 10.1002/9781118697214. Killops, S., and Killops, V., 2005, The Carbon Cycle and Climate Change, in Introduction to Organic Geochemistry, Blackwell Publishing, Oxford, p. 246–294. Killops, S.D., and Massoud, M.S., 1992, Polycyclic aromatic hydrocarbons of pyrolytic origin in ancient sediments: evidence for Jurassic vegetation fires: Organic Geochemistry, v. 18, no. 1, p. 1–7, doi: 10.1016/0146-6380(92)90137-M. Kleber, M., 2010, What is recalcitrant soil organic matter? Environmental Chemistry, v. 7, no. 4, p. 320–332, doi: 10.1071/EN10006. Kleber, M., Sollins, P., and Sutton, R., 2007, A conceptual model of organo-mineral interactions in soils: Self-assembly of organic molecular fragments into zonal structures on mineral surfaces: Biogeochemistry, v. 85, no. 1, p. 9–24, doi: 10.1007/s10533-007-9103-5.

28 Knorr, W., Prentice, I., House, J., and Holland, E., 2005, Long-term sensitivity of soil carbon turnover to warming: Nature, v. 433, no. January, p. 298–301, doi: 10.129/2002PA000837. Kraus, M.J., and Riggins, S., 2007, Transient drying during the Paleocene-Eocene Thermal Maximum (PETM): Analysis of paleosols in the bighorn basin, Wyoming: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 245, no. 3-4, p. 444–461, doi: 10.1016/j.palaeo.2006.09.011. Krull, E.S., Baldock, J.A., and Skjemstad, J.O., 2003, Importance of mechanisms and processes of the stabilisation of soil organic matter for modelling carbon turnover: Functional Plant Biology, v. 30, no. 2, p. 207–222, doi: 10.1071/FP02085. Kuzyakov, Y., Friedel, J.K., and Stahr, K., 2000, Review of mechanisms and quantification of priming effects: Soil Biology and Biochemistry, v. 32, no. 11-12, p. 1485–1498, doi: 10.1016/S0038-0717(00)00084-5. Lal, R., 2004, Soil carbon sequestration impacts on global climate change and food security: Science, v. 304, no. 5677, p. 1623–1627, doi: 10.1126/science.1097396. Lawrence, C.R., Harden, J.W., Xu, X., Schulz, M.S., and Trumbore, S.E., 2015, Long-term controls on soil organic carbon with depth and time: A case study from the Cowlitz River Chronosequence, WA USA: Geoderma, v. 247-248, p. 73–87, doi: 10.1016/j.geoderma.2015.02.005. Lehmann, C.E.R., Archibald, S.A., Hoffmann, W.A., and Bond, W.J., 2011, Deciphering the distribution of the savanna biome: New Phytologist, v. 191, no. 1, p. 197–209, doi: 10.1111/j.1469-8137.2011.03689.x. Leirós, M.C., Trasar-Cepeda, C., Seoane, S., and Gil-Sotres, F., 1999, Dependence of mineralization of soil organic matter on temperature and moisture: Soil Biology and Biochemistry, v. 31, no. 3, p. 327–335, doi: 10.1016/S0038-0717(98)00129-1. Lloyd, J.J., and Taylor, J.A., 1994, On the temperature dependence of soil respiration: Functional Ecology, v. 8, no. 3, p. 315–323, doi: 10.2307/2389824. Lorenz, K., Lal, R., Preston, C.M., and Nierop, K.G.J., 2007, Strengthening the soil organic carbon pool by increasing contributions from recalcitrant aliphatic bio(macro)molecules: Geoderma, v. 142, no. 1-2, p. 1–10, doi: 10.1016/j.geoderma.2007.07.013. Luo, Y., 2007, Terrestrial Carbon–Cycle Feedback to Climate Warming: Annual Review of Ecology, Evolution, and Systematics, v. 38, no. 1, p. 683–712, doi: 10.1146/annurev.ecolsys.38.091206.095808. Von Lützow, M., Kögel-Knabner, I., Ekschmitt, K., Matzner, E., Guggenberger, G., Marschner, B., and Flessa, H., 2006, Stabilization of organic matter in temperate soils: Mechanisms and their relevance under different soil conditions - A review: European Journal of Soil Science, v. 57, no. 4, p. 426–445, doi: 10.1111/j.1365-2389.2006.00809.x. Von Lützow, M., Kögel-Knabner, I., Ludwig, B., Matzner, E., Flessa, H., Ekschmitt, K., Guggenberger, G., Marschner, B., and Kalbitz, K., 2008, Stabilization mechanisms of organic matter in four temperate soils: Development and application of a conceptual model: Journal of Plant Nutrition and Soil Science, v. 171, no. 1, p. 111–124, doi: 10.1002/jpln.200700047. Magioncalda, R., Dupuis, C., Smith, T., Steurbaut, E., and Gingerich, P.D., 2004, Paleocene- Eocene carbon isotope excursion in organic carbon and pedogenic carbonate: Direct comparison in a continental stratigraphic section: Geology, v. 32, no. 7, p. 553–556, doi: 10.1130/G20476.1. Maibauer, B.J., 2013, Carbon Isotope Stratigraphy of Early Eocene Hyperthermals in the Bighorn Basin, Wyoming, USA: Analogues for Modern Anthropogenic Carbon Emissions: The University of Utah, 122 p.

29 Manabe, S., 1997, Early Development in the Study of Greenhouse Warming : The Emergence of Climate Models: Ambio, v. 26, no. 1, p. 47–51. Marschner, B., Brodowski, S., Dreves, A., Gleixner, G., Gude, A., Grootes, P.M., Hamer, U., Heim, A., Jandl, G., Ji, R., Kaiser, K., Kalbitz, K., Kramer, C., Leinweber, P., et al., 2008, How relevant is recalcitrance for the stabilization of organic matter in soils? Journal of Plant Nutrition and Soil Science, v. 171, no. 1, p. 91–110, doi: 10.1002/jpln.200700049. Marynowski, L., and 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, v. 144, no. 3, p. 569–595, doi: 10.1017/S0016756807003317. Marynowski, L., Scott, A.C., Zatoń, M., Parent, H., and Garrido, A.C., 2011, First multi-proxy record of Jurassic wildfires from Gondwana: Evidence from the Middle Jurassic of the Neuquén Basin, Argentina: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 299, no. 1-2, p. 129–136, doi: 10.1016/j.palaeo.2010.10.041. Marynowski, L., and Simoneit, B.R.T., 2009, Widespread Upper Triassic To Lower Jurassic Wildfire Records From Poland: Evidence From Charcoal and Pyrolytic Polycyclic Aromatic Hydrocarbons: Palaios, v. 24, no. 12, p. 785–798, doi: 10.2110/palo.2009.p09- 044r. Masclet, P., Cachier, H., Liousse, C., and Wortham, H., 1995, Emissions of polycyclic aromatic hydrocarbons by savanna fires: Journal of Atmospheric Chemistry, v. 22, no. 1-2, p. 41– 54, doi: 10.1007/BF00708180. Masiello, C.A., 2004, New directions in black carbon organic geochemistry: Marine Chemistry, v. 92, no. 1-4 Spec. Iss., p. 201–213, doi: 10.1029/2002GB001939. May, W.E., Wasik, S.P., and Freeman, D.H., 1978, Determination of the Solubility Behavior of Some Polycyclic Aromatic Hydrocarbons in Water: Analytical chemistry, v. 50, no. 7, p. 997–1000, doi: 10.1021/ac50029a042. Mayer, L.M., 1994, Relationships between mineral surfaces and organic carbon concentrations in soils and sediments: Chemical Geology, v. 114, no. 3-4, p. 347–363, doi: 10.1016/0009- 2541(94)90063-9. McInerney, F.A., and Wing, S.L., 2011, The Paleocene-Eocene Thermal Maximum: A Perturbation of Carbon Cycle, Climate, and Biosphere with Implications for the Future: Annual Review of Earth and Planetary Sciences, v. 39, no. 1, p. 489–516, doi: 10.1146/annurev-earth-040610-133431. Mita, H., and Shimoyama, A., 1999, Distribution of polycyclic aromatic hydrocarbons in the K/T boundary sediments at Kawaruppu, Hokkaido, Japan: Geochemical Journal, v. 33, no. 5, p. 305–315. Muri, G., and Wakeham, S.G., 2009, Effect of depositional regimes on polycyclic aromatic hydrocarbons in Lake Bled (NW Slovenia) sediments: Chemosphere, v. 77, no. 1, p. 74– 79, doi: 10.1016/j.chemosphere.2009.05.023. Murphy, B.H., Farley, K.A., and Zachos, J.C., 2010, An extraterrestrial 3He-based timescale for the Paleocene-Eocene thermal maximum (PETM) from Walvis Ridge, IODP Site 1266: Geochimica et Cosmochimica Acta, v. 74, no. 17, p. 5098–5108, doi: 10.1016/j.gca.2010.03.039. Nobel, P.S., and Cortazar, V.G. De, 1991, Growth and Predicted Productivity of Opuntia ficus- indica for Current and Elevated Carbon Dioxide: Agronomy Journal, v. 83, p. 224–230. Nobel, P.S., and Hartsock, T.L., 1986, Short-Term and Long-Term Responses of Crassulacean Acid Metabolism Plants to Elevated CO2: Plant Physiology, v. 82, p. 604–606.

30 Norris, R.D., and Röhl, U., 1999, climate warming during the Palaeocene / Eocene transition: Nature, v. 401, no. 21, p. 775–778. Oades, J.M., 1988, The Retention of Organic Matter in Soils: Biogeochemistry, v. 5, no. 1, p. 35– 70. Pagani, M., Caldeira, K., Archer, D., and Zachos, J.C., 2006a, An Ancient Carbon Mystery: Science, v. 314, no. 5805, p. 1556–1557, doi: 10.1126/science.1136110. Pagani, M., Pedentchouk, N., Huber, M., Sluijs, A., Schouten, S., Brinkhuis, H., Sinninghe Damsté, J.S., and Dickens, G.R., 2006b, Arctic hydrology during global warming at the Palaeocene/Eocene thermal maximum: Nature, v. 442, no. 7103, p. 671–675, doi: 10.1038/nature05043. Page, D.S., Boehm, P.D., Douglas, G.S., Bence, A.E., Burns, W.A., and Mankiewicz, P.J., 1999, Pyrogenic Polycyclic Aromatic Hydrocarbons in Sediments Record Past Human Activity: A Case Study in Prince William Sound, Alaska: Marine Pollution Bulletin, v. 38, no. 4, p. 247–260, doi: 10.1016/S0025-326X(98)00142-8. Pisani, O., Lin, L.H., Lun, O.O.Y., Lajtha, K., Nadelhoffer, K.J., Simpson, A.J., and Simpson, M.J., 2016, Long-term doubling of litter inputs accelerates soil organic matter degradation and reduces soil carbon stocks: Biogeochemistry, v. 127, no. 1, p. 1–14, doi: 10.1007/s10533-015-0171-7. Ramankutty, N., Foley, J.A., Norman, J., and McSweeney, K., 2002, The global distribution of cultivable lands: Current patterns and sensitivity to possible climate change: Global Ecology and Biogeography, v. 11, no. 5, p. 377–392, doi: 10.1046/j.1466- 822x.2002.00294.x. Ramdahl, T., 1983, Retene - a molecular marker of wood combustion in ambient air: Nature, v. 306, p. 580–582, doi: 10.1038/306580a0. Robert, C., and Kennett, J.P., 1994, Antarctic subtropical humid episode at the Paleocene-Eocene boundary: Clay-mineral evidence: Geology, v. 22, no. 3, p. 211, doi: 10.1130/0091- 7613(1994)022<0211:ASHEAT>2.3.CO;2. Röhl, U., Westerhold, T., Bralower, T.J., and Zachos, J.C., 2007, On the duration of the Paleocene-Eocene thermal maximum (PETM): Geochemistry, Geophysics, Geosystems, v. 8, no. 12, doi: 10.1029/2007GC001784. Schimel, J.P., Wetterstedt, J.?? M., Holden, P.A., and Trumbore, S.E., 2011, Drying/rewetting cycles mobilize old C from deep soils from a California annual grassland: Soil Biology and Biochemistry, v. 43, no. 5, p. 1101–1103, doi: 10.1016/j.soilbio.2011.01.008. Schmidt, M.W.I., and Noack, A.G., 2000, Analysis , distribution , implications , and current challenges: Global Biogeochemical Cycles, v. 14, no. 3, p. 777–793. Schmidt, M.W.I., Torn, M.S., Abiven, S., Dittmar, T., Guggenberger, G., Janssens, I. a., Kleber, M., Kögel-Knabner, I., Lehmann, J., Manning, D. a. C., Nannipieri, P., Rasse, D.P., Weiner, S., and Trumbore, S.E., 2011, Persistence of soil organic matter as an ecosystem property: Nature, v. 478, no. 7367, p. 49–56, doi: 10.1038/nature10386. van de Schootbrugge, B., 2010, A fiery start to the Jurassic: Nature Geoscience, v. 3, no. 6, p. 381–382, doi: 10.1038/ngeo878. Schuur, E.A.G., Bockheim, J., Canadell, J.G., Euskirchen, E., Field, C.B., Goryachkin, S. V, Hagemann, S., Kuhry, P., Lafleur, P.M., Lee, H., Mazhitova, G., Nelson, F., Rinke, A., Romanovsky, V.E., et al., 2008, Vulnerability of permafrost carbon to climate change: Implications for the global carbon cycle: BioScience, v. 58, no. September, p. 701–714, doi: 10.1641/B580807. Scott, A.C., Kenig, F., Plotnick, R.E., Glasspool, I.J., Chaloner, W.G., and Eble, C.F., 2010, Evidence of multiple late Bashkirian to early Moscovian (Pennsylvanian) fire events

31 preserved in contemporaneous cave fills: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 291, no. 1-2, p. 72–84, doi: 10.1016/j.palaeo.2009.06.008. Secord, R., Gingerich, P.D., Lohmann, K.C., and Macleod, K.G., 2010, Continental warming preceding the Palaeocene-Eocene thermal maximum: Nature, v. 467, no. 7318, p. 955– 958, doi: 10.1038/nature09441. Shen, W., Sun, Y., Lin, Y., Liu, D., and Chai, P., 2011, Evidence for wildfire in the Meishan section and implications for Permian-Triassic events: Geochimica et Cosmochimica Acta, v. 75, no. 7, p. 1992–2006, doi: 10.1016/j.gca.2011.01.027. Sloan, L.C., Walker, J.C., and Moore, T.C., 1995, Possible role of oceanic heat transport in early Eocene climate: Paleoceanography, v. 10, no. 2, p. 347–56, doi: 10.1029/94PA02928. Smith, J.J., Hasiotis, S.T., Kraus, M.J., and Woody, D.T., 2008, Relationship of floodplain ichnocoenoses to paleopedology, paleohydrology, and paleoclimate in the Willwood Formation, Wyoming, during the Paleocene-Eocene Thermal Maximum: Palaios, v. 23, no. 9-10, p. 683–699, doi: 10.2110/palo.2007.p07-080r. Smith, F.A., Wing, S.L., and Freeman, K.H., 2007, Magnitude of the carbon isotope excursion at the Paleocene-Eocene thermal maximum: The role of plant community change: Earth and Planetary Science Letters, v. 262, no. 1-2, p. 50–65, doi: 10.1016/j.epsl.2007.07.021. Sollins, P., Homann, P., and Caldwell, B. a., 1996, Stabilization and destabilization of soil organic matter: mechanisms and controls: Geoderma, v. 74, no. 1-2, p. 65–105, doi: 10.1016/S0016-7061(96)00036-5. Stockmann, U., Adams, M.A., Crawford, J.W., Field, D.J., Henakaarchchi, N., Jenkins, M., Minasny, B., McBratney, A.B., Courcelles, V. de R. de, Singh, K., Wheeler, I., Abbott, L., Angers, D.A., Baldock, J., et al., 2013, The knowns, known unknowns and unknowns of sequestration of soil organic carbon: Agriculture, Ecosystems and Environment, v. 164, no. 2013, p. 80–99, doi: 10.1016/j.agee.2012.10.001. Sundquist, E.T., and Visser, K., 2005, Geologic history of the carbon cycle, in Schlesinger, W.H., Holland, H.., and Turekian, K.K. eds., Biogeochemistry, Elsevier, New York, p. 425– 472. Venkatesan, M.I., and Dahl, J., 1989, Organic geochemical evidence fpr global fires at the Cretaceous/Tertiary boundary: Letters of Nature, v. 338, p. 57–60. Vergnoux, A., Malleret, L., Asia, L., Doumenq, P., and Theraulaz, F., 2011, Impact of forest fires on PAH level and distribution in soils: Environmental Research, v. 111, no. 2, p. 193– 198, doi: 10.1016/j.envres.2010.01.008. Westerhold, T., Röhl, U., McCarren, H.K., and Zachos, J.C., 2009, Latest on the absolute age of the Paleocene-Eocene Thermal Maximum (PETM): New insights from exact stratigraphic position of key ash layers + 19 and - 17: Earth and Planetary Science Letters, v. 287, no. 3-4, p. 412–419, doi: 10.1016/j.epsl.2009.08.027. Wing, S.L., and Currano, E.D., 2013, Plant response to a global greenhouse event 56 million years ago: American Journal of Botany, v. 100, no. 7, p. 1234–1254, doi: 10.3732/ajb.1200554. Wing, S.L., Harrington, G.J., Smith, F.A., Bloch, J.I., Boyer, D.M., and Freeman, K.H., 2005, Transient Floral Change and Rapid Global Warming at the Paleocene-Eocene Boundary: Science, v. 310, no. 5750, p. 993–996, doi: 10.1126/science.1116913. Wright, J.D., and Schaller, M.F., 2013, Evidence for a rapid release of carbon at the Paleocene- Eocene thermal maximum: Proceedings of the National Academy of Sciences of the United States of America, v. 110, no. 40, p. 15908–15913, doi: 10.1073/pnas.1309188110/- /DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas.1309188110.

32 Xiang, S.R., Doyle, A., Holden, P.A., and Schimel, J.P., 2008, Drying and rewetting effects on C and N mineralization and microbial activity in surface and subsurface California grassland soils: Soil Biology and Biochemistry, v. 40, no. 9, p. 2281–2289, doi: 10.1016/j.soilbio.2008.05.004. Yunker, M.B., Macdonald, R.W., Vingarzan, R., Mitchell, R.H., Goyette, D., and Sylvestre, S., 2002, PAHs in the Fraser River basin: A critical appraisal of PAH ratios as indicators of PAH source and composition: Organic Geochemistry, v. 33, no. 4, p. 489–515, doi: 10.1016/S0146-6380(02)00002-5. Zachos, J.C., Nicolo, M., Raffi, I., Lourens, L.J., McCarren, H., and Kroon, D., 2005, Rapid Acidification of the Ocean During the Paleocene-Eocene Thermal Maximum: Science, v. 308, p. 1611–1615, doi: 10.1126/science.1109004. Zeebe, R.E., Ridgwell, A., and Zachos, J.C., 2016, Anthropogenic carbon release rate unprecedented during the past 66 million years: Nature Geoscience, v. 9, no. April, p. 325–329, doi: 10.1038/ngeo2681. Zeebe, R.E., and Zachos, J.C., 2013, Long-term legacy of massive carbon input to the Earth system: Anthropocene versus Eocene: Philosophical transactions. Series A, Mathematical, physical, and engineering sciences, v. 371, no. 2001, p. 20120006, doi: 10.1098/rsta.2012.0006.

Chapter 2

Decreased soil carbon in a warming world: Degraded pyrogenic carbon during the Paleocene-Eocene Thermal Maximum

2.1. Abstract

The storage and release of carbon from the biosphere is influenced by temperature and precipitation through changes in plant productivity and in oxidative loss, such as fire and microbial respiration. Global warming will affect both carbon storage and release, but its net impact on the long-term fate of organic carbon in soils is unclear. Changes in climate and soil carbon during past warming events should help elucidate climate-carbon cycle relationships, especially effects that are expressed over long durations (1,000 – 10,000 years).

Abrupt warming during the Paleocene-Eocene Thermal Maximum (PETM) dramatically altered vegetation and hydrologic patterns globally. To assess the consequences for soil carbon we measured total organic carbon (TOC), polycyclic aromatic hydrocarbons (PAHs), and charcoal through two paleo-floodplain depositional sites in the Bighorn Basin, Wyoming, USA.

At both sites TOC, PAHs, and charcoal declined during the PETM. Despite an independently inferred increase in seasonality of precipitation that could have promoted fire, loss of pyrogenic carbon from paleosols obscures any evidence for increased fires during the PETM. Further, the decline in pyrogenic carbon relative to TOC suggests enhanced preservation of more refractory carbon phases such as allochthonous kerogen. The severe loss of soil carbon, and especially pyrogenic carbon, strongly indicates intensified rates of organic matter decay during the PETM.

We propose that soil respiration increased more than primary productivity and hindered soil carbon sequestration during this period of hotter climate with more seasonal precipitation.

34 2.2. Introduction

The long-term fate of soil organic carbon during global warming is difficult to predict, but important because soil carbon is the largest terrestrial carbon reservoir. Further, both temperature and water availability influence the production and preservation of terrestrial organic carbon in the biosphere. The main source of carbon to soils, plant matter, increases with rainfall

(Huxman et al., 2004). The main mechanism of carbon loss from soils, soil respiration, increases with temperature (Lloyd and Taylor, 1994; Knorr et al., 2005; Luo, 2007; Carvalhais et al., 2014;

Karhu et al., 2014) and can also be affected by soil moisture and oxygen concentrations (Leirós et al., 1999; Krull et al., 2003). Organic carbon can be divided into multiple pools based on reactivity and susceptibility to microbial degradation ranging from readily degradable fresh biomass to intermediately degradable refractory soil organic matter to microbial resistant thermally mature fossil organic matter (Knorr et al., 2005). Depending on climate conditions, soils can act as either a source (e.g., permafrost thawing (Schuur et al., 2008)) or a sink (Lal,

2004) for atmospheric carbon. Efforts to enhance the retention of carbon in modern soils and

sequester anthropogenic CO2 (Schmidt and Noack, 2000; Lal, 2004) are tied to the addition of biochar, which is made from biomass via pyrolysis similar to charcoal (Lehmann et al., 2006), even though the duration of such storage is uncertain (e.g., tens to tens of thousands of years).

We evaluated changes in the amount and type of carbon stored in soils during a past global warming event, the Paleocene-Eocene Thermal Maximum (PETM). We used total organic carbon (%TOC) and both particulate and molecular indicators of pyrogenic carbon to assess the impact of changing environmental conditions on the relative preservation of fresh biomass, less labile pyrogenic carbon, and recycled, refractory fossil carbon. Reconstructions of carbon cycling and environmental conditions during past climate warming events, such as the PETM, provide a

35 long-term perspective that can help elucidate terrestrial productivity and organic carbon storage in relation to changes in climate.

The PETM was a geologically abrupt period of climate change 55.5 million years ago

(Westerhold et al., 2012), and is widely invoked as a geologic analog for modern climate change

(Cui et al., 2011; Zeebe and Zachos, 2013; Zeebe et al., 2016). A negative carbon isotope excursion (CIE) marks the PETM, and signifies a major perturbation to the carbon cycle

(McInerney and Wing, 2011). At least 3,000 Pg of 13C-depleted carbon was released into the atmosphere over ~10 ky and global temperatures rose ~5-8°C over ~170 ky (Wing et al., 2005;

Zachos et al., 2005; Röhl et al., 2007; Secord et al., 2010; Cui et al., 2011; McInerney and Wing,

2011). Most likely, a combination of sources and feedbacks, such as thermogenic methane or the dissociation of marine methane clathrates, contributed to the large release of CO2 to the atmosphere (McInerney and Wing, 2011; Bowen et al., 2015). There were dramatic shifts in vegetation and precipitation patterns (Wing et al., 2005; Pagani et al., 2006b; Kraus and Riggins,

2007; Smith et al., 2007; Smith et al., 2008; Jaramillo et al., 2010; Foreman et al., 2012; Handley et al., 2012; Wing and Currano, 2013). For example, increased terrestrial inputs and greater abundance of kaolonite in marginal marine settings during the PETM suggest that there was increased run off and increased variability in rain events (Robert and Kennett, 1994; Crouch et al., 2003). How these changes in the hydrological cycle affected terrestrial productivity, soil respiration, and fire occurrence is not understood.

Terrestrial sediments that span the PETM tend to have low total organic carbon (%TOC), with values reported on the order of 0.01 – 1% and typically lower during the event than the preceding late Paleocene or subsequent early Eocene (Magioncalda et al., 2004; Clechenko et al.,

2007; Domingo et al., 2009; Foreman et al., 2012). The reduced TOC could reflect decreased production (i.e., less productive terrestrial ecosystems), decreased preservation (e.g., enhanced

13 13 oxidation or weathering of soil carbon), or both. The δ C of total organic carbon (δ Corg), which

36 is often used to identify isotope excursions and provide other paleoenvironmental information, can be highly variable on local scales, potentially due to soil respiration (Clechenko et al., 2007;

Cotton et al., 2015) or incorporation of reworked fossil carbon during the PETM (Baczynski et al., 2013; Bataille et al., 2013; Baczynski et al., 2016), which can particularly affect lower TOC samples. Studies have determined that there were inputs of refractory fossil carbon or allochthonous (pre-PETM) carbon to PETM sediments (Baczynski et al., 2013; Bataille et al.,

2013; Baczynski et al., 2016). Reworked Mesozoic organic material (allochthonous carbon) has been observed commonly in PETM sediments of the eastern Bighorn Basin (Wing et al., 2005;

Baczynski et al., 2013; Baczynski et al., 2016). Baczynski et al. (2016) imply that during the

PETM relative to the late Paleocene organic carbon was more degraded by microbes (degradation rates doubled) and had greater inputs of allochthonous material (28-63% of total organic carbon), both of which can affect carbon to different extents at different sites.

We test to what extent organic carbon was degraded with a molecular approach using

%TOC and both particulate and molecular indicators of pyrogenic carbon at two sites within the

Bighorn Basin, Wyoming (Figure 2-1). Understanding the extent of degradation of organics has implications for understanding whether soils served as a sink or source of carbon and determining allochthonous carbon input into the basin (Baczynski et al., 2016). Ultimately, we consider molecular observations, not just as a primary signal (e.g., of environment or fire) but also as a secondary signal to assess the impact of changing conditions on the extent of degradation and relative preservation of fresh biomass, less labile pyrogenic carbon, and refractory fossil or allochthonous carbon.

Pyrogenic carbon is a continuum of combustion products generated as solid residue or volatiles, ranging from slightly charred material to soot (Schmidt and Noack, 2000; Maisello,

2004; Bird et al., 2015). Polycyclic aromatic hydrocarbons (PAHs) are parts of this continuum and representative products of combustion that are released as volatiles and in association with

37 particles (Masclet et al., 1995). PAHs are used as markers for fire in the paleorecord (Ramdahl,

1983; Killops and Massoud, 1992; Page et al., 1999; Yunker et al., 2002; Denis et al., 2012).

Charcoal has a relatively long residence time in soils, estimated on the order of 500 – 10,000 years (Schmidt and Noack, 2000; Forbes et al., 2006; Von Lützow et al., 2006). Like charcoal

(Schmidt and Noack, 2000), high molecular weight PAHs do not degrade readily due to their aromatic structure (Coulon et al., 2005; Johnsen et al., 2005; Von Lützow et al., 2006).

Degradation rates vary with PAH size because lower molecular-weight PAHs (≤ 4 rings) are more soluble in water and, therefore, more susceptible to degradation than higher molecular- weight compounds (≥ 5 rings) (Killops and Massoud, 1992; Maliszewska-kordybach, 1993;

Coulon et al., 2005). Aqueous solubility of PAHs decreases almost logarithmically with increasing molecular mass, and therefore, 5-7 ring PAHs are significantly less soluble than 2-4 ring counterparts (May et al., 1978; Johnsen et al., 2005) (Figure A-1). Because of the overall recalcitrant nature of the fire-derived macerals and compounds (Schmidt and Noack, 2000;

Johnsen et al., 2005; Bird et al., 2014), we use PAHs and charcoal as proxies for refractory carbon.

By examining pyrogenic carbon relative to TOC, we can begin to understand the extent of degradation across the carbon continuum. Here, we develop the use of molecular observations as a metric of degradation by evaluating the quantities of PAHs relative to TOC to make estimates for the percentage of allochthonous, autochthonous, and degraded organic carbon within a given time interval for two paleo-floodplain depositional sites in the Bighorn Basin,

Wyoming, USA (Figure 2-1).

38 2.3. Study Sections

PETM sediments are precisely located and well exposed over wide areas in the Bighorn

Basin and have been well-studied (Gingerich, 1989; Koch et al., 1992; Magioncalda et al., 2004;

Wing et al., 2005; Kraus and Riggins, 2007; Smith et al., 2007; Smith et al., 2008; Baczynski et al., 2013; Clyde et al., 2013). Most deposition occurred on floodplains of avulsive river systems where 20-50 meters of mud accumulated during the PETM. Cores containing unweathered rock were collected as part of the Bighorn Basin Coring Project (BBCP) at two sites, Basin Substation

(BSN, N 44.4162042 W 108.1053154) and Polecat Bench (PCB, N 44.7688571 W 108.8879668)

(Clyde et al., 2013). Polecat Bench is in the northern basin, but near the depositional axis. Here the PETM sequence is dominated by oxidized floodplain paleosols and splay deposits (Kraus et al., 2015). Basin Substation is near the eastern margin of the Bighorn Basin where basin margin folding likely created areas of low topography and poor drainage during the early Cenozoic. Here, fewer oxidized paleosols formed during latest Paleocene and PETM time, probably reflecting wetter, more reduced floodplain soils. At Basin Substation the lowest red paleosols correspond approximately to the onset of the PETM, indicating a shift to better-drained conditions during the event as has been observed farther south along the eastern margin of the basin (Wing et al., 2005;

Rose et al., 2012; Kraus et al., 2013).

2.4. Methods

For lipid analyses from BSN-1B and PCB-2A cores, lipids were extracted from powdered rock using an Accelerated Solvent Extractor (ASE) and separated into compound class fractions by column chromatography using the ASE (Magill et al., 2015). PAHs were analyzed using an

Agilent 6890 GC with an Agilent 5973 quadrupole mass spectrometer (MS). PAHs were

39 identified in Full Scan mode based on authentic standards, NIST 98 spectral library, fragmentation patterns, and retention times and quantified in Selected Ion Monitoring (SIM) mode. The detection limit was 10 pg of PAH. Uncertainty in measurements was determined by treating additional analyses of PAH standards as unknowns and calculating the coefficient of variation (CV) of the concentrations. PAHs had an average CV of 12% of the mean concentration. PAH abundances were normalized to total organic carbon for each sample (μg/g

TOC).

For TOC analyses from BSN-1B and PCB-2A cores, ~2 g aliquots of powdered sample were decarbonated through acidification with 2N HCl (Midwood and Boutton, 1998). TOC was analyzed on a ThermoFinnigan Elemental Analyzer (EA) coupled to a Delta Plus Advantage

Continuous Flow Isotope Ratio Mass Spectrometer (IRMS).

For charcoal from the BSN-1A core, 10 g of uncrushed rock were macerated in acid to dissolve carbonate and silicate materials. The samples were sieved, and the 125-500 μm fraction was examined under a stereo microscope (Belcher et al., 2005). All charred particles that occurred in the samples were counted. Select samples were confirmed by scanning electron microscope (SEM). (See supplemental for details and uncertainties of our methods).

We make a molecular estimate of the proportion of refractory allochthonous carbon in the

PETM interval based on selective preservation and loss of organic phases using percent loss

(%Loss) calculations of intermediate refractory carbon (PAHs) and TOC (Figure 2-5, Table A-1, and Table A-2). For these estimates, if TOC was 100% comprised of intermediately refractory carbon (like PAHs), then points would plot on the 1:1 degradation line; if TOC contains less refractory carbon than PAHs, then points will plot to the left of the 1:1 line; if TOC contains more refractory carbon than PAHs (more resistant to degradation), then points will plot to the right of the 1:1 degradation line.

40 2.5. Results

2.5.1. Basin Substation

The PETM interval was constrained based on litho- and bio-stratigraphy (Baczynski,

2014; Baczynski et al., 2014a; Harrington et al., 2014; Baczynski et al., 2014b) (see supplemental for details). %TOC was variable in pre-PETM Paleocene (0.07 – 3%) and post-PETM Eocene samples (0.07 – 1%) samples, but was low in PETM samples (<0.2% C) (Figure 2-2 and Figure

A-2), where it decreased significantly relative to late Paleocene samples. Total PAH concentrations were on average more abundant in pre-PETM Paleocene samples (~160 μg/g

TOC) relative to Eocene samples (16 μg/g TOC), and there was no trend across the PETM interval (Figure 2-2). Concentrations in total PAHs were highest (2200 μg/g TOC) at 103 meters composite depth (mcd). PAH concentrations did not correlate with TOC (Figure A-3 and Figure

A-4). Charcoal counts showed a similar trend as PAH concentrations (Figure 2-2) with less charcoal observed in PETM samples than pre-PETM Paleocene and post-PETM Eocene samples.

We estimated that refractory allochthonous carbon composed 29% of the late Paleocene TOC and more than doubled to 74% in the PETM interval (Figure 2-5).

2.5.2. Polecat Bench

13 The PETM interval was constrained based on δ Ccarb data, and bio- and litho-stratigraphy

(Harrington et al., 2014; Bowen et al., 2015). %TOC was low and variable throughout the core

(Figure 2-3 and Figure A-2). Median PAH concentrations decreased in all PETM samples relative to both pre-PETM Paleocene and post-PETM Eocene samples (Figure 2-4). PAH concentrations peaked at 138 mcd, but otherwise no trend was observed across the PETM interval (Figure 2-3).

There was no correlation between PAH concentrations and TOC (Figure A-4). Polecat Bench

41 cores were not sampled for charcoal. We estimated that refractory allochthonous carbon composed 24% of TOC in the late Paleocene interval and nearly doubled to 43% in the PETM interval (Figure 2-5).

2.6. Discussion

For both Basin Substation and Polecat Bench cores, organic carbon content was lower within the PETM interval (Figure 2-2, Figure 2-3, and Figure A-2). The decreased %TOC trends in the cores are consistent with observations in outcrop at Polecat Bench (Magioncalda et al.,

2004) and other terrestrial sites (e.g., Piceance Creek Basin, Colorado (Foreman et al., 2012) and

Claret section, Spain (Domingo et al., 2009)) (Figure A-2). Polecat Bench, which is dominated by sequences of oxidized paleosols, has much lower %TOC in the late Paleocene interval than Basin

Substation, a site with wetter and more reducing conditions. The decreased organic carbon is most striking at Basin Substation (Figure 2-2). Here, %TOC decreased and was low throughout

PETM samples (<0.1%), average total PAH concentrations decreased by an order of magnitude and were low in all PETM samples, and a small amount of charcoal was found in the three PETM samples (Figure 2-2). The decreased TOC and PAH concentrations were concurrent with a three orders of magnitude decline in leaf-wax n-alkanes (Baczynski, 2014) and with poor preservation and at least an order magnitude decrease in the abundance of pollen (Harrington et al., 2014).

The decline in organic carbon content in PETM samples at both sites is likely the result of decreased preservation rather than sedimentary dilution or decreased production. Increased dilution by clastics should have resulted in higher rates of deposition during the PETM interval, but this is not supported by sediment accumulation rates. In the Bighorn Basin, long-term average sediment accumulation rates were 0.25 – 0.4 mkyr-1 through the Paleocene and Eocene (Clyde et al., 2013; Bowen et al., 2015) and were also similar (~0.19 – 0.38 mkyr-1) at both sites during

42 PETM (see supplemental for details); therefore, sedimentation rates on average increased no more than 1.5 times during the PETM interval. The reduction in TOC between before (average:

0.58 %TOC) and during the PETM (average: 0.12 %TOC) would require sedimentation rates to increase by a factor of ~5. Hence, sedimentary dilution cannot explain the decreased organic carbon in PETM samples.

The decreased organic carbon and pyrogenic carbon concentrations in PETM samples could be due to decreased biomass abundance and fewer fires, but how biomass abundance changed during the PETM is not clear. CO2-enrichment studies of modern plant productivity found that biomass increased with increasing atmospheric CO2 (Bowes, 1993). With the higher atmospheric CO2 during the PETM relative to the late Paleocene, biomass abundance could have increased during the PETM. Inconsistent with this, however, given that canopy closure increases with greater forest productivity (Larson et al., 2008), paleobotanical and palynological evidence suggests vegetation was more open during the PETM in the Bighorn Basin (McInerney and

Wing, 2011; Wing and Currano, 2013), likely due to drying (Wing et al., 2005; Kraus and

Riggins, 2007; Smith et al., 2007; Smith et al., 2008). A more patchy distribution of vegetation, and, therefore, lower fuel connectivity, could have decreased fire occurrence (Miller and Urban,

2000; Senici et al., 2015). Despite potential fertilization by increased CO2, biomass in the

Bighorn Basin could have declined due to some combination of higher temperatures and drying during the PETM.

TOC, PAH, charcoal, leaf-wax n-alkanes (Baczynski, 2014), and pollen (Harrington et al., 2014) abundance all decreased during the PETM interval relative to the late Paleocene, but not by the same magnitude, which suggests preservation, rather than production of organic matter, controlled patterns of decreased organic carbon. PAH concentrations do not correlate with lithology or TOC content (Figure A-4). Aqueous solubility of PAHs decreases almost logarithmically with increasing molecular mass, and therefore, 5-7 ring PAHs are significantly

43 less soluble than 2-4 ring counterparts (Figure B-1; May et al., 1978; Johnsen et al., 2005).

Hence, low molecular-weight (LMW) PAHs (≤ 4 rings) are more susceptible to degradation than high molecular-weight (HMW) PAHs (≥5 rings) (Killops and Massoud, 1992; Maliszewska- kordybach, 1993; Coulon et al., 2005). Limited studies suggest HMW PAHs (e.g., coronene) are associated with hotter temperature burns and LMW PAHs (e.g., pyrene) are produced at a range of temperatures (McGrath et al., 2003), although many factors influence PAH distributions

(Masclet et al., 1995).

HMW PAHs generally dominated both sections throughout the Paleocene-Eocene

(Figure A-4), which could indicate more intense fires occurred (primary signal) or the loss of

LMW PAHs (secondary signal). Pyrene concentrations decreased more in the PETM interval than coronene concentrations (Figure 2-4), which suggests the dominance of HMW PAHs may be due to preservation issues, such as loss of PAHs due to degradation by microorganisms or water leaching (Killops and Massoud, 1992), not intense fires. In the Bighorn Basin, where mean annual temperature increased ~5°C during the PETM (Wing et al., 2005), increased microbial activity (Lloyd and Taylor, 1994) likely enhanced degradation of organic matter by soil respiration (Baczynski et al.,2016).

There is further evidence of aggressive loss of primary and pyrogenic organic phases. At both sites, PAH concentrations decreased relative to TOC in PETM samples. This suggests that the organic carbon that is present during the PETM interval is even more refractory than PAHs and that fresher, more labile biomass was extensively degraded. The even more refractory carbon would be more resistant to decay than the large PAHs; in addition, this refractory material could have allochthonous inputs as noted by isotope and fossil evidence (Baczynski et al., 2013;

Bataille et al., 2013; Baczynski et al., 2016).

Using TOC and n-alkane isotope dat Baczynski et al. (2016) estimated that allochthonous carbon composed 28 to 63% of the total carbon during the PETM at the southeastern part of the

44 Bighorn Basin (Highway 16 site). Based on separating soil particles by density and TOC isotope data, Bataille et al., (2013) also estimated a higher proportion of allochthonous carbon during the

PETM at the Highway 16, Bighorn Basin site. We estimate increased proportional refractory carbon during the PETM interval at both Polecat Bench and Basin Substation, likely due to both greater inputs of allochthonous carbon and more extensive degradation of fresher, more labile organic carbon (Baczynski et al., 2013; Baczynski et al., 2016).

At Basin Substation, we estimate that refractory allochthonous carbon composed 29% of the late Paleocene TOC and more than doubled to 74% in the PETM interval (Figure 2-5). At

Polecat Bench, we estimate that refractory allochthonous carbon composed 24% of TOC in the late Paleocene interval and nearly doubled to 43% in the PETM interval (Figure 2-5). The varying proportion of allochthonous carbon before, during, and after the PETM event emphasizes that,

13 due to selective loss of organic phases, δ Corg values do not fully represent the carbon isotope excursion (in terms of magnitude and stratigraphic thickness) and that compound-specific isotope

13 13 values (e.g., δ Cn-alkanes) can provide a more robust record than δ Corg values (e.g., Baczynski et al., 2016).

The decreased preservation of organic carbon during the PETM interval suggests that a hotter, more seasonal climate reduced the stability of soil organic matter and facilitated the severe loss of soil carbon, including pyrogenic carbon. Our findings are consistent with modern experiments of short-term (weeks to a few years) temperature sensitivity of labile and non-labile soil organic carbon (Knorr et al., 2005) and provide long-term (thousands to tens of thousands years) evidence of enhanced soil decomposition of both labile and less labile organic carbon in a warming world. Such aggressive soil carbon loss presents challenges for modern long-term sequestration of carbon in soil, such as the addition of biochar. Further, evidence for severely reduced soil carbon, which decreases soil quality and plant productivity, supports concerns of global warming adversely affecting agricultural productivity and food security as plants are

45 increasingly heat and water stressed (Lal, 2004). We propose that intensified soil respiration exceeded terrestrial productivity during the PETM, hindered soil carbon sequestration, and possibly enhanced the warming.

2.7. Conclusion

In the Bighorn Basin, all forms of soil carbon decreased during the PETM, and PAH concentrations decreased even more than TOC, which suggests an even more refractory phase was present, such as allochthonous weathered kerogen. The aggressive loss of both soil organic carbon and pyrogenic carbon (i.e., PAHs and charcoal) during the PETM strongly indicates higher rates of organic carbon decay are linked with a hotter, more seasonal climate during the

PETM (Wing et al., 2005; Kraus and Riggins, 2007; Smith et al., 2007; Smith et al., 2008).

Overall, our results reflect that elevated temperatures caused greater loss rates by soil respiration, even beyond the effects of decreased productivity (i.e., less carbon input). This suggests that storing carbon in soils is difficult and warming soils may be a carbon source rather than a carbon sink, which can reduce soil fertility. The loss of soil carbon during the PETM potentially amplified the atmospheric greenhouse.

2.8. Figures

Figure 2-1. Map of the Bighorn Basin, Wyoming. Map shows exposures of the Willwood Formation (shaded) targeted for coring and the two coring sites (Polecat Bench and Basin Substation; black stars). Index map in lower left corner shows location of the Bighorn Basin in northern Wyoming. Map modified from Baczynski et al., (2013).

Figure 2-2. Core records of the PETM, Basin Substation, Wyoming. Left to right: Stratigraphic column of hole B, where colors represent the general colors of the rock; TOC (grey squares) from hole B with dotted lines for the median value of each interval and logarithmic-scale axis; sum of 12 PAHs from hole B (grey diamonds with black line), note break in axis between 500 ug/g TOC and 2200 ug/g TOC; charcoal from hole A (white circles and grey bar). Charcoal and PAH records are from different stratigraphic intervals in adjacent cores, but holes are on same meters composite depth (mcd) scale. Light grey box delineates the PETM. Conglom.: conglomerate

Figure 2-3. Core records of the PETM, Polecat Bench, Wyoming. Left to right: Stratigraphic column of hole A, where colors represents the general colors of the rock; pedogenic carbonate δ13C from hole A and B from Bowen et al., (2014) (white squares) with 5-point moving average of both (black line); TOC (grey squares) from hole A with dashed lines for the median value of each interval and logarithmic-scale axis; the PAH, coronene, representative PAH and largest PAH quantified, from hole A (grey diamonds with black line). Light grey box delineates the PETM. Conglom.: conglomerate.

Figure 2-4. Box and whisker plots of PAH data (pyrene and coronene) from Polecat Bench and Basin Substation cores. Ends of box represent 25% and 75% quartile; middle line is median; whiskers extend to minimum and maximum values (not including outliers); white squares represent outliers. Number in italics is n samples for each box. Light grey box delineates the PETM; Paleocene is pre-PETM and Eocene is post-PETM samples. Logarithmic-scale axis.

Figure 2-5. Estimate of the relative contributions of soil organic matter degradation and allochthonous carbon to PETM organic carbon as a percent of Paleocene total organic carbon (TOC) for each site. Top: Bivariate plot of TOC and total PAH. If TOC was 100% comprised of intermediately refractory carbon (like PAHs), then points would plot on the 1:1 degradation line (dashed line); if TOC contains less refractory carbon than PAHs, then points will plot to the left of the 1:1 line; if TOC contains more refractory carbon than PAHs, then points will plot to the right of the 1:1 loss line. Polecat Bench values (light-grey square); Basin Substation values (dark-grey circle); estimated percent allochthonous carbon (horizontal arrow*). Bottom: Bar graph shows percent of organic carbon type during the PETM as a percent of Paleocene TOC for each site. Bars represent different forms of carbon: allochthonous (black*), autochthonous (white, solid line), degraded (white, dashed line). Right of bar graphs reports the Paleocene median TOC values for each site. *For a given site, allochthonous horizontal arrow and allochthonous black bar are the same length (same % Paleocene TOC).

2.9. References

Baczynski, A.A., 2014, Evaluating Carbon Cycle Dynamics and Hydrologic Change during the Paleocene-Eocene Thermal Maximum, Bighorn Basin, Wyoming: Northwestern University, 219 p. Baczynski, A.A., Barclay, R.S., Bowen, G.J., Denis, E.H., Freeman, K.H., Harrington, G.J., Jardine, P.E., Maibauer, B., McInerney, F.A., and Wing, S.L., 2014a, Effects of the Paleocene-Eocene Thermal Maximum on terrestrial plants and carbon storage, in 10th North American Paleontological Convention, Paleontological Society, Florida Museum of Natural History. Baczynski, A.A., McInerney, F.A., Wing, S.L., Kraus, M.J., Bloch, J.I., Boyer, D.M., Secord, R., Morse, P.E., and Fricke, H.C., 2013, Chemostratigraphic implications of spatial variation in the Paleocene-Eocene Thermal Maximum carbon isotope excursion, SE Bighorn Basin, Wyoming: Geochemistry, Geophysics, Geosystems, v. 14, no. 10, p. 4133–4152, doi: 10.1002/ggge.20265. Baczynski, A.A., McInerney, F.A., Wing, S.L., Kraus, M.J., Morse, P.E., Bloch, J.I., Chung, A.H., and Freeman, K.H., 2016, Distortion of carbon isotope excursion in bulk soil organic matter during the Paleocene-Eocene thermal maximum: Geological Society of America Bulletin, , no. Xx, p. B31389.1, doi: 10.1130/B31389.1. Baczynski, A.A., McInerney, F.A., Wing, S.L., and the BBCP Science Team, 2014b, n-Alkane PETM records from the Bighorn Basin, Wyoming: A core-outcrop comparison: Rendiconti Online della Società Geologica Italiana, v. 31, p. 19–20, doi: 10.3301/ROL.2014.24. Bataille, C.P., Mastalerz, M., Tipple, B.J., and Bowen, G.J., 2013, Influence of provenance and preservation on the carbon isotope variations of dispersed organic matter in ancient floodplain sediments: Geochemistry, Geophysics, Geosystems, v. 14, no. 11, p. 4874– 4891, doi: 10.1002/ggge.20294. Belcher, C.M., Collinson, M.E., and Scott, A.C., 2005, Constraints on the thermal energy released from the Chicxulub impactor; new evidence from multi-method charcoal analysis: Journal of the Geological Society of London, v. 162, no. 4, p. 591–602, doi: 10.1144/0016-764904-104. Bird, M.I., Wynn, J.G., Saiz, G., Wurster, C.M., and McBeath, A., 2014, The Pyrogenic Carbon Cycle: Annual Review of Earth and Planetary Sciences, v. 43, p. 150223150959000, doi: 10.1146/annurev-earth-060614-105038. Bowen, G.J., Maibauer, B.J., Kraus, M.J., Röhl, U., Westerhold, T., Steimke, A., Gingerich, P.D., Wing, S.L., and Clyde, W.C., 2015, Two massive, rapid releases of carbon during the onset of the Palaeocene-Eocene thermal maximum: Nature Geosci, v. 8, no. 1, p. 44–47, doi: 10.1038/ngeo2316. Bowes, G., 1993, Facing the inevitable: Plants and increasing atmospheric C02: Annual Review of Plant Physiology and Plant Molecular Biology, v. 44, p. 309–332. Carvalhais, N., Forkel, M., Khomik, M., Bellarby, J., Jung, M., Migliavacca, M., Μu, M., Saatchi, S., Santoro, M., Thurner, M., Weber, U., Ahrens, B., Beer, C., Cescatti, A., et al., 2014, Global covariation of carbon turnover times with climate in terrestrial ecosystems: Nature, v. 514, no. 7521, p. 213–217, doi: 10.1038/nature13731.

52 Clechenko, E.R., Kelly, D.C., Harrington, G.J., and Stiles, C.A., 2007, Terrestrial records of a regional weathering profile at the Paleocene-Eocene boundary in the Williston Basin of North Dakota: Bulletin of the Geological Society of America, v. 119, no. 3-4, p. 428– 442, doi: 10.1130/B26010.1. Clyde, W.C., Gingerich, P.D., Wing, S.L., Röhl, U., Westerhold, T., Bowen, G., Johnson, K., Baczynski, A.A., Diefendorf, A., McInerney, F., Schnurrenberger, D., Noren, A., Brady, K., Acks, R., et al., 2013, Bighorn Basin Coring Project (BBCP): A continental perspective on early Paleogene hyperthermals: Scientific Drilling, v. 16, p. 21–31, doi: 10.5194/sd-16-21-2013. Cotton, J.M., Sheldon, N.D., Hren, M.T., and Gallagher, T.M., 2015, Positive feedback drives carbon release from soils to atmosphere during Paleocene/Eocene warming: American Journal of Science, v. 315, no. 4, p. 337–361, doi: 10.2475/04.2015.03. Coulon, F., Pelletier, E., Gourhant, L., and Delille, D., 2005, Effects of nutrient and temperature on degradation of petroleum hydrocarbons in contaminated sub-Antarctic soil: Chemosphere, v. 58, no. 10, p. 1439–1448, doi: 10.1016/j.chemosphere.2004.10.007. Crouch, E.M., Dickens, G.R., Brinkhuis, H., Aubry, M.P., Hollis, C.J., Rogers, K.M., and Visscher, H., 2003, The Apectodinium acme and terrestrial discharge during the Paleocene-Eocene thermal maximum: New palynological, geochemical and calcareous nannoplankton observations at Tawanui, New Zealand: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 194, no. 4, p. 387–403, doi: 10.1016/S0031- 0182(03)00334-1. Cui, Y., Kump, L.R., Ridgwell, A.J., Charles, A.J., Junium, C.K., Diefendorf, A.F., Freeman, K.H., Urban, N.M., and Harding, I.C., 2011, Slow release of fossil carbon during the Palaeocene–Eocene Thermal Maximum: Nature Geoscience, v. 4, no. 7, p. 481–485, doi: 10.1038/ngeo1179. Denis, E.H., Toney, J.L., Tarozo, R., Anderson, R.S., Roach, L.D., and Huang, Y., 2012, Polycyclic aromatic hydrocarbons (PAHs) in lake sediments record historic fire events: Validation using HPLC-fluorescence detection: Organic Geochemistry, v. 45, p. 7–17, doi: 10.1016/j.orggeochem.2012.01.005. Domingo, L., López-Martínez, N., Leng, M.J., and Grimes, S.T., 2009, The Paleocene-Eocene Thermal Maximum record in the organic matter of the Claret and Tendruy continental sections (South-central Pyrenees, Lleida, Spain): Earth and Planetary Science Letters, v. 281, no. 3-4, p. 226–237, doi: 10.1016/j.epsl.2009.02.025. Forbes, M.S., Raison, R.J., and Skjemstad, J.O., 2006, Formation, transformation and transport of black carbon (charcoal) in terrestrial and aquatic ecosystems: Science of the Total Environment, v. 370, no. 1, p. 190–206, doi: 10.1016/j.scitotenv.2006.06.007. Foreman, B.Z., Heller, P.L., and Clementz, M.T., 2012, Fluvial response to abrupt global warming at the Palaeocene/Eocene boundary.: Nature, v. 491, no. 7422, p. 92–5, doi: 10.1038/nature11513. Gingerich, P.D., 1989, New Earliest Wasatchian mammalian fauna from the Eocene of Northwestern Wyoming: Composition and diversity in a rarely sampled high-floodplain assemblage: University of Michigan Papers on Paleontology, v. 28, p. 1–97. Handley, L., O’Halloran, A., Pearson, P.N., Hawkins, E., Nicholas, C.J., Schouten, S., McMillan, I.K., and Pancost, R.D., 2012, Changes in the hydrological cycle in tropical East Africa during the Paleocene-Eocene Thermal Maximum: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 329-330, p. 10–21, doi: 10.1016/j.palaeo.2012.02.002. Harrington, G.J., Jardine, P.E., Wing, S.L., and the BBCP Science Team, 2014, Bighorn Basin Coring Project (BBCP): Pollen floral changes and organic matter from core – outcrop

53 comparisons through the PETM: Rendiconti Online della Società Geologica Italiana, v. 31, p. 95–96. Huxman, T.E., Smith, M.D., Fay, P. a, Knapp, A.K., Shaw, M.R., Loik, M.E., Smith, S.D., Tissue, D.T., Zak, J.C., Weltzin, J.F., Pockman, W.T., Sala, O.E., Haddad, B.M., Harte, J., et al., 2004, Convergence across biomes to a common rain-use efficiency.: Nature, v. 429, no. 6992, p. 651–654, doi: 10.1038/nature02561. Jaramillo, C., Ochoa, D., Contreras, L., Pagani, M., Carvajal-Ortiz, H., Pratt, L.M., Krishnan, S., Cardona, A., Romero, M., Quiroz, L., Rodriguez, G., Rueda, M.J., de la Parra, F., Morón, S., et al., 2010, Effects of rapid global warming at the Paleocene-Eocene boundary on neotropical vegetation: Science, v. 330, no. 6006, p. 957–961, doi: 10.1126/science.1193833. Johnsen, A.R., Wick, L.Y., and Harms, H., 2005, Principles of microbial PAH-degradation in soil: Environmental Pollution, v. 133, no. 1, p. 71–84, doi: 10.1016/j.envpol.2004.04.015. Karhu, K., Auffret, M.D., Dungait, J. a J., Hopkins, D.W., Prosser, J.I., Singh, B.K., Subke, J.-A., Wookey, P. a, Agren, G.I., Sebastià, M.-T., Gouriveau, F., Bergkvist, G., Meir, P., Nottingham, A.T., et al., 2014, Temperature sensitivity of soil respiration rates enhanced by microbial community response.: Nature, v. 513, no. 7516, p. 81–4, doi: 10.1038/nature13604. Killops, S.D., and Massoud, M.S., 1992, Polycyclic aromatic hydrocarbons of pyrolytic origin in ancient sediments: evidence for Jurassic vegetation fires: Organic Geochemistry, v. 18, no. 1, p. 1–7, doi: 10.1016/0146-6380(92)90137-M. Knorr, W., Prentice, I., House, J., and Holland, E., 2005, Long-term sensitivity of soil carbon turnover to warming: Nature, v. 433, no. January, p. 298–301, doi: 10.129/2002PA000837. Koch, P.L., Zachos, J.C., and Gingerich, P.D., 1992, Correlation between isotope records in marine and continental carbon reservoirs near the Palaeocene/Eocene boundary: Nature, v. 358, no. 6384, p. 319–322, doi: 10.1038/358319a0. Kraus, M.J., McInerney, F.A., Wing, S.L., Secord, R., Baczynski, A.A., and Bloch, J.I., 2013, Paleohydrologic response to continental warming during the Paleocene-Eocene Thermal Maximum, Bighorn Basin, Wyoming: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 370, p. 196–208, doi: 10.1016/j.palaeo.2012.12.008. Kraus, M.J., and Riggins, S., 2007, Transient drying during the Paleocene-Eocene Thermal Maximum (PETM): Analysis of paleosols in the bighorn basin, Wyoming: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 245, no. 3-4, p. 444–461, doi: 10.1016/j.palaeo.2006.09.011. Kraus, M.J., Woody, D.T., Smith, J.J., and Dukic, V., 2015, Alluvial response to the Paleocene- Eocene Thermal Maximum climatic event, Polecat Bench, Wyoming (U.S.A.): Palaeogeography, Palaeoclimatology, Palaeoecology, v. 435, p. 177–192, doi: 10.1016/j.palaeo.2015.06.021. Krull, E.S., Baldock, J.A., and Skjemstad, J.O., 2003, Importance of mechanisms and processes of the stabilisation of soil organic matter for modelling carbon turnover: Functional Plant Biology, v. 30, no. 2, p. 207–222, doi: 10.1071/FP02085. Lal, R., 2004, Soil carbon sequestration impacts on global climate change and food security: Science, v. 304, no. 5677, p. 1623–1627, doi: 10.1126/science.1097396. Larson, A.J., Lutz, J.A., Gersonde, R.F., Franklin, J.F., and Hietpas, F.F., 2008, Potential site productivity influences the rate of forest structural development: Ecological Applications, v. 18, no. 4, p. 899–910, doi: 10.1890/07-1191.1.

54 Lehmann, J., Gaunt, J., and Rondon, M., 2006, Bio-char sequestration in terrestrial ecosystems - A review: Mitigation and Adaptation Strategies for Global Change, v. 11, no. 2, p. 403– 427, doi: 10.1007/s11027-005-9006-5. Leirós, M.C., Trasar-Cepeda, C., Seoane, S., and Gil-Sotres, F., 1999, Dependence of mineralization of soil organic matter on temperature and moisture: Soil Biology and Biochemistry, v. 31, no. 3, p. 327–335, doi: 10.1016/S0038-0717(98)00129-1. Lloyd, J.J., and Taylor, J.A., 1994, On the temperature dependence of soil respiration: Functional Ecology, v. 8, no. 3, p. 315–323, doi: 10.2307/2389824. Luo, Y., 2007, Terrestrial Carbon–Cycle Feedback to Climate Warming: Annual Review of Ecology, Evolution, and Systematics, v. 38, no. 1, p. 683–712, doi: 10.1146/annurev.ecolsys.38.091206.095808. Von Lützow, M., Kögel-Knabner, I., Ekschmitt, K., Matzner, E., Guggenberger, G., Marschner, B., and Flessa, H., 2006, Stabilization of organic matter in temperate soils: Mechanisms and their relevance under different soil conditions - A review: European Journal of Soil Science, v. 57, no. 4, p. 426–445, doi: 10.1111/j.1365-2389.2006.00809.x. Magill, C.R., Denis, E.H., and Freeman, K.H., 2015, Rapid sequential separation of sedimentary lipid biomarkers via selective accelerated solvent extraction: Organic Geochemistry, v. 88, p. 29–34, doi: 10.1016/j.orggeochem.2015.07.009. Magioncalda, R., Dupuis, C., Smith, T., Steurbaut, E., and Gingerich, P.D., 2004, Paleocene- Eocene carbon isotope excursion in organic carbon and pedogenic carbonate: Direct comparison in a continental stratigraphic section: Geology, v. 32, no. 7, p. 553–556, doi: 10.1130/G20476.1. Maliszewska-kordybach, B., 1993, The effect of temperature on the rate of disappearance of polycyclic aromatic hydrocarbons from soils: v. 79, p. 15–20. Masclet, P., Cachier, H., Liousse, C., and Wortham, H., 1995, Emissions of polycyclic aromatic hydrocarbons by savanna fires: Journal of Atmospheric Chemistry, v. 22, no. 1-2, p. 41– 54, doi: 10.1007/BF00708180. May, W.E., Wasik, S.P., and Freeman, D.H., 1978, Determination of the Solubility Behavior of Some Polycyclic Aromatic Hydrocarbons in Water: Analytical chemistry, v. 50, no. 7, p. 997–1000, doi: 10.1021/ac50029a042. McGrath, T.E., Chan, W.G., and Hajaligol, M.R., 2003, Low temperature mechanism for the formation of polycyclic aromatic hydrocarbons from the pyrolysis of cellulose: Journal of Analytical and Applied Pyrolysis, v. 66, no. 1-2, p. 51–70, doi: 10.1016/S0165- 2370(02)00105-5. McInerney, F.A., and Wing, S.L., 2011, The Paleocene-Eocene Thermal Maximum: A Perturbation of Carbon Cycle, Climate, and Biosphere with Implications for the Future: Annual Review of Earth and Planetary Sciences, v. 39, no. 1, p. 489–516, doi: 10.1146/annurev-earth-040610-133431. Midwood, A.J., and Boutton, T.W., 1998, Soil carbonate decomposition by acid has little effect on δ13C of organic matter: Soil Biology and Biochemistry, v. 30, no. 10-11, p. 1301– 1307, doi: 10.1016/S0038-0717(98)00030-3. Miller, C., and Urban, D.L., 2000, Connectivity of forest fuels and surface fire regimes: Landscape Ecology, v. 15, no. 2, p. 145–154, doi: 10.1023/A:1008181313360. Pagani, M., Pedentchouk, N., Huber, M., Sluijs, A., Schouten, S., Brinkhuis, H., Sinninghe Damsté, J.S., and Dickens, G.R., 2006, Arctic hydrology during global warming at the Palaeocene/Eocene thermal maximum: Nature, v. 442, no. 7103, p. 671–675, doi: 10.1038/nature05043. Page, D.S., Boehm, P.D., Douglas, G.S., Bence, A.E., Burns, W.A., and Mankiewicz, P.J., 1999, Pyrogenic Polycyclic Aromatic Hydrocarbons in Sediments Record Past Human Activity:

55 A Case Study in Prince William Sound, Alaska: Marine Pollution Bulletin, v. 38, no. 4, p. 247–260, doi: 10.1016/S0025-326X(98)00142-8. Ramdahl, T., 1983, Retene - a molecular marker of wood combustion in ambient air: Nature, v. 306, p. 580–582, doi: 10.1038/306580a0. Robert, C., and Kennett, J.P., 1994, Antarctic subtropical humid episode at the Paleocene-Eocene boundary: Clay-mineral evidence: Geology, v. 22, no. 3, p. 211, doi: 10.1130/0091- 7613(1994)022<0211:ASHEAT>2.3.CO;2. Röhl, U., Westerhold, T., Bralower, T.J., and Zachos, J.C., 2007, On the duration of the Paleocene-Eocene thermal maximum (PETM): Geochemistry, Geophysics, Geosystems, v. 8, no. 12, doi: 10.1029/2007GC001784. Rose, K.D., Chew, A.E., Dunn, R.H., Kraus, M.J., Fricke, H.C., and Zack, S.P., 2012, Earliest Eocene mammalian fauna from the Paleocene-Eocene Thermal Maximum at Sand Creek Divide, Southern Bighorn Basin, Wyoming: University of Michigan Papers on Paleontology, v. 36, p. 1–122. Schmidt, M.W.I., and Noack, A.G., 2000, Analysis, distribution, implications, and current challenges: Global Biogeochemical Cycles, v. 14, no. 3, p. 777–793. Schuur, E.A.G., Bockheim, J., Canadell, J.G., Euskirchen, E., Field, C.B., Goryachkin, S. V, Hagemann, S., Kuhry, P., Lafleur, P.M., Lee, H., Mazhitova, G., Nelson, F., Rinke, A., Romanovsky, V.E., et al., 2008, Vulnerability of permafrost carbon to climate change: Implications for the global carbon cycle: BioScience, v. 58, no. September, p. 701–714, doi: 10.1641/B580807. Secord, R., Gingerich, P.D., Lohmann, K.C., and Macleod, K.G., 2010, Continental warming preceding the Palaeocene-Eocene thermal maximum: Nature, v. 467, no. 7318, p. 955– 958, doi: 10.1038/nature09441. Senici, D., Chen, H.Y.H., Bergeron, Y., and Ali, A.A., 2015, The effects of forest fuel connectivity on spatiotemporal dynamics of Holocene fire regimes in the central boreal forest of North America: Journal of Quaternary Science, v. 30, no. 4, p. 365–375, doi: 10.1002/jqs.2790. Smith, J.J., Hasiotis, S.T., Kraus, M.J., and Woody, D.T., 2008, Relationship of floodplain ichnocoenoses to paleopedology, paleohydrology, and paleoclimate in the Willwood Formation, Wyoming, during the Paleocene-Eocene Thermal Maximum: Palaios, v. 23, no. 9-10, p. 683–699, doi: 10.2110/palo.2007.p07-080r. Smith, F.A., Wing, S.L., and Freeman, K.H., 2007, Magnitude of the carbon isotope excursion at the Paleocene-Eocene thermal maximum: The role of plant community change: Earth and Planetary Science Letters, v. 262, no. 1-2, p. 50–65, doi: 10.1016/j.epsl.2007.07.021. Westerhold, T., Röhl, U., and Laskar, J., 2012, Time scale controversy: Accurate orbital calibration of the early Paleogene: Geochemistry, Geophysics, Geosystems, v. 13, no. 6, p. 1–19, doi: 10.1029/2012GC004096. Wing, S.L., and Currano, E.D., 2013, Plant response to a global greenhouse event 56 million years ago: American Journal of Botany, v. 100, no. 7, p. 1234–1254, doi: 10.3732/ajb.1200554. Wing, S.L., Harrington, G.J., Smith, F.A., Bloch, J.I., Boyer, D.M., and Freeman, K.H., 2005, Transient Floral Change and Rapid Global Warming at the Paleocene-Eocene Boundary: Science, v. 310, no. 5750, p. 993–996, doi: 10.1126/science.1116913. Yunker, M.B., Macdonald, R.W., Vingarzan, R., Mitchell, R.H., Goyette, D., and Sylvestre, S., 2002, PAHs in the Fraser River basin: A critical appraisal of PAH ratios as indicators of PAH source and composition: Organic Geochemistry, v. 33, no. 4, p. 489–515, doi: 10.1016/S0146-6380(02)00002-5.

56 Zachos, J.C., Nicolo, M., Raffi, I., Lourens, L.J., McCarren, H., and Kroon, D., 2005, Rapid Acidification of the Ocean During the Paleocene-Eocene Thermal Maximum: Science, v. 308, p. 1611–1615, doi: 10.1126/science.1109004. Zeebe, R.E., Ridgwell, A., and Zachos, J.C., 2016, Anthropogenic carbon release rate unprecedented during the past 66 million years: Nature Geoscience, v. 9, no. April, p. 325–329, doi: 10.1038/ngeo2681. Zeebe, R.E., and Zachos, J.C., 2013, Long-term legacy of massive carbon input to the Earth system: Anthropocene versus Eocene: Philosophical transactions. Series A, Mathematical, physical, and engineering sciences, v. 371, no. 2001, p. 20120006, doi: 10.1098/rsta.2012.0006.

Chapter 3

Widespread soil carbon loss during the Paleocene-Eocene Thermal Maximum

3.1. Abstract

The Paleocene-Eocene Thermal Maximum (PETM) warming event is marked by a negative carbon isotope excursion (CIE). The magnitude and shape of the CIE varies among

13 carbon archives, but generally bulk organic carbon isotopes (δ Corg) in terrestrial sections are among the most attenuated, and are often highly variable and distorted. Multiple lines of evidence from the Bighorn Basin, Wyoming, USA revealed that extensive organic carbon degradation and increased refractory or allochthonous carbon inputs (~28-74% of total organic carbon) diminished

13 the δ Corg excursion. In order to test if this was spatially widespread and to understand possible underlying mechanisms, we examined the abundance of total organic carbon (%TOC) and polycyclic aromatic hydrocarbons (PAHs) as metrics of soil degradation at a neighboring terrestrial basin, the Piceance Basin, Colorado. Outcrop samples reveal both diminished %TOC and PAH concentrations across the Paleocene-Eocene boundary. Using PAHs as a proxy for intermediately refractory carbon, such as found in mineral soils, evidence from the Piceance

Basin supports both increased organic carbon degradation and enhanced preservation of refractory fossil allochthonous carbon. Correlations between %TOC and elemental oxides (e.g.,

Al2O3 and TiO2) suggest soil organic carbon stabilization was associated with clay minerals. We hypothesize that decreased clay content in the soils (which inhibited stability of fresh carbon) and fluctuations in soil moisture (which destabilized older, refractory carbon), in conjunction with increased temperatures (which increased microbial decomposition rates), contributed to a reduction in soil organic matter preservation during the PETM. These mechanisms destabilized

58 carbon on millennial timescales and, with sustained higher temperatures across the PETM (~150 thousand years), increased soil carbon degradation persisted for tens of thousands of years. As temperatures warmed and remained warmer than the Paleocene, soils served as a sustained source of CO2 to the atmosphere. The loss of more than 50% of TOC relative to before the hyperthermal event implies greater soil organic carbon degradation can be a significant consequence as modern global temperatures rise.

3.2. Introduction

The PETM, a geologically abrupt period of climate change 55.5 million years ago

(Westerhold et al., 2012), provides a useful, but possibly less intense, geologic analog for modern climate change (Cui et al., 2011; Zeebe and Zachos, 2013; Zeebe et al., 2016). A negative carbon isotope excursion (CIE) marks the PETM and signifies a major perturbation to the carbon cycle

(McInerney and Wing, 2011). The CIE varies in magnitude and shape in different carbon

13 archives, and in particular, bulk organic carbon isotopes (δ Corg) values in terrestrial sections have high variability on a local scale (Baczynski et al., 2013; Baczynski et al., 2016). This can cause truncation of the CIE (Baczynski et al., 2013), underestimation of the length of the PETM, and it can even obscure the CIE altogether (Maibauer, 2013). Within the well-studied Paleocene-

Eocene floodplain deposits of the Bighorn Basin, Wyoming (Gingerich, 1989; Koch et al., 1992;

Magioncalda et al., 2004; Wing et al., 2005; Kraus and Riggins, 2007; Smith et al., 2007; Clyde

13 et al., 2013) the variability, reduced magnitude, and temporal truncation of the δ Corg excursion signal has been attributed to soil organic carbon degradation (Cotton et al., 2015; Baczynski et al.,

2016; Denis et al., in prep) and incorporation of refractory allochthonous fossil carbon (i.e., eroded from sedimentary rocks exposed in the hinterland) during the PETM (Baczynski et al.,

2013; Bataille et al., 2013; Baczynski et al., 2016; Denis et al., in prep). Increased soil carbon

59 degradation is a growing concern for modern climate change because CO2 release from soils can enhance warming and because diminished sequestration of carbon in soils can reduce soil fertility and threaten global food security (Lal, 2004; Heath et al., 2005; Knorr et al., 2005; Lehmann et al., 2006).

During the PETM, at least 3,000 Pg of 13C-depleted carbon was released into the atmosphere over ~10 ky and global temperatures rose ~5-8°C over ~170 ky (Wing et al., 2005;

Zachos et al., 2005; Röhl et al., 2007; Secord et al., 2010; Cui et al., 2011; McInerney and Wing,

2011). There were dramatic shifts in hydrologic and vegetation patterns (Wing et al., 2005;

Pagani et al., 2006b; Kraus and Riggins, 2007; Smith et al., 2007; Smith et al., 2008; Jaramillo et al., 2010; Foreman et al., 2012; Handley et al., 2012; Wing and Currano, 2013). There are many indications for the enhancement of the hydrological cycle and increased seasonality in precipitation from increased run-off from continental margins, increased clay deposition (Robert and Kennett, 1994; Crouch et al., 2003), and large extensive sand-bodies in terrestrial settings

(e.g., Claret Conglomerate, Spain (Domingo et al., 2009), Piceance Basin, Colorado (Foreman et al., 2012), and Bighorn Basin (Foreman, 2014)).

High atmospheric CO2 and warming (e.g., ~5°C) during the PETM could have driven increased plant biomass accumulation. Modern field, greenhouse, and incubation plant studies (on the timescale of months to years) found increased plant biomass with higher CO2 and warmer temperatures (Nobel and Hartsock, 1986; Nobel and Cortazar, 1991; Bowes, 1993; Rustad et al.,

2001). In CO2-enrichment studies, plants accumulated 30% more biomass when atmospheric CO2 was doubled (pCO2 was increased from ~350 ppm to ~650 ppm) (Nobel and Hartsock, 1986;

Nobel and Cortazar, 1991; Bowes, 1993). A meta-analysis of data from multiple biomes (e.g., tundra, grassland, forest), without any experimental control on precipitation, observed on average

~20% increased plant productivity with warming up to 6°C over multiple years (Rustad et al.,

2001). Although how these observations would scale to hundreds to tens of thousands of years is

60 unclear, they suggest that PETM conditions had potential to enhance plant growth and biomass input to the soils.

Terrestrial sediments that span the PETM tend to have low %TOC, with values observed on the order of 0.01 – 1% and typically lower during the event than the late Paleocene or early

Eocene (Magioncalda et al., 2004; Clechenko et al., 2007; Domingo et al., 2009; Foreman et al.,

2012; Maibauer, 2013). Dramatic decreases in TOC, leaf-wax n-alkanes, pyrogenic carbon (i.e., polycyclic aromatic hydrocarbons (PAHs) and charcoal), and pollen abundance at Basin

Substation in the Bighorn Basin suggested a severe loss of labile and intermediately refractory carbon and enhanced microbial respiration (Denis et al., in prep). Baczynski et al. (2016) used isotopic evidence to suggest organic carbon degradation rates doubled during the PETM, the proportion of allochthonous organic carbon increased (~28-63%), and that both attenuated the magnitude of the recorded CIE.

Soil carbon dynamics can vary significantly on local scales, and it is unknown whether the aggressive soil carbon loss observed in the Bighorn Basin was a more widespread phenomenon in Laramide foreland basins. Further, the potential geographic extent is hard to estimate without better understanding of mechanisms that contributed to the destabilization of carbon. Previous studies have focused on floodplain-dominated fluvial sites within the Bighorn

Basin (Highway 16, Polecat Bench, and Basin Substation) (Bataille et al., 2013; Baczynski et al.,

2016; Denis et al., in prep), but it is unclear how organic carbon was affected (in terms of soil carbon loss and allochthonous carbon inputs) in other basins and in channel-dominated fluvial sites, such as the Piceance Basin, Colorado. Differences in soil moisture conditions between sites, such as between Polecat Bench and Basin Substation (which had fewer oxidized paleosols), could also affect the extent of organic carbon degradation.

Because organic matter is inherently unstable on the order of days to tens of years, organic matter accumulation requires some form of protection from degradation (Lorenz et al.,

61 2007). The three main types of protection mechanisms that inhibit organic carbon degradation in soils are the inherent chemical recalcitrance of the compound, the accessibility of the compound to degradation, and physical protection through interactions between other organic compounds and minerals (Von Lützow et al., 2006; Von Lützow et al., 2008; Marschner et al., 2008).

Molecular recalcitrance is the resistance of organic matter to degradation due to the structure, chemical, and physical properties of the molecule (Von Lützow et al., 2006; Schmidt et al., 2011). Selective preservation based on inherent chemical compound characteristics is particularly important on the initial phase of degradation (e.g., years), largely dictated by a compound’s solubility in water because most microbial processes entail the soluble phase (Von

Lützow et al., 2006; Marschner et al., 2008). Persistence on longer timescales (hundreds to tens of thousands of years) is not just due to structural characteristics but other stabilization mechanisms

(Von Lützow et al., 2006; Lawrence et al., 2015).

Spatial inaccessibility of organic matter to microorganisms, enzymes, water, and oxygen inhibits carbon degradation. The aggregation of organic matter with soil particles reduces the access of microorganisms and enzymes to the organic carbon and reduces diffusion of enzymes and oxygen (Von Lützow et al., 2006). When soil aggregates are disrupted by bioturbation, erosion, or water, organic matter can be more available for degradation (Von Lützow et al., 2006;

Schmidt et al., 2011). In addition, organic matter is stabilized by interaction with mineral surfaces and metal ions. Finer grained particles (e.g., clays) have more surface area for organic matter to attach to (Christensen, 2001; Von Lützow et al., 2006; Von Lützow et al., 2008). Organic-mineral interactions can persist on 100 – 10,000 year timescales (Torn et al., 1997; Chaopricha and

Marín-Spiotta, 2014).

There are three main categories of destabilizing effects. The first is biogeochemical processes in the litter or depositional zone, due to changing biomass inputs or composition from shifting plant communities, and changing soil micro- or macro-fauna. An additional supply of

62 fresh organic matter from higher plant productivity, called ‘priming’, can enhance microbial growth and enzyme production and expand the ability or likelihood that some microorganisms metabolized less favorable compounds (Fontaine et al., 2003; Marschner et al., 2008; Guenet et al., 2012; Don et al., 2013). The plant community can affect primary productivity, major and trace nutrients (e.g., N and Ca), molecular composition of the organic matter, and the proportions of above and belowground biomass (Kögel-Knabner, 2002; Crow et al., 2009; Pisani et al., 2016).

Belowground productivity (e.g., roots) can help maintain soil structure and reduce soil carbon degradation (Crow et al., 2009). A shift in the microbial community structure can affect the type of compounds that are degraded and the extent of degradation (Schmidt et al., 2011; Pisani et al.,

2016).

The second set of destabilizing effects is changes in chemical-physical conditions in the litter and mineral soil environment, such as due to wet/dry cycles, moisture, kinetics of decomposition with temperature, and oxygen solubility. The drying and rewetting of soils can destabilize the association between organic matter and clay minerals (Oades, 1988). As clay mineral structures expand and contract with the addition and loss of water, organic matter desorbs from the clays and is more accessible to microorganisms (Oades, 1988; Schimel et al., 2011). The third is weathering processes and mineral properties in the mineral soil relevant for long-term storage (>1,000 years) conditions. Soils with a higher clay content retain higher amounts of organic carbon because of increased surface area for organic-mineral interactions (Oades, 1988;

Sollins et al., 1996; Von Lützow et al., 2008). How differences in clay mineralogy affect organic carbon preservation, though, is unclear (Oades, 1988; Sollins et al., 1996; Wattel-Koekkoek et al.,

2003; Von Lützow et al., 2008).

To estimate organic carbon degradation in the Piceance Basin during the PETM, we examined soil carbon and intermediately refractory carbon relationships across the PETM using

TOC and PAHs. PAHs are combustion byproducts that are widely distributed as volatiles and

63 particulates (Masclet et al., 1995). As a primary signal, PAHs can indicate fire in the paleorecord

(Ramdahl, 1983; Killops and Massoud, 1992; Page et al., 1999; Yunker et al., 2002; Denis et al.,

2012). As a secondary signal, PAHs serve as a metric for the degradation of refractory carbon relative to TOC (Denis et al., in prep). Larger PAHs are a form of intermediate refractory carbon

(relatively resistant to degradation) (Johnsen et al., 2005). Like charcoal, which has long residence times in soils on the order of 500 – 10,000 years (Schmidt and Noack, 2000; Forbes et al., 2006; Von Lützow et al., 2006), high molecular weight PAHs (≥ 5 rings) do not degrade readily due to their aromatic structure and low solubility (May et al., 1978; Coulon et al., 2005;

Johnsen et al., 2005; Von Lützow et al., 2006). The PAH and TOC calculation for degradation and refractory allochthonous carbon inputs applied in the Bighorn Basin (Denis et al., in prep) corroborated estimates of allochthonous carbon inputs using n-alkane and TOC isotope data, organic petrography, and fossil evidence (Bataille et al., 2013; Baczynski et al., 2016).

Using outcrop samples collected along a stratigraphic section in the Piceance Basin, we test the hypothesis that increased degradation and allochthonous carbon inputs distorted the CIE recorded in bulk soil organic matter during the PETM. We hypothesize that organic carbon content correlates with major elemental concentrations that are associated with clay minerals

(e.g., Al, Si, Ti) and decreased clay content facilitated organic carbon degradation. If soil moisture is a control on soil organic carbon degradation, then sites with similar soil moisture conditions (e.g., Basin Substation and Piceance Basin) would have a similar percent of TOC loss during the PETM. This study builds a comparative dataset to the Bighorn Basin to provide insights on the effects of environmental and depositional conditions on the preservation of labile and intermediately refractory carbon at a regional scale and the mechanisms that contributed to organic carbon destabilization.

64 3.3. Study Site

The Piceance Basin, located in northwestern Colorado (Figure 3-1), is an intermontane basin that formed during the Laramide orogeny (Dickinson et al., 1988). Samples were collected along the stratigraphic section described in Foreman et al., (2012) (PCBa; N 39°16.838’ W

108°11.667’). Samples were from a ~105 m thick interval that spans the late Paleocene to early

Eocene within the Wasatch Formation (Atwell Gulch and Molina members). The PETM is

13 constrained based on δ Corg negative carbon excursion (CIE) and biostratigraphy (Johnson and

May, 1978; Burger and Honey, 2008; Burger, 2012; Foreman et al., 2012).

The stratigraphic section is composed of fluvial sand-bodies and associated floodplain deposits dominated by well-developed soil horizons. The Atwell Gulch Member (Paleocene and onset of PETM) is a mud-dominated succession with purple, orange, and red paleosols. Thin, laterally restricted fluvial sand-bodies were interpreted as shallow and narrow rivers (Foreman et al., 2012). The Molina Member (PETM into Eocene) is dominated by purple paleosols, which suggested more poorly drained soils (Foreman et al., 2012). Prevalent thick, laterally extensive fluvial sand-bodies, with variable flow depths, were interpreted to potentially record greater mean annual precipitation or increased seasonal precipitation that altered river behavior and deposition

(Foreman et al., 2012).

The Piceance Basin provides an opportunity to evaluate the effect of precipitation and soil moisture conditions on organic carbon preservation by comparing and contrasting our results with those from two sites in the Bighorn Basin: Polecat Bench and Basin Substation. Basin

Substation was inferred to be a wetter site than Polecat Bench based on lower topography and fewer oxidized paleosols (Denis et al., in prep), but both experienced drying and increased seasonal precipitation during the PETM (Wing et al., 2005; Kraus and Riggins, 2007; Smith et al.,

65 2008). In comparision, the Piceance Basin was wetter overall, but also may have experienced increased seasonal precipitation during the PETM (Foreman, 2012; Foreman et al., 2012).

3.4. Methods

The samples analyzed in this study include 21 of those investigated by Foreman et al.

(2012) and 44 additional samples collected along the same stratigraphic section. The additional sampling focused on finer grain materials (clays and silts), especially in sand-dominated sections, in order to assure a variety of lithologies were collected and reduce lithologic bias. Samples were obtained by trenching until fresh rock was exposed and located within the section using distinctive beds and a Jacob’s staff. For geochemical analyses, the exterior of rock samples were rinsed with dichloromethane (DCM), freeze-dried, and ground to a powder with a ball mill.

13 For TOC and δ Corg analyses, ~1 g aliquots of powdered sample were decarbonated through acidification with 1N HCl (Midwood and Boutton, 1998). Samples were reacted with 1N

HCl in centrifuge tubes for 1 hour with occasional vortex mixing until no reaction was visible, and checked for complete reaction by the addition of 20 ml of acid. The mixture was then diluted with deionized water, centrifuged, and decanted repeatedly until the supernatant reached a neutral pH. Decarbonated samples were freeze-dried, and then weighed (5 mg to 40 mg depending on

TOC values) into tin capsules for analysis on a Costech elemental analyzer (EA) coupled to a

Thermo Finnigan Delta XP isotope ratio mass spectrometer (IRMS). Analyses were performed via continuous-flow (He; 120 ml/min), with oxidation at 1020°C over chromium (III) oxide and silvered cobalt (II, III) oxide, followed by reduction over elemental copper at 650°C, and CO2 passed through a water trap before entering the IRMS. For TOC, the mean standard deviation for replicate analyses of samples was 0.01%C (n=56). Measured δ13C values were corrected for sample size and reported relative to Vienna Pee Dee Belemnite (VPDB) based on a two-point

66 calibration (Coplen et al., 2006) and internal standards (international standards: IAEA CH-6 sucrose and USGS24 graphite; lab standards: caffeine, corn starch, and Peruvian mud). Error was determined by analyzing standards as samples in all EA runs and based on replicate analyses of samples. The mean standard deviation for replicate analyses of standards was 0.2‰ (n=94) and for replicate analyses of samples was 0.2‰ (n=56).

Lipids were extracted from ~5 to 20 g of powdered rock using an Accelerated Solvent

Extractor (ASE) with DCM:methanol (MeOH) (v:v, 9:1) at 100°C and 1500 psi. The total lipid extract (TLE) was concentrated with a gentle stream of nitrogen gas and was separated using selective ASE (Magill et al., 2015). ASE 33 ml cell columns were constructed with silver- impregnated alumina (2 g), silica gel (5.5 g), a filter loaded with reconstituted TLE in DCM, silica gel (1 g) and filled with sand. ASE columns were selectively extracted following Magill et al. (2015): saturated hydrocarbons - 100% n-hexane, 30% flush volume (~10 ml), and back-eluted for, unsaturated/aromatics - hexane:DCM (v:v, 9:1) and polar - DCM:MeOH (v:v, 7:3) for three aliquots of solvent at 60% flush volume (~20 ml) each.

PAHs were analyzed from TLE and F2 fraction using an Agilent 6890 GC with an

Agilent 5973 quadrupole mass spectrometer (MS) and a fused silica capillary column (Agilent

J&W DB-5; 30 m, 250 μm, 0.25 μm). The column flow rate was 2.0 ml/min and the oven program started at 60°C for 1 min, ramped to 320°C at 6°C/min, and a final hold time of 15 min.

The MS had an ionization energy of 70 eV with a scanning mass range of m/z 40-700 in Full Scan mode. PAHs were identified in Full Scan mode based on authentic standards, NIST 98 spectral library, fragmentation patterns, and retention times and quantified in Selected Ion Monitoring

(SIM) mode. The following ions were monitored: (m/z) 102, 128; 151, 152, 153, 154; 165, 166,

183, 198; 176, 178; 101, 202, 219, 234; 114, 126, 228, 253, 254; 125, 126, 252; 276, 278; 150,

300. The detection limit was 10 pg of PAH. Compound peak areas were converted to mass quantities using response curves for 19 PAH external standards analyzed in concentrations

67 ranging from 0.01 to 1 ng/μl. Uncertainty in measurements was determined by treating additional analyses of external PAH standards as unknowns and calculating the coefficient of variation (CV) of the concentrations, where CV is the standard deviation divided by the mean multiplied by

100%. The CV for each PAH standard ranged from 8-17% with an average CV of 13% of the mean concentration. PAH abundances were normalized to total organic carbon for each sample

(μg/g TOC).

Degradation of TOC and allochthonous carbon input during the PETM interval relative to the Paleocene was estimated based on molecular percent loss (%Loss) calculations following

Denis et al., (in prep). Calculations were based on median TOC content (g C/g rock) and median

PAH concentrations (ng/g rock) in samples from the Paleocene and from the PETM. We make a molecular estimate of the proportion of refractory allochthonous carbon in the PETM interval based on selective preservation and loss of organic phases using percent loss (%Loss) calculations of intermediate refractory carbon (PAHs) and TOC (Table B-1). For these estimates, if TOC was 100% comprised of intermediately refractory carbon (like PAHs), then points would plot on the 1:1 degradation line; if preserved TOC contains less refractory carbon than PAHs, then points will plot to the left of the 1:1 line; if TOC contains more refractory carbon than PAHs

(more resistant to degradation), then points will plot to the right of the 1:1 degradation line. We compared the observed ratio values to a 1:1 degradation ratio of PAH %Loss to TOC %Loss. The discrepancy between expected and actual TOC %Loss to the right of the 1:1 degradation line represented the percent of refractory allochthonous carbon in Paleocene TOC. We assumed that the absolute amount of allochthonous input did not change from the Paleocene to the PETM in order to estimate the loss of PETM autochthonous carbon, and to infer the relative proportion of allochthonous carbon in the preserved total (Table B-1).

X-ray fluorescence (XRF) elemental weight percents were obtained from Foreman (2012) with corresponding TOC data from Foreman et al., (2012) (Table B-5). Clay and silt outcrop

68 samples were from the same stratigraphic section as this study with 6 pre-PETM Paleocene samples, 10 PETM samples, and 1 post-PETM Eocene sample. Samples were from a variety of heights, but did not correspond to samples analyzed for PAHs.

3.5. Results

13 The δ Corg and %TOC records from the re-sampled section and published in Foreman et al. (2012) displayed similar trends. Overall, TOC was low, but was more variable at the carbon isotope excursion (CIE) onset and decreased further during the CIE recovery (Figure 3-2).

Average %TOC values in samples from the PETM interval (0.09%) were almost half that of the average %TOC values of pre-PETM Paleocene samples (0.17%). Average %TOC values were

0.05% in the post-PETM Eocene sample interval.

PAHs were detected in all samples and their total concentrations ranged from 0.4 to 133

μg/g TOC (Figure 3-2). The highest PAH concentrations were in the Pre-PETM Paleocene interval; the maximum PAH concentration was 133 μg/g TOC at -82.45 m stratigraphic level, which is ~9 m before the designated CIE onset. Total PAH concentrations were on average greater in the pre-PETM Paleocene interval (49 μg/g TOC) than the PETM (17 μg/g TOC) and the post-PETM Eocene (9 μg/g TOC) intervals.

Relative to the Paleocene samples, PETM samples originally contained 22% allochthonous carbon, and 78% autochthonous carbon, of which 68% was degraded and lost (Fig.

3). The proportion of allochthonous carbon tripled to constitute 69% of the preserved inventory of

TOC in the PETM interval (Figure 3-4).

A one-way analysis of variance (ANOVA) was calculated on %TOC values and on PAH concentrations of different lithologies (clay, silt, and sand) with p values calculated at the 5% (α =

0.05) level. At Piceance Basin, both the average %TOC values and the average PAH

69 concentrations between lithologies were not significantly different (p > 0.05). At Polecat Bench there was also no significant difference in TOC based on lithology (p>0.05). At Basin Substation, there was a significant difference between TOC based on lithology. Differences in lithology at

Basin Substation were further assessed using unpaired t-tests. TOC was significantly greater in clays than silts and sands (p<0.05) and silts contained significantly greater TOC than sands

(p<0.05).

From XRF analyses by Foreman (2012) and TOC analyses from Foreman et al., (2012),

2 2 TOC positively correlated with weight percent of TiO2 (R = 0.80) and Al2O3 (R = 0.56) (Figure

2 2 3-5). TOC negatively correlated with weight percent of Na2O (R = 0.68), K2O (R = 0.43), and

2 2 SiO2 (R = 0.42). TOC had no correlation (R <0.1) with weight percent of MgO and CaO.

3.6. Discussion

3.6.1. Organic carbon degradation and the refractory carbon component

Both %TOC values and PAH concentrations decreased in PETM samples relative to

Paleocene samples (Figure 3-2). Lower organic carbon contents were also observed in PETM sections at terrestrial sites in the Western Interior (e.g., Bighorn Basin, Wyoming (Magioncalda et al., 2004; Denis et al., in prep)), and further away in Spain (Claret Conglomerate; (Domingo et al., 2009)). In the Piceance Basin, PAH concentrations normalized to TOC were almost three- times less in PETM samples (17 μg/g TOC) than in Paleocene samples (49 μg/g TOC), similar to those quantified for two Bighorn Basin, Wyoming sections (Basin Substation and Polecat Bench;

Denis et al., in prep).

70 PAHs used as a proxy for intermediate refractory carbon in soils can reveal organic carbon degradation and enhanced allochthonous carbon preservation (Denis et al., in prep). High

13 variability in δ Corg values, even at a local scale, may be due to increased soil respiration (Cotton et al., 2015; Baczynski et al., 2016; Denis et al., in prep) that stripped away primary biomass inputs and favored preservation of recycled older refractory fossil carbon (Baczynski et al., 2013;

Bataille et al., 2013; Baczynski et al., 2016; Denis et al., in prep). Based on compound-specific isotope analyses of n-alkanes, Baczynski et al. (2016) quantified allochthonous carbon as 28 to

63% of the PETM organic carbon at Highway 16 site in the southeastern part of the Bighorn

13 Basin and attributed the variability in the δ Corg CIE record to fossil carbon variability. Denis et al., (in prep) quantified, based on TOC and PAH concentrations, that allochthonous carbon composed 74% of the PETM organic carbon at Basin Substation and 43% at Polecat Bench.

Following Denis et al., (in prep) we used PAH abundances to estimate the proportion of refractory allochthonous carbon in the PETM interval based on the amounts and the ratios of

TOC and PAHs relative to the late Paleocene (Table S1). Relative to the Paleocene samples,

PETM samples originally contained 22% allochthonous carbon and tripled to constitute 69% of the preserved inventory of TOC in the PETM interval (Figure 3-4).

The enhanced allochthonous carbon preservation during the PETM is consistent with observations that pollen in the Piceance PETM interval was mostly Cretaceous recalcitrant material (G. Harrington, pers. comm. 2015). Latest Paleocene pollen has been found just below the PETM interval and earliest Eocene pollen found right above the PETM interval, but samples from the Molina Member within the PETM interval lacked PETM pollen (Johnson and May,

1978; G. Harrington, pers. comm. 2015). The allochthonous carbon is likely sourced from the same units as siliciclastic detritus, the late Cretaceous fluviodeltaic strata from the Uncompahgre

Uplift and Sawatch Range located to the south and southeast of the Piceance Basin (Foreman and

Rasmussen, in review; Figure 3-1).

71 3.6.2. Degradation and the carbon isotope excursion

The extent of organic carbon degradation (and proportion of autochthonous carbon remaining) during the PETM in the Piceance Basin is similar to the proportions estimated for

Basin Substation. The enhanced proportion of allochthonous carbon at both of these sites was almost double of the proportion of allochthonous carbon estimated at Polecat Bench.

13 Interestingly, the δ Corg excursion at Polecat Bench is more consistent across the PETM interval and less variable compared to the other sites (Magioncalda et al., 2004; Foreman et al., 2012;

Maibauer, 2013). Because organics at Polecat Bench were least degraded (43% relative to

Paleocene), the remaining organic carbon had proportionally more autochthonous PETM carbon

13 (almost two-thirds of total PETM organic carbon) and δ Corg values were reflective of that

PETM carbon. In contrast, extensive degradation (60-68% TOC loss) occurred at both Basin

Substation and Piceance Basin, and autochthonous carbon accounts for less than one-third of the

13 remaining PETM TOC. These interpretations are supported by observations that δ Corg values were highly variable and lacked a clear excursion at Basin Substation (Maibauer, 2013) and that

13 δ Corg values were variable and occasionally returned to pre-PETM values across the CIE in the

Piceance Basin (Foreman et al., 2012).

The varying proportion of allochthonous carbon during the PETM at these different sites emphasizes that selective loss of organic carbon phases occurred regionally, although the extent of loss was variable on local scales. This makes it difficult to identify conditions under which

13 there is a greater likelihood the stratigraphic thickness, shape, and magnitude of the δ Corg record

13 is not fully representative of the CIE. Compound-specific isotope values (e.,g., δ Cn-alk), when concentrations allow, are not affected by allochthonous carbon inputs, and provide a direct record of the CIE (Baczynski, 2014; Baczynski et al., 2016). The allochthonous carbon identified in the

72 Bighorn Basin was sourced from Mesozoic marine deposits, which contained little terrestrial leaf wax n-alkane material (Baczynski et al., 2016).

3.6.3. Mechanisms that stabilized and destabilized organic carbon

The extent of organic carbon stabilized is linked to the grain size and mineral composition of the soil matrix (Oades, 1988; Sollins et al., 1996). Clay-size grains have greater surface area for organic-mineral interactions as opposed to sands (Mayer, 1994). The relationship between lithology and TOC, though, varied between sites. At Basin Substation, TOC was significantly greater in clays than silts and sands and in silts than sands. However, TOC and PAH concentrations were not significantly different between finer and coarser lithologies in samples from Piceance Basin and Polecat Bench. This suggests there were additional controls on organic carbon preservation other than grain size, such as clay mineralogy and soil moisture.

Based on TOC (Foreman et al., 2012) and XRF (Foreman, 2012) data from the Piceance

2 2 Basin, TOC positively correlated with weight percents of Al2O3 (R = 0.56) and of TiO2 (R =

0.80) (Figure 3-5). These elemental correlations suggest organic carbon stabilization was associated with clay (aluminosilicate) minerals and particularly with titanium-bearing clays, such as montmorillonite (Essington, 2004). The Al2O3-TOC correlation may not be as strong as the

TiO2-TOC correlation because aluminum is also a structural component of non-clay minerals, like feldspar, which have less surface area for organic-mineral interactions (Mikutta et al., 2009).

In modern soils, differences in clay mineralogy affect organic carbon preservation, but the mechanisms are not fully understood (Sollins et al., 1996; Von Lützow et al., 2006). For example, smectite particles tend to be much smaller than other clay minerals, such as kaolinite and illite, and thus, have more capacity to stabilize organic carbon (Reid-Soukup and Ulery, 2002; Wattel-

Koekkoek et al., 2003); on the other hand, smectite structures expand and contract as they absorb

73 water or dry, which can destabilize organic-mineral interactions (Reid-Soukup and Ulery, 2002).

We propose that aluminum was affiliated with minerals other than titanium-bearing clays. The regression data suggest organic carbon had a stronger interaction with the titanium-bearing clays

(e.g., montmorillonite (a type of smectite)) than the aluminum-bearing minerals (e.g., aluminosilicates and feldspars).

2 2 TOC negatively correlated with weight percent of Na2O (R = 0.68), K2O (R = 0.43),

2 and SiO2 (R = 0.42). These elements are more strongly associated with the non-clay minerals, feldspar and quartz, and Na and K are only minor constituents in some clay minerals as interlayer cations (e.g., kaolinite and illite), but not constituents of other clay minerals (e.g., smectites). All three elements, Na, K, and Si, are part of the chemical structure of feldspars. Feldpsars have less surface area than clays and are more common in coarser silt and sand grain size fractions, and thus, as suggested by the regression, their ability to stabilize organic carbon was minimal

(Essington, 2004; Mikutta et al., 2009).

The similar extent of autochthonous carbon degradation at Basin Substation and Piceance

Basin could reflect similar soil moisture conditions at the two sites. Basin Substation was inferred to be a wetter site than Polecat Bench based on lower topography and fewer oxidized paleosols

(Denis et al., in prep). The Piceance Basin was also a relatively wetter site during the late

Paleocene and early Eocene according to estimates of mean annual precipitation based on the paleosol chemical index of alteration that excludes potassium (CIA-K) (Foreman, 2012), which is based on the relative abundance of aluminum to calcium and sodium in paleosols (Adams et al.,

2011). How precipitation changed across the PETM was inconclusive based on the paleosols

CIA-K data (Foreman, 2012). Increased channel depths and extensive sand bodies suggested

PETM precipitation may have increased or been more seasonal (Foreman et al., 2012).

Sedimentologic, stratigraphic, geochemical, and botanical evidence from other basins in the

Laramide region (e.g., Bighorn Basin, Wyoming and Axehandle Basin, Utah), indicated drier or

74 more seasonally dry conditions during the PETM (Wing et al., 2005; Kraus and Riggins, 2007;

Smith et al., 2007; Smith et al., 2008, Bowen and Bowen, 2008). At Polecat Bench, and two sites in the southeast of the Bighorn Basin (Highway 16 and Sand Creek Divide), the paleosols were described as vertisols with a high illite and smectite clay content (Kraus and Riggins, 2007; Kraus et al., 2013). Vertisols are characterized by high concentrations of shrink-swell clays that cause the soil to crack when dry and expand when moist (Strawn et al., 2015).

The drying and rewetting of soils could have destabilized the organic matter association with clay minerals and disrupted soil particle aggregates that were protecting organic carbon from microbes (Oades, 1988). As clay mineral structures expand and contract with the addition and loss of water, organic matter can desorb from the clays, which makes it more accessible to microorganisms (Oades, 1988; Schimel et al., 2011). The interlayer of some clays (e.g., montmorillonite and smectites) is hydrated or can become hydrated (e.g., illite) and the clay structure swells or shrinks based on the amount of water present in the soil. Upon drying, the shrinking clays form deep cracks in soils, which allows water to penetrate to deeper levels of the soil. It is possible that drying disproportionately increased microbial degradation in originally wetter areas (e.g., Basin Substation and Piceance Basin), where clays contained more water and underwent a greater volume change, compared to initially drier areas (e.g., Polecat Bench), where clays with less water were already less expanded (Pacific et al., 2009).

In our conceptual model of organic carbon destabilization, a combination of higher temperatures, increased variability in soil moisture, and changes in clay mineralogy and clay content contributed to the decreased organic matter preservation during the PETM. During the onset of the carbon isotope excursion (centennial and millennial timescale), as atmospheric CO2 increased and temperatures rose, an initial boost in plant productivity and biomass litter flux (the priming effect) spurred increased degradation of both fresh and older (~100 – 10,000 years) organic carbon. Increased variability in rainfall disrupted aggregates within the soil and

75 subsequently destabilized organic-mineral interactions over hundreds to thousands of years.

Switches in hydrological regimes to drier or more seasonal conditions reduced clay content or altered clay mineralogy to minerals less favorable for organic-mineral interactions (e.g., less surface area or charged surfaces). The reduction in clay surfaces limited a critical organic carbon stabilization mechanism and reduced carbon storage over thousands to tens of thousands of years.

Across the length of the PETM, the persistent high temperatures aided carbon degradation

(increased microbial decomposition rates) for tens of thousands of years.

Overall, this resulted on average nearly 50% less organic carbon preserved during the

PETM than the late Paleocene. If the Paleocene global soil carbon reservoir was equivalent to the modern (1,500 Pg) and half this were degraded, then a total of ~750 Pg of carbon would have been released from soils over the course of the PETM (~170 ky), which is equivalent to ~25% of the estimated 3,000 Pg of carbon released at the onset of the PETM (Zachos et al., 2005; Cui et al., 2011; McInerney and Wing, 2011). The degradation and loss of older, pre-PETM carbon in soils was a potential source of carbon that leaked continually (over tens of thousands of years) during the PETM and contributed to the delayed recovery of the carbon isotope excursion to pre-

PETM δ13C values. A long-term carbon release from soils (~750 Pg) may have enhanced the greenhouse warming.

3.6.4. Implications

Other fluvial sites in the Laramide foreland of the western USA (e.g., Axe Handle, Hanna

Basin, Powder River Basin, and Williston Basin), and in Europe (e.g., Claret Conglomerate,

Spain), could have experienced similar severe degradation of multiple phases of organic carbon.

At the PETM sites for which there are published TOC records (e.g., Axe Handle (Bowen and

Bowen, 2008), Williston Basin (Clechenko et al., 2007), and Claret Conglomerate, Spain

76 (Domingo et al., 2009)), TOC is lower during the PETM relative to the Paleocene. We hypothesize that intermediately refractory carbon decreased as well and suggest widespread soil organic carbon destabilization and subsequent degradation during the PETM.

Soil carbon degradation, including of intermediately refractory carbon, during this past global warming event was likely accentuated by changes in soil moisture conditions, such as increased drying or seasonal precipitation. Although modern soil carbon degradation is spatially variable based on changes in temperature and soil moisture conditions (Heath et al., 2005), sites at two Laramide basins, and other sites in the western USA and Spain, suggest that soil organic carbon degradation was not a local phenomenon contained to the Bighorn Basin, but affected soils regionally and potentially globally. Extensive soil organic degradation during this past global warming event re-emphasizes the modern day concern that soil organic carbon stability and soil fertility will decrease as global temperatures rise (Lal, 2004; Heath et al., 2005; Knorr et al., 2005).

3.7. Conclusion

Decreased TOC and PAH concentrations indicate reduced organic carbon preservation during the PETM. That PAH concentrations decreased relative to TOC suggests that the organic carbon preserved during the PETM was more refractory than PAHs, such as refractory allochthonous fossil carbon. Using PAHs and TOC %Loss calculations as a metric for degradation, the calculated increased degradation and greater proportion of allochthonous carbon in PETM sediments in the Piceance Basin were consistent with other studies at localities in the

Bighorn Basin, Wyoming (Baczynski et al., 2013; Bataille et al., 2013; Baczynski et al., 2016;

13 Denis et al., in prep). These findings explain the high variability δ Corg values in terrestrial sediments and further support previous studies that suggested that caution is required when

13 making interpretations from δ Corg records (Baczynski et al., 2013; Baczynski et al., 2016). The positive correlation between elemental oxide weight percents (e.g., Al2O3 and TiO2) and %TOC suggests organic carbon stabilization was associated with clay minerals. Both Basin Substation and Piceance Basin, which are inferred wetter sites compared to Polecat Bench, had greater percent loss of organic carbon during the PETM. Reduced soil organic matter preservation during the PETM was due to a combination of increased temperatures (which increased microbial decomposition rates), decreased clay content and changes in mineralogy (which inhibited stability of fresh carbon), and fluctuations in soil moisture (which destabilized older, refractory carbon).

Soil carbon loss was widespread beyond the Bighorn Basin, Wyoming and increased soil carbon degradation during a hotter climate is a regional and potentially global concern.

3.8. Figures

Figure 3-1. Map of study site modified from Foreman et al., (2012) with PCBa section marked (red star).

Figure 3-2. Straitigraphic profile of geochemical parameters. Symbols represent binned grainsize; fill represents samples analyzed for PAHs in this work (white) and other values from Foreman et al., (2012) (grey). Left to right: Stratigraphic column 13 (adapted from Foreman et al., (2012)); bulk organic carbon isotope values (δ Corg) with 5-point moving average of both this work and Foreman et al., (2012) data (black line; any replicate analyses were averaged as one sample); Total organic carbon (TOC) with dashed lines indicating the median TOC values for the respective time interval; Total PAH is the sum of 12 PAHs with dashed lines indicating the median total PAH concentration for the respective time interval. Light grey box marks the PETM (Foreman et al., 2012).

Figure 3-3. Box and whisker plots of total organic carbon (TOC) and total PAH by lithology. Ends of box represent 25% and 75% quartile; middle line is median; whiskers extend to minimum and maximum values (not including outliers); white squares represent outliers.

Figure 3-4. Estimate of the relative contributions of soil organic matter degradation and allochthonous carbon to PETM organic carbon as a percent of Paleocene total organic carbon (TOC) for each site. Top: Bivariate plot of TOC and total PAH. If TOC was 100% comprised of intermediately refractory carbon (like PAHs), then points would plot on the 1:1 degradation line (dashed line); if TOC contains less refractory carbon than PAHs, then points will plot to the left of the 1:1 line; if TOC contains more refractory carbon than PAHs, then points will plot to the right of the 1:1 loss line. Polecat Bench values (light-grey square); Basin Substation values (medium-grey circle); Piceance Basin values (dark-grey triangle); estimated percent allochthonous carbon (horizontal arrow*). Bottom: Bar graph shows percent of organic carbon type during the PETM as a percent of Paleocene TOC for each site. Bars represent different forms of carbon: allochthonous (black*), autochthonous (white, solid line), degraded (white, dashed line). Right of bar graphs reports the Paleocene median TOC values for each site. *For a given site, allochthonous horizontal arrow and allochthonous black bar are the same length (same % Paleocene TOC).

Figure 3-5. Total organic carbon correlations (TOC) with TiO2 Al2O3, K2O, Na2O, and SiO2. TOC data is from Foreman et al., (2012) and element weight percent data is from Foreman (2012).

3.9. References

Baczynski, A.A., 2014, Evaluating Carbon Cycle Dynamics and Hydrologic Change during the Paleocene-Eocene Thermal Maximum, Bighorn Basin, Wyoming: Northwestern University, 219 p. Baczynski, A.A., McInerney, F.A., Wing, S.L., Kraus, M.J., Bloch, J.I., Boyer, D.M., Secord, R., Morse, P.E., and Fricke, H.C., 2013, Chemostratigraphic implications of spatial variation in the Paleocene-Eocene Thermal Maximum carbon isotope excursion, SE Bighorn Basin, Wyoming: Geochemistry, Geophysics, Geosystems, v. 14, no. 10, p. 4133–4152, doi: 10.1002/ggge.20265.

83 Baczynski, A.A., McInerney, F.A., Wing, S.L., Kraus, M.J., Morse, P.E., Bloch, J.I., Chung, A.H., and Freeman, K.H., 2016, Distortion of carbon isotope excursion in bulk soil organic matter during the Paleocene-Eocene thermal maximum: Geological Society of America Bulletin, , no. Xx, p. B31389.1, doi: 10.1130/B31389.1. Bataille, C.P., Mastalerz, M., Tipple, B.J., and Bowen, G.J., 2013, Influence of provenance and preservation on the carbon isotope variations of dispersed organic matter in ancient floodplain sediments: Geochemistry, Geophysics, Geosystems, v. 14, no. 11, p. 4874– 4891, doi: 10.1002/ggge.20294. Bowen, G.J., and Bowen, B.B., 2008, Mechanisms of PETM global change constrained by a new record from central Utah: Geology, v. 36, no. 5, p. 379–382, doi: 10.1130/G24597A.1. Bowes, G., 1993, Facing the inevitable: Plants and increasing atmospheric C02: Annual Review of Plant Physiology and Plant Molecular Biology, v. 44, p. 309–332. Burger, B.J., 2012, Northward range extension of a diminutive-sized mammal (Ectocion parvus) and the implication of body size change during the Paleocene-Eocene Thermal Maximum: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 363-364, p. 144–150, doi: 10.1016/j.palaeo.2012.09.008. Burger, B.J., and Honey, J.G., 2008, Plesiadapidae (Mammalia, Primates) from the late Paleocene Fort Union Formation of the Piceance Creek Basin, Colorado: Journal of Vertebrate Paleontology, v. 28, no. September, p. 816–825, doi: 10.1671/0272- 4634(2008)28[816:PMPFTL]2.0.CO;2. Chaopricha, N.T., and Marín-Spiotta, E., 2014, Soil burial contributes to deep soil organic carbon storage: Soil Biology and Biochemistry, v. 69, p. 251–264, doi: 10.1016/j.soilbio.2013.11.011. Christensen, B.T., 2001, Physical fractionation of soil and structural and functional complexity in organic matter turnover: European Journal of Soil Science, v. 52, no. 3, p. 345–353, doi: 10.1046/j.1365-2389.2001.00417.x. Clechenko, E.R., Kelly, D.C., Harrington, G.J., and Stiles, C.A., 2007, Terrestrial records of a regional weathering profile at the Paleocene-Eocene boundary in the Williston Basin of North Dakota: Bulletin of the Geological Society of America, v. 119, no. 3-4, p. 428– 442, doi: 10.1130/B26010.1. Clyde, W.C., Gingerich, P.D., Wing, S.L., Röhl, U., Westerhold, T., Bowen, G., Johnson, K., Baczynski, A.A., Diefendorf, A., McInerney, F., Schnurrenberger, D., Noren, A., Brady, K., Acks, R., et al., 2013, Bighorn Basin Coring Project (BBCP): A continental perspective on early Paleogene hyperthermals: Scientific Drilling, v. 16, p. 21–31, doi: 10.5194/sd-16-21-2013. Coplen, T.B., Brand, W.A., Gehre, M., Gröning, M., Meijer, H.A.J., Toman, B., and Verkouteren, R.M., 2006, New guidelines for δ13C measurements: Analytical Chemistry, v. 78, no. 7, p. 2439–2441, doi: 10.1021/ac052027c. Cotton, J.M., Sheldon, N.D., Hren, M.T., and Gallagher, T.M., 2015, Positive feedback drives carbon release from soils to atmosphere during Paleocene/Eocene warming: American Journal of Science, v. 315, no. 4, p. 337–361, doi: 10.2475/04.2015.03. Coulon, F., Pelletier, E., Gourhant, L., and Delille, D., 2005, Effects of nutrient and temperature on degradation of petroleum hydrocarbons in contaminated sub-Antarctic soil: Chemosphere, v. 58, no. 10, p. 1439–1448, doi: 10.1016/j.chemosphere.2004.10.007. Crouch, E.M., Dickens, G.R., Brinkhuis, H., Aubry, M.P., Hollis, C.J., Rogers, K.M., and Visscher, H., 2003, The Apectodinium acme and terrestrial discharge during the Paleocene-Eocene thermal maximum: New palynological, geochemical and calcareous nannoplankton observations at Tawanui, New Zealand: Palaeogeography,

84 Palaeoclimatology, Palaeoecology, v. 194, no. 4, p. 387–403, doi: 10.1016/S0031- 0182(03)00334-1. Cui, Y., Kump, L.R., Ridgwell, A.J., Charles, A.J., Junium, C.K., Diefendorf, A.F., Freeman, K.H., Urban, N.M., and Harding, I.C., 2011, Slow release of fossil carbon during the Palaeocene–Eocene Thermal Maximum: Nature Geoscience, v. 4, no. 7, p. 481–485, doi: 10.1038/ngeo1179. Denis, E.H., Foreman, B.Z., Maibauer, B.J., Bowen, G.J., Baczynski, A.A., McInerney, F.A., Collinson, M.E., Belcher, C.M., Wing, S.L., and Freeman, K.H. Decreased soil carbon in a warming world: Degraded pyrogenic carbon during the Paleocene-Eocene Thermal Maximum (PETM): Geology,. Denis, E.H., Foreman, B.Z., Maibauer, B.J., Bowen, G.J., Collinson, M.E., Belcher, C.M., and Freeman, K.H., 2014, Carbon burn-down in a greenhouse world: Wildfires and soil carbon loss across the Paleocene-Eocene Thermal Maximum (PETM): American Geophysical Union Annual Meeting, v. Sanfrancis, no. Poster, p. Abstract PP11A–1332. Denis, E.H., Foreman, B.Z., Maibauer, B.J., Bowen, G.J., Baczynski, A.A., McInerney, F.A., Collinson, M.E., Belcher, C.M., Wing, S.L., and Freeman, K.H., in prep. Decreased soil carbon in a warming world: Degraded pyrogenic carbon during the Paleocene-Eocene Thermal Maximum (PETM): Geology. Denis, E.H., Foreman, B.Z., Maibauer, B.J., Bowen, G.J., Collinson, M.E., Belcher, C.M., and Freeman, K.H., 2014, Carbon burn-down in a greenhouse world: Wildfires and soil carbon loss across the Paleocene-Eocene Thermal Maximum (PETM): American Geophysical Union Annual Meeting, Sanfrancisco, CA, Abstract PP11A–1332, Poster. Denis, E.H., Toney, J.L., Tarozo, R., Anderson, R.S., Roach, L.D., and Huang, Y., 2012, Polycyclic aromatic hydrocarbons (PAHs) in lake sediments record historic fire events: Validation using HPLC-fluorescence detection: Organic Geochemistry, v. 45, p. 7–17, doi: 10.1016/j.orggeochem.2012.01.005. Dickinson, W.R., Klute, M.A., Hayes, M.J., Janecke, S.U., Lundin, E.R., Mckittrick, M.A., and Olivares, M.D., 1988, Paleogeographic and paleotectonic setting of Laramide sedimentary basins in the central Rocky Mountain region: Bulletin of the Geological Society of America, v. 100, no. 7, p. 1023–1039, doi: 10.1130/0016- 7606(1988)100<1023:PAPSOL>2.3.CO;2. Domingo, L., López-Martínez, N., Leng, M.J., and Grimes, S.T., 2009, The Paleocene-Eocene Thermal Maximum record in the organic matter of the Claret and Tendruy continental sections (South-central Pyrenees, Lleida, Spain): Earth and Planetary Science Letters, v. 281, no. 3-4, p. 226–237, doi: 10.1016/j.epsl.2009.02.025. Don, A., Rödenbeck, C., and Gleixner, G., 2013, Unexpected control of soil carbon turnover by soil carbon concentration: Environmental Chemistry Letters, v. 11, no. 4, p. 407–413, doi: 10.1007/s10311-013-0433-3. Essington, M.E., 2004, Soil and Water Chemistry: An Integrative Approach: CRC Press LLC, Boca Raton. Fontaine, S., Mariotti, A., and Abbadie, L., 2003, The priming effect of organic matter: A question of microbial competition? Soil Biology and Biochemistry, v. 35, no. 6, p. 837– 843, doi: 10.1016/S0038-0717(03)00123-8. Forbes, M.S., Raison, R.J., and Skjemstad, J.O., 2006, Formation, transformation and transport of black carbon (charcoal) in terrestrial and aquatic ecosystems: Science of the Total Environment, v. 370, no. 1, p. 190–206, doi: 10.1016/j.scitotenv.2006.06.007. Foreman, B.Z., 2014, Climate-driven generation of a fluvial sheet sand body at the Paleocene- Eocene boundary in north-west Wyoming (USA): Basin Research, v. 26, no. 2, p. 225– 241, doi: 10.1111/bre.12027.

85 Foreman, B.Z., 2012, Fluvial Response to the Paleocene-Eocene Thermal Maximum in Western North America: University of Wyoming, 147 p. Foreman, B.Z., Heller, P.L., and Clementz, M.T., 2012, Fluvial response to abrupt global warming at the Palaeocene/Eocene boundary.: Nature, v. 491, no. 7422, p. 92–5, doi: 10.1038/nature11513. Foreman, B.Z., and Rasmussen, D.M. Provenance signals in the Piceance Creek Basin: Uproofing of the Sawatch Range and extent of the Early Paleogene California river system (Colorado, U.S.A.): Journal of Sedimentary Research,. Gingerich, P.D., 1989, New Earliest Wasatchian mammalian fauna from the Eocene of Northwestern Wyoming: Composition and diversity in a rarely sampled high-floodplain assemblage: University of Michigan Papers on Paleontology, v. 28, p. 1–97. Guenet, B., Juarez, S., Bardoux, G., Abbadie, L., and Chenu, C., 2012, Evidence that stable C is as vulnerable to priming effect as is more labile C in soil: Soil Biology and Biochemistry, v. 52, p. 43–48, doi: 10.1016/j.soilbio.2012.04.001. Handley, L., O’Halloran, A., Pearson, P.N., Hawkins, E., Nicholas, C.J., Schouten, S., McMillan, I.K., and Pancost, R.D., 2012, Changes in the hydrological cycle in tropical East Africa during the Paleocene-Eocene Thermal Maximum: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 329-330, p. 10–21, doi: 10.1016/j.palaeo.2012.02.002. Heath, J., Ayres, E., Possell, M., Bardgett, R.D., Black, H.I.J., Grant, H., Ineson, P., and Kerstiens, G., 2005, Rising Atmospheric CO2 Reduces Sequestration of Root-Derived Soil Carbon: Science, v. 309, no. 5741, p. 1711–1713, doi: 10.1126/science.1110700. Jaramillo, C., Ochoa, D., Contreras, L., Pagani, M., Carvajal-Ortiz, H., Pratt, L.M., Krishnan, S., Cardona, A., Romero, M., Quiroz, L., Rodriguez, G., Rueda, M.J., de la Parra, F., Morón, S., et al., 2010, Effects of rapid global warming at the Paleocene-Eocene boundary on neotropical vegetation: Science, v. 330, no. 6006, p. 957–961, doi: 10.1126/science.1193833. Johnsen, A.R., Wick, L.Y., and Harms, H., 2005, Principles of microbial PAH-degradation in soil: Environmental Pollution, v. 133, no. 1, p. 71–84, doi: 10.1016/j.envpol.2004.04.015. Johnson, R.C., and May, F., 1978, Preliminary stratigraphic studies of the upper part of the Mesaverde Group, the Wasatch Formation, and the lower part of the Green River Formation, DeBeque area, Colorado, including environments of deposition and investigation of palynomorph assemblages: USGS Miscellaneous Field Investigations, p. Map MF–1050. Killops, S.D., and Massoud, M.S., 1992, Polycyclic aromatic hydrocarbons of pyrolytic origin in ancient sediments: evidence for Jurassic vegetation fires: Organic Geochemistry, v. 18, no. 1, p. 1–7, doi: 10.1016/0146-6380(92)90137-M. Knorr, W., Prentice, I., House, J., and Holland, E., 2005, Long-term sensitivity of soil carbon turnover to warming: Nature, v. 433, no. January, p. 298–301, doi: 10.129/2002PA000837. Koch, P.L., Zachos, J.C., and Gingerich, P.D., 1992, Correlation between isotope records in marine and continental carbon reservoirs near the Palaeocene/Eocene boundary: Nature, v. 358, no. 6384, p. 319–322, doi: 10.1038/358319a0. Kraus, M.J., and Riggins, S., 2007, Transient drying during the Paleocene-Eocene Thermal Maximum (PETM): Analysis of paleosols in the bighorn basin, Wyoming: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 245, no. 3-4, p. 444–461, doi: 10.1016/j.palaeo.2006.09.011. Lal, R., 2004, Soil carbon sequestration impacts on global climate change and food security: Science, v. 304, no. 5677, p. 1623–1627, doi: 10.1126/science.1097396.

86 Lawrence, C.R., Harden, J.W., Xu, X., Schulz, M.S., and Trumbore, S.E., 2015, Long-term controls on soil organic carbon with depth and time: A case study from the Cowlitz River Chronosequence, WA USA: Geoderma, v. 247-248, p. 73–87, doi: 10.1016/j.geoderma.2015.02.005. Lehmann, J., Gaunt, J., and Rondon, M., 2006, Bio-char sequestration in terrestrial ecosystems - A review: Mitigation and Adaptation Strategies for Global Change, v. 11, no. 2, p. 403– 427, doi: 10.1007/s11027-005-9006-5. Lorenz, K., Lal, R., Preston, C.M., and Nierop, K.G.J., 2007, Strengthening the soil organic carbon pool by increasing contributions from recalcitrant aliphatic bio(macro)molecules: Geoderma, v. 142, no. 1-2, p. 1–10, doi: 10.1016/j.geoderma.2007.07.013. Von Lützow, M., Kögel-Knabner, I., Ekschmitt, K., Matzner, E., Guggenberger, G., Marschner, B., and Flessa, H., 2006, Stabilization of organic matter in temperate soils: Mechanisms and their relevance under different soil conditions - A review: European Journal of Soil Science, v. 57, no. 4, p. 426–445, doi: 10.1111/j.1365-2389.2006.00809.x. Von Lützow, M., Kögel-Knabner, I., Ludwig, B., Matzner, E., Flessa, H., Ekschmitt, K., Guggenberger, G., Marschner, B., and Kalbitz, K., 2008, Stabilization mechanisms of organic matter in four temperate soils: Development and application of a conceptual model: Journal of Plant Nutrition and Soil Science, v. 171, no. 1, p. 111–124, doi: 10.1002/jpln.200700047. Magill, C.R., Denis, E.H., and Freeman, K.H., 2015, Rapid sequential separation of sedimentary lipid biomarkers via selective accelerated solvent extraction: Organic Geochemistry, v. 88, p. 29–34, doi: 10.1016/j.orggeochem.2015.07.009. Magioncalda, R., Dupuis, C., Smith, T., Steurbaut, E., and Gingerich, P.D., 2004, Paleocene- Eocene carbon isotope excursion in organic carbon and pedogenic carbonate: Direct comparison in a continental stratigraphic section: Geology, v. 32, no. 7, p. 553–556, doi: 10.1130/G20476.1. Maibauer, B.J., 2013, Carbon Isotope Stratigraphy of Early Eocene Hyperthermals in the Bighorn Basin, Wyoming, USA: Analogues for Modern Anthropogenic Carbon Emissions: The University of Utah, 122 p. Marschner, B., Brodowski, S., Dreves, A., Gleixner, G., Gude, A., Grootes, P.M., Hamer, U., Heim, A., Jandl, G., Ji, R., Kaiser, K., Kalbitz, K., Kramer, C., Leinweber, P., et al., 2008, How relevant is recalcitrance for the stabilization of organic matter in soils? Journal of Plant Nutrition and Soil Science, v. 171, no. 1, p. 91–110, doi: 10.1002/jpln.200700049. Masclet, P., Cachier, H., Liousse, C., and Wortham, H., 1995, Emissions of polycyclic aromatic hydrocarbons by savanna fires: Journal of Atmospheric Chemistry, v. 22, no. 1-2, p. 41– 54, doi: 10.1007/BF00708180. Mayer, L.M., 1994, Relationships between mineral surfaces and organic carbon concentrations in soils and sediments: Chemical Geology, v. 114, no. 3-4, p. 347–363, doi: 10.1016/0009- 2541(94)90063-9. McInerney, F.A., and Wing, S.L., 2011, The Paleocene-Eocene Thermal Maximum: A Perturbation of Carbon Cycle, Climate, and Biosphere with Implications for the Future: Annual Review of Earth and Planetary Sciences, v. 39, no. 1, p. 489–516, doi: 10.1146/annurev-earth-040610-133431. Midwood, A.J., and Boutton, T.W., 1998, Soil carbonate decomposition by acid has little effect on δ13C of organic matter: Soil Biology and Biochemistry, v. 30, no. 10-11, p. 1301– 1307, doi: 10.1016/S0038-0717(98)00030-3. Mikutta, R., Schaumann, G.E., Gildemeister, D., Bonneville, S., Kramer, M.G., Chorover, J., Chadwick, O.A., and Guggenberger, G., 2009, Biogeochemistry of mineral-organic

87 associations across a long-term mineralogical soil gradient (0.3-4100 kyr), Hawaiian Islands: Geochimica et Cosmochimica Acta, v. 73, no. 7, p. 2034–2060, doi: 10.1016/j.gca.2008.12.028. Nobel, P.S., and Cortazar, V.G. De, 1991, Growth and Predicted Productivity of Opuntia ficus- indica for Current and Elevated Carbon Dioxide: Agronomy Journal, v. 83, p. 224–230. Nobel, P.S., and Hartsock, T.L., 1986, Short-Term and Long-Term Responses of Crassulacean Acid Metabolism Plants to Elevated CO2: Plant Physiology, v. 82, p. 604–606. Oades, J.M., 1988, The Retention of Organic Matter in Soils: Biogeochemistry, v. 5, no. 1, p. 35– 70. Pacific, V.J., McGlynn, B.L., Riveros-Iregui, D.A., Epstein, H.E., and Welsch, D.L., 2009, Differential soil respiration responses to changing hydrologic regimes: Water Resources Research, v. 45, no. 7, p. 6–11, doi: 10.1029/2009WR007721. Pagani, M., Pedentchouk, N., Huber, M., Sluijs, A., Schouten, S., Brinkhuis, H., Sinninghe Damsté, J.S., and Dickens, G.R., 2006, Arctic hydrology during global warming at the Palaeocene/Eocene thermal maximum: Nature, v. 442, no. 7103, p. 671–675, doi: 10.1038/nature05043. Page, D.S., Boehm, P.D., Douglas, G.S., Bence, A.E., Burns, W.A., and Mankiewicz, P.J., 1999, Pyrogenic Polycyclic Aromatic Hydrocarbons in Sediments Record Past Human Activity: A Case Study in Prince William Sound, Alaska: Marine Pollution Bulletin, v. 38, no. 4, p. 247–260, doi: 10.1016/S0025-326X(98)00142-8. Pisani, O., Lin, L.H., Lun, O.O.Y., Lajtha, K., Nadelhoffer, K.J., Simpson, A.J., and Simpson, M.J., 2016, Long-term doubling of litter inputs accelerates soil organic matter degradation and reduces soil carbon stocks: Biogeochemistry, v. 127, no. 1, p. 1–14, doi: 10.1007/s10533-015-0171-7. Ramdahl, T., 1983, Retene - a molecular marker of wood combustion in ambient air: Nature, v. 306, p. 580–582, doi: 10.1038/306580a0. Reid-Soukup, D.A., and Ulery, A.L., 2002, Smectites, in Dixon, J.B. and Schulze, D.G. eds., Soil mineralogy with environmental applications, Soil Science Society of America Book Series, Madison, p. 467–499. Robert, C., and Kennett, J.P., 1994, Antarctic subtropical humid episode at the Paleocene-Eocene boundary: Clay-mineral evidence: Geology, v. 22, no. 3, p. 211, doi: 10.1130/0091- 7613(1994)022<0211:ASHEAT>2.3.CO;2. Röhl, U., Westerhold, T., Bralower, T.J., and Zachos, J.C., 2007, On the duration of the Paleocene-Eocene thermal maximum (PETM): Geochemistry, Geophysics, Geosystems, v. 8, no. 12, doi: 10.1029/2007GC001784. Rustad, L.E., Campbell, J.L., Marion, G.M., Norby, R.J., Mitchell, M.J., Hartley, A.E., Cornelissen, J.H.C., Gurevitch, J., Alward, R., Beier, C., Burke, I., Canadell, J., Callaghan, T., Christensen, T.R., et al., 2001, A meta-analysis of the response of soil respiration, net nitrogen mineralization, and aboveground plant growth to experimental ecosystem warming: Oecologia, v. 126, no. 4, p. 543–562, doi: 10.1007/s004420000544. Schimel, J.P., Wetterstedt, J.Å.M., Holden, P.A., and Trumbore, S.E., 2011, Drying/rewetting cycles mobilize old C from deep soils from a California annual grassland: Soil Biology and Biochemistry, v. 43, no. 5, p. 1101–1103, doi: 10.1016/j.soilbio.2011.01.008. Schmidt, M.W.I., and Noack, A.G., 2000, Analysis, distribution, implications, and current challenges: Global Biogeochemical Cycles, v. 14, no. 3, p. 777–793. Schmidt, M.W.I., Torn, M.S., Abiven, S., Dittmar, T., Guggenberger, G., Janssens, I. a., Kleber, M., Kögel-Knabner, I., Lehmann, J., Manning, D. a. C., Nannipieri, P., Rasse, D.P., Weiner, S., and Trumbore, S.E., 2011, Persistence of soil organic matter as an ecosystem property: Nature, v. 478, no. 7367, p. 49–56, doi: 10.1038/nature10386.

88 Secord, R., Gingerich, P.D., Lohmann, K.C., and Macleod, K.G., 2010, Continental warming preceding the Palaeocene-Eocene thermal maximum: Nature, v. 467, no. 7318, p. 955– 958, doi: 10.1038/nature09441. Smith, J.J., Hasiotis, S.T., Kraus, M.J., and Woody, D.T., 2008, Relationship of floodplain ichnocoenoses to paleopedology, paleohydrology, and paleoclimate in the Willwood Formation, Wyoming, during the Paleocene-Eocene Thermal Maximum: Palaios, v. 23, no. 9-10, p. 683–699, doi: 10.2110/palo.2007.p07-080r. Smith, F.A., Wing, S.L., and Freeman, K.H., 2007, Magnitude of the carbon isotope excursion at the Paleocene-Eocene thermal maximum: The role of plant community change: Earth and Planetary Science Letters, v. 262, no. 1-2, p. 50–65, doi: 10.1016/j.epsl.2007.07.021. Sollins, P., Homann, P., and Caldwell, B. a., 1996, Stabilization and destabilization of soil organic matter: mechanisms and controls: Geoderma, v. 74, no. 1-2, p. 65–105, doi: 10.1016/S0016-7061(96)00036-5. Strawn, D.G., Bohn, H.L., and O’Connor, G.A., 2015, Soil Chemistry: John Wiley & Sons, Ltd, West Sussex. Torn, M.S., Trumbore, S.E., Chadwick, O.A., Vitousek, P.M., and Hendricks, D.M., 1997, Mineral control of soil organic carbon storage and turnover: Nature, v. 389, no. 6647, p. 170–173, doi: Doi 10.1038/38260. Westerhold, T., Röhl, U., and Laskar, J., 2012, Time scale controversy: Accurate orbital calibration of the early Paleogene: Geochemistry, Geophysics, Geosystems, v. 13, no. 6, p. 1–19, doi: 10.1029/2012GC004096. Wing, S.L., and Currano, E.D., 2013, Plant response to a global greenhouse event 56 million years ago: American Journal of Botany, v. 100, no. 7, p. 1234–1254, doi: 10.3732/ajb.1200554. Wing, S.L., Harrington, G.J., Smith, F.A., Bloch, J.I., Boyer, D.M., and Freeman, K.H., 2005, Transient Floral Change and Rapid Global Warming at the Paleocene-Eocene Boundary: Science, v. 310, no. 5750, p. 993–996, doi: 10.1126/science.1116913. Yunker, M.B., Macdonald, R.W., Vingarzan, R., Mitchell, R.H., Goyette, D., and Sylvestre, S., 2002, PAHs in the Fraser River basin: A critical appraisal of PAH ratios as indicators of PAH source and composition: Organic Geochemistry, v. 33, no. 4, p. 489–515, doi: 10.1016/S0146-6380(02)00002-5. Zachos, J.C., Nicolo, M., Raffi, I., Lourens, L.J., McCarren, H., and Kroon, D., 2005, Rapid Acidification of the Ocean During the Paleocene-Eocene Thermal Maximum: Science, v. 308, p. 1611–1615, doi: 10.1126/science.1109004. Zeebe, R.E., Ridgwell, A., and Zachos, J.C., 2016, Anthropogenic carbon release rate unprecedented during the past 66 million years: Nature Geoscience, v. 9, no. April, p. 325–329, doi: 10.1038/ngeo2681. Zeebe, R.E., and Zachos, J.C., 2013, Long-term legacy of massive carbon input to the Earth system: Anthropocene versus Eocene: Philosophical transactions. Series A, Mathematical, physical, and engineering sciences, v. 371, no. 2001, p. 20120006, doi: 10.1098/rsta.2012.0006.

Chapter 4

Fire and ecosystem change in the Arctic across the Paleocene-Eocene Thermal Maximum

4.1. Abstract

Fire, an important component of ecosystems at a range of spatial and temporal scales, affects vegetation distribution, the carbon cycle, and climate. In turn, climate influences fuel composition (e.g., amount and type of vegetation), fuel availability (e.g., vegetation that can burn based on precipitation and temperature), and ignitions (i.e., lightning). Mechanisms that control the relationship between climate and fire are complex, in large part because of the role of vegetation is also an important influence, and it is difficult to predict future changes in wildfire activity and infer past fire activity based only on climate and environmental conditions. Studies project increased wildfire activity in future decades, but the propensity for fire varies on local and regional scales due to variations in vegetation and moisture regimes. Applying short- and long- duration interpretations of fire-climate relationships in the modern world to observations from a past global warming event (e.g., the Paleocene-Eocene Thermal Maximum (PETM)) can inform the characteristics of the past ecosystem and help assess how the feedbacks between climate and tendency for fire operated in the past and on longer timescales (e.g., thousands to tens of thousands of years). The abrupt global warming during the PETM dramatically altered vegetation and hydrologic patterns, and, possibly, fire occurrence. To document coincident changes in climate, vegetation, and fire occurrence, we evaluated biomarkers, including polycyclic aromatic hydrocarbons (PAHs), terpenoids, and alkanes, at a marine depositional site (IODP site 302, the

Lomonosov Ridge) in the Arctic Basin.

90 At this site, precipitation and vegetation changed during the PETM, both of which could be tied to changes in fire propensity. The abundance of the C29 n-alkane tracked trends in the abundance of angiosperms, as quantified by pollen and by the ratio of plant leaf wax n-alkanes relative to diterpenoids. Both pollen and biomarker records indicate angiosperms abundance increased during the PETM relative gymnosperms. Similarly, PAH abundances increased relative to total plant biomarkers throughout the PETM, which suggests PAH production increased relative to plant productivity during the PETM. A time lag between increased moisture transport to the Arctic and increased PAHs and angiosperms suggests wetter conditions, followed by increased temperatures, favored angiosperms and combined to enhance fire occurrence. Hence, in the paleoenvironments, drier conditions do not directly result in increased fire occurrence. Rather, like modern fire dynamics, past fire patterns reflect a balance of variability in precipitation and sufficient fuels produced by vegetation. Increased fire in a wetter Arctic suggests PETM precipitation was seasonal, or more variable on a longer timescale, and that hotter temperatures further facilitated burning.

4.2. Introduction

Many modeling studies project increases in wildfire activity in future decades, but the mechanisms controlling these changes are complex and the primary controls on fire occurrence are not always clear (Hessl, 2011). Today, increased atmospheric CO2 concentrations, higher temperatures, and longer dry seasons are associated with increases in fire activity in the western

USA (Westerling, 2006). Shifts in vegetation, though, can sometimes override the influence of warmer and drier conditions (Higuera et al., 2014). In addition, most empirical evidence, which is also the basis of models, is on an annual to centennial scale, and the changes observed may not translate to climate-vegetation-atmospheric CO2 relationships recorded in the paleorecord on

91 1,000 to 10,000 year scales (Hessl, 2011). Determining the influence of major climate changes on fire occurrence and the feedbacks between precipitation and fire in the past, such as during the

Paleocene-Eocene Thermal Maximum (PETM), can illuminate how current and future climate changes will affect the tendency for fire on longer timescales (e.g., 1,000 to 10,000 years).

The PETM was a geologically abrupt period of global warming that occurred about 55.5 million years ago (Westerhold et al., 2012). This period is widely invoked as a geologic analog for modern climate change, even though modern climate change may be unprecedented in terms of carbon release rate (modern rate (~10 Pg C/yr) is 10 times faster than the PETM) (Cui et al.,

2011; Zeebe and Zachos, 2013; Zeebe et al., 2016). The hyperthermal event is marked by a negative carbon isotope excursion (CIE), signifying a major perturbation to the carbon cycle

(McInerney and Wing, 2011 and references therein). At least 3,000 Pg of 13C-depleted carbon was released into the atmosphere over ~10 ky and global temperatures rose ~5-8°C over ~170 ky

(Wing et al., 2005; Zachos et al., 2005; Röhl et al., 2007; Secord et al., 2010; Cui et al., 2011;

McInerney and Wing, 2011). There were dramatic shifts in vegetation and precipitation patterns

(Wing et al., 2005; Pagani et al., 2006b; Kraus and Riggins, 2007; Smith et al., 2007; Smith et al.,

2008; Jaramillo et al., 2010; Foreman et al., 2012; Handley et al., 2012; Wing and Currano,

2013). For example, increased terrestrial inputs and greater abundance of kaolinite in marginal marine settings during the PETM suggest that there was increased run off and increased variability in rain events (e.g., Robert and Kennett, 1994; Crouch et al., 2003). In addition, there was an inferred increase in moisture transport to the Arctic based on δD of n-alkanes (Pagani et al., 2006b). However, it is difficult to infer how fire occurrence changed in relation to temperature, precipitation, and vegetation before, during, and after the PETM because the mechanisms that control fire are complex and few studies have reported evidence for fire during this time period.

92 Moore and Kurtz (2008) examined graphitic black carbon (GBC), a combustion byproduct, from two International Ocean Drilling Program (IODP) sites, the Bass River section

(New Jersey Margin) and site 1210 (Shatsky Rise). At Shatsky Rise, GBC concentrations were below their method’s detection limit (<0.5 ppm). In the New Jersey Margin, which focused on the onset of the CIE, GBC concentrations varied between ~120-360 ppm before and during the CIE, and there was no clear pattern in GBC flux at the onset or during the CIE. Carbon isotope values of the black carbon recorded a ~3.5‰ negative CIE, which suggested the burning of biomass that had PETM carbon isotope values rather than the burning of older Paleocene peat or coal (Moore and Kurtz, 2008).

Collinson et al. (2007) and Collinson et al. (2009) linked a switch in fire regime to changes in vegetation composition across the PETM in England. Late Paleocene samples, from the Cobham Lignite Bed in Southern England, were dominated by charcoal associated with episodic fires and by fern spores, which suggested a low diversity, fire-prone fern and woody angiosperm community. The PETM vegetation, though, was characterized by the loss of ferns, an increase in wetland plants, and a cessation in fires. This study highlights the importance of vegetation in determining fire propensity.

Denis et al. (2014), Denis et al. (in prep-a), and Denis et al. (in prep-b) investigated pyrogenic carbon (polycyclic aromatic hydrocarbons (PAHs) and charcoal) before, during, and after the PETM event at two coring sites in the Bighorn Basin, Wyoming and an outcrop section in the Piceance Basin, Colorado. The studies found that there was decreased soil carbon stabilization during the PETM, which clouded evidence for changes in fire occurrence at these paleo-floodplain sites. Denis et al. (2014) and Denis et al. (in prep-a) proposed that in the Bighorn

Basin, a more patchy distribution of vegetation, likely driven by a drying climate, lowered fuel connectivity and decreased fire occurrence during the PETM relative to the late Paleocene.

93 Fire history reconstructions of the past decades to 21 thousand years, mainly using sedimentary charcoal and tree ring fire scar analyses, provide information regarding fire frequency, fire extent, and the timing of past fires in relation to climate (Clark, 1990; Whitlock and Larsen, 2001; Whitlock and Anderson, 2003; Margolis and Balmat, 2009). Overall, wet periods allow for the buildup of biomass (fuel) and dry periods facilitate the burning of vegetation

(fuel availability). The length of wet and dry periods can have different effects on fire occurrence depending on fuel type. Holocene fire frequency records in the western United States indicate that enhanced seasonality and anomalously wet years followed by anomalously dry years play a major role in promoting fire conditions for vegetation with annual fuel production, such as grass

(Margolis and Balmat, 2009). Other studies have suggested that extended dry periods lead to widespread fires, such as the 1982-83 and 1997 Indonesian fires that spread wildly during the long El Niño dry season, likely because heavier fuels (e.g., branches and logs) respond to humidity changes more slowly than finer fuels (e.g., grass and small twigs) (Malingreau et al.,

1985; Page et al., 2002). Alternatively, during long-term droughts, fire occurrence could actually decrease if there is insufficient biomass to burn (Flannigan et al., 2009). From analysis of a global compilation of charcoal records covering the last 21,000 years, Daniau et al. (2012) found an overall increase in fire with increasing temperature. In fact, when inferring the effect of climate on paleofires, such as during the PETM, it is often assumed that hotter and drier periods lead to increased fires (Cook et al., 2004; Wing et al., 2005; Secord et al., 2010).

To reconstruct fire occurrence we use pyrogenic carbon, which is a continuum of combustion products generated as solid residue or volatiles, ranging from slightly charred material to soot (Schmidt and Noack, 2000; Masiello, 2004; Bird et al., 2014). Polycyclic aromatic hydrocarbons (PAHs) are parts of this continuum and are byproducts of combustion released as both volatiles and in association with particles (Masclet et al., 1995). PAHs can be a marker of two signals, fire occurrence, in accumulated sediments, and, in paleosols, relative

94 preservation of soil organic carbon (Denis et al., in prep-a; Denis et al., in prep-b). In the sedimentary record, changes in PAH concentrations are usually interpreted to indicate changes in fire occurrence, with more PAHs linked to increased fire occurrence (e.g., Ramdahl, 1983;

Killops and Massoud, 1992; Yunker et al., 2002; Marynowski and Simoneit, 2009; Denis et al.,

2012). Aromatic structures tend to make pyrogenic carbon, including larger PAHs (≥5 rings), relatively resistant to degradation in soil environments and marine sediments (Coulon et al., 2005;

Johnsen et al., 2005; Von Lützow et al., 2006; Knicker, 2011). For example, charcoal, another byproduct of fire, has relatively long residence time in modern soils, estimated on the order of

500 – 10,000 years, and in marine sediments with oxygen exposure, 10,000 – 20,000 years

(Schmidt and Noack, 2000; Forbes et al., 2006; Von Lützow et al., 2006; Kuzyakov et al., 2009;

Knicker, 2011). Thus, in soils, PAHs as an intermediate-phase of refractory carbon, can indicate relative preservation of organic carbon (Denis et al., in prep-a; Denis et al., in prep-b). PAHs, therefore, reflect both production (via fire) and preservation (via degradation) of pyrogenic carbon phases.

We analyze PAHs and plant biomarkers in a sediment core from the central Arctic Ocean

(IODP Hole 302-4A) to determine how fire occurrence changed in relation to vegetation and precipitation in the Arctic before, during, and after the PETM. The combination of hotter and wetter conditions (Pagani et al., 2006b), major changes in vegetation composition (Schouten et al., 2007), and the complexity of mechanisms that control fire make it difficult to infer how fire occurrence changed. If precipitation mainly controlled the tendency for fire in the Arctic, then inferred wetter conditions (Pagani et al., 2006b) might have dampened fire occurrence during the

PETM in the Arctic.

95 4.3. Study Section

Core samples were collected from IODP Hole 302-4A on the Lomonosov ridge in the central Arctic Ocean (Figure 4-1). Several previous studies have analyzed the same samples, or samples from similar stratigraphic intervals, for a variety of geochemical, biomarker, and palynomorph data (Backman et al., 2006; Sluijs et al., 2006; Stein et al., 2006; Pagani et al.,

2006b; Schouten et al., 2007; Sluijs et al., 2008b). Organic-rich siliclastic claystone sediments offer samples with well-preserved biomarkers and palynomorphs (Sluijs et al., 2006; Stein et al.,

2006; Pagani et al., 2006b; Schouten et al., 2007) before, during, and after the PETM interval.

Anoxic bottom-water conditions (supported by laminated sediments and the absence of benthic foraminiferal linings (Sluijs et al., 2006)) and euxinic conditions in the photic zone (supported by a presence of isorenieratene and other isorenieratene derivatives) facilitated organic carbon preservation during the PETM interval (Sluijs et al., 2006; Schouten et al., 2007). Average sedimentation rates from the late Paleocene to the early Eocene were 1 to 3 cm/kyr (Stein et al.,

2006; Sluijs et al., 2008b) and were estimated to have increased during the PETM to 5.0 ± 1.2 cm/kyr (Sluijs et al., 2008b). Sea level rose during the event by about 20-30 m (Sluijs et al.,

2008a; Nagy et al., 2013).

The Arctic region was hotter and wetter during the PETM relative to before and after the event (Sluijs et al., 2006; Pagani et al., 2006b; Sluijs et al., 2008b). Sea surface temperatures increased at least 5 °C during the PETM from ~18 °C (Sluijs et al., 2006). Pagani et al. (2006) suggested a greater export of moisture from the tropics towards higher latitudes. The isotopic composition of Arctic PETM precipitation was considerably deuterium-enriched compared to today, indicating reduced rainout along the source airmass’ trajectory from lower latitudes to the poles. In addition, low-salinity-tolerant dinocyst assemblages (Sluijs et al., 2006) suggest increased precipitation and runoff during the PETM (Pagani et al., 2006b). Sluijs et al. (2006)

96 suggested that higher temperatures and enhanced fluvial runoff increased nutrient inputs, which increased marine productivity, and caused water column stratification. Further, because the Arctic

Basin may have been a restricted basin, high terrestrial runoff during the PETM could have helped create a freshwater lid to facilitate water column stratification (Sluijs et al., 2006).

4.4. Methods

4.4.1. Samples

Sediments were obtained from IODP Expedition 302 Hole 4A on the Lomonosov Ridge

(Backman et al., 2006; Pagani et al., 2006b) (Figure 4-1). An age model for the core was previously determined from palynological data and index events, which put the base of the

Eocene at the top of Core 32x (Backman et al., 2006) (Figure 4-2 and Figure 4-3). The PETM interval was marked based on the negative carbon isotope excursion from δ13C of total organic carbon and of leaf wax n-alkanes (Pagani et al., 2006b; Schouten et al., 2007).

4.4.2. Extraction and analysis

Lipid extracts analyzed in this paper were a subset of those processed and analyzed in

Pagani et al. (2006). Sediments were prepared for analysis as described in Pagani et al. (2006). In brief, sediments were freeze-dried and extracted with dichloromethane (DCM) using accelerated solvent extraction. Total lipid extracts (TLEs) were separated by column chromatography into three fractions using hexane (S1), hexane/DCM (9:1 v:v) (S2), and DCM/methanol (2:1 v:v) (S3).

The first fraction (S1) was further separated into two fractions (adducts and non-adducts) via urea adduction.

97 PAHs were analyzed from a few TLEs and from the S2 fractions using an Agilent 6890

GC with an Agilent 5973 quadrupole mass spectrometer (MS) and a fused silica capillary column

(Agilent J&W DB-5; 30 m, 250 μm, 0.25 μm). The column flow rate was 2.0 ml/min and the oven program started at 60°C for 1 min, ramped to 320°C at 6°C/min, and a final hold time of 15 min. The MS had an ionization energy of 70 eV with a scanning mass range of m/z 40-700 in Full

Scan mode. PAHs were identified and quantified in Full Scan mode based on authentic standards,

NIST 98 spectral library, fragmentation patterns, and retention times. For quantification, extracted ions were (m/z): 202 (pyrene), 237 (simonellite), 252 (simonellite, benzo[a]pyrene, benzofluoranthene, perylene, 1,2,3,4-tetrahydro-2,2,9-trimethylpicene (i.e., “β-amyrin derivative”

(Simoneit, 1986), referred to as “tetra-aromatic triterpane” in Schouten et al., (2007)), 255

(dehydroabietane), 268 (β-amyrin derivative (Wakeham et al., 1980)), 300 (coronene), 324 (β- amyrin derivative), 367 (hope-(17,21)-ene (Sessions et al., 2013)). n-Alkanes were from S1

Adducts (m/z): 43 and 57; pristane and phytane were from S1 Non-adducts (m/z): 57. Relative abundances were determined based on relative peak areas within a given fraction.

4.4.3. Normalized plant biomarker abundance to terrestrial organic carbon inputs

To account for production and preservation changes in TOC and eliminate the influence of changing marine organic carbon (TOCmarine) inputs, we normalize plant biomarker abundances

(Schouten et al., 2007) to terrestrial organic carbon (TOCterr) (Figure 4-3) rather than TOC

(TOCmarine + TOCterr). We use two proxies for terrestrial and marine biomass contributions (Sluijs and Dickens, 2012) to estimate the relative proportions of TOCterr and TOCmarine, the BIT index and the relative portions of terrestrial (pollen and spores) and marine (primarily dinoflagellate cysts) palynomorphs. The BIT index is based on the proportion of specific ether lipids (glycerol dialkyl glycerol tetraethers (GDGTs)) as defined by Hopmans et al. (2004). Distinctive terrestrial

98 GDGTs are produced by bacteria in soils and rivers, while the marine GDGT is primarily produced by pelagic archaea in the ocean. To calculate TOCterr, for each sample we multiply the

TOC value by the percentage of terrestrial inputs based on palynomorphs or the BIT index.

Because both proxies have limitations (Sluijs and Dickens, 2012), we use the average of the

TOCterr calculated by the two proxies. Uncertainty is the difference between TOCterr calculated from the proxies individually for a given sample (Figure 4-2 and Figure C-1).

4.5. Results

PAH abundance increased relative to both diterpenoid and triterpenoid abundances in the

PETM samples compared to the late Paleocene samples (Figure 4-2), and all three compound types were correlated with each other throughout the sampled section (Figure 4-4 and Figure 4-5).

Benzofluoranthene (BF; a representative PAH) is linearly correlated with simonellite and β- amyrin derivative (Figure 4-4) in samples pre-PETM (R2 = 0.85 and 0.78, respectively), PETM

(R2 = 0.70 and 0.57, respectively) and post-PETM (R2 = 0.995 and 0.98, respectively). The slopes of the pre-PETM correlations for both compounds are distinctly shallower than the correlations for PETM or post-PETM samples. The relationship between BF and β-amyrin derivative in

PETM samples is more variable (R2 = 0.17) relative to other time intervals and the BF and simonellite trends. Simonellite and β-amyrin derivative are linearly correlated with each other in pre-PETM (R2 = 0.62) and post-PETM (R2 = 0.98) samples and the relationship is less strong and more variable during the PETM (R2 = 0.38) (Figure 4-5). Coronene/Coronene+Pyrene ratios across the section were an average of 0.3 and ranged from 0.1 to 0.7 (Figure C-1).

The ratio of the sum of odd n-C25 to n-C33 alkanes relative to diterpenoids (simonellite and dehydroabietane) was similar to pollen composition trends and significantly greater than the ratio of β-amyrin derivative relative to diterpenoids (Schouten et al., 2007) (Figure 4-2). The

99 percent of the sum of the n-alkanes ratio ranged from 46% to 58% in the late Paleocene, 55% to

88% in the PETM, and 52% to 77% in the early Eocene. Temperature and percent of angiosperms

(pollen) across the section are linearly correlated (R2 = 0.5, when two outliers are removed), but not correlated within a given time interval (Figure 4-6).

Pristane/Phytane ratios (Pr/Ph) (Figure 4-2) were generally lower in PETM samples than before or after the event and ranged from 0.1 to 3.1 throughout the whole sampled section. Pr/Ph ratios in Paleocene pre-PETM samples ranged from 0.1 to 1.6. Ratios for PETM interval samples were less than 1 for most of the PETM, except for elevated values at ~382.5 mcd. Ratios for

Eocene post-PETM samples ranged from 0.3 to 3.1, with Eocene Pr/Ph ratios less than 1 immediately after the PETM interval, and then greater than 1 at 379 to 378 mcd.

Concentrations (ng/g TOCterr) of dehydroabietane, simonellite, β-amyrin derivative, and

C25-33 n-alkanes had little variation across the PETM event itself (but were more variable before and after the event) except for a peak in biomarker and angiosperm pollen abundance at the end of the body of the CIE (Figure 4-3).

4.6. Discussion

4.6.1. Relationship between PAHs and plant biomarkers

PAH abundances positively correlate with aromatic plant biomarkers but with different slopes before, during, and after the PETM (Figure 4-4). The steeper slopes during the PETM indicate either inputs or preservation of PAHs increased relative to the plant aromatic biomarkers in PETM samples compared to samples from before and after the event. Differential preservation, transportation, or production could attribute to the observed increased PAH abundance, but changes in PAH production is the likely explanation for several reasons.

100 First, transportation differences cannot fully explain the observed increased abundance of

PAHs relative to plant biomarkers because they are similarly transported by air and water

(Poynter et al., 1989; Baek et al., 1991). Both PAHs and the aromatic plant biomarkers have similar chemical structures with multiple aromatic rings. Each class has compounds that cover a range of masses (sizes); for example, PAH masses analyzed here range from 202 g/mol to 300 g/mol, while simonellite has a mass of 252 g/mol and 1,2,3,4-tetrahydro-2,2,9-trimethylpicene

(i.e., “β-amyrin derivative”, referred to as “tetra-aromatic triterpane” in Schouten et al., (2007)) has a mass of 324 g/mol. In our samples within each PAH or plant biomarker group selective preservation of one molecule type was not likely, both because organic matter preservation was high throughout the study interval, and because PAHs and the aromatic plant biomarkers have similar chemical structures and would likely degrade similarly (Sluijs et al., 2006).

Favorable conditions for organic carbon preservation before, during, and after the PETM are supported by the dominance of the 4-ring PAH, pyrene, over the 7-ring PAH, coronene, throughout the section (Figure C-1). Lower molecular weight PAHs are more susceptible to degradation than higher molecular weight PAHs because of their greater solubility and bioavailability (May et al., 1978; Killops and Massoud, 1992; Johnsen et al., 2005). The coronene/coronene+pyrene ratio averaged 30% and had no trend across the section, which suggests that PAHs were well-preserved in the sediments. The dominance of pyrene in the Arctic marine sediments was in stark contrast to the dominance of larger PAH like coronene in PETM terrestrial paleosols (Figure C-2), which experienced increased degradation of soil organic matter during the PETM (Baczynski et al., 2016; Denis et al., in prep-a; Denis et al., in prep-b).

Therefore, the increased PAH abundance relative to aromatic plant biomarkers primarily reflects changes in PAH production and, thus, a general trend of increased fire occurrence. We examine how the increase in PAHs relates to changes in plant composition (percentage of angiosperms relative to gymnosperms) and hydrological patterns during the PETM.

101 4.6.2. Plant biomarkers and pollen: Percent of angiosperms relative to gymnosperms

Schouten et al. (2007) observed that plant biomarker (triterpenoid/diterpenoid) ratios underestimated plant type composition (angiosperms versus gymnosperms) compared to estimates using pollen, and they suggested taphonomic differences accounted for the observed discrepancies in percent of angiosperms relative to gymnosperms. Based on more recent literature, triterpenoids (derived from angiosperms), such as β-amyrin derivative, are not preserved as well as diterpenoids (derived from gymnosperms) in terrestrial sediments

(Diefendorf et al., 2014). Triterpenoid to diterpenoid ratios potentially underestimate the abundance of angiosperms in the source paleovegetation, which would account for the discrepancy in the percent of angiosperms determined from aromatic plant biomarkers versus pollen reported by Schouten et al., (2007). Following the suggested practice of Diefendorf et al.

(2014), we use the ratio of plant wax n-alkanes to diterpenoids to represent the relative composition of angiosperms to gymnosperms contributions to the sediment core samples.

Across the PETM section, the concentration profile of the C29 n-alkane is similar to the angiosperm pollen abundance profile, and is dissimilar to that of the terrestrial plant aromatic biomarkers (Figure 4-3). Although many angiosperms and gymnosperms produce n-alkanes, the conifer families that do are primarily common today in Asia and the Southern Hemisphere (e.g.,

Podocarpaceae), and it is unlikely they lived in the Arctic during the PETM (Basinger et al.,

1994; Hill and Brodribb, 1999). Of the major conifer groups that do produce n-alkanes, hardly any produce the C29 n-alkane (except Podocarpaceae), and thus, the C29 n-alkane provides a strong phylogenetic signal for angiosperm inputs (Diefendorf et al., 2015). The paleovegetation proxy introduced by Diefendorf et al. (2014), which quantifies the ratio of n-alkanes to diterpenoids, yields estimates of angiosperm inputs in the Paleocene-Eocene sediments from the

Arctic that match the pollen record (Figure 4-2). These findings are consistent with the work by

102 Diefendorf et al. (2014) on terpenoid perservational biases, and we suggest the angiosperm contributions were previously underestimated by the triterpenoid to diterpenoid biomarker ratio

(Schouten et al., 2007). Both the Diefendorf proxy (n-alkanes to diterpenoids ratios) and pollen data indicate that the relative contribution of angiosperms increased during the PETM, reflecting a significant ecological shift to angiosperm-dominated vegetation.

4.6.3. Terrestrial plant inputs

Organic geochemical studies typically normalize biomarker abundances to total organic carbon (TOC) in order to account for changes in organic carbon production and preservation.

However, in these sediments there are two sources of carbon, marine-derived (TOCmarine) and terrestrially-derived (TOCterr). Because TOCmarine can vary independently from terrestrial organic contributions (Sluijs et al., 2006; Stein et al., 2006), compounds pertaining to plant inputs normalized to TOCterr, will better represent landscape signals. A variety of evidence indicates the relative proportions of TOCterr and TOCmarine changed before, during, and after the PETM.

Evidence from palynomorphs (dinoflagellate cysts, pollen, and spores), the Branched and

Isoprenoid Tetraether (BIT) index (Hopmans et al., 2004), and the Rock Eval hydrogen index suggested that the uppermost Paleocene sediments were proximal to the coast and were more terrestrially influenced by riverine inputs (Sluijs et al., 2006; Stein et al., 2006). During the PETM interval, evidence indicates aquatic carbon dominated inputs (Sluijs et al., 2006; Stein et al.,

2006). Pr/Ph ratios are consistent with these interpretations (Figure 4-2). Pr/Ph ratios <1 indicate marine inputs and ratios >1 indicate increasing dominance of terrestrial inputs (typically >3)

(Peters et al., 2005). At the end of the PETM, Pr/Ph ratios >1 coincide with low and non- detectable amounts of isorenieratene and monoaromatic isorenieratene derivatives, which signify a return to an oxic photic zone and oxic depositional conditions. Overall, these multiple lines of

103 evidence indicate more marine sourced organic inputs during the CIE interval relative to before and after the PETM (Sluijs et al., 2006; Stein et al., 2006; Figure 4-2). Although there is evidence for increased terrestrial runoff during the CIE (Robert and Kennett, 1994; Crouch et al., 2003), a rise in sea level during the PETM likely reduced the amount of terrestrial material that reached the Lomonosov Ridge (Sluijs et al., 2008a). In addition, increased marine productivity (Kaiho et al., 2006; Sluijs et al., 2008b) likely further diluted the relative proportion of TOCterr preserved in the PETM sediments.

Normalized concentrations (ng/g TOCterr) of dehydroabietane, simonellite, β-amyrin derivative, and C25-33 n-alkanes, reveal plant input did not change very much across the PETM event, except for a peak in biomarker and angiosperm pollen abundance at the end of the body of the CIE (Figure 4-3). As previously noted, biomarkers and pollen data indicate greater inputs from angiosperms during the PETM, and both biomarker (n-alkanes/diterpenoids) and pollen indicators show similar trends for most of the record, although they diverged at the end of the

13 PETM. The discrepancy occurs during the CIE recovery as δ Corg values and temperatures recovered to pre-PETM values and when isorenieratene and monoaromatic isorenieratene derivatives were below detection limit in the samples, which signified a return to an oxic photic zone (Figure 4-2 and Figure 4-3).

4.6.4. Implications for fire and ecosystem change

Changes in angiosperm inputs correlate with proxy evidence for warming temperatures

(Figure 4-6) and followed an inferred increase in moisture to the Arctic (Pagani et al., 2006b).

Pollen increased as temperature increased, and decreased at the end of the PETM as climate cooled. Sluijs et al. (2006) suggested that the increased angiosperm vegetation (Figure C-3) reflected an expanded growing season.

104 We interpret PAH abundances relative to aromatic plant biomarkers to reflect changes in

PAH production. The rise in normalized PAH values suggest increased fire occurrence was associated with the angiosperm vegetation shift, perhaps indicating greater prevalence of more fire-prone species, as was observed in England (Collinson et al., 2007; Collinson et al., 2009).

PAH abundance and plant biomarker abundance correlated before, during, and after the PETM event, but the slope changed during the PETM interval, and PAHs increased more relative to plant biomarkers (Figure 4-4). This suggests that terrestrial productivity and steady-state PAH production are strongly linked, but that during the PETM, PAH production outpaced plant productivity.

The interpreted increased moisture transport to the Arctic, based on the δD of n-alkanes

(Pagani et al., 2006b), preceded the shift toward angiosperm dominance, the increase in normalized PAH abundances and the rise in temperature during the PETM (Figure 4-2). Such a time lag emphasizes that a combination of factors influenced changes in fire occurrence, including a balance of fuel composition (e.g., vegetation amount and type), fuel availability (e.g., amount of vegetation that can burn based on humidity, precipitation, and temperature), and ignitions (i.e., lightning).

To better understand the relationships between fire and ecosystem changes during the

PETM in the Arctic we can combine what we know of climate and environmental conditions in the past, evidence for fire in the past, and relationships observed in studies of the modern. In the modern, forest fires in the Arctic are rarely considered, mainly because of the polar ice caps and because wildfires occur infrequently in modern landscapes north of the Arctic Circle (Higuera et al., 2008). The position of the Arctic near the pole means that the Paleocene-Eocene ecosystems functioned under strong light seasonality with continuous darkness throughout the winter and continuous light throughout summer, and short transitional seasons (e.g., lasting less than 60 days). Despite these extreme natural light conditions, a diverse forest ecosystem can survive, such

105 has been observed from terrestrial sections deposited in the early to mid-Eocene, a different time period but still a warm and humid climate (Jahren and Sternberg, 2003; Jahren, 2007; Zachos et al., 2008). The estimated terrestrial mean annual temperature (MAT) for the early Eocene Arctic was 13.2 ± 2.0 °C based on oxygen-isotope equilibration between environmental water and pedogenic carbonate from Axel Heiberg Island, located north of the Arctic Circle (Jahren and

Sternberg, 2003). In addition, cold month temperatures were above 0 °C, which implies a lack of freeze events (Basinger et al., 1994; Jahren and Sternberg, 2003; Jahren, 2007). The estimated terrestrial MAT for the late Paleocene was ~15°C and increased to ~21°C during the PETM based on the distribution of branched glycerol dialkyl glycerol tetraether membrane lipids (the MBT’-

CBT proxy) (Weijers et al., 2007; Peterse et al., 2012). Similar to Axel Heiberg Island in the early

Eocene, the continental Arctic during the PETM likely had cold month temperatures above freezing.

A time lag between increased moisture transport to the Arctic (Pagani et al., 2006b) and increased PAHs and angiosperms suggests first wetter conditions, followed by increased temperature, favored angiosperms and enhanced fire occurrence. Hence, drier conditions did not directly result in increased fire occurrence. Rather, a balance of variability in precipitation and vegetation composition and availability promotes fire occurrence. In this context, increased fire in a wetter Arctic suggests PETM precipitation was seasonal, or more variable on a longer timescale, and that hotter temperatures favored angiosperms and further facilitated burning.

Increased biomass burning during the PETM added CO2 to the atmosphere and inhibited soil organic carbon storage, which may have enhanced the greenhouse warming. The warmer temperatures and wetter, more variable precipitation may have stemmied gymnosperm growth and, by reducing competition, opened up the ecosystem to angiosperms. Or angiosperms may have migrated, with rising temperatures, from lower latitudes. The angiosperm-dominated fire- prone community may indicate that the PETM angiosperms recovered more rapidly after fire

106 disturbance than gymnosperms, potentially due to higher productivity or higher reproductive rates

(Bond and Midgley, 2012). As temperatures cooled at the end of the PETM, gymnosperm populations were revived and the ecosystem shifted back to a less fire-prone community.

4.7. Conclusions

In the Arctic Basin, even though plant input remained nearly constant despite large climate changes during the PETM, the landscape shifted to an angiosperm-dominated ecosystem.

Similar to observations by Diefendorf et al. (2014), the triterpenoid to diterpenoid ratio for the composition of angiosperms relative to gymnosperms underestimated the percent of angiosperms.

Instead, a ratio of n-alkanes to diterpenoids was similar to the angiosperm composition observed in pollen.

PAH abundance and plant biomarker abundance correlated before, during, and after the

PETM event, although during the PETM, PAHs increased more relative to plant biomarkers. A time lag between increased moisture transport to the Arctic and increased PAHs and angiosperms suggests wetter conditions, followed by increased temperature, favored angiosperms and enhanced fire occurrence. Increased fire in a wetter Arctic suggests PETM precipitation was seasonal, or variable on a longer timescale, and that hotter temperatures and angiosperm- dominated forests further facilitated burning.

107 4.8. Figures

Figure 4-1. Location of IODP Hole 302-4A (star) in the paleogeographical context of the late Paleocene-early Eocene (from Weijers et al., 2007). For reference, circles highlight the Bighorn Basin, Wyoming and Ellesmere Island (includes Axel Heiberg Island).

108

Figure 4-2. Depth profile of geochemical data. Core recovery column, grey represents recovered core and “x” marks intervals without material; error bars connected to Core 31X mark the uncertainty of its stratigraphic position (Sluijs et al., 13 2006). Depth profile of carbon isotope (δ Corg; black circle) and total organic carbon (TOC; white circle) (from Schouten et al., 2007)), terrestrial organic carbon (TOCterr.; black circle); marine organic carbon (TOCmarine; grey circle), Pristane/Phytane (black square) and dashed line at a ratio of 1, Pyrene/Terpenoids (=simonellite and β-amyrin derivative; black diamond), benzofluoranthene (BF)/Terpenoids (white diamond), Benzo[e]pyrene(BeP)/Terpenoids (black diamond), Coronene/Terpenoids (white diamond); δD values (white triangles with line) (Pagani et al., 2006) and sea water temperature based on TEX’86 index (black triangles with line) (Sluijs et al., 2006). %Angiosperm: Pollen (black diamond) and Triterpenoid/Diterpenoid (white triangle) ratios (Schouten et al., 2007); n-alkanes/Diterpenoids ratio (grey diamond). Horizontal dashed lines mark the PETM interval. Horizontal bars connected to TOC represent the uncertainty in TOCterr. or TOCmarine since each was determined from an average of the BIT index and %Terrestrial Palynomorphs (Sluijs and Dickens, 2012). If bar is not visible, uncertainty is less than the size of symbol.

109

Figure 4-3. Depth profile of molecular compounds and pollen. Core recovery column, where grey represents recovered core and “x” marks intervals without recovered material; error bars connected to Core 31X mark the uncertainty of its stratigraphic position (Sluijs et al., 2006). Carbon isotope values and biomarker concentrations were determined by Schouten et al., (2007). Angiosperms and gymnosperms column is pollen abundance (number/g TOCterr). Horizontal dashed lines mark the PETM interval. TOCterr. is the estimated terrestrial organic carbon and TOCmarine is the estimated marine organic carbon. Horizontal bars represent the uncertainty in TOCterr. or TOCmarine since each was determined from an average of the BIT index and %Terrestrial Palynomorphs (from Sluijs and Dickens (2012)). If bar is not visible, uncertainty is less than the size of symbol.

110 A)

30,000,000

25,000,000 y = 0.31x + 8E+04 R² = 0.85 Pre-PETM 20,000,000 Pre-PETM y = 1.8x + 4E+06 15,000,000 PETM R² = 0.64 PETM Post-PETM 10,000,000 ( Area/sample) ( y = 2.0x – 9E+05

Benzofluroanthenes R² = 0.995 5,000,000 Post-PETM

0 0 30,000,000 60,000,000 Simonellite (Area/sample)

30,000,000 y = 0.04x + 3E+05

25,000,000

R² = 0.78 Pre-PETM

Pre-PETM 20,000,000 PETM y = 0.1x + 1E+07 R² = 0.17 15,000,000 Post-PETM PETM

(Area/sample) 10,000,000 y = 0.2x + 3E+05

Benzofluroanthenes R² = 0.98 5,000,000 Post-PETM

0 0 200,000,000 400,000,000 Β-amyrin derivative (Area/sample)

Figure 4-4. A) Cross plot of Benzofluoranthenes (BF) with Simonellite; B) Cross plot of BF with β-amyrin derivative. Symbols represent: Paleocene Pre-PETM samples (solid diamond), PETM samples (open square), Eocene Post-PETM samples (white triangle). Linear trendlines marked by time interval.

111

60,000,000

50,000,000 y = 0.1x + 4E+06

R² = 0.62

40,000,000 Pre-PETM Pre-PETM y = 0.07x + 3E+06 30,000,000 PETM R² = 0.38 PETM

Simonellite Simonellite Post-PETM

(Area/sample) 20,000,000 y = 0.1x + 5E+05 R² = 0.97 10,000,000 Post-PETM

0 0 200,000,000 400,000,000 Β-amyrin derivative (Area/sample)

Figure 4-5. Cross plot of Simonellite with β-amyrin derivative. Symbols represent: Paleocene Pre-PETM samples (solid diamond), PETM samples (open square), Eocene Post-PETM samples (white triangle). Linear trendlines marked by time interval.

100 Pre-PETM

80 PETM (onset+body) PETM (Recovery) 60 Post-PETM Linear (All) 40 y = 5.9x - 54 R² = 0.50

%Angiosperm %Angiosperm (pollen) 20

0 16 18 20 22 24 Temperature (°C)

Figure 4-6. Correlation of percent of angiosperms (%Angiosperms) with temperature. Plot is based on pollen data and sea surface temperature (based on the TEX’86 index) from Sluijs et al., (2006). Symbols represent: Paleocene Pre-PETM samples (solid diamond), PETM samples (square); carbon isotope excursion (CIE) onset and body (open) and CIE recovery (grey fill), Eocene Post-PETM samples (white triangle). Linear trendline for all samples (black line).

112

4.9. References

Backman, J., Moran, K., MicInroy, D.B., Mayer, L.A., and the Expedition 302 Scientists, 2006, Sites M0001–M0004, in Proceedings of the Integrated Ocean Drilling Program, 302, Integrated Ocean Drilling Program Management International, Inc., Edinburgh. Baek, S.O., Field, R.A., Goldstone, M.E., Kirk, P.W., Lester, J.N., and Perry, R., 1991, A review of atmospheric polycyclic aromatic hydrocarbons: Sources, fate and behavior: Water, Air, and Soil Pollution, v. 60, no. 3-4, p. 279–300, doi: 10.1007/BF00282628. Basinger, J.F., Greenwood, D.R., and Sweda, T., 1994, Early Tertiary vegetation of Arctic Canada and its relevance to paleoclimatic interpretation, in Boulter, M.C. ed., Arctic plants and climates: 65 million years of change, Springer-Verlag, Berlin. Bird, M.I., Wynn, J.G., Saiz, G., Wurster, C.M., and McBeath, A., 2014, The Pyrogenic Carbon Cycle: Annual Review of Earth and Planetary Sciences, v. 43, p. 150223150959000, doi: 10.1146/annurev-earth-060614-105038. Bond, W.J., and Midgley, J.J., 2012, Fire and the Angiosperm Revolutions: International Journal of Plant Sciences, v. 173, no. 6, p. 569–583, doi: 10.1086/665819. Clark, J.S., 1990, Fire and Climate Change During the Last 750 Yr in Northwestern Minnesota: Ecological Monographs, v. 60, no. 2, p. 135–159. Collinson, M.E., Steart, D.C., Harrington, G.J., Hooker, J.J., Scott, A.C., Allen, L.O., Glasspool, I.J., and Gibbons, S.J., 2009, Palynological evidence of vegetation dynamics in response to palaeoenvironmental change across the onset of the Paleocene-Eocene Thermal Maximum at Cobham, Southern England: Grana, v. 48, no. 909677517, p. 38–66, doi: 10.1080/00173130802707980. Collinson, M.E., Steart, D.C., Scott, A.C., Glasspool, I.J., and Hooker, J.J., 2007, Episodic fire, runoff and deposition at the Palaeocene-Eocene boundary: Journal of the Geological Society, v. 164, no. 1, p. 87–97, doi: 10.1144/0016-76492005-185. Cook, E.. E.R., Woodhouse, C. a C.A., Eakin, C.M.M., Meko, D.M.D.M., and Stahle, D.W.D.W., 2004, Long-term aridity changes in the western United States: Science, v. 306, no. 5698, p. 1015–1018, doi: 10.1126/science.1102586. Coulon, F., Pelletier, E., Gourhant, L., and Delille, D., 2005, Effects of nutrient and temperature on degradation of petroleum hydrocarbons in contaminated sub-Antarctic soil: Chemosphere, v. 58, no. 10, p. 1439–1448, doi: 10.1016/j.chemosphere.2004.10.007. Crouch, E.M., Dickens, G.R., Brinkhuis, H., Aubry, M.P., Hollis, C.J., Rogers, K.M., and Visscher, H., 2003, The Apectodinium acme and terrestrial discharge during the Paleocene-Eocene thermal maximum: New palynological, geochemical and calcareous nannoplankton observations at Tawanui, New Zealand: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 194, no. 4, p. 387–403, doi: 10.1016/S0031- 0182(03)00334-1. Cui, Y., Kump, L.R., Ridgwell, A.J., Charles, A.J., Junium, C.K., Diefendorf, A.F., Freeman, K.H., Urban, N.M., and Harding, I.C., 2011, Slow release of fossil carbon during the

113 Palaeocene–Eocene Thermal Maximum: Nature Geoscience, v. 4, no. 7, p. 481–485, doi: 10.1038/ngeo1179. Daniau, A.L., Bartlein, P.J., Harrison, S.P., Prentice, I.C., Brewer, S., Friedlingstein, P., Harrison- Prentice, T.I., Inoue, J., Izumi, K., Marlon, J.R., Mooney, S., Power, M.J., Stevenson, J., Tinner, W., et al., 2012, Predictability of biomass burning in response to climate changes: Global Biogeochemical Cycles, v. 26, no. 4, p. 1–12, doi: 10.1029/2011GB004249. Denis, E.H., Foreman, B.Z., and Freeman, K.H., in prep-a, Widespread soil carbon loss during the Paleocene-Eocene Thermal Maximum: Earth and Planetary Science Letters. Denis, E.H., Foreman, B.Z., Maibauer, B.J., Bowen, G.J., Baczynski, A.A., McInerney, F.A., Collinson, M.E., Belcher, C.M., Wing, S.L., and Freeman, K.H., in prep-b. Decreased soil carbon in a warming world: Degraded pyrogenic carbon during the Paleocene- Eocene Thermal Maximum (PETM): Geology. Denis, E.H., Foreman, B.Z., Maibauer, B.J., Bowen, G.J., Collinson, M.E., Belcher, C.M., and Freeman, K.H., 2014, Carbon burn-down in a greenhouse world: Wildfires and soil carbon loss across the Paleocene-Eocene Thermal Maximum (PETM): American Geophysical Union Annual Meeting, Sanfrancisco, CA, Abstract PP11A–1332, Poster. Denis, E.H., Toney, J.L., Tarozo, R., Anderson, R.S., Roach, L.D., and Huang, Y., 2012, Polycyclic aromatic hydrocarbons (PAHs) in lake sediments record historic fire events: Validation using HPLC-fluorescence detection: Organic Geochemistry, v. 45, p. 7–17, doi: 10.1016/j.orggeochem.2012.01.005. Diefendorf, A.F., Freeman, K.H., and Wing, S.L., 2014, A comparison of terpenoid and leaf fossil vegetation proxies in Paleocene and Eocene Bighorn Basin sediments: Organic Geochemistry, v. 71, p. 30–42, doi: 10.1016/j.orggeochem.2014.04.004. Diefendorf, A.F., Leslie, A.B., and Wing, S.L., 2015, Leaf wax composition and carbon isotopes vary among major conifer groups: Geochimica et Cosmochimica Acta, v. 170, p. 145– 156, doi: 10.1016/j.gca.2015.08.018. Flannigan, M.D., Krawchuk, M.A., De Groot, W.J., Wotton, B.M., and Gowman, L.M., 2009, Implications of changing climate for global wildland fire: International Journal of Wildland Fire, v. 18, no. 5, p. 483–507, doi: 10.1071/WF08187. Forbes, M.S., Raison, R.J., and Skjemstad, J.O., 2006, Formation, transformation and transport of black carbon (charcoal) in terrestrial and aquatic ecosystems: Science of the Total Environment, v. 370, no. 1, p. 190–206, doi: 10.1016/j.scitotenv.2006.06.007. Foreman, B.Z., Heller, P.L., and Clementz, M.T., 2012, Fluvial response to abrupt global warming at the Palaeocene/Eocene boundary: Nature, v. 491, no. 7422, p. 92–5, doi: 10.1038/nature11513. Handley, L., O’Halloran, A., Pearson, P.N., Hawkins, E., Nicholas, C.J., Schouten, S., McMillan, I.K., and Pancost, R.D., 2012, Changes in the hydrological cycle in tropical East Africa during the Paleocene-Eocene Thermal Maximum: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 329-330, p. 10–21, doi: 10.1016/j.palaeo.2012.02.002. Hessl, A.E., 2011, Pathways for climate change effects on fire: Models, data, and uncertainties: Progress in Physical Geography, v. 35, p. 393–407, doi: 10.1177/0309133311407654. Higuera, P.E., Brubaker, L.B., Anderson, P.M., Brown, T.A., Kennedy, A.T., and Hu, F.S., 2008, Frequent fires in ancient shrub tundra: Implications of paleorecords for arctic environmental change: PLoS ONE, v. 3, no. 3, p. 1–8, doi: 10.1371/journal.pone.0001744. Higuera, P.E., Brubaker, L.B., Anderson, P.M., Hu, F.S., and Brown, T. A., 2014, Vegetation mediated the impacts of postglacial climate change on fire regimes in the south-central Brooks Range , Alaska: v. 79, no. 2, p. 201–219, doi: 10.1890/07-2019.1.

114 Hill, R.S., and Brodribb, T.J., 1999, Turner Review No. 2. Southern Conifers in Time and Space: Australian Journal of Botany, v. 47, no. 5, p. 639–696, doi: http://dx.doi.org/10.1071/BT98093. Hopmans, E.C., Weijers, J.W.H., Schefuß, E., Herfort, L., Sinninghe Damsté, J.S., and Schouten, S., 2004, A novel proxy for terrestrial organic matter in sediments based on branched and isoprenoid tetraether lipids: Earth and Planetary Science Letters, v. 224, no. 1-2, p. 107– 116, doi: 10.1016/j.epsl.2004.05.012. Jahren, A.H., 2007, The Arctic Forest of the Middle Eocene: Annual Review of Earth and Planetary Sciences, v. 35, no. 1, p. 509–540, doi: 10.1146/annurev.earth.35.031306.140125. Jahren, A.H., and Sternberg, L.S.L., 2003, Humidity estimate for the middle Eocene Arctic rain forest: Geology, v. 31, no. 5, p. 463–466, doi: 10.1130/0091- 7613(2003)031<0463:HEFTME>2.0.CO;2. Jaramillo, C., Ochoa, D., Contreras, L., Pagani, M., Carvajal-Ortiz, H., Pratt, L.M., Krishnan, S., Cardona, A., Romero, M., Quiroz, L., Rodriguez, G., Rueda, M.J., de la Parra, F., Morón, S., et al., 2010, Effects of rapid global warming at the Paleocene-Eocene boundary on neotropical vegetation: Science, v. 330, no. 6006, p. 957–961, doi: 10.1126/science.1193833. Johnsen, A.R., Wick, L.Y., and Harms, H., 2005, Principles of microbial PAH-degradation in soil: Environmental Pollution, v. 133, no. 1, p. 71–84, doi: 10.1016/j.envpol.2004.04.015. Kaiho, K., Takeda, K., Petrizzo, M.R., and Zachos, J.C., 2006, Anomalous shifts in tropical Pacific planktonic and benthic foraminiferal test size during the Paleocene-Eocene thermal maximum: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 237, no. 2-4, p. 456–464, doi: 10.1016/j.palaeo.2005.12.017. Killops, S.D., and Massoud, M.S., 1992, Polycyclic aromatic hydrocarbons of pyrolytic origin in ancient sediments: evidence for Jurassic vegetation fires: Organic Geochemistry, v. 18, no. 1, p. 1–7, doi: 10.1016/0146-6380(92)90137-M. Knicker, H., 2011, Pyrogenic organic matter in soil: Its origin and occurrence, its chemistry and survival in soil environments: Quaternary International, v. 243, no. 2, p. 251–263, doi: 10.1016/j.quaint.2011.02.037. Kraus, M.J., and Riggins, S., 2007, Transient drying during the Paleocene-Eocene Thermal Maximum (PETM): Analysis of paleosols in the bighorn basin, Wyoming: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 245, no. 3-4, p. 444–461, doi: 10.1016/j.palaeo.2006.09.011. Kuzyakov, Y., Subbotina, I., Chen, H., Bogomolova, I., and Xu, X., 2009, Black carbon decomposition and incorporation into soil microbial biomass estimated by 14C labeling: Soil Biology and Biochemistry, v. 41, no. 2, p. 210–219, doi: 10.1016/j.soilbio.2008.10.016. Von Lützow, M., Kögel-Knabner, I., Ekschmitt, K., Matzner, E., Guggenberger, G., Marschner, B., and Flessa, H., 2006, Stabilization of organic matter in temperate soils: Mechanisms and their relevance under different soil conditions - A review: European Journal of Soil Science, v. 57, no. 4, p. 426–445, doi: 10.1111/j.1365-2389.2006.00809.x. Malingreau, J.P., Stephens, G., and Fellows, L., 1985, Remote Sensing of Forest Fires: Kalimantan and North Borneo in 1982-83: Ambio, v. 14, no. 6, p. 314–321. Margolis, E.Q., and Balmat, J., 2009, Fire history and fire-climate relationships along a fire regime gradient in the Santa Fe Municipal Watershed, NM, USA: Forest Ecology and Management, v. 258, no. 11, p. 2416–2430, doi: 10.1016/j.foreco.2009.08.019. Marynowski, L., and Simoneit, B.R.T., 2009, Widespread Upper Triassic To Lower Jurassic Wildfire Records From Poland: Evidence From Charcoal and Pyrolytic Polycyclic

115 Aromatic Hydrocarbons: Palaios, v. 24, no. 12, p. 785–798, doi: 10.2110/palo.2009.p09- 044r. Masclet, P., Cachier, H., Liousse, C., and Wortham, H., 1995, Emissions of polycyclic aromatic hydrocarbons by savanna fires: Journal of Atmospheric Chemistry, v. 22, no. 1-2, p. 41– 54, doi: 10.1007/BF00708180. Masiello, C.A., 2004, New directions in black carbon organic geochemistry: Marine Chemistry, v. 92, no. 1-4 Spec. Iss., p. 201–213, doi: 10.1029/2002GB001939. May, W.E., Wasik, S.P., and Freeman, D.H., 1978, Determination of the Solubility Behavior of Some Polycyclic Aromatic Hydrocarbons in Water: Analytical chemistry, v. 50, no. 7, p. 997–1000, doi: 10.1021/ac50029a042. McInerney, F.A., and Wing, S.L., 2011, The Paleocene-Eocene Thermal Maximum: A Perturbation of Carbon Cycle, Climate, and Biosphere with Implications for the Future: Annual Review of Earth and Planetary Sciences, v. 39, no. 1, p. 489–516, doi: 10.1146/annurev-earth-040610-133431. Moore, E.A., and Kurtz, A.C., 2008, Black carbon in Paleocene-Eocene boundary sediments: A test of biomass combustion as the PETM trigger: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 267, no. 1-2, p. 147–152, doi: 10.1016/j.palaeo.2008.06.010. Nagy, J., Jargvoll, D., Dypvik, H., Jochmann, M., and Riber, L., 2013, Environmental changes during the paleocene-eocene thermal maximum in spitsbergen as reflected by benthic foraminifera: Polar Research, v. 32, p. 19737, doi: 10.3402/polar.v32i0.19737. Pagani, M., Pedentchouk, N., Huber, M., Sluijs, A., Schouten, S., Brinkhuis, H., Sinninghe Damsté, J.S., and Dickens, G.R., 2006, Arctic hydrology during global warming at the Palaeocene/Eocene thermal maximum: Nature, v. 442, no. 7103, p. 671–675, doi: 10.1038/nature05043. Page, S.E., Siegert, F., Rieley, J.O., Boehm, H.-D. V., Jaya, A., and Limin, S., 2002, The amount of carbon released from peat and forest fires in Indonesia during 1997: Nature, v. 420, no. 1, p. 61–66, doi: 10.1038/nature01141.1. Peters, K.E., Walters, C.C., and Moldowan, J.M., 2005, The Biomarker Guide. Volume 2: Biomarkers and Isotopes in Petroleum Exploration and Earth History.: Cambridge University Press, Cambridge. Peterse, F., van der Meer, J., Schouten, S., Weijers, J.W.H., Fierer, N., Jackson, R.B., Kim, J.H., and Sinninghe Damsté, J.S., 2012, Revised calibration of the MBT-CBT paleotemperature proxy based on branched tetraether membrane lipids in surface soils: Geochimica et Cosmochimica Acta, v. 96, p. 215–229, doi: 10.1016/j.gca.2012.08.011. Poynter, J.G., Farrimond, P., Robinson, N., and Eglinton, G., 1989, Aeolian-derived higher plant lipids in the marine sedimentary record: links with palaeoclimate, in Leinen, M. and Samthein, M. eds., Paleoclimatology and paleometeorology: modern and past patterns of global atmospheric transport, Kluwer, Dordrecht, p. 435–462. Ramdahl, T., 1983, Retene - a molecular marker of wood combustion in ambient air: Nature, v. 306, p. 580–582, doi: 10.1038/306580a0. Robert, C., and Kennett, J.P., 1994, Antarctic subtropical humid episode at the Paleocene-Eocene boundary: Clay-mineral evidence: Geology, v. 22, no. 3, p. 211, doi: 10.1130/0091- 7613(1994)022<0211:ASHEAT>2.3.CO;2. Röhl, U., Westerhold, T., Bralower, T.J., and Zachos, J.C., 2007, On the duration of the Paleocene-Eocene thermal maximum (PETM): Geochemistry, Geophysics, Geosystems, v. 8, no. 12, doi: 10.1029/2007GC001784. Schmidt, M.W.I., and Noack, A.G., 2000, Analysis, distribution, implications, and current challenges: Global Biogeochemical Cycles, v. 14, no. 3, p. 777–793.

116 Schouten, S., Woltering, M., Rijpstra, W.I.C., Sluijs, A., Brinkhuis, H., and Sinninghe Damsté, J.S., 2007, The Paleocene-Eocene carbon isotope excursion in higher plant organic matter: Differential fractionation of angiosperms and conifers in the Arctic: Earth and Planetary Science Letters, v. 258, no. 3-4, p. 581–592, doi: 10.1016/j.epsl.2007.04.024. Secord, R., Gingerich, P.D., Lohmann, K.C., and Macleod, K.G., 2010, Continental warming preceding the Palaeocene-Eocene thermal maximum: Nature, v. 467, no. 7318, p. 955– 958, doi: 10.1038/nature09441. Sessions, A.L., Zhang, L., Welander, P. V., Doughty, D., Summons, R.E., and Newman, D.K., 2013, Identification and quantification of polyfunctionalized hopanoids by high temperature gas chromatography-mass spectrometry: Organic Geochemistry, v. 56, p. 120–130, doi: 10.1016/j.orggeochem.2012.12.009. Sluijs, A., Brinkhuis, H., Crouch, E.M., John, C.M., Handley, L., Munsterman, D., Bohaty, S.M., Zachos, J.C., Reichart, G.J., Schouten, S., Pancost, R.D., Sinninghe Damsté, J.S., Welters, N.L.D., Lotter, A.F., et al., 2008a, Eustatic variations during the Paleocene- Eocene greenhouse world: Paleoceanography, v. 23, no. 4, p. 1–18, doi: 10.1029/2008PA001615. Sluijs, A., and Dickens, G.R., 2012, Assessing offsets between the δ13C of sedimentary components and the global exogenic carbon pool across early Paleogene carbon cycle perturbations: Global Biogeochemical Cycles, v. 26, no. 4, p. 1–14, doi: 10.1029/2011GB004224. Sluijs, A., Röhl, U., Schouten, S., Brumsack, H.J., Sangiorgi, F., Sinninghe Damsté, J.S., and Brinkhuis, H., 2008b, Article late Paleocene - Early Eocene paleoenvironments with special emphasis on the Paleocene-Eocene thermal maximum (Lomonosov Ridge, Integrated Ocean Drilling Program Expedition 302): Paleoceanography, v. 23, no. 1, p. 1–17, doi: 10.1029/2007PA001495. Sluijs, A., Schouten, S., Pagani, M., Woltering, M., Brinkhuis, H., Sinninghe Damsté, J.S., Dickens, G.R., Huber, M., Reichart, G.-J., Stein, R., Matthiessen, J., Lourens, L.J., Pedentchouk, N., Backman, J., et al., 2006, Subtropical Arctic Ocean temperatures during the Palaeocene/Eocene thermal maximum: Nature, v. 441, no. 7093, p. 610–613, doi: 10.1038/nature04668. Smith, J.J., Hasiotis, S.T., Kraus, M.J., and Woody, D.T., 2008, Relationship of floodplain ichnocoenoses to paleopedology, paleohydrology, and paleoclimate in the Willwood Formation, Wyoming, during the Paleocene-Eocene Thermal Maximum: Palaios, v. 23, no. 9-10, p. 683–699, doi: 10.2110/palo.2007.p07-080r. Smith, F.A., Wing, S.L., and Freeman, K.H., 2007, Magnitude of the carbon isotope excursion at the Paleocene-Eocene thermal maximum: The role of plant community change: Earth and Planetary Science Letters, v. 262, no. 1-2, p. 50–65, doi: 10.1016/j.epsl.2007.07.021. Stein, R., Boucsein, B., and Meyer, H., 2006, Anoxia and high primary production in the Paleogene central Arctic Ocean: First detailed records from Lomonosov Ridge: Geophysical Research Letters, v. 33, no. 18, p. 2–7, doi: 10.1029/2006GL026776. Wakeham, S.G., Schaffner, C., and Giger, W., 1980, Polycyclic aromatic hydrocarbons in Recent lake sediments - I. Compounds having anthropogenic origins: Geochimica et Cosmochimica Acta, v. 44, p. 403–413. Weijers, J.W.H., Schouten, S., Sluijs, A., Brinkhuis, H., and Sinninghe Damsté, J.S., 2007, Warm arctic continents during the Palaeocene-Eocene thermal maximum: Earth and Planetary Science Letters, v. 261, no. 1-2, p. 230–238, doi: 10.1016/j.epsl.2007.06.033. Westerhold, T., Röhl, U., and Laskar, J., 2012, Time scale controversy: Accurate orbital calibration of the early Paleogene: Geochemistry, Geophysics, Geosystems, v. 13, no. 6, p. 1–19, doi: 10.1029/2012GC004096.

117 Westerling, A.L., 2006, Warming and Earlier Spring Increase Western U.S. Forest Wildfire Activity: v. 1161, no. August, p. 940–943, doi: 10.1126/science.1128834. Whitlock, C., and Anderson, R.S., 2003, Fire History Reconstructions Based on Sediment Records from Lakes and Wetlands, in Swetnam, T.W., Montenegro, G., and Veblen, T.T. eds., Fire and Climate Change in the Americas, Springer, New York, p. 3–31. Whitlock, C., and Larsen, C., 2001, Charcoal as a fire proxy, in Smol, J.P., Birks, H.J.B., and Last, W.M. eds., Tracking environmental change using lake sediments, Kluwer Academic Publishers, Dordrecht, p. 75–97. Wing, S.L., and Currano, E.D., 2013, Plant response to a global greenhouse event 56 million years ago: American Journal of Botany, v. 100, no. 7, p. 1234–1254, doi: 10.3732/ajb.1200554. Wing, S.L., Harrington, G.J., Smith, F.A., Bloch, J.I., Boyer, D.M., and Freeman, K.H., 2005, Transient Floral Change and Rapid Global Warming at the Paleocene-Eocene Boundary: Science, v. 310, no. 5750, p. 993–996, doi: 10.1126/science.1116913. Yunker, M.B., Macdonald, R.W., Vingarzan, R., Mitchell, R.H., Goyette, D., and Sylvestre, S., 2002, PAHs in the Fraser River basin: A critical appraisal of PAH ratios as indicators of PAH source and composition: Organic Geochemistry, v. 33, no. 4, p. 489–515, doi: 10.1016/S0146-6380(02)00002-5. Zachos, J.C., Dickens, G.R., and Zeebe, R.E., 2008, An early Cenozoic perspective on greenhouse warming and carbon-cycle dynamics.: Nature, v. 451, no. 7176, p. 279–283. Zachos, J.C., Nicolo, M., Raffi, I., Lourens, L.J., McCarren, H., and Kroon, D., 2005, Rapid Acidification of the Ocean During the Paleocene-Eocene Thermal Maximum: Science, v. 308, p. 1611–1615, doi: 10.1126/science.1109004. Zeebe, R.E., Ridgwell, A., and Zachos, J.C., 2016, Anthropogenic carbon release rate unprecedented during the past 66 million years: Nature Geoscience, v. 9, no. April, p. 325–329, doi: 10.1038/ngeo2681. Zeebe, R.E., and Zachos, J.C., 2013, Long-term legacy of massive carbon input to the Earth system: Anthropocene versus Eocene: Philosophical transactions. Series A, Mathematical, physical, and engineering sciences, v. 371, no. 2001, p. 20120006, doi: 10.1098/rsta.2012.0006.

118

Chapter 5

Summary and Future Work

5.1. Research Summary

Sedimentary records of polycyclic aromatic hydrocarbons (PAHs), an intermediate phase of refractory carbon, reflect both fire occurrence and relative preservation of organic carbon. I developed a way to use concentrations of both PAHs and total organic carbon (TOC) in a novel metric of degradation, and to estimate the extent of autochthonous organic carbon degradation and the resulting enhancement of more refractory carbonaceous material within soil deposits

(e.g., allochthonous fossil carbon (Bataille et al., 2013; Baczynski et al., 2016)). PAHs and other organic proxies yield lessons about how a warming climate and associated changes in the hydrologic cycle affect the production and preservation of terrestrial organic carbon. This work focused on the abrupt global warming event, the Paleocene-Eocene Thermal Maximum (PETM), at study sites in the western USA and the Arctic.

In Chapter 2, I focused on the source and fate of terrestrial organic carbon and pyrogenic carbon during the PETM at two sites in the Bighorn Basin, Wyoming. In this chapter, I first developed the use of PAHs as a proxy for intermediately refractory carbon to help discern the relative preservation of different carbon phases in soils. I estimated extent of autochthonous carbon degradation based on changes in intermediately refractory compounds relative to TOC, by assuming a conservative relationship between autochthonous carbon and PAH degradation.

119 In this study, all forms of soil carbon decreased during the PETM, and because PAH concentrations decreased even more than TOC this suggests a more refractory phase was preferentially preserved. The decreased organic carbon, including relatively refractory carbon

(i.e., PAHs and charcoal), during the PETM indicates enhanced rates of decay in a hotter and more seasonal climate relative to the late Paleocene (Wing et al., 2005; Kraus and Riggins, 2007;

Smith et al., 2008; McInerney and Wing, 2011). Overall, our results reflect that elevated temperatures caused greater organic carbon loss by soil respiration. These results imply warming soils can lose significant amounts of organic carbon, which will enhance climate warming, and potentially reduce soil fertility.

In Chapter 3, I applied PAH data to estimate soil organic carbon degradation in another fluvial site in the western USA, the Piceance Basin, Colorado. This study offered a regional perspective on the extent to which the aggressive organic carbon degradation during the PETM was a widespread event versus a local phenomenon in the Bighorn Basin, Wyoming (Baczynski et al., 2013; Bataille et al., 2013; Baczynski et al., 2016; Denis et al., in prep). In addition, this chapter helped us understand potential mechanisms that stabilized and destabilized carbon during the PETM. The use of PAHs and TOC in this context was especially informative in a fluvial setting with low organic carbon and variable lithology (clay, silt, sand). I tested the hypothesis that increased organic carbon degradation and refractory (allochthonous) carbon inputs during the

13 PETM contributed to the high variability in bulk organic carbon isotope values (δ Corg), which distorts the recorded carbon isotope excursion (CIE). Results indicated that the high variability in

13 δ Corg values in terrestrial sediments was linked to increased soil carbon degradation and greater proportions of refractory allochthonous carbon during the PETM. These findings provide

13 additional support to previous studies that emphasized that soil carbon δ Corg variability and attenuated CIE magnitude in terrestrial archives is linked to soil organic matter degradation

120 (Baczynski et al., 2013; Baczynski et al., 2016).

The positive correlation between elemental oxide weight percents (e.g., Al2O3 and TiO2) and %TOC suggests organic carbon stabilization was associated with clay minerals. Both Basin

Substation and Piceance Basin, which are inferred wetter sites compared to Polecat Bench, had greater percent loss of organic carbon during the PETM. Reduced soil organic matter preservation during the PETM was due to a combination of increased temperatures (which increased microbial decomposition rates), decreased clay content and changes in mineralogy (which inhibited stability of fresh carbon), and fluctuations in soil moisture (which destabilized older, refractory carbon).

Enhanced soil carbon loss appears to be a widespread phenomenon during the PETM beyond the

Bighorn Basin, Wyoming and is a regional, and potentially global, phenomenon.

In Chapter 4, I examined the occurrence of fire in the Arctic paleoecosystem during the

PETM global warming event and the dynamics between fire, precipitation, and vegetation changes. Plant input was approximately constant across the PETM, but the ecosystem shifted in favor of angiosperms. The ratio of n-alkanes to diterpenoids (Diefendorf et al., 2014) tracked well with the angiosperm composition observed in pollen (Schouten et al., 2007). A time lag between increased moisture transport to the Arctic and increased PAHs and angiosperms suggests wetter conditions, followed by increased temperature, favored angiosperms and enhanced fire occurrence. Thus, in the paleorecord drier conditions did not directly result in greater fire occurrence. Rather, like in the modern, tendency for fire is determined by a balance of variability in precipitation and sufficiently flammable vegetation. In this way, increased fire in a wetter

Arctic suggests PETM precipitation was seasonal, or variable on a longer timescale, and that hotter temperatures and ecosystem shifts further facilitated burning.

Overall, we used PAHs as a primary signal of production (i.e., fire occurrence) in marine

121 sediments and as a secondary signal of preservation (e.g., organic carbon degradation) in ancient soils. Our results highlight that terrestrial organic carbon was better preserved in the marine section than the fluvial sections. Extensive soil organic carbon loss during the PETM suggests soils served as a source of carbon to the atmosphere rather than a sink. The degradation and loss of older, pre-PETM carbon in soils was a potential source of carbon that leaked continually (over tens of thousands of years) during the PETM and contributed to the delayed recovery of the

13 carbon isotope excursion to pre-PETM δ C values. Although CO2 released from microbial respiration enhances the greenhouse warming, increased organic carbon preservation in the marine realm may have counteracted the enhanced carbon output from soils.

5.2. Future Work

Future studies using PAHs and TOC at other terrestrial sites during the PETM are needed to further test the hypothesis that soil carbon degradation was a widespread and potentially global phenomenon. In addition, it would be helpful to explore the extent to which soil carbon

13 degradation and allochthonous carbon inputs affect the bulk organic carbon isotope (δ Corg) values at other sites to help reconstruct a more representative isotope excursion. A compilation of results from sites with a variety of hydrologic regimes (e.g., dry, wet, seasonal) would help to elucidate if soil moisture was a main control on changes in soil carbon degradation in a hotter climate.

Where organic carbon is well preserved, future studies could employ PAHs as an indicator for fire dynamics during climate perturbations. PAH data, in conjunction with environmental data (e.g., temperature, precipitation, and vegetation), can unveil fire, ecosystem, and climate relationships in the past and inform predictions for fire tendency in a warming climate. Examining sites at additional latitudes would provide mechanistic insights to how

122 different precipitation patterns and vegetation compositions affect fire occurrence. These data sets could ultimately help constrain models that predict the tendency for fire based on changing regional to global climate patterns.

My original investigation of fire during the PETM was driven by the idea that combustion of 13C-depleted carbon deposits could have contributed to the negative carbon isotope excursion (see Appendix D). Results from the Arctic suggest that fire was a consequence of global warming rather than the cause. The extensive organic carbon degradation in the terrestrial sites in the western USA, though, obscured evidence for changes in fire occurrence. Nevertheless, the issue of what material burned during the PETM is of interest, and one could use the δ13C

13 value of PAHs (δ CPAH) to help distinguish between the burning of PETM biomass and

Paleocene peat or coal, similar to methods on black carbon by Moore and Kurtz (2008). In

13 addition, the δ CPAH could provide another way to measure the carbon isotope excursion.

Especially in fluvial settings, internal (autogenic) sediment processes may obscure the preservation of paleoenvironmental (allogenic) signals in the stratigraphic record or may even impart apparent stratigraphic patterns (Hajek et al., 2010; Wang et al., 2011). Jerolmack and

Paola (2010) showed that sediment transport processes can destroy environmental signals, if the timescale and amplitude of the signal are within the range of internal system dynamics. Given the high variability in sedimentation in fluvial systems, where floodplain and paleosols deposits are broken by coarse grain channel bodies, a fluvial model combined with an organic carbon signal could help to disentangle the degree to which changes in organic carbon reflect sedimentary process rather than environmental conditions.

The molecular tools developed and applied in this dissertation would be informative during other time periods to understand questions of soil carbon degradation and fire occurrence.

For example, PAH work done at other Eocene hyperthermals could serve as an additional comparison to the PETM and further inform changes in soil carbon degradation and fire

123 occurrence during a warming climate. Limited studies have applied PAHs as an indicator for fire in the paleorecord, but the relationship between fire and climate is important for understanding changes in ecosystems (especially changes in vegetation). As long as the effects of production, transportation, and preservation are considered when results are interpreted, PAHs can be an informative tool to help reconstruct environmental conditions in the past.

5.3. References

Baczynski, A.A., McInerney, F.A., Wing, S.L., Kraus, M.J., Bloch, J.I., Boyer, D.M., Secord, R., Morse, P.E., and Fricke, H.C., 2013, Chemostratigraphic implications of spatial variation in the Paleocene-Eocene Thermal Maximum carbon isotope excursion, SE Bighorn Basin, Wyoming: Geochemistry, Geophysics, Geosystems, v. 14, no. 10, p. 4133–4152, doi: 10.1002/ggge.20265. Baczynski, A.A., McInerney, F.A., Wing, S.L., Kraus, M.J., Morse, P.E., Bloch, J.I., Chung, A.H., and Freeman, K.H., 2016, Distortion of carbon isotope excursion in bulk soil organic matter during the Paleocene-Eocene thermal maximum: Geological Society of America Bulletin, , no. Xx, p. B31389.1, doi: 10.1130/B31389.1. Bataille, C.P., Mastalerz, M., Tipple, B.J., and Bowen, G.J., 2013, Influence of provenance and preservation on the carbon isotope variations of dispersed organic matter in ancient floodplain sediments: Geochemistry, Geophysics, Geosystems, v. 14, no. 11, p. 4874– 4891, doi: 10.1002/ggge.20294. Denis, E.H., Foreman, B.Z., Maibauer, B.J., Bowen, G.J., Baczynski, A.A., McInerney, F.A., Collinson, M.E., Belcher, C.M., Wing, S.L., and Freeman, K.H., in prep, Decreased soil carbon in a warming world: Degraded pyrogenic carbon during the Paleocene-Eocene Thermal Maximum (PETM): Geology. Diefendorf, A.F., Freeman, K.H., and Wing, S.L., 2014, A comparison of terpenoid and leaf fossil vegetation proxies in Paleocene and Eocene Bighorn Basin sediments: Organic Geochemistry, v. 71, p. 30–42, doi: 10.1016/j.orggeochem.2014.04.004. Hajek, E.A., Heller, P.L., and Sheets, B.A., 2010, Significance of channel-belt clustering in alluvial basins: Geology, v. 38, no. 6, p. 535–538, doi: 10.1130/G30783.1. Jerolmack, D.J., and Paola, C., 2010, Shredding of environmental signals by sediment transport: Geophysical Research Letters, v. 37, no. 19, p. 1–5, doi: 10.1029/2010GL044638. Kraus, M.J., and Riggins, S., 2007, Transient drying during the Paleocene-Eocene Thermal Maximum (PETM): Analysis of paleosols in the bighorn basin, Wyoming: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 245, no. 3-4, p. 444–461, doi: 10.1016/j.palaeo.2006.09.011. McInerney, F.A., and Wing, S.L., 2011, The Paleocene-Eocene Thermal Maximum: A Perturbation of Carbon Cycle, Climate, and Biosphere with Implications for the Future: Annual Review of Earth and Planetary Sciences, v. 39, no. 1, p. 489–516, doi: 10.1146/annurev-earth-040610-133431.

124 Moore, E.A., and Kurtz, A.C., 2008, Black carbon in Paleocene-Eocene boundary sediments: A test of biomass combustion as the PETM trigger: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 267, no. 1-2, p. 147–152, doi: 10.1016/j.palaeo.2008.06.010. Schouten, S., Woltering, M., Rijpstra, W.I.C., Sluijs, A., Brinkhuis, H., and Sinninghe Damsté, J.S., 2007, The Paleocene-Eocene carbon isotope excursion in higher plant organic matter: Differential fractionation of angiosperms and conifers in the Arctic: Earth and Planetary Science Letters, v. 258, no. 3-4, p. 581–592, doi: 10.1016/j.epsl.2007.04.024. Smith, J.J., Hasiotis, S.T., Kraus, M.J., and Woody, D.T., 2008, Relationship of floodplain ichnocoenoses to paleopedology, paleohydrology, and paleoclimate in the Willwood Formation, Wyoming, during the Paleocene-Eocene Thermal Maximum: Palaios, v. 23, no. 9-10, p. 683–699, doi: 10.2110/palo.2007.p07-080r. Wang, Y., Straub, K.M., and Hajek, E.A., 2011, Scale-dependent compensational stacking: An estimate of autogenic time scales in channelized sedimentary deposits: Geology, v. 39, no. 9, p. 811–814, doi: 10.1130/G32068.1. Wing, S.L., Harrington, G.J., Smith, F.A., Bloch, J.I., Boyer, D.M., and Freeman, K.H., 2005, Transient Floral Change and Rapid Global Warming at the Paleocene-Eocene Boundary: Science, v. 310, no. 5750, p. 993–996, doi: 10.1126/science.1116913.

Appendix A

Chapter 2 Supplemental Material

Study Section

The BBCP Basin Substation site (BSN, Bighorn Basin, Wyoming) comprises two overlapping holes ~138.5 meters below surface (mbs) (BBCP-BSN11-1A&B; N 44.4162583 W

108.1055664; N 44.4162042 W 108.1053154) drilled 21 m apart. The BBCP Polecat Bench site

(PCB, Bighorn Basin, Wyoming) comprises two holes to 130.0 mbs (BBCP-PCB11-2A; N

44.7688571 W 108.8879668) and 245.1 mbs (BBCP-PCB11-2B; N 44.7688782 W 108.8878902) drilled 6 m apart. Both sites recover late Paleocene-Early Eocene strata deposited in a floodplain and paleosol setting (Clyde et al., 2013).

Basin Substation PETM Stratigraphy

The PETM is marked a negative carbon isotope excursion (CIE) (McInerney and Wing,

2011). At Basin Substation, there was low organic carbon and the bulk organic carbon isotope

13 values (δ Corg) were variable and lacked a discernible CIE (Maibauer, 2013). Most core samples

13 contained insufficient n-alkane concentrations for carbon isotope analyses (δ Cn-alk) and all

13 measurable samples had non-excursion δ Cn-alk values (Baczynski, 2014; Baczynski et al.,

13 2014b). To narrow the PETM window between the δ Cn-alk samples with non-CIE values (42 to

95 mcd), an approximate position of the PETM was determined based on litho- and biostratigraphy using distinctive marker beds and palynological correlations from the core to surface

126 outcrops (~95 to ~50 mcd) (Harrington et al., 2014; Baczynski et al., 2014b). In addition, an increase in the average chain length of n-alkanes (~28 to ~30) coincides with the PETM interval

(~52 to ~94 mcd), which has also been observed in surface outcrop samples from the southeastern part of the Bighorn Basin (Baczynski, 2014; Baczynski et al., 2014; Baczynski et al., 2016).

Therefore, for the purposes of this dissertation, results were reported and interpreted using 52 mcd as the upper bound and 94 mcd as the lower bound of the PETM.

Methods

Rock samples were collected from cores BSN-1B and PCB-2A during coring in the field and from these cores and core BSN-1A during core processing in the Bremen Core Repository,

Bremen, Germany. Sampling protocols were similar to those developed for International Ocean

Drilling Projects (IODP) projects (Clyde et al., 2013).

Lipids were extracted at the Royal Netherlands Institute for Sea Research (NIOZ). For lipid analyses from BSN-1B and PCB-2A cores, rock samples were freeze-dried and homogenized with a mortar and pestle or disk mill. The exterior of rock samples were rinsed with dichloromethane (DCM), freeze-dried, and ground to a powder with a mortar and pestle or disk mill. Lipids were extracted from ~40 to 50 g of powdered rock using an Accelerated Solvent

Extractor (ASE) with DCM:methanol (MeOH) (v:v, 9:1) at 100°C and 1500 psi. The total lipid extract (TLE) was gently concentrated with nitrogen gas, and the TLE was split into thirds to allow BBCP collaborators to work on identical samples. One-third of the TLE was sent to the

Pennsylvania State University, and there, prepared, and analyzed for PAH data in this manuscript.

Additional rock samples that were not extracted at NIOZ were processed similarly to the protocol described above at the Pennsylvania State University. Total organic carbon (TOC) samples were

127 prepared and analyzed at the University of Utah. Charcoal work was done at Royal Holloway

University of London.

At the Pennsylvania State University, TLEs were separated into compound class fractions by a semi-automated chromatographic method using the ASE (Magill et al., 2015). ASE 33 ml cell columns were constructed with silver-impregnated alumina (2 g), silica gel (5.5 g), a filter loaded with reconstituted TLE in DCM, silica gel (1 g), and filled with sand. ASE columns were selectively extracted following Magill et al. (2015): saturated hydrocarbons - 100% n-hexane,

30% flush volume (~10 ml), and back-eluted for, unsaturated/aromatics - hexane:DCM (v:v, 9:1) and polar - DCM:MeOH (v:v, 7:3) for three aliquots of solvent at 60% flush volume (~20 ml) each. PAHs were analyzed from the TLE and F2 fraction using an Agilent 6890 GC with an

Agilent 5973 quadrupole mass spectrometer (MS) and a fused silica capillary column (Agilent

J&W DB-5; 30 m, 250 μm, 0.25 μm). The column flow rate was 2.0 ml/min and the oven program started at 60°C for 1 min, ramped to 320°C at 6°C/min, and a final hold time of 15 min.

(SIM) mode. The following ions were monitored: (m/z) 102, 128; 151, 152, 153, 154; 165, 166,

183, 198; 176, 178; 101, 202, 219, 234; 114, 126, 228, 253, 254; 125, 126, 252; 276, 278; 150,

300. The detection limit was 10 pg of PAH. Compound peak areas were converted to mass quantities using response curves for 19 PAH external standards analyzed in concentrations ranging from 0.01 to 1 ng/μl. Uncertainty in measurements was determined by treating additional analyses of PAH standards as unknowns and calculating the coefficient of variation (CV) of the concentrations, where CV is the standard deviation divided by the mean multiplied by 100%. The

CV for each PAH standard ranged from 6-16% with an average CV of 12% of the mean

128 concentration. PAH abundances were normalized to total organic carbon for each sample (μg/g

TOC).

At the University of Utah, for TOC analyses from BSN-1B and PCB-2A cores, ~2 g aliquots of powdered sample were decarbonated through acidification with 2N HCl (Midwood and Boutton, 1998). TOC was analyzed on a ThermoFinnigan Elemental Analyzer (EA) coupled to a Delta Plus Advantage Continuous Flow Isotope Ratio Mass Spectrometer (IRMS).

Repeatability of analyses averaged 0.01%TOC based on replicate analyses of twenty-two samples.

At Royal Holloway University of London, for charcoal from the BSN-1A core, 10 grams of uncrushed rock were macerated in acid to dissolve carbonate and silicate materials respectively

(Belcher et al., 2005). Samples were treated with 37% HCl followed by 40% HF and 37% HCl, and rinsed in water until a neutral pH was achieved. The samples were sieved at 125 μm and 500

μm, and the 125-500 μm fraction was examined under a stereo microscope. All charred particles that occurred in the samples were counted. Charcoal particles were identified as black with a graphite-like shine, and coaly particles were identified as dark brown or black without the graphite-like shine. Select samples were confirmed by scanning electron microscope (SEM).

PETM Sedimentation Rate Estimate

In the Bighorn Basin, long-term average sediment accumulation rates were 0.25 – 0.4 mkyr-1 through the Paleocene and Eocene (Clyde et al., 2013; Bowen et al., 2015). At Basin

Substation (BSN) and Polecat Bench (PCB), lithology, which varied between clay, silt, and sand, was not substantially different among the late Paleocene, PETM, or early Eocene sample intervals. Given a range for the duration of the PETM as 170±50 kyr (Farley and Eltgroth, 2003;

Röhl et al., 2007; Murphy et al., 2010) and PETM interval thickness of 42 m in the BSN core and

129 of 45 m in the PCB core, we estimate sedimentation rates were on average ~0.19 – 0.38 mkyr-1 at either site. Therefore, sedimentation rates on average increased no more than 1.5 times during the

PETM interval. Hence, sedimentary dilution cannot explain the decreased organic carbon in

PETM samples.

Figures

Figure A-1. Bivariate plot of PAH solubility in water and temperature modified from and using data from May et al., (1978). Symbols represent number of rings (2, circle; 3, triangle; 4, square) and, within each ring-number, shades represent relative solubility on a grey scale, where white is most soluble and black is least soluble.

130

Figure A-2. Box and whisker plots of TOC data from Polecat Bench and Basin Substation cores and compiled TOC data from terrestrial PETM sites published in literature. Ends of box represent 25% and 75% quartile; middle line is median; whiskers extend to minimum and maximum values (not including outliers); white squares represent outliers. Polecat Bench, Wyoming – outcrop (Magioncalda et al., 2004); Piceance Creek Basin, Colorado (Foreman et al., 2012); Claret, Spain (Domingo et al., 2009). Number in italics is n samples for each box. Light grey box delineates the PETM; Paleocene is pre-PETM and Eocene is post-PETM samples. Logarithmic-scale axis.

131

2500

2000

1500

1000

500 Total PAH (ng/gPAHTotal dry sediment) 0 0 1 2 3 TOC (%)

2500

2000

1500

g/g TOC) μ

1000

Total PAH ( PAHTotal 500

0 0 1 2 3 TOC (%)

Figure A-3. Cross plot of PAH concentrations and TOC from Basin Substation.

132

Figure A-4. Cross plots of coronene concentrations and TOC from Basin Substation and Polecat Bench cores in terms of lithology. Symbols represent clay (dark-grey diamond), silt (light-grey square), and sand (white triangle).

133

Figure A-5. Depth profile of the ratio of 6-ring PAH (coronene) to 4-ring PAH (pyrene) from Basin Substation and Polecat Bench. Light grey box marks the PETM. Molecular structures of coronene and pyrene are shown.

134 Tables

Table A-1. Basin Substation percent loss calculations and allochthonous (fossil) carbon estimate. Median values reported PAH %TOC g of C/g rock ng/g rock PETM 0.09 0.0009 24.1 Paleocene 0.23 0.0023 231.9 %Loss 60 90 PETM loss %Paleocene 40 10 Logic If 1:1 ratio of PAH %Loss to TOC %Loss, then expect TOC loss of 90% during PETM. Only 60% of TOC loss because fossil carbon (more refractory than PAH) proportion didn't degrade. If there had been 90% loss of TOC during PETM, how many grams of TOC would remain? g of C/g rock Expected PETM C 0.00024 Fossil Carbon 0.00066

%Fossil Carbon PETM 74 Paleocene 29 % Increase C 45

135

Table A-2. Polecat Bench percent loss calculations and allochthonous (fossil) carbon estimate. Median values reported PAH %TOC g of C/ g rock ng/g rock PETM 0.07 0.0007 1.4 Paleocene 0.12 0.0012 4.3 %Loss 44 68 PETMloss%Paleocene 56 32 Logic If 1:1 ratio of PAH %Loss to TOC %Loss, then expect TOC loss of 68%. Only 44% of TOC loss because fossil carbon (more refractory than PAH) proportion didn't degrade. If there had been 68% loss of TOC during PETM, how many grams of TOC would remain? g of C/g rock Expected PETM C 0.00040 Fossil Carbon 0.00030

%Fossil C PETM 43 Paleocene 24 % Increase C 19

136 Table A-3. Grams of sediment extracted and PAH abundanes (μg/g TOC) for Basin Substation

samples.

)

BSN11

sed

Sample ID Sample (BBCPID depth (mcd) g dry Napthalene Acenaphtylene Acenaphthene Fluorene Cadalene Phenathrene Anthracene Fluoranthene Pyrene 4CC 6.90 9.999 0.008 0.011 0.010 0.029 0.141 0.702 0.047 0.44 0.16 5QCC 8.00 10.009 n/a n/a n/a 0.023 n/a 0.434 0.087 0.11 0.15 13CC 17.32 9.986 0.045 0.118 0.282 0.152 28.72 84.59 0.312 13.30 18.17 14QCC 18.88 9.981 0.020 0.034 0.003 0.005 0.221 2.65 0.029 1.76 2.72 23QCC 33.15 10.419 n/a n/a n/a n/a n/a 2.51 n/a 2.03 3.73 27CC 39.36 10.011 0.013 0.005 0.004 0.014 4.27 1.66 0.033 2.35 3.47 30CC 44.28 10.013 0.009 0.001 0.002 0.003 0.200 0.307 0.010 0.78 1.27 39Q1 55.58 9.996 0.034 0.115 0.009 0.012 0.523 0.226 0.125 0.42 1.11 44CC 62.37 10.021 0.008 n/a n/a 0.004 0.002 0.016 0.006 0.058 0.100 56-2 80.04 10.023 0.047 0.039 0.004 0.011 0.057 0.144 0.048 0.25 0.56 59CC 84.53 13.289 0.012 0.009 0.001 0.005 0.029 0.034 0.008 0.16 0.44 60Q2 86.13 10.016 0.003 0.014 0.000 0.001 n/a 0.073 0.010 0.11 0.31 62CC 89.14 10.029 0.011 0.007 n/a 0.007 0.019 0.257 0.014 0.35 1.08 65Q2 93.93 10.080 0.009 0.008 0.002 0.007 0.019 0.255 0.016 0.21 0.80 68CC 98.84 10.038 0.016 0.017 0.041 0.088 n/a 1.03 0.039 0.62 2.28 71CC 103.45 10.075 0.052 0.134 1.74 4.20 n/a 10.79 0.692 10.03 31.96 75CC 109.99 12.895 0.027 0.019 0.036 0.313 19.04 3.00 0.135 2.31 5.59 89Q1 129.83 10.045 0.020 0.010 0.007 0.020 6.655 2.88 0.060 1.42 3.55 90Q1 130.41 9.266 n/a n/a n/a 0.429 n/a 5.79 n/a 1.06 3.41 99QCC 144.46 9.682 0.124 n/a n/a 0.042 n/a 0.244 n/a 0.057 0.17 100Q1 145.85 10.040 0.026 0.044 0.296 1.490 63.02 11.07 0.198 2.48 6.19

137

Table A-4. Additional PAH abundanes (μg/g TOC) for Basin Substation samples.

)

BSN11

]pyrene

]anthracene

]perylene

]pyrene

]anthracene e a ghi

Sample ID Sample (BBCPID depth (mcd) Retene Benz[ Chrysene Benzofluoranthenes Benzo[ Benzo[ Perylene Indeno[1,2,3 Dibenz[ Benzo[ Coronene 4CC 6.90 0.18 0.02 0.64 1.21 0.43 0.12 0.04 0.25 0.09 0.19 0.34 5CC 8.00 n/a 0.027 0.25 0.40 0.15 0.011 0.020 0.051 0.012 0.059 0.02 13CC 17.32 9.55 4.47 9.44 23.75 7.56 3.38 21.57 2.86 1.01 12.80 0.91 14QCC 18.88 1.12 0.30 1.30 2.88 1.57 0.50 11.95 1.01 0.27 4.99 3.15 23QCC 33.15 n/a 0.59 1.79 4.19 2.15 0.62 19.51 1.82 0.12 16.21 6.01 27CC 39.36 8.05 1.16 2.87 10.14 2.16 0.76 10.53 2.81 0.74 15.57 3.73 30CC 44.28 0.81 0.13 0.49 1.17 0.82 0.28 1.00 0.66 0.18 2.91 1.29 39Q1 55.58 0.13 0.30 1.11 0.94 1.67 1.35 23.08 1.59 1.71 7.23 2.13 44CC 62.37 0.003 0.022 0.127 0.167 0.188 0.052 0.048 0.176 0.094 1.27 0.72 56-2 80.04 0.10 0.10 0.47 0.88 2.12 0.35 1.64 2.31 0.74 11.17 1.98 59CC 84.53 0.013 0.08 0.37 2.01 0.72 0.13 0.19 1.64 0.30 6.83 21.35 60Q2 86.13 0.010 0.05 0.16 0.18 0.33 0.09 0.18 0.33 0.09 1.74 5.20 62CC 89.14 0.009 0.22 0.51 0.87 0.82 0.44 0.90 0.53 0.30 1.43 0.97 65Q2 93.93 0.023 0.37 0.79 2.00 2.50 0.95 3.00 1.67 0.64 10.34 6.38 68CC 98.84 2.70 0.55 0.77 2.48 1.90 1.00 5.01 2.32 2.18 11.58 4.38 71CC 103.45 31.72 15.43 9.15 26.63 41.72 59.31 1816 47.49 12.25 82.04 32.46 75CC 109.99 3.64 4.07 2.65 16.28 3.98 4.54 57.06 10.17 3.95 41.19 19.42 89Q1 129.83 22.61 1.60 2.76 35.15 7.59 5.43 55.15 19.12 5.38 77.93 26.06 90Q1 130.41 n/a 0.31 0.78 0.66 0.25 0.14 5.44 0.23 0.093 0.83 3.01 99CC 144.46 n/a 0.024 0.042 0.070 0.056 0.084 0.788 0.017 0.002 0.082 2.99 100Q1 145.85 8.03 2.06 1.30 5.28 6.73 6.41 359 4.51 0.86 9.58 3.39

138

Table A-5. Basin Substation samples' lithology, grain size, and color from BBCP Science Team.

)

BSN11

Sample ID (BBCP ID Sample (mcd) Depth Lithology Grainsize Color 4CC 6.90 mudstone clay dark yellow brown 5QCC 8.00 mudstone clay yellow brown 13CC 17.32 laminated mudstone clay grey brown 14QCC 18.88 mudstone clay grey green 23QCC 33.15 massive mudstone clay dark grey 27CC 39.36 massive mudstone clay silt gray brown 30CC 44.28 massive mudstone clay gray brown 39Q1 55.58 sandstone very fine sand dark grey 44CC 62.37 mud-siltstone clay silt dark grey green 56-2 80.04 massive mudstone clay silt dark grey 59CC 84.53 massive mudstone clay dark grey 60Q2 86.13 massive mudstone clay silt dark grey 62CC 89.14 massive mudstone clay dark grey 65Q2 93.93 massive mudstone clay dark grey 68CC 98.84 massive mudstone clay grey brown 71CC 103.45 massive mudstone clay silt dark grey brown 75CC 109.99 siltstone silt dark grey 89Q1 129.83 siltstone silt clay dark grey 90Q1 130.41 sandstone very fine sand dark grey 99QCC 144.46 siltstone silt clay dark grey 100Q1 145.85 massive mudstone clay silt dark grey

139 Table A-6. δ13Corg, TOC, PAH values and dry weight of grams extracted for Polecat Bench

samples.

PCB11

g/g TOC) g/g

(‰,VDPB)

org

Sample ID, BBCP ID, Sample (mcd) Depth δ (%) TOC (g) weight dry ( Pyrene ( Coronene UnitType 15Q-2 9.72 -26.75 0.31 8.320 0.001 0.23 mudstone 17Q-CC 12.41 -26.20 0.18 7.421 0.011 0.61 siltstone 19Q-CC 15.86 -23.81 0.07 4.301 0.026 2.50 sandstone 22Q-CC 20.01 -24.77 0.07 4.422 0.036 3.86 siltstone 25Q-CC 23.94 -23.59 0.12 4.906 0.066 2.62 sandstone 27Q-CC 26.67 -23.42 0.11 2.273 0.006 0.48 siltstone 30Q-CC 31.58 -24.47 0.11 4.431 0.595 17.7 mudstone 33Q-2 36.55 -24.78 0.08 1.950 0.299 4.45 sandstone 35Q-1 40.04 -26.83 1.04 1.959 0.138 0.94 mudstone 37Q-CC 43.05 -26.41 0.43 4.351 0.042 3.74 siltstone 41Q-1 49.11 -24.29 0.16 4.371 0.386 11.0 siltstone 44Q-CC 53.57 -24.39 0.07 3.421 0.028 3.14 sandstone 46Q-CC 56.92 -24.38 0.14 0.785 0.118 5.32 mudstone 47Q-CC 58.52 -24.54 0.13 2.183 0.140 1.69 siltstone 51Q-2 64.81 -24.73 0.11 1.332 0.095 3.09 siltstone 54Q-CC 69.38 -25.10 0.11 2.173 0.290 1.71 siltstone 63Q-2 clay 83.50 -27.04 0.37 1.349 0.005 4.63 conglomerate 63Q-2 sand 83.50 -27.04 0.37 4.171 0.005 0.77 conglomerate 66Q-2 88.02 -26.68 0.21 3.636 0.026 0.18 siltstone 67Q-1 89.93 -25.27 0.06 4.437 0.140 0.88 mudstone 68Q-CC 91.42 -25.54 0.05 1.417 0.366 1.32 mudstone 72Q-1 97.83 -27.15 0.66 3.356 0.013 0.06 siltstone 74Q-2 100.99 -26.70 0.07 0.716 0.290 3.74 siltstone 76Q-2 104.13 -26.82 0.27 2.334 0.020 0.03 siltstone 82Q-2 113.89 -25.47 0.12 2.383 0.031 1.70 sandstone 84Q-2 116.94 -27.36 0.40 1.842 0.217 1.18 mudstone 86Q-2 120.19 -26.15 0.06 3.444 0.149 0.78 sandstone 90Q-2 126.78 -24.98 0.12 1.891 0.179 1.99 siltstone 92Q-2 130.09 -26.26 0.28 5.287 0.078 0.54 siltstone

140

PCB11

g/g TOC) g/g

(‰,VDPB)

org

Sample ID, BBCP ID, Sample (mcd) Depth δ (%) TOC (g) weight dry ( Pyrene ( Coronene UnitType 94Q-CC 133.17 -25.89 0.10 3.055 0.086 6.23 sandstone 96Q-CC 136.43 -24.45 0.11 4.122 0.915 35.5 sandstone 98Q-1 138.61 -24.19 1.28 4.638 1.39 5.82 mudstone δ13C and TOC values from Maibauer (2013); Unit type and color from BBCP Science Team.Grey shading marks PETM interval.

References

Baczynski, A.A., 2014, Evaluating Carbon Cycle Dynamics and Hydrologic Change during the Paleocene-Eocene Thermal Maximum, Bighorn Basin, Wyoming: Northwestern University, 219 p. Baczynski, A.A., McInerney, F.A., Wing, S.L., Kraus, M.J., Morse, P.E., Bloch, J.I., Chung, A.H., and Freeman, K.H., 2016, Distortion of carbon isotope excursion in bulk soil organic matter during the Paleocene-Eocene thermal maximum: Geological Society of America Bulletin, , no. Xx, p. B31389.1, doi: 10.1130/B31389.1. Baczynski, A.A., McInerney, F.A., Wing, S.L., and the BBCP Science Team, 2014, n-Alkane PETM records from the Bighorn Basin, Wyoming: A core-outcrop comparison: Rendiconti Online della Società Geologica Italiana, v. 31, p. 19–20, doi: 10.3301/ROL.2014.24. Belcher, C.M., Collinson, M.E., and Scott, A.C., 2005, Constraints on the thermal energy released from the Chicxulub impactor; new evidence from multi-method charcoal analysis: Journal of the Geological Society of London, v. 162, no. 4, p. 591–602, doi: 10.1144/0016-764904-104. Bowen, G.J., Maibauer, B.J., Kraus, M.J., Röhl, U., Westerhold, T., Steimke, A., Gingerich, P.D., Wing, S.L., and Clyde, W.C., 2015, Two massive, rapid releases of carbon during the onset of the Palaeocene-Eocene thermal maximum: Nature Geosci, v. 8, no. 1, p. 44–47, doi: 10.1038/ngeo2316.

141 Clyde, W.C., Gingerich, P.D., Wing, S.L., Röhl, U., Westerhold, T., Bowen, G., Johnson, K., Baczynski, A.A., Diefendorf, A., McInerney, F., Schnurrenberger, D., Noren, A., Brady, K., Acks, R., et al., 2013, Bighorn Basin Coring Project (BBCP): A continental perspective on early Paleogene hyperthermals: Scientific Drilling, , no. 16, p. 21–31, doi: 10.5194/sd-16-21-2013. Farley, K.A., and Eltgroth, S.F., 2003, An alternative age model for the Paleocene-Eocene thermal maximum using extraterrestrial 3He: Earth and Planetary Science Letters, v. 208, no. 3-4, p. 135–148, doi: 10.1016/S0012-821X(03)00017-7. Harrington, G.J., Jardine, P.E., Wing, S.L., and the BBCP Science Team, 2014, Bighorn Basin Coring Project (BBCP): Pollen floral changes and organic matter from core – outcrop comparisons through the PETM: Rendiconti Online della Società Geologica Italiana, v. 31, p. 95–96. Magill, C.R., Denis, E.H., and Freeman, K.H., 2015, Rapid sequential separation of sedimentary lipid biomarkers via selective accelerated solvent extraction: Organic Geochemistry, v. 88, p. 29–34, doi: 10.1016/j.orggeochem.2015.07.009. Maibauer, B.J., 2013, Carbon Isotope Stratigraphy of Early Eocene Hyperthermals in the Bighorn Basin, Wyoming, USA: Analogues for Modern Anthropogenic Carbon Emissions: The University of Utah, 122 p. McInerney, F.A., and Wing, S.L., 2011, The Paleocene-Eocene Thermal Maximum: A Perturbation of Carbon Cycle, Climate, and Biosphere with Implications for the Future: Annual Review of Earth and Planetary Sciences, v. 39, no. 1, p. 489–516, doi: 10.1146/annurev-earth-040610-133431. Midwood, A.J., and Boutton, T.W., 1998, Soil carbonate decomposition by acid has little effect on δ13C of organic matter: Soil Biology and Biochemistry, v. 30, no. 10-11, p. 1301– 1307, doi: 10.1016/S0038-0717(98)00030-3. Murphy, B.H., Farley, K.A., and Zachos, J.C., 2010, An extraterrestrial 3He-based timescale for the Paleocene-Eocene thermal maximum (PETM) from Walvis Ridge, IODP Site 1266: Geochimica et Cosmochimica Acta, v. 74, no. 17, p. 5098–5108, doi: 10.1016/j.gca.2010.03.039. Röhl, U., Westerhold, T., Bralower, T.J., and Zachos, J.C., 2007, On the duration of the Paleocene-Eocene thermal maximum (PETM): Geochemistry, Geophysics, Geosystems, v. 8, no. 12, doi: 10.1029/2007GC001784.

142

Appendix B

Chapter 3 Supplemental Material

Supplemental Text

Sediments within the Wasatch formation are geologically immature (Johnson and Rice,

1990; Johnson, 1992), which means the organic matter was thermally unaltered. Preliminary leaf- wax n-alkane results corroborate that the Paleocene-Eocene samples in this section are from immature sediments. n-Alkane results from a few late Paleocene and early Eocene samples indicate that high molecular weight n-alkanes with a strong odd-over-even carbon preference dominated (Baczynski et al., unpublished results), which is characteristic of vascular plants and immature sediments (Bray and Evans, 1961; Peters et al., 2005; Bush and McInerney, 2013). The carbon preference index (CPI) was greater than 2 and the average chain length (ACL) ranged from ~27 to 28 (Baczynski et al., unpublished results), which are both consistent with non-PETM values in the Bighorn Basin, Wyoming (Baczynski et al., 2016). These samples contained only trace amounts of n-alkanes and were not sufficiently abundant for carbon isotope analyses.

143 Figures

Figure B-1. Depth profile of the ratio of 6-ring PAH (coronene) to 4-ring PAH (pyrene) from Basin Substation, Polecat Bench, and Piceance Basin. Light grey box marks the PETM.

Supplemental Table

144 Table B-1. Fossil (allochthonous) carbon calculations: Piceance Basin, median values reported. PAH TOC TOC (ng/g PAH count %TOC (g C/g rock) rock) count 63 PETM 0.005 0.0005 5.3 38 40 Paleocene 0.17 0.0017 53.7 18 % loss 68 90 PETM (%Paleocene) 32 10

Logic If 1:1 ratio of PAH %Loss to TOC %Loss, then expect TOC loss of 90%. Only 68% loss because fossil carbon (more refractory than PAH) didn't degrade. If 90% loss of TOC, how many grams of TOC would remain:

g C/g rock Expected PETM C 0.00017 Fossil Carbon 0.00038

%Allochthonous Carbon PETM 69 Paleocene 22 % Increase C 47

145

Data Tables

Table B-2. δ13Corg and TOC values and dry weight of grams extracted for Piceance Basin samples. Strat. height δ13C (‰, TOC Dry Sample ID (m) VPDB) (%) weight (g) PCBa-001 -6.00 -24.8 0.01 5.966 PCBa-002 -7.50 -22.0 0.04 3.995 PCBa-006 -15.00 -22.6 0.06 4.620 PCBa-ED43 -18.19 -23.95 0.05 12.620 PCBa-009 -20.00 -22.0 0.07 5.038 PCBa-ED42 -20.1 -23.63 0.06 12.500 PCBa-ED41 -21.91 -24.07 0.08 12.513 PCBa-ED40 -26.7 -23.95 0.04 12.596 PCBa-010 -28.70 -22.5 0.04 4.459 PCBa-011 -32.50 -23.0 0.04 2.733 PCBa-ED39 -33.5 -24.28 0.03 12.155 PCBa-014 -41.20 -25.5 0.02 3.386 PCBa-ED38 -43.4 -25.37 0.02 12.634 PCBa-ED37 -43.9 -26.60 0.03 12.030 PCBa-ED36 -45.08 -24.83 0.03 12.764 PCBa-ED35 -48.98 -25.96 0.05 12.195 PCBa-ED34 -50.5 -28.09 0.16 12.488 PCBa-ED33 -50.8 -28.59 0.12 12.463 PCBa-020 -51.40 -27.2 0.12 3.500 PCBa-ED32 -51.75 -28.40 0.09 12.262 PCBa-022 -58.30 -28.3 0.56 3.529 PCBa-ED31 -60 -24.59 0.06 12.273 PCBa-ED30 -61.8 -24.66 0.07 12.322 PCBa-ED29 -61.95 -24.39 0.05 12.546 PCBa-ED28 -62.25 -24.47 0.03 12.254

146

Strat. height δ13C (‰, TOC Dry Sample ID (m) VPDB) (%) weight (g) PCBa-031 -62.55 -24.2 0.06 4.655 PCBa-ED27 -63.15 -23.11 0.04 12.475 PCBa-ED26 -64.1 -24.91 0.05 12.168 PCBa-ED25 -64.12 -24.40 0.05 12.014 PCBa-ED24 -65.25 -24.73 0.08 12.685 PCBa-ED23 -65.7 -24.47 0.08 12.460 PCBa-ED22 -65.75 -24.66 0.08 12.602 PCBa-035 -65.87 -23.9 0.07 4.618 PCBa-ED21 -66 -26.55 0.09 12.777 PCBa-036 -66.00 -28.9 0.65 3.313 PCBa-ED20 -66.1 -28.46 0.43 12.867 PCBa-ED19 -67.03 -26.05 0.23 12.289 PCBa-039 -67.20 -25.0 0.19 3.141 PCBa-ED18 -67.57 -26.54 0.08 12.385 PCBa-041 -68.30 -24.1 0.04 4.328 PCBa-ED17 -68.65 -24.19 0.15 12.504 PCBa-ED16 -69.93 -24.28 0.04 12.439 PCBa-ED15 -70.23 -25.97 0.08 12.413 PCBa-ED14 -70.75 -24.13 0.10 12.150 PCBa-ED13 -71.4 -25.38 0.07 12.400 PCBa-ED12 -72 -24.35 0.06 12.395 PCBa-047 -73.20 -23.8 0.05 4.356 PCBa-ED11 -73.5 -23.65 0.04 12.425 PCBa-ED10 -73.85 -23.86 0.14 12.156 PCBa-ED9 -74.35 -24.10 0.19 12.279 PCBa-050 -74.75 -22.6 0.12 4.050 PCBa-052 -76.10 -22.9 0.21 6.012 PCBa-ED8 -78.93 -24.58 0.17 12.704 PCBa-ED7 -80.56 -23.53 0.14 12.405 PCBa-059 -82.45 -22.5 0.24 5.813 PCBa-ED6 -83.25 -22.32 0.24 12.595 PCBa-ED5 -88.79 -24.58 0.47 12.440 PCBa-ED4.1 -91.8 -23.67 0.27 12.365 PCBa-ED4 -92.1 -24.24 0.34 12.436 PCBa-072 -92.20 -22.7 0.06 5.872 PCBa-ED3 -95.05 -23.50 0.17 12.134 PCBa-076 -96.25 -22.7 0.25 7.044 PCBa-ED2 -97.62 -22.40 0.27 12.351

147

Strat. height δ13C (‰, TOC Dry Sample ID (m) VPDB) (%) weight (g) PCBa-ED1 -99.05 -24.05 0.26 12.468 PCBa-097 -111.55 -22.4 0.18 6.653 PCBa-0xx δ13C and TOC values from Foreman et al., (2012). Shading represents PETM interval.

Table B-3. PAH concentrations (ng/g TOC) for all Piceance Basin samples

cd]pyrene

Sample ID Sample Phenanthrene Fluoranthene Pyrene Retene Benz[a]anthracene Chrysene Benzofluranthene Benzopyrene Indeno[1,2,3 Dibenz[a,h]anthracene Benzo[g,h,i]perylene Coronene PCBa-001 124 383 170 599 n/a 296 203 161 25 12 16 908 PCBa-002 52 170 117 242 16 941 740 1103 121 36 19 3260 PCBa-006 56 327 180 374 24 209 1971 691 777 158 85 28090 PCBa-ED43 2 27 87 35 4 77 75 154 14 9 10 953 PCBa-009 264 313 139 230 11 497 1096 1093 302 78 31 11055 PCBa-ED42 1 11 22 17 2 128 390 498 114 50 42 5752 PCBa-ED41 5 38 90 20 4 263 413 749 176 69 97 6355 PCBa-ED40 2 35 118 25 2 466 179 488 19 16 17 1620 PCBa-010 166 157 91 337 20 104 112 67 25 14 11 338 PCBa-011 154 845 487 924 146 1218 1365 1081 421 272 47 2312 PCBa-ED39 33 62 182 50 7 57 17 42 n/a n/a n/a 72 PCBa-014 132 338 188 2105 n/a 316 124 84 n/a n/a 18 176 PCBa-ED38 n/a 21 41 30 n/a 170 22 65 n/a n/a n/a 64 PCBa-ED37 9 33 123 29 n/a 135 11 58 n/a n/a n/a 146 PCBa-ED36 n/a 41 101 75 n/a 44 14 37 7 3 8 202 PCBa-ED35 7 79 120 71 13 480 484 659 68 41 25 3086 PCBa-ED34 33 244 93 5 85 4830 5253 6801 1930 774 842 15780 PCBa-ED33 3 49 72 6 3 885 1295 966 160 85 33 3430 PCBa-020 46 237 34 166 34 7467 22578 8884 2313 900 387 36115 PCBa-ED32 2 16 32 20 14 134 635 943 469 189 227 10060 PCBa-022 804 2856 705 145 262 3528 1694 2062 386 250 348 2435 PCBa-ED31 51 490 217 27 81 1029 866 986 306 177 50 2685 PCBa-ED30 157 758 449 41 195 1411 2272 2779 1027 366 453 8269

148

cd]pyrene

Sample ID Sample Phenanthrene Fluoranthene Pyrene Retene Benz[a]anthracene Chrysene Benzofluranthene Benzopyrene Indeno[1,2,3 Dibenz[a,h]anthracene Benzo[g,h,i]perylene Coronene PCBa-ED29 16 413 301 58 103 1053 1265 1359 147 134 40 3381 PCBa-ED28 10 61 54 27 4 120 67 107 4 8 4 188 PCBa-031 114 504 132 125 49 1316 2560 1536 287 162 28 3985 PCBa-ED27 36 252 289 68 14 465 303 464 20 26 12 1093 PCBa-ED26 24 946 699 113 110 2046 1177 1841 91 127 26 2048 PCBa-ED25 15 558 198 20 55 1355 716 1333 231 39 95 1336 PCBa-ED24 152 1088 329 33 76 1115 741 1010 84 93 16 1515 PCBa-ED23 324 2007 694 40 367 2379 2166 3125 628 358 118 5270 PCBa-ED22 177 1168 394 37 179 1370 1202 1691 399 209 130 3999 PCBa-035 134 334 95 354 23 975 1642 806 129 54 10 2758 PCBa-ED21 22 999 415 38 149 1474 1291 1618 200 168 36 3282 PCBa-036 8176 7203 1282 75 743 4215 4100 3076 785 337 219 2212 PCBa-ED20 15135 16727 5234 37 3730 9953 8681 12899 5673 2012 2162 16739 PCBa-ED19 863 7594 2471 40 1971 6126 5047 7697 4153 1039 1822 12663 PCBa-039 176 2826 1012 96 1218 6004 7327 5438 1661 808 102 6038 PCBa-ED18 67 737 405 35 64 795 714 1070 129 92 23 3058 PCBa-041 28 73 35 187 n/a 1296 2249 1499 301 83 37 6260 PCBa-ED17 12 418 159 18 82 658 990 1205 321 172 65 6965 PCBa-ED16 39 303 368 60 12 307 342 454 19 23 7 981 PCBa-ED15 164 717 607 43 55 892 1050 1546 303 113 177 6965 PCBa-ED14 263 1033 317 52 70 1327 1755 3099 806 226 292 12311 PCBa-ED13 22 581 364 21 49 2091 3574 3297 1098 309 185 19332 PCBa-ED12 33 416 250 50 59 964 1348 2580 595 174 378 10659 PCBa-047 296 2299 296 292 28 2994 5928 3852 654 256 81 25128 PCBa-ED11 17 142 128 115 6 77 70 100 10 9 6 400 PCBa-ED10 465 751 319 23 54 1051 2205 2783 1694 291 446 25048 PCBa-ED9 16 888 271 9 94 1207 2291 2854 1737 383 516 15049 PCBa-050 14 277 97 170 n/a 735 2646 2013 1350 195 644 18290 PCBa-052 765 3158 546 86 313 4011 15338 9384 4690 1145 2582 27060 PCBa-ED8 71 2525 944 19 277 5312 9949 13137 9010 1930 5718 36880 PCBa-ED7 9 606 394 23 64 2154 4250 5855 2094 709 1163 21380 PCBa-059 21 810 239 79 52 3631 13351 10994 4425 1261 3963 93759 PCBa-ED6 3 554 139 4 37 1442 3150 4614 1005 434 800 15909

149

cd]pyrene

Sample ID Sample Phenanthrene Fluoranthene Pyrene Retene Benz[a]anthracene Chrysene Benzofluranthene Benzopyrene Indeno[1,2,3 Dibenz[a,h]anthracene Benzo[g,h,i]perylene Coronene PCBa-ED5 40 3482 690 3 637 10100 29389 20350 13665 4248 12116 35104 PCBa-ED4.1 13 298 116 5 17 1367 2294 2555 192 198 199 4672 PCBa-ED4 8 398 197 7 70 1952 1767 2368 296 173 258 6459 PCBa-072 84 384 25 310 28 2338 3290 1309 118 60 57 6449 PCBa-ED3 6 423 478 10 118 2102 4295 3931 1035 227 290 8008 PCBa-076 264 643 301 98 223 3665 26392 12093 3838 1109 2682 24706 PCBa-ED2 80 3468 953 11 710 7290 27004 19262 9990 2322 3703 21631 PCBa-ED1 83 3897 807 10 511 7298 21115 16423 5558 1793 1875 21115 PCBa-097 8 869 161 90 44 3946 17224 7699 1437 348 839 13615 PCBa-0xx δ13C and TOC values from Foreman et al., (2012); grey shading marks PETM

Table B-4. Field descriptions for all samples collected in 2013 (PCBa-ED) Height Sample ID (m) Lithology Color Color # GPS ± 13-25 ft N 39°16.610', PCBa-ED43 -18.19 very fine sand (vfl) pale olive 10Y 6/2 W 108°11.329' N 39°16.610', PCBa-ED42 -20.1 clayey silt olive grey 5Y 4/1 W 108°11.329' N 39°16.610', PCBa-ED41 -21.91 silt olive grey 5Y 4/1 W 108°11.329' N 39°16.610', PCBa-ED40 -22.35 fine sand (vfl-vfu) light olive grey 5Y 5/2 W 108°11.329' N 39°16.610', PCBa-ED39 -33.5 sand green W 108°11.329' 5GY N 39°16.617', PCBa-ED38 -43.4 sandy silt dusky yellow green 5/2 W 108°11.356' N 39°16.617', PCBa-ED37 -43.9 sandy silt pale olive grey 10Y 6/2 W 108°11.356' N 39°16.617', PCBa-ED36 -45.08 silt pale olive grey 10Y 6/2 W 108°11.356' N 39°16.617', PCBa-ED35 -48.98 silt medium grey N5 W 108°11.356' N 39°16.617', PCBa-ED34 -50.5 clay olive grey 5Y 4/1 W 108°11.356' N 39°16.617', PCBa-ED33 -50.8 clay olive grey 5Y 4/1 W 108°11.356' PCBa-ED32 -51.75 silty clay Medium dark grey N4 N 39°16.617',

150

Height Sample ID (m) Lithology Color Color # GPS ± 13-25 ft W 108°11.356' 5GY N 39°16.622', PCBa-ED31 -61.25 sandy silt greyish yellow green 7/2 W 108°11.413' 5GY N 39°16.622', PCBa-ED30 -61.6 clayey silt greyish yellow green 7/2 W 108°11.413' 5GY N 39°16.622', PCBa-ED29 -61.75 very fine sand (vfl) greyish yellow green 7/2 W 108°11.413' 5GY N 39°16.622', PCBa-ED28 -62.05 sandy silt greyish yellow green 7/2 W 108°11.413' N 39°16.622', PCBa-ED27 -62.95 very fine sand (vfl) pale red 10R 6/2 W 108°11.413' 5GY N 39°16.622', PCBa-ED26 -63.9 fine sand (vfl-vfu) greyish yellow green 7/2 W 108°11.413' N 39°16.622', PCBa-ED25 -63.92 clay olive grey 5Y 4/1 W 108°11.413' N 39°16.622', PCBa-ED24 -65.05 very fine sand (vfl) yellowish grey 5Y 7/2 W 108°11.413' N 39°16.622', PCBa-ED23 -65.7 silt pale greenish yellow 10Y 8/2 W 108°11.413' 5GY N 39°16.622', PCBa-ED22 -65.75 fine sand (vfl-vfu) dusky yellow green 5/2 W 108°11.413' N 39°16.622', PCBa-ED20 -65.85 silt medium grey N5 W 108°11.413' 5GY N 39°16.622', PCBa-ED21 -66 fine sand (vfl-vfu) dusky yellow green 5/2 W 108°11.413' 5GY N 39°16.622', PCBa-ED19 -67.03 silty clay greenish grey 6/1 W 108°11.413' N 39°16.622', PCBa-ED18 -67.57 sandy silt yellowish grey 5Y 7/2 W 108°11.413' N 39°16.622', PCBa-ED17 -68.65 silty clay light olive grey 5Y 6/1 W 108°11.413' N 39°16.622', PCBa-ED16 -69.93 fine sand (vfl-vfu) pale green 10G 6/2 W 108°11.413' N 39°16.622', PDBa-ED15 -70.23 fine sand (vfl-vfu) pale green 10G 6/2 W 108°11.413' N 39°16.644', PCBa-ED14 -70.75 silty sand (vfl) W 108°11.420' 10YR N 39°16.644', PCBa-ED13 -71.4 sandy silt pale yellowish brown 6/2 W 108°11.420' N 39°16.644', PCBa-ED12 -72 sandy silt light olive grey 5Y 5/2 W 108°11.420' N 39°16.644', PCBa-ED11 -73.5 fine sand greenish W 108°11.420' N 39°16.644', PCBa-ED10 -73.85 sandy silt light olive grey 5Y 5/2 W 108°11.420' N 39°16.644', PCBa-ED9 -74.35 sandy silt light olive grey 5Y 5/2 W 108°11.420' N 39°16.644', PCBa-ED8 -78.93 clayey silt olive grey 5Y 4/1 W 108°11.420' N 39°16.644', PCBa-ED7 -80.56 sandy silt light olive grey 5Y 5/2 W 108°11.420'

151

Height Sample ID (m) Lithology Color Color # GPS ± 13-25 ft N 39°16.644', PCBa-ED6 -83.25 clay dark grey N3 W 108°11.429' 5GY N 39°16.644', PCBa-ED5 -88.79 clay dark greenish grey 4/1 W 108°11.429' N 39°16.644', PCBa-ED4.1 -91.8 clayey silt medium blueish grey 5B 5/1 W 108°11.429' 5GY N 39°16.644', PCBa-ED4 -92.1 silty clay dark greenish grey 4/1 W 108°11.429' N 39°16.22', W PCBa-ED3 -95.05 clayey silt Medium dark grey N4 108°11.446' 5GY N 39°16.22', W PCBa-ED2 -97.62 silt dark greenish grey 4/1 108°11.446' N 39°16.22', W PCBa-ED1 -99.05 clayey silt Medium dark grey N4 108°11.446' Grey shading marks samples from the PETM interval

Table B-5. Piceance Basin XRF elemental data from Foreman (2012).

(wt (wt %) (wt %)

(wt (wt %)

(wt %)

3 3

O (wt (wt O %)

O O

2 2

TOC TOC (%)

Lithology

Sample ID Sample

CaO (wt (wt %)CaO

SiO

TiO

MgO%) (wt

Soil Horizon

Fe Al Strat. (m) Height

C PCBA2-36 39.95 clay 0.02 4.30 0.98 70.1 13.3 0.23 0.55 0.75 2.24 B PCBA-16 -44.6 silt 0.02 1.83 0.30 81.1 9.8 0.04 0.49 1.08 2.04 B PCBA-17 -45 silt 0.03 4.75 0.59 68.6 14.9 0.04 0.57 0.79 2.39 C PCBA-18 -49.6 silt 0.03 3.04 0.46 74.2 12.8 0.06 0.51 1.13 2.32 C PCBA-19 -51.1 silt 0.04 4.11 0.74 70.1 14.1 0.05 0.62 1.00 2.82 B PCBA-24 -59.82 silt 0.19 3.95 0.91 57.5 27.4 0.07 1.10 0.52 2.37 B PCBA-25 -59.95 silt 0.07 6.88 0.75 66.5 16.6 0.06 0.76 0.71 1.91 B PCBA-26 -60.55 silt 0.07 5.15 0.72 65.2 17.0 0.05 0.74 0.62 1.82 B/C PCBA-27 -61.1 silt 0.05 3.80 0.65 72.3 13.5 0.05 0.60 0.82 1.94 B PCBA-44 -71.8 silt 0.11 2.00 0.35 78.7 10.8 0.04 0.57 0.55 0.75 B PCBA-45 -72.25 silt 0.05 3.89 0.45 75.9 12.7 0.04 0.63 0.77 1.49 B PCBA-66 -89 clay 0.21 5.22 0.59 63.8 19.7 0.08 0.94 0.17 0.54 C PCBA-68 -90.2 clay 0.21 5.36 0.61 64.1 19.3 0.07 0.96 0.17 1.01 B PCBA-69 -91.3 clay 0.14 6.82 0.67 60.0 20.8 0.10 0.95 0.15 1.01 B PCBA-70 -91.6 clay 0.15 7.23 0.71 54.8 25.3 0.16 0.86 0.13 0.74 clayey C PCBA-72 -92.2 0.06 7.50 0.60 65.0 16.4 0.05 0.62 0.39 1.72 silt silty C PCBA-90 -107.1 0.08 3.82 0.50 71.8 13.8 0.06 0.58 0.84 1.45 clay TOC and organic carbon isotope data is from (Foreman et al., 2012).

152 References

Baczynski, A.A., McInerney, F.A., Wing, S.L., Kraus, M.J., Morse, P.E., Bloch, J.I., Chung, A.H., and Freeman, K.H., 2016, Distortion of carbon isotope excursion in bulk soil organic matter during the Paleocene-Eocene thermal maximum: Geological Society of America Bulletin, , no. Xx, p. B31389.1, doi: 10.1130/B31389.1. Bray, E., and Evans, E., 1961, Distribution of n-paraffins as a clue to recognition of source beds: Geochimica et Cosmochimica Acta, v. 22, no. 1, p. 2–15, doi: 10.1016/0016- 7037(61)90069-2. Bush, R.T., and McInerney, F.A., 2013, Leaf wax n-alkane distributions in and across modern plants: Implications for paleoecology and chemotaxonomy: Geochimica et Cosmochimica Acta, v. 117, p. 161–179, doi: 10.1016/j.gca.2013.04.016. Foreman, B.Z., 2012, Fluvial Response to the Paleocene-Eocene Thermal Maximum in Western North America: University of Wyoming, 147 p. Foreman, B.Z., Heller, P.L., and Clementz, M.T., 2012, Fluvial response to abrupt global warming at the Palaeocene/Eocene boundary.: Nature, v. 491, no. 7422, p. 92–5, doi: 10.1038/nature11513. Johnson, S.Y., 1992, Phanerozoic Eeolution of sedimentary basins in the Uinta-Piceance Basin region, northwestern Colorado and northeastern Utah: USGS Bulletin, v. 1787-FF, p. 1– 38. Johnson, R.C., and Rice, D.D., 1990, Occurrence and Geochemistry of Natural Gases, Piceance Basin, Northwest Colorado: The American Association of Petroleum Geogolists Bulletin, v. 74, no. 6, p. 805–829. Peters, K.E., Walters, C.C., and Moldowan, J.M., 2005, The Biomarker Guide. Volume 2: Biomarkers and Isotopes in Petroleum Exploration and Earth History: Cambridge University Press, Cambridge.

153

Appendix C

Chapter 4 Supplemental Material

Figures

Figure C-1. Depth profile from IODP Hole 302-4A of carbon isotope values, total organic carbon (TOC), and Coronene/Pyrene+Coronene ratio. Core recovery column, where grey represents recovered core and “x” marks intervals without recovered material; error bars connected to Core 31X mark the uncertainty of its stratigraphic 13 position (Sluijs et al., 2006). Depth profile of carbon isotope (δ Corg; black circle), total organic carbon (TOC; white circle) (from Schouten et al., 2007)), terrestrial organic carbon (TOCterr.;

154 black circle); marine organic carbon (TOCmarine; grey circle) values, Coronene/Pyrene+Coronene ratio (black diamond) with dashed line at a ratio of 0.5. Horizontal dashed lines mark the PETM interval. Horizontal bars represent the uncertainty in TOCterr. or TOCmarine since each was determined from an average of the BIT index and %Terrestrial Palynomorphs (from Sluijs and Dickens (2012)). If bar is not visible, uncertainty is less than the size of symbol.

Piceance

0.8 Polecat Basin Substation 0.6 Arctic

Pyrene+Coronene 0.4

0.2 Coronene/

0 0 1 2 3 4 5 TOC (%C)

Figure C-2. Cross plot of Coronene/Pyrene+Coronene ratio and %TOC. Includes data from Piceance Basin, Colorado; Polecat Bench, Wyoming; Basin Substation, Wyoming (Denis et al., in prep) and the Arctic (see legend).

155

Percent of Total (%) 0 20 40 60 80 100 376

378

380

382

Spores

Gymnosperms

384 Angiosperms Depth (mcd) Depth 386

388

390

392

Figure C-3. Relative percentage of terrestrial palynomorphs by type: spores (square); angiosperms (black diamond), gymnosperms (white triangle) (data from Sluijs et al., (2006)).

Data Tables

156

Table C-1. Arctic TOCterrestrial and uncertainty with depth

terrestrial

TOC

papers) -

Sample Name (Arctic) for

BIT&%TP average

depth (mcd

terrestrial

UncertaintyBar TOC 302-4A-29X-1W-42-43 halfS2 378.21 1.28 0.51 302-4A-29X-1W-62-63 halfS2 378.41 1.34 0.55 378.61 0.95 0.25 302 -4A-29X-1W-102-103 halfS2 378.81 0.76 0.43 302-4A-29X-CCW-2-3 halfS2 379.01 1.52 1.00 302-4A-30X-1-1-3 SD halfS2 380.31 302-4A-30X-1-21-23 SD halfS2 380.51 302-4A-30X-1W-41-43 halfS2 380.71 302-4A-30X-1W-61-63 halfS2 380.91 302-4A-30X-21-81-83 halfS2 381.11 302-4A-30X-2W-2-3 halfS2 381.82 1.30 0.64 302-4A-30X-2W-22-23 halfS2 382.02 1.30 0.77 302-4A-30X-2W-42-43 halfS2 382.22 1.23 0.53 302-4A-30X-2W-62-63 halfS2 382.42 0.96 0.31 302-4A-30X-2W-82-83 halfS2 382.62 1.85 1.11 302-4A-30X-2W-102-103 halfS2 382.82 1.05 0.49 302-4A-30X-2W-122-123 halfS2 383.02 0.92 0.55 302-4A-30X-3W-2-3 halfS2 383.34 0.62 0.30 302-4A-30X-3W-22-23 halfS2 383.54 0.47 0.22 302-4A-30X-3W-62-63 halfTLE 383.94 0.55 0.26 302-4A-30X-3W-82-83 halfS2 384.14 0.86 0.07 302-4A-30X-3W-102-103 halfS2 384.34 1.26 0.21 302-4A-31X-CC-10-12-SD halfS2 385.8 1.30 0.03 302-4A-31X-CC-23-25 halfS2 385.93 1.02 0.28 302-4A-31X-CC-40-42-SD halfS2 386.1 0.74 0.29 302-4A-32X-1-1-3 SD halfS2 388.01 302-4A-32X-1W-21-23 halfS2 388.21 302-4A-32X-1-41-43 SD halfS2 388.41 302-4A-32X-1W-62-63 halfS2 388.61 1.95 0.08

157

terrestrial

TOC

papers) -

Sample Name (Arctic) for

BIT&%TP average

depth (mcd

terrestrial

UncertaintyBar TOC 302-4A-32X-1W-82-83 halfS2 388.81 2.24 0.68 302-4A-32X-1W-102-103 halfS2 389.01 2.03 0.10 302-4A-32X-1W-122-123 halfTLE 389.21 1.87 0.08 302-4A-32X-1W-141-143 halfS2 389.41 0.68 0.01 302-4A-32X-2W-2-3 halfS2 389.51 0.97 0.00 302-4A-32X-2W-22-23 halfS2 389.71 1.45 0.02 302-4A-32X-2W-42-43 halfTLE 389.91 3.50 0.48 302-4A-32X-2W-62-63 halfTLE 390.11 0.63 0.02 302-4A-32X-2W-82-83 halfS2 390.31 1.57 0.16 302-4A-32X-2W-102-103 halfS2 390.51 3.98 0.01 302-4A-32X-2W-122-123 halfS2 390.71 1.72 0.12 Grey shade are PETM samples (Schouten et al.,2007) Red + green highlights from orginal mastersheet (Nikolai Pedentchouk) Purple highlights TLE samples rather than S2 fraction.

Table C-2. PAH/plant biomarker ratios with depth.

amyrin

+ +

amyrin

papers)

]pyrene/Simonellite+

amyrin deriv.

158

amyrin

+ +

amyrin

papers)

]pyrene/Simonellite+

amyrin deriv.

depth (mcd Fluoranthene/Simonellite amyrin deriv. Pyrene/Simonellite + β deriv. Coronene/Simonellite + deriv. Benzo[ amyrin deriv. Benzofluoranthenes/Simonellite + β Benzo[e]pyrene/Simonellite + amyrin deriv. 380.31 0.0244 0.0636 0.1462 0.0346 0.1067 0.0912 380.51 0.0310 0.1047 0.2205 0.0277 0.1047 0.0328 380.71 0.0650 0.1314 0.0234 0.0335 0.1075 0.0383 380.91 0.0212 0.0773 0.0300 0.0182 0.0716 0.0420 381.11 0.0264 0.0937 0.0910 0.0235 0.0390 0.0528 381.82 0.0206 0.1587 0.0256 0.0213 0.0827 0.0445 382.02 0.0563 0.2272 0.0340 0.0260 0.1177 0.0613 382.22 0.0452 0.1614 0.0372 0.0285 0.1392 0.0694 382.42 0.0253 0.2133 0.0362 0.0361 0.1304 0.0605 382.62 0.0804 0.3101 0.0653 0.0643 0.2303 0.1009 382.82 0.0593 0.4116 0.0862 0.1050 0.4045 0.1550 383.02 0.0616 0.0646 0.1116 0.1354 0.4994 0.1874 383.34 0.0743 0.6404 0.1405 0.1923 0.7055 0.2423 383.54 0.1057 0.2899 0.1872 0.1751 0.7238 0.2408 383.94 0.1018 1.1109 0.1593 0.1634 0.8149 0.2303 384.14 0.1711 0.6109 0.1169 0.1097 0.7324 0.1836 384.34 0.1236 0.2632 0.1487 0.1749 0.9911 0.2575 385.8 0.0718 0.5738 0.1340 0.1498 0.6535 0.1884 385.93 0.1315 0.4019 0.0798 0.1029 0.5349 0.1578 386.1 0.0765 0.1796 0.1151 0.1877 0.9962 0.2661 388.01 0.0317 0.0359 0.0454 0.0013 0.0467 0.0133 388.21 0.0254 0.0429 0.0094 0.0224 0.0697 0.0291 388.41 0.0197 0.0395 0.0343 0.0111 0.0573 0.0169 388.61 0.0116 0.0267 0.0086 0.0048 0.0267 0.0108 388.81 0.0099 0.0433 0.0144 0.0054 0.0372 0.0178 389.01 0.0067 0.0177 0.0302 0.0086 0.0512 0.0216 389.21 0.0061 0.0143 0.0127 0.0032 0.0183 0.0072 389.41 0.0057 0.0225 0.0119 0.0050 0.0289 0.0116 389.51 0.0092 0.0151 0.0121 0.0035 0.0194 0.0082 389.71 0.0095 0.0359 0.0191 0.0072 0.0405 0.0151 389.91 0.0107 0.0295 0.0217 0.0068 0.0340 0.0142 390.11 0.0101 0.0060 0.0156 0.0052 0.0311 0.0129

159

amyrin

+ +

amyrin

papers)

]pyrene/Simonellite+

amyrin deriv.

Table C-3. Coronene/Pyrene ratio, total organic carbon (TOC), plant biomarker (ng/g TOCterrestrial)

and pollen (number/g TOCterrestrial) concentrations.

papers)

alkane

± ± ± ± ± ±

amyrin

29 29

TOC TOC (%)

Simonellite Simonellite

dehydroabietane dehydroabietane

Angiosperm pollen Angiosperm

depth depth (mcd

Gymnosperm pollen Gymnosperm Coronene/pyrene+coronene

378.21 2.56 0.19 19.49 3.91 25.34 5.09 18.94 3.80 6.11 1.23 40 8 37 7 378.41 1.98 0.14 7.08 1.97 10.14 2.83 8.32 2.32 2.56 0.71 134 37 96 27 378.61 1.82 8.62 1.19 17.20 2.37 13.48 1.86 4.13 0.57 47 6 59 8

378.81 1.72 0.22 3.30 0.82 30.86 7.66 20.77 5.16 5.89 1.46 71 18 53 13 379.01 2.80 0.22 1.87 0.66 17.10 6.08 21.80 7.75 5.30 1.88 190 67

380.31 0.70

380.51 0.68

380.71 0.15

380.91 0.28

381.11 0.49 381.82 2.36 0.14 1.42 0.39 3.66 1.00 14.97 4.08 4.92 1.34 58 16 81 22 382.02 2.44 0.13 1.99 0.62 2.99 0.94 10.88 3.42 4.05 1.27 74 23 64 20 382.22 3.35 0.19 3.61 0.58 4.57 0.73 17.33 2.77 4.54 0.73 111 18 99 16

160

papers)

alkane

± ± ± ± ± ±

amyrin

29 29

TOC TOC (%)

Simonellite Simonellite

dehydroabietane dehydroabietane

Angiosperm pollen Angiosperm

depth depth (mcd

Gymnosperm pollen Gymnosperm Coronene/pyrene+coronene

382.42 2.50 0.15 4.82 0.60 6.35 0.80 24.20 3.03 4.94 0.62 40 5 52 6 382.62 4.36 0.17 5.46 1.39 6.14 1.56 11.07 2.81 3.39 0.86 111 28 37 9 382.82 4.47 0.17 11.65 1.28 14.11 1.55 5.94 0.65 5.94 0.65 178 19 118 13 383.02 3.19 0.63 7.88 1.35 8.75 1.50 19.31 3.31 4.90 0.84 188 32 84 14 383.34 3.18 0.18 4.42 0.42 8.09 0.77 19.26 1.83 4.38 0.41 240 23 89 8 383.54 2.69 0.39 3.66 0.31 19.29 1.61 39.06 3.26 10.97 0.91 400 33 66 6 383.94 2.70 0.13 1.29 0.12 13.55 1.29 22.41 2.14 4.11 0.39 273 26 2 0 384.14 3.00 0.16 1.04 0.02 5.87 0.14 11.73 0.27 1.68 0.04 110 3 28 1 384.34 4.80 0.36 1.33 0.06 6.28 0.28 7.04 0.31 1.76 0.08 180 8 27 1 385.8 3.45 0.19 3.08 0.02 9.23 0.07 114 1 157 1 385.93 3.09 0.17 3.62 0.33 12.23 1.13 183 17 20 2 386.1 3.01 0.39 4.08 0.40 14.95 1.46 185 18 15 1 388.01 0.56

388.21 0.18

388.41 0.46

388.61 2.17 0.24 5.46 0.21 9.64 0.38 0.83 0.03 9.04 0.35 16 1 8 0 388.81 3.64 0.25 3.86 0.72 6.15 1.15 5.07 0.94 3.96 0.74 14 3 11 2 389.01 2.33 0.63 4.35 0.19 6.14 0.27 3.66 0.16 0.83 0.04 10 0 6 0 389.21 1.99 0.47 5.70 0.23 6.91 0.28 5.00 0.20 3.33 0.13 11 0 4 0 389.41 0.76 0.35 14.49 0.13 19.42 0.18 11.96 0.11 15.05 0.14 40 0 16 0 389.51 1.14 0.45 5.02 0.00 10.50 0.00 7.87 0.00 7.85 0.00 19 0

389.71 1.66 0.35 12.28 0.11 14.14 0.13 7.78 0.07 8.02 0.07 37 0 14 0 389.91 4.41 0.42 3.72 0.40 4.65 0.51 2.50 0.27 1.82 0.20 6 1

390.11 0.67 0.72 28.55 0.91 31.08 0.99 18.11 0.58 25.83 0.83 84 3 29 1 390.31 1.95 0.24 17.18 1.45 19.11 1.62 8.59 0.73 11.89 1.01 15 1

390.51 4.30 0.17 2.92 0.00 4.20 0.01 2.37 0.00 2.08 0.00 9 0 3 0 390.71 1.96 0.26 8.86 0.54 13.35 0.82 6.44 0.39 3.32 0.20 13 1 31 2 Grey highlights the PETM interval. Uncertainty is from uncertainty in TOCterrestrial. Original plant biomarker data from Schouten et al. (2007) and original pollen data from Sluijs et al. (2006).

161 Table C-4. Arctic Pristane/Phytane ratios. Depth Sample Name (mcd) Pristane/Phytane 302_4A_29X_1W_042_43_S1NA 378.21 2.59 302_4A_29X_1W_062_63_S1NA 378.41 1.55 302_4A_29X_1W_082_83_S1NA 378.61 3.10 302_4A_29X_1W_102_103_S1NA 378.81 2.26 302_4A_30X_1_001_3_SD_S1NA 380.31 0.68 302_4A_30X_1_021_23_SD_S1NA 380.51 0.34 302_4A_30X_1_041_43_SD_S1NA 380.71 0.65 302_4A_30X_1_061_63_SD_S1NA 380.91 0.40 302_4A_30X_1_081_83_SD_S1NA 381.11 0.42 302_4A_30X_2W_002_3_S1NA 381.82 1.14 302_4A_30X_2W_022_23_S1NA 382.02 1.58 302_4A_30X_2W_042_43_S1NA 382.22 2.15 302_4A_30X_2W_062_63_S1NA 382.42 0.58 302_4A_30X_2W_082_83_S1NA 382.62 1.81 302_4A_30X_2W_102_103_S1NA 382.82 0.31 302_4A_30X_2W_122_123_S1NA 383.02 0.71 302_4A_30X_3W_002_3_S1NA 383.34 0.62 302_4A_30X_3W_022_23_S1NA 383.54 0.33 302_4A_30X_3W_062_63_S1NA 383.94 0.18 302_4A_30X_3W_082_83_S1NA 384.14 0.29 302_4A_30X_3W_102_103_S1NA 384.34 0.84 302_4A_31X_CC_010_12_SD_S1NA 385.80 0.12 302_4A_31X_CC_023_25_SD_S1NA 385.93 0.16 302_4A_31X_CC_040_42_SD_S1NA 386.10 0.26 302_4A_32X_1_001_3_SD_S1NA 388.01 0.16 302_4A_32X_1_021_23_SD_S1NA 388.21 0.13 302_4A_32X_1_041_43_SD_S1NA 388.41 0.22 302_4A_32X_1W_062_063_S1NA 388.61 0.56 302_4A_32X_1W_082_83_S1NA 388.81 1.60 302_4A_32X_1W_102_103_S1NA 389.01 0.84 302_4A_32X_1W_122_123_S1NA 389.21 1.27 302_4A_32X_1W_142_143_S1NA 389.41 0.75 302_4A_32X_2W_002_3_S1NA 389.51 0.40 302_4A_32X_2W_022_23_S1NA 389.71 1.59 Grey highlights the PETM interval.

162

Table C-5. Percent of angiosperms based on biomarkers and on pollen.

papers) -

Depth (mcd Depth

~Pollen Depth (mcd) Depth ~Pollen

alkanes/diterpenoids x 100% x alkanes/diterpenoids

%Angiosperm from Pollen from %Angiosperm

n Triterpenoid/diterpenoid x 100% x Triterpenoid/diterpenoid

378.21 12.00 52.99 378.14 51.89 378.41 12.94 58.51 378.33 57.55 378.61 13.79 58.44 378.53 45.28 378.81 14.71 62.94 378.72 56.60 379.01 21.83 77.46 378.91 24.53 380.31 381.38 41.51 380.51 381.64 34.91 380.71 381.70 49.06 380.91 381.77 41.51 381.11 381.89 52.83 381.51 40.11 85.57 382.15 51.89 381.71 45.45 87.74 382.35 43.40 381.82 49.20 88.80 382.54 73.58 382.02 44.87 85.08 382.80 59.43 382.22 35.72 83.42 382.93 67.92 382.42 30.67 83.46 383.19 87.74 382.62 22.60 68.62 383.26 72.64 382.82 18.75 55.70 383.51 84.91 383.02 22.74 73.37 383.71 67.92 383.22 18.73 64.39 383.90 96.23 383.34 25.91 78.77 384.10 76.42 383.54 32.33 77.39 384.29 85.85 383.74 19.66 77.43 385.72 41.51 383.94 21.70 75.18 385.85 89.62 384.14 19.61 79.93 386.04 90.57 384.34 18.76 70.54 388.57 65.09 385.80 388.76 54.72 385.93 388.96 62.26 386.10 389.15 68.87

163

papers) -

Depth (mcd Depth

~Pollen Depth (mcd) Depth ~Pollen

alkanes/diterpenoids x 100% x alkanes/diterpenoids

%Angiosperm from Pollen from %Angiosperm

n Triterpenoid/diterpenoid x 100% x Triterpenoid/diterpenoid

388.01 389.41 66.98 388.21 389.47 59.43 388.41 389.73 68.87 388.61 37.46 53.48 389.86 52.83 388.81 28.35 61.32 390.06 73.58 389.01 7.32 51.22 390.25 50.00 389.21 20.87 55.41 390.51 69.81 389.41 30.74 51.61 390.70 30.19 389.51 33.60 58.45 389.71 23.28 46.23 389.91 17.86 46.11 390.11 30.23 48.72 390.31 24.69 40.89 390.51 22.61 48.87 390.71 13.01 48.39

Grey highlights the PETM interval. Depth color highlights correspond to Mastersheet colors from Nikolai Pedentchouk. Original biomarker and pollen data from Schouten et al. (2007).

164

Appendix D

Carbon isotopes of polycyclic aromatic hydrocarbons (PAHs)

Carbon isotopes of polycyclic aromatic hydrocarbons (PAH) to evaluate pyrogenic carbon source

My original investigation of fire during the Paleocene-Eocene Thermal Maximum was driven by the idea that combustion of 13C-depleted carbon deposits could have contributed to the negative carbon isotope excursion. The Paleocene was a time of extensive terrestrial organic carbon burial, such as in Wyoming as evidenced by present-day coal deposits, and the dramatic change in climate at the PETM (warming and seasonal precipitation extremes or drying), may have induced burning of shallowly buried carbon deposits (Kurtz et al., 2003). Results from the

Arctic, suggest, at least at this site, that fire was a consequence of global warming rather than the cause. The extensive organic carbon degradation in the terrestrial sites in the western USA, though, obscured evidence for changes in fire occurrence. Nevertheless, the issue of what material is burning during the PETM is of interest, and I could use the δ13C value of PAH to help distinguish between the burning of PETM biomass and Paleocene peat or coal, similar to methods on black carbon by Moore and Kurtz (2008). The carbon isotope excursion would be recorded in

PAHs across the PETM if biomass that has isotope signature equilibrated to the atmosphere is burning, whereas, if Paleocene carbon deposits are burning, the isotope excursion will not be recorded in the PAHs. A gas chromatography-isotope ratio mass spectrometer (GC-IRMS) capable of measuring picomolar amounts of carbon for compound-specific isotope analyses

(expected in 2016 for Freeman Lab) will help to measure the δ13C value of PAHs in samples that are otherwise material limited.

165 References

Kurtz, A. C., Kump, L.R., Arthur, M. A., Zachos, J.C., and Paytan, A., 2003, Early Cenozoic decoupling of the global carbon and sulfur cycles: Paleoceanography, v. 18, no. 4, p. 1– 14, doi: 10.1029/2003PA000908. Moore, E.A., and Kurtz, A.C., 2008, Black carbon in Paleocene-Eocene boundary sediments: A test of biomass combustion as the PETM trigger: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 267, no. 1-2, p. 147–152, doi: 10.1016/j.palaeo.2008.06.010.

166

Appendix E

Colombia and Venezuala (Mar2x core)

Introduction

Originally I set out to investigate this low-latitude site to help answer the questions of how fire occurrence changed at a global scale across the Paleocene-Eocene Thermal Maximum

(PETM) and how feedbacks between aridity and fire operated in the past. I was interested in

Colombia and Venezuala because it was the Cerrejon formation core was from a low-latitude site and because n-alkane and botanical data suggested that it was either wetter during the PETM relative to the Paleocene or there was no change in precipitation (Jaramillo et al., 2010). n-Alkane deuterium isotope (δDn-alkanes) values were lower during the PETM relative to the late Paleocene, which suggested conditions were wetter during the PETM (Jaramillo et al., 2010).

Methods

Samples were acquired by Heather Graham (Penn State) from the Mar2x core (Jaramillo et al., 2010; Graham, 2014).

Lithologic descriptions were based on visual inspection and grainsize was determined based on chewing the samples. Samples were considered sandstone if grains were visible with own eyes. Pieces of samples (~3 mm3) were in mouth for 30 – 60 seconds and chewed to break apart particles. Clay, clayey mudstone, silty mudstone, and siltstone were determined based on the grittiness of the sample, where samples that became less gritty with time supported the presence of some clay. Lithology key: (1) Claystone – no grit (no apparent silt); (2) Clayey

167 mudstone – slightly gritty, but mostly disintegrates (mostly clay, but mixture of clay and silt); (3)

Silty mudstone – gritty, but less gritty with time (mostly silt, but mixture of clay and silt); (4)

Siltstone – very gritty, sample doesn’t disintegrate (no apparent clay).

Samples were prepared for lipid analyses on the gas chromatograph-mass spectrometer

13 and for total organic carbon (TOC) and bulk carbon isotope analyses (δ Corg) following the

13 procedures outlined in Denis et al., (in prep). Complementary TOC, δ Corg, n-alkane data was processed and reported on by Heather Graham (Graham, 2014).

Results

There was no trend in individual PAHs or by ring-class before, during, or after the PETM in terms of absolute concentration relative to TOC.

The ratio of large PAH to small PAH (6-ring/3-ring) ranged from ~0.4 to 0.8 in pre-

PETM samples. During the PETM the ratio dropped to ~0.1 – 0.2, with a spike back to ~0.6 mid- way through the PETM. Post-PETM the ratio was consistently between ~0.2 – 0.4.

The ratio of Coronene/Pyrene was highly variable in pre-PETM samples (~0 – 1) until mid-way through the PETM interval. Samples from the upper part of the PETM interval and post-

PETM were less variable and ranged between ~0.3 – 0.6.

Discussion

The Mar2x samples were rich in organic carbon. Chromatograms of total lipid extracts

(TLEs) and the unsaturated/aromatic fraction (F2) were full of compounds. Many had an unresolved complex mixture. The high abundance of many compounds, including a plethora of

168 methylated and akylated PAHs, made quantification challenging. The abundance of substituted

PAHs, often greater than unsubstituted PAHs, suggests PAHs may have been formed by non- pyrogenic sources (Page et al., 1999; Yunker et al., 2002). There is no clear trend in PAHs during the designated PETM interval. The change in ratio of large to small PAHs mid-way through the

PETM is intriguing.

Figures

Figure E-1. Geographical map of study location in Colombia (Mar 2x core, yellow star) from Jaramillo et al. (2010). Inset shows paleogeographic location (Paleocene-Eocene).

169

Figure E-2. Depth profile of total organic carbon (TOC) and polycyclic aromatic hydrocarbons (PAHs) concentrations by ring size.

Figure E-3. Depth profile of total organic carbon (TOC) and polycyclic aromatic hydrocarbon (PAHs) concentrations.

170

Data Tables

Table E-1. Mar2x samples: grams of sediment extracted, TOC, and carbon isotope values. depth depth δ13Ca TOCa δ13Cb TOCb Extracted STRI# (ft) (m) (‰) (%) (‰) (%) (g) 30927 7250 2209.8 -26.1 1.93 9.0364 30939 7271 2216.2 1.35 8.5603 30983 7345 2238.76 -27.01 0.75 9.46633 30983 7345 2238.76 -27.01 0.75 9.46633 31014 7398 2254.91 -27.13 2.624 -26.88 0.71 9.3878 31023 7408 2257.96 -27.03 0.73 9.17723 31042 7429 2264.36 0.88 9.4847 31081 7502 2286.61 -25.94 3.27 -26.63 0.54 9.5457 31092 7513 2289.96 -28.08 1.07 -27.93 1.17 7.6668 31106 7520.33 2292.2 -28.27 1.29 1.11 9.6143 31123 7559.92 2304.26 -27.29 0.67 9.1361 31125 7562 2304.9 -26.43 0.52 9.3523 31126 7563 2305.2 -26.53 0.53 7.1415 31129 7567 2306.42 0.36 9.3326 31135 7572 2307.95 -26.72 1.44 5.8826 31147 7587 2312.52 -27.21 2.92 -26.87 3.14 8.341 31150 7589 2313.13 2.38 8.402 31160 7599 2316.18 -27.11 9.77 6.9 9.5329 31165 7602 2317.09 0.68 9.4989 31168 7616.83 2321.61 0.54 9.19241 31170 7619 2322.27 -26.02 1.4 9.7455 31172 7622.08 2323.21 -17.56 0.97 9.393 31181 7629 2325.32 -24.12 0.66 9.8081 31187 7633 2326.54 -25.33 0.34 -25.03 0.32 9.636 31197 7643 2329.59 0.2 9.0914 31198 7644 2329.89 0.61 9.7415 31202 7648 2331.11 -25.33 0.43 -25.52 3.19 9.38 31209 7676 2339.64 0.32 9.3509 31219 7683 2341.78 0.3 6.9067 31224 7688 2343.3 -25.47 2.41 0.54 8.1352

171

depth depth δ13Ca TOCa δ13Cb TOCb Extracted STRI# (ft) (m) (‰) (%) (‰) (%) (g) 31226 7689 2343.61 -25.38 0.48 6.37 9.4904 31230 7693 2344.83 0.78 9.1776 a b Data from Jaramillo et al., (2010) Data from Graham, (2014)

Table E-2. Mar2x sample lithology information. depth STRI# (ft) Lithology Color Appearance 30927 7250 silty mudstone medium grey white-light grey laminations 30939 7271 claystone dark grey shiney, smooth 30983 7345 claystone dark grey smooth 30983 7345 claystone dark grey smooth 31014 7398 clayey mud dark grey 31023 7408 claystone dark grey shiney, smooth 31042 7429 claystone brown, medium grey 31081 7502 clayey mud dark grey shiney, smooth 31092 7513 clayey mud medium grey pretty smooth (not shiney) 31106 7520.33 silty mudstone medium-dark grey smooth (not shiney) 31123 7559.92 clayey mud medium-dark grey smooth and shiney planes 31125 7562 silty mudstone medium grey 31126 7563 silty mudstone medium grey some sparkle particles 31129 7567 clayey mud medium grey 31135 7572 siltstone light brown, tan 31147 7587 clayey mud dark grey shiney, smooth 31150 7589 silty mudstone dark grey 31160 7599 claystone dark grey 31165 7602 claystone very dark grey 31168 7616.83 clayey mud medium grey 31170 7619 clayey mud dark grey smooth, shiney 31172 7622.08 silty mudstone medium grey 31181 7629 clayey mud dark grey smooth, shiney 31187 7633 claystone dark grey very smooth and shiney 31197 7643 silty mudstone dark grey 31198 7644 siltstone medium-light grey 31202 7648 siltstone light grey sparkle dots 31209 7676 silty mudstone light grey brown spots 31219 7683 silty mudstone light grey sugary texture, sparkly dots 31224 7688 clayey mud dark grey shiney, smooth

172

depth STRI# (ft) Lithology Color Appearance 31226 7689 clayey mud black - darkgrey very shiney and smooth 31230 7693 silty mudstone medium grey sparkly

Table E-3. Mar2x PAH concentrations in μg/g TOC.

(ft)

STRI#

Pyrene

Fluorene

Data file Data depth

Anthracene

Naphthalene

Acenapthene

Fluoranthene

Phenanthrene Acenaphtylene 13072404 30927 7250 0.012 0.016 1.051 0.629 28.05 2.42 9.93 5.07 13072403 30939 7271 0.017 0.022 0.750 0.225 24.23 2.02 9.63 5.46 13072240 30983 7345 0.030 0.044 0.859 0.400 15.31 2.30 4.25 3.01 13072402 30983 7345 0.025 0.035 0.133 0.077 12.53 1.29 3.36 3.29 13072007 31014 7398 0.026 0.035 0.061 0.445 15.47 3.48 5.08 4.64 13072239 31023 7408 0.026 0.039 0.178 0.067 8.39 0.93 2.09 1.86 13072238 31042 7429 0.020 0.030 0.041 0.036 4.57 0.21 0.95 0.55 13072010 31081 7502 0.034 0.046 0.131 0.297 60.94 2.42 12.04 7.47 13072009 31092 7513 0.021 0.028 0.114 0.249 67.12 3.90 11.21 6.27 13072006 31106 7520.33 0.017 0.022 0.027 0.101 58.96 3.08 13.64 6.90 13072113 31123 7559.92 0.039 0.056 0.239 0.373 23.22 1.57 4.22 3.20 13072237 31125 7562 0.034 0.051 0.049 0.062 11.25 0.53 2.17 3.03 13072112 31126 7563 0.056 0.089 0.259 0.286 19.38 0.69 3.01 3.21 13072236 31129 7567 7.60 0.095 0.711 0.736 20.42 0.85 2.90 2.97 13072103 31135 7572 0.026 0.039 0.049 0.155 66.94 2.44 10.34 8.53 13072011 31147 7587 0.009 0.013 1.550 2.353 61.93 1.44 6.73 5.64 13072224 31150 7589 4.25 0.052 6.017 1.147 104.94 1.82 8.70 7.25 13072102 31160 7599 0.004 0.008 1.299 2.399 30.16 1.30 4.02 2.45 13072223 31165 7602 475.57 0.48 34.45 39.89 576.89 27.16 83.65 52.82 13072022 31168 7616.83 0.031 0.045 0.038 0.042 3.10 0.18 2.46 1.07 13072209 31170 7619 0.015 0.007 0.009 0.062 53.28 2.73 8.80 4.07 13072208 31172 7622.08 0.014 0.015 0.299 0.376 20.00 1.21 2.90 0.98 13072222 31181 7629 0.017 0.015 0.023 0.056 15.10 1.55 2.94 2.05 13072013 31187 7633 0.053 0.073 0.062 0.069 1.04 0.16 2.79 1.50 13072207 31197 7643 0.069 0.075 0.417 1.30 49.30 3.42 24.88 12.16 13072235 31198 7644 0.029 0.041 0.037 0.044 3.69 0.25 2.92 1.66

173

(ft)

STRI#

Pyrene

Fluorene

Data file Data depth

Anthracene

Naphthalene

Acenapthene

Fluoranthene

Phenanthrene Acenaphtylene 13072008 31202 7648 0.005 0.009 0.019 0.035 1.17 0.07 0.428 0.23 13072221 31209 7676 0.033 0.031 0.032 0.032 0.060 0.01 0.070 0.04 13072234 31219 7683 0.078 0.117 0.105 0.125 1.06 0.13 0.363 0.29 13072012 31224 7688 0.039 0.052 0.044 0.066 22.57 2.21 7.75 7.42 13072225 31226 7689 0.004 0.026 0.127 0.104 17.05 2.07 3.23 4.26 13072210 31230 7693 0.015 0.013 0.014 0.013 0.415 0.05 0.276 0.12 a Data file is GC-MS data file at Penn State

Table E-4. Additional Mar2x PAH concentrations in μg/g TOC

cd]pyrene

STRI#

Retene

Perylene

depth (ft)

Chrysene

Coronene

Benzo[a]pyrene

Benz[a]anthracene

Benzo[g,h,i]perylene

Dibenz[a,h]anthracene

Indeno[1,2,3 Benzofluoranthens (b+k) 30927 7250 1.83 6.76 6.40 6.29 10.55 30.40 2.47 0.87 5.15 2.75 30939 7271 1.09 6.97 6.11 5.95 9.87 43.62 2.63 0.82 6.13 2.93 30983 7345 1.14 4.84 3.11 3.65 6.55 48.42 2.16 0.36 6.11 4.21 30983 7345 1.20 6.55 4.16 3.08 7.65 25.95 2.48 0.32 5.71 3.38 31014 7398 1.44 5.22 3.84 3.00 5.36 11.02 1.58 0.48 4.62 3.59 31023 7408 0.76 3.78 2.58 3.15 4.51 33.17 1.74 0.23 3.88 2.26 31042 7429 0.14 0.87 0.00 0.38 0.89 2.19 0.16 0.09 0.80 0.31 31081 7502 2.99 7.97 8.66 8.35 7.92 7.25 3.02 0.78 4.45 2.10 31092 7513 2.00 6.26 6.13 7.20 7.62 44.09 2.61 0.85 6.14 4.61 31106 7520.33 2.23 8.52 9.09 11.11 11.46 84.81 3.55 1.17 8.82 6.63 31123 7559.92 1.83 4.91 3.57 6.21 9.91 8.64 6.19 1.25 24.05 16.91 31125 7562 0.46 3.15 2.20 4.12 4.92 3.09 4.15 0.42 20.38 15.20 31126 7563 1.14 3.62 3.18 5.14 6.54 4.89 4.34 1.43 22.75 14.63 31129 7567 0.62 4.64 3.26 6.12 7.31 4.95 6.57 0.66 30.26 19.71 31135 7572 2.48 12.63 9.07 8.43 14.23 7.34 5.15 1.09 13.39 6.21 31147 7587 2.60 8.16 10.79 6.58 12.81 10.60 3.65 0.95 8.36 3.64 31150 7589 3.55 10.58 19.72 10.96 14.29 7.55 5.12 1.99 10.07 5.00 31160 7599 1.36 3.53 3.51 1.86 3.72 17.57 1.00 0.27 1.37 0.19

174

31165 7602 20.99 75.91 80.80 67.53 95.71 220.73 25.63 5.94 31.12 8.23 31168 7616.83 0.18 1.40 1.29 1.96 0.81 3.03 1.74 0.92 4.08 5.99 31170 7619 1.14 6.36 4.88 12.13 7.10 33.92 16.54 3.68 16.49 20.77 31172 7622.08 0.41 2.59 1.90 5.62 2.43 20.93 4.72 1.34 5.12 5.58 31181 7629 0.69 3.30 2.86 9.50 3.54 28.37 4.15 1.64 6.36 9.94 31187 7633 0.69 4.11 3.07 6.99 2.21 1.53 3.01 0.88 2.01 1.79 31197 7643 8.53 30.25 26.87 84.45 40.49 99.57 104.02 29.73 108.60 180.35 31198 7644 0.90 3.07 3.12 8.68 4.30 9.33 7.96 2.43 10.99 19.17 31202 7648 0.11 0.35 0.38 0.89 0.36 0.69 0.55 0.16 0.95 1.70 31209 7676 0.052 0.101 0.266 0.553 0.001 0.018 0.007 0.001 0.053 0.001 31219 7683 0.20 0.40 0.53 0.85 0.29 0.22 0.38 0.32 0.59 1.26 31224 7688 2.69 10.64 7.32 10.97 16.78 41.97 10.38 2.23 25.79 13.45 31226 7689 1.70 6.33 4.54 4.05 7.79 29.81 4.25 0.71 6.13 1.74 31230 7693 0.06 0.28 0.28 0.33 0.03 0.02 0.15 0.05 0.11 0.15

References

Denis, E.H., Foreman, B.Z., Maibauer, B.J., Bowen, G.J., Baczynski, A.A., McInerney, F.A., Collinson, M.E., Belcher, C.M., Wing, S.L., and Freeman, K.H., in prep, Decreased soil carbon in a warming world: Degraded pyrogenic carbon during the Paleocene-Eocene Thermal Maximum (PETM): Geology. Graham, H. V, 2014, Molecular and Isotopic Indicators of Canopy Closure in Ancient Forests and the Effects of Environmental Gradients on Leaf Alkane Expression: The Pennyslvania State University, 307 p. Jaramillo, C., Ochoa, D., Contreras, L., Pagani, M., Carvajal-Ortiz, H., Pratt, L.M., Krishnan, S., Cardona, A., Romero, M., Quiroz, L., Rodriguez, G., Rueda, M.J., de la Parra, F., Morón, S., et al., 2010, Effects of rapid global warming at the Paleocene-Eocene boundary on neotropical vegetation: Science, v. 330, no. 6006, p. 957–961, doi: 10.1126/science.1193833. McInerney, F.A., and Wing, S.L., 2011, The Paleocene-Eocene Thermal Maximum: A Perturbation of Carbon Cycle, Climate, and Biosphere with Implications for the Future: Annual Review of Earth and Planetary Sciences, v. 39, no. 1, p. 489–516, doi: 10.1146/annurev-earth-040610-133431. Page, D.S., Boehm, P.D., Douglas, G.S., Bence, A.E., Burns, W.A., and Mankiewicz, P.J., 1999, Pyrogenic Polycyclic Aromatic Hydrocarbons in Sediments Record Past Human Activity: A Case Study in Prince William Sound, Alaska: Marine Pollution Bulletin, v. 38, no. 4, p. 247–260, doi: 10.1016/S0025-326X(98)00142-8. Yunker, M.B., Macdonald, R.W., Vingarzan, R., Mitchell, R.H., Goyette, D., and Sylvestre, S., 2002, PAHs in the Fraser River basin: A critical appraisal of PAH ratios as indicators of

175 PAH source and composition: Organic Geochemistry, v. 33, no. 4, p. 489–515, doi: 10.1016/S0146-6380(02)00002-5.

176

Appendix F

Tanzania

Introduction

During the PETM, precipitation changes varied globally and regionally. I obtained samples from Marcus Badger and Rich Pancost (Bristol University) from cores collected in

Tanzania as part of the Tanzania Drilling Project (Handley et al., 2008; Handley et al., 2012; Aze et al., 2014). These samples are from a hemipelagic setting and correspond to published n-alkane

δ13C and δD values (Handley et al., 2008; Handley et al., 2012; Aze et al., 2014). Published data suggest the paleoenvironment was drier during the PETM interval relative to before and after the event.

Methods

Polar fractions of extracts were obtained from Marcus Badger and Rich Pancost at Bristol

University. Extracts were concentrated in hexane to 25 μl, and 1 μL was injected on the GC-MS.

There were two sets of extracts from samples from Handley et al., (2008) and from samples from

Aze et al., (2014). The grams of sediment extracted was originally recorded for the Handley et al. samples, but records were not obtainable at the time of this study, and thus data is in mass quantities per sample (ng/μl) rather than in terms of grams of sediment or total organic carbon

(TOC).

PAHs were analyzed from the polar fraction using an Agilent 6890 GC with an Agilent

5973 quadrupole mass spectrometer (MS) and a fused silica capillary column (Agilent J&W DB-

177 5; 30 m, 250 μm, 0.25 μm). The column flow rate was 2.0 ml/min and the oven program started at

60°C for 1 min, ramped to 320°C at 6°C/min, and a final hold time of 15 min. The MS had an ionization energy of 70 eV with a scanning mass range of m/z 40-700 in Full Scan mode. PAHs were identified in Full Scan mode based on authentic standards, NIST 98 spectral library, fragmentation patterns, and retention times and quantified in Selected Ion Monitoring (SIM) mode. The following ions were monitored: (m/z) 102, 128; 151, 152, 153, 154; 165, 166, 183,

198; 176, 178; 101, 202, 219, 234; 114, 126, 228, 253, 254; 125, 126, 252; 276, 278; 150, 300.

Compound peak areas were converted to mass quantities using response curves for 19 PAH external standards analyzed in concentrations ranging from 0.01 to 1 ng/μl.

Discussion

For future work I would normalize the PAHs to another aromatic compound that was in the aromatic fraction, such as a plant biomarker, like was done for PAHs in Arctic samples (Denis et al., in prep). Data normalized to total organic carbon (TOC) should be interpreted cautiously, since it was not clear at the time what the uncertainty or lower limit of quantification was for the

TOC measurements (Handley et al., 2008; Aze et al., 2014). Ratios between PAHs themselves are worth investigating further for trends.

178 Figures

Figure F-1. Geographical map of Tanzania study site (yellow star) from Handley et al. (2008).

179 Data Tables

Table F-1. Tanzania PAH concentrations (pg/μl injected onto GC-MS)

Pyrene Retene

Fluoranthene Phenanthrene

GC data file Sample Name GC Retention time (minutes)  21.57 26.13 26.93 28.41 15090745.D TDP14-2-2_26-36 Polar LH10 20.71 64.02 85.76 9.70 15090744.D TDP14-3-1_20-30 Polar LH10 31.31 69.53 85.36 8.48 15090746.D TDP14-3-1_70-80 Polar LH10 27.40 63.81 86.36 11.31 15090749.D TDP14-3-2_10-20 Polar LH10 20.46 89.61 123.88 2.22 15090741.D TDP14-4-3_10-20 hPolar LH10 1.21 19.01 22.30 3.69 15090750.D TDP14-4-3_40-45 hPolar LH10 2.63 14.10 25.47 61.35 15090909.D TDP14-5-2_40-50 hPolar LH10 4.10 16.00 21.93 68.22 15090740.D TDP14-5-2_77-83 hPolar LH10 312.71 338.04 795.07 1147.75 15090912.D TDP14-6-1_85-95 hPolar LH10 0.46 11.74 41.25 250.17 15091014.D TDP14-6-3_20-22 hPolar LH10 741.92 682.39 2188.86 4163.47 15090721.D TDP14-7-1_10-20 Polar MPSB 639.43 580.38 1700.06 325.99 15090743.D TDP14-7-1_30-40 Polar MPSB 241.82 579.35 1105.50 223.48 15090719.D TDP14-7-1_50-60 Polar MPSB 2431.51 579.36 1226.61 244.26 15090739.D TDP14-7-1_70-80 Polar MPSB 7762.56 1303.18 3695.71 436.75 15090910.D TDP14-7-1_72-80 Polar LH10 39.97 60.66 121.67 6.26 15090728.D TDP14-7-2_10-20 Polar MPSB 3764.56 583.70 1613.49 257.41 15090729.D TDP14-7-2_30-40 Polar MPSB 873.67 809.96 805.24 342.34 15090727.D TDP14-7-2_60-70 Polar MPSB 2096.48 1030.01 2356.68 270.78 15090738.D TDP14-7-2_80-85 Polar MPSB 1719.53 742.35 1601.16 213.19 15090723.D TDP14-8-1_10-20 Polar MPSB 375.59 374.03 401.78 49.21 15090722.D TDP14-8-1_30-40 Polar MPSB 366.91 320.72 391.57 121.72 15090720.D TDP14-8-1_50-60 Polar MPSB 6950.05 937.15 1792.14 250.54 15090726.D TDP14-8-1_70-85 Polar MPSB 1924.80 605.07 1156.76 287.53 15091021.D TDP14-10-2_39-45 Polar LH10 49.57 364.48 437.06 42.00 15091021.D TDP14-10-2_39-45 Polar LH10 48.21 372.41 416.44 43.98 15090742.D TDP14-12-1_33-38 Polar LH10 6.70 91.81 127.31 10.59 15090911.D TDP14-13-3_23-25 Polar LH10 16.53 110.27 376.91 15.35 *Shaded area symbolizes PETM interval

180

Table F-2. Additional Tanzania PAH concentrations (pg/μl injected onto GC-MS)

rylene

cd]pyrene +

Perylene

Chrysene

Coronene

Benzo[e]pyrene Benzo[a]pyrene

Benz[a]anthracene

Benzo[g,h,i]pe

Dibenz[a,h]anthracene Benzofluoranthens (b+k) GC file Indeno[1,2,3 (minutes) 31.63 31.77 35.61 36.36 36.57 36.84 40.05 40.67 44.52 15090745.D 11.59 51.25 13.98 13.50 24.92 10.80 63.46 55.91 49.69 15090744.D 5.21 46.08 40.61 39.16 26.80 12.64 14.95 30.47 221.26 15090746.D 2.83 17.79 27.03 6.53 11.26 6.61 7.09 11.21 23.47 15090749.D 6.16 51.66 35.05 27.63 11.15 97.91 17.79 36.16 82.38 15090741.D 1.12 11.43 27.92 4.55 3.43 3.10 3.50 10.25 13.75 15090750.D 1.69 11.59 24.79 7.27 1.34 25.05 15.57 16.81 21.21 15090909.D 1.66 21.30 n/a n/a n/a n/a 15.39 6.46 3.28 15090740.D 7.05 36.32 n/a n/a n/a n/a 6.15 9.65 13.19 15090912.D 1.08 9.75 n/a n/a n/a n/a 2.83 2.31 3.50 15091014.D 17.44 92.86 91.98 217.76 20.78 885.42 97.19 300.66 152.31 15090721.D 109.43 238.83 402.57 841.06 36.09 1307.00 154.56 467.52 382.97 15090743.D 38.88 185.88 411.12 687.67 46.61 1975.09 99.29 429.56 349.23 15090719.D 42.62 332.32 1005.82 1024.10 118.14 2732.01 221.38 621.37 481.54 15090739.D 116.59 514.93 670.23 332.25 135.73 2741.75 216.69 728.51 652.01 15090910.D 6.70 32.22 92.57 n/a 36.62 533.13 26.64 142.38 89.97 15090728.D 78.84 326.54 2653.86 832.97 116.69 2724.27 278.37 642.16 690.56 15090729.D 15.31 223.46 401.71 880.03 94.80 324.21 102.01 274.60 286.10 15090727.D 39.16 376.16 607.85 1162.14 37.55 1472.41 159.80 430.88 427.96 15090738.D 41.52 219.86 408.33 1040.56 35.07 1488.31 127.46 376.43 392.59 15090723.D 23.09 107.38 341.76 396.23 39.40 611.97 80.36 281.22 246.28 15090722.D 152.40 152.58 401.57 718.79 51.77 892.77 129.54 508.73 488.10 15090720.D 63.98 355.89 1018.38 857.08 46.55 1344.12 251.08 845.53 736.25 15090726.D 54.53 268.13 657.24 721.24 44.30 1410.23 150.17 608.05 516.23 15091021.D 24.49 121.29 232.41 144.12 41.03 203.46 103.36 319.27 346.64 15091021.D 19.28 115.33 217.70 181.57 36.37 183.58 86.80 324.28 334.37 15090742.D 4.32 24.40 80.74 64.84 17.57 95.32 23.32 123.81 207.03

181

rylene

cd]pyrene +

Perylene

Chrysene

Coronene

Benzo[e]pyrene Benzo[a]pyrene

Benz[a]anthracene

Benzo[g,h,i]pe

Dibenz[a,h]anthracene Benzofluoranthens (b+k) GC file Indeno[1,2,3 (minutes) 31.63 31.77 35.61 36.36 36.57 36.84 40.05 40.67 44.52 15090911.D 2.23 48.93 101.95 9.63 n/a n/a 35.48 92.61 121.48

Table F-3. Tanzania sample depth, extracted weight, PAH ratios.

Depth (mbsf)

Dry weight(g) 5+6 ring PAH/Total PAH

Sample Name Coronene/Pyrene+Coronene TDP14-2-2_26-36 Polar LH10 5 0.37 0.38 TDP14-3-1_20-30 Polar LH10 7 0.72 0.23 TDP14-3-1_70-80 Polar LH10 9 0.21 0.13 TDP14-3-2_10-20 Polar LH10 11 0.40 0.33 TDP14-4-3_10-20 hPolar LH10 13.43 0.38 0.19 TDP14-4-3_40-45 hPolar LH10 13.65 0.45 0.29 TDP14-5-2_40-50 hPolar LH10 15.45 0.13 0.14 TDP14-5-2_77-83 hPolar LH10 15.8 0.02 0.01 TDP14-6-1_85-95 hPolar LH10 17.9 0.08 0.02 TDP14-6-3_20-22 hPolar LH10 19.21 0.07 0.14 TDP14-7-1_10-20 Polar MPSB 20.15 39.746 0.18 0.33 TDP14-7-1_30-40 Polar MPSB 20.35 42.004 0.24 0.48 TDP14-7-1_50-60 Polar MPSB 20.55 44.972 0.28 0.39 TDP14-7-1_70-80 Polar MPSB 20.75 45.943 0.15 0.21 TDP14-7-1_72-80 Polar LH10 20.76 0.43 0.67 TDP14-7-2_10-20 Polar MPSB 21.15 44.085 0.30 0.29 TDP14-7-2_30-40 Polar MPSB 21.35 47.411 0.26 0.19 TDP14-7-2_60-70 Polar MPSB 21.65 75.027 0.15 0.24 TDP14-7-2_80-85 Polar MPSB 21.825 45.124 0.20 0.29

182

Depth (mbsf)

Dry weight(g) 5+6 ring PAH/Total PAH

Sample Name Coronene/Pyrene+Coronene TDP14-8-1_10-20 Polar MPSB 23.15 42.630 0.38 0.38 TDP14-8-1_30-40 Polar MPSB 23.35 49.904 0.55 0.45 TDP14-8-1_50-60 Polar MPSB 23.55 51.683 0.29 0.18 TDP14-8-1_70-85 Polar MPSB 23.775 49.053 0.31 0.31 TDP14-10-2_39-45 Polar LH10 27.62 0.44 0.34 TDP14-10-2_39-45 Polar LH10 27.62 0.45 0.34 TDP14-12-1_33-38 Polar LH10 31.16 0.62 0.43 TDP14-13-3_23-25 Polar LH10 34.44 0.24 0.16

Table F-4. PAH concentrations ng/g dry sediment.

__ __

cd]pyrene + + cd]pyrene

Pyrene Retene

Perylene

Chrysene

Coronene

Fluoranthene

Depth Depth (mbsf)

Polar MPSB) Polar

Phenanthrene

Benzo[e]pyrene Benzo[a]pyrene

Benz[a]anthracene

Benzo[g,h,i]perylene

Dibenz[a,h]anthracene

Benzofluoranthens (b+k) Benzofluoranthens

Indeno[1,2,3 Sample Name (TDP14 Sample Name

7-1_10-20 20.15 0.80 0.73 2.14 0.41 0.14 0.30 0.51 1.06 0.05 1.64 0.19 0.59 0.48 7-1_30-40 20.35 0.29 0.69 1.32 0.27 0.05 0.22 0.49 0.82 0.06 2.35 0.12 0.51 0.42 7-1_50-60 20.55 2.70 0.64 1.36 0.27 0.05 0.37 1.12 1.14 0.13 3.04 0.25 0.69 0.54 7-1_70-80 20.75 8.45 1.42 4.02 0.48 0.13 0.56 0.73 0.36 0.15 2.98 0.24 0.79 0.71 7-2_10-20 21.15 4.27 0.66 1.83 0.29 0.09 0.37 3.01 0.94 0.13 3.09 0.32 0.73 0.78 7-2_30-40 21.35 0.92 0.85 0.85 0.36 0.02 0.24 0.42 0.93 0.10 0.34 0.11 0.29 0.30 7-2_60-70 21.65 1.40 0.69 1.57 0.18 0.03 0.25 0.41 0.77 0.03 0.98 0.11 0.29 0.29 7-2_80-85 21.83 1.91 0.82 1.77 0.24 0.05 0.24 0.45 1.15 0.04 1.65 0.14 0.42 0.44 8-1_10-20 23.15 0.44 0.44 0.47 0.06 0.03 0.13 0.40 0.46 0.05 0.72 0.09 0.33 0.29 8-1_30-40 23.35 0.37 0.32 0.39 0.12 0.15 0.15 0.40 0.72 0.05 0.89 0.13 0.51 0.49 8-1_50-60 23.55 6.72 0.91 1.73 0.24 0.06 0.34 0.99 0.83 0.05 1.30 0.24 0.82 0.71 8-1_70-85 23.78 1.96 0.62 1.18 0.29 0.06 0.27 0.67 0.74 0.05 1.44 0.15 0.62 0.53

183

References

Aze, T., Pearson, P.N., Dickson, A.J., Badger, M.P.S., Bown, P.R., Pancost, R.D., Gibbs, S.J., Huber, B.T., Leng, M.J., Coe, A.L., Cohen, A.S., Foster, G.L., 2014, Extreme warming of tropical waters during the Paleocene-Eocene Thermal Maximum, Geology, v. 42, no. 9, p. 739-742. Denis, E.H., Pedentchouk, N., Schouten, S., Pagani, M., and Freeman, K.H., in prep, Fire and ecosystem change in the Arctic across the Paleocene-Eocene Thermal Maximum: Earth and Planetary Science Letters. Handley, L., Pearson, P.N., McMillan, I.K., Pancost, R.D., 2008. Large terrestrial and marine carbon and hydrogen isotope excursions in a new Paleocene/Eocene boundary section from Tanzania. Earth and Planetary Science Letters 275, 17-25. Handley, L., O’Halloran, A., Pearson, P.N., Hawkins, E., Nicholas, C.J., Schouten, S., McMillan, I.K., Pancost, R.D., 2012. Changes in the hydrological cycle in tropical East Africa during the Paleocene-Eocene Thermal Maximum. Palaeogeography, Palaeoclimatology, Palaeocology. 329-330, 10-21.

184

Appendix G Model of how increased soil respiration affects carbon reservoirs

Described below is a simple model thought experiment to test the feasibility of how increasing soil respiration rates due to warming could influence the soil and atmosphere carbon reservoirs. In this case, the atmospheric carbon is at 2,000 Pg (estimate of PETM values) and the system is perturbed 0.5 °C initially.

Model calculations

Arrhenius equations (e.g., Davidson and Janssens, 2006):

Climate sensitivity to temperature (Scheffer et al., 2006):

Conditions and variables

PETM Catm = 2,000 Pg

Csoil stock initial = 1,500 Pg

Csoil input (constant) = 60 Pg C/yr

185

Csoil output (respiration) = k*Csoil

To = 16 °C (Wing et al., 2005)

T1 = 16.5 °C

k is the rate constant

Ea is the activation energy for carbon decay = 50 kJ/mol

R is a constant = 8.314 J/mol*K

Assumptions

There is a constant input of carbon into soils. Carbon output from soils is only changing based on changes in the rate constant (k) from the Arrhenius Equation. k changes based on changes in temperature due to climate sensitivity. The soil carbon reservoir is completely accessible to soil respiration. Soil carbon values, inputs, outputs are based on modern values from

Sundquist and Visser, (2005).

Discussion

Based on this model, due to increased temperature, increased soil respiration rates could have amplified warming during the PETM. By perturbing the system with a 0.5°C increase in temperature, 160 Pg of carbon was released from soils into the atmosphere over 400 years and resulted in ~1°C temperature rise. The time scale of 400 years is likely “too quick”, and feedbacks within in the carbon cycle and organic carbon stabilization mechanisms would slow this response on the order of thousands to tens of thousands of years. Therefore, increased soil respiration rates is another factor to consider in understanding changes in the terrestrial carbon cycle during warming events.

186 Figures

Figure G-1. Schematic of box model.

Soil carbon reservoir

1520

1480

1440

1400

1360

1320 Mass of Soilof(Pg) Carbon Mass 0 100 200 300 400 500 Time (yr)

Figure G-2. Plot of modeled carbon in the soil reservoir with time.

187

Temperature

291

290.6

290.2

289.8

Temperature (K) Temperature 289.4

289 0 100 200 300 400 500 Time (yr)

Figure G-3. Plot of modeled temperature with time.

References

Davidson, E.A., and Janssens, I.A., 2006, Temperature sensitivity of soil carbon decomposition and feedbacks to climate change: Nature, v. 440, no. March, p. 165–173, doi: 10.1038/nature04514. Scheffer, M., Brovkin, V., and Cox, P.M., 2006, Positive feedback between global warming and atmospheric CO2 concentration inferred from past climate change: Geophysical Research Letters, v. 33, no. 10, p. 2–5, doi: 10.1029/2005GL025044. Sundquist, E.T., and Visser, K., 2005, Geologic history of the carbon cycle, in Schlesinger, W.H., Holland, H.., and Turekian, K.K. eds., Biogeochemistry, Elsevier, New York, p. 425– 472. Wing, S.L., Harrington, G.J., Smith, F.A., Bloch, J.I., Boyer, D.M., and Freeman, K.H., 2005, Transient Floral Change and Rapid Global Warming at the Paleocene-Eocene Boundary: Science, v. 310, no. 5750, p. 993–996, doi: 10.1126/science.1116913.

VITA Elizabeth H. Denis

Education Ph.D. Geosciences and Biogeochemistry, Pennsylvania State University, 2016 Advisor: Katherine H. Freeman Dissertation: Production and preservation of organic and fire-derived carbon across the Paleocene- Eocene Thermal Maximum.

B.S. Honors Geology-Chemistry, Brown University, 2010

Fellowships and Awards 2nd Place Oral Presentation by Ph.D. Student (post-comps exam), PSU Geosciences Graduate Colloquium Pennsylvania Space Grant Graduate Student Fellowship (2016) ConocoPhillips Graduate Student Fellowship (2015, 2011) Paul D. Krynine Scholarship (2015, 2014, 2012, 2011) NSF Graduate Research Fellowship (2012-2015) Geological Society of America (GSA) Northeastern Section Student Travel Grant (2014) American Geosciences Institute (AGI) Harriet Evelyn Wallace Fellowship (2014) Geological Society of America (GSA) Student Research Grant (2014) 2nd Place Oral Presentation by Ph.D. Student, PSU Geosciences Graduate Colloquium (2013) Hiroshi and Koya Ohmoto Graduate Fellowship (2013) Charles E. Knopf, Sr., Memorial Scholarship (2012) Biogeochemistry Dual-Degree Program Award (2011) Shell Research Facilitation Award (2011) Brown Geological Sciences Undergraduate Research, Academics and Service Award (2010) Brown Women’s Ice Hockey Academic Excellence Award for the highest GPA (2010) Brown University Research Grant (2009)

Professional and Teaching Experience Geoscience Summer Intern, ConocoPhillips Geological Technology, Houston, TX, (2014) Geoscience Summer Intern, ExxonMobil Upstream Research Company, Houston, TX, (2012) Teaching Assistant, Penn State: The Sea Around Us; Chemical Processes in Geology (2011; 2010) Teaching Assistant, Brown University: The Earth System; Physical Geology (2010, 2009; 2009)

Service Player and mentor, Penn State Women’s Club Ice Hockey Team (2011-2015) Co-founder & Officer, Association of Women Geoscientists (AWG) Penn State Student Chapter (2012-15)

Publications Magill C.R., Denis E.H., Freeman K.H. (2015) Rapid sequential separation of sedimentary lipid biomarkers via selective accelerated solvent extraction. Organic Geochemistry, 88: 29-34. Good S.P., Mallia D.V., Denis E.H., Freeman K.H., Feng X., Li S., Zegre N., Lin J.C., Bowen G.J. (2014) High frequency trends in the isotopic composition of superstorm Sandy. In: Learning from the Impacts of Superstorm Sandy (J.B. Bennington. & E.C. Farmer, eds), Elsevier. Chap. 4, 41-55. Wilkins M.J., Daly R.A., Mouser P.J., Trexler R., Sharma S., Cole D.R., Wrighton K.C., Biddle J.F., Denis E.H., Fredrickson J.K., Kieft T.L., Onstott T.C., Peterson L., Pfiffner S.M., Phelps T.J., Schrenk, M.O. (2014) Trends and future challenges in sampling the deep terrestrial biosphere. Frontiers in Microbiology, 5: 1-8. Denis E.H., Toney J.L., Tarozo R., Anderson R.S., Roach L.D., Huang Y. (2012) Polycyclic aromatic hydrocarbons (PAHs) in lake sediments record historic fire events: Validation using HPLC- fluorescence detection. Organic Geochemistry, 45: 7-17.